The damaging effects of the acidity in PEDOT:PSS on semiconductor device performance and solutions based on non-acidic alternatives

Poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate), PEDOT:PSS, has been widely used as an effective hole transporting material in many different organic semiconductor devices for well over a decade. However, despite having many strong features which make this material such a popular hole transport/injection layer, PEDOT:PSS is well-known to cause degradation in devices and limit their stability due to the acidity of the PSS chain. This review focusses on the attempts that have been made to combat this problem, with different strategies explored, including the development of neutral analogues, use of alternative materials and the introduction of barrier layers to prevent degradation of the electrode. Since solution-processing is a key advantage of using PEDOT:PSS, we concentrate on analogous materials that can also be solution-processed, with particular attention on whether orthogonal processing can be retained. We intend this work to be a useful guide for researchers considering enhanced device lifetimes, an important parameter when considering organic semiconductor devices for commercialisation.


Introduction
Poly (3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS, Figure 1), is widely used in organic electronic devices, such as organic photovoltaics (OPVs) and organic lightemitting diodes (OLEDs), whilst also being used in organic-inorganic hybrid devices such as perovskite-based solar cells, as a hole transport layer (HTL). This material is popular due to its high conductivity and transparency, the ability to reduce surface roughness when coating ITO and its suitability for orthogonal processing. Indeed, it can be considered the default hole transport material (HTM) for solution-processed organic semiconductor devices and has been applied successfully in some benchmark OPV 1-4 (including the current record PCE device) 5 and OLED devices 6-8 over the last 10 years. However, there are disadvantages in using PEDOT:PSS, which include acidity, hygroscopicity, anisotropic charge injection 9 and batch-to-batch variation in electrical and physical properties. For this reason, there has been much research into developing materials to replace PEDOT:PSS in organic electronic devices.  Table 1.
In this review we will focus on the impact of the acidity of PEDOT:PSS on device performance and stability. Furthermore, we will discuss the steps that can be taken to neutralise PEDOT:PSS or prevent degradation due to acidity. The discussion will be mainly concentrated on alternative HTLs, where the materials possess a stabilised ionisation energy and high ptype mobility or hole injection layers (HILs), where often a thin layer is deposited on the anode to modify the work function and improve hole injection. As solution-processing is one of the main advantages of using PEDOT:PSS, this article presents analogues or alternatives that can be solution-processed. If one is looking for a review on how to enhance the conductivity of PEDOT:PSS then this has been nicely summarised by Xu et al, 11 whilst a comprehensive overview of transparent electrode materials in general can be found in a recent article by Cloutet et al. 12 Many of the approaches taken to enhance the conductivity of PEDOT:PSS involve treatment with acids such as formic acid, 13 sulfuric acid, 14 phosphoric acid 15 or the addition of high boiling point solvents, such as DMSO 16 or NMP. 17 The strong acidity of high conductivity grades of PEDOT:PSS could be problematic for stability or limit the choice of compatible active materials, yet as these developments are relatively recent the literature has not reported these obvious concerns.  detected a significant indium content in PEDOT:PSS films after casting, which increased over time with both thermal annealing and exposure to air. 22 Therefore, despite the advantages of acidity in boosting conductivity, there is a need, particularly for applications incorporating an HTL, for nonacidic conductive materials to improve device lifetime.
Whilst ITO is widely used in organic electronics, alternative materials have also been explored, particularly due to the low abundance of indium. However, these technologies are also vulnerable to acid-induced degradation. It should be noted that there has also been a report that PEDOT:PSS has not been detrimental to the stability of silver nanowires. 23 Understanding the exact nature of degradation in other metals/electrodes is therefore of key importance for the development of organic semiconductor devices based on PEDOT:PSS and it analogues.
The acidity of PEDOT:PSS as a hole transport material can have a detrimental consequence on device performance due to its effect on materials coated on top of it. In 2003, Brunner and co-workers 24 studied poly(p-phenylenevinylene) (PPV, Figure 2) and poly (2,7- Degradation of LED performance has also been observed for iridium-based complexes as Baranoff et al. 25 showed a blue phosphorescent iridium complex degrading in the presence of PEDOT:PSS, highlighting that ligand stability can be affected by acidity. 25 The acidic nature of PEDOT:PSS has also been shown to be detrimental to some donor materials for OPV devices. The degradation in the performance of organic semiconductor devices can be extremely difficult to monitor due to the fact there are multiple layers and many competing processes.
Most organic electronic devices are sensitive to degradation by exposure to air and moisture.
Krebs et al. 26 were able to demonstrate phase segregation of PEDOT and PSS chains over time with exposure to oxygen and moisture. In inverted devices this leads to selective oxidation of the PEDOT chains and, subsequently, the layers underneath. Also, in constructing OPV devices based on MDMO-PPV:PC61BM as the active layer with and without a PEDOT:PSS layer, Durrant et al. 27 were able to use a series of different environmental conditions to show that PEDOT:PSS absorbed water from a humid nitrogen atmosphere causing increased resistivity.
Such observations should be considered as part of the overall understanding of the PEDOT:PSS-assisted degradation mechanism. Indeed, it is likely that a combination of the hygroscopic nature and the acidity of the PSS group leads to breakdown of device performance over time. However, this review will discuss the specific problems that have affected device performance due to acidity and how these challenges can be overcome.
Many of the examples discussed show that a simple substitution of an acidic PEDOT:PSS layer for a non-acidic analogue can result in improved device stability. In summarising the different approaches that can be pursued, it is hoped that this review can be a useful source for researchers looking to avoid acid-induced degradation of semiconductor devices.

Barrier layers to prevent electrode degradation
A number of different strategies have been used to avoid diffusion of indium from ITO into active layers of organic semiconductor devices. One common tactic to avoid this degradation is to separate the interface between ITO and PEDOT:PSS by placing a barrier between the two layers. For example, Wang et al. 28 introduced a double layer of PEDOT:PSS, with pH neutral PEDOT:PSS on top of ITO and the regular solution deposited on top of the neutral layer. The indium etching was shown to be reduced by a factor of 6.5 and the device performance was even improved upon. 28 In another approach, there have been attempts to coat the ITO with a self-assembled monolayer (SAM) to prevent etching and subsequent migration of indium into the hole transport and active layers. In fact, Tai et al. 29 showed that even acidic SAMs such as terephthalic acid and derivatives with four and six carboxylic acid groups (Figure 3), could be placed in the interface between ITO and PEDOT:PSS to suppress the migration of indium. The SAM-modified device only showed a 30% reduction in power conversion efficiency (PCE) over 49 days, whereas the reference device with an ITO/PEDOT:PSS interface showed no device activity after this time. The initial performances of the tested devices were also comparable whether or not SAMs were applied.
Lau et al. 30 used silane-based SAMs between PEDOT:PSS and ITO which significantly reduced the atomic concentration of indium measured in the PEDOT:PSS layer from 1.7% to 0.04 -0.06% for the SAM-modified substrates. One of the SAMs used, allyl-triethoxysilane, produced the best blocking ability, despite giving the lowest thickness when deposited, which was attributed to cross-linking of the C=C bonds upon heating. Kara et al. 31   There is also an example, presented by There was similar performance in each of the devices although the devices containing PEDOT:PSS showed the highest efficiency. However, the stability of the devices was greatly enhanced when using either WOx or the WOx/PEDOT:PSS bilayer suggesting that either route can be an effective means of reducing instability whilst maintaining performance.
Using a barrier to reduce the etching of ITO and diffusion of indium ions into the PEDOT:PSS layer (and subsequently the active layer), has been shown to be an effective means of improving the stability of different organic electronic devices. However, any acid-sensitive materials in the active layer may also degrade and ultimately reduce performance and/or stability. However, there are many different solution-processable hole transport layers that are non-acidic and can also be used in place of PEDOT:PSS to improve the overall stability.

Solution-processable alternative materials to PEDOT:PSS
There are many different approaches that have been taken to replace PEDOT:PSS in organic semiconductor devices. However, PEDOT:PSS has been effectively used for over 10 years to give good device performance and it is generally regarded as the best and most convenient material available for electronic device applications. Films of PEDOT:PSS have a high transparency, low surface roughness, reasonably well-aligned energy levels and moderate-to-high conductivity depending on the formulation. Any alternative materials that are developed must match or improve on these properties, in addition to reducing the acidity, if they are to be viable candidates to replace PEDOT:PSS.

pH Neutral PEDOT:PSS solutions
The most intuitive means to mitigate against the degradation of PEDOT:PSS is to use a pH neutral version of the widely-used hole transport material. However, Okuzaki et al. showed that making PEDOT:PSS pH neutral can cause distinct changes in the properties with reduced conductivity, increased absorption in the NIR region and reduced order in the film. 35 Additionally, de Kok and coworkers analysed PEDOT:PSS when treated with sodium hydroxide. Typically PEDOT:PSS shows broad absorption in the NIR region which is attributed to bipolarons (figure 4), the presence of which increase with increased doping level, and this absorption is reduced significantly when the pH of PEDOT:PSS increases as a result of treatment with NaOH. 36 In fact it was shown that the change in absorbance is similar to treatment of PEDOT:PSS with reducing agent hydrazine, which would cause de-doping. 36 Therefore neutralisation of PEDOT:PSS destabilises bipolarons in the polymer backbone and effectively reduces the doping level, explaining the reduction in conductivity observed by Okuzaki and co-workers as a result of increasing pH. 35 In addition to the conductivity, the reduced doping level as a result of increased pH will be expected to affect the work function of the layer as Therefore, direct replacement often results in a reduction in performance due to differences in formulation. For example, Kim and co-workers 38 studied the effect of reacting PEDOT:PSS with NaOH. Increasing the amount of NaOH led to an increase in pH, as expected. However, at NaOH concentrations > 0.2 M, device performance was compromised due to a combination of increased resistance, a reduction in the work function (from 5.3 to 4.9 eV) and increased surface roughness.
However, the devices formed using the layer deposited from the 0.2 M treated solution showed improved stability compared to devices containing pristine PEDOT:PSS when exposed to light for 10 hours at 100 mW cm -2 (a reduction in PCE to 2.19%, compared to 1.65% for the acidic solution, where initially the devices showed PCEs ~ 3.5%). This illustrates that reducing the acidity of PSS is an effective means of improving the lifetime, although other challenges arise as a result of treatment with base.
An aqueous guanidine solution was used by Guo et al. 39  Similarly, Kim et al. 40 reported using imidazole to reduce the acidity of PEDOT:PSS and the resulting solution was deposited onto AgNWs. In this study, the AgNWs showed good stability at room temperature over 36 days, irrespective of whether the PEDOT:PSS used was pH neutral or acidic (PH1000 ; Table 1). However, at high temperature (85°C), the sheet resistance of AgNW/acidic PEDOT:PSS severely increased, whilst the increased temperature had little impact on the neutral film (or the AgNWs alone). The authors reasoned that the increased heat led to the solubilisation of Ag in the acidic environment as Ag is insoluble at room temperature in sulfonic acid, but can dissolve with heating. Interestingly, when subjected to hot and humid conditions (85°C and 85% humidity), degradation occurred at a similar rate in both PEDOT:PSS films, whilst there was degradation of the AgNWs alone too, albeit to a lesser extent. The authors ascribed this degradation process to the formation of hydrogen sulfide (H2S) and carbonyl sulfide (COS).
Wang et al. 28 reported that using a commercial neutral PEDOT:PSS solution, Clevios™, Jet N (Table 1) treated twice with UV ozone/O2 plasma. 28 The authors report that the conductivity and work function of a Jet N layer treated only once with O2 plasma are similar to the double-treated layer but the double treatment leads to a reduction in the roughness of the film, consequently improving the performance (6.11% vs 6.60% PCE).

PEDOT formulations (and analogues) containing different polyelectrolytes
It has been shown that using neutral PEDOT:PSS can be a viable replacement for its acidic analogue in maintaining performance in organic semiconductor devices and improving the stability through treatment of the PSS chain with base. However, it may be possible to take an alternative route and formulate PEDOT with a different polyelectrolyte. This approach has been pioneered by Hadziioannou and co-workers who first reported aqueous PEDOT dispersions with a polystyrenebased polyelectrolyte with (trifluoromethylsulfonyl)imide groups for development of transparent polymer electrodes. 45 Although the main focus of this work was on improving the conductivity and processability for application as a transparent electrode, polymer blends with varying pH were synthesised. It was noted that, in a basic solution, the polymerisation rate was slowed, highlighting a challenge of creating non-acidic PEDOT:polyelectrolyte formulations. The work was extended into a comprehensive analysis of PEDOT formulations with many different polyelectrolytes such as polysaccharides and other polystyrenesulfonylimide derivatives with solutions ranging from pH 3.3 -8.6 ( figure 6). 46 Although it was concluded that strong acidic groups are required for improved doping, and therefore conductivity, the properties determined for the formulations with the higher pHs suggest potential for these to be used as hole transport layers. As far as we know, such studies PEDOT:PSTFSI -pH 5.5 X = Na, K PEDOT:PMSIX -pH 6.5 PEDOT:DS -pH 6.5

Forming PEDOT films in situ
Whilst one strategy to counteract the damaging effects of the acidity of PEDOT:PSS is to formulate PEDOT with a different polyelectrolyte, another consideration is to deposit PEDOT polymer chains without any polyelectrolyte, only corresponding counter ions. This is particularly challenging due to the planar backbone and strong aggregation of PEDOT which makes it poorly soluble in polar and non-polar solvents. However, recently methods have been developed to form a polymer solution insitu or induce solid-state polymerisation of EDOT-based films. The solid-state polymerisation of PEDOT was first reported by Wudl et al. 49 Deposition of PEDOT directly from the EDOT monomer has also been reported by Gleason and co-workers using an oxidative chemical vapour deposition. 50 This involves EDOT being introduced as a vapour and reacting with a sublimed oxidant such as Fe(III)Cl3.

View Article Online
More commonly, PEDOT can be formed in solution with small anionic counter ions and solutionprocessed without polyelectrolyte species. This method is commonly applied when using PEDOT as an electrode material. In depositing the conducting polymer without the need for the insulating PSS chain, these can achieve high conductivities. However, these solutions are often formulated using strong acids such as sulfuric acid; 52 it would be fascinating if such an approach could be adapted in the future to be non-acidic.

Solution-processed metal oxides/metal complexes
There are several non-acidic hole injection materials that are commonly deposited by vacuum deposition, including MoO3, 53, 54 WO3, 55 V2O5 56 and CuI 57 for example. However, many of these materials are not easily solubilised in common organic solvents for solution processing. Therefore, a degree of chemical functionalisation is necessary if these are to be used for solution-processing.
However, due to the attractive properties of metal oxide and metal complexes, including high work function and high transparency, there has been much work into the development of solutionprocessable metal oxides and metal complexes.
Initial methods to deposit metal oxides involved the use of nanoparticles of the materials, but when these were dispersed and deposited as films the roughness could be high due to large aggregates forming. 58 However, methods were improved to produce more uniform surfaces. An early example of this was reported in 2011 by Riedl et al., 59  In devices where MoO3 is used effectively as a vacuum-deposited hole transport material, molybdenum oxide solutions have been used successfully when deposited by solution-processing.
One of the first demonstrations of solution-processed MoO3 being effective in maintaining the performance and improving the stability of organic semiconductor devices was from Rand et al., using a sol-gel method. 60 This involved dissolving MoO3 in H2O2, refluxing for 2 hours and, once cooled, adjusting the concentration and viscosity using polyethylene glycol and 2-methoxyethanol before spin-coating of films. X-ray photoelectron spectroscopy confirmed the presence of MoO3 after thermal treatment at 350°C, but the peaks were broader when compared to the evaporated sample due to the presence of several oxidation states (Figure 8(a)). Nonetheless, when the solution-processed MoO3 was applied to OPV devices containing SubPc/PC61BM and P3HT/PC61BM active layers, the devices gave competitive performance compared to the equivalent PEDOT:PSS containing compounds. Furthermore, whilst the inclusion of PEDOT:PSS caused the device to degrade below 20% of its PCE after 200 hours, the solution-processed MoO3 layer could still achieve 70% of its initial PCE after 1000 hours, which was similar to devices containing the evaporated film ( Figure 8(b)). or better performance with respect to the PEDOT:PSS-containing analogue. 61 Additionally, stability studies showed that devices fabricated using molybdenum oxide or hydrogen molybdenum bronze retained more than 80% of the short-circuit current after 800 hours, whilst the device with PEDOT:PSS showed zero current density after only ~120 hours, illustrating the severity of PEDOT:PSS-induced degradation. 61  They showed higher transmittance than PEDOT:PSS at > 580 nm and although the performance is reduced relative to PEDOT:PSS when used in OPV devices containing P3HT:PC61BM (3.77% vs 3.37%), the layer contributes to significantly improved stability when exposed to air with light soaking. Li et al. 64 were able to show that a tungsten oxide layer deposited from tungsten(VI) isopropoxide could improve performance in OPV devices using P3HT:PC61BM and P3HT:IC60BA blends, highlighting that solution-processed WO3 can be used to improve device performance, as well as lifetime, when used in place of PEDOT:PSS It has also been shown that nickel oxide-based films can be used as hole transport layers. NiOx can impart an electron blocking ability and improve the energy level alignment between the hole injection layer and the donor polymer. 65,66 Olson et al. 67  and co-workers 74 showed, in studies using CuSCN in CdSe@ZnS core-shell-based QLED devices, that when compared to the analogous PEDOT:PSS-containing device, the maximum current efficiency could again be higher when using CuSCN as HIL but the efficiency roll-off was similar to the PEDOT:PSS device.
Copper (I) iodide can also be used as a hole transport layer. Wei et al. 75  There are many metal complexes that have been developed to be easily solution-processed for organic semiconductor devices. Often they have advantages of improved transmittance and an increased work function to improve hole injection. Therefore, such materials should be considered as viable alternatives to PEDOT:PSS with a proven ability to reduce acid-induced degradation of devices.

Solution-processed graphene oxide and transition metal dichalcogenides
The discovery of graphene in 2004 76 has prompted much interest in the use of 2D-materials for many applications, including water filtration, 77 transparent conductive electrodes, 78 field-effect transistors 79 and photocatalytic water splitting 80 for example. Whilst the high conductivity of graphene lends itself to applications such as electrodes due to its high conductivity, 81 2D semiconductors such as transition metal dichalcogenides or insulating graphene oxide can be used in organic semiconductor devices. A summary of 2D materials being used in solution-processed solar cells can be found in recent review articles by Kymakis et al. 82 and Kim et al. 83 These 2D materials have been used in place of PEDOT:PSS with beneficial properties resulting.
Solution-processed graphene oxide was used by Chhowalla and co-workers in place of PEDOT:PSS in OPV devices with P3HT/PC61BM active layer. 84 Due to the insulating nature of graphene oxide, the layer thickness was critical to performance. A 2 nm thick layer in the OPV device resulted in a PCE (3.5%) that was comparable with the PEDOT:PSS reference device (3.6%), but the 4 nm and 10 nm layers caused significant reduction in performance. Hersam and co-workers used graphene oxide  87 There is a drawback from the lithium intercalation and exfoliation method though, which requires stirring of bulk MoS2/WS2 in nbutyllithium, which is pyrophoric. an increase in fill factor and short-circuit current, emphasising the potential for using WS2 hole transporting layers in organic semiconductor devices.

Organic hole transport layers for solution processing
There are many organic HTLs that are commonly used in evaporated OLEDs or OPV devices, for example. However, these materials are often incompatible for solution-processed devices -either they are too soluble in common solvents used to deposit active layers of organic semiconductor devices, making interlayer mixing a problem, or have poor solubility in most solvents due to their molecular design. Materials such as spiro-OMeTAD are commonly used in perovskite solar cells and can be deposited on top of the perovskite layer. These materials have been discussed in recent review articles. 88,89 However, the processability of such materials is limited -depositing organic solutions on top of a layer of spiro-OMeTAD would be challenging due to dissolution and interlayer mixing when solutions of common solvents are used. Therefore, the discussion is focussed on materials with solution-processing properties typically more appropriate for the majority of organic semiconductor devices. Such organic materials should either be polar in nature to allow orthogonal processing or have the ability to be insolubilised after deposition, by cross-linking for example.
Additionally, we have focussed on examples where the materials have been compared directly to PEDOT:PSS reference devices in terms of performance and lifetime. Conventional perovskite solar cells often contain organic materials such as spiro-MeOTAD as hole transport materials deposited on top of the perovskite layer. The perovskite structure is more compatible for orthogonal processing and therefore there are many organic materials soluble in common organic solvents that can be used effectively. There is also great interest in inverted perovskite solar cells due to their ability to suppress the hysteresis effect known to trouble perovskite solar cells. 92 However, inverted PSCs have challenges, of which one is the poor surface coverage of the perovskite film on metal oxide hole transport layers, therefore alternatives to metal oxides and PEDOT:PSS must be sought. Heeger et al. 93 showed the use of CPE-K ( Figure 11 Figure 11. Structure of polyelectrolyte CPE-K Depositing organic materials to be used as hole transport/injection materials is inherently more difficult than processing metal oxides, for example, due to the typical non-polar nature of these materials. However, insolubilising layers via cross-linking can be an effective means of solutionprocessing a layer with post-processing treatment making it compatible for device fabrication.

Conclusions
In this review article we have highlighted many of the challenges and complications that can arise when using acidic solutions of PEDOT:PSS, particularly for device stability. However, many of these problems can be overcome by using several different strategies, including the use of solutionprocessed metal oxides, barrier layers and pH neutral PEDOT:PSS for example. Therefore, it is possible to improve device lifetime and performance by using one of these approaches. The choice of replacement for PEDOT:PSS will depend on properties such as energy level alignment, hole mobility or even stability to flexibility. Often, one of the main advantages quoted for organic semiconductors when used as the active material in organic electronics devices is that they are tuneable and easily modified. The wide range of potential replacements for PEDOT:PSS highlighted in this article demonstrates that there is huge scope to improve the stability of a wide range of different organic electronic devices containing different organic active layers.