Interfacial bonding oxides for glass-ceramic-to-metal seals

Abstract

The present invention relates to structure including an interfacial seal between a glass-ceramic component and a metal component, as well as methods for forming such structures. In particular embodiments, the interfacial seal includes a metal oxide. Such interfacial seals can be beneficial for, e.g., hermetic seals between a glass-ceramic component and a metal component.

Claims

1. A method comprising: providing a glass-ceramic component configured to provide a glass-ceramic composition comprising of from about 65 wt. % to about 80 wt. % of SiO.sub.2; from about 8 wt. % to about 16 wt. % of Li.sub.2O; from about 2 wt. % to about 8 wt. % of Al.sub.2O.sub.3; from about 1 wt. % to about 8 wt. % of K.sub.2O; from about 1 wt. % to about 5 wt. % of P.sub.2O.sub.5; from about 0.5 wt. % to about 7 wt. % of B.sub.2O.sub.3; from 0 wt. % to about 5 wt. % of ZnO; and from about 0.5 wt. % to about 5 wt. % of one or more metal oxidants; providing a metal component, wherein the glass-ceramic component is contacted with a portion of the metal component; heating the glass-ceramic component and metal component to a first temperature T.sub.1 of from about 900 C. to about 1050 C.; initially cooling the glass-ceramic component and metal component to a second temperature T.sub.2 of from about 400 C. to about 750 C.; and further cooling the glass-ceramic component and metal component to a fourth temperature T.sub.4 of from about 10 C. to about 500 C., thereby forming an interfacial seal between the glass-ceramic component and the metal component; and wherein a plurality of reduced copper particles are within the glass-ceramic component and next to the interfacial seal.

2. The method of claim 1, wherein the heating step comprises rapidly cooling at a rate r.sub.2 greater than about 30 C./minute to a second temperature T.sub.2 of from about 400 C. to about 750 C., thereby minimizing formation of a cristobalite SiO.sub.2 phase within the glass-ceramic composition.

3. The method of claim 1, further comprising, after the initially cooling step: reheating the glass-ceramic component and metal component to a third temperature T.sub.3 of from about 750 C. to about 850 C., thereby facilitating formation of a quartz SiO.sub.2 phase within the glass-ceramic composition.

4. The method of claim 1, wherein the glass-ceramic composition comprises of from about 5 wt. % to about 30 wt. % of SiO.sub.2 as a cristobalite phase and from about 5 wt. % to about 30 wt. % of SiO.sub.2 as a quartz phase.

5. The method of claim 1, wherein the one or more metal oxidants is selected from the group consisting of cobalt oxide, nickel oxide, lead oxide, copper oxide, silver oxide, iron oxide, molybdenum oxide, tungsten oxide, and tin oxide.

6. The method of claim 1, wherein the glass-ceramic component comprises of from about 0.2 mol. % to about 2 mol. % of the one or more metal oxidants.

7. The method of claim 1, wherein the metal component comprises a conductive metal, steel, stainless steel, nitronic steel, palladium, gold, silver, copper, nickel, titanium, molybdenum, chrome, iron, chromium, manganese, silicon, cobalt, aluminum, carbon, and/or an alloy thereof.

8. The method of claim 1, wherein the interfacial seal comprises a metal oxide.

9. The method of claim 8, wherein the metal oxide present in the interfacial seal comprises chromium oxide.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 provides a schematic of an exemplary interfacial seal 103 between a metal component 101 and a glass-ceramic component 102.

(2) FIG. 2 is a graph of an exemplary GC-SS sealing profile.

(3) FIG. 3 shows SEM cross-section images of an exemplary GC-SS interface, including (a) grey image, (b) P map, (c) Cr map, and (d) Fe map.

(4) FIG. 4 shows (a) an amplified SEM view of the reaction zone in GC and (b) an electron back scattering image showing the Cr-depleted reaction zone in SS.

(5) FIG. 5 shows SEM images and element mapping of an interface between a SS pin and a BPS_Co2 composition (see Table 4 for the wt. % composition of the BPS_Co2 composition).

(6) FIG. 6 shows an SEM image and element maps of SB_100Cu sessile drop sample (see Table 6 for the wt. % composition of the BPS_Co2 composition).

(7) FIG. 7 shows an illustration of GC-SS bonding from interfacial redox for a SB_100Cu sample.

(8) FIG. 8A-8B shows a schematic of an exemplary fixture 1410 (FIG. 8A) an exemplary GC seal between a GC 1425 and a pin 1420 (FIG. 8B).

(9) FIG. 9A-9B shows schematics of a connector 1500 having an exemplary GC seal 1525 between a casing 1510 and a plurality of pins 1520 in a perspective view (FIG. 9A) and a cross-sectional view (FIG. 9B).

(10) FIG. 10A-10B shows schematics of another exemplary connector 1600 having an exemplary GC seal 1625 between a casing 1610 and a plurality of pins 1620 (FIG. 10A) and an exemplary assembly 1650 to facilitate seal formation (FIG. 10B).

(11) FIG. 11 shows a schematic for an exemplary method of forming a connector 1700.

(12) FIG. 12A-12B shows schematics for an exemplary assembly 1850 to form an electrical header 1800 in a cut-away view having a fixture 1870 (FIG. 12A) and omitting the fixture (FIG. 12B).

DETAILED DESCRIPTION OF THE INVENTION

(13) The present invention relates to seals and methods to add metal oxidants to glass-ceramic (GC) formulations (e.g., a Li.sub.2OB.sub.2O.sub.3K.sub.2OAl.sub.2O.sub.3SiO.sub.2ZnOP.sub.2O.sub.5 composition) for sealing to a metal (e.g., stainless steel (SS)). In one instance, the metal oxidants serve as oxidation agents to oxidize Mn and Cr in SS and form interfacial bonding oxides (e.g., MnCr.sub.2O.sub.4 and/or Cr.sub.2O.sub.3) via a reduction-oxidation (redox) process. The reduced metals precipitate in GC next to the bonding oxides and confine the bonding oxides to the interface.

(14) Any useful metal oxidant can be employed. In one instance, the selection criteria for the metal oxidant (i.e., oxidation agents) are that the changes of Gibbs free energy for their oxidation reactions are lower than the change of Gibbs free energy of forming P.sub.2O.sub.5, allowing thermodynamically favorable reduction of these agents over P.sub.2O.sub.5 during GC-SS interfacial reactions. P.sub.2O.sub.5 is thus preserved to form Li.sub.3PO.sub.4 crystals, which are nuclei for homogeneous growth of Li.sub.2SiO.sub.3, Li.sub.2Si.sub.2O.sub.5, and cristobalite crystal (SiO.sub.2) in the GC. Examples of selected oxidation oxides added to GC formulation include, but are not limited to, cobalt oxide (e.g., CoO), nickel oxide (e.g., NiO), lead oxide (e.g., PbO), copper oxide (e.g., CuO), and silver oxide (e.g., AgO), as well as combinations thereof.

(15) FIG. 1 shows a schematic of an exemplary interfacial seal 103 formed between a metal component 101 and a glass-ceramic (GC) component 102. The metal oxidant present in the GC component reacts with one or more active reducing metal agents (e.g., Si, Mn, and/or Cr) present in the metal component, thereby forming a seal having a metal oxide.

(16) The seal can be formed in any useful manner. In one instance, a GC component (e.g., a GC component, a raw mixture configured to provide any GC composition described herein) is contacted with a metal component and then heated to a sealing temperature (e.g., a temperature T.sub.1, such as any herein). Then, the components can be cooled to form one or more nuclei, promote one or more crystallization phases, as well as to complete bulk crystallization.

(17) FIG. 2 shows a representative thermal cycle for sealing a BPS glass-ceramic. The duration of the seal and nucleate segment controls the number and size of Li.sub.3PO.sub.4 crystallites. Crystalline Li.sub.2SiO.sub.3 and high expansion cristobalite grow preferentially on the Li.sub.3PO.sub.4 facets during the crystallization stage. The size and volume percentage of cristobalite, and thus the overall CTE of the GC, are controlled by the cooling rate from 1000 C. to 750 C. Lower cooling rate allows for growth of more high expansion cristobalite while higher rate undercuts the crystal growth. The BPS thus has a tunable CTE, ranging from 17 ppm/ C. at a cooling rate of 6.5 C./min, to 16 ppm/ C. at a 10.5 C./min, and 15 ppm/ C. at 20 C./min.

(18) The typical peak melting temperature and time of BPS GC is 900 C.-1000 C. and 10 minutes. Upon heating, lithium metasilicate (Li.sub.2SiO.sub.3) forms at 620 C. At 800 C., Li.sub.2SiO.sub.3 reacts with excess Si and forms lithium disilicate (Li.sub.2Si.sub.2O.sub.5). At 890 C., the GC starts to melt and Li.sub.2Si.sub.205 dissolves to form a liquid from which the nucleating crystals, lithium orthophosphate (Li.sub.3PO.sub.4) grow via Ostwald ripening (see, e.g., Hammetter W F et al., Crystallization kinetics of a complex lithium silicate glass-ceramic, J. Am. Ceram. Soc. 1987; 70(8):577-82). It is during the 1000 C. isothermal segment of the heat treatment that the number and size of the Li.sub.3PO.sub.4 crystals is established (see, e.g., Headley T J et al., Crystallization of a glass-ceramic by epitaxial growth, J. Am. Ceram. Soc. 1984; 67(9):620-5). The Li.sub.3PO.sub.4 crystals could grow up to 1 m in size and subsequently serve as nucleation agents for the growth of Li.sub.2SiO.sub.3 and high expansion crystalline silica phases. Upon cooling from 100 C. to 750 C., bulk crystallization of Li.sub.2SiO.sub.3 and cristobalite SiO.sub.2 over the Li.sub.3PO.sub.4 nuclei occurs within the S-GC. Details of the nucleation and crystallization in BPS GC could be found in work by Headley and Loehman (see, e.g., Headley T J et al., Crystallization of a glass-ceramic by epitaxial growth, J. Am. Ceram. Soc. 1984; 67(9):620-5; and Loehman R E et al., Design of high thermal expansion glass-ceramics through microstructural control, in Ceramic Microstructures '86 (Materials Sciences Research Volume 21), eds. J. A. Pask and A. G. Evans, New York and London: Plenum Press, 1987, pp. 33-43).

(19) Metal Oxidants

(20) The glass-ceramic composition can include one or more metal oxidants. In particular embodiments, the metal oxidant is from Group I from Table 1 below. In other embodiments, the metal oxidant is cobalt oxide (e.g., CoO), nickel oxide (e.g., NiO), lead oxide (e.g., PbO), copper oxide (e.g., CuO or Cu.sub.2O), silver oxide (e.g., AgO or Ag.sub.2O), iron oxide (e.g., FeO, or Fe.sub.3O.sub.4, or Fe.sub.2O.sub.3), molybdenum oxide (e.g., MoO.sub.3 or MoO.sub.2), tungsten oxide (e.g., WO.sub.2 or WO.sub.3), and tin oxide (e.g., SnO.sub.2), as well as combinations thereof.

(21) TABLE-US-00001 TABLE 1 Standard Gibbs Free Energy per Mole of O.sub.2 Gibbs Free Energy (G, kJ/mol Oxidation Reaction O.sub.2, 25 C.) Note 4Ag + O.sub.2 = 2Ag.sub.2O 22 Group 1: 2Cu.sub.2O + O.sub.2 = 4CuO 219 Candidates for dopant 2MoO.sub.2 + O.sub.2 = 2MoO.sub.3 275 oxides in GC as 4Cu + O.sub.2 = 2Cu.sub.2O 289 oxidants for preferred 2Pb + O.sub.2 = 2PbO 379 reduction over P.sub.2O.sub.5 2Co + O.sub.2 = 2CoO 425 to form interfacial 2Ni + O.sub.2 = 2NiO 430 bonding oxides by Mo + O.sub.2 = MoO.sub.3 445 redox 2Fe + O.sub.2 = 2FeO 482 4/3Fe + O.sub.2 = Fe.sub.2O.sub.3 495 2W + O.sub.2 = WO.sub.2 505 W + O.sub.2 = WO.sub.3 509 Mo + O.sub.2 = MoO.sub.2 530 Sn + O.sub.2 = SnO.sub.2 535 6FeO + O.sub.2 = 2Fe.sub.3O.sub.4 549 P + O.sub.2 = 2/5P.sub.2O.sub.5 572 Group 2: 2Zn + O.sub.2 = 2ZnO 636 Existing oxidants in GC 4/3Cr + O.sub.2 = Cr.sub.2O.sub.3 695 Group 3: 2Mn + O.sub.2 = 2MnO 726 Alloying elements in Si + O.sub.2 = SiO.sub.2, quartz 856 SS as reducing agents Si + O.sub.2 = SiO.sub.2, 856 cristobalite Si + O.sub.2 = SiO.sub.2, 855 tridymite Si + O.sub.2 = SiO.sub.2, 851 quartz glass

(22) Glass-Ceramic Composition

(23) The glass-ceramic material can be any useful composition. In one embodiment, the GC composition includes a final composition of SiO.sub.2 (e.g., 65-80 wt. %), Li.sub.2O (e.g., 8-16 wt. %), Al.sub.2O.sub.3 (e.g., 2-8 wt. %), P.sub.2O.sub.5 (e.g., 1-5 wt. %), K.sub.2O (e.g., 1-8 wt. %), B.sub.2O.sub.3 (e.g., 0.5-7 wt. %), and ZnO (e.g., 0.1-5 wt. %). High-strength seals can also be made using 304 stainless steel as well as nitronic stainless steel in the metal housing or casing and other stainless steel alloys with the disclosed glass-ceramic of the present invention by belt processing. Additional materials include belt processable S-glass (BPS) (e.g., including Li.sub.2OSiO.sub.2Al.sub.2O.sub.3K.sub.2OB.sub.2O.sub.3P.sub.2O.sub.5ZnO, as described in U.S. Pat. No. 5,820,989), S glass (e.g., including Li.sub.2OSiO.sub.2Al.sub.2O.sub.3K.sub.2OB.sub.2O.sub.3P.sub.2O.sub.5, as described in U.S. Pat. No. 4,414,282), or SB glass (S glass with a lower boron content), as well as any material described in U.S. Pat. Nos. 4,414,282 and 5,820,989, each of which is incorporated herein by reference in its entirety. Table 2 shows an exemplary composition, as well as minimum and maximum wt. %, for S, SB, and BPS glasses.

(24) TABLE-US-00002 TABLE 2 Exemplary composition of BPS GC and S GC S glass (wt. %) SB glass (wt. %) BPS glass (wt. %) Oxide Min Max Target Min Max Target Min Max Target SiO.sub.2 65 80 71.7 71.5 78 74.32 65 80 74.1 Li.sub.2O 8 16 12.6 11 14 13.02 8 16 12.7 Al.sub.2O.sub.3 2 8 5.1 4 5 4.66 2 8 3.8 K.sub.2O 1 8 4.9 4 5 4.04 1 8 2.95 B.sub.2O.sub.3 1.5 7.0 3.2 1 1.5 1.38 0.5 7 1.2 P.sub.2O.sub.5 1 5 2.5 2 3 2.58 1 5 3.15 ZnO 0 5 1.85

(25) The GC composition can be formed from raw materials that are functionally equivalent to the components present in the final composition. For instance, functionally equivalent raw materials include phosphates, carbonates, borates, lithium forms, etc. of any of the components described herein for the GC composition. In one example, the GC composition herein includes from about 8 wt. % to about 16 wt. % of Li.sub.2O, which can be introduced as a raw material that is a carbonate form of the Li.sub.2O component (e.g., as a Li.sub.2CO.sub.3 raw material). Other raw materials include carbonate forms of K.sub.2O (e.g., K.sub.2CO.sub.3 raw materials); lithium forms of P.sub.2O.sub.5 (e.g., Li.sub.3PO.sub.4 raw materials); as well as borate forms of B.sub.2O.sub.3 (e.g., H.sub.3BO.sub.3 raw materials).

(26) Applications

(27) GC compositions can be used in numerous applications. For instance, GC-to-metal seals can be useful for electrical connections (e.g., in co-axial single pin or multiple-pin connectors, headers, as described in U.S. Pat. Nos. 4,414,282 and 8,082,663, which is incorporated herein by reference in its entirety). Electrical connectors with hermetic seals are used in aerospace applications, such as communications satellites, microwave communications equip, and military communications and radar systems; the automotive industry; and commercial communications.

(28) GC-to-metal seals can include any useful metal. Exemplary metals include a conductive metal, stainless steel (e.g., 304L, 330, or 333 stainless steel), nitronic steel, palladium, gold, silver, copper, nickel, titanium, molybdenum, chrome, iron, chromium, manganese, silicon, cobalt, aluminum, carbon, as well as alloys thereof (e.g., a silver-copper nickel alloy, and a nickel-molybdenum chrome alloy (Hastelloy C-276, Inconel 625, or Inconel 718)). Table 3 shows an exemplary composition for 304L stainless steel.

(29) TABLE-US-00003 TABLE 3 Exemplary composition of 304L stainless steel Element Wt. % Element Wt. % Fe 67.41 S 0.003 Cr 19.49 P 0.018 Ni 10.34 C 0.017 Mn 1.77 Nb 0.017 Si 0.62 N 0.01 Cu 0.13 Al 0.006 Mo 0.12 Ti 0.002 Co 0.05 O 0.001

(30) Such metals can be of any useful form, such as a pin, a housing, a casing, etc. In one embodiment, the metal pin is a commercially available pin, referred to as a Paliney 7 pin, with the approximate composition of 34% Pd, 30% Ag, 10% Pt, 10% Au, 15% Cu, and less than 1% Zn. In another embodiment, the metal pin is a commercially available pin, referred to as a Paliney 6 pin, with the approximate composition of 43-45% Pd, approximately 37-39% Ag, approximately 0.8-1.2% Pt, approximately 15-17% Cu, and less than or equal to 1.2% Ni. In one embodiment, these pins have a hardness of greater than or equal to 300 Knoop after heat treatment incorporated into the sealing cycle. The metal pin can also be comprised of 330 or 333 stainless steel.

(31) GC-to-metal seals can be characterized and tested in any useful manner, including sessile drop tests to characterize surface tension and wetting behavior; bonding tests, by way of measuring adhesion strength; and/or general morphology and element mapping by scanning electron microscopy (SEM).

EXAMPLES

Example 1: Design Strategy for Forming GCtSS Seals

(32) Interfacial seals between a GC component and a SS component can be difficult to form. In particular, complex reaction dynamics occur at the interface between the GC and SS components. If the GC component includes a lithium silicate-based composition, then active oxidation agents include P.sub.2O.sub.5 and/or ZnO. In a stainless steel metal, one or more active reducing agents include Si, Mn, and Cr.

(33) In the absence of other energetically preferential reactions, the active oxidation agents and active reducing agents will react during typical sealing conditions, thereby producing oxidized reducing agents (e.g., oxidized Si, Mn, and/or Cr) and reduced oxidation agents (e.g., reduced P.sub.2O.sub.5). Lithium silicate-based glass-ceramics with high coefficients of thermal expansion, designed to form matched hermetic seals in 304L stainless steel housing, show little evidence of interfacial chemical bonding, despite extensive inter-diffusion at the glass-ceramic-stainless steel (GC-SS) interface.

(34) Retention of P.sub.2O.sub.5 nucleic within the GC composition can provide numerous benefits. In one instance, such P.sub.2O.sub.5 nuclei can be the source of Li.sub.3PO.sub.4 in the GC phase, in which Li.sub.3PO.sub.4 in turn serves as a nucleation agent for beneficial crystalline phases in GC (e.g., beneficial high CTE cristobalite SiO.sub.2 phases, lithium silicate (Li.sub.2SiO.sub.3) phases, and lithium disilicate (Li.sub.2Si.sub.2O.sub.5) phases.

(35) To retain P.sub.2O.sub.5 nucleic within the GC composition, the GC composition can be modified with one or more metal oxidants to ensure that reaction of the metal oxidant is thermodynamically favorable, as compared to reaction of the P.sub.2O.sub.5 agent. Particular parameters include a concentration of the metal oxidant in the GC component, a combination of two or more metal oxidants (e.g., a combination of a copper oxide and a cobalt oxide, or a combination of a copper oxide and a lead oxide), as well as mobility of the metal ions (from the metal oxidant) in the GC phase via bulk diffusion.

(36) Metal oxidants can be selected from those that, according to their Gibbs free energy, allow for oxidation of Si, Mn, and/or Cr from stainless steel, and yet to prevent a reduction of P.sub.2O.sub.5 in the glass-ceramic composition. Accordingly, a series of glass-ceramic compositions were modified with a variety of oxidants (e.g., AgO, FeO, NiO, PbO, SnO, CuO, CoO, MoO.sub.3, and WO.sub.3) and then examined for the feasibility of forming bonding oxides through reduction-oxidation (redox) at the GC-SS interface. Additional details are provided in the following Examples.

Example 2: Modified Glass-Ceramic Compositions Based on BPS Formulation

(37) Lithium silicate glass-ceramics, usually modified with a small amount of other oxides (for example, K.sub.2O, B.sub.2O.sub.3, Al.sub.2O.sub.3, ZnO, and often P.sub.2O.sub.5 as the high temperature nucleating agent) have been extensively used for sealing electrical feedthroughs in metal housings (see, e.g., Fluegel A, Thermal expansion calculation for silicate glasses at 210 C. based on a systematic analysis of global databases, Glass TechnologyEur. J. Glass Sci. Technol. A 2010; 51(5):1191-201; and Donald I W, Glass-to-metal seals, Sheffield, UK: Society of Glass Technology, 2009). Glass-ceramic-to-metal (GCtM) seals combine the ease of conventional glass-to-metal (GtM) seals, as well as the performance often achieved in ceramic-to-metal (CtM) seals. The GCtM seals in particular offer several distinct advantages: high temperature mechanical strength of the seals, high dielectric breakdown strength for greater electrical isolation, good hermeticity for environmental protection, and long-term reliability. These unique properties enable GCtM seals to be widely adopted in high performance components and systems that are often deployed in extreme environments.

(38) The interfacial reactions between the GC-SS have been characterized (see, e.g., Donald I W, Preparation, properties and chemistry of glass- and glass-ceramic-to-metal seals and coatings, J. Mater. Sci. 1993; 28:2841-86; Knorovsky G A et al., Interfacial debonding in stainless steel/glass ceramic seals, Sandia Report No. SAND89-1866C, Albuquerque: Sandia National Laboratories; 1989; Loehman R E, Processing and interfacial analysis of glass-ceramic to metal seals, in Technology of glass, ceramic, or glass-ceramic to metal sealing (eds. WE Moddeman, C W Merten, and DP Kramer), American Society of Mechanical Engineers:1987, vol 4, pp. 39-46; Watkins R D et al., Interfacial reaction between a complex lithium silicate glass-ceramic to metal seals, Adv. Ceram. Mater. 1986; 1(1):77-80; Donald I W et al., Interfacial reactions in glass-ceramic-to-metal seals, J. Am. Ceram. Soc. 2008; 91(3):715-20; Kunz S C et al., Thermal expansion mismatch produced by interfacial reactions in glass-ceramic to metal seals, Adv. Ceram. Mater. 1987; 2(1):69-73; Bengisu M et al., Interfacial reactions between lithium silicate glass-ceramics and Ni-based superalloys and the effect of heat treatment at elevated temperatures, J. Mater. Sci. 2004; 39:605-18; and Loehman R E, Interfacial reaction in ceramic-metal systems, Ceram. Bull. 1989; 68(4):891-6).

(39) In part, during the sealing process, Cr diffuses rapidly into the glass-ceramic and is oxidized to form Cr.sub.2O.sub.3 crystals positioning in the glass-ceramics as deep as several tens of m from the GC-SS interface. Iron phosphate (FeP) compounds, due to iron diffusion into the glass-ceramic, are also observed, often next to or coexisting with the Cr.sub.2O.sub.3 crystals. On the other hand, a reduction of P.sub.2O.sub.5 in the glass-ceramic, as well as an accumulation of P near the GC-SS interface, are also observed. FIG. 3a is a SEM cross section image of a GC-SS interface.

(40) There is a distinct P depletion band along the GC-SS interface in the mapping of P (FIG. 3b). Several high P concentration areas indicate a clustering of P in the depletion band, as well as an accumulation of P at the GC-SS interface. The light-colored crystals in glass-ceramic next to the GC-SS boundary in FIG. 3a clearly correlates to the Cr.sub.2O.sub.3, as confirmed by the Cr mapping in FIG. 3c. The size of the Cr.sub.2O.sub.3 crystals is about 1020 m. In the SEM image, the Cr.sub.2O.sub.3 crystals exist at up to 40 m into the glass-ceramic from the interface. In addition, a distance as far as 100 m has also been observed. A close inspection of the Fe mapping in FIG. 3d shows the co-existence of Fe and P clusters, which were further verified by high magnification SEM images as FeP compounds. The compounds are generally seen in close proximity of the Cr.sub.2O.sub.3 crystals.

(41) One of the fundamental driving force for the GC-SS interface reactions is the thermodynamically favorable reduction-oxidation (redox), where P.sub.2O.sub.5 serves as an oxidation agent and the Cr/Mn/Si is the reduction element. Table 1 lists standard Gibbs free energies in descending value for the formation of common oxides relevant to this study. Group 1 lists candidate oxides for glass-ceramic modification (discussed in the experiment procedure section). It is clear that the Gibbs free energy of Cr.sub.2O.sub.3, MnO, and all SiO.sub.2 polymorphs in group 3, representing oxides of metal elements from stainless steel, are lower than those of group 2 P.sub.2O.sub.5 and ZnO in glass-ceramic. The difference in the Gibbs free energy is the driving force of the redox at the GC-SS interface, with the Group 2 oxides being reduced for oxidation of metal elements in Group 3.

(42) During GCtSS sealing, the Cr, Mn, and Si in stainless steel are readily oxidized as part of the interfacial redox reaction(s). However, Cr becomes the dominant reducing agent due to its high weight percentage. In the glass-ceramic, only P.sub.2O.sub.5 and ZnO could serve as the oxidation agents, but the reduction of P.sub.2O.sub.5 is much more thermodynamically favorable over that of ZnO. Iron diffusion from stainless steel into glass-ceramic is common in GCtM seals. However, in the current GC-SS system, the iron appears to combine to the reduced phosphor to form FeP compounds as the formation of Fe.sub.2O.sub.3 is thermodynamically unfavorable due to the relatively higher Gibbs free energy of Fe.sub.2O.sub.3 (Table 1).

(43) Due to the depletion of P and thus, the inadequate formation of the Li.sub.3PO.sub.4 nuclei in the glass-ceramic near the interface, a very course microstructure with large LS crystals are seen in the area. The region with P depletion and abnormal crystal growth is designated as the reaction zone in the glass-ceramic. In general, the CTE of the reaction zone is lower than the CTE of the bulk glass-ceramic. Furthermore, the low CTE reaction zone may lead to tensile stress in the glass-ceramic and could contribute, in part, to interfacial separation.

(44) In the stainless steel, the depletion of Cr was found to be concentrated within 23 m from the interface. FIG. 4a-b show SEM cross sections of the GC-SS interface with a tapered cut and polish. The angle between the GC-SS interface and the polishing plane was 8. The shallow polishing angle allowed an expanded view of the reaction zones with an amplification factor of about 7.2. Multiple Cr.sub.2O.sub.3 crystals are clearly visible in the reaction zone in the glass-ceramic in FIG. 4a. The white needle-like spots on or near some Cr.sub.2O.sub.3 crystals are FeP compounds. The electron back-scattering (EBS) image on the stainless steel in FIG. 4b shows well defined grains in bulk stainless steel but a very different morphology of the reaction zone near the interface. Voids in grains and depleted grain boundaries exist throughout the reaction zone. Previous XRD work on the Cr depletion zone in stainless steel suggested the structure was body-centered cubic (bcc) at room temperature while the bulk stainless steel is face-centered cubic (fcc) (see, e.g., Susan D F et al., Surface alloy depletion and martensite formation during glass to metal joining of austenitic stainless steels, Sci. Technol. Welding Joining 2012; 17(4):321-32; and Donald I W, Glass-to-metal seals, Sheffield, UK: Society of Glass Technology, 2009). The thin bcc layer, it was suggested, was formed on cooling from a high temperature fcc structure as a result of martensitic phase transformation in low Cr steels. The large volume change associated with the fcc-bcc phase transformation was believed detrimental to GC-SS seals because it produced higher stresses and a change of CTE near the interface (see, e.g., Knorovsky G A et al., Interfacial debonding in stainless steel/glass ceramic seals, Sandia Report No. SAND89-1866C, Albuquerque: Sandia National Laboratories).

(45) One critical observation from all of the cross-sectional SEM images is the lack of chemical bonding, defined as an interfacial oxide adhesion layer, between the glass-ceramic and stainless steel. Despite the penetration of glass into the small pores and fine grain boundaries in the stainless steel and the inter-diffusion between the two materials, there is no evidence pointing to the existence of a distinct and continuous redox induced oxide layer at the interface.

(46) It is well recognized that from a chemical or molecular approach, bonding must be accomplished by a transition zone in which the metallic bonding of the metal is gradually substituted for the ionic-covalent bonding of the glass (see, e.g., Bauleke M P et al., How to solve the problems of body cracking and glaze popping in stoneware bodies, Kansas Geological Survey 1978 April; bulletin 211 (part 4):23-7). FIG. 1 illustrates the concept where a metal oxide layer is formed inside glass near the metal-glass interface. Strong chemical bonding can be achieved between a metal and a glass only if the conditions during bonding are such that the glass at the interface becomes and remains saturated with the appropriate metal oxide. Any effects that disrupt and erode this process may negatively impact the quality of the bond. It is worth noting that strong chemical bonds may not form either when there is no such metal oxide present or when the metal oxide diffuses far into the glass without saturation at the interface.

(47) The formation of interfacial oxides requires a source of oxygen. However, the atmosphere for GCtSS sealing needs to be inert, normally a flowing N.sub.2 or Ar gas, to prevent oxidation of the stainless steel. In addition, the graphite, as the most common fixture material to avoid sticking to glasses and glass-ceramics, is chemically reducing in nature. For a GC-SS seal assembly held in graphite fixtures, there is essentially no oxygen available in the local atmosphere surrounding the glass-ceramics throughout a sealing cycle. These competing requirements, a need for oxygen to form an interfacial oxide bonding layer and a lack of oxygen inside the graphite fixture, limit the choice of forming bonding oxides to essentially two approaches: 1) pre-oxidation of SS to form surface oxides, typically C.sub.2O.sub.3, MnCr.sub.2O.sub.4 spinel and SiO.sub.2 for 304L type stainless steel; or 2) doping of glass-ceramic with oxides to serve as oxidants to enable the interface redox reaction to form the bonding oxides. Herein, we focused on the second scenario, where pre-oxidation of the stainless steel housing is not an option, or it is undesirable in real applications.

(48) The goal of modifying glass-ceramics is to add metal oxidants as sacrificial oxides, which could be reduced preferably over the P.sub.2O.sub.5 to preserve P.sub.2O.sub.5 for the formation of Li.sub.3PO.sub.4 nuclei. At the same time, the reduced metal may accumulate in glass-ceramic at the interface to form a barrier to block the diffusion of Cr and Fe from the stainless steel. In principle, any oxidants having higher Gibbs free energy (for example, the oxides in Group 1 in Table 1) than that of P.sub.2O.sub.5 could serve the purpose. The criteria for a viable modification of glass-ceramic are the following, in one non-limiting instance: 1) the metal oxidants could be reduced, preferably over P.sub.2O.sub.5; 2) the metal ions of the dopants have the mobility to quickly diffuse to the interface; so kinetically, the redox is feasible; and 3) the metal oxidants need to be reduced at the GC-SS interface for redox reaction rather than at the free surfaces of the glass-ceramic because of the inert sealing atmosphere or even a reducing atmosphere inside the graphite fixture.

(49) A standard method to form modified GCs is to co-melt all raw materials, including dopant (e.g., metal oxidants), in a single melt to produce glass frits or ribbons for further fabrication of preforms. This fabrication route of modified GCs is designated as co-melted and was adapted as the main method of making modified GCs described herein. Of course, other methods can be employed to form modified GCs including one or more metal oxidants. For instance, an alternative to form modified GCs is to mix powder of the same dopants with that of the base GC, designated as co-mixed, and form preforms thereafter. The co-mixed approach was validated in limited trails, where the partially molten GC dissolved the dopant metal oxide(s) in pre-sintering of the PPPs. The bonding and interfacial reactions of the co-mixed GCs appeared to be the same as those from the co-melted GCs.

(50) The actual selection of oxides for glass-ceramic modification was based on multiple factors, including the Gibbs free energy, ease of processing the glass-ceramic, cost, and the environmental, safety and health (ESH) impact. Each glass-ceramic composition was batched at 1.5 kg weight. The mixed raw oxide materials were melted in a platinum crucible at 1550 C. with extensive stirring. The molten glass was cast to disc ingots at a diameter about 2 inches and thickness 0.25 to 0.50 inch. All modified glass-ceramics were fabricated by Ceradyne Viox Inc. (Seattle, Wash., USA), which is now a division of 3M Corp.

(51) Table 4 provides modified glass-ceramic compositions based on the BPS formulation. Each added oxide was balanced by the same weight reduction on SiO.sub.2.

(52) TABLE-US-00004 TABLE 4 Modified glass-ceramic compositions based on BPS formulation BPS_base BPS_Zn BPS_Co1 BPS_Co2 BPS_Cu1 BPS_Cu2 Oxides (wt. %) (wt. %) (wt. %) (wt. %) (wt. %) (wt. %) SiO.sub.2 74.35 72.2 73.98 73.61 72.95 72.35 Li.sub.2O 12.70 12.7 12.64 12.57 12.55 12.7 Al.sub.2O.sub.3 3.80 3.8 3.78 3.76 3.79 3.8 K.sub.2O 2.95 2.95 2.94 2.92 3.00 2.95 B.sub.2O.sub.3 1.20 1.2 1.19 1.19 1.17 1.2 P.sub.2O.sub.5 3.15 3.15 3.13 3.12 3.27 3.15 ZnO 1.85 4.0 1.84 1.83 1.85 1.85 CoO 0.5 0.99 CuO 1.41 2.0

(53) In Table 4, the BPS_Zn, with more than double the wt. % of the original ZnO, was designed to study whether the extra ZnO could enhance the redox reaction with Cr from stainless steel. The idea was to test whether the relative higher wt. % of ZnO could shield P in glass-ceramic, although thermodynamically, the reduction of P.sub.2O.sub.5 is more favorable. An analogy is the preferred oxidation of Cr over Mn and Si from stainless steel due to the high weight percentage of Cr, although the formation of Cr.sub.2O.sub.3 was less thermodynamically favorable. The BPS_Co1 and BPS_Co2 with 0.50 and 0.99 wt. % CoO were fabricated because the Gibbs free energy of CoO is lower than that of P.sub.2O.sub.5. In addition, the small amount of CoO doping was one of the most commonly used methods to promote the adhesion of glass and glass ceramics to metals. The CuO modification was attractive due to the even lower formation energy of CuO, as compared to CoO.

(54) Images of SEM cross section and element mapping of BPS_Co2 and SS pin are shown in FIG. 5. A reaction zone in the GC was observed next to interface, which was characterized by a coarser crystalline structure, as compared to the much finer crystalline structure in the bulk GC. In addition, we observed Cr diffusion from SS, as well as formation of Cr.sub.2O.sub.3 at the GC-SS boundary and in the GC. While the Cr.sub.2O.sub.3 in GC may not contribute to the adhesion, the Cr.sub.2O.sub.3 at the interface could promote adhesion. There also appears to be metallic Cr in the BPS_Co2 GC, which was not seen in BPS GC. P depletion was present in the GC reaction zone, and an accumulation of P was observed at the interface. The reduction of P.sub.2O.sub.5 could be a source for oxidation of Cr, similar to what occurred in regular BPS GC-SS interface. Finally, reduction of CoO to metallic Co and CoFe alloys were observed in the reaction zone. Co was also observed to accumulate in SS near the interface. It appeared that the intended CoO redox occurred at the GC-SS interface, and helped the GC to SS adhesion. However, the CoO did not shield P.sub.2O.sub.5 from being reduced. The reaction zone in GC with coarse microstructure still existed due to a depletion of P.

(55) Modified glass-ceramics based on SB glass-ceramic are listed in Table 5. All of the dopant metal oxidants belong to Group 1 in Table 1 and have higher Gibbs free energy than that of P.sub.2O.sub.5. The amount of oxide was fixed at 1 mol. % with an exception of CoO modified glass-ceramics, where a composition with 2 mol. % CoO was fabricated.

(56) TABLE-US-00005 TABLE 5 Modified glass-ceramic compositions based on SB formulation SB_Pb SB_Ni SB_Fe SB_Ag SB_Sn SB_W SB_Mo SB_Co1 SB_Co2 Oxides (wt. %) (wt. %) (wt. %) (wt. %) (wt. %) (wt. %) (wt. %) (wt. %) (wt. %) SiO.sub.2 72.81 74.74 74.78 74.09 73.95 72.71 73.83 74.74 73.47 Li.sub.2O 12.3 12.62 12.63 12.51 12.49 12.28 12.47 12.62 12.59 Al.sub.2O.sub.3 3.72 3.82 3.82 3.78 3.78 3.71 3.77 3.82 3.81 K.sub.2O 2.94 3.02 3.02 3.00 2.99 2.94 2.98 3.02 3.01 B.sub.2O.sub.3 1.15 1.18 1.18 1.17 1.17 1.15 1.16 1.18 1.18 P.sub.2O.sub.5 3.2 3.29 3.29 3.26 3.25 3.20 3.25 3.29 3.28 Dopant 3.88 1.33 1.28 2.19 2.38 4.02 2.53 1.34 2.66 Note 1 mol. % 1 mol. % 1 mol. % 1 mol. % 1 mol. % 1 mol. % 1 mol. % 1 mol. % 2 mol. % PbO NiO FeO AgO SnO WO.sub.3 MoO.sub.3 CoO CoO

(57) In particular, the CoO-modified glass-ceramics demonstrated improved GC-SS bonding. But the low mobility of Co.sup.++ ions in the GC limited the amount of CoO that can diffuse to and participate in redox at the interface.

Example 3: CuO-Modified Glass Ceramics

(58) The CuO-doped glass-ceramics were formulated at a finer scale. Table 6 shows a formulation of base SB glass-ceramic and the compositions up to 1 mol. % CuO at an increment of 0.25 mol. % of CuO. The CuO-modified glass-ceramics showed the formation of a continuous layer of strong bonding Cr.sub.2O.sub.3 at the interface in low partial oxygen (PO.sub.2) atmosphere. However, in a local reducing atmosphere, the CuO was preferentially reduced at the surface of glass-ceramic rather than the GC-SS interface for redox.

(59) TABLE-US-00006 TABLE 6 CuO-modified glass-ceramics based on SB glass formulation SB_base SB025Cu SB050Cu SB075Cu SB100Cu Oxides (wt. %) (wt. %) (wt. %) (wt. %) (wt. %) SiO.sub.2 76.01 75.68 75.13 75.00 74.67 Li.sub.2O 12.65 12.64 12.81 12.63 12.61 Al.sub.2O.sub.3 3.83 3.82 3.82 3.82 3.81 K.sub.2O 3.03 3.03 3.03 3.02 3.02 B.sub.2O.sub.3 1.18 1.18 1.18 1.18 1.18 P.sub.2O.sub.5 3.30 3.29 3.32 3.29 3.29 CuO 0.36 0.71 1.06 1.42 Note 0.25 mol 0.50 mol 0.75 mol 1.00 mol % CuO % CuO % CuO % CuO

(60) A modified GC with the addition of CuO as oxidation agent was designed and bonded to SS to prove the concept. Reduction of CuO and precipitation of micron-sized metallic Cu spheres were seen in GC next to the interface. Sub-micron thick MnCr.sub.2O.sub.4 bonding layer via redox was formed at the GC-SS interface. There was no P depletion zone in GC. Homogeneous nucleation and growth of crystals were observed throughout the GC. Our preliminary pull tests indicated the strength of GC to SS bond can reach as high as 14000 psi.

(61) An SEM cross sectional examination of the SB_100Cu sessile drop sample is shown in FIG. 6. Key findings included the presence of metallic Cu particles at sub-micron size in the GC near the GC-SS interface. There was no Cu in the bulk GC. All Cu++ appeared to diffuse towards and accumulated next to the interface. Further, there was a thin (e.g., about 0.1 m) Cr rich oxide layer between the discrete Cu particles and the SS. For comparison, the thickness of oxide layer containing Cr, Mn, and Si from the preoxidation of austenitic stainless steel is about 1 m (see, e.g., Susan D F et al., The effects of pre-oxidation and alloy chemistry of austenitic stainless steels on glass/metal sealing, Oxid. Metals 2010; 73(10):311-35). The Cr depletion layer was seen between the thin Cr rich layer and SS, suggesting an interfacial redox reaction. Most importantly, all Cr depleted from SS appeared to accumulate right at the interface without any indication of Cr diffusion into, and subsequent formation of Cr.sub.2O.sub.3 in, the bulk GC. The CuO apparently shielded the P.sub.2O.sub.5 in GC from being reduced and served as preferred oxidant for interfacial redox. Based on selected area electron diffraction analysis, Cr was the main reducing agent from SS to participate in the interfacial redox, and the interfacial bonding oxide was Cr.sub.2O.sub.3.

(62) FIG. 7 illustrates the microstructure of the SB_100Cu GC-SS interface as a result of the interfacial redox. Starting from GC, the sequence of the materials can be described as bulk GC, a combination of GC and Cu particles, a thin layer assembly of Cr.sub.2O.sub.3 crystals, a thin Cr depletion layer in SS, and bulk SS. However, the observation of reduced Cu particle next to Cr.sub.2O.sub.3 was first observed in the current study. It should be emphasized that this type of bonding could occur only in GC-SS seals, where an interfacial redox proceeded over the surface reduction of the oxidants in GC.

Example 4: Multidopant-Modified Glass Ceramics

(63) A few glass-ceramics doped with two oxidants were fabricated (Table 7). The purpose of composition SB050Cu050Co was to learn the combined redox effect from CoO and CuO. The formation of SB100Cu100Pb allowed examining the combined redox from CuO and PbO, as well as the effect of the anticipated decrease in glass-ceramic viscosity on the wetting behavior with the addition of PbO.

(64) TABLE-US-00007 TABLE 7 Glass-ceramics with multiple dopants SB050Cu050Co SB100Cu100Pb Oxides (wt. %) (wt. %) SiO.sub.2 74.71 71.53 Li.sub.2O 12.61 12.25 Al.sub.2O.sub.3 3.81 3.71 K.sub.2O 3.02 2.93 B.sub.2O.sub.3 1.18 1.14 P.sub.2O.sub.5 3.29 3.19 CuO 0.71 1.38 CoO 0.67 PbO 3.86 Note 0.5 mol % CoO and 1.0 mol % PbO and 0.5 mol % CuO 1.0 mol % CuO

Example 5: GC Preforms for Use in Hermetic Seals

(65) The GC compositions herein can be employed in any useful purpose. In one use, the GC composition (including one or more metal oxidants) is formed into a preform, which is then inserted into a cavity in which a hermetic seal is to be formed. Then, the preform is heated to a sealing temperature (e.g., temperature T.sub.1) and then processed according to any thermal process that produced a near linear strain GC (e.g., any thermal process described herein). This example provides non-limiting embodiments of preforms and their use.

(66) The GC composition herein is particular useful for forming hermetic seals to metals (e.g., stainless steel, as well as other metals described herein). The formation of GC seals can be tested in any useful manner. For instance, FIG. 8A-8B provides a schematic for a fixture 1410 including a bore 1415 through which a metal pin 1420 and a GC preform can be inserted. A seal can be formed between two metal pins 1420 and the GC material 1425, and the strength of this seal can be tested by pulling on the pin (black downward arrow in FIG. 8B).

(67) In one embodiment, the GC-to-metal seal includes a hermetic seal with a plurality of pins. As seen in FIG. 9A, the connector 1500 includes a plurality of pins 1520 (e.g., metal pins) disposed within a casing 1510 (e.g., a metal casing). A hermetic GC seal 1525 is formed between each pin and a cavity within the casing (FIG. 9B). The GC seal can include any useful GC composition described herein.

(68) The GC composition can extend along the entire bore of the casing or within only a portion of the bore. As seen in FIG. 10A, the connector 1600 includes a plurality of pins 1620 (e.g., metal pins) disposed within a casing 1610 (e.g., a metal casing), in which a GC seal 1625 is provided within each bore 1615.

(69) The connector can be formed in any useful way. Generally, a solid GC preform is placed within a cavity or bore requiring the GC seal. Then, the preform is treated to a high temperature sufficient to melt the preform, such that the melted GC composition fills the cavity and forms a hermetic seal. During this melting step, the flow of the GC composition should be controlled (e.g., by employing fixtures and/or inserts). Finally, thermal processes are conducted to crystallize particular polymorphs and solidify the seal.

(70) For instance, as seen in FIG. 10B, an assembly 1650 can be employed to control the melting and sealing process. The solid GC preform 1680 is placed within each bore of the casing 1610 in which a seal is desired. To ensure that the GC composition remains within the bore during the melting step (or sealing step at temperature T.sub.1, as employed herein), a top insert 1665 and a bottom insert 1660 is placed on each end of the bore. The assembly 1650 also includes a top fixture 1675 located on a top surface of the casing 1610, as well as a bottom fixture 1670 located on a bottom surface of the casing 1610. Each of these fixtures includes recesses that align with each pin 1620 and insert 1660,1665. After sealing, the fixtures and inserts can be removed to provide the connector 1600.

(71) In another embodiment, the GC seal is formed by employing a molded casing. As seen in FIG. 11, the casing 1710 includes a molded cavity into which a GC preform 1780 is placed. The casing also allows for placement and alignment of a pin 1720. To ensure that the GC composition remains within the molded cavity, a fixture 1770 can be employed. Then, sealing 1701 occurs by heating the assembly to a first temperature T.sub.1 sufficient to melt the GC preform. After a thermal treatment schedule, the assembly is released 1702 to form the connector 1700.

(72) Any useful connector can be employed. In one instance, the connector can be an electrical header. FIG. 12A-12B shows an exemplary assembly 1850 to form an electrical header 1800. The assembly 1850 includes a hair pin 1820 disposed within a cavity of the header casing 1810 and inserted through bores 1815 located within the cavity. In addition, a removable fixture 1870 with aligned recesses 1875 are employed during sealing, and inserts 1860 are placed within the bores 1815 to maintain the GC composition within the bores. The GC preform 1880 is placed within the cavity of the header casing 1810, and the two preforms sandwich the U-shaped portion of the hair pin 1820.

(73) The GC seal 1825 is formed (e.g., using any thermal treatment described herein), and then the fixture 1870 and inserts 1860 are removed from the assembly. In addition, the hair pin 1820 is ground to form two pins 1825 that are properly spaced and sealed within the header 1800. Additional inspecting and testing steps, as well as electropolishing and cleaning steps, can be performed on the final header.

OTHER EMBODIMENTS

(74) All publications, patents, and patent applications mentioned in this specification are incorporated herein by reference to the same extent as if each independent publication or patent application was specifically and individually indicated to be incorporated by reference.

(75) While the invention has been described in connection with specific embodiments thereof, it will be understood that it is capable of further modifications and this application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure that come within known or customary practice within the art to which the invention pertains and may be applied to the essential features hereinbefore set forth, and follows in the scope of the claims.

(76) Other embodiments are within the claims.