AVS2012 Session EN+AS-ThA: Characterization of Energy Materials and Systems

Capacitance-voltage (CV) measurements are commonly used to characterize semiconductor junctions. A common observation when performing such measurements on polycrystalline CdTe solar cells is that the measured capacitance is a strong function of the voltage scan direction. These results in a noticeable hysteresis in the C-V profile when capacitance data is collected using both forward (fwd) and reverse (rev) voltage scan directions. Similarly, hysteresis curves for the usual derived quantities such as net carrier density, N_a, and depletion width, W, naturally follows. We have recently observed that in particular, the hysteresis in calculated carrier acceptor density, N_a,hys, arbitrarily defined as N_a,fwd – N_a,rev decreases as the CdTe growth temperature is reduced. At higher CdTe growth temperatures, this value is positive and shifts to negative values at lower growth temperatures. This behavior is believed to reflect a transition of the CdTe stoichiometry in CdS/CdTe solar cells from Cd-poor at higher temperatures to Cd-rich at lower temperatures based upon recently published CdTe binary phase diagrams.

The impact of Cd/Te stoichiometry on cell performance is suspected. Cd-poor stoichiometry favors the formation of Cd-vacancies (V_Cd), a beneficial acceptor, but it also simultaneously increases the formation of the Te antisite (Te_Cd), an important recombination center in CdTe. Recent theoretical calculations using the hybrid Heyd-Scuseria-Ernzerhof (HSE) functional also suggests that Te interstitials (Te_i) may be an important recombination center under Cd-poor conditions. Increased Cd chemical potential will reduce the formation of these recombination centers, but also reduce hole carrier density. Thus, an optimal growth condition, which could include extrinsic p-type doping, may be needed for leading to higher performance CdTe solar cells.

Not discussed to date however is whether stoichiometry has an inherent impact on the stability of CdTe solar cells. In this talk, we contrast accelerated lifetime study results for CdTe cells grown at different growth temperatures. Open-circuit voltage (V_oc) stability in cells grown at lower temperatures was noticeably improved. N_a,hys in the latter cells was considerably smaller. Both V_oc and FF were well correlated with this capacitance-derived parameter. N_a,hys was observed to be relatively unchanged in devices where V_oc did not degrade. Finally, time-resolved photoluminescence lifetimes of nearly 10 ns were measured in these cells made at lower temperatures relative to values of around 2 ns measured in higher temperature devices.

The use of organic materials for electronic applications such as Organic Photovoltaics (OPV’s) and Organic Light Emitting Diodes (OLED’s) is rapidly growing. The efficiencies of these devices are widely recognized to depend on the molecular gradients fabricated into the devices. Conversely, the degradation of these devices is suspected to depend on several factors including chemical degradation and chemical migration as a function of use. It is therefore very desirable to develop analytical techniques which can quantify chemical gradients as well as identify degradation products in these films.

Gas Cluster Ion Beam (GCIB) sources with low energy per atom Ar_2,000⁺ ions have recently been demonstrated to provide a depth profiling technique for molecular species (1,2). GCIB depth profiling in an interleaved mode with the surface analysis spectroscopy of X-ray Photoelectron Spectroscopy (XPS) has been applied to model OPV and OLED devices. The 2 to 5 nm information depth of XPS, combined with the demonstrated “non-destructive” chemical information revealed after each GICIB sputter interval, facilitates the chemical gradient analysis of a series of model samples up to several 100 nm in depth with molecular depth resolution < 10 nm.

Selected model OPV and OLED samples were exposed to annealing and environmental degradation testing. The XPS depth profiles measured the migration of organic components and dopants as a function of fabrication processes. This presentation will provide an overview of GICB depth profiling with XPS as well as discuss the insights into efficiencies and degradation processes elucidated by this chemical gradient analysis technique.

1. T. Miyayama, et al. Surf. Interface. Anal. 42 (2010) 1453-1457.

2. T. Miyayama, et al. J. Vac. Sci. Technology A, 28 (2) (2010) L1-L4.

Large area XPS was used for analysis of fresh, conditioned and aged cathode catalyst layers. Spectroscopic analysis, which provides an integral spectrum from approximately 1mm² area, may have a contribution from the GDL sublayer that was not fully removed from the catalyst layer side during separation of the MEA components. The GDL sublayer exhibits a peak in the same BE range as fully fluorinated carbons that are detected in the catalyst layer. Changes that have been detected in CL may not be due to the changes within the ionomer, but rather due to physical intermixing of layers caused by the testing protocols. XPS imaging enabled separation of the differing component contributions. Using Pt 4f imaging, regions are clearly identified where no Pt is present, indicating that GDL pieces adhere to the CL. Fluorine images at two different binding energies (one for the ionomer, and another for the GDL) confirm this. High resolution C 1s spectrum extracted from the area where catalyst is present does not show a high BE component in the C 1s spectrum of the area where GDL is present, confirming that the high BE component detected by large area spectroscopy are contributions from the GDL. High-resolution spectra acquired from the area where only catalysts layer is present shows higher amounts of oxidized forms of carbons. In addition, morphological changes of aged cathode catalyst layers have been evaluated by Digital Image Processing of SEM images for roughness, porosity and texture parameters.

The energy sector is a large, growing field in research and technology. A growing subset of the solar field is organic photovoltaics (OPV). Two areas that have long challenged researchers are efficiency and stability of the OPV devices. An area that is in need of fundamental understanding that may increase efficiency of these devices is the charge transfer. However, a precise nature of the bulk-heterojunction OPV donor/acceptor interface is difficult to pinpoint. Thus, model systems can be used initially to mimic these interactions. In the system presented here, the donor/acceptor interface is well-defined. Phenyl-C61-butyric acid methyl ester (PCBM) was reacted with an amino-terminated organic monolayer on a single crystalline Si(111) surface. Poly 3-hexylthiophene (P3HT) was then deposited onto this complex substrate, and the produced interface can be observed and investigated. Fourier transform infrared spectroscopy was used to identify the functional groups on the surface. The chemical and electronic states of the coating and the substrate were investigated by X-ray photoelectron spectroscopy, and the morphology was determined through atomic force microscopy. Preliminary charge carrier lifetime measurements will also be reported.