Professional Documents
Culture Documents
Textbook Ebook Intelligent Nanotechnology Merging Nanoscience and Artificial Intelligence Yuebing Zheng All Chapter PDF
Textbook Ebook Intelligent Nanotechnology Merging Nanoscience and Artificial Intelligence Yuebing Zheng All Chapter PDF
Edited by
YUEBING ZHENG
Walker Department of Mechanical Engineering and Texas
Materials Institute, The University of Texas at Austin,
Austin, TX, United States
ZILONG WU
Walker Department of Mechanical Engineering and Texas
Materials Institute, The University of Texas at Austin,
Austin, TX, United States
Elsevier
Radarweg 29, PO Box 211, 1000 AE Amsterdam, Netherlands
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
Copyright © 2023 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical,
including photocopying, recording, or any information storage and retrieval system, without permission in writing
from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies
and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency,
can be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as
may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience broaden our
understanding, changes in research methods, professional practices, or medical treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any
information, methods, compounds, or experiments described herein. In using such information or methods they
should be mindful of their own safety and the safety of others, including parties for whom they have a professional
responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for
any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from
any use or operation of any methods, products, instructions, or ideas contained in the material herein.
ISBN: 978-0-323-85796-3
Trevon Badloe
Department of Mechanical Engineering, Pohang University of Science and Technology
(POSTECH), Pohang, Republic of Korea
Yijun Bao
Department of Biomedical Engineering, Duke University, Durham, NC, United States
Keith A. Brown
Department of Mechanical Engineering; Physics Department and Division of Materials Science &
Engineering, Boston University, Boston, MA, United States
Qian Chen
Department of Materials Science and Engineering; Department of Chemistry; Beckman Institute
for Advanced Science and Technology; Materials Research Laboratory, University of Illinois at
Urbana-Champaign, Urbana, IL, United States
Hongrui Cheng
College of Chemistry, Fuzhou University, Fuzhou; Fujian Science & Technology Innovation
Laboratory for Optoelectronic Information, China
Liu Chen-Xu
Beijing Jiaotong University, School of Civil Engineering, Beijing, China
Frank Cichos
Leipzig University, Molecular Nanophotonics Group, Leipzig, Germany
Zhengyang Duan
Department of Electronic Engineering, Tsinghua University, Beijing, China
Zheyu Fang
School of Physics, Academy for Advanced Interdisciplinary Studies, Peking University, Beijing,
China
Paul S. Francis
School of Life and Environmental Sciences, Faculty of Science, Engineering and Built
Environment, Deakin University, Waurn Ponds, VIC, Australia
Sheng Gao
Department of Electronic Engineering, Tsinghua University, Beijing, China
Yi Gao
Department of Electronic Engineering, Tsinghua University, Beijing, China
Yiyang Gong
Department of Biomedical Engineering; Department of Neurobiology, Duke University,
Durham, NC, United States
ix
x Contributors
Aldair E. Gongora
Department of Mechanical Engineering, Boston University, Boston, MA, United States
Grace X. Gu
Mechanical Engineering, University of California, Berkeley, Berkeley, CA, United States
Yu Gui-Lan
Beijing Jiaotong University, School of Civil Engineering, Beijing, China
Che-Lun Hung
Institute of Biomedical Informatics, National Yang Ming Chiao Tung University, Taipei, Taiwan
Santiago Muiños Landin
AIMEN Technology Centre, Smart Systems and Smart Manufacturing–Artificial Intelligence and
Data Analytics Laboratory, Pontevedra, Spain
Chihun Lee
Department of Mechanical Engineering, Pohang University of Science and Technology
(POSTECH), Pohang, Republic of Korea
Fushan Li
Institute of Optoelectronic Technology, Fuzhou University, Fuzhou, China
Yu Li
School of Physics, Academy for Advanced Interdisciplinary Studies, Peking University, Beijing;
Silicon Photonics R&D Center, CUMEC, Chongqing, China
Xing Lin
Department of Electronic Engineering, Tsinghua University, Beijing, China
Zhengchang Liu
School of Physics, Academy for Advanced Interdisciplinary Studies, Peking University, Beijing,
China
Cuicui Lu
Beijing Institute of Technology, Beijing, China
Yongfeng Lu
College of Chemistry, Fuzhou University, Fuzhou; Fujian Science & Technology Innovation
Laboratory for Optoelectronic Information, China
G. Milano
Advanced Materials Metrology and Life Sciences Division, Italian National Institute of Metrology
Research, Torino, Italy
Jaebum Noh
Department of Mechanical Engineering, Pohang University of Science and Technology
(POSTECH), Pohang, Republic of Korea
Ravi Pradip
Universitat Politècnica De Catalunya, Barcelona, Spain
Junsuk Rho
Department of Mechanical Engineering; Department of Chemical Engineering, Pohang
University of Science and Technology (POSTECH), Pohang, Republic of Korea
Contributors xi
C. Ricciardi
Applied Science and Technology Department (DISAT), Politecnico di Torino, Torino, Italy
Verda Saygin
Department of Mechanical Engineering, Boston University, Boston, MA, United States
Kelsey L. Snapp
Department of Mechanical Engineering, Boston University, Boston, MA, United States
Sunae So
Department of Mechanical Engineering; Graduate School of Artificial Intelligence, Pohang
University of Science and Technology (POSTECH), Pohang, Republic of Korea
Zhongwei Xu
Institute of Optoelectronic Technology, Fuzhou University, Fuzhou, China
Charles Yang
Mechanical Engineering, University of California, Berkeley, Berkeley, CA, United States
Lehan Yao
Department of Materials Science and Engineering, University of Illinois at Urbana-Champaign,
Urbana, IL, United States
Hongyi Yuan
Beijing Institute of Technology, Beijing, China
Jooyeong Yun
Department of Mechanical Engineering, Pohang University of Science and Technology
(POSTECH), Pohang, Republic of Korea
Nianen Zhang
Beijing Institute of Technology, Beijing, China
Zhizhou Zhang
Mechanical Engineering, University of California, Berkeley, Berkeley, CA, United States
Yuanhui Zheng
College of Chemistry, Fuzhou University, Fuzhou; Fujian Science & Technology Innovation
Laboratory for Optoelectronic Information, China
Preface
Nanoscience has been an active research field over the past several decades, triggered
by the rising interest in exploring new science and technologies with materials at the
nanoscale. Artificial intelligence (AI) has led to revolutionary development in many
applications such as communication, computer vision, and speech/image recognition.
Bridging the link between nanoscience and AI can provide a new generation of technol-
ogies that will have large impacts on our society. Some of the AI tools are being
implemented by researchers to advance nanoscience and nanotechnology through
AI-enhanced design, manufacturing, measurements, and applications of nanomaterials
and nanodevices. Meanwhile, nanotechnology provides enormous data sets and new
hardware platforms to advance the training of Al models and the implementation of
AI algorithms.
With great contributions from some of the top researchers working in the interdis-
ciplinary field, Intelligent Nanotechnology: Merging Nanoscience and Artificial Intelligence
provides an overview of the advances in science and technology made possible by the
convergence of nanoscience and AI. Section 1 focuses on AI-enhanced design, charac-
terization and manufacturing of nanomaterials, and the use of AI to improve important
material properties, with an emphasis on mechanical, photonic, phononic, electronic,
and magnetic properties. The use of AI in the acquisition and analysis of data in exper-
iments is explored as well. Section 2 includes chapters on AI technologies that have been
enhanced through nanotechnology platforms. For example, memristors, neuromorphic
computing, and artificial neural networks are included. Finally, Section 3 reviews the
advances in applications enabled by the merging of nanotechnology and artificial
intelligence, including examples from biomedicine, chemistry, anticounterfeiting, and
automated research. We hope that this book will provide a bridge to further collabora-
tions among researchers from different fields.
Yuebing Zheng
Zilong Wu
xiii
Acknowledgments
The editors acknowledge the financial support from the National Science Foundation
(NSF-ECCS-2001650) and the National Institute of General Medical Sciences of the
National Institutes of Health (DP2GM128446).
xv
CHAPTER 1
1.1 Gradient-based
Gradient-based optimization techniques rely on the computation of the local gradient of
the objective function in relation to the designated design parameters. This gradient then
guides the optimization algorithm in a certain direction toward a minimum (or maxi-
mum) for the objective function. A problem arises in that there is no guarantee that this
local point of inflection in the gradients of the objective function is the global minimum,
Fig. 1 Visual representations of numerical optimization methods. (A) Gradient descent method. The
blue (gray in the print version) and red (dark gray in the print version) arrows represent optimization
paths from different starting points, leading to different local minima. (B) Genetic algorithm. The
population is made up of N unique solutions to the problem, here with six different parameters.
After each has been given a score and ranked in terms of the fitness function, the best solutions
are used to perform crossover, where the genes of two-parent solutions are spliced together to
create a new child solution. With a specified random chance, some solutions also undergo
mutation to create new solutions. This allows for the optimization to have some chance of
escaping local minima. (C) Particle swarm. Three particles (orange circles (gray in the print version))
with a certain position and momentum explore the parameter space independently and are
updated based on their local information as well as the information of the population as a whole,
to hopefully end in the global minimum, i.e., the optimal solution to the problem. (For
interpretation of the references to color in this figure legend, the reader is referred to the web
version of this article.)
which is, in general, what we wish to uncover to produce the most optimized design. If
the objective function is convex, however, all local minima are also global minima,
meaning that globally optimized solutions can be found. As described in Fig. 1A, after
the gradient has been calculated at a specific location, the optimization algorithm takes
a step in that direction. We need to determine how big of a step to take in the given
direction, and how frequently to measure the gradient. The size of the step is an impor-
tant factor, as too large of a step could lead to overshooting the minimum, while the fre-
quency of measuring the gradient is determined by computational limitations. Of course,
ideally, we would like to know the gradient at every point with infinitesimally small steps
in order to uncover the optimum global solution, but this of obviously unrealistic in prac-
tice. Another important factor in gradient-based optimization methods is the starting
guess in the parameter space. Due to these limitations, it is common to do multiple runs
Inverse design meets nanophotonics 5
with different starting points and step sizes, increasing the computational load. In an
attempt to lighten the computational load on calculating the gradients of the objective
function of high-dimensional parameter’s spaces, the adjoint method has been intro-
duced. This allows the gradients to be calculated using only a singular additional adjoint
simulation, significantly reducing the computational load [1,2].
Fig. 2 Examples of nanophotonic devices designed using the adjoint method. (A) Inverse design of an
achromatic doublet metalens. (i) Schematic of the design. (ii) The widths of the structures that make up
the metalens. (iii) Phase profile across metasurface apertures at λ ¼ 640 nm. (iv) Cross section of focal
plane intensity at all designed wavelengths. (v) The optical efficiency of the metalens. (B) Optical neural
network for handwritten digit classification. (i) Schematic of the design. (ii) Simulated test inputs and
outputs of the optimized design. (iii) Confusion matrix for the MNIST dataset. (iv) The optimized
metasurface designs. (C) A multiwavelength metalens design. (i) Schematic representation of the
forward, (ii) the adjoint simulations in the adjoint optimization technique, and (iii) color maps of
the surface electric current densities used as excitation sources in the adjoint simulations. Panels
A and B are adapted with permission from Mansouree, M., et al., Multifunctional 2.5D metastructures
enabled by adjoint optimization. Optica 7 (2020) 77–84, copyright 2020 Optical Society of America.
Panel C is adapted with permission from Backer, A.S., Computational inverse design for cascaded
systems of metasurface optics. Opt. Express 27 (2019) 30308–30331, copyright 2019 Optical Society of
America.
An FDTD solver was used to simulate the electric fields of plane waves of the desired
wavelengths incident on the metasurface and the corresponding adjoint simulations,
which are defined as the same structure illuminated by sources that are equal to the com-
plex conjugate of the desired output fields (Fig. 2C). This can be understood as propa-
gating the desired output fields back toward the metasurface. Through the calculated
gradients, the dimensions of nanoposts were manipulated to find the optimized param-
eters. This was proved for both two and three layers of metasurfaces, with focusing effi-
ciencies of over 50%. As the fabrication process of multilayer metasurfaces is much more
involved, the performance increase must be worthwhile in practice. Here, the increase of
focusing efficiency from 35% to 48%, to 60% for the single, bilayer, and trilayer meta-
surfaces proved the added complexity of fabrication is indeed worth the extra effort in
order to produce optimized devices.
Expansions of adjoint optimization include using improved update rules [10,11],
automatic differentiation [12], and extension into the nonlinear regime [13]. Limitations
of adjoint optimization come from the discretization of the structure and the meshing
accuracy of the full-wave simulations. Higher accuracy comes with a higher computa-
tional cost for both the forward and adjoint simulations. Although this optimization
technique does indeed simultaneously optimize the whole device at each update step,
full-wave simulations of large-scale devices require servers or clusters with large enough
memory, or other simulation tricks must be utilized, such as making use of symmetry or
simulating a 2D cross section.
Contrary to intuition, it has been shown that random initial geometries produce bet-
ter results than using existing nanophotonic designs as the starting point [26]. TO usually
has no constraints on the constituent parts of the final design, which could possibly lead to
unrealistic designs. For example, if there are voids or floating elements, this may produce
an output that maximizes the objective function, but it would be impossible to realize
experimentally. Therefore, consideration of fabrication feasibility has also become a
key point of improving devices designed using TO [27].
In Ref. [32] the GA was utilized with some modifications in order to design phase-
controlled metasurfaces. In this work, the concept of the crossover was omitted
completely, in its place only mutation ensued. In order to reach an optimal solution,
the amount of mutation was controlled to be smaller and smaller as the iterations were
concluded. An elaborate fitness function was produced to create the exact desired
response for the final solution. Furthermore, physical constraints were included to ensure
that unphysical results were eradicated during the optimization process. This kind of
supervision of the mutation is essential to keep the results in the desired domain. In
another work, strong chiral dichroism in a plasmonic metasurface was demonstrated,
using a micro-genetic algorithm [33]. The key feature of this adaptation is that instead
of having mutation, there is a complete repopulation of the worst-performing solutions
with completely new random solutions for faster convergence. Since the GA has no
bounds in terms of the number of parameters that it can optimize in a single iteration,
it owes well to freeform designs where the surface is discretized into individual pixels
that can be chosen to include or exclude material, similar to the TO method [34].
But since no gradients are needed, there is no limitation on the fitness function to
be differentiable. Here, the surface was discretized into a binary pattern as shown in
Fig. 4A. Circular dichroism of 0.63 and 0.60 was reported for fabricated samples. The
GA has also been used to choose the photonic elements from a library in order to produce
the desired response [35]. The design of the individual structures was done by hand, but
the phase profile was achieved using the optimization power of the GA. Using an adap-
tive GA allows for the problem to be split into subproblems that are solved in sequence, to
reach the final solution [36]. The authors note that the large number of solutions that are
produced with results that are close to the designed fitness function could also be
employed later in data hungry methods such as a deep learning models. In a final example
of using the GA to optimize the performance of nanophotonic applications, it was used to
design perfect absorbers and color filters for single nanostructures and unit cells with mul-
tiple nanostructures (Fig. 4B) [37]. This study showed the limitations that single nano-
particles have when trying to optimize a solution with limited degrees of freedom. The
colors achieved after increasing the unit cell to include multiple structures were greatly
improved, as were the broadband absorption properties. For all of the examples presented
here, it is interesting for researchers to delve into the underlying physics of the optimized
solution to understand why certain designs are better than others, which in turn, helps
researchers to uncover new ideas that can be applied successfully in new applications.
The updates are governed by a learning rate that determines how much the velocity of the
particle is changed at each update. This, along with the number of particles provides the
most important parameters that need to be optimized for each problem and considered
with regard to the available computing power.
In one example of using PSO for nanophotonic applications, Ref. [39] produced a
freestanding metasurface of silicon for high-performance beam steering (Fig. 5A).
A particle swarm algorithm was coupled with a full-wave simulation through the FDTD
method. After optimization, both the forward transmissions were greatly increased, and
the power of the deflected beam in the desired direction was improved dramatically.
However, it was noted that the combination of 3D simulations is a huge bottleneck
in this type of technique, as the total time needed for a single optimization was around
850 h. In another work using the PSO method, a silicon nitride Y-junction was devel-
oped with the goal of converting linearly polarized light into its opposite polarization
(Fig. 5B) [40]. Conversion efficiencies of over 97% for both polarizations were reported
and confirmed with numerical simulations. Many other examples have also been dem-
onstrated, such as waveguides [41], and pixelated surfaces [42] with multiple objective
functions [43].
Fig. 5 Examples of nanophotonic devices designed using particle swarm. (A) (i) Schematic of a
freestanding silicon nanohole array metasurface acting as a beam deflector with a deflection angle θ.
(ii) Top-down view of the silicon nanohole array beam deflector metasurface constructed using
repeating supercells. (B) Conversion efficiency as a function of the length L for the designed mode-
conversion section and the adiabatic taper structure. Panel A is adapted with permission from Ong, J.R.,
Chu, H.S., Chen, V.H., Zhu, A.Y., Genevet, P. Freestanding dielectric nanohole array metasurface for
mid-infrared wavelength applications. Opt. Lett. 42 (2017) 2639–2642, copyright 2017 Optical Society of
America. Panel B is adapted with permission from Zhang, B., et al., Particle swarm optimized polarization
beam splitter using metasurface-assisted silicon nitride Y-junction for mid-infrared wavelengths. Opt.
Commun. 451 (2019) 186–191, copyright 2019 Elsevier.
Another random document with
no related content on Scribd:
"What America owed abroad can never be computed; it is enough
that it reached an enormous sum, to refund which, even under
favorable circumstances, would have taken years of effort;
actually forced payment brought the nation to the brink of a
convulsion. Perhaps no people ever faced such an emergency and
paid, without recourse to war. America triumphed through her
inventive and administrative genius. Brought to a white heat
under compression, the industrial system of the Union suddenly
fused into a homogeneous mass. One day, without warning, the
gigantic mechanism operated, and two hemispheres vibrated with
the shock. In March, 1897, the vast consolidation of mines,
foundries, railroads, and steamship companies, centralized at
Pittsburg, began producing steel rails at $18 the ton, and at
a bound America bestrode the world. She had won her great
wager with fate. … The end seems only a question of time.
Europe is doomed not only to buy her raw material abroad, but
to pay the cost of transport. And Europe knew this
instinctively in March, 1897, and nerved herself for
resistance. Her best hope, next to a victorious war, lay in
imitating America, and in organizing a system of
transportation which would open up the East.
"ARTICLE I.
The High Contracting Parties agree to submit to Arbitration in
accordance with the provisions and subject to the limitations
of this Treaty all questions in difference between them which
they may fail to adjust by diplomatic negotiation.
"ARTICLE II.
All pecuniary claims or groups of pecuniary claims which do
not in the aggregate exceed £100,000 in amount, and which do
not involve the determination of territorial claims, shall be
dealt with and decided by an Arbitral Tribunal constituted as
provided in the next following Article. In this Article and in
Article IV the words 'groups of pecuniary claims' mean
pecuniary claims by one or more persons arising out of the
same transactions or involving the same issues of law and
fact.'
"ARTICLE III.
Each of the High Contracting Parties shall nominate one
arbitrator who shall be a jurist of repute and the two
arbitrators so nominated shall within two months of the date
of their nomination select an umpire. In case they shall fail
to do so within the limit of time above mentioned, the umpire
shall be appointed by agreement between the Members for the
time being of the Supreme Court of the United States and the
Members for the time being of the Judicial Committee of the
Privy Council in Great Britain, each nominating body acting by
a majority. In case they shall fail to agree upon an umpire
within three months of the date of an application made to them
in that behalf by the High Contracting Parties or either of
them, the umpire shall be selected in the manner provided for
in Article X. The person so selected shall be the President of
the Tribunal and the award of the majority of the Members
thereof shall be final.
{578}
"ARTICLE IV.
All pecuniary claims or groups of pecuniary claims which shall
exceed £100,000 in amount and all other matters in difference,
in respect of which either of the High Contracting Parties
shall have rights against the other under Treaty or otherwise,
provided that such matters in difference do not involve the
determination of territorial claims, shall be dealt with and
decided by an Arbitral Tribunal, constituted as provided in
the next following Article.
"ARTICLE V.
Any subject of Arbitration described in Article IV shall be
submitted to the Tribunal provided for by Article III, the
award of which Tribunal, if unanimous, shall be final. If not
unanimous either of the High Contracting Parties may within
six months from the date of the award demand a review thereof.
In such case the matter in controversy shall be submitted to
an Arbitral Tribunal consisting of five jurists of repute, no
one of whom shall have been a member of the Tribunal whose
award is to be reviewed and who shall be selected as follows,
viz:—two by each of the High Contracting Parties, and one, to
act as umpire, by the four thus nominated and to be chosen
within three months after the date of their nomination. In
case they shall fail to choose an umpire within the limit of
time above-mentioned, the umpire shall be appointed by
agreement between the Nominating Bodies designated in Article
III acting in the manner therein provided. In case they shall
fail to agree upon an umpire within three months of the date
of an application made to them in that behalf by the High
Contracting Parties or either of them, the umpire shall be
selected in the manner provided for in Article X. The person
so selected shall be the President of the Tribunal and the
award of the majority of the Members thereof shall be final.
"ARTICLE VI.
Any controversy which shall involve the determination of
territorial claims shall be submitted to a Tribunal composed
of six members, three of whom (subject to the provisions of
Article VIII) shall be Judges of the Supreme Court of the
United States or Justices of the Circuit Courts to be
nominated by the President of the United States, and the other
three of whom (subject to the provisions of Article VIII)
shall be Judges of the British Supreme Court of Judicature or
Members of the Judicial Committee of the Privy Council to be
nominated by Her Britannic Majesty, whose award by a majority
of not less than five to one shall be final. In case of an
award made by less than the prescribed majority, the award
shall also be final unless either Power shall, within three
months after the award has been reported, protest that the
same is erroneous, in which case the award shall be of no
validity. In the event of an award made by less than the
prescribed majority and protested as above provided, or if the
members of the Arbitral Tribunal shall be equally divided,
there shall be no recourse to hostile measures of any
description until the mediation of one or more friendly Powers
has been invited by one or both of the High Contracting
Parties.
"ARTICLE VII.
Objections to the jurisdiction of an Arbitral Tribunal
constituted under this Treaty shall not be taken except as
provided in this Article. If before the close of the hearing
upon a claim submitted to an Arbitral Tribunal constituted
under Article III or Article V either of the High Contracting
Parties shall move such Tribunal to decide, and thereupon it
shall decide that the determination of such claim necessarily
involves the decision of a disputed question of principle of
grave general importance affecting the national rights of such
party as distinguished from the private rights whereof it is
merely the international representative, the jurisdiction of
such Arbitral Tribunal over such claim shall cease and the
same shall be dealt with by arbitration under Article VI.
"ARTICLE VIII.
In cases where the question involved is one which concerns a
particular State or Territory of the United States, it shall
be open to the President of the United States to appoint a
judicial officer of such State or Territory to be one of the
Arbitrators under Article III or Article V or Article VI. In
like manner in cases where the question involved is one which
concerns a British Colony or possession, it shall be open to
Her Britannic Majesty to appoint a judicial officer of such
Colony or possession to be one of the Arbitrators under
Article III or Article V or Article VI.
"ARTICLE IX.
Territorial claims in this Treaty shall include all claims to
territory and all claims involving questions of servitudes,
rights of navigation and of access, fisheries and all rights
and interests necessary to the control and enjoyment of the
territory claimed by either of the High Contracting Parties.
"ARTICLE X.
If in any case the nominating bodies designated in Articles
III and V shall fail to agree upon an Umpire in accordance
with the provisions of the said Articles, the Umpire shall be
appointed by His Majesty the King of Sweden and Norway. Either
of the High Contracting Parties, however, may at any time give
notice to the other that, by reason of material changes in
conditions as existing at the date of this Treaty, it is of
opinion that a substitute for His Majesty should be chosen
either for all cases to arise under the Treaty or for a
particular specified case already arisen, and thereupon the
High Contracting Parties shall at once proceed to agree upon
such substitute to act either in all cases to arise under the
Treaty or in the particular case specified as may be indicated
by said notice; provided, however, that such notice shall have
no effect upon an Arbitration already begun by the constitution
of an Arbitral Tribunal under Article III. The High
Contracting Parties shall also at once proceed to nominate a
substitute for His Majesty in the event that His Majesty shall
at any time notify them of his desire to be relieved from the
functions graciously accepted by him under this Treaty either
for all cases to arise thereunder or for any particular
specified case already arisen.
"ARTICLE XI.
In case of the death, absence or incapacity to serve of any
Arbitrator or Umpire, or in the event of any Arbitrator or
Umpire omitting or declining or ceasing to act as such,
another Arbitrator or Umpire shall be forthwith appointed in
his place and stead in the manner provided for with regard to
the original appointment.
{579}
"ARTICLE XII.
Each Government shall pay its own agent and provide for the
proper remuneration of the counsel employed by it and of the
Arbitrators appointed by it and for the expense of preparing
and submitting its case to the Arbitral Tribunal. All other
expenses connected with any Arbitration shall be defrayed by
the two Governments in equal moieties. Provided, however,
that, if in any case the essential matter of difference
submitted to arbitration is the right of one of the High
Contracting Parties to receive disavowals of or apologies for
acts or defaults of the other not resulting in substantial
pecuniary injury, the Arbitral Tribunal finally disposing of
the said matter shall direct whether any of the expenses of
the successful party shall be borne by the unsuccessful party,
and if so to what extent.
"ARTICLE XIII.
The time and place of meeting of an Arbitral Tribunal and all
arrangements for the hearing and all questions of procedure
shall be decided by the Tribunal itself. Each Arbitral
Tribunal shall keep a correct record of its proceedings and
may appoint and employ all necessary officers and agents. The
decision of the Tribunal shall, if possible, be made within
three months from the close of the arguments on both sides. It
shall be made in writing and dated and shall be signed by the
Arbitrators who may assent to it. The decision shall be in
duplicate, one copy whereof shall be delivered to each of the
High Contracting Parties through their respective agents.
"ARTICLE XIV.
This Treaty shall remain in force for five years from the date
at which it shall come into operation, and further until the
expiration of twelve months after either of the High
Contracting Parties shall have given notice to the other of
its wish to terminate the same.
"ARTICLE XV.
The present Treaty shall be duly ratified by the President of
the United States of America, by and with the advice and
consent of the Senate thereof, and by Her Britannic Majesty;
and the mutual exchange of ratifications shall take place in
Washington or in London within six months of the date hereof
or earlier if possible."
{581}
{582}
"On chinaware the rates of 1890 were restored. The duty on the
finer qualities which are chiefly imported had been lowered to
35 per cent. in 1894, and was now once more put at 60 per
cent. On glassware, also, the general ad-valorem rate, which
had been reduced to 35 per cent. in 1894, was again fixed at
45 per cent., as in 1890. Similarly the specific duties on the
cheaper grades of window-glass and plate-glass, which had been
lowered in 1894, were raised to the figures of 1890. … The metal
schedules in the act of 1897 showed in the main a striking
contrast with the textile schedules. Important advances of
duty were made on many textiles, and in some cases rates went
considerably higher even than those of 1890. But on most
metals, and especially on iron and steel, duties were left
very much as they had been in 1894. … On steel rails there was
even a slight reduction from the rate of 1894—$6.72 per ton
instead of $7.84. On coal there was a compromise rate. The
duty had been 75 cents a ton in 1890, and 40 cents in 1894; it
was now fixed at 67 cents. On the other hand, as to certain
manufactures of iron and steel farther advanced beyond the
crude stage, there was a return to rates very similar to those
of 1890. Thus, on pocket cutlery, razors, guns, we find once
more the system of combined ad-valorem and specific duties,
graded according to the value of the article. … Copper
remained on the free list, where it had been put in 1894. …
For good or ill the copper duty had worked out all its effects
years before. On the other hand, the duties on lead and on
lead ore went up to the point at which they stood in 1890.
Here we have once more the signs of concession to the silver
Republicans of the far West. … The duty on tin plate, a bone
of contention under the act of 1890, was disposed of, with
little debate, by the imposition of a comparatively moderate
duty. …
"A part of the act which aroused much public attention and
which had an important bearing on its financial yield was the
sugar schedule—the duties on sugar, raw and refined. … The act
of 1890 had admitted raw sugar free, while that of 1894 had
imposed a duty of 40 per cent. ad valorem. … The price of raw
sugar had maintained its downward tendency; and the duty of 40
per cent. had been equivalent in 1896 to less than one cent a
pound. In the act of 1897 the duty was made specific, and was
practically doubled. Beginning with a rate of one cent a pound
on sugar tested to contain 75 per cent., it advanced by stages
until on sugar testing 95 per cent. (the usual content of
commercial raw sugar) it reached 1.65 cents per pound. The
higher rate thus imposed was certain to yield a considerable
increase of revenue. Much was said also of the protection now
afforded to the beet sugar industry of the West. That
industry, however, was still of small dimensions and uncertain
future. … On refined sugar, the duty was made 1.95 cents per
pound, which, as compared with raw sugar testing 100 per
cent., left a protection for the domestic refiner,—i. e., for
the Sugar 'Trust,'—of 1/8 of one cent a pound. Some intricate
calculation would be necessary to make out whether this
'differential' for the refining interest was more or less than
in the act of 1894; but, having regard to the effect of the
substitution of specific for ad-valorem duties, the Trust was
no more favored by the act of 1897 than by its predecessor,
and even somewhat less favored. The changes which this part of
the tariff act underwent in the two Houses are not without
significance." In the bill passed by the House. "the so-called
differential, or protection to the refiners, was one-eighth of
a cent per pound. In the Senate there was an attempt at
serious amendment. The influence of the Sugar Trust in the
Senate had long been great. How secured, whether through party
contributions, entangling alliances, or coarse bribery, the
public could not know; but certainly great, as the course of
legislation in that body demonstrated." The Senate attempted