U.S. patent application number 11/835480 was filed with the patent office on 2008-01-17 for large area flat image sensor assembly.
Invention is credited to Mario J. Ciminelli, Michael A. Marcus, Jaime I. Waldman.
Application Number | 20080012082 11/835480 |
Document ID | / |
Family ID | 38948387 |
Filed Date | 2008-01-17 |
United States Patent
Application |
20080012082 |
Kind Code |
A1 |
Waldman; Jaime I. ; et
al. |
January 17, 2008 |
LARGE AREA FLAT IMAGE SENSOR ASSEMBLY
Abstract
A low temperature method for producing a substantially flat
large area image sensor assembly, the method includes the steps of
providing a die attach substrate having a substantially planar
surface; providing a lead frame having a bonding surface and a
plurality of leads extending there from; adhering an imager die to
the substantially planar surface of the die attach substrate with a
low curing temperature first adhesive; and adhering the die attach
substrate with adhered imager die to a bonding surface of the lead
frame with a low curing temperature second adhesive for producing
an image sensor assembly.
Inventors: |
Waldman; Jaime I.;
(Pittsford, NY) ; Ciminelli; Mario J.; (Rochester,
NY) ; Marcus; Michael A.; (Honeoye Falls,
NY) |
Correspondence
Address: |
F-P, Patent Legal Staff;Eastman Kodak Company
343 State Street
Rochester
NY
14650-2201
US
|
Family ID: |
38948387 |
Appl. No.: |
11/835480 |
Filed: |
August 8, 2007 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10922529 |
Aug 20, 2004 |
7276394 |
|
|
11835480 |
Aug 8, 2007 |
|
|
|
09957188 |
Sep 20, 2001 |
|
|
|
10922529 |
Aug 20, 2004 |
|
|
|
Current U.S.
Class: |
257/431 ;
257/E31.019; 257/E31.032 |
Current CPC
Class: |
H01L 27/14683 20130101;
H01L 27/14618 20130101; H01L 2924/16195 20130101; H01L 2224/48227
20130101; H01L 2924/10253 20130101; H01L 2224/73265 20130101; H01L
2924/10253 20130101; H01L 2224/45124 20130101; H01L 2224/73265
20130101; H01L 2224/48091 20130101; H01L 2924/3025 20130101; H01L
2224/32188 20130101; H01L 2224/32245 20130101; H01L 2224/45124
20130101; H01L 2924/1517 20130101; H01L 2924/00 20130101; H01L
2224/32245 20130101; H01L 2924/00 20130101; H01L 2924/00014
20130101; H01L 2224/48091 20130101; H01L 2924/15153 20130101; H01L
2224/48227 20130101; H01L 2924/00 20130101 |
Class at
Publication: |
257/431 ;
257/E31.032; 257/E31.019 |
International
Class: |
H01L 31/0352 20060101
H01L031/0352; H01L 31/0304 20060101 H01L031/0304 |
Claims
1. A large area flat image sensor assembly comprising: (a) a die
attach substrate having a substantially planar surface; (b) an
imager die adhered to the substrate and having an active surface
dimensions of substantially 35 by 35 mm or greater and a deviation
from flatness of the active imager surface of less than
substantially 10 microns.
2. The large area flat image sensor as in claim 1 further
comprising a die attach substrate composed of aluminum nitride.
3. The large area flat image sensor as in claim 1 further
comprising a double side lap polished die attach substrate with a
surface flatness of less than 2 microns total indicated
run-out.
4. The large area flat image sensor as in claim 1 further
comprising a first adhesive layer containing beads.
5. The large area flat image sensor as in claim 1 wherein a z
height of the bonded imager die varies by less than 15 microns with
respect to the second surface of the die attach substrate.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This is a divisional of U.S. Ser. No. 10022,529. which is a
continuation-in-part of application Ser. No. 09/957,188, filed
Sept. 20, 2001 entitled "A Method For Producing An Image Sensor
Assembly" by Jaime I. Waldman et al,
FIELD OF THE INVENTION
[0002] The invention relates generally to the field of large image
sensors and, more particularly, to such large image sensors that
are assembled in a cavity package and are substantially flat over
the entire active imager surface providing improved image capture
capability.
BACKGROUND OF THE INVENTION
[0003] Large area imagers, CCDs and CMOS, are required to be flat
to capture a quality image. For many applications, it is required
that large area imagers be manufactured to form a substantially
flat (with a deviation from flatness of less than 15 microns)
active imaging surface over the entire active imaging area. A large
area image sensor assembly is defined as a packaged imager having
an active sensor area of 20 mm by 20 mm or larger. Currently, these
CCD or CMOS imagers are composed of an imager die mounted on either
a substrate or mounted in an electronic package. When an imager die
is mounted on a substrate, there are several deficiencies that
result. The die and bond wires are not protected from damage or
debris. Additional potting of the wires or additional structural
elements must be added to protect the die and wires. Without these
additions, the imager remains unprotected. When currently available
electronic cavity packages are used, the wires and imager are
protected but the flatness of the imager is not sufficient to meet
the needs of many applications including medical imaging sensors
and large format digital cameras. Current electronic packages use
high temperature methods to join the package components. These high
temperatures approximately 400.degree. C. and higher are used to
either melt glass, braze or co-fire a ceramic package as methods to
join components together. These high temperatures and fastening
techniques cause the critical die attach area to which the imager
is attached to the electronic package to warp. The die attach area
needs to be flat to create a flat imager. Since these are cavity
packages, they impede post grinding of the die attach area to
repair the warping or bowing of the imager plane created during the
high temperature fabrication processes.
[0004] Other methods of packaging an image sensor include mounting
an imager into an injection molded thermoplastic resin package as
disclosed by H. Yamanaka in U.S. Pat. No. 5,529,959. The assembly
process includes injection molding of a base with an incorporated
lead frame. This patent discloses that small imagers can be made
flat by this process without quantifying a definition of flatness.
Although injection molding is highly successful for the
manufacturing of small imager packages, it is extremely difficult
to achieve a base flatness of 10 .mu.m or less in larger areas in
excess of 25 mm by 25 mm or larger. U.S. Pat. No. 5,382,310 by
Ozimek et al. describe a method to make small conventional solid
state image sensors by directly bonding an imager die to a
conductive lead frame. The imager bonding pads are then wire bonded
to the lead frame. The lead frame and imager are then encapsulated
top and bottom with adhesive to provide structural strength. This
approach is not amenable to making large area flat imagers. U.S.
Pat. No. 6,121,675 by Fukamura et al. describes a method which
utilizes silicone to cover the die and wirebonds thus preventing
moisture and dirt egress. The flexible material prevents the wire
bonds from breaking.
[0005] Although the currently known and utilized methods for
producing an image sensor assembly are satisfactory for many
applications, they include drawbacks. The imagers produced by
conventional packaging methods do not have sufficient flatness
after the brazing or glass melting process to meet the requirements
for large size medical imaging sensors and large format digital
cameras. In addition, the flat substrates are not enclosed which
obviously limits their ability to mount optical coverglass or
protect the sensor and wire bonds.
SUMMARY OF THE INVENTION
[0006] The present invention is directed to overcoming one or more
of the problems set forth above. Briefly summarized, according to
one aspect of the present invention, the invention embodies a low
temperature process and method for producing a substantially flat
image sensor assembly, the method comprising the steps of (a)
providing a die attach substrate having a substantially planar
surface; (b) providing a lead frame having a plurality of leads
extending therefrom and a shelf on which a cover glass may be
attached; (c) attaching an imager for collecting incident light to
the substantially planar surface with a low temperature first
adhesive substance; and (d) attaching the imager to a portion of
the lead frame with a low curing temperature second adhesive
substance for producing an image sensor assembly with a flat cavity
package.
[0007] These and other aspects, objects, features and advantages of
the present invention will be more clearly understood and
appreciated from a review of the following detailed description of
the preferred embodiments and appended claims, and by reference to
the accompanying drawings.
Advantageous Effect of the Invention
[0008] The present invention has the following advantage of
providing a substantially flat image sensor in a substantially flat
cavity package while including a shelf on which a cover glass may
be attached. The assembly provides a method to protect the bond
wires, mount the cover glass and prevent contamination egress onto
the sensor without additional structural components.
BRIEF DESCRIPTION OF THE DRAWINGS
[0009] FIG. 1 is a view in vertical cross section of the image
assembly of the present invention;
[0010] FIG. 2 is a view in vertical cross section of an alternative
embodiment of FIG. 1;
[0011] FIG. 3 is a view in vertical cross section of another
alternative embodiment of FIG. 1;
[0012] FIG. 4 is a top view of the first adhesive layer
pattern;
[0013] FIG. 5A is a view in vertical cross section of an imager die
to substrate alignment fixture without the die and substrate;
[0014] FIG. 5b is a view in vertical cross section of an imager die
to substrate alignment fixture with the die and substrate mounted
in place; and
[0015] FIG. 6 is a leveled surface map of an imager mounted in a
package using the inventive method and apparatus of this
invention.
DETAILED DESCRIPTION OF THE INVENTION
[0016] Referring to FIG. 1, the present invention will be described
showing only the right most portions of an image sensor assembly
10, and it is to be understood that an exact duplicate portion or
mirror image of this portion is on the left portion. In other
words, only a portion of the entire image sensor assembly 10 is
shown for clarity of understanding. Furthermore, in the following
description, like reference characters designate like or
corresponding parts throughout the several views of the drawings.
Also in the following description, it is to be understood that such
terms as "top," "bottom," "left," "right," "upwardly,"
"downwardly," and the like are words of convenience and are not to
be constructed as limiting terms.
[0017] The image sensor assembly 10 includes a lead frame 20,
usually rectangular in shape, having a plurality of leads 30 (only
one is shown) along its edge which can be electrically insulated
from each other, and which extend from and are attached to a
rectangular shaped lead frame portion 40, and in combination with
the frame portion on the opposite side (not shown), forms a
hollowed-out portion into which a suitable imager die 60 and
suitable die attach substrate 50 is to be inserted. The lead frame
portion 40 is a multi-tiered portion extending substantially
perpendicular to the leads 30. The lead frame portion 40 of lead
frame 20 is shown to have three tiers (or layers), although it is
to be understood that more or less tiers could be used and each of
these tiers may be made up of more than one layer. The top tier 40a
or shelf provides a shelf for affixing a cover glass 45 to enclose
the imager assembly. The middle tier 40b is slightly longer then
the top tier 40a and it contains metallization such as lead traces
47 used to provide a means of electrically connecting the imager
die 60 to the leads 30. Bond pads 46 are contiguous with the lead
traces 47 and provide a surface to attach wire bonds 80 to the
imager die 60. The bottom portion 40c is used to provide mechanical
features for precisely locating the substrate 50 and imager die 60
within the lead frame 20.
[0018] An imager die bonding surface 50a of a substrate 50 is
ground, lap polished or produced substantially flat, and an imager
die 60 (such as silicon die) with an imager active surface 60a and
an imager bonding surface 60b is affixed to the flat imager die
bonding surface 50a of the substrate 50 at the imager bonding
surface 60b by a first adhesive layer 71, such as room temperature
curing adhesive, which is located between the two surfaces 50a and
60b . After curing the first adhesive layer 71 the imager and
substrate assembly is then attached to the lead frame 20. A second
adhesive layer 72, such as epoxy, is applied between the lead frame
bonding surface of the substrate 50b and the bonding surface 41 of
the lead frame portion 40. The bonding of the lead frame 40 to the
substrate 50 results in a flat cavity package.
[0019] Bond wires 80 are then attached over the upper gap region 91
between the imager die 60 and middle tier 40b for electrically
connecting the two together. As mentioned above, the cover glass 45
is then placed atop the top tier 40a for enclosing the image
assembly. The cover glass may or may not incorporate optical
characteristics to provide enhanced imaging. The cover glass 45 may
also utilize a light shield to block unwanted light from impinging
on the wire bonds 80 thus creating spurious illumination onto the
imager, as illustrated in U.S. Pat. No. 6,075,237.
[0020] Referring to FIG. 2, there is shown an alternative
embodiment of FIG. 1. This embodiment is the same as FIG. 1 except
for the inclusion of a step 92 in the substrate 50 used for
reference positioning and/or constraining the die attach substrate
50 within the lead frame 20. Referring to FIG. 3, there is shown
still another alternative embodiment of FIG. 1. In this embodiment,
the step 92 is inverted from the position of FIG. 2, and the step
92 performs the same functions.
[0021] To provide a proper focal plane, a Charge Coupled Device
(CCD) or Complimentary Metal Oxide Semiconductor (CMOS) image
sensor assembly 10, and more particular the active surface of the
imager die 60, should be as flat as possible. Large image sensors
used in medical applications are required to be flat to less than
15 microns. These large image sensors such as the some Kodak.RTM.
sensors with dimensions 53 mm by 52 mm have less than 15 microns
total indicated run-out on the die surface. The Kodak.RTM. sensor
typically incorporates 4.3 million 24-micron square pixels. For
medical imaging applications, these sensors can have a glass
coherent fiber optic bundle optically adhered to their surface. The
gap between the fiber optic bundle and the imager must be kept to a
minimum in order to provide sufficient imaging properties.
[0022] The actual process to produce a flat 53 mm by 52 mm imager
is described below in detail. First, a substantially flat die
attach substrate 50 must be provided. Double side lap polished
aluminum nitride is one material that provides ideal properties for
use as the substrate 50. Aluminum nitride has ample heat transfer
and dissipation of 179 W/mK and can be plated with a conductive
metal if conductivity is required. Double side lap polished
aluminum nitride substrates of 0.080 inches thickness and 63 mm
square with surface flatness of less than 2 microns total indicated
run-out (i.e. a substantially planar surface) and thickness
uniformity of better than 10 microns have been used as the starting
substrate 50. The term substantially planar surface as used herein
is defined as a surface having a deviation from flatness as less
than 5 microns over the entire surface.
[0023] Room temperature curing adhesive or epoxy is used to adhere
the imager die 60 to the die attach substrate 50. The adhesive must
also be of the type that does not stress the imager die 60 upon
curing which would cause warping of the components. The adhesive
should also be low outgassing in nature. The die attach adhesive
which forms the first adhesive layer 71 can also incorporate beads
of known and tightly controlled size distribution. These beads are
of a material which is of a significantly higher modulus of
elasticity than the die attach adhesive carrier material. These
beads can be conductive in order to allow for electrical and
thermal conduction. An important benefit of the beads is to provide
a uniform first adhesive layer 71 thickness as the die is pressed
against the die attach substrate 50. An appropriate first adhesive
layer thickness is in the range of 0.5 to 1.0 mil. A proper first
adhesive layer pattern 71a (shown in FIG. 4) is made on the die
attach substrate 50. This pattern 71a is typically of a star shape
without excessive loading in the center as shown in FIG. 4. The
pattern is put down with an adhesive dispenser. The shape of the
pattern is designed to allow air to escape upon compression of the
die onto the die attach adhesive and to spread the adhesive
uniformly over the surface of the substrate. The lines forming the
pattern in FIG. 4 are typically 20-40 mils wide. Alternatively, the
pattern could be put down using a template. Zymet ZVR-6000.2, an
ultra-low stress adhesive has been found suitable for use in this
operation.
[0024] A die attach fixture 100 shown in FIGS. 5A and 5B is used to
adhere the imager die 60 to the substrate 50. The die attach
fixture 100 consists of a pair of aligned parallel first and second
metal plates 120 and 110 held in fixed orientation by a set of
linear guides 105 (typically 4). The second plate 110 incorporates
a vacuum port 130 and an imager die recess 65 on its surface facing
the first plate 120. The vacuum port 130 is used to provide vacuum
to hold the imager die in place prior to the pressing operation for
securing the imager die 60 onto the die attach substrate 50 in the
correct x, y locations within 50 microns in x and y. The imager die
60 is cut from a wafer so that it fits snuggly and matingly into
the die recess sidewalls 68 of the imager die recess 65 of the
second plate 110. The first plate 120 includes a substrate recess
55 with substrate recess sidewalls 58 on its surface facing the
second plate 110. This recess similarly constrains the x and y
locations of the substrate 50 to within 50 microns x and y. The
combination of the linear guides 105 and the fixed locations of the
substrate recess 55, substrate recess side walls 58, imager die
recess 65 and die recess side walls 68 provides an alignment
mechanism for precisely positioning the imager die 60 with respect
to the substrate 50. The substrate fits snuggly into the substrate
recess 55 to provide precision placement of the imager die 60 on
the substrate 50. The substrate 50 with first adhesive layer
pattern 71a facing toward the second plate 110 is placed into the
substrate recess 55. Alternatively, the adhesive pattern can be
applied after mounting the substrate 50 into the substrate recess
55. The vacuum pump (not shown) is turned on when the imager die 60
with its imager active surface 60a facing the vacuum port 130 and
its imager bonding surface 60b facing the first plate 120 is
inserted into the imager die recess 65 to provide suction in the
vacuum port 130. The second plate 110 of the die attach fixture 100
is then pressed against the first plate 120 until uniform contact
is made between the imager bonding surface 60b and the imager die
bonding surface 50a of the substrate 50 patterned with first
adhesive 71a . The linear guides 105 and the parallelism of the
plates 110 and 120 ensure that an even force distribution is
obtained over the entire imager surface during assembly. Constant
pressure is applied for a period of time. Typical pressures of 2
psi are utilized during pressing. When the pressure is applied the
adhesive spreads out evenly over the imager bonding surface 60b to
form a continuous layer of first adhesive 71. The vacuum is shut
off after contact is made between the imager die 60 and substrate
50 surfaces. Accurate z position is maintained from the top of the
die attach substrate to the surface of the die with the utilization
of the beads in the first adhesive. Since the die attach substrate
50 is substantially flat with a deviation in thickness of less than
10 microns and a flatness of the imager bonding surface of the
substrate of 2 microns or less the z height of the bonded imager
die 60 varies by less than 15 microns with respect to the second
surface of the die attach substrate 50c . After pressing for about
0.3 -5 minutes the pressure is removed and the die attach adhesive
is then allowed to cure at room temperature for approximately 6
hours. The combination of 2 psi and 30 seconds compression time has
been found to be sufficient to result in a uniform 0.8-1.0 mil
thick first adhesive layer. When cured at room temperature no
thermal expansion or contraction is induced to cause die warping.
Alternatively, it has been found that curing at 100.degree. C. for
one hour does not adversely affect the observed flatness of the
imager using Zymet ZVR 6000.2 or similar adhesives due to the fact
that the modulus of the adhesive is so low. Low curing temperature
as used herein is defined as covering the temperature range of
substantially between 20 to 120 degrees C. Desired properties of
the first adhesive layer include a modulus of elasticity in the
range of 0.5 to 50 MPascal.
[0025] The substrate 50 with adhered imager die 60 is next adhered
to a lead frame 20. The lead frame 20 incorporates a plurality of
leads 30 or pins to connect to a circuit board. The lead frame 20
incorporates a lead frame portion 40 of material such as alumina
and also incorporates bond pads 46 that are electrically connected
to the leads 30 by lead traces 47. Room temperature cure (20-25
degrees C) low out-gassing adhesive (Ablestik brand for example) is
applied as second adhesive layer 72 to secure the alumina frame
portion to the die attach substrate. The second adhesive is
preferably applied as a 2 mil thick film gasket and can also be
dispensed. When cured at low temperature, thermal expansion
mismatches are minimized. Once cured, the second adhesive can
provide a stress relief between the die attach substrate and the
lead frame portion should subsequent temperature fluctuations
occur. The die attach substrate can be designed with a step 92 as
shown in FIG. 2 so that it can fit into the alumina frame portion
for better mechanical fastening and improved x,y positioning. A low
force clamping technique is used that will provide minimal stress
upon the die attach substrate so that exceptional flatness of the
alumina frame portion is not critical when the two surfaces are
co-joined. The second adhesive squeezes out to fill any voids left
by a flatness mismatch of the alumina frame and provides sufficient
mechanical attachment.
[0026] The bond pads on the die are next electrically attached to
the bond pads 46 on the lead frame 20 by wirebonding wires 80.
Aluminum wire wedgebonding is preferred for the assembly of flat
imagers since this can be performed at room temperature. The lead
frame portion 40a can also be used as a mount to secure a
protective optical window 45. This window prevents particles from
coming in contact with the die. The window can be secured with room
temperature cure adhesive or sealed with tape to keep particles
out. Permacel Kapton P-224 tape has been found suitable and allows
for easy removal of the window if required by the customer.
EXAMPLE 1.
[0027] The flatness of typical image sensors (imagers 53 mm by 52
mm) as described and assembled by the above described method were
tested using a low coherence light interferometric apparatus
described in U.S. Pat. No. 6,724,487. Flatness data were obtained
and the maximum bow of a typical set of typical Kodak.RTM. imagers
(53 .times.52 mm) are shown in Table 1. FIG. 6 shows a leveled
surface map of imager #149 mounted in a package obtained using the
low coherence light interferometric apparatus described in U.S.
Pat. No. 6,724,487. TABLE-US-00001 TABLE 1 Imager # Deviation from
flatness (microns) 145 6.26 146 6.84 147 4.69 148 6.82 149 6.41 150
6.51 151 7.89 152 8.50 153 8.02 Average 6.88 Stdev 1.14 Max 8.50
Min 4.69 Range 3.81
Comparative Example 1
[0028] The data shown in Table 2 were obtained on Kodak.RTM.
imagers (53 mm by 52 mm) made with conventional commercially
available packages of alumina with a brazed copper tungsten
substrate. The flatness of these imagers was tested by the same
method described in Example 1. TABLE-US-00002 TABLE 2 Imager #
Deviation from flatness 13 57.6 14 46.1 27 53.4 30 78.3 31 68 32
64.5 45 35.4 47 42.2 65 35.2 68 52.4 74 48.7 83 60.1 86 55.9 91
33.1 93 38.4 Average 51.28 stdev 13.23 max 78.3 Min 33.1 range
45.2
[0029] The invention has been described in detail with particular
reference to certain preferred embodiments thereof, but it will be
understood that variations and modifications can be effected within
the spirit and scope of the invention.
PARTS LIST
[0030] 10 image sensor assembly [0031] 20 lead frame [0032] 30
leads [0033] 40 lead frame portion [0034] 40a top tier or shelf of
frame portion [0035] 40b middle tier portion of frame portion
[0036] 40c bottom tier of frame portion [0037] 41 bonding surface
of lead frame [0038] 45 cover glass/optical window [0039] 46 bond
pads [0040] 47 lead traces [0041] 50 die attach substrate [0042]
50a imager die bonding surface of substrate [0043] 50b lead frame
bonding surface of substrate [0044] 50c second surface of substrate
[0045] 55 substrate recess [0046] 58 substrate recess sidewalls
[0047] 60 imager die [0048] 60a imager active surface [0049] 60b
imager bonding surface [0050] 65 imager die recess [0051] 68 die
recess sidewalls [0052] 71 first adhesive layer [0053] 71a first
adhesive layer pattern [0054] 72 second adhesive layer [0055] 80
wire bonds [0056] 91 upper gap region [0057] 92 step [0058] 100 die
attach fixture [0059] 105 linear guide [0060] 110 second metal
plate [0061] 120 first metal plate [0062] 130 vacuum port/line
* * * * *