A new method for in-situ Sb doping of CdTe that uses a modified vapor transport deposition system is described. This modification enables control of the Sb concentration with a pyrolysis stage to enhance the doping efficiency. CdTe:Sb films under different deposition conditions are characterized by SEM, XRD, and CV measurements for determining morphology, crystal structure, and hole concentration. Variations of the Sb dopant heater and pyrolyzer temperatures do not affect the CdTe morphology and crystal structure. However, CV measurements show that a higher dopant heater or pyrolyzer temperature leads to higher hole concentration. In this study, CdTe: Sb films achieve a hole concentration of 10 16 cm −3 and 10% doping efficiency when the dopant heater is 600C and the pyrolyzer temperature is 1100C. This demonstrates a path to produce high hole concentration polycrystalline CdTe film with a low concentration of dopant-induced defect.

The fabrication of CdZnTe solar cells is complicated by the removal of lattice-bound Zn during CdCl2-based anneal treatments. Fundamental understanding of the chemistry of CdZnTe with chloride treatment is essential to advance the technology. We investigate alternate and manufacturable chloride chemistry using SbCl3 for CdZnTe-based alloy. New SbCl3 treatments of CdZnTe show promising performance while maintaining Zn composition compared to CdCl2. Effects of varying reaction conditions on chemical mechanisms are presented.

The effect of high-temperature annealing (HTA) treatments and Cd-excess during the in-situ doping of CdSeTe:Sb is investigated. Optimized treatments eliminate the wurtzite phase from the as-deposited CdSeTe layer, while facilitating Se intermixing and grain size enhancement before CdC12 treatment. The improved device stack quality results in a VOC improvement of 250 mV. VOC is further improved by tuning the Cd/Sb flux ratio during CdSeTe:Sb deposition. The lowest defect concentration is achieved at Cd/Sb of 1.4:1, which produced the best VOC CdSeTe:Sb cell with VOC = 849mV, despite a decreased carrier concentration due to the harsh CdC12 treatment.

Development of a new SbCl3-based solution treatment for chloride activation of CdTe-based solar cells is described. Activation was confirmed with CdSeTe:Sb devices exhibiting V-OC > 550 mV and J(SC) > 25 mAcm(-2). Treatment optimization showed strong effects of annealing time and temperature on cell properties. Material characterization indicated surface conversion of CdSeTe to CdCl2 and Sb2Te3 with SbCl3 treatment. However, no reaction products were formed with treatment under inert conditions, which was coupled with poor device performance. Drying the SbCl3 films at controlled humidity conditions confirmed that exposure to high relative humidity is critical for CdSeTe activation, through hydrolysis of SbCl3. Mechanistic details of SbCl3 speciation during hydrolysis and annealing to activate CdSeTe are discussed.

The effect of Sb inclusion in CdSeTe thin films was examined. CdSeTe alloys were grown by co-evaporation, either on 5 nm thick electron-beam evaporated Sb layers (Sb/CdSeTe) or coated with Sb post-deposition (CdSeTe/Sb). Sb inclusion improved CdSeTe grain structure, a critical factor for device performance, with average grain size increasing from 0.6 to 1.5 mu m. All samples crystallized on annealing to the cubic structure, with a change in film orientation from (111) to random or (220). The presence of Sb in CdSeTe/Sb or Sb/CdSeTe samples, however, had a limited effect on the carrier concentration compared to Sb-free samples. Time resolved photoluminescence showed Sb inclusion increased decay lifetimes, with tau(1) = 0.8 ns for and tau(2) = 6.9 ns for the Sb/CdSeTe sample.

Cadmium telluride (CdTe)-based cells have emerged as the leading commercialized thin film photovoltaic technology and has intrinsically better temperature co-efficients, energy yield, and degradation rates than Si technologies. More than 30 GW peak (GWp) of CdTe-based modules are installed worldwide, multiple com-panies are in production, modules are shipping at up to 18.6% efficiency, and lab cell efficiency is above 22%. We review developments in the science and technology that have occurred over approximately the past decade. These achievements were enabled by manufacturing innovations and scaling module production, as well as maximizing photocurrent through window layer optimization and alloyed CdSexTe1-x (CST) absorbers. Improved chlorine passivation processes, film microstructure, and serendipitous Se defect passivation significantly increased minority carrier lifetime. Efficiencies >22% have been realized for both Cu and As doped CST-based cells. The path to further efficiency gains hinges primarily on increasing open circuit voltage (Voc) and fill factor (FF) through innovations in materials, fabrication methods, and device stacks. Replacing the longstanding Cu doping with As doping is resulting in better module stability and is being translated to large-scale production. To realize 25% efficiency and >1 V Voc, research and development is needed to increase the minority carrier lifetime beyond 100 ns, reduce grain boundary and interface recombination, and tailor band diagrams at the front and back interfaces. Many of these goals have been realized separately however combining them together using scalable manufacturing approaches has been elusive to date. We review these achievements and outstanding opportunities for this remarkable photovoltaic technology.

a hole concentration of 1016 cm-3 and 10% doping efficiency when the dopant heater is 600C and the pyrolyzer temperature is 1100C. This demonstrates a path to produce high hole concentration polycrystalline CdTe film with a low concentration of dopant-induced defect.

Single‐crystalline Cu2ZnSnSe4 (CZTSe) solar cells with open circuit voltages reaching 500 mV are achieved through a combination of composition control and a low‐temperature thermal ordering treatment. A comparison of the device results for Cu‐poor CZTSe with Cu/Zn + Sn ratios of 0.77 and 0.86 is presented with and without the implementation of a 130°C absorber annealing treatment. An increase in bandgap energy is observed via external quantum efficiency measurements with both the decrease in Cu content and the implementation of the order anneal, the latter of which also leads to a decrease in Urbach energy. Defect characterization performed with admittance spectroscopy on devices is demonstrated as insufficient because of low‐temperature current barriers. Photoluminescence (PL) on crystal surfaces however enables a qualitative comparison of the defect landscape between crystal compositions and annealing treatments. Both a highly compensated and a lightly doped defect model are used to fit PL as a function of laser fluence to identify defects contributing to each observed recombination channel. The PL signatures attributed to the ZnSn defect become unresolvable with a decrease in Cu/Zn + Sn ratio from 0.86 to 0.77. Furthermore, a decrease in Cu–Zn disorder is observed upon the implementation of the annealing treatment through both a comparison of potential fluctuation depths and of both PL models.

The effect of the electron transport layer (ETL) on the growth of methylammonium lead iodide (MAPbI3) thin films by two-step close-space vapor transport (CSVT) is reported. Nanocrystalline CdS, as well as amorphous SnO2 and C60, were selected as ETLs on indium tin oxide-coated glass substrates prior to two-step CSVT. The ETL affected the PbI(2 )growth, leading to different morphological and crystallographic properties. These differences carried over through the methylammonium iodide reaction to the MAPbI(3) phase, but compact films with a reasonable morphology could be made on each ETL. The ETL also affected the PbI2-to-MAPbI(3) reaction rate. Solar cells processed on each ETL showed a low level of residual PbI(2)was important for good photovoltaic conversion efficiency (PCE). The PCEs were similar on average, but trade-offs in J-V parameters depended on the ETL selection. When films on each ETL were reacted beyond an optimal PbI(2)residual content, solar cells had lower performance driven by decreases in different J-V parameters. The ETL also had practical effects on J-V performance, namely, hysteresis and air stability. The hysteresis of solar cells on C(60 )was much less than on SnO2 and CdS. However, the solar cells with C(60 )ETLs were not stable in air, exhibiting FF and J(sc) losses in as little as 15 min of air exposure, while those made on the other ETLs were stable for hours. Thus, the choice of ETL for two-step CSVT affects the growth of PbI(2 )and its reaction to MAPbI3, but interfacial chemistry considerations and effects on current and atmospheric stability appear to be more for device and

Understanding the nature of recombination and its dependence on defects and interfaces is essential for engineering materials and contacts for a higher open-circuit voltage (Voc) and power conversion efficiency in photovoltaic (PV) devices. Time-resolved photoluminescence (TRPL) has conventionally been used to evaluate recombination, but carrier redistribution often dominates the response at short times. Here, we report on the quantification of carrier dynamics and recombination mechanisms by complementary use of both time-resolved terahertz spectroscopy and TRPL combined with numerical modeling of the continuity equations and Poisson's equation. We have demonstrated this approach using CdTe thin films. A thin-film stack with CdTe fabricated by vapor transport deposition and treated with CdCl2 exhibited a bulk lifetime of 1.7 ± 0.1 ns, a negligible CdTe/CdS interface recombination velocity, and a back surface recombination velocity of 6.3 ± 1.3 × 104 cm/s. In contrast, a film stack without CdCl2 treatment had a bulk lifetime of only 68 ± 12 ps and a higher interface recombination velocity of 4 ± 2 × 108 cm/s. By determining the locus and mechanisms of performance-limiting recombination, we can accelerate the development of thin-film PVs with higher Voc and efficiency. While the method has been demonstrated here using CdTe, it is also applicable to perovskites, Cu(InGa)Se2, Cu2ZnSn(S,Se)4, and emerging technologies.

Understanding the nature of recombination mechanisms is essential for higher power conversion efficiency in photovoltaic (PV) devices. Here we use a combination of time-resolved terahertz spectroscopy and numerical modeling to determine the bulk Shockley-Read-Hall lifetime and interface and back surface recombination velocities in CdTe thin film stacks. The measurement was facilitated by fabricating wire-grid device structures using conventional laser scribing. Evaluation of a glass/FTO/SnO 2 /CdS/CdTe stack treated with CdCl 2 allowed separation of the CdTe absorber bulk lifetime, 1.6 ns, from the back surface recombination velocity, ~6x10 4 cm/s, and indicated that CdTe/CdS interface recombination velocity had no significant impact on carrier dynamics.

Methylammonium lead iodide perovskite solar cells were made using an all-vapor process, including two-step close space vapor transport-processed MAPI and evaporated copper phthalocyanine (CuPC). Solar cells fabricated entirely in a nitrogen glovebox had poor performance due to s-shaped J-V curves and fill factors (FF) <45%. Solar cells exposed to dry air in a desiccator for 7 days, or where O2 was flowed into the evaporator during CuPC deposition, had significantly improved performance with reduced or eliminated s-shaped behavior and improved FF up to 71%. Parallel contact conductivity measurements show that exposure to dry air, deposition with oxygen, and MoOx capping layers all increase the conductivity of the CuPC. SCAPS simulations show that increasing hole concentration consistent with oxygen doping effects can explain the J-V behavior of the solar cells.

Understanding the relationship of doping density, carrier lifetime, and interface recombination to device performance is critical to designing solar cells with high power conversion efficiency (PCE). In turn, it is necessary to understand how bulk material composition determines doping density and carrier lifetime. The most efficient kesterite Cu2ZnSn(S,Se)(4) (CZTSSe) thin film solar cells have had Cu-poor, Zn-rich compositions, while more stoichiometric compositions have lower PCEs. However, thin films are grown under highly non-equilibrium conditions, complicating fundamental studies. Here we report on a set of CZTSe monocrystals with varied cation stoichiometry, enabling correlation of bulk composition to material and device properties without the complication of grain boundaries or secondary phases. Copper-poor, zinc-rich compositions (Cu/(Zn + Sn) =0.77-0.90 and Zn/Sn = 1.17-1.25) yield bulk carrier lifetimes longer than 200 ps and PCE >5%. In contrast, near-stoichiometric compositions, with Cu/(Zn + Sn) > 0.90 and Zn/Sn < 1.15, have carrier lifetimes shorter than 20 ps and PCE <2%. CZTSe/CdS interface recombination velocity has a similar value to the CZTSe surface recombination velocity, with values of 10(4)-10(5) cm/s determined by time-resolved terahertz spectroscopy and transport-recombination modeling. Device modeling reveals the dependence of open circuit voltage (V-OC) on doping density, carrier lifetime and interface recombination. For a crystal with low doping density of 10(15) cm(-3), the maximum V-OC is limited by the bulk lifetime. Higher V-OC can be attained with higher doping density, but interface recombination becomes more significant with increased lifetime and doping density. These simulations indicate limitations and potential pathways to high performance.

Admittance spectroscopy has become a commonly used device-level technique to probe the defect structure of kesterite materials. While this technique holds promise, phenomena such as current barriers and metastabilities cause difficulty in the interpretation of results. In this work, devices fabricated on single-crystalline CZTSe absorber layers are used to explore these effects in a more idealized system than thin films. Differences in behavior between admittance signatures of absorbers with Cu/Zn + Sn ratios of 0.86 and 0.77 are observed and mainly attributed to the formation of a low-temperature current barrier at the CZTSe/carbon-based back-contact interface in the more Cu-depleted devices. The effect of this barrier is significantly reduced by the implementation of higher-work function Au or MoO3−x back-contact material. Furthermore, distinct metastable doping densities are observed in each device type with the implementation of 3-h external bias conditioning of white, blue, or red light as well as 1 V applied forward bias. The origin of this behavior is attributed to a VSe–VCu defect complex that is predicted to exist in related chalcogenide materials.

This is the first demonstration of an all-vapor close space vapor transport process to deposit methylammonium lead iodide perovskites.

Close space vapor transport is used to investigate PbI2 deposition and the reaction of PbI2 with CH3NH3I to form CH3NH3PbI3 (MAPI) on nanocrystalline (CdS) or amorphous (SnO2) electron transport layers (ETL). The crystallographic structure and surface morphology of PbI2 thin films and the resultant MAPI films after reaction have strong dependence on ETL substrates. Smoother morphology, stronger crystallographic orientation (001) for PbI2 and (200) for MAPI films - and larger coherency length are observed for films grown on SnO2 ETL as compared to CdS. Solar cells fabricated on both types of ETL exhibit better cell performance for the MAPI absorbers that contain a small amount of residual PbI2. Despite crystallographic differences, solar cells made on both ETLs performed similarly.

Cu2ZnSnSe4 single‐crystal solar cells with open‐circuit voltages reaching 450 mV are demonstrated. The key differences in performance between high‐ and low‐voltage cells are analyzed and compared with state‐of‐the‐art thin film devices. Copper‐poor absorbers of two different compositions were evaluated as a function of surface treatment. Temperature‐dependent JV measurements were used to assess the efficacy of interface passivation. Crystals with lower copper content are shown to require more aggressive chemical treatments to achieve the maximal benefits than does their higher‐Cu counterparts. Hole concentration is confirmed via Hall characterization, with smaller densities corresponding to lower Cu concentration. In conjunction with poor lifetimes, these carrier densities are shown to limit collection at long wavelength, which reduces current in single‐crystal devices. The decrease in Cu concentration is also shown to increase the bandgap from 0.98 to 1.03 eV while maintaining the same level of subbandgap absorption. Postfabrication device annealing was shown to benefit devices with higher Cu/Zn + Sn ratios but hinder devices further deviating from stoichiometry, which is attributed to pn‐junction degradation.

Transparent conductive oxides are used in many technologies and it is important to understand their interfacial chemical reactions. Here, we use recently developed thermomechanical cleaving and X-ray photoelectron spectroscopy to probe oxidation states at the SnO 2 interface of CdTe solar cells. We show that tin oxide promotes the formation of nanometer-scale oxides of tellurium and sulfur, largely during CdCl 2 /O 2 activation. Surprisingly, in copper-doped devices, relatively low temperature anneals (180-260 °C) to diffuse and activate copper acceptors also cause significant oxidation changes at the front interface, providing a heretofore missing aspect of how back contact processes can modify device transport, recombination, and performance. For Group V-doped devices, this same oxidation process causes segregation of the dopants to the SnO 2 interface in their oxidized states, implying that adjacent regions in the absorber have been depleted of dopants. Intriguingly, we demonstrate that because of their layered, van der Walls-bonded crystal structure, spatially segregated Group V oxides may represent a mechanically weak layer in a finished device.