You are on page 1of 53

Electric Double Layer Coupled Oxide

Based Neuromorphic Transistors


Studies Changjin Wan
Visit to download the full and correct content document:
https://textbookfull.com/product/electric-double-layer-coupled-oxide-based-neuromorp
hic-transistors-studies-changjin-wan/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Oxide Based Materials and Structures Fundamentals and


Applications 1st Edition Rada Savkina (Editor)

https://textbookfull.com/product/oxide-based-materials-and-
structures-fundamentals-and-applications-1st-edition-rada-
savkina-editor/

Education, Translation and Global Market Pressures Wan


Hu

https://textbookfull.com/product/education-translation-and-
global-market-pressures-wan-hu/

Omics Applications for Systems Biology Wan Mohd Aizat

https://textbookfull.com/product/omics-applications-for-systems-
biology-wan-mohd-aizat/

Sustainable Electricity Case Studies from Electric


Power Companies in North America 1st Edition Jessica
Fox (Eds.)

https://textbookfull.com/product/sustainable-electricity-case-
studies-from-electric-power-companies-in-north-america-1st-
edition-jessica-fox-eds/
GaN transistors for efficient power conversion Third
Edition De Rooij

https://textbookfull.com/product/gan-transistors-for-efficient-
power-conversion-third-edition-de-rooij/

Forced-Flow Layer Chromatography 1st Edition Tyihák

https://textbookfull.com/product/forced-flow-layer-
chromatography-1st-edition-tyihak/

Quantum Mechanics A Fundamental Approach 1st Edition K.


Kong Wan

https://textbookfull.com/product/quantum-mechanics-a-fundamental-
approach-1st-edition-k-kong-wan/

Electric Circuits Mahmood Nahvi

https://textbookfull.com/product/electric-circuits-mahmood-nahvi/

Nitric Oxide and Signaling in Plants, 1st Edition


Wendehenne

https://textbookfull.com/product/nitric-oxide-and-signaling-in-
plants-1st-edition-wendehenne/
Springer Theses
Recognizing Outstanding Ph.D. Research

Changjin Wan

Electric-Double-Layer
Coupled Oxide-Based
Neuromorphic
Transistors Studies
Springer Theses

Recognizing Outstanding Ph.D. Research


Aims and Scope

The series “Springer Theses” brings together a selection of the very best Ph.D.
theses from around the world and across the physical sciences. Nominated and
endorsed by two recognized specialists, each published volume has been selected
for its scientific excellence and the high impact of its contents for the pertinent field
of research. For greater accessibility to non-specialists, the published versions
include an extended introduction, as well as a foreword by the student’s supervisor
explaining the special relevance of the work for the field. As a whole, the series will
provide a valuable resource both for newcomers to the research fields described,
and for other scientists seeking detailed background information on special
questions. Finally, it provides an accredited documentation of the valuable
contributions made by today’s younger generation of scientists.

Theses are accepted into the series by invited nomination only


and must fulfill all of the following criteria
• They must be written in good English.
• The topic should fall within the confines of Chemistry, Physics, Earth Sciences,
Engineering and related interdisciplinary fields such as Materials, Nanoscience,
Chemical Engineering, Complex Systems and Biophysics.
• The work reported in the thesis must represent a significant scientific advance.
• If the thesis includes previously published material, permission to reproduce this
must be gained from the respective copyright holder.
• They must have been examined and passed during the 12 months prior to
nomination.
• Each thesis should include a foreword by the supervisor outlining the signifi-
cance of its content.
• The theses should have a clearly defined structure including an introduction
accessible to scientists not expert in that particular field.

More information about this series at http://www.springer.com/series/8790


Changjin Wan

Electric-Double-Layer
Coupled Oxide-Based
Neuromorphic Transistors
Studies
The Doctoral Thesis is accepted by the University of
Chinese Academy of Sciences, Beijing, China

123
Author Supervisor
Dr. Changjin Wan Prof. Qing Wan
School of Materials Science and Engineering School of Electronic Science and
Nanyang Technological University Engineering and Collaborative Innovation
Singapore, Singapore Center of Advanced Microstructures
Nanjing University
Nanjing, China

ISSN 2190-5053 ISSN 2190-5061 (electronic)


Springer Theses
ISBN 978-981-13-3313-2 ISBN 978-981-13-3314-9 (eBook)
https://doi.org/10.1007/978-981-13-3314-9

Library of Congress Control Number: 2018961220

© Springer Nature Singapore Pte Ltd. 2019


This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
I do not know what I may appear to the
world, but to myself I seem to have been only
like a boy playing on the seashore, and
diverting myself in now and then finding a
smoother pebble or a prettier shell than
ordinary, whilst the great ocean of truth lay
all undiscovered before me.
Isaac Newton
To my family
Supervisor’s Foreword

A challengeable and rewarding Ph.D. quest should be involved in a topic that is in a


cutting-edge research field with great potential for facilitating the leap beyond
current understanding. Recent decades have witnessed the accelerative and
impressive progresses on digital computer, while they are still suffered from high
energy consumption and low efficiency in dealing with unstructured problems.
Building an electronic brain is daunting but would evoke new computing paradigms
that should complement and/or propose alternative solutions to the classical Von
Neumann/CMOS association. The reemerged attention on neuromorphic engi-
neering that benefit from recent innovative nanotechnologies has motivated Wan
Changjin, to develop transistor-based synaptic devices for synaptic emulations and
to explore the possibility of neuromorphic computing applications by exploiting
them.
Transistors, especially the electric-double-layer (EDL) transistors possess several
advantages as a candidate for building an electronic brain, for example the con-
nectivity. Such transistors could integrate multiple inputs through ionic coupling
evoking a conductance change through the channel, which is inherently similar to
the neurons with massive dendritic terminals. Before fully utilize these properties,
good understanding on the EDL electrostatic coupling as well as electrochemical
properties that underlie synaptic emulations should be studied. This thesis thus
could be divided into two main parts—mechanism studies and applications rooted
on them. In the first part, several of synaptic behaviors involved in memory and
learning have been mimicked, which also propel the modeling of synapse-like
behaviors rooted in the classical theory of EDL transistors. In the second part,
several neuromorphic computing functions were realized such as dendritic inte-
gration, neural arithmetic, and orientation tuning.
In general, Changjin’s thesis opens several new chapters in the exploration of
neuromorphic engineering by exploiting oxide-based EDL transistors. All the phe-
nomena, characterizations and the methodologies are carefully described in the

ix
x Supervisor’s Foreword

thesis. It’s no doubt that this thesis will serve a useful reference both for established
researchers and graduate students working in this cutting-edge area of modern
research.

Nanjing, China Prof. Qing Wan


September 2018
Abstract

Human brain has a massively parallel and reconfigurable architecture with a


complex network of *1011 neurons and *1015 synapses. Such architecture makes
possible more robust, plastic and fault-tolerant learning/memory functions than any
current digital computer. In the neural network, the neuron that receives spike
inputs from thousands of synapses distributed across dendritic trees is often con-
sidered to be the computational engine of the brain. Synaptic plasticity is the
biological process by which specific patterns of synaptic activity result in changes
in synaptic strength and is thought to contribute to learning and memory. Therefore,
emulation of essential synaptic plasticity and computation are viewed as a key step
toward neuromorphic computing. Recently, synaptic electronics which is aimed at
capturing synaptic plasticity and computing power by single electronic device has
aroused widely interesting. At the outset, two-terminal devices such as memristors,
phase change memory, and atom switches, etc, have been explored as the building
blocks of neuromorphic systems. More recently, three-terminal neuromorphic
devices such as ionic/electronic hybrid transistors and ferroelectric transistors have
been demonstrated in the pursuit of the synaptic plasticity and computation in a
single device. The reported three terminal neuromorphic devices reveal alternative
potentials in acting more than a weight tunable connection, which can perform
signal processing/computing serving as synaptic filter, integrator, etc., in neuro-
morphic circuits. The ion-coupled oxide-based electric-double-layer (EDL) tran-
sistors are intrinsically equivalent to the ionic/electronic hybrid transistors, which
could be applied to tune the conducting characteristics of the semiconducting
channel in short-term and long-term, respectively, due to the ion/electron-correlated
electrostatic coupling and electrochemical processes. On this basis, this thesis
focused on the essential synaptic plasticity emulations and neuromorphic com-
puting applications by the ion-coupled oxide-based EDL transistors. The main
content can be summarized from the following aspects:
(1) Preparation and electrical performance research of electrolytes. In this thesis,
nanogranular SiO2, methylcellulose, and graphene oxide were successively
prepared. All of the electrolytes exhibit perfect insulativity. For example the

xi
xii Abstract

maximum leakage current for the nanogranular SiO2 is as low as 0.6 nA.
What’s more, all of the electrolytes are good proton conducting films. The
highest proton conductivity of 4.2  10−4 S/cm was obtained from methyl-
cellulose films. These results indicate the electrolytes are perfect platform for
formation of EDL. Huge EDL capacitances >1 lF/cm2 were observed and the
highest capacitance was measured to 18 lF/cm2 in graphene oxide.
(2) Fabrication and performance characterization of oxide-based EDL transistors.
Oxide-based EDL transistors were fabricated by using the electrolytes men-
tioned in (1) as the gate dielectric. Good transistor performances were achieved
with field-effect mobility higher than 20 cm2 V−1 s−1. The on–off ratio and
subthreshold slope of the oxide-based EDL transistors gated by nanogranular
SiO2 are 2  107 and 114 mV/decade, respectively. The fabrication processes
of the oxide-based EDL transistors gated by Nanogranular SiO2 are compatible
with the CMOS technologies, which indicate that such transistors can have the
great potentials for building neuromorphic circuits/chips. The transistors gated
by graphene oxide or methylcellulose exhibit good transistor performance.
Such transistors deposited on flexible substrates are also demonstrated with
perfect mechanical flexibility. No appreciable degeneration in performance can
be observed in the flexible graphene oxide gated oxide-based EDL transistors
even after bending test for thousands times. In that case, such transistors could
potential applied for large scale flexible neuromorphic circuits.
(3) The short-term behaviors of synapse were successfully mimicked. These
short-term behaviors are excitatory postsynaptic current (EPSC), paired-pulse
facilitation (PPF), spatiotemporal correlated dynamic logic, and short-term
memory. What’s more, a theoretical model for short-term synaptic behaviors
emulations based on EDL modulation and stretched-exponential decay function
was proposed. Such model is consistent well with the experimental results. As
the theoretical model is not limited to the EDL transistors mentioned before,
our results will provide a useful guideline for short-term emulations of other
ion-coupled EDL transistors.
(4) The long-term behaviors of synapse were successfully mimicked. The elec-
trochemical doping/dedoping processes between protons in electrolyte and
electrons in semiconducting channel can be observed under a high gate voltage
(|V| > 4.0 V). The XPS measurements indicate that the oxygen vacancies in
IZO were increased by the electrochemical doping process, which results in a
long-term increase in channel conductance. On the contrary, the electrochem-
ical dedoping process would result in a long-term decrease in channel con-
ductance. On this basis, the long-term behaviors such as spike-timing-
dependent plasticity (STDP), long-term memory, and classical conditioning,
were successfully mimicked.
(5) Dendrite related functions were successfully realized. The huge EDL capaci-
tance is formed at the interface between electrolyte and channel function as a
nanoscale capacitor. Almost all the gate voltage is applied on such capacitor
with nearly no potential difference across the electrolyte. Therefore, multiple
gate inputs could paralleled coupled to channel. Nonlinear dendritic integration
Abstract xiii

functions were realized, which is very similar to the biological experiments.


What’s more, the neural arithmetic behaviors were successfully mimicked by
using additional gate electrode as the modulatory input.
(6) Visual processing functions were successfully realized. In analogy to the
structure of visual neural network, an artificial visual system was built based on
oxide-based EDL transistors. Such visual system shows selectivity in response
to edge orientations around zero degree. The orientation tuning function can be
realized with a full width at half maximum of 44.5°. Then the collision
avoidance behaviors of LGMD neuron were successfully realized based on an
artificial visual system with 20  20 photodetectors and multiple-gate
oxide-based EDL transistors.


Keywords Electric-double-layer (EDL) modulation Oxide-based EDL
 
transistors Neuromorphic transistors Neuromorphic engineering
Acknowledgements

I would firstly like to thank my family, particularly my wife Shu Na who supports
me a lot in both life and research, not only throughout my Ph.D. but at all the times.
I owe a lot to my supervisors, Prof. Wan Qing and Prof. Zhu Liqiang, for all
the invaluable help, concern, and guidance. And I also owe a lot to Prof. Cui Ping,
Prof. Gao Pingqi, Huang Wei, and all the faculties and staffs in NIMTE, for their
kind suggestions and concern. As the same time, I would like to thank the NIMTE
for doctoral training for its training, support, and for providing the basis for many
interesting collaborations.
I also appreciate all group members including Liu Yanghui, Xiao Hui, Liu Ning,
Zhou Jumei, Zhang Hongliang, Guo Liqiang, Chao Jinyu, Liu Lv, Zhu Deming,
Guo Wenhao, Zhang Jin, Guo Zhaojun, Wan Xiang, Zhang Gengming, for many
useful discussions in research as well as lots of activities in life.

xv
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 1
1.2 Brief Introduction of Electric-Double-Layer
Transistors (EDLTs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 2
1.2.1 Overview of EDLTs . . . . . . . . . . . . . . . . . . . . . . ...... 2
1.2.2 Basic Principle of EDL Transistors . . . . . . . . . . . . ...... 3
1.2.3 Gate Dielectric Materials and Applications of EDL
Transistors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 6
1.3 Overview of Neuromorphic Devices . . . . . . . . . . . . . . . . ...... 12
1.3.1 Introduction of Synapses and Their Behaviors . . . . ...... 12
1.3.2 Introduction of Neuromorphic Devices and
Neuromorphic Engineering . . . . . . . . . . . . . . . . . . ...... 15
1.4 Topic Basis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 27
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 29
2 Fabrications and Characterizations of Oxide Based EDL
Transistors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.2 Experimental Materials and Equipment . . . . . . . . . . . . . . . . . . . . 34
2.3 EDL Transistors Gated by Nanogranular SiO2 Electrolyte . . . . . . . 34
2.3.1 Fabrication of Nanogranular SiO2 Electrolyte . . . . . . . . . . 34
2.3.2 Characterizations of Nanogranular SiO2 Electrolyte . . . . . . 37
2.3.3 The Performance of EDLT Gated
by Nanogranular SiO2 . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.4 EDL Transistors Based on Novel Gate Dielectric Materials . . . . . . 41
2.4.1 EDL Transistors Gated by Methylcellulose . . . . . . . . . . . . 41
2.4.2 EDL Transistors Gated by Graphene Oxide . . . . . . . . . . . . 45
2.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

xvii
xviii Contents

3 Oxide Based EDL Transistors for Mimicking Synapse Functions . . . 55


3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.2 Experimental Materials and Equipment . . . . . . . . . . . . . . . . . . . . 56
3.3 Emulations of Short-Term Synaptic Behaviors . . . . . . . . . . . . . . . 57
3.3.1 Emulations and Modulations of Excitatory Postsynaptic
Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 57
3.3.2 Emulations of Paired-Pulse Facilitation . . . . . . . . . . . . . .. 60
3.3.3 Emulations of Spatiotemporal Correlated Dynamic
Logics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.4 Emulations of Long-Term Synaptic Behaviors . . . . . . . . . . . . . . . 63
3.4.1 Emulations of Spike Timing-Dependent Plasticity . . . . . . . 63
3.4.2 Emulations of Memory Behaviors . . . . . . . . . . . . . . . . . . . 69
3.4.3 Emulations of Classical Conditioning . . . . . . . . . . . . . . . . 71
3.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
4 Oxide-Based EDL Transistors for Neuromorphic Computing
Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.2 Experimental Materials and Equipment . . . . . . . . . . . . . . . . . . . . 78
4.3 Nonlinear Dendritic Integration . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.3.1 Dendrite and Dendritic Integration . . . . . . . . . . . . . . . . . . 79
4.3.2 Emulation of Nonlinear Dendritic Integration . . . . . . . . . . 80
4.3.3 Modulation of Nonlinear Dendritic Integration . . . . . . . . . 82
4.3.4 Analog Logic Based on Dendritic Integration . . . . . . . . . . 85
4.4 Neuronal Arithmetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.4.1 Introduction of Neuronal Arithmetic . . . . . . . . . . . . . . . . . 85
4.4.2 Realization of Neuronal Arithmetic . . . . . . . . . . . . . . . . . . 87
4.5 Visual Processing Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.5.1 Introduction of Visual Processing Function . . . . . . . . . . . . 90
4.5.2 Realization of Orientation Tuning Functions . . . . . . . . . . . 91
4.5.3 Realization of Collision Avoidance . . . . . . . . . . . . . . . . . . 92
4.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
5 Conclusion and Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.2 Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Curriculum Vitae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Acronyms

AFM Atomic force microscope


ANN Artificial neural network
CMOS Complementary metal-oxide-semiconductor transistor
CNT Carbon nanotubes
CPU Central Processing Unit
CVD Chemical vapor deposition
DNA Deoxyribonucleic acid
EDL Electric-double-layer
EDLT Electric-double-layer transistor
EGT Electrolyte-gated transistor
EPSP/EPSC Excitatory postsynaptic potential/current
FeFET Ferroelectric-gate field-effect transistor
FET Field-effect transistor
HIFET Hygroscopic insulator field-effect transistor
IBM International Business Machine
IPSP/IPSC Inhibitory postsynaptic potential/current
ISFET Ion-sensitive field-effect transistor
ITO Indium tin oxide
LOD Limit of detection
LTD Long-term depression
LTP Long-term plasticity
LTP Long-term potentiation
OECT Organic electrochemical transistor
PCM Phase change memory
PECVD Plasma enhanced chemical vapor deposition
PET Polyethylene terephthalate
PPF Paired-pulse facilitation
RF Radio frequency
SEM Scanning electron microscopy
STDP Spike-timing-dependent plasticity

xix
xx Acronyms

STP Short-term plasticity


TB Terabyte
TEM Transmission electron microscopy
XPS X-ray photoelectron spectroscopy
Symbols

Ci Specific capacitance
e0 Vacuum permittivity
er Relative dielectric constant
d Thickness
j Dielectric constant
VON Turn-on voltage
VTH Threshold voltage
SS Subthreshold slope
l Carrier mobility
VDS/IDS Voltage/current between drain and source (channel)
VGS Gate voltage
q Quantity of electric charge
jB Boltzmann Constant
T Absolute temperature
lsat Carrier mobility at the saturation region
W Channel width
L Channel length
r Ionic conductivity
sDL Time constant of EDL
s Time constant of the diffusion process
b Extension index depending on the material
A1 Amplitude of the first EPSC peak
A2 Amplitude of the second EPSC peak
A10 Amplitude of the 10th EPSC peak
T Pulse width
F Facilitation ratio
DT Time interval between pulses

xxi
Chapter 1
Introduction

1.1 Background

The human brain is a massively parallel and reconfigurable neural network with
~1011 neurons and ~1015 synapses and operates at a ultralow energy consumption
(<20 W), which is much more robust, plastic, and error-tolerant than any digital com-
puter [1–3]. There are two dominant merits of human brain when compared with dig-
ital computer: (1) massive parallel in structrurally; (2) highly plastic in functionally.
Few decades ago, Carver Mead conceptually proposed the so called neuromorphic
engineering [4]. Neuormorphic engineering is aimed at developing nonbiological
systems that resemble the essential merits of brain, in order to achieve self-learning
and self-cognition computers with ultralow energy consumption ultimately. To real-
ize neuromorphic computing, software and hardware-based approaches have been
proposed. For the former, it’s reported that the IBM team has to employ Blue Gene/P
(BG/P), a super computer equipped with 147 456 CPUs and 144 TB of main mem-
ory, to perform certain cortical simulations at the cat scale even at ~100 times slower
firing rate [5]. It is because the algorithm is essentially run by conventional sequen-
tial machines with limited parallelism for software-based approach. In that case, to
realize neuromorphic computing by using much less resource is considered as a sig-
nificant challenge. Hardware-based approach that physically achieves the massively
parallelism and highly plasticity as the human brain may potentially overcome this
challenge.
Neuron is the basic unit of memory and processing in the brain, and neurons
connect to each other through synapses. Thus, realization of physical devices with
synaptic/neural functions is of great interesting for hardware implementation of the
neuromorphic computation system. Therefore, synaptic electronics [6], a hardware
based approach, has been proposed, which is aimed at developing electronic devices
to capture the essential synaptic plasticity and computing powers. Up to now, a
broad spectrum of electronic devices have been proposed, including two-terminal
synaptic devices such as atom switch, memristor, and phase change memory, and

© Springer Nature Singapore Pte Ltd. 2019 1


C. Wan, Electric-Double-Layer Coupled Oxide-Based Neuromorphic Transistors
Studies, Springer Theses, https://doi.org/10.1007/978-981-13-3314-9_1
2 1 Introduction

three-terminal synaptic devices such as electric-double-layer transistors (EDLTs)


and ferroelectric-gate field-effect transistor (FeFET) [7–23]. Thanks to the achieve-
ments on these synaptic devices, synaptic plasticity such as paired-pulse facilitation
(PPF) and spike-timing-dependent plasticity (STDP), and neuromorphic computing
such as spatiotemporal correlated dynamic logic and pattern recognition, have been
successfully mimicked [7–23].
EDLTs, also called electrolyte-gated transistors (EGTs), are very promising
among various synaptic devices. Previous research on EDLTs spans a spectrum of
issues including flexible electronics, printed electronics and biosensors [24–29]. By
applying a voltage to the gate electrode, the migration and accumulation of ions could
be observed at the gate-electrolyte and semiconductor-electrolyte interfaces, which
will induce the electric-double-layer (EDL) at these interfaces [30–32]. Importantly,
the thickness of the EDL is very small (~1 nm), thus the EDL capacitance could
be very large (>1 μF/cm2 ). In that case, such type transistor could electrostatically
induce a very high density of carrier (>1014 /cm2 ) in the semiconductor channel at a
very low voltage (normally below 3 V). Furthermore, if the semiconducting channel
is permeable, the ions in electrolyte could penetrate into the channel under a relative
high gate voltage, known as the electrochemical doping process [29]. On the con-
trary, we called the electrochemical dedoping process as the return of the ions under
reversed electric-field.
The EDL electrostatic coupling shows a volatile manner and the electrochemi-
cal processes show a nonvolatile manner, which have been proved with the resem-
blance to short-term plasticity (STP) [33, 34] and long-term plasticity (LTP) [15,
16, 23], respectively. More interestingly, transistor-type synaptic devices possess
several merits than two-terminal synaptic devices due to their three-terminal struc-
ture and capacitive coupling working mode. For example, (1) these devices enable
concurrent learning without stopping signal transmission; (2) the synaptic behav-
ior modulation and energy consumption reduction are available by adding another
gate electrodes; (3) artificial neural network could be realized without hard-wired
based on the capacitive coupling [20, 21]. Therefore, EDLTs is of great interesting
in neuromorphic engineering.
This work will start from studies on a series of electrolyte materials and the EDLTs
performance based on these materials. More importantly, synaptic emulations and
neuromorphic computing applications will be discussed based on these EDLTs.

1.2 Brief Introduction of Electric-Double-Layer


Transistors (EDLTs)

1.2.1 Overview of EDLTs

The original EDLTs could date back to 60 years ago when researchers at Bell Labs
employed electrolytes to adjust surface potentials of semiconductors in point contact
1.2 Brief Introduction of Electric-Double-Layer Transistors (EDLTs) 3

devices [35]. Later on, Chao et al. fabricated microelectrochemical transistors that
were employed to amplify small chemical signals and to explore the dependence of
conductivity on electrochemical potential, which could be deemed as the foundation
of EDLTs [36, 37]. Nowadays, EDLTs have drawn a world-wide attention and were
endowed with variety of names [32, 38–43], including electrolyte-gated transistors
(EGTs), organic electrochemical transistors (OECTs), ion-sensitive field-effect tran-
sistors (ISFETs), and hygroscopic insulator field-effect transistors (HIFETs). In this
work, such type of transistors was uniformly named electric-double-layer transistors
(EDLTs).
The utilization of electrolyte is based on two dominant advantages than conven-
tional gate dielectric materials: (1) the huge EDL capacitance and (2) the low tem-
perature fabrication process. Conventional gate dielectric materials, e.g. the SiO2
obtained by thermal oxidation, are good insulator for both electron and ion. The
capacitance between gate electrode and channel could be equivalent to parallel plate
capacitor, the specific capacitance is estimated by the equation: C i  ε0 εr /d, where
ε0 , εr , and d are the vacuum permittivity, relative dielectric constant, and thickness
of the dielectric layer, respectively. Therefore, decreasing the thickness or using
high dielectric constant (high-κ) dielectric materials could significantly increase the
specific capacitance. However, simply decreasing the thickness will increase the
leakage current through the dielectric layer. The operation voltage could be effec-
tively decrease to a low value (≥3 V) by using high-κ dielectric materials such as
zirconia (ZrO2 ) and titanium silicate (TiSiO2 ) [44, 45]. While, it’s still a big chal-
lenge to further increase the specific capacitance. What’s more, the high-temperature
annealing process is always needed for the fabrication of high-κ dielectric materials,
which limits the applications in flexible electronics, printed electronics and so on.
Ideally, electrolyte is an ionic conducting and electronic insulating material. The
removable ions in electrolyte can drift along the electric field and form a dense
electric-double-layer at the electrode/electrolyte interface [30–32]. The schematic
diagram of the EDL structure is shown in Fig. 1.1. At the electrolyte side, ions are
distributed in compact layer and diffuse layer, and the potential drop is mostly applied
on the Helmholtz Plane with thickness of ~1 nm. Such thickness thus enable a huge
specific capacitance of ~1 to 500 μF/cm2 . The ions at the interfaces screen the carriers
with opposite polarity and same concentration in the metal or semiconductor side.
Such strong electrostatic coupling effect could achieve a very high carrier density
of >1014 cm−2 and induce a very large electric field intensity of >1 × 106 V cm−2
[29]. What’s more, most electrolyte materials for gate dielectric could be obtained
at room temperature, e.g. using solution process, which is very suitable for printed
electronics, flexible electronics, and bioelectronics applications [24, 29, 45–51].

1.2.2 Basic Principle of EDL Transistors

EDLTs are one type of field-effect transistors (FETs) [52]. The switch effect of FETs
is realized by the tuning of channel carrier concentration or conductivity. By applying
4 1 Introduction

Fig. 1.1 Schematic diagram


of electric-double-layer

a voltage which vertical to the semiconducting channel, the channel conductivity


could be capacitively modulated through the gate/dielectric/channel capacitor. The
FETs could be divided into N-type and P-type based on the carrier type of the
semiconducting channel (electron for N-type and hole for P-type). In this book, the
N-type FETs are used as the example for working principle demonstration of FETs.
There are several important parameters for FETs such as the turn-on voltage (VON ),
threshold voltage (VTH ), ON–OFF current ratio (ION /IOFF ), subthreshold slope (SS)
and carrier mobility (μ). The precise and detailed definitions of these parameters
will not provide in this book and we simply introduce some essential theories about
the FETs.
We defined the relationship between channel current (IDS ) and swept gate voltage
(VGS ) with a constant drain voltage (VDS ) as transfer characteristics, and the rela-
tionship between channel current and swept drain voltage (VGS ) with different gate
voltage (VGS ) as the output characteristics. Figure 1.2 shows the curve for transfer
characteristics (a) and output characteristics (b), respectively.
We firstly discuss the transfer characteristics of FETs. When the applied gate volt-
age (VGS ) is lower than turn-on voltage (VON ), the carriers are completely depleted,
which induce a very low channel current. In that case, the transistor is on its off
state. When the applied gate voltage is larger than turn-on voltage and lower than the
threshold voltage (VON < VGS < VTH ), the channel can be induced a small number
of carrier, and the IDS at this state is called subthreshold current. At the subthreshold
region, current is dominantly induced by the diffusion of carriers, thus the IDS and
VGS show an exponential relationship:
 
q(VG S − VTH )
I DS ∼ exp (1.1)
kB T
1.2 Brief Introduction of Electric-Double-Layer Transistors (EDLTs) 5

Fig. 1.2 The transfer curves (a) and output curves (b) of the field-effect-transistors

where, κB and T are Boltzmann Constant and absolute temperature, respectively.


When the applied gate voltage is larger than the threshold voltage, the carriers in
channel could be induced and accumulated. Then a positive drain-to-source voltage
(VDS ) will induce a channel current which is similar to applying a voltage on a
resistor. In that case, the IDS is in proportion to VDS , and we call this current region
as linear region. The IDS and VDS thus yield the following equation:
W Ci μ
I DS  (VG S − VTH )VDS (1.2)
L
where, W , L, C i , and μ are the channel width, channel length, specific capacitance
of the gate dielectric, and the field-effect mobility, respectively.
The potential drop between drain and gate will decrease with the increase of VDS .
Therefore, when the VDS reach saturation voltage (VDsat ), the carrier induced by the
gate voltage will be depleted near the drain electrode side. After that, even the VDS
increase to a value larger than VDsat , the current would be almost constant. We define
this region as saturation region and the transition point as pinch-off point. The IDS
versus VDS yield the following equation:
W Ci μsat
I DS  (VG S − Vth )2 (1.3)
2L
where μsat is the carrier mobility at the saturation region. This parameter is a very
important criterion for transistor performance.
6 1 Introduction

Based on the description above, we summarize the essential tips which would be
used in later chapters [52].
a. The turn-on voltage is defined by the voltage when IDS in logarithmic coordinates
significantly increases or the FETs are just completed turned off.
b. The threshold voltage is defined by the lowest voltage for the formation of carrier
accumulation layer at the interface between gate dielectric and channel.
c. The current ON/OFF ratio is defined by the ratio between the ON state current
and OFF state current.
d. The subthreshold slope is the slope of the transfer curve in the plot of logarithmic
at the subthreshold region, which can be calculated by SS  ( dlog I DS
| )−1 .
d VG S max
e. The carrier mobility is the carrier drift velocity under a unit electric field, which
characterizes how quickly an electron (hole) can move through a metal or semi-
conductor.√ The field effect carrier mobility at saturation region is calculated by
d I DS 2
( d VG S )
μsat  1 W .
2 Ci L

There are two working modes for the EDLTs as shown in Fig. 1.3b. The ions in
electrolyte could be driven by electric field and accumulate to the interface between
the channel and electrolyte. For the impenetrable semiconductors (Fig. 1.3b, left
panel), these ions screen the charges in the gate metal and cause accumulation of
carriers in the semiconductor channel formed a layer called the electric-double-layer
(EDL). At steady-state (quasi-static operation), the ionic current from the gate to
source electrodes is near zero ideally and potential drop is almost applied on the EDL
[29]. Such process could be described by the electrostatic modulation as mentioned
before.
While for the penetrable semiconductors, the EDL will form under a low gate
voltage, too. However, the ions could penetrated into the semiconducting channel
and induce carriers in the channel as dopant. Such process is termed electrochemical
doping. Such kind of transistors is also named as electrochemical transistors (ECTs)
[29, 35]. What’s more, when applied a reversed voltage, the ions could be driven out
of the channel, known as the electrochemical dedoping process. Such doping and
dedoping processes would induce a structural change of semiconductor, which may
modify the conducting property of the semiconductor [29, 34].

1.2.3 Gate Dielectric Materials and Applications of EDL


Transistors

The dielectric materials of EDLTs are generally polyelectrolyte [53, 54], ionic liq-
uids [55–57], ion gels [24, 58, 59], inorganic solid electrolytes [60–62], and so on.
One of the earliest studied polyelectrolyte is poly(ethylene oxide) (PEO) doped with
a lithium salt (e.g., LiClO4 ) [63]. More recently, the ionic liquids and ion gels that
possess much higher ionic conductivity, become very promising candidates for EDLs
with high switching speed [55–57, 59]. Besides, ionic liquids and ion gels also exhibit
1.2 Brief Introduction of Electric-Double-Layer Transistors (EDLTs) 7

Fig. 1.3 a Cross section of an EDL transistor. b Two working principles of EDL transistors: elec-
trostatic coupling mode (left) and electrochemical mode (right). Reproduced with permission [29]
Copyright 2013, Wiley-VCH

negligible volatility, non-flammability, exceptional thermal, chemical and electro-


chemical stability. Therefore, these materials are also used as dielectric for abroad
spectrum of applications including dye-sensitized solar cells [64], electromechanical
actuators [65], lithium ion batteries [66], electrochemical capacitors [67]. The high
ionic conductivity is due to the weak electrostatic interaction between ions in the
electrolyte despite of the high ion concentration. Figure 1.4 show the transfer and
output characteristics of an ionic liquid gated organic transistor [58]. The ionic liquid
used in this transistors is polystyrene-poly(methyl methacrylate)-polystyrene (PS-
PMMA-PS). The specific capacitance and ionic mobility are estimated to 30 μF/cm2
and 10−3 S/cm, respectively. Such transistors thus show a high ON/OFF ratio of ~105
and high mobility of >1 cm2 /(V s).
Nanogranular SiO2 is composed of solid nanoscale granular silicon dioxide
with hydrogen, which is deposited by plasma enhanced chemical vapor deposition
(PECVD) method using silane and oxide as reactant gases [60]. Such solid elec-
trolyte proposed by our group in 2009 possesses a high specific capacitance up to 30
μF/cm2 and high ionic conductivity within a range of 10−4 to 0−3 S/cm [60]. The
basic mechanism for ionic conductivity is due to the hoping of hydrogen ion between
hydrogen bonds, as shown in Fig. 1.5 [68].
8 1 Introduction

Fig. 1.4 a The schematic


diagram of EDLT gated by
ionic liquid. The output
characteristics (b) and
transfer characteristics of the
EDLT gated by ionic liquid.
Reproduced with permission
[58]. Copyright 2010,
American Chemical Society

Figure 1.6 shows the ionic conductivity range of some electrolytes. The polar-
ization speed or the EDL formation time of the electrolyte is related to the ionic
conductivity, which can dominate the switching frequency of the devices. Normally,
the polarization speed is higher than microsecond, thus the switching frequency of
1.2 Brief Introduction of Electric-Double-Layer Transistors (EDLTs) 9

Fig. 1.5 The hoping process of the proton between water molecule and hydrogen bond. Reproduced
from [68] with permission from The Royal Society of Chemistry

Fig. 1.6 The ionic conductivity of familiar electrolytes at room temperature. Reproduced with
permission [29] Copyright 2013, Wiley-VCH

the EDLTs is less than megahertz. In this case, the EDLTs is limited for some low fre-
quency applications such as printed electronics [24, 69], flexible electronics [46–48,
50, 70, 71] and sensors [51, 72–75]. These applications will be demonstrated by
several EDLTs as follow.
(1) Printed electronics.
Figure 1.7 shows fabrication process of an all-printed EDLT [69]. Carbon nanotubes
(CNT) networks, ionic gel and PEDOT:PSS utilized as channel, gate dielectric and
gate electrode, respectively, are printed onto polyimide substrate with patterned gold
electrodes in turn through inkjet printing method. The ionic gel possesses a high
specific capacitance of 10 μF/cm2 . The carrier mobility for hole and electron is
10 1 Introduction

Fig. 1.7 The schematic diagram of all printable FET gated by electrolyte. Reproduced with per-
mission [69]. Copyright 2010, American Chemical Society

Fig. 1.8 a Micrograph of the all printable EDLT-based ring oscillator. b The circuit diagram of
the ring oscillator. c The output characteristics of the ring oscillator on SiO2 (top) and polyimide
(bottom) substrates. Reproduced with permission [69]. Copyright 2010, American Chemical Society

30 and 20 cm2 V−1 s−1 , respectively. A printed ring oscillator was fabricated on
polyimide substrate as shown in Fig. 1.8a, and the circuit diagram was shown in
Fig. 1.8b. The output characteristics of the ring oscillator were shown in Fig. 1.8
c. The performance for the oscillator printed on polyimide is very similar to the
oscillator printed on SiO2 substrate. The output frequencies are 2.3 kHz with VDD
 1.5 V and 1.9 kHz with VDD  2.5 V for the former and latter, respectively.
1.2 Brief Introduction of Electric-Double-Layer Transistors (EDLTs) 11

Fig. 1.9 a The schematic diagram and photos of the flexible MoS2 transistor gated by ionic gel on
polyimede substrate. b The transfer characteristics before and after the maximum bending test. c
The drain current and normalized mobility plotted as a function of the curvature radius. Reproduced
with permission [48]. Copyright 2012, American Chemical Society

(2) Flexible electronics.


EDLTs are also widely used in flexible electronics. Figure 1.9 shows a flexible MoS2
transistor gated by ionic gel [48]. The MoS2 active layer were deposited on Polyimide
substrate through chemical vapor deposition (CVD) method. The patterned nickel
electrodes serve as source, drain and gate, and the ionic gel (PS-PMMA-PS) as the
gate dielectric. The transfer curves of the transistor under bent (red line) and after
bent (blue line) were shown in Fig. 1.9b. Negligible degenerations could be observed.
Furthermore, the maximum drain currents and normalized mobility were also shows
a slight decay by increasing the curvature radius. Less than 10% changes could be
obtained when the curvature radius decrease from infinite to 0.75 mm. Therefore,
such devices show very good flexibility by using the electrolyte as the gate dielectric.
Another random document with
no related content on Scribd:
CHAPTER XXVI
SERIOUS THOUGHTS

The young people at The Cedars had taken Garry Knapp right into
the heart of their social life. He knew he was welcome and the
hospitality shown him was a most delightful experience for the young
Westerner.
But “business was business.” He could not see wherein he had
any right to accept a favor from Major Dale because Dorothy wished
her father to aid him. That was not Garry’s idea of a manly part—to
use the father of the girl you love as a staff in getting on in the world.
There was no conceit in Garry’s belief that he had tacit permission,
was it right to accept it, to try to win Dorothy Dale’s heart and hand.
He was just as well assured in his soul that Dorothy had been
attracted to him as he was that she had gained his affection. “Love
like a lightning bolt,” Tavia had called Dorothy’s interest in Garry
Knapp. It was literally true in the young man’s case. He had fallen in
love with Dorothy Dale almost at first sight.
Every time he saw her during that all too brief occasion in New
York his feeling for the girl had grown. By leaps and bounds it
increased until, just as Tavia had once said, if Dorothy had been in
Tavia’s financial situation Garry Knapp would never have left New
York without first learning whether or not there was any possible
chance of his winning the girl he knew he loved.
Now it was revealed to him that he had that chance—and bitterly
did he regret the knowledge. For he gained it at the cost of his peace
of mind.
It is one thing to long for the object forbidden us; it is quite another
thing to know that we may claim that longed-for object if honor did
not interfere. To Garry Knapp’s mind he could not meet what was
Dorothy Dale’s perfectly proper advances, and keep his own self-
respect.
Were he more sanguine, or a more imaginative young man, he
might have done so. But Garry Knapp’s head was filled with hard,
practical common sense. Young men and more often young girls
allow themselves to become engaged with little thought for the
future. Garry was not that kind. Suppose Dorothy Dale did accept his
attentions and was willing to wait for him until he could win out in
some line of industrial endeavor that would afford the competence
that he believed he should possess before marrying a girl used to
the luxuries Dorothy was used to, Garry Knapp felt it would be wrong
to accept the sacrifice.
The chances of business life, especially for a young man with the
small experience and the small capital he would have, were too
great. To “tie a girl up” under such circumstances was a thing Garry
could not contemplate and keep his self-respect. He would not, he
told himself, be led even to admit by word or look that he desired to
be Dorothy’s suitor.
To hide this desire during the few days he remained at The Cedars
was the hardest task Garry Knapp had ever undertaken. If Dorothy
was demure and modest she was likewise determined. Her
happiness, she felt, was at stake and although she could but admire
the attitude Garry held upon this momentous question she did not
feel that he was right.
“Why, what does it matter about money—mere money?” she said
one night to Tavia, confessing everything when her chum had crept
into her bed with her after the lights were out. “I believe I care for
money less than he does.”
“You bet you do!” ejaculated Tavia, vigorously. “Just at present that
young cowboy person is caring more for money than Ananias did.
Money looks bigger to him than anything else in the world. With
money he could have you, Doro Doodlekins—don’t you see?”
“But he can have me without!” wailed Dorothy, burying her head in
the pillow.
“Oh, no he can’t,” Tavia said wisely and quietly. “You know he
can’t. If you could tempt him to throw up his principles in the matter,
you know very well, Doro, that you would be heartbroken.”
“What?”
“Yes you would. You wouldn’t want a young man dangling after
you who had thrown aside his self-respect for a girl. Now, would
you?” And without waiting for an answer she continued: “Not that I
approve of his foolishness. Some men are that way, however. Thank
heaven I am not a man.”
“Oh! I’m glad you’re not, either,” confessed Dorothy with her soft
lips now against Tavia’s cheek.
“Thank you, ma’am. I have often thought I’d like to be of the
hemale persuasion; but never, no more!” declared Tavia, with vigor.
“Suppose I should then be afflicted with an ingrowing conscience
about taking money from the woman I married? Whe-e-e-ew!”
“He wouldn’t have to,” murmured Dorothy, burying her head again
and speaking in a muffled voice. “I’d give up the money.”
“And if he had any sense or unselfishness at all he wouldn’t let you
do that,” snapped Tavia. “No. You couldn’t get along without much
money now, Dorothy.”
“Nonsense——”
“It is the truth. I know I should be hopelessly unhappy myself if I
had to go home and live again just as they do there. I have been
spoiled,” said Tavia, her voice growing lugubrious. “I want wealth—
luxuries—and everything good that money buys. Yes, Doro, when it
comes my time to become engaged, I must get a wealthy man or
none at all. I shall be put up at auction——”
“Tavia! How you talk! Ridiculous!” exclaimed Dorothy. “You talk like
a heathen.”
“Am one when it comes to money matters,” groaned the girl. “I
have got to marry money——”
“If Nat White were as poor as a church mouse, you’d marry him in
a minute!”
“Oh—er—well,” sighed Tavia, “Nat is not going to ask me, I am
afraid.”
“He would in a minute if you’d tell him about those Lance Petterby
letters.”
“Don’t you dare tell him, Dorothy Dale!” exclaimed Tavia, almost in
fear. “You must not. Now, promise.”
“I have promised,” her friend said gloomily.
“And see that you stick to it. I know,” said Tavia, “that I could bring
Nat back to me by explaining. But there should be no need of
explaining. He should know that—that—oh, well, what’s the use of
talking! It’s all off!” and Tavia flounced around and buried her nose in
the pillow.
Dorothy’s wits were at work, however. In the morning she “put a
flea in Ned’s ear,” as Tavia would have said, and Ned hurried off to
the telegraph office to send a day letter to his brother. Dorothy did
not censor that telegraph despatch or this section of it would never
have gone over the wire:

“Come back home and take a squint at the cowboy D.


has picked out for herself.”
CHAPTER XXVII
“IT’S ALL OFF!”

By this time even Ned, dense as he sometimes showed himself to


be, was aware of how things stood between the handsome stranger
from the West and his cousin Dorothy.
Ned’s heart was particularly warm at this juncture. He spent a
good two hours every forenoon writing a long letter to Jennie.
“What under the sun he finds to write about gets me,” declared
Tavia. “He must indite sonnets to her eyebrows or the like. I never
did believe that Ned White would fall so low as to be a poet.”
“Love plays funny tricks with us,” sighed Dorothy.
“Huh!” ejaculated Tavia, wide-eyed. “Do you feel like writing poetry
yourself, Doro Dale? I vum!”
However, to return to Ned, when his letter writing was done he
was at the beck and call of the girls or was off with Garry Knapp for
the rest of the day. Toward Garry he showed the same friendliness
that his mother displayed and the major showed. They all liked the
young man from Desert City; and they could not help admiring his
character, although they could not believe him either wise or just to
Dorothy.
The situation was delicate in the extreme. As Dorothy and Garry
had never approached the subject of their secret attachment for
each other, and now, of course, did not speak of it to the others, not
even Ned could blunder into any opening wherein he might “out with
his opinion” to the Westerner.
Garry Knapp showed nothing but the most gentlemanly regard for
Dorothy. After that first evening on the ice, he did not often allow
himself to be left alone in her company. He knew very well wherein
his own weakness lay.
He talked frankly of his future intentions. It had been agreed
between him and Major Dale that the old Knapp ranch should be
turned over to the Hardin estate lawyers when Garry went back West
at a price per acre that was generous, as Garry said, but not so
much above the market value that he would be “ashamed to look the
lawyers in the face when he took the money.”
Just what Garry would do with these few thousands he did not
know. His education had been a classical one. He had taken up
nothing special save mineralogy, and that only because of Uncle
Terry’s lifelong interest in “prospects.”
“I boned like a good fellow,” he told Ned, “on that branch just to
please the old fellow. Of course, I’d tagged along with him on a burro
on many a prospecting trip when I was a kid, and had learned a lot of
prospector’s lore from the dear old codger.
“But what the old prospector knows about his business is a good
deal like what the old-fashioned farmer knows about growing things.
He does certain things because they bring results, but the old farmer
doesn’t know why. Just so with the old-time prospector. Uncle Terry’s
scientific knowledge of minerals wasn’t a spoonful. I showed him
things that made his eyes bug out—as we say in the West,” and
Garry laughed reminiscently.
“I shouldn’t have thought he’d ever have quarreled with you,” said
Ned, having heard this fact from the girls. “You must have been
helpful to him.”
“That’s the reef we were wrecked on,” said Garry, shaking his
head rather sadly.
“You don’t mean it! How?” queried Ned.
“Why, I’ll tell you. I don’t talk of it much. Of course, you understand
Uncle Terry is one of the old timers. He’s lived a rough life and
associated with rough men for most of it. And his slant on moral
questions is not—well—er—what yours and mine would be, White.”
“I see,” said Ned, nodding. “You collided on a matter of ethics?”
“As you might say,” admitted Garry. “There are abandoned
diggings all over the West, especially where gold was found in rich
deposits that can now be dug over and, by scientific methods, made
to yield comfortable fortunes.
“Why, in the early rush the metal, silver, was not thought of! The
miners cursed the black stuff which got in their way and later proved
to be almost pure silver ore. Other valuable metals were neglected,
too. The miners could see nothing but yellow. They were gold crazy.”
“I see,” Ned agreed. “It must have been great times out there in
those early days.”
“Ha!” exclaimed Garry. “For every ounce of gold mined in the old
times there was a man wasted. The early gold mining cost more in
men than a war, believe me! However, that isn’t the point, or what I
was telling you about.
“Some time after I left the university Uncle Terry wanted me to go
off on a prospecting trip with him and I went—just for the holiday, you
understand. These last few years he hasn’t made a strike. He has
plenty of money, anyway; but the wanderlust of the old prospector
seizes him and he just has to pack up and go.
“We struck Seeper’s Gulch. It was some strike in its day, about
thirty years ago. The gold hunters dug fortunes out of that gulch, and
then the Chinese came in and raked over and sifted the refuse.
You’d think there wasn’t ten cents worth of valuable metal left in that
place, wouldn’t you?”
Ned nodded, keenly interested in the story.
“Well, that’s what the old man thought. He made all kinds of jokes
over a squatter’s family that had picketed there and were digging
and toiling over the played out claims.
“It seemed that they held legal title to a big patch of the gulch.
Some sharper had sawed off the claim on them for good, hard-
earned money; and here they were, broke and desperate. Why!
there hadn’t been any gold mined there for years and years, and
their title, although perfectly legal, wasn’t worth a cent—or so it
seemed.
“Uncle Terry tried to show them that. They were stubborn. They
had to be, you see,” said Garry, shaking his head. “Every hope they
had in the world was right in that God-forsaken gulch.
“Well,” he sighed, “I got to mooning around, impatient to be gone,
and I found something. It was so plain that I wonder I didn’t fall over
it and break my neck,” and Garry laughed.
“What was it? Not gold?”
“No. Copper. And a good, healthy lead of it. I traced the vein some
distance before I would believe it myself. And the bulk of it seemed
to lie right inside the boundaries of that supposedly worthless claim
those poor people had bought.
“I didn’t dare tell anybody at first. I had to figure out how she could
be mined (for copper mining isn’t like washing gold dust) and how
the ore could be taken to the crusher. The old roads were pretty
good, I found. It wouldn’t be much of a haul from Seeper’s Gulch to
town.
“Then I told Uncle Terry—and showed him.”
Ned waited, looking at Garry curiously.
“That—that’s where he and I locked horns,” sighed Garry. “Uncle
Terry was for offering to buy the claim for a hundred dollars. He had
that much in his jeans and the squatters were desperate—meat and
meal all out and not enough gold in the bottom of the pans to color a
finger-ring.”
He was silent again for a moment, and then continued:
“I couldn’t see it. To take advantage of the ignorance of that poor
family wasn’t a square deal. Uncle Terry lost his head and then lost
his temper. To stop him from making any such deal I out with my
story and showed those folks just where they stood. A little money
would start ’em, and I lent them that——”
“But your Uncle Terry?” asked Ned, curiously.
“Oh, he went off mad. I saw the squatters started right and then
made for home. I was some time getting there——”
“You cleaned yourself out helping the owners of the claim?” put in
Ned, shrewdly.
“Why—yes, I did. But that was nothing. I’d been broke before. I got
a job here and there to carry me along. But when I reached home
Uncle Terry had hiked out for Alaska and left a letter with a lawyer for
me. I was the one bad egg in the family,” and Garry laughed rather
ruefully, “so he said. He’d rather give his money to build a
rattlesnake home than to me. So that’s where we stand to-day. And
you see, White, I did not exactly prepare myself for any profession or
any business, depending as I was on Uncle Terry’s bounty.”
“Tough luck,” announced Ned White.
“It was very foolish on my part. No man should look forward to
another’s shoes. If I had gone ahead with the understanding that I
had my own row to hoe when I got through school, believe me, I
should have picked my line long before I left the university and
prepared accordingly.
“I figure that I’m set back several years. With this little bunch of
money your uncle is going to pay me for my old ranch I have got to
get into something that will begin to turn me a penny at once. Not so
easy to do, Mr. White.”
“But what about the folks you steered into the copper mine?”
asked Ned.
“Oh, they are making out fairly well. It was no great fortune, but a
good paying proposition and may keep going for years. Copper is
away up now, you know. They paid me back the loan long ago. But
poor old Uncle Terry—well, he is still sore, and I guess he will remain
so for the remainder of his natural. I’m sorry for him.”
“And not for yourself?” asked Ned, slyly.
“Why, I’d be glad if he’d back me in something. Developing my
ranch into wheat land, for instance. Money lies that way, I believe.
But it takes two or three years to get going and lots of money for
machinery. Can’t raise wheat out there in a small way. It means
tractors, and gangplows and all such things. Whew! no use thinking
of that now,” and Garry heaved a final sigh.
He had not asked Ned to keep the tale to himself; therefore, the
family knew the particulars of Garry Knapp’s trouble with his uncle in
a short time. It was the one thing needed to make Major Dale, at
least, desire to keep in touch with the young Westerner.
“I’m not surprised that he looks upon any understanding with
Dorothy in the way he does,” the major said to Aunt Winnie. “He is a
high-minded fellow—no doubt of it. And I believe he is no namby-
pamby. He will go far before he gets through. I’ll prophesy that.”
“But, my dear Major,” said his sister, with a rather tremulous smile,
“it may be years before such an honorable young man as Garry
Knapp will acquire a competence sufficient to encourage him to
come after our Dorothy.”
“Well—er——”
“And they need each other now,” went on Mrs. White, with
assurance, “while they are young and can get the good of youth and
of life itself. Not after their hearts are starved by long and impatient
waiting.”
“Oh, the young idiot!” growled the major, shaking his head.
Aunt Winnie laughed, although there was still a tremor in her
voice. “You call him high-minded and an idiot——”
“He is both,” growled Major Dale. “Perhaps, to be cynical, one
might say that in this day and generation the two attributes go
together! I—I wish I knew the way out.”
“So do I,” sighed Mrs. White. “For Dorothy’s sake,” she added.
“For both their sakes,” said the major. “For, believe me, this young
man isn’t having a very good time, either.”
Tavia wished she might “cut the Gordian knot,” as she expressed
it. Ned would have gladly shown Garry a way out of the difficulty. And
Dorothy Dale could do nothing!
“What helpless folk we girls are, after all,” she confessed to Tavia.
“I thought I was being so bold, so brave, in getting Garry to come
East. I believed I had solved the problem through father’s aid. And
look at it now! No farther toward what I want than before.”
“Garry Knapp is a—a chump!” exclaimed Tavia, with some heat.
“But a very lovable chump,” added Dorothy, smiling patiently. “Oh,
dear! It must be his decision, not mine, after all. I tell you, even the
most modern of girls are helpless in the end. The man decides.”
Nat came back to North Birchland in haste. It needed only a word
—even from his brother—to bring him. Perhaps he would have met
Tavia as though no misunderstanding had arisen between them had
she been willing to ignore their difficulty.
But when he kissed Dorothy and his mother, and turned to Tavia,
she put out her hand and looked Nat sternly in the eye. He knew
better than to make a joke of his welcome home with her. She had
raised the barrier herself and she meant to keep it up.
“The next time you kiss me it must be in solemn earnest.”
She had said that to Nat and she proposed to abide by it. The old,
cordial, happy-go-lucky comradeship could never be renewed. Nat
realized that suddenly and dropped his head as he went indoors with
his bag.
He had returned almost too late to meet Garry Knapp after all. The
Westerner laughingly protested that he had loafed long enough. He
had to run down to New York for a day or so to attend to some
business for Bob Douglas and then must start West.
“Come back here before you really start for the ‘wild and woolly,’”
begged Ned. “We’ll get up a real house party——”
“Tempt me not!” cried Garry, with hand raised. “It is hard enough
for me to pull my freight now. If I came again I’d only have to—well! it
would be harder, that’s all,” and his usually hopeful face was
overcast.
“Remember you leave friends here, my boy,” said the major, when
he saw the young man alone the evening before his departure.
“You’ll find no friends anywhere who will be more interested in your
success than these at The Cedars.”
“I believe you, Major. I wish I could show my appreciation of your
kindness in a greater degree by accepting your offer to help me. But
I can’t do it. It wouldn’t be right.”
“No. From your standpoint, I suppose it wouldn’t,” admitted the
major, with a sigh. “But at least you’ll correspond——”
“Ned and I are going to write each other frequently—we’ve got
quite chummy, you know,” and Garry laughed. “You shall all hear of
me. And thank you a thousand times for your interest Major Dale!”
“But my interest hasn’t accomplished what I wanted it to
accomplish,” muttered the old gentleman, as Garry turned away.
Dorothy showed a brave face when the time came for Garry’s
departure. She did not make an occasion for seeing him alone, as
she might easily have done. Somehow she felt bound in honor—in
Garry’s honor—not to try to break down his decision. She knew he
understood her; and she understood Garry. Why make the parting
harder by any talk about it?
But Tavia’s observation as Garry was whirled away by Ned in the
car for the railway station, sounded like a knell in Dorothy Dale’s
ears.
“It’s all off!” remarked Tavia.
CHAPTER XXVIII
THE CASTAWAYS

Drifts covered the fences and fitted every evergreen about The
Cedars with a white cap. The snow had come quite unexpectedly
and in the arms of a blizzard.
For two days and nights the storm had raged all over the East.
Wires were down and many railroad trains were blocked. New York
City was reported snowbound.
“I bet old Garry is holed up in the hotel there all right,” said Ned.
“He’d never have got away before the storm.”
Dorothy hoped Garry had not started for the West and had
become snowbound in some train; but she said nothing about it.
It took two full days for the roads to be broken around North
Birchland. And then, of course, to use an automobile was quite
impossible.
The Dale boys were naturally delighted, for there was no school
for several days and snow-caves, snowmen and snow monuments
of all kind were constructed all over the White lawns.
Nor were Joe and Roger alone in these out-of-door activities. The
girls, as well as Ned and Nat, lent their assistance, and Tavia proved
to be a fine snow sculptor.
“Always was. Believe I might learn to work putty and finally
become a great sculptor,” she declared. “At Glenwood they said I
had a talent for composition.”
“What kind of figure do you prefer to sculp, Tavia?” asked Ned,
with curiosity.
“Oh, I think I should just love a job in an ice-cream factory, turning
out works of art for parties and banquets. Or making little figures on
New Year’s and birthday cakes. And then—think of all the nice
‘eats’!”
“Oh! I’d like to do that,” breathed Roger, with round eyes.
“Now, see,” laughed Dorothy, “you have started Roger, perhaps, in
a career. He does love ice-cream and cake.”
At least the joke started something else if it did not point Roger on
the road to fame as an “ice-cream sculptor.” The boy was
inordinately fond of goodies and Tavia promised him a treat just as
soon as ever she could get into town.
A few days before Tavia had been the recipient of a sum of money
from home. When he had any money himself Mr. Travers never
forgot his pretty daughter’s need. He was doing very well in business
now, as well as holding a political position that paid a good salary.
This money she had received was of course burning a hole in Tavia’s
pocket. She must needs get into town as soon as the roads were
passable, to buy goodies as her contract with Roger called for.
The horses had not been out of the stable for a week and the
coachman admitted they needed exercise. So he was to drive Tavia
to town directly after breakfast. It was washday, however, and
something had happened to the furnace in the laundry. The
coachman was general handy man about the White premises, and
he was called upon to fix the furnace just as Tavia—and the horses
—were ready.
“But who’ll drive me?” asked Tavia, looking askance at the spirited
span that the boy from the stables was holding. “Goodness! aren’t
they full of ginger?”
“Better wait till afternoon,” advised Dorothy.
“But they are all ready, and so am I. Besides,” said Tavia with a
glance at Roger’s doleful face, “somebody smells disappointment.”
Roger understood and said, trying to speak gruffly:
“Oh, I don’t mind.”
“No. I see you don’t,” Tavia returned dryly, and just then Nat
appeared on the porch in bearskin and driving gloves.
“Get in, Tavia, if you want to go. The horses need the work,
anyway; and the coachman may be all day at that furnace.”
“Oh—I—ah——” began Tavia. Then she closed her lips and
marched down the steps and got into the cutter. Whatever her
feeling about the matter, she was not going to attract everybody’s
attention by backing out.
Nat tucked the robes around her and got in himself. Then he
gathered up the reins, the boy sprang out of the way, and they were
off.
With the runners of the light sleigh humming at their heels the
horses gathered speed each moment. Nat hung on to the reins and
the roses began to blow in Tavia’s cheeks and the fire of excitement
burn in her eyes.
How she loved to travel fast! And in riding beside Nat the pleasure
of speed for her was always doubled. Whether it was in the
automobile, or behind the galloping blacks, as now, to speed along
the highways by Nat’s side was a delight.
The snow was packed just right for sleighing and the wildly excited
span tore into town at racing speed. Indeed, so excited were the
horses that Nat thought it better not to stop anywhere until the
creatures had got over their first desire to run.
So they swept through the town and out upon the road to The
Beeches.
“Don’t mind, do you?” Nat stammered, casting a quick, sidelong
glance at Tavia.
“Oh, Nat! it’s wonderful!” she gasped, but looked straight ahead.
“Good little sport—the best ever!” groaned Nat; but perhaps she
did not hear the compliment thus wrested from him.
He turned into the upper road for The Beeches, believing it would
be more traveled than the other highway. In this, however, he was
proved mistaken in a very few minutes. The road breakers had not
been far on this highway, so the blacks were soon floundering
through the drifts and were rapidly brought down to a sensible pace.
“Say! this is altogether too rough,” Nat declared. “It’s no fun being
tossed about like beans in a sack. I’d better turn ’em around.”
“You’ll tip us over, Nat,” objected Tavia.
“Likely to,” admitted the young man. “So we’d better both hop out
while I perform the necessary operation.”
“Maybe they will get away from you,” she cried with some fear. “Be
careful.”
“Watch your Uncle Nat,” he returned lightly. “I’ll not let them get
away.”
Tavia was the last person to be cautious; so she hopped out into
the snow on her side of the sleigh while Nat alighted on the other. A
sharp pull on the bits and the blacks were plunging in the drift to one
side of the half beaten track. Tavia stepped well back out of the way.
The horses breasted the deep snow, snorting and tossing their
heads. Their spirits were not quenched even after this long and hard
dash from The Cedars.
The sleigh did go over on its side; but Nat righted it quickly. This,
however, necessitated his letting go of the reins with one hand.
The next moment the sleigh came with a terrific shock into
collision with an obstruction. It was a log beside the road, completely
hidden in the snow.
Frightened, the horses plunged and kicked. The doubletree
snapped and the reins were jerked from Nat’s grasp. The horses
leaped ahead, squealing and plunging, tearing the harness
completely from their backs. The sleigh remained wedged behind the
log; but the animals were freed and tore away along the road, back
toward North Birchland.
Tavia had made no outcry; but now, in the midst of the snow cloud
that had been kicked up, she saw that Nat was floundering in the
drift.
“Oh, Nat! are you hurt?” she moaned, and ran to him.
But he was already gingerly getting upon his feet. He had lost his
cap, and the neck of his coat, where the big collar flared away, was
packed with snow.
“Badly hurt—in my dignity,” he growled. “Oh gee, Tavia! Come and
scoop some of this snow out of my neck.”
She giggled at that. She could not help it, for he looked really
funny. Nevertheless she lent him some practical aid, and after he
had shaken himself out of the loose snow and found his cap, he
could grin himself at the situation.
“We’re castaway in the snow, just the same, old girl,” he said.
“What’ll we do—start back and go through North Birchland, the
beheld of all beholders, or take the crossroad back to The Cedars—
and so save a couple of miles?”
“Oh, let’s go home the quickest way,” she said. “I—I don’t want to
be the laughing stock for the whole town.”
“My fault, Tavia. I’m sorry,” he said ruefully.
“No more your fault than it was mine,” she said loyally.
“Oh, yes it was,” he groaned, looking at her seriously. “And it
always is my fault.”
“What is always your fault?” she asked him but tremulously and
stepping back a little.
“Our scraps, Tavia. Our big scrap. I know I ought not to have
questioned you about that old letter. Oh, hang it, Tavia! don’t you see
just how sorry and ashamed I am?” he cried boyishly, putting out
both gloved hands to her.
“I—I know this isn’t just the way to tell you—or the place. But my
heart just aches because of that scrap, Tavia. I don’t care how many
letters you have from other people. I know there’s nothing out of the
way in them. I was just jealous—and—and mean——”
“Anybody tell you why Lance Petterby was writing to me?” put in
Tavia sternly.
“No. Of course not. Hang Lance Petterby, anyway——”
“Oh, that would be too bad. His wife would feel dreadfully if Lance
were hung.”
“What!”
“I knew you were still jealous of poor Lance,” Tavia shot in,
wagging her head. “And that word proves it.”
“I don’t care. I said what I meant before I knew he was married. Is
he?” gasped Nat.
“Very much so. They’ve got a baby girl and I’m its godmother.
Octavia Susan Petterby.”
“Tavia!” Nat whispered still holding out his hands. “Do—do you
forgive me?”
“Now! is this a time or a place to talk things over?” she demanded
apparently inclined to keep up the wall. “We are castaway in the
snow. Bo-o-ooh! we’re likely to freeze here——”
“I don’t care if I do freeze,” he declared recklessly. “You’ve got to
answer me here and now, Tavia.”
“Have I?” with a toss of her head. “Who are you to command me,
I’d like to know?” Then with sudden seriousness and a flood of
crimson in her face that fairly glorified Tavia Travers: “How about that
request I told you your mother must make, Nat? I meant it.”
“See here! See here!” cried the young man, tearing off his gloves
and dashing them into the snow while he struggled to open his
bearskin coat and then the coat beneath.
From an inner pocket he drew forth a letter and opened it so she
could read.
“See!” Nat cried. “It’s from mother. She wrote it to me while I was
in Boston—before old Ned’s telegram came. See what she says
here—second paragraph, Tavia.”
The girl read the words with a little intake of her breath:

“And, my dear boy, I know that you have quarreled in


some way and for some reason with our pretty, impetuous
Tavia. Do not risk your own happiness and hers,
Nathaniel, through any stubbornness. Tavia is worth
breaking one’s pride for. She is the girl I hope to see you
marry—nobody else in this wide world could so satisfy me
as your wife.”

That was as far as Tavia could read, for her eyes were misty. She
hung her head like a child and whispered, as Nat approached:
“Oh, Nat! Nat! how I doubted her! She is so good!”
He put his arms about her, and she snuggled up against the
bearskin coat.
“Say! how about me?” he demanded huskily. “Now that the Widder
White has asked you to be her daughter-in-law, don’t I come into the
picture at all?”
Tavia raised her head, looked at him searchingly, and suddenly
laid her lips against his eager ones.
“You’re—you’re the whole picture for me, Nat!” she breathed.
CHAPTER XXIX
SOMETHING AMAZING

Now that Garry Knapp had left The Cedars—had passed out of
her life forever perhaps—Dorothy Dale found herself in a much
disturbed state of mind. She did not wish to sit and think over her
situation. If she did she knew she would break down.
She was tempted—oh! sorely tempted—to write Garry Knapp all
that was in her heart. Her cheeks burned when she thought of doing
such a thing; yet, after all, she was fighting for happiness and as she
saw it receding from her she grew desperate.
But Dorothy Dale had gone as far as she could. She had done her
best to bring the man she loved into line with her own thought. She
had the satisfaction of believing he felt toward her as she did toward
him. But there matters stood; she could do no more. She did not let
her mind dwell upon this state of affairs; she could not and retain that
calm expected of Dorothy Dale by the rest of the family at The
Cedars. It is what is expected of us that we accomplish, after all. She
had never been in the habit of giving away to her feelings, even as a
schoolgirl. Much more was expected of her now.
The older people about her were, of course, sympathetic. She
would have been glad to get away from them for that very reason.
Whenever Tavia looked at her Dorothy saw commiseration in her
eyes. So, too, with Aunt Winnie and the major. Dorothy turned with
relief to her brothers who had not much thought for anything but fun
and frolic.
Joe and Roger had quite fallen in love with Garry Knapp and
talked a good deal about him. But their talk was innocent enough
and was not aimed at her. They had not discovered—as they had
regarding Jennie Hapgood and Ned—that their big sister was in the
toils of this strange new disease that seemed to have smitten the
young folk at The Cedars.

You might also like