BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates generally to image sensors and packaging methods thereof. In particular, it relates to the formation of either a CMOS (CIS) or charge-coupled device (CCD) packaged image sensor having embedded microlenses and improved sensitivity as a result of a novel packaging process.
2. Description of the Related Art
Solid state image sensors are necessary components in many optoelectronic devices, including digital cameras, cellular phones, and toys. In the simplest possible terms, such a sensor consists of an array of photosensors (eg. photodiodes), connected to solid state devices that convert the signals generated by the photosensors (typically electrical charge) into forms that can be displayed electronically. Two types of signal converting solid state devices in common use are charge-coupled devices (CCDs) and CMOS image sensors (CISs). In order for these devices to operate at optimal levels, the light from the image being sensed must be focused on the photosensors and any transmission loss along the optical pathway between the entrant surface of the device and the photosensors must be minimized. Potential focusing problems can be solved by the formation of small lens structures (microlenses) above the photosensors. These lenses are embedded within the sensing structure and are formed by photolithography techniques followed by melting or reflowing photoresist squares into hemispheres. The transmission loss problem is addressed by the use of transparent bonding materials (epoxies) and glass (or other transparent media) as connective, filtering, and protective layers.
Examples of various forms of the solid state sensor structures are to be found in the prior art. Okamoto (U.S. Pat. No. 6,545,304 B2) discloses a photoelectric converter element group on one section of a semiconductor substrate and a charge transfer path to transfer accumulated signal charge to a contiguous readout gate region having a readout gate electrode associated therewith. Umetsu et al. (U.S. Pat. No. 6,528,831 B2) discloses a solid state image pickup device in which a matrix array of photoelectric sensors are formed adjacent to charge transfer channels and wherein a read-cum-transfer electrode is formed on an insulating layer and surrounds each photoelectric element. These devices are cited here as examples of a CCD type sensor device.
Methods of enhancing the transmission of light to the active elements of the sensing device are also to be found in the prior art. Teranishi et al. (U.S. Pat. No. 5,844,289) discloses a solid state image sensor incorporating surface microlenses with attached optical fiber-bundles. Abramovich (U.S. Pat. No. 6,362,498 B2) discloses a color CMOS image sensor with microlenses formed in a silicon nitride layer by reactive ion etching (RIE).
Before solid state image sensors can be incorporated within target technologies (digital cameras, phones etc.), they must be properly packaged. Packaging serves several purposes critical to the commercial use of image sensors in various target technologies. Packaging protects delicate solid state elements, it provides a suitable and stable configuration for interfacing with various shapes and designs of target technologies and provides easily accessible electrical interconnects that enable the sensors to be conveniently incorporated within a wide variety of such technologies. The beneficial electrical and mechanical attributes of packaging can, however, adversely impact the required optical properties of the image sensors. The present inventors utilize a particular commercially available packaging technology (manufactured and offered for sale by Shellcase corp.), but the problems to be discussed herein would clearly be associated with other packaging methodologies. In particular, as shown schematically in prior art FIG. 1, the packaged image sensor (Shellcase package) is a glass-silicon-glass laminate. Referring to FIG. 1, there is shown a (prior art) commercially packaged image sensing fabrication of the type which is the subject of the present application. The fabrication comprises an upper glass layer (10) and a lower glass layer (20) that sandwiches an optically and electronically active silicon layer (90) which may advantageously be substantially an entire wafer (comprising, thereby, what is called “wafer level chip scale packaging,” WLCSP). The lower glass layer (20) supports a plurality of solder connectors (39) for connection, by melting and re-solidifying, to external circuitry which is not a part of the device. The solder connectors (39) are internally connected (eg. by wires through the glass layer) to the solid state circuitry of the image sensing device (90) which is integrated on a silicon layer. The silicon layer has an upper surface portion (101) which includes an optically sensitive region containing a matrix array of photodiodes covered by an array of embedded microlenses (30) which are shown schematically and greatly enlarged as small convex “bumps”. A layer of filtering material (not shown) may be formed between the microlenses (30) and the photodiodes. The optically and electronically active silicon layer (90) is bonded above (70) and below (80) by a transparent bonding agent (typically, epoxy) to both glass layers (10, 20). Although the bonding agent is kept to a minimum thickness consistent with requirements of structural integrity, it causes an inevitable degradation of the optical signal impinging on and propagating within the optically sensitive region. The degradation is a result of both absorption of the radiation and defocusing of the light by the microlenses (30) because of mismatching of the indices of refraction of the epoxy bonding agent (typically nB=1.5) and the material of the lenses (typically nL=1.63). Although the radius of curvature of the microlens can be adjusted to shorten the focal length, this method has process and physical limits. The focal length of the microlens is typically shortened by increasing the thickness of the photoresist film to obtain a smaller radius. The thickness of the photoresist film is controlled by rotation rate while spin coating, and the rotation rate cannot too slow due to uniformity of the photoresist film. The object of the present invention is to improve sensor sensitivity by restoring the focus of the microlenses onto the optically sensitive region while still retaining the advantages of packaging.
SUMMARY OF THE INVENTION
Accordingly, it is the object of this invention to provide a packaged image sensor with improved image sensitivity.
This object will be achieved within the context of commercially available packages and packaging technology that are known to be advantageously applied at the wafer level chip scale (WLCSP). In particular three embodiments will be presented. In the first embodiment a special intermediate optically refractive layer will be interposed between the epoxy bonding layer and the microlens. The index of refraction, nI, of this layer will compensate for the defocusing of the incident radiation which results from the index of refraction of the epoxy layer combined with the index of refraction of the lens structure. In the second embodiment, an epoxy with a lower index of refraction will be used as the bonding agent, thus achieving the same end without the use of an additional layer. In a third embodiment, the microlens will be formed in a layer having a greater index of refraction (n>1.7).
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 shows a schematic cross-sectional view of a commercially packaged image sensor in accordance with the prior art.
FIG. 2 shows a detail of a prior art unpackaged image sensor wherein the lens and filter layer combine to produce correctly focused radiation.
FIG. 3 shows a detail of the packaged prior art sensor of FIG. 1, illustrating the defocusing of the incident radiation by a bonding material.
FIG. 4 shows a detail of a first embodiment of the present invention, in which an additional layer is interposed between the packaging epoxy and the microlens.
FIG. 5 shows a detail of a second embodiment of the present invention, in which an epoxy layer of lower index of refraction is used.
FIG. 6 shows a detail of a third embodiment of the present invention, in which the microlens is formed in a higher index of refraction material layer.
FIG. 7 shows the optical conditions for optimization of packaged image sensor.
DETAILED DESCRIPTION OF THE INVENTION
The preferred embodiments of the present invention each teach a method of forming a packaged image sensor so that elements of the packaging structure, such as bonding materials, do not adversely affect the optical performance and sensitivity of the sensor. Of particular concern in the fabrication of image sensors is the focusing of incident radiation on the photosensitive elements of the sensors. Referring to FIG. 2, there is shown a cross-sectional schematic view of portion of a prior art image sensor. A series of lines (20) represent the optical path of rays of incident radiation. The rays pass through air (index of refraction, nA=1) and are incident as an essentially parallel bundle on an optically receptive stacked layer (35) which is formed on the upper surface (101) of the semiconductor substrate (100). An upper portion of the optically receptive stacked layer (35) includes a convex (converging) microlens (30) which is formed, typically by coating a microlens photoresist, patterning the microlens photoresist to form an array of microlens pattern, and heating (for example using a hot plate and maintain the temperature at about 160° C.) the patterned photoresist to induce thermal reflow. The convex microlens (30) having an index of refraction, nL, where nL=1.63 is formed as a result of surface tension. Similarly, if the photopattern is striped, the reflow results in forming hemi-cylindrical lenses. The microlens (30) is in contact with a transparent color filter layer (40) whose index of refraction is nCF=1.51 and thickness is about 2-6 μm. The radius of curvature of the microlens (30) is about 0.5-20 μm, 1-8 μm is preferred. The color filter layer (40) is in contact with an IC stacked layer (50) also of index of refraction nIC=1.51 and thickness of about 2-12 μm which depend on IC design and process. The IC stacked layer (50) comprises passivation layer, inter-metal dielectric layers (IMDs), inter-layer dielectric layer (ILD), metal lines, transistors, junctions and other electric elements. The metal lines shield light, therefore, the light-shielding structures are disposed between pixel boundaries. As shown in the figure, the ray bundle is caused to converge (24) by the surface curvature of the microlens (30) combined with its index of refraction nL and, in accord with the radius of curvature of the lens surface and the indices of refraction of the layers (40), (50) beneath the lens, the rays converge to focus on photosensitive elements (60) disposed on the surface of the semiconductor substrate (100). The photosensitive elements (60) are typically further coupled to integrated device structures within the image sensor, such as charge-coupled devices or CMOS circuitry, which convert the output of the photosensors to electrical signals that can be advantageously used by the target technology which incorporates the image sensor. This associated circuitry is not shown as it is not relevant to the nature of the invention. The present invention does not relate to the particular nature of the signal processing within the image sensor, but only to the optical processes acting on the incident radiation as it travels to the photosensors.
Referring next to FIG. 3, there is shown a cross-sectional schematic illustration of the optical effects of a bonding layer of epoxy (70) with nB=1.5 which is used to bond the sensing device (90) of FIG. 2 within a package such as that shown in FIG. 1 using methods of the prior art. The addition of an epoxy layer with index of refraction different from that of the air in FIG. 2, destroys the convergence of the ray bundles (24) and degrades the signal reaching the photosensitive elements (60). The radius of curvature of the microlens (30) cannot be sufficiently reduced to focus incident radiation on the photosensitive elements (60), because of process and physical limits.
FIG. 7 depicts the spherical boundary surface of a microlens (59) radius R centered at point C. An object or point light source O at an object distance DO from the vertex V along the axis of the microlens will refractively converge a cone of light rays to an image at image distance DI from point V. If the index of refraction in the space between the object light source O and the spherical boundary surface (59) of the microlens is N1 and the index in the space inside the lens (to the right of the spherical lens surface in FIG. 7) is N2, then spherical wavefronts will converge to a real image at I. Using the well-known Fermat's Principle, it can be shown that for spherical refracting surfaces:
N1/DO+N2/DI=(N2−N1)/R
and, that when DO is very large, the image focal length is then given by:
Fi=R(N2/N2−N1).
For fixed Fi and N2, we can have a series of optimal R and N1. In real situation, the process limits the radius of curvature R.
Specifically, if N2=1.63, N1=1 (air), for Fi=12 μm, R=4.63 μm is chosen.
If N2=1.63, N1=1.5 (epoxy), for Fi=12 μm, R=4.63 μm is not suitable and R=0.95 μm should be chosen, however, it is hard to control. When R>1.5 μm, the process can be easily controlled and the microlens can achieve better uniformity.
If N2=1.63, Fi=12, R>1.5 μm, N2−N1 greater than 0.2 is suitable for an image sensor with better sensitivity. In real situation, the focal length Fi may be around 12 μm, for example, 11 μm or above 12 μm, because packaged image sensors are trending toward small areas for embedding within optoelectronic devices, such as digital cameras, cellular phones, toys, or watches.
Referring next to FIG. 4, the packaging method of the first embodiment of the present invention is shown, wherein an additional intermediate optically refractive layer (56) with nI<nB is formed between the microlens (30) and the bonding layer (70). For example, while the bonding layer of epoxy has nB=1.5, the additional intermediate optically refractive layer (56) has nI<1.5, where nI between approximately 1.33 and 1.45 is preferred. This additional intermediate optically refractive layer (56) has the characteristics of high transmittance (>90%, greater than 95% is preferred), thermal resistance, chemical resistance and viscosity greater than 5 mpas (greater than 10 mpas is preferred), and can be a layer of material such as [A] a mixture including Fluororesin derivative, Initatoe, Methylisobutylketone (MIBK), and t-butanol, or [B] a mixture including Fluororesin derivative, Initatoe, Melamine resin, Methylisobutylketone (MIBK), t-butanol, formed to a thickness higher than microlens (30), approximately greater than 0.5 μm, 1 μm is preferred. The convergence of the ray bundles (24) is restored by the interposition of this additional intermediate optically refractive layer (56).
Alternatively, the additional intermediate optically refractive layer (56) satisfies the condition of nL−nI greater than 0.2.
Referring next to FIG. 5, there is shown in cross-sectional schematic form, a second embodiment of the present invention wherein an epoxy layer (55) with an index of refraction less than 1.5 (eg. nB=between approximately 1.33 and 1.45 being preferred) is used to bond the sensor within the package. The use of a lower index of refraction epoxy eliminates the need for the additional intermediate layer (56) used in the first embodiment yet also restores the convergence of the ray bundles (24). In this case, the epoxy layer (55) satisfies the condition of nL−nB greater than 0.2 to focus incident radiation on the photosensitive elements (60). For example, the index of refraction nL of the microlenses equals approximately 1.63, and the epoxy layer (55) has an index of refraction nB which is between approximately 1.33 and 1.45. The epoxy layer (55) can be a transparent material such as organopolysiloxane mixture with nB=1.4.
Referring finally to FIG. 6, there is shown in cross-sectional schematic form, a third embodiment of the present invention wherein the microlens (30) is now formed of a transparent material having an index of refraction nL satisfying nL−nB greater than 0.2. For example, the epoxy layer (70) has an index of refraction nB approximately 1.5, and the index of refraction nL of the microlenses (30) is greater than 1.7, with a range between 1.73 and 1.8 being preferred.
As is understood by a person skilled in the art, the preferred embodiments of the present invention are illustrative of the present invention rather than limiting of the present invention. Revisions and modifications may be made to methods, materials, structures and dimensions employed in fabricating a packaged image sensor having improved sensitivity, while still providing such a packaged image sensor having improved sensitivity as described herein, in accord with the spirit and scope of the present invention as defined by the appended claims.
Patent Application
20050001281
