Optical storage in LiNbo3 : Fe with selective erasure

capability

J.-P. Huignard, J.-P. Herriau, F. Micheron

To cite this version:

J.-P. Huignard, J.-P. Herriau, F. Micheron. Optical storage in LiNbo3 : Fe with se-lective erasure capability. Revue de Physique Appliquee, 1975, 10 (6), pp.417-423.<10.1051/rphysap:01975001006041700>. <jpa-00243939>

HAL Id: jpa-00243939

https://hal.archives-ouvertes.fr/jpa-00243939

Submitted on 1 Jan 1975

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinee au depot et a la diffusion de documentsscientifiques de niveau recherche, publies ou non,emanant des etablissements d’enseignement et derecherche francais ou etrangers, des laboratoirespublics ou prives.

417

OPTICAL STORAGE IN LiNbO3 : FeWITH SELECTIVE ERASURE CAPABILITY

J.-P. HUIGNARD, J.-P. HERRIAU and F. MICHERON

Thomson CSF, Laboratoire Central de Recherches, BP 10, 91401 Orsay, France

Résumé. - Après avoir analysé les différentes causes de déplacements macroscopiques des por-teurs de charges photoexcités dans LiNbO 3, on calcule les variations d’indice photoinduites corres-pondantes, pour les cas du cristal en circuit ouvert et en court-circuit. On montre dans le second casqu’un effacement complet ou sélectif peut être obtenu par une méthode de soustraction cohérente.

Abstract. - After having analyzed the different causes of macroscopic displacements of photo-carriers in LiNbO 3, the corresponding photoinduced changes of refractive index are computed forthe short-circuited and open circuited crystal. It is shown in the second case that total or selectiveerasure can be performed, using a coherent substraction technique.

REVUE DE PHYSIQUE APPLIQUÉE TOME 10, NOVEMBRE 1975, PAGE

1. Introduction. - Three dimensional storage usingphase hologram recording in transparent media is atechnique which allows high storage density andcapacity, since the minimum bit size may be as small asthe cube of the optical wavelength [1]. Photosensitiveelectrooptics belong to the most promissing class ofthree dimensional optical storage materials [2] : illumi-nation causes mainly a change in refractive index, theirspatial resolution is diffraction limited, they don’t

require any revealing process and they are reusable.Some of them require low recording light energy

(100/JlJ/cm2 for K(TaNb)03 using a two photons pro-cess [3], 3 mJ jcm2 for Sro,75Bao,25Nb206 [4]), or showstrongly assymetric recording-erasure cycles, whichallow multiple hologram superimpositions (more than500 holograms superimposed in LiNb03 : Fe [5]).

In the most cases, the storage time is found experi-mentally equal to the dielectric relaxation time (someseconds to some months), and this is the reason whyphotoinduced change of refractive index is attributedto photoinduced electric fields, and not to polar defectsreorientations or photoinduced dipole moments.

Indeed, it was demonstrated that photocarriers excitedfrom impurity centers can move on macroscopic dis-tances in these crystals before being trapped, whichgenerate space charge fields and refractive index

changes via the electrooptic effect.The carrier transport processes can be isotropic

diffusion [6], or anisotropic diffusion in polar materialswhich give rise to a bulk photovoltaic effect [7], or driftunder the influence of an applied field [8, 9], or combi-nations of these three effects. They generate a current

in an external circuit connected to the crystal underillumination such as :

The diffusion current exists only for non uniformilluminations I(x), which cause a gradient of photo-carriers concentration n(x).The photovoltaic current is proportionnal to the

absorbed light power al(x) and can be measured onlyparallel to the polar axis of the crystal. The coefhcient Kis odd function of the spontaneous polarization, anddepends upon the nature and value of the impurityions. The third term in the current J accounts for the

crystal conductivity and photoconductivity (photo-carriers drifts) in presence of the electric field E(x).The field E(x) inside the crystal is given by the elec-

trical displacement D

and the photoinduced refractive index change is

obtained using the appropriate electrooptic coefficient.

Using this model, we derive the operating conditionswhich provide the best use of the full dynamic range inrefractive index changes, especially for hologramssuperimpositions and coherent erasures, in Fe dopedLiNb03 crystals (0.015 % Fe, 3 mm thickness,reduced in Ar atmosphere for 50 % absorption at = 514 nm).

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/rphysap:01975001006041700

418

2. Uniform illuminations of Fe doped LiNbO3crystals. - The LiNbO( crystal is first uniformelyilluminated, and the short circuit current ici is measuredbetween two electrodes perpendicular to the polar caxis (Fig. la). Using eq. (1), Zce is given by

FIG. 1. - Uniform illumination of iron doped - LiNb03crystal and corresponding photocurrents (icc); photoinducedspace charge field (Esc) ; and photoinduced index change :Fig. la : short circuit configuration ; Fig. lb : open circuit

configuration.

where A is the electrodes area. Since no photovoltagecan be generated in this short circuited crystal, it is

expected from the previous model that the refractiveindex remains unchanged. The refractive index wasmeasured using a low power He-Ne laser and crossedpolarizers, and no index change was observed underilluminations by an Ar laser at = 514 nm, up toincident powers of 100 mW/cm’ (the experimentalarrangement limits the measure to index changes largerthan 10-6). At higher powers, or long exposure times,an increasing scattering is observed in the transmittedbeams, whereas no scattering effect can be detectedwith the same energy density, using an incoherent lightsource. This anomalous scattering effect under cohe-rent illumination can be attributed to surface defects orbulk inhomogeneities : the coherent incident light beamis slighly scattered by the defects at the beginning ofillumination, and the randomly scattered beams inter-fere inside the crystal. They cause a randomly nonuniform repartition of illumination, which is thereforerecorded as new refractive index inhomogeneities.

This effect limit to some j/CM3 the energy densitywhich can be absorbed during operations at roomtemperature. Nevertheless, the crystal recovers its lowscattering state after heating at 200 °C for someminutes : the increase in conductivity decreases thedielectric relaxation timer, therefore, complete erasureof all the recorded refractive index changes occurs byheating during times larger than i.

Considering now the arrangement of figure lb, wherethe crystal is uniformely illuminated in open circuit,i. e. J = 0. A space charge field Esc can be developpedby photovoltaic effect, which saturation value is givenby

where J is the conductivity under illumination. Anhomogeneous photoinduced change of refractive indexis now observed, as previously reported by Glass [7],and the same scattering effects appear as shown in theshort circuit configuration.

3. Non uniform illumination with holographic gra-tings. - Using the previous results, one can now

deduce the spatial index distributions during hologramgratings recording, for open circuit and short circuitconditions. The spatial illumination repartition is

where k = 2 03C0/039B, and 039B is the fringe spacing. Thefringes are perpendicular to the polar axis.

In the LiNb03 : Fe crystals used, both diffusionscurrent at A = 10 gm and photoconductivity currentcan be neglected compared to the photovoltaic current.This is not the case for LiNb03 : Fe crystals withhigher ratios Fe3+/Fe2+ (less reduced crystals) in whichphotoconductivity must be taken into account.

Therefore, eq. (1, 2) reduce to

open circuit configuration is described by the limitingcondition :

and solution for E(x, t) is therefore

Both continuous and modulated terms in the holo-

graphic grating are recorded : during hologram super-impositions, the different continuous terms are added,which causes an increase of the average change ofrefractive index (and would contribuate to saturate therefractive index change in photoconductive crystals).Optical erasure by uniform illuminations causes era-sure of the modulated component only, and increasesthe average change of refractive index.

In the short circuit configuration (Fig. 2b), the limit-ing condition is

solution of eq. (6) is therefore (A « crystal length 1)

The continuous term in the holographic grating is nomore recorded, and therefore, the short circuit confi-guration allows multiple hologram superimpositionand provides a true optical erasure. This will be shownin the next section, for the particular case of coherentoptical erasure.

419

FIG. 2. - Spatial index change of refractive index after holo-gram grating recording : Fig. 2a : open circuit configuration ;

Fig. 2b : short circuit configuration.

4. Sélective erasure of superimposed volume holo-

grams. - Information pages can be superimposedat the same location of the storage medium by slightlychanging the beam incidence outside of the Braggselectivity range. Erasure of such recorded hologramsis usually performed by uniform illumination of thestorage crystal at the recording wave-length (or smaller)or by heating. Both of these techniques cause a bulkerasure process which apply indiscriminately to thedifferent holograms and lead to limitations in appli-cations of such a storage process [10]. Selective erasureof any stacked hologram or information bit is demons-trated by using a coherent image substraction method.The storage crystal used for these experiments is iron

doped and reduced LiNb03-O.015 % Fe - whichshows a high resistance to optical erasure by uniformillumination - (assymetric Recording-Optical erasurecycle) [11 ]. The crystals having such a cycle are suitablefor multiple storage, since the recording beams of anew hologram cause a weak erasure of the previouslyrecorded ones in the same volume. The coherent era-sure concept applies to such crystals providing acompact system with a greatly increased storage densityand maintaining the versatility of erasure of any sta-cked hologram or bit in a page.

4.1 OPEN CIRCUIT RECORDING. - When recordingthe Fourier hologram of a transparent object havinga nearly uniform spectrum, the refractive index changeat spatial frequency k is deduced from eq. (8), in opencircuit recording conditions, using the linear electro-optic coefficients (Fig. 3a) :

no : bulk photoinduced index change developped byd. c. term of holographic pattern (low spatialfrequency),

bn : index modulation characteristic of fringes modula-tion (high spatial frequency).

FIG. 3. - Spatial index change of refractive index after record-ing-coherent selective erasure cycle : Fig. 3a : open circuit

configuration ; Fig. 3b : short circuit configuration.

When now recording a complementary index modu-lation : ô’ n such as bn + ô’ n = 0, one can erase thecorresponding modulation of the selected hologram.This is achieved by a second recording of the sameobject with introduction of oc shift on the referencewave. Light energy for coherent erasure cycle is

exactly the same as used for recording.The index change photoinduced during this second

recording takes the form :

After several recording coherent erasure cycles theaverage index change is continuously increasing(contribution of the term no), and this should be alimitation of the number of cycles for a material havinga limited index dynamic range ns (Fig. 3a).

4.2 SHORT CIRCUIT RECORDING. - From the analy-sis performed in the previous section, recording andselective erasure operations should be done with thecrystal short circuited. With such conditions eq. (10)shows that the index change is established around theoriginal value (Fig. 3b).

420

and by coherent erasure process, the new photo-induced change is :

After such a Recording-Coherent erasure cycle, theinitial index value is really recovered (Fig. 3b) and thenumber of cycles should not be limited. Nevertheless,the number of cycles is limited to a tenth by the scatter-ing effect, which increases the noise after each cycle.The non recording of the d. c. terms of hologrampatterns also provides an increase in the number ofsuperimposed images since in practicP a relatively lowfringe modulation is used for high quality imagereconstruction (m 1).

5. Expérimental results. - Complete erasure of aninformation page can be performed using the coherenterasure process. Selective erasure of any informationblock or bit can be also achieved by recording of apartially masked transparency with n shift. Thecommon parts of the two transparencies are erased.This corresponds the logical operation Exclusive OR.Experimental confirmation is demonstrated in figure 4(measured recording sensitivity : 250 mJ/cm2/r¡ = 1 %),which shows complete erasure, selective and only bit bybit erasure in a single recorded hologram.

FIG. 4. - Selective erasure in a single recorded hologram.

Three holograms of the same binary page have beensuperimposed by changing the reference beam inci-dence on the crystal. These angle changes are providedby step mechanical translation of reference beam andconverted into rotation by a lens (Fig. 5). The experi-mental sequence is the following :

- Recording of three stacked holograms of thesame data plane (Bragg angles 0,, 02, 83) (hologramdiameter : 1.7 mm ; Efficiency 11 = 1 % ; Beamratio : 13).- Selective erasure of one single bit in the page

(Fig. 6b).- Repositionning of reference beam at Braggs

angle e2 and selective erasure in the page (Fig. 6b).- Repositionning the reference beam on Bragg

angle 91, and complete erasure of the third hologram(Fig. 6b).

Figure 7 demonstrates the possibility of logical ope-rations between two binary transparencies A and B(metallic grids for avoiding phase distorsion). In theparticular situation where B is a uniform object trans-parency (B = 1) the reconstructed image has a reversedcontrast. The experimental sequence of figure 7 showsthe logical operation A + B.

6. Comments. - For these experiments, precau-tions must be taken to avoid mechanical vibrations ofthe components. The laser stability and n shift are

continuously controlled by projection of a magnificatedportion of the fringes on a vidicon with display on a TVmonitor. The phase shift has been obtained either withan Electro-optic modulator, or by switching a Babinetplate between two positions. The problem encounteredwith our crystals is photoinduced scattering whichrapidly degrades the signal to noise ratio of reconstruct-ed images. After several Recording-Coherent erasurecycles, we observe the scattered light at the same levelas that obtained by continuously reading one holo-gram with the reference beam during the same periodof time. This effect limit actually in our crystals thenumber of superimposed images an selective erasurecycles to about 10 and the potential large dynamicrange of the crystal is not used. As mentioned in [5]recording with heated could reduce the build up of thiscoherent scattering.From system consideration, the ideal material would

be a storage crystal with a completely dissymmetricalcycle and no photoinduced scattering. Informationto be erased should be readed on the photodetectormatrix and rewritten on the page composer for selec-tive erasure.

421

SELECTIVE ERASURE

FIG. 5. - System configuration for selective erasure and processing in stacked holograms.

422

SELECTIVE ERASURE IN STACKED HOLOGRAMS

FIG. 6. - Selective erasure of stacked holograms : Fig. 6a : images reconstruction from three stacked holograms ; Fig. 6b :

selective erasure in three stacked holograms with reference beam repositioning 03, 02, 01, single bit erasure (Bragg angle 93),partial erasure (Bragg angle 92), complete erasure (Bragg angle 01).

423

FIG. 7. - Logical operations between two binary transparencies A and B (operation A + B = A. B).

References

[1] HUIGNARD, J. P., MICHERON, F. and SPITZ, E., In Opticalproperties of solids, Ed. B. O. Seraphin (North Hol-land) 1975.

[2] GLASS, A. M., In : Photonics, Ed. M. Balkanski and P. Lalle-mand (Gauthier-Villars) 1975.

[3] VON DER LINDE, D., GLASS, A. M. and RODGERS, K. F.,Appl. Phys. Lett. 25 (1974) 155.

[4] THAXTER, J. B. and KESTIGIAN, M., Appl. Opt. 13 (1974) 913.[5] STAEBLER, D. L., BURKE, W. J., PHILLIPS, W. and AMODEI, J.

J., Appl. Phys. Lett. 26 (1975) 182.

[6] AMODEI, J. J., Appl. Phys. Lett. 18 (1971) 22.

[7] GLASS, A. M., VON DER LINDE, D. and NEGRAN, T. J.,Appl. Phys. Lett. 25 (1974) 233.

[8] MICHERON, F., ROUCHON, J. M. and VERGNOLLE, M., Appl.Phys. Lett. 24 (1974) 605.

[9] KRATZIG, E. and KURZ, H., To be published in Ferroelec-trics.

[10] D’AURIA, L., HUIGNARD, J. P., SLEZAK, Ch. and SPITZ, E.,Appl. Opt. 31 (1974) 788.

[11] STAEBLER, D. L. and PHILIPS, W., Appl. Opt. 31 (1974) 788.[12] HUIGNARD, J. P., HERRIAU, J. P. and MICHERON, F., Appl.

Phys. Lett. 26 (1975) 256.