A Comparative Study of the Effects of Nontoxic Chloride Treatments on CdTe Solar Cell Microstructure and Stoichiometry

Abstract

either by vapor deposition [ 5,8,10,11 ] or by drop-casting of a CdCl 2 /methanol solution (also used for CdTe nanoparticle solar cells), [ 7,[12][13][14] followed by an air-anneal at temperatures of ≈370-450 °C. While CdTe (as well as CdS) is a highly stable compound and is insoluble in water, the use of water-soluble CdCl 2 in the industrial production of CdTe modules may pose an environmental risk. Recently however, a breakthrough study demonstrated the direct replacement of the toxic (and costly) CdCl 2 with the nontoxic and cheap alternative, MgCl 2 , with no detriment to device performance. [ 15 ] Subsequently, both ZnCl 2[ 16 ] and NH 4 Cl [ 17 ] have also been shown to be potentially suitable alternatives. Interestingly, the use of other salts including NaCl, MnCl 2 , KCl, and HCl was signifi cantly less successful in terms of producing high effi ciencies. [ 15 ] Importantly, these recent studies have reopened a critical parameter space within the fabrication process. In order to accelerate process development, fundamental material studies are now required to investigate the relative effectiveness of different chloride compounds in inducing the desired materials changes in a CdTe thin-fi lm device stack. Moreover, by correlating such studies with the respective device performances, further improvements are anticipated. Reports of desirable materials changes associated with CdCl 2 activation include: Recrystallization and grain growth of CdTe [18][19][20] and CdS [ 5,21 ] fi lms (for minimization of stress and to reduce the density of deleterious defects), p-type doping of CdTe, [ 2,6,8,15 ] S/Te interdiffusion, [ 22,23 ] and grain boundary passivation. [ 5,24 ] However, such reports are often contradictory in terms of identifying which of these is most beneficial in improving effi ciency. [ 25 ] In this work, CdTe solar cell stacks (i.e., CdTe/CdS/ZnO/ SnO 2 /glass) treated with CdCl 2 , MgCl 2 , NaCl, and NH 4 Cl were subjected to extensive X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), and scanning electron microscope (SEM) studies, with the critical aim being to distinguish the reasons as to why CdCl 2 and MgCl 2 treatments can yield higher effi ciencies than NH 4 Cl and, in particular, NaCl. The device performances, reported elsewhere, [ 15,17 ] for each of these Cl treatments, as well as those following MnCl 2 -and KCl treatments, This work presents the fi rst systematic comparison of the effects of a range of chlorides (CdCl 2 , MgCl 2 , NaCl, and NH 4 Cl) on the microstructure and chemical composition of CdTe/CdS/ZnO/SnO 2 solar cells, providing valuable insight to the ubiquitous Cl-activation process. Using X-ray diffraction, it is shown that CdCl 2 induces the greatest extent of recrystallization (standard deviation of texture coeffi cients, σ , reduces from 0.93 for as-grown CdTe to 0.43) and minimizing stress (from 178 MPa for as-grown material to zero). MgCl 2 treatment also yields signifi cant randomization of the CdTe texture ( σ = 0.55) but NaCl treatment does not ( σ = 1....

Building Solar Cells from Nanocrystal Inks

“…From the EQE spectrum, one can see that the device showed a good spectrum response in the short and long wavelength regions, which is in accordance with the previous report [41]. Due to the toxicity of the Cd element, using safer salts (such as NaCl, KCl or NH4Cl) instead of CdCl2 during the processing steps is believed to be important for decreasing the Cd contamination during the NC solar cells fabrication [8,[47][48][49]. On the other hand, finding a new Cd-free n-type partner is also preferred for CdTe NC solar cells.…”

“…The acid molecules will move quickly along the crystal boundary and damage the whole NC thin film, leading to device shunt or thin film peeling off the substrate, which has been confirmed in previous work [33]. For a device with normal structure, by using chemical wet etching, the energy obstacle between ITO and CdTe NC film is eliminated and leads to the highest efficiency CdTe solution processed solar cell with a PCE of 12.7% [8,[47][48][49]. For devices with inverted structure, the use of dielectric materials as a hole transport layer (HTL) before the deposition of back contact is also preferred to obtain low ohmic contact to CdTe NC.…”

In-depth analysis of chloride treatments for thin-film CdTe solar cells

“…In contrast, both the NaCl and KCl treatments give only a minor increase in the chlorine concentration, and even then this is confined to the CdTe nearest to the diffusion surface—deeper in the CdTe, the chloride level does not rise above its background value. This is strongly indicative that the poor performance of these treatments results from an inability of NaCl and KCl to effectively introduce chlorine into the CdTe, in contrast to the case for CdCl 2 and MgCl 2 treatments, possibly due to the higher dissociation energies of the compounds12. In essence, NaCl and KCl treatments have improved the cell performance over as-grown devices in spite of little chlorine being incorporated.…”

“…Utilizing the same treatment ranges (410–450 °C, 20–60 min) gives similar increases in J SC and fill factor but again the V OC is never increased from the as-grown level. There are, however, a number of differences between the ineffective treatments and air annealing: X-ray diffraction analysis shows NaCl treatment causes an increase in <111> preferred orientation, while other treatments generate a decrease12; MnCl 2 and KCl give efficiency levels well below those achievable with air annealing with a pronounced decrease in V OC compared with as-grown devices (Supplementary Table 1). However, as the ineffective treatments fail to incorporate significant amounts of chlorine into the CdTe we may draw some parallels with air annealing.…”

Tuning the Lattice Parameter of InxZnyP for Highly Luminescent Lattice-Matched Core/Shell Quantum Dots

“…Note that the Nelson-Riley method minimises systematic error contributions to the uncertainty on the quoted lattice parameters, as reported in reference. [32][33][34] Electrophoresis measurements: The charge of the synthesized In x Zn y P QDs was determined through measurement of the electrophoretic mobility. 750 µL of a diluted QDs solution in toluene was loaded into the Malvern PCS1115 non-aqueous dip cell and measured on Malvern Zetasizer Z Nano operating at 25°C.…”