Page 29 - Wire Bonding in Microelectronics
P. 29

8     Cha pte r  O n e


              1.2.2  Assessing the Bondability and Reliability
                     of Proposed New Bond Systems
              There are times (particularly in the MCM, MEMS, and sensor indus-
              tries), when a new metallization system or wire must be bonded
              together. The following is a list of steps to consider when approach-
              ing such problems. Table 1-2 outlines the steps and these are described
              in detail below, with examples.

                  1.  Can it be welded by ultrasonic, TS, and or possibly TC
                    methods? As a first cut, see tables of ultrasonically bondable
                    materials in the Welding Handbook, [1-3] Fig. 1-6. For device
                    metallization, consider how any possible dopants (e.g., Cu
                    and Si in Al) might affect bondability and cratering. Very
                    frequently, someone has worked out the problems you face,
                    and their publication will save time (and money). For exam-
                    ple, in order to bond a 50-µm diameter superconducting
                    niobium wire to niobium thin films [1-4], it was necessary to
                    anneal the wire at ~ 2200°C (by passing current through it in
                    a vacuum), which softened (annealed) it from a Vickers
                                                         2
                    hardness value of 180 down to 80 kgf/mm . To prevent oxi-
                     dation of the thin-film pads before using, a thin coating
                     (4 nm) of Pd was sputtered onto the wire. After those






          1)  Is it known to be weldable by ultrasonic, TC, or TS method? (See Fig. 1-6.)
          2) How can possible metal dopants affect bondability cratering or
             reliability? (Cu or Ti in metal?) (See Chaps. 8 and 9.)
          3) Are there any potential bondability or handling problems that could occur
             for high-volume production (Cu oxidizes easily)?
          4) Does the metal form a soft or hard oxide compared to itself ? (Hard is
             best for bonding!)
          5) Is the new wire (or metallization) harder than Al or Au? There may be
             cratering or bondability problems. (See Chap. 8.)
          6) Are there numerous intermetallic compounds that may form—if so are
             their melting points high or low?—look it up in phase diagrams. (High,
             >600°C is more stable.) (See Chap. 5.)
          7) Are the individual materials easily corroded? Does the bond couple form
             a corrosion couple? Check the electrochemical series. (See Table 1-1.)
          8) Is either metal easily attacked by halogens or sulfur compounds? (They
             are everywhere.) (See App. 5B, Chap. 5.)

         TABLE 1-2  Assessing the Reliability of a Proposed New Bond-Metal-System
   24   25   26   27   28   29   30   31   32   33   34