HK1125745B

HK1125745B - Semiconductor radiation detector optimized for detecting visible light

Info

Publication number: HK1125745B
Application number: HK09104463.1A
Authority: HK
Inventors: 阿尔托‧奥罗拉
Original assignee: 阿尔托‧奥罗拉
Priority date: 2006-01-05
Filing date: 2006-02-17
Publication date: 2011-03-18

Abstract

A semiconductor radiation detector comprises a bulk layer of semiconductor material, and on a first surface of the bulk layer in the following order: a modified internal gate layer of semiconductor of second conductivity type, a barrier layer of semiconductor of first conductivity type and pixel dopings of semiconductor of the second conductivity type. The pixel dopings are adapted to be coupled to at least one pixel voltage in order to create pixels corresponding to pixel dopings. The device comprises a first conductivity type first contact. Said pixel voltage is defined as a potential difference between the pixel doping and the first contact. The bulk layer is of the first conductivity type.; On a second surface of the bulk layer opposite to the first surface, there is no conductive back side layer that would transport secondary charges outside the active area of the device or function as the radiation entry window.

Description

Semiconductor radiation detector optimized for detecting visible light

Technical Field

The present invention relates to semiconductor radiation detectors, and in particular to semiconductor radiation detectors having improved internal gates.

Background

In the semiconductor material the radiation is converted into electron-hole pairs. In semiconductor radiation detectors, electron-hole pairs are separated by an electric field. The charge type of the electron-hole pair being measured is referred to as the signal charge and the opposite charge type is referred to as the secondary charge.

Patent applications PCT/FI2004/000492 and PCT/FI2005/000359, which are incorporated herein by reference, disclose a semiconductor radiation detector with an improved internal gate (MIG). This detector will be referred to as MIG detector in the following. The MIG detector is back-illuminated and has a thick fully depleted substrate and a thin conductive layer on the back side of the device. This conductive backside layer has three functions: when properly biased, it enables the thick substrate to be fully depleted, it transports secondary charges outside the active region of the device, and it acts as a thin uniform radiation entry window. MIG detectors have many advantages. The surface generated charges can be separated from the signal charges, which provides small dark current noise. The signal charge can be read out non-destructively, so that the signal charge is read a plurality of times, which reduces read noise. The back-illuminated and thin uniform radiation entrance window enables detection of shallow penetrating radiation (e.g., low energy X-rays and particles) with good energy resolution. The thick fully depleted substrate enables detection of deep penetrating radiation.

The substrate material of the MIG detector is preferably high-resistive, i.e. substantially intrinsic silicon, and the thickness of the substrate is several hundred microns. Such MIG detectors can be used to detect particles from low to medium energies (-100 eV-10 keV), X-rays, and photons from ultraviolet and blue light to near infrared radiation. Here, the near-infrared radiation is referred to as radiation that is invisible to the human eye and has a wavelength below 1.1 μm, which is the detection limit of silicon. Near infrared radiation with wavelengths close to but below this limit has a very large attenuation depth in silicon, up to several hundred microns. Due to back-illumination, due to the thick fully depleted substrate, and due to the thin radiation entrance window, the MIG detector has a high quantum efficiency from near infrared radiation to blue light. Due to the thick substrate, a phenomenon known as edge phenomenon (fringing) is also eliminated. Edge phenomena are a problem in detectors with thin substrates. In such detectors, near infrared radiation is reflected multiple times between the front and back surfaces of the detector before being absorbed, resulting in an undesirable interference pattern. Because the moonless night space contains at least one order of magnitude more near-infrared photons than visible photons, and because the reflectance of many materials is much higher for near-infrared radiation than for visible light (e.g., the reflectance of the leaves is 3 to 6 times higher), MIG detectors may be well-suited for low light detection in night vision devices.

However, MIG detectors are not well suited for the detection of visible light in silicon-based portable user applications for the following reasons. Depletion of a thick substrate requires at least tens of volts. For portable user devices, such voltages are undoubtedly too high and result in too much power consumption. High impedance silicon substrates are expensive and difficult to handle, which increases manufacturing costs. It is also difficult to reliably contact the conductive backside layer from the front side through a thick high impedance substrate, which is very important for mass production. Generating a large bulk (bulk) generated current in a thick fully depleted substrate necessitates the use of cooling. However, in portable user applications, cooling of the detector is often not possible. The sharpness of the image is also somewhat reduced, since visible light is absorbed on the back side of the detector and the signal charge has to drift a long distance before it reaches the front surface. Therefore, there is a problem with using color filters on the back side of the device.

In silicon, the depth of decay of red light is no greater than 10 microns. The attenuation depth is smaller for blue and green light. Thus, it is not necessary to have a thick substrate for visible light detection. Instead of a thick substrate, a thin (typically about 10 μm and less than 50 μm) substrate can be used in a backside thinned MIG detector. However, thin device slow down (break) is very easy, so back side processing must be performed at the end of the manufacturing process. There are two possible ways to perform this process. In the first method, the front side of the substrate is attached to a support substrate, after which the back side of the probe is thinned. In the second approach, the backside of the detector is etched only under the active area containing the pixels, and thicker support areas are left on the sides of the detector. In both methods, the front side processing needs to be completed before the back side is thinned. This situation complicates the manufacturing process of the conductive backside layer. In order to process very thin conductive back-side layers, which can have a good quantum efficiency for blue light, there are two possible processes suitable for mass production. In the first method, the conductive backside layer is realized by an implantation process, which requires a high temperature annealing step. All materials used on the front side of the device (like metal wiring) must have a higher melting point than the annealing temperature. This situation prevents the use of many materials commonly used in integrated circuits, such as aluminum. In the second method, a thin layer is deposited on the back side of the device. However, a lot of dark current is generated at the interface between the conductive layer and the substrate, and cooling is required in order to suppress the current.

There are also inherent problems associated with conductive back side layers in case the MIG detector is used for visible light detection. In order to properly detect a poorly illuminated area of an image, the size of the chip must be large and a large optical aperture (optical aperture) must be used. In order to have a good quantum efficiency also for blue light, the conductive back side layer has to be very thin. If the image also includes very bright areas, a large amount of secondary charge current will flow in the conductive back side layer. However, the large current flowing in the conductive backside layer, and the small thickness and large area of the conductive backside layer, result in a large resistive voltage drop in the conductive backside layer. This resistive voltage drop reduces the image quality and can lead to failure of the detector, especially if the detector is very thin.

Another problem in MIG detectors is that relatively high voltages are required to clear the signal charge in the MIG, especially if a high dynamic range is desired, i.e. if a large signal charge capacity of the MIG is desired. Another problem is that the insulation between surface generated charges and signal charges in a MIG detector should be improved in some cases.

Disclosure of Invention

It is an object of the present invention to provide a semiconductor radiation detector comprising an improved internal gate, wherein the problems caused by the conductive back side layer are eliminated. It is a further object of the invention to provide a structure of a semiconductor radiation detector comprising an improved internal gate, by which structure signal charges can be cleared using only small voltages. It is another object of the invention to provide a method of improving the separation of surface generated charges and signal charges.

The object of the invention is achieved by a semiconductor radiation detector comprising a bulk layer of semiconductor material and further comprising on a first surface of the bulk layer in the following order: a modified internal gate layer of a semiconductor of a second conductivity type; a barrier layer of a semiconductor of a first conductivity type; and a pixel doping (pixel doping) of a semiconductor of the second conductivity type adapted to be connected to at least one pixel voltage to form a plurality of pixels corresponding to the pixel doping. The device is characterized in that it comprises a first conductivity type first contact, such that the pixel voltage is a potential difference between the pixel doping and the first contact, and the bulk layer is of the first type, and the device does not comprise a conductive backside layer on a second surface of the bulk layer opposite to the first surface, wherein the conductive backside layer transports secondary charges outside the active area of the device and serves as a radiation entrance window.

The invention is based on the idea to remove a conductive back side layer from the MIG detector. It is achieved that secondary charges generated in the bulk layer can be collected inside the active region by channel stop (channel stop) without using a conductive back-side layer for transporting it outside the active region. The collection of secondary charges from the bulk layer within the active area can be further improved by gaps (gaps) in the MIG layer and by a suitably filled trench structure. A new MIG detector is also realized which may be front-lit. However, the front illuminated new MIG detector will have a poor blue response due to the fact that. The barrier layer forms a barrier for signal charges between the MIG layer and the surface of the device. Blue light is mainly absorbed between the potential barrier and the front surface of the detector, so that most of the signal charges generated by blue light are collected by the pixel doping and not by the MIG. By realizing that such a barrier is not present under the channel stop and that the channel stop region can act as a radiation entrance window, and by further realizing that the channel stop can be very thin and that the size of the channel stop region can be very large, the blue response of the front illuminated device can be significantly improved. However, the large area of the channel stop reduces the signal charge transport potential gradient in the MIG layer below the channel stop. It is further achieved that the potential gradient for transferring signal charges can be improved by a structure, i.e. by a discontinuous MIG layer. Another possibility is to change the doping concentration in the barrier layer, the MIG layer or the bulk layer next to the MIG layer to improve the potential gradient for transporting the signal charges. In front illuminated detectors, secondary charges generated in the bulk can be collected by channel stops inside the active region, and/or by substrate contacts located outside the active region on the front side of the detector, and/or by substrate contacts located on the edge of the detector chip or on the back side of the detector chip.

The signal charges can be removed using only a small voltage by having a second conductivity type doped region in the first conductivity type barrier layer or a local reduction (reduction) of the barrier layer net doping between the second conductivity type modified internal gate layer and the second conductivity type pixel doping or by having a trench between the MIG layer doping and the front surface of the detector, wherein the gate controls the flow of signal charges from the modified internal gate layer through the second conductivity type doped region, through the trench structure, or through the local reduction of the barrier layer doping to the pixel doping or the front surface of the detector.

The separation of the signal charges and the separation of the surface-generated charges is improved, for example, by a doped region of the second conductivity type between the barrier layer and the front surface of the detector or by a gated structure.

The invention provides a method for detecting radiation, comprising: connecting a plurality of pixels on a surface of a semiconductor radiation detector device to a pixel voltage and illuminating the semiconductor radiation detector with radiation; it is characterized in that it comprises: collecting, from a bulk layer (103), a modified internal gate layer, and a barrier layer of the semiconductor radiation detector, first radiation-induced signal charges to a local minima of a three-dimensional potential function of the first said charges, the local minima positionally coinciding with pixels (111) on the modified internal gate layer (104, 304) located immediately adjacent to the bulk layer (103), and detecting an amount of signal charges collected to the local minima coinciding with the pixels (111).

Preferably, in the method, detecting the amount of the signal charge includes: the electrical characteristics of the pixel specific transistors are observed in relation to their effective channel or base size.

Preferably, in the method, detecting the amount of the signal charge includes: observing an electrical characteristic of the pixel-specific transistor related to the reduced channel or base size of the pixel-specific transistor.

Preferably, in the method, detecting the amount of the signal charge includes: observing an electrical characteristic of the pixel specific transistor associated with the increased channel or base size of the pixel specific transistor.

Preferably, in the method, detecting the amount of the signal charge includes: pixel-dependent charge is transferred through a plurality of pixels to a readout pixel, and the electrical characteristics of the readout pixel are observed.

Drawings

Figure 1 shows an embodiment of the present invention,

figure 2 shows an alternative biasing scheme of the semiconductor radiation detector shown in figure 1,

figure 3 shows another embodiment of the present invention,

fig. 4 shows the electron potential of the detector shown in fig. 1, which uses holes as signal charges,

fig. 5 shows the electron potential of the detector shown in fig. 2, which uses holes as signal charges,

fig. 6 shows the electron potential of the detector shown in fig. 3, which uses holes as signal charges,

figure 7 shows a further embodiment of the invention,

figure 8 shows the semiconductor radiation detector of figure 7 using a protective structure,

fig. 9 shows the electron potential of the detector shown in fig. 7, which uses holes as signal charges,

fig. 10 shows the electron potential of the detector shown in fig. 8, which uses holes as signal charges,

figure 11 shows four pixels of an embodiment of the invention,

figure 12 shows four pixels of another embodiment of the invention,

figure 13 shows a cross-section of the detector shown in figure 11,

figure 14 shows a cross-section of the detector shown in figure 12,

figure 15 shows four pixels of yet another embodiment of the invention,

figure 16A shows a cross-section of the detector shown in figure 15,

figure 16B shows a cross-section of the detector shown in figure 15,

figure 16C shows a cross-section of the detector shown in figure 15,

figure 16D shows a cross-section of the detector shown in figure 15,

figure 17A shows an embodiment of the present invention,

figure 17B shows an embodiment of the present invention,

figure 17C shows an embodiment of the present invention,

figure 17D shows an embodiment of the present invention,

figure 17E shows an embodiment of the present invention,

figure 17F shows an embodiment of the present invention,

figure 17G shows an embodiment of the present invention,

figure 17H shows a cross-sectional view of the detector shown in figure 17G,

figure 17I shows a cross-sectional view of the detector shown in figure 17G,

figure 18A shows an embodiment of the present invention,

figure 18B shows an embodiment of the present invention,

figure 18C shows an embodiment of the present invention,

figure 19 shows an embodiment of the present invention,

figure 20 shows an embodiment of the present invention,

figure 21 shows an embodiment of the present invention,

figure 22 shows an embodiment of the present invention,

figure 23 shows an embodiment of the present invention,

figure 24 shows an embodiment of the present invention,

figure 25 shows a gate signal modified internal gate detector,

figure 26 shows another embodiment of a gate signal modified internal gate detector,

figure 27A shows a cross-section of the detector shown in figures 25 and 26,

figure 27B shows a cross-section of the detector shown in figure 26,

figure 28 shows a gate signal modified internal gate detector,

figure 29 shows another embodiment of a gate signal modified internal gate detector,

figure 30 shows a gate signal modified internal gate detector,

figure 31 shows another embodiment of a gate signal modified internal gate detector,

figure 32A shows one step of one possible detector fabrication process,

figure 32B shows a sequence of one possible detector fabrication process,

figure 32C shows a sequence of one possible detector manufacturing process,

figure 32D shows a sequence of one possible detector fabrication process,

figure 33A shows one step of one possible detector fabrication process,

figure 33B shows a sequence of one possible detector manufacturing process,

figure 33C shows a sequence of one possible detector manufacturing process,

figure 34A shows one step of one possible detector fabrication process,

figure 34B shows a sequence of one possible detector fabrication process,

figure 34C shows a sequence of one possible detector manufacturing process,

figure 34D shows a sequence of one possible detector fabrication process,

figure 35A shows a prior art strip (strip) detector,

figure 35B shows a prior art strip detector,

figure 36 shows the results of the simulation,

figure 37 shows the result of the simulation,

figure 38 shows the result of the simulation,

figure 39 shows the result of the simulation,

figure 40 shows the results of the simulation,

figure 41 shows the result of the simulation,

figure 42 shows the result of the simulation,

figure 43 shows the result of a simulation in which,

figure 44A shows an embodiment of the present invention,

figure 44B shows an embodiment of the present invention,

fig. 44C shows an embodiment of the present invention.

Detailed Description

Fig. 1 is a schematic cross-sectional view of a radiation detector, which is preferably thin and back-illuminated. The detector has a front surface 101 which is directed upwards in the figure. The back surface 102 of the detector through which radiation enters the detector faces downward in the figure. On the rear surface, there may be an optional anti-reflective or scintillator coating. The body layer 103 of the probe is made of a semiconductor material of the first conductivity type. The first and second conductivity types (and vice versa) are referred to herein as positively and negatively doped semiconductors, having an excess of positive and negative charge, respectively. On the front side of the detector, from the back surface towards the front surface, there is first a layer 104 of the second conductivity type, which is referred to as Modified Inner Gate (MIG) layer in the following. In the device of fig. 1 there is a gap in the MIG layer. In front of the MIG layer 104 there is also a layer 105 of the first type, here designated as barrier layer. On top of layer 105 may be a protective insulating layer and a conductor layer forming wires (wiring), gates, capacitors, etc.

Implants 111, 112, 113, 114, 115 of the second conductivity type, preferably pixel-like, patterned in the barrier layer 105 on the front surface of the detector, and referred to as pixel dopings in the following. Biased channel stop dopings (channel stop dopings)121, 122, 123, 124, 125 of the first conductivity type are placed between or in close proximity to the pixels. The dashed line 150 indicates when the bias voltage V is applied_PThe edge of the depletion region when connected between the pixel doping and the channel stop doping. In the embodiment of fig. 1, the depletion regions of the individual pixels are not uniform, so the bulk layer is at the same potential as the channel stop doping. The biased channel stop doping collects all secondary charges generated inside the semiconductor detector, including those generated in the bulk layer, i.e. collects the secondary charges in the active region containing the pixel, and does not have to transfer the secondary charges outside the active region. Thus, no conductive backside layer is required.

The cutting line 160 is perpendicular to the front and rear surfaces, and it passes through the pixel doping. The cut line 170 is also perpendicular to the front and back surfaces and it passes through the channel stop doping. Fig. 4 shows electron potential curves on the cut lines 160 and 170 corresponding to the case where the first conductivity type is n-type and the second conductivity type is p-type. The electron potential curve 403 on the cut line 170 is a straight horizontal line corresponding to the distance axis. The straight horizontal portion of the potential energy curve corresponds to the neutral region and the sloped region corresponds to the depletion region. The electron potential curve 402 corresponds to the cut line 160 and represents a potential difference between the channel stop doping and the pixel doping of V_PThe case (1). Inside the MIG layer, a three-dimensional (3D) potential energy minimum point (minimum) 412 of holes (in this case, signal charges) is formed. The amount of holes in the 3D potential minimum point can be detected as the effective channel width of a Field Effect Transistor (FET) decreases or as the effective base width in a Bipolar Junction Transistor (BJT) decreases. In fig. 4, this corresponds to a reduction in width 415. The location 416 within the barrier layer is a 3D saddle point for electrons and holes. The electron potential curve 401 at the cutting line 160 corresponds to the clear voltage V_CA case of connecting between the channel stop doping portion and the pixel doping portion. In this case, the 3D potential energy minimum point 412 of the hole disappears and the signal charge hole is collected by the pixel doping.

Fig. 2 is a schematic cross-sectional view of a preferably thin back-illuminated semiconductor radiation detector with a gap in the MIG layer, similar to the device of fig. 1. In this case, however, the bias voltage between the channel stop doping and the pixel doping is high, so that there is only one single uniform depletion region 250. The pixel dopings 215 are guard rings that surround the active area. Fig. 5 shows the electron potential curves on the cutting lines 260 and 270. When the potential difference between the channel stop doping and the pixel doping is V_PThe electron potential curve 502 corresponds to the cutting line 260 and the electron potential curve 503 corresponds to the cutting line 270. When the potential difference between the channel stop doping and the pixel doping is V_CThe electron potential curve 501 corresponds to the cutting line 260 and the electron potential curve 504 corresponds to the cutting line 270. The neutral region 513 in curves 503 and 504 corresponds to the channel stop. The neutral region on the right side next to the device backside in the electron potential energy curve 501-504 corresponds to a floating neutral bulk layer. When the potential difference between the channel stop doping and the pixel doping is V_PWhen, i.e. during signal charge integration, there is an electric potential barrier 514 of secondary charge electrons collected by the bulk layer in the curve 503. When the potential difference between the channel stop doping and the pixel doping is V_CWhen the potential energy barrier is not present in curve 504, the secondary charges collected in the bulk layer during signal charge combination can flow freely to the channel stop doping.

Fig. 3 is a schematic cross-sectional view of a preferably thin back-illuminated semiconductor radiation detector with a continuous MIG layer 304. Dashed line 350 is the depletion region edge. The body layer floats in this detector arrangement similar to that in the detector of figure 2. The operating principle of the detector of fig. 3 is shown in fig. 6 and corresponds to the operating principle of the detector of fig. 2.

The device in figures 1 to 3 is a back-illuminated detector which is preferably thin. In thin detectors, the near infrared light should be filtered out to remove edge phenomena. The detectors of figures 1 to 3 may also be front illuminated. In this case the bulk layer is preferably a few hundred microns thick, but the depletion region on the front side of the detector is only a few microns thick. Due to the thick bulk layer, the near infrared radiation does not have to be filtered. The detectors of fig. 1-3 may also have additional layers and structures like anti-reflection coatings, color filters, microlenses, scintillator layers, and the like. It should be noted that in the case of front-illuminated, possible material layers on the back side of the bulk layer are not necessary for the application, and in the case of back-illuminated, possible material layers on the front side of the device are not necessary for the application. In the detectors of fig. 1 to 3, the secondary charges are collected inside the active region by the channel stop doping, i.e. no conductive back side layer is required. Difficulties associated with the fabrication of a thin conductive backside layer on the backside of a thin detector and the operation of such a detector may thus be avoided.

Fig. 7 shows a front illuminated embodiment of the invention where part of the secondary charge is collected by the channel stop doping and part of the secondary charge is collected by the first conductivity type doping 725 acting as a contact to the bulk layer. The contact 725 is on the front side of the probe, but it may also be on the back surface of the probe or on the edge 700 of the probe chip. If the doping 715 forms a pixel, the channel stop doping is preferably at the same potential as the contact 725. Figure 9 shows the working principle of such a detector. Fig. 7 also shows depletion region edge 750.

Fig. 9 shows the case when the first conductivity type is n-type and the second conductivity type is p-type. Curves 901 and 902 of fig. 9 represent the electron potential energy on the cut line 760 through the pixel doping. Curve 901 corresponds to when the pixel doping 111 is connected to the potential V_PThe situation when, and the curve 902 corresponds to when the pixel doping 111 is connected to the clear voltage (clear voltage) V_CThe case (1). Curve 903 represents the electron potential energy on the cut line 770. The 3D saddle points 914 for electrons and holes form a barrier for secondary charge electrons. FromAnd, part of the secondary charges are collected by the contact portion 725. The channel stop dopant and the body layer contact 725 may be at different potentials if the dopant 715 forms a guard ring around the active region. Fig. 10 shows this situation. The curve 1003 in fig. 10 represents the electron potential energy on the cut line 770. In this case, the neutral layer and the channel stop are at different potentials, i.e. the neutral layer is at zero potential and the channel stop is at potential V_CSTo (3).

Fig. 8 shows another front-lit embodiment of the invention. In this detector, additional guard rings 816, 817, and 818 surround the innermost guard ring 215. No trench structures are needed in these guard rings due to the structured MIG layer. Layer 808 is an optional semiconductor layer of the first conductivity type. Layer 808 preferably has a much higher resistivity than the bulk layer and is preferably fabricated by epitaxial growth. Layer 808 may also be a deep well, in which case the layer may be structured. Fig. 8 also shows depletion region edge 850. If the optional layer 808 is not used, the working principle of the probe of fig. 8 corresponds exactly to fig. 10, i.e. the electronic potential energy curves 901 and 902 correspond to the cutting line 860 and the electronic potential energy curve 1003 corresponds to the cutting line 870. If optional layer 808 is used, the only difference from FIG. 10 is that the potential curves 901, 902, and 903 essentially terminate at the interface of layer 808 and the low impedance substrate. Optional layer 808 is preferably made of a first conductivity type semiconductor material, but it may also be made of a second conductivity type semiconductor material. However, this necessitates a process of etching deep trenches through such optional layers to avoid high leakage current at the edge of the detector chip.

It should be noted that the channel stops in the detectors of fig. 7 and 8 may be floating, which means that the secondary current will flow from the channel stop via the potential barrier formed in the MIG layer towards the bulk layer, where it is collected by the bulk layer contact 725. In the case where the channel stop is floating, the semiconductor material is silicon, silicon dioxide is used as the insulator material, and the first conductivity type is n-type, no channel stop doping (hereinafter, silicon dioxide is referred to as oxide) is required. In this case, the positive oxide charge results in an electron accumulation layer at the silicon dioxide interface. This two-dimensional (2D) electron gas layer serves as a channel stop. The 2D electron or hole gas layer can also be artificially formed at the semiconductor-insulator interface by using a suitable biased MOS structure. In this case, the 2D charge gas layer and the MOS structure form a channel stop. Thus, the channel stop region may be formed by the 2D charge gas layer or the channel stop doping or both. The detectors shown in fig. 7 and 8 may also have gaps in the MIG layer as in the detectors in fig. 1 and 2. If the channel stop and the bulk layer are biased at different potentials, the gap in the MIG layer must be such that no current flows between the bulk layer and the channel stop. The gap in the MIG layer can be arbitrarily wide if the channel stops are not biased at different potentials. In this case, the channel stops or is floating or at the same potential as the bulk layer.

It is important to note that figures 7 to 10 are not drawn to scale, as the bulk layer is in fact much thicker than shown in the figures, i.e. the bulk layer is preferably several hundred microns thick. The bulk layer preferably has a low resistivity, i.e. much higher than the resistance of the almost intrinsic substrate shown in documents PCT/FI2004/000492 and PCT/FI 2005/000359. In the detector of fig. 7-8, a portion of the secondary charges are collected in the active region by the channel stop doping and transferred to the substrate contact 725 through the bulk layer. Due to the front lighting and due to the low resistance substrate, no conductive back side layer is needed. Thus, difficulties associated with fabricating a thin conductive backside layer on the backside of a thin detector and with the operation of such a detector are avoided.

The main difference between the previously described embodiments of back-illuminated and front-illuminated detectors is that front-illuminated detectors are cheaper to manufacture than back-illuminated detectors, but have a smaller fill factor and therefore a smaller quantum efficiency in the visible spectrum than front-illuminated detectors.

Fig. 11 shows an embodiment of the invention in which only a small voltage can be used to clear the signal charge, i.e. the pixel doping does not have to be connected to a clear voltage to remove the signal charge. The lack of a MIG layer in region 1191 means that region 1191 corresponds to a MIG layer mask. The gap 1191 in the MIG layer assists the collection of signal charges by improving the signal charge transport potential gradient in the MIG layer. The channel stop doping 1121 collects secondary charges. There are four pixels in fig. 11, where the cut line 1180 partially cuts two pixels. The cut line 1180 corresponds to the cross section shown in fig. 13. The pixel dopings 1131, 1132 and 1133 of the second conductivity type are the source and drain dopings of a dual Metal Oxide Semiconductor Field Effect Transistor (MOSFET) belonging to one pixel, and the conductors 1341 and 1342 are the gates of the dual MOSFET. Pixel dopings 1335, 1336 and 1337 are source and drain dopings and conductors 1344 and 1345 are the gates of dual MOSFETs belonging to another pixel. The signal charge is collected in an optional local enhancement (enhancement)1392 of the MIG layer doping located below the gate. The local enhancement of the MIG layer doping below the gate of the FET or the emitter of the BJT improves the dynamic range of the detector. The signal charge can be removed inside the pixel between the local enhancements 1392 of MIG layer doping by appropriately biasing the source doping, drain doping and gate. This enables signal charges to be read multiple times, reducing read noise.

There is a pixel doping 1334 of the second conductivity type at the front side of the device that serves as a clear contact. Between the MIG layer and the clear contact 1334 there is a doped region 1393 of the second conductivity type which can be produced by an intermediate energy implantation. Optionally, regions 1334 and 1393 represent trenches filled with semiconductor material of the second conductivity type. The flow of signal charge from the MIG layer through the region 1393 is controlled by a gate 1343. This arrangement enables signal charge to be cleared with a low voltage, and can also be used as an anti-blooming (anti-blooming) structure. Layer 1307 is a protective insulator layer, which is preferably silicon dioxide, but can be any other insulator material. It is important to note that the contact openings through the insulator layer 1307, as well as the contacts, are not shown in fig. 13 for clarity.

Fig. 12 shows another embodiment of the present invention in which signal charges can be cleared using only a small voltage. The channel stop doping 1221 collects the secondary charges. Cut line 1280 corresponds to the cross-section shown by fig. 14. The MOS gate 1343 of fig. 11 is replaced by a doping 1443 of the first conductivity type which serves as a junction gate controlling the flow of signal charges from the MIG layer to the front surface of the detector. The doping 1443 is surrounded by a ring-shaped pixel doping 1433 of the second conductivity type which acts as a source/drain doping for the four double MOSFETs. The second conductivity type doping 1434, which serves as a signal charge clear contact, is connected to the MIG layer using an insulator material 1494 deposited on the trench walls. The insulator material 1494 is preferably positively charged silicon dioxide. Due to the positive oxide charge, a 2D electron gas layer forms at the interface of silicon dioxide and silicon. If the first conductivity type is p-type and the second conductivity type is n-type, the signal charge is electrons. Thus, if the gate 1443 is properly biased, signal charge electrons flow in the 2D electron gas layer from the MIG layer to the doping 1434.

If the insulator layer 1307 is positively charged and if the first conductivity type is p-type and the second conductivity type is n-type, the channel stop doping must be very large like the channel stop doping 1221 in fig. 12. In this case the channel stop doping acts as a radiation entry window. The MIG layer is preferably constructed under a large area channel stop doping 1221. If the insulator layer 1307 is positively charged and if the first type doping is n-type and the second type doping is p-type, the channel stop doping can be very small like the channel stop doping in fig. 11. In this case, a 2D electron gas layer exists at the interface between the insulator layer 1307 and the semiconductor material except in the vicinity of the pixel doping 1331-1337 and the gate 1341-1345. The 2D electron gas layer and the insulator layer 1307 in this case act as a channel stop and as a radiation entry window which can be very thin. In addition, the 2D electron gas layer transfers the secondary charges to the channel stop doping 1121. The large area channel stop doping 1221 can also be made very thin due to the very short secondary charge transport distance in this layer. A thin radiation entrance window enables a good quantum efficiency for blue light.

Fig. 15 shows four pixels of still another embodiment of the present invention, in which signal charges can be cleared using only a small voltage. The ring-shaped channel stop doping 1521 collects the secondary charges. Outside this ring-shaped doping 1521, the positively charged insulator material forms a 2D electron gas layer at the insulator semiconductor interface, which acts as a radiation entrance window and as a channel stop. Area 1591 lacks the MIG layer. The cut lines 1580, 1581, 1582, and 1583 correspond to the sections represented by fig. 16A, 16B, 16C, and 16D. The pixel doping 1632 and the gate 1646 forming the source/drain belong to one pixel. The pixel dopings 1635, 1636, and 1637 forming the source and drain and the gates 1644, 1645, and 1647 belong to another pixel. The signal charge clear doping 1634 is connected to the MIG layer through an insulator layer 1494 covered by a conductor layer 1695. The conductor layer 1695 may be biased such that a 2D charge gas layer is formed at the interface of the insulator and the semiconductor material. Thus, the conductor layer 1695 and the gates 1643, 1646 and 1647 may control the flow of signal charges from the MIG layer to the doping 1634. The gate may also be divided into four different sections belonging to each pixel (this also applies to gate 1343). It is also possible to use only the conductor layer 1695 without the insulator material 1494 if the conductor material is chosen appropriately. In this case, the signal charges may be collected by the conductor layer 1695.

It should be noted that the pixels of fig. 11, 12 and 15 are not drawn to scale. The channel stop region, which comprises the region of the channel stop doping and possibly the region of the 2D charge gas layer, should cover a large part of the total area of the pixel to be able to obtain a good quantum efficiency for blue light. The ratio of the channel stop region belonging to one pixel to the total pixel area should be at least 0.3. Advantageously, this ratio should be greater than the ratios 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, with 0.9 corresponding to the most favorable ratio and 0.4 corresponding to the worst ratio.

It is also worth noting that the gap in the MIG layer is not the only way to improve the signal charge transport potential gradient in the MIG layer. The doping concentration in the barrier layer, the MIG layer or the bulk layer next to the MIG layer can also be varied. For example, the MIG layer doping may be reduced or enhanced by appropriate implantation and masking structures. Locally increasing the barrier layer doping, locally decreasing the MIG layer doping and locally increasing the bulk layer doping next to the MIG layer doping, creating a local potential minimum point of signal charge inside the MIG layer. On the other hand, by locally decreasing the barrier layer doping, by locally increasing the MIG layer doping and by locally decreasing the bulk layer doping next to the MIG layer doping, a local potential maximum of the signal charge can be created in the MIG layer. By suitably constructing the enhancement or reduction in the MIG layer the signal charge transport potential gradient in the IMG layer can be improved in a similar way as the gap in the MIG layer doping. The signal charge transport gradient should be such that an appropriate gradient exists at each of the MIG layers, transporting the signal charge towards the desired location, e.g. the local enhancement 1392 of the MIG layer doping. The local enhancement of the MIG layer doping can also be built by adding dots to the doping such that it resembles a star (star) to increase the signal charge transport potential gradient in the MIG layer. If the ratio of channel stop area to total pixel area is large, it may be forced to use several methods simultaneously instead of one of the above to ensure a sufficiently large signal charge transfer potential gradient in the MIG layer.

Yet another important aspect is that: instead of the doping 1393 of the second conductivity type connecting the MIG layer and the clear contact 1393, a local reduction in the doping of the barrier layer can also be used. This local reduction in barrier layer doping should be located at the same position as doping 1393, i.e., below the clear contact 1393 and surrounded by clear gate 1343. The ratio of the net doping concentration in the first conductivity type locally reduced portion of the barrier layer doping to the net doping concentration of the first conductivity type barrier layer doping without the locally reduced portion of the barrier layer doping should be less than 0.9. Advantageously, the ratio should be less than 0.8, 0.7, 0.6, 0.5, 0.4, 0.3, 0.2 and 0.1, with 0.8 being the worst ratio and 0.1 being the best ratio. Removing the signal charge through a local reduction of the first conductivity type barrier layer doping requires a higher voltage to clear the contact 1393 than through the first conductivity type doping 1393. Thus, doping 1393 is more favorable than a local reduction of barrier layer doping.

The embodiments of fig. 17A, 17B, and 17C illustrate methods of improving separation of signal and surface generated charges and methods of improving collection of secondary charges from bulk layers. Fig. 17D, 17E, 17F, 17G, 17H, and 17I illustrate additional methods of improving the separation of the signal and surface generated charges. Pixel dopings 1731, 1732 and 1733 form the source and drain electrodes and conductors 1741 and 1742 form the gates of the dual MOSFETs. The channel stop doping 1721 collects the secondary charges.

Collection of secondary charges from the bulk layer may be improved by the filled trenches. The trenches in fig. 17A are filled with semiconductor material 1726 of the first conductivity type; the trenches in fig. 17B are filled with insulator material 1727 and the trenches in fig. 17C are filled with insulator material 1727 and conductor material 1728. If the conductor material is chosen appropriately, the insulator material can be removed from the device of FIG. 17C. The semiconductor material 1726 of the first conductivity type may be replaced by a plurality of deep implants of the first conductivity type having different energies. The operation principle of the structures 1726, 1727 and 1728 improving the collection of secondary charges from the bulk layer is similar to the operation principle of the structures 1393, 1494 and 1695 clearing signal charges from the MIG layer. In this case, however, the secondary charges are collected instead of the signal charges. In devices with thin bulk layers, the filled trenches 1726, 1727, and 1728 may pass through the entire bulk layer. The collection of secondary charges can be further improved by surrounding the filled trenches 1726, 1727 and 1728 by the gap 1791 in the MIG layer. The filled trenches 1726, 1727, and 1728 may be any shape; they may for example be cylindrical or they may surround the entire pixel. If the trench is deep enough, the neutral layer potential of figures 5 and 6 can reach the channel stop potential, corresponding to the situation of figure 4.

The dopings 1771, 1772 and 1774 of the second conductivity type, the doping 1775 of the first conductivity type and the gates 1773 and 1776 in fig. 17A to 17F improve the separation of the signal and the surface generated charges. The doping 1771 of the second conductivity type in fig. 17A is preferably depleted resulting in a channel for surface generated charges of the second conductivity type. The channel directs surface generated charges of the second conductivity type toward the pixel dopings 1731 and 1733. The second conductivity type dopant 1774 in fig. 17D surrounds the channel stop dopant 1721. The doping 1774 is separate from the pixel doping, but can reach the pixel doping as well as the doping 1771 in fig. 17A. In this case, the doping 1774 is also preferably depleted. The area of the depleted surface can be controlled by the bias of gate 1773 in fig. 17C. Gate 1776 in fig. 17F can be biased such that a channel of surface generated charge of the second type is formed under the gate, improving the separation of the signal and the surface generated charge. The dopings 1772 and 1775 alter the potential distribution in the device to improve the separation of the signal and surface generated charges. Instead of the second conductivity type doping 1771, a first conductivity type doping, which is preferably partially depleted, may also be used.

In order to prevent the doping 1771 of the second conductivity type from forming a conductive path between the pixel dopings 1731, 1732 and 1733, the doping 1771 is preferably configured. Fig. 17G illustrates an embodiment of such a structure, where doping 1777 corresponds to doping 1771. The cut lines 1780 and 1781 correspond to the cross-sections shown in fig. 17H and 17I. In fig. 17G, the ratio of the minimum distance between doping 1777 and gates 1741 and 1742 to source/drain doping 1732 is advantageously greater than 0.1 times the distance L between the source/drain dopings. Advantageously, the ratio should be greater than 0.2L, 0.3L, 0.4L, 0.5L, 0.6L, 0.7L, 0.8L, 0.9L, L, 1.2L, 1.5L and 2L, with the first being the worse ratio and the last being the best ratio.

The embodiments of fig. 18A, 18B and 18C show a way of improving the dynamic range of the detector, i.e. improving the signal charge capacity of the MIG. The pixel dopings 1831 and 1833 are source/drain dopings and the conductors 1841 and 1842 are the gates of a dual MOSFET. The signal charge capacity of the MIG layer has been improved by local enhancements 1392 of the MIG layer doping. In fig. 18A the signal charge capacity of the MIG is further improved by making the source/drain doping 1832 wider and by increasing the gap 1891 in the MIG layer between the two local enhancements 1392 of the MIG layer doping. In fig. 18B, the signal charge capacity of the MIG is improved by splitting the source/drain doping 1832 into two separate parts 1834 and 1835 and by adding a gate 1843 in between. In fig. 18C, the signal charge capacity is further improved by adding a doping 1836 of the second conductivity type and two gates 1844 and 1845 between the two dopings 1834 and 1835.

The MOSFET is not the only possible transistor used in conjunction with the MIG. In fig. 19, the MOSFET is replaced by a BJT. The pixel dopings 1931 and 1932 of the second conductivity type are base dopings and the dopings 1951 and 1952 of the first conductivity type are emitter dopings of the BJT. The channel stop doping 1921 of the first conductivity type serves as a collector of the BJT for collecting charges of the first conductivity type emitted by the emitter. In addition to the three conventional nodes of the BJT, there is a fourth node, i.e., the MIG. The signal charge in the MIG reduces the effective base width. Thus, the signal charge in the MIG increases the emitter current. This effect can be measured and the amount of signal charge can be deduced from the measurement. The pixel doping 1931 includes an additional bend (buckling) 1974. Below the channel stop doping 1321 there is also a local enhancement 1929 of the barrier layer which increases the electric field component in the MIG layer which transports the signal charge towards the local enhancement 1392 of the MIG layer doping. The topography of the local enhancement of the barrier layer doping can be built up in the same way as the gaps 1191 and 1591 in the MIG layer doping.

In fig. 20, the MOSFET is replaced by a Junction Field Effect Transistor (JFET), with the gate doping replaced by MOS gates 2041 and 2042. The pixel dopings 2031 and 2032 serve as source, drain and channel dopings. The pixel dopings also have a curvature 2075.

In all transistors including MIGs that have been described so far, the signal charge in the MIG reduces the effective channel or base width. Fig. 21 and 22 show transistors comprising a MIG in which the signal charge in the MIG increases the effective channel or base width. In fig. 21, the pixel dopings 2131 and 2132 act as collector dopings and the dopings 2151 and 2152 of the first conductivity type act as base dopings. Emitters 2161 and 2162 are formed of, for example, polycrystalline semiconductor material like polysilicon. In fig. 22, the pixel dopings 2206 are continuous layers that seal the channel stop dopings 2221. Inside the pixel doping 2206 of the first conductivity type, there are also source and drain dopings 2251, 2252, 2253 and 2254 of two MOSFETs. Conductors 2241 and 2242 are the gates of the two MOSFETs.

Fig. 23 and 24 show a semiconductor device which can be used as a memory cell or a transistor. The dopings 2331 and 2332 of the second conductivity type are the drain and source and the conductor 2341 is the gate of a MOSFET. The gate 2342 controls the flow of signal charges from the doping 2333 of the second conductivity type through the region 1393 of the second conductivity type to the MIG layer. In fig. 24, the MOSFET is replaced by a BJT having a base 2431 of the second conductivity type and an emitter 2451 of the first conductivity type. In the device of fig. 24, the filled trenches 1494, 1695 have the same function as the doped region 1393 in fig. 23. If the device of fig. 23 and 24 is used as a memory cell, the MIG filled with signal charge and the MIG without signal charge correspond to 1 and 0 and vice versa. If the devices in fig. 23 and 24 are used as transistors, the MIG layer can be very heavily doped, i.e. it can be neutral inside it, and it forms a fourth node in the transistor in addition to the source, drain and gate in the FET and to the emitter, base and collector in the BJT.

There is another way to operate a MIG detector and a new MIG detector not described before. In this case, the source, drain and gate potentials of the FET are such that the channel under the gate is closed, i.e. there is no current path between the source and drain dopings. If the source and drain dopings are at the same potential, only one pixel doping may be used instead of two separate dopings (see, e.g., fig. 25). When e.g. a light pulse is absorbed in the detector, the signal charge will flow to the MIG. This will generate a current pulse in the gate and this current pulse can be used for precise timing of the incidence (incident). A new MIG detector corresponding to the gate signal detection mode is shown in fig. 25 to 31. The cut lines 2580 in fig. 25 and 2680 in fig. 26 correspond to the cross-section shown in fig. 27A. The potential of the optional channel stop doping 2721 of the pixel doping 2731 and gate 2741 causes the semiconductor insulator interface under the gate to be depleted so that a gate signal can be formed. If the semiconducting-insulator interface is locked at the pixel doping potential, i.e. if there is a channel under the gate, no signal or only a very weak signal will be generated in the gate when a number of signal charges reach the MIG. For example, the signal charge may be cleared by applying a clear bias between the channel stop doping and the pixel doping.

The gate signal MIG detector of fig. 25 can be coupled to a read-out chip. In this case, due to the pixelated detector structure, both temporal and 2D position information can be obtained. Another possibility is to connect the gates of a row or column of pixels by metal lines to form a strip detector. The strip detector is capable of detecting both time and one-dimensional (1D) position information. Yet another possibility is to divide the gate into two or three different sections and to connect each section to a different signal line pointing in a different direction to achieve time and 2D position information. The gate signal MIG detector of fig. 26 is a strip detector and the cutting line 2680 corresponds to the cross section shown in fig. 27A. In fig. 27B, the gate 2741 of the detector of fig. 26 is surrounded by an insulator layer 2707 on top of which is a metal cap 2742. Optional metal cover 2742 further reduces noise of the detector.

The strip detectors in fig. 28 to 31 represent different embodiments of gate signal MIG detectors. In fig. 28, the gate 2841 is connected to a second layer metal 2842 for connecting the gates in a row or column of pixels. In this case, the pixel doping is divided into two parts 2831, 2832. The detector in figure 29 is the same as in figure 26 except that the pixel doping 2731 is divided into a plurality of parts 2931, 2932. The dashed lines in fig. 29 correspond to the pixel dopings located under the gates. The same applies to fig. 30 and 31. The detector in figure 30 has only one pixel doping 3031, next to which pixel doping 3031 are two gates 3041 and 3042. In fig. 31, there is only one gate electrode 3141 and a plurality of pixel dopings 3131. The enhancement of the MIG layer doping may be located below the gate in e.g. fig. 28 to 31.

A prior art strip probe is shown in fig. 35A and 35B. In the structure of fig. 35A, signal charges are collected by the doped portion 3531, which changes the potential of the doped portion 3531. This generates a signal to gate 3541, which gate 3541 is held at a constant potential. However, the doped portion 3531 collects surface-generated current in addition to the bulk-generated current. The leakage current also produces the same amount of current in the gate. The noise in the detector may be equal to the square root of the leakage current. In such a strip detector, the surface generated current is typically about 20 times higher than the bulk generated current, and thus the noise of the strip detector in fig. 35A is high. In the prior art device of fig. 35B, dopants 3532 and 3533 collect surface generated charge, which means that dopant 3531 only collects bulk generated current. However, the dopings 3532 and 3533 also collect a portion of the signal charges. Thus, the signal-to-noise ratio in the strip detector of FIG. 35B is typically worse than the signal-to-noise ratio of the strip detector of FIG. 35A. The signal-to-noise ratio of the strip detector in fig. 35A can easily be compared with the signal-to-noise ratio of the gate signal MIG detector. The ratio of parasitic capacitance to total capacitance of the device in fig. 35A is close to 0. In a gate signal MIG device the ratio of parasitic capacitance to total capacitance is about 0.5, resulting in a signal that is half smaller than in prior art devices. However, in MIG detectors the pixel dopings collect surface generated charges and the MIG only collects bulk generated currents. The noise in the prior art device of fig. 35A is thus a factor of the square root of 20 higher than in the MIG detector. Thus the signal-to-noise ratio in the MIG detector is estimated to be 2.2 times higher than the prior art detector of fig. 35A.

The signal charges may also be transferred to the readout pixels through a plurality of pixels, where the amount of the signal charges is measured. The device shown in fig. 44A is an example of such a device, which operates in the same manner as a Charge Coupled Device (CCD). When the potentials of the pixel dopings 4431, 4432 and 4433 of the first conductivity type are changed cyclically, signal charges can be transferred in the MIG layer 104. The pixel dopings may also work as an anti-blooming structure. The signal charges in the MIG layer can be removed by applying a clear voltage between the channel stop doping 4421 and the pixel doping. If the device is front-illuminated, the size of the channel stop region should be large to have good quantum efficiency for blue light. The cutting lines 4480 and 4481 correspond to the cross sections shown in fig. 44B and 44C.

Fig. 32B to 32D, 33A to 33C and 34A to 34D show examples of different manufacturing methods of the new MIG detector. The starting point of these processes is the bare substrate 103 of the first conductivity type shown in fig. 32A. The manufacturing process shown in fig. 32B to 32D is similar to CMOS processing. In fig. 32B, two mask processes, two second type implants and drive in are performed to form a second conductive type well 3204 and a doped region 3292. The well 3204 is used to form the MIG layer and the optional doped region 3292 is used to form an enhancement of the MIG layer doping. In fig. 32C, a mask process, an implant of the first conductivity type, and a drive-in are performed to form the well 3205 of the first conductivity type. The well 3205 serves as a barrier layer. In fig. 32D, at least one mask process and implantation are performed to form the pixel dopings 3231 of the second conductive type. The doping 3234 of the second conductivity type is an optional signal charge clear contact. At this stage, other masking and implantation steps may also be performed to form, for example, channel stops, substrate contacts, and other doped regions as previously described. Thereafter, an annealing step is performed, and then an insulator and a metal layer and a via hole through the insulator layer are formed.

A method of manufacturing the new MIG detector involving deep implantation is shown in fig. 33A to 33C. In fig. 33A, one mask process, one first conductive type implant and drive-in are performed to form a first conductive type well 3305. Well 3305 serves as a barrier layer. In fig. 33B two masking procedures and two high-energy deep implants of the second conductivity type are performed to form the MIG layer 3304 and the optional enhancement 3392 of the MIG layer doping. In fig. 33C, a mask process and a second type implant are performed to form pixel dopings 3331. The second conductivity type doping 3334 is an optional signal charge clear contact. Region 3396 is a local reduction of the net barrier layer doping that is located under the clear contact. Other masking and implantation steps may also be performed at this stage. Next, an annealing step is performed, and then an insulator and a metal layer and a via hole through the insulator layer are formed. The well 3305 forming the barrier layer may be well performed using an intermediate energy implant.

The manufacturing process illustrated in fig. 34A to 34D is similar to a BiCMOS process. In fig. 34A two masking steps and two implantations of the second conductivity type and an optional annealing step are performed to form the MIG layer 3404 and the optional enhancements 3492 of the MIG layer doping. In fig. 34B, an epitaxial layer 3405 of the first conductivity type is grown on top of the semiconductor substrate 103. The epitaxial layer 3405 forms a barrier layer. In fig. 34C, a masking step and a first type implant are performed on epitaxial layer 3405 to form pixel dopings 3431 and optional clear contacts 3434. At this stage, other masking and implantation steps may be performed to form, for example, a channel stop. An optional mask and a second conductivity type intermediate energy implantation step to form a second type doping 3493 between the clear contact 3434 and the MIG layer 3404 is shown in fig. 34D. It should be noted that if the amount of this second type intermediate energy implant is low, only a reduced portion of the net barrier layer doping is created under the clear contact 3434.

It should be noted that the previously shown method of manufacturing the new MIG detector is only an example. In addition to this, there are a number of other methods. The different steps of the different methods described before can be combined in any suitable way or order. Although the substrate contact and the channel stop doping are not shown in fig. 32A to 34D, they may be added to the process flow under appropriate conditions, as already stated. For example, a thin backside-illuminated device can be fabricated from the devices of fig. 32D, 33C, and 34D by grinding the backside of the bulk layer 103 or by fabricating the devices of fig. 32D, 33C, and 34D on an SOI wafer. An SOI wafer has two semiconductor layers and has an insulator layer between them. After processing the other semiconductor surfaces, the other side of the semiconductor wafer can be etched under the active region of the probe until the insulator layer is reached. Thereafter, the insulator layer may be etched away, after which the rear side of the processed semiconductor layer, i.e. the bulk layer, may be covered with, for example, an anti-reflection coating.

The first conductivity type may be n-type and the second conductivity type may be p-type. The embodiments of fig. 11-34D and 44A-44C can be used for both front and rear illuminated detectors, and any combination of them can be used. It is important to note that the embodiments and processes shown in fig. 11-34D and 44A-44C can also be used in MIG detectors as shown by the documents PCT/FI2004/000492 and PCT/FI2005/000359 and having a conductive back side layer. The pixels may have any shape or form instead of the pixels shown in fig. 11, 12, and 15. Instead of a two-transistor pixel, a single-transistor or multi-transistor pixel may be used. However, the speed of reading signal charges multiple times in a two-transistor pixel is twice that in a one-transistor pixel. Instead of MOSFETs, JFETs and BJTs, unipolar or bipolar transistors may be used in the pixels. The source of the FET or the emitter of the bipolar transistor may be floating and may be connected to a capacitor. The pixel may be surrounded by a preferably ring-shaped protective structure formed by a MOS structure or doping to increase the pixel area. The inventive doping can also be made in any possible way with different masks, different energies, different doses, different conductivity types (tailor). In some cases, the doping may also be replaced by an appropriate metal contact (i.e., with an ohmic or schottky type contact). The semiconductor material is preferably silicon, but any other semiconductor material may be used. For example, the semiconductor material may be germanium. Contact openings through the insulator layer 1307 and contacts to different dopings are not shown. The channel stop dopings are optional in the devices of fig. 7 and 8, and they may be floating. Anti-reflection coatings, scintillator coatings or microlenses may be used in front-and back-illuminated detectors.

The amount of signal charge in the MIG of a MIGFET can be obtained for example by measuring a change in threshold voltage, by measuring a change in current flowing through the MIGFET, or by measuring a change in voltage output through a known resistor, which change in voltage output corresponds to a change in current flowing through the MIGFET. For example, the amount of signal charge in the MIG of a MIGBJT can be obtained by measuring the change in emitter current or by measuring the change in voltage output across a known resistor, which corresponds to the change in current flowing through the emitter, or by measuring the change in base or emitter threshold. The base threshold is referred to as the base voltage at which the emitter current starts to flow. The emitter threshold is referred to as the emitter voltage at which the emitter current starts to flow. Other signal charge readout schemes also exist, and all readout schemes may involve, for example, capacitors, transistors, resistors, and the like.

It is important to note that the MIG enables a very small amount of signal charge to be detected. This can be done by taking a measurement when there is signal charge in the MIG, by removing signal charge from the MIG, by taking a measurement when there is no signal charge, by subtracting the first measurement from the second measurement and taking n times. As a result, the read noise will be that of one measurement divided by the square root of n. However, this is not the only way to detect a small amount of signal charge. The new MIG detector (and also the MIG detector) can also be designed in such a way that a suitable operating voltage delivers the signal charge from the MIG and back to the MIG, resulting in an avalanche multiplication of the signal charge. This avalanche multiplication cycle can be performed N times, after which the signal charge is multiplied to Nxm ^ N, where m is the avalanche multiplication gain for a single signal charge transfer. The first of the two methods enables a higher dynamic range. However, the two methods may also be combined, i.e. the first method may be performed first and then the second method. The first method, i.e., the multiple reading method, is performed with applying a lower bias voltage; and a second method, namely, an avalanche gain method is performed with a higher applied bias. This combined approach has the same dynamic range as the multiple read approach.

Groups of four pixels are shown in fig. 11, 12 and 15, and the front or rear surface of each pixel may be covered by a color filter and may be covered with one or more microlenses. For example, the uppermost and lowermost pixels may have green filters, and the left and right pixels may have blue and red filters, respectively. The front or back surface of the detector in fig. 11, 12 and 15 may also be covered by a single color filter and possibly with a microlens. In this case, the light is preferably divided into three different components, e.g. red, green and blue, and preferably three separate chips are available for the camera. It should be noted that the inventive detector should be designed such that no neutral zone is present in the barrier layer between the channel of the FET and the MIG layer and between the base of the BJT and the MIG layer, since such neutral zones increase the noise in the measurement. It should also be noted that all of fig. 1-35B are not to scale and that all of the gate and metal layers shown in fig. 1-35B are formed of transparent conductors. Self-aligned structures are also advantageously used to reduce misalignment of the mask. Instead of square FETs, ring FETs may also be used. The figures are not drawn to scale so that the area of the channel stop doping can be much larger than that shown in the figures. Additional layers may also be present in the device if the principle of operation of the MIG is not affected. Such an additional layer may be, for example, a thin layer of semiconductor material of the first or second conductivity type.

There may also be read and select electronics on the detector chip. Devices including detectors according to embodiments of the invention may also include other semiconductor chips, some of which may have bond connections to pixels of the detector. This enables very compact structures to be built, including probing, amplifying, reading, and in some cases even very small volume storage, e.g., MCM (multi-chip module).

The 2D simulation results of fig. 36 to 43 demonstrate the feasibility of the MIG detector concept. Fig. 36 shows a MIG detector with an n-type bulk layer 103, a p-type MIG layer 104 and an n-type barrier layer 105. P-type pixel dopings 3631, 3632 and 3633 serve as source/drain dopings, gates 3641 and 3642 are used to measure and transfer signal charge, and channel stop contact 3621 is used for signal charge removal. The enhancement 3692 of the MIG layer doping collects the signal charge, which in this case is holes. The case shown in fig. 36 is a case after the signal charges are cleared by applying a clear voltage to the channel stop contact. Fig. 37 shows the situation after accumulation of some bulk generated holes in the local enhancement of the MIG layer doping. In fig. 38 all holes in the left local enhancement of the MIG layer doping are moved to the right local enhancement of the MIG layer doping by applying appropriate voltages instantaneously to the pixel doping and the gate. In fig. 39 the holes in the right enhancement of the MIG layer doping are moved to the left enhancement of the MIG layer doping by applying appropriate voltages momentarily to the pixel doping and the gate. It is important to note that all potentials in fig. 36-39 are the same; only the location of the cavities is different. Fig. 40 shows the hole concentration in the two local enhancements of the MIG layer doping and fig. 41 shows the hole concentration in the combination of the local enhancements of the MIG layer doping.

In fig. 42, the potentials of the p-type pixel doping 4233, the p-type clear contact 4234, the n-type clear gate 4243, and the n-type doping 4221 are such that holes accumulate in the MIG layer under the p-type pixel doping 4233. In fig. 43 the clear gate potential is changed and holes can move freely from the MIG layer to the clear contact through the p-type doping 4293. In addition to the effects described above, simulations show that the change in threshold voltage can be greater than 100 μm.

Claims

1. A semiconductor radiation detector device comprising a bulk layer (103) of semiconductor material and arranged on a first surface of the bulk layer (103) in the following order:

a modified internal gate layer (104, 304) of a semiconductor of a second conductivity type;

a barrier layer (105) of a semiconductor of a first conductivity type; and

-pixel dopings (111, 112, 1331, 1332, 1333, 1334, 2206) of a semiconductor of a second conductivity type adapted to be connected to at least one pixel voltage to create pixels corresponding to the pixel dopings, characterized in that:

the device comprises a first contact of a first conductivity type, the pixel voltage being defined as a potential difference between the pixel doping and the first contact, and,

the bulk layer (103) is of a first conductivity type.

2. A semiconductor radiation detector device according to claim 1, wherein a plurality of pixel dopings (111, 112, 1331, 1332, 1333, 1334, 2206) comprises pixel-specific transistors built on the pixel dopings, the transistors being field effect transistors or bipolar transistors, and the semiconductor radiation detector device comprises a signal charge reader circuit adapted to measure an electrical characteristic of a pixel-specific transistor related to an effective channel or base size of the pixel-specific transistor.

3. A semiconductor radiation detector device according to claim 2, wherein the signal charge reader circuit is adapted to measure electrical characteristics of pixel-specific transistors associated with a reduced channel or base width caused by radiation-induced holes or electrons accumulated in the modified internal gate layer at locations coinciding with pixels containing the pixel-specific transistors.

4. A semiconductor radiation detector device according to claim 2, wherein the signal charge reader circuit is adapted to measure electrical characteristics of pixel-specific transistors associated with increased channel or base width caused by radiation-induced electrons or holes accumulated in the modified internal gate layer at locations coinciding with pixels containing the pixel-specific transistors.

5. A semiconductor radiation detector device according to any of claims 1-4, comprising channel stop regions between pixels.

6. A semiconductor radiation detector device according to claim 5, wherein said channel stop region comprises a doping (121, 122, 1121, 1221, 1521, 1721) of a first conductivity type, thereby exhibiting an opposite type of conductivity compared to said pixel.

7. A semiconductor radiation detector device according to claim 5, wherein the first contact (725) is on the back surface (102) of the detector, or outside the active area on the front surface (101) of the detector, or on the edge (700) of the detector chip.

8. A semiconductor radiation detector device according to claim 5, wherein the detector is illuminated from the first surface.

9. A semiconductor radiation detector device according to claim 5, wherein there is a doped region (1393) of second conductivity type in the barrier layer (105) of the first conductivity type, or a local reduction (3396) of barrier layer doping between the modified internal gate layer (104) of second conductivity type and pixel doping (1334, 1434, 1634, 3334) of second conductivity type, or wherein there is a trench between the modified internal gate layer (104) and the front surface of the detector.

10. A semiconductor radiation detector device according to claim 5, wherein a region or trench structure of a first conductivity type passes through the modified internal gate layer of a second conductivity type to improve the collection of secondary charges from the bulk layer (103).

11. A semiconductor radiation detector device according to claim 5, comprising structures that improve separation of signal and surface generated charges.

12. A semiconductor radiation detector device according to claim 5, wherein the modified internal gate layer is discontinuous.

13. A semiconductor radiation detector device according to claim 5, comprising at least one of: a change in barrier layer doping, a gap in modified internal gate layer doping, an enhancement of the modified internal gate layer doping, a change in the bulk layer doping immediately adjacent the modified internal gate layer to improve a signal charge transport potential gradient in the modified internal gate layer.

14. A semiconductor radiation detector device according to claim 5, comprising a local enhancement (1392) of the modified internal gate layer doping under the gate of a field effect transistor or under the emitter of a bipolar transistor to improve the dynamic range of the detector.

15. A semiconductor radiation detector device according to claim 6, wherein channel stop dopings between the pixels correspond to the first contacts.

16. A semiconductor radiation detector device according to claim 15, wherein secondary charges generated in the bulk layer are collected by the channel stop doping.

17. A semiconductor radiation detector device according to claim 16, wherein the bulk layer is thinned from a second surface opposite the first surface and the semiconductor radiation detector device is illuminated from a back surface (102).

18. A semiconductor radiation detector device according to any of claims 1-4 and 6, wherein the first contact (725) is on the back surface (102) of the detector, or outside the active area on the front surface (101) of the detector, or on the edge (700) of the detector chip.

19. A semiconductor radiation detector device according to claim 18, wherein secondary charges generated in the bulk layer are collected by the first contact (725).

20. A semiconductor radiation detector device according to claim 18, wherein the detector is illuminated from the first surface.

21. A semiconductor radiation detector device according to claim 18, wherein there is a doped region (1393) of second conductivity type in the barrier layer (105) of the first conductivity type, or a local reduction (3396) of barrier layer doping between the modified internal gate layer (104) of second conductivity type and pixel doping (1334, 1434, 1634, 3334) of second conductivity type, or wherein there is a trench between the modified internal gate layer (104) and the front surface of the detector.

22. A semiconductor radiation detector device according to claim 18, wherein a region or trench structure of a first conductivity type passes through the modified internal gate layer of a second conductivity type to improve the collection of secondary charges from the bulk layer (103).

23. A semiconductor radiation detector device according to claim 18, comprising structures that improve separation of signal and surface generated charges.

24. A semiconductor radiation detector device according to claim 18, wherein the modified internal gate layer is discontinuous.

25. A semiconductor radiation detector device according to claim 18, comprising at least one of: a change in barrier layer doping, a gap in modified internal gate layer doping, an enhancement of the modified internal gate layer doping, a change in the bulk layer doping immediately adjacent the modified internal gate layer to improve a signal charge transport potential gradient in the modified internal gate layer.

26. A semiconductor radiation detector device according to claim 18, comprising a local enhancement (1392) of the modified internal gate layer doping under the gate of a field effect transistor or under the emitter of a bipolar transistor to improve the dynamic range of the detector.

27. A semiconductor radiation detector device according to any of claims 1-4, 6, 15-16, and 19, wherein said detector is illuminated from said first surface.

28. A semiconductor radiation detector device according to claim 27, wherein there is a doped region (1393) of second conductivity type in the barrier layer (105) of the first conductivity type, or a local reduction (3396) of barrier layer doping between the modified internal gate layer (104) of second conductivity type and pixel doping (1334, 1434, 1634, 3334) of second conductivity type, or wherein there is a trench between the modified internal gate layer (104) and the front surface of the detector.

29. A semiconductor radiation detector device according to claim 27, wherein a region or trench structure of a first conductivity type passes through the modified internal gate layer of a second conductivity type to improve the collection of secondary charges from the bulk layer (103).

30. A semiconductor radiation detector device according to claim 27, comprising structures that improve separation of signal and surface generated charges.

31. A semiconductor radiation detector device according to claim 27, wherein the modified internal gate layer is discontinuous.

32. A semiconductor radiation detector device according to claim 27, comprising at least one of: a change in barrier layer doping, a gap in modified internal gate layer doping, an enhancement of the modified internal gate layer doping, a change in the bulk layer doping immediately adjacent the modified internal gate layer to improve a signal charge transport potential gradient in the modified internal gate layer.

33. A semiconductor radiation detector device according to claim 27, comprising a local enhancement (1392) of the modified internal gate layer doping under the gate of a field effect transistor or under the emitter of a bipolar transistor to improve the dynamic range of the detector.

34. A semiconductor radiation detector device according to any of claims 1-4, 6, 15-17, and 19, wherein there is a doped region (1393) of second conductivity type in the barrier layer (105) of the first conductivity type, or a local reduction (3396) of barrier layer doping between the modified internal gate layer (104) of second conductivity type and the pixel doping (1334, 1434, 1634, 3334) of second conductivity type, or wherein there is a trench between the modified internal gate layer (104) and the front surface of the detector.

35. A semiconductor radiation detector device according to claim 34, wherein a region or trench structure of a first conductivity type passes through the modified internal gate layer of a second conductivity type to improve the collection of secondary charges from the bulk layer (103).

36. A semiconductor radiation detector device according to claim 34, comprising structures that improve separation of signal and surface generated charges.

37. A semiconductor radiation detector device according to claim 34, wherein the modified internal gate layer is discontinuous.

38. A semiconductor radiation detector device according to claim 34, comprising at least one of: a change in barrier layer doping, a gap in modified internal gate layer doping, an enhancement of the modified internal gate layer doping, a change in the bulk layer doping immediately adjacent the modified internal gate layer to improve a signal charge transport potential gradient in the modified internal gate layer.

39. A semiconductor radiation detector device according to claim 34, comprising a local enhancement (1392) of the modified internal gate layer doping under the gate of a field effect transistor or under the emitter of a bipolar transistor to improve the dynamic range of the detector.

40. A semiconductor radiation detector device according to claim 34, wherein a gate (1343, 1643, 1695) is adapted to control a signal charge flow from the modified internal gate layer (104) to the pixel doping (1334, 1434, 1634, 3334) or from the modified internal gate layer (104) through the doping region (1393) of the second conductivity type, through the local reduction of the barrier layer doping (3396) or through a trench to the front surface of the detector.

41. A semiconductor radiation detector device according to claim 40, wherein the ratio of the net doping concentration of the first conductivity type local reduction of the barrier layer doping to the net doping concentration of the first conductivity type barrier layer doping without the local reduction of the barrier layer doping is less than 0.9.

42. A semiconductor radiation detector device according to claim 40, wherein the gate is formed by a doping (1443) of the first conductivity type.

43. A semiconductor radiation detector device according to claim 40, wherein the gate is formed of a MOS structure.

44. A semiconductor radiation detector device according to any of claims 1-4, 6, 15-17, 19, 40, and 42-43, wherein a region or trench structure of a first conductivity type passes through the modified internal gate layer of a second conductivity type to improve the collection of secondary charges from the bulk layer (103).

45. A semiconductor radiation detector device according to claim 44, comprising structures that improve separation of signal and surface generated charges.

46. A semiconductor radiation detector device according to claim 44, comprising at least one of: a change in barrier layer doping, a gap in modified internal gate layer doping, an enhancement of the modified internal gate layer doping, a change in the bulk layer doping immediately adjacent the modified internal gate layer to improve a signal charge transport potential gradient in the modified internal gate layer.

47. A semiconductor radiation detector device according to claim 44, comprising a local enhancement (1392) of the modified internal gate layer doping under the gate of a field effect transistor or under the emitter of a bipolar transistor to improve the dynamic range of the detector.

48. A semiconductor radiation detector device according to any of claims 1-4, 6, 15-17, 19, 40, and 42-43, comprising structures that improve separation of signal and surface generated charges.

49. A semiconductor radiation detector device according to claim 48, comprising at least one of: a change in barrier layer doping, a gap in modified internal gate layer doping, an enhancement of the modified internal gate layer doping, a change in the bulk layer doping immediately adjacent the modified internal gate layer to improve a signal charge transport potential gradient in the modified internal gate layer.

50. A semiconductor radiation detector device according to claim 48, comprising a local enhancement (1392) of the modified internal gate layer doping under the gate of a field effect transistor or under the emitter of a bipolar transistor to improve the dynamic range of the detector.

51. A semiconductor radiation detector device according to any of claims 1-4, 6, 15-17, 19, 40, and 42-43, wherein the modified internal gate layer is discontinuous.

52. A semiconductor radiation detector device according to claim 51, comprising at least one of: a change in barrier layer doping, a gap in modified internal gate layer doping, an enhancement of the modified internal gate layer doping, a change in the bulk layer doping immediately adjacent the modified internal gate layer to improve a signal charge transport potential gradient in the modified internal gate layer.

53. A semiconductor radiation detector device according to claim 51, comprising a local enhancement (1392) of the modified internal gate layer doping under the gate of a field effect transistor or under the emitter of a bipolar transistor to improve the dynamic range of the detector.

54. A semiconductor radiation detector device according to claim 1, comprising a layer of semiconductor material of the first or second conductivity type between the bulk layer and the modified internal gate layer.

55. A semiconductor radiation detector device according to any of claims 1-4, 6, 15-17, 19, 40, 42-43 and 54, comprising at least one of: a change in barrier layer doping, a gap in modified internal gate layer doping, an enhancement of the modified internal gate layer doping, a change in the bulk layer doping immediately adjacent the modified internal gate layer to improve a signal charge transport potential gradient in the modified internal gate layer.

56. A semiconductor radiation detector device according to claim 55, comprising a local enhancement (1392) of the modified internal gate layer doping under the gate of a field effect transistor or under the emitter of a bipolar transistor to improve the dynamic range of the detector.

57. A semiconductor radiation detector device according to any of claims 1-4, 6, 15-17, 19, 40, 42-43 and 54, comprising a local enhancement (1392) of the modified internal gate layer doping under the gate of a field effect transistor or under the emitter of a bipolar transistor to improve the dynamic range of the detector.

58. A semiconductor radiation detector device according to claim 5, wherein the ratio of the channel stop region to the total pixel area of one pixel is at least 0.3.

59. A semiconductor radiation detector device according to claim 34, wherein the ratio of the net doping concentration of the first conductivity type local reduction of the barrier layer doping to the net doping concentration of the first conductivity type barrier layer doping without the local reduction of the barrier layer doping is less than 0.9.

60. A semiconductor radiation detector device according to claim 1, wherein the channel of the field effect transistor below the gate of the field effect transistor is depleted, and wherein the pulse of signal charge generated by the radiation and entering the modified internal gate layer is detected as a current pulse in the gate.

61. A method for detecting radiation, comprising:

connecting a plurality of pixels on a surface of the semiconductor radiation detector device to a pixel voltage, an

Illuminating the semiconductor radiation detector with radiation;

it is characterized in that it comprises:

collecting first radiation-induced signal charges from a bulk layer (103), a modified internal gate layer, and a barrier layer of the semiconductor radiation detector to local minima of a three-dimensional potential function of the first said charges that are positionally coincident with pixels (111) on a modified internal gate layer (104, 304) located immediately adjacent to the bulk layer (103), and

the amount of signal charges collected in the local minimum portion corresponding to the pixel (111) is detected.

62. The method of claim 61, wherein detecting the amount of signal charge comprises: the electrical characteristics of the pixel specific transistors are observed in relation to their effective channel or base size.

63. The method of claim 62, wherein detecting the amount of signal charge comprises: observing an electrical characteristic of the pixel-specific transistor related to the reduced channel or base size of the pixel-specific transistor.

64. The method of claim 63, wherein detecting the amount of signal charge comprises: observing an electrical characteristic of the pixel specific transistor associated with the increased channel or base size of the pixel specific transistor.

65. The method of claim 64, wherein detecting the amount of signal charge comprises: pixel-dependent charge is transferred through a plurality of pixels to a readout pixel, and the electrical characteristics of the readout pixel are observed.