TW201502075A - Nanoparticles, ink and process for making and using - Google Patents

Abstract

This invention relates to ternary, quaternary or multinary chalcogenide semiconductor nanoparticles (NPs) stabilized by inorganic ions including tetrahalogenoborate, hexahalogenophosphate or hexahalogenoantimonate. The invention further relates to stable nanoparticle inks formulated with these modified nanoparticles, and semiconductor layers and devices generated using these inks. The nanoparticles are preferably semiconducting nanocrystals of the ternary, quaternary or multinary metal chalcogenide type. The specifically stabilized nanoparticles can be produced by a ligand exchange starting from conventional nanoparticles.

Description

Nanoparticles, inks, and methods of manufacture and use

The present invention relates to ternary, quaternary or polychalcogenide semiconductor nanoparticles (NP) stabilized by inorganic ions comprising a tetrahalo borate, a hexahalophosphate or a hexahalogen ruthenate. The invention further relates to stabilized nanoparticle inks formulated with such modified nanoparticles, and semiconductor layers and devices formed using such inks. The nanoparticle is preferably a semiconductor nanocrystal of the ternary, quaternary or polymetallic chalcogenide type. Particularly stable nanoparticles can be produced by ligand exchange starting from the learned nanoparticle.

Based on I-III-VI ₂ type such as CuInSe ₂ (CIS), CuIn(S _y , Se _1-y ) ₂ (CISS), CuIn _x Ga _1-x (Se _y , S _1-y ) ₂ (CIGS) Copper semiconductor (chalcopyrite) is a semiconductor widely studied as an absorption layer for thin film solar cells. CISS and CIGS have a direct band gap that can be adjusted to match the solar spectrum by changing the In/Ga ratio or by changing the S/Se ratio.

Solar cells based on polycrystalline copper indium diselenide (CIS) and its gallium-sulfur alloy (CIGS) films exhibit impressive cell efficiencies and unexpected stability in the module. In recent years, laboratory-scale CIGS cells sputtered onto soft plastics have achieved a new record of 20.4% energy conversion efficiency (www.empa.ch). The performance of CIGS has been higher than the best reported so far in the literature on competitive technologies CdTe (16.5%) and a-Si (12%).

Considering the low profit margins associated with the manufacture of photovoltaic devices, research has also focused on a class of multi-metal chalcogenide materials containing abundant and inexpensive metals in the earth. Copper sulfide tin (Cu ₂ SnS ₃ or "CTS") and copper zinc tin oxide (Cu ₂ ZnSnS ₄ or "CZTS") are particularly useful examples of such materials. Similar to CIGS, CZTS can be doped with an element such as Ge to change its band gap. Although the recording efficiency of CZTS has been lower than the 11% battery efficiency of CIGS exhibited by IBM (Todorov et al., Adv. Energy Mater. 2013 , 3, 34-38) in recent years, zinc is higher than gallium and indium. And the abundance of tin makes it attractive from a cost perspective.

Conventional vacuum co-evaporation and sputtering processes for depositing thin film absorbers are costly, technically demanding, and wasteful, thus hindering the implementation of large scale photovoltaic device fabrication. In order to achieve mass production, major manufacturing challenges must be overcome, such as maintaining a uniform composition of large areas, achieving precise control of flux/deposition rates to avoid intermediate phases, and achieving high material utilization.

It is possible to match the market value of carbon-based energy production via solution-based deposition. Various solution deposition techniques can be used to deposit the precursor to the absorber layer.

A number of non-vacuum operating methods for forming semiconductors have been described in the literature (Hibberd et al., Prog. Photovolt: Res. Appl. 2010 , 18, 434-452). These methods include various solution based methods, including those based on the pyrolysis of metal salts or more complex molecular precursors. An alternative approach relies on the deposition of nanoparticle suspensions.

In the first report of CIGS nanoparticle synthesis (Guo et al., NanoLett. 2008 , 8(9), p. 2982), CIGS nanoparticle inks were deposited and processed to achieve 12.0% photon conversion efficiency within five years. The parameters are optimized. To achieve this, a film formed from a copper indium gallium disulfide (CIGS) nanocrystal in an alkanethiol solvent is selenized (Guo et al., Prog. Photovolt: Res. Appl. 2012 , doi: 10.1002/pip). .2200) Prior to aqueous NaCl exposure treatment. Despite the promising efficiency values, these films contain higher carbon contents due to the aliphatic ligands of oleylamine ( _C18 primary amine) in this case. After heat treatment, the degradation products of the ligand were found to collect near the bottom contact, limiting battery efficiency.

Therefore, a major obstacle to achieving even higher efficiencies from CIGS devices based on nanoparticles is that nanoparticles are conventionally stabilized by organic ligands, usually oleoamine (OLA), which eventually contaminate the semiconductor. Layer and reduce device performance. Ligands are indispensable during the synthesis of nanoparticles and have been shown to be related to the size and morphology of the controlled nanoparticles (Kar et al, J. Am. Chem. Soc. 2011 , 133(43), p. 17239- 17247 pages). Subsequently, the organic ligand is used to dissolve the nanoparticles in an organic solvent before and during deposition. However, after the film is formed, the residual elements of the ligand can affect the device. The residual components of the ligand prevent the formation of the desired micron-sized grains in the absorber layer. The adverse effects of carbonaceous residues that accumulate at the molybdenum interface have been documented (Kaelin et al, Thin Solid Films 2005 , 480-481 (2005), page 486).

The precursor nanoparticles which have been stabilized by ruthenium have been reported in the literature and recently the most efficient solution-processed devices known to be known (Todorov et al., Advanced Materials 2010 , 22, E156-E159). Nonetheless, nanoparticle inks containing less harmful and toxic substances have great advantages for industrial applications.

WO 2009/137637 and WO 2011/066205 are exemplary documents for reporting the use of ternary metal chalcogenide nanoparticles and their use in film formation. Amines are often used as ligands in the synthesis and further use. Inorganic ligands such as tetrafluoroborate, hexafluorophosphate or hexachloroantimonate in the halogen group and the like are not mentioned as ligands.

There is also a need for a ternary or multi-component metal chalcogenide material that can be smoothly converted to a semiconductor film having a high crystalline morphology and good solar conversion.

For industrial processes, the materials and compositions used to create the semiconductor layer should be stable, cost effective, environmentally friendly, and easy to handle without exposure to excessive hazards.

definition

The term "chalcogen" in accordance with the invention is limited to sulfur (S), selenium (Se) and to some extent strontium (Te). Selenium (Se), sulfur (S), and a combination of S and Se are preferred chalcogen elements.

The term "metal chalcogenide" means metal sulfide, metal selenide or metal telluride, and combinations thereof.

A "binary" chalcogenide is composed of a single metal and a chalcogenide such as In ₂ S ₃ or Cu ₂ Se. "Ternary" chalcogenide means a material composed of two metals and a chalcogenide such as CIS (CuInS ₂ ) or CISS (CuIn(S, Se) ₂ ). "Quaternary" chalcogenides similarly represent materials composed of three metals and chalcogenides, such as CIGS. "Diversified" chalcogenides similarly represent materials composed of even more metals.

The chalcogenide semiconductor used throughout the present invention is referred to by its type according to its elemental composition. Number I (for example, for Cu, Ag), II (for example, for Zn, Cd), III (for example, for Ga, In), IV (for example, for Ge, Sn), and VI (for example, The S, Se, Te) are used to represent a certain family element having the same valence electron number, and are added here to the parentheses. Therefore, "I-III-VI" type semiconductor generally means a ternary chalcogenide mainly comprising a metal selected from the group consisting of 11th (also known as IB) and 13 (also known as IIIA) and chalcogenide group (Group I6). . Similarly, the "I-II/VI-VI" type semiconductor means mainly consisting of four metals selected from the 11th (also known as IB) and 12 (also known as IIB), 14 (also known as IVA) and chalcogenide families. Elemental chalcogenide. The stoichiometry varies around the values typically found in this type of semiconductor chalcogenide. An ideal stoichiometric formula, such as CuInS ₂ or Cu ₂ ZnSnSe ₂ , can vary significantly by exchanging metals of different families.

The terms "CIGS" and "CZTS" as used throughout the present invention are similar to the common understandings in the literature. CIGS stands for copper indium gallium selenide with different element distributions. CZTS represents selenized copper zinc tin with different elemental distributions. The compounds are often alternatively described by varying molecular formulas such as Cu(In,Ga)(S,Se) ₂ and Cu ₂ ZnSn(S,Se) ₄ . When the total amount is consistent with the stoichiometric demand of the formula, the elements in parentheses are combined to indicate either or both of the elements. An exemplary (S, Se) _n represents S _nx Se _x and x is a value between 0 and n. For example, when n is 2, x can represent 0, 0.1, 0.2, 0.5, 1.0, 1.5, 1.8, 1.9, or 2.0. In general, weak copper element compositions derived from typical stoichiometry are useful because they exhibit advantageous absorption characteristics. In this sense, the formula throughout the present invention includes such variations in the distribution of elements. In the two absorbents of CIGS and CZTS, there may be some or all of the sulfur instead of selenium or vice versa. Other elements may be present in a smaller amount, for example, instead of Cu, or Sn, Zn, Cd in CIGS, or In, Ge, Cd in CZTS, such as trace elements of Na, Sb, Te, As, etc. .

The term "nanoparticle" is intended to include particles containing chalcogenide characterized by an average maximum dimension of from about 1 nm to about 1000 nm, or from about 1 nm to about 500 nm, or from about 2 nm to about 300 nm. The nanoparticles may have different shapes including polyhedrons, spheres, rods, wires, tubes, sheets, whiskers, rings, discs or prisms. Most nanoparticles are crystalline and often have a single crystal structure. These nano particles are also known as "nano crystals". Unless otherwise specified, the nanoparticles may have a "surface coating" for use as a dispersing aid. Surface coatings, which may also be referred to herein as "surface ligands" or "blocking agents", may be physically adsorbed to the nanoparticles by, for example, an inductive charge or (partial) charge, or by, for example, hydrogen bonding or The covalent bond is more or less chemically bonded to the nanoparticle. The surface coating may also be a solvent or may comprise one or more ligands added to the reaction mixture. Throughout this specification, all references to wt% of nanoparticle are intended to include surface coatings.

The term "ink" and "dispersion" are synonymous in this context unless otherwise specifically defined. The ink is intended to describe a homogeneous mixture of nanoparticles, a liquid medium, and optionally other components.

A "ligand" for use in a nanoparticle as used herein is a neutral or ionic molecule or atom that is suitable for more or less tight adhesion to the surface of the nanoparticle. The interaction between the surface and the ligand can be ionic, covalent or any other type.

In accordance with one or more objects of the present invention, as embodied and broadly described herein, the present invention provides a highly efficient semiconductor, in particular a metal chalcogenide, from a liquid composition (NP ink) of suitable colloidal nanoparticles. Crystalline film. The present invention is further directed to providing nanoparticle and stabilized nanoparticle inks that can be converted to a crystalline absorption layer having a low carbon content for use in high efficiency metal chalcogenide solar cells.

According to one aspect of the invention, the invention relates to the inclusion of tetrahalogenated boric acid An inorganic ion-stabilized ternary, quaternary or multi-component metal chalcogenide nanoparticle (NP) of a root, a hexahalophosphate or a hexahalogenated ruthenate, and a method of preparing the same.

According to another aspect of the present invention, there is disclosed a ternary, quaternary or multicomponent metal comprising an inorganic ion dispersed in a liquid medium and stabilized by an inorganic ion comprising a tetrahalo borate, a hexahalophosphate or a hexahalogen ruthenate Nanoparticle ink of chalcogenide nanoparticle (NP).

Another embodiment of the invention comprises a method of combining a metal chalcogenide nanoparticle according to the invention with a liquid medium to provide an ink.

In another aspect of the invention, a method of preparing ternary, quaternary or polybasic metal chalcogenide nanoparticles (NP) is provided, which is selected from the group consisting of a tetrahalo borate, a hexahalophosphate or The inorganic ions of the hexahalogenated ruthenate treat the NPs such that the NPs are stabilized by the aforementioned inorganic ions.

In another aspect of the invention, there is provided a method of preparing (crystalline) a ternary, quaternary or multi-component metal chalcogenide layer comprising simultaneously or in combination with two or three of the steps or the steps Method steps: a) depositing a nanoparticle ink comprising the nanoparticle and liquid medium on a substrate b) removing the liquid medium, and c) annealing or fusing as appropriate in the presence of additional sources of chalcogen Nano particles.

In another aspect of the invention, the invention provides a photovoltaic device, a thin film transistor, a water-cleaving photoelectrode, a sensor, and other components for an electronic product comprising a metal chalcogenide layer made in accordance with the present invention.

(1) ‧ ‧ absorbing layer

(2) ‧ ‧ molybdenum back contacts

(3) ‧‧‧Substrate glass

The invention will be more fully explained and illustrated by the following description and examples taken in conjunction with the accompanying drawings.

Figure 1: Scanning Electron Photomicrograph (SEM) shows a cross section of a CIGS absorber layer (1) produced as in Examples 1 to 4 below in accordance with the present invention. This picture shows a significantly larger The crystal grains of the CIGS semiconductor are densely distributed in the absorption layer (1) with some small crystal grains in between. A film having a thickness of about 2 microns was produced after selenization in a tube furnace at 550 ° C for 30 minutes. The molybdenum back contact (2) and the substrate glass (3) are shown below the absorbent layer.

Figure 2: This figure shows the photovoltaic device response and the CIGS photovoltaic device of the following example 5.a in accordance with the invention in dark (dashed line) and AM1.5 light conditions (solid line; according to IEC 904-3 (1989), The IV characteristics under the light condition of Part III).

Figure 3: This figure shows the photovoltaic device response and the CZTS photovoltaic device according to the following example 5.b of the present invention under dark (dashed line) and AM1.5 light conditions (IEC 904-3 (1989), part III) The IV feature.

In the development of nanocolloidal inks for device applications, the primary achievement of the present invention is to reduce the carbon loading of the ink without compromising stability in solution. A conventionally produced layer having a higher carbon content forms a double layer having a desired crystalline layer near the surface and a suboptimal layer having smaller particles near the bottom layer after annealing. Analysis of the two layers shows that carbon slag is concentrated in the layer with smaller particles near the bottom layer (substrate or, for example, the Mo layer), which may result in the observation of weak crystallinity. In contrast, a layer having a lower carbon content in accordance with the present invention exhibits a higher degree of crystallinity with dense columnar grains spanning the entire thickness of the layer. It has been found in accordance with the present invention that a highly crystalline layer containing a trace amount of carbon can be produced. As part of the present invention, a nanoparticulate ink is provided which results in the desired low carbon content of the semiconductor layer while the ink still has high stability against precipitation and degradation in combination with ease of processing.

Photovoltaic devices fabricated from particles according to the present invention are superior in device performance. The power conversion efficiency is almost twice that of the comparative example. Other device parameters have also been improved.

In view of the advantage of lowering the carbon content, it has been shown that only the hydrocarbon ink is washed and a stable ink cannot be obtained. It has been shown that inks made from nanoparticles which have been washed 3 to 4 times to achieve a low carbon content to remove the ligands condense into agglomerates within a few hours. Ink that provides only short-term stability is not good for industrial processes.

Unexpectedly, it has been found to replace the convenient route of natively based primarily on carbon ligands. In this aspect of the invention, there is provided a process for the preparation of ternary, quaternary or polybasic metal chalcogenide nanoparticles (NP) by selection from a tetrahalo borate, a hexahalophosphate or a hexa The inorganic ions of the halo citrate treat the NPs such that the NPs are stabilized by the corresponding inorganic ions. The method of treating with inorganic ions is carried out in the following manner: after the treatment, the surface of the nanoparticles is substantially covered by the inorganic ions. The treatment is preferably carried out by using a liquid medium containing inorganic ions or a salt thereof. In this method, most of the previous organic ions are preferably replaced, and more preferably removed from the surface of the nanoparticle. Removal herein means that the resulting particles are as small as the previous ligand, i.e., less than 5% by weight, or even less than 2% by weight or 1% by weight of the nanoparticles are caused by the residual ligand. In the case of often unstable ligands, the amount of this ligand can be measured by thermogravimetric analysis. Therefore, the nanoparticles are stabilized by the new inorganic ligand. The nanoparticle is then preferably in the absence of an organic ligand (a ligand comprising a carbon atom). Preferably, the nanoparticles also do not contain an anthracene ligand. The tetrahalo borate, hexahalophosphate or hexahalodecanoate is provided as a solution containing such an anion which can be provided by dissolving such salts or conversion from the corresponding acid and such agents. The halogen in these anions represents fluorine, chlorine, bromine or iodine. Preferably, the halogen in such anions represents a fluorine or chlorine substituent. More preferably, the tetrahalo borate, the hexahalophosphate and the hexahalogen decanoate are independently tetrafluoroborate, hexafluorophosphate and hexachloroantimonate, respectively.

Useful and suitable cations for association with inorganic ions include trialkyl oxonium, nitrosonium, H ⁺ , ammonium, mono/di/tri/quaternary alkyl ammonium, alkyl pyridinium (eg 1-butyl-) 4-methylpyridinium), alkylimidazolium (such as 1-ethyl-3-methylimidazolium), pyrrolidinium, morpholinium, piperidinium, hydrazine, hydrazine and metal cations, preferably Cu ⁺ . In the trialkyloxindole, alkyl means a straight or branched alkyl group preferably having 1 to 15 carbon atoms, more preferably a linear alkyl group having 1 to 7 carbon atoms, and Most preferred is methyl or ethyl. The alkyl substituent of pyridinium, pyrrolidinium, morpholinium, piperidinium, hydrazine, hydrazine and imidazolium is preferably a linear or branched alkyl group having 1 to 7 carbon atoms. A particularly preferred reagent is trimethyloxonium or triethyloxonium. Triethyloxonium tetrafluoroborate is a well-known Meerwein salt. Some anions such as trialkyloxonium and tetraalkylammonium exhibit particular benefits because of their ability to alkylate oleylamine ligands and thereby inactivate. This results in a more favorable ligand exchange than an alternative system that is not alkylated. In addition, in the case of, for example, H ⁺ and nitrosonium cations, alkylammonium and alkyl oxonium cations are less likely to degrade metal chalcogenide particles.

Solution-based deposition provides many advantages over vacuum-based methods, including high material utilization, excellent composition uniformity, and cost-effective roll-to-roll mass production. (See also Hibberd et al., Prog. Photovolt: Res. Appl., 2009 ). Many solution deposition techniques can be used in accordance with the present invention to deposit an absorbent layer such as spray coating, spin coating, ink jet printing, knife coating, slit coating, flexographic/gravure printing, dispensing, dip coating, and the like. The nanoparticle ink can be easily processed by transfer to a surface to be covered with a semiconductor material by the above deposition technique.

In accordance with the present invention, a stoichiometrically controlled absorber layer can be produced for use in photovoltaic devices by controlling the stoichiometry of the nanoparticles (i.e., the relative amounts of materials comprising the nanoparticles). Under suitable conditions, the single compound nanoparticle dispersed in the ink is converted to a solid single compound absorbent layer by a defined stoichiometry. It is also possible to combine two or more different nanoparticles into the chalcogenide absorption layer.

According to the invention, the semiconductor used in the method and formed as a layer is preferably I-III-VI ₂ or I ₂ -II-IV-VI ₄ type, sometimes I ₂ -IV-VI ₃ type and has a non- A mixed form of integer index stoichiometry. One or more (+III) valence metals are used for the I-III-VI ₂ type semiconductor, preferably the metals are selected from In and Ga, more preferably In and In combined with Ga. The monovalent metal is preferably copper or silver, more preferably copper. A small amount of silver (0.01 to 2 mol%) may be beneficial for the formation of the crystalline absorption layer and the final device performance. The (+III) valence metal is preferably selected from indium or gallium. A mixture of such metals can be used to adjust the band gap of the semiconductor. A typical and preferred stoichiometry is Cu(In,Ga)(S,Se) ₂ . In addition, one or more (+III) valence metals may be combined with (+II) valence and (+IV) valence metal (I-II-IV-VI ₂ type semiconductor, such as Cu ₂ (Zn/Sn) Se ₄ , Cu ₂ (Zn/Sn)S ₄ , Cu ₂ (Zn/Ge)Se ₄ ) or only partially or completely interchangeable with (+IV) metal. The (+II) valence metal is preferably cadmium or zinc, preferably zinc. The (+IV) valence metal is preferably ruthenium or tin, preferably tin. A typical stoichiometric formula for the I ₂ -II-IV-VI ₄ type semiconductor is Cu ₂ ZnSn(S,Se) ₄ . Cu ₂ SnSe ₃ is an example of the stoichiometry of I ₂ -IV-VI ₃ . In general, weak copper element compositions that deviate from typical stoichiometry are preferred and exhibit advantageous absorption characteristics.

The band gap of the resulting semiconductor layer affects the absorption characteristics in the photovoltaic device. The band gap can be adjusted to a desired value by changing the elemental composition of the layer, which is mainly controlled by the elemental composition of the nanoparticle ink and, optionally, the deposition process. A variety of useful semiconductor compositions have been reported in the literature due to the broader spectrum. A preferred stoichiometry for CIGS can be Cu _x In _y Ga _2-yx (S, Se) ₂ , where x can vary from 0.8 to 1.0 and y can vary from 0.2 to 0.6.

I-III-VI ₂ and I-II-IV-VI ₂ type semiconductors can be fabricated and preferably do not contain cadmium, which is an environmental advantage over prior art Cd based semiconductors. The I-II-IV-VI type ₂ semiconductor, especially containing zinc-tin variants, is advantageous from a resource point of view because it uses only elements rich in the earth.

The liquid medium of the nanoparticulate ink typically contains a liquid based material such as a solvent or other liquid. The liquid medium of the nanoparticulate ink preferably contains an organic solvent or a mixture of two or more organic solvents. Typically, the solvent evaporates quickly when the mixture is applied to the substrate and heated to at least above the boiling point of the solvent. Therefore, any solvent evaporation is carried out in one of the steps above and below. The liquid medium is preferably composed of a solvent having a boiling point of less than 180 ° C, more preferably less than 125 ° C.

The liquid medium is used as a carrier vehicle for the metal chalcogenide nanoparticles. The ink comprises from 0.5 wt% to 80 wt%, preferably from 1 wt% to 50 wt%, more preferably from 2 wt% to 30 wt%, most preferably from 3 wt% to 20 wt% of metal chalcogenide nanoparticles, and from 20 wt% to 98 wt% Preferably, from 50% to 95% by weight, optimally from 70% to 95% by weight of the liquid medium. An ink containing more than about 30% by weight of NP is a paste. Paste needs to be compared to lower viscosity inks Some printing technique, such as knife coating. For spray inks, nanoparticle loading is typically low, as low as 1% by weight. The nanoparticle preferably comprises from 80% by weight to 100% by weight, more preferably from 90% by weight to 100% by weight, based on the weight of the granule, of the single-formulated metal chalcogenide moiety.

The liquid medium preferably comprises one or more liquids selected from the group consisting of water, aromatics, heteroaromatics, alkanes, nitriles, ethers, ketones, esters, acetates, Amidoximes, amines, thiols, carboxylic acids, organic halides and alcohols. Suitable liquids include aromatic compounds (toluene, xylene, mesitylene, benzene), water, dimethyl hydrazine (DMSO), pyridine, 2-methylpyridine, 3-methylpyridine, 3,5-dimethyl Pyridine, 5-tert-butylpyridine, hexane, heptane, octane, cyclohexane, 2,2,4-trimethylpentane, acetonitrile, 3-methoxypropionitrile, dimethoxy Ethane, ethylene glycol diethyl ether, tetrahydrofuran, acetone, 2-pentanone, cyclopentanone, cyclohexanone, ethyl acetate, butyl acetate, dimethylformamide (DMF), N-methylpyrrolidine Ketone, tetramethylethylenediamine, 3-methoxypropylamine, dimethylaminoethanol, oleylamine, triethylamine, 1-octylthiol, 1-anthracenethiol, 1-dodecanethiol, mercapto Ethanol, thioglycolic acid, oleic acid, dichloromethane, chloroform, chlorobenzene, 1,2-dichlorobenzene, methanol, ethanol, isopropanol, n-propanol, 2-methoxyethanol, 2-ethoxy Ethyl alcohol, 2-butoxyethanol, monoterpene alcohol (α-, β-terpineol, or a combination of isomers α-, β- and γ-terpineol) and 2, 2, 4-three Methyl-1,3-pentanediol monoisobutyrate (Texanol). Such as pyridine, 2-methylpyridine, 3-methylpyridine, 3,5-lutidine, 5-tributylpyridine, acetonitrile, 3-methoxypropionitrile, tetramethylethylenediamine, 3 - methoxypropylamine, dimethylaminoethanol, oleylamine, triethylamine, 1-octylmercaptan, 1-decyl mercaptan, 1-dodecanethiol, mercaptoethanol, thioglycolic acid, oleic acid, Compounds such as 2-methoxyethanol, 2-ethoxyethanol, 2-butoxyethanol, and 2,2,4-trimethyl-1,3-pentanediol monoisobutyrate (Texanol) can be used. A dispersing or blocking group, and a carrier vehicle as a particle. Preferred liquid media contain one or more selected from the group consisting of volatile solvents, polar solvents, and/or solvents that do not act as NP capping agents. Particularly preferred is a solvent which does not serve as a blocking agent, that is, a solvent composed only of the elements (C, H) and a solvent composed of the elements (C, H, O), such as esters, ketones, alcohols, etc. Wait. These solvents are usually available It is easy to remove in the process without causing residue or voids in the layer.

The ink may further comprise up to about 20% by weight, or up to about 10% by weight, or up to about 5% by weight, or up to about 2% by weight, or up to about 1% by weight of one or more additives, including dispersants, surfactants, binders, Coupling agent, emulsifier, defoaming agent, plasticizer, desiccant, filler, extender, thickener, membrane regulator, antioxidant, leveling agent, leveling agent, ligand, end group Groups, chalcogens, dopants and corrosion inhibitors. While most additives may improve the method of forming the semiconductor layer, care should be taken that the additives do not interfere with the crystallinity of the final layer and the electron (photovoltaic) performance of the layer.

Suitable binders include polymers and oligomers having linear, branched, comb/brush, star, hyperbranched or dendritic structures, and decomposition temperatures below 400 ° C, preferably below 350 ° C Polymers and oligomers. Suitable polymers and oligomers include polyethers, polylactide, polycarbonate, poly[3-hydroxybutyrate], polymethacrylates, poly(methacrylic acid) copolymers, poly(methacrylic acid) Homopolymers and copolymers of poly(ethylene glycol), poly(lactic acid), poly(DL-lactide/glycolide), poly(propylene carbonate) and poly(ethylene carbonate). If present, the polymeric or oligomeric binder is less than 20% by weight of the ink, or less than 10% by weight, or less than 5% by weight, or less than 2% by weight, or less than 1% by weight.

Suitable surfactants include methoxy-, fluoryl-, alkyl-, and alkynyl-substituted surfactants. The choice is usually based on the nature of the coating and dispersion observed and the desired adhesion to the substrate. Suitable surfactants are well known to those skilled in the art and include commercially available BYK® (Byk Chemie), Zonyl® (DuPont), Triton® (Dow), Surfynol® (Air Products), and Dynol® (Air Products). Surfactant.

Suitable dopants include, inter alia, metal cations (eg, Na, Li, and K), Sb, Bi, Cl, and binary semiconductors. If present, the dopant is typically between one part per million and 10% by weight of the ink, preferably between 0.1 and 5% by weight. If present, the chalcogenide additive is typically between 0.1 wt% and 10 wt% of the ink.

Suitable dispersing agents include polyvinylpyrrolidone, polycarboxylates, polyphosphates, polyamines, polyethylene glycols, and peptides comprising cysteine and/or histidine residues.

Nanoparticle ink can be spin coated, spray coated, ink jet printed, dip coated, knife coated, flexographic/gravure printed, slot coated and dispensed onto any substrate, such as glass, coated with metal Or oxide glass, metal foil or plastic. The ink can be deposited on the preheated substrate to deposit a semiconductor layer. The deposition may be followed by a further annealing step to improve the electronic properties, crystallinity, and grain size of the absorber layer.

The semiconductor layer typically has a thickness of from 15 nm to 5 μm, preferably from 500 nm to 3.5 μm. The layer thickness depends on the coating technique used in each case and its parameters. In the case of spin coating, for example, the parameters are the rotational speed and the rotational duration. In the case of spraying, the thickness may increase with the spraying time. In the case of bar coating and blade coating, the thickness can be increased by repeating the deposition step. Coating the semiconductor ink onto the substrate by methods such as dispensing, spraying, bar coating, spin coating, slot coating, dispensing, knife coating, inkjet printing or flexographic/gravure printing to be similar to familiarity This is well known to those skilled in the art (see MAAegerter, M. Menning; Sol-Gel Technologies for Glass Producers and Users, Kluwer Academic Publishers, Dordrecht, Netherlands, 2004).

According to the invention, the substrate can be a rigid substrate such as a glass, ceramic, metal or plastic substrate, or a flexible substrate, in particular a plastic film or a metal foil. According to the present invention, it is preferred to use a substrate coated with a thin molybdenum layer, which serves as an electrode and is very effective for the performance of the solar cell.

In a preferred embodiment of the invention, an additional compound comprising S, Se and/or Te and no metal is added to the process of forming the semiconductor layer. The additional compound can be added by adding a compound to the semiconductor ink or, more preferably, during/after annealing the layer. The S/Se/Te source optionally added to the additional chalcogen element is preferably selected from molecular precursors such as organic compounds containing selenium or sulfur, hydrogen sulfide (H ₂ S) or hydrogen selenide (H ₂ Se). Or; selected from the group consisting of selenium, sulfur or antimony; more preferably selected from selenourea/thiourea or by exchanging hydrogen with other organic groups, thioacetamide or element S/Se/Te derivative. At this point, sulfur and selenium are preferred chalcogen elements. When elemental sulfur or selenium is added to the nanoparticulate ink, it is preferably dissolved or suspended as a powder in an amine (eg, hydrazine, ethylenediamine, ethanolamine, etc.), a phosphine compound (eg, tributyl). Addition in the form of a phosphine, trioctylphosphine, triphenylphosphine, etc.), an organic solvent (eg, an alcohol, DMF, DMSO, etc.), a solvent mixture as described above, or other suitable liquid carrier. When elemental sulfur or selenium is added to the process during/after annealing of the layer, it is preferably added as an elemental vapor preferably diluted in an inert carrier gas. Combined annealing and treatment with selenium/sulfur are referred to as selenization/sulfidation.

To further improve crystallinity and grain size, the layer can be subjected to high temperature treatment in an inert atmosphere or in a vacuum using a conventional high temperature furnace or rapid heat treatment (having lower heat cost and higher yield), that is, 350 ° C to 600 ° C. In addition, annealing may be performed in the presence of Se or S vapor to avoid loss of chalcogen elements from the film during high temperature processing and to promote grain growth and higher crystallinity. At this stage, chalcogen elements may also be interchanged, for example, sulfur may be exchanged with selenium or vice versa.

Post-deposition processing, for example by annealing, selenization or vulcanization of the semiconductor layer, provides additional improved device performance. When additional chalcogen elements (usually Se or S) are optionally supplied to the gas phase at elevated temperatures to maintain a stable chalcogen content, the grain size and grain range can be optimized at these temperatures, as appropriate Replace one chalcogen with another during operation. Typically, additional selenium is provided to the sulfur containing layer, which may replace sulfur in part or almost all. Therefore, in another preferred embodiment of the present invention, the method of preparing a metal chalcogenide semiconductor layer according to the present invention further comprises a selenization step and/or a vulcanization step and/or an annealing step as another step after depositing the semiconductor layer . Annealing can be performed in different time ranges from 10 ^-1 to 10 ⁴ seconds, including so-called rapid thermal annealing. Shorter intervals are suitable to keep the stoichiometry of the layer almost constant. The growth of elemental interchange and expanded single crystal grains may require longer intervals. In a preferred embodiment, the method of producing a chalcogenide layer in accordance with the present invention produces a single crystal layer after step c) of the process without any carbonaceous layer adjacent to the crystalline layer.

A typical photovoltaic device includes a substrate, a back contact layer (eg, molybdenum), an absorber layer (also referred to as a semiconductor layer), a buffer layer (also referred to as a second semiconductor layer), and a top contact layer. Photovoltaic cells can also include metal electrical contacts (such as wires or grids on the top contact layer) and an anti-reflective layer on the light-facing side of the cell to enhance light transmission into the semiconductor layer. Here, the semiconductor layer prepared according to the invention is supplied to the semiconductor together with the contact body in a conventional manner and becomes a desired semiconductor device. In the case of a photovoltaic device, the semiconductor layer is typically deposited on a molybdenum bottom electrode and a buffer layer (CdS), a transparent upper electrode made of, for example, ZnO or indium tin oxide (simplified to ITO), and a metal connected to the electrode The grid is placed over the semiconductor layer.

The buffer layer usually contains inorganic materials such as CdS, ZnS, zinc hydroxide, Zn(S, O, OH), cadmium zinc sulfide, In(OH) ₃ , In ₂ S ₃ , ZnSe, indium zinc selenide, indium selenide. , zinc magnesium oxide or n-type organic materials or a combination thereof. The layers of such materials may be deposited to a thickness of from 5 nm to 500 nm, preferably from 40 nm to 100 nm, by chemical bath deposition, atomic layer deposition, co-evaporation, sputtering or chemical surface deposition.

The top contact layer is typically a transparent conductive oxide such as zinc oxide, aluminum-doped zinc oxide, indium tin oxide or cadmium stannate. Suitable deposition techniques include sputtering, evaporation, chemical bath deposition, electroplating, chemical vapor deposition, physical vapor deposition, and atomic layer deposition. Alternatively, the top contact layer may comprise a transparent conductive polymeric layer, such as poly-3,4-ethylenedioxythiophene (PEDOT) doped with poly(styrene sulfonate) (PSS), which may include Standard method of coating, dip coating or spray coating. In some embodiments, the PEDOT is treated to remove acidic components to reduce the acid induced degradation potential of the photovoltaic cell assembly.

In one embodiment, the absorber layer is placed in a cadmium sulfide bath to deposit a CdS layer. Alternatively, CdS can be deposited on the metal chalcogenide layer by placing the coated substrate in a cadmium sulfide bath containing thiourea.

In one embodiment, the insulating zinc oxide (ZnO) is used instead of the CdS sputtering layer. Photovoltaic devices. In some embodiments, both the CdS and ZnO layers are present in the photovoltaic cell; in other embodiments, only one of CdS and ZnO is present.

In some embodiments, a layer of sodium compound (eg, NaF, Na ₂ S, or Na ₂ Se) is formed on and/or under the metal chalcogenide layer.

Additional means can be employed to optimize the performance of the absorber layer in the photovoltaic device. Selenization/sulfidation (see above), treatment with aqueous cyanide to remove traces of copper or copper sulfide, and thioacetamide/InCl ₃ cleaning for band gap optimization can be used for the semiconductor layer.

The photovoltaic device as described herein preferably has a power conversion efficiency of 8% or higher, more preferably 9% or higher, and still more preferably 12% or higher.

The following abbreviations are used above and below: PCE power conversion efficiency, equivalent to energy conversion efficiency η, FF fill factor, V _OC open circuit voltage, J _SC short circuit current density, NP nanoparticle

The following examples will illustrate the invention without limitation. The characteristics and composition of the materials used in a single instance can be applied to other materials not specifically mentioned in the scope of the patent application.

Instance

The following reagents were used to synthesize materials: 99.99% pure CuCl (Strem), 99.999% pure GaCl ₃ (Sigma), 99.999% pure InCl ₃ (Sigma), 99.999% pure Cu(O ₂ CCH ₃ ) ₂ (Sigma) 99.99% pure Zn(O ₂ CCH ₃ ) ₂ (Sigma), 99.99% pure SnCl ₂ (Sigma), 99.999% pure GeCl ₄ (Sigma), 95% pure NOBF ₄ (Sigma), 80 to 90 % purity of oleic acid (Acros). The reagent Et ₃ OBF ₄ (Mirvin reagent with a purity higher than 97%) and (C ₂ H ₅ ) ₄ N (BF ₄ ) (99% purity) were obtained from Signa and used without further purification.

CIGS nanoparticles were prepared according to the draft standard as described in the literature (Guo et al, Nano Letters, (2009) 9(8), 3060-3065). As described in the literature (Guo et al, J. Am. Chem. Soc., (2009) 131 (33), 11672-11673), this draft can be easily employed to synthesize Cu ₂ ZnSn(S, Se) ₄ nm. Particles. The elemental distribution of the chalcogenide is preferably weaker in copper (see, for example, Guo et al, J. Am. Chem. Soc. (2010), 132, 17384-6).

Example 1.a : Synthesis of Cu(In,Ga)S ₂ Nanoparticles

In a glove box, oleylamine (160 ml), CuCl (20.0 mmol, 1.98 g), GaCl ₃ (6.00 mmol, 1.06 g) and InCl ₃ (14 mmol, 3.01 g) were added to a 500 ml three-necked round bottom flask. The septum was sealed with a septum and transferred to a hood. The pipeline is flushed and a thermocouple and a condenser are attached under a nitrogen stream. The solution was heated to 130 ° C under vacuum and purged three times with nitrogen. Degas for 30 minutes under vacuum. Switch to nitrogen flow and raise the temperature to 225 °C. A sulfur solution (40.0 mmol, 1.23 g) was quickly injected into oleylamine (20 ml) using a plastic syringe and the temperature was allowed to rise rapidly back to 225 °C. Reaction for 30 minutes. After 30 minutes, the heated mantel was removed and cooled to 40 °C. An equal amount of EtOH was added to precipitate the nanoparticles and centrifuged at 5.000 rpm for 5 minutes. Redispersed in toluene. Repeat the washing procedure.

The resulting nanoparticles were surface passivated using oleylamine used during the synthesis. The X-ray diffraction data indicates that the nanoparticle is a pure brass mineral phase free of impurities. Depending on the synthesis, the average particle size is in the range of 7 nm to 10 nm. With a yield of more than 90%, the stoichiometry can be well controlled via the precursor ratio to achieve Cu _x In _y Ga _z S ₂ particles, where x can vary from 0.8 to 1.0 and y can vary from 0.2 to 0.6. The thermogravimetric data indicates that the oleylamine ligand comprises from 15% to 20% of the total mass of the granule.

Example 1.b : Synthesis of Cu ₂ ZnSnS ₄ Nanoparticles

In the glove box, oleylamine (160 ml), Cu(acac) ₂ (7 mmol, 1.83 g), SnCl ₂ (4.0 mmol, 0.758 g) and Zn(OAc) ₂ (4.0 mmol, 0.734 g) were added to 500 ml. In a three-necked round bottom flask. The septum was sealed with a septum and transferred to a hood. The pipeline is flushed and a thermocouple and a condenser are attached under a nitrogen stream. The solution was heated to 130 ° C under vacuum and purged three times with nitrogen. Degas for 30 minutes under vacuum. Switch to nitrogen flow and raise the temperature to 225 °C. A sulfur solution (14.0 mmol, 0.449 g in 20 ml of oleylamine) was quickly injected using a plastic syringe and the temperature was allowed to rise rapidly back to 225 °C. Reaction for 30 minutes. After 30 minutes, the heated mantel was removed and cooled to 40 °C. An equal amount of EtOH was added to precipitate the nanoparticles and centrifuged at 5.000 rpm for 5 minutes. Redispersed in toluene. Repeat the washing procedure.

The resulting nanoparticles were surface passivated using oleylamine used during the synthesis. X-ray diffraction data indicates that the nanoparticles are pure zinc-tin-tin phases without impurities. Depending on the synthesis, the average particle size is in the range of 7 nm to 10 nm. The yield is over 90%, and the stoichiometry can be well controlled via the precursor ratio to achieve Cu _x Zn _y Sn _z S ₄ particles, wherein depending on the ratio of the reactants, x can vary from 1.4 to 2.2, and y can be from 0.6. It varies by 1.5 and z can vary from 0.7 to 1.5. The thermogravimetric data indicates that the oleylamine ligand comprises from 15% to 20% of the total mass of the granule.

Example 2 : Ligand exchange with tetrafluoroborate General procedure for ligand exchange on metal chalcogenide nanoparticles: version 1:

10 ml of chalcogenide nanoparticles (or other chalcogenide NP) according to Example 1 in hexane (20 mg/ml) was added to 10 ml of 0.1 M triethyloxonium tetrafluoroborate in a nitrogen-filled glove box. In a solution of hydrazine (Et ₃ OBF ₄ ), the solution was dissolved in 10 ml of MeCN to form a biphasic solution. After stirring for 30 minutes, the naked nanocrystals will precipitate during this exchange. After 30 minutes, twice the amount of ethanol was added and centrifuged at 3500 rpm for 4 minutes. The pellets were dispersed in DMSO and the NP was washed three times with hexane to remove residual OLA.

Version 2:

10 ml of chalcogenide nanoparticles in toluene (20 mg/ml) were added to 10 ml of a 0.3 M solution of triethyloxonium tetrafluoroborate (Et ₃ OBF ₄ ) in a nitrogen-filled glove box. Dissolved in MeCN in a 250 ml beaker. Stir for 30 minutes or until all NP has settled to the bottom of the beaker. 10 ml of DMSO was added to dissolve the NP. Transfer the solution to two centrifuge tubes and move to a fume hood. 20 ml of acetone was added to each centrifuge tube and centrifuged at 3500 rpm for 4 minutes. The supernatant was discarded and the pellet was dried using a nitrogen sparge to remove any remaining acetone. The pellets were dispersed in DMSO (a stir bar was added to the centrifuge tube to break up the pellets). Transfer the ink to a glass beaker. 10 ml of hexane was added to the NP-BF ₄ ink in DMSO which should form a layer at the top. Use a pipette to remove the hexane layer. Repeat hexane washing twice or more. The remaining hexane at the surface of the DMSO was very gently evaporated using a nitrogen sparge. NP was once again precipitated using acetone to separate and redissolve in another liquid medium, as appropriate.

The resulting dispersion remained stable for several months at a concentration of more than about 50 mg/ml.

Example 2.a

In a particular example, the Cu(In,Ga)S ₂ nanoparticle according to Example 1.a is a ligand exchanged using the general procedure version 2.

Example 2.b

In another specific example, the Cu ₂ ZnSnS ₄ nanoparticle according to Example 1.b is a ligand exchanged using the general procedure version 2.

In each case, thermogravimetric analysis indicated that almost all of the oleylamine was removed from the particles. No decrease in mass was observed in the temperature range of 200 ° C to 500 ° C, while the native oleyl amine terminated nanoparticles lost about 20% by weight when heated.

Example 3 : Nanoparticle ink formulation

The isolated ligand exchange particles of Examples 2.a and 2.b are redissolved in DMSO or, optionally, in a compatible solvent suitable for the chosen deposition method. The ink is stored at ambient concentrations up to 150 mg/ml solid phase load. The dispersion can remain stable for several days at room temperature without degradation or aggregation.

Example 4 : Formation of an absorbent layer

For spraying, the nanoparticle ink is diluted in DMSO to a concentration of 10 mg/ml to 30 mg/ml of nanoparticle. Spray using an automatic sonic wave treatment sprayer. Do not add additives to the ink. Spraying ink onto the substrate at a temperature of 175 ° C to 350 ° C (via Mo Coated soda lime glass) to ensure solvent evaporation and control film morphology. The amount of passing depends on the solids loading of the ink. Achieving the desired film thickness may require a pass of 4 to 20 passes.

For blade coating, a thicker ink is required. The concentration here can be as high as 100 mg/ml, preferably about 50 mg/ml. Depending on the solvent selected, blade coating can be carried out using a substrate heated to 70 ° C to 200 ° C. After 3 to 4 layers have been deposited, the substrate is annealed at a higher temperature (250 ° C to 450 ° C) to ensure complete solvent removal.

Example 5 : Fabrication of photovoltaic devices Example 5.a : Photovoltaic device (CIGS absorber) and comparison with conventional devices

The dried Cu(In,Ga)S ₂ film on the substrate with the Mo back contact of Example 4 (sprayed) was transferred to a graphite box with a lid (non-hermetic) with few selenium bombs. The graphite box assembly was inserted into a quartz tube and the tube was purged three times with argon before being inserted into the preheated tube furnace. The tube furnace was maintained at 550 ° C and selenized for 30 minutes. During the selenization process, the selenium bomb produces selenium vapor over the substrate within the attached graphite box and assists in promoting grain growth and higher crystallinity within the film. In addition, sulfur is more or less replaced by selenium.

To complete the photovoltaic device, a 60 nm thick CdS layer was deposited by a solution process (MAContreras et al., Thin Solid Films 2002, 403-404, 204-211). ZnO (50 nm) and ITO (300 nm) films were sequentially deposited by RF sputtering. Subsequently, silver grid lines were sputtered by DC sputtering. The total device area is 24mm ² .

Figure 2 shows the efficacy of a photovoltaic device with IV characteristics in dark and AM 1.5 light conditions. The device characteristics were as follows: PCE = 12.05%, FF = 61.9%, V _oc = 0.620 V, J _sc = 32.36 mA / cm ² .

The above device characteristics are based on uncorrected light blocked by the silver grid region (about 5%) value.

Compare devices:

Comparative devices were fabricated in a similar manner to the native CIGS nanoparticles according to Example 1, but were not exchanged with the ligands according to Example 2.

The device characteristics were as follows: PCE = 6.71%, FF = 57.8%, V _oc = 0.502 V, J _sc = 23.11 mA / cm ² .

The power conversion efficiency (PCE = 12.05%) of the device of the present invention and other characteristic values are much higher than the value of the device manufactured by standard nanoparticle (PCE = 6.71%).

Example 5.b : Photovoltaic device (CZTS absorber)

Similar to the device of Example 5.a, the Cu ₂ ZnSnS ₄ nanoparticle of Example 2.b was formulated into an ink, sprayed onto Mo-coated soda lime glass, selenized and equipped with a buffer layer and a transparent conductor layer.

The device characteristics were as follows: PCE = 9.53%, FF = 55.3%, V _oc = 0.396 V, J _sc = 43.4 mA/cm ² .

The scope of the following patent application further discloses combinations of embodiments of the invention with variations of the invention.

(1) ‧ ‧ absorbing layer

(2) ‧ ‧ molybdenum back contacts

(3) ‧‧‧Substrate glass

Claims (15)

A ternary, quaternary or polybasic metal chalcogenide nanoparticle (NP) stabilized by an inorganic ion comprising a tetrahalo borate, a hexahalophosphate or a hexahalogen ruthenate. The NP of claim 1, wherein the NPs are not stabilized by the organic ligand. The NP of claim 1 or 2, wherein the NPs comprise copper. An NP as claimed in any one of items 1 to 3, wherein the NPs comprise one, two or more of the elements In, Ga, Zn, Sn and Ge. An NP as claimed in any one of items 1 to 4, wherein the NPs comprise one or both of the elements S and Se. An NP according to any one of claims 1 to 5, wherein the NP comprises one of 95 mol% or more of the following element combinations: (Cu, In, S), (Cu, Ga, S), (Cu, In, Se), (Cu, Ga, Se), (Cu, In, S, Se), (Cu, Ga, S, Se), (Cu, In, Ga, S), (Cu, In, Ga, Se), (Cu, In, Ga, S, Se), (Cu, Sn, S), (Cu, Sn, Se), (Cu, Sn, S, Se), (Cu, Zn, Sn, S) , (Cu, Zn, Sn, Se), (Cu, Zn, Sn, S, Se), (Cu, Zn, Ge, S), (Cu, Zn, Ge, Se), (Cu, Zn, Ge, S, Se), (Cu, Zn, Sn, Ge, S), (Cu, Zn, Sn, Ge, Se), (Cu, Zn, Sn, Ge, S, Se). The NP of any one of claims 1 to 6, wherein the NPs are semiconductors of I-III-VI ₂ , I ₂ -II-IV-VI ₄ , I ₂ -IV-VI ₃ or a mixed type. A method for producing ternary, quaternary or multi-component metal chalcogenide nanoparticles, characterized in that the chalcogenide is treated with an inorganic ion comprising a tetrahalo borate, a hexahalophosphate or a hexahalogen ruthenate Nanoparticles. The method of claim 8, wherein a liquid comprising a combination of a trialkyloxonium ion and an inorganic ion comprising a tetrahaloborate, a hexahalophosphate or a hexahalogenated ruthenate is used. The nanoparticle is treated by a bulk medium, wherein the alkyl group in the trialkyloxonium ion independently represents a linear or branched alkyl group having 1 to 15 carbon atoms. A nanoparticle ink comprising a liquid medium and a nanoparticle dissolved in the liquid medium, wherein the nanoparticle comprises the nanoparticle of any one of claims 1 to 7. The nanoparticle ink of claim 10, wherein it comprises: 0.5 wt% to 80 wt% of metal chalcogenide nanoparticles; and 20 wt% to 98 wt% of a liquid medium. The nanoparticle ink of claim 10 or 11, wherein the liquid medium comprises one or more liquids selected from the group consisting of aromatic compounds, DMF, DMSO, water, heteroaromatics, alkanes, nitriles Classes, ethers, ketones, esters, acetates, guanamines, amines, thiols, carboxylic acids, organic halides and alcohols. A method for producing a chalcogenide layer comprising depositing a ternary, quaternary or multi-component metal chalcogen by the following steps, in that order, or two or all three of the steps simultaneously a nanoparticle according to any one of claims 1 to 7: a) depositing a nanoparticle ink comprising the nanoparticle and a liquid medium on a substrate Upper b) removes the liquid medium, and c) optionally anneals or fuses the nanoparticles in the presence of additional selenium or sulfur sources, as appropriate. The method of claim 13, wherein the obtained chalcogenide layer is a crystalline semiconductor layer. A method for producing a photovoltaic device comprising an absorber layer and an electrode, characterized in that the absorber layer is a chalcogenide layer and is obtained by the method of claim 13 or 14.