OPTICAL REVIEW Vol. 8, No. 5 (2001) 364-367

Time Character of Nanosecond Phase-Conjugate SignalS in an Erythrosin B-Doped Polylner Film Hiromasa TANAKA, Hirofumi FUJIWARA* and Kazuo NAKAGAWA Department ofMaterials Science and Engineering, Muroran Institute of Technology, 27-1, Mizumoto, Muroran, 050-8585 Japan (Received May 7, 2001 ; Accepted June 9, 2001)

The generating mechanisms of nanosecond phase-conjugate (PC) signal and the time delay of a backward pump beam, at which the maximum diffraction efficiency of the PC signal can be obtained, are investigated using degenerate four-wave mixing in an erythrosin B-doped polyvinyl alcohol film. The influence of the coherence time of a pulse source on the measurement of the mechanisms of the PC signal generation is also examined. The population gratings due to saturable absorption of dye molecules mainly contribute to the PC signal generation rather than thermal gratings. The maximum diffraction efficiency is obtained at zero time delay of the backward pump beam, at which the coherence peak is observed. Key words: degenerate four-wave mixing, dye-doped polymer films, saturable absorbing dye, nanosecond phase conjugation, coherence peak

l . Introductron Recently there has been some interest in phase conjugation by degenerate four-wave mixing (DFWM) for applications to real-time optical imag'e processing'sl-3) and phaseconjugate (PC) interferometries4' 5) because of its large spa-

tial bandwidths over a wide field of view. Organic dyedoped polymer films are one of the candidate materials as a phase conjugator, although various types of materials have

been used for producing a PC signal by DFWM. Compared with other nonlinear materials, such dye-doped films have

the probe beams can be precisely measured, if deconvolution is performed using the experimental data and the pulse source profile. On the other hand, in the region of tD < r*, the back-

ward pump beam plays two roles: one is a reading beam and the other is a writing beam which contributes to the grating formation by interfering with the probe beam. Therefore, the experimental data in this region involves information about not only the response time of the grating formation but also the dependence of the diffraction efficiency of the PC signal on the temporal coherence function of the pulse source. Tocho et al.10) and Costela et al.13) measured the temporal

the advantage of easy fabrication of large size elements with

coherence function of a pulse source in the experiments on

tens of um in film thickness. In addition, a wide variety of

PC signal generation using DFWM configuration, in which time delay was introduced between forward pump and probe beams, and the reflectivities of PC signals were measured

dye/polymer combinations enable us to develop functional phase conjugators suitable for many practical applications, in

which the temporal response of the PC signal is a factor as important as its reflectivity. These factors depend strongly on the mechanisms responsible for the PC signal generation. Saturable absorption6) and photoinduced anisotropy7) are the

as a function of the time delay. Eichler et al. 14) proposed a

new method for measuring the temporal coherence function of a pulse source. In this method, a transient grating was formed by two pulses crossing in a dye cell and the inten-

main mechanisms in the phase conjugation by dye-doped

sity of a diffracted probe pulse interacting incoherently with

polymer films using a CW Iaser. From a practical viewpoint, pulse sources are required for the quickness of information

the two pulses was measured as a function of the time delay introduced between the two pulses. In these experiments, the temporal coherence functions having a peak at zero time delay were obtained. This peak is known as the "coherence

processing. Several mechanisms such as saturable absorption 8) thermal effects,9~ll) electrostriction 12) etc. could con-

tribute to the nanosecond PC signal generation in dye-doped

peak" . 1 5)

films.

As far as we know, only a few studies have been reported on nanosecond PC signal generation in xanthene dye-doped polymer filmsl6) and bulk, 13) although dye-doped polymers

The mechanisms with fast response time can be investigated by measuring the diffraction efficiency of the PC signal

as a function of the time delay of a backward pump beam rel-

have great potential in working as phase conjugators. In this

ative to forward pump and probe beams in DFWM configura-

paper, we report on nanosecond PC signals in an erythrosin B-doped polyvinyl alcohol (EB/PVA) film. Our emphasis is on the mechanisms of nanosecond PC signal generation and the time delay of a backward pump bearn, at which the maximum diffraction efficiency can be obtained. In addition, the influence of the coherence time of a pulse source on the measurement of the mechanisms of the PC signal generation is

tion. In this experiment, the time delay of the backward pump

beam, at which a maximum diffraction efficiency can be obtained, is also determined. However, the relation between the time delay (tD) introduced in the backward pump beam and the coherence time (1;c) of a pulse source significantly affects

the measurement of the mechanisms of the PC signal generation. In the region of tD >> 7::., the time response of the grat-

examined.

ings produced by interference between the forward pump and

*E-mail: h-fuji @mmm.muroran-it.ac.jp 3 64

OPTICAL REVIEW Vol. 8, No. 5 (2001)

H. TANAKA et al.

3 65

2. Experimental The experimental setup for the DFWM process is illustrated schematically in Fig. I . An optical parametric oscilla-

tor (OPO) pumped with Q-switched Nd:YAG-1aser was used as a pulse source at 537nm with a linewidth of 0.2cm~1' this wavelength is the line center of erythrosin B and gives

a maximum reflectivity of PC signals. The pulsewidth is - 5 ns (FWHM) and the repetition rate is 10pulses/s. The spolarized laser beam (linearly polarized normal to the plane of the optical table) is first divided by a beam splitter (BS 1) into

M2

two beams. The reflected part is used as a probe beam (Epr).

This beam can be delayed by an optical delay line (TDL1) consisting of two prisms to coincide optical path length with

Frg. 1. Experimental setup for DFWM process. M1, M2, M3, mirrors; BS1, BS2, BS3, beam splitters; TDL1, TDL2, optical delay

lines; AP, aperture; PD, photodetector; Ef, forward pump beam; Eb, that of a forward pump beam (ED•backward Thepump transmitted part is beam; Epr, probe beam. then divided by a beam splitter (BS2) into two counterpropa-

gating pump beams (Ef, Eb). In the optical path of a backward pump beam (Eb), an optical delay line (TDL2) is inserted so that various time delays between Eb and Epr, Ef can be introduced. The angle between beams Epr and Ef is set at 10'. The reason for selecting this angle is that higher refiectivity

of the PC signal due to thermal effect can be obtained for

1 .2

~~

1

~ls

larger grating space. 1 1) An aperture of 2 mm diameter (AP) is

0.8

f

used to transmit only the central part of the probe beam. The

fluences of the two pump beams and the probe beam are set at 2.0mJlcm2 and 0.46 mJlcm2 respectively. The PC signal

~

0.6

l

fl

ji

ij

is detected as time-averaged power by a photodetector (PD) 0.4

calibrated with a pyroelectric detector.

An EB/PVA film was prepared as follows. PVA powder was dissolved in dimethyl sulfoxide (40'C) to which erythrosin B was added. PVA solution of 0.29 wt% in erythrosin

B concentration was selected. The EB/PVA film was made by pouring the solution onto a glass plate and drying it in an

oven, keeping it at a temperature of 34'C and a humidity of 50% for four days. Then, the EB/PVA film was baked in the oven at 100'C for I hour to remove the anisotropy induced when the PVA solution was spread over the glass plate. The

z

0.2

o

-5 o

50

1 ooo 1 500

Time delay

2000 2 500 3000 3soo

/ ps

resultant filrn was about 50 ,hm in thickness and 3.4 in optical density (product of the small-signal absorption coefficient

Fig. 2. Normalized diffraction efficiency of the PC signal as a function of the time delay of the backward pump beam relative to the forward pump and the probe beams. Mean values of the experimental data are shown by solid circles together with maximum and

and the film thickness) at 537 nm.

minimum values.

3. Results and Discussion a maximum at tD - 1200ps. The diffraction efficiencies of the PC signals were measured as a function of the time delay of the backward pump beam relative to the forward pump and the probe beams using the experimental setup for DFWM shown in Fig. I . The diffraction efficiency is defined here as the ratio of the PC signal power to the backward pump power. The experimental results are shown in Fig. 2 where the diffraction efficiencies

are normalized with the maximum mean value at tD = O. The maximum diffraction efficiency obtained in our experiments was O. 1 2%. Since the pulse peak of the backward pump beam coincides with those of the probe and the forward pump beams at the delay time of tD = O in the EB/PVA film, the PC signal can be observed even under the condition of tD < O. The experimental result seems to consist of two characteris-

The PC signal observed in the time delay region close to

tD=0 results from the DFWM process in which two pump beams play roles of both writing and reading beams. The DFWM process means here that the three beams incident on a dye-doped film interact almost coherently with dye molecules. Therefore, the time delay range where the PC signal can be generated by the DFWM process is closely related to the coherence time of the pulse source T:.. In the time delay region of tD < Tc' the diffraction efficiency may decrease with an increase in the time delay depending on the temporal coherence function of the pulse source. In addition, even if there exists a mechanism of PC signal generation with fast response time, the change in the diffraction efficiency reflecting

tic changes in the diffraction efficiency with the different time

the fast mechanism cannot be observed in our experiment, because of the smoothing effect by the convolution between the

delay dependence; one takes a sharp peak at tD = O and the

pulsewidth (-5 ns) and the DFWM process. Consequently,

other varies slowly with an increase in the time delay, taking

H. TANAKA et al.

366 OPTICAL REVIEW Vol 8 No 5 (2001) 1 .2

1 .2

1

111

1

8~

, l

e, . , l

,, ,l

0.8

{ 0.6

,

,

ifjllll

~~

',

0.8

,I

,

,

, ,,,, ,,

{

e

0.6

e

i{, fll

, ,, ,

e ,,

,,

0,4

0.4

" .db e .t

z

0.2

0,2

o

O

-2

-200 -150 -100 -50 O 50 100 150 200

O -1 O -1 O -50

so lo lo 2

Time delay / ps

Time delay I ps Fig. 3. Degree of mutual coherence of the pulse source measured

Fig. 4. Normalized diffraction efficiency of the PC signal. En-

using a Michelson-type interferometer.

largement of area close to tD = O in Fig. 2.

we infer that the change in the diffraction efficiency with a sharp peak corresponds to the "coherence peak" 15) rather than

mal effect was not observed. Therefore, we believe that the PC signals observed in our experiments result mainly from the population gratings due to saturable absorption of dye molecules rather than to thermal effects. Of course, the mechanisms contributing to the PC signal generation in dye-doped films depend strongly on the pump beam intensity, the pulse duration and the chemical and physical characteristics of dye

that reflecting the physical mechanism responsible for the PC signal generation.

In order to confirm this, the temporal coherence function of the pulse source was actually measured using a Michelson-

type interferometer. The coherence time is generally de-

o

fined as the time difference between two interfering beams, at which the contrast of interference fringes falls to 0.5. Accord-

molecules.

ing to this definition, the coherence time of the pulse source

4. Conclusions

is found to be - 100ps (Fig. 3). The enlarged figure close to tD = O in Fig. 2 is shown in Fig. 4 to compare precisely the coherence time with the decay of the PC signal. The time delay at which the normalized diffraction efficiency falls to

the EB/PVA film have been measured as a function of the

0.5 is - 100 ps. Therefore, it is clear that the peak at tD = O

reflects the temporal coherence function of the pulse source.

On the other hand, in the time delay region of tD >> T., the backward pump beam primarily plays the role of the reading beam of the diffraction gratings produced by the probe

and the forward pump beams. We attribute the PC signal varying slowly with an increase in the time delay to the pop-

ulation gratings resulting from saturable absorption of dye molecules. This is because the slow change in the diffraction efficiency takes the maximum value at the time delay of - 1 200 ps corresponding to the fluorescence lifetime of dye molecules (940 ps) actually measured for the BE/PVA film. The convolution effect could introduce the time-1ag between the fiuorescence lifetime and the time delay at which the PC signal takes the maximum diffraction efficiency. The thermally induced gratings may also contribute to the PC signal generation lo, I l, i3, 17) However, the thermal effect is

believed not to have been obvious within the time delay intro-

duced in our experiments, because the rise time of the thermal gratings is not fastl8) and the temperature coefficient of

The diffraction efficiencies of the PC signals generated by

time delay of the backward pump beam relative to the forward

pump and the probe beams. The maximum diffraction efficiency was obtained at the time delay of tD = O, at which the

coherence peak was observed. The population gratings due to saturable absorption of dye molecules were found to contribute mainly to the PC signal generation rather than the ther-

mal gratings, which were, in some cases, the dominant generating mechanism in dye-dispersed polymersl3) or liquids. I l)

This suggests the posibility that the maximum diffraction efficiency could be obtained at a time delay close to the fluorescence lifetime, if the coherence time of a pulse source is suffi-

ciently longer than the fiuorescence lifetime of dye molecules.

Finally, special care should be taken not to mistake the coher-

ence peak for the reflection of the physical mechanisms of PC signal generation. References 1) A. E. Chiou and P. Yeh: Opt. Lett. 11 (1986) 306. 2) D. Z. Anderson and D. M. Lininger: Opt. Lett. 12 (1987) 123. 3) H. Fujiwara, K. Nakagawa and T. Suzuki: Opt. Commun. 79 (1990) 6. 4) K. Nakagawa and H. Fujiwara: Proc. SPIE 954 (1988) I l.

Furthermore, clear evidence of the acoustic wavesl2, 13, 19) in

5) K. Nakagawa and H. Fujiwara: Opt. Commun. 70 (1989) 73. 6) Y. Silberberg and I. Ber-Joseph: Opt. Commun. 39 (1981) 265. 7) L. Nikolova, T. Todorov, N. Tomova and V. Dragostinova: Appl. Opt.

the EB/PVA film due to the electrostrictive effect or the ther-

27 (1988) 1598. 8) R. L. Abrams and R. C. Lind: Opt. Lett. 2 (1978) 94. [Erratum: 3 (1978)

the refractive index in a solid is smaller than that in liquids. 17)

205].

8, No. 5 (2001) OPTICAL REVIEW Vol.

H. TANAKA et al.

9) G. Martin and R. W. Hellwarth: Appl. Phys. Lett. 34 (1979) 371. 10) J. O. Tocho, W. Sibbett and D. J. Bradley: Opt. Commun. 37 (1981) 67.

15) H. J. Eichler, P. Gunter and D. W. Pohl: Laser-Induced Dynamic Grat-

1 1) H. J. Hoffman and P. E. Perkins: IEEE J. Quantum Electron. QE-22

16) R. Y. Choie, T. H. Barness, W. J. Sandle, A. D. Woolhouse and I. T.

(1986) 563. 12) K. A, Nelson, R. J. D. Miller, D. R. Lutz and M. D. Fayer: J. Appl.

17) H. Eichler, G. Enterlein, P. Glozbach, J. Munschau and H, Stahl: Appl.

Phys. 53 (1982) I 144. 13) A. Costela, J. M. Figuera, F. Florido, I. Garcia-Moreno and R. Sastre:

Opt. 11 (1972) 372. 1 8) C. O'Neill. B. Coulombe, P. Galarneau and M.-M. Denariez-Roberge:

Opt. Commun. I19 (1995) 265. 14) H. J. Eichler, U. Klein and D. Ianghams: Appl. Phys. 21 (1980) 215.

ings (Springer-Verlag, Berlin, 1986) Chap. 7, p. 213.

McKinnie: Opt. Commun. 186 (2000) 43.

Appl. Phys. B 51 (1990) 1 16. 19) C. K. Wu, P. Agostini, G. Petite and F. Fabre: Opt. Lett. 8 (1983) 67.

3 67