Effect of electron-beam deposition process variables on the film characteristics of the CrO_x films

The film containing the CrO_x compounds is an important constituent component of the potential industrial devices. The films are potentially applied in the spectrally selective absorber coating for the renewable solar energy applications,^1–3 in the electrochromic coating for the decoration application,⁴ in the protective coating for the harsh environment against the wear and corrosion,^5,6 in the absorptive coating used in attenuated phase-shifting mask for the lithographic applications,^7,8 in the epitaxial magnetoresistance coating for the sensor applications.^9,10 The chromium oxides and their derivatives have unique characteristics which result in a variety of applications. The Cr₂O₃ compound is one of the hardest oxides with a chemical stability of insolubility in water and most organic solutions, corrosive and wear resistance and low friction coefficient.^11,12 The film of CrO₂ is a strong ferromagnetic at room temperature, maintains its ferromagnetism up to temperatures of 120^oC, and has a half-metallic band structure fully spin-polarized at the Fermi level.¹³ Among the oxide derivatives, dichromium trioxide is more stable than chromium dioxide. Metal Cr film is antiferromagnetic at temperatures lower than 38^oC and changes to paramagnetic above that temperature. The tandem film of the multiple layers consisting of Cr_xO_y and metal demonstrates low reflectance, thermal emittance and high absorptance in the visible and infrared spectrum.¹ The tandem film is widely applied in the black absorber application.

A variety of deposition methods have been introduced to prepare the chromium oxide films. These methods include thermal evaporation,¹⁴ chemical vapor deposition,¹⁵ molecular beam epitaxy,^9,10 glow-discharged plasma technology,⁵ spray pyrolysis,⁶ hydrothermal synthesized deposition,¹⁶ pulsed laser deposition,¹⁷ laser CVD,¹³ reactive DC magnetron sputtering,^2,4,7,11,18 RF sputtering,^3,8 high-power impulse magnetron sputtering (HIPIMS)¹ and electron-beam deposition.¹⁹ A few reports have been addressed on the optical properties of the deposited chromium oxide films. The most popular film deposition method in the industry is magnetron sputtering after the reference survey. However, one of the critical limitations for the DC reactive sputtering is that dielectric material film coatings are difficult to produce.¹ The method of RF sputtering is widely accepted for dielectric material films coating, but the hardware investment is relatively high, while the deposition rate is low and the process is complicated. The HIPIMS, an advanced sputtering technology, has advantages of high deposition rate and direct reaction of metal target with required gases to become dielectric compounds, but the delicate power supplier and electronic controller are additional hardware investment. The advantages for the electron-beam deposition are available to deposit the large size of both the metal films and the dielectric films, and affordable hardware investment. Therefore, it is the reason why electron-beam deposition is more adopted in the optical film coating industry. Recently, Al-Kuhaili et al.¹⁹ deposited Cr₂O₃ film using electron-beam evaporation and only investigated the effect of substrate temperatures on optical properties. They deconvoluted the electron status of oxygen into two components corresponding to low and high substrate temperatures. Eight parameters were necessary to be tuned to fit the transmittance spectrum and to retrieve the optical constants within the measured UV/VIS wavelength ranges.

We present the optical constants calculated from the reflectance and transmittance measurements of the chromium oxide films deposited on the glass substrate under a variety of experimental conditions. The X-ray diffractometry (XRD) and sheet resistance of the films are examined. X-ray photoelectron spectroscopy (XPS) was applied to characterize the obtained films and to illustrate the possible causes to the measured data. The film performance results responsible for the process variables of the electron-beam deposition can be contributed as the referred starting point for the manipulation required for a specific characteristic of the film.

The CrO_x films were deposited by an electron-beam deposition system (Fulintec Co., Taiwan) with a cryogenic pump, an ion gauge to monitor the pressure in the chamber, and a RF assisted deposition ion source. The molybdenum crucible was applied to hold the material granules and the thicknesses were monitored by the quartz sensor. The background pressure was pumped down to 8x10^-4 Pa. The raw materials used were Cr₂O₃ granules with 99% purity and approximately 2-3 mm in diameter, B270 glass substrates of 50x50x1 mm³, and oxygen, argon, and nitrogen gases of 5N purity. The glass substrates were cleaned using soap detergent, rinsed in flowing deionized water, and dried using blown nitrogen. The oxygen flow rates varied at 0 standard cubic centimeter per minute (sccm), 5 sccm, 10 sccm, and 25 sccm, which corresponded to the chamber pressures of 1x10^-3 Pa, 5x10^-3 Pa, 2x10^-2 Pa, and 4x10^-2 Pa, respectively to evaluate the effect of oxidation degree on chromium oxide with regard to its structural and optical characteristics when the films were deposited at ambient temperature. The substrate temperatures were set at 25^oC, 100^oC, 150^oC, and 200^oC and the chamber pressure was maintained at 1x10^-3 Pa to evaluate the effect of temperature on the microstructure of the grown films. The effect of the ion-assisted-deposition (IAD) process on the performance of the deposited chromium oxide films with the introduction of oxygen and argon ion sources was carried out at the substrate temperature of 200^oC with a chamber pressure of 4x10^-2 Pa.

The reflectance and transmittance of the films deposited on the glass substrates were measured in the atmospheric environment using a spectrophotometer (Perkin-Elmer Lambda-900) with a wavelength range from 350 nm to 2000 nm. The reflectance measurements were carried out using the V-N type fixture to obtain the absolute reflectance. The microstructure of the deposited films was examined by X-ray diffractometry (XRD, Bruker M18XHF) with Cu K_α irradiation and a bird’s-eye view on the film morphology was observed by a scanning electron microscope (SEM, JOEL JSM-7000F). The sheet resistance of the deposited films was measured by a four-point probe instrument (Sadhu Design EM 4P). The X-ray photoelectron spectroscopy (XPS, ULVAC-PHI Quantera SXM) excited by an aluminum anode with monochromatic K_α X-ray was applied to determine the bonding status and the possible combination compounds of the chromium oxide and their corresponding contributions with the aid of Gaussian curve fitting technique. Before measurement, the binding energy of 284.5 eV for the 1s status of carbon from the hydrocarbon contaminations was calibrated as the reference energy.²⁰ 

The bird’s-eye view images to expose the surface and cross section of the deposited films conducted at a set deposition temperature of 25^oC with various oxygen flow rates of 0 sccm, 5 sccm, 10 sccm, and 25 sccm are exhibited in Fig. 1(a)–1(d), respectively. All deposited films are dense and uniform; no significant morphological differences in the film surfaces and the cross-sections are observed for the films with the studied ranges of oxygen flow rates. The images obtained for the films deposited by electron-beam technique have similar features of films with comparable thicknesses obtained by reactive DC sputtering.¹¹ 

The reflectance and transmittance measurements of the deposited films on the glass substrates varied with different oxygen rates are shown in Fig. 2(a) and 2(b), respectively. The wavelength measurement ranges from 350 nm to 2000 nm. The reflectance and transmittance of the substrate are also exhibited in Fig. 2 for the reference purpose. The maximal reflectance increases and the corresponding valley transmittance decreases with decreasing oxygen flow rates of the deposited films. Supposedly, the deposited films should have the same wavelengths for the maximum and minimum values of reflectance and transmittance measurements if the films have the same optical lengths. However, the wavelengths located for the peak reflectance and their corresponding trough transmittance of the films deposited at lower oxygen flow rates are smaller than that deposited at higher oxygen flow rates. The different locations of the wavelengths for the extremes of the reflectance and transmittance indicate the optical lengths of the deposited films are significantly different. The optical length is defined as the product of the refraction index by the physical thickness of the film and is found to increase with increasing oxygen flow rates. The values of the reflectance measured for the deposited films are higher than that of the glass substrate, indicating that the refraction indices of the chromium oxide films are larger than the glass substrate refraction index within the measurement range. The apparent fluctuations in Figs. 2 occurred in the range of 850-875 nm originate from the signal detector change, whereas the fluctuations occurred in the ranges of 1350-1450 nm and 1800-1950 nm originate from the water vapor absorption.²¹ The data are presented in the pristine form because the fluctuations are systematic and the results have no influence on the data analysis or comparison.

The calculated optical constants, including refraction index (n) and extinction coefficient (k), of the films prepared at varied oxygen flow rates are illustrated in Fig. 3(a) and 3(b), respectively. The mathematical manipulation^22–24 to determine the optical constants of a film deposited on a transparent substrate is as follows; Firstly, the optical constants of the thick glass substrate are obtained by minimizing the differences between the theoretical and measured values for both reflectance and transmittance. The same manipulation is applied for a film stack, comprised of a film deposited on a substrate of specific thickness. The Newton-Raphson numerical iteration is applied to solve simultaneously the two values, n and k for a specific wavelength until the differences are below a pre-determined number when a proper film thickness is given. The refraction index of the chromium oxide films decreases with the increase of the measured wavelength for a film and decreases with increasing oxygen flow rates, i.e. the chamber pressures, for a specific wavelength within the measurement range, shown in Fig. 3(a). The refraction index of the supported substrate is included. The observed phenomenon that the refraction index decreases with increasing the measured wavelength is one of the characteristics of the oxide dielectric materials examined in visible and near infrared spectrum. For instance, the refraction index of a film prepared at chamber pressure of 1x10^-3 Pa decreases monotonically from 2.477 at the wavelength of 350 nm to 2.020 at 2000 nm. The refraction indices of the films deposited at the oxygen flow rates of 0, 5, 10, and 25 sccm are 2.185, 2.084, 2.013, and 1.832 for the measured wavelength of 650 nm. All the refraction indices of the films are larger than 1.528, which is the calculated index of the substrate at the wavelength of 650 nm. The best estimated film thicknesses determined from Figs. 2 are 150 nm, 177 nm, 185 nm and 238 nm for the films prepared at the chamber pressures of 1x10^-3 Pa, 5x10^-3 Pa, 2x10^-2 Pa, and 4x10^-2 pa, respectively. Smoothing was done on the simulation curves around the detector changed regions to prevent the equation from being unsolvable due to the severe signal fluctuations. The extinction coefficients of the deposited films decrease with increasing wavelengths. In general, the values of the extinction coefficient decrease with increasing oxygen flow rate near the UV and visible wavelength spectrum and become indistinguishable near the infrared region. The slight increases of k at the wavelength ranges larger than 1700 nm may result from the environmental water vapor absorption. The trends of the extinction coefficient of the films coincide inversely with the transmittance, indicating that the higher maximum value of transmittance, the smaller the extinction coefficient. But the differences of k among the films are smaller when the wavelengths are longer than visible spectrum because chromium is oxidized.

The XPS results for Cr 2p_3/2 core level regions from the deposited films prepared at chamber pressures of 1x10^-3 Pa and 4x10^-2 Pa are illustrated in Fig. 4a, and 4b, respectively. The Cr 2p_3/2 peaks are characteristic to identify the derivative compounds and the binding energies for the pure compound Cr₂O₃, CrO₂ and metal Cr are 576.8 eV, 576.3 eV and 574.3 eV, respectively.²⁰ The XPS results were deconvoluted into three components, as shown in Figs. 4 when the resulted enveloped curve is assumed to be constructed by their individual components in a Gaussian form. The area ratio of Cr₂O₃, CrO₂ and Cr for the film deposited at the chamber pressure of 1x10^-3 Pa is determined as 0.39:0.41:0.20 and the ratio for the film at pressure of 4x10^-2 Pa is 0.47:0.40:0.13. The component ratio of metal Cr in the films prepared at a high oxygen flow rate is smaller than that of the films prepared at no oxygen flow rate during the deposition process, while the compound Cr₂O₃ is consumed and CrO₂ remained unchanged. The higher metal Cr component in the deposited film obtained at lower chamber pressures or oxygen flow rates corresponds to lower sheet resistance evaluated by a four-point probe technique. The experimental data of the sheet resistance of the films, which increases with increasing the chamber pressures, is shown in Fig. 5. The result may also correspond to the maximum reflectance measurements, in which the higher maximum reflectance of the films is obtained under the experiment conditions of lower chamber pressures.

To estimate the optical absorption edges of the deposited films at different chamber pressures, the Tauc plot,^25,26 in which the (αhν)² versus hν, is shown in Fig. 6. The value α is an optical absorption coefficient calculated by the equation $α=1dln(1−RT)$ from the reflectance and transmittance spectrums and d is the film thickness. The film thicknesses used for the calculation are 150 nm, 177 nm, 185 nm, and 238 nm, respectively, and the values of hν are the incident photon energies derived from the wavelengths. The optical bandgaps of the films are determined ∼ 3.15 eV, 3.18 eV, 3.20 eV, and 3.24 eV by the intercepting the straight line portion with the abscissa in the Tauc plot, assuming that the film is of direct allowed transition. The bandgap values slightly increase with increasing oxygen flow rates. The extinction coefficient of the glass substrate approaching 10^-3∼10^-4 is small and contributes negligible amount to the measured absorptance. The bandgaps of Cr₂O₃ films prepared at different deposition conditions and supported substrates were reported.^{14–17,19,27} However, the reported values were scattered widely from 2.67 eV to 5.238 eV and are heavily dependent on the film deposition methods, the detail microstructure, exponent used in the Tauc equation to fit the data, film thickness and the substrate material used.

The films on a glass substrate deposited at the fixed chamber pressure of 1x10^-3 Pa with varied substrate temperatures of 25^oC, 100^oC, 150^oC, and 200^oC were carried out. The reflectance and transmittance measurements against the wavelength are exhibited in Fig. 7a and 7b, respectively. The substrate temperature at 200^oC has the maximum reflectance, while the optical measurements of the other three temperatures are comparable with one another. The corresponding optical constants, i.e. refraction index and extinction coefficient with the measured wavelengths are exhibited in Fig. 8a and 8b, respectively. The estimated film thicknesses used to match the measured values and theoretical values were 150 nm, 150 nm, 160 nm and 137 nm for the films deposited on various substrate temperatures. The refraction index of the film deposited at the substrate temperature of 200^oC is higher than and nearly parallel to that of the films deposited at the other three substrate temperatures against the measured wavelength, shown in Fig. 8a. The extinction coefficient of film in the visible and near infrared region demonstrates the same tendency with the refraction index and the diagram inserted in Fig. 8b gives a clear view for easy comparison.

The XRD results for the films prepared at substrate temperatures of 25^oC and 200^oC are given in Fig. 9 with the upper curve for 25^oC and the lower for 200^oC. No apparent diffraction peaks are observed, indicating that the structure of the films is amorphous or nanocrystal cluster. The onset crystalline peaks were reported when the substrate temperatures were higher than 300^oC for the films prepared by electron-beam,¹⁹ 450^oC by both chemical vapor deposition¹⁵ and pulsed laser deposition.¹⁷ The relationship of the sheet resistance varied with the substrate temperatures examined by four-point probe technique is shown in Fig. 10. The sheet resistance decreases gradually with increasing substrate temperatures and shows an apparent drop in resistance at the temperature of 200^oC. The XPS was applied to examine the bonding status of the outer shell of the chromium, to identify the constitutive components and to estimate qualitatively the occupied weighting. The enveloped curves of binding energies ranged between 582 eV and 570 eV measured on the films deposited at substrate temperatures of 25^oC and 200^oC are illustrated and their three constitutive components are given in Fig. 11a and 11b, respectively. The occupied area ratio of Cr₂O₃:CrO₂:Cr for the film deposited at 25^oC is 39%:41%:20%, while the film deposited at 200^oC is 34%:36:30%. The component metal Cr in the film deposited at 200^oC and chamber pressure of 1x10^-3 Pa is higher than that deposited at other temperatures and the increase of metal Cr at the expenditure of oxides is observed to explain the sheet resistance shown in Fig. 10.

The optical reflectance and transmittance of the films prepared at the substrate temperature of 200^oC, the chamber pressure of 4x10^-2 Pa with and without IAD process using oxygen and argon as ion sources or chamber gas are illustrated in Fig. 12a and 12b, respectively. The relationship between the refraction index of the deposited films and the measured wavelength is given in Fig. 13 when the film thicknesses used in simulation were 111 nm, 156 nm, 116 nm, and 163 nm for films deposited with an introducing Ar gas, with IAD using an Ar source, with an introducing O₂ gas, and with IAD using an O₂ source. The denoted legends in Figs. 12 and 13 are Ar in, Ar IAD, O2 in, and O2 IAD, respectively. The refraction index of the films decreases with increasing wavelength and this tendency is observed in most dielectric materials measured in the ranges of visible and near infrared spectrum. The refraction indexes of films deposited under the atmospheres of oxygen and argon are similar but the values are higher than that obtained with IAD using Ar ion source and O₂ ion source. This controversy to the general consensus of the IAD resulting in a higher refraction index of the film can be explained using the XPS. The area ratio of the Cr₂O₃:CrO₂:Cr for the films deposited under the four experimental conditions is tabulated in Table I and the XPS for films with two oxygen situations are illustrated in Fig. 14a and 14b. The area weighting order of Cr concentration from high to low is: the film deposited with Ar environment, the film with O₂ environment, the film with IAD using Ar ion and the film with IAD using O₂ ion. The high metal Cr concentration in the films results in high refraction index and the optical characteristic of the film behaves like metallic rather than dielectric. It is easy to solve the area difference of the XPS under the enveloped curves when Fig. 14a and 14b are compared. The film obtained under the IAD with O₂ ions results in significant area increase of Cr₂O₃ and decrease of Cr in XPS data. The compound Cr₂O₃ is more stable in films with O₂ rich environment.

The electronic and optical performances of the CrO_x films on glass substrates deposited by electron-gun deposition were examined. The microstructures of the deposited films are amorphous and the appearances of the films are dense and uniform under the various process variables in this study. The refraction index of the deposited films increases with decreasing chamber pressures or oxygen flow rates and the extinction coefficient is of the same tendency as the refractive index within the visible spectrum range. The XPS data verifies that the film prepared at lower pressure has a higher area ratio of Cr within the structure compared to the film prepared at high pressure. The sheet resistance of the film also increases with increasing chamber pressure, indicating that with a higher fraction of free electrons from Cr, higher refraction index, higher extinction coefficient, and less electrical resistance can be seen in the film. The optical bandgaps obtained using the direct allowed transition of the Tauc plot are determined as 3.15 eV, 3.18 eV, 3.20 eV and 3.24 eV for the pressures of 1x10^-3 Pa, 5x10^-3 Pa, 2x10^-2 Pa, and 4x10^-2 Pa, respectively. The refraction index and extinction coefficient of the films increase generally with increasing substrate temperatures when the films were deposited at 1x10^-3 Pa. The sheet resistance of the films decreases gradually with increasing substrate temperatures and drops abruptly at the substrate temperature of 200^oC. This phenomenon agrees with the refraction index observation. XPS data demonstrate that the area ratio of Cr to the oxides is 20% in the film prepared at the room temperature and increases to 30% in the film prepared at 200^oC. The refraction index of film prepared at the pressure of 4x10^-2 Pa in argon atmosphere without IAD is slightly higher than that prepared at the pressure of 4x10^-2 Pa in oxygen atmosphere without IAD in the visible spectrum, and the phenomenon is reversed in the near infrared region when the substrate temperature was 200^oC. The refraction index of the films prepare with IAD process is lower than that without IAD and is the lowest when prepared with IAD using oxygen as the ion source. The optical properties of the film of the lowest refraction index resemble the stable oxide Cr₂O₃.

The authors thank Taiwan Ministry of Science and Technology for financial support.