JP5661877B2 - Sensor device - Google Patents

Description

  The present invention relates to sensors, and more specifically to sensor devices for use in applications requiring rapid sensing, such as those used in microscopes.

  Different types of radiation sensors are known for different applications, such as optical sensors, X-ray sensors and the like. Depending on the wavelength of radiation to be sensed and the rate of change, different types of sensors are suitable. One specific area for light sensing is confocal microscopy.

  The concept behind a confocal microscope as used in a fluorescence microscope is as follows. Fluorescent dye molecules can be attached to specific portions of the biological sample of interest. When stimulated by light of the appropriate wavelength, the markers fluoresce and only those parts can be observed with a microscope. Fluorescence is usually stimulated by a single or multiple laser beams and detected by a suitable sensor, usually a photomultiplier tube. The sample is scanned by the illumination / detector and a fluorescent image is built across the sample. In order not to detect light from an unintended area of the sample, it is necessary to install a very small hole (pinhole) in the optical path. One arrangement of a typical confocal microscope system is shown in FIG.

  The laser scanning microscope apparatus of FIG. 1 includes a laser 10 arranged to irradiate a sample 12 via a dichroic mirror and two scanning mirrors 16 and 18. With the scanning mirror, the laser beam scans the surface of the sample 12. The imaging optical system 20 is arranged so that the scanning point of the sample light forms an image on the pinhole 22. This arrangement is referred to as a confocal arrangement and ensures that only fluorescence from the particular point being scanned reaches the detector 24. Any light from deep within the sample that is not on the focal point of the optical system does not image accurately on the pinhole (forms a pattern larger than the pinhole) and therefore does not reach the detector 24. In the confocal arrangement, different parts of the sample can thus be sampled over a long period (during the beam scan), which is one form of time-resolved microscope.

  In the confocal microscope embodiment, it is important to be able to quickly sample the light level so that individual points are sampled as the laser scans the surface of the sample.

  One type of sensor known as the Silicon Drift Detector is known to have certain speed advantages in applications such as microscopes. The concept of the silicon drift detector was described in 1983 by E. Gatti and P. Rehak, “Semiconductor Drift Chamber-An application of Novel Cgarge Transport Scheme”, Nucl. Instr. And Meth. A 225, 1984, pp. 608-614. Proposed in It consists of a certain amount of completely depleted high-resistance silicon, where an electric field with a strong component parallel to the surface drives the electrons generated by the absorption of ionizing radiation towards a small current collector anode. Yes. The electric field is created by a number of increasingly energized reversely energized strips covering one surface of the device (creating a pn diode). This concept is illustrated in FIG.

  Various improvements have been proposed, including concentric arrangements where the sensing node is in the center of the device. These devices find special applications in the field of X-ray spectroscopy. The structure is most often based on an improvement with respect to the arrangement shown in FIG. 17, where an appropriate electric field is applied through the diode contacts to sweep all charges out of the device.

  Known variations of Silicon Drift Detectors are known from various previously published literature.

U.S. Pat. No. 4,688,067 describes the operation of a silicon drift detector. This patent shows how to remove charge to the sense node using a bias applied to the p ⁺ contact.

  WO 2006/012764 discloses one type of variable dimension variable aperture sensor based on parallel electrodes arranged to change the effective "slit" size.

  U.S. Pat. No. 4,837,607 improves upon electrode placement.

EP 0383389 describes a modification of the SSD, where the signal produced through the p ⁺ contact is used to obtain position information by measuring the arrival time of incident light.

  U.S. Pat. No. 6,249,033B1 describes the complexity of basic SDD to obtain location information without the need to measure time.

  US 2004/0149919 A1 describes a modification to the SDD for improvements such as response uniformity.

  US Pat. No. 6,794,654 B1 describes an SDD module.

  US 2005/0173733 A1 describes a method of contacting a sensing node in the center of a concentric arrangement.

  WO 2006/053938 A1 describes a very special modification to the SDD to prevent the effects of surface leakage.

US Pat. No. 4,688,067 WO2006 / 012764 US Pat. No. 4,837,607 EP 0383389 US Pat. No. 6,249,033 Patent No. 2004 / 0149919A1 US Pat. No. 6,794,654 B1 US 2005/0173733 A1 WO2006 / 053938A1 E. Gatti, P. Rehak “Semiconductor Drift Chamber-An application of Novel Cgarge Transport Scheme”, Nucl. Instr. And Meth. A 225, 1984, pp. 608-614

  While the benefits of using semiconductor devices as radiation sensors have been recognized, certain limits have been recognized in the sensing speed of clocked charge coupled devices (clocking speed limits).

Furthermore, it has been recognized that the size of the pinhole in a confocal system is selected in a trade-off between resolution and the amount of optical signal that is detected and thus generates noise. A very small pinhole will provide high performance resolution at the expense of high noise.
This is due to the fact that very small pinholes limit the sample area to a very small size, thus reducing the amount of signal detected. Large pinholes reduce noise but reduce resolution. This is due to the fact that large pinholes can detect light from a large sample area and provide a large signal, while large sample areas provide low resolution. In addition, the benefit has been recognized that adjusting the size of the sample area can provide an appropriate balance between resolution and noise.

  The invention is defined in the claims to which reference will now be made.

  In general, the present invention has a plurality of electrodes that can be configured to provide a gradient of the electric field across the plurality of electrodes, and the charge in the sensor generated by incident radiation extends across the electrodes. As the electric field gradient sweeps the output, at least one of the electrodes can be connected to a DC voltage, thus providing a barrier to charge in the imaging area so that the sensor has a variable sample area. Provides a new sensor arrangement.

  Since the electric charge is swept by an electric field gradient across a plurality of electrodes, the signal electric charge can be moved quickly as compared with the known clocking method of the CCD device. Since the sample area is variable, the sample area will be selected to balance the need for sensor sensitivity and resolution. This aspect of the invention uses a technique that provides an arrangement of electrodes that are placed on the sensor such that when an appropriate voltage is applied, charge is swept to the output only from selected sample areas. .

  One embodiment of the present invention can replace the confocal photomultiplier tube and pinhole combination described above with a single sensor. In addition to providing a much more robust arrangement, the user will also be able to adjust the size of the sensing area quickly and accurately. Since the doubling resistor is included on the same monolithic substrate as the plurality of electrodes, the charge received from the sample area can be doubled to provide the required sensitivity.

  While this sensor can be used for a variety of wavelengths of radiation, including x-rays, the primary application is sensing the visible spectrum of light. Variations to sense other wavelengths are possible without significant changes to the structure of the described embodiment, but the dimensions and types of materials used will vary. The substrate is preferably a p-type silicon substrate having a buried n-type channel. Other substrates are possible, including an n-type with a buried p-type channel and either an n-type or p-type surface channel arrangement.

  The electrodes may be placed directly on the substrate, as in the structure of an n-type substrate with a p-type electrode. However, it is desirable to arrange the electrodes across the surface of the substrate, separated by an insulating layer in a conductor / insulator / semiconductor arrangement.

  The electrodes themselves can be arranged in a variety of shapes and patterns and define a variable size sample area. In a preferred arrangement, each electrode substantially surrounds the previous electrode, and the size of the sample area is expanded and reduced by selecting an electrode that defines the outer edge of the sample area. A preferred embodiment of this arrangement is a series of concentric electrode rings. It is also possible for the electrodes to define other patterns of sample areas.

  The electrodes are connectable such that, depending on the applied voltage, the electric field across the imaging area sweeps charge across the plurality of electrodes from the imaging area to the output. Thus, the sensor outputs charge from the area defined by a single electrode to the largest area defined by all electrodes, as well as any area defined by adjacent placement between the electrodes. Is configured to sweep. The electrode defining the sample area is continuous in sensing the passage of charge directly from one electrode region to another under the influence of a potential gradient created by the voltage difference between the electrodes. Must.

  Both electrodes are connected to a certain level of DC voltage and provide a barrier to charge within the imaging area. The barrier is a potential gradient that prevents charge from reaching the output from outside the sample area and ensures that only the charge in the selected sample area is sampled. The DC voltage applied to the electrodes all creates an electric field where one of the electrodes provides a barrier, and the voltage applied to the other electrode has an appropriate voltage to sweep charge from the sample area to the output. Have been adjusted. To dispel doubts, the DC voltage is variable and not fixed.

  The size and number of electrodes can be selected to suit the intended use of the sensor. However, the maximum electrode width is limited because the potential gradient under the electrode decreases as the electrode width increases. A preferred size is a 5 micron wide electrode so that the surrounding electric field provides a suitable voltage gradient. The applied voltage can be selected for each individual electrode or group of electrodes.

It is a block diagram of a known laser scanning microscope apparatus. 1 is a schematic diagram of a sensor embodying the present invention. It is a graph which shows the electric potential distribution of a buried channel sensor. It is a graph which shows the channel potential of the sensor of FIG. FIG. 3 is a graph showing a channel potential of the sensor of FIG. 2 in a state where a bias is applied to the fifth ring. FIG. 3 is a graph showing the channel potential of the sensor of FIG. 2 with a bias applied to sweep charge to the edge of the device. It is a graph which shows the channel potential with respect to a channel width. FIG. 3 is a schematic diagram of the sensor of FIG. 2 showing a tapered gate structure. It shows how the output signal varies with the size of the detector. It shows how the output signal changes with sampling time. Fig. 5 shows an alternative embodiment in the form of a seven-sided concentric sensor. Fig. 5 shows an alternative embodiment in the form of a four-sided concentric sensor. Fig. 5 shows an alternative embodiment in the form of a non-concentric sensor. Fig. 5 shows an alternative embodiment in the form of a non-concentric sensor with the output going to one side. Fig. 3 shows a schematic diagram of the output arrangement for the sensor of Fig. 2; Fig. 4 illustrates an alternative embodiment of the invention having four output stages. A known drift detector is shown.

  Various embodiments of the present invention described above will be described with reference to the accompanying drawings.

  Embodiments of the present invention comprise a device having electrodes arranged so that charge can be collected and sensed from a variable size area by applying an appropriate bias voltage. Although the electrodes can be arranged in a variety of patterns, the preferred embodiment has a series of electrodes, each electrode substantially surrounding the first electrode in succession. The described embodiment has an annular concentric electrode arrangement. The rings are preferably uniformly spaced, but other arrangements and shapes are possible. Various different embodiments having other arrangements and shapes will be described later. Basically, embodiments provide a sensor that is variable in size or “aperture”. The sensor is capable of high speed signal capture and high sensitivity when a charge multiplication resistor is used in its structure. Control of the voltage applied to the electrode is performed by a control bias circuit.

  A preferred embodiment of the present invention is shown in FIG. Embodiments of the present invention are based on a structure similar to a CCD and include a sensor having a circular shaped optical sensing area 30 covered in concentric electrodes 32. This structure is a CCD in that the electrode arrangement is of the type having a conductive electrode covering an insulator on the surface of the semiconductor (often referred to as MIS, which is an abbreviation for metal / insulator / semiconductor). Similar to structure. The number of rings from the center that contribute to the working area of the device can be selected by the user by selecting the appropriate potential applied to the gate. Thus, the sample area is selectable. The electrodes are sometimes referred to as gate electrodes, but to eliminate doubts, the gate electrodes are not timed when reading the charge from the device, unlike the gate electrodes in a conventional area array CCD device, which is timed per line. Please note that. Therefore, although there are similarities in structure, the operation of the embodiment of the present invention is greatly different from the known CCD device.

  The arrangement of the gate electrode of this embodiment is applied to the gate insulating layer on the p-type silicon substrate as part of the embedded n-type channel CCD device. The potential received by the electrons created in the structure is determined by the potential applied to the gate. The structure initially has an empty signal charge. When the light to be sensed is poured into the sensing area 30, the required signal charge that is released flows towards the center of the device due to the potential difference across the gate. Regardless of where the electrons generated by the incident photons are initially generated, if φn is a potential applied to the nth gate from the center and the potential applied to the gate is φn> φn + 1, it will be at the center of the structure. Flows quickly toward.

  Note that the above description applies to n-channel devices where charge is collected in the center of the structure. An alternative arrangement includes a structure in which the signal is moved to the outer periphery of the device. Another alternative is a p-channel CCD device where the potential applied to the gate is φn <φn + 1 and the generated holes flow toward the center of the device.

  A dimensional potential distribution for a buried channel device is shown in FIG. As shown, if there is no influence from surrounding electrodes, the electrodes tend to gather at the maximum potential point. This maximum potential is known as the channel potential. When the two gates are arranged adjacent to each other, the signal electrons tend to flow under the gate with the highest channel potential.

FIG. 4 shows the channel potential along the diameter of the structure shown in FIG. 2 when the signal charge is emptied and a DC voltage is applied to the electrode to sweep the charge from the entire sensing area. In this case, the p-type silicon substrate has a resistivity of 20 Ohm / cm. The n-type buried channel implant has a uniform dopant concentration of 2.5 × 10 ¹⁶ / cm ⁻³ extending over a depth of 0.54 microns. The gate oxide is 130 nm thick. Each ring electrode is 4 microns wide. The diameter of the center electrode is 10 microns. A DC voltage is applied to the electrode and provides a potential difference from one electrode to the next. The voltage is not timed and provides an electric field such that electrons generated at the edge of the device are swept to the center within about 20 ns. As the difference between successive gate potentials increases, faster movement occurs. Note that the potential distribution is not a steep step from one electrode to another. This is because the edge electric field generates a potential gradient under each electrode.

  With the channel potential shown in FIG. 4, the signal generated over the entire device area will be collected in the center of the device. When one of the gates is at a low potential, only signals generated close to the center will be collected. The user can select which gate is to be low potential and what is the effective size of the working area. FIG. 5 shows the channel potential when the fifth ring is set to a low level. Any charge generated close to the edge of the device will be prevented from reaching the center by the potential barrier created under the low bias ring. This unwanted signal accumulates and eventually overflows the drain structure outside the working area and is removed.

  If this unwanted signal needs to be removed at high speed, the bias applied to the edge gates can be increased so that the electric field directs the signal directly to the drain. A specific example of the channel potential distribution in such a case is shown in FIG. As shown, the charge generated within the low voltage biased ring (within about 28 microns from the edge) is swept to the center of the sensor and outside the low voltage biased ring. The generated charge is swept to the edge of the sensor by the increasing voltage.

  When the charge reaches the center of the working area, it needs to be removed and measured. The simplest method is to have a conventional output node connected directly to the amplifier. However, it has been found that this sensor is useful for low light level applications, so it is desirable to perform charge multiplication before the output node. It has also been found that it is necessary to remove the charge from the device for amplification. Embodiments of the present invention can use a CCD register that extends to the center of the device and is installed to clock the charge emitted from the sensor. Alternatively, other arrangements can be used that create an electric field suitable for moving the signal charge from the center of the device to the edge through the channel without having to clock the gate. This arrangement consists of multiple gates biased at the appropriate DC level, or a single gate modified in structure to create a suitable electric field within the device. Such modifications include a change in gate insulator thickness along the channel, a change in buried channel implant concentration along the channel, or a change in gate electrode / channel width along the channel.

The change in channel width is a preferred embodiment and will be described below. Using a single gate structure eliminates the need for multiple gate electrodes and consequently simplifies the design. Furthermore, if the potential change along the channel is sufficiently high, a signal can be extracted at a higher speed than the clock register. The channel potential in the buried channel device is determined by the gate potential. However, the channel potential under the gate is also affected by the adjacent structure due to the edge electric field. The influence of the surrounding structure is determined by the width of the gate. FIG. 7 shows the channel potential with respect to the gate width in the buried channel device. Here, the gate is surrounded by an insulating structure, which in this case is a surface channel region, but may be a p ⁺ channel stopper, or another gate electrode biased to a low potential, for example. When a constant potential is applied to the gate where one end is narrower than the other end, the voltage increases from the narrow end to the wide end and is introduced into the narrow end. This shows that all the electrons are moving to the wide end quickly due to the influence of the electric field. Thus, a simple gate structure can be introduced into the structure of FIG. 2 to extract the signal charge.

  A preferred output gate structure is shown in FIG. The structure 34 includes a gate electrode 36 that forms an output channel with channel insulators 38 on either side of the channel. The narrow end A is located at the center of the structure and the wide end B is located at the edge of the structure. When 1 volt is applied to this gate and the gate width changes from 4 microns at the center A of the device to 8 microns at the outer periphery B, the channel potential changes from 9.2 volts to 10.2 volts. This potential change is sufficient to extract approximately 20 ns of signal electrons if the gate length is 50 microns. Note that the gate itself need not vary in width along the length. Instead of making the width of the gate constant, the width of the channel under the gate may be changed, and in this case, the change in the channel width causes a potential change.

It has been found that the resolution of the sensor can be reduced by non-ideal detector behavior. First, consider an ideal detector that scans the entire sample by detecting the emitted light. This fictitious detector samples light from an object over an infinitesimal area and measures the signal as soon as the light hits the detector. When the area of interest is scanned by such a detector, the measured signal matches the emitted light and the constructed imaging shows ideal resolution. However, two main things reduce resolution. The first is that the detector cannot be infinitesimal and light is collected over the area of the detector as it is scanned. To illustrate this result, consider the case of a one-dimensional object and a one-dimensional detector.
The object is composed of a bar pattern of period W. The signals from the detectors with widths W / 2, W / 4, and W / 8 when the entire object is scanned are shown in FIG.

  The second thing that reduces the resolution is the finite time it takes to extract the entire signal from the detector. After the initial momentary stimulus, it takes time T to extract the entire signal, and the signal extraction rate during this time is the case for a hypothetical detector that is constant after the initial stimulus. Consider. The result from this detector after scanning the bar pattern for period W is shown in FIG. The detector has a width of W / 2. Compared to the ideal case of instantaneous signal acquisition, the results for various T values are represented by curves. T is selected to be the time taken to scan the distances W / 2 and W / 4.

  As is apparent from FIG. 10, T must be much shorter than the time to scan the distance W / 4, otherwise the contrast at the highest resolution will be degraded. A useful value for the maximum value of T is the time to scan a distance of W / 10. Actually, a typical time for scanning the distance W is about 1 μs at which the maximum value of T is 100 ns. The new output channel arrangement shown in FIG. 8 can achieve such a fast travel time, but it is feasible to use a CCD clock to sweep the signal out of the imaging area at this ratio. Not a suggestion.

  As soon as the signal charge is extracted from the working area, it must be converted to a measurable potential. A typical CCD output circuit is attached to the output channel output. However, it is desirable to have a CCD gate at the output of the output channel, so that the signal is sent to the output circuit until this gate pulsates to a high level or the output channel gate pulsates to a low level. Prevent entry.

  As pointed out above, the charge is preferably multiplied in a multiplication register before it is sensed and converted by the output amplifier. This is shown in FIGS. 15 and 16 and will be described later, and may optionally include a transfer gate to move charge from the output channel to the multiplication register. The moving gate is a CCD gate at the input of the multiplication register and pulsates when signal charge is required to enter the multiplication register. This is not strictly necessary, but provides additional flexibility to the user. For example, it may be required to send a signal to the multiplication register in 1 μs. If the multiplication register can be clocked at 1 MHz, the signal will enter every 1 μs without using a moving gate. However, it is convenient to run the multiplication register at high speed to suppress the rise of the dark signal. When a moving gate is employed, the multiplication register can be run at, for example, about 10 MHz, and the moving gate can be pulsated every 1 μs. In this case, every 10th doubling element contains a signal charge.

  Several alternative imaging area arrangements are shown in FIGS. Note that the figures show a tapered output channel. Other arrangements may exist, including the use of a CCD clocking system to extract the signal from the center of the imaging area. The advantage of defining a concentric gate as shown in FIGS. 11 and 12 is that the detector area can be changed without moving the detector to ideally align with the optical axis. However, this is not an important point and other non-concentric arrangements are possible as shown in FIGS.

  In many applications, including confocal microscopy, it is necessary to detect a single photon. This is known in the art and requires the use of an electron multiplier register as shown in FIG. Various different shapes and arrangements of electrodes are possible. One such embodiment is shown in FIG. 16 and is a split arrangement with signals from each segment extracted into different output circuits. For example, the detector is divided into quadrants and has four output circuits. The signal from each quadrant is monitored and used to align the optical system.

  Control of the voltage applied to the electrode may be performed by any suitable control bias circuit. The control bias circuit takes various forms. In one simple example, the control lines coming out of each electrode are provided with a row of switches so that the operator can set the appropriate voltage when using the device. However, the control line from the electrode is preferably connected to a control bias circuit that can be switched by software control. This allows the imaging area to be selected based on signals received from the device or other inputs, either under operator control or automatically.

  Since the preferred structure of the new sensor device is a buried channel conductor / insulator / semiconductor type structure and the charge enhancement resistor is of the same type structure, they are conveniently formed on a single silicon substrate. Can be convenient.

DESCRIPTION OF SYMBOLS 10 Laser 12 Sample 16, 18 Scanning mirror 20 Imaging optical system 22 Pinhole 24 Detector 30 Sensing area 32 Concentric electrode 36 Gate electrode 38 Channel insulator

Claims (9)

In a sensor device of the type that senses incident radiation by charge generation,
A substrate in which charges are generated internally by incident radiation;
A plurality of electrodes disposed on the substrate to define an imaging area of the substrate, wherein the first electrode and each subsequent electrode are disposed to surround the previous electrode. And,
The electrodes are connectable to a dc voltage and are arranged such that an electric field is created across the imaging area and the charge across multiple electrodes of the substrate can be swept from the imaging area to the output; At least one of the electrodes is connectable to a DC voltage so as to provide a barrier to charge in the imaging area, so that the sensor has a variable sample area, and the charge is swept. The electrode facing away is located in the center of the sample area,
The output comprises an output channel, the output channel having a channel insulator on both sides of the channel and having a gate electrode connectable to a DC voltage so as to form the output channel;
The output channel is connected to a region of the substrate below the electrode in the sample area where charge is swept to;
The output channel has a tapered width, a narrow end, and a wide end, at which the charge is swept away from the sample area to a region of the substrate under the electrode. Are connected,
The tapered width causes the channel potential to rise along the channel so that the output channel sweeps charge from the central region of the sample area to the edge of the imaging area outside the sample area. It has become,
The sensor device, wherein the electrode is connectable to a DC voltage, and an electrode outside the sample area has a voltage level that sweeps charge away from the barrier .   The sensor device according to claim 1, wherein the electrodes are concentric.   3. The electrodes of claim 2, wherein each electrode is an annular ring, each subsequent electrode surrounds the previous electrode, and further surrounds a central circular electrode that is the electrode to which charge is being swept. Sensor device.   The sensor device according to claim 3, wherein the electrodes have equal widths.   The electrode of claim 1, wherein the electrode is connectable to a DC voltage, and an electrode outside the sample area has a voltage level that sweeps charge toward a drain outside the sample area. Sensor device. Coupled to said sample region by said output channel, it said further comprises the placed electrical charge multiplication register to multiply the charge from the sample zone, in any one of one to 5 above claims The sensor device described. The laser scanning microscope apparatus provided with the sensor apparatus of any one of the said Claims 1-6 . A method for operating a sensor device of a type that senses incident radiation by charge generation, the sensor device comprising: a substrate on which charge is generated by incident radiation; and an imaging area of the substrate to define the imaging area A plurality of electrodes disposed on the substrate, each subsequent electrode being disposed so as to surround the previous electrode, and the sensor device further comprises an output channel The output channel has a channel insulator on both sides of the channel and a gate electrode connectable to a DC voltage to form the output channel, the output channel sweeping charge Connected to a region of the substrate under the electrode in the sample area that is being directed, and the electrode to which the charge is swept is located in a central region of the sample area, A method for operating the capacitors device,
Connecting the electrodes to a DC voltage such that an electric field is created across the imaging area and charge across multiple electrodes is swept from the imaging area to the output;
Connecting at least one of the electrodes to a DC voltage so as to provide a barrier to charge within the imaging area, thereby allowing the sensor to have a variable sample area;
Connecting the gate electrode to a DC voltage to form the output channel, the output channel having a tapered width, a narrow end, and a wide end, the narrow At the end, charge is coupled to the region of the substrate under the electrode that is swept away from the sample area, so that the channel potential rises along the channel from the narrow end to the wide end. The output channel sweeps charge from the central region of the sample area to the edge of the imaging area outside the sample area;
Connecting the electrode to a DC voltage such that an electrode outside the sample area has a voltage level that sweeps charge away from the barrier;
A method consisting of: Connecting the electrode to a DC voltage so that an electrode outside the sample area has a voltage level that sweeps charge towards a drain outside the sample area. 9. The method according to 8 .

Images