METHOD AND DEVICE FOR WAVELENGTH-SENSITIVE PHOTO- SENSING USING MULTIPLE PN-JUNCTIONS
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application claims benefit and priority from U.S. provisional application no. 60/561,152 filed April 12, 2004, the contents of which are incorporated herein by reference.
FIELD OF THE INVENTION
[0002] The present invention relates generally to photo-sensing, and more particularly to methods and devices for wavelength-sensitive photo-sensing using multip le pn-} uncti ons .
BACKGROUND OF THE INVENTION
[0003] Wavelength-sensitive photo-sensing (often referred to as color sensing) has applications in a wide range of fields such as medicine and biology, the food, printing, and cosmetics industries, and the like. For example, in the study of cells and tissues, it may be necessary to monitor or detect the transmission and absorption of light of a certain bandwidth by cells and tissues under study.
[0004] Color of light can be sensed using «-junction based photo-diodes. As is known, a/rø-junction can form a photo-diode. When light is absorbed in a depletion region around a reversely biased /rø-junction, free charge carriers are generated due to photoelectric effect. The photo-generated carriers can be detected by sensing either a current or a voltage across the pn-] uncti on. The sensed current or voltage can indicate the intensity of light absorbed in the depletion region. Since longer wavelength light can penetrate deeper below the surface than shorter wavelength light, different wavelength bands of an incident light can be detected by positioning multiple j_w-junctions at different depths in a semiconductor. In a spectrometric approach, the spectrum of light is divided into many narrow bands. In a colorimetric approach, the color of light is resolved
into three primary color components, such as blue, green and red (RGB), which penetrate increasingly deeper into a semiconductor. Thus, the color of light can be detected using a sensor having three or more appropriately positioned pn-] unctions.
[0005] U.S. Patent No. 5,965,875 to Richard B. Merrill, entitled "Color separation in an active pixel cell imaging array using a triple-well structure", which is incorporated herein by reference (referred herein as "Merrill I") discloses a known color sensing device. Each photo-sensing cell or pixel of the device includes three well-shaped pn- junctions (referred to as a triple-well structure) formed in a semiconductor substrate. Each junction is at an appropriate depth so that the three spectral components of incident light can be detected by detecting currents from the respective /w-junctions. An array of pixels is formed in the device. The pixels are isolated by doped well regions. However, this device has some shortcomings. For example, the triple-well structures are formed by diffusion. As is known, it can be difficult to form a well-defined abrupt /rø-junction by diffusion, particularly when it is deep. Thus, a color sensor based on the triple-well structure may have poor performance.
[0006] Moreover, the performance of such a device can be negatively affected by interference between adjacent /w-junctions in a cell or between adjacent cells. Such a cross-junction or cell interference can occur between two adjacent and closely disposed /7«-junctions due to movement of free carriers from an area near one of the pn-] unctions to an area near the other /w-junction. Further, it may be desirable that the size of each pixel in a sensing array is small and that the pixels are well isolated. It can be difficult to achieve good isolation and a small pixel size with triple-well structures. Manufacture limitations or other performance considerations may also make it difficult to reduce the size of a triple-well structure.
[0007] U.S. Patent No. 6,632,701 to Richard B. Merrill, entitled "Vertical color filter detector group and array", which is incorporated herein by reference, discloses a device in which the vertically-stacked /w-junctions are isolated by buried layers. In this device, the vertically stacked junctions in each pixel are well isolated. However, this device still suffers some of the shortcomings discussed above - it can be difficult to keep adjacent
pixels well isolated while limiting the size of each pixel. Further, it can be difficult and cumbersome to fabricate this device without advanced technology and specialized equipments.
[0008] Accordingly, there is a need for improved methods and devices for sensing color based on vertically-stacked «-junctions.
SUMMARY OF THE INVENTION
[0009] A photo-sensing cell can include one or more insulation regions for laterally isolating adjacent regions in the cell and/or for laterally isolating the cell from an adjacent cell. The photo-sensing cell can be used alone or in an array of cells such as forming light sensing pixels in a color sensing device.
[0010] An insulation region can provide good isolation between different regions in a cell or between adjacent cells in a sensing array even when the cell or cells are small in size and closely disposed. Since insulation regions can be made laterally small, such as by employing the shallow trench isolation (STl) technique, the color sensing pixels can have small sizes with low or no cross-interference.
[0011] A sensing cell can be formed primarily by epitaxial growth to achieve good control of layer thickness, doping level, doping profile, and very abrupt pn-} unctions with very well controlled features.
[0012] Accordingly, in an aspect of the present invention, a photo-sensing cell is provided. The photo-sensing cell has a substrate with a surface. A plurality of regions are formed in the substrate beneath the surface one below another and are alternatively doped of first and second conductive types, thus forming a plurality ofpw-junctions at different depths relative to the surface. The depths are chosen for detection of light of respective wavelength bands incident on the surface. A conductive region is formed in the substrate and extends from the surface to an associated one of the plurality of regions. One or more insulation regions are formed in the substrate, each extending from the surface proximate and along the conductive region for laterally isolating the conductive region from an adjacent depletion region.
[0013] In another aspect of the present invention, there is provided a photo-sensing cell comprising a plurality of semiconductor layers formed epitaxially one on top of another. An insulation layer having a surface is formed on the semiconductor layers. The semiconductor layers are alternately doped of first and second conductive types, thus forming a plurality of pn-] unctions at different depths relative to the surface of the insulation layer. The depths are selected for detection of light of respective predetermined wavelength bands incident on the surface. A plurality of conductors is also provided. Each conductor extends from one of the layers through the insulation layer for interconnection with an electrical interconnect.
[0014] In a further aspect of the present invention, there is provided a semiconductor device comprising one or more photo-sensing cells, each according to one of the two preceding paragraphs.
[0015] In yet another aspect of the present invention, there is provided a method of forming a semiconductor device. In this method, a plurality of layers are epitaxially grown on a wafer, one on top another and alternatively doped of first and second conductive types, thus forming a plurality of /rø-junctions. An insulation layer is formed on the plurality of layers. The insulation layer has a surface and the depths of the pn- junctions relative to the surface are respectively chosen for detection of light of different wavelength bands incident on the surface. A conductor is placed on each one of one or more of the plurality of layers. The conductor extends from the each layer through the insulation layer.
[0016] In another aspect of the present invention, there is provided another method of forming a semiconductor device. This method includes epitaxially growing, on a first layer of a first conductive type, a second layer of an opposite, second conductive type, thus forming a first /rø-junction. A first well region doped of the first conductive type is formed in the second layer, thus forming a seconds-junction. A second well region doped of the second conductive type is formed in the first well region, thus forming a third pn-} unction. The pn-} unctions are at respective depths relative to a surface of the second layer, chosen for detection of light of different wavelength bands incident on the
surface. The method further includes forming, in the second layer and away from the first well region, a conductive region doped of the first conductive type extending from the first layer through the second layer, and forming an isolation trench in the second layer proximate and along the conductive region for isolating the conductive region from a depletion region within in the second layer.
[0017] Other aspects and features of the present invention will become apparent to those of ordinary skill in the art upon review of the following description of specific embodiments of the invention in conjunction with the accompanying figures.
BRIEF DESCRIPTION OF THE DRAWINGS
[0018] In the figures, which illustrate, by way of example only, embodiments of the present invention,
[0019] FIG. IA is a schematic cross-sectional view of a photo-sensing cell, exemplary of an embodiment of the present invention;
[0020] FIG. IB is a schematic, partial cross-sectional view of a semiconductor device having a plurality of photo-sensing cells of FIG. IA;
[0021] FIG. 2 is a schematic, partial cross-sectional view of a semiconductor device having an exemplary implementation of the photo-sensing cell of FIG. IA;
[0022] FIG. 3 is a schematic cross-sectional view of another photo-sensing cell, exemplary of an embodiment of the present invention; and
[0023] FIGS. 4 to 6 are respectively schematic, partial cross-sectional views of semiconductor devices incorporating the exemplary photo-sensing cell of FIG. 3.
DETAILED DESCRIPTION
[0024] In this description, the letters "N" and "P" are used to indicate the conductive types of doped semiconductor regions. The relative doping concentrations are indicated with trailing signs "+" and "-". Alternatively, the terms "strongly doped", "normally
doped" and "weakly doped" are used to indicate the relative doping concentrations.
[0025] FIG. IA depicts schematically a photo-sensing cell 100 in a semiconductor device, exemplary of an embodiment of the present invention. Cell 100 includes a substrate 102 (partially shown) having a surface 104. A plurality of alternately doped regions 106, 108, 110 and 112 are formed in substrate 102 one below another beneath surface 104. Regions 106 and 108 are of «-type and regions 110 and 112 are ofp-type. Thus, three /w-junctions 114, 116 and 118 are formed at respective depths, d\, d2 and d3, relative to surface 104, where d\ < dι < d^. The depths are chosen so that incident light at three different wavelengths of surface 104 generates charge carriers in the depletion regions developed from the respective junctions in response to the light of the three different wavelength bands. For example, the depths may be chosen for detection of blue (B), green (G) and red (R) components of the incident light.
[0026] A conductive region 120 extends from region 118 to surface 104. Conductive region 120 can include a semiconductor more strongly doped than region 112.
[0027] Insulation regions 122A, 122B and 122C (also collectively and individually referred to as 122) each extends from surface 104 to region 118. Insulation region 122A extends proximate and along conductive region 120 to laterally isolate conductive region 120 from the depletion region developed from pn-] unction 116.
[0028] As illustrated in FIG. IB, which depicts two photo-sensing cells 100 arranged side by side, insulation regions 122B and 122C disposed on opposite ends of a cell 100 can laterally isolate each cell 100 from a neighbouring cell or cells. While only two cells 100 are depicted, it is understood that the number of photo-sensing cells formed on a device may vary. The photo-sensing cells may be arranged in a desirable pattern for imaging in one or two lateral dimensions.
[0029] Each insulation region 122 can include a trench filled with an insulating material such as undoped polysilicon or SiO2. For example, the trench can have a SiO2 coated wall and be filled with undoped polysilicon for easy fabrication and good insulation. One or two of insulation regions 122A, 122B and 122C may be omitted. For
example, when cell 100 is to be used alone, insulation regions 122B and 122C may be omitted. Each insulation region may extend deeper or shallower than is shown in FIG. IA, depending on the application. A deeper insulation region may provide better isolation.
[0030] Insulation regions 122 may be formed using a standard or modified Shallow Trench Isolation (STl) technique.
[0031] Photo-sensing cell 100 has a number of electrical interconnects 124 respectively in communication with regions 106, 108, 110, and 112 for detecting electrical signals respectively derived from the depletion regions developed around junctions 114, 116, and 118 so that incident light on surface 104 can be detected through electrical interconnects 124. Each electrical interconnect 124 may include a conductor such as a metal. For example, interconnects 124 may be made of aluminium.
[0032] Cell 100 can be formed using any suitable materials and semiconductor fabrication techniques. Exemplary techniques that can be used in the fabrication of cell 100 are disclosed in Hong Xiao, Introduction to Semiconductor Manufacturing Technology, Prentice Hall, 2001, which is incorporated herein by reference, and Merrill I and Merrill II, supra. Suitable materials and fabrication techniques in a particular application can be readily selected and implemented by persons skilled in the art, after reviewing this description and references cited herein.
[0033] For example, region 112 may be formed by epitaxial growth on a semiconductor wafer (not shown). Region 108 may also be formed by epitaxial growth on region 112. Each of regions 106, 110, 120 and 124 may be formed by diffusion, or by ion implantation followed with drive-in diffusion.
[0034] One of regions 108 and 112 may have a strongly-doped sub-region (not shown) formed at the interface of regions 108 and 112 between insulating trenches 122A and 122C, as will be discussed below, for limiting expansion of the depletion region developed from pn-] unction 118 within the sub-region. As can be appreciated, an electrical path is formed in the sub-region which has a lower resistivity than those of the
adjacent regions.
[0035] Other regions or layers such as light shields, antireflective layers and insulation layers may be formed on substrate 102 above surface 104. Such additional regions and layers can be readily determined by persons skilled in the art.
[0036] In operation, an applied voltage may reverse bias one or more of the pn- junctions 114, 116 and 118. A beam of light is incident on surface 104 and absorbed at different depths in substrate 102, depending on the wavelengths of light. Free carriers may be generated in a depletion regions developed from one of pn-} unctions 114, 116 and 118 in response to incident light and produce an electrical signal such as a current or voltage in the electrical interconnects 124 in communication with the depletion region. The output signal is dependent on the amount of light absorbed in the depletion region. Since there are three depletion regions at three different depths, three different spectral components of the incident light can be detected.
[0037] The applied biasing voltages can be chosen to control the size of the depletion regions, for example, to adjust the spectral responses of the depletion regions. The operation of cell 100 is similar to the triple-well photo-sensor disclosed in U.S. Patent No. 5,965,875, supra, which provides further details on operating a triple-junction photo- sensing cell. Thus, detailed operation of cell 100 will not be discussed herein.
[0038] Conveniently, because conductive path 120 is provided between region 112 and surface 104, region 108 can be formed by epitaxial growth on region 112. Moreover, since it is not necessary to form region 108 by diffusion, the lateral size (in the horizontal direction as depicted) of cell 100 can be small, as compared to the size of a conventional triple-well color sensing cell. Further, as a conductive path such as conductive region 120 is isolated from adjacent >«-j unctions extending along it by insulation regions 122, cross- interference can be limited even when the lateral size of cell 100 is small. Thus, cell 100 can be made more compact than a photo-sensing cell based on the triple-well structure and still have good performance. Cells 100 can be advantageously used in a photo- sensing array such as an imaging device as pixels, where cross-interference between adjacent pixels are limited by insulation regions 122B and 122C. The photo-sensing
array or imaging device can thus have a high density of small pixels, resulting in high resolution.
[0039] As can now be appreciated, regions 122B and 122C can extend deeper than shown in FIG. IA. For example, they may extend through region 112 for better insulation between adjacent cells 100. In addition, 122C can be located closer to 122A by cutting through well regions 106 and 110 to further reduce the later size of cell 100. It will be apparent to one skilled in the art that cell 100 has a simple structure and can be readily modified to suit various applications.
[0040] FIG. 2 illustrates an example physical embodiment of cell 100. Substrate 202 has a surface 204 and includes doped regions 206 (N+), 208 (N), 210 (P) and 212 (P~). The conductive region necessary for contacting region 212 includes a strongly doped contact region 220 (?*). The insulating regions include insulating trenches 222. The electrical interconnects include metal contacts 224.
[0041] The additional regions or layers include a strongly doped p-type substrate layer 230 (P+), a strongly doped w-type buried layer 232 (N*) which can be considered a sub-region of region 208, strongly doped contact regions 234 (βt) and very strongly doped contact regions 236 (N *) and 238 (P-1^) in substrate 202 each extending from surface 204 for interconnection with a respective electrical interconnect 224, and relatively strongly doped edge regions 240, the benefit of which will become clear below.
[0042] For example, the layers and regions may have the following respective doping concentrations: about 1E18 to 1E19 cm-3 for region 206; 1E15 to 1E16 cm-3 for region 208; 1E17 to 1E18 cm-3 for region 210; 1E14 to 1E15 cm-3 for region 212; 1E17 to 1E19 cm-3 for layer 230; IE 18 to IE 19 cm-3 for layer 232; and at least 1E16 to 1E17 cm-3 for edge regions 240.
[0043] Region 206 may be relatively thin, such as about 0.04 to 0.06 microns thick. As can be appreciated by one skilled in the art, because it is so shallow, region 206 can have high sheet resistance. In such a case, providing two contact regions 236 at the opposite ends of region 206 can guard region 206 to optimize its breakdown voltage and
reduce the travel distance of photo-generated carriers within region 206.
[0044] Layer 208 may be about 4 to 6 microns thick. Layer 232 may be thinner, for example, having a thickness of about 1.8 microns.
[0045] It should be understood that the doping levels and thicknesses of the layers and regions in cell 200 may vary depending on the particular application and will not always be specified. When specific values are given, they are exemplary and for illustration purposes only.
[0046] As alluded to earlier, layer 232 is more strongly doped than its adjacent regions to limit expansion of the depletion region developed from pn-} unction 218 within layer 232 and to form an electrical path in layer 232 which has a lower resistivity than those of the adjacent regions. Thus, photo-generated charge carriers can be more effectively collected from region 208.
[0047] As can be understood, cell 200 can be fabricated in various suitable manners. For example, cell 200 may be formed by first providing a strongly doped p-type wafer to form layer 230. The wafer may have any suitable size and thickness depending on the application. A relatively weakly doped p-type layer is epitaxially grown on layer 230 to form layer 212. A strongly doped w-type layer is implanted in a selected region in layer 212 to form buried layer 232. Due to the nature of the implantation technique, the implanted region may extend into layer 212, as depicted in FIG. 2. A weakly doped n- type layer is epitaxially grown on layers 212 and 232 to form layer 208. A shallow, doped p-type well is formed in layer 208 by diffusion to form region 210. A strongly doped n- type well is formed in region 210 by diffusion to form region 206. Region 210 has edge regions 240, which are also ofp-type and are formed by a field implantation technique. The edge regions can prevent accidental field-induced inversion at surface 204.
[0048] Strongly doped contact regions such as contact regions 220, 234, 236 and 238 may be formed by diffusion with materials of appropriate conductive type and may extend to desired depths in a given application. Each strongly doped contact region may have a doping concentration higher than the surrounding region. For example, the deep
contact regions 220 and 234 may each have an average doping concentration of about
1E18 to 1E19 cm .-"3 and the shallow contact regions 236 and 238 may each have a surface doping concentration of about 1E20 cm-3. As can be understood, the depth and doping level of each contact region depend on its purpose. For contacting low-lying regions (such as regions 208 and 212), the contact regions (such as contact regions 220 and 234) may be deeper. It is not necessary for a contact region to have a uniform doping concentration. For example, a deeper portion (e.g. regions 220 and 234) of a contact region may have a lower doping concentration than a shallower portion (e.g., region 236 above region 234, and region 238 above region 220) of the contact region. The higher doping of the shallower portion allows good ohmic contact with a metal interconnect and the lower doping of the deeper portion allows for easier fabrication, such as by diffusion.
[0049] Conductive contact region 220 is formed away from well region 210 and extends through region 208.
[0050] Insulating trenches 222 are formed in layer 208 by a STl technique or any other such similar method. Each insulating trench 222 may be formed by first etching a trench in layer 208, oxidizing the trench wall to form a SiO2 layer 242, and filling the trench with a filler 244 such as polySi. Filling with polySi may be advantageous because relatively thick polySi can be quickly deposited. Furhter, polysilicon has a coefficient of thermal expansion very close to that of monocrystalline silicon, which is typically used for forming the semiconductor layers and regions. Therefore, mechanical stress in substrate 202, which can be induced by different thermal expansion of different regions, can be reduced. Layer 242 may provide electrical insulation as well as make the surface the trench wall smooth. Trench 222 may be first etched using a dry etching technique and followed by a series of short, wet etching steps. After each etching step, oxidation growth may be performed to remove sharp edges. As can be appreciated, sharp edges may be undesirable because they produce strong electrical fields. After fabrication of trenches 222, surface 204 may be planarized using a suitable technique such as Chemo- Mechanical Planarization (CMP).
[0051] In some applications, to minimize the lateral size of cell 200, trenches 222
may be formed before any diffusion is performed, for example, to form conductive region 220 or contact region 234. The trenches 222 can be positioned to limit the lateral diffusion. In some other applications, trenches 222 may be formed late in the fabrication process. When formed late in the process, trenches 222 can still be positioned to limit cell size to a certain degree, for example, by "cutting through" regions formed by diffusion.
[0052] The metal contacts 224 are formed using a suitable metal such as an aluminium-based alloy including AlSiCu.
[0053] Additional layers and electronic structures including a protective cover (not shown) may be formed on surface 204. For example, the cover may include an insulation layer, an antireflective (AR) filter for preventing internal reflection of light, and a metal shield for shielding external light. The cover may also have a window area with a reduced thickness such as about 100 A or thinner, which is above and in substantial alignment with/w-junctions 214, 216, and 218 to allow more light to pass through. The window area should be made of a suitable material such as metal-oxide silicon (MOS) gate oxide, to not only provide good insulation and but also minimize generation-recombination phenomena at the Si-SiO2 interface.
[0054] A further exemplary photo-sensing cell 300, exemplary of another embodiment of the present invention is schematically illustrated in FIG. 3. Photo- sensing cell 300 includes a substrate 302. Substrate 302 includes a plurality of semiconductor layers 306, 308, 310 and 312 formed one on top of another, which include w-type regions 306 and 308 and p-type regions 310 and 312. Substrate 302 also has an insulation layer 322 formed on layers 306, 308, 310 and 312. Insulation layer 322 has a surface 304. The three pn-} unctions 314, 316 and 318 are formed at respective depths, d\, d and < ., relative to surface 304, where d\ < d2 < d^, chosen as in cell 100.
[0055] Each of layers 306, 308, 310 and 312 has an exposed top portion in contact with insulation layer 322. Conductors 320A-320D (also referred to collectively or individually as 320) each extends from one of layers 306, 308, 310 and 312 to surface 304. Each conductor 320 defines a conductive region and can be formed of a metal such as aluminium, copper, or the like. Conductors 320 form part of electrical interconnects in
communication with regions 306, 308, 310 and 312, respectively, for detecting electrical signals respectively derived from the depletion regions so that incident light on surface 304 can be detected through conductors 320.
[0056] Insulation layer 322 laterally isolates conductors 320 from each other and from the semiconductor layers, as well as isolating cell 300 from its environment. Insulation layer 122 can be formed of an insulating material such as SiO2. As can be appreciated, a conductor 320 does not have to extend to surface 304, and may otherwise extend through insulation layer 322 for interconnection with an electrical interconnect. For example, in an alternative embodiment, conductor 320D may turn toward the right hand side.
[0057] Cell 300 can also be formed using any suitable materials and semiconductor fabrication techniques. Suitable materials and fabrication techniques in a particular application can be readily selected and implemented by persons skilled in the art, after reviewing this description and references cited herein. In particular, layers 306, 308, 310 and 312 may be formed by epitaxial growth followed by exposing top portions such as by etching so that insulation layer 322 can be next formed on the exposed portions.
[0058] Cell 300 can be operated in a similar manner as for cell 100 to detect wavelength components of light.
[0059] Cell 300 may have certain advantages over cell 100. For example, the doping profiles in the regions of cell 300 can change more quickly than in cell 100 and thus can form much more abrupt n-junctions. In general, cell 300 can be realized to have regions with well controlled properties such as thickness and doping profile. Further, it is not necessary to form insulating trenches in cell 300, which may be time-consuming and expensive to form. One or more additional »-junctions can be conveniently formed, if desired, without significantly increasing the size of the cell, such as by forming additional semiconductor layers on top of layer 306.
[0060] As can be appreciated, it is not necessary that all of the semiconductor layers 306, 308, 310 and 312 are exposed to insulation layer 322. For example, only the top two
layers 306 and 310 may be exposed. The electrical contacts for layers 308 and 312 can be provided from the side and the backside, respectively.
[0061] FIG. 4 illustrates an example physical embodiment 400 of cell 300. A substrate 402 includes a wafer 430 (P*) strongly doped of p-type. To form vertically aligned p«-junctions, a plurality of alternatively doped layers is formed in substrate 402 on wafer 430, including doped w-regions 406 (N ") and 408 (NT to N), and doped p- regions 410 (P to P ) and 412 (P_). As indicated, region 406 is strongly doped, regions 408 and 410 have different subregions with different doping concentrations, and region 412 is normally doped. Region 408 includes a buried layer 432 (N ) more strongly doped than adjacent regions. Regions 410 and 412 respectively include contact regions 438 (P^ more strongly doped than adjacent regions. An insulation layer 422, which can be made of SiO2, is formed on top of the doped layers. Layer 422 has a portion 442 with reduced thickness at a window area 444. Conductors 420 respectively extend from the doped layers through insulation layer 422 for electrical interconnect. Each of regions 406, 408, 410, and 412 is in electrical communication with two contactors 420. Conductors 420 can be made of metal.
[0062] Again, a protective cover and/or other electronic structures (not shown) may be formed on substrate 402.
[0063] The layers and regions may have the exemplary doping concentrations and thicknesses list in Table I.
TABLE I.
[0064] As mentioned above, regions 408 and 410 each has different sub-regions with different doping concentrations. The more strongly doped lower portion in each region can limit undesirable extension of the depletion region developed from thepw-junction below and allow increased conductivity within the lower half for more effective collection of photogenerated carriers.
[0065] Cell 400 can be similarly fabricated as cell 100, 200, or 300 with some modifications that will be apparent to persons skilled in the art. In particular, doped layers may be epitaxially grown and then etched (such as by dry etching) to form regions 406, 408, 410 and 432, so that a portion of each layer is exposed. Contact regions 438 may be formed by diffusion or implantation. It may be convenient to form all contact regions 438 in a single implantation drive-in sequence. A SiO layer 422 can be deposited on top of regions 406, 410, 412 and 432 after they have been formed. Layer 422 can be planarized using a suitable planarization technique such as CMP. The thick portions of layer 422 may be formed of field oxide (FOX) while thin portion 442 may be formed of MOS gate oxide. Thin portion 442 may be about 100 A thick. Layer 422 can be etched to form window area 444. Openings in layer 422 can be made and conductors 424 can be placed on each one of regions 406, 410, 412 and 432 through the openings with a suitable metallization technique. Contact regions 438 may be formed before forming layer 422 or after etching the openings.
[0066] A protective cover (not shown) may be formed on layer 422 in a suitable manner, as discussed above.
[0067] FIG. 5 illustrates a possible cell 500, a variation of cell 400. The difference between cell 500 and cell 400 is that the region 502 below the bottom p«-j unction 504 is
isolated by isolation trenches 506 and SiO2 insulating layer 508. Cell 500 may be formed on a silicon-on-insulator (SOI) wafer, which has layer 508 and a semiconductor layer 510. Thus, each trench 506 extends between insulation layers 508 and 522 for laterally isolating regions on its opposite sides from each other. Trenches 506 may be formed before buried layers 512 and contact regions 514 have been formed by diffusion. As can be appreciated, when region 502 is formed on top of a SOI wafer, region 502 can have a very different doping concentration or different doping type than that of layer 510. Layer 510 can be weakly, normally, or strongly doped, while region 502 can still have a doping concentration within a wide range, including one that is much different from that of region 412 in cell 400. For example, region 502 may include a thin but very strongly doped bottom sub-region and a much thicker but very weakly doped sub-region above it.
[0068] Cells 500 can form pixels in a photo-sensing device with good electrical insulation between neighboring pixels.
[0069] FIG. 6 illustrates another possible variation of cell 400. A difference between cell 600 and cell 400 is that the contact 602 for the bottom p-type region 604 is below and in contact with the contact region 606 below region 604. Contact region 606 is strongly doped. Consequently, a buried layer 608 can extend over the full width of cell 600 and the number of contact regions formed by diffusion and the number of openings in the top insulation layer can be reduced. Cell 600 may be advantageous. It may be easier to fabricate. In some applications, it may be permissible or desirable that electrical contact be made from the backside (i.e. the bottom side in FIG. 6) of the photo-sensing cell. For example, as can be seen by comparing FIGS. 4 and 6, the lateral size of cell 600 may be reduced in comparison to cell 400 because fewer contacts are required at the front side.
[0070] As can be understood, the exemplary embodiments described herein may be modified. For example, the number of vertically-stacked pn-} unctions may be increased. The doping types of the layers or regions in a photo-sensing cell may be reversed. The operation of such a reversed cell is similar to the operation of the original cell, except that all polarities, biasing voltages or current directions may also be reversed.
[0071] Each of the photo-sensing cells described herein can be used either alone or in
an array of cells. In an array of cells, the cells may be advantageously isolated from each other by the isolation regions such as isolation trenches or insulation layers.
[0072] Other features, benefits and advantages of the embodiments described herein not expressly mentioned above can be understood from this description and the drawings by those skilled in the art.
[0073] Of course, the above described embodiments are intended to be illustrative only and in no way limiting. The described embodiments are susceptible to many modifications of form, arrangement of parts, details and order of operation. The invention, rather, is intended to encompass all such modification within its scope, as defined by the claims.