TRANSPARENT LUMINESCENT SOLAR CONCENTRATORS
 
By
 
Yimu Zhao
 
 
 
 
 
 
 
 
 
A THESIS
 
Submitted to
 
Michigan State University
 
in partial fulfillment of the requirements 
 
for the degree of 
 
Chemical Engineering
-
 
Doctor of Philosophy
 
2015 
 
 
 
 
A
BSTRACT
 
TRANSPARENT LUMINESCENT SOLAR CONCENTRATORS
 
By
 
Yimu Zhao
 
    
Integrating solar
-
harvesting systems into the built environment is a transformative 
route to capturing large areas of solar energy, lowering effective solar cell installation 
costs, and improving building efficiency.  Indeed, the idea of luminescent so
lar 
concentrators (LSC), which were first introduced in the 1970s to reduce solar cell costs, 
are now regaining attention as low
-
cost solar harvesting systems to deploy around the 
building envelope. However, the visible absorption and emission of these LSC
s result in 
highly colored systems that hamper their widespread adoptability in many applications, 
including windows. Here, we introduce the concept of transparent luminescent solar 
concentrators (TLSC) that can selectively harvest ultraviolet (UV) and nea
r
-
infrared (NIR) 
without impacting visible light, and which create an entirely new paradigm for power
-
producing transparent surfaces that can be deployed in windows, displays, and beyond.  
In the first configuration, we have designed systems composed of no
vel metal halide 
phosphorescent luminophore blends; these nanoclusters enable selective harvesting of 
UV photons with absorption cutoff positioned at the edge of visible spectrum (430nm) 
and massive
-
downconverted emission in the near
-
infrared (800nm) with 
very high 
quantum yields for luminescence. In the second configuration, we have developed 
transparent luminescent solar concentrators employing fluorescent organic salts with both 
efficient NIR absorption and emission.  We describe the photophysical proper
ties and 
electronic performance of both classes of devices, the impact of ligand
-
host control, and 
architecture optimization. These TLSCs present new opportunities for clear solar
-
harvesting surfaces and windows that can translate into improved building en
ergy 
efficiency, autonomous mobile electronics, and lower cost solar harvesting systems.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
iv
 
 
ACKNOWLEDGEMENTS
 
 
   
     
 
First and foremost I would like to thank my advisor, Dr. Richard Lunt. 
I cannot 
express my appreciation 
for your
 
guidance, 
encouragement and support. 
You 
have 
witnessed and guided my growth over the 
course of this thesis
. 
I could not have asked for 
a more 
perfect advisor
. I 
also 
am appreciative to my thesis committee members Dr. 
Pengpeng Zhang, Dr. Ilsoon Le
e, and Dr. Scott Barton for taking the time to contribute to 
this work. 
 
    
I owe thanks to the MOE members who 
filled 
my PhD life 
with 
wonderful
 
memor
ies
. 
Chris, you helped me in
 
so many ways, 
in research, and in
 
adapting
 
to
 
US life. I
 
a
m 
convinced 
that 
you have a great sense of humor,
 
even thou
gh I sometimes 
miss
 
the
 
punchline
.
 
Peggy, o
ur group is more like a big family because of your devotion. 
Also, I 
owe thanks to Paddy, Pei, Yunhua, Dhanashree, 
Lily, 
Kevin, John, Brian, Juan, Lucas, 
Tyler, Jorge. 
I w
ish everyone great success
 
and 
the 
best of luck
 
in the future. 
 
    
Also, I am thankful to my collaborators,
 
Dr. Baker, Dr.
 
Levine,
 
Dr. Hogan,
 
and their 
students. I 
am 
also thankful
 
to Brian
 
Rook
 
in the Composite Materials & Structures 
Center
, who helped 
me many times
. 
 
I
 
a
m especially grat
eful to my great friends at MSU
: 
Hao, 
Xinting, 
Yan, Peng, Yang, Yi,
 
Miao,
 
Catlin, Dena, Oishi, Chetan, 
Jing, Saisi, Anna. 
I will never forget all the fun 
we have had 

 
all the 
picnic
s
, camping, festival celebration
s
, 
cards
,
 
and 
I will 
never forget how we 
made it 
through these four years. 
 
 
 
v
 
 
    
I am eternal
ly indebted to my parents. Your
 
persistent encouragement, support and 
confidence in me 
always keeps me going
, no matter when you
 
are in the US or China
, no 
matter 
when things go up or down
.
 
I
 
a
m proud to be your pride always. 
 
 
    
Finally, I 
would l
ike to thank Mr. 
G.
 
Chen. 
Because of you, I 
have a deeper love for 
this world.
 
 
 
 
 
 
 
 
 
 
 
vi
 
 
T
ABLE OF CONTENTS
 
 



 
viii
 



.

ix
 
 
Chapter 1 Introduction to Organic and Excitonic Semiconductors


 
 


1
 
 

..............
3
 
 


6
 
       


11
 
       


12
 
       

.............
14
 
Chapter 2 I
ntroduction to Luminescent Solar Concentrators


16
 
 


16
 
 
2.2 Luminophore


.
...
.............
20
 
 


20
 
 
2.4 Overall LSC efficiency


23
 
       


24
 
       


24
 
       


26
 
Chapter 3 Transparent photovoltaics



 
28
 


 
28
 


 
29
 
3.3 Theoretical and practical limits of transparent luminescent solar concentrators33
 
Chapter 4  
Experimental Techniques


 
.
37
 


 
..
37
 

.............
...
39
 


.
..
...
42
 
4.
4  Electrical, J
-


.
...
...
45
 

......
.
.....
...
48
 
Chapter 5 UV
-
absorbing Transparent Luminescent Solar Concentrators
..
. 

52
 
 



.

..
52
 
 


.


53
 
       


.


53
 
vii
 
 
       

.

...
54
 
 

.

...
55
 
 


.
...
...........
....
63
 
        

.



63
 
        

...........
.
....
65
 
        

.



66
 


............
....
66
 
Chapter 6 NIR
-


68
 
 

.


...
68
 
 

.



69
 
       

.
.
...
...........
..
.
69
 
       

.


.
70
 
       


.

..
70
 
 

.

.

..
71
 
 
6.4 Di

.


.


78
 
        

.


.


78
 
        


.

..
79
 
        

.......
..
......
.
.
83
 
 


..

.
85
 
 


.
..
.
.
.
86
 
 

.


89
 
Chapter 7 Transparent Luminescent Solar Concentrators Cost
.



..
.
90
 
  

.

.


.
92
 
  



.
.93
 
Chapter 8 Conclusions and Future Outlooks


...
.
96
 
 

.



96
 
       

.


.
96
 
       

.


.
98
 
       



100
 
       
8.1.4 Transparent mirror design for photon management in TLSCs

.
.
101
 
 
APPEN

3 
 
BIBLIOGR

7
 
 
 
 
 
viii
 
 
L
IST OF TABLES
 
 
Table
 
1.1
 

 
Table 1.2
 

 
Table 1.3
 

 
Table 2.1
 

 
Table 2.2
 

 
Table 7.1
 

 
Table 
A
1 
Quantum yield and absorption of organic luminescent 
materials in 
solution
s


..104
 
Table A
2 
 
Quantum yield and absorption for luminescent materials in films

.
..106
 
 
 
 
 
 
 
 
ix
 
 
L
IST OF FIGURES
 
 
Figure 1.1
 
Structural diagram of several organic compounds, (top from left to right) di
-
indenoperylene (DIP), carbon nanotube, C
60
, (bottom from left to right) ClAlPc,  
naphthalene through pentacene, Ru
-

..

 
Figure 1.2
 
Schematic illustration of a bilayer heterojunctions at open circuit voltage. 
Excitons are created an
d diffuse to the donor/ acceptor interface where the excitons are 
dissociated. The charge carriers are then transported to the electrodes to generate power. 
The work functions for cathode and anode are 

C
 
and 

A
, respectively.
.

..

 
Figure 1.3
 
 
Current density (
J
) 

voltage (
V
) characteristics of a GaAs solar cell. The red 
point is the operating point. The FF is defined by the ratio of 
A
1
/A
2
 
where 
A
1
 
is 
maximized along the 
J
-
V
 



 
Figure 1.4
 
 
Simplified schematic of spins in the ground and excited states 


 
Figure 1.5
 
Jablonski diagram outlining the excitation and emission processes for an 
organic or excitonic molecule 1) excitation to the first excited state; 2) excitation to a 
higher ex
cited state; 3) and 4) vibrational relaxation; 5) non
-
radiative decay to the ground 
state; 6) fluorescent emission; 7) intersystem crossing; and 8) phosphorescent emission
..8
 
Figure 1.6
 
Normalized absorption and emission spectra for a fluorescent emitter (
up) and 
phosphorescent emitter (bottom). The difference between the absorption and emission 
peak is Stokes shift. In the case of the phosphorescent emitter the Stokes shift is 
alternatively referred to as the down
-


 

 
Figur
e 
2.1
 
Photograph of focusing
-
optics concentrators
[
56
]
 
a) parabolic troughs b) 
heliostatic
 




...17
 
Figure 
2.2
  
Schematic of a luminescent solar concentrator composed of luminescent dyes, 
waveguide, and attached PVs.  PVs are typically mounted around every edge, but have 
been omitted
 
from the front and back for schematic clarity.  The component efficiencies 
shown in the configuration are 1) solar spectrum absorption efficiency 2) luminophore 
photoluminescence efficiency (quantum yield efficiency) 3) waveguide efficiency 4) re
-
absorpti
on efficiency 5) solar cell quantum efficiency



...18
 
x
 
 
Figure  2.3
 
Photograph of typical colorful luminescent solar concentrators highlighting 
the waveguided emission that is brightly focused at the edge. (b) Example of a traditional 
LSC in fron
t a window, highlighting the difficulty in installing these systems in a variety 
of applications such as windows.   



 
Figure 2.4
 
Absorption coefficients of standard glass, quartz, BF33 glass, low iron glass, 
polycarbonate (PC) and 
poly(methyl methacrylate) (PMMA) 
[
71
]
 
as a function of 
wavelength.



 
Figure 2.5
 
Photos of cut Si cell (left) and GaAs cell (right) for attaching to the edge of 



 
Figure 2.6
 
Plot of the AM1.5G
 
Solar Cell  Efficiency  as  a  function  of  bandgap (black),  
and  efficiency  normalized  by  the AM1.5 absorption efficiency
 
(gold).
 
Highest      
reported  efficiencies  are highlighted
 
 
in red squar



 
Figure 3.1
 
Overlap of the AM1.5G solar flux and the photonic response of the eye shows 
that about 1/3 of the photon flux is in the visible part of the spectrum with the remaining 



30
 
Figure 3.2
 
Diagram of the point of installation for building glasses with transparent 


.33
 
Figure 3.3
 
Transparent PV theoretical efficiency limits as a function of bandgap and 
level of visible transpa
rency (VT).  Note that these limits are the same for transparent 

.
33
 
Figure 3.4
 
Representative absorption and emission characteristics of TLSC emitters.  
The Stokes shift (
S
) and e
mission width (
W
), two key emitter 
parameters, are 
highlighted. 


.34
 
Figure 3.5
 
TLSC practical efficiency limits with no ideal absorption/emission profiles 
with no spectral overlap as a function of the a) emission width (
W
) with 
S
 
= 0 or b) 
Stokes shift (
S
) with 
W
 
= 0.  Note that the decrease in efficiency stems from a lower 
theoretical voltage and FF from the PV cell mounted around the edge. However, the 
situation in b) has a distinct advantage in practice since it will su
ppress reabsorption 
losses in non
-
ideal absorption emission overlap configurations.   That is, narrow emitters 
will always provide greater flexibility for larger Stokes shifts and minimized reabsorption 



 
xi
 
 
Figu
re 4.1
 



 
Figure 4.2
 
Images of the solid film (left) and solution (right) sample holder for 
photoluminescent excitation and emission measurements.  Red glow from a chromophore 
solution can be seen wit



 
Figure 4.3
 
P
hoto of an integrating sphere for quantum yield measurement
 

 
Figure 4.4
 
Quantum yield data for a 9,10
-
Diphenylanthracene (DPA) film. DPA is 
dissolved in toluene and spin
-
coated as film. Th
e background glass sample (blue) and 
DPA film sample (red) spectra are plotted. (inset) The emission peaks for the two 
films. 


...42
 
Figure 4.5
 
Schematic of processses of absorption, reflection, scattering, and transmission 
which occur when incident light passes through a sample with a thickness of 
L
 
(left); 
Visualization of the transmission, absorption and reflection from a rhodamine solution 
illuminated with a green laser (right) 

 
note that the exponential absorption profi
le can be 
seen in the solution as. 



 
Figure 4.6
 
P
hotos of transmission measurement of liquid sample in cuvette (left) and 



 
Figure 4.7
 
Configuration of 
JV 
testing for a TLSC 
system.  An additional mask is used to 


.46
 
Figure 4.8
 
Photo of EQE measurement a) monowavelength signal illuminating on the 
surface or the test sample b) the test fixture connected to th
e Keithley picoameter for 
current readings and c) 
the Keithley sourcemeter and channel switcher for testing 
multiple devices.   



 
Figure 4.9
 
Correction for EQE measurement at distance of 
d
 
to the edge when one PV 
cell is attach
ed. 
L
 
is the plate width and 
g
 
is geometric factor. The measured 
EQE
 
is 



 
Figure 5.1
 
Quantum yield of core material MH (Mo
6
Cl
12
) with different cations (K, H
2
O, 
TBA) in solvents acetone (ACE) or 



 
xii
 
 
Figure 5.2
 
(a) Schematic of the transparent luminescent solar concentrator and (b) 
M
6
X
12

2 
nanocluster
-
blend structure (dark green atom = M, light green atom = X); note 
that the salt compounds can be represented in (b) 
by replacing L with A
-
X or L
-
A
-
X.
..
.55
 
Figure 5.3
 
(a) Normalized absorption(solid blue line) and emission (dotted red line) 
spectra of the UV absorbing luminophores overlaid with the measured EQE of the edge
-
mounted Si PVs (grey line). (b) Photograph of th
e LSC system incorporating the 
massive
-
Stokes shift UV nanoclusters under intense UV illumination where  only  the  
tail
-
end  of  red  NIR  emission  can  barely  be  seen  except around the edge. Note that 
some blue/violet light can be seen through the LS
C that originates from tail lamp 
emission in the range of 410
-
450nm.  (inset) Comparison of the transparent luminescent 

-

right) highlighting the need for both absorption 
and emission outside the visible 



 
Figure 5.4
 
Measured quantum yield (QY) of K
2
Mo
6
Cl
14
 
and (TBA)
2
Mo
6
Cl
14
 
in 
acetonitrile as a function of concentration; solid lines are fits to Equation. 1. (inset) QY of 
the TBA f
ilm as a function of time under shelf
-



 
Figure 5.5
 
External quantum efficiency (
EQE
) of the LSC system as a function of 
wavelength; (inset) measured 
EQE 

..

 
Figure 5.6
 
Current densi
ty as a function of voltage for the fully assembled LSC systems 



 
Figure 5.7
 
Light intensity dependence of the 
Voc
, 
FF
 
and power efficiency of the Si 
cell 

.61
 
Figure 
5.8
 
Calculation of the overal LSC system efficiency as a function of plate size and 
luminophore concentration, without (solid lines) and with (dotted lines) the intensity 
dependence of the Si PV.  Note that the solid lines are calculated using the 1
-
sun Si
 
PV 
paramaters and that the plate length of 0.2m roughly translates to a luminescence 
equivalent intensity of 1sun around the edges.  The system efficiency is impacted by the 
luminophore concentration accounting for the solar absorption efficiency, reabsor
ption 
loss, and the quantum yield.   Improving the quantum yield to 100% would result in a 
peak system efficiency of up to 1.3
-
1.5%.  
 



 
Figure 5.9
 
Comparison of transparent UV
-
absorbing luminescent solar concentrators 
with traditional 
color LSCs.



 
xiii
 
 
Figure 6.1
 
(a) Schematic of the transparent luminescent solar concentrator (b) 
Photograph of the transparent LSC system incorporating CY 
luminophore
 
in front of
 
a 


 
Figure 6.2
 
CY and HITC molecular cation structures (top); the natural excited
-
state 
transition orbital pairs for HITC (left) and CY (right).  The hole orbitals are shown on the 



 
Figure 6.3
 
(a) N
ormalized absorption (circle symbols) and emission (square symbols) 
spectra of the NIR
-
absorbing luminophores CY (blue line) and HITCI (black line) films.  
(b)  Measured quantum yield (QY) and absorption of CY (blue line) and HITCI (black 
line) solutions i
n dichloromethane as a function of concentration; solid lines are fits to 
model for the QY 
[
72
]
 
and the Beer
-



 
Figure 6.4
 
(a) Light intensity dependence of the 
Voc
, 
FF
, p
ower efficiency and 
responsivity of the CY TLSC device. (b) Current density as a function of voltage for the 



 
Figure 6.5
 
(a) External quantum efficiency (EQE) of the CY TLSC system as
 
a function 
of wavelength measured from 0.02m to 0.07m, with 1cm increments.  (b) Calculated 
EQE as a function of  CY LSC length in the range of 0.02m to 0.08m.  Measured EQE 
(circle symbols)  are also shown here.  (inset) Calculated and measured EQE as a 



 
Figure 6.6
 
Calculation of the CY TLSC optical efficiency as a function of LSC length 
with a function of simulated Stokes shift.  Stokes shift are 10nm (black, triangle), 20nm 
(black, square), 30nm (black, sphere), 50nm (gray, triangle), 80nm (gray, square), 130nm 
(g
ray, sphere).  For reference, CY and HITCI have Stokes shifts of
 
20nm, and 30nm, 

 

 
Figure 6.7
 
Lifetime test: absorption of CY7 acid films under different conditions as a 
function of time. Film 1 is CY7 acid PM
MA film and Film 2 is CY7 acid fluoroshield 

-



 
Figure 6.8
 
L
ifetime test: quantum yield of CY7 acid film under different conditions as a 
f
unction of time. Films are same as shown in figure 6.7
..



 
Figure 6.9
 
Photograph of a NIR
-
absorbing TLSC demo
. 



 
xiv
 
 
Figure 6.10
 
Flexible transparent luminescent solar concentrators on flexible plastic 
substrate

..

 
Figure 6.11
 
(a) Performance metrics (
Voc,
 
FF, 
efficiency and responsivity) of the TLSC 
system as a function of illumination intensity.  (b) 
J
-
V
 
curve of the sample at different 
incident intensities. (c) 
EQE
 
as a function of waveleng
th measured at distance of 5, 10, 
15, 20, 25 mm from the GaAs cell. (d) Transmission of the overall system. 

8
 
 
Figure 7.1
 
Overall  calculated  LSC  system  efficiency  for  the  UV  only concentrator 
utilizing the spectral properties of Fig. 4, Si PVs, low
-
iron substrate with  index  1.7,  and  
quantum  yields  of  a)  40%  and  b)  80%  as  a  function  of luminophore conce
ntration 
embedded in a  5 mm thin host.  Due to the massive Stoke shifts, the LSC performance 



 
Figure 7.2
 
Overall  calculated  LSC  system  efficiency  for  a)  the  NIR
-
only 
concentrator,
  
and  b)  combined  with  spatially  segmented  50

m  Si  solar  cells  
utilizing  the  spectral  properties, low
-
iron  substrate  with  index  1.7,  and quantum 
yields of a) 40% and b) 80% as a function of luminophore concentration embedded in a 
50

m thi
n host.  Note that with the transparent NIR luminophore embedded in the 
segmented PVs, the overall system efficiency can be more than triple  that  of  the  
segmented  PVs  alone  (grey  line)  for  the  same  level  of transparency and only 
utilizing a sl
iver of the NIR (675
-



 
Figure 7.3
 
Projection for the module cost in $/WP of fabricating transparent LSCs on 
window surfaces a), b) for the UV luminophore and c), d) for the NIR CY luminophore 
using their respective spectral pro
perties. Estimates are for a substrate index of n = 1.7, a 
commercially
-
available Si PV module efficiency of 18% under AM1.5G, and quantum 
yields of a), c) 0.4 and b, d) 0.8 and  modest Stokes shift  < 50nm


 
Figure 8.1
 
Vertical excitation energies of the lowest singlet (
S
1
) and triplet (
T
1
) excited 
states of three candidates for singlet fission. (b) Schematic of the singlet fission
-
energy 



 
Figure 8.2
 
IR 780 in dimethylf
ormamide (DMF) reacted with 1,4 benzenedithiol 


 
Figure 8.3
 

...

 
Figure 8.4
 
New chemical variants with shifted absorption spectrum and modified Stokes 
shif
ts after reaction for 

 
xv
 
 
Figure 8.5
 
(
left) Measured and simulated reflection curves for two 
transparent mirrors
 
designed to accept short
-
wave NIR and trap longer
-
wave NIR emission (shown 
schematically, middle).  (right) Schematic for the ideal mirr
or configuration for the NIR 
and UV LSCs used to eliminate waveguiding losses.


 


 
 
1
 
 
Chapter 1
 
Introduction to Organic and Excitonic Semiconductors
 
 
    
The goal of
 
this chapter is to introduce readers to the field of organic and excitonic 
semiconductors
. The 
underlying 
principles 
of the 
basic properties 
of 
organic and 
excitonic materials 
and solar cells will be
 
discussed
. Then the 
photophysics 
and 
categories of
 
luminescent materials
 
will
 
be introduced
.
 
1.1 
O
rganic and excitonic 
materials
 
 
 
Organic materials
 
generally
 
refer to compounds 
composed of
 
carbon 
and hydrogen
.
 
(
examples are 
shown in Figure 1.1)
.
 
While
 
many 
carbon
-
containing compounds, such as 
C
60
, carbon nanotubes, simple carbon oxides, are not 
strictly 
organic
 
molecules because 
they lack C
-
H bonds, 
they
 
nonetheless 
act 
like organic small molecules
.  
Organic 
compounds consist of both 
intramolecular 
covalent bonding and 
intermolecular 
molecular bon
ding. 
An important subsection of molecular compounds, not limited to 
organic compounds, is called conjugated small molecule compounds. The highly 

-
 

desirable 
electronic 
excitonic properties. The d

-
 

 
electron systems result in electronic 
excitation energies in the range of 
several 
eV, 
which is of great interest when considered 
for applications in semiconductors
, solar cells, and light generation
. 
 
2
 
 
Figure 1.1
 
Structural diagram 
of several organic compounds, (
top from left to right) di
-
indenoperylene (DIP), carbon nanotube, C
60
, (bottom 
from left to right) ClAlPc, 
 
naphthalene through pentacene, Ru
-
complex
 
 

,
 
which 
are 
delocalized above and 
below the atom planes
,
 
create 

, 
and impart
 
many of the
 
optical and electrical 
semiconductor characteristics. 
Organic materials have nearly infinite tunability of the 
optical/electrical properties by functional group and chemic
al substitution. 
    
The 
existence of excitonic states is 
one of 
the most defining characteristic of conjugated 
molecule compounds. An exciton is defined as a bond state of an electron and a hole 
which are attracted to each other by th
e electrostatic Coulo
mb force.
 
The radius of an 
exciton is defined as the average distance between the electron and the 
bound 
hole. 
 
    
  
 
 
In addition
, a lack of necessity for lattice
-
matching characteristic of inorganic film 
growth enables 
greater 
convenience and low cost 
processability
.  Low temperature and 
large area deposition on light
-
weight flexible substrates such as 
metal foils, 
paper, 
and 
plastics ha
ve
 
recently
 
been realized
 
making these materials excellent candidates for a 
range of low cost electronics
.
[
1
-
4
]
 
Organic and excitonic semiconductors have 
already 
been deployed in
 
a variety of applications 
such as
 
trans
istors, sensors, solar cells, organic 
light emitting diodes (OLEDs), 
lasers, and 
circuits.
 
3
 
 
1.2
 
Organic and excitonic photovoltaics 
  
 
Solar energy has gained tremendous attention 
in the past few decades
 
as solar energy 
has 
been 
shown to be 
one of a few 
alternative
 
energy sources capable
 
of 
offsetting 
the 
current worldwide energy consumption. 
 
Solar cells convert 
various 
wavelengths of 
absorbed 
photon flux 
to electrical power. 
 
The first practical photovoltaic device was demonstrated in the 1950s. From th
e 
1980s, research 
on 
silicon solar cells increased the efficiency of photovoltaics
 
substantially
. 
In 1985, silicon 
photovoltaics
 
reached a milestone efficiency of 20%. Over 
the next decade, the efficiency 
continued 
increasing promoted by the power supply 
m
arket
 
and now sits around 25% at the lab 
scale
 
and 
17
% in commercially available 
modules
. 
[
5
]
 
In contrast, t
he first efficient o
rganic solar cell
 
above 1% 
was  demonstrated by Tang 
in the 1980s
.
[
6
]
 
An efficient organic solar cell incorporates a bilayer structure composed 
of a donor layer and an acceptor layer
 
shown in 
Figure 1.2
.
 
The working prin
ciple of an 
organic an
d excitonic cell is as follows
. An incident photon is absorbed by either the 
donor layer or the acceptor 
layer and a tightly bound exciton
 
is produced. The potential 
difference between contact layers provides a weak driving force to separate the exciton. 
If 
a heterojunction 
with 
adequate 
energy level offset 
is in
troduced, it 
can then
 
enhance the 
driv
ing force to dissociate exciton
 
that 
diffu
se to the 
heterojunction 
interface. If the 
exciton dissociate
s
 
before it recombines, free carries (electrons and
 
hole
s) are produced 
and move to electrodes where they are collected to generate current. 
 
4
 
 
 
Figure 1.2
 
Schematic illustration of 
a
 
bilayer heterojunctions
 
at open circuit voltage
. 
Excitons are created and diffuse to the donor/ acceptor interface where the excitons are 
dissociated
. The charge carriers are 
then 
transported to the electrodes
 
to generate power
. 
The work functions for cat
hode and anode are 

C
 
and 

A
, respectively.
 
 
The 
quantum 
conversion efficiency, defined as the 
ratio between 
the number of 
generated 
electrons
 
to number of
 
incident
 
photon
s
, is com
posed of 
the constituent 
mechanism
s as:
  
 
                         
       
  
             
 
      
 
(1.1)
 
Here, 
 
is the 
quantu
m
 
conversion efficiency, 
 
is the absorption efficiency, 
is 
the exciton diffusion efficiency, 
 
is the charge
-
transfer efficiency at the interface, 
 
is the charge dissociation efficiency (i.e. dissociation prior to
 
geminate
 
recombination
)
5
 
 
is the charge collection efficiency
 
(i.e. collection prior to bimolecular recombination). 
 
Figure 1.3
 
Current density (
J
) 

voltage (
V
) characteristics of a GaAs solar cell
. The red 
point is the operating point. The FF is defined by the ratio of 
A
1
/A
2
 
where 
A
1
 
is 
maximized along the 
J
-
V
 
curve. 
 
 
    
The curve showing 
the relationship between 
current density through an electronic 
device and voltage across the terminal is called 
current density (
J
) 

 
voltage (
V
)
 
curve
, 
which
 
is key in the defining the 
output power generation 
of a solar 
cell.
  
It 
is the 
superposition of 
a
 
diode in the dark with the light
-
generated current density. The light 
shifts the 
J
-
V
 
curve down into the fourth quadrant where output power is extr
acted from 
the diode. Figure 1.3 
shows a typical 
J
-
V
 
curve for a
 
solar ce
ll.   
The energy conversion 
efficiency for a solar cell is defined as the fraction of output electricity to the incident 
power, which can
 
be calculated with Equation 1.2
 
 
                         
(1.2
) 
 
6
 
 
where 
 
is the input power, 
 
is the open circuit voltage, 
is the short circuit 
current, and 
is the fill factor.  Here, 
 
is defined as:
 
                   
           
          
(1.3
)
 
and 
 
are the voltage and current density where the output power is 
maximum. The relationship between short circuit current (
Jsc
) and external quantum 
efficiency (
EQE
) is 
 
                                
      
 
   
 
(1.4
)
 
where 

 
is the photon flux. 
The power efficiency of a solar cell is then proportional to 
the area under the 
JV
 
curve in the fourth quadrant.   
 
 
1.3
 
Luminescent m
aterials 
 
     
Excitonic materials by their nature are good photon emitters.  
The term luminescence 
describes those processes that involve 
the release of a photon from an excited molecule
. 
Input energy from different physical and chemical 
energy sources result in a transition 
from the ground state to excited states.
 
Table 1.1
 
shows various source
s
 
of energy and the 
corresponding luminescence terminology.
 
 
 
 
 
7
 
 
Table 1.1
 
Luminescence categories defined based on energy source.
 
 
Source of energy
 
Luminescence terminology
 
Electrical
 
Electroluminescence, Galvanoluminescence
 
Chemical
 
Chemiluminescence,  Bioluminescence, 
Electrochemiluminescence
 
Light
 
Photoluminescence, Fluorescence, Phosphorescence
 
Thermal
 
Thermoluminescence, Pyroluminescence
 
Me
chanical
 
Triboluminescence, Mechanoluminescence, 
 
 
     
For 
species excited by a
bsorbance of a photo
n, fluorescence and phosphorescence
 
are 
the 
two forms 
of luminescence
. 
In general, fluorescence is 
a 
fast 
emission 
process 
that 
occurs 
at 
the 
nanosecond time scale, 
while 
phosphorescence 
occurs at 
longer time scale
s 
from microseconds 
up to hours or even days.
  
The differences between these two forms 
of luminescence, often tied to differences in spin states, will
 
be
 
discussed in greater detail 
th
roughout this thesis.
 
 
       
An electron can have a spin quantum number of 1/2 or 
-
1/2. 
A s
inglet state 
is an 
excited state in which 
all electrons in the molecule are spin
-
paired
 
(up and down)
,
 
so
 
the
 
total spin for the system is 
S
s
 
=
 
0. 
 
In contrast, t
riplet state
s have
 
one set of electron spins
 
which 
is unpaired
, so 
S
s
 
=
 
1
.
 
The spin
-
angular momentum vector 
L
 
associated with these 
electrons and systems can be calculated as 
L
=2
 
S
s
 
+1.
 
For a singlet state, 
L
=2(0)+1=1 so 
only one singlet state exists. For 
a triplet state, 
L
=2(1)+1=3 so three triplet states exist. 
Figure
 
1.4
 
shows the 
simpli
fi
ed 
spin 
configuration 
for singlet and triplet states.
 
Photons 
8
 
 
releas
ed
 
fr
om excited singlet state result
 
in fluorescence emission while 
photons 
from 
excited triplet state 
resul
t
 
in
 
phosphorescence emission. 
 
 
  
 
Figure 1.4
 
Simplified schematic of s
pins in the ground and excited states
 
 
Figure 1.5
 
Jablonski d
iagram 
outlining the 
excitation and emission process
es
 
f
or
 
a
n 
organic or 
excitonic
 
molecule
 
1) excitation to 
the 
first excited state
;
 
2) excitation to 
a 
higher excited state
;
 
3) and 4) vibrational relaxation
;
 
5) non
-
radiative decay
 
to the ground 
state;
 
6) fluorescent emission
;
 
7) intersystem crossing
; and
 
8) phosphorescent emission
 
Figur
e 1.5
 
diagrams the 
major 
processes that can 
occur during excitation and emission 
from 
an organic semiconductor
.  
The ground state (S
0
)
 
can be 
excited to 
a 
first electronic 
9
 
 
excited state S
1
 
with the absorption of a photon, 
or 
an 
even higher energy excited 
state S
2
 
 
 
if the photon energy is large enough
. In 
all cases
, the excited 
state 
rapidly decays to 
lowest vibrational level of S
1
 
through relaxation
. Th
is
 
excited 
state 
may
 
then
 
further 
relax 
to ground state by nonradiative
 
decay, releasing extra energy as heat
 
or phonons or, 
a
lternatively, 
the excitation energy is re
-
emitted as a photon
 
(
fluorescence
)
. 
In addition, 
an intersystem crossing 
process 
may occur 
that transfers the excited state 
from 
the S
1
 
to 
the triplet
, 
T
1
.
 
The T
1
 
state can decay to the ground state S
0
 
by emitting a photon 
(
phosphorescence
)
 
or can
 
relax to ground state by a non
-
radiative decay
.
 
Because the 
transitions between the singlet ground state and excited triplets states are spin
-
forbidden
,
 
phosphores
cence
 
is typically not observed.   However, there are several mechanisms that 
can make this spin
-
forbidden process weakly allowed such as spin
-
orbit coupling and 
symmetry.
  
Indeed, the discovery of efficiency phosphorescent emitters has played an 
important
 
role in the commercial success of organic light emitting diodes.   
      
 
 
    
Thus, t
he 
key 
characteristics 
that define these
 
system
s
 
are the spectral distribution of 
the emission (the emission spectrum)
, 
the excitation (the excitation spectrum; or simply 
as absorption spectrum) and the ratio of the radiative and the nonradiative rates of return 
to the ground state (quantum yield efficiency).
 
The difference between absorption and 
emission spectral maxima is
 
defined as Stokes shift
, and is another key parameter
. The 
Stokes shift may vary from 
a few nanometers to hundreds of nanometers depending on 
the materials. 
Because triplet levels are (almost) always lower in energy than singlet 
energies due to the electr
on repulsion interactions from electron exchange, 
phosphorescent materials 
typically 
have 
substantially 
larger Stokes shift than fluorescent
 
10
 
 
material
s
. 
Fi
gure 1.6
 
show
s
 
an example of 
absorption and
 
fluorescent
 
emission 
spectra 
and Stokes shift
 
for a
 
fluorescent and a phosphorescent
 
emitter
, respectively. 
 
 
 
Figure 1.6
. No
r
malized absorption and emission spectra for 
a fluorescent 
emitter 
(up) 
and phosphorescent 
emitter 
(bottom)
. 
T
he difference between the absorption and 
emission peak is Stokes shift. 
In the case of the phosphorescent emitter the Stokes shift is 
alternatively referred to as the down
-
co
n
version shift.
 
 
11
 
 
1
.3
.1
 
Organic dyes
 
     

. 
By engineering 
the conjugation, the properties of the materials can be varied
 
over a large spectral range
. 
For example, increasing conjugation typically reduces the bandgap of organi
c molecules. 
 
 
    
Organic dyes have been studied as luminophores because of their high quantum yield 
that stems from their excitonic nature, good solubility, and absorption efficiency.
 
Several 
representative dyes investigated 
previously for
 
LSC
s
 
include: 
 
bipyridines
,
[
7
]
  
naphtalimid
es
,
[
8
-
10
]
 
dicarbocyan
ine  iodide
,
[
11
]
, dicyanomethylenes
,
[
11
-
14
]
 
coumarins
[
15
-
17
]
  
and perylenebisimides
,
[
8
-
10
, 
15
, 
18
-
21
]
 
 
rhodamines
,
[
8
, 
11
, 
13
, 
22
, 
23
]
 
sulforhodamines
,
[
11
, 
24
]
 
 
etc.
  
Many properties of these common chromophores and luminophores are 
highlighted 
in 
Table 
1.2
.
  
While the quantum yields of many visible chromophores are close to 1, the 
quantum yields of emitters in the infrared and near
-
infrared are typically much lower 
(<50%), representing an important materials challenge moving forward.
  
 
Table 1.2
 
Absorption, emission and quantum yield for common 
visible 
chromophores. 
   
 

 


 



 

 



 
 



 

 

 

 

 

 



 

 

 

 

 



 


 

 

 

 



 


 

 

 

 



 


 

 

 

 



 
 
12
 
 

)
 




 

 

 

 



 

 

 

 

 

 

 

 

 

 



 

 

 

 

 



 

 

 

 

 



 

 

 

 

 



 

 

 

 

 



 

 

 

 

 



 

 

 

 

 



 


 

 

 

 





 


 

 

 

 



 
 
1.3
.2 
Nanoclusters and q
uantum dots
 
    
Nanoscale materials exhibit optical and electric properties that differ significantly from 
their bulk counterparts. In terminology, nanoparticles, nanocrystals
 
and quantum dots 
all 
refer to
 
nanoscale
 
crystalline ensembles of bulk semiconductors
. In contrast, nanoclusters
 
are an inorganic molecular species, exactly defined in chemical composition and structure 
at the nanometer scale, analogous to organic molecules without organic bonding.
 
 
 
Quantum con
finement is 
a
 
key characteristic of nanomaterials. In a more general form, 
a confined system is 
one 
where a particle 
surface defines
 
the quantum 
potential
 
well 
barrier
. 
 
Quantum confinement effect
s
 
arise when
 
the particle size is comparable to the de 
Broglie wavelength 
of the confined particle. Consequently, nanomaterials have discrete 
13
 
 
energy levels and exhibit a variety of optical and electrical properties which are tunable 
through particle size, surface struc
ture, and elemental constituents.
 
     
Q
uantum dots are widely investigated as luminescent dyes
 
in a variety of applications 
as sensors, lasers, and LEDs
.
[
31
-
33
]
  
The sizes of the quantum dots determine optical 
behavior of the dot 

 
both the absorption and emission spectra 
can 
be tuned as a result 
of size 
variations
. Particle size of the dots can be manipulated by varying the reaction 
conditions
,
[
34
-
36
]
 
but some size distribution is always present. As a result, the bandgap will 
decrease and the photoluminescence spectra are red
-
shifted
.
[
37
]
 
There is a quantitative 
relationship between the red
-
shift and the spread of the quantum dot sizes, providin
g an 
additional strategy to minimize the re
-
absorption losses
.
[
15
]
 
Another strategy that has 
recently emerged with QDs is the synthesis of core/shell structures that afford greater 
tunability and protection.  
The core/shell quantum dots 
also can 
allow for 
spectral 
separation that confin
es absorption and emission processes to different parts of the 
heterostructure resulting in large Stokes shift. 
The increase in the Stokes shift for the 
recently demonstrated core
/
shell QDs can help to reduce re
-
ads
orption of photons, which 
greatly avoids the surface loss and increases the efficiency
.
[
38
, 
39
]
 
However, QDs are quite 
s
ensitive to oxygen and light when they are kept outside a solid matrix
.
[
35
, 
36
]
 
Man
y 
properties of common QDs 
are 
highlighted 
in Table 
1.3
.
 
 
 
 
 
14
 
 
Table 1.3
 
Absorption, emission and 
quantum yield for QD
s from the literature
.    
 


 


 



 

 



 
 


 

 


 

 

 

 

 
[
40
]
 

 

 

 

 
[
41
]
 

 

 

 

 
[
38
]
 





 

 

 

 
[
39
]
 

 

 

-


 

 
[
42
, 
43
]
 

 

 

 

 
[
44
]
 



 

 

-


 

 
[
41
]
 



 

 

 
-
 
[
45
]
 



 

-

 

 

 
[
46
]
 




-


 

-

 

 

 
[
47
]
 
 
1
.3
.3 Rare earth ions
 
    
 
Rare earth 
ions
 
are
 
a group of 
chemical elements in the periodic table, 
most notably  
the 
lanthanides as well as scandium and yttrium.
   
Good photostability and large Stokes 
shift make rare earth ions 
candidate 
dyes for LSCs.
  
However, absorption spectra are 
often notably discreet with sharp and narrow absorption peaks. 
 
Neodymium
 
(Nd
3+
) 
was the earliest rare earth ion
[
48
]
 
used 
in LSCs 
but the 
efficiency was 
low
.
[
49
]
 
It 
exhibits sharp peak absorption 
around 580nm
, but also absorbs 
at longer wavelengths
. Emitted 
photons have wavelength of both 880nm and 1060nm. To 
increase the efficiency
 
of neodymium
-
doped LSCs
, ytterbium(Yb
3+
) is 
usually 
co
-
doped 
15
 
 
with Nd
3+
.  In this way, energy absorbed by Nd
3+
 
can be transferred to the Yb
3+
, the 
emission energy of which is 
higher than the bandgap
.
 
 
     
Rare earth organic complexes are composed of a ligand as an antenna and an emitting 
lanthanide ion. These complexes often
 
feature large Stokes shift and
 
avoid self
-
absorption losses
 
but have 
limited 
absorption ranges making it difficult to harvest 
photons over a large 
spectral 
range
. 
The
 
luminescence process
 
in organolanthanide 
complex
 
occurs as follows
:
 
light is absorbed by a properly selected ligand. Intra
-
energy 
will be transferred from the singlet
 
to the triplet states by inter
-
system crossing. Then the 
energy transfer will be from triplet state to the excited state of the Ln
3+
 
ion. 
Tb
3+
 
based 
organic complexes have a maximum emission efficiency around 500nm, which match the 
spectrum of amorphous Si solar cell
,
[
50
, 
51
]
 
while NIR
-
emitting metal
-
organic Yb
3+
 
or 
Nd
3+
 
complexes are more suitable for c
-
Si solar cells
.
[
52
-
55
]
 
However, the wide
 
utilization 
of rare earth elements 
ultimately been 
limited 
by issues associated with overall quantum 
yield, abundance, and cost
.  
 
 
 
16
 
 
Chapter 
2
 
Introduction
 
to Luminescent Solar Concentrators
 
 
   
 
In this chap
ter, t
he principles underlying the electrical 
and optical 
properties of 
luminescent solar concentrators (
LSC
)
 
will be reviewed. 
Specifically
, 
the operational 
mechanisms of 
traditional 
luminescent solar concentrators will be 
described.  
Component 
efficienc
ies
 
of the LSC 
operating 
process 
will be 
outlined 
in detail
.
 
2.1 Overview of l
uminescent solar concentrators
 
There are 
generally 
two 
categories 
of concentrators being widely developed, 
i) 
focusing
-
optics
 
concentrators
 
(often thermal concentrators)
 
and 
ii) 
luminescent
 
concentrators. A geometric 
focusing 
solar concentrator takes advantage of reflective or 
refractive surface to 
reflect, 
redirect
,
 
or focus incident radiation onto solar cells
 
or 
energy
-
harvesting systems
 
including
 
parabolic troughs, heliostatic
 
arrays, 
sterling
 
d
ishes, 
and Fresnel reflectors
.
[
56
]
 
This approach may work well 
in some configurations
, but it 
suffers from two shortcomings. One is the cost for robust element
s
 
are often 
more 
expensive
 
than the solar cells themselves
. The other is 
that they
 
often 
require solar 
tracking to harvest the direct components of the solar spectrum
 
most effectively
, which is 
often very cost prohibitive
.
  
Second
, the use of tracking systems often prevents their 
incorporation into the building environment as t
hey are typically bulky
, heavy,
 
and 
difficult to integrate seamlessly.
 
17
 
 
 
Figure 2.1
 
Photograph of focusing
-
optics concentrators
[
56
]
 
a
) 
parabolic troughs
 
b
)
 
heliostatic arrays
 
c) 
sterling dishes, and d) Fresnel reflectors 
 
    
The 
luminescent
 
concentrator, 
on the other hand, is a good alternative
 
for its simplicity
, 
ease of integration, and 
low cost
. 
The basic LSC 
design allows sunlight to be absorbed by 
chromophores embedded in an inexpensive plastic or glass waveguide.  The absorbed 
light is re
-
emitted at longer wavelength and a fraction of the emission light is trapped in 
waveguide by total internal reflection. S
mall PV cells are attached at the edge of the 
waveguide to collect the light and convert it to electricity
, thus only requiring a small area 
of the expensive PV material
. 
 
18
 
 
 
Figure 2.2
 
Schematic of a luminescent solar 
concentrator
 
composed of
 
luminescent
 
dyes, 
waveguide, and attached PV
s.  PVs are typically mounted around every edge, but have 
been omitted from the front and back for schematic clarity.  
The component efficiencies
 
shown in the configuration are 1) solar spectrum absorption efficiency 2) lum
inophore 
photoluminescence efficiency (quantum yield efficiency) 3) waveguide efficiency 4) re
-
absorption efficiency 5) solar cell quantum efficiency
 
 
    
The 
concept of luminescent solar concentrators (LSCs) has existed for over 30 years, 
and with recent 
advances in phosphorescent and fluorescent luminophores efficiencies, 
the power conversion efficiency for LSC modules have increased to 7.1% using multi
-
dye systems with GaAs photovoltaics.
[
57
]
 
Although optical funneling limits the 
realistic 
overall system conversion efficiency of LSC
s
 
to less than
 
20
%,
[
58
]
 
it can dramatically 
reduce the area of expensive solar cells needed, driving down the overall module cost.
 
[
15
, 
59
-
62
]
 
A review 
of the high
est
 
efficiency LSCs performance 
metrics 
are shown in Table 
2.1.
 
Note that 
metrics are
 
for 
strongly colored (or black) LSCs,
 
which 
absorb
 
and 
emi
t
 
in 
visible spectrum.
 
19
 
 
 
 
Figure 
2.3
  
Photograph
 
of typical colorful luminescent solar concentrator
s
 
highlighting 
the waveguided emission that is brightly focused at the edge
.
 
(b) Example of a traditional 
LSC in front a window, highlighting the difficulty in installing these systems in a variety 
of appl
ications such as windows.   
 
 
Table 
2.1
 
E
fficiency of high
est
 
performance LSC
s
 
from the literature
.
 
Dye
 
Cell
 
LSC
 
size
 
(m x m)
 
Efficiency
 
(%)
 
Reference
 
Red305, CRS040
 
 
GaAs
 
0.05 x 0.05
 
7.1
 
[
57
]
 
BA241, BA856
 
GaInP
 
0.05 x 0.05
 
6.7
 
[
63
]
 
Rubrene, DCJTB
 
CdTe
 
0
.1 x 0.1
 
4.7
 
[
64
]
 
CdSe/CdS/CdZnS/ZnS
 
Si
 
0.05 x 0.03
 
2.8
 
[
65
]
 
CdSe/CdS
 
GaAs
 
1.4 x 1.4
 
4.5
 
[
66
]
 
Red305, perylene perinone
 
Si
 
0.05 x 0.05
 
4.5
 
[
67
]
 
SrB
4
O
7
 
Si
 
0.05 x 0.05
 
2.5
 
[
67
]
 
EuTT
 
Si
 
0.1 x 0.1
 
0.28
 
[
68
]
 
   
 
 
20
 
 
2.2 
 
Luminophores     
 
 
  
 
L
uminophore
s
 
in LSC system
s
 
are
 
the molecules which absorb and emit photons. 
Here, 


 
chromophore has historically implied the absorption or emission of light in the generation 
of color whereas a luminophore does not necessarily impart color.  
A 
 
luminopho
re is 
regarded as effective if it meets following 
requirements 
[
69
]
 
for LSC applications
:
  
1) 
it 
should have
 
broad spectral absorption, which means it can make maximum use of the 
incoming light
 
of interest
 
or key wavelengths of particular interest
;
 
2) 
the 
quantum yields 
for luminescence
 
should be high
 
(as close to 1 as possible)
;
 
3) 
the 
Stokes shift
 
should be 
large enough 
so that the overlap of th
e absorption and emission is small
 
or 
negligible
;
 
and 
4) 
the 
luminescence wavelength is well
-
match
ed to
 
the solar cell spectral response
.
 
     
Based on the requirements above, a variety of lu
min
ophores are widely being 
explored 
for LSC applications
. Generally, effective lu
min
ophores fall into 
several 
categories: 1) 
organic dyes 2) quantum dots
, 3) nanoclusters,
 
and 
4
) rare earth ions
, which are 
described 
above.
 
 
2.3 
Waveguide and 
Hosts
 
   
The host is the matrix that surrounds the 
lumi
no
phore 
and 
is critical for 
1) creating light 
trapping of the luminescent emission; 2
) physically separating molecules to increase 
quantum yields
;
 
3
) 
interacting
 
directly and electronically with the molecules to increase 
or decrease quantum yields through 
various 
interaction
s;
 
4
) and encapsulate the 
chromophore to act as a barrier to air, moisture, and increase longevity. Because host
-
21
 
 
guest interactions are diffic
ult to predict 
a
-
priori
, we necessarily must screen a wide 
range to optimize for all system components.  
A key 
criterion
 
for host selection is the 
absorption coefficient at both the absorbing and emitting wavelengths of the luminophore.  
Because the host w
ill transport the emitted photons over long distances (cm to m), even 
small absorption coefficients can act to absorb this energy.  S
hown in Fi
gure 2.4
 
is a 
comparison of absorption coefficients for glass and several representative polymers.   
Poly
 
(methyl
 
methacrylate) (
PMMA
)
, for example,
 
is a common host polymer material 
for LSC
 
for its highly transparency and stable chemical properties. 
 
 
    
The waveguide in LSCs is the key structural support that is also used to transport the 
optical energy over large areas.   
Waveguide efficiency indicates how much light can be 
transmitted within the waveguide by total internal reflection. 
According to the S

if photons arrive at the interface at an angle larger than the critical angle, there is total 
internal reflection and the photons are trapped within the waveguide.  The critical angle is 
defined as 
 
 
                  
          
(2.1
)
 
 
Here 
n
 
is the refractive index of the waveguide
 
(and the
 
refractive index of air is 1
)
. 
The 
refractive index for representativ
e materials is shown in Table 2.2
.
 
 
 
 
22
 
 
Table
 
2.2
 
Refractive index for 
common
 
materials
[
70
]
 
Materials
 
Refractive index
 
Air
 
1.00
 
Water
 
1.33
 
Titanium Dioxide
 
2.61
 
Ethanol 
 
1.36
 
Acrylic glass
 
1.49
 
Polycarbonate
 
1.60
 
PMMA
 
1.49
 
Crown glass
 
1.52
 
Flint glass
 
1.61
 
 
 
The 
waveguiding effi
ciency as a function of index can 
then 
be shown as
 
[
69
]
   
 
 
                     
(2.2
)  
 
If the refractive index for the substrate is around 1.5, which 
is typical
 
for most 
industrial 
glass
es
, 
nearly 75
% 
of 
emitted photons 
will be waveguided
 
to the edge
. 
However, t
he 
larger the refractive index, the more light will be reflected at the interface.
  
Thus to 
optimize the product of the reflection and waveguiding 
efficienc
ies
, the optimal index is 
around 1.8
-
2.2. 
Waveguide 
roughness 
and optical transparency als
o play an important 
role as waveguides are scaled to around m
2
 
areas
, where both can act as reabsorption or 
scattering losses.
 
Absorption coefficients for glass, quartz are shown i
n figure 2.4
 
and 
the 
low absorption coefficients lead to low reabsorption losses, which have significant effect 
when scaled up to larger area. 
 
 
23
 
 
 
Figure 
2.4
 
Absorption coefficients of 
standard 
glass, quartz, 
BF33
 
glass
, low iron glass, 
polycarbonate (PC) 
and 
poly(methyl methacrylate) (PMMA) 
[
71
]
 
as a function of 
wavelength
.
 
 
2
.
4
 
Overall 
LSC efficiency
 
 
    
The performance of the LSCs can be understood by the component efficiencies: 
1) 
solar spectrum absorption 2
) 
luminophore photoluminescence efficiency 3) waveguide 
(trapping) efficiency 4) solar cell
 
quantum
 
efficiency and 5) transport (re
-
absorption) 
efficiency.  These processes are highlighted in the overall system LSC system efficiency 
as
: 
[9]
 
          
(2.3)
 
24
 
 
where  
R
  
is  the  
fraction of 
front  face  reflection, 
is the solar spectrum absorption 
efficiency of the luminophore, 
 
is the luminescence efficiency of the luminophore, 
 
is the waveguiding
 
efficiency of the light, 
 
 
is the PV quantum efficiency 
(reported for AM1.5G) normalized by the solar spectrum absorption efficiency and the 
quantum efficiency at the luminophore wavelength,
 
and
 
  
is the 
efficiency of 
transporti
ng photons without reabsorption loss
.
[10]
  
The luminescence efficiency and 
reabsorption efficiency are the most important parameters, especially for large area 
devices.
 
2.
4
.1 
Solar spectrum a
bsorption efficiency
 
For LSCs 
using 
down
-
converting luminophore dyes, the absorption efficiency is 
necessarily 
equal 
to, 
or less than
,
 
the absorption efficiency of the attached PVs. For 
reference, Si solar cells have 50.3% absorption efficiency for the solar spectrum. 
To 
maxim
ize
 
the absorption efficiency, 
it is important for the
 
sheet to absorb as 
much
 
of the 
incident light as possible, 
while
 
also matching the 
emission 
with 
the absorption range of 
the 
mounted 
solar cells.
 
2
.
4.2
 
Solar cell
 
efficiency
 
 
A range of solar cells can be utilized in the fabrication of LSC devices.  By selecting a 
PV with quantum efficiency cutoff that closely matches the spectral emission of the 
luminophore, it is possible to minimize thermal losses (that is voltage). However,
 
cost 
    
and availability will 
ultimately drive the cell desig
n.
    
 
25
 
 
 
Figure 
2.5
 
Photos of cut Si cell (left) and GaAs cell (right) for attaching to the edge of 
the 
LSC 
waveguide.
 
Solar cells attached to the edge 
have
 
quantum efficiency describing how many photons 
at a specific wavelength will be converted to electrons
. 
The best 
Si cell
s
 
have 
a quantum 
efficiency as high as around 85% in the range of 
600
-
1100
 
nm. 
The best 
GaAs cell
s
 
ha
ve
 
quantum efficien
cy close to 95%
 
at wavelengths < 
900 nm. 
Beyond the quantum 
efficiency, a
dditional 
PV losses stem from the intrinsic thermodynamically
-
limited shape 
of the current
-
voltage curve. As the solar cell band gap decreases, voltage and fill
-
factor 
losses 
generally 
increase. Due
 
to the monochromatic emission nature of the LSC, only 
single junction PVs can be considered around each individual LSC, defining the upper 
bound for the solar cell efficiency to that of the single
-
junction defined by Shockley
-
Quiesser. 
Figure 
2.6
 
shows the solar cell efficiency normalized by the absorption 
efficiency for LSC uses.
 
26
 
 
 
Figure
 
2.6
 
Plot of the AM1.5G
 
Solar Cell  Efficiency  as  a  function  of  bandgap (black),  
and  efficiency  normalized  by  the AM1.5 absorption 
efficiency
 
(gold).
 
Highest      
reported  efficiencies  are highlighted
 
 
in red squares.  
 
 
2.
4.3
 
Reabsorption efficiency
 
    
It has long been recognized that LSCs are most limited by reabsorption losses, 
particularly for larger plate sizes. Indeed, much of the research with LSCs has focused on 
the reduction of these reabsorption losses. 
Reabsorption efficiency
 
describes the 
prob
ability that an emitted photon will be transmitted to solar cells without being re
-
absorbed by another luminophore. 
These losses are critically dependent on the quantum 
yield of the dye, the overlap (or Stokes shift)
 
of the dye emission
-
absorption
, and the
 
overall waveguide dimensions.
 
 
Al
though reemission may 
occur 
after reabsorption
 
(particularly for high QY luminophores)
, this still leads to energy loss because the 
emitted light is 
radiated 
in random direction
s
 
and 
a small 
fraction will leave the 
wavegui
de after each event
.
 
It has been shown theoretically, that low QY massive 
Stokes
-
shift materials can be outperform fluorophores with unity quantum yields over 
27
 
 
large LSC dimensions. That is, even for luminophore with 100% quantum yield, 
reabsorption losses 
can become dominant for luminophores with small Stokes shift in 
large waveguides since each absorption/emission event leads to a reduction of photon 
flux through cone emission from the front of the waveguide that effectively act as 
scattering events.
 
 
Reab
sorption losses introduced by the luminophore have been reduced in previous 
work 
through 1) increas
ing
 
Stokes shift
s
 
with
 
organic phosphors,
[
72
]
 
2) 
integrating 
multiple
-
dye 
optimizations
 
to artificially increase the Stokes shift
[
73
]
 
or 3) reso
nance 
shifting,
[
74
, 
75
]
 
applicable only to neat
-
film dye layers less than several microns thick. 
However, t
he clearest paths to fully maximize the large
-
area scaling of these devices is 
through embedded collectors or incorporation of high
-
quantum
-
yield massive
-
Stokes
-
shift
-
lumophores (MSSL) where the former is beneficial for greater power output at the 
expense of poorer cost scaling and the latter is the most effective means to the realization 
of low
-
cost scaling. 
 
 
 
 
28
 
 
Chapter 3
 
Transparent 
photovoltaics
 
     
In this chapter, we 
turn to
 
the concept of
 
transparent solar cells. 
We then introduce the 
idea 
of 
a 
clear, transparent luminescent solar concentrator and contrast 
this 
with recently 
demonstrated transparent organic solar cells.  This wil
l be followed by a discussion about 
application of transparent luminescent solar concentrators in a range of applications 
including as solar windows
 
and
 
the t
heoretical efficiency limit
s
 
for
 
transparent 
luminescent solar concentrator (
TLSC
)
 
are 
described
. 
 
3.1 Transparent solar cells 
 
Transparent solar cells are a transformative new approach
 
to deploy solar harvesting 
surfaces seamlessly around buildings, mobile electronics, and automobiles to generate 
power without knowing they are there.   These system
s
 
c
an be achieved by
 
selectively 
harvest
ing the
 
UV or NIR 
parts of the solar spectrum, the parts we cannot see with our 
eye
s
.
  
They can 
convert any 
surface 
into an
 
electricity
 
generating surfaces
 
while 
maintain
ing
 
the underlying 
aesthetic and design freedom. 
Previous efforts in constructing 
transparent or semi
-
transparent power
-
producing surface have focused on 1) optically
-
thin photovoltaics that have significant tinting or limited transmission,
[11, 12] 
 
or have an 
inherent tradeoff between efficiency and tra
nsparency, 
2)
 
 
LSCs incorporating colored 
chromophores that have either absorption or emission in the visible spectrum, again 
creating significant tinting,
[13, 14]
 
or 3) focusing
-
optics systems using direct
-
light only that 
requires bulky solar tracking or 
optics.
[15, 16]
 
These approaches 
have been 
severely limited 
in their potential for window 
and display 
applications due to aesthetic properties, 
29
 
 
bulkiness, or considerably limited transparency. 
They suffer from an inherent tradeoff 
between power conversion 
efficiency and visible transparency, since both parameters 
cannot be simultaneously optimized.
 
Architectural adoption is impeded further with 
typical PVs that have non
-
uniform absorption within the visible spectrum, resulting in 
poor color rendering index,
 
CRI (high colored tinting), and poor natural lighting quality. 
Recently, selectively absorbing near
-
infrared photovoltaics (PV)
 
with high transparency
 
have been fabricated by exploiting the excitonic character of molecular and organic 
semiconductors with efficiencies in the range of 2
-
4% over small areas (mm
2
-
cm
2
).
[17
-
19]
 
While these transparent photovoltaics offer the highest possible combination of 
tr
ansparency and efficiency, they also have strict defect tolerances, challenges with 
partial shading, and a different set of challenges to scale
-
up.
  
In contrast, visibly 
transparent, UV/NIR
-
selective LSCs avoid aesthetic tradeoffs (low 
visible transmittanc
e
 
or CRI) that hinder architectural adoption and provide a clear route to large area scaling. 
 
3.2
 
Transparent luminescent solar concentrator for solar w
indows
 
    
According to the U.S. Energy information Administration (EIA)
,
 
the
 
buildings
 
sector 
consumes nearly half
 
(47.
6%
) of total energy produced in the Unit States.
 
About 75%
 
of 
all the 
electricity 
produced is consumed by operating buildings. 
While more than half the 
generated electr
icity is lost in distribution, 
it is critical to design
 
sustainable and energy
-
efficient buildings 
or
 
exploit 
existing
 
infrastructure to reduce cost for renewable energy
.
 
Accordingly, many participants in the built environment are looking for solutions to curb 
energy demand and generate renewable energy on
-
sit
e. Solar energy offers a viable 
solution to our growing energy need, but adoption has been slow due to cost and 
difficulties with seamless integration.
 
 
30
 
 
 
Figure 
3
.1
 
Overlap of the AM1.5G solar flux and the photo
n
ic response of the eye shows 
that about 1/3
 
of the photon flux is in the visible part of the spectrum with the remaining 
2/3 of the flux in the NIR
 
    
Roughly 
20
% of all building electricity in the U.S. is used for lighting. Incentivizing 
solar window adoption can encourage architects to utilize m
ore window space, increasing 
natural lighting, and reducing lighting costs. Reproducing the solar spectrum is the goal 
of most artificial lighting applications. 
 
For example, despite significant enhancements in 
power efficiency, fluorescent bulbs have fail
ed to eliminate incandescent bulbs in large 
part due to their reduced color rendering quality that has created 
social barriers to 
implementation. In contrast, natural lighting through windows and skylights provides 
high quality light, reduces additional da
ytime lighting requirements and generates a 
pleasing ambience. For non
-
niche applications it is important to assure that transmitted 
light remains high color quality (as good as or better than, e.g.
 
ceiling lighting) since it 
will offset some form of artif
icial lighting. Each solar window will impart a small 
modification to the spectrum transmitted. Thus, we utilize the color rendering index (CRI
, 
31
 
 
se
e Section 4.4
) evaluation by convoluting the transmission spectra with the solar 
spectrum (e.g. AM1.5G)
.
 
The color rendering framework is useful for defining the visible 
part of the spectrum that should be transmitted and the components that can be utilized 
for power generation. 
 
    
There is significant interest in the solar field in creating three
-
dimensio
nal structures, 
such as solar towers and obelisks to enhance solar collection.
[
76
]
 
These structures can 
collect substantially more flux than solar tracking uni
ts of an equivalent footprint. 
Likewise, buildings can act as excellent solar collectors, collecting sunlight effectively 
throughout the day on each face. 
 
When energy
-
scavenging devices are installed to the 
window can also help cool the buildings. A signi
ficant fraction (~15%) of building 
electricity is utilized for air
-
conditioning. Conversion of electricity to heat is highly 
efficient (nearly 100%), while cooling efficiencies are significantly lower. Solar windows 
can dramatically reduce unwanted solar h
eating from infrared flux by utilizing this 
energy for electricity generation and rejecting the rest from transparent, NIR
-
selective 
mirrors. In fact, many researchers and companies are exploring coating technologies to 
achieve IR
-
rejection functionality a
lone, even in colder climates.
[
77
]
 
 
It is an added benefit 
to utilizing this unwanted resource for electricity generation that could compliment or 
replace current low solar heat gain coefficient coating technologies. This hear
-
rejection 
functionally can enhance the effective LSC from direct
 
power generation by 50
-
100%.
 
    
Integrating solar
-
harvesting systems into the building envelope is a transformative 
route to improving building efficiency, capturing large areas of solar energy, and 
lowering effective solar cell installation costs by piggybacking on the installation, 
fra
ming, and maintenance of the existing building envelope. 
There has been 
32
 
 
demonstrated interest in utilizing LSCs as architectural windows.
[
69
, 
78
-
82
]
 
 
However, the 
widespread adoption of such a pathway is typically hampered by difficulties associated 
with mounting traditional photovoltaic (PV) modules in non
-
standard configurations on
 
and around buildings due to added structural cost, architectural impedance, and most 
importantly, aesthetics.   To overcome these hurdles we have developed UV absorbing 
and NIR absorbing luminescent solar concentrators
 
which are highly transparent in the 
visible spectrum and enable integration onto window and glazing systems in the building 
envelope
 
without disrupting the visible components or require solar tracking
. These 
transparent LSCs offer a different route to large area scaling with high defect tole
rances 
compared to other transparent photovoltaic devices.  These 
LSC 
systems
 
present 
opportunities for non
-
tinted, highly
 
transparent solar
-
scavenging windows that can 
translate into improved building energy efficiency and lower cost solar harvesting 
syst
ems beyond the building skin including
 
1) energy scavenging electronics displays, 2) 
autonomous electrochromic
-
windows, 3) visible
-
blind detectors, and 4) coatings for 
improved UV response and protection of traditional solar installations.
 
These LSCs can 
g
enerate electricity from both direct and diffuse light, and enhance the low
-
light 
efficiency through solar concentration while maintaining window glass aesthetics and 
architectural design freedom. Successful commercialization of this technology will 
result
s in affordable net
-
zero
-
energy buildings which can be widespread throughout the 
build environment.
 
33
 
 
 
Figure 3.2
 
Diagram of the point of installation for building glasses with transparent 
luminescent solar concentrators
 
 
 
3.3 
Thereoetical and Practical 
Limits of 
Transparent Luminescent Solar 
Concentrators
 
Figure 3.3
 
Transparent PV theoretical efficiency limits
 
as a function of bandgap and 
level of visible transparency (VT)
.  Note that these limits are the same for transparent 
LSCs
 
with ideal absorption/emission profiles with no spectral overlap.
 
    
The theoretical limits of transparent luminescent solar concentrators are fundamentally 
constrained by
 
those of the transparent photovoltaic
 
(see Figure 3.3)
.
[
83
]
 
Interestingly, 
without harvesting any visible light the theoretical efficiency limit for transparent devices 
(single junction) is over 20%.   For reference, the single junction opaque limit is 33.1%.   
Using this as a starting point the limits of TLSC are 
derived as a function of key idealized 
34
 
 
parameters outlined below
 
(see Figure 
3.4
)
.  It should be noted that the practical 
considerations of the performance of the TLSC is dependent on a greater number of 
parameters and subsystems and thus follows a linear 
progression in this analysis.  We 
start from the most idealized case and 
analyze
 
a number of cases as the idealized 
parameters are relaxed.         
 
 
Figure 3.4
 
Representative absorption and emission characteristics of TLSC emitters.  
The Stokes shift (
S
)
 
and emission width (
W
), two key emitter parameters, are highlighted.    
 
                                                                                                                             
                                                         
                                                                                                                             
                                                                                                                             
      
                                                                                                                             
                                                                                                                             
      
                                                                                                                             
                      
 
The ideal transparent solar concentrator
 
requires the following conditions
 
be met
: 1) 
the PL efficiency is 100%; 2) the PV mounted around the edge has an efficiency and 
quantum efficiency equal to the 
Shockley

Queisser
 
limit; 3) there are no reflection losses 
into the luminescent absorber; 4) there are no waveguiding losses (perfect li
ght trapping); 
5) there is no overlap between the emission and absorption (no reabsorption losses); 5) 
35
 
 
there is no intensity dependence to the edge mounted PV; and 6) the emission width (
W
) 
is perfectly narrow. 
 
In this 
case, the efficiency limit is 20%
, i
dentical to TPVs
.
 
 
Figure 3.5
 
TLSC practical efficiency limits with no ideal absorption/emission profiles 
with no spectral overlap as a function of the a) emission width (
W
) with 
S
 
= 0 or b) 
Stokes shift (
S
) with 
W
 
= 0.  Note that the decrease in efficiency stems from a lower 
theoretical voltage and FF from the PV cell mounted around the edge. However, the 
situation in b) has a distinct advantage in practice since it will suppress reabsorption 
losses in non
-
ideal ab
sorption emission overlap configurations.   That is, narrow emitters 
will always provide greater flexibility for larger Stokes shifts and minimized reabsorption 
losses.   
 
Th
e ideal
 
case is extended to
 
luminophores
 
with
 
a finite emission width (
W
) as a 
fun
ction of 
S
.   This is one of the first key areas of loss as it confines the solar spectral 

required in selecting a particular PV cell.  For example, given a particular molec
ule 
emitter with a defined absorption, the smaller the 
W
, the higher voltage PV
 
can be 
selected. A range of emission widths have been demonstrated for organic molecules (50
-
100nm),
[
84
]
 
nanoclusters (50
-
300nm),
[
72
, 
85
]
 
J
-
aggrega
tes (10
-
20nm),
[
86
]
 
nanocrystals 
(10nm),
[
87
]
 
and carbon nanotubes (10nm)
[
88
]
 
suggesting that such losses can, in fact, be 
minimized.  Nonetheless, for the most efficient emitters, this value of 
W
 
is typically 
around 100nm
. 
    
 
36
 
 
For practical efficiency limits we consider the case where: 1) the PV quantum 
efficiency is 90% 
(typical for GaAs and Si) with a 1
-
sun photovoltage that is 90% of the 
SQ limit
;
[
89
]
 
 
2) the waveguid
ing efficiency is 85% (
n 
= 1.7
); 3) W = 80nm and S = 80nm, 
and 
4) the emitter is 100%
 
with no reabsorption losses.
  
This results in a practical 
efficiency limit of 11.6% for an emitter bandgap of 1.12eV
.
  
The 
practical
 
efficiency 
limits
 
track surprisingly close to 
that 
for transparent photovoltaics
 
which was estimated
 
to 
be
 
1
1
%
,
 
also with a bandgap of 1.12eV
.
[
83
]
 
Moving forward, i
t then becomes a materials developme
nt challenge to create emitters 
with high quantum yield, optimal bandgap and wavelength selectivity, and a combination 
of both narrow emission and moderate Stokes shifts.   Provided these properties are met, 
highly efficient and exceptionally low cost tran
sparent solar harvesting systems can be 
developed and deployed.   
 
 
 
 
 
 
 
37
 
 
Chapter 4
 
Experimental Techniques
 
 
    
In this chapter, 
key 
experimental techniques utilized to measure chemical, optical, and 
electrical properties of materials and devices
 
are discussed
. We 
begin 
with a discussion 
of 
luminescence spectroscopy
, and 
luminescence 
quantum yield measurement
s
. 
W
e 
then 
describe
 
optical absorption, reflection and transmission measurement
s, followed by 
electrical 
solar cell 
testing and 
quantum efficiency 
measurement
s
. Finally
,
 
simulation 
methods 
are
 
described and discussed
. 
 
 
4.1
 
Photoluminescent e
mission 
and excitation 
testing
 
Photoluminescent 
spectra are measured
 
with 
a 
spectrofluorometer
.
 
 
The 
spectrofluorometer
 
uses an 
optical
 
excitation source to
 
pump a thin film or solution 
sample while recording the intensity and wavelength of photons emitted from the sample.
  
Emission intensity of a particular wavelength (as selected by the output monochromator) 
is detected
 
for emission sca
ns
.  
For excitation scans, t
he pump wavelength is scanned 
through a range, with the emission 
wavelength 
of the output monochromator 
fixed at one 
specific emission wavelength
.  As with the emission scan, a background scan 
is 
p
er
formed
 
for each sample over the same wavelength range.
 
     
Filters are used to help reduce the influence of the background and prevent false 
signals (e.g. wavelength doubling) in the monochromaters. There are two types of filters 
commonly utilized in these expe
riments, short pass and long pass filters. They are 
38
 
 
typically fabricated with absorptive molecules or multilayers of dielectric films. Long 
pass filters enable any wavelength longer than wavelength cutoff to be transmitted, as the 
name implies, while
 
short
 
pass filters enable any wavelength shorter than the cutoff 
wavelength to be transmitted. Filters can be placed both excited side and emitted side and 
their selection is important in the design of the experiments. 
 
Example test fixtures for 
both solution a
nd solid films are shown in Fig 4.2  
 
.
 
 
Figure 4
.1
 
Schematic of spectrofluorometer system
 
39
 
 
 
Figure 4.2
 
Images of the 
solid film (
left
) and solution (
right
) sample holder for 
photoluminescent 
excitation
 
and emission measurement
s
.  
Red glow
 
from 
a 
chromophore
 
solution can be seen 
with
 
green excitation at
 
550 nm. 
 
 
4.2 Quantum yield measurement
 
    
Photoluminescence quantum yield
 
(QY)
 
can be measured 
in a variety of ways
:
 
using 
reference dyes of known quantum yield and index, by concentration curves,
 
or in 
integrating spheres.
[
90
-
94
]
  
The most reliable method is the use of integrating sphere which 
does not require any other 
reference dye 
samples (which can easily be contaminated or 
diminished performance) and can be made with two simple scans and minimal error bars. 
I
n an int
egrating sphere
, all light from t
he luminescent sample including 
waveguided and 
forward emitted/reflected light is uniformly redistributed throughout
 
the sphere by a 
reflective coating with nearly 100% reflectivity
. (
Gamma Scientific)
 
Because of this 
uniform redistribution, the excitation or luminescence emission can be detected at any 
point of the sphere, typically as a small exit aperture 
directed toward 
the output 
monochromater. 
Two scans 
are performed
 
for each sample over the 
full 
w
avelength 
range and with the same configuration
; the first scan is 
performed 
with 
a 
reference
 
40
 
 
substrate 
or 
cuvette without dye to measure the total excitation intensity.  The second 
scan is performed for the desired sample
.  
T
he 
reference or 
background sca
n is 
then 
generally subtracted from the sample scan
 
around the luminescence region of the spectra
.  
By recording the change in the 
excitation 
intensity with a blank and the sample under test, 
the quantum yield is calculated as
:
[
28
]
 
 
        
(4
.1
)
 
where 
 
is the luminescent intenstiy from the film, 
 
is the 
excitation 
inensity from a 
bare substrate, 
is the
 
excitation
 
intensity 
from the 
tested 
film, 
is the calibration 
factor (for s
pectrometer, intergrating sphere and fiber coupling) at the wavelength range 
of the emission, and 
is the calibration factor at the waveleng
th range of the excitation 
light
. 
 
To measure the quantum yield, an integrating sphere is 
l
oaded
 
directly in the 
spectrofl
uo
rometer 
to ensure that all light illuminating the sample is collected (overall 
wavelengths and emission angles).  The testing procedure is 
similar to
 
the procedure for 
emission 
spectra 
except that the emission range 
must in
clude 
the excit
ation
 
wavelength.
 
41
 
 
 
Figure 4.3
 
P
hoto of an integrating sphere for quantum yield measurement
 
Figure 4.4
 
shows an example or quantum yield measurement data for 
a 
diphenylanthracene (DPA) 
film
. The quantu
m yield
 
can be calculated by dividing the 
difference of the absorption and emission peak
 
for the two films
, respectively.
 
 
42
 
 
 
Figure 4.4
 
Quantum yield 
data 
for 
a 
9,10
-
D
i
phenylanthracene (DPA)
 
film. 
DPA is 
dissolved in toluene and spin
-
coated as film.
 
The 
background
 
glass
 
sample
 
(
blue
) and 
DPA
 
film
 
sample
 
(
red
) spectra are plotted.
 
(
inse
t) 
The emission peaks 
for the two films. 
 
 
4.3
 
Optical absorption, reflection and transmission
 
     
When 
photons are incident 
on the surface of the materials,
 
they can be absorbed
, 
reflect
ed
, scatter
ed,
 
or
 
transmi
tted
.  
 
43
 
 
 
Figure 4.5
 
S
chematic
 
of 
p
rocess
ses
 
of absorption
, reflection
, scattering, and transmission 
which
 
occur 
when incident light pass
es
 
through a
 
sample with a thickness of 
L
 
(left); 
Visualiz
ation
 
of the
 
transmission, absorption and reflection 
from a r
hodamine solution 
illuminated with a
 
green
 
laser 
(right)
 

 
note that the exponential absorption profile can be 
seen in the solution as
. 
 
  
Figure 4.5 shows absorption, transmission, scattering 
process
es that can occur
 
at 
a 
sample 
surface. 
 
The light enters the sample 
at 
X
=0 is 
 
 
 
      
(
4.2
)
 
  
is the total incident light intensity and 
R
 
is the fraction of the light reflected at the 
front surface. For 
unpolarized light at normal incident to a flat surface
 
(with one surface)
, 
the fraction of light reflected is given by:
 
  
   
(
4.3
)
 
w
he
re
 
n
 
is the 
re
fractive index of materials, while the 
n
 
for air is 1.
 
   
The light intensity, which is a measure of the flux of photons in the optical beam, 
decreases through the sample as the photons are absorbed. Assuming there is a uniform 
probability of absorption throughout the sample, the intensity reaching distance 
X
 
int
o the 
sample 
is described simply 
by 
the 
Beer
-
Lamber
t
 
equation
:
 
 
 
     
 
(4.4)
 
44
 
 
w
here 
 
is the optical absorption coefficient
 
and is
 
linearly 
related to the concentration 
of the sample
 
and
 
a strong function of wavelength. 
T
ransmi
ssion
, 
T
, is then
 
is defined as
:
 
   
 
        
(
4.5
)
 
    
The relationship 
between 
absorption (
A
), transmission (
T
) and reflection (
R
) is
 
then 
related by the population (or energy) balance as
:
 
 
   
     
 
 
(
4.6
)
 
  
 
 
In many experiments, a
 
reference 
sample is 
typically utilized in dual beam 
spectrometers to 
correct for 
reflection
s 
and 
simplify the measurement or calculation of 
the absorption or absorption coefficient
 
(
f
or
 
solution sample, a pure solvent sample is as 
controls and for films substrate film is the control sample
)
.
  
However, because of the 
complexity of accounting for reflections in thin samples, reference samples should 
generally not be utilized with samples w
ith thin films.  In this case it is better to directly 
measure 
R
 
and 
T
 
independently with different test fixtures.   
 
Nonetheless, it is also 
important to note that when measuring 
T
 
for devices, it is often the overall transmission 
that is of greatest conc
ern, and reference samples 
are not 
utilized for such absolute 
measurements.  
 
  
 
45
 
 
 
Figure 4.6
 
P
hotos of transmission measurement of liquid sample in cuvette (left) and 
film samples on glass (right)
 
 
 
4.4
 
Electrical, 
J
-
V
 
and EQE
 
measurement
s
 
    
Current 
density versus voltage (
J
-
V
) measurements were obtained under simulated 
AM1.5G solar illumination corrected for solar spectrum mismatch.
[
95
]
 
L
ight from a xenon 
arc lamp is collimated and simulated as AM1.5G solar spectrum. 
The illumination area, 
with a diameter of
 
2.5
 
cm, uniformly 
covers a masked region of 
the sample
, but limits the 
overall 
LSC test size
. 
A range of neutral density
 
filters 
positioned between the light 
source and the sample
 
enable
d measurement of 
different 
illumination 
intensity. 
The 
J
-
V
 
curve of the LSC device or module is measured with a Keithley sourcemeter. 
 
Figure 4.7
 
sho
ws device set for 
J
-
V
 
testing.
 
Two cells were mounted on orthogonal 
edges and connected in parallel.  The remaining two edges were covered with enhanced 
specular reflector (Vikuiti, 3M).   
 
 
46
 
 
 
Figure 4.7
 
Configuration of 
JV 
testing 
for a TLSC system
.  An additional mask is used to 
block the edge of the cells from direction illumination.  
 
 
External quantum efficiency (
EQE
) is an important characteristic of all solar cells
 
that 
describes the efficiency of electron generation at each wavelength in the 
solar spectrum
. 
The external quantum efficiency is measured as follows. The light from a xenon lamp 
is 
directed through a
 
filter wheel
 
chopped at 75Hz, into a monochromater 
to produce 
monochromatic 
beam that is fiber coupled to a sample.
 
The monochromatic light 
illuminate
s
 
a small area 
of the sample 
(
under filling
) 
and the generated 
current
 
is detected
 
with a lockin amplifier
.
 
 
When testing the EQE for a
n
 
LSC 
device,
 
only one solar cell was mounted around the 
edge.  The remaining 3 
edge
s
 
were blocked
 
with black electrical tape to eliminate edge 
reflections and 
simplify the geometric configuration for the EQE measurements
.
.  To 
avoid any direct illumination of the solar cell, a thin 
mask was used to define the active 
area around the LSC edg
es 
and shield the edge
-
mounted PVs
.  Due to the illumination 
47
 
 
area of the solar simulator (67005 Xe arc lamp, Newport) plate lengths for power 
efficiency measurements were limited to < 0.05m and for EQE measurements < 0.2m.
 
 
Figure 4.8
 
Photo of EQE measurement
 
a) monowavelength signal illuminating on the 
surface or the test sample b) the 
test fixture connected 
to the Keithley 
picoameter 
for 
current readings 
and 
c)
 
the Keithley sourcemeter
 
and channel switcher for testing 
multiple device
s.   
 
   
 
To predict scalability, reabsorption losses, etc,
 
EQE measurements were obtained by 
directing a monochromatic excitation beam from a fiber perpendicular to the LSC at 
various distances (
d
) from a single edge
-
mounted PV.  The measured EQE was 
then
 
corrected by the factor, 
,
[
12
]
 
due to the different angle subtended by 
the solar cell at each spot distance, where 
g
 
is the geometric collection correction for 
collecting photons only on one edge of the concentrator (the
 
rest of the edges are 
blackened), 
d
 
is the distance from the illumination spot on the LSC and 
L
 
is the length of 
the collection edge of the LSC. The EQE was measured utilizing a Newport calibrated Si 
detector.
[
12
]
  
 
48
 
 
 
Figure 4.9
 
Correction for EQE measurement at distance of 
d
 
to the edge when 
one PV 
cell
 
is
 
attached. 
L
 
is the 
p
late width
 
and
 
g
 
is geometric factor. The 
measured 
EQE 
 
is 
corrected
 
by the geometric factor. 
 
 
4.5 LSC simulations
 
As discussed in previous chapters, the efficiency an LSC system is comprised of 
component efficiencies defined by equation 2.3. 
Here we derive the reabsorption losses 
as a function of key parameters to simulate the scalability of a range of LSC devices.   
First, a collection efficiency 
Q
 
is introduced to calculate the limit of reabsorption losses 
along the whole waveguide. 
Q,
 
defined as the fraction of the absorbed solar photons that 
are transported to the PV, can be calculated as the sum of the collection 
efficiency of 
each absorption/emission event
, Q
i
, where 
i
 
is the remission event number. 
[
96
]
 
 
                                         
(4.7)
 
Each time a photon is reabsorbed prior to reaching the PV, it is still possible for the next 
remitted photon to reach the PV provided there is a high PL efficiency and the emitted 
photon is not emitted with the angle range of the escape cone.  If the photo
n is emitted 
49
 
 
within this escape cone it can be lost out of the front or back of the LSC.  
 
is then
 
defined as the probability that luminescence outside the critical cone will be re
-
absorbed 
before it reaches the PV on the edge, an
d 
 
is the probability of reabsorption for 
photons emitted within the critical cone.  
Q 
can be written as an infinite series as:
 
 
(4.8)
 
Because 4.12 is a convergent series, it can be written more succinctl
y as:
 
                           
(4.9)
 
Here we assume 
 
= 0 (there is no self
-
absorption in the critical cone) since we focus 
on thin films of luminophores. In this case,
 
                
(4.10)
 
where 
is the  total reabsorption efficiency in the waveguide: 
 
                                  
(4.11)
 
Again, the energy conversion efficiency of the LSC is: 
 
 
                      
(4.12) 
 
With each reabsorption
-
emission event, the efficiency will be impacted if both the 
quantum yield and waveguiding efficiencies are < 100%. Unless reflectors can be 
50
 
 
attached to the front and back side of
 
the waveguide to reflect escaped emission light 
back to the waveguide, the waveguide efficiency will also impact the total reabsorption 
efficiency. Because of the light dependence of 
this correction is dependent on the 
geometrical gain of the collector, 
where the incident photon flux on the LSC is related to 
the photon flux around the edge as
 
. 
 
Appreciating that the path length for reabsorption events depends on the emission 
angle within the waveguide, the attenuation along the photon trajectory is then
 
modified 
from 
exp(
-

 
to 
exp(
-


))
 
where 

 
is the molar absorptivity, 
L 
is the 
plate length

 
is azimuth relative to the normal of the LSC plane, and 

 
is the normal 
rotation coordinate, 
y
 
is distance along the plate direction parallel
 
to 
L.
  
The reabsorption 
efficiency  
 
can then be obtained by integrating the probability of arrival along all 
paths 
outside of the escape cone 
from a fixed volume element around the plate length, 
and over all emission wavelengths weighted by the normalized luminescence spectrum of 
the luminophore. 
 
(4.13) 
 
where 
C
 
is the concentration, and the luminescence spectrum, 

),
 
is the normalized 
luminescence spectrum. The two exponential terms result from the fact that the light is 
guided in both directions along the length with 
y
 
= 0 defined in the center of the 
51
 
 
waveguide. The integral over 
y
 
can be obtained analytically, and
 
rearranging the order of 
integration considering variable 
dependencies
 
is: 
 
       
  
       
 
(
4.14)
 
Finally, accounting for the fractional time each photon spends in a compound waveguide 
(
t
) with a thin 
luminophore film (
t
0
), gives: 
 
   
 
(4.15)
 
The results of numerically integrating this equation as a function of luminophore 
parameters and LSC plate sizes are described in Chapter 5.   
 
 
 
52
 
 
Chapter 5
 
UV
-
A
bsorbing Transparent Luminescent Solar Concentrators
 
 
5
.
1
 
Introduction
 
Transparent luminescent solar concentrator can be enabled 
by 
selectively 
harvesting in 
different spectral ranges: in the UV or NIR
. 
Despite the lower overall fraction of the solar 
spectrum in the UV (~6% photons, ~10% energy), there is si
gnificant potential in 
harvesting the UV in a transparent luminescent solar concentrator (TLSC) configuration 
up to 4%
 
system energy conversion 
efficiency
. 
In this chapter
, we 
demonstrate
 
a 
TLSC
 
employing novel 
nanocluster
-
polymer blends that allow for 
selective ul
traviolet light 
harvesting that 
results in a high degree of visible light transmittance
 
demonstrating this 
pathway as a viable route to the production of transparent LSCs
. 
 
To efficiently capture and transform UV light around the visible spectr
um, we focus on 

6
(II)X
12
.  
While the parent compounds of M = Mo and W, and X = Cl, Br, I etc. have been known 
for some time, their use has been limited to photophysical studies and 
oxygen sensors 
with quantum yields typically less than 20%.
[
95
]
 
These materials are highly stable, highly 
luminescent, US abundant, and non
-
toxic.
 
Here, we have synthesized 
nanocluster 
complex
-
host blends with quantum yields > 75% and anticipate reaching near
-
unity 
quantum yields through further chemical and ligand mo
difications.
 
Luminophore
-
ligand
-
host pairing is shown to increase the solid
-
state quantum yields (over solution) through 
polar interactions, while also encapsulating the nanoclusters.  
We subsequently exploit 
the massive Stokes shift (~400nm), near
-
infrare
d emission, and perfect UV
-
VIS 
53
 
 
absorption cutoff to 1) reduce nearly all reabsorption losses in the LSC configuration, 2) 
maintain a high degree of visible transparency and 3) eliminate visible glow.  
The TLSC 
demonstrated here based on inorganic 
nanoclust
er 
luminophores provides an alternative 
strategy for transparent solar harvesting systems over very large areas.
 
5
.2 
Experimental s
ection 
 
5
.2.1 
Nano
cluster preparation
 
Metal(II)  halide  clusters  were  modified  from  the  parent  salt  compounds through
 
ligand exchange reactions. The parent cluster was synthesized through reduction of 
commercially available M(V)X
5,
 
following previously outlined procedures
[
97
-
99
]
 
or 
purchased directly in the form of M(II)X
2
.
  
For Mo
-
based clusters, the parent compound 
was then purified by conversion  to  the  HCl  salt,  which  was  returned  to  M
6
(II)X
12
  
via  heat
-
treatment at > 200ºC under  vacuum.  For K
2
M
6
X
14
, a 5mg/mL solution of KCl 
in 6M HCl was added to a stirred solution of 2mg/mL of Mo
6
Cl
12
 
in 6M HCl at a volume 
ratio of 1:6.   The resulting solution was concentrated by boiling in a hotplate wit
h 
stirring in an inert atmosphere.   The solution was slowly cooled below room temperature 
resulting in yellow needles.   The TBA salt was prepared by adding tetrabutylammonium 
chloride to a solution of 2mg/mL of Mo
6
Cl
12
 
in 6M HCl which precipitated solid 
needles.  
Ligand exchange was carried out in a Soxhlet extractor with the parent compound for 
several days.   Variations of M, X, and L in M
6
X
12

2, 
A
2
M
6
X
14
 
and A
2
M
6
X
14

2 
including M = Mo and X = Cl,  and   L   =   Cl, H
2
O   (hydrate),   acetonitrile (ACN
),   
and   A   =   K,   and tetrabutylammonium (TBA) were characterized, where ammonium  
salts can be readily anchored to polymer chain backbones. Solutions for optical 
54
 
 
characterization were p
repared by directly dissolving 
each compound in acetonitrile at 
various concentrations up to 4mg/ml.
 
 
Figure 5.1
 
Quantum 
yield 
of core material MH (Mo
6
Cl
12
) with different 
cations 
(K, H
2
O, 
TBA) in solvents acetone (ACE) or acetonitrile (ACN)
 
 
5
.2.2
 
Module fabrication:  
 
    
To fabricate modules, a 4mg/ml A
2
M
6
X
14
 
acetonitrile solution was mixed with a 
30mg/ml 
Poly
-
(ethyl methacrylate) (
PEMA) acetonitrile solution (Sigma
-
Aldrich) and 
(
poly)
-
butyl methacrylate
-
co
-
methyl methacrylate (PBMMA)
 
(Sigma
-
Aldrich) at a 
volume ratio of 1:2:1, yielding the target dye concentr
ation of 1mg/mL in the polymer 
composite film. This mixture was drop cast on 2.5cm 
x
 
2.5cm x 0.1cm substrates and 
allowed to dry for 2 hours for each layer under flowing nitrogen, and repeated 7 times, 
resulting in a layer thickness of approximately 200µm.
 
On the mounted edge, a laser
-
cut, 
2.5cm 
x
 
0.1cm Si cell (Narec Solar) with an efficiency of 15±1% @ 1 sun (see Figure 
5.7
) 
was attached using nearly
-
indexed matched epoxy.  For the EQE measurements, the other 
55
 
 
three edges were taped with black electrical t
ape to block edge reflection and simplify the 
geometric configuration.  For the power efficiency measurements, two cells were edge
-
mounted on orthogonal edges and connected in parallel.  The remaining two edges were 
covered with enhanced specular reflector
 
(Vikuiti, 3M).   A thin border area around the 
LSC edges was masked to avoid any direct illumination of the solar cell.  Due to the 
illumination area of the solar simulator (67005 Xe arc lamp, Newport
)
 
plate lengths were 
limited to < 0.05m.
 
5
.3
 
Results
 
 
Figure 
5
.
2
 
 
(a) Schematic of the transparent luminescent solar concentrator and (b) 
M
6
X
12

2 
nanocluster
-
blend structure (dark green atom = M, light green atom = X); note 
that the salt compounds can be represented in (b) by replacing L with A
-
X or L
-
A
-
X.
 
 
   
Details of the hexanuclear 
nanocluster 
preparation are discussed in the Experimental 
section and the core cluster structure is shown in Figure 
5
.
2
(b). The absorption and 
emission spectra of the hexanuclear metal hali
d
e clusters are shown in Figure 5.3
(a
).  The 
spectra show band absorption from 300
-
430nm with near perfect UV
-
VIS absorption 
56
 
 
cutoff and broad near infrared (NIR) emission centered at 800nm that stems from highly 
efficient phosphorescent down conversion and emission.  Nanocluster
-
ligand variat
ions 
of 
M
6
X
12
, M
6
X
12

2
 
and A
2
M
6
X
14

2
 
exhibit nearly identical absorption
-
emission 
properties, where emission spectra can be tuned roughly 50
-
100nm with metal/halogen 
(M/X) substitution, and terminal cations/ligands (A/L) are found to primarily impact the
 
overall non
-
radiative rates without impacting spectral shape.  
 
 
            
Fig
ure 5.3
 
(a
) Normalized absorption(solid blue line) and emission (dotted red line) 
spectra of the UV absorbing luminophores overlaid with the measured EQE of the edg
e
-
mounted 
Si PVs (grey line). (b
) Photograph of the LSC system incorporating the 
massive
-
Stokes shift UV nanoclusters under intense UV illumination where  only  the  
tail
-
end  of  red  NIR  emission  can  barely  be  seen  except around the edge. Note that 
some blue
/violet light can be seen through the LSC that originates from tail lamp 
emission in the range of 410
-
450nm.  (inset) Comparison of the transparent luminescent 

-

right) hi
ghlighting the need for both absorption and emission outside the visible spectrum.
 
     
To understand the photophysical behavior of these nanoclusters, we explore the 
luminescence quantum yield (QY) dependence on the concentration.  The measured QY 
of hexa
nuclear metal halide clusters with different cations/ligands in both acetonitrile 
solutions and blended in (
poly)
-
butyl methacrylate
-
co
-
methyl methacrylate (PBMMA) / 
57
 
 
(poly)
-
ethyl methacrylate (PEMA) solid
-
film com
posites are shown in Figure 4.3
 
as a 
functi
on of 
nanocluster 
concentration for K
2
Mo
6
Cl
14
 
and (TBA)
2
Mo
6
Cl
14
. 
 
We fit the QY 
data of Figure 5.4
 
with the model:
 
 
            
(5
.3)
 
where 
 
is the rate for luminescence (phosphorescence), 
 
is the intrinsic non
-
radiative rate for isolated clusters, 
C
 
is molar concentration, and 
a
 
is concentration 
quenching scaling constant.  For K
2
Mo
6
Cl
14
,
,
, while for 
(TBA)
2
Mo
6
Cl
14
 
, 
. 
 
 
 
58
 
 
 
Figure 5.4
 
Measured quantum yield (QY) of K
2
Mo
6
Cl
14
 
and (TBA)
2
Mo
6
Cl
14
 
in 
acetonitrile as a function of concentration; solid lines are fits to Equation. 1. (inset) QY of 
the TBA film as a function of time under shelf
-
life conditions.
 
Initial shelf
-
life testing also indicate that the inorganic phosphorescent luminophore
-
blend
s are highl
y stable as shown in Figure 5
.
4
, and as has been indicated elsewhere for 
anologous cluster
-
host systems used in oxyen sensors.
[
98
]
  
In Figure 5.3
(b
) we show the 
assembled transparent LSC with a single
-
edge mounted Si 
solar cell incorporating the 
TBA luminophore where the measured EQE of the attached Si
 
solar cell is shown in 
Figure 5.3
 
for reference. 
 
59
 
 
 
Figure 5.5
 
External quantum efficiency (
EQE
) of the LSC syste
m as a function of 
wavelength; 
(inset) measured 
EQE 
at 
325nm as a function of geometric gain.
 
The spectrally resolved EQE of the overall LS
C system is shown in Figure 5.5
, which 
exhibits a cut
-
off at the edge of visible spectrum around 400
-
430nm, matching the 
a
bsorption spectrum (see Figure 5.3
(a)). The EQE at
 
325nm as a function of the 
geometric gain 
G
 
(the ratio of the area of the concentrator and the attached solar cells) 
remains unchanged, within error. 
 
60
 
 
 
Figure 5.6
 
Current density as a function of voltage for the fully assembled LSC systems 
with two of th
e luminophores.  
 
Shown in Figure 5.6
 
is the current voltage characteristic of the 2.5cm
 
x
 
2.5cm 
x
 
0.1cm 
LSC system for two edge mounted Si solar PVs (see Experimental section for details); 
 
 
 
61
 
 
 
Figure 5.7
  
Light intensity dependence of the 
Voc
, 
FF
 
and power efficiency of the Si cell.    
 
the measured intensity dependence of th
e Si PVs is shown in Figure 5.7
 
and subsequently 
utilized in device
-
scaling modeling.  The measured short circuit current density (
Jsc
) of 
the overall system under 1.0±0.1 sun
 
was 1.3±0.1mA/cm
2
, with an open circuit voltage 
(
Voc
) of 0.52±0.01V and a fill factor of 0.65±0.02 leading to an efficiency 0.44±0.05% 
for the TBA luminophore.  To confirm the measured photocurrent results, we integrate 
the product of the EQE and the 
AM1.5G solar spectrum resulting in a short circuit current 
density of 1.5±0.1mA/cm
2
. The corresponding average visible transmittance and color 
rendering index
[
83
]
 
for the LSC system is 
84% ±1% and 94, respectively, compared to 
90%±1% and 100 for the glass substrate alone.   
 
62
 
 
 
Figure 5.8
 
Calculation of the overal
 
LSC system efficiency as a function of plate size and 
luminophore concentration, without (solid lines) and with (dotted lines) the intensity 
dependence of the Si PV.  Note that the solid lines are calculated using the 1
-
sun Si PV 
paramaters and that the p
late length of 0.2m roughly translates to a luminescence 
equivalent intensity of 1sun around the edges.  The system efficiency is impacted by the 
luminophore concentration accounting for the solar absorption efficiency, reabsorption 
loss, and the quantum y
ield.   Improving the quantum yield to 100% would result in a 
peak system efficiency of up to 1.3
-
1.5%.  
 
 
In Figure 5.8
 
we show area
-
scaling calculations of the system efficiency as a function 
of square plate size.  These simulations are, in part, validat
ed by the EQE dista
nce 
dependence shown in Figure 5.5
.  For the 2.5cm
 
x
 
2.5cm 
x
 
0.1cm substrate, where 
G 
= 6, 
the light intensity that is absorbed by the Si cell is approximately equivalent in 
monochromatic intensity to 0.3 suns. In this case, the calculat
ed efficiency is 0.5%, which 
is in good agreement to the measured value.  For 
G > 
250, the estimated efficiency of this 
TLSC system peaks at ~ 1.0%. Due to the massive Stokes shift, the calculated LSC 
63
 
 
performance remains high even for sheets as large as 1
-
10m
2
.  Accounting for the 
intensity dependent Si PV performance, the efficiency is expected to increase slightly 
(with a constant optical efficiency) as the LSC size increases due to increased intensity 
concentration around the edges, while the optical eff
iciency starts to decrease for LSC 
sizes > 10 m
2
 
due to reabsorption losses.  
 
 
 
Figure 5.9
 
C
omparison of transparent UV
-
absorbing luminescent solar concentrators 
with 
traditional
 
color LSC
s
.
 
 
 
5
.4
. Discussion
 
5
.4.1 
Nanocluster p
hotophysics
 
The presence of a non
-
zero 
 
from the fits in Figure 5.4
 
implies that individual 
nanoclusters have non
-
radiative modes that weakly compete with radiative rates.  At 
higher concentrations, the quantum yield decreases with increasing concentration, 
suggesting that particle
-
particle interactions further lead to ex
cited state (non
-
radiative) 
64
 
 
quenching, common to many organic dye systems.
[
100
]
 
Therefore complet
e particle 
isolation is important to realize the full potential of these devices.  The triplet excited 
state of organic compounds can also be efficiently deactivated by the presence of oxygen. 
[
101
-
103
]
 
In our case, we tested the luminescent in absence of oxygen so quenching between 
luminophores is indeed the dominant mechanism.  Accordingly, it becomes necessary to 
use thicker layers of lower concentration to maintain both high absorption efficiency and
 
high quantum yield, following the Beer
-
Lambert 
law. 
Therefore, the balance between 
concentration and thickness on the overall absorption efficiency is linearly related and 
this relationship was utilized as a design criteria. For example, the critical conc
entration 
(
C
C
) where the quantum yield starts to drop is 
C
C
 
= 1.0mg/mL for the (TBA)
2
Mo
6
Cl
14
 
cluster in PBMMA
-
PEMA blend, translating to a required blend thickness of 200µm to 
maintain a high UV absorption efficiency.   
 
    
In contrast to other luminescen
t systems, these are very high quantum yields for such 
large Stokes shift.  Typical quantum yields (QY) for down
-
converting chromophores 
(with Stokes shift < 200nm and visible emission/absorption) now readily reach values 
approaching 100% for a variety of 
materials including organic phosphors, organic 
fluorophores, and colloidal quantum dots.
[
69
, 
104
, 
105
]
  
Up
-
converters or anti
-
Stokes 
materials still have notably low luminescent efficiencies < 
5%.
[
106
]
  
While a limited 
number of chromophores with  massive Stokes shift have been demonstrated, the 
quantum yields of these materials have
 
been notably limited at < 1
-
2% or typically have 
emission centered in the visible part of the spectrum.
[
71
, 
107
, 
108
]
 
 
65
 
 
 
5
.4.2 
LSC d
esign
 
   
For the Si modules used here, the 
monochromatic
 
efficiency translates to 
 
 
= 19% 
under 1 sun equivalent illumination for the emission centered at 800nm.  A range of other 
solar cells could be utilized to reduce additional thermal losses.  For example, 
comparing 
the 1
-
sun solar cell parameters for several PV technologies
,
 
[
109
]
 
we estimate the LSC 
system
 
efficiency could be improved up to 1.3% with GaAs cell attached at the peak gain.   
Due to the monochromatic emission nature of the LSC, only single junction PVs can be 
considered around each individual LSC, limiting the overall system efficiency (without
 
LSC stacking) to that below the Shockley
-
Queisser limit.  
However, cost and availability 
will ultimately drive the cell design.  For example, GaAs cells could boost the efficiency 
for 
 
over 30% compared to Si, particularly for the LSCs demonstrated here,
 
but the 
estimated costs for these cells are more than double or triple that of Si.
[
109
]
 
Accordingly, 
we chose to couple lower
-
cost Si PVs with AM1.5G solar efficiency of 14
-
16% for our 
proof
-
of
-
principle demonstration, which can be further improved as cheaper, higher 
bandgap cells (GaAs, and GaInP) become available. 
 
    
Re
-
absorption losses
 
are critically dependent on the overlap of the emission and 
absorption spectrum and the overall dimension of the waveguide.  
It has long been 
recognized that LSCs are most limited by reabsorption losses, particularly for larger plate 
sizes.  Indeed, much 
of the research with LSCs has focused on the reduction of these 
reabsorption losses through 1) increasing Stokes shifts with organic phosphors,
[
12
]
 
2) 
multiple dye optimization to artificially increase the Stokes shift
[
12
, 
73
, 
110
]
 
or 3) reso
nance 
shifting limited to neat
-
film dye layers less than several microns thick.
 
[
111
]
   
The clearest 
66
 
 
path to fully maxi
mize the large
-
area scaling of these devices is through incorporation of 
high quantum
-
yield massive Stokes shift (MSS) luminophores.
[
68
]
 
With the use of the 
MSS luminophores developed here, there is essentially no overlap between the absorption 
and emission spectrum 
and therefore little reabsorption loss. Accordingly, these LSCs 
can be expanded to very large sizes (> 1m
 
x
 
1m) provided the substrates are sufficiently 
transparent in the NIR.
[
71
]
 
This is highlighted in Figure 5.8
, w
here at lower 
nanocluster 
concentrations the system (and optical) efficiencies only begin to drop off due to 
reabsorption loses for plate sizes > 10m 
x
 
10m, which 
is larger than most typical 
windows
.  These results indicate a clear route to large
-
area TLSC scalability.  Despite the 
low relative efficiency for these types of TLSCs (maximum efficiency of approximately 
2%), the ability to maintain this efficiency to very large areas with high defect tol
erances 
provides tremendous potential for window applications.  
 
5.4.3
  
Understanding blend configuration via XPS
  
 
 
       
There still remain questions about the precise nature of chemical interactions 
between luminophore and host that lead to quantum yie
ld enhancements.    X
-
ray 
Photoemission 
Spectroscopy (XPS) 
could be used to understand the nature of the 
chemical bonding that leads to enhanced/suppressed luminescence efficiency in these 
nanocluster and molecular systems.   For example, we have already 
found that the 
nanocluster salts of tetra butyl ammonium coupled with particular hosts lead to enhanced 
QYs, where this could be due to polymer anchoring of the entire nanocluster salt to create 
better isolation or could act t
o detach
 
the cation from the a
nionic nanocluster via 
anchoring of the cation.    
With XPS a narrow line width X
-
ray is used to probe a 

-
ejected core electrons the 
67
 
 
atomic composition of the surface for all elements excluding 
H and He is determined.  
Because the sampling depth is limited to several monolayers on the surface, chemical 
information regarding the composition of a surface and how it may interact with various 
interfaces can be determined without the dilution by the b
ulk of a material as seen with 
many other forms 
of spectroscopy
.  
 
 
5
.
5
 
Conclusion
 
    
In conclusion, we have designed and fabricated novel transparent luminescent solar 
concentrators devices composed of phosphorescent metal halide nanocluster
s
 
with a 
powe
r efficiency of 0.44% ±0.05%  (projected to reach 1.0%  for 
G
 
> 250), system 
external quantum efficiency of 60%, and an average transmittance of 84% ±1%, that are 
achieved through optimization of 
nanocluster
-
host interactions.  The near perfect 
absorption 
cutoff at the edge of the visible spectrum (430nm) and the massive Stokes 
shift to the near
-
infrared (800nm) of these nanoclusters allows for efficient and selective 
harvesting of ultraviolet (UV) photons, improved reabsorption efficiency and non
-
tinted 
tr
ansparency in the visible spectrum.  We show through both experiments and modeling 
that this architecture can exhibit a sustained power conversion of 0.5
-
1.5% over module 
areas > 1 m
2
 
with an upper efficiency limit of 5%
. These novel concentrators present 
an 
entirely new 
paradigm 
for very large
-
area, non
-
tinted, and highly
-
adoptable solar 
windows that can translate into improved building efficiency, enhanced UV
-
barrier layers, 
and lower cost solar harvesting systems.
 
68
 
 
C
hapter 6
 
NIR
-
absorbing Transparent Lum
inescent Solar Concentrators
 
 
6
.
1 Introduction
 
In this chapter, w
e 
exploit 
the structured absorption of 
organic excitonic 
semiconductors
 
 
to produce near
-
infrared (NIR) LSC architectures that selectively harvest 
NIR photons by waveguiding deeper
-
NIR 
luminophore emission to high efficiency 
segmented solar cells. These transparent NIR LSCs can eliminate the visual impact and 
minimize the amount of expensive solar materials required while extending the photon 
harvesting range into the NIR.  
 
In the previ
ous chapter
, ultraviolet (UV) harvesting hexanuclear metal halide clusters 
with downconverted emission into the NIR were demonstrated for UV TLSCs.
[
72
]
 
The 
large phosphorescent Stokes shift and high quantum yield allowed for power conversion 
efficiency of > 0.5% over large module area (m
2
) but these LSCs are ultimately l
imited to 
efficiencies 
up to 
5% due to the limited UV fraction in the solar spectrum.  To increase 
the overall potential of these systems, we look to selectively harvest NIR photons, where 
there is a substantially greater fraction of the solar photon flux 
(~74%). NIR fluorescent 
dyes, especially phthalocyanines, cyanines, and squaraine dyes have been widely used in 
fluorescence microscopy, bioimaging, organic light emitting diodes and other light 
emission applications. However, the quantum efficiency has mo
stly been limited to < 40% 
and most exhibit visible absorption. 
[
11
2
-
116
]
  
Previous research on NIR
-
emitting LSCs 
employing inorganic compounds such as semiconducting quantum dots and nanocrystals 
as active materials typically have improved quantum yields
[
33
, 
41
, 
117
-
119
]
 
but also present 
69
 
 
continuous band absorbance (with only minor excitonic features near the band edge).  
Accordingly, these systems all exhibit visible absorption or coloring despite emitting NIR.
 
   
 
In this 
chapter
 
we demonstrate the first transparent NIR
-
absorbing luminescent solar 
concentrator with high transparency and minimal tinting.  
We focus on the development 
of luminophore blends of cyanine and cyanine salts and have synthesized cyanine 
salt
-
host blends with quantum yields of > 20%, combined with spectrally
-
selective NIR 
harvesting.  We show the impact of Stokes shift on the ultimate performance and 
demonstrate transparent power conversion efficiencies > 0.
8
%.   
The near
-
infrared 
TLSCs
 
ba
sed on organic salts provide
 
an alternative strategy for transparent solar 
harvesting systems that can ultimately enhance the overall system efficiency of combined 
UV and NIR TLSCs.
 
 
6
.2 Experimental s
ection 
 
6
.2.1 
Organic salt solution preparation:  
 
    
1
-
(6
-
(2,5
-
dioxopyrrolidin
-
1
-
yloxy)
-
6
-
oxohexyl)
-
3,3
-
dimethyl
-
2
-
((E)
-
2
-
((E)
-
3
-
((E)
-
2
-
(1,3,3 trimethylindolin
-
2
-
ylidene)ethylidene)cyclohex
-
1
-
enyl)vinyl)
-
3H
-
indolium 
chloride
 
(CY) (Lumiprobe) and 2
-
[7
-
(1,3
-
dihydro
-
1,3,3
-
trimethyl
-
2H
-
indol
-
2
-
ylidene)
-
1,3, 5
-
he
ptatrienyl]
-
1,3,3
-
trimethyl
-
3H
-
indolium iodide (HITCI) (Exciton) were 
characterized as received without further purification. Solutions for optical 
characterization were prepared by directly dissolving each compound in dichloromethane 
at various concentrat
ions up to 4mg/ml.
 
70
 
 
6
.2.2
 
Module fabrication: 
 
    
 
A 5mg/L CY dichloromethane solution was mixed with (
poly)
-
butyl methacrylate
-
co
-
methyl methacrylate (PBMMA)
 
(Sigma
-
Aldrich) at a volume ratio of 1:1, yielding the 
target dye concentration (5mg/L) in the polymer composite film. This mixture was drop 
cast on 2cm 
x
 
2cm x 0.1cm (for efficiency measurements) or 7cm x 7cm x0.1cm (for 
EQE measurements) glass substrate
s and allowed to dry for 2 hours for each layer, and 
repeated 3 times, resulting in a layer thickness of approximately 1mm. On the mounted 
edge, a laser
-
cut, 7cm 
x
 
0.1cm Si cell (Narec Solar) with an efficiency of 15±1% @ 1 sun 
was attached within a clip t
hat fit around the edge of the waveguide and film.   
 
6
.2.3
 
Electronic structure c
alculations: 
 
All DFT calculations
,
 
done 
in collaboration by
 
the Levine group at MSU,
 
were 
performed with the Q
-
Chem 
ab initio
 
electronic structure 
package 
[
120
-
122
]
 
while EOM
-
CCSD calculations were performed in GAMESS.
[
120
, 
121
, 
123
-
125
]
 
In order to assess the 
magnitude of the error in our CAM
-
B3LYP
[
126
]
/ 6
-
31G* calculations of the vertical 
excitation energies of the cyanine dyes, a truncated model dye: (CH
3
)
2
N(CH)
9
N(CH
3
)
2
+
 
was evaluated.  This dye contains the same 9
-
carbon polymethine
 
chain terminated by 
tertiary amines as HITC
+
 
and CY
+
, but smaller dimethylamine groups replace the larger 
heterocycles to make more accurate quantum chemical calculations feasible.  The vertical 
excitation energy of this model was calculated at the same T
DDFT level of theory as 
applied to the larger dyes and with the more accurate equation
-
of
-
motion coupled cluster 
singles and doubles (EOM
-
CCSD) level of theory 
[
127
]
 
with the cc
-
pVDZ basis set.
[
128
]
  
The computed vertical excitation energies at the TDDFT and EOM
-
CCSD levels of 
71
 
 
theory are 3.09 and 2.40 eV, respectively.  The difference, 0.69 eV, is subtracted from the 
TDDFT
-
computed vertical excitation energ
ies of the larger dyes to estimate the more 
accurate EOM
-
CCSD excitation energies of these dyes.  Note that the lowest excitation in 

lowest excitations in HITC and CY.
 
6
.
3
 
Results
 
    
We focus on two promising cyanine derivatives: 
2
-
[7
-
(1,3
-
dihydro
-
1,3,3
-
trimethyl
-
2H
-
indol
-
2
-
ylidene)
-
1,3,5
-
heptatrienyl]
-
1,3,3
-
trimethyl
-
3H
-
indolium (HITC) iodide (HITCI) 
and 
1
-
(6
-
(2,5
-
dioxopyrrolidin
-
1
-
yloxy)
-
6
-
oxohexyl)
-
3,3
-
dimethyl
-
2
-
((E)
-
2
-
((E)
-
3
-
((E)
-
2
-
(1,3,3
-
trimethylindolin
-
2
-
ylidene)ethylidene)cyclohex
-
1
-
enyl)vinyl)
-
3H
-
indolium 
chloride (CY) and utilize these systems to explore the impact of the Stokes shift
.  The CY 
an
d HITCI molecular 
structures are shown in Figure 6
.2
. The absorption spectra peaks at 
742±1 nm for CY and 733±1 nm for HITCI with little visible absorption, and NIR 
emission peaks at 772±1 nm and 753±1 nm for CY and HITCI, respectively.  The Stokes 
shift, 
defined as the wavelength difference between the absorption and emission peaks, is 
30±2 nm for CY and 20±2 nm for HITCI.  The Stokes shift of the two materials helps 
demonstrate the difference in assembled LSCs performance and is also an important 
paramete
r to predict large
-
area scalability.
[
12
]
 
72
 
 
 
Figure 6.1
 
(a) Schematic of the transparent luminescent solar concentrator (b) 
Photograph of the transparent LSC system incorporating CY 
luminophore
 
in front of a 
picture of beach scenery.    
 
 
73
 
 
 
Figure 
6
.2
 
CY and HITC molecular cation structures (top); the natural excited
-
state 
transition orbital pairs for HITC (left) and CY (right).  The hole orbitals are shown on the 
top of the excited electron orbitals.
 
 
    
We explore the luminescence quantum yield 
(QY) dependence on the concentration to 
understand the photophysical behavior of these luminophores.  The measured QY and 
absorption of CY and HITCI in dichloromethane
 
solutions are shown in Figure 5.3
(b) as 
a function of luminophore concentration
. 
 
 
74
 
 
 
Fig
ure 6
.3
 
(a) Normaliz
ed absorption (circle symbols) 
and emission (square symbols) 
spectra of the NIR
-
absorbing luminophores CY (blue line) and HITCI (black line) films.  
(b)  Measured quantum yield (QY) and absorption of CY (blue line) and HITCI (black 
line
) solutions in dichloromethane as a function of concentration; solid lines are fits to 
model for the QY 
[
72
]
 
and the Beer
-
Lambert law for the absorption.
 
 
  
  
We fit the QY data of Figure 6
.3
(b) to a model with a 
concentration dependent non
-
radiative rate
[
72
]
 
and calculate the relationships between
 
intrinsic non
-
radiative rate for 
75
 
 
isolated clusters 



0,
 
rate for luminescence 


, and concentration quenching scaling 
constant 

.
[
72
]
 
For CY
, 



0 



=5.8
±0.2
, 




=1.4
±0.1 
ml/mg
, 
while for HITCI
, 



0 



=2.5
±0.2
, 




=2.2
±0.1 
ml/mg.   For both CY and HITCI, the non
-
radiative rate (



0
) 
is much larger than the radiative rate (


). The larger ratio 




 
for HITCI implies a 
slightly larger luminophore
-
luminophore interaction distance for HITCI; compared to the 
metal halide nanoclusters these concentration dependent terms are nea
rly an order of 
magnitude larger and thus require very dilute concentrations to suppress detrimental 
interactions.
[
72
]
 
The critical concentrations for CY and HITCI, defined here as the 
concentration where the QY is half of the maximum,
 
are 5mg/L and  1mg/L,  respectively
. 
In Figure 
6
.
1(b) we show the transparent LSC waveguide incorporating the CY 
luminophore. The measured intensity dependence of the assem
bled CY LSC is shown in 
Figure 6
.4
(a) which is largely dependent on the intensity dependence of the edge
-
mounted Si cells. 
 
 
            
Figure 6
.4 
(a) Light intensity dependence of the 
Voc
, 
FF
 
, power efficiency and 
responsivity of the CY TLSC device. (b) Current density as a function of
 
voltage for 
the fully assembled TLSC systems with two of the luminophores.
 
 
76
 
 
     
The spectrally resolved external quantum efficiency (EQE) of the overall LSC system 
of different p
late sizes are shown in Figure 6
.5
(a), which exhibit a peak at around 760 
nm, 
matching the absorption spectrum an
d the calculated EQE in Figure 6
.5
(b). 
 
 
            
Figure
 
6
.5
 
(a) External quantum efficiency (EQE) of the CY TLSC system as a 
function of wavelength measured from 0.02m to 0.07m, with 1cm increments.  (b) 
Calculat
ed EQE as a function of  CY LSC length in the range of 0.02m to 0.08m.  
Measured EQE (circle symbols)  are also shown here.  (inset) Calculated and measured 
EQE as a function of CY LSC length over a larger plate range.
 
77
 
 
     
Shown in Figure 6
.4
(b) are the 
current voltage characteristics of the 2cm
 
x
 
2cm 
x
 
0.1cm CY and HITCI LSC systems for two edge mounted Si solar PVs (see Experimental 
section for details). The measured short circuit current density (
Jsc
) of the overall system 
for CY under 1.0±0.1 sun was 
1.2±0.1mAcm
-
2
, with an open circuit voltage (
Voc
) of 
0.50±0.01V and a fill factor of 0.66±0.02 leading to an efficiency 0.4±0.03% for the CY 
luminophore.  The calculated short circuit current density from integrating the product of 
the EQE and the AM1.5G s
olar spectrum, is 1.0±0.1mAcm
-
2
, which is within error of the 
measured photocurrents. 
The corresponding average visible transmittance and color 
rendering index
[
83
]
 
for the CY LSC system is 
86%±1% and 94, respectively, compared to 
90%±1% and 100 for the glass substrate alone and is slightly better  in aesthetic quality 
compared to the UV
-
only TLSCs.
[
72
]
   
 
 
 
Figure 6.
6
 
Calculation of the CY TLSC optical efficiency as a function of LSC length 
with a function of simulated Stokes shift.  Stokes shift are 10nm (black, triangle), 20nm 
(black, square), 30nm (black, sphere), 50nm (gray, triangle), 80nm (gray, square), 130nm 
(gr
ay, sphere).  For reference, CY and HITCI have Stokes shifts of 20nm, and 30nm, 
respectively.   
 
 
78
 
 
    
In Figure 6
.6
 
we show area
-
scaling calculations of optical efficiency of CY LSC with 
modeled Stokes shift as a function of LSC length.  When the Stokes sh
ift (
S)
 
is below 
30nm, the critical plate length,
 
defined here as the LSC length where the optical 
efficiency is half of the maximum,
 
is around 1
-
2 cm while increasing the 
S
 
can 
significantly increase the critical plate length to > 1 m for 
S
 
> 100 nm. 
 
   
   
To explore the potential for modifying the two luminophores, the electronic structures 
of isolated HITC and CY molecular cations were optimized using density functional 
theory (DFT)
[
129
]
 
with the CAM
-
B3LYP functional 
[
126
]
 
and 6
-
31G* basis set.  The 
excitation energies, oscillator strengths, and natural transition orbitals 
[
130
]
 
were then 
calculated at the time
-
dependent (TD
-
) DFT level of theory
[
131
, 
132
]
 
in the Tamm
-
Dancoff 
approximation
[
133
]
 
using the same functional and basis set.  
 
6
.4
. Discussion
 
6
.4.1 
Luminophore
 
p
hotophysics
 
    
The individual non
-
radiative rates for luminophore are larger than the radiative rates, 
leading to moderately
-
low 
quantum yields.  This is true for many of the demonstrated 
NIR fluorophores and there continues to be significant effort to improve QY in this 
spectral range both for medical applications and light emitting diodes.
[
112
, 
113
, 
116
]
   
Here, 
the quantum yield decreases with increasing concentration due to excited state (non
-
radiative) quenching caused by particle
-
particle interactions at
 
higher concentrations 

 
these interactions persist even into dilute solutions likely through long
-
range dipole
-
dipole interactions such as Förster energy transfer due to the significant absorption
-
emission overlap. Accordingly, we design our concentrators
 
using thicker layers of dilute 
concentrations to maintain both high quantum yield and high absorption efficiency, 
79
 
 
following the Beer Lambert law
.  
Utilizing this criterion we start with a design 
concentration of 5mg/L for high quantum yield as well as a r
easonable thickness, a molar 
absorption coefficient of 1.45
 
x
 
10
8
 
Lmol
-
1
m
-
1
 
for CY at 760nm, leading to blend 
thicknesses around 1mm for near complete NIR absorption.  By utilizing thicker layers of 
dilute concentrations this also aids in reducing reabsorption losses discussed below.   
 
6
.4.2 
LSC d
esign
 
     
The efficiency of the
 
transparent LSCs is governed by: the solar spectrum absorption 
efficiency, luminophore photoluminescence efficiency, waveguide (trapping) efficiency,   
transport (re
-
absorption) efficiency
 
and solar cell quantum efficiency
.
[
96
]
 
The optical 
efficiency consists of waveguiding efficiency
, transport efficiency, and luminescence 
efficiency. The EQE consists of the optical efficiency and the quantum efficiency of the 
PV at the emission wavelength. D
ue to the monochromatic emission nature of the LSC, 
only
 
single junction PVs
 
can be attached
 
a
round each individual LSC
, which ultimately
 
limi
ts
 
the overall system efficiency without LSC stacking to that below the PV efficiency 
directly.
 
Reabsorption losses are critically dependent on the quantum yield of the 
luminophore, the Stokes shift of the dy
e emission
-
absorption, and the overall waveguide 
dimensions.
[
69
]
  
The reabsorption losses limit the performance of LSC fabricated in this 
work, due to t
he moderately low Stokes shift. For example, 
although HITCI exhibits a 
slightly higher quantum yield than CY, the CY device has enhanced LSC performance 
due to the larger Stoke shift. 
Indeed, 
it has long been recognized that LSC performance is 
often limite
d by reabsorption losses, particularly for dyes with modest 
S
 
and larger plate 
sizes. The calculated
 
optical
 
efficiency in Figure 6.6
 
shows that 
an increase of 
Stokes 
shift 
from 30nm to 80nm 
can improve the critical 
length, 
from 3cm to 30cm, where an 
80
 
 
LSC 
size of 30cm would be appropriate for many LSC applications
. To
 
fully maximize 
the large
-
area scaling of these devices, the design of high quantum yield molecules with a 
larger Stokes shift is favorable
[
68
]
 
such as with phosphorescent emitters.
[
12
]
  
However,
 
when designing appropriate fluorescent and phosphorescent luminophores for NIR LSCs 
there is a balance in considering the ideal Stokes shift since we are limited to PVs that 
have high quantum efficiency at the luminophore emission peak, and also must cons
ider 
maximizing the bandgap to minimize PV voltage losses.   With Si PVs the maximum 
Stokes shift is limited to < 200nm with the expectation of harvesting a 200
-
300nm slice 
of the NIR spectrum.  For GaAs this maximum 
S
 
is even more restricted.    
 
One appr
oach to obtain better scalability is through the improvement of the quantum 
yield closer to 100% t
hrough optimization of luminophore
-
host interactions and 
molecular design
. 
TDDFT shows that the lowest singlet transitions of HITC and CY are 
very bright, with oscillator strengths of 3.5 and 3.3, respectively.  Natural transition 

transitions loc
alized on the polymethine chain of both dyes.  NTOs corresponding to the 
dominant excitation amplitude (>0.95) for each 
dye are presented in Figure 6.2
.  The 
predicted vertical excitation energies for these transitions are 2.56 and 2.52 eV for HITC 
and CY,
 
respectively; the overestimation of the excitation energies is not surprising given 
that the range
-
corrected functional used here is known to overestimate the vertical 
excitation energies of cyanine dyes.
[
134
]
  
To determine a correction to our TDDFT vertical 
excitation energies, the highly accurate equation
-
of
-
motion coupled cluster singles and 
doubles (EOM
-
CCSD) level of theory
[
127
]
 
was applied to a similar but truncated cyanine 
dye. Comparison to the TDDFT
-
computed excitation energy of the same truncated 
81
 
 
system suggests that TDDFT overestimates the vertical excitation energies of these dyes 
by approximately 0.69 eV.  Taking this as 
a correction to the TDDFT vertical excitation 
energies of HITC and CY yields energies of 1.87 and 1.83 eV, respectively, which are in 
much better agreement with experiment.  Thus we are confident in the assignment of 
these NIR absorptions to polymethine
-
lo

Section for details of these EOM
-
CCSD calculations.) These calculations help us to 
understand molecular motif targets to manipulate bandgap, oscillator strength, radiative 
rates and potentially improve QY.
 
 
Our 
present work and the past work of other groups both point toward rational 
principles for the design of NIR cyanine emitters with larger Stokes shifts, and thus 
promising candidates to decrease in reabsorption losses.  Modified cyanine dyes which 
exhibit ef
ficient fluorescence with Stokes shifts of ~100 nm have been reported.
[
135
, 
136
]
  
These molecules, which emit from intramolecular charge transfer excited states, absorb in 
the visible rang
e and thus are not useful in the present application.  However, their 
existence demonstrates that the engineering of cyanines with intramolecular charge 
transfer states may allow us to achieve NIR emitters with large Stokes shifts through 
s
substitutions in
 
the 
central methine group with bridgehead amine groups (that could be 
applied to the molecules here)
.  Another strategy is suggested by the increased Stokes 
shift of CY compared to HITC.  The side chain or the central methine group of CY must 
be responsib
le for this increase, given that these are the only differences between the two 
luminophores.  Considering the flexibility of the side chain, we hypothesize that it may 
interact with the polymethine chain electrostatically, reorienting to stabilize the 
red
istribution of charge upon excitation in the same way that a solvent molecule would 
82
 
 
respond to excitation of a solvated dye.  Increasing the number and/or polarity of these 
side chains as well as central methine substitution are two rational strategies to 
engineer 
visibly
-
transparent NIR luminophores wit
h larger Stokes shifts, and can
 
be pursued in 
future work.
 
    
 
For materials with QY close to 100% each reabsorption event leads to another 
emission event, reducing transport losses even for small 
S
.  
However, it should be noted 
that for each absorption
-
remission event this appears as additional scattering since 
radiative emission is typically isotropic and therefore would eventually lead to greater 
front/back
-
side losses for larger plate lengths.  Prov
ided high quantum yields can be 
achieved, multiple cascaded luminophores could also be utilized to increase the effective 
Stokes shift.
[
12
, 
63
]
 
 
Another important approach to improve the scalability for low Stokes 
shift materials, is 
to embed highly
-
segmented solar cell micro
-
arrays as meshes throughout the LSC 

-

[
14
, 
137
]
 
allowing for minimal 
reabsorption losses and additional contribution from the segmented PV.  Considering this 
last approach, and combining these demonstrated NIR LSCs with the UV
-
TLSCs, 
efficiencies >1% are readily achievable, and efficiencies approaching
 
10% are possible 
with enhancements in QY, Stokes shifts, and micro
-
LSCs.    
 
Combining current UV and NIR LSCs in series enables LSC system with efficiency 
potential much above 1%. Such demonstration could allow us to realize the full 
efficiency, and larg
e
-
area potential. 
 
 
 
83
 
 
6.4.3
 
TLSC lifetime measurements 
 
    
Degradation can occur in fluorescent molecules under certain conditions due to oxygen 
exposure, photobleaching, triplet reaction (necessarily coupled with O
2
), and host 
degradation.      Many   of 
  
these   effects   are   readily   overcome,   however,   through   
material encapsulation,    purification, triplet scavenging (e.g.  incorporation  of  
cycloheptatriene, cyclooctatetraene,  nitrones,  etc.),  and  advantageous  host  pairings.    
Indeed
,  LSC  dye  lifetimes incorporating  small  molecules  have  already  been  
demonstrated  with  >  10  year  lifetime  with projections of > 20 year lifetime.  
Nonetheless, this
 
is
 
a critical parameter that must 
be characterized  and 
  
each  host,  dye,  
and  host
-
dye  combination  by  monitoring absorbance  and  edge  luminescence  as  a  
function  of  solar illumination, humidity, tem
perature and  oxygen exposure should be 
screened.
   
 
84
 
 
 
Figure 6.7
 
Lifetime test: absorption of CY7 ac
id films under different conditions as a 
function of time. Film 1 is CY7 acid 
PMMA 
film and Film 2 is CY7 acid fluoroshield 
film. 

AO

 
refers to the addition of 
an antioxidant
 
DACBO (1.4
-
Diazabicyclo[2.2.2]octane)
  
 
85
 
 
 
Figure 6.8
 
L
ifetime test: quantum yiel
d of CY7 acid film under different conditions as a 
function of time. Films are same as shown in figure 6.
7
. 
 
 
Preliminary lifetime tests show that transparent luminescent solar concentrators 
can keep high
-
level performance in long time (weeks to years). UV
 
absorbing TLSCs can 
be kept at least 1 year while NIR TLSCs can be kept around months, if kept in dark in the 
room. Oxygen is detrimental for TLSCs lifetime while capsulation can help reduce the 
impact. Some lifetime test
 
results are shown in Figure 6.7 a
nd 6.8
.
 
6.5 Demo
 
    
To demonstrate the potential of these TLSCs we fabricated modules with a plate 
length typical for a mobile electronic device that was capable of power small electronic 
devices.  
Glass substrate
s
 
with a thickness of 0.5cm
 
were 
sequentially cleaned with soap, 
deionized water, and acetone for 5 min in a sonicating bath. Substrates were then rinsed 
86
 
 
in boiling isopropanol for 5 min and oxygen plasma treated for 30s. 
Luminophore
-
host 
mixture
s were
 
drop
-
cast onto the surface of the su
bstrate and 
allowed to 
fully 
dry 
over
 
12 
hrs
.
 
Then the sample was encapsulated with a second piece of glass.
 
 
    
 
The edge of the glass 
was 
cleaned by Solar Flux pen for solar panel (Kester 186)
.  The 
PV was then g
lued
 
using 
index
-
matching epoxy 
(
AMS Tech
nology
)
.  
The demo was 
stored in an air and light free environment during the all the drying processes, but once 
encapsulated were safe to bring into air.   
 
 
Figure 6.
9
 
Photograph 
of a NIR
-
absorbing TLSC demo
.   
 
6.6 Flexib
le t
ransparent LSCs
 
     
In addition to LSC devices developed directly on glass above, we have demonstrated 
flexible LSC architectures that can enable retrofittable integration.  This is a key 
demonstration as it can open u
p many markets and applications.
[
14
]
  
We have already 
identified nearly ideal flexible substrat
e waveguide/hosts (see Figure 6.10
) that can be 
coated with luminophore films or directly embedded with the luminophores through 
87
 
 
extrusion techniques.  Additionally, we identified and integrated hi
gh efficiency flexible 
thin film GaAs cells (24.5% 1
-
sun efficiency) to make these systems high performance 
and complete flexibility
. One flexible GaAs cell is attached at the bottom of the 
waveguide with index
-
matching epoxy. The measured energy conversio
n efficiency 
reaches 
0.8%
, nearly double that of our previous glass/Si TLSC devices
.  
 
 
 
 
 
F
igure 6.10
 
Flexible transparent luminescent solar concentrators on flexible 
plastic 
substrate
s.
 
 
  
 
88
 
 
 
Figure 6.11
 
(a) Performance metrics (
Voc,
 
FF, 
efficiency and
 
responsivity) of the TLSC 
system as a function of illumination intensity
.  
(b) 
J
-
V
 
curve of the sample at different 
incident intensities. (c) 
EQE
 
as a function of wavelength measured at distance of 5, 10, 
15, 20,
 
25 mm from the GaAs cell. (d) 
Transmission 
of the overall system. 
 
The measured energy conversion efficiency and EQE (external quantum yield) of the 
flexib
le TLSCs are shown in Figure 6.11
 
as a function of illumination intensity with the 
flexible GaAs cells (2 edge mounted cells).  Not
e that the GaAs cells provide two distinct 
benefits: 1) higher voltage and overall efficiency (24% compared to 15% under 1 sun), 
and 2) greater mechanical flexibility over Si cells.  The measured short circuit current 
density (Jsc) of the overall system un
der 1.0±0.1 sun was 1.3±0.1mAcm
-
2
, with an open 
circuit voltage (Voc) of 0.93±0.03V and a fill factor of 0.57±0.02 leading to an efficiency 
of 0.70±0.02%. The peak EQE at 760nm can reach 8%, which is double that of the NIR 
TLSC we previously developed. The
 
corresponding average visible transmittance and 
color rendering index for the TLSC system is 86±1% and 95, respectively
.
 
 
8
9
 
 
6.
7 
Conclusion
 
 
    
We have designed and fabricated the first visibly
-
transparent luminescent solar 
concentrator devices which selectively harvest NIR photons based on fluorescent organic 
salts.  These transparent TLSCs exhibit
 
a non
-
tinted transparency of 84
% ±1% in the 
visible spectrum co
mbined with an efficiency of 0.8% ±0.01
% and have the potential for 
efficiencies up to 10% due to the large fraction of photon flux in the near
-
infrared.  We 
show through both experiments and modeling that the development of larger Stoke
s shift 
luminophores, optimization of luminophore
-
host interactions and fabrication of 
embedded segmented
-
PV configurations could further reduce reabsorption losses and 
increase the system efficiency over large areas.  These transparent NIR LSCs provide a 
new route to transparent light
-
harvesting systems with tremendous potential for high 
defect tolerances and processability. 
 
 
 
 
 
 
90
 
 
Chapter 7
 
T
ransparent Luminescent 
S
olar 
C
oncen
trators
 
Cost
 
 
     
One of the advantages with transparent luminescent
 
solar concentrators is the low cost 
compared with traditional expensive solar cells. In this chapter, calculations of 
manufacturing and module cost for TLSC systems are 
described and used 
to
 
estimate the 
large scale 
potential
 
for both types of TLSCs
.
 
The 
evaluation 
  
of 
LSC
 
system
 
efficiencies   for   two   of 
the   
representative   
LSC materials 
(UV Phosphor and
 
NIR  Luminophore) are  shown  in  
Figure 
7
.1 and 7.2
.    
The 
impact of the massive Stokes shift 
is highlighted in the optical efficiency   for   the   
UV and   NIR   scaling.   At moderate phosphor loading, the UV LSC can retain 
efficiencies beyond  
 
10m,   which   is larger than 
most typical windows. In contrast for
 
the NIR emitters
 
(in the absence 
of enhanced Stokes shifts described in
 
chapter 6
 
, the 

-

-
10
 
cm, defining the ideal embedded solar cell 
spacing.
 
 
91
 
 
 
Figure 7.1
 
Overall  calculated  LSC  system  efficiency  for  the  UV  only concentrator 
utilizing the spectral properties of Fig. 4, Si PVs, low
-
iron substrate with  index  1.7,  and  
quantum  yields  of  a)  40%  and  b)  80%  as  a  function  of luminophore 
conce
ntration 
embedded in a  5 m
m thin host.  Due to the massive Stoke shifts, the LSC performance 
remains high even for sheets as large as 10ft.
 
 
Figure 7.2
  
Overall  calculated  LSC  system  efficiency  for  a)  the  NIR
-
only 
concentrator,  and  b)  combined
  
with  spatially  segmented  50

m  Si  solar  cells  
utilizing  the  spe
ctral  properties,
 
low
-
iron  substrate  with  index  1.7,  and quantum 
yields of a) 40% and b) 80% as a function of luminophore concentration embedded in a 
50

m thin host.  Note that 
with the transparent NIR luminophore embedded in the 
segmented PVs, the overall system efficiency can be more than triple  that  of  the  
segmented  PVs  alone  (grey  line)  for  the  same  level  of transparency and only 
utilizing a sliver of the NIR (67
5
-
800nm).
 
 
92
 
 
7.1
 
TLSC manufacturing cost
 
The manufacturing of LSCs reduces the number of 
coatings 
to one, resulting in highly 
simplified processing analogous to car
-
paint coating, with a final additional labor step 
required to assemble the solar cells around the LSC edge or within the LSC. We estimate 
the materials costs with small
-
scale commercial pri
ces for reagents from commercial 
sources (Sigma, Dow, etc.), scaled by one
-
third to account for potential bulk material 
price reductions for both the luminophores and host chemicals. For labor and capital costs 
we assume similar practices to other thin
-
fil
m coating technology estimates (e.g. CdTe 
and OPV) adjusted for the number of processing steps: LSCs for example have 
approximately 1/6 the steps required, therefore requiring lower capital to that of OPV 
(estimated at $3.30/m
2
).
[
138
]
 
Hence a reasonable estimate for the capital cost is $0.55/m
2
, 
which is in accord with the upper limit estimate of a capital investment of roughly $50M 
for a typ
ical new 10,000,000m
2
/yr low
-
e sputtering coating facility, which translates to 
$5/m
2
. The direct labor costs are assumed to be for a semi
-
autonomous factory with three 
shifts per day (3 operators/shift) at 8hr/shift, and a wage rate of $15/hr and an avera
ge 
yearly output of 0.5km
2
 
translates to $0.84/m
2
. This is likely a conservative estimate, as 
compared to the magnitude estimates for thin
-
film production of 0.73
-
10 $/m
2
.
[
66
, 
138
]
 
Here we assume that our production will be directly integrated (in
-
line) with existing 
glass
-
manufacturing facilities so that facility costs (rent, etc.) are neg
ligible. Because of 
the room
-
temperature and low utility needs of the LSC fabrication, the utility costs 
estimated here ($0.36/m
2
) are roughly 1/6 the cost for other vacuum
-
based production 
($2.18/m
2
). The final manufacturing cost (excluding materials) of 
$2/m
2
 
is comparable to 
our estimate for automotive spray coating of $3.49/m
2
 
(excluding paint costs). A 
93
 
 
summary of the upper/lower manufacturing cost estimate for the LSCs is provided in 
Table 2.1. For comparison, the cost of typical BF33 (Schott) glass an
d low
-
iron glass 
(Saint Gobain 
-
 
Diam
a
nt, Planilux) range from $22/m
2
 
and 83/m
2
, respectively.
[
139
]
 
Because the LSCs designed here will be fabricated on glass that is already being installed 
as windows, this cost can be neglected in analyzing the incremental cost for transforming 
these surfaces into power producing surfaces, thereby eliminating the domi
nate cost for 
these systems (see Table 6.3).
 
 
Table 7.1
 
Estimated manufacturing (left) and materials costs (right) for the transparent 
LSCs by an inline coating process, integrated with a low e
-
coating facility. The glass and 
framing cost are included for
 
reference
 
 
7.2
 
TLSC m
odule cost
 
The cost of LSC systems strongly depends on the overall 
performance and 
system 
scale. The module cost, in $/WP, can be estimated as:
 
    
(
7
.1
)
 
 
94
 
 
where 
 
is the overall system efficiency, 
P
0
 
is the average solar power density 
accounting for the orientation dependent loss, 
, 
PV 
is the 1
-
sun solar cell efficiency, 
G 
= 
L 
/4
t
0
 
is the geometric gain, and 
L 
is the plate length. This mode
l is an estimate for 
the modular cost of adding solar harvesting functionality to the window that is to be 
installed, not including BOS costs. The breakdown of the BOS systems can widely vary, 
but typically run in the range of $1
-
2/WP with an approximate b
reakdown for a rooftop 
installation of $1.0/WP Racking and Prep., $0.3/WP Inverter, $0.1
-
0.3/WP Installation, 
and $0.2/WP Wiring and Transformer. 
[
138
]
 
Since as much as 1/2 of the BOS cost can be 
for installation
-
framing
-
land, it is expected the effective BOS costs could be less than 
half of that for systems integrated elsewhere. That is, under certain plate dimensions and 
configurations, the
 
module cost can be well under $1/WP with an overall installed $/WP 
with BOS to be < $1/WP. The resulting module costs are shown in 
Figure 
7.3
 
for both the 
UV and NIR focused LSCs. Importantly, the lowest cost/power ($/WP) systems are the 
ones with the low
est optical losses (larger Stoke shift material, moderate concentrations), 
not necessarily the outright highest efficiency configurations. Both the optical 
performance and cost analysis highlight the extreme benefit of the massive Stoke
-
shift 
materials bei
ng developed. Despite the limited solar flux in the UV the module cost in 
$/WP clearly can reach values approaching $0.25/ WP (at the highest performance 
metrics), indicating the levelized energy cost would be primarily dictated by the 
remaining balance of
 
systems costs. This indicates the payback period for adding such 
functionality would be minimal, particularly since the incremental cost in $/m
2
 
could be 
lower than ubiquitous multilayer low
-
e coatings. 
Given a 
total 
LSC system 
cost of $5/m
2
 
and a system efficiency of 1%, the payback time is around 
3 yrs 
if electricity cost 
is fixed 
95
 
 
at 
$0.1/kW
-
hr. 
Despite lower overall system potentials, the low
-
cost potential makes 
these UV concentrators an excellent stepping stone for initiating the solar w
indow market 
with high aesthetic
-
quality on route to implementation of higher efficiency UV and NIR 
LSCs. 
 
 
Figure 7.3
 
 
Projection for the module cost in $/WP of fabricating transparent LSCs on 
window surfaces a), b) for the UV luminophore and c), d) for 
the NIR CY luminophore 
using 
their respective 
spectral properties. Estimates are for a substrate index of n = 1.7, a 
commercially
-
available Si PV module efficiency of 18% under AM1.5G, and quantum 
yields of a), c) 0.4 and b, d) 0.8
 
and  modest Stokes shift
  
< 50nm.   
 
 
 
 
 
 
 
 
 
 
 
 
 
 
96
 
 
Chapter 8
 
Conclusions and Future Outlook
 
8
.1 
Future outlook
 
    
The aim of this work 
was to 
demonstrate transparent luminescent solar concentrator
s
 
as 
a new pathway 
for widespread window deployment that can be scaled to commercial 
window products. Such a 
technology has the benefit of
 
facile production, high defect 
tolerances, and exceptional scalability without many of drawbacks of scaling a 
transparent PV technology 
(resistive losses, partial shading losses, device yield, current 
matching, etc.).  
While the work presented so far has opened up an exciting new field and 
has demonstrated a key starting point for these new types of devices, there are a number 
of opportuni
ties yet to be explored including:  
  
 
    
  
 
8
.
1.
1
.
 
Singlet fission h
osts for UV TLSCs
 
          
Singlet fission is a process that
 
singlet excited state converts to two 
triplet
 
states
.
  
The goal here is to find a singlet fission host material, capable of producing two triplets 
for one high energy singlet, and demonstrate energy transfer coupling to
 
nanocluster 
emitters (Figure 8
.1
).  
 
97
 
 
 
Figure 8
.1
 
(a)
 
Vertical excitation 
energies of the lowest singlet (
S
1
) and triplet (
T
1
) 
excited states of three candidates for singlet fission. (b) Schematic of the singlet fission
-
energy transfer process
 
   
This would reduce the amount of required cluster and increase the limiting efficien
cy of 
the overall system by a factor of two by minimizing thermal losses in the Stoke shift from 

the nanocluster acts as the phosphorescent emitter, rather than the pr
imary absorber. 
Candidates for single fission hosts, including 9,10
-
diphenylanthracene, anthracene, 
diphenylisobenz
ofuran, phenanthrene, etc., should be
 
characterized for singlet fission 
98
 
 
yield and cluster coupling with the use of PL measurements.   With a perfectly coupled 
system, the QY of the cluster can be increased to 200%.  
 
8
.1
.2
 
Molecular design for NIR TLSCs
 
Modified cyanine dyes which exhibit efficient fluorescence with Stokes 
shifts of ~100 
nm have been reported.
[
135
]
  
These molecules, which emit from intramolecular charge 
transfer excited states, absorb in the visible range and thus are not useful in the present 
application.  However, their existence demonstrates that the engineering of cyanines with 
intramolecular ch
arge transfer states may allow us to achieve NIR emitters with large 
Stokes shifts through substitutions in the 
central methine group with bridgehead amine 
groups (that could be applied to the molecules here)
.  Another strategy is suggested by 
the increase
d Stokes shift of CY compared to HITC.  The side chain or the central 
methine group of CY must be responsible for this increase, given that these are the only 
differences between the two luminophores.  Considering the flexibility of the side chain, 
we hypo
thesize that it may interact with the polymethine chain electrostatically, 
reorienting to stabilize the redistribution of charge upon excitation in the same way that a 
solvent molecule would respond to excitation of a solvated dye.  Increasing the number 
a
nd/or polarity of these side chains as well as central methine substitution are two rational 
strategies to engineer visibly
-
transparent NIR luminophores with larger Stokes shifts, and 
should 
be pursued in future work.
 
Alternatively, nucleophilic substituti
onal attack at 
targeted Cl positions provides another pathway to break the symmetry of the molecular 
orbitals and enhance the Stokes shift
 
as shown in Figures 8.2 and 8
.3.  
 
 
99
 
 
 
Figure
 
8
.2
 
IR 780 in dimethylformamide (DMF) reacted with 1,4 benzenedithiol
 
 
 
Figure 8
.3
 
Examples of candidates for molecular design
 
for NIR TLSCs
 
Additional polymethine derivatives that 
are candidates 
for
 
Stokes shift engineer
ing 
are 
shown in Figure 8.3
. 
 
Shown in
 
F
igure 8
.4 
are examples of these new compounds with 
their 
corresponding
 
absorption.   
 
100
 
 
 
Figure 8
.4
 
 
New chemical variants with shifted 
absorption spectrum 
and modified Stokes 
shifts after reaction for 10min.    
 
 
8
.1
.3
 
Micro LSC Arrays for NIR TLSCs:
 
      
There are
 
two parallel approaches to the segmentation of the inorganic PVs for 
implantation into waveguides for the NIR
-
harvestin
g LSC:
 
i) directly position ultrathin 
bifacial Si PVs (50
-
100
 

m thick, 1mm wide) that are commercially available, in a
 
vertical orienta
tion within 50
 

m slits in a PMMA substrates so that the PV faces are 
parallel to the edge of the waveguide.  The spacing of the segmented PVs is defined by 
the reabsorption length of 
the NIR emitter.  PV wiring can
 
be explored in combinations 
of parallel and series
 
connections to maximize output
 
ii) 
import
 
laser
-
pattern Si PVs into 
meshes that are nearly invisible that are then embedded in the plane of the LSC 
101
 
 
waveguide.  Laser
-
cut openings 
can then be
 
encapsulated 
with the NIR harvester/emitter 
with the host acting as the waveguide.   We have already demonstrated that such 
patterning is feasible without loss in performance analogous to the demonstration of 


-
Stokes shift emitters that can eliminate reabsorption losses.    Simultaneously these 
components add photocurrent from solar harvesting proportional to the absorption 
fraction (1 
-
 
<
T
>), where segmented PVs create neutral filtering t
uned for overall 
transmission requirements (typically 80
-
85%).   
 
8
.1.4
 
Transparent mirror design for photon m
anagement in TLSCs
 
     
  
It has already been
 
shown that incorporation of visibly transparent, selective NIR 
reflectors (based on distributed Bragg reflector designs) in transparent photovoltaics can 
substantially improve power conversion efficiencies by 50
-
100%.  Similarly we see in 
our simulations 
that the incorporation of these coatings can improve the optical efficiency 
at low plate dimension by > 20% while reducing the quantity of dye needed by half for a 
given optical density. 
 
 
102
 
 
 
Figure 8
.5
 
(
left) Measured and simulated reflection curves for two 
transparent mirrors
 
designed to accept short
-
wave NIR and trap l
onger
-
wave NIR emission 
(shown 
schematically, middle).  (right) Schematic for the ideal mirror configuration for the NIR 
and UV LSCs used
 
to eliminate waveguiding losses.
 
 
 
For LSC sizes > 0.5m
2
 
these mirrors are likely to be helpful in mitigating any surface and 
bulk scattering imperfections that could reduce system efficiencies. Moreover, these 
coating layers are very similar to low
-
e
-
coa
tings that are already ubiquitously deployed 
and can complement or replace nearly all of their functionality for heat rejection and 
optical insulation. 
 
Figure 8.5
 
shows the mirror design for the UV
-
only and NIR
-
only 
LSCs.  For complete optimization these 
architectures
 
can 
be simulated
 
as a function of 
emission angle for simultaneously maximized performance and aesthetics.   
 
 
 
 
 
103
 
 
 
 
 
 
 
 
 
 
 
 
 
APPENDIX
 
 
 
104
 
 
A1 Quantum yield for solutions and films
 
 
Table A
1 
Quantum yield and absorption of organic luminescent 
materials in solutions
 

 
 

 
 

 

 


 


 

 

 

 

 

 
 
 

 

 

 
 

 
 

 

 
 
 
 

 

 
 
 
 

 

 

 

 

 

 

 
 
 

 

 

 

 

 

 

 

 
 
 

 

 

 

 

 

 

 

 
 

 
 

 

 
 
 
 

 

 

 

 

 

 

 
 
 

 

 

 
 

 
 

 

 
 
 
 

 

 

 

 

 

 

 
 
 




 

 

 

 

 

 

 

 
 

 

 

 

 
 

 

 

 

 
 

 

 

 

 
 

 

 

 

 
 

 

 

 

 
 

 

 

 

 
 

 

 

 

 
 

 

 

 

 

 

 

 

 

 
 
 

 

 

 
 
 

 

 

 
 

 

 

 

 
105
 
 
Table A
1
 

 
 
 

 

 

 
 
 

 

 

 

 

 

 

 

 
 
 

 

 

 
 
 

 

 

 
 

 

 

 

 
 
 

 

 

 
 
 

 

 

 

 

 

 

 

 
 

 

 

 

 
 

 

 

 

 
 
 
 
 
 

 

 

 

 

 
 

 

 

 

 
 

 

 

 

 
 
 
 
 
 

 

 

 

 

 
 

 

 

 

 
 

 

 

 

 

 

 

 

 

 
 
 

 

 

 
 
 

 

 

 
 
 

-

 

 

 
 
 

-

 

 

 
 
 

-

 

 

 
 
 

-

 

 

 
 
 

-

 

 

 
 
 

-

 

 

 
 

 

 

 
 
 
 

 

 
 
 
 

 

 
 
 
 

-

 

 
 
 
 

-

 

 
 
 
 

-

 

 
 
 
 

-

 

 
 
 
 
106
 
 
Table A
2 
 
Quantum yield and absorption for luminescent materials in films
 

 
 

 
 

 
 

 


 

 
 


 


 

 

 
 

 

 

 

 
 
 
 
 

 

 

 

 

 
 

 

 

 

 
 

 
 

 

 

 

 

 

 
 

 

 

 

 
 
 
 

 

 

 

 
 
 
 

 

 

 

 

 

 

 

 

 

 

 
 
 
 
 

 

 

 
 
 
 
 

 

 

 
 
 
 

 

 

 

 
 
 
 
 

 

 

 
 
 
 
 

 

 

 
 
 
 

 

 

 

 
 
 
 
 

 

 

 

 

 
 

 

 

 

 
 
 
 
 

 

 

 
 
 
 

 

 

 

 
 
 
 
 

 

 

 
 
 
 

 

 

 

 
 
 
 
 

 

 

 

 

 
 

 

 

 

 
 
 
 
 

 

 
 
 
 
 

 

 

 

 
 
 
 
107
 
 
 
 
 
 
 
 
 
 
BIBLIOGRAPHY
 
 
 
108
 
 
BIBLIOGRAPHY
 
 
[1]
 
S. R. Forrest, 
Nature
 
2004
, 
428
, 911.
 
[2]
 
M. C. Barr, R. M. Howden, R. R. Lunt, V. Bulovic, K. K. Gleason, 
Advanced 
Energy Materials
 
2012
, 
2
, 1404.
 
[3]
 
M. Kaltenbrunner, M. S. White, E. D. Glowacki, T. Sekitani, T. Someya, N. S. 
Sariciftci, S. Bauer, 
Nature Communications
 
2012
, 
3
.
 
[4]
 
Y. S. Park, K. H. Choi, H. K. Kim, 
Journal of Physics D
-
Applied Physics
 
2009
, 
42
.
 
[5]
 
M. A. Green, K. Emery, Y. Hishikawa,
 
W. Warta, E. D. Dunlop, 
Progress in 
Photovoltaics
 
2014
, 
22
, 701.
 
[6]
 
C. W. Tang, 
Applied Physics Letters
 
1986
, 
48
, 183.
 
[7]
 
F. Vollmer, W. Rettig, 
Journal of Photochemistry and Photobiology a
-
Chemistry
 
1996
, 
95
, 143.
 
[8]
 
I. Baumberg, O. Berezin, A. Drabki
n, B. Gorelik, L. Kogan, M. Voskobojnik, M. 
Zaidman, 
Polymer Degradation and Stability
 
2001
, 
73
, 403.
 
[9]
 
R. Kinderman, L. H. Slooff, A. R. Burgers, N. J. Bakker, A. Buchtemann, R. 
Danz, J. A. M. van Roosmalen, 
Journal of Solar Energy 
Engineering
-
Transactions of the 
Asme
 
2007
, 
129
, 277.
 
[10]
 
L. R. Wilson, B. S. Richards, 
Applied Optics
 
2009
, 
48
, 212.
 
[11]
 
J. S. Batchelder, A. H. Zewail, T. Cole, 
Applied Optics
 
1981
, 
20
, 3733.
 
[12]
 
M. J. Currie, J. K. Mapel, T. D. Heidel, S. Goffri, M. A
. Baldo, 
Science
 
2008
, 
321
, 
226.
 
109
 
 
[13]
 
J. M. Drake, M. L. Lesiecki, J. Sansregret, W. R. L. Thomas, 
Applied Optics
 
1982
, 
21
, 2945.
 
[14]
 
J. Yoon, L. F. Li, A. V. Semichaevsky, J. H. Ryu, H. T. Johnson, R. G. Nuzzo, J. 
A. Rogers, 
Nature Communications
 
2011
, 
2
, 343.
 
[15]
 
W. G. J. H. M. van Sark, K. W. J. Barnham, L. H. Slooff, A. J. Chatten, A. 
Buchtemann, A. Meyer, S. J. McCormack, R. Koole, D. J. Farrell, R. Bose, E. E. Bende, 
A. R. Burgers, T. Budel, J. Quilitz, M. Kennedy, T. Meyer, C. D. M. Donega, A. 
Meij
erink, D. Vanmaekelbergh, 
Opt Express
 
2008
, 
16
, 21773.
 
[16]
 
A. M. Hermann, 
Solar Energy
 
1982
, 
29
, 323.
 
[17]
 
L. H. Slooff, R. Kinderman, A. R. Burgers, A. Buechtemann, R. Danz, T. B. 
Meyer, A. J. Chatten, D. Farrell, K. W. J. Barnham, J. A. M. van Roosmalen
, 
Photonics 
for Solar Energy Systems
 
2006
, 
6197
.
 
[18]
 
M. G. Debije, P. P. C. Verbunt, P. J. Nadkarni, S. Velate, K. Bhaumik, S. 
Nedumbamana, B. C. Rowan, B. S. Richards, T. L. Hoeks, 
Applied Optics
 
2011
, 
50
, 163.
 
[19]
 
R. Reisfeld, M. Eyal, V. Chernyak, R. Zusman, 
Solar Energy Materials
 
1988
, 
17
, 
439.
 
[20]
 
A. F. Mansour, H. M. A. Killa, S. A. El
-
Wanees, M. Y. El
-
Sayed, 
Polymer 
Testing
 
2005
, 
24
, 519.
 
[21]
 
W. E. Benjamin, D. R. Veit, M. J. Perkins, E. Bain, K. Scharnhorst,
 
S. McDowall, 
D. L. Patrick, J. D. Gilbertson, 
Chemistry of Materials
 
2014
, 
26
, 1291.
 
[22]
 
M. A. ElShahawy, A. F. Mansour, 
Journal of Materials Science
-
Materials in 
Electronics
 
1996
, 
7
, 171.
 
[23]
 
R. F. Kubin, A. N. Fletcher, 
Journal of Luminescence
 
1982
, 
2
7
, 455.
 
[24]
 
A. F. Mansour, 
Polymer Testing
 
1998
, 
17
, 333.
 
[25]
 
A. Sanguineti, M. Sassi, R. Turrisi, R. Ruffo, G. Vaccaro, F. Meinardi, L. 
Beverina, 
Chemical Communications
 
2013
, 
49
, 1618.
 
110
 
 
[26]
 
C. Botta, P. Betti, M. Pasini, 
Journal of Materials Chemistry 
A
 
2013
, 
1
, 510.
 
[27]
 
A. P. Green, K. T. Butler, A. R. Buckley, 
Applied Physics Letters
 
2013
, 
102
.
 
[28]
 
Y. Kawamura, H. Sasabe, C. Adachi, 
Japanese Journal of Applied Physics Part 1
-
Regular Papers Short Notes & Review Papers
 
2004
, 
43
, 7729.
 
[29]
 
G. Weber, F. W. J. Teale, 
Transactions of the Faraday Society
 
1957
, 
53
, 646.
 
[30]
 
W. Holzer, A. Penzkofer, T. Tsuboi, 
Chemical Physics
 
2005
, 
308
, 93.
 
[31]
 
O. I. Micic, C. J. Curtis, K. M. Jones, J. R. Sprague, A. J. Nozik, 
Journal of 
Physical Chemistry
 
199
4
, 
98
, 4966.
 
[32]
 
D. J. Norris, M. G. Bawendi, 
Physical Review B
 
1996
, 
53
, 16338.
 
[33]
 
A. J. Chatten, K. W. J. Barnham, B. F. Buxton, N. J. Ekins
-
Daukes, M. A. Malik, 
Solar Energy Materials and Solar Cells
 
2003
, 
75
, 363.
 
[34]
 
S. J. Gallagher, B. Norton, P.
 
C. Eames, 
Solar Energy
 
2007
, 
81
, 813.
 
[35]
 
X. Y. Wang, J. Y. Zhang, A. Nazzal, M. Xiao, 
Applied Physics Letters
 
2003
, 
83
, 
162.
 
[36]
 
S. M. Reda, 
Acta Materialia
 
2008
, 
56
, 259.
 
[37]
 
A. Schuler, M. Python, M. V. del Olmo, E. de Chambrier, 
Solar Energy
 
2007
, 
81
, 
1159.
 
[38]
 
I. Coropceanu, M. G. Bawendi, 
Nano Letters
 
2014
, 
14
, 4097.
 
[39]
 
C. S. Erickson, L. R. Bradshaw, S. McDowall, J. D. Gilbertson, D. R. Gamelin, D. 
L. Patrick, 
Acs Nano
 
2014
, 
8
, 3461.
 
111
 
 
[40]
 
Z. Krumer, S. J. Pera, R. J. A. van Dijk
-
Moes, Y. M. Zhao, A. F. P. de Brouwer, 
E. Groeneveld, W. G. J. H. M. van Sark, R. E. I. Schropp, C. D. Donega, 
Solar Energy 
Materials and Solar Cells
 
2013
, 
111
, 57.
 
[41]
 
G. V. Shcherbatyuk, R. H. Inman, C. Wang, R. 
Winston, S. Ghosh, 
Applied 
Physics Letters
 
2010
, 
96
.
 
[42]
 
S. Kim, B. Fisher, H. J. Eisler, M. Bawendi, 
Journal of the American Chemical 
Society
 
2003
, 
125
, 11466.
 
[43]
 
R. G. Xie, X. H. Zhong, T. Basche, 
Advanced Materials
 
2005
, 
17
, 2741.
 
[44]
 
P. T. K. Chin, C. D. M. Donega, S. S. Bavel, S. C. J. Meskers, N. A. J. M. 
Sommerdijk, R. A. J. Janssen, 
Journal of the American Chemical Society
 
2007
, 
129
, 
14880.
 
[45]
 
Y. P. Du, B. Xu, T. Fu, M. Cai, F. Li, Y. Zhang, Q. B. Wang, 
Journal of the 
American Ch
emical Society
 
2010
, 
132
, 1470.
 
[46]
 
K. E. Knowles, T. B. Kilburn, D. G. Alzate, S. McDowall, D. R. Gamelin, 
Chemical Communications
 
2015
, 
51
, 9129.
 
[47]
 
L. R. Bradshaw, K. E. Knowles, S. McDowall, D. R. Gamelin, 
Nano Letters
 
2015
, 
15
, 1315.
 
[48]
 
J. A. Levitt, W. H. Weber, 
Applied Optics
 
1977
, 
16
, 2684.
 
[49]
 
R. Reisfeld, Y. Kalisky, 
Chemical Physics Letters
 
1981
, 
80
, 178.
 
[50]
 
T. Jin, S. Inoue, K. Machida, G. Adachi, 
Journal of the Electrochemical Society
 
1997
, 
144
, 4054.
 
[51]
 
T. Jin, S. Inoue, S. 
Tsutsumi, K. Machida, G. Adachi, 
Chemistry Letters
 
1997
, 
171.
 
[52]
 
L. N. Sun, J. B. Yu, H. J. Zhang, Q. G. Meng, E. Ma, C. Y. Peng, K. Y. Yang, 
Microporous and Mesoporous Materials
 
2007
, 
98
, 156.
 
112
 
 
[53]
 
K. A. White, D. A. Chengelis, K. A. Gogick, J. Stehman,
 
N. L. Rosi, S. Petoud, 
Journal of the American Chemical Society
 
2009
, 
131
, 18069.
 
[54]
 
J. Zhang, P. D. Badger, S. J. Geib, S. Petoud, 
Angewandte Chemie
-
International 
Edition
 
2005
, 
44
, 2508.
 
[55]
 
A. Sanguineti, A. Monguzzi, G. Vaccaro, F. Meinardi, E. Ronc
hi, M. Moret, U. 
Cosentino, G. Moro, R. Simonutti, M. Mauri, R. Tubino, L. Beverina, 
Physical 
Chemistry Chemical Physics
 
2012
, 
14
, 6452.
 
[56]
 
H. Bengtson,  2011.
 
[57]
 
L. H. Slooff, E. E. Bende, A. R. Burgers, T. Budel, M. Pravettoni, R. P. Kenny, E. 
D. Dun
lop, A. Buchtemann, 
Physica Status Solidi
-
Rapid Research Letters
 
2008
, 
2
, 257.
 
[58]
 
F. Galluzzi, E. Scafe, 
Solar Energy
 
1984
, 
33
, 501.
 
[59]
 
W. H. Weber, J. Lambe, 
Applied Optics
 
1976
, 
15
, 2299.
 
[60]
 
A. Goetzberger, W. Greubel, 
Applied Physics
 
1977
, 
14
, 
123.
 
[61]
 
V. Wittwer, W. Stahl, A. Goetzberger, 
Solar Energy Materials
 
1984
, 
11
, 187.
 
[62]
 
G. Smestad, H. Ries, R. Winston, E. Yablonovitch, 
Solar Energy Materials
 
1990
, 
21
, 99.
 
[63]
 
J. C. Goldschmidt, M. Peters, A. Bosch, H. Helmers, F. Dimroth, S. W. Glu
nz, G. 
Willeke, 
Solar Energy Materials and Solar Cells
 
2009
, 
93
, 176.
 
[64]
 
M. A. Faist, T. Kirchartz, W. Gong, R. S. Ashraf, I. McCulloch, J. C. de Mello, N. 
J. Ekins
-
Daukes, D. D. C. Bradley, J. Nelson, 
Journal of the American Chemical Society
 
2012
, 
134
, 
685.
 
[65]
 
J. Bomm, A. Buchtemann, A. J. Chatten, R. Bose, D. J. Farrell, N. L. A. Chan, Y. 
Xiao, L. H. Slooff, T. Meyer, A. Meyer, W. G. J. H. M. van Sark, R. Koole, 
Solar Energy 
Materials and Solar Cells
 
2011
, 
95
, 2087.
 
113
 
 
[66]
 
G. Smestad, C. Bignozzi, R. Ar
gazzi, 
Solar Energy Materials and Solar Cells
 
1994
, 
33
, 253.
 
[67]
 
L. Desmet, A. J. M. Ras, D. K. G. de Boer, M. G. Debije, 
Optics Letters
 
2012
, 
37
, 
3087.
 
[68]
 
T. X. Wang, J. Zhang, W. Ma, Y. H. Luo, L. J. Wang, Z. J. Hu, W. X. Wu, X. 
Wang, G. Zou, Q. J. Zh
ang, 
Solar Energy
 
2011
, 
85
, 2571.
 
[69]
 
M. G. Debije, P. P. C. Verbunt, 
Advanced Energy Materials
 
2012
, 
2
, 12.
 
[70]
 
Wikipedia.
 
[71]
 
L. R. Wilson, 
Ph.D. Thesis
, Heriot
-
Watt University, May 
2010
.
 
[72]
 
Y. Zhao, R. R. Lunt, 
Advanced Energy Materials
 
2013
, 
3
, 
1143.
 
[73]
 
S. T. Bailey, G. E. Lokey, M. S. Hanes, J. D. M. Shearer, J. B. McLafferty, G. T. 
Beaumont, T. T. Baseler, J. M. Layhue, D. R. Broussrd, Y. Z. Zhang, B. P. Wittmershaus, 
Solar Energy Materials and Solar Cells
 
2007
, 
91
, 67.
 
[74]
 
R. W. Olson, R. F
. Loring, M. D. Fayer, 
Applied Optics
 
1981
, 
20
, 2934.
 
[75]
 
N. C. Giebink, G. P. Wiederrecht, M. R. Wasielewski, 
Nature Photonics
 
2011
, 
5
, 
695.
 
[76]
 
A. L. Bhatnagar, W, 
Fundamentals of solar astronomy
,  
2005
.
 
[77]
 
D. H. W. Li, J. C. Lam, C. C. S. Lau, T. W.
 
Huan, 
Renewable Energy
 
2004
, 
29
, 
921.
 
[78]
 
A. Kerrouche, D. A. Hardy, D. Ross, B. S. Richards, 
Solar Energy Materials and 
Solar Cells
 
2014
, 
122
, 99.
 
[79]
 
S. W. Leow, C. Corrado, M. Osborn, S. A. Carter, 
High and Low Concentrator 
Systems for Solar Electric
 
Applications Viii
 
2013
, 
8821
.
 
114
 
 
[80]
 
J. W. E. Wiegman, E. van der Kolk, 
Solar Energy Materials and Solar Cells
 
2012
, 
103
, 41.
 
[81]
 
G. Kocher
-
Oberlehner, M. Bardosova, M. Pemble, B. S. Richards, 
Solar Energy 
Materials and Solar Cells
 
2012
, 
104
, 53.
 
[82]
 
F. Meinardi, A. Colombo, K. A. Velizhanin, R. Simonutti, M. Lorenzon, L. 
Beverina, R. Viswanatha, V. I. Klimov, S. Brovelli, 
Nature Photonics
 
2014
, 
8
, 392.
 
[83]
 
R. R. Lunt, 
Applied Physics Letters
 
2012
, 
101
, 043902.
 
[84]
 
N. Takada, T. Tsutsui, S. Saito, 
Ap
plied Physics Letters
 
1993
, 
63
, 2032.
 
[85]
 
J. Zheng, C. W. Zhang, R. M. Dickson, 
Physical Review Letters
 
2004
, 
93
.
 
[86]
 
M. Reers, T. W. Smith, L. B. Chen, 
Biochemistry
 
1991
, 
30
, 4480.
 
[87]
 
X. H. Zhong, Y. Y. Feng, W. Knoll, M. Y. Han, 
Journal of the Americ
an 
Chemical Society
 
2003
, 
125
, 13559.
 
[88]
 
J. A. Misewich, R. Martel, P. Avouris, J. C. Tsang, S. Heinze, J. Tersoff, 
Science
 
2003
, 
300
, 783.
 
[89]
 
R. R. Lunt, T. P. Osedach, P. R. Brown, J. A. Rowehl, V. Bulovic, 
Advanced 
Materials
 
2011
, 
23
, 5712.
 
[90]
 
D. G. Goebel, 
Applied Optics
 
1967
, 
6
, 125.
 
[91]
 
J. A. Jacquez, H. F. Kuppenheim, 
Journal of the Optical Society of America
 
1955
, 
45
, 460.
 
[92]
 
K. Suzuki, A. Kobayashi, S. Kaneko, K. Takehira, T. Yoshihara, H. Ishida, Y. 
Shiina, S. Oishic, S. Tobita, 
Physic
al Chemistry Chemical Physics
 
2009
, 
11
, 9850.
 
[93]
 
L. Porres, A. Holland, L. O. Palsson, A. P. Monkman, C. Kemp, A. Beeby, 
Journal of Fluorescence
 
2006
, 
16
, 267.
 
115
 
 
[94]
 
R. Rottgers, S. Gehnke, 
Applied Optics
 
2012
, 
51
, 1336.
 
[95]
 
V. Shrotriya, G. Li, Y. Yao, 
T. Moriarty, K. Emery, Y. Yang, 
Advanced 
Functional Materials
 
2006
, 
16
, 2016.
 
[96]
 
J. S. Batchelder, A. H. Zewail, T. Cole, 
Applied Optics
 
1979
, 
18
, 3090.
 
[97]
 
W. C. Dorman, R. E. Mccarley, 
Inorganic Chemistry
 
1974
, 
13
, 491.
 
[98]
 
D. J. Osborn, G. L. Baker, R. N. Ghosh, 
Journal of Sol
-
Gel Science and 
Technology
 
2005
, 
36
, 5.
 
[99]
 
M. Newsham, 
PhD Thesis
, Michigan State University,  
1988
.
 
[100]
 
J. Kalinowski, 
Organic Light
-
Emitting Diodes: Principles, Characteristics, and 
Processes.
, CRC Press,  
2005
.
 
[101]
 
J. R. Lakowicz, 
Principles of fluorescence spectroscopy
, Springer,  
2006
.
 
[102]
 
R. V. Ziolkowski, K.; Kamat.P.V., 
Langmuir
 
1997
,
 
13 
3124.
 
[103]
 
Y. Amao, K. Asai, I. Okura, 
Journal of Porphyrins and Phthalocyanines
 
2000
, 
4
, 
292.
 
[1
04]
 
H. Lin, D. Q. Chen, Y. L. Yu, Z. F. Shan, P. Huang, A. P. Yang, Y. S. Wang, 
Journal of Alloys and Compounds
 
2011
, 
509
, 3363.
 
[105]
 
X. A. Y. Huang, X. A. H. Ji, Q. Y. A. Zhang, 
Journal of the American Ceramic 
Society
 
2011
, 
94
, 833.
 
[106]
 
C. Strumpel, M.
 
McCann, G. Beaucarne, V. Arkhipov, A. Slaoui, V. Svrcek, C. 
del Canizo, I. Tobias, 
Solar Energy Materials and Solar Cells
 
2007
, 
91
, 238.
 
[107]
 
B. C. Rowan, B. S. Richards, 
Proceedings of the 23rd European Photovoltaic 
Solar Energy Conference and 
Exhibition, Valencia,
 
2008
, 700.
 
116
 
 
[108]
 
O. Moudam, B. C. Rowan, M. Alamiry, P. Richardson, B. S. Richards, A. C. 
Jones, N. Robertson, 
Chemical Communications
 
2009
, 6649.
 
[109]
 
M. A. Green, K. Emery, Y. Hishikawa, W. Warta, E. D. Dunlop, 
Progress in 
Photovol
taics
 
2012
, 
20
, 12.
 
[110]
 
G. Calzaferri, R. Meallet
-
Renault, D. Bruhwiler, R. Pansu, I. Dolamic, T. Dienel, 
P. Adler, H. R. Li, A. Kunzmann, 
Chemphyschem
 
2011
, 
12
, 580.
 
[111]
 
N. C. Giebink, G. P. Wiederrecht, M. R. Wasielewski, 
Nature Photonics
 
2011
, 
5
, 
69
4.
 
[112]
 
V. J. Pansare, S. Hejazi, W. J. Faenza, R. K. Prud'homme, 
Chemistry of Materials
 
2012
, 
24
, 812.
 
[113]
 
J. O. Escobedo, O. Rusin, S. Lim, R. M. Strongin, 
Current Opinion in Chemical 
Biology
 
2010
, 
14
, 64.
 
[114]
 
G. M. Fischer, M. Isomaki
-
Krondahl, I. 
Gottker
-
Schnetmann, E. Daltrozzo, A. 
Zumbusch, 
Chemistry
-
a European Journal
 
2009
, 
15
, 4857.
 
[115]
 
R. Weissleder, 
Nature Biotechnology
 
2001
, 
19
, 316.
 
[116]
 
J. C. Ostrowski, K. Susumu, M. R. Robinson, M. J. Therien, G. C. Bazan, 
Advanced Materials
 
2003
, 
15
, 
1296.
 
[117]
 
K. Barnham, J. L. Marques, J. Hassard, P. O'Brien, 
Applied Physics Letters
 
2000
, 
76
, 1197.
 
[118]
 
M. H. V. Werts, J. W. Hofstraat, F. A. J. Geurts, J. W. Verhoeven, 
Chemical 
Physics Letters
 
1997
, 
276
, 196.
 
[119]
 
V. Sholin, J. D. Olson, S. A. Carter, 
Journal of Applied Physics
 
2007
, 
101
.
 
[120]
 
M. W. Schmidt, K. K. Baldridge, J. A. Boatz, S. T. Elbert, M. S. Gordon, J. H. 
Jensen, S. Koseki, N. Matsunaga, K. A. Nguyen, S. J. Su, T. L. Windus, M. Dupuis, J. A. 
Montgom
ery, 
Journal of Computational Chemistry
 
1993
, 
14
, 1347.
 
117
 
 
[121]
 
M. S. Gordon, M. W. Schmidt, in 
Theory and Applications of Computational 
Chemistry: the First Forty Years
,  (Eds: C. E. Dykstra, G. Frenking, K. S. Kim, G. E. 
Scuseria), Elsevier, Amsterdam 2005
, 1167.
 
[122]
 
Y. Shao, L. F. Molnar, Y. Jung, J. Kussmann, C. Ochsenfeld, S. T. Brown, A. T. 
B. Gilbert, L. V. Slipchenko, S. V. Levchenko, D. P. O'Neill, R. A. DiStasio, R. C. 
Lochan, T. Wang, G. J. O. Beran, N. A. Besley, J. M. Herbert, C. Y. Lin, T. Van
 
Voorhis, 
S. H. Chien, A. Sodt, R. P. Steele, V. A. Rassolov, P. E. Maslen, P. P. Korambath, R. D. 
Adamson, B. Austin, J. Baker, E. F. C. Byrd, H. Dachsel, R. J. Doerksen, A. Dreuw, B. D. 
Dunietz, A. D. Dutoi, T. R. Furlani, S. R. Gwaltney, A. Heyden, S. H
irata, C. P. Hsu, G. 
Kedziora, R. Z. Khalliulin, P. Klunzinger, A. M. Lee, M. S. Lee, W. Liang, I. Lotan, N. 
Nair, B. Peters, E. I. Proynov, P. A. Pieniazek, Y. M. Rhee, J. Ritchie, E. Rosta, C. D. 
Sherrill, A. C. Simmonett, J. E. Subotnik, H. L. Woodcock,
 
W. Zhang, A. T. Bell, A. K. 
Chakraborty, D. M. Chipman, F. J. Keil, A. Warshel, W. J. Hehre, H. F. Schaefer, J. 
Kong, A. I. Krylov, P. M. W. Gill, M. Head
-
Gordon, 
Physical Chemistry Chemical 
Physics
 
2006
, 
8
, 3172.
 
[123]
 
P. Piecuch, S. A. Kucharski, K. Kow
alski, M. Musial, 
Computer Physics 
Communications
 
2002
, 
149
, 71.
 
[124]
 
K. Kowalski, P. Piecuch, 
Journal of Chemical Physics
 
2004
, 
120
, 1715.
 
[125]
 
M. Wloch, J. R. Gour, K. Kowalski, P. Piecuch, 
Journal of Chemical Physics
 
2005
, 
122
.
 
[126]
 
T. Yanai, D. P. Tew, N. C. Handy, 
Chemical Physics Letters
 
2004
, 
393
, 51.
 
[127]
 
J. F. Stanton, R. J. Bartlett, 
Journal of Chemical Physics
 
1993
, 
98
, 7029.
 
[128]
 
T. H. Dunning, 
Journal of Chemical Physics
 
1989
, 
90
, 1007.
 
[129]
 
W. Kohn, L. J. Sham, 
Physical 
Review
 
1965
, 
140
, 1133.
 
[130]
 
R. L. Martin, 
Journal of Chemical Physics
 
2003
, 
118
, 4775.
 
[131]
 
E. K. U. Gross, W. Kohn, 
Physical Review Letters
 
1985
, 
55
, 2850.
 
118
 
 
[132]
 
M. E. Casida, C. Jamorski, K. C. Casida, D. R. Salahub, 
Journal of Chemical 
Physics
 
1998
, 
108
, 4439.
 
[133]
 
S. Hirata, M. Head
-
Gordon, 
Chemical Physics Letters
 
1999
, 
314
, 291.
 
[134]
 
D. Jacquemin, E. A. Perpete, G. Scalmani, M. J. Frisch, R. Kobayashi, C. Adamo, 
Journal of Chemical Physics
 
2007
, 
126
.
 
[135]
 
X. J. Peng, F. L. Song, E. Lu, Y. N. Wan
g, W. Zhou, J. L. Fan, Y. L. Gao, 
Journal 
of the American Chemical Society
 
2005
, 
127
, 4170.
 
[136]
 
C. A. Berolino, G. Caputo, C. Barolo, G. Viscardi, S. Coluccia, 
J. Fluoresc.
 
2006
, 
16
, 221.
 
[137]
 
X. Sheng, L. Shen, T. Kim, L. F. Li, X. R. Wang, R. Dowdy, P. Froeter, K. 
Shigeta, X. L. Li, R. G. Nuzzo, N. C. Giebink, J. A. Rogers, 
Advanced Energy Materials
 
2013
, 
3
, 991.
 
[138]
 
J. Kalowekamo, E. Baker, 
Solar Energy
 
2009
, 
83
, 1224.