Interface Engineering for High Performance Bulk-Heterojunction Polymeric Solar Cells

Abstract

The nature of electrical contact between the bulk-heterojunction (BHJ) layer and the electrodes is critical to the overall performance and stability of polymer solar cells (PSCs). In recent years, considerable efforts have been devoted to the interface engineering of BHJ–PSCs. In this chapter, we reviewed the development of interface modification materials and their applications in BHJ–PSCs. The discussions were divided into several sections according to the device configuration of BHJ–PSCs, i.e., conventional device, inverted device, and tandem device. For single junction devices (including conventional and inverted devices), we highlighted the interface modification materials which were applied at both electron-collecting and hole-collecting electrodes, including inorganic materials (such as metals, salts, semiconducting oxides, etc.), self-assembled monolayers (SAMs) or SAMs-modified functional layers, and organic/polymeric materials. For tandem devices, the discussions were focused on the functional materials utilized as intermediate connectors. As discussed in this chapter, the development of interface modification materials has resulted in great progress in BHJ–PSCs, which can potentially pave the way for the commercialization of BHJ–PSCs.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

3.1 Introduction

Bulk-heterojunction (BHJ) polymer solar cell (PSC) is one of the most promising photovoltaic techniques to directly convert the terrestrial solar radiation into electricity because of their compatibility with low-cost, large-scale fabrication by solution processing [1–3]. In the last two decades, great progress in the field of BHJ–PSCs has been achieved with the power conversion efficiency (PCE) increasing from around 1 % to more than 9 % [4]. Such great advancement was driven by the efforts in various aspects including the development of new low-band-gap conjugated polymers and new fullerene-based acceptor materials [5–7], the usage of new device processing methods [8–10], the invention of new device structures [11, 12], the application of innovative interface modification materials [13–15], and so forth. The simplest structure of BHJ–PSC is the so-called sandwich device structure, where the blend of conjugated polymer donor and fullerene derivative acceptor was sandwiched between a transparent metal oxide, such as indium tin oxide (ITO), and a metal electrode [16]. For conventional devices, ITO and low work-function metal are used as anode and cathode, respectively. While in an inverted device, the modified ITO and high work-function metal are cathode and anode, respectively. The configurations of conventional and inverted device for BHJ–PSCs are demonstrated in Fig. 3.1. It is well-known that the nature of electrical contact between the organic active layer and the electrodes is critical to the overall device performance, for both the conventional and inverted devices [13–15]. The ideal Ohmic contact in the organic blend layer/electrode interface with an energy barrier height as low as possible is thus desired for efficient charge carrier extraction and transport, and thereby reduce the recombination loss caused by the accumulation of charge in the interface. Indeed, a barrier height of several tens of mV may lead to severe charge accumulation. Therefore, the modification of organic active layer/electrode interface is critical to achieve high efficiency and long-term stability for BHJ–PSCs. As a result, in recent years, a lot of efforts from research groups all around the world are devoted to the development of new interfacial materials with different functions and their applications. The aim is to achieve high PCE and long-term stability for BHJ–PSC devices [13–15].

Fig. 3.1

Comparison of conventional device structure (left) and inverted device structure (right) for BHJ–PSCs

In principle, ideal interfacial materials should first function as charge extraction or transporting layer to minimize the barrier height in organic active layer/electrode interfaces, which will help to form Ohmic contact in interface to reduce recombination loss and consequently obtain enhanced short circuit current density (J _sc). Additionally, this layer should also function as exciton blocking layer to selectively transport only one type of carrier and block the other, i.e., an electron extraction/transporting layer with hole blocking property, and a hole extraction/transporting layer with electron-blocking ability. Moreover, the interfacial layer usually functions as buffer layer to prevent the chemical reaction between organic active layer and metal electrode, and the diffusion of metal ions into organic layer. Furthermore, it was found that some interfacial materials can also function as optical spacers to modulate the incident light distribution in BHJ–PSCs, so that the maximum incident light intensity located around the center of the active layer to ensure as much as incident light harvested by the device and consequently to obtain maximum J _sc [13, 15].

In this chapter, we will review interfacial materials applied at the interfaces between organic active layer and electrodes, including both anode and cathode, and their influence on the photovoltaic performance of BHJ–PSC devices. The design of interfacial materials and their applications will be divided into several sections by the device structure, including conventional device, inverted device, and tandem device. For single junction devices (including conventional and inverted devices), the interfacial materials applied at both cathode and anode will be separately reviewed in the categories of inorganic materials (including metals, salts, semiconducting oxides, etc.), self-assembled monolayers (SAMs) or SAMs modified functional layers, as well as organic/polymeric materials. For tandem devices, we will focus on the functional materials used as intermediate connectors.

3.2 Interface Engineering for Conventional Solar Cell Devices

3.2.1 Cathode Interlayer

For a conventional BHJ–PSC, the efficient electron extraction at cathode is of critical importance for the high J _sc and the overall photovoltaic performance. Aluminum (Al) is a most widely used electrode materials for BHJ–PSCs because of its abundance and proper work function of 4.3 eV, which is very close to the lowest unoccupied molecular orbital (LUMO) energy levels of acceptor materials such as (6,6)-phenyl-C₆₁-butylric acid methyl ester (PC₆₁BM) and (6,6)-phenyl-C₇₁-butylric acid methyl ester (PC₇₁BM) (see Scheme 3.1). However, the organic optoelectronic devices with Al cathode usually suffer from poor stability, because the reactive hot Al atoms generated at the thermal evaporation process can diffuse into the organic layer and result in chemical reaction at the metal/organic interface [17, 18]. Moreover, it was reasoned that Al–C bond formation at the Al/organic interface can inevitably break the π-conjugated system of organic layer, and thereby lead to poor stability of Al cathode [19, 20]. Hence, various interfacial materials were inserted into the organic active layer/Al junction to improve the interface contact properties and thereby photovoltaic performance of BHJ–PSCs. The summary of performance of conventional PSCs using different cathode interlayer designs are listed in Table 3.1 and the chemical structures of organic materials used as cathode interlayer in conventional PSCs are shown in Scheme 3.2.

Scheme 3.1

Chemical structures of the most widely used (a) acceptors (PC₆₁BM and PC₇₁BM) and (b) donors (P3HT, PCDTBT, MEH-PPV, MDMO-PPV, PCPDTBT, and PSBTBT)

Table 3.1 Summary of device characteristics of representative conventional PSCs employing different cathode interlayers

Scheme 3.2

Chemical structures of organic materials for cathode interlayer

3.2.1.1 Metal and Inorganic Compounds as Cathode Interlayer

Low work-function metals, such as calcium (Ca) or barium (Ba), which were two widely used interfacial materials for polymeric light-emitting diodes (PLEDs) [21, 22], were also usually inserted into the interface between Al and organic active layer to further improve the device performance via the formation of ideal Ohmic contact at interface. The BHJ–PSCs with a thin layer of Ca or Ba between Al and organic active layer usually exhibited superior device performance, especially open circuit voltage (V _oc) with respect to PSCs with bare Al as cathode. However, both Ca and Ba are active metals, and thus are sensitive to the moisture in the ambient atmosphere. Although the stability of the solar cell devices could be improved in some degree by depositing a thick layer of Al or Ag onto Ca or Ba to form bilayer cathodes, such as Ca/Al and Ba/Al, the development of new interfacial materials to use as cathode interlayer is still required.

Inorganic fluorides, such as lithium fluoride (LiF), are promising electron extraction materials for BHJ–PSCs. A bilayer electrode of LiF/Al is usually chosen to replace a pristine Al cathode, because a thin layer of LiF can guarantee a good Ohmic contact between the metal and the organic active layer [23]. The formation of a favorable dipole moment or a LiF buffer layer at the organic active layer/LiF/Al interface is usually considered to be the reason why LiF could improve the performance of BHJ–PSCs. It was considered that a monolayer covered LiF can be decomposed upon Al deposition and resulting in Li-doping of the organic layer to deliver a low work-function contact; while a thick LiF usually generates a dipole layer to lower the work function of electrode [24–26].

Caesium carbonate (Cs₂CO₃) is another efficient electron extraction material used to fabricate bilayer cathode for BHJ–PSCs [27, 28]. Interestingly, the Cs₂CO₃ layer can be deposited by either thermal evaporation or solution spin coating. The implementation of Cs₂CO₃ is beneficial for reducing the electron extraction barrier and series resistance (R _s). Compared to bare Al device, the device with inserted Cs₂CO₃ exhibited enhanced V _oc and fill factor (FF) [29, 30]. However, whether the actual product of thermally evaporated Cs₂CO₃ is Cs₂O or Cs₂CO₃ is still uncertain[27, 31–33]. Anyway, it is widely accepted that the formation of Al–O–Cs complex yields the low work-function contact and thus facilitates the electron extraction [27].

n-Type semiconducting metal oxides such as titanium suboxide (TiO_x), which is a well-known photocatalyst and is widely used in dye-sensitized solar cells [34, 35], had attracted considerable attention recently due to its solution processability from sol to gel process [36], nontoxicity, and transparency. TiO_x possesses a conduction band edge and valence band edge of −4.4 and −8.1 eV, respectively, which would endow TiO_x good electron extraction ability from active layer and outstanding hole-blocking ability [36]. Hence, TiO_x is an ideal cathode modification material for BHJ–PSCs application [36]. The TiO_x layer can also resist the permeation of oxygen and moisture into the organic active layer, because the formation of robust titanium oxide film via annealing the titanium oxide film deposited by sol–gel process at 150 °C. Moreover, the TiO_x layer can also function as an optical spacer via spatially redistributing the incident light intensity to further improve the light harvesting property of BHJ–PSCs [36]. Recently, Park et al. reported a promising PCE of 6.1 % for BHJ–PSCs with poly[N-9′-heptadecanyl-2,7-carbazole-alt-5,5-(4′,7′-di-2-thienyl-2′,1′,3′-benzothiadiazole)] (PCDTBT, see Scheme 3.1):PC₇₁BM blend as active layer by inserting TiO_x as an electron extraction layer and optical spacer. Notably, the device exhibited internal quantum efficiency (IQE) closing to 100 % around 450 nm, which indicates the high efficient conversion of incident photon into charge carrier [37]. However, the intrinsic electrical properties of the TiO_x film are dominated by the processing conditions. Additionally, the electron mobility of TiO_x is 1.7 × 10⁻⁴ cm² V⁻¹ s⁻¹, which is almost two orders of magnitude lower than that of PC₆₁BM, and this may potentially limit the device performance. To further improve the property of a TiO_x interlayer, Park et al. fabricated Cs-doped film TiO₂ by mixing Cs₂CO₃ solution with a nanocrystalline TiO₂ solution prepared from a sol to gel process. The work function of TiO₂ was successfully downshifted by Cs doping to yield a better Ohmic contact between organic active layer and metal electrode. The improved electron extraction combining the preserved hole-blocking ability of the TiO_x interlayer afforded a much better photovoltaic performance for the poly(3-hexylthiophene) (P3HT, see Scheme 3.1)/PC₆₁BM solar cell than that of cell using sole TiO_x interlayer [38].

ZnO is another n-type metal oxide, which possesses a high electron mobility of 0.066 cm² V⁻¹ s⁻¹ [39] and a Fermi level of 4.4 eV [40]. The high electron mobility, ideal Fermi level, and solution processability of ZnO nanoparticles (NPs) indicate that ZnO would be a good interfacial material to facilitate not only the efficient charge transfer from PC₆₁BM (or PC₇₁BM) to ZnO, but also the effective electron collection at cathode. In addition, the wide band gap of ZnO (3.2 eV) would endow it good hole blocking ability and good transparency. Moreover, Gilot et al. had proved that ZnO can also serve as an effective optical spacer for thin (<60 nm) active layers [41]. These advantages of ZnO render its widely application in inverted BHJ–PSCs. However, the direct use of ZnO as cathode buffer layer for conventional PSCs is scarce.

3.2.1.2 Self-Assembled Monolayer on Metal Oxide Surface

The contact property of ZnO/metal bilayer cathode and resulted device performance could be further improved by spin coating the functional SAM on the ZnO surface [42, 43]. Modifying the ZnO NPs surface with a benzoic acid group constituted with electron donating end groups (such as –OCH₃, –CH₃, –H), which form a negative dipole (toward metal and away from ZnO), the band offset between the conduction band of ZnO and the work function of metal can be decreased, and thereby the Ohmic contact was formed. Compared to the unmodified devices, the devices with SAMs-modified ZnO NPs interlayers showed significant improvement in all the J–V characteristics including J _sc, V _oc, FF, and PCE. On the other hand, opposite effect was observed when the ZnO NPs layer was modified with electron-withdrawing end groups substituted SAM molecules [40]. PCE of 4.2 % was achieved for device with SAM modification with respect to 3.2 % for the control device with unmodified cathode [40]. A further enhanced PCE of 4.6 % was obtained when the ZnO NPs layer was modified with captoundecanoic acid [44]. More importantly, the stable high work-function metals such as Ag and Au can also be used as cathodes, and the PCE of 3.65 % and 3.22 % were achieved for P3HT:PC₆₁BM devices with ZnO/SAM/Ag and ZnO/SAM/Au cathodes, respectively [40].

3.2.1.3 Water/Alcohol Soluble Polymers as Cathode Interlayer

Although the above-mentioned inorganic materials exhibited outstanding cathode modification functions, the thermal evaporation of most of them requires high vacuum, which leads to increases in fabrication costs and thereby limits their applications. Organic interfacial materials were thus widely developed to improve the interface property of metal electrode, because of their solution processability, facile modification of chemical structures, and in turn electrical and electronic properties. The commonly used organic interfacial materials for cathode modification are summarized in Scheme 3.2. Zhang et al. had reported the use of environment friendly aqueous solution processable poly(ethylene oxide) (PEO) as interlayer to modify the cathode in BHJ–PSCs, which showed similar function as LiF. The PSCs with PEO modified cathode showed not only enhanced V _oc, but also J _sc and FF [45].

Water/alcohol soluble conjugated polyelectrolytes (CPEs) and related neutral polymers, which bear pendant hydrophilic polar groups, were widely used as promising interfacial materials in organic optoelectronic devices such as polymer light-emitting diodes, field effect transistors (FETs), and solar cells. Their special solubility in environment friendly solvents such as alcohol and water provide the convenience of fabricating multilayer organic devices by avoiding the problem of corrupting the low-lying neutral organic semiconducting layers which are usually dissolved in nonpolar solvents. It was generally considered that the interfacial dipoles were formed by inserting a CPE layer, and thereby the reduction of electron extraction barrier to electron collection and transporting at cathode. In addition, UV photoelectron spectroscopy has shown that this kind of polymers can effectively influence the work function of adjacent electrodes. Moreover, a large amount of water/alcohol soluble fluorene-based CPEs and related neutral polymers were successfully used as electron injection/transporting materials in PLEDs [46, 47]. Luo et al. reported the use of a thin layer of alcohol-soluble fluorene-based CPE PFNBr-DBT15 between the active layer and metal electrode, and the V _oc of the resulted PSCs can be enhanced by 0.3 V [48]. Based on Luo’s work, He et al. has insightfully investigated five different alcohol-soluble fluorene-based conjugated polymers or CPEs as cathode interlayers for PSCs based different donor materials system. He’s results revealed that the V _oc enhancement is only observed in devices that contain a polyfluorene copolymer as the donor material, while no significant enhancement in V _oc was found for P3HT and poly(2-methoxyl-5-((2′-ethylhexyl)oxyl)-1,4-phenylenevinylene) (MEH-PPV, see Scheme 3.1) devices [49]. Similar works of using other fluorene-based alcohol polymers as interfacial materials for cathode modifications were also demonstrated by Na et al. and Zhao et al. almost at the same time [50, 51]. Notably, BHJ–PSCs with high work-function metals such as Ag, Au, and Cu also showed encouraging device performances with PCE above 3 % with the implementation of fluorene-based CPE WPF-6-oxy-F [52]. After that, the use of alcohol-soluble conjugated polymer as cathode interlayer gradually become a commonly applied approach to further improve the performance of PSCs [53, 54], and the encouraging PCE of 6.07 % was achieved by inserting a thin layer of PFN (see Scheme 3.2) for the BHJ–PSC based on a newly developed low-band-gap donor material [54]. Recently, new water-/alcohol-soluble conjugated polymers and ionic conjugated di-block copolymer based on carbazole or thiophene were also developed to use as cathode interfacial materials [55, 56]. Compared to the device with sole Al as cathode, the PCE of PSCs can be enhanced from 5 to 6.5 % by inserting an ultra-thin, ionic conjugated, di-block copolymer poly[9,9-bis(2-ethylhexyl)-fluorene]-b-poly[3-(6-trimethylammoniumhexyl)thiophene] (PF2/6-b-P3TMAHT) [56]. These encouraging results indicate that water-/alcohol-soluble conjugated polymers would be good candidates of cathode interfacial materials for the future development of printable large-scale solar cells.

3.2.1.4 n-Type Organic Semiconductors as Cathode Interlayer

n-Type organic semiconductors having low-lying LUMO levels, which are close to the LUMO of PC₆₁BM and the work function of Al cathode, are potentially electron extraction and transporting materials for BHJ–PSCs. Recently, Zuo et al. had reported the use of n-type phthalocyanine tin (IV) dichloride (SnCl₂Pc) as cathode interlayer to form SnCl₂Pc/LiF/Al trilayer cathode for PSCs. The integration of SnCl₂Pc into the solar cell not only enhances the electron transporting and collection efficiencies due to the step-like electron injection barrier to cathode after the implementation of SnCl₂Pc interlayer, but also improves the exciton dissociation efficiency because of the formation of additional MEH-PPV/SnCl₂Pc exciton dissociation junction. Consequently, the 15.2 % enhancement of PCE up to 2.49 % was achieved for PSCs with the thermally evaporated SnCl₂Pc as cathode interlayer [57].

3.2.1.5 Organic Cathode Interlayer Formed by Surface Self-Segregation

Compared to the thermal evaporation, the formation of a interfacial layer of n-type organic semiconductor by controlling the surface segregation of the materials during spin coating may be a much milder approach for the low-lying organic active layer. Wei et al. had reported that when a small amount of a fullerene derivative with a fluorinated alkyl chain (F-PCBM) is mixed with the blend of P3HT and PC₆₁BM, the F-PCBM spontaneously migrates to the surface of the organic layer during spin coating due to the low surface energy of the fluorinated alkyl chain. The interfacial dipole moment toward the Al cathode caused by the perfluoroalkyl chains is favorable for the reduction of the work function of Al for better alignment with PC₆₁BM energy levels. The energy barrier for electron collection and the R _s of the modified device were hence reduced, and consequently the enhanced PCE of the device was achieved [58]. The surface self-segregation approach was also applied to incorporate a thin layer of poly(dimethylsiloxane)-b-poly(methylmethacrylate) (PDMS-b-PMMA) di-block copolymer at the interface between organic active layer and Al electrode to function as a buffer layer for improving the performance of PSC. An enhancement of the PCE of the PSCs from 3.05 % to 3.56 % on average, and the highest PCE of 3.86 % were achieved by inserting a PDMS-b-PMMA into the P3HT:PC₆₁BM solar cells due to the reduction of charge carrier recombination at the organic/metal interface [59].

3.2.1.6 Ferroelectric Polymer as Cathode Interlayer

It is well-known that a sufficient intrinsic internal electric field is required to efficiently dissociate the charge transfer excitons generated in active layer [60]. However, most widely used electrode materials deliver a work function offset of <2 eV, affording a much smaller internal electric field for efficient charge transfer excitons dissociation in BHJ–PSCs [36]. Therefore, an external bias voltage is generally needed to efficiently dissociate the electrons and holes. Most recently, Yuan et al. demonstrated another innovative method to provide sufficient intrinsic internal electric field in BHJ–PSCs by inserting a ultra-thin ferroelectric polymer layer of vinylidene fluoride-trifluoroethylene copolymer (P(VDF-TrFE)) between organic active layer and Al electrode [61]. After poling, an induced net polarization electric field is generated, which is ten times-higher than that achieved by the use of electrodes with different work functions. As a consequence, an enhanced PCE from 1–2 % without the ferroelectric film to 4–5 % was demonstrated for BHJ–PSCs based on several different active layers. Also note that these improved PCEs are higher than those achieved by other methods, including active layer morphology and electrode work-function optimization [61].

3.2.2 Anode Interlayer

3.2.2.1 PEDOT: PSS as Anode Interlayer

Interface engineering at the hole-collecting electrode is also of paramount importance for the improvement of the performance of BHJ–PSCs. The summary of performance of conventional PSCs using different anode interlayer designs is listed in Table 3.2 and the chemical structures of organic materials used as anode interlayer in conventional PSCs are shown in Scheme 3.3. In conventional device configuration using ITO as the anode, the polymer complex of poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) has been commonly used as the anode interlayer to improve the contact property between ITO and organic active layer, and to increase the work function of ITO for effective hole collection and transporting [62, 63]. However, it was revealed by Rutherford backscattering (RBS) studies and XPS that the acidic nature of PEDOT:PSS etches the ITO and results in the poor chemical stability at the ITO/PEDOT:PSS interface [64–66]. Moreover, the electrical inhomogeneities of PEDOT:PSS limit its electron-blocking ability and thus usually lead to electron leakage at the anode [65, 67]. Therefore, many efforts were devoted to modify PEDOT:PSS in order to achieve the improvement of PSCs performance. Recently, Xiao et al. reported the enhanced PCE of 4.7 % of P3HT:PC₆₁BM solar cell when a layer of ethylene glycol was spin coated over PEDOT:PSS. The enhancement was achieved by the increase of PEDOT:PSS conductivity that improves charge extraction, and enhanced PEDOT:PSS transparency that contributes to enhanced P3HT:PC₆₁BM absorption [68]. Soon after, similar work was also reported by Peng et al., who treated PEDOT:PSS layer by ethanol and 2-propanol. Similar enhancement of PCE was also observed when using this treated PEDOT:PSS as anode buffer layer, and this enhancement was attributed to higher conductivity and optimized surface morphology of the PEDOT:PSS [69].

Table 3.2 Summary of device characteristics of representative conventional PSCs employing different anode interlayers

Scheme 3.3

Chemical structures of organic materials for anode interlayer

3.2.2.2 Semiconducting Oxides as Anode Interlayer

Aside from the modification of PEDOT:PSS, the development of new materials for hole-collecting/transporting is also vigorously conducted. In principle, an efficient anodic interlayer to substitute PEDOT:PSS should first possess the capability to withstand the organic solvent erosion. Second, a good transparency is desired for the anodic interlayer to guarantee the efficient incident light absorption in organic active layer. Moreover, it should be taken into account as well that the surface properties of the anodic interlayer can significantly influence the phase separation process and morphology of the BHJ active layer.

p-Type semiconducting transition metal oxides such as molybdenum oxide (MoO₃), vanadium oxide (V₂O₅), nickel oxide (NiO), and tungsten oxide (WO₃) were hence widely employed as anodic buffer layers to modify the interface between ITO and organic active layer in the conventional BHJ–PSCs. These oxides have relatively large band gap, which guarantee the good optical transparency in visible and near infrared light region of the anode, and consequently allow incident solar photons to reach the organic active layer. More importantly, the Fermi level of these oxides usually positions in the range between 5.0 and 5.4 eV, which align well with the HOMO energy level of most donor conjugated polymers to form ideal Ohmic contact for efficient hole extraction and transporting. Furthermore, the lowest energy level of the conduction band of these p-type oxides usually located above 2.5 eV, which is much higher than the LUMO energy level of most organic photovoltaic (OPV) materials (including both acceptors and donors), indicating the good electron-blocking ability of these oxides.

Shrotriya et al. first reported the use of thermally-evaporated MoO₃ as the anodic interlayer for BHJ–PSCs to replace PEDOT:PSS [70]. The BHJ–PSCs with 5 nm of MoO₃ as anodic interlayer exhibited slightly better performance (3.36 % PCE) than that of PEDOT:PSS device (3.10 % PCE). They revealed as well that an optimized thickness is critical to obtain ideal device performance: a thinner layer of MoO₃ results in a smaller V _oc and leakage current because of the incomplete coverage, whereas a thicker layer of MoO₃ increases the R _s and in turn leads to a smaller J _sc and FF [70]. Similar solar cell performance enhancement was also observed by Kim et al. when MoO₃ was inserted as anodic buffer layer between ITO and organic active layer [71]. Most recently, Sun et al. reported encouraging results about the use of thermally-evaporated MoO₃ as the anodic buffer layer of BHJ–PSCs with the configuration of ITO/MoO₃/PCDTBT:PC₇₁BM/TiO_x/Al. The implementation of MoO₃ improved the light absorption with the organic active layer, and thereby resulted in a PCE over 6 % at BHJ layer thickness up to 200 nm. A further enhancement in PCE up to 7.2 % was achieved by using an antireflection coating. In addition, BHJ–PSC with MoO₃ as anodic buffer layer demonstrated much better long-term air stability than that of solar cell fabricated with PEDOT:PSS. The PCE remains at approximately 50 % of the original value after the storage in air for 720 h, while the PCE of control PEDOT:PSS device fell to <10 % of the original value after storage in air for 480 h [72].

Deposition of a thin layer of p-type NiO by pulsed laser onto ITO to replace PEDOT:PSS as anodic interlayer led to significant performance enhancement of BHJ–PSC based on P3HT:PCBM blend. A 5–10 nm NiO layer gives rise to the PCE as high as 5.2 %, while the PCE of control PEDOT:PSS device is only 2.4 %. The enhancement was initially attributed to the ideal work function of NiO (5.0–5.4 eV) to match well with the HOMO level of P3HT (5.0), and large band gap (ca. 3.6 eV) of NiO to deliver high transparency and sufficient barrier for electron blocking [73]. After that, Irwin et al. via a multifaceted analysis further revealed that NiO grows as smooth, crystalline, and oriented thin films on ITO substrates to form an optically transparent, electrically uniform, and passivated semiconducting anode coating, which prevents anode electron injection and facilitates anode hole injection [74]. However, the pulsed laser deposition of NiO layer is neither scalable nor a cost effective method. Steirer et al. deposited a thin layer of NiO onto ITO by spin coating a diluted nickel metal organic ink followed by thermal annealing at 250 °C. The BHJ–PSCs from this solution-processed NiO exhibited comparable performances with that of PSCs from pulsed laser deposited NiO and PEDOT:PSS [75].

Thermally evaporated thin layers of V₂O₅ and WO₃ were employed as effective buffer layers on ITO to improve the performance of BHJ–PSCs based on P3HT:PC₆₁BM blend. The devices based on both oxides exhibited comparable performances with those of PEDOT:PSS control device. It was suggested that the ideal work function (4.7 eV) and the relatively high-positioning of the lowest energy level of the conduction band (2.4 eV) of V₂O₅ are beneficial for forming efficient hole-collection injection contact with the organic active layer and to provide sufficient barrier for electron leakage at anode [70]. Han et al. revealed that the uniform amorphous film of WO₃ can effectively planarize an originally rough ITO. P3HT films, grown on WO₃ film, have a higher degree of ordering and larger hole mobility than those grown on PEDOT:PSS [76]. An ultra-thin layer of AgO_x generated through plasma oxidized Ag (1 nm) deposited on ITO was found to be able to improve the contact property of ITO/PEDOT:PSS interface. The enhancement of device performance is suggested to the formation of an interface energy step between ITO and PEDOT:PSS that could improve the charge collection efficiency and the overall efficiency of solar cell devices [77].

3.2.2.3 Organic/Polymer Materials as Anode Interlayer

Despite that p-type transition metal oxides exhibited outstanding anode interface modification functions, the fabrication of thin films based on these oxides usually involves high-cost thermal evaporation, which limits their application in large-area devices. Organic hole-collecting/transporting materials were thus greatly desired due to their solution processabilities and their facile tunable properties. A cross-linkable blend of poly[9,9-dioctylfluorene-co-N-[4-(3-methylpropyl)]-diphenylamine] (TFB) and 4,4′-bis[(p-trichlorosiylpropylphenyl)phenylamino]biphenyl (TPDSi₂) was spin coated onto a ITO substrate to form a robust, optical transparent, homogeneous film after thermal annealing, which can be used as an effective PEDOT:PSS alternative. TFB:TPDSi₂ possesses HOMO and LUMO energy levels of −5.3 and −2.3 eV, respectively, which endow the buffer layer good hole-collecting and electron-blocking ability. As a result, BHJ–PSCs of poly(2-methoxyl-5-((3′,7′-dimethyloctyl)oxyl)-1,4-phenylenevinylene) (MDMO-PPV):PCBM with TFB:TPDSi₂ anodic buffer layer delivered much better photovoltaic performance and thermal stability than those of PEDOT:PSS control devices [78]. Interestingly, the use of a double interfacial layer of PEDOT:PSS + TFB:TPDSi₂ could decrease the FF of the solar cells, but increase electron blocking to suppress charge leakage and thereby enhance V _oc more than using either PEDOT:PPS or TFB:TPDSi₂ independently [79]. After that, Subbiah et al. demonstrated that a double interlayer of MoO₃/TFB could also improve the photovoltaic performance of the BHJ–PSC with respect to solar cells with a sole PEDOT:PSS or MoO₃ anodic buffer layer, due to the enhanced electron blocking and hole collecting from organic active layer to the anode [80].

Li et al. demonstrated the use of a novel self-doped polymer of sulfonated poly(diphenylamine) (SPDPA) as the anodic buffer layer to replace PEDOT:PSS in P3HT:PC₆₁BM solar cell. The polar surface of SPDPA film induces the oriented arrangement of P3HT in the active layer during the spin-coating and film-growing processes, which is beneficial for enhancing the hole mobility, producing a better Ohmic contact at the anode junction and thereby gives rise to an enhanced PCE [81]. The implementation of a thin layer of polytetrafluoroethylene (PTFE) between ITO and P3HT:PC₆₁BM was found to form an dipole layer at the anode junction, and thereby facilitate the hole extraction. Compared to the PEDOT:PSS control device, solar cells based on PTFE showed obviously enhanced photovoltaic performance. Note that such an insulated PTFE thin layer was deposited by thermally evaporation, and it thereby can be prepared at low substrate temperatures, which is compatible to the flexible polymer substrates for BHJ–PSCs applications [82].

In addition to the modification of PEDOT:PSS or development of PEDOT:PSS alternatives, the implementation of hole selective transporting materials below or above PEDOT:PSS layers is another effective approach to facilitate hole-collecting and electron-blocking at anode junction, and thereby to improve the performance of BHJ–PSCs. It is well-known that there is a vertical composition gradient in organic blend layer with a profile of PC₆₁BM- or PC₇₁BM-rich blend adjacent to PEDOT:PSS and conjugated polymer-rich blend near the cathode side, which is possibly detrimental to charge extraction efficiency [83, 84]. Liang et al. hence simply inserted a thin layer of high molecular weight P3HT between PEDOT:PSS and P3HT:PC₆₁BM blend layer to increase the electron-blocking ability of PEDOT:PSS. The extra donor/acceptor interfaces adjacent to the bottom of the composite were created as well, leading to an enhanced photoinduced electron transfer efficiency and photocurrent density. Relative to the 3.98 % PCE of the reference device, the extra P3HT inserted PSC delivered PCE of 5.05 % [85]. The insertion of a discotic liquid crystal of hexabutoxytriphenylene (HAT4) at the interface between anodic buffer layers (PEDOT:PSS, MoO₃ or NiO) and organic active layer was also found to be an effective method to improve the photovoltaic performance of P3HT:PC₆₁BM solar cell. Atomic force microscopy (AFM) measurement indicate that the ordered hexagonal columnar phase formed in HAT4 layer provides a more efficient pass way for hole transporting, and thereby leads to an enhanced J _sc and FF [86].

The insertion of an in situ polymerized triphenylamine-containing polyperfluorocyclobutane (TPA-PFCB) thin layer between ITO and PEDOT:PSS was revealed to be an effective approach to block the electron leakage at anode. After the coverage of TPA-PFCB, the surface roughness of ITO was substantially reduced. Compared to the control PEDOT:PSS solar cell, the implementation of the TPA-PFCB layer increases both V _oc and J _sc, and thereby gives rise to enhanced PCE. The HOMO and LUMO of TPA-PFCB was measured to be −5.2 and −1.7 eV, respectively, which verified its good charge selective hole-transporting and electron-blocking properties. In addition, FET measurement confirmed that electron transporting is completely blocked by implementation of TPA-PFCB between the organic active layer and the Al source and drain electrodes [87].

3.2.2.4 Carbon Nanotubes as Anode Interlayer

Carbon nanotubes (CNTs) possess high electrical conductivity, approximate work function of ~5.0 eV which match well the work function of ITO and HOMO of most donor polymers, and outstanding optical transparency in a broad spectral range from UV to deep infrared region, making CNTs potential hole-collecting materials for BHJ–PSCs [88]. Chaudhary et al. had inserted CNTs at different interfaces (ITO/PEDOT:PSS, PEDOT:PSS/P3HT:PC₆₁BM, and P3HT:PC₆₁BM/Al) of P3HT:PC₆₁BM solar cell, and had found that only CNTs were incorporated into the interfaces of ITO/PEDOT:PSS or PEDOT:PSS/P3HT:PC₆₁BM and the solar cells exhibited obvious enhancement in PCE [89]. Most recently, Hatton et al. had demonstrated that the use of partially oxidized CNTs as anodic buffer layer to replace PEDOT:PSS can effectively facilitate the hole extraction. P3HT:PC₆₁BM solar cells with this neutral aqueous processed CNTs anodic buffer layer exhibited comparable photovoltaic performance with that of PEDOT:PSS control device [90].

3.2.2.5 Self-Assembled Monolayers as Anode Interlayer

SAMs with different end groups (–CH₃, –NH₂, –CF₃) were also employed to modify the work function of ITO, and thereby to improve the performance of P3HT:PC₆₁BM solar cell. The work function of ITO increased initially from 4.7 to 5.16 eV after the insertion of a –CF₃ substituted SAM molecules, affording a much better match with the HOMO of P3HT for efficient hole extraction at anode junction. Moreover, it was found that the surface properties of the inserted SAMs can greatly influence the morphology of the upper organic active layer. Active layer spin coated onto the hydrophobic surface of –CF₃ SAM exhibited little undesired phase separation, leading to a high PCE of 3.15 % for the corresponding solar cell [91].

3.3 Interface Engineering for Inverted Solar Cell Devices

The conventional device structure of BHJ–PSCs has some inherent device stability problems. The transparent conducting ITO used as the hole-collecting contact can be etched over time by the acidic PEDOT:PSS hole-transporting layer [64]. The thermo-deposition of low work-function cathode usually requires high vacuum, thus leading to increases in fabrication costs. Moreover, to avoid the exposure of low work-function cathode to air, encapsulation technologies have to be applied which further complicate the fabrication process. Therefore, device architectures that can remove the need of PEDOT:PSS (or at least separate it from ITO) and use nonvacuum-deposited air-stable high work-function metal electrodes at the top interface are desired. Based on these considerations, inverted solar cell device architecture was proposed. The device structure is shown in Fig. 3.1. This architecture has recently attracted considerable attention due to the device stability and processing advantages compared to the conventional architecture. In an inverted device, the polarity of charge collection is the opposite of the conventional architecture, allowing the use of higher work function and air-stable materials (Au, Ag, and Cu) as the top anode which is exposed to air. The use of higher work-function metals offer better ambient device stability and the possibility for using low-cost solution-processed techniques such as spray coating [92] or screen printing [93] to deposit the top anode. The focus of current research for the inverted device architectures is to understand how to improve the device efficiency and stability through the design and processing of the different interfacial layers in the device structure. In this section, we will review the recent progress of interface engineering for inverted solar cells in two aspects: the progress in anode interlayer design and the progress in cathode interlayer design.

3.3.1 Cathode Interlayer

In principle, ITO is capable of collecting either electrons or holes since its work function (4.5–4.7 eV) lies between the typical HOMOs and LUMOs of common OPV materials. Therefore, the polarity of ITO can be modified to efficiently collect either electrons or holes by coating functional interlayers of different work functions onto its surface [29]. For inverted solar cell application, electron transporting layers (ETL, or electron selective layer) with low work function are formed on top of ITO to modify the ITO interface for efficient electron extraction and collection from active layer. Such ETL materials typically include inorganic ones like Cs₂CO₃ [29, 33], Ca [94], Al₂O₃ [95], ZnO [96–98], and TiO_x [99, 100]. Since light needs to pass through the ETL to reach the active layer to generate photocurrent, the layers are typically very thin to maintain high optical transmittance. The electron mobility and work function of the ETL layer should also be optimized in order to realize efficient electron collection. The summary of performance of inverted PSCs using different ETL designs is listed in Table 3.3 and the chemical structures of organic materials used as ETL in inverted PSCs are shown in Scheme 3.4 (WPF-6-oxy-F is shown in Scheme 3.2).

Table 3.3 Summary of performances of inverted polymer solar cell devices using different ETL designs

Scheme 3.4

Chemical structures of organic materials used in ETL for inverted polymer solar cells

3.3.1.1 Metal Oxide Thin Film as ETL

ZnO and TiO_x are the most commonly utilized ETL materials for inverted solar cells due to the high optical transparency in the visible and near infrared region, high carrier mobility, and its solution processibility. Many demonstrations of using these n-type metal oxide layers as the electron selective layer for inverted solar cells have been reported in literatures.

An efficient P3HT:PC₆₁BM BHJ inverted solar cell from a high-temperature processed sol–gel ZnO interlayer on ITO and an Ag electrode as the top hole-collecting contact was first demonstrated by White et al. [96]. The zinc acetate (ZnAc) sol–gel precursor was directly spin cast onto ITO and then thermally annealed at 300 °C for 5 min to hydrolyze and crystallize into amorphous ZnO thin film. Improved conductivity and mobility after annealing led to conversion efficiencies of 2.97 % of the inverted device. Interestingly, it was found that when these devices were exposed to air, the performance gradually improved. They attribute the improvement to the oxidation of Ag which formed Ohmic contact with P3HT. State-of-the-art inverted solar cell using ZnO as ETL and a low-band-gap donor material PCDTBT can yield PCE as high as 6.33 % [97]. To further improve the electronic properties of ZnO, Al doping was explored and inverted devices with Al-doped ZnO (AZO) as ETL were thus fabricated [101, 102]. Although the AZO device performance is not greatly improved compared to undoped ZnO devices, AZO layer can be made much thicker (>100 nm) than ZnO layer without hampering the solar cell performance; thus, it is more robust and easier to process.

TiO_x sol–gel layer (∼10 nm) was also demonstrated as an effective ETL for inverted solar cells [99]. The devices with structure of ITO/TiO_x/P3HT:PC₆₁BM/PEDOT:PSS/Au showed a PCE of 3.1 %. o-Xylene was used as solvent for P3HT:PC₆₁BM instead of the commonly used chlorobenzene or dichlorobenzene to make the vertical phase segregation of donor and acceptor more favorable for inverted devices. Like in the case of conventional devices, controlling the bulk blend vertical phase segregation is also important in inverted devices; however, the optimal phase segregation is different from the case of conventional devices [84].

3.3.1.2 Metal Oxide Nanoparticle Layer

One of the problems of sol–gel process of ZnO and TiO_x is that they require high temperature annealing processing conditions in order to improve the crystallinity of the material to minimize resistive losses in the solar cell devices. These high temperature processing conditions can reach as high as 500 °C, which is not compatible with industry scale roll-to-roll process. To overcome this problem, ZnO NPs are introduced as ETL materials for inverted solar cells [41, 103]. The devices fabricated from the ZnO NPs on ITO-coated glass show an average PCE of ∼3.6 %. This value is very similar to that obtained from the high temperature processed ZnO sol–gel devices on glass/ITO which show an average efficiency of ∼3.5 %. This demonstrates that sol–gel ZnO and ZnO NPs thin film layers can both act as a good electron selective layer in the inverted device architecture.

3.3.1.3 Self-Assembled Monolayer on Metal Oxide Surface

Although reasonable efficiencies have been reached with n-type metal oxides as the electron selective layer in inverted solar, it still has great room for improvement, as the surface of metal oxides have hydroxyl groups that can cause charge trapping at the metal oxide/active layer interface [104]. These hydroxyl groups terminated surfaces lead to high-interface charge recombination due to poor charge transfer. One approach that can improve the electrical and morphological properties of metal oxide/active layer interface is to utilize a SAM between the inorganic and organic interface [40, 91, 105–109]. SAMs can be utilized to significantly modify the interfaces of oxide and metallic surfaces to improve adhesion, compatibility, charge-transfer properties, energy level alignment, and affect the upper layer growth of materials. It was demonstrated that modifying the metal oxide surfaces of TiO₂ and ZnO-based inverted solar cells with a fullerene-based SAM (C₆₀-SAM) can improve the device performance. The C₆₀-SAM affects the photo-induced charge transfer at the interface to reduce the recombination of charges, passivate inorganic surface trap states, improve the exciton dissociation efficiency at the polymer/metal oxide interface as well as act as a template to influence the overlayer BHJ distribution of phases and crystallinity leading to higher efficiency inverted solar cells [105, 110].

3.3.1.4 Polymer and Cross-Linked Interlayer on Metal Oxide Surface

The potential drawbacks for SAM formation on metal oxide surface are incomplete coverage at the molecular scale and probable desorption of this monolayer during wet processing, creating localized defects in this interlayer [109]. The other approach that can improve the metal oxide/active layer interface is to insert an organic ETL interlayer in between. In order to resist the solvent washing from the over-layer, this organic layer should have orthogonal solubility with active layer, or it should be cross-linkable. Choi et al. reported a remarkable improvement in inverted solar cell performance by employing a thin layer of CPE on top of TiO_x. The TiO_x/CPE composite ETL improves the electron injection and transport at the cathode and blocks the hole transport to the cathode, leading to a PCE improvement from 2.65 to 3.55 % [111]. The CPE material was alcohol soluble, and thus not affected by the solvent of the active layer. Hsieh et al. reported a PC₆₁BM-based n-type material [6,6]-phenyl-C₆₁-butyric styryl dendron ester (PCBSD) functionalized with a dendron containing two styryl groups as thermal cross-linkers [112]. By heating at 160 °C for 30 min a robust, adhesive, and solvent-resistant thin film can be generated on top of ZnO layer. An inverted solar cell device based on ITO/ZnO/cross-linked PCBSD/P3HT:PC₆₁BM/PEDOT:PSS/Ag configuration not only achieves enhanced device characteristics (PCE 4.4 %), but also exhibits an exceptional device lifetime without encapsulation; it greatly outperforms a reference device (PCE 3.5 %) based on an ITO/ZnO/P3HT:PC₆₁BM/PEDOT:PSS/Ag configuration without the interlayer. Changing the acceptor in the active layer from PC₆₁BM to a novel fullerene derivative indene-C₆₀ bis-adduct (ICBA) can further improve the PCE of inverted solar cell device with PCBSD interlayer [113], yielding a record efficiency of 6.2 % for inverted organic solar cell devices at the time.

3.3.1.5 Polymer or Cross-linkable Organic Thin Film as Single ETL

Besides acting as an interlayer between metal oxide layer and active layer, polymer or cross-linkable organic thin films can also be implemented as independent ETL for inverted solar cells to completely replace the metal oxide layer. Zhou et al. reported the use of PEO thin film as single ETL on ITO for inverted solar cells. The insertion of PEO interlayer between ITO and APFO3:PC₆₁BM active layer improved the PCE of the devices from 0.5 % to 0.7 % [114]. Na et al. later reported the use of a water-soluble polyfluorene CPE as a single ETL which can improve the PCE of P3HT:PC₆₁BM inverted devices from 1.04 % without interlayer to 3.38 % [115]. The same group found out that the CPE material can also be an effective ETL for multilayered graphene (MLG) cathode [116]. An inverted solar cell device with configuration of MLG/CPE/P3HT:PC₆₁BM/PEDOT:PSS/Al has a PCE of 1.23 %, while the reference device without CPE interlayer has virtually no photovoltaic effect at all. In these studies, the increase in PCE is explained by the enhancement of built-in voltage due to work function lowering effect of the interlayer, and this is generally supported by work function measurement of ITO substrate with or without interlayer through ultraviolet photoelectron spectroscopy (UPS) or Kelvin probe.

Besides common solution process, polymer ETL can also be fabricated by layer-by-layer (LBL) method. Rider et al. reported an ETL generated by repeated LBL deposition of cationic water-soluble polythiophene poly[3-(6-pyridiniumylhexyl) thiophene bromide] (P3PHT⁺Br⁻) with anionic (PEDOT:PSS)⁻Na⁺ on ITO [117]. The inverted solar cell device of ITO/(P3PHT⁺/(PEDOT: PSS)⁻)₅/P3HT:PC₆₁BM/V₂O₅/Al has an efficiency of ~2 % and operation lifetime >500 hours.

As the organic ETL materials generally suffer from their low electron conductivity compared to metal oxides, some researchers try to compensate this by doping the ETL layer. Cho et al. demonstrated that the n-doping of a cross-linkable PC₆₁BM derivative PCBM-S using decamethylcobaltocene (DMC) can significantly increase the electron conductivity of the cross-linked ETL layer [118]. In the configuration of ITO/PCBM-S/P3HT:PC₆₁BM/PEDOT:PSS/Ag, device with PCBM-S layer of optimum DMC doping level yield a PCE of 2.53 %, compared to 1.24 % for devices with undoped PCBM-S interlayer.

Before the end of this section it should be noted that the electronic properties of the interface of ITO and metal oxide is complicated and our understanding is far from complete. For example, Wang et al. reported an interesting case that the PCE of the inverted solar cell device with the structure of ITO/P3HT:PC₆₁BM/PEDOT:PSS/Ag changes irreversibly from 1.46 to 4.1 % when the device is exposed to continuous illumination of simulated sunlight for 2 h [119]. Though the mechanisms behind the large enhancement in PCE is not thoroughly explained, the fact that high PCE can be achieved in a device without ETL poses the question of how and to what extent an ETL can improve the electron-collecting ability of ITO in inverted devices.

3.3.2 Anode Interlayer

Direct contact of the top high work-function metal to the active layer can lead to degradation of the solar cell performance, therefore, a hole-transporting/electron-blocking layer (HTL/EBL) is generally deposited between these two layers to improve charge selectivity and collection of holes. The materials that have been utilized as HTL are various high work-function transition metal oxides such as MoO₃, or denoted as MoO_x because of oxygen vacancy formed during thermo-evaporation [97, 120, 121], WO₃ [122, 123] and V₂O₅ [29, 124], and solution-processed organic materials including PEDOT:PSS [99, 112, 113], SPDPA [125], dextran-doped poly(allylamine hydrochloride) (PAH-D) [126], and cross-linkable molecule N,N′-diphenyl-N,N′-bis-(3-methylphenyl)-(1,1′)-biphenyl -4,4′- diamine- bis (vinyl benzyl ether) (TPD-BVB) [127]. The chemical structures of the organic HTL materials are shown in Scheme 3.5 (PEDOT:PSS and SPDPA are shown in Scheme 3.3).

Scheme 3.5

Chemical structures of organic HTL materials for inverted polymer solar cells

3.3.2.1 PEDOT:PSS as HTL or Anode

The literature on HTL design of inverted solar cells is relatively scarce compared to ETL. For solution-processed HTL PEDOT: PSS is often used. The challenge of using PEDOT:PSS as HTL in an inverted device is that PEDOT:PSS is an aqueous dispersion and it is difficult to coat it onto an organic active layer due to its hydrophobic property. Cosolvent [99] or surfactants [112] are added to improve interfacial compatibility of PEDOT: PSS and active layer. However, the conformation of PEDOT:PSS using these methods was different and consequently its work function and conductivity would be changed. Another way to uniformly deposit PEDOT:PSS onto an active layer is to use spray coating technique [128]. By incorporating D-sorbitol into PEDOT:PSS, a transparent “electric glue” can be formed, which is capable of laminating films together both mechanically and electrically [129]. Implementing the unique property of the “electric glue” into the inverted device, a semitransparent polymer solar cell based on the P3HT:PC₆₁BM blend was fabricated by the lamination process with a 3 % PCE [130]. This method took advantage of the solution process, which also featured self-encapsulation and provided an alternative to the roll-to-roll production of inverted solar cells.

PEDOT:PSS in its high conductivity form can be directly used as anode for inverted solar cells. Lim et al. spray coated a layer of modified PH500 (300 S cm⁻¹) high conductivity PEDOT:PSS onto ITO/Cs₂CO₃/P3HT:PC₆₁BM to form the anode, and the PCE of the device was 2 % [63]. Though the conductivity of PEDOT:PSS layer formed by spray coating was not optimized compared to spin-coating method, the advantage of spray deposition is that very thick films can be obtained, which makes up for the low conductivity. Zhou et al. went one step further to use PH1000 (900 S cm⁻¹) high conductivity PEDOT:PSS as both anode and cathode for semitransparent inverted solar cells [131]. The device structure was PH1000/ZnO/P3HT:PC₆₁BM/CPP-PEDOT (a low conductivity PEDOT: PSS)/PH1000, and the PCE was 1.8 %.

3.3.2.2 Solution Processable Metal Oxide HTL

Though the best inverted solar cells are fabricated with thermo-evaporated transitional metal oxide HTL such as MoO₃ or V₂O₅, solution processable HTL is preferred since vacuum evaporation could detract from the advantage of the ease of polymer solar cell fabrication. Therefore, some efforts are made to design solution-processing route for transitional metal oxide materials. Huang et al. dispersed V₂O₅ powder in isopropanol through ultrasonic agitation and spin cast the dispersion onto P3HT: PC₆₁BM active layer to form HTL [124]. The device with configuration ITO/ZnO nanorod/P3HT:PC₆₁BM/V₂O₅/Ag has a PCE of 3.56 % under optimum V₂O₅ thickness, compared to 2.24 % for reference device without the V₂O₅ layer. Later, it was found that NiO NPs can also be dispersed in isopropanol and spin cast onto active layer to form HTL. Lim et al. reported solution-processed NiO as single HTL [132] and Lin et al. reported the solution process of NiO/plexcore HTL double layer HTL [133]. A sol–gel solution process of copper (I) oxide (Cu₂O) HTL similar to the sol–gel process of ZnO was recently reported [134]. Deionized water was added into the copper (II) acetate monohydrate precursor solution to trigger sol–gel reaction and the solution is then spin cast onto active layer to form Cu₂O layer in situ. PCE of the devices with the Cu₂O HTL are improved from 3.34 to 4.02 %.

3.4 Interface Engineering for Organic Tandem Solar Cell Devices

To improve the absorption of the sunlight photons by organic solar cells, active materials with a broader absorption spectrum have to be designed. This is generally achieved by reducing the band gap of the material which could lead to lower V _OC in the devices. Another strategy is to apply tandem device structure in which multiple absorbers with different band gaps are to be stacked or mixed in multiple junctions. When two or more donor materials with nonoverlapping absorption spectra are used in a tandem solar cell, broader range of the solar spectrum can be covered without sacrificing V _OC of the device, as the thermalization of hot carriers can be minimized [135]. The most commonly employed tandem cell structure is a two terminal monolithic device in which two subcells are connected in series through an interconnecting layer (ICL), as shown in Fig. 3.2. Several approaches for organic tandem cells have been reported in recent years, depending on the materials used for the active layer and the respective interconnection layers. In general, the organic tandem solar cells can be divided into three classes [136]: (1) Tandem organic solar cells in which low-molecular-weight molecules are used for both the bottom and the top cells; (2) Hybrid tandem organic solar cells in which the bottom cell is processed from polymers by solution-processing, while the top cell is made of vacuum-deposited low-molecular-weight molecules; (3) Fully solution-processed tandem organic solar cells in which both the bottom and the top cells are BHJ–PSCs. The third class of organic tandem solar cells began to receive enormous attentions since Kim et al. successfully fabricated an all solution-processed (except the cathode) tandem cell device with a record efficiency of 6.5 % in 2007 [12]. The interface problems in such tandem solar cell devices are much more complicated than single junction devices, as more layers are incorporated into a single device.

Fig. 3.2

Schematics of common double junction tandem solar cell device structure

The ICL, both physically and electrically, connects the two subcells in an organic tandem solar cell device and it is critical in the fabrication of highly efficient tandem devices. Inappropriate design of the ICL could lead to electrical or optical losses which would in turn result in the loss of performance, and most notably the loss of V _OC. The role of the ICL in tandem devices is threefold: (1) it should efficiently collect electrons from one subcell and holes from another subcell; (2) it should act as an efficient recombination zone so that charge accumulation in this layer is minimized; (3) for all solution-processed tandem devices, it should also act as effective separation layer to protect the bottom cell from dissolving during processing of the top cell.

In this section, we focus on the design and fabrication of ICLs in organic tandem solar cells since the most important interfacial engineering problems are related to this layer. The summary of organic tandem solar cell performance using different ICL designs is listed in Table 3.4.

Table 3.4 Summary of performances of organic tandem solar cell devices using different ICL designs

3.4.1 Thermally Evaporated ICL

In small molecule tandem cells, the acceptor layer of bottom cell and the donor layer of the top cell can be completely separated to prevent formation of an inverse p–n junction by the simple incorporation of an ultra-thin layer of Au [137] or Ag [138]. However, if a thin layer of metal is used as ICL in a tandem cell with at least one polymer BHJ as subcell, the interlayer contacts both the acceptor and donor domains of the BHJ, making the selective extraction of electrons or holes from the polymer BHJ impossible. To overcome this problem, diffused bilayer polymer heterojunction can be used instead of a BHJ, so that only PC₆₁BM contacts the metal layer [139]. Another approach is to have additional hole-blocking or electron-blocking layers in the ICL. Heavily doped p–n junction as ICL was proposed to ensure Ohmic contact between the subcells [140, 141]. This approach allowed the addition of the V _OC of individual subcells in tandem architectures with minimal loss due to absorption or reflection in the interlayer. More importantly, the good electrical contact of the ICL does not depend on the choice of photoactive materials.

More commonly used ICL design for tandem cells is the bilayer composite consisting of a low work-function (n-type) layer and a high work-function (p-type) layer [142]. Typical materials used for p-type layer are high work-function metal oxides such as MoO₃ or WO₃. These metal oxides can easily form Ohmic contact with organic active layers and they are also highly transparent. Zhao et al. used MoO₃ as high work-function contact for the top cell and a bilayer of LiF and Al forming a low work-function contact with the bottom cell [143]. Negligible optical loss in the ICL was observed because of almost 100 % transmittance of the interlayer and no potential losses were observed either.

A major challenge in fabricating polymer tandem cells is the processing of the top cell without damaging the underlying layers. Thus, the criterion for choosing the materials for ICL, besides Ohmic contact forming ability, is the robustness of the thin film so that it can withstand any solution process and protect the bottom cell layers. Kawano et al. demonstrated a polymer tandem cell using ITO/PEDOT:PSS as ICL [144]. The ITO layer in the ICL was sputtered onto bottom cell active layer in 1 Pa of argon gas so that the damage in active layer can be prevented during sputtering. It was found that 20 nm of ITO was sufficient to protect the underlying active layer from solution process of the top cell. Hadipour et al. used an efficient multilayer ICL consisting of n-type and p-type layers [145]. The n-type layers were LiF/Al and the p-type layers were Au/PEDOT:PSS. Au layer was deposited to protect the bottom LiF/Al contact from the solution process of the top cell. V _OC of the tandem cell was equal to the sum of V _OC of the component cells which shows the effectiveness of the ICL. However, the disadvantage of using metal layers is the loss of incident photons for the top cell.

3.4.2 Solution-Processed ICL

The real advantage of polymer solar cells is the ease of fabrication using solution process, thus the development of solution-processed ICL for tandem cells is necessary in order not to compromise this advantage. There were many reports on solution-processed electron transport layers such as ZnO and TiO₂ (or TiO_x), that can be used as n-type layers in the ICL. These electron transport layers are wide-band-gap semiconductors and are highly transparent to sunlight. PEDOT:PSS, being a high work-function conducting polymer, is predominantly used to form Ohmic contact for holes with the polymer BHJ and is an alternative to thermally evaporated high work-function metal oxides. The first demonstration of all solution-processed polymer tandem cells was using ZnO NPs dissolved in acetone to deposit the n-type layer, and aqueous-based neutral pH PEDOT:PSS as the p-type layer [146]. ZnO has been shown to form an efficient contact for electron extraction from polymer BHJ because of its matching energy level with PC₆₁BM. The alcohol dispersed ZnO NPs does not damage the bottom polymer layer and aqueous-based neutral PEDOT:PSS layer does not affect the ZnO layer. Such an ICL is robust enough to protect the bottom polymer cell from any subsequent solution process. It was thus guaranteed that during the entire fabrication process, none of the solution process steps damaged the previously formed layers. Using these solution-processed ICLs, double and triple junction tandem cells were shown with only minor losses in V _OC.

Sol–gel derived TiO₂ is another potential candidate as an n-type contact. Kim et al. and Sista et al. used sol–gel based TiO₂ layer for efficient electron extraction from the bottom cell and PEDOT:PSS as a p-type layer [12, 147]. Though the synthesis approaches of TiO₂ employed by the two groups are totally different, highly efficient tandem cells were both demonstrated without V _OC losses. Kim et al. used a TiO_x precursor solution to coat 20–30 nm thick dense films on polymer layer that was followed by baking in air for hydrolysis of precursor into solid state TiO_x. The purpose of the TiO_x layer was threefold: being an electron transport layer, a hole-blocking layer, and an optical spacer. On the other hand, Sista et al. used crystalline NPs of TiO₂ dispersed in alcohol solvent to form an electron transport layer [148]. Several other reports on solution-processed tandem cells have used ZnO or TiO₂ via various processes as n-type contact and PEDOT:PSS for p-type contact, yielding efficient tandem cells.

3.4.3 Efficient Recombination Inside ICL

The charge recombination rate inside the ICL should match the charge extraction rate from the two sub cells, otherwise it will lead to accumulation of charges in the ICL, increasing the R _s and even induces significant loss in the photocurrent near V _OC, resulting in low efficiency devices with S-shaped J–V curves.

Gilot et al. observed such S-shaped J–V curves in their tandem solar cell devices and they attributed the problem to the non-Ohmic contact between ZnO and PEDOT:PSS in the ICL, which would form a counterdiode, and thus hinders charge recombination [146]. To restore efficient charge recombination in the ICL, a thin Ag layer was inserted between ZnO and PEDOT layer, and the S-shaped curve disappeared. Alternatively, an Ohmic contact between n-type and p-type layers inside ICL can be formed by heavily doping the two layers. In the case of PEDOT:PSS/ZnO ICL, further doping of PEDOT:PSS is not necessary, and the doping level of ZnO was increased by UV irradiation. Gilot et al. observed that the S-shaped curve disappears soon after UV irradiation and the V _OC of the tandem cell increases. Similar cases were reported by Sista et al. in tandem solar cell devices using PEDOT:PSS/TiO₂ ICL [147]. They observed S-shaped curves in such tandem devices, and when the devices were irradiated with UV light of wavelength below 400 nm, the J–V characteristics returned to normal and the PCE was thus greatly improved.

The UV activation phenomenon described above is useful to study the interface interaction between transitional metal oxides and heavily-doped conjugated polymers. A similar behavior was observed for single junction devices with TiO₂/PEDOT:PSS/Al as the composite cathode, confirming the theory that this transition from a high resistance state to a low resistance state is the result of Schottky-to-Ohmic transition of PEDOT:PSS/TiO₂ contact [147]. The p-type PEDOT:PSS and n-type TiO₂ form a metal–semiconductor contact with a triangular barrier at the TiO₂/PEDOT:PSS interface. At a low doping level of TiO₂, the triangular barrier width is large and blocks electrons in TiO₂ to recombine with holes from PEDOT:PSS. After irradiating with UV light, the free carrier concentration in TiO₂ increases significantly and the barrier width decreases to an extent that electrons can tunnel through the barrier. However, this transition to Ohmic contact is not permanent, as the device reverts back to the high resistance state after prolonged storage in the dark. Thus, methods that can realize stable doping of metal oxides are desired. One approach is the chemical doping using electron-donating species. Park et al. reported the doping of TiO₂ NPs by Cs₂CO₃ [148], causing a significant energy level shift of TiO₂. It was observed that Ti ions were partially reduced by Cs ion through charge transfer, thus increasing the n-type doping.

3.4.4 ICL for Inverted Tandem Solar Cells

As reported by Chou et al., the advantages of the tandem and inverted structure can be combined by employing a metal oxide-only interlayer (MoO₃/Al/ZnO) to connect two inverted BHJs devices [149]. An inverted tandem solar cell with P3HT:PC₆₁BM and poly[(4,4′-bis(2-ethylhexyl)dithieno[3,2-b:2′,3′-d] silole)-2,6-diyl-alt-(2,1,3-benzothiadiazole)-4,7-diyl] (PSBTBT, see Scheme 3.1):PC₇₁BM as bottom cell and top cell respectively can reach a PCE of 5.1 %. A low temperature hydrolysis process was developed to form a dense and smooth amorphous ZnO layer with excellent diode properties, while the crystalline ZnO appears rougher and more porous. In addition, the MoO₃ provides a much robust resilience against the sol–gel process compared to V₂O₅ or Al. As a result, the materials and configurations of the interlayer for the tandem architecture were no longer limited by the acidic PEDOT:PSS. Compared to PEDOT:PSS, the absorption of the metal oxide-based interlayer is also smaller, resulting in higher photocurrent for both single and tandem cells. Sun et al. also reported a multilayered ICL in inverted tandem cells [150]. MoO₃/Ag/Al/Ca interlayer was vacuum deposited, and followed by deposition of the rear cell via spin-coating process. Such an interlayer structure features high transparency and low R _s, as well as effective charge recombination, rendering an exact summation of V _oc (1.18 V) of the two subcells and a high FF (61.8 %). Maximum PCE of 2.78 % is mainly due to the fact of using identical subcells.

3.5 Summary and Outlook

The overall performance of BHJ–PSCs (for either conventional device, inverted device, or tandem device) is determined by a series of sequential optoelectronic processes including charge separation, transporting, and extraction, which can be controlled by the manipulation of intrinsic properties of the photoactive materials, the morphology of active layer and the electrical contact of interfaces between different layers, etc. The interface problems between different thin film layers inside a polymer solar cell device is neglected for a long time by researchers as most of their efforts were devoted to the development of new active materials, the optimization of active layer morphology. Thus only until recently did the mainstream researchers realize the importance of interface engineering on the photovoltaic performance of devices. Great improvements in performance were thus achieved for BHJ–PSCs field via the understanding of interface function and the development of highly efficient interface modification materials. As discussed in this chapter, the interlayers function in many aspects, such as improving interfacial electrical contact, passivating charge trapping states, altering electrode work function, controlling energy alignment, enhancing charge collection, inducing active layer phase separation, redistributing the incident light field, and even improving the yield and quality of device fabrication process. However, most of the reported interfacial materials function only the above-mentioned aspects partially and may result in some side effects. Moreover, some interfacial materials are sensitive to the active layer, i.e., functional for some certain photovoltaic materials but not for others. Therefore, the integration of all kinds of different functions into one material is still challenging and in the near future, the development of more efficient interfacial materials is still required. Nevertheless, the great progress in the field of BHJ–PSCs gives the researchers opportunity and confidence to develop more powerful interfacial materials and roll-to-roll compatible fabrication methods, and thereby pave the way for commercialization of polymer solar cells.