1. Introduction

In recent years, large research efforts have been made to find fast and cheap materials that can be used in holographic applications. Typically, photorefractive inorganic crystals like iron doped lithium niobate are used in these applications [1, 2]. But beside their cost intensive and complex growth processes, their photorefractive performance is rather slow and they can only be manufactured in small dimensions. Photorefractive composites are one type of materials, which feature the possibility to replace inorganic crystals in holographic applications [3]. Especially if large area devices, fast response times, or easy and cheap fabrication are needed, photorefractive composites outmatch their inorganic counterparts. Due to these properties, photorefractive composites are ideal candidates to be employed as optical beam couplers, fast storage devices [4], holographic 3D displays [5, 6], and tomographs for living tissue [7].

The photorefractive (PR) effect describes a reversible refractive index change, caused by inhomogeneous illumination of the material [8]. This refractive index change has two underlying processes: photoconductivity and the electro-optic effect. When a photorefractive composite is illuminated with a light pattern, absorbing sensitizer molecules are excited in regions with high light intensity. Under the influence of an applied external electric field, these excitons are separated, and the mobile holes move through the (polymeric) hole transporter to the regions with low light intensity, while the electrons remain immobile in the excited sensitizer molecules. In the dark regions, holes can be trapped and hence the redistributed charges lead to the formation of a space-charge field. Nonlinear optical units, also called chromophores, provide the electro-optic effect and therefore can translate the space-charge field into the desired refractive index change [9].

One challenge in the optimization process of PR composites is the reduction of the external electric field to achieve a refractive index change [10]. Currently, several tens to hundred V μm⁻¹ have to be applied. This challenge can be addressed in different ways. One possibility is to find improved nonlinear optical units, that can generate the refractive index change more efficient and do not need large electric fields for preorientation. Another possibility is the optimization of the photoconductivity. As the photoconductivity is a combined phenomenon of charge generation and charge mobility [11], optimization of both effects can improve it. An enhanced charge generation will probably lead to a stronger space-charge field with reduced build-up time. Also an improvement of the mobility will lead to a faster redistribution of the charges and therefore a more rapid generation of the space-charge field. Recently, we achieved dramatically improved charge generation when we replaced C₆₀ by the perylene bisimide dimer DiPBI as sensitizer in the well-known composite of the electro-optic liquid crystal 4-cyano-4-n-pentylbiphenyl (5CB) and the conductive polymer poly-n-vinylcarbazole (PVK) (5CB:PVK:C₆₀) [12–14]. In this paper, we demonstrate an approach to further improve the photoconductivity by employing the highly conductive N, N′-diphenyl-N, N′-bis(3-methylphenyl)-[1, 1′-biphenyl]-4, 4′-diamine (TPD) embedded in polystyrene (PS) as charge transporting agent [15, 16] in our composites. TPD based charge transport has been studied for several years. Due to its high mobility, it is often used as a functional group in charge transporting polymers [6, 17, 18]. Additionally, charge transport in pure films of TPD and TPD doped polystyrene (PS:TPD) has been studied [16, 19]. Also the formation of charge transfer complexes in PS:TPD with C₆₀ has been observed [19].

2. Sample preparation and experimental techniques

To analyze the influence of PS:TPD on the photoconductivity, we apply our well characterized composition 5CB:PVK:DiPBI with a weight ratio of 40.18:59.80:0.02 wt% as reference system, and replace PVK by PS:TPD in a weight ratio of 80.00:20.00 wt%. If PVK is replaced by pristine TPD or higher doped PS, composites tend to unwanted crystallization and the composite loses its transparency [20]. Because of the electro-optic as well as the plasticizing properties of 5CB, we have chosen this material as a nonlinear optical unit. As an advantage, no further plasticizing component is needed to adjust the glass transition temperature T_g of the composites. Referring to the publication of Zhang and Singer [12], T_g for the PVK comprising samples is about 40 °C. Khan et al. reported T_g ≈ 70 °C for PS:TPD at a ratio of 80.00:20.00 wt% [20]. Assuming that 5CB further lowers the glass transition temperature, we estimate T_g for the PS:TPD comprising samples to be close to room temperature. The energy levels depicted in Table 1 illustrate that the differences between PVK and TPD are rather low and both should be suitable to build an efficient charge transporting complex with DiPBI. The difference between the HOMO levels of DiPBI and TPD is even larger than between DiPBI and PVK. This enhanced energetic difference can contribute to an improved charge separation between both components and therefore both can build a more efficient charge transfer complex than DiPBI and PVK. Because of the lower HOMO level of 5CB compared to the charge transporting agents, trapping of holes by 5CB molecules would be unlikely. Polystyrene works as an inert host material for TPD [20], thus trapping by PS is not expected and all traps are considered as intrinsic and caused by the conformation of the components. Higher amounts of PS may lead to an increased spatial disorder and thus broadening of the density of states and more intrinsic traps, because TPD molecules are no longer surrounded by other TPD molecules, and the conductivity of PS:TPD is reduced. At the chosen ratio of PS and TPD, we assume that PS does not incorporate large amounts of traps into the composite.

Table 1. Energy Levels of the Components 5CB [21], PVK [22], TPD [23], and DiPBI [24]

View Table

All components except DiPBI are bought from Aldrich and used without further purification. The synthesis of DiPBI is described in a previous report [25]. For sample preparation, all components are dissolved in chloroform at the desired ratio. After this step, the composite is dropped onto indium tin oxide (ITO) coated glass substrates and annealed in an oven at 55 °C for four hours. Finally, a d = 50 μm thick spacer foil and the second ITO coated glass is added and the sample is melt-pressed at 90 °C. Absorption spectra measured with a Jasco V-530 UV/VIS spectrometer are depicted in Fig. 1. The DiPBI doped PS:TPD sample provides the strongest absorption, while the unsensitized version possesses the weakest absorption. Although there are small differences, the absorption of the composites is clearly governed by the sensitizer DiPBI. Employing a modified version of the Beer-Lambert law [26], absorption co-efficients of the sensitized samples at a wavelength of 532 nm are calculated by α = A/d to α_PS:TPD = (20 ± 1) cm⁻¹ and α_PVK = (16 ± 1) cm⁻¹. Here, A is the measured absorption and d is the sample thickness.

 

Fig. 1 Absorption spectra of samples containing PS:TPD or PVK as hole transporter.

Download Full Size | PDF

The characterization of the photoelectric properties of the samples is carried out by measurements of the photocurrent I_ph with a Keithley 6485 picoammeter under homogeneous illumination by an expanded laser beam. The hole electrode area a = 39 mm² of the samples is irradiated with an intensity of W = 16 mWcm⁻² at a wavelength of 532 nm, and an applied external electric field E_ext. During the measurement procedure, the dark current I_dark as well as the current under homogeneous illumination I are investigated. The photocurrent I_ph is given by I_ph = I − I_dark. Then the photoconductivity is calculated by

The photorefractive qualities are investigated for different applied electric fields by the well-known two-beam coupling technique [28]. We illuminate the sample with two interfering laser beams, each with an intensity of W_1,2 = 8 mWcm⁻². The sample normal is tilted with respect to beam 1 by 40° and beam 2 by 60°, resulting in a grating spacing of 2 μm, if a refractive index of n = 1.7 is assumed. The energy transfer between both p-polarized beams is given by the gain coefficient Γ, calculated with

3. Photoconductive and photorefractive properties

The photocurrent measurements reveal stronger dark- and photocurrents for the PS:TPD comprising samples. These results can be directly transferred to the photo- and dark conductivities, as their relation is linear. At an applied field of 30 V μm⁻¹, the dark conductivities are σ_d,PVK = (0.013±0.001) pScm⁻¹ and σ_d,PS:TPD = (0.036 ± 0.002) pScm⁻¹, and the corresponding photoconductivities σ_ph,PVK = (0.009 ± 0.001) pScm⁻¹ and σ_ph,PS:TPD = (0.112 ± 0.004) pScm⁻¹. Therefore, the dark conductivity in the PS:TPD comprising composites is about three times larger than in composites incorporating PVK as hole transporter. In contrast, the photoconductivity is 12 times larger for PS:TPD than for PVK containing samples. The internal photocurrent efficiencies, calculated by Eq. (2), in dependence on the applied electric field are depicted in Fig. 2. It can be observed, that both compositions provide an explicit increase of ϕ_int with rising E_ext. For the PS:TPD composition, an enhancement of the internal photocurrent efficiency by a factor 11 at an electric field of 30 V μm⁻¹ is observed. The values at this field are (5.0 ± 1.0) × 10⁻⁶ and (5.5 ± 1.1) × 10⁻⁵ for PVK and PS:TPD comprising samples, respectively. This enhancement is attributed to an improved charge separation between DiPBI and TPD molecules, and a much larger mobility of the holes in TPD, caused by a reduced trap concentration compared to PVK. Further aspects, such as the higher mobility in pristine TPD relative to pristine PVK [17], or the change in the ionization potential between PVK and TPD [19] also contribute to the enhancements. Due to the high mobility, charges are transported more effective through the composite and therefore the enhanced photoconductivity can be explained. The change in the ionization potential crucially enhances the charge separation between sensitizer and charge transporting agent and hence, also contributes strongly to the enhanced photoconductivity. The increased dark conductivity hints to a decreased trap concentration in PS:TPD. Regarding to these aspects of the PS:TPD system, it allows for effective charge generation and transport at low applied electric fields.

 

Fig. 2 Comparison of the internal photocurrent efficiency of samples containing PS:TPD or PVK as hole transporter in dependence on the applied electric field.

Download Full Size | PDF

Considering the photorefractive properties during the two-beam coupling experiment, we observe increasing light intensity in one beam, while the intensity in the other beam decreases. When we change the direction of E_ext, the direction of the energy coupling also changes. Applying Eq. (3), we calculate the gain coefficient of the two-beam coupling. Figure 3 illustrates the dependence of Γ on the applied electric field. The gain coefficient rises superlinear with increasing E_ext for both compositions. At an applied field of E_ext = 30 V μm⁻¹, Γ is calculated to (8.7 ± 1.1) cm⁻¹ and (4.8 ± 0.6) cm⁻¹ for PVK and PS:TPD comprising samples, respectively. Therefore, the gain for the PVK containing composite is about 1.8 times larger. Including the absorption coefficient calculated in section 2 into the determination of gain, no net gain is observed for both composites. At an applied field of 70 V μm⁻¹, we investigated a net gain of 85 cm⁻¹ in PVK comprising samples with similar content of DiPBI in our previous work [14]. We expect an improvement in PS:TPD comprising samples by a reduction of the absorption. This can be achieved by an optimized sensitizer content in the samples. In an enhanced composition, an observation of net gain would be likely.

 

Fig. 3 Dependence of the two-beam coupling gain coefficient on the applied electric field for PS:TPD and PVK containing samples.

Download Full Size | PDF

The gain coefficient of the two-beam coupling experiment strongly depends on the strength of the space-charge field and on the phase shift of the refractive index grating. For highest gain coefficients, strong space-charge fields and a 90° phase shift of the refractive index grating with respect to the incident light pattern are desired. Therefore, the low gain of the samples can be caused by weak space-charge fields or too small phase shifts. The lower performance of the PS:TPD samples is explained by a too low trap concentration in the samples. For a strong and fast PR performance not only the mobility has to be high, but also there has to be enough trapping of redistributed charges for a strong space-charge field. High dark currents and therefore low trapping inevitably lead to low photorefractivity because of weak space-charge fields. Whereas TPD is known to provide a high mobility, PS is an inert material in PS:TPD. As reported earlier, charge transport operates well in the chosen ratio of PS and TPD [16]. Therefore, an increased trapping rate compared to PVK resulting in a reduced phase shift of the refractive index grating during the two-beam coupling experiment is not expected. Therefore we do not estimate the phase shift to be smaller than 90° and therefore, the small gain coefficients in the PS:TPD samples are attributed to the small amplitude of the space-charge field. To further improve PR properties of the composites, the contrast between photo- and dark conductivity has to be enlarged. In the presented composition, this may be realized by a reduction of the TPD content or doping with a further component to incorporate more traps. Due to these steps, the mobility of the holes will be lowered and so the photoconductivity, resulting in the need of higher applied fields. Therefore, a trade-off between the photo- and dark conductivity has to be found.

When the overall performance in a photorefractive material is low, it is difficult to observe a fast photorefractive response. Due to the small amplitudes of the two-beam coupling, we could not determine the time constants of the two-beam coupling, and therefore, could not prove the assumption that PS:TPD would help in the space charge build-up and increase photorefractive speed.

4. Conclusion

We characterize the photorefractive composite 5CB:PS:TPD:DiPBI by absorption spectroscopy, photocurrent measurements, and two-beam coupling and compare the results with the reference composition 5CB:PVK:DiPBI. Incisive improvements of the photoconductive properties in PS:TPD comprising composites are observed. These enhancements are attributed to the outstanding properties of the charge generator DiPBI and the hole transporting PS:TPD system, namely: broad absorption, highly effective charge generation and transport. Consequently, the choice of a TPD based charge transporting agent in combination with DiPBI as sensitizer is a promising approach for photorefractive composites operating at reduced electric fields. The two-beam coupling gain coefficient in TPD comprising samples remains lower than in the PVK comprising composites. We attribute the smaller PR performance to the lower trap concentration in PS:TPD and so to the reduced space-charge field.