SAM PHOTODIODE WITH MULTIPLICATION OF A SINGLE TYPE OF CARRIER IN A PERIODIC MULTILAYER REGION

Abstract

SAM PHOTODIODE WITH MULTIPLICATION OF A SINGLE CARRIER TYPE IN A PERIODIC MULTILAYER REGION This description relates to an avalanche photodiode (5) comprising, in order, an absorption region (10) for receiving radiation (6), a first multiplication region (20a), a second multiplication region (20b), and a collection region (30). The first and second multiplication regions have a lower dopant concentration than the absorption and collection regions. The first multiplication region comprises two to five multilayer structures (22) where each multilayer structure comprises a first layer having a first bandgap and then a second layer having a second bandgap strictly smaller than the first bandgap.The second multiplication region has a constant bandwidth or one that varies in only one direction from the first multiplication region to the collection region. Figure for the abbreviation: Fig. 1.

Description

Title of the invention: SAM PHOTODIODE WITH MULTIPLICATION OF A SINGLE TYPE OF CARRIER IN A PERIODIC MULTI-LAYER REGION technical field

[0001] The present description relates in general to avalanche photodiodes, and more particularly to photodiodes with separate absorption and multiplication regions (SAM-APD, English acronym for "Separate Absorption and Multiplication region - Avalanche PhotoDiode") which perform carrier multiplication by impact ionization of a single type of carrier. Previous technique

[0002] Certain applications of light detection using an avalanche photodiode, such as LIDAR (Light Detection and Ranging) laser remote sensing, quantum optics, or free-space telecommunications, require the detection of a small number of photons per observation time. However, the conservation of information contained in a photon stream is fundamentally limited by the quantum efficiency of the photodiode, which describes its ability to transform each incident photon into an electrical signal, and by the excess noise factor of the photodiode, which reflects the noise added by the multiplication of electrical charges performed by the photodiode.

[0003] US patent 11322637B2 describes a photodiode with separate absorption and multiplication zones based on CdHgTe mercury-cadmium tellurides, having a reduced noise excess factor and comprising an amplification region consisting of repeated multilayer structures. Such a structure allows for the localization of multiplication events and a reduction in the noise excess factor. One drawback is that implementing a stacking arrangement with a large number of multilayer structures can be complex. Another drawback is that high and / or fluctuating doping levels can be observed in the multilayer structures.This results in a spatial variation of the operating electric field in the amplification region, which can induce a large variation in the probability of electric charge multiplication in the amplification region and can, in an extreme case, block carrier transport in layers with a small band gap. Summary of the invention.

[0004] An embodiment overcomes all or part of the disadvantages of photodiodes with known separate absorption and multiplication zones.

[0005] An object of an embodiment is to provide a photodiode with absorption zones and separate multiplication which presents a reduced noise excess factor.

[0006] One embodiment provides for an avalanche photodiode comprising, in order, an absorption region for receiving radiation and doped with a first type of conductivity, a first multiplication region, a second multiplication region, and a collection region doped with a second type of conductivity different from the first type of conductivity, the first and second multiplication regions carrying out carrier multiplication by impact ionization of a single type of carrier and having a lower dopant concentration than those of the absorption region and the collection region, the first multiplication region comprising two to five multilayer structures where each multilayer structure comprises, from the absorption region to the collection region, a first layer having a first band gap and then a second layer having a second band gap,the first bandwidth being strictly greater than the second bandwidth, the second multiplication region having a constant bandwidth or one that varies in only one direction from the first multiplication region to the collection region and having a thickness greater than each second layer.

[0007] According to one embodiment, the first multiplication region comprises two to four multilayer structures, preferably three multilayer structures.

[0008] According to one embodiment, the second multiplication region has a different bandgap width than the second bandgap width.

[0009] According to one embodiment, the second multiplication region has a band gap strictly less than the second band gap.

[0010] According to one embodiment, the second multiplication region has a band gap greater than strictly the second band gap.

[0011] According to one embodiment, the concentration of dopants in the absorption region and the collection region is strictly greater than 1016 atoms / cm3 and the concentration of dopants in the first and second multiplication regions is strictly less than 1016 atoms / cm3.

[0012] According to one embodiment, the first layer of the multilayer structure in contact with the absorption region is thicker than the first layer of the other multilayer structure or of the other multilayer structures.

[0013] According to one embodiment, each multilayer structure comprises, from the absorption region to the collection region, a negative gradient band gap layer between the first layer and the second layer and a positive gradient band gap layer after the second layer, the negative gradient band gap passing from the first band gap to the second band gap.

[0014] According to one embodiment, the positive band gap gradient goes from the second band gap to the first band gap for at least the multilayer structure closest to the absorption region.

[0015] According to one embodiment, the positive gradient layer with band gap width is thicker than the negative gradient layer with band gap width.

[0016] According to one embodiment, the absorption region has a band gap that decreases towards the first multiplication region. Brief description of the drawings

[0017] These features and advantages, as well as others, will be described in detail in the following description of particular embodiments, given by way of non-limiting example, in relation to the accompanying figures, among which:

[0018] [Fig.1] is a partial and schematic cross-sectional view of an embodiment of an avalanche photodiode with separate absorption and multiplication zones;

[0019] [Fig.2] represents a curve of variation of the proportion of cadmium and a curve of variation of the cutoff wavelength in the layers of the photodiode of [Fig.1];

[0020] [Fig.3] illustrates the variation of the band gap within a multilayer structure of the multiplication region of the photodiode [Fig.1];

[0021] Figure 4 shows curves showing the evolution of the total noise excess factor of an amplification region of the photodiode in Figure 1 for different values ​​of the gain of an amplification region; and

[0022] [Fig.5], [Fig.6], [Fig.7], and [Fig.8] are partial and schematic cross-sectional views of other embodiments of an avalanche photodiode with separate absorption and multiplication zones. Description of the implementation methods

[0023] The same elements have been designated by the same reference numerals in the different figures. In particular, structural and / or functional elements common to the different embodiments may have the same reference numerals and may have identical structural, dimensional and material properties.

[0024] For the sake of clarity, only the steps and elements useful for understanding the described embodiments have been shown and are detailed. In particular, the electronic circuits used to bias an avalanche photodiode with separate absorption and multiplication zones and to process the signals supplied by the photodiode are known and are not described further.

[0025] Unless otherwise specified, when referring to two interconnected elements, this means directly connected without intermediate elements other than conductors, and when referring to two connected (in English "coupled") elements between them, this means that these two elements can be connected or linked via one or more other elements.

[0026] In the following description, when reference is made to absolute position qualifiers, such as the terms "front", "back", "top", "bottom", "left", "right", etc., or relative position qualifiers, such as the terms "above", "below", "superior", "inferior", etc., or to orientation qualifiers, such as the terms "horizontal", "vertical", etc., reference is made, unless otherwise specified, to the orientation of the figures.

[0027] Unless otherwise specified, the expressions "approximately", "roughly", and "in the order of" mean within 10%, preferably within 5%.

[0028] According to one embodiment, a SAM-APD photodiode with separate absorption and multiplication zones is considered, which performs carrier multiplication by impact ionization of a single type of carrier (SCM for "Single Carrier Multiplication"). By "carrier multiplication by impact ionization of a single type of carrier," it is understood that the impact multiplication of one of the carrier types (holes, for example) is negligible compared to the impact multiplication of the other carrier type (electrons in this example), that is to say, the ratio between the two multiplication rates is greater than 50, preferably 100, or even 1000.

[0029] Fig. 1 is a partial, schematic cross-sectional view of an embodiment of an avalanche photodiode 5 for the detection and measurement of electromagnetic radiation 6. The wavelength of the radiation 6 is within a given range of wavelengths, for example a wavelength within the range of wavelengths of the near-infrared, for example equal to about 1.55 pm.

[0030] The photodiode 5 is for example produced by means of molecular beam epitaxy growth of the CdHgTe material whose cadmium composition can be modulated in order to vary the bandgap of the material.

[0031] The photodiode 5 comprises an absorption region 10, a collection region 30, and a multiplication region 20 between the absorption region 10 and the collection region 30. The multiplication region 20 performs carrier multiplication by impact ionization, predominantly for electrons (i.e., hole multiplication is negligible compared to electron multiplication). The absorption region 10 comprises a face 11 receiving the radiation 6.

[0032] The photodiode 5 is made from a material that allows for predominant multiplication for one type of carrier. In one embodiment, the photodiode 5 is made from mercury-cadmium telluride, that is to say, from a material comprising telluride and at least one element chosen from cadmium and... mercury and conforming to the following formulation CdxHgi ​​xTe with the value x, also called xCd hereafter, corresponding to the proportion of cadmium relative to mercury, which is between 0 and 1, inclusive. According to another embodiment, the photodiode 5 is made based on indium arsenide-antimonide of the type InAs, v Sby, with the value y corresponding to the proportion of antimony relative to arsenic, which is between 0 and 1, inclusive.

[0033] In operation, the photodiode 5 is subjected to a reverse bias, such as a bias between 10 V and 15 V, for example equal to 13 V. The electrons generated in the absorption region 10 are transported by drift or diffusion to the multiplication region 20. The application of a high potential difference between the absorption region 10 and the collection region 30 will cause a multiplication of the carriers in the multiplication region 20. This multiplication amplifies the initial photocurrent and corresponds to the establishment of a gain, also called multiplication gain or avalanche gain.

[0034] Fig. 2 represents a curve of variation of the proportion of cadmium xCd relative to mercury and a curve of variation of the cutoff wavelength Xc in the layers of the photodiode 5 of Fig. 1 as a function of the distance D measured from face 11 along a direction orthogonal to face 11 when the photodiode 5 is made from mercury-cadmium telluride.

[0035] The absorption region 10 has a first type of conductivity, for example type P, and thickness and bandgap characteristics adapted to absorb most of the light radiation 6 received by the photodiode 5. In one embodiment, the absorption region 10 has a doping level typically greater than 10¹⁶ atoms / cm³, for example between 10¹⁶ atoms / cm³ and 10¹⁸ atoms / cm³. Preferably, the doping is obtained by adding arsenic (As), gold (Au), or antimony (Sb). The thickness and the proportion of cadmium in the absorption region 10 are adapted to the target wavelength for each application. To detect photons with a wavelength of 1.55 pm, the thickness is typically between 1 pm and 3 pm, preferably between 1 pm and 2 pm. The proportion of cadmium xCd is typically greater than 0.3, for example greater than 0.45.According to one embodiment, the proportion of cadmium xCd in the absorption region 10 decreases from face 11 to multiplication region 20. According to one embodiment, the proportion of cadmium xCd is between 0.6 and 0.42.

[0036] The multiplication region 20 is characterized by a low doping level compared to the absorption region 10, typically less than 1016 atoms / cm3 and preferably less than 1015 atoms / cm3, in order to establish a uniform electric field across the multiplication region 20.

[0037] According to one embodiment, the multiplication region 20 comprises the stacking of a first multiplication region 20a and a second multiplication region 20b. The second multiplication region 20b is in direct physical contact with the first multiplication region 20a. The first multiplication region 20a is in direct physical contact with the absorption region 10 and the second multiplication region 20b is in direct physical contact with the collection region 30. According to one embodiment, the first multiplication region 20a is not in direct physical contact with the collection region 30 and the second multiplication region 20b is not in direct physical contact with the absorption region 10.

[0038] According to one embodiment, the first multiplication region 20a comprises a plurality of multilayer structures 22. The material characteristics of the multilayer structures 22 favor the localization of impact ionization events within a fraction of the thickness of each multilayer structure 22 and the multiplication of a single type of carrier corresponding to the minority carriers in the absorption region 10, for example, electrons. The first multilayer structure 22 is the multilayer structure 22 of the first multiplication region 20a closest to the absorption region 10, and the last multilayer structure 22 is the multilayer structure 22 of the first multiplication region 20a closest to the second multiplication region 20b.

[0039] Fig. 3 illustrates the variation of the band gap separating the valence bands Ev and conduction bands Ec of an embodiment of a multilayer structure 22 from the first unpolarized multiplication region 20a as a function of the distance D measured along a direction orthogonal to the face 11 from the edge of the multilayer structure 22 located on the side of the absorption region 10.

[0040] Each multilayer structure 22 comprises successively, in a direction from the absorption region 10 to the collection region 30, a first layer with a constant bandwidth 22a having a substantially constant bandwidth equal to a first value AEb, a first transition layer 22b, a second layer with a constant bandwidth 22c having a second substantially constant bandwidth equal to a second value AE2, and a second transition layer 22d. In the first transition layer 22b, the bandwidth varies from the first bandwidth AEi to the second bandwidth AE2.In the second transition layer 22d, the bandgap varies from the second bandgap AE2 to the first bandgap AEi when the multilayer structure 22 is followed by another multilayer structure 22 and varies from the second bandgap AE2 to the bandgap of the second multiplication region 20b for the last multilayer structure 22.

[0041] According to one embodiment, the first forbidden bandwidth AEi is su greater than the second bandgap AE2. In such a way, the second constant bandgap layer 22c has an average carrier multiplication rate per micrometer higher than the average carrier multiplication rate per micrometer of the first constant bandgap layer 22a. To achieve this, according to one embodiment, the first and second constant bandgap layers 22a, 22c are made of mercury-cadmium tellurides of the type CdxHgi ​​xTe with different proportions of cadmium x, and more particularly a maximum proportion of cadmium xCdi in the first constant bandgap layer 22a greater than the minimum proportion of cadmium xCd2 in the second constant bandgap layer 22c.According to one embodiment, the maximum proportion of cadmium xCdi in the first constant band gap layer 22a is between 0.4 and 0.5, for example equal to 0.42. According to one embodiment, the minimum proportion of cadmium xCd2 in the second constant band gap layer 22c is between 0.33 and 0.4, for example equal to 0.34.

[0042] According to one possible embodiment, the first transition layer 22b has a negative bandgap gradient and the second transition layer 22d has a positive bandgap gradient. The negative bandgap gradient passes from the first bandgap AEi to the second bandgap AE2, and the positive bandgap gradient passes from the second bandgap AE2 to the first bandgap AEi when the multilayer structure 22 is followed by another multilayer structure 22, and passes from the second bandgap AE2 to the bandgap of the second multiplication region 20b for the last multilayer structure 22.

[0043] In one embodiment, the second transition layer 22d with a positive gradient band gap is thicker than the first transition layer 22b with a negative gradient band gap. The negative gradient is thus steeper than the positive gradient. In one embodiment, between the first band gap AE1 and the second band gap AE2, the relative change, in absolute value, of the conduction band Ec is greater than the relative change, in absolute value, of the valence band Ev. This ensures fast drift transport and response time of the photodiode. In one embodiment, the change in the band gap of the second transition layer 22d is substantially linear.

[0044] According to one embodiment, the thickness of each multilayer structure 22 is between 400 nm and 800 nm. According to another embodiment, for each multilayer structure 22, the ratio between the thickness of the first constant bandgap layer 22a and the thickness of the multilayer structure 22 is between 45% and 80%, for example, equal to approximately 60%. According to one embodiment, for each multilayer structure 22, the ratio between the thickness of the second constant-bandwidth layer 22c and the thickness of the multilayer structure 22 is between 20% and 50%, for example, equal to approximately 25%. According to one embodiment, for each multilayer structure 22, the ratio between the thickness of the second transition region 22d and the thickness of the multilayer structure 22 is between 5% and 20%, for example, equal to approximately 10%. According to one embodiment, for each multilayer structure 22, the ratio between the thickness of the first transition region 22b and the thickness of the multilayer structure 22 is between 1% and 10%, for example, less than 5%.

[0045] According to one embodiment, the number Nb of multilayer structures 22 in the first absorption region 20a is greater than or equal to 2 and less than or equal to 5, preferably equal to 3. The reduced number Nb of multilayer structures 22 results in a more homogeneous electric field in this part of the structure. This allows for a similar multiplication probability in the different multilayer structures 22, thus reducing the noise excess factor to an approximate value close to one. In one embodiment, the thickness of the first multiplication region 20a is between 0.8 µm and 4 µm. The first multiplication region 20a provides a high gain and a low noise excess factor with reduced sensitivity to variations in the residual doping in the first multiplication region 20a.

[0046] According to one possible embodiment, the positive bandwidth gradient goes from the second bandwidth AE2 to the first bandwidth AEi, such that the multilayer structures 22 have the same bandwidths AEi and AE2 within the first and second layers with constant bandwidth 22a and 22c.

[0047] According to one embodiment, the last multilayer structure 22 of the first multiplication region 20a differs from the other multilayer structures 22 by the characteristic of the second transition region 22d. In this region, the thickness and the gradient of cadmium proportion must be adapted to allow the carriers to transition between this region and the second multiplication region 20b.

[0048] The second multiplication region 20b has a substantially constant proportion of cadmium xCd or a proportion of cadmium xCd that exhibits a gradient in one direction from the first multiplication region 20a to the collection region 30. In one embodiment, the thickness of the second multiplication region 20b is between 1 µm and 5 µm. The proportion of cadmium xCd and the thickness of the second multiplication region 20b are adapted to obtain a suitable level of gain and dark noise for each application. In one embodiment, In the case of an application requiring a high operating temperature, the bandwidth of the second multiplication region 20b can be greater than the second bandwidth AE2. According to another embodiment, in the case where the aim is to maximize the gain, for example in the case of applications requiring the detection of single photons, the bandwidth of the second multiplication region 20b can be less than the second bandwidth AE2.

[0049] The dopant concentration in the first multiplication region 20a and in the second multiplication region 20b is lower than the dopant concentration in the absorption region 10 and in the collection region 30, preferably less than 10¹⁵ atoms / cm³. The first multiplication region 20a may be doped with the same type of conductivity as the second multiplication region 20b or with the opposite type of conductivity to that of the second multiplication region 20b. In one embodiment, the doping level in the first multiplication region 20a and in the second multiplication region 20b corresponds to the residual doping level imposed by the material composing these regions and the growth process implemented for the formation of the multiplication region 20. In the case where the photodiode 5 is based on CdHgTe, the residual doping is of type N at a level less than 10¹⁵ atoms / cm³.The level of residual doping will influence the field profile across multiplication regions 20a and 20b. Advantageously, the impact of this modulation of the electric field across photodiode 5 is reduced compared to a photodiode whose multiplication region 20 comprises a number Nb of multilayer structures greater than 4. The presence of the second multiplication region 20b reduces the sensitivity of photodiode 5 to a variation in residual doping in the first multiplication region 20a.

[0050] The collection region 30, with which the multiplication region 20 is in direct physical contact in the embodiment of [Fig. 1], is characterized by a second type of conductivity, for example, type N. According to one embodiment, the collection region 30 has a doping level typically greater than 10¹⁶ atoms / cm³, preferably greater than 10¹⁷ atoms / cm³. The proportion of cadmium xCd in the collection region 30 is, for example, between 0.35 and 0.5. According to one embodiment, the thickness of the collection region 30 is between 0.5 pm and 2 pm, and is, for example, equal to 0.8 pm. According to one embodiment, the gap width in the collection region 30 is preferably greater than the gap width of the second multiplication region 20b in order to reduce the generation of a dark current on faults or by tunneling in this region.In this case, a bandwidth gap gradient is also created between the second multiplication region 20b and the collection region 30 to facilitate the transport of charges. between the two regions.

[0051] In operation, the photodiode 5 is subjected to a reverse bias, such as a bias between 10 V and 15 V, for example equal to 13 V. The electrons generated in the absorption region 10 are transported by drift or diffusion to the multiplication region 20. The application of a high potential difference between the absorption region 10 and the collection region 30 will cause a multiplication of the carriers in the multiplication region 20. This multiplication amplifies the initial photocurrent and corresponds to the establishment of an avalanche gain.

[0052] In operation, the first constant-bandgap layer 22a forms a carrier acceleration layer that increases the electron energy without initiating electron multiplication events, while the second constant-bandgap layer 22c forms a multiplication layer in which electron multiplication events are localized. The first multiplication region 20a with a plurality of multilayer structures 22 thus allows for significant localization of multiplication events within the second constant-bandgap layers 22c. In this way, the multiplication randomness is reduced and consequently the noise excess factor is decreased.

[0053] The reduced number Nb of multilayer structures 22 in the first multiplication region 20a reduces the achievable gain M2a in this part of the structure. For example, the gain M2a will approach 8 for Nb equal to 3 in the optimal case with a multiplication probability equal to 1 per multilayer structure 22.

[0054] The second multiplication region 20b of the multiplication region 20 is designed to introduce a strong gain M2b, which makes it possible to increase the total gain Mtot in the photodiode 5 which is equal to the product of the gain M2a of the first multiplication region 20a and the gain M2b of the second multiplication region 20b.

[0055] By the order of implementation of the first multiplication region 20a and the second multiplication region 20b and by the fact that only one type of carrier is multiplied in the first multiplication region 20a, the first multiplication region 20a acts as a preamplifier whose gain fluctuation characteristics and noise excess factor are dominant for the total gain fluctuations through the two multiplication regions 20a and 20b of the photodiode 5.

[0056] By calling F2a the noise excess factor of the first multiplication region 20a and F2b the noise excess factor of the second multiplication region 20b, the total noise excess factor Ftot of the amplification region 20, considering independent multiplication processes in the two multiplication regions 20a and 20b, is given by the following formula: [Math 1] T? _1 ,%«_1 , , ( tj l\ , (P2b'D ^tot“ M?o, + MiM, -1+kr2a-JJ+ M2a

[0057] Figure 4 shows curves C1, C2, C3, C4, and C5 representing the evolution of an estimate of the total noise excess factor Ftot of the amplification region 20 of the photodiode 5 for different values ​​of the gain M2a of the first amplification region 20a and for different values ​​of the noise excess factor F2b of the second multiplication region 20b. Curves C1, C2, C3, C4, and C5 are obtained with the gain M2b of the second multiplication region 20b equal to 30 and the noise excess factor F2a of the first multiplication region 20a equal to 1.01. Curves C1, C2, C3, C4, and C5 are obtained with the noise excess factor F2b equal to 10, 3, 2, 1.5, and 1.24, respectively. The values ​​obtained on line A correspond to a number Nb equal to 1, which corresponds to a gain M2a equal to 2. The values ​​obtained on line B correspond to a number Nb equal to 2, which corresponds to a gain M2a equal to 4.The values ​​obtained on line C correspond to a number Nb equal to 3, which corresponds to a gain M2a equal to 8. The values ​​obtained on line D correspond to a number Nb equal to 4, which corresponds to a gain M2a equal to 16.

[0058] The estimates in [Fig.4] highlight that a low excess noise factor Ftot can be obtained even with a second multiplication region 20b with a very random multiplication characterized by a multiplication of the two types of carriers and, consequently, an excess noise factor F2b greater than 5. In this typical case, for a multiplication layer made with a III-V semiconductor, the use of at least 3 multilayer structures 22 allows the total excess noise factor at Ftot to be less than 1.2. In the case of the implementation of a second multiplication region 20b with a less random multiplication with an excess noise factor F2b less than 2, as is the case when the photodiode 5 is made with the HgCdTe semiconductor, the total excess noise factor at Ftot is less than 1.1 already when Nb is equal to 2 and tends towards Fia, i.e. close to unity, when Nb is greater than or equal to 3.

[0059] Estimates in [Fig. 4] show that the multiplication region 20, comprising the first multiplication region 20a and the second multiplication region 20b, allows for a randomness dominated by gain fluctuations in the first multiplication region 20a. The second multiplication region 20b can thus be adjusted to obtain the required total gain Mtot for the application, without regard to the potential excess noise generated in this region. For example, for single-photon detection applications, it is important to obtain a very high total gain Mtot to generate signals above the noise of the electronic circuits connected to the photodiode and, consequently, to achieve good detection efficiency.

[0060] According to one embodiment, if the second multiplication region 20b is made of HgCdTe, the gain in the second multiplication region 20b can be increased by decreasing the band gap in this region compared to the minimum band gap AE2 in the first multiplication region 20a. In this case, a very high total gain Mtot is obtained, associated with a total noise excess factor Ftot close to unity. However, using a smaller band gap in the second multiplication region 20b will lead to an increase in the generation of thermal generation events, which will induce an increase in detector noise in the dark.The impact of this noise will be limited by the architecture of photodiode 5, which has the advantage of amplifying generation events, thermal or photonic, less in the second multiplication region 20b compared to signals that are amplified by the total gain. This allows for a better signal-to-noise ratio compared to a homogeneous multiplication layer with a low bandwidth gap.

[0061] Fig. 5, Fig. 6, Fig. 7, and Fig. 8 are partial, schematic cross-sectional views of other embodiments of a photodiode.

[0062] Figure 5 is a partial, schematic cross-sectional view of another embodiment of a photodiode 60. The photodiode 60 comprises all the elements of the photodiode 5 shown in Figure 1 and further includes a passivation layer 40 covering the collection region 30 on the side opposite the second multiplication region 20b, thus protecting the photodiode 60 from electrical degradation induced by a mechanical or chemical mechanism. The photodiode 70 further includes a metal pad 50 located in and around a hole 42 in the passivation layer 40, in order to provide an electrical contact, ideally ohmic, with the collection region 30. An electrical contact (not shown) is also made with the absorption region 10 in order to apply a polarization between the absorption region 10 and the collection region 30 through the multiplication region 20.

[0063] Figure 6 is a partial, schematic cross-sectional view of another embodiment of a photodiode 70. The photodiode 70 comprises all the elements of the photodiode 60 shown in Figure 5, except that the multiplication region 20 and the collection region 30 have a mesa-type structure on the absorption region 10. That is, the stacking 62 of the multiplication region 20 and the collection region 30 forms a raised element with inclined sides 64 resting on the absorption region 10 with a lateral dimension, measured parallel to the face 6, that is smaller than the lateral dimension of the absorption region 10. The passivation layer 40 further extends over the inclined sides 64 of the collection region 30 and the multiplication region 20, and extends over the absorption region 10. direct physical contact of the absorption region 10 around the stack 62. The stack 62 can be formed by an engraving defining a mesa-shaped pillar.

[0064] Figure 7 is a partial, schematic cross-sectional view of another embodiment of a photodiode 80. The photodiode 80 comprises all the elements of the photodiode 70 shown in Figure 6, except that the collection region 30 includes a central region 31 and a peripheral region 32 surrounding the central region 31 and interposed between the flanks 64 and the central region 31. The doping level of the peripheral region 32 is lower than the doping level of the central region 31. According to one embodiment, the doping level of the peripheral region 32 is less than 10¹⁵ atoms / cm³, and is, for example, substantially equal to the residual doping level imposed by the material composing the collection region 30 and the growth process implemented for the formation of the collection region 30.In the case where the photodiode 80 is based on CdHgTe, the residual doping in the peripheral region 32 is of the N type at a level below 10¹⁵ atoms / cm³. This advantageously limits the concentration of the electric field on the stack's flanks 64. According to one embodiment, the doping level of the central region 31 is greater than 10¹⁷ atoms / cm³. The central region 31 can be formed by ion implantation.

[0065] Fig. 8 is a partial, schematic cross-sectional view of another embodiment of a photodiode 90. The photodiode 90 comprises all the elements of the photodiode 60 shown in Fig. 5, the collection region 30 further comprising the central region 31 surrounded by the peripheral region 32 as for the photodiode 80 shown in Fig. 7.

[0066] According to one embodiment, P-type doping by mercury vacancies in the structure is first induced throughout the photodiode following molecular beam epitaxial growth of the layers composing the absorption region 10, the multiplication region 20, and the collection region 30. When the photodiode 90 is CdHgTe-based, this doping can be obtained by vacuum annealing at a temperature close to 300°C to achieve a doping level close to 10¹⁶ atoms / cm³. The formation of the central N-type region 31 at a level typically greater than 10¹⁷ atoms / cm³ and localized in the collection region 30 will induce the removal of mercury vacancies and, consequently, the formation of a zone 91 in the multiplication region 20 and the collection region 30 characterized by N-type doping at a level less than 10¹⁵ atoms / cm3, corresponding to the residual doping of the material composing photodiode 90.The remaining zone 92 of the absorption region 10, the multiplication region 20, and the collection region is still characterized by P-type doping. The boundary between zones 91 and 92 is schematically represented by a dashed line 93 in [Fig. 8]. In this case, the multiplication function with very low noise excess takes place in the part of the multiplication region 20 contained within it. in zone 91, that is to say in the center of photodiode 90. This embodiment has the advantage of eliminating the need for a mesa-type structure to delimit the active zone of the photodiode.

[0067] Various embodiments and variations have been described. A person skilled in the art will understand that certain features of these various embodiments and variations could be combined, and other variations will become apparent to a person skilled in the art.

[0068] Finally, the practical implementation of the embodiments and variants described is within the reach of a person skilled in the art, based on the functional indications given above.

Claims

Demands

1. Avalanche photodiode (5; 60; 70; 80; 90) comprising, in order, an absorption region (10) for receiving radiation (6) and doped with a first type of conductivity, a first multiplication region (20a), a second multiplication region (20b), and a collection region (30) doped with a second type of conductivity different from the first type of conductivity, the first and second multiplication regions (20a, 20b) carrying out carrier multiplication by impact ionization of a single type of carrier and having a lower dopant concentration than the absorption region (10) and the collection region (30), the first multiplication region (20a) comprising two to five multilayer structures (22) where each multilayer structure comprises, from the absorption region (10) to the collection region (30),a first layer (22a) having a first bandgap (AEJ) and then a second layer (22c) having a second bandgap (AE2), the first bandgap being strictly greater than the second bandgap, the second multiplication region (20b) having a constant bandgap or one that varies in only one direction from the first multiplication region (20a) to the collection region (30) and having a thickness greater than each second layer (22c).

2. Avalanche photodiode according to claim 1, wherein the first multiplication region (20a) comprises two to four multilayer structures (22), preferably three multilayer structures (22).

3. Avalanche photodiode according to claim 1 or 2, wherein the second multiplication region (20b) has a band gap different from the second band gap (AE2).

4. Avalanche photodiode according to claim 3, wherein the second multiplication region (20b) has a band gap strictly less than the second band gap (AE2).

5. Avalanche photodiode according to claim 3, wherein the second multiplication region (20b) has a band gap greater than strictly the second band gap (AE2).

6. Avalanche photodiode according to any one of claims 1 to 5, wherein the concentration of dopants in the absorption region (10) and of the collection region (30) is strictly greater than 1016 atoms / cm3 and in which the concentration of dopants of the first and second multiplication regions (20a, 20b) is strictly less than 1016 atoms / cm3.

7. Avalanche photodiode according to any one of claims 1 to 6, wherein the first layer (22a) of the multilayer structure (22) in contact with the absorption region (10) is thicker than the first layer of the other multilayer structure or of the other multilayer structures.

8. Avalanche photodiode according to any one of claims 1 to 7, wherein each multilayer structure (22) comprises, from the absorption region (10) to the collection region (30), a layer (22b) with a negative band gap gradient between the first layer (22a) and the second layer (22c) and a layer (22d) with a positive band gap gradient after the second layer (22c), the negative band gap gradient passing from the first band gap (AEi) to the second band gap (AE2).

9. Avalanche photodiode according to claim 8, wherein the positive bandgap gradient goes from the second bandgap (AE2) to the first bandgap (AEi) for at least the multilayer structure (22) closest to the absorption region (10).

10. Avalanche photodiode according to claim 8 or 9, wherein the positive band gap gradient layer (22d) is thicker than the negative band gap gradient layer (22b).

11. Avalanche photodiode according to any one of claims 1 to 10, wherein the absorption region (10) has a band gap that decreases in the direction of the first multiplication region (20a).

Images