Hydroiodic Acid Additive Enhanced the Performance and Stability of PbS-QDs Solar Cells via Suppressing Hydroxyl Ligand

Highlights The hydroiodic acid was explored systematically to modify PbS quantum dots (QDs) ink process, which could remove trap states by hydroxyl ligand and improve iodine passivation on the PbS-QDs surface. This strategy solved the essential question of PbS-QDs ink process and showed the favorable application prospects in QDs technology. Electronic supplementary material The online version of this article (10.1007/s40820-020-0372-z) contains supplementary material, which is available to authorized users.


S1.2.1 Preparation of ZnO Film by Sol-gel Method
The ZnO sol-gel precursor was prepared according to a previously reported procedure [S1]. The ZnO precursor was spin-coated on ITO glass at 4000 r min -1 for 30 s under ambient environment, followed by annealing 320 °C for 12 min. This process was repeated a few times to reach the required thickness.

S1.2.2 Lead Sulfide Colloidal Quantum Dots (PbS CQDs) Synthesis
Oleate-capped PbS CQDs were synthesized under Schlenk-line conditions according to previous reports [S2] with slight modifications [S1]. A mixture of lead oxide (4 mmol, 0.9 g), oleic acid (9.5 mmol, 3 mL) and 1-octadecene (20 mL) in a flask was gradually increased to target temperature and degassed for 12 h. After that, 2 mmol TMS was dissolved in 20 mL ODE, and then injected quickly into lead oleate solution under vigorous stirring at 120 °C. The reaction took a few minutes and then it cooled to room temperature naturally. QDs were purified in air by adding octane to dissolve and precipitated by acetone with centrifugation. This process was repeated three times, and the final separated QDs were re-dispersed in Octane with a 30 mg mL -1 for solar cell fabrication.

S1.2.3 Device Fabrication
PbS CQD films fabricated by solution-phase ligand-exchange process serves as the main light-absorbing layer, the oleic acid-capped CQDs (OA-CQDs) could be changed into halide-passivated CQDs under air as described in previous reports [S3]. Halide ligand was prepared by PbI2-DMF solution (691 mg/5 mL) for ligand exchange. 5 mL of OA-CQDs in octane (15 mg mL -1 ) were mixed with the as-prepared DMF solution. For complete transfer of CQD toward DMF, this mixed solution was vortexed for 1.5 min at room temperature. The ligand-exchanged solution, CQD dispersed in DMF was washed by octane three times for removing remained residues. After washing, toluene was added to the ligand exchanged solution for precipitation of CQD, and the CQDs were totally collected by centrifugation. The PbS CQDs were dried in vacuum for 20 min to get CQD powder. The obtained iodide-passivated PbS CQDs were re-dispersed in mixed solvent Butylamine (BTA) and N,N-dimethylformamide (DMF) with desired concentrations for absorber deposition. The volume ratio of this mixed solvent is VBTA: VDMF = 9:1. For modified QDs-ink process, it is similar to above process and the only difference is the amount of HI additive in PbI2 ligand solution. We adding HI (45%) with different mole ratio related to PbI2 in PbI2-DMF solution. For example, the 2% HI need add 3 μL HI solution into 5 mL PbI2-DMF solution. After that, two PbS-EDT layers as an hole extraction layer were fabricated via a layer-by-layer method. OA-CQDs were dropped onto pre-deposited PbS-PbI2 layer and spin-cast at 2500 rpm for 18 s, and then a 0.01 vol% 1,2-Ethanedithiol in Acetonitrile solution was loaded on the film for 30 s, then spin-casted at 2500 rpm for 10s, followed by washing with acetonitrile for 2 times. This process repeated one time more. Finally, ~80 nm Au was deposited by thermal evaporation at low pressure (< 4 × 10 -3 Pa).
The w/o HI-PbS+TMAH powders or films were similar to above modified QDs-ink process, the only difference is the HI additive are replaced by TMAH (with 2% mole ratio related to PbI2).
For EMII and TMAH treated device, the EMII-methanol solution (10mg/ml) or TMAH-methanol solution (10 mg mL -1 ) was dropped on QD surface, kept for 1 min and 20 s respectively, and then span for 20 s. The EMII (TMAH) ligand exchanged film was washed three times by methanol to remove the non-reacted ligand residue. The above EMII (TMAH) treated PbS QD was repeated for 8-10 times to reach the required absorber thickness. At the end of active layer, another two layers of PbS-EDT QDs were spin-coated similar to above process.

S1.2.4 Device Characterization
The current density-voltage (J-V) characteristics were measured using Keithley 2400 (J-V) digital source meter under simulated AM1.5G (100 mW cm -2 ) illumination from a 450 W Xenon lamp (Oriel, Model 9119, Newport) as the light source in air at room temperature. The light intensity was calibrated with a standard Si solar cell (Oriel, Model 91150V, Newport). The device was covered with a metal mask with an aperture area of 0.09 cm [S2] during efficiency measurement. The external quantum efficiency (EQE) measurements were taken by a home-made setup containing a Keithley 2400 Source Measure unit and Newport monochromator.

S1.2.5 Characterization of Material and Device
The absorbance of PbS CQDs was obtained using UV-visible spectroscopy (PerkinElmer instrument, Lambda 950). The mobility of the CQD films was measured using space-charge-limited current (SCLC) model [S4] with an electron-only or a hole-only device of ITO/CQDs/Al and ITO/CQDs/Au respectively. The dark J-V curve wasrecorded using a B1500A semiconductor characterization system and used to calculate the mobility of carriers by Eq. S1: where L is the thickness of the single carrier device, ε0 is the vacuum permittivity, ε is the relative dielectric constant for the PbS-CQDs, which is obtained from C-f measurements (Fig. S16), the trap-state density can be extracted by form Eq. S2 [S5] where ntrap is the trap density of the PbS-QDs layer, The VTFL and Vb is extracted from Fig. S13.
X-ray photoelectron spectroscopy (XPS) and Ultraviolet photoelectron spectroscopy (UPS) measurements were carried out in a Kratos AXIS Ultra-DLD with an Al Kα radiation source. Scanning electron microscopy (SEM) images were obtained using FEI Nova Nano SEM 450. Photoluminescence (PL) measurements were carried out using home-made system with 800 nm laser (Ti-Sapphire laser) from School of optical and electronic information. FT-IR spectra were obtained by performing FT-IR spectrometer (Bruker Vertex 70), the transmission mode FT-IR samples were taken from PbS-QDs powders mixed with potassium bromide. All the solution-state 1 H NMR experiments were carried out on a Bruker Avance III 600 instrument with scanned 128 times, all the Solid-state 1 H NMR were recorded on a Bruker Avance III 400M spectrometer with scanned 156 times. All of solid NMR samples were dewatered several hours under vacuum and heated.
The capacitance-voltage (Cp -V) measurements were acquired with an Agilent 4200A at a frequency of 10 kHz and AC signal of 50 mV, scanning from -1 to +1 V, with a step size of 50 mV. The built-in potentials (Vbi) of the devices were obtained from a Mott-Schottky plot. The depleted width of QD layer (WD) at zero bias was accorded to Eqs. S3 and S4: where NA and εQD were the PbS CQD carrier density and dielectric constant respectively, εQD was taken from c-f measurement (Fig. S16) like previous work [S1, S6], ε0 is the vacuum permittivity, ND is the ZnO carrier density from the Hall Effect results, and Vbi is the built in potential of QDs devices.
The transient photovoltage (TPV) and transient photocurrent decay (TPC) measurements were carried out using home-made system in dark condition. A ring of red light-emitting pulse diode (LED, Lumiled) controlled by a fast solid-state switch (the pulse widths were 1 ms). The transient photocurrent was measured using 40 ohm external series resistance to operate the device in short-circuit condition. Similarly, transient photovoltage was applied using 1 M ohm external series resistance to operate the device in open-circuit condition. The voltage output was recorded using an oscilloscope directly with connecting the measured solar cells. The TPC and TPV results were fitted to a mono-exponential decay function to extract the transport and recombination time [S7], respectively. The charge transport (τt) and recombination time (τr) were defined as the time interval during which the photocurrent or photovoltage decays to 1/e of their initial value immediately after excitation.
Drive-level capacitance profiling (DLCP) measurement of the devices (ITO/ZnO/CQD/Au) was performed with variant amplitude (14-140 mV) and frequency (100-100 kHz) [S3]. Capacitance with respect to amplitude was obtained to correct the capacitance value at higher orders, following Eq. S5: Solar cell device Simulations were performed using SCAPS software [S9] to ensure the energy band gap schematic and accept the guidance for device performance improvement. The complete set of simulation parameters in this simulation based on previous work [S3, S10] with slightly modification [S6], as shown in Fig. S12 and Table S2.   The w/ and w/o PbS-QDs were firstly dispersed in BTA. After completely drying the QDs solution, the treated PbS-QDs powders were re-dissolved in d6-DMSO and were filtered in glass test tube for measurement. Comparing with backgrounds NMR spectra (Fig. S3), the most notable changes were no free oleic acid peak at 11.94 ppm and significantly reduced oleate at 5.32 ppm for both w/ and w/o PbS-QDs (Fig. S4), indicating the original oleate or oleic acid was significantly removed. The peak for OH (∂ = 4.48 ppm) could be easily found in w/o HI-PbS but not clear in w/ HI-PbS, and then, the w/ QDs obtained a peak at 3.6-4.2 ppm is H signal from H2O, which further prove the existence of H2O from the deprotonation reaction between OH group and HI.

Fig. S5
Absorbance spectra of PbS-QDs films with different ratio of a 0%-HI, b 5%-HI, c 10%-HI and d 50%-HI treatment As shown in XRD spectra, it should be noted that CQDs treated by HI solution with more ratio (50%) showed broader peak and similar intensity of (111) and (200) peak, which was generally observed in smaller QDs [S11, S12]. For PL measurement, there was a part of short wavelength PL appearing in w/ HI-50% PbS-QDs solution, which was considered as an indication of decomposition in PbS-CQDs films (Fig. S6b). Therefore, it further illustrated more HI treatment would decompose the CQDs.     According to Eqs. S1 and S2, we can extract the electron mobility and bulk trap density of w/o and w/ HI PbS-QDs films, the μe are 1.9×10 -3 and 5.27×10 -3 cm 2 V -1 S -1 , the bulk trap density of both devices are Nw/o HI = 1.1×10 16 cm 3 and Nw/ HI = 4.1×10 15 cm 3 respectively.  The TMAH treated PbS-QDs were performed X-ray photoelectron spectroscopy (XPS) to investigate the species in QD films. The O 1s signal of XPS spectrum for EMII treated PbS-QDs film was referred to the previous work [S13], the Pb-OH peak was greatly suppressed and the COOpeak was the highest signal for O 1s peak in EMII treated QDs film. By comparison, the highest signal for O 1s was from Pb-OH peak in TMAH treated PbS-QDs film, indicating TMAH treated PbS-QDs film had more OH ligands than EMII treated one [S13]. According to Eq. S1, we can extract the bulk trap density of TMAH and EMII treated PbS-QDs films, the bulk trap density of both devices are NTMAH = 4.3×10 16 cm 3 and NEMII = 2.1×10 16 cm 3 , respectively.  From the XPS results, the I:Pb ratio had increased by w/ HI treatment. This change may affect the energy band alignment for device, we performed UPS measurements on w/o and w/ HI films to investigate it. As shown in Fig. S15a, the Fermi level of QDs film treated by HI shifted to shallower energy levels ascribed to higher iodine bonding content compared to w/o HI ones. The conduction band edge, Fermi level and valence band edge values for w/ HI film were -4.24, -4.54, and -5.60 eV, respectively. The energy level alignment at the PbS-QDs was plotted in Fig. S15b.

Fig. S18
The XPS spectra of three pure reference materials. O 1s XPS showing peaks at 529.1 eV from PbO, at 531 eV from O in Pb-OH and at 532.8 eV from O in carboxyl group. All of these peaks is broader caused by background, this broader peak for pure material is also found in previous work [S14]