Some content of this application is unavailable at the moment.
If this situation persist, please contact us atFeedback&Contact
1. (GB1181581) Process for Treating Glass
Note: Text based on automatic Optical Character Recognition processes. Please use the PDF version for legal matters
PATENT SPECIFICATION
NO DRAWINGS 1,181,581 /^/.4ixSY Date of Application (No. 8671/67) and filing (Complete Specification: 23 Feb., 1967.
Application made in United States of America (No. 529,215) on 23 Feb., 1966.
Application made in United States of America (No. 604,169) on 23 Dec., 1966.
Complete Specification Published: 18 Feb., 1970.
Index at acceptance:-Cl M(IlBl, 11B3, 11C1, 11C4, 11C5, 11C6, 11C7, 11C9, 11F1, 11F2, 11F3, 11F14, 11F15, 1L1F28, 11F29, 11F33, 1IF31, 11J2, 11J3, 11K1, 11K5, 11K8, 13C, 13E, 13J, 13U2) International Classification: -C 03 c 21/00 COMPLETE SPECIFICATION
Process for Treating Glass We, OWENS-ILLINOIS, INc., a Corporation of the State of Ohio, United States of America of Toledo, Ohio, United States of America (Assignee of EVERETT FRANKLIN GRUBB, AUGUSTUS WILLIAM LADuE and PAUL WHITENER LINIC GRAHAM), do hereby declare the invention, for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement: This invention relates to a process for treating articles of glass, including glass components of articles, to improve the strength of the glass articles, and also relates to the articles resulting from the process. The present invention especially relates to a process for treating silicate glass composed of silica and alkali metal oxide or oxides, with or without one or more of other compatible constituents such as alkaline earth metal oxides, alumina zirconia, titania, and boron oxide, glass-coloring oxides such as oxides of iron, cobalt, nickel, manganese, chromium and vanadium, and fining agents, and also especially relates to the silicate glass article resulting from the present process.
As used herein, the term "glass" means those inorganic glasses that (1) are not controllably crystallizable, and thus can be devitrified as the term is normally used, to form crystalline material usually in a matrix of a glass, the matrix having a composition determined by the initial composition and by the composition of the crystalline material; (2) are controllably crystallized by a heat treatment; or (3) have been controllably crystallized by a heat treatment. Glass that is controllably crystallizable is commonly referred to as thermally crystallizable glass composition. A crystallized glass is commonly referred to as a glass-ceramic.
As described later in detail many types of silicate glasses, including glass-ceramics, that contain alkali metal ions have been treated at an elevated temperature by contact with an alkali metal inorganic salt, for exchange of alkali metal ions in a surface portion of the glass with alkali metal ions of the inorganic salt. The usual process is an immersion of the glass in a molten bath of alkali metal inorganic salt or of a mixture of the alkali metal inorganic salt with other inorganic salts. The time of immersion is usually sufficient to cause this exchange only in a surface layer of the glass article. Lithium ions in a glass have been exchanged alternately with sodium and potassium ions in molten inorganic salt baths. Sodium ions in glass have been exchanged with lithium and potassium ions of molten salt baths containing lithium and potassium inorganic salts.
Some molten alkali metal salts, such as potassium nitrate, are dangerous when used with the salt temperature above 7500F., because they can decompose spontaneously. Also metal nitrate salts, such as potassium nitrate, at such high temperature react very vigorously with organic material. In the manufacture of some glass articles, such as glass containers, molds are used. These molds are periodically lubricated by sprays of hydrocarbon material.
Unless such material is removed from the glass surface prior to contact with molten potassium nitrate for ion exchange treatment, vigorous reaction can occur. It is not feasible to remove such organic material from the outer surface of a glass container immediately after it is taken from the mold, because the temperature of the container is very high.
Also the container must be subjected to an annealing treatment before being cooled to room temperature or to a moderate temperature at which the organic material can be removed by washing.
According to the present invention there rT 4 rGI) F-4 r1,181,581 is provided a process for treating an article composed of an inorganic glass containing al least 2% by weight expressed as soda mole equivalent of an alkali metal oxide which comprises:
(1) applying to the surface of said article a solution or suspension, in water, an organic solvent inert to the salt, or a mixture thereof, of at least one salt of a different alkali metal, thereby to form on at least an area of a surface of said glass of the article a substantially continuous layer of material comprising said salt of said different alkali metal; (2) maintaining said surface area of glass and said layer of material at an elevated temperature above the boiling point of water or the organic solvent or the mixture thereof for a period of time sufficient for an ion exchange to take place between said alkali metal of the glass and said different alkali metal to provide a compressive stress surface layer in, and a strengthening of, the glass article, but for a time insufficient to provide such ion exchange to a substantial degree in the interior portion of the glass of the article and for a time insufficient to provide substantial stress relaxation of said glass with consequent loss of the strengthening; and (3) cooling the glass article to a temperature at which said ion exchange does not occur, said layer of material and said alkali metal salt therein being solid at said elevated temperature and being removable from said glass after said maintenance at said elevated temperature.
The alkali metal salt or mixture of these salts preferably constitutes at least 90% by weight of said layer of material formed on the glass, and preferably the glass of at least the surface layer contains at least 5% alkali metal oxide. Preferably the alkali metal salt is a salt, such as carbonate, that is alkaline in water and that is used alone or with another alkali metal salt, and in the latter case can constitute substantially less than 90% by weight of the layer of material when the molar ratio of alkali metal carbonate to the other salt of the same alkali metal is at least 1: 5.
The layer of material can be removed before or after completion of the cooling step.
In the embodiment in which the elevated temperature used is above the strain point, especially about or above the annealing point, further strength can be imparted to the glass by thermal tempering, which is a process that is well known in the art, and which involves directing substantially cool air to the surface of the hot glass to rapidly cool the surface layer, to provide thereby a compressive stress in the layer with tensile stress in the interior of the glass article. This stress will supplement that provided by the ion-exchange treatment at the elevated temperature.
The elevated temperature maintained in this process for the ion exchange is preferably a maximum of 50 deg. C (90 deg. F) above the Lannealing point of the particular glass being treated, if the glass is of other than glassceramic, which has no annealing temperature as the term is defined. In the case of glassceramic the maximum teperature that is maintained for the ion exchange is about 7500C.
(about 14000F.). The preferred maximum elevated temperature is that of the annealing point of the glass being treated by the process 75 of this invention. The elevated temperature at which the layer of material, containing alkali metal salt, and the glass article on which it is formed are maintained, as mentioned above, is usually a temperature of at least 2000C. 80 but is preferably a temperature at or above the strain point of the glass being treated.
It is necessarily above the strain point when the alkali metal in the glass is larger, as described below. It is especially preferred that 85 this elevated temperature be between the annealing point of the glass being used and a temperature between 50 deg. C (90 deg. F) below the annealing temperature.
The layer and the alkali metal salt are such 90 that neither melts nor decomposes at or below the elevated temperature used and neither the glass layer nor the salt melts nor decomposes at a temperature of 4250C. (about 8000F.). The alkali metal of the salt is preferably an alkali metal that is immediately adjacent to said alkali metal in glass in the arrangement of alkali metals as Group I of the Periodic Table of the Elements.
When the alkali metal in the glass has a 100 larger atomic radius than the different alkali metal of the salt, the elevated temperature of the process is above the strain point of the glass being used, to avoid the formation of a surface layer that will crack and separate 105 from the main body of the glass article, for example, when sodium in a glass is replaced by lithium of a lithium salt salt layer on the glass surface.
When the alkali metal in the glass has a 110 smaller atomic radius than the different alkali metal in the salt, the elevated temperature can be below the strain point of the glass but is preferably at least that of the strain point of the glass being used, and it is especially preferred that the elevated temperature be at about, but not above, the annealing point of the glass being used.
The period of time for which the layer and the area of the glass article are maintained 120 at the elevated temperature is dependent on various factors, such as (1) the glass composition, including the particular alkali metal in the surface layer, (2) the salt of the different alkali metal, including the nature of 125 the anion of the salt; whether the alkali metals of the glass and the salt are immediately adjacent (in the Periodic Table) and whether the alkali metal of the glass has the smaller or larger ionic radius, (3) the 130 1,181,581 elevated temperature used, including whether this temperature is (a) below the strain point or (b) at or above the strain point, (4) the depth desired for the compressive stress to be obtained by the process, and (5) whether the salt or mixtures of salts at the highest portion of the range of elevated temperature etches the glass surfaces and whether such etch is desired or is to be avoided or minimized.
Thus the time can range between a few seconds to several minutes at temperatures near 14001F. for glass-ceramic or at a temperature above the annealing point by a maximum of 90 deg. F for the other types of glasses. When the temperature for ion exchange is in the range of from the strain point to about the annealing point of the glass being used, the time can range between a few minutes to a few hours, and the time varies inversely with the temperature. At about the annealing temperature the time is preferably more than five minutes and less than 30 minutes. When the temperature is below the strain point of the glass being treated, the time is between about one-half hour and about 24 hours, with the time also varying inversely with temperature. Usually time periods ranging from 5 minutes to about 4 hours are employed.
In the process of this invention the layer of material consisting essentially of the salt of the different alkali metal can be formed on the area of the glass article by any of numerous techniques. For example, this layer can be formed by spraying this surface area of the glass article with a liquid stream or a spray of liquid droplets in a gas stream, that is a mixture, slurry, suspension, sol or solution containing the alkali metal salt and a liquid, such as water, an organic material or mixtures thereof. The organic material is liquid at the temperature at which the mixture is sprayed onto the glass surface.
The organic material, when used in place of or with water, is an organic compound or mixture of organic compounds that will impart adherence of the alkali metal salt to the glass surface and yet will volatilize, with or without decomposition, or bum off at atemperature at or below the temperature maintained for the ion exchange step of the process, without adversely affecting the adherence of the alkali metal salt to the surface of the glass article so that the alkali metal salt will form a substantially continuous layer of solid material on the sprayed area of the glass surface.
Of course, more than one alkali metal salt may be used, provided the mixture of these salts does not have a melting point at or below the elevated temperature at which the layer and the glass article are maintained for the ion exchange.
Water is the preferred liquid that is used admixed with the alkali metal salt to form via the spray technique a substantially continuous layer of material. If the glass is at the elevated temperature used for the ion exchange step, the water of the spray will volatilize from the glass surface and leave the solid material, consisting essentially of alkali 70 metal salt, as a solid layer on the glass surface. Also, if the liquid spray is used on the glass article at a temperature below that elevated temperature, the water content of the spray will be removed by volatilization as the 75 article and the spray coating thereon are raised to the elevated temperature. It is preferred that the fluid mixture be applied to the glass article which is at a temperature above the boiling point of water to permit 80 by one spray application the formation of the solid layer with a sufficient depth to provide enough alkali metal salt for the desired degree of ion exchange. Thus the temperature of the surface of the glass to which the spray, 85 containing water and alkali metal salt, is applied is preferably above 300OF (above about 1500C), and for certain utilizations of the process it is preferred that the temperature be 6001F. to 9000F. (3150C. to 4850C.). 90 The mixture of liquid, either water and/or organic material, and alkali metal salt contains a sufficient concentration of the latter to provide by a minimum amount of spraying a layer of solid material that will provide 95 the ion exchange to the desired degree. This mixture, that is sprayed, is preferably a solution of the salt in the liquid, either water and/or liquid organic material. Slurries of alkali metal salt and liquid, either water and/ 100 or liquid organic material may be used as the spray material. The preferred solutions are aqueous solutions that are saturated with the salt or the salts at the temperature at which the aqueous solution is used in the spray directed to the glass article.
Another way of forming this layer containing alkali metal salt on the surface area of the glass article is by dipping or immersing this surface portion of the glass article in the 110 fluid mixture, followed by removal of the glass article before its maintenance at the elevated temperature for ion exchange. In this case, fluid mixture adheres to the surface area by dipping or immersion and before or at 115 the elevated temperature water or volatile organic material leaves to leave an adherent layer of solid material consisting essentially of one or more salts of said different alkali metal with or without (preferably without) 120 one or more salts of another alkali metal and/ or other metals. In any event, the solid layer must be of such composition that it remains solid at the elevated temperature utilized for the ion exchange step of the process. 125 For the dipping or immersion technique of formation of the solid layer on the glass, aqueous solutions of the alkali metal salt or 1,181,581 salts are preferred. Especially preferred are saturated aqueous solutions, as mentioned above for the spray technique.
Organic materials used to provide an adherent layer of alkali metal salt that is substantially continuous on the surface area may be chosen from a wide variety of organic compounds, individually or as mixtures. The preferred organic material is a mixture of nitrocellulose and amyl acetate, that is used to provide a suspension or slurry of the alkali metal salt. This suspension or slurry, when it is sprayed on a surface area of a hot glass article, will burn clean and there will be formed a solid, adherent film or layer of alkali metal salt. The organic material used is required, of course, to be a material that does not react with the alkali metal salt.
Other techniques of applying the layer of solid material is by applying the mixture of salt and liquid by a brush or squeegee or through a screen or by placing a porous sheet containing the mixture in its pores against the glass surface.
As described later, there are many types of glass that contain alkali metal oxide that can have alkali metal ions in the surface layer replaced by other alkali metal ions of the salt in the layer formed on the glass. Some of these alkali metal salts, that are suitable for use in the present process, have been used in other ion exchange processes, but only when admixed with further alkali metal salts so that the mixture is liquid at the elevated temperature used heretofore for the ion exchange.
Potassium nitrate has been used in the prior art to replace sodium in a glass with potassium at a temperature at which the potassium nitrate is liquid. Mixtures of potassium nitrate with other potassium salts, such as potassium chloride and potassium sulfate, have been used to provide the liquid ion-exchange bath medium. In such mixtures the potassium nitrate constitutes the predominant mole percentage.
In the process of the present invention some alkali metal salts are used alone, such as potassium sulfate and potassium chloride, which have heretofore been admixed with nitrate salts of the same alkali metal or another alkali metal to provide a liquid ionexchange bath. They have melting points above the elevated temperature at which the ion-exchange step is conducted. They are also used in the present invention with salts of alkali metal that have not been used heretofore, such as carbonate salts of alkali metal, that have melting points above this elevated temperature, provided that the mixture as a layer is solid at the elevated temperature of the present process. For example, in the replacement of at least part of the sodium ions in the surface layer of glass with potassium ions by using potassium salt or salts, the present invention utilizes such potassium salts 65 as potassium carbonate (m.p. of 8910C.), potassium chloride (m.p. of 7900C.), potassium sulfate (m.p. of 10690C.), potassium bromide (m.p. of 7300G.), potassium iodide (m.p. of 7230C.), potassium tribasic phosphate 70 (m.p. of 13400G.) and potassium metaphosphate (m.p. of 8070C.). They are salts of potassium that have melting points above the annealing point of most of the glasses that are ion exchanged by the present process and 75 some of these salts have melting points above the annealing point of all the glasses treated by the process of the present invention that have an annealing point, and above 14000F.
for those glasses that are glass-ceramics. Examples of lithium salts that have suitable melting points are lithium carbonate (m.p. of 6180C.), lithium chloride (m.p. of 6140C.), lithium bromide (m.p. of 5470C.), lithium sulfate (m.p. of 8600C.), and lithium tribasic 85 phosphate (m.p. of 8370C.). When using sodium salt to replace, e.g., lithium or potassium in glass by the present process, suitable salts- are sodium carbonate (m.p. of 8510C.), sodium chloride (m.p. of 8000C.), sodium 90 bromide (m.p. of 7550C.), sodium iodide (m.p.
of 6510C.), sodium sulfate (m.p. of 8840G.), sodium pyrophosphate (m.p. of 9880C.) and sodium metaphosphate (m.p. of 6280C.). The comparable salts of rubidium and cesium can 95 be used, with their suitable high melting points, to replace other alkali metals or each other.
According to another embodiment, the process of the present invention comprises forming on the surface of an article of glass, containing at least 2%, preferably at least 5% by weight of soda, a substantially continuous layer or coating of dipotassium hydrogen orthophosphate; maintaining said surface 105 and said layer of dipotassium hydrogen orthophosphate, at an elevated temperature for a period of time sufficient for some potassium ion to diffuse into and replace a portion of the sodium ion in said glass to provide a 110 compressive stress in said glass, i.e. not all the Na iron need be replaced by K; and cooling the glass article to a temperature at which ion exchange does not occur and removing the layer of potassium salt. The salt 115 layer can be removed prior to or after the completion of the cooling step. The salt can be removed completely by a water wash and no acid treatment need be provided.
More specifically, the dipotassium hydrogen orthophosphate (K2HPO,) is applied to the glass in the form of an aqueous solution having a concentration ranging from 0.1 to 1.3 grams of K1HPO4 per cc of water. Prefera'oly the concentration is greater than 0.4 125 gram per cc of water. The aqueous potassium salt solution should be applied to the glass article at a temperature above the boiling that will not be subjected in service to this severity of abrasion. With such severe abrasion it is necessary that the depth of the compressive stress layer be at least 50 microns. However for certain products, such as glass containers, a depth of compressive stress layer of at least 10 microns is sufficient to retain a substantial degree of increase of strength, afforded by the ion-exchange treatment, during the use and reuse of such products for a reasonable period of time.
Glass bottles or containers can be tested for impact strength using the Preston impact testing machine, Model 400, which is marketed by American Glass Research, Inc., Butler, Pennsylvania. It is described in a pamphlet originally issued by Preston Laboratories, Butler, Pennsylvania. This machine has a pivotally mounted hammer which is raised to an angle relative to a vertical plane. The striking face is a hardened steel ball. The machine can be adjusted to provide impact at any of several locations of a bottle placed in position on a machine. For example, impact can be against the shoulder of the bottle to provide in degrees of pendulum drop, i.e., degrees of drop of the hammer as a measure of the strength. The maximum number of degrees of arcuate movement is the minimum angle that produces breaking of the bottle at the shoulder. Instead of expressing the result in degrees it can be expressed as shoulder impact strength in inch-pounds.
To determine the loss of the strength of bottles as a result of service use, there has been developed a machine that simulates the type of abuse received by bottles when they are fed to and taken away from the filling station in a plant where the bottles are filled with products. This machine is known as the Consumer Line Simulator, also referred to as the CLS Abuser, and is described in report No. 62-127 of December 13, 1962, issued by American Glass Research, Inc., Butler, Pennsylvania. Abuse for one minute using this device is said to be equivalent to the amount of abuse that glass containers receive in one year of service. Of course, this refers to returnable bottles which go through the product filling station a number of times within a year.
The following examples illustrate preferred embodiments of the present invention using various alkali metal salts and using various types of glass, illustrative of those that are ion exchangeable by the present process. The percentages used in this specification are by weight unless otherwise specified.
point of water to permit, by one spray application, the formation of the solid layer with sufficient depth to provide enough potassium ion for the desired degree of ion exchange.
Usually the glass temperature is above 3000F, and preferably is 6000F to 11000F. During the spray application, care should be taken to avoid fracturing the glass article in thermal shock.
The terms " annealing point " and " annealing temperature " as used herein mean the same thing.
The flexural strength or modulus of rupture of glass can be determined by a number of testing methods. A common method uses the Tinius-Olsen testing machine. The glass is tested as a rod that is, for example, 5 inches long and has a diameter of about 136 inch. This machine applies a measured load through a single knife edge to the center of the sample rod supported on two knife edges that are four inches apart (3-point loading). The load is applied at a constant rate of 24 lbs. per min.
until failure occurs with a marker indicating the highest load applied to the point of failure. A dial micrometer calibrated in inches and equipped with a bar contact instead of a point contact is used to measure the maximum and minimum diameters at the center of the sample to an accuracy of 0.0005 inch.
Since few sample rods are perfectly round, the load is applied normal to the maximum diameter and the standard formula for an elliptical cross-section is used in calculating the modulus of rupture as follows:
(10.185) X Load MRD2 XD, MR=Modulus of Rupture D,=Major axis D2=Minor axis The modulus of rupture (=MR) in this formula gives the flexural strength in pounds per square inch of cross-sectional area at failure.
The glass rods are obtained by cutting cane pulled from molten glass.
The flexural strength of rods, with or without ion-exchange treatment, can be determined by this method and compared with the flexural strength of such rods after subjecting them to a substantial degree of abrasion.
Several methods of abrasion have been developed. One example of abrasion comprises tumbling the rods for 15 minutes in a ball mill containing No. 30 silicon carbide grit. This type of abrasion can substantially reduce or eliminate the increase of flexural strength afforded by ion exchange. It is believed that this type of abrasion is much too severe for the evaluation of ion-exchange strengthening of glass that is to be utilized in the form of certain products, such as glass containers, EXAMPLE I
One-way beer bottles, that were made of a 120 conventional flint container glass were sprayed with aqueous salt solutions immediately after the formation of these bottles and thus prior to their annealing. The bottles at the place I us 1,1 81,581 6 1,181,581 of spraying are normally at temperatures between 600cF. and 9000F., as previously determined. The temperature of bottles at the spraying locations varied in temperature profile because they were from different molds of an IS machine and were different styles of bottles.
Some bottles were sprayed on their outside surface with an aqueous solution of potassium carbonate. Some were sprayed on their outside surface with an aqueous solution containing two salts, using 80 mole percent potassium carbonate and 20 mole percent potassium chloride. These solutions were prepared by heating water to boiling temperature and adding salt or a mixture of salts to dissolve as much salt as possible. The aqueous solution was cooled to room temperature to provide saturated solutions in the presence of solid salt. The hot, newly-formed bottles, after being sprayed with these saturated salt solutions, were immediately run through the annealing lehr to receive the normal annealing-temperature treatment that unsprayed bottles receive in their manufacture.
A temperature profile in a normal annealing lehr has been determined using a travelling thermocouple attached to the bottom of a bottle. This temperature profile determination indicates that for about the last one-half of the initial 5 minutes the temperature of the bottom of bottle was rising from about 9801F.
and then in the next 5 minutes the bottom of the bottle was at a temperature between 10000F. and 10250F. After the first 10 minutes the temperature decreases. At the end of about 15 minutes overall, the temperature was reduced to 9000F. and at the end of minutes overall it was reduced to about 6000F. followed by still further cooling. The total travel time through the lehr is about minutes. The temperature in the sidewallportion of the bottle passing through the lehr will be ahead of the temperature of the bottom during the heat-up period and will retain that temperature of the bottom during the cooling period.
The glass of these bottles has an annealing point of 10330F. and a strain point of 9860F.
Untreated This glass has the following theoretical composition, expressed as oxides in weight percent:
Sio2 A1202 CaO MgO Na20 K20 72.0 1.9 9.6 4.2 11.5 0.4 This is a conventional soda-lime-silica container glass. A typical batch composition for 60 this glass is as follows on a weight percent basis:
Sand Soda Ash High Calcite Lime Raw Dolomite Nepheline Syenite Salt Cake 57.1 15.8 5.9 14.6 6.1 0.5 The bottles after exiting from the annealing lehr and cooling to room temperature were washed with dilute nitric acid. An examination of the bottles indicated that they had on their outer surface a 15-micron depth of surface compressive stress layer.
Part of the bottles from each type of sprayand-heat treatment were subjected to 10 minutes of abuse with the CLS Abuser. These and the other ion-exchange bottles that were not subjected to abuse were tested for shoulder impact strength. Bottles of the same type, but that had been through the annealing lehr and had not been sprayed with aqueous salt solution, were tested also for shoulder impact strength. Some of these untreated bottles were abused by the CLS Abuser for 10 minutes prior to testing for impact strength. The degree of pendulum drop increases with increase in shoulder impact strength. The following tabulates the data for shoulder impact strength of the one-way beer bottles untreated and those sprayed with two different salt solutions and heat treated, all without abuse. The other values are the strengths on other bottles of these groups after the 10minute abuse.
K2CO3 4K2CO3/KCI No abuse 790 670 490 Abused 440 630 520 The foregoing data show clearly the increase of strength obtained by the alkali metal salt treatment at the elevated temperature and its substantial retention after a greater amount of abuse than the bottles would be expected to receive. The data also show the advantages of using potassium carbonate alone as compared with the specific mixture of it and 1,181,581 1,181,581 potassium chloride as the alkali metal salt layer on the glass during the elevated temperature treatment.
EXAMPLE II
One-way beer bottles, having the glass composition described above in Example I, were broken into pieces that were melted in a pot.
The molten glass was used to draw cane from which rods were prepared for treatment in accordance with this process. Of course, this glass has the strain point and annealing point mentioned in Example I. The rods were preheated to 7001F. in an oven. Immediately upon removal from the oven they were sprayed with one of two aqueous salt solutions. These were saturated solutions, prepared as described above in Example I, of potassium carbonate and potassium sulfate. The sprayed Salt Used K.CO3 K.CO3 Temp., OF.
1050 rods were then placed in an oven maintained at the desired temperature. For each temperature, some rods were maintained in the oven for a longer period of time than others. The heat-treated, coated rods then were cooled slowly to room temperature and washed for removal of the salt coating.
The rods coated with potassium carbonate, were washed with the dilute nitric acid. Those coated with potassium sulfate were washed with water. The rods were examined for depth of compressive stress surface layer. All had a compressive stress in that layer. Samples from the rods and having a thickness of about 0.0008 inch were examined for compressive retardation expressed in millimicrons. The data are tabulated below for the different times and temperature treatment.
Time, Minutes 1025 K2CO3 1000 K2C03 975 K2SO 1025 The greater is the depth of the layer of compressive stress, the greater is the resistance to loss of increased strength by abuse.
This depth of layer should be at least 10 microns, as mentioned above, and is preferably at least 15 microns. The greater is the compression retardation, the greater is the increase in strength. However, a bottle having a greater Depth of Stress Layer, t Retardation 210 210 220 compression retardation as a result of the ionexchange treatment than another bottle that has a greater depth of compressive stress layer is not necessarily better from the standpoint 50 of commercial use. Retention of increased strength, during use, is assured by layer depth and not by amount of compression retardation. The data for potassium sulfate treatS 1,181,SS1 0 ment indicate that it is less effective as an ion exchange medium than potassium carbonate. At the temperature used potassium sulfate did not provide a sufficient depth of compressive stress surface layer to afford a retention of satisfactory increase in impact strength during a service-simulated abuse.
However, the data for potassium sulfate treatment indicates that there can be a definite increase in strength of the glass article as a result of this potassium sulfate treatment in accordance with the present invention. This strength increase can be retained by coating the treated and washed bottles with an organic coating to impart lubricity and thus avoid or minimize damage during use, especially in a product filling line.
Although the potassium carbonate treatment at 10501F. for 30 minutes had a compressive stress layer, other experiments indicate that this is about the period of time of treatment beyond which the rate of compressive stress retardation at that temperature begins to decrease. The stress provided by the ion-exchange treatment for a longer period of time eventually provides a surface layer that has a tensile stress. It will be noted that the stress retardation, as compared with the 15-minute treatment, is lower for the 30minute treatment. This indicates that at 1025cF., additional time of the heat treatment will result in loss of compressive stress along with the creation of a tensile stress.
The data show that the desired depth of compressive stress layer is attained faster using temperatures above the strain point and greater depths are also obtained as compared with a temperature slightly below the strain point, namely, 975 F.
EXAMPLE III
One-way beer bottles were made but using a molten glass having the following composition expressed as oxides on a weight percent basis:
SiO2 A1203 MgO Na2O KO 72 3 0.5 This glass has an annealing point of 10280F.
and a strain point of 9800F. This glass is of a type useful in the practice of the invention; the type being an alkali-alkaline earth silicate glass composition comprising on a weight basis about 12% to about 20% NaO, about 5% to about 20% MgO, between 0% and less than 10% ALO, 0% to about 2% Li20, 0% to about 5% K20, 0% to about 5% CaO, and SiO2, said Na2O and MgO constituting at least 23% of the glass, said Na2O, MgO, A120, and SiO2 constituting at least 90% of the glass, the weight ratio of CaO: MgO being a maximum of 1:1, said Li20 content being a maximum of about 1% in the absence of Al2O, and said glass composition having in the absence of A1203 content a mole ratio of MgO: Na2O between 0.4: 1 and 1.25: 1 and in the presence of A120 content a mole ratio of total of MgO and A1,03 to Na.O between 0.5: 1 and 70 1.25: 1. This glass preferably has a maximum weight ratio of CaO: MgO of 0.5:1 and a maximum CaO content of about 2.5%. The total of Na20, MgO, A1203 and SiO2 contents preferably constitutes 95% of the glass. 75 These bottles, directly from the bottle forming station of the IS machine, were sprayed on their outside surface with aqueous potassium carbonate solution, prepared as described in Example I, to provide a substantially uniform 80 coating or layer of potassium carbonate upon evaporation of.the water content. These bottles were also passed through the annealing lehr with the temperature profile indicated in Example I. At the exit end of the lehr the 85 bottles were removed, cooled to room temperature and washed with dilute nitric acid for removal of the salt layer. Again the shoulder impact strengths for bottles were determined using the Preston standard impact go machine, with and without prior abuse using the CLS Abuser. Other bottles of this glass composition, that were made at about the same time but were not sprayed, were passed through the annealing lehr and were also 95 tested for shoulder impact strength. None of these bottles was subjected to abuse. The bottles, that were subjected to the potassium carbonate treatment at the lehr temperature, had compressive stress layers with a depth 100 that varied from one bottle to another between 10 and 14 microns and they had an average compressive retardation value of 170 millimicrons. The average impact strength of the untreated bottles was 4.8 inch-pounds. 105 The potassium carbonate treated bottles, without any abuse, had an average impact strength of 14 inch-pounds. The average impact strength for the bottles treated with potassium carbonate at the elevated temperature and subjected later to abuse was 16 inch-pounds.
These are average values. The fact that the value after abuse is higher than the treated bottles without abuse is not indicative. However, it does indicate that the layer depth 115 is sufficient to provide retention of increase of strength under the longer time of abuse than conventionally used to test bottles.
EXAMPLE IV
Rods of the flint container glass were made 120 as described in Example II and each subjected to a spray of an aqueous salt solution.
1,181,581 1,lgl,581 The rods had been preheated as in Example II. They were either placed in an oven as described for the bottles in Example II or were immediately placed in the annealing lehr described above in connection with Examples I and III for the temperature treatment as described in Example I. The salt content of the aqueous solutions was saturated, and these solutions were prepared as described in Example I. Five types of aqueous solutions were used and three of them contained only one potassium salt, whereas the other two contained a mixture of potassium chloride and potassium carbonate in a molar ratio of Salt in Solution Sprayed Temperature OF.
K2CO,: KCL of 1: 1 and 1:4, respectively.
The salt of the solution, the temperature maintained for the salt layer to react with the glass for ion exchange, the time for this maintenance in the case of the oven treatment and the modulus of rupture that was determined as described above using the TiniusOlsen machine without any abrading treatment are tabulated below. The time for the treatment in the annealing lehr is not shown.
The overall time was about 40 minutes, as described above, but the rods were at the various temperatures during their travel.
Time in Minutes Flexural Strength P.s.i.
K2CO3 1025 30 33,000 K2COa 975 30 36,000 K.CO3 Anneal. lehr 40,000 KC1,, ,, 17,000 K2SO4,) a, 15,400 K2CO3:KCl,, ,, 28,000 K2CO3:4KCl,, ,, 20,000 Rods of this glass, but without the foregoing salt-and-heat treatment and without any abrading, had a flexural strength of 16,000 p.s.i. Other samples of the rods were not sprayed but were passed through the annealing lehr, and they were found to have a flexural strength of 13,500 p.s.i. Thus the foregoing data show the improved strengths obtained by treatment with some of the salts of the present invention applied as a uniform or substantially continuous layer of salt solid at the elevated temperature maintenance that is followed by cooling and coating removal.
It would appear from the first two lines of data that the temperature below the strain point provides a higher strength than the temperature near the annealing point. However, it is believed that 30 minutes at 10250F.
is too long. A few minutes more would result in a lower flexural strength. This is substantiated by the work reported above in Example II.
It is noted that the treatment up to the annealing point using the annealing lehr resulted in the highest strength using potassium carbonate solution as the spray. It must be kept in mind that in the lehr this high temperature is not imparted to the rod for the full travel through the lehr. As a matter of fact, the time is substantially shorter than 30 minutes as described above in Example I.
EXAMPLE V
A glass having the following theoretical composition, on a weight percent basis was prepared in a glass melting tank using as the batch 1,328 lbs. Spruce Pine feldspar, 260 lbs.
raw dolomite limestone, 245 lbs. high-calcite Mississippi limestone, 5 lbs. arsenic trioxide, 4 lbs. niter and 4.8 lbs. sodium antimonate to make a large number of glass containers as described in that application. The analyses of the first three batch ingredients are presented in that application.
SiO2 AlO, CaO Na_.0 K20 MgO AsOs Sb20, Fe!'O3 56.3 15.6 14.8 5.5 3.7 3.5 0.25 0.23 0.08 This glass had an annealing point of 12150F.
and a strain point of 11600F.
A gob was taken from the glass tank and 1,181,581 remelted in a platinum pot to obtain molten glass from which cane was pulled to produce rods, as described earlier. These rods at a temperature of about 700 F. were sprayed with potassium carbonate solution prepared as described in Example I. Then the rods were maintained in an oven at specific temperatures Temperature OF.
Time, Minutes and for specific times, followed by gradual cooling to avoid thermal tempering. The rods were examined for depth in microns of their compressive stress surface layer and for their compression retardation in millimicrons. The data presented are tabulated below.
Depth Compression Retardation 1200 30 33 150 1175 5 10 130 1175 30 27 175 1175 60 33 175 This type of glass has an oxide composition consisting essentially, on a percent by weight basis, of:
Total Na2O and K20 expressed as NaO mole equivalent SiO2 A1203 CaO MgO 43-63 14-25 0O-30 0-20 Total CaO and MgO, expressed as CaO mole equivalent Total alkali metal oxide, expressed as Na2O mole equivalent 10-30 and the log viscosity is at least 4.
EXAMPLE VI
A glass was made in a large continuous furnace or tank lined with a high-alumina refractory (monofrax M) to make a pressed glassware product. The glass had the following analyzed composition on a weight percent basis:
5-15 and the log viscosity of the glass at its liquidus temperature is at least 2.3, and preferably consisting essentially of:
SiO2 A1203 Total CaO and MgO, expressed as CaO mole equivalent Total Na2O and K20 expressed as Na2O mole equivalent 51-63 15-22 10-22 7-14 and the log viscosity is at least 3.4.
The glass that is especially preferred is the composition consisting essentially of:
SiO2 A1203 54-63 17-22 Total CaO and MgO, expressed as CaO mole equivalent 10-12 SiO2 A1203 MgO Li2O3 ZrO2 TiO2 P205 F Na20 As20O 70.4 16.8 4 3.5 1.3 1.8 1.5 0.09 0.5 0.15 The glass was made by melting at a temperature of 2900 F. for about 43 hours a mixture of the following batch materials using a slight excess of air at an oxidizing atmosphere: Petalite (contains 77.7% SiO2, 16.2% A120,, 4.2% Li2O, and minor amounts of other alkali metal oxides and other impurities); flint (99.9+% SiO2); Alcoa A-10 alumina (99.5% AO1203 and minor impurities); periclase (95.3% MgO, 0.5% Fe2O,, 2.8% SiO2, 0.3% A120,, 1.1% CaO); Florida zircon (66% ZrO2, 33.5% SiO2, 0.25% TiO2, 0.1% Fe203O); titanox (substantially pure TiO2); aluminum metaphosphate (substantially pure, except about 1% ignition loss); lithium fluoride (essentially pure LiF); arsenic trioxide, niter and water, 8-13 1,181,581 This glass had an annealing point of about 12200F. Cane was pulled from the glass melt and a number of glass rods about 2 inch in diameter were prepared from the cane. These rods at temperatures between 700 and 9000F.
were sprayed with a saturated sodium chloride aqueous solution to produce a very fine, thin coating of sodium chloride on the rods. The coated rods were heated for one hour at 9000F. The rods were cooled slowly and then washed with water to remove the salt layer.
An ion exchange occurred whereby lithium ions in the glass were replaced by sodium ions. The depth of the compressive stress surface layer was approximately 50 microns.
These rods were found to have an average flexural strength of 45,000 p.s.i. whereas glass rods were not subjected to the sodium chloride treatment at the elevated temperature had an average flexural strength of only 19,800 p.s.i.
None of these rods was subjected to any abrasion treatment prior to testing for strength.
The glass of the composition of this example is thermally crystallizable. This glass has been ion exchanged using a molten bath of sodium nitrate maintained at 7500F. for one-half hour and for 3 hours.
EXAMPLE VII
The batch materials from the corresponding column of Table I (headed Example VII) were melted and refined with mechanical agitation in a platinum crucible under oxidizing conditions in a gas fired furnace at 29000F for about 20 hours. Cane was drawn from the homogeneous molten glass after it had melted and refined. The cane samples, having an average diameter of about 0.2 inches, were cut into about 5 inch lengths to make sample rods. Several of these rods were preheated to 7000F and sprayed with various aqueous solutions of KZHPO4 and heat treated and tested as set forth below in Table III. The spray apparatus utilized was an atomizing handoperated spray gun using compressed air as the propellant. The preheated rods were slowly rotated while the K2HPO4 was sprayed onto the hot rods. As the spray contacted the rods, the water evaporated depositing a uniform coating of the potassium salt on the rods.
The spray application required less than 30 seconds to achieve the desired thickness of salt coating. The rods were then heat treated by placing them in a preheated furnace according to the time schedule set forth below in Table III.
/mu Stress in psi=optical retardation inch, Upon completion of the heat treatment the rods were removed from the furnace, cooled to room temperature, and washed with water to remove the adhering salt layer. The treated rods were clear and bright, and had the appearance of untreated glass.
The sample rods were cut into cross sectional pieces and examined by well known optical techniques to determine the depth, type, and degree of stress throughout the cross section. The stress characteristics were measured as a function of birefringence using a graduated quartz wedge (prism) calibrated in millimicrons. Cross sectional pieces of the exchanged rods were cut from the rods using an ordinary circular diamond saw. Such saws are common in the glass industry. The thickness of the cross sectional piece was measured in the direction parallel to the original axis of the rod sample. The measured cross sectional piece was then mounted in matched index fluid (e.g., oil having the same index of refraction as glass) on the stage of a petrographic microscope so that the polarized light would pass through the measured dimension.
The polarizer was contained in the optic system below the microscopic stage.
The polarized light passing through the measured cross sectional test piece was received in an eye piece equipped with the calibrated quartz prism. The optic lag (retardation), expressed in terms of millimicrons retardation per unit cross sectional thickness of the test specimen, was then obtained by reading the calibrated prism.
The depth of the stress layer was measured to the neutral stress axis with an eye piece calibrated in microns.
The degree of surface stress was approximated based on the assumption that the stressoptical coefficient for the compositions studied was about pounds 2 (mu) (inch) The degree of surface stress reported in Table III is then only an approximation because the stress optical coefficient is known to depend on the glass composition, and there is a compositional gradient established by the ion exchange treatment itself. The calculation, then, expressing the degree of stress can be expressed as follows:
(2 pounds X stress optical coefficient 2 mu inch I f al 1,181,581 The treated sample rods were tested with and without prior abrasion to determine their modulus of rupture.
The abrasion method consists of tumbling several sample rods for 15 minutes in a ball mill in intimate contact with number 30 grit silicon carbide. This is a very severe abrading condition and is more drastic than the abuse encountered in most commercial applications.
The flexural strengths (modulus of rupture) values were determined using a Tinius-Olsen testing machine. This machine applied a measured load through two knife edges spaced - inch apart, to the center of the sample rod supported on two knife edges 4y' inches apart (4-point loading). The load is applied at a constant rate of 24 pounds per minute until failure occurs, with a marker indicating the highest load applied to the point of failure.
A dial micrometer calibrated in inches and equipped with a bar contact instead of a point contact, is used to measure the maximum and minimum diameters at the center of the sample to an accuracy of 0.0005 inch. Since few cane samples are perfectly round, the load is applied normal to the maximum diameter and the standard formula for an elliptical cross section is used in calculating the modulus of rupture as hereinbefore explained.
In this case, the formula is:
MR Load (lbs.) X 8 X span (in.) (D12 XD2) r The modulus of rupture in this formula gives the flexural strength in pounds per square inch of cross sectional area at failure. The results of these tests are reported in Table III. For the purposes of comparison, the five glasses listed in these tables without any ion exchange treatment would have modulus of rupture values of about 10,000-12,000 psi after being subjected to the abrasion method described above.
To further demonstrate the principles of the present invention, the various glass compositions described in the tables under the columns headed Example VIII, Example IX and Example X, were prepared- and treated according to the methods of Example VII.
These examples illustrate the applicability of this invention to a wide range of ion exchangeable glass compositions. Example VII and IX are soda aluminosilicate compositions, Example VIII is -a soda magnesia silicate composition while Example X is a soda-lime silica composition.
The properties, compositions and results are listed in Tables I, II and III, the performance of an ordinary soda-lime-silica glass is included in Tables II and III because of its commercial importance.
TABLE I
Batch Materials for Exemplary Glasses Material Ex. VII Ex. VIII Ex. IX Nepheline syenite 3826.5 635.6 Periclase 104.3 511.2 Sand (99.5k SiO2) 667.3 3213.2 0290.0 Sodium borate (anhydrous) 256.2 - Soda Ash 318.6 1144.7 910.5 Alumina (99.5+ A120 ) 0.5 - 887.3 Sodium Antimonate - - 47.4 _12 1,181,581 TABLE II
Theoretical Composition and Properties of Exemplary Glasses Constituent Ex. VII Ex. VIII Ex.- IX Ex. X SiO2 59.4%/ 71.8% 59.5% 72.0% A1203 18.0 3.0 25.2 2.0 MgO 2.0 10.0 - 3.5 CaO Na20 K20 Sb2O3 B203 Liquidus ( F.) Log = 4 ( F) Annealing Point F.
10.0 13.2 3.7 14.6 0.6 15.3 12.0 0.5 1.0 3.5 1725 2150 1765 1950 1985 2650 1370 1940 1800 1025 TABLE III
Summary of Exemplary Results
Concentration K2HPO4 in water (gm/cc H20) Heat Treatment temperature ( F.) Heat treatment time (minutes) Maximum optical retardation in mu/0.01 inch Approximate surface compressive in PSI Depth of compressive stress in microns Flexural strength (unbraided) Flexural strength (abraded) 0.1 1025 Example VII
Ion Exchange Treatment 0.2 0.4 0.7 1025 1025 1025 5 5 Optical Properties of Exchanged Glasses 1.0 1025 1.3 1025 0 85 260 265 315 265 0 17,000 52,000 53,000 63,000 53,000 - 19 39 44 42 39 Flexural Strength Modulus of Rupture of Exchanged Glasses in PSI (Average of 5 samples - 1f1 A sNn 9) Ana Inn AG rAAf A R7ANN 13,400 18,900 cc }..
] -'j00 16,300 9,2 --1. 13,900 Concentration K2HPO4 in water (gm/cc H20) Heat treatment temperature ( F).
Heat treatment time (minutes) Maximum optical retardation in mu/0.01 inch Approximate surface compressive stress in PSI Depth of compressive stress in microns Flexural strength (unabraded) Flexural strength (abraded) TABLE III (Continued) Summary of Exemplary Results
Example VIII
Ion Exchange Treatment 0.1 0.2 0.4 0.7 1.0 1.3 980 980 980 980 980 980 5 5 5 5 5 Optical Properties of Exchanged Glasses 0 30 90 90 90 110 0 6,000 18,000 18,000 18,000 22,000 - 12 20 10 15 21 Flexural Strength Modulus of Rupture of Exchanged Glasses in PSI (average of 5 samples) 14,900 15,400 31,500 56,500 40,900 31,200 6,900 8,100 11,200 13,600 10,400 15,500 th 00 TABLE III (Continued) Concentration K2HPO4 in water (gm/cc H20) Heat treatment temperature ( F) Heat treatment time minutes)) Maximum optical retardation in mu/0.01 inch Approximate surface compressive stress in PSI Depth of compressive stress in microns Flexural strength (unabraded) Flexural strength (abraded) Summary of Exemplary Results
Example IX
Ion Exchange Treatment 0.1 0.2 0.4 0.7 1.0 1.3 1030 1030 1030 1030 1030 1030 30 30 30 30 30 Optical Properties of Exchanged Glasses 0 0 80 70 110 160 -]-- 16,000 14,000 22,000 32,000 0 0 78 62 16 64 Flexural Strength Modulus of Rupture of Exchanged Glasses in PSI (average of 5 samples) 16,400 11,100 19,800 40,000 35,900 52,91 10,700 31,300 27,200 37,31 44,300 37,400 It, Oo 1,181,581 EXAMPLE X
To further demonstrate the principles of the present invention, the ion exchange technique was practiced in conjunction with ordinary soda-lime glass. It should be noted that sodalime glass of itself has very poor ion exchange properties, but nevertheless, it was possible to achieve ion exchange by the present spray technique that is comparable to the ion exchange achieved by the molten bath technique.
Small soda-lime glass jars of conventional design, similar to those used in packaging baby food, were formed in conventional glass forming equipment and sprayed with an aqueous solution of K2HPO4 (1.3 gm K2HPO4/cc of H20) immediately after their formation, and prior to their annealing. The temperature of the jars at the point of application was 700-9000F. The hot, freshly formed coated jars were then passed through the annealing lehr described in Example I to receive the normal annealing temperature treatment that untreated bottles receive in their manufacture.
The soda-lime glass used was the same as that used and described in Example I.
The jars after exiting from the annealing lehr and cooling to room temperature were washed with tap water. The treated jars were clear and bright, and generally had the appearance of untreated jars.
The surface compressive layer of the treated jars was evaluated by the methods described above. A compressive stress layer of about 10 microns in depth with a maximum optical retardation of about 100 mu/0.01 inch was observed. This is equivalent to a compressive stress of approximately 20,000 psi. A corresponding increase in strength was realised in the treated glass jar.
The process of the present invention is not limited to the specific glass compositions that were used for the foregoing examples. The process is applicable to many other types of glasses that have been ion exchanged using alkali metal salts of inorganic acids in molten form and to other types of glass, especially silicate glass containing alkali metal ions capable of ion exchanging.
It is seen from the foregoing that there are many types of silicate glasses that contain silica and alkali metal oxide. Some contain one or more other oxides that are real or probable glass formers and some contain other oxides as glass modifiers. These silicate glasses containing alkali metal oxide have compositions that contain the following components in the indicated percent ranges:
Sio2 M20 Al2sO CaO MgO BaO SrO B203 ZrO, TiO2 SnO2 P20Q As2Q, Sb201 Percent by Weight 35-88 1-48 0-40 0-15 0-28 0-15 0-15 0-15 0-25 0-12 0-2 0-10 0-3 0-3 wherein M20 refers to the total of alkali metal oxide and, when the alkali metal oxide is 75lithium oxide, potassium oxide, rubidium oxide or cesium oxide, it normally constitutes a maximum of 25% by weight of the glass composition. The content of alkali metal oxide to be at least partially replaced in a surface layer 80 by another alkali metal oxide preferably constitutes at least 2%/, and for glasses, other than glass-ceramics, it is especially preferred that it constitutes at least 5%.
For those glass compositions that are thermally crystallizable to form glass-ceramics, antimony oxide or arsenic oxide is normally part of the batch material to form the glass.
Up to about 1% by weight of either or both is used. They are used as fining agents or 90 oxidizing agents. Most of these oxides are lost by vaporization in the glass-making furnace so that the final glass composition will actually contain at most only a few tenths of one percent. When arsenic oxide is used as 95 fining agent there is commonly used also in the batch, a small amount of sodium nitrate, but it is not normally shown in the theoretical composition.
Fluorine as a salt is commonly used in 100 batch material as an additive in an amount usually not exceeding 0.3% by weight in the final composition. Fluorine is believed to aid crystallization; but its content of the composition is limited to a low value, because it 105 accelerates the crystallization, sometimes with an undesirable exothermic effect.
Within this glass composition, it will be apparent to one skilled in the art that there are narrower limits to the ranges of the individual oxides depending upon which ones are present to form a compatible mixture as a melt that when cooled will be a glass. These glasses are per se no part of the present invention. Instead, they are the materials that 115 are treated by the process of this invention to 1,181,581 form the improved glass articles. However various classes of glasses within this broac type are presented below for the purpose o:
illustrating the cited variation of glasses usefu in the present invention.
The following presents various examples o0 multicomponent glass systems.
One example is the class of glasses composed of silica, one or more alkali meta.
oxides, and one or more alkaline earth metal oxides. A common glass representative of this class is the alkali-lime-silica glass, such as used for window sheet glass, plate glass and container glass. In these commercial glasses the alkaline earth metal oxide content is usually lime or a mixture of calcia and magnesia such as is present in a dolomitic lime.
The approximate composition of such commercial glasses on a weight basis is as follows:
70-74% silica, 12-16% soda, either 1013%/o calcia and magnesia total or 8-12%/, calcia and 1-4% magnesia. Alumina is present in 0.5-1.5% by weight for sheet and plate glass while for container glass it is usually 1.5-2.5%, but in some cases exceeds 5%.
This glass with the low alumina content can be ion exchanged to improve its strength as evidenced by some of the foregoing examples.
Another class of glasses within the broad type of alkali metal silicate glasses is the leadalkali metal silicate glass, in which the alkali metal oxide is potassium oxide alone or with soda, i.e., sodium oxide.
Another class of glasses useful in the present invention is the alkali aluminosilicate glass compositions: 50-75% silica; at least 5% and preferably from 10-25% alumina; and at least 5%, preferably 10-25%, NaO, with the alumina and Na2O content preferably constituting at least 15% of the glass composition and with these two plus the silica constituting at least 85% of the glass composition. Divalent metal oxides, potassium oxide, boron oxide, titania, phosphorus pentoxide and fluorine may be present up to a maximum individiual content of 10% and collectively up to a maximum of 15%. Lithium oxide may be present but should not exceed 1%. Because some of these limitations are based upon the attaining of the high strength even after abrasion, such limitation, although preferred, is not a limitation on the present invention.
Another class of glasses of the broad alkali metal oxide-silica type is the lithium silicate glass (on a weight basis) 46-88% silica and 4-29% lithia. This glass may contain alumina to constitute the remainder, if any, but the ratio of silica to alumina should be at least 2: 1. Thus it is seen that this class of glasses can be the binary type, but when alumina is present it is the alkali metal aluminosilicate. Instead of alumina, or for part of it, there may be present one or more of the following constituents: zirconia; titania; and boron oxide. In addition other alkali metal oxides, namely, sodium oxide and potassium oxide, may be present along with lead oxide (PbO) and fluorine up to a total of 15 mole percent. Of course, some of these limitations relates to the compositions which provide the maximum mechanical strength after abrasion, but such is not a limitation for the present invention in its broadest sense.
A further class of glasses that contains ion exchangeable alkali metal ions is the glass composition constituting at least 10%, preferably at least 20%, by weight of sodium oxide, at least 5% by weight of zirconia and the balance silica, except for lithia (lithium oxide), if present, which normally should not exceed 1% by weight and except for optional compatible ingredients including divalent metal oxides, potassium oxide, boron oxide, phosphorus pentoxide, titania and fluorine which individually may be present in an amount up to 10% by weight and collectively may be present in an amount up to 15% by weight. In the ternary glass system the composition can be, e.g., 60 to 75% by weight of silica, 5 to 20% by weight of zirconia and 20% by weight of sodium oxide.
Alkali-alkaline earth metal silicate glasses, which may contain alumina, boron oxide and various compatible inorganic oxides, can be ion exchanged using alkali metal salts. These glasses contain by weight in excess of 40%, e.g., 65-75% silica, 0-15% boron oxide, 0-35% alumina, 0-25% calcium oxide, magnesia, strontia, barium oxide, lead oxide and/or zinc oxide and combinations thereof, 0-10% titania, 0-10% potassium oxide and 2-20% sodium oxide and/or lithium oxide.
Typical glass compositions are described and these are ion exchanged for strengthening of the glass.
Examples of thermally crystallizable silicate glass compositions are as follows:
SiO2 Al20, Li2O Na20 K20 CaO MgO TiO0 ZrO2 Percent by Weight 56.1-73.1 12.1-15.3 3.0- 5.2 0 - 1.7 o - 0.2 0 -11.1 0 - 8.8 4.5-13.8 0 - 3.9 In some of these compositions fluorine is present as a fining agent. These compositions after controlled thermal crystallization are glass-ceramics and some of these, provided there is suitable heat treatment, are capable of ion exchanging lithium in the glass-ceramic with an alkali metal in an inorganic salt bath and thus also capable of being ion exchanged by the method of the present invention.
Another example is as follows:
I - 418 is 1,181,581 Sio2 Al2,O, LiO20 TiO2 SiO2 and TiO,2 Percent by Weight 55-75 12-36 2-15 3-7 58-82 etc. Metal oxides as colorants may be present in an amount of 0.005-2%o by weight. To provide lower expansion characteristics to the glass-ceramic that can be formed from the glass composition, the components are as follows:
with the recited ingredients constituting at least 95% of the composition and the weight ratio of Li0: Al,203 being between 0.1: 1 and 0.6: 1.
Another class of thermally crystallizable glass composition that can be ion exchanged in the glass form and by proper heat treatment can be exchanged as a glass-ceramic is as follows:
SiO2 A120, LiO ZrO, Percent by Weight 48-73 14-35 4-10 2-6 and wherein the sum of recited ingredients, other than zirconia, is greater than 85% of the composition.
Another thermally crystallizable glass composition having the following composition:
Sio2 Al,20, LiO20 TiO2 ZrO.
Percent by Weight 48-73 14-25 4-10 0-1.8 2-6 wherein the total of the recited ingredients, other than titania and zirconia, constitutes at least 85% of the glass. Many of the specific compositions that are disclosed contain 3% by weight of BO,.
Another class of glass compositions as thermally crystallizable glass and as glassceramics is illustrated by Example VI, supra.
The composition on a percent by weight basis consists essentially of:
SiO2 AlOa LiO20 MgO ZrO2 TiO= SnO2 P205 BaO ZnO 66-73 15-19 2.5-4 3-7.7 1-1.7 1-< 1.9 0-1.7 0-3 0-5 0-3 where the total weight percent of ZrO2, TiO,, SnO2 and P,20, is at least 2.8, and the total weight percent Li20 and MgO is 6.3 to 10.5.
The amount of nucleating agent, such as titania and zirconia, depends upon the particular composition and the particular nucleating agent or combination of nucleating agents, SiO, Percent A120, 56-68 LiO 18-27 CaO 3.4-4.
ZnO 0-3 B20sO, 0-2 TiO2 0-4 ZrO2 0-6 MgO 0-3 NaO20 0-3 P20O 0-1 (SiO, & Al,20,) 0-3 (SiO,, A2I,O,, B,20s & at least P20,) (CaO, MgO, ZnO & 86-91 Na,20) (SiO2, Al,203, P,20., & 2.5-6 Li,20) TiO, & ZrO2 no more 2-6 by Weight than 93 where the ratio of (CaO, MgO, ZnO, Na20 & B20) to Li20 is less than 2.4 and the ratio of SiO2 to Al20, is no more than 3.3 and preferably no more than 3.8.
When the ion-exchange process of the present invention is carried out by spraying to form the layer while the glass is at or near its annealing point and then maintaining it at that temperature, some etching of glass occurs if the salt is potassium carbonate and the glass is conventional flint glass as used in Examples I, II and IV. This can be avoided or minimized in a number of ways. One way is to use a mixture of alkali metal carbonate and another salt of the alkali metal, e.g., the chloride. A further technique is to spray the glass while it is at a substantially lower temperature and then raise its temperature for the ion exchange. Still another way is to use a lower temperature for the ion-exchange treatment. Also the appearance of the etched surface, if undesirable for certain products can be improved by an overcoat of organic material with substantially matching refractive index. Furthermore, for certain products, etching in predetermined areas by this process is possible and will give increased strength in those areas and/or enhancing ornamental effects.
The process of the present invention can provide in addition to the improved strength a coloration to the glass. This coloration is accomplished by incorporating, in the layer of material consisting essentially of salt or salts of alkali metal, as defined above, a salt of a colored heavy metal ion, such as cobalt, that will provide a diffusion of the colored 1,181,581 metal in the glass during the elevated temperature treatment for the ion exchange, preferably with the alkali metal salt (carbonate) and at the temperature for ion exchange with etching, mentioned above. Other heavy metals are iron and nickel. Their heavy salts, that are used, do not melt at the elevated temperature used for the ion exchange. Thus their colored salts are, for example, chlorides, carbonates and sulfates of the heavy metals.
Such salts constitute less than 10(%, and preferably less than 5%., by weight of the alkali metal salt content of the solid layer formed on the glass in the first step of the ionexchange process.
Dilute nitric acid has been mentioned in various examples as being used for removal of alkali metal carbonate, specifically potassium carbonate, from the glass surface after the ion exchange. It was 3N nitric acid. Other concentrations of aqueous nitric acid have been used. As a matter of fact, potassium carbonate, for example, can be removed merely by washing with water but such removal takes longer and uses a larger volume of wash liquid. Of course, therefore, the concentration of nitric acid can be varied widely to provide easy removal of salt layer from the surface of the glass.