Open Access
Add to BibSonomyAbstract
Electrical and thermal modulation of porous silicon microcavities is demonstrated based on a change in the refractive index of liquid crystals infiltrated in the porous silicon matrix. Positive and negative anisotropy liquid crystals are investigated, leading to controllable tuning to both longer and shorter wavelengths. Extinction ratios greater than 10 dB have been demonstrated. Larger attenuation can be achieved by increasing the Q-factor of the microcavities.
©2005 Optical Society of America
1. Introduction
The need for alternative interconnect technologies that can efficiently deal with large bandwidths of information has led to the investigation of photonic devices suitable for integration into an optical interconnect platform. Photonic crystals are particularly attractive due to their compactness and ability to control light propagation. Moreover, when fabricated in silicon, photonic crystal devices can be directly integrated into current microelectronics technology. Active modulation of the optical properties of silicon-based photonic crystals provides the foundation for a variety of tunable components.
Photonic crystals have been a significant research topic since the early 1990s [1,2]. More recently, there has been an increasing emphasis on tuning the optical properties of these photonic bandgap structures. Liquid crystals were initially proposed to enable tuning in 1999 [3]. Thermal tuning of porous silicon [4] and GaAs-based [5] two-dimensional photonic bandgap structures infiltrated with liquid crystals has been demonstrated. Electrical tuning of liquid crystals on top of an InGaAsP two-dimensional photonic crystal laser [6] and in the central cavity layer of SiO2/TiO2 [7] and SiO2/Si [8] one-dimensional photonic bandgap structures has also been demonstrated. Additionally, thermal [9,10] and electrical [11,12] tuning of liquid crystals in photonic crystal fiber and inverse opal structures has been shown. Moreover, polarization-dependent tuning of glancing-angle deposition films infiltrated with liquid crystals was also investigated [13].
We present electrical and thermal modulation of liquid crystals in a one-dimensional porous silicon photonic bandgap microcavity. Both the electrical and thermal tuning of the optical properties is achieved by rotating the birefringent liquid crystals that are infiltrated throughout the microcavity. The advantages of these structures include ease of fabrication, flexibility in design wavelength, precise control over the material morphology, and compatibility with silicon microelectronics technology. Issues related to surface anchoring effects and liquid crystals in confined geometries require further attention to improve the performance of the devices beyond that which is demonstrated in this paper.
2. Porous silicon photonic bandgap microcavities
The material basis for the tunable microcavities is porous silicon. Porous silicon is a unique material in that it retains the advantages of silicon technology but adds versatility in the control of optical properties.
2.1 Fabrication
Porous silicon microcavities are formed by electrochemical etching of silicon wafers using a hydrofluoric acid-based electrolyte. The choice of wafer resistivity and electrolyte composition influences the pore size and morphology. Adjusting the applied current density changes the porosity, or percentage of void space in the material. No prepatterning of the surface is done so the pores are randomly distributed on the surface. Mesoporous silicon, with an average pore diameter of 20 nm, is formed on (100) p-type (0.01Ω-cm) silicon using a solution of 15% hydrofluoric acid in ethanol [14]. Macroporous silicon, with an average pore diameter of 150 nm, is formed on (100) n-type (0.01Ω-cm) silicon using an electrolyte composed of H2O:HF (17:1) and a few drops of surfactant [15].
Since the pore sizes are much smaller than the incident near-infrared wavelengths used in this study, effective media approximations are employed to relate the porosity and refractive index. Therefore, the refractive index profile of porous silicon structures can be carefully tailored by simply choosing the proper current density profile. Porous silicon microcavities are formed by first etching a top Bragg mirror, followed by a defect layer, and then a bottom Bragg mirror [14]. Figure 1 shows a schematic of the microcavity design and the respective morphologies of the mesoporous and macroporous silicon microcavities. Distinct layers of different porosity are formed because the etching preferentially proceeds at the pore tips where the field is concentrated. Layers previously formed are unaffected during subsequent etching since they are depleted of the necessary charge carriers. The Bragg mirrors are composed of quarter-wavelength optical thickness layers of alternating refractive index and the defect layer is half-wavelength optical thickness. For the mesoporous silicon microcavities, the layers have porosities of 50% and 75%, corresponding to refractive indices of 2.15 and 1.43 in the near-infrared, respectively. For the macroporous silicon microcavities, the layers have porosities of 50% and 80%, corresponding to refractive indices of 2.15 and 1.32 in the near-infrared, respectively. In both cases, the defect layer has the same porosity as the higher porosity layers. The resonance wavelength of the microcavities can be adjusted from the visible to the infrared by simply modifying the applied current density and duration of the etching.
Fig. 1. Schematic of porous silicon microcavity (a) and SEM images showing the morphology of (b) mesoporous silicon layers and (c) macroporous silicon layers. In the schematic, the yellow layers are low porosity (high refractive index) and the red layers are high porosity (low refractive index). In the SEM images, the darker regions represent the void space and the bright area is the silicon matrix. Therefore, the brighter layers have lower porosity than the darker ones. The pore openings of the macropores are much larger than those of the mesopores.
After each porous silicon microcavity is formed, it is oxidized at 900°C for approximately 10 minutes inside a tube furnace. During oxidation, a fraction of the silicon pore wall is converted into silicon dioxide, which possesses a lower refractive index than silicon. Because the refractive index of the porous silicon layers decreases due to oxidation, the resonance wavelength blue shifts. The Q-factor is not significantly degraded for the oxidation conditions used in this work. After oxidation, liquid crystals are introduced into the porous silicon matrix under vacuum in order to ensure uniform infiltration throughout the microcavity. Prior to infiltration, the liquid crystals are heated above their phase transition temperature to reduce their viscosity and further facilitate infiltration. The presence of the liquid crystals in the pores increases the refractive index of the porous silicon layers and red shifts the resonance wavelength. The Q-factor decreases after liquid crystal infiltration due to the resulting decrease of the refractive index contrast between high and low porosity microcavity layers. With liquid crystals aligned axially in the pores (i.e., parallel to the pore walls), the mesoporous silicon microcavities have layers of refractive index 2.27 and 1.65 and the macroporous silicon microcavities have layers of refractive index 2.27 and 1.55 in the near-infrared. Increasing the number of periods in the Bragg mirrors of the microcavity or increasing the initial refractive index contrast of the microcavity layers will lead to increased Q-factors after liquid crystal infiltration.
2.2 Characterization
Reflectance spectra of the porous silicon microcavities are taken using a Perkin-Elmer Lambda 900 spectrophotometer and an Ando AQ6317 optical spectrum analyzer with a fiber coupled Xenon arc lamp. Typical spectra are shown in Fig. 2. For the mesoporous silicon microcavities, Q-factors exceeding 1000 without liquid crystals and a few hundred after liquid crystal infiltration are regularly obtained. For the macroporous silicon microcavities, further optimization of the etching conditions is required to increase the measured Q-factors above 100 with liquid crystals.
Fig. 2. Reflectance spectra, before liquid crystal infiltration, of (a) mesoporous silicon microcavity measured by a spectrophotometer and (b) macroporous silicon microcavity measured by an optical spectrum analyzer with a fiber coupled Xenon arc lamp. Tuning of the resonance wavelength after liquid crystal infiltration enables optical modulation.
3. Electrical modulation
3.1 Device details
In order to enable electrical modulation of the optical properties of the porous silicon microcavities, ITO-coated glass is attached to the top surface of the porous silicon. After the liquid crystals are infiltrated into the pores under vacuum, the ITO-coated glass is tightly pressed against the porous silicon using a clamp, and then epoxy is applied along the outside to seal the device. Upon visual inspection, fringes are apparent, indicating that a liquid crystal layer of a few microns remains on the surface of the porous silicon. Wires are attached, using a conductive epoxy, to both the ITO coating and the backside of the crystalline silicon wafer (see Fig. 3(a) inset). An a.c. electric field at 1 kHz is applied in order to prevent random current fluctuations and dynamic scattering that arise in the liquid crystals when d.c. electric fields are used [16]. The temperature of the microcavity is monitored during the experiment and no heating occurs. It is critical that the temperature does not increase when the voltage is applied due to the thermally induced drift of the resonance wavelength that occurs when both the silicon matrix [17] and the liquid crystals are heated.
3.2 Experimental results and discussion
Figure 3 shows the shift in the resonance wavelength of the mesoporous and macroporous silicon microcavities as a function of applied voltage with unpolarized incident light. The resonance shifts due to a physical rotation of the birefringent nematic liquid crystals in the porous silicon matrix and is reversible when the voltage is turned off. For positive anisotropy E7 liquid crystals, with Δε~13.8 in the kHz frequency regime (Fig. 3(a)), the liquid crystal molecules align with the applied field, leading to a refractive index decrease and resonance shift to shorter wavelengths. For negative anisotropy ZLI-4788 liquid crystals, with Δε~-5.7 in the kHz frequency regime (Fig. 3(b)), the long axis of the liquid crystal molecules aligns perpendicular to the field lines, leading to a refractive index increase and a resonance shift to longer wavelengths. No change in temperature is detected when the electric field is applied. This reported observation of electrical tuning within a silicon matrix is significant because it was previously thought that electrical tuning in a doped silicon matrix would not be possible due to screening of the electric field by a high density of free carriers [4]. Moreover, our reported electrical tuning to shorter or longer wavelengths, depending on the choice of liquid crystal, is especially important because it adds flexibility to the design and enhances the operation of tunable filter devices. The initial alignment of the liquid crystals and the placement of electrodes for an integrated device determines whether positive or negative anisotropy liquid crystals will lead to a larger birefringence and, hence, larger extinction ratio.
There are likely several reasons why the magnitude of the resonance shift is larger for the macropores. First, the overall porosity is slightly higher in the macroporous silicon microcavity, which allows more liquid crystals to infiltrate the structure. Additionally, the geometry is more constricted in the mesoporous silicon. Therefore, a larger fraction of the liquid crystals are influenced by surface anchoring [18]. Within the first one or two monolayers, it is very difficult to dislodge the liquid crystal molecules. As the distance between the liquid crystal molecules and the surface increases, the anchoring energy decreases [19]. In the macropores, the liquid crystals in the center of the pores are more likely to be randomly oriented and, consequently, able to rotate more freely in the applied field.
Fig. 3. (a) Resonance wavelength red shift as a function of applied voltage for mesoporous and macroporous silicon microcavities with positive anisotropy E7 liquid crystals. The liquid crystals rotate more freely in the macroporous silicon, which leads to the larger wavelength shift. Electrical contact is made to the crystalline silicon and ITO-coated glass on top of the microcavity (inset). (b) Resonance wavelength red shift as a function of applied voltage for mesoporous silicon microcavities with negative anisotropy ZLI-4788 liquid crystals.
4. Thermal modulation
4.1 Device details
Mesoporous and macroporous silicon microcavities are thermally modulated by mounting the structures on a simple heating and temperature sensing apparatus. Using thermally conductive epoxy, each sample is attached to an aluminum plate. A resistor and thermistor are attached on the back of the aluminum plate. The microcavities are heated by passing a current through the resistors and the temperature is monitored by measuring the resistance of the thermistor. A control experiment was performed to ensure that the temperature measured on the aluminum plate is the same as that measured on the front surface of the porous silicon microcavities.
4.2 Experimental results and discussion
Fig. 4. Thermal tuning of mesoporous and macroporous silicon microcavities with liquid crystals. The resonance shift takes place at different temperatures depending on the phase transition temperature of the infiltrated liquid crystal, as shown in (a)-(c). The refractive index increases when liquid crystals change from the ordered nematic phase to disordered isotropic phase (inset (a)). (d) For a resonance with a measured Q-factor of 400, the resonance shift resulting from the E7 phase transition corresponds to a 14 dB extinction ratio.
Based on the phase transition from the nematic (ordered) to isotropic (disordered) state, the refractive index of liquid crystals can be changed with temperature. While the resonance wavelength of porous silicon microcavities without liquid crystals can be shifted by direct heating to enable optical switching, the bandwidth of the tuning, even over large temperature ranges, is limited to a few nanometers [17]. More efficient thermal tuning (a larger bandwidth over a smaller temperature range) is achieved by infiltrating liquid crystals inside the porous silicon microcavities. Figure 4 summarizes the resonance wavelength shift achieved by mesoporous and macroporous silicon microcavities using liquid crystals with different phase transition temperatures. It should be noted that the resonance wavelength of empty porous silicon microcavities oxidized at 900°C exhibit a blue shift with heating while a red shift with heating is observed when liquid crystals are infiltrated into the microcavities. As shown in Fig. 4, 5CB has a phase transition temperature of 35°C, E7 has a phase transition temperature of 58°C, and ZLI-4788 has a phase transition temperature of 81°C. All three liquid crystals can be obtained from Merck. The birefringence of 5CB and E7 at optical frequencies is approximately Δn=0.2 and the birefringence of ZLI-4788 at optical frequencies is approximately Δn=0.15. Flexibility in phase transition temperature allows devices to operate at a variety of temperatures, depending on the ambient environment.
5. Current performance and future potential of silicon-based PBG modulators
For optical switching applications, the key parameter is the extinction ratio between the on and off states of a particular channel (see Fig. 5 inset). Since the tuning range of the microcavity resonance is limited by the liquid crystal birefringence, the only way to increase the extinction ratio is to increase the Q-factor, which is the ratio between the resonance wavelength and the full width at half maximum (FWHM) of the resonance. Figure 5 illustrates the relationship between the Q-factor of a microcavity and the attenuation achievable for a given change of the infiltrated liquid crystal refractive index. Porous silicon microcavities with Q-factors greater than 7,000 have been fabricated [20]. After liquid crystal infiltration, such microcavities would have Q-factors of approximately 600.
The resonance wavelength shifts accomplished by the electrical tuning of mesoporous and macroporous silicon microcavities, shown in Fig. 3(a), correspond to refractive index changes of 0.005 and 0.01, respectively. Assuming a Q-factor of 600, electrical tuning would lead to extinction ratios of approximately 7 dB and 12 dB. The achieved birefringence is significantly smaller than the full E7 birefringence due to the initial liquid crystal alignment and surface anchoring effects, which inhibit complete 90° liquid crystal rotation in the applied electric field. Nevertheless, the largest extinction ratio is achievable for a birefringence of Δn=0.1, not the maximum E7 birefringence of Δn=0.2, since signal contrast decreases as the resonance begins to shift beyond the stopband. Furthermore, since the electrical tuning effect shown in Fig. 3 is not saturated, which would be the case if the liquid crystals achieved their maximum rotation in the applied field, larger birefringence and hence larger extinction ratios are expected with increasing voltage. Reducing the surface anchoring energy by adjusting the surface chemistry of the porous silicon and eliminating the liquid crystal layer between the ITO-coated glass and the porous silicon may allow for lower voltage requirements in the future.
Fig. 5. The extinction ratio is calculated as the change in transmission at the resonance wavelength (inset). As the Q-factor of the resonance increases, the achievable extinction ratio increases, for a given refractive index change. The magnitude of the refractive index change determines the magnitude of the resonance wavelength shift. When the refractive index change becomes too large, the resonance wavelength begins to shift beyond the stopband, which reduces the extinction ratio. For a microcavity with a Q-factor of 600, the maximum attenuation is achieved for Δn=0.1.
As shown in Fig. 4(d), thermal tuning of the liquid crystal-filled mesoporous silicon microcavity causes the resonance to shift by a distance greater than the FWHM, resulting in a large extinction ratio. While the measurement is performed in reflection, assuming negligible absorption for silicon near 1.5 µm, an extinction ratio of 14 dB is obtained for this microcavity with a Q-factor of 400. Furthermore, with a Q-factor of 600, extinction ratios of 17 dB and 24 dB for mesoporous and macroporous silicon microcavities, respectively, could be obtained with the measured changes in refractive index of 0.02 and 0.05 shown in Fig 4. While the response time of liquid crystal-based devices is limited to milliseconds or, at best, tens of microseconds, the achievable extinction ratios could enable applications in reconfigurable networks such as optical network protection and restoration [21].
6. Conclusion
Electrical and thermal modulation of liquid crystal-filled porous silicon one-dimensional photonic bandgap microcavities is demonstrated. Active tuning to shorter or longer wavelengths can be achieved based on the choice of liquid crystal. The extinction ratio of the devices is linked to both the achievable liquid crystal birefringence and Q-factor of the microcavity. Extinction ratios of 14 dB were realized. The microcavities can serve as building blocks for tunable filters in microelectronics, microphotonics, and display technologies.
Acknowledgments
The authors gratefully acknowledge Dr. Stephen Jacobs and Ken Marshall of the Laboratory for Laser Energetics at the University of Rochester for their liquid crystal expertise. This work is supported in part by the Air Force Office of Scientific Research and Intel Corporation. The Electron Microscopy Facility at the University of Rochester is supported by the National Science Foundation.
References and links
1. E. Yablonovitch, “Inhibited spontaneous emission in solid-state physics and electronics,” Phys. Rev. Lett. 58, 2059–2062 (1987). [CrossRef] [PubMed]
2. S. John, “Strong localization of photons in certain disordered dielectric superlattices,” Phys. Rev. Lett. 58, 2486–2489 (1987). [CrossRef] [PubMed]
3. K. Busch and S. John, “Liquid-crystal photonic-band-gap materials: the tunable electromagnetic spectrum,” Phys. Rev. Lett. 83, 967–970 (1999). [CrossRef]
4. S. W. Leonard, J. P. Mondia, H. M. van Driel, O. Toader, S. John, K. Busch, A. Birner, U. Gösele, and V. Lehmann, “Tunable two-dimensional photonic crystals using liquid-crystal infiltration,” Phys. Rev. B 61, R2389–R2392 (2000). [CrossRef]
5. Ch. Schuller, F. Klopf, J. P. Reithmaier, M. Kamp, and A. Forchel, “Tunable photonic crystals fabricated in III–V semiconductor slab waveguides using infiltrated liquid crystals,” Appl. Phys. Lett. 82, 2767–2769 (2003). [CrossRef]
6. B. Maune, M. Lončar, J. Witzens, M. Hochberg, T. B. Jones, D. Psaltis, A. Scherer, and Y. Qiu, “Liquid-crystal electric tuning of a photonic crystal laser,” Appl. Phys. Lett. 85, 360–362 (2004). [CrossRef]
7. R. Ozaki, T. Matsui, M. Ozaki, and K. Yoshino, “Electrically color-tunable defect mode lasing in one-dimensional photonic-band-gap system containing liquid crystals,” Appl. Phys. Lett. 82, 3593–3595 (2003). [CrossRef]
8. G. Pucker, A. Mezzetti, M. Crivellari, P. Bellutti, A. Lui, N. Daldosso, and L. Pavesi, “Silicon-based near-infrared tunable filters filled with positive or negative dielectric anisotropic liquid crystals,” J. Appl. Phys. 95, 767–769 (2004). [CrossRef]
9. T. T. Larsen, A. Bjarklev, D. S. Hermann, and J. Broeng, “Optical devices based on liquid crystal photonic bandgap fibres,” Opt. Express 11, 2589–2596 (2003), http://www.opticsexpress.org/abstract.cfm?URI=OPEX-11-20-2589. [CrossRef] [PubMed]
10. K. Yoshino, Y. Shimoda, Y. Kawagishi, K. Nakayama, and M. Ozaki, “Temperature tuning of the stop band in transmission spectra of liquid-crystal infiltrated synthetic opal as tunable photonic crystal,” Appl. Phys. Lett. 75, 932–934 (1999). [CrossRef]
11. F. Du, Y.-Q. Lu, and S.-T. Wu, “Electrically tunable liquid-crystal photonic crystal fiber,” Appl. Phys. Lett. 85, 2181–2183 (2004). [CrossRef]
12. Y. Shimoda, M. Ozaki, and K. Yoshino, “Electric field tuning of a stop band in a reflection spectrum of synthetic opal infiltrated with nematic liquid crystal,” Appl. Phys. Lett. 79, 3627–3629 (2001). [CrossRef]
13. K. Robbie, D. J. Broer, and M. J. Brett, “Chiral nematic order in liquid crystals imposed by an engineered inorganic structure,” Nature 399, 764–766 (1999). [CrossRef]
14. S. M. Weiss, M. Haurylau, and P. M. Fauchet, “Tunable photonic bandgap structures for optical interconnects,” Opt. Mat. 27, 740–744 (2005). [CrossRef]
15. H. Ouyang, M. Christophersen, R. Viard, and P. M. Fauchet, Center for Future Health, University of Rochester, Rochester, N.Y. 14627, are preparing a manuscript to be called “Macroporous silicon microcavities for macromolecule detection.”
16. G. Meier, E. Sackmann, and J. G. Grabmaier, Applications of Liquid Crystals, (Springer Verlag, Berlin, Germany, 1975), pp. 29–30.
17. S. M. Weiss, M. Molinari, and P. M. Fauchet, “Temperature stability for silicon-based photonic band-gap structures,” Appl. Phys. Lett. 83, 1980–1982 (2003). [CrossRef]
18. G. P. Crawford and S. Žumer, eds., Liquid Crystals in Complex Geometries Formed by Polymer and Porous Networks (Taylor & Francis Ltd., London, 1996), pp. 21–52.
19. J. Cognard, Alignment of Nematic Liquid Crystals and Their Mixtures (Gordon and Breach, New York, 1982), p. 59.
20. G. Lérondel, P. Reece, A. Bruyant, and M. Gal, “Strong light confinement in microporous photonic silicon structures,” in Engineered Porosity for Microphotonics and Plasmonics , R. Wehrspohn, F. Garcial-Vidal, M. Notomi, and A. Scherer, eds., Mat. Res. Soc. Proc. 797, W1.7.1–W1.7.6 (2004).
21. R. Ramaswami and K. N. Sivarajan, Optical Networks: A Practical Perspective (Morgan Kaufmann Publishers, San Francisco, 1998), pp. 423–462.
