Integrated electro-optic lens scanner in a LiTaO 3 single crystal

Transcription

1 Integrated electro-optic lens scanner in a LiTaO 3 single crystal Kevin T. Gahagan, Venkatraman Gopalan, Jeanne M. Robinson, Quanzi X. Jia, Terence E. Mitchell, Matthew J. Kawas, Tuviah E. Schlesinger, and Daniel D. Stancil We report what we believe to be the first stand-alone integrated electro-optic lens and scanner fabricated on a single crystal of Z-cut LiTaO 3. The independently controlled lens and scanner components consist of lithographically defined domain-inverted regions extending through the thickness of the crystal. A lens power of cm 1 kv 1 and a deflection angle of mrad kv 1 were observed at the output of the device Optical Society of America OCIS codes: , , Introduction The ability to control both the angular position and the spot size of a laser beam with high speed is of interest for many applications such as optical data storage, laser printing, and heads-up display technology. These functions have traditionally been performed with separate elements for focusing and scanning, thus requiring multistep manufacturing processes and sometimes difficult alignment procedures. Integration of these components into a single manufacturing step promises a great reduction in the cost of producing such devices as well as an improvement in reliability. In addition, a recent report of an integrated quasi-phase-matched second-harmonic generator and electro-optic scanner for blue laser light 1 suggests that integration of such a device with an electro-optic lens to couple the second-harmonic waveguide output to the scanner may greatly enhance device performance. Selective inversion to create shaped 180 ferroelectric domains by electric-field poling 2 is a common technique for fabricating nonlinear and electro-optic devices in LiTaO 3. For example, second-harmonic generation gratings, 3,4 electro-optic scanners, 1,5 7 K. T. Gahagan, V. Gopalan, J. M. Robinson, Q. X. Jia, and T. E. Mitchell are with the Los Alamos National Laboratory, Los Alamos, New Mexico M. J. Kawas, T. E. Schlesinger, and D. D. Stancil are with the Department of Electrical and Computer Engineering, Carnegie Mellon University, Pittsburgh, Pennsylvania Received 5 October 1998; revised manuscript received 15 December $ Optical Society of America and electro-optic lenses 8,9 have all been realized with this technique. Here, we report the first, to the best of our knowledge, successful integration of an independently controlled electro-optic lens and scanner on a singlecrystal LiTaO 3 wafer. Although more powerful lens 8,9 and scanner 10 devices have been demonstrated individually, we believe this device is the first to the best of our knowledge to integrate both functions. Design and fabrication of the device by photolithography and room-temperature electric-field poling are reported in Section 2. A discussion of device performance, including focusing and deflection characteristics, is given in Section 3, followed by our conclusions in Section Design A schematic of the integrated device is shown in Fig. 1. The lens component consists of a series of N l 10 cylindrical plano-convex 180 domain-inverted lenses. The scanner component is separated from the lens component by a 1-mm spacing and consists of a series of N s 7 domain-inverted triangles or prisms. When an electric field, E V t, is present across either the lens or the scanner component, the electro-optic effect induces a change in refractive index by an amount n 1 2 n e 3 r 33 V t, (1) where n e is the linear refractive index of the medium, r 33 is the electro-optic coefficient in the vertical polarization direction, V is the applied voltage, and t is the thickness of the wafer. When the electric field, E, is parallel antiparallel to the spontaneous polarization direction, P s, of the ferroelectric domains, the 1186 APPLIED OPTICS Vol. 38, No. 7 1 March 1999

2 Fig. 1. Diagram of the ferroelectric domains of the integrated electro-optic lens scanner. The side view top depicts the path of the laser through the device and the position of electrodes used to induce the electro-optic effect. The view from above middle shows the domain boundaries defining the lens stack and the scanner that extend through the crystal. Six devices were fabricated on a single 17 mm 10 mm 225 m crystal. Dimensions of the domains in a single stack are also shown bottom. index n e f n e n. Thus for the lens to focus, E should be antiparallel to P s inside the lens-shaped domains. However, the scanner functions with E of either polarity, with one polarity deflecting the beam to the right, for example, while the the other polarity deflects the beam to the left. Thus the beam may be focused by the lens or deflected by the scanner by an amount proportional to the applied field. A computer simulation Fig. 2 was performed with a one-dimensional fast Fourier-transform beampropagation method. 11,12 An electric field through both the lens and the scanner of 8 kv mm and the domain dimensions measured from images of the actual device Fig. 3 were used. This field is greater than the typical fields used in the experiment 4 kv mm and was chosen simply to illustrate device Fig. 2. Simulated propagation of a nm He Ne laser beam through the lens scanner. With an 8 kv mm field applied to the crystal, the beam is focused and deflected to a point approximately 17 mm from the output face of the device at an angle of 25 mrad. Measured device dimensions were used in the fast Fouriertransform beam-propagation method simulation. Fig. 3. Images of ferroelectric domains of portion a of the lens and b of the scanner obtained with polarized light microscopy. performance. Under these conditions, the 600- mdiameter incident beam is focused at a distance of approximately 17 mm from the output of the device to a spot diameter of approximately 60 m and is deflected by an angle of approximately 25 mrad Fig. 2. The device was fabricated in a m thick, Z-cut single crystal of LiTaO 3 wafer. Domain microengineering to form the lens and the scanner was performed by chemical patterning of the surface followed by electric-field application, as described in detail by Baron et al. 2 Briefly, a photoresist pattern resembling the device shown in Fig. 1 was defined on the Z face of the crystal surface with photolithography. A uniform Ta metal was then deposited on this pattern with DC sputtering, followed by liftoff of the metal by dissolving the underlying photoresist pattern in acetone. This results in a Ta-film pattern forming on the crystal surface and resembling the device in Fig. 1, with the Ta surrounding the lens and the prism features and no Ta inside these features. The crystal is next immersed in pyrophosphoric acid and heated to 235 C for 1.5 h to cause selective diffusion of hydrogen near the surface 0.1 m in the lens and the prism regions where no Ta pattern is present. This process is called proton exchange. An electric field of approximately 21 kv mm is then applied across the crystal between the Ta-film pat- 1 March 1999 Vol. 38, No. 7 APPLIED OPTICS 1187

3 tern, used as the negative electrode, and an ionic liquid saturated KNO 3 solution in water, used as a uniform positive electrode, to invert the ferroelectric domain at room temperature. We then stripped the Ta pattern by etching in a HF HNO 3 ::1:2 ratio for a few seconds. The domain pattern is observed with polarized light microscopy. 13 Figure 3 shows the domain walls defining the lenses Fig. 3 a and the prisms Fig. 3 b as seen between cross polarizers. The edges of the Ta-film pattern used as a negative electrode for domain reversal have fringing fields at the edges, outlining the shapes of the lenses and the scanner. These fringe fields induce a slight domain reversal inside the lens and the scanner, causing irregular edges as well as a decrease in the overall size of the patterned regions, as seen in Fig. 3. The dimensions of the Ta pattern were R 575 m, W l 1150 m, and d l 50 m for the lens component and were L s 1mmandW s 775 m for the scanner component with no separation between then prisms. However, the dimensions of the domains measured from these images are reduced to R 540 m, W l 1080 m, and d l 84 m for the lens component and to L s 940 m, W s 720 m, and a separation distance of d s 40 m between the prisms. Thus the domains are reduced in size by approximately m on all sides. However, without the proton-exchange step, which is found to suppress the effect of fringing fields to some extent and to provide smoother domain boundaries, the reduction would be even greater. As a final step, the surfaces of the poled crystal are coated with an electrode layer over each component Fig. 1. An uncoated border of 1-mm thickness is maintained around each electrode to inhibit discharge between neighboring electrodes. Contact with the electrodes is established with copper tape, and the device is then mounted between two insulating rubber layers to further inhibit discharge or corona formation during operation. Voltages of as much as 2 kv across the lens and as much as 1.2 kv across the scanner, corresponding to the limits of our power supplies, were applied independently without inducing breakdown. 3. Testing Independent tests of the scanner and the lens were performed to determine the scan angle and the lens power as a function of applied voltage with the apparatus depicted in Fig. 4. Collimated, vertically polarized light from a nm He Ne laser is focused in the vertical y direction with a cylindrical lens such that the beam passes cleanly through the crystal without scattering from the top or the bottom surfaces. A beam-profiling CCD camera Coherent LaserCam placed at the output of the device is used to record the intensity profile of the beam. Applying a voltage across the lens component of the device focuses the beam in the horizontal x direction. For example, the intensity profile observed at a distance of D 132 mm from the output face of the device for three different values of the applied voltage Fig. 4. Diagram of device-testing apparatus. A vertically polarized beam from the He Ne laser is first collimated with a circular lens not shown and then focused in the vertical direction by a cylindrical lens f 100 mm through the electro-optic device. The electric fields are applied independently to the lens and the scanner components to focus and deflect the beam observed with a CCD camera placed some distance from the output. V l equal to 0, 100, and 260 V left to right is shown in Fig. 5. The width of the beam along the x axis defined here as the horizontal distance at which the intensity drops to 1 e 2 of the peak value decreases from w x 549 m atv l 0tow x 226 m atv l 260 V; this is a ratio of 2.42:1. This decrease in spot size more than doubles the number of resolvable spots in this plane. To quantify the focusing properties of the electrooptic lens component, we placed the CCD array at a range of distances from 10 to 30 cm from the output of the device. For each distance, we recorded the voltage, V l, at which the horizontal x width of the intensity profile was minimized. Treating the device as a thin lens and assuming the beam is collimated in the x direction at the entrance to the device yields, by the distance to the CCD array, the effective focal length, f, of the electro-optic lens stack at the recorded voltage. The lens power, 1 f, is plotted in Fig. 6 as a function of the applied voltage, V l. The lens power was found to increase linearly with the applied voltage with a dependence of cm 1 kv 1. The error bars indicate the uncertainty of 20 V in the determination of the voltage corresponding to the minimum spot diameter. One can derive a theoretical estimate of the power dependence solid line in Fig. 6 by modeling the array as a series of thin lenses, each contributing a power of l 2 n R. 8 The power dependence of the Fig. 5. CCD images of the beam profile at a distance of 132 mm from the output face of the device for three different values of the lens voltage, V l equal to 0, 100, and 260 V. The corresponding 1 e 2 diameter of the beam along the x axis, through the centroid of the intensity distribution, is shown below the images APPLIED OPTICS Vol. 38, No. 7 1 March 1999

4 Fig. 6. Measured lens power as a function of applied voltage, V l. The solid line indicates the theoretical dependence of the lens power with a thin-lens approximation, 2 nn l R. Fig. 7. Measured deflection angle as a function of applied voltage, V s. The solid line indicates the theoretical dependence of the deflection angle with d 2 nl W. Fig. 8. Multiple-exposure image of the beam profile at D 143 mm over a range of scanning voltages from V s equal to 750 to 750 V. A lens voltage of V l 300 V is applied to maximize the number of resolvable spots in this plane. A line plot below the image shows the intensity profile along the horizontal dotted line in the image. electro-optic lens stack is then l N l cm 1 kv 1, in agreement with the measured dependence. To determine the deflection angle of the scanner, we measured the change in the position of the centroid of the intensity distribution in the horizontal plane of the camera located a distance of D 194 mm from the device. For a centroid displacement of x, the deflection angle is given by tan 1 x D. The deflection angle measured as a function of applied voltage is plotted in Fig. 7. The measured voltage dependence of the deflection angle is mrad kv. A theoretical estimate solid line in Fig. 7 for the deflection angle is given by 2 nl W mrad kv, where L N s L s is the total length of the scanner prisms, 14 and is in good agreement with the experimental results. The error bars reflect the relative uncertainty in the measurement of the deflection distance and the distance from the device. The scanner component may be operated separately or in concert with the lens. To illustrate the latter regime, a multiple-exposure image of the beam profile measured at a distance of 143 mm from the output of the device is shown in Fig. 8. The voltage for each exposure was varied incrementally from 750Vto 750 V such that each profile was separated by approximately one spot diameter from the next with a maximum angular deflection of approximately 9.6 mrad. In addition, the lens voltage was set at V l 300 V to minimize the spot diameter at this distance. With V l 0, the spot diameter at this distance is nearly 2.5 times larger than the minimum, and thus fewer spots may be resolved. The spreading of the profiles in the vertical direction is a consequence of cylindrical lens focusing before entering the device and may easily be compensated by placement of a second cylindrical lens after the device to recollimate the beam in the vertical direction. We have not done so here since our primary concern is with horizontal focusing and deflection of the beam. 4. Conclusions We have successfully demonstrated the first, to the best of our knowledge, stand-alone integrated electrooptic lens and scanner fabricated in a single-crystal LiTaO 3 wafer. Independent control of the lens and the scanner components permits optimization of the device so that a maximum number of resolvable spots at an arbitrary distance from the device, an important characteristic for printing and optical read write applications, can be achieved. We report a measured deflection angle of mrad kv 1 and a lens power of cm 1 kv 1 for a device patterned in a 225- m-thick crystal. These values are typical of those reported previously for individual lens and scanner devices in LiTaO 3. More important, we have demonstrated that both functions may be integrated in a single-crystal wafer and that the lens and the scanner may be independently controlled. We thank Yi Chiu for providing us the fast Fouriertransform beam-propagation method simulation code. References 1. V. Gopalan, M. J. Kawas, M. C. Gupta, T. E. Schlesinger, and D. D. Stancil, Integrated quasi-phase-matched 2nd-harmonic 1 March 1999 Vol. 38, No. 7 APPLIED OPTICS 1189

5 generator and electro-optic scanner on LiTaO 3 single-crystals, IEEE Photonics Technol. Lett. 8, C. Baron, H. Cheng, and M. C. Gupta, Domain inversion in LiTaO 3 and LiNbO 3 by electric-field application on chemically patterned crystals, Appl. Phys. Lett. 68, K. Mizuuchi and K. Yamamoto, Highly efficient quasi-phasematched 2nd-harmonic generation using a 1st-order periodically domain-inverted LiTaO 3 wave-guide, Appl. Phys. Lett. 60, C. Baron, H. Cheng, and M. C. Gupta, Periodic domain inversion in ion exchanged LiTaO 3 by electric field Application, in Nonlinear Frequency Generation and Conversion, M. C. Gupta, W. J. Kozlovsky, and D. C. MacPherson, eds., Proc. SPIE 2700, Q. B. Chen, Y. Chiu, D. N. Lambeth, T. E. Schlesinger, and D. D. Stancil, Guided-wave electro-optic beam deflector using domain reversal in LiTaO 3, J. Lightwave Technol. 12, J. Li, H. C. Cheng, M. J. Kawas, D. N. Lambeth, T. E. Schlesinger, and D. D. Stancil, Electro-optic wafer beam deflector in LiTaO 3, IEEE Photonics Technol. Lett. 8, N. Ramanujam and J. J. Burke, Optimizing KTP and LiTaO 3 channel wave-guides for quasi-phase-matched 2nd-harmonic generation with high conversion efficiency, IEEE J. Quantum Electron. 33, M. J. Kawas, T. E. Schlesinger, D. D. Stancil, and V. Gopalan, Electro-optic lens stacks on LiTaO 3 by domain inversion, J. Lightwave Technol. 15, M. Yamada, M. Saitoh, and H. Ooki, Electric-field-induced cylindrical lens; switching and deflection devices composed of the inverted domains in LiNbO 3 crystals, Appl. Phys. Lett. 69, V. Gopalan, T. E. Mitchell, Q. X. Jia, J. M. Robinson, M. J. Kawas, T. E. Schlesinger, and D. D. Stancil, Ferroelectrics as a versatile solid-state platform for integrated-optics, Integr. Ferroelectr. 22, J. A. Fleck, J. R. Morris, and M. D. Feit, Time-dependent propagation of high-energy laser-beams through atmosphere, Appl. Phys. 10, M. D. Feit and J. A. Fleck, Jr., Light propagation in gradedindex optical fibers, Appl. Opt. 17, V. Gopalan and M. C. Gupta, Origin of internal field and visualization of 180-degree domains in congruent LiTaO 3 crystals, J. Appl. Phys. 80, J. F. Lotspeich, Electro-optic light-beam deflection, IEEE Spectr. 5, APPLIED OPTICS Vol. 38, No. 7 1 March 1999