Method for manufacturing an image sensor
Patent Information
- Application Number
- EP2023841606
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2022-12-28
- Filing Date
- 2023-12-22
- Publication Date
- 2025-11-05
AI Technical Summary
The existing methods for manufacturing image sensors sensitive to both visible light and near-wave infrared rays require the use of multiple layers of different semiconductor materials, leading to reduced quantum efficiency and limited resolution for visible light pixels due to the use of materials like InGaAs for both visible and infrared detection.
A method involving a silicon-on-insulator substrate where cavities are formed in the first semiconductor material to create visible light pixels, followed by the growth of a second semiconductor material within these cavities for infrared detection, allowing both types of pixels to be in the same plane without vertical stacking, with an insulating protective layer to isolate them electrically.
This approach maintains quantum efficiency and allows for independent optimization of pixel dimensions for visible and infrared light detection, eliminating the resolution limitations of traditional methods.
Smart Images

Figure 1.1
Abstract
Description
[0001] Description
[0002] Title: Manufacturing process of an image sensor
[0003] Technical field
[0004] The present disclosure relates to a method of manufacturing an image sensor sensitive to visible light as well as near-wave infrared rays.
[0005] State of the art
[0006] The manufacture of image sensors sensitive to visible light as well as to near infrared rays or short-wave infrared rays (SWIR) involves using distinct materials to form two different types of pixels: a first series of pixels sensitive to visible light, i.e., whose wavelength is between 380 and 780 nm, and a second series of pixels sensitive to near infrared rays, i.e., whose wavelengths range from 780 nm to approximately 3 pm.
[0007] Silicon is commonly used in the manufacture of visible light-sensitive sensors, due to its semiconductor properties. Silicon-on-insulator (SOI) substrates are used, comprising a doped silicon base substrate, an intermediate layer of silicon oxide, called the buried oxide layer, and a so-called active silicon layer, with a doping that may be different from that of the base substrate, in which the pixels are formed.
[0008] However, silicon is not suitable for detecting infrared rays, whose photon energy is too low to be detected by wide bandgap materials such as silicon. Therefore, narrower bandgap materials are used, such as III-V semiconductors, including for example indium phosphide (InP) alloys, or germanium (Ge) alloys, for the manufacture of infrared cameras or sensors.
[0009] InP alloys are particularly used, however InP is a relatively rare material, only occurring in the form of small diameter substrates (less than 15cm), brittle and relatively difficult to handle.
[0010] Therefore, the manufacture of sensors sensitive simultaneously to visible light and infrared rays according to the state of the art requires the use of several layers of different materials, originating from different substrates. The sensors obtained by this type of method typically comprise at least one silicon substrate, a layer of a semiconductor material sensitive to visible light, and a layer of a different semiconductor material, sensitive to infrared rays. An alternative is to use III-V semiconductors for the detection of visible light but also infrared rays, as envisaged by document WO 2015 / 048304A2. For this type of sensors, illustrated in Figure 1, a layer 104 of indium gallium arsenide (InGaAs) serves as a detection layer for visible waves as well as near infrared waves, the thickness of this layer being between 0.5 pm and 6 pm.This layer 104 of InGaAs is formed by epitaxial growth on a substrate 101 of InP, and is covered with an additional layer of InP 105 serving as a contact layer. The surface layer 105 of indium phosphide is particularly thin, so as to allow the passage of visible light towards the layer 104 of InGaAs.
[0011] However, the presence of the 105 indium phosphide layer still reduces the quantum efficiency of the sensor at certain visible light wavelengths, typically around 600nm. In addition, using the same 104 InGaAs layer for both visible and infrared detection means that the dimensions and spacing of the visible pixels must be the same as for the SWIR pixels. The resolution of the sensor is therefore limited by the resolution of the infrared pixels, which is typically much lower than that desired for visible pixels. This solution therefore does not provide satisfactory resolution for visible light pixels.
[0012] Summary
[0013] In order to overcome these drawbacks, the present disclosure provides a method of manufacturing an image sensor for detecting visible light and shortwave infrared rays, the method comprising the following steps: a. providing a support substrate comprising a first semiconductor material; b. forming, in the support substrate, cavities to define visible light-sensitive pixels in the first semiconductor material between said cavities; c. forming an electrically insulating protective layer at least on the lateral surfaces of each visible light-sensitive pixel; d. growing, in the cavities, a second material, different from the first material, so as to form shortwave infrared-sensitive pixels.
[0014] This produces a sensor sensitive to both visible light rays and shortwave infrared rays, with pixels sensitive to both types of rays located in the same plane, without having to vertically stack successive layers of semiconductor materials. The resulting sensor has no loss of quantum efficiency, and the resolution of its pixels sensitive to visible light rays is not limited by that of its pixels sensitive to shortwave infrared rays.
[0015] According to one embodiment, the method comprises, between step a and step b, a step a' of forming a protective film of the support substrate, to protect the pixels sensitive to visible light during step b of forming the cavities; and between step c and step d, a step c' of removing the protective film in order to expose the upper surface of the pixels sensitive to visible light.
[0016] According to one embodiment, the method comprises, between step a and step b, a step a' of forming a protective film of the support substrate, to protect the pixels sensitive to visible light during step b of forming the cavities; and after step d, a step d' of removing the protective film in order to expose the upper surface of the pixels sensitive to visible light.
[0017] According to one embodiment, step c comprises the formation of the protective layer by rapid thermal annealing.
[0018] According to one embodiment, the first semiconductor material is silicon.
[0019] According to one embodiment, the second semiconductor material is a III-V semiconductor, preferably chosen from indium phosphide, indium gallium arsenide, germanium alloys, or quantum dots, for example colloidal quantum dots based on lead(II) sulfide.
[0020] According to one embodiment, the support substrate is of the silicon-on-insulator type, and successively comprises a base substrate, an electrically insulating intermediate layer and a monocrystalline layer of the first semiconductor material.
[0021] According to one embodiment, the method comprises, between step b and step c, a step b' of thickening the monocrystalline layer by homoepitaxy.
[0022] According to one embodiment, step b is an etching step carried out in the support substrate until the electrically insulating intermediate layer is reached.
[0023] According to one embodiment, the method comprises, after step c, and optionally step c', a step c” of removing, in the cavities, the material forming the electrically insulating intermediate layer, and in that the pixels sensitive to short-wave infrared rays are formed, in step d, on the face of the base substrate revealed by the removal step c”.
[0024] According to one embodiment, the support substrate is a solid substrate of the first semiconductor material.
[0025] The present disclosure further provides an image sensor for detecting visible light and shortwave infrared rays, comprising: a support substrate; a plurality of pixels sensitive to visible light, formed in the support substrate, and pixels sensitive to shortwave infrared rays, said pixels sensitive to infrared rays being arranged on a surface region of the support substrate; a protective layer arranged at least on the side walls of each pixel so as to electrically insulate each pixel from adjacent pixels; the pixels forming an arrangement arranged on the same plane, the support substrate being of the semiconductor on insulator (SOI) type, in particular silicon on insulator, the support substrate successively comprising a base substrate in a semiconductor material, an electrically insulating intermediate layer and a monocrystalline layer in a semiconductor material.
[0026] Brief description of the figures
[0027] Figure 1 illustrates a state-of-the-art visible light and near infrared sensitive sensor.
[0028] Figure 2 illustrates a typical SOI substrate.
[0029] Figure 3 illustrates an epitaxial growth step of the SOI monocrystalline layer, according to one embodiment.
[0030] Figures 4a, 4b, 4c, 4d, 4e and 4f illustrate the steps of the proposed method, according to a first embodiment.
[0031] Figures 5a, 5b, 5c, 5d and 5e illustrate the steps of the proposed method, according to a second embodiment.
[0032] Figure 6 shows an arrangement of pixels on a sensor, known as a "Bayer arrangement". To make the figures easier to read, the various elements have not necessarily been shown to scale.
[0033] The same reference sign used from one figure to another designates the same element, which will not be described in detail again.
[0034] Detailed description of embodiments
[0035] The present disclosure relates to a method of manufacturing an image sensor, capable of detecting visible light as well as shortwave infrared rays, as well as to the sensors obtained by this method.
[0036] The method comprises a first step of providing a support substrate. This substrate comprises at least a first semiconductor material, in which the pixels sensitive to visible light will be formed. More particularly, the first semiconductor material may be silicon. According to a first embodiment, the support substrate is a silicon-on-insulator (SOI) type substrate, which successively comprises a base substrate, an electrically insulating intermediate layer and a monocrystalline layer of a first semiconductor material.
[0037] If the monocrystalline layer of the first semiconductor material has a thickness less than the desired thickness for the pixels sensitive to visible light, the method may comprise a growth step aimed at thickening said layer. This step may be an epitaxial growth step, in particular homoepitaxy. For example, the thickness of the pixels sensitive to visible light is advantageously between 1 μm and 10 μm. According to a second embodiment, the support substrate is formed entirely from the first semiconductor material, and may in particular be a bulk silicon substrate.
[0038] The method comprises a step of forming cavities in the first semiconductor material of the support substrate. This makes it possible to delimit in the first semiconductor material, between the cavities, a plurality of pixels sensitive to visible light. As will be seen below, the cavities will subsequently be filled with a second semiconductor material sensitive to infrared rays. Consequently, the size of the cavities is advantageously chosen to optimize the size of the pixels sensitive to infrared rays and the distance between adjacent cavities is chosen to optimize the size of the pixels sensitive to visible light. For example, but in a non-limiting manner, the distance between two adjacent cavities (corresponding substantially to the width of a pixel sensitive to visible light is between 0.25 and 2 μm, and the width of the cavities (corresponding substantially to the width of a pixel sensitive to infrared rays) is between 1 and 10 μm.
[0039] Prior to this cavity formation step, the proposed method may comprise a protective film formation step. Such a step aims to protect certain parts of the support substrate, and to leave other parts of the support substrate unprotected, the unprotected parts then forming the cavities. The cavity formation step may be a plasma-assisted dry etching step.
[0040] The protective film can then be used to protect certain parts of the support substrate during etching. It can be a film that acts as a mask, protecting the support substrate from the etching agent.
[0041] The protective film can be formed from silicon nitride, which has the advantage of exhibiting high etching selectivity towards silicon and III-V materials.
[0042] After forming the cavities, the method further comprises a step of forming an electrically insulating protective layer on the pixels sensitive to visible light, in particular on their lateral surfaces. When the bottom of the cavities is made of semiconductor material, the protective layer also extends over this surface. This layer can be obtained by rapid thermal annealing, in particular in an oxidizing atmosphere.
[0043] The proposed method also comprises growing, in the cavities, a second semiconductor material, different from the first semiconductor material, which is sensitive to near infrared rays. The second material is advantageously chosen from III-V semiconductors, such as alloys of indium phosphide (InP), gallium-indium arsenide (InGaAs) or germanium (Ge), or else be formed from colloidal quantum dots such as colloidal quantum dots based on lead(II) sulfide or lead selenide.
[0044] The second semiconductor material grows by heteroepitaxy on the first semiconductor material. For this purpose, if the bottom of the cavities is covered with the electrically insulating protective layer, it is necessary to first remove said layer to expose the first semiconductor material.
[0045] To grow the second semiconductor material in the cavities, the surface of the pixels sensitive to visible light is advantageously protected by means of a mask, for example made of silicon nitride.
[0046] The first embodiment uses, as support substrate 1, an SOI, as shown in FIG. 2. Such an SOI successively comprises a base substrate 13 made of semiconductor material, generally silicon, an intermediate layer 14 made of electrically insulating material, in particular silicon oxide, and a layer 15 made of monocrystalline semiconductor material, also generally silicon. The thickness of the monocrystalline semiconductor layer is typically between 50 nm and a few pm.
[0047] As illustrated in Figure 4a, the proposed method may comprise a step of forming a protective film 3, which will selectively protect certain parts of the substrate 1 and leave other parts unprotected.
[0048] Figure 4b represents a step of forming pixels 11 sensitive to visible light. During this step, cavities 5 are made in the monocrystalline layer 15 of the support substrate 1, until the insulating intermediate layer 14 is reached. According to a preferred embodiment, these cavities are made by plasma-assisted dry etching, so as to etch the unprotected parts of the substrate 1, while the parts protected by the protective film 3 do not undergo etching.
[0049] Such etching is generally anisotropic, essentially oriented along the thickness direction of the substrate 1, making it possible to obtain cavities 5 with sides substantially parallel to this direction.
[0050] The distance between the cavities, which corresponds substantially to the width of the parts of the support substrate 1 covered by the protective film 3, is chosen so as to obtain the pixels 11 sensitive to visible light of a desired dimension. When the monocrystalline layer 15 of the SOI is too thin to allow the formation of pixels 11 sensitive to visible light of the desired size, an epitaxial growth step can be provided so as to thicken the monocrystalline layer 15, for example to a thickness of the order of a few micrometers. Such a step is illustrated in FIG. 3. This layer being preferably formed of silicon, this step will in this case be a step of homoepitaxy of monocrystalline silicon on the monocrystalline layer 15.
[0051] This step makes it possible in particular to choose the thickness of the pixels 11 sensitive to visible light, this thickness depending directly on that of the monocrystalline layer 15 of the SOI.
[0052] As illustrated in Figure 4c, once the cavities are obtained, an electrically insulating protective layer 4 is formed so as to protect at least the side walls of each pixel sensitive to visible light. According to one embodiment, this protective layer is obtained by thermal oxidation which can implement rapid thermal annealing. When such thermal annealing is carried out in an oxidizing atmosphere, a layer of silicon oxide forms on the surface of the pixels 11 sensitive to visible light, forming the protective layer 4. Furthermore, the protective layer can also be formed by deposition. The function of this protective layer 4 is to create electrical insulation between the pixels 11 sensitive to visible light and pixels 12 sensitive to near infrared rays which will be formed according to a later step of the method.
[0053] The use of an SOI as a support substrate thus makes it possible to obtain a layer of silicon oxide present both on the side walls of the pixels 11 sensitive to visible light and between these pixels and the base substrate 13 of the support substrate 1. Indeed, the protective layer 4 of silicon oxide is already partially formed by the intermediate layer 14 of the SOI serving as support substrate 1. This not only allows electrical insulation of the pixels 11 sensitive to visible light with respect to the pixels 12 sensitive to near infrared, but also with respect to the rest of the support substrate 1, which limits the transconductance phenomena between pixels.
[0054] The method then comprises a step of removing the parts of the protective layer 4 (or 14) extending at the bottom of the cavities, so as to reveal the base substrate 13 of the support substrate 1 as illustrated in FIG. 4d. The protective layer is then located only around the side walls of each pixel 11 sensitive to visible light, as well as between these pixels 11 and the base substrate 13 of the support substrate 1.
[0055] This makes it possible to reveal an upper surface of the support substrate 1 in order to form the pixels 12 sensitive to infrared rays there, as will be explained later.
[0056] Finally, the method comprises a step of forming pixels 12 sensitive to infrared rays, on the parts of the support substrate 1 located between the pixels 11 sensitive to visible light. This step is illustrated in Figure 4e. Preferably, this formation is done by epitaxy, a superficial surface of the support substrate 11 then serving as a seed layer for the growth of the material sensitive to infrared rays. According to one embodiment, the material forming the pixels 12 is chosen from 11 lV semiconductors, such as alloys of indium phosphide (InP), gallium-indium arsenide (InGaAs) or germanium (Ge). The material forming the pixels 12 can also be formed from colloidal quantum dots such as colloidal quantum dots of lead(ll) sulfide. The substrate 1 being generally made of silicon, this step is preferably a hetero-epitaxial growth step.
[0057] When a protective film has been used to form the cavities 5, the film is removed following the step of forming pixels 12 sensitive to infrared rays. Thus, the protective film 3 also makes it possible to prevent the growth of the material forming the pixels 12 on the pixels 11 sensitive to visible light. The protective film 3 can in particular be removed using an additional etching step.
[0058] The formation of the pixels 12 sensitive to near infrared rays in cavities 5 arranged between the pixels 11 sensitive to visible light makes it possible to have these two types of pixels arranged in the same plane of the sensor. Thus, the formation of the sensor does not require a vertical stack of photosensitive materials configured respectively for the detection of visible light and near infrared radiation. Unlike the sensors of the prior art, neither of the two materials used is therefore buried, and all the pixels are directly exposed to the light rays.
[0059] Thus, the sensors obtained using the proposed method do not exhibit the loss of quantum efficiency observed in state-of-the-art sensors at certain wavelengths.
[0060] Furthermore, since the formation of the pixels 12 sensitive to infrared rays is independent of that of the pixels 11 sensitive to visible light, these pixels 12 are formed in desired dimensions (in particular desired width and thickness), independently of the dimensions of the pixels 11 sensitive to visible light. Each type of pixel is formed in the most suitable material.
[0061] As illustrated in Figure 4f, it is possible to provide, before the step of forming the pixels 12 sensitive to infrared rays, the formation of a buffer layer 6 on the parts of the support substrate 1 located at the bottom of the cavities 5, so as to absorb the dislocations possibly generated by the epitaxial growth process of the pixels 12.
[0062] According to a second embodiment of the method, shown in Figures 5a to 5e, a bulk silicon substrate (called "bulk silicon" in English) is used as the support substrate 1', so as to obtain a sensor as shown in Figure 5e. According to this embodiment, and as illustrated by Figure 5b, cavities 5 are made in the silicon layer 1', preferably by etching using a protective film 3 as described previously. The cavities are made over only a portion of the thickness of the support substrate 1 so as to achieve a desired thickness of the pixels 11' sensitive to visible light. A protective layer 4 is then formed, as shown in Figure 5c. Then, a step of removing the portions of the protective layer 4 extending to the bottom of the cavities may be provided, so as to reveal the support substrate T. This step is shown in Figure 5d.
[0063] For this embodiment, the protective layer 4 finally obtained will only be arranged on the side walls of the pixels 11' sensitive to visible light.
[0064] Finally, the pixels 12' sensitive to infrared rays are formed, in particular by hetero-epitaxial growth on the surface of the support substrate 1' arranged in the cavities 5. As for the previous embodiment, a III-V semiconductor such as those mentioned above, or colloidal quantum dots such as lead(II) sulfide colloidal quantum dots will preferably be chosen as the material sensitive to infrared rays.
[0065] As for the previous embodiment, when a protective film 3 has been used to form the cavities 5, the film is removed following the step of forming the pixels 12' sensitive to infrared rays, in particular using an additional etching step.
[0066] As described previously, a buffer layer 6 may be provided between the support substrate 1' and the material sensitive to infrared rays, so as to absorb any dislocations generated by the epitaxial growth process of the pixels 12'.
[0067] Figure 6 illustrates, in top view, an example of a sensor pixel arrangement, in the form of a Bayer matrix type matrix comprising pixels sensitive to visible light of blue, green or red color (respectively denoted R, G and B for the English "red, green, blue") as well as pixels sensitive to clean infrared rays (denoted SWIR). All the pixels are located on the same plane. The shape and size of the pixels are given for information purposes only and are not representative of the actual size of the two types of pixels.
Claims
Claims 1. A method of manufacturing an image sensor for detecting visible light and short-wave infrared rays, characterized in that it comprises the following steps: a. providing a support substrate (1, 1') comprising a first semiconductor material; b. forming, in the support substrate (1, 1'), cavities (5) in order to define pixels (11, 11') sensitive to visible light in the first semiconductor material between said cavities (5); c. forming an electrically insulating protective layer (4) at least on the lateral surfaces of each pixel (11, 11') sensitive to visible light; d. growing, in the cavities (5), a second semiconductor material, different from the first material, so as to form pixels (12, 12') sensitive to short-wave infrared rays.
2. Method according to the preceding claim, characterized in that it comprises: - between step a and step b, a step a' of forming a protective film (3) of the support substrate (1, 1'), to protect the pixels (11, 11') sensitive to visible light during step b of forming the cavities (5); - after step d, a step d of removing the protective film (3) in order to expose the upper surface of the pixels (11, 11') sensitive to visible light.
3. Method according to any one of the preceding claims, characterized in that step c comprises the formation of the protective layer (4) by thermal oxidation.
4. Method according to any one of the preceding claims, characterized in that the first semiconductor material is silicon.
5. Method according to any one of the preceding claims, characterized in that the second semiconductor material is a III-V semiconductor, preferably chosen from indium phosphide, indium gallium arsenide, germanium alloys, or colloidal quantum dots based on lead(II) sulfide or lead selenide.
6. Method according to any one of the preceding claims, characterized in that the support substrate (1) is of the silicon on insulator (SOI) type, and successively comprises a base substrate (13), an electrically insulating intermediate layer (14) and a monocrystalline layer (15) of the first semiconductor material.
7. Method according to the preceding claim, characterized in that it comprises, between step b and step c, a step b' of thickening the monocrystalline layer (15) by homoepitaxy.
8. Method according to any one of claims 6 and 7, characterized in that step b is an etching step carried out in the support substrate (1) until the electrically insulating intermediate layer (14) is reached.
9. Method according to any one of claims 6 to 8, characterized in that it comprises, after step c, a step c” of removing, in the cavities, the material forming the electrically insulating intermediate layer (14), and in that the pixels (12) sensitive to short-wave infrared rays are formed, in step d, on the face of the base substrate (13) revealed by the removal step c”.
10. Method according to any one of claims 1 to 5, characterized in that the support substrate (1') is a solid substrate of the first semiconductor material.
11. Image sensor for detecting visible light and short-wave infrared rays, comprising: a support substrate (1), a plurality of pixels (11) sensitive to visible light, formed in the support substrate, and pixels (12) sensitive to short-wave infrared rays, said pixels (12) sensitive to infrared rays being arranged on a surface region of the support substrate; a protective layer (4) arranged at least on the side walls of each pixel so as to electrically insulate each pixel from adjacent pixels; wherein the pixels (11, 12) form an arrangement arranged on the same plane, and wherein the support substrate (1) is of the semiconductor on insulator (SOI) type, successively comprising a base substrate (13) in a semiconductor material, an electrically insulating intermediate layer (14) and a monocrystalline layer (15) in a semiconductor material.