Open Access
Add to BibSonomyAbstract
A method for suppressing the formation of optical damage in quadratic electrooptic devices operated at short wavelengths is presented. Formation of optical damage is attributed to the generation of a trapped space charge induced by photoionization of impurity ions by the propagating beam. It is shown that in potassium lithium tantalate niobate where the electrooptic effect is quadratic, operating the electrooptic device by a bipolar driving voltage prevents the space charge from accumulating, which inhibits the formation of the optical damage. A 6 hours continuous operation of electrooptic modulator for a 30 W/cm2 at λ = 445 nm input beam is demonstrated.
© 2015 Optical Society of America
1. Introduction
In recent years we have been witnessing a growing need for electrooptical (EO) devices that are operable at short visible wavelengths. These include in particular devices for lab-on-a-chip (LOC) modules [1], i.e. miniscule spectrometers that operate in the visible-IR wavelengths for analytical purposes. Several EO crystals such as lithium niobate (LNB), and lithium tantalate (LT) which exhibit a strong EO effect and are transparent at the relevant wavelengths, are prone to develop optical damage (OD) which renders their applicability at short visible wavelengths to be limited [2].
Optical damage is defined as the inhibition of the EO effect, and the random scattering of a light beam as it propagates in the medium due to spatially distributed changes in the refractive index induced by the photorefractive effect [3]. EO crystals such as LNB and LT always contain traces of non-stoichiometric impurities (e.g. Fe ions in LNB) which make them photorefractive. Photons at sufficiently short wavelengths that are absorbed in the crystals photo-ionize the impurities and produce free charge carriers (e.g. in LNB with traces of Fe ions these are electrons in the conduction band) [4]. The latter are transported by the bulk photovoltaic effect [5], diffusion [6], and drift under the applied electric field [7], and eventually are retrapped. This creates a trapped space charge that is spatially correlated with the intensity of the propagating beam [8]. The electric field of this space charge acts to screen the applied electric field, and in addition, through the electrooptic effect, superimposes spatially distributed changes in the refractive index [9]. The former inhibits the EO changes induced by the applied field, whereas the latter causes the propagating light beam to scatter. This renders the spectral range at which devices with EO medium that manifest this phenomenon to be limited to long wavelengths that do not photo-excite charge carriers.
Several attempts to suppress optical damage in EO crystals were made over the last years [10]. The most common method is Mg doping [11]. However, this method is costly and complicates domain engineering [12]. Other methods are based on purification of the material from photo-active electrons using thermo-electrical cleaning [13], or optical cleaning [14]. The potential of these methods has been demonstrated experimentally [15]. However, they require a long time and the optimization of the cleaning process parameters is sample specific [12], rendering their applicability to be limited. Additionally, these methods are of relevance for noncentrosymmetric optical materials only [15].
We henceforth demonstrate a technique for inhibiting the formation of OD in the centrosymmetric EO crystal potassium lithium tantalate niobate (KLTN) operated at the paraelectric phase. KLTN is an oxygen perovskite crystal with a composition given by K1-xLixTa1-yNbyO3. KLTN undergoes a ferroelectric phase transition switching from a cubic structure to a tetragonal structure at a temperature which depends on x and y (Tc). This enables to set Tc by selecting the composition of the crystals [16, 17]. At the paraelectric phase KLTN manifests a strong quadratic EO effect given by [18]:
where Δn is the change in the birefringence, no≈2.2 is the refractive index at the paraelectric phase, geff is the effective quadratic electrooptic coefficient, and P is the dc (or low frequency) polarization induced by the applied electric field Eo. The induced polarization at the paraelectric phase (for T>Tc + 6°C where the “mean field approximation” is in effect [19]) is given by P = εo(εr-1)Eo≈εEo, where εo is the electric permeability, εr is the relative dielectric constant, ε is the dielectric constant, and where it is assumed that εr>>1. For a light beam polarized in parallel to the applied field geff=g11=0.16 m4/C2 [20]; Hence, for εr = 2·104, and Eo = 5 kV/cm the induced change in the refractive index is Δn≈7·10−3.In general KLTN is transparent in the spectral range of 0.4 - 6 µm. However, as pointed out above, exploiting this strong EO effect is limited to long wavelengths due to the formation of OD as explained above. However, the fact that in KLTN at the paraelectric phase the EO effect is quadratic opens the way to inhibit the formation of the space charge which causes the OD.
As pointed out above the space charge is formed by retrapping of the charge carriers that are photo-excited by the illumination, and are redistributed in the crystal in spatial correlation with the illumination. In KLTN the dominant transport mechanism is the drift under the applied electric field. Hence, applying an alternating bipolar driving voltage does not enable the space charge to accumulate, and consequently the formation of OD is avoided. At the same time, the electrically induced changes in the refractive index are not affected by the bipolarity of the applied field due to the fact that the EO effect is quadratic.
This principle is henceforth put to the test in a KLTN based EO modulator. It should be noted that the method of applying an alternating voltage has been used in the field of optical modulators with a ferroelectric liquid crystal medium for avoiding fatigue effects which are characteristic to these crystals [21]. However, hitherto, this method has not been employed for the inhibition of space charge induced optical damage in EO modulators.
2. Methodology of the measurement
The measurements were done in an electrooptical modulator constructed in a “crossed polarizers” configuration [18]. The experimental set-up is illustrated in Fig. 1. The system included two input channels. The primary channel was fit with two alternative lasers at short wavelengths that cause OD: (i) A 1 W laser operating at λ = 445 nm; and (ii) a 50 mW laser operating at λ = 488 nm. The secondary channel was fit with a cw DFB laser operating at 1550 nm which acted as the probe beam as this wavelength does not cause OD in KLTN. The beams from the two input channels were co-aligned by a beam splitter to form one input beam which propagates along the optical axis of the system (along the z direction in Fig. 1). The sample was placed between two polarizers. The input polarizer was oriented at 45° to the crystallographic [100] axes (i.e. parallel to direction x + y in Fig. 1). The output polarizer was oriented at −45° to the crystallographic [100] axes, and at 90° to the input polarizer (i.e. parallel to direction x-y in Fig. 1). In addition, a quarter-wave plate was placed in front of the output polarizer in order to compensate for the natural birefringence (BR) induced by dipolar clusters which are created spontaneously in the vicinity of the phase transition [22].
The output intensity for a “crossed polarizers” set-up is given by [18]:
where Iin is the input intensity, Iout is the output intensity, d is the length of the crystal (in the z direction in Fig. 1), λ is the wavelength, and Δn is the electrooptically induced BR in the sample (assuming the natural BR in the sample is completely compensated by the λ/4 plate). Plugging Eq. (1) into Eq. (2) yieldswhere g11 and g12 are the effective quadratic electrooptic coefficients for TM and TE propagating beams respectively, and the electric field Eo is applied along the y direction as illustrated in Fig. 1.The KLTN crystals used in the experiments presented below were grown using the top-seeded solution growth method [17]. Two plate shaped samples of 3.3x4.3x1.9 mm3 and 3.3x2.2x1.7 mm3 (samples S1 and S2 respectively) were cut from the grown boule along the crystallographic [100] axes, and polished to optical grade. Gold electrodes were deposited on the 3.3x4.3 mm2, and the 3.3x2.2 mm2 faces of samples S1 and S2 respectively. The ferroelectric - paraelectric phase transition temperature of the samples was found by measurement of the low frequency dielectric constant to occur at the Tc = 17°C and Tc = 27.2°C for samples S1 and S2 respectively. The samples temperature was stabilized during the measurements. The working points were set well above Tc in order to minimize the effect of the depolarization that occurs in the vicinity of the phase transition while maintaining a large EO effect [22].
It should be emphasized that experiments in over 20 different samples were done in variety of conditions for the investigation of the formation and the suppression of OD in KLTN crystals. All experiments have produced similar behavior. Representative typical results are presented henceforth in section 3.
3. Experimental results and analysis
3.1 Optical damage formation
The formation of OD when a unipolar voltage is applied to the crystal is shown in Fig. 2. Here sample S1 was stabilized at T = 24°C and illuminated with a blue laser beam at λ = 445 nm with intensity of 30 W/cm2. The output intensity of the beam as function of time is shown in Fig. 2a immediately after the modulation was started and in Fig. 2b one second later. The amplitude of the driving voltage was set in order to achieve maximal modulation (i.e. polarization rotation of π/2).
Fig. 2 OD formation when a unipolar voltage is applied: The intensity at the output of the modulator as function of time (a) Immediately at the beginning of the operation; and (b) one second later.
It can be seen that at the beginning the intensity at the output of the modulator tracks the applied voltage function. However, after a few milliseconds of operation, the quality of the modulation is significantly degraded, i.e. OD is formed in the EO modulator substrate. One second after the measurement began the output intensity was stable with extinction ratio (i.e., IMax/IMin) that was significantly lower than the extinction ratio before the OD was formed.
3.2 Investigation of the optical damage formation mechanism
In order to facilitate the investigation of the physical mechanism of the OD formation, the low power laser that produces 50 mW at λ = 488 nm was installed in the primary input channel. For this low power source the optical damage time formation is considerably slower, enabling accurate monitoring of the OD formation. First, the effect of OD formation in sample S2 on the probe beam was investigated. The observed intensity at the output vs. the applied voltage are presented in Fig. 3 at t = 0 as the input beam was switched on and the modulation was started, and 300 seconds later.
Fig. 3 Intensity at the output of the modulator vs. the applied voltage at the beginning of the operation and 300 seconds later, measured data and curve fitting according to Eq. (3).
As can be seen the output intensity at t = 0 is centered around Eo = 0 as predicted in Eq. (3). However, during the operation, as the OD builds up, the output curve gradually shifts. This is attributed to the formation of an internal electric field induced by a space charge generated by photo-excitation of charge carriers by the λ = 488 nm photons. This internal electric field partially negates the applied electric field.
It should be noted that the electric field induced by the optical damage mechanism (EOD) depends on the spatial distribution of the incident (excitation) beam, and on the spatial distribution of the impurities in the crystal. As the incident beam is switched on, OD is starting to build up until it reaches a steady state level. Note that the charge carriers that form the OD space charge are photo-excited from and trapped by impurity ions that reside deep within the band gap. Therefore, this space charge keeps accumulating whenever the illumination is switched on. For a given distribution of the illumination intensity and level of the applied field the space charge will evolve in space and time inducing a field given by EOD(t,r), that eventually will reach a steady state level that will remain stable for long periods of time after the illumination is switched off. Neglecting spatial variations in the induced field within the volume traversed by the propagating field, the output intensity is given by introducing the EOD(t) into Eq. (3), which yields
where EOD is the field induced by the optical damage mechanism as it evolves.3.3 The optical damage time evolution
The time evolution of the OD was also investigated quantitatively by monitoring a probe beam at λ = 1550 nm that was co-aligned with the pump beam (the latter at λ = 488 nm). In addition, a notch filter was installed at the output of the system in order to prevent pump photons from reaching the detector. In this configuration the probe beam at the output of the system manifests the development of the OD following the activation of the pump beam. In particular, the power of the probe beam at the output of the system Iout(t), can be used to derive the time evolution of EOD - the field produced in the EO medium when the pump beam is activated under the application of an external field Eo. Given the power at the output of the system is given by Eq. (4), the relation between Iout(t) and EOD(t) is given by
Note that it is assumed in Eq. (5) that EOD(t = 0) = 0, and that Eo is chosen so that the sine in Eq. (4) does not reach its first maximum point. The buildup of EOD in sample S2 following the activation of the pump beam under a continuous dc external field of Eo = 1200 V/cm is shown in Fig. 4.
It was previously shown in [23, 24] that in KLTN at the paraelectric phase the creation and the time evolution of EOD have two origins with different time scales. It can be seen in Fig. 4 that the time evolution of EOD is characterized by two time constants and can accordingly be described by
where A = 187.6 ± 0.8 V/cm; B = 369.2 ± 0.9 V/cm; = 1.42 ± 0.04 hour; and = 11·10−3 ± 0.5·10−3 hour. The fast time constant is attributed to the formation of the space charge [cf. 25]. The slow time constant is attributed to the relaxation of the dipolar clusters which appear in the vicinity of the ferroelectric phase transition below the Burns temperature [26], and manifest dynamics of a glass forming liquid [27]. The clusters are reoriented in correlation with the field in the crystal, namely Eo-EOD, with a relaxation time which becomes longer upon approaching Tc [22].It should be noted that the optical damage is formed by the accumulated space charge. Hence, the optical damage is proportional to the accumulated absorbed energy of the exciting illumination. In the latter experiment, we used the low power laser that produces 50 mW at λ = 488 nm in order to slow down the formation of the space charge so that we can monitor the formation of the damage more accurately. But in principle the conclusion extracted from the two sets of experiments (the first presented in section 3.1 whereas the second presented in section 3.2 and 3.3), are based on the same physical mechanism.
3.4 Suppression of the formation of optical damage in KLTN
The OD in KLTN can be inhibited by exploiting the fact that the EO effect in the paraelectric phase is quadratic. As can be seen in Eq. (3) the electrically induced changes in the refractive index are independent of the polarity of the applied electric field. However, the photo-excited charge carriers drift in the direction of the applied electric field. Thus, when applying a series of electric pulses with alternating polarity, the photorefractive space charge will not accumulate, and therefore OD will not be formed.
This prediction was put to the test in a crossed polarizers EO modulator with a KLTN medium using the set-up illustrated in Fig. 1. The sample used in the experiment was S1 stabilized at T = 24°C = Tc + 7°C. The input light source was a 1 W laser operating at λ = 445 nm that produced an intensity of 30 W/cm2 at the input plane of the KLTN crystal. The results are presented in Fig. 5. First, in Fig. 5(a) the bipolar driving voltage vs. time is presented. Note that each half cycle of the applied voltage is a two steps function. This can be used for applying the voltage needed for compensating the effect of the natural birefringence that occurs at the paraelectric phase close to the phase transition obviating the need for a quarter-wave plate. The EO modulator was operated continuously for 6 hours. The light intensity at the output of the modulator is presented in Fig. 5(b). The output signal during 5 miliseconds of operation immediately after the laser is turned on and 6 hours later are shown. There is no discernible difference between the signals. The observed extinction ratio is IMax/IMin>500. It should be mentioned that at the beginning of the experiment sample S1 manifested residual OD that remained from previous experiments. (The latter were done with unipolar driving voltage). Once the input beam was switched on under bipolar driving voltage the residual OD was inhibited.
Fig. 5 Inhibition of OD formation by applying bipolar voltage: (a) The bipolar driving voltage; (b) The output intensity of the modulator vs. time at the beginning of the operation, and after 6 hours of continuous operation.
4. Conclusion and discussion
In conclusion, it was observed that KLTN at the paraelectric phase develops optical damage when illuminated at short wavelengths under unipolar driving voltage. This is similar to the OD developed in LNB and LT at the ferroelectric phase. Here, however, the space charge that is generated by the photorefractive process is accompanied by a redistribution of the dipolar clusters that are formed in KLTN upon approaching the phase transition. The dipolar clusters are not affected directly by the excitation, but rather by the space charge that is formed by the photorefractive effect which is the source of the optical damage and causes them to reorient [24]. As was shown in [24], the relaxation time of the dipolar cluster close to the phase transition is very slow in comparison to the modulation cycle. Thus, under bipolar driving field, the photorefractive space charge will not accumulate, and therefore the dipolar clusters will not contribute to the formation of the OD.
It should be noted that as shown above, in KLTN the dominant transport mechanism of the retrapped charge carriers that were photo-excited by the illumination is the drift under the applied electric field. Hence, in EO devices and integrated optics applications, where similar electric fields are applied, OD is formed at short visible wavelengths. However, it was shown that in KLTN the formation of the optical damage can be inhibited at the paraelectric phase by applying an alternating bipolar driving voltage without affecting the electrically induced changes in the refractive index due to the fact that the EO effect is quadratic.
The effectiveness of the suppression of OD in KLTN based quadratic EO modulators was tested in over 20 samples which exhibited no OD when a bipolar voltage was applied. Such OD immune quadratic EO modulators have been used for real applications such as four dimensional EO biosensor array [28]. In these applications the EO modulators were operated at the visible light range continuously for a lot of days without manifesting any OD.
Finally, the method presented above for the suppression of OD in KLTN, together with the methodology for constructing complex integrated photonic circuits in KLTN by the implantations of high energy ions [29], opens the way for a new class of EO devices and circuits that operate at the entire visible-near IR spectral range.
Acknowledgment
This research was supported in part by Israeli Ministry of Science, Technology, and Space Grant No. 310892.
References and links
1. A. Manz, N. Graber, and H. M. Widmer, “Miniaturized total chemical analysis systems - a novel concept for chemical sensing,” Sens. Actuators B Chem. 1(1-6), 244–248 (1990). [CrossRef]
2. A. Ashkin, G. D. Boyd, J. M. Dziedzic, R. G. Smith, A. A. Ballman, J. J. Levinstein, and K. Nassau, “Optically‐induced refractive index inhomogeneities in LiNbO3 and LiTaO3,” Appl. Phys. Lett. 9(1), 72–74 (1966). [CrossRef]
3. L. Solymar, D. Webb, and A. Grunnet-Jepsen, The Physics and Applications of Photorefractive Materials (Clarendon, 1996).
4. K. Buse, “Light-induced charge transport processes in photorefractive crystals I: Models and experimental methods,” Appl. Phys. B 64(3), 273–291 (1997). [CrossRef]
5. A. M. Glass, D. von der Linde, and T. J. Negran, “High‐voltage bulk photovoltaic effect and the photorefractive process in LiNbO3,” Appl. Phys. Lett. 25(4), 233–235 (1974). [CrossRef]
6. R. L. Townsend and J. T. LaMacchia, “Optically induced refractive index changes in BaTiO3,” J. Appl. Phys. 41(13), 5188–5192 (1970). [CrossRef]
7. F. S. Chen, “Optically induced change of refractive indices in LiNbO3 and LiTaO3,” J. Appl. Phys. 40(8), 3389–3396 (1969). [CrossRef]
8. K. Buse, J. Imbrock, E. Kratzig, and K. Peithman, “Photorefractive effects in LiNbO3 and LiTaO3,” in Photorefractive Materials and Their Applications 2, P. Gunter, J.-P. Huig-nard ed. (Springer, 2007).
9. N. V. Kukhtarev, V. B. Markov, S. G. Odulov, M. S. Soskin, and V. L. Vinetskii, “Holographic storage in elctrooptic crystals II. Beam coupling-light amplification,” Ferroelectrics 22(1), 961–964 (1978). [CrossRef]
10. P. Günter and J.-P. Huignard, Photorefractive Materials and Their Applications I, II (Springer, 2006, 2007).
11. D. A. Bryan, R. Gerson, and H. E. Tomaschke, “Increased optical damage resistance in lithium niobate,” Appl. Phys. Lett. 44(9), 847–849 (1984). [CrossRef]
12. B. Sturman, M. Kösters, D. Haertle, C. Becher, and K. Buse, “Optical cleaning owing to the bulk photovoltaic effect,” Phys. Rev. B 80(24), 245319 (2009). [CrossRef]
13. M. Falk and K. Buse, “Thermo-electric method for nearly complete oxidization of highly iron-doped lithium niobate crystals,” Appl. Phys. B 81(6), 853–855 (2005). [CrossRef]
14. M. Kösters, B. Sturman, P. Werheit, D. Haertle, and K. Buse, “Optical cleaning of congruent lithium niobate crystals,” Nat. Photonics 3(9), 510–513 (2009). [CrossRef]
15. F. Lüdtke, N. Waasem, K. Buse, and B. Sturman, “Light-induced charge-transport in undoped LiNbO3 crystals,” Appl. Phys. B 105(1), 35–50 (2011). [CrossRef]
16. C. H. Perry, R. R. Hayes, and N. E. Tornberg, “Ferroelectric soft modes and anti resonance behavior in KTN mixed crystals,” in Proceeding of the International Conference on Light Scattering in Solids, M. Balkansky, ed. (Wiley, 1975), pp. 812–815.
17. R. Hofmeister, A. Yariv, and A. J. Agranat, “Growth and characterization of the perovskite K1-yLiyTa1-xNbxO3:Cu,” J. Cryst. Growth 131(3-4), 486–494 (1993). [CrossRef]
18. A. Yariv and P. Yeh, Photonics: Optical Electronics in Modern Communications (Oxford University, 6th Edition, 2007).
19. S. M. Rytov, “Electromagnetic properties of a finely stratified medium,” Sov. Phys. JETP 2, 605–616 (1956).
20. A. Bitman, N. Sapiens, L. Secundo, A. J. Agranat, G. Bartal, and M. Segev, “Electroholographic tunable volume grating in the g44 configuration,” Opt. Lett. 31(19), 2849–2851 (2006). [CrossRef] [PubMed]
21. L. Vicari, Optical Applications of Liquid Crystals (Taylor & Francis, 2003).
22. A. Gumennik, Y. Kurzweil-Segev, and A. J. Agranat, “Electrooptical effects in glass forming liquids of dipolar nano-clusters embedded in a paraelectric environment,” Opt. Mater. Express 1(3), 332–343 (2011). [CrossRef]
23. G. Bitton, M. Razvag, and A. J. Agranat, “Formation of metastable ferroelectric clusters in K1-xLixTa1-yNbyO3:Cu,V at the paraelectric phase,” Phys. Rev. B 58(9), 5282–5286 (1998). [CrossRef]
24. A. J. Agranat, M. Razvag, M. Balberg, and V. Leyva, “Dipolar holographic gratings induced by the photorefractive process in potassium lithium tantalate niobate at the paraelectric phase,” J. Opt. Soc. Am. B 14(8), 2043–2048 (1997). [CrossRef]
25. A. Yariv, S. Orlov, G. Rakuljicand, and V. Leyva, “Holographic fixing, readout, and storage dynamics in photorefractive materials,” Opt. Lett. 20(11), 1334–1336 (1995). [CrossRef] [PubMed]
26. G. Burns and F. H. Dacol, “Polarization in the cubic phase of BaTIO3,” Solid State Commun. 42(1), 9–12 (1982). [CrossRef]
27. S. E. Lerner, P. Ben Ishai, A. J. Agranat, and Y. Feldman, “Percolation of polar nanoregions: A dynamic approach to the ferroelectric phase transition,” J. Non-Cryst. Solids 353(47-51), 4422–4427 (2007). [CrossRef]
28. Y. Kabessa, V. Korouma, H. Ilan, S. Yagur-Kroll, S. Belkin, and A. J. Agranat, “Simultaneous quantification of the fluorescent responses of an ensemble of bacterial sensors,” Biosens. Bioelectron. 49, 394–398 (2013). [CrossRef] [PubMed]
29. A. J. Agranat, A. Gumennik, and H. Ilan, “Refractive index engineering by fast ion implantations: a generic method for constructing multi-components electro-optical circuits,” Proc. SPIE 7604, 1–17 (2010). [CrossRef]
