Semiconductor device manufacturing: process – Making device or circuit responsive to nonelectrical signal – Responsive to electromagnetic radiation
Reexamination Certificate
2000-08-09
2003-02-11
Thomas, Tom (Department: 2811)
Semiconductor device manufacturing: process
Making device or circuit responsive to nonelectrical signal
Responsive to electromagnetic radiation
C438S073000, C257S440000
Reexamination Certificate
active
06518085
ABSTRACT:
BACKGROUND OF THE INVENTION
(1) Field of the Invention
The present invention relates to the fabrication of a CMOS color imager on a semiconductor substrate using CMOS technology, and more particularly relates to a method for making photodiodes with improved (more uniform) spectral response across the frequency band for the red, green, and blue (wavelength) pixel cells in the array of photodiodes.
(2) Description of the Prior Art
In camcorders and early digital cameras, charge coupled devices (CCDs) were used as optical detectors for detecting and processing color images. The CCDs detect the light on an array of photosites on the surface of an image sensor, and then the induced charge at each site in a row of the array of photosites is sequentially transferred to a readout register where it is detected, amplified, and processed through an analog/digital converter and stored as digital information for further processing. Although the CCD is a convenient device for photoimaging, it is not as economically practical as the CMOS image sensor for integrating other (camera) circuit functions on the same sensor chip for image processing. Since the CMOS technology is more advanced because of its extensive use in computers for logic functions, such as central processing units (CPUs) and for data storage and the like, the CMOS technology is the technology of choice for color sensors in future digital cameras. The CMOS technology allows more function to be integrated directly on the sensor chip. Therefore fewer CMOS chips are required than for the CCD technology, and the CMOS color imager is more manufacturing cost effective.
To better understand the limitations of the present CMOS color imager technology, a portion of a CMOS color imager, having photodiodes for the primary color pixels for red, green, and blue (R/G/B) wavelengths, is depicted in the schematic cross-sectional view of FIG.
1
. In the conventional process, a single diffused photodiode is typically used for each of the colored (R/G/B) pixels. These photodiodes are formed in the substrate
10
by forming N doped wells
12
, one for each color pixel, in an array of pixels.
FIG. 2A
shows one of the CMOS circuits associated with each photodiode, commonly referred to as an active photodiode circuit. The function of the circuit is to sample the change in output voltage (delta V
out
) on the photodiode at the output node
22
that is a function of the light intensity (number of photons nhv) impinging on the surface of the diode
12
. Briefly, the photodiodes
12
, one of which is shown in
FIG. 2A
, are reversed biased and the diodes are charged to a reset voltage V
reset
. When an optical image is impressed on the photodiodes, the light intensity is measured as the number n of photons (having energy=hv)
20
generating the photocurrent I
ph
and charging the diode
12
. The h is Planck's constant. The v is the frequency of the light, and is related to wavelength lamda by lamda=c/v, where c is the speed of light. The diode
12
is then interrogated using the row select (row_s) gate
24
to determine the change in voltage delta V at the output voltage V
out
, which is proportional to the light intensity or energy nhv. In
FIG. 2B
this delta V is depicted as the change in V
out
in the chart for voltage vs. time, where the vertical axis is a measure of the change in voltage (delta voltage) and the horizontal axis is time in milliseconds. The reset voltage V
reset
26
waveform is shown offset above the plot of the change in output voltage (delta V
out
)
22
′ also plotted in FIG.
2
B. After interrogating the array of photodiodes (pixels) and processing and storing the digital data, the active photodiode circuit is then reset for recording the next optical image. The primary colors (R,G,B) in the image imposed on the CMOS image sensor are determined by using a separate diode for each of the color pixels. This is achieved by using color filters (or dyes), that is red, green, and blue (R/G/B) filter
16
′,
16
″, and
16
′″, respectively, over each of the three diodes, as depicted in FIG.
1
.
The color filter response curve for the photodiodes
12
with the three color filters
16
′ (red),
16
″ (green), and
16
′″ (blue) of
FIG. 1
used to detect the color image is depicted in FIG.
3
A. The color filter response is also shown in arbitrary units. The ideal or preferred response profile is shown as the solid curves
4
, and the actual response profile for a conventional photodiode
12
is depicted by the dashed curves
6
′,
6
″, and
6
′″, respectively, for the red, green, and blue pixel cells
16
′,
16
″, and
16
′″ of FIG.
1
. It is clearly seen that the color filter response is substantially reduced at the shorter wavelength blue pixel cells
6
′″ at a wavelength of 450 nm. This results in poor color fidelity of the original color image. This poor color fidelity is a result of the nonuniform quantum efficiency (QE) across the optical waveband from 450 to 650 nm. This variation in QE is depicted by the curve
2
in
FIG. 3B
, where the y-axis is the QE (the ratio of the number of photoelectrons to the number of photons) measured in arbitrary units, and the x-axis is the wavelength of the light in nm.
Several methods of forming photodiodes for CMOS color imagers have been reported in the literature. For example, Drowley et al., U.S. Pat. No. 6,023,081, describe a semiconductor image sensor in which the red and blue pixels are made in the same type of well area (N well). The blue and red pixels are made after forming the FET gate electrodes. In U.S. Pat. No. 5,965,875 to Merrill a triple diffused well structure is used to separate out and to detect and measure the intensity of the three primary colors red, green, and blue. The method is based on the principle that the absorption length of the light in silicon is a function of the different frequencies. U.S. Pat. No. 6,040,593 to Park et al. describes a method for making a buried diffused photodiode structure with a self-aligned silicide layer for making CMOS image sensors. U.S. Pat. No. 5,122,850 to Burkey describes a method for making CCD image sensors, which include P-stripes (diffused regions) under the transfer gate of the CCD devices and adjacent to but not touching the photodiode which provides effective anti-blooming control while effectively transferring the photocharge to the CCD. U.S. Pat. No. 5,514,620 to Aoki et al. describes a method of using solid state diffusion for making shallow PN junction devices that include photo-electric conversion devices.
However, there is still a strong need in the digital imaging industry to provide photodiodes for color imagers with a more uniform spectral response curve (QE) across the optical spectral range for red, green, and blue pixels (photodiodes).
SUMMARY OF THE INVENTION
A principal object of this invention is to provide photodiodes for color imagers with a more uniform spectral response (QE) for the red, green, and blue (R/G/B) pixels for better color fidelity.
It is another object of this invention to improve the spectral response by varying the individual junction depths of the diffused photodiodes for the R/G/B pixels to provide a more uniform spectral response curve for the optical bandwidth.
Still another object of this invention is to form this more uniform spectral response curve by forming deep photodiode diffused (metallurgical) junctions for the longer wavelength red pixels, and shallower diffused photodiode junctions for the shorter wavelength green and blue pixels.
A further object of this invention is to provide a process that is compatible with the standard CMOS processes to form CMOS color imagers for a more cost-effective manufacturing process.
In accordance with the above objects, a method for fabricating photodiodes for CMOS color imagers with a more uniform spectral response for the red, green, and blue pixel cells is now described. The method utilizes photodiodes with diffe
Tseng Chien-Hsien
Wang Ching-Chun
Wuu Shou-Gwo
Yaung Dun-Nian
Ackerman Stephen B.
Gebremariam Samuel A
Saile George O.
Taiwan Semiconductor Manufacturing Company
Thomas Tom
LandOfFree
Method for making spectrally efficient photodiode structures... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for making spectrally efficient photodiode structures..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for making spectrally efficient photodiode structures... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3129273