Method for manufacturing a sensor chip for the direct conversion of X-rays, a sensor for the direct conversion of X-rays and the dental radiology apparatus for using such a sensor

Abstract

This invention relates to a method to manufacture a chip to detect the direct conversion of X-rays. It also relates to a direct conversion detector for X-rays using such a chip and dental radiology equipment using at least one such detector. The method to manufacture the wafer comprises a step for applying pressure (3, 4, 4 a) to a powdered polycrystalline semiconductor material and a step for heating (5-9) during a set time period. It comprises a preliminary step for providing an impurity level of at least 0.2% in the polycrystalline semiconductor material.

Claims

1. Method of manufacturing a direct conversion X-ray detecting semiconductor chip for a direct conversion X-ray detector, the method comprises a step for applying pressure (3, 4, 4a) to a powdered polycrystalline semiconductor material and a step for heating (5-9) the powdered polycrystalline semiconductor material during a set time period, where the purity of the powdered polycrystalline semiconductor material is equal to or greater than 98% and less than 99.8%.

2. Method according to claim 1, characterized in that the step for applying pressure consists of applying (3) an axial compression force (4, 4a) to the powdered polycrystalline semiconductor material of a value which will ensure an axial orientation (C) of the polycrystalline semiconductor material grains according to the direction of the application (C) of said axial compression force.

3. Method according to claim 2, characterized in that the value of said axial compression force is comprised between 100 and 1,000 MPa, and in that the duration of the pressurizing is equal to at least an hour.

4. Method according to claim 3, characterized in that the temperature of the heat is comprised between 70 C. and 200 C., and in that the duration of the heating is equal to at least one hour.

5. Method according to claim 1, characterized in that the step of applying pressure is implemented upon starting the heating step.

6. Method according to claim 5, characterized in that the step of applying pressure is implemented throughout the entire heating step.

7. Method according to claim 1, characterized in that the powdered semiconductor material comprises at least one of the constituents chosen from among: PbI.sub.2, HgI.sub.2, PbO.

8. Method according to claim 1, characterized in that the step for applying pressure is preceded by a step in which a dopant is incorporated into the polycrystalline semiconductor material, the dopant being preferably selected for HgI.sub.2 or PbI.sub.2 from halogenated compounds, and especially selected from among: CsI, CdI.sub.2, SnCl.sub.2, AgI or BiI.sub.3, and for PbO selected from oxide compounds.

9. Direct conversion X-ray detector, characterized in that the direct conversion X-ray detector comprises a semi-conductor wafer (13) manufactured in accordance with claim 1.

10. Direct conversion X-ray detector according to claim 9, characterized in that the direct conversion X-ray detector is associated with an integrated semiconductor circuit, and in that the integrated semiconductor circuit comprises a first continuous electrode (14) in contact with an entry surface of the wafer (13), and a second electrode (15) constituted by a plurality of conductive patches in contact with the opposite surface of the wafer (13), so as to provide a one-dimensional or two dimensional array of pixels, said first and second electrodes being electrically (16) connected with the integrated semiconductor circuit (17) associated with a surface upon which the semiconductor wafer (13) has been deposited, the integrated semiconductor circuit (17) being arranged (18, 19) so that the integrated semiconductor circuit produces a plurality of electrical signals (20) representative of the intensity of the X-rays received in the different pixels of said semiconductor chip.

11. Radiological apparatus, in particular, for dental radiology, using at least one direct conversion X-ray detector according to claim 10, characterized in that the dental radiological apparatus further comprises at least one controlled X-ray source (21,22) and a control circuit (29, 30) to execute at least one X-ray exposure in the direction of said at least one direct conversion X-ray detector (28) and for deducing (33) therefrom, by viewing (34), printing (35) and/or recording (36) at least one graphical representation based on the plurality of electrical signals generated by said at least one direct conversion X-ray detector (28).

12. Radiological apparatus in accordance with claim 11, characterized in that said apparatus is an intraoral dental X-ray apparatus or an extraoral dental X-ray apparatus.

13. Method according to claim 1, further comprising: producing a semiconductor wafer made of only the powdered polycrystalline semiconductor material, and depositing an electrode on a surface of the semiconductor wafer, wherein the producing a semiconductor wafer comprises applying pressure between 100 and 1,000 MPa to only the powdered polycrystalline semiconductor material while heating with a temperature between 70 C. and 200 C. during a set time period.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) Other characteristics and advantages of this invention will be better understood using the description and the accompanying drawings in which:

(2) FIG. 1 shows a schematic representation of a device with which certain steps of the method in accordance with the invention may be implemented;

(3) FIGS. 2a and 2b are diagrams explaining the effect resulting from the axial compression;

(4) FIG. 3 is a schematic representation of a direct conversion X-ray detector using a semiconductor wafer manufactured in accordance with the invention; and

(5) FIG. 4 is a diagram of a particular embodiment of a dental X-ray apparatus using a direct conversion X-ray detector in accordance with the invention.

(6) According to the invention, the manufacturing method uses a powdered semiconductor material.

(7) According to the scientific literature, PbI.sub.2, HgI.sub.2 and/or PbO based materials are potentially the most effective. In one preferred embodiment, the material chosen is HgI.sub.2. Furthermore, when the material possesses an anisotropic electric mobility which is the case with the aforementioned materials, it is preferable that it be shaped so that the electric current flows in the direction of greater mobility.

(8) In the prior art, as was described above, the powders used to create a detection chip by sintering must be subjected to a series of purification steps to use the most pure semiconductor material possible.

(9) On the contrary and surprisingly, the inventors have observed that the presence of a minimum level of impurities makes it possible to obtain a much higher yield of electrical conversion of X-photons.

(10) The inventors find that, in the case of an HgI.sub.2 wafer manufactured from a less pure powder, better results are obtained in terms of detection sensitivity.

(11) The desirable and admissible impurities in accordance with the invention are in particular, metal waste products contained in the initial powder or those that come from tools used to reduce the powder and to measure or calibrate the average diameter of the particles of the powder and from the manufacturing process of the industrial powder producer.

(12) Preferably, the powdered semiconductor material to be used to create the sintering of the chip should be 99.0% pure, with a rate of impurities, especially metals, lower than 1.0%.

(13) Such a non-purified powder is advantageously commercially available, thus allowing one to remove all the costs involved in a series of purification steps. Specifically, manufacturers of powders commercially offer gradations of purity in their products by groups of ten. Tests conducted by the inventors showed improved results in the sintered powders for direct X-photon conversion from the group of ten corresponding to 0.2% impurities. This represents a gain of more than two thousand times the impurity required in the prior art including the aforementioned document U.S. Pat. No. 5,892,227.

(14) The inventors then realized that the application of an axial compressive load to such a material having a predetermined minimum rate of impurities which also favors an improvement in the yield of the electrical conversion by reorienting the grains along a crystallographic axis C common to the grains making up the initial powder. This characteristic is combined with the minimal impurity rate characteristic in a preferred embodiment.

(15) In FIG. 1, a device which makes it possible to implement the manufacturing method is shown.

(16) A furnace 1 is made up by a heat conducting body 2, which has a central opening 6 limited by a lower wall 4a and by an axial compression press 4 placed above the powdered semiconductor material to be sintered.

(17) A mold whose cavity has the shape to be taken by the semiconductor wafer material is placed in the central opening of the housing 6 or furnace 1.

(18) For example, this mold makes it possible to produce a wafer with at least two opposing and parallel surfaces and with a circular or rectangular shape.

(19) The heat conducting body 2 is pierced by channels 5 for circulation of a heated oil from a reserve 7. The reserve oil is heated by an electric resistor 8 connected by a circuit controlling the duration and/or the temperature for sintering (not shown).

(20) The heated oil is aspirated by an electric pump 9 controlled by the control circuit for the duration and/or the temperature for sintering (not shown). The hot oil enters into the channels 5 of the body 2. These channels are connected in a spiral shape and the flow of the heating oil is then routed towards the reserve 7.

(21) Preferably, the upper press 4 is activated by a mechanism 3 which regulates the value, and the duration of the application, of the force of the axial compression applied to the powder, then to the semiconductor wafer housed in the chamber 6 of the furnace 1.

(22) This implementation of the axial compression is preferably performed upon startup of the heating, i.e., of the heating for sintering the powder and then the wafer being manufactured.

(23) In another embodiment, interesting results were also obtained by performing cold axial compression, which is carried out before activating the heating for the manufacture by sintering of the wafer made of semiconductor material. In particular, good results were obtained by performing shell-firing after prior cold axial compression for up to an hour.

(24) FIG. 2a is a schematic illustration of the grains in a wafer made of semiconductor material. These grainsshown, for example, by the grains 10, 11have a parallelepiped form. This form does not predetermine the actual shape of the grains in the pellet, but its purpose is to highlight the anisotropy of their electrical properties, in the case of an anisotropic material. The direction C of greater electrical mobility corresponds to the smallest dimension of the parallelepiped.

(25) The grains shown in FIG. 2a are oriented in any direction. As a result, if electrical charges flow through the material, they will cross the grains according to their direction of greater electric mobility, and will then travel to other grains in a lower electric mobility direction. This will result in the overall electrical mobility of the material not being maximized.

(26) In FIG. 2b, for implementation of axial compression in a given direction C, the grain surfaces 10, 11 of the particle in the powdered semiconductor material tends to virtually align such that direction C of the higher electrical mobility is parallel to the axis of the axial compression.

(27) The overall electric mobility of the ceramic is increased along this direction. This results in an improved flow of electrical charges along this direction, and hence an improved yield in the electrical conversion of X-photons.

(28) The compression, or application of pressure, or sintering under pressure of the powder to be sintered has already been practiced in the prior art. But in particular, it is noted that the desired effect was then a simple compaction of the powder to impart better mechanical strength before the pellet thus pre-compacted is placed in the sintering furnace. In particular, the compaction only results in a reduction in the powder volume, the voids between the powder particles being reduced by applying this pressure.

(29) According to the invention, applying the axial compression at the same time as heating exceeds that single state of volume reduction to promote the growth of grains having the orientation of the axis C parallel to the axis of compression at the expense of the other grains and results in a semiconductor chip having a preferred orientation of the grains.

(30) The compression performed in the manufacturing method according to the invention should preferably take place at the beginning of the heating step. It occurs with a compression load or force at a set value, during a set time period, and at a set temperature value.

(31) After the finishing the axial compression step, the heating step for sintering can be stopped immediately or may be continued for a set time period or at a maintained temperature or at another predetermined temperature, in particular according to the gradation of the powdered semiconductor material and/or the thickness of the wafer that one wishes to produce.

(32) Good results were also obtained by first practicing an axial cold compression step, and by performing it just after the sintering heating step itself. In such an embodiment, it was found advantageous to use a technique of shell-firing; the axially compressed pellet is placed on a layer of non-compacted powder and finally covered with a layer of non-compacted powder.

(33) The heating step makes it possible for the grains to grow within the wafer. The presence of the powder bed during sintering can limit contact between the wafer and the atmosphere during sintering. The phenomena of transport of matter on the surface are limited by the gradual evaporation and recondensation of the material on the surface.

(34) Shell-firing makes it possible to obtain a better surface, i.e., better flatness by reducing roughness from the surfaces of the opposing sides of the wafer. Shell-firing also makes it possible to reduce the formation of a surface parasite phase which may occur if the wafer had been directly in contact with the furnace atmosphere during annealing.

(35) The surfaces of the press 4 and the bottom of the mold 4a have surface states which make it possible to comply with a smooth surface state free of open surface porosity that was possible to obtain with axial compression and heating during the sintering itself.

(36) It is thus perfectly possible to directly make an electrode deposit on the two active surfaces opposing the wafer made of semiconductor material and to deposit the assembly directly, of the electrodes and of the wafer, on an integrated CMOS circuit, which will be described later.

(37) In one particular embodiment, the axial compression step is preceded by a step during which a dopant is incorporated in the semiconductor material. In one preferred embodiment, the dopant is chosen from among iodide compounds and is for example, CsI or BiI.sub.3.

(38) In FIG. 3, a direct conversion X-ray detector is illustrated which is produced by using a ceramic or semiconductor material wafer 13 obtained with the manufacturing method of the invention. The wafer made of a semi-conductor material 13 has two parallel and opposing surfaces, one lower and one upper when viewing FIG. 3. They have a surface state which is perfectly flat and free of roughness.

(39) It is then possible to deposit one or more metal layers, such as palladium, by evaporation, or by means of any suitable technique, which will constitute respectively a first electrode 14 on the upper surface and a second electrode 15 on the lower surface of the wafer made of semiconductor material.

(40) In one preferred embodiment, the first electrode 14 is a continuous, flat and two dimensional, which may be crossed by the incident X-photon flow. Depending on the energy of the incident X-photons, and the wavelength, the axial thickness separating the upper surface of the lower surface of the wafer 13 is determined by maximizing the absorption of the X-photons through the second electrode 15.

(41) This second electrode 15 is not produced in a two dimensional continuous fashion, but rather in the form of patches defining the right to a pixel image sensor for each patch.

(42) Depending on the case, the second electrode 15 is then configured in such a way that the conductor patches are arranged: in one linear direction or another, but one dimensional, to create a one dimensional image sensor; or in two dimensions, under various arrangements chosen during the design of the detector, to create a two-dimensional sensor for a 2D image.

(43) Each conductive patch on the second electrode is then connected by a network of conductive lines to a set of amplifiers for signals detected for each pixel. As is known, the 2D pixel arrangement thus formed can be operated directly or by multiplexing, including a 2D addressing mechanism in rows and columns. To this end, the direct conversion detector of the invention also comprises a integrated CMOS circuit 17, upon which the upper surface of the wafer made of semiconductor material 13, with its electrodes 14 and 15, is deposited and fixed.

(44) The integrated CMOS circuit 17 is mainly comprised of a detection circuit 18 and a signal processing circuit 19. The circuit 18 is connected to each electrical charge pixel sensor. Each pixel sensor comprises a preamplifier and a formatting circuit responsive to the charge produced on each pixel of the wafer 13. The circuit 19 is provided with means for producing electrical signals representative of the X-ray intensity received in the various pixels of the semiconductor chip.

(45) For this purpose, the electrodes 14 and 15 of the semiconductor chip 13 are electrically connected to the conductive input patches (not shown) of the integrated CMOS circuit 17 by electrical connections. The conductive input patches of the integrated CMOS circuit 17 are electrical connected to the properly polarized sensor circuits 18 as is known.

(46) Not all the necessary circuits to be developed for the integrated CMOS circuits 17 are described herein, but only those necessary so that the invention may be understood. The electrical signals representative of the of the received X-ray intensity produced by the circuits 18 are available on the output terminals 20 of the integrated complex circuit 13-20 thus formed which constitute a direct conversion X-ray detector.

(47) In FIG. 4, a dental X-ray apparatus is schematically shown that advantageously uses at least one direct conversion X-ray detector as described in reference to FIG. 3.

(48) The sensor equipped with the wafer of the invention can be used in an extraoral dental imaging system. One source 21 of X-rays is mounted on a mobile arm 23 on a support 24 upon which it can be moved. With regard to the X-ray source 21, the arm 23 carries an apparatus 25 designed to produce at least one image signal produced by using at least one sensor with a direct conversion X-ray detector 28 constituted in accordance with that described above.

(49) The sensor equipped with the wafer of the invention can be used in an intraoral dental imaging system. The sensor is then positioned inside the patient's mouth behind the tooth to be X-rayed. The X-ray source is positioned on the adjustable arm. It is positioned against the patient's cheek just before exposure.

(50) The direct conversion X-ray detector 28 is opposite the X-ray source 21 equipped with an assembly 22 providing filtering and collimating of the incident X-rays. X-rays 27 then pass through an analysis region 26 such as the head of a patient, or a part of his jaw, to be examined by involving the direct conversion 28 X-ray detector 27.

(51) The dental radiology apparatus thus constituted also comprises a calculator 29 for performing such a control circuit 29, 31 to perform at least one X-ray exposure. The calculator 29 mainly comprises: means for generating control signals 30 from the X-ray source 21 and the filter and collimation unit 22; means for generating control signals 31 for activating the arm 23, as desired, so as to perform determined scans of the region to be analyzed 26 in synchronism with the X-ray exposures determined by the X-ray source 21; means for receiving and processing various electric signals 32 representative of the intensity of X-rays received in the various pixels of the semiconductor sensor chip direct conversion of X-rays 28; means to derive by viewing, printing and/or recording at least one graphic representation on the basis of the plurality of electrical signals generated by said at least one direct conversion X-ray detector.

(52) These last means of calculating 29 and are connected to a graphic display device 34, a graphics printer 35 and/or a storage system 36 and direct consultation or by means of a communication network.

(53) We noted that this type of radiology apparatus operates with a recurrence of high frequency that require the performance that the wafer made of semiconductor material obtained by the manufacturing method of the invention which makes it possible to attain what we shall show in a later section.

EXAMPLES

Example 1

(54) For the polycrystalline semiconductor materials, the following parameters are used:

(55) Sintering temperature: 100 C.

(56) Axial compression load: 300 Mega (3.10.sup.8) Pascals.

(57) Duration of pressurizing at this temperature: 20 hours.

Example 2

(58) For the polycrystalline semiconductor materials, the following parameters are used:

(59) Sintering temperature: Comprised between 70 C. and 130 C.

(60) Axial compression load: Comprised between 100 Mega Pascals and 800 Mega Pascals.

(61) Duration of pressurizing at this temperature: At least 1 hour.

Example 3

(62) For mercury iodide HgI.sub.2, the following parameters are used: Sintering temperature: Lower than 200 C.

(63) Axial compression load: Lower than 1000 Mega Pascals.

(64) Duration of pressurizing at this temperature: At least 1 hour.

Example 4

(65) For mercury iodide HgI.sub.2, the following parameters are used:

(66) Sintering temperature: Comprised between 70 C. and 200 C.

(67) Axial compression load: Comprised between 100 and 1000 Mega Pascals.

(68) Duration of pressurizing at this temperature: At least 1 hour.

Example 5

(69) For semiconductor materials selected from PbI.sub.2, HgI.sub.2 a dopant of a halogenated compound was added selected from CsI, BiI.sub.3, CdI.sub.2, SnCl.sub.2 and AgI. For semiconductor materials such as PbO, doping was performed by adding an oxide compound.

(70) The concentration indicated is on the order of a few percentages.

(71) Testing the Wafers and Direct Conversion Detectors

(72) Effects of the Purity of the Starting Powder on the Sensitivity of the X-Ray Detection.

(73) In a series of tests, the powder used to make the ceramics was a commercial powder sold by the company Sigma Aldrich with a purity equal to 99.0% (reference 221090 ACS reagent, 99.0%).

(74) For comparison, wafers were also created using a commercial powder with a purity of 99.999% (supplier reference 203785, 99.999% trace metal basis).

(75) Table 1 below shows the measurements of the dark current and the sensitivity for the wafers made with powder having a purity of 99.0% and 99.999%. For physical reasons, it is not possible to know the exact X-ray dose received by the sample. Only the dose emitted by the source can be known. However, sensitivity measurements performed for two samples of the same size and thickness are comparable. Therefore, we compared the relative sensitivity of a sample in relation to the other one.

(76) TABLE-US-00001 TABLE 1 Purity Study Dark Current (nA/cm.sup.2) Relative Sensitivity Purity 99.0% 251.6 (25%) 3.94 Purity 99.999% 215.9 (19%) 1.0

(77) The dark current must be as weak as possible to have the best signal to noise ratio. Sensitivity refers to the number of loads collected based on the X-ray dose and the electrode surface. Maximizing this quantity is the goal.

(78) Wafers made from the 99.0% pure powder, i.e., with a poor purity with a dark current equivalent to the wafers prepared with a very pure powder, and especially with a greater sensitivity.

(79) Crystallographic Analysis

(80) Crystallographic analysis has shown that the method according to the invention has the effect of promoting grain growth and resorption of the porosity. The grains are not just welded to each other. The smaller grains are more absorbed by the larger ones. This growth of the larger grains at the expense of the smaller grains occurs in any sintering operation, but the method according to the invention makes it possible to increase its efficiency due to the applied load.

(81) Consequently, the polycrystalline material thus obtained has a porosity that is much lower than the material obtained by simple sintering without pressure or by the simple application of pressure without sintering.

(82) Second, when the grains have an anisotropic crystalline structure, maintaining an axial compression or load during sintering can promote the growth of the grains having a minimal mechanical energy and thus making it possible to obtain a polycrystalline material in which the grains are in the preferred orientation.

(83) In the case of HgI.sub.2, this makes it possible to obtain a material having an orientation such that the direction of greater electrical conductivity is parallel to the axis of pressure. The lower porosity and the preferential orientation of the grains produced by sintering under pressure makes it possible to improve the electrical transport properties in comparison with a simple sintering or compaction without sintering.

(84) Third, placing the wafer under pressure, by avoiding evaporation, makes it possible to obtain a better evenness and less surface roughness of the polycrystalline material, which subsequently allows better contact between the electrodes and the surfaces of the material, and to decrease defects on the surface of the material that could degrade the efficiency of the electrical charge collection.

(85) Micro-Structural Study

(86) Comparing the microstructure of various wafers made of semiconductor material to derive the benefit of the orientation by axial compression. The HgI.sub.2 powder, a sintered wafer, and a wafer sintered under pressure are compared. The preferred orientation of the grains in a wafer sintered under pressure and a powder to be analyzed by X-line diffraction. To compare the proportion of a structure oriented along the crystallographic axis C, the area under the diffraction lines characteristic of the crystallographic axis was measured under the diffraction lines characteristic of this crystallographic axis (type (00x) lines, where x is a nonzero positive integer). Table 2 below shows the results for an HgI.sub.2 powder, a sintered wafer from the prior art, and a wafer sintered under pressure in accordance with the invention. The relative intensity is calculated as the ratio between the measurement line (type (00x)) to a reference line (line (102)).

(87) TABLE-US-00002 TABLE 2 Intensity relative to the characteristic lines of the axis C Relative intensity Powder Sintered Sintered under pressure Reference Line (102) 1 1 1 Line (002) 0.27 1.52 5.18 Line (004) 0.22 1.23 3.13 Line (006) 0.06 1.37 1.57

(88) In the case of the ceramic or sintered wafer made under pressure according to the invention, one obtains the greatest values for the characteristic lines for the crystallographic axis C, i.e., of the type (00x) (2a and 2b). There is therefore a preferential orientation which is the most important according to this axis.

(89) Return to Equilibrium Time after an X-Ray Pulse (Lag)

(90) As was disclosed in the description of FIG. 4, of significant importance in the case of an imaging pulse detector, for example, in 3D imaging is the time to return to equilibrium or lag, i.e., the time required for the current produced by an X-ray pulse to return to zero after the end of the pulse. To be able to produce rapid images, this lag must be as short as possible.

(91) The lag on the semiconductor wafers produced according to the invention, as well as references: The CdTe monocrystal and HgI.sub.2 monocrystal to be measured. Table 3 below shows the results.

(92) TABLE-US-00003 TABLE 3 Lag Measurement Hgl.sub.2 Wafers CdTe Monocrystal Hgl.sub.2 Monocrystal Lag 1 ms 66 ms 15 ms

(93) The HgI.sub.2 wafers produced by axial sintering under pressure according to the manufacturing method in accordance with the invention exhibit the lowest lag.

(94) Linearity in Time.

(95) Another performance criterion is the linearity of the direct conversion X-ray detector over time.

(96) The electric charge collected by the semiconductor chip must be proportional to the dose of X-rays received. The linearity of the wafers was measured by subjecting them to a series of X-rays, i.e., a series of X-ray exposures with interposed pauses. We measured the amount of charges collected by the wafer made of semiconductor material, the cumulative quantity over time, by exposing the direct conversion X-ray detector, the X-ray wave stream, composed of a sequence of pulses. The characteristics of the X-ray wave stream process were:

(97) Duration of a pulse: 50 Ms.

(98) Duration of the following dark current: 50 Ms.

(99) The accumulation of charges collected CCC versus time complies with high precision for a linear relationship during the exposure time interval [0.28 s] of the direct conversion X-ray detector:

(100) CCC=a*t, with a=35000/28 in units of charges collected per unit of time (in seconds).

(101) The amount of accumulated collected charges CCC increases from a value of zero before exposition to the X-ray wave stream begins in a linear fashion over time. This is an advantage for use in 3D imaging, that is to say, all the X-ray exposures performed make it possible to collect the same number of charges.

(102) Production of a Direct Conversion X-Ray Detector

(103) The thickness of the wafer made from semiconductor material produced by the method of the invention depends on the particle size of the starting powder and the value of the applied compressive force.

(104) In one example of an embodiment of a direct conversion X-ray detector, a density which was very close to that of a HgI.sub.2 monocrystal (98%) was attained.

(105) The amount of powder was determined by the ratio: Powder mass=sectionthicknessdensity (with a density of 6.36 g/cm.sup.3).

(106) In this embodiment, it was not necessary to make a preform. The mold placed in the furnace (FIG. 1) was used during the sintering under pressure to give the final shape to the wafer manufactured in accordance with the invention.

(107) It was noted that the longer the sintering is, the greater the density and the larger the size of the grains will be.

(108) The wafer exits the furnace in a cylindrical form. It was cut and trimmed to form a rectangular parallelepiped with dimensions of 15 cm15 cm500 m (thickness). The direct conversion X-ray detector was produced by evaporating two conductive electrodes of 50 nm on the two opposing faces of the wafer whose surface state was previously cleaned properly and the unit was deposited on an integrated CMOS circuit as was described using FIG. 3.