The Advantages of Thermodynamic Modeling in Hydrogen Production

Complimentary Contributor Copy

ADVANCES IN MATERIALS SCIENCE RESEARCH
ADVANCES IN MATERIALS
SCIENCE RESEARCH
VOLUME 35
No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or
by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no
expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No
liability is assumed for incidental or consequential damages in connection with or arising out of information
contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in
rendering legal, medical or any other professional services.
SCIENCE RESEARCH
Additional books and e-books in this series can be found on Nova’s

website under the e-book or Series tab.

ADVANCES IN MATERIALS SCIENCE RESEARCH
SCIENCE RESEARCH
VOLUME 35
MARYANN C. WYTHERS
EDITOR

Copyright © 2019 by Nova Science Publishers, Inc.
All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted
in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying,
recording or otherwise without the written permission of the Publisher.
We have partnered with Copyright Clearance Center to make it easy for you to obtain permissions to
reuse content from this publication. Simply navigate to this publication’s page on Nova’s website and
locate the “Get Permission” button below the title description. This button is linked directly to the title’s
permission page on copyright.com. Alternatively, you can visit copyright.com and search by title, ISBN,
or ISSN.
For further questions about using the service on copyright.com, please contact:
Copyright Clearance Center
Phone: +1-(978) 750-8400 Fax: +1-(978) 750-4470 E-mail: info@copyright.com.
NOTICE TO THE READER

The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or
implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is
assumed for incidental or consequential damages in connection with or arising out of information
contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary
damages resulting, in whole or in part, from the readers’ use of, or reliance upon, this material. Any parts
of this book based on government reports are so indicated and copyright is claimed for those parts to the
extent applicable to compilations of such works.
Independent verification should be sought for any data, advice or recommendations contained in this
book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons
or property arising from any methods, products, instructions, ideas or otherwise contained in this
publication.
This publication is designed to provide accurate and authoritative information with regard to the subject
matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in
rendering legal or any other professional services. If legal or any other expert assistance is required, the
services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS
JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A
COMMITTEE OF PUBLISHERS.
Additional color graphics may be available in the e-book version of this book.
Library of Congress Cataloging-in-Publication Data
ISBN: H%RRN

ISSN: 2159-1997
Published by Nova Science Publishers, Inc. † New York

CONTENTS
Preface vii
Chapter 1 Self-Healing of Structural Composites
by Incorporating Diels-Alder Base
Polymers as Healing Agent 1
Vassilis Kostopoulos and Athanasios Kotrotsos
Chapter 2 Controlling Structural and Electrochemical
Properties of CNF through Calcination
of PAN Nanofibers 79
Reyna Ojeda-López, Isaac J. Pérez-Hermosillo,
Juan M. Esparza-Schulz,
Guadalupe Ramos-Sánchez
and Armando Domínguez-Ortiz
Chapter 3 Purification of Gas and Liquid Streams
Using Composite Sorbents Embedded
in a Polyacrylonitrile Matrix 119
Brian J. Riley and Troy G. Garn

vi Contents
Chapter 4 Dewetting Behavior of Liquid Metal Film

on Nanopillared Graphene and Its Effect
on the Coalescence Dynamics 149
Hui Li, Tao Li, Yanyan Jiang and Xiongying Li
Chapter 5 The Advantages of Thermodynamic Modeling in
Hydrogen Production, a Case Study: Ethanol
Steam Reforming Combined with CO2 Absorption 177
Virginia Collins-Martinez,
Miguel Escobedo Bretado,
Miguel Meléndez Zaragoza,
Jesús Salinas Gutiérrez
and Alejandro López Ortiz
Chapter 6 New Piperazine-Based Polymerizable
Monomers for the Development
of Stimuli-Responsive Polymers 195
G. Roshan Deen and Chin Hao Mah
Contents of Earlier Volumes 221
Index 227

PREFACE
Advances in Materials Science Research. Volume 35 opens with an

examination of the effect of bismaleimides based on the Diels-Alder reaction
mechanism on high performance aerospace carbon fibre reinforced plastics.
More precisely, the interlaminar fracture toughness of unidirectional
bismaleimide modified composites and their healing capabilities were
measured under mode I and mode II fracture loading conditions.
Following this, polyacrylonitrile nanofibers are first obtained from
polyacrylonitrile in dimethylformamide by electrospinning. Then, the
carbon nanofibers are obtained by calcination of polyacrylonitrile
nanofibers. Additionally, the capacity of these materials to promote oxygen
reduction reaction is evaluated.
As new sorbents are developed for purifying gases or liquids via
physisorption and/or chemisorption processes, the objective is typically to
increase the specific surface area to maximize the available binding sites for
capturing various species of interest. One drawback to making a sorbent
more porous is that it often leads to a product that is fragile or friable and
can be damaged, thus imposing process limitations. Thus, the authors
propose an alternative approach for improving the mechanical integrity of
the sorbent by binding it within a porous matrix that passively holds the
sorbent in place.

viii Maryann C. Wythers
Next, evidence that the nanopillared graphene surface can induce strong
dewettability is been reported by molecular dynamics simulations. The
retracting and detaching behaviors of the liquid film can be effectively
controlled by tuning the geometric parameters of the liquid film or nanopillar
as well as the temperature.
This compilation also includes a study which aims to present the
advantages of using a thermodynamic analysis for the production of high
purity hydrogen by the ethanol steam reforming combined with CO2
absorption (ESR-CO2-ABS). For this purpose, a thermodynamic analysis is
used to find optimal reaction conditions for the ESR-CO2-ABS system.
Thermodynamic analysis of the ESR-CO2-ABS system was performed
using CaO, CaO•MgO, Na2ZrO3, Li2ZrO3 and Li4SiO4 as CO2 absorbents
to determine optimal operating conditions to generate high production and
concentration of hydrogen in the product gas.
In conclusion, the authors discuss how recently, research emphasis has
been directed towards integrating multiple functions into polymer materials.
Among these new advancements in materials science are functional
polymers with structural designs for versatile applications. Piperazine is an
interesting heterocyclic diamine which possess excellent medicinal and
chelating property. Polymers and surfactants based on piperazine and
surfactants have shown good hydrophilic, chelating anti-bacterial and anti-
helminthic properties
Chapter 1 - Bismaleimides (BMIs) exhibit healing functionalities on
polymer level. In the present chapter, the effect of BMIs that are based on
Diels-Alder (DA) reaction mechanism into high performance aerospace
carbon fibre reinforced plastics (CFRPs) is assessed. More precisely, the
interlaminar fracture toughness of unidirectional (UD) BMI modified
composites (containing BMI interleaves or BMI pre-impregnated fibre
layers (prepregs)) and their healing capability were measured under mode I
and mode II fracture loading conditions. BMI prepregs were fabricated in
order to facilitate the introduction of the self-healing agent (SHA) into the
composite laminated structure. During testing, these modified samples
exhibited extended fibre bridging between the interlaminar crack flanks,
which considerably enhanced their interlaminar fracture toughness. In

Preface ix
addition, BMI interleaves and prepregs were further utilized to modify

quasi-isotropic CFRPs in a symmetric fashion in order the damage tolerance
and the healing capability of these composites to be investigated. To that
direction, the BMI modified laminates were tested under quasi-static tensile
loading, tension-tension fatigue loading, low velocity impact (LVI)
conditions while compression after impact (CAI) tests were conducted prior
and after the activation of the healing. Finally, examination of the
morphology of the fracture surface led to qualitative conclusions regarding
the involved failure and healing mechanisms.
Chapter 2 - In recent decades the use of carbon nanofibers (CNF) has
increased notably as a result of the development of techniques that allow the
manipulation of their chemical and textural properties. This has allowed the
expansion of the fields of application, from components for batteries and
fuel cells to vapor adsorption and catalysts. In this context, CNF have been
developed to satisfy diverse industrial requirements, taking advantage of
some of their important properties; such as high surface area, high strength
to weight ratio, excellent chemical resistance, superior electrical
conductivity and good thermal conductivity.
In this work, polyacrylonitrile nanofibers (PANNF) are first obtained
from polyacrylonitrile (PAN) in dimethylformamide (DMF) by
electrospinning. Then, the CNF are obtained by calcination of PANNF. This
process is comprised of two steps: i) stabilization at temperatures between
523 to 573 K in an oxidizing atmosphere (air), and ii) carbonization at
temperatures above 873 K in an inert atmosphere (nitrogen). These steps
control both, microporosity and amount (and type) of nitrogen species. The
study is then divided in two stages, the first consists of varying the
temperature in step one while maintaining the temperature in step two
constant, and vice versa for stage two. The consequences of four
stabilization temperatures (523, 543, 553 and 573 K) on these properties are
evaluated while maintaining the carbonization temperature constant at 1173
K; the most adequate stabilization temperature to keep the fibrillary structure
is 553 K. To study the second stage, different portions of material stabilized
at 553 K are carbonized at 873, 973, 1073, 1173 and 1273 K. Finally, the

x Maryann C. Wythers
capacity of these materials to promote oxygen reduction reaction (ORR) is

evaluated.
Chapter 3 - As new sorbents are developed for purifying gases or liquids
via physisorption and/or chemisorption processes, the objective is typically
to increase the specific surface area (SSA) to maximize the available binding
sites for capturing various species of interest. One downfall to making a
sorbent more porous is that it often leads to a product that is fragile or friable
and can be damaged (resulting in unwanted powders), thus imposing process
limitations. Techniques can be used to enhance the mechanical integrity of
a sorbent, including heat treatments or pelletization with extrusion or
granulation processes, but these can result in collapsed, unavailable pore
structures within the sorbent, which decreases the diffusivity and eventually
reduces the capacities of target species. An alternative approach for
improving the mechanical integrity of the sorbent is to bind it (or encapsulate
it) within a porous matrix that passively holds the sorbent in place.
Typically, this approach leads to a decrease in sorption efficiency on the
basis of starting sorbent mass, because the active sorbent is diluted by the
binding matrix material and the active surfaces of the sorbent are somewhat
obscured by the binding matrix; however, it comes with the advantage that
the active sorbent is held in place and better protected from disintegrating
during operation. One approach to securing the active sorbent is to embed it
in a macroporous, passive polyacrylonitrile (PAN) matrix to create a
composite sorbent. When optimized, the PAN matrix is highly porous and
allows for gaseous or aqueous media transport through the product,
providing adequate binding site access between the active sorbent and the
species of interest. Including a polymer in the composite sorbent limits the
maximum operating temperature, above which the composite sorbent incurs
loss of structural integrity (e.g., decomposes) and/or reduction in SSA.
Composite sorbents have been demonstrated with sulfide aerogels (SnS-
PAN) for capturing I2(g), hydrogen mordenite (HZ-PAN) and silver-
exchanged mordenite (AgZ-PAN) for capturing Kr and Xe, nanoparticle
metal oxides (Nano-Composite Arsenic Sorbent, or N-CAS) for removing
arsenic from water, ammonium molybdophosphate (AMP-PAN) for
removing cesium-137 (137Cs) from aqueous nuclear waste streams, and The

Preface xi
Phosphate Sponge (TPS) composite, which is being used to remove

inorganic/ortho-phosphates in water treatment processes to mitigate
introduction to the environment, which impedes toxic algae bloom
formations. This chapter provides details on production of these materials
along with options to optimize the synthesis process. It also includes
experimental techniques to characterize products and assess their
performance.
Chapter 4 - Dewetting has attracted considerable attention in recent
years because of its significant importance in many applications, including
anti-ice, anti-fog, self-clean and waterproof. In this chapter, evidence that
the nanopillared graphene surface can induce strong dewettability has been
reported by molecular dynamics (MD) simulations. The retracting and
detaching behaviors of the liquid film can be effectively controlled by tuning
the geometric parameters of the liquid film or nanopillar, as well as the
temperature. Additionally, the dewetting-induced coalescence dynamics
between liquid metals is also discussed. This work provides a
comprehensive view of the molecular-level interactions involved in
dewetting and promotes the authors’ understanding on how to control liquid
dewetting behavior using nanopillared structure, which offers potential in a
wide range of applications, especially in the fabrication of self-cleaning
materials.
Chapter 5 - Research related to thermodynamic studies of steam
reforming reactions has been an invaluable tool for establishing an
experimental basis for further evaluation of reaction systems. This study
aims to present the advantages of using a thermodynamic analysis for the
production of high purity hydrogen by the ethanol steam reforming
combined with CO2 absorption (ESR-CO2-ABS). For this purpose, a case
study is presented where the thermodynamic analysis is used to find optimal
reaction conditions for the ESR-CO2-ABS system. Thermodynamic analysis
of the ESR-CO2-ABS system was performed using CaO, CaO•MgO,
Na2ZrO3, Li2ZrO3 and Li4SiO4 as CO2 absorbents to determine optimal
operating conditions to generate high production and concentration of
hydrogen in the product gas. Conditions studied were at 300-850°C and 1
atm. Steam to ethanol feed molar ratio (S/EtOH) was varied from 3:1

xii Maryann C. Wythers
(stoichiometric) to 6:1. Results indicate that S/EtOH ≤ stoichiometric

resulted in prevention of carbon formation. The highest production and
concentration of H2 was obtained with CaO and MgO•CaO and Na2ZrO3 at
600°C and S/EtOH = 6:1. Sodium zirconate (Na2ZrO3) as CO2 absorbent
under the ESR-CO2-ABS process scheme can be considered a promising
alternative due to its greater kinetics and stability.
Chapter 6 - In recent years, new research emphasis is directed towards
integrating multiple functions into polymer materials. Among this new
advance in materials science are the functional polymers with structural
designs for versatile applications. Piperazine is an interesting heterocyclic
diamine, which possesses excellent medicinal and chelating property.
Polymers based on piperazine (polyamido-amines), surfactants based on
piperazine have shown good hydrophilic, chelating anti-bacterial and anti-
helminthic properties. However, piperazine-based homopolymers and
copolymers have not been investigated for stimuli-responsive behavior in
aqueous solutions and in bulk. Stimuli-responsive polymers have the ability
to respond to changes in external stimuli such as pH, temperature, salts,
surfactants, light, magnetic field, pressure, and biomolecules. These
materials are synthesized in various architectures such as linear polymers,
block polymers, hydrogels, nanogels, inter-penetrating networks, plum-
pudding type gels, and dendrimers. The behavior of this class of polymeric
materials resembles that of living systems in response to changing external
environment. These materials find wide range of applications in biomedical
science, pharmaceuticals, cosmetics, chemical sensors etc. In continuation
of the authors’ research interest in piperazine-based polymers of poly
(amido-amine) architectures, the authors describe the synthesis of new
piperazine-based polymerizable monomers in the development of multiple
stimuli-responsive polymer systems.

In: Advances in Materials Science Research ISBN: 978-1-53614-720-9
Editor: Maryann C. Wythers © 2019 Nova Science Publishers, Inc.
Chapter 1
SELF-HEALING OF STRUCTURAL
COMPOSITES BY INCORPORATING
DIELS-ALDER BASE POLYMERS
AS HEALING AGENT
Vassilis Kostopoulos* and Athanasios Kotrotsos, PhD

Department of Mechanical Engineering and Aeronautics,
Patras University, Patras, Greece
ABSTRACT
Bismaleimides (BMIs) exhibit healing functionalities on polymer

level. In the present chapter, the effect of BMIs that are based on Diels-
Alder (DA) reaction mechanism into high performance aerospace carbon
fibre reinforced plastics (CFRPs) is assessed. More precisely, the
interlaminar fracture toughness of unidirectional (UD) BMI modified
composites (containing BMI interleaves or BMI pre-impregnated fibre
layers (prepregs)) and their healing capability were measured under mode
I and mode II fracture loading conditions. BMI prepregs were fabricated in
order to facilitate the introduction of the self-healing agent (SHA) into the
*
Corresponding Author Email: kostopoulos@mech.upatras.gr.

2 Vassilis Kostopoulos and Athanasios Kotrotsos
composite laminated structure. During testing, these modified samples

exhibited extended fibre bridging between the interlaminar crack flanks,
which considerably enhanced their interlaminar fracture toughness. In
addition, BMI interleaves and prepregs were further utilized to modify
quasi-isotropic CFRPs in a symmetric fashion in order the damage
tolerance and the healing capability of these composites to be investigated.
To that direction, the BMI modified laminates were tested under quasi-
static tensile loading, tension-tension fatigue loading, low velocity impact
(LVI) conditions while compression after impact (CAI) tests were
conducted prior and after the activation of the healing. Finally, examination
of the morphology of the fracture surface led to qualitative conclusions
regarding the involved failure and healing mechanisms.
Keywords: composites, self-healing, diels-alder, bismaleimides, fracture

mechanics, fatigue, impact, damage tolerance
1. INTRODUCTION
1.1. Background
During the last decades, the use of composites in aeronautics, wind

energy, automotive, and other mechanical engineering applications has been
significantly increased. The need for new materials with enhanced
mechanical properties, low density and corrosion resistance combined with
the necessity for greener operation of structures, were the main driver for
intensive research and application of polymer matrix fibre reinforced
composites in many fields as structural materials. Figure 1, shows the
evolution in use of materials versus time. In this picture, the rapid increase
of the use of polymers and composite materials during the last decades is
evident [1].
In addition, the use of composites and especially carbon/epoxy
composites in aerospace applications has been considerably increased since
1960. Characteristic examples are Boeing 787 Dreamliner and Airbus 350-
XWB in which their structural mass is made by 50% and 55% of composites
respectively. As it is expected the extensive use of composites results to
more fuel efficient and environmental friendly aircrafts. However, a primary

Self-Healing of Structural Composites … 3
limitation of composites is their poor interlaminar strength and fracture

toughness that make them prone to delaminations [2, 3]. Figure 2 illustrates
possible damage modes into a polymer composite in detail [4]. Delamination
is the result of out of plane LVI of a composite structure. Joining of the
microcracks, developed into the polymer matrix under service loads is
another option for delaminations. It is obvious that delaminations can lead
to significant suppression of the load-bearing capacity of the composite
components and structures.
Figure 1. Schematic representation of the importance of four classes of materials

(ceramics, composites, polymers and metals) in mechanical and civil engineering as a
function of the time [1].
During the last decades, the focus of various researchers has been the
improvement of the fracture performance of structural composites and
especially of CFRPs. A variety of methodologies have been proposed in the
literature to prevent composites from delaminations. These include chemical
modification of epoxy matrix systems (either by modifying the resin or the
hardener), interleaving, hybridization, stitching, short-fibres, z-pinning,

optimization of staking sequence, reactive rubbers and edge cap

reinforcement [5-14].
Figure 2. Damage modes appearing into polymer composites. Indentation, impact,

corrosive environments, ballistic punctures, surface scratching, and fatigue can lead to
various damage modes in polymer composites, as shown here [4].
Figure 3. (a) Scarf cavity formation prior to patch repair [15], (b) depiction of the resin
injection process into a honeycomb panel skin's composite.
However, although the interlaminar fracture properties have been

enhanced, delamination remains the main damage mode in the case of FRPs.
Furthermore, conventional repair techniques [15-16] of composites are time
consuming, require extensive labor cost and the repair defects face certain
restrictions. These repair techniques involve patch repair (e.g., bonded

external patch, bonded flush patch) or non-patch repair (e.g., resin injection,
potting or filling and surface coating) [15-16]. In the case of repair of
primary structural components, the certification authorities required
additional mechanical joining. Figure 3-a illustrates a representative
example of patch repair method in which the damage material is carefully
removed by using a high-speed grinder [15]. After material removal, new
composite plies are utilized to fill the created cavity. On the other hand,
Figure 3-b shows non-patch repair which includes resin injection through an
access hole into the damage area. This method eliminates the removal of
undamaged plies and results in higher recovery of the strength [16].
Thus, utilization of composites in human safety critical components has
always to be accompanied with non-destructive testing techniques (NDT) as
damage monitoring tool. These techniques involve ultrasonics, surface
waves, acoustic emission, infrared thermography and X-ray radiography
[16]. It is also of note that conventional damage detection equipment is in
some cases not able to detect tiny defects deep inside the material. However,
these defects could rapidly propagate between two periodical inspections
and lead the composite part to significant deterioration. This challenging
situation acted as an inspiration for seeking of new repair methods; cheaper
and applicable at the early stages of damage evolution. With an aim to
address some principal weak points of conventional repair techniques, an
emerging approach called “self-healing materials” [17, 18] has been
proposed but not yet been applied to commercial composites. This smart
technology aims to in-situ and autonomously repair damage and thus to lead
to extension of the effective life-span of composite structures. Self-healing
of composites promise to mitigate the importance for detecting damage and
to reduce the frequency of scheduled inspections.
Self-healing mechanisms incorporated into existing materials is an
emerging alternative to conventional repair methods to remove cracks and
restore performance. It is also considered as an intense research field over
the last decades. A self-healing material [19], is a class of biomimetic
material that can autonomously repair damage and regain its initial
properties after damage. Self-healing technologies could extend the service
life and reliability of epoxy resins and their composites. Based on published

research work, a typical way to classify self-healing systems is as: (a)

autonomous (extrinsic) [20-21] (capsule-based systems [20] or vascular-
based systems [21]), where a container is prefilled with a liquid SHA. (b)
Non-autonomous [22-23], where self-healing is created either by using a
modifier with inherently reversible behaviour [22] or by incorporating
functional groups directly into a thermoset structure that can form reversible
bonds (intrinsic [23]). While for non-autonomous self-healing an external
stimulus (heating) is required, autonomous self-healing operates at service
temperature (not external stimulus is required).
A promising approach to incorporate an intrinsic or non-autonomous
self-healing functionality to the epoxy resin and its composites is based on
the DA and Retro-DA reactions (R-DA) [24-34]. The first report on DA
reaction mechanism by Kennedy and Castner dates back in 1979 [24]. Then,
thermally reversible crosslinked polymer through DA reaction was
introduced by Wudl and co-workers [25, 26]. The main mechanism behind
the healing is the thermally reversible DA reaction, where a
dicyclopentadiene unit in the polymer backbone breaks apart into two
cyclopentadiene terminal groups, which then reunite upon cooling. At room
temperature (RT) the material is rigid and presents resin-type behaviour
while upon heating the physical properties deteriorate due to
“intermonomer” disconnection. Then, upon cooling the properties revert
back. DA-based techniques possess several advantages over other self-
healing techniques; among these, they are simpler, they can offer multiple
temperature induced healing events and there is no need for catalysts or other
additional compounds to the parent material that may compromise the
structural integrity of the latter.
Various self-healing composites based on polymers whose molecular
structure consists of reversible DA crosslinks have been developed during
the last decades [27-30]. The healing capabilities of a similar SHA coating
via resistive heating in a CFRP substrate was evaluated under short three-
point bending tests in [31, 32]. A comprehensive study related to the
toughening and healing performance of a high performance CFRP
containing two carbon fibre/BMI prepreged plies on its mid-plane, under
mode I loading conditions was recently published [33]. The aforementioned

investigation is analytically presented in section 4 of the current chapter.

BMI was proven to be able to significantly increase the fracture toughness
of the host laminate and it also incorporated a healing functionality. In [34],
the self-healing process of the same type of SHA which was used as adhesive
in a metal-to-metal single lap-joint tension testing was evaluated non-
destructively. The study confirmed the success of the healing performance
of the developed healing system.
In the present chapter, interleaves or prepregs comprising self-healing
materials based on DA base BMI polymers were successfully incorporated
into high performance UD carbon/epoxy composites. The interlaminar
fracture toughness of these modified composites and their healing capability
were measured under mode I and mode II fracture loading conditions
(section 3, 4 and 5 respectively). During testing, the BMI modified samples
exhibited fibre bridging phenomena, which considerably enhanced their
interlaminar fracture toughness properties. Potential knock down effect due
to the incorporation of the BMI as SHA was quantified through three-point
bending tests (3PB). BMI carbon fibre prepregs were fabricated in order to
to simplify the incorporation process of SHA into the composite laminates.
BMI interleaves or prepregs were further utilized to modify quasi-isotropic
CFRPs. To that direction BMI modified CFRPs were tested under quasi-
static tensile and tension-tension fatigue loading conditions (section 6). In
addition, potential knock down effects and the induced healing capability of
these modified composites were investigated under compression before
impact (CBI), LVI and CAI tests (section 7). Finally, examination of the
morphology of fracture surface led to qualitative conclusions regarding the
involved failure and healing mechanisms.
2. EXPERIMENTAL
2.1. Materials and Methods
2.1.1. Raw Materials and Self-Healing Agent (SHA) Synthesis

The BMI modified composite laminates, which are used in the present
chapter, were fabricated by UD carbon fibre/epoxy resin prepreg CE-1007

150-38. The prepreg tape material supplied by SGL Group, Germany having
tensile strength of 2.4 GPa and axial Young modulus of 140 GPa. Two types
of SHA were utilized for the needs of the current chapter; the BMI
prepolymer (BMI pp) and the pure "in-house" BMI polymer. More
information about the synthesis of both types is provided below.
The DA prepolymer based epoxy resin in grains form was used in parts
of this work. For the synthesis of the SHA the readily available cost effective
commercial raw materials furfuryl amine, diglycidyl ether bisphenol A
(DGEBA) and BMI were utilized. Firstly, a prepolymer from DGEBA with
furfuryl amine (5:4 moles respectively) was synthesized. The obtained
prepolymer was a powder. Then, BMI (20 g BMI per 100 g prepolymer,
corresponding to 0.64 equivalents of maleimide groups per furfuryl group),
which is the reversible (DA) crosslinker, was added. Both powders were
mixed and homogenized through extrusion and injected into silicon moulds.
Details for the synthesis can be found in [35]. Finally, the bulk polymer was
converted into powder with average particle size of 100 μm by utilizing a
powder machine.
The "in-house" developed BMI polymer blend was used to impregnate
UD carbon fabrics and play the role of the SHA in parts of the current work.
The “in-house” BMI polymer was synthesized according to the following
process: a trifurane monomer (TF) was prepared by reacting furfuryl
glycidylether with furfuryl amine and was allowed to react with
commercially available BMI oligomers (BMI-1700), that were supplied by
Designer Molecules Inc., San Diego, USA. The TF and BMI were mixed in
stoichiometric analogy and the resulting polymer was used to impregnate
the carbon fabrics.
The melting temperature (Tm) of both types o f SHAs was measured to
be approximately 110°C, using the Perkin-Elmer DSC 8500 differential
scanning calorimeter. The DSC samples were heated from ambient
temperature to 150°C at a rate of 5°C/min. Above Tm, the interactions
between adjacent polymer chains of the material start to significantly
diminish.

2.1.2. Sieving Process and Preparation of the Pure BMI Prepreg Plies
The integration process of the BMI pp grains through sieving onto the
CFRP plies is illustrated in Figure 4-a. The BMI pp (Figure 4-a-i) was sieved
onto prepregs surface as Figure 4-a-ii suggests. The resulting prepreg ply
containing BMI pp grains on its surface is illustrated in Figure 4-a-iii. On
the other hand the preparation process of the BMI prepreg plies is illustrated
in Figure 4-b. The “in-house” pure BMI polymer (Figure 4-b) was poured in
the centre of the UD carbon fabric and the system was vacuum bagged.
Using a spatula the polymer was spread from the centre to the edges of the
fabric and then vacuum was applied to the system.
Figure 4. (a) Illustration of the sieving process during the manufacturing of the
modified CFRPs containing BMI prepolymer (BMI pp) grains. Depiction of (i) the
BMI pp grains, (ii) the sieving process at RT (iii) the final BMI pp covered prepreg’s
surface. (b) Preparation process of the pure BMI modified prepreg plies. (i) The bulk
“in-house” pure BMI polymer in high viscosity liquid form, (ii) impregnation process
of the carbon fabric with BMI polymer at 110°C, (iii) depiction of the final modified
prepreg ply.
Due to the large viscosity of the ΒΜΙ resin the impregnation was carried
out by medium temperature heating at 110°C, using a heating gun (Figure

4-b-ii). Above this temperature, the R-DA reaction takes place as the DA
reactions start to decrease significantly and thereby the interactions between
adjacent polymer chains of the material. After the impregnation of the
fabrics the fibre volume fraction (Vf) of the BMI prepregs was calculated to
be close to 55% (Figure 4-b-iii). The handling of the BMI pp grains and of
the CF/BMI prepreg plies at RT and the incorporation of them into
composites’ architecture did not create any concerns. BMI seems to combine
unique characteristics of traditional thermosets (good fracture toughness
properties) and typical behaviour of common thermoplastics (good
processability, softening and flow at elevated temperatures, etc). With the
viscosity of BMI remaining low during production of the CF/BMI prepregs,
the preparation of them was simple.
2.1.3. Mode I Interlaminar Fracture Toughness Testing

Quasi-static interlaminar fracture toughness tests described in section 3
and 4 respectively were performed at a 25 kN Instron Universal testing
machine (Instron, High Wycombe, UK) at RT conditions. The mode I
interlaminar fracture toughness was measured using the double cantilever
beam (DCB) method according to the AITM 1.0005 standard issued by
Airbus. DCB specimen dimensions and experimental set-up are both
illustrated in Figure 7 and Figure 18 respectively. The 25-mm-long PTFE
starter film was placed at the central plane, at one edge of the samples to
facilitate crack initiation is also shown. Two aluminium tabs were glued on
the outer surfaces of the pre-cracked end of the specimens, for the
application of the opening load. The specimens were loaded in tension at a
cross-head velocity of 10 mm/min until the crack was propagated from
starting point considered as zero to 75 mm. Five samples were tested each
time for the calculation of the mode I interlaminar fracture toughness energy
value (GIC). The GIC value of the CFRPs was calculated by using the
Equation (1),
A kJ
GIC = aw ∙ 103 (m2 ) (1)

where, A is the required energy to achieve the total propagated crack length,
a is the propagated crack length (final crack length minus initial crack length,
a=75 mm) and w is the specimen’s width.
2.1.4. Mode II Interlaminar Fracture Toughness Testing

The mode II interlaminar fracture toughness tests described in section 5
were measured by using the three-point end-notched flexure (ENF) method
according to the AITM 1.0006 standard issued by Airbus. Firstly, the DCB
specimens were loaded under mode I in alignment with the AITM 1.0005 to
create a natural sharp pre-crack. Then, the delaminated specimens were cut
into ENF specimens as shown in Figure 29, according to AITM 1.0006. ENF
specimen dimensions and experimental set-up are both illustrated in Figure
29. The pre-cracked ENF specimens were subjected to flexural loading that
generated a shear crack driving force at the crack tip region. The cross-head
velocity was fixed at 1 mm/min. Five samples were tested for each fracture
toughness assessment. The mode II interlaminar fracture toughness energy
value (GIIC) was determined by using the Equation (2),
9Pa2 d kJ
GIIC = 1 ( ) (2)
2w( L3 +3𝛼3 ) m2
4
where, d is the cross-head displacement at crack delamination onset, P is the

critical load to start the crack, α is the initial crack length (from the support
point to the end of the crack, α=35 mm), w is the width of the specimen and
L is the span length (L=100 mm). For the accurate calculation of the GIIC the
simple methodology described was applied.
2.1.5. Open-Hole Tensile Tensing

Uniaxial tension tests of a balanced and symmetric laminate are
described in section 6 and are performed in accordance with test method D
3039/D 3039M, despite the fact that the sample contains a centrally located
hole. The ultimate tensile strength (UTS), is calculated based on the gross
cross-sectional area disregarding the presence of the hole. While the hole
causes a stress concentration and reduced net section, it is common

aerospace practice to develop notched design allow-able strengths based on

gross section stress to account for various stress concentrations (fastener
holes, free edges etc.) The cross-head velocity was fixed at 2 mm/min. Five
samples were tested for each material type, at an Instron (250 kN) hydraulic
machine. The ultimate open-hole tensile strength was determined by using
the Equation (3),
P
σ = A (MPa) (3)
where, σ is the ultimate open-hole (notched) tensile strength, P is the

maximum force carried by test specimen prior to failure and A is the gross-
sectional area (disregarding hole).
2.1.6. Open-Hole Fatigue Testing

Tension-tension fatigue tests were performed according to D 3479/D
3479M – 96 (section 6) by using an Instron (250 kN) hydraulic machine.
These tests were conducted under load amplitude control, having a loading
ratio of R=0.1 with identical testing conditions to quasi static tests described
above (paragraph 2.1.5). The recommended loading frequency (f) was that
of 5 Hz and ensures a self-heating of specimens (ΔΤ) of less than 10 0C
during the test. Three types of fatigue tests were performed at a loading level
of 0.8 of the UTS; tests at continuous loading conditions, tests with “stops”
every 10 k cycles and tests with "stops" every 10 k cycles and application of
the healing cycle. All tests have been stopped at specimen's failure.
2.1.7. Low Velocity Impact and Post-Impact Testing

The core objective of the current characterization programme which is
described in section 7, is the determination of the LVI resistance and the
CAI strength of the proposed modified CFRP systems, since these
characteristics dominate the damage tolerance of the composite material
system. The LVI and CAI tests were performed at RT conditions according
to AITM1-0010: 2015 standard issued by Airbus. A guided drop weight

tester was utilized for this purpose. A drop tower equipped with a 16-mm
diameter hemispherical aluminum impactor weighing 2.5 kg was employed.
The selected impact energy was that of 25 J and was delivered by adjusting
the initial height of the impactor. The impactor was manually left to fall from
the pre-determined height and was arrested automatically after rebounding
to avoid a second strike. The residual compressive strength of the damaged
CFRP plates before and after the healing process was measured through CAI
tests. For this purpose, an anti-buckling jig was employed to support the
specimen edges and to inhibit Euler buckling. All tests were conducted at an
Instron (250 kN) hydraulic machine. The specimens were loaded under
displacement control of cross head velocity of 0.5 mm/min.
2.1.8. Three-Point Bending Testing

3PB tests were carried out following the ASTM D7264M-07 by using
an Instron (25 kN) hydraulic machine (sections 3 and 4). Five specimens
from each type of plates were extracted and tested. Centre loading was
applied on a simple supported beam. The load and corresponding mid-span
deflection were continuously recorded by the test machine until the failure
occurred. The flexural stiffness (Eflex) and flexural strength (σmax) of the three
material types were determined. σmax is calculated according to beam theory
from Equation 4(a), where P is the maximum applied load recorded during
testing. The maximum strain, developed at the outer surface, is calculated
by Equation 4(b), where d is the mid-span deflection of the beam. Eflex
corresponds to the linear portion of the resulting σ-ε curve.
3PL
σ = 2wh2 (MPa) (4(a))
6dh
ε= (4(b))
L2
where, σ is the stress at the outer surface in the load span region, P is the
applied load, L is the support span, w is the specimen width, h is the
specimen thickness, ε is the maximum strain at the outer surface and d is the
mid-span deflection.

2.1.9. Composites Quality Issues

C-scan inspection was performed on all the manufactured plates. A
Physical Acoustics Corporation (PAC) UT C-Scan system was used with a
5 MHz transducer. C-scan images of plates intended for mode I, mode II,
tensile, fatigue, LVI and compression tests confirmed good qualities and
showed absence of porosity and delaminations due to the manufacturing
process.
2.1.10. Healing Procedure and Healing Efficiency Calculations

After first crack propagation under mode I or mode II or delamination
damage due to fatigue and LVI described bellow (sections 3, 4, 5, 6 and 7
respectively), the specimens were subjected to a simple healing cycle of
heating under controlled through-the-thickness compression. The cycle
comprised a 5 min dwell at 150°C for CFRPs containing (BMI pp) grains
while 15 min dwell at 130°C for CFRPs containing the pure “in-house” BMI
in prepreg form. Both healing cycles were applied under compressive
loading of 1 kN for mode I, mode II and open holed samples while 5 kN
through the thickness compressive force was applied to LVI samples by
using a heat press machine. The applied temperature was chosen to be 20-
40°C higher than the Tm value (approximately 110°C) of the SHA material,
in order to be sure that the BMI will flow within the cracks (damage area)
and to achieve the healing effect. The aforementioned healing profiles were
chosen according to SHA manufacturer's guidelines. The compressive
loading value was chosen as the minimum necessary to ensure that the
adjacent crack flanks were kept in close proximity during healing activation.
The loading values were chosen as the minimum nessesary to ensure that the
adjacent crack flanks were kept in intimate contact during the healing
procedure. Then, the samples were left to cool-down at RT. After the healing
cycle, the samples were tested again using the same configurations. The
calculations of the healing efficiency (H.E.) of every system were based on
Equation (5) for mode I and II tests and on Equation (6) for LVI tests:
Mhealed
HE = M ∙ 100 (%) (5)
modified

Shealed
HE = S ∙ 100 (%) (6)
damaged
where M or S is the property under examination. Mhealed and Mmodified are the
values of the property after healing and before healing, respectively. Shealed
and Sdamaged are the values of the property after healing and after LVI testing
respectively.
3. TOUGHENING AND HEALING OF CONTINUOUS

CFRPS BY BISMALEIMIDE BASED POLYMER BLENDS:
EFFECTS OF HEALING AGENT CONCENTRATION
AND CURING CYCLE
3.1. Experimental
3.1.1. Test Programme Outline

The scheduled experimental test campaign organized and realized for
the needs of the current work is outlined in Figure 5. The work was divided
into three study levels. At first, CFRP plates containing 0 (reference), 60,
120, 180 or 240 grams per square meter (gsm) SHA in grains form in their
mid-plane interlaminar region were manufactured and tested under mode I
fracture loading conditions. Then, a simple healing cycle was realized to
repair the test induced delamination and the mechanical tests were repeated.
Thus, the effect of varying SHA concentrations into the toughening and
healing performance of the host laminate was examined. The highest
performing material group was selected for further assessment in the second
study level, where an alternative curing cycle was employed. The first curing
cycle (at 130°C and 6 bars for 2 h) was the manufacturer’s recommended
while the second one (at 100°C and 4 bars for 5 h) was based on previous
in-house DSC measurements. From the second study level, the effect of the
curing cycle on the toughening and healing performance of CFRPs was
verified. Finally, as a parallel third study, the effect of both the SHA

concentration and the curing cycle on the in-plane mechanical performance

of the (un-healed) modified composites was studied via 3PB tests. Here,
research work conducted in [35-37] is progressed a step forward. Previous
works [35-37] studied the nature of the utilized BMI pp and also preliminary
attempts to impregnate it in lab-scale polymer of composite coupons were
achieved.
Figure 5. Schematic presentation of the scheduled characterization campaign realized

in the current work consisting of three parallel individual studies.
3.1.2. Composites Production and Material Morphology

Examinations
The composite laminates were manufactured in-house by
prepreg/autoclave technologies. Five types of laminated plates with 22
unidirectional plies each were prepared; the reference laminate and four
modified laminates containing 60, 120, 180 and 240 gsm SHA in their mid-
thickness interlaminar region (Figure 6-a). The SHA created an “interleaf”
in the mid-plane of the composite, a configuration which is favorable for
future implementation purposes in CFRP structural components. The
photograph of Figure 6-b presents the mid-plane of a modified laminate
containing 120 gsm of SHA. All plates had dimensions of (300 × 150 × 3)
mm3. During the manufacturing two 13-μm-thick sheets of PTFE were

positioned in the mid-plane of all plates as shown in Figure 6-a, b to initiate

crack according to the referenced mode I test standard.
Figure 6. a) Design of the reference CFRP plate and of the modified CFRP plates
containing 60, 120, 180 or 240 gsm of SHA. b) A representative photograph of the
mid-plane of modified CFRP plates containing 60, 120, 180 or 240 gsm of SHA.
Figure 7. Schematic depiction of the modified DCB specimen configuration. a) Side

view. b) Top view. c) A snapshot just after the end of the test. Dimensions in mm.

Following lamination, the composites were vacuum bagged and cured

in an autoclave. As already mentioned, two curing cycles were applied for
the needs of the current study; (a) curing cycle 1: at 130°C and 6 bars for 2
h, according to manufacturer’s recommendation and (b) curing cycle 2: at
100°C and 4 bars for 5 h, according to in-house DSC data. The lower applied
pressure during the second curing cycle was selected in order the diffusion
of the SHA from the mid-thickness plane to the upper and lower neighboring
plies to be minimized. The diffusion process was clearly observed in [33].
Five Double Cantilever Beam (DCB) test specimens were precisely cut
out from each plate, whose geometry is illustrated in Figure 7-a, b. Two
aluminum tabs were appropriately glued using a two-component epoxy
adhesive to introduce opening load to the specimens during testing.
3.2. Results and Discussion
3.2.1. Manufacturing
The handling of the SHA powder at RT and the integration of it into
composite mid-plane did not create any concerns as previously mentioned
in paragraph 2.1.2. The thickening effect of the SHA in the entire composite
laminate was not significant; only 0.2 mm increase (2.9 vs. 3.1 mm) was
measured in the extreme case of the 240 gsm plate. Finally, it was shown
that both curing pressures applied did not significantly affect the thicknesses
of the plates. Thus, the Vf was detained close to 60% for all types of plates.
3.2.2. Effect of the Concentration of the Self-Healing Agent (SHA) on

the Toughening and Healing Performance of CFRP
It is reasonable that the concentration of the SHA in the mid-thickness
region of the composite laminates greatly affects the mode I interlaminar
fracture toughness of the laminates before and after healing. It is noted that
for the study presented in this paragraph, the curing temperature of the
composites was higher than the Tm of the SHA and it is believed that will
cause improvement to the adhesion between the epoxy matrix and the
healing material.

3.2.3. Toughening Performance of CFRP

In this paragraph, the effect of the SHA concentration (0 (reference), 60,
120, 180, or 240 gsm) located in the central region on the mode I
interlaminar fracture toughness characteristics (Pmax and GIC) is examined.
For this purpose load (P) vs. displacement (d) measurements were performed
as described in paragraph 2.1.3 for the 5 material groups.
The resulting P vs. d curves are illustrated in Figure 8-a. For all material
groups, the general trend does not present any dramatic difference; the
applied load was initially increased linearly followed by a visible deviation
from linearity and ended with a load drop just after crack propagation onset.
However, except for the 240 gsm modified composite, all the other modified
composites exhibited higher interlaminar fracture toughness performance
than the reference CFRP, as Figure 8-a suggests. Thus, the incorporation of
the BMI powder SHA into composite architecture has a potential to toughen
the final composite. Analogous test results were reported in preliminary
study [33] utilizing an analogous type of SHA. It is noted that no
considerable difference is observed between the various composites with
regard to their apparent stiffnesses (slopes of the linear portions of the
curves).
Figure 8. a) Representative load (P) vs. displacement (d) curves for the reference
CFRPs (0 gsm) and for the modified CFRPs with 60, 120, 180 or 240 gsm of SHA
during the mode I interlaminar fracture toughness tests. b) Synopsis of the results of
the mode I interlaminar fracture toughness tests of the CFRPs in terms of the
maximum load (Pmax) and the mode I interlaminar fracture toughness energy (GIC), as a
function of the SHA concentration (0 (ref.), 60, 120, 180 and 240 gsm).

Figure 8-b compares the critical mode I interlaminar fracture toughness

characteristics Pmax and GIC of the 5 material groups. Samples with 60 gsm
SHA increased the Pmax and GIC at the amount of 27% and 47%, respectively.
Samples with 120 gsm exhibited increases of 28% and 10% for the Pmax and
GIC, respectively. Samples with 180 gsm showed increases of 20% and 7%
for the Pmax and GIC, respectively. On the contrary, the concentration of 240
gsm slightly decreased the Pmax and GIC values (4% and 11%, respectively).
Thus, the incorporation of 60, 120, or 180 gsm SHA in the mid-plane region
increases both Pmax and GIC. Similar satisfactory results have also been
reported in literature for CFRPs containing polymers with thermally
activated healing functionalities, e.g., [33, 38-40].
Figure 9. a) Representative mode I crack opening resistance curves (R-curves) for the
reference CFRP (0 gsm) and for the modified CFRPs with 60, 120, 180 or 240 gsm of
SHA, where mode I interlaminar fracture toughness (GI) is given as a function of the
crack length (a). b) Snapshot of a typical DCB specimen during the test, in which the
characteristic orange colour of the SHA in the mid-thickness area is clearly illustrated.
Figure 9-a compares the crack propagation resistance curves (R-curves)

of all material groups. The capacity of the BMI-prepolymer in
concentrations of 60, 120, and 180 gsm to enhance the fracture toughness of
the final composite is prominent in this figure. Firstly, the R-curve of the
reference composite shows very limited damage process zone, lower than 5
mm; the plateau was showed just after reaching a crack length of 30 mm. On
the other hand, the 60, 120, and 180 gsm modified composites exhibited
slightly larger damage process zones (2-4 mm larger than the reference one).

240 gsm modified samples exhibited equal damage process zone with the
reference one. Finally, all BMI modified composites promoted relatively
stable delamination crack growths, as Figure 9-a implies.
Figure 10. Representative optical microscopy photographs from the cross-section area
of the manufactured CFRPs. a) 0 gsm (reference). b) 60 gsm. c) 120 gsm. d) 180 gsm.
e) 240 gsm. f) 120 gsm, after the curing cycle 2. Scale bar: 120 μm.
Figure 11. Representative optical microscopy photographs from the fracture surface
area of the manufactured CFRPs. a) 0 gsm (reference). b) 60 gsm. c) 120 gsm. d) 180
gsm. e) 240 gsm. f) 120 gsm, after the curing cycle 2. Scale bar: 120 μm.
The increase of the mode I interlaminar fracture toughness for samples

containing BMI powder of 60, 120, or 180 gsm is strongly attributed to the
developed fibre bridging phenomenon within the top and bottom crack

flanks (Figure 9-b), which in its turn suppressed the crack tip opening
stresses and thus reduced the crack opening displacement at a given applied
load. Thus, the resistance of the modified composites to both the initiation
and the propagation of the delamination damage was enhanced. Also, it was
observed during testing that with increasing SHA concentration the bridging
phenomena during testing were slightly decreased.
The interlaminar fracture performances of the various composites
(Figures 8, 9-a, b) are heavily relying to the internal structure of the materials
in the vicinity of the mid-plane interlaminar region. After tests, careful
microscopy studies on the cross-sections and fracture surfaces of the
samples were realized. First of all, it was observed that by increasing the
SHA concentration from 60 to 240 gsm both the thickness of SHA
interlayered zone and the thickness of the infiltrated neighboring plies are
increased, as shown in Figure 10. Ιn Figure 11 the lighting of the colour scale
is observed with increasing SHA interlayer thickness. In addition, the
previously mentioned slight increase of the damage process zone for
samples containing 60-180 gsm SHA (Figure 9-a) is attributed to the
enhanced ductile behaviour of the BMI material. Also, diffusion of the
epoxy system into the melted BMI-prepolymer powder, clearly observed in
previous work [33], intensified the plastic deformation phenomena during
the test. Additionally, the lower cleavage resistance of the 240 gsm samples
than the reference samples (Figure 8) could be attributed to the local increase
of the polymer volume fraction in the mid-thickness area. Thus, it is believed
that despite the fact that the BMI modification of the composites did not
affected the overall Vf of the plate (paragraph 3.2.1), it seems that the local
increase of the polymer phase in the mid-plane decreased the fibre bridging
phenomena during testing.
3.2.4. Healing Performance of CFRP

For the examination of the H.E. of the BMI modified CFRPs after first
testing, specimens were subjected to a repair process consisting of heating
under compression as described in paragraph 2.1.10. As already mentioned,
the healing parameters (i.e., temperature, pressure, and time) were selected
according to manufacturer’s guidelines.

Figure 12. Representative load (P) vs. displacement (d) curves of the modified CFRPs
with a) 60, b) 120, c) 180 or d) 240 gsm of SHA before healing and after two
consecutive healing cycles.
Figure 12 presents the P vs. d curves of the 4 modified composites before

and after two consecutive healing cycles while Figure 13 shows the
corresponding Pmax and GIC values respectively. Firstly, samples with 60 gsm
of BMI did not exhibit any measured healing functionality (see Figure 12-
a). On the other hand, samples with 120, 180 and 240 gsm of BMI exhibited
moderate healing performances after heating at 150°C for 5 min (see Figures
12-b, c, d), with the 120 gsm modified samples to exhibit the best H.E.
proportions. More precisely, the H.E. values of the 120 gsm samples for the
Pmax and GIC values were amounted to be 15% and 2% respectively after the
first healing cycle. These values were calculated to be 4% and 0.5%
respectively after the second healing cycle. The 180 and 240 gsm modified
samples exhibited slightly lower H.E. values. Samples with 180 gsm
exhibited H.E. values at the amount of 9% and 1% for Pmax and GIC
respectively, after the first healing cycle while 8% and 0.5% for P max and GIC
respectively after the second one. Finally, samples with 240 gsm exhibited
H.E. values at the amount of 14% and 1% for the P max and GIC values

respectively, after the first healing cycle while 7.5% and 0.5% for the P max
and GIC values respectively after the second one.
Based on theory, the ability of the BMI pp material to heal the cracks
and partially recover the interlaminar fracture toughness of the composites
depends upon the DA reaction that makes the material’s chains to reconnect
after the mechanical rupture that fracture tests introduce [41-42]. In addition,
the BMI-prepolymer material acquires sufficiently low viscosity above its
Tm that enables it to easily flow within the crack flanks and to reposition
itself equally over the debonded fracture surfaces. Nevertheless, the poor
H.E. values of all modified composites are mostly attributed to the high
temperature of the curing cycle (beyond the Tm of the SHA) that caused
infiltration of the SHA in the mid-thickness composite layers, as Figures 10
and 11 suggest.
Figure 13. Results of the mode I interlaminar fracture toughness tests of the reference
CFRP (0 gsm) and of CFRPs modified with 60, 120, 180 and 240 gsm SHA before
healing and after two consecutive healing cycles in terms of the maximum load (Pmax)
and the mode I interlaminar fracture toughness energy (GIC).
In addition, another reason related to this behaviour is the degradation

of the structural performance of the SHA due to the heat. Especially, for
samples with 60 gsm of (BMI pp) in their mid-plane being clearly
inadequate to heal the crack flanks after fracture, the SHA was disappeared

through infiltration during the curing. This response is consistent with the
microscopy photographs of Figures 10 and 11 in which the absence of the
reversible polymer is implied.
3.3. Effect of the Curing Cycle on the Toughening

and Healing Performance of CFRP
The interlaminar fracture toughness and the H.E. of the composites are
evaluated in this study with respect to the utilized curing temperature. The
curing cycle in low temperatures (i.e., lower than the Tm of the SHA) is
expected to increase the H.E. values up to a certain extend while adversely
affect the virgin fracture toughness of the composites. The 120 gsm
composite, exhibiting satisfactory toughening performance and the best H.E.
values was utilized as a representative one in order to determine how the
curing cycle affects the interlaminar fracture toughness of the composites,
before and after the healing process. The new curing cycle was that of 100°C
at 4 bars for 5 h. It is underlined that the new curing temperature of 100°C
remains 11°C lower than the Tm of the utilized SHA. DSC study showed that
both curing cycles exhibited the same polymerization proportions to the final
composites, something that was also confirmed by mechanical tests on the
reference specimens arose from both applied curing cycles (1 and 2).
3.3.1. Toughening Performance of CFRP

Figure 14-a, illustrates P vs. d curves under mode I fracture loading for
samples modified with 120 gsm BMI and cured in two different curing
cycles. It is shown that the curing cycle at 100°C slightly undermines the
mode I fracture toughness of the composite if compared with that at 130°C;
the Pmax and GIC values were decreased at the amount of 11% and 3%,
respectively (see Figure 16). This behaviour is attributed to the weakening
effect between the solid SHA and the epoxy system that the 30°C lower
curing temperature caused. In addition, during the second curing cycle the
SHA was not effectively infiltrated into the neighboring layers as Figures

10-f and 11-f indicate and thus the BMI-prepolymer remained intact in the
mid-plane of the composite.
3.3.2. Healing Performance of CFRP

Figure 14-b compares the typical P vs. d curves before and after healing
for samples cured according to the curing cycle 1 and 2 respectively.
According to these curves, it was shown that the healing functionality of the
BMI-prepolymer has been significantly increased after the second curing
cycle. The Pmax and GIC values for samples cured in the two different curing
cycles before and after two consecutive healing cycles are presented in
Figure 15. The H.E. values obtained for the Pmax and GIC values are 26% and
32% respectively after the first healing cycle while 17% and 23%
respectively after the second healing cycle. These H.E. values are 11% and
30% higher for the Pmax and GIC values respectively after the first healing
cycle while 13% and 22.5% higher after the second healing cycle (if
compared with the H.E. values achieved following the curing cycle 1).
The significant increase of the H.E. values is attributed to the strong
presence of the SHA on the mid-plane that led to an increased healing
functionality after the curing cycle 2. In addition, the curing parameters of
the curing cycle 2 (lower applied temperature and pressure) caused the SHA
not to be infiltrated through the mid-thickness neighboring composite plies,
as occurred in the case of the curing cycle 1. Furthermore, the degradation
effect of the SHA after two consecutive healing cycles is prominent in
optical microscopy photographs arose from the fracture surface area of the
DCB samples after testing (Figure 14-c). This behaviour is attributed to the
heat that was applied during the consecutive healing cycles. In these
photographs, the darkening of the excess of the SHA on the mid-plane after
the second healing cycle due to the heat is clearly illustrated (Figure 14-c-ii)
while the initial shiny form after fracture is also illustrated in Figure 14-c-i.
To sum up, the curing cycle 2 at 100°C slightly decreased the
toughening performance of the composite while increased the healing
functionality. Based on this, an engineer has the option to choose the curing
cycle in order to achieve the desired trade-off between these two competing
featured of the composites under investigation.

Figure 14. Representative load (P) vs. displacement (d) curves during the mode I
interlaminar fracture toughness tests of the modified CFRPs with 120 gsm of SHA: a)
Cured in two different cycles. b) (left) Before and after two consecutive healing cycles
for samples cured in curing cycle 1. (right) Before and after two consecutive healing
cycles for samples cured in curing cycle 2. c) Representative optical microscopy
photographs from the fracture surface area of the CFRPs containing 120 gsm, after the
curing cycle 2: (i) Before healing, and (ii) after the second healing cycle. Scale bar:
120 μm.
Figure 15. Illustration of the maximum load (P max) and the mode I interlaminar fracture
toughness energy (GIC) values for samples containing 120 gsm of SHA and cured in
two different curing cycles, before and after two consecutive healing cycles.

3.4. Effect of the SHA Concentration and of the Curing Cycle on

the in-Plane Mechanical Performance of CFRP
The incorporation of the BMI pp grains in various concentrations into

composites’ architecture and the application of two different curing cycles
are expected to have an impact on in-plane mechanical properties of the
entire composite. For this purpose, 3PB tests were conducted for reference
and BMI-modified CFRPs cured according to the curing cycle 1 or 2, as
described in paragraph 2.1.8. All tests were performed before the healing
process.
Firstly, test results show that following the curing cycle 1 both flexural
properties (flexural modulus (Eflex) and flexural strength (σmax)) exhibited
decreasing trend with increasing SHA concentration. Samples with 240 gsm
exhibited reduced Eflex and σmax by almost 10% and 20%, respectively (see
Figure 16-a). In addition, the flexural properties of reference composite were
not affected by the application of both curing cycles (see Figure 16-b). On
the other hand, samples containing 120 gsm of BMI pp grains showed
slightly reduced flexural properties by the application of the curing cycle 2
(see Figure 16-b), probably due to poor bonding created between SHA and
the epoxy system
Figure 16. Determination of the (a) knock down effect that the BMI prepolymer (BMI
pp) grains in different concentrations (60-240 gsm of SHA) caused to the entire
composite in terms of the flexural modulus (Eflex) and flexural strength (σmax) after
curing cycle 1, b) knock down effect of reference and selected modified CFRPs (with
120 gsm of SHA) on the flexural modulus (E flex) and flexural strength (σmax) values
after both curing cycles and comparison.

4. MODE I INTERLAMINAR FRACTURE TOUGHENING

AND HEALING OF CARBON FIBRE/EPOXY COMPOSITES
WITH BIS-MALEIMIDE BASED PREPREGS
4.1. Experimental
4.1.1. Composites Manufacturing

Two UD laminated plates with 22-layers each were manufactured for
the needs of the current study; the reference laminate and the modified
laminate with two BMI prepregs on the mid-plane, both intended for mode
I interlaminar fracture tests. Figure 17 shows schematically the plate’s
configuration. The dimensions of the plates were 300 mm x 150 mm x 3
mm. During the manufacturing process two 13-μm-thick sheets of PTFE
film were placed in the mid-thickness plane of both laminates as shown in
Figure 17 to act as initial pre-crack according to the request of the
interlaminar fracture test.
Figure 17. Design of the (a) reference and (b) modified composite plate. (c)
Photograph of the CF/BMI ply placement on modified plate’s mid-plane, together with
the PTFE initial crack formation.

Following the lay-up, the laminates were vacuum bagged and cured in
an autoclave for 2 h at 130°C under 6 bars applied pressure, according to the
prepreg manufacturer guidelines. Thickness measurements were also
performed. Five DCB samples were cut from both the reference and the
modified plate. Two aluminum tabs were glued on the DCB specimen outer
surfaces (Figure 18) using a two-component epoxy adhesive in order to
transfer the desired loading.
Figure 18. Schematic depiction of the modified double cantilever beam (DCB) test
specimen configuration showing the interply location of the modified prepreg plies. (a)
Side view. (b) Top view. Dimensions in mm.
4.2.1. Mode I Interlaminar Fracture Resistance

Two indicative load vs. displacement curves from the mode I
interlaminar fracture toughness testing of the reference and the BMI-
modified CFRPs are compared in Figure 19-a. For both material types, a
linear increase of the load is initially observed followed by a notable
deviation from linearity, ending with a load drop after the crack initiation.
The capacity of BMI to significantly increase the recorded test Pmax (which

corresponds to the onset of the crack growth) is clearly observed from these
curves. Additionally, the load remains relatively high during the propagation
of the crack as compared to the load of the reference specimen. In addition
to Pmax, the calculated GIC is significantly increased in the case of the BMI-
modified material. The synopsis of the processed experimental results that
show the considerable increase of the recorded Pmax (~150%) and the
computed GIC (~330%) is presented in the histogram of the Figure 19-b.
In Figure 20, the extended fibre bridging and BMI ligaments phenomena
by the presence of pure BMI between the adjacent crack flanks of the
specimen are clearly visible. The developed bridging carbon fibres and BMI
ligaments within the top and bottom crack surfaces reduced the crack
opening displacement (COD) at a given applied load (P) and thus resulted
to lower stresses at the crack tip, concluding to an increased GIC. Thus, the
BMI is characterized by fine adhesion properties with carbon fibres and also
high peel resistance. It is noted that concerning the stiffness (slope of the
linear portion of the curve), no considerable difference has been observed
between the two material groups for all the specimens tested.
Figure 19. (a) Representative load (P) vs. crack opening displacement (d) curves
during the mode I interlaminar fracture toughness testing of the reference and the
modified CFRPs. (b) Comparison between the reference and the modified CFRPs in
terms of the maximum load (Pmax) and the mode I interlaminar fracture toughness
energy (GIC).
Typical R-curves for the reference and the modified composites, under
mode I loading conditions are depicted in Figure 21. As it is shown in Figure
21, the reinforcing effect of the BMI polymer is prominent. The presence of

the BMI polymer promotes the development of a stronger bridging traction

zone at the interlaminar region, which suppresses the crack tip opening
stresses and as a result increases both the resistance of the composite to the
initiation and propagation of the delamination damage.
Figure 20. (a) Snapshot during the mode I interlaminar fracture toughness testing of a
modified DCB specimen. (b) Zoom in the marked region of part (a) where the
extended fibrebridging phenomena between top and bottom crack flanks are visible.
Figure 21. Indicative interlaminar crack growth resistance curves (R-curves) of the
reference and the ΒΜΙ modified CFRPs, where the mode I fracture toughness energy
(GI) is given as a function of the crack length (a).
The R-curve for the reference composite shows a plateau value after
reaching a crack length of 30 mm, having a very limited damage process
zone (around 5 mm) indicative of stable delamination crack growth process.
On the other hand, the R-curve for the BMI modified composite reaches a

plateau value at a crack length of 45 mm, having a larger damage process

zone close to 20 mm. The BMI modified composite prior to healing not only
achieves to significantly increase the mode I fracture toughness but it also
promotes a relatively stable delamination crack growth as it is shown in
Figure 21. The R-curve for the modified composite was also steady due to
the continuous distribution of BMI along the delamination crack plane.
Figure 22. Cross-section micrographs of the (a) reference and (b) modified CFRPs.
The extended diffusion of the BMI into the upper adjacent reference plies and the
diffusion of the epoxy system into the BMI composite plies are visible.
Figure 23. Fracture surfaces of the (a) reference and (b) modified CFRPs. The
darkened surface of the modified CFRP indicates the strong presence of the BMI
reversible polymer.
The significant toughness increase obtained by the introduction of the

BMI composite layers is attributed to the BMI material’s ductile behaviour.

In Figure 22 the mid-plane cross section area of both materials is examined

via optical microscopy. The thickness of the two BMI-impregnated carbon
fibre layers was calculated to be approximately 360 μm. The interface
created between the BMI and the epoxy matrix seems coherent and strong;
a strong bonding between the BMI and the epoxy seems to be established at
first glance. This behaviour is visible in Figure 22-b, where the BMI material
diffuses into the closest upper carbon epoxy prepreg layers (dark region of
the matrix next to the BMI prepregs) during the autoclave process.
Additionally, an analogous diffusion process was carried out for the BMI-
impregnated layers, in which the epoxy system has been diffused. The
diffusion of the epoxy system into the BMI-impregnated layers caused
greater plastic deformation during the mode I experiments, resulting in
increased fracture toughness characteristics.
It was observed that the failure of the matrix for the reference and the
modified materials was clearly performed in brittle and ductile manner
respectively (see Figure 23). As can be seen in the photograph of the fracture
surface for the reference samples (see Figure 23-a) the carbon fibres remain
intact. On the other hand in the case of the BMI modified samples the
fracture surface has been darkened due to the presence of the BMI polymer,
while cleavage fracture surface and pulled out carbon fibres are observed
(see Figure 23-b). This behaviour, is attributed to the intermediate adhesion
strength that was created between the BMI polymer and the carbonfibre that
resulted the crack propagation to be held along the fibre-matrix interfaces
and caused extended fibre bridging. The fibre bridging in its turn increased
the fracture toughness characteristics.
In the present work, the effect of the incorporation of the two BMI/CF
layers on the mid-plane with precise thickness (360 μm) was investigated.
The modified layers which manufactured had comparable Vf (approximately
55%) with the final reference plate (approximately 60%). Nevertheless,
further assessment of an optimum thickness for the impregnated layers needs
to be performed in order the maximum mode I fracture characteristics to be
achieved.

4.2.2. Mode I Interlaminar Fracture Recovery

The use of the BMI polymer, further to the increase of the interlaminar
fracture toughness is to investigate the concept of self-healing within the
composites. To that direction, after the completion of mode I interlaminar
fracture tests in the case of BMI modified composites, the tested samples
followed a healing cycle as described in paragraph 2.1.10. After the healing
process, the samples were tested again using the same mode I interlaminar
fracture configuration.
In Figure 24-a typical P vs. d curves for mode I interlaminar fracture
tests are depicted before and after the application of the healing cycles.
According to experimental results, it was shown that the BMI material
provided a partial recovery of mode I interlaminar fracture toughness of the
composite after the application of the healing process. As it can be seen in
the resulting P vs. d curves for the BMI composites before and after the
consecutive healing cycles (see Figure 24-a), recoveries around 29.4% and
19.2% for the Pmax and GIC values respectively after the first healing cycle
were observed. In the subsequent two healing cycles on the same samples, a
very slight increase in H.E. values was observed. More precisely the Pmax
and GIC values after the third healing cycle were calculated to be 29.9% and
23.4% respectively. Finally, after the fifth healing the recoveries of Pmax and
GIC values amounted at the level of 23.7% and 16.5% respectively, when
compared to the pristine samples (see Figure 24-a and 24-b for the Pmax and
GIC values respectively). When compared to the values after the fourth
healing cycle a reduction at the amount of 20% and 15% for the Pmax and GIC
values respectively was observed.
Although the bridging phenomenon was observed again during the
mode I experiments for the healed samples, its efficiency appeared to
decrease with increasing the number of the healing cycles. This behaviour is
attributed to the more and more fractured and pulled out fibres during the
mode I fracture events that are not able to be healed. The only one material
that takes part in the healing process and heals the crack is the BMI
reversible polymer.
The ability of the BMI material to heal the cracks and the partial
recovery of the fracture properties depends upon the DA reaction that makes

the material’s chains to reconnect after mechanical rupture occurs during the
mode I tests. In addition, the BMI material acquires sufficiently low
viscosity above its Tm that enables it to easily flow into the crack flanks and
to reposition itself equally over the debonded fracture. Therefore, the slight
increase of the Pmax and the GIC values while increasing the number of
healing cycles up to third one can be attributed to the crosslinking reactions
occurred between functional groups into the reversible material (post curing
effect). After the fourth healing cycle a type of degradation of the healing
properties of the BMI material was observed because of the successive
heating imposed for the healing activation. During the fourth and fifth
healing cycle the degradation of the BMI material occurs due to the fact that
more and more chain bonds cannot be repaired.
Figure 24. (a) Representative load (P) vs. crack opening displacement (d) curves
during the mode I interlaminar fracture toughness testing of the modified CFRP before
and after the healing cycles, (b) The effect of the number of healing cycles on the
maximum load (Pmax) and the mode I interlaminar fracture toughness energy (GIC) of
the modified CFRP.
The typical R-curves under mode I loading conditions before and after
the application of the healing cycles are depicted in Figure 25 and they are
consistent with the load-displacement data. The general trend is that with
increasing number of healing cycles up to third one, the GIC values increase
slightly while after the fourth healing cycle decrease. The plateau value of
the GIC is reached at a later stage than the corresponding plateau value of the
pristine composite; at 90 mm instead of 45 mm. In line with the pristine
sample, the plateau value remains almost constant until a crack growth

length of about 60 mm. After that crack length, a slight decrease is observed
until the end of the test. However, this decrease is smoother for the healed
BMI samples and their fracture process zone is greater when compared to
the pristine samples.
Figure 25. Representative interlaminar crack growth resistance curves (R-curves) of

the modified CFRP where the mode I fracture toughness energy (GI) is given as a
function of the crack length (a), before healing and after each healing cycle.
4.3. Knock-Down Effect on in-Plane Mechanical Properties
The knock down effect that the two pure BMI prepregs caused on the
flexural performance of the composites is assessed in this paragraph. For this
purpose, 3PB tests were conducted for reference and BMI modified CFRPs
as described in paragraph 2.1.6. According to these experiments, it was
shown that the Eflex value was decreased by almost 15% (from approx. 55.5
GPa to 47 GPa) while σmax value was reduced by 35% (from approx. 638
MPa to 420 MPa). These results are illustrated in Figure 26 and fully aligns
with other investigations [43] in which it was proven that the incorporation
of thermoplastic additives into the composite’s matrix cause softening to the

entire composite as a result the degradation of their in-plane properties.

More information about the current section (section 4) can be found in [33].
Figure 26. Identification of the knock down effect that the two pure BMI prepregs
caused to the flexural modulus (Eflex) and flexural strength (σmax) of the CFRP.
5. MODE II FRACTURE TOUGHENING AND HEALING

OF CONTINUOUS CFRPS CONTAINING POLYMERS
BASED ON DIELS-ALDER REACTION
5.1. Experimental

the needs of the current work is outlined in Figure 27. The current work is
divided into two study levels. The first study level investigates plates
containing 0, 60, 120, 180 or 240 gsm of BMI pp SHA in grains form while
the second level study investigates plates containing pure BMI prepregs. The
aforementioned material groups were utilized and extensively tested under
mode I loading conditions, as described in previous sections (3 and 4). Both
types of SHAs were placed in their mid-thickness region. For both studies
the manufactured plates were intended to mode II fracture loading tests.

Then, a simple healing cycle was realized to repair the delamination and the
mechanical tests were repeated. Thus, the effect of varying SHA form and
concentrations into the toughening and healing performance of the host
laminate was examined.

in the current work.

The composite laminates were manufactured in-house by
prepreg/autoclave technologies. Six types of laminated plates with 22 plies
each stacked unidirectionally were prepared; the control laminate (see
Figure 28-a), four modified laminates containing 60, 120, 180 and 240 gsm
of BMI pp grains respectively and the modified laminate containing two
BMI prepregs. Both types of SHAs were incorporated into their mid-
thickness interlaminar region. The BMI pp SHA created an “interleaf” in the
mid-plane of the composite while primary layers of the composite were
replaced by pure BMI prepregs. Both configurations are favorable for future
implementation purposes in FRP structural components.
Figure 28-b, c present the geometry of both modified plates as well as
the position of both SHAs. All plates had dimensions of (300 × 150 × 3)
mm3. During the manufacturing two 13-μm-thick sheets of PTFE were
positioned in the mid-plane of all plates as shown in Figure 28-a, b, c to
initiate crack according to the referenced standard. The photograph of Figure
28-d, e presents the mid-plane of both modified laminates containing BMI
pp grain zones or pure BMI prepregs as SHAs.
Following lamination, the composites were vacuum bagged and cured
in an autoclave at 130°C and 6 bars for 2 h, according to manufacturer’s
recommendation. Five end-notched flexure (ENF) test specimens were

precisely cut out from each plate, whose geometry is illustrated in Figure
29-a, b. Two aluminum tabs were appropriately glued using a two-
component epoxy adhesive to introduce opening load to the specimens
during testing.
Figure 28. Design of (a) the reference CFRP plate, (b) the modified CFRP plates
containing 60, 120, 180 or 240 gsm of SHA, (c) the modified CFRP plates containing
two pure BMI prepregs. (d) Representative photograph of the mid-plane of (a) the
modified CFRP plates containing 60, 120, 180 or 240 gsm of SHA in powder form, (e)
the mid-plane of modified CFRP plates containing two pure BMI prepregs.
5.1.3. Mode II Interlaminar Fracture Toughness Testing

Mode II interlaminar fracture toughness tests were performed according
to specifications described in paragraph 2.1.4. Based on Equation (2), the
crack length (a) has to be known in order the GIIC value to be calculated. For
brittle materials the moment of unstable crack growth can be clearly
determined; Pmax is attributed to the critical load at the onset of crack growth.
On the other hand, for pure BMI modified composites containing ductile

polymer in which the crack is growing through the ductile material, the crack
initiation (Pini) occurs long before Pmax. In addition, in mode II loading of the
pre-cracked specimen the crack tends to close which hinder a clear
visualization of its tip. Thus, mode II fracture characterization of CFRPs
with a ductile interlayer remains a major challenge including uncertainties
on the results. Figure 29, depicts the modified ENF test specimen
configuration as derived from DCB specimen.
Figure 29. Schematic depiction of both types of modified end-notched flexure (ENF)
specimen configurations, as derived from (a, b-i) double cantilever beam (DCB)
specimens. Depiction of the modified samples with (a) BMI prepolymer (BMI pp)
grains and (b) two pure BMI prepregs. (a, b-ii) Side view, (a, b-iii) Top view.
Dimensions in mm.
Figure 30. Experimental set-up of the mode II interlaminar fracture toughness tests
where the high-resolution camera monitoring of the crack evolution is seen.

Based on these, whereas in the case of reference and BMI pp case

CFRPs the GIIC calculations were extracted directly from the P-d data
without any concern, for the pure BMI modified CFRPs the monitoring of
the crack growth using the high resolution digital camera system was carried
out. It is suggested that the crack initiation point of the pure BMI modified
CFRPs corresponds to the onset of the visual deviation of linearity in P-d
curves. In addition, a vertical line with pencil was marked at the end of the
starter film, as shown in Figure 30. It is suggested that the separation of the
line into two lines corresponds to the crack initiation.

The handling of both SHAs (BMI pp and pure BMI prepregs) at RT and
the incorporation of them into composite's mid-plane did not create any
concerns as previously mentioned in paragraph 2.1.2. The thickening effect
of the SHA in the entire composite laminate was not significant; only 0.2
mm difference (2.9 vs. 3.1 mm) was measured in the extreme case where
240 gsm or BMI prepregs were incorporated into composite mid-plane.
Finally, it was shown that Vf of the manufactured plates for all types of plates
was retained close to 60%.
5.2.2. Effect of the SHA Concentration and Form on the Toughening

and Healing Performance of CFRP
It is reasonable, that the concentration and the form of the SHA in the
mid-thickness interlaminar region of the composite laminate greatly affects
the mode II fracture toughness performance of the laminate in before and
after the application of the healing cycles. In the present paragraph, the effect
of the SHA on composites' architecture is assessed.
5.2.2.1. Toughening Performance of the CFRPs

In this section, the effect of the concentration and the form of SHA on
the mode II interlaminar fracture toughness characteristics Pmax and GIIC is

examined. P vs. d measurements were performed for all material groups;

reference and modified by various concentrations of BMI pp as SHA or pure
BMI prepregs in their mid-plane area, as described above. The resulting P
vs d curves are illustrated in Figure 31-a and c for samples containing BMI
pp grains and pure BMI prepregs respectively. The general trend does not
present any dramatic difference for reference and all BMI pp modified
material groups; the applied load increases linearly until the onset of the pre-
crack propagation where the load drops abruptly. On the other hand, despite
the similarity in failure mode of all BMI pp material groups, pure BMI
prepreg modified samples exhibited ductile behaviour during testing as the
initially linear P-d relation is followed by a distinguishable deviation from
linearity and much later the load drop appears. Whereas the load drop for
the reference and all BMI pp cases is attributed to the crack evolution onset,
for the pure BMI prepreg modified CFRP the Pmax indicates the load bearing
capacity of the composite before undergoing major damages.
For BMI pp modified samples, it was shown that the higher the SHA
concentration (gsm) the higher the mode II interlaminar fracture toughness
characteristics (approx. 25% and 60% increase for the Pmax and GIIC values
respectively, for samples containing 240 gsm). Only samples containing 60
gsm of BMI pp exhibited slightly decreased properties (8% and 10%
reduction for the Pmax and GIIC values respectively). However, except for the
modified composite with 60 gsm in the mid-plane, all the other modified
composites exhibited higher measured Pmax and calculated GIIC from the
reference CFRP (Figure 31-b). Also it was observed the higher the
concentration the higher the displacement at failure (60% higher for
modified samples with 240 gsm). Additionally, the laminate bending
stiffness was increased with the increase of the SHA concentration (17%
higher for samples containing 240 gsm). Thus, one can say that the
incorporation of the SHA in powder form into the composite generally
caused toughening to the final composite.
Interestingly, modified samples with pure BMI prepregs exhibit a
significantly higher Pmax and respective displacement. As it is seen in Figure
31-d, Pmax has been increased by 44% if compared with the reference CFRP,
while the displacement at failure is approximately 80% larger. From the

recorded mechanical data the GIIC was also calculated. For the calculation,
the displacement (d) at crack propagation onset is needed to be known. The
crack initiation points for the reference and the pure BMI modified CFRP
was determined according to specifications described in paragraph 5.1.3 and
are marked in Figure 31-c. The GIIC for both material sets was calculated
using in Equation (2) the (P, d) values that correspond to the indicated points
of the curve. The average values and standard deviation are shown in Figure
31-d. The GIIC values for the pure BMI modified CFRP was retained at the
same levels which is related to almost the same corresponding load (P) and
displacement (d) recorded at that instance. In addition, the laminate bending
stiffness is slightly increased by 10%.
Figure 31. (a) Representative load (P) vs. displacement (d) curves for reference and
modified CFRP specimens with 60, 120, 180 or 240 gsm of BMI prepolymer (BMI pp)
grains, during mode II interlaminar fracture toughness tests. (b) Synopsis of the results
of the mode II interlaminar fracture toughness tests of the CFRPs in terms of the
maximum load (Pmax) (left axis) and the mode II interlaminar fracture toughness energy
(GIIC) (right axis), as a function of the BMI pp concentration (ref., 60, 120, 180 and
240 gsm). (c) Representative load (P) vs. displacement (d) curves for reference and
modified CFRP specimens with pure BMI prepregs, during the mode II interlaminar
fracture toughness tests. (d) Synopsis of the results of the mode II interlaminar fracture
toughness tests of the reference and pure BMI prepreg modified CFRPs.

In Figure 31-b, d the general reinforcing effect of both SHAs is

prominent. As post-tests inspection for BMI pp modified samples revealed
that the in-plane crack propagated through the BMI pp rather than along the
SHA epoxy interface. For samples containing two BMI prepregs in the mid-
thickness area it is prominent that the in-plane crack propagated through the
pure ductile BMI polymer.
According to optical microscopy examinations conducted and provided
in previous subchapters (3 and 4), it was shown that both SHAs have been
infiltrated into epoxy matrix which in its turn resulted to stronger interfaces.
This observation indicates that the interface of the BMI pp interleaf to the
host matrix is stronger than the properties of the interleaf, thus guiding the
crack to a lower energy path through the interleaf. This behaviour essentially
maximizes the effectiveness of the interleaf and explains the higher
resistance to crack propagation. An analogous behaviour occurred for
samples containing pure BMI material into the mid-plane in combination
with a diffusion process that was carried out for the BMI-impregnated
layers, in which the epoxy system has been diffused. The diffusion of the
epoxy system into the BMI-impregnated layers caused greater plastic
deformation during mode II experiments resulting in increased fracture
toughness characteristics.
In addition, high displacements are recorded for both BMI modified
samples while the load for BMI pp samples to increase steadily with
increasing the concentration. According to mode I experiments conducted in
these composites, presented in section 3, it was shown the higher the
concentration the higher the bridging phenomenon created during testing.
Even when the crack initiated, the presence of the BMI pp interleaf arrested
the crack and prevented propagation due to the developed bridging with
increase the prepolymer concentration within the crack sides. All these
mechanisms described above, resulted in a lowering of the stress at the crack
tip and consequently in an increased mode II interlaminar fracture toughness
characteristics for both types of BMI modified samples.
As a matter of fact, similar findings have been reported in by Aksoy and
Carlsson where thermoplastics were investigated for interleave toughening.
The high deformation of the interleaf resulted in the creation of ‘shear-lips’.

The deformation mechanism of the interleaf is believed to be similar to the

formation of “shear lips” in mode II fracture of ductile thermoplastic
polymers [44]. In Groleau et al. investigation [45] a similar mechanism is
described in which the interlayered nylon particles played the role of
toughening agent into the composite while exhibited plastic deformation and
bridging. During the bridging process the particles absorbed energy through
extensive plastic deformation and thus the GIIC value was increased.
5.2.2.2. Healing Performance of CFRP

For the examination of the healing performance of both BMI modified
CFRPs (BMI pp and pure BMI prepregs), after first testing the ENF
specimens were subjected to a repair cycles consisting of heating under
compression as described in 2.1.10. As already mentioned, the healing
characteristics (i.e., temperature, time and loading) were selected according
to manufacturer’s guidelines.
The curves and the HE values of all BMI pp modified composites (60-
240 gsm) before and after the application of four consecutive healing cycles
are shown in Figure 32. Figure 32-a, 32-b, 32-c, 32-d and 33 illustrate results
from samples containing 60, 120, 180 and 240 gsm of BMI pp grains
respectively into the mid-thickness area. According to these results, it was
shown that all material groups were able to withstand two consecutive
healing cycles. Before and after two consecutive healing cycles, curves
exhibited the same trend; the applied load increases linearly until the onset
of the pre-crack propagation where the load drops abruptly. After the third
and fourth healing cycles the P-d curves exhibited linear behaviour (3PB
mode) as crack initiation point did not occur (samples were not healed). It is
prominent that the healing functionality of the SHA after two consecutive
cycles was degraded due to the heat during healing cycle above 130°C. In
addition, as mentioned in section 3 the curing cycle of these plates at 130°C
made the SHA grains to be infiltrated into epoxy matrix as a result the SHA
to be inadequate to fully repair or withstand more healing cycles. As it was
proved in this chapter curing at lower temperatures offers better healing
functionality of these BMI modified CFRPs. This behaviour was also
confirmed by optical microscopy examinations presented in section 3.

Despite the poor healing functionality of the BMI pp grains, all material
groups presented in Figure 32 exhibited high HE values for the two applied
consecutive healing cycles due to the nature of the mode II testing. An
analogous behaviour was also observed in previous works by the author in
which conventional CFPRs were modified with supramolecular interlayers
[39] or nylon micro-particles [46]. More precisely, Pmax values for all
material groups exhibited decreasing trend except for samples containing
240 gsm of BMI pp (i.e., from approx. 83% after 1st healing cycle to approx.
87% after 2nd healing cycle). This phenomenon was also observed in
previous section (section 4) and it is a common behaviour for polymers
based on DA reaction. This behaviour is attributed to crosslinking reactions
occurred between functional groups into the reversible material.
On the contrary, GIIC values were slightly increased after two
consecutive healing cycles as a result increased H.E. values. This behaviour
is also prominent in [39] and is attributed to the increased displacement
values (d) after the consecutive healing cycles, in combination with the
slight decrease of the Pmax values that led to increased GIIC values according
to Equation (2). In addition, the apparent bending stiffness of all material
groups gradually decreases with increasing the number of healing
activations. Such behaviour could be attributed to the extended
load/displacements which progressively damaged the highly loaded outer
plies of the specimens an observation was also reported in similar studies of
[39] and [46].
The P vs. d curves and the H.E. values of the modified composites with
two pure BMI prepregs into their mid-plane area, before and after the
application of two consecutive healing cycles are shown in Figure 34-a.
There are three curves; one for the initial test (before healing) and one after
each consecutive healing test. All curves exhibited ductile behaviour and the
same trend; during testing as the initially linear P-d relation is followed by a
distinguishable deviation from linearity and much later the load drop appears
(especially in the before healing case, in which the load drop is more
evident). The crack initiation points for all curves are given in Figure 34-a
while the results for the Pmax and GIIC values in Figure 34-b. According to
these results it was shown that both values exhibited a decreased trend and

reduced H.E. values after two consecutive healing cycles. More precisely,
recoveries around 66%, 88% and 60% for the Pini, Pmax and GIIC values
respectively after the first healing cycle were observed, while 53%, 81% and
45% respectively for these properties after the application of the second
healing cycle.
Figure 32. Representative load (P) vs. displacement (d) curves of the modified CFRP
with (a) 60, (b) 120, (c) 180 or (d) 240 gsm of SHA before healing and after four
consecutive healing cycles.
The ability of the BMI material to heal the cracks and the partial
recovery of the fracture properties depends upon the DA reaction that makes
the material’s chains to reconnect after mechanical rupture (that occurs
during the mode II interlaminar fracture toughness tests).
In addition, the BMI material acquires sufficiently low viscosity above
its Tm that enables it to easily flow within the crack flanks and to reposition
itself equally over the debonded fracture.

Figure 33. Experimental results arose from mode II interlaminar fracture toughness
tests of modified CFRPs with 60, 120, 180 and 240 gsm of SHA before and after two
consecutive healing cycles, in terms of the maximum load (P max) and the mode II
interlaminar fracture toughness energy (GIIC).
Figure 34. (a) Depiction of the representative load (P) vs. displacement (d) curves for
modified CFRPs containing two pure BMI prepregs, before healing and after two
consecutive healing cycles. (b) Results of mode II interlaminar fracture toughness tests
for modified CFRPs containing two pure BMI prepreg before healing and after two
consecutive healing cycles, in terms of the maximum load (P max) and the mode II
interlaminar fracture toughness energy (GIIC).
The decreasing trend that the mode II interlaminar fracture properties

exhibit with the application of the consecutive healing cycles is attributed to
the degradation of the BMI material due to the heat. Degradation occurs due
to the fact that more and more chain bonds cannot be repaired after each
fracture event.

6. FATIGUE BEHAVIOUR OF OPEN-HOLED CARBON

FIBRE/EPOXY COMPOSITES CONTAINING
BIS-MALEIMIDE BASED POLYMER BLENDS
6.1. Experimental

the needs of the current work is outlined in Figure 35. The work was divided
into three study levels as already mentioned in paragraphs 2.1.5 and 2.1.6.
The first study level investigates the fatigue life performance of CFRPs at
continuous loading conditions. On the other hand, at the second study level
the fatigue experiments were paused every 10 k cycles in order the damage
evolution within the composite to be determined through C-scan inspections.
Finally, in the third study level the same route as in the second level study
was followed together with the application of the healing cycles. All samples
were tested to failure. Thus, this section thoroughly investigates the effect
of the healing cycles on the fatigue life performance of CFRPs.

in the current work.

Two types of quasi-isotropic laminated plates containing 16 plies, with
[45/-45/0/90]2S stacking sequence each were manufactured for the needs of
the current study; the reference laminate, the modified laminate with BMI
pp grains between the primary layers of the CFRP at the amount of 120 gsm.
Both material groups (reference and modified) were tested under quasi-static

tensile and tension-tension fatigue loading conditions. Figure 36, shows

schematically the configuration of the plates (Figure 36-a) and the position
where the BMI pp (Figure 36-b, 37-b, c) was placed. The BMI pp grains
were placed in symmetrical fashion, in between all -450/00 group layers (4
zones 38 mm wide into the centre of the prepreg layer as analytically
presented in Figure 37-b, c) within the composite structure. Following the
lay-up, the laminates were vacuum bagged and cured in an autoclave for 2 h
at 130°C under 6 bars applied pressure, according to the prepreg
manufacturer guidelines. The dimensions of the final plates were 300 mm ×
150 mm × 2.1 mm. Five samples for each type of CFRP were exposed to
quasi-static tensile tests while fifteen samples were tested under fatigue
loading conditions (five samples for each category of fatigue test described
above respectively).
Figure 36. Design of (a) the reference CFRP and (b) the modified CFRP with BMI
prepolymer (BMI pp).
Open hole test specimens were precisely cut out from each plate, whose
geometry is illustrated in Figure 38-a. The hole was made by using a vertical
hole drill with diameter size of 6 mm for all material sets. Special care was
taken during the drilling process in order to ensure that creation of the
specimen hole does not delaminate or otherwise damage the material
surrounding the hole. C-scan images of samples after the drilling process
showed minimal damage in areas and confirmed that samples were not
affected by the mechanical loading. Figure 39, provides optical microscopy

image of the cross-section area of the BMI modified CFRP, in which the
incorporated SHA is clearly illustrated.
Figure 37. Design of (a) the reference CFRP plate, (b) the modified CFRP plate
containing 120 gsm of SHA and c) a representative photograph of the area between -
450 and 00 prepreg layers of the modified CFRP plates, where the SHA has been placed
on.
Figure 38. (a) Schematic depiction of the modified open hole specimen configuration.
b) A snapshot during the test. Dimensions in mm.

Figure 39. Optical microscopy photograph of the cross section area of the BMI
prepolymer (BMI pp) modified CFRP, in which the incorporated SHA is clearly
illustrated.
6.2.1. Quasi-Static Tensile Tests

In this paragraph, initial quasi-static tensile tests of open holed reference
and BMI modified CFRPs are examined. For this purpose, stress (σ) vs.
strain % (ε) measurements were performed according to specifications
described in paragraph 2.1.5. These tests were carried out by using a batch
of open holed specimens manufactured from the same plates used in the
fatigue work that will be presented within the next paragraphs of the current
section. The resulting σ vs. ε curves are illustrated in Figure 40-a. For both
material sets the general trend does not present any dramatic difference; the
applied stress was increased linearly to failure. Figure 40-b compares the
critical quasi-static tensile characteristics σmax and E for the two material
sets. According to these results, it was shown that the tensile properties of
CFRPs were not affected by the incorporation of the BMI pp that plays the
role of the SHA. More precisely, BMI modified CFRPs exhibited slightly
decreased σmax value at the amount of 2.3% (from 661 MPa to 645 MPa)
while the apparent E value was retained at the same levels for both material
sets (45.4 GPa). This behaviour is attributed to the rigid and resin type nature
of the SHA that made the composites not to deteriorate their mechanical
properties. After failure both type of samples showed a pull-out type failure
mechanism which was caused by fibre failure within the 00 plies,
accompanied by delamination between some of the off-axis, one or more of
which failed via matrix cracking. The splitting failure of the off-axis plies

propagated from the hole and occurred roughly where the fibre direction was
tangential to the hole boundary.
Figure 40. (a) Representative stress (σ) vs. strain % (ε) curves during quasi-static
tensile testing of open holed reference and BMI modified CFRPs, (b) Comparison
between the reference and the BMI modified CFRPs in terms of the tensile strength
(σmax) and young's modulus (E) values, respectively.
6.2.2. Fatigue Testing

Tension-tension fatigue tests were performed for both material sets
according to specifications described in paragraph 2.1.6. The introduction of
the BMI pp as SHA into the CFRP architecture is expected to have an impact
on the materials' fatigue performance. At this part of work the fatigue
behaviour of reference and BMI modified CFRPs is investigated. More
precisely, fatigue tests were performed following three different routes; at
continuous loading conditions, with "stops" every 10 k cycles and with
"stops" together with the application of the healing cycles.
6.2.2.1. Tension-Tension Fatigue Behaviour of CFRPs at Continuous

Loading Conditions
According to fatigue experimental results for tests conducted at
continuous loading conditions, it was shown that BMI modified samples
exhibited slightly improved fatigue life by almost 8% (from 34300±6000 to
37000±5000 cycles respectively). This behaviour is believed to be attributed
to the increase of the fracture toughness properties (both mode I and II) that
the BMI causes to the entire composite according to extended work
presented in previous sections (3, 4 and 5). This behaviour made the

composites to withstand more fatigue cycles as slightly delayed the damage

evolution within the composite structure at interfaces where high stiffness
mismatch appears. Typical stress-strain hysteresis loops of the early
(n/Nf=0) and last loading cycles (n/Nf=1) of reference and BMI modified
CFRPs are presented in Figure 41-a and Figure 41-b respectively. In general,
both material sets present similar behaviours; the loops move towards higher
strains and the apparent modulus exhibit degradation due to damage
evolution, for constant stress levels applied for the needs of the current study
(0.8 UTS). As it is seen in both pictures the final hysteresis loops shows
significant apparent modulus degradation and the area of the loop constantly
increases as expected, since it is directly related to the fatigue damage.
The plots of the maximum (εmax) and minimum (εmin) strains with respect
to the life ratio for both material sets, are presented in Figure 42-a and Figure
42-b. These figures highlight a continuous increase of both strain values for
all material sets (reference and BMI modified ones). This behaviour is
supposed to be related to the hot creep effect due to the fact that the fatigue
loading consists of a positive mean stress superimposed with sinusoidal
variations [47]. Because both extreme strains present almost similar trends,
only the εmin, indicating the permanent deformation of the samples during
the test is of significance. The overall development of the εmin was calculated
to be 0.04% and 0.08% for reference and BMI modified CFRPs respectively.
One can observe that εmin evolves into an almost quasi-linear stage for the
whole life ratio up to failure.
Figure 41. Depiction of the first and the final hysteresis loops of (a) the reference and
(b) the BMI modified CFRPs at 0.8 of UTS.

Figure 42. Typical minimum (εmin) and maximum (εmax) cycle strains evolutions of (a)
the reference and (b) the BMI modified CFRPs at 0.8 of the UTS.
Figure 43-a and b present the normalized dynamic modulus evolution

(E/E0) for both material sets. The evolution of the specimens' dynamic
modulus is important factor as it gives valuable information on the material's
damage evolution. This modulus was calculated as the slope of the straight
line connecting the maximum and minimum stress tips of a hysteresis loop.
Figure 43. Modulus ratio evolution of (a) the reference and (b) the BMI modified
CFRPs at 0.8 of the UTS.
Figure 43-a and b show typical and global evolutions of the E/E0 with
respect to the specimens life ratio, where E and E0 are respectively the actual
cyclic and average modulus measured during the first cycles. As expected
and according to these figures, it was shown that the modulus in both cases
follows a common pattern and decreases in three stages; a first and third
stage with a relatively sharp decrease of the modulus and a second steady

stage (the middle one) between the other two where there modulus appears
a slight reduction. The maximum modulus decrease was calculated to be
close to 20% for both the reference and BMI modified CFRPs. This range
for the decrease of the dynamic modulus is similar to the general behaviour
of common CFRPs loaded under fatigue conditions according to other
investigations [48].
6.2.2.2. Tension-Tension Fatigue Tests with “Stops”

In this section, tension-tension fatigue experiments with interruptions
every 10 k cycles were performed according to specifications described in
paragraph 2.1.6, for both material sets. Results for reference and BMI
modified CFRPs are presented in Figure 44 and 45 respectively.
Figure 44. Tension-tension fatigue experiments with “stops” every 10 k cycles at 0.8 of
the UTS for reference CFRPs. (a) First and final hysteresis loops after each
consecutive “stop” during fatigue experiments, (b) typical maximum (εmax) and
minimum (εmin) cycle strains evolutions, (c) modulus ratio evolution, (d) C-scan
inspection images after each consecutive “stop” for the identification of the damage
area within the composite structure.

In this type of experiments, the damage evolution within the composite

was determined through C-scan inspections after each consecutive stop
applied to the fatigue loading of the samples. The white regions represent
the damage areas induced by the fatigue loading.
Typical stress-strain hysteresis loops of the early and last loading cycles
after each consecutive stop for reference CFRPs are illustrated in Figure 44-
a. In general, the trend does not significantly differ with that observed at
continuous loading conditions. The hysteresis loops move towards higher
strains for the constant stress levels applied to the specimens during the
fatigue loading. Also as it is seen, the hysteresis loops gradually moved very
far from the first cycle. The plots of the εmax and εmin strains with respect to
the life ratio are presented in Figure 44-b. The overall development of the
εmin was calculated to be 0.03% for reference CFRPs. One can observe that
εmin evolves into an almost quasi-linear stage for the whole life ratio up to
failure. Global evolution of the E/E0 with respect to the specimens life ratio
is illustrated in Figure 44-c. According to this Figure, it is seen that the
modulus decreases again in three stages; a first and third stage with a sharp
decrease of the modulus and a second steady stage between the other two
where the rate of decrease is much lower. The loss of the modulus was
calculated to be approximately 20%. Finally, typical C-scans for samples
prior and after each consecutive stop during the fatigue experiments are
depicted in Figure 44-d. The damage evolution within the composite
structure is prominent according to these images provided.
after each consecutive stop for BMI modified CFRPs are illustrated in Figure
45-a. In general, the trend does not significantly differ with that observed at
continuous loading conditions. The hysteresis loops move towards higher
strains for the constant stress levels applied to the specimens during the
fatigue loading. Also as it is seen, the hysteresis loops gradually moved very
far from the first cycle. The plots of the εmax and εmin strains with respect to
the life ratio are presented in Figure 45-b. The overall development of the
εmin was calculated to be 0.084% for BMI modified CFRPs. One can observe
that εmin evolves into an almost quasi-linear stage for the whole life ratio up
to failure. Global evolution of the E/E0 with respect to the specimens life

ratio is illustrated in Figure 45-c. According to this figure, it is seen that the
modulus decreases in three stages; a first and third stage with a sharp
decrease of the modulus and a second steady stage between the other two.
The loss of the modulus was calculated to be approximately 20%. Finally,
typical C-scans for samples prior and after each consecutive stop during the
fatigue experiments are depicted in Figure 44-d. The damage evolution
within the composite structure is prominent according to these images
provided.
the UTS for BMI modified CFRPs. (a) First and final hysteresis loops after each
consecutive “stop” during fatigue experiments, (b) typical maximum (εmax) and
minimum (εmin) cycle strains evolution, (c) modulus ratio evolution, (d) C-scan
inspection images after each consecutive “stop” for the identification of the damage
area within the composite structure.
To summarize, both material sets exhibited similar behaviours at both

continuous loading and loading with stop and go according to results
discussed above. The main difference was that both material sets during
fatigue tests with stops exhibited slightly decreased fatigue life, if compared
with them tested under continuous loading conditions (13%-18% reduction).

This is not a usually expected behaviour, where under interrupted fatigue

loading the life of the material normally extends. However, in the present
case one has to take into consideration the used high σmax which is at the
level of 0.8 of the UTS. This behaviour (scatter in results) which is clearly
visible is also attributed to the nature of fibre reinforced composites.
Composite materials are known to show scatter in their fatigue lives due to
statistical variations in the rate of damage development [49].
6.2.2.3. Tension-Tension Fatigue Tests with “Stops” and

Healing Activation
In the current section, tension-tension fatigue experiments with stops
every 10 k cycles together with the application of the healing cycles are
provided. These tests were performed only for samples having healing
functionality (BMI modified samples), according to specifications described
in paragraph 2.1.6. Results for this type of fatigue experiments are presented
in Figure 46. In this type of experiments, the damage evolution within the
composite was determined through C-scan inspections after each
consecutive stop. The damage recovery was also determined through C-scan
inspections prior and after the each healing cycle applied to the samples.
after each consecutive stop and healing activation are illustrated in Figure
46-a. In general, the trend does not significantly differ with that observed at
other type of fatigue tests described previously (at both continuous loading
conditions and with stops). All hysteresis loops captured move towards
higher strains for the constant stress levels applied to the specimens during
the fatigue loading. Also as it is seen, the hysteresis loops moved far from
the first cycle. In this case, the main difference is that the application of the
healing cycle after each consecutive stop retarded the damage evolution
within the composite structure as a result the fatigue life of the CFRP to be
extended by 75% (see and compare with Figure 45-a). This behaviour is also
confirmed by C-scan inspections presented in Figure 46-d, where is clearly
visible that part of initial damage area has been healed. The plots of the εmax
and εmin strains with respect to the life ratio are presented in Figure 46-b. The
overall development of the εmin was calculated to be 0.081% for these

mendable composites. One can observe that εmin evolves into an almost
quasi-linear stage for the whole life ratio up to failure. Global evolution of
the E/E0with respect to the specimens life ratio is illustrated in Figure 46-c.
According to this figure, it is seen that the modulus decreases in three stages;
a first and third stage with a sharp decrease of the modulus and a second
steady stage between the other two. The loss of the modulus was calculated
to be approximately 35%.
the UTS for BMI modified CFRPs, together with the application of the healing cycles.
(a) First and final hysteresis loops after each consecutive “stop” and healing cycle
during fatigue experiments, (b) typical maximum (εmax) and minimum (εmin) cycle
strain evolutions, (c) modulus ratio evolution, (d) C-scan inspection images after each
consecutive “stop” for the identification of the damage area within the composite
structure.
6.2.3. Assessment of Healing Functionality of Composites and

Extension of the Fatigue Life
In Figure 47-a, the global evolution of the E/E0 with respect to the
specimens life cycles (Nf) is illustrated for reference and BMI modified
samples, prior and after the healing activations. These curves represent
tension-tension fatigue tests with stops conducted for both material sets and

are compared. According to these curves, it is clearly shown the reduction

of the normalized modulus with the increase of the number of the fatigue
cycles for all material types with reference CFRPs to exhibit the higher one.
The incorporation of the SHA within the composite slightly enhanced the
dynamic modulus of the samples while the fatigue live was extended by
12%. On the other hand, the activation of the healing cycles made the
modified CFRPs to extend their fatigue life to a great extend up to 75%, if
compared with the same unhealed materials. In addition, healed samples
exhibited 100% extended fatigue life if compared with reference material
sets (see Figure 47-a).
Figure 47. (a) Modulus ratio evolution s (E/E0) for the reference and the BMI modified
CFRPs (with and without the application of the healing cycles) as a function of the
number of fatigue cycles, (b) depiction of the fatigue life of the reference and the BMI
modified CFRPs for all fatigue test types; at continuous loading conditions, with
"stops" every 10 k cycles and with "stops" every 10 k cycles together with the
application of the healing cycles.
7. LOW-VELOCITY IMPACT RESPONSE AND COMPRESSION

AFTER IMPACT ASSESSMENT OF CARBON FIBRE/EPOXY
COMPOSITES CONTAINING BIS-MALEIMIDE
BASED POLYMERS
7.1. Composites Manufacturing
Three types of quasi-isotropic laminated plates containing 16 plies, with

[45/0/-45/90]2S stacking sequence each were manufactured for the needs of

the current study; the reference laminate, the modified laminate with BMI
pp grains between the primary layers of the CFRP and the modified laminate
with two pure BMI prepregs. All material groups (reference and both
modified) were tested under compression prior of impact, in order to identify
possible knock down in compression strength. After exposed to LVI, CAI
tests were performed for the pristine CFRP plates as well as for the BMI
modified ones, but in this case before and after the application of the healing
activation cycle. Figure 48, shows schematically the configuration of the
plates (Figure 48-a) and the position where the BMI pp (Figure 48-b) and
pure BMI prepregs (Figure 48-c) were placed. The BMI pp grains were
positioned in between the 0/-45 plies while the BMI prepregs replaced the
6th and 11th UD 0° layers of the composite (Figure 48-c).
Following the lay-up, the laminates were vacuum bagged and cured in
an autoclave for 2 h at 130°C under 6 bars applied pressure, according to the
prepreg manufacturer's guidelines. The dimensions of the final plates were
150 mm × 100 mm × 2.1 mm. Ten reference and thirty modified impact test
samples (fifteen samples for each type of modified CFRPs) were
manufactured respectively. Initially, five samples of each material set were
directly tested to compression in order potential knock-down effect due to
the incorporation of the SHA to be determined. The other five samples of
pristine CFRPs and twenty samples of BMI modified (ten samples for each
type of modified CFRPs) CFRPs were exposed to impact tests. All the
samples were tested to CAI. In the case of BMI modified samples one group
was tested to CAI just after initial impact, while a second group was passed
through a healing cycle (following the earlier described cycle) and after this
step they were also tested to CAI.
The modified prepreg plies did not thicken the entire CFRP laminate
(i.e., 2.1±0.1 mm for all material sets); the Vf of both composite groups was
calculated to be about the same (approximately 60%). They also did not
disrupt the architecture. In Figure 49, optical microscopy cross-section
photographs of both modified CFRPs with SHAs based on the DA reaction
mechanism are illustrated. Both photographs describe the location in which
the SHA has been incorporated symmetrically for both BMI modified
CFRPs.

Figure 48. Design of (a) the reference CFRP, (b) the modified CFRP with BMI
prepolymer (BMI pp) and (c) the modified CFRP with pure BMI prepregs.
Figure 49. Optical microscopy photographs of the cross-section area of (a) the
reference CFRPs, (b) the modified CFRPs with BMI prepolymer (BMI pp) grains, (c)
the modified CFRPs with pure BMI prepregs. For both modified CFRPs, the SHA area
is clearly illustrated.
7.2.1. Resistance to Low-Velocity Impact (LVI)

Reference and both modified CFRPs were subjected to LVI tests, as
described in section 2.1.5. After testing, C-scan inspection measurements
were performed in order to evaluate the impact induced damage. The typical
C-scans for all material types in the before and after impact situation are
shown in Figure 50-a, b and c. The white region in the centre of the images
in Figure 50- a, b and c represents the damage area (mainly delamination)
induced by the impact event. For the determination of the impact damage
area, a post-processing software was utilized. The software was able to
quantify the white region area (damage area) of the recorded C-scan images
after the LVI event for both material sets. These areas were calculated to be

approximately 592, 1015 and 950 mm2 for the reference, the BMI pp
modified and the pure BMI prepreg modified CFRPs respectively.
Figure 50. Representative C-scan inspection images of (a) the reference CFRP, (b) the
modified CFRP with BMI prepolymer (BMI pp) grains, (c) the modified CFRP with
pure BMI prepregs, in the before and after the LVI situation. (d) Indicative top- and
bottom-face photographs after LVI for the modified composites.
Thus, the reference composite exhibits higher resistance to

delamination, if compared with that observed in both BMI modified
composite cases. Thus, the incorporation of the SHA into composites
architecture decreases the LVI-induced damage resistance of the final
material. In addition, samples containing BMI pp grains exhibited slightly
higher damage area (10% higher) when compared with the damage are
exhibited the pure BMI modified samples. It is believed, that in both
modified composite cases BMI caused softening to the entire composite. In
the current work impact energy of 25 J was utilized. It is considered a

moderate amount of impact energy that does not occurs fibre breakage and
the main damage mode that is promoted is delamination. In fact, no visible
or detectable damage was shown on the top and bottom faces of the reference
samples after LVI testing. On the contrary, optical inspection of all modified
samples after LVI tests revealed that indentation occurred on the top face
(Figure 50-d-i), while ply splitting and fibre fracture on the bottom face is
prominent (Figure 50-d-ii).
In general, during an impact event with moderate impact energy the
impact energy introduced into the composite structure is mainly absorbed
through elastic deformation and through different failure modes, while
plastic deformation does not take place [50]. Samples containing SHAs with
a ductile nature such as BMI, plastic deformation takes place. On the other
hand BMI modified samples was expected to exhibit not only a common
CFRP failure mode but also a plastic deformation due to the presence of the
ductile BMI reversible polymer. According to Figure 51-a and b, the
formation of multiple delaminations sites in the reference sample is
apparent. On the contrary, modified CFRPs exhibited different behaviour
after LVI. Extended damage has been occurred into the composite in the
form of delaminations and significant transverse cracks that lead to off axis
ply split (see Figure 53-b) due to the softening effect that both SHAs caused
to the entire composite.
Figure 51. Illustration of the reference CFRP cross section area, just after the low
velocity impact (LVI) test. (a) Optical microscopy and (b) SEM image (zoom in the
marked region).

7.2.2. Repair of the Impact Induced Damages via the

Healing Treatment
After LVI tests, one group of both BMI modified samples (containing
BMI pp or pure BMI prepregs) were subjected to heating under controlled
through-the-thickness compression, as described in paragraph 2.1.10. After
healing, C-scan was performed in order to evaluate the reduction of the
impact induced damage in the areas where the SHA have been placed.
Interestingly, it was found that after healing the modified samples exhibited
macroscopically a H.E. of about 55% and 100% for the BMI pp and BMI
prepreg modified samples respectively (accounting by the measurement of
the impact damage area before and after healing, as it demonstrated globally
in the C-Scan plots (Figure 52-a, b).
Figure 52. Representative C-scan inspection images of the modified CFRP plates (a)
with BMI prepolymer (BMI pp), (b) with two pure BMI prepregs, both before impact,
after impact and after the healing process.
The H.E. values were calculated according to Equation (6). According

to C-scan inspections provided in Figure 52, it was shown that samples

containing BMI prepregs were able to heal the entire damage within the
composite structure after the activation of healing cycle. On the other hand,
modified samples containing BMI pp were able to heal only the 55% of the
initial induced damage. This behaviour was also validated by optical
microscopy examination of the damaged samples after the healing process.
Materials micro-structure examination under optical microscopy was used
to validate the results from non-destructive C-scan inspections. In Figure 53
the optical microscopy photos of the cross-section areas of the repaired
modified CFRPs are illustrated. Figure 53 zooms in the SHA area and the
upper and lower adjacent layers of the laminate. For the BMI pp modified
samples (Figure 53-a) the presence of many delaminations after the healing
process is still apparent, as it is expected in all the interfaces that do not
contain BMI pp material. Only areas adjacent to BMI were healed and this
is the reason why low H.E. value was achieved. On the contrary, BMI
prepreg modified samples after healing were able to heal the entire damage
as cracks disappeared though heating during the healing process due to the
nature of this type of SHA.
Figure 53. Optical microscopy photographs for both BMI modified CFRPs. Cross
section areas after the healing process: (a) modified CFRP with BMI prepolymer (BMI
pp) grains (b) modified CFRP with two pure BMI prepregs.

7.2.3. Compressive Behaviour of CFRPs before Impact

The incorporation of both BMI based SHAs via both ways into
composite architecture is expected to have an impact (i.e., knock-down
effect) on final composites’ mechanical performance. In the current
subsection, the effect on the compressive properties is assessed. Typical
stress vs. strain curves under CBI for the reference and both modified CFRPs
in their initial (unimpacted) situation are illustrated in Figure 54-a. In all
curves, the initially linear stress vs. strain response is followed by a visual
deviation from linearity and load drop just before the end of the test. The
compressive strength (σmax) and the effective compressive modulus (Ecomp)
values of the two material types are summarized in Figure 54-b. Both BMI
modified CFPRs exhibited lower compression characteristics than the
reference ones. The σmax was decreased by 20% and 10% for the BMI pp
and the pure BMI prepreg modified CFRPs respectively. On the other hand,
modified samples with BMI pp exhibited decreased Ecomp by 20% while the
the Ecomp for the pure BMI prepreg modified samples was retained at the
same levels. Thus, both types of BMI SHAs deteriorate the in-plane
(compressive) properties of the final CFRP.
Figure 54. (a) Representative compressive stress vs. compressive strain (%) curves
from the CBI testing of the reference and both BMI modified CFRPs. (b) Compressive
strength (σmax) and compressive modulus (Ecomp) of the three material types (average
values and standard deviations (SSD)).
7.2.4. Residual Compressive Properties of the CFRPs after LVI

As already described above, after LVI tests reference and both BMI
modified samples were loaded in compression to assess their CAI properties.

Also, the introduced healing functionality of the BMI modified samples was
quantified by subjecting these samples to a simple healing cycle as described
in paragraph 2.1.10 and then, repeating the CAI tests in order to investigate
the post-healing CAI performance of these modified plates.
Typical compressive stress vs. compressive strain (%) curves after LVI
for reference and both BMI modified samples (prior and after healing) are
illustrated in Figure 55-a and b. In addition, σmax and Ecomp values of the
reference and both modified CFRPs in prior impact, after impact and after
healing situation are summarized in Figure 55-c. According to these results,
it is shown that after LVI the σmax of the reference and both BMI modified
samples was reduced by 16%, 5% (prepolymer) and 21% (prepreg). The
reduction of the apparent Ecomp was calculated to be 5% for the reference
while 24% and 10% for the BMI pp and BMI prepreg modified CFRPs
respectively. Furthermore, both BMI modified samples passed through
healing process as it is has been already described earlier, and after that they
exposed to CAI tests in order the post-healing performance of the composite
plates to be investigated. According to these tests it was shown that healed
samples presented higher σmax and Ecomp values of approximately 4% and
15% (prepolymer) and 4% and 2% (prepreg) respectively compared against
the unhealed BMI modified ones (impacted samples). Therefore, modified
samples improved CAI properties via healing treatment with BMI prepreg
modified samples to exhibit comparable Ecomp with the reference impacted
one (see Figure 55-b and C).
As previously mentioned, the BMI pp SHA was not able to heal all the
cracks within the composite structure. In addition, during impact testing both
modified samples exhibited ply splitting and fibre breakage (that cannot be
healed). After healing process samples containing both types of SHAs
showed better load transfer characteristics. Finally, it is of note that whereas
reference and BMI pp modified composites fail in a brittle manner BMI
prepreg modified CFRPs fail in a ductile or less brittle manner (both the
virgin and healed ones). It is suggested that this result is a reflection of the
ductile nature of the pure BMI and of the softening effect induced by the
presence of this type of plies as SHA into composites' architecture.

Figure 55. Representative compressive stress vs. compressive strain (%) curves from
the compression after impact (CAI) testing of (a) both the reference and the BMI
prepolymer (BMI pp) modified CFRPs and (b) both the reference and the BMI prepreg
modified CFRPs (prior and after healing), (c) Compressive strength (σmax) and
compressive modulus (Ecomp) of both the reference and both modified CFRPs prior
impact, after impact and after healing.
8. OUTLOOK
Polymers based on DA polymers are very promising for future

applications due to their healing functionality. Future work must focus on
making these materials to withstand higher loading conditions. Prepregs
containing these reversible polymers seems to be very promising way for
composites during the manufacturing process due to the easier
incorporation. In addition the increase of the BMI polymer (DA reaction)
into the BMI pp resin (i.e., 25%, 30%) may be beneficial for FRPs in terms
of toughening performance and healing capability. Self-healing technology
based on reversible polymers can be placed in known critical regions of
structures where the damage predominately occurs, such as around drilled
holes or on skin-stringer run-outs. These regions will take advantage of these
repair mechanisms and do not represent the overall host matrix mechanical
performance due to the knock down effect of these polymers to the in-plane

mechanical properties of the whole composite structure. The scope of all

these modified systems is considered for future implementation as parts in
aerospace structures.
CONCLUSION
In the current chapter, new healable carbon/epoxy composites were

developed by incorporating SHAs that are based on DA reaction mechanism
in grains and in prepreg form, locally. The effect of these SHAs on the: (a)
mode I interlaminar fracture toughness, (b) mode II interlaminar fracture
toughness, (c) tensile, (d) fatigue, (e) LVI and (f) CAI of the CFRPs, and the
provided healing capability was extensively studied. BMI polymer in
interlayer or prepreg form were locally incorporated into the area that the
cracks are expected to propagate. BMI prepregs were fabricated in order the
incorporation of the SHA into the composite to be simplified. According to
experimental results, it was shown that in both cases BMI SHAs were able
to considerably enhance the mode I and mode II interlaminar fracture
toughness of the composites laminates through extended fibre bridging.
Characteristic example is that of GIC value (for samples containing pure BMI
prepregs) which was significantly increased by 330%. These modified
samples, also exhibited moderate recoveries after first fracture for their
mode I interlaminar fracture toughness properties with increasing trend after
curing these samples at lower temperatures. High recoveries after mode II
interlaminar fracture tests were achieved due to the nature of the precise
experiment. Potential knock down effects on the in-plane properties of these
modified CFRPs were also assessed. 3PB experiments revealed that the
incorporation of the BMI SHAs into composites' architecture made the in-
plane properties to slightly decrease. Fatigue experiments showed that by
the incorporation of the BMI SHA the fatigue life was improved by 8%
while extended by 75% after the activation of the healing cycles. In addition,
samples containing BMIs were examined under CBI, LVI and CAI tests
while compared with the benchmark ones. Initially, it was shown that by the
incorporation of these SHAs into composites architecture the compression

properties were degraded. After LVI tests, samples containing BMIs

exhibited lower resistance to damage when compared with the reference
ones. Healing process revealed that BMI pp SHA was not able to fully
restore (even locally) the damage into the composite while samples
containing BMI prepregs fully restored the entire damage. Finally, both BMI
modified samples exhibited slightly improved compression characteristics
after the activation of the healing cycle. An analogous investigation that is
related to the healing performance of modified CFRPs containing
dendrimers as SHA has been recently published by the authors in [51-52].
ACKNOWLEDGMENTS
The present work has been funded by: (a) the EU FP7 Transport
(including Aeronautics) Programme within the frame of the project: Self-
healing polymers for concepts on self-repaired aeronautical composites
HIPOCRATES (ACP3-GA-2013-605412). (b) Clean Sky 2 Programme
within the frame of the project: Development of innovative and ECO-
friendly airframe TECHnologies from design to manufacturing to improve
aircraft life cycle environmental footprint ecoTECH, Grant Agreement
number: 807083 — GAM AIR 2018 — H2020-IBA-CS2-GAMS-2017.
Also, the authors thank Hartmut Fisher (TNO) and Aggelos Christopoulos
(NTUA) for the material supply.
REFERENCES
[1] Ashby, M. F. (1987). Technology in the 1990s: Advanced materials

and predictive design. Philosophical Transactions of the Royal Society
of London, 332: 393-407.
[2] O’Brian, Τ. K. (1990). Towards a damage tolerance philosophy for
composite materials and structures. Composite Materials: Testing and
Design, ASTM special technical publication, 1059:7-33.

[3] Wisnom, M. R., (2012). The role of delamination in failure of fibre-

reinforced composites. Philosophical Transactions of the Royal
Society A, 370:1850-1870. https://doi.org/10.1098/rsta.2011.0441.
[4] Blaiszik, B. J., Kramer, S. L. B., Olugebefola, S. C., Moore, J. S.,
Sottos, N. R. and White, S. R. (2010). Self-healing polymers and
composites. Annual Review of Materials Research, 40:179-211.
[5] Mezzenga, R., Boogh, L. and Manson, J. (2001). A review of dentritic
hyperbranched polymer as modifiers in epoxy composites. Composites
Science and Technology, 61:787-795.
[6] Masters, J. E. (1989) Improved impact and delamination resistance
through interleaving. Key Engineering Materials, 37:317-348.
[7] Stevanovic, M. M. and Stecenko, T. B. (1992). Mechanical behaviour
of carbon and glass hybrid fibre reinforced polyester composites.
Journal of Materials Science, 27:941-946.
[8] Dransfield, K. A., Jain, L. K. and Mai, Y. W. (1998). On the effects of
stitching in CFRPs–I. mode I delamination toughness. Composites
[9] Gilbert, E. N., Hayes, B. S. and Seferis, J. C. (2003). Interlayer
toughened unidirectional carbon pre-preg systems: effect of preformed
particle morphology. Composites Part A: Applied Science and
Manufacturing, 34:245-252.
[10] Mouritz, A. P. (2007) Review of z-pinned composite laminates.
Composites Part A: Applied Science and Manufacturing, 38:2383-
2397.
[11] Chan, W. C. (1991). Design approaches for edge delamination
resistance in laminated composites. Journal of Composites,
Technology and Research, 13:91-96.
[12] Chan, W. C. and Ochoa, O. O. (1989). Edge delamination resistance
by a critical ply termination. Key Engineering Materials, 37:285-304.
[13] Choi, J., Yee, A. F. and Laine, R. M. (2004). Toughening of cubic
silsesquioxane epoxy nanocomposites using core-shell rubber
particles: A three-component hybrid system. Macromolecules,
37:3267-3327.

[14] Howard, W. E., Gossard, J. T.and Jones, R. M. (1989). Composite

Laminate Free-Edge Reinforcement with U-Shaped Caps Part II:
Theoretical-Experimental Correlation. AIAA Journal, 27:617-623.
[15] Whittingham, B., Baker, A. A., Harman, A. and Bitton, D. (2009).
Micrographic studies on adhesively bonded scarf repairs to thick
composite aircraft structure. Composites Part A: Applied Science and
[16] Baker, A., Dutton, S. and Kelly, D. (2004). Composites materials for
aircraft structures. 2nd ed. American Institute of Aeronautics and
Astronautics (AIAA) Publication Series, 601 p.
[17] Bekas, D.G., Tsirka, K., Baltzis and D., Paipetis, A. S. (2016). Self-
healing materials: A review of advances in materials, evaluation,
characterization and monitoring techniques. Composites Part B:
Engineering, 87:92-119.
[18] Diesendruck, C. E., Sottos, N. R., Moore, J. S. and White, S. R. (2015).
Biomimetic self-healing. Angewandte Reviews, 127:10572-10593.
[19] Williams, G., Trask, R. and Bond, I. (2007). A self-healing carbon
fibre reinforced polymer for aerospace applications. Composites Part
A: Applied Science and Manufacturing, 38:1525-1532.
[20] White, S. R., Moore, J. S., Sottos, N. R., Krull, B. P., Santa Cruz, W.
A. and Gergely, R. C. R. (2014). Restoration of large damage volumes
in polymers. Science, 344:620-623.
[21] Varley, R. J., Craze, D. A., Mouritz, A. P. and Wang, C. H. (2013).
Thermoplastic healing in epoxy networks: Exploring performance and
mechanism of alternative healing agents. Macromolecules Materials
and Engineering, 298:1232-1242.
[22] van Gemert, G. M. L., Peeters, J.W., Sontjens, S. H. M., Janssen, H.
M. and Bosman, A. W. (2012). Self-healing supramolecular polymers
in action. Macromolecules Chemistry and Physics, 213:234-242.
[23] Bai, N., Saito, K. and Simon, G. P. (2013). Synthesis of a diamine
cross-linker containing Diels-Alder adducts to produce self-healing
thermosetting epoxy polymer from a widely used epoxy monomer.
Polymer Chemistry, 4:724-730.

[24] Kennedy, J. P. and Castner, K. F. (1979). Thermally reversible

polymer systems by cyclopentadienylation. I. A model for termination
by cyclopentadienylation of olefin polymerization. Journal of Polymer
Science Part A: Polymer Chemistry, 17:2039-2054.
[25] Chen, X., Dam, M. A., Ono, K., Mal, A., Shen, H., Nutt, S.R., Sheran,
K. and Wudl, F. (2002). A thermally remendable cross-linked
polymeric material. Science, 295:1698-1702.
[26] Chen, X., Wudl, F., Mal, A. K., Shen, H. and Nutt, S. R. (2003). New
thermally remendable highly cross-linked polymeric materials.
Macromolecules, 36:1802-1807.
[27] Aubert, J. H. (2003). Thermally removable epoxy adhesives
incorporating thermally reversible diels-alder adducts. Journal of the
Adhesives, 79:609-616.
[28] Ghezzo, F., Smith, D. R., Starr, T. N., Perram, T., Starr, A. F.,
Darlington, T. K., Baldwin, R. K. and Oldenburg, S. J. (2010).
Development and characterization of healable carbon fiber composites
with a reversibly cross linked polymer. Journal of Composite
Materials, 44:1587-1603.
[29] Pratama, P. A., Peterson, A. M. and Palmese, G. R. (2013). The role
of maleimide structure in the healing of furan-functionalized epoxy-
amine thermosets. Polymer Chemistry, 4:5000-5006.
[31] Park, J. S., Darlington, T. A., Starr, F., Takahashi, K., Riendeau and J.
Hahn, H. T. (2010). Multiple healing effect of thermally activated self-
healing composites based on Diels-Alder reaction. Composites Science
and Technology, 70:2154-2159.
[32] Park, J. S., Takahashi, K., Guo, Z., Wang, Y., Bolanos, E., Hamann-
Schaffner, C., Murphy, E., Wudl, F. and Hahn, H. T. (2008). Towards
development of a self-healing composite using a mendable polymer
and resistive heating. Journal of Composite Materials, 42:2869-2881.

[33] Kostopoulos, V., Kotrotsos, A., Tsantzalis, S., Tsokanas, P.,

Christopoulos and A. C., Loutas, T. (2016). Toughening and healing
of continuous fibre reinforced composites with bis-maleimide based
pre-pregs. Smart Materials and Structures, 25:84011 (12 pp).
[34] Bekas, D. G., Baltzis, D., Tsirka, K., Exarchos, D., Matikas, T.,
Meristoudi, A., Pispas, S. and Paipetis, A. S. (2016). Self-healing
polymers: evaluation of self-healing process via non-destructive
techniques. Plastics, Rubber and Composites, 45:147-156.
[35] Turkenburg, D. H. and Fischer, H. R. (2015). Diels-Alder based,
thermo-reversible cross-linked epoxies for use in self-healing
composites. Polymer, 79:187-194.
[36] Coope, T. S., Turkenburg, D. H., Fischer, H. R., Luterbacher, R., van
Bracht, H. and Bond, I. P. (2016). Novel Diels-Alder based self-
healing epoxies for aerospace composites. Smart Materials and
Structures, 25:084010-084019.
[37] Turkenburg, D. H. and Fischer, H. R. (2016). Self-healing mechanism
based on dispersed solid particles of various monomeric
bismaleimides. Materials Today Communications, 8:148-155.
[38] Kostopoulos, V., Kotrotsos, A., Tsantzalis, S., Tsokanas, P., Loutas,
T. and Bosman, A. W. (2016). Toughening and healing of continuous
fibre reinforced composites by supramolecular polymers. Composites
[39] Kostopoulos, V., Kotrotsos, A., Baltopoulos, A., Tsantzalis, S.,
Tsokanas, P., Loutas, T. and Bosman, A. W. (2016). Mode II fracture
toughening and healing of composites using supramolecular polymer
interlayers. Express Polymer Letters, 10: 914-926.
[40] Wang, C. H., Sidhu, K., Yang, T., Zhang, J. and Shanks, R. (2012).
Interlayer self-healing and toughening of carbon fibre/epoxy
composites using copolymer films. Composites Part A: Applied
Science and Manufacturing, 43:512-518.
[41] Pratama, P. A., Peterson, A. M. and Palmese, G. R. (2013). The role
of maleimide structure in the healing of furan-functionalized epoxy-
amine thermosets. Polymer Chemistry, 4:5000-5006.


[43] Pingkarawat, K., Wang, C. H., Varley, R. J. and Mouritz, A. P. (2014).
Mechanical properties of mendable composites containing self-
healing thermoplastic agents. Composites Part A: Applied Science and
[44] Aksoy, A. and Carlsson, L. A. (1992). Interlaminar shear fracture of
interleaved graphite/epoxy composites. Composites Science and
Technology, 43:55-69.
[45] Groleau, M. R., Shi, Y. B., Yee, A. F., Bertram, J. L., Sue, H. J. and
Yang, P. C. (1996). Mode II fracture of composites interlayered with
nylon particles, Composites. Science and Technology, 56:1223-1240.
[46] Kostopoulos, V., Kotrotsos, A., Tsokanas, P. and Tsantzalis, S. (2018).
Toughening and healing of composites by CNTs reinforced copolymer
nylon micro-particles. Materials Research Express, 5:025305-025332.
[47] Petermann, J. and Schulte, K. (2002). The effects of creep and fatigue
stress ratio on the long-term behaviour of angle-ply CFRP. Composites
Sructures, 57:205-210.
[48] Case, S. W. and Reifsnider, K. L. (2008). Fatigue of composite
materials. Comprehensive structures integrity, Chapter 4:405-441.
[49] Arnold N. T. and Martin P. A. (2008). Fatigue of sisal fibre reinforced
composites: Constant-life diagrams and hysteresis loop capture.
Composites Science and Technology, 68:915-924.
[50] Shen, G. A., Namilae, S. and Chandra, N. (2006). Load transfers issues
in the tensile and compressive behaviour of multiwall carbon
nanotubes. Material Science and Engineering A, 429:66-73.
[51] Kostopoulos, V. and Kotrotsos, A. (2018). Self-healing of Structural
Composites Containing Dendrimers as Healing Agent, Dendrimers-
Fundamentals and Applications. INTECH open science, Chapter 6,
101-132.
[52] Kotrotsos, A. (2017). Development of self-healing techniques for fiber
reinforced composites, PhD Thesis, University of Patras, May 2017.

Chapter 2
CONTROLLING STRUCTURAL AND

ELECTROCHEMICAL PROPERTIES OF
CNF THROUGH CALCINATION OF
PAN NANOFIBERS
Reyna Ojeda-López1, Isaac J. Pérez-Hermosillo1,

Juan M. Esparza-Schulz1,
Guadalupe Ramos-Sánchez1,2 and
Armando Domínguez-Ortiz1,*
1
Departamento de Química,
Universidad Autónoma Metropolitana, Mexico
2
CONACyT fellow
ABSTRACT
In recent decades the use of carbon nanofibers (CNF) has increased

notably as a result of the development of techniques that allow the
manipulation of their chemical and textural properties. This has allowed
the expansion of the fields of application, from components for batteries
*
Corresponding Author Email: doar@xanum.uam.mx.

80 Reyna Ojeda-López, Isaac J. Pérez-Hermosillo et al.
and fuel cells to vapor adsorption and catalysts. In this context, CNF have
been developed to satisfy diverse industrial requirements, taking advantage
of some of their important properties; such as high surface area, high
strength to weight ratio, excellent chemical resistance, superior electrical
conductivity and good thermal conductivity.
In this work, polyacrylonitrile nanofibers (PANNF) are first obtained
from polyacrylonitrile (PAN) in dimethylformamide (DMF) by
electrospinning. Then, the CNF are obtained by calcination of PANNF.
This process is comprised of two steps: i) stabilization at temperatures
between 523 to 573 K in an oxidizing atmosphere (air), and ii)
carbonization at temperatures above 873 K in an inert atmosphere
(nitrogen). These steps control both, microporosity and amount (and type)
of nitrogen species. The study is then divided in two stages, the first
consists of varying the temperature in step one while maintaining the
temperature in step two constant, and vice versa for stage two. The
consequences of four stabilization temperatures (523, 543, 553 and 573 K)
on these properties are evaluated while maintaining the carbonization
temperature constant at 1173 K; the most adequate stabilization
temperature to keep the fibrillary structure is 553 K. To study the second
stage, different portions of material stabilized at 553 K are carbonized at
873, 973, 1073, 1173 and 1273 K. Finally, the capacity of these materials
to promote oxygen reduction reaction (ORR) is evaluated.
1. INTRODUCTION
Carbon nanofibers have been inadvertently produced from natural
cellulosic fibers, such as cotton or linen, for thousands of years. However, it
was Thomas Alva Edison who, in 1879, intentionally took fibers from cotton
and bamboo, and turned them into carbon, in his search of filaments for
incandescent lamps (Chand, 2000; Zhang et al., 2014; Park, 2015). In the
last years, the study and application of carbon nanofibers have increased
considerably, due to their enhanced physical and chemical properties such
as: large specific surface area, high electronic conductivity and remarkable
mechanical resistance and thermal resistance (Zhang et al., 2014; Park,
2015).
CNF are filamentous materials composed mainly of hexagonal networks
of carbon atoms with sp2 hybridization (structure similar to graphite), in
which the carbon atoms are grouped into fine structures with diameters from
a few tens of nanometers to ~ 200 nm and lengths in the order of microns,

Controlling Structural and Electrochemical Properties of CNF … 81
having a graphite planar spacing of 0.335 – 0.342 nm (Rodriguez, 1993;

Park, 2015). CNF contain at least 90 wt.% of carbon, derived from the
controlled pyrolysis of the appropriate precursors. The principal features that
the precursors must possess are the following: easy conversion to carbon
nanofibers, high carbon content and cost-effective processing. From this
perspective, the most commonly used precursors are: i) acrylic, ii) cellulosic
and, iii) pitch-based (Park, 2015).
The carbon sheets in CNF are arranged to form a determinate angle with
respect to the nanofibers length axis, which make up the classification into
3 types: i) platelet, ii) fishbone, and iii) parallel (Rodriguez, 1993). The type
of arrangement of the carbon sheets is mainly a consequence of the synthesis
method; depending on the application of the CNF, this arrangement can be
modified leading to a customized functionality. For instance, Zheng et al.
synthesized carbon nanofibers with controlled microstructures, i.e., platelet,
fish-bone and tubular CNF; they also investigated their activity towards the
electrocatalytic oxygen reduction reaction (ORR) in acid media, concluding
that the parallel CNF has the highest ORR peak current due to the higher
density of catalytic activity sites (Zheng et al., 2008).
Generally to fabricate CNF the most used methodologies are: laser
ablation, chemical vapor deposition (CVD), plasma with radio frequency,
synthesis by flame, floating catalysis and electrospinning (Park, 2015).
Despite the preparation methods for CNF it is recognized that they are
complicated and overpriced; on the other hand, the electrospinning process
is a simple and low-cost method to obtain CNF from polymeric solutions.
This process consists in the use of polymer solutions subjected to a high
voltage difference to force the generation of nanofibers. PAN is the most
used precursor for obtaining carbon nanofibers, mainly due to its high
thermal deformation resistance. PAN is a polymer formed by large carbon
nitrogen and hydrogen chains, in the form of ladder, with highly polar nitrile
groups and a vitreous transition temperature of 358 K (Minagawa et al.,
2002). Due to the strong interactions between the chains, the PAN can only
be dissolved in strong polar solvents, the most used are DMF
(dimethylformamide) and DMSO (dimethyl sulfoxide) (Iovleva et al., 2001;
Wu et al., 2012b; Eom and Kim, 2014).

This chapter analyzes the preparation of CNF via the calcination of

PANNF synthesized by electrospinning. In section 1, the necessary concepts
to a comprehensive development of this work are presented, i.e.,
electrospinning process, chemical reactions in the calcination process, and
microporosity. In section 2, the characterization techniques are described. In
section 3, the calcination stages are analyzed. In subsection 3.1, it is
analyzed the impact of the stabilization temperature on the structural
properties and the fibrillary shape of the CNF. In subsection 3.2, it is
evaluated the effect of the carbonization temperature on the structural
properties, degree of graphitization, carbon and nitrogen content of CNF.
Finally, in subsection 3.3, these materials are evaluated as possible catalysts
for the oxygen reduction reactions in fuel cells.
1.1. Electrospinning
The electrospinning method was patented in 1900 by J.F. Cooley (Patent

GB 06385 “Improved methods of and apparatus for electrically separating
the relatively volatile liquid component from the component of relatively
fixed substances of composite fluids” 19th May 1900). From then until the
present, the initial prototype has undergone several improvements, making
it a very accessible and economical process. The assembly of an
electrospinning equipment requires a high voltage power source that
provides the electrical charge, a syringe with a metal needle (without
beveling) for the polymer solution, a metal collector (aluminum, copper, and
others) to deposit the nanofibers and an infusion pump to control the flow
velocity between the tip of the needle and the collector.
The electrospinning technique involves applying a high voltage
difference between the tip of the syringe and the metal collector. The
polymer solution at the tip of the needle is subjected to capillary and
electrostatic forces. During the electro-hydrodynamic flow, the charges are
induced in the fluid by the electric field strength between the two electrodes
(syringe and collector). When the capillary forces are equal to the electric
force, the so-called Taylor cone is formed; the electric field reaches a critical

value and overcomes the force of the surface tension of the polymer solution,
then the drop begins to elongate resulting in the formation of a jet. The
formation of the jet consists of two stages: i) stable stage, it occurs
immediately to the formation of the Taylor cone, due to a stretching process
and by means of high resolution devices it is possible to observe a single
thread, and ii) unstable stage, known as a whipping region, in this stage the
jet is destabilized because the solution begins to evaporate and the thread
formed in the stable stage is divided into finer threads that fall in the collector
in a disorderly manner. As the jet accelerates, i.e., the polymer solution is
expelled towards the collector (Huang et al., 2003; Wang and Kumar, 2006;
Mottaghitalab and Haghi, 2011), it occurs a reduction in the diameter of jet
due to evaporation of the solvent, and once in the collector, the jet solidifies
creating nanometric fibers. (Subbiah et al., 2005).
The working variables in the electrospinning process are important to
understand how the conversion of polymeric solutions into micrometric
fibers occurs. The morphology and size of the nanofibers in electrospinning
process are governed by three types of variables (Shin et al., 2001; Theron
et al., 2004; Jalili et al., 2006; Wang et al., 2007; He et al., 2008; Yördem et
al., 2008; Beachley and Wen, 2009; Yu et al., 2010):
1. solution variables:
(a) precursor concentration
(b) electrical conductivity of the solution
(c) molecular mass
(d) surface tension
(e) solvent volatility
(f) viscosity
2. process variables:
(a) distance between the tip of the needle and metal collector
(b) volumetric flow velocity
(c) voltage
3. environmental variables
(a) humidity
(b) temperature

The most important solution variable is the precursor concentration; this

variable affects the viscosity and surface tension of the system. The surface
tension tends to deform the jet in drops, while the viscous forces give
cohesion to the jet, causing the solidification before it breaks into drops,
resulting the formation of nanofibers (Gomes et al., 2007; He et al., 2008).
As a general rule, the diameter of nanofibers increases with the solution
concentration, i.e., with the increase in viscosity. The electrical conductivity
of the solution allows the availability of charged ions in the solution,
required to make a solution useful for electrospinning (Panapoy et al., 2008).
The volatility of solvent determines the time of flight necessary for
evaporate the solvent in the path between the needle tip and the collector,
solidifying the jet in form of a nanofiber. When the solvent is very volatile
it can generate the solidification of Taylor cone, covering the needle and
stopping the process (Ramakrishna et al., 2005).
For the process variables, the distance between needle tip and the metal
collector affects the nanofibers structure and morphology, due to its effect
on the flight time, altering the whipped and evaporation times. When the
distance is too short (less than 8 cm), the nanofibers do not have enough time
to solidify and the solvent evaporation is incomplete (Yördem et al., 2008).
The volumetric flow velocity of solution from the needle towards the current
collector affects the amount and quality of nanofibers obtained in the
electrospinning process. If the speed is fast, the efficiency of the process
decreases, since not all the material is electrospun, most of it falls before
reaching the collector and the one that arrives has not completely solidified,
i.e., the solvent has not been totally evaporated; on the contrary, very slow
speeds cause the solidification of the solution before reaching the tip of the
syringe (Wang and Kumar, 2006). The voltage is responsible for the
electrical charge of the fluid, in the electrospinning process a minimum
voltage value is necessary to start the ejection of jet from the Taylor cone,
this value depends on the polymeric solution used. High values, increase the
strength of electric field reducing the flight time of jet, generating an
incomplete evaporation of the solvent (Guo et al., 2013).
Finally, the environmental variables as humidity and temperature, also
affect the morphology and the size of the nanofibers. Higher temperatures

favor the reduction of nanofibers diameter, generating nanofibers of PAN

with diameters lower than 100 nm (Wang et al., 2007). With respect to
humidity, a low humidity value helps a better evaporation of solvent, while
high humidity values lead to the formation of thicker nanofibers (De Vrieze
et al., 2009).
1.2. Calcination Process
1.2.1. Stabilization in Air Atmosphere

This process involves chemical reactions such as: cyclization,
dehydrogenation, aromatization, oxidation, and others; resulting in the
formation of a structure in form of a ladder. In general, the linear molecules
of polymer precursors (PAN) are first converted into cyclic structures. The
reaction mechanism proposed in the literature is presented in Figure 1
(Rahaman et al., 2007). The PAN nanofibers stabilization is carried out at
temperatures between 523 and 573 K, in air atmosphere. The air in the
ambient produces the oxidation and dehydrogenation of polymeric
nanofibers, possibly due to the activation of cyclization centers by oxygen
and the promotion of the dehydrogenation process. During the cyclization
process, the nitrile groups transform their triple bond structure to a double
bond structure (C ≡ N → C = N), producing pyridine rings with six carbons
(Dalton et al., 1999; Wu et al., 2012a; Xue et al., 2013).
1.2.2. Carbonization in Nitrogen Atmosphere

To obtain CNF, the stabilized PAN nanofibers are subjected to a heating
process at temperatures above 873 K; this process is known as carbonization.
The carbonization of stabilized nanofibers is carried out in an inert
atmosphere (N2). In this step, dehydrogenation and denitrogenation take
place, where adjacent polymer chains are bound by C-C bonds due to the
aromatization process, resulting in broader carbon sheets, as shown in Figure
2 (Ko, 1991; Liu et al., 1994; Wangxi et al., 2003; Rahaman et al., 2007).
Temperatures higher than 3000 K allow obtaining pure carbon structure.

Figure 1 and 2 represent the stages of an ideal process, in which the

carbon sheets formed during the carbonization process contain only carbon,
considering that the nitrogen and oxygen are gradually eliminated. However,
it has been observed that not all reactions proceed at 100% efficiency and in
the carbonization process take place the formation of intermediates and sub-
products. Some sub-products eliminated during the carbonization process
are: H2, CH4, CO, CO2, H2O, N2, NH3 and HCN (Rahaman et al., 2007).
Jointly, during the chemical transformation of carbon nanofibers different
nitrogen species are formed, some of the most important are (Dalton et al.,
1999; Laffont et al., 2004; Rahaman et al., 2007; Kumar et al., 2012):
pyridine-N (N-6), pyrrolic-N (N-5), pyridone (N-P), quaternary-N or
graphitic-N (N-Q), and pyridinic-N-oxide (N-X), these types of nitrogen
species are illustrated in Figure 3.
Figure 1. Reaction mechanism of PANNF stabilization process in an oxidizing

atmosphere (air).
Figure 2. Reaction mechanism of PANNF carbonization process in an inert

atmosphere (nitrogen).

Figure 3. The different nitrogen species highly probable to be formed in

carbon nanofibers.
The control of the content of these species during the carbonization of

nanofibers, provides a structure with specific structural and electrocatalytic
properties; for example, the introduction of nitrogen as a dopant, which
functions as catalytic centers in fuel cells (Ojeda-López et al., 2017). The
evolution of the hetero-atom content and the chemical bonding along the
various steps of the PAN conversion process shown in Figure 3, it has been
evaluated by different authors through an XPS (X-ray photoelectron
spectroscopy) analysis and EELS (electron energy loss spectroscopy)
(Laffont et al., 2004; Rahaman et al., 2007).
1.3. Vapor Adsorption
The elucidation of textural properties of CNF is usually carried out by

vapor adsorption. The characterization involves the use of two fluids at
lower temperatures than their critical; nitrogen at 77 K and carbon dioxide
at 273 K.
In vapor adsorption, the solid is denominated adsorbent and the gas,
which is potentially capable of being adsorbed, is called adsorptive; while
the gas adsorbed by the solid, is called adsorbate. During the adsorption

process, it is important to distinguish two terms: “adsorption - ascending

limit curve (ALC)” and “desorption - downward limit curve (DLC)”. The
first shows the increase in the amount adsorbed with respect to the increase
in the relative pressure, while the second is the inverse process, i.e., denotes
the progressive decrease of the amount adsorbed.
When the adsorption and desorption curves do not coincide, a loop
known as hysteresis is formed. The graphical representation of both curves
is known as adsorption isotherm, which shows the relationship in the balance
between the amount of gas adsorbed and the gas pressure at constant
temperature. In order to illustrate these processes, Figure 4 shows an
adsorption isotherm for carbon nanofibers synthesized in this work. The
capacity of adsorption at low pressures is usually due to the presence of
micropores.
Figure 4. N2 physisorption isotherm at 77 K of CNF stabilized at 553 K and carbonized

at 1173 K.
In the context of physisorption, the pores are classified based on their

diameter (Thommes et al., 2015): micropores (𝑑 < 2.0 𝑛𝑚), mesopores

(2.0 𝑛𝑚 < 𝑑 < 50.0 𝑛𝑚) and macropores (𝑑 > 50.0 𝑛𝑚); within the
micropores; the last classification have two subdivision: ultramicropores
(𝑑 < 2.0 𝑛𝑚) and supermicropores (0.7 𝑛𝑚 < 𝑑 < 2.0 𝑛𝑚).
BET model (Equation 1) is the most common methodology for the
calculation of specific surface area. Due to the considerations made in
obtaining the BET equation, the application of the BET method to
microporous adsorbents is widely debated in the literature (Rouquerol et al.,
2007). However, several criteria have been proposed for obtaining the BET
area for different types of microporous materials, one of which is described
below.
𝑃 ⁄𝑃 0 (𝐶𝐵𝐸𝑇 −1) 𝑃 1
=[ ] + (1)
𝑛𝑎 (1−𝑃⁄𝑃0 ) 𝑛𝑚 𝐶𝐵𝐸𝑇 𝑃0 𝑛𝑚 𝐶𝐵𝐸𝑇
𝑦 = 𝑚𝑥 + 𝑏
where P/P0 is the relative pressure of equilibrium, na is moles number of gas

adsorbed on 1 gram of adsorbent (mol/g), nm is specific capacity of adsorbate
monolayer, i.e., amount of adsorbate necessary to cover the surface with a
complete monolayer of molecules (mol/g), and CBET: BET constant. When
equation 1 is applied to an adsorption isotherm, the data obtained are known
as BET transformed.
The isotherm of Figure 4, corresponding to a CNF, based on the
recommendations of IUPAC (Thommes et al., 2015) corresponds to a
compound isotherm Type I (due to microporosity) and Type IV (due to
mesoporosity, hysteresis cycle). In these cases, Rouquerol recommends
criteria to be taken into consideration for the selection of analysis region of
BET transformed in the case of microporous materials.
1. The straight portion selected should have a positive intercept on the

ordinate, i.e., no negative value of constant BET “CBET”, which
would be meaningless.
2. The term na (1 − P⁄Po ) should continuously increase together with
P/P°; if not, the pressure range should be narrowed.

Figure 5 shows the BET transformation of the isotherm presented in

Figure 4, corresponding to a CNF. Using the commonly criterion of the
linear region gives rise to a dilemma; due to the fact that several portions of
the graph meet this requirement, specifically three intervals are noted:
0.0002 - 0.13, 0.05 - 0.21 and 0.22 - 0.31. To solve this conflict and select a
single linear portion of BET graph the Rouquerol criteria are used, their
effect is described below (Rouquerol et al., 2007):
1. In Figure 5, one sees that the first criterion eliminates the upper
portion of the BET plot represented there.
2. In Figure 6, the second criterion now eliminates the whole relative
pressure range above 0.08. This criterion can be considered as a self-
consistency criterion for the modified BET equation proposed by
Keii et al. (Keii et al., 1961) and used by Parra et al. in the case of
microporous carbons (Parra et al., 1995).
Considering the criteria mentioned above, all the calculations performed

for the BET area in this work are limited in the relative pressure interval
from 0.002 to 0.08. The selected interval makes possible to corroborate that
micropores filling by physisorption always occurs at low relative pressures
and the interval of low pressure depends on the shape and dimensions of the
micropores, the size of the adsorption molecules and their interactions with
the adsorbent. The adsorption in ultramicropores takes place at very low
relative pressures and this process has been called “primary micropores
filling”, while the supermicropores are filled by a secondary process in a
relative pressure interval, 𝑃⁄𝑃𝑜 ≈ 0.01 − 0.15, for the adsorption of Argon
at 87 K and Nitrogen at 77 K (Thommes et al., 2015).
On the other hand, in the last decades, the adsorption of Nitrogen at 77
K has been accepted as the standard method for the analysis of average size
of micropores and mesopores; however, recent studies have shown that
nitrogen is not a fully satisfactory adsorbent to evaluate the micropore size
distribution (Thommes et al., 2015). One of the reasons is the quadrupole
nature of nitrogen molecule, responsible for the interaction with functional
surface groups and exposed ions.

Figure 5. BET transformation of isotherm presented in Figure 4, corresponding to

a CNF.
Figure 6. Graphic Vads (Po – P) versus p/po (data from Figure 4).
This not only affects the orientation of nitrogen molecule adsorbed on

the surface of the adsorbent, it also strongly affects the filling pressure of
micropores. An alternative to solve this problem is the use of CO2 (kinetic
diameter 0.33 nm) as adsorbent at 273 K, whose diffusion is rapid and allows

access to pores up to 0.4 nm, for this reason the adsorption of CO2 at 273 K
has been become an accepted method for the study of carbon materials with
very narrow micropores (Lowell et al., 2004; Thommes and Cychosz, 2014).
2. MATERIALS AND METHODS
To analyze the effect of the calcination process, it is studied the impact

of the stabilization (first stage) and carbonization (second stage)
temperatures on CNF structural and electrochemical properties.
2.1. Materials
The polymer precursor, polyacrylonitrile (PAN) was acquired from

Sigma-Aldrich, with a molecular mass average of 150,000 uma, Tg = 358 K
and transition temperature of 590 K, the precursor was used as received
without additional purification. The solvent N, N-dimethylformamide
(DMF) grade anhydrous, with a purity of 99.8% and density = 0.944 g/mL,
was also acquired from Sigma-Aldrich.
2.2. Experimental Methodology
PAN nanofibers were synthesized using a concentration of 10% (wt.%)

of PAN, the electrospinning apparatus was set at a flow rate of 0.5 mL/h, a
voltage of 15 kV and 10 cm distance between the tip of the syringe and the
collector. The calcination process was carried out in two stages and the effect
of the temperature was analyzed. First, the effect of stabilization temperature
was evaluated; for this, the PAN nanofibers are divided into four portions
and each one stabilized at 523, 543, 553 and 573 K, respectively, afterwards
all of them were carbonized at 1173 K. In the second stage the impact of
carbonization temperature was studied, with base on the results of stage one,
a stabilization temperature was selected; then, five portions of PANNF were

stabilized at the temperature determined in stage 1 and subsequently

carbonized to 873, 973, 1073, 1173 and 1273 K, respectively. For a clear
description, Figure 7 is presented, in the shaded boxes the treatment
temperatures are indicated and in the white boxes the labels for the materials
obtained from the process is named.
Figure 7. Diagrams of the experimental methodology for the evaluation of the effect of
calcination temperature on the structural properties of PANNF.
2.3. Characterization Techniques
For SEM analysis, the samples were mounted on a double-sided carbon

conductive tape in an aluminum holder (the PANNF were deposited in the
aluminum collector and a small proportion was placed on the carbon tape,
while the CNF were pulverized and placed directly on the carbon tape). Later
CNF were observed in a Scanning Electron Microscope JEOL JSM-6010LA
at 20 kV of acceleration voltage in conditions of high vacuum to 2000X and
10000X. In TEM analysis, 1 μg CNF was mixed with 1 mL of 2-propanol
and immersed in an ultrasound bath (sonicator Branson 2800) during 15 min

to achieve a better dispersion of material, then a drop of this dispersion is

placed on a grid filmed with carbon and left to dry for 24 hours for further
analysis in an Atomic Resolution Transmission Electron Microscopy
equipment, JEOL JEM-ARM200CF, with acceleration voltage of 80–
200kV, using a Cold Field Emission Gun (CFEG) and a spherical aberration
corrector CEOS for the mode STEM.
The FTIR experiments were carried in an equipment Perkin Elmer
Paragon 1000 and the Raman spectra in an equipment with triple
monochromator T64000 Horiba Jobin Yvon, at a power of 20 mW at the
laser output, wavelength 532.1 nm, target x100, 15 accumulations of 60 s in
the region from 50 to 1850 cm-1.
Nitrogen adsorption was performed on a Micromeritics ASAP 2020
Surface Area and Porosity Analyzer, while for CO2 adsorption
Quantachrome Autosob 1 equipment was used with the adaptation of thermal
re-circulator to control the temperature to 273 K.
X-ray photoelectron spectroscopy (XPS) was performed with a Thermo
Scientific K-alpha spectrometer using monochromatic radiation AlKa
source (1487 eV). Instrument base pressure was 1 x 10-9 mbar. The high-
resolution spectra of the target elements were collected using an analysis
area of 400 μm2 at pass energy of 20 eV, with a resolution of 0.6 eV. The
signal of C1s (248.6 eV) was considered as reference to correct charging
effects. The quantitative analysis was carried out using average spectra
collected from different sample regions. The data were analyzed using
Thermo Avantage v5.932. High resolution spectra were deconvoluted using
Gaussian-Lorentzian mixed functions.
3. RESULTS
3.1. Stage 1: Stabilization of PANNF at 523, 543, 553 and 573 K

and Carbonization at 1173 K
The proposed interval for the stabilization temperature was selected

based on thermogravimetric analysis performed on the precursor PANNF.

The main purpose of this analysis is to determine a temperature interval

where the material is not thermally degraded. For the analysis were used 10
mg of PANNF and a heating rate of 5 K/min.
3.1.1. Characterization by FTIR

To observe the effects of stabilization temperature on the chemical
groups of PAN nanofibers, Figure 8 presents the infrared spectra of the
samples in the wavenumber interval from 3000 to 500 cm-1. The PANNF
precursors show particular vibrations in ~ 2245, ~ 1670 and ~ 1454 cm-1,
characteristics of nitrile groups (−C ≡ N), carbonyl groups (-C=O) and
aliphatic groups (CH, CH2 and CH3), respectively (Dalton et al., 1999;
Wangxi et al., 2003; Xue et al., 2013). After performing the heat treatments
in the 523 to 573 K temperature interval, the intensity of signal in 2245 cm-
1
decreases significantly, while the signal intensity of carbonyl groups (-
C=O, ~ 1670 cm-1) increases, forming a wider signal, in which the signals
corresponding to the carbon-carbon double bonds (-C=C-, ~ 1620 cm-1) and
carbon-nitrogen double bonds (-C=N-) are overlapped.
Figure 8. FTIR spectra of PANNF and PANNF stabilized at 523, 543, 553 and 573 K.
In addition, the appearance of a new signal in ~ 810 cm-1 is characteristic

of the aromatic rings (C=C-H). The reduction of nitrile groups, loss of

hydrogen and formation of aromatic structures are the result of cyclization

(-C=N-), dehydrogenation (-C=C-) and oxidation (-C=O) of PAN
nanofibers.
Figure 9. SEM images of PANNF and CNF: (a) PANNF, (b) S523-C1173, (c) S543-
C1173, (d) S553-C1173 and (e) S573-C1173.

When the stabilization temperature is 553 K, the infrared spectroscopy

technique is incapable to detect the presence of nitrile groups, meaning that
most of these groups have been converted into aromatic structures; thus, at
this temperature the objective of the stabilization process has been achieved.
The characterization results presented in this section correspond to PANNF
precursors and PANNF stabilized at 523, 543, 553 and 573 K. The structural
characterization of these stabilized nanofibers and after carbonized at 1173
K is shown below. The separation of the results in this way will allow
appreciating the effect of the stabilization temperature on the structure and
morphology of the resulting CNF.
3.1.2. Characterization by SEM and TEM

Figure 9 shows SEM images of PANNF (Figure 9a) and CNF (Figure
9b-e). Figure 9a corresponds to the PANNF without any thermal treatment;
the use of a fixed metal collector generates fibers with orientations
crisscrossed. Figure 9b presents the material stabilized at 523 K; note that
low stabilization temperatures do not allow to preserve the fibrous structure
of the material after carbonization, resulting in merged nanofibers, without
interstitial spaces between them.
Figure 10. TEM micrographs of PANNF stabilized at 553 K and carbonized at 1173 K.
The materials S543-C1173 and S553-C1173 preserve both, the fibrous

structure and the length of nanofibers (Figure 9c and Figure 9d). Finally, the

material S573-C1173, stabilized at temperatures closer to the PANNF

degradation temperature, presents a break in the length; in this case, the fiber
structure is also preserved, but the reduction in length causes the
agglomeration of nanofibers and as a result the interstitial space between the
nanofibers is lost. Figure 10 shows TEM images of the material stabilized at
553 K and carbonized at 1173 K. In this is evident a random ordering of
carbon sheets forming the structure of the nanofiber. The distance between
the different carbon sheets was determined by image analysis; the values
obtained oscillate between 0.4 and 1.0 nm (the distance between the sheets
of graphitic carbons is ~ 0.4 nm). These results indicate that carbon
nanofibers have a disorderly structure, i.e., amorphous stack ordering.
3.1.3. Characterization by Raman Spectroscopy

The Raman spectra in Figure 11 are characteristic of carbon
nanostructured materials, with two predominant peaks: i) D (disorder) to ~
1360 cm-1 and ii) G (graphitic) to ~ 1580 cm-1, both attributed to sp2 bonded
species (Wang et al., 2003; Kim et al., 2004). The first peak, D, is related to
the disorder, i.e., the presence of fragments of amorphous carbon; this peak
is usually associated with the imperfect graphitic structure. The second peak,
G, is attributed to the longitudinal vibration C-C of the graphite layers and
indicates the degree of graphitization in carbon materials. After all materials
are stabilized, D peak decreases and G peak increases; in general, the
increase in the stabilization temperature generates a material with a lower
graphitization degree. The Raman spectra deconvolution was performed
using a Gaussian-Lorentzian mix of functions and were fitted to obtain the
match the spectra. The relative intensity relationship between bands D and
G, (R = ID/IG), can be calculated by obtaining the area under each
deconvoluted peak. The R value is generally lower when the degree of
structural ordering of carbon is higher. The material stabilized at 573 K
presents a higher value of R, 2.42, this value decreases to 1.82 when the
material is stabilized at 523 K. These results show that the stabilization
process is necessary to reduce the disorder in the arrangement of the
structures of carbon nanofibers. In addition, the amorphous structure
observed by TEM images corroborates these results.

3.1.4. Characterization by N2 and CO2 Adsorption

Figure 12 shows the nitrogen physisorption isotherms of PANNF
stabilized at 523, 543, 553 and 573 K and carbonized at 1173 K. Based on
the IUPAC classification, they correspond to a combination of Type I and
Type IV isotherms, the first one is due to the presence of microporosity and
the second to mesoporosity. These isotherms also have a small hysteresis
cycle, correspondent to Type H4, characteristic of micro-mesoporous
carbons (Thommes et al., 2015). At low relative pressures, nitrogen
adsorption is mainly attributed to the presence of micropores; on the other
hand, at relative pressures greater than 0.35, the materials S543-C1173,
S553-C1173 and S573-C1173 present a hysteresis cycle, associated with the
presence of mesopores (zoom in Figure 12).
Gas physisorption in CNF can occur in two spaces: i) between the
carbon sheets that form the structure of a nanofiber, or ii) between the
interstitial spaces that are formed by the arrangement of several nanofibers.
This explains why the material stabilized at 523 K (S523-C1173) does not
present a hysteresis cycle, as shown in Figure 9, the nanofibers in this
material have completely fused, forming a solid without interstitial spaces.
The hysteresis cycle presented by the materials S543-C1173, S553-C1173
and S573-C1173 (zoom in the lower part of Figure 12) is reduced with the
increase of stabilization temperature, confirming the observations made by
SEM. The material S573-C1173 shows shattered nanofibers, which causes
nanofibers agglomeration and reduction in the interstitial spaces; these
changes originate the decrease in the adsorbed nitrogen capacity.
To obtain information from smaller pores, CO2 adsorption was utilized,
this molecule is smaller than N2 leading to more detailed information about
the pores in CNF. The CO2 adsorption isotherms are obtained at 273 K. The
N2 and CO2 adsorption isotherms allow the calculation of the specific
surface area and the average pore size distribution. The BET method was
used for the calculation of the specific surface area, while the software
Autosorb 1 of Quantachrome equipment was utilized for the calculations of
average pore size, considering the NLDFT model for slit pores.

Figure 11. Raman spectra of CNF at the different stabilization temperatures (a) S523-
C1173, (b) S543-C1173, (c) S553-C1173 and (d) S573-C1173.
Figure 12. Nitrogen physisorption isotherms of the materials (a) S523-C1173, (b)
S543-C1173, (c) S553-C1173, and (d) S573-C1173.

Figure 13 shows the pore size distribution (PSD) obtained from the N2
and CO2 adsorption isotherms. In PSD obtained from N2 all materials have
a preferential pore diameter of ~ 0.8 nm, although it is also possible to
observe a small region of pores in the interval of 1.1 and 1.3 nm (left Figure
13). On the other hand, the PSD obtained from CO2 shows four preferential
pore sizes, one of which coincides with the average pore diameter obtained
by N2 adsorption isotherms, 0.8 nm, which is indicated in the box at the right
of Figure 13. The CO2 adsorption allows determining pore sizes that oscillate
between 0.45 and 1.00 nm. The values obtained for the pore diameters in
both cases, viz Table 1, are comparable with the distances between the
carbon sheets shown in the TEM micrographs of Figure 10 in section 3.1.2,
fluctuating between 0.4 and 1.0 nm. Results confirming that adsorption is
favored among carbon sheets.
Table 1. Structural properties of PANNF stabilized at different

temperatures and carbonized at 1173 K by N2 and CO2 adsorption
Adsorbate Textural S523- S543- S553- S573-

Properties C1173 C1173 C1173 C1173
N2 SBET (m2 g-1) 966.0 740.0 728.0 635.0
SMICRO (m2 g-1) 921.0 699.0 693.0 605.0
SEXT (m2 g-1) 45.0 41.0 35.0 30.0
VT (cm3 g-1) 0.42 0.32 0.29 0.27
VMICRO (cm3 g-1) 0.36 0.27 0.26 0.24
VMESO (cm3 g-1) 0.06 0.05 0.03 0.03
DNLDF (nm) 0.79 0.80 0.81 0.82
CO2 DNLDF (nm) 0.51 0.51 0.51 0.51
DNLDF (nm) 0.59 0.59 0.59 0.59
DNLDF (nm) 0.68 0.71 0.69 0.71
DNLDF (nm) 0.82 0.83 0.82 0.83
SBET ≡ BET area; SMICRO ≡ micropore area (calculated using the t-plot method); SEXT ≡ external
surface area; VT ≡ total volume; VMICRO ≡ micropores volume; VMESO ≡ mesopore volume;
DNLDFT ≡ preferential pore diameter calculated by the NLDFT model; in the case of CO2
adsorption, four pore sizes observed in Figure 14 are reported (DNLDFT).
The textural properties obtained from the N2 and CO2 isotherms are
shown in Table 1. As observed in the isotherms of Figure 12, these materials
present a highly amount of micropores, conjecture corroborated with the

BET area and micropores and mesopores volume. The material stabilized at
523 K presents a BET area of 966 m2/g; this value decreases when the
stabilization temperature increases, decreasing to 635 m2/g for material
stabilized at 573 K. In this way, the increment in the stabilization
temperature plays an important role on the materials microporosity, which
modifies the BET area by 35%. On the other hand, more than 80% of the
volume in each material corresponds to microporosity and decreases with
the increase of the stabilization temperature.
Figure 13. PSD using NLDFT for N2 and CO2 isotherms. (a) S523-C1173, (b) S543-
C1173, (c) S553-C1173, and (d) S573-C1173.
3.2. Stage 2: Stabilization at 553 K and Carbonization at 873,

973, 1073, 1173 and 1273 K
The effect of the stabilization temperature on the structural properties of

CNF allowed to choose temperatures needed to preserve the CNF structure.

543 and 553 K seem to be the most suitable; however, as the FTIR analysis
revealed, the material stabilized at 543 K still shows traces, being the
removal of these groups (nitrile groups) one of the main objectives of
stabilization stage. Thus, it is concluded that the most suitable stabilization
temperature is 553 K; this stabilization temperature is used in stage 2.
In the second calcination stage, 5 g of PANNF are stabilized at 553 K
and then divided in 5 fractions; each fraction is carbonized at different
temperature: 873, 973, 1073, 1173 and 1273 K.
3.2.1. Characterization by N2 and CO2 Adsorption

Figure 14 shows the nitrogen physisorption isotherms for the carbonized
materials. The isotherms present a mixture between Type II and Type IV
isotherms with a H4 hysteresis cycle, corresponding to micro-mesoporous
carbon materials (Thommes et al., 2015). Note that all materials at relative
pressures below 0.1 adsorb more than 80% of their total adsorbed volume,
this phenomenon is characteristic of the presence of narrow micropores (d <
~ 1.0 nm), caused by strong adsorbent-adsorptive interactions. The materials
hysteresis loops present two behaviors: i) carbonized materials at
temperatures above 1073 K show a small closed hysteresis cycle, between
0.35 and 0.70 of relative pressure, associated with capillary condensation
interstitial spaces and, ii) materials carbonized at 873 and 973 K do not
present a closed hysteresis cycle. The first behavior is known as force closed
phenomenon and occurs by pore network effect, due to the presence of
supermicropores, ultramicropores, mesopores and even macropores in
carbon nanofibers, this leads to interconnected pores. When the
experimental isotherm shows signs of pore network effects, it is
recommended to use the adsorption curve for average pore size calculations
(Groen et al., 2003). The unclosed hysteresis loop might be due to (Mastalerz
et al., 2012; Monson, 2012; Thommes and Cychosz, 2014; Qi et al., 2017):
i) Carbon nanofibers which do not possess a rigid structure and can suffer
deformation by adsorption or pore filling, ii) trapped nitrogen which cannot
be released due to the higher affinity for the heterogeneous properties of the
CNF surface, i.e., the adsorption potential of the pore wall traps the nitrogen
molecules (Tang et al., 2015), or iii) the existence of the ink-bottle pores in

carbon nanofibers and the possible permanence of small amounts of

precursor polymer.
Based on the adsorption isotherms of Figure 14, the average pore size
distribution was determined using the NLDFT model. In Figure 15 (left) the
curves corresponding to the micropores interval are presented (diameters
between 0.7 and 2.0 nm). Note that the material carbonized at 873 K presents
a maximum in pore diameter around ~ 1.3 nm, which means that the
separation between the carbon sheets in the carbon nanofibers is not
homogeneous (fluctuate from 0.7 nm to 2.0 nm). Carbonized material at 973
K presented principally pore sizes between 0.7 and 1.0 nm, although a small
proportion corresponding to pore sizes between 1.0 and 2.0 nm can be
observed. As the carbonization temperature increases it is possible to
observe a decrease in the number of pores with pore diameter higher than
1.0 nm; thus, carbonization temperatures above 1173 K generate a material
with a more uniform pore size.
Figure 14. N2 physisorption isotherms of the CNF stabilized at 553 K and calcined at
several temperatures: (a) 873, (b) 973, (c) 1073, (d) 1173 and (e) 1273 K.
To determine the smaller pore sizes (less than 0.7 nm) CO2 adsorption
isotherms at 273 K are used again. The pore size distribution obtained using
CO2 adsorption and applying the NLDFT model is shown in Figure 15
(right). It can be observed more than one maximum in the interval of 0.4 to

1.0 nm pore size, indicating the presence of multiple preferential pore sizes
(Table 2). The results using the N2 adsorption curves do not allow the
determination of pore sizes smaller than 0.7 nm. Even though it is important
to note that the pore size of ~0.83 nm coincides in both cases.
Figure 15. PSD using NLDFT for N2 and CO2 isotherms of PANNF stabilized at 553 K
and carbonized at (a) 873, (b) 973, (c) 1073, (d) 1173 and (e) 1273 K.
Table 2 shows the textural properties of carbon nanofibers. The

calculated specific surface area of the obtained materials is 296, 431, 635,
710 and 822 m2/g for S553-C873, S553-C973, S553-C1073, S553-C1173
and S553-C1273, respectively. Contrary to the observed in the stabilization
stage, the increase in the carbonization temperature leads to an increase in
the specific surface area; similar behavior is observed in the total volume
calculated by nitrogen adsorption, 0.14 to 0.34 cm3.g-1 for S553-C873 and
S553-C1273, respectively. All materials have a contribution of more than
80% of their total adsorbed volume due to micropores present in carbon
nanofibers, which has been previously reported (Chen et al., 2012; Lee et
al., 2013; Wang et al., 2016). Regarding the values determined by the
average pore size distribution curves, the use of N2 adsorption isotherms

allows to determine supermicropores (0.7 to 2.0 nm) and the CO2 adsorption
isotherms allow to determine ultramicropores (< 0.7 nm).
3.2.2. Characterization by XPS

The elemental surface composition of CNF and PAN was investigated
using XPS. The XPS survey scan of CNF and PAN revealed the presence of
carbon, nitrogen, and oxygen (Figure 16). The composition of C, O, and N
(wt.%) were calculated from the corresponding peak areas of the XPS
spectra (Table 3).
Table 2. Textural properties of PANNF carbonized at

873, 973, 1073, 1173 and 1273 K
Adsorbate Textural S553- S553- S553- S553- S553-

Properties C873 C973 C1073 C1173 C1273
N2 SBET (m2 g-1) 296 431 635 710 822
SMICRO (m2 g-1) 266 398 597 665 779
SEXT (m2 g-1) 30 33 38 45 43
VT (cm3 g-1) 0.14 0.19 0.27 0.33 0.34
VMICRO (cm3 g-1) 0.11 0.16 0.24 0.29 0.30
VMESO (cm3 g-1) 0.03 0.03 0.03 0.04 0.04
DNLDF (nm) 1.15 0.80 0.80 0.83 0.83
CO2 DNLDF (nm) 0.51 0.51 0.51 0.51 0.51
DNLDF (nm) 0.59 0.59 0.59 0.59 0.59
DNLDF (nm) 0.73 0.72 0.72 0.70 0.72
DNLDF (nm) 0.83 0.83 0.83 0.83 0.83
SBET ≡ BET area; SMICRO ≡ micropore area (calculated using the t-plot method); SEXT ≡ external
surface area; VT ≡ total volume; VMICRO ≡ micropore volume; VMESO ≡ mesopore volume;
DNLDFT ≡ preferential pore diameter calculated by the NLDFT model; in the case of CO2
adsorption, four pore sizes observed in Figure 17 are reported (DNLDFT).
Due to the scale, in Figure 16 the signals corresponding to C 1s and O

1s show no significant change between the different carbonization
temperatures; therefore, Table 3 presents the percentage content of these
elements. When the material is carbonized at higher temperatures, the O 1s
has a variation small, the C 1s present a gradual increase from 73 to 91%,
and the N 1s decreases from 25 to 5%. Note that the signal of N 1s for the
PAN presents the highest value with respect to all carbonized materials; this

behavior is related to the lack of thermal treatment, as it is taken as a

reference to compare changes in nitrogen concentration.
Figure 16. XPS general spectra of (a) precursor material (PAN) and PANNF stabilized
at 553 K and carbonized at (b) 873, (c) 973, (d) 1073, (e) 1173 and (f) 1273 K.
The general XPS spectra allow identifying the surface composition of

the samples. To obtain a more specific study on some of elements, it is
necessary to carry out a high resolution analysis. In this way, the different
nitrogen species present in the carbon nanofibers are determined. Figure 17
shows high resolution spectra for N 1s. Observe that for the PAN, the
spectrum presents a single maximum, centered on 399.5 eV, characteristic
of nitrile bonds (C ≡ N). As already mentioned, the stabilization of PAN
nanofibers causes a break in the C ≡ N bonds to form C=N; bonds; while
with the carbonization temperature occurs the elimination of H2(g) and N2(g),

generating different intermediate species. In the spectra of Figure 18, mainly

three species predominant of nitrogen are appreciated. In the materials
carbonized to 873 and 973 K predominate pyridine nitrogen and pyrrole
nitrogen; temperatures higher than 1073 K show a decrease in both species,
probably due to the conversion of pyridine nitrogen to quaternary nitrogen,
which are favored at high temperatures (> 1073 K) and which are now the
predominant groups. Therefore, carbonization at temperatures above 1073
K induces the formation of new nitrogen species such as: pyridone,
quaternary nitrogen or graphite, pyridine nitrogen with oxidized species and
chemisorbed nitrogen (Fitzer and Müller, 1975; Pels et al., 1995; Laffont et
al., 2004; Kumar et al., 2012; Roldán et al., 2012) and causes disappearance
of the original functional groups.
For the purpose of quantification, the deconvolution of the spectra
shown in Figure 17 was performed. Seven binding regions can be assigned
to the different nitrogen groups (peaks), these peaks could be assigned to: i)
residual PAN (397.8 eV), ii) pyridine-N (398.8 eV), iii) cyanide functional
group, C ≡ N (399.5 eV), iv) pyrrolic-N/pyridone (400.5 eV), v) quaternary-
N or graphitic-N (401.8 eV), vi) pyridine-N-oxide (403.1 eV), vii)
chemisorbed nitrogen oxide species/𝜋 → 𝜋 ∗ transitions (405.12 eV). Pels et
al. mentioned that the chemical environment of the nitrogen atom in
pyridone is similar to pyrrolic-N, in both, the nitrogen atom contributes with
two p-electrons to the system, while the hydrogen atom is bound in the plane
of the ring; therefore, within the accuracy of XPS measurements, pyridone
cannot be distinguished from pyrrolic-N (Pels et al., 1995).
Table 3. Surface composition of PAN and CNF derived

of from the XPS spectra (wt.%)
Material C 1s N 1s O 1s
PAN 73 25 2
S553-C873 78 16 6
S553-C973 82 13 5
S553-C1073 85 9 6
S553-C1173 90 5 5
S553-C1273 91 5 4

Figure 17. High resolution XPS of N 1s. (a) PAN, (b) S553-C873, (c) S553-C973, (d)
S553-C1073, (e) S553-C1173, and (f) S553-C1273.
Figure 18. Nitrogen functional groups obtained through high resolution XPS spectra
for N 1s: (a) residual PAN, (b) chemisorbed nitrogen oxide species or 𝜋 → 𝜋 ∗
transitions, (c) C ≡ N, (d) pyrrolic or pyridone, (e) pyridine-N-oxide, (f) pyridine, and
(g) quaternary/graphitic.
Figure 18 shows the contribution of nitrogen species for all materials,

calculated through the deconvolution of the high resolution XPS spectra for
N1s. Three main species predominate in CNF. The first of these are pyridine
nitrogen, which originally come from the precursor used and decrease with

rising carbonization temperatures. Pyrrolic and pyridone nitrogen are

abundant at temperatures below 1073 K. Finally, quaternary/graphitic
nitrogen grows proportionally with the increase in carbonization
temperature. The number and types of nitrogen species are determinant in
the selection of the application for CNF. Fox example, as catalysts for the
oxygen reduction reactions (ORR).
3.3. Electrocatalytic ORR Activity of the CNFs
Preliminary analyses of CNF in ORR are shown in Figure 19. Figure 19

shows the CV curves of CNFs in O2 saturated electrolyte. When the
experiment is performed in oxygen-saturated electrolyte, the presence of a
well-defined reduction peak at potentials between 0.8 and 1.0 V vs RHE
indicates activity for the oxygen reduction for materials carbonized at
temperatures above 1073 K. On the other hand, the samples carbonized at
lower temperatures do not present a well-defined reduction peak but a slight
increment on the current, indicative of a much lower activity towards the
ORR. Connecting these results with the XPS experiments, it is evident that
although the amount of nitrogen species found in the nanofibers prepared at
lower calcination temperatures is higher (873 and 973 K), the lack of
quaternary nitrogen (or the lower amount of them) is the main cause of the
poor electrocatalytic activity, and also, connected to the lower extent of
graphitic domains which influences the material conductivity. The position
of the peaks associated to the ORR in the CV experiments, is positively
shifted as consequence of the increment on the carbonization temperature,
thus the catalytic effect of such nitrogen groups is directly being observed
i.e., the higher the carbonization temperature the lower the overpotential for
the ORR to occur. For purposes of application in fuel cells, the optimal
amount of nitrogen as a dopant that leads to greater catalytic activity in ORR
in graphitic materials, remains a debatable issue; however, some authors
have agreed that the total amount of nitrogen is not very important, the
important is the amount of quaternary or graphitic nitrogen, and these groups
should be present in a proportion of around 3% in order to present
advantages for the ORR (Lai et al., 2012; Vazquez-Arenas et al., 2016).

Figure 19. Cyclic voltammetry in saturated oxygen for carbon nanofibers, ν = 5 mV/s.
(a) S553-C873, (b) S553-C973, (c) S553-C1073, (d) S553-C1173 and (e) S553-C1273.
Carbonized materials at 1173 and 1273 K present ~ 5% total nitrogen,

of which ~ 2% corresponds to quaternary/graphitic nitrogen; this indicates
that both materials can have a positive effect on the catalytic properties
towards the ORR (Ojeda-López et al., 2017). In this work, the adequate
carbonization temperature is 1173 K, because the increase in the
carbonization temperature at 1273 K did not represent substantial increases
in the catalytic activity, any changes in the amount of quaternary nitrogen.
The above allows elucidating that carbonization method to transform
PANNF to CNF can be used to control the amount and type of nitrogen
species, the degree of graphitization and the specific surface area; which
influence the electrocatalytic properties in the ORR. These materials can
catalyze this reaction to lower overpotentials in comparison with non-doped
graphitic materials and, very importantly, without the use of noble metal
electrocatalysts.
CONCLUSION
In this work, N-doped CNF were synthesized using PAN as the carbon
and nitrogen source. The synthesis of PANNF by electrospinning followed
by PANNF calcination, allows the control of the microporosity and chemical

composition of the obtained CNF. The main results of this work can be
summarized by the following:
First, the stabilization step controls the degree of graphitization,
fibrillary structure, and porosity in CNF. A stabilization temperature of 523
K generates a material with a high microporosity; however, nanofibers
stabilized at this temperature melt when they are carbonized at 1173 K. The
stabilization at 573 K produces brittle structures with lower microporosity.
The stabilization at 543 and 553 K allows the preservation of the fibrous
structure of CNF and present a greater degree of graphitization. FTIR
analysis of the materials stabilized at 543 K show the presence of nitrile
groups, which should have been removed at this step. Therefore, the
stabilization temperature of 553 K turns out to be the most viable, to achieve
a better graphitization and greater microporosity.
Second, the carbonization step caused different structural and chemical
changes: i) better degree of graphitization, ii) increases microporosity, iii)
higher carbon content and decreases of nitrogen content, and iv)
predominance of quaternary/graphitic nitrogen species. This last property
allows the use of CNF in ORR, because these species increase the catalytic
activity of CNF. These species are most abundant in the material carbonized
at 1173 K. The synthesis herein is competitive and adaptable relative to other
techniques reported in the literature because it allows the tailoring of the
presence of nitrogen functional groups by controlling the calcination
temperature.
REFERENCES
Beachley, V; Wen, X. Effect of Electrospinning Parameters on the Nanofiber

Diameter and Length. Mater. Sci. Eng. C, 2009, 29 (3), 663–668.
Chand, S. Carbon Fibers for Composites. J. Mater. Sci., 2000, 35 (6), 1303–
1313.
Chen, LF; Zhang, XD; Liang, HW; Kong, M; Guan, QF; Chen, P; Wu, ZY;
Yu, SH. Synthesis of Nitrogen-Doped Porous Carbon Nanofibers as an

Efficient Electrode Material for Supercapacitors. ACS Nano, 2012, 6 (8),

7092–7102.
Dalton, S; Heatley, F; Budd, PM. Thermal Stabilization of Polyacrylonitrile
Fibres. Polymer, 1999, 40 (20), 5531–5543.
Eom, Y; Kim, BC. Solubility Parameter-Based Analysis of Polyacrylonitrile
Solutions in N,N-Dimethyl Formamide and Dimethyl Sulfoxide.
Polymer, 2014, 55 (10), 2570–2577.
Fitzer, E; Müller, DJ. The Influence of Oxygen on the Chemical Reactions
during Stabilization of Pan as Carbon Fiber Precursor. Carbon N. Y.,
1975, 13 (1), 63–69.
Gomes, DS; Silva, ANRda; Morimoto, NI; Mendes, LTF; Furlan, R; Ramos,
I. Characterization of an Electrospinning Process Using Different
PAN/DMF Concentrations. Polímeros: Ciência e Tecnologia, 2007, 17
(3), 206–211. [Polymers: Science and Technology]
Groen, JC; Peffer, LAA; Pérez-Ramírez, J. Pore Size Determination in
Modified Micro- and Mesoporous Materials. Pitfalls and Limitations in
Gas Adsorption Data Analysis. Microporous Mesoporous Materials,
2003, 60 (1–3), 1–17.
Guo, C; Zhou, L; Lv, J. Effects of Expandable Graphite and Modified
Ammonium Polyphosphate on the Flame-Retardant and Mechanical
Properties of Wood Flour-Polypropylene Composites. Polym. Polym.
Compos., 2013, 21 (7), 449–456.
He, JH; Wan, YQ; Yu, JY. Effect of Concentration on Electrospun
Polyacrylonitrile (PAN) Nanofibers. Fibers Polym., 2008, 9 (2), 140–
142.
Huang, ZM; Zhang, YZ; Kotaki, M; Ramakrishna, S. A Review on Polymer
Nanofibers by Electrospinning and Their Applications in
Nanocomposites. Compos. Sci. Technol., 2003, 63 (15), 2223–2253.
Iovleva, MM; Smirnova, VN; Budnitskii, GA. The Solubility of
Polyacrylonitrile. Fibre Chem., 2001, 33 (4), 262–264.
Jalili, R; Morshed, M; Ravandi, SAH. Fundamental Parameters Affecting
Electrospinning of PAN Nanofibers as Uniaxially Aligned Fibers. J.
Appl. Polym. Sci., 2006, 101 (6), 4350–4357.

Keii, T; Takagi, T; Kanetaka, S. A New Plotting of the BET Method. Anal.

Chem., 1961, 33 (13), 1965.
Kim, C; Park, SH; Cho, JI; Lee, DY; Park, TJ; Lee, WJ; Yang, KS. Raman
Spectroscopic Evaluation of Polyacrylonitrile-Based Carbon Nanofibers
Prepared by Electrospinning. J. Raman Spectrosc., 2004, 35 (11), 928–
933.
Ko, TH. The Influence of Pyrolisis Properties and Microstructure of
Modified PAN Fiber during Carbonization. J. Appl. Polym. Sci., 1991,
43, 589–600.
Kumar, A; Ganguly, A; Papakonstantinou, P. Thermal Stability Study of
Nitrogen Functionalities in a Graphene Network. J. Phys. Condens.
Matter, 2012, 24 (23), 235503-235508.
Laffont, L; Monthioux, M; Serin, V; Mathur, RB; Guimon, C; Guimon, M.
F. An EELS Study of the Structural and Chemical Transformation of
PAN Polymer to Solid Carbon. Carbon N. Y., 2004, 42 (12–13), 2485–
2494.
Lai, L; Potts, JR; Zhan, D; Wang, L; Poh, CK; Tang, C; Gong, H; Shen, Z;
Lin, J; Ruoff, RS. Exploration of the Active Center Structure of
Nitrogen-Doped Graphene-Based Catalysts for Oxygen Reduction
Reaction. Energy Environ. Sci., 2012, 5 (7), 7936.
Lee, HM; Kang, HR; An, KH; Kim, HG; Kim, BJ. Comparative Studies of
Porous Carbon Nanofibers by Various Activation Methods. Carbon
Lett., 2013, 14 (3), 180–185.
Liu, J; Wang, PH; Li, RY. Continuous Carbonization of Polyacrylonitrile‐
based Oxidized Fibers: Aspects on Mechanical Properties and
Morphological Structure. J. Appl. Polym. Sci., 1994, 52 (7), 945–950.
Lowell, S; Shields, JE; Thomas, MA; Thommes, M. Characterization of
Porous Solids and Powders: Surface Area, Pore Size and Density.
Netherlands: Kluwer Academic Publishers, 2004.
Mastalerz, M; He, L; Melnichenko, YB; Rupp, JA. Porosity of Coal and
Shale: Insights from Gas Adsorption and SANS/USANS Techniques.
Energy and Fuels, 2012, 26 (8), 5109–5120.
Minagawa, M; Kanoh, H; Tanno, S; Nishimoto, Y. Glass-Transition
Temperature (Tg) of Free-Radically Prepared Polyacrylonitrile by

Inverse Gas A Study on Tg of Atactic. Monodisperse Polystyrenes, 2002,

17, 2475–2480.
Monson, PA. Understanding Adsorption/Desorption Hysteresis for Fluids in
Mesoporous Materials Using Simple Molecular Models and Classical
Density Functional Theory. Microporous Mesoporous Materials, 2012,
160, 47–66.
Mottaghitalab, V; Haghi, AK. A Study on Electrospinning of
Polyacrylonitrile Nanofibers. Korean J. Chem. Eng., 2011, 28 (1), 114–
118.
Ojeda-López, R; Ramos-Sánchez, G; Esparza-Schulz, JM; Lartundo, L;
Domínguez-Ortiz, A. On Site Formation of N-Doped Carbon
Nanofibers, an Efficient Electrocatalysts for Fuel Cell Applications. Int.
J. Hydrogen Energy, 2017, 2, 2–11.
Panapoy, M; Dankeaw, A; Ksapabutr, B; Technology, I; Campus, SP.
Electrical Conductivity of PAN-Based Carbon Nanofibers Prepared by
Electrospinning Method. Mater. Sci., 2008, 13, 11–17.
Park, SJ. Carbon Fibers., 210. New York, London: Springer; 2015.
Parra, JB; de Sousa, JC; Bansal, RC; Pis, JJ; Pajares, JA. Characterization
of Activated Carbons by the BET Equation — An Alternative Approach.
Adsorpt. Sci. Technol., 1995, 12 (1), 51–66.
Pels, JR; Kapteijn, F; Moulijn, JA; Zhu, Q; Thomas, KM. Evolution of
Nitrogen Functionalities in Carbonaceous Materials during Pyrolysis.
Carbon N. Y., 1995, 33 (11), 1641–1653.
Qi, L; Tang, X; Wang, Z; Peng, X. Pore Characterization of Different Types
of Coal from Coal and Gas Outburst Disaster Sites Using Low
Temperature Nitrogen Adsorption Approach. Int. J. Min. Sci. Technol.,
2017, 27 (2), 371–377.
Rahaman, MSA; Ismail, AF; Mustafa, A. A Review of Heat Treatment on
Polyacrylonitrile. Fiber. Polym. Degrad. Stab., 2007, 92 (8), 1421–
1432.
Ramakrishna, S; Fujihara, K; Teo, WE; Lim, TC; Ma, Z. An Introduction to
Electrospinning and Nanofibers., 48. Singapore: World Scientific
Publishing Co. Pte. Ltd, 2005.

Rodriguez, NM. A Review of Catalytically Grown Carbon Nanofibers. J.

Mater. Res., 1993, 8 (12), 3233–3250.
Roldán, L; Armenise, S; Marco, Y; García-Bordejé, E. Control of Nitrogen
Insertion during the Growth of Nitrogen-Containing Carbon Nanofibers
on Cordierite Monolith Walls. Phys. Chem. Chem. Phys., 2012, 14 (10),
3568.
Rouquerol, F; Rouquerol, J; Sing, KSW; Llewellyn, P; Maurin, G.
Adsorption by Powders and Porous Solids: Principles, Methodology
and Applications. Marseille, France: Academic Press, 2014.
Rouquerol, J; Llewellyn, P; Rouquerol, F. Is the Bet Equation Applicable to
Microporous Adsorbents? Studies in Surface Science and Catalysis,
2007, 160, 49–56.
Shin, YM; Hohman, MM; Brenner, MP; Rutledge, GC. Experimental
Characterization of Electrospinning: The Electrically Forced Jet and
Instabilities. Polymer, 2001, 42 (25), 9955–9967.
Subbiah, T; Bhat, GS; Tock, RW; Parameswaran, S; Ramkumar, SS.
Electrospinning of Nanofibers. J. Appl. Polym. Sci., 2005, 96 (2), 557–
569.
Tang, X; Wang, Z; Ripepi, N; Kang, B; Yue, G. Adsorption Affinity of
Different Types of Coal: Mean Isosteric Heat of Adsorption. Energy and
Fuels, 2015, 29 (6), 3609–3615.
Theron, SA; Zussman, E; Yarin, AL. Experimental Investigation of the
Governing Parameters in the Electrospinning of Polymer Solutions.
Polymer, 2004, 45, 2017–2030.
Thommes, M; Cychosz, KA. Physical Adsorption Characterization of
Nanoporous Materials: Progress and Challenges. Adsorption, 2014, 20
(2–3), 233–250.
Thommes, M; Kaneko, K; Neimark, AV; Olivier, JP; Rodriguez-Reinoso, F;
Rouquerol, J; Sing, KSW. Physisorption of Gases, with Special
Reference to the Evaluation of Surface Area and Pore Size Distribution
(IUPAC Technical Report). Pure Appl. Chem., 2015, 87 (9–10), 1051–
1069.
Vazquez-Arenas, J; Ramos-Sanchez, G; Franco, AA. A Multi-Scale Model
of the Oxygen Reduction Reaction on Highly Active Graphene

Nanosheets in Alkaline Conditions. J. Power Sources, 2016, 328, 492–

502.
De Vrieze, S; Van Camp, T; Nelvig, A; Hagström, B; Westbroek, P; De
Clerck, K. The Effect of Temperature and Humidity on Electrospinning.
J. Mater. Sci., 2009, 44 (5), 1357–1362.
Wang, C; Chien, HS; Hsu, CH; Wang, YC; Wang, CT; Lu, HA.
Electrospinning of Polyacrylonitrile Solutions at Elevated
Temperatures. Macromolecules, 2007, 40 (22), 7973–7983.
Wang, MX; Guo, ZY; Huang, ZH; Kang, FY. Preparation of Porous Carbon
Nanofibers with Controllable Pore Structures for Low-Concentration
NO Removal at Room Temperature. New Carbon Materials, 2016, 31
(3), 277–286.
Wang, T; Kumar, S. Electrospinning of Polyacrylonitrile Nanofibers. J.
Appl. Polym. Sci., 2006, 102 (2), 1023–1029.
Wang, Y; Serrano, S; Santiago-Avilés, JJ. Raman Characterization of Carbon
Nanofibers Prepared Using Electrospinning. Synthetic Metals, 2003,
138, 423–427.
Wangxi, Z; Jie, L; Gang, W. Evolution of Structure and Properties of PAN
Precursors during Their Conversion to Carbon Fibers. Carbon N. Y.,
2003, 41 (14), 2805–2812.
Wu, M; Wang, Q; Li, K; Wu, Y; Liu, H. Optimization of Stabilization
Conditions for Electrospun Polyacrylonitrile Nanofibers. Polym.
Degrad. Stab., 2012, 97 (8), 1511–1519.
Wu, QY; Chen, XN; Wan, LS; Xu, ZK. Interactions between
Polyacrylonitrile and Solvents: Density Functional Theory Study and
Two-Dimensional Infrared Correlation Analysis. J. Phys. Chem. B,
2012b, 116 (28), 8321–8330.
Xue, Y; Liu, J; Liang, J. Correlative Study of Critical Reactions in
Polyacrylonitrile Based Carbon Fiber Precursors during Thermal-
Oxidative Stabilization. Polym. Degrad. Stab., 2013, 98 (1), 219–229.
Yördem, OS; Papila, M; Menceloğlu, YZ. Effects of Electrospinning
Parameters on Polyacrylonitrile Nanofiber Diameter: An Investigation
by Response Surface Methodology. Mater. Des., 2008, 29 (1), 34–44.

Yu, DG; Branford-White, CJ; Chatterton, NP; White, K; Zhu, LM; Shen,
XX; Nie, W. Electrospinning of Concentrated Polymer Solutions.
Macromolecules, 2010, 43 (24), 10743–10746.
Zhang, L; Aboagye, A; Kelkar, A; Lai, C; Fong, H. A Review: Carbon
Nanofibers from Electrospun Polyacrylonitrile and Their Applications.
J. Mater. Sci., 2014, 49 (2), 463–480.
Zheng, JS; Zhang, XS; Li, P; Zhou, XG; Yuan, WK. Microstructure Effect
of Carbon Nanofiber on Electrocatalytic Oxygen Reduction Reaction.
Catal. Today, 2008, 131 (1–4), 270–277.

Chapter 3
PURIFICATION OF GAS AND LIQUID

STREAMS USING COMPOSITE
SORBENTS EMBEDDED IN
A POLYACRYLONITRILE MATRIX
Brian J. Riley1,*, PhD and Troy G. Garn2

1
Pacific Northwest National Laboratory, Richland, WA, US
2
Idaho National Laboratory, Idaho Falls, ID, US
ABSTRACT
As new sorbents are developed for purifying gases or liquids via
physisorption and/or chemisorption processes, the objective is typically to
increase the specific surface area (SSA) to maximize the available binding
sites for capturing various species of interest. One downfall to making a
sorbent more porous is that it often leads to a product that is fragile or
friable and can be damaged (resulting in unwanted powders), thus
imposing process limitations. Techniques can be used to enhance the
mechanical integrity of a sorbent, including heat treatments or pelletization
with extrusion or granulation processes, but these can result in collapsed,
unavailable pore structures within the sorbent, which decreases the
*
Corresponding Author Email: brian.riley@pnnl.gov.

120 Brian J. Riley and Troy G. Garn
diffusivity and eventually reduces the capacities of target species. An

alternative approach for improving the mechanical integrity of the sorbent
is to bind it (or encapsulate it) within a porous matrix that passively holds
the sorbent in place. Typically, this approach leads to a decrease in sorption
efficiency on the basis of starting sorbent mass, because the active sorbent
is diluted by the binding matrix material and the active surfaces of the
sorbent are somewhat obscured by the binding matrix; however, it comes
with the advantage that the active sorbent is held in place and better
protected from disintegrating during operation. One approach to securing
the active sorbent is to embed it in a macroporous, passive polyacrylonitrile
(PAN) matrix to create a composite sorbent. When optimized, the PAN
matrix is highly porous and allows for gaseous or aqueous media transport
through the product, providing adequate binding site access between the
active sorbent and the species of interest. Including a polymer in the
composite sorbent limits the maximum operating temperature, above
which the composite sorbent incurs loss of structural integrity (e.g.,
decomposes) and/or reduction in SSA. Composite sorbents have been
demonstrated with sulfide aerogels (SnS-PAN) for capturing I2(g),
hydrogen mordenite (HZ-PAN) and silver-exchanged mordenite (AgZ-
PAN) for capturing Kr and Xe, nanoparticle metal oxides (Nano-
Composite Arsenic Sorbent, or N-CAS) for removing arsenic from water,
ammonium molybdophosphate (AMP-PAN) for removing cesium-137
(137Cs) from aqueous nuclear waste streams, and The Phosphate Sponge
(TPS) composite, which is being used to remove inorganic/ortho-
phosphates in water treatment processes to mitigate introduction to the
environment, which impedes toxic algae bloom formations. This chapter
provides details on production of these materials along with options to
optimize the synthesis process. It also includes experimental techniques to
characterize products and assess their performance.
1. INTRODUCTION
Sorbents are used for a variety of purposes, including the selective
capture and retention of specific target species from gaseous and/or liquid-
based process streams. The sorbent of choice for a given application must
maintain physical integrity within the active environment, which can pose
issues if a particular sorbent is effective for capturing the target species but
fragile, in fine particulate form, or otherwise mobile. One way to improve
the stability of the sorbent is to embed the active sorbent in a passive and
macroporous matrix; polyacrylonitrile [also called PAN or poly(1-
acrylonitrile)] is an option for this role. PAN is an organic polymer available

Purification of Gas and Liquid Streams … 121
in granular or fibrous solid form with the chemical formula of (C3H3N)n.

PAN exhibits numerous beneficial properties that make it a suitable choice
for use as a binding agent, including the following:
1) PAN has fairly high temperature stability (~180–300°C) [1, 2], and
thus is naturally suited for gaseous phase capture at elevated
temperatures.
2) PAN is soluble in a variety of media including organic solvents such
as dimethyl sulfoxide (C2H6SO or DMSO), dimethyl acetamide
(C4H9NO), dimethyl formamide (C3H7NO), ethylene carbonate
[(CH2O)2CO], sulfolane [(CH2)4SO2 or tetrahydrothiophene 1,1-
dioxide], and N-methylpyrrolidone (C5H9NO or N-methyl-2-
pyrrolidone); mineral acids such as >5 M phosphoric (H3PO4), >5
M sulfuric (H2SO4), and >8 M nitric (HNO3); and concentrated
inorganic salt solutions such as sodium thiocyanate (NaSCN),
lithium bromide (LiBr), and zinc chloride (ZnCl2) [3].
3) PAN has a macroporous structure that allows access to the surface
area of the active sorbent [4].
4) PAN is chemically stable, or insoluble, in neutral to moderately
acidic solutions, making it a good choice as a scaffold for aqueous
applications [5].
5) PAN is mechanically robust and particularly stable in radioactive
environments [4].
Table 1 provides a summary of some PAN-composite sorbents

previously developed that are discussed in the literature. However, only the
following composite sorbents will be discussed here in detail: tin sulfide
aerogels (SnS-PAN) [6] for capturing I2(g) [7], hydrogen mordenite (HZ)
[8, 9] and silver mordenite (AgZ) [10, 9] (i.e., HZ-PAN and AgZ-PAN,
respectively) for capturing Kr and Xe from used nuclear fuel reprocessing
streams [7], Nano-Composite Arsenic Sorbent (N-CAS) for removing toxic
metals from drinking water (e.g., As, Se, Sb) [11, 3, 12-14], The Phosphate
Sponge (TPS) for removing inorganic/ortho-phosphates from water to
remediate toxic algae blooms [15], and ammonium molybdophosphate

137
(AMP-PAN) for removing Cs from highly acidic aqueous nuclear waste
streams [16, 17].
Table 1. PAN-composite sorbents from the literature along with

the active sorbent, whether the sorbent can be regenerated (“Reg.”),
the applications where each composite sorbent has been demonstrated
with moderate-to-high selectivities for specific contaminants,
and corresponding references
Sorbent name Active sorbent Reg.* Application(s) Reference(s)

AgZ-PAN Silver mordenite Y Xe, Kr [10, 9]
AMP-PAN Ammonium molybdophosphate Y Cs, Sr, Ln, Am [16, 17, 19, 23]
Ba(Ca)SO4-PAN Barium/calcium sulfate U Sr, Ce, Y [19, 23]
CLIN-PAN Natural clinoptilolite U Sr [26]
CSbA-PAN Crystalline antimonic acid U Sr [20, 23]
CuFC-PAN Copper ferrocyanide U Cs [25, 27]
HZ-PAN Hydrogen mordenite Y Xe, Kr [8, 9]
KCoFC-PAN Potassium cobalt ferrocyanide U Cs [28]
KCuHCF-PAN Potassium copper ferrocyanide U Cs [29]
KNiFC-PAN Potassium nickel ferrocyanide U Cs [30]
KZnHCF-PAN Potassium zinc ferrocyanide U Cs [29]
M315-PAN Synthetic mordenite M315 U Cs, Sr, Ce [19, 23]
MgO-PAN Magnesium oxide U Sr, Am, Y [19, 23]
MnO-PAN Manganese dioxide U Pu, Ce, Am, Sr, Y [19, 23]
N-CAS Metal oxides, hydroxides, U As, Se, Sb [11, 3, 12-14]
oxyhydroxides, or salts
Na-Y-PAN Synthetic Na-Y zeolite U Pu [19]
NaTiO-PAN Sodium titanate Y U, Sr, Am [19, 23]
NiFC-PAN Nickel hexacyanoferrate N Cs, Sr, Am, Y [31, 19, 23]
NM-PAN Manganese dioxide and nickel U Sr [23]
ferrocyanide
SnS-PAN Tin sulfide aerogel N I [7]
SnSbA-PAN Tin(IV) antimonate U Sr [23]
TiO-PAN Titanium oxide Y Sr, U, Am [19, 23]
TiP-PAN Titanium phosphate U Sr, U [23]
TiSbA-PAN Titanium(IV) antimonate U Sr [23]
TPS Proprietary (unspecified) Y PO43− [15]
ZrO-PAN, Zirconium oxide and/or U Sr, Ce, Am [23]
ZrP-PAN, and phosphate
ZrOP-PAN
An overview of several composites from this list is provided elsewhere [4, 18-24] and the reader is
referred to additional references for more composites not listed here (some with unspecified
applications) [4, 25]. *Y = yes, N = no, U = unknown

2. GENERAL PROCEDURE FOR SYNTHESIZING

COMPOSITE SORBENTS
Figure 1. Conceptual design of PAN-composite sorbent synthesis. This figure was

modified from the original by Riley et al. [6] and reprinted with permission from the
American Chemical Society.
Figure 1 shows the general process for generating a composite sorbent

using PAN, DMSO, and a sulfide-based aerogel (chalcogel) sorbent. In this
general process, the steps are as follows:
1) The active sorbent is added at a known concentration (i.e., mass of

sorbent per unit volume of solvent) to a solvent (e.g., DMSO) in
vessel-1 and is properly suspended in the solvent using agitation
(e.g., magnetic stirring, vortexing).
2) Then, the PAN is added to the solvent-sorbent mixture in vessel-1,
typically as a granular or fibrous material; gentle heating at ~40–
80°C can be used to accelerate dissolution of PAN into the solvent.
3) Once the PAN dissolves, the mixture (i.e., PAN + solvent + sorbent)
is removed from vessel-1 using a pipette and added dropwise into
vessel-2, which contains water or another fluid in which the PAN is
insoluble, thus creating a granular composite product.
4) The product is rinsed in water and the granular composite product
is dried in an oven (e.g., ~50°C) and/or under vacuum.

Note that many modifications can be made to this process, including (a)
the order of steps (1) and (2) can be reversed [6]; (b) instead of adding drops
of the mixture into a second solvent as described in Step-3 above, different
shapes could be made using molds, and then these could be immersed in
water (or other liquid), where the product can float out of the mold; (c) these
could be made in a more continuous process in which the mixture in Step-2
is poured into water; (d) the water could be replaced with acidic or basic
solutions (0.1–8 M NaOH); or (e) alternative polymers with high specific
surface areas (i.e., SSA, typically in m2 g-1) can be used in place of PAN as
the passive matrix.
3. SORBENT CHARACTERIZATIONS
Once a composite sorbent is synthesized, it can be characterized with a

variety of analytical techniques that can be broadly categorized into (1)
physical/structural analyses, (2) chemical analyses, and (3) performance
analyses. Analytical characterization techniques to consider for a given
sorbent are listed below:
 Physical/structural analyses
 Mechanical strength testing to determine the crush resistance
and attrition/abrasion resistance of sorbents to validate their use
in process applications.
 Microscopy to look at changes in pore structure as a function of
active sorbent loading in the PAN matrix, e.g., scanning and/or
transmission electron microscopies (i.e., SEM, STEM, TEM),
optical microscopy.
 Crystallographic analysis to assess crystalline phase changes in
the active sorbent after capture, e.g., X-ray diffraction, selected
area diffraction.
 Specific surface area, pore volume, and pore size distribution
analyses using adsorption/desorption isotherms, e.g., Brunauer-

Emmett-Teller method [32], Barret-Joyner-Halenda method

[33].
 Chemical analyses
 Assessment of actual versus targeted composition using surface
techniques, e.g., energy-dispersive spectroscopy, electron
energy loss spectroscopy.
 Assessment of actual composition versus targeted composition
using bulk techniques, e.g., sample dissolution followed by
analysis of liquid product with inductively coupled plasma mass
spectroscopy, ion chromatography, and/or ion-selective
electrodes.
 Performance analyses
 Determination of maximum capacity for target species without
known competing species present [e.g., I2(g) capture without
Cl2(g) present] using batch or continuous-stream processes.
 Determination of the influence of sorption competition with
other species present that may or may not interfere with the
capture of the target species analytically evaluated with gas
chromatography.
 Breakthrough experiments where the active sorbent
performance is evaluated under various test conditions, e.g., bed
temperature, gas/liquid flow rate, target species concentration,
bed interaction time (residence time of sorbent bed in contact
with the target species).
4. CAPTURE OF GASEOUS IODINE WITH

SULFIDE AEROGEL COMPOSITES
Capturing iodine gas [I2(g)] is one of the greatest concerns associated

with reprocessing of used nuclear fuels because one of the most prominent
iodine isotopes, 129I, has a very long half-life (t½) of 1.6 × 107 y (ß-emitter)
and because it can incorporate into the human metabolic process through the

thyroid, potentially causing cancer in affected individuals [34-36]. Methods

for the capture and immobilization of gaseous iodine have been investigated
for decades and generally include alkaline scrubbing options (e.g.,
hydroxides) and solid sorbents; a more comprehensive review is provided
elsewhere [37].
The most promising technologies are porous sorbents that use a metal
getter to chemisorb the I2(g) into a metal-iodide complex. Typical metals
used for gettering I2(g) include Ag, Cd, Cu, Hg, Mn, Pb, Pd, and Tl [38-43].
Some of the more common sorbents evaluated to date include silver-loaded
ceramics (e.g., AgZ [44], silver faujasite (AgX) [45]) and silver-loaded
aerogels (e.g., Ag0-functionalized silica aerogel [46], silver-loaded
aluminosilicate aerogels [47]).
In work done by Riley et al. [6], sulfide-based aerogels (or chalcogels)
were used to capture I2(g). The specific chalcogel composition used for this
study was Sn2S3, but several others have been demonstrated to capture I2(g)
and could be used to make composite sorbents as well including PtGe2S5
[48, 49], (CoNi)MoS4 [49], NiMoS4 [50], CoMoS4 [50], Sb4Sn3S12 [50],
Zn2Sn2S6 [50], and K0.16CoSx (x = 4−5) [50]. For this particular study, three
different composites (i.e., SnS33, SnS50, and SnS70) were synthesized with
the chalcogel masses (mCg), PAN masses (mPAN), and DMSO volumes
(VDMSO) listed in Table 2. The general process used is shown in Figure 1 and
was described previously.
Since PAN is a lightly colored polymer, it was found that the easiest
method for visualizing the porosity of these materials was through analysis
of cross-sectional views of the granular product with SEM. The PAN matrix
is generally soft, so the granular product can be easily cut in half for cross-
sectional polishing. Prior to SEM analysis, the granular products were
mounted in resin and polished. The general pore structure of the three
composite sorbents along with a pure PAN granule are shown in Figure 2.
Through the contrast of backscattered electrons, the active sorbent shows up
much brighter than the PAN matrix, and the increased concentrations of
chalcogel can be easily observed through the increases in brightness as SnS33
< SnS50 < SnS70. Also, it is apparent from Figure 2 that the pore structure
does collapse (fewer open pores) as the active sorbent concentration within

the composite increases, in the order of SnS70 < SnS50 < SnS33. Figure 2 also
provides visual comparisons between the as-made and iodine-loaded
composites; since the SnS70 composite was not iodine loaded using the same
approach as the others, it was not shown in the comparison collage. The as-
made PAN granules were white. The as-made SnS33, SnS50, and SnS70 were
light tan, medium gray, and dark gray, respectively. For the iodine-loaded
sorbents, the PAN+I, SnS33+I, and SnS50+I were light yellow, light brown,
and dark purple, respectively.
Table 2. Recipes for making PAN-chalcogel composite sorbents
Sample mCg mPAN mTotal m%Cg mCg: VDMSO

(g) (g) (g) (mass%) mPAN (mL)
SnS33 0.20 0.40 0.60 33 0.33 6.00
SnS50 0.13 0.13 0.26 50 0.50 2.50
SnS70 0.19 0.08 0.27 70 0.70 1.20
“m,” “m%,” “V,” “Total,” and “Cg” denote mass, mass%, volume, chalcogel+PAN, and chalcogel,
respectively
Figure 2. (a) Scanning electron micrographs of pure PAN form (“PAN”) along with
Sn2S3-loaded PAN (SnS-PAN) at 33 mass% Sn2S3 (i.e., SnS33), 50 mass% Sn2S3 (i.e.,
SnS50), and 70 mass% Sn2S3 (i.e., SnS70) in cross-sectional views. Photographs of (b)
as-made and (c) iodine-loaded composites in full form views.

The maximum iodine uptake was determined under static conditions by

placing granules of the different samples into pre-tared glass vials, putting a
known mass of iodine (99.9999%) into a separate pre-tared glass vial,
loading these vials (along with a blank vial) into a vacuum desiccator,
evacuating the desiccator, and periodically breaking vacuum to weigh all of
the vials. The instantaneous iodine mass uptake (i.e., mI) of the i-th sampling
interval (i.e., mI,i) was monitored as the difference between the total mass
after sitting in the desiccator in the presence of iodine (i.e., mCg+I,i) and the
starting sample mass (i.e., mCg) using Equation (1) (all accounting for the
tare mass of each vial). The cumulative iodine uptake (i.e., mI) is defined as
the difference between the final sample mass containing iodine after the
experiment (i.e., mCg+I) and mCg, as shown in Equation (2). Additionally, the
mass% of iodine in the final iodine-loaded specimen can be calculated either
incrementally at different intervals using Equation (3) as m%I,i for the i-th
interval, or at the end of the experiment using Equation (4) as m%I.
mI,i = mCg+I,i – mCg (1)
mI = mCg+I – mCg (2)
m%I,i = 100 × (mI,i) / mCg+I,i (3)
m%I = 100 × (mI) / mCg+I (4)
Figure 3 provides some comparisons of the iodine uptake in the different

chalcogel-PAN composite sorbents. Equation (3) was used to calculate the
incremental mass uptake as a function of time for the PAN, SnS33, SnS50,
and SnSp (PAN-free) sorbents, and these data are plotted in Figure 3a with
the final maximum iodine concentrations, which were calculated with
Equation (4), presented as the numbers shown in the boxes to the right of
each dataset. Figure 3b shows the utilization of the sorbent in terms of iodine
loading (i.e., mass or moles) as a function of Sn2S3 mass and moles of Sn in
the composite. Figure 3c shows the iodine concentration in the final
composite after maximum iodine uptake as a function of PAN concentration

in the composite; in the comparison are the pure chalcogel powder (SnSp+I)
and the PAN granules without chalcogel (PAN+I). Figure 3d shows the mass
of iodine uptake as a function of chalcogel mass in the starting sorbent. Here,
it can be seen from the 1:1 and 2:1 loading trend lines that the iodine loading
is closer to 2:1 (by mass) of iodine per mass of Sn2S3 that is present within
the composite.
Figure 3. (a) Summary of maximum iodine uptake experiments in a vacuum desiccator

using powdered chalcogel (i.e., SnSp; see powder in Figure 1), two PAN-chalcogel
composites (i.e., SnS33 and SnS50), and pure PAN (i.e., PAN) granules for comparison.
Values are shown as the mass% of iodine in the final form. (b) Comparison of Sn
utilization including I/Sn2S3 (by mass) and I/Sn (by mole). (c) Summary of iodine
uptake (in mass% of total mass after capture) as a function of PAN content in the
sorbent (the PAN-free sorbent, SnSp, and pure PAN are included for comparison). (d)
Summary of iodine mass loading as a function of Sn2S3 mass in the composite along
with 1:1 and 2:1 loading lines (mass basis) for comparison of sorbent utilization. Part
A of this figure was modified from the original by Riley et al. [6] and reprinted with
permission from the American Chemical Society.

5. CAPTURE OF XENON AND KRYPTON
The capture of Xe and Kr present in nuclear fuel reprocessing off-gas

streams is necessary to mitigate their release to the environment. Some of
the more common technologies historically used include cryogenic
distillation [51-55] and solid sorbents such as activated carbon (or charcoal)
[56], metal-organic frameworks (MOFs) [57-59], metal-impregnated
zeolites [60], and chalcogels [61]. In recent work by Garn, Greenhalgh, and
Law, selective volatile noble gas (i.e., Xe and Kr) capture was demonstrated
with both HZ-PAN [8] and AgZ-PAN [10]; these are described in more
detail below. Xenon exhibits characteristics useful to numerous commercial
applications [62]. If Xe can be selectively partitioned and purified, an
economic supply to the commercial sector is anticipated. Examples of
industry uses for xenon include (1) lamp manufacturing, which consumes
60–80% of the available quantities, (2) computerized tomography, (3)
anesthesiology, (4) lasers, and (5) the space industry.
5.1. HZ-PAN for Krypton Capture
An HZ-PAN sorbent was fabricated by mixing 10 g CBV90A hydrogen

mordenite (Zeolyst International) and 2 g 3.3 dtex semi-dull PAN fibers
(Dralon GmbH) in 30 mL DMSO at 80°C. Spherical droplets were formed
using a glass apparatus fitted with an air-assisted nozzle design. The
homogeneous mixture was poured into the apparatus and gravity fed to the
air-assisted nozzle, where droplets were formed and dropped into an agitated
nanopure water bath, which removed the water-soluble solvent to form
spherical granules. The granules were dried in a vacuum oven overnight at
60°C. While the SSA of the starting HZ sorbent was 493 m2 g-1, that of the
final HZ-PAN sorbent was 336 m2 g-1. Considering the additional mass of
PAN added, minimal loss of SSA was reported during the HZ-PAN sorbent
preparation.
The Kr capacity of the HZ-PAN was then evaluated using a cryostat
system for thermal-swing adsorption operations. Prior to sorbent capacity

measurements, the HZ-PAN was activated at 150°C using a He flow of 50

standard cm3 min-1 (sccm) for 12 h to remove any entrained water. Then, the
column containing the HZ-PAN sorbent was cooled to −82°C while flowing
He across the bed. A feed gas of 150 µL Kr L-1 (in He balance) was
introduced at 50 sccm, using a mass flow controller to obtain the desired Kr
concentration ranging from 75–2544 µL L-1. Using a Langmuir equilibrium
isotherm fit for the experimental dataset, the model had the following form:
𝑞max 𝐾eq 𝐶
𝑞= 1+𝐾eq 𝐶
(5)
where q is the sorbent capacity at equilibrium with feed concentration (C),

Keq is the Langmuir adsorption equilibrium constant, and qmax is the
predicted maximum capacity of the sorbent at the specified temperature.
Figure 4 shows a Kr adsorption isotherm. The Kr capacity was determined
for a series of 24 breakthrough tests, between which the sorbent was
thermally cycled up to 100°C for desorption, with capacity values ranging
from 100–112 mmol kg-1 (baseline conditions). The Langmuir parameters,
Keq and qmax, were determined to be 23.4 m3 mol-1 and 0.94 mol kg-1,
respectively.
Figure 4. Kr adsorption isotherm showing Kr capacity as a function of feed

concentration for HZ-PAN composite (circles) along with the Langmuir prediction
(dotted line) [8].

Figure 5. SEM micrographs HZ-PAN granules depicting (a) an exterior view of a

whole granule, (b) a cross-sectional view of a granule, and (c) a higher-magnification
view of the cross section.
In addition to capacity evaluations, thermal cycling effects were also

evaluated to determine the robustness of the sorbent and, more importantly,
the ability of the PAN to withstand aggressive temperature conditions. The
consistent baseline condition capacity measurements obtained throughout
the course of the evaluation gave strong indication the HZ-PAN composite
was unaffected by thermal cycling. SEM micrographs showing the porosity
of the exterior and interior of the HZ-PAN composites are shown in
Figure 5.
5.2. AgZ-PAN for Xenon Capture
Research findings for silver-exchanged zeolite materials suggest a

strong adsorptive property, presumably due to a strong interaction between
silver and Xe [63], which is possible because the large Xe atoms are
polarizable. The Xe polarization, along with highly porous structures found
within zeolites, can improve Xe capture, reducing the need for wide-range
thermal-swing operations. The AgZ-PAN composite sorbent provides the
highly porous structure, allowing for strong Ag-Xe interactions, even at
temperatures approaching ambient.

The AgZ-PAN sorbents were made following procedures similar to

those described previously for the HZ-PAN. For the AgZ-PAN sorbent,
some synthesis parameters were varied in an attempt to produce a composite
sorbent with high SSA; these included converting the sodium mordenite
(NaZ) to AgZ before making the PAN composite with (a) DMSO, (b)
dimethylacetamide, and (c) 1-methyl-2-pyrrolidinone as the solvents; and
(d) making the composite with NaZ followed by exchanging the Na for Ag
within the composite by soaking the NaZ-PAN composite in 0.1 M silver
nitrate (AgNO3) solution for 24 h at 90°C under mild agitation to obtain
AgZ-PAN. The final synthesis method resulted in an acceptable, high-SSA
material with a ~10 mass% Ag loading.
The AgZ-PAN demonstrated a selectivity for Xe > Kr over a
temperature range of ambient to 240°C. Selectivities (S) were calculated
using Equation (6), where XXe and XKr are molar fractions of Xe and Kr in
the adsorbed phase (capacities), and YXe and YKr are molar fractions of Xe
and Kr in the bulk phase (feed gas concentrations 1020 ppm Xe and 150 ppm
Kr, by volume; ppm = parts per million), respectively. Using the capacity
values from Table 3, the selectivities at ambient conditions for Xe over Kr
were 7.8 in the He balance gas and 6.9 in the air balance test gas using the
minimum determinable krypton capacities.
X /Y
SXe,Kr = 𝑋 Xe /𝑌Xe (6)
Kr Kr
Table 3. Experimental results from AgZ-PAN sorption in different

Xe-Kr feed gas mixtures under a 50 sccm flow rate
Feed gas T Xe capacity Kr capacity

composition (±1 K) (mmol kg-1) (mmol kg-1)
Xe-Kr-Air 296 27.8 0.6*
Xe-Kr-He 295 31.6 0.6*
Xe-He 296 31.5 N/A
“*” denotes minimum determinable capacity; “N/A” denotes not applicable
Thermal cycling effects were also evaluated for the AgZ-PAN

composite similar to those for HZ-PAN. After numerous thermal cycles with

intermittent baseline testing, the composites experienced no decline in SSA

or capacities throughout the testing. The results from the HZ-PAN and AgZ-
PAN testing indicates that the PAN-composite structures experience no
deleterious effects from extensive thermal cycle capacity tests.
6. REMOVAL OF CONTAMINANTS FROM WATER
The purification of water supplies (i.e., drinking water or industrial

streams) from toxins generated during industrial activities including
agriculture, mining, semiconductors, batteries, and petroleum production is
a major concern throughout the world. The following sections discuss heavy
metal removal (e.g., As, Se, Sb) and radionuclide removal (i.e., cesium-137)
from aqueous streams using PAN-based composites.
6.1. Arsenic Removal from Aqueous Streams
Removing heavy metals from aqueous streams will be increasingly

important as populations across the world continue to grow. This topic has
been a focus in the scientific community for many decades. Sorbents for
heavy metal capture from aqueous streams include iron compounds [64-66],
manganese oxides [67], and chelating polymers [68]. The N-CAS composite
uses metal oxides (e.g., Fe2O3, Fe3O4), metal oxyhydroxides [e.g.,
FeO(OH)], hydrated metal oxides or hydroxides (e.g., 5Fe2O3•9H2O,
Fe5HO8•4H2O), or metal salts (e.g., chlorides, oxychlorides, sulfates,
nitrates, acetates) embedded in a PAN polymer matrix for removing toxins
such as As, Se, and Sb from aqueous streams [11, 3, 12-14]. This product is
also referred to as a PAN-HIO composite (HIO = hydrated iron oxide) [3].
The primary sorbent is 10–85 mass% of ferrihydrite (i.e., Fe3+ 2 O2 • 0.5H2 O
or Fe(III) hydrous oxide), the active binding material, with the balance being
PAN; this can be adjusted based on the application.
This material is produced as follows: (1) a metal compound is dissolved
in a solvent to create a metal solution, (2) PAN is then dissolved into the

metal solution, and (3) the mixture is placed into a quenching bath.
Depending on the type of metal being used, the quenching bath in Step 3
contains an alkaline pH achieved with 0.1–8 M NaOH (the optimal
conditions for ferrihydrite are in the range of 0.5–6 M NaOH, with 1 M being
a good starting concentration). During quenching, depending on how the
process is conducted, the morphology produced can be a granular (i.e., bead-
type) shape or a filter can be impregnated with the mixture, where the active
sorbent is homogeneously distributed throughout the composite.
For reference, the SSA of the 85 mass% ferrihydrite composite is 275–
350 m2 g-1.
The lifetime estimate for this material is that 1 L of the composite
sorbent could likely be used to purify 350,000–400,000 L of contaminated
water. When this composite was patented, it was also estimated that, at the
cost of $1.00 US, over 37,000 L of water could be treated, demonstrating the
cost effectiveness of this sorbent. A demonstration of the arsenic loadings as
a function of active sorbent concentration in the composite sorbents is
summarized in Figure 6. The sorbent can be regenerated multiple times by
soaking it in dilute caustic solution, without significant reduction in the
arsenic loading capacities [3].
Figure 6. Equilibrium arsenic-loading isotherms for N-CAS sorbent at different

ferrihydrite concentrations in the composite sorbent [11].

6.2. Phosphate Removal for Remediating Toxic Algae Blooms
If not controlled, increases in nitrogen and phosphate levels in

waterways can accelerate algae growth. Significant algae increases will
reduce oxygen levels, which can kill fish and other aquatic marine life. If
these algae growths become very large, they can be referred to as algae
blooms. These blooms are harmful to humans due to the increased toxins
and bacterial growth present in polluted waters. Contact of these blooms
with humans, including activities such as swimming, drinking contaminated
water, or consuming tainted fish, result in sickness. In 2008, researchers at
Kansas State University research came to the conclusion that pollution
caused by the overabundance of nitrogen and phosphate in water costs the
US $4.3B annually [69]. Toxic algae blooms are problematic throughout the
world. From large population fish mortality in Chile and Texas to the large
blooms found in Lake Erie and Lake Okeechobee, these environmental
challenges post direct threats to human health and high cost restoration to
government agencies.
Global Phosphate Solutions is a company formed to combat the algae
problem [15]. They have developed a composite sorbent that has been shown
to remove inorganic/ortho-phosphates from water. The composite known as
TPS consists of a proprietary adsorbent material, which is bound in a PAN
matrix. The PAN provides a stable engineering form that prevents the loss
of the active adsorbent. This material has been installed in water treatment
facilities and has demonstrated effluent water phosphate levels below 10
ppb.
The captured phosphate can be eluted from the sorbent and resold to
commercial markets such as the fertilizer industry. After elution, the
material can be reused with no reduced capacity.
6.3. Cesium Removal from Aqueous Nuclear Waste Streams
Cesium-137 is one of the high-activity radionuclides that is released

from fission of isotopes found in both nuclear reactors and weapons. It is of

high concern for several reasons, the most prevalent being that it has a very
high activity due to its t½ of 30.2 yr, primarily by ß decay, and the high
mobility of Cs+ throughout the environment. Different technologies have
been evaluated for removing Cs from aqueous solutions, including ion-
exchange resins (e.g., resorcinol-formaldehyde, SuperLig® 644) [70, 71] and
metal-based sorbents (e.g., Fe-, Ni-, and/or Cu-based materials) [72-74].
An alternative sorbent option is a PAN-based composite that uses
an active sorbent of ammonium molybdophosphate [i.e., (NH4)3
P(Mo3O10)4•3H2O, or AMP] [17, 75, 76, 16]; this composite sorbent is
referred to as AMP-PAN and was developed at the Czech Technical
University in Prague (Czech Republic) [16]. The application discussed here
is the use of AMP-PAN for removing 137Cs+ from acidic nuclear waste
streams. The AMP-PAN sorbent is ideal for acidic nuclear waste streams
because it meets the following criteria:
1) The sorbent should be stable in acidic conditions.

2) The sorbent should be stable in oxidizing conditions.
3) The sorbent should have a high capacity for Cs+ in the presence of
other waste constituents (e.g., Na+, K+, Al3+, Fe3+) and in acidic
conditions, because the concentration of Cs+ is low in the waste
streams.
4) The sorbent should be radiation tolerant (i.e., resistant to radiolytic
damage).
5) The sorbent should be temperature tolerant (i.e., heat from
radionuclide decay).
The AMP-PAN composite uses an ion-exchange mechanism to remove

Cs+ from aqueous solutions: Cs+ ions are exchanged for NH4+ and H+ ions
from within the hollow sphere formed by (PMo12O40)3− in the AMP crystal
structure [77, 78] (see Figure 7), resulting in the release of NH4+ and H+ into
the solution [79]. In acidic solutions, AMP can also be used to ion exchange
other species including K+, Rb+, and Tl+, where some NH4+ ions always
remain no matter what type of exchange is conducted [79, 78].

Figure 7. Crystal structure of AMP according to Boeyens et al. [78] shown on N(100)
so that the cages can be seen. This is Inorganic Crystal Structure Database entry
number 212 with the chemical formula of (NH4)2.6(H3O)0.4(PMo12O40) of cubic space
group 𝑃𝑛3̅𝑚𝑧 (224).
Additional work through a collaboration with US-based Department of

Energy national laboratories demonstrated that the AMP-PAN sorbent
worked well at removing Cs+ from acidic solutions as well as basic solution,
but Cs was co-adsorbed with lanthanides and/or actinides [19].
CONCLUSION
A wide variety of polyacrylonitrile-based composite sorbents have been

produced (or envisioned) over the past few decades and are reported
throughout the literature. The compositions described here are only a small
reflection of the wide variety of sorbent types that can be produced using
this basic process, which can support a diverse range of applications for
purification of gaseous and/or liquid-based process streams.
Polyacrylonitrile is a highly resilient material that is porous, stable in a wide
range of pH values, and radiolytically stable, making it a great scaffold

material for an assortment of composite sorbents for use in numerous

process applications.
REFERENCES
[1] Paiva, M. C., Kotasthane, P., Edie, D. D., and Ogale, A. A. “UV
stabilization route for melt-processible PAN-based carbon fibers.”
Carbon 2003, 41 (7):1399-1409.
[2] Rahaman, M. S. A., Ismail, A. F., and Mustafa, A. “A review of heat
treatment on polyacrylonitrile fiber.” Polym. Degrad. Stabil. 2007, 92
(8):1421-1432.
[3] Tranter, T. J., Mann, N. R., Todd, T. A., and Herbst, R. S. High
capacity adsorption media and method of producing. Patent
US7807606B2, 2010.
[4] Šebesta, F., John, J., Motl, A., and Štamberg, K. Evaluation of
Polyacrylonitrile (PAN) as a Binding Polymer for Absorbers Used to
Treat Liquid Radioactive Wastes. SAND95-2729. Albuquerque, NM:
1995.
[5] Šebesta, F. “Composite sorbents of inorganic ion-exchangers and
polyacrylonitrile binding matrix. I. Methods of modification of
properties of inorganic ion-exchangers for application in column
packed beds.” J. Radioanal. Nucl. Chem. 1997, 220 (1):77-88.
[6] Riley, B. J., Pierce, D., A, Chun, J., Matyáš, J., Lepry, W. C., Garn, T.,
Law, J., and Kanatzidis, M. G. “Polyacrylonitrile-chalcogel hybrid
sorbents for radioiodine capture.” Environ. Sci. Technol. 2014, 48
(10):5832-5839.
[7] Garn, T. G., Law, J. D., Greenhalgh, M. R., Tranter, T. J., and Tranter,
R. Composite media for fluid stream processing, a method of forming
the composite media, and a related method of processing a fluid
stream. Patent US8686083B2, 2014.
[8] Greenhalgh, M., Garn, T. G., and Law, J. D. “Development of a
hydrogen mordenite sorbent for the capture of krypton from used

nuclear fuel reprocessing off-gas streams.” J. Nucl. Sci. Tech. 2014, 51

(4):476-481.
[9] Welty, A. K., Greenhalgh, M., and Garn, T. G. Preliminary Desorption
Studies for HZ-PAN and AgZ-PAN. NTRD-MRWRD-2017-000207,
INL/EXT-17-43231. Idaho Falls, ID: 2017.
[10] Garn, T. G., Greenhalgh, M., and Law, J. D. “Development and
evaluation of a silver mordenite composite sorbent for the partitioning
of xenon from krypton in gas compositions.” J. Nucl. Sci. Technol.
2015, 53 (10):1484-1488.
[11] Tranter, T. J., Herbst, R. S., Mann, N. R., and Todd, T. A. High
capacity adsorption media and method of producing. Patent
US7368412B2, 2008.
[12] Tranter, T. J., Mann, N. R., Todd, T. A., and Herbst, R. S. Methods of
removing a constituent from a feed stream using adsorption media.
Patent US7947861B2, 2011.
[13] Tranter, T. J., Herbst, R. S., Mann, N. R., and Todd, T. A. Methods of
using adsorption media for separating or removing constituents.
Patent US8043586B2, 2011.
[14] Mann, N. R., and Tranter, T. J. Methods of producing adsorption
media including a metal oxide. Patent US8664150B2, 2014.
[15] 2018. “Global Phosphate Solutions, 410 May St. Idaho Falls, ID
83401.” www.clearwatergps.com.
[16] Šebesta, F., and Štefula, V. “Composite ion exchanger with amm-
onium molybdophosphate and its properties.” J. Radioanal. Nucl.
Chem. 1990, 140 (1):15-21.
[17] Brewer, K. N., Todd, T. A., Wood, D. J., and Tullock, P. A. “AMP-
PAN column tests for the removal of 137Cs from actual and simulated
INEEL high-activity wastes.” Czech. J. Phys. 1999, 49:959-964.
[18] Šebesta, F., John, J., and Motl, A. Development of Composite Ion-
Exchangers and Their use in Treatment of Liquid Radioactive Wastes.
XA9744278. Prague, Czech Republic: 1996.
[19] Marsh, S. F., Svitra, Z. V., and Bowen, S. M. Distributions of 14
Elements on 63 Absorbers from Three Simulant Solutions (Acid-

Dissolved Sludge, Acidified Supernate, and Alkaline Supernate) for

Hanford HLW Tank 102-SY. LA-12654. Los Alamos, NM: 1994.
[20] Marsh, S. F., Svitra, Z. V., and Bowen, S. M. Effects of Soluble
Organic Complexants and Their Degradation Products on the
Removal of Selected Radionuclides from High-Level Waste. Part III:
Distributions of Sr, Cs, Tc, Pu, and Am onto 33 Absorbers from Four
Variations of a 3:1 Dilution of Hanford Complexant Concentrate (CC)
Simulant Solutions. Part IV: The Effects of Varying Dilution Ratios on
the Distrubtions of Sr, Cs, Tc, Pu, and Am onto 12 Absorbers from
Hanford Complexant Concentrate (CC) Simulant Solutions. LA-
13000. Albuquerque, NM: 1995.
[21] Svitra, Z. V., Marsh, S. F., and Bowen, S. M. Distributions of 12
Elements on 64 Absorbers from Simulated Hanford Neutralized
Current Acid Waste (NCAW). LA-12889. Los Alamos National
Laboratory: 1995.
[22] Marsh, S. F., Svitra, Z. V., and Bowen, S. M. Effects of Soluble
Organic Complexants and Their Degradation Products on the
Removal of Selected Radionuclides from High-Level Waste. Part II:
Distributions of Sr, Cs, Tc, and Am onto 32 Absorbers from Four
Variations of Hanford Tank 101-SY Simulant Solution. LA-12943. Los
Alamos, NM: 1995.
[23] Marsh, S. F., Svitra, Z. V., and Bowen, S. M. Distributions of 15
Elements on 58 Absorbers from Simulated Hanford Double-Shell
Slurry Feed (DSSF). LA-12863. Los Alamos, NM: 1994.
[24] Marsh, S. F., Svitra, Z. V., and Bowen, S. M. Effects of Aqueous-
Soluble Organic Compounds on the Removal of Selected
Radionuclides from High-Level Waste. Part I: Distribution of Sr, Cs,
and Tc onto 18 Absorbers from an Irradiated, Organic-Containing
Leachate Simulant for Hanford Tank 101-SY. LA-12862. Los Alamos,
NM: 1995.
[25] John, J., Šebesta, F., and Motl, A. 1999. “Application of New
Inorganic-Organic Composite Absorbers with Polyacrylonitrile
Binding Matrix for Separation of Radionuclides from Liquid
Radioactive Wastes.” In Chemical Separation Technologies and

Related Methods of Nuclear Waste Management, edited by G.R.

Choppin and M. K. Khankhasayev, 155-168. Dordrecht: Springer.
[26] Šebesta, F. 1990. “Composite Inorganic-Organic Ion Exchangers,
Their Application for Control of Radioactive Contamination of Water
and Treatment of Radioactive Waste Water.” Radionuclides and
Ionizing Radiation in Water Management, Harrachov, Czech
Republic.
[27] Nilchi, A., Saberi, R., Moradi, M., Azizpour, H., and Zarghami, R.
“Adsorption of cesium on copper hexacyanoferrate-PAN composite
ion exchanger from aqueous solution.” Chem. Eng. J. 2011, 172
(1):572-280.
[28] Kameník, J., and Ŝebesta, F. “Comparison of some commercial and
laboratory prepared caesium ion-exchangers.” Czech. J. Phys. 2003,
53 (1):A571-A576.
[29] Someda, H. H., ElZahhar, A. A., Shehata, M. K., and El-Naggar, H.
A. “Supporting of some ferrocyanides on polyacrylonitrile (PAN)
binding polymer and their application for cesium treatment.” Sep.
Purif. Technol. 2002, 29 (1):53-61.
[30] Breier, C. F., Pike, S. M., Sebesta, K. Tradd, K., Breier, J. A., and
Buesseler, K. O. “New applications of KNiFC-PAN resin for broad
scale monitoring of radiocesium following the Fukushima Dai-ichi
nuclear disaster.” J. Radioanal. Nucl. Chem. 2016, 307 (3):2193-2200.
[31] Motl, A., John, J., and Sebesta, F. “Composite absorbers of inorganic
ion-exchangers and polyacrylonitrile binding matrix.” J. Radioanal.
Nucl. Chem. 1997, 222 (1-2):205-207.
[32] Brunauer, S., Emmett, P. H., and Teller, E. “Adsorption of gases in
multimolecular layers.” J. Am. Chem. Soc. 1938, 60 (2):309-319.
[33] Gregg, S., and Sing, K. 1982. Adsorption, Surface Area, and Porosity.
2nd ed. Orlando: Academic Press.
[34] Grossman, C. M., Morton, W. E., and Nussbaum, R. H.
“Hypothyroidism and spontaneous abortions among Hanford,
Washington, downwinders.” Arch. Environ. Health: Int. J. 1996, 51
(3):175-176.

[35] Grossman, C. M., Nussbaum, R. H., and Nussbaum, F. D.

“Thyrotoxicosis among Hanford, Washington, Downwinders: a
community-based health survey.” Arch. Environ. Health: Int. J. 2002,
57 (1):9-15.
[36] Grossman, C. M., Nussbaum, R. H., and Nussbaum, F. D. “Cancers
among residents downwind of the Hanford, Washington, plutonium
production site.” Arch. Environ. Health: Int. J. 2003, 58 (5):267-274.
[37] Riley, B. J., Vienna, J. D., Strachan, D. M., McCloy, J. S., and Jerden,
J. L. “Materials and processes for the effective capture and
immobilization of radioiodine: A review.” J. Nucl. Mater. 2016,
470:307-326.
[38] Ackley, R. D., and Combs, Z. Applicability of Inorganic Sorbents for
Trapping Radioiodine from LMFBR Fuel Reprocessing Off-Gas.
ORNL/TM-4227. Oak Ridge, TN: 1973.
[39] Gal, I. J., Muk, A., Todorovic, M., Rajnvajn, J., Cvjeticanin, D.,
Vujisic, L., and First, M. W. 1974. “Adsorption of methyl iodide on
impregnated alumina.” the 13th AEC Air Clean. Conf., CONF 740807,
San Francisco, CA, 12 Aug 1974.
[40] Maeck, W. J., and Pence, D. T. 1970. “Application of the metal
zeolites to radioactive air cleaning problems.” 11th AEC Air Cleaning
Conference, CONF-700816, Richland, WA, 31 Aug 1970.
[41] Pence, D. T., Duce, F. A., and Maeck, W. J. 1970. “A study of the
adsorption properties of metal zeolites for airborne iodine species.”
11th AEC Air Cleaning Conf., CONF 700816, Richland, WA,
September 1970.
[42] Pence, D. T., Duce, F. A., and Maeck, W. J. 1972. “Developments in
the removal of airborne iodine species with metal-substituted
zeolites.” 12th AEC Air Cleaning Conf., CONF 720823, Oak Ridge,
TN, 28 Aug 1972.
[43] Thomas, T. R., Murphy, L. P., Staples, B. A., and Nichols, J. T.
Airborne elemental iodine loading capacities of metal zeolites and a
method for recycling silver zeolite. ICP-1119. Idaho Falls, ID: 1977.
[44] Jubin, R. T. 1982. “Organic Iodine Removal from Simulated Dissolver
Off-Gas Systems Using Partially Exchanged Silver Mordenite.” 17th

DOE/NRC Nucl. Air Cleaning Conf., CONF-820833, Denver, CO,

August 3, 1982.
[45] Wilhelm, J., and Puppe, L. Process for removal of iodine and iodine
compounds from gases and vapours with silver containing zeolite X.
Patent US5075084A, 1989.
[46] Matyáš, J., Fryxell, G. E., Busche, B. J., Wallace, K., and Fifield, L.
S. 2011. “Functionalized silica aerogels: advanced materials to capture
and immobilize radioactive iodine.” Ceramic Materials for Energy
Applications: Ceramic Engineering and Science Proceedings, Volume
32, Daytona Beach, FL.
[47] Riley, B. J., Kroll, J. O., Peterson, J. A., Matyáš, J., Olszta, M. J., Li,
X., and Vienna, J. D. “Silver-Loaded Aluminosilicate Aerogels As
Iodine Sorbents.” Environ. Sci. Technol. 2017, 9 (38):32907-32919.
[48] Riley, B. J., Chun, J., Ryan, J. V., Matyáš, J., Li, X. S., Matson, D. W.,
Sundaram, S. K., Strachan, D. M., and Vienna, J. D. “Chalcogen-based
aerogels as a multifunctional platform for remediation of radioactive
iodine.” RSC Adv. 2011, 1:1704-1715.
[49] Riley, B. J., Chun, J., Um, W., Lepry, W. C., Matyáš, J., Olszta, M. J.,
Li, X., Polychronopoulou, K., and Kanatzidis, M. G. “Chalcogen-
based aerogels as sorbents for radionuclide remediation.” Environ. Sci.
Technol. 2013, 47 (13):7540-7547.
[50] Subrahmanyam, K. S., Sarma, D., Malliakas, C. D.,
Polychronopoulou, K., Riley, B. J., Pierce, D. A., Chun, J., and
Kanatzidis, M. G. “Chalcogenide aerogels as sorbents for radioactive
iodine.” Chem. Mater. 2015, 27 (7):2619-2626.
[51] Baetsle, L. H., and Broothaerts, J. 1977. “Reprocessing Off-Gas
Treatment Research in Belgium.” In Seminar on Radioactive Effluents
from Nuclear Fuel Reprocessing Plants. Karlsruhe, Germany: Schule
für Kerntechnik Kernforschungszentrum.
[52] Chesne, A., Miquel, P., Goumondy, J. P., and Leseur, A. 1977.
“Progress Report on French Research and Development
in the Treatment of Off-Gases from Reprocessing Plants.” In Seminar
on Radioactive Effluents from Nuclear Fuel Reprocessing Plants.
Karlsruhe, Germany: Schule für Kerntechnik Kernforschungszentrum.

[53] Henrich, E., and von Ammon, R. “Das Konzept Der Abgasreinigung
in Wiederaufarbeitungsanlagen.” Atomkernenergie/Kerntechnik 1985,
46 (2):81-88. ["The concept of waste gas purification in reprocessing
plants." Atomic nuclear energy / nuclear technology]
[54] Laser, M., Beaujean, H., Bohnenstingl, J., Filss, P., Heidendael, M.,
Mastera, S., Merz, E., and Vygen, H. 1973. “Off-Gas Treatment and
Krypton Disposal in HTGR-Fuel Element Reprocessing.” Symposium
on the Management of Radioactive Wastes from Fuel Reprocessing,
Paris, France.
[55] Yarbro, O. O., Mailen, J. C., and Stephenson, M. J. 1977. Retention of
Gaseous Isotopes. Vienna, Austria: International Atomic Energy
Agency. IAEA International Conference on Nuclear Power and Its
Fuel Cycle, Salzburg, Austria, May 2-13, 1977.
[56] Ringel, H. D., Brodda, B. G., and Burbach, T. 1990. “Development of
a process for adsorptive separation of Kr-85 from off-gas of nuclear
facilities.” Proceedings of the 21st DOE/NRC Nuclear Air Cleaning
Conference. NUREG/CP-0016-Vol. 2.
[57] Wang, C., Liu, D., and Lin, W. “Metal–organic frameworks as a
tunable platform for designing functional molecular materials.” J. Am.
Chem. Soc. 2013, 135 (36):13222-13234.
[58] Lu, W., Wei, Z., Gu, Z.-Y., Liu, T.-F., Park, J., Park, J., Tian, J.,
Zhang, M., Zhang, Q., Gentle Iii, T., Bosch, M., and Zhou, H.-C.
“Tuning the structure and function of metal-organic frameworks via
linker design.” Chem. Soc. Rev. 2014, 43 (16):5561-5593.
[59] Nouar, F., Eubank, J. F., Bousquet, T., Wojtas, L., Zaworotko, M. J.,
and Eddaoudi, M. “Supermolecular building blocks (SBBs) for the
design and synthesis of highly porous metal-organic frameworks.” J.
Am. Chem. Soc. 2008, 130 (6):1833-1835.
[60] Deliere, L., Coasne, B., Topin, S., Gréau, C., Moulin, C., and
Farrusseng, D. “Breakthrough in xenon capture and purification using
adsorbent-supported silver nanoparticles.” Chem. Eur. J. 2016,
22:9660-9666.
[61] Subrahmanyam, K. S., Spanopoulos, I., Chun, J., Riley, B. J.,
Thallapally, P. K., Trikalitis, P. N., and Kanatzidis, M. G. “Chalco-

genide aerogels as sorbents for noble gases (Xe, Kr).” ACS Appl.
Mater. Interfaces 2017, 9 (39):33389-33394.
[62] Hammarlund, N. “The krypton and xenon markets up to the year
2000.” Nucl. Inst. Meth. Phys. Res. Sect. A 1992, 316 (1):83-87.
[63] Grosse, R., Gedeon, A., Watermann, J., Fraissard, J., and Boddenberg,
B. “Adsorption and 129Xe n.m.r. of xenon in silver-exchanged Y
zeolites: Application to the location of silver cations.” Zeolites 1992,
12 (8):909-915.
[64] Ghosh, M. M., and Yuan, J. R. “Adsorption of inorganic arsenic and
organoarsenicals on hydrous oxides.” Environ. Prog. 1987, 6 (3):150-
157.
[65] Driehaus, W., Jekel, M., and Hildebrandt, U. “Granular ferric
hydroxide - A new adsorbent for the removal of arsenic from natural
water.” J. Water SRT Aqua 1998, 47 (1):30-35.
[66] Lackovic, J. A., Nikolaidis, N. P., and Dobbs, G. M. “Inorganic arsenic
removal by zero-valent iron.” Environ. Eng. Sci. 2000, 17 (1):29-39.
[67] Driehaus, W., Seith, R., and Jekel, M. “Oxidation of arsenate(III) with
manganese oxides in water treatment.” Water Res. 1995, 29 (1):297-
305.
[68] Ramana, A., and Sengupta Arup, K. “Removing selenium(IV) and
arsenic(V) oxyanions with tailored chelating polymers.” J. Environ.
Eng. 1992, 118 (5):755-775.
[69] 2008. “Freshwater Pollution Costs US At Least $4.3 Billion A Year”
ScienceDaily, November 17, 2008. Kansas State University. www.
sciencedaily.com/releases/2008/11/081112124418.htm.
[70] Hassan, N. M., and Adu‐Wusu, K. “Cesium removal from Hanford
tank waste solution using resorcinol‐formaldehyde resin.” Solv. Extr.
Ion Exch. 2005, 23 (3):375-389.
[71] Hassan, N. M., King, W. D., McCabe, D. J., Hamm, L. L., and
Johnson, M. E. “SuperLig 644 equilibrium sorption data for cesium
from Hanford tank waste supernates.” J. Radioanal. Nucl. Chem.
2002, 253 (3):361-367.

[72] Orechovská, J., and Rajec, P. “Sorption of cesium on composite

sorbents based on nickel ferrocyanide.” J. Radioanal. Nucl. Chem.
1999, 242 (2):387-390.
[73] Lin, Y., Fryxell, G. E., Wu, H., and Engelhard, M. “Selective sorption
of cesium using self-assembled monolayers on mesoporous supports.”
Environ. Sci. Technol. 2001, 35 (19):3962-3966.
[74] Milyutin, V. V., Kononenko, O. A., Mikheev, S. V., and Gelis, V. M.
“Sorption of cesium on finely dispersed composite ferrocyanide
sorbents.” Radiochem. 2010, 52 (3):281-283.
[75] Tranter, T. J., Herbst, R. S., Todd, T. A., Olson, A. L., and Eldredge,
H. B. “Evaluation of ammonium molybdophosphate-polyacrylonitrile
(AMP-PAN) as a cesium selective sorbent for the removal of 137Cs
from acidic nuclear waste solutions.” Adv. Environ. Res. 2002, 6
(2):107-121.
[76] Miller, C. J., Olson, A. L., and Johnson, C. K. “Cesium absorption
from acidic solutions using ammonium molybdophosphate on a
polyacrylonitrile support (AMP-PAN).” Separ. Sci. Technol. 1997, 32
(1-4):37-50.
[77] Van, J., and Smit, R. “Ammonium salts of the heteropolyacids as
cation exchangers.” Nature 1958, 181:1530-1531.
[78] Boeyens, J. C. A., McDougal, G. J., and Van R. Smit, J.
“Crystallographic study of the ammonium/potassium 12-moly-
bdophosphate ion-exchange system.” J. Solid State Chem. 1976, 18
(2):191-199.
[79] Buchwald, H., and Thistlethwaite, W. P. “Some cation exchange
properties of ammonium 12-molybdophosphate.” J. Inorg. Nucl.
Chem. 1958, 5 (4):341-343.

Chapter 4
DEWETTING BEHAVIOR OF LIQUID

METAL FILM ON NANOPILLARED
GRAPHENE AND ITS EFFECT ON
THE COALESCENCE DYNAMICS
Hui Li, Tao Li, Yanyan Jiang and Xiongying Li

Key Laboratory for Liquid-Solid Structural Evolution
and Processing of Materials, Ministry of Education,
Shandong University, Jinan, People’s Republic of China
ABSTRACT
Dewetting has attracted considerable attention in recent years because

of its significant importance in many applications, including anti-ice, anti-
fog, self-clean and waterproof. In this chapter, evidence that the
nanopillared graphene surface can induce strong dewettability has been
reported by molecular dynamics (MD) simulations. The retracting and
detaching behaviors of the liquid film can be effectively controlled by
tuning the geometric parameters of the liquid film or nanopillar, as well as

Corresponding Author Email: lihuilmy@hotmail.com.

150 Hui Li, Tao Li, Yanyan Jiang et al.
the temperature. Additionally, the dewetting-induced coalescence

dynamics between liquid metals is also discussed. This work provides a
comprehensive view of the molecular-level interactions involved in
dewetting and promotes our understanding on how to control liquid
dewetting behavior using nanopillared structure, which offers potential in
a wide range of applications, especially in the fabrication of self-cleaning
materials.
Keywords: molecular dynamics simulation, dewetting, nanopillared

structure, liquid metals, coalescence dynamics
1. INTRODUCTION
The control of droplet dynamics is widely recognized as one of the

quintessential technologies for applications in the manipulation of science
and engineering, especially in the drop wettability, which has attracted a lot
of attention [1, 2]. Wettability is considered as an essential property of a
solid surface, and thus plays an important role in the areas of industry and
agriculture, as well as in our daily life. It has been reported that the desired
wettability can be obtained by modifying solid surfaces with micro/nano-
structures [3, 4] or chemically modifying surfaces to alter their surface
energy [5, 6]. In nature, the vertical climbing ability of geckos [7-9] and the
self-cleaning ability of lotus leaves [10] are all related to the unique micro-
and nanostructures on their surfaces. Inspired by these interesting animals,
functional surfaces with special wettability have attracted great research
interest and have been synthesized by various methods [11-13]. For
example, the designed superhydrophobic surfaces with a contact angle larger
than 150 are fabricated to remove contamination, corrosive, and ice. During
these processes, the dewetting behavior always occurs.
Due to the potential applications in fabricating self-cleaning materials,
much work has been conducted to investigate the dewetting behavior of
liquids, including polymer films, [14-17] water, [18-20] and metal films [2,
21-23]. It is believed that the deformed liquid droplets could convert their
surface deformation energy into kinetic energy, which makes them contract

Dewetting Behavior of Liquid Metal Film … 151
to a sphere shape and finally detach from the non-wettable substrate. During
this process, the dewetting dynamics undergoes two distinct stages: the
retraction of film and the detachment of droplet, in which the inertia plays a
dominant role [2, 24]. Studies have confirmed that the dewetting begins at
the edges or corners of films where the surface tension is high [25], so the
shape of initial film and its contact angle (CA) must have a strong effect on
the dewetting dynamics. Recently, Fuentes-Cabrera et al. [26] have
investigated the dewetting-induced ejection of nanoscale liquid copper
nanostructures that are in the shape of a circle, square, equilateral, and
isosceles triangle deposited on a graphitic substrate, and pointed out that the
ejected nanodroplets showing different detachment velocities depend on
their initial shape, which is ascribed to the temporal asymmetry of the mass
coalescence during the droplet formation. Additionally, other influence
factors, such as the size, temperature and composition of the nanodroplets,
[21, 27-30] as well as the intensity of laser irradiation that provides energy
for melting [2, 22, 31-33], also have a great effect on the dewetting
dynamics. It is worth noting that although there has been much progress in
the understanding of dewetting, most existing studies mainly concentrate on
the effect of droplet itself. However, how the surface structure of the
substrate affects the dewetting dynamics has gotten less attention, and how
to use modified substrates to control the dewetting-induced detachment of
nanodroplets is also poorly understood.
From a practical point of view, the study of dewetting behavior can pave
the way for self-assembly nanotechnology [34-37] which may involve the
drop coalescence behavior. The phenomenon of coalescence where two
liquid droplets merge to form a bigger one exhibits a range of incredible
behavior that is important to understand from the theoretical viewpoint as
well as the practical aspect. The dynamics of coalescing drops is central for
a whole host of processes, including micro/nanofluidic devices or
technologies, [38, 39] inkjet printing, [40, 41] spray coating, [42] viscous
sintering, [43] and the raindrop formation in clouds [44]. Therefore, the
detailed physical understanding of the coalescence dynamics can provide
significant guidance for controlling the drop movement. As of yet, although
the scaling law about the growth of liquid bridge during coalescence

dynamics has been well developed [45-47], the study of the dewetting-
induced coalescence dynamics is severely rare. Motivated by this and
inspired by the amazing wettability of the nanofabricated surfaces
mentioned above, molecular dynamics (MD) simulations are performed to
explore the dewetting behavior of metallic thin films on pillared graphene
(PG) and how the dewetting process affects coalescence dynamics, with the
aim of controlling the dynamics of liquid films by tuning the sizes of
nanopillars.
2. SIMULATION METHOD
MD simulations are carried out to study the dewetting and coalescence

behaviors of liquid metal films on PG substrates, using the large-scale
atomic/molecular massively parallel simulator (LAMMPS) package in the
NVT ensemble (The number of particles N, volume V, and temperature T
are constant). The temperature is held by the Nose–Hoover thermostat
[48, 49]. The time integration of Newton’s equation of motion is calculated
by the velocity Verlet algorithm [50] with a time step of 1.0 fs. To improve
the computational efficiency, all the substrates are fixed during simulation
progress. Periodic boundary conditions are applied in the x, y and z
directions.
As shown in Figure 1, the capped carbon nanotubes (CNTs) are
perpendicularly placed on the graphene (G) to form the pillared graphene
(PG) to mimic the lotus leaves which possess superhydrophobicity and
produce the dewetting behavior. The liquid metal film, extracted from a bulk
liquid that is melted to 1500K and relaxed for 500ps, is deposited at a
distance of 3.0 Å above the PG surface. Interactions between metal atoms
are described by an embedded atom method (EAM) potential, [51, 52] which
can be written as:
1
Ei  F (   (rij ))    (rij )
j i 2 j i

where F is the embedding energy,  is a pair potential interaction. C-C

interactions are modeled by an adaptive intermolecular reactive empirical
bond order (AIREBO) potential [53]. Due to the fact that metal and carbon
can only form soft bonds via the charge transfer from the π electrons in the
sp2 hybridized carbon to the empty 4s states of metal, [54] the 12-6 Lennard-
Jones (L-J) potential is utilized to describe the interactions between metal
and carbon atoms with the following parameters: εAl-C = 0.0309eV, σAl-C =
3.422 Å, [55] εPb-C = 0.01751eV, σPb-C = 3.288 Å [56](the Lorentz-Berthelot
combining rules [57,58] is used to calculate the Pb-C interactions [59]), εCu-
C =0.018eV, σCu-C = 3.0 Å. [60,61], εAu-C = 0.00873 eV, σAu-C = 2.99943 Å
[62]. With these parameters, we calculate the equilibrium contact angle (CA)
as 83.43° for Al, 112.12° for Pb, and 132° for Cu on the flat graphene at
1500K, which accord well with the experimentally measured contact angle
of Al (about 85°) [63], Pb (about 110°) [64] and Cu (about 140°) [34]. The
Au–C interaction described with a 12-6 Lennard–Jones (LJ) potential can be
fitted to ab initio and experimental data.
Figure 1. Schematic plot of the simulated system. The nanopillared graphene is formed
by placing the CNTs at the corners and center of the regular hexagons on the G. The
interval distance, diameter, height of pillars and the diameter and thickness of the
liquid disk are defined as a, d, h, D, H, respectively.
The CA of a liquid droplet on surfaces is calculated using the Werder

model [65]. Firstly, the isochore line of the time-averaged metal number
density being about one fourth of the bulk density is defined as liquid-air
interface. Secondly, we fit a circle to the interface, and then measure the CA.
The velocity of the liquid nanodroplet along the z direction vertical to the
substrate is defined as

dZCM
vz (t ) 
dt ,
where ZCM is the center-of-mass displacement of the nanodroplet, and t is

time. The moment when the C-metal potential energy EC-metal equals zero is
defined as the detaching time td. After detachment, the liquid droplets remain
at a constant velocity, which is defined as the detaching velocity vd.
3. DEWETTING DYNAMICS
3.1. The Size Effect of Nanopillars and Metal Films

on Dewetting Dynamics
Figure 2 shows selected snapshots of the dewetting dynamics and

potential energy distributions of liquid Cu films. It should be noted that, for
H = 3 Å, several holes form in the Cu film. These holes gradually merge into
a large hole, then decay and disappear. When H increases to 5 Å, no holes
can be observed. Compared with the G substrate, the nanodroplet can detach
from the PG surface with either the H = 3 Å or 5 Å. Like PG, another
substrate PG2 displays similar dewetting properties, as presented in Figure
2(c). The nanodroplets under such microscopic size detach from the PG and
PG2 surfaces at 41 and 51 ps, respectively, which are far smaller than those
in previous studies [2, 22].
The potential energy distributions are shown to further understand the
dewetting behaviors. It is observed that the dewetting begins at the edges of
films and holes. As the simulation proceeds, the edge retracts toward the
center of the disk, which makes mass and energy flow and finally forms a
spherical droplet when the system reaches the equilibrium. The total
potential energy, Etot, is defined as Etot = EC-C + ECu-Cu + EC-Cu, where EC-C,
ECu-Cu and EC-Cu represent the C-C, Cu-Cu and C-Cu interaction,
respectively. It is worth noting that ECu-Cu is the driving force, while the EC-
Cu is the resisting force, and the EC-C is equal to 0 eV due to the fixed

substrate. Therefore, the energy difference, ΔECu-Cu, is partially transformed

into kinetic energy that induces retraction and detachment. Figure 3 presents
theΔECu-Cu, EC-Cu and vz(t) as a function of time. In terms of those systems
without detachment, theΔECu-Cu (EC-Cu) of the Cu-PG system is much larger
(lower) than the Cu-G system, indicating that the PG exhibits a stronger
dewettability than the G, as displayed in Figures 3 (a) and (b).
By contrast, the systems with detachment present remarkably different
results. As shown in Figures 3 (d)–(f), on the one hand, it is observed that
on the PG substrate the driving forceΔECu-Cu is much larger either at the
beginning or at the equilibrium, which is one of the key factors that the
droplet could detach from the pillared surfaces. On the other hand, the
hindering force EC-Cu is much lower, which is quite beneficial for the droplet
to retract and detach. Finally, the much higher vz implies that a large
maximum vz is essential for the subsequent detachment. Therefore, from the
above analysis, we regard that low EC-Cu and high equilibriumΔECu-Cu are the
key factors for producing dewetting. The highΔECu-Cu could generate a large
detachment velocity vz, and the low EC-Cu ensures that the droplet detaches
from the substrate with a constant vz.
Figure 2. Dewetting dynamics and potential energy distributions of Cu films (D=135.6

Å) on substrates, (a) G. (b) PG, a =10.2 Å, h = 13.67 Å. (c) PG2, a =3.4 Å, h =10.92 Å.

Figure 3. ΔECu-Cu, EC-Cu and vz of the systems without (left) and with (right)
detachment.
So what plays the dominant role in retraction and detachment? To gain

insight into this question, the Reynolds number (Re), a dimensionless
quantity that gives a measure of the ratio of inertial forces to viscous forces
during the liquid film evolution, is introduced to quantify the relative
importance of these two types of forces. Re is given as
 vL
Re 
 ,
where ρ, v, L, μ are the liquid density, liquid velocity, liquid dynamic

viscosity, and characteristic linear dimension, respectively. For these
systems, Re < 6.10 indicates that the inertial force plays an important role in
determining the retraction and detachment.
The size effect of nanopillars is analyzed for the first time. As shown in
Figures 4(a) and (b), the td exhibits a nearly exponential decrease and the vd

shows a exponential increase with the increase of a. However, for the h

increasing from 13.67 to 17.22Å, both td and vd change slightly, indicating
that the dewetting properties of the droplet is not sensitive to the height of
the pillars, but is greatly affected by the interval of the pillars. It is concluded
that the arrangement density of the nanopillar on the smooth surface is even
more important than its height.
Figure 4. The size effects of nanopillars on detachment: (a) td and (b) vd.
Figure 5. The H effects on detachment: (a) td and (b) vd.
Then the height effect of liquid films (H) is shown in Figure 5. Notably,
as H increases from 5 to 10 Å, td increases monotonously, but vd slightly
decreases with some fluctuations. At every point of H, the td (vd) decreases
(increases) significantly with the increasing a. These phenomena indicate
that the dewetting dynamics of the Cu liquid film is more sensitive to a than
H. That is because increasing the H results in the increase of both the EC-Cu
and the ECu-Cu, while increasing a only leads to the decrease of the hindering
force EC-Cu. Therefore, from the analysis above, we can conclude that it is an

effective approach to control the dewetting dynamics of the liquid film by

tailing the interval distance of pillars rather than tuning the height of the
films or nanopillars.
3.2. The Effect of the System Temperature

The influence of temperature (T) on the dewetting properties is

investigated in Figure 6. Here, the Cu disk is placed on the PG substrate with
the temperature ranging from 1500 to 2000 K. It can be found that increasing
T from 1500 to 2000 K leads to a 24.44% decrement of td and 82.76%
increment of vd for a = 10.2 Å. According to the Re, the inertial force plays
an important role. Moreover, the temperature effects can also be attributed
to the enhanced inertial effects, which results from the decrease of density
and viscosity [26]. As an example, the Re is calculated as the temperature
ranges from 1500 to 1900 K and its value increases from 6.10 to 10.13,
indicating that the role of inertial effects is very obvious at high temperature.
Therefore, the results suggest that the dewetting behavior can be controlled
by adjusting the system temperature.
Figure 6. The effect of system temperature on detachment. (a) td and (b) vd. The curves
correspond to those systems with H = 10 Å, h = 13.67 Å, and a ranging from 6.8 to
10.2 Å. The dashed lines are fit curves.

3.3. The Effect of the Initial Shape of Liquid Metal Films

A series of simulations are carried out to explore how the shape of the
liquid film affects the dewetting behavior, and three kinds of shapes
including disk, square and triangle are considered. As shown in Figure 7,
during the dewetting evolution process, the Cu disk film always keeps its
initial shape when its size shrinks. Interestingly, it takes 23 ps for the square
Cu to retract to a round shape step by step, and more than 23 ps for the
triangle Cu film. The snapshot of the triangle Cu clearly shows that the
retraction at the line area is faster than that at the corner area. Importantly,
the td (62 ps) for the triangle Cu is the largest, followed by 58 ps for the
square, then by 49 ps for the round. The results strongly demonstrate that the
sharp corner cannot facilitate the retraction and detachment of the Cu film.
That is, the dewetting behavior of liquid films shows significant dependence
on their initial shapes, as is further confirmed by Figures 6 (d)–(f). In Figure
6(d), after 12 ps, the vz of the disk Cu films is the largest and reaches the
maximum of about 128 m/s at 21 ps, but the vz for the square Cu slightly
decreases, and then increases to a much lower maximum of about 114 m/s
at 24 ps. Especially, the vz of the triangle Cu greatly decreases, and then
slightly increases to the lowest maximum of about 93 m/s at 28 ps. From the
Figures 6(e) and (f), at every point of D, the td (vd) of the triangle and square
Cu is much larger (smaller) than those of the round Cu. The shape effect
may possibly be explained by the fact that the sharp corners consume more
formation energy to form a round (the lowest energy state) during the
dewetting process. So there is less energy that can be transformed into
kinetic energy. Additionally, we also note that both td and vd increase with
the increase of D, indicating that increasing D would delay the detachment
because more time is required to achieve the mass redistribution, but
increase the detachment velocity vd due to the increase of the surface
energy(on account of the increase of the ECu-Cu).

Figure 7. The initial shape effects of Cu films on detachment. The initial shape of films
are (a) disk (b) square (c)triangle, respectively, the height H is 10 Å. (d) The vz(t) of the
liquid Cu droplet when D=162.72 Å. (e) td and (f) vd as functions of D. The black, red
and blue lines represent the disk, square and triangle shape, respectively.
3.4. The Height Effect of PG Confinement

Based on the above studies, it is learned that the PG substrate can induce
the dewetting behavior. And here the droplet dynamics in the PG
confinement is discussed in order to understand how to control the dewetting
dynamics. As shown in Figure 8, the Cu liquid films with D = 108.48 Å are
placed on the bottom of the wall in the confined conditions. As the
simulation proceeds, the film retracts rapidly to form a droplet. This
contracted droplet, however, displays diverse states in the confined space
with different confinement heights (H’). When H’ = 30 Å, the retracted drop
exhibits a non-detachment state, that is, no dewetting occurs. The droplet
contacts the top wall during the retracting process. Due to the limited space,
the retracted droplet is unable to jump off the bottom wall, finally displaying
a drum-like shape and having a strong interaction with the bottom wall,
which suggests that the droplet is unable to dewet the surface. As the
confinement height increases to 50 Å, a semi-detachment state appears,
accompanied with a quasi-dewetting behavior. The retracted droplet

displays a completely different shape in which it is spherical and seems to

detach from the bottom wall. However, there exists a weak attractive force
between the drop and the bottom wall. Namely, the droplet contacts with the
bottom wall at a point, indicating that the dewetting is not normal. Therefore,
this state is defined as a semi-detachment state. When H’ = 100 Å, a full-
detachment state occurs, exhibiting a normal dewetting behavior and no
effect of confined conditions. In the second state, the retracted droplet fully
detaches form the bottom wall at a certain time and collides with the top
wall, and then it would bounce back after fully spreading and retraction on
the top wall, which also presents a clear dewetting behavior. In this case, the
confined condition actually has no effect on the dewetting dynamics. The
full-detachment process had been used by Boneberg et al. [66] to collect
droplets. Additionally, they found that the collected particle of a small size
was in a spherical shape, while the largest particle, which required the
longest time to cool, exhibited a spherical calotte shape. Therefore, the
collected droplets of different sizes finally displayed different shapes due to
the different cooling rates. At this stage, actually, the impacting droplets
should finally display spherical shapes similar to that in the semi-detachment
state due to the constant temperature of the droplet if the time is long enough.
In a word, one can control the dewetting dynamics and shape of droplets by
setting confined conditions with different height, which may provide the
potential in a wide range of applications.
To qualitatively determine the critical height among these states, a series
of MD simulations are performed on different systems, i.e., different Cu
films (D = 108.48 and 204.40 Å), different carbon walls including double
graphene(DG), vertical placed PG(V-PG) and horizontal placed PG(H-PG),
and different confined spaces (range from 10 to 240 Å). As shown in the
graphs in Figure 8, the critical height that determine three states in different
carbon-based walls are concluded. When H’≤H1,’ the retracted droplet is in
the non-detachment state, while H’≥H2,’ it is in the full-detachment state, and
there would be in the semi-detachment state when H’ is between H1’ and H2.’
It is indicated that the states of the liquid droplet can be precisely controlled
by tailoring the height of the confined space. In Boneberg’s work, the non-

detachment and full detachment states were obtained by decreasing or

increasing the laser energy density. However, discussion about the semi-
detachment state was missing in their studies, which is possibly due to the
difficulty in precisely controlling the laser energy density to obtain the semi-
detachment state. The semi-detachment state introduced here may benefit
for our further understanding the control of dewetting dynamics.
Figure 8. Snapshots of dynamics of Cu droplets in the confinement. The droplets

finally display three states, (a) non-detachment, (b) semi-detachment and (c) full-
detachment. The right two graphs show the critical heights, H1’ and H2,’ between the
three states in different systems. The final droplet displays the non-detachment state in
the region below the black line, the full-detachment state in the region above the red
line and the semi-detachment state in the region between the two lines.

4. DEWETTING-INDUCED COALESCENCE DYNAMICS
4.1. The Effect of Dewetting on the Coalescence of Homogeneous

Liquid Metals
The coalescence behavior of two identical liquids on PG substrate is

firstly studied. Here, the liquid Au is selected due to its dewetting ability on
PG surface. As shown in Figure 9(a), it can be found that some familiar
coalescence process occurs, as follows:
The generation and rapid motion of the tiny liquid bridge, a slow
rearrangement of the combined droplet shape from elliptical to more circular
with longer time. However, there also exists an abnormal process that the
Au liquid detaches from the substrate at 120 ps during the forming process
of a big spherical drop and then is suspended, quite different from that on
the other flat surfaces, which indicates the effect of dewetting properties on
the coalescence dynamics. Huang et al. [67] and Habenicht et al. [2] have
experimentally shown that Au films deposited on a substrate can be
vertically ejected under the condition of laser pulses heating. As is discussed
before, the presence of this detachment is mainly because of the obtained
kinetic energy converted from the surface deformation energy. In their
experiment, the detachment only takes place when the laser fluence is
slightly above the required energy value for the films to stay melted long
enough to retract to spheres. Therefore, it is noted that the control of laser
density is difficult due to the hysteresis of the decrease and increase of
temperature. This study would provide a new approach to remove the
superfluous or corrosive liquid by this dewetting-induced coalescence
behavior. Generally, with the help of inertia, the Au atoms on such surface
would have a strongly upward motion, which leads to the detachment of
droplets from the surface. The role of the pillars is thought to increase the
surface roughness which can significantly improve the dewettability. The
dewetting-induced coalescence behavior of the liquid Au on PG substrate is
important for industrial production such as smelting and slagging. As a case
in point, in gold ores, there may be many metal impurities, such as Ag, which
is bad for refinement. However, the different wettability of different metal

can lead to detachment of part of the metal, which is instructive for the
smelting. Figure 9 (b) shows the coalescence behavior of spherical Au drop,
which is a main process of the coalescence without retraction, meaning that
the inertia plays a dominant in dewetting dynamics. It can be observed that
the final coalesced drop still touches the top of pillars, and does not leave
the substrate. This is because the obtained drop deformation energy during
the coalescence process is less and as a result the kinetic energy is not
enough to support the bounce of the droplets. These results indicate that the
dewetting of the droplets is not only related to the surface structure but also
the liquid shape.
Figure 9. Coalescence process of two liquid Au on the PG substrate. (a) The side-view
images of Au films on PG. The droplets detach from the PG surfaces at 120 ps, while
this detachment would not appear on smooth surfaces. (b) Coalescence of two
spherical droplets on PG. The coalesced drop is in contact with the substrate, not
showing the dewetting behavior, but its contact angle is very larger.
To further illustrate how the PG substrate affects the coalescence

behavior, we show the time-evolution diffusion coefficient of Au atom in
Figure 10. It can be seen that the diffusion along the x direction plays a
dominant role in the process of two spherical drops due to the much larger
Dx than the Dy and Dz, Moreover, the time-evolution diffusion coefficient
curves of Au films early overlap on both two substrates, suggesting that the
contraction on PG is equal to that on double graphene (DG).

Figure 10. Diffusion coefficient of Au atom along different directions on different

substrates, PG and DG. (a)–(c) The diffusion coefficient for the case of two films. (d)–
(f) The diffusion coefficient for the case of two drops.
However, in terms of spherical drops, there are still quite great

differences in the diffusions along the x direction, and the diffusion speed
on the PG substrate preferentially reach to the maximum, indicating the
coalescence process is faster than that on the DG substrate. As shown in
Figures 5(d) and (e), the top of the curve on PG surface occurs nearly 100
ps prior to the DG surface.
4.2. The Effect of Dewetting on the Coalescence

of Heterogeneous Liquid Metals
The dewetting property of substrates has a significant effect on the

coalescence dynamics of two identical liquid metals, and is able to change
the final state of the coalesced drop. So what is the induced effect of
dewetting on two heterogeneous liquid metals, especially two with different
wettability. The coalescence of liquid Al and Pb is investigated. On one
hand, the wettabilty of these two kinds of liquids is different. As shown in
Figure 11(c), the wetting contact angle of Pb is larger than 90° but that of

liquid Al is smaller than 90° after reaching the equilibrium state, implying
the wettability of liquid Al on graphene is stronger than liquid Pb. Specially,
on the PG surface, the Pb film would exhibit the dewetting behavior, quite
different from the Al film whose atoms can penetrate into the interval of
nanopillars. On the other hand, as is known, liquid Al and Pb are immiscible
in liquid phase [68], which may restrict the synthesis of their alloy and the
relevant applications. Therefore, it is important to determine whether or not
liquid Al and Pb droplets can coalesce in the immiscible region and
understand the behavior of coalescence. Therefore, liquids Al and Pb are
chosen in this chapter.
The early-time coalescence evolution of Al and Pb films on DG and PG
substrates is presented in Figures 11(a) and (b). Initially, the two liquid films
are adjacent to each other and are located on the carbon-based substrate.
Once the simulation starts, these two immiscible films tend to coalesce at a
high speed and finally convert into one larger droplet. Interestingly, the
droplet forming time is quite different between two substrates. On DG
substrate, about 60 ps is required to form a new spherical droplet, while on
PG substrate, the coalescing time is approximately twice of the DG, which
is 120ps, suggesting that PG does not promote the merging, which verifies
the importance of dewetting. The side–view images of the coalescence
process are shown to explain the difference in coalescing time on two
substrates. Apparently, on DG, atoms merge into each other instantly in the
x direction. However, on PG, Pb atoms exhibit a strongly upward motion (z
direction) and a great detachment tendency from the substrate due to the
dewetting. But because there is the interaction between Al and Pb, Pb cannot
detach from the surface, and then the Pb droplet begins to blend with the Al
droplet along the x direction. Different behavior of two droplets on DG and
PG is ascribed to the different wettability of Pb films on these two substrates.
Figure 11(c) shows that the wetting contact angle of Pb is larger than 90°,
but it is less than 90 °for liquid Al, revealing that the wettability of Pb on
DG is weaker than Al. However, both of them still attach to the graphene
surface. When applied to the surface containing some microconvexities or
greater roughness as shown in Figure 11(d), the Pb droplet can detach from
the surface showing dewetting behavior. Liquid Al penetrates the spacing of

pillared structures, and does not display dewetting behavior. From this, it is
concluded that the wettable liquid that can be transformed into the dewetting
state by altering the substrate determines the coalescing process, which
reveals the importance of dewetting in coalescence dynamics.
Figure 11. Coalescing process of Al and Pb films in contact with (a) DG and (b) PG
surface (gray atoms indicate Pb atoms, yellow atoms indicate Al, and red atoms
indicate the carbon-based substrate). The drop-forming time is indicated in the figure.
(c) Singular Al and Pb film on DG at 300 ps. (d) Dewetting behavior of Pb film on PG.
The droplet detaches from the PG surface after 70 ps.
To have a better understanding of the movement of droplets during the

coalescence process and the effect of substrates, the center-of-mass
displacement of the Al and Pb droplets (CMD) as a function of time is shown
in Figure 12. Along the x direction, it is worth noting that the dash in the
figure represents the middle position between the Al and Pb films, which is
the boundary between the Al and the Pb droplets. Both on the DG and PG
substrates, all Pb droplets finally approach the Al after 600 ps, suggesting
that Pb atoms have a large displacement. Because weaker interaction
between Pb and carbon decreases the wettability of Pb compared to Al, the
movement of Pb atoms is not impeded by the substrate. It is found that the
CMDX of Al (Al-CMDX) exhibits the largest peak on DG, but there is a

valley on PG. Due to the interaction between Al and Pb, Al and Pb atoms
move towards each other.
Figure 12. Center-of-mass displacement (CMD) of the Al and Pb atoms along the x
and z directions as a function of time. The dashed line represents the middle position
between the Al and Pb films. The x and z directions correspond with the X-axis and Z-
axis in Figure 11. (a) and (c) DG, (b) and (d) PG.
The movement of Al atoms has been limited in the right direction of the
X-axis because of the restriction from Pb atoms. On DG, the surface is
smooth, so the Al drop has a greater tendency to move towards Pb with the
driving force, leading to a larger peak of Al-CMDX. In the case of PG
substrate, the Al atoms can penetrate the interval of the pillared nanotubes
and are fixed on the surface. Because of this pinning effects caused by
pillared nanotubes, the movement of the Al is restricted, hindering its ability
to move close to the Pb droplet. Subsequently, when the two metallic
droplets become one larger droplet, they begin to move to the left in the X-
axis direction because of the inertia force caused by the Pb atoms. Thus, a
valley forms in the Al-CMDX. With the curves close to each other, two
droplets become one larger drop. Overall, the substrate can significantly

affect the movement of atoms to determine the coalescence behavior,

especially the one that can induce the dewetting of droplets.
Along the z direction (CMDZ), Pb atoms exhibit a strong upward motion
on the two substrates in the beginning, giving rise to a big gap between Al-
and Pb-CMDZ. All the Pb-CMDZ show a peak, implying that there is a
distance limitation for atoms to detach from the surface due to the better
wettability of Al on graphene. However, the distance between the Al peak
and the Pb peak on the same substrate is quite different. It is about 17 Å on
DG, and almost 29 Å on PG. Obviously the PG substrate has a bigger
influence on the motion of atoms. On the DG substrate, two drops move
almost along the horizontal direction paralleled to the substrate, which is
considered as the “horizontal coalescence.” However, on the PG substrate
that can induce the dewetting behavior of Pb droplet, a “vertical
coalescence” way would occur, indicating that the PG-induced dewetting
also affects the way of coalescence dynamics.
CONCLUSION
In summary, MD simulations have been carried out to study the

dewettability of liquid metals on the nanopillared graphene. The MD results
show that the nanopillars act as the effective dewetting components, in
which the inertia force plays a dominant role. This dewetting behavior can
be controlled by simply tuning the geometrical parameters of the films and
the surface nanostructures, as well as the temperature. Additionally, the
dewetting of Cu droplets subjected to two-wall confinement is also
investigated. Three different dewetting states occur depending on the height
of confinement, including non-detachment, semi-detachment and full-
detachment, and the critical heights among these states are determined on
different substrates.
More importantly, the dewetting-induced coalescence dynamics is also
studied. For homogeneous liquid metals, the anomalies of the drop
formation on PG surface highlight that the dewettability of graphene
surfaces determines the processes of coalescence, and the coalesced drop

would detach from the surface. In terms of the heterogeneous liquid metals,
the nanopillared substrate can greatly affect the coalescence behavior of
metal films via transforming the weak wetting on DG into dewetting on PG
for Pb metal in the early-time stage, and the coalescence way is also changed
in this process.
These studies not only shed new light on understanding and controlling
the dewetting behavior on nanopillared surfaces, but also pave the way for
promising applications in some technologies, such as the fabrication of self-
cleaning materials, nanofluidic devices and homogeneous coatings.
ACKNOWLEDGMENTS
We would like to acknowledge the support from the National Natural

Science Foundation of China (Grant No. 51671114) and National Key
Research Program of China (Grant No. 2016YFB0300501). This work is
also supported by the Special Funding in the Project of the Taishan Scholar
Construction Engineering and Qilu Young Scholar Program of Shandong
University.
REFERENCES
[1] Liu, T. L. & Kim, C. J. (2014). Repellent surfaces. Turning a surface

superrepellent even to completely wetting liquids. Science 346: 1096-
1100.
[2] Habenicht, A., Olapinski, M., Burmeister, F., Leiderer, P. & Boneberg,
J. (2005). Jumping nanodroplets. Science 309: 2043-2045.
[3] Koishi, T., Yasuoka, K., Fujikawa, S. & Zeng, X. C. (2011).
Measurement of contact-angle hysteresis for droplets on nanopillared
surface and in the Cassie and Wenzel states: a molecular dynamics
simulation study. ACS nano 5: 6834-6842.

[4] Yang, C., Tartaglino, U. & Persson, B. N. J. (2006). Influence of

surface roughness on superhydrophobicity. Physical review letters 97:
116103.
[5] Feng, X. J. & Jiang, L. (2006). Design and Creation of Superwetting/
Antiwetting Surfaces. Advanced materials 18: 3063-3078.
[6] Rafiee, J., Rafiee, M. A., Yu, Z. Z. & Koratkar, N. (2010).
Superhydrophobic to superhydrophilic wetting control in graphene
films. Advanced materials 22: 2151-2154.
[7] Autumn, K. et al. (2000). Adhesive force of a single gecko foot-hair.
Nature 405: 681.
[8] Geim, A. K. et al. (2003). Microfabricated adhesive mimicking gecko
foot-hair. Nature Materials 2: 461-463.
[9] Lee, H., Lee, B. P. & Messersmith, P. B. (2007). A reversible wet/dry
adhesive inspired by mussels and geckos. Nature 448: 338-341.
[10] Barthlott, W. & Neinhuis, C. (1997) Purity of the sacred lotus, or
escape from contamination in biological surfaces. Planta 202: 1-8.
[11] Meng, S., Zhang, Z. & Kaxiras, E. (2006). Tuning solid surfaces
from hydrophobic to superhydrophilic by submonolayer surface
modification. Physical review letters 97: 036107.
[12] Xie, X. et al. (2013). Large-range control of the microstructures and
properties of three-dimensional porous graphene. Scientific reports 3:
2117.
[13] Chen, P. C. & Xu, Z. K. (2013). Mineral-Coated Polymer Membranes
with Superhydrophilicity and Underwater Superoleophobicity for
Effective Oil/Water Separation. Scientific reports 3: 2776.
[14] Reiter, G. (1992). Dewetting of thin polymer films. Physical review
letters 68: 75.
[15] Xie, R., Karim, A., Douglas, J. F., Han, C. C. & Weiss, R. A. (1998).
Spinodal Dewetting of Thin Polymer Films. Physical review letters
81: 1251-1254.
[16] Mukherjee, R., Bandyopadhyay, D. & Sharma, A. (2008). Control of
morphology in pattern directed dewetting of thin polymer films. Soft
Matter 4: 2086-2097.

[17] Saulnier, F., Raphaël, E. & De Gennes, P. G. (2002). Dewetting of thin

polymer films near the glass transition. Physical review letters 88:
196101.
[18] Jensen, T. R. et al. (2003). Water in contact with extended hydrophobic
surfaces: direct evidence of weak dewetting. Physical review letters
90: 086101.
[19] Brochard-Wyart, F. & Gennes, P. G. D. Dewetting of a water film
between a solid and a rubber.
[20] Severin, N., Lange, P., Sokolov, I. M. & Rabe, J. P. (2012). Reversible
dewetting of a molecularly thin fluid water film in a soft graphene-
mica slit pore. Nano Letters 12: 774.
[21] Müller, C. M. & Spolenak, R. (2010). Microstructure evolution during
dewetting in thin Au films. Acta materialia 58: 6035-6045.
[22] Afkhami, S. & Kondic, L. (2013). Numerical simulation of ejected
molten metal nanoparticles liquified by laser irradiation: Interplay of
geometry and dewetting. Physical review letters 111: 034501.
[23] Trice, J., Thomas, D., Favazza, C., Sureshkumar, R. & Kalyanaraman,
R. (2007). Pulsed-laser-induced dewetting in nanoscopic metal films:
Theory and experiments. Physical Review B 75: 195-204.
[24] Rack, P. D., Guan, Y., Fowlkes, J. D., Melechko, A. V. & Simpson,
M. L. (2008). Pulsed laser dewetting of patterned thin metal films: A
means of directed assembly. Applied Physics Letters 92: 59.
[25] Krishna, H., Favazza, C. & Kalyanaraman, R. S. R. (2006). Laser-
induced Dewetting Nanomorphologies in Single and Bilayer Metal
Films. Mrs Proceedings 960.
[26] Fuentes-Cabrera, M. et al. (2011). Controlling the velocity of jumping
nanodroplets via their initial shape and temperature. ACS nano 5:
7130-7136.
[27] Yang, S. et al. (2011). Template‐ Confined Dewetting Process to
Surface Nanopatterns: Fabrication, Structural Tunability, and
Structure‐ Related Properties. Advanced Functional Materials 21:
2446-2455.
[28] Krishna, H. et al. (2010). Thickness-dependent spontaneous dewetting
morphology of ultrathin Ag films. Nanotechnology 21: 155601.

[29] Kosinova, A., Kovalenko, O., Klinger, L. & Rabkin, E. (2015).

Mechanisms of solid-state dewetting of thin Au films in different
annealing atmospheres. Acta materialia 83: 91-101.
[30] Ruffino, F. & Grimaldi, M. G. (2014). Self-organized patterned arrays
of Au and Ag nanoparticles by thickness-dependent dewetting of
template-confined films. Journal of Materials Science 49: 5714-5729.
[31] Ruffino, F. et al. (2012). Formation of nanoparticles from laser
irradiated Au thin film on SiO 2 /Si: Elucidating the Rayleigh-
instability role. Materials Letters 84: 27-30.
[32] Makarov, S. V. et al. (2016). Controllable femtosecond laser‐ induced
dewetting for plasmonic applications. Laser & Photonics Reviews 10:
91–99.
[33] Fowlkes, J. D., Wu, Y. & Rack, P. D. (2010). Directed Assembly of
Bimetallic Nanoparticles by Pulsed-Laser-Induced Dewetting: A
Unique Time and Length Scale Regime. ACS Applied Materials &
Interfaces 2 (7): 2153-2161.
[34] Fuentescabrera, M. et al. (2011). Molecular dynamics study of the
dewetting of copper on graphite and graphene: implications for
nanoscale self-assembly. Phys Rev E Stat Nonlin Soft Matter Phys 83:
041603.
[35] Zhang, J., Li, Y., Zhang, X. & Yang, B. (2010). Colloidal self-
assembly meets nanofabrication: from two-dimensional colloidal
crystals to nanostructure arrays. Advanced materials 22: 4249-4269.
[36] Fowlkes, J. D., Kondic, L., Diez, J., Wu, Y. & Rack, P. D. (2011).
Self-assembly versus directed assembly of nanoparticles via pulsed
laser induced dewetting of patterned metal films. Nano Letters 11:
2478-2485.
[37] Wang, X., Feng, X., Ma, G., Yao, L. & Ge, M. (2016). Amphiphilic
Janus Particles Generated via a Combination of Diffusion‐ Induced
Phase Separation and Magnetically Driven Dewetting and Their
Synergistic Self‐ Assembly. Advanced materials 28: 3131-3137.
[38] Ahn, K. et al. (2006). Dielectrophoretic manipulation of drops for
high-speed microfluidic sorting devices. Applied Physics Letters 88:
768.

[39] Squires, T. (2005). Microfluidics: Fluid physics at the nanoliter scale.

Reviews of Modern Physics 77: 977-1026.
[40] Eggers, J. & Villermaux, E. (2008). Physics of liquid jets. Rep. prog.
phys 71: 036601.
[41] Belgardt, C. et al. (2013). Inkjet printing as a tool for the patterned
deposition of octadecylsiloxane monolayers on silicon oxide surfaces.
Physical Chemistry Chemical Physics Pccp 15: 7494-7504.
[42] Steenkiste, T. H. V. et al. (1999). Kinetic spray coatings. Surface &
Coatings Technology 111: 62-71.
[43] Bellehumeur, C. T., Bisaria, M. K. & Vlachopoulos, J. (1996). An
experimental study and model assessment of polymer sintering.
Polymer Engineering & Science 36: 2198-2207.
[44] Berry, E. X. & Reinhardt, R. L. (1978). An Analysis of Cloud Drop
Growth by Collection: Part I. Double Distributions. Journal of the
Atmospheric Sciences 31: 1814-1824.
[45] Ristenpart, W. D., Mccalla, P. M., Roy, R. V. & Stone, H. A. (2006).
Coalescence of Spreading Droplets on a Wettable Substrate. Physical
review letters 97: 064501.
[46] Eddi, A., Winkels, K. G. & Snoeijer, J. H. (2013). Influence of droplet
geometry on the coalescence of low viscosity drops. Physical review
letters 111: 144502.
[47] Case, S. C. & Nagel, S. R. (2008). Coalescence in low-viscosity
liquids. Physical review letters 100: 084503.
[48] Shinoda, W., DeVane, R. & Klein, M. L. (2007). Multi-property fitting
and parameterization of a coarse grained model for aqueous
surfactants. Molecular Simulation 33: 27-36, doi: 10.1080/089270206
01054050.
[49] Shinoda, W., DeVane, R. & Klein, M. L. (2008). Coarse-grained
molecular modeling of non-ionic surfactant self-assembly. Soft Matter
4: 2454-2462.
[50] Naidich, Y. V. & Kolesnichenko, G. (1968). A study of wetting of
diamond and graphite by fused metals and alloys VII. The effect of
vanadium, niobium, manganese, molybdenum, and tungsten on

wetting of graphite by copper-based alloys. Soviet Powder Metallurgy

and Metal Ceramics 7: 563-565.
[51] Zhao, H., Min, K. & Aluru, N. (2009). Size and chirality dependent
elastic properties of graphene nanoribbons under uniaxial tension.
Nano letters 9: 3012-3015.
[52] Klaver, T. P. C., Zhu, S.-E., Sluiter, M. H. F. & Janssen, G. C. A. M.
(2015). Molecular dynamics simulation of graphene on Cu (100) and
(111) surfaces. Carbon 82: 538-547, doi: 10.1016/j.carbon.2014.
11.005.
[53] Zhou, X. et al. (2001). Atomic scale structure of sputtered metal
multilayers. Acta materialia 49: 4005-4015.
[54] Sutter, P., Hybertsen, M., Sadowski, J. & Sutter, E. (2009). Electronic
structure of few-layer epitaxial graphene on Ru (0001). Nano letters 9:
2654-2660.
[55] Fang, R. R., He, Y. Z., Zhang, K. & Li, H. (2014). Melting Behavior
of Aluminum Nanowires in Carbon Nanotubes. The Journal of
Physical Chemistry C 118: 7622-7629, doi: 10.1021/jp410805q.
[56] Li, T., Li, J., Wang, L., Duan, Y. & Li, H. (2016). Coalescence of
Immiscible Liquid Metal Drop on Graphene. Scientific reports 6:
34074.
[57] Stuart, S. J., Tutein, A. B. & Harrison, J. A. (2000). A reactive
potential for hydrocarbons with intermolecular interactions. The
Journal of Chemical Physics 112: 6472, doi: 10.1063/1.481208.
[58] Hudson, G. & McCoubrey, J. (1960). Intermolecular forces between
unlike molecules. A more complete form of the combining rules.
Transactions of the Faraday Society 56: 761-766.
[59] Heinz, H., Vaia, R., Farmer, B. & Naik, R. (2008). Accurate simulation
of surfaces and interfaces of face-centered cubic metals using 12− 6
and 9− 6 Lennard-Jones potentials. The Journal of Physical Chemistry
C 112: 17281-17290.
[60] Deng, B., Xu, A. W., Chen, G. Y., Song, R. Q. & Chen, L. (2006).
Synthesis of copper-core/carbon-sheath nanocables by a surfactant-
assisted hydrothermal reduction/carbonization process. Journal of
Physical Chemistry B 110: 11711-11716.

[61] Li, X., He, Y., Wang, Y., Dong, J. & Li, H. (2014). Dewetting
properties of metallic liquid film on nanopillared graphene. Scientific
reports 4: 3938.
[62] Wang, L., Li, X., Zhou, X., Li, Y. & Li, H. (2016). Drop formation
and coalescence of liquid Au on nano carbon substrate. RSC Advances
6: 41053-41059.
[63] Léger, A., Weber, L. & Mortensen, A. (2015). Influence of the wetting
angle on capillary forces in pressure infiltration. Acta materialia 91:
57-69.
[64] Métois, J. & Heyraud, J. (1982). Mechanisms of morphological change
during the establishment of the equilibrium shape; Lead on graphite.
Journal of Crystal Growth 57: 487-492.
[65] T. Werder†, J. H. Walther, R. L. Jaffe, T. Halicioglu, a. & P.
Koumoutsakos † (2003). On the Water−Carbon Interaction for Use in
Molecular Dynamics Simulations of Graphite and Carbon Nanotubes.
Journal of Physical Chemistry B 107: 1345-1352.
[66] Boneberg, J. et al. (2008). Jumping nanodroplets: a new route towards
metallic nano-particles. Applied Physics A Materials Science &
Processing 93: 415-419.
[67] Huang, W., Qian, W. & El-Sayed, M. A. (2006). Gold nanoparticles
propulsion from surface fueled by absorption of femtosecond laser
pulse at their surface plasmon resonance. Journal of the American
Chemical Society 128: 13330-13331.
[68] Wang, C. P., Liu, X. J., Ohnuma, I., Kainuma, R. & Ishida, K. (2002).
Formation of immiscible alloy powders with egg-type microstructure.
Science 297: 990.

Chapter 5
THE ADVANTAGES OF THERMODYNAMIC

MODELING IN HYDROGEN PRODUCTION,
A CASE STUDY: ETHANOL STEAM
REFORMING COMBINED WITH CO2
ABSORPTION
Virginia Collins-Martinez1, Miguel Escobedo Bretado2,

Miguel Meléndez Zaragoza1, Jesús Salinas Gutiérrez1
and Alejandro López Ortiz1,
1
Centro de Investigación en Materiales Avanzados, S. C.,
Miguel de Cervantes 120, Chihuahua, Chih., México
2
Facultad de Ciencias Químicas, Universidad Juárez del Estado de
Durango, Ave. Veterinaria s/n, Circuito Universitario,
Durango, México

Corresponding Author Email: alejandro.lopez@cimav.edu.mx.

178 V. Collins-Martinez, Miguel E. Bretado, Miguel M. Zaragoza et al.
ABSTRACT
Research related to thermodynamic studies of steam reforming

reactions has been an invaluable tool for establishing an experimental basis
for further evaluation of reaction systems. This study aims to present the
advantages of using a thermodynamic analysis for the production of high
purity hydrogen by the ethanol steam reforming combined with CO2
absorption (ESR-CO2-ABS). For this purpose, a case study is presented
where the thermodynamic analysis is used to find optimal reaction
conditions for the ESR-CO2-ABS system. Thermodynamic analysis of the
ESR-CO2-ABS system was performed using CaO, CaO•MgO, Na2ZrO3,
Li2ZrO3 and Li4SiO4 as CO2 absorbents to determine optimal operating
conditions to generate high production and concentration of hydrogen in
the product gas. Conditions studied were at 300-850°C and 1 atm. Steam
to ethanol feed molar ratio (S/EtOH) was varied from 3:1 (stoichiometric)
to 6:1. Results indicate that S/EtOH ≤ stoichiometric resulted in prevention
of carbon formation. The highest production and concentration of H2 was
obtained with CaO and MgO•CaO and Na2ZrO3 at 600°C and S/EtOH =
6:1. Sodium zirconate (Na2ZrO3) as CO2 absorbent under the ESR-CO2-
ABS process scheme can be considered a promising alternative due to its
greater kinetics and stability.
1. INTRODUCTION
Nowadays, the capacity to satisfy the demand of liquid fuels depends on

different factors: a fast and immediate diversification of these, a transition
towards alternative energy vectors, where possible, and on timely measures
and actions to supply their demand [1]. Two crucial aspects that are
considered in the replacement of fossil fuels are the clean nature of the fuel
(environmentally friendly) and a high efficiency of its energy conversion. In
recent years, fuel cells have become an important option for the generation
of high efficiency electric power based on renewable hydrogen.
Currently, 95% of hydrogen production is based on the steam reforming
of natural gas [2]. However, an increase in the demand for hydrogen is
expected due to fuel cell applications that will promote the development of
new methods for the production of hydrogen, especially from renewable
sources. The use of biofuels (biogas, ethanol, etc.) for the production of

The Advantages of Thermodynamic Modeling … 179
hydrogen, through conventional processes such as steam reforming, partial

oxidation and dry reforming (currently well established in the industry) have
the advantage of a zero CO2 balance, since the carbon dioxide released to
the environment during their consumption can be re-absorbed in the process
of biomass formation [3].
Among biofuels, ethanol is considered a potential candidate for the
production of hydrogen from biomass [4]. Ethanol can be obtained from the
fermentation of various sources of biomass, including non-edible plants,
agro-industrial waste and organic matter from municipal solid waste, to
name a few [5]. It is important to point out that from ethanol it is possible to
generate a gas with a high hydrogen/carbon (H/C) molar ratio and this means
that, in principle, its conversion could produce a stream of high purity
hydrogen through the steam reforming process. In addition, the steam
reforming of ethanol, which is an oxygenated hydrocarbon, is
thermodynamically favored at lower temperatures than non-oxygenated
hydrocarbons, such as methane [6].
Ethanol steam reforming consists of a mixture of superheated steam with
alcohol according to a certain steam/ethanol molar ratio. After this gas
mixture is preheated, it is passed through tubes packed with catalyst,
commonly Ni/Al2O3. The gas mixture reacts on the catalyst to form
hydrogen and carbon dioxide according to the following reaction:
C2H6O(g) + 3H2O(g) = 2CO2(g) + 6H2(g) ΔH°298 = 171.2 kJ/mol (1)
Although this reaction looks simple, it is accompanied by a very

complex system of secondary reactions.
Generally, under the steam reforming of biofuels, the selectivity of
hydrogen is dominated and limited by the thermodynamic equilibrium of the
reaction system. For polymer electrolyte fuel cell (PEMFC) applications, the
maximum hydrogen production can be carried out in three steps. The first is
the conversion of biofuels into H2, CH4, CO and CO2 (steam reforming)
followed by a low temperature step where most of the CO produced is
converted to additional CO2 and H2 through the water-gas-shift reaction
(WGS):

CO(g) + H2O(g) = CO2(g) + H2(g)ΔH°298 = -41.2 kJ/mol (2)
The third step involves the use of high purity hydrogen and oxygen as
raw materials for a PEMFC followed by energy generation. Since the WGS
reaction is limited by the thermodynamic equilibrium, the CO concentration
after this step is still high to be used at the anode of a PEMFC where CO
concentrations greater than 50 ppm cause platinum catalyst poisoning. The
need for a high purity hydrogen stream for PEMFC applications without the
requirement of several steps, performed by conventional processes, has led
the scientific community to propose alternative methods, such as the steam
reforming combined with CO2 absorption (ESR-CO2-ABS). This ESR-CO2-
ABS process offers a promising alternative in a single step for the production
of high purity hydrogen [7]. The fundamental concept on which this
technique is based is the principle of Le Chatellier, in which the equilibrium
of the reaction can be shifted towards the production of hydrogen when CO2
gas is eliminated in situ within the reforming reactor. Thus, the carbon
dioxide generated during the steam reforming step is removed from the gas
phase using a CO2 solid absorbent, which is carbonized by trapping the
carbon dioxide within its structure. In the ESR-CO2-ABS reactor the mixing
of a CO2 absorbent (for example CaO) and a reforming catalyst will
theoretically produce a high purity hydrogen stream in a single step. The
combination of the absorption of CO2 by CaO (exothermic reaction) through
the carbonation reaction:
CaO + CO2(g) = CaCO3 ΔH°298 = -178.3 kJ/mol (3)
and the ethanol steam reforming (exothermic reaction, 2) results in a heat

balance in the ESR-CO2-ABS reactor that implies potential energy savings
with the combined use of the absorbent and the biofuel to produce H2.
However, this absorbent must be regenerated if a continuous process is
desired, since the reverse reaction (3) is highly endothermic energy, which
eventually will be required to provide additional heat to the process.
Several authors [8, 9] have described thermodynamic and experimental
studies for the production of hydrogen through steam reforming of ethanol

combined with simultaneous CO2 removal. However, the use of calcium

oxide-based sorbents produces the loss of more than half of their absorption
capacity after several absorption/regeneration cycles (normally five), with
the exception of calcined dolomite (CaO • MgO), which is capable of
withstanding several cycles with minor deterioration. The loss of capacity in
these absorbents is due to the sintering of the materials as a consequence of
the high temperature required for their regeneration (T > 900°C) [10].
Therefore, the use of an absorbent that combines low regeneration
temperatures with high performance in multi-cycles is highly desirable.
Recent studies have reported the development of synthetic CO2
absorbents. Compounds such as lithium orthosilicate (Li4SiO4), lithium
zirconate (Li2ZrO3) [11] and sodium zirconate (Na2ZrO3) [12], which are
able to withstand many carbonation/regeneration cycles without significant
loss of capacity and activity at high temperatures. Due to these
characteristics, those synthetic absorbents have become very attractive to be
used in the ESR-CO2-ABS process.
Moreover, a research related to thermodynamic studies of steam
reforming reactions has been an invaluable tool for establishing an
experimental basis for further evaluation of the reaction systems under
study. Experimentation in this area requires continuous use of specialized
equipment such as ovens with packed or fluidized tubes at high temperatures
and pressures, as well as water dosing systems for evaporation (high-
pressure pumps) within the evaluation systems. Therefore, performing
previous thermodynamic studies allows establishing ranges of reaction
variables in which it is expected to obtain satisfactory experimental results
with the premise of reducing and/or eliminating possible kinetic reaction
limitations, along with, saving time and decreasing experimental work and
associated costs.
2. OBJECTIVE
The objective of this chapter is to present the advantages of the use of

thermodynamic modeling for the production of high-purity hydrogen by

ethanol steam reforming combined with CO2 absorption. Through this case
study, a thermodynamic analysis of the ESR-CO2-ABS system by means of
the Gibbs free energy minimization technique is used to determine favorable
experimental conditions to produce a high-purity hydrogen product stream.
This case consisted in the ESR-CO2-ABS system by the steam reforming of
ethanol, while the CO2 absorbents studied were CaO, CaO•MgO, Na2ZrO3,
Li2ZrO3 and Li4SiO4. The influence of the steam to ethanol (S/EtOH) feed
molar ratio and the temperature at atmospheric conditions on the
composition of the product gas was investigated for each material. Results
were compared with experimental and theoretical data obtained from
existing literature. In addition, an analysis of the carbon deposition
conditions is included, due to its relevancy to the continuous operation of
the process, since carbon deposition could be the main cause of catalyst
deactivation, resulting in a low durability and loss of activity of this.
3. GIBBS FREE ENERGY MINIMIZATION TECHNIQUE
In a reaction system, in which many simultaneous reactions take place,

equilibrium calculations can be performed using the Gibbs energy
minimization approach (also called the non-stoichiometric method). In this
technique, the total free energy of the system consists of an ideal gas phase
and condensed pure phases, which can be expressed as:
𝐺 𝐺° 1
𝑅𝑇
= {∑𝑁 𝑖
𝑖=1 𝑛𝑖 𝑅𝑇 + ln(𝑦𝑖 𝑃)} + 𝑅𝑇 {∑𝑁 °
𝑖=1 𝑛𝑖 𝐺𝑖 }𝑐𝑜𝑛𝑑𝑒𝑛𝑠𝑒𝑑 (4)
𝑔𝑎𝑠
The technique is based on the search for different values of ni that

minimizes the objective function (5) and is subjected to the restrictions of
the elemental mass balance:
∑𝑁
𝑖=1 𝑎𝑖𝑗 𝑛𝑖 = 𝐴𝑗 , 𝑗 = 1, 2, 3 … , 𝑘 (5)

where aij is the number of atoms of the jth element in a mole of the i-th species.
Aj is defined as the total number of atoms of the j th element in the reaction
mixture [13]. All calculations in the two cases presented were made using
the equilibrium module of the HSC Chemistry® software [14]. HSC
calculates the equilibrium composition of all possible combinations of
reactions that are capable of occurring within the thermodynamic system.
These thermodynamic calculations make use of the equilibrium composition
module of the HSC program that is based on the Gibbs free energy
minimization technique. The Gibbs program of this module finds the most
stable phase combination and looks for compositions where the Gibbs free
energy of the system reaches its minimum (equation 4) in a fixed mass
equilibrium (a minimization problem constrained by equation 5), while
fixing pressure and temperature. In this non-stoichiometric approach, each
species in the system must be defined. The selection of viable products
should be based on previous experimental results found in the literature. For
each system, the possible species are specified based on thermodynamic and
experimental analyses previously reported.
Species analyzed during the development of this case study included
ethanol, ethylene, ethane, acetone, acetaldehyde, acetic acid, C, CO, CH4,
CO2, H2 and H2O and these were based on the experimental species reported
in the literature [15-17]. Additionally, appropriate conditions were used for
cases where CO2 absorbents were employed and the possible formation of
solid and elemental carbon was expected. For the case of the CaO absorbent,
Ca(OH)2 and CaCO3 species were considered. For the calcined dolomite
(CaO•MgO) absorbent, CaO, Ca(OH)2, CaCO3, MgCO3 and MgO species
were included in the calculations. When calculations involved the sodium
zirconate absorbent Na2ZrO3, Na2CO3 and ZrO2 species were considered,
while for the lithium zirconate absorbent Li2ZrO3, Li2CO3 and ZrO2 species
were involved. For the lithium orthosilicate (Li4SiO4) absorbent, Li2CO3 and
Li2SiO3 species were added to the thermodynamic system. Finally,
carbonation reactions involved in the CO2 absorption process considered in
the present study are presented below:
CaO•MgO + CO2(g) = CaCO3 + MgO ΔH°298 = -170.9 kJ/mol (6)

Na2ZrO3 + CO2(g) = Na2CO3 + ZrO2 ΔH°298 = -151.5 kJ/mol (7)
Li2ZrO3 + CO2(g) = Li2CO3 + ZrO2 ΔH°298 = -162.5 kJ/mol (8)
Li4SiO4 + CO2(g) = Li2CO3 + Li2SiO3 ΔH°298 = -142.2 kJ/mol (9)
4. CASE STUDY: ETHANOL STEAM REFORMING PROCESS

COMBINED WITH CO2 ABSORPTION FOR
THE PRODUCTION OF HYDROGEN
The ethanol steam reforming process combined with CO2 absorption

(ESR-CO2-ABS) for the production of H2 is based on the combination of
reaction (1) with reactions (3) and (6) to (9) for each of the absorbents, which
causes an increase in the amount of hydrogen produced by eliminating CO2
in the reaction system. During the thermodynamic modeling, the reaction
temperature was varied in the range of 300-850°C at 1 atm. The
steam/ethanol (S/EtOH) feed molar ratio was varied from 3:1
(stoichiometric) to 6:1. For the carbon formation study, the S/EtOH feed
molar ratio was varied from 0.2 to 0.5. Additionally, thermal efficiency of
the process was calculated based on the approach developed by Da Silva and
Muller et al. [8, 13]. The details of the equations and calculation methods
can be found in their studies. The calculated LHV values for hydrogen and
ethanol are 239.2 and 1240 kJ/mol respectively. According to equation (1),
the maximum thermal efficiency for the conversion of a biofuel (ethanol)
with steam can be found assuming that one mole of ethanol produces 6 moles
of H2. For the ESR-CO2-ABS system, the procedure for calculating the
thermal performance is also described according to similar thermodynamic
studies [8, 13]. The only difference with respect to these studies is that HSC
takes into account the fact that the reactor feed is composed of biofuel, water
vapor and a solid CO2 absorbent (typically with a 50% excess). Therefore,
the thermal efficiency depends on the production of hydrogen and the

required heat consumption (heat load) to be used to regenerate the solid CO2
absorbent.
4.1. Thermodynamic Analysis Results for the ESR-CO2-ABS

Reaction System
Table 1 exhibits a summary of the thermodynamic equilibrium results

for the ESR-CO2-ABS system with and without the use of a CO2 absorbent.
The conditions included in this table approximated the maximum hydrogen
production obtained for each absorbent, which were S/EtOH = 6:1 and 600
°C. This last temperature was chosen because it represents the average of the
maximum hydrogen production in all the performed calculations. This table
shows the evident limitation of the ethanol steam reforming (ESR) without
CO2 absorption, since only 4.6 moles of H2 were produced per mole of
ethanol. Even at this relatively high temperature, the amounts of CO, CH4
and CO2 are relatively high, with 0.23, 0.54 and 1.23 moles at equilibrium,
respectively. In addition, the concentration of hydrogen was only 69%. The
expected improvement with the use of a CO2 absorbent is clear when a
comparison is made with respect to the values obtained for CaO.
According to Table 1 the increase in hydrogen production was 19.2%
higher using CaO and consequently, the other gaseous species were reduced.
The CO produced was reduced by 3.3 times, while CO2 decreased 15.4
times. In addition, methane diminished 10.8 times. All this translates into a
97% concentration of H2 for this absorbent. Other absorbents behaved
similarly to the CaO results. For example, calcined dolomite (CaO•MgO)
exhibited a minor difference in results with respect to CaO with a hydrogen
production of 3.5% below this, while CO and CO2 slightly increased.
However, the production and concentration of hydrogen (94%) was 16% and
25% greater compared to the non-absorbent reforming conditions (w/o
Absorbent, 69%). Results for absorbents such as Na2ZrO3 and Li2ZrO3 were
very similar between them to those of CaO•MgO, as can be seen in Table 1.
For example, hydrogen production for Na2ZrO3 and Li2ZrO3 was 5.4 and 5.3
moles at equilibrium, respectively, compared to 5.5 for CaO•MgO. Moles of

CO at equilibrium followed the same trend, while moles of CO2 for the
zirconate absorbers showed only a slight decrease compared to those
produced by calcined dolomite (0.14 vs. 0.20 and 0.22, see Table 1) and the
same trend occurred with the formation of methane. However, the hydrogen
concentration was marginally lower for the zirconates (92%) compared to
the concentration produced by calcined dolomite (94%). Finally, Li 4SiO4
was the absorbent that produced the lowest amount of hydrogen and higher
concentrations of byproducts (CO, CO2 and CH4). This resulted in a reduced
hydrogen concentration of 85%.
Therefore, a crucial feature in the production of hydrogen through the
ESR-CO2-ABS system resides in the nature of the solid CO2 absorbent,
besides presenting a favorable thermodynamic equilibrium, must have an
adequate absorption capacity and a fast absorption-regeneration kinetics.
Several studies have focused their research on the effects of pressure,
temperature and composition of reactive gas on calcium oxide based
absorbents using the thermogravimetric (TGA) technique [18-19]. However,
the sintering of these materials reduces their performance after several
absorption-regeneration cycles. On the other hand, calcined dolomite
(CaO•MgO) has shown a better performance in the absorption of CO2 at high
temperatures compared to CaO in multi-cyclic tests [20]. Unfortunately, this
absorbent of mineral origin requires high regeneration temperatures (T >
950°C) that produce the degradation of the material after 10 cycles of
absorption-regeneration. Bandi et al. [21] proposed the use of mineral
huntite (Mg3Ca(CO3)4) that exhibits a good performance in its regeneration.
However, this absorbent has several disadvantages, such as high
regeneration temperature and low CO2 capacity. Also, of mineral origin, the
hydrotalcite Mg6Al2(CO3)(OH)16•4H2O was proposed by Hufton et al. [22]
and Ding and Alpay [23], who used this CO2 adsorbent at moderate
temperatures (400-500°C), resulting in low adsorption capacity. Studies
conducted by López Ortiz et al. [12] have demonstrated superior
performance of Na2ZrO3 as a synthetic solid CO2 absorbent, alternative to
those expensive lithium-based absorbents (Li2ZrO3 and Li4SiO4) reported by
Nakagawa and Ohashi [24] and Kato et al. [25]. This enhanced behavior was

attributed to its excellent thermal stability, fast absorption kinetics and

consistent CO2 capture capacity.
Recently, Ochoa Fernández et al. [26] experimentally evaluated several
synthetic CO2 absorbents, under the steam methane reforming with CO2
absorption as reaction scheme, among these included: Li2ZrO3, Li4SiO4 and
Na2ZrO3 and concluded that Na2ZrO3 is the absorbent that presents the
highest methane conversions, hydrogen purity and reaction kinetics. In
addition, Jakobsen and Halmøy [27] modeled a methane reforming reactor
with CO2 absorption using CaO, Li4SiO4 and Na2ZrO3 and also concluded
that Na2ZrO3 is the most efficient absorbent with the highest hydrogen
production ratio (92.6%) compared to CaO (79.3%) and Li4SiO4 (82.1%) at
the same reaction conditions in a temperature range of 600°C to 800°C.
Therefore, from the thermodynamic analysis previously presented of the
ESR-CO2-ABS system, it can be concluded that Na2ZrO3 is an absorbent
that has a favorable thermodynamic equilibrium comparable to CaO and
greater kinetics and stability performance.
Table 1. Summary of thermodynamic equilibrium results

of ethanol steam reforming at 600°C
Absorbent Moles in Equilibrium Parameter

H2 CO CO2 CH4 S/EtOH ratio %H2
CaO 5.7 0.07 0.08 0.05 6:1 97
CaO•MgO 5.5 0.11 0.14 0.09 6:1 94
Na2ZrO3 5.4 0.13 0.20 0.12 6:1 92
Li2ZrO3 5.3 0.13 0.22 0.13 6:1 92
Li4SiO4 5.0 0.18 0.45 0.24 6:1 85
w/o Absorbent 4.6 0.23 1.23 0.54 6:1 69
4.2. Carbon Formation
For the present case study, the formation of carbon was examined in a
Steam/EtOH feed molar ratio from 0.2 to 0.5. Results indicate that without
the use of a CO2 absorbent carbon formation is favored and its maximum is

reached at lower temperatures than with the use of a CO2 absorbent. The
steam reforming without absorbent (w/o Absorbent in Table 2) produced a
maximum carbon formation of 0.96 kmol per mole of ethanol fed to the
system at a Steam/EtOH = 0.5 ratio and a temperature of 577°C. At higher
temperatures and S/EtOH ratios, lower amounts of carbon are produced,
whereas greater S/EtOH ratios than 2.5 and temperatures above 512°C will
ensure the prevention of coal deposits in this system. Surprisingly, the use
of CO2 absorbents produced the highest carbon formation at the same
temperatures and S/EtOH ratios, with the exception of Li4SiO4, which was
the most likely to produce larger amounts of carbon than other absorbents
studied. Carbon-free operation can be found at higher S/EtOH ratios and
temperatures of 2.0 and 958°C, respectively for CaO-based absorbents (CaO
and MgO•CaO). While, higher temperatures than 760°C are needed for
synthetic absorbents. It is important to mention that synthetic absorbents
(Na2ZrO3, Li2ZrO3 and Li4SiO4) have the tendency to present lower carbon-
free operating temperatures than with the use of CaO-based absorbents. The
behavior related to a lower carbon formation that was observed with the use
of a CO2 absorbent, is directly related to the reduction of the CO content. Li
[28] confirmed in his thermodynamic study that the formation of graphite is
suppressed with the absorption of CO2. According to this author, Boudouard
reaction:
2CO(g) = CO2(g) + C(s) ΔH°298 = -172.5 kJ/mol (10)
is promoted towards the reverse reaction (10) because its equilibrium

constant is related to the square of the CO concentration.
4.3. Thermal Efficiency Analysis
Table 2 shows the influence of the S/EtOH ratio on the thermal

efficiency (η) at 600°C. In this Table, the temperature of 600°C was selected
because at this condition, most of the hydrogen reaction systems reached
their maximum production and purity.

Table 2. Summary of the Thermal Efficiency Results

of the ESR-CO2-ABS at 600° C
Absorbent S/EtOH Molar Ratio Thermal Efficiency (η, %)

w/o Absorbent 5 64.0
6 66.3
6.5 67.0
CaO 6 82.6
6.5 81.9
CaO*MgO 6 80.0
Na2ZrO3 6 78.3
Li2ZrO3 6 77.9
Li4SiO4 6 73.5
For the case where no absorbent was used, the thermal efficiency of the
process increases as the S/EtOH ratio also increases. However, as suggested
by He et al. [29] the cost of energy for the generation of additional steam
can be compensated by promoting the production of hydrogen in the
reforming process. For example, at a Steam/EtOH ratio of five the thermal
efficiency is 64%, while the increase to a value of six yielded an efficiency
of 66.3%, which reflects the behavior described above. However, a greater
increase in the S/EtOH ratio to 6.5 only produces a marginal increase in
thermal efficiency (67%). These values agree well with the studies reported
by Lima da Silva and Müller [8]. Therefore, here it can be concluded that
the S/EtOH = 6 and 600°C is a good choice of conditions in which all the
absorbents can be compared in terms of thermal efficiency. Furthermore, in
Table 2 the values for the thermal efficiency of the ESR-CO2-ABS system
are presented for each studied absorbent. The use of CaO produced a thermal
efficiency of 82.6%, while an increase in the S/EtOH ratio of 6.5 produced
a slight reduction of its efficiency to a value of 81.9%. This means that the
amount of steam generated can eventually reduce the efficiency of the
system, since not only the gaseous species absorb the heat provided by the
water vapor, but also the solid absorbent retains part of that heat. However,
the advantage in the use of a CO2 absorbent (CaO) results in an increase in
thermal efficiency of 16.3%, which represents significant energy savings
generated by the ESR-CO2-ABS reaction scheme. The use of calcined

dolomite only reduces the thermal efficiency by 2% with respect to CaO.

This is reasonable if it is considered that, an inert (MgO) is used together
with the main absorbent compound (CaO). The same behavior is observed
for Na2ZrO3, Li2ZrO3 and Li4SiO4, which resulted in efficiencies slightly
lower than CaO with values of 78.3, 77.9 and 73.5, respectively. The
differences observed can be explained in terms of their different
regeneration temperatures, their heats of reaction and regeneration and the
amount of hydrogen produced. Therefore, a complex combination of
parameters generates a specific thermal efficiency. From the results
presented in Table 2 it can be seen that the synthetic absorbent that presents
a better performance in terms of thermal efficiency is Na2ZrO3, followed by
Li2ZrO3 with a small difference, while Li4SiO4 showed a limited thermal
efficiency.
4.4. Optimal Operating Conditions for the ESR-CO2-ABS

Reaction System
For ethanol under the ESR-CO2-ABS reaction system, it can be

observed that it is possible to obtain a concentration of hydrogen of 97%
purity at 1 atm, 600°C and S/EtOH = 6. At these conditions, a yield of 5.7
moles of H2 per mole of ethanol fed is generated with the use of CaO as a
CO2 absorbent. It is worth mentioning that the thermal performance is
closely related to the concentration and hydrogen production in the product
gas. Therefore, this set of conditions guarantees high thermal performances
for the ESR-CO2-ABS reaction system.
CONCLUSION
Thermodynamic modeling of the ESR-CO2-ABS reaction system was

carried out to determine favorable operating conditions to produce a high
purity H2 stream. Results indicate that without the use of a CO2 absorbent
there is no carbon formation at S/EtOH ratios less than or equal to the

stoichiometric values (S/EtOH ≤ 3). However, for the ethanol reforming

system using a CO2 absorbent, carbon-free operation can be found at S/EtOH
ratios greater than 2.0 and temperatures above 958°C. Generally, carbon
formation is suppressed with the absorption of CO2 compared to the
operation of conventional ethanol steam reforming.
At optimum operating conditions for the ESR-CO2-ABS system it is
possible to produce a concentration of 99% of hydrogen at 1atm, at 600°C
and S/EtOH = 6. The CaO•MgO and CaO absorbents showed similar results
with high levels of hydrogen purity and yields, while Na2ZrO3 and Li2ZrO3
resulted only in slightly lower values than those found with CaO. While,
thermal efficiency of conventional reforming and under the ESR-CO2-ABS
system is strongly related to the concentration of hydrogen in the product
gas. Thermal efficiency was also improved with the use of a CO2 absorbent,
since with the ESR-CO2-ABS system increased from 66% to 81% with
respect to the conventional reforming process. Finally, Na2ZrO3 and Li2ZrO3
can be considered as promising alternative absorbents with thermodynamics
comparable to those based on CaO. The limited durability of the CaO and
MgO•CaO absorbents makes the zirconate-based absorbents the ideal
materials to be used in the ESR-CO2-ABS system. Finally, of the two
zirconates, Na2ZrO3 has the highest kinetics and thermal stability. Therefore,
Na2ZrO3 should be considered the absorbent with the highest potential under
the ESR-CO2-ABS reaction system for future experimental evaluations.
REFERENCES
[1] Owen N. A., Inderwildi O. R., Kinga D. A., Energy Policy, (2010) 38,
4743-4749.
[2] Dunn S., Int. J. Hydrogen Energy, (2002) 27, 235-264.
[3] Schmidt J., Sylvain L., Dotzauer E., Kindermann G., Schmid E., Appl.
Energ. (2010) 87, 2128-2141.
[4] Brown L. F., Int. J. Hydrogen Energy, (2001) 26, 381-397.
[5] Amigun B., Gorgens J., Knoetze H., Energy Policy, (2010) 38, 312-
322.

[6] Lysikov A. I., Trukhan S. N., Okunev A. G., Int. J. Hydrogen Energy,
(2008) 33, 3061-3066.
[7] Davda R., Shabaker J., Huber G., Cortright R., Dumesic J., Appl Catal
B: Environ, (2005) 56, 171-186.
[8] Lima da Silva A., Müller I. L., Int. J. Hydrogen Energy, (2011) 36,
2057-2075.
[9] Ochoa-Fernández E., Berntsen H., Zhao T., Hi L., Blekkan E. A., Chen
D., North American Catalysis Society 20th North American Meeting,
(2007) June 17-22, Abstract P-S12-34B.
[10] Hildengrand N., Readman L., Dahl I. M., Blom R., Appl. Catal., A,
(2006) 303, 131-138.
[11] Kato M., Yoshikawa S., Nakawaga K., J. Mater Sci. Lett., (2002) 21,
485.
[12] López A., Pérez N., Reyes A., Lardizábal D., Sep. Sci. Technol. (2004)
39, 3563–3579.
[13] Jarungthammachote S., Dutta A., Energy Convers. Manage, (2008)
49, 1345–1356.
[14] Roine A., Outokumpu. HSC 6.0 Chemistry for windows 2010.
[15] Benito M., Sanz J. L., Isabel R., Padilla R., Arjona R., Daza L., J.
Power Sources, (2005) 151, 11-17.
[16] Fatsikostas A. N., Verykios X. E. J. Catal, (2004) 225, 439-452
[17] Biswas P., Kunzru D., Int. J. Hydrogen Energy, (2007) 32, 969-980
[18] Han C, Harrison D. P. Sep. Sci. Technol. (1997) 32, 681-97.
[19] Wu Y, Alvarado F. D., Santos J. C, Gracia F, Cunha A F, Rodrigues
A. E., Chem. Eng. Technol. (2012) 35, 847-58.
[20] Silaban A, Harrison D. P. Chem. Eng. Comm. (1996) 47, 149-62.
[21] Bandi A, Specht M, Sichler P, Nicoloso N. In situ gas conditioning in
fuel reforming for hydrogen generation. In: 5th international
symposium on gas cleaning at high temperature. Morgantown West
Virginia; (2002) Available at: http://www.zsw-bw.de/en/docs/
research/REG/pdfs/REG_5th_ISGC_2002.pdf.
[22] Hufton J. R., Mayorga S., Sircar S., AIChE Journal, (1999) 45, 248-
256.
[23] Ding Y., Alpay E., Process Saf. Environ. Prot., (2001) 79, 45-51.

[24] Nakagawa K., Ohashi T. J., J. Electrochem. Soc., (1998) 145, 1344-
1346
[25] Kato M., Yoshikawa S., Essaki K., Nakagawa K., Toshiba
Corporation. INTERMAC, Japan Electric Measuring Instruments
Manufacturers’ Association, Joint Technical Conference. (2001), SE-
3, 1021.
[26] Ochoa-Fernández E., Lacalle-Vilà C., Zhao T., Rønning M., Chen D.,
Stud. Surf. Sci. Catal., (2007) 167, 159-164.
[27] Jakobsen J. P., Halmøy E., Energy Procedia, (2009) 1, 725-732.
[28] Li M., Int. J. Hydrogen Energy, (2009) 34, 9362-9372.
[29] He L., Parra J. M. S., Blekkan E. A., Chen D., Energy Environ. Sci.,
(2010) 3, 1046-1056.
Reviewed by:
Dra. Rosa de Guadalupe González Huerta

ESIQIE-IPN, México
Alternate Energy Laboratory
rosgonzalez_h@yahoo.com.mx
Dr. Miguel Ángel Valenzuela Zapata

ESIQIE-IPN, México
Laboratory of Catalysis and Hydrogen Production
mavalenz@ipn.mx

Chapter 6
NEW PIPERAZINE-BASED POLYMERIZABLE

MONOMERS FOR THE DEVELOPMENT
OF STIMULI-RESPONSIVE POLYMERS
G. Roshan Deen and Chin Hao Mah

Soft Materials Laboratory, Natural Sciences and Science
Education, Nanyang Technological University,
Singapore
ABSTRACT
In recent years, new research emphasis is directed towards integrating

multiple functions into polymer materials. Among this new advance in
materials science are the functional polymers with structural designs for
versatile applications. Piperazine is an interesting heterocyclic diamine,
which possesses excellent medicinal and chelating property. Polymers
based on piperazine (polyamido-amines), surfactants based on piperazine
have shown good hydrophilic, chelating anti-bacterial and anti-helminthic
properties. However, piperazine-based homopolymers and copolymers
have not been investigated for stimuli-responsive behavior in aqueous
solutions and in bulk. Stimuli-responsive polymers have the ability to

Corresponding Author Email: Roshan.gulam@nie.edu.sg.

196 G. Roshan Deen and Chin Hao Mah
respond to changes in external stimuli such as pH, temperature, salts,

surfactants, light, magnetic field, pressure, and biomolecules. These
materials are synthesized in various architectures such as linear polymers,
block polymers, hydrogels, nanogels, inter-penetrating networks, plum-
pudding type gels, and dendrimers. The behavior of this class of polymeric
materials resembles that of living systems in response to changing external
environment. These materials find wide range of applications in
biomedical science, pharmaceuticals, cosmetics, chemical sensors etc. In
continuation of our research interest in piperazine-based polymers of
poly(amido-amine) architectures, we describe the synthesis of new
piperazine-based polymerizable monomers in the development of multiple
stimuli-responsive polymer systems.
Keywords: piperazine, stimuli-responsive polymers, piperazine-based

monomers, heterocyclic diamine, pH-responsive polymers
INTRODUCTION
The functions of living cells are regulated by natural biopolymers such

as carbohydrates, proteins, lipids, and nucleic acids. The important feature
of these biopolymers is their ability to respond to changes in local
environment leading to the formation of molecular assemblies [1-3]. Many
synthetic polymers exhibit this behavior in response to changes in external
stimuli, and such polymers are termed, as ‘smart’, ‘intelligent’ or ‘stimuli-
responsive’ polymers. Due to the reversible stimuli-responsive property,
these polymers find applications in targeted drug delivery systems,
nanomedicine, bioadhesion mediators, sensors/actuators, microfluidic
devices etc., [4-10]. Although the well-known glass transition and melting
temperature of solid polymers fit within the definition of stimuli-responsive
polymers, recent interest in these systems focus on aqueous solutions,
hydrogels, microgels and interfaces.
The external stimuli is either physical, chemical or biochemical in nature
in the form of temperature, pH, ionic strength, light intensity, pressure,
biomolecules, magnetic field etc., [10-15]. The potential response to these
external stimuli are in the form of changes in phase, shape, volume, or
optical properties of the polymer. A schematic representation of stimuli-

New Piperazine-Based Polymerizable Monomers… 197
responsive polymer system is shown in Figure 1. For a polymer system to

respond strongly to slight changes in the external environment, a phase
transition (first-order) accompanied by a sharp decrease in the specific
volume of the polymer should take place.
Of interest are materials that respond to changes in more than one
external stimuli (multiple stimuli). The multiple stimuli-responsive
polymers are usually developed through various chemical synthesis such as
[16-20]:
1. copolymerization of functional monomers

2. combination of temperature-responsive polymers with
polyelectrolytes
3. polymerization of monomers through inter-penetrating networks
4. polymerization of new functional monomers
Cationic polymers contain tertiary amine functional group that are

protonated under acidic pH conditions. This leads to the formation of fixed
electric charges on the polymer network that undergoes electrostatic
interactions. The charges interact with anionic biomolecules such as nucleic
acids, and proteins through the formation of electrostatic complexes. These
materials thus show promise in targeted drug and gene delivery applications
[21-23].
Figure 1. Stimuli-response behavior of polymers in response to various external

stimuli.

Although there are, a variety of stimuli-responsive cationic polymers

available, polymers based on piperazine has received considerably less
attention. Piperazine is a heterocyclic diamine with excellent biocidal,
hydrophilic and chelating property. In this chapter, piperazine based
polymers that show stimuli-responsive properties is discussed.
STIMULI-RESPONSIVE CATIONIC POLYMERS
Polymers that respond to changes in external pH are termed as pH-

responsive polymer. These contain ionisable functional groups (weak acids
or weak base) in their structure. The weakly acid and basic functional groups
either release protons or become protonated in response to external pH
changes. For example, the polymer polyacrylic acid (PAa) ionizes to
polyacrylate anions in solutions of basic pH [24], while the polymer
poly(dimethyl aminoethyl methacrylate) (PDMAEMA) becomes protonated
in solutions of acidic pH [25]. This pH-responsive ionization is illustrated in
Figure 2.
Figure 2. Effect of external pH on ionization of pH-responsive polymers.

Cationic Polymers (Natural Origin)
Cationic polymers that are of natural origin are chitosan and gelatin.
Chitosan is a linear cationic copolymer composed of randomly distributed
D-glucosamine and N-acetyl glucosamine linked at the β-(1→ 4) position.
This polymer is derived from the exoskeleton of crustaceans [26] and the
chemical structure is shown in Figure 3.
The acid-dissociation constant (pKa) of the primary amine group is 6.5
and the materials is pH responsive. The low solubility of chitosan limits the
application in drug delivery systems [27].
Other than the pH-sensitive characteristic, this weak polybase has
beneficial properties such as antibacterial, antioxidant, biocompatible, non-
toxic and biodegradable that impart versatility in drug release, cell
transfection and muco-adhesion applications [28, 29]. However, the
limitation of the usage on chitosan in drug delivery and therapeutic
applications is due to its pKa that is near physiological pH, which results in
low solubility. The solubility is improved by quaternization reaction to form
quaternary ammonium ions as shown in Figure 4.
Figure 3. Chemical structure of chitosan.
Figure 4. Effect of pH on chitosan.

The N, N, N-trimethyl chitosan, N-propyl-N, N-dimethyl chitosan, N-

dodecyl-N, N-dimethyl chitosan and N-diethylmethylamino chitosan are few
examples of reported quaternary ammonium salts of chitosan that have
improved muco-adhesive, solubility, and bioactive properties such as
necrosis of bacterial cells, Staphylococcus aureus [30, 31].
Gelatin is a thermally denatured collagen extracted from connective
tissues in skin, tendon and bone. The denaturing process breaks up the
tertiary structure of collagen producing gelatin, a biomacromolecule that is
composed of 18 non-uniformly distributed amino acids which contain
inherent cationic characteristic mainly due to the lysine and arginine
residues [32, 33]. A typical chemical structure of gelatin containing amino
acid groups –Ala-Gly-Pro-Arg-Gly-Glu-4Hyp-Gly-Pro-177 is shown in
Figure 5 [34].
Two types of gelatins are generally obtained: (i) Type A, obtained by
acid pre-treatment and (ii) Type B, obtained by alkaline pre-treatment. The
acid pre-treated type A gelatin has an isoelectric point (IEP) of 6-9 with no
significant change to its intrinsic properties [34], while the alkaline-treated
type B gelatin has IEP of 4.7-5.4 but has its amide groups of asparagines and
glutamine hydrolyzed to aspartate and glutamate converting the type B
gelatin to contain mainly carboxyl group residues [32]. Hence, cation
density is higher for acid-treated type A gelatin and lower for alkaline-
treated type B gelatin.
Figure 5. Chemical structure of gelatin containing amino acid groups.
Generally, the gelatin exhibits cationic behavior at pH values below its

IEP. Despite this pH limiting factor, the gelatin or gelatin derivatives are

widely used in pharmaceutical and medical applications [35-37] due to it

biodegradability, and biocompatibility under physiological conditions. Due
to its biocompatible and bioactive nature, Ovsianikov and coworkers
modified the gelatin with methacrylamide to produce a strong and rigid
scaffolding [38]. In order to improve the bioactive property of gelatin, Jiang
and coworkers successfully synthesized quarternized ammonium-gelatin
with 2,3-epoxy propyltrimethyl ammonium chloride. This material, showed
improved anti-bacterial properties with minimum inhibitory concentration
(MIC) against two types of bacteria such as of bacterium Staphylococcus
aureus and Escherichia coli [39].
Cationic Polymers (Synthetic Origin)
The polyethyleneimine (PEI) is most widely used cationic polymer. It is

composed of primary, secondary and tertiary amine functional groups and
exists in both linear and branched forms with varying molecular mass (Mw)
[40, 41]. The linear polyethyleneimine (LPEI) is synthesized via ring
opening polymerization of 2-ethyl-2-oxazoline followed by hydrolysis,
while the branched polyethyleneimine (BPEI) is obtained through acid-
catalyzed polymerization of aziridine [42]. Figure 6 (A and B) depict the
synthesis of LPEI and BPEI respectively.
Both LPEI and BPEI contain several amine functional groups that can
be protonated and developed as antimicrobial agents [43]. These materials
are used as gene carriers to attract negatively charged nucleic acids [44].
Despite the numerous applications, the high molecular mass PEIs form
enzymatically stable polyelectrolyte complexes. This decreases its
degradability and hence increases the toxicity and lowers the haemo-
compatibility. This drawback is resolved by incorporating biodegradable
disulfide, ester or acid-labile crosslinkers in the PEI backbone. Kim and
coworkers through the usage of acid-labile crosslinker glutaraldehyde [45]
investigated the biodegradable property of this material.

Figure 6. Synthesis of linear (A) and branched (B) polyethylenimines.
The poly(amino-ester)s PAEs belongs to the class of synthetic and

hydrolytically degradable polyamine [46]. These PAEs are synthesized
through Michael addition reaction of primary or secondary amines with
diacrylate esters as shown in Figure 7. The amino group in the PAEs
backbone makes it responsive to pH changes while the ester group
contributes to its degradability. This type of easily labile, ionizable and pH
-responsive PAEs have been widely used in transfection, drug delivery and
antimicrobial applications [47-50].

Figure 7. Synthesis pathway of poly(amino-ester) PAE.
The poly(amido-amine)s PAAs are synthetic cationic polymers that are

obtained by Michael-type polyaddition reaction with primary amines or
secondary amines to bis-acrylamides as shown in Figure 8. This synthesis
pathway leads to polymers containing tertiary amine and amido functional
groups arranged in the main backbone chain. The first linear PAA was
reported by Ferruti and coworkers with a comparative study on the rate of
polymerization of PAAs from monomers containing di-amine, tri-amine and
piperazine [51, 52]. The synthesized PAAs have pKa values ranging from
2.5 to 9.2, which contributes to the pH-responsive behavior. Beside the
inherent pH-responsive characteristic, the PAAs are regarded as
customizable and versatile materials as the side-substituents contain other
chemical functional groups such as alkyl, allyl, amino, carboxyl, thio and
hydroxyl groups [53, 54]. The PAAs specifically those containing
heterocyclic piperazine molecule, are of importance as piperazine forms an
important component in numerous medicinal drugs [55].

Figure 8. Michael type addition reaction in the synthesis of poly(amido amine).
Figure 9. Chemical structure of piperazine
PIPERAZINE AND DERIVATIVES
Piperazine, is a heterocyclic diamine (Figure 9) and was first introduced

in 1953 as an anthelmintic veterinary drug to effectively expel parasites from
the body of the host [56]. It is a weak base with two pKbs of 5.35 and 9.73.
Since then it has been widely used in medicinal products. Piperazine-
crosslinked with ciprofloxacin dimers (Figure 10a) are potent antibacterial
agents against resistant strains including Staphylococcus aureus and
Escherichia coli [57-59]. The N-alkyl and N-aryl substituted piperazine
derivatives are also potent against bacterial cells such as Staphylococcus
epidermidis, Staphylococcus mutans, Bacillus subtilis and Pseudomonas

aeruginosa [60, 61]. Piperazine derivatives containing quinolones [62]

(Figure 10b) and tetrazole [63] (Figure 10c) moieties have antimalarial and
antifungal properties. The diphenyl alkyl piperazine derivatives (Figure 10d)
inhibit dopamine neurotransmitter dopamine [64] and show promise in the
treatment of Parkinson’s disease [65], depression [66] and cocaine addiction
[67]. A few medicinal drugs containing piperazine moiety that are marketed
is listed in Table 1.
Figure 10. Chemical structure of piperazine-derivatives used in pharmaceutical drugs.

Table 1. Chemical structure of piperazine-derivatives

in medicinal products [68-73]
No. Chemical Trade Chemical structure Condition

name name
1 Cetirizine Zyrtec antihistamine
2 Trimetazidine Vastarel antianginal
3 Amoxapine Demolox antidepressant
4 Indinavir Crixivan HIV protease

inhibitor
5 meta- Ecstasy agonist

Chlorophenyl antimigraine
-piperazine

POLY(AMIDO-AMINE)S BASED
ON PIPERAZINE
The free diamine on 1, 4 position of the piperazine ring is capable of

chelating or coordinating metal ions and ionic salts and can be used in
separation science. The surface-treated poly(vinyl alcohol) PVA membranes
with coating of poly(piperazine-amide) (Figure 11a) developed by Jahnshahi
and coworkers, is effective in rejecting ionic salts as compared to the non-
treated PVA membranes [74]. Likewise, Religa and coworkers utilized the
pH-sensitive nature of poly(piperazine-amide) by manipulating the pH of
the bulk solution to improve the recovery efficiency of chromium(III) [75].
The polystyrene-supported piperazine-substituted traizoles developed by
Girard and coworkers, evaluated and emphasized the addition of piperazine
moiety improves the efficiency of metal ions (Fe3+, Cu2+ and Pb2+) extraction
[76].
Apart from water purification processes, the piperazine moiety improves
adhesive and mechanical strength for materials used as flame-retardants.
Singh and coworkers developed bis-Phenol (bP) (Figure 11b) with additive
piperazine as corrosion inhibition coating on Mild steel. The piperazine-
based bP is protonated by 1M HCl to serve as adhesive through
physisorption with acid-treated metal surface [77]. This mode of action
ensures that there is donor-acceptor interactions between the π-electrons of
aromatic ring on bP and vacant d-orbital of surface metal, causing a layer of
adsorbed electrons on the inhibitor preventing electrochemical reaction from
occurring on the metal surface and thus protects the metal from corrosion
[77, 78]. Besides being used as adhesive, the piperazine is used as
enhancement additive to existing polymer systems. Tan and coworkers [79]
modified ammonium polyphosphate with piperazine (Figure 11c), to
enhance mechanical strength for flame retardance / suppression property
from fire growth rate (FIGRA) of 14.8 kWm-2s-1 to 4.4 kWm-2s-1.
The initial synthesis and development of piperazine-based polymer by
Ferruti and coworkers, was utilized in pharmaceutical and therapeutic
industries [51, 52]. Ferruti and coworkers studied the reactivity of

polyaddition of primary amines, disecondary diamines and amino-acids with

bis-acrylamides and disecondary diamine with bis-acrylic ester to produce a
series of PAAs [51]. Figure 12 illustrates a series of PAA piperazine-based
polymers synthesized by Ferruti and coworkers [52]. The study reveals the
increasing order of basicity: piperazine ≤ 2-methylpiperazine < N,N’-
dimethylethylenediamine ≤ N,N’-diisopropyl- ethylenediamine, i.e., higher
the basicity of amine, the lower is the reactivity.
Figure 11. Chemical structure of polymer-based piperazine derivatives.
Ferruti and coworkers later developed piperazine-based polymers

through polyaddition of N, N’-dimethylethylenediamine with 1,4-bis-

acryloylpiperazine, glycine (Figure 13a) and 2, 3-trans-piperazine-

dicarboxylic acid (Figure 13b) [80, 81]. The study concluded that both
piperazine-based PAA are good heparin absorbers in blood regardless in-
vitro or in-vivo, which did not exhibit any haemolytic action and remained
non-thrombogenic in blood [81-83].
Figure 12. Chemical structure of piperazine-based polymers.
Ferruti and coworkers further extended the application of piperazine-

based PAAs in industry. They synthesized a series of various piperazine-
based PAAs and compared against structurally similar morpholine-
derivatives for heavy metal (Cu2+) chelation [84, 85]. Figure 14 shows the
chemical structure of piperazine-based PAAs and morpholine-derivatives.
The study concluded that piperazine-based PAAs is more effective in
coordination with Cu2+ ions in comparison with morpholine containing
polymers.

Figure 13. Chemical structure of piperazine-based PAAs used as heparin antagonist.
Figure 14. Chemical structure of piperazine-based PAAs and morpholine-derivatives.

PIPERAZINE-BASED STIMULI-RESPONSIVE POLYMERS
Due to the presence of tertiary amine, piperazine containing PAA show

pH-responsive behavior as function of external pH changes. The first study
of pH-responsive polymers (linear and hydrogels) based on piperazine is by
Gan and coworkers [86]. They developed a series of water-soluble
piperazine-derivatives based on poly(N-acryloyl-N’alkyl piperazine) (alkyl
= methyl, ethyl and propyl). The acid dissociation constant (pKa) of poly(N-
acryloyl-N’-methyl piperazine) (AcrNMP) pKa = 4.51, poly(N-acryloyl-
N’ethyl piperazine) (AcrNEP) pKa = 4.58, and poly(N-acryloyl-N’-propyl
piperazine) pKa = 4.61 (AcrNPP), are established through potentiometric
titration [87]. The chemical structure of poly(N-acryloyl-N’-alkyl
piperazine) (alkyl = methyl, ethyl and propyl) are shown in Figure 15.
Figure 15. Chemical structure of poly(N-acryloyl-N’-alkylpiperazine).
Of the three polymers, only PAcrNPP was responsive to changes in

multiple stimuli. The pH-sensitive PAcrNMP swelled four folds in acidic
pH than in basic pH solutions due to protonation of tertiary amine groups.
The hydrogel was effective in complexing with various metal ions (Co2+,
Cu2+, Ni2+, Zn2+ and Ca2+). The efficiency of the metal-uptake improved with
increasing pH of the external media [88]. When AcrNMP was
copolymerized with methyl methacrylate (MMA), the resulting copolymer
was responsive to both pH and temperature changes. Rivas and coworkers
copolymerized N-acetyl-α-aminoacrylic acid (AacA) with AcrNMP to
enhance the uptake efficiency of metal ions (Pb2+, Ni2+, Cu2+, Co2+, Cd2+,

Cr2+, Zn2+ and Ag+) The copolymerized poly(AcrNMP-co-AacA) exhibited

the same metal uptake trend as PAcrNMP [89].
Stimuli-responsive polyampholytes (linear and hydrogels) based on
AcrNEP and maleic acid (MA) showed pH- and temperature responsive
behavior. With 10 mol % of MaA in the copolymer exhibited a LCST at 80.5
°C [90]. However, due to the hydrophilicity of MaA, increasing the content
increased the LCST above 85 °C. The polyampholytes hydrogel showed an
efficiency of 93 % removal of a toxic dye Congo red from water.
Copolymers of AcrNEP and NIPAM responded to changes in both pH and
temperature [18]. The LCST depended on the ratio of the corresponding
monomer units in the copolymer. For a ratio of AcrNEP 10 % and NIPAM
90 % in the copolymer, the LCST was 36.5 °C, while for AcrNEP 50 % and
NIPAM 50 % the LCST was 51.4 °C. It is known that the homopolymer of
PNIPAM exhibits phase-transition at 32 °C39 but PAcrNEP does not.
Therefore, the ionizable AcrNEP moiety increases the hydrophilicity of the
copolymer thereby raising the LCST. The copolymers exhibited salting-out
effect with salts such as NaCl, NaBr, and Na2SO4
The copolymer hydrogels was responsive to both pH and temperature
changes. Below the pKa of AcrNEP the gels swelled in acidic medium due
to electrostatic repulsions. The copolymers of AcrNEP and vinyl
caprolactam (VCL) responds to changes in both pH and temperature [91].
The LCST was highly dependent on the composition of AcrNEP and VCL
in the copolymer. Interpenetrating polymer networks based on AcrNEP and
VCL exhibited pH and temperature response behavior [92].
In this series of piperazine-based polymer, the homopolymer poly(N-
acryloyl-N’-propylpiperazine) (PAcrNPP) responds to multiple external
stimuli such as pH (pKa 4.61), temperature, ionic strength and surfactants.
The LCST of this polymer in water is 37 °C, which is close to that that of
human body temperature. The hydrogels based on this polymer swells in
solutions of acididc pH and deswells in basic solutions, with volume phase
transition temperature close to the LCST of the corresponding
homopolymers. The multiple response property of this polymer is based on
hydrophilic and hydrophobic groups of the polymer. As such, the need for
copolymerization to achieve this property is overcome [93].

Wang and coworkers [94] utilized the multiple stimuli response

behavior of PAcrNPP in the development of layer-by-layer multilayer films.
The films showed improved adsorption properties in response to external
stimuli. Based on the pKa values evaluated by Wang, polyacrylic acid (pKa
= 5.0) and PAcrNPP is (pKa = 6.3) the modified film experiences low
adhesive when either of the components experiences complete protonation.
Urban and coworkers [95] copolymerized 2-ethoxyethyl methacrylate
(EEMA) with AcrNPP to develop polymers with multiple responsive
properties. The stimuli responsive properties was studied by FTIR and DSC
methods. Substituting the alkyl chain of the piperazine ring with phenyl ring
imparts pH- and photo-responsive properties [96].
Goh and coworkers studied the protonation and ionic interaction of
PAcrNPhP with the following acidic polymers such as poly(ρ-vinylphenol)
(PVPh), poly(vinylphosphonic acid) (PVPA) and poly(acrylic acid) (PAA),
to form ionic complexes [96]. Swelling of hydrogels of PAcrNPhP in
solutions of various pH indicates the role of steric hindrance in swelling. The
shifting of IR bands, spectrum intensity and increase of binding energy
indicated strong interactive forces of ionic and hydrogen bonds.
CONCLUSION
In summary, stimuli-responsive polymers based on piperazine show

interesting properties in response to changes in external pH, temperature,
and ionic strength of salt solutions. The presence of tertiary amine in poly(N-
acryloyl-N’-alkyl piperazine) is prone to easy protonation in solutions of
acidic pH making the material responsive to pH changes. The alkyl group
attached to this tertiary amine (N’-alkyl) provides the correct hydrophilic-
hydrophobic balance for the material to become temperature-responsive thus
exhibiting he critical solution behavior. Overall, stimuli-responsive
polymers based on piperazine that are responsive to multiple external stimuli
can be developed either through modification of N’-alkyl group or through
copolymerization with hydrophobic comonomers. Based on the types of

piperazine-based polymers listed in this article, these materials show

promise for the development of polymer-based biomaterials.
REFERENCES
[1] Galaev, I. Y.; Mattiason, B. Trends Biotechnol 1999, 17, 335-340.

[2] Senaratne, W.; Andruzzi, L.; Ober, C. K. Biomacromolecules 2005, 6,
2427-2448.
[3] Cohen Stuart, M. A.; Huck, W. T. S.; Genzer, J.; Müller, M.; Ober, C.;
Stamm, M.; Sukhorukov, G. B.; Sleifer, I.; Tsukruk, V. V.; Urban, M.;
Winnik, F.; Zauscher, S.; Luzinov, I.; Minko, S. Nature Mater 2010,
9, 101-113.
[4] Kikuchi, A.; Okano, T. Adv. Drug. Delivery Rev. 2002, 54, 53-77.
[5] Li, S. K.; D’Emanuele. A. J. Controll Rel. 2001, 75, 55-67.
[6] Roshan Deen G.; Loh. X. J. Gels 2018, 4, 1-13.
[7] Kost, J.; Langer, R. Adv. Drug Delivery Rev. 2001, 46, 128-148.
[8] Peppas, N. Curr Opin Colloid Interface Sci. 1997, 2, 531-537.
[9] Ista, L. K.; Lopez, G.P. J Ind Microbiol Biotechnol 1998, 20, 121-125.
[10] Hoffmann, J.; Plotner, M.; Kuckling, D.; Fischer, W. J. Sens. Actuators
A 1999, 77, 139-144.
[11] Roshan Deen, G.; Mah, C. H. Polymer 2016, 89, 55-68.
[12] Stayton, P. S.; Shimoboji, T.; Long, C. Nature 1995, 378, 472-474.
[13] Jeong, B.; Bae, Y. H.; Kim, S. W. Nature 1997, 338, 860-862.
[14] Tanaka, T. Phys. Rev. Lett. 1978, 40, 820-823.
[15] Caliceti, P.; Veronese, F. M. Adv. Drug Delivery Rev. 2003, 55, 1261-
1277.
[16] Bulmus, V.; Ding, Z.; Long, C. J.; Stayton, P. S.; Hoffmann, A. S.
Bioconjug. Chem. 2000, 11, 78-83.
[17] Roshan Deen, G.; Gan, L. H. Polymer 2006, 47, 5025-5034.
[18] Roshan Deen, G.; Chua, V.; Ilyas, U. J. Polym. Sci. Polym. Chem.
2012, 50, 3363-3372.
[19] Verestiuc, L.; Ivanov, C.; Barbu, E.; Tsikbouklis, J. Int. J. Pharm.
2004, 269, 185-194.

[20] Gonzalez, N.; Elvira, C.; Roman, S.J. Macromolecules 2005, 38, 9298-
9303.
[21] Xu, X.; Capito, R. M.; Spector, M. J. Biomed. Mater. Res. A 2008, 84,
73-83.
[22] Rigby, P. G. Nature 1969, 221, 968-969.
[23] Antony Raj, P.; Dentino, A. R. Future Med. Chem. 2013, 14, 1635-
1645.
[24] Gudeman, L. F.; Peppas, N. A. J. Appl. Polym. Sci. 1995, 55, 919-928.
[25] Gua, S.; Qiao, Y.; Wang, W.; He, H.; Deng, L.; Xing, J.; Xu, J.; Liang,
X. J.; Dong A. J. Mater Chem. 2010, 20, 6935-6941.
[26] Dash, M.; Chiellini, F.; Ottenbrite, R. M.; Chiellini, E. Prog. Polym.
Sci. 2011, 36, 981-1014.
[27] Pillai, C. K. S.; Paul, W.; Sharma, C. P. Prog. Polym. Sci. 2009, 34,
641-678.
[28] Ping, Y.; Liu, C-D.; Tang, G-P.; Li, J-S.; Li, J.; Yang, W-T.; Xu, F-J.
Adv. Funct. Mater. 2010, 20, 3106-3116.
[29] Nimesh, S.; Thibault, M. M.; Lavertu, M.; Buschmann, M. D. Mol.
Biotechnol. 2010, 46, 182-196.
[30] Jia, Z.; Shen, D.; Xu, W.; Carbohydrate Res. 2001, 22, 1-6.
[31] Pei, L.; Cai, Z.; Shang, S.; Song, Z. J. Appl. Polym. Sci. 2014, 131,
40052-40059.
[32] Su, K.; Wang, C. Biotechnol. Lett. 2015, 37, 2139-2145.
[33] Samal, S. K.; Dash, M.; Vlierberghe, S. V.; Kaplan, D. L.; Chiellini,
E.; van Blitterwijk, C.; Mornoni, L.; Dubruel, P. Chem. Soc. Rev. 2012,
41, 7147-7194.
[34] Hamman, J. H. Marine Drugs 2010, 8, 1305-1322.
[35] Morimoto, K.; Chono, S.; Kosai, T.; Seki, T. Drug Delivery 2008, 15,
113-117.
[36] Xu, X.; Capito, R. M.; Spector, M. J. Biomed. Mater. Res. A 2008, 84,
73-83.
[37] Inada, S.; Fujiwara H.; Atsuji, K.; Takashima, K.; Araki, Y.; Kubota,
T.; Tabata, Y.; Yamagishi, H. Anticancer Res. 2006, 26, 1957-1963.
[38] Ovsianikov, A.; Deivick, A.; van Vlierberghe, S.; Dubruel, P.; Möller,
L.; Dräger, G.; Chickov, B. Biomacromolecules 2011, 12, 851-858.

[39] Jiang, Q.; Xu, J.; Li, T.; Qiao, C.; Li, Y. J. Macromol. Sci. B 2014, 53,
133-141.
[40] Tauhardt, L.; Kempe, K.; Knop, K.; Altunatas, E.; Jäger, M.
Macromol. Chem. Phys. 2011, 17, 1918-1924.
[41] Jones, G. D.; Langsjoem, A.; Neuman, M. M. C.; Zomlefer, J. J. Org.
Chem. 1994, 9, 125-147.
[42] Brissault, B.; Kichler, A.; Guis, C.; Leborgne, C.; Danos, O.;
Cheradame, H. Bioconjug. Chem. 2003, 14, 581-587.
[43] Lin, J.; Qiu, S.; Lewis, K.; Slibanov, A. M. Biotechnol. Prog. 2002,
18, 1082-1086.
[44] Neu, M.; Fischer, D.; Kissel, T. J. Gene Med. 2005, 7, 992-1009.
[45] Kim Y. H.; Park, J. Y.; Lee, M.; Kim, Y. H.; Park, T. G.; Kim, S. W.
J. Controlled Release 2005, 103, 209-219.
[46] Deka, S. R.; Sharma, A. K.; Kumar, P. Curr. Topic Med. Chem. 2015,
15, 1179-1195.
[47] Akinc, A.; Anderson, D. G.; Lunn, D. M.; Langer, R. Bioconjugate
Chem. 2003, 14, 979-988.
[48] Lynn, D. M.; Langer R. J. Am. Chem. Soc. 2000, 122, 10761, 10768.
[49] Anderson, D. G.; Akinc, A.; Hossain, N.; Langer, R. Mol. Ther. 2005,
11, 426-434.
[50] Sunshine, J. C.; Sunshine, S. B.; Bhutto, I.; Handa, J. T. PLoS One
2012, 7, 37543-37554.
[51] Danusso, F.; Ferruti, P. Polymer 1970, 11, 88-113.
[52] Barbucci, R.; Ferruti, P.; Improta, C.; Delfini, M.; Segre, A. L.; Conti,
F. Polymer 1978, 19, 1329-1334.
[53] Ferruti, P.; Marchisio, M. A.; Dunca, R. Macromol. Rapid Commun.
2002, 23, 332-355.
[54] Ferruti, P. J. Polym. Sci. Part. A: Polym. Chem. 2013, 51, 2319-2353.
[55] Siddiqui, N.; Andalip, S. B.; Ali, R.; Afzal, O.; Akhtar, M. J.; Azad,
B.; Kumar, R. J. Pharm. Bioallied. Sci. 2011, 3, 194-212.
[56] Lyons, E. T.; Tolliver, S. C.; Drudge J. H. Vet Parasitol. 1999, 85, 97-
112.
[57] Kerns, R. J.; Rybak, M. J.; Kaatz, G. W.; Vaka, F.; Cha, R.; Grucz, R.
G.; Diwadkar, V. U. Bioorg. Med. Chem. Lett. 2003, 13, 2109-2112.

[58] Kerns, R. J.; Rybak, M. J.; Kaatz, G. W.; Vaka, F.; Cha, R.; Grucz, R.
G.; Diwadkar, V. U.; Ward, T. D. Bioorg. Med. Chem. Lett. 2003, 13,
1745-1749.
[59] Zhao, X.; Quinn, B.; Kerns, R.; Drlica, K. J. Antimicrob. Chemo-
therapy 2006, 58, 1283-1286.
[60] Singh, K. K.; Joshi, S. C.; Mathela, S. Ind. J. Chem. 2011, 50B, 196-
200.
[61] Chaudhary, P.; Nimesh, S.; Ydav, V.; Verma, A. K.; Kumar, R. Eur.
J. Med. Chem. 2007, 42, 471-476.
[62] Ryckebusch, A.; Deprez-Poulain, R.; Maes, L.; Debreu-Fontaine, M.;
Mouray, E.; Grellier, P.; Sergheraert, C. J. Med. Chem. 2003, 46, 542-
557.
[63] Upadhayaya, R. S.; Sinha, N.; Jain, S.; Kishore, N.; Chadra, R.; Arora,
S. K. Bioorg. Med. Chem. 2004, 12, 2225-2238.
[64] Kimura, M.; Masuda, T.; Yamada, K.; Mitani, M.; Kubota, N.;
Kawakatsu, N.; Kishii, K.; Inazu, M.; Kiuchi, Y.; Oguchi, K.; Namiki,
T. Bioorg. Med. Chem. 2003, 11, 3953-3963.
[65] Bernheimer, H.; Birkmayer, W.; Hornykiewicz, O.; Jellinger, K.;
Seitelberger, F. J. Neurol. Sci. 1973, 20, 415-455.
[66] Brown, A. S.; Gershon, S. Gen. Sect. 1993, 91, 75-109.
[67] Carrol, F. L.; Howell, L. L.; Kuhar, M. J. J. Med. Chem. 1999, 42,
2721-2736.
[68] Zhang, L.; Cheng, L.; Hong, J. Pharmacology 2013, 92, 14-25.
[69] Dézsi, C. A. Am. J. Ther. 2016, 23, e871-e879.
[70] Apiquian, R.; Fresan, A.; Ulloa, R-E.; de la Fuente-Sandoval, C.;
Herrer-Estrella, M.; Vazquez, A.; Nicolini, H.; Kapur, S. Neuro-
psycho. Pharmacol. 2005, 30, 2236-2244.
[71] Boyd, M. Expert Opin. Pharmacother. 2007, 8, 957-964.
[72] Pettibone, D. J.; Williams, M. Biochem. Pharmacol. 1984, 33, 1531-
1535.
[73] Leone, M.; Attanasio, A.; Croci, D.; Filippini, G.; D’Amico, D.;
Grazzi, L.; Nespolo, A.; Bussone, G. Neurology 2000, 55, 136-139.
[74] Jahanshahi, M.; Rahimpour, A.; Peyravi, M. Desalination 2010, 257,
129-136.

[75] Religa, P.; Kowalik-Klimczak, A.; Giercycz, P. Desalination 2013,

315, 115-123.
[76] Religa, P.; Kowalik-Klimczak, A. Water Sci. Tech. 2015, 72, 1803-
1809.
[77] Singh, P.; Quraishi, M.A.; Ebenso, E.E. Int. J. Electrochem. Sci. 2013,
8, 10890-10902.
[78] Quraishi, M. A.; Singh, A.; Singh, V. K.; Yadav, D. K.; Singh, A. K.
Mater. Chem. Phys. 2010, 122, 114-122.
[79] Tan, Y., Shao, Z-B., Yu, L-X., Long, J-W, Qi, M., Chen, L., Wang, Y-
Z. Polym. Chem. 2016, 7, 3003-3012.
[80] Marchisio, M. A.; Longo, T.; Ferruti, P. Eur. J. Pharmacol. 1970, 12,
236-242.
[81] Marchisio, M. A.; Longo, T.; Ferruti, P. Experimentia 1973, 29, 93-
95.
[82] Ferruti. P.; Arnoldi, D.; Marchisio, M. A.; Martuscelli, E.; Palma, M.;
Riva, F.; Provenzale, L. J. Polym. Sci. Polym. Chem. Ed. 1977, 15,
2151-2162.
[83] Barbucci, R.; Benvenuti, M.; Casini, G.; Ferruti, P.; Tempesti, F.
Biomaterials 1985, 6, 102-105.
[84] Barbucci, R.; Ferruti, P. Polymer 1979, 20, 1061-1062.
[85] Ferruti, P.; Oliva, L.; Barbucci, R. Inorg. Chem. Acta 1980, 41, 25-29.
[86] Gan, L. H.; Roshan Deen, G.; Loh, X. J.; Gan, Y. Y. Polymer 2001,
42, 65-69.
[87] Roshan Deen, G.; Gan, L. H, J. Appl. Polym. Sci. 2008, 107, 1449-
1458.
[88] Loh, X. J.; Roshan Deen, G.; Gan, Y. Y.; Gan, L. H. J. Appl. Polym.
Sci. 2001, 80, 268-273.
[89] Rivas, B. L.; Pooley, A. S.; Luna, M. J. Appl. Polym. Sci. 2002, 83,
2556-2561.
[90] Roshan Deen, G.; Teo, T. W.; Lee, K. F. Polymer 2016, 104, 91-103.
[91] Roshan Deen, G.; Lim, E. K.; Mah, C. H.; Heng, K. M. Ind. Eng.
Chem. Res. 2012, 51, 13354-13365.
[92] Roshan Deen, G. Encyclopedia of Biomedical Polymers and Poly-
meric Biomaterials, Taylor and Francis, USA, 2014; Vol. 1, pp 1-10.

[93] Gan, L. H.; Roshan Deen, G.; Gan, Y. Y. Macromolecules 2000, 33,
7893-7897.
[94] Chang, L.; Kong, X.; Wang, F.; Wang, L.; Shen, J. Thin Solid Films
2008, 516, 2125-2129.
[95] Liu, F.; Jarrett, W. L.; Urban, W.M. Polym. Chem. 2011, 2, 963-969.
[96] Goh, S. H.; Liu, Y.; Lee, S. Y. Macromolecules 1999, 32, 8595-8602.

CONTENTS OF EARLIER VOLUMES
Advances in Materials Science Research. Volume 34
Chapter 1 Modelling of Plasma-Assisted Growth of

Vertically Aligned Carbon Nanostructures
Igor B. Denysenko, Kostya (Ken) Ostrikov
and Nikolay A. Azarenkov
Chapter 2 Rayleigh and Brillouin Scattering Spectroscopy
for Optical Material Science
Leonid V. Maksimov and Anatoli V. Anan’ev
Chapter 3 Computer Study of the Removal
of Heavy Metals from Graphene
Alexander Y. Galashev
Chapter 4 Flexible Resistive Switching Memories:
From Materials to Devices
Shuang Gao, Xiaohui Yi, Jie Shang, Bin Chen,
Gang Liu and Run-Wei Li

222 Contents of Earlier Volumes
Chapter 5 Improving the Octane Number of Gasolines

through the Transformation of n-Hexane
into Branched Isomers
Angela Martins and Nelson Nunes
Chapter 6 The Application of Strontium Dodeceyl
Sulfate, Sr(DS)2 in the Synthesis of 4,4'-
Diaminotriarylmethanes-Leuco
Malachite Materials
Najmeh Dashtaki and Farahnaz K. Behbahani
Chapter 1 High Strain Response in Pb-Free Bi0.5

(Na0.84K0.16)0.5TiO3–0.04SrTiO3
Piezoelectric Ceramics
Rizwan Ahmed Malik, Ali Hussain,
Adnan Maqbool, Soonil Lee and Myong-Ho Kim
Chapter 2 Cultured Cell-Derived Decellularized Matrices as
In Vitro Extracellular Matrix Models for
Comprehensive Research
Takashi Hoshiba, Naoki Kawazoe
and Guoping Chen
Chapter 3 Transmission Electron Microscopy Investigation
of Microbands, Micro-Shear Bands and
Transition Bands in Hot Deformed Austenite
Pavel Cizek
Chapter 4 Carbon Aerogel: Properties,
Preparation, and Applications
Soo-Jin Park, Yifan Zhang and Guijun Yang
Chapter 5 Biodegradable Protein-Tannic Acid Multilayers
Hooi Hong Lau, Chui Ping Ooi
and Maxim V. Kiryukhin

Chapter 6 Analysis of Stress Concentration in

Fiber Reinforced Polymer Composites
Néstor F. Ortega and Sandra I. Robles
Chapter 7 Thermal Diffusivity Characterization of
Dysprosium Zirconate, Cerate and Hafnate
Marta Mikuśkiewicz
Chapter 8 Temperature and Dislocations Distribution in
Liquid-Phase-Pulse-Impact Diffusion Welded
Joints of Particle Reinforcement Aluminum
Matrix Composites SiCp & Al2O3p/AlMg1SiCu
Kelvii Wei Guo
Chapter 9 Nonequilibrium Thermodynamics and
Kinetics of Metals Fragmentation during
Severe Plastic Deformation
Alexei Khomenko and Daria S. Troshchenko
Chapter 1 Graphene Oxide: A Great Platform

for the Synthesis of Highly Efficient
Electrocatalytic Nanocomposites
Hiep S. T. Pham, Peter J. Mahon and Aimin Yu
Chapter 2 The Thermal Control of Tribolab:
A Materials Experiment on Board the
International Space Station
Iñaki Garmendia and Eva Anglada
Chapter 3 Facsimile: How to Create It by Using
Non-Invasive Methods
Tanja Urbanc, Sabina Bračko
and Klementina Možina

Chapter 4 Novel Route for Rapid Sol-Gel Synthesis of

Bioglass and Bioceramics, Using Fast Drying
with a 100-Fold to 200-Fold Reduction in
Processing Time
Basam A. E. Ben-Arfa and Robert C. Pullar
Chapter 5 Diboronic Acids and Their Derivatives:
New Perspectives in Sensing and
Materials’ Chemistry
Agnieszka Adamczyk-Woźniak
and Andrzej Sporzyński
Chapter 1 Electromagnetic Metamaterials and

Their Applications for Designing Novel
Microwave Antennas
Yahong Liu, Kun Song and Xiaopeng Zhao
Chapter 2 Electrically Conductive Polymer Composites:
Preparation and Properties
Anett Király and Ferenc Ronkay
Chapter 3 Protective Composite Helmet Shells with
Continuous Textile Reinforcements
Xiaogang Chen and Shengnan Min
Chapter 4 Optimization of the Vibration Isolation
Performance of Machines Using
Carbon Nanotube Reinforced Elastomers
G. Mansour, K. Tsongas and D. Tzetzis
Chapter 5 Cu – Sb Based Chalcogenide Thin Films:
An Alternative Absorber Material for
Photovoltaic Applications
Vineetha Vinayakumar, Sadasivan Shaji,
David Avellaneda and Bindu Krishnan

Chapter 6 Carbon Fiber/Epoxy Structural Composites

Infused with Graphene Nanoplatelet
Mohammad K. Hossain,
Mahmudur R. Chowdhury
and Nydeia W. Bolden
Chapter 7 Temperature Dependent Properties of CF/FF
and BF/FF Reinforced Hybrid Epoxy Composites
D. Luca Motoc
Chapter 8 Terahertz Kinetics Spectroscopy for
Laminated Helmet Materials Characterization
Anis Rahman and Aunik K Rahman
Chapter 1 Transfer Processes at the Manufacture

of Metal Matrix Composite Materials
Ioan Carcea and Raluca Maria Florea
Chapter 2 Characterization, Manufacturing and
Application of Metal Matrix Composites
Blaža Ž. Stojanović and Saša T. Milojević
Chapter 3 Geocells Reinforcement Pavement
Samo Peter Medved, Bojan Žlender
and Stanislav Lenart
Chapter 4 Conditions for Necking in Neat
and Filled Polyolefines
Maria A. Guseva, Victor A. Gerasin,
Boris F. Shklyaruk and Vadim A. Dubinskiy

Chapter 5 The Application of Fibrous Ion-Exchange

Sorbents for Water Treatment
and the Purification of Gaseous Mixtures
Renat Khaydarov, Rashid Khaydarov,
Olga Gapurova, Ilnur Garipov, Lutfi Firdaus
and Emiliya Nikolova

INDEX
amine, xii, 8, 76, 77, 196, 197, 199, 201,

A
203, 204, 208, 211, 213
amine group, 199, 211
acetaldehyde, 183
amino acid, 200
acetic acid, 183
ammonium, x, 120, 121, 137, 147, 201, 207
acetone, 183
annealing, 173
acrylic acid, 213
antidepressant, 206
acrylonitrile, 120
antioxidant, 199
activated carbon, 130
application, 225, 226
additives, 37
aqueous solutions, xii, 137, 195, 196
adhesion, 18, 31, 34, 199
arginine, 200
adhesion properties, 31
aromatic rings, 95
adhesion strength, 34
arsenic, x, 120, 135, 146
adhesives, 76
asparagines, 200
adsorption, ix, 80, 87, 88, 89, 90, 92, 94, 99,
aspartate, 200
101, 103, 104, 105, 106, 124, 130, 131,
atmosphere, ix, 80, 85, 86
139, 140, 143, 186, 213
atoms, 80, 132, 152, 163, 166, 167, 168,
adsorption isotherms, 99, 101, 104, 105
169, 183
advancements, viii
aerogels, x, 120, 121, 126, 144, 146
aerospace, vii, viii, 1, 2, 12, 72, 75, 77 B
Al2O3p, 223
AlMg1SiCu, 223 Bacillus subtilis, 204
alternative energy, 178 bacteria, 201
aluminium, 10 bacterial cells, 200, 204
aluminum matrix composites (AMCs), 223 bacterium, 201

228 Index
basicity, 208 carbon nanostructures, 221

binding energy, 213 carbon nanotubes, 78, 152
biocompatibility, 201 carbonization, ix, 80, 82, 85, 86, 87, 92, 97,
biodegradability, 201 104, 105, 106, 107, 110, 111, 112, 175
biodegradable, 222 carbonyl groups, 95
biofuel, 180, 184 carboxyl, 200, 203
biogas, 178 case study, xi, 178, 182, 183, 187
bioglass, 224 catalyst deactivation, 182
biomass, 179 catalytic activity, 81, 110, 111, 112
biomaterials, 214 catalytic effect, 110
biomolecules, xii, 196, 197 catalytic properties, 111
biopolymers, 196 C-C, 85, 98, 153, 154
bisphenol, 8 cesium, x, 120, 134, 142, 146, 147
block polymers, xii, 196 chemical properties, 80
BMI, viii, 1, 6, 7, 8, 9, 10, 14, 16, 19, 20, chemical reactions, 82, 85
21, 22, 23, 24, 25, 26, 28, 29, 30, 31, 32, chemical vapor deposition, 81
33, 34, 35, 37, 38, 39, 40, 41, 42, 43, 44, chemisorption, vii, x, 119
45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, chirality, 175
56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, chitosan, 199, 200
67, 68, 69, 70, 71, 72 chromatography, 125
bonds, 6, 36, 49, 85, 95, 107, 153 chromium, 207
Brillouin, L., 221 cleavage, 22, 34
CO2, vi, viii, xi, 86, 91, 94, 99, 101, 102,
103, 104, 105, 106, 177, 178, 179, 180,
C
181, 182, 183, 184, 185, 186, 187, 188,
189, 190, 191
Ca2+, 211
coal, 188
calcination temperature, 93, 110, 112
cobalt, 122
calcium, 122, 181, 186
cocaine, 205
cancer, 126
collage, 127
caprolactam, 212
collagen, 200
carbohydrates, 196
commercial, 5, 8, 130, 136, 142
carbon, vii, viii, ix, xii, 1, 2, 6, 7, 8, 9, 31,
community, 134, 143, 180
34, 72, 74, 75, 76, 77, 78, 79, 80, 81, 82,
compatibility, 201
85, 86, 87, 88, 92, 93, 95, 98, 99, 101,
competition, 125
103, 104, 105, 106, 107, 111, 112, 139,
compilation, viii
152, 153, 161, 166, 167, 175, 176, 178,
composite, 224
179, 180, 182, 183, 184, 187, 190
composites, vii, viii, 1, 2, 3, 4, 5, 6, 7, 10,
carbon aerogel, 222
16, 18, 19, 20, 22, 23, 24, 25, 26, 28, 31,
carbon atoms, 80, 153
35, 37, 39, 40, 42, 43, 45, 46, 47, 53, 55,
carbon dioxide, 87, 179, 180
60, 61, 65, 69, 70, 71, 72, 73, 74, 76, 77,
carbon materials, 92, 98, 103
78, 122, 126, 127, 129, 132, 134

Index 229
composition, 106, 107, 108, 112, 125, 126,

D
133, 151, 182, 183, 186, 212
compounds, 6, 134, 144
decay, 137, 154
comprehensive research, 222
deconvolution, 98, 108, 109
compression, ix, 2, 7, 14, 22, 46, 63, 67, 69,
defects, 4, 5
71, 72
deformation, 45, 55, 66, 103, 150, 163
condensation, 103
degradation, 24, 26, 36, 38, 49, 55, 98, 186
conductivity, ix, 80, 110
Department of Energy, 138
configuration, 16, 17, 29, 30, 35, 41, 51, 52,
deposition, 174, 182
63
derivatives, 200, 204, 205, 206, 208, 209,
confinement, 160, 162, 169
210, 211
connective tissue, 200
desorption, 88, 124, 131
constituents, 137, 140
detachment, 151, 154, 155, 156, 157, 158,
consumption, 179, 185
159, 160, 161, 162, 163, 164, 166, 169
contaminated water, 135, 136
DGEBA, 8
contamination, 150, 171
diamines, 208
continuous, 224
diboronic acids, 224
copolymer, 77, 78, 199, 211, 212
differential scanning calorimeter, 8
copolymerization, 197, 212, 213
diffraction, 124
copolymers, xii, 195, 212
diffusion, 18, 22, 33, 34, 45, 91, 164, 165
copper, 82, 122, 142, 151, 173, 175
diffusion process, 18, 34, 45
corrosion, 2, 207
diffusivity, x, 120
cosmetics, xii, 196
dimethylformamide, vii, ix, 80, 81, 92
cost effectiveness, 135
displacement, 11, 13, 19, 22, 23, 27, 30, 31,
cotton, 80
36, 43, 44, 47, 48, 49, 154, 167, 168
cracks, 5, 14, 24, 35, 48, 66, 68, 70, 72
dissociation, 199, 211
creep, 55, 78
distillation, 130
crystal structure, 137
distribution, 33, 90, 99, 101, 104, 105, 124
crystalline, 124
diversification, 178
curing cycle, 15, 18, 21, 24, 25, 26, 27, 28,
DMF, ix, 80, 81, 92, 113
46
dopamine, 205
CVD, 81
dosing, 181
cyanide, 108
double bonds, 95
cycles, 12, 14, 23, 24, 25, 26, 27, 28, 35, 36,
drinking water, 121, 134
42, 46, 47, 48, 49, 50, 54, 56, 57, 58, 59,
drug delivery, 196, 199, 202
60, 61, 62, 72, 133, 181, 186
drug release, 199
cycling, 132, 133
DSC, 8, 15, 18, 25, 213
cyclopentadiene, 6
DSC method, 213
durability, 182, 191
dynamic viscosity, 156

230 Index
E F
elastic deformation, 66 facsimile, 223

electric charge, 197 fiber reinforced polymer, 223
electric field, 82, 84 fibers, 80, 83, 97, 130, 139
electrical conductivity, ix, 80, 83, 84 films, 77, 150, 152, 154, 155, 157, 159, 160,
electrocatalytic, 223 161, 163, 164, 165, 166, 167, 168, 169,
electrodes, 82, 125 170, 171, 172, 173, 213
electrolyte, 110, 179 fission, 136
electron(s), 87, 108, 124, 125, 126, 153, flexible, 221
127, 207 fluid, 82, 84, 123, 139, 172
electrospinning, vii, ix, 80, 81, 82, 83, 84, force, 11, 12, 14, 81, 82, 103, 154, 155, 156,
92, 111 157, 158, 161, 168, 169, 171
elucidation, 87 formaldehyde, 137, 146
embedded atom method, 152 formamide, 121
emission, 5 formation, xii, 4, 29, 46, 66, 83, 84, 85, 86,
endothermic, 180 96, 108, 151, 159, 169, 176, 178, 179,
energy, 2, 10, 11, 19, 24, 27, 31, 32, 36, 37, 183, 184, 186, 187, 190, 196, 197
44, 45, 46, 49, 66, 87, 94, 125, 145, 150, formula, 121, 138
153, 154, 155, 159, 162, 163, 178, 180, fracture toughness, vii, viii, 1, 3, 7, 10, 11,
183, 189 18, 19, 20, 21, 24, 25, 27, 30, 31, 32, 33,
energy density, 162 34, 35, 36, 37, 40, 41, 42, 43, 44, 45, 48,
engineering, 2, 3, 136, 150 49, 54, 72
environmental variables, 83, 84 fragments, 98
epoxy polymer, 75, 76, 78 free energy, 182, 183
epoxy resins, 5 FRP, 39
equilibrium sorption, 146 FTIR, 94, 95, 103, 112, 213
ESR-CO2-ABS, viii, xi, 178, 180, 181, 182, fuel cell, ix, 80, 82, 87, 110, 178, 179
184, 185, 186, 187, 189, 190, 191 furan, 76, 77
ester, 201, 202, 203, 208
ethanol, viii, xi, 178, 179, 180, 182, 183,
G
184, 185, 187, 188, 190, 191
ethylene, 121, 183
geocells, 225
evaporation, 83, 84, 85, 181
geometrical parameters, 169
evolution, 2, 5, 41, 43, 50, 55, 56, 57, 58,
geometry, 18, 39, 40, 51, 172, 174
59, 60, 61, 62, 87, 156, 159, 164, 166,
Gibbs energy, 182
172
glass transition, 172, 196
exoskeleton, 199
glutamate, 200
experimental condition, 182
glutamine, 200
external environment, xii, 196, 197
glycine, 209
extrusion, x, 8, 119
granules, 127, 128, 129, 130, 132

Index 231
graph, 90
I
graphene, 221
graphene oxide, 223
image analysis, 98
graphite, 78, 80, 98, 108, 173, 174, 176, 188
immobilization, 126, 143
growth rate, 207
Impact Assessment, 62
impact energy, 13, 65, 66
H impregnation, 9
inertia, 151, 163, 168, 169
healing, vii, viii, 1, 2, 5, 6, 7, 12, 13, 14, 15, inertial effects, 158
18, 20, 22, 23, 24, 25, 26, 27, 28, 33, 35, infrared spectroscopy, 97
36, 37, 39, 42, 46, 47, 48, 49, 50, 54, 60, inhibition, 207
61, 62, 63, 67, 68, 70, 71, 72, 73, 74, 75, inhibitor, 206, 207
76, 77, 78 Instron, 10, 12, 13
heating rate, 95 integration, 9, 18, 152
heavy metals, 134 intermolecular interactions, 175
helmet, 224, 225 International Atomic Energy Agency, 145
HIV, 206 International Space Station, 223
homopolymers, xii, 195, 212 iodine, 125, 127, 128, 129, 143, 144
hot deformed austenite, 222 ion-exchange, 137, 139, 142, 147
human body, 212 ionization, 198
human health, 136 ions, 84, 90, 137, 199, 207, 209, 211
humidity, 83, 84 iron, 134, 146
hybrid, 74, 139 irradiation, 151, 172
hybridization, 3, 80 isotherms, 99, 100, 101, 102, 103, 104, 105,
hydrocarbons, 175, 179 106, 124, 135
hydrogels, xii, 196, 211, 212, 213
hydrogen, viii, x, xi, 81, 96, 108, 120, 121,
K
130, 139, 178, 179, 180, 181, 184, 185,
186, 187, 188, 189, 190, 191, 192, 213
kinetics, xii, 178, 186, 187, 191
hydrogen bonds, 213
krypton, 133, 139, 140, 146
hydrolysis, 201
hydrophilicity, 212
hydroxide, 146 L
hydroxyl, 203
hydroxyl groups, 203 lamination, 18, 39
hysteresis, 55, 56, 57, 58, 59, 60, 61, 78, 88, laser ablation, 81
89, 99, 103, 163, 170 lasers, 130
hysteresis loop, 55, 56, 57, 58, 59, 60, 61, life cycle, 61, 73
78, 103 light, xii, 127, 170, 196
linear molecules, 85
linear polymers, xii, 196

232 Index
liquid fuels, 178 modulus, 8, 28, 38, 54, 55, 56, 57, 58, 59,
liquid phase, 166 61, 62, 69, 71
liquid-air interface, 153 mole, 129, 183, 184, 185, 188, 190
liquids, vii, x, 119, 150, 163, 165, 170, 174 molecular dynamics, viii, xi, 149, 150, 152,
lithium, 121, 181, 183, 186 170
low temperatures, 25 molecular mass, 83, 92, 152, 201
lysine, 200 molecular structure, 6
molecules, 89, 90, 103, 175
molybdenum, 174
M
monolayer, 89
monomers, xii, 196, 197, 203
macropores, 89, 103
mordenite, x, 120, 121, 122, 130, 133, 139,
magnetic field, xii, 196
140
manganese, 134, 146, 174
morphology, ix, 2, 7, 83, 84, 97, 135, 171,
manipulation, ix, 79, 150, 173
172
manufacturing, 9, 14, 16, 29, 39, 71, 73, 130
mortality, 136
mass, x, 2, 120, 123, 125, 127, 128, 129,
multilayer films, 213
130, 131, 133, 134, 135, 151, 154, 159,
municipal solid waste, 179
167, 168, 182, 183
mussels, 171
materials science, viii, xii, 195
matrix, vii, x, 3, 18, 34, 37, 45, 46, 53, 71,
120, 124, 126, 136, 139, 142, 225 N
mechanical properties, 2, 28, 53, 72
melting temperature, 8, 196 Na+, 137
membranes, 207 Na2SO4, 212
metal ion, 207, 211 NaCl, 212
metal matrix composite, 225 nanocomposites, 74, 223
metal nanoparticles, 172 nanofabrication, 173
metal oxides, x, 120, 134 nanofibers, vii, ix, 79, 80, 81, 82, 83, 84, 85,
metal salts, 134 86, 87, 88, 92, 95, 96, 97, 98, 99, 103,
metals, xi, 3, 126, 150, 165, 169, 174, 175 104, 105, 107, 110, 111, 112
methodology, 11, 89, 93 nanomedicine, 196
methyl methacrylate, 211 nanometers, 80
microbands, 222 nanoparticles, 145, 173, 176
microgels, 196 nanoribbons, 175
microporous materials, 89 nanostructured materials, 98
microscopy, 22, 25, 53, 64, 66, 68 nanostructures, 150, 151, 169, 221
micro-shear bands, 222 nanotechnology, 151
microstructures, 81, 171, 176 natural gas, 178
Ministry of Education, 149 necrosis, 200
MMA, 211 neurotransmitter, 205
modelling, 221 neutral, 121

Index 233
nickel, 122, 147 pH, xii, 135, 138, 196, 197, 198, 199, 200,
niobium, 174 202, 203, 207, 211, 212, 213
nitrates, 134 pharmaceuticals, xii, 196, 201, 205, 207
nitrogen, ix, 80, 81, 82, 86, 87, 90, 91, 95, phosphate, 122, 136
99, 103, 105, 106, 107, 108, 109, 110, phosphates, xi, 120, 121, 136
111, 112, 136 photographs, 21, 25, 26, 27, 63, 64, 65, 68
noble gases, 146 physical properties, 6
NRC, 144, 145 plants, 145, 179
nucleic acid, 196, 197, 201 plastic deformation, 22, 34, 45, 46, 66
plastics, vii, viii, 1
platform, 144, 145
O
platinum, 180
plutonium, 143
oligomers, 8
polarization, 132
operations, 130, 132
pollution, 136
optical microscopy, 21, 26, 27, 34, 45, 46,
polyacrylonitrile fiber, 139
51, 63, 68, 124
polyamine, 202
optical properties, 196
polyelectrolyte complex, 201
optimization, 4
polymer, viii, x, xii, 1, 2, 3, 4, 6, 8, 9, 10,
ores, 163
16, 22, 25, 31, 33, 34, 35, 41, 45, 66, 71,
organic matter, 179
72, 74, 75, 76, 77, 81, 82, 85, 92, 104,
organic solvents, 121
120, 124, 126, 134, 142, 146, 150, 171,
oxidation, 85, 96, 179
172, 174, 179, 195, 196, 197, 198, 199,
oxygen, vii, x, 80, 81, 82, 85, 86, 106, 110,
201, 203, 207, 208, 212, 213, 214
111, 136, 180
polymer chain, 8, 10, 85
polymer composites, 4
P polymer films, 150, 171, 172
polymer materials, viii, xii, 195
PAA, 203, 208, 209, 211, 213 polymer matrix, 2, 3, 134
PAN, v, ix, x, 79, 80, 81, 85, 87, 92, 95, 96, polymer networks, 212
106, 107, 108, 109, 111, 113, 114, 115, polymer solutions, 81
117, 120, 121, 122, 123, 124, 126, 127, polymer systems, xii, 76, 196, 207
128, 129, 130, 131, 132, 133, 134, 136, polymeric materials, xii, 76, 196
137, 138, 139, 140, 142, 147 polymerization, 25, 76, 197, 201, 203
parallel, 15, 16, 81, 152 polystyrene, 207
parasites, 204 porosity, 14, 112, 126, 132
particle, 223 potassium, 147
particle morphology, 74 propagation, 14, 19, 20, 22, 31, 32, 34, 43,
particle reinforcement, 223 44, 45, 46
pathway, 203 protective, 224
petroleum, 134 proteins, 196, 197, 222
protons, 198

234 Index
Pseudomonas aeruginosa, 205 silica, 126, 144

PTFE, 10, 16, 29, 39 silicon, 8, 174
pumps, 181 silver, x, 120, 121, 126, 132, 133, 140, 143,
purification, 92, 134, 138, 145 144, 145, 146
purity, viii, xi, 92, 178, 179, 180, 181, 187, sintering, 151, 174, 181, 186
188, 190, 191 sodium, 121, 133, 181, 183
PVA, 207 solid polymers, 196
pyrolysis, 81 solid surfaces, 150, 171
solidification, 84
solvents, 81, 133
Q
sorption, x, 120, 125, 133, 147
specific surface, vii, x, 80, 89, 99, 105, 111,
quaternary ammonium, 199, 200
119, 124
spectroscopy, 87, 125
R spontaneous abortion, 142
Sr(DS)2, 222
radiation, 94, 137, 142 SSA, x, 119, 124, 130, 133, 134, 135
radio, 81, 139, 143 stabilization, ix, 80, 82, 85, 86, 92, 94, 95,
radiography, 5 97, 98, 99, 100, 102, 105, 107, 112, 139
Raman spectra, 94, 98, 100 standard deviation, 44, 69
raw materials, 8, 180 steel, 207
Rayleigh, 221 stress, 11, 13, 45, 53, 54, 55, 56, 58, 60, 69,
reaction mechanism, vii, viii, 1, 6, 72, 85 70, 71, 78
reaction temperature, 184 strong interaction, 81, 132, 160
resorcinol, 137, 146 substrate, 6, 151, 153, 154, 155, 158, 160,
rubber, 4, 74, 172 163, 164, 165, 166, 167, 168, 169, 170,
176
S sulfate, 122
surface area, ix, 21, 26, 27, 80, 99, 101, 105,
salts, xii, 122, 147, 196, 200, 207, 212 106, 121, 124
SANS, 114 surface energy, 150, 159
scaling law, 151 surface modification, 171
scattering, 221 surface structure, 151, 164
selenium, 146 surface tension, 83, 84, 151
self-assembly, 151, 173, 174 surfactants, viii, xii, 174, 195, 212
self-consistency, 90 synthesis, xi, xii, 8, 81, 111, 112, 120, 123,
self-repair, 73 133, 145, 166, 196, 197, 201, 203, 204,
SEM micrographs, 132 207, 223, 224
semiconductors, 134 synthetic polymers, 196
sensors, xii, 196
SiCp, 223

Index 235
T U
tannic acid, 222 ultrasound, 93

temperature, viii, ix, x, xi, xii, 6, 8, 9, 14, uniaxial tension, 175
18, 22, 24, 25, 26, 46, 80, 82, 83, 84, 88,
92, 94, 95, 97, 98, 99, 102, 103, 104,
V
105, 107, 110, 111, 112, 120, 121, 125,
131, 132, 133, 137, 150, 151, 152, 158,
vacuum, 9, 18, 30, 39, 51, 63, 93, 123, 128,
161, 163, 169, 172, 179, 181, 182, 183,
129, 130
185, 186, 187, 188, 192, 196, 197, 211,
vanadium, 174
212, 213
velocity, ix, 2, 10, 11, 12, 13, 66, 82, 83, 84,
tendon, 200
152, 153, 154, 155, 156, 159, 172
tensile strength, 8, 11, 12, 54
viscosity, 9, 24, 36, 48, 83, 84, 158, 174
tension, ix, 2, 7, 10, 11, 12, 51, 54, 57, 59,
volatility, 83, 84
60, 61, 84
textile reinforcements, 224
TGA, 186 W
thermal deformation, 81
thermal resistance, 80 water purification, 207
thermal stability, 187, 191 water supplies, 134
thermal treatment, 97, 107 water treatment, 226
thermodynamic calculations, 183 water vapor, 184, 189
thermodynamic equilibrium, 179, 180, 185, waterways, 136
186, 187 weight ratio, ix, 80
thermodynamics, 191, 223 wettability, 150, 152, 163, 165, 166, 167,
thermogravimetric analysis, 94 169
thermoplastics, 10, 45 wetting, 165, 166, 170, 171, 174, 176
thermosets, 10, 76, 77
thin films, 152 X
tin, 121
TiP, 122 X-axis, 168
titanate, 122 xenon, 130, 140, 145, 146
toxic metals, 121 X-ray diffraction, 124
transducer, 14 X-ray photoelectron spectroscopy (XPS),
transfection, 199, 202 87, 94, 106, 107, 108, 109, 110
transfer processes, 225
transition bands, 222
transition temperature, 81, 92, 212 Z
transmission electron microscopy (TEM) ,
zeolites, 130, 132, 143, 146
94, 222
zinc, 121, 122
tungsten, 174

The Advantages of Thermodynamic Modeling in Hydrogen Production

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Advantages of Thermodynamic Modeling in Hydrogen Production

Uploaded by

Copyright:

Available Formats

Complimentary Contributor Copy

Complimentary Contributor Copy

Additional books and e-books in this series can be found on Nova’s

Complimentary Contributor Copy

Complimentary Contributor Copy

NOTICE TO THE READER

Library of Congress Cataloging-in-Publication Data

ISBN:  H%RRN

Published by Nova Science Publishers, Inc. † New York

Complimentary Contributor Copy

Complimentary Contributor Copy

Chapter 4 Dewetting Behavior of Liquid Metal Film

Complimentary Contributor Copy

Advances in Materials Science Research. Volume 35 opens with an

Complimentary Contributor Copy

Complimentary Contributor Copy

addition, BMI interleaves and prepregs were further utilized to modify

Complimentary Contributor Copy

capacity of these materials to promote oxygen reduction reaction (ORR) is

Complimentary Contributor Copy

Phosphate Sponge (TPS) composite, which is being used to remove

Complimentary Contributor Copy

(stoichiometric) to 6:1. Results indicate that S/EtOH ≤ stoichiometric

Complimentary Contributor Copy

Vassilis Kostopoulos* and Athanasios Kotrotsos, PhD

Bismaleimides (BMIs) exhibit healing functionalities on polymer

Complimentary Contributor Copy

composite laminated structure. During testing, these modified samples

Keywords: composites, self-healing, diels-alder, bismaleimides, fracture

During the last decades, the use of composites in aeronautics, wind

Complimentary Contributor Copy

limitation of composites is their poor interlaminar strength and fracture

Figure 1. Schematic representation of the importance of four classes of materials

Complimentary Contributor Copy

optimization of staking sequence, reactive rubbers and edge cap

Figure 2. Damage modes appearing into polymer composites. Indentation, impact,

However, although the interlaminar fracture properties have been

Complimentary Contributor Copy

Complimentary Contributor Copy

research work, a typical way to classify self-healing systems is as: (a)

Complimentary Contributor Copy

investigation is analytically presented in section 4 of the current chapter.

2.1. Materials and Methods

2.1.1. Raw Materials and Self-Healing Agent (SHA) Synthesis

Complimentary Contributor Copy

Complimentary Contributor Copy

Complimentary Contributor Copy

2.1.3. Mode I Interlaminar Fracture Toughness Testing

Complimentary Contributor Copy

2.1.4. Mode II Interlaminar Fracture Toughness Testing

where, d is the cross-head displacement at crack delamination onset, P is the

2.1.5. Open-Hole Tensile Tensing

Complimentary Contributor Copy

aerospace practice to develop notched design allow-able strengths based on

where, σ is the ultimate open-hole (notched) tensile strength, P is the

2.1.6. Open-Hole Fatigue Testing

2.1.7. Low Velocity Impact and Post-Impact Testing

Complimentary Contributor Copy

2.1.8. Three-Point Bending Testing

Complimentary Contributor Copy

2.1.9. Composites Quality Issues

2.1.10. Healing Procedure and Healing Efficiency Calculations

Complimentary Contributor Copy

3. TOUGHENING AND HEALING OF CONTINUOUS

ISBN: H%RRN