Bulk heterojunctions

The limitation in a bilayer heterojunction may be overcome by preparation a bulk heterojunction cell. Bulk heterojunction is a blend of the donor and acceptor components in a bulk volume (Figure 2.7). The both components are usually prepared separately, blended in organic solvent and then grafted on the substrate. The blend exhibits a donor-acceptor phase separation in a 5-15 nm length scale. In such a nanoscale interpenetrating network, each interface is within a distance less than the exciton diffusion length from the absorbing site. The bulk heterojunction concept has heavily increased (orders of magnitude) the interfacial area between the donor and acceptor phases and resulted in improved efficiency solar cells [82]. Meanwhile, separated charges require percolated pathways for the hole and electron transporting phases to the contacts. In other words, the donor and acceptor phases have to form a nanoscale, bicontinuous, and interpenetrating network [83]. Therefore, the bulk heterojunction devices are much more sensitive to the nanoscale morphology in the blend. Optimal morphology is achieved by a careful balance between the nanoparticle aggregation needed for good charge transport and phase separation needed for efficient exciton dissociation. Wang et al. [84] have postulated that the capped pyridine molecules at the CdS surface facilitate dissolution of CdS nanocrystals in the pyridine solvent. Moreover, pyridine acts as a surfactant which improves compatibility of CdS nanocrystals with the polymer chains (MEH-PPV) both in the solution and in the solid film. Homogeneous dispersion of nanoparticles in the polymer blends leads to the larger area of donor–acceptor interface which could improve the charge separation efficiency.

22

Figure 2.7 Bulk heterojunction configuration in organic solar cells.

The performance of photovoltaic devices may be also controlled by changing the size, shape and concentration of nanocrystals. The influence of semiconductor nanoparticle content on the efficiency of PV hybrid cell has been studied by Greenham at al. [52]. It has been shown that at low concentrations of spherical 5 nm CdSe nanoparticles in MEH-PPV only part of electrons have a continuous pathway to the external circuit. At higher concentrations (65 wt%) the nanocrystals begin to form a connected network and both components provide continuous ways for the charge carriers to the electrodes. The highest EQE (12%) was obtained at 90 wt% of CdSe.

However, even at the optimum concentration, the EQE is still low due to trapping of the electrons at “dead ends” in the nanocrystal network. Therefore, the much effort was focused on preparation of the hybrid cells containing nanorods or branched semiconductor structures.

It has demonstrated that one-dimensional nanorods [50, 64] or branched semiconductors [85–89] are used to create the percolation networks in the polymer/semiconductor blend improving the charge transport through the semiconductor phase to the electrode. Huynh at al. [50, 64] have demonstrated that change of the aspect ratio of CdSe nanocrystals from 1 to 10 results in the increase of

ITO glass PEDOT-PSS Aluminum

Light N-type + P-type

23

the external quantum efficiency at the wavelength of 500 nm from about 18% to 54%, respectively (Figure 2.8). By contrast, the cells containing hyperbranched CdSe approached maximum PCE values at lower nanoparticle concentration values than the other nanoparticle geometries [85]. Gur et al. [85] argued this was because of more efficient formation of charge transfer pathways. Furthermore, the cells based on hyperbranched particles revealed near-linear rise of JSC and power conversion efficiency with increased loading of CdSe, also at a low content of nanostructures.

This suggests that single incorporated hyperbranched particle can contribute independently to the output of the cell [85].

In another paper Gur et al. have shown that a unique tetragonal structure of CdTe nanocrystals gives rise to a natural ordering in the deposited films [90, 91]. Namely, three arms of each tetrapod contact the substrate at its base, while the fourth arm points up, perpendicular to the substrate. This ordering remains in the composite films created by spin-coating of the polymer–nanocrystal mixture. However, the efficiency of the device based on CdTe tetrapods and P3HT spin cast from chloroform solution are still low (less than 1%), suggesting the need of further improvement by optimization of device composition and morphology [90].

Figure 2.8 Influence of the increasing length (7, 30 and 60 nm) of 7-nm diameter CdSe nanorods on the external quantum efficiency of CdSe (90 wt%)/P3HT 200-nm thick solar cell. [50]

24

One of the major difficulties in preparation of the bulk heterojunction cell is the phase segregation of organic and inorganic phases. The key parameters are processing conditions as the type of the solvents used for preparation of the polymer/semiconductor mixture and temperature which control the rate of drying of the hybrid system. The influence of the rate of solvent evaporation on the formation of the π-stacked aggregates and in consequence, on the charge carrier mobility, has been also confirmed [92–94]. Yang et al. [94] have shown that drop-casting of P3HT reinforces the lamellar π-π stacking perpendicular to the substrate, depending on the type of solvent. The charge mobility in the film decreases in the order CHCl3 > THF >

toluene > CH2Cl2. The significantly lower charge mobility in the films deposited from toluene and dichloromethane was attributed to the formation of distinct grain boundaries and the short nanorod-like structures. In contrast, the polymer chains cast from CHCl3 form highly interconnected nanofibrillar networks without grain boundaries providing fast charge transport through the film.

Solvent evaporation causes a transition from a one phase to a two phase region.

Extensive nanoparticle aggregation and phase separation of the polymer phase is favored as nanoparticle and polymer volume fraction increase. In well-controlled solvent evaporation processes this gives a bicontinuous solid dispersion. It is in this way that non-equilibrium morphologies can be frozen (kinetically trapped) into the photoactive layers of nanoparticle–polymer PV cells.

Phase segregation during casting may lead to formation of “dead ends” and isolated domains trapping the charge carriers. The polymer wetting and charge transfer dynamics may be improved by means of interface modifiers. One option is to use a surfactant containing the head group of a high affinity to semiconductor and the end group providing solubility in the chosen solvent. Milliron et al. [95] have

25

demonstrated that pentathiophene phosphonic acid (T5-PA), served as a complexing agent with respect to the CdSe nanocrystals, could be used as a third component in nanocrystal–polymer devices mediating the charge transport between these two materials.

Alternative approach applied by Liu et al. [96] consisted in functionalization of the polymer chain ends with amine groups, which are strongly adsorbed onto CdSe nanoparticles. The endamino groups attached to the P3HT replace the pyridine surfactant on the surface of nanorods and thereby enhance the miscibility of the polymer with the nanocrystals and improve the power conversion efficiency of hybrid device. The ligand-exchange method has been also applied to graft the carbodithioate-containing oligo- and polythiophenes on the CdSe surface [97].

Another strategy, which utilizes direct attachment of P3HT onto the CdSe nanorod surface, has been applied by Zhang et al. [98] by coupling of vinylterminated P3HT to arylbromide-functionalized CdSe. The recent advances in preparation the organic–inorganic nanohybrids through tailoring of semiconductor nanocrystals with conjugated polymers have been discussed by Lin in the review paper [99].

An efficient way of impeding the phase segregation is also to induce strong interaction between organic–inorganic components through molecular recognition phenomena. This concept has been successfully used by De Girolamo et al. [100] in preparation of the composites of diamino-pyrimidine functionalized P3HT with thymine-capped CdSe nanocrystals. This molecular processing method has been also exploited for the alternating deposition of thymine-capped nanocrystals and functionalized polymer monolayers by layer-by-layer assembly [101].

Thermal treatment of the polymer in the vicinity of its melting temperature (Tm) was proven to be an effective method to increase the hole transport velocity due to enhanced chain ordering (π-stacking) [102,103] and self-organization of the polymer

26

[64]. According to Wang et al. [84], improvement of the device performance after annealing is mainly attributed to reorganization of the inorganic nanocrystal/polymer interface, leading to the increase of exciton dissociation efficiency and reduction of the recombination loses. It may be also a result of reduced interface defects [104] and improved phase-structured morphology [105]. Dittmer et al. [106] have postulated that annealing of hybrid system promotes the equilibrium morphology of a spin-coated film, improves crystallinity within the phase-separated networks and thereby facilitates the charge transport to the electrodes. Thermally induced crystallization enhances also of the efficiency of the charge transport across the interface between the bulk of material and the collecting electrode (η_tr). Most of the reports in the literature on the influence of annealing on the efficiency of hybrid cells concern the polymer/PCBM systems but the mechanism of performance enhancement for the polymer/inorganic semiconductor hybrid systems is less clear.

2.4.3 Porous and Vertically ordered semiconductor–polymer systems