FR2629592A1 - Adaptive pressure sensor - Google Patents

Abstract

The invention proposes the use of a TEGFET as a hydrostatic pressure sensor. By adjusting the gate voltage, it is possible to vary the conditions of operation of the sensor, so as to adapt it dynamically to the particular measurement conditions (DV and DV1 operating characteristics).

Description

 ADAPTIVE PRESSURE SENSOR.

 The present invention relates to a device sensitive to a hydrostatic pressure to which it is subjected, comprising at least a first layer of gallium-aluminum arsenide AlxGal, xAs whose relative aluminum content (x) is less than 0.25, and a second layer of gallium-aluminum arsenide A1, Gal, yAs whose relative aluminum content (y) is greater than (x) and between 0.15 and 1, and in which is formed, at the interface of these two layers, a conduction channel whose electrical resistance varies as a function of the pressure undergone by this device.

 The pressure sensitivity of structures composed of semiconductor elements from columns III and V of the Xendeleiev table is known today.

In particular, aluminum gallium arsenide structures on a gallium arsenide substrate have already been the subject of several publications, and their behavior in pressure or in temperature has been described in particular in the following articles: - "Pressure and compositional dependences of the Hall coefficient in AlyGal, yAs and their significance ", N. Lifshitz, A.Jayaraman, RALogan, Physical Review B, Vol, 21, No 2, (January 5, 1980), pp. 670 to 678;
- nThe conduction band structure and deep levels in Gal, yAlyAs alloys from a high pressure experiment,
A.Saxena, J. Phys.C: Solid State Phys. 13, (1980), pp. 4323 to 4334;
- "Comprehensive analysis of Si-doped AlyGa î-yA5 (x = O to 1); Theory and experiments", N.Chand, T.Henderson, J. Klem, WTMasselink, R.Fisher, The American Physical
Society, Physical Review B, Vol.30, No 8, (October 15, 1984), pp. 4481 to 4492; and
- "Hydrostatic Pressure Control of the Carrier Density in
GaAs / GaAlAs Heterostructures; role of the metastable deep levels ", JMMercy, C.Bousquet, JLRobert, A.Raymond et al., Surface Science 142, (1984), pp 298-305.

 A characteristic of the known structures of this family is that a determined pressure variation leads to a corresponding relative variation in the resistivity of their sensitive layer, which is fixed once and for all by the nature of the material used. At room temperature, the limit is 45% per kilobar. As a result, the resolution of such structures, that is to say the smallest value of pressure variation that they are capable of detecting, is inextricably linked to the smallest detectable relative variation in the resistance of their sensitive layer.

 Pressure sensors with strain gauges are also known, using piezoresistivity, in which the electrical resistance of each gauge varies as a function of the stress which it undergoes by application of the pressure.

 It is possible, by judiciously choosing the manufacturing parameters of these sensors, to produce structures, in particular pressure-sensitive structures, capable of operating within predetermined ranges of measurement conditions, for example in a previously chosen temperature range. and pressure.

 However, these structures are most often fragile, limited in their performance, and made complex by the fact that the gauges, which are not directly sensitive to hydrostatic pressure, must be associated with a particular conformation of the sensor, capable of converting the hydrostatic pressure to be measured in an anisotropic stress to which the gauges are only sensitive.

 These devices lead to the appearance of many problems, including long-term stability and hysteresis.

 In addition, the adaptation of such a sensor to average operating conditions being carried out once and for all at the time of its manufacture, a sensor of this type is not capable of adapting to several different ranges of average conditions Operating.

 In this context, one of the aims of the present invention is to propose a pressure sensor capable of adapting dynamically to different average measurement conditions and to present, for each of them, a resolution capable of being adjusted and to take high values.

 To this end, the invention provides a device, sensitive to hydrostatic pressure, comprising in known manner at least one first layer of gallium-aluminum arsenide A1, Gal, xAs whose relative aluminum content (x) is less than 0.25, and a second layer of gallium-aluminum arsenide AlyGa1 ~ yAs whose relative aluminum content (y) is greater than (x) and between 0.15 and 1, and in which is formed at the interface of these two layers, a conduction channel whose electrical resistance varies as a function of a pressure undergone by this device, the latter comprising a field effect transistor and two-dimensional electronic distribution (TEGFET), comprising a drain, a source and a gate .

The AlyGa1 # yAs gallium-aluminum arsenide layer is selectively doped with impurities from column IV
A and / or column VI A of the periodic table, for example at least one of the elements Si, Ge, Sn, S, Se, and
Te, at a total concentration, integrated over the thickness of this layer, which is advantageously between 1012 and 1013 atoms per square centimeter.

 The invention further provides a method for measuring hydrostatic pressure comprising a first operation of subjecting a pressure sensor to the pressure to be measured, a second operation of collecting a response signal from the sensor, and a third operation of deducing of this signal the value of this hydrostatic pressure thanks to a known law connecting this signal to this value, this method being essentially characterized in that the first and second operations are carried out using a sensor which comprises at least one field effect transistor and two-dimensional electronic distribution (TEGFET), comprising a drain, a source and a grid.

 A device for implementing the second mentioned operation may include means for keeping the gate, drain and source potentials constant, and means for measuring the current flowing between the drain and the source.

 However, this second operation can also be implemented by means of keeping the drain and source potentials constant, and by means of adjusting the gate potential so as to keep the current flowing between the drain and the source constant. .

 By these characteristics, the device and the method of the invention are capable of adapting perfectly to the case where the conditions for measuring the hydrostatic pressure change, in particular by variation of this pressure itself or by variation of the temperature, neighborhood of average measurement conditions, the method then comprising an operation consisting in adjusting, for these average measurement conditions, the gate potential to a value barely greater than the threshold voltage of the transistor, and the drain-source voltage to a value greater than the difference between the gate voltage and the threshold voltage, so that the transistor operates in saturation mode.

 As is well known to those skilled in the art, the threshold voltage of the transistor is the gate voltage for which the drain-source current is canceled. By "a value barely higher than the threshold voltage" is meant here a value exceeding by more than 1 volt the threshold voltage.

 The method of the invention can also be implemented in another form, in which the first operation consists essentially in subjecting to the hydrostatic pressure to be measured a set of two-dimensional electronic field effect transistors integrated in a circuit. , known as a ring oscillator, produced by the connection of an odd number of elementary gates, themselves composed of transistors and Schottky diodes produced on the same substrate. The third operation then essentially consists in collecting the signal produced by this oscillator, and the frequency of which is representative of the pressure to be measured.

Other characteristics and advantages of the invention will emerge clearly from the description which follows, for information and in no way limitative, with reference to the appended drawings, among which:
- Figure 1 is a sectional view of a TEGFET transistor usable in the context of the present invention;
- Figure 2 is a sectional view of an assembly using the principles of the present invention;
- Figure 3 is a diagram showing the variation of the drain-source current of a TEGFET transistor as a function of the gate voltage applied to the latter;
FIG. 4 is a diagram representing the variations in the drain-source current of the channel of a TEGFET transistor as a function of the hand of the gate voltage applied to the latter and on the other hand of the hydrostatic pressure at which it is submitted;;
- Figure 5 is an enlargement of part of Figure 4;
- Figure 6 is a diagram of a device using a first pressure measurement technique; and
- Figure 7 is a diagram of a device using a second pressure measurement technique.

TEGFET transistors, such as that shown in Figure 1, are known in the field of fast electronics. They derive their designation from an abbreviation of their definition in English, "Two-dimensional
Electron Gas Field Effect-Transistor, that is to say a field effect transistor with two-dimensional electronic distribution.

 They can be produced, for example, by the process described in the article in the article Growth of extremely uniform layers by rotating substrate holder with molecular beam epitaxy for applications to electro-optic and microwave devices ", AYCho and KYCheng, Physical application Lett.

38 (5), March 1, 1981, pp. 360 to 362.

 The structure of TEGFET 1 in Figure 1 is more precisely as follows.

 The substrate 2 is formed from a semi-insulating gallium arsenide GaAs wafer, with a thickness of the order of 300 micrometers for example.

 On this substrate is deposited by epitaxy a layer 3 of gallop arsenide GaAs undoped, with a thickness of 1 to a few micrometers, constituting a buffer layer intended to offer a good surface condition to the subsequent layers.

 On this buffer layer 3 can be deposited by epitaxy an additional layer (not shown), in gallium-aluminum arsenide A1, Gal, xAs undoped, whose relative aluminum content (x) is less than 0.25 and whose thickness is for example of the order of a few tenths of a micrometer. Overall, the buffer layer 3, whether or not covered with this additional layer, therefore constitutes a layer of gallium arsenide - AlxGal aluminum, undoped xAs, the relative aluminum content (x) of which, possibly zero, is less than 0.25.

 On the buffer layer 3 or on this optional additional layer, not shown, is deposited, also by epitaxy, a separation layer 4 of gallium-aluminum arsenide AlyGa1 # yAs undoped, the relative aluminum content (y) of which is greater than (x) and between 0.15 and 1. The thickness of this separation layer 4 is for example of the order of 0 to 30 nanometers.

 Then on the separation layer 4 is deposited, still by epitaxy, a layer 5 of doped gallium-aluminum arsenide AlyGa1 # yAs, the relative aluminum content (y) of which is substantially the same as that of the separation layer 4, and whose thickness is for example between 30 and 100 nanometers.

The layer 5 can be doped with impurities from column IV A and / or from column VI A of the periodic table, in particular by at least one of the elements Si,
Ge, Sn, S, Se, and Te, at a total concentration, integrated over the thickness of this layer, between 1012 and 1013 atoms per square centimeter-.

 As symbolically shown by dashes in Figure 1, a two-dimensional electron gas appears at the interface formed by layers 4 and 5 of gallium-aluminum arsenide AlyGa1 # yAs with layer 3, and its additional layer possible, in AlxGai # xAs gallium-aluminum arsenide.

 Layer 5 is, in known manner, covered with a contact recovery layer of gallium arsenide GaAs, then etched to present a central depression Sa, after which metal studs 6a, 6b, 6c, intended respectively to constitute the drain , the gate, and the source of the transistor, are deposited.

 FIG. 2 represents a possible mounting mode of the TEGFET transistor 1 as a hydrostatic pressure sensor.

The support of the transistor 1 is constituted by a piece of steel 7 which, with the exception of a thread 8 and a flat head 9 allowing its screwing, has a flat transverse support surface 10 and an internal peripheral recess 11, is of substantially cylindrical shape
The transducer 1 is, on the side of the step, bonded by means of a flexible adhesive to the bearing surface 10, in which three bores 10a, 10b, 10c are made, two of which are visible in FIG. 2.

 A through electrical terminal, such as 12a, 12b, 12c is maintained in each of the bores by means of a glass bead such as 13a, 13b, 13c, which ensures on the one hand the solidification of each terminal with the surface of support 10, on the other hand the electrical insulation of this terminal and this surface, and finally a sealed closure of the bores lOa to vioc.

 On the side of the recess 11, each electrical terminal is connected to a pad such as 6a, 6b, or 6c of the transistor 1, the assembly being embedded in a volume of insulating oil closed by a deformable membrane of steel 14 welded on the recess 11.

 The part 7 thus designed can be screwed onto the wall of any enclosure in which there is a pressure P to be measured. The latter, transmitted through the deformable membrane 14 to the volume of oil bathing the transistor, subjects the latter to hydrostatic pressure and temperature conditions which it is possible, in accordance with the teaching of the invention, to measure optimally as described below with reference to Figures 3 and 4.

 FIG. 3 represents the variation of the current IDS flowing between the drain and the source of a TEGFET as a function of the voltage VG applied to its gate, the voltage VDS applied between the drain and the source of the transistor being constant.

 This figure shows a first zone Zî, called blocking zone, corresponding to the values of the gate voltage VG for which no current IDS flows in the transistor.

 A second zone Z2, called the saturation regime zone, appears for values of the gate voltage greater than a limit value VT, called the threshold value, but less than the value of the gate voltage VL beyond which a third zone Z3, called the linear regime zone, characterized by an affine law of variation of the current IDS as a function of the voltage VG.

 FIG. 4 also represents the variation of the drain-source current IDS as a function of the voltage VG applied to the gate for a constant value of the polarization VDS, but for two distinct values of a hydrostatic pressure P applied to the transistor.

 This figure immediately highlights the fact that the IDS current decreases when the applied pressure P increases, a phenomenon whose optimal exploitation is the basis of the invention.

The value of the current IDES, measured by keeping the polarization of the transistor 1 subjected to the pressure, that is to say by keeping constant the gate, drain and source potentials of this transistor, is in fact usable as representative signal pressure
P.

 This measurement technique is illustrated by the vertical line DV in FIG. 4, which, at two different values of the pressure, unequivocally corresponds two different values of the current IDs.

 A device for the implementation of this technique is shown in FIG. 6. In addition to the transistor 1, this device comprises an own adjustable voltage source 15 & provide a constant but adjustable drain potential VD, an own adjustable voltage source 16 to provide a constant but adjustable LV gate potential, and an ammeter 17 indicating the value of the IDES current.

The source is also connected to a constant potential, for example that of the mass.

However, as shown in Figure 4, another measurement technique is also possible since in fact the value to which the gate potential must be adjusted
VG so as to keep the IDS current constant, keeping the drain and source potentials constant, is also representative of the hydrostatic pressure to which the transistor is subjected.

This second measurement technique is illustrated by the horizontal line DH of FIG. 4 which, at two different values of the pressure, unequivocally matches two different values of the gate voltage
VG for the same IDES current value.

 A device for implementing this second technique is shown in FIG. 7. In addition to the transistor 1, this device comprises an adjustable voltage source 18 capable of providing a constant but adjustable drain potential VD, a resistance of value R, connected between the source of the transistor and the ground, and crossed by the current IDs, an adjustable voltage source 19 capable of providing a reference potential equal to the product R.IDS, a differential amplifier 20 whose inputs are connected to the source of the transistor l and at the voltage source 19, and the output of which is connected to the gate of this transistor, and a voltmeter 21 indicating the value of the potential VG.

 In such an arrangement, the amplifier 20 adjusts the gate potential so that the voltage across the resistor R is constant, which fixes on the one hand the value of the potential of the source of transistor 1, and on the other hand share the value of the IDES current.

A first, considerable advantage, of using a
TEGFET compared to the use of a passive component such as the gridless heterostructure described in the article "Hydrostatic Pressure Control of the Carrier Density in GaAs / GaAlAs Heterostructures; role of the metastable deep levels", previously cited, and which is directly obtained by the implementation of one or the other of the two measurement techniques described with reference to FIG. 4, resides in the fact that the control of the gate potential, which it is important not to making it floating, ensures good immunity to parasites by shielding any electrostatic influence from the external environment on the conduction channel sensitive to pressure.

A second advantage resulting from the use of a TEGFET compared to the use of a different transistor, a
MOSFET for example, lies in the fact that TEGFET is directly sensitive to hydrostatic pressure, so that structures that are both complex, fragile and limited in their performance, in particular by the problems of long-term stability and hysteresis that they show, but which must nevertheless be used to convert a hydrostatic pressure to be measured into a stress to which the sensor is only sensitive, become useless.

 However, the use of a TEGFET, compared to the use of a passive sensor, leads to yet another advantage, if it is used according to the invention, according to the following procedure.

As shown in FIG. 4, the relative variation of the current IDS is maximum at the most negative values of the gate voltage, that is to say in the zone Z2 of saturation regime. For example, the relative variation of the IDS current for an absolute variation of pressure ikbar, is of the order of 14% when the transistor operates in linear mode (zone Z3), for a value of
VG close to -0.5 volts, while it can be greater than 50% when the transistor is operating in saturation mode, that is to say with a gate voltage barely greater than the threshold value VT.

 Thus, since the adjustment of the gate voltage allows a displacement of the measurement point along each of the curves of FIG. 4, it is possible to dynamically adapt the TEGFET sensor to the pressure range in which it works and even, in since each curve may possibly vary as a function of other physical parameters, for example the temperature, it is possible to adapt the TEGFET sensor to its entire measurement environment.

 In particular, in the case where the conditions for measuring the hydrostatic pressure change, in particular by variation of this pressure itself or by variation of the temperature, in the vicinity of average measurement conditions, it is advantageous to adjust, for these average measurement conditions, the gate potential at a value barely greater than the threshold voltage of the transistor, and the drain-source voltage at a value greater than the difference between the gate voltage and the threshold voltage, so that the transistor operates in saturation regime.

 This measurement technique is illustrated by the vertical DV1 in FIG. 4, in the vicinity of which the resolution of the sensor, for pressure variations around an average hydrostatic pressure of 1 kbar, is excellent, as shown even better by the enlargement made in figure 5.

 According to another measurement technique, it is possible to use a TEGFET to obtain a hydrostatic pressure measurement in the form of a frequency signal.

 To this end, the sensor comprises an odd number of inverters produced by means of TEGFET transistors and Shottky diodes connected so as to constitute a ring oscillator, delivering a signal whose frequency is representative of the pressure conditions supported by these transistors.

Although the possibility of obtaining, by means of active pressure sensors connected in a ring to form an oscillating circuit, a frequency signal representative of a pressure to be measured has already been described, for example in European patent EP 0.040.795 , the use of
TEGFET, in accordance with the teaching of the present invention, allows a direct measurement of the hydrostatic pressure, without the need to resort to the use of membranes.

Claims (7)

1. Device sensitive to hydrostatic pressure, comprising at least a first layer of gallium-aluminum arsenide AlxGal, xAs whose relative aluminum content (x) is less than 0.25, and a second layer of gallium-aluminum arsenide AIyGa1 # yAs whose relative aluminum content (y) is greater than (x) and between 0.15 and 1, and in which is formed, at the interface of these two layers, a conduction channel whose electrical resistance varies in function of a pressure undergone by this device, the latter comprising a field effect transistor and two-dimensional electronic distribution (TEGFET), comprising a drain, a source and a gate. 2. Device according to claim 1, characterized in that said layer of gallium-aluminum arsenide AlyGa1 # yAs is selectively doped with impurities from column IV A and / or from column VI A of the periodic table at a total concentration , integrated over the thickness of this layer, between 1012 and 1013 atoms per square centimeter. 3. Device according to claim 1 or 2, characterized in that it comprises means for maintaining the gate, the drain and the source at constant potentials, and means for measuring the current flowing between the drain and the source. 4. Device according to claim 1 or 2, characterized in that it comprises means for maintaining the drain and the source at constant potentials, and means for adjusting the gate potential so as to keep the current flowing between the drain and source. 5. A method for measuring a hydrostatic pressure comprising a first operation consisting in subjecting a pressure sensor to the pressure to be measured, a second operation consisting in collecting a response signal from the sensor, and a third operation consisting in deducing from this signal the value of this hydrostatic pressure thanks to a known law connecting this signal to this value, characterized in that the first and second operations are carried out using a sensor which comprises at least one field effect transistor and two-dimensional electronic distribution (TEGFET) , comprising a drain, a source and a grid. 6. Method according to claim 5, applicable in the case where the conditions for measuring said hydrostatic pressure change, in particular by variation of this pressure itself or by variation of the temperature, in the vicinity of average measurement conditions, characterized in that , for these average measurement conditions, the gate potential is adjusted to a value barely greater than the threshold voltage of the transistor, and the drain-source voltage to a value greater than the difference between the gate voltage and the threshold, so that the transistor operates in saturation mode.  7. Method according to claim 5, characterized in that the first operation essentially consists in subjecting to the hydrostatic pressure to be measured a set of two-dimensional electronic field effect transistors and electronic distribution integrated into a ring oscillator, and in that said third operation essentially consists in collecting a frequency signal representative of the pressure to be measured.