Integrated particulate matter sensor with cavity

Abstract

The particulate matter (PM) sensor comprises a substrate containing a semiconductor chip 4, forming a cavity 5, and a light source 1 arranged within the cavity 5. The light source 1 is adapted to emit a light beam 7. The light beam 7 forms a detection volume 8 for particulate matter 9 outside the cavity 5. Optionally, the particulate matter sensor comprises an optical element 2 that bounds the cavity 5 at one end. The optical element 2 is configured to shape the light beam 7. Furthermore, the particulate matter sensor comprises at least one photodetector 3 integrated on a surface of the semiconductor chip 4. The surface on which the at least one photodetector 3 is integrated faces the detection volume 8. The at least one photodetector 3 is adapted to detect light 10 scattered by particulate matter 9 within the detection volume 8.

Description

[Technical field]

[0001] The present invention relates to a particulate matter sensor and a particulate matter sensor module including a particulate matter sensor. [Background technology]

[0002] Particulate matter (PM) refers to solid particles and / or liquid droplets in a fluid. PM may pose a health risk, for example if inhaled, or may cause poor visibility, known as haze. Common categories of PM include PM10 and PM2.5, i.e. particles with diameters less than 10 μm and 2.5 μm, respectively.

[0003] A conventional PM sensor module comprises a light source that emits light into a detection volume and a photodetector that detects light scattered by particulate matter in the detection volume. A conventional PM sensor module is built from separate components, namely a light source assembly with a laser diode, optical elements, a photodetector, a printed circuit board (PCB), separate amplifiers, a microprocessor, and a housing. An airflow for sampling particles is generated using a fan or alternatively a heater element. An example is disclosed in, for example, US Pat. No. 5,399,363.

[0004] Such PM sensor modules are of macroscopic scale, i.e. have dimensions of the order of a few centimeters. The form factor and size of conventional PM sensor modules stems from the discrete nature of the optoelectronic components used, i.e. laser diodes, optics, mounting aids and photodetectors.

[0005] US Patent No. 5,999,336 discloses a PM sensor comprising a light source and a photodetector disposed adjacent to each other within a body portion. The light source emits light toward air introduced into the body portion. An optical lens disposed on the light source focuses the emitted light. The scattered light is detected by the photodetector.

[0006] Patent document 3 discloses a mobile device configured to detect particulate matter. A sensor in the mobile device comprises a light emitter and a light receiver arranged at an angle.

[0007] Patent Document 4 discloses a mobile device for measuring the concentration of particulate matter. A flash of the mobile device emits light. A collecting lens collects the backscattered light, which is filtered and detected by a photodetector.

[0008] The problem to be solved by an embodiment of the present invention is to provide a compact PM sensor that provides particularly reliable and high quality measurements. [Prior art documents] [Patent documents]

[0009] [Patent Document 1] International Publication No. 2018 / 100209 [Patent Document 2] US Patent Application Publication No. 2015 / 0153275 [Patent Document 3] US Patent Application Publication No. 2016 / 0025628 [Patent Document 4] Chinese Patent No. 106483051 Summary of the Invention

[0010] This problem is solved by a particulate matter (PM) sensor according to claim 1. Advantageous embodiments are obtained in the dependent claims.

[0011] therefore, a substrate including a semiconductor chip and defining a cavity (5) at least a portion of which is formed within the semiconductor chip; at least one photodetector integrated on the surface of the semiconductor chip; a light source disposed within the cavity and emitting a light beam toward one end (hereinafter referred to as the "first" end) of the cavity that defines a detection volume for particulate matter outside the cavity; Equipped with a surface of the semiconductor chip on which the at least one photodetector is integrated faces the detection volume, A particulate matter sensor is provided, wherein at least one light detector is adapted to detect light scattered by particulate matter within a detection volume.

[0012] By providing a cavity that is at least partially formed in one and the same semiconductor chip in which the at least one photodetector is integrated and by locating the light source in the cavity, a very compact PM sensor can be obtained.

[0013] The detection volume includes that portion of the light beam where the light intensity is high enough to allow light scattered by PMs in the light beam to be detected by at least one photodetector. In particular, the detection volume can be defined as a volume within which PMs present will result in a clear (i.e., statistically significant) signal above the noise level at the PM sensor. The detection volume therefore depends on various factors such as the size of the PM, the optical power of the light source, and the geometry of the light beam.

[0014] In some embodiments, the semiconductor chip comprises a CMOS stack. One or more layers of the CMOS stack may form a membrane that spans the first end of the cavity. The membrane may have a thickness of less than 20 μm, in particular less than 10 μm. The membrane may thus protect the light source. In particular, the cavity may be completely closed at the first end by the membrane, making the cavity fluid-tight at the first end.

[0015] In some embodiments, the particulate matter sensor may include an optical element that bounds the cavity at the first end, the optical element configured to shape the light beam and form the detection volume. In other embodiments, the optical element may be omitted. For example, the light source itself may be configured to generate a sufficiently collimated or focused light beam such that the light beam has sufficient intensity to form the detection volume outside the cavity. In some embodiments, the optical element includes a membrane formed by one or more layers of a CMOS stack. In other embodiments, the cavity is open at the first end, and the optical element is disposed at the open first end of the cavity.

[0016] The cavity is preferably open at a second end opposite the first end.The light source is preferably disposed within the cavity at the second end of the cavity.

[0017] In some embodiments, the substrate may be formed entirely by the semiconductor chip, i.e., the substrate may consist solely of the semiconductor chip, while in other embodiments, the substrate may include a spacer to which the semiconductor chip is bonded, as described further below.

[0018] When the substrate consists only of a semiconductor chip and when optical elements are present, the particulate matter sensor has the following features: A semiconductor chip forming a cavity; at least one photodetector integrated on the surface of the semiconductor chip; a light source disposed within the cavity and adapted to emit a light beam; an optical element that defines a cavity at one end (the "first" end);

[0033] a light source disposed to direct a light beam onto the optical element; the optical element is configured to shape the light beam such that the light beam forms a detection volume for particulate matter outside the cavity; a surface of the semiconductor chip on which the at least one photodetector is integrated faces the detection volume, The at least one light detector is adapted to detect light scattered by particulate matter within the detection volume.

[0019] Advantageous embodiments of the PM sensor are described below. A PM sensor typically includes the following elements: - a substrate forming the cavity: the substrate comprises or consists of a semiconductor chip, advantageously the semiconductor chip comprises a complementary metal oxide semiconductor (CMOS) stack. The PM sensor function is therefore preferably integrated in the semiconductor chip. The cavity has side walls formed by the substrate. At least a part of each side wall is formed by the semiconductor chip. The cavity can have essentially the shape of a rectangular parallelepiped, for example with side lengths in the range 0.3 mm to 1 mm, or it can have the shape of a cylinder, for example with a diameter in the range 0.3 mm to 1 mm. In other embodiments, the cavity can have the shape of a truncated cone or pyramid. More generally, each side wall of the cavity can have at least one inclined portion inclined with respect to the optical axis or the axis of symmetry of the cavity. At least a part of the cavity can be manufactured, for example, by etching the semiconductor chip, preferably from the bottom side of the substrate, or by alternative processing techniques. In some embodiments, the cavity can reach the entire thickness of the semiconductor chip, while in other embodiments, the cavity can not reach the entire thickness of the semiconductor chip and can take the form of a recess bounded by a membrane formed by the remainder of the semiconductor chip, typically by one or more layers of the CMOS stack. - a light source adapted to emit a light beam: the light source is arranged within the cavity. In an advantageous embodiment, the light source is a laser diode, for example a vertical cavity surface emitting laser (VCSEL). The term "light" is not limited to visible light, but is rather intended to include at least ultraviolet and infrared light. Typically, the wavelength of the emitted light is in the range of 500 nm to 1100 nm, in particular 640 nm to 950 nm. The light source emits a light beam towards a first end of the cavity. The light source can be arranged at a second end of the cavity. - Optionally, an optical element bounding the cavity at a first end: typically the optical element is located at the end of the cavity opposite the light source. The light source is arranged within the cavity so as to direct a light beam through at least a portion of the cavity towards the optical element. The optical element is adapted to shape the light beam and thereby form the detection volume. In an advantageous embodiment, the optical element focuses the light beam, as will be explained in more detail below. - at least one photodetector integrated in the substrate: at least one photodetector is integrated in the surface of the semiconductor chip. The at least one photodetector may comprise at least one photodiode. The at least one photodetector may be formed in the semiconductor chip by a CMOS process. The at least one photodetector faces the detection volume and is adapted to detect light scattered by the PM in the detection volume. In particular, the at least one photodetector is arranged at a distance of up to 2 mm (measured edge to edge) from the optical element, more particularly adjacent to the optical element. Furthermore, as will be explained in more detail below, it is advantageous for the optical element to be arranged within a tolerance of +1 mm / -0.1 mm in the plane defined by the at least one photodetector, more precisely in the plane defined by the surface of the semiconductor chip in which the at least one photodetector is integrated.

[0020] Such PM sensors can be constructed with small form factors, i.e., less than 7 mm×7 mm×2 mm, in particular less than 5 mm×5 mm×1.6 mm, and can be integrated into PM sensor modules or portable electronic devices, such as smartphones or Internet of Things (IoT) devices. Furthermore, such PM sensors have the advantage of low current consumption, which also makes them well suited for integration into battery-powered devices.

[0021] In some advantageous embodiments, the PM sensor further comprises: a control unit electrically connected to the at least one photodetector: the control unit is adapted to receive a signal from the at least one photodetector caused by light scattered by PM in the detection volume. Furthermore, the control unit is adapted to evaluate the signal with respect to a physical quantity related to the PM, i.e. to determine a physical quantity related to the particulate matter based on the signal. In particular, the physical quantity may comprise at least one of the number concentration, the size and the size distribution of the PM. Advantageously, at least a part of the control unit may be integrated in a semiconductor chip. In particular, at least a part of the control unit may be formed in a CMOS stack. More particularly, at least a part of the control unit may be an ASIC formed in the CMOS stack. The control unit may be fully implemented in the semiconductor chip or may be partly implemented separately from the semiconductor chip, for example in a separate signal processor or computing unit.

[0022] Further advantageous technical features will become apparent from the following description, and it is obvious to those skilled in the art that these features can be combined in various ways to form embodiments of the present invention.

[0023] light source The amount of light scattered by the PM in the detection volume and received by the at least one photodetector depends, inter alia, on the optical power of the light source. It is therefore important to quantify the optical power. The following embodiments are particularly advantageous when the light source comprises a VCSEL, because the optical power of a VCSEL is typically not controlled, since the exact optical power is not important for applications such as time-of-flight (TOF) measurements.

[0024] To quantify the light output of the light source, the PM sensor may comprise a photosensitive auxiliary detector arranged to receive the light emitted by the light source and not scattered by the PM. The auxiliary detector may be integrated in a semiconductor chip. The auxiliary detector may be manufactured by the same technology as the at least one photodetector used to detect the light scattered by the PM. The auxiliary detector may in particular be a photodiode, in particular a photodiode manufactured by a CMOS process. The auxiliary detector has a surface area in the photodetector plane that is significantly smaller than the total surface area of ​​the photodetectors used to detect the light scattered by the PM, for example a surface area of ​​1% or less of the surface area of ​​the photodetectors, thereby ensuring that the signal from the auxiliary detector is not significantly affected by the light scattered by the PM and minimizing sensitivity to ambient light.

[0025] If the auxiliary detector is integrated in a semiconductor chip with a CMOS stack, at least three different optical paths are possible between the light source and the auxiliary detector. The first optical path extends through the semiconductor chip. Although light may be strongly attenuated by semiconductors such as silicon, the penetration depth of the light is usually not so small that it can be neglected. If the auxiliary detector is placed in the semiconductor chip close enough to the wall of the cavity (e.g., at a lateral distance of 200 μm or less), a sufficient amount of light can reach the auxiliary detector through the semiconductor chip. The second optical path extends through the CMOS stack, which can act as a light guide. In this way, stray light can be guided laterally to the auxiliary detector. The third optical path, if present, extends through the optical element. In this way, stray light scattered in or at the surface of the optical element can reach the auxiliary detector. Depending on the design, one or more of these optical paths may be enabled.

[0026] Thus, in a first embodiment, the auxiliary detector, in particular the photodiode, is arranged adjacent to the optical element. In this way, the auxiliary detector can receive stray light from the optical element. The auxiliary detector can be arranged in the cavity, for example on the wall of the cavity facing the optical element. By integrating the auxiliary detector into the semiconductor chip, the manufacturing process is simplified, for example because the photosensitive auxiliary detector is formed during a normal CMOS process.

[0027] The auxiliary detector is adapted to measure stray light from the optical element, i.e. light that does not leave the optical element towards the detection volume but is reflected or scattered in other directions, e.g. in the reverse direction. It has been found that the amount of stray light is indicative of the light output of the light source, in particular proportional to the light output of the light source. Therefore, the control unit is further electrically connected to the auxiliary detector and adapted to determine the light output of the light source from the stray light and to evaluate a physical quantity related to PM depending on the determined light output. Alternatively or additionally, the control unit is adapted to control the light source depending on the determined light output.

[0028] In a second embodiment, an auxiliary detector, in particular a photodiode, is arranged in or adjacent to the cavity and is adapted to measure the spontaneous emission of the light source. This may be particularly important when the light source is a VCSEL. It has been found that VCSELs exhibit spontaneous emission of light at one or more side walls, i.e. at one or more walls other than the main emission surface of the VCSEL. Furthermore, it has been found that the amount of spontaneous emission is again indicative of the light output of the light source, in particular proportional to the light output of the light source. Therefore, a control unit is further electrically connected to the photodiode and is adapted to determine the light output of the light source from the measured spontaneous emission and to evaluate a physical quantity related to PM depending on the determined light output. Alternatively or additionally, the control unit is adapted to control the light source depending on the determined light output. Here again, it is advantageous for the photodiode to be integrated into the substrate.

[0029] The described embodiments facilitate more accurate measurement of physical quantities related to PM, particularly when the optical output of a light source is otherwise unknown, such as a VCSEL.

[0030] Optical Elements Generally, the described PM sensor is optimized for a large detection volume since the PM count is directly proportional to the detection volume as defined above. As explained above, a PM particle must generate enough scattered light in the direction of at least one photodetector so that the signal from the scattered light detected by the at least one photodetector exceeds the noise level, e.g. dark current noise. The volume where this condition is met is called the detection volume. The light scattered by PM can be approximately described by Mie theory. To explain the optimization of the detection volume, further approximations can be useful. It is clear that the detection volume is limited by geometrical effects, in particular the diffusion of the light emitted by a point source or scattered by a particle. Due to this diffusion, the light intensity decreases with the distance d from the point source or scattering particle as 1 / d, corresponding to the increase of the surface of the outgoing spherical wave. 2This has implications for the design of PM sensors in general and the optics in particular.

[0031] The optical element generally defines an optical axis, which is preferably perpendicular to the surface of the semiconductor chip in which the at least one photodetector is integrated. In an advantageous embodiment, the optical element focuses the light beam, for example at a focal point or focal region. The intensity of the light beam therefore varies with the distance I from the optical element to the focal point along the optical axis, corresponding to a decrease in the surface of the conical light beam, I 2 It can be seen that this effect of increasing the light intensity in the detection volume counteracts and to some extent balances the diffusion effect of the light scattered by the PM particles as described above. In this way, the detection volume is maximized for a given light source and a given photodetector.

[0032] In particular, the detection volume therefore extends from the optical element to at least the focal point of the light beam. The distance I0 between the optical element and the focal point may be at least 1 mm. Typically, the optimal focal distance depends on the threshold of at least one photodetector that resolves the particle-scattered light to noise, the light output of the light source, and the numerical aperture of the optical element. Depending on the size of the PM particles, the detection volume may extend beyond the focal point, for example, it may be 1.2 or 1.5 times I0 for large particles. In this way, the PM sensor is adapted to detect PM within a range of at least 1.5 mm from the optical element.

[0033] Instead of focusing, the optical element can be adapted to collimate the light beam, i.e. to shape the light beam so that the different rays in the light beam are essentially parallel outside the cavity. In this case, the light intensity theoretically remains constant along the light beam, assuming no scattering and no attenuation. Such a configuration with collimating optics instead of focusing optics also allows to obtain a larger detection volume, e.g. up to 3 cm from the optical element.

[0034] Again, in terms of the maximum size of the detection volume, it is advantageous for the optical element to be located in the same plane as the at least one photodetector (i.e. in the same plane as the surface of the semiconductor chip in which the at least one photodetector is integrated) or very slightly above or below as described above. Thus, the optical element can in one embodiment protrude from the plane of the at least one photodetector or in another embodiment protrude very slightly, for example by a maximum of 0.6 mm. It is also advantageous that the thickness of the optical element perpendicular to the light beam is small, i.e. less than 2 mm, in particular less than 1 mm. In this way, shadowing of the at least one photodetector by the optical element from light scattered by PM particles in the vicinity of the optical element can be prevented. In other words, the detection volume can be increased towards the optical element or optimally up to the optical element.

[0035] At the same time, it is advantageous for the cavity height, i.e. the distance between the light source and the optical element, to be at least 0.25 mm, in particular at least 0.45 mm. This makes the PM sensor more robust against manufacturing errors, such as small deviations from the optimum dimensions. Together with the above considerations regarding shadowing, it follows that the thickness of the optical element should advantageously be small.

[0036] In some embodiments, the substrate is disposed on a base substrate, which may be, for example, a carrier made of glass, semiconductor, ceramic, etc. In such embodiments, the cavity may be bounded at one (first) end by the optical element and at the other (second) end by the base substrate. The light source may be disposed on the base substrate and may emit light in the direction of the optical element.

[0037] Typically, the optical element may form a refractive optical element, in particular a lens, or a diffractive optical element. A refractive optical element shapes a light beam by refraction, whereas a diffractive optical element shapes a light beam by diffraction. These principles may also be combined. In some embodiments, the optical element comprises an imprinted polymer lens or an injection molded lens.

[0038] In some embodiments, the optical element can comprise a glass carrier substrate and an optical structure, in particular a polymer lens, formed on the glass carrier substrate. The optical structure can be formed, for example, by imprinting a UV-curable polymer with a stamp and then UV curing, or can be formed by photolithography.

[0039] In the process of manufacturing the imprinted lens, a polymer lens is formed on a glass carrier substrate. In particular, multiple polymer lenses can be formed on the glass carrier substrate. The glass carrier substrate is then diced to form a single lens unit. The polymer lens is then placed on the cavity together with the glass carrier substrate.

[0040] In one embodiment, the glass carrier substrate has a thickness of less than 1000 μm, such as less than 800 μm, in particular less than 750 μm or less than 600 μm.

[0041] In another embodiment, the optical element comprises a membrane formed by one or more layers of the CMOS stack. The optical structure can be arranged on the membrane forming the optical element. Additionally or alternatively, the membrane itself can have at least one CMOS layer structured to form the optical element. The membrane therefore functions as a diffractive optical element (DOE). In particular, the membrane can function as a metamaterial transparent to the light beam, the metamaterial having a structure that effectively shapes the light beam. To generate such a DOE, the membrane is manufactured from a substrate, for example in the form of a thin layer, and is formed, for example, by etching the substrate from the bottom side almost over the entire thickness of the substrate, leaving the membrane covering the cavity on the front side of the substrate. The structure that shapes the light beam can be manufactured in a prior step during the processing of the CMOS stack or in a subsequent step, for example by structuring the membrane, for example by etching or by applying a structure to the membrane. The thickness of the membrane or metamaterial is advantageously less than 20 μm, in particular less than 10 μm.

[0042] Typically, thin optical elements as proposed in the above embodiments minimize the shadowing of scattered light and therefore make it possible to provide a large detection volume. In other words, thin optical elements facilitate minimizing the distance required between the optical elements and the at least one photodetector while preventing shadowing. Thin optical elements also facilitate reducing the overall form factor of the PM sensor.

[0043] Another aspect related to the optical element concerns stray light that exits the optical element in a direction other than the desired light beam, e.g., sideways, particularly towards the at least one photodetector. If such stray light reaches the at least one photodetector, the noise level increases significantly, thereby reducing the signal-to-noise ratio of the PM sensor and therefore effectively reducing the detection volume.

[0044] In order to prevent stray light from the optical element, particularly in the direction towards the at least one photodetector, the PM sensor advantageously comprises a light barrier between the optical element and the at least one photodetector.

[0045] In some embodiments, the light barrier comprises a blackening or silvering of the sidewall of the optical element facing at least one photodetector. In particular, the blackening or silvering may comprise a selective coating that reacts only with the carrier glass layer and not with the polymer lens mentioned above. A prime example is the application of a mirroring layer by the known silver nitrate process. The term "silvering" is understood as a reflective coating that acts as a light barrier, but is not necessarily made of silver. Other materials that impede the passage of light may also be used.

[0046] In some embodiments, the light barrier comprises a diaphragm formed by a coating on the optical element, the diaphragm defining an aperture for the light beam. The diaphragm can be formed, for example, by a chrome coating on a glass carrier substrate. In particular, in the case of an imprinted polymer lens on a glass carrier substrate, it is advantageous that the diaphragm with the aperture can be located on at least one of the top side or the bottom side of the glass carrier substrate.

[0047] Such a light barrier makes it possible to block stray light from reaching at least one photodetector, while the light beam passes through the optical beam unhindered, and stray light exiting the optical element towards the cavity can be kept largely unaffected, such that the above-mentioned method of quantifying the light output of the light source is still feasible with such an embodiment.

[0048] The present invention also provides an optical element having a light barrier as described herein, whether or not the optical element is integrated into a PM sensor.

[0049] Photodetector The following disclosure relating to one or more photodetectors will be considered to be disclosed in combination with a PM sensor, but will also be considered to be disclosed outside of such PM sensor application, i.e., independent of the PM sensor, but rather as a photodetector device including a photodetector integrated within a semiconductor chip, which may include CMOS metallization and dielectric layers on top of the semiconductor chip.

[0050] The at least one photodetector can be any type of photodetector, but is advantageously a silicon-based photodetector. Such photodetectors can preferably be manufactured in the same process step, e.g., a CMOS process step, as the control unit, represented by the electronic circuitry, which is integrated in the preferred silicon substrate. Such photodetectors are easier to handle during manufacturing and have lower cost than other semiconductor photodetectors. Thus, such photodetectors are well suited to manufacturing a large number of PM sensors, e.g., for IoT devices.

[0051] In an advantageous embodiment, the particulate matter sensor comprises a number of photodetectors integrated on the same surface of the semiconductor chip. The photodetectors can be arranged in an array, i.e. a number of photodetectors can be arranged in a regular pattern. This is useful since it is desirable to have at least one photodetector cover a large area while minimizing the distance to the first end of the cavity, and in particular to the optical element, if present. The photodetectors can be arranged at different locations around the cavity or optical element, preferably on diametrically opposite sides of the cavity or optical element, and more preferably distributed over multiple locations along the periphery of the cavity or optical element. In particular, when the photodetectors are arranged in one or more arrays, the array or arrays can be distributed around the cavity or optical element. As an example, four photodetectors can be distributed in the same plane as the optical element.

[0052] Each photodetector can form a pixel. A photodetector pixel is 1 x 1 mm2 Less than 0.5 x 0.5 mm, preferably 2 less than 0.3×0.3 mm, and even more preferably 2 Preferably, each of the pixels has planar dimensions less than 100 mm. The same square area also applies in the case of non-square shaped pixels, such as circular shaped pixels.

[0053] In this way, the yield of light scattered by the PM within the detection volume and impinging on the photodetector is maximized.

[0054] Here too, the geometrical considerations mentioned above apply. It is advantageous to have the at least one photodetector as close as possible to the cavity or optical element. In this way, the optical path length from the scattering particles in the detection volume to the at least one photodetector is minimized, and therefore the signal-to-noise ratio is maximized.

[0055] Optionally, the PM sensor comprises an optical filter on the at least one photodetector. This means that the optical filter covers the surface of the at least one photodetector opposite to the substrate. The optical filter can be arranged on the surface of the semiconductor chip on which the at least one photodetector is integrated. The optical filter advantageously filters out light and radiation outside the main wavelength range of the light source. In this way, background rejection is achieved since spurious light or radiation emissions do not reach the at least one photodetector. The optical filter can be an interference filter with multiple layers with different refractive indices, causing destructive interference outside the desired wavelength range.

[0056] In an advantageous embodiment, the photodetectors are separated by a conductive material, for example having the shape of a grid, where the photodetectors are arranged in the voids of the grid, for example in the form of tiles or pixels, as already mentioned above. In particular, the photodetectors can be separated by a metallization of the substrate. In this way, the manufacture of the conductive material can be integrated into the normal processing of the semiconductor chip, the metallization of the top of the CMOS stack being manufactured in such a way that this metallization, in particular its compartment borders, serves as a conductive material separating the photodetectors. Such a conductive material between the photodetectors can be grounded, i.e. functioning as a Faraday cage and exposed towards the measurement volume. To enable the grounding of the conductive material, the conductive material can be connected to a ground connector of the sensor device. In particular, the conductive material is adapted to protect the photodetectors from electromagnetic interference, for example with other electronic devices in the surrounding environment of the PM sensor. The above range of pixel dimensions therefore defines the distance between the metallizations, facilitating shielding from electromagnetic interference.

[0057] To further reduce the electromagnetic interference, the at least one photodetector is advantageously partitioned into a first section facing the detection volume and a second section shielded from light scattered by PM in the detection volume. For example, the at least one photodetector in the second section can be covered by an opaque layer that is opaque at least in a wavelength range including the main wavelength of the light source. The opaque layer is preferably electrically insulating so as to ensure that both sections are exposed to the same level of electromagnetic interference. For example, the opaque layer can be formed by inkjet printing. The two separate sections can be used to detect and cancel signals in the photodetector that are due only to undesired electromagnetic interference and not due to light scattered by PM in the detection volume. For this purpose, the control unit is adapted to perform a differential measurement of the first section and the second section. Thereby, in particular spurious effects of electromagnetic interference with the first and second sections of the at least one photodetector are canceled.

[0058] Further Aspects As already mentioned above, the substrate can be provided with a spacer. The semiconductor chip can be bonded to the spacer, in particular on the back side of the semiconductor chip, which does not face the surface on which the photodetector is integrated. The cavity can be formed both in the spacer and in the semiconductor chip. By using the spacer, the distance between the light source and the optical element can be increased. Increasing the distance between the light source and the optical element makes it possible to use optical elements with a larger focal length. This can have several beneficial effects, in particular on the size of the detection volume and on the sensitivity to manufacturing tolerances.

[0059] As already mentioned, the PM sensor may comprise a base substrate. The light source may be mounted on the base substrate. The substrate may also be arranged on the base substrate such that the light source is arranged in the cavity. If the substrate consists of a semiconductor chip, the semiconductor chip may be directly connected to the base substrate. If the substrate comprises a spacer, the spacer may be arranged between the base substrate and the semiconductor chip. The base substrate preferably extends in a plane parallel to the surface of the semiconductor chip on which the photodetector is integrated. The base substrate may form or include a land grid array.

[0060] An opaque coating can be applied to the sidewalls of the cavity to reduce the amount of light from the light source reaching the at least one photodetector through the sidewalls of the cavity. Similarly, an opaque coating can be applied to the backside of the substrate or the semiconductor chip, which does not face the surface on which the at least one photodetector is integrated, to reduce the effect of ambient light. If the substrate comprises a semiconductor chip and a spacer, the opaque coating can be applied to the backside of the spacer, the backside of the semiconductor chip, or both. The opaque coating can include metallization and / or a coating applied by an inkjet process.

[0061] The cavity may have an axis of symmetry. In particular, the cavity may have a discrete or continuous rotational symmetry around the axis of symmetry. The axis of symmetry is preferably perpendicular to the surface of the semiconductor chip in which the photodetector is integrated. The axis of symmetry is preferably parallel to an optical axis defined by the optical element. The axis of symmetry may coincide with the optical axis.

[0062] The PM sensor may further comprise a light blocking element, the light blocking element being arranged on the surface of the semiconductor chip in which the light detectors are integrated, the light blocking element selectively blocking a part of one or more of the light detectors from light scattered by the particle in the detection volume, depending on the distance of the particle of particulate matter from the surface of the semiconductor chip in which the at least one light detector is integrated, while one or more other light detectors are not blocked by the light blocking element. The light blocking element may be formed by an asymmetric extension of the optical element. The control unit may be configured to determine a measure of the distance of the particle from the surface of the semiconductor chip in which the at least one light detector is integrated by comparing signals from the light detectors that are partially blocked by the light blocking element with signals from the light detectors that are not blocked by the light blocking element. The control unit may further be configured to take the determined distance into account when determining the physical quantity related to the particulate matter. In particular, a size parameter of the PM may be determined more reliably by taking said distance into account.

[0063] To mechanically protect the substrate, the PM sensor may include an enclosure that laterally surrounds the substrate, the enclosure being made from a molding compound.

[0064] According to a further aspect of the invention, a PM sensor module comprises a housing and a flow channel disposed within the housing. The PM sensor module further comprises a fan or heater disposed within the housing and adapted to move air through the flow channel, and a PM sensor as described in any of the above embodiments or one of the following embodiments. The PM sensor is disposed within the housing such that a portion of the flow channel coincides with the detection volume.

[0065] According to another aspect, the present invention provides a method for determining a physical quantity of particulate matter using a particulate matter sensor as described herein, the method comprising: operating a light source to emit a light beam; operatively operating at least one photodetector to detect light scattered by particulate matter intersecting the light beam; analyzing the signal from the at least one optical detector to determine at least one parameter indicative of a physical quantity of the particulate matter; Includes.

[0066] Determining the at least one parameter may include determining the light output of the light source, as described above, and / or determining the distance of the particle from a surface of the semiconductor chip in which the at least one photodetector is integrated, and / or performing a differential measurement of the signal from the shielded and unshielded areas.

[0067] According to another aspect, the present invention provides a method of manufacturing a particulate matter sensor as described herein, comprising the steps of: a) forming at least one photodetector on a surface of a semiconductor chip; b) etching the semiconductor chip along a direction perpendicular to said surface to form at least a portion of a cavity; c) optionally bonding the semiconductor chip to a spacer forming another part of the cavity; d) disposing a light source within the cavity, the light source being configured to emit a beam of light toward a first end of the cavity; e) optionally providing an optical element on the semiconductor chip, the optical element defining a cavity at a first end and configured to shape the light beam; The present invention provides a method comprising:

[0068] Step b) usually occurs after step a), but can occur before step a). Step c), if present, usually occurs after steps a) and b). Step d) usually occurs after steps a) and b), but if present, occurs after step c). Step e) can occur simultaneously with steps a) and b) or after any of steps b) to d) if the DOE is integrated into a film formed by a layer of the CMOS stack.

[0069] The method may include further steps, such as forming at least a part of a control unit, i.e., an ASIC, in a semiconductor chip as described herein, arranging an optical filter on the semiconductor chip as described herein, applying a coating to the cavity walls and / or the backside of the semiconductor chip and / or spacer as described herein, forming an optical element by any one of the methods described herein, arranging a light source and a substrate on a base substrate, forming wire bonds between the ASIC and the base substrate and / or between the light source and the base substrate, enclosing the substrate in an enclosure made from a molding material, and integrating a PM sensor in a PM sensor module as described herein.

[0070] Further advantageous embodiments are recited in the dependent claims and in the following description.

[0071] The invention will be better understood, and objects other than those mentioned above will become apparent, from the following detailed description, which is to be read in conjunction with the accompanying drawings, in which: [Brief description of the drawings]

[0072] [Figure 1] FIG. 1 is a schematic longitudinal section through a PM sensor according to one embodiment of the present invention. [Diagram 2] FIG. 2 is a perspective view of the PM sensor of FIG. [Diagram 3] FIG. 3 is a diagram showing an embodiment of the PM sensor of FIG. [Figure 4] FIG. 4 is a diagram showing an embodiment of the PM sensor of FIG. [Diagram 5] FIG. 5 is a diagram showing an embodiment of the PM sensor of FIG. [Figure 6] FIG. 6 is a diagram showing an embodiment of the PM sensor of FIG. [Figure 7] FIG. 7 is a schematic longitudinal section through a PM sensor comprising different optical elements according to embodiments of the present invention. [Figure 8] FIG. 8 is a schematic longitudinal section through a PM sensor comprising different optical elements according to embodiments of the present invention. [Figure 9] FIG. 9 is a schematic longitudinal section through a PM sensor comprising different optical elements according to embodiments of the present invention. [Figure 10] FIG. 10 is a schematic longitudinal section through a PM sensor according to another embodiment of the present invention. [Figure 11A] FIG. 11A is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the invention, showing the PM sensor in more detail than FIGS. 1-10. [Figure 11B] FIG. 11B is a functional schematic diagram of the PM sensor of FIG. 11A. [Figure 12] FIG. 11 is a schematic longitudinal section through a PM sensor according to yet another embodiment of the present invention. [Figure 13A] FIG. 13A is a schematic diagram showing the signal processing of the PM sensor according to FIG. [Figure 13B] FIG. 13B is a schematic diagram showing the signal processing of the PM sensor according to FIG. [Figure 14]FIG. 14 is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the present invention. [Figure 15] FIG. 15 is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the present invention. [Figure 16] FIG. 16 is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the present invention. [Figure 17] FIG. 17 is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the present invention. [Figure 18] FIG. 18 is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the present invention. [Figure 19] FIG. 19 is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the present invention. [Figure 20] FIG. 20 is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the present invention. [Figure 21] FIG. 21 is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the present invention. [Figure 22] FIG. 22 is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the present invention. [Figure 23] FIG. 23 is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the present invention. [Figure 24] FIG. 24 is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the present invention. [Diagram 25] FIG. 25 is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the present invention. [Figure 26] FIG. 26 is a schematic longitudinal cross-section through a PM sensor according to an embodiment of the present invention. [Figure 27] FIG. 27 is a perspective view of a PM sensor having two sections of a photodetector. [Figure 28] FIG. 28 is a schematic longitudinal cross-sectional view passing through the sensor module according to the first embodiment. [Figure 29] FIG. 29 is a schematic longitudinal cross-sectional view passing through the sensor module according to the second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0073] Throughout this specification and the claims, the terms "in particular," "preferably," and "optionally" are to be understood as expressing that the object of their corresponding use is optional.

[0074] Overall configuration of PM sensor (Fig. 1 and Fig. 2) FIG. 1 shows a schematic cross section through a PM sensor according to an embodiment, and FIG. 2 shows a perspective view of the PM sensor. A cavity 5 is formed in a substrate, in this example formed by a semiconductor chip 4, on a base substrate 6. Alternatively, the base substrate 6 can also be part of the semiconductor chip 4. The cavity is bounded by side walls 26 formed by the substrate. The cavity 5 is preferably formed from the bottom side (back side) of the substrate and can therefore also show sloping side walls, as indicated by dashed lines. A light source 1 is arranged in the cavity 5 at its bottom end, i.e. at the end facing the base substrate 6. An example of a light source 1 is a laser diode, in particular a vertical cavity surface emitting laser (VCSEL). At the top end of the cavity 5, i.e. the end opposite to the bottom end, an optional optical element 2 is arranged, thus closing the cavity 5. The optical element 2 defines an optical axis 20. Furthermore, photodetectors 3 , for example photodiodes, are integrated in the semiconductor chip 4 on two or more sides of the optical element 2 .

[0075] As shown in FIG. 1, the photodetector 3 is integrated on the top surface of the semiconductor chip 4, not facing the base substrate 6. This surface defines a plane, hereafter referred to as the "photodetector plane". The photodetector plane extends perpendicular to the optical axis 20. In the embodiment of FIG. 1, the optical element 2 is essentially located in the photodetector plane. In particular, the optical element 2 shall not protrude more than 1 mm above the photodetector plane. The reasons for this have been discussed above and are illustrated in FIG. 9. The protruding lens 15 provides shadowing such that scattered light pulses 10 scattered by PM particles much closer to the lens 15 than the particles 9 do not reach the photodetector 3 and are therefore not detected.

[0076] As shown in FIG. 2, the top surface of the semiconductor chip 4 may include an array of photodetectors 3, for example four arrays of photodetector pixels. Metallization 18 is provided between or around the individual photodetector pixels. The metallization 18 may be made of any conductive material. Advantageously, the metallization 18 is formed during normal processing of the CMOS stack of the semiconductor chip 4 by exposing one of the metal layers at the surface. When grounded, the metallization 18 acts as a Faraday cage, shielding the photodetectors 3 from electromagnetic interference and thus from spurious signals. In particular, some of the metallization layers of the CMOS stack may form connections for the readout of the photodetectors, while at least one of the metallization layers (preferably the top layer) may be grounded to act as a Faraday cage. Ground contacts may be formed on the semiconductor chip to connect the corresponding layers to ground. The photodetector pixels 3 may have a size of 1×1 mm 2 Less than 0.5 x 0.5 mm, preferably 2 less than 0.3×0.3 mm, and even more preferably 2 Preferably, each of the pixels has planar dimensions less than 100 mm. The same square area also applies in the case of non-square shaped pixels, such as circular shaped pixels.

[0077] The optical element 2 is not shown in Figure 2. Indeed, in some embodiments, the optical element 2 can be omitted, for example if the light source 1 itself already produces a light beam with a sufficiently small divergence.

[0078] PM sensor operation (Fig. 3 to 6) Figures 3 to 6 show various embodiments of the PM sensor of Figure 1. When the light source 1 is switched on, it emits light towards the optical element 2, which shapes the light beam 7 and in particular focuses it to a focal point 21. In the measurement configuration, as shown in Figure 4, a PM particle 9 approaches the light beam 7. This situation can be obtained, for example, by placing the PM sensor on the wall of the flow channel. Air with PM is blown through the flow channel by a fan or alternatively by a heater (see the discussion of Figures 28 and 29 below).

[0079] Figures 4 to 6 show a portion of the light beam 7 with different shading, which defines a detection volume 8, in which a PM particle 9 present will generate a scattered light pulse 10 that is large enough to be detected by at least one of the photodetectors 3, i.e. the resulting signal at the photodetectors 3 will be above the noise level, e.g. dark current noise.

[0080] The proposed configuration with the focusing optics 2 and the photodetector 3 in the same plane has the advantage that the detection volume 8 reaches at least as far as the focal point 21 of the optics 2. The geometrical reasons for this have been discussed above. For particularly large PM particles 9, the detection volume 8 can also extend beyond the focal point 21, i.e. the scattering particle height 11 may be larger than the focal length of the optics 2, and the particles will still be detected.

[0081] Typically, the size and shape of the detection volume 8 can be optimized or tailored for a particular application by varying one or more of the following parameters: the light output of the light source 1, the focal length of the optical element 2, the distance between the light source 1 and the optical element 2, the distance between the optical element 2 and the photodetector 3, the sensitivity of the photodetector 3, electromagnetic shielding to reduce noise levels, etc.

[0082] Optical Element Design Figures 7 to 9 show an embodiment of a PM sensor similar to Figure 1 but with a different optical element 2. In Figure 7, the optical element is a conventional optical lens 12, for example made of glass or polymer. Depending on the optical index of the lens material, the conventional optical lens 12 must have a certain thickness to exhibit a desired focal length according to the law of refraction.

[0083] An alternative optical element is shown in Figure 8. For example, a diffractive optical element (DOE) 13 disposed on the membrane 22 can be configured with a smaller thickness for the same focal length. The membrane 22 of the DOE 13 can be exposed from the CMOS stack of the semiconductor chip 4 during fabrication. In certain embodiments, the DOE 13 can be a metamaterial, for example, with the membrane surface configured to effectively function as an optical lens.

[0084] 9 shows yet another optical element. A lens is placed on a carrier substrate, for example a polymer lens 15 is placed on a glass carrier substrate 23 as described above. Such a lens 15 on the glass carrier substrate 23 can be manufactured as an imprinted polymer lens on the glass carrier substrate 23 and then attached on the semiconductor chip 4, for example by using an adhesive. The thermal expansion coefficients of the glass carrier substrate 23 and the semiconductor chip 4 are matched or at least similar, which reduces the stress on the adhesive during temperature cycling.

[0085] 9 further illustrates the geometrical considerations related to the spreading of the spherical wavefront of the scattered light, as explained above. Advantageously, the distance 14 of the photodetector 3 from the optical axis 20, and hence from the optical elements, is minimized. This allows for a large detection volume 8, or in other words a large number of PM counts.

[0086] Figures 9 and 10 show another advantageous feature of an embodiment of the invention. The sidewall of the optical element 2, in this example the sidewall of the glass carrier substrate 23 carrying the lens 15, is provided with a light barrier 19, e.g. a blackening or silvering, to prevent stray light of the optical element from reaching the photodetector 3 in a direct path, thereby significantly increasing the noise level and reducing the ability of the PM sensor to detect PM particles, i.e. significantly reducing the detection volume 8. The blackening or silvering can be obtained by applying a selective coating to the glass carrier substrate 23, e.g. a chemical that does not bond to the polymer lens 15 but to the glass of the glass carrier substrate 23, blackening / silvering the glass. In addition, such a coating is thin by its own nature and therefore does not add material to the lateral thickness of the optical element and therefore does not exacerbate the shadowing mentioned above.

[0087] FIG. 10 shows a schematic cross section through a PM sensor according to another embodiment. This PM sensor shares most of the features with the PM sensor of FIG. 1. However, this PM sensor has a photodetector 3 only on one side of the optical element, which in the illustrated embodiment is also a lens 15 on a carrier substrate 23. On the other side, the optical element is supported by a support 16, which can be, for example, a molded frame formed by a substrate or a dummy substrate spacer. It is clear that the PM sensor of FIG. 10 can have a smaller signal-to-noise ratio than the PM sensors of FIGS. 1 to 9. However, this PM sensor can be built with a much smaller form factor, making it well suited for miniaturization applications.

[0088] Details of the PM sensor (Figure 11A) FIG. 11A shows an embodiment of a PM sensor similar to the embodiment of FIG. 9, but shown in more detail than in FIGS. 1-10.

[0089] In this embodiment, the semiconductor chip 4 is a silicon chip carrying a CMOS stack 24. The photodetectors 3 are formed in the semiconductor material by a CMOS process. For example, each photodetector 3 can be a photodiode formed by creating an n-type doped well in a p-type doped part of the silicon chip. The CMOS stack above the photodiode is removed by etching so that light can reach this photodiode. The anode and cathode of the photodiode are connected to the metallization layers of the CMOS stack 24.

[0090] The analog-digital electronic circuitry is formed in the CMOS stack 24. This electronic circuitry forms an application specific integrated circuit (ASIC), which functions inter alia as a control unit 27, as will be explained below with reference to FIG.

[0091] A photosensitive auxiliary detector 25 is formed in the semiconductor chip 4, which determines the light output of the light source 1. This detector can also be called a "feedback detector" since it can provide feedback to regulate the output of the light source 1 in a closed loop. The auxiliary detector 25 can be of the same type as the main photodetector 3. The auxiliary detector 25 is preferably a photodiode. The auxiliary detector 25 can be formed in the semiconductor chip 4 in the same way as the main photodetector 3. The auxiliary detector 25 is preferably located very close to the surface of the sidewall 26 that bounds the cavity 5. The auxiliary detector preferably has a very small surface area compared to the total surface area of ​​the main photodetector 3. The auxiliary detector can for example cover a surface area of ​​less than 100 μm×100 μm, for example 50 μm×50 μm. In this way, the auxiliary detector 25 is relatively insensitive to light scattered by PM in the detection volume 8. On the other hand, it is highly sensitive to light that can reach the auxiliary detector 25 through other optical paths.

[0092] The light received by the auxiliary detector 25 can have at least two different origins. On the one hand, the auxiliary detector 25 can receive light emitted by the light source 1 generally along the main emission direction, i.e. generally along the direction of the light beam 7, and scattered from this direction before the light leaves the optical element. Such scattering can occur, for example, at the surface of the optical element or within the optical element. In the present disclosure, this type of light is referred to as "stray light". On the other hand, the auxiliary detector 25 can receive light emitted by the light source along a direction other than the main emission direction. For example, if the light source is a VCSEL, the light source has a main emission surface from which the light beam 7 is emitted and has lateral sidewalls. The main emission surface faces the direction of the light beam. VCSELs usually generate some light also on the lateral sidewalls by spontaneous emission. This light can also propagate to the auxiliary detector 25.

[0093] In the example of FIG. 11A, at least three optical paths from the light source 1 to the auxiliary detector 25 are possible. The first optical path extends through the semiconductor chip 4. The penetration depth of the light into silicon depends on the wavelength. For a typical IR wavelength of 940 nm, the light intensity decreases by about 10% (1 / e 2 ) This light path is particularly important for light generated by spontaneous emission. Therefore, if it is desired to monitor the light output of the light source 1 by monitoring the intensity of the light generated by spontaneous emission, the auxiliary detector 25 should be located as close as possible to the surface of the sidewall 26 that bounds the cavity 5, which surface should not be covered by an opaque material. In addition, to prevent light from entering the auxiliary detector 25 from above, the top surface of the auxiliary detector 25 can be covered by at least some of the layers of the CMOS stack. - the second optical path extends through the CMOS stack 24. The CMOS stack can act as a light guide perpendicular to the optical axis of the optical element, i.e. in the photodetector plane or parallel to the photodetector plane. Light entering the CMOS stack is thus guided to the auxiliary detector 25. In some embodiments, similar to the embodiment of FIG. 11A, stray light and / or light due to spontaneous emission can enter the CMOS stack at the cavity wall 26. However, in practice this mechanism can often be neglected. In other embodiments, the membrane 22 is formed by one or more layers of the CMOS stack, similar to the embodiment of FIG. 8, and stray light can arise due to scattering in the membrane 22 and / or at its boundaries, so that it can directly enter the CMOS stack. This optical path is valid even if there is an opaque layer on the surface of the sidewall 26 of the cavity 5. - A third optical path extends through the optical element. In the example of Fig. 11A, the optical element is a polymer imprinted lens 15 on a glass carrier substrate 23. The glass carrier substrate 23 is capable of directing stray light to a part of its bottom surface that is located directly above the auxiliary detector 25. Light leaving this part of the glass carrier substrate 23 can hit the auxiliary detector 25 from above. Similar considerations may apply to other types of optical elements. This optical path is also valid in the case of the presence of an opaque layer on the surface of the side wall 26 of the cavity 5.

[0094] Depending on the arrangement of the auxiliary detector 25 relative to the cavity 5 (in particular the lateral distance of the auxiliary detector 25 from the cavity walls), depending on the presence, type and arrangement of optical elements (e.g. whether an optical element is present or not, whether the optical element comprises a film made from the CMOS stack, whether there is a direct light path from the optical element to the top of the auxiliary detector, whether the auxiliary detector is covered by one or more opaque layers of the CMOS stack), as well as further measures such as adding opaque layers to the surface of the cavity side walls, the light received by the auxiliary detector 25 may be dominated by either stray light or light generated by spontaneous emission. Thus, the control unit 27 can determine the light output of the light source from stray light, from spontaneous emission or both.

[0095] An optical filter 28 is placed on the top surface of the semiconductor chip 4 and covers both the photodetector 3 and the auxiliary detector 25. The optical filter 28 is an optical bandpass filter that allows only light within a wavelength range that includes the wavelength of the light source 1 to pass. The optical filter 28 is an interference filter applied using a wafer level process in which multiple layers with different refractive indices are stacked on top of each other. The thickness of each layer can be in the range of approximately a quarter wavelength of the main wavelength of the light source. This leads to destructive interference for all wavelengths except the desired wavelength range. In the case of the main photodetector 3, the optical filter 28 helps to avoid DC saturation and increased noise levels due to ambient light, thereby improving the signal-to-noise level of the signal pulses resulting from the PM. In the case of the auxiliary detector 25, the optical filter 28 reduces the contribution of ambient light to the output signal, making the auxiliary detector 25 relatively less sensitive to ambient light.

[0096] Bonding and packaging of the PM sensor can be done as follows: The base substrate 6 can be, for example, a Land Grid Array (LGA). The light source 1 and the controller 27 can be connected to the lands of the LGA by wire bonds 31. During the manufacture of the PM sensor, the light source 1 can be first attached to the LGA and wire bonded to the appropriate lands of the LGA. In some embodiments, the LGA with the light source bonded can be provided as a pre-assembled unit. For example, a VCSEL may be provided in the form of a pre-assembled unit on the LGA. Then, the semiconductor chip 4 with the cavity 5 and the CMOS stack 24 can be attached (e.g., glued to the LGA) to the LGA such that the light source 1 is located in the cavity 5. The circuitry in the CMOS stack 24 can then also be wire bonded to the LGA. The resulting assembly can then be partially enclosed in an enclosure 32 by open cavity molding, while leaving the photodetector 3, the auxiliary detector 25, and the cavity 5 accessible from above. In some embodiments, the optical element is finally attached to the semiconductor chip 4 to cover the cavity 5. In other embodiments, the optical elements may be formed or attached to the semiconductor chip in a prior manufacturing step.

[0097] In the embodiment of FIG. 11A, as in some other embodiments, the optical element is a polymer lens 15 on a glass carrier substrate 23. In particular, the lens may be a "wafer level optics lens", or "WLO lens" for short. In wafer level optics, a carrier substrate is provided in the form of a wafer, the optical structures are formed on the wafer, and the wafer is then diced. In particular, a polymer lens may be formed on the wafer by coating the wafer with a UV-curable polymer, imprinting the uncured polymer with a wafer-sized stamp, and UV-curing the polymer. In this disclosure, a polymer lens formed in this manner is referred to as an imprinted polymer lens.

[0098] The side walls of the glass carrier substrate 23 can be provided with a light barrier 19, i.e. an opaque coating, to prevent stray light from reaching the photodetector 3. The light barrier 19 can be a mirror-like silvering that can be applied to the side walls of the glass carrier substrate 23 using a known silver nitrate mirroring process. For this purpose, after the polymer lenses 15 are formed on the glass wafer forming the glass carrier substrate, the wafer is mounted on a dicing foil and diced. The diced wafer is then treated by the silver nitrate mirroring process. The mirror is formed only on the diced side walls of the glass carrier substrate 23, since the back side of the wafer is protected by the dicing foil and the polymer forming the lenses does not react with the chemicals.

[0099] To further reduce the risk of stray light problems, a further coating 29 can be applied to the top and / or bottom surface of the glass carrier substrate 23, forming a diaphragm defining an aperture. The coating 29 can for example be a chrome coating. In the example of Fig. 11A, a chrome coating is applied to the top surface of the glass carrier substrate 23. This coating forms a diaphragm defining an aperture, within which the polymer lens 15 is located. Instead of or in addition to forming a diaphragm on top of the glass carrier substrate, a diaphragm defining an aperture can also be formed on the bottom of the glass carrier substrate.

[0100] Although the principles of wafer level optics have been illustrated using the example of a glass wafer forming a glass carrier substrate, the carrier substrate can be formed from materials other than glass.

[0101] Control unit (Fig. 11B) FIG. 11B is a functional schematic diagram of the PM sensor of FIG. 11A. A control unit 27 receives signals from both the photodetector 3 and the photosensitive auxiliary detector 25. The control unit 27 processes the signal from the photodetector 3 so as to detect signal pulses corresponding to light pulses originating from PM within the detection volume 8 of the light beam 7. The control unit 27 further analyzes these signal pulses to derive at least one parameter indicative of a physical quantity of PM, such as a parameter indicative of PM concentration, at least one PM size parameter (e.g. at least one parameter characterizing the average size and / or size distribution), and / or at least one PM velocity parameter. For example, the PM concentration parameter can be determined based on the number of pulses per unit time and a known, measured or estimated flow rate of the fluid flow passing the PM sensor, as known per se. The PM size parameter can be determined based on the amplitude of the pulses, as also known per se. When calculating the PM parameters, the control unit 27 can take into account the light output of the light source 1, as represented by the signal from the auxiliary detector 25. The control unit 27 can further use the signal from the auxiliary detector 25 to control the optical output of the light source 1 by means of a closed loop control algorithm. As will be explained in more detail in relation to Figure 12, the control unit can also take into account the distance of the PM particles from the photodetector plane.

[0102] In summary, the control unit 27 has two main objectives: a) to process the signal from the photodetector 3 to derive at least one parameter characteristic of the PM, and b) to monitor and optionally control the output of the light source 1 using the auxiliary detector 25.

[0103] In some embodiments, the control unit 27 can be implemented completely "on-chip" in an ASIC formed by the CMOS stack 24. In other embodiments, some of the functions of the control unit 27 can be implemented in said ASIC, and other functions can be implemented "off-chip" in external circuitry. The external circuitry can be connected to the ASIC, for example, via the base substrate 6. Some initial processing steps of the signals received from the main photodetector 3 and / or the auxiliary detector 25, such as signal amplification, analog-to-digital conversion, and filtering, can be performed on-chip by the ASIC formed by the CMOS stack 24, while subsequent processing steps, such as computation steps to calculate parameters characteristic of the PM and / or to calculate control signals to control the light source 1, can be performed off-chip by the external circuitry. The external circuitry can include a general-purpose or special-purpose processor configured to execute a computer program that causes the processor to perform one or more processing steps to determine said parameters.

[0104] Distance determination (Figure 12, Figure 13A, Figure 13B) 12 shows an embodiment of a PM sensor that makes it possible to determine the distance from the photodetector plane at which a detected particle intersects the light beam 7. For this purpose, the optical element can be complemented by an asymmetric extension 41.

[0105] The asymmetric extension 41 selectively extends laterally towards one or more of the photodetectors, partially shading these photodetectors but not the others. In the example of FIG. 12, the partially shading photodetector is shown as photodetector 3b, while the unshading photodetector is shown as photodetector 3a. The asymmetric extension 41 shades the photodetector 3b, which is affected by a portion of the light of a PM particle that intersects the light beam 7 near the optical element. This is illustrated in FIG. 12 using the example of two PM particles 9, 9' passing through the PM sensor at different distances from the photodetector plane. Particle 9 passes through the PM sensor at a relatively large distance. The asymmetric extension 41 does not prevent the light scattered by this particle from reaching the photodetector 3b. Thus, photodetectors 3a and 3b receive the same amount of scattered light. In contrast, particle 9' passes through the PM sensor at a relatively small distance from the photodetector plane. Because the asymmetric extension 41 blocks some of the light scattered from the particle 9', the photodetectors 3a and 3b receive different amounts of scattered light.

[0106] FIG. 13A shows diagrammatically the signal pulses recorded by photodetectors 3a and 3b due to scattered light received from particles 9 and 9', respectively. At time t1, photodetectors 3a and 3b receive scattered light from particle 9. The resulting signal pulses have approximately the same amplitude. At time t2, photodetectors 3a and 3b receive scattered light from particle 9'. The resulting signal pulse from photodetector 3a is much larger than the pulse from photodetector 3b.

[0107] 13B shows the resulting ratio of the signals from photodetectors 3a and 3b, which is a direct measure of the distance of the particle from the photodetector plane as it traverses the light beam 7. In particular, this ratio exhibits the following behavior: a) The closer this ratio is to 1 (equal signal levels), the farther the particle is from the photodetector plane as it crosses the light beam. b) The closer this ratio is to 0 (no light reaches the photodetector 3b facing the asymmetric extension due to shadowing), the closer the particle is from the photodetector plane as it crosses the light beam. c) Ratios between 0 and 1 correspond to different distances of the particle from the photodetector plane.

[0108] This information can be used by the control unit 27 to compensate for undesirable effects that said distance may have on the signal level. For example, if the intensity distribution of the light beam 7 along the optical axis is known, the control unit 27 can correct the measured pulse amplitude against the known intensity distribution. As a result, a better estimate of the particle's size can be obtained. In general, better sensor performance can be achieved.

[0109] In FIG. 12, the effect of the asymmetric extension 41 of the optical element is illustrated using the example of a polymer lens 15 on a glass carrier substrate 23, however, the same concept can be employed for other types of optical elements.

[0110] More generally, the asymmetric extension 41 is an example of a light blocking element disposed on the semiconductor chip 4 to selectively block one or more portions of the light detectors from light scattered by PM particles in the light beam 7, depending on the distance of the particle from the light detector plane. The light blocking element may be separate from the optical element. The light blocking element may be disposed laterally adjacent to the optical element. The light blocking element may also be provided when no optical element is present at all.

[0111] Opaque coating on the sidewalls of the cavity (Figures 14 and 15) Forming a "light blocker", i.e. a coating layer (opaque to the main emission wavelength of the light source 1) on the side walls of the cavity and / or on the bottom side of the semiconductor chip 4 facing the base substrate 6, prevents direct light and / or stray light from the light source 1 from passing through the semiconductor chip to the photodetector. In this way, saturation of the detector can be prevented and / or the (Schottky) noise can be reduced. A lower noise level means that a lower threshold can be chosen for PM detection, which results in improved performance. In particular, smaller particles can be detected. The statistics of the data evaluation are increased, which results in better accuracy.

[0112] In the embodiment of FIG. 14, such a coating layer is formed by backside metallization 51 of the semiconductor chip 4. Such metallization can be formed, for example, by sputter deposition. If the deposition process is performed after etching the cavity into the semiconductor chip, the cavity sidewalls are also automatically covered by the metallization 51. Examples of suitable materials for the metallization include, but are not limited to, Al, Cu, Ag, Ti, and TiN. Practical metallization thicknesses range from 50 nm up to 1 μm or more.

[0113] However, such metallization processes are not suitable for product designs that include a membrane formed by one or more CMOS layers, such as in the example of Figure 8, because the metallization would make the membrane opaque, and therefore other processes that form an opaque coating on the chip surface should be used for such embodiments.

[0114] An embodiment in which the opaque coating 52 is formed by an alternative process is shown in FIG. 15. In this embodiment, the membrane 22 formed by one or more layers of the CMOS stack 24 spans the cavity 5. The membrane can be part of the optical element, as described in more detail below, or it can simply be provided to protect the light source from contamination. The opaque coating 52 can be formed by wafer-level inkjet printing into the cavity 5. This process is very cost-effective. This process is suitable for product designs using the membrane 22 due to the small drop size. A flow stop structure 53 (e.g., a ring made of oxide in the membrane) can be designed in the membrane 22 to prevent the ink from spilling towards the center of the membrane, which should remain transparent to allow the passage of the light beam.

[0115] It is noted that the ink disposed radially outward of the flow stopping structure 53 on the membrane 22 can be considered to represent another example of a diaphragm defining an aperture, as discussed above using the example of a chrome coating.

[0116] In both embodiments (metallization or inkjet coating), it is advantageous if the cavity sidewalls slope towards the backside of the semiconductor chip 4, as indicated by the dashed lines in FIG.

[0117] Spacer (Fig. 16) 16 shows an embodiment in which a spacer 61 is disposed between the base substrate 6 and the semiconductor chip 4. The spacer 61 and the semiconductor chip 4 together form a substrate 60.

[0118] The spacer 61 is preferably also made of silicon. The spacer 61 has a central opening (through hole) that extends from a bottom side of the spacer 61 facing the base substrate 6 to a top side facing the semiconductor chip 4. The central opening is arranged coaxially with a cavity in the semiconductor chip 4. The cavity in the semiconductor chip 4 and the central opening in the spacer 61 together form a cavity 5 in which the light source 1 is located.

[0119] 16, the central opening in the spacer 61 has a lateral dimension that is slightly larger than the cavity in the semiconductor chip 4. However, in other embodiments, the lateral dimension of the opening in the spacer 61 can be the same as or smaller than the cavity in the semiconductor chip 4.

[0120] The spacer 61 increases the distance H between the light source 1 and the optical element along the optical axis 20. A larger distance H allows the use of optical elements with a larger focal length. On the one hand, such optical elements can be manufactured more easily. On the other hand, a larger focal length of the optical element allows an increase in the distance between the optical element and the focus of the light beam. It should be noted that this distance is not necessarily the same as the focal length of the optical element, since the focal position usually depends on the emission characteristics of the light source (e.g. divergent or collimated) and, in the case of divergent emission, on the distance H between the light source and the optical element. By increasing the distance between the optical element and the focus of the light beam, the size of the detection volume 8 can be increased. A larger distance H between the light source 1 and the optical element also reduces the sensitivity of the device to differences in manufacturing processes and materials, in particular differences in the thickness of the semiconductor chip 4, which increases the manufacturing stability and reduces the performance differences from device to device. This is particularly important when a collimated (cylindrical) light beam is used as opposed to a converging light beam, as collimation is particularly sensitive to tolerances in the distance between the light source and the optical element.

[0121] If spacers are not used, the distance H is limited by the maximum available thickness of the wafer on which the semiconductor chip 4 is fabricated. For example, the maximum thickness of a commercially available 8 inch silicon wafer is typically 720 μm. By using spacers, also fabricated from silicon wafers, the total thickness of the substrate 60 can easily be doubled. If an even greater thickness is desired, two or more spacers can be stacked or a thicker spacer can be obtained by using larger wafers, which may be available in larger thicknesses.

[0122] As a result, the thickness of the semiconductor chip 4 in which the photodetector 3 is integrated becomes a freely adjustable design parameter: for example, it is possible to use thin silicon wafers (typically about 300 μm) for the manufacture of the photodetector 3 and the electronic circuitry, and to use spacers 61 of the desired thickness to compensate for the remaining part of the required distance between the light source 1 and the optical element.

[0123] The wafer on which the photodetectors are integrated and the silicon spacer wafer can be connected before dicing by readily available bonding techniques such as the "direct bonding" process, in which two Si wafers are bonded together using van der Waals forces. Alternatively, "adhesive bonding" can also be used, using structured foils as the bonding interface.

[0124] If desired, an opaque coating may be applied to the sidewalls of the central opening and / or the backside of the spacer, as described above with respect to the semiconductor chip in which the photodetector is integrated.

[0125] FIG. 16 shows an optical element in the form of a polymer lens 15 on a glass carrier substrate 23, however, any optical element can be used in conjunction with a spacer.

[0126] Imprinting in membranes (Figures 17 to 19) Optical elements can be created by depositing structures functioning as refractive optical elements (ROEs) or diffractive optical elements (DOEs) directly onto the film formed by one or more layers of the CMOS stack. Examples are shown in Figures 17-19.

[0127] The membrane 22 is typically formed by forming a CMOS stack on a silicon wafer and then etching the wafer from the backside to form the cavity 5. An etch stop can be formed by the lowest oxide layer in the stack. Further thinning of the membrane can be done by further etching from within the cavity and / or from above the CMOS stack. In the region of the membrane, the CMOS stack should preferably only contain SiO and / or SiN layers to make the membrane transparent to light.

[0128] The optical structures are then formed directly on the film 22 by wafer level optical processes. Wafer level optical processes applied directly to the wafer on which the semiconductor chips 4 are formed have several advantages: the manufacturing tolerances are reduced because the optical elements are formed by wafer level processes; the optical elements can be closer to the photodetector plane where the photodetectors are located; the shadowing caused by the optical elements is minimized, which increases the measurement volume; the flow over the sensor is more laminar; there is an overall better measurement accuracy in laminar flow; in addition, the sensor is less susceptible to dirt accumulation, so an extended lifespan can be achieved. This design is inherently fluid-tight at the first end of the cavity, so it is particularly well suited for applications such as wearables where some level of water resistance is required.

[0129] In the embodiment of Figure 17, a polymer lens 71 is formed by wafer level imprinting directly on the membrane 22. It is also possible to imprint a Fresnel lens 72 (Figure 18) or a diffractive optical element (DOE) pattern 73 (Figure 19) to obtain an even flatter device topography, as shown in Figures 18 and 19.

[0130] It is also possible to generate ROE or DOE patterns on the film 22 by related wafer level techniques such as nanoimprint lithography or grayscale lithography. In grayscale lithography, photoresist is applied to the wafer surface in a spin-coating process. Standard photolithography equipment is used in conjunction with a grayscale mask to partially harden the photoresist. The unhardened resist is removed, leaving behind the shape of the optical element that remains on the wafer. Grayscale lithography can also be performed using direct laser writing, where the grayscale hardening intensity is adjusted by varying the laser power as it scans the photoresist surface on the wafer.

[0131] In all these techniques, the optical polymer or photoresist should not cover the photodetectors 3 or wire bond pads on the semiconductor chip 4. This can be accomplished by selectively UV curing the optical polymer or photoresist through a partially transparent stamp. After forming the optical elements, the uncured, still liquid polymer can be removed from the surface of the semiconductor chip 4.

[0132] Dispensing or Droplet Microlenses (Figure 20) In some embodiments, the optical element can include a dispensed or droplet microlens 74, as shown in FIG. 20. The dispensed or droplet microlens acquires its shape by the action of phenomena such as surface tension, wetting or dewetting, and gravity. The shape of the dispensed or droplet microlens depends on several factors, including the surface energy of the film, the volume of the droplet, the structure of the film surface, and the surface tension of the droplet. These parameters can be modified to some extent to affect the optical properties of the microlens. For example, the surface energy of the film can be changed by a plasma process. The surface tension of the droplet can be changed by choosing a different lens material.

[0133] Droplet lenses can also be obtained using a fused photoresist process, in which (binary) photolithography is used to form polymer pillars on top of a membrane, after which a reflow process is used to melt the polymer.

[0134] Optical elements on the bottom side of the membrane It is also possible to place the optical structures 75 on the bottom side of the membrane 22 facing the cavity 5, as shown in Fig. 21. The same techniques as described above can be used to form the optical structures. For this, the wafer can be flipped over so that the cavities are open towards the top, and the optical structures can be formed from above by imprinting and UV curing or by photolithography. The side walls of the cavities act as an inherent flow stop for the polymer or photoresist. In an alternative process, the wafer is not flipped over and the optical structures are formed from below by a process of filling a stamp with an optical polymer and pressing the wafer against the stamp from above.

[0135] DOE in membranes (Figure 22) FIG. 22 shows an embodiment in which the optical structure 76 is formed directly inside the CMOS membrane, i.e. inside the membrane formed by the layers of the CMOS stack 24. This can be achieved by fabricating the optical structure 76 using the CMOS process or by a subsequent wafer level process applied above or below the CMOS membrane. In this way, manufacturing tolerances can be further reduced and costs can also be reduced. The flow over the sensor becomes more laminar, further reducing the accumulation of dirt on the optical elements. Again, such a design is inherently fluid-tight at the first end of the cavity.

[0136] When a membrane is structured in this way, the membrane's material can be considered to be transformed into a metamaterial, i.e., a material that is structured so that it has properties that the material did not have prior to structuring.

[0137] A method for fabricating a DOE using a CMOS process is disclosed in the following publication: Dai, Ching-Liang & Chen, Hunglin & Lee, Chi-Yuan & Chang, Pei-Zen, "Fabrication of diffractive optical elements using the CMOS process", Journal of Micromechanics and Microengineering. 12(1):22 (2001), DOI: 10.1088 / 0960-1317 / 12 / 1 / 304.

[0138] Another possible manufacturing method is as follows: the optical structures can be formed on the wafer by imprinting photoresist or by photolithography. Afterwards, the photoresist can be removed and an etching process can be applied that transforms the optical structures of the photoresist into the film topography. This can be done from either side of the film.

[0139] In these embodiments, stray light for monitoring the light output of the light source 1 can reach the auxiliary detector 25 directly through the CMOS layer of the membrane. An opaque coating can therefore be safely applied to the side walls 26 of the cavity 5.

[0140] 2K molded lens (Figure 23) 23 shows one embodiment in which the optical element is a 2K molded lens 77. A 2K molded lens comprises an injection molded lens frame where the lens itself is replicated by dispensing lens material into a replication mold and UV curing.

[0141] The embodiments of Figures 17-23 are shown with spacers 61, however the spacers may be omitted. As described in relation to Figure 15, an opaque layer may be applied to the cavity sidewalls to reduce stray light at the photodetector.

[0142] Shape of cavity sidewall (Figures 24 to 26) Figures 24 to 26 show some examples of possible shapes of the cavity sidewalls 26. In Figure 24, the sidewalls 26 slope towards the bottom, i.e. the lateral dimension of the cavity 5 increases towards the bottom. Such an embodiment is particularly advantageous when an opaque coating is applied to the cavity sidewalls 26. In Figure 25, the sidewalls slope upwards. In Figure 26, the sidewalls have an upper apex that slopes upwards and a bottom that opens downwards in a convex shape. As known in the art, different shapes can be easily formed by suitable etching methods.

[0143] In all these embodiments, the cavity 5 has an axis of symmetry perpendicular to the photodetector plane. For example, in the case of a cavity with a square cross section, the cavity may have four-fold rotational symmetry around the axis of symmetry. In the case of a cavity with a circular cross section, the cavity may have cylindrical symmetry. The axis of symmetry preferably coincides with the optical axis 20.

[0144] Photodetector sectioning (Figure 27) 27 shows an embodiment in which some of the photodetectors are covered by an opaque layer, e.g. a black layer formed by inkjet printing. The opaque layer shields these photodetectors from light scattered by PM in the detection volume, while the other photodetectors face the detection volume without being shielded. The unshielded photodetectors form a first section 81 and the shielded photodetectors form a second section 82. The control unit receives signals from both the shielded and unshielded photodetectors and can apply differential processing to cancel signals due to electromagnetic interference.

[0145] Sensor module (Fig. 28 and Fig. 29) FIG. 28 shows one embodiment of a complete PM sensor module 90. The sensor module 90 comprises a PM sensor 91 according to any one of the embodiments described above. The PM sensor 91 is housed in a housing 92 that defines a flow channel 97. A heater 93 generates a convection current 98 in the flow channel 97. The PM sensor emits a light beam 7 into the flow channel 97. In this embodiment, the direction of the light beam 7 is perpendicular to the direction of the convection current 98. The light beam is deflected by a mirror 94 into a beam dump 95.

[0146] In this example, the light beam 7 is a focused beam having a focal point 21. The focal point is located within the flow channel 97. The detection volume is therefore located within the flow channel 97.

[0147] Figure 29 shows another embodiment of the complete PM sensor module. In this embodiment, there is no heater. Instead, the flow 98 is generated by a fan 99. [Explanation of symbols]

[0148] 1 light source 2. Optical elements 3, 3a, 3b Photodetector 4. Semiconductor chips 5 Cavity 6 Base Board 7. Light Beam 8 Detection Volume 9,9' PM particles 10, 10' scattered light pulse 11 Scattering particle height 12 Conventional Optical Lenses 13 Diffractive Optical Element (DOE) on a Film 14 Distance from the optical axis of the photodetector 15 Polymer lenses on glass substrates 16 Support 17 Optical elements, e.g. glass windows 18 Metallization 19 Light Barrier 20 Optical axis 21 focus 22 membrane 23 Glass Substrate 24 CMOS stack 25 Auxiliary Detector / Photodiode 26 Cavity Wall 27 Control Unit 28 Optical Filters 29 Chrome Coating 30 Aperture 31 Wire Bond 32 Enclosure 41 Asymmetrical extension 51 Metallization 52 Inkjet Coating 53 Inkjet flow stop 60 Substrates 61 Spacer 71 Wafer-level polymer imprinted lens directly on the membrane 72 Wafer-level polymer imprinted Fresnel lens directly above the membrane 73 Wafer-level polymer imprint pattern directly on the membrane 74 Wafer-level droplet microlenses directly above the membrane 75 Wafer-level polymer imprint pattern on bottom side of membrane 76 Optical Structures in Films (DOE) 77 2K molded lens 81 First Section 82 Second Section 90 PM Sensor Module 91 PM sensor 92 PCB 93 Heater 94 Mirror 95 Beam dump 96 Housing 97 Flow Channel 98 Flow 99 Fans

Claims

1. a substrate including a semiconductor chip (4) forming a cavity (5), at least a part of which is formed in the semiconductor chip (4) itself; At least one photodetector (3) integrated on the surface of said semiconductor chip (4); a light source (1) arranged within the cavity (5) and emitting a light beam (7) towards a first end of the cavity (5) defining a detection volume (8) for particulate matter (9) outside the cavity (5); Equipped with said surface of said semiconductor chip (4) on which at least one said photodetector (3) is integrated faces said detection volume (8), A particulate matter sensor, wherein the at least one light detector (3) is adapted to detect light (10) scattered by particulate matter (9) within the detection volume (8).

2. The semiconductor chip (4) has a CMOS stack (24), 2. The particulate matter sensor of claim 1, wherein one or more layers of the CMOS stack (24) form a membrane (22) spanning a first end of the cavity (5).

3. 3. A particulate matter sensor as claimed in claim 1 or 2, comprising an optical element (2) arranged at the first end of the cavity (5), the optical element (2) configured to shape the light beam (7) so that the light beam (7) forms the detection volume (8).

4. A particulate matter sensor as described in claim 3, wherein the optical element (2) defines a boundary of the cavity (5) at the first end and the light beam (7) forms the detection volume (8).

5. 5. A particulate matter sensor according to claim 3 or 4, wherein the optical element defines an optical axis (20), the optical axis (20) being perpendicular to the surface of the semiconductor chip (4) in which at least one photodetector (3) is integrated.

6. The particulate matter sensor according to any one of claims 3 to 5, wherein the optical element (2) is configured to focus the light beam (7).

7. the optical element (2) is arranged in a plane defined by the surface of the semiconductor chip (4) on which at least one photodetector (3) is integrated, and / or The particulate matter sensor according to any one of claims 3 to 6, wherein the optical element (2) protrudes by no more than 1 mm from the surface of the semiconductor chip (4) in which the at least one photodetector (3) is integrated.

8. A particulate matter sensor according to any one of claims 3 to 7, wherein the at least one photodetector (3) is arranged at a distance of maximum 2 mm from the optical element (2).

9. The particulate matter sensor according to any one of claims 3 to 8, wherein the optical element (2) comprises a refractive lens (12, 15, 71, 72, 74) or a diffractive optical element (13, 73, 75).

10. The optical element (2) includes a glass carrier substrate (23) and a polymer lens (15) formed on the glass carrier substrate (23); A particulate matter sensor according to any one of claims 3 to 9, wherein the glass carrier substrate (23) has a thickness of less than 1000 μm.

11. The semiconductor chip (4) has a CMOS stack (24), The optical element (2) comprises a membrane (22) formed by one or more layers of the CMOS stack (24), The particulate matter sensor according to any one of claims 3 to 10, wherein an optical structure (71, 72, 73, 74, 75) is arranged on the membrane (22) and forms, together with the membrane (22), the optical element (2).

12. The semiconductor chip (4) has a CMOS stack (24), The optical element (2) comprises a membrane (22) formed by one or more layers of the CMOS stack (24), The particulate matter sensor according to any one of claims 3 to 10, wherein the membrane (22) comprises at least one CMOS layer (76) structured to form the optical element (2).

13. A particulate matter sensor according to any one of claims 3 to 12, further comprising a diaphragm (29) formed by a coating on the optical element (2) and defining an aperture for the light beam (7).

14. The particulate matter sensor according to any one of claims 1 to 13, wherein the particulate matter sensor comprises a plurality of the optical detectors (3), the optical detectors (3) being arranged at different locations around the cavity (5).

15. 15. The particulate matter sensor according to claim 14, wherein on the surface of the semiconductor chip (4) on which the photodetectors (3) are integrated, the photodetectors (4) are separated by a conductive material (18).

16. A particulate matter sensor as described in any one of claims 1 to 15, wherein an optical filter (28) is arranged on the surface of the semiconductor chip (4) in which at least one of the photodetectors (3) is integrated, the optical filter (28) covering the at least one of the photodetectors (3), and the optical filter (28) is configured to remove light having wavelengths outside a wavelength range including a main wavelength of the light source (1).

17. 17. The particulate matter sensor according to claim 1, further comprising a control unit (27) electrically connected to at least one of the photodetectors (3) and adapted to receive a signal from the at least one photodetector (3) caused by light (10) scattered by particulate matter (9) in the detection volume (8) and to determine a physical quantity related to the particulate matter (9) based on the signal.

18. A particulate matter sensor as described in claim 17, wherein the control unit (27) is integrated into the semiconductor chip (4).

19. a light-sensitive auxiliary detector (25) integrated in the semiconductor chip (4) and configured to receive light emitted by the light source (1) and not scattered by particulate matter (9) in the detection volume (8); the control unit (27) is connected to the auxiliary photosensitive detector (25) and is configured to determine the light output of the light source (1) based on a signal of the auxiliary photosensitive detector (25); 19. A particulate matter sensor as claimed in claim 17 or 18, wherein the control unit (27) is configured to take into account the determined light output when determining the physical quantity related to the particulate matter or to control the light source (1) depending on the determined light output.

20. A particulate matter sensor as described in claim 19, wherein the photosensitive auxiliary detector (15) is a photodiode.

21. A particulate matter sensor as described in claim 19, comprising an optical element (2) arranged at the first end of the cavity (5), the optical element (2) configured to shape the light beam (7), and the photosensitive auxiliary detector (25) arranged adjacent to the optical element (2) and configured to measure stray light from the optical element (2).

22. The photodetector (3) is divided into a first section (81) facing the detection volume (8) and a second section (82) shielded from light (10) scattered by particulate matter (9) in the detection volume (8); The particulate matter sensor according to any one of claims 17 to 21, wherein the control unit (27) is configured to perform a differential measurement of the first section (81) and the second section (82) thereby to counteract spurious effects of electromagnetic interference with at least one of the optical detectors (3).

23. The particulate matter sensor according to any one of claims 1 to 22, further comprising a light blocking element (41), the light blocking element (41) being arranged on the surface of the semiconductor chip (4) in which at least one of the photodetectors (3) is integrated, so as to selectively block one or more parts of the photodetectors (3b) from light scattered by the particles in the detection volume (8), the particulate matter sensor being determined depending on a distance of a particle of particulate matter from the surface of the semiconductor chip (4) in which the at least one photodetector (3) is integrated, while one or more other photodetectors (3a) are not blocked by the light blocking element (41).

24. A control unit (27) electrically connected to at least one of said photodetectors (3) and adapted to receive a signal from at least one of said photodetectors (3) resulting from light (10) scattered by particulate matter (9) in said detection volume (8) and to determine a physical quantity related to said particulate matter (9) based on said signal, the control unit (27) is configured to determine a measure of the distance of the particle from the surface of the semiconductor chip (4) in which at least one of the photodetectors (3) is integrated by comparing signals from photodetectors (3b) that are partially blocked by the light blocking element (41) with signals from photodetectors (3a) that are not blocked by the light blocking element (41), 24. A particulate matter sensor according to claim 23, wherein the control unit (27) is configured to take into account the determined distance when determining the physical quantity related to the particulate matter.

25. The particulate matter sensor according to any one of claims 1 to 24, wherein the substrate comprises the semiconductor chip (4).

26. The substrate (60) includes a spacer (61), and the semiconductor chip (4) is bonded to the spacer; A particulate matter sensor according to any one of claims 1 to 24, wherein the cavity (5) is formed both in the spacer (61) and in the semiconductor chip (4).

27. A base substrate (6) is provided, and the light source (1) is attached on the base substrate (6); the substrate in which the cavity (5) is formed is disposed on the base substrate (6), the second end of the cavity (5) being open towards the base substrate (6); A particulate matter sensor according to any one of the preceding claims, wherein the base substrate (6) extends in a plane parallel to the surface of the semiconductor chip (4) in which the photodetector (3) is integrated.

28. A particulate matter sensor as claimed in any one of claims 1 to 27, wherein the cavity (5) is laterally bounded by a sidewall (26), the sidewall (26) being formed by the substrate and at least a part of the sidewall (26) being formed by the semiconductor chip (4).

29. A particulate matter sensor as described in any one of claims 1 to 28, wherein the cavity (5) has an axis of symmetry, the cavity (5) has rotational symmetry around the axis of symmetry, and the axis of symmetry is perpendicular to the surface of the semiconductor chip (4) on which the photodetector (3) is integrated.

30. 30. A particulate matter sensor as claimed in any one of claims 1 to 29, further comprising an opaque coating (51, 52) covering at least one side wall (26) of the cavity (5) to prevent light from the light source (1) from passing through the at least one side wall (26) to reach the at least one photodetector (3).

31. A housing (92); a flow channel (97) disposed within the housing (92); a fan (99) or heater (93) disposed within the housing (92) and configured to move air through the flow channel (97); A particulate matter sensor according to any one of claims 1 to 30, wherein the flow channel (97) is disposed within the housing (92) such that the flow channel (97) comprises at least a portion of the detection volume (8); A particulate matter sensor module comprising:

32. A method for determining a physical quantity of particulate matter using the particulate matter sensor according to any one of claims 1 to 30, comprising: operating the light source (1) to emit the light beam (7); operating at least one said light detector (3) to detect light scattered by particulate matter intersecting said light beam (7); - analysing the signal from at least one said optical detector (3) to obtain at least one parameter indicative of said physical quantity of said particulate matter; The method comprising:

33. A method for manufacturing a particulate matter sensor according to any one of claims 1 to 30, comprising the steps of: a) forming at least one photodetector (3) on a surface of a semiconductor chip (4); b) etching the semiconductor chip (4) along a direction perpendicular to the surface so as to form at least a part of a cavity (5); c) optionally bonding said semiconductor chip (4) to a spacer (61) forming another part of said cavity (5); d) placing a light source (1) within said cavity (5) configured to emit a light beam (7) towards a first end of said cavity (5); The method comprising:

Images