Effects of chemical etching with nitric acid on glass surfaces - AVS

Effects of chemical etching with nitric acid on glass surfaces H. K. Jang,a) Y. D. Chung, S. W. Whangbo, T. G. Kim, and C. N. Whang Atomic-scale Surface Science Research Center and Department of Physics, Yonsei University, Seoul 120-749, Korea

S. J. Lee Research and Development Department, Advanced Lab Solution, Seoul 120-140, Korea

S. Lee Photonics Research Center, Korea Institute of Science and Technology, Seoul 130-650, Korea

共Received 8 December 1999; accepted 23 October 2000兲 Glass slides were chemically etched with nitric acid using five different methods. We investigated the effects of chemical etching treatments on such properties as chemical composition, surface roughness, transmittance, and thermal stability of the glass. Sodium, calcium, and aluminum atoms in the surface of the glass are effectively reduced by chemical etching with nitric acid. Especially, boiling the glass in a 70% nitric acid solution for 30 min is more effective in the reduction of sodium, calcium, and aluminum atoms at the surface of the glass than other etching methods with nitric acid. The surface morphologies of the glasses were very similar regardless of the chemical etching treatments with nitric acid etchant. Root-mean-square surface roughness of the bare glass was 0.58 nm but that of the glass etched with nitric acid was ranged from 5.4 to 6.8 nm. Sodium concentration at the glass surface was largely reduced from 2.44% to 0.25%. In order to investigate the thermal stability of the glass, the bare glass and the glass samples boiled for 30 min in a 70% HNO3 solution were annealed in air at 300, 400, 500, and 600 °C for 1 h. Sodium, aluminum, and calcium at the surface of the glass increased with an annealing temperature regardless of chemical etching with nitric acid, but the Na, Al, and Ca content at the surface of the glass boiled in 70% HNO3 for 30 min was greatly reduced relative to that of the surface of the bare glass at an annealing temperature of 500 °C. © 2001 American Vacuum Society. 关DOI: 10.1116/1.1333087兴

I. INTRODUCTION Glass is of considerable importance for the production of scientific, technical, architectural, and decorative objects as well as many articles used in daily life.1 In most glass products, sodium is added to reduce the melting temperature during glass production. However, the sodium at the glass surface reacts with water vapor, causing degradation of glass durability2 and the failure of the insulated gate field-effect transistor.3 It has also been recognized that the reaction of sodium oxide migrating from a soda glass into TiO2 films produces either a brookite TiO2 phase or an incompletely characterized phase, referred to as Na2O•xTiO2 . 4,5 Various surface treatments of glassware that remove alkali from the surface have been found to greatly improve the durability of the glassware. The treatment with the longest history involves exposing the hot glass to air containing a small percentage of sulphur dioxide.6,7 Paz et al. tried to eliminate sodium at the glass surface by boiling it for 30 min in a 50% sulfuric acid solution, but they only investigated the reaction between glass and TiO2 films for a study of the photoactivity.8,9 To date, this phenomenon has not been reported in detail. In this article, we report on observations related to the reduction of sodium, calcium, and aluminum in glass by chemical etching with nitric acid at various treatments and

the thermal stability of the glass. Effects of chemical etching on such properties of glass as surface roughness, chemical composition, transmittance, and thermal stability of the glass were investigated with x-ray photoelectron spectroscopy 共XPS兲, atomic force microscope 共AFM兲, transmittance, and Rutherford backscattering spectroscopy 共RBS兲. II. EXPERIMENT Commercial slide glasses 共Halfwhite glass, Superior Co.兲 were used for this experiment, and the plate was cut into 12 mm⫻12 mm samples. The composition of the glass slide samples used in this study was SiO2 共72.8%–73.0%兲, Na2O 共14.6%–14.8%兲, K2O 共0.70%–0.80%兲, CaO 共5.80%– 6.00%兲, MgO 共3.9%–4.1%兲, Al2O3 共1.35%–1.40%兲, and Fe2O3 共0.045%–0.047%兲. Five glass samples were prepared with different chemical etching treatments using nitric acid etchant. One sample was not etched, but set aside to be compared with etched samples. The samples were etched in the following manner: 共a兲共b兲共c兲共d兲

a兲

Electronic mail: [email protected]

267

J. Vac. Sci. Technol. A 19„1…, JanÕFeb 2001

bare glass; rinsed glass with isopropanol, soaked in a 50% HNO3 solution at room temperature for 6 h; rinsed glass with isopropanol, soaked in a 70% HNO3 solution at room temperature for 6 h; rinsed glass with isopropanol, sonicated for 30 min in a 50% HNO solution;

0734-2101Õ2001Õ19„1…Õ267Õ8Õ$18.00

©2001 American Vacuum Society

267

268

Jang et al.: Effects of chemical etching with nitric acid

268 TABLE I. Atomic composition of the glass surface as a function of chemical etching conditions.

FIG. 1. XPS survey spectra of the glass surface after various etching treatments: 共a兲 bare glass, 共b兲 soaked for 6 h in 50% HNO3 , 共c兲 soaked for 6 h in 70% HNO3 , 共d兲 sonicated for 30 min in 50% HNO3 , 共e兲 boiled for 30 min in 50% HNO3 , and 共f兲 boiled for 30 min in 70% HNO3 .

共e兲共f兲

rinsed glass with isopropanol, boiled for 30 min in a 50% HNO3 solution; and rinsed glass with isopropanol, boiled for 30 min in a 70% HNO3 solution.

After chemical etching, all samples were washed thoroughly with deionized water, then dried in pure N2 gas. XPS was used to investigate the composition of the glass surface. The XPS data were obtained with a Physical Electronics PHI 5700 electron spectroscopy for chemical analysis spectrometer using a Al K ␣ (h ␯ ⫽1486.6 eV) x-ray source, with an energy resolution of 0.86 eV. The x-ray source setting was always 300 W, 15 kV, and 20 mA. An AFM 共PSI Co.兲 was used to study the surface morphology of the etched glass and to measure the root-meansquare 共rms兲 surface roughness. Optical characteristics of the glass were deduced from the transmission spectra measured with a UV-3100 共Shimadzu Co.兲, which was used to investigate transmittance of the bare glass (T bare) and the etched glass (T etch). In order to investigate the thermal stability of the glass, the etched glass sample boiled for 30 min in a 70% HNO3 solution and the bare glass were annealed in air at 300, 400, 500, and 600 °C for 1 h. Atomic migration at the glass surface during annealing treatment were measured by RBS and XPS. For RBS observations, the target was tilted to an angle of 60° from the incident direction of 2 MeV He⫹ and the energy of the backscattered He⫹ was analyzed at a laboratory scattering angle of 170° with the detector of 14 keV energy resolution. This arrangement allowed a surface depth resolution in the range of 10 nm. The sodium migration of the glass was evaluated by using the RUMP program to simulate the experimental spectra.10 III. RESULTS AND DISCUSSION Figure 1 shows the XPS survey spectra of the glass surface after chemical etching with HNO3 . As shown in Fig. 1, the Na and carbon peaks from the glass surface largely deJ. Vac. Sci. Technol. A, Vol. 19, No. 1, JanÕFeb 2001

Sample

C

O

Na

Mg

Al

Si

Ca

Fe

a b c d e f nominal

52.49 24.73 13.87 13.07 17.54 20.10 0

29.18 45.94 54.22 51.70 52.73 50.91 60.22

2.70 2.30 1.77 2.00 0.25 0.34 9.69

0 0 0 0 0 0 1.996

1.24 0.56 1.04 0.84 0.41 0.52 0.546

12.63 25.34 28.63 30.85 28.48 27.78 25.03

0.34 0.63 0.43 0.32 0.21 0.33 2.14

0 0 0 0 0 0 0.0116

creased after chemical etching with HNO3 . Samples 共e兲 and 共f兲, boiled for 30 min in 50% HNO3 and 70% HNO3 , show only very weak Na Auger peaks. We have taken the XPS core-level lines of Si 2p 3/2 共99.3 eV兲, O 1s 共531.0 eV兲, C 1s 共284.5 eV兲, Mg 2p 共49.8 eV兲, Al 2 p 共72.9 eV兲, Ca 2p 3/2 共346.7 eV兲, N 1s 共398.1 eV兲, and Na 1s 共1071.8 eV兲 to obtain atomic composition at the glass surface. XPS spectra were obtained without sputtering to eliminate carbon contamination, since the sodium and calcium are very easily removed by ion beam sputtering.11 The atomic composition at the glass surface was estimated from the XPS peak areas, using a relative sensitivity factors.12 Table I shows the atomic composition at the glass surface as a function of chemical etching conditions. As shown in Table I, sodium composition from all etched glass surfaces are lower than the nominal sodium composition 共9.7%兲 of the glass. These differences may be due to the carbon contamination of the surface and the free sodium present on the internal surface of the glass.13 Sodium composition at the surface that was not chemically etched was 2.7%, but the sodium composition of the glass surface etched at conditions 共b兲, 共c兲, 共d兲, 共e兲, and 共f兲 are 2.3%, 1.77%, 2.0%, 0.25%, and 0.34%, respectively. Chemical etching with HNO3 is a very effective method for Na elimination in the glass surface. Boiling the glass in HNO3 extracts sodium from the glass more efficiently than other techniques. Also, aluminum and calcium at the glass surface were eliminated by chemical etching with HNO3 . Figures 2 and 3 show the XPS Na 1s and Ca 2p peaks from the glass surface with various etching conditions. As shown in Fig. 2, the intensity of Na 1s peaks decreased with chemical etching with nitric acid. The decrease of the Na 1s peak was especially notable when the sample was boiled in 50% and 70% HNO3 solution. But the decrease of the Ca 2p peaks was notable when the sample was boiled in concentrated HNO3 共refer to Fig. 3兲. The depth profiles of the sample Figs. 1共a兲 and 1共f兲 are shown in Figs. 4共a兲 and 4共b兲, respectively. We have only taken the XPS core level lines of Si 2p, O 1s, C 1s, Na 1s, and Ca 2p to obtain atomic concentration in the glass. The profile was obtained by rastering a 3 keV Ar⫹ beam over a 2 mm⫻2 mm surface. As shown in Fig. 4共a兲, the sodium concentration at the surface of the bare glass was found to be 2.2%, it has been reduced to 0% after sputtering for 1 min, and then the sodium concentration slightly increased with increasing sputter time and the concentration approached 5.8%. The depth distribution of the

269


269

FIG. 2. Na 1s XPS spectra from the glass surface after various chemical etching treatments: 共a兲 bare glass, 共b兲 soaked for 6 h in 50% HNO3 , 共c兲 soaked for 6 h in 70% HNO3 , 共d兲 sonicated for 30 min in 50% HNO3 , 共e兲 boiled for 30 min in 50% HNO3 , and 共f兲 boiled for 30 min in 70% HNO3 .

calcium was very similar trends to the sodium distribution in the glass. For the etched glass Fig. 4共b兲 shows that the sodium concentration is less than 1.0% throughout the depth profile. This value is lower compared to the sodium concentration of 2.7% at the bare glass surface. Boiling the glass in HNO3 extracts sodium from the glass. All carbon peaks from the bare and etched glass surface almost disappear after the first sputtering runs which were attributed to organic surface contamination.

FIG. 3. Ca 2p XPS spectra from the glass surface after various chemical etching treatments: 共a兲 bare glass, 共b兲 soaked for 6 h in 50% HNO3 , 共c兲 soaked for 6 h in 70% HNO3 , 共d兲 sonicated for 30 min in 50% HNO3 , 共e兲 boiled for 30 min in 50% HNO3 , and 共f兲 boiled for 30 min in 70% HNO3 . JVST A - Vacuum, Surfaces, and Films

FIG. 4. Depth profiles of the glass: 共a兲 bare glass, and 共b兲 glass boiled in 70% HNO3 solution for 30 min.

Figure 5 shows the AFM surface images of the glass after various chemical etchings. A significant difference is observed in their surface roughness. Surface roughness of the glass etched with HNO3 is rougher than that of the bare glass. In order to get reliable surface roughness data, five different areas of each sample were scanned. The rms surface roughness of bare glass 共a兲 and glass etched at 共b兲, 共c兲, 共d兲, 共e兲, and 共f兲 was 0.58, 6.06, 5.91, 5.44, 6.78, and 5.79 nm, respectively. The surface morphology of the glass samples etched with HNO3 is very similar regardless of etching treatments. Figure 6 shows the dependence of relative transmittance (T etch /T bare) on the etching condition with nitric acid in the wavelength range of 400–800 nm. As shown in Fig. 6, the variation of T etch /T bare could not be detected within the resolution of 0.01%. The optical transmittance of the glass was not affected by the chemical etching with nitric acid. To investigate the thermal stability of the glass, the bare glass 共a兲 and the glass boiled in 70% HNO3 for 30 min 共f兲 were annealed at 300, 400, 500, and 600 °C, respectively for 1 h in an air environment. Figure 7 shows the AFM surface images of the bare glass as a function of annealing temperature. As shown in Fig. 7, the surface roughness of the bare glass was roughened with increasing annealing temperature. The rms surface roughness of the bare glass and the glass annealed at 300, 400, 500, and 600 °C was 0.58, 1.4, 1.8, 2.0, and 3.0 nm, respectively. As shown in Fig. 7共a兲, surface morphology of the bare glass was

270


270

FIG. 5. AFM surface images of the glass after various etching treatments: 共a兲 bare glass, 共b兲 soaked for 6 h in 50% HNO3 , 共c兲 soaked for 6 h in 70% HNO3 , 共d兲 sonicated for 30 min in 50% HNO3 , 共e兲 boiled for 30 min in 50% HNO3 , and 共f兲 boiled for 30 min in 70% HNO3 .

very smooth. But the surface of the glass annealed at 300 °C came to show small protrusions 关refer to Fig. 7共b兲兴. The diameters of the protrusions ranged from 0.15 to 0.25 ␮m and their height ranged from 3.5 to 8.5 nm. At the annealing temperature of 400 °C, the number of small protrusions at the glass surface were increased 关refer to Fig. 7共c兲兴 and their size

FIG. 6. Dependence of relative transmittance (T etch /T bare) in the wavelength range of 400–800 nm on etching treatments with nitric acid. J. Vac. Sci. Technol. A, Vol. 19, No. 1, JanÕFeb 2001

was slightly decreased to 0.1–0.18 ␮m, but the height of the protrusions was greatly increased to 9–20 nm. Small protrusions at the surface of the glass were greatly increased with an annealing temperature of 500 °C 关refer to Fig. 7共d兲兴. Their size was slightly decreased to 0.1–0.15 ␮m, and the height of the protrusions greatly reduced to 1–5.4 nm. The surface of the glass annealed at 600 °C was dramatically changed. The protrusions at the glass surface completely disappeared, and tiny voids with a diameter of about 0.15 ␮m and a depth ranging from 2 to 4 nm appeared. The change of the surface morphology might be due to the volatility of the sodium atom at the glass surface. However, a significant difference is observed in the surface morphology of the glass boiled for 30 min in a 70% HNO3 solution with an annealing temperature 共refer to Fig. 8兲. These significant differences in their surface morphology with annealing temperature might be due to the chemical etching with nitric acid, so that the surface composition of the glass is different with that of the bare glass. The rms surface roughness of the glass boiled in a 70% HNO3 solution for 30 min and the glass annealed at 300, 400, 500, and 600 °C was 7.5, 10.2, 2.3, 29, and 4.6 nm, respectively. Figures 9 and 10 show the XPS survey spectra of the bare

271


271

FIG. 7. AFM surface images of the bare glass as a function of annealing temperature: 共a兲 as received, 共b兲 300 °C, 共c兲 400 °C, 共d兲 500 °C, and 共e兲 600 °C.

glass and the glass boiled for 30 min in a 70% HNO3 solution with an annealing temperature. As shown in Figs. 9 and 10, the Na 1s XPS peaks and Na Auger peaks from the glass surface increased with annealing temperature and the Na peak at the glass surface which was boiled for 30 min in a 70% HNO3 solution is smaller than that of the bare glass up to annealing temperature of 500 °C. We have taken the XPS core level lines of Si 2p 3/2 共99.3 eV兲, O 1s 共531.0 eV兲, C 1s 共284.5 eV兲, Al 2p 共72.9 eV兲, Ca 2p 3/2 共346.7 eV兲, and Na 1s 共1071.8 eV兲 to obtain atomic composition at the glass surface. Figures 11–13 show the XPS spectra of Na 1s, Ca 2p, and Al 2p, respectively. As shown in the Fig. 11, the height of the Na 1s peaks from the bare glass greatly increases with an annealing temperature of 300 °C 关Fig. 11共A兲兴, but the height of the Na 1s peaks from the glass boiled for 30 min in a 70% HNO3 solution slightly increases with an annealing temperature of 500 °C 关Fig. 11共B兲兴. Ca 2p peaks from the glass and the glass boiled for 30 min in a 70% HNO3 solution shows the same trends as those observed for the Na 共refer to Fig. 12兲. Also, Al 2p peaks from the bare glass appeared at room temperature and increased with an annealing temperature up to 600 °C, but Al 2p peaks from the glass boiled for 30 min in a 70% HNO3 solution did not appear until the annealing temperature reached 500 °C. Tables II共a兲 JVST A - Vacuum, Surfaces, and Films

and II共b兲 show the variation of the chemical composition of the bare glass and the glass boiled for 30 min in a 70% HNO3 solution with annealing temperature. As shown in Table II共a兲 and II共b兲, sodium, aluminum, and calcium at the surface of the glass tend to increase with the annealing temperature regardless of chemical etching with nitric acid, but the Na content at the surface of the glass boiled for 30 min in a 70% HNO3 solution is much lower than that of the surface of the bare glass with an annealing temperature of 500 °C. Also, the carbon content at the surface decreased with the annealing temperature regardless of chemical etching with nitric acid. This is due to the desorption of various hydrocarbon molecules from the glass surface. Figures 14 and 15 show that the RBS spectra of the bare glass and the bare glass annealed at 600 °C for 1 h in air, and the RBS spectra of the glass boiled for 30 min in a 70% HNO3 solution and the glass annealed in air at 600 °C for 1 h after boiling the glass for 30 min in a 70% HNO3 solution, respectively. The surface of the O 共0.725 MeV兲, Na 共0.995 MeV兲, Si 共1.133 MeV兲, and Ca 共1.344 MeV兲 are displayed in Figs. 14 and 15. As shown in Fig. 14, the straight line from the bare glass is slightly different from the dashed line from the bare glass annealed at 600 °C for 1 h in air on the range about 0.98 MeV. The stoichiometry and the concentration

272


272

FIG. 8. AFM surface images of the glass boiled for 30 min in 50% HNO3 as a function of annealing temperature: 共a兲 as received, 共b兲 300 °C, 共c兲 400 °C, 共d兲 500 °C, and 共e兲 600 °C.

across the glass were obtained from these spectra using the RUMP code. As shown in Fig. 14 共inset兲 and Fig. 15 共inset兲, the sodium from the bare glass and the glass boiled in 70% HNO3 were outdiffused to the surface during the annealing temperature of 600 °C, and the sodium content of the bare glass and the glass boiled for 30 min in a 70% HNO3 solution were about 5% and 2.5% within 50 nm from the surface.

FIG. 9. XPS survey spectra of the bare glass with annealing temperatures: 共a兲 as received, 共b兲 300 °C, 共c兲 400 °C, 共d兲 500 °C, and 共e兲 600 °C. J. Vac. Sci. Technol. A, Vol. 19, No. 1, JanÕFeb 2001

In addition, the sodium content of the bare glass and the glass boiled for 30 min in a 70% HNO3 solution annealed at 600 °C in air for 1 h were almost the same value of 9%. The difference between the values from RBS and XPS are due to the depth resolution of RBS and XPS. The depth resolution of XPS is only a few nanometers.14 XPS can measure the

FIG. 10. XPS survey spectra of the glass boiled for 30 min in 70% HNO3 with annealing temperatures: 共a兲 as received, 共b兲 300 °C, 共c兲 400 °C, 共d兲 500 °C, and 共e兲 600 °C.

273


273 TABLE II. Chemical composition of glass with annealing temperature. Part 共a兲 is bare glass. Part 共b兲 is sample 共f兲共glass boiled for 30 min in 70% HNO3 ).

FIG. 11. Na 1s XPS spectra of the bare glass 共A兲 and the glass boiled for 30 min in 70% HNO3 共B兲 with annealing temperatures: 共a兲 as received, 共b兲 300 °C, 共c兲 400 °C, 共d兲 500 °C, and 共e兲 600 °C.

C

O

Na

Mg

Al

Si

Ca

Fe

共a兲 RTa 300 °C 400 °C 500 °C 600 °C

52.49 40.49 26.89 35.65 25.81

29.18 39.71 46.28 41.32 48.55

2.7 3.60 5.55 4.55 4.28

0 0 0 0 0

1.24 1.33 1.09 1.11 1.06

12.63 14.88 19.77 16.73 19.87

0.34 0 0.42 0.65 0.43

0 0 0 0 0

共b兲 RT 300 °C 400 °C 500 °C 600 °C

35.91 20.97 30.79 19.98 25.03

43.60 52.20 45.72 53.06 50.14

1.20 2.66 2.63 2.82 4.63

0 0 0 0 0

0.17 0.23 0.45 0.32 0.26

19.06 23.85 19.97 23.49 19.58

0.06 0.08 0.45 0.33 0.37

0 0 0 0 0

a

RT is room temperature.

carbon contamination of the surface, however, it cannot measure the free sodium present on the internal surface of the glass.13 Boiling glass in a concentrated HNO3 solution effectively reduces, Na, Al, and Ca in the glass surface, thus improving the thermal stability of the glass. IV. CONCLUSIONS

FIG. 12. Ca 2p XPS spectra of the bare glass 共A兲 and the glass boiled for 30 min in 70% HNO3 共B兲 with annealing temperatures: 共a兲 as received, 共b兲 300 °C, 共c兲 400 °C, 共d兲 500 °C, and 共e兲 600 °C.

FIG. 13. Al 2p XPS spectra of the bare glass 共A兲 and the glass boiled for 30 min in 70% HNO3 共B兲 with annealing temperatures: 共a兲 as received, 共b兲 300 °C, 共c兲 400 °C, 共d兲 500 °C, and 共e兲 600 °C. JVST A - Vacuum, Surfaces, and Films

Slide glass specimens were chemically etched with nitric acid under various etching conditions. Sodium, aluminum, and calcium at the glass surface were effectively reduced by boiling the glass in nitric acid for 30 min. Surface morphology of the glass samples were very similar regardless of chemical etching condition with nitric acid. The dependence of the relative transmittance (T etch /T bare) on etching condition with nitric acid in the wavelength range of 400–800 nm is nearly constant at unity. Sodium, aluminum, and calcium at the surface of the glass increase with the annealing temperature regardless of chemical etching with nitric acid, but the sodium, aluminum, and calcium content at the surface of

FIG. 14. RBS spectra of the bare glass 共 600 °C for 1 h in air 共– – –兲.

兲 and the bare glass annealed at

274


274

ACKNOWLEDGMENT This work was supported by the Brain Korea 21 Project.

H. K. Pulker, Coating on Glass 共Elsevier, Amsterdam, 1984兲. H. Rawson, Properties and Applications of Glass 共Elsevier, Amsterdam, 1980兲. 3 I. Blech, H. Sello, and L. V. Gregor, in Thin Films in Integrated Circuits, edited by L. I. Maissel and R. Glang 共McGraw-Hill, New York, 1970兲, pp. 22–23. 4 I. Keesmann and Z. A. Allgem, Chem. 346, 30 共1966兲. 5 H. Bach and H. Schroeder, Thin Solid Films 1, 255 共1967/68兲. 6 A. Cousen, J. Soc. Glass Technol. 20, 418 共1936兲. 7 E. L. Mochel, M. N. Nordberg, and T. E. Elmer, J. Am. Ceram. Soc. 49, 585 共1966兲. 8 Y. Paz and Z. Luo, J. Mater. Res. 10, 2842 共1995兲. 9 Y. Paz and A. Heller, J. Mater. Res. 12, 2759 共1997兲. 10 L. R. Doolittle, Nucl. Instrum. Methods Phys. Res. B 15, 227 共1987兲. 11 H. K. Pulker, Coating on Glass 共Elsevier, Amsterdam, 1984兲, p. 93. 12 J. F. Moulder, W. F. Stickle, P. E. Sobol, and K. D. Bomben, Handbook of X-ray Photoelectron Spectroscopy 共Perkin-Elmer, Eden Prairie, MN, 1992兲. 13 S. M. Budd and J. J. Kirwan, Advanced in Glass Technology 共Plenum, New York, 1962兲, Vol. 1, pp. 527–540. 14 F. Abulfotuh and L. L. Kazmerski, in Thin Film Technology Handbook, edited by A. A. R. Elshabini-Riad and F. D. Barlow III 共McGraw-Hill, New York, 1997兲, pp. 6–48. 1 2

FIG. 15. RBS spectra of the glass boiled for 30 min in 70% HNO3 共兲 and the glass boiled for 30 min in 70% HNO3 annealed at 600 °C for 1 h in air 共– – –兲.

the glass boiled for 30 min in a 70% HNO3 solution is smaller than that of the bare glass annealed at a temperature of 500 °C. Boiling glass in a concentrated HNO3 solution effectively reduces, Na, Al, and Ca in the glass surface, thus improving the thermal stability of the glass.

J. Vac. Sci. Technol. A, Vol. 19, No. 1, JanÕFeb 2001