WO2004109235A2

WO2004109235A2 - Optical sensor element and sensor array

Info

Publication number: WO2004109235A2
Application number: PCT/EP2004/006247
Authority: WO
Inventors: Peter Deimel; Ulrich Prechtel
Original assignee: Daimlerchrysler Ag
Priority date: 2003-06-11
Filing date: 2004-06-09
Publication date: 2004-12-16
Also published as: WO2004109235A3

Abstract

The invention relates to an optical sensor element (10) which comprises, in a semiconductor substrate (1), a light-sensitive region (18) in which charge carriers can be released by irradiation, and two doped regions (15, 16) for receiving the charge carriers released in the light-sensitive region (18). The invention is characterized in that electrodes (13, 14) for generating a field gradient in the light-sensitive region (18) are insulated from the light-sensitive region (18) and are disposed in trenches formed in the surface of the substrate (1).

Description

Optical sensor element and sensor arrangement

The present invention relates to an optical sensor element in which a light-sensitive region in which charge carriers can be released by exposure and two doping zones for receiving charge carriers released in the light-sensitive region are formed in a semiconductor substrate, and electrodes with which are insulated from the light-sensitive region for generation a field gradient in the photosensitive area.

Conventional sensor elements of this type have the structure shown schematically in FIG. 1. 1 shows a section through a semiconductor substrate 1, in which doping zones 2, 3 are formed by diffusion or implantation of foreign atoms. A translucent oxide layer 4 covers part of the doping zones 2, 3 and an intermediate substrate area with intrinsic conductivity. Two translucent electrodes 5, 6 are applied to the oxide layer 4. The structure is similar to that of a conventional MOSFET, the gate of which is divided into two parts corresponding to the electrodes 5, 6 by a narrow window 7.

Light penetrates through the electrodes 5, 6 and the window 7 lying between them and through the oxide layer 4 into the semiconductor substrate 1 and generates pairs of charge carriers therein. The electrodes 5, 6 are transparent in order to be able to use the entire substrate surface between the doping zones 2, 3 for light absorption. The electrodes 5, 6 are alternately connected to a potential which creates a potential gradient in the region of the semiconductor substrate 1 lying between the electrodes 5, 6, which, depending on the polarity, “shovels” the charge carriers to one of the two doping zones 2, 3 _. Charge carriers of the appropriate type that reach one of the doping zones 2 or 3 thus result in a photocurrent.

The use of such sensor elements lies in particular in their suitability for carrying out an optical distance measuring method. For this purpose, a light source, such as a laser diode, is modulated on-off with the same signal that is also applied to one of the electrodes 5, 6 in order to generate a potential gradient with an alternating direction between them. The laser diode radiates the light onto an object , the distance of which is to be measured, and light reflected by the object strikes the electrodes 5, 6 and / or the window 7 and generates charge carrier pairs in the semiconductor material underneath. If the distance of the object is zero, there is no phase difference between the light striking the window 7 and the signal, for example, applied to the electrode 5; Whenever light hits the window 7, there is a potential gradient at the electrodes 5, which leads the charge carriers generated in the substrate to the doping zone 2. In the time intervals in which the direction of the potential gradient is reversed and the charge carriers are guided to the doping zone 3, no light falls on the window 7, so that a maximum photocurrent is detected at the doping zone 2 and no photocurrent at the doping zone 3. With increasing distance of the object to be detected, the transit time-related phase shift between the signals applied to the electrodes and the light incident on the window 7 increases, and the distance of the object can be deduced from the ratio of the photocurrents tapped at the doping zones 2, 3 become. A problem with the known structure in FIG. 1 is that the light penetrates into a silicon semiconductor substrate 1 a few micrometers deep (approx. 20 μm at a wavelength of 850 nm), but that the space charge zone and thus the field gradient serve is generated in the substrate 1 by the potentials applied in phase opposition to the electrodes 5, 6, and is required to allow the charge carriers to migrate to one of the doping zones 2, 3 and has a considerably lower penetration depth. That is, only charge carriers with good effectiveness are collected and directed into the doping zones which are generated in the space charge zone close to the surface of the semiconductor substrate 1 and at a short distance from the electrodes 5, 6; However, a large part of the charge carriers generated arise in deeper areas of the substrate 1 outside the space charge zone, where there is no potential gradient. With these charge carriers, there is a high probability that they will not reach a doping zone or will only reach a doping zone by thermal diffusion after the potential gradient has reversed its direction. The distance information contained in these charge carriers is thus lost.

In addition, it can be assumed that only a small part of the surface of the substrate can be used effectively for the detection of light. The arrangement of the electrodes 5, 6 on the surface of the substrate leads to a strengthening of the electric field at the mutually facing edges of the electrodes. The electrodes themselves shield large parts of the substrate 1 against the electrical field of the potential gradient, so that charge carriers from there also slowly reach one of the doping zones 2, 3 by thermal diffusion.

The object of the present invention is to provide a sensor element of the type defined at the outset, which has a high sensitivity. The object is achieved by a sensor element having the features of claim 1. According to the invention, since the insulated electrodes are arranged in trenches formed in the surface of the substrate, they are able to generate an electric field which drives a charge carrier drift between adjacent trenches, which penetrates to a considerable depth into the substrate and also detects charge carriers generated in regions of the substrate remote from the surface and quickly dissipates them to one of the doping zones. The arrangement of the electrodes prevents local elevation of the potential gradient; shielding by channel formation can be avoided. In addition, due to the arrangement of the electrodes, a high percentage of the substrate surface can be used for signal generation. Ideally, the electric field extends from one trench to the other, ie the potential gradient between the trenches is sufficient to pull almost all of the charge carriers generated out of the space charge zone.

Each doping zone should expediently touch an insulation layer of one of the insulated electrodes, so that if a conductive channel forms on the insulation layer as a result of a pull-off potential applied to the insulated electrode, this channel has contact with the doping zone and charge carriers of the doping zone collected in the channel can be supplied without losses. Since, in contrast to the conventional structure, the channels in the structure according to the invention are practically perpendicular to the desired drift direction, they do not shield the areas of the semiconductor substrate lying between two electrodes to any appreciable extent against the electrical field. The entire semiconductor mass between the two electrodes thus contributes to the sensitivity of the sensor element.

If the depth of the trenches is greater than the thickness of the doping zones, charge carriers can also be supplied to the doping zones via the channels which form on the electrodes and which lie in deep zones of the semiconductor substrate below of the doping zones are generated. Since the thickness of the doping zones is generally much smaller than the penetration depth of the light, even the semiconductor material below the doping zones can contribute to the sensitivity of the sensor element.

The preferred depth of the trenches is between 5 and 40 μm, preferably between 12 and 25 μm. In general, the deeper the trenches, the greater the depth of penetration of the light to be detected into the semiconductor substrate 1.

In order to achieve a good utilization of the substrate area, two sensor elements adjacent in a first direction are expediently arranged on both sides of a common insulated electrode. Doping zones of the two sensor elements adjacent to the common insulated electrode can be connected in an electrically conductive manner. Two sensor elements with conductively connected doping zones are expediently combined to form a pixel, wherein a pixel can have more than two sensor elements.

In order to create a spatially resolving sensor arrangement, there should be at least individual pairs of sensor elements in which doping zones of the two sensor elements adjacent to the common insulated electrode are electrically isolated from one another, so that the photocurrents collected in the two doping zones can be processed separately from one another.

Such insulation of doping zones lying opposite one another on both sides of an insulated electrode can be achieved, for example, by the insulated electrode lying between them having a thicker insulating layer at the bottom of its trench than at its side walls. This measure prevents the creation of a conductive channel Prevented floor of the trench, which could otherwise be a conductive connection between the doping zones.

According to another embodiment, two adjacent sensor elements belonging to different pixels do not have a common insulated electrode, but a zone which isolates the electrodes from one another is formed between two such electrodes of the adjacent sensor elements. Such an insulating zone can be, for example, the semiconductor substrate itself if, for example, the two electrodes are each housed in their own trenches.

The charges discharged from the doping zones are stored on two capacitors. The distance of an object imaged on the pixels can be determined from the difference in the lengths of these two capacitors. To save substrate surface, these capacitors are ^'as the insulated electrodes are preferably housed in trenches so that its plates are perpendicular to the substrate surface orientation aids advantage.

Further features and advantages of the invention result from the following description of exemplary embodiments with reference to the attached figures.

Show:

1 already treated, a section through a semiconductor substrate with a conventional sensor element;

2, partly in section, partly in a perspective top view looking at the surface, a sensor element according to the invention;

Fig. 3 is a plan view of a pixel of a spatial resolution

Sensor arrangement formed from several of the sensor elements shown in FIG. 2; 4 shows a plan view of a plurality of pixels of a second spatially resolving sensor arrangement;

5 shows a schematic section through a sensor element according to a second embodiment of the invention; and FIG. 6 shows a further example of a sensor arrangement;

7, partly in section, partly in a perspective top view of the surface, an inventive sensor element with a metal semiconductor structure and Schottky barrier; Fig. 8 several sensor elements with diffused p ⁺ contact.

2 shows a single sensor element 10 according to the invention. It comprises two parallel trenches 11 anisotropically etched in a silicon substrate 1, which after the etching have been oxidized on the surface to form an insulating oxide layer 12, and which are subsequently coated with electrically conductive material such as metal or highly doped polysilicon have been filled in to form electrodes 13, 14 insulated against the substrate 1. The electrodes 13, 14 face each other like parallel plates of a capacitor. The depth of the trenches 11 is typically approximately 25 μm, their length is largely arbitrary and, depending on the size of a pixel formed by one or more sensor elements 10, can be, for example, in a range from 20 to 200 μm.

Two doping zones 15, 16 are formed between the two electrodes 13, 14 and in each case in contact with the oxide layer 12 of one of them. The thickness of the doping zones 15, 16 is a few hundred nanometers and is therefore significantly less than the depth of penetration of the light into the semiconductor substrate 1, so that not only light that falls on an undoped surface area 17 between the zones 15, 16, but also light that penetrates the doping zones 15, 16 can release charge carriers in the sensitive area 18 of the substrate lying between the trenches 11 , These charge carriers are drawn off toward the electrode 13 or 14, each of which has a pull-off potential. If the applied pull-off potential is high enough to attract charge carriers to the region of the substrate 1 adjoining the oxide layer 12 of the electrode 13 or 14, a channel 19 is formed in this region in which the charge carriers can move freely. The charge carriers flow via this channel 19 to the adjacent doping zone 15 or 16.

The charge carriers are derived from the doping zones 15, 16 via an ohmic contact attached thereto, e.g. to (not shown) collecting capacitors, the plates of which, like the electrodes 13, 14, are each formed by electrically conductive material which is introduced into a trench etched into the semiconductor substrate 1 and is electrically insulated from the substrate 1.

FIG. 3 shows a plan view of a pixel of a sensor arrangement, which is composed of four sensor elements 10, as shown in FIG. 2. An individual sensor element 10 in FIG. 3 corresponds to the area identified by a dashed rectangle. There are two insulated electrodes 13, designated 13 ′ in FIG. 3, which each belong to two adjacent sensor elements 10 and on whose two long sides doping zones 15, 16 extend. The two doping zones 15, 16 on each of the electrodes 13 'are extended over a longitudinal end of the electrode 13' and thus fused together in an electrically conductive manner. Only the outer electrodes, designated 13 ", have a doping zone 15, 16 only on one long side. The electrodes 13 ', 13 "are each connected alternately to two supply lines 20, 21, via which they each receive the pull-off potential out of phase. Accordingly, the doping zones 15, 16 are alternately connected to two signal lines 22, 23 via which drain the charge carriers to collecting capacitors and / or other evaluation circuits.

In the sensor arrangement shown in FIG. 4, each individual doping zone 15 or 16 surrounding an insulated electrode 13 or 14 is provided with its own signal line 24. This means that when the electrodes 13 are connected to the pull-off potential, the doping zones 15 surrounding them each collect charge carriers from the two sensor elements which are combined in the figure under the reference numeral 24, while when the electrodes 14 receive the pull-off potential, they each Collect load carriers from the pairs labeled 25. Two sensor elements 10 each form a pixel, the position of the pixels fluctuating periodically by half a pixel width or the assignment of the sensor elements 10 to a pixel varying depending on which electrodes have the pull-off potential. With such a sensor arrangement, very high-resolution images can be generated, especially in a field mode; In order to use these images for a spatially resolving distance measurement, however, a greater processing outlay is required than with stationary pixels according to the embodiment in FIG. 3.

Small stationary pixels can be created with the design of FIG.

5 can be obtained. The sensor element 10 ′ shown in this figure differs from the sensor element 10 in FIG. 2 in that the oxide layer 12 of the insulated electrodes 13, 14 is in each case made significantly wider at the bottom 26 of the trench in which the electrodes are arranged than on its side flanks 27. As a result, the electric field strength in the semiconductor material adjacent to the oxide layer 12 is rial in each case lower on the bottom 26 than on the side flanks 27. This makes it possible to apply a pull-off potential to one of the electrodes 13, 14 that is strong enough to produce two channels 19 on both sides of the electrode, but not one the bottom 26 bridging channel that would short these two channels 19. Since in this embodiment the doping zones 15, 16 on both sides of an isolated electrode 13, 14 are also not connected to one another on the substrate surface, adjacent sensor elements 10 'do not influence one another, so that each sensor element 10' is a pixel that is independent of the other represents.

Another possibility of decoupling neighboring sensor elements in order to use them each as a pixel is shown in FIG. 6. The individual sensor elements 10 here are identical to those from FIG. 2, but unlike in FIG. 3, each insulated electrode 13, 14 belongs to exactly one sensor element 10, and there is one between adjacent electrodes 13, 14 of different sensor elements 10 insulating layer 28, here in the form of material of the semiconductor substrate 1.

In order to reduce the capacitance of the entire sensor arrangement, the insulating layer can also be a further trench, which electrically separates the trenches of adjacent electrodes 13, 14 from one another. Such a trench can surround the entire pixel and thus contribute to the optical and electrical separation of the individual pixels from one another.

A single sensor element according to the invention is shown in FIG. 7 in accordance with FIG. The sensor element shown in FIG. 7 differs from the sensor element shown in FIG. 2 in that a simple metal semiconductor structure 31 is introduced in the etched trench 11 instead of a metal oxide semiconductor (MOS). Instead of an O In this case, the trenches 11 etched into the photosensitive region 18 are filled with a metal layer, analogously to the highly doped polysilicon 14. The oxide layer 12 shown in FIG. 2 is not present here. The trenches 11 filled with the metal are metal-semiconductor contacts which form Schottky barriers 30 with respect to the silicon in the photosensitive region 18. The charge carriers generated in silicon become phase-correct, ie the two Schottky diodes are driven by a phase shift of 180 °, pulled out of the silicon via the two vertical Schottky diodes and passed on to the collecting capacitors. In a particularly advantageous manner, the sensor element has no doping zones (15, 16). As a result, the sensor element can be manufactured easily, with fewer work steps being required.

Fig. 8 shows a plurality of sensor elements according to the invention with the surface of the light-sensitive areas 18-diffused p ⁺ contact closure 32. By means of the wide ohmic p ⁺ - contacts 32 can direct the light, such as sunlight in the environment, are eliminated. In addition to the signal photons, photons from the ambient light of the surroundings are also absorbed on the surface of the light-sensitive areas 18. The short-wave portion of the constant light photons has only a small penetration depth in silicon and is therefore absorbed on the surface of the photosensitive region 18. Signal photons of the near infrared wavelength range, however, penetrate deeper into the silicon. By applying a reverse voltage to the diffused p ⁺ contacts connected to the conductor track 33, those charge carriers are removed which are generated by the short-wave direct light radiation near the surface of the photosensitive region 18 and do not contribute to the signal.

Claims

claims

1. Optical sensor element (10), in which in a semiconductor substrate (1) a light-sensitive area (18) in which charge carriers can be released by exposure, and two doping zones (15, 16) for recording in the light-sensitive area (18) Charge carriers are formed, and with electrodes (13, 14) insulated against the photosensitive region (18) for generating a field gradient in the photosensitive region (18), characterized in that the insulated electrodes (13, 14) in the surface of the substrate ( 1) formed trenches are attached.

2. Optical sensor element according to claim 1, so that each doping zone (15, 16) touches an insulation layer (12) of one of the insulated electrodes (13, 14).

3. An optical sensor element according to claim 1 or 2, so that an ohmic contact is formed at each doping zone (15, 16).

4. Optical sensor element according to one of the preceding claims, characterized in that that the depth of the trenches is greater than the thickness of the doping zones (15, 16).

5. Optical sensor element according to one of the preceding claims, that the depths of the trenches are between 5 and 40 μm, preferably between 12 and 25 μm deep.

6. Optical sensor element according to one of the preceding claims, so that each doping zone (15, 16) is assigned a collecting capacitor for collecting charge carriers drawn from the doping zone (15, 16).

7. The optical sensor element as claimed in claim 6, so that each collecting capacitor comprises two conductive plates which are arranged in trenches of the substrate.

8. Optical sensor element according to one of the preceding claims, d a d u r c h g e k e n n z e i c h n e t that instead of isolated electrodes (13, 14) made of metal semiconductor structures (31) electrodes are present, which form Schottky barriers (30) with respect to the light-sensitive region (18).

9. The optical sensor element as claimed in claim 8, so that the sensor element has no doping zones (15, 16).

10. Optical sensor element according to one of the preceding claims, characterized in that an ohmic p-contact (32) has diffused into the surface of the photosensitive region (18).

11. Optical sensor arrangement with a plurality of sensors according to one of the preceding claims, so that two sensor elements (10) adjacent in a first direction are arranged on both sides of a common insulated electrode (13 ').

12. Optical sensor arrangement according to claim 11, so that doping zones (15, 16) of the two sensor elements (10) adjoining the common insulated electrode (13 ') are electrically conductively connected.

13. An optical sensor arrangement according to claim 12, so that the two sensor elements (10) are combined to form a pixel.

14. An optical sensor arrangement according to claim 11, so that doping zones (15, 16) of the two sensor elements (10, 10 ') adjoining the common insulated electrode (13') are electrically insulated from one another.

15. Optical sensor arrangement according to claim 14, so that an insulating layer (12) of one of the insulated electrodes (13, 14) is thicker on the bottom (26) of its trench than on its side walls (27).

16. Optical sensor arrangement with a plurality of sensors according to one of claims 1 to 7, characterized in that between adjacent insulated electrodes (13, 14) is formed by two adjacent sensor elements (10) in a first direction, a zone (28) isolating the electrodes (13, 14) from each other.

17. Optical sensor arrangement according to claim 16, so that the insulating zone (28) is formed by the semiconductor substrate (1) or a trench.