Surface potential of chalcopyrite films measured by KPFM

Abstract

Atomic force microscopy is widely used to characterize the surface topography of a variety of samples. Kelvin probe force microscopy (KPFM) additionally allows determining images of the surface potential with nanometer resolution. The KPFM technique will be introduced and studies on surfaces of chalcopyrite semiconductors for solar cell absorbers will be presented. It is shown that operation in ultra-high vacuum (UHV) is required to obtain meaningful work function values. Different methods for obtaining UHV-clean surfaces are presented and KPFM studies on these are compared. Surfaces where prepared by in-vacuum deposition, inert-gas transfer, in-vacuum decapping of a protective Se-cap and a peel-off method. Finally, a sputter-annealing cycle also allows to obtain well-suited surfaces for KPFM studies. Employing KPFM, variations in the local surface potential at grain boundaries of polycrystalline CuGaSe2 films were observed. A potential drop indicates the presence of charged defects at grain boundaries. Furthermore, different electronic activity was found for different grain boundaries, as concluded from studies under illumination. Using laterally resolved surface photovoltage, a Cu2−xSe impurity phase could be observed in CuGaSe2.Copyright line will be provided by the publisher 1 Introduction Photovoltaic energy conversion is one route followed in the quest for a sustainable energy supply in the future. Thin film solar cell technologies offer a high cost reduction potential compared to standard silicon wafer technology. With their high power conversion efficiencies on the laboratory scale, thin film solar cells from the Cu-chalcopyrite materials system are a promising candidate for this technology; in fact, pilot production lines have recently been started. Currently, the highest efficiency of nearly 20% is obtained for Cu(In, Ga)Se 2 containing ≈ 30% of Ga [1]. A typical solar cell consists of a metallic Mo layer on a float glass substrate, a polycrystalline p-type Cu-chalcopyrite absorber layer with a typical thickness of 2 µm, a thin buffer layer (typically CdS) and the n-type double layer window (ZnO). A NiAl grid provides the front contact. A wide variety of growth techniques for the chalcopyrite absorber have been studied, including evaporation, sputtering and subsequent rapid thermal processing in chalcogen atmosphere, chemical vapor deposition, metal organic vapor phase epitaxy and electrochemical deposition. In such multilayered structures the various interfaces play a crucial role, as they are prone to contain electronically active defects causing electronic losses in the device. The polycrystalline character of the materials introduces additional effects, as for example varying composition in different grains and grain boundaries [2]. Recently, also compositional variations on the nanometer scale have been proposed [3]. Therefore, studies of surfaces and interfaces are highly important in order to further the understanding of these materials and eventually increase their efficiency towards the theore...

Electronic properties of grain boundaries in Cu(In,Ga)Se2 thin films with various Ga-contents

“…In the following we present the results, comparatively to as grown samples. This value is in good agreement with the results in Figure 1(b) as well as with literature [19].…”

Junction formation of Cu3BiS3 investigated by Kelvin probe force microscopy and surface photovoltage measurements

“…The work function of the surfaces was obtained from the measured CPD by calibration of the tip against a sample of highly oriented pyrolytic graphite with a known work function. The surface photovoltage was determined by illuminating the samples with red laser light (λ = 675 nm, 70 mW/cm 2 ) and subtracting the work function in the dark [ 26 ]. Also for KPFM, samples were NH 3 etched and transferred through air into the UHV system within less than 5 min.…”

Microscopic characterization of individual grain boundaries in Cu-III–VI2 chalcopyrites

“…In a previous publication [24] we have studied the development of the work function with different surface cleaning steps, consisting in subsequent annealing and sputtering cycles. It was shown that after an initial annealing and sputter cycle the work function increased substantially, whereas upon subsequent cycles it remains rather unchanged, indicating a fairly reproducible surface condition; this conclusion was supported by a low SPV [24]. In Fig.…”