L. Stolt

Uppsala University, Uppsala, Uppsala, Sweden

Are you L. Stolt?

Claim your profile

Publications (83)175.14 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: Light-induced metastabilities of Cu(In,Ga)Se2 (CIGS)-based thin film solar cells have been extensively researched for many years. It is commonly observed that junction capacitance and p-doping level in CIGS absorbers increase considerably after light soaking (LS). In this work, we focus on the LS behaviors of cells with different minority carrier lifetimes (τn). Experiments show that high efficiency cells with long τn lose open circuit voltage (Voc) and fill factor (FF) upon LS, whereas low efficiency cells with short τn lose less or even gain Voc and FF. The sodium content measured with glow discharge optical emission spectroscopy (GD-OES) increases in the region close to the CdS/CIGS interface with LS and may contribute to the observed LS behaviors. The change in electrical parameters is explained with simulations, which relate the Voc and FF changes to a reduced recombination rate in the space charge region due to the light-induced doping increase. The simulations also suggest that cells with higher n-doping in the CdS are less sensitive to changes in interface recombination rate and doping of CIGS, which agrees with the hypothesis that the CdS buffer deposition is important for the LS behavior.
    Thin Solid Films 05/2015; 582. DOI:10.1016/j.tsf.2014.10.046 · 1.87 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper we present results of the efficiency optimization study performed in our production line for CIGS solar modules. For three different Ga/(Ga+In) ratios two different grading settings were tested. Results indicate high impact of these parameters on efficiency, in the range of 0.5% (abs.). Increasing the Ga content in the film, results in an increase of module voltage. This is almost completely compensated by the decrease of Isc. The resulting change in efficiency is FF driven. The FF increases slightly with higher overall Ga content, but the grading plays a more important role. For the grading setting 1, which was steeper than the setting 2, the FF and consequently efficiency were significantly improved. Optimum Ga content and grading has been defined accordingly.
    2013 IEEE 39th Photovoltaic Specialists Conference (PVSC); 06/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: We report a new certified world-record efficiency for thin-film Cu(In,Ga)Se2-based photovoltaic sub-modules of 17.4% (aperture area). The record efficiency of the 16 cm2, monolithically integrated, sub-module has been independently confirmed by Fraunhofer ISE. The record device is the result of extensive co-optimization of all processing steps. During the optimization process, strong focus has been put on the scalability of processes to cost-effective mass production, as reflected, for example, in Cu(In,Ga)Se2 deposition time and substrate temperature. Device manufacturing as well as results of electrical and material characterization is discussed. Copyright © 2012 John Wiley & Sons, Ltd.
    Progress in Photovoltaics Research and Applications 11/2012; 20(7). DOI:10.1002/pip.2246 · 9.70 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The influence of the average Se-to-metal overpressure during three-stage co-evaporation of Cu(InxGa1−x)Se2 solar absorber layers has been investigated. Solar cell devices were fabricated using a baseline process consisting of chemical bath deposited CdS, magnetron sputtered intrinsic and Al-doped ZnO, and e-beam evaporated Ni/Al/Ni current collection grid. For the higher Se-to-metal rate ratios studied, an increased short-circuit current in combination with a decreased fill factor is observed, while the open-circuit voltage stays fairly constant. Based on quantum efficiency measurements, fitting of IV data to a one-diode model, and simulations, we suggest the observed effects to be due to a decreased effective doping in combination with a decreased bulk recombination with higher average Se-to-metal overpressures. This could e.g. be explained by a decreasing number of Se-vacancies or VSe–VCu divacancies with higher Se rate. For the range of lower average Se-to-metal rate ratios studied, device performance drops due to a decreased open-circuit voltage and, for the lowest Se rate investigated, fill factor. In addition to electrical characterization, the effects on absorber microstructure are discussed based on results obtained from X-ray diffraction and scanning electron microscopy.
    Thin Solid Films 08/2011; 519(21):7237–7240. DOI:10.1016/j.tsf.2010.12.182 · 1.87 Impact Factor
  • MRS Online Proceeding Library 01/2011; 275. DOI:10.1557/PROC-275-299
  • [Show abstract] [Hide abstract]
    ABSTRACT: The effect of deposition pressure and controlled oxygen dosing on the diffusion barrier performance of thin film Ta to Cu penetration was investigated. In-situ resistivity, Auger compositional profiling, scanning electron microscopy and cross-sectional transmission electron microscopy were used to determine the electrical, chemical and structural changes that occur in Cu/Ta bilayers on Si upon heating. A 20 nm Ta barrier allowed the penetration of Cu at temperatures ranging from 320 to 630°C depending on processing conditions. Barrier failure temperature is dependent upon the deposition pressure and oxygen contamination at the Ta/Cu interface. This indicates the importance of understanding how deposition conditions affect diffusion barrier performance.
    MRS Online Proceeding Library 01/2011; 203. DOI:10.1557/PROC-203-387
  • [Show abstract] [Hide abstract]
    ABSTRACT: Thin film solar cells with the structure soda lime glass/Mo/Cu(In,Ga)Se2/Zn(O,S)/ZnO/ZnO:Al are studied for varying thickness and sulfur content of the Zn(O,S) buffer layer. These Zn(O,S) layers were deposited by atomic layer deposition (ALD) at 120 °C. Devices with no or small concentrations of sulfur in the buffer layer show low open-circuit voltages. This is explained by the cliff, or negative conduction-band offset (CBO), of −0.2 eV measured by photoelectron spectroscopy (PES) and optical methods for the Cu(In,Ga)Se2 (CIGS)/ZnO interface. Devices with ZnS buffer layers exhibit very low photocurrent. This is expected from the large positive CBO (spike) of 1.2 eV measured for the CIGS/ZnS interface. For devices with Zn(O,S) buffer layers, two different deposition recipes were found to yield devices with efficiencies equal to or above reference devices in which standard CdS buffer layers were used; ultrathin Zn(O,S) layers with S/Zn ratios of 0.8–0.9, and Zn(O,S) layers of around 30 nm with average S/Zn ratios of 0.3. The sulfur concentration increases towards the CIGS interface as revealed by transmission electron microscopy and in vacuo PES measurements. The occurrence of this sulfur gradient in ALD‐Zn(O,S) is explained by longer incubation time for ZnO growth compared to ZnS growth. For the Zn(O,S) film with high sulfur content, the CBO is large which causes blocking of the photocurrent unless the film is ultrathin. For the Zn(O,S) film with lower sulfur content, a CBO of 0.2 eV is obtained which is close to ideal, according to simulations. Efficiencies of up to 16.4% are obtained for devices with this buffer layer.
    Journal of Applied Physics 08/2006; 100(4):044506-044506-9. DOI:10.1063/1.2222067 · 2.19 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: CdS films were deposited with chemical bath deposition (CBD) onto Cu(In,Ga)Se<sub>2</sub>/Mo/soda-lime glass structures. The basic ingredients for the CBD were ammonia (NH<sub>4</sub>OH), cadmium acetate (CdAc), and thiourea CS(NH<sub>2</sub>)<sub>2</sub>. Two recipes with very low thiourea concentration were compared to the baseline recipe. From quantum efficiency measurements we observe an increased carrier collection for longer wavelengths for the modified buffer recipes as compared to baseline. The higher carrier collection is explained by an increase in the depletion region width. A lower bandgap is observed for the modified buffer. The solar cells with modified buffer layers have lower voltages and fill factors than the solar cells with our baseline buffer, leading to a lower efficiency for the modified cell structures in spite of slightly higher current densities. A tentative explanation is given assuming an inverted CIGS layer at the CIGS/buffer interface, possibly with Cd as a donor element
    Photovoltaic Energy Conversion, Conference Record of the 2006 IEEE 4th World Conference on; 06/2006
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: CuIn1−xGaxSe2 (CIGS) and CuInSe2 (CIS) thin-film solar cells, with ZnO buffer layers deposited by Atomic Layer Deposition (ALD), are examined with respect to dominant recombination path. They are compared with reference cells with CdS buffer layers. The principal method of examination is temperature-dependent J–V characterization (J(V)T), and the analysis of the J(V)T data has been modified in order to more reliably discern the dominant recombination path.Compared to the CIS cells with the traditional CdS buffer layer, the CIS cells with ALD–ZnO buffer layer exhibit the same dominant recombination path, i.e., recombination in the bulk of the absorber. For the CIGS cells (with [Ga]/([Ga]+[In])=0.3), however, the analysis of the cells with ALD–ZnO buffer points to dominant interface recombination, while the CdS buffer cells are dominated by bulk recombination.For CIGS, the difference between the recombination in ALD–ZnO and CdS cells is consistent with the negative conduction band offset found by photoelectron spectroscopy in these ALD–ZnO cells in a previous study. This offset leads to increased interface recombination.For CIS/ALD–ZnO, it was previously found that there is no negative conduction band offset since the conduction band minimum of the absorber is lower. Consistently there is no difference in dominant recombination path between ALD–ZnO buffer cells and traditional CdS buffer cells.
    Thin Solid Films 06/2005; DOI:10.1016/j.tsf.2004.11.008 · 1.87 Impact Factor
  • O. Lundberg, M. Edoff, L. Stolt
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper the effect of an in depth variation of the Ga/(In+Ga) ratio in GIGS based thin film solar cells is presented. The conclusions made are based on a review of earlier publications, theoretical considerations and results from a large set of new devices. For standard devices with normally thick CIGS films (1.5-2 mu m) deposited at a relatively long deposition time (60 min) an improved efficiency of around 0.4% units for the devices with an increased Ga/(In+Ga) ratio towards the back contact is observed. This improvement is due to a field assisted carrier collection resulting in an improved QE response at long wavelengths. When the CIGS thickness is reduced the importance of the increased Ga/(In+Ga) ratio towards the back contact is enhanced and at a GIGS thickness of 0.5 mu m a gain of 2.5% units is obtained. The gain is due to an improved V proportional to and FF. The main reason for the improvement is passivation of the back contact, which becomes increasingly detrimental for the device performance as the CIGS thickness is reduced. Also for pure CIS a significant improvement of the device performance is obtained by introducing an increased Ga concentration towards the back contact. This improvement is, however, more related to the introduction of Ga itself than the gradient of the Ga-concentration. From many simulations the largest gain is predicted for an increased Ga/(In+Ga) ratio towards the GIGS surface. However, neither in the literature nor from our own experiments we can find evidence for an improved device performance due to an increased Ga-concentration towards the GIGS surface. (c) 2004 Elsevier B.V. All rights reserved.
    Thin Solid Films 06/2005; 480-481:520-525. DOI:10.1016/j.tsf.2004.11.080 · 1.87 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The formation of the interface between In2S3 grown by atomic layer deposition (ALD) and co-evaporated Cu(In,Ga)Se2 (CIGS) has been studied by X-ray and UV photoelectron spectroscopy. The valence band offset at 160°C ALD substrate temperature was determined as −1·2±0·2 eV for CIGS deposited on soda-lime glass substrates and −1·4±0·2 eV when a Na barrier substrate was used. Wavelength dependent complex refractive index of In2S3 grown directly on glass was determined from inversion of reflectance and transmittance spectra. From these data, an indirect optical bandgap of 2·08±0·05 eV was deduced, independent of film thickness, of substrate temperature and of Na content. CIGS solar cells with ALD In2S3 buffer layers were fabricated. Highest device efficiency of 12·1% was obtained at a substrate temperature of 120°C. Using the bandgap obtained for In2S3 on glass and a 1·15±0·05 eV bandgap determined for the bulk of the CIGS absorber, the conduction band offset at the buffer interface was estimated as −0·25±0·2 eV (−0·45±0·2 eV) for Na-containing (Na-free) CIGS. Copyright © 2005 John Wiley & Sons, Ltd.
    Progress in Photovoltaics Research and Applications 04/2005; 13(3):179 - 193. DOI:10.1002/pip.595 · 9.70 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: A reactively sputtered ZrN reflector layer on top of the conventional Mo back contact yields enhanced absorber/back contact reflectance in Cu(In,Ga)Se2 thin film solar cells. Improved long wavelength quantum efficiency is demonstrated with a ZrN reflector at a Cu(In,Ga)Se2 thickness of 0.5 μm. The optical gain with respect to a standard Mo back contact is initially offset by increased back contact recombination and contact resistance, but these electronic losses can be suppressed by Ga grading of the absorber or by inclusion of a contact layer of MoSe2. This allows for a significantly improved power conversion efficiency of devices with sub-micron Cu(In,Ga)Se2 thickness.
    Applied Physics Letters 09/2004; 85(13):2634-2636. DOI:10.1063/1.1794860 · 3.52 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The reversible persistent changes of the fill factor (FF) induced by the illumination and voltage bias along with changes in the electronic properties of the ZnO/CdS/Cu(In,Ga)Se2 photovoltaic devices have been studied. Admittance spectroscopy and capacitance–voltage characterization reveal a correlation between the FF and the space charge distribution within the absorber. Our experiments provide evidence that a major source of FF loss in efficient devices is caused by excess negative charge close to the interface. We explain the persistent changes in the net acceptor concentration in the interface region by the relaxation effects due to compensating donors—the same mechanism, which leads to metastable changes of the doping level in the bulk of the absorber.
    Solar Energy Materials and Solar Cells 10/2003; 80(2):195-207. DOI:10.1016/j.solmat.2003.06.006 · 5.03 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Minority carrier traps in the absorber layer of the ZnO/CdS/Cu(In,Ga)Se2 photovoltaic devices have been investigated by use of deep level transient spectroscopy. In the efficient baseline structures a recombination process involving deep electron trap is revealed in the presence of blue light introducing holes from the buffer into absorber. A large capture cross-section for minority carriers and high concentration exceeding net acceptor concentration suggest that this trap plays a significant role as a recombination center in these devices. Its specific features indicate that it might also be a center involved in the metastable phenomena characteristic for these devices. Another deep electron trap, observed in the cells of inferior performance, has been investigated by double pulse DLTS. We conclude, that its concentration and values of capture cross-sections for holes and electrons are too low to account for the low efficiency of these structures.
    Journal of Physics and Chemistry of Solids 09/2003; 64:2041-2045. DOI:10.1016/S0022-3697(03)00092-1 · 1.59 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The diffusion of indium and gallium in polycrystalline thin film Cu(In,Ga)Se-2 layers has been investigated. Bilayer structures of CuInSe2 on top of CuGaSe2 and vice versa have been fabricated in both a Cu-rich and Cu-poor process (in relation to the ideal stoichiometry). In each process molybdenum coated soda-lime glass with and without a sodium barrier was used. These bilayers were analyzed with secondary ion mass spectrometry, X-ray diffraction, scanning electron microscope and transmission electron microscope equipped with energy dispersive X-ray spectroscopy. It was found that the grain boundary diffusion was not significantly higher than the diffusion inside the grains, also for Cu-rich layers. The diffusion is suggested to mainly proceed via vacant metal sites in the lattice structure. In sodium free films a higher diffusion into the bottom layers, compared to films with sodium, was seen in all cases. This observation was explained with a larger number of vacancies, that facilitates indium and gallium diffusion, in the sodium free films. The difference in diffusion between indium in CGS layers and gallium in CIS layers, in both Cu-rich and Cu-poor processes, was small for layers with sodium.
    Journal of Physics and Chemistry of Solids 09/2003; 64(9-10):1499-1504. DOI:10.1016/S0022-3697(03)00127-6 · 1.59 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Thin films of Cu(In,Ga)Se2 are grown by a co-evaporation process in which the In, Ga, and Se fluxes, as well as the substrate temperature, are constant and the only variable is the Cu flux. This Cu flux varies in three steps in such a way that the growing film evolves from Cu-poor to Cu-rich and then back to Cu-poor. The film growth is monitored by the ‘end point detection’ method, and film thicknesses of the order of 2 μm are deposited in less than 20 min. Quality devices (efficiencies above 15%) are produced in our baseline processes for all of the other synthesis steps. The Cu(In,Ga)Se2 layers are studied from a (112) versus (220) (204) orientation and recrystallization point of view. Including the results from a previous study on the influence of the substrate temperature to the present X-ray diffraction and scanning as well as tunneling electron microscopy data, a five-stage growth model for the films is described. The specific features of these films are that they are weakly (220) (204) oriented and exhibit crevices in their top fractions. The growth model hypothesizes about the origins of these crevices and on how to avoid them. Copyright © 2003 John Wiley & Sons, Ltd.
    Progress in Photovoltaics Research and Applications 07/2003; 11(5):319 - 331. DOI:10.1002/pip.495 · 9.70 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Cd-free solar cells based on Cu(In,Ga)Se/sub 2/, with efficiencies of up to 16.0%, are achieved by replacing the (CBD)CdS layer with a Zn(O,S)/ZnO bilayer deposited by ALD. Problems with reproducibility of the device results are observed and are found to be correlated with thickness variations of the Zn(O,S) layer, probably induced by differences of the CIGS surface. An ultra-thin Zn(O,S) layer, possibly not completely covering the CIGS surface, is observed by XPS analysis for high efficiency devices. Degradation of fill factor is observed on some, but not all, Zn(O,S) devices after about 2 months of storage, but these devices recover after light-soaking at elevated temperature. In order to improve the reproducibility, a new Zn(O,S) process is developed that includes longer ALD pulses. Using this new process, a promising average result of 12.1% (without AR coating) for 10 consecutive ALD runs is obtained with a maximum spread of /spl plusmn/1% unit.
    Photovoltaic Energy Conversion, 2003. Proceedings of 3rd World Conference on; 06/2003
  • [Show abstract] [Hide abstract]
    ABSTRACT: We have analyzed the potential for light trapping in Cu(In,Ga)Se/sub 2/ cells. Key quantities such as back contact reflectance and reflectance at the absorber-window interface are discussed and calculated from measured optical properties. Two model cases, perfectly specular interfaces and Lambertian scattering from the absorber front and back surfaces, are compared in terms of integrated AM 1.5 absorption in the absorber as a function of absorber thickness d/sub a/ for Ag, TiN, Mo back reflectors. Relative to the Mo reflector in the specular model, the potential gain of an Ag reflector at d/sub a/ = 0.5 /spl mu/m is estimated to 2.0 and 3.5 mA/cm/sup 2/ in the specular and scattering cases, respectively. Improved long wavelength quantum efficiency is experimentally demonstrated with a TiN back reflector. The resulting gain in J/sub sc/ is 0.8 mA/cm/sup 2/ at 0.45 /spl mu/m absorber thickness, in accordance with our model results.
    Photovoltaic Energy Conversion, 2003. Proceedings of 3rd World Conference on; 06/2003
  • [Show abstract] [Hide abstract]
    ABSTRACT: The electronic properties of the interface region in the ZnO/CdS/Cu(In,Ga)Se2 devices have been investigated in the various metastable states induced by voltage bias and illumination. Capacitance spectroscopy has been used to gain information about the level spectrum in the interface region of absorber and space-charge distribution within the structures. The results of capacitance spectroscopy are analyzed in conjunction with the current–voltage characteristics. We have differentiated between the metastable effect due to the changes of the space-charge distribution in the absorber and a process involving the persistent changes of the Fermi-level position at the interface. We attribute the first one to the electronic processes due to relaxing donor-type VSe centers. The second one in our opinion involves the process of forming a quasi-equilibrium between the positive and negative charges in the immediate vicinity of the interface. In the admittance and DLTS spectra of interface levels a signal belonging to bulk donors (most probably to InCu defects) has been identified.
    Thin Solid Films 05/2003; 431:153-157. DOI:10.1016/S0040-6090(03)00221-9 · 1.87 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Solar cells based on CuInSe2 (CIS) and Cu(In,Ga)Se2 (CIGS), with a ZnO buffer layer deposited by atomic layer deposition (ALD), are compared to their reference cells; (CIS or CIGS)/CdS/ZnO. While the CIS/ZnO devices show only slightly lower efficiencies compared to their reference cells, the difference between the CIGS/ZnO devices and their reference cells is larger. In the latter case, the main difference is the lower open circuit voltage of approximately 200 mV of the direct ZnO devices. The valence band offset between CIS and ZnO is determined by ultraviolet photoelectron spectroscopy to −2.2 eV which gives a conduction band offset, ΔEc, of +0.1±0.2 eV. This slightly positive offset is in contrast to our previous result for the CIGS/ZnO interface of ΔEc=−0.2±0.2 eV, and is a possible explanation for the much lower voltage loss observed for the CIS/ZnO devices. Zn diffusion into the different absorbers is investigated by energy dispersive X-ray spectroscopy on transmission electron microscope cross-sections prepared from direct ZnO devices. These cross-sections also show very good coverage of the absorber surface by the (ALD)ZnO layer.
    Thin Solid Films 05/2003; s 431–432:321–325. DOI:10.1016/S0040-6090(03)00229-3 · 1.87 Impact Factor

Publication Stats

2k Citations
175.14 Total Impact Points

Institutions

  • 1990–2006
    • Uppsala University
      • Department of Engineering Sciences
      Uppsala, Uppsala, Sweden
  • 1996
    • Alimetrics Ltd, Espoo, Finland
      Helsinki, Uusimaa, Finland
    • University of Illinois, Urbana-Champaign
      • Department of Materials Science and Engineering
      Urbana, Illinois, United States
  • 1993–1994
    • KTH Royal Institute of Technology
      Tukholma, Stockholm, Sweden
  • 1990–1991
    • IBM
      Armonk, New York, United States
  • 1989
    • Lawrence Berkeley National Laboratory
      Berkeley, California, United States