Photomechanical Response of Amorphous Carbon Nitride Thin Films on SiO 2 Substrate

Photo-induced deformation of amorphous carbon nitride (a-CNx) thin films was observed under visible light irradiation. This phenomenon shows the energy conversion of photon energy to mechanical energy. The a-CNx films were prepared on rectangular ultrathin SiO2 substrates by reactive radio frequency magnetron sputtering method at different deposition temperatures. The graphite like films were obtained with increasing the deposition temperature. In order to evaluate the photomechanical response of a-CNx, the time resolved bending deformation of the a-CNx/SiO2 specimens was measured using optical-lever technique when the incident light was turned on and off. The absolute amount of bending deformation was found to be the maximum value with the specimen deposited at 573 K. As a result of the time resolved measurement of the photomechanical response of the a-CNx/SiO2 specimens, all of the specimens began to be bent immediately when the light was turned on and off, and then the deformation reached to the saturation value after about 4 s. The photomechanical response speed increased with increasing the nitrogen concentration, and the graphite like specimen showed low photomechanical responsivity. [DOI: 10.1380/ejssnt.2015.352]


INTRODUCTION
Microactuator is one of the key components in micromachines.The driven systems of the actuator are various types: electrostatic [1], piezoelectric [2], heat [3] and so on.Among them, considerable attention is paid to light driven actuators [4][5][6][7][8].Because the light driven system does not need the wiring for energy supply, which is able to actuate in noncontact.Additionally, the advantages of light driven systems are wireless control, miniaturization of devices, and less electric noise.
Recently, the search for photo-responsive materials has been carried out.And very recently, photoactuators based on composite materials containing liquid crystalline elastomers and carbon nanotubes are reported [7].When the composite is illuminated by light, the carbon nanotubes generate heat locally because of absorbing photon energy, and then the elastomers show the contraction as the phase transition because the heat is transferred to the elastomers.This phenomenon can be reversibly changed by increasing and decreasing temperature.The time constant of the photomechanical response of this material is about 4 to 5.5 s.In addition, single-walled carbon nanotube/polycarbonate bilayers is also photoresponsive materials, which show the reversible bending when visible light was turned on and off [8].The time constant of the photomechanical response was reported to be about 0.5 s, which shows the very fast response.However, the photoresponse of both materials causes photo-thermal conversion.
On the other hand, the polymer containing azobenzene can be deformed reversibly with ultraviolet and visible light irradiation [5,6].This phenomenon is caused by cis-trans photoisomerization of azobenzene, and the pho-tomechanical response of this material is slow of about 10 s.
In contrast, photoinduced deformation of amorphous carbon nitride (a-CN x ) thin films under visible light irradiation is completely reversible [9,10].The deformation of a-CN x films can be obtained by the bending behavior of a-CN x /ultrathin SiO 2 substrate specimens.When the light was switched on and off, the specimens bent reversibly according to the light irradiation.Although the a-CN x films has been originally developed as a coating material [11,13], we are focusing on the biocompatibility [14] of a-CN x films and intend to apply new characteristics of phtoactuation of a-CN x films to medical devices which can be used in human bodies.In this study, we investigate influence of the nitrogen concentration and the chemical bonding states on the photomechanical response speed of a-CN x films after the light irradiation.

A. Sample preparation
Amorphous-CN x films were prepared by reactive radio frequency magnetron sputtering method.The target and reactive gas were graphite plate and pure nitrogen, respectively.The RF power was 85 W.And the reactive gas flow rate and pressure were kept constant with 3 sccm and 0.12 Torr, respectively.The deposition temperature, Ts, was 473, 573, 673, 773 and 873 K, which was controlled by the stage heater.The film thickness of all films was within 780∼1060 nm.The rectangular shaped ultrathin SiO 2 of 30×2×0.05mm 3 was used as a substrate.The photomechanical response was evaluated from the displacement change of the a-CN x /SiO 2 specimens under light irradiation using optical-lever technique.The irradiation area was about 2×3 mm 2 of film side.

B. Characterization
The nitrogen concentration, N/C, and the N-sp 2 C/Nsp 3 C bonding fraction of the a-CN x were obtained by Xray photoelectron spectroscopy (XPS; PHI ESCA1600).XPS spectra were obtained using Mg K α (1253.6 eV) radiation.The N/C ratio was calculated from the area ratio of N1s to C1s core level spectra.The N-sp 2 C/N-sp 3 C bonding fraction was calculated from the result of decomposition of N1s spectra.The N1s spectrum consisted of three peaks centered at 398.8, 400.6, and 402.5 eV of Nsp 3 C, N-sp 2 C, and N-O bonding states, respectively [15].Gaussian function was used to decompose the N1s spectra.
To evaluate the C-C bonding state, Raman scattering spectroscopy (Seishin Trading RA-07) was performed.The Nd:YAG laser (wavelength 532 nm) was used as excitation light and its power was 50 mW.The resolution is 4 cm −1 .
Fourier transform infrared spectroscopy (FT-IR; Thermo Scientific Nicolet 6700 FT-IR Spectrometer) was performed to evaluate C≡N triple bond.The accumulated number of measurements was more than 128 scans.The resolution is 4 cm −1 .

C. Photomechanical response measurement
The measurement system was developed by using the optical-lever technique.One end of the specimens was clamped by holder and the amount of displacement, δ, of the other free end of the film/substrate specimens was measured using probe He-Ne laser (wavelength 632.8 nm).The detailed measurement system was shown in Ref. 10.A Xe lamp (power 150 W) was used to irradiate the a-CN x films, where the light was concentrated to the width of 3 mm.The surface temperature change of the films caused by absorption of visible light was measured using an IR camera (NEC, TH9100MR).The temperature resolution of the camera is 0.02 • C.

A. Nitrogen concentration and chemical bonding states
Figure 1 shows the N/C ratio and the N-sp 2 C/N-sp 3 C bonding fraction as a function of the deposition temperature.The N/C ratios were calculated from the area ratio of N1s to C1s core level spectra of XPS.In order to evaluate the detailed chemical bonding states, we decomposed the N1s spectra [15].And the N-sp 2 C/N-sp 3 C bonding fractions were calculated from the area ratio of N-sp 2 C bonding to N-sp 3 C bonding.As shown in Fig. 1, the N/C ratio decreased and N-sp 2 C/N-sp 3 C bonding fraction increased with increasing the deposition temperature.
Figure 2 shows the I(D)/I(G) ratio, which correlates to the graphite cluster size in the films, as a function of the deposition temperature.The I(D)/I(G) ratios were calculated from the intensity ratio of the D peak to the G peak in Raman spectra.The G peak is located at around 1584 cm −1 and reflects the bond-stretching mode of all pairs of sp 2 bonding including the structures of rings and chains.The D peak is located at around 1360 cm −1 and reflects strongly the breathing mode of sixfold aromatic rings [16,17].The Raman spectra were decomposed into the G and D peaks with a mixture of Gaussian and Lorentzian functions.As a result, the I(D)/I(G) ratio increased with increasing the deposition temperature.This indicates that the graphite cluster size in the a-CN x films increased with increasing the deposition temperature.
In the FT-IR spectra, two peaks which centered at 2178 cm −1 and 2220 cm −1 derived from C≡N triple bonds were observed.The isonitrile (-N≡C) and nitrile (-C≡N) structures can be assigned by the wavenumbers at around 2178 cm −1 and 2220 cm −1 , respectively.The A i /A n ratios were calculated from the area ratio of the isonitrile peak http://www.sssj.org/ejssnt(J-Stage: http://www.jstage.jst.go.jp/browse/ejssnt/) (A i ) to the nitrile peak (A n ).As a result, the A i /A n ratios of a-CN x films deposited at different temperatures were obtained to be 1.20 (474 K), 1.33 (573K), 0.87 (673 K), 1.05 (773 K) and 1.09 (873 K).

B. Photomechanical response of a-CNx films
Figures 3(a) and (b) show the time dependence of the amount of displacement (δ t ) after light irradiation of the a-CN x /SiO 2 specimens deposited at temperatures from 473 to 873 K when the incident light was turned on and off, respectively.The maximum value of δ t was obtained by the specimen deposited at 573 K when the incident light was turned on.Although the reason for the maximum value of δ t observed in the a-CN x films deposited at 573 K has not been elucidated yet, we think that the existence of the C≡N triple bonds is necessary for the displacement and, in particular, the structures with the high A i /A n ratio and relatively small graphite cluster size play an important role for the displacement [9,10].
From Fig. 3(a), it can be seen that on the beginning of light irradiation, all of the specimens started to be displaced toward the film contraction immediately, and then the displacement of the specimens was saturated after about 2 to 7 s.And the time dependence of the displacement after turning off the light, as shown in Fig. 3(b), indicates that all of the specimens started to be displaced immediately after turning off, and then returned to the roughly original state after about 3 to 10 s.
In order to quantitatively evaluate the photomechanical response of a-CN x films, the time constant (τ ) after turning on and off the light irradiation was obtained by fitting the experimental data of Fig. 3 to exponential functions, and the results were shown in Fig. 4. Hereafter, the time constants after turning on and off are denoted as τ off-on and τ on-off , respectively.From this figure, it is found that τ off-on is smaller than τ on-off in all of the specimens deposited at different temperatures.We think that this behavior is related to the fact that the a-CN x films intrinsically have compressive stress [9,18].The shape of the as-deposited a-CN x films is upper convex to the film side due to ultrathin substrates.The a-CN x /SiO as shown in Fig. 5.Although the origin of intrinsic stress has not been cleared yet, the ratio of the photo-induced stress (σ p ) to the intrinsic stress (σ 0 ), σ p /σ 0 , was found to be about 15±3%, except for the a-CN x film deposited at 873 K.Both σ p and σ 0 of the film deposited at 873 K were estimated to be below the detection limit in our system.On the other hand, the difference between τ off-on and τ on-off , ∆τ , decreased slightly as a function of the deposition temperature.
When the light is turned off, the shape of the specimen changes slowly compared to turning on.Since the band gap of the a-CN x films is 0.42∼1.30eV estimated from UV-vis transmittance spectroscopy and corresponds to the photon energy of the incident light, the existence of ∆τ may suggest influence of the heat converted from the absorbed light.From the surface temperature change measurements using the IR camera, it is found that the surface temperature of the specimens becomes almost constant after about 7 to 8 s irradiation, and the speed of surface temperature change after turning on the light is somewhat faster than that after turning off the light.Therefore, ∆τ as shown in Fig. 4 would include the effect of the heat.However, although the amount of heat generation is almost proportional to T s and the surface temperature change of the film deposited at 873 K is the highest, the amount of δ t of the film is quite small.Therefore, it is reasonable to surmise that the occurrence of δ t was caused mainly by light.In the case of using the Si substrate, the surface temperature change of the a-CN x films was almost zero because of the higher thermal diffusivity, but the change of δ t was almost similar to the case of SiO 2 substrates [10].
In Fig. 4, both τ off-on and τ on-off decreased as a function of the deposition temperature.Although the displacement, δ t , has the maximum at the deposition temperature of 573 K, the time constant, τ , was negatively correlated with T s .The time constant of the photomechanical response after the white light irradiation of the present a-CN x /SiO 2 specimens was about 1.0 to 2.0 s.On the other hand, Zhang et al. [8] reported that the time constant of the photomechanical response after the white light irradiation of the single-walled carbon nanotube/polycarbonate bilayers was about 0.5 s.Comparing with both the time constants, the photomechanical response of the present a-CN x /SiO 2 specimens is found to be slow.
Figures 6(a), (b) and (c) shows the photomechanical response speed (off→on) of a-CN x films as a function of the N/C ratio, N-sp 2 C/N-sp 3 C bonding fraction and the I(D)/I(G) ratio respectively, where the photomechanical response speed of a-CN x films (ν) is defined as the amount of the saturated displacement (δ max ) divided by the elapsed time until the saturation (t).In Fig. 6(a), ν increased with increasing the N/C ratio.From Figs. 6(b) and (c), ν decreased with increasing the sp 2 bonding fraction.Therefore, these results suggest that the graphite like sp 2 bonding structures interfere with the movement of the films under light irradiation.The values of ν were not correlated with the A i /A n ratio clearly.

IV. CONCLUSIONS
We investigated the photomechanical response of the a-CN x films prepared by reactive radio frequency magnetron sputtering on ultrathin SiO 2 substrate at various temperatures, and revealed the relationship between the photomechanical response and both nitrogen concentration, and chemical bonding states.In order to evaluate the photomechanical response, the time constants after turning on and off were calculated, respectively.As a re- sult, the photomechanical response of a-CN x films takes about 4 s to be saturated.And comparing with bending speeds of a-CN x /SiO 2 bimorph and nitrogen concentration and chemical bonding states, the film which has more nitrogen and less sp 2 bonds shows the high response.Although the photomechanical response of a-CN x film is not so fast, the films have a possibility to be used safely in hu-man bodies.
FIG. 1. Nitrogen concentration (N/C) and N-sp 2 C/N-sp 3 C bonding fraction of the a-CNx films as a function of the deposition temperature.Lines are a guide to the eye.

FIG. 2 .
FIG. 2. The intensity ratio of the D peak to the G peak, I(D)/I(G), of the a-CNx films as a function of the deposition temperature.

FIG. 3 .
FIG. 3. Time dependence of the amount of displacement in the a-CNx/SiO2 specimens deposited at 473 to 873 K after the light was turned (a) on and (b) off.

FIG. 5 .
FIG. 5. Schematic illustration of the change in shape of the a-CNx/SiO2 specimen when visible light is turned on and off.
FIG. 6.The photomechanical response speed of the a-CNx/SiO2 specimens as a function of (a) N/C ratio, (b) Nsp 2 C/N-sp 3 C bonding fraction, and (c) I(D)/I(G) ratio.