High-resolution, high-speed photorefractive incoherent-to ... - lmops

3 downloads 55852 Views 889KB Size Report
We demonstrate a photorefractive incoherent-to-coherent optical converter driven by ultraviolet light that ..... B. Ai, D. S. Glassner, R. J. Knize, and J. P. Partanen,.
February 15, 1999 / Vol. 24, No. 4 / OPTICS LETTERS

199

High-resolution, high-speed photorefractive incoherent-to-coherent optical converter ¨ P. Bernasconi, G. Montemezzani, M. Wintermantel, I. Biaggio, and P. Gunter Nonlinear Optics Laboratory, Institute of Quantum Electronics, Swiss Federal Institute of Technology, ETH Honggerberg, ¨ CH-8093 Zurich, ¨ Switzerland Received October 5, 1998 We demonstrate a photorefractive incoherent-to-coherent optical converter driven by ultraviolet light that provides a 35-ms response time and an optical resolution of 124 line pairsymm. The device, implemented in KNbO3 , operates with a modulating intensity of 85 mWycm2 , which corresponds to an optical switching energy per bit of 0.5 pJ. A conversion rate of the order of 90 Gbitsyss cm2 d is achieved. The conversion between the ultraviolet light and the visible laser beam at l ­ 532 nm occurs through anisotropic Bragg diffraction at a modulated interband photorefractive grating. Our device has a better optical resolution and conversion rate than optically addressed solid-state spatial light modulators based on the photorefractive effect and multiple quantum wells, and it is also faster than devices based on liquid crystals.  1999 Optical Society of America OCIS codes: 330.6120, 070.4550, 160.5320, 090.7330.

High resolution and fast response are two fundamental requirements for all-optical image-processing devices. Important results have been achieved with devices employing liquid crystals1 or multiplequantum-well (MQW) structures.2 Optically driven spatial light modulators (SLM’s) based on the photorefractive effect have also been proposed3 – 6; however, they have a slower response and lower optical resolution. We demonstrate an optically addressed SLM that transfers images carried by incoherent light to a coherent beam by use of the interband photorefractive effect.7,8 In a photorefractive optical converter (PICOC) the information transfer between incoherent and coherent light occurs through diffraction at a modulated holographic phase grating. The grating is the key element that determines the time response and inf luences the optical resolution. In contrast with previous work,3 – 6 in which the grating was recorded through the conventional photorefraction,10 we produce the required illumination pattern by using a light wavelength shorter than the fundamental absorption edge of the material. The three main advantages are faster response, thinner and stronger gratings, and shorter operating light wavelengths.7,8 The resonant absorption drastically increases the photoexcitation eff iciency and the photoconductivity. This implies response times typically 2–3 orders of magnitude faster than with conventional photorefraction.7 The large absorption conf ines the photoexcitation processes to a thickness of the order of the light penetration depth, so grating thicknesses of a few tens of micrometers can be achieved.8 The thickness can also be adjusted by means of wavelength tuning inside the absorption band. The interband processes give rise to strong phase gratings with index modulations as high as Dn ­ 1024 . Holographic sensitivities of several square centimeters per joule have been demonstrated.7 Further, the interband gratings are robust under nonresonant illumination, i.e., they are not affected by light with wavelengths longer than the 0146-9592/99/040199-03$15.00/0

absorption edge. So the readout wavelength, lro , at which the image is converted can be chosen to be inside the whole transparency range of the material with no restriction on its intensity. Such freedom is not available in devices operating with the conventional photorefractive effect, in which both lro and intensity have to be set long enough and low enough that the hologram does not deteriorate during the readout process. Short operating wavelengths and small grating thicknesses also contribute to a better optical resolution. In fact, in a grating with thickness d, diffraction p limits the minimal feature p size to dl , so the maximal linear resolution is R ~ 1y dl. For our experimental conditions we calculated the resolution limits5,10 of the PICOC according to the coupled-wave equations in anisotropic materials11 and by taking into account the wavelength dispersion. The curves are shown in Fig. 1. The optical resolution can be also limited by the optical projection system to R ­ Ays2f ld, where A and f are the aperture and the focal length of the system. This limit becomes R ­ f ysdAd when it matches the diffraction-limited resolution.5 The experimental implementation of the interband PICOC was performed in a 47-mm-thick sample of nominally pure, single-domain KNbO3 crystal. This material was chosen because of its excellent photorefractive properties in the interband regime.7,8 As shown in Fig. 2, the photorefractive grating was recorded by two interfering laser beams. Note that the interference light pattern can also be produced by incoherent light spatially modulated by an appropriate mask or grating. The s-polarized laser beams are at lrec ­ 364 nm shn ­ 3.4 eVd, so the photon energy exceeds the material energy gap sEg ­ 3.2 eVd. For this wavelength the absorption is a ­ 540 cm21 . The grating wave vector was oriented along the crystallographic b axis, and the grating period Lg was 0.6 mm. In this geometry the readout beam at lro ­ 532 nm is diffracted anisotropically and perpendicularly to the crystal surface, as is required for the best optical resolution.12 This orientation also offers the largest electro-optic coefficient and thus the largest  1999 Optical Society of America

200

OPTICS LETTERS / Vol. 24, No. 4 / February 15, 1999

Fig. 1. Resolution limit R versus the ratio nylro for different grating thicknesses d in KNbO3 . R is defined at 50% of the modulation transfer function, and n is the refractive index at lro . The lines are calculated assuming Lg ­ 0.6 mm and a diffraction efficiency of h ­ 0.01. s, resolution measured with a 0.83-mm-thick conventional photorefractive grating5 at lro ­ 633 nm; d, resolution measured in this work with a 47-mm-thick interband grating at lro ­ 532 nm. Inset, detail of the converted image showing optical resolution of 124 lpymm. The size of the image on the crystal is 5 mm 3 5 mm.

Fig. 2. Schematic setup of a PICOC. The two interfering recording waves produce a photorefractive grating, which is modulated by the projected image carried by incoherent light. The readout laser beam is diffracted, revealing the contrast-reversed image.

diffraction efficiency. We set the total intensity of the recording beams to Irec ­ 85 mWycm2 to ensure an almost-homogeneous interband photorefractive grating.8 The readout intensity was Iro ­ 5 Wycm2 . Because of the robustness of the interband grating, the wavelength of the incoherent light has to be shorter than the fundamental absorption edge as well. A mercury-arc lamp provided the incoherent light filtered by a bandpass filter (transmission, 300–400 nm), which was first modulated by a resolution chart and then imaged onto the crystal. Finally, the image carried by the output wave was monitored by a CCD camera. The best optical resolution of 124 line pairsymm (lpymm) was achieved with Iinc ­ 70 mWycm2 of incoherent light. The overall contrast was better than 10:1 (Fig. 1), the response time t ­ 35 ms, and the diffraction eff iciency was h ­ 0.4%. For stronger intensities of the incoherent or the recording waves the resolution did not show any im-

provement. We noticed instead an increase of noise owing to light scattering at lro . For lower intensities the optical resolution decreased because of a worse image contrast when Iinc was reduced and because of a less-homogeneous grating8 when Irec was reduced. For instance, with Irec ­ 10 mWycm2 and Iinc ­ 70 mWycm2 , the resolution fell below 96 lpymm. For higher intensities the speed and the diffraction eff iciency increase. As shown in Fig. 3(a), stronger Iinc accelerate the recording, i.e., the time needed to modulate the grating, whereas no inf luence of Iinc on the recovery time needed to restore the original homogeneous grating is noticed. The resolution of 124 lpymm is slightly less than the 148 lpymm expected for our projection system. The next limiting factor would be the Bragg diffraction process, for which 160 lpymm are calculated at 50% of the modulation transfer function. The resolution obtained with our interband PICOC exceeds the performance of all solid-state materials and is comparable with that of the best optically addressed SLM’s using ferroelectric liquid crystals (Table 1). The write –read – erase cycle, tc ­ 70 ms in our device, is almost as fast as for MQW structures and much faster than for SLM’s based on ferroelectric liquid crystals. In our case speed and resolution combine to give, to our knowledge, the highest incoherent-to-coherent conversion rate G ; s2Rd2 ytc ­ 88 Gbitsyss cm2 d demonstrated to date in the cw regime (Table 1). By defining the optical switching energy per bit as Enybit ; sIrec 1 Iinc d 3 tys2Rd2 , we obtain Enybit ­ 0.5 pJ, one of the smallest values measured in optically addressed SLM’s. However, this value does not take into consideration the readout intensity Iro or the output power in the coherent beam carrying the image. An appropriate figure of merit is then given by ppin yppout , i.e., the number of total input photons per pixel necessary to generate one photon per pixel at the output.17 For our experimental conditions this is expressed by ∂1/3 ∑

∏ s2Rd2 hnro 1/3 Iro t µ ∂∏ Irec t Iinc Iro , 3 1 1 s2Rd2 hnrec hninc hnro

ppin yppout ­

µ

3 h ∑

(1)

Fig. 3. (a) Recording time and (b) grating recovery time versus incoherent intensity for different recording intensities Irec : s, 8 mWycm2 ; ¶, 25 mWycm2 ; h, 85 mWycm2 . The lines are guides to the eye.

201

February 15, 1999 / Vol. 24, No. 4 / OPTICS LETTERS

Table 1. Experimental Results of an Illustrative Comparison among Selected Optically Addressed SLM’s Operated in the Diffraction s Ddd or Transmission s T d Modea Modulating Material Pure KNbO3 Fe:KNbO3 BSO thin film GaAsb Bacteriorhodopsin GaAs-AlGaAs MQWb FLCb Liquid crystal (nematic)b Cs atom vapor correlator

R slpymmd tc smsd 124 35 27 17 100 70 150 70 15

0.07 100 10 2 100 0.035 2 2 –

h s%d ­ 0.4 ­ 1.2 ­ 3.8 ­ 70 ­ 10 ­ 1.5 ­ 0.2 – D ­ 7 3 1024 D D T T T D D

GsGbits s21 cm21 d 88 0.005 0.036 0.06 0.04 56 4.5 1 –

Enybit s pJd 0.5 103 170 175 104 0.2 0.02 0.05 –

ppin yppout

Ref.

5 3 105 4 3 108 107 4 3 106 4 3 107 3 3 105 106 – 104

This work 5 13 14 15 2 1 16 17

a

Note that the values reported in the R, tc , and h columns correspond to the best experimental device performances obtained through a single measurement leading to an optimum for G, Enybit, and ppin yppout . The individual parameters can then be further improved by optimization of the device under the specific aspects. The quantities are def ined in the text. b Operated with an external voltage.

where I and hn are the intensities and the photon energies of the different waves. This figure of merit allows us to compare our converter with optical correlators. In fact, a PICOC can also be used as a correlator by direct modulation of the recording waves with the object and the reference images and by placement of the crystal in the Fourier plane. The diffraction of the readout beam reveals the correlation function of the two input images in the object plane. With Eq. (1), we obtain ppin yppout ­ 5 3 105 for our PICOC correlator, a value that compares well even with those of correlators implemented in vapor gases17 (Table 1). We point out that a complete and detailed comparison also requires other parameters that are not considered in Table 1. In the table we present experimental results for an illustrative comparison among selected optically addressed SLM’s. The reported values do not always ref lect the best limiting performances of each parameter, but they correspond to the best global results obtained through a single measurement in which we try to optimize G, Enybit, and ppin yppout . The interband PICOC presented here may be further improved by use of a shorter lrec , which would decrease the response time further and improve the resolution by reducing the grating thickness. However, owing to a smaller holographic thickness, a smaller diffraction eff iciency h is expected. Nevertheless, h could be enhanced if the grating recording were assisted with an external electric field so that eff iciencies in the few-percent range could be measured.7 The resolution would also benefit from a further shift of lro closer to the absorption edge of the material. One should also note that for the SLM based on the interband photorefractive effect no strict resonance requirements on the readout wavelengths lro exist, in contrast with SLM’s based on MQW’s. We have demonstrated a technique based on the interband photorefraction used to implement a solidstate high-resolution incoherent-to-coherent optical converter. Our KNbO3 -based device is driven by nearultraviolet wavelengths in the range 300–400 nm, which modulates visible light. The operation can be

shifted toward longer wavelengths in the visible or even the infrared range by substitution of KNbO3 with another photorefractive material with a smaller bandgap. ¨ We are very grateful to H. Wuest and to J. Hajf ler for excellent sample preparation. References 1. S. Fukushima, T. Kurokawa, and M. Ohno, Appl. Phys. Lett. 58, 787 (1991). 2. S. R. Bowman, W. S. Rabinovich, G. Beadie, S. M. Kirkpatrick, D. S. Katzer, K. Ikossi-Anastasiou, and C. L. Adler, J. Opt. Soc. Am. B 15, 640 (1998). 3. A. A. Kamshilin and M. P. Petrov, Sov. Tech. Phys. Lett. 6, 144 (1980). 4. Y. Shi, D. Psaltis, A. Marrakchi, and A. R. Tanguay, Appl. Opt. 22, 3665 (1983). ¨ 5. P. Amrhein and P. Gunter, J. Opt. Soc. Am. B 7, 2387 (1990). 6. J. Zhang, H. Wang, S. Yoshikado, and T. Aruga, Opt. Lett. 22, 1612 (1996). ¨ 7. G. Montemezzani, P. Rogin, M. Zgonik, and P. Gunter, Phys. Rev. B 49, 2484 (1994). ¨ 8. P. Bernasconi, G. Montemezzani, and P. Gunter, ‘‘Of fBragg angle light dif fraction and structure of dynamic photorefractive gratings,’’ Appl. Phys. B (to be published). ¨ 9. P. Gunter and J. P. Huignard, eds., Photorefractive Materials and Their Applications (Springer-Verlag, Berlin, 1988), Vols. I and II. 10. A. Marrakchi, A. R. Tanguay, J. Yu, and D. Psaltis, Opt. Eng. 24, 124 (1985). 11. G. Montemezzani and M. Zgonik, Phys. Rev. E 35, 1035 (1997). ¨ 12. E. Voit and P. Gunter, Opt. Lett. 12, 679 (1987). 13. Y. Nagao, H. Sakata, and Y. Mimura, Appl. Opt. 31, 3966 (1992). 14. Y. Bitou and T. Minemoto, Appl. Opt. 37, 4347 (1998). 15. Q. Wang-Song, C. Zhang, R. Blumer, R. B. Gross, Z. Chen, and R. R. Birge, Opt. Lett. 18, 1373 (1993). 16. B. S. Lowans, B. Bates, and R. G. H. Greer, Opt. Commun. 109, 29 (1994). 17. B. Ai, D. S. Glassner, R. J. Knize, and J. P. Partanen, Appl. Phys. Lett. 64, 951 (1994).