1. INTRODUCTION

Gray-scale lithography is an efficient technique for the fabrication of microlens arrays, gray-scale photomasks, high-resolution optical images for micro/nano image storage, and micro-artworks [1–6]. In order to fabricate high-quality gray-scale images, material systems need to possess continuous-tone gray-scale levels. Thus, high-energy beam-sensitive (HEBS) glass was used for gray-scale lithography [1]. However, gray-scale patterns on the HEBS glass were determined by expensive electron beam exposure, which suffers from high cost. Organic photoresists such as SU-8 also were utilized to fabricate micro/nano structures, for instance, bridges, cantilevers, and micromixers by gray-scale lithography [7]. However, the molecular weight of organic photoresists was large, and the boundary between unexposed and exposed areas is unclear after laser exposure [8].

Chalcogenide material, such as ${\mathrm{Ge}}_2{\mathrm{Sb}}_2{\mathrm{Te}}_5$ and AgInSbTe, have been widely applied to data storage because of the obvious difference in electrical and optical reflectivity between its amorphous and crystalline states [9–11]. Chalcogenide thin films have been experimentally verified to be excellent inorganic resists through the etching selectivity between crystalline and amorphous states [12–14]. Besides, a transient superresolution effect also can be realized on chalcogenide materials based on nonlinear saturation absorption characteristics [15–17]. ${\mathrm{TeO}}_x$ thin film, as one of chalcogenide materials, has been used for an optical disk memory and heat-mode inorganic photoresist [18–20]. Telluride suboxide ${\mathrm{TeO}}_x$ with $x<2$, which was prepared by either the co-evaporation of Te and ${\mathrm{TeO}}_2$ or radio-frequency reactive magnetron-control sputtering technique, was found to have sufficient sensitivity to laser energy and excellent stability.

In this work, it is found that multilevel reflectivity can be obtained by controlling the “virtual” optical bandgap of ${\mathrm{TeO}}_x$ thin film through the laser writing method. It is noted that the optical bandgap can be defined only for single phase compositions; however, for our ${\mathrm{TeO}}_x$ thin film in which Te and ${\mathrm{TeO}}_2$ phases may be involved, a “virtual” bandgap is defined as the average response of ${\mathrm{TeO}}_x$ thin film. Arbitrary gray-scale image has been written on ${\mathrm{TeO}}_x$ thin film, accordingly.

2. EXPERIMENTAL

${\mathrm{TeO}}_x$ thin films with 150 nm thickness were deposited on K9 glass substrates by a radio-frequency magnetron-controlling sputtering system (JGP560 type) at room temperature, where the coating chamber was bumped to the base pressure of $7.5\times {10}^{-4}\mathrm{Pa}$ before allowing the mixture of ${\mathrm{O}}_2$ and Ar gas to flow into it. The sputtering power was 40 W. Samples were deposited with different ${\mathrm{O}}_2/\mathrm{Ar}$ ratios with a constant sputtering pressure of 0.8 Pa. The Ar gas flow was fixed at 90 sccm, and ${\mathrm{O}}_2$ gas flow varied from 0.5 to 0.8, 1.2, 1.5, and 1.8 sccm.

The as-deposited ${\mathrm{TeO}}_x$ thin film was patterned by a direct laser writing system (LW405B+) with the laser wavelength of 405 nm and the numerical aperture (NA) of 0.65 for focusing lens. The laser energy density was $0\sim 5600\mathrm{mJ}/{\mathrm{cm}}^2$. X-ray diffraction (XRD) data were recorded via 18KW-D/MAX2500V type equipment produced by Rigaku. The written images were observed by optical microscope (Olympus BX51 microscopy). Micro-reflectivity spectra were obtained by reflection spectrum (PG2000-Pro, Idea Optics Company, China). X-ray photoelectron spectroscopy (XPS) measurement was performed by an XPS spectrometer (Microlab 310F) produced by VG Scientific, and the exciting source is Mg $\mathrm{K}\alpha$. Spectroscopic ellipsometry data were measured using an ellipsometer with an automatic rotating analyzer (VASE, J. A. Woollam Co., Inc.). All measurements were carried out at room temperature.

3. RESULTS AND DISCUSSION

In this study, the pure tellurium reacted with ${\mathrm{O}}_2$ during the sputtering process in the mixture of ${\mathrm{O}}_2$ and Ar gas. ${\mathrm{O}}_2/\mathrm{Ar}$ ratio has an important effect on the structure of ${\mathrm{TeO}}_x$ thin films. Figure 1 shows the XRD patterns of the as-deposited ${\mathrm{TeO}}_x$ thin films. The peaks for metallic Te could be seen in the patterns for the samples with ${\mathrm{O}}_2$ gas flow of 0.5 and 0.8 sccm. When ${\mathrm{O}}_2$ gas flow was higher than 1.2 sccm, the diffraction peaks due to Te crystals disappeared; thus ${\mathrm{TeO}}_x$ thin films presented an amorphous state. Different ${\mathrm{O}}_2/\mathrm{Ar}$ ratios ($0.5:90$, $0.8:90$, $1.2:90$, $1.5:90$, $1.8:90$) corresponded to different stoichiometry, i.e., the value $x$ in ${\mathrm{TeO}}_x$ thin film. As ${\mathrm{O}}_2/\mathrm{Ar}$ ratio increased, the value of $x$ generally increased accordingly. However, the increment of ${\mathrm{O}}_2$ gas flow would reduce the absorption of ${\mathrm{TeO}}_x$ thin film at 405 nm wavelength due to the formation of visible transparent ${\mathrm{TeO}}_2$ [20]. It is required that ${\mathrm{TeO}}_x$ thin film has higher absorption at 405 nm wavelength for image lithography. Therefore, ${\mathrm{TeO}}_x$ thin film with ${\mathrm{O}}_2$ gas flow of 1.2 sccm was selected for further study.

 

Fig. 1. XRD patterns of ${\mathrm{TeO}}_x$ thin film with various ${\mathrm{O}}_2$ gas flows.

Download Full Size | PDF

In order to verify the atomic ratio of O: Te in as-deposited ${\mathrm{TeO}}_x$ thin film with ${\mathrm{O}}_2$ gas flow of 1.2 sccm, XPS of ${\mathrm{TeO}}_x$ thin film was measured, and the O/Te value was obtained to 0.7 from the peak areas of the O 1s and Te 3d states and the formula that the content of one element = (peak area of that element/sensitivity factor of that element)/[the sum of (peak area of each element/sensitivity factor of each element)], as shown in Fig. 2(a). For confirming the uniformity of the thickness-wise direction, the depth profile of O/Te ratio in ${\mathrm{TeO}}_x$ thin film was determined by XPS measurement as presented in Fig. 2(b). As can be seen, the O/Te ratio was $\sim 0.7$ and was basically stable in thickness direction. Thus the composition of ${\mathrm{TeO}}_{0.7}$ thin film was uniform in a thickness-wise direction.

 

Fig. 2. (a) XPS of ${\mathrm{TeO}}_x$ thin film with ${\mathrm{O}}_2$ gas flow of 1.2 sccm. (b) Depth profile of O/Te ratio in ${\mathrm{TeO}}_x$ thin film by XPS measurement.

Download Full Size | PDF

The as-deposited ${\mathrm{TeO}}_{0.7}$ thin film was heated at the protection of Ar gas up to 200°C with the heat rate of 5°C/min in order to study its thermal stability. Figure 3 shows the dependence of reflective power at 658 nm wavelength on temperature in as-deposited ${\mathrm{TeO}}_{0.7}$ thin film. The relevant details of measurement can be found elsewhere [20,21]. As shown in the figure, the curve exhibited a starting flat region, corresponding to the reflective power of as-deposited ${\mathrm{TeO}}_{0.7}$ material and then an obvious increase beyond a critical transition temperature ($T_x$). The $T_x$ value is $\sim 94°\mathrm{C}$ for ${\mathrm{TeO}}_{0.7}$ thin film, which is similar to the results of [20]. Results indicate that the as-deposited thin film had better thermal stability at room temperature. It is believed that the lower transition temperature lowered the writing threshold, and the film was more sensitive to 405 nm laser.

 

Fig. 3. Dependence of reflective power at 658 nm wavelength on temperature in as-deposited ${\mathrm{TeO}}_{0.7}$ thin film, where the heat rate is 5°C/min and Ar gas as protection gas at heating process.

Download Full Size | PDF

Figure 4 shows the XPS of ${\mathrm{Te}}_{3\mathrm{d}}$ state with various laser energy densities. For the as-deposited sample, the Te $3{\mathrm{d}}_{5/2}$ of Te-Te bonding and Te-O bonding is located at 571.08 eV and 574.28 eV, respectively, whereas Te $3{\mathrm{d}}_{3/2}$ of Te-Te bonding and Te-O bonding is located at 581.48 eV and 584.68 eV, respectively [20]. Peak position shifts to higher energy after laser exposure. This case is due to the segregation and crystallization of Te particles. Furthermore, with increasing laser energy from 1700 to $5600\mathrm{mJ}/{\mathrm{cm}}^2$, the peaks of Te-Te and Te-O bonds shift to low energy slightly, and the signal intensity of Te-O bond becomes weaker. This indicates the different local structures of ${\mathrm{TeO}}_{0.7}$ thin film with changing laser energy.

 

Fig. 4. XPS of Te 3d states in ${\mathrm{TeO}}_{0.7}$ thin film with different energy densities.

Download Full Size | PDF

For the purpose of more clearly understanding phase evolution, Fig. 5 shows XRD patterns of ${\mathrm{TeO}}_{0.7}$ thin film by different laser energy irradiations. It is clearly seen that the as-deposited ${\mathrm{TeO}}_{0.7}$ thin film was amorphous while diffraction peaks corresponding to crystalline Te occurred when samples were irradiated by laser energy. Moreover, the diffraction peaks become stronger with the increment of laser energy from 0 to $4200\mathrm{mJ}/{\mathrm{cm}}^2$. When the laser energy was higher than $4200\mathrm{mJ}/{\mathrm{cm}}^2$, the diffraction peaks of Te crystals become weaker. It is reported that the as-deposited ${\mathrm{TeO}}_{0.7}$ thin film consisted of finely dispersed, partly agglomerated crystalline Te particles embedded in an amorphous ${\mathrm{TeO}}_2$ matrix [22,23]. However, laser irradiation led to the melting of Te and ${\mathrm{TeO}}_2$ composites, followed by segregation and crystallization of Te. When laser energy was too high, the large temperature gradient of the melting of Te and ${\mathrm{TeO}}_2$ composites resulted in partly amorphization Te, again. Thus, weaker diffraction peaks appeared.

 

Fig. 5. XRD patterns of ${\mathrm{TeO}}_{0.7}$ thin film by different laser energy irradiations.

Download Full Size | PDF

In order to verify the influence of crystallization degree on its optical characteristics in ${\mathrm{TeO}}_{0.7}$ thin film, optical constant of ${\mathrm{TeO}}_{0.7}$ thin film had been determined. Figures 6(a) and 6(b) show the refractive index and extinction coefficient of ${\mathrm{TeO}}_{0.7}$ thin film with different laser energy irradiations. The refractive index and extinction coefficient changed greatly with increasing laser energy. The difference of refractive index and extinction coefficient may lead to apparent optical reflectivity contrast.

 

Fig. 6. Plots of (a) refractive index and (b) extinction coefficient values in the ${\mathrm{TeO}}_{0.7}$ thin films for various laser energy irradiations.

Download Full Size | PDF

To further investigate quantitatively laser-induced structure changes, “virtual” bandgap energy was calculated based on spectroscopic ellipsometry data, as displayed in Figs. 7(a) and 7(b). The nature of the transition can be investigated by the dependence of absorption coefficient with the incident photon energy. The absorption coefficient of ${\mathrm{TeO}}_{0.7}$ thin films can be calculated from the extinction coefficient and wavelength using the formula $\alpha =4\pi k/\lambda$, where $\alpha$ is absorption coefficient, $k$ is extinction coefficient, and $\lambda$ is optical wavelength. It is known that, in the vicinity of the fundamental absorption edge, the absorption coefficient is described by [24]

 

Fig. 7. (a) Plot of ${(\alpha h\nu )}^2$ versus $h\nu$ for ${\mathrm{TeO}}_{0.7}$ thin film with various laser energy irradiations. (b) Dependence of laser energy on “virtual” bandgap energy in ${\mathrm{TeO}}_{0.7}$ thin films.

Download Full Size | PDF

Figure 7(b) plots the dependence of laser energy on “virtual” bandgap energy in ${\mathrm{TeO}}_{0.7}$ thin films derived from Fig. 7(a). It can be seen that the bandgap energy reduced from 2.45 to 1.69 eV with the increment of laser energy from 0 to $2800\mathrm{mJ}/{\mathrm{cm}}^2$. For further increasing laser energy to $5600\mathrm{mJ}/{\mathrm{cm}}^2$, however, the bandgap energy of ${\mathrm{TeO}}_{0.7}$ thin film increased from 1.69 to 2.55 eV. It is worthwhile that the bandgap of pure Te crystal was 0.32 eV while that of ${\mathrm{TeO}}_2$ single crystal was 3.5 eV. It is suggested that ${\mathrm{TeO}}_{0.7}$ thin film consisted of Te and ${\mathrm{TeO}}_2$ composites. Tuning laser energy changed the relative content between Te and ${\mathrm{TeO}}_2$ composites, resulting in different “virtual” bandgap energy of ${\mathrm{TeO}}_{0.7}$ thin film. Accordingly, the optical constant of ${\mathrm{TeO}}_{0.7}$ thin film generated obvious changes after laser irradiations. Therefore, we maybe realize multilevel gray-scale tones by tuning the “virtual” optical bandgap of ${\mathrm{TeO}}_{0.7}$ thin film.

Thus, gray-scale levels were fabricated as well as reflectivity spectra of ${\mathrm{TeO}}_{0.7}$ thin film irradiated by various laser energy densities, as shown in Fig. 8. The inset of Fig. 8 displays gray-scale tones. It is clear that the gray-scale become lighter first and then was deeper with increasing laser energy. It was found from reflectivity spectra that the reflectivity in visible light range enhanced with the increment of laser energy from 0 to $2800\mathrm{mJ}/{\mathrm{cm}}^2$. However, further increasing laser energy from 2800 to $5600\mathrm{mJ}/{\mathrm{cm}}^2$ led to the gradual reduction of reflectivity. The segregation of Te from ${\mathrm{TeO}}_2$ matrix was responsible for the reflectivity changes [23]. In general, thin films showing Te segregation tended to have larger reflectivity increases after laser irradiations than films not showing such segregation. Moreover, the reflectivity changes were in good agreement with “virtual” bandgap energy of ${\mathrm{TeO}}_{0.7}$ thin films, as shown in Fig. 7(b). Thereby, we were capable of achieving a gray-scale image recording by adjusting the bandgap of ${\mathrm{TeO}}_{0.7}$ thin films, and adjusting “virtual” optical bandgap can be realized through changing laser irradiation energy.

 

Fig. 8. Reflectivity spectra of ${\mathrm{TeO}}_{0.7}$ thin film irradiated by various laser energy densities. The inset is gray-scale tones.

Download Full Size | PDF

Figure 9 presents the images recorded in ${\mathrm{TeO}}_{0.7}$ thin film, where the laser energy was changed from 0 to $5600\mathrm{mJ}/{\mathrm{cm}}^2$. The recorded patterns were clear, and the appearance was lifelike. Results suggested that as-deposited ${\mathrm{TeO}}_{0.7}$ thin film can be used as image recording material.

 

Fig. 9. Recording images in ${\mathrm{TeO}}_{0.7}$ thin film.

Download Full Size | PDF

4. CONCLUSIONS

In conclusion, ${\mathrm{TeO}}_x$ thin films at various ${\mathrm{O}}_2/\mathrm{Ar}$ ratios were fabricated by the reactive magnetron-controlling sputtering technique. XRD, XPS, and temperature-dependent reflective power results indicated that ${\mathrm{TeO}}_{0.7}$ thin film had better thermal stability at room temperature and lower writing threshold. Complicated images were successfully fabricated by laser adjusting “virtual” optical bandgap energy. It is found that gray-scale become lighter first with increasing laser energy and then becomes deeper with further increasing laser energy. This tendency was consistent with “virtual” bandgap energy and reflectivity evolution. Thus, images recording can be realized by adjusting the optical bandgap of ${\mathrm{TeO}}_{0.7}$ thin film, and ${\mathrm{TeO}}_{0.7}$ thin film has potential applications in the field of micro/nano image lithography.