Active solid-state devices (e.g. – transistors – solid-state diode – Combined with electrical contact or lead – Solder wettable contact – lead – or bond
Reexamination Certificate
2000-07-17
2002-06-11
Chaudhari, Chandra (Department: 2813)
Active solid-state devices (e.g., transistors, solid-state diode
Combined with electrical contact or lead
Solder wettable contact, lead, or bond
C257S772000, C257S778000, C257S786000, C438S108000, C438S612000
Reexamination Certificate
active
06404064
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to FCCSP (Flip-Chip Chip-Scale Package) technology, and more particularly, to a flip-chip bonding structure on substrate for flip-chip package application, on which solder bumps can be bonded for electrically coupling a flip chip to the substrate.
2. Description of Related Art
FCCSP (Flip-Chip Chip Scale Package) is an advanced type of integrated circuit packaging technology that allows the overall package size to be made very compact. By FCCSP, the semiconductor chip is mounted in an upside-down manner over a substrate formed with an array of solder-bump pads, and which is mechanically bonded and electrically coupled to the substrate by means of solder bumps. A conventional flip-chip package configuration is schematically depicted in the following with reference to 
FIGS. 1A-1B
.
FIG. 1A
 shows a schematic sectional view of a conventional flip-chip package configuration. As shown, the flip-chip package configuration includes a substrate 
10
 which is provided with an array of solder-bump pads 
20
 on the front surface thereof. As shown in the top view of 
FIG. 1B
, these solder-bump pads 
20
 are arranged on the periphery of the substrate 
10
. By the FCCSP technology, a semiconductor chip 
30
 is mounted in an upside-down manner over the substrate 
10
 (i.e., with its circuit surface facing down against the front surface of the substrate 
10
), and which is mechanically bonded and electrically coupled to the substrate 
10
 by means of solder bumps 
40
.
Conventionally, there are many various ways of making the solder-bump pads 
20
, including, for example, the SMD (Solder Mask Define) method shown in FIG. 
2
A and the NSMD (Non-SMD) method shown in FIG. 
2
B.
Referring to 
FIG. 2A
, the SMD method includes the use of a solder mask 
51
 having a circularly-shaped opening 
52
 to define solder-bump pad location. Further, a circularly-shaped electrically-conductive layer 
53
 is formed within the circularly-shaped mask opening 
52
 to serve as the intended solder-bump pad, and which is connected to an electrically-conductive trace 
54
.
The SMD method has the benefit of easy fabrication. However, one drawback to the SMD method is that it has poor routability to the electrically-conductive trace 
54
. One solution to the poor-routability problem is to use the NSMD method.
Referring to 
FIG. 2B
, the NSMD method includes the use of a solder mask 
61
 having a circularly-shaped opening 
62
 to define solder-bump pad location. Further, a circularly-shaped electrically-conductive layer 
63
 is formed within the circularly-shaped mask opening 
62
, and an elongated electrically-conductive trace 
64
 is connected to the circularly-shaped electrically-conductive layer 
63
. The electrically-conductive trace 
64
 has an uncovered part 
64
a 
exposed through the mask opening 
62
 and a covered part 
64
b 
extending into the underneath of the solder mask 
61
.
It should be noted that, in the case of 
FIG. 2B
, the combined surface area of the electrically-conductive layer 
63
 and the exposed portion 
64
a 
of the electrically-conductive trace 
64
 collectively serves as the intended solder-bump pad. In other words, the wetted solder would be spread over all exposed electrically-conductive surfaces, including the electrically-conductive trace 
64
 and the exposed portion 
64
a 
of the electrically-conductive trace 
64
.
The NSMD method has the benefit of allowing good routability to the electrically-conductive traces interconnecting the solder-bump pads on the substrate. One drawback to the NSMD method, however, is that the overall pad surface area would be affected by a positional deviation of the solder mask due to misalignment. This problem is described in more details in the following.
Theoretically, when a fixed amount of solder is wetted on a solder-bump pad, the degree of solder collapse would be substantially proportional to the overall surface area of the solder-bump pad. For this sake, in an array of solder-bump pads, if some pads are different in surface area from the others, the resulted array of solder bumps on these solder-bump pads would be non-coplanarized, which would lead to the problem of package warpage.
In addition, since the array of solder bumps are typically highly densified, any misalignment that causes the solder-bump pads to be increased in surface area would cause the attached solder bumps to be more horizontally extended, which would result in a short-circuit problem and a flip-chip underfill problem as illustratively depicted in the following with reference to FIG. 
3
.
FIG. 3
 is a schematic diagram showing two neighboring solder-bump pads 
71
, 
72
 on which two respective solder bumps 
81
, 
82
 are formed. In the example of 
FIG. 3
, the two neighboring solder-bump pads 
71
, 
72
 are circularly-shaped, which are fabricated through the NSMD method.
In 
FIG. 3
, assume P represents the pitch between the two neighboring solder-bump pads 
71
, 
72
; D represents the diameter of each of the circularly-shaped solder-bump pads 
71
, 
72
: W represents the diameter of each of the solder bumps 
81
, 
82
; and G represents the width of the gap between the two solder bumps 
81
, 
82
. In a typical FCCSP layout, P=125 &mgr;m, D=75 &mgr;m, and W=93 &mgr;m; and therefore, G=32 &mgr;m. After reflow, however, the solder bumps 
81
, 
82
 will be further spread out horizontally, thus increasing W to about 105 &mgr;m; and as a result, G is reduced to only about 20 &mgr;m.
The narrowing of the gap G between the neighboring solder bumps 
81
, 
82
 would undesirably result in two problems. First, it would make the neighboring solder bumps 
81
, 
82
 more likely to be short-circuited to each other; and second, the narrowed gap G between the neighboring solder bumps 
81
, 
82
 would make subsequent flip-chip underfill process very difficult to implement.
It can be seen from 
FIG. 4A
 that, for the SMD-type of solder-bump pad shown in 
FIG. 2A
 (which includes only the circularly-shaped electrically-conductive layer 
53
), a positional deviation in the solder mask 
51
 would not affect the overall pad surface area.
However, for the NSMD-type of solder-bump pad shown in 
FIG. 2B
 (which includes the circularly-shaped electrically-conductive layer 
63
 and the exposed portion 
64
a 
of the electrically-conductive trace 
64
), it can be seen from 
FIG. 4B
 that a downward positional deviation in the solder mask 
61
 would cause part of the should-be-covered portion 
64
b 
of the electrically-conductive trace 
64
 to be exposed, thus giving an additional surface area to the solder-bump pad.
As mentioned earlier, a larger solder-bump pad would undesirably cause the attached solder bump to be more collapsed to a lower height, thus resulting a non-coplanarity problem to the resulted array of solder bumps.
Related patents are partly listed in the following:
U.S. Pat. No. 5,834,849 entitled “HIGH DENSITY INTEGRATED CIRCUIT PAD STRUCTURES”.
U.S. Pat. No. 5,637,832 entitled “SOLDER BALL ARRAY AND METHOD OF PREPARATION”;
U.S. Pat. No. 5,783,865 entitled “WIRING SUBSTRATE AND SEMICONDUCTOR DEVICE”;
U.S. Pat. No. 5,915,977 entitled “SYSTEM AND INTERCONNECT FOR MAKING TEMPORARY ELECTRICAL CONNECTIONS WITH BUMPED SEMICONDUCTOR COMPONENTS”;
U.S. Pat. No. 5,535,101 entitled “LEADLESS INTEGRATED CIRCUIT PACKAGE”;
U.S. Pat. No. 5,011,066 entitled “ENHANCED COLLAPSE SOLDER INTERCONNECTION”.
U.S. Pat. No. 5,926,694 entitled “SEMICONDUCTOR DEVICE AND A MANUFACTURING METHOD THEREOF”; and
U.S. Pat. No. 5,489,750 entitled “METHOD OF MOUNTING AN ELECTRONIC PART WITH BUMPS ON A CIRCUIT BOARD”.
None of the above-listed patents, however, teach how to retain a solder-bump pad at its predefined surface area when the solder mask is deviated in position due to misalignment.
SUMMARY OF THE INVENTION
It is therefore an objective of this invention to provide a flip-chip bonding structure on substrate for flip-chip package application, which allows each solder-bump pad to retain predefined surface area even when
Chiu Shih-Kuang
Mao Kuo-Liang
Suo Chao-Dung
Tsai Ying-Chou
Berezny Nema
Chaudhari Chandra
Corless Peter F.
Edwards & Angell LLP
Jensen Steven M.
LandOfFree
Flip-chip bonding structure on substrate for flip-chip... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Flip-chip bonding structure on substrate for flip-chip..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Flip-chip bonding structure on substrate for flip-chip... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2900301