Supplemental Materials
Photothermal Deflection Spectroscopy
Photothermal Deflection Spectroscopy (PDS) is a sensitive spectroscopy technique that measures
optical absorption from the change in refractive index due to monochromatic light heating of a
sample surface. The schematic structure of instrument is as shown in Fig S1. The sample to be
measured is immersed in a quartz container with CCl4, a liquid with a highly temperaturedependant refractive index. As the sample absorbs incident radiation from a monochromatic beam,
absorption cause the sample to heat, and the flux of heat into the adjoining CCl4 solution
generates a temperature gradient in the space close to the sample surface, which is accompanied
by a gradient of the index of refraction. A monochromatic beam, chopped at 10Hz, produces a
time-dependent refractive index variation profile in the direction normal to the sample surface,
which deflects a focused He-Ne laser beam that passed tangentially to the surface of the sample.
The beam deflection at a given wavelength is monitored by position sensitive detector and
normalized by the incident light intensity, which is monitored by a pyroelectric detector. Our PDS
system allows contactless measurements for over a range of 4 orders of magnitude in absorption
for an energy range of 0.5 eV to 3 eV for 0.01 - 100µm thick films at room temperature. The PDS
measurement technique allows the real band gap of the semiconductor to be measured as well as
Urbach energy and the density of midgap states through analysis of the band gap and the band tail
absorption.
FIG. S1. (Color online) Schematic structure of photothermal deflection spectrometer
Impact of Transparent Conductor (ITO vs. FTO)
We also fabricated TiO2/PbS (1.1eV) device on fluorine-doped tin oxide (FTO) /glass substrate
and compared it with indium tin oxide (ITO) device due to concern that the sintering process would
impact the ITO. The results are shown below in Table SI and Figure S2. While the control device (ITO
used as cathode) is lower than our baseline performance, the comparison between devices is valid. No
obvious difference was observed between them.
TABLE SI. Performance parameters of the TiO2/PbS (1.1eV) devices fabricated on the ITO and FTOcoated glass substrates. The values listed here are the averages for all devices on one panel.
Cathode
ITO
FTO
Jsc
(mA/cm2)
14.9
14.6
Voc
(V)
0.362
0.362
FF
(%)
27.7
28.0
η
(%)
1.49
1.48
FIG. S2. (Color online) J-V characteristics of TiO2/PbS (1.1eV) devices with ITO and FTO cathodes
under 100mW/cm2 of AM 1.5 illumination.
Air exposure experiments with and without pump process
We evaluated whether exposure to vacuum in the antechamber during box transfer impacted our results.
When the devices were transferred back into the glove box, the cycle of evacuate (about -0.9- -1.0 bar)refill-evacuate was process for 3 times. The whole process was finished in less than 1.5 min. The results
are shown in Table SII and Figure S3. The similar trend with air exposure time on two devices was
observed though the device with flush only process has a slightly smaller performance improvement after
5-min air exposure and degraded a slightly more quickly than the device with pump process. This effect
is due to residual trapped air in the porous film.
FIG. S3. (Color online) J-V characteristics for TiO2/PbS devices under 100mW/cm2 AM 1.5 illumination
up to one hour of air exposure with (left) and without (right) pump process.
TABLE SII. Performance parameters of the corresponding TiO2/PbS devices over the air exposure time
with and without pump process.
Pump
Flush
Time
(minute)
0
5
20
60
Jsc
(mA/cm2)
20.0
19.7
17.3
10.9
Voc
(V)
0.326
0.451
0.463
0.482
FF
(%)
28.6
36.7
39.4
29.4
η
(%)
1.86
3.26
3.16
1.54
0
5
20
60
20.7
19.7
17.2
8.70
0.337
0.457
0.472
0.485
27.3
32.8
33.6
26.6
1.90
2.95
2.73
1.12
EDT Treatment
After 24-hour of air exposure in normal fluorescent light, the device was soaked again in the
same EDT solution used in the ligand exchange for 5 minutes and then J-V curves labeled as “24hour air + 5-min EDT” in FIG.2 and FIG.S4 were measured in nitrogen-filled glove box. After
that, the device was soaked in the same EDT solution for 25 more minutes (total 30 minutes),
followed by the measurements of J-V curves labeled as “24-hour air + (5+25)-min EDT”. A
similar EDT soak operation was applied in PDS experiments. For these experiments, all
measurements were performed in series on the same device, with air exposure followed by
subsequent soaks in EDT. The results are reproducible on all devices studied in our experiments
(>20 devices). As can be observed, the exposure to air caused the dark currents to decrease. The
currents are mostly recovered upon exposure to EDT.
FIG.S4. (Color online) Dark J-V curves for the same TiO2/PbS (1.1eV) device as that shown in
the FIG.2 in the main text before and after 24-hour air exposure as well as followed 5-min and
30-min of EDT soak.
TiO2 Treatment
To check if the filling of oxygen vacancies in the TiO2 layer during the air exposure caused the
observed behavior in the work, we intentionally filled some oxygen vacancies by sintering the
TiO2 film in oxygen and by treating with an oxygen plasma in the process of TiO2 films
preparation. For the sintering experiment in pure oxygen, the applied sintering temperature is
450 °C and the duration time is 30min. Compared to the control device with TiO2 layer sintered
in air at the same temperature for same time, no increase in Voc was observed.
Oxygen plasma treatment was applied on TiO2 films annealed in air just before dip-coating PbS
films. The oxygen plasma is composed of 30% oxygen and 70% nitrogen. Different plasma
treatment time (1min, 5min, 10min, and 60min) was applied in the experiments; however, none of
them caused similar results as those observed in the air exposure experiments.
Absorption Spectra
The absorption data was measured as a function of air exposure time. This data shows limited
changes in the absorption over the first 2 hours of exposure under normal room lighting; however,
significant shifts are observed over 1 day (1440 minutes). These shifts are due to an effective
reduction in the quantum dot size due to the formation of an insulating shell on the PbS surface.
This reduction in effective quantum dot size shifts the exciton peak and reduces the overall
absorption.
FIG. S5. (Color online) Absorption change of EDT exchanged PbS film over air exposure time. The
spectra labeled as 0, 5min, 20min, 60min and 120min overlap.
Stability of ZnO/PbS Devices
We also made some solar cell devices from ZnO nanoparticles and two different sizes of PbS quantum
dots (1.1eV and 1.35eV) with a Jsc up to15 mA/cm2. A slightly slower, though similar, decrease in the
ZnO/PbS devices is observed in comparison to TiO2/PbS. These results are different than the stability
results observed by others and suggest that the stability may depend on the initial conditions of the PbS
solutions. If the insulating shell on the QD is formed during synthesis, it can protect the PbS from further
oxidation, but lead to the lower initial currents observed by NREL. If the insulating shell is formed after
the ligand exchange, as is the case for all of our devices, the shell will be impacted by the presence of the
ligands and therefore not form a fully protective barrier. All of our PbS quantum dot solutions are
prepared under inert conditions and are never exposed to oxygen.
FIG. S6. (Color on line) J-V characteristics of a ZnO/PbS (1.1eV) device (a) and a ZnO/PbS (1.35eV)
device (b) before and after 24-hour air exposure under 100mW/cm2 of AM 1.5 illumination. Insets: The
specific parameters of two devices.
Device Performance Tables
Here we list device performance data in Tables that correspond to the Fig. 1, 2 and 4 in the text.
TABLE SIII. Performance parameters for a TiO2/PbS (1.1eV) device under 90mW/cm2 AM 1.5
illumination in one hour of air exposure. (Corresponding to Fig. 1)
Time
(minute)
0
5
20
60
Jsc
(mA/cm2)
15.5
14.1
13.1
11.3
Voc
(V)
0.401
0.508
0.519
0.525
FF
(%)
30.3
38.8
39.9
40.3
η
(%)
2.09
3.09
3.01
2.66
TABLE SIV. Performance change of a TiO2/PbS (1.1eV) device over long-term air exposure and
different EDT soak time under 100mW/cm2 AM 1.5 illumination. (Corresponding to Fig. 2)
No air
24-hour air
24-hour air +5-min EDT
24-hour air+(5+25)-min EDT
Jsc
(mA/cm2)
20.4
0.603
10.6
15.3
Voc
(V)
0.343
0.621
FF
η
(%)
(%)
28.5 1.99
23.5 0.0880
0.540
0.468
36.9
28.9
2.11
2.07
TABLE SV. Performance parameters of solar cells composed of two different sizes of PbS QDs before
and after 5-min of air exposure under 100mW/cm2 AM 1.5 illumination. (Corresponding to Fig. 4)
Parameters
Jsc
(mA/cm2)
Voc (V)
FF (%)
η (%)
Device (1.1eV)
No air 5-min air
20.7
18.6
Device (1.7eV)
No air 5-min air
9.04
8.03
0.452
34.1
3.19
0.497
32.1
1.44
0.517
41.8
4.02
0.655
35.0
1.84