Dual-mode image sensor with a signal-separating color filter array, and method for same

Abstract

A method for fabricating a signal-separating CFA includes forming a multi-height CFA on a substrate. The multi-height CFA includes a plurality of tall spectral filters and a plurality of short spectral filters. Each of the tall spectral filters is taller than each of the short spectral filters. The method also includes disposing a spectral-blocking layer on the multi-height CFA, and planarizing the spectral-blocking layer to expose a top surface of each of the plurality of tall spectral filters.

Claims

1. A method for fabricating a signal-separating color filter array (CFA), the method comprising: forming a multi-height CFA on a substrate, the multi-height CFA comprising a plurality of short spectral filters and a plurality of tall spectral filters, each tall spectral filter being taller than any short spectral filter; disposing a spectral-blocking layer on the multi-height CFA; and planarizing the spectral-blocking layer to expose a top surface of each of the plurality of tall spectral filters.

2. The method of claim 1, further comprising forming a microlens array on the signal-separating CFA to form an image sensor with a signal-separating color filter array.

3. The method of claim 1, the step of forming the multi-height CFA comprising: depositing a photoresist layer on a uniform-height CFA, the uniform-height CFA comprising the plurality of tall spectral filters and a plurality of etch-pending spectral filters; exposing the photoresist layer through a photomask; and etching the plurality of etch-pending spectral filters to form the plurality of short spectral filters.

4. The method of claim 1, the step of forming the multi-height CFA comprising: forming the plurality of short spectral filters on a surface of the substrate; depositing a spectral filter layer upon each of the short spectral filters and upon exposed substrate regions adjacent to each of the plurality of short spectral filters, the spectral filter layer having a height exceeding a height of each short spectral filter, the respective heights being relative to the surface; and removing regions of the spectral filter layer disposed upon each of the plurality of short spectral filters.

5. The method of claim 4, the step of removing comprising: depositing a photoresist layer on the spectral filter layer; exposing the photoresist layer through a photomask, the photomask comprising a plurality of apertures each aligned with a respective one of the plurality of short spectral filters; developing the exposed photoresist regions to remove photoresist disposed on each short spectral filter; and etching the spectral filter layer to yield the plurality of tall spectral filters.

6. The method of claim 1, the step of disposing further comprising forming a thickness of the spectral-blocking layer on each of the plurality of short spectral filters such that the combined height of a short spectral filter and spectral blocking layer formed thereon equals or exceeds the height of any tall spectral filter adjacent to the short spectral-blocking filter.

7. The method of claim 1, the step of disposing comprising depositing the spectral-blocking layer using a thin-film deposition technique.

8. The method of claim 1, the step of planarizing comprising leveling the spectral-blocking layer using a chemical mechanical planarization process.

9. The method of claim 1, the step of planarizing comprising blanket-etching the spectral-blocking layer.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) FIG. 1 shows representative transmission spectra of visible and IR spectral filters used in CMOS image sensors.

(2) FIG. 2 is a cross-sectional view of a representative prior-art VIS-IR image sensor 200 with reduced IR contamination.

(3) FIG. 3 is a cross-sectional view of a dual-mode image sensor with a signal-separating CFA for visible and IR imaging, in an embodiment.

(4) FIG. 4 shows representative transmission spectra of blue, green, red, and IR spectral filters of a signal-separating CFA, in an embodiment.

(5) FIG. 5 is a flowchart illustrating a method for fabricating a signal-separating CFA, in an embodiment.

(6) FIG. 6 shows cross-sectional views of CFAs on a substrate, the views corresponding to steps of the method of FIG. 5, in an embodiment.

(7) FIG. 7 is a flowchart illustrating a first method for fabricating a multi-height CFA, in an embodiment.

(8) FIG. 8 shows cross-sectional views of CFAs on a substrate, the views corresponding to steps of the method of FIG. 7, in an embodiment.

(9) FIG. 9 is a flowchart illustrating a second method for fabricating a multi-height CFA, in an embodiment.

(10) FIG. 10 shows cross-sectional views of CFAs on a substrate, the views corresponding to steps of the method of FIG. 9, in an embodiment.

(11) FIG. 11 shows cross-sectional views of CFAs on a substrate, the views corresponding to steps of the method of FIG. 9, in an embodiment.

DETAILED DESCRIPTION

(12) FIG. 3 is a cross-sectional view of an exemplary dual-mode image sensor 300 with a signal-separating CFA 334 for visible and IR imaging. Dual-mode image sensor 300 includes an array of pixel microlenses 316 disposed on signal-separating CFA 334, which is disposed on a substrate 320. Microlenses 316 are transparent to visible and IR light, and may be formed of materials known in the art. Microlenses 316 may be formed by a reflow process, an embossing replication process, or other methods known in the art.

(13) Signal-separating CFA 334 includes visible spectral filters 314 and IR-pass filters 391. Substrate 320 includes photodiode regions 322 and is formed of a semiconductor such as silicon. Substrate 320 may include layers and regions of different materials without negating its function as a substrate for filters 314 and 391.

(14) Herein, the terms spectral filter and color filter are synonymous. A spectral filter with a pass band in the visible portion of the electromagnetic spectrum is called a visible spectral filter or visible filter. Similarly, a spectral filter with a pass band in the IR portion of the electromagnetic spectrum is called an IR spectral filter or IR filter. An array of spectral filters containing both visible and IR spectral filters is referred to as a CFA.

(15) The CFA pattern of signal-separating CFA 334 may be a modified Bayer pattern where half of the green filters are replaced by IR-pass filters 391. Candidates for IR-pass filters 391 include Color Mosaic materials by Fujifilm Corporation, such as the IR pass filter described in WIPO publication numbers WO2014084147 and WO2014103628.

(16) As FIG. 3 is a cross-sectional view along a Bayer pattern row or column, visible spectral filters 314 are either all blue filters or all red filters, depending on the particular row or column corresponding to the cross-sectional view of FIG. 3. Other CFAs using primary-color filters (RGB), such as those described in by R. Lukac (IEEE Trans. Consum. Electr., 51, 4, 2005), may be modified to incorporate IR spectral filters without departing from the scope hereof. In such an embodiment, each visible spectral filter 314 in the cross-sectional view of FIG. 3 need not be the same type of filter, but may be one of a red, a green, or a blue spectral filter. Similarly, CFAs that include secondary-color filters (CMY), visibly clear filters, primary color filters, and combinations thereof may be modified to incorporate IR spectral filters without departing from the scope hereof.

(17) Dual-mode image sensor 300 also includes pixel-localized IR-cut filters 390 between visible spectral filters 314 and their corresponding pixel microlenses 316. Pixel-localized IR-cut filters 390 are for example formed of a single-layer of material. Pixel-localized IR-cut filters 390 may be formed of materials known in the art, such as Color Mosaic materials by Fujifilm Corporation. Candidate IR-cut filter materials described in the patent literature include materials containing one or more of a resin, a UV-curable resin, a metal oxide, and dyes, such as those described in WIPO publication numbers WO2014104136 and WO2014104137 by Fujifilm Corporation, U.S. Pat. No. 5,807,511 by Kunimatsu and Yamazaki, and U.S. Pat. No. 6,191,884 by Takizawa and Takahashi. For clarity of illustration, not all visible spectral filters 314, pixel microlenses 316, photodiode regions 322, or pixel-localized IR-cut filters 390 are labeled in FIG. 3.

(18) An appropriate height 340 of pixel-localized IR-cut filters 390 may be determined by the Beer-Lambert law: I(z)=I.sub.0e.sup.4n.sup.i.sup.z/.sup.0, where I(z) is the light intensity after propagating a distance z in the filter, I.sub.0 is the incident light intensity, n.sub.i is the imaginary part of the IR-cut filter's refractive index, and .sub.0 is the free-space light wavelength.

(19) Each of IR-pass filters 391 has an equal height 341 such that the top surfaces of the IR-pass filters are coplanar. Height 341 exceeds a height 344 of visible spectral filters 314. IR-cut filters have height 340 such that the sum of height 344 of visible spectral filters 314 and height 340 of pixel-localized IR-cut filters 390 is substantially equal to (i.e., between 90% and 110% of) the height 341 of IR-pass filters 391. In an embodiment, height 341 is between 1.0 m and 1.4 m and height 344 is between 0.5 m and 0.7 m. In an embodiment, the heights of visible spectral filters 314, pixel-localized IR-cut filters 390, and IR-pass filters 391, may be 0.7 m, 0.5 m, and 1.2 m, respectively.

(20) Within signal-separating CFA 334, the heights 340 and 344 may vary while satisfying the following two constraints: For a given visible spectral filter 314 (with height 344) and IR-cut filter 390 (with height 340) disposed thereon, (1) the sum of heights 344 and 340 is substantially equal to IR-pass filter height 341, and (2) height 341 exceeds height 344.

(21) FIG. 4 shows representative transmission spectra 445, 455, 465, and 491 of blue, green, red, and IR spectral filters, respectively of a signal-separating CFA 334. At visible wavelengths the transmission spectra 445, 455, and 465 resemble transmission spectra 145, 155, and 165 of FIG. 1, respectively. Beyond 0.8 m, pixel-localized IR-cut filters 390 suppress transmission of the blue, green, and red filters, as shown by transmission spectra 445, 455, and 465, respectively.

(22) FIG. 5 is a flowchart illustrating an exemplary method 500 for fabricating a signal-separating CFA. FIG. 6 shows cross-sectional views of CFAs on a substrate, the views corresponding to steps of method 500. FIG. 5 and FIG. 6 are best viewed together with the following description.

(23) In step 502, method 500 forms a multi-height CFA on a substrate. The multi-height CFA includes a plurality of tall spectral filters and a plurality of short spectral filters. Each tall spectral filter is tall by virtue of having a height that exceeds any height of a short spectral filter. In an example of step 502, method 500 forms a multi-height CFA 624 on a substrate 620, as shown in cross-sectional view 602, FIG. 6.

(24) Multi-height CFA 624 includes visible spectral filters 614, each with a height 644, and IR-pass filters 691, each with a height 641 that exceeds height 644. Visible spectral filters 614 are similar to visible spectral filters 314. Substrate 620 is similar to substrate 320, and includes photodiode regions 622. Each color filter of multi-height CFA 624 is aligned to, e.g., centered above, a respective photodiode region 622.

(25) Step 502 may involve methods and materials used for forming CFAs known in the art. These methods and materials include those described in the following patents: U.S. Pat. No. 4,239,842 to Sandhu, U.S. Pat. No. 4,565,756 to Needs and Blok, U.S. Pat. No. 5,059,500 to Needham, Chiulli, and Clark, U.S. Pat. No. 5,268,245 to Chiulli, U.S. Pat. No. 5,321,249 to Nomura, and U.S. Pat. No. 6,004,704 to Ko. Spectral filter materials for image sensors may also include Color Mosaic materials by Fujifilm Corporation.

(26) In step 504, method 500 disposes a spectral-blocking layer on the multi-height CFA. In an example of step 504, method 500 disposes an IR-cut layer 670 on the multi-height CFA 624, as shown in cross-sectional view 604, FIG. 6. Before the planarization of step 506, IR-cut layer 670 has a minimum height 671 above a substrate top surface 621 that equals or exceeds height 641 of IR-pass filters 691. IR-cut layer 670 has, for example, a stop band at wavelengths between 0.75 m and 1.1 m and is formed of the same material as pixel-localized IR-cut filters 390, FIG. 3. Step 504 may employ a thin-film deposition technique known in the art.

(27) In step 506, method 500 planarizes the spectral-blocking layer to expose top surfaces of the tallest spectral filters in the multi-height CFA and form pixel-localized spectral-blocking filters within a signal-separating CFA. In an example of step 506, method 500 planarizes IR-cut layer 670 to expose top surfaces 681 of IR-pass filters 691 and form pixel-localized IR-cut filters 690 within a signal-separating CFA 634, as shown in cross-sectional view 606. Signal-separating CFA 634 is an embodiment of signal-separating CFA 334. Each IR-pass filter 691 has a top surface 681, of which just one is shown in cross-sectional view 606 for clarity. Step 506 may be performed with blanket-etching, chemical mechanical planarization (a.k.a. chemical mechanical polishing), or other planarization methods known in the art.

(28) In step 506, exposing the surfaces of the tallest filters in the signal-separating CFA may result in the IR-cut filters having a non-planar top surface without departing from the scope hereof. For example, in cross-sectional view 606, pixel-localized IR-cut filter 690(1) has a concave top surface 680(1), whereas the other pixel-localized IR-cut filters 690 each have a flat top surface, for example, a top surface 680(2).

(29) In optional step 507, method 500 deposits a transparent separating layer on the top surface of the signal-separating CFA. The transparency of the separating layer to visible and IR light is similar to that of the microlenses 316. In an example of step 507, method 500 deposits a transparent separating layer, not shown in FIG. 6, on top surfaces 680 and 681 of signal-separating CFA 634. The transparent separating layer may be considered part of signal-separating CFA 634.

(30) In optional step 508, method 500 forms a microlens array on the signal-separating CFA to form a dual-mode image sensor with a signal-separating color filter array. Each microlens of the microlens array is axially aligned to a respective individual filter of the signal-separating CFA. In an example of step 508, method 500 forms an array of microlenses 616 on signal-separating CFA 634, which results in a dual-mode image sensor 608. Microlenses 616 are similar to microlenses 316 of FIG. 3.

(31) In an embodiment of method 500 that includes optional step 507, the transparent separating layer separates the bottom surfaces of microlenses 616 and top surfaces 680 and 681. The transparent separating layer may function to optimize the focus of each microlens 616 with respect to each respective photodiode region 622.

(32) Dual-mode image sensor 608 and signal-separating CFA 634 are similar to dual-mode image sensor 300 and signal-separating CFA 334, respectively. Each microlens 616 is an axially aligned to a respective individual spectral filter 614 or 691 of signal-separating CFA 634.

(33) FIG. 7 is a flowchart illustrating a first exemplary method 700 for fabricating a multi-height CFA, for example, the multi-height CFA of method 500, step 502, and multi-height CFA 624, FIG. 6. Method 700 is involves photolithographically processing a uniform-height CFA. A uniform-height CFA is a standard CFA used in the art with color filters of equal height, to within fabrication tolerances. FIG. 8 shows cross-sectional views of CFAs on a substrate, the views corresponding to steps of method 700. FIG. 7 and FIG. 8 are best viewed together with the following description.

(34) In step 702, method 700 deposits a photoresist layer on a uniform-height CFA. For the purpose of describing method 700, the color filters within the uniform-height CFA are categorized into two groups: tall filters and etch-pending filters.

(35) In an example of step 702, method 700 deposits a photoresist layer 830 on a uniform-height CFA 823. Uniform-height CFA 823 is disposed on a substrate 820, as shown by cross-sectional view 802, FIG. 8. Uniform-height CFA 823 includes two categories of filters: tall spectral filters 891 and etch-pending spectral filters 815. Substrate 820 is similar to substrate 320, FIG. 3.

(36) In step 704, method 700 exposes the photoresist layer through a photomask, the photomask blocking photoresist regions disposed on the tall filters and having apertures aligned with the etch-pending filters. In an example of step 704, method 700 exposes photoresist layer 830, resulting in exposed photoresist regions 831 covering etch-pending spectral filters 815, as shown by cross-sectional view 804.

(37) In step 706, method 700 develops the exposed photoresist regions to remove exposed photoresist regions on each of the etch-pending filters. In an example of step 706, method 700 develops exposed photoresist regions 831 to remove photoresist regions on each of the etch-pending spectral filters 815, as shown by cross-sectional view 806.

(38) In step 708, method 700 etches the etch-pending filters to yield a plurality of short spectral filters. In an example of step 708, method 700 etches etch-pending spectral filters 815 to a height 844, which results in short spectral filters 814, as shown in cross-sectional view 808. Height 844 is smaller than a height 841 of each tall spectral filter 891.

(39) In step 710, method 700 removes remaining photoresist layer from the tall spectral filters. In an example of step 710, method 700 removes remaining photoresist layer 830 from tall spectral filters 891, as shown in cross-sectional view 810. Short spectral filters 814 and tall spectral filters 891 constitute a multi-height CFA 824.

(40) FIG. 9 is a flowchart illustrating a second exemplary method 900 for fabricating a multi-height CFA. FIG. 10 shows cross-sectional views of color filter arrays on a substrate, the views corresponding to steps of method 900. FIG. 9 and FIG. 10 are best viewed together with the following description.

(41) In step 902, method 900 forms a plurality of short spectral filters on a substrate. In an example of step 902, method 900 forms an array of short spectral filters 1014 on a substrate 1020, as shown in cross-sectional view 1002, FIG. 10. Spectral filters 1014 are designated as short to distinguish them from tall filters, formed in subsequent steps of this method, which are taller than short spectral filters 1014.

(42) In cross-sectional view 1002, distances 1071 between adjacent short spectral filters 1014 are equal or substantially equal to a width 1074 of each short spectral filter 1014. Subsequent steps of method 900 will form spectral filters between adjacent short spectral filters 1014.

(43) In step 904, method 900 deposits a spectral filter layer upon each of the plurality of short spectral filters and upon exposed substrate regions adjacent to each of the plurality of the short spectral filters. The height of the spectral filter layer exceeds the height of each short spectral filter, where each height is relative to the plane of the substrate.

(44) In an example of step 904, method 900 deposits a spectral filter layer 1081 upon each short spectral filters 1014 and upon exposed regions of substrate 1020 adjacent to a short spectral filter 1014, as shown by cross-sectional view 1004. Spectral filter layer 1081 has a height 1041 that exceeds a height 1044 of each short spectral filter 1014. In cross-sectional view 1004, selected regions of spectral filter layer 1081 disposed on short spectral filters 1014 are labeled as filter regions 1084.

(45) In step 905, method 900 removes regions of the spectral filter layer disposed upon each of the plurality of short spectral filters. In an example of step 905, method 900 removes regions 1084 of spectral filter layer 1081 disposed upon each of the plurality of short spectral filters 1014 to form tall spectral filters 1091 as shown in cross-sectional view 1012. Short spectral filters 1014 and tall spectral filters 1091 constitute a multi-height CFA 1024.

(46) In an embodiment of method 900, step 905 includes steps 906, 908, 910, and 912, described below. FIG. 11 shows cross-sectional views of color filter arrays on a substrate, the views corresponding to steps 906, 908, and 910. FIG. 9 and FIG. 11 are best viewed together with the following description.

(47) In step 906, method 900 deposits a photoresist layer on the spectral filter layer. In an example of step 906, method 900 deposits a photoresist layer 1030 on spectral filter layer 1081, as shown by cross-sectional view 1106, FIG. 11.

(48) In step 908, method 900 exposes the photoresist layer through a photomask, the photomask having apertures aligned with the short spectral filters. In an example of step 908, method 900 exposes photoresist layer 1030, resulting in exposed photoresist regions 1031 covering respective short spectral filters 1014, as shown by cross-sectional view 1108.

(49) In step 910, method 900 develops the exposed photoresist regions to remove photoresist regions above each of the short spectral filters. In an example of step 910, method 900 develops exposed photoresist regions 1031 to remove photoresist regions above each of the short spectral filters 1014, as shown by cross-sectional view 1110.

(50) In step 912, method 900 etches the spectral filter layer to yield a plurality of tall spectral filters. In an example of step 912, method 900 etches spectral filter layer 1081 to yield a plurality of tall spectral filters 1091, as shown by cross-sectional view 1012, FIG. 10. Short spectral filters 1014 have a height 1044, which is smaller than height 1041 of tall spectral filters 1091. Tall spectral filters 1091 and short spectral filters 1014 constitute a multi-height CFA 1024.

(51) Changes may be made in the above methods and systems without departing from the scope hereof. It should thus be noted that the matter contained in the above description or shown in the accompanying drawings should be interpreted as illustrative and not in a limiting sense. The following claims are intended to cover all generic and specific features described herein, as well as all statements of the scope of the present method and system, which, as a matter of language, might be said to fall there between.