Metal fusion bonding – Process – Metal to nonmetal with separate metallic filler
Reexamination Certificate
2000-06-28
2001-11-13
Dunn, Tom (Department: 1725)
Metal fusion bonding
Process
Metal to nonmetal with separate metallic filler
C228S124500, C228S124100, C228S208000
Reexamination Certificate
active
06315188
ABSTRACT:
CROSS-REFERENCE TO RELATED APPLICATIONS
This application is related to co-pending application, “Coating System to Permit Direct Brazing of Ceramics”, by Charles H. Cadden and F. Michael Hosking, both of Albuquerque, N.Mex.; commonly assigned to Sandia Corporation, Albuquerque, N.Mex.
BACKGROUND OF THE INVENTION
This invention relates generally to the field of brazing, and more specifically to a method for preparing the surface of a high purity alumina ceramic or sapphire specimen that enables direct brazing in a hydrogen atmosphere using an active braze alloy. The present invention also relates to a method for directly brazing a high purity alumina ceramic or sapphire specimen to a ceramic or metal member using this method of surface preparation, and to articles produced by using this brazing method. In this context “brazing” is defined as the joining of two or more members using a filler metal that is heated above it's liquidus temperature (e.g. melted) in a reduced-oxygen atmosphere furnace at a temperature greater than 450 C.
In U.S. Pat. No. 4,542,073, Tanaka discloses a method for bonding a ceramic first member to a second member made of ceramic or metal, where the method comprises two sequential thermal treatments. The first thermal treatment is performed by heating the ceramic first member in an oxidizing atmosphere from 1200 C to 1400 C, whereby an oxide surface layer is formed, only if the bulk ceramic contains an oxide-forming element, such as silicon (e.g. Silicon Carbide, Silicon Nitride, or Sialon). Next, a metal sheet (e.g. copper) is placed in-between the two members and a second thermal treatment is performed in a nitrogen atmosphere, whereby the oxidized ceramic surface reacts with the metal sheet to form a chemical bond (e.g. copper oxide). Tanaka teaches that during the second thermal treatment step the peak temperature is not higher than the melting point of the metal filler sheet. Tanaka's limitation precludes the possibility of brazing the two members together, because brazing requires melting of the filler metal, as is practiced by the present invention.
Traditionally, the brazing of ceramic components requires costly surface preparation steps to ensure adequate bonding to the non-metallic substrate, especially when joining a ceramic to a metal (e.g. Kovar, copper, or steel). The conventional preparation process involves formation of a “metallization layer”, typically a mixture of molybdenum and manganese powders fired at a high temperature (>1400 C). Sintering of this mixture produces a well-bonded, glassy coating having a thickness of about 10-40 microns. This is then followed by a nickel-plating step (e.g. electroplating, electroless coating, or nickel oxide painting), following optionally by a sintering step. This provides a continuous metallic surface of about 2-10 microns thick that can be wet by conventional, non-active brazes. Commonly referred to as the “Mo—Mn—Ni” or the “moly-manganese” metallization process, these costly and environmentally unfriendly manufacturing steps require close process control and inspection.
Active braze alloys (ABA's) were developed as an alternative means to directly wet a ceramic, without the necessity for prior metallization of its surface. The ABA's ability to perform successfully as a direct braze depends upon the chemical reaction(s) that occur between the active element(s) in the braze and the ceramic surface. Early active alloy systems primarily utilized titanium (Ti) as the active component, although Zr, Cr, V, Hf, Ta, and Nb have also been used. Small additions of titanium to common face centered cubic (FCC) filler metals (e.g. Cu, Ni, Ag, Au, and their alloys) have been shown to facilitate wetting on alumina through the formation of a titanium-oxide chemical reaction layer (e.g. TiO and Ti
2
O
3
). Unfortunately, Ti-ABA's (e.g. TiCuSil or CuSil-ABA made by Wesgo, Inc., San Carlos, Calif.) are limited to brazing in either vacuum or inert atmospheres, since titanium forms undesirable hydrides in a reactive hydrogen atmosphere.
Recently, a new family of braze alloys has been developed that are more compatible with brazing in a hydrogen atmosphere. These so-called “second-generation”ABA's employ vanadium (V) as the active constituent. They expand the application of active brazing technology to areas where the preferred, or required, furnace atmosphere is dry hydrogen. One example of a second generation V-ABA braze alloy is based on modifying the eutectic binary AWS BAu-4 alloy (82 Au-18Ni) with additions of Mo and V to form a quarternary alloy (75-98% gold, 0.5-20% nickel, 0.5-6% vanadium, 0.25-6% molybdenum, and 0-5% chromium, wt. %). These ductile braze alloys have a liquidus temperature in the range of approximately 960 C to 1100 C. See U.S. Pat. Nos. 5,385,791, 5,301,861, and 5,273,832 by Mizuhara and Huebel. Such braze joints have yielded generally acceptable hermeticity and tensile strength, but the reaction product(s) responsible for adhesion, a necessary condition for fully realizing the optimum joint properties, is not well understood.
In previous research, the authors fabricated ASTM F-19 tensile specimens of alumina brazed to alumina using Au-16Ni-0.7Mo-1.7V filler metal, that were brazed at 1000 C in slightly positive dry hydrogen. All of these samples were air-fired (e.g. re-sintered) at 1575° C. in air prior to brazing, in order to heal surface flaws from grinding. See Mizuhara, H. and Mally, K. 1985, “Ceramic-to-Metal Joining with Active Brazing Filler Metal”, Welding Journal 64(10): 27-32. The air-fired/brazed 94% alumina tensile specimens had an average tensile strength of 100 MPa, compared to only 40 MPa average strength for the air-fired/brazed high purity 99.8% alumina samples. The increased joint strengths can be attributed to favorable reactions or synergisms with the 6% balance of other constituents in 94% alumina.
It is well known that direct brazing to alumina using titanium as the active element results in a well-defined titanium-oxide chemical reaction layer. However, the authors could not find a comparable well-defined reaction layer in alumina ceramic joints brazed with a V-containing, gold-based filler alloy. Rather, electron microscopy of those braze joints revealed the presence of a discontinuous spinel-type (Al-V-O) compound of limited thickness, typically less than 30 nm at the metal/ceramic interface. However, machined alumina specimens that were air-fired at 1575° C. prior to brazing exhibited small (<50 angstroms thick) quantities of a silicon-rich constituent on external surfaces. This was particularly evident on 94% alumina specimens, which contain 6% of a deliberately added glassy binder phase consisting of a large fraction of silica (SiO
2
). Silicon was also detected, albeit in reduced quantities, on high purity 99.8% alumina samples after the air-firing step.
It is possible that the presence of silicon on the surface of the 94% alumina samples during brazing may have contributed to the superior bond strength (100 MPa), as compared to the 99.8% alumina samples strength (40 MPa). This speculation is supported by wetting trials, where surface analysis of an Au-16Ni-0.7Mo-1.7V sessile drop on 94% alumina (melted at 1000 C in dry hydrogen) revealed solidified braze material concentrated along the exposed glassy phase between the alumina grain boundaries. Thermodynamic calculations of the Gibbs free energy values for simple binary reactions between alumina (or silica) and V (or Mo) indicate that the combination of silica and V produced the lowest free energy value, which is consistent with the observed wetting behavior of the glass grain boundary phase (mostly silica) in the 94% alumina binder. These calculations indicated that silica has only marginal thermodynamic stability in the presence of vanadium at the brazing temperature. Furthermore, SiO
2
is reduced at 1000 C in a hydrogen atmosphere having a dewpoint of approximately −80 C, only slightly drier than the environment used for these experiments (approximately −60 C).
Cadden Charles H.
Hosking Floyd M.
Yang Nancy Yuan Chi
Cooke Colleen
Dunn Tom
Sandia Corporation
Watson Robert D.
LandOfFree
Surface preparation for high purity alumina ceramics... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Surface preparation for high purity alumina ceramics..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Surface preparation for high purity alumina ceramics... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2596696