1PapfsvmrontaswtwgctT�

s=�[8r�spt=

444 J. Opt. Soc. Am. B/Vol. 26, No. 3 /March 2009 Mosimann et al.

Photorefractive waveguides in He+ implanted pureand Te-doped Sn2P2S6

Roger Mosimann,1,* Flurin Juvalta,1 Mojca Jazbinsek,1 Peter Günter,1 and Alexander A. Grabar2

1Nonlinear Optics Laboratory, Swiss Federal Institute of Technology, ETH Zurich, CH-8093 Zurich, Switzerland2Institute of Solid State Physics and Chemistry, Uzhgorod National University, 88 000 Uzhgorod, Ukraine

*Corresponding author: nlo@phys.ethz.ch

Received October 31, 2008; accepted December 1, 2008;posted December 15, 2008 (Doc. ID 103506); published February 9, 2009

We have demonstrated for the first time, to the best of our knowledge, photorefractive two-wave mixing in He+

implanted waveguides in one of the most promising materials for infrared photorefractive applications, theferroelectric semiconductor Sn2P2S6. The high optical nonlinearity is preserved after implantation and at thetelecommunication wavelength �=1.55 �m, a maximal two-wave mixing gain of 2.5 cm−1 has been measured inTe-doped waveguides. In the nominally pure material an increase of the effective number of traps after im-plantation has been observed, resulting in an increase of the two-beam coupling gain by a factor of almost 2 inthe 633–1064 nm spectral range. In 1% Te-doped Sn2P2S6 the effect of ion implantation to the photorefractiveresponse is completely different than in pure materials. While the dominant contribution by holes is not con-siderably affected, a strong, thermally induced charge compensation is observed in the He+ implanted Te-dopedwaveguides. © 2009 Optical Society of America

OCIS codes: 160.5320, 190.5330, 190.7070, 230.7390.

a1tlachliMqpct

tSpscte

2A

1Aiaevhf

. INTRODUCTIONhotorefractive materials are interesting for a variety ofpplications, for example, laser beam cleanup, opticalhase conjugation, signal processing, or spatial solitonormation [1]. Many of these applications require a re-ponse in the near infrared wavelength region, whereery few materials fulfill most of the materials require-ents. Semiconductors are, for example, sensitive in this

ange but they usually need high external fields or shouldperate close to resonance conditions to reach the desiredonlinearity [2,3]. Many conventional photorefractive ma-erials such as the ferroelectric oxides LiNbO3 or KNbO3re very sensitive at shorter wavelengths, but do nothow a high photosensitivity in the infrared region. Oneay of solving this is doping the pure material, which led

o some promising results—mostly with BaTiO3 and alsoith KNbO3 [4–6]. Another successful, but less investi-ated approach may be ion implantation of samples, thusreating a waveguide and additionally introducing newrap levels through deposition of energy in the process [7].his method allowed for a photorefractive sensitivity at=1.55 �m in Fe-doped KNbO3 [8].Tin thiohypodiphosphate �Sn2P2S6� is a ferroelectric

emiconductor with a high electro-optic coefficient (r111170 pm/V at �=633 nm) and a narrower bandgap

2.3 eV� compared to conventional photorefractive oxides9]. It has a wide transparency range from 530 nm up to

�m, making it very attractive for applications at infra-ed wavelengths. A very fast photorefractive response10–100 ms� in the near infrared at moderate light inten-ities of about 1 W/cm2 can be achieved [10,11]. Recently,hotorefractive sensitivity has been demonstrated at theelecommunications wavelength �=1.55 �m ��2.8 cm−1� in the bulk material without an electric field

0740-3224/09/030444-6/$15.00 © 2

pplied, but high light intensities of the order of00 W/cm2 have been required. This is due to the rela-ively large dark conductivity in this material so that theow absorption makes high intensities necessary to induce

photoconductivity significantly higher than the darkonductivity [12]. Recently ion implanted waveguidesave been demonstrated in pure Sn2P2S6 material with

osses of about 10 dB/cm [13]. These waveguides are verynteresting for integrated optical application, such as

ach–Zehnder interferometers or for phase-matched fre-uency conversion [14], but may be also interesting forhotorefractive applications, particularly at telecommuni-ations wavelengths, to reduce power requirements forhe wave-mixing applications [12].

In this work we investigate the photorefractive proper-ies of the ion implanted planar waveguides in puren2P2S6 crystals and in Te-doped Sn2P2S6 crystals im-lanted along different crystallographic directions. Wehow that ion implantation does not affect the high opti-al nonlinearity of this crystal but considerably influenceshe concentration of traps in Sn2P2S6, however in a differ-nt way in pure and doped crystals.

. EXPERIMENTAL. Samples and Ion Implantation

. Pure Sn2P2S6lthough the excellent performance of Sn2P2S6 makes it

nteresting for many applications, only a little is knownbout the origins of the photorefractive response, consid-ring both charge transport and energy (trap) levels in-olved [10,15]. Nominally pure (yellow) Sn2P2S6 crystalsave been divided into two groups depending on their per-ormance [10]. Type I crystals show a strong charge com-

009 Optical Society of America

pTspdnp

wm�t[eipcvAcfuertotmmpfiFcrefl

2Tpsatctiprltcfcancrct=b

BF67la=irtcb�lmtmipwa

osmwam

Fmgetcci

Mosimann et al. Vol. 26, No. 3 /March 2009/J. Opt. Soc. Am. B 445

etition and can be influenced by preillumination [16].his charge competition was used for example to demon-trate a coherent optical oscillator with periodic zero-�hase modulation [17] or the photorefractive slowingown of light [18]. Type II crystals do not show a pro-ounced charge compensation and are preferable for ap-lications requiring high steady-state gain.The pure type II Sn2P2S6 crystal studied in this work

as grown using the conventional vapor transportethod [10]. The sample proportions were 4.88 mm2.92 mm�1.8 mm, along the x, y, and z axes, respec-

ively, where the Cartesian axes x ,y ,z are defined as in10]. The crystal was then irradiated in a TANDEM accel-rator at ETH Zurich on the polished z face by 2 MeV He+

ons at a fluence of 0.25�1015 cm−2. The implantationrocess was performed at room temperature with the ionurrent density being kept to extremely low values to pre-ent excessive heating of the crystals during the process.fter implantation the sample was poled; during this pro-ess it was heated to 100 °C and kept at this temperatureor several hours for annealing purposes. Annealing issually used for ion implanted samples to restore theventual decreased electro-optic coefficient in the guidingegion due to ion-induced crystalline damage [7]. Usinghe barrier coupling method [19] and by approximatingur refractive index profile by a step-index profile we de-ermined a waveguide thickness of 6.0±0.1 �m, whichatches very well with SRIM simulations [www.sri-.org]. This approach is much simpler than the model

resented in [13] considering a more accurate index pro-le, but can describe our results with sufficient precision.urthermore, from these measurements we obtained thehange of the refractive index �n=0.08±0.01 in the bar-ier region, which is in good agreement with the param-ters obtained from implantation experiments at higheruences [13].

. Te-Doped Sn2P2S6he crystal was grown using conventional vapor trans-ort technique with 1% Te in the initial compound. Theample proportions were 4.85 mm�2.3 mm�4.55 mmlong the x, y, and z axes, respectively. The ion implanta-ion was done the same way as for the pure Sn2P2S6, ex-ept for the fluence that was higher, 0.6�1015 cm−2, andhe implantation direction, in this case the y surface, wasrradiated. This implantation direction in Sn2P2S6 hasreviously not been examined yet. Note that this configu-ation makes the end-face polishing process more chal-enging, because the cleavage plane is perpendicular tohe y axis. Figure 1(a) shows the results of the barrieroupling measurement. Whenever the reflectivity as aunction of the angle drops, a mode was coupled into therystal. Figure 1(b) shows the corresponding measurednd theoretical mode indices, used to determine the thick-ess of the waveguide and the barrier refractive indexhange. The line is just a guide to the eye, connecting theesulting mode points. The thickness of the waveguidealculated from this measurement was 5.52±0.05 �m andhe barrier had a refractive index change of �n0.09±0.01. The Te-doped sample was investigated bothefore and after the annealing following the implantation.

. Two-Wave Mixing Measurement Setupor the two-wave mixing experiments a HeNe laser at33 nm �Pmax=10 mW�, single-mode laser diodes at80 nm (Pmax=100 mW, Rainbow Photonics), a Nd:YAGaser at 1064 nm (Pmax=400 mW, Lightwave Electronics),nd a single-frequency fiber laser at 1550 nm (Pmax400 mW, NP Photonics) were used. The beam was split

nto the signal and the pump beam using a glass plate,esulting in a ratio of 1:100 between the two beams. Thewo beams were then coupled into the waveguide using aylindrical lens with focal length of 19 mm to focus theeam in the vertical direction resulting in a spot area of�w0z�w0x�=��4 �m�500 �m� at �=633 nm, with simi-

ar dimensions being used at the other wavelengths. Theaximum power of the respective lasers was used to ob-

ain the highest possible intensity and for the measure-ent in the bulk crystal the beams were focused as well

n order to be in a similar intensity regime. The outputlane of the crystal for the signal beam was visualizedith a second spherical lens with focal length of 25 mmnd imaged onto a CCD camera or a photodiode.For the pure Sn2P2S6 sample the beams were incident

n the xz crystal surface, which is parallel to the mirrorymmetry plane of the crystal with the point group sym-etry m. Because of this the polarization of the beamsas adjusted with respect to the rotation of the indicatrixt the respective wavelength in order to have an eigen-ode propagating in the waveguide [13]. The beams were

ig. 1. (a) Measured reflectivity using the barrier couplingethod [19] as a function of the angle for the Te-Sn2P2S6 wave-

uide implanted with He+ ions with ion energy 2 MeV and flu-nce 0.6�1015 cm−2. (b) The corresponding effective indices forhe measured (dots) and calculated (solid curve) profiles. The cal-ulated modes of the best-fit profiles have been connected forlarity and are shown by the solid curve. The resulting thicknesss 5.52±0.05 �m and the barrier strength �n=0.09±0.01.

iyTetrato

Sptw

3AFuivfsrct

trmtovet

wtseista

tpuTlts0sw=a

eet

tit1sshtef

aaawcfsarorwtw

b

Tc

Fsit

446 J. Opt. Soc. Am. B/Vol. 26, No. 3 /March 2009 Mosimann et al.

ncident in the xy plane symmetrically with respect to theaxis so that the grating vector was parallel to the x axis.his means that the electro-optic coefficient used in thexperiment was not the one usually employed in this ma-erial (r111=170 pm/V at �=633 nm) but a combination of111, r131, and r331 resulting in a coefficient r=180 pm/Vt �=633 nm for light polarized along the x3 main axis ofhe Fresnel ellipsoid and the electric field along the x axisf the chosen Cartesian system [9].

Due to the different orientation of the Te-dopedn2P2S6 sample during implantation the usually em-loyed electro-optic coefficient r111 could be used. Thushe beams entered the crystal through the xy surface andere polarized along the x-axis.

. RESULTS AND DISCUSSION. Pure Sn2P2S6or the evaluation of the amplification data we used thesual plane-wave approximation. If we use the model tak-

ng into account the Gaussian beam shape and stronglyarying intensities in the focused area [20] the photore-ractive gain constant increases by about 10% with re-pect to the value reported, which is within the errorange and therefore neglected. The whole length of therystal was used as the interaction length for the calcula-ion of the two-beam coupling gain.

We first measured the photorefractive gain � as a func-ion of intensity to make sure that we are in the saturatedegion and the dark conductivity does not decrease theeasured gain [12]. The input angle � of the beams onto

he crystal was varied to measure the gain as a functionf the grating spacing �=� / �2 sin ��, where � is theacuum light wavelength. The two-beam coupling gain co-fficient can be described in the diffusion case and withinhe weak probe beam approximation as follows [21]:

�0 =2�reffn

3

�

kBT

e

cos�2���

cos����

2�

��1 + �2�ls

��2� , �1�

here n is the refractive index at the wavelength � ;�� ishe internal angle between the signal beam and theample normal; reff is the effective electro-optic coefficient;is the electron charge; kB is the Boltzmann constant; T

s the temperature; and ls= �0kBT / �e2Neff��1/2 is thecreening length, where is the dielectric constant, 0 ishe vacuum permeability, Neff is the effective trap density,nd reff is the effective electro-optic coefficient.Figure 2(a) shows the grating spacing dependence of

he two-wave mixing gain coefficient � in the nominallyure Sn2P2S6 bulk crystal. The measurement was done bysing the same setup as for coupling into the waveguide.he direction of the amplification with respect to the po-

arization of the crystal allows to determine that holes arehe dominant charge carriers, as is expected [10]. The re-ulting maximal gain coefficients were 3.5, 1.8, and.61 cm−1 at the wavelengths 633, 780, and 1064 nm, re-pectively. The solid curves were obtained using Eq. (1)ith the following data: effective dielectric constant eff230 and the refractive indices n1=3.02 at 633 nm, 2.91t 780 nm, and 2.82 at 1064 nm [10]. With these param-

ters we obtained values for the effective electro-optic co-fficient reff and the effective concentration of traps Neff;hese are listed in Table 1.

Figure 2(b) shows the grating spacing dependence ofhe two-wave mixing gain coefficient � in the waveguiden the same pure Sn2P2S6 crystal. The resulting maximalwo-wave mixing gain coefficients are 6.8, 3.3, and.0 cm−1 at the wavelengths 633, 780, and 1064 nm, re-pectively. The results of the analysis with Eq. (1) arehown in Table 1. The effective concentration of traps Neffas increased in the waveguide region. This is attributedo the ions deposing a small amount of their implantationnergy in the guiding region and thus introducing new de-ects.

In photorefractive measurements, one has to considern effective electro-optic coefficient reff=reff� , where reff� iscombination of the strain-free electro-optic contribution

nd the elasto-optic contribution. is a reduction factor,hich lowers the apparent effective electro-optic coeffi-

ient reff� , due to, e.g., electron–hole competition, differentrom the charge compensation observed at longer timecales. This reducing factor is responsible for the discrep-ncy of reff in Table 1 and the value obtained through di-ect interferometric measurements. The effective electro-ptic coefficient does not change in the waveguidingegion compared to the bulk within the error margins,hich is an important result considering the reduction of

he nonlinear-optic properties in several materials afteraveguide formation [7].The temporal evolution of the buildup can be described

y a single exponential function with the time constant �,

� = �0�1 − exp�− t/���. �2�

he time constants � are shown in Table 1. The directomparison of the time constants obtained in the bulk

ig. 2. Two-wave mixing gain � as a function of the gratingpacing � in nominally pure Sn2P2S6: (a) in the bulk crystal, (b)n the waveguide. The solid curves are according to Eq. (1) andhe resulting parameters are listed in Table 1.

cecsorsortdi

BAtncvbiicmcttip

anilc=piwbb

radadftistoietcKswc

Fptti�

Mosimann et al. Vol. 26, No. 3 /March 2009/J. Opt. Soc. Am. B 447

rystal and in the waveguide region is difficult since thexact coupling efficiency and thus the resulting intensityannot be determined with high precision. Due to thetrong confinement and thus high intensity of the beamne would expect faster time constants in the waveguideegion, which is confirmed by the measurements. At timecales of several minutes no decrease of the gain could bebserved, neither in the waveguide region nor in the bulkegion of the crystal. The amplification was in the direc-ion of the +x crystalline axis, implying that holes are theominant charge carriers in both regions of the crystal, ass usual in Sn2P2S6 crystals [10].

. Te-Doped Sn2P2S6s for pure Sn2P2S6 the grating spacing dependence of

he peak gain coefficient was measured and the effectiveumber of traps Neff and the effective electro-optic coeffi-ient reff determined according to Eq. (1). The respectivealues are listed in Table 1. Neff is almost the same for theulk and the waveguiding region, which is different thann the case of undoped Sn2P2S6, where the implantationncreases its value. In Te-doped Sn2P2S6 the trap levelsreated through implantation do not interfere with theain trapping channel. This might indicate that the con-

entration of these traps has already reached the satura-ion in our sample (1% Te in the initial compound). Some-hing similar is observed if the Te concentration in thenitial compound is increased to 2%–3%. In this case thehotorefractive gain remains about the same as well [11].The direction of the amplification again implies holes

s being the dominant charge carriers. In the bulk crystalo charge compensation was observed, which can be seen

n Fig. 3. We measured the amplified signal beam ononger time scales, of the order of several minutes, andould not observe any compensating effect except at �1.55 �m. At this wavelength a weak and very slow com-ensation effect with �s of the order of 400 s was observed,n good agreement with [12]. On the other hand in theaveguiding region a strong charge compensation coulde observed at all the wavelengths, similar as observed inulk type I samples. Therefore we refer to the charge car-

Table 1. Photorefractive Para

�

(nm)

Nominally pure 633 wg(yellow) bulk

780 wgbulk

1064 wgbulk

SPS:Te 1% 633 wgbulk

780 wgbulk

1064 wgbulk

1550 wgbulk

aThe error of the parameters is about 15%.

iers responsible for the first increase of the signal beams “fast charge carriers” and the ones responsible for theecrease as “slow charge carriers.” Figure 3 shows an ex-mple at �=780 nm. These slow charge carriers are 1 or-er of magnitude slower than those responsible for theormation of the first grating at room temperature, andhus essentially faster than, for example, the compensat-ng effect observed in the bulk at �=1.55 �m, whichhowed compensation effects at a time scale of minutes. Ifhe nature of the compensating charges is the same andnly the concentration is increased by the implantation orf different levels are created is not clear. It is also inter-sting that the effect does not occur in nominally pureype II Sn2P2S6 samples. The creation of new types ofharge carriers through ion implantation is known fromNbO3. Here an even more pronounced effect was ob-

erved, since the direction of the amplification in theaveguiding region was opposite to the one in the bulk

rystal the charge separation process was changed from p

rs of Ion Implanted Sn2P2S6a

��

(ms)reff

(pm/V)Neff

�1016 cm−3�

2 36 0.897 25 0.458 25 0.94

14 24 0.2638 30 0.1364 23 0.0783 36 1.14 33 0.98 46 0.81

11 45 0.6120 76 0.4923 61 0.5185 74 0.3175 72 0.34

ig. 3. Two-beam coupling signal measured after opening theump at t=0 s in Te-doped Sn2P2S6, with crosses representinghe signal in the waveguiding region and squares representinghe bulk region. It can be clearly seen that only in the waveguid-ng region charge compensation occurs, with a time constant

=140 ms (�=780 nm, �=2.0 �m).

mete

�max�cm−1

6.83.53.31.81.00.67.56.45.85.254.12.52.6

wgs

tpSeto

wtvccts

ctlhf

wtcts=cte0tpi

iltcsco

rb

we��cociedg

Tt4alTri

hatdTs

4Wpucr=bcc

cwiwdpct

Fs�cswda

448 J. Opt. Soc. Am. B/Vol. 26, No. 3 /March 2009 Mosimann et al.

o n type, which was attributed to the reduction of Fe im-urities by proton irradiation [8]. The ion implantation inn2P2S6:Te gives us a method to create the compensatingffect if wanted for applications such as the coherent op-ical oscillator with periodic zero-� phase modulation [17]r the photorefractive slowing down of light [18].

To investigate the nature of this compensating effecte measured the temperature dependence of the build-up

imes of the fast and slow charge carriers, as it was pre-iously observed, that the slow carriers in type I Sn2P2S6rystals are rather temperature induced than by light ex-itation [10,16]. Peltier elements were used to cool or heathe crystal and the two-beam coupling signal was mea-ured.

The compensating charge carriers’s build-up time de-reases exponentially with increasing temperature, whilehe fast charge carriers’s build-up time remains more oress constant. This is shown in Fig. 4. The exponential be-avior of the slow charge carriers can be described by theollowing Arrhenius equation:

�wgs = �0 exp��E

kb

1

T� , �3�

here �wgs is the slow time constant, �0 is the normaliza-ion constant, T is the temperature in Kelvin and �E theharacteristic activation energy for the thermal motion ofhe slow charge carriers. The activation energy �E corre-ponding best to the measurement performed at �780 nm in the waveguide is 0.95±0.02 eV for the slowharge carriers. This matches very well with the activa-ion energy measured for the compensating charge carri-rs in bulk Sn2P2S6:Te at �=1.55 �m, where �E is.96±0.2 eV, indicating their common origin, althoughhe response times differ by 3 orders of magnitude. Com-ared to normal bulk type I Sn2P2S6 crystals this energys almost three times larger in the Sn2P2S6:Te He+-ion

ig. 4. Temperature dependence of the inverse build-up time ishown on a logarithmic scale for Sn2P2S6:Te at �=780 nm and=2.0 �m. �bulkf and �wgf are the build-up times for the fast

harge carriers in the bulk material and in the waveguide, re-pectively, and �wgs is the one for the slow charge carriers in theaveguide. Only the slow charge carriers show a strong depen-ence on the temperature. The solid line is according to Eq. (3)nd results in an activation energy of �E=0.95 eV.

mplanted waveguide [16]. The origin could still be simi-ar, since in doped samples we have additional levels inhe gap and so complementary channels for excitation ofharge carriers. At 40 °C the two gratings form at theame speed and by evaluating the magnitude of the gainontribution of the slow charge carriers even exceeds thene from the fast charge carriers.

From the intensity dependence of the charge excitationate it is possible to determine the dark conductivity �darky using the following relation:

1

�=

�photo�I� + �dark

0, �4�

here �photo�I� is the photoconductivity depending lin-arly on the intensity. In the bulk region �dark is 2.710−8 1/ m and in the waveguide region �dark is 7.410−8 1/ m at room temperature. This increased dark

onductivity due to ion implantation increases the speedf the process but may reduce the two-wave mixing effi-iency at low intensities, most strongly at 1.55 �m, wheret has been shown that extremely high intensities are nec-ssary �400 W/cm2� to induce a large enough photocon-uctivity to reach the saturated photorefractive gain re-ion [12].

Of special interest are the photorefractive properties ofe-doped Sn2P2S6 at �=1.55 �m. As mentioned, to reachhe saturated gain region a very high intensity of about00 W/cm2 is necessary [12], which is more easily achiev-ble in a waveguide, especially for a longer interactionength if compared to a focused beam. As can be seen inable 1 the photorefractive properties in the waveguideegion are about the same as in the bulk material, mak-ng it well suited for applications.

To investigate the effect of annealing, the sample waseated to a temperature of 100 °C for several hours anddditionally repoled along the x axis. Measurements ofhe two-beam coupling gain did not show any significantifference compared to before the annealing was done.he compensating effect of the slow charge carriers wastill present and not influenced by the annealing.

. CONCLUSIONe have measured for the first time the photorefractive

roperties of ion implanted waveguides in Sn2P2S6. In thendoped crystal implanted along the z direction an in-rease of the effective number of traps Neff was observed,esulting in a higher two-beam coupling gain �. At �633 nm � increased from 3.5 to 6.8 cm−1. Neither in theulk material nor in the waveguide region could chargeompensation be observed. The high effective electro-opticoefficient reff is preserved by He+ implantation.

In Te-doped Sn2P2S6 for the first time the y face of therystal was implanted allowing the use of r111 and TEaveguiding modes. No increase of Neff could be observed

n the waveguiding region compared to the bulk crystal,hich may be attributed to the high trap concentration inoped samples as compared to the ones created by im-lantation. In the waveguiding region a strong chargeompensation was observed, created by the ion implanta-ion, since in the bulk crystal no such effect was observed.

Tscteol

fThcew

AWMre

R

1

1

1

1

1

1

1

1

1

1

2

2

Mosimann et al. Vol. 26, No. 3 /March 2009/J. Opt. Soc. Am. B 449

he nature of these ion-induced defects is presumablyimilar to those previously observed in type I Sn2P2S6rystals. The temperature dependence of the build-upime of this compensation resulted in an activation en-rgy of �E=0.95±0.02 eV, which is larger than previouslybserved in type I Sn2P2S6 crystals but may have a simi-ar origin.

At �=1.55 �m in Te-doped Sn2P2S6 a similar photore-ractive response was observed as in the bulk material.his is especially interesting since at this wavelength aigh intensity is needed to reach the saturated two-beamoupling region, making waveguides the ideal choice,ven though a slightly higher dark conductivity in theaveguide was measured.

CKNOWLEDGMENTSe thank J. Hajfler for his expert crystal preparation and. Döbeli for the ion-implementation experiments. This

esearch has been supported by the Swiss National Sci-nce Foundation.

EFERENCES1. P. Günter and J.-P. Huignard, eds., Photorefractive

Materials and Their Applications III: Applications(Springer-Verlag, 2007).

2. A. Partovi, J. Millerd, E. M. Garmire, M. Ziari, W. Steier, S.B. Trivedi, and M. B. Klein, “Photorefractivity at 1.5 �m inCdTe-V,” Appl. Phys. Lett. 57, 846–848 (1990).

3. K. Shcherbin, “Recent progress in semiconductorphotorefractive crystals,” in Photorefractive Materials andTheir Applications II, P. Günter and J.-P. Huignard, eds.(Springer, 2007), pp. 391–418.

4. G. W. Ross, P. Hribek, R. W. Eason, M. H. Garrett, and D.Rytz, “Impurity enhanced self-pumped phase conjugationin the near infrared in ‘blue’ BaTiO3,” Opt. Commun. 101,60–64 (1993).

5. C. Medrano, M. Zgonik, N. Sonderer, Ch. Beyeler, S.Krucker, J. Seglins, H. Wüest, and P. Günter,“Photorefractive effect in Cu- and Ni-doped KNbO3 in thevisible and near infrared,” J. Appl. Phys. 76, 5640–5645(1994).

6. M. Kaczmarek, R. W. Eason, and I. Mnushkina, “The effectof doping and processing conditions on the opticalperformance of Rh:BaTiO3,” Appl. Phys. B 68, 813–817(1999).

7. D. Kip, “Photorefractive waveguides in oxide crystals:fabrication, properties, and applications,” Appl. Phys. B 67,

131–150 (1998).

8. S. Brülisauer, D. Fluck, P. Günter, L. Beckers, and C.Buchal, “Photorefractive effect in proton-implanted Fe-doped KNbO3 waveguides at telecommunicationwavelengths,” J. Opt. Soc. Am. B 13, 2544–2548 (1996).

9. D. Haertle, G. Caimi, A. Haldi, G. Montemezzani, P.Günter, A. A. Grabar, I. M. Stoika, and Y. M. Vysochanskii,“Electro-optical properties of Sn2P2S6,” Opt. Commun. 215,333–343 (2003).

0. A. Grabar, M. Jazbinsek, A. N. Shumelyuk, Y. M.Vysochanskii, G. Montemezzani, and P. Günter,“Photorefractive effects in Sn2P2S6,” in PhotorefractiveMaterials and Their Applications II, P. Günter and J.-P.Huignard, eds. (Springer, 2007), pp. 327–362.

1. T. Bach, M. Jazbinsek, G. Montemezzani, P. Günter, A. A.Grabar, and Y. M. Vysochanskii, “Tailoring of infraredphotorefractive properties of Sn2P2S6 crystals by Te and Sbdoping,” J. Opt. Soc. Am. B 24, 1535–1541 (2007).

2. R. Mosimann, P. Marty, T. Bach, F. Juvalta, M. Jazbinsek,P. Günter, and A. A. Grabar, “High-speed photorefractionat telecommunication wavelength 1.55 �m in Sn2P2S6:Te,”Opt. Lett. 32, 3230–3232 (2007).

3. A. Guarino, M. Jazbinsek, C. Herzog, R. Degl’Innocenti, G.Poberaj, and P. Günter, “Optical waveguides in Sn2P2S6 bylow fluence MeV He+ ion implantation,” Opt. Express 14,2344–2358 (2006).

4. D. Haertle, M. Jazbinsek, G. Montemezzani, and P. Günter,“Nonlinear optical coefficients and phase-matchingconditions in Sn2P2S6,” Opt. Express 13, 3765–3776 (2005).

5. B. Sturman, P. Mathey, H. R. Jauslin, S. Odoulov, and A.Shumelyuk, “Modeling of the photorefractive nonlinearresponse in Sn2P2S6 crystals,” J. Opt. Soc. Am. B 24,1303–1309 (2007).

6. S. Odoulov, A. Shumelyuk, U. Hellwig, R. Rupp, A. A.Grabar, and I. Stoyka, “Photorefraction in tinhypothiodiphosphate in the near infrared,” J. Opt. Soc. Am.B 13, 2352–2360 (1996).

7. A. N. Shumelyuk, A. Hryhorashchuk, and S. G. Odoulov,“Coherent optical oscillator with periodic zero-� phasemodulation,” Phys. Rev. A 72, 023819 (2005).

8. A. N. Shumelyuk, K. Shcherbin, S. Odoulov, B. Sturman,E. Podivilov, and K. Buse, “Slowing down of light inphotorefractive crystals with beam intensity couplingreduced to zero,” Phys. Rev. Lett. 93, 243604 (2004).

9. A. Guarino and P. Günter, “Nondestructive method for thecharacterization of ion-implanted waveguides,” Opt. Lett.30, 2412–2414 (2005).

0. D. Fluck, J. A. Weiss, S. Brülisauer, and P. Günter, “Two-wave mixing of focused Gaussian beams in photorefractivewaveguides,” Opt. Lett. 19, 2080–2082 (1994).

1. N. V. Kukhtarev, V. B. Markov, S. G. Odoulov, M. S. Soskin,and V. L. Vinetskii, “Holographic storage in electro-opticcrystals. II. Beam coupling–light amplification,”Ferroelectrics 22, 961–964 (1979).