Coupled Recycling of PVC and Glass Wastes Producing Chlorine

Nov 21, 2008 - Coupled Recycling of PVC and Glass Wastes Producing Chlorine-Free Fuels and Cement Feed Stock. Ho-jin Sung*, Reiji ... Corresponding au...
0 downloads 3 Views 4MB Size
Environ. Sci. Technol. 2009, 43, 47–52

Coupled Recycling of PVC and Glass Wastes Producing Chlorine-Free Fuels and Cement Feed Stock H O - J I N S U N G , * ,†,‡ R E I J I N O D A , †,§ A N D MASAYUKI HORIO† Plant Engineering Center, Institute for Advanced Engineering, Woncheon-dong, Yeongtong-gu, Suwon-si, Gyeonggi-do, 443-749, South Korea, Chemical and Environmental Engineering, Graduate School of Engineering, Gunma University, 1-5-1 Tenjin-cho, Kiryu, Gunma, 376-0052, Japan, and Department of Chemical Engineering, Graduate School of Bio-Applications and Systems Engineering, Tokyo University of Agriculture and Technology, 24-16 Naka-machi 2, Koganei-shi, Tokyo 184-8588, Japan

Received May 13, 2008. Revised manuscript received September 8, 2008. Accepted October 20, 2008.

To utilize PVC and glass wastes in a landfill, a kinetic study was conducted for neutralization of HCl derived from PVC pyrolysis with sodium in soda glass. The effective diffusion coefficient of sodium at 550 °C was 3.3 × 10-16 m2/s in steam atmosphere, but was 1.2 × 10-17 m2/s in dry atmosphere. It was confirmed from experimental results in which considerable NaCl crystals were deposited on the surface of glass particles after 6 h reaction with particles of 25 µm in diameter that NaCl crystals’ growth on the glass surface does not affect the neutralization rate rate in our experimental conditions. The effect of hydrothermal treatment was studied for the glass treated at 250 °C under a vapor pressure of 3.6 MPa for 5 h. Approximately 20 times higher rate than that of original glass was caused by the formation of the porous surface layer through which sodium ions can readily diffuse out. The effect was not clear until steam pressure reached the above value. The absence of chlorine within the glass matrix was confirmed by EDS analysis on the cross section of glass cullets reacted with HCl gas. Neutralization of HCl gas with soda-glass conducted under steam atmosphere to increase the reaction rate is effective to recover energy and material from PVC and glass wastes.

Introduction

free flue gas and char, and found that concentrations of sodium, calcium, and potassium in leachate, which is obtained by washing the glass beads reacted with HCl gas containing steam at 550 °C, were 98.2, 1.6, and 0.2 mol%, respectively. Furthermore, Horio et al. (3) proposed to apply the HCl neutralization process with glass wastes to the cement manufacturing process where low level waste heat good for PVC pyrolysis is available and fuel gas, char, and silica-rich glass byproduct can all be utilized in the cement process. The size of the neutralizer, which is a core component of the proposed process, is highly dependent on the neutralization kinetics. Although it has been known that the neutralization rate is controlled by diffusion, some information is still lacking about factors affecting the kinetics, such as reaction atmosphere or treatment prescriptions. The self-diffusion of sodium ions in soda-lime-silicate glasses in the atmosphere without steam have been studied with a radioactive tracer by Johnson et al. (4), Wilson and Carter (5), Williams and Heckman (6), and Terai et al. (7). However, since the surface of a glass particle is attacked by HCl gas and steam in the neutralizer, the diffusion behavior of sodium ions in the glass is not purely a self-diffusion but a complex one with an interaction between the glass matrix and the mixed gas of steam and HCl. Dealkalization processing of bottles and flat glasses with SO2, HF, and HCl gases for refinement of their surfaces has been well-known for decades (8-16). However, to the best of our knowledge, there is no literature available to determine the neutralization rate of sodium ions in soda-lime silicate glass particles in the coexistence of HCl and steam. Therefore, the primary objective of the present work is to determine the effective diffusion rate of sodium ions in the glass particles under the steam atmosphere containing HCl gas and subsequently to investigate the effect of different factors affecting its rate.

Experimental Procedure A single quartz tube of 30 mm i.d. and 710 mm long (heating area, 470 mm) with a porous quartz plate in its middle part is used as a differential reactor to evaluate the neutralization kinetics of hydrogen chloride and soda-lime-silicate glass (Figure 1). Glass samples used are glass cullets and glass beads with a nonspherical and a spherical shape, respectively, to evaluate similarity in their reaction rates. The chemical compositions of both of glass samples determined by X-ray fluorescence spectrometer (RIX 3000, Rigaku Corp.) indicated their similarity with soda–lime-silicate glasses, except for the contents of MgO and Cr2O3 (Table 1).

By examining the material flow of sodium and chlorine in the society, Horio et al. (1) suggested that PVC and glass wastes so far disposed to landfill sites can be used to neutralize each other. From stoichiometry of sodium and chlorine in the material flow, Sung et al. (2) found that the amount of glass wastes can neutralize about 60% of PVC wastes in Japan on the national level. By means of a laboratory scale glassfixed bed, it was successfully demonstrated that HCl generated from flash pyrolysis of PVC could be completely neutralized with soda–lime-silicate glass producing a chlorine* Corresponding author phone: +82-31-219-2688; fax: +82-31216-9125; e-mail: [email protected]. † Tokyo University of Agriculture and Technology. ‡ Institute for Advanced Engineering. § Gunma University. 10.1021/es800599y CCC: $40.75

Published on Web 11/21/2008

 2009 American Chemical Society

FIGURE 1. Schematic of the neutralization reaction apparatus. VOL. 43, NO. 1, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

47

FIGURE 2. Elemental distributions on glass surface and within glass matrix after the reaction with 0.28% HCl gas containing steam for 3600 s at 550 °C.

TABLE 1. Chemical Compositions of Glass Samples Used concentration [wt%]

a

component

glass cullets

glass beads

SiO2 Al2O3 Na2O K2O CaO MgO Cr2O3 Fe2O3 TiO2 MnO Co2O3 CuO SO3 Cl Others

68.1 2.27 13.2 1.21 13.7 0.58 0.12 0.29 0.06 0.02 0.03 0.02 0.15 0.09 0.16

68.3 2.14 12.4 0.72 11.1 3.54 0.02 0.17 0.09 0.02 0.02 a ND 0.25 0.11 1.12

)Not detected.

HCl Neutralization Reaction with Glass Cullets. The experimental conditions for glass cullets to evaluate the reaction rate of sodium are shown in Table 2. Particle size distribution of glass cullets used for the experiment is taken in the range of 355-600 µm in mesh opening size (-31 mesh and +44 mesh). When the effective diffusion coefficient of sodium is determined, glass cullets are assumed as spherical particles to simply the calculation. Although the calculation of effective diffusion coefficient of sodium with glass beads is more accurate than that with glass cullets, we need to know it from glass cullets for commercial consideration. However, the affordability of the reaction rate derived has to be confirmed by comparison between glass cullets and glass beads in the rate. Approximately 900 mg of samples are filled on the quartz plate to form a thin glass bed and heated up to 550 °C under N2 atmosphere to avoid their softening and agglomeration. Since in the previous experiment (2) for the temperature of 450, 500, 550 °C the amount of chlorine captured by glass beads increases with temperature, the experimental temperature for the neutralization kinetics is limited to 550 °C. The temperature of glass on the quartz plate is then measured by a K-type thermocouple, which is covered with the quartz 48

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 1, 2009

FIGURE 3. Fraction of extracted sodium from the glass cullets reacted with the mixed gas of HCl and steam at 550 °C together with effective diffusion coefficient of sodium. tube of 3 mm of inner diameter to prevent corrosion. As soon as the temperature of glass is reached at 550 °C, N2 gas in the reactor is replaced with steam for 300 s and then the mixed gas of steam, HCl and N2 (balance gas) is passed through the glass bed from the top of the reactor at a constant velocity. Constant velocity is calculated with total gas flow rate divided by area of the reactor, and the total gas flow rate is determined so that the mixed gas can be arrived at the fixed-bed of glass within 10 s. Each flow rate of HCl gas, N2, and water is controlled with gas flow meters and peristatic tube pump, respectively. Effect of HCl concentration on the neutralization rate in the presence of steam is evaluated with 0.07, 0.14 and 0.28 vol% HCl. At the end of a predetermined residence time, glass samples are cooled down under N2 atmosphere. Sodium chloride deposited on the surface of the reacted glass is washed off with 50 mL of purified water in a shaker for 600 s. The leachates are then filtered with membrane syringe filter (pore size: 0.45 µm) and the amount of chlorine is determined by ion chromatography (DX120, Dionex Corp.). Surface morphology and elemental distributions on the surface and inside of the glass are observed by FE-SEM (JSM6335F, Jeol Ltd.) equipped with an energy dispersive X-ray system (EDS) at an accelerating voltage in the range of 4-10 kV and at an emission current in the range of 7-12 µA. HCl Neutralization Reaction with Glass Beads. The experimental conditions for glass beads to evaluate similarity between glass cullets and glass beads in the reaction rate (no. GB-1), reaction atmosphere (no. GB-2),

TABLE 2. Experimental Conditions for Glass Cullets GC # 1 sample reaction temperature [°C] reaction time [s] mixed gas velocity at 550 °C [m · s-1] concentration [vol%] HCl steam N2 glass cullets size [µm] glass cullets quantity [kg]

GC # 2

GC # 3

glass cullets 550 60-3600 1.43 × 10-1

glass cullets 550 60-3600 1.43 × 10-1

glass cullets 550 60-3600 1.43 × 10-1

0.07 62.04 37.89 355-600 about 9 × 10-4

0.14 62.04 37.82 355-600 about 9 × 10-4

0.28 62.04 37.68 355-600 about 9 × 10-4

GB no. 1

GB no. 2

GB no. 3

glass beads 550 300-7200 1.37 × 10-1

glass beads 550 900-3600 1.37 × 10-1

glass beads 550 3600-21600 1.37 × 10-1

1.93 61.56 36.51 125-150 about 5 × 10-4

1.93 0 98.07 125-150 about 5 × 10-4

1.93 61.56 36.51 3-43 about 5 × 10-4

TABLE 3. Experimental Conditions for Glass Beads

sample reaction temperature [°C] reaction time [s] mixed gas velocity at 550 °C [m · s-1] concentration [vol.%] HCl steam N2 glass beads size [µm] glass beads quantity [kg]

NaCl crystals deposited on the fine glass particles with long reaction time (no. GB-3) on the neutralization kinetics are shown in Table 3. Particle size distribution of coarse glass used for the experimental condition no. GB-1 is taken in the range of 125-150 µm. Dry and steam atmospheres containing 1.93% HCl are adopted for the experimental condition no. GB-2. Particle size distribution of fine glass

used for the experimental condition no. GB-3 is taken in the range of 3-43 µm, and the reaction time is up to 6 h. Hydrothermal Treatment of Glass Beads. In order to improve the neutralization rate of sodium in glass, three kinds of preconditioning treatments are conducted as follows: (1) Hydrothermal treatment with water at high vapor pressure (HWH treatment): coarse glass beads with an average particle size of 460 µm are heated with ultra purified water at 3.5 MPa and 250 °C for 5 h in an autoclave. (2) Hydrothermal treatment with water at low vapor pressure (HWL treatment): coarse glass beads are heated with ultra purified water at 1.5 MPa and 300 °C for 5 h in an autoclave. (3) Hydrothermal treatment with alkali solution at low vapor pressure (HAL treatment): coarse glass beads are heated with 0.5N NaOH solution at 1.3 MPa and 200 °C for 5 h in an autoclave. Due to the difference in amount of water for each preconditioning treatment, pressure in an autoclave does not increase proportional to temperature. As mentioned previously, the treated glasses together with original coarse glass are reacted with the mixed gas of 1.93% HCl and steam at 550 °C.

FIGURE 4. Fraction of extracted sodium and effective diffusion coefficient from the glass cullets and glass beads in steam atmosphere and the glass beads in dry atmosphere.

Results and Discussion

FIGURE 5. Fraction of extracted sodium and effective diffusion coefficient from the coarse and fine glass particles.

Elemental Distributions on the Surface and within the Glass Particle. Figure 2 shows distribution of sodium, chlorine, and silicon on the surface and within the glass cullets reacted with 0.28% HCl gas containing steam at 550 °C for 3600 s. Samples were prepared using a Cross-Section Polisher (SM-09010, JEOL Ltd.), comprising of a broad argon ion beam to polish cross-sections of specimens. It is important to note that even though NaCl crystalline solids are present on the outlet of the glass, the inner region (the cross-sectional part of glass as shown in Figure 2) shows absolutely little chlorine. This indicates that washing-off of NaCl from its surface, the dealkalized glass cullets can be utilized as a feedstock for cement production. Amount of Extracted Sodium from Glass Cullets. Fraction of extracted sodium from the glass cullets against square root of reaction time under the coexistence of steam and VOL. 43, NO. 1, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

49

FIGURE 6. NaCl crystal growth on the surface of the fine glass beads reacted with 1.93% HCl gas containing steam at 550 °C for up to 6 h. Solving eq 3 subject to the boundary condition, C (R, t) ) 0 (t > 0), C (0, t) ) finite and the initial condition, C (r, 0) ) Ci (r ) 0 ∼ R), the following solution is obtained: (-1) exp(( 2R πr )∑ n ∞

n

C(r, t) ) -Ci

n)1

(nπ)2Defft R2

)

sin

(4) ( nπr R )

By integrating the surface flux over t ) 0 ∼ t, the fraction of extracted sodium from the glass is given by ∞

X(t) ) FIGURE 7. Fraction of extracted sodium from thermally treated-glass beads by water and alkali solution at high or low vapor pressure. HCl gas at 550 °C is shown in Figure 3. The fraction of extracted sodium (X) is defined as follows: amount of sodium deposited on the surface of a glass(mol ⁄ g) X[-] ) amount of sodium in a glass(mol ⁄ g)

(1)

It is clearly seen that the fraction of extracted sodium increases proportional to t1/2 for most of the region studied, regardless of the concentration of HCl. A widely accepted interpretation of the correlation between X and t1/2 is a diffusion controlled process (4-7, 10, 15, 16, 18, 19). For such process, the lower fraction of extracted sodium from a coarse glass particle may increase with a decrease in the particle size of the glass. Therefore, the determination of diffusion coefficient of sodium ions in glass particles gives useful information to design a neutralizer with a glass bed, such as should the glass particle size be desirable from a viewpoint of process engineering and how much glass is necessary to completely neutralize HCl gas. Determination of Effective Diffusion Coefficients of Sodium Ions. To determine the effective diffusion coefficient of sodium ions in glass cullets, the fraction of extracted sodium is calculated as follows (details are given in the Supporting Information of this article): The effective diffusion coefficient of sodium ions Deff is defined by the one-dimensional diffusion model J ) -Deff

∂C ∂r

(2)

where J is the flux of sodium ions and ∂C/∂r is concentration gradient at a potion of radial direction r. From the equation of continuity, ∂2C Deff ∂ 2 ∂C ∂C ∂(-J) r ) ) Deff 2 ) 2 ∂t ∂r ∂r r ∂r ∂r

( )

(3)

where t is the reaction time and Deff is independent of concentration. 50

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 1, 2009

[

(

(nπ)2Defft 6 1 1 exp π2 n)1 n2 R2



)]

(5)

Effective diffusion coefficient of sodium ions Deff in the glass cullets calculated by fitting eq 5 to the profiles of extracted sodium (cf. Figure 4) is about 3.3 × 10-16 m2/s at 550 °C under the atmosphere containing steam and HCl gas. Experimental error of the fraction of extracted sodium is within (18% as compared with those predicted by eq 5 indicating that Deff calculated may well represent the neutralization kinetics in the conditions studied. However, it should be mentioned that the effective diffusion coefficient of sodium calculated at 550 °C under steam atmosphere was lower by several orders of magnitude than the self-diffusion coefficients of sodium obtained by a tracer-diffusion study in soda-lime-silicate glasses, about 8.0 × 10-13 m2/s at 547 °C (4) and 10-15 - 10-14 m2/s at 350 °C (4-7). This may be attributed to the difference in the mechanism between the two diffusion behaviors. In the case of a dealkalization process with acid gases (e.g., HCl, SO2 + steam) or water, the electrical neutrality of glass is preserved by the inward diffusion of hydrogen ions (or hydronium ions). Therefore, the fraction of extracted sodium is supposed to be controlled by the rate of interdiffusion of ions in the glass. Similarity between Glass Cullets and Glass Beads in the Neutralization Kinetics. Glass beads (dp ) 125 - 150 µm) with spherical shapes were compared with glass cullets with nonspherical shapes in the fraction of extracted sodium. The fraction of extracted sodium from both the glass cullets and the glass beads by HCl gas under steam atmosphere is plotted in Figure 4 against square root of time, which is divided by particle sizes used to simplify the neutralization rate. It is quite interesting to see that the fractions of extracted sodium for both glasses lie in the same line holding the relation between X and t1/2 and the effective diffusion coefficient of sodium calculated (Deff ) 3.3 × 10-16 m2/s, error range: (9%) is very similar to that for the glass cullets. These data suggest that the diffusion constant determined by assuming that glass cullets are spherical particles can be applied to both the glass cullets and the glass beads. Effect of Reaction Atmosphere on the Neutralization Kinetics. Effect of reaction atmospheres, i.e., dry atmosphere and steam atmosphere, on the fraction of extracted sodium was evaluated with 1.93% HCl gas. The fraction of extracted sodium from the glass beads at 550 °C for both reaction

FIGURE 8. Morphological images of the surface of the glass beads and NaCl crystals before and after reaction with 1.93% HCl gas containing steam at 550 °C. The glass beads are treated with ultra purified water at 3.5 MPa and 250 °C for 5 h in an autoclave. atmospheres are also plotted in Figure 4 against t1/2/dp. When steam atmosphere is compared with dry atmosphere in the fraction of extracted sodium, it must be emphasized that the fraction of extracted sodium for the steam atmosphere is significantly higher than that for the dry atmosphere, even though the concentration of HCl is the same for both reaction atmospheres. Moreover, the effective diffusion coefficient of sodium calculated for the dry atmosphere is approximately 1.24 × 10-17 m2/s (error range: -20 ∼ +10%), which is about 27 times lower than that of the steam atmosphere. Therefore, neutralization of HCl gas with soda-glass should be conducted under steam atmosphere to increase the yield of extracted sodium. This difference may be attributed to dehydration reaction accompanying a compacting of the glass network in atmosphere containing less amount of hydrogen ions. Theoretically, the removal of positive sodium ions from the glass at elevated temperature would give rise to a negative charge in the glass, and a positive charge would accumulate in polar compounds (e.g., water and HCl). The unbalanced positive ions (proton) in the polar compound will be attracted to the glass surface, and the electrical neutrality is preserved (10). After the exchange of sodium ions in glass with hydrogen ions from hydrogen chloride is preceded, the formation of ≡Si-OH groups takes place as follows: 2HCl + 2 ≡ Si-ONa T 2NaCl + 2 ≡ Si-OH

(6)

However, Douglas and Isard (10) after studying the action of water and SO2 on glass surfaces reported that the hydrogenglass may separate into a water phase and a silicate phase and H2O molecules formed from the H+ ions and the oxygen of the glass network evaporate at the surface resulting in a compacting of the glass network: 2 ≡ Si-OH T ≡ Si-O-Si ≡ +H2O

(7)

They suggested that the interdiffusion of sodium ions with hydrogen ions was much slower than the dehydration process and that the compacting process did not alter the relation between the amount of extracted sodium and t1/2. If dealkalization process with SO2/steam mixed gas is similar to that with HCl gas, the fast compacting process and the high resistance of the compacted layer can explain the reduced fraction of extracted sodium under the dry atmosphere, while holding the X - t1/2 relation true. If the atmosphere surrounding the glasses is saturated with steam, the dehydration will be inhibited and compacting will not take place eq 8. H2O + HCl + ≡ Si-ONa T NaCl + ≡ Si-OH + H2O

(8)

Effect of NaCl Crystals Deposited on the Neutralization Kinetics. The fraction of extracted sodium from two kinds of glass particles with different particle sizes is plotted against t1/2/dp in Figure 5. Corresponding SEM images of fine glasses reacted with HCl gas are shown in Figure 6. It is evident that the fraction of extracted sodium from the fine glass particles of 25 µm is about 5 times higher than that for the coarse glass particles of 138 µm. Moreover, even though NaCl crystals are considerably deposited on the surface of glass particles (cf. Figure 6), the fraction of extracted sodium is proportional to the square root of time for a long reaction time. These data indicate that NaCl crystals deposited on the glass surface does not affect the neutralization reaction rate in our experimental conditions. Therefore, the fine particles should be utilized to increase the amount of extracted sodium as long as the cost of crushing of glass is economically viable. Effect of Hydrothermal Treatment of Glass on the Neutralization Kinetics. Figure 7 shows effects of hydrothermal treatments with water and alkali solution at high or low vapor pressure on the fraction of extracted sodium from glass beads. It is interesting to note that after hydrothermal treatment of glass beads with water at high vapor pressure (HWH), the surface of glass beads was found to be porous and hazy in nature as well as their leachate indicated a high pH value (Figure 8a). Clearly, after reacting with the mixed gas of HCl and steam, the fraction of extracted sodium and the effective diffusion coefficient from the glass treated at high vapor pressure (HWH treatment) was about 5 and 20 times higher than those for the original glass and the glasses treated at low vapor pressure (HWL and HAL treatments). The increased reaction rate seems to be attributed to the porous surface layer through which sodium ions can be readily diffused out to the glass (cf., Figure 8a). However, the porous layer is formed only on the surface of glass. Thus after the reaction, it is found that glass surface has been clearly separated into the porous layer having a lot of NaCl crystals and the nonporous layer having small amount of them (cf., Figure 8b). If reactor size and cost of the hydrothermal treatment are economically viable, the glass which has a porous surface layer and alkali solution can be utilized to increase the amount of neutralization of HCl.

Acknowledgments We gratefully acknowledge the financial support from Grantsin-Aid for Scientific Research (A) (No. 17201017), Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan and New Energy and Industrial Technology Development Organization (NEDO) under Matching Fund Project VOL. 43, NO. 1, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

51

on “Establishment of technology for cement feedstock by neutralization of glass and PVC waste”. We also thank Mr. A. Honya for cooperation with the experiment.

NOMENCLATURE C) concentration of sodium [mol · kg-1] C (r, t) ) concentration of sodium at r ) r, t ) t [mol · m-3] initial sodium concentration of glass particle Ci ) [mol · kg-1] sodium concentration at glass surface [mol · kg-1] Cs ) particle diameter [µm] dp ) effective diffusion coefficient of sodium ions Deff [m2 · s-1] F (t) ) total amount of sodium deposited on a glass surface [mol] J (t) ) molar flow rate of sodium on a glass surface [mol · s-1] r) potion of radial direction [m] R) radius of glass particle [m] t) reaction time [s] X (t) ) fraction of extracted sodium from a glass [-] ξ) dimensionless radius [-] θ) dimensionless time () tD/R2) [-] dimensionless concentration [-] φC )

Supporting Information Available Mathematical derivations of the effective diffusion coefficient of sodium ions in soda-lime-silicate glass. This material is available free of charge via the Internet at http://pubs.acs. org.

Literature Cited (1) Horio, M.; Noda, R.; Sung, H.-J. Method and equipment to produce useful gas from halogen-containing combustible materials and alkali-containing substances. Japan Patent 3878994, November 5, 2006. (2) Sung, H.-J.; Noda, R.; Horio, M. Thermal treatment of waste PVC and chlorine neutralization by waste glass. J. Chem. Eng. Jpn. 2005, 38, 220–228.

52

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 1, 2009

(3) Horio, M.; Noda, R.; Sung, H.-J., Sato, K.; Sueoka, T. Production process of fuel and raw material by neutralization of glass wastes and halogen- containing plastic wastes. Japan Patent Disclosure 2006- 229642, August 25, 2006. (4) Johnson, J. R.; Bristow, R. H.; Blau, H. H. Diffusion of ions in some simple glasses. J. Am. Ceram. Soc. 1951, 34, 165–172. (5) Wilson, C. G.; Carter, A. C. Self-diffusion of sodium ions in a borosilicate glass and a soda-lime glass. Phys. Chem. Glasses 1964, 5, 111–112. (6) Williams, E. L.; Heckman, R. W. Na diffusion in soda-limealuminosilicate glasses. Phys. Chem. Glasses 1964, 5, 166–171. (7) Terai, R.; Kitaoka, T.; Ueno, T. Self-diffusion of sodium ions in some silicate glasses. Yogyo Kyokai Shi 1969, 77, 88–94. (8) Coward, J. N.; Turner, W. E. S. The clouding of soda-lime-silica glass in atmospheres containing sulphur dioxide. J. Soc. Glass Technol. 1938, 22, 309–323. (9) Williams, H. S.; Weyl, W. A. Surface dealkalization of finished glassware. Glass Ind. 1945, 26, 290–292. (10) Douglas, R. W.; Isard, J. O. The action of water and of sulphur dioxide on glass surfaces. J. Soc. Glass Technol. 1949, 33, 289– 335. (11) Mochel, E. L.; Nordberg, M. E.; Elmer, T. H. Strengthening of glass surfaces by sulfur trioxide treatment. J. Am. Ceram. Soc. 1966, 49, 585–589. (12) Mahoney, W. P. Treatment of newly formed glass articles. U.S. patent 3 249 246, May 3, 1966. (13) Poole, J. P.; Snyder, H. C.; Ryder, R. J. Corrosion retarding fluorine treatment of glass surfaces. U.S. patent 3 314 772, April 18, 1967. (14) Schaeffer, H. A.; Mecha, J.; Freude, E.; Weiβkopf, K.; Erlangen; Kno¨dler, H. Mu ¨ nchen Entalkalisierung von Na2O-CaO-SiO2glasoberfla¨chen bei einwirkung chlorhaltiger gase. Glastech. Ber. 1981, 54, 247–256. (15) Brow, R. K.; LaCourse, W. Fluorine treatment of soda-lime silicate glass surfaces. J. Am. Ceram. Soc. 1983, 66, C-123. (16) Schaeffer, H. A.; Stengel, M.; Mecha, J. Dealkalization of glass surfaces utilizing HCl gas. J. Non-Cryst. Solids 1986, 80, 400– 404. (17) Senturk, U.; Varner, J. R.; LaCourse, W. C. Structure-hardness relation for high-temperature SO2-dealkalized float glass. J. NonCryst. Solids 1997, 222, 160–166. (18) Terai, R.; Hayami, R. Ionic diffusion in glasses. J. Non-Cryst. Solids 1975, 18, 217–264. (19) Doremus, R. H. Diffusion-controlled reaction of water with glass. J. Non-Cryst. Solids 1983, 55, 143–147.

ES800599Y