Supplementary Information
Rationalization of Donor-Acceptor Ratio in Bulk Heterojunction Solar
Cells using Lateral Photocurrent Studies
Sabyasachi Mukhopadhyay, K. S. Narayan*1
Jawaharlal Nehru Centre for Advanced Scientific Research, Jakkur,
Bangalore, India – 560064
1. Fabrication and Electrical measurement
Optical absorption spectra of Si-PCPDTBT:PC71BM blend (1:1) film and individual
constituents are measured at 450 – 1000 nm range (Figure S1a). A significant optical
absorption is achieved even at visible range due to the presence of PC71BM absorption band
in green region. Donor polymer has broad emission in 650-950 nm range which fully
quenched when PC71BM is incorporated in solution.
S1.Optical absorption coefficient of PC71BM, Si-PCPDTBT and blend film spun on quartz
1
E-mail: narayan@jncasr.ac.in
substrate. (b) BHJ characterization of PCDTBT:PC70BM with different blend ratio and inset
shows the variation of VOC and JSC with blend ratio. (c) FF, series resistance (RS) and
parallel resistance (RP) variation
We utilized optimum solvent evaporation technique in device fabrication process
which generally differs depending upon chemical composition of donor polymer and used
solvent. The process-optimized PCDTBT:PC71BM typically showed the following
characteristics: PCE  1.5 %, JSC  6 mAcm-2, FF  0.42 and VOC = 0.63 V at AM 1.3
illumination. J-V characteristics of devices with different blend ratio are shown in Figure
S1(b). 1:3 blend ratio furnished an optimal range for interfacial area and largest short circuit
current. A variation in the series and parallel resistances also observed with blend
composition as plotted in Figure S1(c). For all the devices, parallel resistance is 103 times
higher as compared to series resistance. Variation in blend morphology has less effect on the
series resistance and it reveals that a less efficient continuous percolation path for carrier
transport is always present in blend morphology. Blend morphology significantly modify the
carrier recombination kinetics as observed in parallel resistance variation. With increasing
PC71BM fraction in blend, donor-acceptor phase separated domains are formed which reduce
carrier loss and results in a balanced carrier transport in blend film (e  h 10-4cm2/Vs for
1:3 ratio).1-3
S2: Photocurrent response of Si-PCPDTBT:PC71BM BHJ solar cells with different blend
composition.
All the device fabrication process is optimized by varying blend ratio with the best
combinations of all the above mentioned parameters. The device thickness is optimized by
comparing dark and light J-V, which in other way helps to avoid space charge limited current
(SCLC) regime.
2. Morphological Study: AFM Measurement
Tapping mode AFM (JPK Instruments, hexagonal shape silicon-nitride tip  15 nm)
imaging method is utilized to monitor the surface morphology of active layers. Topological
AFM image near vicinity of the Al electrode, depicts the existence of morphological length
scale of < 20 nm. The approximate rms roughness of the films is  6  2.5 nm. Phase contrast
images of Si-PCPDTBT:PC71BM films depict the pure phase segregated donor and acceptor
material on BHJ film surface and their interconnected network. Studies show that the active
polymer-cathode interface plays an important role in dictating the solar cell efficiency4. AFM
studies with different D-A ratios reveal that distribution of donor-acceptor component on the
film surface varies with blend ratio which finally affects the overall solar cell performance.
(Figure S3) Photo-conducting AFM at the same region is performed with Pt/Ir coated Si
probes with +/- 2 bias range keeping ITO electrode as ground. Local current variation is
observed with different tip bias correlating with the aggregation of PC71BM on the film
surface.5-7
S3: (a) Schematic of photo conducting AFM measurement, (b and c) topographical and
phase contrast measurements of blend film ( 70 nm) with 1:3 ratio
Phase contrast AFM images (Nanonics MultiView4000TM) presented in Figure S4
shows spin coated, thermal annealed, polymer:PC71BM phase-topology on PEDOT:PPS
coated ITO substrate. Phase contrast AFM images help to distinguish these individual islands
of different chemical constituent on the film surface. The height of the domains is in between
10-15 nm and not much of a variation is observed in rms roughness ( 2 nm) for varying
PC71BM ratios. A homogeneous phase distribution is observed for films with low PC71BM
concentration ( 50%). As PC71BM fraction increases, the phase distribution becomes
inhomogeneous such that spatial phase separation between two components is clearly visible.
Aggregation of acceptor material at cathode-blend film interface directly improves the solar
cell performance since no additional blocking layer is used at the top electrode. So along with
bulk morphology of the spin coated films, enrichment of acceptor material on film surface
also plays a crucial role in charge extraction. The lateral phase distribution is attributed to the
diffusion of PC71BM moieties within the polymer matrix under thermal annealing process
and substituent crystallization to large lateral dimension.
S4: Phase contrast image of blend films with different donor-acceptor ratio. Domains with
larger phase contrast reflect the distribution of donor chemical composition
3. Spatial decay length measurement
The experimental schematic of spatial photocurrent decay measurement is shown in
Figure S5. Laser light spot is expended using 10  optical expander and a uniform intensity
over light spot is obtained. A 60 objective with 0.95 NA (working distance  250 m with
cover slip correction) is used to focus the light spot on the sample with  2-3 m spot size. In
order to maintain the normal incidence and focusing at Al electrode-polymer interface, 50/50
beam splitter is used in the optical path. Manual translation stages are used for optical
alignment such that large back reflection from the Al coating of sample is obtained. Finally
photocurrent from the device is monitored at different distance from the electrode edge by
preciously translating computer controlled stepper motor stage with 1m resolution.
Photocurrent and transmitted signal is simultaneously measured by standard Lock-in
technique with optical chopper ( 170 – 700 Hz) utilizing LabView software.
S5: Schematic of the lateral decay length measurements
Size and intensity of the illumination spot is controlled by neutral density optical
filters. The sharp electrode edge ( 1-4 m) is accurately determined monitoring transmission
over the line profile and with the 1st derivative of the transmission profile.
S6: Spatial photocurrent decay profile for electron (a) and hole (b) current and
corresponding theoretical fits in Si-PCPDTBT:PC71BM devices. Decay profile is obtained by
varying PC71BM composition in blend ratio from 33% (a), 40% (b), 50% (c) and 66% (d).
The derivative of optical transmission profile provides a Gaussian distribution over
the electrode edge. It appears originally from the finite dimension of the optical spot (2 – 5
m) and scattering at the sharp electrode edge. As the edge modes decays within a 10 m, the
photocurrent persists over few hundred microns (LD  80 – 120 m) for low bandgap
donor:fullerene-71 blend as compared to a few tens of microns (LD  10 – 20 m) for
P3HT:fullerene-60 blend. This affirm the efficient long lived trap assisted carrier transport
connectivity and less carrier loss in these low band gap polymer blends as compared to the
P3HT blend. Transmission contrast through device is taken to remove optically induced
artifacts in the measurements. All the decay profile measurements are carried out for at least
10-15 devices for each blend ratio.
S7: Spatial photocurrent decay profile for electron and hole current and corresponding
theoretical fits in PCDTBT-PC71BM devices. Decay profile is obtained by varying PC71BM
composition in blend ratio from 50% (a), 66% (b), 75% (c) and 80% (d).
Summary of the theoretical fits to decay profiles: Variation in 
Table 1 - 1
PCDTBT:PC70BM blend
PC70BM %
Electron
Hole
a
e (m)

a
h (m)

50
0.0132
0.082
0.22
0.015
0.48
0.29
66
0.0091
0.09
0.23
0.015
0.73
0.23
75
0.0073
0.11
0.25
0.015
0.51
0.27
80
0.0083
0.03
0.21
0.003
0.31
0.31
Si-PCPDTBT:PC70BM blend
PC70BM %
Electron
Hole
a
e (m)

a
h (m)

0
0.0096
0.38
0.35
0.0102
0.13
0.26
33
0.0084
0.5
0.34
0.0072
0.53
0.31
40
0.0075
0.57
0.35
0.0123
0.42
0.29
50
0.0111
0.35
0.36
0.0067
0.19
0.31
66
0.0089
0.42
0.35
0.0076
0.2
0.31
S8: Equivalent circuit representation of asymmetric electrode device configurations (Figure
S6 and 7) for electron (a) and hole (b) transport study respectively.
Modelling and theoretical studies on phase separation of blend films implies that there
exists a distribution in periodicity of phase separated domains, which is directly correlated to
the persistence transport lengths. The interconnected donor/acceptor networks are represented
by a series resistor whereas a parallel resistor R'R represents the recombination during the
carrier transport. The current source, I0 represents the saturation value near to the overlap
region. For electron transport measurements, the hole current is directly collected by the
bottom (ITO) electrode, which can be represented by a fixed electrode resistance RBulk(Figure
S8a). Similarly modelling can be done for hole current also (Figure S8b). The predominant
lateral current flows through a spreading resistance Rsp(x) = Rs g(x), where Rs is the sheet
resistance of PC70BM (e) or polymer network (h). Here ‘x’ is the distance of the light beam
from the Al-electrode edge along the scanning direction. The measured Iph is represented by
Thus, for the normalized Iph shown in Figure S7 and S8, we can fit to an expression of the
form Iph(x) = {1 + a exp [(x/ )]}−1 where a Rs/RR (Rbulk << Rsp , RR) represent the
recombination process during the carrier transport.
4. Power Spectral Density analysis
We have employed power spectral density analysis (PSD) on surface morphology,
optical transmission mapping of blend films to determine the strength of the periodic domain
variations as a function of spatial frequencies8,9. It provides the information about the relevant
length scales which used to quantify the changes with different composition ratio, and phase
separation upon thermal annealing. In our analysis, first all the contrast images are
transformed to the 8 bit color contrast image (0 to 256) with 512  512 pixels image
resolution. A taper window function is applied to reduce the edge effects and to minimize the
spectral leakage on all of the images before calculating PSD. Then commercially available 2dimentaion (2D) Fourier transform program is applied to all the images. Finally few of these
obtained PSD data are cross-verified with our MatLab simulation. In the present analysis the
computation for the 2D average discrete PSD function can be express as
1 N N
2
PSD2 D k x , k x   2  I mn e 2iLkm xkn y  L  
L m1 n1

2
Here Imn represents the profile information contrast (height, current or transmitted
light intensity) for (m, n)th pixel with scan surface area (L2) at sampling dimension L (L/N)
and ki represents the spatial frequency along i direction. Finally average angular power
spectral density is calculated which is represented as the radial frequencies k (= k/2) to
generate radial one-dimensional PSD as
PSD2 D k  
1
2

2
0
PSDk , d
AFM and T-NSOM scans are carried out for different films at different region with
15m  15m, 10m  10m, 5m  5m area at 512  512 data points which provides  10
nm (5m/512) sampling rate. The minimum spatial frequency corresponds to kmin = 1/15m =
0.066 m-1 and spatial frequency resolution is limited by the scan sampling size kmax =
1/10nm = 100 m-1. These form the lower and upper bandwidths of limitation of our PSD
plots. PSD spectra obtained upon Fourier analysis of images recorded upon illuminating
different areas away from the Al electrode resulted in a similar PSD distribution indicating
that the phase separation is quite consistent over a large device area.
5. T-NSOM images of Si-PCPDTBT:PC71BM film
Transmission near field scanning mapping of Si-PCPDTBT:PC71BM blends are
performed with varying PC71BM concentration and are shown in Figure S9. In the T-NSOM
images, brighter (red) regions corresponds to the PC71BM network with higher photon count
rates as compared to the absorbing donor domains (blue) regions.8 Films with PC71BM
concentration
up
to
20%
demonstrated
a
very
homogeneous
morphology
in
PCDTBT:PC71BM bulk films. At 40% PC71BM concentration, phase separated PC71BM
crystalline regions with smaller length scales (20-40 nm), mixed amorphous regions along
with large polymer domains ( 200 nm) are clearly resolved in T-NSOM images. At higher
concentration of PC71BM, both donor and acceptor domain size gets larges ( 200 nm) which
prevails an unbalanced carrier transport in the bulk film.
S9: T-NSOM images of Si-PCPDTBT:PC70BM films with different acceptor ratios
Power spectral density analysis of 3:2 Si-PCPDTBT:PC71BM film consists of four dominated
periodicities. The maximum corresponds to the 29.7 m-1   35 nm. Figure S10 shows the
spatial frequency distribution obtained from film with blend ratio of 3:2 (D:A).
S10: Power spectral density analysis of Si-PDPDTBT:PC71BM film at optimum ratio.
References:
1
M. C. Scharber, M. Koppe, J. Gao, F. Cordella, M. A. Loi, P. Denk, M. Morana, H.-J.
Egelhaaf, K. Forberich, G. Dennler, R. Gaudiana, D. Waller, Z. Zhu, X. Shi, and C. J.
Brabec, Adv. Mater. 22, 367 (2010).
2
M. Morana, H. Azimi, G. Dennler, H.-J. Egelhaaf, M. Scharber, K. Forberich, J.
Hauch, R. Gaudiana, D. Waller, Z. Zhu, K. Hingerl, S. S. van Bavel, J. Loos, and C. J.
Brabec, Adv. Funct. Mater. 20, 1180 (2010).
3
A. Maurano, R. Hamilton, C. G. Shuttle, A. M. Ballantyne, J. Nelson, B. O’Regan,
W. Zhang, I. McCulloch, H. Azimi, M. Morana, C. J. Brabec, and J. R. Durrant, Adv.
Mater. 22, 4987 (2010).
4
D. Gupta, S. Mukhopadhyay, and K. S. Narayan, Solar Energy Mater. Solar Cells 94,
1309 (2010).
5
C. Groves, O. G. Reid, and D. S. Ginger, Acc. Chem. Res. 43, 612 (2010).
6
L. S. C. Pingree, O. G. Reid, and D. S. Ginger, Adv. Mater. 21, 19 (2009).
7
L. S. C. Pingree, O. G. Reid, and D. S. Ginger, Nano Lett. 9, 2946 (2009).
8
S. Mukhopadhyay, S. Ramachandra, and K. S. Narayan, J. Phys. Chem. C 115, 17184
(2011).
9
I. Horcas, R. Fernandez, J. M. Gomez-Rodriguez, J. Colchero, J. Gomez-Herrero, and
A. M. Baro, Rev. of Scient. Instrum. 78, 013705 (2007).