We propose to encode optical information through the localized depoling of polar chromophores in thin films of grafted polymeric materials with a femtosecond near IR laser source. This disorientation is promoted through the photoisomerization of the azo-dye component induced by a two-photon absorption process. We show that the resulting localized loss in second harmonic generation efficiency can be exploited in data storage applications. The low irradiation powers used allow for a recycling by reheating and repoling the films leading to a rewritable system.
©2006 Optical Society of America
Azo-dye doped or side-chain polymers have attracted tremendous attention over the years because of their photophysical properties which can be exploited in many non linear optical applications [1–3]. In this context, azo-doped poly(methyl methacrylate) (PMMA) polymer film is a model system which has been widely exploited for optical switching , holographic storage [5,6] or optical memories  where the required order of the chromophores can be achieved not only by applying an electric field but also via optical poling . In the field of optical data storage, the crucial parameter is spatial resolution. This need has been specifically addressed by the use of femtosecond laser sources, which, because of their high peak power, are able to give rise to multiphoton processes localized within a sub-wavelength volume in the vicinity of the focal point. This has been achieved by Maeda and coworkers  who performed orientational hole burning through two-photon absorption in films of poly(methyl methacrylate) doped with Disperse Red 1 (DR1) that can subsequently be detected through confocal differential reflexion microscopy. A more sophisticated approach has recently been proposed by the Zyss group  which encodes information by an all-optical poling technique in which the angles of polarization of the two irradiating fields are varied. The resulting spatial changes in the symmetry of the quadratic susceptibility tensor lead to a modulation of the detected SHG intensity when scanning the sample with the IR beam alone thus using a nonlinear optical phenomenon for the read-out stage as well.
One possible way to combine the advantages of these two distinct techniques would consist in using two-photon isomerisation in a photo-assisted poling scheme followed by a simple SHG read-out stage. However, the critical step would remain the orientation of chromophores in a small volume and over a limited time span. Therefore, we propose to start out with the even simpler method of writing optical data into a previously corona poled film by locally disorienting the polar order, now using two photon isomerisation to randomize the initial orientation of the chromophores. Again, data retrieval will be performed by monitoring SHG intensity while scanning the sample with an IR beam. In our approach, the information is being encoded into the succession of localized areas which have been disordered or not. This takes advantage of the fact that it is by far easier to induce disorder than to create order, and that the former is more irreversible. In addition, we will show that the intensity thresholds will be low enough to allow the erasing of the data by heating the sample and the rewriting of new data after repoling it.
2. Materials and experimental setup
The samples were prepared by spin coating on transparent microscope slides from a 20% wt solution of polymethyl-methacrylate (PMMA) grafted at 10% with DR1, in 1,1,2- trichloroethan solvent. The 400 nm thick film, measured by a Dektak profilometer, is placed under nitrogen atmosphere in the corona poling device. A high voltage (5kV) is applied to two 30 µm diameter tungsten wires placed on the top of the sample. The uniformly poled 2×2 cm2 sample is then placed on X, Y, Z, θ stages and in the focal point of a ×50 objective microscope (numerical aperture=0.45 yielding an Airy disk 2.1 µm in diameter at 800 nm) as shown in Fig. 1. The angle θ, defined between the normal to the surface of the sample and the optical axis, is chosen to maximize the SHG signal. The X, Y, Z motorized displacement stages are controlled so as to keep the focal point in the plane of the sample. Fig. 1 also shows the scanning SHG imaging microscope set up. The exciting beam is provided by a Tsunami Ti:saphire tunable laser (670–1100 nm) with 120 fs pulse duration and 80 MHz repetition rate. The polarization and the power of the incident beam are adjusted with a half-wave plate and a Glan-Taylor polarizer. The light exiting the sample is collected by a second, identical, microscope objective and then directed to the input slit of a spectrometer. The spectrum is then recorded by a fast cooled CCD camera. A Schott BG 39 filter eliminates the fundamental wave. At a wavelength of 800 nm, when the power is set to 2 mW (corresponding to 25 pJ per pulse and 400W peak power), we obtain nearly uniform SHG signals over the surface of the corona poled samples.
3. Experimental results
A first series of measurements has been performed to follow the dynamics of SHG intensity loss at a fixed location on the sample while irradiating at 800 nm. The DR1 chromophore does not undergo one photon transitions at this wavelength but two-photon absorption is possible although the cross section is only about 100 GM . This feature is actually helpful here since it allows recording the SHG intensity decay curves shown in Fig. 2 in spite of the high repetition rate of the laser and the 40 ms integration time of the CCD camera. The evolution of the SHG signal has been recorded for several values of the incident power to illustrate the rate of disorientation of chromophore alignment that can be achieved.
The inset in Fig. 2 shows the typical spectrum of the SHG signal analyzed by the spectrometer. The signal to noise ratio is very satisfactory and the spectral width of this harmonic signal correlates well with the characteristics of the fundamental femtosecond pulses. Using the XYZ stages, the sample can be repositioned at regularly spaced intervals to execute a series of exposures for fixed durations while incrementing the IR power. This stage corresponds to writing an array of dots with increasing disorder among chromophores. Then, in a second stage, the sample area is continuously scanned at a high speed of displacement (20 µm/s) after the incident power has been lowered down to 2 mW in order to avoid, as far as possible, further isomerization of the azo-dyes while keeping a reasonable level of SHG signal.
In the top of Fig. 3, we present a 3-D reconstitution of this SHG signal of the sample after the recording process. The bottom of Fig. 3 shows a cross section of the depth in the contrast of SHG signal. In the remainder of this work, we have raised the intensity of the reading beam to 4 mW to be able to scan SHG images at a faster rate. In Fig. 4, a 20×20 array of dots separated by 10 µm has been written into the sample using incident powers of 50 mW (to the left of the dotted line) and 100 mW (to the right of the dotted line). The reading step is performed repeatedly by imaging the SHG signal at several scanning speeds and inter-lines spacing settings in order to increase the spatial resolution in steps, from Fig. 4(a) to Fig. 4(c). In addition, Fig. 4(d) shows a numerically enlarged 3D representation of the SHG data corresponding to the lower right hand corner of Fig. 4(b).
The results presented in Fig. 4 prove that a given patterned area can be imaged repeatedly without detectable losses in neither contrast nor resolution. Furthermore, the sample used for these scans was of below average quality as one can distinguish inhomogeneities in Fig 4(a) and 4(b). Nevertheless, the optical information retrieval performed extremely well, indicating that the proposed technique is rather immune to defects in the film.
Figure 5 shows a different 3D representation constructed using the SHG data from a sample showing a more uniform thickness, which had been written with the help of another microscope objective (NA: 0.85) to obtain a higher density of 1012 dots/cm2. Indeed, the SHG image neatly displays dots of 1µm in radius, separated by 4.5 µm. The high quality of this picture can also be ascribed to the writing power of 25 mW which, combined with an exposure time of 200 ms, provides a good contrast to SHG background intensity.
To assess the impact of the writing process on the film morphology, we have carried out a separate series of measurements where a grating has been burned with a low incident power of 10 mW. As shown in Fig. 6(a), the test strip has been framed by a groove burned in at 105 mW.
This marking allows us to remove the sample from the holder for other characterization techniques such as optical microscopy as shown in Fig. 6(b), and to put it back on the SHG microscopy set-up as in Fig. 6(c) and 6(d). In optical microscopy, the framing groove is clearly visible while the grating is not apparent. The writing process at 10 mW has not lead to material transport inducing local changes in the refractive index, in contrast to the burning of the frame. To verify that the written grating can be erased, the sample has been heated and poled again under the same experimental conditions as the first time. Fig. 6(c) shows that the thus erased strip has recovered its capacity to generate a uniform SHG signal and is ready to undergo a second writing procedure, the results of which are displayed in Fig. 6(d). The new grating has been written with a different pitch on purpose, to show that new optical information can be encoded after the erasing procedure. This demonstrates that our technique could be the basis of a erasable/rewritable storage device. Other trials have shown us that writing at powers between 20 and 60 mW leads to changes which are visible in optical microscopy but tend to disappear during the erasing stage. At powers exceeding 60 mW the markings tend to be permanent.
In this study, we take advantage of the efficiency of SHG microscopy around 800 nm in imaging thin corona-poled films of a DR1-PMMA copolymer. This technique allows to both characterize (at low incident powers) and modify (at higher irradiations) the SHG efficiency on a sub-micron scale. Indeed, at these wavelengths we benefit from near resonance SHG and from the finite two-photon absorption cross section to induce photo-isomerization of the azo-dye to locally randomize the polar order of the chromophores. Both processes being well localized in the vicinity of the focal point because of their quadratic dependence in intensity, they can be associated to provide a simple technique towards optical storage of information. We have also demonstrated that this storage can be performed at such levels of intensities that the polymeric surface remains largely unaffected. This opens the way for a rewritable device since the film can be erased by repoling to recover its initial uniform SHG efficiency which can be used to store new data.
References and links:
1. H. Rau, “Photoisomerization of azobenzenes,” in Photochemistry and Photophysics, J. F. Rabek, ed. (CRC, Boca Raton, FL1990), 2, 119–141.
3. S. K. Yesodha, C. K. S. Pillai, and N. Tsutsumi, “Stable polymeric materials for non linear optics: a review based on azobenzene systems,” Prog. Polym. Sci. , 29, 45–74 (2004). [CrossRef]
4. A. Donval, E. Toussaere, R. Hierle, and J. Zyss, “Polarization insensitive electro-optic polymer modulator,” J. Appl. Phys. 87, 3258–3262 (2000). [CrossRef]
5. W. Zhang, S. Bian, S. I. Kim, and M. G. Kuzyk, “High-efficiency holographic volume index gratings in DR1-dye-doped poly(methyl methacrylate),” Opt. Lett. 27, 1105–1107 (2002). [CrossRef]
6. R. Birabassov, N. Landraud, T. V. Galstyan, A. Ritcey, C. G. Bazuin, and T. Rahem, “Thick dye-doped poly(methyl methacrylate) films for real-time holography,” Appl. Opt. 37, 8264–8269 (1998). [CrossRef]
7. M. Maeda, H. Ishitobi, Z. Sekkat, and S. Kawata, “Polarization storage by nonlinear orientational hole burning in azo dye-containing polymer films,” Appl. Phys. Lett. 85, 351–353 (2004). [CrossRef]
8. G. Xu, Q. G. Yang, J. Si, X. Liu, P. Ye, Z. Li, and Y. Shen, “Application of all-optical poling in reversible optical storage in azopolymer films,” Opt. Commun. 159, 88–92 (1999). [CrossRef]
9. S. Bidault, J. Gouya, S. Brasselet, and J. Zyss, “Encoding multipolar polarization patterns by optical poling in polymers: towards nonlinear optical memories,” Opt. Express 13, 505–510 (2005), http://www.opticsexpress.org/abstract.cfm?id=82388 [CrossRef] [PubMed]