SEMICONDUCTOR DOPING PROCESS

FIELD OF THE INVENTION

The present invention relates a process for producing a doped organic semiconductor, a process for producing a layer comprising a doped organic semiconductor and a process for producing a semiconductor device. The invention also relates to a composition useful in the described process.

The project leading to this application has received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No. 653184. The work leading to this invention has received funding from the European Union Seventh Framework (FP7 2007-2013) under grant agreement No. 604032.

BACKGROUND OF THE INVENTION

Organic semiconductors are an important component in semiconductor devices such as light- emitting diodes, organic field-effect transistors and photovoltaic devices. High quality polycrystalline organic thin films are often used as active materials. Organic semiconductor materials typically have insufficient intrinsic conductivity and often require doping to enhance their charge transporting properties. For hole transporting materials (HTMs), numerous p-type dopants are known. These p-type dopants have been used to improve the device performances. Examples of p-type dopants include elemental species, covalent solids, Bronsted acids, Lewis acids and small molecules (Liissem et al, Chem. Rev., 2016, 116 (22), pp 13714-13751; Salzmann et al, Acc. Chem. Res. 2016, 49, 370-378).

However, there are several problems associated with known techniques for doping organic semiconductors and hole transport materials. Often, p-type dopant materials used are expensive due to their complex chemical structure. In certain cases, the dopants are susceptible to the presence of moisture and the ambient atmosphere which means the extent of doping can degrade over time. For some dopants, by-products are formed with any metals or oxygen that may be present which then have an impact on the stability of devices containing the doped semiconductor. Some p-type dopants can have low efficiency when doping organic semiconductors, i .e. only a fraction of the added dopant molecules lead to the oxidization of the organic semiconductor. Furthermore, it has been observed that partial charge transfer via the formation of ground state charge-transfer complexes can occur for individual dopant-host systems meaning that substantial amounts of dopant are required in order to increase the free carrier concentration, which can have adverse effects on the morphology of the host organic semiconductor. The dopant molecules become the counter ions after the doping process is complete, which remain in the film after the doping process. This therefore makes it difficult to optimize their role as both dopant and benign counterion separately. Of critical importance, due to the Coulombic interaction when a dopant molecule transfers charge to the organic semiconductor host, is that the remaining anion (or cation in the case of n-type doping) can act as a trap for charge of opposite polarity, which can result in a very weak influence upon increasing the density of free carriers and conductivity with increasing dopant concentration.

A further challenge, especially for solution processed doped semiconductors, is the ability to achieve graded doping. A graded doped film is a doped film of a semiconducting host wherein the concentration of dopant ions is higher on one side of the film that the other. For vapour deposited organic semiconductors, this is simply achieved by co-evaporation of different fractions of host and dopant molecules as the film is built up. However, for solution processed materials, it is not possible to simply coat two subsequent layers on top of each other, due to the common solubility of the host, without introducing more complicated chemistries such as molecular cross linking to make the underlying film insoluble. The later may introduce further adverse affects upon the semiconducting nature of the host. There is therefore a need to develop a new route to achieving graded doping in solution processed organic semiconductors.

In general, there is a need to develop an inexpensive yet effective dopant system for the doping of organic semiconductors.

SUMMARY OF THE INVENTION

The inventors have surprising found that a doping system for organic semiconductors using a compound comprising a sulfoxide group and an activator has several advantages over known doping processes. The advantages of this system over the existing methods and materials include: (a) the availability of flexible doping methods (i.e. either solution or vapour phase may be used); (b) a clean chemical reaction which does not form undesirable by-products; (c) the ability to dope a wide range of different organic semiconductors; (d) low cost doping materials; (e) easy handling of the process; and (f) reduced toxicity. Furthermore, it is possible to separately select the counter ion which is introduced into the doped

semiconductor allowing other properties of the semiconductor to be tuned.

The present invention accordingly provides a process for producing a doped organic semiconductor, which process comprises treating an organic semiconductor with (i) a compound comprising a sulfoxide group and (ii) an activator.

The invention also provides a process for producing a layer comprising a doped organic semiconductor, which process comprises producing the doped organic semiconductor by a process according to the invention.

The invention further provides process for producing a semiconductor device, which process comprises producing a doped organic semiconductor by a process according to the invention.

Also provided by the invention is a composition comprising: (i) a compound comprising a sulfoxide group; (ii) an acid of formula HX, where X is an anion; and (iii) an organic semiconductor.

Further provided by the invention is use of a composition comprising (i) a compound comprising a sulfoxide group and (ii) an activator to dope an organic semiconductor.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 shows: (a) the chemical structure of spiro-OMeTAD, DMSO and HBr; (b) UV-vis absorption spectra of a spiro-OMeTAD solution in chlorobenzene with the different dopant concentration (top) and UV-vis absorption spectra of the spiro-OMeTAD thin-film with the different exposure time (bottom) [absorption peaks at 370 (1), 502 (2) and 690 (3) nm are related to the oxidation state of spiro-OMeTAD (1 : spiro-OMeTAD, 2: spiro-OMeTAD ⁺¹ and 3 : spiro-OMeTAD ⁺² )]; and (c) the conductivity of a spiro-OMeTAD thin-film prepared by solution phase doping (top) and the vapour phase doping (bottom).

Figure 2 shows: (a) UV-vis absorption spectra of several organic hole-transporting materials in chlorobenzene before (dotted line) and after (solid line) doping.

Figure 3 shows: J-V curves of planar devices with several organic hole-transporting materials before (dotted line) and after (solid line) doping on OFET chips with bottom-gate and bottom-contact geometry. Figure 4 shows: (a) the FTIR spectra of DMSO in Chlorobenzene (CB), DMSO after exposure to the FIBr vapour and then after the addition of spiro-OMeTAD (Absorption peak for DMSO:HBr complex decreases after the addition of Spiro-OMeTAD); (b) formation of H2O during the doping process as the water peak increases; and (c) a schematic of the mechanism of the doping process.

Figure 5 shows: Ellipsometry data ! and A of an un-doped MeO-TPD thin film on top of a heavily doped MeO-TPD thin film under thermal stress (at 50 °C for 100 h).

Figure 6 shows: (a) UV-vis absorption spectra of a spiro-OMeTAD solution in chlorobenzene doped with DMSO-CSA and DMSO-HBr-CSA; and (b) conductivity of DMSO-HBr-CSA doped Spiro-OMeTAD

Figure 7 shows: thermal tolerance test by conductivity measurement of DMSO-HBr and DMSO-HBr-CSA doped in MeO-TPD

Figure 8 shows: (a) Schematic of the perovskite photovoltaic device; (b) current density- voltage curves of the solar cells with different dopants and counter ions; and (c) stabilized power output (SPO) for the respective perovskite solar cells at the fixed maximum power point voltage.

Figure 9 shows: trace of stabilized power output (SPO) of the perovskite solar cells with doped spiro-OMeTAD by 0 ₂ , DMSO-HBr and DMSO-HBr-CSA.

Figure 10 shows: (a) Schematic of the organic light emitting device (OLED); (b) current- voltage curves of the OLEDs; and (c) luminescence-voltage curve for the OLEDs with different dopants and counter ions.

Figure 11 shows: (a) Evolution of absorption spectra of the spiro-OMeTAD solution in chlorobenzene as a function of added DMSO-FIBr adduct. (b) Correlation plots for the consumption of Spiro-OMeTAD against the total amount of DMSO for the estimation of doping efficiency determined by UV-vis absorption Straight line shows the fitting result of plots by a linear fit (y=ax) where a indicates the doping efficiency.

Figure 12 shows: Evolution of UV-vis absorption spectrum of a spiro-OMeTAD thin film with the increasing exposure time in the DMSO and HBr vapour phase doping. Figure 13 shows: UV-vis absorption spectra of Spiro-OMeTAD solutions with only DMSO, HBr or the DMSO-HBr adduct.

Figure 14 shows: UV-vis absorption spectra of Spiro-OMeTAD solution doped with DMSO- HBr-CSA and DMSO-CSA.

Figure 15 shows: Evolution of UV-vis spectrum of spiro-OMeTAD doped with DMSO-HBr- CSA by the solution process, (a) Absorption spectra of spiro-OMeTAD solution in chlorobenzene with an increase in concentration of DMSO-HBr. (b) Correlation plots for the consumption of Spiro-OMeTAD against a total amount of DMSO for the estimation of doping efficiency determined by UV-vis absorption. Straight line shows fitting result of plots by a liner fit (y=ax) where a indicates the doping efficiency.

Figure 16 shows: Gradient doping in a hole-only device, (a) Schematic of the hole-only device with doping at Poly-TPD: Au interface (left) and at FTO: Poly-TPD interface (right). Current density-voltage curves of the hole-only devices (b) doping at Poly-TPD: Au interface (c) at FTO: Poly-TPD interface. Accepter density (NA) as function of distance from the FTO side, (d) homogeneous doping, (e) doping at the Poly-TPD: Au interface and (f) doping at the FTO: Poly-TPD interface. The dashed line in panel (d), (e) and (f) represents NA values of the un-doped sample.

Figure 17 shows: Homogeneous doping in hole-only device, (a) Current density-voltage curves and (b) accepter density-approximated depth profile of the hole-only devices before and after DMSO-HBr-CSA doping (0.3 mol% of A.F.A.).

Figure 18 shows: Usage of adduct-based dopants in optoelectronic devices, (a) Schematic of the perovskite photovoltaic device, (b) current density-voltage curves of the solar cells with different dopants and counterions. (c) stabilized power output (SPO) for the respective perovskite solar cells at a fixed maximum power point voltage, (d) Schematic of the organic light emitting device (OLED). (e) Current density-voltage curves of the OLEDs and (f) luminescence-voltage curve for the OLEDs with different dopants and counterions Inset: Optical image of a forward biased OLED in which the HTM is doped with DMSO-HBr-CSA.

Figure 19 shows: PV performance parameters of the perovskite solar cells doped spiro- OMeTAD with different doping methods. Dependence of device performance on the dopants. The performance parameters are extracted from forward bias to short-circuit. Current density- voltage curves of perovskite solar cells were measured under simulated AMI .5 sunlight at 100 mW cm ^"2 . The data are represented as a standard boxplot where the box range is defined by the standard deviation (n>30) Ninety percent of all data points fall within the upper and lower whisker. 1 : doped by Li-TFSI, 2: doped by Li-TFSI-oxygen in air for 12-15 hrs, 3 : doped with Co[III]TFSI, 4: doped with DMSO-HBr by the solution process (1 mol% of adduct forming agent (A.F.A)), 5 : doped with DMSO-HBr by the vapor process (30 sec reaction), 6: doped with DMSO-HBr by the simultaneous of solution (0.8 mol% of A.F.A) and vapor process (20 sec reaction), and 7: doped with DMSO-HBr-CSA by the solution process (1 mol% of A.F.A).

Figure 20 shows: External quantum efficiency of the perovskite solar cells. External quantum efficiency (EQE) spectra of perovskite solar cells with different dopants for the spiro- OMeTAD layer where the HTM layers are doped with different methods.

Figure 21 shows: Hysteresis property of the perovskite solar cell. Both forward-bias to short- circuit (FB-SC) and short-circuit to forward-bias (SC-FB) current density-voltage sweeps of the devices with different doping methods where O2 and DMSO-adduct doped spiro- OMeTAD films.

Figure 22 shows: PV performance of the perovskite solar cells fabricated with polymer hole- transporting material. The JV curves of the p-i-n structure device. FAo.83Cso.i7Pb(Io.8Bro.2)3 was the photo-absorber and was sandwiched between poly-TPD as the hole transport layer and a PCBM layer as the n-type electron transport layer.

DETAILED DESCRIPTION OF THE INVENTION

The term "semiconductor" or "semiconducting material", as used herein, refers to a material with electrical conductivity intermediate in magnitude between that of a conductor and a dielectric. A semiconductor may be a negative (n)-type semiconductor, a positive (p)-type semiconductor or an intrinsic (i) semiconductor. A semiconductor may have a band gap of from 0 5 to 3.5 eV, for instance from 0.5 to 2.5 eV or from 1.0 to 2.0 eV (when measured at 300 K). An organic semiconductor is a semiconductor which comprises an organic compound (for instance a compound comprising a hydrogen atom bonded to a carbon atom).

The term "semiconductor device", as used herein, refers to a device comprising a functional component which comprises a semiconducting material. This term may be understood to be synonymous with the term "semiconducting device". Examples of semiconductor devices include a photovoltaic device, a solar cell, a photo detector, a photodiode, a photosensor, a chromogenic device, a transistor, a light-sensitive transistor, a phototransistor, a solid state triode, a battery, a battery electrode, a capacitor, a super-capacitor, a light-emitting device and a light-emitting diode. The term "optoelectronic device", as used herein, refers to devices which source, control, detect or emit light. Light is understood to include any

electromagnetic radiation, but often refers to light in the visible range. Examples of optoelectronic devices include photovoltaic devices, photodiodes (including solar cells), phototransistors, photomultipliers, photoresistors, light emitting devices, light emitting diodes and charge injection lasers.

The term "n-type region", as used herein, refers to a region of one or more electron- transporting (i.e. n-type) materials. Similarly, the term "n-type layer" refers to a layer of an electron-transporting (i.e. an n-type) material. An electron-transporting (i.e. an n-type) material could, for instance, be a single electron-transporting compound or elemental material. An electron-transporting compound or elemental material may be undoped or doped with one or more dopant elements.

The term "p-type region", as used herein, refers to a region of one or more hole-transporting (i.e. p-type) materials. Similarly, the term "p-type layer" refers to a layer of a hole- transporting (i.e. a p-type) material. A hole-transporting (i.e. a p-type) material could be a single hole-transporting compound or elemental material, or a mixture of two or more hole- transporting compounds or elemental materials. A hole-transporting compound or elemental material may be undoped or doped with one or more dopant elements.

The term "alkyl", as used herein, refers to a linear or branched chain saturated hydrocarbon radical. An alkyl group may be a Ci-20 alkyl group, a CI-H alkyl group, a Ci-10 alkyl group, a Ci-6 alkyl group or a C1-4 alkyl group. Examples of a Ci-10 alkyl group are methyl, ethyl, propyl, butyl, pentyl, hexyl, heptyl, octyl, nonyl or decyl Examples of Ci-6 alkyl groups are methyl, ethyl, propyl, butyl, pentyl or hexyl. Examples of C1-4 alkyl groups are methyl, ethyl, i-propyl, n-propyl, t-butyl, s-butyl or n-butyl. If the term "alkyl" is used without a prefix specifying the number of carbons anywhere herein, it has from 1 to 6 carbons (and this also applies to any other organic group referred to herein). The term "aryl", as used herein, refers to a monocyclic, bicyclic or polycyclic aromatic ring which contains from 6 to 14 carbon atoms, typically from 6 to 10 carbon atoms, in the ring portion. Examples include phenyl, naphthyl, indenyl, indanyl, anthrecenyl and pyrenyl groups. The term "aryl group", as used herein, includes heteroaryl groups. The term

"heteroaryl", as used herein, refers to monocyclic or bicyclic heteroaromatic rings which typically contains from six to ten atoms in the ring portion including one or more

heteroatoms. A heteroaryl group is generally a 5- or 6-membered ring, containing at least one heteroatom selected from O, S, N, P, Se and Si. It may contain, for example, one, two or three heteroatoms. Examples of heteroaryl groups include pyridyl, pyrazinyl, pyrimidinyl, pyridazinyl, furanyl, thienyl, pyrazolidinyl, pyrrolyl, oxazolyl, oxadiazolyl, isoxazolyl, thiadiazolyl, thiazolyl, isothiazolyl, imidazolyl, pyrazolyl, quinolyl and isoquinolyl.

The term "substituted", as used herein in the context of substituted organic groups, refers to an organic group which bears one or more substituents selected from Ci-io alkyl, aryl (as defined herein), cyano, amino, nitro, Ci-io alkylamino, di(Ci-io)alkylamino, arylamino, diarylamino, aryl(Ci-io)alkylamino, amido, acylamido, hydroxy, oxo, halo, carboxy, ester, acyl, acyloxy, Ci-io alkoxy, aryloxy, halo(Ci-io)alkyl, sulfonic acid, thiol, Ci-io alkylthio, arylthio, sulfonyl, phosphoric acid, phosphate ester, phosphonic acid and phosphonate ester. Examples of substituted alkyl groups include haloalkyl, perhaloalkyl, hydroxyalkyl, aminoalkyl, alkoxyalkyl and alkaryl groups. When a group is substituted, it may bear 1 , 2 or 3 substituents. For instance, a substituted group may have 1 or 2 substitutents.

Process

The invention provides a process for producing a doped organic semiconductor, which process comprises treating an organic semiconductor with (i) a compound comprising a sulfoxide group and (ii) an activator. The doped organic semiconductor is typically a p- doped organic semiconductor.

Typically, treating the organic semiconductor with (i) and (ii) comprises exposing the organic semiconductor to (i) and (ii) (either simultaneously or sequentially). The organic

semiconductor may be contacted with (i) the compound comprising a sulfoxide group and (ii) the activator. For instance, the organic semiconductor could be mixed with (i) and (ii) in solution. Alternatively, the organic semiconductor could be exposed to a vapour comprising (i) and a vapour comprising (ii). Typically, the activator is a compound which can form an adduct with the compound comprising a sulfoxide group. The activator may be an inorganic acid, a carboxylic acid, an acid anhydride, an acyl halide or a metal organic framework.

Typically, the activator is an acid. The acid may be a strong acid (for instance having a pKa of less than -2.0). The acid may be a compound of formula HX, wherein X is an anion selected from Br , CI , I , F , CH ₃ COO , HCOO , CF3COO , (CF ₃ S0 ₂ ) ₂ N , C10H14SO4 (camphorsulfonate), HSO4 , NC ^" and H2PO4 .

The activator is typically an acid which is selected from HBr, HC1 and HI. Preferably the acid is Ffflr.

The compound comprising a sulfoxide group is typically an organic compound comprising a sulfoxide group. The compound comprising a sulfoxide group is a typically a compound of formula R2S=0, wherein each R group is independently a substituted or unsubstituted Ci-20 alkyl group or a substituted or unsubstituted aryl group, and wherein the two R groups are optionally j oined together to form a ring.

A sulfoxide group may be represented either as >S=0 or >S ⁺ -0 . Often, each R group is an unsubstituted Ci-6 alkyl group which may optionally be joined together to form a ring. Each R group is typically independently selected from unsubstituted C1-3 alkyl and unsubstituted aryl. For instance, R may be methyl, ethyl, propyl or phenyl.

Examples of compounds comprising a sulfoxide group include dimethylsulfoxide, ethylmethylsulfoxide, diethylsulfoxide, dipropylsulfoxide, methylphenylsulfoxide, methyltolylsulfoxide, diphenylsulfoxide, ditolylsulfoxide and dibenzylsulfoxide. Preferably, the compound comprising a sulfoxide group is dimethylsulfoxide.

Typically, the process comprises treating the organic semiconductor with a composition comprising: (i) the compound comprising a sulfoxide group and (ii) the activator. The composition may be solid, liquid or gaseous composition. For instance, the composition may be a solution comprising (i) the compound comprising a sulfoxide group and (ii) the activator. The compound comprising a sulfoxide group may act as a solvent in the composition. The process typically comprises treating the organic semiconductor with an adduct produced from (i) the compound comprising a sulfoxide group and (ii) the activator. For instance, the composition comprising (i) the compound comprising a sulfoxide group and (ii) the activator typically comprises an adduct of the compound comprising a sulfoxide group and the activator. An adduct is a molecular species formed by the direct addition of two compounds. For instance, a lone pair on one compound may interact with the empty orbital (i.e. lone pair acceptor) on another compound to form an adduct. For instance, H3 and BF3 may form an adduct which may be described as ¾N:BF3 or H3N→BF3. An adduct between a sulfoxide compound and an activator which is an acid HX may take the form R2S=0:H-X (which may also be displayed as R2S=0→H-X).

The adduct between the compound comprising a sulfoxide and an activator which is a compound of formula HX is typically an adduct of formula R2S=0:HX or R2S=0:(HX)2. Often, the adduct of the compound comprising a sulfoxide group and the acid is the adduct DMSO:HBr or DMSO:(HBr) ₂ .

The process of the invention typically comprises contacting: (a) a first composition comprising the compound comprising a sulfoxide group and the activator; and (b) a second composition comprising the organic semiconductor. The process may comprise solution processing. The first and second compounds may be solutions. For instance, the process may comprise mixing: (a) a first solution comprising a first solvent, the compound comprising a sulfoxide group and the activator; and (b) a second solution comprising a second solvent and the organic semiconductor; and producing a composition comprising the first solvent, the second solvent and the doped organic semiconductor. The compound comprising a sulfoxide may be a liquid and may act as a solvent in the first composition.

The first and second solvents may be any suitable solvents (for instance solvents in which the other components are soluble). Each solvent may be a polar solvent or a non-polar solvent. Each solvent is typically a polar solvent. For instance the first and second solvents may be independently selected from water, an alcohol (such as methanol or ethanol),

dichloromethane, chlorobenzene, acetonitrile, or dimethylformamide (DMF). For instance, the first solvent may be water and the second solvent may be dichloromethane or chlorobenzene. The first or second solvent may comprise 2-ethoxyethanol (ethylene glycol). In a solution-based process, any additional solvents are typically removed after mixing of (a) the first solution comprising a first solvent, the compound comprising a sulfoxide group and the activator; and (b) the second solution comprising a second solvent and the organic semiconductor. The solvent is typically removed to produce a solid composition comprising the doped organic semiconductor (for instance a layer comprising the doped organic semiconductor). The solvent may be removed by heating or by allowing the solvent to evaporate. The solid composition comprising the doped organic semiconductor may be subsequently annealed, for instance to remove any additional solvent.

The process may alternatively comprise vapour processing. For instance, the process may comprise exposing the organic semiconductor to a vapour comprising: (i) the compound comprising a sulfoxide group and (ii) the activator. The process may comprise (a) disposing a layer of the organic material on a substrate and (b) exposing the layer of the organic semiconductor to a vapour comprising: (i) the compound comprising a sulfoxide group and (ii) the activator.

The organic semiconductor may be exposed, simultaneously or sequentially, to a first vapour comprising the compound comprising a sulfoxide group and a second vapour comprising the activator. For instance, the organic semiconductor may be placed in a low pressure chamber with a source of the compound comprising the sulfoxide group and a source of the activator. The source of the compound comprising the sulfoxide group may be heated. The source of the activator may be heated, or may be left at ambient temperature (for instance from 15 °C to 30°C). The organic semiconductor is typically exposed to the vapour comprising (i) the compound comprising a sulfoxide group and (ii) the activator for from 10 seconds to 1 hour.

The molar ratio of the compound comprising a sulfoxide group and the activator is typically from 1.0:5.0 to 5.0: 1.0 (sulfoxide compound:activator). For instance, the molar ratio of the compound comprising a sulfoxide group and the activator may be from 1.0: 1.0 to 1.0:3.0. In some cases, the molar ratio of the compound comprising a sulfoxide group and the activator is about 1 :2. The molar ratio of the adduct of the sulfoxide compound and the activator and the organic semiconductor is typically from 1.0:50.0 to 5.0: 1.0 (adduct: organic

semiconductor), for instance from 1.0: 10.0 to 2.0: 1.0. In the process of the invention, the doping concentration may be from 1 to 90 mol%, for instance from 5 to 70 mol% or from 10 to 30 mol%. The concentration of the adduct in a composition (for instance the combined first and second compositions) comprising one or more solvents with which an organic semiconductor is treated is typically from 0.01 M to 10 M, for instance from 2 0 M to 7.0 M.

The process may comprise treating the organic semiconductor with (i) the compound comprising a sulfoxide group and (ii) the activator sequentially. For instance, the process may comprise: (a) disposing on a substrate a composition (for example a solution) comprising the organic semiconductor and the compound comprising a sulfoxide group to form a layer comprising the organic semiconductor and the compound comprising a sulfoxide group and (b) treating the layer with the activator. Treating the layer with the activator may comprise: exposing the layer to a vapour comprising the activator; or disposing on the layer a composition comprising the activator. The process may comprise exposing a layer comprising the organic semiconductor and the compound comprising a sulfoxide group to a vapour comprising the activator.

The process of the invention can dope a wide range of organic semiconductors. Typically, the organic semiconductor comprises a compound selected from: a compound comprising an aryl amine group; a conjugated polymer; a conjugated oligomer; and a compound comprising a polycyclic aromatic hydrocarbon. Compounds comprising aryl amine groups are typically compounds which comprise a nitrogen atom substituted with one or more aryl groups. The semiconductor may be a compound comprising a triarylamine group. A compound comprising a triarylamine group typically refers to an organic semiconductor which comprises one or more moieties which are a nitrogen atom substituted with three substituted or unsubstituted aryl or heteroaryl rings. The simplest such semiconductor is triphenylamine, N(C ₆ H ₅ ) ₃ .

Often, the organic semiconductor comprises a compound of formula (I), (II), (ΠΙ), (IV) or (V):

wherein: each Ai is the same or different and is an unsubstituted or substituted aryl ring or an unsubstituted or substituted heteroaryl ring and n is 3 or 4. For instance, each Ai may be phenyl or methoxyphenyl. In formula (II), the central Ai ring may in some cases be replaced by a biphenyl group.

Examples of compounds comprising an aryl amine group include triphenylamine, tri(4- methylphenyl)amine, l,4-bis(diphenylamino)benzene, l,3-bis(N-carbazolyl)benzene, 4,4'- bis(N-carbazolyl)- 1, 1 '-biphenyl, 4,4'-bis(3-ethyl-N-carbazolyl)-l,l '-biphenyl, N,N'-bis(3- methylphenyl)-N,N'-diphenylbenzidine, N,N'-bis(phenanthren-9-yl)-N,N'-diphenylbenzidine, 4,4'-cyclohexylidenebis[N,N-bis(4-methylphenyl)benzenamine], 2,2'-dimethyl-N,N'-di-[(l- naphthyl)-N,N'-diphenyl]-l, 1 '-biphenyl -4,4 '-diamine, 9,9-dimethyl-N,N'-di(l-naphthyl)- N,N'-diphenyl-9H-fluorene-2,7-diamine, N,N,N',N'-tetrakis(4-methoxyphenyl)benzidine (MeO-TPD), N,N,N',N'-tetrakis(3 -methylphenyl)-3,3 '-dimethylbenzidine, 4- (diphenylamino)benzaldehyde diphenylhydrazone, N,N,N',N'-tetraphenylnaphthalene-2,6- diamine, tris(4-carbazoyl-9-ylphenyl)amine, tris[4-(diethylamino)phenyl]amine, 1,3,5- tris(diphenylamino)benzene, l,3,5-tris[(3-methylphenyl)phenylamino]benzene, 2,7- bis(carbazol-9-yl)-9,9'-spirobifluorene, 2,2',7,7'-tetrakis(N,N-diphenylamino)-9,9- spirobifluorene, 2,2',7,7'-tetrakis[N,N-di(4-methoxyphenyl)amino]-9,9-spirobi fluorene and poly[N,N-diphenyl-4-methoxyphenylamine-4',4"-diyl] (PMeOTPA).

Examples of conjugated polymers include polyacetylene, polyphenylene, polyparaphenylene, polyparaphenylene vinylene, polyparaphenylene acetylene, polyazulene, polynaphthalene, polypyrene, polyaniline, polyparaphenylene sulphide, polyfluorene, polypyrrole,

polythiophene, polythieno[3,2-b]thiophene, polycarbazole, polyazepine and polyindole, each of which may be unsubstituted or substituted.

The organic semiconductor typically comprises a compound selected from: 2,2',7,7'- tetrakis[N,N-di(4-methoxyphenyl)amino]-9,9-spirobifluorene (spiro-OMeTAD); Ν,Ν,Ν',Ν'- tetrakis(4-methoxyphenyl)benzidine (MeO-TPD); 4,4',4' ^r -tris[phenyl(m- tolyl)amino]triphenylamine (m-MTDATA); poly(4-butylphenyldiphenylamine); poly(3- hexylthiophene-2,5-diyl) (P3HT); poly(triaryl amine); poly[bis(4-phenyl)(2,4,6- trimethylphenyl)amine] (PTAA); N,N'-di(l-naphthyl)-N,N'-diphenyl-(l, 1 '-biphenyl)-4,4'- diamine (NPD); polyvinyl carbazole (PVK); and methoxydiphenylamine-Substituted

Carbazole Twin Derivative (V886). Often, the organic semiconductor comprises spiro- OMeTAD or MeO-TPD.

The activator typically comprises a species X, which may be an anion. Once the

sulfoxide: activator adduct has oxidised the organic semiconductor, the resulting oxidised semiconductor (which may be referred to as [OS ⁺ ]) may form a salt with X. For instance, if the adduct is DMSO:HBr, the resulting doped semiconductor typically comprises a bromide salt of the organic semiconductor (for instance [spiro-OMeTAD ⁺ ][Br ^" ]).

For instance, the activator may comprise a compound of formula HX, and the doped organic semiconductor comprises a compound of formula [OS ⁺ ][X ^_ ], wherein OS is the organic semiconductor. X may be as defined herein, for instance Br or camphor sulfonate. Some activators comprise small anions (for instance bromide) which may migrate out of the organic semiconductor after doping and potentially leading to de-doping of the organic semiconductor The inventors have discovered that this de-doping may be avoided if the anion X taken from the activator is exchanged with a second anion which does not diffuse out of the organic semiconductor. Accordingly, the process of the invention may comprise treating the organic semiconductor with (i) the compound comprising a sulfoxide group, (ii) the activator which is a first acid, and (iii) a second acid. The first acid is typically a compound of formula HX as described herein. The organic semiconductor may be treated with (i), (ii) and (ii) simultaneously or sequentially.

Typically, the second acid is an organic acid. The organic acid may be an organic acid having a molecular weight of greater than or equal to 50 gmol ^"1 . The organic acid may have a molecular weight of greater than or equal to 100 gmol ^"1 . The organic acid may be a carboxylic acid or a sulfonic acid. For instance, the organic acid may be a compound of formula RCOOH or RSO3H where R is an organic group. R typically comprises from 1 to 20 carbon atoms and may be substituted or unsubstituted. For instance, R may be a bicyclic organic group comprising from 6 to 12 carbon atoms and optionally one to three oxo groups. R may alternatively be a substituted or unsubstituted CMS alkyl group or a CMS alkenyl group. The second acid may be a compound of formula HY, wherein Y is an anion having a molecular weight of at least 100 gmol ^"1 . Y is typically an organic anion.

Examples of the second acid include camphorsulfonic acid, methanesulfonic acid, triflic acid, sulfoacetic acid, ethane sulfonic acid, 1 -propane sulfonic acid, benzoic acid, phthalic acid, salicylic acid, benzene sulfonic acid (optionally substituted with one or more groups selected from hydroxyl, nitro and amino), phenylsulfonic acid, p-toluenesulfonic acid, formic acid, acetic acid, trifluoracetic acid, propionic acid, malonic acid, citric acid, tartaric acid, barbituric acid, cinnamic acid, fumaric acid, glutaric acid, gluconic acid, hexanoic acid, lactic acid, malic acid, folic acid and oleic acid. The second acid may for instance be selected from camphorsulfonic acid, benzoic acid, phthalic acid, salicylic acid, p-toluenesulfonic acid, acetic acid and trifluoracetic acid. Preferably, the second acid is camphorsulfonic acid (camphor- 10-sulfonic acid). Camphor- 10-sulfonic acid has the structure:

If a second acid is present, then the doped organic semiconductor may comprise a compound of formula [OS ⁺ ][Y ^_ ], wherein OS is the organic semiconductor and Y is as defined herein for the second acid.

In some cases, the process comprises mixing: (a) a first solution comprising a first solvent, the compound comprising a sulfoxide group, the activator which is a first acid, and the second acid; and (b) a second solution comprising a second solvent and the organic semiconductor; and producing a composition comprising the first solvent, the second solvent and the doped organic semiconductor.

The molar ratio of the compound comprising a sulfoxide group and the second acid is typically from 1.0:5.0 to 5.0: 1.0. Preferably the molar ratio of the compound comprising a sulfoxide group and the second acid is from 1.0: 1.0 to 1.0:3.0. For instance, the molar ratio of the compound comprising a sulfoxide group and the second acid may be 1 :2. The molar ratio of the first acid to the second acid may be from 5.0: 1.0 to 1.0:5.0.

The process may comprise (a) disposing on a substrate a composition comprising the organic semiconductor and the second acid to form a layer comprising the organic semiconductor and the second acid and (b) treating the layer with the compound comprising a sulfoxide group and the activator which is a first acid. Treating the layer with the compound comprising a sulfoxide group and the activator may comprise: exposing the layer to a vapour comprising the compound comprising a sulfoxide group and the activator; or disposing on the layer a composition comprising the compound comprising a sulfoxide group and the activator.

The process may comprise (a) disposing on a substrate a solution comprising the second acid and the organic semiconductor to produce a layer and (b) exposing the layer to a vapour comprising the compound comprising a sulfoxide group and the activator. The process may comprise (a) disposing on a substrate a solution comprising the second acid, the activator and the organic semiconductor to produce a layer and (b) exposing the layer to a vapour comprising the compound comprising a sulfoxide group. The process may comprise (a) disposing on a substrate a solution comprising the second acid, the compound comprising a sulfoxide group and the organic semiconductor to produce a layer and (b) exposing the layer to a vapour comprising the activator.

The process of the invention is typically conducted at a temperature from 10°C to 60°C. The organic semiconductor is typically treated with the compound comprising a sulfoxide group and the activator for from 1 second to 1 hour, for instance from 10 seconds to 30 minutes.

The process of the invention is effective and can produce heavily doped organic

semiconductors. "Heavily doped" means that the majority (e.g. greater than 80 mol% or greater than 90 mol%) of the molecules of organic semiconductor have been oxidised by the sulfoxide compound and activator. Typically, organic semiconductors are used in a moderately doped state. A moderately doped semiconductor (e.g. the p-doped

semiconductor) can be produced by mixing the heavily doped organic semiconductor obtained from the process of the invention with an additional amount of undoped organic semiconductor. According, the doped organic semiconductor obtained from the process of the invention may be a heavily doped organic semiconductor and the process may further comprise combining the heavily doped organic semiconductor with an additional amount of the organic semiconductor to produce a moderately doped organic semiconductor. This may occur in the solid phase or the solution phase. For instance, after preparation of a solution comprising the heavily doped organic semiconductor, an additional amount of undoped organic semiconductor may be added.

The process for producing a doped organic semiconductor may be a process for producing a gradient doped organic semiconductor (which may also be referred to as a graded doped organic semiconductor). A gradient doped organic semiconductor is an organic

semiconductor in which there is an increasing or decreasing extent of doping of the organic semiconductor in a direction through the organic semiconductor (for instance in a direction away from the surface and into the interior of the semiconductor). For example, a gradient doped film may be a doped film of an organic semiconducting wherein the concentration of dopant ions is higher on one side of the film that the other.

A process for producing a gradient doped organic semiconductor typically comprises disposing on a layer of the organic semiconductor a gradient doping composition comprising one or more of (i) the compound comprising a sulfoxide group, (ii) the activator and (iii) the second acid (as defined herein). The amount of the gradient doping composition disposed affects the extent of gradient doping of the organic semiconductor. For instance, a thicker layer of the gradient doping composition may be disposed on the layer of the organic semiconductor in order to cause a greater extent of gradient doping.

The gradient doping composition is typically a liquid and typically further comprises one or more solvents. The gradient doping composition comprises one or more of components (i), (ii) and (iii). Any of components (i), (ii) and (iii) which is not present in the gradient doping composition may be subsequently disposed on the layer of the organic semiconductor by vapour deposition. For instance, the process may comprise: disposing on a layer of an organic semiconductor a gradient doping composition which is a solution comprising (i) the compound comprising a sulfoxide group, (ii) the activator and (iii) the second acid. The process may alternatively comprise: (a) disposing on a layer of an organic semiconductor a gradient doping composition which is a solution comprising (i) the compound comprising a sulfoxide group and (iii) the second acid; and (b) subsequently exposing the layer of an organic semiconductor to a vapour comprising (ii) the activator. The process may alternatively comprise: (a) disposing on a layer of an organic semiconductor a gradient doping composition which is a solution comprising (ii) the activator and (iii) the second acid; and (b) subsequently exposing the layer of an organic semiconductor to a vapour comprising (i) the compound comprising a sulfoxide group. The process may alternatively comprise: (a) disposing on a layer of an organic semiconductor a gradient doping composition which is a solution comprising (iii) the second acid; and (b) subsequently exposing the layer of an organic semiconductor to a vapour comprising (i) the compound comprising a sulfoxide group and (ii) the activator. The compound comprising a sulfoxide group may be DMSO and the activator may be FIBr. The second acid may be camphorsulfonic acid (CSA). Exposing a layer of an organic semiconductor to a vapour comprising one or more of (i), (ii) and (iii) typically comprises exposing the layer to the vapour at a pressure of from 1 to 200 kPa. Typically the layer is exposed for from 1 second to 10 minutes In some cases, the order of steps (a) and (b) may be reversed with the vapour deposition occurring before the solution deposition.

The one or more solvents present in the gradient doping composition may be as defined herein. Typically, the solubility of the organic semiconductor in the one or more solvents is low (e.g. less than 1 mg/mL or less than 0.01 mg/mL). For instance, the gradient doping composition may comprise one or more of acetonitrile, 2-ethoxyethanol and water. The concentration of the second acid in the gradient doping composition, if present, is typically from 0.01 to 1.0 mg/mL.

The invention also provides a process for producing a layer comprising a doped organic semiconductor, which process comprises producing the doped organic semiconductor by a process according to the invention. The process for producing a layer comprising a doped organic semiconductor typically comprises disposing on a substrate a composition comprising: (i) a compound comprising a sulfoxide group, (ii) an activator and (iii) an organic semiconductor. Typically, the process comprises (a) disposing on a substrate a solution comprising: (i) a compound comprising a sulfoxide group, (ii) an activator; (iii) an organic semiconductor and (iv) a solvent, and (b) removing the solvent to form a layer comprising a doped organic semiconductor. The solvent may be removed by heating, by applying a vacuum or by allowing the solvent to evaporate.

Alternatively, the process for producing a layer comprising a doped organic semiconductor may comprise contacting a layer of the organic semiconductor with (i) a compound comprising a sulfoxide group and (ii) an activator. For instance, the layer of the organic semiconductor could be dipped in a solution comprising the compound comprising a sulfoxide group and the activator or the layer of the organic semiconductor could be exposed to a vapour comprising the compound comprising a sulfoxide group and the activator.

The layer comprising the doped organic semiconductor may for instance have a thickness of from 0.1 μπι to 100 μπι.

The invention also provides a process for producing a semiconductor device, which process comprises producing a doped organic semiconductor by a process according to the invention. The process for producing a semiconductor device may comprise producing a layer comprising a doped organic semiconductor by a process according to the invention.

The semiconductor device typically comprises a layer of the doped organic semiconductor. The semiconductor device is typically an optoelectronic device, a photovoltaic device, a solar cell, a photo detector, a photodiode, a photosensor, a chromogenic device, a transistor, a light-sensitive transistor, a phototransistor, a solid state triode, a battery, a battery electrode, a capacitor, a super-capacitor, a light-emitting device or a light-emitting diode (for instance an organic light-emitting diode). Typically, the semiconductor device is an optoelectronic device such as a photovoltaic device or a light-emitting device. The process for producing a semiconductor device typically comprises (i) providing a substrate comprising a layer of a photoactive material and (ii) disposing on the photoactive material a layer of a doped organic semiconductor by a process according to the invention.

The photoactive material may be any suitable photoactive material. Typically, the photoactive material comprises a perovskite compound. Examples of suitable photoactive materials are described in WO 2013/171517, WO 2013/171518, WO 2017/037448 and WO 2017/089819, the entirety of which are incorporated by reference. The semiconductor device may be as defined in WO 2013/171517, WO 2013/171518, WO 2017/037448 and WO 2017/089819. The photoactive material may be a compound of formula

FAo.83Cso.i7Pb(Io.8Bro. ₂ )3 or FAo.83Cso.i7Pb(Io.66Bro.44)3 where FA is formamidinium.

Composition

The invention also provides a composition comprising: (i) a compound comprising a sulfoxide group; (ii) an acid of formula HX, where X is an anion; and (iii) an organic semiconductor. Each component may be as defined for the process of the invention.

Typically, X is an anion selected from Br, Or, Γ, F , CH3COO , HCOO , CF3COO , C10H14SO4- (camphorsulfonate), (CF ₃ S0 ₂ )2N-, HS0 ₄ ^~ NO3 and H ₂ P0 ₄ . X may be Br or camphorsulfonate.

The composition typically further comprises one or more solvents. The one or more solvents may be selected from water, an alcohol (such as methanol or ethanol), dichloromethane, chlorobenzene, acetonitrile or dimethylformamide (DMF). The composition may further comprise water and/or chlorobenzene.

Typically, the concentration of the compound comprising a sulfoxide group is from 0.005 mM to 14 M; and the concentration of the acid is from 0.01 mM to 8.8 M. For instance, the concentration of the compound comprising a sulfoxide group may be from 1 mM to 1 M. The concentration of the acid may be from 1 mM to 1 M.

Preferably, the composition comprises: (i) dimethylsulfoxide; (ii) FIBr; and (iii) an organic semiconductor comprising a compound which comprises an aryl amine group. The composition may further comprise a second acid as described herein, for instance camphorsulfonic acid. The organic semiconductor may be spiro-OMeTAD or MeO-TPD. The invention also provides use of a composition comprising: (i) a compound comprising a sulfoxide group; and (ii) an activator, to dope an organic semiconductor. The composition may be as defined herein. The process for doping the organic semiconductor may be as defined herein.

EXAMPLES

Example 1 Experimental

DMSO-HBr doping by solution process

150 \L of DMSO (Aldrich; anhydrous, >99.9%) was added into 500 μL of HBr (Aldrich: 48 wt. % in H2O, >99.99%) and the solution was mixed by a vortex mixer for 1 min. 10 μL of DMSO-HBr solution was added to 2.5 mL of spiro-OMeTAD (0.28 mM) solution in chlorobenzene (Aldrich; anhydrous, >99.9%). The solution was mixed by a vortex mixer for 1 min and then kept for 20 min at room temperature to prepare a heavily doped spiro- OMeTAD solution. 3-15 \\L of the heavily doped spiro-OMeTAD solution was added to 1 mL of un-doped spiro-OMeTAD solution (70 mM) in chlorobenzene and the solution was kept for at least 10 minutes before the thin film preparation.

DMSO-HBr doping by vapor process

70 mM of spiro-OMeTAD in chlorobenzene was spin coated onto the required substrate at 2500 rpm for 45 sec (acceleration: 500 rpm). 1 5 μL of DMSO and 5 \\L of HBr were put in a vacuum chamber (pressure -70 kPa) along with the un-doped spiro-OMeTAD thin film. The vial containing DMSO was heated at 55 °C in the vacuum chamber while the substrate and HBr container remained at the ambient temperature. The exposure time was varied to control the doping concentration in the film.

DMSO-HBr-CSA doping

30 mg of CSA (Aldrich, (±)-10-camphorsulfonic acid, 98%) was dissolved in 20 μL of DMSO and 300 \L of chlorobenzene. The solution was vortexed until CSA was dissolved completely. 3 μL of HBr (48 wt% in water) was added to the CSA solution and the solution was vortexed for 30 seconds. 50 uL of DMSO-HBr-CSA solution was added into 500 \\L of spiro-OMeTAD solution (7.78 mM) in chlorobenzene to prepare a heavily doped solution. 10 - 50 \L of heavily doped spiro-OMeTAD solution was added to the un-doped spiro- OMeTAD chlorobenzene solution (70 mM) and was kept for 10 min before the preparation of thin films. The spiro-OMeTAD solution was spin-coated on to the perovskite layer at 2500 rpm for 45 sec (acceleration: 500 rpm).

O2 doping for spiro-OMeTAD

A thin film of spiro-OMeTAD was stored in a low humidity desiccator overnight. This led to doping of the thin film by O2.

Doping of other hole-transporting materials

Other hole-transporting materials were doped in an equivalent manner as described for spiro- OMeTAD.

Fabrication of perovskite solar cell

FTO (F-doped tin oxide) -coated glass sheets (TEC 7, 7Q/sheet, Pilkington) were etched with zinc powder and 2M HC1 to obtain the required electrode pattern. The sheets were then washed with 2% Hellmanex in water with sonication for 10 minutes and then washed with deionized water, acetone, ethanol and isopropanol. The last traces of organic residues were removed by oxygen plasma cleaning for 10 minutes.

A Sn02 compact layer was fabricated from Sn0 ₂ precursor, which were fabricated via spin coating and chemical bath deposition. For spin coating Sn02, 17.5 mg of SnCU 5H2O (Sigma-Aldrich, 98%) was dissolved in 1 ml of iso-propanol. After 30 minutes of stirring at room temperature, a fully dissolved, clear solution was obtained. This solution was then filtered with a 0.2 μπι pore size polytetrafluoroethylene filter. The SnCU solution was spun on FTO substrate at 3000 rpm for 30 s (acceleration: 200 rpm), and annealed at 180 °C for 60 min to form a thin layer of Sn02. For chemical bath deposition, 108 mg of SnCh 2(H20) (Sigma-Aldrich, 98%), 500 mg of Urea (Sigma-Aldrich, 98%), 500 \L of HC1 (37% solution in water, Fisher Scientific) and 10 \\L of 3-Mercaptopropionic acid (Sigma-Aldrich, 99%) were dissolved in 40 mL of H2O. The spin coated Sn02 substrates were immersed in the above mentioned SnC solution at 70 °C for 180 min. After the incubation, the substrates were rinsed immediately in a deionized water bath and then sonicated for 120 s in deionized water. The substrates were then dried with a nitrogen gun, and annealed at 180 °C for 60 min. To obtain a FAo.83Cso.i7Pb(Io sBro.2)3 precursor solutions, FAI (formamidinium iodide), Csl, PbBr2 and Pbh were dissolved in a mixed solvent of anhydrous N,N-dimethylformamide (DMF; Aldrich) and anhydrous dimethyl sulfoxide (DMSO; Aldrich) in 4: 1 volume ratio to obtain a stoichiometric solution with desired composition and a molar concentration of 1.4 M.

The perovskite precursor solution was coated onto the FTO-SnC substrate by a consecutive two-step spin-coating process at 1000 rpm for 10 sec and 5000 rpm for 25 sec under low humidity (15-20% at 20°C) condition with a dry compressed air purge in a dry box. 200 \L of anisole and chlorobenzene in 9: 1 volume ratio was dropped onto the pre-crystallized perovskite film for 8- 10 seconds during the second step of spin coating (i.e. when the spin speed was at 5000 rpm). The perovskite films were dried at 80 °C for 5 min and then 100 °C for 60 min in a box oven.

(a) DMSO -HBr doping by solution process

150 μL· of DMSO (Aldrich; anhydrous, >99.9%) was added into 500 μΐ, of HBr (Aldrich: 48 wt. % in H2O, >99.99%) and the solution was mixed by a vortex mixer for 1 min. 10 μΐ _^ of DMSO-HBr solution was added to 2.5 mL of spiro-OMeTAD (0.28 mM) solution in chlorobenzene (Aldrich; anhydrous, >99.9%). The solution was mixed by a vortex mixer for 1 min and then kept for 20 min at room temperature to prepare a heavily doped spiro- OMeTAD solution. 3-15 of the heavily doped spiro-OMeTAD solution was added to 1 mL of un-doped spiro-OMeTAD solution (70 mM) in chlorobenzene and the solution was kept for at least 10 minutes before the thin film preparation. 30 μΐ. of tert-butly pyridine (tBP) and 20 of bis(trifluoromethane)sulfonimide lithium salt (Li-TFSI) in acetonitrile were added into 1 mL of spiro-OMeTAD solution as the extra additives along with DMSO- HBr.The spiro-OMeTAD solution was spin-coated on to the perovskite layer at 2500 rpm for 45 sec (acceleration: 500 rpm).

(b) DMSO-H£r-(±)-10-camphorsulfonic acid (CSA) doping

30 mg of CSA (Aldrich, (±)-10-camphorsulfonic acid, 98%) was dissolved in 20 \L of DMSO and 300 \L of chlorobenzene. The solution was vortexed until CSA was dissolved completely. 3 μL of HBr (48 wt% in water) was added to the CSA solution and the solution was vortexed for 30 seconds. 50 μΐ _^ of DMSO-HBr-CSA solution was added into 500 μί _^ of spiro-OMeTAD solution (7.78 mM) in chlorobenzene to prepare a heavily doped spiro- OMeTAD solution. 10 - 50 μΐ _^ of the heavily doped spiro-OMeTAD solution was added to the un-doped spiro-OMeTAD chlorobenzene solution (70 mM) and was kept for 10 min before the preparation of thin films. 30 μΐ _^ of tert-butly pyridine (tBP) and 20 of bis(trifluoromethane)sulfonimide lithium salt (Li-TFSI) in acetonitrile were added into 1 mL of the doped spiro-OMeTAD solution as the extra additives along with DMSO-HBr-CSA. The spiro-OMeTAD solution was spin-coated on to the perovskite layer at 2500 rpm for 45 sec (acceleration: 500 rpm).

80 nm of Au metal contact layer was deposited as the counter electrode on the hole- transporting material layer by thermal evaporation.

Fabrication of perovskite solar cell for stability test

FTO (F-doped tin oxide) -coated glass sheets (TEC 15, 15Q/sheet, Pilkington) were etched with zinc powder and 2M HCl to obtain the required electrode pattern. The sheets were then washed with 2% Hellmanex in water with sonication for 10 minutes and then washed with deionized water, acetone, ethanol and isopropanol. The last traces of organic residues were removed by oxygen plasma cleaning for 10 minutes.

An Sn0 ₂ compact layer was fabricated from Sn02 nanoparticles, which were synthesized via hydrothermal synthesis. 467 mg of SnCU 5H ₂ 0 (Sigma-Aldrich, 98%) was dissolved in 20 ml of deionized water. After 10 minutes of stirring at room temperature, a fully dissolved, clear solution was obtained. This solution was then transferred to a Teflon-lined stainless steel autoclave, and heated for 2 hours at 250 °C. The autoclave was quenched to room temperature using cold water. The reacted solution was centrifuged at a speed of 10000 rpm for 10 minutes to collect Sn0 ₂ precipitates. The nanoparticles were re-dispersed in ethanol by sonication in order to wash out the unreacted SnCU and were then centrifuged at a speed of 10000 rpm for 10 minutes. This washing treatment was repeated 3 times. After the final washing treatment, the nanoparticles were also re-dispersed in ethanol. 2mg ml ^"1 Sn0 ₂ nanoparticle solution was spin coated on FTO substrate at 2000 rpm for 45 s, and annealed at 150 °C for 120 s to form a thin layer of Sn0 ₂ . Ceo (10 mg/ 1 ml of 1,2, di -chlorobenzene) layer was coated on the Sn0 ₂ compact layer (at 2000 rpm for 30 sec) and the films were annealed at 100 °C for 10 min.

FAI (formamidinium iodide), Csl, PbBr ₂ and Pbl ₂ were dissolved in anhydrous N,N- dimethylformamide (DMF; Aldrich) to get 1.3 M FAo.s3Cso.i7Pb(Io.66Bro.44)3 . 75 of hydroiodic acid (57 wt% in H2O, Aldrich) and 37 μΐ of hydrobromic acid (48 wt% in H2O, Aldrich) were added to 1ml of the 1.3M above mentioned salt solution. After the addition of the acids, the solution was stirred for ~ 48 hours and used as precursor solution for the

FAo.83Cso.i7Pb(Io.66Bro.44)3 films.

For the FAo.83Cso.nPb(Io.66Bro.44)3 film formation, the perovskite precursor solution was coated onto the SnC /Ceo substrate by a consecutive two-step spin-coating process (spin speed 1200 rpm for 20 sec followed by 3000 rpm for 20 sec) under a low humidity (15-20% at 20°C) condition with a dry compressed air purge in a dry box. The perovskite films were quickly dried by compressed air blowing for 20 sec. The perovskite films were dried at 20 °C for 15 min and then 70 °C for 10 min on a hot plate. The dried films were annealed at 175 °C for 90 min in the box oven without controlling the humidity inside the oven.

(a) DMSO-HBr doping by solution process

150 μΐ, of DMSO (Aldrich; anhydrous, >99.9%) was added into 500 μΐ, of HBr (Aldrich: 48 wt. % in H2O, >99.99%) and the solution was mixed by a vortex mixer for 1 min. 10 \L of DMSO-HBr solution was added to 2.5 mL of spiro-OMeTAD (0.28 mM) solution in chlorobenzene (Aldrich; anhydrous, >99.9%). The solution was mixed by a vortex mixer for 1 min and then kept for 20 min at room temperature to prepare a heavily doped spiro- OMeTAD solution. 3-15 μΐ _^ of the heavily doped spiro-OMeTAD solution was added to 1 mL of un-doped spiro-OMeTAD solution (70 mM) in chlorobenzene and the solution was kept for at least 10 minutes before the thin film preparation. The spiro-OMeTAD solution was spin-coated on to the perovskite layer at 2500 rpm for 45 sec (acceleration: 500 rpm).

(b) Z ⁾ MSO-HSr-(±)-10-camphorsulfonic acid (CSA) doping

30 mg of CSA (Aldrich, (±)-10-camphorsulfonic acid, 98%) was dissolved in 20 \L of DMSO and 300 uL of chlorobenzene. The solution was vortexed until CSA was dissolved completely. 3 \L of HBr (48 wt% in water) was added to the CSA solution and the solution was vortexed for 30 seconds. 50 μΐ. of DMSO-HBr-CSA solution was added into 500 uL of spiro-OMeTAD solution (7.78 mM) in chlorobenzene. 10 - 50 μί of heavily doped spiro- OMeTAD solution was added to the un-doped spiro-OMeTAD chlorobenzene solution (70 mM) and was kept for 10 min before the preparation of thin films. The spiro-OMeTAD solution was spin-coated on to the perovskite layer at 2500 rpm for 45 sec (acceleration: 500 rpm).

80 nm of Au metal contact layer was deposited as the counter electrode on the hole- transporting material layer by thermal evaporation in vacuum ( 10 ^"6 Torr)

Fabrication of Organic Light Emitting Diode (OLED):

ITO (Indium tin oxide) -coated glass sheets (15 Ω/sheet) were washed with 2% Hellmanex in water with sonication for 10 minutes and then washed with deionized water, acetone, ethanol and isopropanol. The last traces of organic residues were removed by oxygen plasma cleaning for 10 minutes.

(a) Undoped MeO-TPD films:

50 mg of N,N,N',N'-Tetrakis(4-methoxyphenyl)benzidine (MeO-TPD) in 1ml

chlorobenzene was stirred at 80 °C for 30 min. The MeO-TPD solution was spin-coated on to the ITO substrate at 2500 rpm for 45 sec (acceleration: 500 rpm).

(b) Doping of MeO-TPD film by solution process by DMSO:HBr :

150 \L of DMSO (Aldrich; anhydrous, >99.9%) was added into 500 \L of HBr (Aldrich: 48 wt. % in H2O, >99.99%) and the solution was mixed by a vortex mixer for 1 min. 10 \L of DMSO-HBr solution was added to 2.5 mL of MeO-TPD (0.32 mM) solution in

chlorobenzene (Aldrich, anhydrous, >99.9%). The solution was mixed by a vortex mixer for 1 min and then kept for 20 min at room temperature to prepare a heavily doped MeO-TPD solution. 50 μΐ, of the heavily doped MeO-TPD solution was added to 150 iL of un-doped MeO-TPD solution (80 mM) in chlorobenzene and the solution was kept for at least 10 minutes before the thin film preparation. The doped MeO-TPD solution was spin-coated on to the ITO substrate at 2500 rpm for 45 sec (acceleration: 500 rpm).

(c) Doping of MeO-TPD film by solution process by Z)MSO-Hi?r-(±)-10-camphorsulfonic acid (CSA) .

30 mg of CSA (Aldrich, (±)-10-camphorsulfonic acid, 98%) was dissolved in 20 \L of DMSO and 300 \L of chlorobenzene. The solution was vortexed until CSA was dissolved completely. 3 μΐ, of HBr (48 wt% in water) was added to the CSA solution and the solution was vortexed for 30 seconds. 50 xL of DMSO-HBr-CSA solution was added into 500 \L of MeO-TPD solution (6.87 mM) in chlorobenzene. 5 \L of heavily doped MeO-TPD solution was added to 105 μΐ, of un-doped MeO-TPD chlorobenzene solution (80 mM) and was kept for 10 min before the preparation of thin films. The MeO-TPD solution was spin-coated on to the perovskite layer at 2500 rpm for 45 sec (acceleration: 500 rpm).

Tris-(8-hydroxyquinoline)aluminum (Alq3), Bathophenanthroline (BPhen) and Al were deposited as the emitting layer, electron transporting layer (hole blocking layer) and the counter electrode, respectively by thermal evaporation in vacuum (10 ^"7 Torr). The thickness of MeO-TPD, Alq3, BPhen, Al layers were 70 nm, 100 nm, 8 nm, and 100 nm, respectively.

Characterization

Conductivity measurement

For conductivity measurements, OFET chips from Fraunhofer IPMS (Dresden, Germany) with bottom-gate and bottom contact geometry were used. The gate oxide was 90 nm thick on top of an n-doped Si wafer. There was no applied bias at the gate during the measurements. The gold interdigitated fingers of the source and drain contacts had varying channel widths of 2.5 μιτι, 5 μπι, 10 μιτι and 20 μπι and a channel length of 10 mm. Under ambient conditions, a Keithley 2400 source measurement unit was used to preform current-voltage (J-V) measurements across the varying channel widths.

UV-visible absorption measurements

UV-vis absorption spectra were measured by a commercial spectrophotometer (Varian Cary 300 UV-Vis, USA). ~ 0.05 mg/ml of hole-transporting material in chlorobenzene or toluene was usually used as a solution phase. -100 nm of hole-transporting material thin films on glass substrates were used for the thin film measurements.

Kelvin probe measurement:

The spiro-OMeTAD thin films with the different dopant concentration were prepared on an ITO coated glass substrate. A kelvin probe set up (KP technology, UK) was used to measure the surface potential to determine work function. The probe workfunction was calibrated with HOPG. The measurements were done in ambient condition. FTIR measurements:

FTIR spectrometer with liquid nitrogen cooled Mercury Cadmium Telluride (MCT) detector was used for IR spectroscopy. For multiple reflections of incident IR light, a modified ATR accessory (GladiATR, PIKE Technologies) placed inside the spectrometer. The IR cell was developed in-house for in-situ experiments (Hidalgo et al, Agnew Chemie Int Ed 127, 7216- 7219 (2015)).

Characterization for PV devices:

The current density-voltage (J-V) measurements and stabilized power output (SPO) measurements were measured (2400 series source meter, Keithley Instruments) under the simulated solar light (AAB ABET technologies Sun 2000 solar simulator) with its light intensity at 99.2 mW cm ^"2 (AM 1.5). The simulated light was calibrated against a standard amorphous-silicon PV cell (NREL-calibrated KG5 filtered silicon reference cell). The mismatch factor was estimated to be M=l .035405 and the lamp intensity was adjusted to account for this mismatch. The J-V curves were measured between 1.2 V to 0 V. The cell aperture area of light incidence was set to 0.0913 cm ² photoactive area by employing an opaque mask.

Stability test of the perovskite cells:

The device stability test was carried out under full spectrum simulated AMI .5, 76 mA cm ^"2 irradiance at Foc using an Atlas SUNTEST XLS+ (1700W air-cooled Xenon lamp). All devices were aged without encapsulation. The chamber temperature is kept -50 °C. The light source was pulsed at 100Hz .

Measurement of OLED:

The current density-voltage (J-V) measurements were measured (2400 series source meter, Keithley Instruments). The luminescence intensity was measured along with bias voltage from a source meter, and luminescence was collected using with a fiber-coupled detector (Ocean Optics USB 2000+) in an integrating sphere (Oriel Instruments 70682NS). Thermal stress test:

The thin films of hole-transporting materials were fabricated on top of a glass substrate for Uv-vis absorption spectra and on top of an OFET chip (without the biasing the gate) for conductivity measurement. The thermal tolerance test was carried out on a hot plate in ambient atmospheric condition for both of UV-vis and conductivity measurements. For conductivity measurement, the substrates were heated at 50, 80, 100 and 120 °C for 10 min at each heating step. Measurements were done after heating at each temperature.

Ellipsometry measurement:

50 mg of N,N,N',N'-Tetrakis(4-methoxyphenyl)benzidine (MeO-TPD) in chlorobenzene was stirred at 80 °C for 30 min. The DMSO-HBr doped MeO-TPD solution was spin-coated on to a glasssubstrate at 2500 rpm for 45 sec (acceleration: 500 rpm) which gave -70 nm thinfilm. Undoped MeO-TPD film (70nm) was deposited on top of the doped thinfilm by thermal evaporation in vacuum ( 10 ^"6 torr) . Elliposmetric measurements of the doped-undoped film under thermal stress were done using a J.A. Woollam ellipsometer. The thermal tolerance test was carried out on a hot plate in ambient atmospheric condition

Results and discussion

In this Example, it has been shown that a dimethyl sulfoxide (DMSO)-based metastable adduct can be effectively used as a p-dopant for several organic semiconductors. The adduct is formed in situ by the interaction of DMSO with an activator. The advantages of this system over the existing methods and materials include: (a) the availability of flexible doping methods (i.e. either solution or vapour phase may be used); (b) a clean chemical reaction which does not form undesirable by-products; (c) the ability to dope a wide range of different organic semiconductors; (d) low cost doping materials; (e) easy handling of the process; and (f) reduced toxicity. Furthermore, it is possible to separately select the counter ion which is introduced into the doped semiconductor allowing other properties of the semiconductor to be tuned.

A commonly used hole transporting material (HTM) in organic electronics, spiro-OMeTAD, was chosen as a testbed to demonstrate the doping, understand the mechanism and the usage in device application. In Figure la, the chemical structure of DMSO, HBr (one of the activator) and Spiro-OMeTAD are shown. To dope spiro-OMeTAD, DMSO and HBr were prepared in 1 :2 molar ratio and added the solution to a solution of spiro-OMeTAD in chlorobenzene. The detailed procedure is provided in the experimental section above. Figure lb shows the evolution of the absorbance spectra as a function of added doping material. It is observed that an absorption peak related to a neutral state of spiro-OMeTAD at 370 nm decreases with the increase of concentration of DMSO-HBr mixture, and the peak for singly oxidized spiro-OMeTAD appears at 502 nm. On further increasing the DMSO-HBr mixture concentration, the absorption peak at 690 nm which is due to doubly oxidized spiro- OMeTAD appears. For vapor phase doping, a thin film of spiro-OMeTAD is exposed to the vapor comprising DMSO and HBr in a sealed chamber at -500 torr. The absorbance spectrum of the spiro-OMeTAD before and after the exposure is shown in Figure lb.

Figure l c shows the conductivity values of the spiro-OMeTAD thin-films prepared with the solution phase oxidation (top) and the vapor phase oxidation (bottom). The highest conductivity values of doped spiro-OMeTAD are higher or comparable to other methods of doping such as with ionic liquid (HTFSI), spiro-(TFSI)2, and Co(III)TFSI complex (J. Am. Chem. Soc. 2013, 135, 13538-13548; J. Am. Chem. Soc. 2014, 136, 10996-1 1001 ; J. Am. Chem. Soc, 201 1, 133 (45), pp 18042-18045).

The doping effect with HBr and DMSO is not only limited to spiro-OMeTAD. It has been found that the method can dope a variety of hole transporting materials. Figure 2 shows the absorbance spectra of solutions after doping of various organic semiconducting materials. Conductivity measurements were performed for all the HTMs (see Figure 3) and increments in conductivity by 2-3 orders of magnitude were found on doping. From Kelvin probe measurements, spiro-OMeTAD thin films doped with DMSO-HBr showed higher work function values with respect to the undoped film (see Table 1).

Dopant concentration (mol%) WF (eV) Table 1

0 4.537 ± 0.004

0.16 4.594 + 0.001

3.25 4.625 ± 0.002

9.67 4.695 ± 0.009

18.5 4.711 ± 0.008

33.8 4.730 ± 0.003

58.0 4.730 ± 0.003 Work was undertaken to understand the mechanism of the doping. It is thought that neither DMSO nor HBr can dope HTMs when used alone. This indicates that the dopant forms in- situ when DMSO-HBr comes together. In-situ ATR-Fourier transform infrared spectroscopy (ATR-FTIR) spectroscopy and a gas chromatography mass spectrometry (GC-MS) were used to unravel the mechanism. Figure 4a shows the FTIR spectrum of DMSO in chlorobenzene. After the addition of the activator (i .e. HBr) to the solution, the absorbance peak around -1050 cm ^"1 that corresponds to absorbance of S=0 bond shifts to a lower wavenumber and broadens. This shift of absorbance peak to a lower energy indicates that the bond order of S=0 bond in DMSO has changed due to the formation of a DMSO-HBr adduct. On addition of spiro-OMeTAD, it can be seen that the adduct peak decreases which indicates the consumption of the DMSO-HBr adducts during the doping process. When observing the water absorbance region (Figure 4b), it is found that there is an increase in the water concentration in the solution. From the FTIR measurement it can be concluded that the DMSO-HBr adducts are consumed during the doping process and water molecules are generated as by-products. To further investigate what other compounds form during the doping process, GC-MS analysis of the gaseous products that come from the solution was performed. It was found that dimethyl sulfide is another by product of the reaction. Based on the UV-vis, FTIR and GC-MS studies, the following mechanism for the doping is proposed. The DMSO-HBr adduct is generated in-situ which act as the dopant. The DMSO-HBr adduct takes electron(s) from the organic material and the oxidation state of sulfur in DMSO changes from (+4) to (+2) state. After this oxidation process, the available bromide ions from HBr serve as a counter anion for the oxidized organic species. Finally, the O from S=0 and H ⁺ from HBr forms H ₂ 0 and DMSO reduces to DMS. Both of the by-products, H ₂ 0 and DMS can leave the organic matrix even at room temperature. The doping mechanism is shown schematically in Figure 4c.

In the above demonstration, bromide (Br ^" ) has been used as the counter ion. Bromide can in some cases diffuse across the organic matrix and can leave the organic matrix at higher temperatures by a de-doping process which is often not desirable. The mobile property of Br- into the N,N,N',N'-Tetrakis(4-methoxyphenyl)benzidine (MeO-TPD) matrix was investigated by spectroscopic ellipsometry measurements and both of A and lvalues does not change over 100 h even the films were kept at 50 °C as shown in Figure 5. To further improve the thermal stability of the doping , DMSO-HBr-10-camphorsulfonic acid (CSA) doping system was introduced. The rationale behind the modified doping system is that while DMSO-HBr does the doping as explained earlier, CSA provides camphorsulfonate, a bigger counter ion instead of Br ^" . 4.5 wt% of HBr was used into the DMSO-CSA (1 :2 molar ratio) to dope spiro- OMeTAD. Figure 6a shows the absorption spectra of the spiro-OMeTAD solutions with DMSO-CSA and DMSO-HBr-CSA. The DMSO-HBr-CSA doping reaches to similar conductivity level with the reported dopant (Figure 6b). Figure 7 show a comparative study of the thermal tolerance test of doped organic thin films doped by two different systems. The thin films show greater thermal stability when DMSO-HBr-CSA is used as dopant. In the case of DMSO-HBr dopant only, a decrease in conductivity happened in between 50 ~ 85 °C.

In order to assess the impact of p-doping for an optoelectronic application, perovskite solar cells with organic-inorganic lead halide perovskite material and organic light emitting devices (OLED) were fabricated. FAo.83Cso.nPbIo sBro.2 was used as the absorber and this was sandwiched between spiro-OMeTAD as the hole transport layer and Sn0 ₂ layer as the n-type electron transport layer as shown in Figure 8a. Figures 8b and 8c show the current density voltage (JV) curves and stabilized power output (SPO) with measured under simulated sun light (99.2 mWcm ^"2 , AMI .5) of the best performing devices to compare the influence of different p-type dopants. The power conversion efficiency (and solar cell performance parameters) was determined from the forward-bias to short-circuit (FB-SC) JFcurves to be 21.1 % (Jsc = 23.2 mA cm ^"2 , V _oc = 1.14V, d FF = 79.3%) and 21.3 % (J _sc = 23.5 mA cm ^"2 , Voc = 1.11 V, and FF = 81 0%) for the cell fabricated with DMSO-HBr doping and DMSO- HBr-CSA doping, respectively. In comparison, 18.5 % (J _sc = 22.6 mA cm ^"2 , V _oc of 1.07 V, and FF = 76.0) for the devices incorporating conventional O2 doping was showed. SPO measurement was also determined by measuring the photocurrent density over time at a fixed voltage at the maximum power point which was determined from the FB-SC JV curves (Figure 8b). It was found that the efficiency and photocurrent for the devices fabricated from the O2 doping stabilized at 17.2 % (J _mpp : 19.8 mAcm ^"2 and V _mpp : 0.87 V). Further improvement in the SPO was observed for the devices fabricated from the DMSO-HBr and DMSO-HBr-CSA doping with efficiency values stabilizing at 19.7 % (J _mpp : 21.4 mA/cm ² and Vm _PP : 0.92 V) and at 20.1 % (J _mpp : 20.9 mA/cm ² and V _mpp : 0.96 V), respectively. A stability test of the doped hole-transporting material films with perovskite material as the solar cell wer shown under full spectrum simulated (76 mA cm ^"2 , AMI .5) irradiance at Voc without a UV filter at -50 °C (Figure 9). For the stability test, the perovskite devices were fabricated without capsulation and without the extra additives (i.e. tBP and Li-TFSI) into the spiro- OMeTAD solution. Cells containing FA083Cso.17PbI0 668Bro.44 as the perovskite absorber were also fabricated. It was observed that the use of DMSO-HBr and DMSO-HBr-CSA p- doping for spiro-OMeTAD showed longer stability than the device with O2 doping.

For OLED, the device architecture Tris-(8-hydroxyquinoline)aluminum (Alq3) was used as the emitting material, which is sandwiched between MeO-TPD as the hole transport layer and a bathophenanthroline (Bphen) as the electron transport layer as shown in Figure 10a. Figure 10b shows the JV curves of the OLEDs with and without doped MeO-TPD layer. Doped HTM layer improved the current density at a given applied voltage. Figure 10c shows the luminescence-voltage curve of the devices. The un-doped device showed a luminescence intensity of 60 cd/m ² at -17 V. On the other hand, the devices doped with DMSO-HBr and DMSO-HBr-CSA displayed further improvement in the luminescence intensity at the same voltage (i.e. 17 V), which reached 400 cd/m ² .

In conclusion, a sulfoxide-activator based p-type doping method for organic hole transporting material has been developed. It has been shown that a DMSO-adduct based reaction can be effectively used to electronically dope organic semiconductors. It has also been demonstrated that this adduct based doping method works by both solution and vapour phase processing. A mechanism for the doping process has been suggested. The by-products of the sulfoxide doping process are non-toxic and volatile and easily escape the organic matrix after doping. It has also been demonstrated that this method gives an opportunity to choose different counter ions for the doped organic species. In addition, use of the doped organic semiconductor in photovoltaic and light emitting applications have been demonstrated an improvement of their performance with the DMSO-adduct based dopants, particularly for photovoltaic application has achieved PCEs of over 21%. This clean, cheap and simple doping method can be adopted in a wide range of organic electronics application as a FET transistor, light emitting diode and PV devices. Example 2

General experimental

Preparation ofDMSO-HBr adduct forming agent (A.F.A): 150 pL of DMSO (Aldrich;

anhydrous, >99.9%) was added into 500 pL of HBr (Aldrich: 48 wt. % in H ₂ 0, >99.99%) and the solution was mixed by a vortex mixer for 1 min. The concentration of DMSO in the A.F.A. is (4.23 molar).

DMSO-HBr adduct Doping of HTM in solution: 1-5 pL of A.F.A. was added to the 1 mL HTM solutions in chlorobenzene or toluene and the solution was mixed by a vortex mixer. The solution was kept at room temperature for 20 minutes. The concentration of the A.F.A. in the HTM solutions was varied by adding a different amount of A.F.A..

DMSO-HBr adduct Doping of HIM thin films in solution: 10 pL of A.F.A. were added to 2.5 mL of 2,2(7,7(-tetrakis-(N,N-di-p-methoxyphenylamine)9,9(-spirobif luorene))) (spiro- OMeTAD, 0.28 mM: Luminescence technology) solution in chlorobenzene (Aldrich;

anhydrous, >99.9%). The solution was mixed by a vortex mixer for 1 min and then kept for 20 mins at room temperature to prepare a heavily doped spiro-OMeTAD solution. 3-15 pL of the heavily doped spiro-OMeTAD solution was added to 1 mL of un-doped spiro-OMeTAD solution (70 mM) in chlorobenzene and the solution was kept for at least 15 minutes before thin film preparation. A typical example for the preparation of doped thin films spiro- OMeTAD is provided here. Similar protocols were adapted for other HTMs: The

concentration of A.F.A. in the HTM solution was varied by changing the amount of A.F.A. in the HTM solution.

DMSO-HBr adduct Doping of Spiro-OMeTAD thin films by vapor process: Spiro-OMeTAD in chlorobenzene (concentration = 70 mM) was spin coated onto the substrate at 2500 rpm for 45 sec (acceleration: 500 rpm). 1.5 pL of DMSO and 5 pL of HBr were put in separate crucibles in a vacuum desiccator (PELCO 2245; pressure -70 kPa) along with the un-doped spiro-OMeTAD thin films on glass substrates. The vial containing DMSO was heated at 55 °C in the vacuum desiccator while the substrate and HBr container remained at the ambient temperature. The exposure time was varied to control the doping concentration in the film. DMSO-HBr adduct doping for MeO-TPD by solution process. 150 μΐ, of DMSO (Aldrich; anhydrous, >99.9%) was added into 500 of HBr (Aldrich: 48 wt. % in H ₂ 0, >99.99%) and the solution was mixed by a vortex mixer for 1 min. 10 of the DMSO-HBr solution (A.F .A.) was added to 2.5 mL of MeO-TPD (0.32 mM) solution in chlorobenzene (Aldrich; anhydrous, >99.9%). The solution was mixed by a vortex mixer for 1 min and then kept for 20 min at room temperature to prepare a heavily doped spiro-OMeTAD solution. 50 μΐ _^ of the heavily doped MeO-TPD solution was added to 150 μΐ. of un-doped MeO-TPD solution (80 mM) in chlorobenzene and the solution was kept for at least 10 minutes before the thin film preparation. The doped MeO-TPD solution was spin-coated onto the perovskite layer at 2500 rpm for 45 sec (acceleration: 500 rpm).

Z ⁾ M¾9-H5r-(±)-10-camphorsulfonic acid (CSA) doping: 30 mg of CSA (Aldrich, (±)-10- camphorsulfonic acid, 98%) was dissolved in 20 μΙ_, of DMSO and 300 μΐ _^ of chlorobenzene. The solution was vortexed until CSA was dissolved completely. 3 μΐ, of HBr (48 wt% in water) was added to the CSA solution and the solution was vortexed for 30 seconds. 50 μ L of DMSO-HBr-CSA solution was added into 500 μ L of spiro-OMeTAD solution (7.78 mM) in chlorobenzene. 5 - 50 μΐ _^ of heavily doped spiro-OMeTAD solution was added to the un- doped spiro-OMeTAD chlorobenzene solution (70 mM) and was kept for 15 min before the preparation of thin films. The spiro-OMeTAD solution was spin-coated onto the perovskite layer at 2500 rpm for 45 sec (acceleration: 500 rpm).

DMSO-HBr-CSA doping for gradient doping (method i): 30 mg of CSA (Aldrich, (±)-10- camphorsulfonic acid, 98%) was dissolved in 250 μΐ, of acetonitrile and 50 2- Ethoxyethanol. The solution was vortexed until CSA was dissolved completely. 3 μΐ. of HBr (48 wt% in water) and 20 μΤ, of DMSO were added to the CSA solution and the solution was vortexed for a further 30 seconds. Varied amount of the prepared DMSO-HBr-CSA solution was diluted to acetonitrile and 2-Ethoxyethanol (5: 1 volume ratio) mixed solution to control thickness of CSA layer. The DMSO-HBr-CSA solution was spin-coated on substrate or non- doped poly-TPD layer at 4000 rpm for 45 sec (acceleration: 500 rpm). The doped film was finally annealed on the hot plate at 100 °C for 10 min in ambient air.

DMSO-HBr-CSA doping for gradient doping (method ii): 30 mg of CSA (Aldrich, (±)-10- camphorsulfonic acid, 98%) was dissolved in 250 μL· of acetonitrile and 50 μΐ. 2- Ethoxyethanol. The solution was vortexed until CSA was dissolved completely. 10 μΧ of HBr (48 wt% in water) was added to the CSA solution and the solution was vortexed for a further 30 seconds. Varied amount of the prepared CSA-HBr solution was added to acetonitrile and 2-Ethoxyethanol (5 : 1 volume ratio) mixed solution to control thickness of CSA layer. The HBr-CSA solution was spin-coated on substrate or non-doped poly-TPD layer at 4000 rpm for 45 sec (acceleration: 500 rpm). 1.5 μL· of DMSO was put in crucibles in a vacuum desiccator (PELCO 2245; pressure -70 kPa) along with the HBr-CSA deposited on poly-TPD films. The crucible containing DMSO was heated at 55 °C in the vacuum desiccator. The exposure time was depended on the dopant thickness on the film (usually ~ 1 min). The doped film was finally annealed on the hot plate at 100 °C for 10 min in ambient air.

DMSO-HBr-CSA doping for gradient doping (method in): 30 mg of CSA (Aldrich, (±)-10- camphorsulfonic acid, 98%) was dissolved in 250 \L of acetonitrile and 50 \\L 2- Ethoxyethanol. The solution was vortexed until CSA was dissolved completely. 20 \\L of DMSO was added to the CSA solution and the solution was vortexed for a further 30 seconds. Varied amount of the prepared DMSO-CSA solution was added to acetonitrile and 2-Ethoxyethanol (5 : 1 volume ratio) mixed solution to control thickness of CSA layer. The DMSO-CSA solution was spin-coated on substrate or non-doped poly-TPD layer at 4000 rpm for 45 sec (acceleration: 500 rpm). 5 iL of HBr (48 wt% in water) was put in crucibles in a vacuum desiccator (PELCO 2245; pressure -70 kPa) along with the DMSO-CSA deposited on poly-TPD films. The HBr container remained at the ambient temperature, although the vacuum desiccator was heated at 55 °C on the hotplate. The exposure time was depended on the dopant thickness on the film (usually ~ 1 min). The doped film was finally annealed on the hot plate at 100 °C for 10 min in ambient air.

DMSO-HBr-CSA doping for gradient doping (method iv): 30 mg of CSA (Aldrich, (±)-10- camphorsulfonic acid, 98%) was dissolved in 250 of acetonitrile and 50 μΕ 2- Ethoxyethanol. The solution was vortexed until CSA was dissolved completely. Varied amount of the prepared CSA solution was added to acetonitrile and 2-Ethoxyethanol (5: 1 volume ratio) mixed solution to control thickness of CSA layer. The CSA solution was spin- coated on substrate or non-doped poly-TPD layer at 4000 rpm for 45 sec (acceleration: 500 rpm). 1.5 μΕ of DMSO and 5 μΕ of HBr were put in separate crucibles in a vacuum desiccator (PELCO 2245; pressure -70 kPa) along with the un-doped spiro-OMeTAD thin films on glass substrates. The vial containing DMSO was heated at 55 °C in the vacuum desiccator while the substrate and HBr container remained at the ambient temperature. The exposure time was depended on the dopant thickness on the film (usually ~ 1 min). The doped film was finally annealed on the hot plate at 100 °C for 10 min in ambient air.

DMSO-HBr-CSA doping for MeO-TPD: 30 mg of CSA (Aldrich, (±)-10-camphorsulfonic acid, 98%) was dissolved in 20 μΐ. of DMSO and 300 of chlorobenzene. The solution was vortexed until CSA was dissolved completely. 3 μί of HBr (48 wt% in water) was added to the CSA solution and the solution was vortexed for a further 30 seconds. 50 μL· of DMSO- HBr-CSA solution was added into 500 μΐ. of MeO-TPD solution (6.87 mM) in

chlorobenzene. 5 uL of the heavily doped MeO-TPD solution was added to 105 μΐ. of un- doped MeO-TPD chlorobenzene solution (80 mM) and was kept for 10 min before the preparation of thin films. The MeO-TPD solution was spin-coated on to the perovskite layer at 2500 rpm for 45 sec (acceleration: 500 rpm).

O2 doping for spiro-OMeTAD: A thin film of spiro-OMeTAD was stored in a desiccator overnight. This lead to doping of the thin film by O2.

Preparation of doped Spiro-OMeTAD film with DMSO-HBr adduct and extra additives for the optimized perovskite solar cell: 150 \L of DMSO (Aldrich; anhydrous, >99.9%) was added into 500 μΐ. of HBr (Aldrich: 48 wt. % in H ₂ 0, >99.99%) and the solution was mixed by a vortex mixer for 1 min. 10 L of the DMSO-HBr solution was added to 2.5 mL of spiro- OMeTAD (0.28 mM) solution in chlorobenzene (Aldrich; anhydrous, >99.9%). The solution was mixed by a vortex mixer for 1 min and then kept for 20 min at room temperature to prepare a heavily doped spiro-OMeTAD solution. 3 \L of the heavily doped spiro-OMeTAD solution was added to 1 mL of un-doped spiro-OMeTAD solution (70 mM) in chlorobenzene and the solution was kept for at least 10 minutes before the thin film preparation. 30 μΐ. of tert-butly pyridine (tBP) and 20 iL of bis(trifluoromethane)sulfonimide lithium salt (Li- TFSI) in acetonitrile were added to 1 mL of the spiro-OMeTAD solution as the extra additives along with DMSO-HBr. The spiro-OMeTAD solution was spin-coated onto the perovskite layer at 2500 rpm for 45 sec (acceleration: 500 rpm). Then 1.5 uL of DMSO and 5 μL of HBr were put in a vacuum chamber (pressure -70 kPa) along with the pre-doped spiro- OMeTAD thin film. The vial containing DMSO was heated at 55 °C in the vacuum chamber while the substrate and HBr container remained at the ambient temperature. The exposure time was 20 sec for the optimum condition. Fabrication of perovskite solar cell with FAo.83Cso.nPb(Io.8sBro.i 5)3: FTO (F-doped tin oxide) coated glass sheets (TEC7, 7Q/sq, Pilkington) were etched with zinc powder and 2M HCl to obtain the required electrode pattern. The sheets were then washed with a 2% Hellmanex solution in water followed by sonication for 10 minutes and then washed with deionized water, acetone, ethanol and isopropanol. The last traces of organic residues were removed by oxygen plasma cleaning for 10 minutes. Sn0 ₂ compact layers were then fabricated on the cleaned FTO substrates via spin coating and chemical bath deposition from an Sn0 ₂ precursor solution. The method for the preparation of Sn0 ₂ layer is as follows, 17.5 mg of SnCl4- 5H ₂ 0 (Sigma-Aldrich, 98%) was dissolved in 1 ml of iso-propanol. After 30 minutes of stirring at room temperature, a fully dissolved, clear solution was obtained. This solution was then filtered with a 0.2 mm pore size polytetrafluoroethylene filter. The SnCU solution was spun on FTO substrate at 3000 rpm for 30 s (acceleration: 200 rpm), and annealed at 180 °C for 60 min to form a thin layer of Sn0 ₂ . For chemical bath deposition, 108 mg of SnCl ₂ -2(H ₂ 0) (Sigma-Aldrich, 98%), 500 mg of Urea (Sigma-Aldrich, 98%), 500 mL of HCl (37% solution in water, Fisher Scientific) and 10 mL of 3-Mercaptopropionic acid (Sigma- Aldrich, 99%) were dissolved in 40 mL of H ₂ 0. After a fully dissolved the solution was obtained the spin-coated Sn0 ₂ substrates were immersed in the SnCl ₂ solution at 70 °C for 180 min. The substrates were immediately rinsed with deionized water and then sonicated for 120 s in a deionized water bath. The substrates were then dried with a nitrogen gun and annealed at 180 °C for 60 min. To obtain a FAo.s3Cso i7Pb(Io.85Bro.i5)3 precursor solutions, FAI (formamidinium iodide), Csl, PbBr ₂ and Pbl ₂ were dissolved in a mixed solvent of anhydrous Ν,Ν-dimethylformamide (DMF; Aldrich) and anhydrous dimethyl sulfoxide (DMSO; Aldrich) in 4: 1 volume ratio to obtain a stoichiometric solution with desired composition and a molar concentration of 1.4 M. The perovskite precursor solution was coated onto the Sn0 ₂ substrate by a consecutive two-step spin-coating process, step one at 1000 rpm for 10 sec and step two at 5000 rpm for 25 sec under low humidity (15-20% at 20°C) condition with a dry compressed air purge in a dry box. 200 μΐ _^ of anisole and chlorobenzene in 9: 1 volume ratio was dropped onto the substrate on the onto the pre- crystallized perovskite film for 8- 10 seconds during the second step of spin coating (i.e., at 5000 rpm). The perovskite films were annealed at 80 °C for 5 min and then 100 °C for 60 min in a box oven in ambient humidity condition. For both perovskite films, Spiro-OMeTAD thin films were deposited on the perovskite films as described in the methods mentioned above. 80 nm thick Au metal contact layers were deposited as the counter electrodes on the HTM layers by thermal evaporation.

Fabrication of perovskite solar cell with FAo.83Cso.i7Pb(Io.6Bro.4)3 ^' . FTO-coated glass sheets (TEC 15, 15Q/sq, Pilkington) were cleaned and etched with the procedure described above. Sn02 compact layers were fabricated from SnC>2 nanoparticles, which were synthesized via hydrothermal synthesis. Briefly, 467mg of SnCU 5Η ₂ 0 (Sigma-Aldrich, 98%) was dissolved in 20 ml of deionized water. After 10 minutes of stirring at room temperature, a fully dissolved, clear solution was obtained. This solution then transferred to a Teflon-lined stainless-steel autoclave and heated for 2 hours at 250 °C. The autoclave was brought to room temperature using cold water. The reacted solution was centrifuged at a speed of 10000 rpm for 10 min to collect Sn0 ₂ precipitates. The nanoparticles were re-dispersed in ethanol. To wash out the unreacted SnC , the solution was sonicated for 20 minutes and then centrifuged at a speed of 10000 rpm for 10 min. This washing treatment was repeated thrice. After the final washing treatment, the nanoparticles were again re-dispersed in ethanol. 2mg ml ^"1 Sn0 ₂ nanoparticle solution was spin coated on FTO substrates at 2000 rpm for 45 s, and annealed at 150 °C for 120 s to form a thin layer of SnC . Ceo (10 mg/ml in 1,2, di-chlorobenzene) layers were coated on SnC compact layers at 2000 rpm for 30 sec and the films were annealed at 100 °C for 10 min.

To obtain a FAo.83Cso.i7Pb(Io eBro.4)3 precursor solutions, FAI, Csl, PbBr ₂ and Pbl ₂ were dissolved in anhydrous Ν,Ν-dimethylformamide (DMF) to obtain a stoichiometric solution with desired composition and a molar concentration of 1.3 M. 75 of hydroiodic acid (57 wt% in H ₂ 0, Aldrich) and 37 ul of hydrobromic acid (48 wt% in H ₂ 0, Aldrich) was added to lml of 1.3M precursor solutions. After the addition of the acids, the solution was stirred for approximately 48 hours. The perovskite precursor solution was coated onto the SnO ₂ /C60 substrate by a consecutive two-step spin-coating process, step one at 1200 rpm for 20 sec and step 2 at 3000 rpm for 20 sec under low humidity (15-20% at 20°C) maintained by dry compressed air purge in a dry box. The perovskite films were quickly dried by the compressed air blowing for 20 sec. The perovskite films were kept at 20 °C for 15 min and then at 70 °C for 10 min on a hot plate. The dried films were then annealed at 175 °C for 90 min in a box oven without controlling the humidity of the ambient. For both perovskite films, Spiro-OMeTAD thin films were deposited on the perovskite films as described in the above- mentioned methods. 80 nm thick Au metal contact layers were deposited as the counter electrodes on the HTM layers by thermal evaporation.

Fabrication of organic light emitting diodes (OLEDs) with doped and undoped MeO-TPD as HTM: ITO (Indium tin oxide) -coated glass sheets (15 Ω/sheet) were sonicated with 2% Hellmanex in water for 10 minutes and then washed with deionized water, acetone, ethanol and isopropanol. The last traces of organic residues were removed by oxygen plasma cleaning for 10 minutes 50 mg of N, Ν,Ν ⁷ ,N ^r -Tetrakis(4-methoxyphenyl)benzidine (MeO-TPD) in chlorobenzene was stirred at 80 °C for 30 min. The MeO-TPD solutions (undoped or doped with DMSO-HBr/ DMSO-HBr-CSA) were spin-coated on to the ITO substrates at 2500 rpm for 45 sec (acceleration: 500 rpm). The substrate was then transferred to a vacuum chamber (CreaPhys GmbH, Germany) where Alq3, BPhen (both Luminescence Technologies, Taiwan; sublimed) and Al (K.I Lesker) were thermally sublimed at rates of 0.4 A/s, 0.4 A/s and 1.5 A/s respectively at a base pressure of < 10 ^"6 mbar. During deposition, the substrate was held at room temperature. The thickness of each layer was 100, 8 and 100 nm, respectively.

Characterization

Conductivity measurement: For conductivity measurements, organic field-effect transistor (OFET) chips from Fraunhofer IPMS (Dresden, Germany) with bottom-gate and bottom- contact geometry were used. The gate oxide was 90 nm thick on top of an n-doped Si wafer. The gold interdigitated fingers of the source and drain contacts had varying channel lengths of 2.5 pm, 5 pm, 10 pm and 20 pm and a channel length of 10 mm. Under ambient conditions, a Keithley 2400 source measurement unit was used to perform J-V measurements across the varying channel lengths with no applied bias at the gate.

UV-visible absorption spectroscopy: UV-vis absorption spectra were measured by a commercial spectrophotometer (Varian Cary 300 UV-Vis, USA). ~ 0.05 mg/ml of HTM in chlorobenzene or toluene was usually used as a solution phase. -100 nm of HTM thin films were used as a solid-state. The HTM thin films (thickness = -100 nm) were deposited on glass substrates by spin coating method.

Morphological characterization: A scanning electron microscope (SEM; Hitachi, S-4300) was used to acquire SEM images. The thickness of hole-transporting layers was measured by a stylus surface profiler (Dektak 150 Veeco Instruments, Inc.). Attenuated total reflectance Fourier transform infrared spectroscopy (ATR-IR) measurement: Agilent FTIR spectrometer with liquid nitrogen cooled Mercury Cadmium Telluride (MCT) detector was used for in situ IR spectroscopy. For multiple reflections of incident IR light a modified ATR accessory (GladiATR, PIKE Technologies) was placed inside the

spectrometer.

A Si trapezoidal ATR reflection element with a dimension of 5 x 8 x 1 mm3 and 39° angle of incidence at both short edges (Crystal GmbH, Germany) was used as the ATR crystal. The IR cell was developed in-house for in-situ experiments. The Si crystal was attached to the baseplate of the IR cell with epoxy glue to prevent leakage in the cell.

For reaction mechanism study by ATR-IR: First, the spectrum of DMSO in Cholobenzene was recorded, then vapors of HBr (direct addition of HBr in H2O solution was avoided to minimize the water content in the starting solution) allowed to interact with the DMSO in Chlorobenzene solution and IR spectrum was recorded. Then solution of Spiro-OMeTAD in chlorobenzene (70 mM) was added to the IR cell and IR spectrum was measured at different time intervals. IR data are presented as absorbance spectra with reference spectra collected at a bare Si ATR reflection element. Spectra were collected at a spectral resolution of 4 cm ^"1 and presented as 250 co-added scans.

For the detection of remnant DMSO/ DMSO:HBr adduct in the doped film by ATR-IR: 10 \L of A.F.A. were added to 2.5 mL of spiro-OMeT D, 0.28 mM solution in chlorobenzene (Aldrich; anhydrous, >99.9%). The solution was mixed by a vortex mixer for 1 min and then kept for 20 mins at room temperature to prepare a heavily doped spiro-OMeTAD solution. 100 of the doped spiro-OMeTAD was drop casted on a Si ATR crystal and the film was allowed to dry by keeping the ATR crystal at 40°C on a hot plate. The ATR crystal then brought to the room temperature and FTIR spectrum of the film recorded.

Gas chromatography-mass spectrometry (GC-MS) measurement: A.F.A. was added to Spiro- OMeTAD, 0.28 mM solution in chlorobenzene and the gaseous by-products from the solution was analyzed using a GC MS (Agilent Technologies 6890N Gas Chromatograph with an Agilent 5973 Mass Selective Detector and a 50QC2/BPI column from SGE Ltd) set up.

Quantum Chemical calculation: The geometry optimization and normal mode vibrational frequency calculations of DMSO, DMSO:HBr and DMSO:2HBr were performed at M06- 2X/aug-cc-pVDZ level. The optimized structures are the energy minimum structures as verified with all positive vibrational frequencies. The Gaussian 09 software was used for the calculations.

Van der Waals volume calculation: Marvin sketch software was used for the calculation.

Kelvin probe measurement: Samples for the Kelvin probe measurement was prepared by a spin coating method on an F-doped tin oxide (FTO) glass substrate. A Kelvin probe setup (probe diameter = 2 mm, KP Technology, UK) was used to measure the surface potential. Calibration of the Kelvin probe was done by a freshly cleaved highly ordered pyrolytic graphite surface which has known work function of 4.65 eV. The measurements were done in 20-25 °C in air.

UPS and XPS measurement: UPS and XPS measurements were carried out in an ultra-high vacuum (UHV) chamber with a base pressure of <2 xlO ^"10 Torr. Samples were prepared in nitrogen and transferred under nitrogen directly to UHV without air exposure. UPS was performed with both the He I (21.22 eV) and He II (40.81 eV) photon lines. For each UPS spectrum, the satellite lines of the He discharge lamp were carefully subtracted. The vacuum level was determined from the low-energy secondary-electron cutoff seen in the UPS He I spectra. XPS was performed with non-monochromatized Al Ka X-rays (1486.70 eV), and Origin Lab Pro was used for XPS peak fitting. A double-pass cylindrical mirror analyzer was used to measure the kinetic energy of electrons from UPS He I, UPS He II and XPS with an overall measurement resolution of 0.15 eV, 0.25 eV and 0.80 eV, respectively. The Fermi level reference for both UPS and XPS were determined with a clean Au substrate in electrical contact with the sample. Sample measurements were kept as short as possible to avoid degradation of the organic materials.

Ellipsometry measurement: 50 mg of N, N,N',N'-Tetrakis(4-methoxyphenyl)benzidine (MeO- TPD) in chlorobenzene was stirred at 80 °C for 30 min. The DMSO-HBr or DMSO-HBr- CSA doped MeO-TPD solution was spin-coated onto a glass substrate at 2500 rpm for 45 sec (acceleration: 500 rpm) which gave ~70 nm thin-film. An undoped MeO-TPD film (70nm) was deposited on top of the doped thin-film by thermal evaporation in a vacuum (base pressure < 10 ^"6 mbar). Ellipsometric measurements of the resulting films under thermal stress were carried out using an RC2 ellipsometer (J. A. Woollam). The thermal tolerance test was carried out on a hot plate in ambient atmospheric conditions. Thermal stress test: The thin films of hole-transporting materials were fabricated on top of a glass substrate for UV-vis absorption spectra and on top of an OFET chip for the conductivity measurement. The thermal tolerance test was carried out on a hot plate in ambient atmospheric conditions for both of UV-vis, ellipsometry and conductivity measurements. For the thermal stress test, the substrates were heated at 50, 80, 100 and 120 ^° C for 10 min at each heating step. Measurements were done after heating at each temperature.

Mott- Schottky analysis: Mott-Schottky analysis on thick devices has been performed according to (Kirchartz, J. Phys. Chem. C 2012, 116 (14), 7672). The measurement of the capacitance was conducted between -10 V to 10 V with steps of 50 mV using an Autolab PGSTAT302F (Metrohm) at 1 kHz. During all the measurements, the devices are kept in an air.

Characterization of PV devices: The current-voltage (J-V) measurements and stabilized power output (SPO) measurements were measured (2400 series source meter, Keithley Instruments) under both dark and simulated solar light (AAB ABET technologies Sun 2000 solar simulator) with its light intensity, 100 mW cm ^"2 (AM 1.5). The simulated light was calibrated against a standard amorphous-silicon PV cell (NREL-calibrated KG5 filtered silicon reference cell). The mismatch factor was estimated to be M=l .035405 and the lamp intensity was adjusted to account for this mismatch. The J-V curves were measured between 1.4V to 0V. The cell aperture area of light incidence was set to 0.0913 cm ² by employing an opaque mask with an aperture. The device stability test was carried out under full spectrum simulated AMI .5, 76 mA cm ^"2 irradiance at Voc using an Atlas SUNTEST XLS+ (1700W air-cooled Xenon lamp). All devices were aged without capsulation. The chamber temperature is ~ 50 ^° C. The light source is pulsed at 100Hz.

External quantum efficiency (EQE) measurement: EQE spectra were evaluated via custom- built Fourier transform photocurrent spectroscopy based on the Bruker Vertex 80v Fourier transform spectrometer. A Newport AAA sun simulator was used as the light source and the light intensity was calibrated with a Newport-calibrated reference silicon photodiode.

Characterization of OLED: The current density -voltage (J-V) measurements were measured using a source meter (2400 series source meter, Keithley Instruments). The luminescence intensity was measured along with bias voltage from source meter, and luminescence was collected using with a fiber-coupled detector (Ocean Optics USB 2000+) in an integrating sphere (Oriel Instruments 70682NS).

Results and Discussion

The ability of the proposed agents to dope a variety of hole transport materials (HTMs) in solution was tested. The structures of the six HTMs tested are shown below.

To dope these HTMs, a mixture of DMSO and HBr in 1 :2 molar ratio was prepared (this mixture is referred to as adduct forming agent (AFA)). The AFA was added to solutions of the HTM in chlorobenzene as described in the experimental sections above. Absorbance spectra of the HTM solutions before and after the addition of AFA show absorption features corresponding to the oxidized organic molecules after the addition of AFA. 2,2',7,7'- Tetrakis N,N-di(4-methoxyphenyl)amino]-9,9'-spirobifluorene (Spiro-OMeTAD) was taken as the testbed and used to investigate the evolution of the charged species in the solution. It was observed that the absorption peak related to spiro-OMeTAD ⁺ appears at ~510 nm as shown in Figure 11(a). On further increasing the concentration of AFA in the solution, the absorption peak at -690 nm corresponding to spiro-OMeTAD ⁺⁺ appears (Figure 11(a)). It was estimated that the doping efficacy defined by the molar ratio of oxidized spiro-OMeTAD and added DMSO was—17.5 % (see Fig. 11 (b)). The disappearance of the neutral spiro- OMeTAD peak at 389 nm on the absorption measurement was tracked in Figure 11(a), which allows for the calculation of the molar concentration of oxidized spiro-MeOTAD in chlorobenzene. The Ionization potentials of the HTMs and the conductivity of the HTM thin films before and after doping is shown in Table 2.

bisphenylbenzidine] (Poly-TPD) j

To explore the possibility of doping the HTM using vapor phase techniques, a spiro- OMeTAD thin-film of thickness -100 nm was exposed to a vapour of DMSO and HBr. The absorbance spectra of the thin-film was observed. As shown in Figure 12, there is a continuous increase in the absorbance corresponding to the doped species in the spiro- OMeTAD thin-film at 510nm and 690 nm. This indicates that the doping strategy is compatible with both solution and vapor phase processing.

To confirm that the change in absorbance spectra, characteristic of oxidized spiro-OMeTAD, is p-doping, conductivity measurements of the HTM thin-films doped with DMSO-HBr adduct in planar devices were made using inter-digitated Au electrodes. In Table 2, the conductivity values of the HTM thin-films before and after the doping are shown. A 2-3 order of magnitude increase in the conductivity values was observed for all the HTMs upon addition of the dopant. The doped spiro-OMeTAD reaches a maximum conductivity of around l-2xl0 ^"3 S cm ^"1 at a doping concentration of 50 mol% in solution phase. It is noted that the highest conductivity for the doped spiro-OMeTAD films for the present invention is similar to that obtained from a method of doping with a Co(III) TFSI complex. This indicates that the doping process has had a minimal negative impact upon charge carrier mobility.

To gain further insight into the doping effect, ultraviolet photoelectron spectroscopy (UPS), X-ray photoelectron spectroscopy (XPS) and Kelvin probe (KP) measurements were performed on HTM thin films. As expected for p-type doping, an increase in the work function of the thin-film and a shift of the fermi levels closer to the highest occupied molecular orbital (HOMO) of the OS were observed. The XPS spectra of the doped films showed C-ls and O-l s peaks shifting to lower binding energies with the increase in the concentration of AFA, which provides further confirmation of p-doping the material. Another HTM, N,N,N,N-Tetrakis(4-methoxyphenyl)benzidine (MeO-TPD), was tested which also shows similar UPS and XPS features on doping.

As shown in the absorbance spectra in Figure 13, it has been found that neither DMSO nor HBr can dope the HTMs alone. This indicates that it is essential for DMSO-HBr to come together for the doping process to happen. To understand the mechanism and the resulting products, the reaction process of DMSO-HBr doping was assessed by in-situ Attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR) and gas chromatographic mass spectrometry (GC-MS). The peak at around -1060 cm ^"1 in the FTIR spectrum corresponds to the absorbance of S=0 bond of DMSO. After the addition of HBr to DMSO in CB, it was observed that the absorbance of S=0 broadens and shifts to a lower wavenumber (around 1020-1050 cm ^"1 ), indicating the lowering in the bond strength of the S=0. Quantum chemical calculations of the possible interactions between DMSO and HBr were performed. It was determined that DMSO:HBr and DMSO:(HBr)2 molecular adducts can form and would be expected to have the observed IR absorbance due to the elongation of S=0 bond. The peaks at around -1035-1055 cm ^"1 in the FTIR were assigned to S=0 from the molecular adduct. Spiro-OMeTAD solution was added to the adduct solution and the change in the absorbance spectra over time was recorded. A continuous decrease in the peak intensity at -1050 cm ^"1 was observed indicating the consumption of the molecular adduct and breaking of the bond between sulfur and oxygen. Concurrently, an increase in the absorbance of OH stretching was also observed in the water absorbance region indicating that H2O is one of the by-products of the process. Dimethyl sulfide (DMS) was identified as another by-product via GC-MS of the released gases during the doping process

Based on the information corresponding to the by-products and the oxidation states of Sulfur containing molecules, the following mechanism is proposed for the doping. The DMSO: HBr adduct acts as the dopant in the process. The adduct accepts electron(s) from the organic molecule where the activated DMSO reduces to DMS, one of the by-products. The oxidation states of sulfur in DMSO and DMS are 0 and -2, respectively. The positive charge on the organic materials is counter balanced by the bromide ion from HBr. The oxygen from DMSO and H from HBr molecule then form H ₂ 0, another by-product. The probable mechanism is shown schematically in the Fig 2 (c). It is noted that at this point it is unclear whether the reaction proceeds via one-step two-electron transfer process or two steps with a one-electron transfer at each step, which is a matter of ongoing investigation.

As the by-products DMS, H ₂ 0, DMSO and HBr have relatively high vapour pressures compared to the HTM matrix, it is expected that the by-products and the un -reacted AFA (if any) should escape the OS matrix after the doping process under normal processing conditions. The JK and XPS spectroscopy can detect any unreacted DMSO or the trapped byproduct DMS. The FTIR spectrum of the spiro-OMeTAD film before and after doping with DMSO-HBr did not show any detectable absorbance from S=0, S-0 and OH bonds.

Absorbance for sulfur was also not observed in XPS spectra of doped films. This indicates that the doping process is clean, in that it does not leave behind by-products or unreacted dopants in the organic semiconductor matrix. Only doped HTM and counter ions remain after the doping reaction.

With conventional p-dopants like, F4TCNQ, 2,2'-(perfluoronaphthalene-2,6- diylidene)dimalononitrile (FeTC Q), 2-(3-(adamantan- l -yl)propyl)-3,5,6-trifluoro-7,7,8,8- tetracyano-quinodimethane (F3-TCNQ-Adl), and 3,6-difluoro-2,5,7,7,8,8-hexacyano- quinodimethane (F2-HCNQ), the doping process leaves behind a counter ion, the negatively charged dopant molecule, which can act a Coulombic trap to the newly-created hole, inhibiting its contribution to charge transport. In the above-proposed adduct -based doping, the HBr molecule plays a dual role: firstly as an activator and then as a provider of the anion. Crucially, the proposed doping mechanism should allow introduction of further, specifically- chosen counter ions, and hence decouple the dopant from the counter ions. To demonstrate that such a decoupling of the dopant and counter ion is feasible, 10-camphorsulfonic acid (CSA) and DMSO was used with the addition of HBr (see Figure 14). The doping ability of the DMSO-HBr-CSA system in Spiro-OMeTAD and MeO-TPD facilitated with HBr was measured by absorbance, conductivity, UPS and XPS studies. The evolution of the UV-vis spectrum of spiro-OMeTAD doped with DMSO-HBr-CSA is shown in Figure 15.

If DMSO-CSA is added on its own, no doping of the HTMs is observed. Surprisingly, only a very small concentration of HBr is required to activate the doping process. This may be rationalized by considering that, while DMSO and HBr form the adduct that does the doping, CSA can provide the counter ion in the form of 10-camphorsulfonate and the HBr may be regenerated for the further doping. It has been determined that the doping efficiency of the DMSO-HBr-CSA system is -22% for Spiro-OMeTAD and the conductivity values of the highest doped films are up to 2 ^χ 10 ^"4 , comparable to the best commercially available dopants. The use of CSA is one example, but in principle many other counter ions tailored to specific needs could also be used.

A critical factor for the utility of a doping process is that it must be stable under the sort of conditions it might experience during manufacturing of further layers in the device stack and in typical end use. In most optoelectronic devices, stability at elevated temperatures is required, which could be due to joule heating in LED operation, heating due to solar light exposure in a photovoltaic (PV) cell, or simply the requirement for a thermal processing step of one of the subsequent electronic layers or packaging. Here, conductivity and ellipsometric measurements were used to investigate the stability of the doped films under thermal stress. The sample preparation procedure is as follows: a first layer of -70 nm thick MeO-TPD doped with DMSO-HBr or DMSO-HBr-CSA is spin-coated, and on top of this another -70 nm thick layer of un-doped MeO-TPD is evaporated. If de-doping or migration of the doped organic molecule/counter ion occurs due to the thermal stress, it would be expected that the conductivity values and ellipsometric profile of the films would change. The phase difference (Δ) and amplitude ratio (Ψ) as a change in polarization as light reflects or transmits from the MeO-TPD films doped with DMSO-HBr over a period of 100 hours at 50 °C was measured. No change in Δ and Ψ was observed upon thermal stress, nor any change in the conductivity indicating that no de-doping and/or migration of the dopant occurred for either case.

However, when the stressing temperature is raised to 85 °C and a decrease in the conductivity for the DMSO-HBr doping is observed which indicates de-doping. In contrast, for the films doped with DMSO-HBr-CSA no change in the conductivity of the doped HTM was observed with prolonged stressing, even at a further elevated temperatures of 100 °C. It can be estimated via a Van der Waals surface calculation that the counterion camphorsulfonate is -7.7 times bulkier than the bromide. This is the likely origin of its reduced diffusivity and volatility and resultant thermal stability of the doping.

It is estimated that the cost of DMSO-HBr and DMSO-HBr-CSA dopant is -2200 times and 700 times lower than that of the commonly used dopant F4TCNQ, respectively.

Graded doping

Graded doping has been central to enabling organic light emitting devices, which are mainly prepared by vacuum deposition techniques, to reach the market. Here, the possibility of gradient doping by a solution process is explored. To that end, hole-only devices with poly- TPD (thickness =-700 nm) as the OS and F-doped Sn02 (FTO) and Au as electrodes (Figure 16a) were prepared. For gradient doping, the adduct activated CSA layers of different thicknesses were added either at the Poly-TPD:Au interface or at FTO:Poly-TPD interfaces. In Figures 16b and 16c, the current-voltage curves of the hole-only devices are shown. For the device with doped Poly-TPD:Au interface, for the positive bias (i.e., hole injection from the Au side), the current density increases with the increase of doped layer which is controlled via the thickness of the CSA layer. For the negative bias (i.e. hole injection from the FTO side), the current density remains almost of similar magnitude. The opposite trend is seen in the current-voltage curve for the devices with doped FTO:Poly-TPD interfaces. If the doping is universal across the film due to the diffusion of the dopants, then the increase in the current density due to the doping is expected to be same for both negative and positive bias which is not the case here which confirms the localized doping at the interfaces.

To further analyze the profile of doping, devices where 300 nm thick poly-TPD layer is sandwiched between FTO substrate and Au electrode were prepared. The OS in the devices is un-doped, universally doped or doped at the interfaces. Capacitance-voltage measurements were made and Mott-Schottky analysis was used to calculate the doping density (acceptor density (NA)) as a function of the distance from the metal-OS junction.

Figure 16d shows the NA profile of the un-doped and homogeneously doped sample. As expected, homogenous NA values are seen for both the samples but with a higher value for the doped sample. For the devices with doping at the interfaces, it is found that the NA values follow a gradient profile (Figure 16e). Closer to the interface, NA values are higher than that in the un-doped device and gradually decreases as the distance from the interface increases and finally become closer to the values for the un-doped samples. Figure 17 shows (a) current density-voltage curves and (b) accepter density-approximated depth profile of the hole-only devices before and after DMSO-HBr-CSA doping (0.3 mol% of A.F.A.).

Devices

Having established this doping method and economic feasibility, demonstration of its utility was carried out in two types of optoelectronic devices which benefit from doped charge transport layers: perovskite solar cells and organic light emitting diodes. For the photovoltaic (PV) cells, FAo.83Cso.i7Pb(Io.8Bro.2)3 was used as a photo-absorber sandwiched between doped spiro-OMeTAD as the hole transport layer and a compact SnC layer as the electron transporter, a schematic of which is shown in Figure 18(a). The spiro-OMeTAD is doped with all of the doping methods described above and compared to other methods commonly used, such as doping with Li-TFSI and oxygen or tris(2-(lH-pyrazol-l-yl)-4-tert- butylpyridine)cobalt(III) tri[bis(trifluoromethane)sulfonimide](Co(III)TFSI). In Figure 18(b) and (c) the current density-voltage (JV) curves and stabilized power output (SPO) best- performing devices measured under simulated sun light are shown. It is found that the devices where the spiro-OMeTAD are doped with DMSO-HBr (both solution and vapor process) and DMSO-HBr-CSA (solution process) show significantly improved power conversion efficiency (PCE) as compared to other types of common dopants (Figure 19). PCE was determined from the forward-bias to short-circuit (FB-SC) J curves to be 21.1 % (J _sc = 23.2 mA cm ^"2 , V _oc = 1.14 V, and FF = 0.79) and 21.3 % (J _c = 23.5 mA cm ^"2 , V _oc = 1.11 V, and FF = 0.81) for the device fabricated with DMSO-HBr and DMSO-HBr-CSA doping, respectively, in comparison to 18.7 % (J _sc = 22.8 mA cm ^"2 , V _oc of 1.08 V, and FF = 0.76) for the devices incorporating conventional Li-TFSI-oxygen doping. The J _sc measured under simulated sun light was corroborated by the external quantum efficiency (EQE) of the devices (See Figure 20). JV curve hysteresis still persists for all devices (see Figure 21 and Table 3), therefore the stabilized power output (SPO) for the devices was presented by measuring the photocurrent density over time at the maximum power point voltage that was determined from the FB-SC JV curves (Figure 18(c)). Table 3

Sweeps Jsc (mA cm ^"2 ) Voc (V) FF (%) PCE (%)

Doping method

FB→ SC 22.8 1.05 77.2 18.4

Li-TFSI-oxygen

SC→ FB 22.8 0.93 69.2 14.6

FB→ SC 22.8 1.14 78.6 20.6

DMSO-HBr

SC→ FB 22.8 1.08 72.9 18.1

FB→ SC 23.0 1.11 80.9 20.8

DMSO-HBr-CSA

SC→FB 23.0 1.05 66.8 16.3

It was determined that the PCE and photocurrent for the devices fabricated from the oxygen doping stabilize at 18.2 % and 19.8 mA cm ^"2 (at 0.88 V) respectively. Further improvement in the SPO for the devices fabricated from both of DMSO-HBr and DMSO-HBr-CSA doping were observed stabilizing at 19.8 % (J _mpp : 21.4 mA cm ^"2 and V _mpp : 0.92 V) and at 20.1 % (Jmpp. 20.9 mA cm ^"2 and V _mpp . 0.96 V), respectively. The doping capability of the DMSO- HBr on the Poly[N,N'-bis(4-butylphenyl)-N,N'-bisphenylbenzidine] (Poly-TPD) thin-film in an inverted architecture were also demonstrated for the perovskite cell (Figure 22). The PCE of the cell is -17.8 %.

The stability of perovskite solar cells fabricated with FAo.83Cso.i7Pb(Io.6Bro 4)3 as the absorber was investigated, where the HTMs are doped by Li-TFSI-oxygen and the DMSO-HBr DMSO-HBr-CSA . The non-encapsulated cells were stressed in ambient air under simulated AMI .5 76mWcm ^"2 sun light at 60°C. It was observed that devices where DMSO-HBr or DMSO-HBr-CSA dopes the HTM show significantly improved stability in comparison to the Li-TFSI-oxygen doped counterpart.

As proof of concept, simple OLEDs with Tris-(8 -hydroxy quinoline)aluminum (Alq3) as the emitting material sandwiched between MeO-TPD as the hole transport layer and

bathophenanthroline (BPhen) as the electron transport layer were also fabricated. The structure of the OLED is shown in Figure 18(d). The JV curves for the devices are shown in Figure 18(e), where larger current density is observed for any given applied voltage in the forward bias for OLEDs where the MeO-TPD layer is doped with DMSO-HBr or DMSO- HBr-CSA, in comparison to the un-doped device. The electroluminescence (EL) spectra of the OLEDs stacks is shown in Figure 18(f), which show an EL peak at 552 nm corresponding to the emission from Alq3. The luminescence-voltage curve is shown in Figure 18(f). At -18 V applied voltage, EL intensity is -500 and -600 cd/m ² for DMSO-HBr and DMSO-HBr- CSA doping, respectively. In contrast, the OLED with an un-doped HTM demonstrates only -100 cd/m ² . The improvement in EL may be attributed to the improved charge injection with doped HTMs. Although these OLEDs are not optimized, it is clear that the doping method according to the invention is useful in the manufacture of OLEDs via both vapour and solution processing.

In conclusion, the method of the invention provides a DMSO-adduct based p-dopant applicable for a variety of organic HTMs ranging from small molecules to polymers. The inexpensive dopant is compatible with both solution and vapor phase processing and can be handled in ambient conditions. The by-products of the reactions easily escape the organic matrix after the doping process is complete making it a very clean way of doping. From understanding the mechanism, the doping method of the invention is not restricted to DMSO- HBr adduct. In principle, it can be extended to other adduct systems where different sulfoxide containing molecules in combination with different activators can be used to exploit the clean and inexpensive doping method. It has also been shown that this method decouples the dopant from the counter ions allowing the electronic properties and thermal stability of the doped HTM to be tuned separately. Finally, it has been demonstrated that the usage of the doped HTM in perovskite-based devices and in OLED applications shows significant improvements in performance in both cases.