Correlation between properties of HfO2 films and preparing parameters by ion beam sputtering deposition Huasong Liu,1 Yugang Jiang,1 Lishuan Wang,1 Jian Leng,1 Peng Sun,1 Kewen Zhuang,1 Yiqin Ji,1,* Xinbin Cheng,2 Hongfei Jiao,2 Zhanshan Wang,2 and Bingjun Wu3 1
Tianjin Key Laboratory of Optical Thin Film, Tianjin Jinhang Institute of Technical Physics, Tianjin 300192, China 2
Institute of Precision Optical Engineering, Tongji University, Shanghai 200092, China
3
College of Physics and Material Science, Qinzhou University, Qinzhou 535000, China *Corresponding author: [email protected] Received 6 August 2013; revised 23 December 2013; accepted 27 December 2013; posted 3 January 2014 (Doc. ID 195308); published 24 January 2014
Ion beam sputtering is one of the most important technologies for preparing hafnium dioxide thin films. In this paper, the correlation between properties of hafnium dioxide thin films and preparing parameters was systematically researched by using the orthogonal experiment design method. The properties of hafnium oxide films (refractive index, extinction coefficient, deposition rate, stress, and inhomogeneity of refractive index) were studied. The refractive index, extinction coefficient, physical thickness, and inhomogeneity of refractive index were obtained by the multiple wavelength curve-fitting method from the reflectance and transmittance of single layers. The stress of thin film was measured by elastic deformation of the thin film–substrate system. An orthogonal experimental strategy was designed using substrate temperature, ion beam voltage, ion beam current, and oxygen flow rate as the variables. The experimental results indicated that the temperature of the substrate is the key influencing parameter on the properties of hafnium oxide films, while other preparing parameters are also correlated with specific properties. The experimental results are significant for selecting proper parameters for preparing hafnium oxide films with different applications. © 2014 Optical Society of America OCIS codes: (310.0310) Thin films; (310.1620) Interference coatings; (310.6860) Thin films, optical properties. http://dx.doi.org/10.1364/AO.53.00A405
1. Introduction
Hafnium oxide film is one of the most important highrefractive-index materials for optical coatings due to its high laser-induced damage threshold, relatively wide band gap, transparency from the infrared to the ultraviolet range, and high thermodynamic stability [1–3]. There are many methods for the deposition of hafnium oxide film in addition to vacuum deposition methods, including electron beam 1559-128X/14/04A405-07$15.00/0 © 2014 Optical Society of America
evaporation [4], ion beam sputtering (IBS) deposition [4], ion beam assisted deposition [5], atomic layer deposition [6], and pulsed laser deposition [7]. The obtained properties of HfO2 films at the short wavelength and at the absorption band of H2 O are not better with those conventional deposition methods. IBS has also been used in a number of studies aimed at obtaining high-quality optical film properties, such as low scattering, low absorption, high packing density, low contamination, high refractive index, and environmental stability [8]. The key processing parameters of HfO2 film were relatively less researched systematically. In 1 February 2014 / Vol. 53, No. 4 / APPLIED OPTICS
A405
the literature, the research almost all focused on one or two types of properties. For the high-damagethreshold films, absorption and stress properties should be optimized. For low-loss films, the refractive index and absorption properties should be optimized. For the batch production, the deposition rate should be optimized. So in this paper, the properties of HfO2 films (refractive index, extinction coefficient, deposition rate, stress, and inhomogeneity of refractive index) were studied. In this paper, in order to obtain HfO2 films that can be used as high-refractive-index material to prepare low-loss optical films or high-damage-threshold films, the parameters for preparing HfO2 films were researched. In our experiments, HfO2 films were deposited by the IBS technique in different conditions; the preparing parameters were selected according to the character of the ion source. Through the orthogonal experimental results, the proper process parameters can be chosen to obtain the optimum properties for HfO2 film. The properties of hafnium oxide thin films, such as refractive index, extinction coefficient, physical thickness, and inhomogeneity of refractive index, were obtained by the multiple wavelength curvefitting method from their reflectance and transmittance spectra. The stress of hafnium oxide thin films was measured by a ZYGO interferometer from deformation of the film–substrate system. In the experiments, the orthogonal experiment method was chosen so that the correlation between properties of thin films and preparing parameters can be determined through analyzing the experimental results. 2. Experiment A.
Film Deposition Process
In our experiments, the IBS deposition equipment is commercial type from VEECO. When the HfO2 film was first prepared by this deposition machine, its stress was so much larger that the surface shape of prepared HfO2 film was poor. Poor surface shape of HfO2 film limits its applied aperture. Besides, the absorption property in short wavelength also needs to be further optimized. The hafnium oxide films were deposited on superpolished silica (Schott Lithosil Q1; 6 mm thick, 25 mm in diameter) and thin silica (1 mm thick, 25 mm in diameter) substrates by the IBS technique in different conditions. The corresponding average value of rms surface roughness for super-polished silica substrate is 0.336 nm. The deposition system was equipped with a 16 cm ion source and a 12 cm RF assist ion source, which were both commercial type and manufactured by VEECO Company. The alternative voltage range of the 16 cm ion beam source was set from 300 to 1300 V. The alternative current range of the 16 cm ion beam source was set from 150 to 650 mA. The parameters of ion beam voltage (300 V) and ion beam current (150 mA) were used for the assist source. In the deposition A406
APPLIED OPTICS / Vol. 53, No. 4 / 1 February 2014
Fig. 1. Scheme of the dual IBS system.
equipment, four disks with a diameter of 200 mm are fixed on the substrate holder, and the substrate holder is set at the middle of the IBS machine. The substrate holder can be rotated, and the four disks can not only be rotated, but also autorotated. Substrate heating can enhance the refractive index of HfO2 film and the stress, so in the experiments, the proper substrate temperature should be selected. There were quartz lamps applied as heat sources in the deposition system, where the temperature range can be adjusted from room temperature to 250 deg. Before deposition, the substrates were first cleaned to remove surface contaminants by the standard procedure, followed by assist ion beam bombardment for 10 min. The base pressure before deposition was pumped as low as 3 × 10−4 Pa by CTI-400 cryopump. Ar and O2 gases were introduced into the ion source and on the target surface, respectively. According to the literature, HfO2 films were all almost prepared by the reaction of Hf target and O2 gas. When Hf target was used, the splash of impurities could be prevented from being smaller compared with the HfO2 target. In the IBS machine, the diameter of the hafnium target (the purity of 99.9%) is 340 mm. The film thickness was controlled by time. The IBS system used for the deposition of hafnium oxide films is shown schematically in Fig. 1. B. Design of Orthogonal Experiment
In the experiment, the orthogonal experiment method [9–12] was employed. The merits of this method are that the strategy establishes the minimum number of runs needed to evaluate a given parameter space. In the experiments, substrate temperature, ion beam voltage, ion beam current, and oxygen flow were set as the independent preparing parameters. Details of the selection of factors and levels of preparing parameters are shown in Table 1. The orthogonal table L934 with four factors and three levels was used for arranging experiments,
Table 1.
Substrate Temperature (°C) Level 1 2 3
Selection of Factors and Levels of Preparing Parameters
Ion Beam Voltage (V)
Ion Beam Current (mA)
Volume
Symbol
Volume
Symbol
Volume
Symbol
Volume
Symbol
20 120 200
A1 A2 A3
650 950 1250
B1 B2 B3
300 450 600
C1 C2 C3
20 30 40
D1 D2 D3
Table 2.
Oxygen Flow (sccm)
Design of Orthogonal Experiment for Hafnium Oxide Films by IBS
Substrate Temperature
Ion Beam Voltage
Number
Experiment
(°C)
(V)
1 2 3 4 5 6 7 8 9
A1B1C1D1 A1B2C2D2 A1B3C3D3 A2B1C2D3 A2B2C3D1 A2B3C1D2 A1B1C3D2 A1B2C1D3 A1B3C2D1
20 20 20 120 120 120 200 200 200
650 950 1250 650 950 1250 650 950 1250
Ion Beam Current
Oxygen Flow
Deposition Time
(mA)
(sccm)
(s)
300 450 600 450 600 300 600 300 450
20 30 40 40 20 30 30 40 20
15000 8000 5000 11000 5500 11000 8000 14000 7500
as shown in Table 2. At least nine experiment groups are needed to systemically research the correlation between properties and preparing parameters of hafnium dioxide thin films.
method, span analysis was carried out based on Table 3. The values of range (R) were calculated for each index column to sequence the influence of factors on the property indices, as listed in Table 4.
C.
B. Analysis of Orthogonal Experiment
Sample Characterization
A spectro-photometer (Lambda 900) was employed in measuring the transmittance and reflectance spectra. In our experiments, reflectance measurements were done with additional equipment (L631 0200 VW Directional Reflectance Set). The basic measurement method is the V-W method, as also named the Strong method. When the beam path was calibrated, the V type was chosen. When the sample was measured, the W type was chosen. When the reflectance spectra of HfO2 films were measured, the incidence angle was about 8 deg. The transmittance accuracy of the equipment was 0.08%, and the reflectance accuracy of the equipment was 0.5%. The measured wavelength ranged from 250 to 1200 nm. The refractive index, extinction coefficient, inhomogeneity, and physical thickness of hafnium oxide films on super-polished silica substrates can be accurately determined from their reflectance and transmittance in the spectral region from the near IR (1000 nm) to the near UV (250 nm). The determination of optical constants can be accomplished by OptiChar software, and the Cauchy model was used to calculate the refractive index and extinction coefficient [13]. The stress of thin films was evaluated from deformation of thin film– substrate measured by a ZYGO interferometer. 3. Results and Discussion A.
Results of Experiments
The measured transmittance and reflectance spectra of HfO2 films and fused silica substrate are shown in Figs. 2 and 3. According to the orthogonal test
According to range analysis of the orthogonal table, the largest influencing factors on refractive index, extinction coefficient, deposition rate, stress, and inhomogeneity of refractive index of HfO2 films are individually substrate temperature, oxygen flow, ion beam current, substrate temperature, and substrate temperature. After obtaining the impact sequence of the factors on characteristics of hafnium oxide films, the impact of factor levels can be further analyzed. First, except for the very factor to be studied, average characteristics were calculated from other factors. Second, by choosing the very factor levels as the horizontal ordinate, a graph of the relation between average characteristics and this factor could be drawn. Third, the key level for controlling thin film characteristics can be seen from the curve tendency in the figures. 1. Refractive Index Figure 4 shows the refractive index of hafnium oxide films at various preparing parameters. It can be seen from Fig. 4 that with the enhancement of substrate temperature, the refractive index of hafnium oxide films increases; the impact of ion beam voltage on the thin films’ refractive index can be ignored, but the thin films’ refractive index will have minimum point in the ion beam current level C2 and level D2. Therefore, in order to reduce the refractive index of hafnium oxide films, the substrate temperature should be decreased first, and then the oxygen flow and ion beam current. This indicated a combination of A1D3C2 for processing parameters in Table 1. The 1 February 2014 / Vol. 53, No. 4 / APPLIED OPTICS
A407
Fig. 2. Transmittance spectra of HfO2 films and fused silica substrate.
Fig. 3. Transmittance spectra of HfO2 films and fused silica substrate.
influence of ion beam voltage on refractive index needs further verification. 2. Extinction Coefficient Figure 5 shows the extinction coefficient of hafnium oxide films at various preparing parameters. From Table 3.
Number
Experiment
Refractive Index at 633 nm
1 2 3 4 5 6 7 8 9
A1B1C1D1 A1B2C2D2 A1B3C3D3 A2B1C2D3 A2B2C3D1 A2B3C1D2 A1B1C3D2 A1B2C1D3 A1B3C2D1
1.9467 1.9404 1.9345 1.9736 1.9874 1.9956 2.0514 2.0304 2.0334
A408
Fig. 5, one can deduce that when enhancing the substrate temperature, the extinction coefficient of hafnium oxide films tends to increase. With the increase of ion beam voltage and ion beam current, the extinction coefficient of hafnium oxide films first increases and then decreases. With the increase of oxygen flow, the extinction coefficient of hafnium
Results of Orthogonal Experiments
Extinction Coefficient at 300 nm
Physical Thickness (nm)
Deposition Rate (nm/s)
Inhomogeneity (%)
Stress (GPa)
0.00653 0.00570 0.00489 0.00738 0.00498 0.01037 0.01104 0.00705 0.00556
466.5 492.4 485.9 477.5 448.6 479.1 417.5 470.4 483.1
0.031 0.062 0.097 0.043 0.082 0.044 0.052 0.034 0.064
0.73 0.63 0.70 1.02 0.92 1.58 4.76% 7.16% 6.05%
−0.446 −0.697 −0.730 −1.093 −1.218 −1.009 −2.124 −2.533 −2.340
APPLIED OPTICS / Vol. 53, No. 4 / 1 February 2014
Table 4.
Refractive index Extinction coefficient Deposition rate (nm/s) Inhomogeneity Stress (GPa)
Span Analysis of Orthogonal Experiments
Substrate Temperature (°C)
Ion Beam Voltage (V)
Ion Beam Current (mA)
Oxygen Flow (sccm)
0.097874 2.18E − 03 0.013209 0.053012 1.707938
0.004478 2.41E − 03 0.026151 0.007324 0.261945
0.008617 1.77E − 03 0.040892 0.010331 0.047428
0.016275 3.35E − 03 0.006594 0.006396 0.175499
oxide films first decreases and then increases. Therefore, in order to reduce the extinction coefficient of hafnium oxide films, we should first reduce the substrate temperature and choose the proper ion beam voltage, ion beam current, and oxygen flow. The specific adjustment for improving the thin films’ extinction coefficient needs further verification. 3. Deposition Rate Figure 6 shows the deposition rate of hafnium oxide films at various preparing parameters. From Fig. 6, we can deduce that with the enhancement of substrate temperature, the deposition rate of hafnium oxide films tends to be monotone down. With the enhancement of ion beam voltage and ion beam current, the deposition rate of hafnium oxide films tends to be monotone increasing. In order to achieve high deposition rate, the combination of A1B3C3 for processing parameters is the optimum selection. The impact of oxygen flow on deposition rate exhibits
a slight inflection point, so it indicates that oxygen flow near level 2 should be avoided in the processing. 4. Stress Figure 7 shows the stress of hafnium oxide films at various preparing parameters. Figure 7 tells us that the stresses of hafnium oxide films are all compressive stress. For the IBS deposition system, with the increase of the deposition time, the density of the surface layer increases and the surface layer has the tendency of expansion, so the prepared HfO2 films all exhibit high compressive stress. High compressive film stress should be prevented according to the practical application. According to the small size mirror, high compressive stress cannot be considered. When HfO2 film was applied in the large size mirror, we needed to consider the stress of the matching low-refractive-index film material. When low-refractive-index film exhibits low stress, high compressive stress for HfO2 films is not desirable.
Fig. 4. Impact of processing parameters level on the refractive index of hafnium oxide films.
Fig. 5. Impact of processing parameters level on the extinction coefficient of hafnium oxide films. 1 February 2014 / Vol. 53, No. 4 / APPLIED OPTICS
A409
Fig. 6. Impact of processing parameters level on the deposition rate of hafnium oxide films.
Fig. 7. Impact of processing parameters level on the stress of hafnium oxide films.
When low-refractive-index film exhibits high tensile stress, high compressive stress for HfO2 films is desirable and the stress of multilayer coatings can be reduced. With the enhancement of substrate temperature, the compressive stress of hafnium oxide films tends to increase. A maximum point of stress appeared at level 2 of ion beam voltage. Ion beam current has little impact on the stress, while minimum stress points appeared at level 2 of oxygen flow. Therefore, in order to control the stress of hafnium oxide films, one should first decrease the substrate temperature, and then choose the proper ion beam voltage and oxygen flow. Ion beam voltage should avoid values near level 2, while oxygen flow should be set near the value of level 2.
5. Inhomogeneity of Refractive Index In this paper, the inhomogeneity of refractive index was only considered in the direction of film thickness. Positive inhomogeneity means that the refractive index of the film surface is larger than the bottom of the film layer. Negative inhomogeneity means that the refractive index of the film surface is lower than the bottom of the film layer. Inhomogeneity of refractive index is common positive for HfO2 films. Figure 8 shows the refractive index inhomogeneity of hafnium oxide films at various preparing parameters. It can be seen that the refractive index inhomogeneity of HfO2 films is all positive. The IBS deposition process is a strongly nonequilibrium state with high energy/momentum; the film layer
Fig. 8. Impact of processing parameters level on the refractive index inhomogeneity of hafnium oxide films A410
APPLIED OPTICS / Vol. 53, No. 4 / 1 February 2014
was grown on a “cold” surface with a high energy/ momentum particle impact. In the setting time, the high energy/momentum deposition process was suddenly stopped, so the mutation area appeared in the near surface layer and the refractive index in the near surface layer was higher than other positions. Besides, with the increase of the deposition time, the temperature of the target surface increases and the surface layer density increases, so the refractive index in the near surface layer is higher. So the inhomogeneity of refractive index for HfO2 films under different preparing parameters is all positive. According to HfO2 films, high thin film inhomogeneity is not desirable. For example, positive thin film inhomogeneity can impact the properties of antireflective coatings, such as central wavelength and residual reflectance. It can be seen from Fig. 8 that as the substrate temperature increases, the refractive index inhomogeneity of hafnium oxide films tends to increase. A maximum point of refractive index inhomogeneity appeared at level 2 of ion beam voltage. With the enhancement of ion beam current, the refractive index inhomogeneity of hafnium oxide films tends to decrease. A minimum refractive index inhomogeneity appeared at level 2 of oxygen flow. Therefore, in order to decrease refractive index inhomogeneity, one should first decrease the substrate temperature, and then increase ion beam voltage and ion beam current. The combination of A1B1C3 for processing parameters is the optimum selection. The proper oxygen flow should be set around level 2 and may need more experiments to be further determined. In the practical application, HfO2 films are all needed to match the low-refractive-index material to prepare the multilayer coatings (highly reflective coating, antireflective coating), and the experimental results for the application of multilayer coatings need further analysis. 4. Conclusion
In summary, the correlation between properties of hafnium dioxide thin films deposited by IBS and preparing parameters was systematically studied. The results indicated that a decrease of substrate temperature is good for all the properties of hafnium dioxide thin films. Low ion beam voltage is in favor of reducing refractive index inhomogeneity of hafnium dioxide thin films, while high ion beam voltage is in favor of increasing the deposition rate hafnium dioxide thin films. Meanwhile the impact of ion beam voltage on the stress, extinction coefficient, and refractive index needs further verification. High ion beam current is in favor of increasing the deposition rate and reducing the refractive index inhomogeneity of hafnium dioxide thin films and has less impact on the
stress, and the impact on the extinction coefficient and refractive index needs further verification. High oxygen flow is in favor of reducing the refractive index and the stress of hafnium dioxide thin films. But the impact of oxygen flow on the extinction coefficient, deposition rate, and refractive index inhomogeneity needs further verification. Comprehensively speaking, in order to improve the refractive index and stress of hafnium oxide films, which are the two key characteristics of optical films, it is contradictory for the choosing of processing parameters. Therefore, one should carefully choose the proper parameters for the deposition of hafnium oxide films with different application requirements. This research was supported by the National Natural Science Foundation of China (No. 61235011) and the Science and Technology Commission of Tianjin Municipality (Nos. 13JCYBJC17300, 12JCQNIC01200). References 1. R. Chow, S. Falabella, G. E. Loomis, F. Rainer, C. J. Stolz, and M. R. Kozlowski, “Reactive evaporation of low defect density hafnia,” Appl. Opt. 32, 5567–5574 (1993). 2. C. J. Stolz, L. M. Sheehan, M. K. Gunten, R. P. Bevis, and D. Smith, “The advantages of evaporation of hafnium in a reactive environment to manufacture high damage threshold multilayer coatings by electron-beam deposition,” Proc. SPIE 3738, 318–324 (1999). 3. M. Alvisi, M. Di Giulio, S. G. Marrone, M. R. Perrone, M. L. Protopapa, A. Valentini, and L. Vasanelli, “HfO2 films with high laser damage threshold,” Thin Solid Films 358, 250–258 (2000). 4. L. Gallais, J. Capoulade, J. Yves, M. Commandré, M. Cathelinaud, C. Koc, and M. Lequime, “Laser damage resistance of hafnia thin films deposited by electron beam deposition, reactive low voltage ion plating, and dual ion beam sputtering,” Appl. Opt. 47, C107–C113 (2008). 5. P. André, L. Poupinet, and G. Ravel, “Evaporation and ion assisted deposition of HfO2 coatings: some key points for high power applications,” J. Vac. Sci. Technol. 18, 2372–2377 (2000). 6. J. J. Ganem, I. Trimaille, I. C. Vickridge, D. Blin, and F. Martin, “Study of thin hafnium oxides deposited by atomic layer deposition,” Nucl. Instrum. Methods Phys. Res. B 219–220, 856–861 (2004). 7. M. Ratzke, D. Wolfframm, M. Kappa, S. Kouteva-Arguirova, and J. Reif, “Pulsed laser deposition of HfO2 and PrxOy high-k films on Si(100),” Appl. Surf. Sci. 247, 128–133 (2005). 8. T. W. David, “Ion beam interference coating for ultralow optical loss,” Appl. Opt. 28, 2813–2816 (1989). 9. The writing group of “Handbook of Mathematics”, Handbook of Mathematics,” in Chinese (High Education, 1979). 10. H. Xu, “An algorithm for constructing orthogonal and nearly orthogonal arrays with mixed levels and small runs,” Technometrics 44, 356–368 (2002). 11. R. V. Leon and C. F. J. Wu, “A theory of performance measures in parameter design,” Statist. Sinica 2, 353–358 (1992). 12. T. R. Jensen, J. Warren, and R. L. Johnson, Jr., “Ion-assisted deposition of moisture-stable hafnium oxide films for ultraviolet applications,” Appl. Opt. 41, 3205–3210 (2002). 13. http://www.optilayer.com.
1 February 2014 / Vol. 53, No. 4 / APPLIED OPTICS
A411
Tianjin Key Laboratory of Optical Thin Film, Tianjin Jinhang Institute of Technical Physics, Tianjin 300192, China 2
Institute of Precision Optical Engineering, Tongji University, Shanghai 200092, China
3
College of Physics and Material Science, Qinzhou University, Qinzhou 535000, China *Corresponding author: [email protected] Received 6 August 2013; revised 23 December 2013; accepted 27 December 2013; posted 3 January 2014 (Doc. ID 195308); published 24 January 2014
Ion beam sputtering is one of the most important technologies for preparing hafnium dioxide thin films. In this paper, the correlation between properties of hafnium dioxide thin films and preparing parameters was systematically researched by using the orthogonal experiment design method. The properties of hafnium oxide films (refractive index, extinction coefficient, deposition rate, stress, and inhomogeneity of refractive index) were studied. The refractive index, extinction coefficient, physical thickness, and inhomogeneity of refractive index were obtained by the multiple wavelength curve-fitting method from the reflectance and transmittance of single layers. The stress of thin film was measured by elastic deformation of the thin film–substrate system. An orthogonal experimental strategy was designed using substrate temperature, ion beam voltage, ion beam current, and oxygen flow rate as the variables. The experimental results indicated that the temperature of the substrate is the key influencing parameter on the properties of hafnium oxide films, while other preparing parameters are also correlated with specific properties. The experimental results are significant for selecting proper parameters for preparing hafnium oxide films with different applications. © 2014 Optical Society of America OCIS codes: (310.0310) Thin films; (310.1620) Interference coatings; (310.6860) Thin films, optical properties. http://dx.doi.org/10.1364/AO.53.00A405
1. Introduction
Hafnium oxide film is one of the most important highrefractive-index materials for optical coatings due to its high laser-induced damage threshold, relatively wide band gap, transparency from the infrared to the ultraviolet range, and high thermodynamic stability [1–3]. There are many methods for the deposition of hafnium oxide film in addition to vacuum deposition methods, including electron beam 1559-128X/14/04A405-07$15.00/0 © 2014 Optical Society of America
evaporation [4], ion beam sputtering (IBS) deposition [4], ion beam assisted deposition [5], atomic layer deposition [6], and pulsed laser deposition [7]. The obtained properties of HfO2 films at the short wavelength and at the absorption band of H2 O are not better with those conventional deposition methods. IBS has also been used in a number of studies aimed at obtaining high-quality optical film properties, such as low scattering, low absorption, high packing density, low contamination, high refractive index, and environmental stability [8]. The key processing parameters of HfO2 film were relatively less researched systematically. In 1 February 2014 / Vol. 53, No. 4 / APPLIED OPTICS
A405
the literature, the research almost all focused on one or two types of properties. For the high-damagethreshold films, absorption and stress properties should be optimized. For low-loss films, the refractive index and absorption properties should be optimized. For the batch production, the deposition rate should be optimized. So in this paper, the properties of HfO2 films (refractive index, extinction coefficient, deposition rate, stress, and inhomogeneity of refractive index) were studied. In this paper, in order to obtain HfO2 films that can be used as high-refractive-index material to prepare low-loss optical films or high-damage-threshold films, the parameters for preparing HfO2 films were researched. In our experiments, HfO2 films were deposited by the IBS technique in different conditions; the preparing parameters were selected according to the character of the ion source. Through the orthogonal experimental results, the proper process parameters can be chosen to obtain the optimum properties for HfO2 film. The properties of hafnium oxide thin films, such as refractive index, extinction coefficient, physical thickness, and inhomogeneity of refractive index, were obtained by the multiple wavelength curvefitting method from their reflectance and transmittance spectra. The stress of hafnium oxide thin films was measured by a ZYGO interferometer from deformation of the film–substrate system. In the experiments, the orthogonal experiment method was chosen so that the correlation between properties of thin films and preparing parameters can be determined through analyzing the experimental results. 2. Experiment A.
Film Deposition Process
In our experiments, the IBS deposition equipment is commercial type from VEECO. When the HfO2 film was first prepared by this deposition machine, its stress was so much larger that the surface shape of prepared HfO2 film was poor. Poor surface shape of HfO2 film limits its applied aperture. Besides, the absorption property in short wavelength also needs to be further optimized. The hafnium oxide films were deposited on superpolished silica (Schott Lithosil Q1; 6 mm thick, 25 mm in diameter) and thin silica (1 mm thick, 25 mm in diameter) substrates by the IBS technique in different conditions. The corresponding average value of rms surface roughness for super-polished silica substrate is 0.336 nm. The deposition system was equipped with a 16 cm ion source and a 12 cm RF assist ion source, which were both commercial type and manufactured by VEECO Company. The alternative voltage range of the 16 cm ion beam source was set from 300 to 1300 V. The alternative current range of the 16 cm ion beam source was set from 150 to 650 mA. The parameters of ion beam voltage (300 V) and ion beam current (150 mA) were used for the assist source. In the deposition A406
APPLIED OPTICS / Vol. 53, No. 4 / 1 February 2014
Fig. 1. Scheme of the dual IBS system.
equipment, four disks with a diameter of 200 mm are fixed on the substrate holder, and the substrate holder is set at the middle of the IBS machine. The substrate holder can be rotated, and the four disks can not only be rotated, but also autorotated. Substrate heating can enhance the refractive index of HfO2 film and the stress, so in the experiments, the proper substrate temperature should be selected. There were quartz lamps applied as heat sources in the deposition system, where the temperature range can be adjusted from room temperature to 250 deg. Before deposition, the substrates were first cleaned to remove surface contaminants by the standard procedure, followed by assist ion beam bombardment for 10 min. The base pressure before deposition was pumped as low as 3 × 10−4 Pa by CTI-400 cryopump. Ar and O2 gases were introduced into the ion source and on the target surface, respectively. According to the literature, HfO2 films were all almost prepared by the reaction of Hf target and O2 gas. When Hf target was used, the splash of impurities could be prevented from being smaller compared with the HfO2 target. In the IBS machine, the diameter of the hafnium target (the purity of 99.9%) is 340 mm. The film thickness was controlled by time. The IBS system used for the deposition of hafnium oxide films is shown schematically in Fig. 1. B. Design of Orthogonal Experiment
In the experiment, the orthogonal experiment method [9–12] was employed. The merits of this method are that the strategy establishes the minimum number of runs needed to evaluate a given parameter space. In the experiments, substrate temperature, ion beam voltage, ion beam current, and oxygen flow were set as the independent preparing parameters. Details of the selection of factors and levels of preparing parameters are shown in Table 1. The orthogonal table L934 with four factors and three levels was used for arranging experiments,
Table 1.
Substrate Temperature (°C) Level 1 2 3
Selection of Factors and Levels of Preparing Parameters
Ion Beam Voltage (V)
Ion Beam Current (mA)
Volume
Symbol
Volume
Symbol
Volume
Symbol
Volume
Symbol
20 120 200
A1 A2 A3
650 950 1250
B1 B2 B3
300 450 600
C1 C2 C3
20 30 40
D1 D2 D3
Table 2.
Oxygen Flow (sccm)
Design of Orthogonal Experiment for Hafnium Oxide Films by IBS
Substrate Temperature
Ion Beam Voltage
Number
Experiment
(°C)
(V)
1 2 3 4 5 6 7 8 9
A1B1C1D1 A1B2C2D2 A1B3C3D3 A2B1C2D3 A2B2C3D1 A2B3C1D2 A1B1C3D2 A1B2C1D3 A1B3C2D1
20 20 20 120 120 120 200 200 200
650 950 1250 650 950 1250 650 950 1250
Ion Beam Current
Oxygen Flow
Deposition Time
(mA)
(sccm)
(s)
300 450 600 450 600 300 600 300 450
20 30 40 40 20 30 30 40 20
15000 8000 5000 11000 5500 11000 8000 14000 7500
as shown in Table 2. At least nine experiment groups are needed to systemically research the correlation between properties and preparing parameters of hafnium dioxide thin films.
method, span analysis was carried out based on Table 3. The values of range (R) were calculated for each index column to sequence the influence of factors on the property indices, as listed in Table 4.
C.
B. Analysis of Orthogonal Experiment
Sample Characterization
A spectro-photometer (Lambda 900) was employed in measuring the transmittance and reflectance spectra. In our experiments, reflectance measurements were done with additional equipment (L631 0200 VW Directional Reflectance Set). The basic measurement method is the V-W method, as also named the Strong method. When the beam path was calibrated, the V type was chosen. When the sample was measured, the W type was chosen. When the reflectance spectra of HfO2 films were measured, the incidence angle was about 8 deg. The transmittance accuracy of the equipment was 0.08%, and the reflectance accuracy of the equipment was 0.5%. The measured wavelength ranged from 250 to 1200 nm. The refractive index, extinction coefficient, inhomogeneity, and physical thickness of hafnium oxide films on super-polished silica substrates can be accurately determined from their reflectance and transmittance in the spectral region from the near IR (1000 nm) to the near UV (250 nm). The determination of optical constants can be accomplished by OptiChar software, and the Cauchy model was used to calculate the refractive index and extinction coefficient [13]. The stress of thin films was evaluated from deformation of thin film– substrate measured by a ZYGO interferometer. 3. Results and Discussion A.
Results of Experiments
The measured transmittance and reflectance spectra of HfO2 films and fused silica substrate are shown in Figs. 2 and 3. According to the orthogonal test
According to range analysis of the orthogonal table, the largest influencing factors on refractive index, extinction coefficient, deposition rate, stress, and inhomogeneity of refractive index of HfO2 films are individually substrate temperature, oxygen flow, ion beam current, substrate temperature, and substrate temperature. After obtaining the impact sequence of the factors on characteristics of hafnium oxide films, the impact of factor levels can be further analyzed. First, except for the very factor to be studied, average characteristics were calculated from other factors. Second, by choosing the very factor levels as the horizontal ordinate, a graph of the relation between average characteristics and this factor could be drawn. Third, the key level for controlling thin film characteristics can be seen from the curve tendency in the figures. 1. Refractive Index Figure 4 shows the refractive index of hafnium oxide films at various preparing parameters. It can be seen from Fig. 4 that with the enhancement of substrate temperature, the refractive index of hafnium oxide films increases; the impact of ion beam voltage on the thin films’ refractive index can be ignored, but the thin films’ refractive index will have minimum point in the ion beam current level C2 and level D2. Therefore, in order to reduce the refractive index of hafnium oxide films, the substrate temperature should be decreased first, and then the oxygen flow and ion beam current. This indicated a combination of A1D3C2 for processing parameters in Table 1. The 1 February 2014 / Vol. 53, No. 4 / APPLIED OPTICS
A407
Fig. 2. Transmittance spectra of HfO2 films and fused silica substrate.
Fig. 3. Transmittance spectra of HfO2 films and fused silica substrate.
influence of ion beam voltage on refractive index needs further verification. 2. Extinction Coefficient Figure 5 shows the extinction coefficient of hafnium oxide films at various preparing parameters. From Table 3.
Number
Experiment
Refractive Index at 633 nm
1 2 3 4 5 6 7 8 9
A1B1C1D1 A1B2C2D2 A1B3C3D3 A2B1C2D3 A2B2C3D1 A2B3C1D2 A1B1C3D2 A1B2C1D3 A1B3C2D1
1.9467 1.9404 1.9345 1.9736 1.9874 1.9956 2.0514 2.0304 2.0334
A408
Fig. 5, one can deduce that when enhancing the substrate temperature, the extinction coefficient of hafnium oxide films tends to increase. With the increase of ion beam voltage and ion beam current, the extinction coefficient of hafnium oxide films first increases and then decreases. With the increase of oxygen flow, the extinction coefficient of hafnium
Results of Orthogonal Experiments
Extinction Coefficient at 300 nm
Physical Thickness (nm)
Deposition Rate (nm/s)
Inhomogeneity (%)
Stress (GPa)
0.00653 0.00570 0.00489 0.00738 0.00498 0.01037 0.01104 0.00705 0.00556
466.5 492.4 485.9 477.5 448.6 479.1 417.5 470.4 483.1
0.031 0.062 0.097 0.043 0.082 0.044 0.052 0.034 0.064
0.73 0.63 0.70 1.02 0.92 1.58 4.76% 7.16% 6.05%
−0.446 −0.697 −0.730 −1.093 −1.218 −1.009 −2.124 −2.533 −2.340
APPLIED OPTICS / Vol. 53, No. 4 / 1 February 2014
Table 4.
Refractive index Extinction coefficient Deposition rate (nm/s) Inhomogeneity Stress (GPa)
Span Analysis of Orthogonal Experiments
Substrate Temperature (°C)
Ion Beam Voltage (V)
Ion Beam Current (mA)
Oxygen Flow (sccm)
0.097874 2.18E − 03 0.013209 0.053012 1.707938
0.004478 2.41E − 03 0.026151 0.007324 0.261945
0.008617 1.77E − 03 0.040892 0.010331 0.047428
0.016275 3.35E − 03 0.006594 0.006396 0.175499
oxide films first decreases and then increases. Therefore, in order to reduce the extinction coefficient of hafnium oxide films, we should first reduce the substrate temperature and choose the proper ion beam voltage, ion beam current, and oxygen flow. The specific adjustment for improving the thin films’ extinction coefficient needs further verification. 3. Deposition Rate Figure 6 shows the deposition rate of hafnium oxide films at various preparing parameters. From Fig. 6, we can deduce that with the enhancement of substrate temperature, the deposition rate of hafnium oxide films tends to be monotone down. With the enhancement of ion beam voltage and ion beam current, the deposition rate of hafnium oxide films tends to be monotone increasing. In order to achieve high deposition rate, the combination of A1B3C3 for processing parameters is the optimum selection. The impact of oxygen flow on deposition rate exhibits
a slight inflection point, so it indicates that oxygen flow near level 2 should be avoided in the processing. 4. Stress Figure 7 shows the stress of hafnium oxide films at various preparing parameters. Figure 7 tells us that the stresses of hafnium oxide films are all compressive stress. For the IBS deposition system, with the increase of the deposition time, the density of the surface layer increases and the surface layer has the tendency of expansion, so the prepared HfO2 films all exhibit high compressive stress. High compressive film stress should be prevented according to the practical application. According to the small size mirror, high compressive stress cannot be considered. When HfO2 film was applied in the large size mirror, we needed to consider the stress of the matching low-refractive-index film material. When low-refractive-index film exhibits low stress, high compressive stress for HfO2 films is not desirable.
Fig. 4. Impact of processing parameters level on the refractive index of hafnium oxide films.
Fig. 5. Impact of processing parameters level on the extinction coefficient of hafnium oxide films. 1 February 2014 / Vol. 53, No. 4 / APPLIED OPTICS
A409
Fig. 6. Impact of processing parameters level on the deposition rate of hafnium oxide films.
Fig. 7. Impact of processing parameters level on the stress of hafnium oxide films.
When low-refractive-index film exhibits high tensile stress, high compressive stress for HfO2 films is desirable and the stress of multilayer coatings can be reduced. With the enhancement of substrate temperature, the compressive stress of hafnium oxide films tends to increase. A maximum point of stress appeared at level 2 of ion beam voltage. Ion beam current has little impact on the stress, while minimum stress points appeared at level 2 of oxygen flow. Therefore, in order to control the stress of hafnium oxide films, one should first decrease the substrate temperature, and then choose the proper ion beam voltage and oxygen flow. Ion beam voltage should avoid values near level 2, while oxygen flow should be set near the value of level 2.
5. Inhomogeneity of Refractive Index In this paper, the inhomogeneity of refractive index was only considered in the direction of film thickness. Positive inhomogeneity means that the refractive index of the film surface is larger than the bottom of the film layer. Negative inhomogeneity means that the refractive index of the film surface is lower than the bottom of the film layer. Inhomogeneity of refractive index is common positive for HfO2 films. Figure 8 shows the refractive index inhomogeneity of hafnium oxide films at various preparing parameters. It can be seen that the refractive index inhomogeneity of HfO2 films is all positive. The IBS deposition process is a strongly nonequilibrium state with high energy/momentum; the film layer
Fig. 8. Impact of processing parameters level on the refractive index inhomogeneity of hafnium oxide films A410
APPLIED OPTICS / Vol. 53, No. 4 / 1 February 2014
was grown on a “cold” surface with a high energy/ momentum particle impact. In the setting time, the high energy/momentum deposition process was suddenly stopped, so the mutation area appeared in the near surface layer and the refractive index in the near surface layer was higher than other positions. Besides, with the increase of the deposition time, the temperature of the target surface increases and the surface layer density increases, so the refractive index in the near surface layer is higher. So the inhomogeneity of refractive index for HfO2 films under different preparing parameters is all positive. According to HfO2 films, high thin film inhomogeneity is not desirable. For example, positive thin film inhomogeneity can impact the properties of antireflective coatings, such as central wavelength and residual reflectance. It can be seen from Fig. 8 that as the substrate temperature increases, the refractive index inhomogeneity of hafnium oxide films tends to increase. A maximum point of refractive index inhomogeneity appeared at level 2 of ion beam voltage. With the enhancement of ion beam current, the refractive index inhomogeneity of hafnium oxide films tends to decrease. A minimum refractive index inhomogeneity appeared at level 2 of oxygen flow. Therefore, in order to decrease refractive index inhomogeneity, one should first decrease the substrate temperature, and then increase ion beam voltage and ion beam current. The combination of A1B1C3 for processing parameters is the optimum selection. The proper oxygen flow should be set around level 2 and may need more experiments to be further determined. In the practical application, HfO2 films are all needed to match the low-refractive-index material to prepare the multilayer coatings (highly reflective coating, antireflective coating), and the experimental results for the application of multilayer coatings need further analysis. 4. Conclusion
In summary, the correlation between properties of hafnium dioxide thin films deposited by IBS and preparing parameters was systematically studied. The results indicated that a decrease of substrate temperature is good for all the properties of hafnium dioxide thin films. Low ion beam voltage is in favor of reducing refractive index inhomogeneity of hafnium dioxide thin films, while high ion beam voltage is in favor of increasing the deposition rate hafnium dioxide thin films. Meanwhile the impact of ion beam voltage on the stress, extinction coefficient, and refractive index needs further verification. High ion beam current is in favor of increasing the deposition rate and reducing the refractive index inhomogeneity of hafnium dioxide thin films and has less impact on the
stress, and the impact on the extinction coefficient and refractive index needs further verification. High oxygen flow is in favor of reducing the refractive index and the stress of hafnium dioxide thin films. But the impact of oxygen flow on the extinction coefficient, deposition rate, and refractive index inhomogeneity needs further verification. Comprehensively speaking, in order to improve the refractive index and stress of hafnium oxide films, which are the two key characteristics of optical films, it is contradictory for the choosing of processing parameters. Therefore, one should carefully choose the proper parameters for the deposition of hafnium oxide films with different application requirements. This research was supported by the National Natural Science Foundation of China (No. 61235011) and the Science and Technology Commission of Tianjin Municipality (Nos. 13JCYBJC17300, 12JCQNIC01200). References 1. R. Chow, S. Falabella, G. E. Loomis, F. Rainer, C. J. Stolz, and M. R. Kozlowski, “Reactive evaporation of low defect density hafnia,” Appl. Opt. 32, 5567–5574 (1993). 2. C. J. Stolz, L. M. Sheehan, M. K. Gunten, R. P. Bevis, and D. Smith, “The advantages of evaporation of hafnium in a reactive environment to manufacture high damage threshold multilayer coatings by electron-beam deposition,” Proc. SPIE 3738, 318–324 (1999). 3. M. Alvisi, M. Di Giulio, S. G. Marrone, M. R. Perrone, M. L. Protopapa, A. Valentini, and L. Vasanelli, “HfO2 films with high laser damage threshold,” Thin Solid Films 358, 250–258 (2000). 4. L. Gallais, J. Capoulade, J. Yves, M. Commandré, M. Cathelinaud, C. Koc, and M. Lequime, “Laser damage resistance of hafnia thin films deposited by electron beam deposition, reactive low voltage ion plating, and dual ion beam sputtering,” Appl. Opt. 47, C107–C113 (2008). 5. P. André, L. Poupinet, and G. Ravel, “Evaporation and ion assisted deposition of HfO2 coatings: some key points for high power applications,” J. Vac. Sci. Technol. 18, 2372–2377 (2000). 6. J. J. Ganem, I. Trimaille, I. C. Vickridge, D. Blin, and F. Martin, “Study of thin hafnium oxides deposited by atomic layer deposition,” Nucl. Instrum. Methods Phys. Res. B 219–220, 856–861 (2004). 7. M. Ratzke, D. Wolfframm, M. Kappa, S. Kouteva-Arguirova, and J. Reif, “Pulsed laser deposition of HfO2 and PrxOy high-k films on Si(100),” Appl. Surf. Sci. 247, 128–133 (2005). 8. T. W. David, “Ion beam interference coating for ultralow optical loss,” Appl. Opt. 28, 2813–2816 (1989). 9. The writing group of “Handbook of Mathematics”, Handbook of Mathematics,” in Chinese (High Education, 1979). 10. H. Xu, “An algorithm for constructing orthogonal and nearly orthogonal arrays with mixed levels and small runs,” Technometrics 44, 356–368 (2002). 11. R. V. Leon and C. F. J. Wu, “A theory of performance measures in parameter design,” Statist. Sinica 2, 353–358 (1992). 12. T. R. Jensen, J. Warren, and R. L. Johnson, Jr., “Ion-assisted deposition of moisture-stable hafnium oxide films for ultraviolet applications,” Appl. Opt. 41, 3205–3210 (2002). 13. http://www.optilayer.com.
1 February 2014 / Vol. 53, No. 4 / APPLIED OPTICS
A411
