« journal navigation
Light-induced symmetry breaking and related giant enhancement of nonlinear properties in CdZnTe:V crystals
Optics Express, Vol. 14, Issue 20, pp. 9385-9390 (2006)
http://dx.doi.org/10.1364/OE.14.009385
Acrobat PDF (117 KB)
Article Navigation
▸ Article Outline
– Abstract
1. Introduction
2. Experiments and results
3. Discussion
4. Conclusion
– References and links
– Abstract
1. Introduction
2. Experiments and results
3. Discussion
4. Conclusion
– References and links
Article Tools
Full Text
Article Info
References
Cited By
Figures
Metrics
Download PDF
Viewing Equations in the
HTML Article
Optimal Viewing Recommendations
Abstract
We report on enormous light-induced reversible strain effects in CdZnTe:V crystals, which lead to a remarkable enhancement of their nonlinear properties, such as electrostriction and electro-optic effects. Using both high resolution x-ray diffraction and optical interferometry we measure light-induced relative deformation of the initial crystalline lattice (changes in d-spacings) up to 0.15%. The experimental findings are attributed to light-induced breaking of the initial cubic crystalline symmetry. Our results point to a family of inorganic materials whose nonlinear properties can be remarkably enhanced by light, offering new possibilities for nonlinear frequency conversion, generation of Terahertz radiation, electro-optic modulation, and self-deflection of optical beams.
© 2006 Optical Society of America
1. Introduction
Semiconductor crystals are known to display a very small electro-optic response [1]. For example, the coefficients of the electro-optic (Pockels) effect in GaAs, InP, CdTe, and CdZnTe (CZT) are all in the range of 1–7 pm/V, which should be compared with 1640 and 1340 pm/V found in BaTiO3 and Sr0.75Ba0.25Nb2O6, respectively [1]. This 2–3 orders of magnitude difference is fundamental, and is caused by the dissimilar origin of the electro-optic effect in these materials. In cubic semiconductors, the electro-optic effects are primarily of electronic origin, whereas in dielectric crystals of non-cubic symmetry (such as BaTiO3 and Sr0.75Ba0.25Nb2O6), the strong electro-optic response is associated with ionic displacements. Recently, however, we have reported on an anomalously large electro-optic response in CdZnTe doped by vanadium (CZT:V) [2]. The index change depends linearly on light intensity and reaches Δn≈0.01. This corresponds to an effective electro-optic coefficient of ~900 pm/V. This finding cannot be understood through the conventional electro-optic effect in CZT, because to support such a giant index change the electric field would have to be of the order of the local atomic field in the crystal. Moreover, we observed large refractive index changes in experimental geometries for which the electro-optic coefficients should be zero according to the initial cubic symmetry of the crystalline lattice.
S. Shwartz, M. Segev, and U. El-Hanany, “Self-deflection and all-optical beam steering in ZnCdTe,” Opt Lett. 29, 760–762 (2004). [CrossRef] [PubMed]
Here, we investigate the physical origin of the large refractive index change detected in our previous experiment [2]. We demonstrate experimentally that light, combined with an electric field induces enormously large reversible strain in CZT: V crystals, breaking the crystalline symmetry and leading to a drastic enhancement of their nonlinear properties, such as electrostriction and electro-optic effects. Using high-resolution x-ray diffraction and optical interferometry we measure light-induced deformations of the crystalline lattice which depend linearly on light intensity and quadratically upon the applied electric field. We measure values as large as 0.15% in lattice deformations (change in d-spacings), which are three orders of magnitude higher than the value calculated with the tabulated piezoelectric coefficients [3]. We emphasize that the lattice deformations observed in our experiments are induced by light and are reversible, in sharp contradistinction with the results of all previous studies on CZT, where structural deformations have been identified as a consequence of elemental replacements or ordered growth of epitaxial layers, both irreversible (see, e.g., [4,5]). We therefore conclude that the large enhancement of the nonlinear electro-optic and electrostriction effects observed in CZT: V under illumination arises from light-induced crystalline symmetry-breaking.
S. Shwartz, M. Segev, and U. El-Hanany, “Self-deflection and all-optical beam steering in ZnCdTe,” Opt Lett. 29, 760–762 (2004). [CrossRef] [PubMed]
R. Weil, R. Nkum, E. Muranevich, and L. Benguigui, “Ferroelectricity in Zinc Cadmium Telluride,” Phys. Rev. Lett. 62, 2744–2746 (1989). [CrossRef] [PubMed]
M. C. Mondoloni, R. Triboulet, and J. Rioux, “Structural distortions from Ferroelectric Origin in Zn 1-x Cd x Te Semiconductor Alloys,” Solid State Commun. 75, 275–277 (1990). [CrossRef]
2. Experiments and results
Our Cd1-xZnxTe crystals are grown by the modified version of the horizontal Bridgman technique [6]. The nominal Zn concentration is x=0.01, and the nominal value of the V doping is ~1017 cm-3. We measure the in-situ lattice strain by using high-resolution x-ray diffraction, while the crystal is biased by an electric field, and illuminated (uniformly) with a laser beam of λ=980 nm wavelengths. The x-ray measurements are carried out in the Bragg scattering geometry using a Bede D3 multi-purpose diffractometer combined with a 18 kW Rigaku rotating anode generator and a channel-cut crystal-monochromator and analyzer for the incident and diffracted beams. High accuracy (Δa/a≈5·10-5) absolute measurements of the lattice parameters are performed by the Bond method [7]. The samples for this study are of a 5×5×1 mm3 rectangular shape, with faces oriented in directions perpendicular to the crystallographic directions <110>, <110̄>, and <001>. Metallic electrodes were deposited on the 5×1 mm2 faces. In this geometry, the bias electric field is applied along the <110> crystallographic direction, while the light propagates along <110̄>. High-resolution (004) x-ray diffraction profiles are taken from the (001) atomic planes parallel to the 5×5 mm2 faces. Figures 1(a) and 1(b) show the (004) d-spacing (left-hand axis) and the induced strain (right-hand axis) as functions of the applied electric field and light intensity, respectively.
P. Cheuvart, U. El-Hanany, D. Schneider, and R. Triboulet, “CdZnTe Crystal Growth By Horizontal Bridgman Technique,” J. Crystal Growth 101, 270–274 (1990). [CrossRef]
W. L. Bond, “Precision Lattice Constant Determination,” Acta Cryst. 13, 814–816 (1960). [CrossRef]
Fig. 1. Experimental results displaying light-induced tensile strain (lattice expansion) measured through x-ray diffraction (a, b) and optical interferometry (c, d). The interferometric method is illustrated at the bottom right panel. (a) Light-induced change (blue diamonds) and change in the absence of light illumination (pink squares) in the (004) d-spacing (left-hand axis) and strain (right-hand axis) vs. an applied electric field. The light-induced strain measurements are taken at a constant light intensity of 253 mW/cm2. (b) Light-induced change in the (004) d-spacing (left-hand axis) and strain (right-hand axis) vs. light intensity at a fixed applied electric field of 6 kV/cm. In both (a) and (b) experiments, x-ray diffraction is taken from the (001)-plane. (c) Light-induced expansion of the crystal (left-hand axis) and strain (right- hand axis) vs. an applied electric field at a constant light intensity of 730 mW/cm2. (d) Light-induced expansion of the crystal (left-hand axis) and strain (right- hand axis) vs. light intensity at a fixed applied electric field of 6 kV/cm. All interferometric measurements represent light reflection from the (001)-plane, and are taken with an electric field applied in the direction and the light beam propagating through the crystal along the <110̄> direction. The intensity of Beam 1 is roughly 1 mW/cm2, which is negligible compared with the intensity of Beam 2.
The results of x-ray measurements shown in Figs. 1(a) and 1(b) provide direct evidence for the strong modifications of the lattice d-spacings under the experimental conditions, and exemplify the crucial role played by the light beam causing these changes. Specifically, in the absence of illumination [pink squares in Fig. 1(a)], the measured strain is zero within the accuracy of x-ray measurement. With illumination, the strain increases quadratically with the applied field and linearly with the light intensity, reaching the value of Δa/a≈4·10-4 (blue diamonds in Figs. 1(a) and 1(b). However, the x-ray penetration depth (under our experimental conditions) is only several microns (~10,000 atomic layers) beneath the surface of the sample. To obtain information on the light-induced deformation in the sample bulk, we use optical interferometry-by replacing one of the mirrors in a Michelson interferometer with the CZT: V crystal while keeping the far end of the crystal fixed (see lower right panel in Fig. 1). A direct comparison between the strain measured via x-ray diffraction and the strain as deduced from the interferometric measurements can establish whether the induced strain is indeed a bulk effect, in which all unit cells are deformed collectively. Typical results are shown in Fig. 1(c) and 1(d), depicting the absolute crystal expansion (left-hand axes) and relative deformation (right-hand axes) as functions of applied field and light intensity. The interferometric results are indeed consistent with those obtained by x-ray diffraction to within 20%. The remaining difference is attributed to surface charge, which to some extent reduces the electric field near the surface, thereby influencing the x-ray diffraction results. These experiments unequivocally prove that the light-induced deformation is a bulk effect.
Fig. 2. Experimental results of eff Δn measurements. (a) eff Δn vs. applied electric field at a constant light intensity of 730 mW/cm2. (b) eff Δn vs. light intensity at a fixed applied electric field of 6 kV/cm. All interferometric measurements represent eff Δn in the <001> direction, and are taken with the electric field applied in the <110> direction and the light beam inducing eff Δn propagating through the crystal along the <110̄> direction.
Another way to track the effects of light-induced crystalline deformation is by monitoring the phase changes accumulated by a wave propagating through the crystal, which experiences simultaneously a light-induced length change and an electro-optic index change. Such an experiment compares the magnitude of two nonlinear effects, electrostriction and the electro optic effect, which in CZT: V are both enhanced by light. To perform a direct comparison, we construct a Mach-Zehnder interferometer and insert the crystal in one of its arms (see Fig. 2). When both the refractive index and the length of the crystal vary with light intensity, the effective index change is Δneff=Δn+n(ΔL/L), where Δn the index change is, L is the crystal length, n is the refractive index, and ΔL is the length change. Figure 2 shows Δneff as a function of applied electric field and light intensity. Comparing the measured values of Δneff and strain, we find that both depend linearly on the light intensity and quadratically on the applied electric field.
Finally, we repeat both interferometric experiments in a different crystalline geometry: applying electric field along <001>, light propagation along<110̄>, the strain and phase changes (electro-optic+strain) along <110>. The results are shown in Figs. 3(a)–3(d). The values of the observed strain are very close to those obtained in the geometry depicted in Fig. 1, with one exception: now the crystal exhibits contraction rather than expansion [8].
Fig. 3. Experimental results presenting light-induced compressive strain (lattice contraction) in the crystal and Δneff in an experimental geometry (with respect to the crystalline axes) different than that of Figs. 1 and 2. Results are taken through interference measurements, with the electric field applied in the <001> direction and the light beam propagating through the crystal in the <110̄> direction. The data in (a, b) is taken with the interferometer of Fig. 1, through light reflection from the (110) plane. The data in (c, d) is taken with the interferometer of Fig. 2, by propagating the probe beam through the crystal in the <110> direction (a) Contraction of the crystal (left-hand axis) and the induced strain (right-hand axis) vs. applied electric field measured at a constant light intensity of 331 mW/cm2. (b) Contraction of the crystal (left-hand axis) and the induced strain (right-hand axis) vs. light intensity measured at a fixed applied electric field of 4 kV/cm. (c) Δneff vs. applied electric field measured at a constant light intensity of 378 mW/cm2.
3. Discussion
Our experiments have shown that, under illumination, the changes in the refractive index and lattice strain depend quadratically on the applied electric field. Formally, one could attribute the quadratic field dependence to the “conventional” dc Kerr effect and electrostriction, respectively, which are both described by the fourth rank tensors. However, these tensors have certain zero components for the CZT 4̄3m point group [9]. This should prohibit the observed electrostriction and dc Kerr effect in some of the experimental geometries we have used. For example, the g41 coefficient of the dc Kerr effect and the e14 tensor element of electrostriction are both zero for the CZT point group, implying that in the experiment of Fig. 3 (electric field applied along <001> with the light propagating along <110>), neither electrostriction nor electro-optic index changes should have been observed. Our experiments, however, reveal the opposite: the measured light-induced effects are in fact very strong, and they are insensitive to the polarization of both light beams (i.e., the beam causing the effect and the probe beam, marked as Beams 1 and 2 in Fig. 1, respectively). The fact that the observed effects are insensitive to the light polarization fundamentally contradicts the properties of the dc Kerr effect.
We emphasize that the observed strain cannot be attributed simply to thermal effects induced by the incident light beam, because we find that the temperature increase in the presence of the most intense beams we used (~1W/cm2) is only 2°C. By using the tabulated values for the thermo-optic effect (1.4·10-4K-1) [10] and the thermal expansion coefficient (5.9·10-6K-1) [3], one should expect to observe Δn=2.8·10-4 and a tensile strain of 1.2·10-5. These values are about two orders of magnitude smaller than those found by us experimentally. Moreover, the thermal expansion coefficient is a second rank tensor, and hence is isotropic in cubic crystals. This means that in cubic crystals, heating can only lead to isotropic lattice expansion, which clearly stands in a sharp contrast to our results. For example, thermo-optic effects cannot explain the lattice contraction displayed in Fig. 3. In a similar vein, thermo-optic effects in CZT can account only for expansion of the crystal unit cell, hence they cannot explain the contraction effect displayed in Fig. 3. The refractive index change cannot be attributed to elasto-optic effects for the same reasons that rule out attributing the observations to the dc Kerr effect. Namely, the elasto-optic effect is highly sensitive to light polarization, and is also described by a fourth-rank tensor with the same zero components as the dc Kerr effect, both are with a sharp contradiction with our experimental findings.
D. O. Caplan, G. S. Kanter, and P. Kumar, “Characterization of dynamic optical nonlinearities by continuous time-resolved Z-scan,” Opt. Lett. 21, 1342–1344 (1996). [CrossRef] [PubMed]
Our experimental findings suggest that the large refractive index changes and the huge strain we observe arise from light-induced breaking of the initial cubic symmetry of the CZT crystals. The light excites charge carriers that are subsequently separated by the electric field and then re-trapped, thereby establishing a strong internal space-charge field. The space-charge-field leads to collective ion displacements within the crystalline lattice, thereby breaking the initial cubic symmetry of the CZT crystals.
All experimental results described above are obtained for CZT crystals doped with V. To elucidate the role of the doping, we have performed a series of experiments with nominally undoped CZT crystals, in the same configurations and under the same experimental conditions described above. We find that the electro-optic response of undoped CZT is much weaker than in doped CZT. In fact, undoped CZT crystals behave exactly as expected from the electro-optic coefficient r41=5.5pm/V, the values provided by conventional electro-optic effects in cubic semiconductors.
4. Conclusion
In conclusion, we have demonstrated that laser light causes significant lattice deformation within the bulk of CdZnTe: V crystals, which increase linearly with light intensity and quadratically with an applied electric field up to 0.15%. This light-induced electrostriction is accompanied by an enormous enhancement of the electro-optic effect, resulting in a giant refractive index change of about 0.01. To our knowledge, such light-induced deformation and refractive index changes are the highest ever reported for a bulk semiconductor, and cannot be explained through intrinsic electrostriction and the dc Kerr effects. The CZT: V is, most probably, not the sole crystal in which the observed phenomena can occur. We believe that such light-induced symmetry-breaking effects will take place in the family of inorganic materials of cubic symmetry, containing ions weakly bonded and easy to move, and doped with midgap dopants. The symmetry breaking itself, being of ionic nature, occurs in time scales of the order of the dielectric relaxation time, which, for the ~1W/cm2 intensity used in our experiments, is of the order of a few msec. However, once the initial symmetry is broken and the crystal acquires strong internal dipole moments, all of its nonlinear properties arising from ultrafast electronic origins are enormously enhanced as well. We therefore anticipate that, under illumination, CZT: V will be an efficient medium for nonlinear frequency conversion, optical rectification, and for generation of THz electromagnetic waves.
References and links
A. Yariv, in Optical Electronics in Modern Communication , Fifth Edition, (Oxford University Press, Oxford, 1997), Chap. 9. | |
S. Shwartz, M. Segev, and U. El-Hanany, “Self-deflection and all-optical beam steering in ZnCdTe,” Opt Lett. 29, 760–762 (2004). [CrossRef] [PubMed] | |
H. Landolt and R. Bornstein , Numerical Data and Functional Relationships in Science and Technology-New Series, Group III, Volume 11, K. H Hellwege,ed. (Springer-Verlag Berlin, 1979). | |
R. Weil, R. Nkum, E. Muranevich, and L. Benguigui, “Ferroelectricity in Zinc Cadmium Telluride,” Phys. Rev. Lett. 62, 2744–2746 (1989). [CrossRef] [PubMed] | |
M. C. Mondoloni, R. Triboulet, and J. Rioux, “Structural distortions from Ferroelectric Origin in Zn 1-x Cd x Te Semiconductor Alloys,” Solid State Commun. 75, 275–277 (1990). [CrossRef] | |
P. Cheuvart, U. El-Hanany, D. Schneider, and R. Triboulet, “CdZnTe Crystal Growth By Horizontal Bridgman Technique,” J. Crystal Growth 101, 270–274 (1990). [CrossRef] | |
W. L. Bond, “Precision Lattice Constant Determination,” Acta Cryst. 13, 814–816 (1960). [CrossRef] | |
Note that Δneff is positive in both geometries, suggesting that the net (electro-optic) refractive index change in the geometry depicted in Fig. 3 is somewhat larger than the one measured in the geometry shown in Fig. 2 (because Δneff>0 in Fig. 2 whereas ΔL<0 in Fig. 3). | |
J. F. Nye, Physical Properties of Crystals , (Oxford University Press, Oxford, 2001). | |
D. O. Caplan, G. S. Kanter, and P. Kumar, “Characterization of dynamic optical nonlinearities by continuous time-resolved Z-scan,” Opt. Lett. 21, 1342–1344 (1996). [CrossRef] [PubMed] |
OCIS Codes
(190.4400) Nonlinear optics : Nonlinear optics, materials
(190.5940) Nonlinear optics : Self-action effects
ToC Category:
Nonlinear Optics
History
Original Manuscript: August 22, 2006
Revised Manuscript: September 25, 2006
Manuscript Accepted: September 26, 2006
Published: October 2, 2006
Citation
Sharon Shwartz, Raoul Weil, Mordechai Segev, Eugene Lakin, Emil Zolotoyabko, Vinod M. Menon, Stephen R. Forrest, and Uri EL-Hanany, "Light-induced symmetry breaking and related giant enhancement of nonlinear properties in CdZnTe:V crystals," Opt. Express 14, 9385-9390 (2006)
http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-14-20-9385
Sort: Year | Journal | Reset
References
- A. Yariv, in Optical Electronics in Modern Communication, Fifth Edition, (Oxford University Press, Oxford, 1997), Chap. 9.
- S. Shwartz, M. Segev and U. El-Hanany, "Self-deflection and all-optical beam steering in ZnCdTe," Opt Lett. 29, 760-762 (2004). [CrossRef] [PubMed]
- H. Landolt and R. Bornstein, Numerical Data and Functional Relationships in Science and Technology-New Series, Group III, Volume 11, K. H Hellwege, ed. (Springer-Verlag Berlin, 1979).
- R. Weil, R. Nkum, E. Muranevich, and L. Benguigui, "Ferroelectricity in Zinc Cadmium Telluride," Phys. Rev. Lett. 62, 2744-2746 (1989). [CrossRef] [PubMed]
- M. C. Mondoloni, R. Triboulet, and J. Rioux, "Structural distortions from Ferroelectric Origin in Zn1-xCdxTe Semiconductor Alloys,"Solid State Commun. 75, 275-277 (1990). [CrossRef]
- P. Cheuvart, U. El-Hanany, D. Schneider and R. Triboulet, "CdZnTe Crystal Growth By Horizontal Bridgman Technique," J. Crystal Growth 101, 270-274 (1990). [CrossRef]
- W. L. Bond, "Precision Lattice Constant Determination," Acta Cryst. 13, 814-816 (1960). [CrossRef]
- Note that is positive in both geometries, suggesting that the net (electro-optic) refractive index change in the geometry depicted in Fig. 3 is somewhat larger than the one measured in the geometry shown in Fig. 2 (because > 0 in Fig. 2 whereas ΔL< 0 in Fig. 3).
- J. F. Nye, Physical Properties of Crystals, (Oxford University Press, Oxford, 2001).
- D. O. Caplan, G. S. Kanter, and P. Kumar, "Characterization of dynamic optical nonlinearities by continuous time-resolved Z-scan,"Opt. Lett. 21, 1342-1344 (1996). [CrossRef] [PubMed]
Cited By |
 Alert me when this paper is cited |
OSA is able to provide readers links to articles that cite this paper by participating in CrossRef's Cited-By Linking service. CrossRef includes content from more than 3000 publishers and societies. In addition to listing OSA journal articles that cite this paper, citing articles from other participating publishers will also be listed.
« Previous Article | Next Article »
OSA is a member of 