Polarization holography reveals the nature of the grating ......Polarization holography reveals the...

Polarization holography reveals the nature of the grating inpolymers containing azo-dye

P.-A. Blanche a,*, Ph.C. Lemaire a, C. Maertens b, P. Dubois b,1, R. J�erome b

a Centre Spatial de Li�ege, Universit�e de Li�ege, Parc Scienti®que du Sart Tilman, Avenue du Pr�e-Aily, B-4031 Angleur-Li�ege, Belgiumb Center for Educational and Research on Macromolecules, Universit�e de Li�ege, Sart Tilman B6a, B-4000 Li�ege, Belgium

Received 4 July 2000; accepted 6 September 2000

Abstract

To study the origin of reversible holographic recording in three polymers containing the same azo-dye, we have

measured the di�raction e�ciency and analyzed the gratings characteristics for various writing beams polarizations.

The amplitude of the holographic grating, as well as the ratio between index and absorption modulations, have been

investigated by gratings shifting. The total amount of di�racted power and the di�raction e�ciency versus the reading

beam polarization has been measured by non-degenerated four waves mixing. These experiments have revealed that the

molecular mechanisms of holographic recording in the studied compounds are di�erent. The photoinduced orientation

of the chromophores is predominant for C6±C11±DMNPAA; so, in C11±C6±DMNPAA (DMNPAA: 2,5-dimethyl-4-

(p-nitrophenylazo)anisole), the refractive index variation comes from the presence of both trans and cis populations

generated by photoisomerization. The behavior of the PVK:DMNPAA is included between these extreme cases since

both phenomena act. Ó 2000 Elsevier Science B.V. All rights reserved.

Keywords: Polymer; Holography; Molecular orientation; Photoinduced birefringence

1. Introduction

Polymers containing azo-dye are known since along time for their possibilities of holographic re-cording [1,2]. The latter seems to be well under-stood and come from the reorientation or by

trans±cis photoisomerization of the molecules bylight polarization [3±5]. Molecular reorientationinduces phenomena such as photoinduced bire-fringence, dichroism, and allows to write polar-ization holograms [5±11]. Since these propertiesare of signi®cant interest for applications, azo-dyeare currently intensively studied.

In a previous paper [11], we have analyzed thein¯uence of the temperature on the photoinducedbirefringence as well as on the holographic re-cording of three polymers containing the sameazo-dye. Mathematical models have been devel-oped in order to ®t the transmitted and di�ractede�ciency versus temperature measurements as wellas to de®ne parameters able to characterize the

1 November 2000

Optics Communications 185 (2000) 1±12

www.elsevier.com/locate/optcom

* Corresponding author. Present address: Optical Sciences

Center, University of Arizona, Tucson, AZ 85721, USA. Tel.:

+1-520-621-2229; fax: +1-520-621-9610.

E-mail address: [email protected] (P.-A.

Blanche).1 Present address: SMPC, Universit�e de Mons-Hainaut, B-

7000 Mons, Belgium.

0030-4018/00/$ - see front matter Ó 2000 Elsevier Science B.V. All rights reserved.

PII: S00 3 0-4 0 18 (0 0 )0 09 7 5- 5

compounds. During this study, we have found thatone of the three polymers (C11±C6±DMNPAA)(DMNPAA: 2,5-dimethyl-4-(p-nitrophenylazo)ani-sole) has a strange behavior facing both others andits e�ciency cannot be ®tted properly by ourmodels. We have then postulated that the molec-ular behavior of the chromophores in this poly-mer could be di�erent from the one previouslyexpected. The present paper details our investi-gations to understand the molecular mechanismstaking place into these three polymers. The methodused is polarization holography.

The three polymers synthesized are poly[N-vi-nyl carbazole] (PVK) doped with 15 weight per-cent (wt.%) of DMNPAA and plasticized with 30wt.% of N-ethyl-carbazole (ECZ, the PVK mono-mer). Both others are copolymers constitutedof [x-(N-carbazolyl)alkylmethacrylate] and [4-(11-methacryloylalkyloxy)-2,5-dimethylphenyl](4-nitrophenyl)diazene where alkyl spacers lengthhas been set to be hexyl and undecyl for the C6±C11±DMNPAA and undecyl and hexyl for theC11±C6±DMNPAA. The proportion of the co-polymer is such that there is also 15 wt.% of azo-dye in these compounds. Their chemical structureis presented in Fig. 1. The glass transition tem-peratures (Tg) have been measured by di�eren-tial scanning calorimetry to be 15°C for thePVK:DMNPAA, 56°C for the C6±C11±DMN-PAA and ÿ20°C for the C11±C6±DMNPAA.Polymers and chromophore synthesis can befound elsewhere [12].

PVK:DMNPAA polymer is well known for itsphotorefractive e�ect when a little amount of tri-nitro¯uorenone sensitizer is added [13,14]. In ourexperiments, no photorefractive e�ect could beinduced in the polymers since there is no sensitizerto form a charge transfer complex with the car-bazole group present in the three compounds.

Samples have been made by pressing polymerpowder between two glass plates heated around20°C above the compound glass transition tem-perature. Thickness is determined by 70 lm spac-ers. No crystallization has been observed.

In polymers containing azo-dye, two mecha-nisms allow to write holograms: molecular reori-entation and presence of both trans and cis isomerpopulation. The fact that one or the other take

place in the compound depend on the moleculartime constants.

Molecular reorientation results of the combi-nation of three processes: the selective trans to cisphotoisomerization, the angular di�usion and thecis to trans relaxation [10,15±18]. We have plottedin Fig. 2 the angular distribution of both theground and excited levels of the chromophoresaccording to the process acting (Fig. 2 does notcome from any calculation but are based on Du-mont's model [17]). The excited level includesmolecules that has absorbed a photon, they areeither in the cis form or in another conformation(i.e. triplet state or photodecomposition). Mole-cules that have relaxed in the trans form but stillhaving vibrational energy allowing them to rotatein the matrix are considered to be in the excitedlevel. Thus, ground level only contains moleculesthat have not absorbed any photon since a certaintime and have dissipated their vibrational energy.

At rest, all the molecules are in the ground leveland the distribution is isotropic (Fig. 2(a)). When

Fig. 1. Molecular structure representation of the three com-

pounds studied in the order of their acronym given (n � 15 wt.%).

2 P.-A. Blanche et al. / Optics Communications 185 (2000) 1±12

polarized beam is switched on, selective photo-isomerization begins and molecule oriented alongthe polarization axis (0°) are more likely pumpedto the excited level. Levels become anisotropic(Fig. 2(b)). Angular di�usion is due to thermalagitation, it happens in both levels but at di�erentrates due to the di�erent molecular conformations.This mechanism tends to restore the isotropybroken by the optical pumping (Fig. 2(c)). Cis±trans relaxation occurs either by another photo-

absorption or by thermal relaxation. Thermal re-laxation is isotropic: the rate does not depend onthe molecular orientation (Fig. 2(d)). Multiplecycles of these three mechanisms could lead toanisotropy of the molecular distribution. Indeed,if angular di�usion in ground state is slower thanin excited state and slower than a excitation±relaxation cycle, this mechanism combined withmultiple selective photoexcitation and isotropicrelaxation modi®es the angular distribution of the

Fig. 2. Angular distribution of the ground and the excited levels illustrating the di�erent mechanisms taking place during illumination

of the chromophores with a linearly polarized light.

P.-A. Blanche et al. / Optics Communications 185 (2000) 1±12 3

molecules. Fig. 2(e) shows the distribution in suchconditions when the polarized light has beenswitched o� and the molecules have relaxed.

However, since the three above mentionedprocesses can have di�erent time constants, analternative mechanism can happen. If angulardi�usion is much more faster than excitation andrelaxation, angular distributions in both levels willbe smoothed rapidly and no anisotropy will ap-pear. The ®nal distribution, presented in Fig. 2(f),will be isotropic. In these conditions, the refractiveindex can change since there is two molecularpopulations (trans and cis), but there will be noanisotropic phenomena such as birefringence anddichroism.

To put forward the molecular mechanisms,speci®c polarizations of the writing beams havebeen used to record holograms in the samples.These polarizations lead to various interferencepatterns that we have tried to represent graphicallyin Fig. 3, together with the polarization of theincident beams. The precise calculation of thepatterns according to the incoming beams polar-ization as well as their incidence angle h can befound in Ref. [19]. The de®nition and the utility ofthe selected con®gurations are the following:

VV 8h (Fig. 3(a)). Both incident beams arepolarized linearly and their polarization vector isperpendicular to the incident plane. The interfer-ence pattern is modulated in intensity and consti-tuted by linear polarizations oriented like the onesof the writing beams. This con®guration will beused as a reference for the other polarizations.

HH h �0 (Fig. 3(b)). Incoming beams arelinearly polarized along the incident plane. The®gure obtained is quite similar to the one of theVV case with polarization vectors turned at 90°.This con®guration is useful to determine whetheror not there is surface grating caused by polymermigration. Indeed, this kind of grating has alreadybeen shown in azo-dye doped polymers and theiramplitude varies with the polarization orientation[2,20,21].

VH h �0 (Fig. 3(c)). Incident beams are lin-early polarized in perpendicular directions. So, oneis along the plane of incidence and the other isperpendicular to this plane. The resulting inter-ference pattern is the more complex. It is consti-

Fig. 3. Graphic representation of the interference patterns ob-

tained for various polarization con®gurations of the incident

beams. Di�erent dashes represent left or right circular polar-

izations. Numbers on the top of the graphs are relative to the

phase shift between the two incident waves. See text for more

details.


tuted by linear polarizations oriented at 45° fromthose of incident beams, separated by left and theright circular polarizations. It has to be noted thatthere is no intensity modulation in this case. Lin-ear polarizations will induce molecular orientationand/or presence of cis isomers according to themolecular ability since, circular polarizations donot produce any molecular orientation in the po-larization plane and could only isomerize thechromophores isotropically.

RR h �0 (Fig. 3(d)). Both incident beams arecircularly polarized in the same direction (right inthe present case). Pattern is modulated in intensitywith circular polarizations oriented as the one ofthe incoming beams. Only few anisotropy could beproduced by this pattern, but it can reveal presenceof trans and cis populations.

RL h �0 (Fig. 3(e)). One beam is right circu-larly polarized, the other is left circularly polar-ized. The interference pattern is only constitutedby linear polarizations that turn with the beamsphase. There is no intensity modulation. Isomer-ization is identical over all the pattern surface andso, cannot induce grating. However, anisotropy isproduced along di�erent axes. Only photoinducedorientation allows holographic recording with thiscon®guration.

2. Grating characterization by two waves coupling

Two beams coming from an argon laser(514 nm), linearly or circularly polarized, crosseach other in the sample to form an interference

pattern. The latter is recorded by the polymer ®lmas molecular modi®cations (isomerization or re-orientation). By shifting the interference patternaccording to the grating written into the sample,one obtains a modulation of the output beamsintensity. Mathematical process allows to calculatethe amplitude of the holographic grating, and todetermine if there is an absorption or an indexgrating. It is also possible to calculate the phaseshift between the recorded gratings and the inter-ference pattern. This method which has been in-troduced by Sutter and G�unter [22] and re®ned byWalch and Moerner [23] is really powerful sinceone can determine various parameters and detectgratings with very low di�raction e�ciencies (i.e.10ÿ5) due to the synchronous detection.

The setup used is presented in Fig. 4. The phaseshifting requested between the written grating andthe illumination pattern is produced by moving themirror re¯ecting one of the laser beam. The inci-dent angle of the laser beams is approximately 5°.Thus, we are in the conditions of small angles re-quired for obtaining the interfering patterns de-scribed above �h � 0�; but still in the Bragg regimesince no other di�raction order is generated.

In order to not induce thermal grating orchemical degradation of the azo molecules, theintensity of the argon beams has been set at 1 mW/cm2 [24]. Experiments have been done at ambienttemperature.

Our setup geometry is fully symmetrical,i.e. both laser beams intensities are identical, theirincidence angles are equal and the sample is per-pendicular to the bisector of the angle formed by

Fig. 4. Gratings shifting setup. B.S.: 50=50 beams splitter, S.F.: spatial ®lter, k=x: retardation wave plate, Pz: piezoelectric transducer.


the beams. We are also in the Bragg regime. Inthese conditions, when the interference pattern ismoved at the constant speed v along the sampleplane, the intensities of the output beams are givenby:

I1 � I0 exp

�ÿ ad

cos h

�1

�ÿ 2Ad cos uA

�� 2pvt

KG

�

� 2Pd sin uP

�� 2pvt

KG

��; �1a�

I2 � I0 exp

�ÿ ad

cos h

�1

�ÿ 2Ad cos uA

�� 2pvt

KG

�

ÿ 2Pd sin uP

�� 2pvt

KG

��; �1b�

where I0 is the incident intensity, a is the absorp-tion coe�cient of the sample, d is its thicknessand KG acts as the grating spacing. The 2pvt=KG

expression is the phase induced by the mirrordisplacement. A and P are, respectively, the dif-fraction amplitude of the absorption grating andof the index grating. They are de®ned by Eqs. (2a)and (2b)

A � pk cos h

Da; �2a�

P � pk cos h

Dn; �2b�

where k is the reading wavelength, Da the ab-sorption modulation and Dn the refractive indexmodulation.

As it can be seen in Eqs. (1a) and (1b), theoutput beams intensity will oscillate with thegrating shifting. In the case of a pure absorp-tion grating, I1 and I2 will be in phase. For an in-dex grating, they will be p phase shifted. Whenboth gratings exist in the sample, the sum and thedi�erence of I1 and I2 allow to calculate the am-plitude (A and P) as well as the phase (uA and uP )of the gratings:

I �� I1 � I2 � I0 exp

�ÿ ad

cos h

�

� 2

�ÿ 4Ad cos uA

�� 2pvt

KG

��; �3a�

I �ÿ� � I1 ÿ I2 � I0 exp

�ÿ ad

cos h

�� 0

�� 4Pd sin uP

�� 2pvt

KG

��: �3b�

2.1. Modulation amplitude

To perform the measurement, the interferencepattern has been let a su�ciently long time (severalminutes) to ensure the grating is fully recorded inthe sample. The grating has been shifted at a speedmuch faster than the writing time of the hologramto be sure to not erase the latter. Displacement isdone within a second. We have recorded the am-plitude of the output beams modulation.

Geometrical factors as well as absorption co-e�cient of the sample are introduced in Eqs. (3a)and (3b), it is then di�cult to determine the ab-solute amplitude of the holographic grating.Moreover, we do not need absolute measurementbut only relative values to compare the variouspolarization con®gurations.

The graphics shown in Fig. 5 present normal-ized amplitude of the holographic grating ac-cording to ®ve polarization con®gurations for thethree polymers. We understand by the ``amplitudeof the holographic grating'' the amplitude of themodulations recorded on the detectors, which isproportional to the grating amplitude if there isonly one grating recorded in the material (see Eqs.(1a) and (1b)). From these measurements, con-clusions are:

(1) Gratings amplitudes for the VV and HHcon®gurations are identical for the three com-pounds studied. So, there is no surface gratingwritten in these polymers. At least, in our experi-mental conditions (intensity and writing angle).

(2) The C6±C11±DMNPAA (Fig. 5(a)) shows aweak modulation in the RR con®guration. Thelatter generates a pattern only constituted by cir-cular polarizations (Fig. 3(d)). So, this con®gura-tion can only produce grating by changing thecis:trans population ratio, only minor orienta-tional e�ects can happen. The fact that there isonly very weak grating recorded in this con®gu-ration indicate that the holographic recording in


this polymer is driven mainly by the reorientationof the chromophores, exactly as expected.

(3) Conversely of the precedent case, for theC11±C6±DMNPAA sample (Fig. 5(b)), there mustbe an intensity modulation to record holograms.Indeed, the pure polarization patterns do not giveany output modulation (con®gurations VH andRL). So, there is no molecular reorientation in theC11±C6±DMNPAA. The holographic recordingcertainly comes from the presence of both transand cis populations that have di�erent refractiveindices.

(4) Concerning the PVK:DMNPAA (Fig. 5(c)),its behavior is intermediate. All con®gurations canbe recorded with approximately the same ampli-tude. We believe there must be two recordingmechanisms: one responsible of the modulationobserved with the RR polarization; the other, in-ducing grating with orthogonally polarized writingbeams.

2.2. Index and absorption gratings

In Fig. 6, the ratio between the amplitude of theabsorption grating and the one of the index grat-ing is plotted:

DI ��

DI �ÿ�� A

P: �4�

One can see that, if the index modulation pre-dominates for all the three compounds, there is anabsorption component that is not negligible invarious cases. The ratio between the gratings isquite similar for the VV, HH and RR con®gura-tions, those are the ones for which intensity mod-ulation exists. For the VH and RL polarizations,there is a signi®cant modi®cation: the ratio in-creases for the C6±C11±DMNPAA and for thePVK:DMNPAA since, there is no grating (so noratio) for the C11±C6±DMNPAA (see Section2.1).

The PVK:DMNPAA case is very interestingsince it brings information on the imaginary partof the trans and cis dipolar moments �Il�. RRcon®guration induces very weak absorption grat-ing. Now, predominant recording mechanism inthis con®guration is isomerization. So, one canconclude that Il of the cis form is quite the samethan the mean of the trans form. At the opposite,since in VH and RL con®gurations absorption

Fig. 5. Amplitude of the holographic grating according to the

polarization con®guration of the writing beams for the three

compounds studied.


gratings are more important, and these polariza-tions favor chromophore orientation, one can saythat longitudinal and transversal Il of the transisomers are very di�erent. This can also be notedby comparing the C6±C11±DMNPAA and theC11±C6±DMNPAA. For the latter, precedent ex-periment has shown that there is no molecularorientation, and there is only few absorptiongrating. So, trans and cis mean Il are similar. Forthe C6±C11±DMNPAA, the absorption grating ismore predominant and chromophores reorienta-tion has been observed. DMNPAA perpendicularand parallel trans Il are di�erent.

No temporal modi®cation of the ratio betweenthe gratings amplitude has to be mentioned. Thus,for a given compound, the phenomena responsibleof the index and the absorption variations have thesame time constants and, more likely, come fromthe same mechanism.

We have not measured any phase shifting be-tween the recorded gratings and the illuminationpattern: uA � uP � 0 or p. Both are local. For allthe compounds, uA � p since the movement of theinterference pattern decreases the total output in-tensity (I ��). So, the absorption grating is due tophotobleaching of the polymer.

3. Di�raction e�ciency measurement

Measurements of the di�raction e�ciency ac-cording to the various interference patterns as wellas the polarization angle of the reading beam havebeen carried out. The experimental setup is pre-sented in Fig. 7. Exactly the same writing beamsgeometry has been reused from the phase shiftingexperiments. The hologram is read in situ bymeans of a linearly polarized helium±neon laserbeam (633 nm). Its polarization axis can be turnedby means of a polarizer, in order to determine ifthe holographic grating written is isotropic or not.The incident angle of the HeNe beam is equal tothe Bragg angle.

We have plotted, in Fig. 8, the normalized dif-fraction e�ciency of the HeNe beam according toits polarization angle; this for the ®ve interferencepatterns. For each graphs, the di�racted intensityhas been normalized to its highest value. The

Fig. 6. Ratio between the amplitudes of the absorption and the

index gratings for ®ve polarization con®gurations of the inci-

dent beams.


maximum di�raction e�ciency measured is 3:8�10ÿ2 for the C6±C11±DMNPAA, 6:8� 10ÿ4 forthe C11±C6±DMNPAA and 1:34� 10ÿ3 for thePVK:DMNPAA.

The comparison of the three graphics presentedin Fig. 8 shows that the behavior of the di�ractione�ciency, according to the polarization con®gu-ration, is very di�erent from one compound toanother. However, we can distinguish the owncharacteristics of the interference patterns: thedi�raction e�ciency is modulated according to asine square for the HH and the VV con®gurationsin a symmetrical way; then, the di�racted ampli-tude is constant for the VH, RR and RL incidentpolarizations. These properties come directly fromthe polarization symmetry of the interferencepatterns that can be remarked at Fig. 3.

Speci®c information can be extracted by ana-lyzing the ratio of the di�racted intensity at 0° and90° in the VV and HH con®gurations. Indeed, ifthe molecular orientation is the only actingmechanism, we know that parallel to the writingbeam polarization, there is a depletion in the transangular distribution and thus a large modulationof the holographic grating. Instead, in the or-thogonal direction, chromophore can be orientedrandomly in the plane de®ned by the propagationbeams direction. The grating modulation is con-siderably reduced. The ratio between both dif-fracted intensities at 0° and 90° polarizations isimportant. That is what appears in the Fig. 8(a):for the C6±C11±DMNPAA compound, the dif-fraction e�ciency falls practically to zero whenwriting and reading polarizations are orthogonal.

When it is the presence of both trans and cispopulations that characterize the holographicgrating, the following reasoning can be done: inthe illuminated zones of the interference pattern,the cis isomers are present and their dipolar mo-ments is quite homogeneous; then, in the darkareas, trans isomers are randomly distributed. Thehologram is so more isotropic in polarizationcompared to what we have for the molecular re-orientation. Meanwhile, it must be noted that thetrans to cis isomerization is still selective and so,the angular distribution of the trans molecules isnot homogeneous. This explains the remaininganisotropy. It is exactly what is observed for theC11±C6±DMNPAA, where the curves VV andHH of Fig. 8(b) are only weakly modulated.

Once again, the case of the PVK:DMNPAAcompound is intermediate, since the di�racted in-tensity ratio is around 40%. We can conclude thatboth recording mechanisms are present: reorien-tation induced anisotropy but trans and cis dis-tributions are also smoothed by di�usion.

To analyze the di�raction e�ciency for eachpolymer, it is also convenient to plot the totalamount of di�racted energy. This can be calcu-lated by integrating the di�raction e�ciency overthe angles of the reading beam polarization (0±p),or by performing the di�raction e�ciency experi-ments with an unpolarized reading beam. In Fig.9, this quantity has been normalized, for eachpolymer, to the maximum value obtained.

Fig. 9(a) shows that, for the C6±C11±DMNPAA, the di�raction e�ciency in the con-®guration RL is higher than for the other cases.

Fig. 7. Experimental setup used to measure the di�raction e�ciency according to the polarization of the writing and reading beams.

B.S.: 50=50 beams splitter, S.F.: spatial ®lter, k=x: retardation wave plate, P: polarizer.


This is due to the fact that the interference patternfor these writing polarizations is constituted bylinear polarizations with axes turn according to thephase (Fig. 3(e)). This is the more e�cient con-®guration for molecular reorientation since the

Fig. 8. Di�raction e�ciency of the HeNe beam according to its

polarization angle and, for di�erent polarization con®gurations

of the writing beams.

Fig. 9. Total amount of energy di�racted according to the po-

larization of the writing beams.


latter takes place in all the holographic area. Itshould be noted that, in the VH mode, since po-larizations at p=2 and 3p=2 are circular, the indexmodulation is not necessary sinusoidal and dif-fraction is then reduced.

In the case of the C11±C6±DMNPAA (Fig.9(b)), one can see that the di�raction e�ciencyincreases in the RR con®guration compared to theVV and HH states. This polymer shows only iso-merization; the more e�cient con®guration is thenobtained when the polarization of the interferencepattern is circular (RR con®guration). Indeed,isomerization trans±cis is realized isotropically inall the directions of the polarization plan. In thecase of linear polarizations (VV and HH), mole-cules placed perpendicularly do not undergo thelight electric ®eld in¯uence.

For the PVK:DMNPAA (Fig. 9(c)), the dif-fraction e�ciency in VH con®guration is an orderof magnitude lower than in the other cases. Thiscan be interpreted as follows: since there are mo-lecular reorientation (because there is di�raction inRL case) and isomerization (since there is di�rac-tion in RR case), the VH interference pattern inthis compound induces some reorientation of thechromophores where the polarizations are linearand some isomerization where the polarizationsare circular (see Fig. 3(d)). Both gratings overlayand cancel their e�ects.

We have just seen that the molecular mecha-nisms of holographic recording can perfectlyexplain the di�erent behaviors of di�raction e�-ciency observed according to the writing polar-izations.

4. Conclusions

Phase shifting and the di�raction e�ciencyexperiments, realized according to various polar-ization of the writing beams, have shown, in a self-reliant way, that the holographic recording processinto the three compounds C6±C11±DMNPAA,C11±C6±DMNPAA and PVK:DMNPAA is dif-ferent although they contain the same chromo-phore. We have found that:

(1) For the C6±C11±DMNPAA, the photoin-duced reorientation of the azo-dye molecules acts

for the major part in the holographic recordingprocess. The presence of cis isomers populationonly plays a minor role. This is the well knownholographic recording process for azo-dye dopedpolymers.

(2) In an opposite way, the C11±C6±DMNPAAcompound does not show any molecular reorien-tation, and the anisotropy is very weak. Holo-grams recording comes from the isomerization ofthe chromophores. This is the presence of bothtrans and cis isomer populations that induces in-dex change. This explains why we have previouslyobserved so many discrepancies between thephotoinduced birefringence and the di�raction ef-®ciency for this polymer [11].

(3) The analysis of the PVK:DMNPAA poly-mers is really interesting. It reveals an ``interme-diary'' behavior: the presence of cis isomers is asimportant as the molecular reorientation of thetrans form for the holographic recording.

The comparison of these results lets us thinkthat the di�erent behaviors observed are due tothe di�erent glass transition temperatures (Tg) ofthe compounds. Indeed, we have discussed in theintroduction the two extreme cases that can beobserved: photoinduced anisotropy and photoin-duced isomerization. They can both be explainedby the three same mechanisms: excitation, di�u-sion and relaxation but with di�erent time con-stants. Thus, for the C6±C11±DMNPAA, whoseTg is above the ambient temperature, the angulardi�usion due to thermal agitation is low com-pared to isomerization and relaxation; moleculescan then be aligned by the light, what leads tophotoinduced anisotropy. The Tg of the C11±C6±DMNPAA is under the ambient temperature andthe thermal agitation can easily randomize themolecular distribution. So, di�usion is faster thanexcitation and/or relaxation and smoothes theangular anisotropy of the levels (see Fig. 2(c)).Only the e�ects of the populations are observable.For the PVK:DMNPAA whose Tg is near theambient temperature, di�usion also smoothes thedistribution but not fast enough, anisotropy staysobservable and both mechanisms can be detected.

We have shown that the Tg in¯uences muchmore the chromophores behavior than the chemi-cal link between the dye and the polymer main


chain. The temporal analysis of multi-wave-lengths dichroism has con®rmed all these con-clusions [10].

References

[1] C.M. Verber, R.E. Schwerzel, P.J. Perry, R.A. Craig,

Holographic recording materials development, Technical

Report No. NASA-CR-144991, National Aeronautics and

Space Administration, 1976; reproduction: N76-23544

National Technical Information Service, Spring®eld, VA,

USA.

[2] P.S. Ramanujam, N.C.R. Holme, L. Nikolova, R.H. Berg,

S. Hvilsted, E.T. Kristensen, C. Kulinna, A.B. Nielsen,

M. Pedersen, Erasable holographic storage in azobenzene

polyesters and peptides, in: S.A. Benton, T. Trout (Eds.),

Practical Holography XI and Holographic Material III,

SPIE Proceedings, vol. 3011, San Jose, 1997, pp. 319±326.

[3] A.M. Makushenko, B.S. Neporent, O.V. Stolbova, Re-

versible orientation photodichroism and photoisomeriza-

tion of aromatic azo compounds I: model of the system,

Optical Spectroscopy (USSR) 31 (1971) 295±299.

[4] G.S. Kumar, D.C. Neckers, Photochemistry of azoben-

zene-containing polymers, Chemical Review 89 (1989)

1915±1925.

[5] P.-A. Blanche, Ph.C. Lemaire, C. Maertens, P. Dubois,

R. J�erome, Polarised light induced birefringence in azo dye

doped polymer: a new model and polarised holographic

experiments, Optics Communication 139 (1997) 92±98.

[6] T. Todorov, L. Nikolova, N. Tomova, Polarization holog-

raphy. 1: a new high-e�ciency organic material with re-

versible photoinduced birefringence, Applied Optics 23

(1984) 4309±4312.

[7] S. Calixto, R.A. Lessard, Holographic recording and

reconstruction of polarized light with dyed plastic, Applied

Optics 23 (1984) 4313±4318.

[8] L. Nikolova, T. Todorov, M. Ivanov, F. Andruzzi,

S. Hvilster, P.S. Ramanujam, Polarization holographic

gratings in side-chain azobenzene polyesters with linear

and circular photoanisotropy, Applied Optics 35 (1996)

3835±3840.

[9] C. Wang, H. Fei, Y. Yang, Z. Wei, Y. Qiu, Y. Chen,

Photoinduced anisotropy and polarization holography in

azobenzene side-chain polymer, Optics Communication

159 (1999) 58±62.

[10] P.-A. Blanche, Ph.C. Lemaire, M. Dumont, M. Fischer,

Photoinduced orientation of azo dye in various polymers

matrices, Optics Letters 24 (1999) 1349±1351.

[11] P.-A. Blanche, Ph.C. Lemaire, C. Maertens, P. Dubois,

R. J�erome, Photoinduced birefringence and di�raction

e�ciency in azo dye doped or grafted polymers: theory

versus experiment of the temperature in¯uence, Journal of

Optical Society of America B 17 (2000) 729±740.

[12] C. Maertens, P. Dubois, R. J�erome, P.-A. Blanche, Ph.C.

Lemaire, Synthesis and polarized light induced birefrin-

gence of new polymethacrylates containing carbazolyl and

azobenzene pendant groups, Journal of Polymer Science

Part B: Polymer Physics 38 (2000) 205±213.

[13] B. Kippelen, N. Peyghambarian, S.R. Lyon, A.B. Padias,

H.K. Hall Jr., New highly e�cient photorefractive polymer

composite for optical storage and image-processing appli-

cations, Electronics Letters 29 (1993) 1873±1874.

[14] K. Meerholz, B.L. Volodin, Sandalphon, B. Kippelen, N.

Peyghambarian, A Photorefractive polymer with high

optical gain and di�raction e�ciency near 100%, Nature

371 (1994) 497±500.

[15] S. Morino, S. Machida, T. Yamashita, K. Horie, Photo-

induced refractive index change and birefringence in

poly(methyl methacrylate) containing p-(dimethyl-

amino)azobenzene, Journal of Physical Chemistry 99 (1995)

10280±10284 and references therein.

[16] M. Dumont, A common model for optical ordering of

photoisomerizable molecules, in: F. Kajzar, V.M. Agrano-

vich, C.Y.-C. Lee (Eds.), Photorefractive Organic Materi-

als. Science and Applications, High Technology, vol. 3,

NATO Adcanced Sciences Institutes Series, vol. 9, Kluwer,

Dordrecht, 1996, pp. 501±511.

[17] M. Dumont, A general model for optically induced

molecular order in amorphous materials via photoisomer-

ization, Nonlinear Optics 15 (1996) 69±72.

[18] S. Hosotte, M. Dumont, Photoassisted poling and orien-

tational relaxation of dye molecules in polymers. the case

of spiropyran, in: G.R. M�ohlmann (Ed.), Nonlinear

Optical Properties of Organic Materials IX, SPIE Proceed-

ings, vol. 2852, Denver, Colorado, 1996, pp. 53±63.

[19] T. Huang, K.H. Wagner, Coupled mode analysis of

polarization volume hologram, IEEE Journal of Quantum

Electronics 31 (1995) 372±390.

[20] P. Le®n, C. Fiorini, J.-M. Nunzi, Anisotropy of the

photoinduced translation di�usion of azo-dyes, Optical

Materials 9 (1998) 323±328.

[21] N.C.R. Holme, L. Nikolova, P.S. Ramanujam, S. Hvilsted,

An analysis of the anisotropic and topographic gratings in

a side-chain liquid crystalline azobenzene polyester, Ap-

plied Physics Letters 70 (1997) 1518±1520.

[22] K. Sutter, P. G�unter, Photorefractive grating in organic

crystal 2-cyclooctylamino-5-nitropyridine doped with

7,7,8,8-tetracyano-quinodimethane, Journal of the Optical

Society of America B 7 (1990) 2274.

[23] C.A. Walsh, W.E. Moerner, Two-beam coupling measure-

ments of grating phase in a photorefractive polymer,

Journal of the Optical Society of America B 9 (1992) 1642±

1647 � erratum.

[24] P.-A. Blanche, Ph.C. Lemaire, C. Maertens, P. Dubois, R.

J�erome, Limiting factor of the di�raction e�ciency in azo

dye doped polymers, in: R.A. Lessard (Ed.), Photopolymer

Device Physics, Chemistry, and Applications IV, SPIE

Proceedings, vol. 3417, Qu�ebec City, Qu�ebec, Canada,

1998.


Polarization holography reveals the nature of the grating ......Polarization holography reveals the...

Documents

Transcript of Polarization holography reveals the nature of the grating ......Polarization holography reveals the...