Examples from Current Research

Direct Observation of an Inhomogeneous Chlorine Distribution in CH3NH3PbI3-xClx Layers: Surface Depletion and Interface Enrichment

We have used hard X-ray photoelectron spectroscopy (HAXPES) at different photon energies and fluorescence yield X-ray absorption spectroscopy (FY-XAS) to non-destructively investigate CH3NH3PbI3-xClx perovskite thin films on compact TiO2. This combination of spectroscopic techniques allows the variation of information depth from the perovskite layer surface to the top-most part of the underlying compact TiO2 layer. We have taken advantage of this to understand the distribution of chlorine throughout the perovskite/TiO2 layer stack. No Cl is detected using HAXPES, indicating surface depletion of Cl and allowing us to place an upper limit on the amount of Cl in the perovskite layer: x < 0.07 and x < 0.40 to depths of ~10 nm and ~26 nm, respectively, beneath the perovskite film surface. Our FY-XAS results, however, demonstrate that there is a higher average concentration of Cl throughout the perovskite layer than at the surface (x > 0.40) consistent with both enhanced concentrations of Cl deep beneath the perovskite film surface and near the CH3NH3PbI3-xClx perovskite/TiO2 interface. The consequences of this distribution of Cl in the CH3NH3PbI3-xClx perovskite layer on device performance are discussed.

Cl 2p and I 4s spectra taken with a photon energies of 2 keV (black) and 6 keV (red) of a 60 nm thick CH3NH3PbI3-xClx film on compact TiO2. Note the absence of any detectable Cl in both spectra.

Cl K edge FY-XAS spectra of a 60 nm thick CH3NH3PbI3-xClx layer on compact TiO2 (blue) and the compact TiO2 layer without a CH3NH3PbI3-xClx layer on top (green). Note the distinctively different shapes of the two spectra.

For more details see:

D.E. Starr, G. Sadoughi, E. Handick, R.G. Wilks, J.H. Alsmeier, L. Köhler, M. Gorgoi, H.J. Snaith, and M. Bär,  Energy Environ. Sci. 8, 1609 (2015)

The impact of Cu2-xS surface phases and their impact on the electronic structure of CuInS2

Combining surface and near-surface bulk sensitive x-ray spectroscopies to study CuInS2 (CIS) thin-film solar cells absorbers, we could identify spatially separated areas with different surface properties. The dark and light areas (distinguishable by eye before KCN etching) are dominated by CuS and Cu2S like secondary surface phases. After the removal of the Cu2−xS phases by KCN etching the CIS samples appear similar to the eye but the surface composition still signifficantly differs; in particular in the degree of Cu-deficiency. We also see a change in the surface band gap (see publication for details). Such (hidden) inhomogenieties might be an explanation why respective solar cells fall behind compared to their low-gap counterparts w.r.t. device performance.

XPS survey spectra of initially dark and light CIS material after KCN etching. Inset: Respective XPS spectra of the valence band region compared to that of an In2S3 reference. We observe a significant variation of the Cu 3p line intensity, indicating a change in the In/Cu surface ratio of the two CIS samples. By fitting the respective peaks, we are able to reveal that while a Cu deficiency is found at the surface of both samples, it is significantly more pronounced for the (initially) light CIS sample.

For more details see:

M.Bär, J. Klaer, L. Weinhardt , R.G. Wilks, S. Krause, M. Blum, W. Yang, C. Heske, and H.-W. Schock Adv. Energy Mater. 2013, 3, 777–781

Si-O formation at the Si/ZnO thin-film solar cell interface

State-of-the-art thin-film silicon solar cells often feature ZnO:Al as a transparent window layer. Hard x-ray photoelectron spectroscopy (HAXPES) was used to investigate the resulting buried Si/ZnO interface. It was found that the solid phase crystallization (SPC) of the (initially) amorphous silicon leads to an (additional) formation of Si-O bonds at the Si/ZnO interface. Furthermore, the ZnO could be identified as the oxygen source for this process. Altogether, this may shed new light on why solar cells based on Si/ZnO layer stacks fall behind w.r.t. the expected device efficiencies.

Si 1sSi–Ox/Si 1sSi–Si intensity ratio of the Si/ZnO layer stack compared to that of a reference system in which Si–OX bonds are exclusively present at the surface (an oxidized Si wafer).for ex-situ (a) and in-situ (b) SPC-treated samples. The increase of this intensity ratio with increasing excitation energy (i.e., inelastic mean free path) does not agree with the behavior predicted by the surface oxidation model (dotted grey lines) and thus can be interpreted as a strong indication of an oxidation at the Si/ZnO interface.

For more details see:

M.Wimmer, M.Bär, D.Gerlach, R.G.Wilks, S.Scherf, C.Lupulescu, F.Ruske, R.Felix, J.Hüpkes, G.Gavrila, M.Gorgoi, K.Lips, W.Eberhardt, and B.Rech, Hard x-ray photoelectron spectroscopy study of the buried Si/ZnO thin-film solar cell interface: Direct evidence for the formation of Si–O at the expense of Zn-O bonds, Appl. Phys. Lett. 99, 152104 (2011).

Band alignment at the CdS/Cu2ZnSnS4 thin-film solar cell heterojunction

Earth-abundant Cu2ZnSnS4 (“CZTS”) kesterites are promising absorbers for thin-film solar cells. A combination of direct and inverse photoelectron spectroscopy was used to study the band alignment at the CdS/CZTS interface. A “cliff”-like conduction band offset was revealed and a significant etch-induced enhancement of the energetic barrier for charge carrier recombination across the CdS/CZTS interface was found.

Schematic presentation of the band positions at the kesterite and CdS/kesterite sample surfaces and resulting band alignment at the interface. The red schemes and numbers represent the situation for the KCN-etched CZTS absorber. The experimental uncertainty for the VBM and CBM is ± 0.10 eV, and that for the Egsurf and interface band gap Eg,i values is 0.15 eV.

For more details see:

M. Bär, B.-A. Schubert,, B. Marsen, R. G. Wilks, S. Pookpanratana, M. Blum, S. Krause, T. Unold, W. Yang, L. Weinhard, C. Heske, and H.-W. Schock, Cliff-like conduction band offset and KCN-induced recombination barrier enhancement at the CdS/Cu2ZnSnS4 thin-film solar cell heterojunction, Appl. Phys. Lett. 99, 222105 (2011).

Depth-resolved band gap in Cu(In,Ga)(S,Se)2 thin films

The surface composition of Cu(In,Ga)(S,Se)2 (“CIGSSe”) thin films intrinsically deviates from the corresponding bulk composition, which also modifies the electronic structure and thus the optical properties. We used a combination of photon and electron spectroscopies with different information depths to gain depth-resolved information on the band gap energy (Eg) in CIG(S)Se thin films. We find an increasing Eg with decreasing information depth, indicating the formation of a surface region with significantly higher Eg. This Eg-widened surface region seems to extend further into the bulk of the sulfur-free CIGSe thin film compared to the CIGSSe thin film.

Schematic presentation of the determined CIGSe (top) and CIGSSe (bottom) band gap energies using the three different spectroscopic approaches (optical reflection, left; XES and XAS, center; UPS and IPES, right). The respective information depths of the applied spectroscopies are also shown. The gray area represents the measurement error. CBM (VBM) denotes the conduction band minimum (valence band maximum).

For more details see:

M. Bär, S. Nishiwaki, L. Weinhardt, S. Pookpanratana, O. Fuchs, M. Blum, W. Yang, J.D. Denlinger, W.N. Shafarman, and C. Heske, Depth-resolved band gap in Cu(In,Ga)(S,Se)2 thin films, Appl. Phys. Lett. 93, 244103 (2008).