heterojunction pbs nanocrystal solar cells with oxide
TRANSCRIPT
1
Heterojunction PbS Nanocrystal Solar Cells
with Oxide Charge-Transport Layers
Byung-Ryool Hyun*, Joshua J. Choi+, Kyle L. Seyler*, Tobias Hanrath
+, and Frank W. Wise*
School of Applied and Engineering Physics*, School of Chemical and Biomolecular Engineering+,
Cornell Univeristy, Ithaca, NY 14853
PbS NC Synthesis and Purification : PbS NCs were synthesized based on a variation of a previous
literature report1. Briefly, 20 mL of 1-octadecene (ODE) was added to a three-neck flask, which was
vacuumed and purged with nitrogen gas 3 times to allow it to degas. The ODE was then heated to 110
°C, followed by two vacuum/purge cycles for 5 and 2 minutes to remove water and allow it to degas
further. Afterwards, the ODE was allowed to cool to room temperature in a nitrogen environment. 420
µL of bis(trimethylsilyl)sulfide (TMS) was added to the 20 mL of ODE and allowed to mix thoroughly.
In another three-neck flask, PbO, oleic acid, and ODE were mixed, degassed, and purged at 110 °C as
before. Once the solution turned clear, indicating the formation of lead oleate, the flask temperature was
lowered to a certain temperature. The mixture of TMS and ODE was then extracted and rapidly injected
into the lead oleate solution using a syringe. After injection, the temperature dropped and the solution
turned dark brown, indicating the formation of PbS NCs. The nanocrystal solution was left on the
heating mantle for 3 minutes and afterwards, was cooled rapidly to room temperature with an ice bath.
The nanocrystals were collected into centrifuge tubes, precipitated by adding twice the amount acetone
as NC solution, and centrifuged for 5 minutes under ambient conditions. The supernatant was discarded
and the NCs were redispersed in hexane. Acetone was once again added until the entire solution turned
2
cloudy, and then the NCs were centrifuged, rinsed with acetone, and dried. The NCs were washed this
way one more time for a total of 3 precipitations, and finally, they were redispersed at 40 mg/mL in
chlorobenzene for solar cell fabrication and filtered through a 0.2 µm PVDF syringe filter to remove any
dust, particulates, and NC aggregates.
PbS NC-sensitized mesoporous SiO2 films: 4 g of colloidal SiO2 nanoparticles in water (Nalco 2327,
diameter = 20 nm, 40 wt. % aqueous dispersion) was diluted with 3.2 mL of water and then, 0.64 g of
polyethylene glycol (MW 20000) was added into the diluted dispersion of SiO2 nanoparticles. The
mixture was stirred for at least 2 hrs. Thin films of SiO2 nanoparticles were prepared on the glass
substrate by using a spin coating technique (1000 rpm, 30 sec). Films were subjected to 450 °C heat
treatment for 2 h. The temperature was then lowered to room temperature. Repeated steps of spin
coating and temperature treatment yielded the transparent mesoporous SiO2 film of ~10 µm-thicknesses.
At the last thermal treatment, films were lowered to 120 °C. The hot films were placed directly in a
0.20M 3-mercaptoacetic acid (MPA) (or 3-mercaptopropyl-trimethoxysilane, MPS) in toluene solution
for 12 hrs. Films were washed three times with toluene and placed directly in PbS NCs in toluene for 24
hrs to sensitize PbS NCs. The PbS NC-SiO2 films were washed three times with a toluene to remove the
unbound PbS NCs and then, were dipped in a 1M 2-mercaptoethanol (MEtOH) in toluene to remove the
oleic acid at the surface of PbS NCs. NiO nanoparticles solution was prepared by dispersing 1 g of NiO
nanopoweder (Inframat Advanced Materials, diameter = 20 nm) in 10 mL of water. After 0.5 hr
sonication, the NiO dispersion was left overnight to settle down the big aggregations. The light brown
supernatant was carefully pipetted. To fill NiO nanoparticles out inside the PbS NC-SiO2 mesoporous
film, a few 100 µL of NiO nanoparticles in water was dropped into the PbS NC-SiO film and water was
removed by an evaporation vacuum pump. It was repeated two or three time to fill the empty space
inside of mesoporous film. The NiO nanoparticle covered film was sealed by putting another cleaned
glass slide on top and sealing the sides with UV-curable epoxy. Encapsulated samples were stable
against oxidation for several months.
3
Absorption and Emission Spectrum Measurements: Absorption spectra were measured on a
Shimadzu UV-3101PC spectrophotometer at room temperature. Emission spectra were recorded at room
temperature with an infrared fluorometer equipped with a 200-mm focal length monochromator, a single
mode fiber coupled laser source (S1FC635PM, 635 nm, Thorlabs, Inc) as the excitation source, and an
Si and InGaAs photodiode (New Focus Femtowatt model 2151 and 2153).
Fluorescence lifetime measurements: The samples were excited at a repetition rate of 1 kHz by
femtosecond pulses of a Ti:sapphire laser system with an optical parametric amplifier (OPA) (at 680
nm) or regenerative amplifier (at 800 nm). The sample was exposed to intensity levels below 10
mW/cm2 so that the excitation level was always well below one electron-hole pair per dot. Fluorescence
below 1000 nm was monitored with a Si avalanche photodiode (APD) single photon counting module
4
(PerkinElmer, SPCM-AQRH-44-FC). The time–resolved fluorescence measurements were performed in
the time-correlated single photon counting (TCSPC) mode with a TCSPC board (PicoQuant, TimeHarp
200) and a digital delay generator (Stanford Research Systems, DG645) under right-angle sample
geometry. The fluorescence decay curves were analyzed by means of iterative re-convolution using the
FluoFit. Fluorescence lifetime measurements between 1000 and 1600 nm were performed with an
InGaAs photomultiplier tube module (H10330-75, Hamamatsu, Inc.). The output was fed into a digital
oscilloscope (Tektronix TDS 520A) and averaged, which provides adequate temporal resolution of
decay times in the microsecond range.
Figure S2. Absorption and emission spectra of PbS NCs with diameters of 4.0, 4.9 and 6.0 nm in TCE
Figure S3. Typical transient fluorescence traces of PbS NCs in TCE and SiO2-PbS NC-NiO composites.
The time constants of decay for PbS NCs with D = 4.0, 4.9, and 6.0 nm in TCE were 2.2, 1.6, and 1.4
5
µs, respectively. The 1/ decay times for SiO2-MPA-PbS NC composites were 0.082 (green line), 0.42
(red line), and 0.16 (blue line) µs, respectively.
Solution processed solar cells:
a) NiO layer : We followed the procedure of ref [2]. Briefly, the nickel acetate tetrahydrate (Aldrich)
was dissolved in methanol to a concentration of 0.4 M. An equimolar quantity of diethanolamine
(Aldrich) was added under stirring. ITO substrate was cleaned by successive ultrasonic treatment in
detergent, purified water, acetone and propanol before drying under nitrogen. Thin films were
prepared by spin-coating the sol–gel precursor onto ITO substrate under ambient conditions at a spin
speed of 3000 rpm for 30 s. Samples were then placed immediately into a tube furnace and annealed
under air at 425 °C for 10 min, forming a dense, polycrystalline material. Film thicknesses were
measured by a stylus profiler to be 30–40 nm thick.
b) ZnO layers : ZnO nanoparticles were synthesised in ambient air. 1.76 g of zinc acetate dihydrate
was dissolved in 150 mL of ethanol with vigorous stirring and heated to 60 oC for 1 hour. Parafilm
was used to block solvent evaporation during the heating and the entire reaction time. In a separate
container, 6.4 mL of a tetramethyl ammonium hydroxide (TMAH) solution (28% in MeOH) was
added to 50 mL of ethanol. Over 10 minutes, the TMAH solution was slowly added to the zinc
acetate solution at regular intervals. During this time the temperature of the zinc acetate solution was
maintained at 60 oC with continuous stirring. Subsequently, the solution was heated at 60 oC for 30
minutes after which the heater and stirrer were removed. After cooling to room temperature, the
solution was kept refrigerated at ~ 5 oC. This mother-liquor solution was stable for over ~5 months.
ZnO nanoparticles were collected from the refrigerated solution immediately before device
fabrication. To prepare devices, 5 mL of the solution was typically mixed with 20 mL of hexane to
precipitate the ZnO particles. After centrifugation, the supernatant was removed and the white
precipitate was dissolved in isopropyl alcohol.
6
c) EDT-treated PbS NC Layer Deposition: PbS NCs at 40 mg/mL in chlorobenzene were deposited by
spin-coating from a Laurell WS-400A-6NPP-LITE. All spin steps were carried out at 1000 rpm for
30 seconds. First, 250 µL of NC solution was dispensed onto the substrate using an automatic
pipette and then the substrate was spun. Second, 1 mL of 0.1 M ethanedithiol (EDT) in acetonitrile
was swiftly dispensed on the substrate, left for 30 seconds, and then spun. The short EDT ligands
replace the longer oleic acid ligands, improving conductivity through improved inter-NC coupling.
Third, the substrate was rinsed with pure acetonitrile, spun, rinsed with pure chlorobenzene, and
spun to remove residual ligand molecules and NCs. This three-step spin process of deposition,
ligand exchange, and rinsing was one cycle. Three cycles were performed to produce ~ 100 nm thick
NC films.
d) Fabrication solar cells: The samples were loaded into a metal evaporator and the evaporation
chamber was pumped down to ultrahigh vacuum (~10-6 Torr). Au, Ag or Al was thermally
evaporated at a rate of 0.5 Å/second for the first 100 Å and at 2~3 Å/second for the rest of
deposition. Typically total of 400 ~ 600 Å of metal electrode was deposited.
e) Device characterization: Device testing was performed with a source measurement unit (Keithley
236). Samples were irradiated with 100mW/cm2 AM1.5 illumination from a Solar Light 16S-002
solar simulator. Light output power was calibrated with a Newport 818P-010-12 thermopile high
power detector. Small spectral mismatch was not taken into account in these measurements.
ITO Substrate Cleaning: Pre-patterned ITO coated glass substrates (1" × 1", Kintec) were cleaned by
scrubbing for over 30 seconds with a swab soaked a solution of 1% Alconox in distilled H2O. The
following steps were then taken, with the substrate kept wet at all times. Rinse with distilled H2O,
sonicate in 1% Alconox solution for 10 min, rinse with distilled H2O, rinse with acetone, sonicate in
acetone for 10 min, rinse with acetone, rinse with isopropanol (IPA), sonicate in IPA for 10 min, and
rinse with IPA. Finally, the IPA-wetted substrates were dried with a blast of nitrogen gas.
7
Air annealing and photodoping
Room-temperature sputtered ZnO and NiO layers: ZnO was deposited at a rate of 1.5 Å/s by
reactively sputtering at a base pressure of 10-6 Torr with a 2″ diameter Zn target in a 4:1 Ar:O2 pressure
ratio at 5 mTorr and 65W of DC power. NiO was deposited at a rate of 1.7 Å/s by reactively sputtering
at a base pressure of 10-6 Torr with a 2″ diameter Ni target in a 99:1 Ar:O2 pressure ratio at 10 mTorr
and 57W of DC power. Substrates sat on a rotating stage with a mask over the edges to keep the ITO
bare for good electrical contact.
X-ray diffraction (XRD):
8
Transmittance spectrum:
9
AFM image of DC-sputtered NiO film: RMS = 3 nm
AFM image of DC-sputtered ZnO film: RMS = 3 nm
10
Material Characterization: Surface morphology measurements were obtained by using a Veeco
Dimension 3100 AFM in tapping mode. Transmittance and absorbance spectra were obtained using a
Shimadzu UV-Vis-NIR Spectrometer. Although film thicknesses were not measured for each device,
care was taken to calibrate tooling factors. Film thicknesses were measured using a Tencor Alpha Step
500 profiler.
Performance data for PbS NCs-based solar cells:
Series resistance and shunt resistance were taken as the inverse slope at 0.9-1V and as the inverse slope
of dark JV curve near 0V before it breaks down, respectively. The values in the table were obtained
from averaging values from several devices and at least 2 measurements have been performed for each
device.
1) Sol-gel processed devices
Diameter
[nm]
VOC
[V]
JSC
[mA/cm2]
FF Efficiency
[%]
Series resistance
[Ω⋅cm2]
Shunt resistance
[Ω⋅cm2]
2.75
0.65
± 0.01
3.82
± 0.15
0.38
± 0.004
0.94
±0.03
16.9
± 0.6
6925000
± 14035000
3
0.58
± 0.01
8.15
± 1.42
0.44
± 0.01
2.08
±0.36
8.5
± 0.1
3688000
± 1053000
3.5
0.49
± 0.01
9.53
± 0.33
0.47
± 0.02
2.21
±0.16
9.80
± 0.02
150000
± 8000
4.4
0.36
± 0.01
7.67
± 0.50
0.43
± 0.03
1.19
±0.12
10.2
± 0.01
57000
± 3000
5.7
0.32
± 0.01
5.38
± 0.28
0.43
± 0.023
0.73
±0.07
12.95
± 0.03
6400
± 100
6.4
0.19
± 0.01
2.60
± 0.26
0.32
± 0.01
0.16
±0.03
13.17
± 0.05
400
± 100
11
2) The sputtered devices
The thickness of ZnO = 10 nm
NiO
thickness
[nm]
VOC
[V]
JSC
[mA/cm2]
FF Efficiency
[%]
Series resistance
[Ω⋅cm2]
Shunt resistance
[Ω⋅cm2]
0
0.63
± 0.01
2.1
± 0.2
0.28
± 0.01
0.36
± 0.03
33.46
± 3.03
113334
± 1520
5
0.61
± 0.01
10.7
± 0.1
0.43
± 0.01
2.83
± 0.02
8.51
± 0.01
870
± 4
10
0.625
± 0.01
13.9
± 1.3
0.29
± 0.01
2.53
± 0.27
4.83
± 0.30
1514
± 15
20
0.63
± 0.01
6.7
± 0.7
0.27
± 0.01
1.13
± 0.06
4.31
± 0.07
15761
± 268
30
0.62
± 0.02
5.1
± 0.1
0.26
± 0.01
0.82
± 0.07
8.60
± 0.10
595
± 1
40
0.58
± 0.01
1.7
± 0.2
0.30
± 0.01
0.31
± 0.03
7.90
± 0.10
870
± 7
The thickness of NiO = 10 nm
ZnO
thickness
[nm]
VOC
[V]
JSC
[mA/cm2]
FF Efficiency
[%]
Series resistance
[Ω/cm2]
Shunt resistance
[Ω/cm2]
5
0.35
± 0.04
3.9
± 1.1
0.28
± 0.01
0.25
± 0.04
4.83
± 0.04
469
± 5
10
0.63
± 0.01
13.9
± 1.3
0.29
± 0.01
2.53
± 0.27
4.83
± 0.30
1514
± 15
20 0.65 11.2 0.33 2.42 4.59 16641
12
± 0.01 ± 0.4 ± 0.01 ± 0.10 ± 0.10 ± 89
30
0.66
± 0.01
11.1
± 0.3
0.33
± 0.02
2.50
± 0.20
4.46
± 0.10
97359
± 4729
40
0.67
± 0.01
11.6
± 0.7
0.33
± 0.02
2.53
± 0.33
4.51
± 0.10
64334
± 1479
Optical model calculation : Scattering matrix formalism is based on (1) linearity of equations for
electric field propagation and (2) continuity of tangential component of the electric field at every
interface. A generalized 2 by 2 matrix is formulated by accounting for phase accumulation during
propagation and phase change upon reflection at various interfaces. Assuming that the structure to be
modeled is multi-layered thin films with perfectly flat interfaces and the plane wave is incident in the
direction perpendicular to the interfaces, reflection and transmission that occur at each interface is
described by an interface matrix:
1
1 1
where tjk and rjk are the Fresnel complex transmission and reflection coefficients at the interface jk. In
addition to the changes at the interfaces, the propagating wave accumulates phase while traveling
through each layer. This is described by the phase matrix:
00
where is the refractive index dependent wavevector and dj is the thickness of the layer j. Therefore,
is the amount of phase change due to the propagation through layer j. The scattering matrix, S, now
can be obtained by multiplying all the matrices required to account for every interface and layer in the
structure to be modeled. Using the scattering matrix, transmission and reflection coefficients and optical
field throughout the entire structure can be calculated. Detailed derivations and equations can be found
in Ref. [3]. For the device structure model, we used the stack of Glass (1.1 mm) / ITO (80 nm) / ZnO
13
(variable thickness) / PbS_EDT (100 nm) / NiO (variable thickness) / Au (60 nm). The experimentally
measured complex refractive indices for each film were used as input for the calculation.
1. Hines, M. A.; Scholes, G. D. Colloidal PbS Nanocrystals with Size-Tunable Near-Infrared Emission: Observation of Post-Synthesis Self-Narrowing of the Particle Size Distribution. Adv. Mater.
2003, 15, 1844-1849. 2. Mashford, B. S.; Nguyen, T. L.; Wilson, G. J.; Mulvaney, P. All-Inorganic Quantum-Dot Light-Emitting Devices Formed via Low-Cost, Wet-Chemical Processing. J. Mater. Chem. 2010, 20, 167-172. 3. Pettersson, L. A. A.; Roman, L. S.; Inganas, O. Modeling Photocurrent Action Spectra of Photovoltaic Devices Based on Organic Thin Films. J. Appl. Phys. 1999, 86, 487-496.