The Current-Voltage Trade-Off
in Organic Solar Cells and How
to Get Around It
Mark Thompson
Department of Chemistry
University of Southern California
The trade-off between VOC and JSC
-
hn
CT
CS
EDA
-
hn
CT
CS
+
EDA
+
JSC is related to the
absorption profile,
minimum energy is
related to the
HOMO-LUMO gap
VOC is related to the
energy difference
between the hole and
the electron, EDA
HOMO-LUMO gap is smaller:
collect a larger fraction of
solar spectrum, increase JSC,
BUT VOC will suffer
Device Performance of subPC acceptors
ITO/CuPc (400Å)/Acceptor (250Å)/BCP (100Å)/Al (1000Å)
D/A
EDA
(eV)
Voc
(V)
CuPc/FSubPc
1.0
0.13
CuPc/NO2SubPc
1.3
0.22
-2
CuPc/C60
1.5
0.46
-4
CuPc/ClSubPc
1.7
0.59
SubPc/C60
1.9
0.97
N
A
2
N
N
Cl
B
N
N
N
Current Density (mA/cm )
4
A
A
2
F-SubPc
NO2-SubPc
C60
Cl-SubPc
0
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
Voltage (V)
Voc (V)
FF
hp (%)
F-SubPc
NO2-SubPc
C60
Cl-SubPc
Jsc
(mA/cm2)
1.8
2.5
3.6
1.6
0.13
0.22
0.45
0.59
0.34
0.43
0.54
0.55
0.08
0.2
0.9
0.5
SubPc/C60
4.0
0.97
0.57
2.1
Acceptor
Motivation for using Pt(TPBP) in photovoltaics
V
BCP buffer layer
C60 Acceptor
Donor
hn
Absorbance
1.0
Thin Film Spectra (20 nm)
CuPC
Pt(TPBP)
Pd(TPBP)
aPt = 5.9 x 105 cm-1
aPd = 6.0 x 105 cm-1
aCuPc = 2.1 x 105 cm-1
0.5
High absorption coefficient
 long exciton lifetime: high LD ??
0.0
300
400
500
600
Wavelength (nm)
700
800
LD  D
Short Exciton Diffusion Length Drives Device Design
• Very large LD for single crystals, but need
amorphous films
• Common OPV structure and materials
vs.
– Buffer or blocking layer is commonly used
– Donor: Phthalocyanines/porphyrins, oligo and
polythiophenes, acenes (tetracene and pentacene)
– Acceptor: perylene derivatives, C60 and derivatives
– Most have relatively short exciton diffusion lengths,
and short measured exciton lifetimes
• Most OPV materials have
high ISC efficiencies
LD  D
– CuPC: 0.55, C60: > 0.9,
acenes: > 0.9
– Triplet excitons may be the
active species in OPV
Cpd.
LD (Å)
CuPC
150
C60
400
tetracene
700
PPV
120
P3HT
70
Singlet vs. Triplet Exciton
S1: D*  D+ + e1
-
-
T1: D*  D+ + e2
3
4
+
1.
2.
3.
4.
5.
D + hn  D* S1/T1
D* + D  D + D*
D* + A  [D+A-]
[D+A-]  D+ + AD+ + A-  current
5
+
absorption + ISC
exciton migration
charge transfer
charge separation
conduction
+
Motivation for using Pt(TPBP) in photovoltaics
V
BCP
C60 Acceptor
Donor
hn
-1
5
-1
2.0
Absorption
Emission 77K
Emission RT
1.5
1.0
0.8
0.6
1.0
0.4
0.5
0.2
0.0
0.0
350 400 450 500 550 600 650 700 750 800
Wavelength (nm)
Emission Intensity (a.u.)
Absorptivity (10 M cm )
Exciton energies
S1 = 1.9 eV
T1 2.5
= 1.6 eV
C60 + e-  C60-
S1: PtP*  PtP+ + eT1: PtP*  PtP+ + e-
D
A
Net: Por* + C60  Por+ + C60Driving force for exciton separation:
S1 = 0.5 eV, T1 = 0.2 eV
Pt(TPBP) OPV performance
6
Pt (TPBP)
CuPc
N
2
N
N
N Cu N
2
J (mA/cm )
4
0
N
N
N
-2
Pt(TPBP)
-4
-6
-0.2
0.0
0.2
0.4
0.6
CuPC
0.8
V(volt)
QE%
Abs., PtTPBP
Abs., C60
30
FF
CuPc
5.51
0.482
0.60
1.59
PtTPBP
4.48
0.685
0.63
1.93
0.8
QE (%)
Voc
1.0
0.6
10
0.4
0.2
0
0.0
400
500
600
Wavelength (nm)
700
Absorbance (a.u.)
h(%)
Jsc
20
1.2
Exciton diffusion length limitation
ITO/Pt(TPBP)(xÅ)/C60(400Å)/BCP(100Å)//Al
-2.4
100 Å
150 Å
200 Å
300 Å
400 Å
-0.5
2
0.0
1.0
Jsc at 0.5 Sun (mA/cm )
2
0.5
-1.0
-1.5
-2.0
-2.2
0.8
-2.0
0.6
-1.8
0.4
JSC
-1.6
Abs.
-2.5
-0.4
-0.2
0.0
0.2
Voltage (V)
0.4
0.6
0.8
0.2
-1.4
100
200
300
PtTPBP Thickness (A)
• PtTPBP is a good hole conductor
– FF is not significantly affected by the thickness
• Optimal thickness = 150 Å
• Measured exciton diffusion length = 57 Å
400
PtTPBP Absorbance at 625 nm
Current density (mA/cm )
1.0
Exiton trapping in PtTPBP films
• Film of PtTPBP emits from a excimer or aggreate
PL Intensity (a.u.)
Absorption
Emission
Crystalline
Pt(tBu)
8TPBP
1.0
1.0
E = 0.34eV
1.0
exc 77K
em 77K
RT
0.8
0.5
0.6
0.5
0.4
0.2
0.0
0.0
0.0
400
600
800
1000
1200
Wavelength (nm)
300
400
500
600
700
800
PtTPBP
doped
thin
film
Wavelength (nm)
PtTPBP thin film
Normalized emission intensity, a.u.
Normalized absorption, a.u.
– OLED = doped film: no excimer present
– Thin film exciton trapped 0.34 eV below molecular exciton,
disordered solid may lead to deeper traps
– Exciton/aggregate lifetimes are short
PL lmax = 754 nm
 = 18 sec
(77K) = 68 sec
LD = 30 nm
M.D. Perez, et. al., Adv. Mat., 2009
Pt(TPBP) OPV performance
PtTPBP
CuPc
C60
6
Pt (TPBP)
CuPc
2
2
J (mA/cm )
4
0
300 mV
-2
-4
-6
-0.2
0.0
0.2
0.4
0.6
0.8
V(volt)
h(%) EDA
JS
Jsc
Voc
FF
CuPc
5.51
0.482
0.60
1.59
1.7
1.1
PtTPBP
4.48
0.685
0.63
1.93
1.4
0.02
h+
e-
(A/cm2)
M.D. Perez, et. al., J. Am. Chem. Soc., 2009
Kinetic Control of VOC
At VOC the photocurrent is cancelled out by the injected current: steady state
kct
D+ + A- Light
D* + A
e-
D+ + A-
krec
hν
D +A
VOC
upper = EDA
limit
Dark
h+
+
-
Light
+
-
Dark
Voc
Current, a.u.
DONOR
15
10
5
0
-5
-10
-15
-0.5
ACCEPTOR
Photocurrent
0.0
0.5
Voltage, V
1.0
1.5
Pt(TPBP) OPV performance
PtTPBP
CuPc
C60
6
Pt (TPBP)
CuPc
300 mV
2
2
J (mA/cm )
4
0
-2
-4
h+
-6
-0.2
0.0
0.2
0.4
0.6
e-
0.8
V(volt)
h(%) EDA
JS
Jsc
Voc
CuPc
5.51
0.482
1.59
1.7
1.1
PtTPBP
4.48
0.685
1.93
1.4
0.02
JS  recombination rate
(A/cm2)
+
+
Pt(TPBP)/C60
CuPc/C60
M.D. Perez, et. al., J. Am. Chem. Soc., 2009
Chemical annealing of porphyrin films
CN
N
H
N
N
N
N
N
N
N
N
N
N
N
N
py
pyCN
dmap
pyz
triazine
Himi
N
Meimi
N
Phimi
UV vis spectra of 150Å ZnTPP film treated with ligands
• Film treated with ligand
vapor only
• Reaction complete in mins.
• Film composition:
ZnTPP:Ligand = 1:1
ZnTPP
coated
substrate
N-base
– Based on NMR spectra of
dissolved thin films
N
1.4
t=0
10 s
20 s
40 s
1 min
2 min
4 min
Absorbance
1.2
1.0
0.8
0.6
0.4
0.16
0.12
N
0.08
0.04
0.00
• Not clear isosbestic
behavior
• Peak sharpening - film
reorganization
540 560 580 600 620 640
0.2
0.0
400
450
500
550
600
650
Wavelength (nm)
C. Trinh, et al., Chem. Mater. (2012)
Thin film structure – XRD measurement
300
250
N
ZnTPP film
200
ZnTPP•pz film
N
150
100
(1,-1,3)
50
Single crystal ZnTPP•pyz
Intensity (a.u.)
Intensity (a.u.)
300
0
5
10
15
2 (  )
20
25
30
N
200
ZnTPP•triazine film
N
N
(202)
100
Single crystal ZnTPP•triazine
0
5
10
15
20
25
30
2 (o)
(1,-1,3)
(202)
C. Trinh, et al., Chem. Mater. (2012)
Grazing Incidence XRD – relative orientation
ZnTPP-pz, 50 nm
(1,-1,3)
• Orientation of the molecules is
parallel to the substrate!
• Intensity spread for the (1,-1,3)
as a function of polar angle 
shows that the crystallites are
+/- 8.2º to the surface
C. Trinh, et al., Chem. Mater. (2012)
Current density (mA/cm2)
Device performance
3
2
D1
D2
Al
1
0
-1
Al
BCP (10nm)
BCP (10nm)
C60(40 nm)
C60(40 nm)
ZnTPP 15nm
ZnTPP-pz
15nm
ITO
ITO
-2
-3
-1.0
-0.5
0.0
0.5
1.0
Voltage (V)
• High leakage, low VOC
• Crystalline  high conductivity
– High conductivity = high FF
D1
D2
AFM images of 100Å films on glass substrates
N
N
N
N
ZnTPP•pyz
ZnTPP
Section analysis
0
nm
-5
0μ
1
2
ZnTPP•triazine
Section analysis
5
Section analysis
20
7
0
nm
-20
0
nm
3
0μ
N
-7
1
2
3
0μ
1
2
3
10 nm
50 nm
15 nm
5 nm
25 nm
7.5 nm
0 nm
0 nm
0 nm
Aggregates are formed on chemical annealing
C. Trinh, et al., Chem. Mater. (2012)
Current density (mA/cm22)
Device performance
3
2
Al
D1
D2
D4
D5
Al
BCP (10nm)
BCP (10nm)
C60(40 nm)
C60(40 nm)
-2
ZnTPP 15nm
ZnTPP-pz
15nm
-3
ITO
ITO
D1
D2
Al
Al
1
0
-1
-1.0
-0.5
0.0
0.0
Voltage
Voltage (V)
(V)
0.5
0.5
1.0
1.0
NPD
• Using NPD to prevent leakage
– Recovers ½ of the VOC
– NPD lmax < 400 nm
• Structure at the D/A interface
affects rate of recombination
BCP (10nm)
BCP (10nm)
C60(40 nm)
C60(40 nm)
NPD (10nm)
ZnTPP-pz
15nm
NPD (10nm)
ITO
ITO
ZnTPP 15nm
D4
D5
Kinetic Control of VOC
At VOC the photocurrent is cancelled out by the injected current: steady state
kct
D+ + A- Light
D* + A
e-
D+ + A-
krec
hν
D +A
VOC
upper = EDA
limit
Dark
h+
+
-
Light
+
-
Dark
Voc
Current, a.u.
DONOR
15
10
5
0
-5
-10
-15
-0.5
ACCEPTOR
Photocurrent
0.0
0.5
Voltage, V
1.0
1.5
Electron transfer rates for Pentacene/C60
Perpendicular
107
104
Parallel
Dashed, solid = CT: pent/C60  pent+/C60Dotted = Recom.: pent+/C60-  pent/C60
• Rates of forward and back electron transfer depend on orientation
• Parallel orientation gives rise to a high recombination rate
Y. P. Yi, V. Coropceanu and J. L. Bredas, J. A. Chem. Soc., 2009, 131, 15777-15783.
CT state energy correlates with Voc
-
h+
e–
C60
C60
C60
C60
CT
+
High ECT
Low ECT
e–
h+
Plot courtesy of Koen Vandewal
Current density (mA/cm2)
Device performance
3
2
D1
D2
D4
D5
Al
Al
BCP (10nm)
BCP (10nm)
C60(40 nm)
C60(40 nm)
-1
ZnTPP 15nm
-2
ZnTPP-pz
15nm
ITO
ITO
D1
(D4)
D2
(D5)
1
0
-3
-1.0
-0.5
0.0
0.5
1.0
Voltage (V)
• High recombination rate 
high dark current  low VOC
• Amorphous interface is
better than the wrong one
C60
C60
C60
C60
C60
C60
Current density (mA/cm2)
Device performance
3
Al
Al
D6
D7
D8
BCP (10nm)
BCP (10nm)
C60(40 nm)
C60(40 nm)
ZnTPP 20nm
ZnTPP-pz
20nm
-1
ITO
ITO
-2
D6
D7
2
1
0
-3
-1.0
Al
-0.5
0.0
0.5
1.0
Voltage (V)
BCP (10nm)
C60(40 nm)
• Using thicker amorphous ZnTPP to
recover most of VOC, improve FF/JSC
ZnTPP 10nm
ZnTPP-pz 10nm
ITO
D8
C. Trinh, et al., Chem. Mater. (2012)
Device performance
Al
Al
1.4
t=0
10 s
20 s
40 s
1 min
2 min
4 min
Absorbance
1.2
1.0
0.8
0.6
0.4
0.16
BCP (10nm)
BCP (10nm)
C60(40 nm)
C60(40 nm)
ZnTPP 20nm
ZnTPP-pz
20nm
ITO
ITO
D6
D7
0.12
0.08
0.04
0.00
540 560 580 600 620 640
0.2
Al
0.0
400
450
500
550
600
650
Wavelength (nm)
BCP (10nm)
C60(40 nm)
• Using thicker amorphous ZnTPP to
recover most of VOC, improve FF/JSC
ZnTPP 10nm
ZnTPP-pz 10nm
ITO
D8
C. Trinh, et al., Chem. Mater. (2012)
Ag, 100 nm
OH O HO
2
As Cast
Pre-C60
Post-C60
Schottky
• DPSQ: a  105 cm-1
•Solvent vapor annealing (SVA): film exposed to a
saturated solvent vapor (CH2Cl2): improves molecular
diffusivity (and formation of nanoXstals)
•Annealing done before and after C60 deposition.
1
DPSQ
0.2 0.4 0.6 0.8
Voltage [V]
ITO, 150 nm
OH O HO
MoO3, 8 nm
N
DPSQ, 16 nm
+
C60, 40 nm
N
PTCBI, 8 nm
1.2 1.4
Achieving the “ideal” morphology
Morphological control through Solvent Vapor Annealing
•
Device
Crystallinity
VOC
JSC
As Cast
Least
Good
Poor
Pre C60
Best
Poor
Poor
Post C60
Middle
Good
Good
Device performance not
monotonically related to crystallinity.
• DPSQ crystallization in post-C60
evident from absorption
spectrum, but not XRD.
• Crystallized DPSQ templates
C60.
•
•
Improved bulk crystallinity LD.
Highly crystalline interfaces kPPr.
J. D. Zimmerman, et al. Nano Letters (2012)
Solvent Vapor Annealing pre- & post-C60 deposition
• SVA in DCM pre-C60:
– 0.12 V in VOC.
– EQE  DPSQ  improved exciton
transport.
– EQE for C60
• SVA post-C60:
– DPSQ EQE  by 80% , JSC  by 25% .
– No loss in VOC  interface is the same
Process
JSC [mA
cm-2]
VOC [V]
FF
[%]
hP
[%]
As Cast
6.1
0.96
74
4.3
Pre-C60
6.0
0.84
71
3.6
Post-C60
7.7
0.97
72
5.5
J. D. Zimmerman, et al. Nano Letters (2012)
Solvent Vapor Annealing pre- & post-C60 deposition
As Cast
• SVA in DCM pre-C60:
– 0.12 V in VOC.
– EQE  DPSQ  improved exciton
transport.
– EQE for C60
• SVA post-C60:
– DPSQ EQE  by 80% , JSC  by 25% .
– No loss in VOC  interface is the same
Pre-C60
Post-C60
Process
JSC [mA
cm-2]
VOC [V]
FF
[%]
hP
[%]
As Cast
6.1
0.96
74
4.3
Pre-C60
6.0
0.84
71
3.6
Post-C60
7.7
0.97
72
5.5
J. D. Zimmerman, et al. Nano Letters (2012)
VOC summary
60
• The structure at the D/A interface influences
Doped C
40
VOC
30
50
Undoped C60
EQE (%)
60
– Sterics can increase spacing
and thus VOC
20
– Ordering can be good or10bad
• Chemical annealing ZnTPP0 orders
400 500to the
600 wrong
700 800
Wavelength (nm)
structure
• Disordered is better than the wrong order
• Can we increase JSC without altering VOC?
– Doping as seen yesterday, broaden absorbance below Eg
– Singlet fission
Singlet fission routes around the Shockley-Queisser limit
2.0
T1T1
5
S1 + S0  T1 + T1
2
1
Global tilt incident power (W/m nm )
S0S1
1.6
4
1.2
3
0.8
2
0.4
1
0.0
500
1000
1500
Wavelength (nm)
2000
0
2500
Photon Energy in eV
Area corrected for exciton
energy drop to bandgap
Photons not
absorbed
S0
• Incomplete absorption and thermolysis to
band gap energy limit h < 31%
Shockley and Queisser, J. Appl. Phys. (1961)
Singlet fission routes around the Shockley-Queisser limit
1 photon in
2 excitons out
1
2
T1
5
1.6
1.6
Area corrected for exciton
energy drop to bandgap
1.2
1.2
4
3
0.8
0.8
2
0.40.4
1
0.00.0
500400
1000
600
8001500
1000
2000
1200
0
14002500
Wavelength
(nm)(nm)
Wavelength
Theoretical efficiency > 45%!
A. Nozik, Appl. Phys. Lett. (2006)
Photon Energy in eV
Global tilt incident power (W/m nm )
2
1
Global tilt incident power (W/m nm )
2.0
x2
Photons not
absorbed
S0
• Incomplete absorption and thermolysis to
band gap energy limit h < 31%
Shockley and Queisser, J. Appl. Phys. (1961)
• Singlet Fission solution: 45%
• You MUST absorb ALL photons
• Efficient Singlet Fission materials are
largely limited to crystalline/polycrystalline
materials
Singlet Fission Materials
• Several materials systems have SF = 200% (triplet yield)
• High efficiencies only observed for crystalline materials
• Tetracene
– M. Pope, J.Chem.Phys (1965), A.J. Taylor, et. al.,
Phys.Rev.Lett.(2009); Damrauer, ibid. (2010); Bardeen,
J.Chem.Phys. (2011)
• Pentacene
– R.H. Friend, et. al., J. Am. Chem. Soc. (2010); A.A. Villaeys et. al.
Chem. Phys. Lett. (1995)
• Diphenyl-benzofuran
– J. Michl, et. al., J. Am. Chem. Soc. (2010)
• Carotenoids
– M.J. Tauber, et. al., J. Am. Chem. Soc. (2010)
Absorption/Emission of Acenes
Tetracene 1000 Å
Amorphous thin film by X-ray and e- diffraction
DPT 1000 Å
Transient Absorption Spectroscopy
Abs  AbsON  AbsOFF
Singlet Fission in DPT
SnS1
S.T. Roberts
et. al. JACS
134
(2012)
C. Burgdorff
et. al. Spectrochim.
Acta.
44A
(1988)6388.
1137.
Singlet Fission in amorphous DPT films
QT = 122%
fast = 1.3 ps
slow = 105 ps
Crystalline Tetracene: SF = 40-80 ps
A.J. Taylor, et. al., Phys.Rev.Lett.(2009); Damrauer, ibid.
Bardeen,
J.Chem.Phys.
(2011)6388.
S.T.(2010);
Roberts
et. al. JACS
134 (2012)
DPT Crystal Structure
3.68 Å Pair
4.00 Å
4.00 Å Pair
3.68 Å
NAQMD: Kinetic Monte Carlo Simulations
• Utilize Surface
Hopping for State
Transitions
• Wavefunctions from
TD-DFT
• Fermi’s Golden
Rule for Transition
Probabilities
• 5000 individual
trajectories
Weiwei Mou & Aiichiro Nakano, Appl. Phys. Lett. (2013)
NAQMD: Preferred Fission Sites
Simulations suggest
3.9% of molecules give
91% of fission events!
NAQMD: Preferred Fission Sites
Experiment
NAQMD-KMC
Simulations suggest
3.9% of molecules give
91% of fission events!
Singlet fission routes around the Shockley-Queisser limit
1.6
2
1
Global tilt incident power (W/m nm )
x2
S1/T1
1.2
T1
0.8
S0
0.4
0.0
400
600
800
1000
1200
1400
Wavelength (nm)
• Incomplete absorption and thermolysis to
band gap energy limit h < 31%
Shockley and Queisser, J. Appl. Phys. (1961)
• You MUST absorb ALL photons
Theoretical efficiency > 45%!
A. Nozik, Appl. Phys. Lett. (2006)
Red Dye to Fill the SF Gap
Spectra in THF solution
Materials studied as
co-deposited films
2.0
1.5
0.8
PtTPBP
S1-T1
DPT Triplet
1.0
0.5
0.0
1.2
DPT S1-T1
0.4
0.0
400
600
800
Wavelength (nm)
Wavelength
(nm)
1000
Emission (arb. units)
Absorbance (arb. units)
DPT
DPT
PtTPBP
PtTPBP phosphorescence
Sensitizing to the red in SF materials
SS = 4.6 ps
ISC = 400 fs
TT = 35 ps
Singlet Fission 1 + 105 ps
S1 + S0  2 T1
< 50% SF comes from prompt
ANODE
Acceptor
SF + Sensitizer
CATHODE
S1,Pt(TPBP)  T1,DPT: 85% Efficient
S1,DPT  T1,DPT: 80% Efficient
Singlet fission: 61%
Need to eliminate delayed SF
to make this structure work.
S. Roberts, et. al., J. Phys. Chem. Lett. (2011)
Combining SF with mid-band absorption
S1
• Problems:
– We did not shut off S11  S12.
– Low doping level for PtTPBP
• Spatially separating the SF
material from all singlet traps will
eliminate transfer (layered rather
than mixed structure)
S1
T1
T1
Donor 2
(PtTPBP)
CATHODE
Acceptor
SF + Sensitizer
T1
SF Donor
(DPT)
CATHODE
Acceptor
SF material
Sensitizer
LUMO
Acceptor
Singlet Fission Conclusions
• Singlet fission can be observed for both crystalline and
amorphous materials
– DPT shows SF in both thin films and NP
• SF takes place at specific dimer sites
– Prompt and diffusive SF
• What is the preferred dimer structure?? Is efficient
“unimolecular” singlet fission possible? Can we create
systems with only prompt SF?
Acknowledgements
Kristen Mutolo, M. Dolores Perez, Cong Trinh, Steve Bradfoth, Sean
Roberts, Eric McAnally
Department of Chemistry
University of Southern California
Stephen Forrest, Jeramy Zimmerma
Departments of Physics and Electrical Engineering
University of Michigan
Michael Toney, Christopher Tassone, Materials Science Department and
Stanford Synchrotron Radiation Lightsource
Funding:
CAMP
Center for Advanced Molecular Photovoltaics
STAN FORD UN IVERSITY
King Abdulla University of
Science and Technology
+
Acceptor
Donor
P0
(1 sun, AM1.5)
8
Light
Dark
4
0
Photocurrent
-4
-8
Jsc
-12
-0.4
0.0
6
5
4
3
2
Voc1
0
-1
-2
-3
-4
-5
-6
2
2
Current Density (mA/cm )
12
Power Density (mW/cm )
Solar Cell Efficiency
0.4
Applied Test Voltage (V)
Pmax J sc  Voc  FF

Cell Efficiency =h p 
Po
Po
Solar Cell Efficiency
+
Acceptor
Donor
P0
(1 sun, AM1.5)
8
-4
-2
4
Voc
0
0
-4
Pmax = VmJm
2
4
-8
Jsc
-12
-0.4
0.0
2
-
-6
Light
Dark
Power Density (mW/cm )
2
Current Density (mA/cm )
12
6
0.4
Applied Test Voltage (V)
Pmax J sc  Voc  FF
hp 

Po
Po
Solar Cell Efficiency
+
Acceptor
Donor
P0
(1 sun, AM1.5)
8
Light
Dark
-4
-2
4
Voc
0
0
-4
2
-8
4
Jsc
-12
-0.4
0.0
2
-
-6
Power Density (mW/cm )
2
Current Density (mA/cm )
12
6
0.4
Applied Test Voltage (V)
Pmax
J mVm
FF 

 Pmax  J scVoc FF
J scVoc J scVoc
Pmax J sc  Voc  FF
hp 

Po
Po