Low noise InGaAs photodiode array

Abstract

A photodiode pixel structure for imaging short wave infrared (SWIR) and visible light built in a planar structure and may be used for one dimensional and two dimensional photodiode arrays. The photodiode arrays may be hybridized to a read out integrated circuit (ROIC), for example, a silicon complementary metal-oxide-semiconductor (CMOS) circuit. The photodiode in each pixel is buried under the surface and does not directly contact the ROIC amplification circuit. Charge is transferred form the detector using a junction field effect transistor (JFET) in each pixel. Disconnecting the photodiode from the ROIC amplification circuit enables low dark current as well as double correlated sampling in the pixel.

Claims

1. A focal plane array comprising: a photodiode array that includes a plurality pixels, each of which includes: a buried photodiode, comprising a p-type region of an In.sub.xGa.sub.1-xAs layer and at least one cap layer, wherein the p-type region is doped by diffusing or ion-implanting a p-type dopant through the at least one cap layer, and wherein the p-type region is buried underneath an n-type region such that the p-type region does not contact the surface of the at least one cap layer; and a first junction field effect transistor (JFET) proximate to the buried photodiode and operative to remove charge from the buried photodiode; and a second JFET proximate to the first JFET and operative to remove charge from the first JFET; and a read out integrated circuit (ROIC) including an amplification and storage circuit, wherein ROIC is hybridized to the photodiode array, and wherein the ROIC has circuitry that is capable of controlling the gate voltages of gates of the first and second JFETs to cause charge to flow from buried photodiode through at least one of the first and second JFETs without an ohmic connection between the buried photodiode and the amplification and storage circuit.

2. The focal plane detector array according to claim 1, further comprising: a substrate; at least one buffer layer, formed on the substrate, of n-type material at least partly composed of indium; an In.sub.xGa.sub.1-xAs layer, formed a topmost of the at least one buffer layers, of intrinsic or undoped material; a first cap layer, formed on the In.sub.xGa.sub.1-xAs layer, of n-type material at least partly composed of indium; a second cap layer, formed on the first cap layer, of p-type material at least partly composed of indium; and a third cap layer, formed on the second cap layer of n-type material at least partly composed of indium.

3. The photodiode array according to claim 2, wherein a first and second JFET drains are formed by p-type regions which are deep enough to reach the second cap layer but not deep enough to enter the first cap layer.

4. The focal plane detector array according to claim 1, wherein the second JFET is linearly distributed across a row or a column of the photodiode array.

5. The focal plane detector array according to claim 1, wherein each pixel of the plurality of pixels includes an independent second JFET.

6. The focal plane detector array according to claim 1, wherein the flow of charge from the buried photodiode is controlled by varying gate voltages of the first and second JFETs using the ROIC.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Embodiments will be described with reference to the following drawing figures, in which like numerals represent like items throughout the figures, and in which:

(2) FIG. 1A is a cross section diagram of an example buried photodiode structure suitable for In.sub.xGa.sub.1-xAs material lattice matched to InP for visible and SWIR light detection;

(3) FIG. 1B is a cross section diagram of an example structure of In.sub.xGa.sub.1-xAs photodiode structure for visible and extended SWIR wavelength non-lattice matched to InP;

(4) FIG. 2 is a diagram of an example a two dimensional photodiode array as viewed from the top.

(5) FIGS. 3A-B is a flowchart showing a method of manufacture for a buried photodiode;

(6) FIG. 4 is diagram showing the deposition of the semiconductor layers of (A) an extended wavelength photodiode and (B) a lattice matched photodiode;

(7) FIG. 5 is a diagram showing the deposition of a passivation layer, SiNx in this case but others can be used, with a hole (75) for adding the p-type dopant in a lattice matched photodiode;

(8) FIG. 6 is a diagram showing the region of the diffused or ion-implantation of a p-type dopant to form the photodetector in a lattice matched photodiode;

(9) FIG. 7 is a diagram showing a larger area opening (76) then (75) and above (6) in the passivation layer to allow the deposition of an n-type semiconductor region by diffusion or ion implantation to bury the photodetector in a lattice matched photodiode;

(10) FIG. 8 is a diagram showing the deposition of a another passivation layer (77) to fill in the hole over the buried photodetector (6,7) in a lattice matched photodiode;

(11) FIG. 9 is a diagram showing new openings in the passivation layer (77) to enable the diffusion or ion-implantation of p-type dopants to form the drains of the junction field effect transistors (JFETs) in a lattice matched photodiode; and

(12) FIG. 10 is a diagram showing the deposition of the p-contacts (14, 15) to the drains (9,10) in a lattice matched photodiode; and.

(13) FIG. 11 is a diagram showing the channel etch to the cathode contact or first epitaxy layer (1) in a lattice matched photodiode; and

(14) FIG. 12 is a diagram showing the n-contacts for the gates for the JFETs (8,11) as well as the cathode contact (60) for the photodiode in a lattice matched photodiode.

DETAILED DESCRIPTION

(15) The present disclosure describes planar devices, which allow a complete interconnected array of pixels with no loss of material between pixels thus no loss in signal. In various implementations, an InGaAs detector array that has a buried photodiode (either p-n or p-i-n) and a junction field effect transistor (JFET) or multiple JFETs in each pixel to allow the movement of charge from the buried photodiode to a read out integrated circuit (ROIC) without the ROIC being directly connected to the photodiode. In some scenarios, the JFETs may be integrated to the photodetector in each pixel to allow movement of charge to the ROIC. A second JFET may be used to remove charge from the first JFET. This second JFET may be in each individual pixel or it may be a larger device that works an entire column or row simultaneously or 2 rows or 2 columns simultaneously. The multiple JFETs combined with the buried photodiode enables double correlated sampling in the pixel as well as lower capacitance in the connection to the ROIC, thereby enabling low read noise readouts. This scheme enables us to move charge noiselessly from the photodiode to the ROIC similar to a Charge Coupled Device (CCD). In addition it lowers the dark current of the photodiode because there is no bias applied to the photodiode from the ROIC and the photodiode does not contact the surface limiting the number of defects. Some embodiments may be entirely monolithic and may be used for multiple alloys of In.sub.xGa.sub.1-xAs (e.g., In.sub.0.53Ga.sub.0.47As, also known as InGaAs) detector arrays from, for example, In.sub.0.50Ga.sub.0.50As to In.sub.0.95Ga.sub.0.05As. The JFET may be formed in the InAs.sub.yP.sub.1-y cap layers, enabling the charge to be moved from the buried photodiode.

(16) In Si CMOS imagers, for example, buried photodiodes may be built with transfer gates to move the charge from the buried photodiode. Buried photodiodes have not been developed in infrared imagers because these are traditionally hybrid devices. A hybrid device or focal plane array is a device where the infrared imager is a detector material(s) (InGaAs, InSb, HgCdTe, Ge, etc.) and it is attached to Read Out Integrated Circuit (ROIC) with electronic circuitry this is a second material, typically Si CMOS. The attachment between the detector and ROIC uses indium bump or another hybridization technology, for example, each pixel or picture element is the connection from the detector to the ROIC. It is difficult to hybridize two materials, such as InGaAs and Si CMOS ROIC, especially on a device with small pitch, spacing between pixels. Each pixel may require a separate contact to the ROIC. Buried photodiode technology requires multiple contacts for a single pixel to integrate the switching circuits to move the charge from the buried photodiode adding additional complication to the hybridization. In addition to the hybridization issues, it is necessary to build a circuit beyond a photodiode in every pixel, which significantly adds difficulty, complexity, and cost to the process. The inventive photodiode arrays described herein overcome these and other deficiencies in the field of infrared photodiode arrays.

(17) Referring now to FIG. 1A, a first embodiment of a photodiode is shown. In the first embodiment, the detector structure may be grown epitaxialy on InP substrates 13 or other substrates like InAs or GaAs which may be n-type or semi-insulating. The detector structure is developed by placing an InP n-type layer 1 followed by an In.sub.xGa.sub.1-xAs layer, for example an In.sub.0.53Ga.sub.0.47As alloy (which is lattice matched to the InP substrate, it is abbreviated as InGaAs in drawing), which is used as the detector layer 2. The In.sub.xGa.sub.1-xAs layer may be undoped or intrinsic. The In.sub.xGa.sub.1-xAs layers may be doped n type but it is not required. The thickness of the In.sub.xGa.sub.1-xAs may be from 0.5 m to 6 m thick. On top of the In.sub.0.53Ga.sub.0.47As 2 an InP cap (3, 4, 5) layer may be grown. The cap is composed of three InP layers each doped differently. The layer 3 is n type semiconductor having a carrier concentration in the range of about 110.sup.15 to 310.sup.19 per cm.sup.3 and a thickness ranging from about 0.05 m to 2 m. The layer 4 may be about 0.05 m to 2 m thick of p-type material with a doping of 110.sup.15 to 310.sup.19 per cm.sup.3. The layers 4, 5 may be grown in the same growth run as layers 1 and/or 2, 3 or it may be grown afterward in a regrowth following diffusion of the dopants to form the buried photodiode 6. The fifth layer 5 may be 0.1 m to 2 m thick and may be composed of n-type InP with an n-type doping of 110.sup.15 to 310.sup.19 per cm.sup.3. A silicon-nitride SiN.sub.x or other passivation layer (17) may also be applied after the growth is completed to passivate the surface. Holes are opened in the passivation to diffuse or ion-implant the various dopant constituents (6, 7,9, 10) as well as the n-contact for gates (8, 11, 26, 32) and cathode contacts (60) and p-contacts for drains of the JFET (14, 15, 27, 34) to the JFET gates and drains. Sidewall passivation (63) may be needed to protect the structure near the edge of the array and a metal path (62) to bring the cathode contact (60) to the array surface for hybridization (21).

(18) To form the buried photodiode 6 a p-type dopant, e.g., Zn, will be diffused into or ion implanted through the InP cap layers 3, 4, 5 to the InGaAs layer 2. This is done by opening a hole in layer 17 to allow the p-type dopant to be diffused or ion-implanted. The dopant may be deposited from 0.01 m to 4 m deep in the InGaAs layer 2 through layers 3, 4, and 5. Alternatively the p-type dopant may be diffused or ion implanted before layer 4, and/or 5 are grown. The depth of the dopant should be the same, 0.01 m to 4 m deep in the InGaAs, then layer 4, and 5 of the InP may be grown in, a regrowth process as described above. Alternatively if InP layer 5 already exists from the initial growth the p-type dopant is added to form 6 then an n-type dopant 7 may be placed in layer 5 above the buried photodiode 6. The n-type dopant 7 should be ion implanted or diffused into the surface above the photodiode to achieve a similar or greater n-type doping profile to the rest of the cap layer 5. The area of region 7 needs to be wider then region 6 so the p-type volume 6 is completely covered, no p-type material may contact the surface. To create a wider area of coverage a wider hole would be opened in layer 17 to enable the ion-implant or diffusion of n-type dopant. The depth of the ion implant may be from the surface to 0.05 m to 1 m deep depending on the layer thickness of the InP layer 5. A new passivation layer is applied to cover 7. Holes are open in the passivation layer 17 to allow a p-type dopant to be diffused or ion-implanted into the material through the n-type layer 5 to the InP p-type second layer 4 to form JFETs drains 9, 10. The p-type dopant must enter the p-type layer 4 but may not diffuse beyond the p-type layer to the n-type layer 3 below it. Following the addition of the p-type and n-type dopants 6, 7 the holes in the passivation 17 for the p-type diffusion or ion-implantation and n-type diffusion or ion-implantation is closed.

(19) Following the manufacture of the buried photodiode the JFET drains 9, 10 are formed. Transfer JFET drain (JFET-T) 10 may allow charge to be removed from the photodiode while refresh JFET drain (JFET-D) 9 may enable charge to be cleared from JFET-T 10. Once the diffusions are complete ohmic contacts 14, 15 may be made to both JFETs drains. After the p-contacts are formed the n-contacts for the gates for the JFETs 8 and 11 and the cathode contact 60 may be established. The n-contact to establish the gate 8 controlling the flow of charge between the ROIC circuit 50 and the buried photodiode 6. The n-contact 11 establishes the gate controlling the flow of charge between 10 and 9. Applying or removing applied voltages to these gates 8 and 11 allow the ROIC to control the flow of charge between the photodiode and the various JFET drains. Before the n-contacts are formed by making holes in the passivation layer 17, a hole is formed in the passivation layer at the end of the array and material is removed from the surface layer 5 down to the substrate 13 or the first InP epitaxial layer 1. Passivation is applied to cover the sidewalls 63. N-contacts are then formed at 60, 8 and 11 simultaneously. The n-contact 60 on the initial InP n-type layer 1 or on the substrate 13 is the cathode connection allows the completion of the circuit for attachment to the ROIC 35 (shown in FIG. 2). Metal 62 is generally applied to allow a connection between the cathode 60 and the top surface of the photodetector array to make hybridization of the ROIC 18 to the photodetector array simpler by placing all of the connection points (21) on a single planar surface.

(20) The photodiode array may be attached to the ROIC 18 through various techniques available for hybridization, i.e. Indium bumps 21. The ROIC 18 will control the gate voltages to 8 and 11 allowing charge to be moved from the photodiode to 10 and then 50 or 9 and thus 51. The ROIC 18 can then either store the charge collected from the charge removed at 12 in the amplification and storage circuit 50 or move charge from 10 to 9 and discard it to ground 51 depending on the operation. This setup allows for zero bias on the photodiode and minimizes dark current from surface defects in the photodiode. Charge may be moved from the photodiode 6 similar to a CCD thus allowing for correlated double sampling in the pixel and other techniques to reduce the noise i.e. kTC noise in the ROIC 18 at the circuit 50 and at the same time minimize dark charge from the photodetector 6.

(21) Referring now to FIG. 1B, the second embodiment of a photodiode is shown with non-lattice matched In.sub.xGa.sub.1-xAs. In the second embodiment, the detector structure is also grown epitaxialy on InP substrates 13 or other substrates like InAs or GaAs which may be n-type or semi-insulating. The detector structure is developed by placing an InP n-type layer 1 followed by any/multiple InAs.sub.yP.sub.1-y (e.g., 52, 53, 54) that are needed to grade the material to modify its lattice constant to enable a lattice match to the In.sub.xGa.sub.1-xAs alloy used as the detector layer 55. The In.sub.xGa.sub.1-xAs layer 55 may be undoped or intrinsic. The In.sub.xGa.sub.1-xAs layers can be doped n type but it is not required. The thickness of the In.sub.xGa.sub.1-xAs may be from 0.5 m to 6 m thick. On top of the In.sub.xGa.sub.1-xAs 55 an InAs.sub.yP.sub.1-y cap 56, 57, 58 layer may be grown that has the same lattice constant as the In.sub.xGa.sub.1-xAs 55. The cap is composed of three InAs.sub.yP.sub.1-y layers each doped differently and with different thicknesses. The layer 56 is n type semiconductor having a carrier concentration between 110.sup.15 to 310.sup.19 per cm.sup.3 and may range from 0.05 m to 2 m thick. The layer 57 may be 0.05 m to 2 m thick of p-type material with a doping of 110.sup.15 to 310.sup.19 per cm.sup.3. The layers 57 and 58 may be grown in the same growth run as layers 1 and/or 52, and 56 or it may be grown afterward in a regrowth following diffusion of the dopants to form the buried photodiode 6. The top cap layer 58 may be 0.1 m to 2 m thick and may be composed of n-type InAs.sub.yP.sub.1-y with an n-type doping of 110.sup.15 to 310.sup.19 per cm.sup.3. A SiN.sub.x or other passivation layer (17) may also be applied after the growth is completed to passivate the surface. Holes are opened in the passivation to diffuse or ion-implant the various dopant constituents (6,7,9,10) as well as the n-contact for gates (8,11,26,32) and substrate connection 60, and p-contacts for drains of the JFETs (14,15,27,34). Similarly to FIG. 1A passivation of the side walls is conducted 63 and metal is applied 62 to bring the substrate contact 60 to the surface for hybridization 21.

(22) This detector structure as described above is attached to a ROIC. The buried photodiode will require this device to operate differently then a standard InGaAs focal plane array. Typically the photodiode is directly attached to the ROIC. The photodiode is also typically put in reverse bias or near reverse bias. Light is shined on the detector array and if energy of that photon is high enough it will allow the creation of an electron-hole pair. This electron-hole pair is in addition to dark current or electron-hole pairs which naturally form in the pixel as long as one is above absolute zero in temperature. The charge created by the photon or by dark current is then trapped in the field created by the reverse bias at the p-i-n or p-n junction and the charge is then moved to the ROIC. The ROIC has an amplification circuit, which then converts the charge to a voltage and allows one to measure the amount of charge created by photons and dark current in a given amount of time. The photodiode is connected to the ROIC circuit and charge created by any mechanism is moved directly to the ROIC circuit. With this new detector structure the detector is not directly connected to the ROIC amplification circuit. A switch is placed in the detector material to control the charge flow from the photodiode to the ROIC 18 which has an amplification/storage circuit 50 at each pixel or set of pixels. The charge movement is now controlled by the ROIC 18 through the JFETs. The switch is a JFET (8, 10, 14) in this case. The photodiode formed by the p-i-n junction at 6, 2, and 1 still collects dark charge as well as charge created from photons landing in the photodiode array. The charge is stored in the p-i-n or p-n junction. After a specific amount of time, i.e the integration time or a part of the integration time if one is trying to achieve a greater dynamic range, the charge will need to be moved from the buried photodiode to the drain of the JFET-T (10). To move the charge the gate of JFET-T 8 must be opened (no voltage applied to the gate of JFET-T (8)). To prevent the flow of charge from the photodiode to the drain of JFET-T (10,12) a voltage is applied to the gate (8), this closes the switch. The voltage is applied to 8 during the integration time to build charge on the photodiode (6). When the switch 8 is closed one can also measure the charge on the drain of JFET-T at 10. This allows double correlated sampling and enables the reduction of read noise in the circuit by measuring the kTC noise before and after the charge from the photodiode is moved. To ensure all of the charge is removed from the drain of JFET-T 10 a second JFET, JFET-D (9, 11, 15) is used. When JFET-T has its gate closed 8 between the photodiode 6 and its' drain (10) JFET-D could have its gate open 11. This allows charge to move from JFET-T to JFET-D. When JFET-T is open and is accepting charge from the photodiode or when JFET-T drain 10 is being sampled by the circuit 50 to measure the amount of charge captured JFET-D's gate (11) has a voltage applied to close its switch so the charge does not move to JFET-D drain 9. When charge is being moved from the photodiode to JFET-T drain (10) it is desired that no charge is lost to the JFET-D drain (9) or ground 51. The gate on JFET-D (11) only has its voltage turned off when charge is looking to be removed from JFET-T drain (10).

(23) Referring to FIG. 2, it is a top view of a photodiode array (one-dimensional or two-dimensional with 2-dimensional shown in FIG. 2 in InGaAs detector alloys with a buried photodiode (33) and a Junction Field Effect Transistor(s) in each pixel to allow for movement of charge to a Read Out Integrated Circuit. It is comprised of a plurality of photodiodes as described above in relation to FIG. 1. The following layers are included: a substrate (layer 13 of FIG. 1) of InP that is either n-type or semi-insulating; first layer (layer 1) of InP buffer layer of n-type material; a second layer (layer 2 of FIG. 1) of In.sub.0.53Ga.sub.0.47As layer of intrinsic or undoped material that is 1 m to 6 m thick; a third layer (layer 3 FIG. 1) of InP layer of n-type material (110.sup.14 to 510.sup.19/cm.sup.3) that is 0.5 m to 2 m thick; a fourth (layer 4 of FIG. 1) of InP layer of p-type material (110.sup.14 to 510.sup.19 per cm.sup.3) that is 0.05 to 1 m thick; a fifth layer (layer 5 of FIG. 1) of InP layer of n-type material (110.sup.14 to 510.sup.19 per cm.sup.3) that is 0.1 m to 2 m thick.

(24) A photodiode array comprised of the above structure can be manufactured in a single epitaxy growth run or through a regrowth process. Either after the deposition of the fourth layer or the fifth layer, holes are opened in the passivation layer 17, a diffused or ion-implanted p-junction (buried photodiode 6 of FIG. 1) through the third, fourth, and fifth layers and into the second layer (layer 2 of FIG. 1) to form a p-i-n or p-n photodetector. The p-type dopant (buried photodiode 6 of FIG. 1) may be diffused from 0.1 m to 5.5 m in depth of the second layer. The position of buried photodiode 33 is shown but would not be directly viewable since it is buried under n-type region 36, which is wider in area and is touching the surface. If the p-type region is implanted or diffused after the fifth layer is deposited, a diffused or ion-implanted n-type region (region 7 of FIG. 1) is deposited above the diffused p-type region that buries the junction below the surface of the fifth layer. The depth of the n-type implant is 0.1 m to 2 m into the fifth layer. It should not enter the p-type layer of InP (layer 4 of FIG. 1). If the p-type region is implanted after the fourth layer, the above n-type implant is not required since the regrowth of the fifth layer accomplishes a similar structure. The holes in the passivation layer 17 for the p-type diffusion or ion-implantation and n-type diffusion or ion-implantation are closed.

(25) After the deposition of the fifth layer (or of the n-type implant) a new set of holes are opened in the passivation layer 17 to form the JFETs by diffusing or ion implanting a p-type region for JFET drains 23 and 30 (p-type implant 10 of FIG. 1) and (p-type implant 9 of FIG. 1). The diffused p-type region should be deep enough so it reaches the InP p-type region of the fourth layer (layer 4 of FIG. 1) but not be so deep that it enters the n-type InP zone of the third layer (layer 3 of FIG. 1).

(26) Referring again to FIG. 2, a 2-D photodiode array is shown. Photodiode arrays such as those described above need not have two JFETs in every pixel. A single JFET drain 23 and gate 24 may be included with each buried photodiode (JFET-T) and the second JEFT for clearing charge (JFET-D) from the first JFET can be a row or column based JFET with the gate control 31 running the length of the row or column and the JFET-D drain 30 running the length of the row or column. The connection to the gate 26 and drain 27 of the row or column JFET can be made on the end of the array by the substrate connection 35 (60 in FIG. 1A) of the array allowing the circuit to be completed by attaching to the ROIC 18. This simplifies the hybridization of the photodiode array to the ROIC by having only two connections at each pixel versus four. Less is needed in each individual pixel, this allows for a smaller pitch device since fewer connections at each pixel.

(27) Referring to FIG. 3, a method 300 is described to form an array of photodiodes in InGaAs detectors for detecting visible and shortwave infrared light that has a buried photodiode to reduce the dark current that is not directly connected to the ROIC. A JFET may also be integrated in every pixel to allow the charge to be removed from the photodiode to a ROIC. The JFET may be connected to the ROIC.

(28) The method includes depositing, on a substrate, a first n-type doped semiconductor layer 302. The first n-type doped semiconductor layer may be, for example, InP. Optionally, one or more additional layers may be deposited on the semiconductor layer 304. The additional layers may be composed of InAs.sub.yP.sub.1-y and deposited to transition from the lattice constant of the first n-type doped semiconductor layer to that of an In.sub.xGa.sub.1-xAs alloy deposited as a higher layer, as shown in FIG. 4A. An undoped or intrinsic semiconductor layer is deposited on the first n-type doped semiconductor layer 306 (or the top additional layer). The undoped or intrinsic semiconductor layer may be, for example, an In.sub.xGa.sub.1-xAs alloy. As described above, the one or more additional layers can be deposited to transition from the first n-type doped semiconductor layer to the In.sub.xGa.sub.1-xAs. When In.sub.0.53Ga.sub.0.47As is used, no additional layers are required to provide lattice matching, as shown in FIG. 4B. Dopants for p-type may include Zn, Be, and/or C for example. Dopants for the n-type may include S, Si, Ge, Se, and/or Sn for example.

(29) A cap comprised of alternating n-type and p-type doped semiconductor layers is deposited on top of the undoped or intrinsic semiconductor layer, as shown in both FIGS. 4A (in extended wavelength cases where additional layers are required to match lattice constants) and 4B (in cases where In.sub.0.53Ga.sub.0.47As is used). A first cap layer comprising second n-type doped semiconductor layer is deposited on the undoped or intrinsic semiconductor layer 308. The first cap layer may be, for example, comprised of InAs.sub.yP.sub.1-y, doped n-type to a carrier concentration from about 110.sup.15 to 310.sup.19 per cm.sup.3 and may have a thickness of 0.25 m to 2 m. In cases where In.sub.0.53Ga.sub.0.47As is used, the cap layers are composed of InP, i.e. InAs.sub.yP.sub.1-y where y=0. A second cap layer is deposited over the first cap layer and comprises a p-type doped semiconductor layer 310. The second cap layer may be, for example comprised of InP, doped to a carrier concentration from about 110.sup.15 to 310.sup.19 per cm.sup.3 and may have a thickness of 0.05 m to 2 m. A third cap layer is deposited over the second cap layer and comprises a third n-type doped semiconductor layer 312. The third cap layer may be, for example InP doped n-type to a carrier concentration from about 110.sup.15 to 310.sup.19 per cm.sup.3 and may have a thickness of 0.1 m to 2 m. In cases where In.sub.0.53Ga.sub.0.47As is used, the cap layers are composed of InP, i.e. InAs.sub.yP.sub.1-y where y=0.

(30) In the embodiments shown in FIGS. 4A, 4B, and 5, the method 300 continues to step 314 after depositing the third, fourth, and fifth layers. However, after depositing any of the third, fourth, or fifth layers (i.e. after any of steps 308 or 310), a passivation layer may also be deposited and/or photolithography can be used to designate where the ion-implantation or diffusion will occur. This is shown in FIG. 5 as passivation layer 17 and one or more holes 75. Alternatively once the third layer is deposited (layer 3 of FIGS. 4A and 4B), a passivation layer (e.g. a layer composed of SiN.sub.x) is deposited 314 and/or basic photolithography techniques maybe utilized to designate the p-type region of 6. Alternatively, the passivation layer may be deposited and/or photolithography techniques maybe used after the fourth layer is deposited to designate the p-type region of 6. The embodiments are not limited in this regard. If a passivation layer is deposited, one or more holes are opened in the passivation layer using photolithographic techniques 316. The formation of a photodetector p-n or p-i-n in the undoped or intrinsic layer may be performed 318 using any of the aforementioned techniques. The passivation layer or photolithographic material, with the one or more holes (i.e., holes 75), provides a mask to precisely place the photodetector in the lower layers, as described below.

(31) Referring now to FIG. 6, the buried photodiode 6 is formed by diffusing or ion-implanting a p-type dopant through the fifth, fourth, and/or the third layers (5, 4, 3) and into the undoped or intrinsic semiconductor layer. The photodiode may be a diffused or ion-implanted p-type region through and forming a photodetector of p-i-n or p-n. The p-type dopant may be diffused to a depth of about 0.1 to 5.5 m into the undoped or intrinsic semiconductor layer 2. If either the fourth or the fifth layer has not yet been deposited, the passivation layer deposited in 314 is removed and the fourth and/or fifth layers are deposited 320. The process 300 then continues on FIG. 3B at point A. Alternatively, if the fourth and fifth layers have been previously deposited, the process continues on FIG. 3B at point B where an n-type region is diffused or ion-implanted over the p-type dopant diffusion to provide an n-type doped surface 322. This is shown in FIG. 7 as n-type region 7.

(32) After either of steps 320 of FIG. 3A or 322 of FIG. 3B, a passivation layer is reapplied over at least the n-type layer deposited over the photodetector or over the entire surface, depending on when the photodetector was formed in the above process 324. This is shown in FIG. 8 as passivation layer 77 extending over n-type layer 7. Referring to FIG. 9, holes 78, 79 are opened in passivation layer 77 (step 326), and a p-type dopant is diffused in the open areas to form the JFETs drains 9, 10 (step 328). As shown in FIG. 10, the contacts 14, and 15 are deposited using standard photolithographic techniques to make openings in the passivation layer 77 to form the p-contacts for the JFET drains (step 330).

(33) After step 330 photolithographic steps (or other semiconductor processing technique) would be used to make a new hole on the end of the array in the passivation 77. After making a hole in the passivation then material would be removed from layer 5, 4, 3, and 2 to allow a connection to the first InP layer 1 or the substrate 13 as seen in FIG. 11 (step 332). This hole will enable an n-contact for the cathode connection, 60.

(34) Holes would then be opened in 77 using photolithography (or other technique) to form the n-contacts for the JFET gates (8 and 11). The n-contact would then be applied to form n-contacts for the JFET gates (8 and 11) as well as the cathode contact 60 on the substrate 13. (step 334)