5566 PDF

ABSTRACT
ZHAO, JUNJIE. Metal-Organic Framework Thin Films on Fibers Enabled by Atomic Layer
Deposition. (Under the direction of Dr. Gregory N. Parsons).
Metal-organic frameworks (MOFs) are crystalline porous materials that consist of metal
clusters and organic bridging linkers. With ultra-high surface area and large porosity, MOFs
are promising for many applications such as gas adsorption, separations and catalysis.
However, the powder form of MOFs obtained from common solvothermal synthesis requires
further handling and manipulation, and is difficult to be integrated into functional materials
and devices for applications such as gas filtration, membrane separations and chemical
sensing. Surface-immobilized MOF thin films, especially on natural and synthetic fiber
scaffolds, could simplify the deployment, enable new device configurations, and expand the
applications of MOFs.
This research is motivated to develop new methods to synthesize MOF thin films on
polymer fibers and other functional substrate materials, and to understand the growth
mechanism of MOF thin films on functionalized surfaces. Metal oxide thin films including
Al2O3, ZnO and TiO2 deposited via atomic layer deposition (ALD) were used as nucleation
layers or templating materials for MOF thin films. Both macroscopic uniformity and
microscopic conformality of the MOF coatings are improved compared to previous methods.
The functional properties of MOFs are also fully maintained, as the MOF-coated fibers show
excellent performance for capturing toxic industrial chemicals and catalytically degrading
chemical warfare agents. The methods developed in this work can also be generally applied
to a wide range of MOFs and fibrous substrates.

Through an extensive investigation for the reactions of MOF precursors on ALD metal
oxide surfaces, we also discovered an ultra-fast reaction path for MOF formation at room
temperature. Hydroxy double salts (HDSs) converted from ZnO were found as key
intermediate species to enable the rapid formation of MOFs via anion exchange. The
space-time-yield of this process is over 3×104 kg∙m-3∙d-1, a significant enhancement for
advancing the industial applications and commertialization of MOFs. New MOF-based
composites can also be synthesized using this facile convertion from HDSs to MOFs.
The synthesis routes we have developed represent breakthroughs for MOF thin films and
scalable MOF production. Several applications of MOF-functionalized fibrous materials in
gas adsorption and heterogeneous catalysis will be presented in this dissertation. The
synthesis method and growth mechanism of milimeter-scale HKUST-1 [Cu3(BTC)2] single
crystals and the ZnO 3D hierarchical nanostructures for photocatalysis will also be discussed.
© Copyright 2016 by Junjie Zhao
All Rights Reserved

Metal-Organic Framework Thin Films on Fibers Enabled by Atomic Layer Deposition
by
Junjie Zhao
A dissertation submitted to the Graduate Faculty of

North Carolina State University
in partial fulfillment of the
requirements for the Degree of
Doctor of Philosophy
Chemical Engineering
Raleigh, North Carolina
2016
APPROVED BY:
_____________________________ ___________________________
Dr. Gregory N. Parsons Dr. Saad A. Khan
Chair of Advisory Committee
_____________________________ ___________________________
Dr. Michael D. Dickey Dr. Chih-Hao Chang
DEDICATION
To my parents, Jiannong Zhao and Xueying Wu,
and my fiancée Yizhou Chen
ii
BIOGRAPHY
Junjie Zhao was born in Suzhou, Jiangsu Province, China in 1988. He obtained a
Bachelor of Engineering degree in Chemical Engineering & Techonology as well as a
Bachelor of Arts degree in English from Zhejiang Univeristy, Hangzhou, China in June 2011.
In August 2011, Junjie was enrolled in the Ph.D. program in the Department of Chemical &
Biomolecular Engineering at North Carolina State Univeristy. His research focus is on
metal-organic framework (MOF) thin films, MOF-functionalized polymer fibers for
adsorption and catalysis, scalable rapid room-temperature synthesis for MOFs and the
application of atomic layer deposition for nanostructures. During his Ph.D. studies, Junjie has
achieved the American Vacuum Society Thin Film Division Graduate Student Award, the AIF
Annual Best Paper Award at NCSU and the 2nd Place Oral Presentation in 2016 Schoenborn
Graduate Research Symposium. He also visited the National Institute of Standards and
Technology as a guest researcher for research collabration in 2016. He will continue his
research career as a postdoctoral scholar at Massachusetts Institute of Technology in
Semptember 2016.
iii
ACKNOWLEDGEMENT
I am very grateful to my advisor, Prof. Gregory N. Parsons, for his tremendous support
and guidance for my research and helpful advice for my career development. I also
appreciate the help and inspiring comments from the members of my committee, Prof. Saad
A. Khan, Prof. Michael D. Dickey and Prof. Chih-Hao Chang.
I would like to thank former members in Parsons research group, Dr. Mark Losego, Dr.
Bo Gong, Dr. Do Han Kim, Dr. Kyoungmi Lee, Dr. Christina McClure, Dr. Will Sweet, Dr.
Berc Kalanyan, Dr. Sarah Atanasov, Dr. Philip Williams, Dr. Moataz Mousa for their
mentoring and support. Dr. Mark Losego helped me a lot in both research and technical
writing. I am thankful to current group members, Dr. Chris Oldham, Dr. Alex Brozena, Mike
Maniti, Paul Lemaire, Zack Mundy, Erinn Needham, James Daubert, Eric Stevents, Jennifer
Ovental, Wenyi Xie, Mariah Ritz, Dennis Lee and Heather Barton, for their helpful
discussion and research collaboration. I also want to thank Trent Blevins, Fahim Sidi,
William Nunn, Anqi Qu and Ian Woodward, for their help in the past projects.
I also would like to acknowledge the staff at the analytical instrumentation facility (AIF)
at NCSU, Roberto Garcia, Ching-Chang Chung, Chuck Mooney, Yang Liu, Xiahan Sang and
Chuanzhen Zhou for their training and help with some characterization. I appreciate the
research collaboration with Gregory Peterson (ECBC), Dr. Howard Walls (RTI), Dr. Sarah
Shepherd (RTI), Robert Yaga (RTI), Dr. Pamela Chu (NIST), Dr. Brent Sperling (NIST), Dr.
M. Douglas LeVan (Vanderbilt), Trent Tovar (Vanderbilt), Dr. Michael Dickey, Yiliang Lin,
Dr. Chih-hao Chang, Abhijeet Bagal and Xu Zhang.
Last but not least, I am very grateful to my family. Without your support this would not
iv
be possible.
v
TABLE OF CONTENTS
LIST OF TABLES ................................................................................................................................ ix

LIST OF FIGURES ................................................................................................................................ x
CHAPTER 1. Introduction – Synthesis, Characterization and Applications of Metal-Organic
Framework Functionalized Polymer Fibers ........................................................................................... 1
1.1 Introduction .................................................................................................................................. 2
1.2 Synthetic Methods ........................................................................................................................ 5
1.2.1 Solvothermal Synthesis ......................................................................................................... 5
1.2.1.1 Direct Solvothermal Growth .............................................................................................. 5
1.2.1.2 Solvothermal Synthesis on Chemically Functionalized Surface ........................................ 6
1.2.1.3 Solvothermal Synthesis on MOF-Seeded Surface.............................................................. 8
1.2.1.4 Solvothermal Synthesis on Atomic Layer Deposited Nucleation Layers .......................... 9
1.2.2 Layer-by-Layer Synthesis ................................................................................................... 14
1.2.2.1 Direct Layer-by-Layer Synthesis ..................................................................................... 14
1.2.2.2 Layer-by-Layer Synthesis on Chemically Functionalized Surface .................................. 16
1.2.2.3 Layer-by-Layer Synthesis on Atomic Layer Deposited Nucleation Layers ..................... 19
1.2.3 MOF Conversion from Thin Film Templates ..................................................................... 21
1.2.3.1 Conversion from ALD Metal Oxide Thin Films .............................................................. 21
1.2.3.2 Room-Temperature Conversion from Hydroxy Double Salts .......................................... 23
1.2.4 Electrospinning .................................................................................................................... 25
1.2.5 Co-Extrusion Spinning ........................................................................................................ 33
1.2.6 Inkjet Printing ...................................................................................................................... 35
1.2.7 Spay Drying Method ........................................................................................................... 36
1.3 Characterization Methods ........................................................................................................... 39
1.3.1 Attenuated Total Reflection (ATR) Infrared spectroscopy ................................................. 39
1.3.2 UV-Vis Spectroscopy .......................................................................................................... 41
1.3.3 Analysis of MOF Mass Fraction ......................................................................................... 42
1.3.4 Permeation Test ................................................................................................................... 44
1.3.5 Surface Roughness Analysis ............................................................................................... 46
vi
1.3.6 Adhesion Test ...................................................................................................................... 46
1.4 Applications................................................................................................................................ 49
1.4.1 Capture of Hazardous Gases ............................................................................................... 49
1.4.2 Membrane Separation .......................................................................................................... 52
1.4.3 Destruction of Chemical Warfare Agents ........................................................................... 54
1.4.4 Antibacterial Protection ....................................................................................................... 57
1.4.5 Catalytic NO Release .......................................................................................................... 58
1.5 Outlook ....................................................................................................................................... 59
References ........................................................................................................................................ 60
CHAPTER 2. Highly Adsorptive, MOF-Functionalized Nonwoven Fiber Mats for Hazardous Gas
Capture Enabled by Atomic Layer Deposition .................................................................................... 68
Abstract ............................................................................................................................................ 69
2.1 Introduction ................................................................................................................................ 69
2.2 Experimental Methods ............................................................................................................... 71
2.3 Results and Discussion ............................................................................................................... 74
2.4 Conclusion .................................................................................................................................. 84
References ........................................................................................................................................ 85
Supporting Information .................................................................................................................... 87
CHAPTER 3. Conformal and highly adsorptive metal–organic framework thin films via
layer-by-layer growth on ALD-coated fiber mats ................................................................................ 96
Abstract ............................................................................................................................................ 97
3.1 Introduction ................................................................................................................................ 98
3.2 Experimental Methods ............................................................................................................. 100
3.3 Results and Discussion ............................................................................................................. 103
3.4 Conclusions .............................................................................................................................. 114
References ...................................................................................................................................... 116
Supporting Information .................................................................................................................. 120
CHAPTER 4. Facile Conversion of Hydroxy Double Salts to Metal-Organic Frameworks Using
Metal Oxide Particles and ALD Thin Film Templates....................................................................... 121
Abstract .......................................................................................................................................... 122
4.1 Introduction .............................................................................................................................. 122
vii
4.3 Conclusion ................................................................................................................................ 134
References ...................................................................................................................................... 135
CHAPTER 5. Conformal Metal-Organic Framework Thin Films on Nanofibers for Ultra-Fast
Degradation of Chemical Warfare Agents ......................................................................................... 160
Abstract .......................................................................................................................................... 161
5.1 Introducion ............................................................................................................................... 161
5.3 Conclusion ................................................................................................................................ 175
References ...................................................................................................................................... 176
CHAPTER 6. Diffusion of CO2 in large single crystals of Cu-BTC MOF ........................................ 199
Abstract .......................................................................................................................................... 200
6.1 Introduction .............................................................................................................................. 200
6.4 Conclusion ................................................................................................................................ 216
References ...................................................................................................................................... 217
CHAPTER 7. ALD-Assisted Synthesis of ZnO 3D Hierarchical Nanostructures for Enhanced
Photocatalytic Dye Degradation ......................................................................................................... 228
Abstract .......................................................................................................................................... 229
7.1 Introduction .............................................................................................................................. 229
7.4 Conclusion ................................................................................................................................ 243
References ...................................................................................................................................... 244
viii
LIST OF TABLES
Table 1.1 Summary of synthesis methods and properties of MOF-functionalized fibers ................. 38
Table 1.2 CO2 permeance and permselectivity for CO2/N2 for pure Ultem and Ultem/ZIF-8 hollow
fiber membranes ................................................................................................................................... 54
Table 2.1 Characterization of MOF-functionalized fiber mats: fiber mass gain (mg/cm2) after MOF
integration, BET surface area (m2/g) of MOF-fiber materials and NH3 dynamic loading (mol/kg)
calculated from breakthrough curves. .................................................................................................. 82
Table 4.1 Routes for HDS-Driven Room-Temperature Synthesis of Various MOF Materials. ..... 131
Table S4.1 ICP-OES for the concentration of Cu and Zn in the HKUST-1 crystals prepared through
the rapid room-temperature MOF synthesis. ...................................................................................... 149
Table S4.2 ICP-OES for the concentration of Cu and Zn in the filtrate after collecting the MOF
powder by filtration. ........................................................................................................................... 149
Table 5.1 Material properties and catalytic performance towards CWA simulant degradation. ..... 172
Table S5.1 MOF mass fraction in MOF-nanofiber composites calculated from BET results and
ICP-OES analysis ............................................................................................................................... 188
Table S5.2 Kinetic constants and half-lives of DMNP hydrolysis with UiO MOF powders and UiO
MOF thin films on PA-6@TiO2 nanofibers. ...................................................................................... 193
Table 6.1 Micropore diffusion fitting parameters for various Cu-BTC crystal sizes at 0.5% CO2. 213
Table 6.2 Micropore diffusion fitting parameters for 0.7 mm Cu-BTC single crystals and BPL
activated carbon at different CO2 concentrations. .............................................................................. 215
Table 7.1 Feature sizes and photocatalytic performance of ZnO 3D hierarchical nanostructrues. . 242
ix
TABLE OF FIGURES
Figure 1.1 Schematic illustration of the constituents for metal-organic frameworks. ......................... 3
Figure 1.2 Some common and representative MOF structures. .......................................................... 4
Figure 1.3 SEM image of HKUST-1 grown on chemithermomechanical pulp fibers using direct
solvothermal synthesis ........................................................................................................................... 6
Figure 1.4 Carboxymethylation on cellulose for anchoring HKUST-1 MOF coatings ...................... 7
Figure 1.5 SEM of HKUST-1 grown on cotton fibers pretreated via carboxymethylation. Insert
image is an EDX spectrum showing the presence of Cu ........................................................................ 7
Figure 1.6 SEM images of (a) ZIF-90 seeds deposited onto Torlon hollow fiber surface by dip
coating, (b-e) polycrystalline ZIF-90 thin film on Torlon hollow fiber. (f) XRD patterns for
Torlon/ZIF-90 membrane and the bare Torlon substrate ....................................................................... 9
Figure 1.7 Schematic illustration of the process sequence for ALD Al2O3 ....................................... 10
Figure 1.8 (a-c) schematic of the synthesis approach using ALD nucleation layers for solvothermal
synthesis. (a) polymer fiber, (b) polymer fiber coated with ALD Al2O3 with rich surface hydroxyl
groups, (c) conformal MOF coatings formed on ALD-coated fiber. (d) SEM image of conformal
HKUST-1 crystals grown on ALD-coated PP fiber ............................................................................. 11
Figure 1.9 Illustration of the synthesis route for solvothermal growth of Zr-based UiO MOFs on
nylon nanofibers with ALD TiO2 nucleation layers ............................................................................. 13
Figure 1.10 SEM images and cross-sectional TEM images of (a,d) UiO-66 grown on PA@TiO2
nanofibers, (b,e) UiO-66-NH2 grown on PA@TiO2 nanofibers, and (c,f) UiO-67 grown on PA@TiO2
nanofibers ............................................................................................................................................. 13
Figure 1.11 SEM image of HKUST-1 particles grown silk fibers using ultrasound-enhanced
layer-by-layer method .......................................................................................................................... 14
Figure 1.12 SEM images of HMTI-1 grown on silk fibers using (a-b) LbL method without
irradiation enhancement and (c-d) sonication-assisted LbL synthesis ................................................. 15
Figure 1.13 SEM images of silk fibers coated with MOF-5 using direct LbL synthesis (a) without
ultrasound irradiation and (b) with sonication ...................................................................................... 16
Figure 1.14 Surface functionalization of polyester fibers using polyvinylamine (PVA) and
bromoacetic acid (BAA) for layer-by-layer deposition of HKUST-1 thin films ................................. 17
Figure 1.15 SEM images of (a-b) untreated polyester fibers and (c-d) 40 cycles of HKUST-1
layer-by-layer growth on chemically functionalized polyester fibers .................................................. 18
Figure 1.16 SEM images of (a) untreated cotton fibers and (b-f) HKUST-1 coated cotton fibers ... 19
x
Figure 1.17 Illustration of the synthesis approach using ALD Al2O3 as nucleation layer for LbL
synthesis of HKUST-1 conformal thin films........................................................................................ 20
Figure 1.18 (a-b) SEM images of ALD-coated PP fibers with 40 cycles of LbL HKUST-1. (c-d)
Cross-sectional TEM images showing the conformal HKUST-1 thin films deposited on ALD-coated
PP fibers ............................................................................................................................................... 20
Figure 1.19 Schematic for MOF conversion from ALD metal oxide thin films ............................... 22
Figure 1.20 SEM images of conformal ZIF-8 and MIL-53-NH2(Al) thin films grown on PAN
nanofibers using ALD-to-MOF conversion ......................................................................................... 23
Figure 1.21 Schematic for MOF conversion from hydroxy ALD Metal Oxide Thin Films ............. 24
Figure 1.22 SEM Images and cross-sectional TEM image of HKUST-1 thin films grown on (a-b)
PP microfibers and (c-d) PAN nanofibers using HDS-to-MOF conversion ........................................ 25
Figure 1.23 Schematic of a horizontal electrospinning apparatus with rotational counter electrode 26
Figure 1.24 (a-b) Optical micrograph and SEM image of electrospun HKUST-1/PS fibers. (c)
Photo of MIL-100(Fe)/PVP fibers on a polypropylene nonwoven substrate. (d) SEM image of
electrospun MIL-100(Fe)/PVP fibers. (e-f) SEM images of HKUST-1/PAN fibers on a PAN
nonwoven substrate .............................................................................................................................. 27
Figure 1.25 (a) Optical photograph, (b) SEM image and (c) TEM image of ZIF-8–PVP nanofibers28
Figure 1.26 Schematic of electrospinning process and secondary growth for synthesizing MOF
membranes ........................................................................................................................................... 29
Figure 1.27 SEM image of (a) as-spun PS nanofibers with HKUST-1, (b) HKUST-1 membrane
after 4 cycles of secondary solvothermal synthesis on the PS/HKUST-1 nanofiber scaffold, (c)
as-spun PS nanofibers with ZIF-8, (d) ZIF-8 membrane after 5 cycles of secondary growth on
PS/ZIF-8 electrospun fibers.................................................................................................................. 30
Figure 1.28 SEM images of electrospun (a) UiO-66/PAN and (b) MIL-101(Cr)/PAN nanofiber
composites ............................................................................................................................................ 31
Figure 1.29 (a) N2 isotherms for UiO-66 powder and PVCi/UiO-66 nanofiber membrane
synthesized by electrospinning plus 3 cycles of secondary solvothermal growth at 100℃ ................. 33
Figure 1.30 Schematic of the setup for co-extrusion spinning to synthesize hollow fibers .............. 34
Figure 1.31 SEM images of (a) pure Ultem (b-c) Ultem/ZIF-8 hollow fiber membrane .................. 35
Figure 1.32 Photographs and SEM images of HKUST-1 printed on paper fibers for (a) 3 cycles and
(b-d) 8 cycles, respectively ................................................................................................................... 36
Figure 1.33 SEM images of UiO-66 sprayed on silk microfibers ..................................................... 37
Figure 1.34 Schematic of IR beam path in an ATR-IR ..................................................................... 40
Figure 1.35 ATR-IR spectra for cotton fabric coated with different number of cycles of HKUST-1 . .
.............................................................................................................................................................. 41
xi
Figure 1.36 (a) UV-Vis absorption spectra for different numbers of HKUST-1 layer-by-layer
cycles on polyester fibers pretreated with polyvinylamine and bromoacetic acid. (b) absorbance at
704 nm plotted as a function of cycle number ..................................................................................... 42
Figure 1.37 (a) Permeance for UiO-66/PVCi nanofiber membrane for He, N2, Ar and SF6. (b) a
schematic of the apparatus for permeation test .................................................................................... 45
Figure 1.38 AFM images and roughness profiles of (a) untreated polyester fiber surface and (b)
chemically modified polyester fibers deposited with 40 cycles of HKUST-1 layer-by-layer growth . 46
Figure 1.39 Mass change of ALD-coated PBT fiber mats with HKUST-1 thin films during the
compress air blowing test ..................................................................................................................... 47
Figure 1.40 (a) Photographs of cotton fiber mats printed with HKUST-1 dot pattern (5
printing-drying cycles) before and after exposure to NH3, HCl and H2S. (b) Relative mass change for
the HKUST-1 thin film printed on QCM during NH3 adsorption/desorption tests. (c) Adsorption
capacities of the HKUST-1 films printed on QCM for different gases. ............................................... 49
Figure 1.41 NH3 breakthrough curves for bare PP fibers (PP, green blank circle), PP fibers with
ALD Al2O3 coatings (PP/ALD, orange blank diamond), HKUST-1 grown on untreated PP fibers
(MOF-PP, blue circle), HKUST-1 grown on PP/ALD fiber mats (MOF-PP/ALD, red diamond) ...... 50
Figure 1.42 (a) Ammonia breakthrough curves for LbL HKUST-1 thin films grown on
ALD-coated PP fibers and control sample without MOF coating. (b) Ammonia dynamic loading for
PP/ALD fiber mats with 0 ~ 40 cycles of LbL HKUST-1. Black square points were calculated from
the breakthrough data before saturation, and red circle points were calculated based on the
breakthrough curves with desorption parts. Error bar shows standard deviation ................................. 51
Figrue 1.43 Ammonia dynamic loading on HKUST-1 MOF coated PP microfibers and PAN
nanofibers ............................................................................................................................................. 52
Figure 1.44 (a-c) Single gas permeances for Torlon/ZIF-90 hollow fiber membranes measured at (a)
35 oC and 50 psia, (b) 70 oC with varying pressure, (c) 50 psia with varying temperature. (d)
Permeances of hydrocarbon molecules measured at 22 oC by pervaporation ...................................... 53
Figure 1.45 (a) SEM images of silk fibers coated with UiO-66@LiOtBu. (b-d) Catalytic hydrolysis
profiles of CWA simulants using silk@[UiO-66@LiOtBu] composite catalysts ................................ 55
Figure 1.46 (a) Reaction schematic of DMNP catalytic hydrolysis using MOF-nanofiber catalyst.
(b) UV/Visible absorption spectra to monitor DMNP degradation. (c-e) DMNP conversion during
hydrolysis with different MOF powders and MOF-functionalized nanofiber catalysts ....................... 56
Figure 1.47 Photos of the petri dishes for antibacterial tests using HKUST-1 coated silk fibers. The
results were compared with untreated silk, gentamycin (GM), amoxicillin (AMX), cefepime (FEP)
and oxacillin (OX). Samples in Group I were prepared via LbL synthesis without sonication, while
Group IV were synthesized with ultrasound irradiation ....................................................................... 57
Figure 1.48 (Top) Catalytic reaction to form NO from S-nitrosocysteamine. (Bottom) NO release
as function of time with different catalytic materials ........................................................................... 58
xii
Figure 2.1 (a-c) Schematic illustration of the synthesis route. (a) Polymer fiber substrate. (b)
Al2O3-coated polymer fiber via atomic layer deposition (ALD). The cross section in the dashed
square illustrates the conformal coating of ALD Al2O3 with hydroxyl surface termination. (c) MOFs
integated on Al2O3-coated polymer fiber using solvothermal MOF synthesis. (d) SEM image of
HKUST-1 MOF crystals grown on MOF-PP/ALD.............................................................................. 75
Figure 2.2 Comparison of HKUST-1 MOF grown on untreated and ALD-coated polypropylene
fiber mats. (a) SEM image of HKUST-1 MOF grown on untreated polypropylene fiber mats
(MOF-PP). Insert image is a photograph of MOF-PP (scale bar represents 1cm). b) SEM image of
HKUST-1 MOF grown on ALD-Al2O3-coated polypropylene fiber mats (MOF-PP/ALD). Insert
image is a photograph of MOF-PP/ALD (scale bar represents 1cm). c) X-ray diffraction of MOF-PP
and MOF-PP/ALD. d) Mass increase percentage based on substrate dry weight. Interquartile range
and average value were calculated based on 16 MOF-PP samples and 12 MOF-PP/ALD samples. e)
Brunauer–Emmett–Teller (BET) surface area of MOF-fiber materials and calculated BET surface
area for MOF part (error bars represent standard deviation). BET surface area for PP fiber substrates
is 1.3~1.5 m2/g. The values were measured for uncoated and ALD coated fibers, and the values were
indistinguishable. .................................................................................................................................. 77
Figure 2.3 NH3 breakthrough curves for untreated polypropylene fiber mats (PP, ○ ),
ALD-Al2O3-coated polypropylene fiber mats (PP/ALD, ◇), HKUST-1 MOF grown on untreated
polypropylene fiber mats (MOF-PP, ● ), HKUST-1 MOF grown on ALD-Al2O3-coated
polypropylene fiber mats (MOF-PP/ALD, ◆). .................................................................................. 79
Figure 2.4 MOF-fiber mats with different ALD coating thicknesses. a) MOF crystal size
distribution on the top surface. b) MOF crystal size distribution on the cross-section. Crystal size
distributions were analyzed based on 50 measured crystal sizes on each corresponding SEM image. c)
Mass increase of the fiber mats after MOF integration. d) BET surface area of the
MOF-functionalized fiber mats and the surface area of MOF component in the fiber mats. e) NH 3
dynamic loading on MOF-functionalized fiber mats. f) NH3 breakthrough time on
MOF-functionalized fiber mats. ........................................................................................................... 81
Figure 2.5 HKUST-1 MOF grown on different ALD coatings and different polymer fibers. a) SEM
image of MOF on ALD-ZnO-coated polypropylene fiber mats (MOF-PP/ALD(ZnO)). b) SEM
image of MOF on ALD-TiO2-coated PP fiber mats (MOF-PP/ALD(TiO2)). c) SEM image of MOF
on ALD-Al2O3- coated PP fiber mats (MOF-PP/ALD(Al2O3)). d) SEM image of MOF on
ALD-Al2O3-coated polybutylene terephthalate (PBT) fiber mats (MOF-PBT/ALD(Al2O3)). e) SEM
image of MOF on ALD-Al2O3-coated cotton fiber mats (MOF-Cotton/ALD(Al2O3)). ....................... 83
Figure S2.1 Surface wetting ability of nonwoven polypropylene fiber mats with different number
of ALD Al2O3 cycles (0~500 cycles). 50 ALD cycles generates a very thin (~6nm) Al2O3 coating,
but this film was “thick” enough to change the wettability of PP substrates from originally
hydrophobic to hydrophilic. Contact angles of PP in ethanol are 0 degree regardless of ALD cycles.87
xiii
Figure S2.2 (a) Mass increase of PP/ALD fiber mats due to MOF growth at different temperatures
(100~140 ℃ ). (b) total BET surface area and calculated MOF surface area of HKUST-1
functionalized PP/ALD fiber mats prepared at different temperatures (100~140℃). ......................... 88
Figure S2.3 (a) Optical image of MOF-PP/ALD fiber mats prepared at different temperatures
(100~140℃). (b) SEM image of the side products found in MOF-PP/ALD fiber mats synthesized at
100℃. (c) SEM image of the flower-like CuO poly-crystals found in MOF-PP/ALD fiber mats
synthesized at 140℃. ........................................................................................................................... 89
Figure S2.4 X-ray diffraction data of MOF-PP/ALD fiber mats prepared at different temperatures
(100~140ºC) and PP substrate. ............................................................................................................ 90
Figure S2.5 Mass change of MOF funbctionalized fiber mats during the adhesion tests (data shown
are based on MOF-PBT/ALD fiber mats). The mass loss stabilizes after <4 min of air flow, so
empirically we take m/mo(t=4min) to evaluate the attachment of MOF crystals to the fiber substrate.
Mass change of MOF-PP/ALD during the adhesion tests is similar to MOF-PBT/ALD. ................... 91
Figure S2.6 Rapid micro-breakthrough analysis equipment. ............................................................ 92
Figure S2.7 Representative SEM images showing the top surface and the cross-section of
MOF-functionalized fibrous materials based on polypropylene fibers with different ALD coating
thickness. .............................................................................................................................................. 93
Figure S2.8 MOF crystal size distribution in PP fiber mats with different ALD thickness. ............. 94
Figure S2.9 SEM image of MOF crystals grown inside ALD-Al2O3 coated cotton fibers. .............. 95
Figure S2.10 (a) SEM image of UiO-66 MOF grown on PP/ALD fiber mats. (b) SEM image of
Zn-MOF-74 grown on PP/ALD fiber mats. (c) SEM image of Mg-MOF-74 grown on PP/ALD fiber
mats. ..................................................................................................................................................... 95
Scheme 3.1 Schematic of the synthesis route. Polymer fiber substrates were coated with 50 cycles
of ALD-Al2O3, forming a core@shell structure of “Fiber@Al2O3”. HKUST-1 MOF thin film was
grown onto ALD-coated polymer fibers via layer-by-layer synthesis method. ................................... 99
Scheme 3.2 Schematic of the rapid micro-breakthrough analysis system for NH3 and H2S
breakthrough tests............................................................................................................................... 103
Figure 3.1 (a,b,e) SEM images for ALD-Al2O3-coated PP fibers with 40 cycles of LbL HKUST-1
MOF (PP@ALD@LbL40). (c,d) Cross-sectional TEM images for PP@ALD@LbL40 showing the
core@shell structure. (f-i) Energy Dispersive X-ray analysis for PP@ALD@LbL40 showing the
presence of carbon (f) from the polypropylene and the HKUST-1 MOF, oxygen (g) from the ALD
Al2O3 and the HKUST-1 MOF, aluminum (h) from the ALD coating and copper (i) from the MOF. ....
............................................................................................................................................................ 104
xiv
Figure 3.2 SEM images of (a) 20 cycles of LbL HKUST-1 MOF on untreated PP fibers and (b) 20
cycles of LbL HKUST-1 MOF on ALD-Al2O3-coated PP fibers (PP@ALD). MOF thin film grown
on untreated PP substrate exhibits poor uniformity, while MOF coating on PP@ALD shows good
conformality with complete coverage on fibers. ................................................................................ 106
Figure 3.3 (a) Optical images of ALD-Al2O3-coated PP fibers with 0~40 cycles of LbL HKUST-1
MOF. (b) X-ray diffraction data of ALD-Al2O3-coated PP fibers with 0, 20, 40 cycles of LbL
HKUST-1 MOF. Green triangles (▼) represent the peaks for HKUST-1 MOF. (c) Thickness of the
MOF coating (0~40 LbL cycles) on ALD-Al2O3-coated PP fibers measured from cross-section TEM
images. Error bars represent a 95% confidence interval based on 25 data points measured for each
fiber mat. Solid line is a linear-fitted line to the data points. (d) Percent mass increase of LbL
HKUST-1 MOF (0~40 cycles) based on the dry weight of the fiber substrates (△m/mo). Dashed line
is a linear-fitted line to the data points. .............................................................................................. 107
Figure 3.4 (a) Fourier transform infrared (FTIR) spectra for HKUST-1 thin films deposited on
ALD-Al2O3 coated silicon wafers using layer-by-layer (LbL) method. Al2O3 coated silicon wafers
were dipped sequentially in LbL precursor solutions for 1 hour with 5-minute ethanol wash steps in
between. (b) Plot of IR absorbance at ~1370 cm-1 (carboxylate symmetric stretching vibrations) vs.
number of LbL cycles. A linear increase was observed for the absorbance at ~1370 cm -1 after 2 LbL
cycles. ................................................................................................................................................. 110
Figure 3.5 Brunauer-Emmett-Teller (BET) surface area (in units of m2/gMOF+Fiber) of
ALD-Al2O3-coated PP fibers with 0~40 cycles of LbL HKUST-1 MOF. ......................................... 111
Figure 3.6 (a) NH3 breakthrough curves for ALD-coated PP fiber mat with no LbL MOF (■), 20
cycles of LbL MOF (▲) and 40 cycles of LbL MOF (●). (b) NH3 dynamic loading on ALD-coated
PP fiber mats with 0, 20, 40 cycles of LbL MOF. Square points (■) were calculated based on the
corresponding breakthrough curve before saturation, and circle points (●) were calculated with the
desorption part. Error bar represents standard deviation. ................................................................... 111
Figure 3.7 (a-c) SEM images of 20 cycles of LbL HKUST-1 MOF grown on ALD-Al2O3-coated
polypropylene (a), polyethylene terephthalate (b) and cotton (c). (d) Percent mass gain for 20 LbL
cycles on PP@ALD, PET@ALD and Cotton@ALD. (e) BET surface area of PP@ALD@LbL20,
PET@ALD@LbL20 and Cotton@ALD@LbL20. ............................................................................ 113
Figure S3.1 Schematic of the homemade hot-wall viscous-flow ALD reactor used for ALD Al2O3
coatings on fibers. In an ALD cycle, trimethylaluminum (TMA) and water are dosed sequentially
into the chamber, with a purge step of inert gas (N2) in between. Deposition temperature is controlled
by the furnace, and all the gas lines and valves are wrapped with heating tapes to prevent precursor
condensation. ...................................................................................................................................... 120
xv
Figure 4.1 (a) Schematic of the rapid room-temperature synthesis route for Cu3(BTC)2. ZnO reacts
with Cu(NO3)2 to form (Zn,Cu) hydroxy double salt. The (Zn,Cu) HDS converts to Cu3(BTC)2 via
fast anion exchange. (b) Powder XRD patterns for Cu3(BTC)2 synthesized in rapid room-temperature
method (black) and solvothermal method (red), and simulated Cu3(BTC)2 pattern (blue). (c) Percent
yield in 1 minute of reaction (black circle) and space-time-yield for the rapid room temperature
synthesis (red diamond). Insert image in (c) is an SEM image of a Cu3(BTC)2 crystal showing the
octahedral shape. ................................................................................................................................ 125
Figure 4.2 (a) XRD patterns for Cu3(BTC)2 powder (black), ALD ZnO surface after exposure to
Cu(NO3)2 for 1 min (blue) and subsequently to H3BTC for 30 s (red). Blue dot lines represent the
(Zn,Cu) hydroxy double salt. (b) FTIR difference spectra for ALD ZnO in the as-deposited form
(black, Si as background), after exposure to Cu(NO3)2 for 1 min (magenta, previous spectrum as
background), and after exposure to H3BTC for 30 s (green, previous spectrum as background) and
the final Cu3(BTC)2 spectrum (orange, Si as background). (c-d) HAADF STEM images for the cross
section of the Cu3(BTC)2 grown on ALD ZnO coated silicon wafer. The green box in (c) indicates
the location of (d-i). (e-i) High resolution EDX mapping images of the cross section. Scale bars in
(b-g) represent 50nm. ......................................................................................................................... 128
Figure 4.3 (a) Schematic of the fabrication procedure for HKUST-1 patterns. (b-e) SEM image and
EDX mapping images for a star-shape HKUST-1 pattern. ................................................................ 130
Figure 4.4 (a) Schematic of the rapid room-temperature synthesis route for MOF coatings onto
various form factors. (b-d) SEM images of HKUST-1 deposited onto PS spheres, silicon wafer and
PAN nanofibers, respectively. ............................................................................................................ 132
Figure 4.5 (a) BET surface area and (b) NH3 dynamic loading for untreated PP microfibers and
PAN nanofibers, and MOF-coated PP and PAN fibers (MOF-PP, MOF-PAN, respectively)........... 133
Figure S4.1 Size distribution of ZnO nanoslurries dispersed in deionized water measured by
dynamic light scattering (DLS). The average particle size is 252±23 nm, and the polydispersity index
is 0.732. .............................................................................................................................................. 145
Figure S4.2 (a-b) SEM images of HKUST-1 crystals obtained from the rapid room-temperature
synthesis. (c) Crystal size distribution analyzed from SEM images (100 measured data points). The
average crystal size is 1.17±0.40 μm. ................................................................................................. 146
Figure S4.3 (a) SEM image of HKUST-1 crystals prepared via rapid synthesis (dispersed on a
silicon wafer). (b-d) Energy dispersive X-ray (EDX) mapping images for C, O and Cu in the
HKUST-1 crystals. (e) EDX spectrum showing that no Zn can be detected by EDX (blue dash lines
indicate where Zn can be expected). .................................................................................................. 147
Figure S4.4 (a-c) ToF-SIMS surface mapping (negative ion mode) images for a layer of densely
packed HKUST-1 crystals. These results confirm the presence of (a) oxygen, (b) carbon and (c)
copper. (d) ToF-SIMS depth profile (positive ion mode) for C+, Cu+ and 65Cu+ in a layer of densely
packed HKUST-1 crystals. Zn-containing residue was not found by either positive or negative
detection mode. .................................................................................................................................. 148
xvi
Figure S4.5 N2 adsorption and desorption isotherm for the HKUST-1 powder prepared via rapid
synthesis. ............................................................................................................................................ 150
Figure S4.6 SEM images for HKUST-1 grown on top of (Zn,Cu) hydroxy double salt. ................ 151
Figure S4.7 Powder XRD pattern for the Cu-BDC MOF converted from (Zn,Cu) hydroxy nitrate
HDS at room temperature (red). The simulated Cu-BDC pattern is shown in black. The XRD pattern
of the Cu-BDC obtained from HDS agrees well with the simulated pattern and reported powder XRD
patterns for this MOF.6,7 The results from the synthesis of HKUST-1 and Cu-BDC from (Zn,Cu)
HDS indicate that (Zn,Cu) hydroxy nitrate is an intermediate that preferentially converts to Cu-based
MOFs. ................................................................................................................................................. 152
Figure S4.8 Powder XRD patterns for the (Zn,Zn) HDS (blue) and IRMOF-3 converted from
(Zn,Zn) HDS at room temperature (red). The simulated IRMOF-3 pattern is shown in black. ......... 153
Figure S4.9 Powder XRD patterns for the (Zn,Zn) hydroxyl acetate HDS synthesized with DMF
(blue) and ZIF-8 converted from (Zn,Zn) HDS (red). The simulated ZIF-8 pattern is shown in black.
............................................................................................................................................................ 154
Figure S4.10. (a) Photo of circular and star-shape HKUST-1 patterns. The red box in (a) represents
the location of Image (b). (b-d) Optical micrograph and SEM images of star patterns made from
HKUST-1. The red box in (c) shows the location of (d). ................................................................... 155
Figure S4.11 SEM images for (a) untreated PS microspheres, (b) PS microspheres with ALD ZnO
coating, and (c-d) HKUST-1 grown on ZnO-coated PS microspheres. ............................................. 156
Figure S4.12 (a-c) SEM images for ZnO-coated PP fibers (a) before and (b-c) after HKUST-1
rapid synthesis. Insert photo in (b) shows the macroscopic uniformity of MOF growth on the PP fiber
mat. (d) Cross-sectional TEM image shows uniform HKUST-1 coating on ZnO-coated PP fibers.
(e-f), SEM images for ZnO-coated PAN nanofibers (e) before and (f) after HKUST-1 rapid synthesis.
Insert optical image in (f) shows the uniform MOF growth on the PAN nanofiber mat. .................. 157
Figure S4.13 (a) XRD patterns for ALD ZnO coated polypropylene fiber mat (PP/ZnO, red) and
HKUST-1 grown on PP/ZnO (MOF-PP, blue). (b) XRD patterns for ALD ZnO coated PAN
nanofiber mat (PAN/ZnO, red) and HKUST-1 grown on PAN/ZnO (MOF-PAN, blue) .................. 158
Figure S4.14 (a) NH3 breakthrough curves for untreated PP and MOF-PP fiber mats. (b) H2S
breakthrough curves for untreated PP and MOF-PP fiber mats. (c) NH3 breakthrough curves for
untreated PAN and MOF-PAN fiber mats. ........................................................................................ 159
Figure 5.1 Schematic of the synthetic procedure for Zr-based MOF functional coatings on
polyamide-6 nanofibers. The MOF crystal structures are illustrated in the dashed box .................... 165
Figure 5.2 (a) Photo of a free-standing PA-6@TiO2@UiO-66-NH2 nanofiber mat. (b-d) SEM
images, (e-i) energy dispersive X-ray mapping images. .................................................................... 165
xvii
Figure 5.3 (a-c) SEM images of (a) PA-6@TiO2@UiO-66, (b) PA-6@TiO2@ UiO-66-NH2 and (c)
PA-6@TiO2@UiO-67. (d-f) Cross-sectional TEM images of (d) PA-6@TiO2@UiO-66, (e)
PA-6@TiO2@UiO-66-NH2 and (f) PA-6@TiO2@UiO-67. (g-i) XRD patterns of PA-6 nanofibers
before and after ALD, MOF-coated nanofibers and MOF powders. (j-l) N2 adsorption and desorption
isotherms for PA-6@TiO2 nanofibers with and without MOF coatings and Zr-based MOF powders. ...
............................................................................................................................................................ 169
Figure 5.4 (a) Catalytic reaction of DMNP hydrolysis using Zr-based MOF powder and MOF
functionalized nanofiber catalyst. (b) UV/Visible absorption spectra for monitoring degradation of
DMNP. (c-e) Conversion of DMNP to p-nitrophenoxide versus reaction time using MOF powder
and functionalized nanofiber catalyst. ................................................................................................ 171
Figure 5.5 (a) Catalytic reaction of GD hydrolysis using MOF-nanofiber catalyst. (b) Conversion
of GD versus reaction time during catalysis. Dashed lines are fitted results assuming first order
reaction kinetics. ................................................................................................................................. 173
Figure S5.1 SEM images of UiO-66-NH2 grown on (a-b) untreated PA-6 nanofibers and (c-d) ALD
TiO2 coated PA-6 nanofibers. Depositing a thin ALD TiO2 nucleation layer significantly improves
the growth uniformity and crystal coverage on the fiber surface. ...................................................... 186
Figure S5.2. SEM images of (a-b) electrospun PA-6 nanofibers and (c-d) ALD TiO2 coated PA-6
nanofibers. The average diameter of untreated PA-6 nanofibers measured from SEM images is
37±16 nm. ........................................................................................................................................... 187
Figure S5.3 DMNP percent conversion as a function of time during the hydrolysis with untreated
PA-6 and ALD TiO2 coated PA-6 (PA-6@TiO2) nanofibers. Estimated t1/2 values are 3950 min and
1170 min for PA-6 and PA-6@TiO2, respectively. ............................................................................ 190
Figure S5.4 Kinetic analysis of DMNP degradation with (a-b) UiO-66, (c-d) UiO-66-NH2, (e-f)
UiO-67. The red curves in (a), (c), and (e) are plotted using the rate constants derived from the linear
fitting in (b), (d), and (f), respectively, based on the assumption of first order reaction kinetics. We
found that the reaction kinetics with UiO-66-NH2 does not fit well to first-order rate equation. ...... 191
Figure S5.5 Kinetic analysis of DMNP degradation with (a-b) PA-6@TiO2@UiO-66, (c-d)
PA-6@TiO2@UiO-66-NH2, (e-f) PA-6@TiO2@UiO-67. The red curves in (a), (c), and (e) are
plotted using the rate constants derived from the linear fitting in (b), (d), and (f), respectively, based
on the assumption of first order reaction kinetics. The reaction kinetics with
PA-6@TiO2@UiO-66-NH2 does not fit well to first-order rate equation. ......................................... 192
Figure S5.6 SEM images and EDX spectra of (a-b) PA-6@TiO2@UiO-66 nanofibers, (c-d)
PA-6@TiO2@UiO-66 nanofibers and (e-f) PA-6@TiO2@UiO-67 nanofibers after DMNP
degradation experiment. SEM images and EDX results confirm that significant amounts of MOF
coatings remain in the MOF-nanofiber composites even after strong agitation during the DMNP
hydrolysis tests. .................................................................................................................................. 194
31
Figure S5.7 P NMR spectra of GD during hydrolysis with PA-6@TiO2@UiO-66 nanofibers.... 195
xviii
Figure S5.8 31P NMR spectra of GD during hydrolysis with PA-6@TiO2@UiO-66-NH2 nanofibers.
............................................................................................................................................................ 196
31
Figure S5.9 P NMR spectra of GD during hydrolysis with PA-6@TiO2@UiO-67 nanofibers.... 197
Figure 6.1 (a-b) Molecular representation of the Cu-BTC MOF along the [100] direction and [111]
direction. Color code: Cu (yellow); O (red); C (black); H (not shown). (c-d) Optical microscopic and
(e-f) SEM images of Cu-BTC single crystals..................................................................................... 206
Figure 6.2 XRD patterns of a Cu-BTC single crystal for the planes parallel to (100) (black) and the
planes parallel to (111) (red), and simulated powder diffraction pattern for Cu-BTC. ...................... 207
Figure 6.3 Increase of Cu-BTC crystal size as a function of reaction time ..................................... 209
Figure 6.4 Gravimetric CO2 isotherm on large Cu-BTC single crystals at 298 K fit by a Toth
isotherm model. Also shown are data points from literature isotherms of CO2 on powder Cu-BTC at
298 K. ................................................................................................................................................. 210
Figure 6.5 Amplitude ratio curves for CSFR experiments on large Cu-BTC single crystals versus a
powder sample. ................................................................................................................................... 211
Figure 6.6 CSFR curves for different Cu-BTC crystal sizes at 0.5% CO2 concentrations. ............. 212
Figure 6.7 CSFR curves for various CO2 concentrations on Cu-BTC crystals of approximately 1.5
mm. ..................................................................................................................................................... 214
Figure S6.1 Schematic of the concentration-swing frequency response apparatus ......................... 220
Figure S6.2 Optical microscopic images of Cu-BTC crystals obtained from a three-day
solvothermal synthesis. The crystal size ranges from 500 μm to 1.3 mm. (Crystal size measurements
are labelled on the images.) ................................................................................................................ 221
Figure S6.3 (a) N2 isotherms for Cu-BTC single crystals and fine powder. (b) BET surface area
and (c) pore volume calculated from N2 isotherms. ........................................................................... 222
Figure S6.4 Thermogravimetric analysis (TGA) for Cu-BTC single crystals (red) and fine powder
(black)................................................................................................................................................. 223
Figure S6.5 The correlation between the mass per crystal and the crystal dimension. The error bar
indicates the standard deviation of the crystal sizes in a mass measurement group. .......................... 224
Figure S6.6 Crystal mass increase rate (dm/dt) as a function of reaction time based on Eqnation S3.
............................................................................................................................................................ 226
Figure S6.7 CO2 CSFR curves on BPL activated carbon at varying concentrations....................... 227
Figure 7.1 Schematic of the fabrication procedure for the ZnO 3D hierarchical nanostructures. (a)
Silicon wafer spun-coated with antireflection layer and positive photoresist. (b) Periodic
nano-pillars fabricated using laser interference lithography. (c) Conformal ZnO thin films deposited
onto the nano-pillars via ALD. (d) ZnO 3D hierarchical nanostructure obtained from hydrothermal
synthesis of ZnO nano-wires on ALD ZnO coated nano-pillars. ....................................................... 231
xix
Figure 7.2 Water droplet contact angles of the periodic nano-pillars (500 nm high) (a) before ZnO
ALD and (b) after ZnO ALD (200 cycles). Insert SEM images represent the nano-pillars before and
after ZnO ALD, respectively. Scale bars in SEM images represent 1μm. ......................................... 236
Figure 7.3 (a-d) SEM images of ALD ZnO coated nano-pillars (a) before and after (b) 60 min, (c)
120 min, and (d) 150 min of hydrothermal synthesis. (e) ZnO nano-wire length plotted as a function
of hydrothermal reaction time. ........................................................................................................... 237
Figure 7.4 (a) Schematic of ALD ZnO coated flat Si wafer (control) and nano-pillars with different
heights. (b) Normalized concentration as a function of time during the photocatalytic dye degradation.
(c) Apparent first order rate constant (black square) and half-life of MO (red circle) as a function of
the normalized illuminated surface area. ............................................................................................ 238
Figure 7.5 Comparison of photocatalytic performance between nanostructure with and without
ZnO nano-wires. (a) MO degradation curves for flat ZnO surface on Si wafer and ZnO nano-wires
on ZnO coated Si wafer. (b-d) MO degradation curves for ALD ZnO coated nanopillars with and
without ZnO nano-wires. The heights of nanopillars in (b-d) are 250 nm, 500 nm and 800 nm,
respectively......................................................................................................................................... 240
xx
CHAPTER 1 is a reprint of a manuscript in preparation.
CHAPTER 1. Introduction – Synthesis, Characterization and
Applications of Metal-Organic Framework Functionalized
Polymer Fibers
Junjie Zhao, Dennis T. Lee, Gregory N. Parsons*
Department of Chemical & Biomolecular Engineering, North Carolina State University,
Raleigh, NC 27695, U.S.A
*Corresponding author: gnp@ncsu.edu
1
1.1 Introduction
Metal organic frameworks (MOFs) are crystalline porous networks composed of
metal-containing clusters and organic bridging ligands.1 Common building blocks for MOFs
are shown in Figure 1.1, and representative MOF structures are shown in Figure 1.2. Due to
the wide variety of constituents, over 20,000 different MOF structures have been reported in
the past decade.2 These porous materials have garnered tremendous interest because MOFs
generally exhibit ultrahigh surface area (typically 1000 ~ 10,000 m2/g) and large pore
volume.2–6 The pore size, geometry and internal functionality can also be designed rationally
owing to the advantage of significant synthetic versatility.2,7 Post-synthetic processes have
also been established for modifying MOF structures to further enhance the properties.8–10
With these advantages, MOFs have been applied to many applications. For example,
adsorption of harmful gases,11–13 storage of hydrogen and methane,14,15 and sequestration of
carbon dioxide.16 Since new features (luminosity, structural flexibility, pH-sensitivity, and
electronic properties) of MOFs were discovered and taken advantage of, applications have
been extended to separations, drug delivery, chemical sensing and catalysis.7
MOFs obtained from common solvothermal synthesis are in powder forms that are not
soluble in most organic solvents. Therefore further handling and manipulation are required
for practical use. In addition, it is difficult to be integrated MOF powders into functional
materials and application-orientated devices.17,18 Surface-immobilized MOF thin films can
potentially overcome the inherent limitation of powders, and simplify the deployment. MOF
thin films also enable new composite structure and novel device configurations, and expand
the applications of MOFs as well.19,20
2
There has been growing interest in integrating MOFs with natural and synthetic fibers
since 2009,21 and this field is rapidly developing. Fibrous substrates are flexible and porous,
and often exhibit substantial external surface area and good gas permeability. These features
make fibers promising candidates as scaffolds for MOF thin films. A wide variety of
synthetic approaches have been developed in the past few year, and MOF-functionalized
fibers have also been demonstrated for the applications in hazardous gas capture,22–25
membrane separations,26–28 catalytic degradation of toxic compounds,29,30 and functional
biomedical materials.31,32
Figure 1.1 Schematic illustration of the constituents for metal-organic frameworks.
While recent reviews on topics of MOF composites and MOF thin films have partly
discussed about MOF integration on fibers,17,18,20 there is not a comprehensive review to date
3
to cover all the thoughtful strategies of synthesis, the characterization techniques and the
growing number of applications for MOF-coated fibers. In this review, our focus is on the
discussion of different synthesis approaches as well as the associated characterization
methods and application for MOF thin films on textiles and polymer hollow fibers.
Properties of the MOF-fiber composites are compared and listed in Table 1.1, including MOF
mass fraction, overall BET surface area (per unit mass of MOF plus fiber), microscopic
heterogeneity and crystal density on fiber surfaces observed in electronic microscopic
images.
Figure 1.2 Some common and representative MOF structures.
4
1.2 Synthetic Methods
1.2.1 Solvothermal Synthesis
1.2.1.1 Direct Solvothermal Growth
Küsgens et al. were the first few researchers who investigated MOF integration on fibers.
They investigated the growth of HKUST-1 MOF on chemithermomechanical pulp (CTMP),
bleached southern pine kraft pulp and unbleached kraft pulp fibers using direct solvothermal
synthesis.21 8 mmol of Cu(NO3)2 was dissolved in 50 mL of water, and 8 mmol of H3BTC
was dissolved in 100 mL DMF/ethanol (1:1 v/v). 1 g of pulp fibers were added to the mixed
precursor solution. After 15 min of stirring, the suspension was heated to 85oC for 24 h. The
mass fraction of HKUST-1 in the composite was determined by TGA, and the MOF weight
percent was 20.0 wt% and 10.7% for CTMP pulp fibers and unbleached pulp fibers,
respectively. Küsgens et al. found that higher lignin (containing COOH groups) content in
the pulp fibers results in higher mass loading for HKUST-1, consistent with the previous
study of HKUST-1 growth on COOH-terminated SAMs.33 However, the carboxylic acid
terminal groups in the pulp fibers were still not enough for a dense MOF coating (Figure 1.3).
Consequently, the MOF mass fraction and the surface coverage of HKUST-1 on the pulp
fibers were both low.
5
Figure 1.3 SEM image of HKUST-1 grown on chemithermomechanical pulp fibers using
direct solvothermal synthesis.21 (Scale bar adapted from Ref. 17)
Ullah et al. reported carbon nanofibers (CNFs) functionalized with MIL-53(Al) using
direct solvothermal synthesis.34 2.0 g of Al(NO3)2 and 0.949 g of H2BDC were mixed in 40
mL of DMF. 0.4 g of carbon nanofibers was then added to the mixed solution. The reaction
proceeded at 150 oC for 5 h. The resulting CNF@MIL-53 composite exhibits a BET surface
area of 140 m2/g. However, the BET surface area of their bulk MIL-53 powder (138 m2/g) is
quite low compared to reported values for this MOF (1140 – 1235 m2/g),35,36 probably due to
insufficient activation. The CNF@MIL-53 was characterized for CO2 adsorption, and shows
0.10 mmol/g CO2 uptake at 100 kPa.
1.2.1.2 Solvothermal Synthesis on Chemically Functionalized Surface
Carboxymethyltation has been used to add COOH terminal groups to cellulose fiber
surfaces, providing anchoring sites for Cu2+ ions as seeds for Cu-based MOFs.37 This surface
functionalization was based on the reaction between sodium chloroacetate and the hydroxyl
6
groups on cellulose with NaOH catalyst (Figure 1.4). Subsequently, Cu2+ ions from
Cu(OAc)2 precursor can be bound to the carboxylate groups on the treated cellulose surface
via ligand exchange. These Cu-carboxylate seeds enabled the nucleation and growth of
HKUST-1 crystals on cotton fibers when H3BTC linkers were further added to the growth
solution. XRD, XPS and ATR-FTIR results confirm the formation of HKUST-1 on chemical
modified cotton. 37 However, SEM images (Figure 1.5) show low density of MOF crystals on
the fiber surface. There was still space for improving the microscopy growth uniformity.
Figure 1.4 Carboxymethylation on cellulose for anchoring HKUST-1 MOF coatings.37
Figure 1.5 SEM of HKUST-1 grown on cotton fibers pretreated via carboxymethylation.
Insert image is an EDX spectrum showing the presence of Cu. 37
7
1.2.1.3 Solvothermal Synthesis on MOF-Seeded Surface
Using MOF crystals as seeds for secondary solvothermal synthesis has been
demonstrated for achieving continuous MOF films or membranes with high MOF mass
loading.27,38,39 MOF seeds can be deposited onto fiber surface by dip coating or incorporated
in polymer via electrospinning. The latter method will be discussed in the part on
electrospinning processes. Brown et al. synthesized ZIF-90 crystals with narrow size
distribution and deposited these seeds onto a Torlon (a polyamide-imide) hollow fiber by dip
coating the fiber in a ZIF-90 ethanolic suspension (4g/L).27 The seeded fiber was then
transferred to the mixed methanol solution (100 mL) containing imidazole carboxyaldehyde
(ImCA, 0.10 M) and Zn(NO3)2∙6H2O (0.025 M). The solvothermal synthesis proceeded at 65
o
C for 4 h, and a continuous polycrystalline ZIF-90 thin film was obtained on the Torlon
hollow fiber (Figure 1.6). In comparison, the attempts to grow ZIF-90 direct on Torlon fibers
failed to form uniform intergrown ZIF-90 crystals. Therefore seeding plays an important role
to achieve high-quality defect-free MOF thin films. SEM images show the thickness of the
ZIF-90 thin film is about 5 μm.27
8
Figure 1.6 SEM images of (a) ZIF-90 seeds deposited onto Torlon hollow fiber surface by
dip coating, (b-e) polycrystalline ZIF-90 thin film on Torlon hollow fiber. (f) XRD patterns
for Torlon/ZIF-90 membrane and the bare Torlon substrate.27
1.2.1.4 Solvothermal Synthesis on Atomic Layer Deposited Nucleation Layers
Inspired by the early work on MOF thin films grown on inorganic substrates,33,40 our
group developed a general approach to form metal oxide thin films on polymer fibers and use
the deposited inorganic surface for MOF nucleation and growth.23,24,30,41 Atomic layer
deposition (ALD) was used to deposit conformal metal oxide coatings onto fibrous substrates
in a vapor-phase process. Figure 1.7 shows a typical cycle of an example ALD reaction
9
sequence. In the first half-reaction, the precursor reacts with the substrates and leads to a
change in surface terminal groups which are ready to react with the co-reactant in the second
half reaction. Between each half-reaction, vapor byproduct and precursor residue are
removed from chamber using an inert gas purge step. The cycle repeats to build up a film
with thickness controlled by the number of ALD cycles used.42 Previous work in our group
has demonstrated that ALD is powerful in modifying polymers as well as enhancing the
wetting ability and conductivity of fibers.43–46 Recent work in our group has shown that ALD
nucleation layers significantly improve the conformality of MOF coatings on polymer fiber.
Figure 1.7 Schematic illustration of the process sequence for ALD Al2O3.
Zhao et al. showed that ALD Al2O3 (200 cycles deposited at 60 oC, approximately 24 nm)
thin films deposited on polypropylene (PP) fibers enable conformal growth of HKUST-1
micron-scale crystal on the fiber surface using solvothermal synthesis (Figure 1.8).23
10
Compared to untreated PP fibers, the MOF mass loading is increased from 9.61 mg/cm2 to
14.78 mg/cm2 with the ALD Al2O3 nucleation layer. The MOF mass fraction (75.3%) is also
among the highest values ever reported. The thickness of ALD Al2O3 films was found to
affect the MOF crystal size distribution on the fibers. The MOF mass loadings also vary on
different ALD surfaces (Al2O3, ZnO and TiO2) on PP fibers. This method was also applied to
polybutylene terephthalate (PBT) and cotton fibers, and further extended to MOF-74 and
UiO-66 coatings.23
Figure 1.8 (a-c) schematic of the synthesis approach using ALD nucleation layers for
solvothermal synthesis. (a) polymer fiber, (b) polymer fiber coated with ALD Al2O3 with
rich surface hydroxyl groups, (c) conformal MOF coatings formed on ALD-coated fiber. (d)
SEM image of conformal HKUST-1 crystals grown on ALD-coated PP fiber. 23
11
Our group have also investigated the growth mechanism of HKUST-1 on different metal
oxide nucleation surfaces on fibers during solvothermal synthesis.25,41 HKUST-1 growth rate
was found fastest on ALD ZnO coated PP fibers, because (Zn,Cu) hydroxy double salt
intermediate was quickly formed on the ZnO coatings after exposing to Cu(NO3)2. Further
anion exchange leads to the fast kinetics for HKUST-1 formation on ALD ZnO. A rapid
room-temperature synthesis route has also been developed based on this mechanism, which
will be discussed in the later part of this review.
Zhao et al. also applied ALD nucleation layers to nanofiber systems for solvothermal
sysnthesis of Zr-based MOFs including UiO-66, UiO-66-NH2, and UiO-67 (Figure 1.9).30
ALD TiO2 (100 cycles at 100 oC, approximately 5 nm) was deposited onto electrospun PA-6
nanofibers. Solvothermal synthesis (85 oC) with HCl modulator was performed on the
TiO2-coated PA-6 nanofibers (PA-6@TiO2). SEM and cross-sectional TEM images show that
conformal MOF coatings were formed onto the nanofiber surface for all the three types of
Zr-based UiO MOFs (Figure 1.10). The fact that the MOF crystals were grown directly on
and around the nanofibers indicates the strong attachment to the substrates. These
UiO-coated nanofibers were also applied to catalytic degradation of chemical warfare agents
and showed excellent activity comparable to the UiO MOF powders.30
12
Figure 1.9 Illustration of the synthesis route for solvothermal growth of Zr-based UiO
MOFs on nylon nanofibers with ALD TiO2 nucleation layers.30
Figure 1.10 SEM images and cross-sectional TEM images of (a,d) UiO-66 grown on
PA@TiO2 nanofibers, (b,e) UiO-66-NH2 grown on PA@TiO2 nanofibers, and (c,f) UiO-67
grown on PA@TiO2 nanofibers.30
13
1.2.2 Layer-by-Layer Synthesis
1.2.2.1 Direct Layer-by-Layer Synthesis
Similar to direct solvothermal synthesis, substrate materials with rich COOH groups are
desired for direct layer-by-layer (LbL) method without pretreatment on fiber surfaces. Abbasi
et al. investigated ultrasound-enhanced LbL process to grow HKUST-1 on silk.31 The silk
substrates were sequentially dipped in Cu(OAc)2 aqueous solution and H3BTC in
DMF/ethanol mixed solution, with washing steps in water between the exposure to reactants.
Sonication was used to accelerate the LbL reactions and to improve the homogeneity of MOF
coatings on fibers. Although the MOF growth uniformity was improved compared to the
early reports using direct solvothermal synthesis, this method was not able to form
continuous MOF films (Figure 1.11). While the authors claimed that the COOH groups are
the anchoring sites for the MOFs, the density of the carboxylic acid groups present in the
chain terminals of polypeptides in natural silk is actually very low.47 As a result, silk may not
be an ideal substrate for direct MOF synthesis.
Figure 1.11 SEM image of HKUST-1 particles grown silk fibers using
ultrasound-enhanced layer-by-layer method.31
14
Direct LbL synthesis was also applied for growing [Pb(NO3)2(4-bpdh)(H2O)]n (HMTI-1)
on silk fibers.48 0.033 M of Pb(NO3)2 aqueous solution and 0.033 M of
2,5-bis(4-pyridyl)-3,4-diaza-2,4-hexadiene (4-bpdh) methanol solution were used as the
precursor solutions for LbL. In each LbL cycle, the silk substrate was dipped sequentially in
the two reactant solutions for 5 min, with rinsing steps (2 min) in water between the
immersion steps in precursors. 20 LbL cycles were deposited onto silk fibers. SEM images
(Figure 1.12) show high crystal density in the MOF film, but the film growth is not
conformal on the fiber surface. Using ultrasound irradiation to enhance the LbL synthesis,
the uniformity of HMTI-1 coating was improved. Decreased MOF crystal size was also
observed. 48
Figure 1.12 SEM images of HMTI-1 grown on silk fibers using (a-b) LbL method without
irradiation enhancement and (c-d) sonication-assisted LbL synthesis .48
Khanjani et al. also investigated the synthesis of MOF-5 coatings on silk fibers using
direct LbL method.49 7.74 mmol Zn(OAc)2 DMF solution and 3.05 mmol H2BDC DMF
15
solution were used as precursor solutions. SEM images indicate that the morphology of the
MOF-5 coatings synthesized with sonication is more uniform than that obtained without
ultrasound irradiation (Figure 1.13). The MOF-5 coated silk fibers were tested for dye
adsorption, and the large uptake for cargo red dye confirms the porosity of the MOF-5
coatings. 49
Figure 1.13 SEM images of silk fibers coated with MOF-5 using direct LbL synthesis (a)
without ultrasound irradiation and (b) with sonication. 49
1.2.2.2 Layer-by-Layer Synthesis on Chemically Functionalized Surface
Meilikhov et al. showed a chemical functionalization approach to modify polyester
microfiber surface for layer-by-layer deposition of HKUST-1 thin films.50 The low density of
16
COOH groups in the polymer chain terminals was improved by the sequential reactions with
polyvinylamine (PVA) and bromoacetic acid (BAA) (Figure 1.14). The enriched carboxylic
acid terminal groups provide anchoring sites for the secondary building unites of HKUST-1.
In a layer-by-layer growth cycle, the modified polyester fibers were first immersed in 2 mM
Cu(OAc)2 ethanol solution for 10 min, and then washed with pure ethanol for 10 min.
Subsequently, the fibers were exposed to 0.2 mM H3BTC ethanol solution for 20 min, and
washed with pure ethanol again for 10 min. 40 cycles of LbL HKUST-1 thin films were
deposited onto the polyester fibers.
Figure 1.14 Surface functionalization of polyester fibers using polyvinylamine (PVA) and
bromoacetic acid (BAA) for layer-by-layer deposition of HKUST-1 thin films.50
XRD and IR spectroscopy confirm the formation of HKUST-1 on the modified polyester
fibers, and UV-Vis absorbance spectra show linear absorbance at 704 nm increases as a
function of the number of LbL cycles. MOF mass loading was determined by dissolving the
MOF thin films and measuring the Cu concentration via ICP-OES. Meilikhov et al. found
that for the polyester fibers treated with PVA and BAA the Cu uptake is 69 ± 9
17
μg∙gFiber-1∙cycle-1. Taking the Cu concentration in bulk HKUST-1 into account, this value
corresponds to MOF mass uptake of ~0.22 mg∙gFiber-1∙cycle-1. With 40 LbL cycles, the MOF
weight percent in the composite is approximately 0.88%. The authors claimed that
continuous HKUST-1 thin films have been formed on the modified polyester fibers. However,
large MOF agglomerates were also observed in the SEM images (Figure 1.15). The film
roughness is also not as good as those grown on SAM-functionalized Au substrates.50
Figure 1.15 SEM images of (a-b) untreated polyester fibers and (c-d) 40 cycles of
HKUST-1 layer-by-layer growth on chemically functionalized polyester fibers.50
Neufeld et al. describes a synthesis approach similar to the previously reported
procedures for functionalizing cotton surfaces with carboxylate groups37 and growing
Cu-BTC using layer-by-layer method.24,31,50 In this work, high concentration of the reactant
solutions were used, and the solvents for each precursors were different from typical
layer-by-layer synthesis. Specifically, 0.40 M of Cu(OAc)2 aqueous solution and 0.20 M of
H3BTC DMF/ethanol/water (equal parts) solution were used for the sequential dipping
18
process. In contrast to the continuous smooth thin films grown with low concentration
precursors,24 the modified recipe forms dense micron-scale crystals on the pretreated cotton
surface (Figure 1.16).
Figure 1.16 SEM images of (a) untreated cotton fibers and (b-f) HKUST-1 coated cotton
fibers.32
1.2.2.3 Layer-by-Layer Synthesis on Atomic Layer Deposited Nucleation Layers
Since ALD metal oxide thin films on polymer fibers have been demonstrated as good
nucleation layers for MOF solvothermal growth, our group have also investigated the growth
of MOF films on ALD-coated fibers using LbL method.24 Zhao et al. deposited ALD Al2O3
thin films on PP fibers (50 cycles at 60 oC, approximately 6 nm) and applied LbL synthesis
with ultrasound irradiation (Figure 1.17). Conformal and smooth HKUST-1 thin films were
19
obtained, and no agglomerates of MOF crystals were observed in SEM images (Figure 1.18).
The sub-micron crystal size of the LbL HKUST-1 is also much smaller than the crystals
synthesized via solvothermal methods. MOF film thickness and mass loading were found to
increase linearly with the number of LbL cycles, indicating an excellent control over the
MOF film growth. Low concentration of precursor solution and sufficient rinsing between
the precursor exposure steps are critical for high-quality MOF thin films.24
Figure 1.17 Illustration of the synthesis approach using ALD Al2O3 as nucleation layer for
LbL synthesis of HKUST-1 conformal thin films.24
Figure 1.18 (a-b) SEM images of ALD-coated PP fibers with 40 cycles of LbL HKUST-1.
(c-d) Cross-sectional TEM images showing the conformal HKUST-1 thin films deposited on
ALD-coated PP fibers.24
20
Zhao et al. also investigated the growth rate of HKUST-1 thin films in the LbL synthesis.
Film thicknesses were analyzed from cross-sectional TEM images, and a growth rate was
found to be 3.0 nm/cycle (or ~288 ng∙cm-2 per cycle). This value is close to the dimension of
one monolayer of HKUST-1 MOF. The growth rate measured from MOF mass uptake on
fibers ~290 ng∙cm-2 per cycle is also similar to the results from film thickness analysis. This
synthesis method was also applied to polyethylene terephthalate and cotton fibers and
generated uniform HKUST-1 films on these substrates as well.24
1.2.3 MOF Conversion from Thin Film Templates
1.2.3.1 Conversion from ALD Metal Oxide Thin Films
Metal oxides have been previously shown as possible precursors than can be directly
converted to MOF structures in the reaction with organic linkers under solvothermal
conditions.51,52 The same strategy can also be applied to MOF thin films on fibers.41,53 Figure
1.19 shows the general approach for MOF conversion from ALD metal oxide thin films.
Bechelany et al. reported the synthesis of ZIF-8 and MIL-53-NH2(Al) on PAN nanofibers
using this method.53 50 nm of ALD ZnO and 42 nm of ALD Al2O3 were deposited onto PAN
nanofibers respectively prior to MOF conversion. 2-methylimidazole methanolic solution
was used for ZIF-8 conversion from ZnO with microwave-assisted heating at 100 oC for 1.5 h,
while 2-aminoterephthalic acid aqueous solution was used for MIL-53-NH2(Al) conversion
from Al2O3 with the same heating condition and reaction time. 53
21
Figure 1.19 Schematic for MOF conversion from ALD metal oxide thin films.
The excellent conformality of ALD thin films can be translated into good microscopic
homogeneity of the MOF coatings. Bechelany et al. showed dense crystalline ZIF-8 and
MIL-53-NH2(Al) thin films formed on PAN nanofibers using this ALD-to-MOF conversion
(Figure 1.20). Since this conversion is localized,52 no agglomerates of MOFs are formed in
the fiber networks. Conventional heating and microwave-assisted heating were also
compared for this type of synthesis. The growth uniformity and crystal coverage on fibers for
samples prepared using conventional heating are not as good as those with microwave
irradiation. Nevertheless, this process provides opportunities for new types of conformal
MOF thin films.53
22
Figure 1.20 SEM images of conformal ZIF-8 and MIL-53-NH2(Al) thin films grown on
PAN nanofibers using ALD-to-MOF conversion. 53
Our group also observed ALD-to-MOF conversions when we were investigating surface
reactions between MOF precursors and ALD metal oxides.41 Lemaire et al. found that the
ALD Al2O3 coating on PP fibers can be converted to MIL-96 after exposure to H3BTC
solution with mixed ethanol/water (1:1 v/v) solvent at 120 oC for 20 h. Similarly, ALD ZnO
films on PP fibers were converted to Zn-BTC. These results further confirmed the general
applicability of ALD-to-MOF conversion for MOF thin films on fibrous substrates.41
1.2.3.2 Room-Temperature Conversion from Hydroxy Double Salts
Hydroxy double salts (HDSs) are layered compounds containing 2-D cationic sheets and
interlamellar anions, and can exhibit fast anion exchange rate.54 Zhao et al. found that HDSs
23
can also be converted to MOFs at room temperature through rapid anion exchange.25 (Zn, Cu)
hydroxy nitrate formed in situ from ZnO powder or films can further react with the organic
linkers and form HKUST-1 or Cu(BDC) (Figure 1.21). This process significantly reduces the
synthesis temperature and exhibits ultra-high space-time-yield (up to 3.6 × 104 kg·m−3·d−1).25
Figure 1.21 Schematic for MOF conversion from hydroxy ALD Metal Oxide Thin Films.
Using this fast HDS-to-MOF conversion, MOF thin films were synthesized on PP
microfibers and PAN nanofibers. SEM and TEM images show that dense HKUST-1 crystal
coatings were formed on the fiber surfaces (Figure 1.22). Similar to ALD-to-MOF
conversion, the conversion from HDS intermediates is also localized. The quality of the
MOF films as well as the microscopic uniformity is comparable to the conformal MOF
coatings synthesized via other method. Since this process can be performed at room
temperature, it is very promising for scale-up production of MOF-functionalize composites.25
24
Figure 1.22 SEM Images and cross-sectional TEM image of HKUST-1 thin films grown on
(a-b) PP microfibers and (c-d) PAN nanofibers using HDS-to-MOF conversion.25
1.2.4 Electrospinning
In a typical electrospinning process, the polarization of polymer solution draws fibers
from a feed droplet at the outlet of the metallic needle to the counter electrode collector
(Figure 1.23). The solvent evaporates during the process, and dry fiber mat can be obtained.55
This technique has been well developed for making polymer micro- and nanofibers. Recently,
electrospinning has also been demonstrated for the synthesis of MOF-incorporated fibers.
25
Figure 1.23 Schematic of a horizontal electrospinning apparatus with rotational counter
electrode.55 Vertical electrospinning setups are similar, since the basic principle is the same.
Rose et al. are the first few researchers to look into the synthesis of MOF-functionlized
textiles using electrospinning.55 Three polymer systems were investigated including
polyvinylpyrrolidone (PVP) in ethanol, polystyrene (PS) in THF and polyacrylonitrile (PAN)
in DMF, and the mass ratios in the mixed solutions for electrospinning are
HKUST-1/PAN/DMF = 1:0.25:4.7, HKUST-1/PS/THF = 1:1.6:4.3 and MIL-100(Fe)/PVP/
EtOH = 1:0.25:5.3. The feed flow rate was 4 mL/h. A high voltage of 10 – 20 kV was applied
and the collector was 10 – 15 cm from the needle. Up to 40 wt% was achieved for
HKUST-1/PS fibers, but the huge size difference between the MOF crystal and the fiber
diameter resulted in non-uniform MOF distribution in the fiber (Figure 1.24b). The
investigators further studied electrospinning PVP fibers with reduced diameters (100 – 500
nm) for physically trapping the MOF crystals in the fiber web (Figure 1.24d). However, the
instability of PVP in polar solvent was a great concern. The attachment of these micron-scale
MOF crystals to the nanofibers may also be weaker in this case than those chemically-bound
26
MOF coatings or incorporated MOF particles. While a high MOF weight percent (80%) was
achieved for HKUST-1/PAN system, large chunks of MOFs were observed in SEM images,
indicating poor uniformity of MOF distribution.
Figure 1.24 (a-b) Optical micrograph and SEM image of electrospun HKUST-1/PS fibers.
(c) Photo of MIL-100(Fe)/PVP fibers on a polypropylene nonwoven substrate. (d) SEM
image of electrospun MIL-100(Fe)/PVP fibers. (e-f) SEM images of HKUST-1/PAN fibers
on a PAN nonwoven substrate.55
Ostermann et al. found that MOF nanocrystals with narrow distribution are critical in
order to achieve a homogeneous MOF distribution in the electrospun MOF-polymer fibers.56
Using previous solvothermal synthesis method, they prepared ZIF-8 nanoparticles with
27
narrow distribution. 400 mg ~ 2000 mg of ZIF-8 was mixed with 12 wt% PVP methanol
solution. After dilution or concentration, the final solution with 3.5 wt% PVP concentration
was fed through a needle at the rate of 0.35 mL/h. The voltage for electrospinning was 5 kV,
and the aluminum foil collector was placed 6-8 cm from the needle. SEM and TEM images
(Figure 1.25) show that ZIF-8 nanoparticles are homogeneously embedded in the PVP
nanofibers. 56% ZIF-8 mass fraction was achieved in the MOF-nanofiber structure. As a
result, the overall BET surface area for the PVP-ZIF-8 exceeds 500 m2/g.56
Figure 1.25 (a) Optical photograph, (b) SEM image and (c) TEM image of ZIF-8–PVP
nanofibers.56
To avoid the disadvantage of PVP dissolution in polar solvents, Ostermann et al. also
applied the same process for polystyrene-ZIF-8 nanofibers. With 25 wt% MOF mass fraction,
the PS-ZIF-8 nanofiber composite exhibits overall BET surface area of 210 m2/g.
28
Polyethylene oxide (PEO) was also attempted for electrospinning MOF-nanofibers. However,
PEO-ZIF-8 shows no accessible MOF surface area because the polymer completely blocks
the MOF pores. The polymer layer covering the MOF particles was found to be an additional
diffusion barrier, so a low thickness is crucial for the polymer layer to maintain the
accessibility of the MOF pores in these electrospun MOF-nanofiber structures.56
Wu et al. also investigated electrospinning and explored membrane synthesis using
electrospun MOF-fibers as scaffold. PS in THF solution was used for HKUST-1 and ZIF-8,
and PVP in ethanol solution was used for MIL-101(Fe) and Zn2(bpdc)2(bpee)
(bpdc=4,4’-biphenyldicarboxylate; bpee=1,2-bipyridylethene). The electrospinning process is
similar to the method developed by Ostermann et al (Figure 1.26). With approximately 20 wt%
MOF mass fraction, the electrospun PS/HKUST-1 fibers exhibit BET surface area of 311
m2/g, and PS/ZIF-8 fibers have BET surface area of 303 m2/g. The homogeneous distribution
of MOF in the fibers is similar to the MOF-nanofiber structures Ostermann et al. have
reported.
Figure 1.26 Schematic of electrospinning process and secondary growth for synthesizing
MOF membranes.38
29
Using the as-spun nanofibers with MOF seeds embedded, further secondary
solvothermal growth can improve the MOF mass fraction up to 50% and enable the
formation of MOF membranes (Figure 1.27).38 Since MOF crystals uniformly cover the
entire surface of the fiber mat, nanofiber skeleton is no longer visible in SEM images after
repeated secondary growth. As a result of increased MOF weight percent, the overall BET
surface area for the MOF membranes is 944 m2/g for PS/HKUST-1 membrane after 4 cycles
of secondary solvothermal synthesis on the PS/HKUST-1 nanofiber scaffold. Similarly, the
PS/ZIF-8 membrane obtained after 5 cycles of secondary growth exhibits surface area of 935
m2/g. Wu et al. also investigated the potential of the PS/ZIF-8 membrane for CO2/N2
separation. However, due to macroscopic pores and defects in the membrane, the separation
factor of N2/CO2 was limited to 2.4. 38
Figure 1.27 SEM image of (a) as-spun PS nanofibers with HKUST-1, (b) HKUST-1
membrane after 4 cycles of secondary solvothermal synthesis on the PS/HKUST-1 nanofiber
scaffold, (c) as-spun PS nanofibers with ZIF-8, (d) ZIF-8 membrane after 5 cycles of
secondary growth on PS/ZIF-8 electrospun fibers.38
30
Figure 1.28 SEM images of electrospun (a) UiO-66/PAN and (b) MIL-101(Cr)/PAN
nanofiber composites.57
Ren et al. reported the fabrication of MOF-nanofiber composites containing Zr- and
Cr-based MOFs using similar electrospinning method. 100 mg of UiO-66 and MIL-101(Cr)
were mixed with 500 mg of 10 wt% polyacrylonitrile (PAN) DMF solution respectively, and
used for electrospinning.57 The feed rate of mixed suspension was controlled at 1.25 – 2.5
mL/h, and the voltage for electrospinning was set to 10 kV. Electrospun nanofibers were
collected on an aluminum foil which was 10 cm away from the solution outlet. The author
pointed out that the MOF mass fraction in the mixed suspension has to be controlled under
40 wt% in order to obtain proper viscosity for electrospinning. Although the authors
31
calculated the MOF weight percent in the MOF-fiber composites based on the MOF mass
fraction in the mixed suspension, they neglected the evaporation of DMF during the
electrospinning and post-synthetic drying processes. The actual MOF mass fraction in the
final composite form could be as high as 67%, assuming the removal of all the solvent in the
electronspun MOF-fibers. The ratio of the overall (MOF+fiber) BET surface area to the MOF
surface area is another good estimation of the actual MOF mass fraction in the composites.
The MOF weight percent calculated based on their BET surface area is 68.7% for UiO-66 on
PAN and 43.3% for MIL-101(Cr) on PAN, respectively. As a result of the high MOF mass
loading, the MOF-fiber composites also exhibited over 50% H2 adsorption capacity
compared with the bulk MOF powders. 57
Armstrong et al. reported a similar approach using electrospinning technique. UiO-66
crystals were mixed with 10 wt% poly(vinyl cinnamate) (PVCi) dichloromethane solution,
and fed through a needle at 4.2 mL/h rate. The voltage for electrospinning was 22 kV, and the
collector was 6 cm away from the needle. In addition to electrospinning, UV irradiation and
secondary growth were also applied to the MOF-nanofiber mats. During the UV irradiation,
PVCi was crosslinked via a photo-induced cycloaddition reaction, which was confirmed by
FTIR. As a result, the stability of PVCi nanofibers in DMF was significantly improved after
photo-crosslinking. Using the UiO-66 embedded in the PVCi nanofibers as the crystal seed,
the author also performed repeated secondary growth to obtain high MOF loading. After 3
cycles of secondary solvthermal growth at 100℃ on the electrospun MOF-nanofiber mat,
overall BET surface area of 430±1 m2/g was obtained. Based the overall surface area and
that for the UiO-66 powder (1188 m2/g), we estimate the MOF mass fraction in this
32
UiO-66/PVCi composite structure to be 36.2%. 39
Figure 1.29 (a) N2 isotherms for UiO-66 powder and PVCi/UiO-66 nanofiber membrane
synthesized by electrospinning plus 3 cycles of secondary solvothermal growth at 100℃.39
1.2.5 Co-Extrusion Spinning
Husain et al. demonstrated a spinning technique by co-extruding two polymer solutions
and applied this method for asymmetric polymer/zeolite hollow fibers.58 Later this process
was used to synthesize polymer/MOF hollow fibers.28 Figure 1.30 illustrates the
experimental setup for co-extrusion spinning. Basically, sheath dope (for the
33
MOF-containing skin) and core dope (for the porous polymer hollow fiber substrate)
solutions are extruded simultaneously with the presence of bore fluid for supporting the
nascent hollow fiber. The active sheath layer is formed when the solvent is evaporated in the
air gap, and the rapid phase separation in the subsequent wet quench process enables the
formation of large porosity in the core polymer layer. 28
Figure 1.30 Schematic of the setup for co-extrusion spinning to synthesize hollow fibers.58
Dai et al. synthesized Ultem (a polyetherimide) hollow fibers with a functional
ZIF-8-containing sheath layer using the co-extrusion spinning process.28 SEM images show
no obvious boundary between the core and sheath layer, indicating good adhesion. Since the
ZIF-8 crystals were embedded in the sheath layer like common flat polymer mixed matrix
membranes, this type of MOF/fiber structure is essentially designed for gas separation
applications.
34
Figure 1.31 SEM images of (a) pure Ultem (b-c) Ultem/ZIF-8 hollow fiber membrane.28
1.2.6 Inkjet Printing
Taking advantage of evaporation/solvent induced crystalization,22,59 Zhuang et al.
developed a precursor solution and used it as the ink for printing MOFs on papers and
textiles.60 Ethylene glycol was found to stablize the precursor solution with
ethanol/dimethylsulfoxide (DMSO) mixted solvents. The precursor ink were printed onto
PET foils, office paper and cotton textiles using a commertial printer, and the substrate
materials were dried at 80℃ for 2 mins. The thickness of the MOF coating and the density
of MOF crystals can be controled by the number of printing-drying cycles. This unique
process can achieve very good growth uniformity on the top surface (Figure 1.32) of the
substrates and can be easily applied for patterning on large substrates. However, due to the
limited penetration depth of the precursor ink into the complex fiber networks, it may be an
issue for this process to obtain good spacial homogeneity for textiles. Large MOF
agglomerates were also observed in the voids of paper substrates. 60
35
Figure 1.32 Photographs and SEM images of HKUST-1 printed on paper fibers for (a) 3
cycles and (b-d) 8 cycles, respectively. 60
1.2.7 Spay Drying Method
López-Maya et al. reported a process to physically disperse UiO-66 particles onto
fibrous substrates using spray drying method.29 Methanolic suspension containing
[Zn4O(dmcapz)3] or UiO-66 was sprayed through an atomizer during the electrospinning
process for silk fibers. The MOF particles spayed onto the substrates in each process cycle
are physically immobilized in the as-spun layers of silk fibers.29 While this method can
achieve high MOF mass fraction (50%), the MOF crystals are not chemically bound to the
fiber surface. The agglomerates of MOF particles (Figure 1.33) in the composites also exhibit
a smaller external surface area that is accessible during heterogeneous catalysis, compared
with the conformal MOF thin films grown onto the fiber scaffolds.
36
Figure 1.33 SEM images of UiO-66 sprayed on silk microfibers.29
37
Table 1.1 Summary of synthesis methods and properties of MOF-functionalized fibers.
Type of MOF Substrate Material Synthesis Method MOF wt% Overall BET Microscopic Crystal References
Surface Area Homogeneity Density on

2
(m /g) Fibers
HKUST-1 Cotton microfiber Layer-by-layer synthesis after carboxymetylation (sodium 18 ±2 N/A Good High 32
chloroacetate + NaOH)
HKUST-1 Chemithermomechanical pulp Direct Solvothermal 20.0 314 Good Low 21
(CTMP) microfiber
HKUST-1 Unbleached kranft pulp Direct Solvothermal 10.7 165 N/A N/A 21
microfibers
HKUST-1 Polyester microfiber Layer-by-layer synthesis after surface functionalization 0.88 <b> N/A Poor High 50
HKUST-1 Polystyrene microfiber Electrospinning 40 N/A Poor Low 55
HKUST-1 Polyacrylonitrile nanofiber Electrospinning 80 1200 Poor Low 55
HKUST-1 Polystyrene nanofiber Electrospinning 20 <c> 311 Good Low 38
HKUST-1 Polystyrene nanofiber Electrospinning + secondary solvothermal growth 50 <a> 944 Good High 38
HKUST-1 Cotton microfiber Solvothermal synthesis after carboxymetylation N/A N/A Poor Low 37
HKUST-1 Silk microfiber Direct layer-by-layer synthesis N/A N/A Good Low 31
HKUST-1 Paper Inkjet printing 10.1 N/A Good High 60
HKUST-1 Cotton microfiber Inkjet printing 2.7 N/A Good High 60
HKUST-1 Polypropylene microfiber Solvothermal on ALD nucleation layers 75.3 695 Good High 23
HKUST-1 Polybutylene terephthalate Solvothermal on ALD nucleation layers 52.5 <a> 485 Good High 23
microfiber
HKUST-1 Cotton microfiber Solvothermal on ALD nucleation layers 13.1 <a> 121 Poor Low 23
HKUST-1 Polypropylene microfiber Layer-by-layer on ALD nucleation layers 17 93 Good High 24
HKUST-1 Polyethylene terephthalate Layer-by-layer on ALD nucleation layers 1.2 12.2 Good High 24
microfiber
HKUST-1 Cotton microfiber Layer-by-layer on ALD nucleation layers 5.3 18.2 Good High 24
HKUST-1 Polypropylene microfiber MOF conversion from hydroxy double salt 10.6 <a> 201 Good High 25
HKUST-1 Polyacrylonitrile nanofiber MOF conversion from hydroxy double salt 27.7 <a> 524 Good High 25
MIL-100(Fe) Polyvinylpyrrolidone nanofiber Electrospinning N/A N/A Poor Low 55
MIL-101(Cr) Polyacrylonitrile nanofiber Electrospinning 43.3 <a> 1134 Good High 57
MIL-101(Fe) Polyvinylpyrrolidone nanofiber Electrospinning 67 <c> N/A Good Low 38
MIL-53(Al) Carbon nanofiber Direct Solvothermal N/A 140 N/A N/A 34
MIL-53-NH2(Al) Polyacrylonitrile nanofiber MOF Conversion from ALD Metal Oxide N/A N/A Good High 53
MOF-5 Silk microfiber Direct layer-by-layer synthesis N/A N/A Good Low 49
MOF-74(Mg) Polypropylene microfiber Solvothermal on ALD nucleation layers N/A N/A Good High 23
MOF-74(Zn) Polypropylene microfiber Solvothermal on ALD nucleation layers N/A N/A Good High 23
Pb(4-bpdh)(NO3)∙ Silk microfiber Direct layer-by-layer synthesis 61.5 N/A Poor High 48
(H2O)
UiO-66 Polyacrylonitrile nanofiber Electrospinning 68.7 <a> 815 Good High 57
UiO-66 Poly(vinyl cinnamate) nanofiber Electrospinning + UV cross-linking, + secondary growth 36.2 <a> 430 Poor Low 39
UiO-66 Polypropylene microfiber Solvothermal on ALD nucleation layers N/A N/A Good High 23
UiO-66 Silk microfiber Spray Drying Method 50 670 Poor Low 29
UiO-66 Polyamide-6 nanofiber Solvothermal on ALD nucleation layers 8.6 140 Good High 30
UiO-66@LiOtBu Silk microfiber Spray Drying Method N/A 130 Poor Low 29
UiO-66-NH2 Polyamide-6 nanofiber Solvothermal on ALD nucleation layers 13 196 Good High 30
UiO-67 Polyamide-6 nanofiber Solvothermal on ALD nucleation layers 19.4 447 Good High 30
ZIF-8 Polyvinylpyrrolidone nanofiber Electrospinning 56 530 Good High 56
ZIF-8 Polystyrene nanofiber Electrospinning 25 210 Good High 56
ZIF-8 Polystyrene nanofiber Electrospinning 20 <c> 303 Good Low 38
ZIF-8 Polystyrene nanofiber Electrospinning + secondary solvothermal growth 50 935 Good High 38
ZIF-8 Ultem (a polyetherimide) hollow Co-extrusion spinning 0.10 <b> N/A Good High 28
fiber
ZIF-8 Polyacrylonitrile nanofiber MOF Conversion from ALD Metal Oxide N/A N/A Good High 53
ZIF-90 Torlon® (a polyamide-imide) Solvothermal on MOF Seeds N/A N/A Good High 27
hollow fiber
Zn2(adc)2(dabco)2 Paper Inkjet printing N/A N/A N/A N/A 60
Zn2(bpdc)2(bpee) Polyvinylpyrrolidone nanofiber Electrospinning 50 <c> N/A Good Low 38
Zn4O(dmcapz)3 Silk microfiber Spray Drying Method 50 690 Poor Low 29
<a> MOF wt% calculated based on BET surface area. <b> MOF wt% estimated from metal
concentration. <c>MOF wt% estimated based on the MOF to polymer mass ratio in the
mixed solution for electrospinning.
38
1.3 Characterization Methods
Common characterization techniques for bulk MOF powders, including XRD, BET,
SEM, TEM and TGA, are generally applicable to MOF-functionalized fibers. In this part, we
will highlight a few analysis methods that are particularly useful for fibrous materials. The
methods presented here can be applied for investigation of surface composition, thin film
growth rate, MOF mass fraction, permeation property, surface roughness and mechanical
stability (MOF attachment).
1.3.1 Attenuated Total Reflection (ATR) Infrared spectroscopy
Developed by Harrick and Fahrenfort more than 50 years ago,61,62 ATR technique relies
on the total reflection of IR beam in an internal reflection element (IRE) and the evanescent
wave formed perpendicular to the reflecting surface in the IRE (Figure 1.34).63,64 To form
total reflection within the IRE, the angle of incidence θ needs to be larger than the critical
angle (θcrit):
𝑛
𝜃𝑐𝑟𝑖𝑡 = 𝑠𝑖𝑛−1 ( 2 ) (1)
𝑛1
where n1 and n2 are the refractive indices of the IRE and the sample, respectively.64 The
evanescence wave can penetrate a small distance (dp) into the sample and be absorbed by the
sample. The penetration depth is given by:
𝜆
𝑑𝑝 = (2)
2𝜋√𝑛12 𝑠𝑖𝑛2 𝜃−𝑛22
where λ is the wavelength of incident IR beam, θ is the angle of incidence, n1 and n2 are the
refractive indices of the IRE and the sample, respectively.64
39
Figure 1.34 Schematic of IR beam path in an ATR-IR.
Compared to the common transmission FTIR spectroscopy, ATR FTIR is advantageous
for probing the surface of the sample (rather than measuring the bulk) and particularly
powerful for in situ analysis of chemisorption and reactions at solid/liquid and solid/gas
interfaces.64 For thin MOF-nanofiber composites, transmission FTIR may still be useful for
characterization.39 However, ATR FTIR is generally suitable for chemical analysis and
qualitative studies of the MOF thin film growth for all MOF-functionalized fibers.
Neufeld et al. reported using ATR-IR to monitor the growth of HKUST-1 on cotton
fabrics by tracking the peak intensity at 728 cm-1 (δ(CH)), 1371 cm-1 (νs(COO)) and 1644
cm-1 (νa(COO)) (Figure 1.35). Increased peak intensity was observed at these wavenumbers,
indicating the additional MOF loadings onto the substrate during each cycle of deposition.
40
Figure 1.35 ATR-IR spectra for cotton fabric coated with different number of cycles of
HKUST-1. 32
1.3.2 UV-Vis Spectroscopy
UV-Vis absorption spectroscopy and diffuse reflectance UV-Vis spectroscopy can be
used to qualitatively and quantitatively (with other complementary analysis) monitor the
growth of MOF thin films on fibers. Especially for MOF thin films grown by layer-by-layer
method which are often too thin for XRD measurements, UV-Vis has been demonstrated as a
useful alternative approach to analyze the film growth.32,50 Meilikhov et al. used UV-Vis to
monitor the growth of HKUST-1 thin films deposited onto chemically functionalized
polyester fibers using layer-by-layer method (Figure 1.36). The absorbance at 704 nm
increases linearly as a function of the number of cycles after the second cycle, indicating the
linear growth of film thickness.50 Similarly, Neufeld et al. used diffuse reflectance UV-Vis
spectroscopy for characterizing HKUST-1 layer-by-layer synthesis on surface-functionalized
cotton fibers. The increase of absorbance at 704 nm was also found as more cycles were
deposited.32
41
Figure 1.36 (a) UV-Vis absorption spectra for different numbers of HKUST-1
layer-by-layer cycles on polyester fibers pretreated with polyvinylamine and bromoacetic
acid. (b) absorbance at 704 nm plotted as a function of cycle number.50
1.3.3 Analysis of MOF Mass Fraction
Analyzing the MOF mass fraction in the MOF-fiber composites is sometimes
challenging, when the actual mass change is very small to measure directly by weighing
methods and/or there is water adsorption on MOFs and hygroscopic fabric materials.
Meilikhov et al. used inductively coupled plasma optical emission spectroscopy
(ICP-OES) to measure the Cu concentration in the MOF-fiber composite after dissolving the
HKUST-1 layers. Using the known Cu concentration in the bulk MOF structure, MOF weight
percent can be calculated from the measured Cu concentration.50 Neufeld et al. described a
similar method using atomic absorption spectroscopy (AAS) to determine the mass fraction
of HKUST-1 in the MOF-fiber composites.32 MOF coated fabrics were sonicated in 3.5×
10-3 M HCl aqueous solution for 30 mins to fully dissolve the HKUST-1 coatings. The
42
copper concentration measured by AAS was used for the calculation of MOF mass fraction.
This method provides an accurate approach to quantify the amount of MOFs assuming all the
MOF coatings can be effectively removed. However, it is still necessary to measure the total
dry mass of MOF-coated fibers for calculating the weight percent of MOFs.
Our group developed a method to use BET surface area of MOF-coated fibers and that of
bulk MOFs to calculate the MOF mass fraction. Previously, we reported the following
equation to correlate the MOF mass fraction with the overall BET surface per unit mass of
MOF plus fibers:
𝑆𝐴𝑀𝑂𝐹+𝑓𝑖𝑏𝑒𝑟 × 𝑚𝑀𝑂𝐹+𝑓𝑖𝑏𝑒𝑟 −𝑆𝐴𝑓𝑖𝑏𝑒𝑟 × 𝑚𝑓𝑖𝑏𝑒𝑟

𝑆𝐴𝑀𝑂𝐹 ≈ (3)
𝑚𝑀𝑂𝐹+𝑓𝑖𝑏𝑒𝑟 − 𝑚𝑓𝑖𝑏𝑒𝑟
where m is the mass and SA is the surface area for each component.23 With known surface
area for the uncoated fiber substrate, the MOF-coated fibers and the bulk MOF material, the
MOF mass fraction (ω) can be calculated using Equation X:
𝑆𝐴𝑀𝑂𝐹+𝑓𝑖𝑏𝑒𝑟 = 𝑆𝐴𝑀𝑂𝐹 ∙ 𝜔 + 𝑆𝐴𝑓𝑖𝑏𝑒𝑟 ∙ (1 − 𝜔) (4)
𝑆𝐴𝑀𝑂𝐹+𝑓𝑖𝑏𝑒𝑟 −𝑆𝐴𝑓𝑖𝑏𝑒𝑟
or 𝜔= (5)
𝑆𝐴𝑀𝑂𝐹 − 𝑆𝐴𝑓𝑖𝑏𝑒𝑟
We found that the MOF weight percentage calculated in this approach is similar to what
was measured by the ICP-OES method.30 This method provides a simple and straightforward
way to calculate the MOF weight percent when instruments for elemental analysis are not
available. One prerequisite is that both the MOF coatings on the fibers and the bulk MOF
powder are fully activated for BET measurements. It is also recommended to use the MOF
crystals collected from the same batch for MOF synthesis on fibers for BET analysis in order
43
to obtain a more accurate MOF mass fraction.
1.3.4 Permeation Test
The gas permeation property is important for MOF-fiber composites especially in
applications of garments, membranes and gas filters. The permeance (P/l)i (in unites of gas
permeation unit or GPU, 1 GPU = 1×10–6cm3(STP)/cm2∙s∙cmHg) is defined with the equation
below:
𝑃 𝑄𝑖
( )𝑖 = (3)
𝑙 ∆𝑃 ∙ 𝐴
where P is permeability in units of Barrer (1 Barrer = 1×10–10 cm3(STP)∙cm / cm2∙s∙cmHg), l
(cm) is membrane thickness, Qi (cm3/s) is permeation rate for gas species i at STP, ∆𝑃
(cmHg) is pressure drop across the membrane, A (cm2) is area of membrane.65
The ideal separation factor (or selectivity) αA/B can be calculated based on the
permeability or permeance of species A and B:65

𝑃
𝑃𝐴 ( 𝑙 )𝐴
αA/B = = 𝑃 (4)
𝑃𝐵 ( 𝑙 )𝐵
Armstrong et al. tested the gas permeation properties of their UiO-66/PVCi nanofiber
membrane using an apparatus similar to those typically used for polymer mixed matrix
membranes (Figure 1.37b).39 The nanofiber membrane was fitted with rubber o-rings in a
stainless steel permeation cell. The transmembrane pressure drop was controlled at 1 psi, and
the permeation rate was measured by a 100 mL bubble flow meter. The results (Figure 1.37a)
indicate that the permeation in the UiO-66/PVCi nanofiber membrane follows Knudsen flow
mechanism.39 Knudsen flow mechanism is given in Eqn. 5:
44
𝜀 1 1
𝐹𝐾 = ( ) ( ) ( ) 𝐷 (5)
𝜏 𝐿 𝑅𝑇
where FK is permeance, 𝜀 is porosity, 𝜏 is tortuosity, L is membrane thickness, R is ideal
gas constant, T is temperature, D is pore diffusivity.66 If the Knudsen diffusion is the
dominating mechanism for the mass transport, the pore diffusivity of Eqn. 5 can be replaced
with Knudsen diffusivity (DK): 66
2 8𝑅𝑇 1/2
𝐷𝐾 = ( ) 𝑟𝑃 (6)
3 𝜋𝑀𝑤
where Mw is the gas molecular weight, 𝑟𝑃 is the pore radius. Substitution of D in Eqn. 5
with DK gives Eqn. 7: 39
𝜀 𝑟𝑃 1 1
𝐹𝐾 = 1.06 ∙ ∙ ∙ ∙ (7)
𝜏 𝐿 √𝑅𝑇 √𝑀𝑤
Figure 1.37 (a) Permeance for UiO-66/PVCi nanofiber membrane for He, N2, Ar and SF6.
(b) a schematic of the apparatus for permeation test.39
45
1.3.5 Surface Roughness Analysis
Meilikhov et al. has shown using AFM to characterize the surface roughness of the MOF
thin films.50 AFM images were acquired by scanning a surface area of 10 × 10 μm2, and the
roughness profiles were reported for both untreated fiber surface and MOF-coated fiber
surface (Figure 1.38). The author pointed out that bending roughness curve is a result of the
cylindrical fiber shape. Consequently, it is still challenging to interpret the AFM data and
report the roughness values for the MOF thin films alone. Further research is needed in this
area to address these challenges and develop a protocol for analyzing roughness for MOF
thin films on fibers.
Figure 1.38 AFM images and roughness profiles of (a) untreated polyester fiber surface and
(b) chemically modified polyester fibers deposited with 40 cycles of HKUST-1
layer-by-layer growth.50
1.3.6 Adhesion Test
The adhesion of MOF particles or thin films to the fiber substrate is a critical property
46
that can affect the application and may cause particle shedding issues with weak attachment.
Several thoughtful ideas have been reported in literature including blowing test, tensile test
and so on.23,32
Zhao et al. reported a novel blowing test to quantify the attachment of MOF crystals to
the fiber substrate, as traditional bending and rubbing tests for textiles resulted negligible
amount of mass loss due to good MOF adhesion.23 For the blowing test, compressed air (~ 40
psi) was forced through a MOF-coated fiber mats for more than 4 min. The mass change was
recorded as a function of blowing time. The loosely attached MOF crystals are expected to be
knocked off by the compress air. For example, Figure 1.39 shows the mass change of a
HKUST-1 coated PBT fiber mats during the blowing tests. The remaining mass percentage
after 4 min was used to analyze the MOF adhesion, since the mass loss tends to stabilize after
4 min of compress air blowing.23
Figure 1.39 Mass change of ALD-coated PBT fiber mats with HKUST-1 thin films during
the compress air blowing test.23
47
Neufeld et al. used cyclic tensile test to analyze the adhesion property.32 The samples
were stretched with a tensile force of 2 N (maximum strain = ~7.1% (2 mm / 28 mm)) and
32
relaxed afterwards. After 5 cycles of tensile test, the materials were imaged using SEM.
While this measurement demonstrates that the mechanical robustness of the cotton substrate
was maintained, the actual stress on MOF particles is much smaller than that applied
macroscopically on the fabric and may not be enough to cause particle shedding. SEM
Images after the tensile tests cannot provide quantitative information about the MOF mass
loss either.
48
1.4 Applications
1.4.1 Capture of Hazardous Gases
Since MOFs have large surface area, porosity and high adsorption capacity for toxic
industrial chemicals (TICs),11,12,67 integrating MOFs onto fabrics enables new composite
structures for filters, masks and suits for TIC protection.23–25,60 Zhuang et al. printed
HKUST-1 on cotton textiles and reported the color change as an indication of TIC uptake.
After exposure to NH3, HCl and H2S, the original turquoise color of HKUST-1 coatings
changes to dark blue, yellow and brown, respectively (Figure 1.40). Quantitative analysis of
TIC adsorption capacity was performed on the HKUST-1 thin films printed on quartz crystal
microbalance (QCM). The printed HKUST-1 coating shows large uptake for several TICs
and CO2. Zhuang et al. also observed the deformation of HKUST-1 films after the adsorption
tests for NH3, HCl and H2S, as the XRD patterns show no MOF peaks.60
Figure 1.40 (a) Photographs of cotton fiber mats printed with HKUST-1 dot pattern (5
printing-drying cycles) before and after exposure to NH3, HCl and H2S. (b) Relative mass
change for the HKUST-1 thin film printed on QCM during NH3 adsorption/desorption tests.
(c) Adsorption capacities of the HKUST-1 films printed on QCM for different gases. 60
49
Our group have also looked into the applications of MOF-functionalized fiber mats for
TIC protection, and quantified the gas adsorption capacity for the MOF-coated fabrics
directly using micro-breakthrough analysis (Figure 1.41).23–25 Zhao et al. reported that the
dynamic loading of NH3 for HKUST-1 coated PP fibers (pretreated with ALD Al2O3) is 5.28
± 0.40 molNH /kg(MOF+fiber). The adsorption capacity for the MOF film calculated based the
3
MOF mass fraction is 6.9 ± 0.5 mol/kg (MOF), similar to that for bulk HKUST-1.23 This result
demonstrates that the functionality of bulk MOF crystals is fully maintained after integration
on fibers, and is very promising for TIC protection for environmental preservation and
industrial safety.
Figure 1.41 NH3 breakthrough curves for bare PP fibers (PP, green blank circle), PP fibers
with ALD Al2O3 coatings (PP/ALD, orange blank diamond), HKUST-1 grown on untreated
PP fibers (MOF-PP, blue circle), HKUST-1 grown on PP/ALD fiber mats (MOF-PP/ALD,
red diamond). 23
Zhao et al. also evaluated the TIC removal performance for the HKUST-1 thin films
grown on ALD-coated PP fibers via layer-by-layer process.24 Both breakthrough time and
50
NH3 dynamic loadings were found to increase as a function of the number of LbL cycles
(Figure 42). With 40 LbL cycles (~17 wt%), the MOF-coated PP/ALD fibers show an
ammonia dynamic loading of 1.37 molNH3/kgMOF+fiber (equivalent to 7.63 molNH3/kgMOF) and
a H2S dynamic loading of 1.49 molH2S/kgMOF+fiber (equivalent to 9.46 molH2S/kgMOF). The high
adsorption capacities indicate the excellent performance of these MOF-functionalized fibers
for TIC removal.24
Figure 1.42 (a) Ammonia breakthrough curves for LbL HKUST-1 thin films grown on
ALD-coated PP fibers and control sample without MOF coating. (b) Ammonia dynamic
loading for PP/ALD fiber mats with 0 ~ 40 cycles of LbL HKUST-1. Black square points
were calculated from the breakthrough data before saturation, and red circle points were
calculated based on the breakthrough curves with desorption parts. Error bar shows standard
deviation. 24
MOF-functionalized nanofibers have also been evaluated for TIC removal. The
HKUST-1 thin films grown on PAN nanofibers via rapid room-temperature HDS-to-MOF
conversion were tested and compared to the HKUST-1 coated PP microfibers prepared using
51
the same approach. The higher MOF mass loading on PAN nanofibers leads to significant
improvement of NH3 adsorption capacity (Figure 1.43). The high dynamic loadings of NH3
as well as H2S on these MOF-functionalized fiber mats suggest that these composites are
very promising for gas filtration and protective suites. 25
Figrue 1.43 Ammonia dynamic loading on HKUST-1 MOF coated PP microfibers and
PAN nanofibers.25
1.4.2 Membrane Separation
In addition to the MOF-incorporated polymer mixed matrix membranes (MMM), free
standing MOF membranes supported by MOF-fiber scaffolds and MOF thin films on hollow
fibers are another two types of composite structures for MOF-based membranes. Polymer
MMMs with MOF fillers have been reviewed previously,68,69 and are outside the scope of
MOF-fiber composites. So the latter two membrane structures will be mainly discussed here
for separation applications.
Brown et al. synthesized continuous ZIF-90 thin films on Torlon (a polyamide-imide)
hollow fiber and measured the single-gas permeation properties using a time-lag method.27
52
They found that the permeance through the membrane decreases with increased kinetic
diameter of gas molecules (Figure 1.44a), indicating that the mass transport within ZIF-90
pores but not through defects is the permeation mechanism. The selectivity for CO2/N2 (3.5)
and CO2/CH4 (1.5) is higher than the corresponding Knudsen selectivity (0.8 and 0.6
respectively). Brown et al. also found that the permeance for all gases increases with feed
pressure and decreases as a function of temperature (Figure 1.44b). They also investigated
the permeation of hydrocarbon molecules through the Torlon/ZIF-90 membranes by
pervaporation method, and found that the membranes are selective for n-hexane over
benzene and cyclohexane. 27
Figure 1.44 (a-c) Single gas permeances for Torlon/ZIF-90 hollow fiber membranes
measured at (a) 35 oC and 50 psia, (b) 70 oC with varying pressure, (c) 50 psia with varying
o
temperature. (d) Permeances of hydrocarbon molecules measured at 22 C by
pervaporation.27
53
Dai et al. investigated the separation performance of their Ultem/ZIF-8 hollow fibers
synthesized by co-extrusion spinning.28 It is very interesting that the ZIF-8 functionalized
Ultem hollow fiber shows higher CO2 permeance and selectivity towards CO2/N2 at various
test conditions (Table 1.2). The enhancement of selectivity was ascribed to the active ZIF-8
fillers in the sheath layer. Mixed gas permeation was also tested at multiple temperatures and
pressures for this hollow fiber composite. While the selectivity for CO2/N2 decreased slightly,
increased permeance of CO2 in the Ultem/ZIF-8 membrane was observed.28
Table 1.2 CO2 permeance and permselectivity for CO2/N2 for pure Ultem and Ultem/ZIF-8
hollow fiber membranes.28
1.4.3 Destruction of Chemical Warfare Agents
The applications of MOFs in heterogeneous catalysis have been widely investigated. The
catalytic effect of the metal nodes and organic linkers in the bulk MOF structures has been
demonstrated, while post-synthetic modification to introduce addition functionalities or
incorporate nanoparticles can further enhance the catalytic performance of MOFs.70,71 One
specific application of MOF catalyst is destruction of highly toxic compounds like chemical
warfare agents (CWA), and a few MOFs have been reported with high activity towards these
reactions.72–78 Integrating MOFs onto textiles enables new composite materials for protecting
54
first-responders, soldiers and the general public from CWA threats.29,30
López-Maya et al. investigated the degradation of CWA simulants using
silk@[UiO-66@LiOtBu] composite catalysts. MOF particles (50 wt%) sprayed onto the silk
fibers were further modified with LiOtBu. The half-lives of diisopropylfluorophosphate
(DIFP), dimethylmethylphosphonate (DMMP) and 2-chloroethylethylsulfide (CEES) are 20
min, 50 min and 8 min, respectively (Figure 1.45). Experimental results also confirmed the
heterogeneous nature of these composite catalysts.29
Figure 1.45 (a) SEM images of silk fibers coated with UiO-66@LiOtBu. (b-d) Catalytic
hydrolysis profiles of CWA simulants using silk@[UiO-66@LiOtBu] composite catalysts.29
Our group synthesized conformal Zr-based MOF thin films, including UiO-66,
UiO-66-NH2 and UiO-67, on nylon nanofibers, and applied them for CWA degradation.30
Hydrolylsis of DMNP (a CWA simulant) in an aqueous buffer solution containing
N-ethylmorpholine (pH = 10) was evaluated with UiO MOF powders and MOF-coated
55
nanofiber catalysts. Both PA-6@TiO2@UiO-66-NH2 and PA-6@TiO2@UiO-67 enable
ultra-fast reaction rates for DMNP degradation with half-lives less than 8 min (Figure 1.46).
The catalytic activity of the UiO thin films were found comparable to that of the bulk UiO
powders. Catalytic degradation of O-Pinacolyl Methylphosphonofluoridate (also known as
Soman or GD) was further investigated with PA-6@TiO2@UiO-66-NH2. The results in drop
test show GD conversion in 1 h is up to 46%. This is also the first demonstration for
destruction of real nerve agents using MOF-functionalized fibers.30
Figure 1.46 (a) Reaction schematic of DMNP catalytic hydrolysis using MOF-nanofiber
catalyst. (b) UV/Visible absorption spectra to monitor DMNP degradation. (c-e) DMNP
conversion during hydrolysis with different MOF powders and MOF-functionalized
nanofiber catalysts.30
56
1.4.4 Antibacterial Protection
Textiles with immobilized nanoparticles can be potentially applied to hygienic clothing
for protection from bacterial hazard. Abbasi et al. have investigated the antimicrobial activity
of HKUST-1 coated silk fibers.31 The inhibition zones in the Müller-Hinton agar were used to
evaluate the antibacterial performance (Figure 1.47). HKUST-1/Silk fibers synthesized via
sonication-enhanced layer-by-layer method (IV-cycle 4 and IV-cycle 10) exhibit distinct
zones of inhibition against E. coli (Gram-negative bacteria) and S. aureus (Gram-positive
bacteria). The increased MOF mass loading as a result of increased number of LbL cycles
was also found to promote the antibacterial activity. The leaching of Cu ions into the aqueous
medium was considered to be possible mechanism. 31
Figure 1.47 Photos of the petri dishes for antibacterial tests using HKUST-1 coated silk
fibers. The results were compared with untreated silk, gentamycin (GM), amoxicillin (AMX),
cefepime (FEP) and oxacillin (OX). Samples in Group I were prepared via LbL synthesis
without sonication, while Group IV were synthesized with ultrasound irradiation. 31
57
1.4.5 Catalytic NO Release
Nitric oxide (NO) has biomedical application due to its antiplatelet, antibacterial and
anti-inflammatory effects.32 Reynolds and coworkers have demonstrated HKUST-1 as
effective catalyst for producing NO from S-nitrosothiols,79 applied HKUST-1 coated cotton
fabrics for catalytic NO release (Figure 1.48). Cotton swatches with HKUST-1 coatings were
found to generate a large total amount of NO (7.1 ± 1.2 μmolNO / mgMOF) over a long period
of release time. The authors also found that the surface-immobilized HKUST-1 crystals
exhibit a shorter induction period compared to mixed matrix membrane with the same type
of MOF, indicating fast diffusion rate in the MOF-fiber structure. Diffuse reflectance UV-Vis
spectra and ICP-AES analysis for the solvated Cu2+ confirm the structural integrity of the
MOF-coated fabric after the catalytic NO release. This unique application is promising for
localized NO release for biomedical uses such as wound healing.32
Figure 1.48 (Top) Catalytic reaction to form NO from S-nitrosocysteamine. (Bottom) NO
release as function of time with different catalytic materials.32
58
1.5 Outlook
A wide variety of synthesis methods have been discussed and compared in this review.
Surface functionalization processes including chemical modification and ALD for metal
oxide nucleation layers are critical for solvothermal and layer-by-layer synthesis in order to
obtain highly conformal MOF thin films. MOF Conversion from ALD oxides and HDSs
opens doors for developing novel MOF thin films on fibers especially at low processing
temperature. Electospinning, co-extrusion, and spay-drying approaches can take advantage of
commercially available MOF powders. However, challenges are still remaining. Compared to
the thousands of MOF structures reported in literature so far, the number of MOFs that have
been explored for MOF thin films on fibers is still very limited. There is huge opportunity for
the future development in this field. To date, standard metrics have not been fully established.
In addition, many applications are yet to explore for MOF-functionalized fibers.
59
References
(1) James, S. L. Metal-organic frameworks. Chem. Soc. Rev. 2003, 32 (5), 276–288.
(2) Furukawa, H.; Cordova, K. E.; O’Keeffe, M.; Yaghi, O. M. The Chemistry and
Applications of Metal-Organic Frameworks. Science 2013, 341 (6149), 1230444.
(3) Chui, S. S. A Chemically Functionalizable Nanoporous Material [Cu3(TMA)2(H2O)3]n.

Science 1999, 283 (5405), 1148–1150.
(4) Park, K. S.; Ni, Z.; Côté, A. P.; Choi, J. Y.; Huang, R.; Uribe-Romo, F. J.; Chae, H. K.;
O’Keeffe, M.; Yaghi, O. M. Exceptional chemical and thermal stability of zeolitic
imidazolate frameworks. Proc. Natl. Acad. Sci. 2006, 103 (27), 10186–10191.
(5) Cavka, J. H.; Jakobsen, S.; Olsbye, U.; Guillou, N.; Lamberti, C.; Bordiga, S.; Lillerud,
K. P. A New Zirconium Inorganic Building Brick Forming Metal Organic Frameworks
with Exceptional Stability. J. Am. Chem. Soc. 2008, 130 (42), 13850–13851.
(6) Farha, O. K.; Eryazici, I.; Jeong, N. C.; Hauser, B. G.; Wilmer, C. E.; Sarjeant, A. A.;
Snurr, R. Q.; Nguyen, S. T.; Yazaydın, A. Ö.; Hupp, J. T. Metal–Organic Framework
Materials with Ultrahigh Surface Areas: Is the Sky the Limit? J. Am. Chem. Soc. 2012,
134 (36), 15016–15021.
(7) Meek, S. T.; Greathouse, J. A.; Allendorf, M. D. Metal-Organic Frameworks: A

Rapidly Growing Class of Versatile Nanoporous Materials. Adv. Mater. 2011, 23 (2),
249–267.
(8) Wang, Z.; Cohen, S. M. Postsynthetic Covalent Modification of a Neutral

Metal−Organic Framework. J. Am. Chem. Soc. 2007, 129 (41), 12368–12369.
(9) Kim, M.; Cahill, J. F.; Fei, H.; Prather, K. A.; Cohen, S. M. Postsynthetic Ligand and
Cation Exchange in Robust Metal–Organic Frameworks. J. Am. Chem. Soc. 2012, 134
(43), 18082–18088.
(10) Brozek, C. K.; Dincă, M. Cation exchange at the secondary building units of
metal-organic frameworks. Chem. Soc. Rev. 2014, 43 (16), 5456–5467.
(11) Britt, D.; Tranchemontagne, D.; Yaghi, O. M. Metal-organic frameworks with high
capacity and selectivity for harmful gases. Proc. Natl. Acad. Sci. U. S. A. 2008, 105
(33), 11623–11627.
(12) Glover, T. G.; Peterson, G. W.; Schindler, B. J.; Britt, D.; Yaghi, O. MOF-74 building
unit has a direct impact on toxic gas adsorption. Chem. Eng. Sci. 2011, 66 (2), 163–
60
170.
(13) Li, J.-R.; Kuppler, R. J.; Zhou, H.-C. Selective gas adsorption and separation in
(14) Rosi, N. L.; Eckert, J.; Eddaoudi, M.; Vodak, D. T.; Kim, J.; O’Keeffe, M.; Yaghi, O.
M. Hydrogen storage in microporous metal-organic frameworks. Science 2003, 300
(5622), 1127–1129.
(15) Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O’Keeffe, M.; Yaghi, O. M.
Systematic design of pore size and functionality in isoreticular MOFs and their
application in methane storage. Science 2002, 295 (5554), 469–472.
(16) Xie, Y.; Wang, T.-T.; Liu, X.-H.; Zou, K.; Deng, W.-Q. Capture and conversion of CO2
at ambient conditions by a conjugated microporous polymer. Nat. Commun. 2013, 4.
(17) Boehringer, B.; Fischer, R.; Lohe, M. R.; Rose, M.; Kaskel, S.; Kuesgens, P. MOF
Shaping and Immobilization; Farrusseng, D., Ed.; Wiley-V C H Verlag Gmbh:
Weinheim, 2011.
(18) Bradshaw, D.; Garai, A.; Huo, J. Metal-organic framework growth at functional
interfaces: thin films and composites for diverse applications. Chem. Soc. Rev. 2012,
41 (6), 2344–2381.
(19) Allendorf, M. D.; Schwartzberg, A.; Stavila, V.; Talin, A. A. A roadmap to

implementing metal-organic frameworks in electronic devices: challenges and critical
directions. Chem. Weinh. Bergstr. Ger. 2011, 17 (41), 11372–11388.
(20) Furukawa, S.; Reboul, J.; Diring, S.; Sumida, K.; Kitagawa, S. Structuring of metal–
organic frameworks at the mesoscopic/macroscopic scale. Chem. Soc. Rev. 2014, 43
(16), 5700–5734.
(21) Kuesgens, P.; Siegle, S.; Kaskel, S. Crystal Growth of the Metal-Organic Framework
Cu3(BTC)2 on the Surface of Pulp Fibers. Adv. Eng. Mater. 2009, 11 (1-2), 93–95.
(22) Zhuang, J.-L.; Ceglarek, D.; Pethuraj, S.; Terfort, A. Rapid Room-Temperature
Synthesis of Metal–Organic Framework HKUST-1 Crystals in Bulk and as Oriented
and Patterned Thin Films. Adv. Funct. Mater. 2011, 21 (8), 1442–1447.
(23) Zhao, J.; Losego, M. D.; Lemaire, P. C.; Williams, P. S.; Gong, B.; Atanasov, S. E.;
Blevins, T. M.; Oldham, C. J.; Walls, H. J.; Shepherd, S. D.; et al. Highly Adsorptive,
MOF-Functionalized Nonwoven Fiber Mats for Hazardous Gas Capture Enabled by
Atomic Layer Deposition. Adv. Mater. Interfaces 2014, 1 (4), 1400040.
61
(24) Zhao, J.; Gong, B.; Nunn, W. T.; Lemaire, P. C.; Stevens, E. C.; Sidi, F. I.; Williams, P.
S.; Oldham, C. J.; Walls, H. J.; Shepherd, S. D.; et al. Conformal and highly adsorptive
metal–organic framework thin films via layer-by-layer growth on ALD-coated fiber
mats. J. Mater. Chem. A 2015, 3 (4), 1458–1464.
(25) Zhao, J.; Nunn, W. T.; Lemaire, P. C.; Lin, Y.; Dickey, M. D.; Oldham, C. J.; Walls, H.
J.; Peterson, G. W.; Losego, M. D.; Parsons, G. N. Facile Conversion of Hydroxy
Double Salts to Metal-Organic Frameworks Using Metal Oxide Particles and Atomic
Layer Deposition Thin-Film Templates. J. Am. Chem. Soc. 2015, 137 (43), 13756–
13759.
(26) Wu, Y.; Li, F.; Liu, H.; Zhu, W.; Teng, M.; Jiang, Y.; Li, W.; Xu, D.; He, D.; Hannam,
P.; et al. Electrospun fibrous mats as skeletons to produce free-standing MOF
membranes. J. Mater. Chem. 2012, 22 (33), 16971–16978.
(27) Brown, A. J.; Johnson, J. R.; Lydon, M. E.; Koros, W. J.; Jones, C. W.; Nair, S.
Continuous Polycrystalline Zeolitic Imidazolate Framework-90 Membranes on
Polymeric Hollow Fibers. Angew. Chem. Int. Ed. 2012, 51 (42), 10615–10618.
(28) Dai, Y.; Johnson, J. R.; Karvan, O.; Sholl, D. S.; Koros, W. J. Ultem((R))/ZIF-8 mixed
matrix hollow fiber membranes for CO2/N-2 separations. J. Membr. Sci. 2012, 401,
76–82.
(29) López-Maya, E.; Montoro, C.; Rodríguez-Albelo, L. M.; Aznar Cervantes, S. D.;
Lozano-Pérez, A. A.; Cenís, J. L.; Barea, E.; Navarro, J. A. R. Textile/Metal–
Organic-Framework Composites as Self-Detoxifying Filters for Chemical-Warfare
Agents. Angew. Chem. Int. Ed. 2015, 54 (23), 6790–6794.
(30) Zhao, J.; Lee, D. T.; Woodward, I. R.; Oldham, C. J.; Walls, H. J.; Yaga, R. W.; Hall,
M. G.; Peterson, G. W.; Parsons, G. N. Conformal Metal-Organic Framework Thin
Films on Nanofibers for Ultra-Fast Degradation of Chemical Warfare Agents. 2016, in
Preparation..
(31) Abbasi, A. R.; Akhbari, K.; Morsali, A. Dense coating of surface mounted CuBTC
Metal-Organic Framework nanostructures on silk fibers, prepared by layer-by-layer
method under ultrasound irradiation with antibacterial activity. Ultrason. Sonochem.
2012, 19 (4), 846–852
.
(32) Neufeld, M. J.; Harding, J. L.; Reynolds, M. M. Immobilization of Metal–Organic
Framework Copper(II) Benzene-1,3,5-tricarboxylate (CuBTC) onto Cotton Fabric as a
Nitric Oxide Release Catalyst. ACS Appl. Mater. Interfaces 2015, 7 (48), 26742–
26750.
62
(33) Zacher, D.; Baunemann, A.; Hermes, S.; Fischer, R. A. Deposition of microcrystalline
[Cu3(btc)2] and [Zn2(bdc)2(dabco)] at alumina and silica surfaces modified with
patterned self assembled organic monolayers: evidence of surface selective and
oriented growth. J. Mater. Chem. 2007, 17 (27), 2785–2792.
(34) Ullah, S.; Shariff, A. M.; Bustam, M. A.; Elkhalifah, A. E. I.; Murshid, G.; Riaz, N.;
Shimekit, B. Effect of Modified MIL-53 with Multi-Wall Carbon nanotubes and nano
Fibers on CO2 Adsorption. In Process and Advanced Materials Engineering; Ahmed, I.,
Ed.; Trans Tech Publications Ltd: Stafa-Zurich, 2014; Vol. 625, pp 870–873.
(35) Loiseau, T.; Serre, C.; Huguenard, C.; Fink, G.; Taulelle, F.; Henry, M.; Bataille, T.;
Férey, G. A Rationale for the Large Breathing of the Porous Aluminum Terephthalate
(MIL-53) Upon Hydration. Chem. Eur. J. 2004, 10 (6), 1373–1382.
(36) Patil, D. V.; Rallapalli, P. B. S.; Dangi, G. P.; Tayade, R. J.; Somani, R. S.; Bajaj, H. C.
MIL-53(Al): An Efficient Adsorbent for the Removal of Nitrobenzene from Aqueous
Solutions. Ind. Eng. Chem. Res. 2011, 50 (18), 10516–10524.
(37) Pinto, M. da S.; Augusto Sierra-Avila, C.; Hinestroza, J. P. In situ synthesis of a

Cu-BTC metal-organic framework (MOF 199) onto cellulosic fibrous substrates:
cotton. Cellulose 2012, 19 (5), 1771–1779.
P.; et al. Electrospun fibrous mats as skeletons to produce free-standing MOF
membranes. J. Mater. Chem. 2012, 22 (33), 16971–16978.
(39) Armstrong, M. R.; Arredondo, K. Y. Y.; Liu, C.-Y.; Stevens, J. E.; Mayhob, A.; Shan,
B.; Senthilnathan, S.; Balzer, C. J.; Mu, B. UiO-66 MOF and Poly(vinyl cinnamate)
Nanofiber Composite Membranes Synthesized by a Facile Three-Stage Process. Ind.
Eng. Chem. Res. 2015, 54 (49), 12386–12392.
(40) Stavila, V.; Volponi, J.; Katzenmeyer, A. M.; Dixon, M. C.; Allendorf, M. D. Kinetics
and mechanism of metal-organic framework thin film growth: systematic investigation
of HKUST-1 deposition on QCM electrodes. Chem. Sci. 2012, 3 (5), 1531–1540.
(41) Lemaire, P. C.; Zhao, J.; Williams, P. S.; Walls, H. J.; Shepherd, S. D.; Losego, M. D.;
Peterson, G. W.; Parsons, G. N. CuBTC MOF Nucleation Mechanisms on Metal Oxide
Powders and Thin Films formed by Atomic Layer Deposition. ACS Appl. Mater.
Interfaces 2016.
(42) Parsons, G. N.; George, S. M.; Knez, M. Progress and future directions for atomic
layer deposition and ALD-based chemistry. Mrs Bull. 2011, 36 (11), 865–871.
63
(43) Peng, Q.; Sun, X.-Y.; Spagnola, J. C.; Hyde, G. K.; Spontak, R. J.; Parsons, G. N.
Atomic layer deposition on electrospun polymer fibers as a direct route to Al2O3
microtubes with precise wall thickness control. Nano Lett. 2007, 7 (3), 719–722.
(44) Hyde, G. K.; Scarel, G.; Spagnola, J. C.; Peng, Q.; Lee, K.; Gong, B.; Roberts, K. G.;
Roth, K. M.; Hanson, C. A.; Devine, C. K.; et al. Atomic Layer Deposition and Abrupt
Wetting Transitions on Nonwoven Polypropylene and Woven Cotton Fabrics.
Langmuir 2010, 26 (4), 2550–2558.
(45) Gong, B.; Spagnola, J. C.; Arvidson, S. A.; Khan, S. A.; Parsons, G. N. Directed
inorganic modification of bi-component polymer fibers by selective vapor reaction and
atomic layer deposition. Polymer 2012, 53 (21), 4631–4636.
(46) Kalanyan, B.; Losego, M. D.; Oldham, C. J.; Parsons, G. N. Low-Temperature Atomic
Layer Deposition of Tungsten using Tungsten Hexafluoride and Highly-diluted Silane
in Argon. Chem. Vap. Depos. 2013, 19 (4-6), 161–166.
(47) Liu, X.; Zhang, K.-Q. Silk Fiber — Molecular Formation Mechanism, Structure-
Property Relationship and Advanced Applications. In Oligomerization of Chemical
and Biological Compounds; Lesieur, C., Ed.; InTech, 2014.
(48) Khanjani, S.; Morsali, A. Layer by layer growth of nano porous lead(II) coordination
polymer on natural silk fibers and its application in removal and recovery of iodide.
Crystengcomm 2012, 14 (23), 8137–8142.
(49) Khanjani, S.; Morsali, A. Ultrasound-promoted coating of MOF-5 on silk fiber and
study of adsorptive removal and recovery of hazardous anionic dye “congo red.”
Ultrason. Sonochem. 2014, 21 (4), 1424–1429.
(50) Meilikhov, M.; Yusenko, K.; Schollmeyer, E.; Mayer, C.; Buschmann, H.-J.; Fischer, R.
A. Stepwise deposition of metal organic frameworks on flexible synthetic polymer
surfaces. Dalton Trans. 2011, 40 (18), 4838–4841.
(51) Zhan, W.; Kuang, Q.; Zhou, J.; Kong, X.; Xie, Z.; Zheng, L. Semiconductor@Metal–
Organic Framework Core–Shell Heterostructures: A Case of ZnO@ZIF-8 Nanorods
with Selective Photoelectrochemical Response. J. Am. Chem. Soc. 2013, 135 (5),
1926–1933.
(52) Khaletskaya, K.; Turner, S.; Tu, M.; Wannapaiboon, S.; Schneemann, A.; Meyer, R.;
Ludwig, A.; Van Tendeloo, G.; Fischer, R. A. Self-Directed Localization of ZIF-8 Thin
Film Formation by Conversion of ZnO Nanolayers. Adv. Funct. Mater. 2014, 24 (30),
4804–4811.
64
(53) Bechelany, M.; Drobek, M.; Vallicari, C.; Chaaya, A. A.; Julbe, A.; Miele, P. Highly
crystalline MOF-based materials grown on electrospun nanofibers. Nanoscale 2015, 7
(13), 5794–5802.
(54) Meyn, M.; Beneke, K.; Lagaly, G. Anion-exchange reactions of hydroxy double salts.
Inorg. Chem. 1993, 32 (7), 1209–1215.
(55) Rose, M.; Boehringer, B.; Jolly, M.; Fischer, R.; Kaskel, S. MOF Processing by
Electrospinning for Functional Textiles. Adv. Eng. Mater. 2011, 13 (4), 356–360.
(56) Ostermann, R.; Cravillon, J.; Weidmann, C.; Wiebcke, M.; Smarsly, B. M.
Metal-organic framework nanofibers via electrospinning. Chem. Commun. 2011, 47
(1), 442–444.
(57) Ren, J.; Musyoka, N. M.; Annamalai, P.; Langmi, H. W.; North, B. C.; Mathe, M.
Electrospun MOF nanofibers as hydrogen storage media. Int. J. Hydrog. Energy 2015,
40 (30), 9382–9387.
(58) Husain, S. Mixed Matrix Dual Layer Hollow Fiber Membranes For Natural Gas
Separation.
(59) Ameloot, R.; Gobechiya, E.; Uji-i, H.; Martens, J. A.; Hofkens, J.; Alaerts, L.; Sels, B.
F.; De Vos, D. E. Direct Patterning of Oriented Metal-Organic Framework Crystals via
Control over Crystallization Kinetics in Clear Precursor Solutions. Adv. Mater. 2010,
22 (24), 2685 – 2688.
(60) Zhuang, J.-L.; Ar, D.; Yu, X.-J.; Liu, J.-X.; Terfort, A. Patterned Deposition of
Metal-Organic Frameworks onto Plastic, Paper, and Textile Substrates by Inkjet
Printing of a Precursor Solution. Adv. Mater. 2013, 25 (33), 4631–4635.
(61) Harrick, N. J. Study of Physics and Chemistry of Surfaces from Frustrated Total
Internal Reflections. Phys. Rev. Lett. 1960, 4 (5), 224–226.
(62) Fahrenfort, J. Attenuated total reflection. A new principle for the production of useful
infra-red reflection spectra of organic compounds. Spectrochim. Acta 1961, 17, 698–
709.
(63) Hind, A. R.; Bhargava, S. K.; McKinnon, A. At the solid/liquid interface: FTIR/ATR
— the tool of choice. Adv. Colloid Interface Sci. 2001, 93 (1–3), 91–114.
(64) Andanson, J.-M.; Baiker, A. Exploring catalytic solid/liquid interfaces by in situ

attenuated total reflection infrared spectroscopy. Chem. Soc. Rev. 2010, 39 (12), 4571–
4584.
65
(65) Ismail, A. F.; Rana, D.; Matsuura, T.; Foley, H. C. Examples of CMSM Preparation,
Characterization and Testing. In Carbon-based Membranes for Separation Processes;
Springer New York, 2011; pp 93–108.
(66) Ismail, A. F.; Rana, D.; Matsuura, T.; Foley, H. C. Other Carbon-Based Membranes. In
Carbon-based Membranes for Separation Processes; Springer New York, 2011; pp
145–246.
(67) DeCoste, J. B.; Peterson, G. W. Metal-Organic Frameworks for Air Purification of

Toxic Chemicals. Chem. Rev. 2014, 114 (11), 5695–5727.
(68) Qiu, S.; Xue, M.; Zhu, G. Metal–organic framework membranes: from synthesis to
separation application. Chem. Soc. Rev. 2014, 43 (16), 6116–6140.
(69) Seoane, B.; Coronas, J.; Gascon, I.; Etxeberria Benavides, M.; Karvan, O.; Caro, J.;
Kapteijn, F.; Gascon, J. Metal-organic framework based mixed matrix membranes: a
solution for highly efficient CO2 capture? Chem. Soc. Rev. 2015, 44 (8), 2421–2454.
(70) Lee, J.; Farha, O. K.; Roberts, J.; Scheidt, K. A.; Nguyen, S. T.; Hupp, J. T.
Metal-organic framework materials as catalysts. Chem. Soc. Rev. 2009, 38 (5), 1450–
1459.
(71) Metal Organic Frameworks as Heterogeneous Catalysts; Llabrés i Xamena, F.,

Gascon, J., Eds.; RSC Catalysis Series; Royal Society of Chemistry: Cambridge, 2013.
(72) Montoro, C.; Linares, F.; Quartapelle Procopio, E.; Senkovska, I.; Kaskel, S.; Galli, S.;
Masciocchi, N.; Barea, E.; Navarro, J. A. R. Capture of Nerve Agents and Mustard Gas
Analogues by Hydrophobic Robust MOF-5 Type Metal-Organic Frameworks. J. Am.
Chem. Soc. 2011, 133 (31), 11888–11891.
(73) Padial, N. M.; Quartapelle Procopio, E.; Montoro, C.; Lopez, E.; Enrique Oltra, J.;
Colombo, V.; Maspero, A.; Masciocchi, N.; Galli, S.; Senkovska, I.; et al. Highly
Hydrophobic Isoreticular Porous Metal-Organic Frameworks for the Capture of
Harmful Volatile Organic Compounds. Angew. Chem.-Int. Ed. 2013, 52 (32), 8290–
8294.
(74) Katz, M. J.; Mondloch, J. E.; Totten, R. K.; Park, J. K.; Nguyen, S. T.; Farha, O. K.;
Hupp, J. T. Simple and Compelling Biomimetic Metal-Organic Framework Catalyst
for the Degradation of Nerve Agent Simulants. Angew. Chem.-Int. Ed. 2014, 53 (2),
497–501.
(75) Mondloch, J. E.; Katz, M. J.; Isley Iii, W. C.; Ghosh, P.; Liao, P.; Bury, W.; Wagner, G.
66
W.; Hall, M. G.; DeCoste, J. B.; Peterson, G. W.; et al. Destruction of chemical
warfare agents using metal–organic frameworks. Nat. Mater. 2015, 14 (5), 512–516.
(76) Moon, S.-Y.; Liu, Y.; Hupp, J. T.; Farha, O. K. Instantaneous Hydrolysis of
Nerve-Agent Simulants with a Six-Connected Zirconium-Based Metal-Organic
Framework. Angew. Chem.-Int. Ed. 2015, 54 (23), 6795–6799.
(77) Peterson, G. W.; Moon, S.-Y.; Wagner, G. W.; Hall, M. G.; DeCoste, J. B.; Hupp, J. T.;
Farha, O. K. Tailoring the Pore Size and Functionality of UiO-Type Metal-Organic
Frameworks for Optimal Nerve Agent Destruction. Inorg. Chem. 2015, 54 (20), 9684–
9686.
(78) Liu, Y.; Moon, S.-Y.; Hupp, J. T.; Farha, O. K. Dual-Function Metal–Organic
Framework as a Versatile Catalyst for Detoxifying Chemical Warfare Agent Simulants.
ACS Nano 2015, 9 (12), 12358–12364.
(79) Harding, J. L.; Reynolds, M. M. Metal Organic Frameworks as Nitric Oxide Catalysts.
J. Am. Chem. Soc. 2012, 134 (7), 3330–3333.
67
CHAPTER 2 is a reprint of a manuscript published Advance Materials Interfaces.
CHAPTER 2. Highly Adsorptive, MOF-Functionalized Nonwoven
Fiber Mats for Hazardous Gas Capture Enabled by Atomic Layer
Deposition
Junjie Zhao†, Mark D. Losego†, Paul C. Lemaire†, Philip S. Williams†, Bo Gong†,
Sarah E. Atanasov†, Trent M. Blevins†, Christopher J. Oldham†, Howard J. Walls‡,
Sarah D. Shepherd‡, Matthew A. Browe#, Gregory W. Peterson#
and Gregory N. Parsons†*
†
Department of Chemical and Biomolecular Engineering, North Carolina State University,
911 Partners Way, Raleigh, NC 27695, U.S.A (E-mail: gnp@ncsu.edu)

‡
RTI International, 3040 East Cornwallis Road, Research Triangle Park, NC 27709, U.S.A.
#
Edgewood Chemical Biological Center, 5183 Blackhawk Road, Aberdeen Proving Ground,
MD 21010, U.S.A.
68
Abstract
While metal-organic frameworks (MOFs) show great potential for gas adsorption and
storage, their powder form limits deployment opportunities. Integration of MOFs on
polymeric fibrous scaffolds will enable new applications in gas adsorption, membrane
separation, catalysis, and toxic gas sensing. Here, we demonstrate a new synthesis route for
growing MOFs on fibrous materials that achieves high MOF loadings, large surface areas
and high adsorptive capacities. We find that a nanoscale coating of Al2O3 formed by atomic
layer deposition (ALD) on the surface of nonwoven fiber mats facilitates nucleation of MOFs
on the fibers throughout the mat. Functionality of MOFs is fully maintained after integration,
and MOF crystals are well attached to the fibers. Breakthrough tests for HKUST-1 MOFs
[Cu3(BTC)2] on ALD-coated polypropylene fibers reveal NH3 dynamic loadings up to
5.93±0.20 mol/kg(MOF+fiber). Most importantly, this synthetic approach is generally applicable
to a wide range of polymer fibers (e.g., PP, PET, cotton) and MOFs (e.g., HKUST-1,
MOF-74, and UiO-66).
2.1 Introduction
Hazardous gas adsorption has important implications for human health, industrial safety,
and environmental protection. Traditional adsorbents including zeolites, activated carbons,
and silica gels can abate hazardous gas emission, but there is growing need for materials with
higher capacity and better selectivity, while simultaneously providing multifunctional
performance (i.e. capture and alert capability).1,2 Metal organic frameworks (MOFs),
composed of metal cluster secondary building units and organic bridging ligands,3 exhibit
69
high surface area, good thermal stability, and have significant synthetic versatility, enabling
structures with tunable pore sizes and adjustable internal functionality.4 MOF synthesis
usually follows wet solvothermal batch methods, producing powders that require further
manipulation and handling, and may be difficult to implement for some applications.5
Integration of MOFs on polymeric fibrous scaffolds could simplify handling, regeneration
and deployment. It could also broaden practical use for filtration, separations, catalysis,
sensing and other applications.6
Several methods have been used to integrate MOF crystals on polymer fiber matrices,
including encapsulation in electrospun fibers7,8 and immobilization on fibers via solvothermal
synthesis9,10, layer-by-layer method6,11 or microwave irradiation12. Unfortunately, these
methods generally lead to small MOF loading fractions, poor MOF crystal quality, and low
Brunauer–Emmett–Teller (BET) surface areas—limiting applicability. For example,
Kuesgens et al. grew HKUST-1 crystals on pulp fibers using direct solvothermal synthesis
and found the surface area to be limited to 314 m2/g.9 Wu et al. encapsulated HKUST-1
MOFs in electrospun fibers, and also only achieved a value of 311 m2/g before repeating
secondary growth.8 These values are at least 2 times lower than pure HKUST-1 powders
grown solvothermally (692~1460m2/g).13-17 To date, no prior report has assessed the
hazardous gas adsorption capacity for MOF-functionalized fibers.
In this work, we introduce a novel synthesis route for achieving high-quality MOF
growth on polymer fibers. We use atomic layer deposition (ALD) to form a nanoscale-thick
conformal coating on polymer fibers for nucleation of solvothermally grown MOF crystals.
This approach is first demonstrated with HKUST-1 on ALD-Al2O3-coated polypropylene
70
fibers. MOF nucleation on fibers is compared between untreated and ALD-coated substrates,
while the quality of MOFs is benchmarked to pure powders. Hazardous gas (ammonia)
adsorption capacity was characterized via breakthrough tests. We also show the general
applicability of this technique to a wide range of polymer fibers (e.g., PP, PET, and cotton)
and MOFs (e.g., HKUST-1, MOF-74, and UiO-66).
2.2 Experimental Methods
Polymeric Fibrous Materials
Non-woven polypropylene (PP), polybutylene terephthalate (PBT) and cotton fiber mats
were acquired from Nonwovens Cooperative Research Center (NCRC), North Carolina State
University. Non-woven PP fiber mats are 0.30mm thick, with fiber diameter ranging from
0.6μm to 9.0μm. Non-woven PBT fiber mats are 0.76mm thick, with fiber diameter ranging
from 8.2μm to 18μm. Non-woven cotton fiber mats are 1.25mm thick, with fiber diameter
ranging from 13μm to 16μm.
Atomic layer deposition (ALD) of Al2O3
ALD Al2O3 was deposited onto polymeric fibrous substrates using a homemade hot-wall
viscous-flow vacuum reactor. Deposition pressure was kept at ~1Torr, and the temperature
was 60℃. In a typical ALD Al2O3 cycle, tri-methyl aluminum (TMA) was first dosed to the
reaction chamber for 1s, followed with 30s of N2 purge between doses. After TMA dose and
N2 purge, deionized water was dosed with another 30s of N2 purge. 200 cycles of ALD Al2O3
were selected as the standard coating thickness, while 50, 100 and 500 cycles were also used
71
to study the effect of ALD thickness on MOF growth.
Atomic layer deposition (ALD) of ZnO and TiO2
ALD ZnO and TiO2 were conducted in the same hot-wall viscous-flow vacuum reactor
as that used for ALD Al2O3. The deposition pressures were both ~1Torr, and the temperatures
were kept at 100℃. In an ALD ZnO cycle, the substrate was exposed to 2s of di-ethyl zinc
(DEZ) and 2s of deionized water alternatively, with 60s of N2 purge between dose steps. In
an ALD TiO2 cycle, precursors (TiCl4 and H2O) were dosed alternatively to the reaction
chamber for 1s, with 40s of N2 purge between dose steps. 200 cycles of ALD ZnO or TiO2
were deposited onto the fiber mats as the promoter of MOF growth.
Solvothermal growth of HKUST-1 MOF on fibrous materials
HKUST-1 MOFs were synthesized solvothermally using copper nitrate trihydrate and
1,3,5-benzene-tricarboxylic (BTC) acid as precursors. 0.42g BTC was dissolved in 12mL of
ethanol, and 0.87g copper nitrate trihydrate was dissolved in 12mL of deionized water. Fiber
mats were placed in a Teflon-lined pressure vessel then covered by the liquid reagents which
readily soaked into the substrates. The pressure vessel was sealed and crystal growth
proceeded at 120°C for 20 hours. After MOF growth, the fiber mats were dried in vacuum
over at 120°C for 12 hours. Dry samples were weighed, and mass increases were calculated
based on the dry weights of the starting substrate materials and the products.
72
Materials characterization
Scanning electron microscope (SEM) images of MOF-fiber materials were taken with an
FEI Phenom® bench-top SEM. Samples were sputter-coated with 5~10nm of Au-Pd before
imaging. X-ray diffraction (XRD) was measured with a Rigaku SmartLab X-ray diffraction
tool (Cu Kα X-ray source). BET surface areas of the MOF-fiber materials were measured
using a Quantachrome Autosorb-1C surface area and pore size analyzer. Samples were
vacuum dried at 120°C for 12h before the measurement of nitrogen adsorption isotherm.
BET surface areas were calculated based on isotherm within the P/Po range of 0.05~0.31.
Ammonia breakthrough test
Rapid micro-breakthrough analysis equipment (shown in Fig. S6) was used to NH3
adsorption of our MOF-fiber materials. NH3 was added into a ballast and pressurized, and
subsequently mixed with moisturized air stream to reach the challenge concentration of 1000
mg/m3 with 50% relative humidity (RH). The challenge gas mixture then passed through an
adsorbent column containing ~40mg of MOF-fiber material. The adsorbent column was kept
at 20°C in a themostatted water bath. The effluent stream was analyzed with a continuously
measuring gas chromatograph (HP5890 Series II) equipped with a photoionization detector.18
Ammonia breakthrough data analysis
Ammonia breakthrough time is defined as the time for the effluent signal to reach 5% of the
feed concentration. Dynamic loadings (DL in units of mol/kg) were calculated based on the
following equations.18
73
𝐶𝑓𝑒𝑒𝑑 𝐹𝑓𝑒𝑒𝑑 ∙𝑡𝑡𝑜𝑡𝑎𝑙
𝑁𝑓𝑒𝑒𝑑 = (1)
𝑀𝑤
𝑡𝑡𝑜𝑡𝑎𝑙 𝐶𝑜𝑢𝑡 𝐹𝑓𝑒𝑒𝑑
𝑁𝑜𝑢𝑡 = ∫0 𝑑𝑡 (2)
𝑀𝑤
𝑁𝑓𝑒𝑒𝑑 −𝑁𝑜𝑢𝑡
𝐷𝐿 = (3)
𝑚𝑎𝑑𝑠
Nfeed is defined as total moles of adsorbate in the feed, while Nout is defined as the moles of
adsorbate detected in the effluent stream. Both Nfeed and Nout have units of mol. Cfeed and Cout
are the concentrations of adsorbate in the feed and in the downstream, and have unites of
g/m3. Ffeed is defined as the feed flow rate in units of m3/min. Time (t) has units of min. Mw
is the molecular weight of the adsorbate in g/mol. mads is defined as the mass of adsorbent
and has units of kg.
2.3 Results and Discussion
Figure 2.1a-c illustrates the process scheme for the case of HKUST-1 grown on
ALD-Al2O3-coated polypropylene fibers. Polypropylene (PP) fibers (Figure 2.1a) in a
nonwoven mat were conformally coated with ALD Al2O3 (200 cycles at 60°C), creating a
core/shell “PP/ALD” structure with hydroxyl surface termination (Figure 2.1b).19 This ALD
coating can change wettability of fibers for thorough permeation of solvothermal solvents
(Figure S2.1),17 and improve reactivity of the fiber surfaces for MOF nucleation.20
Depending on the the deposition conditions, ALD coatings can increase surface roughness,21
which may also promote MOF nucleation. HKUST-1 [Cu3(BTC)2] was grown onto
ALD-coated fibers solvothermally using copper nitrate trihydrate and
1,3,5-benzene-tricarboxylic (BTC) acid as precursors in a water/ethanol (50/50 vol%)
74
solution (Figure 2.1c). Fiber mats were placed in a Teflon-lined pressure vessel and
immersed in the liquid reagents which readily wetted the coated fiber substrates. The vessel
was sealed, and MOF growth proceeded under the optimal condition at 120°C for 20 hours
(optimization shown in SI). Figure 2.1d shows HKUST-1 cyrstals grown densely and
conformally on an ALD-coated PP fibers. This resulting MOF-functionalized fiber mat is
referred to as“MOF-PP/ALD”.
Figure 2.1 (a-c) Schematic illustration of the synthesis route. (a) Polymer fiber substrate. (b)
Al2O3-coated polymer fiber via atomic layer deposition (ALD). The cross section in the
dashed square illustrates the conformal coating of ALD Al2O3 with hydroxyl surface
termination. (c) MOFs integated on Al2O3-coated polymer fiber using solvothermal MOF
synthesis. (d) SEM image of HKUST-1 MOF crystals grown on MOF-PP/ALD.
75
In Figure 2.2 we compare the growth of HKUST-1 on untreated and ALD-coated PP
fiber mats. MOFs grown on the untreated PP fibers (referred to as “MOF-PP”) show
macroscopic non-uniformity and appear to have nucleated homogeneously (i.e. in solution)
throughout the fiber matrix (Figure 2.2a). When coated with ALD Al2O3, the fiber mats show
uniform MOF coverage indicative of heterogeneous nucleation on the fiber surface. X-ray
diffraction data in Figure 2.2c confirms HKUST-1 crystal formation for both structures.
Figure 2.2d shows the percent mass increase after MOF growth on untreated and
ALD-coated PP fiber substrates. Mass increase percentage (m/m0) is 336±38% for
PP/ALD, and 228±83% for untreated PP respectively. The results show that the ALD
treatment enhances both the amount and uniformity of MOF growth. Normalizing the MOF
mass loading per unit substrate external surface area, we find m = 14.78 mg/cm2 on
PP/ALD to be 54% larger than on untreated PP (m = 9.61 mg/cm2). Mass gain is also more
reproducible on the ALD-Al2O3-coated fiber mats, as indicated by the boxplot’s interquartile
range (IQR) = 69% for PP/ALD vs 145% for untreated PP. Figure 2.2e shows results from
N2 isotherms and Brunauer–Emmett–Teller (BET) surface area analysis. The MOF-PP/ALD
fiber mats show an average surface area (SAMOF+fiber) of 695 ±76 m2/g(MOF+fiber), which is ~60%
higher than that of MOF-PP fibers (434 ± 198 m2/g(MOF+fiber)). This is consistent with the
higher MOF mass gain on PP/ALD substrates. The surface area of MOF-PP/ALD fiber mat
is also >2× larger than prior reports for MOFs on fibers.8,9
76
Figure 2.2 Comparison of HKUST-1 MOF grown on untreated and ALD-coated
polypropylene fiber mats. (a) SEM image of HKUST-1 MOF grown on untreated
polypropylene fiber mats (MOF-PP). Insert image is a photograph of MOF-PP (scale bar
represents 1cm). b) SEM image of HKUST-1 MOF grown on ALD-Al2O3-coated
polypropylene fiber mats (MOF-PP/ALD). Insert image is a photograph of MOF-PP/ALD
(scale bar represents 1cm). c) X-ray diffraction of MOF-PP and MOF-PP/ALD. d) Mass
increase percentage based on substrate dry weight. Interquartile range and average value
were calculated based on 16 MOF-PP samples and 12 MOF-PP/ALD samples. e) Brunauer–
Emmett–Teller (BET) surface area of MOF-fiber materials and calculated BET surface area
for MOF part (error bars represent standard deviation). BET surface area for PP fiber
substrates is 1.3~1.5 m2/g. The values were measured for uncoated and ALD coated fibers,
and the values were indistinguishable.
77
We compare the quality of the MOFs nucleated on fibers to that of pure powders. The
surface area of the MOF component in the MOF-fiber material (SAMOF) is given by:

𝑆𝐴𝑀𝑂𝐹 ≈ (4)
where SA is the surface area and m is the mass of each component. From this analysis (Figure
2.2e), we find SAMOF = ~716 ± 305 m2/g(MOF) for MOF-PP, while SAMOF = 924 ± 94 m2/g(MOF)
for MOF-PP/ALD. HKUST-1 powders similarly produced in our lab show SAMOF = 1066
m2/g. It is consistent with as-synthesized high quality HKUST-1 crystals,13-17 and essentially
similar to the MOFs we grew on PP/ALD fibers. Although post-synthesis activation maybe
further increases the MOF surface area, vacuum drying at high temperature is not suitable for
the polymer fiber substrate. Lower surface area on untreated PP is likely the result of
nonstoichiometric crystal phases or other unwanted side products.
To assess the quality of interaction between MOFs and the fiber substrates, we
performed compressed air blowing tests. We forced compressed air at ~40 psi for 4 min
through a standard-size MOF-fiber mat (10 cm2) and measured the mass change versus
air-flow time. Results are plotted in Figure S2.5. Generally, the mass loss stabilizes within a
few minutes, and the overall loss of MOF-PP/ALD is limited to ~15% of the starting mass.
Many samples show >85% MOF retention. Subsequent laboratory handling of MOF-fiber
mats after forced air testing resulted in no noticeable MOF detachment. In addition, we tested
our MOF-fiber mats in bending and rubbing tests. The amount of particles coming off during
these tests was too small to quatifiy. Rubbing test may also lead to fiber abrasion. So
compressed air blowing test was chosen to be our standard methods for MOF attachement
78
testing.
PP PP/ALD
1
0.8
0.6
C/Co
0.4
0.2
MOF-PP MOF-PP/ALD
0
0 5000 10000 15000 20000
Weighted Time (min/g)
Figure 2.3 NH3 breakthrough curves for untreated polypropylene fiber mats (PP, ○),
ALD-Al2O3-coated polypropylene fiber mats (PP/ALD, ◇ ), HKUST-1 MOF grown on
untreated polypropylene fiber mats (MOF-PP, ● ), HKUST-1 MOF grown on
ALD-Al2O3-coated polypropylene fiber mats (MOF-PP/ALD, ◆).
To evaluated MOF-functionalized fibers for hazardous gas adsorption, we used ammonia
breakthrough analysis (Figure S2.6).18 Results in Figure 3 show downstream NH3
concentration changes as a function of time after the feed flow starts. Ammonia breakthrough
is defined as the time at which the downstream signal reaches 5% of the feed concentration.
Dynamic loading (calculation shown in SI) indicates the total ammonia sorption capacity for
NH3 at saturation. The decay in NH3 signal after feed flow is terminated shows NH3
desorption rate, indicating the retention and therefore strength of sorption. Figure 2.3 shows
that the breakthrough time for MOF-PP/ALD is up to 58% longer than MOF-PP. Control PP
mats without MOFs exhibit immediate breakthrough. From the breakthrough results, the
ammonia dynamic loadings are 5.28±0.40 molNH3/kg(MOF+fiber) for MOF-PP/ALD and 4.57±
79
0.27 molNH3/kg(MOF+fiber) for MOFs on untreated fibers. Calculation results show the NH3
dynamic loading for the MOF component in MOF-PP/ALD fiber mats is 6.9 ± 0.5
mol/kg(MOF), similar to pure MOF powders synthesized in our lab (7.23 mol/kg) and
consistent with the reported values.[16] The similar values for the NH3 dynamic loadings
indicate the adsorption takes place in the bulk MOF crystals on the fibers.
In addition to ammonia, we also evaluated the performance MOF-functionalized fibers
for H2S removal. Compared with H2S adsorption by MOF powders,22 the H2S dynamic
loading for our MOF-fiber mats show good H2S loading capacity, demonstrating that the
MOF-fiber constructs have functional capacity to adsorb hazardous gases beyond NH3.
We investigated the effect of ALD coating thickness, and find that MOF crystal size
distribution and extent of growth into the fiber mat can be systematically changed with ALD
thickness. Figure S2.7 shows top-view and cross-sectional SEM images of MOFs grown on
PP fiber mats with 0, 50, 100, 200 and 500 cycles of Al2O3 ALD (roughly correspond to 0 nm,
6 nm, 12 nm, 24 nm, and 60 nm of Al2O3 respectively). MOF crystal size distributions
analyzed from both top-view and cross-sectional SEM are plotted in Figure 2.4a,b and Figure
S2.8. On the top layer of fibers (Figure 2.4a), MOF crystal sizes decrease from 24±8μm to 9
±3μm with narrowing distributions, as ALD coating thickness increases from 0 to 500
cycles. Similar trend is also found in the cross-sections of the MOF-fiber mats with different
ALD thickness (Figure 2.4b). The amount of MOFs grown into the fiber mat rises as the
number of ALD cycles increases to ~100 cycles (Figure S2.7), leading to the similar trends of
MOF loading and BET surface area shown in Figure 2.4c,d. The NH3 breakthrough tests,
80
summarized in Figure 2.4e,f show that NH3 adsorption capacity correlates closely with BET
surface area, regardless of the different crystal size distribution.
Figure 2.4 MOF-fiber mats with different ALD coating thicknesses. a) MOF crystal size
distribution on the top surface. b) MOF crystal size distribution on the cross-section. Crystal
size distributions were analyzed based on 50 measured crystal sizes on each corresponding
SEM image. c) Mass increase of the fiber mats after MOF integration. d) BET surface area of
the MOF-functionalized fiber mats and the surface area of MOF component in the fiber mats.
e) NH3 dynamic loading on MOF-functionalized fiber mats. f) NH3 breakthrough time on
MOF-functionalized fiber mats.
In addition to ALD coating thickness, we also tested the effect of ALD chemistry. ALD
81
ZnO and TiO2 coated PP fiber mats were functionalized with HKUST-1 MOFs, and the
results shown in Figure 2.5a,b are compared with MOFs on ALD Al2O3 coated PP. We find
that MOF crystals on ZnO are small, whereas those on TiO2 are generally larger than the
MOFs on Al2O3. This indicates that MOF crystal morphology can be adjusted using different
ALD coatings. We hypothesize that the wettability, surface roughness and isoelectric points
(IEP) of different ALD coatings may all affect the nucleation of MOF crystals. The
mechanism of MOF nucleation on different ALD materials is currently under investigation.
The resulting mass increase, BET surface area, and NH3 dynamic loading of these MOF-fiber
materials are shown in Table 2.1. The NH3 dynamic loading of MOF-PP/ALD(ZnO)
(5.93±0.20 mol/kg(MOF+fiber)) represents the best adsorptive capacity of MOF-fiber materials
so far. MOF-PP/ALD(TiO2) fiber mats exhibit less adsorptive capacity than
MOF-PP/ALD(Al2O3) and MOF-PP/ALD(ZnO). This may be explained with the low surface
area of the MOF component in this material (~670 m2/g(MOF)), which may suggest the
formation of impurities or the necessity of careful post-synthesis activation.
Table 2.1 Characterization of MOF-functionalized fiber mats: fiber mass gain (mg/cm2) after
MOF integration, BET surface area (m2/g) of MOF-fiber materials and NH3 dynamic loading
(mol/kg) calculated from breakthrough curves.
Fiber Mass Gain Total BET Surface NH3 Dynamic

MOF-Fiber Composites
(mg/cm2) Area (m2/g) Loading (mol/kg)
MOF-PP/ALD(ZnO) 25.21 764±11 5.93±0.20
MOF-PP/ALD(TiO2) 12.85 509±151 2.75±0.02
MOF-PP/ALD(Al2O3) 14.78 694±76 5.28±0.40
MOF-PBT/ALD(Al2O3) 17.17 485 2.52±0.23
MOF-Cotton/ALD(Al2O3) 13.59 121 1.70±1.14
82
Figure 2.5 HKUST-1 MOF grown on different ALD coatings and different polymer fibers.
a) SEM image of MOF on ALD-ZnO-coated polypropylene fiber mats
(MOF-PP/ALD(ZnO)). b) SEM image of MOF on ALD-TiO2-coated PP fiber mats
(MOF-PP/ALD(TiO2)). c) SEM image of MOF on ALD-Al2O3- coated PP fiber mats
(MOF-PP/ALD(Al2O3)). d) SEM image of MOF on ALD-Al2O3-coated polybutylene
terephthalate (PBT) fiber mats (MOF-PBT/ALD(Al2O3)). e) SEM image of MOF on
ALD-Al2O3-coated cotton fiber mats (MOF-Cotton/ALD(Al2O3)).
Finally, we demonstrate our approach can readily extend to various fibers and MOFs.
Figure 2.5d,e and Table 2.1 show HKUST-1 growth on ALD Al2O3 coated polybutylene
terephthalate (PBT) and cotton fiber mats. Cotton cellulose has hydroxyl groups that promote
ALD Al2O3 growth,23 but our results show HKUST-1 mass uptake on ALD treated cotton is
less than on other ALD-modified fibers. Initial MOF loading on cotton/ALD fiber mats is
83
comparable with that on PP/ALD and PBT/ALD fiber mats (Table 2.1), but MOFs grown on
cotton/ALD fibers tend to show poor adhesion. We hypothesize that the ALD layer enhances
nucleation, but nucleii attachement is less robust, possibly resulting from the smooth surface
after ALD treatment. This is supported by Figures 2.5(e) and S2.9 showing MOF growth in
the crevaces on the cotton fibers. The details of MOF nucleation on different surfaces are
currently under investigation. Preliminary examples of Zn-MOF-74, Mg-MOF-74 and
UiO-66 on ALD-coated nonwoven polymer fibrous substrates are given in Figure S2.10.
Integration of these MOFs on nonwoven fiber mats could potentially broaden the practical
use for filtration, separations, catalysis, sensing and other applications.
2.4 Conclusion
In conclusion, we demonstrate a new synthesis route using ALD coatings to facilitate
MOF nucleation on fibrous materials. Al2O3 ALD coatings improve the MOF macroscopic
uniformity and its coverage on fiber surfaces. MOF-PP/ALD fiber mats exhibit higher BET
surface areas and higher NH3 dynamic loading capacities than MOFs grown on untreated PP
fiber mats. These results significantly advance capability for functional MOFs on fiber
substrates. Variations in ALD surface modification are found to affect MOF crystallite size
and distribution but ultimately have minimal effect on ammonia removal performance. The
process scheme presented here provides a robust platform technology for integrating various
MOFs onto polymeric fibrous substrates for the use in human protection, industrial safety,
and environmental preservation.
84
References
(1) Li, J.-R.; Kuppler, R. J.; Zhou, H.-C. Chem. Soc. Rev. 2009, 38, 1477–1504.
(2) Lu, Z.-Z.; Zhang, R.; Li, Y.-Z.; Guo, Z.-J.; Zheng, H.-G. J. Am. Chem. Soc. 2011, 133,
4172–4174.
(3) James, S. L. Chem. Soc. Rev. 2003, 32, 276–288.
(4) Meek, S. T., Greathouse, J. A.; Allendorf, M. D. Adv. Mater. 2011, 23, 249–267.
(5) Bradshaw, D.; Garai, A.; Huo, J. Chem. Soc. Rev. 2012, 41, 2344–2381.
(6) Meilikhov, M.; Yusenko, K.; Schollmeyer, E.; Mayer, C.; Buschmann, H.-J.; Fischer,
R. A. Dalton Trans. 2011, 40, 4838–4841.
(7) Ostermann, R.; Cravillon, J.; Weidmann, C.; Wiebcke, M.; Smarsly, B. M. Chem.
Commun. 2011, 47, 442–444.
P.; Li, G. J. Mater. Chem. 2012, 22, 16971–16978.
(9) Kuesgens, P.; Siegle, S.; Kaskel, S. Adv. Eng. Mater. 2009, 11, 93–95.
(10) Pinto, M. da S.; Augusto Sierra-Avila, C.; Hinestroza, J. P. Cellulose 2012, 19, 1771–
1779.
(11) Abbasi, A. R.; Akhbari, K.; Morsali, A. Ultrason. Sonochem. 2012, 19, 846–852.
(12) Centrone, A.; Yang, Y.; Speakman, S.; Bromberg, L.; Rutledge, G. C.; Hatton, T. A. J.
Am. Chem. Soc. 2010, 132, 15687–15691.
(13) Chui, S. S. Science 1999, 283, 1148–1150.
(14) Basu, S.; Maes, M.; Cano-Odena, A.; Alaerts, L.; De Vos, D. E.; Vankelecom, I. F. J. J.
Membr. Sci. 2009, 344, 190–198.
(15) Britt, D.; Tranchemontagne, D.; Yaghi, O. M. Proc. Natl. Acad. Sci. U. S. A. 2008, 105,
11623–11627.
(16) Peterson, G. W.; Wagner, G. W.; Balboa, A.; Mahle, J.; Sewell, T.; Karwacki, C.J. J.
Phys. Chem. C 2009, 113, 13906–13917.
85
(17) Levasseur, B.; Petit, C.; Bandosz, T. J. ACS Appl. Mater. Interfaces 2010, 2,
3606-3613.
(18) Glover, T. G.; Peterson, G. W.; Schindler, B. J.; Britt, D.; Yaghi, O. Chem. Eng. Sci.
2011, 66, 163-170.
Roth, K. M.; Hanson, C. A.; Devine, C. K.; Stewart, S. M.; Hojo, D.; Na, J.-S.; Jur, J.
S.; Parsons, G. N. Langmuir 2010, 26, 2550–2558.
(20) Zacher, D.; Baunemann, A.; Hermes, S.; Fischer, R. A. J. Mater. Chem. 2007, 17,
2785–2792.
(21) Jur, J. S.; Spagnola, J. C.; Lee, K.; Gong, B.; Peng, Q.; Parsons, G. N. Langmuir 2010,
26, 8239–8244.
(22) Petit, C.; Mendoza, B.; Bandosz, T. J. ChemPhysChem 2010, 11, 3678-3684.
(23) Hyde, G. K.; Park, K. J.; Stewart, S. M.; Hinestroza, J. P.; Parsons, G. N. Langmuir
2007, 23, 9844-9849.
86
Supporting Information
160
Water Contact Angle (degree)

140
120
100
80
60
40
20
0
0 100 200 300 400 500
Number of ALD Cycles
Figure S2.1 Surface wetting ability of nonwoven polypropylene fiber mats with different
number of ALD Al2O3 cycles (0~500 cycles). 50 ALD cycles generates a very thin (~6nm)
Al2O3 coating, but this film was “thick” enough to change the wettability of PP substrates
from originally hydrophobic to hydrophilic. Contact angles of PP in ethanol are 0 degree
regardless of ALD cycles.
87
Synthesis Optimization
The temperature of MOF growth on fibers was optimized based on PP fiber mats with 200
cycles of ALD Al2O3. We studied the temperatures ranging from 100℃ to 140℃. MOF
crystals start to form on fibers when T≥110℃, and mass gain increases almost linearly with
temperature from 39% to 336% within the range of 100℃~120℃ (Shown in Fig. S2.2a).
The amount of MOF crystals grown on fabric substrates decreases when T≥120℃. This is
reasonable because at increased temperature the reaction rate is raised and may be higher
than the diffusion rate of precursor molecules into the fiber networks. Crystals quickly
forming on the outer surfaces of the substrate may block the path for molecular diffusion,
giving rise to low internal MOF growth. BET surface area of fibrous substrate plus MOFs
shown in Fig. S2.2b is consistent with the mass gains.
Figure S2.2 (a) Mass increase of PP/ALD fiber mats due to MOF growth at different
temperatures (100~140℃). (b) total BET surface area and calculated MOF surface area of
HKUST-1 functionalized PP/ALD fiber mats prepared at different temperatures
(100~140℃).
88
50μm 50μm
Figure S2.3 (a) Optical image of MOF-PP/ALD fiber mats prepared at different
temperatures (100~140℃). (b) SEM image of the side products found in MOF-PP/ALD fiber
mats synthesized at 100℃. (c) SEM image of the flower-like CuO poly-crystals found in
MOF-PP/ALD fiber mats synthesized at 140℃.
XRD (shown in Fig. S2.4) and SEM (Fig. S2.3) results show that one-dimensional crystals
dominantly formed at 100℃ are not MOF-199. These low dimensional crystals are likely to
be the reported chain polymer [Cu(BTC-H)-(H2O)3]n.1
89
140℃
130℃
Intensity (a.u.)
120℃
110℃
100℃
PP substrate
5 10 15 20 25 30 35 40 45
2θ (deg)
Figure S2.4 X-ray diffraction data of MOF-PP/ALD fiber mats prepared at different
temperatures (100~140ºC) and PP substrate.
Characteristic peaks of PP are not readily observed after MOF growth at 110 ~130 °C,
because the densely packed MOF crystals on fibers could diminish the X-ray diffraction
signal from the substrate. In a control experiment, we characterized the crystallinity of PP
fiber substrates after soaking them in ethanol/water (50:50 vol%) for 20 hours. XRD results
show the crystallinity of these PP fibers does not change in the solvothermal conditions. This
is also consistent with PP peaks being present after MOF growth at 100 ºC and 140 ºC where
MOF loading is relatively small (Figure S2.2a and S2.3a).
90
1.00
m/mo 0.90
0.80
0.70
0.60
0 1 2 3 4 5 6 7
Time (min)
Figure S2.5 Mass change of MOF funbctionalized fiber mats during the adhesion tests
(data shown are based on MOF-PBT/ALD fiber mats). The mass loss stabilizes after <4 min
of air flow, so empirically we take m/mo(t=4min) to evaluate the attachment of MOF crystals to
the fiber substrate. Mass change of MOF-PP/ALD during the adhesion tests is similar to
MOF-PBT/ALD.
91
Figure S2.6 Rapid micro-breakthrough analysis equipment.
92
Figure S2.7 Representative SEM images showing the top surface and the cross-section of
MOF-functionalized fibrous materials based on polypropylene fibers with different ALD
coating thickness.
93
Figure S2.8 MOF crystal size distribution in PP fiber mats with different ALD thickness.
94
Figure S2.9 SEM image of MOF crystals grown inside ALD-Al2O3 coated cotton fibers.
10μm
Figure S2.10 (a) SEM image of UiO-66 MOF grown on PP/ALD fiber mats. (b) SEM
image of Zn-MOF-74 grown on PP/ALD fiber mats. (c) SEM image of Mg-MOF-74 grown
on PP/ALD fiber mats.
(1) Pech, R.; Pickardt, J. Acta Crystallogr. C 1988, 44, 992.
95
CHAPTER 3 is a reprint of a manuscript published Journal of Materials Chemistry A.
CHAPTER 3. Conformal and highly adsorptive metal–organic
framework thin films via layer-by-layer growth on ALD-coated
fiber mats
Junjie Zhao†, Bo Gong†, William T. Nunn†, Paul C. Lemaire†, Eric C. Stevens†,
Fahim I. Sidi†, Philip S. Williams†, Christopher J. Oldham†, Howard J. Walls‡,
Sarah D. Shepherd‡, Matthew A. Browe‡, Gregory W. Peterson‡, Mark D. Losego§,
and Gregory N. Parsons†,*
†

‡
#
MD 21010, U.S.A.
§
School of Materials Science and Engineering, Georgia Institute of Technology, 771 Ferst Dr.
NW, Atlanta, GA 30332, U.S.A.
96
Abstract
Integration of metal-organic frameworks (MOFs) on textiles shows promise for enabling
facile deployment and expanding MOF applications. While MOFs deposited on flat
substrates can show relatively smooth surface texture, most previous reports of MOFs
integrated on fibers show poor conformality with many individual crystal domains. Here we
report a new low-temperature (<70°C) method to deposit uniform and smooth MOF thin
films on fiber surfaces using an energy enhanced layer-by-layer (LbL) method with an ALD
Al2O3 nucleation layer. Cross-sectional TEM images show a well-defined core@shell
structure of the MOF-functionalized fiber, and SEM shows a flat MOF surface texture. We
analyze the thickness and mass increase data of LbL HKUST-1 MOF thin films on
ALD-coated polypropylene fibers and find the growth rate to be 288~290 ng∙cm-2 per LbL
cycle. Unlike planar LbL MOF embodiments where adsorption capacities are difficult to
quantify, the large volume quantity on a typical fiber mat enables accurate surface area
measurement of these unique MOF morphologies. After 40 LbL cycles the MOFs on fibers
exhibit N2 adsorption BET surface areas of up to 93.6 m2/gMOF+fiber (~535 m2/gMOF) and
breakthrough test results reveal high dynamic loadings for NH3 (1.37 molNH3/kgMOF+fiber) and
H2S (1.49 molH2S/kgMOF+fiber). This synthesis route is applicable to many polymer fibers, and
the fiber@ALD@MOF structure is promising for gas filtration, membrane separation,
catalysis, chemical sensing and other applications.
97
3.1 Introduction
Metal-organic frameworks (MOFs) are crystalline organic/inorganic materials consisting
of nanoscale metal ions or cluster building blocks coordinated by organic linkers. MOFs
exhibit large surface area and high porosity,1 and many allow for post-synthetic chemical
modification.2–4 Utilizing these characteristics, MOFs have been applied to gas adsorption
and storage,5–8 separations,9–11 and catalysis.12–16 However, since MOFs are generally
synthesized in the form of insoluble powders, methods for depositing MOF thin films are in
high demand for applications such as adsorptive gas filters, smart membranes, chemical
sensors and catalytic coatings.17,18 Integration of MOFs into application-oriented
configurations could also enable new device fabrication and simplify their deployment,
pushing forward the commercialization of MOFs.
The layer-by-layer (LbL) method (or so-called “liquid phase epitaxy”) generates thin
MOF coatings of defined thickness with good homogeneity.19,20 This synthetic approach is
advantageous for controlling MOF structural interpenetration,21 and has been used to build
chemical sensors22 and photo-switches for molecule release23. While most studies of the LbL
method are based on planar substrate surfaces, including SAM surfaces on Au substrates,24–27
silicon20,22 and alumina substrates20, little is known about the LbL process on fibers.
In most examples, even relatively modest MOF loading requires substrates selected
specifically from carboxylic-group-consisting polymers (polyester28 and silk fibers29),
presumably to help promote MOF nucleation. For example, polyester fibers can be enriched
with -COOH surface groups using polyvinylamine and bromoacetic acid sequentially to
improve MOF growth.28 MOFs formed on silk via LbL method appear as isolated crystals
98
with a rough surface texture.29 Until recently,30 very few reports quantify BET surface areas
or gas adsorption capacities of MOFs on fibers, and no reports are available that analyze LbL
MOFs on fibers.
Our recent work30 introduced atomic layer deposition (ALD) as a MOF nucleation layer
to dramatically improve solvothermal MOF growth on various polymer fibrous materials.
Based on self-limiting reactions, ALD enables thin film coatings on complex 3D surfaces
with a control of thickness at the sub-nanometer scale.31 With abundant hydroxyl groups,
ALD coatings improve the wettability of hydrophobic polymers,32 and also provide
anchoring sites for the metal-containing units in MOFs.20 In this work, we show that ALD
layers on fibers also promote LbL MOF growth, yielding very uniform and smooth MOF thin
films on flexible polymer fibers.
Scheme 3.1 Schematic of the synthesis route. Polymer fiber substrates were coated with 50
cycles of ALD-Al2O3, forming a core@shell structure of “Fiber@Al2O3”. HKUST-1 MOF
thin film was grown onto ALD-coated polymer fibers via layer-by-layer synthesis method.
99
Nonwoven fiber mat materials
Nonwoven polypropylene (PP), polyethylene terephthalate (PET) and cotton fiber mats
were used as received from Nonwovens Cooperative Research Center (NCRC), North
Carolina State University. Fiber diameters of PP, PET, and cotton fiber mats are 0.6μm ~
9.0μm, 33μm ~ 35μm and 13μm ~ 16μm respectively.
Atomic layer deposition (ALD) of Al2O3
Nonwoven fiber mats were coated with Al2O3 by ALD using a homemade hot-wall
viscous-flow vacuum reactor (Figure S3.1). Deposition temperature was kept at 60℃, and
pressure was ~1Torr. A typical ALD Al2O3 cycle started with trimethyl aluminum (TMA, 98%
STREM Chemicals, Inc.) dose for 1s and subsequent N2 purge for 30s. After TMA dose and
N2 purge, deionized water was dosed to the chamber for 1s, followed with 60s of N2 purge.
50 cycles of ALD Al2O3 were selected as the standard coating thickness for layer-by-layer
MOF growth.
Layer-by-layer (LbL) growth of HKUST-1 MOF on fibrous materials
1mM 1,3,5-benzenetricarboxylic (BTC, 98%, Acros Organics) acid and 1mM copper
acetate monohydrate (99%, Sigma Aldrich) were dissolved separately in two vessels with
150 mL ethanol. Precursor solutions and rinsing solvent (ethanol) were placed in a sonicated
water bath during the LbL process. In a typical LbL cycle, non-woven fiber substrates were
dipped in BTC solution for 5 minutes, followed with ethanol rinse for 1 minute. The
100
substrates were then transferred to Cu(OAc)2 solution for 5 minutes and subsequently rinsed
in ethanol for 1 minute. During each transfer between vessels, the samples were dried in air
for ~10 s. Precursor solutions were refreshed every 10 cycles to avoid a temperature effect
when the sonication bath became heated. We observed heated sonication (>70°C) led to poor
quality of MOF thin films. The rinse solution was refreshed every 4 cycles to maintain
cleanliness. A good rinse process is critical for fully removing any unreacted species and/or
unattached nuclei and maintaining controlled LbL growth.
Scanning electron microscopy (SEM) and energy dispersive X-ray analysis (EDX) were
performed on a JEOL JSM 6010 SEM. The synthesized fiber@ALD@MOF material
structures were sputter-coated with a thin layer of Au-Pd (5~10nm) before SEM imaging.
Microtomed fiber mats were imaged in cross section using a JEOL 2010F transmission
electron microscope (TEM). Rigaku SmartLab X-ray diffraction (XRD) tool (Cu Kα X-ray
source) was used for crystalline phase analysis. Fourier transform infrared spectrometer
(FTIR, Thermo Scientific Nicolet 6700) was used to monitor LbL MOF growth on ALD-
Al2O3 coated silicon wafers. Brunauer–Emmett–Teller (BET) surface area was measured on a
Quantachrome Autosorb-1C surface area and pore size analyzer. Fiber@ALD@MOF
samples were dried in vacuum (~1 × 10−5 𝑇𝑜𝑟𝑟)at room temperature for 12h before BET
measurement. An eleven-point nitrogen adsorption isotherm was measured for MOF-coated
fiber mats at 77 K within P/Po range of 0.05~0.30.
101
Breakthrough test for NH3 and H2S
A custom-built rapid, micro-breakthrough system (shown in Scheme 3.2) was used for
characterizing NH3 and H2S adsorption of our fiber@ALD@MOF materials. Challenge gas
(NH3 or H2S) injected into a ballast was pressurized and subsequently mixed with a
moisturized air stream to achieve the target concentration of 1000 mg/m3 with 50% relative
humidity (RH). The challenge gas mixture then flowed through an adsorbent column loaded
with fiber@ALD@MOF material (~40 mg). The temperature of the adsorbent column was
maintained at 20°C in a water bath. The downstream concentration of the challenge gas was
detected with a continuously measuring gas chromatograph (HP5890 Series II) equipped
with a photoionization detector for NH3 (or a flame photometric detector for H2S).33
Breakthrough data analysis
Dynamic loadings (DL in units of mol/kg) of NH3 or H2S on MOF-coated fibers were
calculated from the breakthrough curves using the following equations.33
𝐶𝑓𝑒𝑒𝑑 𝐹𝑓𝑒𝑒𝑑 ∙ 𝑡𝑡𝑜𝑡𝑎𝑙

𝑀𝑤
𝑡𝑡𝑜𝑡𝑎𝑙 𝐶
𝑜𝑢𝑡 𝐹𝑓𝑒𝑒𝑑
𝑁𝑜𝑢𝑡 = ∫ 𝑑𝑡 (2)
0 𝑀𝑤
𝑁𝑓𝑒𝑒𝑑 − 𝑁𝑜𝑢𝑡
𝐷𝐿 = (3)
𝑚𝑎𝑑𝑠
Nfeed is defined as the total moles of challenge gas flowing through the adsorbent, while Nout
is the total moles of target gas detected in the effluent stream. Cfeed and Cout (in g/m3) are the
concentrations of challenge gas in the feed and the downstream respectively. Feed flow rate
(Ffeed) has units of m3/min, and time (t) has units of min. Mw (g/mol) is the molecular weight
102
of the challenge gas, and mads (kg) is the adsorbent mass.
Scheme 3.2 Schematic of the rapid micro-breakthrough analysis system for NH3 and H2S
breakthrough tests.
Conformal LbL MOF Thin Films Enabled by ALD-Al2O3 Nucleation Layer
Figure 3.1 displays SEM and TEM images showing the microscopic morphology of
MOF thin films grown on ALD-Al2O3 coated polypropylene (PP) fibers. 50 cycles of ALD
Al2O3 was deposited on PP fibers at 60℃, forming a Fiber@ALD core@shell structure
(PP@ALD). Subsequently, 40 cycles of LbL HKUST-1 MOF was synthesized on top of the
ALD coating, adding an extra shell layer to PP@ALD (shown in Scheme 3.1). We refer to
these MOF-functionalized materials as PP@ALD@LbL40. Figures 3.1a and 3.1b show good
uniformity and complete fiber coverage of the LbL MOF thin film. During solvothermal
MOF film growth on ALD-coated fibers, MOF crystals may nucleate homogeneously within
the fiber mat in the voids between fibers.30 However, the LbL method shows only MOFs
directly deposited on the fibers. Sufficient ethanol rinsing in sonication bath between dipping
103
steps in LbL process can remove unreacted precursor molecules and/or unattached nuclei
from the fiber mesh, ensuring MOF formation happens only on the reactive surface of the
fibers. A high magnification SEM image shown in Figure 3.1b suggests the average MOF
crystal size is ≤1 μm. In comparison, MOF crystals grown on ALD-coated PP fibers via
solvothermal synthesis are typically >5 μm.30 This indicates the LbL method is more
advantageous in controlling the surface roughness of the MOF coating.
Figure 3.1 (a,b,e) SEM images for ALD-Al2O3-coated PP fibers with 40 cycles of LbL
HKUST-1 MOF (PP@ALD@LbL40). (c,d) Cross-sectional TEM images for
PP@ALD@LbL40 showing the core@shell structure. (f-i) Energy Dispersive X-ray analysis
for PP@ALD@LbL40 showing the presence of carbon (f) from the polypropylene and the
HKUST-1 MOF, oxygen (g) from the ALD Al2O3 and the HKUST-1 MOF, aluminum (h)
from the ALD coating and copper (i) from the MOF.
Figures 3.1c-d show cross-sectional TEM images of the PP@ALD@LbL40 core@shell
structures. The well-defined thin layer (~10 nm thick) sandwiched between the fiber substrate
104
and the thick coating corresponds to the ALD-Al2O3 thin film. The ALD growth rate (~2
Å/cycle) and the smoothness of the coating deposited at 60℃ are consistent with previous
analysis of low temperature ALD on these fiber materials.34 With abundant hydroxyl
termination, this ALD layer is expected to facilitate MOF nucleation.30 On top of the ALD
coating, a layer with less uniform TEM contrast corresponds to the LbL HKUST-1 MOF thin
film. This MOF layer exhibits good conformality, with an average thickness of 117 nm. The
higher resolution TEM image in Figure 1d shows that the LbL MOF thin film is porous and
consists of nanocrystals with dimensions of about 5 nm to 8 nm. Similar nanoparticles were
also observed and reported by Wöll and co-workers for their work on LbL HKUST-1 coating
on magnetic nanoparticles.35 These nanoparticles are likely CuO produced from the
HKUST-1 during electron beam irradiation in the TEM.
The energy dispersive X-ray analysis (EDX) images for PP@ALD@LbL40 are shown in
Figures 3.1(f-i). The results show a uniform aluminum signal from the fibers, indicating the
ALD Al2O3 coating is conformal and is fully maintained after MOF integration. Copper is
also detected uniformly on the fibers, further confirming complete and uniform MOF
coverage. Carbon is expected from the polypropylene and from the MOF organic linker
(1,3,5-benzenetricarboxylic acid). The ALD thin films and the MOF coatings are both likely
to contribute to the oxygen signal detected by EDX.
To examine MOF nucleation mechanisms, we performed 20 MOF LbL cycles on PP
fibers with and without Al2O3 ALD pretreatment. SEM images of these coated fibers are
compared in Figure 2. PP fibers without an Al2O3 coating (Fig. 3.2a) exhibit non-uniform and
patchy MOF growth, whereas MOFs grown PP@ALD fibers (Fig. 3.2b) are uniform and
105
smooth. The improved nucleation and growth on the ALD-treated substrate is consistent with
that reported for solvothermal MOF growth on polymer fibers.30 Virgin PP fibers are devoid
of reactive functional groups and, therefore, are relatively inert to MOF nucleation.
Consequently, MOF nucleation likely occurs on random surface defects. Once the nucleation
seed is attached to the untreated fiber, the MOF will continue to grow preferentially at this
site leading to a patchy coverage.
(a) (b)
2μm 2μm
Figure 3.2 SEM images of (a) 20 cycles of LbL HKUST-1 MOF on untreated PP fibers and
(b) 20 cycles of LbL HKUST-1 MOF on ALD-Al2O3-coated PP fibers (PP@ALD). MOF
thin film grown on untreated PP substrate exhibits poor uniformity, while MOF coating on
PP@ALD shows good conformality with complete coverage on fibers.
Growth Rate of LbL HKUST-1 MOF on ALD Al2O3 Coatings
Figure 3.3a is an optical image showing the color change due to the growth of LbL
MOFs on ALD-coated PP fiber mats. The ALD Al2O3 (50 cycles) does not change the visual
color of the PP fiber substrate. With increased number of LbL cycles, the fiber mats transition
from white to turquoise blue. If dried in vacuum, the MOF functionalized fiber mats would
turn deep purple, consistent with the loss of water ligands on the copper paddle wheels, as
observed for solvothermally prepared HKUST-1 MOFs during vacuum degassing. The X-ray
106
diffraction data shown in Figure 3.3b further confirm the HKUST-1 crystal structure. The
diffraction signal intensity increases with the number of LbL cycles.
Figure 3.3 (a) Optical images of ALD-Al2O3-coated PP fibers with 0~40 cycles of LbL
HKUST-1 MOF. (b) X-ray diffraction data of ALD-Al2O3-coated PP fibers with 0, 20, 40
cycles of LbL HKUST-1 MOF. Green triangles (▼) represent the peaks for HKUST-1 MOF.
(c) Thickness of the MOF coating (0~40 LbL cycles) on ALD-Al2O3-coated PP fibers
measured from cross-section TEM images. Error bars represent a 95% confidence interval
based on 25 data points measured for each fiber mat. Solid line is a linear-fitted line to the
data points. (d) Percent mass increase of LbL HKUST-1 MOF (0~40 cycles) based on the dry
weight of the fiber substrates (△m/mo). Dashed line is a linear-fitted line to the data points.
107
To quantify the growth rate of the LbL process, we analyzed MOF film thickness using
high-resolution cross-sectional TEM (Fig. 3.3c) and characterized the mass change vs.
number of LbL cycles (Fig. 3.3d). Figure 3.3c shows the thickness of 0~40 cycles of LbL
MOF coatings on ALD-Al2O3 coated PP fibers. For the TEM analysis, at least 5
cross-sectional images were collected for each sample type, and 5 data points were measured
on each image. The MOF coating thickness increases linearly with the number of LbL cycles,
and the slope corresponds to a growth rate of 3.0 nm/cycle. Using the reported density for
HKUST-1 (0.96 g∙cm-3),36 the change in thickness corresponds to a mass gain of ~288
ng∙cm-2/cycle. The thickness/cycle value we find is close to 2.634 nm, the periodicity of the
HKUST-1 MOF unit cell in the [100] direction,19,36 suggesting the LbL method produces one
MOF “monolayer” per cycle. Previous results show HKUST-1 LbL growth on alumina to
be ~2.5 nm/cycle.20 Small differences in these values may point to subtle but important
differences in the LbL growth reactions. We believe that more detailed studies of half-cycle
reaction saturation and effects of process conditions on MOF growth rate need to be
performed to define true “layer by layer” deposition reaction conditions.
We also measured the mass gain for different numbers of LbL cycles on ALD-coated PP
fibers, and calculated percent mass increase (∆𝑚/𝑚𝑜 ) based on the dry mass of the
substrates. The percent mass gain (Fig. 3.3d) also scales linearly with the number of LbL
cycles, with a slope of 0.435% per cycle (or 4.35× 10−3 gMOF/gfiber per cycle). Considering
the BET surface area of ALD-coated PP fibers is ~1.5 m2/g, the mass gain corresponds to
~290 ng ∙ cm-2/cycle, showing excellent consistency between the mass gain and TEM
thickness analysis results.
108
In addition to PP fibers, we studied the growth rate of LbL HKUST-1 thin films on
ALD-Al2O3 coated silicon wafers by monitoring the characteristic peaks for carboxylate in
Fourier transform infrared (FTIR) spectra. In a typical LbL cycle, Al2O3 coated silicon wafers
were dipped sequentially in BTC and Cu(OAc)2 ethanolic solutions for 1 hour, with 5-minute
ethanol wash steps in between. FTIR spectra were collected for 1, 2, 4, 6, 8 and 10 cycles of
LbL MOF growth. The asymmetric (1645 and 1590 cm-1) and symmetric (1450 and 1370
cm-1) stretching vibration modes shown in Figure 3.4a represent the carboxylate linkers in
the MOF thin films.37 The absorbance of the carboxylate peak at 1370 cm-1 was plotted vs.
number of LbL cycles in Figure 3.4b. A linear increase in IR absorbance at 1370 cm-1 was
observed after the first 2 LbL cycles, indicating a layer-by-layer growth fashion after the
initial nucleation delay.
Adsorption Capacity of Fiber@ALD@MOF
Figure 3.5 shows the BET surface area of PP@ALD@MOF fibers (determined from
BET using an eleven point N2 isotherm, p/po=0.05~0.30, at 77K) plotted versus number of
MOF LbL cycles. ALD Al2O3 coated PP fiber mats had a surface area of ~1.5 m2/g. The
MOF coating added substantial surface area to the substrate, with values exceeding 93
m2/gMOF+fiber after 40 LbL cycles. Note that this surface area is normalized to the total mass of
MOF+fiber, so the surface area per unit mass of MOF is substantially larger. Following
procedures developed previously,30 we analyzed the MOF mass fraction on
fiber@ALD@MOF samples using careful drying and weighing protocols before and after
MOF LbL growth. With this method, after 40 LbL cycles the MOF mass fraction is ~17%,
109
giving a surface area of ~535 m2/gMOF. While the surface area in the range of ~500 m2/gMOF
shows good porosity, the value is still ~2× smaller than typical surface area for bulk
HKUST-1 crystals prepared via solvothermal synthesis. The trend in surface area vs. LbL
cycle in Figure 5 shows the surface area increases non-linearly, with a larger increase during
later LbL cycles. This indicates that the surface area of the growing MOF crystals improves
as growth proceeds, suggesting that MOF thin film deposited in initial LbL cycles may be
amorphous or have a poorer crystallinity than that coated in subsequent LbL cycles.
Figure 3.4 (a) Fourier transform infrared (FTIR) spectra for HKUST-1 thin films deposited
on ALD-Al2O3 coated silicon wafers using layer-by-layer (LbL) method. Al2O3 coated
silicon wafers were dipped sequentially in LbL precursor solutions for 1 hour with 5-minute
ethanol wash steps in between. (b) Plot of IR absorbance at ~1370 cm-1 (carboxylate
symmetric stretching vibrations) vs. number of LbL cycles. A linear increase was observed
for the absorbance at ~1370 cm-1 after 2 LbL cycles.
110
Figure 3.5 Brunauer-Emmett-Teller (BET) surface area (in units of m2/gMOF+Fiber) of
ALD-Al2O3-coated PP fibers with 0~40 cycles of LbL HKUST-1 MOF.
(a) (b)
Figure 3.6 (a) NH3 breakthrough curves for ALD-coated PP fiber mat with no LbL MOF
(■), 20 cycles of LbL MOF (▲) and 40 cycles of LbL MOF (●). (b) NH3 dynamic loading on
ALD-coated PP fiber mats with 0, 20, 40 cycles of LbL MOF. Square points (■) were
calculated based on the corresponding breakthrough curve before saturation, and circle points
(●) were calculated with the desorption part. Error bar represents standard deviation.
111
We also evaluate the adsorption capacity of the Fiber@ALD@MOF materials for
hazardous gases via breakthrough tests. Figure 3.6a shows the NH3 breakthrough curves for 0,
20, 40 cycles of LbL HKUST-1 MOF on ALD-coated PP fibers. Ammonia concentration
detected downstream changes as a function of time, and breakthrough is defined as the time
when the downstream concentration reaches 5% of the feed concentration (Co). Without
MOF coating, PP@ALD fibers show immediate breakthrough when exposed to ammonia,
consistent with zero NH3 adsorption. As the MOF mass increases, breakthrough time
increases, indicating a larger adsorption capacity. PP@ALD@LbL40 fiber mat exhibits
breakthrough time of ~700 min/gMOF+fiber and ~4200 min/gMOF, which is ~80% of the
previously reported values for MOF-fiber mats prepared via solvothermal method.30
Dynamic loadings are calculated for both saturation adsorption with and without
desorption. Saturation dynamic loadings (without desorption) represents the total sorption
capacity, which includes the physisorption and chemisorption capacity. Dynamic loading
with desorption reveals the amount of NH3 retained in the adsorbent even after the feed gas is
terminated, i.e. the amount of NH3 that is chemisorbed. Results from dynamic loading with
and without desorption are shown in Figure 6b. For both measurements, the NH3 dynamic
loading increase almost linearly with the number of LbL cycles. With 40 LbL cycles,
PP@ALD@LbL40 exhibits a dynamic loading (without desorption) of 1.37
molNH3/kgMOF+fiber, equivalent to 7.63 molNH3/kgMOF. With no MOF present, the PP@ALD
samples show nearly zero (0.12 molNH3/kgfiber) dynamic NH3 loading without desorption.
Dynamic loading of 7.63 molNH3/kgMOF agrees well with reported data for bulk HKUST-1
powder (6.6~8.9 molNH3/kgMOF),38 and similar to values reported for solvothermal MOFs
112
grown on fibers.30 With consideration of desorption, PP@ALD@40 can still retain 0.92 mole
of ammonia per kilogram of MOF+fiber.
In addition to ammonia adsorption, we also tested PP@ALD@LbL40 fiber mats for H2S
absorption. The dynamic loading without desorption is up to 1.49 molH2S/kgMOF+fiber (or 9.46
molH2S/kgMOF), and that dynamic loading with desorption is up to 1.44 molH2S/kgMOF+fiber.
Figure 3.7 (a-c) SEM images of 20 cycles of LbL HKUST-1 MOF grown on
ALD-Al2O3-coated polypropylene (a), polyethylene terephthalate (b) and cotton (c). (d)
Percent mass gain for 20 LbL cycles on PP@ALD, PET@ALD and Cotton@ALD. (e) BET
surface area of PP@ALD@LbL20, PET@ALD@LbL20 and Cotton@ALD@LbL20.
LbL MOF on ALD-coated PET and Cotton Fibers
The Fiber@ALD@MOF core@shell structure is not limited to polypropylene substrates.
113
We also deposited LbL MOF thin films on ALD-coated polyethylene terephthalate (PET) and
cotton non-woven fiber mats. Figures 7a-c compare the microscopic morphology of the LbL
MOF thin films grown on PP@ALD, PET@ALD and Cotton@ALD. MOF coating on these
substrates all exhibits good uniformity, high coverage and smooth surface texture. Figure 7d
shows percent mass increase for 20 LbL cycles on different polymer substrates. We find the
percent mass gain data decreases as the diameter of the substrate fiber increases. Assuming
the MOF thickness increase per cycle (∆ℎ, in units of cm/cycle) is same for different fibers,
the percent mass gain will be inversely proportional to the fiber diameter (Eqn. 4).
∆𝑚 𝜌𝑀𝑂𝐹 ∙𝑉𝑀𝑂𝐹 𝜌𝑀𝑂𝐹 ∙(∆ℎ∙𝑆𝐴𝑓𝑖𝑏𝑒𝑟 ) 4 𝜌𝑀𝑂𝐹 ∆ℎ

= = 𝑑𝑓𝑖𝑏𝑒𝑟 = ∙ (4)
𝑚𝑜 𝜌𝑓𝑖𝑏𝑒𝑟 ∙𝑉𝑓𝑖𝑏𝑒𝑟 𝜌𝑓𝑖𝑏𝑒𝑟 ∙(𝑆𝐴𝑓𝑖𝑏𝑒𝑟 ∙ ) 𝜌𝑓𝑖𝑏𝑒𝑟 𝑑𝑓𝑖𝑏𝑒𝑟
4
where ∆𝑚 is mass increase per cycle due to MOF deposition in units of mg/cycle, 𝑚𝑜 is
the substrate weight of ALD-coated fibers in mg, 𝜌𝑀𝑂𝐹 and 𝜌𝑓𝑖𝑏𝑒𝑟 (g∙cm-3) are the densities
of the MOF thin film and the ALD-coated fibers respectively, 𝑆𝐴𝑓𝑖𝑏𝑒𝑟 is the surface area of
the fibers in cm2, 𝑑𝑓𝑖𝑏𝑒𝑟 is the diameter of the fibers in cm. The trend predicted by
equation (4) is consistent with results in Figure 7d, showing decreasing MOF mass fraction
for cotton and PET substrates with larger average fiber diameter. The total BET surface
area (m2/gMOF+fiber) shown in Figure 7e also scales with MOF mass fraction, indicating that
the MOF surface area on these substrates is similar to the MOFs formed on polypropylene.
3.4 Conclusions
We demonstrate fabrication of fiber@ALD@MOF core@shell structures with
conformal and smooth MOF surfaces using a combination of controlled inorganic atomic
114
layer deposition and MOF layer-by-layer synthesis on natural and synthetic polymer fibers.
We use ALD Al2O3 thin film as a nucleation layer for LbL HKUST-1 MOF crystals, and
SEM and cross-sectional TEM images confirm good uniformity and high fiber coverage of
the MOF thin films. After 40 LbL cycles on Al2O3-coated PP fibers, the BET surface area for
the MOF thin films is ~535 m2/gMOF. Although the value is not as high as crystals prepared
via solvothermal methods, we expect the MOF quality and surface area per gram to increase
as growth proceeds further. Analysis of the MOF coating thickness shows the growth rate is
3.0 nm/cycle. Calculation based on both the thickness and the percent mass gain data reveal a
consistent value of 288~290 ng∙cm-2/cycle for the growth rate of LbL HKUST-1 on
ALD-coated PP fibers. Furthermore, the PP@ALD@MOF fiber mats formed by ALD/LbL
have high adsorption capacity for NH3 and H2S. The dynamic loadings of the MOF thin film
are comparable to those of bulk MOF powder, indicating the good quality of the MOF
coating. We also confirm this synthesis route is applicable to other polymer fibers, such as
PET and cotton. In addition to hazardous gas removal, this fiber@ALD@MOF structure is
also promising for catalysis, chemical sensing and many other applications.
115
References
(1) Furukawa, H.; Cordova, K. E.; O’Keeffe, M.; Yaghi, O. M. The Chemistry and
Applications of Metal-Organic Frameworks. Science 2013, 341 (6149), 1230444.
(2) Kim, M.; Cahill, J. F.; Fei, H.; Prather, K. A.; Cohen, S. M. Postsynthetic Ligand and
Cation Exchange in Robust Metal–Organic Frameworks. J. Am. Chem. Soc. 2012, 134
(43), 18082–18088.
(3) Bloch, W. M.; Burgun, A.; Coghlan, C. J.; Lee, R.; Coote, M. L.; Doonan, C. J.;
Sumby, C. J. Capturing snapshots of post-synthetic metallation chemistry in metal–
organic frameworks. Nat. Chem. 2014, 6 (10), 906–912.
(4) Brozek, C. K.; Dincă, M. Cation exchange at the secondary building units of metal–
organic frameworks. Chem Soc Rev 2014, 43 (16), 5456–5467.
(5) Rosi, N. L.; Eckert, J.; Eddaoudi, M.; Vodak, D. T.; Kim, J.; O’Keeffe, M.; Yaghi, O.
M. Hydrogen storage in microporous metal-organic frameworks. Science 2003, 300
(5622), 1127–1129.
(6) Britt, D.; Tranchemontagne, D.; Yaghi, O. M. Metal-organic frameworks with high
capacity and selectivity for harmful gases. Proc. Natl. Acad. Sci. U. S. A. 2008, 105
(33), 11623–11627.
(7) Murray, L. J.; Dinca, M.; Long, J. R. Hydrogen storage in metal-organic frameworks.
Chem. Soc. Rev. 2009, 38 (5), 1294–1314.
(8) Sumida, K.; Rogow, D. L.; Mason, J. A.; McDonald, T. M.; Bloch, E. D.; Herm, Z. R.;
Bae, T.-H.; Long, J. R. Carbon Dioxide Capture in Metal-Organic Frameworks. Chem.
Rev. 2012, 112 (2), 724–781.
(9) Pan, L.; Olson, D. H.; Ciemnolonski, L. R.; Heddy, R.; Li, J. Separation of
Hydrocarbons with a Microporous Metal–Organic Framework. Angew. Chem. Int. Ed.
2006, 45 (4), 616–619.
(10) Li, J.-R.; Kuppler, R. J.; Zhou, H.-C. Selective gas adsorption and separation in
(11) Couck, S.; Denayer, J. F. M.; Baron, G. V.; Remy, T.; Gascon, J.; Kapteijn, F. An
Amine-Functionalized MIL-53 Metal-Organic Framework with Large Separation
Power for CO2 and CH4. J. Am. Chem. Soc. 2009, 131 (18), 6326–6327.
(12) Lee, J.; Farha, O. K.; Roberts, J.; Scheidt, K. A.; Nguyen, S. T.; Hupp, J. T.
116
Metal-organic framework materials as catalysts. Chem. Soc. Rev. 2009, 38 (5), 1450–
1459.
(13) Wang, C.; Xie, Z.; deKrafft, K. E.; Lin, W. Doping Metal-Organic Frameworks for
Water Oxidation, Carbon Dioxide Reduction, and Organic Photocatalysis. J. Am.
Chem. Soc. 2011, 133 (34), 13445–13454.
(14) Xie, Y.; Wang, T.-T.; Liu, X.-H.; Zou, K.; Deng, W.-Q. Capture and conversion of CO2
at ambient conditions by a conjugated microporous polymer. Nat. Commun. 2013, 4.
(15) Sun, D.; Fu, Y.; Liu, W.; Ye, L.; Wang, D.; Yang, L.; Fu, X.; Li, Z. Studies on
Photocatalytic CO2 Reduction over NH2-UiO-66(Zr) and Its Derivatives: Towards a
Better Understanding of Photocatalysis on Metal–Organic Frameworks. Chem. – Eur.
J. 2013, 19 (42), 14279–14285.
(16) Wang, S.; Yao, W.; Lin, J.; Ding, Z.; Wang, X. Cobalt Imidazolate Metal–Organic
Frameworks Photosplit CO2 under Mild Reaction Conditions. Angew. Chem. Int. Ed.
2014, 53 (4), 1034–1038.
(17) Bo, L.; Fischer, R. A. Liquid-phase epitaxy of metal organic framework thin films. Sci.
China-Chem. 2011, 54 (12), 1851–1866.
(18) Bradshaw, D.; Garai, A.; Huo, J. Metal-organic framework growth at functional
interfaces: thin films and composites for diverse applications. Chem. Soc. Rev. 2012,
41 (6), 2344–2381.
(19) Munuera, C.; Shekhah, O.; Wang, H.; Wöll, C.; Ocal, C. The controlled growth of
oriented metal–organic frameworks on functionalized surfaces as followed by
scanning force microscopy. Phys. Chem. Chem. Phys. 2008, 10 (48), 7257–7261.
(20) Stavila, V.; Volponi, J.; Katzenmeyer, A. M.; Dixon, M. C.; Allendorf, M. D. Kinetics
and mechanism of metal-organic framework thin film growth: systematic investigation
of HKUST-1 deposition on QCM electrodes. Chem. Sci. 2012, 3 (5), 1531–1540.
(21) Shekhah, O.; Wang, H.; Paradinas, M.; Ocal, C.; Schuepbach, B.; Terfort, A.; Zacher,
D.; Fischer, R. A.; Woell, C. Controlling interpenetration in metal-organic frameworks
by liquid-phase epitaxy. Nat. Mater. 2009, 8 (6), 481–484.
(22) Lu, G.; Hupp, J. T. Metal-Organic Frameworks as Sensors: A ZIF-8 Based Fabry-Perot
Device as a Selective Sensor for Chemical Vapors and Gases. J. Am. Chem. Soc. 2010,
132 (23), 7832–7833.
(23) Heinke, L.; Cakici, M.; Dommaschk, M.; Grosjean, S.; Herges, R.; Bräse, S.; Wöll, C.
117
Photoswitching in Two-Component Surface-Mounted Metal–Organic Frameworks:
Optically Triggered Release from a Molecular Container. ACS Nano 2014, 8 (2),
1463–1467.
(24) Shekhah, O.; Wang, H.; Kowarik, S.; Schreiber, F.; Paulus, M.; Tolan, M.; Sternemann,
C.; Evers, F.; Zacher, D.; Fischer, R. A.; et al. Step-by-step route for the synthesis of
metal-organic frameworks. J. Am. Chem. Soc. 2007, 129 (49), 15118–15119.
(25) Shekhah, O.; Wang, H.; Zacher, D.; Fischer, R. A.; Wöll, C. Growth Mechanism of
Metal–Organic Frameworks: Insights into the Nucleation by Employing a
Step-by-Step Route. Angew. Chem. Int. Ed. 2009, 48 (27), 5038–5041.
(26) Liu, B.; Ma, M.; Zacher, D.; Betard, A.; Yusenko, K.; Metzler-Nolte, N.; Woell, C.;
Fischer, R. A. Chemistry of SURMOFs: Layer-Selective Installation of Functional
Groups and Post-synthetic Covalent Modification Probed by Fluorescence Microscopy.
J. Am. Chem. Soc. 2011, 133 (6), 1734–1737.
(27) Arslan, H. K.; Shekhah, O.; Wohlgemuth, J.; Franzreb, M.; Fischer, R. A.; Woell, C.
High-Throughput Fabrication of Uniform and Homogenous MOF Coatings. Adv.
Funct. Mater. 2011, 21 (22), 4228–4231.
(28) Meilikhov, M.; Yusenko, K.; Schollmeyer, E.; Mayer, C.; Buschmann, H.-J.; Fischer, R.
A. Stepwise deposition of metal organic frameworks on flexible synthetic polymer
surfaces. Dalton Trans. 2011, 40 (18), 4838–4841.
(29) Abbasi, A. R.; Akhbari, K.; Morsali, A. Dense coating of surface mounted CuBTC
Metal-Organic Framework nanostructures on silk fibers, prepared by layer-by-layer
method under ultrasound irradiation with antibacterial activity. Ultrason. Sonochem.
2012, 19 (4), 846–852.
Atomic Layer Deposition. Adv. Mater. Interfaces 2014, 1, 1400040 DOI:
10.1002/admi.201400040.
(31) Parsons, G. N.; George, S. M.; Knez, M. Progress and future directions for atomic
layer deposition and ALD-based chemistry. MRS Bull. 2011, 36 (11), 865–871.
Langmuir 2010, 26 (4), 2550–2558.
118
(33) Glover, T. G.; Peterson, G. W.; Schindler, B. J.; Britt, D.; Yaghi, O. MOF-74 building
unit has a direct impact on toxic gas adsorption. Chem. Eng. Sci. 2011, 66 (2), 163–
170
.
(34) Jur, J. S.; Spagnola, J. C.; Lee, K.; Gong, B.; Peng, Q.; Parsons, G. N.
Temperature-Dependent Subsurface Growth during Atomic Layer Deposition on
Polypropylene and Cellulose Fibers. Langmuir 2010, 26 (11), 8239–8244.
(35) Silvestre, M. E.; Franzreb, M.; Weidler, P. G.; Shekhah, O.; Woell, C. Magnetic Cores
with Porous Coatings: Growth of Metal-Organic Frameworks on Particles Using
Liquid Phase Epitaxy. Adv. Funct. Mater. 2013, 23 (9), 1210–1213.
(36) Chui, S. S. A Chemically Functionalizable Nanoporous Material [Cu3(TMA)2(H2O)3]n.

Science 1999, 283 (5405), 1148–1150.
(37) Petit, C.; Mendoza, B.; Bandosz, T. J. Hydrogen Sulfide Adsorption on MOFs and
MOF/Graphite Oxide Composites. Chemphyschem 2010, 11 (17), 3678–3684.
(38) Peterson, G. W.; Wagner, G. W.; Balboa, A.; Mahle, J.; Sewell, T.; Karwacki, C. J.
Ammonia Vapor Removal by Cu3(BTC)2 and Its Characterization by MAS NMR. J.
Phys. Chem. C 2009, 113 (31), 13906–13917.
119
Pressure
Gauge
Mass Flow
Controller Samples
N2 Load
Lock
Gate
Valve
Furnace
TMA
H2O
To Pump
Figure S3.1 Schematic of the homemade hot-wall viscous-flow ALD reactor used for ALD
Al2O3 coatings on fibers. In an ALD cycle, trimethylaluminum (TMA) and water are dosed
sequentially into the chamber, with a purge step of inert gas (N2) in between. Deposition
temperature is controlled by the furnace, and all the gas lines and valves are wrapped with
heating tapes to prevent precursor condensation.
120
CHAPTER 4 is a reprint of a manuscript published Journal of American Chemical Society.
CHAPTER 4. Facile Conversion of Hydroxy Double Salts to
Metal-Organic Frameworks Using Metal Oxide Particles and
ALD Thin Film Templates
Junjie Zhao †, William T. Nunn †, Paul C. Lemaire †, Yiliang Lin †, Michael D. Dickey †,
Howard J. Walls ‡, Gregory W. Peterson #, Mark D. Losego §, and Gregory N. Parsons †,*
†

‡
#
MD 21010, U.S.A.
§
School of Materials Science and Engineering, Georgia Institute of Technology, 771 Ferst Dr.
NW, Atlanta, GA 30332, U.S.A.
121
Abstract
Rapid room-temperature synthesis of metal-organic frameworks (MOFs) is highly
desired for industrial implementation and commercialization. Here, we find that a (Zn,Cu)
hydroxy double salt (HDS) intermediate formed in situ from ZnO particles or thin films
enables rapid growth (< 1 min) of HKUST-1 (Cu3(BTC)2) at room temperature. The
space-time-yield reaches over 3×104 kg∙m-3∙d-1, which is at least one order of magnitude
greater than any prior report. The high anion exchange rate of (Zn,Cu) hydroxy nitrate HDS
drives the ultra-fast MOF formation. Similarly, we obtained Cu-BDC, ZIF-8 and IRMOF-3
structures from HDS’s, demonstrating synthetic generality. Using ZnO thin films deposited
via atomic layer deposition (ALD), MOF patterns are obtained on pre-patterned surfaces, and
dense HKUST-1 coatings are grown onto various form factors including polymer spheres,
silicon wafers, and fibers. Breakthrough tests show the MOF-functionalized fibers have high
adsorption capacity for toxic gases. This rapid synthesis route is also promising for new
MOF-based composite materials and applications.
4.1 Introduction
Metal-organic frameworks (MOFs) are a class of crystalline porous materials that exhibit
high surface area and large pore volume.1 Versatile combinations of MOF constituents enable
structural design and pore size control,2 and post synthetic modification can introduce
additional internal functionality.3 While MOFs are promising for gas adsorption and
separation,4 catalysis/photocatalysis,5 chemical sensing6 and other applications,7 the poor
synthesis rates (space-time-yield typically less than 300 kg∙m-3∙d-1) and harsh processing
122
conditions (high temperature and pressure) of traditional solvothermal methods still remain
as major hurdles for industrial implementation of MOFs and MOF-functionalized
composites.8 Therefore, new synthetic routes are highly desired to enable fast MOF
formation at low reaction temperature.
Several strategies have been attempted to tackle this challenge, including
mechanochemical methods,9 irradiation-assisted synthesis,10 and electrochemical
approaches.11 While these methods can reduce the reaction time and synthesis temperature
for MOF production, large amounts of external energy input are required to enable the
chemical reactions. Recently, metal oxides and hydroxides have been reported to act as
nucleating agents or sources of cations for room-temperature MOF synthesis.12-13 However,
the incorporation of residual oxide/hydroxide seeds within the MOF crystals due to
insufficient conversion affect the purity and properties of these MOFs. Although extending
reaction time to over 1 hour12 or increasing reaction temperature to >95℃13 can reduce the
oxide/hydroxide residue in the final product, alternative approaches with rapid conversion at
~25 °C are still needed.
Here we report a novel synthesis method using hydroxy double salts (HDS’s) as
intermediates. HDS’s are layered compounds consisting of cationic sheets connected by
inorganic/organic interlamellar anions.14 These materials are generally synthesized by
reacting one divalent metal oxide (MeO) with another different divalent cation (M2+). Similar
to layered double hydroxides, HDS’s show excellent anion exchangeability, and are
promising for catalysis15 and pharmaceutical applications.16 We find, for the first time, that
HDSs can undergo facile conversion and enable rapid formation of MOFs at room
123
temperature. Our space-time-yield (STY) for HKUST-1 (Cu3(BTC)2) reaches up to 3.6×104
kg∙m-3∙d-1, which is at least one order of magnitude greater than any prior report.8,11-12
Figure 4.1a depicts our rapid room-temperature approach for synthesizing HKUST-1
powder. Cu(NO3)2 aqueous solution and DMF are added to ZnO nanoslurries (dispersed in
H2O, average particle size = 252±23 nm, Figure S4.1) at room temperature to form a light
blue suspension of (Zn,Cu) hydroxy nitrate HDS. H3BTC ethanolic solution is immediately
added to the suspension, turning it turquoise in about 10 seconds (Video S4.1 in the
supporting information). This turquois color is a definitive visual indication of HKUST-1
formation. The powder product was immediately filtered and washed with ethanol after 1
minute of reaction, and subsequently dried in a vacuum oven at 120℃ for 6 hours.
Figure 4.1b compares simulated and experimental X-ray diffraction (XRD) patterns for
Cu3(BTC)2 powders obtained from rapid room-temperature synthesis and 85 °C solvothermal
synthesis. All peak positions match the simulated pattern, indicating the product synthesized
from HDS is crystalline HKUST-1. The MOF powder synthesized at 85℃shows a somewhat
larger (111) peak, possibly a result of dehydration and rehydration. The sharp XRD peaks
confirm good crystallinity of the HKUST-1.
124
Figure 4.1 (a) Schematic of the rapid room-temperature synthesis route for Cu3(BTC)2.
ZnO reacts with Cu(NO3)2 to form (Zn,Cu) hydroxy double salt. The (Zn,Cu) HDS converts
to Cu3(BTC)2 via fast anion exchange. (b) Powder XRD patterns for Cu3(BTC)2 synthesized
in rapid room-temperature method (black) and solvothermal method (red), and simulated
Cu3(BTC)2 pattern (blue). (c) Percent yield in 1 minute of reaction (black circle) and
space-time-yield for the rapid room temperature synthesis (red diamond). Insert image in (c)
is an SEM image of a Cu3(BTC)2 crystal showing the octahedral shape.
125
SEM images (insert image in Figure 4.1c and Figure S4.2a-b) reveal the MOF particles
obtained from the rapid synthesis have an average size of 1.17±0.40 μm (Figure S4.2c). The
octahedron crystal shape is also consistent with previous observations for this fcc-type MOF
crystal.17 After the rapid MOF synthesis, ZnO residue was not observed in SEM images or in
X-ray diffraction patterns, and no Zn was detected on the MOF surface by energy dispersive
X-ray (EDX) analysis (Figure S4.3) or by time-of-flight secondary ion mass spectroscopy
(ToF-SIMS) (Figure S4.4). Inductively coupled plasma optical emission spectroscopy
(ICP-OES) was also used to characterize the concentration of Zn2+ residue in the HKUST-1
powder after degrading the crystals in strong acids and H2O2.18 ICP-OES data (Table S4.1)
reveal Zn2+ concentration is generally less than 0.70 wt%, demonstrating high purity within
the HKUST-1 crystals.
We evaluated the BET surface area of the HKUST-1 powder based on the N2 adsorption
isotherm measured at 77 K. Figure S4.5 shows the N2 adsorption and desorption curves for
the powder prepared via rapid synthesis. The average BET surface area (1895±84 m2/g) is
high compared to previous reports,17,19 also suggesting the high purity of the HKUST-1
powder, as any residue with low surface area will precipitously reduce the overall BET
surface area.
Figure 4.1c shows STY (kg∙m-3∙d-1) for the rapid room-temperature synthesis and the
yield after 1 min of reaction. Both the STY and the yield increase linearly as a function of the
n(ZnO):n(Cu(NO3)2) molar ratio (n(H3BTC):n(Cu(NO3)2) was kept at 1:1.8). Previous
high-throughput synthesis methods for HKUST-1 show STY ranging from 225 to 1842
kg∙m-3∙d-1.11-12 Our synthesis strategy reaches an STY up to 3.6×104 kg∙m-3∙d-1, an
126
improvement of more than one order of magnitude. The yield also reaches up to 98% in just
1 minute of reaction, showing significant promise for scale-up processing. While a further
increase of the n(ZnO):n(Cu(NO3)2) ratio can lead to a higher apparent yield, a corresponding
drop in the BET surface area indicates incomplete HDS conversion due to insufficient
H3BTC reactant. By maintaining the ratio n(ZnO):n(Cu(NO3)2) ≤ 0.8, all HKUST-1 samples
show BET surface area exceeding 1800 m2/g.
We hypothesized that HDS could be the critical intermediate during the rapid
room-temperature MOF synthesis, because (Zn,Cu) hydroxy nitrate HDS can be converted
from ZnO and has been reported with high reaction rate for anion exchange.14 To understand
the reaction mechanism for the rapid MOF growth, we investigated particularly how HDS
forms and how it converts to MOF. Specifically, ZnO thin films (34.4±0.6nm) were deposited
on IR-transparent silicon wafers via atomic layer deposition (ALD) and soaked in Cu(NO3)2
aqueous solution for 1 min, followed by XRD analysis. The resulting XRD pattern (Figure 2a)
matches well with prior reports for (Zn,Cu) hydroxy nitrate.14 Subsequently, the HDS was
exposed to H3BTC solution (DMF : H2O : EtOH = 1 : 1 : 1 volume ratio) for only 30 s, and
analyzed by XRD. The XRD peaks representative for Cu3(BTC)2 in the corresponding
pattern and octahedron crystals found on the surface in SEM images (Figure S6) confirm the
formation of the targeting MOF. The decreased intensity of HDS XRD peaks indicates HDS
is partially consumed after the short exposure to H3BTC.
127
Figure 4.2 (a) XRD patterns for Cu3(BTC)2 powder (black), ALD ZnO surface after
exposure to Cu(NO3)2 for 1 min (blue) and subsequently to H3BTC for 30 s (red). Blue dot
lines represent the (Zn,Cu) hydroxy double salt. (b) FTIR difference spectra for ALD ZnO in
the as-deposited form (black, Si as background), after exposure to Cu(NO3)2 for 1 min
(magenta, previous spectrum as background), and after exposure to H3BTC for 30 s (green,
previous spectrum as background) and the final Cu3(BTC)2 spectrum (orange, Si as
background). (c-d) HAADF STEM images for the cross section of the Cu3(BTC)2 grown on
ALD ZnO coated silicon wafer. The green box in (c) indicates the location of (d-i). (e-i) High
resolution EDX mapping images of the cross section. Scale bars in (b-g) represent 50nm.
128
Figure 4.2b shows FTIR difference spectra collected for ALD ZnO thin film in the
as-deposited form and after sequential exposure to Cu(NO3)2 and H3BTC solutions. After
soaking in Cu(NO3)2 aqueous solution for 1 min, the negative–going mode between 400 cm-1
and 480 cm-1 confirms loss of ZnO, and the appearance of υ(NO− -1 -1
3 ) (1360 cm and 1420 cm )
and the distinct O-H group modes (3300 cm-1 ~ 3620 cm-1) indicate the formation of hydroxy
nitrate.20 After subsequent exposure to the H3BTC solution for 30 s, the peaks for NO−
3 and
O-H diminish while the symmetric (1378 cm-1) and asymmetric (1649 cm-1) stretching
modes for the carboxylate groups appear. This spectrum change clearly reveals the anion
exchange process in the (Zn,Cu) HDS, and further supports our proposed reaction path
(Figure 4.1a); ZnO reacts with Cu(NO3)2 to form (Zn,Cu) hydroxy nitrate HDS, followed by
rapid anion exchange between NO−

3 and OH
−
in the HDS and BTC3-, yielding ultra-fast
formation of HKUST-1 at room temperature.
We further studied the reactions on ZnO surface when it is exposed simultaneously to
both Cu(NO3)2 and H3BTC. ALD ZnO (58.0±0.6 nm) deposited on a silicon wafer was
soaked in the mixed precursor solution for 5 minutes without stirring. Cross-sections of the
resulting thin films were imaged using high angle annular dark field (HAADF) scanning
transmission electron microscope (STEM) (Figure 4.2c-d), and high-resolution EDX
mapping (Figure 4.2e-i) was used to analyze the elemental distribution in the cross section.
Figure 4.2c shows the abrupt interface between the nucleation layer and the silicon substrate,
and Figure 4.2f indicates Zn is only present in the nucleation layer. The nucleation layer also
reveals a uniformly distributed Cu signal in addition to Zn and O, further evidence for the
formation of (Zn,Cu) HDS. The C, O and Cu signals in the MOF layer are consistent with
129
Cu3(BTC)2. These results therefore indicate that the HDS is an important intermediate in the
rapid MOF synthesis, even in a one–pot batch procedure.
Figure 4.3 (a) Schematic of the fabrication procedure for HKUST-1 patterns. (b-e) SEM
image and EDX mapping images for a star-shape HKUST-1 pattern.
To demonstrate the generality of HDS for MOF synthesis, we synthesized three other
MOFs via HDS intermediates. Table 1 summarizes these reaction routes (XRD in Figure
S4.7-4.9). We formed (Zn,Zn) hydroxy acetate HDS by reacting ZnO with
Zn(CH3CO2)2∙2H2O in deionized water at room temperature for 24h.21 ZIF-8 and IRMOF-3
were obtained from this (Zn,Zn) HDS within 10 min at room temperature. Furthermore,
(Zn,Cu) HDS converts quickly (within 2 min) to Cu-BDC MOF. In addition to ZnO, CuO
and NiO can also react with Zn2+, Cu2+, Ni2+ and Co2+ salts to form HDS.14 However, most
reported methods to synthesize these HDS’s are slow (several days) and may require elevated
temperature. Therefore these HDS are not desirable intermediates for rapid room temperature
MOF synthesis. Important factors for analogous schemes will also include the solubility of
130
organic linkers and the mobility of linkers in HDS lattices.
The rapid room-temperature synthesis method is not limited to forming bulk MOF
powder, but also applicable for MOF patterns and thin film coatings. Figure 3a briefly
describes the fabrication procedure for patterning HKUST-1. We prepared SU-8 (negative
photoresist) patterns on the ALD ZnO surface using photolithography, and then exposed the
patterned ZnO surface to Cu(NO3)2 and H3BTC solutions sequentially. SEM and EDX
mapping images (Figure 4.3b-e) of a star-shape pattern reveal the selective HKUST-1 growth
on ZnO surface. More optical micrographs of the MOF patterns are shown in Figure S4.10.
Table 4.1 Routes for HDS-Driven Room-Temperature Synthesis of Various MOF
Materials.
Metal Oxide Metal Salt HDS Organic Linker MOF HDS-to-MOF Time (min)
ZnO Cu(NO3)2 (Zn,Cu) HDS H3BTC HKUST-1 [Cu3(BTC)2] 1

ZnO Cu(NO3)2 (Zn,Cu) HDS H2BDC Cu-BDC [Cu(BDC)] 2
ZnO Zn(OAc)2 (Zn,Zn) HDS 2-mIM ZIF-8 [Zn(mIM)2] 10
ZnO Zn(OAc)2 (Zn,Zn) HDS H2BDC-NH2 IRMOF-3 [Zn4O(BDC-NH2)3] 10
131
Figure 4.4 (a) Schematic of the rapid room-temperature synthesis route for MOF coatings
onto various form factors. (b-d) SEM images of HKUST-1 deposited onto PS spheres, silicon
wafer and PAN nanofibers, respectively.
Figure 4.4a illustrates the general approach to grow HKUST-1 thin films onto various
form factors. Taking advantage of ALD,22 we can deposit conformal ZnO thin films with
controlled thickness on substrate materials with varied morphologies, such as polystyrene
(PS) spheres, silicon wafers and polyacrylonitrile (PAN) fibers. Within 1 minute of exposure
to the HKUST-1 mother solution, dense coatings of HKUST-1 are also obtained on the
abovementioned ZnO-coated substrates (Figure 4.4b-d). Note that the substrate morphology
is maintained with the conformal ALD ZnO thin films (200 cycles, ~36 nm). The coatings of
densely packed crystals shown in Figure 4.3b-d therefore solely represent the HKUST-1
MOF. More SEM and TEM images for MOF coated PS spheres, polypropylene (PP)
microfibers and PAN nanofibers are shown in Figure S4.11-4.12.
132
Previously, we showed that ALD Al2O3 promotes HKUST-1 solvothermal synthesis and
layer-by-layer growth.23 Here, using HDS formed from ALD ZnO thin films, we reduced the
synthesis time to the scale of minutes. This is also the first example showing fast fabrication
of MOF-functionalized fibrous materials at room temperature. By avoiding long exposure to
heated organic solvents, we expect this synthesis route will enable more MOF-based
composite structures, especially for delicate substrate materials that degrade at high
temperature.
To demonstrate the functionality of these MOF-integrated materials, we characterized
and compared the adsorption capacity of MOF-coated PP microfibers and PAN nanofibers
(referred to as MOF-PP and MOF-PAN, respectively). BET analysis based on N2 isotherm
(Figure 4.5a) reveals the overall surface area is 201 m2/gMOF+Fiber and 524 m2/gMOF+Fiber for
MOF-PP and MOF-PAN, respectively. The larger external surface area on nanofibers enables
higher MOF mass loading, leading to a larger overall BET surface area.
Figure 4.5 (a) BET surface area and (b) NH3 dynamic loading for untreated PP microfibers
and PAN nanofibers, and MOF-coated PP and PAN fibers (MOF-PP, MOF-PAN,
respectively).
133
Breakthrough tests were also performed to quantify the performance for hazardous gas
removal. Figure 4.5b compares the NH3 dynamic loadings on PP and PAN fiber mats with
and without MOF coatings. Untreated PP and PAN fibers can barely retain the NH3 challenge
gas, while MOF-PP and MOF-PAN exhibit 36 and 18 higher dynamic loadings towards
ammonia than the corresponding untreated fibers. In addition to NH3, our
MOF-functionalized fibers also show high adsorption capacity for H2S (Figure S4.14),
another highly toxic industrial chemical. These results all suggest the fibrous materials with
MOF coatings are promising for gas filtration and protective garments.
4.3 Conclusion
In summary, we have discovered an ultra-fast room-temperature MOF synthesis strategy
using hydroxy double salt intermediates. (Zn,Cu) hydroxy nitrate HDS formed in situ from
ZnO shows a high rate of anion exchange in the linker solution and drives the rapid
formation of HKUST-1. The space-time-yield for HKUST-1 reaches 3.6×104 kg∙m-3∙d-1,
more than one order of magnitude higher than previous published reports.11-12 Similar
synthetic strategy has been applied to Cu-BDC, ZIF-8 and IRMOF-3, demonstrating the
synthetic generality. Using ALD ZnO thin films, we have successfully obtained HKUST-1
patterns and dense MOF coatings on PS microspheres, silicon wafers, and PP and PAN fiber
mats in a fast processing rate at room temperature. The HDS-driven MOF synthesis approach
reported here is expected to dramatically improve MOF production rates and widely expand
the material set of MOF-functionalized composites.
134
References
(1) Furukawa, H.; Cordova, K. E.; O’Keeffe, M.; Yaghi, O. M. Science 2013, 341, 1230444.
(2) Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O’Keeffe, M.; Yaghi, O. M.
Science 2002, 295, 469.
(3) (a) Sadakiyo, M.; Yamada, T.; Kitagawa, H. J. Am. Chem. Soc. 2014, 136, 13166; (b)
Brozek, C. K.; Dincă, M. Chem. Soc. Rev. 2014, 43, 5456.
(4) (a) Britt, D.; Tranchemontagne, D.; Yaghi, O. M. Proc. Natl. Acad. Sci. U. S. A. 2008, 105,
11623; (b) Li, J. -R.; Kuppler, R. J.; Zhou, H. -C. Chem. Soc. Rev. 2009, 38, 1477; (c) Li, K.;
Olson, D. H.; Seidel, J.; Emge, T. J.; Gong, H.; Zeng, H.; Li, J. J. Am. Chem. Soc. 2009, 131,
10368.
(5) (a) Lee, J.; Farha, O. K.; Roberts, J.; Scheidt, K. A.; Nguyen, S. T.; Hupp, J. T. Chem. Soc.
Rev. 2009, 38, 1450; (b) Wang, C.; Xie, Z.; deKrafft, K. E.; Lin, W. J. Am. Chem. Soc. 2011,
133, 13445.
(6) Kreno, L. E.; Leong, K., Farha, O. K.; Allendorf, M.; Van Duyne, R. P.; Hupp, J. T. Chem.
Rev. 2012, 112, 1105.
(7) (a) Meek, S. T.; Greathouse, J. A.; Allendorf, M. D. Adv. Mater. 2011, 23, 249; (b)
Horcajada, P.; Gref, R.; Baati, T.; Allan, P. K.; Maurin, G.; Couvreur, P.; Ferey, G.; Morris,
R.E.; Serre, C. Chem. Rev. 2012, 112, 1232.
(8) Stock, N.; Biswas, S. Chem. Rev. 2011, 112, 933.
(9) (a) Beldon, P. J.; Fábián, L.; Stein, R. S.; Thirumurugan, A.; Cheetham, A. K.; Friščić, T.
Angew. Chem. Int. Ed. 2010, 49, 9640; (b) Klimakow, M.; Klobes, P.; Thünemann, A. F.;
Rademann, K.; and Emmerling, F. Chem. Mater. 2010, 22, 5216.
(10) (a) Khan, N. A.; Jhung, S. H. Bull. Korean Chem. Soc. 2009, 30, 2921; (b) Zou, F.; Yu,
R.; Li, R.; Li, W. ChemPhysChem 2013, 14, 2825.
(11) (a) Müller, U.; Pütter, H.; Hesse, M.; Wessel, H.; Schubert, M.; Huff, J.; Guzmann, M.
International Patent WO/2005/049892, 2005; (b) Müller, U.; Schubert, M.; Teich, F.; Pütter,
H.; Schierle-Arndt, K.; Pastré, J. J. Mater. Chem. 2006, 16, 626.
(12) Majano, G.; Pérez-Ramírez, J. Adv. Mater. 2013, 25, 1052.
(13) Zanchetta, E.; Malfatti, L.; Ricco, R.; Styles, M. J.; Lisi, F.; Coghlan, C. J.; Doonan, C.
J.; Hill, A. J.; Brusatin, G.; Falcaro, P. Chem. Mater. 2015, 27, 690.
135
(14) Meyn, M.; Beneke, K.; Lagaly, G. Inorg. Chem. 1993, 32, 1209.
(15) Hara, T.; Ishikawa, M.; Sawada, J.; Ichikuni, N.; Shimazu, S. Green Chem. 2009, 11,
2034.
(16) (a) Bull, R. M. R.; Markland, C.; Williams, G. R; O’Hare, D. J. Mater. Chem. 2011, 21,
1822. (b) Taj, S. F.; Singer, R.; Nazir, T.; Williams, G. R. Rsc Adv. 2013, 3, 358.
(17) Chui, S. S.-Y.; Lo, S. M.-F.; Charmant, J. P. H.; Orpen, A. G.; Williams, I. D. Science
1999, 283, 1148.
(18) Gotthardt, M. A.; Schoch, R.; Wolf, S.; Bauer, M.; Kleist, W. Dalton Trans. 2015, 44,
2052.
(19) (a) Peterson, G. W.; Wagner, G. W.; Balboa, A.; Mahle, J.; Sewell, T.; Karwacki, C. J. J.
Phys. Chem. C 2009, 113, 13906; (b) Wong-Foy, A. G.; Matzger, A. J.; Yaghi, O. M. J. Am.
Chem. Soc. 2006, 128, 3494.
(20) Secco, E. A.; Worth, G. G. Can. J. Chem. 1987, 65, 2504.
(21) Morioka, H.; Tagaya, H.; Karasu, M.; Kadokawa, J.; Chiba, K. Inorg. Chem. 1999, 38,
4211.
(22) Parsons, G. N.; George, S. M.; Knez, M. MRS Bull. 2011, 36, 865.
(23) (a) Zhao, J.; Losego, M. D.; Lemaire, P. C.; Williams, P. S.; Gong, B.; Atanasov, S. E.;
Blevins, T. M.; Oldham, C. J.; Walls, H. J.; Shepherd, S. D.; Browe, M. A.; Peterson, G. W.;
Parsons, G. N. Adv. Mater. Interfaces 2014, 1, 1400040; (b) Zhao, J.; Gong, B.; Nunn, W. T.;
Lemaire, P. C.; Stevens, E. C.; Sidi, F. I.; Williams, P. S.; Oldham, C. J.; Walls, H. J.;
Shepherd, S. D.; Browe, M. A.; Peterson, G. W.; Losego, M. D.; Parsons, G. N. J. Mater.
Chem. A 2015, 3, 1458.
136
Experimental Section
Synthesis of HKUST-1 MOF Using (Zn,Cu) Hydroxy Double Salt Intermediate
0.293 g of ZnO powder (≥99.0%, Sigma-Aldrich) was dispersed in 8 mL of deionized
water using sonication for 5~10 min to form the nanoslurry. 1.74 g of Cu(NO3)2∙3H2O
(99-104%, Sigma-Aldrich) was dissolved in 8 mL deionized water
(n(ZnO):n(Cu(NO3)2)=1:2), and 0.840g of 1,3,5-benzenetricarboxylic acid (H3BTC, 95%,
Aldrich) was dissolved in 16 mL ethanol (200 proof, Koptec) separately. After the Cu(NO3)2
aqueous solution and H3BTC ethanolic solution were prepared, ZnO nanoslurries were mixed
with 16 mL of DMF (Fisher Scientific), to which the Cu(NO3)2 aqueous solution was added
and next the H3BTC ethanolic solution under magnetic stirring. After 1 min of reaction, the
powder product was immediately filtered with a polypropylene membrane filter (0.45 μm
pore size, Whatman) and washed with ethanol (50 mL, 3 times). The HKUST-1 product was
dried in a fume hood for 30 min and then dried in vacuum oven (~0.01 inHg) at 120℃ for 6
hours. No HKUST-1 formed in the control experiments without ZnO nanoslurries.
Synthesis of Cu-BDC MOF Using (Zn,Cu) Hydroxy Double Salt Intermediate
The synthesis of Cu-BDC MOF form (Zn,Cu) HDS is similar to HKUST-1. 0.293 g of
ZnO powder was dispersed in 6 mL of deionized water using sonication for 5~10 min. 1.74 g
of Cu(NO3)2∙3H2O was dissolved in 6 mL deionized water (n(ZnO):n(Cu(NO3)2)=1:2), and
0.673g of 1,4-benzenedicarboxylic acid (H2BDC, 98%, Aldrich) was dissolved in 25 mL
DMF separately. After the Cu(NO3)2 aqueous solution and H2BDC DMF solution were
137
prepared, ZnO nanoslurries were mixed with 6 mL of DMF, to which the Cu(NO3)2 aqueous
solution was added and next the H2BDC DMF solution under fast magnetic stirring. After 2
min of reaction, the turquoise powder product was immediately filtered and dried in air.
Powder XRD pattern for the Cu-BDC product was measured when the sample was still
damp.
Synthesis of IRMOF-3 using (Zn,Zn) Hydroxy Double Salt Intermediate
(Zn,Zn) hydroxy double salt was synthesized at room temperature using a previously
reported method.1 Specifically, 0.407 g (5 mmol) of ZnO was dispersed in 5 mL of deionized
water, and 1.10 g (5 mmol) of Zn(CH3CO2)2∙2H2O (≥99%, Sigma-Aldrich) was dissolved in
5 mL of deionized water. Zinc acetate aqueous solution was added to the ZnO suspension
under strong magnetic stirring at room temperature. After 24 hours, the mixture converted
into a gel-like viscous fluid as the (Zn,Zn) hydroxy acetate HDS formed. 1.05 g of (Zn,Zn)
HDS suspension was transferred to a beaker, and mixed with 25 mL of 2-aminoterephthalic
acid (0.272 g, 1.5 mmol, ≥99%, Acros Organics) DMF solution at room temperature for 10
min. The product was then filtered and dried in air. Powder XRD pattern for the product was
collected when the sample was still damp.
Synthesis of ZIF-8 using (Zn,Zn) Hydroxy Double Salt Intermediate
The synthesis of ZIF-8 from (Zn,Zn) HDS is similar to IRMOF-3. (Zn,Zn) hydroxy
double salt was synthesized at room temperature by mixing 5 mL of ZnO aqueous suspension
(1 M) with 5 mL of Zn(CH3CO2)2 aqueous solution (1 M) and 5 mL of DMF for 24 h. 3 mL
138
of (Zn,Zn) HDS suspension was added to 9 mL of 2-methylimidazole (0.493 g, 6 mmol, 99%,
Aldrich) DMF solution under fast magnetic stirring at room temperature for 10 min. The
product was then filtered and dried in air. Powder XRD pattern for the product was collected
when the sample was still damp.
Atomic Layer Deposition (ALD) for ZnO Thin Films
ALD ZnO thin films were deposited onto polystyrene microspheres, silicon wafers,
polypropylene and polyacrylonitrile fiber mats using a homemade hot-wall viscous-flow
ALD reactor that has been described in our previous work.2 The deposition pressure was ~2
Torr, and the temperature was 100 ℃. In a typical ALD ZnO cycle, diethyl zinc (DEZ, 95%,
STREM Chemicals) was first dosed to the reactor chamber for 2 s, followed by 60 s of N2
(dried with an Entegris gatekeeper) purge. After DEZ dose and N2 purge, deionized water
was dosed for 2 s, and another 60s of N2 purge completed the ALD cycle.
HKUST-1 Precursor Solutions
0.870 g of Cu(NO3)2∙3H2O was dissolved in 12 mL deionized water (Solution A), and
0.420g of H3BTC was dissolved in 12 mL ethanol (Solution B).
For reaction mechanism studies and MOF patterning, Solution B was mixed with equal
volumes of DMF and deionized H2O, making a H3BTC solution (referred to as Solution B’)
with mixed solvents (DMF:EtOH:H2O=1:1:1).
In the dip coating processes for growing MOFs onto different substrates, 3 mL of
Solution A was first mixed with 3mL DMF and then 3mL of Solution B in a 20 mL
139
scintillation vial (VWR International) under mild magnetic stirring for 1 min. The mixed
precursor solution (Solution M) was then used for growing HKUST-1 coatings on
ZnO-coated substrates.
Fabrication Procedure for HKUST-1 Patterns
Negative photoresist SU-8 2050 (Microchem) was used as received and spun-coated
(3000 rpm for 30 s) onto the ALD ZnO coated Si wafer. After soft baking at 65℃ for 1 min
and 95℃ for 7 min, the wafer was exposed to UV lamp (INTELLI-RAY 400, 60% intensity)
for 5 s. The wafer was then baked at 65 ℃ for 1 min and 95 ℃ for 6 min, and
subsequently dipped into SU-8 developer. The pre-patterned sample was rinsed in IPA and
ethanol, and dried in compressed air. During MOF patterning, the sample was soaked in
Solution A for 1 min, and washed in ethanol for 1 min. The wafer was then transferred to
Solution B’ for 45~60s, followed with gentle ethanol rinse for 5 min. The patterned sample
was finally dried in compressed air.
Synthesis of HKUST-1 Coatings on Polystyrene Spheres
80 μL Polybead® polystyrene (PS) microspheres in aqueous suspension (3 μm diameter,
2.5% solids in water, Polysciences Inc.) were spun-coat onto silicon wafers pretreated with
O2 plasma. The prepared sample was then coated with 200 cycles of ALD ZnO, and then
dipped in Solution M for 1 min. The sample was then slowly lifted out of the MOF precursor
solution, and carefully transferred and soaked in ethanol for 5 min. After ethanol rinsing, the
140
samples were dried in air.
Synthesis of HKUST-1 Coatings on Silicon Wafers
200 cycles of ALD ZnO were deposited onto silicon wafers using the abovementioned
process. After ALD coating, Si wafers were soaked in Solution M for 1 min, and then
carefully transferred to ethanol for 5 min of soaking and finally dried in air.
Synthesis of HKUST-1 Coatings on Fibers
Nonwoven PP microfiber mats were used as received from the Nonwovens Cooperative
Research Center (NCRC) at North Carolina State University. Electrospun PAN nanofibers
were used as received from RTI International. 200 cycles of ALD ZnO were deposited onto
these fiber mats using the process described above. After ALD coating, the PP/ZnO and
PAN/ZnO mats were soaked in the Solution M for 1 min and 5 min respectively, dried in air
for 1 hour, and rinsed in methanol for solvent exchange for 1 day. The MOF coated fibers
were finally dried at room temperature under vacuum.
Space-Time-Yield (STY) Calculation
The space-time-yield (STY, in units of kg·m-3·d-1) was calculated using the following
equation:
𝑚𝑀𝑂𝐹
𝑆𝑇𝑌 = × 1.44 × 106 (1)
𝑉𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛 ∙ 𝜏
where mMOF is the dry mass (g) for the MOF powder obtained from the rapid synthesis,
141
Vsolution is the total volume (cm3) for the mixed precursor solution, and 𝜏 is the residence
time (min).
Material Characterization
Scanning electron microscopic (SEM) images were taken using a JEOL JSM 6010 SEM
and an FEI Verios 460L field emission SEM. A thin layer of Au-Pd (5~10 nm) was
sputter-coated onto all samples before SEM imaging. Energy dispersive X-ray (EDX)
analysis for HKUST-1 powder was performed using an FEI Verios 460L field emission SEM
equipped with an Oxford energy dispersive X-ray spectrometer. The cross sections of the
MOF-coated silicon wafers were prepared with an FEI Quanta 3D FEG focused ion beam
(FIB), and imaged using an FEI Titan 80-300 probe aberration corrected scanning
transmission electron microscope (STEM). High-resolution EDX was analyzed using the
SuperX Energy Dispersive Spectrometry (SuperX EDS) system installed on the FEI Titan
STEM. Time-of-flight secondary ion mass spectroscopy (TOF-SIMS) analysis for HKUST-1
powder was performed using a TOF-SIMS V Instrument (ION TOF, Inc. Chestnut Ridge,
NY). Perkin Elmer 8000 Inductively-coupled plasma-optical emission spectroscopy
(ICP-OES) was used for analyzing the concentration of Cu and Zn in the MOF powder and
the residue solution collected after filtration. Before ICP-OES analysis, the MOF powder was
digested in a mixture of 5 mL HNO3, 3 mL HCl and 0.5 mL H2O2 for 1 hour using a CEM
Mars 5 microwave digesting system.3 X-ray diffraction (XRD) was conducted with a Rigaku
SmartLab X-ray diffraction tool (Cu Kα X-ray source) for crystalline phase analysis. MOF
powder diffraction patterns were also simulated using Mercury 3.0 software and the
142
crystallographic information files from Cambridge Crystallographic Data Centre (CCDC
112954 for HKUST-1, CCDC 687690 for Cu-BDC, CCDC 175574 for IRMOF-3, and CCDC
602542 for ZIF-8). A Quantachrome Autosorb-1C surface area and pore size analyzer was
used for measuring N2 isotherm at 77K. Samples were dried in vacuum (~1 × 10−5 𝑇𝑜𝑟𝑟) at
room temperature for 12h before N2 adsorption measurement. BET surface area was
calculated based on the N2 adsorption data within a relative pressure range of P/Po = 0.02 ~
0.10.4 A Thermo Scientific Nicolet 6700 Fourier transform infrared spectrometer was used
for analyzing MOF growth on IR silicon wafers. Dynamic light scattering (Zetasizer Nano,
Malvern Instruments Ltd) was used to measure the size distribution of ZnO nanoslurries.
Breakthrough Test for NH3 and H2S
The adsorption performance of our MOF-functionalized fiber mats was characterized
with a custom-built rapid, micro-breakthrough system.1 Challenge gas (NH3 or H2S in
moisturized air, 1000 mg/m3 concentration, 50% relative humidity) was injected into an
adsorbent column loaded with MOF-fiber material (~20 mg). The column temperature was
kept at 20°C in a water bath. The downstream concentration was analyzed with a
continuously measuring gas chromatograph (HP5890 Series II) equipped with a
photoionization detector for NH3 (or a flame photometric detector for H2S).5
Breakthrough Data Analysis
Dynamic loading (DL in units of mol/kg) of challenge gas on MOF-functionalized fibers
was calculated using the following equations.5
143
𝐶𝑓𝑒𝑒𝑑 𝐹𝑓𝑒𝑒𝑑 ∙ 𝑡𝑡𝑜𝑡𝑎𝑙
𝑀𝑤
𝑡𝑡𝑜𝑡𝑎𝑙 𝐶
𝑜𝑢𝑡 𝐹𝑓𝑒𝑒𝑑
𝑁𝑜𝑢𝑡 = ∫ 𝑑𝑡 (3)
0 𝑀𝑤
𝑁𝑓𝑒𝑒𝑑 − 𝑁𝑜𝑢𝑡
𝐷𝐿 = (4)
𝑚𝑎𝑑𝑠
Nfeed (mol) is the total moles of challenge gas injected into the adsorbent column, Nout (mol) is
the total moles of challenge gas detected in the downstream. Cfeed and Cout (g/m3) are the
concentrations of challenge gas in the feed and the downstream respectively. Ffeed (m3/min) is
the feed flow rate, t (min) is test time. Mw (g/mol) is the molecular weight of the challenge
gas, and mads (kg) is the adsorbent mass.
144
Figure S4.1 Size distribution of ZnO nanoslurries dispersed in deionized water measured
by dynamic light scattering (DLS). The average particle size is 252±23 nm, and the
polydispersity index is 0.732.
145
Figure S4.2 (a-b) SEM images of HKUST-1 crystals obtained from the rapid
room-temperature synthesis. (c) Crystal size distribution analyzed from SEM images (100
measured data points). The average crystal size is 1.17±0.40 μm.
146
Figure S4.3 (a) SEM image of HKUST-1 crystals prepared via rapid synthesis (dispersed
on a silicon wafer). (b-d) Energy dispersive X-ray (EDX) mapping images for C, O and Cu
in the HKUST-1 crystals. (e) EDX spectrum showing that no Zn can be detected by EDX
(blue dash lines indicate where Zn can be expected).
147
Figure S4.4 (a-c) ToF-SIMS surface mapping (negative ion mode) images for a layer of
densely packed HKUST-1 crystals. These results confirm the presence of (a) oxygen, (b)
carbon and (c) copper. (d) ToF-SIMS depth profile (positive ion mode) for C+, Cu+ and 65Cu+
in a layer of densely packed HKUST-1 crystals. Zn-containing residue was not found by
either positive or negative detection mode.
148
Table S4.1 ICP-OES for the concentration of Cu and Zn in the HKUST-1 crystals prepared
through the rapid room-temperature MOF synthesis.
Reaction Cu Concentration* Zn Concentration n(Zn):n(Cu)

n(ZnO):n(Cu(NO3)2) Time (min) in MOF (wt%) in MOF (wt%) in MOF
1:5 1 20.9 0.034 1:628
1:3 1 28.3 0.43 1:67
2:5 1 24.2 0.40 1:62
1:2 1 29.2 0.67 1:45
*The difference of Cu concentration in the MOF is probably a result of solvent inclusion.
Higher Cu concentration is observed for the samples completely activated after vacuum
drying in a BET tube.
Table S4.2 ICP-OES for the concentration of Cu and Zn in the filtrate after collecting the
MOF powder by filtration.
Reaction Cu Concentration Zn Concentration n(Zn):n(Cu)

n(ZnO):n(Cu(NO3)2)
Time (min) in filtrate* (wt%) in filtrate* (wt%) in filtrate
1:3 1 14.8 6.60 1:2.3
1:2 1 11.3 9.71 1:1.2
*The solvent was further evaporated from the filtrate. While the actual concentration of Cu2+
and Zn2+ both increased during solvent removal, the ratio (n(Zn):n(Cu)) did not change in the
concentrated filtrate assuming Cu2+ and Zn2+ are not volatile.
ICP-OES analysis for the filtrate collected after synthesis shows that Zn2+ concentration
in the filtrate is at least one order of magnitude greater than that in the MOF powder, and that
n(Zn):n(Cu) in the filtrate is larger than the initial molar ratio in the mixed reactants. These
results can be explained by the conversion of ZnO to (Zn,Cu) HDS and the release of
Zn-containing species into the solution during the HDS-to-MOF conversion.
149
Figure S4.5 N2 adsorption and desorption isotherm for the HKUST-1 powder prepared via
rapid synthesis.
150
Figure S4.6 SEM images for HKUST-1 grown on top of (Zn,Cu) hydroxy double salt.
151
Figure S4.7 Powder XRD pattern for the Cu-BDC MOF converted from (Zn,Cu) hydroxy
nitrate HDS at room temperature (red). The simulated Cu-BDC pattern is shown in black.
The XRD pattern of the Cu-BDC obtained from HDS agrees well with the simulated pattern
and reported powder XRD patterns for this MOF.6,7 The results from the synthesis of
HKUST-1 and Cu-BDC from (Zn,Cu) HDS indicate that (Zn,Cu) hydroxy nitrate is an
intermediate that preferentially converts to Cu-based MOFs.
152
Intensity (a.u.)
(Zn,Zn) hydroxy acetate HDS
IRMOF-3 converted from (Zn,Zn) HDS
(731)
(200)
(220)
Simulated IRMOF-3 Pattern

(440)
(420)
(600)(531)
(751)
(533)
(444)
(711)
(642)
(111)
(311)
(222)
(620)
(733)
(753)
(400)
(511)
5 10 15 20 25 30 35 40 45
2degree)
Figure S4.8 Powder XRD patterns for the (Zn,Zn) HDS (blue) and IRMOF-3 converted
from (Zn,Zn) HDS at room temperature (red). The simulated IRMOF-3 pattern is shown in
black.
153
Intensity (a.u.) (Zn,Zn) Hydroxy Acetate HDS
(211) ZIF-8 Converted from (Zn,Zn) HDS

(110)
(310)
(222)
(332)
Simulated ZIF-8 Pattern
(411)
(321)
(220)
(200)
5 10 15 20 25 30 35 40 45
2(degree)
Figure S4.9 Powder XRD patterns for the (Zn,Zn) hydroxyl acetate HDS synthesized with
DMF (blue) and ZIF-8 converted from (Zn,Zn) HDS (red). The simulated ZIF-8 pattern is
shown in black.
154
Figure S4.10. (a) Photo of circular and star-shape HKUST-1 patterns. The red box in (a)
represents the location of Image (b). (b-d) Optical micrograph and SEM images of star
patterns made from HKUST-1. The red box in (c) shows the location of (d).
155
Figure S4.11 SEM images for (a) untreated PS microspheres, (b) PS microspheres with
ALD ZnO coating, and (c-d) HKUST-1 grown on ZnO-coated PS microspheres.
156
Figure S4.12 (a-c) SEM images for ZnO-coated PP fibers (a) before and (b-c) after
HKUST-1 rapid synthesis. Insert photo in (b) shows the macroscopic uniformity of MOF
growth on the PP fiber mat. (d) Cross-sectional TEM image shows uniform HKUST-1
coating on ZnO-coated PP fibers. (e-f), SEM images for ZnO-coated PAN nanofibers (e)
before and (f) after HKUST-1 rapid synthesis. Insert optical image in (f) shows the uniform
MOF growth on the PAN nanofiber mat.
157
Figure S4.13 (a) XRD patterns for ALD ZnO coated polypropylene fiber mat (PP/ZnO, red)
and HKUST-1 grown on PP/ZnO (MOF-PP, blue). (b) XRD patterns for ALD ZnO coated
PAN nanofiber mat (PAN/ZnO, red) and HKUST-1 grown on PAN/ZnO (MOF-PAN, blue)
158
Figure S4.14 (a) NH3 breakthrough curves for untreated PP and MOF-PP fiber mats. (b)
H2S breakthrough curves for untreated PP and MOF-PP fiber mats. (c) NH3 breakthrough
curves for untreated PAN and MOF-PAN fiber mats.
(1) Morioka, H.; Tagaya, H.; Karasu, M.; Kadokawa, J.; Chiba, K. Inorg. Chem.
1999, 38, 4211.
S.; Oldham, C. J.; Walls, H. J.; Shepherd, S. D.; Browe, M. A.; Peterson, G. W.; Losego,
M. D.; Parsons, G. N. J. Mater. Chem. A 2015, 3, 1458.
(3) Gotthardt, M. A.; Schoch, R.; Wolf, S.; Bauer, M.; Kleist, W. Dalton Trans. 2015, 44,
2052.
(4) Wong-Foy, A. G.; Matzger, A. J.; Yaghi, O. M. J. Am. Chem. Soc. 2006, 128, 3494.
(5) Glover, T. G.; Peterson, G. W.; Schindler, B. J.; Britt, D.; Yaghi, O. M. Chem. Eng. Sci.
2011, 66, 163.
(6) Carson, C. G.; Hardcastle, K.; Schwartz, J.; Liu, X.; Hoffmann, C.; Gerhardt, R. A.;
Tannenbaum, R. Eur. J. Inorg. Chem. 2009, 2009, 2338.
(7) Rodenas, T.; Luz, I.; Prieto, G.; Seoane, B.; Miro, H.; Corma, A.; Kapteijn, F.; Llabrés i
Xamena, F. X.; Gascon, J. Nat. Mater. 2015, 14, 48.
159
CHAPTER 5. Conformal Metal-Organic Framework Thin Films
on Nanofibers for Ultra-Fast Degradation of Chemical Warfare
Agents
Junjie Zhao†, Dennis T. Lee†, Robert W. Yaga‡, Morgan G. Hall#,
Ian R. Woodward†, Christopher J. Oldham†, Howard J. Walls‡,
Gregory W. Peterson#,* and Gregory N. Parsons†,*
†

‡
#
MD 21010, U.S.A. (E-mail: gregory.w.peterson.civ@mail.mil)
160
Abstract
The threat associated with chemical warfare agents (CWAs) motivates the development
of novel materials to provide enhanced protection in a reduced burden. Metal-organic
frameworks (MOFs) have recently been shown as highly effective catalysts for detoxifying
CWAs, but challenges still remain for integrating MOFs into functional filter media and/or
protective garments. Here, we report a series of MOF-based nanofiber composites for fast
degradation of CWAs. We found TiO2 coatings deposited via atomic layer deposition (ALD)
onto polyamide-6 nanofibers enable the formation of conformal Zr-based MOF thin films
including UiO-66, UiO-66-NH2 and UiO-67. Cross-sectional TEM images show that these
MOF crystals nucleate and grow directly on and around the nanofibers, with strong
attachment to the substrates. These MOF-functionalized nanofibers exhibit excellent
reactivity for detoxifying CWAs. The half-lives of a CWA simulant compound and nerve
agent soman (GD) are as short as 7.3 min and 2.3 min, respectively. These results therefore
provide the earliest report of MOF/nanofiber textile composites capable of ultra-fast
degradation of CWAs.
5.1 Introducion
Chemical warfare agents (CWAs) are highly toxic compounds that can injure,
incapacitate or even kill human beings. Detoxification of chemical warfare agents (CWAs)
has been of great social significance due to the past accidental or deliberate emissions and
remaining threat posed to civilian and military personnel.1,2 Materials that can efficiently
degrade these lethal chemicals are therefore highly desired to protect soldiers,
161
first-responders and the general public. Activated carbon loaded with various impregnants
including metal salts, amines and acids are currently used for broad spectrum protection.3
However, the complementary functionalities of impregnants can react with each other within
the carbon pores, leading to a decay of performance over time.1 In this regard, new sorbent
materials are of great interest to provide prolonged period of protection in a reduced volume
and weight.
Metal-organic frameworks (MOFs) are a class of crystalline networks consisting
metal-containing building blocks and organic linking units. Their ultra-high surface area,
large porosity and amenability to design and tailor internal chemical functionalities make
MOFs promising candidates for sorption and detoxification of CWAs.4 The high
concentration of metal ion/clusters in the MOF structures also provides numerous
Lewis-acidic sites that mimic the active moieties in enzymes to promote CWA destruction.5
Early studies were focused on the adsorption and catalytic reaction toward CWA and
simulants using Cu-BTC [Cu3(BTC)2] (also referred to as HKUST-1).6–8 While Cu-BTC
shows reasonable removal rate toward sulfur mustard (HD), sarin (GB) and their simulants,
the high water affinity and instability in ambient moisture affect its long-term use.6–8 Later, a
few hydrophobic or water-stable MOFs have been developed and assessed for their catalytic
activity for CWA destruction. Navarro et al. developed several hydrophobic MOFs including
[Zn4O(3,5-dimethyl-4-carboxypyrazolato)3] and [Ni8(OH)4(H2O)2(L)6]n (L =
pyrazolate-based ligands), and evaluated their adsorption of volatile organic compounds.6,9
Farha, Hupp and their co-workers have reported a series of stable MOFs containing
[Zr6O4(OH)4] clusters for fast degradation of soman (GD) and VX and simulants including
162
dimethyl 4-nitrophenyl phosphate (DMNP) and 2-chloroethyl ethyl sulfide (CEES) with
half-lives as short as 0.5 min. These MOFs include UiO-66(67) and their derivatives,
NU-1000, MOF-808 and PCN-222/MOF-545.4,5,10–13
While bulk MOF powders exhibit exceptional catalytic properties for CWA destruction,
many practical issues still need to be addressed before these materials can be widely applied.
For example, the catalytic reactions to degrade CWAs were observed to occur mainly on the
external surface of MOF crystals,11 indicating that the diffusion of CWAs is limited into the
MOF pores. Meanwhile, particles tend to aggregate and consequently lose many accessible
catalytic sites. Thus, it is a challenge to minimize the amount of MOF powder while
maintaining substantial long-term protection. Simplifying the handling and deployment is
also an important issue for these MOF catalysts.
Polymeric nanofibers, often acquired from electrospinning, can exhibit very high
external surface area, excellent water vapor transport property and good mechanical
strength.14,15 Therefore nanofibers are promising scaffolds for MOF-functionalized
composites designed for CWA protection. As far as we know, only one report about
MOF-textile composites for catalytic destruction of CWA simulants has appeared in
literature, where UiO-66 particles are physically attached to silk microfibers.16 Conformal,
high-quality MOF thin films covalently bound to nanofibers is yet to be explored for CWA
destruction application. In this paper, we report a series of novel MOF-nanofiber structures
for this application. Half-lives of nerve agent soman (O-Pinacolyl
Methylphosphonofluoridate, also known as GD) are as short as 2.3 min using our
MOF-functionalized nanofiber catalysts. This is also the first demonstration to detoxify real
163
CWA compound with MOF-fabric composites, as previous work has only investigated
simulants.16
Figure 5.1 describes the procedure to synthesize our MOF-nanofiber composites.
Free-standing polyamide-6 (PA-6) nanofiber mat obtained from electrospinning was coated
with a conformal thin layer (~ 5 nm thick) of TiO2 using atomic layer deposition (ALD). This
ALD TiO2 layer is expected to promote MOF heterogeneous nucleation on fibers,17 and
provide some contribution to catalytic CWA detoxification.18,19 Three Zr-based MOFs
including UiO-66, UiO-66-NH2 and UiO-67 (crystal structures shown in the dashed box in
Figure 5.1) were chosen because previous reports has demonstrated the excellent stability
and good catalytic properties of these MOFs.5,10,11,20 We used concentrated HCl as the
modulator for the solvothermal synthesis which is similar to reported recipes for these
MOFs,21 but developed our own conditions to achieve optimized growth of these Zr-based
MOF thin films onto the nanofibers. Specifically, 0.343 mmol of ZrCl4 and 0.343 mmol of
dicarboxylic acid linkers (1:1 molar ratio) were dissolved in 20 mL of DMF. 25 μL of
deionized water and concentrated HCl (1.33 mL for UiO-66 and UiO-66-NH2, 0.67 mL for
UiO-67 respectively) were added to the solution. Subsequently, TiO2-coated PA-6 nanofiber
(PA-6@TiO2) mat was transferred into the mixed solution, which was then heated to 85 °C
for 24 h. MOF-coated nanofiber mat was collected, washed and activated after the
solvothermal synthesis.
164
Figure 5.1 Schematic of the synthetic procedure for Zr-based MOF functional coatings on
polyamide-6 nanofibers. The MOF crystal structures are illustrated in the dashed box.
Figure 5.2a is a photo of a free-standing PA-6@TiO2 nanofiber mat coated with
UiO-66-NH2 MOF thin films (referred to as PA-6@TiO2@UiO-66-NH2). The structural
integrity and flexibility were fully maintained after the solvothermal synthesis (Video S5.1).
SEM images (Figures 5.2b-d) shows that UiO-66-NH2 nanocrystals (average size = 126±25
nm) were grown conformally on the PA-6@TiO2 nanofibers. While we and other groups
have attempted various methods to deposit UiO-66 onto polymeric fibers in the past,16,17,22-24
our results in Figure 5.2 uniquely demonstrate good growth uniformity and high crystal
coverage on the fibers. The challenge of integrating Zr-based MOFs onto nanofibers is
solved using ALD TiO2 nucleation layers and modified HCl-modulated solvothermal
synthesis. In contrast, the growth of UiO-66-NH2 on PA-6 nanofibers without TiO2
165
nucleation layers is patchy and does not result in a conformal thin film on the nanofiber
surface as that on TiO2 coated PA-6 nanofibers using the same synthesis conditions (Figure
S5.1).
Figure 5.2 (a) Photo of a free-standing PA-6@TiO2@UiO-66-NH2 nanofiber mat. (b-d)
SEM images, (e-i) energy dispersive X-ray mapping images.
Energy dispersive X-ray analysis (Figures 5.2e-i) also confirms uniform MOF growth on
the nanofibers. C, O and N signals originate from the polyamide substrate and the
2-aminoterephthalate linker in the MOF structures. The ALD TiO2 thin film deposited onto
the PA-6 nanofibers also contributes to the EDX intensity of O. In addition, Zr and Ti
mapping images (Figures 5.2g-h) agree well with the secondary electron image (Figure 2d),
showing excellent conformality of the UiO-66-NH2 MOF coatings and the ALD TiO2 thin
166
films, respectively.
The uniform growth of UiO-66 and UiO-67 was also achieved on PA-6@TiO2
nanofibers using our solvothermal synthesis (these two MOF-nanofiber composites are
referred as PA-6@TiO2@UiO-66 and PA-6@TiO2@UiO-67, respectively). Figures 5.3a-c are
SEM images of PA@TiO2 nanofiber mat coated with three different Zr-based MOF thin films.
Although the crystal shapes shown in our MOF-nanofiber structure are distinct from the
octahedron morphology often reported for these fcu-type Zr-based MOFs, X-ray diffraction
patterns of the MOF thin films on nanofibers match well with those of MOF powders
(Figures 5.3g-i), confirming the formation of targeted MOF structures. We speculated that the
spherical shapes of these crystals reflect the rapid kinetics for growing the MOF coatings
because slow reactions could more readily allow formation of crystal facets.21
Figures 5.3d-h are cross-sectional TEM images of the MOF-coated PA@TiO2 nanofibers.
Tubular features observed in these images represent the core@shell structures of PA@TiO2
nanofibers sliced along the axial direction. The diameters of PA-6 nanofibers measured from
TEM images range from 15 nm to 55 nm, consistent with the average fiber diameter (37±16
nm) measured from SEM images (Figure S5.2). The ALD TiO2 thin films deposited onto the
PA-6 nanofibers is conformal and uniform as shown in Figures 5.3d-h. The average thickness
of the TiO2 coatings is 5.7±1.3 nm, corresponding to an ALD growth rate of ~ 0.6 Å per
cycle. The spherical MOF crystals are found to nucleate and grow directly on or around the
PA-6@TiO2 nanofibers, indicating formation of strong covalent binding to the substrates. No
noticeable particle shedding during the handling after synthesis also confirms good adhesion
of our MOF thin films to the nanofibers.
167
The quality of the Zr-based MOF thin films grown on nanofibers were characterized
using XRD and BET. The sharp XRD peaks associated with the targeted MOFs (Figures
5.3g-i) demonstrate the good crystallinity of our MOF coatings. BET surface area analysis
was performed on the MOF powders collected from the solvothermal synthesis as well as on
the MOF-nanofiber composites. All samples were washed and activated, and characterized
with N2 adsorption/desorption measurement (Figure 5.3j-l). BET surface area of the
UiO-type MOF powders (Table 5.1) is consistent reported values for all the three MOFs. The
large surface area is a result of proper solvent exchange and activation processes to fully
remove solvent and other guest molecules within the pores of the MOFs.
The BET surface area of PA-6@TiO2 nanofibers coated with UiO-66, UiO-66-NH2 and
UiO-67 is also given in Table 5.1. The surface area for the MOF-nanofiber composites is in
excess of 10 times larger than the nanofiber substrates alone, demonstrating the high porosity
of the MOF coatings on the nanofibers. We find that analyzing the MOF mass fraction in the
composites is challenging with weighing methods. The net mass increase due to MOF
loading is relatively small and often comparable to the expected mass change due to water
uptake by the hygroscopic nylon nanofibers. However, assuming surface area of the MOF
component in the MOF-nanofiber system is equivalent to that of the simultaneously
produced MOF powder, we estimate the MOF mass loading using the calculation methods
described previously.17 The calculated MOF mass fraction is 8.8%, 14.7%, and 15.4% for
PA-6@TiO2@UiO-66, PA-6@TiO2@UiO-66-NH2 and PA-6@TiO2@UiO-67, respectively.
These results were further confirmed by elemental analysis using inductively coupled plasma
optical emission spectroscopy (ICP-OES) (Table S5.1).
168
Figure 5.3 (a-c) SEM images of (a) PA-6@TiO2@UiO-66, (b) PA-6@TiO2@ UiO-66-NH2
and (c) PA-6@TiO2@UiO-67. (d-f) Cross-sectional TEM images of (d)
PA-6@TiO2@UiO-66, (e) PA-6@TiO2@UiO-66-NH2 and (f) PA-6@TiO2@UiO-67. (g-i)
XRD patterns of PA-6 nanofibers before and after ALD, MOF-coated nanofibers and MOF
powders. (j-l) N2 adsorption and desorption isotherms for PA-6@TiO2 nanofibers with and
without MOF coatings and Zr-based MOF powders.
169
To evaluate the catalytic property of our MOF-nanofiber composites for CWA
destruction, we first analyzed the hydrolysis of simulant 4-nitrophenyl phosphate (DMNP).
Since nerve agents are extremely toxic and requires special caution to handle, investigation
of simulants with phosphate ester bonds is a more accessible approach for almost all
laboratories to test the efficacy of the catalysts and compare the kinetics.4 Here, we
investigated the catalytic degradation of DMNP (Figure 5.4a) in an aqueous buffer solution
of N-ethylmorpholine (0.45M, pH = 10). 2.5 mg of MOF powders or 14 mg of MOF-coated
nanofibers was used the catalysts, and the reaction kinetics was characterized using a
procedure similar to the methods described in previous reports.4,5
Figure 5.4b shows a typical serious of spectra measured during the catalytic hydrolysis
of DMNP. The decreased absorption at 270 nm represents the reduced concentration of
DMNP, and the increase of absorbance at 407 nm corresponds to the formation of
p-nitrophenoxide.5 We monitored the reaction progress by tracking the absorbance at 407 nm,
and calculated the concentration of p-nitrophenoxide based on Lambert-Beer Law (Equation
1). The percent conversion of DMNP during catalytic degradation using MOF powders or
MOF-nanofiber composites is plotted as a function of time in Figures 5.4c-e. For Zr-based
MOF powders, we observed 95%, 98% and 96% DMNP conversion in 60 min of reaction
when UiO-66, UiO-66-NH2 and UiO-67 were used respectively. Half-life (t1/2) of DMNP
with MOF powder catalysts (Table 5.1) shows similar trend to reported data. UiO-66-NH2
exhibits the fastest degradation rate (t1/2 = 2.8 min) among the three MOF powders, while
UiO-67 also significantly reduces the half-life of DMNP compared to UiO-66. We speculate
that the amine moieties in UiO-66-NH2 synergistically enhance the catalytic activity, while
170
the large pore size of UiO-67 may allow further diffusion of DMNP molecules into the
catalyst.
Figure 5.4 (a) Catalytic reaction of DMNP hydrolysis using Zr-based MOF powder and
MOF functionalized nanofiber catalyst. (b) UV/Visible absorption spectra for monitoring
degradation of DMNP. (c-e) Conversion of DMNP to p-nitrophenoxide versus reaction time
using MOF powder and functionalized nanofiber catalyst.
171
Table 5.1 Material properties and catalytic performance towards CWA simulant
degradation.
BET Surface MOF wt% DMNP t1/2 Ref.

2
Area (m /g) (min)
1512 100 22 This work
UiO-66 a
1580 100 24, 45 5, 11
1334 100 2.8 This work
UiO-66-NH2
1200a 100 0.7 11
2250 100 7.7 This work
UiO-67
2500a 100 3.8 11
PA-6@TiO2@UiO-66 143.9 8.8 135 This work
PA-6@TiO2@UiO-66-NH2 205.9 14.7 7.3 This work
PA-6@TiO2@UiO-67 356.2 15.4 7.4 This work
PA-6@TiO2 11.5 0 1170 This work
PA-6 11.5 0 3950 This work
*BET surface area values are adapted from Ref. 21.
We also characterized the catalytic performance of the MOF-functionalized nanofibers
for degrading CWA simulant DMNP. The t1/2 values of DMNP using MOF-nanofiber
composite catalysts and control materials are summarized in Table 5.1. For untreated PA-6
nanofibers, DMNP shows negligible rate of hydrolysis with an estimated t1/2 value over 65 h
(Figure S5.3). With ALD TiO2 coatings, PA-6@TiO2 reduces the half-life to ~20 h, consistent
with the reported catalytic effect of TiO2 for degrading CWAs.18,19 Compared with PA-6 and
PA-6@TiO2, MOF-coated PA-6@TiO2 nanofibers exhibit significantly enhanced catalytic
performance. Both PA-6@TiO2@UiO-66-NH2 and PA-6@TiO2@UiO-67 enable short
half-life of DMNP (< 8 min) and high conversion (> 90%) in 60 min, demonstrating the
172
excellent catalytic properties of these MOF-nanofiber composites. PA-6@TiO2@UiO-66
shows a slower DMNP hydrolysis rate than the other two composite structures, because of
the smaller MOF mass loading and the lower catalytic activity of UiO-66 compared to
UiO-66-NH2 and UiO-67. Notably the DMNP t1/2 values for the PA-6@TiO2@UiO-66-NH2
and PA-6@TiO2@UiO-67 are all within the same magnitude as those for the corresponding
MOF powder catalysts, indicating that the functionality of MOFs are well maintained after
growth on nanofibers. Detailed analyses of the reaction kinetics are shown in Figure
S5.4-S5.5 and Table S5.2. SEM images and EDX spectra taken for the MOF-nanofiber
composites after DMNP degradation (Figure S5.6) show that significant amounts of Zr-based
MOF coatings remain in the composite structures even after strong agitation during the
DMNP hydrolysis tests. These results also demonstrate the good adhesion of our MOF thin
films to the nanofiber substrates.
Figure 5.5 (a) Catalytic reaction of GD hydrolysis using MOF-nanofiber catalyst. (b)
Conversion of GD versus reaction time during catalysis. Dashed lines are fitted results
assuming first order reaction kinetics.
173
In addition to simulant DMNP, we further tested our MOF-nanofiber composites for the
destruction of the highly toxic nerve agent O-pinacolyl methylphosphonofluoridate (GD)
(Figure 5.5a). For this test, 2.6 µL of GD was added to approximately 14 mg of
MOF-nanofiber catalyst in an NMR tube, followed with vigorous shaking in order to mix GD
well with MOF-nanofibers in the solution. The tube was then analyzed for GD degradation
31
and product formation using P NMR. Figure 5b is a plot of the conversion of GD during
catalytic degradation using MOF-functionalized nanofibers. The doublet peaks at
approximately 40 and 33 ppm associated with GD immediately decrease upon exposure to
the MOF-nanofibers (Figures S5.7-S5.9). The pinacolyl methylphosphonic acid (PMPA)
peak (approximately 27 ppm) begins growing at the same time, indicating detoxification of
the CWA to a non-toxic product. The half-lives of GD are 3.0 min (with
PA-6@TiO2@UiO-66), 3.7 min (with PA-6@TiO2@UiO-66-NH2) and 2.3 min (with
PA-6@TiO2@UiO-67), respectively. All three MOF-nanofiber composites show fast GD
destruction (t1/2 ≤ 4 min) and high conversion (≥80%) within 10 min. The fastest reaction
rate with PA-6@TiO2@UiO-67 is possibly because the large pore size of UiO-67 allows
diffusion of reactants into the pores while the catalytic reactions occur mainly on the external
surface of UiO-66 and UiO-66-NH2.11 As far as we know, this is the first demonstration of
MOF-fiber composites for detoxifying real CWA compounds. Our results are also very
promising since as the catalytic performance of these MOF thin films on nanofibers is
comparable or even exceeds the reported bulk MOF powders.4,11 These results clearly reveal
that conformal MOF thin films grown onto nanofiber substrates can achieve excellent
catalytic activity even with small MOF loadings (< 20%). This advantage will eventually
174
benefit the end users by providing substantial catalytic efficacy at a reduced burden.
5.3 Conclusion
In conclusion, we have demonstrated that nanofibers functionalized with Zr-based MOF
thin films are active catalytic materials for decomposing the chemical warfare agent simulant
DMNP and nerve agent GD. ALD TiO2 coatings deposited onto PA-6 nanofibers enabled
direct nucleation and crystal growth of MOFs onto the fiber substrates, leading to conformal
MOF thin films. Cross-sectional TEM images show that these MOF crystals nucleate and
grow directly on or around the nanofibers, indicating strong attachment to the substrates.
XRD and BET confirm that the MOF coatings are highly crystalline with large adsorptive
capacity. We have shown that our MOF-nanofiber composite textiles enable ultra-fast
detoxification of simulant DMNP and nerve agent GD. The half-lives of DMNP and GD are
as short as 7.3 min and 2.3 min, respectively, indicating great promise of our
MOF-functionalized nanofibers for CWA protection. The synthesis method and the
MOF-nanofiber composite structures we have presented here will also offer new
opportunities to advance the development of gas filters, chemical sensors, and potentially
smart textile materials to protect against harmful air pollutants.
175
References
(1) DeCoste, J. B.; Peterson, G. W. Metal–Organic Frameworks for Air Purification of

Toxic Chemicals. Chem. Rev. 2014, 114 (11), 5695–5727.
(2) Enserink, M. U.N. Taps Special Labs to Investigate Syrian Attack. Science 2013, 341
(6150), 1050–1051.
(3) Romero, J. V.; Smith, J. W. H.; Sullivan, B. M.; Mallay, M. G.; Croll, L. M.; Reynolds,
J. A.; Andress, C.; Simon, M.; Dahn, J. R. Gas Adsorption Properties of the Ternary
ZnO/CuO/CuCl2 Impregnated Activated Carbon System for Multigas Respirator
Applications Assessed through Combinatorial Methods and Dynamic Adsorption
Studies. ACS Comb. Sci. 2011, 13 (6), 639–645.
W.; Hall, M. G.; DeCoste, J. B.; Peterson, G. W.; et al. Destruction of chemical
warfare agents using metal–organic frameworks. Nat. Mater. 2015, 14 (5), 512–516.
Hupp, J. T. Simple and Compelling Biomimetic Metal-Organic Framework Catalyst
for the Degradation of Nerve Agent Simulants. Angew. Chem.-Int. Ed. 2014, 53 (2),
497–501.
(6) Montoro, C.; Linares, F.; Quartapelle Procopio, E.; Senkovska, I.; Kaskel, S.; Galli, S.;
Masciocchi, N.; Barea, E.; Navarro, J. A. R. Capture of Nerve Agents and Mustard Gas
Analogues by Hydrophobic Robust MOF-5 Type Metal-Organic Frameworks. J. Am.
Chem. Soc. 2011, 133 (31), 11888–11891.
(7) Roy, A.; Srivastava, A. K.; Singh, B.; Shah, D.; Mahato, T. H.; Gutch, P. K.; Halve, A.
K. Degradation of sarin, DEClP and DECNP over Cu-BTC metal organic framework.
J. Porous Mater. 2013, 20 (5), 1103–1109.
(8) Peterson, G. W.; Wagner, G. W. Detoxification of chemical warfare agents by CuBTC.

J. Porous Mater. 2014, 21 (2), 121–126.
(9) Padial, N. M.; Quartapelle Procopio, E.; Montoro, C.; Lopez, E.; Enrique Oltra, J.;
Colombo, V.; Maspero, A.; Masciocchi, N.; Galli, S.; Senkovska, I.; et al. Highly
Hydrophobic Isoreticular Porous Metal-Organic Frameworks for the Capture of
Harmful Volatile Organic Compounds. Angew. Chem.-Int. Ed. 2013, 52 (32), 8290–
8294.
(10) Moon, S.-Y.; Liu, Y.; Hupp, J. T.; Farha, O. K. Instantaneous Hydrolysis of
Nerve-Agent Simulants with a Six-Connected Zirconium-Based Metal-Organic
176
Framework. Angew. Chem.-Int. Ed. 2015, 54 (23), 6795–6799.
Farha, O. K. Tailoring the Pore Size and Functionality of UiO-Type Metal-Organic
Frameworks for Optimal Nerve Agent Destruction. Inorg. Chem. 2015, 54 (20), 9684–
9686.
(12) Moon, S.-Y.; Wagner, G. W.; Mondloch, J. E.; Peterson, G. W.; DeCoste, J. B.; Hupp, J.
T.; Farha, O. K. Effective, Facile, and Selective Hydrolysis of the Chemical Warfare
Agent VX Using Zr-6-Based Metal-Organic Frameworks. Inorg. Chem. 2015, 54 (22),
10829–10833.
(13) Liu, Y.; Moon, S.-Y.; Hupp, J. T.; Farha, O. K. Dual-Function Metal–Organic
Framework as a Versatile Catalyst for Detoxifying Chemical Warfare Agent Simulants.
ACS Nano 2015, 9 (12), 12358–12364.
(14) Gibson, P.; Schreuder-Gibson, H.; Rivin, D. Transport properties of porous membranes
based on electrospun nanofibers. Colloids Surf. Physicochem. Eng. Asp. 2001, 187–
188, 469–481.
(15) Huang, Z. M.; Zhang, Y. Z.; Kotaki, M.; Ramakrishna, S. A review on polymer
nanofibers by electrospinning and their applications in nanocomposites. Compos. Sci.
Technol. 2003, 63 (15), 2223–2253.
(16) López-Maya, E.; Montoro, C.; Rodríguez-Albelo, L. M.; Aznar Cervantes, S. D.;
Lozano-Pérez, A. A.; Cenís, J. L.; Barea, E.; Navarro, J. A. R. Textile/Metal–
Organic-Framework Composites as Self-Detoxifying Filters for Chemical-Warfare
Agents. Angew. Chem. Int. Ed. 2015, 54 (23), 6790–6794
Atomic Layer Deposition. Adv. Mater. Interfaces 2014, 1 (4), 1400040 DOI:
10.1002/admi.201400040.
(18) Wagner, G. W.; Chen, Q.; Wu, Y. Reactions of VX, GD, and HD with Nanotubular
Titania. J. Phys. Chem. C 2008, 112 (31), 11901–11906.
(19) Wagner, G. W.; Peterson, G. W.; Mahle, J. J. Effect of adsorbed water and surface
hydroxyls on the hydrolysis of VX, GD, and HD on titania materials: the development
of self-decontaminating paints. Ind. Eng. Chem. Res. 2012, 51 (9), 3598–3603.
(20) Cavka, J. H.; Jakobsen, S.; Olsbye, U.; Guillou, N.; Lamberti, C.; Bordiga, S.; Lillerud,
177
K. P. A New Zirconium Inorganic Building Brick Forming Metal Organic Frameworks
with Exceptional Stability. J. Am. Chem. Soc. 2008, 130 (42), 13850–13851.
(21) Katz, M. J.; Brown, Z. J.; Colon, Y. J.; Siu, P. W.; Scheidt, K. A.; Snurr, R. Q.; Hupp, J.
T.; Farha, O. K. A facile synthesis of UiO-66, UiO-67 and their derivatives. Chem.
Commun. 2013, 49 (82), 9449–9451.
(22) Armstrong, M. R.; Arredondo, K. Y. Y.; Liu, C.-Y.; Stevens, J. E.; Mayhob, A.; Shan,
B.; Senthilnathan, S.; Balzer, C. J.; Mu, B. UiO-66 MOF and Poly(vinyl cinnamate)
Nanofiber Composite Membranes Synthesized by a Facile Three-Stage Process. Ind.
Eng. Chem. Res. 2015, 54 (49), 12386–12392.
(23) Ren, J.; Musyoka, N. M.; Annamalai, P.; Langmi, H. W.; North, B. C.; Mathe, M. Int.
J. Hydrog. Energy 2015, 40 (30), 9382–9387.
(24) Zhang, Y.; Yuan, S.; Feng, X.; Li, H.; Zhou, J.; Wang, B. J. Am. Chem. Soc. 2016, 138
(18), 5785–5788.
178
Experimental Section
Synthesis of Zr-based MOFs
UiO-66, UiO-66-NH2 and UiO-67 were synthesized using slightly modified recipes
based on previous reports.1,2
For UiO-66, 0.080 g (0.343 mmol) of ZrCl4 (≥99.5%, Alfa Aesar) was added to 20 mL of
DMF (Fisher) in a glass scintillation vial, followed with 1 min of sonication and 5 min of
magnetic stirring to dissolve ZrCl4 well. 0.057 g (0.0343 mmol) of benzene-1,4-dicarboxylic
acid (H2BDC, 98%, Sigma) was then added to the ZrCl4 solution under stirring. Subsequently,
25 μL of deionized water and 1.33 mL of concentrated hydrochloric acid (NF/FCC grade,
Fisher) were added to the mixed solution. The vial was then placed in a box furnace (Thermo
Scientific) and heated to 85 °C for 24 h. After the solvothermal synthesis, UiO-66 powder
product was filtered using polypropylene membrane (0.45 μm pore size, Whatman) and
washed with DMF twice. Anhydrous ethanol (200 proof, VWR) was used for solvent
exchange, and the solvent was replaced every 24 h for a total of two times. After solvent
exchange, the UiO-66 powder was collected via filtration and dried in a dessicator at room
temperature at reduced pressure for 12 h. To further activate the MOF, ~ 0.3 g of powder
product dried in BET at 80 °C for 12 hour and then at 110 °C for 12~18 h (heating
temperature was slowly ramped from 80 °C to 110 °C).
For UiO-66-NH2, the recipe and synthesis procedure are similar to UiO-66. 0.062 g
(0.343 mmol) of 2-aminoterephthalic acid (H2BDC-NH2, 99%, Sigma) was used as the
organic linker instead of H2BDC. The amounts of ZrCl4, DMF, H2O and HCl were kept same
179
as UiO-66. Reaction temperature was controlled at 85 °C for 24 h. After synthesis, the
washing step, solvent exchange and drying process for UiO-66-NH2 were similar to UiO-66.
For UiO-67, 0.083 g (0.343 mmol) of biphenyl-4,4’-dicarboxylic acid (H2BPDC, 97%,
Sigma) was used as ligand. Same amounts of ZrCl4, DMF, H2O were used as UiO-66. The
addition of HCl modulator was reduced to 0.67 mL for 20 mL of mixed solution. Reaction
proceeded at 85 °C for 24 h. After synthesis, the washing step, solvent exchange and drying
process for UiO-66-NH2 were similar to UiO-66.
Electrospinning of Polyamide-6 Nanofiber Mat
Polyamide-6 (referred to as PA-6, Scientific Polymer Products) was dissolved overnight
in a solvent mixture of glacial acetic acid (Sigma Aldrich) and formic acid (spectroscopy
grade, Fluka) (glacial acetic acid : formic acid = 2 : 1 wt/wt) to reach a final PA-6
concentration of 12 wt%. The polymer solution was then loaded into a syringe with a
27 gauge stainless steel needle attached and placed into a custom-built electrospinning
system described previously.3 An atmosphere of bone dry grade carbon dioxide gas (Airgas)
was mixed with water vapor to form ~50% RH at an ambient temperature of ~22°C. This RH
controlled, CO2 atmosphere provides a stable environment for electrospinning the fibers. The
spinning conditions were a voltage gradient of 1.57 kV/cm, a 25 cm spinning distance
(needle to collection substrate distance), and 90 minutes elapsed spinning time. The polymer
solution was continuously fed through the needle at a rate to maintain a stable
electrospinning jet. PA-6 nanofibers were deposited onto a paper ring (10 cm outside
diameter and 6.5 cm inside diameter) sitting on a 3.5-inch (~8.9 cm) petri dish filled with
180
deionized water. The water provided a collection surface for the forming fibers during
electrospinning. After electrospinning the paper ring with deposited nanofiber mat was
carefully removed and dried overnight at ambient conditions.
Atomic Layer Deposition of TiO2 on PA-6 Nanofibers
A thin film of TiO2 was deposited onto free-standing PA-6 nanofiber mat using atomic
layer deposition (ALD) before growing Zr-based MOF coatings. A custom-built hot-wall
viscous-flow reactor that has been described previously was used for ALD TiO2 processes.4
In a typical ALD TiO2 cycle, titanium (IV) chloride (TiCl4, 99%, STREM Chemicals) was
first dosed to the reactor chamber for 1 s, followed with 40 s of N2 (99.999%, Airgas, further
purified with an Entegris gatekeeper). After TiCl4 dose and N2 purge, deionized H2O was
dosed to the chamber for 1 s, and the chamber was subsequently purged with N2 for 60 s. The
process temperature was controlled at 100°C, and the pressure was about 1.3 Torr. All PA-6
nanofibers were coated with 100 cycles of ALD TiO2 before solvothermal MOF growth.
Synthesis of Zr-based MOF Coatings on Nanofibers
Precursor solutions for Zr-based MOF coatings (i.e. UiO-66, UiO-66-NH2 and UiO-67)
were prepared using the recipe and procedure described above for Zr-based MOF powders.
~10 mg of free standing PA-6 nanofiber mat coated with ALD TiO2 was transferred into a 20
mL scintillation vial and soaked in the MOF precursor solution. The vial was then placed in a
box furnace and heated to 85 °C for 24 h. After the solvothermal synthesis, the MOF-coated
PA-6 nanofiber mat was transferred into a fine aluminum mesh and washed with 100 mL of
181
DMF twice. After DMF wash, the MOF-coated nanofiber mat was further exchanged in
anhydrous ethanol, and the solvent was replaced every 12 h for a total of three times. To fully
activate the MOF coating, the MOF-coated nanofiber mat was first dried in a dessicator at
room temperature at reduced pressure for 12 h, and then in a BET at 80 °C for 12 hour and
subsequently at 110 °C for 12~18 h (heating temperature was slowly increased from 80 °C to
110 °C).
Material Characterization
Scanning electron microscopic (SEM) images were obtained using an FEI Verios
460L field emission SEM. All the samples were sputter-coated with Au-Pd (5~10 nm) before
imaging. Energy dispersive X-ray (EDX) mapping images was taken using an Oxford energy
dispersive X-ray spectrometer attached to the FEI Verios 460L FESEM. Cross sections of the
MOF-coated nanofibers were obtained via microtoming, and imaged using a JEOL 2010F
field emission transmission electron microscope (TEM). X-ray diffraction (XRD) was
performed using a Rigaku SmartLab X-ray diffraction tool (Cu Kα X-ray source) for
crystalline phase analysis. Glass sample holder was used for powder samples, and aluminum
holder was used for nanofiber samples. Powder diffraction patterns for Zr-based MOFs were
also simulated using Mercury 3.0 software based on the crystallographic information files
from Cambridge Crystallographic Data Centre (CCDC 733458 for UiO-66, CCDC 733458
for UiO-67, no CIF available for UiO-66-NH2). BET surface area of MOF powders was
measured with a Micromeritics ASAP 2020 surface area and porosimetry analyzer, and
MOF-coated nanofiber samples were analyzed using a Quantachrome Autosorb-1C surface
182
area and pore size analyzer. Samples were dried in vacuum at 80 °C for 12 hour and
subsequently at 110 °C for 12~18 h (heating temperature was slowly increased from 80 °C to
110 °C) before N2 adsorption measurement. BET surface area was calculated based on the
N2 adsorption data within a relative pressure range of P/Po = 0.02 ~ 0.07 (this range meets
the criteria for determining BET surface area of microporous materials described in
literature).5,6
Degradation of Dimethyl 4-Nitrophenyl Phosphate (DMNP)
The kinetics of catalytic degradation of dimethyl 4-nitrophenyl phosphate (DMNP) was
characterized using a procedure similar to the methods described in previous reports.7,8 2.5
mg of MOF powder sample or 14 mg of MOF-coated nanofiber sample was first dispersed in
1 mL of N-ethylmorpholine aqueous solution (0.45 M) in a 2 mL Eppendorf tube under fast
magnetic stirring (1100 rpm stir rate set on a Thermo Scientific stir plate) for 20 min. 4 μL
(~5.6 mg, 0.023 mmol) of DMNP was then added to the MOF suspension. The Eppendorf
tube was kept on the stir plate (1100 rpm stir rate) during the reaction. A 20 μL aliquot was
taken from the reaction mixture each time and diluted in 10 mL of N-ethylmorpholine
aqueous solution (0.15 M) for UV-visible absorbance spectroscopy. The reaction progress
was evaluated by monitoring the p-nitrophenoxide (degraded product from DMNP)
absorbance at 407 nm. The concentration of p-nitrophenoxide was calculated based on
Lambert-Beer Law (Equation S1).

𝐼
A = −lg = 𝜀𝐶𝑙 (S1)
𝐼0
183
Where A is the absorbance in units of absorbance unit (a.u.), I0 is the incident light intensity,
I is the transmitted light intensity, 𝜀 is the molar absorptivity coefficient
(𝜀 = 18330 𝑀−1 𝑐𝑚−1 ), C is the analyte concentration in unites of M, l is the length of light
path in units of cm.
The percent conversion of DMNP was obtained from the ratio of the p-nitrophenoxide
concentration before dilution to the initial DMNP concentration in the reaction mixture.
Degradation of O-Pinacolyl Methylphosphonofluoridate (GD)
Approximately 14 mg of MOF-coated PA-6@TiO2 nanofibers was weighed on a Mettler
Toledo AB104 calibrated digital scale. Once weighed, the samples were placed into a
humidity chamber (Thunder Scientific Corporation Series 2500) at 25oC to condition them at
50% relative humidity (RH) for at least 16 h. Upon humidification completion, the samples
were transferred into a 4-mm glass NMR tube. 700 µL of deionized water and 47 µL of
N-ethylmorpholine (buffer) were then added to the NMR tube. After 2.6 µL of O-pinacolyl
methylphosphonofluoridate (GD) was added onto the inner wall of the NMR tube, the tube
was capped and vigorously shaken in order to ensure mixture of GD with MOF-nanofibers in
the solution. The time of solution agitation was recorded in order to account for the Δ𝑡 of mix
31
time vs initial scan time. P NMR spectra were obtained over time at ambient temperature
(24-25°C) using a Varian INOVA 400 NB NMR spectrometer equipped with a Doty
184
Scientific 4-mm Liquid NMR probe to monitor the reaction and to identify the products.
The samples were not spun in order to limit slurry formation and obtain good reproducibility.
Samples scan times employed using a 90o pulse widths of 4μs and an overall acquisition time
of 8 min per acquisition. We noticed that the half-lives of GD in the reactions without
N-ethylmorpholine buffer exceed 300 min. Under same reaction conditions, the half-lives of
GD are less than 4 min with the presence of N-ethylmorpholine. We speculate that the HF
formed during the reaction is likely to poison the active sites on the MOF coatings, while the
N-ethylmorpholine buffer can neutralize HF and maintain the catalytic properties of the
MOFs.
185
Figure S5.1 SEM images of UiO-66-NH2 grown on (a-b) untreated PA-6 nanofibers and
(c-d) ALD TiO2 coated PA-6 nanofibers. Depositing a thin ALD TiO2 nucleation layer
significantly improves the growth uniformity and crystal coverage on the fiber surface.
186
Figure S5.2. SEM images of (a-b) electrospun PA-6 nanofibers and (c-d) ALD TiO2 coated
PA-6 nanofibers. The average diameter of untreated PA-6 nanofibers measured from SEM
images is 37±16 nm.
187
Table S5.1 MOF mass fraction in MOF-nanofiber composites calculated from BET results
and ICP-OES analysis
Overall MOF Substrat MOF Measure Measure MOF

BET Surfac e wt% d Zr wt% d Zr wt% wt%
Surfac e Area Surface Calculate in in MOF Calculate
e Area (m2/g) Area d from composit (%) d from
(m2/g) (m2/g) BET e (%) ICP
Results Results
(%) (%)
PA-6@TiO2
143.9 1512 11.5 8.82 2.81 33.0 8.51
@UiO-66
PA-6@TiO2
@UiO-66-N 205.9 1334 11.5 14.7 4.80 29.7 16.2
H2
PA-6@TiO2
356.2 2250 11.5 15.4 4.51 27.2 16.6
@UiO-67
Previously, we reported the following equation to correlate the MOF mass fraction with
the overall BET surface per unit mass of MOF plus fibers:
𝑆𝐴𝑀𝑂𝐹 ≈ (𝑆2)
where m is the mass and SA is the surface area for each component.10 With known surface
area for the uncoated fiber substrate, the MOF-coated fibers and the bulk MOF material, the
MOF mass fraction (ω) can be calculated using Equation X:
𝑆𝐴𝑀𝑂𝐹+𝑓𝑖𝑏𝑒𝑟 = 𝑆𝐴𝑀𝑂𝐹 ∙ 𝜔 + 𝑆𝐴𝑓𝑖𝑏𝑒𝑟 ∙ (1 − 𝜔) (𝑆3)
𝑆𝐴𝑀𝑂𝐹+𝑓𝑖𝑏𝑒𝑟 −𝑆𝐴𝑓𝑖𝑏𝑒𝑟
or 𝜔= (𝑆4)
𝑆𝐴𝑀𝑂𝐹 − 𝑆𝐴𝑓𝑖𝑏𝑒𝑟
We found that the MOF weight percentage calculated in this approach is similar to what
was measured by the ICP-OES method. This method provides a simple and straightforward
way to calculate the MOF weight percent when instruments for elemental analysis are not
188
available. One prerequisite is that both the MOF coatings on the fibers and the bulk MOF
powder are fully activated for BET measurements. It is also recommended to use the MOF
crystals collected from the same batch for MOF synthesis on fibers for BET analysis in order
to obtain a more accurate MOF mass fraction.
189
Figure S5.3 DMNP percent conversion as a function of time during the hydrolysis with
untreated PA-6 and ALD TiO2 coated PA-6 (PA-6@TiO2) nanofibers. Estimated t1/2 values
are 3950 min and 1170 min for PA-6 and PA-6@TiO2, respectively.
190
Figure S5.4 Kinetic analysis of DMNP degradation with (a-b) UiO-66, (c-d) UiO-66-NH2,
(e-f) UiO-67. The red curves in (a), (c), and (e) are plotted using the rate constants derived
from the linear fitting in (b), (d), and (f), respectively, based on the assumption of first order
reaction kinetics. We found that the reaction kinetics with UiO-66-NH2 does not fit well to
first-order rate equation.
191
Figure S5.5 Kinetic analysis of DMNP degradation with (a-b) PA-6@TiO2@UiO-66, (c-d)
PA-6@TiO2@UiO-66-NH2, (e-f) PA-6@TiO2@UiO-67. The red curves in (a), (c), and (e)
are plotted using the rate constants derived from the linear fitting in (b), (d), and (f),
respectively, based on the assumption of first order reaction kinetics. The reaction kinetics
with PA-6@TiO2@UiO-66-NH2 does not fit well to first-order rate equation.
192
Table S5.2 Kinetic constants and half-lives of DMNP hydrolysis with UiO MOF powders
and UiO MOF thin films on PA-6@TiO2 nanofibers.
k Calculated Observed MOF wt% TOF
(min-1) t1/2 (min) t1/2 (min) (s-1)
UiO-66 0.046 15 22 100 0.0054
UiO-66-NH2 0.23 3.0 2.8 100 0.045
UiO-67 0.076 9.1 7.7 100 0.020
PA-6@TiO2@UiO-66 0.0061 113 135 8.8 0.0018
PA-6@TiO2@UiO-66-NH2 0.078 8.9 7.3 14.7 0.021
PA-6@TiO2@UiO-67 0.097 7.1 7.4 15.4 0.024
*The amounts of catalysts used in DMNP degradation are 2.5 mg for UiO MOF powders and
14 mg for MOF-nanofiber composites. Turnover frequency (TOF) was calculated at t1/2,
although initial reaction rates give slightly higher TOF values.
193
Figure S5.6 SEM images and EDX spectra of (a-b) PA-6@TiO2@UiO-66 nanofibers, (c-d)
PA-6@TiO2@UiO-66 nanofibers and (e-f) PA-6@TiO2@UiO-67 nanofibers after DMNP
degradation experiment. SEM images and EDX results confirm that significant amounts of
MOF coatings remain in the MOF-nanofiber composites even after strong agitation during
the DMNP hydrolysis tests.
194
31
Figure S5.7 P NMR spectra of GD during hydrolysis with PA-6@TiO2@UiO-66
nanofibers.
195
31
Figure S5.8 P NMR spectra of GD during hydrolysis with PA-6@TiO2@UiO-66-NH2
nanofibers.
196
31
Figure S5.9 P NMR spectra of GD during hydrolysis with PA-6@TiO2@UiO-67
nanofibers.
197
(1) Schaate, A.; Roy, P.; Godt, A.; Lippke, J.; Waltz, F.; Wiebcke, M.; Behrens, P. Chem. –
Eur. J. 2011, 17 (24), 6643–6651.
(2) Katz, M. J.; Brown, Z. J.; Colon, Y. J.; Siu, P. W.; Scheidt, K. A.; Snurr, R. Q.; Hupp, J.
T.; Farha, O. K. Chem. Commun. 2013, 49 (82), 9449–9451.
(3) Andrady, A. L.; Ensor, D. S. Electrospinning in a controlled gaseous environment. US

Patent 7,297,305.
S.; Oldham, C. J.; Walls, H. J.; Shepherd, S. D.; et al. J. Mater. Chem. A 2015, 3 (4),
1458–1464.
(5) Walton, K. S.; Snurr, R. Q. J. Am. Chem. Soc. 2007, 129 (27), 8552–8556.
(6) Thommes, M.; Kaneko, K.; Neimark, A. V.; Olivier, J. P.; Rodriguez-Reinoso, F.;
Rouquerol, J.; Sing, K. S. W. Pure Appl. Chem. 2015, 87 (9–10), 1051–1069.
Hupp, J. T. Angew. Chem. Int. Ed Engl. 2014, 53 (2), 497–501.
W.; Hall, M. G.; DeCoste, J. B.; Peterson, G. W.; et al. Nat. Mater. 2015, 14 (5), 512–
516.
Farha, O. K. Inorg. Chem. 2015, 54 (20), 9684–9686.
Blevins, T. M.; Oldham, C. J.; Walls, H. J.; Shepherd, S. D.; et al. Adv. Mater.
Interfaces 2014, 1400040.
198
CHAPTER 6. Diffusion of CO2 in large single crystals of Cu-BTC
MOF
Trenton M. Tovar, †,§ Junjie Zhao‡,§William T. Nunn‡, Heather F. Barton‡,
Gregory W. Peterson#, Gregory N. Parsons‡,*, and M. Douglas LeVan†,*
†
Department of Chemical and Biomolecular Engineering, Vanderbilt University, Nashville,
Tennessee 37212, U.S.A. (m.douglas.levan@vanderbilt.edu)

‡

#
MD 21010, U.S.A.
§
T. M. Tovar and J. Zhao contributed equally to this work.
199
Abstract
Carbon dioxide adsorption in metal-organic frameworks has been widely studied for
applications in carbon capture and sequestration. A critical component that has been largely
overlooked is the measurement of diffusion rates. This paper describes a new reproducible
procedure to synthesize millimeter-scale Cu-BTC single crystals using concentrated reactants
and an acetic acid modulator. Microscopic images, XRD patterns, BET surface area, and
TGA results all confirm the high quality of these Cu-BTC single crystals. The large crystal
size aids in the accurate measurement of micropore diffusion coefficients.
Concentration-swing frequency response performed at varying gas-phase concentrations
gives diffusion coefficients that show very little dependence on loading up to pressures of 0.1
bar. The measured micropore diffusion coefficient for CO2 in Cu-BTC is 1.7 x 10-9 m2/s.
6.1 Introduction
The continuously increasing amount of CO2 emissions due to anthropogenic activities
has caused a sharp rise of CO2 levels in the atmosphere.1 While the search for alternative
clean energy sources continues, developing new technologies for CO2 capture and
sequestration is still necessary and has significant impact for controlling global warming.
Metal-organic frameworks (MOFs) are highly porous crystalline networks containing metal
clusters interconnected by organic linkers.2 MOFs typically have high surface area that
exceeds traditional adsorbent materials.3-7 In addition, the numerous options of metal centers
and linkers enable rational design and synthesis of MOF structures to achieve enhanced
adsorption capacities and catalytic properties. For example, MOFs have shown promising
200
performance for CO2/N2 separation for post-combustion CO2 removal from flue gas, CO2/H2
separation for pre-combustion capture, CO2/CH4 separation for natural gas upgrading, and
direct capture of CO2 from air.1,8,9 MOFs also show excellent catalytic activities for
converting CO2 to valuable compounds.10-13 As a result, MOFs are promising candidates for
CO2 capture and sequestration.
Despite the amount of research on CO2 adsorption and separations and catalysis using
MOFs, a major deficiency in the literature is the measurement of diffusion rates, a critical
property that can significantly affect the overall performance for the abovementioned
applications. CO2 diffusion constants have been reported for a few MOFs. Salles et al.14
measured diffusion rates ranging from 10-8 to 10-9 m2/s over a range of loadings on
MIL-47(V). Sabouni et al.15 reported 7x10-12 m2/s at 298 K for CPM-5, and Saha et al.16
reported diffusion constants on the order of 10-9 m2/s for MOF-177. Even for the same MOF
(MOF-5), the reported diffusion coefficients show a wide range from 10-9 m2/s to 10-8
m2/s.16,17
A big challenge for obtaining accurate diffusion coefficients is to decouple the presence
of different mass transfer mechanisms related to adsorbent morphology. For example, the
impact of surface barriers on MOF thin films was explored by Heinke et al.18 Fletcher et al.19
found a linear driving force behavior for CO2 in Ni2(4,4’-bipy)3(NO3)4, but macropore
diffusion was determined to be the controlling resistance for Co/DOBDC and Ni/DOBDC
pellets by Hu et al.20 using zero-length column techniques and for Cu-BTC pellets by Liu et
al.21 using concentration-swing frequency response (CSFR).
In this paper, we report a novel method to measure CO2 micropore diffusion coefficients
201
using CSFR to analyze millimeter-scale MOF single crystals. We use Cu-BTC (HKUST-1) as
a model to demonstrate this approach, as it shows one of the highest CO2 capacities (at
hydrated conditions) among all MOFs.9 High selectivity for CO2/H2 and CO2/N2 separations
have also been reported for this MOF.22-25
We developed a new synthesis technique to achieve mm-scale Cu-BTC single crystals,
and the crystal sizes can be well controlled. Microscopic images, XRD patterns, BET surface
area and TGA results were used to characterize the quality of these Cu-BTC single crystals.
Diffusion of CO2 is measured by the CSFR technique to determine rate controlling
mechanisms and diffusion constants, and rate behavior is compared with that in an activated
carbon.
Materials
Commercial adsorbents used in this study were Cu-BTC powder purchased from Sigma
Aldrich and BPL activated carbon (lot no 4814-5) in 6x16 mesh form obtained from Calgon
Carbon Corp. Research grade helium, pure CO2, and 1% CO2 in He were purchased from Air
Liquide.
Cu-BTC Single Crystal Synthesis
A low temperature solvothermal method was used to synthesize large Cu-BTC single
crystals. 0.49 g copper nitrate trihydrate (Cu(NO3)2∙3H2O) was dissolved in 3 mL of
deionized water. 0.24 g benzene-1,3,5-tricarboxylic acid (H3BTC) was dissolved in 3 mL of
202
ethanol (slight heating was needed to fully dissolve this organic linker). The Cu(NO3)2
solution was first mixed with 3 mL of N,N-dimethylformamide (DMF) in a 20 mL
scintillation vial. The H3BTC solution and 12 mL of glacial acetic acid (modulator) were
subsequently added to the mixed solution as well. The scintillation vial was then placed in a
furnace where the reaction was allowed to proceed at 55 °C for 3 days. After the
solvothermal synthesis, the Cu-BTC crystals were removed by hand from the glass surface of
the vial. All the crystals were soaked in ethanol for at least 3 days for solvent exchange.
Crystal Characterization
Cu-BTC single crystals were imaged using an Olympus BX60 optical microscope
equipped with a ProgRes C5 camera. Optical microscopic images were captured and
analyzed with ProgRes® CapturePro 2.8.8 software. SEM images of Cu-BTC crystal were
with an FEI Phenom® bench-top SEM. Crystals were mounted on conductive carbon taps
and sputter-coated with 5-10nm of Au-Pd before imaging. X-ray diffraction patterns of the
Cu-BTC single crystals were collected using a PANalytical Empyrean X-ray diffractometer.
BET surface area of Cu-BTC single crystals and powders was measured using a
Quantachrome Autosorb-1C surface area and pore size analyzer. 10 mg of Cu-BTC was
vacuum dried at room temperature for 12-18 h before the measurement of N2 adsorption
isotherm. BET surface area was calculated based on the isotherm within the P/Po range of
0.02-0.10. Thermogravimetric analysis (TGA) was measured with a Discovery TGA from TA
Instruments.
Pure CO2 isotherms were measured using a Rubotherm gravimetric analyzer. A sample
203
was degassed at 25 °C for approximately 12 hours using a turbo molecular pump until no
weight change was detected. The sample was then dosed with CO2 at a specified pressure and
allowed to equilibrate until no more weight change was detected. After equilibration, the CO2
pressure was increased to the next specified value. This process was repeated for the pressure
range of the isotherm, from 0.5 to 800 mmHg. The CO2 loading was calculated by the weight
gain of the sample at each pressure step.
Diffusion Measurement
CO2 diffusion was measured using a concentration-swing frequency response apparatus
(Figure S1). Approximately 15-20 mg of adsorbent was regenerated for 8 hours under
vacuum (at 25 °C for Cu-BTC cyrstals, at 250 °C for BPL activated carbon) for accurate
weight measurement. The sample was then loaded into a shallow bed and placed under
vacuum at room temperature with a 1 sccm He flow for 16 hours. Then a CO2 stream was
mixed with the He stream upstream of the adsorbent bed. Both streams were controlled with
MKS mass flow controllers, and the ratio of the two flow rates determined the gas-phase
concentration. The pressure in the adsorbent bed was controlled at 1 bar by a MKS Baratron
pressure controller, and effluent gas from the adsorbent bed was sampled by an Agilent 5975
mass spectrometer.
To run an experiment, the system was allowed to reach steady-state as determined by a
constant signal from the mass spectrometer. Then the mass flow controllers were used to
introduce sinusoidal perturbations to the flowrates of each stream but 180 degrees out of
phase. This resulted in a feed stream to the adsorbent bed that has a constant flow rate with a
204
sinusoidal concentration swing around the desired steady-state gas-phase concentration.
The perturbations were performed at different frequencies in the range of 0.001 to 0.1 Hz.
The collected data are presented as plots of the amplitude ratio (AR) as a function of the
perturbation frequency, where the AR is calculated from the amplitude of the gas exiting the
adsorbent bed divided by the amplitude of the gas entering the adsorbent bed. The diffusion
mechanism and rate parameters can be extracted by fitting the data to a mathematical model
derived from transfer functions. Detailed mathematical models for a CSFR apparatus have
been described elsewhere.26,27
Crystal Characterization.
Figures 6.1c-f show the optical micrographs and SEM images of the Cu-BTC single
crystals. Truncated cube and truncated octahedron shapes were observed for this fcu-type
MOF crystal. The octagon (or square) facets represent the planes viewed down the <100>
directions, while the hexagon-shape facets correspond to the planes along the <111>
directions. The size of Cu-BTC crystals obtained from a 3-day synthesis ranges from 500 μm
to 1.3 mm (Figure S6.2). These crystals were sorted by size for later CO2 diffusion studies.
The quality of the Cu-BTC single crystals was characterized using XRD, BET and TGA.
Figure 6.2 shows the XRD patterns for the crystal planes parallel to (100) and (111)
respectively. The peak positions match well with the corresponding peaks present in the
simulated powder diffraction pattern and reported powder patterns in the literature.37 The fact
that only peaks associated with targeted parallel crystal planes appear in the diffraction
205
patterns and that optical micrographs show negligible defects confirm the formation of
high-quality single crystals.
Figure 6.1 (a-b) Molecular representation of the Cu-BTC MOF along the [100] direction
and [111] direction. Color code: Cu (yellow); O (red); C (black); H (not shown). (c-d)
Optical microscopic and (e-f) SEM images of Cu-BTC single crystals.
206
BET surface area and pore volume for the Cu-BTC single crystals were measured and
compared with the Cu-BTC fine powder prepared from 85 °C solvothermal synthesis (Figure
S6.3). The surface area is 1980 and 1825 m2/g for the single crystals and the fine powder
respectively. Pore volumes are 0.85 and 0.79 cm3/g for the large crystals and powder samples
respectively. Both the surface area and pore volume compare well with values from the
literature.22
Figure 6.2 XRD patterns of a Cu-BTC single crystal for the planes parallel to (100) (black)
and the planes parallel to (111) (red), and simulated powder diffraction pattern for Cu-BTC.
TGA was used to evaluate the thermal stability of the Cu-BTC single crystals (Figure
S6.4). Cu-BTC powder shows a steeper initial mass drop compared to the large crystals due
to the shorter diffusion length in small particles for solvent evaporation. We also noticed that
the mm-scale crystal exhibits a sharper mass decrease than the fine powder in the
207
decomposition regime (300-385 °C), possibly because of slower heat transfer and less
impurities in the large crystals. Our measured decomposition temperature regime also agrees
well with reported values,29,30 indicating the good quality of these MOF crystals.
Crystal Growth Mechanism
The crystal size was measured after specific growth times (1-5 d), and Figure 6.3 shows
that the size of the largest crystal increases as a function of synthesis time. No crystals form
during the nucleation period (1 d), and the crystal size quickly ramps up to ~1.3 mm at 3 days.
After 3 days, the change in crystal size is very small, indicating the termination of crystal
growth. We used a Gompertz function (Equation S1), a common mathematical model to
describe such growth behavior, and the data fit well to the model. Taking the mass-size
correlation into account (Figure S6.5 and Equation S2), this growth function can be
converted into a plot similar to the classic La Mer diagram (Figure S6.6).31
At the beginning of the synthesis, secondary building units (SBUs) of Cu-BTC MOF
start to form in the solution. As the reaction proceeds and the SBU concentration reaches
supersaturation (after ~1 d), heterogeneous nucleation occurs on the wall of the scintillation
vial as it is more thermodynamically favorable. The crystal growth rate depends on the SBU
concentration and stops once the concentration drops below the supersaturation point (after
~4 d).
While previous work reported the use of nitric acid to inhibit the deprotonation of
H3BTC linker and thus slow down the crystal growth rate in order to achieve large single
crystals,32 our method is more effective and reproducible by using a supersaturated
208
concentration of MOF reactants and an acetic acid modulator. The formation of an acetate
complex competes with the generation of SBU in the solution, allowing the slow and steady
growth of our single crystals. By avoiding high acidity of the precursor solution, we also
obtained high crystal quality as shown above.
Figure 6.3 Increase of Cu-BTC crystal size as a function of reaction time.
Diffusion Measurements
To ensure that CO2 adsorption in the single crystals is similar to Cu-BTC powder, a CO2
isotherm was measured and compared with literature values. Figure 6.4 shows the CO2
isotherm measured at 25 °C on Cu-BTC crystals of various sizes. The isotherm fits well to a
Toth model.33 Figure 6.4 also shows some of the representative CO2 isotherms reported in the
literature. We plotted these curves based on parameters for a Sips fit from Aprea et al.34 and a
Langmuir fit from Hamon et al.35 Select data points at 1 bar were also adapted from Yazaydin
et al.,22 Millward and Yaghi,36 and Wang et al.37 While most isotherms in the literature were
209
measured at high pressure, due to the large capacity of Cu-BTC, CO2 isotherms at low
pressures are important for comparison to diffusion measurements at low CO2 concentrations.
It is clear that the isotherm for the large Cu-BTC single crystals compares favorably with the
MOF powder at low CO2 pressure.
Figure 6.4 Gravimetric CO2 isotherm on large Cu-BTC single crystals at 298 K fit by a
Toth isotherm model. Also shown are data points from literature isotherms of CO2 on powder
Cu-BTC at 298 K.
Figure 6.5 compares the CSFR curves for the Cu-BTC single crystals and the powder at
a gas-phase concentration of 0.5% CO2 and total pressure of 1 bar. The CSFR curve for the
Cu-BTC powder has a steeper slope than the single crystals, indicating faster CO2 uptake in
the powder. This is reasonable because the diffusion length in the powder (crystal size < 50
μm) is much shorter than for the mm-scale single crystals. We used the micropore diffusion
model reported previously to fit the curves,26 and found that this model describes the single
crystals well but shows noticable deviation from the powder data. The fact that the micropore
210
diffusion model fails to fit the powder data well is possibly because other diffusive
mechanisms, such as surface barriers or external mass transfer, may also be involved in
addition to micropore diffusion. For the mm-scale Cu-BTC single crystals, micropore
diffusion is the dominant factor that limits mass transfer rates, ensuring that the measured
diffusion coefficients truly represent the diffusion in the MOF micropores. Therefore, single
crystals are excellent candidates for micropore diffusion measurements.
1.0
0.8
Amplitude Ratio
0.6
0.4
Cu-BTC crystals
Cu-BTC powder
0.2
0.0 2 4 6 8 2 4 6 8 2 4 6 8
0.001 0.01 0.1 1
Frequency (Hz)
Figure 6.5 Amplitude ratio curves for CSFR experiments on large Cu-BTC single crystals
versus a powder sample.
The micropore diffusion model uses two parameters to fit the CSFR curves; the
regressed values are the local isotherm slope, K, and the micropore diffusion coefficient
parameter Ds/R2, where R is the diffusion length scale. As the regressed diffusion coefficient
211
is dependent on particle size, accurate diffusion measurements require precise knowledge of
the crystal size, which is well controlled by the synthesis technique. Three sets of Cu-BTC
crystals with particle sizes of 0.7±0.1 mm, 1.0±0.1 mm, and 1.3±0.1 mm were prepared,
where particle size corresponds to the longest crystal dimension. These were used for CSFR
measurements using 0.5% CO2 concentration. Figure 6 shows the CSFR curves for each set
of crystals, and the fitted parameters for the micropore diffusion model are given in Table 1.
The regressed isotherm slopes show good agreement between the different crystal sizes. The
Ds/R2 values decrease as the crystal size increases, as expected.
Figure 6.6 CSFR curves for different Cu-BTC crystal sizes at 0.5% CO2 concentrations.
The micropore diffusion coefficients are calculated by multiplying the regressed
diffusion parameter by the square of the diffusion length scale. However, the model assumes
a spherical particle shape that does not accurately represent the crystals. We found that the
height of crystals is approximately half of the longest dimension (L), indicating that the
212
effective diffusion length should be smaller than L. We measured the average crystal mass
per crystal size and correlated it as a function of L (Figure S6.5). A power law fit the data
well with an exponent of 2.14 instead of the expected cubic function for a perfect sphere. To
account for the shape factor, an effective diffusion length (Reff) was defined as
4
mass   Reff3  L2meas
.14
(1)
3
where Lmeas (in units of mm) is the length of the longest crystal dimension, ρ (in units of
g/cm3) is the crystal density, α = 0.52 mg∙mm-2.14. For single crystals, the measured density is
1.29 g/cm3, similar to reported values for Cu-BTC.1,10 When normalized by Reff, the
micropore diffusion coefficients agree fairly well and give an average value of 0.0017 mm2/s
(or 1.7 x 10-9 m2/s). This value is on the same order of magnitude as several other CO2
diffusion rates reported in other MOFs.
Table 6.1 Micropore diffusion fitting parameters for various Cu-BTC crystal sizes at 0.5%
CO2.
Crystal Size K Ds/R2 Ds

(mm) (mol/kg∙bar) (1/s) (mm2/s)
0.7 7.1 0.011 0.0014
1 7.1 0.0082 0.0017
1.3 6.6 0.0066 0.0020
An important aspect of diffusion in nanoporous materials is the impact of surface
concentration on diffusivity. Figure 6.7 shows the CSFR curves measured at various
gas-phase concentrations using 18 mg of Cu-BTC single crystals with diameters of ~0.7 mm.
The steady-state CO2 concentrations used in this series of experiments were 0.1%, 0.5%, and
213
10%. We found that the micropore diffusion model accurately fits the data at each tested CO2
concentration. The curves are nearly identical for the 0.1 and 0.5% experiments while the 10%
curve is shifted to higher amplitude ratios indicating a decrease in the local isotherm slope.
Figure 6.7 CSFR curves for various CO2 concentrations on Cu-BTC crystals of
approximately 1.5 mm.
Table 2 summarizes the fitted parameters for the CSFR curves at each concentration as
well as for a similar set of experiments on BPL activated carbon (Figure S7), a
predominantly microporous material that follows the micropore diffusion model well for
various adsorbates.38 For comparison, we also calculated the K values from the Toth fit of the
CO2 isotherm for the Cu-BTC crystals (Figure 4) and list them in Table 2. The K values
measured from CSFR experiments are slightly higher than those from the CO2 isotherm, but
the trends are very similar. For the 0.1% and 0.5% experiment, the isotherm slopes are
similar, which agrees with the near linear shape of the CO2 isotherm at such low
214
concentrations. For the 10% experiment, the K value significantly decreases for both the
CSFR and adsorption isotherm measurements.
Table 6.2 Micropore diffusion fitting parameters for 0.7 mm Cu-BTC single crystals and
BPL activated carbon at different CO2 concentrations.
Cu-BTC BPL Cu-BTC

CSFR CSFR isotherm
Conc K Ds/R2 K Ds/R2 K
% (mol/kg∙bar) (1/s) (mol/kg∙bar) (1/s) (mol/kg∙bar)
0.1 6.9 0.0090 4.9 0.0024 5.9
0.5 6.7 0.0086 4.2 0.0072 5.8
10 5.8 0.0089 2.8 0.0142 5.1
It is very interesting that the diffusion coefficients for Cu-BTC single crystals and BPL
activated carbon particles show completely different trends. For the single crystal Cu-BTC
samples there is no significant difference from tests using CO2 concentrations from 0.1% to
10%. However, for BPL activated carbon, the diffusion coefficient increases dramatically as
a function of CO2 concentration. These different trends are reasonable since the Cu-BTC
single crystals have narrow distribution of pore sizes, ordered lattice structure, and therefore,
more homogenous adsorption sites. Thus, the adsorbate-adsorbent interaction at low partial
pressure remains similar at increased CO2 pressures and loadings. This agrees well with
literature, as Wang et al.37 reported the heat of adsorption of CO2 remains virtually constant
as loading increases after an initial decrease. In contrast, the BPL activated carbon is very
heterogeneous, so the first CO2 molecules will associate with stronger adsorption sites and
will diffuse more slowly than at higher pressures.
215
In comparison with the MOF, the BPL activated carbon particles were larger,
approximately 2 mm in diameter. After normalizing by the diffusion length scale, the
diffusion coefficient for the single MOF crystals is significantly slower than for the activated
carbon. This could be due to strong intermolecular interaction between CO2 and the open
metal sites of Cu-BTC. It could also be due to the small uniform 9 Å pores of Cu-BTC,
whereas the activated carbon has a wide distribution of pore sizes.
6.4 Conclusion
A new synthesis procedure to produce large millimeter size single crystals of Cu-BTC
has been reported. Crystal size could be accurately controlled by length of the synthesis time.
Properties measured by XRD, BET, TGA, and CO2 isotherms on the Cu-BTC crystals agree
well with Cu-BTC powders produced by traditional solvothermal synthesis. The increased
length scale of the microporous domain makes diffusion measurements easier as it limits the
impact of multiple mass transfer mechanisms. CSFR experiments were performed which
verified micropore diffusion as the rate controlling diffusive mechanism. Diffusion
coefficients measured at varying concentrations show very little dependence on loading up to
about 0.1 bar, the highest pressure considered. At 0.5% gas-phase concentration of CO2 in He
at 1 bar, the measured diffusion coefficient is approximately 1.7 x 10-9 m2/s.
216
References
(1) Sumida, K.; Rogow, D. L.; Mason, J. A.; McDonald, T. M.; Bloch, E. D.; Herm, Z. R.;
Bae, T.-H.; Long, J. R. Chem. Rev. 2012, 112 (2), 724-781.
(2) Kim, J.; Chen, B.; Reinike, T. M.; Li, H.; Eddaoudi, M.; Moler, D. B.; O’Keeffe, M.;
Yaghi, O. M. J. Am. Chem. Soc. 2001, 123 (34), 8239-8247.
(3) Furukawa, H.; Ko, N.; Go, Y. B.; Aratani, N.; Choi, S. B.; Choi, E.; Yazaydin, A. O.;
Snurr, R. Q.; O’Keeffe, M.; Kim, J.; Yaghi, O. M. Science 2010, 329 (5990), 424-428.
(4) Farha, O. K.; Yazaydin, A. O.; Eryazici, I.; Malliakas, C. D.; Hauser, B. G.; Kanatzidis,
M. G.; Nguyen, S. T.; Snurr, R. Q.; Hupp, J. T. Nat. Chem. 2010, 2, 944-948.
(5) Farha, O. K.; Eryazici, I.; Jeong, N. C.; Hausert, B. G.; Wilmer, C. E.; Sarjeant, A. A.;
Snurr, R. Q.; Nguyent, S. T.; Yazaydin, A. O.; Hupp, J. T. J. Am. Chem. Soc. 2012, 134
(36), 15016-15021.
(6) Grunker, R.; Bon, V.; Muller, P.; Stoeck, U.; Krause, S.; Mueller, U.; Senkovska, I.;
Kaskel, S. Chem. Commun. 2014, 50, 3450-3452.
(7) Wang, T. C.; Bury, W.; Gomez-Gualdron, D. A.; Vermeulen, N. A.; Mondloch, J. E.;
Deria, P.; Zhang, K.; Moghadam, P. Z.; Sarjeant, A. A.; Snurr, R. Q.; Stoddart, J. F.;
Hupp, J. T.; Farha, O. K. J. Am. Chem. Soc. 2015, 137 (10), 3585-3591.
(8) Li, J.-R.; Ma, Y.; McCarthy, M. C.; Sculley, J.; Yu, J.; Jeong, H.-K.; Balbuena, P. B.;
Zhou, H.-C. Coordin. Chem. Rev. 2011, 255, 1791-1823.
(9) Keskin, S.; van Heest, T. M.; Sholl, D. D. ChemSusChem, 2010, 3, 879-891.
(10) Wang, C.; Xie, Z; deKrafft, K. E.; Lin, W. J. Am. Chem. Soc. 2011, 133 (34),
13445-13454.
(11) Fu, Y.; Sun, D.; Chen, Y.; Huang, R.; Ding, Z.; Fu, X.; Li, Z. Angew. Chem.-Int. Ed.
2012, 51 (14), 3364-3367.
(12) Sun, D.; Fu, Y.; Liu, W.; Ye, L.; Wang, D.; Yang, L.; Fu, X.; Li, Z. Chem.-Eur. J. 2013,
19 (42), 14279-14285.
(13) Wang, S.; Yao, W.; Lin, J.; Ding, Z.; Wang, X. Angew. Chem.-Int. Ed. 2014, 53 (4),
1034-1038.
217
(14) Salles, F.; Jobic, H.; Devic, T.; Llewellyn, P. L.; Serre, C.; Ferey, G.; Maurin, G. ACS
Nano 2010, 4 (1), 143-152.
(15) Sabouni, R.; Kazemian, H.; Rohani, S. Micropor Mesopor Mater. 2013, 175, 85-91.
(16) Saha, D.; Bao, Z.; Jia, F.; Deng, S. Environ. Sci. Technol. 2010, 44, 1820-1826.
(17) Zhao, Z.; Li, Z.; Lin, Y. S. Ind. Eng. Chem. Res. 2009, 48, 10015-10020.
(18) Heinke, L.; Gu, Z.; Woll, C. Nature Communications 2014, 5, 4562, 1-6.
(19) Fletcher, A. J.; Cussen, E. J.; Prior, T. J.; Rosseinsky, M. J.; Kepert C. J.; Thomas, K. M.
J. Am. Chem. Soc. 2001, 123 (41), 10001-10011.
(20) Hu, X.; Brandani, S.; Benin A. I.; Willis, R. R. Ind. Eng. Chem. Res. 2015, 54,
5777-5783.
(21) Liu, J.; Wang, Y.; Benin, A. I.; Jakubczak, P.; Willis, R. R.; LeVan, M. D. Langmuir
2010, 26 (17), 14301-14307.
(22) Yazaydin, A. O.; Benin, A. I.; Faheem, S. A.; Jakubczak, P.; Low, J. J.; Willis R. R.;
Snurr, R. Q. Chem. Mater. 2009, 21, 1425-1430.
(23) Yang, Q.; Zhong, C. J. Phys. Chem. B 2006, 110, 17776-17783.
(24) Liu, B.; Smit, B. Langmuir 2009, 25 (10), 5918-5926.
(25) Yang, Q.; Xue, C.; Zhong C.; Chen, J.-F. AIChE J. 2007, 53 (11), 2832-2840.
(26) Wang, Y.; LeVan, M. D. Ind. Eng. Chem. Res. 2008, 47, 3121-3128.
(27) Wang, Y.; LeVan, M. D. AIChE J. 2011, 57 (8), 2054-2069.
(28) Zhao, J.; Nunn, W. T.; Lemaire, P. C.; Lin, Y.; Dickey, M. D.; Oldham, C. J.; Walls, H.
J.; Peterson, G. W.; Losego, M. D.; Parsons, G. N. J. Am. Chem. Soc. 2015, 137 (43),
13756–13759.
(29) Chowdhury, P.; Bikkina, C.; Meister, D.; Dreisbach, F.; Gumma, S. Micropor. Mesopor.
Mat. 2009, 117, 406-413.
(30) Liang, Z.; Marshall, M.; Chaffee, A. L. Energ. Fuel. 2009, 23, 2785-2789.
218
(31) Stock, N.; Reinsch, H.; Schilling, L.-H. CHAPTER 2. Synthesis of MOFs. In RSC
Catalysis Series; Llabrés i Xamena, F., Gascon, J., Eds.; Royal Society of Chemistry:
Cambridge, 2013; pp 9–30.
(32) Li, L.; Sun, F.; Jia, J.; Borjigin, T.; Zhu, G. Growth of large single MOF crystals and
effective separation of organic dyes. Crystengcomm 2013, 15 (20), 4094–4098.
(33) Do, D. D. Adsorption Analysis: Equilibrium and Kinetics, Imperial College Press:
London, 1998.
(34) Hamon, L.; Jolimaitre, E.; Pirngruber, G. D. Ind. Eng. Chem. Res. 2010, 49,
7497-7503.
(35) Aprea, P.; Caputo, D.; Gargiulo, N.; Iucolano, F.; Pepe, F. J. Chem. Eng. Data 2010, 55,
3655-3661.
(36) Millward, A. R.; Yaghi, O. M. J. Am. Chem. Soc. 2005, 127, 17998-17999.
(37) Wang, Q. M.; Shen, D.; Bulow, M.; Lau, M. L.; Deng, S.; Fitch, F. R.; Lemcoff, N. O.;
Semanscin, J. Micropor. Mesopor. Mat. 2002, 55, 217-230.
(38) Wang, Y.; Mahle, J. J.; Furtado, A. M. B.; Glover, T. G.; Buchanan, J. H.; Peterson G.
W.; LeVan, M. D. Langmuir 2013, 29, 2935-2945.
219
Figure S6.1 Schematic of the concentration-swing frequency response apparatus.
220
Figure S6.2 Optical microscopic images of Cu-BTC crystals obtained from a three-day
solvothermal synthesis. The crystal size ranges from 500 μm to 1.3 mm. (Crystal size
measurements are labelled on the images.)
221
Figure S6.3 (a) N2 isotherms for Cu-BTC single crystals and fine powder. (b) BET surface
area and (c) pore volume calculated from N2 isotherms.
222
Figure S6.4 Thermogravimetric analysis (TGA) for Cu-BTC single crystals (red) and fine
powder (black).
223
Figure S6.5 The correlation between the mass per crystal and the crystal dimension. The
error bar indicates the standard deviation of the crystal sizes in a mass measurement group.
224
Gompertz function for fitting the growth curve:
𝐿 = 𝑎 ∗ exp(− exp(−𝑘(𝑡 − 𝑡𝑐 ))) Equation S1
Where L (mm) is the size of the large crystal obtained from particular reaction time, 𝑎 (mm)
is the pre-exponential factor, k (d-1) is the kinetic factor, t and tc (in units of d) are the growth
time and transition time (for fitting purpose), respectively. After fitting the data to the
Gompertz function, we get 𝑎 = 1.34 mm, tc = 1.78 d, k = 2.18 d-1.
Figure S4 shows the correlation between the mass and crystal dimension. We used a power
law to fit the data:
𝑚 = 𝜌 ∗ 𝐿𝑛 Equation S2
Where m (mg) is the mass per crystal, L (mm) is the size of the large crystal (crystal
dimension), 𝑛 is power, 𝜌 (𝑚𝑔 ∙ 𝑚𝑚−𝑛 ) is the density factor. After fitting, we get n = 2.14,
𝜌 = 0.52 𝑚𝑔 ∙ 𝑚𝑚−2.14 .
The reaction kinetics can be derived by substituting L in Eqn S2 with Eqn 1 and taking a
derivative of Eqn S2:
𝑑𝑚 𝑑𝐿
= 𝜌𝑛𝐿 𝑛−1 ∗
𝑑𝑡 𝑑𝑡
𝑑𝑚
= 𝜌𝑛𝐿 𝑛−1 ∗ 𝐿 ∗ {− exp[−𝑘(𝑡 − 𝑡𝑐 )]}′
𝑑𝑡
𝑑𝑚
= 𝜌𝑛𝐿 𝑛 ∗ {− exp[−𝑘(𝑡 − 𝑡𝑐 )]} ∗ (−𝑘)
𝑑𝑡
𝑑𝑚
= 𝜌𝑛𝐿 𝑛 ∗ exp[−𝑘(𝑡 − 𝑡𝑐 )] Equation S3
𝑑𝑡
225
Figure S6.6 Crystal mass increase rate (dm/dt) as a function of reaction time based on
Eqnation S3.
226
Figure S6.7 CO2 CSFR curves on BPL activated carbon at varying concentrations.
227
CHAPTER 7. ALD-Assisted Synthesis of ZnO 3D Hierarchical
Nanostructures for Enhanced Photocatalytic Dye Degradation
Junjie Zhao†, Abhijeet Bagal‡, Qiaoyin Yang‡, Xu Zhang‡ and Chih-Hao Chang‡,
Mark D. Losego† and Gregory N. Parsons†,*
†

‡
Department of Mechanical and Aerospace Engineering, North Carolina State University,
911 Oval Drive, Raleigh, NC 27695, U.S.A.
228
Abstract
ZnO 3D hierarchical nanostructures can overcome limitations of free-standing
nano-slurries, and show promising photocatalytic performance. However, very few reports
have systematically investigated how surface–attached nanostructure dimensions affect
photocatalyic performance. In this work, interference lithography was used to fabricate
nano-pillars with controllable heights (250 ~ 800 nm) as templates for photocatalytic ZnO
coatings via atomic layer deposition (ALD). Conformal ALD ZnO thin films promote
uniform nucleation of ZnO nano-wires synthesized using hydrothermal method. The length
of nano-wires increase systematically with growth time (60 ~ 150 min). We find that taller
nano-pillars enhance the rate for photocatalytic degradation of methyl orange, and the
addition of hierarchical nanowires increases the first order degradation rate constant by up to
61%.
7.1 Introduction
Semiconductor-based photocatalysts have attracted tremendous interest over the past
few decades for organic synthesis, water splitting, and decontamination and disinfection
applications.1–4 ZnO is a II-VI semiconductor (large direct bad gap of 3.37eV at room
temperature)5 with similar photocatalytic activity compared to TiO2, and is often considered
as a possible alternative photocatalyst to TiO2 because of its low cost and non-toxic
property.4,6 Specifically in applications for waste water treatment, ZnO has shown promising
photocatalytic performance for degrading azo dyes,6,7 and other organic contaminants.8-10
While using ZnO nano-slurries as photocatalyst can achieve high reaction rate, particle
229
aggregation in solution and difficulty in separating catalyst after use for recycling are major
practical challenges.6,11 Immobilizing ZnO nanostructure photocatalysts on substrate
materials could potentially overcome these limitations.
Many thoughtful strategies have been reported for growing ZnO nanostructures onto
functional substrate materials.11-17 Among them, hierarchical nanostructures are of particular
interest as these structures can exhibit large surface area and light trapping properties.12 For
example, Gong et al., Lee et al. and Colmenares et al. reported ZnO nano-wires/nano-leaves
grown on ZnO seeded fiber surfaces using hydrothermal method.11,16,17 Hung et al.
fabricated ZnO pore-array films using nanosphere mask and electrochemical deposition, and
further decorated the structure with ZnO nano-wires and Ag nanoparticles.12 Xu et al.
synthesized ZnO nanoplate-nanowire hierarchical structures using eletrodeposition and
hydrothermal synthesis.13 Compared with flat ZnO surfaces, nano-wires, nano-plates or
nanosheets, higher photocatalytic activity has been reported for ZnO 3D hierarchical
nanostructures.11-13 However, because of challenges in systematic tuning of periodic
hierarchical nanostructures, very few reports have investigated the role of nanostructure
dimension on photocatalyic activity and overall reaction rates.
Here, we report a new interference lithography–based procedure to synthesize ZnO 3D
hierarchical nanostructures with controllable periodicity and feature size. A series of nano-pillar
structures with varied heights were coated with conformal ALD ZnO thin films. The pillars were
further decorated with hydrothermally grown ZnO nano-wires. The rate of photocatalytic
degradation of methyl orange (MO) was investigated vs feature height and nanowire length.
230
Figure 7.1 Schematic of the fabrication procedure for the ZnO 3D hierarchical
nanostructures. (a) Silicon wafer spun-coated with antireflection layer and positive
photoresist. (b) Periodic nano-pillars fabricated using laser interference lithography. (c)
Conformal ZnO thin films deposited onto the nano-pillars via ALD. (d) ZnO 3D hierarchical
nanostructure obtained from hydrothermal synthesis of ZnO nano-wires on ALD ZnO coated
nano-pillars.
Fabrication of Periodic Nano-Pillars
Periodic nano-pillar structures (500 nm periodicity) were patterned using laser
interference lithography (shown in Figure 7.1(a-b)).18 An antireflection coating (ARC
i-CON-16, Brewer Science, Inc.) and a layer of positive photoresist (Sumitomo PFi88A7)
were spun-coated on silicon wafers. The thickness of the antireflection coating was
231
controlled at 97 nm, while the thickness of the photoresist layer was tuned to control the
height of the nano-pillars (250 nm, 500 nm and 800 nm, respectively). A HeCd laser source
(wavelength = 325 nm) and a Lloyd’s mirror interferometer were used for interference
lithography. The periodic nano-pillar patterns were obtained after two orthogonal exposures
and post-exposure development, and the periodicity was controlled constant at 500 nm.
Atomic Layer Deposition of ZnO
ALD ZnO was deposited onto the nano-pillars (Figure 7.1c) using a customized hot-wall
viscous-flow vacuum reactor. The deposition pressure was controlled at ~2 Torr, and the
temperatures were kept at 100℃. In an ALD ZnO cycle, diethyl zinc (DEZ) was first dosed
to the chamber for 2 s, followed with 60 s of purge with dry N2 (99.999%, further filtered
through an Entegris gatekeeper). Deionized water was subsequently dosed to the chamber for
2 s, followed with another 60 s of N2 purge. 200 cycles of ALD ZnO were deposited onto the
nano-pillars as nucleation layers for hydrothermal synthesis of ZnO nano-wires.
Hydrothermal Growth of ZnO Nano-Wires
To prepare the precursor solution for the hydrothermal synthesis, 0.02 M zinc nitrate
hexahydrate (Zn(NO3)2·6H2O, 99%, Aldrich) and 0.02 M hexamethylene tetramine (C6H12N4,
99%, Aldrich) were mixed in 100 mL of deionize water in a 250 mL PTFE beaker.
232
Substrates with ZnO coated nano-pillars were transferred into the aqueous precursor solution.
The surface with nanostructures was kept facing down in the solution using a PTFE sample
holder during the synthesis. The PTFE beaker was covered with a piece of aluminum foil
(with several holes made on the foil), and placed in a furnace with a set reaction temperature
of 80 ◦C. The reaction time was controlled (60 min, 90 min, 120 min and 150 min) to tune the
lengths of nanowires. After hydrothermal growth, the samples were rinsed with deionized
water for 2 min and dried in N2 flow at room temperature.
The thicknesses of ALD thin films on flat silicon wafers were measured with a
fixed-angle spectroscopic ellipsometer (J.A. Woollam Co., Inc). Contact angle of static water
on sample surfaces were measured using a Ramé-Hart Model 200 contact angle goniometer
with a CCD camera. At least three different spots were measured on each sample and the
average value of the contact angles was calculated. Scanning electron microscopic (SEM)
images of the ZnO 3D hierarchical nanostructures were taken using a field emission SEM
(JEOL JSM-6400F). Samples were sputter-coated with a thin layer of Au-Pd (5~10 nm)
before imaging. The dimension of the nano-pillars and nano-wires were measured using
ImageJ (1.48v) software.
233
Photocatalytic Characterization
ZnO nanostructure photocatalyst sample on silicon substrate (1  1 cm2) was placed into
a 50 mL Pyrex glass vial containing 30 mL of methyl orange (MO) in deionized water (15
mg/L). The solution was continuously stirred for 30 minutes in the dark to promote
equilibrium adsorption of the dye onto the catalytic surface. Photocatalysis was tested under
UV exposure (Intelli-Ray 400, Uvitron International, 245 μW/cm2) with continuous stirring
at constant temperature ~ 30 ◦C using a water bath. Every 30 min, 0.5 mL of solution aliquot
was collected to characterize MO concentration via UV–Vis absorption at 466 nm. After
each UV-Vis measurement, the sampled solution was added back to the Pyrex vial to
maintain constant reactant volume.
Analysis of Photocatalytic Reaction Kinetics
The photocatalytic degradation data was modeled using pseudo-first-order reaction
kinetics:
−𝑟𝑀𝑂 = 𝑘𝑎𝑝𝑝 𝐶𝑀𝑂 (1)
𝑑𝐶𝑀𝑂
𝑟𝑀𝑂 = (2)
𝑑𝑡
where 𝑟𝑀𝑂 (in units of 𝑚𝑜𝑙 ∙ 𝐿−1 ∙ ℎ−1 ) is the rate of the photocatalytic reaction,
𝑘𝑎𝑝𝑝 (in units of ℎ−1 ) is the apparent first order rate constant. 𝐶𝑀𝑂 (in units of 𝑚𝑜𝑙 ∙ 𝐿−1 )
is the concentration of MO during the reaction. t (in units of h) is the reaction time.
Replacing the 𝑟𝑀𝑂 in Equation 1 with 𝑑𝐶𝑀𝑂 /𝑑𝑡 and integrating the both sides of the
equation, we derived Equation 3:
234
𝐶𝑀𝑂
− 𝑙𝑛 ( ) = 𝑘𝑎𝑝𝑝 𝑡 (3)
𝐶0
where 𝐶0 (in units of 𝑚𝑜𝑙 ∙ 𝐿−1 ) is the initial concentration of MO. By plotting −𝑙𝑛(𝐶𝑀𝑂 /
𝐶0 ) as a function of t and linearly fitting the data, we obtained 𝑘𝑎𝑝𝑝 from the slope of the
fitted line. The half-life of MO (𝑡1/2 , in units of h) during the photocatalytic degradation is
given by:
𝑙𝑛2
𝑡1/2 = (4)
𝑘𝑎𝑝𝑝
Surface Wettability Tuned by ZnO ALD
Periodic nano-pillars were fabricated on silicon wafers using laser interference
lithography. Insert SEM image in Figure 7.2 shows the cylindrical morphology of these
nano-pillars. The nano-pillars used for wettability comparison have a periodicity of ~ 700 nm,
and the height of the pillars was controlled at 500 nm. Figure 7.2(a-b) compare the water
droplet contact angles of the nano-pillars before and after ZnO ALD. Because of the
hydrophobicity of the photoresist, the surface before ALD ZnO exhibits a contact angle of
103◦. In comparison, 200 cycles of ALD ZnO (~36 nm thick) improve the surface wettability
significantly, reducing the contact angle to less than 15◦. This wettability change is a result of
–OH terminal groups on the ALD ZnO surface, consistent with previous reports.19,20 The
improved surface wettability is important for the hydrothermal synthesis to achieve
conformal ZnO nano-wire growth on the nanostructure substrate.11
235
Figure 7.2 Water droplet contact angles of the periodic nano-pillars (500 nm high) (a)
before ZnO ALD and (b) after ZnO ALD (200 cycles). Insert SEM images represent the
nano-pillars before and after ZnO ALD, respectively. Scale bars in SEM images represent
1μm.
ZnO Nano-wire Growth on ALD Coated Nano-pillars
Figure 7.3(a-d) show SEM images of ALD ZnO coated nano-pillars before and after
hydrothermal synthesis at 80 °C with reaction time from 60 to 150 min. With the ALD ZnO
nucleation layer, conformal and uniform ZnO nano-wires were successfully grown onto the
nano-pillars, producing a hierarchical 3D structure. Using the SEM images in Figures 3(b-d),
the average nanowire length vs growth time is plotted in Figure 3e. The results show the
nano–wire length increases linearly with growth time, possibly with some delay during
growth initiation. This linear growth trend is promising for controlling the feature size of
these nano-wires and for tuning the void space between the nano-pillars within the
hierarchical nanostructure. It is important to note that the ZnO nano-wires maintained a
constant diameter (32.0±1.5 nm) for reaction time between 60 min and 150 nm. The
preferential growth of ZnO nano-wires in the axial direction also agrees with previous
reports.11
236
Figure 7.3 (a-d) SEM images of ALD ZnO coated nano-pillars (a) before and after (b) 60
min, (c) 120 min, and (d) 150 min of hydrothermal synthesis. (e) ZnO nano-wire length
plotted as a function of hydrothermal reaction time.
Effect of Nano-pillar Height on Photocatalytic Property
Taking advantage of controlled lithography, conformal ALD and well-tuned ZnO
nano-wire length, we fabricated a series of 3D hierarchical nanostructures with various
feature sizes and characterized their photocatalytic performance. More specifically, we
investigated the effect of nano-pillar height for ALD ZnO coated nano-pillars without
nano-wires, and the effect of nano-wires for the hierarchical nanostructures. In the following
discussion, we will refer to ALD coated nano-pillars as “NPxxx/ALD” (xxx is the height of
237
the nano-pillars in nm), while nano-pilar/nano-wire hierarchical structures will be referred to
as “NPxxx/ALD-NWyyy” (yyy is the hydrothermal growth time in min for ZnO nano-wires).
Similarly for the control group, ALD coated Si wafer will be referred to as “Si/ALD”, and
nano-wire decorated Si/ALD will be “Si/ALD-NWyyy”.
Figure 7.4 (a) Schematic of ALD ZnO coated flat Si wafer (control) and nano-pillars with
different heights. (b) Normalized concentration as a function of time during the
photocatalytic dye degradation. (c) Apparent first order rate constant (black square) and
half-life of MO (red circle) as a function of the normalized illuminated surface area.
238
Figure 7.4(a) schematically illustrates the repeating units of the nano-pillars with
different heights. By simply changing the thickness of the positive photoresist layer (Figure
1(a)) and using the same interference lithography process, nano-pillar height can be tuned
between 250 nm and 800 nm. After 200 cycles of ALD ZnO, these nanostructures were
tested for photocatalytic MO degradation.
Figure 7.4(b) shows the normalized MO concentration vs UV exposure time with and
without photocatalyst present. With only MO solution present, the MO concentration
decreases slightly, likely due to UV assisted MO hydrolysis. Adding a planar ZnO layer on Si
produced a small rate enhancement, whereas the nanostructured catalysts (NP250/ALD,
NP500/ALD or NP800/ALD) all showed significant enhancement, with faster rates for the
taller nano-pillars.
To quantify the kinetic data, we used a first-order reaction model for MO degradation to
derive kapp and t1/2. The external geometric surface area of a repeating unit (S, in units of nm2)
under illumination was estimated from the measured feature geometry then normalized it to a
unit flat Si wafer (S0, in units of nm2). Figure 7.4(c) plots kapp from the data fits vs S/S0 and
shows that kapp increases from 0.25 to 0.45 h–1 as S/S0 increases from 1.0 to 4.0. This can be
explained as the illuminated surface area of the nanostructures (or available reaction sites) is
a contributing factor to kapp.21 As t1/2 is inversely proportional to kapp (Equation 4), the
half-life of MO decreases when the surface area of the nanostructure photocatalysts
increases.
239
Effect of Hierarchical Structure on Photocatalytic Reaction Rate
We further investigated how photocatalytic performance can be improved when ZnO
nano-wires are decorated onto the nano-pillars. Figure 5 compares the MO degradation rate
with a group of nano-pillar structures with and without nano-wires. The growth time for the
ZnO nano-wires were controlled at 120 min (~ 105 nm long) for comparison within this
group of samples. kapp and t1/2 are summarized in Table 7.1.
Figure 7.5 Comparison of photocatalytic performance between nanostructure with and
without ZnO nano-wires. (a) MO degradation curves for flat ZnO surface on Si wafer and
ZnO nano-wires on ZnO coated Si wafer. (b-d) MO degradation curves for ALD ZnO coated
nanopillars with and without ZnO nano-wires. The heights of nanopillars in (b-d) are 250 nm,
500 nm and 800 nm, respectively.
240
Figure 7.5(a) shows that ZnO nano-wires improve the photocatalytic reaction kinetics
significantly for ALD ZnO coated flat Si wafer substrates. The apparent first order rate
constant is increased from 0.25 h-1 to 0.37 h-1, as the nano-wires provide more reaction sites
compared to the flat Si/ALD. The kapp of Si/ALD-NW120 lies within the range between
NP500/ALD and NP800/ALD (Table 7.1), indicating the effective surface area of
Si/ALD-NW120 should be in the range of 2.76 ~ 3.81S0. While the external geometric
surface area of the nano-wires is at least one order of magnitude higher than that of flat
Si/ALD surface (S0), the limited diffusion of dye molecules into the nm-scale pores between
the nano-wires hinders the overall MO degradation rate. In addition, the UV intensity decays
exponentially along the axial direction of the nano-wires (parallel to the direction of UV
source) according to the Lambert-Beer Law.21 The effective surface area of the nano-wires is
therefore less than the external geometric surface area.
Figure 7.5(b) compares short ALD-coated nano-pillars (NP250/ALD) with nano-wire
decorated nano-pillar hierarchical nanostructures (NP250/ALD-NW120). Similar to the
comparison between Si/ALD and Si/ALD-NW120, NP250/ALD-NW120 also enables a
faster MO degradation than NP250/ALD, enhancing the kapp from 0.26 h-1 to 0.42 h-1 (Table
7.1). This 61% increase in reaction rate demonstrates that the 3D hierarchical nanostructures
promote the kinetics of photocatalyic MO degradation. Improved photocatalytic activity was
also observed for NP500/ALD-NW120 and NP800/ALD-NW120 (Figure 7.5(c-d)) compared
to the corresponding nano-pillar structures without nano-wires. However, the extent of
increase in kapp decreases to 25% and 8% for NP500/ALD-NW120 and NP800/ALD-NW120,
respectively (Table 7.1). This indicates that effective reaction sites are not significantly
241
improved when the height of these 3D hierarchical nanostructures is increased, possibly a
combined result of diffusion limitation and UV intensity decay from the top of the structure
to the bottom.
We also studied the effect of nano-wire length on photocatalytic property using
nano-pillar substrates with same feature sizes (NP500/ALD). The synthesis time of
nano-wires on NP500/ALD ranged from 60 min to 120 min. The kapp and t1/2 values are
summarized in Table 1. An improved photocatalytic performance is observed when the
nano-wire length is increased. We noticed that the kapp values of NP500/ALD-NW60 and
NP500/ALD-NW90 are slightly less than NP500/ALD. This could be because of the light
scattering and reflection in the hierarchical nanostructure. An optical model will be needed to
understand the UV intensity profile in this type of complex structures.
Table 7.1 Feature sizes and photocatalytic performance of ZnO 3D hierarchical
nanostructrues.
Nano-pillar Nano-wire
Photocatalyst kapp (h-1) t1/2 (h)
Height (nm) Length (nm)
No Catalyst N/A N/A 0.0084 83
Si/ALD 0 0 0.25 2.8
Si/ALD-NW120 0 120 0.37 1.9
NP250/ALD 250 0 0.26 2.6
NP250/ALD-NW120 250 120 0.42 1.6
NP500/ALD 500 0 0.34 2.1
NP500/ALD-NW60 500 60 0.27 2.5
NP500/ALD-NW90 500 90 0.33 2.1
NP500/ALD-NW120 500 120 0.42 1.6
NP800/ALD 800 0 0.45 1.6
NP800/ALD-NW120 800 120 0.48 1.4
242
7.4 Conclusion
In summary, we have demonstrated a new fabrication procedure for synthesizing ZnO
3D hierarchical nanostructures. Using laser interference lithography, we have prepared a
series of nano-pillars with desired heights. Taking advantage of ALD, conformal ZnO thin
films have been deposited onto the nano-pillars in order to both make the surface hydrophilic
and serve as nucleation layers for ZnO nano-wires. We have shown that within the linear
growth window we can easily tune the nano-wire length by adjusting the hydrothermal
synthesis time. We have fabricated a series of hierarchical nanostructures and studied the
effect of nano-pillar height and the effect of hierarchical structure on the photocatalytic
kinetics of MO degradation. Without nano-wires, taller nano-pillars show higher
photocatalytic activity as a result of increased reaction sites. When ZnO nano-wires are
introduced to the system, the apparent first order rate constant can be further increased up to
61% compared to nano-pillars without nano-wires. Limitation of diffusion of dye molecules
into the nm-scale pores and UV light intensity decay in the hierarchical nanostructures affects
the overall photocatalyic reaction rate. Mathematic modeling of mass transport coupled with
reaction kinetics and light intensity profile in the complex structure will be needed in the
future to further understand photocatalysis in hierarchical nanostructures. This work provides
insights for designing and controlling nano-features for enhancing reactivity, which can also
be applied to similar structures for dye sensitized solar cells, nanostructure batteries and
other photocatalytic systems.
243
References
(1) Hoffmann, M. R.; Martin, S. T.; Choi, W.; Bahnemann, D. W. Environmental

Applications of Semiconductor Photocatalysis. Chem. Rev. 1995, 95 (1), 69–96
(2) Khan, S. U. M.; Al-Shahry, M.; Ingler, W. B. Efficient photochemical water splitting
by a chemically modified n-TiO2 2. Science 2002, 297 (5590), 2243–2245.
(3) Maeda, K.; Takata, T.; Hara, M.; Saito, N.; Inoue, Y.; Kobayashi, H.; Domen, K.
GaN:ZnO Solid Solution as a Photocatalyst for Visible-Light-Driven Overall Water
Splitting. J. Am. Chem. Soc. 2005, 127 (23), 8286–8287.
(4) Hernández-Alonso, M. D.; Fresno, F.; Suárez, S.; Coronado, J. M. Development of

alternative photocatalysts to TiO2: Challenges and opportunities. Energy Environ. Sci.
2009, 2 (12), 1231–1257.
(5) Janotti, A.; Van de Walle, C. G. Fundamentals of zinc oxide as a semiconductor. Rep.
Prog. Phys. 2009, 72 (12), 126501.
(6) Akyol, A.; Bayramoglu, M. Preparation and characterization of supported ZnO

photocatalyst by zincate method. J. Hazard. Mater. 2010, 175 (1-3), 484–491.
(7) Daneshvar, N.; Salari, D.; Khataee, A. . Photocatalytic degradation of azo dye acid red
14 in water on ZnO as an alternative catalyst to TiO2. J. Photochem. Photobiol. Chem.
2004, 162 (2–3), 317–322.
(8) Evgenidou, E.; Fytianos, K.; Poulios, I. Photocatalytic oxidation of dimethoate in

aqueous solutions. J. Photochem. Photobiol. Chem. 2005, 175 (1), 29–38.
(9) Hariharan, C. Photocatalytic degradation of organic contaminants in water by ZnO

nanoparticles: Revisited. Appl. Catal. Gen. 2006, 304, 55–61.
(10) Daneshvar, N.; Aber, S.; Seyed Dorraji, M. S.; Khataee, A. R.; Rasoulifard, M. H.
Photocatalytic degradation of the insecticide diazinon in the presence of prepared
nanocrystalline ZnO powders under irradiation of UV-C light. Sep. Purif. Technol.
2007, 58 (1), 91–98.
(11) Gong, B.; Peng, Q.; Na, J.-S.; Parsons, G. N. Highly active photocatalytic ZnO
nanocrystalline rods supported on polymer fiber mats: Synthesis using atomic layer
deposition and hydrothermal crystal growth. Appl. Catal. Gen. 2011, 407 (1–2), 211–
216.
(12) Hung, S.-T.; Chang, C.-J.; Hsu, M.-H. Improved photocatalytic performance of ZnO
244
nanograss decorated pore-array films by surface texture modification and silver
nanoparticle deposition. J. Hazard. Mater. 2011, 198, 307–316.
(13) Xu, F.; Shen, Y.; Sun, L.; Zeng, H.; Lu, Y. Enhanced photocatalytic activity of
hierarchical ZnO nanoplate-nanowire architecture as environmentally safe and facilely
recyclable photocatalyst. Nanoscale 2011, 3 (12), 5020–5025.
(14) Yang, J.; Fei, G. T.; Li, H.; Ouyang, H. Facile Synthesis of 3D Porous Flower-like
ZnO Micro/nanostructure Films and Their Photocatalytic Performance. Chin. J. Chem.
Phys. 2012, 25 (3), 339–344.
(15) Iqbal, D.; Kostka, A.; Bashir, A.; Sarfraz, A.; Chen, Y.; Wieck, A. D.; Erbe, A.
Sequential Growth of Zinc Oxide Nanorod Arrays at Room Temperature via a
Corrosion Process: Application in Visible Light Photocatalysis. Acs Appl. Mater.
Interfaces 2014, 6 (21), 18728–18734.
(16) Lee, H. U.; Park, S. Y.; Lee, S. C.; Seo, J. H.; Son, B.; Kim, H.; Yun, H. J.; Lee, G. W.;
Lee, S. M.; Nam, B.; et al. Highly photocatalytic performance of flexible 3
dimensional (3D) ZnO nanocomposite. Appl. Catal. B Environ. 2014, 144, 83–89.
(17) Colmenares, J. C.; Kuna, E.; Jakubiak, S.; Michalski, J.; Kurzydłowski, K.
Polypropylene nonwoven filter with nanosized ZnO rods: Promising hybrid
photocatalyst for water purification. Appl. Catal. B Environ. 2015, 170–171, 273–282.
(18) Smith, H. I. Low cost nanolithography with nanoaccuracy. Phys. E Low-Dimens. Syst.
Nanostructures 2001, 11, 104–109.
Langmuir 2010, 26 (4), 2550–2558.
(20) Lee, K.; Jur, J. S.; Kim, D. H.; Parsons, G. N. Mechanisms for
hydrophilic/hydrophobic wetting transitions on cellulose cotton fibers coated using
Al2O3 atomic layer deposition. J. Vac. Sci. Technol. A 2012, 30 (1) DOI:
10.1116/1.3671942.
(21) Lee, K.; Losego, M. D.; Kim, D. H.; Parsons, G. N. High performance photocatalytic
metal oxide synthetic bi-component nanosheets formed by atomic layer deposition.
Mater. Horiz. 2014, 1 (4), 419–423.
245

5566 PDF

Uploaded by

Document Information

Original Title

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

5566 PDF

Uploaded by

Copyright:

Available Formats

ABSTRACT

to a wide range of MOFs and fibrous substrates.

space-time-yield of this process is over 3×104 kg∙m-3∙d-1, a significant enhancement for

advancing the industial applications and commertialization of MOFs. New MOF-based

scalable MOF production. Several applications of MOF-functionalized fibrous materials in

synthesis method and growth mechanism of milimeter-scale HKUST-1 [Cu3(BTC)2] single

All Rights Reserved

A dissertation submitted to the Graduate Faculty of

Raleigh, North Carolina

To my parents, Jiannong Zhao and Xueying Wu,

and my fiancée Yizhou Chen

Bachelor of Engineering degree in Chemical Engineering & Techonology as well as a

Biomolecular Engineering at North Carolina State Univeristy. His research focus is on

metal-organic framework (MOF) thin films, MOF-functionalized polymer fibers for

research career as a postdoctoral scholar at Massachusetts Institute of Technology in

A. Khan, Prof. Michael D. Dickey and Prof. Chih-Hao Chang.

Dr. Chih-hao Chang, Abhijeet Bagal and Xu Zhang.

LIST OF TABLES ................................................................................................................................ ix

CHAPTER 1. Introduction – Synthesis, Characterization and

Applications of Metal-Organic Framework Functionalized

Junjie Zhao, Dennis T. Lee, Gregory N. Parsons*

Department of Chemical & Biomolecular Engineering, North Carolina State University,

Raleigh, NC 27695, U.S.A

*Corresponding author: gnp@ncsu.edu

Metal organic frameworks (MOFs) are crystalline porous networks composed of

owing to the advantage of significant synthetic versatility.2,7 Post-synthetic processes have

adsorption of harmful gases,11–13 storage of hydrogen and methane,14,15 and sequestration of

been extended to separations, drug delivery, chemical sensing and catalysis.7

materials and application-orientated devices.17,18 Surface-immobilized MOF thin films can

the applications of MOFs as well.19,20

membrane separations,26–28 catalytic degradation of toxic compounds,29,30 and functional

Figure 1.1 Schematic illustration of the constituents for metal-organic frameworks.

discussion of different synthesis approaches as well as the associated characterization

heterogeneity and crystal density on fiber surfaces observed in electronic microscopic

Figure 1.2 Some common and representative MOF structures.

1.2.1 Solvothermal Synthesis

1.2.1.1 Direct Solvothermal Growth

They investigated the growth of HKUST-1 MOF on chemithermomechanical pulp (CTMP),

synthesis.21 8 mmol of Cu(NO3)2 was dissolved in 50 mL of water, and 8 mmol of H3BTC

study of HKUST-1 growth on COOH-terminated SAMs.33 However, the carboxylic acid

fibers were both low.

direct solvothermal synthesis.21 (Scale bar adapted from Ref. 17)

0.10 mmol/g CO2 uptake at 100 kPa.

1.2.1.2 Solvothermal Synthesis on Chemically Functionalized Surface

Figure 1.4 Carboxymethylation on cellulose for anchoring HKUST-1 MOF coatings.37

Insert image is an EDX spectrum showing the presence of Cu. 37

ZIF-90 thin film is about 5 μm.27

for Torlon/ZIF-90 membrane and the bare Torlon substrate.27

1.2.1.4 Solvothermal Synthesis on Atomic Layer Deposited Nucleation Layers

SEM image of conformal HKUST-1 crystals grown on ALD-coated PP fiber. 23

will be discussed in the later part of this review.

and showed excellent activity comparable to the UiO MOF powders.30

MOFs on nylon nanofibers with ALD TiO2 nucleation layers.30

grown on PA@TiO2 nanofibers.30

1.2.2.1 Direct Layer-by-Layer Synthesis

substrates were sequentially dipped in Cu(OAc)2 aqueous solution and H3BTC in

be an ideal substrate for direct MOF synthesis.

ultrasound-enhanced layer-by-layer method.31

on silk fibers.48 0.033 M of Pb(NO3)2 aqueous solution and 0.033 M of

2,5-bis(4-pyridyl)-3,4-diaza-2,4-hexadiene (4-bpdh) methanol solution were used as the

irradiation enhancement and (c-d) sonication-assisted LbL synthesis .48

without ultrasound irradiation and (b) with sonication. 49

1.2.2.2 Layer-by-Layer Synthesis on Chemically Functionalized Surface