Demand for cheap renewable energy sources is driving the development of organic photovoltaic devices (OPVs), which are generally lower in cost to develop than their silicon counterparts. In comparison with silicon, however, OPVs have a relatively low conversion efficiency.
An OPV typically consists of an active material sandwiched between two electrodes (see Figure 1). When the system is illuminated, photons are absorbed inside the active layer, generating pairs of charge carriers, or excitons. To be an effective power source, these photogenerated excitons need to be dissociated into electrons and holes, to allow charge collection at the electrodes. This process must occur within the lifetime of the excitons, to avoid recombination.1 The active layer in organic solar cells is commonly around 100–300nm thick, because this results in practically complete light absorption in the spectral region near the solar maximum at 530nm. However, in organic materials, performance is then hampered by short exciton diffusion lengths (the distance an exciton can travel before the electron and hole recombine) and poor charge transport (especially electrons), leading to disappointing charge collection. As a result, an OPV’s efficiency is compromised by its relatively low photon absorption in ultrathin (∼50nm) active layer structures. One possible solution for capturing the light more efficiently is the application of metal nanoparticles (MNPs).2 MNPs enhance the absorption of solar radiation by near-field enhancement—when the large electric field (E-field) around the particles increases the probability of excitons dissociating into electrons and holes—and by increasing the forward scattering cross section (a measure of how much of the incoming light is scattered in the forward direction). The optical absorption spectra of MNPs depend on their size, so by tuning the size of the nanoparticles, we can achieve significant absorption in different parts of solar spectrum.3
We have devised a solar cell structure that takes advantage of E-field enhancement from 50nm-diameter gold nanoparticles (AuNPs).4To calculate the conversion efficiency of OPVs we consider three factors: photon absorption efficiency (PAE), exciton dissociation efficiency (EDE), and charge collection efficiency (CCE). When we incorporated the AuNPs, we saw an increase in both PAE and EDE, and we could increase CCE using a thinner active layer. In total, we achieved a 65% increase in total absorption using AuNPs. In finite-difference time-domain simulations the E-field intensity in immediate proximity to the AuNPs at their resonant frequency was enhanced by a factor of up to 120 (see Figure 2), which increases the EDE. In Figure 2, we see that the enhanced E-field around 50nm-diameter AuNPs occurs only at the surface, and does not extend beyond 15nm. With this key observation, and considering that the exciton generation and dissociation occur within the photoactive layer, we have confined this E-field enhancement inside the active layer by choosing the appropriate thickness of the buffer layer.
To incorporate the MNPs in organic photovoltaic devices, we developed a solution-processed method (where the polymer material is deposited in a solution on the substrate) that is low-cost and can be scaled up. We used silanized indium tin oxide (ITO) electrodes to self-assemble MNPs and showed that by employing an atypical silane—N ′ -(3-trimethoxysilylpropyl) diethylenetriamine (DETA)—we could uniformly deposit the MNPs onto the ITO surface with virtually no aggregation compared with conventional methods, and could adjust the surface coverage at will.4 Scanning electron microscope images show that we can achieve uniform AuNP-coated ITO with different surface coverages (see Figure 3). The advantages of this method are its practical simplicity, strong attachment of AuNPs to ITO, and uniform surface coverage tunable between 5 and 50%. Since the MNP-modified ITO electrodes are solution processed, the fabrication technique is compatible with large-area, low-cost, roll-to-roll processing.
Using the self-assembled AuNP-modified ITO we have fabricated plasmonic solar cells. Comparing the performance of 50nm active-layer plasmonic solar cells with that of 120nm active-layer devices without AuNPs (‘standard’) reveals the short-circuit current of the devices with nanoparticles is enhanced by at least 30% (see Figure 4). In future work, we may use the method to incorporate AuNPs in other optical resonance-based fields, such as in nonlinear photonics, where large E-field enhancements would allow the use of low-power lasers for optical switching applications.