**4.1 Introduction**

The photoinduced reversible volume changes of the lithographically patterned features of such materials are also used for realizing diffraction gratings, operating in the Raman-Nath regime, with controlled and reversible diffraction efficiency (DE). In particular, the alternating pulsed UV and green laser irradiation on the gratings causes the doped photochromic molecules to undergo transformations, which induce reversible dimensional changes to the samples. These volume changes cause reversible changes to the diffraction efficiency of the gratings, which is increased upon UV irradiation and decreased after irradiation with green laser light for various irradiation cycles. The experimental results are confirmed by a theoretical diffraction model. It is proved that the diffraction efficiency changes are attributed exclusively to the reversible dimensional changes of the imprinted structures (Fragouli et al 2008), and not to the refractive index changes as is the case in the majority of previous works. Specifically, most of the gratings with similar thickness as the here presented, are produced after irradiation with interfering beams that cause variation of the refractive index inside the samples mainly consisting of azobenzene polymers and liquid crystals. In these studies the switching procedure in the DE is due to the formation and deletion of the gratings (Yamamoto et al 2001, Fu et al 2005). Moreover, although the thickness of the produced gratings is very small, they exhibit higher DE than most of the volume gratings prepared in this regime (Tong et al 2005, Yamamoto et al 2001, Fu et al. 2005).

### **4.2 Results and discussion**

A grating of period 4 μm and thickness of ca. 240 nm was formed on the photochromic polymeric surface by the SM technique as described at section 2. The experimental setup used for the study of the diffraction efficiency (DE) of the formed grating, is demonstrated in Figure 10. As shown, a continuous wave diode laser operating at λ=822nm, with an incident angle adjusted to have the maximum intensity of the first-order diffracted line (I1), was used as a reading beam. This wavelength was chosen in order not to be absorbed by the photochromic sample. The zero-order transmitted (I0) and the I1 lines, are measured by two photodiodes, and were used for the calculation of the DE (equation 3).

$$\text{DE} = \text{I}\_1 / \text{I}\_0 \tag{3}$$

The DE was calculated for three irradiation cycles, and Figure 11 shows the DE relative change in each case. Initially, the DE of the grating was ca 2.2%, and upon UV laser irradiation it was increasing until it reached a final value. After two minutes in the dark in order to leave time to the system to relax, subsequent irradiation with green laser pulses

In conclusion, at this section it is presented the possibility to create both hydrophobic and hydrophilic surfaces starting from the same photochromic polymeric sample by changing the topological parameters of its surface features using soft molding lithography. Due to the photochromic transformations taking place upon alternating UV and green irradiation, these surfaces can reversibly change their wettability. By careful control of the surface topology these changes can be fully controlled and tuned, in such a way that the surfaces

The photoinduced reversible volume changes of the lithographically patterned features of such materials are also used for realizing diffraction gratings, operating in the Raman-Nath regime, with controlled and reversible diffraction efficiency (DE). In particular, the alternating pulsed UV and green laser irradiation on the gratings causes the doped photochromic molecules to undergo transformations, which induce reversible dimensional changes to the samples. These volume changes cause reversible changes to the diffraction efficiency of the gratings, which is increased upon UV irradiation and decreased after irradiation with green laser light for various irradiation cycles. The experimental results are confirmed by a theoretical diffraction model. It is proved that the diffraction efficiency changes are attributed exclusively to the reversible dimensional changes of the imprinted structures (Fragouli et al 2008), and not to the refractive index changes as is the case in the majority of previous works. Specifically, most of the gratings with similar thickness as the here presented, are produced after irradiation with interfering beams that cause variation of the refractive index inside the samples mainly consisting of azobenzene polymers and liquid crystals. In these studies the switching procedure in the DE is due to the formation and deletion of the gratings (Yamamoto et al 2001, Fu et al 2005). Moreover, although the thickness of the produced gratings is very small, they exhibit higher DE than most of the volume gratings prepared in this regime (Tong et al 2005, Yamamoto et al 2001, Fu et al.

A grating of period 4 μm and thickness of ca. 240 nm was formed on the photochromic polymeric surface by the SM technique as described at section 2. The experimental setup used for the study of the diffraction efficiency (DE) of the formed grating, is demonstrated in Figure 10. As shown, a continuous wave diode laser operating at λ=822nm, with an incident angle adjusted to have the maximum intensity of the first-order diffracted line (I1), was used as a reading beam. This wavelength was chosen in order not to be absorbed by the photochromic sample. The zero-order transmitted (I0) and the I1 lines, are measured by two

The DE was calculated for three irradiation cycles, and Figure 11 shows the DE relative change in each case. Initially, the DE of the grating was ca 2.2%, and upon UV laser irradiation it was increasing until it reached a final value. After two minutes in the dark in order to leave time to the system to relax, subsequent irradiation with green laser pulses

DE= I1/I0 (3)

photodiodes, and were used for the calculation of the DE (equation 3).

**4. Reversible diffraction efficiency changes of photochromic polymer** 

can be wetted in a reversible manner.

**gratings** 

2005).

**4.2 Results and discussion** 

**4.1 Introduction** 

caused the DE to recover close to its initial value. It is worth noticing that the relative changes of the DE during the first irradiation cycle exhibit big variations between the various examined samples, in contrary with the following irradiation cycles, where the changes are similar for all the examined gratings. This is mainly attributed to internal stresses of the polymer matrix, produced during the preparation of the gratings, that are released in a random way upon irradiation (Liang et al 2007). Nevertheless, at the second cycle during UV irradiation the DE increases with increasing number of pulses, until it stabilizes to an average value of approximately 7.4±3.0% with respect to its initial value. After green irradiation the DE is slowly reaching its initial value with increasing number of pulses. This behavior is repeated also at the third cycle.

Fig. 10. Experimental setup for the measurement of the diffraction efficiency of the photochromic gratings (10% wt. SP in PEMMA) (λUV=355nm, FUV=20 mJ cm-2, λgreen=532 nm, Fgreen=35 mJ cm-2, λreading beam=822 nm.)

Fig. 11. Diffraction efficiency changes of the grating upon UV-green irradiation. (Fragouli et al 2008)

In order to examine the effect of the refractive index (*n*) change of the photochromic polymer sample upon UV-green irradiation on the observed change to the DE, ellipsometric

Photocontrolled Reversible Dimensional Changes of Microstructured Photochromic Polymers 161

after the photoisomerization process, by 8.7%. The agreement between the calculated and experimental (ca. 7.4±3.0%) values is notable, taking into account the experimental error due to factors such as the exact value of the laser beam spot size, which leads to an approximate value of the observed stripes, the imperfections of the surface introduced during the grating formation, and so on. The theoretical calculations presented, demonstrate that the decrease of the dimensions of the stripes of the gratings and of the period, are the main parameters

Fig. 12. AFM images of a grating before any irradiation (a) after irradiation with 20 UV pulses (b), and after exposure to 600 green pulses (c). The insets demonstrate a single stripe in magnification. 3D images (d–f) of the grating shown in (a–c), respectively. (α is the period and β is the distance between the stripes, measured at the full width at half-maximum of

In conclusion, it is demonstrated how soft molding lithography can be employed for the preparation of microstructured photochromic polymeric films, which undergo light controlled photomechanical changes responsible for the control of some functional characteristics of the patterned surfaces, namely the wetting properties and the diffraction

each feature. (Fragouli et al 2008)

**5. Conclusions** 

that define the change in the DE upon irradiation (Fragouli et al 2008).

measurements on a similar sample were conducted. The results show that before any irradiation for λ=822 nm, the refractive index of the sample is *n*=1.509. After UV irradiation, the *n* is higher by Δ*n*=0.029. This difference is very small compared to the periodic refractive index variations in the grating between the photochromic polymer (1.509) and the air (1), which is actually what causes diffraction to occur. Thus it is believed that this change plays a negligible role in the measured DE relative changes. Furthermore, the small thickness of the gratings (ca. 240 nm) reduces the importance of the Δ*n* even more.

Thus, the reversible DE changes can be attributed exclusively to the light-induced macroscopic deformations of the gratings. Specifically, Figure 12 illustrates the reversible macroscopic changes of the grating before and after UV-green irradiation as taken by AFM microscopy. As shown, the width of the stripes of the grating (α−β) is decreased by ca. 13% after UV irradiation while the distance between the two stripes (β) is increased. A small decrease is also observed in the period of the grating (α) (α and β before UV, 3.971 μm and 2.366 μm respectively; α and β after UV, 3.842 μm and 2.449 μm respectively). After the subsequent irradiation with green light the values recover very close to the initial ones. It is worth noticing that, as shown at Figure 12, there is a dip separating each stripe in two equal parts. As already mentioned at section 2.2, the SM technique which is followed for realizing the grating relies mainly on the capillarity that allows the viscous polymer to spontaneously fill the vertical channels that are made of the recessed features of the elastomeric mold, since the wetting lowers the overall free energy. There is always the possibility that the photochromic polymer may not fill completely such regions, thus may be mostly accumulated in the regions that are adjacent to the protruding areas of the mold, forming thus dips in the central part of the growing capillarity features. This behavior is common to different imprint lithography methods (Zankovych et al 2001, Hong and Lee 2003, Pisignano et al 2004). However, this dip is useful for the AFM morphological analysis of the patterned surfaces, since it makes the volume changes upon UV-green irradiation cycles much clearer. Moreover, it is too narrow to give any contribution to the diffracted light from the grating. In order to compare the experimental result with the existing theory, the basic equation that describes the intensity distribution of monochromatic light passing through a grating, was used (equation 5). (Born and Wolf 1999)

$$\frac{I}{I\_0} = \left[\frac{\sin(\pi \mathcal{J} \, p \,/\, \lambda)}{\pi \mathcal{J} \, p \,/\, \lambda}\right]^2 \left[\frac{\sin(N \pi \alpha \, p \,/\, \lambda)}{\sin(\pi \alpha \, p \,/\, \lambda)}\right]^2 \tag{5}$$

I and I0 are the intensities of the light after the grating at various orders of diffraction and at zero order respectively, β is the distance between two successive stripes, α is the period of the grating, N is the number of stripes, λ is the wavelength of the reading beam, and p = sinθ -sinθ0 =mλ/α (m=0, ±1, ±2 etc) where θ*<sup>0</sup>* is the angle of incidence and θ the angle of diffraction. The number of the stripes covered by the reading beam was calculated by dividing the diameter of the spot of the beam by the period of the grating in each case. The angle of incidence of the reading beam was θ0 =20°. In each case, by the AFM images it was measured the different value of α and β before and after UV-green irradiation. Taking into account the measured parameters by the experiment and using the equation 5, it was calculated the ratio I1/I0, and consequently the DE for the gratings before and after the UV– green irradiation. The theoretical calculations confirm that there is an increase of the DE after the photoisomerization process, by 8.7%. The agreement between the calculated and experimental (ca. 7.4±3.0%) values is notable, taking into account the experimental error due to factors such as the exact value of the laser beam spot size, which leads to an approximate value of the observed stripes, the imperfections of the surface introduced during the grating formation, and so on. The theoretical calculations presented, demonstrate that the decrease of the dimensions of the stripes of the gratings and of the period, are the main parameters that define the change in the DE upon irradiation (Fragouli et al 2008).

Fig. 12. AFM images of a grating before any irradiation (a) after irradiation with 20 UV pulses (b), and after exposure to 600 green pulses (c). The insets demonstrate a single stripe in magnification. 3D images (d–f) of the grating shown in (a–c), respectively. (α is the period and β is the distance between the stripes, measured at the full width at half-maximum of each feature. (Fragouli et al 2008)
