Chemical bath process for highly efficient Cd-free chalcopyrite thin-film-based solar cells

Abstract

The highest efficiency for Cu(Ga,In)Se 2 (CIGS) thin-film-based solar cells has been achieved with CdS buffer layers prepared by a solution growth method known as the chemical bath deposition (CBD). With the aim of developing Cd-free chalcopyrite-based thin-film solar cells, we describe the basic concepts involved in the CBD technique. The recipes developed in our laboratory for the heterogeneous deposition of good-quality thin films of ZnO, ZnSe, and MnS are presented. In view of device optimization, the initial formation of chemical-bathdeposited ZnSe thin films on Cu(Ga,In)(S,Se) 2 (CIGSS) and the subsequent development of the ZnSe/CIGSS heterojunctions were investigated by X-ray photoelectron spectroscopy (XPS). The good surface coverage was controlled by measuring changes in the valence-band electronic structure as well as changes in the In4d, Zn3d core lines. From these measurements, the growth rate was found to be around 3.6 nm/min. The valence band and the conduction band-offsets E V and E C between the layers were determined to be 0.60 and 1.27 eV, respectively for the CIGSS/ZnSe interface. The energy-band diagram is discussed in connection with the band-offsets detemined from XPS data. A ZnSe thickness below 10 nm has been found to be optimum for achieving a homogeneous and compact buffer layer on CIGSS with a total area efficiency of 13.7%.Résumé : Le plus haut rendement pour des cellules solaires en couche mince de Cu(Ga,In)Se 2 (CIGS) a été obtenu avec des couches tampons de CdS préparées par une méthode de croissance en solution, appelée dépôt en bain chimique (CBD). Le but est de développer des cellules solaires chalcopyrites en couches minces sans Cd et nous décrivons ici les concepts de base de la technique CBD. Nous présentons les recettes développées dans notre laboratoire pour faire des dépôts hétérogènes de haute qualité de couches minces de ZnO, ZnSe et MnS. Afin d'optimiser le processus, nous utilisons la spectroscopie X des photoélectrons (XPS) pour étudier la formation initiale de couches minces de ZnSe déposées en bain chimique sur du Cu(Ga,In)(S,Se) 2 (CIGSS) et le développement subséquent des hétérojonctions ZnSe/(CIGSS). Le bon recouvrement de la surface est contrôlé en mesurant les changements dans la structure électronique de la bande de valence aussi bien que dans les lignes du coeur In4d, Zn3d. Ces mesures nous indiquent que le taux de croissance est d'environ 3,6 nm/min. Les décalages E V et E C des bandes de valence et de conduction entre les couches sont de 0,60 et 1,27 eV respectivement pour l'interface CIGSS/ZnSe. Nous analysons le diagramme des bandes

Formation of a ZnS∕Zn(S,O) bilayer buffer on CuInS2 thin film solar cell absorbers by chemical bath deposition

“…[10][11][12] In this case, the obtained device efficiencies were comparable to those of corresponding CBD-CdS buffered referencesindependently confirmed by the National Renewable Energy Laboratory ͑NREL͒ ͑Golden, CO, U.S.A.͒. 12 Note that at this time the CBD process used still contained the highly toxic reactant hydrazine.…”

“…12 Note that at this time the CBD process used still contained the highly toxic reactant hydrazine. [10][11][12] However, when transferring this approach to wide-gap CIS-based solar cells, resulting devices with CBDZnS x O y H z buffers have yielded ͑active area͒ efficiencies of up to 10.7%, 13 whereas corresponding CBD-CdS buffered references reached 11.9%, showing again the efficiency gap of ϳ1% ͑abs.͒. Recent additional efforts initially aiming at a single-layer, nominal ZnS buffer CBD without the need of any toxic reactants such as hydrazine have helped us to close a͒ Authors to whom correspondence should be addressed.…”

“…14 Since the parameters of the chemical bath deposition ͑pH value, temperature, complexing agents, deposition time, reactants and their concentration, application of ultrasound, rinsing solution, etc.͒ determine the properties of the prepared layer, there are as many different possible compositions ͑and consequently also notations͒ as CBD processes ͑see above͒. For a chemical bath consisting mainly of a Zn salt, thiourea, in some cases hydrazine, [10][11][12][13] and aqueous ammonia one can find in the literature notations for the composition of the deposited layers from as simple as CBD-ZnS ͑Ref. 6͒ over Zn͑S,OH͒, [10][11][12] ZnS͑O,OH͒, 7,9 Zn͑O,S,OH͒ x ͑Ref.…”

“…For a chemical bath consisting mainly of a Zn salt, thiourea, in some cases hydrazine, [10][11][12][13] and aqueous ammonia one can find in the literature notations for the composition of the deposited layers from as simple as CBD-ZnS ͑Ref. 6͒ over Zn͑S,OH͒, [10][11][12] ZnS͑O,OH͒, 7,9 Zn͑O,S,OH͒ x ͑Ref. 8͒ to ZnS x O y H z .…”

Intermixing at the heterointerface between ZnS∕Zn(S,O) bilayer buffer and CuInS2 thin film solar cell absorber

“…͑Note that these values are based on the active area of the solar cell.͒ Recently, this efficiency gap could be closed for production scale low-gap Cu͑In, Ga͒͑S,Se͒ 2 absorbers provided by Shell Solar GmbH ͑Munich, Germany͒ applying CBD-Zn-compound buffer layers developed at the Hahn-Meitner-Institut Berlin ͑HMI͒ ͑Berlin, Germany͒. [5][6][7] In this case, the obtained total area device efficiency ͓ max ͑Zn-S͒ = 14.4% ͒ was comparable to that of corresponding CBD-CdS buffered references ͓ max ͑CdS͒ = 14.6% ͔-independently confirmed by NREL. 7 However, when transferring this approach to wide-gap ͑E g = 1.54 eV͒ CuInS 2 ͑CIS͒ based devices-which, according to theoretical considerations, 12 promise a higher efficiency than the current low-gap world record chalcopyrite solar cell and are additionally of particular interest in terms of a possible application as top cell in a prospective chalcopyrite tandem cell-resulting devices with CBDZnS x O y H z buffers have yielded active area efficiencies of up to 10.7%, 11 whereas corresponding CBD-CdS buffered references reached 11.9%, again showing the efficiency gap of ϳ1% ͑abs.͒.…”