KR102239613B1 - Bulk annealing of glass sheets - Google Patents

Abstract

Surface modification layer 30, which may be provided on sheet 20, carrier 10, or both to control both room temperature van der Waals (and/or hydrogen) bonding and hot covalent bonding between the thin sheet and the carrier. ) And associated heat treatment. Room temperature bonding is controlled to be sufficient to hold the thin sheet and carrier together, for example during vacuum processing, wet processing and/or ultrasonic cleaning processing. And at the same time, the hot covalent bonding is controlled to prevent permanent bonding between the thin sheet and the carrier during hot processing, as well as to maintain sufficient bonding to prevent delamination during hot processing.

Description

Bulk annealing of glass sheets {BULK ANNEALING OF GLASS SHEETS}

This application claims priority to U.S. application serial number US14/047251, filed October 7, 2013 under 35 USC §120, and 35 U.S.C. Priority is claimed to U.S. Provisional Application Serial No. 61/791,418, filed March 15, 2013 under § 119, the contents of which are trusted and incorporated herein by reference in their entirety.

Field of invention

The present invention relates to articles and methods for processing flexible sheets on a carrier, and more particularly to articles and methods for processing flexible glass sheets on a glass carrier.

Flexible substrates offer the prospect of cheaper devices using roll-to-roll processing, and the possibility of manufacturing thinner, lighter, more flexible, and more durable displays. However, the technologies, equipment and methods required for roll-to-roll processing of high-quality displays have still not been fully developed. Since panel manufacturers have already invested heavily in toolsets for processing large sheets of glass, it is thinner, lighter, and more flexible to laminate flexible substrates to carriers and fabricate display devices by sheet-to-sheet processing It provides a more short-term solution for developing valuable propositions of sexuality display. The display was demonstrated on a polymer sheet, for example polyethylene naphthalate (PEN), where the device fabrication was sheet-to-sheet with PEN laminated to a glass carrier. The upper temperature limit of PEN limits the equipment quality and process that can be used. In addition, the high permeability of the polymer substrate leads to environmental degradation of OLED devices that require an almost hermetic package. Thin film encapsulation offers the potential to overcome this limitation, but it has not yet been demonstrated to provide acceptable yields at large doses.

In a similar manner, a display device can be manufactured using a glass carrier laminated to one or more thin glass substrates. It is expected that the low permeability and improved temperature and chemical resistance of thin glass will enable higher performance longer life flexible displays.

However, thermal, vacuum, solvent and acidic, and ultrasonic, flat panel display (FPD) processes require a strong bond to the thin glass bonded to the carrier. FPD processes are typically vacuum deposition (chemical vapor deposition (CVD) deposition of sputtered metals, transparent conductive oxides and oxide semiconductors, amorphous silicon, silicon nitride, and silicon dioxide, and dry etching of metals and insulators), thermal processes (~300-400 ℃ CVD deposition, 600 ℃ or less p-Si crystallization, 350-450 ℃ oxide semiconductor annealing, 650 ℃ dopant annealing, and ~200-350 ℃ contact annealing), acid etching (metal etch, oxide semiconductor etch), solvent exposure ( Stripping photoresist, deposition of polymer encapsulation), and ultrasonic exposure (in solvent stripping and aqueous cleaning of the photoresist, typically in an alkaline solution).

Adhesive wafer bonding has been widely used in micromechanical systems (MEMS) and semiconductor processing for later stages where the process is less harsh. Commercial adhesives by Brewer Science and Henkel are typically 5-200 microns thick, thick polymeric adhesive layers. The large thickness of these layers causes a large amount of volatiles, trapped solvents, and the possibility of adsorbed paper contaminating the FPD process. These substances are thermally decomposed and outgassing above -250°C. The material can also cause contamination in the downstream stages by acting as a sink for gases, solvents and acids, which can outgas in subsequent processes.

U.S. Provisional Application Serial No. 61/596,727 (hereinafter US'727) filed on February 8, 2012 under the title Processing Flexible Glass with a Carrier, was first described by Van der Waals forces. A thin sheet, for example a flexible glass sheet, is bonded to a carrier, followed by a device (e.g., an electronic or display device, a component of an electronic or display device, an organic light emitting device (OLED) material, a photovoltaic (PV) structure or (A thin film transistor) after processing the thin sheet/carrier to form a thin sheet/carrier, while retaining the ability to remove portions of the thin sheet, discloses a concept that includes increasing the bond strength in a specific area. At least a portion of the thin glass is attached to the carrier to prevent the device process fluid from entering between the thin sheet and the carrier, reducing the likelihood of contamination of the downstream process, i.e. the sealing portion bonded between the thin sheet and the carrier is hermetic, some desirable. In embodiments the sealing covers the exterior of the article to prevent penetration of liquid or gas into or out of any area of the sealed article.

US '727 continues that in low temperature polysilicon (LTPS) (low temperature compared to solid phase crystallization processes, which can be about 750° C. or less) temperature, vacuum, and wet etch environments close to 600° C. or higher can be used in the device manufacturing process. Start. These conditions limit the materials that can be used and place high demands on the carrier/thin sheet. Thus, it is possible to process thin glass, i.e. glass having a thickness of ≤ 0.3 mm thick, without contamination or loss of the bonding strength between the thin glass and the carrier at high process temperatures, and the thin glass is easily from the carrier at the end of the process. A carrier approach that utilizes the manufacturer's existing capital infrastructure, which is seamlessly de-coupled, is desirable.

One commercial advantage to the approach disclosed in US '727 is that, as mentioned in US '727, manufacturers can use thin glass sheets for, for example, PV, OLED, LCD and patterned thin film transistor (TFT) electronics. It will be possible to use their existing capital infrastructure in process equipment while taking advantage. Additionally, the approach is for cleaning and surface preparation of thin glass sheets and carriers to facilitate bonding; To strengthen the bond between the carrier and the thin sheet in the bonding area; To maintain the releasability of the thin sheet from the carrier in the non-bonding (or reduced/low-strength bonding) area; And for the cutting of thin sheets to facilitate extraction from the carrier, and the like.

In the glass-to-glass bonding process, the glass surface is cleaned to remove all metal, organic and particulate residues, leaving mostly silanol-terminated surfaces. The glass surface is first brought into intimate contact so that van der Waals and/or hydrogen-bonding forces pull the glass surface together. Using heat and optionally pressure, surface silanol groups are condensed to form strong covalent Si-O-Si bonds across the interface, permanently fusing the glass pieces. Metallic, organic and particulate residues will prevent bonding by masking the surface, preventing the tight contact required for bonding. Since the number of bonds per unit area will be determined by the probability that two silanol species on opposite surfaces will react and condense with water, a high silanol surface concentration is also required to form strong bonds. Zhuravlel reported that the average number of hydroxyls per ^{nm 2} for well hydrated silica was 4.6 to 4.9. (Zhuravlel, LT, The Surface Chemistry of Amorphous Silika, Zhuravlev Model, Colloids and Surfaces A: Physiochemical Engineering Aspects 173 (2000) 1-38). In US '727, non-bonding regions are formed within the bonded periphery, and the main way described for forming such non-bonding regions is to increase the surface roughness. An average surface roughness of greater than 2 nm Ra can prevent glass-to-glass bond formation during elevated temperatures of the bonding process. U.S. Provisional Application Serial No. 61/736,880, filed December 13, 2012 under the title Facilitated Processing for Controlling Bonding Between Sheet and Carrier by the same inventor. In the following US '880), the controlled bonding region is formed by controlling the van der Waals and/or hydrogen bonding between the carrier and the thin glass sheet, but the covalent bonding region is also still used. Thus, articles and methods for processing thin sheets in US '727 and US '880 using a carrier can withstand the harsh environment of the FPD process, but are undesirable for some applications, such as covalent bonding (e.g., Si-O -Si) bonded area, for example, in the bonding area ^{bonded with an adhesive force of ~1000-2000 mJ/m 2} , which is approximately the breaking strength of the glass, the carrier can be reused by strong covalent bonding between the thin glass and the glass carrier. There will be no. Using frying or peeling, it is not possible to separate the covalently bonded portions of the thin glass from the carrier, and thus the entire thin sheet cannot be removed from the carrier. Instead, the non-bonding area with the device above it is lined and excised, leaving only the bonded periphery of the thin glass sheet on the carrier.

In view of the above, it is possible to withstand the harshness of FPD processing, including high-temperature processing (without outgassing which may be incompatible with the semiconductor or display manufacturing process to be used), but the use of thin sheets to allow reuse of carriers for processing other thin sheets. There is a need for a thin sheet-carrier article that allows the entire area to be removed from the carrier (all at once or in portions). This specification describes a way to control the adhesion between the carrier and the thin sheet to create a temporary bond that is strong enough to withstand FPD processing (including LTPS processing), but weak enough to allow the debonding of the sheet from the carrier even after a high temperature process. . Such controlled bonding can be used to create articles with re-usable carriers, or alternatively articles with patterned regions of controlled bonding and covalent bonding between the carrier and the sheet. More specifically, the present disclosure relates to a surface modification layer (various materials) that may be provided on the thin sheet, carrier or both to control room temperature van der Waals and/or both hydrogen bonding and hot covalent bonding between the thin sheet and the carrier. And associated surface heat treatment). Even more specifically, room temperature bonding can be controlled to be sufficient to hold the thin sheet and carrier together during vacuum processing, wet processing and/or ultrasonic cleaning processing. And at the same time, the hot covalent bonding can be controlled to prevent permanent bonding between the thin sheet and the carrier during hot processing, as well as to maintain sufficient bonding to prevent delamination during hot processing. In an alternative embodiment, a covalent bonding region to provide further processing options using a surface modification layer, e.g., maintaining the hermeticity between the carrier and the sheet even after dicing the article into smaller pieces for further device processing. Together, it is possible to create a variety of controlled bonding regions, where the carrier and sheet remain sufficiently bonded during various processes including vacuum processing, wet processing and/or ultrasonic cleaning processing. Additionally, some surface modification layers provide control of the bonding between the carrier and the sheet at the same time while reducing outgassing during harsh conditions in an FPD (e.g. LTPS) processing environment, including, for example, high temperature and/or vacuum processing. do.

Additional features and advantages will be described in the following detailed description, and will be apparent to those skilled in the art from the detailed description in part, or will be recognized by practicing various aspects as illustrated in the described description and the accompanying drawings. It is to be understood that both the above general description and the following detailed description are merely illustrative of various aspects and are intended to provide an overview or framework for understanding the nature and features of the invention as claimed.

The accompanying drawings are included to provide a further understanding of the principles of the invention, and are incorporated into and constitute a part of this specification. The drawings illustrate one or more embodiment(s) and together with the description serve to illustrate the principles and operation of the invention by way of example. It is to be understood that the various features disclosed in the specification and drawings may be used in any and all combinations. As a non-limiting example, various features may be combined with each other as described as aspects later in the specification.

1 is a schematic side view of an article having a carrier bonded to a thin sheet with a surface modification layer therebetween.
2 is an exploded view and a partial perspective view of the article of FIG. 1;
3 is a graph of surface hydroxyl concentration for silica as a function of temperature.
4 is a graph of the surface energy of an SC1-cleaned sheet of glass as a function of annealing temperature.
5 is a graph of the surface energy of a thin fluoropolymer film deposited on a sheet of glass as a function of the percentage of one of the constituent materials from which the film is made.
6 is a schematic top view of a thin sheet bonded to a carrier by a bonding area.
7 is a schematic side view of a stack of glass sheets.
8 is an exploded view of one embodiment of the stack of FIG. 7.
9 is a schematic diagram of a test setup.
10 is a collection of graphs of surface energy versus time (of different portions of the test setup of FIG. 9) for various materials under different conditions.
11 is a graph of change in% bubble area versus temperature for various materials.
12 is another graph of change in% bubble area versus temperature for various materials.

In the following detailed description, for purposes of explanation and not limitation, exemplary embodiments disclosing specific details are set forth to provide a thorough understanding of the various principles of the present invention. However, it will be apparent to those skilled in the art having the benefit of the present disclosure that the present invention may be practiced in other embodiments departing from the specific details disclosed herein. Further, descriptions of well-known devices, methods, and materials may be omitted so as not to obscure the description of the various principles of the present invention. Finally, where applicable, the same reference numbers refer to the same elements.

Ranges may be expressed herein as “about” one particular value and/or to “about” another particular value. Where this range is expressed, another embodiment includes one particular value and/or to another particular value. Similarly, when values are expressed as approximations using the prefix “about”, it will be understood that certain values form another embodiment. It will be further understood that the endpoints of each range are significant both in relation to the other endpoints and independently of the other endpoints.

Orientation terms as used herein-for example up, down, right, left, front, back, top, bottom-are made only in connection with the drawing as shown and are not intended to indicate an absolute orientation.

As used herein, the singular form includes the plural referent unless the context clearly dictates otherwise. Thus, for example, a reference to “element” includes aspects having two or more such elements unless the context clearly dictates otherwise.

US 61/596,727 and US61/736,880 filed on February 8, 2012 under the title Processing Flexible Glass with a Carrier (facilitated to control the bonding between the sheet and the carrier). In both processes (filed December 13, 2012 under the title Facilitated Processing for Controlling Bonding Between Sheet and Carrier), at least a portion of the thin glass sheet remains "non-bonded" and is processed onto the thin glass sheet A solution is provided that allows the processing of thin glass sheets on the carrier, allowing the device to be removed from the carrier. However, the periphery of the thin glass is permanently (or covalently or hermetically) bonded to the carrier glass through the formation of covalent Si-O-Si bonds. Since the thin glass cannot be removed without damaging the thin glass and the carrier in the permanently bonded area, the covalently bonded periphery prevents reuse of the carrier.

In order to maintain advantageous surface shape characteristics, the carrier is typically a display grade glass substrate. Thus, in some situations, simply discarding the carrier after one use is wasteful and expensive. Therefore, in order to reduce the cost of manufacturing the display, it is desirable to be able to process more than one thin sheet substrate by reusing the carrier. The present disclosure is that thin sheets are processed at high temperature-where the high temperature processing is processing at a temperature of ≥ 400° C., which can vary greatly depending on the type of device being manufactured, for example amorphous silicon or amorphous indium gallium zinc oxide (IGZO) backplane. The harshness of the FPD processing line, including a temperature of about 450°C or less as in the case of processing, about 500-550°C or less as in the case of crystalline IGZO processing, or about 600-650°C or less as typical in the LTPS process. Allows to be processed through the environment, but also allows the thin sheet to be easily removed from the carrier without damage to the thin sheet or carrier (e.g., one of the carrier and the thin sheet breaks or splits into two or more pieces). So that the carrier can be reused.

1 and 2, the glass article 2 has a thickness 8, a carrier 10 having a thickness 18, a thin sheet 20 having a thickness 28 (i.e. 10-50 micrometer, 50-100 micrometer, 100-150 micrometer, 150-300 micrometer, 300, 250, 200, 190, 180, 170, 160, 150, 140, 130, 120, 110, 100, A sheet having a thickness of ≤ 300 micrometers, including, but not limited to, a thickness of 90, 80, 70, 60, 50, 40, 30, 20, or 10 micrometers), and a surface having a thickness 38 And a modified layer 30. The glass article 2 is a thicker sheet (i.e., approximately ≥ .4 mm, for example .4 mm, .5 mm, .6 mm, .7 mm, although the thin sheet 20 itself is ≤ 300 micrometers) , .8 mm, .9 mm, or 1.0 mm). That is, the thickness 8, which is the sum of the thicknesses 18, 28, and 38, is the thicker sheet that one piece of equipment was designed to process-for example, equipment designed to place electronic device components on a substrate sheet. It is designed to be equal to the thickness of For example, if the processing equipment is designed for a 700 micrometer sheet, and the thin sheet has a thickness 28 of 300 micrometers, then the thickness 18 is, assuming that the thickness 38 is negligible. , 400 micrometers will be chosen. That is, the surface modification layer 30 is not drawn to scale; Instead, it is heavily exaggerated for illustration purposes only. In addition, the surface modification layer is shown in perspective. In practice, the surface modification layer will be evenly disposed over the bonding surface 14 provided it provides a reusable carrier. Typically, the thickness 38 will be on the order of nanometers, such as 0.1 to 2.0, or 10 nm or less, and may be 100 nm or less in some examples. The thickness 38 can be measured with an ellipsometer. In addition, the presence of the surface modification layer can be detected by surface chemical analysis, for example ToF Sims mass spectrometry. Thus, the contribution of the thickness 38 to the article thickness 8 is negligible and to determine the thickness 18 of the carrier 10 suitable for processing a given thin sheet 20 having a thickness 28. Can be neglected in the calculation for. However, to the extent that the surface modification layer 30 has any significant thickness 38, such that the thickness 18 of the carrier 10 is a given thickness 28 of the thin sheet 20, and the processing equipment is devised. It can be considered in determining for a given thickness that has been made.

The carrier 10 has a first surface 12, a mating surface 14, a perimeter 16, and a thickness 18. Additionally, the carrier 10 can have any suitable material, including, for example, glass. The carrier need not be glass, but can instead be ceramic, glass-ceramic, or metal (since the surface energy and/or bonding can be controlled in a manner similar to that described below with respect to the glass carrier). When made of glass, the carrier 10 may have any suitable composition, including alumino-silicate, boro-silicate, alumino-boro-silicate, soda-lime-silicate, and contains alkali or It may be alkali-free. The thickness 18 may be about 0.2 to 3 mm or more, such as 0.2, 0.3, 0.4, 0.5, 0.6, 0.65, 0.7, 1.0, 2.0, or 3 mm or more, and as mentioned above, the thickness 28 If, and thickness 38 are not negligible it will depend on such. Further, the carrier 10 may be made of a single layer, or a plurality of layers (including a plurality of thin sheets) bonded together, as shown. In addition, the carrier may have a Gen 1 size or more, for example Gen 2, Gen 3, Gen 4, Gen 5, Gen 8 or more (e.g., sheet size 100 mm x 100 mm to 3 meters x 3 meters or more). have.

The thin sheet 20 has a first surface 22, a bonding surface 24, a perimeter 26, and a thickness 28. The peripheries 16 and 26 may have any suitable shape, may be the same as each other, or may be different from each other. In addition, the thin sheet 20 may have any suitable material including, for example, glass, ceramic, or glass-ceramic. When made of glass, the thin sheet 20 can have any suitable composition, including alumino-silicate, boro-silicate, alumino-boro-silicate, soda-lime-silicate, and alkali depending on its end use. It may be containing or alkali free. The coefficient of thermal expansion of the thin sheet could be matched relatively close to the coefficient of thermal expansion of the carrier to prevent warping of the article during processing at elevated temperature. The thickness 28 of the thin sheet 20, as mentioned above, is less than or equal to 300 micrometers. In addition, thin sheets may have a Gen 1 size or greater, e.g. Gen 2, Gen 3, Gen 4, Gen 5, Gen 8 or greater (e.g., sheet size 100 mm x 100 mm to 3 meters x 3 meters or greater). I can.

The article 2 will not only need to have the correct thickness processed in existing equipment, but it will also need to be able to withstand the harsh environment in which the processing is performed. For example, flat panel display (FPD) processing can include wet ultrasonic, vacuum, and high temperature (eg,> 400° C.) processing. In some processes, as mentioned above, the temperature can be ≥ 500°C, or ≥ 600°C, and up to 650°C.

In order to withstand the harsh environment in which the article 2 will be processed, such as during FPD manufacturing, for example, the bonding surface 14 has a sufficient strength to prevent the thin sheet 20 from separating from the carrier 10. ) Must be combined. And the strength should be maintained throughout processing so that the thin sheet 20 does not separate from the carrier 10 during processing. In addition, in order to allow the thin sheet 20 to be removed from the carrier 10 (so that the carrier 10 can be reused), the bonding surface 14 may be, for example, by means of an initially designed bonding force and/or It should not be too tightly bonded to the bonding surface 24 by the bonding force caused by the modification of the initially designed bonding force, such as may occur when the article is subjected to processing at high temperatures, for example> 400°C. The surface modification layer 30 can be used to control the strength of the bond between the bonding surface 14 and the bonding surface 24 to achieve both of these purposes. The controlled bonding force is achieved by controlling the contribution of van der Waals (and/or hydrogen bonding) and covalent attraction energy to the total bonding energy controlled by adjusting the polar and non-polar surface energy components of the thin sheet 20 and carrier 10. Is achieved. The controlled bonding is to withstand FPD processing (including wet, ultrasonic, vacuum, and thermal processes including processing temperatures of ≥ 400° C., and in some cases ≥ 500° C., or ≥ 600° C., and 650° C. or less). It is strong enough and remains debondable by application of sufficient separating force and, however, by force that will not cause catastrophic damage to the thin sheet 20 and/or carrier 10. This debonding allows the removal of the thin sheet 20 and the devices fabricated thereon, and also allows reuse of the carrier 10.

Although the surface modification layer 30 is shown as a solid layer between the thin sheet 20 and the carrier 10, that need not be the case. For example, layer 30 may be approximately 0.1 to 2 nm thick and may not completely cover all of the bonding surfaces 14. For example, the coverage may be ≤ 100%, 1% to 100%, 10% to 100%, 20% to 90%, or 50% to 90%. In other embodiments, layer 30 may be 10 nm thick or less, or in other embodiments 100 nm thick or less. The surface modification layer 30 can be considered to be disposed between the carrier 10 and the thin sheet 20 even though it cannot contact either the carrier 10 or the thin sheet 20. In any case, an important aspect of the surface modification layer 30 modifies the ability of the bonding surface 14 to bond with the bonding surface 24, thereby increasing the strength of the bonding between the carrier 10 and the thin sheet 20. To control. The material and thickness of the surface modification layer 30, as well as the treatment of the bonding surfaces 14, 24 prior to bonding, can be used to control the strength of the bonding between the carrier 10 and the thin sheet 20 (adhesive energy). have.

In general, the adhesion energy between two surfaces is given by ("A theory for the estimation of surface and interfacial energies. I. derivation and application to interfacial tension", LA Girifalco and RJ Good, J. Phys. Chem., V 61, p904):

<Equation 1>

는 각각 표면 1, 표면 2의 표면 에너지, 및 표면 1 및 2의 계면 에너지이다. 각각의 표면 에너지는 보통 두 항; 분산 성분 γ^d, 및 극성 성분 γ^p의 조합이다.In the above formula,

Is the surface energy of the surface 1, the surface 2, and the interfacial energy of the surfaces 1 and 2, respectively. Each surface energy is usually two terms; It is a combination of a dispersion component γ ^d and a polar component γ ^p.

<Equation 2>

If the adhesion was mainly due to London dispersing ^{force (γ d} ) and polar force, for example hydrogen bonding (γ ^p ), the interfacial energy could be given by (Girifalco and RJ Good, as mentioned above) :

<Equation 3>

After substituting Equation 3 into Equation 1, the adhesion energy could be roughly calculated as follows:

<Equation 4>

In Equation 4, only van der Waals (and/or hydrogen bonds) component of the adhesion energy is considered. These include polar-polar interactions (Keesom), polar-nonpolar interactions (Debye) and nonpolar-nonpolar interactions (London). However, other attractive energies such as covalent and electrostatic bonds may also be present. Thus, in a more generalized form, the above equation is described as follows:

<Equation 5>

Where w _c and w _e are covalent and electrostatic adhesion energies. Covalent adhesion energy is rather common, as in silicon wafer bonds where the initial hydrogen bonded pair of wafers is heated to a higher temperature to convert many or all silanol-silanol hydrogen bonds into Si-O-Si covalent bonds. Initial, room temperature, hydrogen bonding ^{produces an adhesion energy of approximately ~100-200mJ/m 2} allowing separation of the bonded surfaces, while fully covalently bonded wafers such as those achieved during high temperature processing (approximately 400-800°C) The pair has an adhesion energy of ^{~1000-3000 mJ/m 2 which} does not allow separation of the bonded surfaces; Instead, the two wafers act as a single piece. On the other hand, when both surfaces are fully coated, with a thickness large enough to shield the effect of the underlying substrate, using a low surface energy material, such as a fluoropolymer, the adhesion energy is the adhesion of the coating material. Energy will be, and will be very low resulting in low or no adhesion between the bonding surfaces 14, 24, whereby the thin sheet 20 will not be able to be processed on the carrier 10. Consider two extreme cases: (a) twice standard clean 1 (SC1, as known in the art), the glass surface saturated with the cleaned silanol groups is hydrogen bonded (thereby the adhesion energy is ~100-200 mJ/). m ² ) and then heated to high temperature to convert the silanol groups into covalent Si-O-Si bonds (thereby the adhesion energy is 1000-3000 mJ/m ² ). The adhesive energy of the latter is too high to make the pair of glass surfaces detachable; (b) Two glass surfaces completely coated with fluoropolymer are bonded ^{at room temperature with low surface adhesion energy (~12 mJ/m 2} per surface) and heated to high temperature. In the case of the latter (b) above, the surface is ^{not only not bonded (because the total adhesion energy of ~24 mJ/m 2} is too low when the surfaces are combined), but also because the surface has no polar reactive groups (or too little) at high temperature. Is also not combined. Between these two extremes, the range of adhesion energy ^{exists, for example 50-1000 mJ/m 2} , which can produce the desired degree of controlled bonding. Therefore, the present inventors induce the adhesion energy between these two extremes, and maintain a pair of glass substrates bonded to each other (e.g. glass carrier 10 and thin glass sheet 20) during the harshness of FPD processing. It is not only sufficient as necessary for the following, but also allows a controlled bond to be created that allows for the detachment of the thin sheet 20 from the carrier 10 after the processing is complete (even after high-temperature processing, for example ≥ 400°C). Various ways of providing the surface modification layer 30 have been found. Further, the detachment of the thin sheet 20 from the carrier 10 is by mechanical force and at least there is no catastrophic damage to the thin sheet 20, so preferably also no catastrophic damage to the carrier 10. Can be done in that way.

Equation 5 states that the adhesion energy is a function of the four surface energy parameters + covalent and electrostatic energy if present.

Appropriate adhesion energy can be achieved by careful selection of the surface modifier, i.e. the surface modifying layer 30, and/or the heat treatment of the surface prior to bonding. Appropriate adhesion energies can be achieved by the selection of chemical modifiers of one or both of the bonding surface 14 and the bonding surface 24, which in turn is van der Waals (and/or hydrogen bonding, these terms are used throughout the specification). Used interchangeably throughout) as well as the similar covalent bonding energy resulting from high temperature processing (eg, approximately ≧400° C.). For example, taking the bonding surface of SC1 cleaned glass (which is initially saturated with silanol groups with high polar components of surface energy) and coated with a low energy fluoropolymer to determine the functional coverage of the surface by polar and non-polar groups. Provides control. This not only provides control of the initial van der Waals (and/or hydrogen) bonds at room temperature, but also provides control of the degree of covalent bonding/degree of covalent bonding at higher temperatures. Control of the initial van der Waals (and/or hydrogen) bonds at room temperature allows for vacuum and or rotation-rinse-dry (SRD) type processing of one surface to another, and in some cases also easily formed other It is carried out to provide a bonding of one surface to the surface-here the easily formed bonding is a thin sheet, such as was done when pressing the thin sheet 20 onto the carrier 10 using a squeegee, or using a depressurized environment. It can be carried out at room temperature without the application of an externally applied force over the entire area of (20). That is, the initial van der Waals bonding provides at least a minimal degree of bonding that holds the thin sheet and carrier together so that one is fixed and the other does not separate when gravity is applied. In most cases, the initial van der Waals (and/or hydrogen) bonding will have such a degree that the article will also be able to undergo vacuum, SRD, and ultrasonic processing without delamination of the thin sheet from the carrier. Van der through the surface modification layer 30 (including the surface treatment of the material from which the surface modification layer is made and/or the surface to which the surface modification layer is applied), and/or by heat treatment of the bonding surface prior to bonding them together. This precise control at the appropriate level of both Vals (and/or hydrogen bonding) and covalent interactions allows the thin sheet 20 to bond with the carrier 10 during FPD style processing, while simultaneously FPD style processing. The desired adhesion energy is then achieved which allows the thin sheet 20 to be separated from the carrier 10 (by an appropriate force preventing damage to the thin sheet 20 and/or the carrier). Also, under appropriate circumstances, an electrostatic charge could be applied to one glass surface or both to provide another level of control of the adhesion energy.

FPD processing, e.g. p-Si and oxide TFT fabrication, typically exceeds 400° C. which will lead to glass-to-glass bonding bonding the thin glass sheet 20 with the glass carrier 10 in the absence of the surface modification layer 30. , Thermal processes at temperatures above 500°C, and in some cases above 600°C and below 650°C. Thus, control of the formation of Si-O-Si bonds leads to reusable carriers. One way to control the formation of Si-O-Si bonds at elevated temperatures is to reduce the concentration of surface hydroxyls on the surfaces to be bonded.

Hydroxyl per square nm, as shown in Figure 3, which is a plot of Iler of surface hydroxyl concentration on silica as a function of temperature (RK Iller: The Chemistry of Silica (Wiley-Interscience, New York, 1979). The number of (OH groups) decreases as the temperature of the surface increases. Thus, heating of the silica surface (and by analogy to the glass surface, for example the bonding surface 14 and/or the bonding surface 24) By reducing the concentration of hydroxyl, the likelihood of the hydroxyls interacting on the two glass surfaces is reduced This reduction in the surface hydroxyl concentration consequently reduces the Si-O-Si bonds formed per unit area, resulting in lower adhesion. However, removal of surface hydroxyls requires long annealing times at high temperatures (> 750° C. to completely remove surface hydroxyls) These long annealing times and high annealing temperatures are expensive processes, and the strain point of typical display glasses. There is a possibility of exceeding, resulting in an impractical process.

From the above analysis, the inventors have found that an article comprising a thin sheet and a carrier, suitable for FPD processing (including LTPS processing), can be manufactured by balancing the following three concepts:

(1) Intermediate adhesion sufficient to control van der Waals (and/or hydrogen) bonding to facilitate initial room temperature bonding and to withstand non-hot FPD processes such as vacuum processing, SRD processing, and/or ultrasonic processing. Carrier and/or thin sheet bonding surfaces by controlling the initial room temperature bonding, which can be accomplished by generating ^{energy (e.g., having a surface energy of >40 mJ/m 2} per surface prior to bonding the surfaces) Modification of s);

(2) in a thermally stable manner to withstand the FPD process without delamination and/or unacceptable contamination in device fabrication, e.g., outgassing that may cause unacceptable contamination to the semiconductor and/or display manufacturing process in which the article may be used. The surface modification of the carrier and/or the thin sheet of the material; And

(3) It can be carried out by controlling the concentration of hydroxyl on the carrier surface, and of other species that can form strong covalent bonds at elevated temperatures (e.g., temperatures of ≥ 400°C), whereby even after high temperature processing (especially the FPD process). As in, the adhesion between the carrier and the thin sheet (during the thermal process in the range of 500-650° C.) at least does not damage the thin sheet (preferably does not damage the thin sheet or carrier), but the carrier and the thin sheet during processing The bonding energy between the bonding surface of the carrier and the thin sheet can be controlled so that it is within a range that allows debonding of the thin sheet from the carrier with sufficient separating force as necessary to maintain the bonding between them so as not to be delaminated, Control of bonding at high temperatures.

In addition, the inventors have found that the use of the surface modification layer 30 allows the controlled bonding area, i.e. article 2, to be processed in FPD-type processes (including vacuum and wet processes), with appropriate bonding surface preparation. Provides room temperature bonding between the thin sheet 20 and the carrier 10, which is sufficient for processing, but after the article 2 has finished high-temperature processing, e.g., FPD type processing, or LTPS processing, the thin sheet 20 is transferred to the carrier 10 A bond that controls the covalent bonding between the thin sheet 20 and the carrier 10 (even at an elevated temperature of ≥ 400° C.) so that it can be removed (at least without damage to the thin sheet, preferably also without damage to the carrier). It has been found that it is possible to balance this concept with ease to achieve the realm. Each suitability was evaluated using a series of tests to evaluate the potential bonding surface preparation, and surface modification layer, which would provide a reusable carrier suitable for FPD processing. Different FPD applications have different requirements, but the LTPS and oxide TFT processes appear to be the most stringent at this point, and thus the tests representing the steps in this process were chosen because they are the desired application for article 2. Vacuum processes, wet cleaning (including SRD and ultrasonic type processes) and wet etching are typical for many FPD applications. Typical aSi TFT fabrication requires processing below 320°C. Annealing at 400°C is used in the oxide TFT process, while crystallization and dopant activation steps above 600°C are used in LTPS processing. Thus, the following five tests are used to ensure that the specific bonding surface preparation and surface modification layer 30 allows the thin sheet 20 to remain bonded to the carrier 10 during FPD processing, while such processing (≥ 400° C. The possibility of allowing the thin sheet 20 to be removed from the carrier 10 (without damage to the thin sheet 20 and/or the carrier 10) after processing at temperature) was evaluated. The tests were run in sequence, and samples were run from one test to the next unless there was a type of failure that would not allow subsequent testing.

(1) Vacuum test. Vacuum compatibility testing was performed on a STS Multiplex PECVD loadlock (available from SPTS, Newport, UK)-the loadlock was an Evara A10S dry pump (Sacramento, CA) with a soft pump valve. Pumped by Ebara Technologies Inc.). The sample was placed in a load lock, then the load lock was pumped from atmospheric pressure to 70 mTorr within 45 seconds. The failure, indicated by the notation of "F" in the "vacuum" column of the table below, was considered to have occurred when the following was present: (a) loss of adhesion between the carrier and the thin sheet (by visual inspection by the naked eye, where Failure was believed to have occurred when the thin sheet fell from the carrier or was partially debonded therefrom); (b) Bubbling between the carrier and the thin sheet (as determined by visual inspection with the naked eye-pictures of the samples were taken before and after processing and then compared, failure occurred when the defect increased in size by the visible dimension. Decided); Or (c) movement of the thin sheet relative to the carrier (as determined by visual inspection with the naked eye-a photograph of the sample was taken before and after the test, where the failure was the movement of the bonding defect, e.g. in the presence of bubbles, or the edge Is believed to have occurred if there is a debonding, or if there is a movement of the thin sheet on the carrier). In the table below, the notation of “P” in the “vacuum” column indicates that the sample did not fail according to the above criteria.

(2) Wet process test. Wet process suitability testing was performed using a Semitool model SRD-470S (available from Applied Materials, Santa Clara, Calif.). The test consisted of a 60 second 500 rpm rinse, a Q-rinse with 15 MOhm-cm at 500 rpm, a 10 second purge at 500 rpm, 90 seconds drying at 1800 rpm, and 180 seconds drying at 2400 rpm under warm flowing nitrogen. Failure, indicated by the notation of "F" in the "SRD" column of the table below, was considered to have occurred when the following was present: (a) loss of adhesion between the carrier and the thin sheet (by visual inspection by the naked eye, where Failure was believed to have occurred when the thin sheet fell from the carrier or was partially debonded therefrom); (b) Bubbling between the carrier and the thin sheet (as determined by visual inspection with the naked eye-pictures of the samples were taken before and after processing and then compared, failure occurred when the defect increased in size by the visible dimension. Decided); Or (c) movement of the thin sheet relative to the carrier (as determined by visual inspection with the naked eye-a photograph of the sample was taken before and after the test, where the failure was the movement of the bonding defect, for example in the presence of bubbles, or the edge Was believed to have occurred when there was debonding, or when there was migration of the thin sheet on the carrier; or (d) penetration of water under the thin sheet (as determined by visual inspection at 50x with an optical microscope, where the failure was the liquid Or the residue was determined to have occurred if observable) In the table below, the designation of “P” in the “SRD” column indicates that the sample did not fail according to the above criteria.

(3) Temperature test up to 400°C. 400° C. process suitability tests were performed using Alwin21 Accuthermo610 RTP (available from Alwin21, Santa Clara, Calif.). The carrier to which the thin sheet was bonded thereto was heated in a chamber of a cycle that was made to 400° C. at room temperature at 6.2° C./min, maintained at 400° C. for 600 seconds, and cooled to 300° C. at 1° C./min. The carrier and thin sheet were then allowed to cool to room temperature. The failure, as indicated by the notation of “F” in the “400° C.” column of the table below, was considered to have occurred when the following was present: (a) loss of adhesion between the carrier and the thin sheet (visual inspection by the naked eye) Whereby the failure was believed to have occurred when the thin sheet fell from the carrier or was partially debonded therefrom); (b) Bubbling between the carrier and the thin sheet (as determined by visual inspection with the naked eye-pictures of the samples were taken before and after processing and then compared, failure occurred when the defect increased in size by the visible dimension. Decided); Or (c) increased adhesion between the carrier and the thin sheet, wherein this increased adhesion (by insertion of a razor blade between the thin sheet and the carrier, and/or a Kapton™ tape, 1" wide x 6" long By sticking the piece to a thin sheet and pulling the tape so that it is attached to a 2-3" 100 mm square thin glass (K102 series from Saint Gobain Performance Plastic, NY Dessert)) Prevents the debonding of the thin sheet from the carrier without damage, wherein the failure is if there is damage to the thin sheet or carrier in an attempt to separate the thin sheet or carrier, or the thin sheet and carrier are either of the above debonding methods. It was believed to have occurred if it could not be debonded by the performance of). In addition, after bonding the thin sheet with the carrier, and before thermal cycling, a debonding test was performed on a representative sample to perform any associated surface treatment. It was determined that the specific material it contained allowed the debonding of the thin sheet from the carrier prior to temperature cycling In the table below, the designation of “P” in the “400° C.” column indicates that the sample did not fail according to the above criteria. .

(4) Temperature test up to 600°C. The 600° C. process suitability test was performed using an Alwin 21 Accu Thermo 610 RTP. The carrier comprising the thin sheet was circulated from room temperature to 600° C. at 9.5° C./min, held at 600° C. for 600 seconds, and then heated in a chamber cooled to 300° C. at 1° C./min. The carrier and thin sheet were then allowed to cool to room temperature. The failure, as indicated by the notation of “F” in the “600° C.” column of the table below, was considered to have occurred when the following was present: (a) loss of adhesion between the carrier and the thin sheet (visual inspection by the naked eye) Whereby the failure was believed to have occurred when the thin sheet fell from the carrier or was partially debonded therefrom); (b) Bubbling between the carrier and the thin sheet (as determined by visual inspection with the naked eye-pictures of the samples were taken before and after processing and then compared, failure occurred when the defect increased in size by the visible dimension. Decided); Or (c) increased adhesion between the carrier and the thin sheet, wherein this increased adhesion is achieved (by inserting a razor blade between the thin sheet and the carrier, and/or by attaching a piece of Kapton™ tape as described above to the thin sheet and applying the tape. By pulling) to prevent debonding of the thin sheet from the carrier without damaging the thin sheet or carrier, where the failure is if there is damage to the thin sheet or carrier when attempting to separate the thin sheet or carrier, or the thin sheet and carrier It was believed to have occurred when it could not be detached by performing any of the above debonding methods). In addition, after bonding the thin sheet with the carrier, and prior to thermal cycling, desorption tests were performed on representative samples to ensure that certain materials, and any associated surface treatments, allowed debonding of the thin sheet from the carrier before temperature cycling. I decided. In the table below, the notation of “P” in the “600° C.” column indicates that the sample did not fail according to the above criteria.

(5) Ultrasonic test. The ultrasonic compatibility test was performed by cleaning the article in four tank lines, where the article was processed in each tank sequentially from tank #1 to tank #4. The tank dimensions for each of the four tanks were 18.4"L x 10"W x 15"D. Two cleaning tanks (#1 and #2) were in DI water at 50° C., Yokohama Oils & Pats, Yokohama, Japan. Containing 1% Semiclean KG, available from Yokohama Oils and Fats Industry Co Ltd. Cleaning tank #1 was a NEY Prosonic 2 104 kHz ultrasonic generator (Blackstone, Jamestown, NY). NEY Ultrasonics (available from Blackstone-NEY Ultrasonics) and rinse tank #2 was stirred with a NEY Prosonic 2 104 kHz ultrasonic generator, two rinse tanks (tank #3 and tank #4) at 50°C. Rinse tank #3 was stirred with a NEY SweepSonic 2D 72 kHz ultrasonic generator and rinse tank #4 was agitated with a NEY SweepSonic 2D 104 kHz ultrasonic generator The process was 10 in each tank #1-4. After running for a minute, a rotary rinse drying (SRD) was performed after the sample was removed from tank # 4. As indicated by the notation of “F” in the “ultrasonic” column of the table below, the failure is when the following is present. It was believed to have occurred: (a) loss of adhesion between the carrier and the thin sheet (by visual inspection, where the failure was believed to have occurred if the thin sheet fell from the carrier or partially debonded therefrom); ( b) bubbling between the carrier and the thin sheet (as determined by visual inspection with the naked eye-pictures of the samples were taken before and after processing and then compared, and the failure was determined to have occurred when the defect increased in size by the visible dimension. Or (c) formation of other significant defects (as determined by visual inspection at 50x with an optical microscope, where the failure occurred when there were particles trapped between the thin sheet and the carrier that were not previously observed); or (c) the formation of other significant defects (as determined by visual inspection at 50x with an optical microscope); Was considered); Or (d) penetration of water down the thin sheet (as determined by visual inspection at 50x with an optical microscope, where the failure was determined to have occurred if the liquid or residue was observable). In the table below, the notation of "P" in the "ultrasonic" column indicates that the sample did not fail according to the above criteria. In addition, in the table below, a blank or "?" in the "ultrasonic" column indicates that the sample was not tested by the above method.

Preparation of bonding surface through reduction of hydroxyl by heating

The advantage of modifying one or more of the bonding surfaces 14, 24 with a surface modification layer 30 so that the article 2 can successfully undergo FPD processing (i.e., the thin sheet 20 is applied to the carrier 10 during processing. An article (2) having a glass carrier (10) and a thin glass sheet (20) without a surface modification layer (30) therebetween, which remain bonded but can be separated from the carrier (10) after processing, including high-temperature processing It was proved by processing. In particular, it was attempted to prepare the bonding surfaces 14 and 24 by first heating to reduce the hydroxyl groups, but without the surface modification layer 30. The carrier 10 and thin sheet 20 were cleaned, the bonding surfaces 14 and 24 were bonded together, and then the article 2 was tested. A typical cleaning process for making bonding glasses is in dilute hydrogen peroxide and base (usually ammonium hydroxide, but a solution of tetramethylammonium hydroxide, for example JT Baker JTB-100 or JTB-111 can also be used). It is an SC1 cleaning process that is cleaned. Cleaning removes particles from the bonding surface and creates a known surface energy, ie it provides a baseline for the surface energy. Other types of cleaning may also be used, since the mode of cleaning need not be SC1, and the type of cleaning can have only very little effect on the silanol groups on the surface. The results of various tests are shown in Table 1 below.

Strong but separable initial, room temperature or van der Waals and/or hydrogen-bonds, respectively, of approximately 0.2 nm, available from Eagle XG® display glass (Corning Incorporated, Corning, NY). 100 mm square x 100 micron thick thin glass sheet, including alkali-free, alumino-boro-silicate glass, having an average surface roughness Ra, and a 150 mm diameter single average flat (SMF) wafer 0.50 or 0.63 It was created by simply cleaning the mm thick glass carrier. In the above example, the glass was washed in a 65° C. bath of 40:1:2 DI water:JTB-111:hydrogen peroxide for 10 minutes. The thin glass or glass carrier may or may not be annealed at 400° C. for 10 minutes in nitrogen to remove residual water-the designation “400 in the “carrier” column or “thin glass” column in Table 1 below. "°C" indicates that the sample was annealed at 400°C for 10 minutes in nitrogen. The FPD process suitability test demonstrates that the SC1-SC1 initial, room temperature, bond is mechanically strong enough to pass vacuum, SRD and ultrasonic tests. However, heating at 400° C. and above resulted in a permanent bond between the thin glass and the carrier, ie the thin glass sheet could not be removed from the carrier without damage to one or both of the thin glass sheet and the carrier. And this was also true for Example 1c, where each carrier and thin glass had an annealing step to reduce the concentration of surface hydroxyl. Thus, the above-described preparation of the bonding surfaces 14, 24 through heating alone and subsequently bonding of the carrier 10 and the thin sheet 12, without the surface modification layer 30, is a FPD process where the temperature is ≥ 400°C. Is not suitable for controlled binding.

<Table 1>-Process suitability test of SC1-treated glass bonding surface

Preparation of bonding surface by hydroxyl reduction and surface modification layer

Hydroxyl reduction, such as, for example, by heat treatment, and the surface modification layer 30 can be used together to control the interaction of the bonding surfaces 14 and 24. For example, the binding energy of the bonding surfaces 14, 24 (both van der Waals and/or hydrogen-bonds at room temperature due to the polar/dispersive energy component, and covalent bonds at high temperature due to the covalent energy component) It can be controlled to provide varying bond strength, from difficult room temperature bonding to allowing easy room temperature bonding and separation of the bonding surface after high temperature processing,-after high temperature processing-preventing the surface from separating without damage. have. In some applications, it may be desirable to have no or very weak bonds (when the surface is as described in the thin sheet/carrier concept of US '727 and is present in a “non-bonding” region as described below). have. In other applications that provide re-usable carriers, for example for FPD processes, etc., where process temperatures ≥ 500° C., or ≥ 600° C., and 650° C. or less can be achieved, initially at room temperature with thin sheets and It is preferred to have sufficient van der Waals and/or hydrogen-bonds to bond the carriers together, but to prevent or limit hot covalent bonds. For another application, sufficient room temperature to initially join the thin glass and carrier together (if the surface is as described in the thin sheet/carrier concept of US '727 and is present in the “bonding region” as discussed below). It may be desirable to have bonds and also generate strong covalent bonds at high temperatures. While not wishing to be bound by theory, in some cases the room temperature bonding where the thin sheet and the carrier are initially joined together can be controlled using a surface modification layer, while the hydroxyl groups on the surface (e.g., surface heating Reduction by or by reaction of hydroxyl groups with the surface modification layer can be used to control covalent bonding, especially at high temperatures.

The material for the surface modification layer 30 will provide a bonding surface 14, 24 having energy (e.g., energy <40 mJ/m ² as measured on one surface, and including polar and dispersive components). And thus the surface only creates weak bonds. In one example, hexamethyldisilazane (HMDS) can be used to cause this low energy surface by reacting with surface hydroxyls, leaving a trimethylsilyl (TMS) terminal surface. Both room temperature and high temperature bonding can be controlled by using HMDS together as a surface modification layer while surface heating to reduce the hydroxyl concentration. By selecting a bonding surface preparation method suitable for each bonding surface 14, 24, articles having various capabilities can be achieved. More particularly, among those of interest in providing reusable carriers for LTPS processing, each vacuum SRD, 400° C. (parts a and c), and 600° C. (parts a and c), to withstand (or pass) processing tests. ) A suitable bonding can be achieved between the thin glass sheet 20 and the glass carrier 10.

In one example, HMDS treatment of both thin glass and carrier after SC1 cleaning produces a weakly bonded surface that is difficult to bond at room temperature with Van der Waals (and/or hydrogen bonding) forces. Mechanical force is applied to bond the thin glass to the carrier. As shown in Example 2a of Table 2, the bonding was sufficiently weak that the warpage of the carrier was observed in the vacuum test and SRD processing, and bubbling (possibly due to outgassing) was observed in the thermal process at 400°C and 600°C Defects were observed after ultrasonic processing.

In another embodiment, HMDS treatment of only one surface (carrier in the cited example) produces a stronger room temperature adhesion that withstands vacuum and SRD processing. However, thermal processes above 400° C. permanently bonded the thin glass to the carrier. This indicates that the maximum surface coverage of trimethylsilyl groups on silica is described in Sindorf and Maciel, J. Phys. Chem. 1982, 86, 5208-5219], it was calculated to be ^{2.8/nm 2 and described in Suratwala et.} al., Journal of Non-Crystalline Solids 316 (2003) 349-363], compared to the hydroxyl concentration of ^{4.6-4.9/nm 2} in the case of completely hydroxylated silica, it was unexpected because it was measured as ^{2.7/nm 2.} It's not a job. That is, the trimethylsilyl group will bond with some surface hydroxyl, but some unbound hydroxyl will remain. Therefore, considering sufficient time and temperature, condensation of surface silanol groups that permanently bonds the thin glass and the carrier will be expected.

The altered surface energy can be created by heating the glass surface to reduce the surface hydroxyl concentration prior to HMDS exposure, leading to an increased polar component of the surface energy. This not only reduces the driving force for the formation of covalent Si-O-Si bonds at high temperatures, but also leads to stronger room temperature bonds, such as van der Waals (and/or hydrogen) bonds. 4 shows the surface energy of the Eagle XG® display glass carrier after annealing and after HMDS treatment. The elevated annealing temperature prior to HMDS exposure increases the total (polar and disperse) surface energy (line 402) after HMDS exposure by increasing the polarity contribution (line 404). It is also shown that the dispersion contribution to the total surface energy (line 406) remains largely unchanged by the heat treatment. While not wishing to be bound by theory, the increase in the polar component at the surface after HMDS treatment, and hence the increase in total energy, is that some exposed glass surface areas exist even after HMDS treatment because of the sub-monolayer TMS coverage by HMDS. It seems to be caused by.

In Example 2b, the thin glass sheet was heated under vacuum at a temperature of 150° C. for 1 hour before bonding with a non-heat-treated carrier with a coating of HMDS. This heat treatment of the thin glass sheet was not sufficient to prevent the permanent bonding of the thin glass sheet to the carrier at a temperature of ≧400°C.

As shown in Examples 2c-2e of Table 2, changing the annealing temperature of the glass surface prior to HMDS exposure can change the binding energy of the glass surface to control the bonding between the glass carrier and the thin glass sheet.

In Example 2c, the carrier was annealed for 1 hour under vacuum at a temperature of 190° C., and then HMDS exposure was performed to provide the surface modification layer 30. In addition, the thin glass sheet was annealed for 1 hour under vacuum at 450° C. before bonding with the carrier. The resulting article survived the vacuum, SRD, and 400° C. tests (parts a and c, but failed to pass part b due to increased bubbling), but failed the 600° C. test. Thus, there is an increased resistance to hot bonding compared to Example 2b, but this is not sufficient to make an article for processing at temperatures ≧600° C. where the carrier is reusable (eg LTPS processing).

In Example 2d, the carrier was annealed for 1 hour under vacuum at a temperature of 340° C., followed by HMDS exposure, to provide a surface modification layer 30. Again, the thin glass sheet was annealed under vacuum at 450° C. for 1 hour before bonding with the carrier. The results were similar to those of Example 2c, and the article withstood the vacuum, SRD, and 400°C tests (parts a and c, but did not pass part b due to increased bubbling), but failed the 600°C test. .

As shown in Example 2e, after annealing both the thin glass and the carrier under vacuum at 450° C. for 1 hour, the HMDS exposure of the carrier was performed, and then bonding the carrier and the thin glass sheet was performed at the temperature for permanent bonding. Improves resistance. Annealing of both surfaces to 450° C. prevents permanent bonding after RTP annealing at 600° C. for 10 minutes, ie the sample failed to pass 600° C. processing tests (parts a and c, but part b due to increased bubbling; Similar results were observed in the case of the 400° C. test).

<Table 2>-Process suitability test of HMDS surface modified layer

In Examples 2a-2e above, each carrier and thin sheet was Eagle XG® glass, wherein the carrier was a 150 mm diameter SMF wafer 630 micrometers thick, and the thin sheet 100 mm square 100 micrometers thick. HMDS was applied by pulse evaporation in a YES-5 HMDS oven (available from Yield Engineering Systems, San Jose, Calif.), which was one atomic layer thick (i.e., about 0.2 to 1 nm), but only the surface The coating may be less than a single layer, ie some of the surface hydroxyls may be described by Maciel and not covered by HMDS as discussed above. Due to the small thickness in the surface modification layer, there is little risk of outgassing which can cause contamination in device fabrication. In addition, as indicated in Table 2 by the notation "SC1", each carrier and thin sheet was cleaned using the SC1 process prior to heat treatment or any subsequent HMDS treatment.

The comparison of Example 2b and Example 2a shows that the bonding energy between the thin sheet and the carrier can be controlled by varying the number of surfaces comprising the surface modification layer. And the bonding force between the two bonding surfaces can be controlled using the control of the bonding energy. Further, the comparison of Examples 2b-2e shows that the bonding energy of the surface can be controlled by changing the parameters of the heat treatment conducted on the bonding surface prior to application of the surface modifying material. Again, heat treatment can be used to reduce the number of surface hydroxyls, and thus, the degree of covalent bonding, especially at high temperatures, can be controlled.

Other materials that can act in different ways to control the surface energy on the bonding surface can be used for the surface modification layer 30 to control the room temperature and high temperature bonding forces between the two surfaces. For example, by modifying one or both bonding surfaces with a surface modification layer that sterically hinders or coats a species such as hydroxyl to prevent the formation of strong permanent covalent bonds between the carrier and the thin sheet at elevated temperature. Reusable carriers can also be created if they create a degree of bonding force. One way to generate variable surface energy and to coat surface hydroxyls to prevent the formation of covalent bonds is the deposition of a plasma polymer film, for example a fluoropolymer film. Plasma polymerization can be carried out at atmospheric pressure or reduced pressure, and source gases such as fluorocarbon sources (including CF4, CHF3, C2F6, C3F6, C2F2, CH3F, C4F8, chlorofluorocarbons, or hydrochlorofluorocarbons), hydrocarbons, e.g. For example, from alkanes (including methane, ethane, propane, butane), alkenes (including ethylene, propylene), alkynes (including acetylene), and aromatics (including benzene, toluene), hydrogen, and other gas sources, such as SF6. A thin polymer film is deposited under plasma excitation (DC or RF parallel plate, inductively coupled plasma (ICP) electron cyclotron resonance (ECR) downstream microwave or RF plasma). Plasma polymerization produces a layer of highly crosslinked material. The reaction conditions and control of the source gas can be used to control the film thickness, density, and chemistry to tailor the functional groups to the desired application.

FIG. 5 shows the overall (line (line) of plasma polymerized fluoropolymer (PPFP) film deposited from a CF4-C4F8 mixture using an Oxford ICP380 etch tool (available from Oxford Instruments, Oxfordshire, UK). 502)) surface energy (including polarity (line 504) and dispersion (line 506) components). The film was deposited on a sheet of Eagle XG® glass, and spectroscopic ellipsometry indicated that the film was 1-10 nm thick. As can be seen from Figure 5, the glass carrier treated with a plasma polymerized fluoropolymer film containing less than 40% C4F8 ^{exhibits >40 mJ/m 2} surface energy, and is thin by van der Waals or hydrogen bonding at room temperature. It creates a controlled bond between the glass and the carrier. Facilitated bonding is observed when initially bonding the carrier and thin glass at room temperature. That is, if thin sheets were placed on the carrier and pressed together at one point, the wavefront would cross the entire carrier, but at a slower rate than observed in the case of an SC1 treated surface that did not have a surface modification layer thereon. Move. Controlled bonding is sufficient to withstand all standard FPD processes including vacuum, wet, ultrasonic, and thermal processes below 600°C, i.e., the controlled bonding is subjected to 600°C processing tests without delamination or motion of thin glass from the carrier. Passed. Debonding was achieved by peeling with a razor blade and/or Kapton™ tape as described above. The process suitability of two different PPFP films (deposited as described above) is shown in Table 3. PPFP 1 of Example 3a was formed with C4F8/(C4F8+CF4)=0, that is, formed with CF4/H2 instead of C4F8, and PPFP 2 of Example 3b was deposited with C4F8/(C4F8+CF4)=0.38. . Both types of PPFP films withstood vacuum, SRD, 400°C and 600°C processing tests. However, after 20 minutes of ultrasonic cleaning of PPFP 2, delamination was observed showing insufficient adhesion to withstand this processing. Nevertheless, the surface modification layer of PPFP2 can be useful for some applications, such as when ultrasonic processing is not required.

<Table 3>-Process suitability test of PPFP surface modified layer

In Examples 3a and 3b above, each carrier and thin sheet was Eagle XG® glass, wherein the carrier was a 150 mm diameter SMF wafer 630 micrometers thick, and the thin sheet was 100 square mm 100 micrometers thick. Due to the small thickness in the surface modification layer, there is little risk of outgassing which can cause contamination during device fabrication. Also, since the surface modification layer did not appear to degrade again, the risk of outgassing is less. In addition, as shown in Table 3, each thin sheet was cleaned using an SC1 process before heat treatment at 150° C. for 1 hour under vacuum.

Another material that can act in different ways to control the surface energy can be used as the surface modification layer to control the room temperature and high temperature bonding forces between the thin sheet and the carrier. For example, bonding surfaces capable of producing controlled bonding can be created by silanizing a glass carrier and/or a thin sheet of glass. Silanes are chosen to produce suitable surface energy and to have sufficient thermal stability for the application. The carrier or thin glass to be treated is cleaned by a process, e.g., O2 plasma or UV-ozone, and SC1 or standard clean-to (SC2 as known in the art) cleaning to prevent silane reacting with surface silanol groups Organics and other impurities (eg metals) can be removed. Other chemical based washings, such as HF, or H2SO4 washing chemicals, can also be used. The carrier or thin glass can be heated to control the surface hydroxyl concentration prior to silane application (as discussed above with respect to the surface modification layer of HMDS) and/or complete the silane condensation with the surface hydroxyl after silane application. Can be heated to do so. The concentration of unreacted hydroxyl groups after silanization can be made low enough prior to bonding to prevent permanent bonding between the thin glass and the carrier at temperatures> 400° C., ie to form a controlled bond. This approach is described below.

Example 4a

Subsequently, the glass carrier with O2 plasma and SC1 treated bonding surface was treated with 1% dodecyltriethoxysilane (DDTS) in toluene and annealed at 150° C. under vacuum for 1 hour to complete condensation. The DDTS-treated surface showed a surface energy ^{of 45 mJ/m 2.} As shown in Table 4, a glass thin sheet (SC1 cleaned and heated at 400° C. under vacuum for 1 hour) was bonded to a carrier bonding surface with a DDTS surface modification layer thereon. The article survived wet and vacuum process tests, but failed to withstand thermal processes above 400° C. without bubble formation under the carrier due to thermal decomposition of the silane. The thermal decomposition results in all linear alkoxy and chloroalkylsilanes R1 _x Si(OR2) _y (Cl) _z (where x=1-3, and y+z=4-x, methyl, which yields a coating with good thermal stability, dimethyl-, and trimethyl-silane (x = 1 to _3, R1 = CH 3) is expected in the case of not including).

Example 4b

Then, the glass carrier having the bonding surface treated with O2 plasma and SC1 was treated with 1% 3,3,3 in toluene, trifluoropropyltriethoxysilane (TFTS), and annealed at 150° C. under vacuum for 1 hour. Condensation is complete. The TFTS-treated surface showed a surface energy ^{of 47 mJ/m 2.} As shown in Table 4, a glass thin sheet (SC1 cleaned and then heated at 400° C. for 1 hour under vacuum) was bonded to a carrier bonding surface with a TFTS surface modification layer thereon. The article survived vacuum, SRD, and 400° C. process tests without permanent bonding of the glass thin sheet to the glass carrier. However, the 600° C. test resulted in bubble formation under the carrier due to thermal decomposition of the silane. This was not unexpected because of the limited thermal stability of the propyl group. Although the sample failed the 600° C. test due to bubbling, the material and heat treatment of the sample can be used in some applications where bubbling and its side effects, such as a decrease in surface flatness, or an increased surface waveform may be tolerated. .

Example 4c

Subsequently, the glass carrier with O2 plasma and SC1 treated bonding surface was treated with 1% phenyltriethoxysilane (PTS) in toluene and annealed at 200° C. under vacuum for 1 hour to complete condensation. The PTS-treated surface showed a surface energy ^{of 54 mJ/m 2.} As shown in Table 4, a glass thin sheet (SC1 cleaned and then heated at 400° C. for 1 hour under vacuum) was bonded to the carrier bonding surface with the PTS surface modification layer. The article survived vacuum, SRD, and thermal processes up to 600° C. without permanent bonding of the glass carrier and the glass thin sheet.

Example 4d

Subsequently, the glass carrier with O2 plasma and SC1 treated bonding surface was treated with 1% diphenyldiethoxysilane (DPDS) in toluene and annealed at 200° C. under vacuum for 1 hour to complete condensation. The DPDS-treated surface showed a surface energy ^{of 47 mJ/m 2.} As shown in Table 4, a glass thin sheet (SC1 cleaned and then heated at 400° C. for 1 hour under vacuum) was bonded to a carrier bonding surface with a DPDS surface modification layer. The articles survived vacuum and SRD tests, as well as thermal processes up to 600° C. without permanent bonding of the glass carrier and the glass thin sheet.

Example 4e

Subsequently, the glass carrier with O2 plasma and SC1 treated bonding surface was treated with 1% 4-pentafluorophenyltriethoxysilane (PFPTS) in toluene and annealed at 200° C. under vacuum for 1 hour to complete condensation. . The PFPTS-treated surface showed a surface energy ^{of 57 mJ/m 2.} As shown in Table 4, a glass thin sheet (SC1 cleaned and then heated at 400° C. for 1 hour under vacuum) was bonded to the carrier bonding surface with the PFPTS surface modification layer. The articles survived vacuum and SRD tests, as well as thermal processes up to 600° C. without permanent bonding of the glass carrier and the glass thin sheet.

<Table 4>-Process suitability test of the silane surface modification layer

In Examples 4a-4e above, each carrier and thin sheet was Eagle XG® glass, wherein the carrier was a 150 mm diameter SMF wafer 630 micrometers thick, and the thin sheet 100 mm squared 100 micrometers thick. The silane layer was a self-assembled monolayer (SAM) and thus was approximately less than about 2 nm thick. In this example, the SAM was created using an organosilane having an aryl or alkyl nonpolar tail group and a mono, di, or tri-alkoxide head group. They react with the silanol surface on the glass to directly attach organic functional groups. The weaker interactions between the non-polar head groups make up the organic layer. Due to the small thickness in the surface modification layer, there is little risk of outgassing which can cause contamination in device fabrication. In addition, since the surface modification layer did not appear to degrade again in Examples 4c, 4d, and 4e, the risk of outgassing was even less. In addition, as shown in Table 4, each glass thin sheet was cleaned using an SC1 process before heat treatment at 400° C. for 1 hour under vacuum.

As can be seen from the comparison of Examples 4a-4e, controlling the surface energy of the bonding surface to be greater ^{than 40 mJ/m 2} to facilitate initial room temperature bonding will withstand the FPD processing and nevertheless thin sheets from the carrier without damage. It is not just the discussion of creating controlled associations that will allow them to be eliminated. In particular, as can be seen from Examples 4a-4e, each carrier ^{had a surface energy greater than 40 mJ/m 2} , which facilitated the initial room temperature bonding so that the article withstands vacuum and SRD processing. However, Examples 4a and 4b did not pass the 600°C processing test. As mentioned above, for certain applications, the bonds are processed below high temperatures (e.g., ≥ 400°C, ≥ 500°C, or ≥600°C, 650°C or below, as appropriate to the process in which the article is designed to be used). It is also important to hold the thin sheet and carrier together without decomposition of the bond, and to withstand to a point insufficient to control the covalent bonds occurring at such high temperatures so that there is no permanent bond between the thin sheet and the carrier.

The above-described separation in Examples 4, 3, and 2 is carried out at room temperature without the addition of any additional heat or chemical energy to modify the bonding interface between the thin sheet and the carrier. The only energy input is mechanical pulling and/or peeling force.

The materials described above in Examples 3 and 4 can be applied to a carrier, to a thin sheet, or to both the carrier and the thin sheet surface to be bonded together.

Uses of controlled coupling

Reusable carrier

One use of controlled bonding through a surface modification layer (including material and associated bonding surface heat treatment) is to provide reuse of carriers in articles that have undergone a process requiring a temperature of ≥ 600° C., for example in LTPS processing. will be. Surface modification layers (including material and bonding surface heat treatment), as illustrated by Examples 2e, 3a, 3b, 4c, 4d, and 4e above, can be used to provide reuse of carriers under these temperature conditions. In particular, these surface modification layers can be used to change the surface energy of the overlapping area between the bonding area of the thin sheet and the carrier, whereby the entire thin sheet can be separated from the carrier after processing. The thin sheets may be separated all at once, or separated by dividing, for example, first removing the device manufactured on the part of the thin sheet and then removing the remaining part to clean the carrier for reuse. You may. In the case of removing the entire thin sheet from the carrier, the carrier can be reused as it is by simply placing another thin sheet on it. Alternatively, it can be made to have a thin sheet again by cleaning the carrier and forming a new surface modification layer. Since the surface modification layer prevents permanent bonding of the carrier and the thin sheet, they can be used in processes with temperatures ≥ 600°C. Of course, these surface modification layers can control the bonding surface energy during processing at temperatures ≥ 600° C., but they can also be used to make thin sheet and carrier combinations that will withstand processing at lower temperatures, and such lower temperature It can be used to control coupling in applications. In addition, if the thermal processing of the article does not exceed 400° C., a surface modification layer as exemplified by Examples 2c, 2d, 4b can also be used in this same manner.

To provide a controlled bonding area

A second use of controlled bonding via a surface modification layer (including material and associated bonding surface heat treatment) is to provide a controlled bonding area between the glass carrier and the glass thin sheet. More particularly, the use of a surface modification layer allows sufficient separating force to separate the thin sheet portion from the carrier without damage to the thin sheet or carrier caused by bonding, but sufficient bonding force to hold the thin sheet against the carrier throughout the processing. It is possible to form an area of controlled bonding that is maintained throughout. Referring to FIG. 6, the glass thin sheet 20 can be bonded to the glass carrier 10 by means of a bonded region 40. In the bonded region 40, the carrier 10 and the thin sheet 20 are covalently bonded to each other to act as a unitary body. In addition, there is a controlled bonding area 50 with a periphery 52, in which the carrier 10 and the thin sheet 20 are connected, but after processing at a high temperature, for example at a temperature of ≥ 600° C. Can be separated from Although ten controlled bonding regions 50 are shown in FIG. 6, any suitable number may be provided, including one. A carrier 10 and a thin sheet using a surface modification layer 30, including material and bonding surface heat treatment, as illustrated by Examples 2a, 2e, 3a, 3b, 4c, 4d, and 4e above ( 20) it is possible to provide a controlled bonding area 50 between. In particular, these surface modification layers may be formed in the periphery 52 of the controlled bonding area 50 on the carrier 10 or thin sheet 20. Thus, when the article 2 is processed at high temperature to form a covalent bond in the bonding area 40 or during device processing, control between the carrier 10 and the thin sheet 20 within the area bonded to the periphery 52 A bonded bond can be provided, whereby the separating force can separate the thin sheet and carrier in the area (without catastrophic damage to the thin sheet or carrier), but the thin sheet and carrier will not delaminate during processing, including ultrasonic processing. will be. Thus, the controlled bonding herein provided by the surface modification layer and any associated heat treatment can improve on the carrier concept in US '727. In particular, carriers of US '727 with bonded periphery and unbonded center regions have been proven to withstand FPD processes, including hot temperature processes of ≥ about 600° C., while ultrasonic processes such as wet cleaning and resist strip processes are still It is difficult. In particular, since there is little or no adhesion to bond the thin glass and the carrier in the non-bonding region, the pressure wave in solution has been shown to induce resonance in the thin glass in the non-bonding region (non-bonding is US ' 727). Standing waves can form in thin glass, where these waves can generate vibrations that can lead to breakdown of the thin glass at the interface between the bonded and unbonded regions if ultrasonic agitation is of sufficient strength. This problem can be eliminated by minimizing the spacing between the thin glass and the carrier, and by providing sufficient adhesion in these regions 50, or a controlled bonding between the carrier 20 and the thin glass 10. The surface modification layer of the bonding surface (including the material as exemplified by Examples 2a, 2e, 3a, 3b, 4c, 4d and 4e and any associated heat treatment) controls the bonding energy to control these desired values in the controlled bonding area. Provides sufficient bonding between the thin layer 20 and the carrier 10 to avoid non-vibration.

Subsequently, while extracting the target portion 56 having the periphery 57, the portion of the thin sheet 20 in the periphery 52 is processed and after separating the thin sheet along the periphery 57, the carrier 10 Can be simply separated from Since the surface modification layer controls the bonding energy to prevent permanent bonding of the carrier and the thin sheet, they can be used in processes where the temperature is ≧600°C. Of course, while these surface modification layers can control the bonding surface energy during processing at temperatures ≥ 600° C., they can also be used to make thin sheet and carrier combinations that will withstand processing at lower temperatures, and such lower temperature Can be used in applications. In addition, if the thermal processing of the article does not exceed 400° C., the surface modification layer as illustrated by Examples 2c, 2d, 4b is also applied in this same way-in some cases, according to other process requirements-the bonding surface. Can be used to control energy.

To provide a bonding area

A third use of controlled bonding via a surface modification layer (including material and any associated bonding surface heat treatment) is to provide a bonding area between the glass carrier and the glass thin sheet. Referring to FIG. 6, the glass thin sheet 20 can be bonded to the glass carrier 10 by means of a bonded region 40.

In one embodiment of the third use, in the bonded region 40, the carrier 10 and the thin sheet 20 may be covalently bonded to each other to act as a unitary body. In addition, there is a controlled bonding area 50 with a periphery 52, where the carrier 10 and the thin sheet 20 are sufficient to withstand the processing, but high temperature processing, for example processing at a temperature of ≥ 600°C. They are joined together to allow separation of the thin sheets from the carrier even afterwards. Thus, using the surface modification layer 30 (including material and bonding surface heat treatment) as exemplified by Examples 1a, 1b, 1c, 2b, 2c, 2d, 4a, and 4b above, the carrier 10 and A bonding region 40 may be provided between the thin sheets 20. In particular, these surface modification layers and heat treatments can be formed outside the perimeter 52 of the controlled bonding area 50 on the carrier 10 or thin sheet 20. Thus, when the article 2 is processed at a high temperature, or is processed at a high temperature to form a covalent bond, the carrier and the thin sheet 20 are within the bonding area 40 outside the area bonded to the periphery 52. Will be combined with each other. Subsequently, while it is desirable to dicing the thin sheet 20 and the carrier 10 while extracting the target portion 56 having the periphery 57, these surface modification layers and heat treatments are applied to the thin sheet 20. The article can be separated along the line 5 as it is covalently bonded with the carrier 10 and it acts as a monolith in this area. Since the surface modification layer provides a permanent covalent bonding of the carrier and the thin sheet, they can be used in processes with temperatures ≥ 600°C. In addition, when the thermal processing of the article, or the initial formation of the bonding region 40, is ≥ 400°C or less than 600°C, the surface modification layer, as illustrated by the material and heat treatment in Example 4a, is also prepared in this same way. Can be used.

In a second embodiment of the third use, in the bonded region 40, the carrier 10 and the thin sheet 20 can be bonded to each other by controlled bonding via the various surface modification layers described above. In addition, there is a controlled bonding area 50, with a periphery 52, where the carrier 10 and the thin sheet 20 are sufficient to withstand the processing, but at a high temperature processing, for example at a temperature of ≥ 600°C. They are joined together to allow separation of the thin sheets from the carrier even after processing. Thus, if the processing will be carried out at a temperature of 600° C. or lower, and it is desired to have no permanent or covalent bonds in the region 40, as illustrated by Examples 2e, 3a, 3b, 4c, 4d, and 4e above. The same surface modification layer 30 (including material and bonding surface heat treatment) can be used to provide the bonding region 40 between the carrier 10 and the thin sheet 20. In particular, these surface modification layers and heat treatments may be formed outside the periphery 52 of the controlled bonding region 50 and may be formed on the carrier 10 or on the thin sheet 20. The controlled bonding region 50 may be formed of the same or different surface modification layer that was formed in the bonding region 40. Alternatively, if processing will only be carried out at temperatures below 400° C., and it is desired to have no permanent or covalent bonds in region 40, the above examples 2c, 2d, 2e, 3a, 3b, 4b, 4c, A surface modification layer 30 (including material and bonding surface heat treatment) as illustrated by 4d and 4e may be used to provide bonding region 40 between carrier 10 and thin sheet 20.

Instead of controlled bonding in region 50, there may be a non-binding region in region 50, wherein the non-binding region may be a region with increased surface roughness as described in US '727, or A surface modification layer as illustrated by Example 2a may be provided.

For bulk annealing or bulk processing

A fourth use of the above-described way of controlling bonding is for bulk annealing of stacks of glass sheets. Annealing is a thermal process to achieve compression of the glass. Compression involves reheating the glass body to a temperature below the glass softening point but above the maximum temperature leading to a subsequent processing step. This achieves structural rearrangement and dimensional relaxation in the glass before subsequent processing rather than during subsequent processing. Annealing before subsequent processing maintains the correct orientation and/or flatness in the glass body during subsequent processing, as in the manufacture of flat panel display devices in which many layered structures must be aligned with very tight tolerances, even after being applied in a high temperature environment. It is advantageous to do. When the glass is compressed in one high temperature process, the layers of the structure deposited on the glass before the hot process cannot accurately align with the layers of the structure deposited after the hot process.

Compressing glass sheets into a stack is economically attractive. However, this caused adjacent sheets to be sandwiched or separated to avoid tackiness. At the same time, it is advantageous to keep the sheet very flat and optically-characterized, or have a very clean surface finish. Also, for certain stacks of glass sheets, e.g. sheets with small surface areas, the glass sheets can be "sticky" together during the annealing process so that they can be easily moved without separating them as units, but after the annealing process (e.g. It may be advantageous to be easily separated from each other and to be able to use the sheets individually. Alternatively, a glass sheet, which prevents selected one of the glass sheets from permanently bonding to each other, while at the same time allowing the other of the glass sheet, or portions of these other glass sheets, for example their peripheries, to be permanently bonded to each other. It may be advantageous to anneal the stack of. As another alternative, it may be advantageous to laminate the glass sheets in bulk to selectively permanently bond the peripheries of selected adjacent pairs of sheets in the stack. The bulk annealing and/or selective bonding can be achieved using the above-described manner of controlling bonding between the glass sheets. In order to control bonding at any particular interface between adjacent sheets, a surface modification layer can be used on at least one of the major surfaces facing that interface.

One embodiment of a stack of glass sheets suitable for bulk annealing or bulk permanent bonding in a selected area (eg, around the periphery) will be described in connection with FIGS. 7 and 8. 7 is a schematic side view of a stack 760 of glass sheets 770-772, and FIG. 8 is an exploded view thereof for the purposes of further explanation.

The stack 760 of glass sheets may include glass sheets 770-772, and a surface modification layer 790 that controls bonding between the glass sheets 770-772. Further, the stack 760 may include cover sheets 780 and 781 disposed above and below the stack, and may include a surface modification layer 790 between the cover and adjacent glass sheets.

As shown in FIG. 8, each glass sheet 770-772 includes a first major surface 776 and a second major surface 778. The glass sheet can be made of any suitable glass material, for example alumino-silicate glass, boro-silicate glass, or alumino-boro-silicate glass. Further, the glass may be alkali-containing or may be alkali-free. Each of the glass sheets 770-772 may have the same composition, or the sheets may have a different composition. Also, the glass sheet can have any suitable kind. That is, for example, the glass sheets 770-772 may all be carriers as described above, all thin sheets as described above, or alternatively may be carriers and thin sheets. It is advantageous to have a stack of carriers, and a separate stack of thin sheets, if bulk annealing requires different time-temperature cycling for the carrier than for the thin sheet. Alternatively, using the right surface modification layer material and arrangement, it may be desirable to have a stack comprising alternately carriers and thin sheets, whereby the target pairs of carriers and thin sheets, i.e. those forming the article, are then It can be covalently bonded to each other in bulk for processing of, while at the same time preserving the ability to separate adjacent articles from each other. Additionally, there may be any suitable number of glass sheets in the stack. That is, although only three glass sheets 770-772 are shown in FIGS. 7 and 8, any suitable number of glass sheets may be included in the stack 760.

Any one glass sheet in any particular stack 760 may not include a surface modification layer, or may include one surface modification layer, or two surface modification layers. For example, as shown in FIG. 8, sheet 770 does not include a surface modification layer, and sheet 771 includes one surface modification layer 790 on its second major surface 778. And the sheet 772 includes two surface modification layers 790 with one surface modification layer on each of its major surfaces 776 and 778.

The cover sheets 780, 781 can be any material that will adequately withstand the time-temperature cycle for a given process (not only in terms of time and temperature, but also with respect to other suitable considerations, e. have. Advantageously, the cover sheet can be made of the same material as the glass sheet being processed. If cover sheets 780, 781 are present and have a material that will undesirably bond with the glass sheet when attaching the stack over a given time-temperature cycle, the surface modification layer 790 is applied with the glass sheet 771 Between the sheet 781 and/or between the glass sheet 772 and the cover sheet 780, it can be included suitably. When present between the cover and the glass sheet, the surface modification layer may be on the cover (as shown with the cover 781 and adjacent sheet 771), or may be on the glass sheet (cover ( 780) and sheet 772), or on both the cover and adjacent sheets (not shown). Alternatively, if cover sheets 780, 781 are present but have a material that will not bond with adjacent sheets 772, 772, the surface modification layer 790 need not be present between them.

There is an interface between adjacent sheets in the stack. For example, between adjacent ones of the glass sheets 770-772, an interface is defined, that is, an interface 791 exists between the sheet 770 and the sheet 771, and the sheet 770 and the sheet ( There is an interface 792 between 772). In addition, when the cover sheets 780 and 781 are present, not only the interface 793 exists between the cover 781 and the sheet 771, but also an interface 794 exists between the sheet 772 and the cover 780. do.

A surface modification layer 790 can be used to control bonding at a given interface 791, 792 between adjacent glass sheets, or at a given interface 793, 794 between a glass sheet and a cover sheet. For example, as shown, at each interface 791, 792, a surface modification layer 790 is present on at least one of the major surfaces facing the interface. For example, for the interface 791, the second major surface 778 of the glass sheet 771 includes a surface modification layer 790 to control the bonding between the sheet 771 and an adjacent sheet 770. . Although not shown, the first major surface 776 of the sheet 770 may also include a surface modification layer 790 thereon to control bonding with the sheet 771, i.e. facing any particular interface. There may be a surface modification layer on each major surface.

A specific surface modification layer 790 (and any associated surface modification treatment) at any given interface 791-794-e.g., a heat treatment for a specific surface, or surface modification prior to applying a specific surface modification layer to that surface. The surface heat treatment of the surfaces the layer can contact), the major surfaces (776, 778) facing that particular interface (791-794) control the bonding between adjacent sheets, and are applied to the stack 760 accordingly. Can be selected to achieve the desired result for a given time-temperature cycle.

If it is desirable to bulk anneal the stack of glass sheets 770-772 at a temperature of 400° C. or less and separate each glass sheet from each other after the annealing process, then at any particular interface, for example interface 791 The binding of can be controlled using a material according to any one of Examples 2a, 2c, 2d, 2e, 3a, 3b, or 4b-4e, with any associated surface preparation. More specifically, the first surface 776 of the sheet 770 will be treated as "thin glass" in Table 2-4, while the second surface 778 of the sheet 771 will be treated as "carrier" in Table 2-4. Will be treated as "and vice versa. Then a suitable time-temperature cycle, having a temperature of 400° C. or less, is selected based on the desired degree of compression, the number of sheets in the stack, as well as the size and thickness of the sheets, to achieve the required time-temperature throughout the stack. Could.

Similarly, if it is desirable to bulk anneal the stack of glass sheets 770-772 at a temperature of 600° C. or lower, and separate each glass sheet from each other after the annealing process, then any specific interface, e.g. 791) could be controlled using a material according to any one of Examples 2a, 2e, 3a, 3b, 4c, 4d, 4e, with any associated surface preparation. More specifically, the first surface 776 of the sheet 770 will be treated as "thin glass" in Table 2-4, while the second surface 778 of the sheet 771 will be treated as "carrier" in Table 2-4. Will be treated as "and vice versa. Then a suitable time-temperature cycle, having a temperature of 600° C. or less, is selected based on the desired degree of compression, the number of sheets in the stack, as well as the size and thickness of the sheets, to achieve the required time-temperature throughout the stack. Could.

In addition, bulk annealing and bulk article formation can be performed by appropriately configuring a stack of sheets and a surface modification layer between each pair of them. If it is desirable to bulk anneal the stack of glass sheets (770-772) at a temperature of 400° C. or less, and then to bulk covalently bond adjacent pairs of sheets to each other to form article 2, a suitable material to control bonding And associated surface preparation. For example, around the periphery (or in another preferred bonding area 40), the bonding at the interface between a pair of glass sheets that will be formed into the article 2, for example sheets 770 and 771, is (i ) The material according to any one of Examples 2c, 2d, 4b around the periphery of the sheets 770, 771 (or in other preferred bonding areas 40) with any associated surface treatment; And (ii) the material according to any one of Examples 2a, 2e, 3a, 3b, 4c, 4d, 4e on the inner region of the sheets 770, 771 (i.e., the inside of the periphery as treated in (i)). The area, or separation of one sheet from another, can be controlled by use with any associated surface treatment (in the desired controlled bonding area 50) desired. In this case, subsequent processing of the device in the controlled bonding region 50 could be performed at a temperature of 600° C. or less.

Materials and heat treatments could be appropriately selected for mutual compatibility. For example, any material (2c, 2d, or 4b) could be used for the bonding area 40 with the material according to Example 2a for a controlled bonding area. Alternatively, the heat treatment for the bonding area and the controlled bonding area could be properly controlled to minimize the effect of heat treatment in one area adversely affecting the desired degree of bonding in adjacent areas.

After appropriate selection of the surface modification layer 790 and the associated heat treatment for the glass sheets in the stack, the sheets are properly arranged into a stack and then heated to 400° C. or lower to bulk anneal all sheets into the stack without permanently bonding them to each other. Could. The stack could then be heated to 600° C. or lower to form covalent bonds of a pair of adjacent sheets in the desired bonding area to form an article 2 having a pattern of bonding areas and controlled bonding areas. The bonding at the interface between a pair of sheets that must be covalently bonded by the bonding region 40 to form an article 2, and another pair of such sheets forming a separate, but adjacent article 2, is an example The material of 2a, 2e, 3a, 3b, 4c, 4d, 4e and the associated heat treatment can be controlled so that adjacent articles 2 will not be covalently bonded to each other. In this same way of controlling the bonding between adjacent articles, the bonding between the articles present in the stack and any cover sheets could be controlled.

Additionally, similar to the above, the article 2 from the stack 760 can be formed in bulk without prior annealing of the same stack 760. Instead, the sheets could be annealed individually, or annealed to another stack and separated therefrom, prior to making the article bulk by configuring the sheets into a stack for the desired controlled bonding. In the immediately-described manner of forming articles into bulk from one and the same stack after bulk annealing, bulk annealing is simply omitted.

The only way to control bonding at interface 791 has been detailed above, but of course the same is at interface 792, or for any other interface that may be present in a particular stack-of more than three glass sheets in the stack. This can be done as in the case, or as in the case where there is a cover sheet to be undesirably bonded to the glass sheet. In addition, the same way of controlling bonding can be used at any interface (791, 792, 793, 794) present, but other of the above-described ways of controlling bonding are the same or different in terms of the type of bonding desired. It can also be used at other interfaces to produce results.

In the above process of bulk annealing or forming the article 2 into bulk, HMDS is used as a material to control bonding at the interface, and when HMDS is exposed to the outer periphery of the stack, covalent bonding is prevented in the area of HMDS. If it is desirable to do so, heating above about 400° C. should be carried out in an oxygen-free atmosphere. That is, if HMDS is exposed to an amount of oxygen in the atmosphere sufficient to oxidize HMDS (at temperatures above about 400° C.), bonding in any such area where HMDS has been oxidized will result in covalent bonds between adjacent glass sheets. will be. Other alkyl hydrocarbon silanes such as ethyl, propyl, butyl, or steryl silanes can likewise be affected by exposure to oxygen at higher temperatures, such as above about 400°C. Similarly, when using other materials for the surface modification layer, the environment for bulk annealing should be chosen so that the material will not decompose during the time-temperature cycle of the annealing. As used herein, oxygen free can mean an oxygen concentration of less than 1000 ppm by volume, more preferably less than 100 ppm by volume.

Once the stack of sheets is bulk annealed, each sheet can be separated from the stack. Each sheet may be treated to remove the surface modification layer 790 (eg, by an oxygen plasma heating in an oxygen environment at a temperature> 400° C., or by chemical oxidation, SC1, or SC2). Each sheet can be used, depending on the purpose, for example as an electronic device, for example an OLED, FPD, or a PV device substrate.

The above-described method of bulk annealing, or bulk processing, has the advantage of maintaining a clean sheet surface in an economical way. More particularly, the sheet does not have to be maintained in a clean environment from start to finish as in clean-room annealing rares. Instead, the stack can be formed in a clean environment where the sheet surface is not contaminated with particles because there is no fluid flow between the sheets, and then processed in standard annealing rares (i.e., where cleanliness is not controlled). Thus, the sheet surface is protected from the environment in which the stack of sheets is annealed. After annealing, the stack of sheets can be easily transported to additional processing areas (at the same or different facilities) as the sheets retain some degree of adhesion, but remain separable from each other without damaging the sheets when sufficient force is applied. have. That is, a glass manufacturer can assemble and anneal a stack of glass sheets (for example), and then transport the sheets as a stack, the sheets sticking together during shipping (without fear of their separation in transit) and their destination. Upon arrival, the sheets can be separated from the stack by a consumer who can use the sheets individually or in smaller groups. Once separation is desired, the stack of sheets can be processed again in a clean environment (after washing the stack as needed).

Examples of bulk annealing

The glass substrate was used as it was obtained from the fusion drawing process. The fusion-drawn glass composition was (in mole %): SiO2 (67.7), Al2O3 (11.0), B2O3 (9.8), CaO (8.7), MgO (2.3), and SrO (0.5). Seven (7), 0.7 mm thick x 150 mm diameter, fusion drawn glass substrates were patterned by a lithographic method with a 200 nm depth fiducial/scale using HF. As a surface modification layer, this (2) nm plasma deposited fluoropolymer was coated on all bonding surfaces of all glass substrates, i.e. each surface of the substrate facing another substrate was coated, resulting in the creation of each sheet surface. The resulting surface energy was approximately 35m J/m ² . Each of the seven coated glass substrates was put together to form a single thick substrate (referred to as a “glass stack”). The glass stack was ramped from 30° C. to 590° C. over a 15 minute period in a nitrogen purged tubular furnace, held at 590° C. for 30 minutes, then lowered to about 230° C. over a 50 minute period, and then the glass stack was furnace And annealed by cooling to room temperature of about 30° C. in about 10 minutes. After cooling, the substrate was removed from the furnace and easily separated into each sheet using a razor wedge (i.e., the sample was not permanently bonded, either entirely or locally). Compression was measured on each individual substrate by comparing the glass fiducial to an unannealed quartz reference. Each substrate was found to compress about 185 ppm. Two of the substrates (not laminated together) as each sample were subjected to a second annealing cycle as described above (hold at 590° C./30 min). Compression was measured again and the substrate was found to have further compressed less than 10 ppm (actually 0 to 2.5 ppm) due to the secondary heat treatment (change in glass dimensions after secondary heat treatment-compared to the original glass dimensions- Subtracting the change in glass dimensions after the primary heat treatment). Therefore, the inventors of the present invention can achieve compression, cooling, separation into individual sheets by coating individual glass sheets, laminating and heat treatment at high temperature, and dimensions of <10 ppm, even <5 ppm after the second heat treatment. It has been demonstrated that it can have a change (compared to their size change after the first heat treatment).

Although the furnace was purged with nitrogen in the above-described annealing examples, the annealing furnace may also be air, argon, oxygen, CO ₂ , or a combination thereof, depending on the annealing temperature and the stability of the surface modification layer material at that temperature in a particular environment. It can be purged with other gases including.

Also, although not shown, the glass can be annealed in the form of a spool instead of a sheet. That is, a suitable surface modification layer can be formed on one or both sides of the glass ribbon, and the ribbon can then be rolled. The entire roll may be subjected to the same treatment as mentioned above for the sheet, so that the entire spool of glass will be annealed without sticking one wrap of glass to the adjacent one. Upon unfolding, the surface modification layer can be removed by any suitable process.

Outgassing

Polymeric adhesives used in typical wafer bonding applications are generally 10-100 microns thick and lose about 5% of their mass at or near their temperature limits. For these materials generated from thick polymer films, it is easy to quantify the amount of mass loss, or outgassing, by mass-spectrometry. On the other hand, it is more difficult to measure outgassing from thin surface treatments of about 10 nm or less thickness, for example from the plasma polymer or self-assembled monolayer surface modified layers described above, as well as by thin layers of pyrolyzed silicone oil. For these materials, mass-spectrometry is not sufficiently sensitive. However, there are many other methods of measuring outgassing.

The first method of measuring small outgassing is based on the measurement of surface energy, and will be described in connection with FIG. To perform this test, a setup as shown in Fig. 9 can be used. A first substrate, or carrier 900, having a surface modification layer to be tested thereon provides a surface 902, ie the surface modification layer corresponds in composition and thickness to the surface modification layer 30 to be tested. The second substrate, or cover 910, is placed such that its surface 912 is adjacent to, but not in contact with, the surface 902 of the carrier 900. Surface 912 is an uncoated surface, that is, the surface of the bare material from which the cover is made. The spacers 920 are placed at various points between the carrier 900 and the cover 910 and keep them in a spaced relationship. Spacer 920 is thick enough to separate cover 910 from carrier 900 to allow material transfer from one to another, but minimizes the amount of contamination from the chamber atmosphere on surfaces 902 and 912 during testing. It should be thin enough to do. The carrier 900, the spacer 920, and the cover 910 together form a test article 901.

Before assembling the test article 901, the surface energy of the bare surface 912 is measured, as is the surface energy of the surface 902, that is, the surface of the carrier 900 having a surface modification layer provided thereon. Both of the surface energy, polarity and dispersion components as shown in Fig. 10 are S. The theoretical model developed by S. Wu (1971) was used as three test liquids; It was measured by fitting to the three contact angles of water, diiodomethane and hexadecane. (See: S. Wu, J. Polym. Sci. C, 34, 19, 1971).

After assembly, test article 901 is placed in heating chamber 930 and heated throughout the time-temperature cycle. The heating is carried out at atmospheric pressure and under flowing N2 gas, i.e., N2 gas flowing in the direction of arrow 940 at a rate of 2 standard liters/minute.

During the heating cycle, changes in the surface 902 (including, for example, changes to the surface modification layer due to evaporation, pyrolysis, decomposition, polymerization, reaction with carriers, and dewetting) are the surface energy of the surface 902 Manifested by a change in The change in the surface energy of the surface 902 by itself does not necessarily mean that the surface modification layer has been outgassed, but since its properties are changing, for example due to the above-mentioned mechanisms, it indicates the general instability of the material at that temperature. . Thus, the less the change in the surface energy of the surface 902, the more stable the surface modification layer. On the other hand, because of the proximity of the surface 912 to the surface 902, any material outgassed from the surface 902 will collect on the surface 912 and change the surface energy of the surface 912. will be. Thus, the change in the surface energy of the surface 912 becomes a proxy for outgassing of the surface modification layer present on the surface 902.

Thus, one test for outgassing uses a change in the surface energy of the cover surface 912. Specifically, if there is a change in the surface energy-of the surface 912 of ≥ 10 mJ/m ^{2, then there is outgassing.} Changes in surface energy on this scale are consistent with contamination that can lead to loss or degradation of film adhesion in material properties and device performance. The change in surface energy of ≤ 5 mJ/m ² is close to the repeatability of the surface energy measurement and the non-uniformity of the surface energy. This small change is consistent with minimal outgassing.

During the tests that resulted in the results of FIG. 10, the carrier 900, cover 910, and spacer 920 were Eagle XG glass, alkali-free alumino-available from Corning, Inc., Corning, NY. Although made of boro-silicate display-grade glass, that need not be the case. The carrier 900 and the cover 910 were 150 mm in diameter and 0.63 mm thick. In general, carrier 910 and cover 920 will each be made of the same material as carrier 10 and thin sheet 20, for which outgassing testing is desirable. During this test, the silicone spacer was 0.63 mm thick, 2 mm wide, and 8 cm long, thereby forming a gap of 0.63 mm between surfaces 902 and 912. During the test, the chamber 930 is cycled from room temperature to the test limit temperature at a rate of 9.2° C./min, maintained at the test limit temperature for various times as indicated by the “annealing time” in the graph, and then 200 at the furnace speed. Included in MPT-RTP600s rapid thermal processing equipment cooled to °C. After cooling the oven to 200° C., the test article was removed, and after cooling the test article to room temperature, the surface energy of each surface 902 and 912 was measured again. Thus, for example, for material #1, using data on the change in cover surface energy, line 1003, tested at a limit temperature of 450°C, the data was collected as follows. The data point at 0 minutes ^{represents a surface energy of 75 mJ/m 2} (milli-joules per square meter), and is the surface energy of the bare glass, i. The data points at 1 minute represent the surface energy as measured after a time-temperature cycle performed as follows: (with material #1 used as a surface modification layer on carrier 900 to provide surface 902) ) Put the article 901 into the heating chamber 930 at room temperature and atmospheric pressure; The chamber was heated to a test-limit temperature of 450° C. under 2 standard liters/min of N2 gas flow, even at a rate of 9.2° C./min, and held at a test-limit temperature of 450° C. for 1 minute; The chamber is then allowed to cool to 300° C. at a rate of 1° C./min, and the article 901 is then removed from the chamber 930; The article is then allowed to cool to room temperature (without a flowing atmosphere of N2); The surface energy of the surface 912 was then measured and plotted on line 1003 as a point over 1 minute. Then the remaining data points for material #1 (line(1003, 1004)), as well as material #2 (line(1203, 1204)), material #3 (line(1303, 1304)), material #4 (line (1403, 1404)), Material #5 (Lines 1503, 1504), and Material #6 (Lines 1603, and 1604)), appropriately in minutes of annealing time, test-limit temperature, 450°C. , Or in a similar manner corresponding to the holding time at 600°C. Lines 1001, 1002, 1201, 1202, 1301, 1302, 1401, 1402, 1501, 1502, 1601, representing the surface energy of the surface 902 for the corresponding surface modification layer material (Material #1-6) And 1602) were determined in a similar manner, except that the surface energy of the surface 902 was measured after each time-temperature cycle.

The assembly process, and time-temperature cycle, were performed for six different materials as described below, and the results are plotted in FIG. 10. Of the six materials, materials #1-4 correspond to the surface modification layer materials described above. Materials #5 and #6 are comparative examples.

Material #1 is a CHF3-CF4 plasma polymerized fluoropolymer. This material is consistent with the surface modification layer in Example 3b above. As shown in Fig. 10, lines 1001 and 1002 show that the surface energy of the carrier has not changed significantly. Thus, the material is very stable at temperatures of 450°C to 600°C. Further, as shown by lines 1003 and 1004, the surface energy of the cover also did not change significantly, that is, the change was ≤ 5 mJ/m ² . Thus, there was no outgassing associated with the material between 450°C and 600°C.

Material #2 is phenylsilane, a self-assembled monolayer (SAM) deposited from a 1% toluene solution of phenyltriethoxysilane and cured at 190° C. for 30 minutes in a vacuum oven. This material is consistent with the surface modification layer of Example 4c above. As shown in Fig. 10, lines 1201 and 1202 represent slight changes in the surface energy of the carrier. As mentioned above, this shows a slight change in the surface modification layer, and in comparison, material #2 is somewhat less stable than material #1. However, as noted by lines 1203 and 1204, the change in the surface energy of the carrier is ≦5 mJ/m ² , showing that the change to the surface modification layer did not cause outgassing.

Material #3 is pentafluorophenylsilane, a SAM deposited from a 1% toluene solution of pentafluorophenyltriethoxysilane and cured at 190° C. for 30 minutes in a vacuum oven. This material is consistent with the surface modification layer of Example 4e above. As shown in Fig. 10, lines 1301 and 1302 represent slight changes in the surface energy of the carrier. As mentioned above, this shows a slight change in the surface modification layer, and in comparison, material #3 is somewhat less stable than material #1. However, as noted by lines 1303 and 1304, the change in the surface energy of the carrier is ≦5 mJ/m ² , showing that the change to the surface modification layer did not cause outgassing.

Material #4 is hexamethyldisilazane (HMDS) deposited from steam at 140° C. in a YES HMDS oven. The material is consistent with the surface modification layer of Example 2b in Table 2 above. As shown in Fig. 10, lines 1401 and 1402 represent slight changes in the surface energy of the carrier. As mentioned above, this shows a slight change in the surface modification layer, and in comparison, material #4 is somewhat less stable than material #1. In addition, the change in the surface energy of the carrier for Material #4 is greater than that of any of Materials #2 and #3, indicating, in comparison, that Material #4 is somewhat less stable than Materials #2 and #3. However, as noted by lines 1403 and 1404, the change in the surface energy of the carrier is ≤ 5 mJ/m ^2, and the change to the surface modification layer does not cause outgassing which had an effect on the surface energy of the cover. Shows that it didn't. However, this is consistent with the way HMDS outgasses. In other words, HMDS releases ammonia and water that do not affect the surface energy of the cover and cannot affect some electronic device manufacturing equipment and/or processing. On the other hand, if the product of outgassing is trapped between the thin sheet and the carrier, there may be other problems, as mentioned below with respect to the secondary outgassing test.

Material #5 is glycidoxypropylsilane, a SAM deposited from a 1% toluene solution of glycidoxypropyltriethoxysilane and cured at 190° C. for 30 minutes in a vacuum oven. This is a comparative example material. As indicated by lines 1501 and 1502, there is a relatively small change in the surface energy of the carrier, but there is a significant change in the surface energy of the cover as indicated by lines 1503 and 1504. See FIG. 10. That is, although material #5 was relatively stable on the carrier surface, a significant amount of material was actually outgassed on the cover surface, and the cover surface energy was changed ^{by ≥ 10 mJ/m 2.} At the end of 10 minutes at 600° C., the surface energy is ^{within 10 mJ/m 2} , but the change during that time exceeds ^{10 mJ/m 2.} See, for example, data points at 1 and 5 minutes. While not wishing to be bound by theory, a slight increase in surface energy from 5 to 10 minutes has the potential to cause some of the outgassed material to decompose and fall off the cover surface.

Material #6 is prepared by dispensing 5 ml Dow Corning 704 diffusion pump oil tetramethyltetraphenyl trisiloxane (available from Dow Corning) onto a carrier and placing it on a 500° C. hot plate in air for 8 minutes. It is a silicone coated DC704. The completion of sample preparation is noted by the end of the visible smoke. After preparing the sample by the above method, the outgassing test described above was performed. This is a comparative example material. As shown in Fig. 10, lines 1601 and 1602 represent slight changes in the surface energy of the carrier. As mentioned above, this shows a slight change in the surface modification layer, and in comparison, material #6 is less stable than material #1. Also, as noted by lines 1603 and 1604, the change in the surface energy of the carrier is ≧10 mJ/m ² , indicating significant outgassing. More particularly, at a test-limit temperature of 450°C, the data points for 10 minutes show ^{a decrease in surface energy of about 15 mJ/m 2} , and for the data points at 1 and 5 minutes, the surface energy is much larger. Indicates a decrease. Similarly, the decrease in the surface energy of the cover, which is the change in the surface energy of the cover during cycling at 600° C. test-limit temperature, was about 25 mJ/m ² at the 10 min data point, slightly greater at 5 min, 1 It was somewhat less in minutes. Overall, there was a significant amount of outgassing for this material over the entire range of the test.

Significantly, for material #1-4, the surface energy over the time-temperature cycle remains at the surface energy where the cover surface matches the surface energy of the bare glass, i.e. no collected material outgassed from the carrier surface. Shows. For Material #4, as mentioned in connection with Table 2, the manner in which the carrier and thin sheet surfaces are made depends on whether the article (thin sheet bonded together with the carrier via a surface modification layer) will withstand FPD processing. There is a big difference. Thus, although the example of material #4 shown in FIG. 10 is not outgassing, the material may or may not withstand 400° C. or 600° C. tests as mentioned in connection with the discussion of Table 2.

A second way of measuring small outgassing is based on the assembled article, i.e. a thin sheet is bonded to the carrier through a surface modification layer, and the change in percent bubble area is used to determine outgassing. That is, while heating the article, bubbles formed between the carrier and the thin sheet indicate outgassing of the surface modification layer. As mentioned above in connection with the primary outgassing test, it is difficult to measure outgassing of very thin surface modification layers. In this secondary test, outgassing under the thin sheet can be limited by the strong adhesion between the thin sheet and the carrier. Nevertheless, layers <10 nm thick (e.g. plasma polymerized material, SAM, and pyrolyzed silicone oil surface treatment) can continue to bubble during heat treatment, despite their less absolute mass loss. have. And the creation of bubbles between the thin sheet and the carrier can cause problems with pattern generation, photolithographic processing, and/or orientation during device processing on the thin sheet. In addition, bubbling at the boundary of the bonded area between the thin sheet and the carrier can cause problems with the process fluid from one process contaminating the downstream process. A change in the% bubble area of ≥ 5 indicating outgassing is significant, and is not preferable. On the other hand, the change in% bubble area of ≤ 1 is not significant and is an indication that there was no outgassing.

In a class 1000 clean room, the average bubble area of thin glass bonded by passive bonding is 1%. The% bubble in the bonded carrier is a function of the cleanliness of the carrier, thin glass sheet, and surface preparation. Since these initial defects act as nucleation sites for bubble growth after heat treatment, any change in bubble area of less than 1% upon heat treatment is within the variability of sample preparation. To perform this test, a commercially available desktop scanner (Epson Expression 10000XL photo) with a perspective view unit was used to create a first scan image of the area in which the thin sheet and carrier were bonded immediately after bonding. Portions were scanned using standard Epson software and using 508 dpi (50 microns/pixel) and 24 bit RGB. The image processing software first, if necessary, prepares the image by stitching the images of different sections of the sample into a single image and removing the scanner artifact (using a calibration calibration reference scan performed without the sample in the scanner). The bound regions are then analyzed using standard imaging techniques such as bordering, hole filling, erosion/expansion, and blob analysis. The newer Epson Expression 11000XL photo can also be used in a similar manner. In the transmission mode, bubbles in the combined region are visible in the scanned image, and a value for the bubble region may be determined. The bubble area is then compared to the total bonding area (i.e., the total overlapping area between the thin sheet and the carrier) to calculate the% area of the bubbles in the bonding area relative to the total bonding area. The samples are then heat treated for up to 10 minutes at test-limit temperatures of 300° C., 450° C., and 600° C. in an N2 atmosphere in an MPT-RTP600s rapid thermal processing system. In particular, the time-temperature cycle carried out includes: inserting the article into the heating chamber at room temperature and atmospheric pressure; The chamber is then heated to the test-limit temperature at a rate of 9° C./min; Holding the chamber at the test-limit temperature for 10 minutes; The chamber is then cooled to 200° C. at a furnace speed; The article is removed from the chamber and allowed to cool to room temperature; It then included rescanning the article with an optical scanner. The% bubble area from the second scan was then calculated as above and compared to the% bubble area from the first scan to determine the change in% bubble area (Δ% bubble area). As mentioned above, a change in bubble area of ≥ 5% is significant and is an indication of outgassing. The change in% bubble area was originally chosen as a measure because of the variability in% bubble area. That is, most of the surface modification layers have a bubble area of about 2% in the first scan after making thin sheets and carriers and before bonding them due to handling and cleanliness. However, variations can occur between substances. The same substances #1-6 described in relation to the primary outgassing test method were used again in the secondary outgassing test method. Of these materials, Materials #1-4 showed about 2% bubble area in the first scan, while Materials #5 and #6 showed significantly larger bubble areas, ie, about 4% in the first scan.

The results of the secondary outgassing test will be described with reference to FIGS. 11 and 12. The results of the outgassing test for Substance #1-3 are shown in Fig. 11, while the results of the outgassing test for Substance #4-6 are shown in Fig. 12.

The results for Material #1 are shown as square data points in FIG. 11. As can be seen from the figure, the change in% bubble area was close to zero for the test-limit temperatures of 300°C, 450°C, and 600°C. Thus, Material #1 does not exhibit outgassing at these temperatures.

The results for material #2 are shown as rhombus data points in FIG. 11. As can be seen from the figure, the change in% bubble area is less than 1 for the test-limit temperatures of 450°C and 600°C. Thus, Material #2 does not exhibit outgassing at these temperatures.

Results for Material #3 are shown as triangular data points in FIG. 11. As can be seen from the figure, similar to the results for Material #1, the change in% bubble area was close to zero for the test-limit temperatures of 300°C, 450°C, and 600°C. Thus, Material #1 does not exhibit outgassing at these temperatures.

Results for Material #4 are shown as circular data points in FIG. 12. As can be seen from the figure, the change in% bubble area is close to zero for the test-limit temperature of 300°C, but closes to 1% for the test-limit temperature of 450°C and 600°C for some samples, For other samples of the same material it is about 5% at test-limit temperatures of 450°C and 600°C. The results for Material #4 are very inconsistent and depend on how the thin sheet and carrier surfaces are made to bond with the HMDS material. The manner in which the sample performance depends on the manner in which the sample is prepared is consistent with the examples of these materials described in connection with Table 2 above, and the related discussion. Note that for these materials, samples with a change in% bubble area close to 1% for the 450° C. and 600° C. test-limit temperatures did not allow separation of the thin sheet from the carrier according to the separation test described above. I did. That is, the strong adhesion between the thin sheet and the carrier may have limited bubble generation. On the other hand, samples with a change in% bubble area close to 5% allowed separation of the thin sheet from the carrier. Thus, samples that did not have outgassing had an undesirable result of increased adhesion (preventing removal of the thin sheet from the carrier) after temperature treatment to stick the carrier and the thin sheet together, while allowing the removal of the thin sheet and the carrier. The sample had an undesirable result of outgassing.

Results for Material #5 are shown as triangular data points in FIG. 12. As can be seen from the figure, the change in% bubble area is about 15% for the test-limit temperature of 300°C, and significantly exceeds that value for the higher test-limit temperature of 450°C and 600°C. Thus, Material #5 exhibits significant outgassing at these temperatures.

Results for material #6 are shown as square data points in FIG. 12. As can be seen from this figure, the change in% bubble area is more than 2.5% for the test-limit temperature of 300°C and more than 5% for the test-limit temperature of 450°C and 600°C. Thus, Material #6 exhibits significant outgassing at test-limit temperatures of 450°C and 600°C.

conclusion

It should be emphasized that the above-described embodiments of the invention, in particular any "preferred" embodiments, are only possible examples of the described implementation for a clear understanding of the various principles of the invention. Many changes and modifications can be made to the above-described embodiments of the invention without substantially departing from the spirit and various principles of the invention. All such changes and modifications are intended to be included within the scope of this disclosure and the invention and to be covered by the following claims.

For example, the surface modification layer 30 of many embodiments has been shown and discussed as being formed on the carrier 10, but instead, or in addition, may be formed on the thin sheet 20. That is, the material as described in Examples 4 and 3 can be applied to the carrier 10 on the sides to be bonded together, to the thin sheet 20, or to both the carrier 10 and the thin sheet 20.

In addition, although some surface modification layer 30 has been described as controlling the bonding strength so that the thin sheet 20 can be removed from the carrier 10 even after processing the article 2 at a temperature of 400°C, or 600°C. , Of course, the article 2 can be processed at a temperature lower than the temperature of the specific test the article has passed, and still remove the thin sheet 20 from the carrier 10 without damaging the thin sheet 20 or You can achieve the same ability to eliminate.

Additionally, although the concept of controlled bonding has been described herein as being able to be used with carriers and thin sheets, in certain circumstances they may be applied to control bonding between thicker sheets of glass, ceramic, or glass ceramic, and sheets ( Or a portion thereof) from each other may be desirable.

Additionally, although the controlled bonding concept has been described herein as useful with a glass carrier and a glass thin sheet, the carrier can be made of other materials, such as ceramic, glass ceramic, or metal. Similarly, the sheet controllably bonded to the carrier can be made of another material, for example ceramic or glass ceramic.

According to the first aspect, the following:

Laminating a plurality of glass layers;

There is provided a method of annealing glass comprising exposing a stack of glass layers to a time-temperature cycle sufficient for each glass layer to be compressed,

Each of the glass layers has two major surfaces, thus defining an interface between adjacent ones of the glass layers in a plurality of glass layers, and a surface modification layer on at least one of the major surfaces facing one of the interfaces. Is being placed,

The surface modification layer is sufficient to control the bonding between adjacent ones of the glass layers defining one of the interfaces in the stack during a time-temperature cycle, and the bonding is fixed in one layer and gravity is applied to the other layer. Is controlled to have the force to allow the layers to separate without breaking one of the adjacent ones of the glass layer into two or more pieces.

According to a second aspect, the method of aspect 1 is provided, wherein the time-temperature cycle comprises a temperature of ≧400° C. but below the strain point of the glass sheet.

According to a third aspect, the method of aspect 1 is provided, wherein the time-temperature cycle comprises a temperature of ≧600° C. but below the strain point of the glass sheet.

According to a fourth aspect, the method of any one of aspects 1 to 3 is provided, wherein the surface modification layer is one of HMDS, plasma polymerized fluoropolymer and aromatic silane.

According to the fifth aspect, when the surface modification layer comprises a plasma polymerized fluoropolymer, the surface modification layer is plasma polymerized polytetrafluoroethylene; And a plasma polymerized fluoropolymer surface modification layer deposited from a CF4-C4F8 mixture having <40% C4F8.

According to a sixth aspect, the method of aspect 4 is provided, wherein the surface modification layer is phenyl silane when the surface modification layer comprises an aromatic silane.

According to a seventh aspect, when the surface modification layer comprises an aromatic silane, the surface modification layer is phenyltriethoxysilane; Diphenyldiethoxysilane; And 4-pentafluorophenyltriethoxysilane.

According to an eighth aspect, there is provided a method of any one of aspects 1-7, wherein the time-temperature cycle is performed in an oxygen-free environment.

According to a ninth aspect, the method of any one of aspects 1-8 is provided, wherein the stack of glass layers comprises a rolled glass sheet.

Claims (9)

Laminating a plurality of glass layers;
A method of annealing glass comprising exposing a stack of glass layers to a time-temperature cycle sufficient for each glass layer to be compressed,
The time-temperature cycle includes a temperature ≥ 400° C. but below the strain point of the glass,
Each of the glass layers has two major surfaces, thus defining an interface between adjacent ones of the glass layers in a plurality of glass layers, and a surface modification layer on at least one of the major surfaces facing one of the interfaces. Is being placed,
The surface modification layer is sufficient to control the bonding between adjacent ones of the glass layers defining one of the interfaces in the stack during the time-temperature cycle, the bonding being fixed in one layer and not fixed in the glass layer. Directing the stack of unfixed layers down prevents one layer from separating from another, but the force that allows the layers to separate without destroying one of the adjacent ones of the glass layer into two or more pieces. Is controlled to have,
The surface modification layer is one of HMDS, plasma polymerized fluoropolymer and aromatic silane. delete The method of claim 1, wherein the time-temperature cycle comprises a temperature> 600° C. but below the strain point of the glass. delete The method of claim 1, wherein when the surface modification layer comprises a plasma polymerized fluoropolymer, the surface modification layer comprises plasma polymerized polytetrafluoroethylene; And a plasma polymerized fluoropolymer surface modification layer deposited from a CF4-C4F8 mixture having a ratio of the amount of C4F8 in the CF4-C4F8 mixture <0.40. The method of claim 1, wherein when the surface modification layer comprises an aromatic silane, the surface modification layer is a phenyl silane. The method of claim 1, wherein when the surface modification layer comprises an aromatic silane, the surface modification layer comprises phenyltriethoxysilane; Diphenyldiethoxysilane; And 4-pentafluorophenyltriethoxysilane. The method of claim 1, wherein the time-temperature cycle is carried out in an oxygen-free environment. The method of claim 1, wherein the stack of glass layers comprises a rolled glass sheet.

Images