Hybrid Organic/Inorganic Nanocomposites for Photovoltaic Cells

Inorganic/organic hybrid solar cells have attracted a lot of interest due to their potential in combining the advantages of both components. To understand the key issues in association with photoinduced charge separation/transportation processes and to improve overall power conversion efficiency, various combinations with nanostructures of hybrid systems have been investigated. Here, we briefly review the structures of hybrid nanocomposites studied so far, and attempt to associate the power conversion efficiency with these nanostructures. Subsequently, we are then able to summarize the factors for optimizing the performance of inorganic/organic hybrid solar cells.


Introduction
In searching for renewable energy sources, solar cells represent a class of promising candidates. In the past decades, a tremendous amount of research work has been carried out in the relevant fields and many types of photovoltaic cells have been developed and tested. So far, all of these photovoltaic cells are still a long way from satisfying the various requirements demanded, and many critical issues remain or are being studied. Most of the issues being studied target the key goal, the improvement of the power conversion efficiency of the photovoltaic cells.
Very many types of photovoltaic materials, e.g., inorganic semiconductor nanoparticles, silicon, organic dyes, conducting polymers, C 60 derivatives, and graphene etc. as well as their combinations have already been attempted. Each type of these materials has its merits and drawbacks. Organic conducting polymers with a variety of chemical structures, for instance, have promising properties, OPEN ACCESS such as easy processing, possible recyclability, relatively low cost, scalability and applicability as sustainable materials. Inorganic semiconductors, on the other hand, possess better electronic properties, e.g. a high dielectric constant, a high charge mobility and thermal stability etc., while their nanoparticles exhibit enhanced electronic, photo-conducting and luminescent properties. To take advantage of both materials, hybrid nanocomposites of inorganic semiconductors and conducting polymers are of great interest, especially as candidates for photovoltaic-cell materials, where the combined absorption band of both materials can also better harvest sun light. With this concept, conducting polymers have been combined with different inorganic semiconductors. In addition, photo-induced charge separation mainly takes place at the interfaces between inorganic semiconductors and conducting polymers in these hybrid materials, where electrons are injected from the conducting polymers into inorganic semiconductors and holes remain in the polymers. Actually, this interfacial charge separation can to some extent prevent the recombination of separated electrons and holes. The reviews by Saunders et al. [1], Zhou et al. [2] and Wright et al. [3] have given a fairly complete picture of this concept of hybrid solar cells and have summarized a full list of issues, especially many experimental details, targeting a better power conversion efficiency in these hybrid systems. So far, a lot of studies to find good combinations of inorganic/organic materials and optimal hybrid architectures have been attempted, and significant progress has been achieved in recent years. However, the power conversion efficiency has not yet caught up to the level that the dye-sensitized photovoltaic materials achieved, ~12%. Nevertheless, we believe that there is still room for improvement of the hybrid organic/inorganic photovoltaic materials. For example, good combinations of the paired components and well-designed hybrid structures should be among the ways of achieving this target.
In this brief review, we look into recent progress witnessed in the field of hybrid organic/inorganic photovoltaic materials for solar cells from the physical point of view and focus on their architectural designs. By comparing the various pairs of organic/inorganic components and nano-structures proposed, we can summarize factors affecting their photovoltaic conversion efficiency and architectures achieved for these hybrid systems in order to clarify the relationship between structures and underline the physical mechanism therein, and then outline guidelines for future improvement.

Charge Separation at the Interfaces between Inorganic Semiconductors and Conducting Polymers
The interfacial charge separation is one key step in the photovoltaic processes of the hybrid materials. However, factors controlling the interfacial charge separation are not fully understood. It is commonly believed that a photogenerated exciton dissociates to generate a bound complex with charge transfer character at a donor:acceptor heterojunction and then separates into fully dissociated charge carriers [4][5][6] as illustrated in Figure 1. Here, the efficiency of charge generation depends on whether the interfacial charge separation can successfully compete with the recombination of the bound charge transfer complex and with the energy transfer to singlet or triplet excitons of either component. During this process, factors, such as the interface area, the nominal interfacial driving energy (the differences in energy level across the interface), the morphology of the interfaces/domains, the interfacial layer/molecules/bridges, the crystallinity of the components, the mixture homogeneity, and the migration of separated charges from the interfaces, can influence the charge separation rate.

Figure 1.
A scheme of the charge separation process at the Donor:Acceptor interface in a hybrid solar cell. The major photovoltaic steps include: photo-excitation into excitons (1), excitons migration to interfaces (2), charge transfer from the donor to the acceptor at the interface (3), charge migration to electrodes (4) and charge injections into electrodes. There are also competition processes, which can reduce the photovoltaic conversion efficiency, including the recombination of the excitons in the donor and separated charges at the interfaces.
Bansal et al. demonstrated that the dissociation of photogenerated polaron pairs at the heterojunctions of CdS and several conducting polymer nanocomposites is assisted by the presence of crystalline electron acceptor domains [6]. In this case, the crystallinity and larger size (>1 nm) of CdS domains can encourage delocalization of the geminate pair state and attenuate the required interfacial driving energy. Oosterhout et al. suggested that the morphology of the hybrid system is important [108], and the inorganic component therein improves the thermal stability. In their experiments, the internal quantum efficiency of the hybrid P3HT/ZnO nanocomposites increased with the thickness of the hybrid layer until the optimal layer thickness, which was significantly larger than 100 nm. However, the morphology analysis indicated that larger domains of P3HT and ZnO, and larger distance (>5 nm) between them in the thinner sample likely led to the lower internal quantum efficiency. Our recent study, using scanning electrostatic potential microscopy to directly monitor the interfacial charge separation in the hybrid nanocomposites of PPV and TiO 2 , showed that the holes were only able to travel around 20 nm away from the interfaces after charge separation [109]. These results suggested that there should be an optimal domain size for the hybrid nanocomposites to maximize the efficiency of photo-induced charge separations; in turn the structural design of the hybrid photovoltaic system is critical.

The Nano Architecture of Hybrid Photovoltaic Materials
Because the interfacial charge separation is the critical step in the whole photovoltaic process, the larger the interfaces, the more the opportunity for the excitons to reach the interfaces, and probably the higher the conversion efficiency. Therefore, most of the hybrid photovoltaic materials that have been studied are nano-structured composites. There are many types of nanostructures, including particles, rods, tubes, tetrapods, sheets, needles, and porous network, etc., and the way to mix the components together can be either disordered or ordered as shown in Figure 2.

Quantum Dots/Nanocrystal Based Systems
Quantum dots/nanocrystals of inorganic semiconductors based systems for solar cells have been extensively studied over the past decades [22,44,[110][111][112][113][114][115], because quantum dots (QDs)/nanocrystals are expected for a larger surface area than bulk materials. In blending with a conducting polymer, these hybrid systems can also take advantage of the large-area solution processing. In addition, the quantum size effect leads to a tunable energy gap [54] of QDs to allow energy level matching between the components in the solar cell systems, while a better coverage of the spectrum of sunlight is also expected.
The blending of QDs/nanocrystals with conducting polymers is not as simple as just mixing and coating. The procedure details of blending influence the final nano-structure and hybrid interfaces, which can determine the photovoltaic processes therein. There is still considerable room for optimization of their power conversion efficiency.
Because capping is important in the synthesis of QDs/nanocrystals and different cappings are known to alter the contact between QDs and conducting polymers, several capping strategies have emerged in the literature: the ligand exchange [110,116,117], the use of novel surfactants [118,119], the synthesis of QDs with thermally cleavable ligands [120], the synthesis of QDs in solutions of conducting polymers [52,121], the polymerization of conducting polymers on the surface of QDs [122], and the deposition of a conducting polymer containing a soluble precursor of the QDs [123]. More details on the role of capping ligands/conducting polymers can be found in recent reviews [124][125][126][127][128]. Reynolds et al. reported in situ growth of CdS QDs from the precursor in a P3HT polymer film, and found that the solar cells from the blends of P3HT and in situ growth of CdS QDs displayed much higher open-circuit voltage (V OC ) and short-circuit current (J SC ) than that from the blend of P3HT and CdS QDs capped by oleic acid or hexylamine. They proposed that the difference is due to the morphology/interfaces, where the in situ grown CdS QDs formed an interconnected layer, in sharp contrast to the discrete, unaggregated nanocrystals in the pre-synthesized and capped QDs [123]. This indicates a physical fact that the contact between the conducting polymer and the QDs, as well as between QDs themselves, is essential for the photovoltaic performance, because interfacial charge separation and charge migration throughout the photovoltaic process can eventually be affected by these contacts. Similarly, the solution conditions during QD synthesis and QDs/polymer mixing, and later annealing of the blend films can also influence the phase separation, the interfacial conditions and in turn the power conversion efficiency.

Nanorods/Nanotetrapods Based Systems
Nanorods/nanotetrapods of inorganic semiconductor based systems have also received considerable interest. [21,77,[129][130][131] Though nanorods/nanotetrapods are not so easy to synthesize as nanoparticles/QDs and the processes of the corresponding hybrid systems require more harsh conditions, the larger surface area to volume ratio of nanorods ( Figure 2b)/nanotetrapods (Figure 2c), the smaller number of interparticle hops necessary for electrons to leave the device, and thus the better connected paths are suggested as enhancing the power conversion efficiency [132][133][134]. Sun et al. [132] compared blends of branched CdSe nanoparticles and conjugated polymers with similar blends of nanorods/polymer, and found a significant increase in external quantum efficiency when branched nanoparticles were used. Lee et al. [133] further compared hybrid solar cells based on long-armed nanotetrapods and short-armed nanotetrapods, and found that the nanotetrapods with longer arms lead to greater short-circuit current density and fill factor than those with short arms. They suggested that the higher charge transport efficiency in long-armed nanotetrapods increases both the quantum efficiency and the fill factor, and thus leads to better power conversion efficiency in the hybrid systems. Morphologically, the hybrid system based on long-armed nanotetrapods presents a longer percolation pathway and larger domains of P3HT, where more favorable hole drift and diffusion can be expected. This can minimize exciton recombination events along the charge transport pathways. Of this type of hybrid system, Dayal et al. [41] reported a high power conversion efficiency of 3.13% in the nanotetrapods of a CdSe based system, where a low band gap (~1.4 eV) conjugated polymer PCPDTBT was incorporated. Soci et al. [135] claimed that a band gap of 1.5 eV was optimum for conjugated polymers in hybrid solar cells.
Another interesting nanostructure is based on inorganic semiconductor quantum dots and nanorods of the conducting polymer. Ren et al. [136] used solvent-assisted chemical grafting and ligand-exchange to synthesize P3HT nanowires and decorated them with CdS QDs, and the power conversion efficiency was as high as 4.1%. The high efficiency in this hybrid system can be attributed to the enhanced charge collection efficiency due to continuous percolated nanomorphology in both components, where more efficient transport pathways for both electrons and holes are available. In addition, the charge separation efficiency may also be enhanced between P3HT nanowires and CdS QDs. This is an interesting hybrid structure, as an invert to the nanorods of the inorganic semiconductor based system discussed above. In both nanorod based systems, the long nanorods provide better pathways for charge transport and reduce the need of inter-particle hops for charge carriers.

Nanoporous Inorganic Semiconductor-Based Hybrid Systems
There are other various choices of hybrid structures for the disordered nanocomposite system. Among them, the nanoporous structure (Figure 2d) is of great interest, and many attempts have been made based on porous inorganic semiconductors, such as TiO 2 [137][138][139], CdS [140], ZnO [141], CdSe(ZnS) [142] etc. De Freitas et al. compared films of P3HT:aerogel of CdSe(ZnS) with films of P3HT:quantum dots with a similar ratio of each component, where the films of the P3HT:aerogel showed increased photocurrent and charge generation. This was attributed to the existence of a more interconnected network, which extended the lifetime of the photoinduced charge separated state and reduced the chance of recombination of the separated charges [142]. The idea of an interconnected network of the hybrid components in the systems is the same as using branched inorganic nanoparticles when mixing with conducting polymers. However, the performance of this type of photovoltaic cell achieved is still not as good as expected, because there are still other factors that can limit the power conversion efficiency, e.g., poor contact between the porous semiconductor and the conducting polymer. Therefore, improvements are still eagerly anticipated for this type of photovoltaic material, particularly in obtaining better interfaces between the polymer and the porous inorganic base.

Ordered Nanorod Array (NRA) Based System
In the above hybrid systems, the very large interfacial area achieved between the inorganic semiconductor and the conjugated polymer can lead to efficient dissociation of excitons into holes and electrons. However, the charge carrier recombination inevitably takes place here or there during the charge transport through the random networks of the material, and eventually limits the PCE. From the above discussion, it is quite clear that improved pathways for charge transport can greatly help the performance of the solar cells. In this regard, ordered hybrid solar cells are proposed as a promising structure, in which the donor and acceptor materials are interdigitated at the nanometer scale, providing more direct pathways for charge transport and avoiding the need for inter-nanoparticle hops. The ordered nanorod array based systems have been proposed and examined in many research laboratories [9,61,[143][144][145][146][147][148][149][150][151][152][153][154][155][156].
Lee et al. [157] compared a hybrid system based on a vertically aligned 180 nm ZnO NRA with a bilayer one, and found an ~2.7 enhancement of the short-circuit current. Similarly, Choi et al. [144] reported a better PCE in a photovoltaic device based on a ZnO nanostructure length of 100 nm than in an equivalent device without ZnO nanorods. However, the achieved power conversion efficiency of bicomponent hybrid systems (ZnO NRA/P3TH) is not as good as those based on disordered nanorods. The introduction of third party molecule/materials at the interfaces can greatly enhance the performance of the final heterojunctions, e.g., using ZnO/CdS-core/shell heterostructures [158], adding a third component of a small dye molecule [141,147,159] or [6,6]-phenyl-C 61 -butyric acid methyl ester (PCBM) [156]. Compared with the original bicomponent hybrid system of the conjugated polymer/inorganic semiconductor NRAs, an increment in the power conversion efficiency of the multicomponent NRA based system was found. In the case of ZnO NRA/P3HT system adding PCBM, the increment of the PCE is more significant, but it is still far below the best performance found for the PCBM polymer solar cell, which is now over 10% [160]. Therefore, it is hard to draw a conclusion that the increment of PCE is due to adding a third component to these solar cells. Instead, it may be just a "half pure" PCBM/conducting polymer hybrid system. Nevertheless, the maximal PCE for these ordered NRA hybrid systems is only comparable to that of their equivalent disordered systems. Though the interfacial area in the ordered NRA hybrid system is larger than that in the plane bilayer hybrid system, it is much smaller than the disordered nanorods based equivalent. The impact of the relatively limited interfacial area on the PCE counteracts that of the better pathways for charge transport.

Ordered Nanotube Array (NTA) Based System
To increase the interfacial area in the ordered nano array hybrid system while keeping the effective charge transport pathways, ordered nanotube arrays as shown in Figure 2f were used. [15,88,90,[161][162][163][164][165][166][167][168][169][170][171][172] However, experimentally those nanotube arrays were not as perfect as displayed in Figure 2f, e.g., a typical FESEM of TiO 2 nanotube arrays as shown in Figure 3 [170], so improvements in the chemical/electrochemical synthesis are needed. Nevertheless, a better performance of the ordered nanotube array based system is expected because of the larger interfaces than their nanorod equivalents. However, similar to the ordered nanorod array based system, the PCE for TiO 2 NTA based systems reported by different groups varied by more than one order of magnitude (Table 1). [15,88,90,[161][162][163][164][165][166][167][168][169] Again, the introduction of third party molecules/materials at the interfaces can greatly increase the PCE of these hybrid systems. Li et al. [15] reported that the PCE of the hybrid system Au/P3HT/CdS-NTA/FTO has a ~4 fold increase in the PCE compared with the hybrid system Au/P3HT/NTA/FTO. Besides the CdS shell, quite a few dyes [164,165,169] and PCBM [88,161,163] have also be incorporated into the P3HT/TiO 2 -NTA system, and in most cases the PCE can exceed 1%. For example, Shankar et al., reported a PCE of 4.1% in a hybrid system of P3HT-PCBM-TiO 2 [163]. Interestingly, to improve the infiltration of conducting polymers into NTAs and their contact, in situ electrochemical polymerization of polythiophene was utilized, but the PCE of 1.46% obtained was not as good as expected. Thus, although this improvement may not be significant, it should still be the right direction to get better interfaces between inorganic and organic components in these ordered hybrid systems. However, there are still quite a few other factors that cannot be neglected.

Important Factors in the Performance of Photovoltaic Materials
Some representative hybrid systems as described above with reported power conversion efficiency are summarized in Table 1. The best performance achieved among these hybrid systems is one combining P3HT and nanotube arrays of TiO 2 [163]. However, several hybrid systems of other types of structures can also be very close in performance [136,156]. From the power conversion efficiency among these systems as listed in Table 1, we can confirm the influence of the factors discussed in the previous section on the charge separation processes. However, their importance varies, and we suggest that the following factors are critical and need optimization for the improvement of these hybrid systems. First, the charge transportation along both components can determine the current density and so affect the final performance. A continuous network of the components, e.g., long nanorods and branched nanorods, can greatly reduce the need for charge carriers to hop between nanoparticles during their transportation to the electrode, so that the nanorods and nanotube array-based hybrid systems showed overall better performances in reports. However, the carrier transportation in conducting polymers is especially one of the bottlenecks in these hybrid systems, and the improvement on the quality of the conducting polymer network can greatly enhance the power conversion efficiency of the final system as reported by Ren et al. [136] Second, the interfacial area and contact between inorganic and organic components determine the charge separation efficiency. The nanotube array-based systems show on average better performance than nanorod array-based systems because the surface area of nanotubes is larger than that of nanorods of similar sizes. At the same time, a 3D network of porous structure-based systems, although it would be expected to enhance the carrier transportations, is not so successful, which is likely due to the fact that the inorganic-organic interfaces are far away from good contacts. Third, the energy-level alignment at the interfaces is also very critical for optimal hybrid systems, so their components should be carefully chosen. According to our understanding from the literature, the following guidelines can be stated for the improvement of hybrid photovoltaic systems of inorganic semiconductors/conducting polymers.
First, the right combination of inorganic and organic semiconductors should be chosen. For this purpose, the LUMO of the conducting polymer needs to be aligned with the conducting band of the inorganic semiconductor, or the band-gap of the inorganic semiconductor can be tuned by quantum confinement effect. In addition, an optimal band-gap of ~1.5 eV is suggested for conducting polymers in the hybrid system. This is physically sound and reasonable, because such a band-gap may balance a large enough absorption of sunlight (thus a good photocurrent) and a high enough output photo-voltage.
Second, nanostructures should be used to provide a large interface for the enhancement of the charge separation process, but at the same time the structural design has also to consider the connection in each component to facilitate charge transportation. Thus, long nanorods/nanotubs, branched nanopods, or a porous network may be utilized, and to accelerate the charge transportation, an ordered nanorod/nanotube array can be considered.
Third, good contact between organic and inorganic components should also be considered as a prerequisite due to the improvement in charge separation at the contact interface. Here, optimal conditions for blending of components (e.g., in situ synthesis of mixed components and annealing etc.), exchanging of the capping group of the inorganic nanoparticles, and incorporation of small organic dye molecules can be used. A detailed review in this respect on the experimental conditions of synthesis was given by Zhou et al. [177]. In addition, we think that the hybrid system is not necessary to be bi-component and suggest that another inorganic or organic semiconductor can be introduced as a third component to enhance the light absorption, to optimize the energy-level alignment and to improve the contacts between components. For instance, PCBM and its derivatives have been used as the third component, and improvements in PCE have been reported, which are however still not comparable to that of the PCBM/derivative and conducting polymer hybrid systems. The reason may be that the introduction of a third component may also break the continuity of charge transportation along each component. Good comparisons of these two types of hybrid systems have been reviewed in detail by Saunders [178] and Reiss et al. [124] However, inorganic semiconductors as one of the components in these hybrid systems have their unique advantages too, and a better PCE can be expected. Nevertheless, in future architectural designs of these hybrid solar cells, the negative effect, i.e., the breakage of the continuity of the charge transportation pathway, on the introduction of multi-components should be minimized.
Fourth, a nano-structured network of a conducting polymer in the hybrid system may be more helpful, because the mobility of holes in conducting polymers is much more limited than that of electrons in inorganic semiconductors.
Last but not least, the inorganic/organic blending layer in the photovoltaic system is more or less symmetrical, so physically, the random diffusion of the charge carriers will give them similar chances to move to both electrodes and thus decrease the effective power conversion of sunlight. Therefore, if the structural design can break or limit the symmetry in this blending layer, the power conversion efficiency of the resulted hybrid system can also be improved. In the work of Liu et al. [174], the highest PCE (~5.5%) of this type of solar cell has been reported so far, the vertical phase segregation of PDTPBT and PbS x Se 1−x has been claimed as one of the reasons for the PCE reached. In fact, this vertical phase segregation resulted in a nanostructure more or less breaking the symmetry of the components usually seen in other reports, although they did not claim the effect from symmetry breakage.

Conclusions
Here, we briefly reviewed recent progress on hybrid photovoltaic systems of inorganic semiconductors and organic conducting polymers. We focused on nanostructures and their impact on the photovoltaic performance of these hybrid materials. From the results from the literature, we summarized and clarified the critical factors for optimization of the hybrid inorganic/organic semiconductors. Finally, based on our knowledge, we suggested a few guidelines for structural designs of these hybrid photovoltaic systems in order to enhance their power conversion efficiency.

Acknowledgments
We gratefully acknowledge support from the Research Start Fund for Talent Recruitment, Chongqing University, and the fundamental and advanced research program of Chongqing (grant # cstc2013jcyjA10047), China.

Conflicts of Interest
The authors declare no conflict of interest.