Advanced Architecture for Colloidal PbS Quantum Dot Solar Cells

Sep 20, 2016 - Advanced architectures are required to further improve the performance of colloidal PbS heterojunction quantum dot solar cells. Here, w...
0 downloads 8 Views 836KB Size
Download PDF
Advanced Architecture for Colloidal PbS Quantum Dot Solar Cells Exploiting a CdSe Quantum Dot Buffer Layer Tianshuo Zhao,† Earl D. Goodwin,‡ Jiacen Guo,† Han Wang,§ Benjamin T. Diroll,‡ Christopher B. Murray,†,‡ and Cherie R. Kagan*,†,‡,§ †

Department of Materials Science and Engineering, ‡Department of Chemistry, and §Department of Electrical and Systems Engineering, University of Pennsylvania, Philadelphia, Pennsylvania 19104, United States S Supporting Information *

ABSTRACT: Advanced architectures are required to further improve the performance of colloidal PbS heterojunction quantum dot solar cells. Here, we introduce a CdI2treated CdSe quantum dot buffer layer at the junction between ZnO nanoparticles and PbS quantum dots in the solar cells. We exploit the surface- and size-tunable electronic properties of the CdSe quantum dots to optimize its carrier concentration and energy band alignment in the heterojunction. We combine optical, electrical, and analytical measurements to show that the CdSe quantum dot buffer layer suppresses interface recombination and contributes additional photogenerated carriers, increasing the open-circuit voltage and short-circuit current of PbS quantum dot solar cells, leading to a 25% increase in solar power conversion efficiency. KEYWORDS: PbS, CdSe, quantum dot, solar cell, buffer layer, interface electrophoretic deposition,24 passivating the junction with doped polymers25 or modifying the interface with selfassembled monolayers.21 Akin to copper indium gallium selenide (CIGS) thin film solar cells, where thin film CdS is often introduced at the CIGS/oxide junction,26 here, we fabricate ZnO nanoparticle (NP)/PbS QD solar cells with a buffer layer of CdSe QDs at the heterojunction interface. In contrast to the wide band gap and high-resistivity CdS buffer material used in the CIGS system, we n-dope the CdSe QD layer through treatment with the metal salt CdI2, as confirmed by current−voltage and capacitance−voltage measurements. We optimize the band alignment by exploiting the size-dependent electronic structure of CdSe QDs and study the resulting carrier transport across the junction under different illumination conditions. Using external quantum efficiency (EQE) and time-resolved microwave conductivity (TRMC) measurements, we show that the CdSe QD buffer layer reduces interface recombination and provides additional photogenerated carriers, resulting in significantly higher open-circuit voltage (VOC) and short-circuit current density (JSC) and therefore a 25% enhancement in PCE

C

olloidal semiconductor quantum dots (QDs) are an exciting materials class to serve as building blocks for next-generation thin film electronic and optoelectronic devices.1,2 They stand out from other candidates for their sizetunable optical and electronic properties3 and their lowtemperature, solution-based device fabrication, and compatibility with flexible substrates. The large surface-to-volume ratio of QDs also provides opportunities to manipulate material properties. Modifying the surface chemistry of the QDs allows the energy levels4,5 and the carrier concentration,6,7 mobility,8−11 and lifetime12−14 in QD thin films to be engineered. PbS QDs have tunable band gap energies across the nearinfrared region of the electromagnetic spectrum, which makes them appealing for solar cell applications.15 Over the past decade, progress made in improving the mobility-lifetime product of minority carriers in the QD thin films,12,16 passivating the QD surface,17−19 and designing the device structure20 has boosted the power conversion efficiency (PCE) of PbS QD solar cells to as high as 10%21 and has improved their air stability.20,22 The device architecture of the most widely studied PbS QD solar cells consists of a heterojunction with a wide band gap oxide, typically ZnO or TiO2. The junction interface is tailored to reduce carrier recombination and engineer energy levels, for example, by inserting an additional window layer by atomic layer deposition23 or © 2016 American Chemical Society

Received: May 12, 2016 Accepted: September 20, 2016 Published: September 20, 2016 9267

DOI: 10.1021/acsnano.6b03175 ACS Nano 2016, 10, 9267−9273

Article

www.acsnano.org

Article

ACS Nano

Figure 1. (a) SEM cross-sectional image and schematic of a ZnO NP/PbS QD solar cell with the CdSe QD buffer layer. (b) Current density− voltage (J−V) characteristics of representative solar cells without (black) and with (red) a CdSe QD buffer layer under AM 1.5 illumination.

fill factor (FF). We perform a two-sample, two-tail unpaired t test on each of the statistical results, and the resulting p values are all much smaller than 0.05, indicating that the difference in device performance is statistically significant. To understand the role of the CdSe QD buffer layer in enhancing the PbS QD solar cell performance, we characterize the thin film constructed with 8 nm CdSe QDs optically and electronically. UV−vis absorption and Fourier transform infrared (FTIR) spectroscopy measurements show that the CdI2 treatment successfully displaces 82% of the long organic ligands introduced in synthesis and leads to a red shift and broadening of the first excitonic resonance (Figure 2a),

for ZnO NP/CdSe QD/PbS QD solar cells in comparison to reference cells without the CdSe QD layer.

RESULTS AND DISCUSSION The PbS QD solar cell device structure developed in this work is imaged in cross section under a scanning electron microscopy (SEM) and depicted schematically (Figure 1a). Compared to the commonly explored depleted heterojunction PbS QD solar cells, we introduce an extra buffer layer made of CdSe QDs at the interface between the ZnO NP and PbS QD layers. From the bottom up, the ZnO NPs (5 nm in diameter) are deposited by spin-coating in air to form a 100 nm thick layer. Using a robotic dip-coater mounted in a nitrogen glovebox, we fabricate CdSe QD (4 or 8 nm in diameter) layers by sequentially dipping samples into CdSe QD dispersions in hexane, methanolic CdI2 solutions, and pristine methanol. Multiple dip-coating cycles build up the CdSe QD layer to controlled thicknesses ranging from 20 to 60 nm. The PbS QD (3 nm in diameter) thin films are fabricated by sequentially building up multiple layers to realize an optimal thickness of 250 nm. Each PbS QD layer is deposited by spin-coating a PbS QD dispersion in octane and hexane mixture and immersing the samples in a methanolic solution of 3-mercaptopropionic acid as previously reported.17 Twelve nanometers of molybdenum trioxide (MoO3) and 65 nm of Au are deposited by thermal evaporation to create an electron-blocking layer and to form the anode, respectively, completing the device. Current density−voltage (J−V) characteristics of representative solar cells without and with the CdSe QD buffer layer are shown in Figure 1b. The device incorporating a 20 nm thick CdSe QD buffer layer constructed from 8 nm QDs has a larger VOC and JSC. The statistics and a histogram describing device performance for 20 independent devices fabricated on different substrates is reported in Table 1 and in Supporting Information Figure S1. Devices with the CdSe QD buffer layer show a 25% improvement in the PCE of the solar cells from 6.0% (±0.5) to 7.5% (±0.4). The enhanced PCE is attributed to the significantly increased JSC and an improved average VOC and

Figure 2. Characterization of the 8 nm CdSe QD buffer layer. (a) Absorption spectra of CdSe QD films as-deposited with organic ligands (gray), after treatment with CdI2 (black), and upon annealing at 250 °C (red). Inset: Corresponding IR absorption spectra of the CdSe QD film normalization by their absorption. (b) ID−VG (red) and C−V (blue) characteristics of CdI2-treated CdSe QD field-effect transistors.

consistent with an increase in electronic coupling. Further annealing of the film gives rise to a negligible change in absorption in both the UV−vis and the IR spectra. The elemental composition of the CdSe QD film determined by energy-dispersive X-ray (EDX) spectroscopy shows that CdI2 treatment enriches the CdSe QD surface in Cd and iodine (Supporting Information Table S1), which passivates the QD surface and reduces recombination as previously reported for II−VI and IV−VI QD thin films treated with metal halide salts.18,27 Figure 2b shows the current−voltage and capacitance− voltage (C−V) characteristics of a representative 8 nm CdSe QD thin film probed in the platform of the field-effect transistor (FET) to characterize the doping achieved by the CdI2 treatment. Both the drain current versus gate voltage (ID−VG) and C−V characteristics are consistent with n-doped CdSe QD films. We calculate the total concentration of accumulated charges in the FET channel by integrating the capacitance from

Table 1. Device Parameter Statistics N = 20

VOC (V)

JSC (mA/cm2)

FF

PCE (%)

ZnO NP/ PbS QD ZnO NP/ CdSe QD/ PbS QD

0.586 ± 0.017

25.0 ± 1.8

0.41 ± 0.03

6.0 ± 0.5

0.600 ± 0.017

29.5 ± 2.0

0.43 ± 0.02

7.5 ± 0.4

9268

DOI: 10.1021/acsnano.6b03175 ACS Nano 2016, 10, 9267−9273

Article

ACS Nano the extrapolated threshold voltage VTH of VG = −18 V to VG = 0 V and divide it by the volume defined by the channel area and the Debye length to obtain the electron concentration in the CdSe QD thin film.28 We find an electron concentration of ∼4 × 1018 /cm3 and therefore estimate the Fermi level to be 0.8 eV above midgap. The high electron concentration in the CdI2treated CdSe QD film is consistent with previous reports of Cd and iodine serving as donors.29,30 We apply the same analysis to the ZnO NP film (Supporting Information Figure S2) and find a higher electron concentration of ∼1 × 1020 /cm3. To study and correlate the effect of inserting the CdSe QD buffer layer on the band alignment and device performance, we exploit the size-dependent electronic structure of CdSe QDs and fabricate solar cells with smaller (4 nm in diameter) and larger (8 nm in diameter) CdSe QD buffer layers. We construct equilibrium band diagrams for heterojunctions with 8 and 4 nm CdSe QD buffer layers (Figure 3a,b). To find the band

and one at the CdSe QD/PbS QD interface of the junction (Figure 3b). We investigate the influence of the band alignment introduced by the CdSe QD buffer layer on carrier transport. We measure the J−V characteristics of the ZnO NP/CdSe QD/ PbS QD junctions in the dark and then under “red” light and standard AM 1.5 white light illumination. The “red” light is obtained from AM 1.5 solar simulated light that is additionally filtered by an 800 nm long-pass filter, such that the red light can be selectively absorbed by the PbS QD layer and not by the CdSe QD layer. Figure 3c shows representative J−V characteristics of a PbS QD solar cell with the 8 nm CdSe QD buffer layer. The device exhibits well-behaved, rectifying characteristics of a diode in the dark and under both red and white light illumination conditions, consistent with the smooth transition in the conduction band of the junction shown in Figure 3a. The device with the 4 nm CdSe QD buffer layer has lower current densities both in dark and under illumination (Figure 3d), consistent with barriers to electron transport in the conduction band (Figure 3b). Under red light illumination, the J−V curve has an “s shape”, which is similar to that seen and commonly referred to in CIGS solar cells as a “red kink”.32 The distorted J−V characteristics in CIGS solar cells is attributed to the secondary barrier created by the conduction band offset between the CdS buffer layer and the CIGS active material, akin to the conduction band offset in the PbS QD solar cell with the 4 nm CdSe QD buffer layer reported here. Red photons have insufficient energy to generate carriers within the wide band gap buffer layer, and at small forward bias, the interfacial barriers are too high for photogenerated carriers to surmount the barrier and transit the junction; thus the total current density decreases, leading to a kink. Under AM 1.5 white light illumination, compared to the well-behaved J−V characteristics from the device with the 8 nm CdSe QD buffer layer, the one constructed with the 4 nm CdSe QD buffer layer stills show a kink. We study the J−V characteristics after white light soaking of the junction for 15 min. In CIGS cells, the high-energy photons of white light are used to increase the carrier concentration in the buffer layer and reduce the barrier height. For devices fabricated with the 8 nm CdSe QD buffer layer, the J−V characteristics are unchanged after AM 1.5 white light soaking for 15 min. However, for devices constructed with the 4 nm CdSe QD buffer layer, the kink in the J−V characteristics nearly disappears, similar to the behavior observed in CIGS solar cells.33 We optimize the thickness of the CdSe QD layer to maximize the device PCE. The J−V characteristics of solar cells with CdSe QD buffer layers of varying thickness are shown in Supporting Information Figure S4. As the thickness of the CdSe QD layer increases from 20 to 60 nm, the enhancement in the J−V characteristics weakens and ultimately degrades the device performance. This is consistent with (1) a smaller depletion region width in the CdSe QD layer compared to that in the PbS QD layer, as the CdSe QD layer has a higher carrier concentration and lower dielectric constant; (2) increased recombination in the neutral region of thicker CdSe QD layers and therefore lower VOC; and (3) addition of device resistance with increasing semiconductor layer thickness. CdSe QD layers with a thickness less than 20 nm are not uniform, and therefore, thinner buffer layers are not included in this study. To distinguish any effect of CdI2 treatment at the PbS QD/ ZnO NP interface from incorporating the CdI2-treated CdSe QD interface layer, we fabricate devices in which we replace the

Figure 3. Band alignment of the ZnO NP/CdSe QD/PbS QD heterojunction with (a) 8 nm CdSe QD buffer layer and (b) 4 nm CdSe QD buffer layer. (c) J−V characteristics of devices constructed with (c) 8 nm CdSe QD buffer and (d) 4 nm CdSe QD buffer layer in the dark (black), under red light illumination (red), upon AM 1.5 white light illumination (dashed line, purple) and after 15 min of white light soaking (blue line).

energies, we combine UV−vis absorption and cyclic voltammetry measurements (Supporting Information Figure S3). To identify the length scales of band bending at each of the device heterointerfaces, we use dielectric constants14,31 and carrier concentrations found in literature6 and from C−V measurements (Supporting Information Table S2). For the 8 nm CdSe QD interface layer, at the CdSe QD/ZnO NP isotype interface, we calculate an electron accumulation (depletion) region of 2 nm (0.2 nm) in the CdSe QD (ZnO NP) layer. At the PbS QD/CdSe QD anisotype interface, we find an 8 nm (162 nm) depletion region in the CdSe QD (PbS QD) layer. A ∼10 nm quasi-neutral region remains in the CdSe QD layer away from the device interfaces. Since the conduction band of the 8 nm CdSe QD layer lies below that of the PbS QD layer relative to the vacuum level, that is, the conduction band offset between the CdSe and PbS QD layer is negative (ΔEC < 0), once the junction is formed, band bending provides a smooth transition in the conduction band across the heterointerface (Figure 3a). In contrast, wider band gap, 4 nm CdSe QDs introduce a larger and positive band offset (ΔEC > 0) that creates two barriers in the conduction band, one at the ZnO NP/CdSe QD interface 9269

DOI: 10.1021/acsnano.6b03175 ACS Nano 2016, 10, 9267−9273

Article

ACS Nano

optimized for the solar cells. The ZnO NP/CdSe QD/PbS QD stacks show an increase in the QY−mobility product in comparison to those without the CdSe QD layer. Since the carriers being probed exist predominately in the PbS QD or ZnO NP layers, as the CdSe QD layer is relatively thin, we expect carrier mobility to remain unchanged and, therefore, the increase in the QY−mobility product to arise from an increase in QY. We hypothesize that the increase in QY with the additional CdSe QD layer is caused by a reduction in interface recombination and more efficient charge extraction across the heterojunction. Noting that the photoexcitation is introduced by a 532 nm laser, we normalize the QY−mobility product to the film absorption, so the increase is also not caused by the added absorption from the CdSe QD layer. TRMC photoconductance lifetimes are shown for completeness in Supporting Information Figure S7. To quantitatively study the role of the CdSe QD layer in changing device parameters, we analyze the J−V characteristics of the ZnO NP/PbS QD and ZnO NP/CdSe QD/PbS QD solar cells. Due to the relatively low fill factor of these devices, the uncertainty from analysis would be too large to make clear statements for J−V data under illumination.34 As a result, all the following operations are only applied to the dark J−V characteristics. Simplified from the general diode equation, the total diode current in the dark is expressed as

CdSe QDs with the same PbS QDs as the rest of the active layer. All other parameters are identical during the processing. From the characterization of 12 different samples (Supporting Information Table S3), devices with CdI2-treated PbS QD interface layers show a lower average VOC and JSC and therefore a lower average PCE. A representative J−V curve for a device with a CdI2-treated PbS QD interface layer is shown in Supporting Information Figure S5 and compared to that of a ZnO NP/PbS QD reference device. The reduced device performance upon incorporating the CdI2-treated PbS QD interface layer is consistent with literature reports of the influence of this metal halide treatment.18 We study the underlying mechanism of the PCE enhancement by the CdSe QD buffer layer. Figure 4a shows that the

Figure 4. (a) EQE of PbS QD solar cells without the CdSe QD buffer layer (black) and with the CdSe QD buffer layer (red). Change of EQE (blue), ΔEQE = EQE (red) − EQE (black). (b) TRMC measurements of the product of the carrier quantum yield (Φ) and the sum of the carrier mobilities in ZnO NP/PbS QD (black) and ZnO NP/CdSe QD/PbS QD (red) films.

⎡ q ⎤ J = J0 exp⎢ (V − RJ )⎥ + GV ⎣ nkT ⎦

where J0 is the saturation current density, q is elementary charge, n is the ideality factor, k is Boltzmann’s constant, T is temperature, R is series resistance, and G is shunt conductance. By plotting dJ/dV against V, dV/dJ against (J−GV)−1, and J− GV against V−RJ (Supporting Information Figure S8), we extract G, R, n, and J0, respectively. Table 2 summarizes the device parameters obtained from the analysis of the dark J−V characteristics. With the CdSe QD buffer layer, G and J0 are reduced, indicating that the buffer layer prevents shunt leakage and suppresses recombination in dark operation. For the J−V characteristics under illumination, the open-circuit voltage VOC obeys the relation

device with the CdSe QD buffer layer has higher EQE over the entire measured wavelength range compared to an otherwise identical cell without the CdSe buffer layer. For photon energies smaller than the band gap energy of the CdSe QD layer (1.85 eV, wavelengths > 670 nm), the net EQE enhancement (Figure 4a, blue) is roughly constant, consistent with the CdSe QD buffer layer reducing interface recombination and facilitating carrier collection. For photon energies larger than the CdSe QD band gap, with wavelengths < 670 nm, the ΔEQE increases, consistent with higher absorbance in the device stack (Supporting Information Figure S6) and generation and collection of carriers in the CdSe QD buffer layer. Unlike the loss that occurs in the highly resistive and wide band gap buffer layer in the analogous CIGS thin film devices,26 the 8 nm CdSe QD layer in the PbS QD solar cells acts to contribute photogenerated carriers to the device efficiency. We use time-resolved microwave conductivity measurements to characterize the product of the carrier quantum yield (QY) (Φ) and the sum of the carrier mobilities (∑μ) in the PbS QD heterojunction without and with the CdSe QD buffer layers. Figure 4b shows the QY−mobility product at various photoexcitation densities from three independent samples of ZnO NP/PbS QD and ZnO NP/CdSe QD/PbS QD stacks assembled on (3-mercaptopropyl)trimethoxysilane (MPTS)treated float glass following the same fabrication procedure

VOC =

nkT ⎛ JSC ⎞ ln⎜⎜ ⎟⎟ q ⎝ J0 ⎠

when JSC ≫ J0. Based on the VOC from the J−V curve under illumination and the calculated J0, we find n for the ZnO NP/ PbS QD and ZnO NP/CdSe QD/PbS QD devices under illumination to be 1.9 ± 0.1 and 1.7 ± 0.1, respectively. The smaller n is consistent with reduced recombination in the device with the buffer layer.

CONCLUSIONS In conclusion, we improve the performance of PbS QD solar cells by introducing a CdI2−CdSe QD buffer layer at the ZnO NP/PbS QD heterojunction. The 8 nm CdSe QD buffer layer enhances the device JSC, VOC, and FF. We attribute these

Table 2. Device Parameters Based on the Dark Characteristics

ZnO NP/PbS QD ZnO NP/CdSe QD/PbS QD

G (mS/cm2)

R (Ω·cm2)

n

J0 (mA/cm−2)

1.2 ± 0.2 0.6 ± 0.2

6.1 ± 2.2 7.3 ± 1.7

2.4 ± 0.4 2.2 ± 0.2

2.9 ± 0.3 × 10−4 1.0 ± 0.1 × 10−4

9270

DOI: 10.1021/acsnano.6b03175 ACS Nano 2016, 10, 9267−9273

Article

ACS Nano

coater mounted in a nitrogen glovebox. With a withdrawal rate of 50 mm/min, the substrate is dipped into the CdSe QD dispersion in hexane (10 mg/mL), CdI2 solution (24 mg/mL in methanol), and pristine methanol sequentially, during which the substrate is held for 60 s in the CdI2 ligand solution and 30 s in the methanol rinse. Between each step, there is enough waiting time to allow the solvent to evaporate. This process is repeated to build up a CdSe QD film of desired thickness. Once the CdSe QD film is deposited, it is annealed at 250 °C for 10 min before being transferred out of the glovebox. On top of the CdSe QD layer, 250 nm thick PbS QD films are deposited by sequentially spin-coating the PbS QD dispersion at 2500 rpm and exchanging the long organic ligands used in the synthesis for shorter MPA ligands. Ligand exchange is carried out by immersing samples in an MPA solution (1% volume ratio in methanol) and then rinsing three times in methanol. The fabrication of the PbS QD layers is done in a fume hood with humidity below 30% at 25 °C. The top metal contact, 12 nm of MoO3 and 65 nm of Au, is deposited by thermal evaporation, using an evaporator mounted inside a glovebox, through shadow masks, to define active device areas of 2 mm by 2 mm. The device cross-section sample for SEM imaging is prepared using the focused ion beam (FIB, FET Strata DB235). We first deposit platinum (400 nm) on top of the device to prevent damage from the Ga+ ion beam that is used to mill trenches on the device and cut an area free from the substrate. After a typical lift-out process, the crosssection specimen is attached to a copper grid for final thinning to remove undesired platinum in the area of interest. We then observe the sample under a high-resolution SEM (JEOL 7500F). CdSe FET Preparation. A highly n-doped silicon wafer with a 250 nm thick thermally grown oxide (Silicon Inc.) and a 20 nm layer of Al2O3 (20 nm), applied by atomic layer deposition, is used to fabricate FETs. Before use, the wafer is cleaned using acetone and isopropyl alcohol and treated by UV−ozone for 25 min. Similarly to the solar cell fabrication, CdSe QDs are deposited by dipping the substrate sequentially in the QD dispersion, CdI2 methanolic solution, and pristine methanol using the robotic dip-coater. The FET channel (10 μm × 150 μm) is defined as evaporation of Au electrodes through a shadow mask. Characterization. UV−vis absorption and IR absorption spectra are collected using a Cary 5000 and a Thermo-Fisher FTIR spectrometer (model 6700), respectively. Samples are prepared on float glass substrates and tested in air. The plotted IR spectrum is obtained by normalization to the visible absorption of the same sample. EDX measurements are carried out in an FEI Quanta 600 ESEM using an Oxford Instruments energy-dispersive spectrometer operating at an accelerating voltage of 10 kV and collected using a 30 s integration time. The result is averaged over five areas (2000 μm × 1500 μm) on each sample. FET measurements are performed on a Karl Suss PM5 probe station mounted in a nitrogen glovebox with a model 4156C semiconductor parameter analyzer (Agilent). Using the same probe station setup, the capacitance−voltage measurements are conducted using a model 4192A LF impedance analyzer (Agilent). The low terminal of the LCR meter is connected to source and drain electrodes which are electrically shorted, while the high terminal is connected to the gate electrode. The data are collected at 10 kHz. Current density−voltage characteristics are obtained with a Keithley 2420 sourcemeter and under illumination using a solar simulator (Oriel instruments model 96000, Newport Co.) that is mounted inside a N2 glovebox. The simulated AM 1.5 light is brought into the glovebox through a liquid light guide feedthrough. The intensity at the device is calibrated to be 1 sun, 100 mW/cm2 by a Si reference cell and meter from Newport (model 91150). Solar cells are illuminated through an aperture of 1.6 mm by 1.6 mm in size, smaller than the active device active area. EQE spectra are measured by illuminating samples mounted in a Horiba FluoroLog with light from a 450 W xenon short arc-lamp (Ushio Inc.) filtered through its monochromator and collected with a Keithley 2400 sourcemeter. The illumination intensity is measured by a power meter (Coherent, FieldMaxII). Since the measurement is performed under ambient atmosphere, the sample

improvements to the optimized band alignment across the junction, passivation of interface traps by the buffer layer, and additional carriers photogenerated in the CdSe QD layer. We perform a complementary mathematical analysis of device J−V characteristics, and the improved device parameter is in agreement with our hypothesis based on our experimental observations. For future applications, the CdSe QD buffer layer can be applied to air-stable PbS/Se QD solar cells20,22,35 and to devices with structured, nonplanar geometries.36

EXPERIMENTAL SECTION Materials. Oleic acid (OA, 90%), 1-octadecene (ODE, 90%), lead oxide (PbO, 99.999%), cadmium chloride (CdCl2, 99.99%), oleylamine (70%), selenium powder (99.99%), zinc acetate dehydrate (98%), 3-mercaptopropionic acid (MPA, 99%), (3-mercaptopropyl)trimethoxysilane (MPTS, 95%), anhydrous hexane, anhydrous 2propanol, anhydrous methanol, anhydrous toluene, and anhydrous acetone are purchased from Sigma-Aldrich. Bis(trimethylsilyl) sulfide (TMS, 95%) is purchased from Acros. Cadmium oxide (CdO, 99.99%) is purchased from Strem. Tetradecylphosphonic acid (TDPA, 98%) and molybdenum oxide (MoO3, 99.9995%) are purchased from Alfa Aesar. Potassium hydroxide (KOH, 85%) is purchased from Fisher. QD Synthesis. PbS QDs are synthesized following a modified literature recipe.17 Using standard Schlenk line techniques, a mixture of 0.47 g of PbO, 23 mL of ODE, and 2 mL of OA is degassed at 120 °C for 2 h under vacuum in a reaction flask. Then the flask is switched to N2 and stabilized at 100 °C. Meanwhile, a 42 μL TMS/2 mL ODE solution is prepared inside a nitrogen-filled glovebox, and a predried CdCl2 solution (0.3 g of CdCl2 and 0.033 g of TDPA in 5 mL of oleylamine) is heated to 70 °C inside the glovebox. Next, 5 mL of TMS precursor is quickly injected into the reaction flask with the heat turned off. During the slow cooling process, 1 mL of the CdCl2 solution is injected once the reaction mixture reaches 70 °C, and then after 5 min, the reaction is quenched by being cooled in a water bath. The reaction flask is transferred into the glovebox without air exposure. The PbS QDs are purified by washing four times with acetone, 2-propanol, and methanol and finally redispersed in octane/ hexane (4:1) at 40 mg/mL. CdSe QDs (4 and 8 nm diameter) are synthesized according to a previous report.27 Briefly, to make the 4 nm CdSe QDs, CdO (77 mg) is dissolved by myristic acid (275 mg) in octadecene (37 mL) at 250 °C. Next, selenium powder (24 mg) is added to the solution after the solution has been evacuated for 30 min at 100 °C, and the new mixture is evacuated for an additional 10 min. The reaction temperature is increased to 240 °C under nitrogen atmosphere. After the mixture is stabilized at 240 °C for 3 min, a pre-degassed solution of oleylamine (1 mL), oleic acid (1 mL), and octadecene (4 mL) is injected into the reaction at 1 mL/min. The reaction is maintained at 240 °C for 30 min and then cooled to quench the reaction. For the synthesis of the larger 8 nm CdSe QDs, the reaction is heated to 280 °C and an extra mixture (36 mL, 1:5 by volume) of Cd-oleate in oleic acid (0.5 M) and Se dissolved in octadecene (0.1 M) is introduced at 0.2 mL/min. The reaction is quenched by being cooled to room temperature after completion of the injection. CdSe QDs are transferred to and purified in a glovebox by being washed with acetone as the antisolvent and stored in hexane. Solar Cell Fabrication. Prepatterned ITO/glass substrates (Thin Film Devices) are cleaned by sonication in 5% Hellmanex in DI water, pure DI water, and ethanol consecutively, followed by UV−ozone treatment for 30 min and MPTS (5% in toluene) soaking for 10 h before use.37 The substrates are rinsed with toluene and sonicated in ethanol to remove excess unbound MPTS molecules. ZnO NPs, synthesized according to the literature procedure,38 are deposited by being spin-coated onto the ITO substrates at 1800 rpm for 1 min to achieve a film thickness of about 100 nm. The film is annealed at 250 °C for 20 min to dry. The CdSe QD film is assembled over the ZnO layer by dip-coating, using a KSV NIMA robotic dip9271

DOI: 10.1021/acsnano.6b03175 ACS Nano 2016, 10, 9267−9273

Article

ACS Nano is encapsulated by epoxy and cover glass before being taken out of the glovebox to avoid degradation. Cyclic voltammetry measurements are made using an electrochemistry workstation (Epsilon, C-3 cell stand) mounted inside a N2 glovebox. The electrolyte used is 10 mM tetrabutylammonium hexafluorophosphate in acetonitrile. The three-electrode system is satisfied by depositing QD or NP films on conductive substrates that serve as the working electrode and using Ag/AgNO3 in the electrolyte solution (10 mM) as the reference electrode and a Pt wire as the auxiliary electrode. To make the conductive substrates, Si wafers are coated by 5 nm of Cr and 40 nm of Pd following the MPTS treatment. All potentials are calibrated by a ferrocene/ferrocenium redox couple measured under identical conditions. The setup and theory of TRMC measurements are described in refs 39 and 14. Briefly, the sample is loaded into a microwave cavity with a resonant frequency of about 9 GHz. Pulsed photoexcitation from a 532 nm laser (Continuum Minilite II) is used to illuminate the thin film stack of interest, generating photocarriers. The photocarriers change the film conductivity, ΔG, which proportionally changes the reflected microwave power, ΔP. The quantum yield mobility product is calculated as

⌀(μe + μ h ) =

and Engineering, under Award No. DE-SC0002158. Scanning electron microscopy was performed in facilities supported by the NSF MRSEC Program under Award No. DMR-1120901.

REFERENCES (1) Talapin, D. V.; Lee, J.-S.; Kovalenko, M. V.; Shevchenko, E. V. Prospects of Colloidal Nanocrystals for Electronic and Optoelectronic Applications. Chem. Rev. 2010, 110, 389−458. (2) McDonald, S. A.; Konstantatos, G.; Zhang, S.; Cyr, P. W.; Klem, E. J. D.; Levina, L.; Sargent, E. H. Solution-Processed PbS Quantum Dot Infrared Photodetectors and Photovoltaics. Nat. Mater. 2005, 4, 138−142. (3) Murray, C. B.; Kagan, C. R.; Bawendi, M. G. Synthesis and Characterization of Monodisperse Nanocrystals and Close-Packed Nanocrystal Assemblies. Annu. Rev. Mater. Sci. 2000, 30, 545−610. (4) Brown, P. R.; Kim, D.; Lunt, R. R.; Zhao, N.; Bawendi, M. G.; Grossman, J. C.; Bulović, V. Energy Level Modification in Leadsulfide Quantum Dot Thin Films through Ligand Exchange. ACS Nano 2014, 8, 5863−5872. (5) Oh, S. J.; Wang, Z.; Berry, N. E.; Choi, J.-H.; Zhao, T.; Gaulding, E. A.; Paik, T.; Lai, Y.; Murray, C. B.; Kagan, C. R. Engineering Charge Injection and Charge Transport for High Performance PbSe Nanocrystal Thin Film Devices and Circuits. Nano Lett. 2014, 14, 6210−6216. (6) Voznyy, O.; Zhitomirsky, D.; Stadler, P.; Ning, Z.; Hoogland, S.; Sargent, E. H. A Charge-Orbital Balance Picture of Doping in Colloidal Quantum Dot Solids. ACS Nano 2012, 6, 8448−8455. (7) Oh, S. J.; Berry, N. E.; Choi, J.-H.; Gaulding, E. A.; Lin, H.; Paik, T.; Diroll, B. T.; Muramoto, S.; Murray, C. B.; Kagan, C. R. Designing High-Performance PbS and PbSe Nanocrystal Electronic Devices through Stepwise, Post-Synthesis, Colloidal Atomic Layer Deposition. Nano Lett. 2014, 14, 1559−1566. (8) Lee, J.-S.; Kovalenko, M. V.; Huang, J.; Chung, D. S.; Talapin, D. V. Band-like Transport, High Electron Mobility and High Photoconductivity in All-Inorganic Nanocrystal Arrays. Nat. Nanotechnol. 2011, 6, 348−352. (9) Choi, J.-H.; Fafarman, A. T.; Oh, S. J.; Ko, D.-K.; Kim, D. K.; Diroll, B. T.; Muramoto, S.; Gillen, J. G.; Murray, C. B.; Kagan, C. R. Bandlike Transport in Strongly Coupled and Doped Quantum Dot Solids: A Route to High-Performance Thin-Film Electronics. Nano Lett. 2012, 12, 2631−2638. (10) Liu, Y.; Tolentino, J.; Gibbs, M.; Ihly, R.; Perkins, C. L.; Liu, Y.; Crawford, N.; Hemminger, J. C.; Law, M. PbSe Quantum Dot FieldEffect Transistors with Air-Stable Electron Mobilities above 7 cm2 V−1 S−1. Nano Lett. 2013, 13, 1578−1587. (11) de Kergommeaux, A.; Fiore, A.; Faure-Vincent, J.; Chandezon, F.; Pron, A.; de Bettignies, R.; Reiss, P. Highly Conductive CuInSe2 Nanocrystals with Inorganic Surface Ligands. Mater. Chem. Phys. 2012, 136, 877−882. (12) Jeong, K. S.; Tang, J.; Liu, H.; Kim, J.; Schaefer, A. W.; Kemp, K.; Levina, L.; Wang, X.; Hoogland, S.; Debnath, R.; Brzozowski, L.; Sargent, E. H.; Asbury, J. B. Enhanced Mobility-Lifetime Products in PbS Colloidal Quantum Dot Photovoltaics. ACS Nano 2012, 6, 89− 99. (13) Thon, S. M.; Ip, A. H.; Voznyy, O.; Levina, L.; Kemp, K. W.; Carey, G. H.; Masala, S.; Sargent, E. H. Role of Bond Adaptability in the Passivation of Colloidal Quantum Dot Solids. ACS Nano 2013, 7, 7680−7688. (14) Straus, D. B.; Goodwin, E. D.; Gaulding, E. A.; Muramoto, S.; Murray, C. B.; Kagan, C. R. Increased Carrier Mobility and Lifetime in CdSe Quantum Dot Thin Films through Surface Trap Passivation and Doping. J. Phys. Chem. Lett. 2015, 6, 4605−4609. (15) Carey, G. H.; Abdelhady, A. L.; Ning, Z.; Thon, S. M.; Bakr, O. M.; Sargent, E. H. Colloidal Quantum Dot Solar Cells. Chem. Rev. 2015, 115, 12732−12763. (16) Carey, G. H.; Levina, L.; Comin, R.; Voznyy, O.; Sargent, E. H. Record Charge Carrier Diffusion Length in Colloidal Quantum Dot

ΔGmax AqI0FA

where A is a dimension factor of the resonant cavity, q the electron charge, I0 the photon fluence, and FA the fraction of photons absorbed in the film at 532 nm. The TRMC transients are filtered using a Butterworth filter of the fifth order, and the cavity response is deconvolved from the traces.

ASSOCIATED CONTENT S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.6b03175. Histogram of device PCEs, CdSe QD thin film elemental analysis, optical and electrical characterization of ZnO NP thin films, cyclic voltammetry measurements, schematic energy level diagrams, electrostatic calculations of band bending across device heterointerfaces, J−V characteristics as a function of CdSe QD buffer layer thickness and analysis, device parameter statistics and representative J−V characteristics with the CdI2-treated PbS QD interface layers, and absorption and photoconductance decay lifetime for PbS QD solar cells (PDF)

AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS The authors are grateful for primary support of this work from Center for Advanced Solar Photophysics, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences for electrical characterization of CdSe QD and ZnO NP thin films and PbS QD solar cell fabrication, characterization, and analysis. PbS QD synthesis, CdSe QD and PbS QD ligand exchange processes, and electronic band structure characterization of NP and QD films were supported by NSF Award No. CBET126406. CdSe QD synthesis and SEM and EDX characterization were supported by the U.S. Department of Energy Office of Basic Energy Sciences, Division of Materials Science 9272

DOI: 10.1021/acsnano.6b03175 ACS Nano 2016, 10, 9267−9273

Article

ACS Nano Solids via Mutual Dot-To-Dot Surface Passivation. Adv. Mater. 2015, 27, 3325−3330. (17) Ip, A. H.; Thon, S. M.; Hoogland, S.; Voznyy, O.; Zhitomirsky, D.; Debnath, R.; Levina, L.; Rollny, L. R.; Carey, G. H.; Fischer, A.; Kemp, K. W.; Kramer, I. J.; Ning, Z.; Labelle, A. J.; Chou, K. W.; Amassian, A.; Sargent, E. H. Hybrid Passivated Colloidal Quantum Dot Solids. Nat. Nanotechnol. 2012, 7, 577−582. (18) Crisp, R. W.; Kroupa, D. M.; Marshall, A. R.; Miller, E. M.; Zhang, J.; Beard, M. C.; Luther, J. M. Metal Halide Solid-State Surface Treatment for High Efficiency PbS and PbSe QD Solar Cells. Sci. Rep. 2015, 5, 9945. (19) Lan, X.; Voznyy, O.; Kiani, A.; García de Arquer, F. P.; Abbas, A. S.; Kim, G.-H.; Liu, M.; Yang, Z.; Walters, G.; Xu, J.; Yuan, M.; Ning, Z.; Fan, F.; Kanjanaboos, P.; Kramer, I.; Zhitomirsky, D.; Lee, P.; Perelgut, A.; Hoogland, S.; Sargent, E. H. Passivation Using Molecular Halides Increases Quantum Dot Solar Cell Performance. Adv. Mater. 2016, 28, 299−304. (20) Chuang, C.-H. M.; Brown, P. R.; Bulović, V.; Bawendi, M. G. Improved Performance and Stability in Quantum Dot Solar Cells through Band Alignment Engineering. Nat. Mater. 2014, 13, 796−801. (21) Kim, G.-H.; Garcia de Arquer, F. P.; Yoon, Y. J.; Lan, X.; Liu, M.; Voznyy, O.; Yang, Z.; Fan, F.; Ip, A. H.; Kanjanaboos, P.; Hoogland, S.; Kim, J. Y.; Sargent, E. H. High Efficiency Colloidal Quantum Dot Photovoltaics via Robust Self-Assembled Monolayers. Nano Lett. 2015, 15, 7691−7696. (22) Zhang, J.; Gao, J.; Church, C. P.; Miller, E. M.; Luther, J. M.; Klimov, V. I.; Beard, M. C. PbSe Quantum Dot Solar Cells with More than 6% Efficiency Fabricated in Ambient Atmosphere. Nano Lett. 2014, 14, 6010−6015. (23) Kemp, K. W.; Labelle, A. J.; Thon, S. M.; Ip, A. H.; Kramer, I. J.; Hoogland, S.; Sargent, E. H. Interface Recombination in Depleted Heterojunction Photovoltaics Based on Colloidal Quantum Dots. Adv. Energy Mater. 2013, 3, 917−922. (24) Tan, F.; Wang, Z.; Qu, S.; Cao, D.; Liu, K.; Jiang, Q.; Yang, Y.; Pang, S.; Zhang, W.; Lei, Y.; Wang, Z. A CdSe Thin Film: A Versatile Buffer Layer for Improving the Performance of TiO2 Nanorod array:PbS Quantum Dot Solar Cells. Nanoscale 2016, 8, 10198− 10204. (25) Yuan, M.; Voznyy, O.; Zhitomirsky, D.; Kanjanaboos, P.; Sargent, E. H. Synergistic Doping of Fullerene Electron Transport Layer and Colloidal Quantum Dot Solids Enhances Solar Cell Performance. Adv. Mater. 2015, 27, 917−921. (26) Jackson, P.; Hariskos, D.; Lotter, E.; Paetel, S.; Wuerz, R.; Menner, R.; Wischmann, W.; Powalla, M. New World Record Efficiency for Cu(In,Ga)Se2 Thin-Film Solar Cells beyond 20%. Prog. Photovoltaics 2011, 19, 894−897. (27) Goodwin, E. D.; Diroll, B. T.; Oh, S. J.; Paik, T.; Murray, C. B.; Kagan, C. R. Effects of Post-Synthesis Processing on CdSe Nanocrystals and Their Solids: Correlation between Surface Chemistry and Optoelectronic Properties. J. Phys. Chem. C 2014, 118, 27097−27105. (28) Oh, S. J.; Berry, N. E.; Choi, J.-H.; Gaulding, E. A.; Paik, T.; Hong, S.-H.; Murray, C. B.; Kagan, C. R. Stoichiometric Control of Lead Chalcogenide Nanocrystal Solids to Enhance Their Electronic and Optoelectronic Device Performance. ACS Nano 2013, 7, 2413− 2421. (29) Kim, D. K.; Fafarman, A. T.; Diroll, B. T.; Chan, S. H.; Gordon, T. R.; Murray, C. B.; Kagan, C. R. Solution-Based Stoichiometric Control over Charge Transport in Nanocrystalline CdSe Devices. ACS Nano 2013, 7, 8760−8770. (30) Zhitomirsky, D.; Furukawa, M.; Tang, J.; Stadler, P.; Hoogland, S.; Voznyy, O.; Liu, H.; Sargent, E. H. N-Type Colloidal-QuantumDot Solids for Photovoltaics. Adv. Mater. 2012, 24, 6181−6185. (31) Grinolds, D. D. W.; Brown, P. R.; Harris, D. K.; Bulovic, V.; Bawendi, M. G. Quantum-Dot Size and Thin-Film Dielectric Constant: Precision Measurement and Disparity with Simple Models. Nano Lett. 2015, 15, 21−26.

(32) Kanevce, A.; Gloeckler, M.; Pudov, A. O.; Sites, J. R. Conduction-Band-Offset Rule Governing J-V Distortion in CdS/ CI(G)S Solar Cells. MRS Online Proc. Libr. 2005, 865, 32. (33) Pudov, A. O.; Kanevce, A.; Al-Thani, H. A.; Sites, J. R.; Hasoon, F. S. Secondary Barriers in CdS−CuIn1−xGaxSe2 Solar Cells. J. Appl. Phys. 2005, 97, 064901. (34) Hegedus, S. S.; Shafarman, W. N. Thin-Film Solar Cells: Device Measurements and Analysis. Prog. Photovoltaics 2004, 12, 155−176. (35) Cao, Y.; Stavrinadis, A.; Lasanta, T.; So, D.; Konstantatos, G. The Role of Surface Passivation for Efficient and Photostable PbS Quantum Dot Solar Cells. Nat. Energy 2016, 1, 16035. (36) Labelle, A. J.; Thon, S. M.; Masala, S.; Adachi, M. M.; Dong, H.; Farahani, M.; Ip, A. H.; Fratalocchi, A.; Sargent, E. H. Colloidal Quantum Dot Solar Cells Exploiting Hierarchical Structuring. Nano Lett. 2015, 15, 1101−1108. (37) Katari, J. E. B.; Colvin, V. L.; Alivisatos, A. P. X-Ray Photoelectron Spectroscopy of CdSe Nanocrystals with Applications to Studies of the Nanocrystal Surface. J. Phys. Chem. 1994, 98, 4109− 4117. (38) Beek, W. J. E.; Wienk, M. M.; Kemerink, M.; Yang, X.; Janssen, R. a J. Hybrid Zinc Oxide Conjugated Polymer Bulk Heterojunction Solar Cells. J. Phys. Chem. B 2005, 109, 9505−9516. (39) Goodwin, E. D.; Straus, D. B.; Gaulding, E. A.; Murray, C. B.; Kagan, C. R. The Effects of Inorganic Surface Treatments on Photogenerated Carrier Mobility and Lifetime in PbSe Quantum Dot Thin Films. Chem. Phys. 2016, 471, 81−88.

9273

DOI: 10.1021/acsnano.6b03175 ACS Nano 2016, 10, 9267−9273