Page 75 - Wire Bonding in Microelectronics

54 Cha pte r T h ree

bonding wires. These were stored on 5 cm (2 in) spools at 22.8 ± 1.6°C
(73 ± 3 ΕF) for 2 years and tested periodically. In conjunction with
updating the ASTM bonding wire standards, wire manufacturers
were recently contacted and they agreed that the general aging char-
acteristics of current wires are expected to remain generally valid
today. This should hold even with the major metallurgical changes
that have been made in wires for gold ball bonding, as discussed
extensively in the next paragraph.
The original data was published in ASTM standards, F 72 and F 487.
In 2006, those documents were updated and reballotted, soliciting
comments from four current wire manufacturers and other experts.
Aluminum fine wire still primarily consists of Al, 1% Si and has
approximately the same characteristics as shown in the original mea-
surements. The main difference being in more controlled/reproducible
characteristics due to manufacturing improvements. Gold bonding
wire for high-speed autobonders now contains differing and often
much higher percentages (up to 1%) of stabilizing/intermetallic-inhib-
iting additives. Although no equivalent multicompany aging tests
have been made, these new alloys are expected to be at least as stable
over time as the wires of the mid-80s period.
In general, the breaking load of hard, as-drawn wire decreased
rapidly (from 5 to 15%) within 6 weeks after manufacture (thus, hard-
wire is seldom recommended for volume production, where repro-
ducibility is required). It continued to decrease, but more slowly, over
the 2-year period as it self-annealed at room temperature. All stress-
relieved and annealed wires of both gold and aluminum stayed within
their breaking load specification for the entire 2-year test period. The
elongation characteristics for Al, 1% Si wire were more ambiguous
than the breaking load, changing upward or downward but within
the specification extremes, and generally recovering to the median by
the end of the test. The data were compiled and published in ASTM
standards, F 487 (for Al, 1% Si) and in F 72 (for Au alloys, Be and Cu
doped). The aging data for aluminum (1% Si) and gold (<10 ppm
beryllium doped) are given in Figs. 3-2 and 3-3, respectively. Other
aging charts, Al, 1% Mg, and Au + Cu, are not reproduced here
because these wires are seldom used today and have been dropped
from the ASTM standards. However, they were included in the origi-
nal ASTM study (ASTM F 638 and F 72, respectively). The reader is
directed to the ASTM standards for more detailed data on the specifi-
cations and properties of bonding wire (see App. 3A for a listing).
The conclusion drawn from the shelf-life study is that, in general,
small-diameter annealed or stress-relieved wire (not hard, as drawn)
can be used for up to 2 years with only minimal change in its break-
ing load, although Al wire elongation may vary over its entire speci-
fied range. The caveat is that the wire must be stored at approximately
constant room temperature, and exposure to direct sunlight, drafts
from an open door, or possible heat sources must be avoided.