- Email: [email protected]
Molecular migration mechanism for laser induced surface relief grating formation
Synthetic Metals 115 (2000) 121±125
Molecular migration mechanism for laser induced surface relief grating formation C. Fiorini*, N. Prudhomme, G. de Veyrac, I. Maurin, P. Raimond, J.-M. Nunzi LETI (CEA-Technologies AvanceÂes), CEA Saclay, DEIN-SPE, Groupe Composants Organiques, 91191 Gif-Sur-Yvette Cedex, France
Abstract Direct and reversible holographic recording of surface-relief gratings in azo-dye polymers was recently evidenced using atomic force microscopy. Irradiation with an interference pattern of polarized laser beams was observed to lead to substantial mass-transport. In typical experiments, the wavelength of the laser was chosen to be near the absorption of the chromophores. The surface gratings have a negative amplitude which can be as large as twice the polymer ®lm thickness. The origin of such photo-driven mass transport is still unclear. We provide here experimental evidence of a migration of the chromophores from high to low intensity regions. It indicates that, in such doped or grafted guest-host systems, polymer chain migration is a result of the guest chromophores photoinduced movements. This is in good agreement with our recently developed model of azo-dye photoinduced translation diffusion. # 2000 Elsevier Science S.A. All rights reserved. Keywords: Azo-dye polymers; Holography; Photoinduced gratings; Molecular movements
1. Introduction Azo-dye aromatic polymers have been shown to offer interesting prospects for material engineering using light matter interactions [1±6]. It is now well-established that following excitation and cis±trans isomerization, thermal diffusion enables a molecular rotation within the matrix leading ®nally to a full reorientation of the dye molecules [3±5]. Molecular ordering can be optically controlled using appropriate beam combinations [7]. More recent results on photoinduced surface-relief gratings have also opened the way to molecular translation control using optical ®elds. It was shown that irradiation of such materials with an interference pattern of coherent light could induce not only an alignment of the chromophores throughout the volume of the material but also a controlled modi®cation of the ®lm surface, in conjunction with the light interference pattern [6,8,9]. Such result was quite unexpected as it implies a substantial mass transport occurring at temperatures much below the glass transition temperature. After the ®rst demonstration in the case of amorphous polymeric matrices grafted with azo-dyes, same observations were also reported in the case of liquid-crystalline
*
Corresponding author. Tel.: 33-1-69-08-52-34; fax: 33-1-69-08-76-79. E-mail address: [email protected] (C. Fiorini).
polymers [11] or sol±gel matrices [12]. In typical experiments, the wavelength of the laser was chosen to be near the maximum absorption of the chromophores, typically l 488 nm or 514 nm of an Argon laser, with intensities ranging from I 10ÿ1000 mW/cm2. The main characterization methods for photoinduced surface-relief grating have mainly consisted in diffraction ef®ciency measurements or atomic force microscopy (AFM) studies. Raman spectroscopy measurements were also developed to monitor the photoinduced changes at the molecular level following surface-relief formation [10]. Surface gratings have a negative amplitude: the position of maximum height corresponds to light intensity minima [13]. Modulation amplitude of the surface-relief gratings can be as large as twice the ®lm thickness, which corresponds to very high diffraction ef®ciencies for the visible Argon wavelengths. Heating above the glass transition temperature is mandatory for erasure of the surface-relief gratings. The in¯uence of experimental parameters such as laser intensity, beam polarization, grating spacing have been explored [9,10,14]. Several explanations were proposed such as ablation or swelling but none could explain the latest results, namely, the dependence of the grating amplitude on the laser polarization that excludes thermal effects due to light absorption as the main contribution to this process. Indeed thermal effects are not polarization dependent and p-polarized laser interferences were observed to produce surface gratings with a diffraction ef®ciency at least ten times larger
0379-6779/00/$ ± see front matter # 2000 Elsevier Science S.A. All rights reserved. PII: S 0 3 7 9 - 6 7 7 9 ( 0 0 ) 0 0 3 3 2 - 5
122
C. Fiorini et al. / Synthetic Metals 115 (2000) 121±125
than s-polarized ones. Additionally the process is reversible, which excludes ablation. Different models have been proposed to explain such phenomena [11,14±18]. Indeed, understanding of the relevant parameters for photoinduced mass-transport process is a key issue for optimization of opto-mechanical effects and further application to the realization of devices for photonic applications such as optical memories or 2D planar photonic microstructures used for instance as ®lters or couplers. After studying the in¯uence of the dye grafting on the ef®ciency of the recording process, we provide here experimental evidence of a migration of the chromophores from high to low intensity regions. This indicates that polymer chain migration is a result of the guest chromophores photoinduced movements, which is in good agreement with our recently developed model of azo-dye photoinduced translation diffusion. 2. Surface-relief grating formation: influence of the dye grafting onto the polymer chain In order to record surface-relief gratings, we made use of the now classical Lloyd mirror set-up initially developed by Rochon and coworkers [6]: the incident expanded beam from an Argon laser (l 514 nm; I 100 mW/cm2) is divided into two parts: ®rst half is directly incident onto the sample and second half re¯ects ®rst onto a metallic mirror rigidly held at 908 to the sample. This permits to reduce side effects due to mechanical instabilities throughout the writing process (a few tens of minutes). The incident angle is approximately 78, leaving a 2 mm periodicity of the grating. Samples were spin-coated ®lms of the azo-dye molecule Disperse Red 1 [4-(N-(2-hydroxyethyl)-N-ethyl-)amino-40 nitroazobenzene] (DR1) in poly(methyl-methacrylate) (PMMA). Experiments were performed both in grafted or guest-host systems. Because of phase segregation effects, concentration in guest-host systems is generally limited to 10% in weight. For this purpose, DRPR [4-(dibutylamino)40 -nitroazobenzene], a DR1 derivative showing enhanced solubility was synthetized, enabling a 30% dissolution. Different copolymers have been studied (DR1±MMA x= 100 ÿ x). There are obtained by free radical polymerization of a 100 ÿ x=x molar mixture of methyl methacrylate (MMA) and N-ethyl-N-(methacryloxyethyl)-40 amino-4-nitroazobenzene (a DR1 derivative). The glass transition temperature TG, measured by differential scanning calorimetry was found comparable in each system: T G 1058C for the doped systems (PMMA doped with 10% in weight DR1 and PMMA doped with 30% in weight DRPR) and T G 1308C for the grafted systems. A modulation amplitude up to 500 nm (peak to valley) was measured in the grafted system DR1±MMA 35/65 using atomic force microscopy. Using the same experimental conditions and comparable sample thicknesses, a much
weaker modulation (about 20 nm peak to valley) was measured in the case of the 30% in weight DRPR guest-host system (DRPR30) whereas no modulation could be detected in the other low concentrated 10% in weight DR1 guest-host system. Grafting of the azo-dye chromophores to the polymeric matrix is thus not a mandatory parameter for surfacerelief grating formation, although, the surface relief grating formation seems to be favored in the case of grafted systems. Other parameters such as the host polymer chain length were shown to have a great in¯uence on the recording process [14]. In order to study exclusively the in¯uence of the dye grafting onto the polymer chain, a grafted PMMA system corresponding to 30% in weight DR1 doping was synthetized (DR1G). Comparable glass transition temperature was measured in both cases of the grafted DR1G (T G 1108C) and doped DRPR30 (T G 1058C) systems. After spincoating, the two polymeric ®lms were shown to have almost the same thicknesses and absorption spectra. Fig. 1 shows the real-time growth of the photoinduced modulation recorded in the same experimental conditions in these two ®lms DR1G and DRPR30 using AFM. Although, the dynamics of the process appears comparable both in the grafted and doped systems, the surface-relief recorded after saturation was shown to be about seven times greater in the grafted than in the doped system. Dye grafting onto the polymer chain appears thus as an important parameter for improvement of the surface-relief grating formation. In the case of DR1±MMA copolymers grafted with different concentration of Disperse Red 1, the relative modulation depth Dh=2l, measured using AFM with Dh the height difference between peaks and valleys and l, the ®lm thickness, was independent on the chromophores concentration [13]. It indicates that, unlike the case of liquidcrystalline based systems [11], in such systems, the in¯uence of any cooperative intermolecular movements is negligible in the surface-relief grating formation process [13].
Fig. 1. Real-time growth of the photoinduced modulation recorded in the same experimental conditions in two identically concentrated dye-grafted (DR1G) and dye-doped (DRPR30) polymeric films: lDR1G 360 nm and DRPR30 ODDRIG 340 nm and ODDRPR30 514 nm 0:78; l 514 nm 0:76. Recording was performed using p-polarized Argon laser beams (l 514 nm) with I 100 mW/cm2. Surface-relief gratings were analyzed using AFM.
C. Fiorini et al. / Synthetic Metals 115 (2000) 121±125
123
Fig. 2. Optical microscopy analysis of surface-relief gratings recorded in two spin-coated PMMA based thin films using white illumination. Both cases correspond to identical surface-relief grating recording conditions. Whatever the reading beam polarization, the same transmission spectrum was observed in both systems. Left: DR1±MMA 35/65 copolymer system. A relative modulation of 100% was measured using atomic force microscopy. Right: PMMA doped with Disperse Red 1 (10% in weight). No surface modulation could be detected.
3. Experimental evidence for chromophores migration
4. Modeling
The main characterization methods for photoinduced surface relief gratings have mainly consisted in diffraction ef®ciency measurements or AFM studies. As a complementary characterizing tool aiming more particularly at a better study of the dye concentration following irradiation, optical transmission analyzed were performed. Fig. 2 shows the optical transmission images obtained under white illumination in the case of surface-relief gratings induced both in doped and grafted systems. As it is well known from photoinduced dichroõÈsm measurements, following excitation and cis±trans isomerization, molecules preferentially reorient in a plane perpendicular to the exciting light [3,4]. In order to take this effect into consideration, optical transmission analyzed were performed using different reading beam polarizations. Whatever the reading beam polarization, the same modulated transmission spectrum was observed, indicating an isotropic distribution of the chromophores orientation in the transmission microscopy conditions. Referring to Beer±Lambert law: OD ecl, variation of the optical density OD can result from either a variation of sample thickness l or a variation of chromophores concentration c. Indeed, the lifetime of the chromophores cis-form being quite short (from a few milliseconds to a few seconds) and the formation of photoinduced surfacerelief grating being reversible, no modi®cation of the chromophores extinction coef®cient E can be invoked. Most important is that modulated transmission are observed even in the case of the 10% in weight doped DR1/PMMA guesthost system where no thickness modulation could be detected. This evidences the existence of regions with different chromophores concentration, indicating that surface modulation is inherent to chromophores migration, molecules presumably pushing or pulling the polymeric host as motors when diffusing from high intensity to low intensity regions.
Such experimental evidence of chromophores migration is in good agreement with our recently developed model of photoinduced translation diffusion of the azo-dyes [14,15]. As illustrated in the insert of Fig. 1, we may indeed assume that following excitation and isomerization from trans to cis, molecules relax back to their stable Trans form after an average translation L parallel to the molecule axis. In such a model, surface effects are thus governed by molecular diffusion, azo-dye molecules playing the role of molecular motors (Fig. 3). Using this model, we have been able to reproduce typical experimental results. For instance, numerical simulations with this model can fairly well predict the in¯uence of the (s) or (p) polarized light on the grating diffraction ef®ciency [15], p-polarized laser interferences producing surface gratings with diffraction ef®ciency typically 10 times larger than s-polarized ones.
Fig. 3. Photoinduced caterpillar-like DR1 molecule motion following excitation. In the cis form, the molecule is bleached. When it relaxes back to the trans form, after orientation redistribution, it has undergone translation by an average amount L.
124
C. Fiorini et al. / Synthetic Metals 115 (2000) 121±125
Fig. 4. Surface-relief gratings recorded after 1 h irradiation through 3 mm periodicity binary chromium mask. The copolymer film was about 300 nm thick. Irradiation was performed with I 50 mW/cm2 at l 514 nm using a 458 polarization of the beam.
In order to get the order of magnitude of the maximum total diffusion length after recording of the surface-relief grating, we performed some experiments through a binary chromium mask. As schematically illustrated in Fig. 4, in order to avoid diffraction effects over propagation after irradiation through the mask, the azo-dye DR1±MMA copolymer was directly spin-coated onto the mask, AFM measurement being performed on the back side of the ®lm. The result obtained after 1 h irradiation of a 0.3 mm thick ®lm is given in Fig. 4. As shown in the AFM scan before irradiation (left part of Fig. 4), the square modulation is mainly due residual surface modulation of the copolymer ®lm following extra-thicknesses due to partial chromium deposition. In fact, photoinduced surface modulation is thus limited to the small pits and peaks appearing at places of intensity changes: intensity modulation being step-like, we may thus deduce that the maximum diffusion length corresponds to
the width of these photoinduced pits and peaks, which is of the order of 400 nm. 5. Single-beam experiments In situ dynamic studies of the chromophores concentration-grating formation was also performed using confocal microscopy. In such experiments, irradiation of the sample was performed using one single focused beam at the Argon laser wavelength l 514 nm. Following the previous experiments, a concentration grating was expected in conjunction with the Airy diffraction disc forming at the beam focus. Left and right parts of Fig. 5 correspond to the different cases of horizontal and nearly vertical writing beam polarizations. In both cases, recording of the confocal microscopy
Fig. 5. In situ recording of the confocal microscopy transmission images observed in a DR1±MMA 35/65 copolymer film, following single beam irradiation at the beam focus using I 400 nW/ mm2 at l 514 nm. Left and right pictures correspond to horizontal and nearly vertical beam polarization P, respectively.
C. Fiorini et al. / Synthetic Metals 115 (2000) 121±125
transmission images were made at different writing times using the same experimental recording parameters (I 400 nW/mm2, l 514 nm). As can be deduced from Fig. 5, transmission is much higher at the beam center, which con®rms that chromophores migrate from high to low intensity regions. Importantly, the concentration grating can only be observed along the direction of the writing beam polarization, giving further evidence of the strong polarization dependence of the surface-relief grating formation. This is in good agreement with the experimental photoinduced surface deformation results recently obtained by S. Bian and coworkers [19] under comparable singlebeam irradiation conditions. 6. Conclusion We have given experimental evidence that together with the surface-relief grating formation, a chromophores concentration grating could also be simultaneously evidenced optically. Such result indicates that surface-relief grating formation obviously follows chromophores migration from high intensity to low intensity regions, photoinduced molecular movements leading to polymer chain migration. This is in good agreement with our recently developed model of azo-dye photoinduced translation diffusion [13,16,17,20]. Most interesting is that it opens the way of molecular translation control using optical ®elds. Deeper understanding of the relevant parameters of molecular translation processes is now a key issue for optimization of such an opto-mechanical effect: indeed, one challenging issue in the ®eld of the design of devices for photonic applications would be to achieve complete manipulation of the molecular order. In this respect, such photoinduced molecular-motor effects represent an interesting route toward device nanostructuring. A more extensive study of photoinduced molecular movements should enable optimization of the molecular structure. Control of the relative weight of photoinduced translation and rotation processes [21] appears also as an important step to control such optical microstructure patterning.
125
Acknowledgements We thank Dr. R. Pansu from Ecole Nationale SupeÂrieure in Cachan for introducing us to confocal microscopy and kindly assisting us during our experiments. References [1] T. Todorov, L. Nikolova, N. Tomova, Appl. Opt. 23 (1984) 4309. [2] P. Rochon, J. Gosselin, A. Natansohn, S. Xie, Appl. Phys. Lett. 60 (1992) 4. [3] Z. Sekkat, M. Dumont, Appl. Phys. B 54 (1992) 486. [4] R. Loucif-Saibi, K. Nakatani, J.A. Delaire, M. Dumont, Z. Sekkat, Mol. Cryst. Liq. Cryst. 235 (1993) 251. [5] C. Fiorini, F. Charra, J.-M. Nunzi, P. Raimond, J. Opt. Soc. Am. B 14 (1997) 1984. [6] P. Rochon, E. Batalla, A. Natansohn, Appl. Phys. Lett. 66 (1995) 136. [7] A.-C. EtileÂ, C. Fiorini, F. Charra, J.-M. Nunzi, Phys. Rev. A 56 (1997) 3888. [8] D.Y. Kim, L. Li, J. Kumar, S.K. Tripathy, Appl. Phys. Lett. 66 (1995) 1166. [9] D.Y. Kim, L. Li, X.L. Jiang, V. Shivshankar, J. Kumar, S.K. Tripathy, Macromolecules 28 (1996) 8835. [10] F. LagugneÂ-Labarthet, T. Buffeteau, C. Sourisseau, J. Phys. Chem. 102 (1998) 2654. [11] T.G. Pedersen, P.M. Johansen, N.C.R. Holme, P.S. Ramanujam, Phys. Rev. Lett. 80 (1998) 89. [12] B. Darracq, F. Chaput, K. Lahlil, Y. Levy, J.P. Boilot, Adv. Mater. 10 (14) (1998) 1133. [13] C. Fiorini, N. Prudhomme, A.-C. EtileÂ, P. Lefin, P. Raimond, J.-M. Nunzi, Macromol. Symp. 137 (1999) 105. [14] C.J. Barrett, A.L. Natansohn, P.L. Rochon, J. Phys. Chem. 100 (1996) 8836. [15] C.J. Barrett, P.L. Rochon, A.L. Natansohn, J. Chem. Phys. 109 (1998) 1505. [16] P. Lefin, C. Fiorini, J.-M. Nunzi, Pure Appl. Opt. 7 (1997) 71. [17] P. Lefin, C. Fiorini, J.-M. Nunzi, Opt. Mat. 9 (1998) 323. [18] J. Kumar, L. Li, X.L. Jiang, D.Y. Kim, T.S. Lee, S.K. Tripathy, Appl. Phys. Lett. 72 (17) (1998) 2096. [19] S. Bian, L. Li, J. Kumar, D.Y. Kim, J. Williams, S.K. Tripathy, Appl. Phys. Lett. 73 (13) (1998) 1817. [20] M. Itoh, K. Harada, H. Matsuda, S. Ohnishi, A. Parfenov, N. Tamaoki, T.J. Yatagai, J. Phys. D: Appl. Phys. 31 (1998) 463. [21] D.B. Hall, A. Dhinojwala, J.M. Torkelson, Phys. Rev. Lett. 79 (1997) 103.
Molecular migration mechanism for laser induced surface relief grating formation C. Fiorini*, N. Prudhomme, G. de Veyrac, I. Maurin, P. Raimond, J.-M. Nunzi LETI (CEA-Technologies AvanceÂes), CEA Saclay, DEIN-SPE, Groupe Composants Organiques, 91191 Gif-Sur-Yvette Cedex, France
Abstract Direct and reversible holographic recording of surface-relief gratings in azo-dye polymers was recently evidenced using atomic force microscopy. Irradiation with an interference pattern of polarized laser beams was observed to lead to substantial mass-transport. In typical experiments, the wavelength of the laser was chosen to be near the absorption of the chromophores. The surface gratings have a negative amplitude which can be as large as twice the polymer ®lm thickness. The origin of such photo-driven mass transport is still unclear. We provide here experimental evidence of a migration of the chromophores from high to low intensity regions. It indicates that, in such doped or grafted guest-host systems, polymer chain migration is a result of the guest chromophores photoinduced movements. This is in good agreement with our recently developed model of azo-dye photoinduced translation diffusion. # 2000 Elsevier Science S.A. All rights reserved. Keywords: Azo-dye polymers; Holography; Photoinduced gratings; Molecular movements
1. Introduction Azo-dye aromatic polymers have been shown to offer interesting prospects for material engineering using light matter interactions [1±6]. It is now well-established that following excitation and cis±trans isomerization, thermal diffusion enables a molecular rotation within the matrix leading ®nally to a full reorientation of the dye molecules [3±5]. Molecular ordering can be optically controlled using appropriate beam combinations [7]. More recent results on photoinduced surface-relief gratings have also opened the way to molecular translation control using optical ®elds. It was shown that irradiation of such materials with an interference pattern of coherent light could induce not only an alignment of the chromophores throughout the volume of the material but also a controlled modi®cation of the ®lm surface, in conjunction with the light interference pattern [6,8,9]. Such result was quite unexpected as it implies a substantial mass transport occurring at temperatures much below the glass transition temperature. After the ®rst demonstration in the case of amorphous polymeric matrices grafted with azo-dyes, same observations were also reported in the case of liquid-crystalline
*
Corresponding author. Tel.: 33-1-69-08-52-34; fax: 33-1-69-08-76-79. E-mail address: [email protected] (C. Fiorini).
polymers [11] or sol±gel matrices [12]. In typical experiments, the wavelength of the laser was chosen to be near the maximum absorption of the chromophores, typically l 488 nm or 514 nm of an Argon laser, with intensities ranging from I 10ÿ1000 mW/cm2. The main characterization methods for photoinduced surface-relief grating have mainly consisted in diffraction ef®ciency measurements or atomic force microscopy (AFM) studies. Raman spectroscopy measurements were also developed to monitor the photoinduced changes at the molecular level following surface-relief formation [10]. Surface gratings have a negative amplitude: the position of maximum height corresponds to light intensity minima [13]. Modulation amplitude of the surface-relief gratings can be as large as twice the ®lm thickness, which corresponds to very high diffraction ef®ciencies for the visible Argon wavelengths. Heating above the glass transition temperature is mandatory for erasure of the surface-relief gratings. The in¯uence of experimental parameters such as laser intensity, beam polarization, grating spacing have been explored [9,10,14]. Several explanations were proposed such as ablation or swelling but none could explain the latest results, namely, the dependence of the grating amplitude on the laser polarization that excludes thermal effects due to light absorption as the main contribution to this process. Indeed thermal effects are not polarization dependent and p-polarized laser interferences were observed to produce surface gratings with a diffraction ef®ciency at least ten times larger
0379-6779/00/$ ± see front matter # 2000 Elsevier Science S.A. All rights reserved. PII: S 0 3 7 9 - 6 7 7 9 ( 0 0 ) 0 0 3 3 2 - 5
122
C. Fiorini et al. / Synthetic Metals 115 (2000) 121±125
than s-polarized ones. Additionally the process is reversible, which excludes ablation. Different models have been proposed to explain such phenomena [11,14±18]. Indeed, understanding of the relevant parameters for photoinduced mass-transport process is a key issue for optimization of opto-mechanical effects and further application to the realization of devices for photonic applications such as optical memories or 2D planar photonic microstructures used for instance as ®lters or couplers. After studying the in¯uence of the dye grafting on the ef®ciency of the recording process, we provide here experimental evidence of a migration of the chromophores from high to low intensity regions. This indicates that polymer chain migration is a result of the guest chromophores photoinduced movements, which is in good agreement with our recently developed model of azo-dye photoinduced translation diffusion. 2. Surface-relief grating formation: influence of the dye grafting onto the polymer chain In order to record surface-relief gratings, we made use of the now classical Lloyd mirror set-up initially developed by Rochon and coworkers [6]: the incident expanded beam from an Argon laser (l 514 nm; I 100 mW/cm2) is divided into two parts: ®rst half is directly incident onto the sample and second half re¯ects ®rst onto a metallic mirror rigidly held at 908 to the sample. This permits to reduce side effects due to mechanical instabilities throughout the writing process (a few tens of minutes). The incident angle is approximately 78, leaving a 2 mm periodicity of the grating. Samples were spin-coated ®lms of the azo-dye molecule Disperse Red 1 [4-(N-(2-hydroxyethyl)-N-ethyl-)amino-40 nitroazobenzene] (DR1) in poly(methyl-methacrylate) (PMMA). Experiments were performed both in grafted or guest-host systems. Because of phase segregation effects, concentration in guest-host systems is generally limited to 10% in weight. For this purpose, DRPR [4-(dibutylamino)40 -nitroazobenzene], a DR1 derivative showing enhanced solubility was synthetized, enabling a 30% dissolution. Different copolymers have been studied (DR1±MMA x= 100 ÿ x). There are obtained by free radical polymerization of a 100 ÿ x=x molar mixture of methyl methacrylate (MMA) and N-ethyl-N-(methacryloxyethyl)-40 amino-4-nitroazobenzene (a DR1 derivative). The glass transition temperature TG, measured by differential scanning calorimetry was found comparable in each system: T G 1058C for the doped systems (PMMA doped with 10% in weight DR1 and PMMA doped with 30% in weight DRPR) and T G 1308C for the grafted systems. A modulation amplitude up to 500 nm (peak to valley) was measured in the grafted system DR1±MMA 35/65 using atomic force microscopy. Using the same experimental conditions and comparable sample thicknesses, a much
weaker modulation (about 20 nm peak to valley) was measured in the case of the 30% in weight DRPR guest-host system (DRPR30) whereas no modulation could be detected in the other low concentrated 10% in weight DR1 guest-host system. Grafting of the azo-dye chromophores to the polymeric matrix is thus not a mandatory parameter for surfacerelief grating formation, although, the surface relief grating formation seems to be favored in the case of grafted systems. Other parameters such as the host polymer chain length were shown to have a great in¯uence on the recording process [14]. In order to study exclusively the in¯uence of the dye grafting onto the polymer chain, a grafted PMMA system corresponding to 30% in weight DR1 doping was synthetized (DR1G). Comparable glass transition temperature was measured in both cases of the grafted DR1G (T G 1108C) and doped DRPR30 (T G 1058C) systems. After spincoating, the two polymeric ®lms were shown to have almost the same thicknesses and absorption spectra. Fig. 1 shows the real-time growth of the photoinduced modulation recorded in the same experimental conditions in these two ®lms DR1G and DRPR30 using AFM. Although, the dynamics of the process appears comparable both in the grafted and doped systems, the surface-relief recorded after saturation was shown to be about seven times greater in the grafted than in the doped system. Dye grafting onto the polymer chain appears thus as an important parameter for improvement of the surface-relief grating formation. In the case of DR1±MMA copolymers grafted with different concentration of Disperse Red 1, the relative modulation depth Dh=2l, measured using AFM with Dh the height difference between peaks and valleys and l, the ®lm thickness, was independent on the chromophores concentration [13]. It indicates that, unlike the case of liquidcrystalline based systems [11], in such systems, the in¯uence of any cooperative intermolecular movements is negligible in the surface-relief grating formation process [13].
Fig. 1. Real-time growth of the photoinduced modulation recorded in the same experimental conditions in two identically concentrated dye-grafted (DR1G) and dye-doped (DRPR30) polymeric films: lDR1G 360 nm and DRPR30 ODDRIG 340 nm and ODDRPR30 514 nm 0:78; l 514 nm 0:76. Recording was performed using p-polarized Argon laser beams (l 514 nm) with I 100 mW/cm2. Surface-relief gratings were analyzed using AFM.
C. Fiorini et al. / Synthetic Metals 115 (2000) 121±125
123
Fig. 2. Optical microscopy analysis of surface-relief gratings recorded in two spin-coated PMMA based thin films using white illumination. Both cases correspond to identical surface-relief grating recording conditions. Whatever the reading beam polarization, the same transmission spectrum was observed in both systems. Left: DR1±MMA 35/65 copolymer system. A relative modulation of 100% was measured using atomic force microscopy. Right: PMMA doped with Disperse Red 1 (10% in weight). No surface modulation could be detected.
3. Experimental evidence for chromophores migration
4. Modeling
The main characterization methods for photoinduced surface relief gratings have mainly consisted in diffraction ef®ciency measurements or AFM studies. As a complementary characterizing tool aiming more particularly at a better study of the dye concentration following irradiation, optical transmission analyzed were performed. Fig. 2 shows the optical transmission images obtained under white illumination in the case of surface-relief gratings induced both in doped and grafted systems. As it is well known from photoinduced dichroõÈsm measurements, following excitation and cis±trans isomerization, molecules preferentially reorient in a plane perpendicular to the exciting light [3,4]. In order to take this effect into consideration, optical transmission analyzed were performed using different reading beam polarizations. Whatever the reading beam polarization, the same modulated transmission spectrum was observed, indicating an isotropic distribution of the chromophores orientation in the transmission microscopy conditions. Referring to Beer±Lambert law: OD ecl, variation of the optical density OD can result from either a variation of sample thickness l or a variation of chromophores concentration c. Indeed, the lifetime of the chromophores cis-form being quite short (from a few milliseconds to a few seconds) and the formation of photoinduced surfacerelief grating being reversible, no modi®cation of the chromophores extinction coef®cient E can be invoked. Most important is that modulated transmission are observed even in the case of the 10% in weight doped DR1/PMMA guesthost system where no thickness modulation could be detected. This evidences the existence of regions with different chromophores concentration, indicating that surface modulation is inherent to chromophores migration, molecules presumably pushing or pulling the polymeric host as motors when diffusing from high intensity to low intensity regions.
Such experimental evidence of chromophores migration is in good agreement with our recently developed model of photoinduced translation diffusion of the azo-dyes [14,15]. As illustrated in the insert of Fig. 1, we may indeed assume that following excitation and isomerization from trans to cis, molecules relax back to their stable Trans form after an average translation L parallel to the molecule axis. In such a model, surface effects are thus governed by molecular diffusion, azo-dye molecules playing the role of molecular motors (Fig. 3). Using this model, we have been able to reproduce typical experimental results. For instance, numerical simulations with this model can fairly well predict the in¯uence of the (s) or (p) polarized light on the grating diffraction ef®ciency [15], p-polarized laser interferences producing surface gratings with diffraction ef®ciency typically 10 times larger than s-polarized ones.
Fig. 3. Photoinduced caterpillar-like DR1 molecule motion following excitation. In the cis form, the molecule is bleached. When it relaxes back to the trans form, after orientation redistribution, it has undergone translation by an average amount L.
124
C. Fiorini et al. / Synthetic Metals 115 (2000) 121±125
Fig. 4. Surface-relief gratings recorded after 1 h irradiation through 3 mm periodicity binary chromium mask. The copolymer film was about 300 nm thick. Irradiation was performed with I 50 mW/cm2 at l 514 nm using a 458 polarization of the beam.
In order to get the order of magnitude of the maximum total diffusion length after recording of the surface-relief grating, we performed some experiments through a binary chromium mask. As schematically illustrated in Fig. 4, in order to avoid diffraction effects over propagation after irradiation through the mask, the azo-dye DR1±MMA copolymer was directly spin-coated onto the mask, AFM measurement being performed on the back side of the ®lm. The result obtained after 1 h irradiation of a 0.3 mm thick ®lm is given in Fig. 4. As shown in the AFM scan before irradiation (left part of Fig. 4), the square modulation is mainly due residual surface modulation of the copolymer ®lm following extra-thicknesses due to partial chromium deposition. In fact, photoinduced surface modulation is thus limited to the small pits and peaks appearing at places of intensity changes: intensity modulation being step-like, we may thus deduce that the maximum diffusion length corresponds to
the width of these photoinduced pits and peaks, which is of the order of 400 nm. 5. Single-beam experiments In situ dynamic studies of the chromophores concentration-grating formation was also performed using confocal microscopy. In such experiments, irradiation of the sample was performed using one single focused beam at the Argon laser wavelength l 514 nm. Following the previous experiments, a concentration grating was expected in conjunction with the Airy diffraction disc forming at the beam focus. Left and right parts of Fig. 5 correspond to the different cases of horizontal and nearly vertical writing beam polarizations. In both cases, recording of the confocal microscopy
Fig. 5. In situ recording of the confocal microscopy transmission images observed in a DR1±MMA 35/65 copolymer film, following single beam irradiation at the beam focus using I 400 nW/ mm2 at l 514 nm. Left and right pictures correspond to horizontal and nearly vertical beam polarization P, respectively.
C. Fiorini et al. / Synthetic Metals 115 (2000) 121±125
transmission images were made at different writing times using the same experimental recording parameters (I 400 nW/mm2, l 514 nm). As can be deduced from Fig. 5, transmission is much higher at the beam center, which con®rms that chromophores migrate from high to low intensity regions. Importantly, the concentration grating can only be observed along the direction of the writing beam polarization, giving further evidence of the strong polarization dependence of the surface-relief grating formation. This is in good agreement with the experimental photoinduced surface deformation results recently obtained by S. Bian and coworkers [19] under comparable singlebeam irradiation conditions. 6. Conclusion We have given experimental evidence that together with the surface-relief grating formation, a chromophores concentration grating could also be simultaneously evidenced optically. Such result indicates that surface-relief grating formation obviously follows chromophores migration from high intensity to low intensity regions, photoinduced molecular movements leading to polymer chain migration. This is in good agreement with our recently developed model of azo-dye photoinduced translation diffusion [13,16,17,20]. Most interesting is that it opens the way of molecular translation control using optical ®elds. Deeper understanding of the relevant parameters of molecular translation processes is now a key issue for optimization of such an opto-mechanical effect: indeed, one challenging issue in the ®eld of the design of devices for photonic applications would be to achieve complete manipulation of the molecular order. In this respect, such photoinduced molecular-motor effects represent an interesting route toward device nanostructuring. A more extensive study of photoinduced molecular movements should enable optimization of the molecular structure. Control of the relative weight of photoinduced translation and rotation processes [21] appears also as an important step to control such optical microstructure patterning.
125
Acknowledgements We thank Dr. R. Pansu from Ecole Nationale SupeÂrieure in Cachan for introducing us to confocal microscopy and kindly assisting us during our experiments. References [1] T. Todorov, L. Nikolova, N. Tomova, Appl. Opt. 23 (1984) 4309. [2] P. Rochon, J. Gosselin, A. Natansohn, S. Xie, Appl. Phys. Lett. 60 (1992) 4. [3] Z. Sekkat, M. Dumont, Appl. Phys. B 54 (1992) 486. [4] R. Loucif-Saibi, K. Nakatani, J.A. Delaire, M. Dumont, Z. Sekkat, Mol. Cryst. Liq. Cryst. 235 (1993) 251. [5] C. Fiorini, F. Charra, J.-M. Nunzi, P. Raimond, J. Opt. Soc. Am. B 14 (1997) 1984. [6] P. Rochon, E. Batalla, A. Natansohn, Appl. Phys. Lett. 66 (1995) 136. [7] A.-C. EtileÂ, C. Fiorini, F. Charra, J.-M. Nunzi, Phys. Rev. A 56 (1997) 3888. [8] D.Y. Kim, L. Li, J. Kumar, S.K. Tripathy, Appl. Phys. Lett. 66 (1995) 1166. [9] D.Y. Kim, L. Li, X.L. Jiang, V. Shivshankar, J. Kumar, S.K. Tripathy, Macromolecules 28 (1996) 8835. [10] F. LagugneÂ-Labarthet, T. Buffeteau, C. Sourisseau, J. Phys. Chem. 102 (1998) 2654. [11] T.G. Pedersen, P.M. Johansen, N.C.R. Holme, P.S. Ramanujam, Phys. Rev. Lett. 80 (1998) 89. [12] B. Darracq, F. Chaput, K. Lahlil, Y. Levy, J.P. Boilot, Adv. Mater. 10 (14) (1998) 1133. [13] C. Fiorini, N. Prudhomme, A.-C. EtileÂ, P. Lefin, P. Raimond, J.-M. Nunzi, Macromol. Symp. 137 (1999) 105. [14] C.J. Barrett, A.L. Natansohn, P.L. Rochon, J. Phys. Chem. 100 (1996) 8836. [15] C.J. Barrett, P.L. Rochon, A.L. Natansohn, J. Chem. Phys. 109 (1998) 1505. [16] P. Lefin, C. Fiorini, J.-M. Nunzi, Pure Appl. Opt. 7 (1997) 71. [17] P. Lefin, C. Fiorini, J.-M. Nunzi, Opt. Mat. 9 (1998) 323. [18] J. Kumar, L. Li, X.L. Jiang, D.Y. Kim, T.S. Lee, S.K. Tripathy, Appl. Phys. Lett. 72 (17) (1998) 2096. [19] S. Bian, L. Li, J. Kumar, D.Y. Kim, J. Williams, S.K. Tripathy, Appl. Phys. Lett. 73 (13) (1998) 1817. [20] M. Itoh, K. Harada, H. Matsuda, S. Ohnishi, A. Parfenov, N. Tamaoki, T.J. Yatagai, J. Phys. D: Appl. Phys. 31 (1998) 463. [21] D.B. Hall, A. Dhinojwala, J.M. Torkelson, Phys. Rev. Lett. 79 (1997) 103.