Professional Documents
Culture Documents
Full Download PDF of Dynamics of Molecular Excitons: Theories and Applications (Nanophotonics) 1st Edition Seogjoo J. Jang - Ebook PDF All Chapter
Full Download PDF of Dynamics of Molecular Excitons: Theories and Applications (Nanophotonics) 1st Edition Seogjoo J. Jang - Ebook PDF All Chapter
http://ebooksecure.com/product/ebook-pdf-nonequilibrium-
molecular-dynamics-theory-algorithms-and-applications/
https://ebooksecure.com/download/molecular-dynamics-simulation-
ebook-pdf/
https://ebooksecure.com/download/spectroscopy-and-dynamics-of-
single-molecules-methods-and-applications-ebook-pdf/
https://ebooksecure.com/download/analysis-of-composite-laminates-
theories-and-their-applications-ebook-pdf/
(eBook PDF) Theories of Development: Concepts and
Applications 6th Edition
http://ebooksecure.com/product/ebook-pdf-theories-of-development-
concepts-and-applications-6th-edition/
https://ebooksecure.com/download/theories-of-personality-ebook-
pdf/
https://ebooksecure.com/download/theories-of-development-
concepts-and-applications-ebook-pdf/
http://ebooksecure.com/product/original-pdf-molecular-
biotechnology-principles-and-applications-of-recombinant-dna-5th-
edition/
http://ebooksecure.com/product/ebook-pdf-theories-and-
applications-of-counseling-and-psychotherapy-relevance-across-
cultures-and-settings/
DYNAMICS OF
MOLECULAR EXCITONS
DYNAMICS OF
MOLECULAR EXCITONS
SEOGJOO J. JANG
Queens College, City University of New York
Department of Chemistry and Biochemistry
Queens, NY, 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 © 2020 Elsevier Ltd. 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: 978-0-08-102335-8
Contents
About the author . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv
Chapter 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Motivation and objective. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Frenkel and Wannier excitons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Disorder, fluctuations, and measure of delocalization. . . . . . . . . . . . 16
1.4 Utility and limitations of exciton models . . . . . . . . . . . . . . . . . . . . . . . 18
Chapter 2 Microscopic derivation of Frenkel exciton-bath Hamiltonian. 21
2.1 Aggregates of chromophores. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.2 Aggregates of chromophores embedded in host media . . . . . . . . . 45
2.3 Summary and additional remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Chapter 3 Linear spectroscopy of molecular excitons . . . . . . . . . . . . . . . . 53
3.1 Absorption lineshape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.2 Stimulated emission lineshape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.3 Model calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.4 Summary and additional remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Chapter 4 Exciton transfer rates and hopping dynamics . . . . . . . . . . . . . . . 83
4.1 Transfer between two exciton states: Förster theory’s and its
generalizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.2 Transfer between groups of exciton states. . . . . . . . . . . . . . . . . . . . . . 93
4.3 Master equation approaches and long range exciton hopping
dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
4.4 Summary and additional remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Chapter 5 Quantum dynamics of molecular excitons. . . . . . . . . . . . . . . . . 107
5.1 Projection operator formalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
5.2 Second order approximations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
5.3 Fourth order approximations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
viii Contents
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
About the author xi
Preface
Seogjoo J. Jang
New York
December 6, 2019
Acknowledgments xv
Acknowledgments
I thank the late Bob Silbey who taught me a great deal about
molecular excitons when I was a postdoctoral researcher at MIT.
He also continued inspiring me for many years afterwards as a
colleague. Interactions with him were always special. He could
easily explain important ideas and issues in a few minutes, which
helped me understand their implications so quickly and pursue
new ideas efficiently. Writing this book gave me another chance
to interact with him indirectly through his publications, and I am
still learning from works he has completed decades ago.
Although not directly related to this topic, I am heavily in-
debted to Marshall Newton, one of the best critical thinker and
theoretician I am aware of. His broad knowledge and expertise in
electron transfer theories helped me envision a roadmap for de-
veloping theories of exciton transfer and dynamics, which have
remained relatively unexplored in comparison to the field of elec-
tron transfer theories. Communications and interactions with him
helped me evolve through my early formative period as an inde-
pendent theoretician.
There are also many others that I should not forget mention-
ing names for various contributions related to this book. I am
indebted to my mentors at Seoul National University, Sangyoub
Lee and Kook Joe Shin, who first introduced me to Förster theory
and guided my first research on this topic, and to my PhD mentor
Gregory Voth from whom I learned the fundamentals of quantum
dynamics. I would like to thank Jeff Cina, David Reichman, Chris
Bardeen, Yuan-Chung Cheng, Benedetta Mennucci, Joe Subotnik,
Dorthe Eisele, Greg Scholes, Young Min Rhee, David Coker, Jian-
shu Cao, Jasper Knoester, Graham Fleming, and K. Birgitta Whaley
for discussions and help at various stages leading to this work,
and Frank Spano, Ari Chakraborty, Qiang Shi, and Andrew Mar-
cus for allowing me use their images. Support from National Sci-
ence Foundation, the Department of Energy, and Camille Dreyfus
Foundation have made it possible to continue projects that con-
stitute various parts of this book.
I am grateful to the editor of the Nanophotonics Series, Prof.
David Andrews, and the editorial office of Elsevier for support. Fi-
nally, I cannot thank enough my wife who has been truly support-
ive of writing this book. I am happy to dedicate this book to her.
Seogjoo J. Jang
January 28, 2020
Introduction
1
Contents
1.1 Motivation and objective 1
1.2 Frenkel and Wannier excitons 3
1.2.1 A brief overview 3
1.2.2 Model Hamiltonians for Frenkel excitons 6
1.2.2.1 Linear chain 7
1.2.2.2 Circular chain 11
1.2.3 Semiclassical model Hamiltonians for Wannier excitons 15
1.3 Disorder, fluctuations, and measure of delocalization 16
1.4 Utility and limitations of exciton models 18
data, and are perceived as important objects for a few key applica-
tions. However, while data informing the existence and contribu-
tion of excitons abound, accurate and complete information on
excitons is still difficult to get in general. In particular, accurate
measurement of both spatial patterns and temporal dynamics of
excitons, which is crucial for satisfactory microscopic characteri-
zation, remains a challenging area of research even to date.
The difficulty in detailed microscopic characterization of ex-
citons stems in part from their indefinite spatial extents (in
nanometer length scales) and transient dynamical nature (in time
scales typically as short as tens of femtoseconds). Resolving this
issue requires experimental tools that can simultaneously probe
nanometer length scales at femtosecond time resolution, which
remain challenging despite recent advances. The other reason, al-
though less appreciated, is the fact that exciton has often been
perceived by experimentalists as a loosely defined simple object of
semi-empirical nature. This is in contrast to the current status of
computational molecular quantum mechanics, which has come a
long way of constructing sophisticated methods and rigorous mi-
croscopic formulations that can account for various many-body
quantum effects. The gap between these modern practices of
molecular quantum mechanics and the conceptual basis of most
experimental investigation based on old exciton theories, is often
the source of confusion and miscommunication.
For both practical and fundamental reasons, there is increas-
ing demand for accurate characterization and control of excitons
in complex environments. However, for many interesting systems,
the scale and accuracy of calculations necessary for reliable com-
putation of excitonic properties is in general beyond the reach of
rigorous computational methods at present. Significant theoreti-
cal progress is needed for more accurate and microscopic descrip-
tion of excitons as well as for further advances in experimental
probe techniques. In promoting the theoretical progress, it is im-
portant to establish a general theoretical framework for excitons
that is both rigorous and practical. This is one of the main motiva-
tions for this book.
The first objective of this book is to help consolidating many
versions and ideas of excitons into a single theoretical formalism.
While this is an ambitious objective difficult to achieve at present,
this book can set the stage for initiating a long term effort of such
nature. In addition, many assumptions and approximations clar-
ified in this book will help gain better understanding of the dy-
namics and spectroscopic data. The second objective is to offer
comprehensive description of various theories of exciton dynam-
ics along with their applications. There are multiple approaches
Chapter 1 Introduction 3
possible for achieving this goal. The approach taken in this book
is to begin with molecular Hamiltonians. This approach makes
it straightforward to clarify assumptions and approximations im-
plicit in different forms of excitons from the outset. The formal
foundation for this will be provided in the next chapter. In the
rest of this chapter, for motivational purposes, a brief overview of
Frenkel and Wannier excitons is presented. This is then followed
by introduction of simple model Hamiltonians used for describ-
ing such excitons.
A few words on the convention of notations. For exciton states
and Hamiltonians, the first order quantization and the Dirac no-
tation for states and operators will be used in most cases. The sec-
ond quantization will be primarily reserved only for the descrip-
tion of electromagnetic radiation within the quantum electrody-
namics formalism. When appropriate, semiclassical description
of electromagnetic radiation will also be used for the description
of matter-radiation interaction.
Higher order excitons such as double and triple excitons can also
be defined in a similar manner.
The original formulation of Frenkel exciton [7,9] used a single
orbital to define each state |ej . However, this is not in fact an es-
sential assumption. Each electronic state at molecule j can be any
general molecular electronic state accounting for all many-body
electronic correlations within the molecule as has been suggested
by Agranovich [14]. The only assumption necessary for the con-
struction of Frenkel exciton basis, as will become clear in the next
chapter, is that ej |ek = 0 for j = k. Consequently,
the exciton basis, which is different from the site excitation basis.
The exciton states are fully delocalized and also form a band of vir-
tually continuous states in the limit of large N . This exciton band
is however constructed out of two electron processes and is differ-
ent from the conduction and valance band of the single electron
picture that form the basis of the Wannier excitons as described
below.
The concept of Wannier exciton is built on the band theory
of electronic states in semiconducting/insulating solid materials,
which represent almost continuum states of delocalized electrons
(conduction band) and holes (valance band). Namely, it is based
on the following effective electronic Hamiltonian:
Ĥe = Ec (k)ek† êk + Ev (k)ĥ†k hk
k
1
+ V (p, k, k )êk+p
†
ĥ†p−k ĥp−k êp+k , (1.5)
Nt
p,k,k
Figure 1.1. Illustration of the formation of a Frenkel exciton. Each ellipse with +
and − represents intramolecular excitation, and the whole exciton state (black
line) is formed as a linear combination of these localized excitations. C1 , · · · , CN
are complex coefficients such that |C1 |2 + |C2 |2 + · · · + |CN |2 = 1.
N
N−1
Ĥe = E |sj sj | + J |sj sj +1 | + |sj +1 sj | . (1.6)
j =1 j =1
On the other hand, the highest value of the exciton energy and the
corresponding eigenstate correspond to p = N as follows:
π
EN = E − 2J cos , (1.11)
N +1
8 Chapter 1 Introduction
1/2
N
2 nπ
|ϕN = sin (−1)n+1 |sj . (1.12)
N +1 N +1
n=1
Therefore, the energies of all the exciton states are within a band
of the following width:
π
= 4|J | cos ≤ 4|J | , (1.13)
N +1
μ2
J = J0 + 3
sin θj sin θj +1 cos(φj − φj +1 ) − 2 cos θj cos θj +1 ,
d
(1.15)
(i) In this case where all the transition dipoles are along the
z direction in head-to-tail, Eq. (1.15) simplifies to J = J0 −
2μ2 /(d 3 ). Assume that J0 < 0. Then, J < 0, and Eqs. (1.9)
and (1.10) represent the lowest exciton energy and eigen-
state. The corresponding transition dipole of each exciton
state has zero components along the x and y directions, but
only the following z-component:
N
μ̃p,z = μ ϕp |sj
j =1
1/2
N
2 pj π
=μ sin
N +1 N +1
j =1
(1 − (−1)p ) pπ
=μ √ cot , (1.16)
2(N + 1) 2(N + 1)
where p = 1, · · · , N . In
√ the large N limit and for p << N , this
approaches μ̃p,z ≈ μ 2(N + 1)(1 − (−1)p )/(pπ).
(ii) In this configuration of an alternating head-to-head and
tail-to-tail configuration along the z direction, Eq. (1.15)
simplifies to J = J0 + 2μ2 /(d 3 ). Depending on the sign of
J0 and its magnitude relative to 2μ2 /(d 3 ), the order of exci-
ton energy levels may or may not be the same as case (i). Let
us assume for now that J > 0. Then, Eqs. (1.11) and (1.12)
represent the lowest single exciton energy and eigenstate.
As in case (i), the transition dipole of each exciton state has
zero components along the x and y directions, and has the
10 Chapter 1 Introduction
N
μ̃p,z = μ (−1)j −1 ϕp |sj
j =1
1/2
N
2 pj π
=μ sin + (j − 1)π
N +1 N +1
j =1
(1 − (−1)p ) pπ
=μ √ tan . (1.17)
2(N + 1) 2(N + 1)
Note that this has the same expression as Eq. (1.16) except
that the direction is now along the x-direction. This is ex-
pected because the transition dipoles are parallel in both
cases and exactly the same expression is used for summing
up the contribution to the transition dipole of the exciton
state. However, there is major difference from case (i) in that
the order of exciton state is reversed here, as in case (ii). For
even N , the transition dipole of the lowest exciton state is
zero. For odd N , the transition dipole of the lowest exciton
state is finite but of the minimum magnitude.
√ In the limit of
large odd N , it approaches μ̃N,x ∼ π 2μ/(N + 1)3/2 , which is
vanishingly small. Thus, the lowest single exciton state has
either zero or almost zero transition dipole in this case.
Chapter 1 Introduction 11
Note that this has the same expression as Eq. (1.17) except
that it is now along the x-direction. However, note the dif-
ference of exciton energy ordering in this case. While the
transition√dipole of the lowest exciton state is nonzero, its
value is π 2μ/(N + 1)3/2 , which is vanishingly small.
The details described above for the cases (i) and (ii) are al-
most equivalent except for the phase factor, which can be incor-
porated into the definition of each state |sj . The same is true for
the cases (iii) and (iv). Thus, both the cases (i) and (ii) correspond
to J -aggregates [8,15,16] in the sense that the major oscillator
strength are concentrated on the lowest exciton state. On the other
hand, the cases (iii) and (iv) correspond to H -aggregates [8,15,16]
in the sense the lowest exciton state has virtually zero oscillator
strength. These also serve as counter examples for often referred
simple picture for J - and H aggregates, i.e., the former as result-
ing from “head-to-tail” and the latter from “parallel” configura-
tion of transition dipoles. For example, (ii) is not in “head-to-tail”
but the pattern of transition dipoles will empirically classify it as
J -aggregate. On the other, (iv) is not “parallel” but its pattern of
transition dipoles also makes it an example of H -aggregate.
The summary provided above is of course based on the as-
sumption that the sign of J0 is the same as that of the transition
dipole interaction. If the sign of J0 is different and plays a domi-
nant role, the classification given above should be altered. In addi-
tion, for more general values of θj and φj − φj +1 = δφ, the division
into J or H aggregates will be less clear.
N
N
Ĥe = J (j − k)|sj sk | , (1.20)
j =1 k=1
12 Chapter 1 Introduction
1 2πipj/N
N
|ϕp = √ e |sj , (1.21)
N n=1
1
N N
ϕp |Ĥe |ϕp = sj |Ĥ0 |sk e2πipj /N e−2πip j/N
N
j =1 j =1
1
N N
= J (j − j )e2πip(j −j )/N e2πi(p−p )j/N
N
j =1 j =1
⎛ ⎞
N
1 N
= J (k)e−2πipk/N ⎝ e2πi(p−p )j/N ⎠
N
k=1 j =1
N
= δpp J (k)e−2πipk/N , (1.22)
k=1
where the index k = j − j was introduced in the third line and the
fact that J (j ) = J (j + N ) and e−2πipj/N = e−2πip(j +N )/N was used to
replace j < 0 with N + j in the summation. As a result, the exciton
Hamiltonian and eigenvalue can be expressed as
N −1
Ĥe = Ep |ϕp ϕp | , (1.23)
p=0
N −1
Ep = E + J (k)e−2πipk/N . (1.24)
k=1
are given by
N
1 2πipk/N
N
μ̃p = ϕp |sk μk = √ e μk
k=1
N k=1
√
N
=μ sin θμ eiφμ δp,−1 (1, −i, 0) + e−iφμ δp,1 (1, i, 0)
2√
+ μ Nδp,0 cos θμ (0, 0, 1) . (1.29)
Unlike the case with the linear chain, non-zero values of transition
dipoles in this model are confined to only p = 0 and ±1. Whether
these correspond to low-lying or high-lying exciton states depend
on details such as J0 , θμ , φμ , and N .
14 Chapter 1 Introduction
(ii) In this case of transition dipoles aligned along the radial di-
rection of the circle in the x − y plane, the exciton energy
level is given by
2πp
Ep = E + 2J0 cos
N
2 (3 − cos(2πk/N ))
N−1
μ 2πpk
+ cos .
16R 3 | sin(πk/N )|3 N
k=1
where m∗e and m∗h are effective masses of electron and hole that are
determined by the dispersion relations of respective bands, Ve (r̂e )
and Vh (r̂h ) are effective confining potentials, and is the effective
dielectric constant of the medium, and Vnc (re , rh ) represent other
contributions. Even at this two-particle picture, exact analytic so-
lution is not possible in general.
In the simplest case where the boundary effects and Vnc (re , rh )
can be ignored, Eq. (1.34) reduces to that of hydrogen-like Hamil-
tonian with effective reduced mass μ∗ = m∗e m∗h /(m∗e + m∗h ). How-
ever, this is no longer true if anisotropy due to crystalline environ-
ments need to be taken into consideration. An elegant way to solve
this is to map the Hamiltonian to an isotropic one in fractional-
dimension [17]. Solution of this leads to the following expression
for the energy of bound exciton state:
Ee
En = Eg − , (1.35)
(n + (α − 3)/2)2
1
LIDP R = , (1.36)
p wp j |Ujp |4
where wp is the Boltzman factor for each eigenstate |ϕp and Ujp =
sj |ϕp , which is the component of the unitary transformation be-
tween the exciton state and site excitation state. For the case where
|Ujp |2 = 1/N , which is true for the lowest exciton state of the linear
chain in Sec. 1.2.2.1 and for all exciton states for the circular chain
in Sec. 1.2.2.2, LIDP R = N . For the case where the disorder signif-
icantly lowers the energy of a specific site j such that Ujp = δjj ,
LIDP R = 1. Thus, LIDP R approaches correct values in the two oppos-
ing limits. In between, the value of LIDP R serves as a reasonable
quantitative measure for the average size of Frenkel excitons, but
its exact physical meaning is not always clear.
The measure of delocalization is not in fact unique and there
are many other alternatives. One simple alternative measure is the
following one [19]:
LD =
J DS
wp min 1, N |Ujp |2
. (1.37)
p j
This produces the same values as LIDP R in the two fully localized
and delocalized limits. However, Eqs. (1.36) and (1.37), and also
other alternative measures, in general have different values in the
intermediate situation.
In fact, it is not unexpected that there is no unique value for the
measure of delocalization in the intermediate situation consider-
ing that it is typically used to give qualitative or semi-quantitative
estimate of the size of the exciton. This leaves the possibility for
the definition of a new delocalization length with clear opera-
tional implication in quantum mechanical sense. However, this is
again expected to depend on a specific type of measurement.
In the presence of additional time dependent fluctuations or
dynamical interactions with environments, Eqs. (1.36) and (1.37)
can be used in the same manner. The quantity being averaged over
the static disorder now becomes a dynamical variable in this case
and it offers information on the spatio-temporal evolution of the
extent of delocalization.
For Wannier excitons, the contribution of disorder and fluc-
tuations enter at two stages. The first is at the level of forming
the semiclassical electron and hole quasiparticles near the band
18 Chapter 1 Introduction
Figure 1.5. Examples of exciton hosting systems. (A) Dithiaanthracenophane (DTA) – a dimeric exciton system
(adapted with permission from [20]. © (2010) American Chemical Society); (B) Light harvesting 2 (LH2) complex of
purple bacteria (reprinted with permission from [13]. © (2018) by the American Physical Society); (C) Self-assembled
nanotube consisting of a double layer (adapted with permission from [21]. © (2012) Springer Nature); (D) Dimeric
exciton on a DNA scaffold (adapted with permission from [22]. © (2018) AIP Publishing.); (E) Two-dimensional
herringbone lattice of oligothiophenes (adapted with permission from [23]. © (2007) AIP Publishing); (F) Cd1012 Se1063
quantum dot (adapted with permission from [24]. © (2013) American Chemical Society).
Figure 2.1. The procedure of deriving exciton-bath Hamiltonian starting from the
microscopic molecular Hamiltonian.
Nc
1c
N c N
Ĥc = Ĥj + Ĥj k , (2.1)
2
j =1 j =1 k=j
where, in the first line on the right hand side, T̂ and V̂ represent
kinetic and potential energy operators. The subscripts n and e in
these terms respectively refer to nuclear and electronic degrees of
freedom. For example, T̂j,n represents the kinetic energy opera-
tor of the nuclei of the j th chromophore, and V̂j,nn the nuclear-
nuclear potential energy operator of the j th nuclei. The second
24 Chapter 2 Microscopic derivation of Frenkel exciton-bath Hamiltonian
Figure 2.2. Schematic of an aggregate with 4 chromophore units (Nc = 4). Each
chromophore can be a separate molecule embedded in a host media or part of a
supramolecular system with well defined number of electrons and nuclei.
where, each term with subscript n/e on the right hand side of the
first line represents the interaction potential term between nu-
clei/electrons of the j th chromophore and nuclei/electrons of the
kth chromophore. The expressions in the second and third lines
above again show detailed expressions of all these terms in the
same order. Fig. 2.2 shows a schematic of an aggregate with 4 chro-
mophore units, for which there are 6 distinctive Ĥj k ’s.
Chapter 2 Microscopic derivation of Frenkel exciton-bath Hamiltonian 25
where
Lj
2
Tj,n (∇j ) = − ∇2 , (2.5)
2Mj,l j,l
l=1
where T̂j,e and V̂j,ee have been defined by Eq. (2.2), and
Lj Nj
Zj,l e2
V̂j,en (Rj ) = − , (2.7)
|Rj,l − r̂j,i |
l=1 i=1
which is the same as V̂j,en defined in Eq. (2.2) except that R̂j,l has
been replaced with Rj,l . All the terms presented above are stan-
dard quantum mechanical expressions.
Let us denote the αj th eigenstate of Eq. (2.6) with eigenvalue
Eαj (Rj ) as |Eαj (Rj ), which is an adiabatic electronic state of the
j th chromophore defined at Rj . Thus,
as follows:
Rj |j = Cαj αj (Rj )|Eαj (Rj ) , (2.9)
αj
2
+ 2
α (Rj )∇j,l |Eαj (Rj ) , (2.10)
2Mj,l j
where Uαj (Rj ) is the effective nuclear potential energy for the αj th
adiabatic state of the pigment j defined as
which is the sum of the nuclear potential energy and the elec-
tronic energy for the αj th state. The terms in the third line in
Eq. (2.10) represent intrachromophore nonadiabatic couplings.
For closely packed molecular aggregates and natural light harvest-
ing pigment-protein complexes, displacements of nuclear degrees
of freedom from their equilibrium coordinates are small in gen-
eral. Therefore, the contribution of these nonadiabatic terms are
expected to be insignificant. The success of simple exciton models
in the interpretation of recent experimental results, where these
nonadiabatic terms are not taken into consideration, indeed val-
idate this assumption. At least, such nonadiabatic terms do not
appear to be explicitly needed for modeling the spectroscopy or
exciton dynamics. However, for flexible macromolecules in flexi-
ble environments, there can be significant displacements in nu-
clear degrees of freedom following excitation, making regions of
significant nonadiabatic effects as illustrated in Fig. 2.3 more ac-
cessible. Intra-chromophore nonadiabatic effects in such systems
can influence the formation, energetics, and dynamics of excitons
Chapter 2 Microscopic derivation of Frenkel exciton-bath Hamiltonian 27
Figure 2.3. Illustration of electronic energies for two different electronic states for
one dimensional nuclear coordinate. R0 is the most likely position for the nuclear
coordinate where the non-adiabatic effect is expected to be insignificant. The
region marked by dot-dashed line is the region where nonadiabatic effects can be
significant.
1
+ αj (Rj )Eαj (Rj )| ∇j,l
2
|Eαj (Rj ) . (2.12)
2
− αj (Rj ) ∇j,l Eαj (Rj )| · ∇j,l |Eαj (Rj ) . (2.14)
2
Lj
Sαj αj (Rj ) = ∇j,l Eαj (Rj )| · ∇j,l |Eαj (Rj ) . (2.16)
2Mj,l
l=1
G αj αj ,l (Rj ) = dr ψα∗ (r; Rj )∇j,l ψαj (r; Rj ) , (2.19)
j
Lj
2 ∗
Sαj αj (Rj ) = dr ∇j,l ψα (r; Rj ) · ∇j,l ψαj (r; Rj ) .
2Mj,l j
l=1
(2.20)
In the above expression, note that ∇j,l is the gradient with respect
to Rj , which serves as the parameter of the electronic wavefunc-
tion. For the case where ψαj (r; Rj ) is a real valued function, which
can be assumed to be true for bound states in the absence of mag-
netic field, the combination of Eq. (2.19) and the condition (2.17)
shows that Gαj αj ,l (Rj ) = 0. If not, Gαj αj ,l (Rj ) is not necessarily
zero. On the other hand, Sαj αj (Rj ) is nonzero in general according
to Eq. (2.20). Because the formulation here is for all possible situ-
ations, from now on, I will assume that Gαj αj ,l (Rj ) is also nonzero.
Employing Eqs. (2.15) and (2.16), Eq. (2.14) can be expressed as
Eαj (Rj )|Rj |Ĥj |j = Cαj Tj,n (∇j ) + Uαj (Rj ) αj (Rj )
Lj
2
+ Cαj − Gαj αj ,l (Rj ) · ∇j,l αj (Rj )
αj
2Mj,l
l=1
+ ∇j,l · Gαj αj ,l (Rj )αj (Rj )
Note that Eαj (Rj )|Rj |Ĥj |j = Rj |Eαj (R̂j )|Ĥj |j , where
Eαj (R̂j )| or |Eαj (R̂j ) represents the adiabatic state in the space
of electronic degrees of freedom except that nuclear coordinate
Rj is replaced with R̂j . Employing this in Eq. (2.23) and using the
fact that the equation remains true for arbitrary value of Rj , it sim-
plifies to
Eαj (R̂j )|Ĥj |j = Cαj T̂j,n + Uαj (R̂j ) |αj
Lj
i
+ Cαj − Gαj αj ,l (R̂j ) · P̂j,l + P̂j,l · Gαj αj ,l (R̂j )
α
2Mj,l
j l=1
The above expression is valid only for the order shown on the right
hand side because |Eαj (R̂j ) function as an operator with respect
to the nuclear state.
Since the set of all the eigenstates |Eαj (R̂j )’s form a complete
basis in the electronic space of the chromophore j , in a simi-
lar manner, we can define the identity operator in the electronic
space of the chromophore j as follows:
1̂j,e = |Eαj (R̂j )Eαj (R̂j )| . (2.26)
αj
+ Cαj |Eαj (R̂j ) Fαj αj (R̂j , P̂j ) + Sαj αj (R̂j ) |αj ,
αj αj
(2.27)
where
Lj
i
Fαj αj (R̂j , P̂j ) = − Gαj αj ,l (R̂j ) · P̂j,l + P̂j,l · Gαj αj ,l (R̂j ) .
2Mj,l
l=1
(2.28)
Since G†α α ,l (R̂j ) = −Gαj αj ,l (R̂j ), as can be seen from Eq. (2.17), it
j j
is straightforward to show that Fα† α (R̂j , P̂j ) = Fαj αj (R̂j , P̂j ). From
j j
Eq. (2.18), it is also clear that Sα† α (R̂j ) = Sαj αj (R̂j ).
j j
Eq. (2.27) can now be converted to an operator identity em-
ploying the fact that Eαj (R̂j )|Eαj (R̂j ) = δαj αj . Inserting this iden-
tity along with dummy indices in front of |αj ’s on the right hand
side of Eq. (2.27), we find that
⎛ ⎞
Ĥj ⎝ Cαj |Eαj (R̂j )|αj ⎠
αj
= |Eαj (R̂j ) T̂j,n + Uαj (R̂j ) Eαj (R̂j )|
αj
+ |Eαj (R̂j ) Fαj αj (R̂j , P̂j ) + Sαj αj (R̂j ) Eαj (R̂j )|
αj αj
⎛ ⎞
×⎝ Cαj |Eαj (R̂j )|αj ⎠ . (2.29)
αj
The above equation is true for arbitrary Cαj ’s and |αj ’s. This im-
plies that the Hamiltonian operator for the chromophore j can be
32 Chapter 2 Microscopic derivation of Frenkel exciton-bath Hamiltonian
expressed as:
Ĥj = |Eαj (R̂j ) T̂j,n + Uαj (R̂j ) Eαj (R̂j )|
αj
+ |Eαj (R̂j ) Fαj αj (R̂j , P̂j ) + Sαj αj (R̂j ) Eαj (R̂j )| .
αj αj
(2.30)