High work function with reduced phase separation of PSS in metal oxide modified PEDOT:PSS interlayers for organic photovoltaics

Abstract

A simple and effective approach to reduce leakage currents in organic photovoltaics (OPVs) is achieved by doping a commonly used hole transporting interlayer (HTL), poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate), PEDOT:PSS, with a metal oxide. The effects of three metal oxide dopants, WO_x, VO_x, and NiO_x, are investigated. From Raman spectra analysis it is found that conversion of the benzoid structure of pristine PEDOT:PSS to a quinoid structure increases after metal oxide doping during thin-film processing. X-ray photoelectron spectroscopy also reveals that metal oxides reduce phase separation between PEDOT⁺ and PSS⁻ polymeric chains, consistent with the decrease in thickness of the insulating phase separated PSS thin-film that forms on the surface of pristine PEDOT:PSS, from 62 Å to 40 ± 5, 57 ± 5, and 55 ± 5 Å, respectively for WO_x, VO_x, and NiO_x incorporation. For the case of WO_x doping, ultraviolet photoelectron spectroscopic depicts work function enhancement from 4.84 eV (PEDOT:PSS) to 5.15 eV (PEDOT:PSS:WO_x). By employing metal oxide incorporated PEDOT:PSS thin films as hole transporting layers in a bulk heterojunction (BHJ) organic photovoltaic (OPV) (poly(3-hexylthiophene) (P3HT) and C₆₀ derivative, [6,6]-phenyl-C₆₀-butyric acid methyl ester (IC₆₀BA)), leakage current decreased from ∼30 × 10⁻⁸ to ∼4 × 10⁻⁸ A cm⁻². The reproducibility of these effects of metal oxide doped PEDOT:PSS on OPV performance is also confirmed in two other BHJs (PTB7:PC₇₀BM and PBDTTT:EFT:PC₇₀BM).

---

1. Introduction

The doped semiconductor, poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate), PEDOT:PSS, is an extensively studied polymer, which is commonly used as a solution-processed electrode^1–3 that is applied in many optoelectronic devices such as organic photovoltaics (OPVs) and organic light emitting diodes^4,5 because of its relatively high conductivity and smooth morphology.^6,7 The conductivity and morphology of PEDOT:PSS basically rely on electrostatically attached PEDOT and PSS rings. PSS anions in PEDOT:PSS tend to electrostatically stabilize by formation of PSS⁻Na⁺, PSS⁻H⁺ or PSS⁻PEDOT⁺.⁸ PSS–H and PSS–Na are neutral species, and do not participate in coulombic interactions. As hydrophilic PSS–H, hydrophilic PSS–Na, and hydrophobic PEDOT may undergo transformation with selective treatment or doping, electrical properties and work functions of PEDOT:PSS can be regulated. Among numerous treatments, high boiling solvents and organic solvents have been successfully introduced, but only for conductivity enhancement of PEDOT:PSS thin films. Among high boiling solvents ethylene glycol (EG)² and dimethyl sulfoxide (DMSO)^9–11 were successfully added to an aqueous solution of PEDOT:PSS and among organic solvents sulfuric acid,³ EG,¹² acetone and ethanol,¹³ and methanol¹⁴ were successfully added to PEDOT:PSS, either by dripping or the solvent treatment method. In principle, structural transformations of the PEDOT ring, from a benzoid to quinoid structure, and phase separation of PSS from PEDOT seem to be the most acceptable interpretations that are used to explains the modifications in electrical properties.^11,15,16 The benzoid to quinoid phase transformation can easily be identified by a blue shift in Raman spectra, whereas the PSS phase separation can be identify by monitoring the S2p binding energy of PSS and PEDOT in X-ray photoelectron spectroscopy (XPS). However, solvent mixing or post treatment may affect the stability of PEDOT:PSS, as they disturb the weakly bonded PEDOT and PSS, leading to their rearrangement.

A recent study by the authors shows significant enhancement in the performance and stability of organic photovoltaics (OPVs) employing PEDOT:PSS that is doped with WO_x.¹ WO_x has an anionic behaviour that is similar to that of PSS, making it possible for WO_x to replace PSS in PEDOT:PSS. In the same study, it was also revealed that OPVs with WO_x doped PEDOT:PSS have reduced leakage currents. In this study, the origin of the of leakage current reduction by metal oxide (MO) doping is investigated. The effects of three metal oxide dopants, WO_x, VO_x, and NiO_x, are investigated. Results are analysed by means of X-ray photoelectron spectroscopy (XPS), Raman phase spectra, and ultraviolet photoelectron spectroscopic (UPS). Charge transport properties of metal oxide doped PEDOT:PSS thin films are studied by dark J–V characteristics of OPVs, utilizing, (poly({4,8-bis[(2-ethylhexyl)oxy]benzo[1,2-b:4,5-b′]dithiophene-2,6-diyl}{3-fluoro-2-[(2-ethylhexyl)carbonyl]thieno[3,4-b]thiophenediyl}) (PTB7):([6,6]-phenyl-C₇₁-butyric acid methyl ester) PC₇₀BM, (poly[4,8-bis(5-(2-ethylhexyl)thiophen-2-yl)benzo[1,2-b:4,5-b′]dithiophene-2,6-diyl-alt-(4-(2-ethylhexyl)-3-fluorothieno [3,4-b] thiophene)-2-carboxylate-2-6-diyl)]) PBDTTT:EFT:PC₇₀BM, and poly(3-hexylthiophene) (P3HT):indene-C₆₀ bisadduct (IC₆₀BA) bulk heterojunctions (BHJs).

2. Experimental

2.1. Synthesis and preparation of WO_x, VO_x, NiO_x, and PEDOT:PSS:metal oxide mixture

Preparation of WO_x, NiO_x, VO_x solutions. Aqueous WO_x solution is prepared by the hydration method described previously by Dong et al.^17,18 Initially, an ammonium metatungstate hydrate (AMH) powder purchased from Sigma Aldrich, containing 85% WO₃, is dissolved in water to form 1 mM solution, and then 2 M of hydrochloric acid are added, until the pH of the mixed solution reaches between 1 and 1.5. As previously reported,¹⁹ NiO_x 0.5 M prepared from nickel acetate of 0.5 M is dissolved in 2 M of LiOH by controlling the pH value to around 9.0. With centrifugal process, first green precipitate that occurs during the first phase is removed, followed by peptizing the solution with glacial acetic acid by controlling the pH to around 4.5. Viscosity of the prepared solution is controlled by the addition of a few drops of deionized water. A VO_x solution is prepared using vanadium pentoxide powder (V₂O₅, 99.99% purity) purchased from Sigma-Aldrich. 1 wt% of solution is prepared in 2-propanol.²⁰ These solutions are finally mixed with PEDOT:PSS (Heraeus Clevios™ 4083). Detailed information for fabrication of OPVs is given in ESI.†

2.2. Thin films and fabrication

PEDOT:PSS is mixed with sol–gel prepared WO_x solution in the ratio, (20 : 1, 10 : 1, 1 : 1, 1 : 3). The optimized PEDOT:PSS : WO_x ratio is 1 : 1, and the wt% of WO_x in the thin films is 1.11%. Similarly, the optimized blending v/v ratio for PEDOT:PSS : VO_x and PEDOT:PSS : NiO_x are found to be, 1 : 1 and 5 : 1, respectively. The wt% of VO_x and NiO_x are 4.99 and 2.46%, respectively. The prepared metal oxide:PEDOT:PSS mixture is stirred for 4 h followed by filtering through a 0.45 mm hydrophilic filter.

2.3. Thin films characterization and measurement

Raman spectra are collected with Renishaw Raman microscope in confocal imaging Raman mode using a 532 nm Nd-YAG green (E_g = 2.33 eV) laser operating at constant power. Raman data are analyzed using the WiRE project software. The Raman phase image of the selected peak is plotted relative to the x, y position. UPS is carried out by AXIS Ultra DLD (KRATOS Inc.). The measurements are performed using the He I photon line (hν = 21.22 eV) of a He discharge lamp under UHV conditions (4 × 10⁻⁸ Torr). Emitted photoelectrons are collected using a semi spherical channeltron with analyzer pass energy set to 5 eV, dwell time 100 ms, and energy resolution of 0.02 eV on a 55 μm area. Sample bias is set at −15 V. Fermi edge is calibrated using gold metal deposited on a high n-doped Si substrate. XPS measurements are performed in a PHI 5000 VersaProbe (Ulvac-PHI) with a background pressure of 6.7 × 10⁻⁸ Pa using a monochromatized Al Kα (hν – 1486.6 eV) anode (25 W, 15 kV). The spot size for XPS measurement is 100 μm × 100 μm. The current density–voltage (J–V) characteristics are recorded with a Keithley 2410 source unit.

3. Result and discussion

Raman spectroscopy is favourable for observing the chemical modification owing to its chemical selectivity. Characteristics of pristine PEDOT:PSS and PEDOT:PSS:MO are shown in Fig. 1a. The strong vibrational Raman band of pristine PEDOT:PSS occurs at 1442 cm⁻¹, and it shifts to 1443, 1441, and 1437 cm⁻¹, after WO_x, VO_x, and NiO_x doping, respectively. These bands contribute to the symmetric stretching of the aromatic C_α=C_β band which has about 13–20 cm⁻¹ red shift, consistent with previous reports.²¹ The red shift of C_α=C_β band indicates an increase in the conjugation length, which relate to the phase transformation of the PEDOT chain from the benzoid to quinoid structure (inset Fig. 1a) by means of doping. Asymmetry C_α=C_β peaks are missing, except for the peak at 1570 cm⁻¹.²² Additional Raman bands observed at relatively high wavenumbers, 1495, 1494, and 1502 cm⁻¹, after respectively doping PEDOT:PSS with WO_x, VO_x, and NiO_x, indicate de-doping of the PEDOT ring.²² The doping and de-doping of the PEDOT ring depend on the bipolaron (PEDOT²⁺), polaron (PEDOT⁺) or neutral state of PEDOT, as previously studied by Luo et al.²³ Luo et al. analyzed increased absorbance of post treated PEDOT:PSS thin films in the infra-red region via addition of an ionic liquid, EMMBF₄, in which PEDOT undergoes a change from the bipolaron, polaron to neutral state. There is thus a belief in the existence of different states of PEDOT. To elucidate this further, the symmetric C_α=C_β band is deconvoluted and the transformation from the benzoid to quinoid phase of PEDOT is confirmed by utilizing the Gaussian fit. Fig. 1b depicts the ratio of area and full width half maximum (FWHM) of the transformation from the benzoid to quinoid phase. The Gaussian fits and the Raman phase maps of symmetric C_α=C_β bands for pristine PEDOT:PSS and WO_x, VO_x and NiO_x doped PEDOT:PSS are given in Fig. 1c, d, e, f, g, h, i and j, respectively. Deconvolution of the C_α=C_β band reveals that the benzoid structure of PEDOT undergoes transformation after metal oxide incorporation. Furthermore, the Raman phase maps depict the phase of the C_α=C_β band of the PEDOT ring, which changes with metal oxide doping as indicated by the color changes. The phase of PEDOT is depicted by red color, it consists of relative small grains of PEDOT. After WO_x doing the Raman phase changes to dark red and bright red with VO_x and NiO_x mixture, and the PEDOT grains is enhanced by metal oxide doping. The color change in Raman phase spectra is observed due to Raman peak intensity of PEDOT grains.

Fig. 1 (a) Raman spectra of pristine PEDOT:PSS and metal oxide mixture. (b) Benzoid to quinoid phase transformation of pristine PEDOT:PSS thin films and after metal oxide doping. Deconvolution and Raman phase imaging of C_α=C_β symmetric peak of (c and d) PEDOT:PSS, (e and f) PEDOT:PSS:WO_x, (g and h) PEDOT:PSS:VO_x, and (i and j) PEDOT:PSS:NiO_x.

Three others important bands of PEDOT in a neat PEDOT:PSS thin film are occurred at 1377, 1261, 991 cm⁻¹, indicate the C_β–C_β stretching, C_α–C_α inter ring stretching, and oxyethylene ring deformation, respectively. However, in WO_x doped thin film, oxyethylene ring band is missing. Raman phase maps of C_β–C_β, C_α–C_α, and oxyethylene ring are depicted in Fig. S1 (ESI†). C_α–C_α band depicts the similar phase intensities. The grain size and the brightness of conductive PEDOT island is enhanced with MO doping. The significant phase change in C_β–C_β stretching band of PEDOT indicates the modification of C_β–C_β band with MO doping. Therefore, it is assume that the PEDOT ring undergoes chemical modifications, in results PSS interacted PEDOT modifies with MO doping.

The atomic force microscopy (AFM) images (Fig. S2†) of PEDOT:PSS and MO doped PEDOT:PSS are coated on a bare ITO. The smooth clusters of grains with few sub grains are dominating in neat PEDOT:PSS, and with MO doping, enlargement of these grains without sub grains have been noticed. The roughness of PEDOT:PSS, 2.08 nm, reduces to 1.70, 1.74, 1.95 nm with WO_x, VO_x, and NiO_x doping, respectively.

The XPS binding energy of neat PEDOT:PSS and PEDOT:PSS:MO thin films are discussed thoroughly. The presence of W4f, Ni2p, and V2p peaks confirm the metal oxide in PEDOT:PSS as shown in Fig. S3 (ESI†). C1s, O1s, S2p, and Na1s binding energy shifts are noticed with MO doping as shown in Fig. S4 (ESI†). Elemental concentration of C1s, S2p, and Na1s are reduced, and O1s is increased with MO doping. The differences in the chemical binding energy and the elemental concentration in MO doped PEDOT:PSS thin films indicate the phase separation of neat PEDOT:PSS due to MO doping. The elemental concentrations are summarized in Table S1.† Further, to elucidate the phase separation, S2p binding energy spectra are discussed in details, owing to its presence in both PEDOT and PSS.

The higher binding energy component with S(2p_3/2) peaks between 170 ± 0.2 and 166 ± 0.2 eV are associated with PSS–H and PSS–Na.²⁴ And, the lower binding energy S2p peaks between 166 ± 0.2 eV and 162 ± 0.2 indicate the presence of PEDOT⁺.²⁴ As depicted in Fig. S4c,† the PSS to PEDOT intensity ratio on the surface of thin film is very high and the ratio decreases with MO doping. This, predicts that the insulating PSS has partially separated from PEDOT and segregate on the surface. Thus, to calculate the PSS thickness, S2p of both PSS and PEDOT are deconvoluted by Gaussian fit, as shown in Fig. S5.† Phase segregated PSS of each films are calculated by using the following expression, d = λ ln[{(I_PSS/I⁰_PSS)/(I_PEDOT/I⁰_PEDOT)} + 1], where d is the thickness of PSS, λ represents the inelastic mean free path, I_x and I⁰_x are measured and expected signal intensities. The value of inelastic mean free path (λ) of ∼27 ± 3 Å (PEDOT:PSS) is taken from Clark et al.²⁵ PSS thickness in a neat PEDOT:PSS thin film is found to be 62 ± 5 Å, which is 20 Å thicker than the previous studies.^25,26 The difference of 20 Å could arise due to processing condition, annealing, storage, thin film fabrications or characterization (XPS sputtering rate).^26,27,28 Therefore, this number should be treated as a rough estimation.

Fig. 2a depicts the thickness of PSS after MO doping. With WO_x, VO_x, and NiO_x doping, it reduced to 40 ± 5, 57 ± 5, and 55 ± 5 Å, respectively. Furthermore, PSS to PEDOT ratio of neat PEDOT:PSS and PEDOT:PSS:WO_x thin films are examined by means of XPS depth profile. Binding energy spectra of C1s, O1s, S2p, and Na1s on the surface of thin films and at each 10 nm of sputtering are depicted in Fig. S6 (ESI†). PSS covered thin films, recovered to PEDOT rich region after 10 nm of sputtering as depicted in Fig. 3b. Also, a small concentration of S2p diffuses into ITO. The thickness of PEDOT:PSS and PEDOT:PSS:WO_x thin films are 40 and 28 nm, respectively.

Fig. 2 Thickness of PSS calculated from S2p of PSS and S2p of PEDOT in PEDOT:PSS and after metal oxide (WO_x, VO_x, and NiO_x) incorporation. (a) PSS thickness at the interface of PEDOT:PSS, and PEDOT:PSS mixture with WO_x, VO_x, or NiO_x. (b) S2p (PSS) concentration gradient from the top interface (BHJ) to the bottom interface (ITO) with and without WO_x doped PEDOT:PSS layers. The optimized thickness of pristine PEDOT:PSS and WO_x doped PEDOT:PSS thin films are 40 and 28 nm respectively. Thin films for XPS measurement were deposited on the bare glass.

Fig. 3 UPS spectra of pristine PEDOT:PSS and metal oxide doped PEDOT:PSS thin films. Two different ratio (10 : 1, 1 : 1) of WO_x doped PEDOT:PSS along with the optimized wt% of VO_x and NiO_x in PEDOT:PSS showed the blending effect on the (a) Fermi level, E_f, (b) HOMO level, and (c) work function of these thin films. Based on the UPS energy alignment (d) HOMO energy alignment, dipole moment (Δ), were calculated. Thin films were coated on ITO. Energy alignment of donor materials have been taken from the literatures.^29,30

The UPS spectra of MO doped PEDOT:PSS is shown in Fig. S7.† PEDOT:PSS is denoted as PP, and PEDOT:PSS doped WO_x is denoted as PP:WO_x and so on. The UPS is carried out of PP, PP : WO_x (10 : 1), PP : WO_x (1 : 1), PP:VO_x, and PP:NiO_x thin films deposited on ITO. Work function (ϕ), dipole formation (Δ), Fermi energy (E_f) and least unoccupied molecular orbital (LUMO) of MO doped PEDOT:PSS are measured as shown in Table 1. The low binding energy spectra (Fig. 3a and b) depict the E_f and LUMO, and higher binding energy spectra (Fig. 3c) depicts the ϕ. The E_f shifts of MO doped PEDOT:PSS are depicted in Fig. 3a. Heavily doped PEDOT in PEDOT:PSS should be metallic (no shift in E_f). However, PEDOT:PSS thin film is covered by 62 ± 5 Å of PSS, shifts the E_f by 0.42 eV. And, with WO_x, VO_x, and NiO_x doping, where, PSS thickness is thinner, it shifts the E_f by 0.10, 0.11 and 0.35 eV, respectively. The values of E_f, ϕ, Δ, and LUMO are reported in Table 1. The LUMO of MO doped PEDOT:PSS thin films are depicted in Fig. 3b. LUMO of PEDOT:PSS is observed at 2.16 eV, and this is shifted to 1.98, 2.04, 2.20 eV with WO_x, VO_x, and NiO_x doping, respectively. PSS–ITO interactions results in the formation of dipole moment of dipole energy of 0.68 eV with PEDOT:PSS thin film. Dipole energy increases to 0.99 eV with WO_x doping, probably due to PSS⁻/WO_x⁻ or WO_x⁻/PEDOT⁺ interaction with ITO. Furthermore, secondary cut off region of UPS spectra (Fig. 3c) depicts the work functions of PEDOT:PSS:MO thin films. Work function of neat PEDOT:PSS is found to be 4.84 eV. With WO_x, NiO_x and VO_x doping, this increased to 5.15, 4.79 and 4.96, respectively. The cyclic voltammetry (Fig. S8†) of neat WO_x doped PEDOT:PSS is carried out to examine the HOMO levels. Measurements are carried out by fixing the HOMO level of PEDOT:PSS to 5.0 eV. Further, the slope of current–voltage plots, intersecting at voltage axis are noted. The increment in the slope on the voltage axis indicates the deeper HOMO level. The HOMO of PP : WO_x (1 : 1) thin films increased by 0.36 eV (5.36) wrt. neat PEDOT:PSS (5.0 eV). The similar results are observed in the UPS, as the work function of pristine PEDOT:PSS (4.84 eV) increase to 5.15 eV after WO_x doping. Schematic in Fig. 3d is plotted by extracting UPS parameters.

Table 1 Work function (ϕ), dipole (Δ), Fermi shift (E_f) and least unoccupied molecular orbital (LUMO) of PEDOT:PSS and metal oxide incorporated PEDOT:PSS. ITO was used as the reference

Parameters	PEDOT:PSS (PP)	PP : WOx (10 : 1)	PP : WOx (1 : 1)	PP:VOx	PP:NiOx
ϕ (eV)	4.84	5.08	5.15	4.79	4.96
Δ (eV)	−0.68	−0.92	−0.99	−0.63	−0.80
Ef (eV)	0.42	0.24	0.10	0.11	0.35
LUMO (eV)	2.16	2.10	1.98	2.04	2.20

---

Dark current–voltage (J–V) characteristics of regular geometry OPVs with three different BHJs are shown in Fig. 4. PTB7:PC₇₀BM, PBDTTT:EFT:PC₇₀BM, and P3HT:IC₆₀BA are named as BHJ1, BHJ2 and BHJ3, respectively. OPVs are fabricated for BHJ1–BHJ3, where doping concentration of WO_x in PEDOT:PSS is varied from 1 : 0, 20 : 1, 10 : 1, 1 : 1, and 1 : 3. Blend 1 : 0 depicts the devices with neat PEDOT:PSS interlayer (Fig. 4a–c). Furthermore, OPVs with optimized blend of WO_x (1 : 1), VO_x (10 : 1), and NiO_x (5 : 1) are fabricated utilizing BHJ3 (Fig. 4d). Similarly, J–V characteristics of regular geometry OPVs under light illumination are depicted in Fig. 5. The leakages currents of these devices are collected at −0.9 V (V_oc ∼ 0.9 ± 0.05 V). OPVs depict consistent reduction in the leakages with increasing WO_x concentrations and optimized at 1 : 1 blending ratio. OPVs fabricated with neat PEDOT:PSS interlayer yields leakage currents of 30, 95.2, and 107.0 × 10⁻⁸ A cm⁻² for BH1, BHJ2 and BHJ3 respectively. While OPVs with PEDOT:PSS : WO_x interlayers yielded leakage currents of 4.0, 10.30, and 5.41 × 10⁻⁸ A cm⁻² for BHJ1, BHJ2, and BHJ3 at 1 : 1 blending ratio, respectively. OPVs fabricated with PEDOT:PSS:WO_x yielded close to one order (BHJ1, BHJ2) and two order (BHJ3) of reduction in the leakage currents (Table S2†). OPVs fabricated with VO_x and NiO_x doping, produce leakage currents of 42.8, and 22.69 × 10⁻⁸ A cm⁻², utilizing BHJ3. PCE of regular OPVs with BHJ1, BHJ2 and BHJ3 for neat PEDOT:PSS devices are found to be 6.45, 6.12 and 3.75%, respectively (Table S3†). With WO_x doping, PCE is improved by 13.18% (7.30%, BHJ1), 13.89% (6.97%, BHJ2) and 16% (4.35%, BHJ3). The improved performances of OPVs are determined by increased shunt resistance (R_sh), decreased series resistance (R_s), low leakages (J₀), improved ideality factor (n), and high rectification ratio (RR). The detailed information of R_sh, R_s, J₀, n, and RR are given in Table 2. High external quantum efficiency (EQE) with MO doping depicts the efficient collections of charges (Fig. S9†).

Fig. 4 Dark current–voltage characteristics of regular OPVs depicting leakages at reverse bias of pristine PEDOT:PSS device and after WO_x mixture in the following ratio, 20 : 1, 10 : 1, 1 : 1, and 1 : 3. Reproducibility of WO_x mixed PEDOT:PSS for an effective anode hole transport interlayer was examined through three different active layers, (a) PTB7:PC₇₀BM, (b) PBDTTT-EFT:PC₇₀BM, (c) P3HT:IC₆₀BA in a regular OPV. Insert in (b) depicts the regular geometry structure. Dark J–V characteristics of the regular OPVs with (d) pristine PEDOT:PSS hole transport materials and optimized metal oxides (WO_x, VO_x, and NiO_x) doped PEDOT:PSS interlayers utilizing P3HT:IC₆₀BA as the active layer.

Fig. 5 Current–voltage characteristics of regular OPVs under light illumination of pristine PEDOT:PSS device and after WO_x mixture in the following ratio, 20 : 1, 10 : 1, 1 : 1, and 1 : 3. WO_x mixed PEDOT:PSS was examined through three different active layers, (a) PTB7:PC₇₀BM, (b) PBDTTT-EFT:PC₇₀BM, (c) P3HT:IC₆₀BA in a regular OPV. J–V characteristics of the regular OPVs with (d) pristine PEDOT:PSS hole transport materials and optimized metal oxides (WO_x, VO_x, and NiO_x) doped PEDOT:PSS interlayers utilizing P3HT:IC₆₀BA as the active layer.

Table 2 (a) Shunt resistance, R_sh, series resistance, R_s, saturation current density, J₀, ideality factor, n, and rectification ratio, RR, in PTB7:PC₇₀BM, PBDTTT:EFT:PC₇₀BM, and P3HT:IC₆₀BA BHJ in regular OPVs, utilizing PEDOT:PSS and metal oxide:PEDOT:PSS mixture used as a HTL

Regular geometry OPVs		1 : 0	20 : 1	10 : 1	1 : 1	1 : 3	PP:VOx	PP:NiOx
PTB7:PC70BM	Rsh (Ω)	2099	6439	6199	8748	4472
	Rs (Ω)	1.80	1.64	1.41	1.02	2.23
	log J0 (mA cm^−2)	−7.40	−7.78	−7.87	−8.01	−7.68
	n	1.60	1.56	1.54	1.49	1.61
	RR	78 220	128 870	156 339	206 524	53 580
PBDTTT:EFT:PC70BM	Rsh (Ω)	1309	1372	5293	8134	2842
	Rs (Ω)	1.55	1.45	1.42	1.32	1.89
	log J0 (mA cm^−2)	−8.27	−8.71	−8.69	−8.72	−8.24
	n	1.59	1.53	1.49	1.42	1.58
	RR	10 346	10 623	46 381	92 619	9056
P3HT:IC60BA	Rsh (Ω)	1369	2505	4374	5055	2124	2586	2512
	Rs (Ω)	3.10	2.98	2.63	2.39	8.47	2.76	2.43
	log J0 (mA cm^−2)	−8.78	−8.94	−9.28	−9.75	−6.74	−9.11	−9.34
	n	1.59	1.68	1.57	1.42	1.93	1.47	1.45
	RR	2283	9653	38 367	59 879	81	10 184	8902

---

The origin of reduction in the leakage currents is likely due to very high increment in the shunt resistance as depicted in the Table 2. Reduction in the series resistance with MO doping, explains improved ohmic contact as depicted in Fig. S10.† Hole only devices (HOD) has the following structure; ITO/PEDOT:PSS:MO/Al. Fundamentally, in diode equivalent circuit, the net dark current density (J_dark) accounts shunt current density (J_sh), and diode current density (J_d). Current density in a diode can be expressed as, eqn (1);

J = J_ph − J_dark = J_ph − (J_sh + J_d). (1)

Where J_ph is photocurrent density. Further, the diode current can be expressed by the following equation, eqn (2);

J = J_ph − [{(qV − JR_s)/R_sh} + J₀ exp{(qV − JR_s)/nK_bT} − 1]. (2)

The J₀ is the reverse saturation current density. R_s and R_sh are denoted to series and shunt resistances, and, n, K, and T are diode quality factor, Boltzmann constant, and temperature, respectively. R_sh, R_s, J₀, and n of the regular geometry OPVs for the different bulk heterojunctions (BHJ1–BHJ3), are shown in Table 2. Current through the shunt can be ignored if the R_sh is high (eqn (2)) and V_oc (J = 0) can be written as:

V_oc = (nK_bT/q)ln{(J_ph/J₀) + 1} ∼ ln(J_ph).

Where, J_sc correlates linearly and V_oc shows logarithmic dependency on photo generated current density. From the dark J–V curves, leakage current is defined as current density at arbitrary bias and R_sh is calculated from the slope of the J–V curve at reverse bias. A high R_sh device (such as BHJ3, PP : WO_x (1 : 1)) has high fill factor (FF), clearly reduces the OPV performances due to decrease in FF as R_sh decreases drastically (such as BHJ3, PP : WO_x (1 : 3)). Whereas, V_oc and J_sc decreases to a lesser extent because at short circuit conditions, no current flow through R_sh. However, the dark leakage current increases by more than one order of magnitude on going from R_sh = 5055 to 2124 Ω. Therefore, it implies that the R_sh is the limiting resistance in reverse polarity to control the leakage current in MO doped PEDOT:PSS devices. Series resistance in OPVs with neat PEDOT:PSS interlayers is likely due to a thick insulting layer of PSS. PSS thickness of neat PEDOT:PSS and MO doped PEDOT:PSS at BHJ interface may significantly alter the shunt current. PSS thickness due to phase separation in pristine and MO doped PEDOT:PSS increases with increase in the shunt resistance.

4. Conclusion

In this study we produces a simple and efficient anode buffer layer for the optoelectronic devices. The metal oxide (WO_x, VO_x, and NiO_x) doped PEDOT:PSS is used as an anode buffer layer in OPVs. Metal oxide doping reduces the leakage currents significantly in OPV devices. MO doped PEDOT:PSS are highly reproducible, they play a crucial rule to reduces the leakage current and improve the PCE of OPVs by more than 13% for PTB7:PC₇₀BM, and PBDTTT:EFT:PC₇₀BM, and 16% for P3HT:IC₆₀BA bulk heterojunction. Spectroscopic study of MO doped PEDOT:PSS reveals that the reduction in shunt currents in OPVs is related to the phase separation of insulating PSS at the interface. WO_x doping in PEDOT:PSS enhances the dipole moment at the interface, increases the work function of PEDOT:PSS by 0.29 eV. PSS forms a thick insulating barrier of 62 Å on the surface of neat PEDOT:PSS, reduces to 40 ± 5, 57 ± 5, and 55 ± 5 Å with WO_x, VO_x, and NiO_x doping, respectively. This contributes in the reduction of leakage current from ∼30 × 10⁻⁸ to ∼4 × 10⁻⁸ A cm⁻².

References

1. A. Kanwat and J. Jang, J. Mater. Chem. C, 2014, 2, 901–907.
2. Y. H. Kim, C. Sachse, M. L. Machala, C. May, L. Muller-Meskamp and K. Leo, Funct. Mater., 2011, 21, 1076.
3. Y. J. Xia, K. Sun and J. Ouyang, Adv. Mater., 2012, 24, 2436.
4. G. Heywang and F. Jonas, Adv. Mater., 1992, 4, 116–118.
5. Q. Pei, G. Zuccafrello, M. Ahlskog and O. Inganas, Polymer, 1994, 35, 1347–1351.
6. A. N. Aleshin and S. R. Williams, Synth. Met., 1998, 94, 173–177.
7. X. Crispin, F. L. E. Jakobsson, A. Crispin, P. C. M. Grim, P. Andersson, A. Volodin, C. V. Haesendonck, M. V. D. Auweraer, W. R. Salaneck and M. Berggren, Chem. Mater., 2006, 18, 4354–4360.
8. J. S. Yeo, J. M. Yun, D. Y. Kim, S. Park, S. S. Kim, M. H. Yoon, T. W. Kim and I. S. Na, ACS Appl. Mater. Interfaces, 2012, 4, 2551–2560.
9. S. H. Lee, H. Park, S. Kim, W. Son, I. W. Cheong and J. H. Kim, J. Mater. Chem. A, 2014, 2, 7288.
10. Y. H. Zhou, H. Cheun, S. Choi, W. J. Potscavage, C. F. Hernandez and B. Kippelen, Appl. Phys. Lett., 2010, 97, 153304.
11. K. Ruit, R. I. Cohen, D. Bollen, T. Mol, R. Y. Rozen, R. A. J. Janssen and M. Kemerink, Adv. Funct. Mater., 2013, 23, 5778.
12. G. H. Kim, L. Shao, K. Zhang and K. P. Pipe, Nat. Mater., 2013, 12, 719.
13. Y. Xia and J. Quyang, J. Mater. Chem., 2011, 21, 4927–4936.
14. K. Sun, Y. Xia and J. Quyang, Sol. Energy Mater. Sol. Cells, 2012, 97, 89–96.
15. S. Ashizawa, R. Horikawa and H. Okuzaki, Synth. Met., 2005, 153, 5.
16. N. Kim, B. H. Lee, D. Choi, G. Kim, H. Kim, J. R. Kim, J. Lee, Y. H. Kahng and K. Lee, Phys. Rev. Lett., 2012, 109, 106405.
17. X. Dong, J. Liu, J. Yan, J. Wang and G. Hong, Rare Met. Mater. Eng., 2005, 34(3), 421–424.
18. R. Ganesan and A. Gedanken, Nanotechnology, 2008, 19, 025702.
19. A. Kanwat, W. Milne and J. Jang, Sol. Energy Mater. Sol. Cells, 2015, 132, 623–631.
20. J. Kim, A. Kanwat, H. M. Kim and J. Jang, Phys. Status Solidi A, 2015, 212, 640–645.
21. X. Zhang, D. Chang, J. Liu and Y. Luo, J. Mater. Chem., 2010, 20, 5080–5085.
22. A. A. Farah, S. A. Ruledge, A. Schaarschmidt, R. Lai and J. P. Freedman, J. Appl. Physiol., 2012, 112, 113709.
23. J. Luo, D. Billep, T. Waechtler, T. Otto, M. Toader, O. Gordan, E. Sheremet, J. Martin and M. Hietschold, J. Mater. Chem. A, 2013, 1, 7576–7589.
24. J. Hwang, F. Amy and A. Kahn, Org. Electron., 2006, 7, 387–396.
25. D. T. Clark and H. R. Thomas, J. Polym. Sci., Polym. Chem. Ed., 1977, 15, 2843.
26. G. Greczynski, T. Kugler and W. R. Salaneck, Thin Solid Films, 1999, 354, 129–135.
27. C. H. Chang and S. A. Chen, Appl. Phys. Lett., 2007, 91, 103514.
28. Y. J. Lin, F. M. Yang and C. Y. Huang, Appl. Phys. Lett., 2007, 91, 092127.
29. S. Zhang, L. Ye, W. Zhao, D. Liu, H. Yao and J. Hou, Macromolecules, 2014, 47, 4653–4659.
30. Y. Liang, Z. Xu, J. Xia, S. T. Tsai, Y. Wu, G. Li, C. Ray and L. Yu, Adv. Mater., 2010, 22, 135–138.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra23170e

---