Radio frequency device

Abstract

There is provided a radio frequency, RF, device (100, 102, 104, 106) comprising: an active semiconductor layer (110) comprising a front surface (112) and a rear surface (114); and an RF transmission line (120) coupled to the front surface (112) of the active semiconductor layer (110), the RF transmission line (120) comprising an active region (121), the active region (121) causing the RF transmission line (120) to comprise a first state in which the RF transmission line (120) has a first radio frequency transmission property, and a second state in which the RF transmission line (120) has a second radio frequency transmission property, wherein the RF transmission line (120) is selectively switchable between the first state and the second state in response to illuminati on of the active semiconductor layer (110) by a light source (210) when in use, and wherein the active semiconductor layer (110) comprises an intrinsic semiconductor material with a thickness, between the front surface (112) and the rear surface (114), of less than 100 microns.

Description

[0001] RADIO FREQUENCY DEVICE

[0002] FIELD OF THE INVENTION

[0003] The present invention relates to radio frequency (RF) devices and methods of manufacturing such devices.

[0004] BACKGROUND

[0005] Radio frequency (RF) devices, such as RF switches and variable RF attenuators, are essential components in many electronic systems, particularly for communication and radar-based applications. Radio frequencies include those that are typically referred to as radio-waves, microwaves, millimetre-waves, and terahertz-waves - ranging from 3 kHz to 300 GHz.

[0006] There are many types of RF devices that have been developed over the course of decades, but each has at least one drawback associated with it. RF devices that are operable at mmWave frequencies may be referred to as mmWave devices. Electromechanical devices (e.g., relay switches) are usually physically large and have slow switching times. Conventional solid-state devices - for example diodes, field effect transistors (FETs) and monolithic microwave integrated circuits (MMICs) - typically have non-linearities in their response that lead to intermodulation distortion. Meanwhile, micro-electromechanical systems (MEMS) often face repeatability issues, high actuation voltage requirements , and relatively slow switching speeds.

[0007] More recently, optoelectronic RF devices have gained interest. Optoelectronic RF devices use an optical signal to control an RF signal. The lack of mechanical parts and / or electronic control in such devices may help address some of the drawbacks discussed above. However, the implementation of optoelectronics also poses challenges. Charge carrier velocity and recombination within semiconductors provide constraints to switching speeds and power handling capabilities. Thermal challenges may also arise, wherein the heat generated by RF signals and light sources can place limitations on the construction of the device, which may reduce effectiveness or efficiency.

[0008] Therefore, further development of optoelectronic RF devices is desirable in order to address at least some of these issues and to enable their wider use in RF applications. SUMMARY

[0009] According to a first aspect of the invention, there is provided a radio frequency, RF, device. The RF device comprises: an active semiconductor layer comprising a front surface and a rear surface; and an RF transmission line coupled to the front surface of the active semiconductor layer, the RF transmission line comprising an active region, the active region causing the RF transmission line to comprise a first state in which the RF transmission line has a first radio frequency transmission property, and a second state in which the RF transmission line has a second radio frequency transmission property, wherein the RF transmission line is selectively switchable between the first state and the second state in response to illumination of the active semiconductor layer by a light source when in use, and wherein the active semiconductor layer comprises an intrinsic semiconductor material with a thickness, between the front surface and the rear surface, of less than 100 microns.

[0010] When the active semiconductor layer is illuminated (by photons with an energy greater than the bandgap of the semiconductor material), a high density of charge carriers (electrons and holes), which may be referred to as a plasma, is created within the active semiconductor layer. The charge carrier plasma creates a conducting region within the active semiconductor layer in the area in which light is incident. An active region of the RF transmission line may be defined, corresponding with (overlying) the conducting region of the active semiconductor layer that is generated when illuminated. The optically-induced charge carriers are therefore able to modify an RF transmission property of the RF transmission line that is coupled to the active semiconductor layer. The RF transmission property may be transmission coefficient or transfer impedance for example, depending upon the configuration of the RF transmission line.

[0011] The RF device may be an optoelectronic RF switch or form part of an optoelectronic variable RF attenuator, for example. When the active semiconductor layer is not illuminated, it behaves as an insulating material and the RF transmission line is in a first state. When the active semiconductor layer is illuminated by appropriate light, the optically-induced charge carriers cause the active semiconductor layer in the illuminated area to behave as an electrically conducting material, thereby causing the RF transmission line to switch to a second state. The first state and second state of the RF transmission line may correspond to two distinct transmission coefficients (measured via an S21 measurement, for example). In the first state (which may be where the RF device is unilluminated), the transmission coefficient may be approximately -34 dB, or between -10 and -50 dB across the range of 0 to 30 GHz, for example. In the second state (which may be where the RF device is illuminated), the transmission coefficient may be approximately -1.4 dB, or between - 0.1 and -5dB across the range of 7 to 40 GHz, for example. When the transmission coefficient in the first state is smaller than the transmission coefficient in the second state, the RF device (or the RF transmission line) may be said to be in a series configuration. When the transmission coefficient in the first state is greater than the transmission coefficient in the second state, the RF device (or the RF transmission line) may be said to be in a shunt configuration. The reciprocal of the smaller transmission coefficient may be referred to as the isolation of the RF device, while the reciprocal of the greater transmission coefficient may be referred to as the insertion loss of the RF device.

[0012] Compared to electronic devices of the prior art (e.g., diodes and transistors), the use of optical control may advantageously enable higher power handling and / or lower non-linear distortion, due to the fact that there no additional electrical control signal is used in switching the RF transmission line from the first state to the second state. There may be no PN junctions, heterostructures and / or Schottky contacts which give rise to non-linear behaviour. The optically-induced charge carrier plasma may be inherently linear.

[0013] An ‘intrinsic’ semiconductor material may be taken to mean that it has a low intrinsic-point-defect density. The intrinsic-point-defect density is less than I xlO15 / cm3for silicon, for example. The intrinsic-point-defect density is less than 1.5xl014 / cm3for gallium arsenide, for example.

[0014] An intrinsic semiconductor material may be taken to mean that it has a low impurity density. The impurity density of carbon, nitrogen, or oxygen may be less than I xlO15 / cm3and the impurity density of any other element may be less than I xlO12 / cm3for silicon, for example. The impurity density of boron may be less than 1.5xl014 / cm3and the impurity density of any other element may be less than 1.5xl013 / cm3for gallium arsenide, for example.

[0015] The relatively low intrinsic-point-defect density and impurity density of an intrinsic semiconductor material may reduce the thermal -equilibrium charge carrier concentration and increase the optically -induced charge carrier concentration within the active semiconductor layer. This may reduce the insertion loss and increase the isolation of the RF device.

[0016] The relatively low intrinsic-point-defect density and impurity density of an intrinsic semiconductor material may reduce the interaction between the RF signal and the semiconductor material. This may improve the linearity and reduce the dispersion of the RF device.

[0017] The relatively low intrinsic-point-defect density and impurity density of an intrinsic semiconductor material may increase the breakdown voltage of the semiconductor material. This may improve the power handling capabilities of the RF device.

[0018] An intrinsic semiconductor material may be a type of undoped, semi-insulating semiconductor material. An ‘undoped’ semiconductor material may be taken to mean that it has not been intentionally doped. A ‘semi -insulating’ semiconductor material may be taken to mean that its room-temperature electrical resistivity is relatively high (e.g., > IxlO4Q cm for silicon and > 5xl07Q cm for gallium arsenide).

[0019] The intrinsic / undoped semi-insulating semiconductor material may be manufactured from a bulk semiconductor crystal (e.g., a float-zone silicon ingot or a vertical-gradient- freeze-grown gallium arsenide ingot).

[0020] Positioning the RF transmission line at the front surface of the active semiconductor layer while providing the selective illumination from the rear surface may provide for a more efficient RF device. Using conventional front illumination, some of the incident light will be reflected by the RF transmission line (approximately proportion al to the proportion of the area of the front surface covered by the RF transmission line). Providing illumination from the rear surface may therefore allow more of the provided light to be incident upon the active semiconductor layer, therefore enabling lower power light sources to be used. Illumination from the rear surface may enable complete absorption (i.e., no metal-incurred reflection, though the semiconductor material may have some inherent reflectance). However, it will be understood that the RF device may be illuminated from the front surface of the active semiconductor layer; this may still result in sufficient charge carrier plasma being formed so as to switch the RF transmission line.

[0021] In order for illumination at the rear surface (sometimes referred to as bottom illumination) to be effective, the active semiconductor layer must be sufficiently thin. The incident light will only penetrate a certain distance into the active semiconductor layer (determined by the penetration / absorption depth of the semiconductor material at the wavelength of the incident light). Meanwhile, the optically -induced charge carrier plasma will only exist over certain volume / depth (determined by the diffusion coefficient, lifetime and surface recombination velocity of the semiconductor material). Thus, if the active semiconductor layer is too thick, the charge carrier plasma generated at the rear surface will not extend to the RF transmission line at the front surface, and there will not be sufficient coupling of the plasma to the RF transmission line (and therefore switching between the first state and second state will be minimal). However, the active semiconductor layer cannot be too thin, in order to prevent it being substantially transparent to the incident light.

[0022] Therefore, the active semiconductor layer must be “adequately thin”. An “adequately thin” semiconductor layer in the context of the RF device may be defined as: i) half the absorption depth of the semiconductor material at the wavelength of the incident light is smaller (or preferably much smaller) than the diffusion length of the semiconductor material at the wavelength of the incident light; ii) the thickness of the active semiconductor layer is greater than half the absorption depth of the semiconductor material at the wavelength of the incident light ; and iii) the thickness of the active semiconductor layer is smaller than the diffusion length of the semiconductor material at the wavelength of the incident light.

[0023] Evidently, this definition of ‘ adequately thin’ is dependent upon the wavelength of the incident light and the optical and semiconductor properties of the semiconductor material. The absorption depth D may be defined as D = A / 4nk, where A is the wavelength of the incident light and k is the extinction coefficient of the semiconductor material at A. The absorption depth D may be defined as the depth at which l / e2 intensity is reached. The diffusion length L may be defined as L = j DaT, where Dais the diffusion coefficient of the semiconductor material and T is the lifetime of the semiconductor material.

[0024] For an intrinsic semiconductor material, half the absorption depth of the semiconductor material at the wavelength of the incident light is usually much smaller than the diffusion length of the semiconductor material at the wavelength of the incident light. For example, half the absorption depth and diffusion length of intrinsic silicon (commercially often referred to as float -zone, undoped silicon with a resistivity greater than 10000 Q-cm) at 850 nm may be 9 microns and 20 mm, respectively.

[0025] For materials / wavelengths of interest for the RF device, the thickness of the active semiconductor layer (i.e., the distance between the front surface and the rear surface) is less than 100 microns. The thickness of the active semiconductor layer may be between 0.1 and 100 microns thick, between 5 microns and 100 microns thick, between 10 microns and 75 microns thick, between 25 microns and 75 microns thick, or between 30 and 60 microns thick, for example. The thickness of the active semiconductor layer may be between 0.1 and 1 microns, between 5 microns and 20 microns thick, between 10 microns and 50 microns thick, between 50 microns and 75 microns thick, or between 75 and 100 microns thick, for example. The thickness of the active semiconductor layer may be 0.5 microns, 5 microns, 10 microns, 25 microns, 50 microns, 75 microns, or 100 microns, for example.

[0026] The ‘adequately thin’ active semiconductor layer may also advantageously enable a lower insertion loss and a higher isolation of the RF device. Firstly, the ‘adequately thin’ semiconductor material may reduce the effective thermal -equilibrium conductance of the active semiconductor layer compared to devices of the prior art. Secondly, the ‘adequately thin’ semiconductor material may provide for greater confinement of the optically-induced charge carriers within the active semiconductor layer compared to devices of the prior art (i.e. although the effective photoconductance of the active semiconductor layer is not increased, more of the optically -induced charge carriers will be close enough to the active region of the RF transmission line so that they can affect its properties). The ‘adequately thin’ active semiconductor layer may advantageously enable greater switching speeds of the RF transmission line between the first state and second state. The ‘adequately thin’ active semiconductor layer may allow the charge carrier plasma conducting region to permeate to the RF transmission line at the front surface faster than a thick semiconductor layer, therefore enabling a faster non -illuminated to illuminated switching time. The greater influence of surface interactions provided by the ‘adequately thin’ active semiconductor layer may allow a faster illuminated to non-illuminated switching time; charge carriers may more rapidly recombine at the surfaces of the ‘adequately thin’ active semiconductor layer (due to the reduced distance between the centre of the charge carrier plasma and the surfaces of the active semiconductor layer), thereby causing a faster decay of the charge carrier plasma conducting region.

[0027] Furthermore, the ‘adequately thin’ active semiconductor layer may advantageously enable greater dissipation of the heat accumulated in the active region of the RF transmission line that is generated by a light source and / or a RF source to a potential heatsink.

[0028] The active semiconductor layer may comprise silicon and / or gallium arsenide. Silicon may be a preferable semiconductor material for the active semiconductor layer given its widespread use in electronics and well-established manufacture techniques and methods. However, dependent upon the application, different semiconductor materials may be used. Semiconductor materials with a smaller bandgap (e.g., germanium or indium nitride) may be used, which may allow lower power light sources to be used, for example. Semiconductor materials with a larger bandgap (e.g., gallium arsenide) may be used, which may allow a faster switching speed of the RF device to be achieved, for example.

[0029] The RF device may further comprise at least one of: a front passivation layer coupled to the front surface of the active semiconductor layer; and a rear passivation layer coupled to the rear surface of the active semiconductor layer. The front and / or rear passivation layer may comprise a passivation material. The passivation material may be an oxide of the semiconductor material (e.g., in the case that the semiconductor material is silicon, the passivation material may be silicon oxide , SiOx) such that the passivation layer may be easily formed by oxidation of the active semiconductor layer. Alternatively, the passivation material may be a non -similar oxide. For example, the active semiconductor layer may comprise silicon and the front and / or rear passivation layers may comprise aluminium oxide. Such passivation materials may be actively deposited onto the active semiconductor layer using any suitable deposition technique (compared to a native oxide layer formed via naturally occurring oxidation of the semiconductor material’s surface).

[0030] The passivation layers may reduce the influence of surface interactions at the surfaces of the active semiconductor layer and therefore increase the optically -induced charge carrier concentration within the active semiconductor layer. The passivation layers may protect the active semiconductor layer from the environment, thereby increasing the stability and / or the durability of the RF device. The passivation layers may reduce the chemical and / or electrochemical reactivity of the active semiconductor layer, for example. The front and / or rear passivation layers may be thin enough that the coupling of the active semiconductor layer and the RF transmission line is not substantially impacted. The passivation layers may be less than 100 nm in thickness, less than 50 nm in thickness, less than 30 nm in thickness or less than 10 nm in thickness, for example. However, the passivation layers may still be sufficiently thick so as to provide adequate surface passivation and / or protection. The passivation layers may each have a thickness of at least 1 nm, at least 2.5 nm, at least 5 nm or at least 10 nm.

[0031] The front passivation layer and / or the rear passivation layer may comprise silicon oxide or aluminium oxide. The use of A1OXmay be preferable to SiOx, as A1OXmay provide improved surface passivation, greater RF capacitance and / or greater thermal conductance for any given processing temperature and thickness.

[0032] The RF device may further comprise a substrate disposed at the rear surface of the active semiconductor layer, the substrate comprising a large bandgap material that is transparent to light from the light source. Where the RF device comprises a rear passivation layer, the substrate may be provided on a rear surface of the rear passivation layer. The material may be regarded as large -bandgap if its bandgap is larger than the semiconductor material of the active semiconductor layer. The material may be regarded as transparent to light if its extinction coefficient is smaller than 1x105from 400 nm to 2000 nm, for example.

[0033] The substrate may advantageously enable more convenient manufacture or handling of the RF device. Whereas the active semiconductor layer must be sufficiently thin (as discussed above), the substrate may be thicker than the active semiconductor layer so long as it is sufficiently transparent. The substrate may be at least 100 microns, or 250 microns, or 500 microns thick, depending upon the tensile strength of the substrate material, for example. The active semiconductor layer and other layers of the RF device may be manufactured whilst coupled to the substrate. The substrate may make it easier to handle the other layers of the RF device during manufacturing steps. The substrate may minimize the likelihood of device breakage in manufacture and packaging. A transparent-to-light substrate, if existing temporarily, may enable double -sided photomask alignment in manufacture.

[0034] The substrate may advantageously provide for greater thermal conduction of the other layers of the RF device. The substrate may act as a heatsink for the active semiconductor layer and / or the RF transmission line, which may reduce the likelihood of the RF device being thermally damaged and / or reduce the high-temperature-incurred degradation of the optically-induced charge carrier concentration within the active semiconductor layer when the RF device is illuminated with high-intensity light and / or provided with a high- power RF signal. The substrate may therefore enable higher power light sources and / or RF sources to be used.

[0035] The substrate may reduce the influence of surface interactions at the rear surface of the active semiconductor layer and therefore increase the optically -induced charge carrier concentration within the active semiconductor layer. When the substrate has a very large bandgap and has a very high purity, itself (without using a rear passivation layer) may provide for greatest reduction of the influence of surface interactions at the rear surface of the active semiconductor layer and therefore increase the optically -induced charge carrier concentration within the active semiconductor layer most significantly . The substrate may provide for greater electromagnetic -field confinement and therefore reduce the likelihood of having undesirable, transient RF transmission modes.

[0036] The substrate may comprise: glass (e.g. silicate glass, comprising mainly silicate) ; sapphire (i.e. crystalline alumina, AI2O3); silicon carbide; aluminium nitride; gallium nitride; and / or diamond.

[0037] Glass may be preferable in some applications due to it being a widely available and cheap material. Sapphire may be preferable in some applications due to it having a greater tensile strength, greater thermal conductivity and a greater refractive index than glass.

[0038] Materials other than glass or sapphire (e.g., silicon carbide, aluminium nitride gallium nitride, or diamond) may be preferable. Such materials may provide one or more of the following advantages: i) enabling > 300°C cleanroom processing, (e.g., high-temperature atomic layer deposition (ALD), high-temperature ion implantation, high -temperature plasma enhanced chemical vapor deposition (PECVD), high-temperature permanent bonding, high-temperature smart cut, and high-temperature annealing) which may provide improved processing quality and hence better semiconductor performance. Sapphire and glass may be limited to < 250°C processing due to having thermal expansion coefficients that are significantly different from the semiconductor materials (notably silicon), which may cause cracks at high temperatures due to temperature -induced stresses between the active semiconductor layer and the substrate; ii) incurring less light diffraction and reducing thickness -variation-incurred reflectance fluctuation due to the refractive indices of SiC, AIN, GaN and diamond being greater than those of sapphire and glass. This may mean that the incident light beam can be more focused and / or the substrate can be thinner (so as to let the light source be situated closer to the rear surface of the active semiconductor layer); iii) providing greater thermal dissipation than sapphire and glass, which may be used to enable greater light source powers and / or RF source powers to be used. The thermal conductivity of each of SiC, AIN, GaN and diamond is higher than that of either of sapphire or glass.

[0039] The RF device may further comprise an anti -reflection (AR) layer disposed at the rear surface of the active semiconductor layer. The AR layer may be provided only below the active region of the RF transmission line. The AR layer may comprise silicon oxide (which may be applied using PECVD) or a polymer coating, for example. Where the RF device comprises a substrate, the AR layer may be provided on the rear surface of the substrate. The AR layer may comprise any suitable material and may be configured to match a wavelength of the light source used to illuminate the RF device (e.g., a wavelength of an LED or laser light source may be within a waveband of the AR layer). Where the RF device comprises a rear passivation layer, the AR layer may be provided on a rear surface of the rear passivation layer. Where the RF device comprises a substrate, the AR layer may be provided on a rear surface of the substrate. The AR layer may increase the optical power that is provided to the active semiconductor layer by the incident light, thereby improving the optical efficiency of the RF device.

[0040] The RF device may further comprise a rear ground plane disposed at the rear surface of the active semiconductor layer. The rear ground plane may only be provided in the area which is not below the active region of the RF transmission line. The rear ground plane may have a thickness greater than 1 micron, for example. The rear ground plane may comprise at least two stacked transition metal materials / electrically conducting materials (e.g., gold on titanium or copper on chromium) . Where the RF device comprises a rear passivation layer, the rear ground plane may be provided on a rear surface of the rear passivation layer. Where the RF device comprises a substrate, the rear ground plane may be provided on a rear surface of the substrate.

[0041] The rear ground plane may provide for greater thermal conduction of the other layers of the RF device. The rear ground plane may act as a heatsink for the active semiconductor layer and / or the RF transmission line, which may reduce the likelihood of the RF device being thermally damaged and / or reduce the high-temperature-incurred degradation of the optically-induced charge carrier concentration within the active semiconductor layer, when the RF device is illuminated with high-intensity light and / or supplied with a high-power RF signal. The rear ground plane may therefore enable higher power light sources and / or RF sources to be used. In addition, the rear ground plane may serve as a rear ground plane of the RF transmission line and / or an extension of a potential front cathode (or anode) contact of the light source to which the front cathode (or anode) contact of the light source can be permanently bonded.

[0042] The RF device may further comprise a reflective layer disposed at the front surfaces of the active semiconductor layer and the RF transmission line, the reflective layer being reflective to light incident from the direction of the rear surface of the active semiconductor layer. The reflective layer may comprise a dielectric mirror layer. The reflective layer may comprise a Bragg reflector (comprising a series of sub -layers with alternating refractive index which may be applied using PECVD), for example. Where the RF device comprises a front passivation layer, the reflective layer may be provided on the front surfaces of the front passivation layer and the RF transmission line. Providing a reflective layer may increase the optical efficiency of the RF device and / or enable a thinner active semiconductor layer to be used, as any incident light that passes through the active semiconductor layer from the rear surface without being ab sorbed may be reflected back into the active semiconductor layer from the direction of the front surface, thereby providing a ‘second pass’ of the light. Even without a separate reflective layer, the RF transmission line may provide for a similar effect. The reflective layer may protect the RF transmission line from the environment, thereby increasing the stability and / or the durability of the RF device. The reflective layer may reduce the chemical and / or electrochemical reactivity of the RF transmission line, for example.

[0043] The active semiconductor layer may comprise a region which is doped with deep-level impurities (e.g., vanadium), the doped region encompassing a volume of the active semiconductor layer beneath the active region of the RF transmission line . The doped region may be non-photoconductive.

[0044] In this context, the doped region “encompasses” the volume of the active semiconductor layer beneath the active region of the RF transmission line in the sense that it surrounds or encloses the cross-sectional area of said volume in the plane of the active semiconductor layer. For example, the doped region may be a toroidal ring or an annulus that is doped. Said doped region may surround a cylindrical region of the active semiconductor layer, said cylindrical region being beneath the active region of the RF transmission line. Although the doped region comprises dopants, the region of the active semiconductor layer that is surrounded by the doped region may still be an intrinsic semiconductor material, as discussed above.

[0045] The doped region may surround the volume of the active semiconductor layer directly below the active region of the RF transmission line (or an illuminated region of the active semiconductor layer) such that the optically-induced charge carrier plasma is confined to said volume (or said illuminated region). Deep-level impurities may introduce defect states / energy levels near the centre of the bandgap of the semiconductor material, which can not only provide an efficient recombination path for the optically-induced free electrons but also minimise thermal excitation of defect -state electrons. By contrast, a shallow-level impurity introduces energy levels near to the edges of the valance band or conduction band and may therefore be readily filled by thermally excited charge carriers. Deep-level impurities may introduce energy states at least, for example, 0.1 eV, 0.2 eV or 0.3 eV from either the valance or conduction band edge and therefore experience minimal thermal recombination . The energy levels provided by the deep-level impurities may therefore enable the optically -induced charge carriers to recombine in the doped region, thereby trapping the charge carrier plasma within the doped region. This may result in a region with high resistivity in both the dark state and the illuminating state.

[0046] Confining the optically-induced charge carrier plasma within the volume of the active semiconductor layer below the active region of the RF transmission line may advantageously enable a smaller insertion loss and a greater isolation of the RF device. Firstly, isolating the said volume of the active semiconductor layer may reduce the effective thermal-equilibrium conductance of the active semiconductor layer compared to devices of the prior art. Secondly, isolating the said volume of the active semiconductor layer may provide for greater confinement of the optically -induced charge carriers within the active semiconductor layer compared to devices of the prior art (i.e. although the effective photoconductance of the active semiconductor layer is not increased, more of the optically -induced charge carriers will be close enough to the active region of the RF transmission line so that they can affect its properties).

[0047] The active region of the RF transmission line may be shaped so as to substantially conform to an area corresponding to the area of the active semiconductor layer that is illuminated by the light beam when in use. The active region of the RF transmission line may be shaped so as to substantially conform to the shape of the optically -induced charge carrier plasma. In other words, the boundary of the active region of the RF transmission line and the boundary of the area of the active semiconductor layer that is illuminated by the light beam when in use (i.e., where the value of the intensity drops to 1 / e2of the maximum value) may have a substantially similar centroid, shape, perimeter, and / or size. Preferably, the centroid of the former boundary and the centroid of the latter boundary may exactly overlap; the shape of the former boundary and the shape of the latter boundary may be identical; the ratio of the perimeter of the former boundary to the perimeter of the latter boundary may be between 1 and 2. By providing a ‘matching’ between the active region of the RF transmission line and the area of the active semiconductor layer that is illuminated by the light beam when in use, the active region of the RF transmission line and the optically-induced charge carrier plasma may interact more efficiently. This may provide better coupling between the active semiconductor layer and the RF transmission line.

[0048] The active region of the RF transmission line may be configured to be substantially circular in shape. Ordinarily, a light source beam (e.g., a laser beam) and its optically-induced charge carrier plasma may be circular, so providing a circularly shaped active region of the RF transmission line may improve the efficiency of its interaction with the optically-induced charge carrier plasma. The active region of the RF transmission line may be looped or meandered over a circular region.

[0049] The active region of the RF transmission line may comprise a meandered RF transmission line structure comprising : a meandered signal line; a front ground plane separated from and surrounding the meandered signal line; and a pair of meandered gaps formed between the meandered signal line and the front ground plane.

[0050] The active region of the RF transmission line may comprise a circular meandered RF transmission line structure comprising : a circular meandered signal line; a front ground plane separated from and surrounding the circular meandered signal line; and a pair of circular meandered gaps formed between the circular meandered signal line and the front ground plane.

[0051] The meandered RF transmission line structure (e.g., the circular meandered RF transmission line structure) may have a first terminal and a second terminal to which electrical components may be connected.

[0052] Such a meandered RF transmission line structure (e.g., a circular meandered RF transmission line structure) may be used to provide or form part of an optoelectronic RF shunt switch. An RF signal may be connected across the first and second ports of the RF transmission line (or the RF device). When the RF device is not illuminated, the meandered signal line is electrically isolated from the front ground plane so that the RF signal passes between the first and second ports of the RF transmission line (or the RF device). However, when the RF device is appropriately illuminated, the generation of the optically-induced charge carrier plasma shorts the meandered signal line to the front ground plane, thereby preventing the RF signal from passing between the first and second ports of the RF transmission line (or the RF device). Providing a meandered RF transmission line structure, rather than just a straight RF transmission line, may enable stronger coupling between the active region of the RF transmission line and the optically-induced charge carrier plasma.

[0053] The active region of the RF transmission line may comprise an interdigitated-gap- loaded RF transmission line structure comprising a first part and a second part, each of which comprises a plurality of interdigitated elements. The first part and second part of the interdigitated-gap-loaded RF transmission line structure are separated by at least one interdigitated gap.

[0054] The active region of the RF transmission line may comprise a circular, periodic, interdigitated-gap-loaded RF transmission line structure comprising: a plurality of “trees”, each of which comprises a “trunk” and a plurality of “branches”; and a plurality of interdigitated gaps .

[0055] Such an interdigitated-gap-loaded RF transmission line structure (e.g., a circular, periodic, interdigitated-gap-loaded RF transmission line structure) may be used to provide or form part of an optoelectronic RF series switch. An RF signal may be connected to the first and second ports of the RF transmission line (or the RF device). When the RF device is not illuminated, the first and second parts of the interdigitated- gap-loaded RF transmission line structure are electrically isolated so that the RF signal does not pass across the first and second ports of the RF transmission line (or the RF device). When the RF device is appropriately illuminated, the generation of the optically-induced charge carrier plasma enables current to flow between the first and second parts, thereby turning the switch to an ‘on’ state. E.g., the generation of the optically-induced charge carrier plasma shorts the first and second parts of the interdigitated-gap-loaded RF transmission line structure and thus the RF signal passes between the first and second ports of the RF transmission line (or the RF device).

[0056] The first part and the second part of an interdigitated-gap-loaded RF transmission line structure may comprise a first plurality of interdigitated elements and a second plurality of interdigitated elements, respectively, each plurality comprising at least 10 interdigitated elements that are interdigitated with one another. Using RF transmission line manufacturing techniques, more than 10, or more than 50 or more than 100 elements (from each plurality) may be interdigitated. Providing many interdigitated elements may improve the insertion loss of the RF device.

[0057] A circular periodic-interdigitated-gap-loaded RF transmission line structure may comprise 4 trees, each of which comprises 1 trunk and 80 branches, and 3 interdigitated gaps; may comprise 6 trees, each of which comprises 1 trunk and 40 branches, and 5 interdigitated gaps; may comprise 8 trees, each of which comprises 1 trunk and 120 branches, and 7 interdigitated gaps, for example. Providing many trees and / or interdigitated gaps may improve the isolation of the RF device. Providing many branches for each tree may improve the insertion loss of the RF device. Providing many trees and / or interdigitated gaps, and many branches for each tree may improve the isolation and the insertion loss of the RF device simultaneously.

[0058] According to a second aspect of the invention, there is provided a radio frequency, RF, system comprising: an RF device comprising: an active semiconductor layer comprising a front surface and a rear surface; and an RF transmission line coupled to the front surface of the active semiconductor layer, the RF transmission line comprising an active region, the active region causing the RF transmission line to comprise a first state in which the RF transmission line has a first radio frequency transmission property, and a second state in which the RF transmission line has a second radio frequency transmission property, wherein the RF transmission line is selectively switchable between the first state and the second state in response to illumination of the active semiconductor layer by a light source when in use, and wherein the active semiconductor layer comprises an intrinsic semiconductor material with a thickness, between the front surface and the rear surface, of less than 100 microns; a light source configured to illuminate the active semiconductor layer; a controller configured to selectively activate the light source such that the RF transmission line of the RF device is selectively switchable between the first state and the second state.

[0059] The RF device may be according to the first aspect of the invention.

[0060] By selectively activating the light source of the system (and thereby switching the RF transmission line of the RF device between the first state and second state), the RF device may be an optoelectronic RF series or shunt switch. The RF device may be used as a component and integrated with other RF devices, RF sources, and / or RF loads in a number of wider RF circuits based upon the configuration of the RF device and the implementation of the RF circuits. The RF device may be an optoelectronic multiple- pole-multiple-throw RF switch, an optoelectronic variable RF attenuator, an optoelectronic tuneable RF filter, an optoelectronic tuneable RF phase shifter, an optoelectronic tuneable RF antenna, a radar system, a communications system, or an RF test instrumentation, for example.

[0061] An RF circuit may be connected in series or in shunt across the ports of the RF device. The controller may be configured to receive an input from the RF circuit or from some other control means, such that RF system can be used to provide feedback control of the RF circuit. For example, the RF device may be configured in a shunt configuration (the RF device is an optoelectronic RF shunt switch in this case) and the RF circuit may be connected in series across the ports of the RF device, an RF signal ordinarily passing through with the light source not activated. When a predetermined parameter is reached (for example, the current across the RF transmission line exceeds a certain threshold), the light source may be activated, thereby shunting the RF signal to ground.

[0062] The light source may be configured to illuminate the active semiconductor layer from the rear surface of the active semiconductor layer. Illuminating the active semiconductor from the rear may enable the advantages discussed above (e.g., reduced optical reflection from the RF transmission line, thereby enabling greater efficiency) . The light source may be used to illuminate the active semiconductor layer from the front surface of the active semiconductor layer.

[0063] The light source may be monochromatic and configured to have a photon energy between the bandgaps of the semiconductor material and the substrate material. For example, where the semiconductor material is silicon (bandgap = 1.12 eV) and the substrate material is 4H-silicon carbide (bandgap = 3.20 eV), the light source may have a wavelength between 400 nm and 1100 nm (e.g., near infrared light at a wavelength of 940 nm).

[0064] The light source may comprise: a light emitting diode, LED; or a laser (including but not limited to an edge-emitting laser (EEL) and a vertical-cavity surface-emitting laser (VCSEL)).

[0065] Due to the efficiencies / capabilities of the RF device discussed above, an optical power of 1 mW, 10 mW, 25 mW, 50 mW or 100 mW may be adequate to generate sufficient charge carrier plasma to switch the RF transmission line from the first state to the second state. Such optical powers may be achievable using LED devices, which may be cheap, easily obtainable and safer to use. Higher optical-power devices (e.g., collimated lasers with optical powers of 1 W or greater) may be used and may provide for higher performance.

[0066] The RF device may also enable the light source to be operated in a continuous wave (CW) mode when activated. The low thickness of the active semiconductor layer and / or the good thermal coupling to the substrate may provide sufficient thermal dissipation to handle high-optical-power light sources. The strong coupling between the active region of the RF transmission line and the optically -induced charge carrier plasma may reduce the required photoconductance and optical power to switch the RF transmission line. Therefore, the light source, e.g., a laser, may be used in a continuous wave mode, rather than having to use a pulsed beam. This may unlock more application scenarios of the RF device.

[0067] According to a third aspect of the invention, there is provided a method of manufacturing a radio frequency, RF, device. The method comprises: bonding a semiconductor wafer to a substrate; thinning the semiconductor wafer such that the thickness of the semiconductor wafer is less than 100 microns; forming an RF transmission line on a surface (e.g., a front surface) of the thinned semiconductor wafer.

[0068] A plurality of RF transmission lines may be formed on the front surface of the thinned semiconductor wafer. Each of the RF transmission lines may correspond to a separate RF device. The method may further comprise dicing the stacked wafer (i.e., the plurality of RF transmission lines, the thinned semiconductor wafer, and the substrate) into a plurality of RF devices. This may allow a plurality of RF devices to be manufactured simultaneously on the same stacked wafer.

[0069] Bonding the semiconductor wafer to the substrate may comprise surface-activated bonding (SAB). During SAB, surfaces of the semiconductor wafer and the substrate are cleaned by employing fast atom bombardment (FAB) prior to bonding. Such a bonding method may be regarded as ‘permanent bonding’, and the substrate may form part of the final RF device.

[0070] Bonding the semiconductor wafer to the substrate may comprise using a dissolvable adhesive (e.g., a solvent-dissolvable adhesive). Such a bonding method may be regarded as ‘temporary bonding’, where the substrate is only provided to assist with manufacturing of the RF device and will be removed after manufacturing, the substrate not forming part of the final RF device.

[0071] SAB may be advantageous for the permanent bonding of the semiconductor wafer to the substrate. Compared to traditional permanent bonding techniques, SAB may not require as high temperatures during and / or after the bonding process to obtain sufficient bonding strength for the subsequent thinning of the semiconductor wafer. This may reduce purity degradation and cracking of the semiconductor material. SAB does not require any auxiliary intermediate layers and does not generate any unintentional intermediate layers (e.g. containing oxygen, hydrogen, and / or one or more hydroxy groups). This may reduce the influence of surface interactions at the rear surface of the semiconductor wafer and increase thermal conduction between the semiconductor wafer and the substrate compared to other permanent bonding techniques (where intermediate layers may make the passivation of the rear surface of the semiconductor wafer worse and function as a thermal insulator).

[0072] Although SAB may be preferred, it will be understood that any permanent bonding technique may be used. The use of permanent bonding may provide for a number of advantages. The use of semiconductor wafers may enable multiple RF devices to be manufactured from the same wafer, while the wafers themselves may be widely commercially available. Permanent bonding to a substrate may provide for easier handling of the semiconductor wafer during manufacturing and easier handling of the RF device during operation, as the semiconductor wafer and the RF device are supported by the substrate. This may enable the use of very thin active semiconductor layers (i.e., less than 100 nm), which may provide for the operational advantages discussed above.

[0073] Bonding the semiconductor wafer to the substrate may be temporary, the substrate not forming part of the final RF device (i.e., where the substrate is only provided to assist with manufacturing of the RF device and will be removed after manufacturing). Any suitable bonding technique may be used. This may involve the use of adhesives (which may be dissolved by solvents), for example.

[0074] Thinning the semiconductor wafer may comprise any suitable mechanical or chemical technique. The thinning may comprise mechanical grinding, chemical mechanical planarization (CMP), wet chemical etching, dry chemical etching (DCE), smart cut, or any suitable grinding, polishing, etching, or hybrid method, for example.

[0075] Forming the plurality of RF transmission lines may comprise: depositing a conducting layer on the surface; and using photolithography to pattern the conducting layer. Patterning the conducting layer may comprise performing a lift-off process, in which the photolithographic patterning is used to define a photoresist pattern on which the conducting layer is deposited, and the photoresist is lifted off to pattern the conducting layer.

[0076] Patterning the conducting layer may comprise performing an etch, masked by a photoresist layer patterned by the photolithographic process.

[0077] The conducting layer may comprise an electrically conducting material which may be a metal material, for example gold. The conducting layer may comprise at least two stacked transition metal materials (e.g., gold on titanium or copper on chromium) . The conducting layer may be between 0.5 and 5 microns thick. The conducting layer may be 1 micron, or 2.5 microns, or 5 microns thick, for example. The conducting layer may be deposited using any suitable thin-layer deposition technique, such as: physical vapour deposition; chemical vapour deposition; electroplating; or screen printing. A number of well-known processes exist for forming patterned conducting layers of suitable thickness - from printed circuit board manufacturing to integrated circuit manufacturing. Any suitable process can be used to form the plurality of RF transmission lines.

[0078] Dicing the stacked wafer (i.e., the plurality of RF transmission lines, the thinned semiconductor wafer, and the substrate) into a plurality of RF devices may comprise scribing and breaking, mechanical sawing, and / or laser cutting, for example.

[0079] Where the substrate does not form part of the final RF device, dicing will stop before reaching the substrate so that the substrate may be reused.

[0080] The method may further comprise modifying at least one surface of the semiconductor wafer so as to produce a passivation layer.

[0081] A rear passivation layer may be formed on the rear surface of the semiconductor wafer prior to bonding to the substrate. A front passivation layer may be formed on the front surface of the semiconductor wafer after thinning the semiconductor wafer, and before forming the plurality of RF transmission lines. Modifying the semiconductor wafer may comprise growing a passivation layer on the front and / or rear surfaces of the active semiconductor layer by oxidising the semiconductor (e.g., by thermal oxidation). For example, where the wafer is a silicon wafer, the front and / or rear passivation layers may comprise silicon oxide. Modifying the semiconductor wafer may comprise depositing a passivation layer - for example a silicon oxide layer may be deposited on silicon, rather than grown by thermal oxidation of silicon. Any suitable deposition technique, for example PECVD or ALD, may be used to deposit a passivation layer upon the front and / or rear surfaces of the semiconductor wafer. The front and / or rear passivation layers may be an aluminium oxide layer, for example (which may be deposited by ALD).

[0082] The method may further comprise selectively doping a region of the semiconductor wafer to contain deep-level impurities, said doping comprising implanting the semiconductor wafer via an ion beam.

[0083] The doped region of the semiconductor wafer may form a non-photoconductive / high resistivity region. The dopant may be vanadium (where the semiconductor material is silicon in particular) or chromium (where the semiconductor material is gallium arsenide), for example. Other methods for form a non-photoconductive / high resistivity region of the semiconductor wafer may be used. However, compared to other methods (e.g., selectively etching a region of the semiconductor wafer with chemicals), selectively doping a region of the semiconductor wafer with deep-level impurities may provide for greater reliability of following manufacture, a lower insertion loss of the RF device, and / or better lateral surface passivation of the undoped / unetched region of the semiconductor wafer, for example.

[0084] The method may be used to manufacture an RF device according to the first aspect of the invention.

[0085] BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Embodiments of the invention will be described, purely by way of example, with reference to the accompanying drawings, in which:

[0087] Figure 1 is a schematic diagram of a cross-section of an example RF device; Figure 2 is a schematic diagram of an example RF system;

[0088] Figure 3 is a schematic diagram of an example RF circuit;

[0089] Figure 4 is a schematic diagram of an example RF circuit;

[0090] Figure 5 is a schematic diagram of a top-down view of an example RF device;

[0091] Figure 6 is a schematic diagram of an example RF system;

[0092] Figure 7 is a schematic diagram of a top-down view of an example RF device with a circular, meandered RF transmission line structure;

[0093] Figure 8 is a schematic diagram of a cross -section of an example RF device with an RF transmission line structure like that of Figure 7;

[0094] Figure 9 is a schematic diagram of a top-down view of an example RF device with a circular, periodic-interdigitated-gap-loaded RF transmission line structure;

[0095] Figure 10 is a schematic diagram of a cross -section of an example RF device with an RF transmission line structure like that of Figure 9; and

[0096] Figure 11 shows the transmission coefficient and insertion phase shift data for an example RF device in both the unilluminated and the illuminated states .

[0097] DETAILED DESCRIPTION

[0098] Referring to Figure 1 , a schematic diagram of a cross-section of an example RF device 100 is shown. The RF device 100 corresponds to a series configuration. The RF device 100 comprises a plurality of different layers. The RF device 100 comprises an active semiconductor layer 110, said layer 1 10 having a front surface 112 and rear surface 114. The front surface 112 and rear surface 114 define a front and rear side of the RF device 100 respectively. The active semiconductor layer 110 comprises an intrinsic semiconductor material (e.g., intrinsic silicon) - so has a high resistivity when it is not exposed to light. When the active semiconductor layer 110 is exposed to sufficient intensity of light of an appropriate wavelength (with sufficient energy to excite electrons into the conduction band), a high density of charge carriers can be generated - enough to lower its resistivity so that it is a sufficient electrical conductor.

[0099] On the front surface of the active semiconductor layer 110 is disposed an RF transmission line 120. In the example of Figure 1 , the RF transmission line 120 comprises a first element 120a and a second element 120b (the outmost portions of which, 122a and 122b, may also be referred to as a first port and a second port of the RF transmission line 120 (or the RF device 100), respectively), said elements being separated from one another by a gap. The portions of the first and second elements, 120a and 120b, that are adjacent to the gap, and the gap itself, comprise an active region 121 of the RF transmission line 120. When light is incident upon the rear side of the RF device 100, a charge carrier plasma is generated within the active semiconductor layer 110. The charge carrier plasma creates a conducting region within the active semiconductor layer 110 that couples to the RF transmission line 120, thereby altering an RF transmission property of the active region 121 of the RF transmission line 120 (and thus altering the transmission of an RF signal across the RF transmission line 120 as a whole). The active region 121 of the RF transmission line 120 may be regarded as the portion of the RF transmission line 120 around the gap that couples to the optically - induced charge carrier plasma, which will be dependent on the properties of the active semiconductor layer 110 and the light source. The optically -induced charge carrier plasma may reduce the resistance between the first element 120a and the second element 120b, thereby allowing current to flow therebetween and allowing an RF signal to pass between the ports 122a, 122b. The RF transmission line 120 comprises an electrically conducting material, for example a transition metal material such as gold, copper or aluminium. The RF transmission line 120 may comprise at least one adhesion material, for example a transition metal material such as chrome, titanium, or tungsten.

[0100] The active semiconductor layer 110 may comprise a non-photoconductive region 116, which may be formed by deep-level impurity doping of the active semiconductor layer 110. The non-photoconductive region 116 encompasses a volume of the active semiconductor layer 110 below the active region 121 , thereby confining the optically - induced charge carrier plasma to this volume.

[0101] A front passivation layer 130 and a rear passivation layer 132 are provided on the front surface 112 and the rear surface 114 of the active semiconductor layer 110 respectively. At the rear side of the active semiconductor layer 110 (e.g., on a rear surface of the rear passivation layer 132), a substrate 140 is provided (the substrate 140 comprising a large bandgap material that is transparent to light from the light source). The substrate 140 may provide a base onto which the other layers are deposited and / or arranged.

[0102] At the rear-most side of the RF device 100, coupled to the rear surface of the substrate 140, is an AR layer 150. The AR layer 150 may be disposed across all or only a part of the rear surface of the substrate 140, but is at least across the area below the active region 121 , such that reflection of light incident from the rear side of the RF device 100 onto the active region 121 is reduced. A rear ground plane 160 is also provided at the rear-most side of the RF device 100, coupled to the rear surface of the substrate 140. The rear ground plane 160 may not be provided in the area below the active region 121 , such that light incident from the rear side of the RF device 100 onto the active region 121 is not reflected. The rear ground plane 160 comprises an electrically conducting material, for example a transition metal material such as gold, copper or aluminium. The rear ground plane 160 may comprise at least one adhesion material, for example a transition metal material such as chrome, titanium, or tungsten.

[0103] At a front-most side of the RF device 100, there is provided a reflective layer 170 (e.g., a dielectric mirror layer). The reflective layer 170 is provided at the front surfaces of the front passivation layer 130 and the RF transmission line 120. Thus, light incident from the rear side of the RF device 100 that is not absorbed by the active semiconductor layer 110 may be reflected back, thereby increasing optical absorption within the active semiconductor layer 110. The RF transmission line 120 may itself reflect light incident from the rear side of the RF device 100 back into the active semiconductor layer 110. The reflective layer 170 may only be functional in the areas of the front surface of the front passivation layer 130 that are not covered by the RF transmission line 120. The existence of the other parts of the reflective layer 170 is mainly due to ease of manufacturing. The reflective layer 170 may improve the interaction of light with the active region 121 of the RF transmission line 120. It will be understood from the present disclosure that some of the above layers may be omitted. Generally, the RF device 100 comprises an active semiconductor layer 110 comprising a front surface 112 and a rear surface 114, and an RF transmission line 120 coupled to the front surface 112 of the active semiconductor layer 110.

[0104] Referring to Figure 2, a schematic diagram of an example RF system 200 is shown. The RF system 200 comprises an RF device 100, with features like that of the RF device 100 shown in Figure 1. The RF device 100 corresponds to a series configuration. The RF device 100 generally comprises an active semiconductor layer 110, said layer comprising a front surface 112 and a rear surface 114. An RF transmission line 120 is coupled to the active semiconductor layer 110 at the front surface 112.

[0105] The RF transmission line 120 comprises a first element 120a and a second element 120b (the outmost portions of which, 122a and 122b, may also be referred to as a first port and a second port of the RF transmission line 120 (or the RF device 100), respectively), said elements being separated from one another by a gap. The portions of the first and second elements, 120a and 120b, that are adjacent to the gap, and the gap itself, comprise an active region 121 of the RF transmission line 120.

[0106] The RF system 200 further comprises a light source 210, the light source 210 being selectively controlled by a controller 212 (e.g., a waveform generator). When the light source 210 is activated, a light beam 211 is produced. The light source 210 may be a laser, in which case the light beam 211 may be a collimated, monochromatic laser beam, for example. Alternatively, the light source 210 may be a monochromatic LED. Additional optical elements such as fibres, lenses, filters and / or mirrors (not shown) may be provided to collimate, focus, and / or redirect the light beam 211.

[0107] The light beam 211 is incident upon the rear surface 114 of the active semiconductor layer 110 (i.e., the surface opposite to that on which the RF transmission line 120 is disposed). Illumination from the rear side may be advantageous (e.g., it may be more efficient as light is not partially reflected by the RF transmission line 120). However, it will be understood that the RF device 100 may also be operated using illumination from the front side. The incident light beam 211 is at least partially absorbed by the active semiconductor layer 110, thereby generating charge carriers (due to valence electrons being excited across the semiconductor bandgap to the conduction band). The optical excitation generates a plasma of charge carriers (i.e., pairs of free electron and free hole), resulting in a conducting region 111 within the active semiconductor layer 110. The active semiconductor layer 110 comprises an intrinsic semiconductor material with a relatively long diffusion length and has a thickness d that is sufficiently thin so charge carriers generated by the rear-side illumination are able to interact with the RF transmission line 120 - i.e., the conducting region 111 produced near to the rear surface 114 extends sufficiently close to the RF transmission line 120 at the front surface 112.

[0108] The RF device 100 of the system 200 may be an optoelectronic RF series switch. The RF device 100 may be used as a component in a wider RF circuit, in which the RF device 100 is integrated with other RF devices, RF sources, and / or RF loads. The RF circuit may be an optoelectronic multiple-pole-multiple-throw RF switch, an optoelectronic variable RF attenuator, an optoelectronic tuneable RF filter, an optoelectronic tuneable RF phase shifter, or an optoelectronic tuneable RF antenna, a radar system, a communications system, or an RF test instrumentation, for example. In the example of Figure 2, an RF source 220 is connected to a first port 122a of the RF device 100 and an RF load 221 is connected to a second port 122b of the RF device 100 (the RF source 220 and load 221 may be a 50 source and load, for example), which forms an RF circuit. When the light source 210 is activated so as to produce the conducting region 111 , the optically -induced charge carriers may carry charge between the first port 122a and the second port 122b, thereby completing the RF circuit .

[0109] The RF device 100 may be packaged or integrated into a housing or package. The package may help protect the RF device 100 (from physical damage and also electrical and / or optical interference). The package may include one or more ports that assist in coupling of the RF source 220, the RF load 221, and / or the light source 210 to the RF device 100. The housing may have a pair of input and output ports coupled to the first and second ports 122a, 122b of the RF device 100. The housing may have an input port for receiving the light beam 211 , e.g., comprising a connector which may comprise a screw thread or bayonet coupler for coupling to a fibre optic. In some embodiments the packaged device may include a light source 210 (e.g., an LED or VCSEL) arranged to produce light to switch the RF device 100 on receipt of an external control signal. Where the light source 210 comprises a front cathode (or anode) contact (e.g., a bare-die VCSEL) and the RF device 100 comprises a rear ground plane 160, the front cathode (or anode) contact of the light source 210 may be permanently bonded to the rear ground plane 160.

[0110] Referring to Figure 3, another embodiment of an RF circuit is shown. The circuit comprises an RF device 100 like that describe above with regard to Figure 1. However, unlike the circuit shown in Figure 2 (where the RF source 220 and RF load 221 are connected across the ports 122a, 122b of the RF device 100 in series), the RF load 220 and RF load 221 are both connected to the first port 122a of the RF device 100. The second port of the RF device 100 is connected an RF short 230. When the light source 210 is off the RF circuit may be complete ; i.e., an RF signal may flow between the RF source 220 and the RF load 221. However, illumination of the RF device 100 via light beam 211 causes the RF signal to be redirected to the RF short 230 via the optically- induced conducting region 111.

[0111] An RF circuit may comprise two RF devices 100 like that described above with regard to Figure 1. Referring to Figure 4, an RF circuit comprising two RF devices 100a, 100b is shown. Each RF device 100a, 100b is provided with a separate control mean respectively, such as a light source 210a, 210b and a controller 212a, 212b (though said control means may still be interconnected - e.g., only one light source 210a, 210b may be turned on at one time). The two RF devices 100a, 100b are provided with a common RF source 220. Each RF device 100a, 100b is provided with a separate RF load 221a, 221b respectively. When the first light source 210a is activated to illuminate the first RF device 100a with light beam 211a (the second light source 210b is deactivated) , an RF signal may flow between the RF source 220 and the first RF load 221a via the optically-induced conducting region 211a. Likewise, illumination of the second RF device 100b by the second light source 210b (the first light source 210a is deactivated) may allow an RF signal to flow between the RF source 220 and the second RF load 221b (not shown).

[0112] The skilled person will understand that any number of RF devices 100 may be coupled in parallel and controlled accordingly, which may form an optoelectronic single-pole- multiple-throw RF switch. If there are more than one RF source 220, the optoelectronic single-pole-multiple-throw RF switch may be modified into an optoelectronic multiple- pole-multiple-throw RF switch. The skilled person would also understand that any number of RF devices 100, RF sources 220, and RF loads 221 may be interconnected in any topological configurations (each RF device 100 provided with a separate control mean respectively), which allow more sophisticated RF signal routing.

[0113] Referring to Figure 5, a schematic top-down view of an example RF device 102 is shown. The RF device 102 is identical to the RF device 100 shown in Figure 1 , except for the configuration of the RF transmission line 120 (shown in lighter grey). The RF device 102 corresponds to a shunt configuration.

[0114] In the example of Figure 5, the RF transmission line 120 comprises a first element 120a and a second element 120b, said elements being separated from one another by a pair of gaps 120c. The RF transmission line 120 (or the RF device 102) comprises two ports 122a, 122b. The portions of the first and second elements, 120a and 120b, that are adjacent to the pair of gaps 120c and the pair of gaps 120c itself comprise a region of the RF transmission line 120, the part of which enclosed by an area corresponding to the area of the active semiconductor layer that is illuminated by the light beam when in use comprises an active region 121 of the RF transmission line 120.

[0115] When the RF device 102 is not illuminated by an appropriate light source (e.g., an LED / laser, with a photon energy greater than the bandgap of the semiconductor material), an RF signal provided across the first element 120a will flow between the ports 122a, 122b (with an acceptably small insertion loss). Essentially, in this first, nonilluminated state, the RF device 102 functions as an RF transmission line disposed on an insulating substrate (the active semiconductor layer 110). However, when the RF device 102 is illuminated from the rear side (or front side) by an appropriate light source over the active region 121 of the RF transmission line 120, absorption of photons by the active semiconductor layer 110 gives rise to an optically -induced charge carrier plasma within the active semiconductor layer 110. The induced charge carriers are able to electrically couple the first element 120a to the second element 120b, thereby shunti ng the RF signal to ground. Essentially, in this second, illuminated state, the active semiconductor layer 110 in the illuminated area has a sufficiently high density of charge carriers that it acts as a sufficiently electrically conducting material, thereby altering the impedance between the components of the RF transmission line 120 and creating a conducting path between the first and second elements 120a, 120b, across the pair of gaps 120c. Referring to Figure 6, a schematic diagram of an example RF system 202 is shown. The RF system 202 comprises an RF device 102, with features like that of the RF device 102 shown in Figure 5. The RF device 102 corresponds to a shunt configuration. The RF device 102 generally comprises an active semiconductor layer 110, said layer comprising a front surface 112 and a rear surface 114. An RF transmission line 120 is coupled to the active semiconductor layer 110 at the front surface 112.

[0116] The RF transmission line 120 comprises a first element 120a and a second element 120b, said elements being separated from one another by a pair of g aps 120c. The RF transmission line 120 (or the RF device 102) comprises two ports 122a, 122b. The portions of the first and second elements, 120a and 120b, that are adjacent to the pair of gaps 120c and the pair of gaps 120c itself comprise a region of the RF transmission line 120, the part of which enclosed by an area corresponding to the area of the active semiconductor layer that is illuminated by the light beam 211 when in use comprises an active region 121 of the RF transmission line 120.

[0117] The RF system 202 further comprises a light source 210, the light source 210 being selectively controlled by a controller 212 (e.g., a waveform generator). When the light source 210 is activated, a light beam 211 is produced. The light source 210 may be a laser, in which case the light beam 211 may be a collimated, monochromatic laser beam, for example. Alternatively, the light source 210 may be a monochromatic LED. Additional optical elements such as fibres, lenses, filters and / or mirrors (not shown) may be provided to collimate, focus, and / or redirect the light beam 211.

[0118] Although the light beam 211 appears to be injected into the active semiconductor layer 110 laterally in Figure 6, it is incident upon the rear surface 114 of the active semiconductor layer 110 (i.e., like that of Figure 2). Illumination from the rear side may be advantageous (e.g., it may be more efficient as light is not partially reflected by the RF transmission line 120). However, it will be understood that the RF device 102 may also be operated using illumination from the front side. The incident light beam 211 is at least partially absorbed by the active semiconductor layer 110, thereby generating charge carriers (due to valence electrons being excited across the semiconductor bandgap to the conduction band). The optical excitation generates a plasma of charge carriers (i.e., pairs of free electron and free hole), resulting in a conducting region 111 within the active semiconductor layer 110. The active semiconductor layer 110 comprises an intrinsic semiconductor material with a relatively long diffusion length and has a thickness that is sufficiently thin so charge carriers generated by the rear -side illumination are able to interact with the RF transmission line 120 - i.e., the conducting region 111 produced near to the rear surface 114 extends sufficiently close to the RF transmission line 120 at the front surface 112.

[0119] The RF device 102 may be an optoelectronic RF shunt switch. The RF device 102 may be used as a component in a wider RF circuit, in which the RF device 102 is integrated with other RF devices, RF sources, and / or RF loads. The RF circuit may be an optoelectronic multiple-pole-multiple-throw RF switch, an optoelectronic variable RF attenuator, an optoelectronic tuneable RF filter, an optoelectronic tuneable RF phase shifter, or an optoelectronic tuneable RF antenna, a radar system, a communications system, or an RF test instrumentation, for example. In the example of Figure 6, an RF source 220 is connected to a first port 122a of the RF device 102 and an RF load 221 is connected to a second port 122b of the RF device 102 (the RF source 220 and load 221 may be a 50 source and load, for example), which forms an RF circuit. When the light source 210 is activated so as to produce the conducting region 111, the optically -induced charge carriers may carry charge between the first port 122a and the second port 122b, thereby completing the RF circuit. In addition, the RF device 102 may be integrated with other RF devices, RF sources, and / or RF loads to form wider RF circuits like those of Figure 3 and Figure 4.

[0120] Likewise, the RF device 102 and / or the light source 210 may be packaged or integrated into a housing or package as for the RF device 102.

[0121] Referring to Figure 7, a schematic top-down view of an example RF device 104 is shown. The RF device 104 corresponds to a shunt configuration; said RF device 104 may be used as an optoelectronic RF shunt switch. The optoelectronic RF shunt switch may be used as a component in a wider RF circuit, in which the optoelectronic RF shunt switch is integrated with other RF devices, RF sources, and / or RF loads. The RF circuit may be an optoelectronic multiple-pole-multiple-throw RF switch, an optoelectronic variable RF attenuator, an optoelectronic tuneable RF filter, an optoelectronic tuneable RF phase shifter, an optoelectronic tuneable RF antenna, a radar system, a communications system, or an RF test instrumentation, for example. The RF device 104 comprises an RF transmission line 120 (shown in lighter grey) comprising a first element 120a and a second element 120b , said elements being separated from one another by a pair of gaps 120c. The RF transmission line 120 (or the RF device 104) has two ports 122a, 122b. The RF transmission line 120 is coupled to a front surface of an active semiconductor layer 110 (shown in darker grey). The RF transmission line 120 comprises an electrically conducting material, for example a metal material such as gold. A conducting layer is patterned to define the first element 120a and the second element 120b.

[0122] In the example of Figure 7, the active region 121 of the RF transmission line 120 comprises a circular meandered RF transmission line structure comprising a circular meandered signal line 123a separated from a front ground plane 123b by a pair of circular meandered gaps 123c. Part of the first element 120a of the RF transmission line

[0123] 120 is configured to form a circular meandered signal line 123a. Part of the second element 120b of the RF transmission line 120 forms a front ground plane on both sides of the circular meandered signal line 123a. The circular meandered signal line 123a, front ground plane 123b and pair of circular meandered gaps 123c form the active region

[0124] 121 of the RF transmission line 120.

[0125] When the RF device 104 is not illuminated by an appropriate light source (e.g., an LED / laser, with a photon energy greater than the bandgap of the semiconductor material), an RF signal provided across the circular meandered signal line 123a will flow between the ports 122a, 122b (with an acceptably small insertion loss). Essentially, in this first, non-illuminated state, the RF device 104 functions as an RF transmission line disposed on an insulating substrate (the active semiconductor layer 110). However, when the RF device 104 is illuminated from the rear side (or front side) by an appropriate light source over the circular active region 121 , absorption of photons by the active semiconductor layer 110 gives rise to an optically -induced charge carrier plasma within the active semiconductor layer 110. The induced charge carriers are able to electrically couple the circular meandered signal line 123a to the front ground plane 123b, thereby shunting the RF signal to ground. Essentially, in this second, illuminated state, the active semiconductor layer 110 in the illuminated area has a sufficiently high density of charge carriers that it acts as a sufficiently electrically conducting material, thereby altering the impedance between the components of the RF transmission line 120 and creating a conducting path between the circular meandered signal line 123a and the front ground plane 123b, across the pair of circular meandered gaps 123c.

[0126] The RF transmission line 120, or at least a part thereof, may be configured so that the shape of the active region 121 of the RF transmission line 120 substantially matches an area of the active semiconductor layer 110 that is illuminated. The area that is illuminated (particularly in the case that the light source is a focused LED, a fibre - coupled EEL, or a single-aperture VCSEL) may be a circle, as depicted in Figure 7 as the dashed circle. The active region 121 of the RF transmission line 120 may be configured to substantially match the said area; as shown in Figure 7, the circular meandered signal line 123a, the front ground plane 123b, and the pair of circular meandered gaps 123c are shaped so as to criss-cross the dashed circle (or the said area). The optically-induced charge carrier plasma will mostly subsist in the area that is illuminated. Therefore, the coupling between the active region 121 of the RF transmission line 120 and the optically-induced charge carrier plasma may be maximised, which results in a more efficient RF device 104. If the light beam area were larger than the active region 121 of the RF transmission line 120, some of the induced plasma would be wasted in regions of the active semiconductor layer 110 that do not enable switching. By contrast, if the light beam area were smaller than the active region 121 of the RF transmission line 120, some parts of the active region 121 would not switch from the first state to the second state (as they would not be coupled to the plasma).

[0127] Although Figure 7 shows a top-down view of the RF device 104, illustrating the stratified RF transmission line / active semiconductor layer pattern, it will be understood that said RF device 104 may comprise all / some of the additional layers shown in Figure 1. In a preferred example of the shunt configuration, the RF device 104 comprises: i) a 5 micron (or 1 to 10 micron) thick RF transmission line, with a pattern as shown in Figure 7. The active region of the transmission line has a diameter of 600 microns. The RF transmission line may comprise gold and / or titanium (i.e. gold on a titanium adhesion layer) , for example; ii) a 75 micron (or 50 to 100 micron) thick active semiconductor layer comprising intrinsic silicon; iii) a non-photoconductive region encompassing a volume of the active semiconductor layer beneath the active region of the RF transmission line. In the example embodiment the non-photoconductive region may have an inner diameter of 600 microns. The non-photoconductive region may be doped with vanadium, for example; iv) a 1.3 nm (or 0.5 to 30 nm) thick front passivation layer and / or a 1.3 nm (or 0.5 to 30 nm) thick rear passivation layer. The passivation layers may comprise aluminium oxide, for example; v) no additional substrate; and vi) a 50 to 500 nm thick AR layer. In the example embodiment the AR layer may has a diameter of 600 microns. The AR layer may comprise PECVD silicon oxide, with a thickness of 146 nm (or 130 to 160 nm). A 5 micron (or 1 to 10 micron) thick rear ground plane may be provided around the AR layer, the rear ground plane for example comprising gold and / or titanium (e.g., gold on a titanium adhesion layer) with an internal diameter of 600 microns.

[0128] Said RF device 104 may be used with an 850 nm (or 800 to 900 nm) wavelength collimated Gaussian beam, for example. Said RF device 104 may be used with a collimated single-aperture VCSEL with a wavelength of 850 nm (or 800 to 900 nm) and a Gaussian beam which has an optical power of 7 mW (or 5 to 9 mW) and a beam waist of 150 microns (or 100 to 200 microns).

[0129] Referring to Figure 8, a schematic diagram of a cross -section of an embodiment of an RF device 104 comprising a circular meandered RF transmission line structure, like that of Figure 7, is shown. The cross-section of the RF device 104 is taken along line a, as shown inset. Features with reference numerals correspond similarly to those of Figure 1 and 7.

[0130] The circular meandered RF transmission line structure comprises a circular meandered signal line 123a and a front ground plane 123b, separated by a pair of circular meandered gaps 123c (i.e. , on the cross-sectional plane taken along line a). The RF transmission line 120 is disposed on a front surface 112 of an active semiconductor layer 110, which also has a rear surface 114. The active semiconductor layer 110 comprises a non - photoconductive region 116, said region configured to encompass an active region 121 of the RF transmission line 120 (i.e., the circular meandered signal line 123a, the front ground plane 123b, and the pair of circular meandered gaps 123c). A front passivation layer 130 is provided at the front surface 112 of the active semiconductor layer 110. A rear passivation layer 132, an AR layer 150 and a rear ground plane 160 are provided at the rear surface 114 of the active semiconductor layer 110.

[0131] Referring to Figure 9, a schematic top-down view of an example RF device 106 is shown. The RF device 106 corresponds to a series configuration; said RF device 106 may be used as an optoelectronic RF series switch. The optoelectronic RF series switch may be used as a component in a wider RF circuit, in which the optoelectronic RF series switch is integrated with other RF devices, RF sources, and / or RF loads. The RF circuit may be an optoelectronic multiple-pole-multiple-throw RF switch, an optoelectronic variable RF attenuator, an optoelectronic tuneable RF filter, an optoelectronic tuneable RF phase shifter, an optoelectronic tuneable RF antenna, a radar system, a communications system, or an RF test instrumentation, for example .

[0132] The RF device 106 comprises an RF transmission line 120 (shown in lighter grey) comprising a first interdigitated structure and a second interdigitated structure. The first interdigitated structure is electrically coupled to a first port 122a of the RF transmission line 120 (or the RF device 106), while the second interdigitated structure is electrically coupled to the second port 122b of the RF transmission line 120 (or the RF device 106). The RF transmission line 120 is coupled to a front surface of an active semiconductor layer 110 (shown in darker grey). The RF transmission line 120 comprises an electrically conducting material, for example gold or titanium. Gaps are left between the various components of the RF transmission line 120 so as to define a pattern which may be used for electronic circuitry.

[0133] In the example of Figure 9, the active region 121 of the RF transmission line 120 comprises a circular, periodic, interdigitated-gap-loaded RF transmission line structure comprising a first interdigitated structure and a second interdigitated structure being separated by a plurality of interdigitated gaps 126.

[0134] As shown inset, each of the first and second interdigitated structures comprise a plurality of vertical ‘trunks’ 124, off of which extend a plurality of horizontal ‘branches’ 125. The branches 125a of the first interdigitated structure are interdigitated with the branches 125b of the second interdigitated structure, the first and second interdigitated structures being separated by a plurality of interdigitated gaps 126. When the optoelectronic RF series switch is integrated into an RF circuit, RF components may be connected to both ports 122a, 122b such that the RF device 106 is in series, with the first and second interdigitated structures forming two separated terminals. The first interdigitated structure, the second interdigitated structure, and the plurality of interdigitated gaps 126 form the active region 121 of the RF transmission line 120.

[0135] When the RF device 106 is not illuminated by an appropriate light source (e.g., an LED / laser, with a photon energy greater than the bandgap of the semiconductor material), an RF signal provided across the ports 122a, 122b will be stopped. Essentially, in this first, non-illuminated state, the RF device 106 functions as an off or open switch. However, when RF device 106 is illuminated from the rear side (or the front side) by an appropriate light source, absorption of photons by the active semiconductor layer 110 gives rise to an optically-induced charge carrier plasma within the active semiconductor layer 110. The induced charge carriers are able to electrically couple the first plurality of branches 125a and second plurality of branches 125b, thereby allowing the RF signal to pass between the ports 122a, 122b. Essentially, in this second, illuminated state, the active semiconductor layer 110 that is in the illuminated area behaves as an electrically conducting material, thereby forming an on or closed switch; a conducting path is created between the first and second interdigitated structures across the plurality of interdigitated gaps 126.

[0136] As discussed above for Figure 7, the RF transmission line 120, or at least a part thereof, may be configured so that the shape of the active region 121 of the RF transmission line 120 substantially matches an area of the active semiconductor layer 110 that is illuminated. The area that is illuminated (particularly in the case that the light source is a focused LED, a fibre-coupled EEL, or a single-aperture VCSEL) may be a circle, as depicted in Figure 9 as the dashed circle. The active region 121 of the RF transmission line 120 may be configured to substantially match the said area; as shown in Figure 9, the first interdigitated structure, the second interdigitated structure, and the plurality of interdigitated gaps 126 are substantially circular so as to match the dashed circle (or the said area). The optically-induced charge carrier plasma will mostly subsist in the area that is illuminated. Therefore, the coupling between the active region 121 of the RF transmission line 120 and the optically -induced charge carrier plasma may be maximised, which may result in a more efficient RF device 106. Although Figure 9 shows a top-down view of the RF device 106, illustrating the stratified RF transmission line / active semiconductor layer pattern, it will be understood that said RF device 106 may comprise all / some of the additional layers shown in Figure 1. In a preferred example of the series configuration, the RF device 106 comprises: i) a 125 nm (or 50 to 1000 nm) thick RF transmission line, with a pattern as shown in Figure 9. The active region of the RF transmission line may have a diameter of 50 microns. The RF transmission line may comprise gold and / or titanium (e.g., gold on a titanium adhesion layer) , for example; ii) a 10 micron (or 8 to 12 microns) thick active semiconductor layer comprising intrinsic silicon; iii) a non-photoconductive region encompassing a volume of the active semiconductor layer beneath the active region of the RF transmission line. In the example embodiment the non-photoconductive region may has an inner diameter of 50 microns. The non-photoconductive region may be doped with vanadium, for example; iv) a 1.3 nm (or 0.5 to 30 nm) thick front passivation layer and / or a 1.3 nm (or 0.5 to 30 nm) thick rear passivation layer. The passivation layers may comprise aluminium oxide, for example; v) a 100 micron (or 75 to 125 microns) thick substrate comprising high-purity semi-insulating 4H-silicon carbide, for example; vi) a 50 to 500 nm thick AR layer. In the example embodiment the AR layer may has a diameter of 50 microns. The AR layer may comprise PECVD silicon oxide, with a thickness of 134 nm (or 120 to 150 nm). A 5 micron (or 1 to 10 micron) thick rear ground plane may be provided around the AR layer, the rear ground plane for example comprising gold and / or titanium (e.g., gold on a titanium adhesion layer) with an internal diameter of 50 microns; and vii) a reflective layer, e.g., a dielectric mirror layer. The dielectric mirror layer may comprise alternating SiOx / SiNxsub-layers with a total thickness of approximately 2 to 3 microns.

[0137] Said RF device 106 may be used with a 780 nm (or a 730 - 830 nm) wavelength Gaussian beam, for example. Said RF device 106 may be used with a single-aperture VCSEL with a wavelength of 780 nm (or a 730 - 830 nm) and a Gaussian beam which has an optical power of 20 mW (or 1 to 30 mW) and a beam waist of 6 microns (or 2 to 10 microns). Referring to Figure 10, a schematic diagram of a cross-section of an embodiment of an RF device 106 comprising a circular, periodic, interdigitated-gap-loaded RF transmission line structure, like that of Figure 9, is shown. The cross-section of the RF device 106 is taken along line b, as shown inset. Features with reference numerals correspond similarly to those of Figure 1 and 9.

[0138] The circular, periodic, interdigitated-gap-loaded RF transmission line structure comprises a first plurality of branches 125a and a second plurality of branches 125b, separated by an interdigitated gap 126 (i.e., on the cross-sectional plane taken along line b). For similarity, only 1 repeat is shown, but it will be understood that the RF transmission line structure may comprise many branches, as shown in Figure 9. The RF transmission line 120 is disposed on a front surface 112 of an active semiconductor layer 110, which also has a rear surface 114. The active semiconductor layer 110 comprises a non-photoconductive region 116, said region configured to encompass an active region 121 of the RF transmission line 120 (i.e., the first and second pluralities of branches 125a, 125b and the interdigitated gap 126). A front passivation layer 130 and a reflective layer 170 are provided at the front surface 112 of the active semiconductor layer 110. A substrate 140, an AR layer 150 and a rear ground plane 160 are provided at the rear surface 114 of the active semiconductor layer 110.

[0139] Referring to Figure 11 , experimental results for a device according to the example of Figures 7 and 8 is shown. A control voltage, a control current, and a DC power dissipation of 2.1 V, 12 mA, and 25.2 mW, respectively, are required to activate the aforementioned light source.

[0140] The left Y-axis of plot a) shows the measured transmission coefficients (insertion loss) of the unilluminated RF device across a range of 0 to 20 GHz, which is lower than 2.24 dB from 0-18 GHz. The right Y-axis of plot a) shows the measured transmission coefficients (isolation) of the illuminated RF device across the range of 0 to 20 GHz, which is higher than 17 dB over a frequency range of 0.2-2 GHz, is higher than 34.3 dB over a frequency range of 2-8 GHz, and is higher than 49.1 dB over a frequency range of 8-18 GHz.

[0141] Plot b) shows the measured insertion phase shift curves of the RF device across a range of 0 to 20 GHz for both the unilluminated and illuminated cases, both of which are very linear from 0-18 GHz and indicate low dispersion of the RF device in this frequency range.

[0142] The on-state input intercept point (IIP3) is typically greater than 100 dBm over a frequency range of 0-20 GHz, which indicates high on-state linearity of the RF device in this frequency range.

[0143] The average switching speed of the device is 15 microseconds; this comprises an on - switching time of 23 microseconds and an off-switching time of 6 microseconds. This shows a 65% improvement compared to thick-intrinsic-silicon-based devices of the prior art.

[0144] The device may be capable of handling 1 W of RF source power or greater over a frequency range of 0-20 GHz; no on-state compression was observed with 1 W of RF source power over a frequency range of 0-20 GHz.

[0145] In one example, a method of manufacturing an RF device, according to the present invention, where the substrate forms part of the final RF device, comprises the following steps:

[0146] 1. modifying the rear surface of a semiconductor wafer so as to produce a rear passivation layer;

[0147] 2. permanently bonding the rear surface of the rear passivation layer to a substrate;

[0148] 3. thinning the semiconductor wafer (using any suitable mechanical or chemical technique, or any combination of suitable mechanical and / or chemical techniques);

[0149] 4. selectively doping the thinned semiconductor wafer with deep -level impurities;

[0150] 5. modifying the front surface of the thinned semiconductor wafer so as to produce a front passivation layer;

[0151] 6. forming a plurality of RF transmission lines on the front surface of the front passivation layer;

[0152] 7. depositing a reflective layer (e.g., a dielectric mirror layer) on the front surface of the front passivation layer. The reflective layer also be deposited on the front surface of the plurality of RF transmission lines during this process. 8. thinning the substrate (using any suitable mechanical or chemical technique, or any combination of suitable mechanical and / or chemical techniques);

[0153] 9. depositing an AR layer on the rear surface of the substrate ;

[0154] 10. depositing or mounting a rear ground plane on the rear surface of the substrate;

[0155] 11. dicing the stacked wafer (i.e., the plurality of RF transmission lines, the thinned semiconductor wafer, the thinned substrate, and any other layers) into a plurality of RF devices.

[0156] It will be understood that some or all of these steps may be used during the manufacture of the said RF device. Generally, the method of manufacturing comprises steps 2, 3, 6, and 11 (i.e., bonding to a substrate, thinning the semiconductor wafer, forming a plurality of RF transmission lines, and dicing the stacked wafer). The RF device may not comprise a doped region, a front passivation layer, a rear passivation layer, an AR layer, a rear ground plane, and / or a reflective layer, for example.

[0157] Any suitable method may be used for modifying at least one surface of the semiconductor wafer so as to produce a passivation layer. For example, modifying at least one surface of the semiconductor wafer so as to produce a passivation layer may comprise depositing aluminium oxide on intrinsic silicon comprising: pre-annealing the semiconductor wafer (e.g., in O2 for 30 minutes at 1000 °C) using a tube furnace; removing any thermally-grown oxide layers using the RCA clean; depositing an aluminium oxide layer on a surface of the semiconductor wafer using ALD; post-annealing the semiconductor wafer (e.g., in O2 for 30 minutes at 460 °C) using a tube furnace; depositing another aluminium oxide layer on the other surface of the semiconductor wafer using ALD; post-annealing the semiconductor wafer (e.g., in O2 for 30 minutes at 460 °C) using a tube furnace.

[0158] Any suitable method may be used for permanently bonding the semiconductor wafer to the substrate. For example, permanently bonding the semiconductor wafer to the substrate may comprise SAB comprising: polishing the rear surface of the semiconductor wafer and the front surface of the substrate to mirror-like (e.g., arithmetic average roughness < 1 nm) using CMP; positioning the semiconductor wafer and the substrate in an ultra-high vacuum chamber (e.g., ambient pressure < 5x106Pa), the rear surface of the semiconductor wafer facing the front surface of the substrate ; cleaning the rear surface of the semiconductor wafer and the front surface of the substrate using FAB with an inert gas (e.g., argon) target ; depositing an auxiliary semiconductor layer on the front surface of the substrate and cleaning the front surface of auxiliary semiconductor layer using FAB with a semiconductor target and an inert gas (e.g., argon) target, respectively (i.e., this step being required only if the arithmetic average roughness of one to-be-bonded surface is greater than 5 nm, for example); bonding the semiconductor wafer to the substrate (with a bond temperature, a bond pressure, and a bond time of 23 °C, 3 MPa, and 3 minutes, respectively, for example).

[0159] Any suitable method of forming the plurality of RF transmission lines may be used (generally, depositing a conducting layer on the front surface of the front passivation layer and then patterning it, or depositing a conducting layer directly in a pattern). For example, forming the plurality of RF transmission lines may comprise a lift-off process comprising: depositing a photoresist layer on the front surface of the front passivation layer; exposing the photoresist layer to a defined optical pattern and / or a defined electron-beam pattern; developing a pattern of the photoresist layer with a photoresist developer; depositing a conducting layer on the front surfaces of the front passivation layer and the photoresist layer; and removing the photoresist layer with a photoresist remover to pattern the conducting layer (thereby forming the plurality of RF transmission lines).

[0160] In other embodiments, the conducting layer may be patterned using an etch that is masked by a photolithographically defined masking layer. Any suitable method may be used for selectively doping the thinned semiconductor wafer with deep-level impurities (e.g., via the use of an ion beam, masked by a patterned layer comprising silicon oxide for example).

[0161] The deep-level impurities for selectively doping intrinsic silicon may be vanadium, for example. The deep-level impurities for selectively doping intrinsic gallium arsenide may be chromium, for example. The temperature, the ion energy, and the implantation dose for selectively doping a semiconductor wafer comprising intrinsic silicon with vanadium may be 300 °C, 500 keV, and IxlO12cm2, respectively, for example.

[0162] The substrate may not form part of the final RF device. Instead, the substrate may be provided only for support during the manufacturing of the RF device. Where the substrate does not form part of the final RF device the bonding of the semiconductor wafer to the substrate may be regarded as ‘temporary’ . When the substrate does not form part of the final RF device, some changes may be made to the method of manufacturing discussed above: a. step 9 and step 10 are moved backward to be subsequent to step 1 ; b. bonding to a substrate is temporary in step 2 ; c. dicing the stacked wafer in step 11 stops before reaching the substrate; and d. a step 12 is added as removing the substrate.

[0163] In this case, the method of manufacturing generally comprises steps 2, 3, 6, 11, and 12. The RF device may not comprise a doped region, a front passivation layer, a rear passivation layer, an AR layer, a rear ground plane, and / or a reflective layer, for example.

[0164] Any suitable method may be used for temporarily bonding the semiconductor wafer to the substrate. For example, temporarily bonding the semiconductor wafer to the substrate may comprise the use of a dissolvable adhesive comprising: spin-coating a dissolvable adhesive layer on the rear surface of the semiconductor wafer; baking the semiconductor wafer (with a bake temperature and a bake time of 160 °C and 9 minutes, respectively, for example) ; the two above-stated steps may be conducted in turn for a second time; bonding the semiconductor wafer to the substrate (with a bond temperature, an ambient pressure, a bond pressure, and a bond time of 200 °C, 5 mbar, 0.055 MPa, and 3 minutes, respectively, for example). Any suitable method may be used for removing the substrate. For example, removing the substrate may comprise soaking the stacked wafer in a chemical bath containing solvents for a few hours.

[0165] Although specific examples have been described, the skilled person will appreciate that variations are possible, within the scope of the invention, which should be determined with reference to the accompanying claims.

Claims

CLAIMS1. A radio frequency, RF, device comprising: an active semiconductor layer comprising a front surface and a rear surface ; and an RF transmission line coupled to the front surface of the active semiconductor layer, the RF transmission line comprising an active region, the active region causing the RF transmission line to comprise a first state in which the RF transmission line has a first radio frequency transmission property, and a second state in which the RF transmission line has a second radio frequency transmission property , wherein the RF transmission line is selectively switchable between the first state and the second state in response to illumination of the active semiconductor layer by a light source when in use, and wherein the active semiconductor layer comprises an intrinsic semiconductor material with a thickness, between the front surface and the rear surface, of less than 100 microns.

2. The RF device of any preceding claim, wherein the active semiconductor layer comprises: silicon; and / or gallium arsenide.

3. The RF device of any preceding claim, further comprising at least one of: a front passivation layer coupled to the front surface of the active semiconductor layer; and a rear passivation layer coupled to the rear surface of the active semiconductor layer.

4. The RF device of claim 3, wherein the front passivation layer and / or the rear passivation layer comprises silicon oxide or aluminium oxide.

5. The RF device of any preceding claim, further comprising a substrate disposed at the rear surface of the active semiconductor layer, the substrate comprising a large bandgap material that is transparent to light from the light source.

6. The RF device of claim 5, wherein the substrate comprises: glass; sapphire; silicon carbide; aluminium nitride; gallium nitride; and / or diamond.

7. The RF device of any preceding claim, further comprising an anti-reflection layer disposed at the rear surface of the active semiconductor layer.

8. The RF device of any preceding claim, further comprising a rear ground plane disposed at the rear surface of the active semiconductor layer.

9. The RF device of any preceding claim, further comprising a reflective layer disposed at the front surface of the active semiconductor layer, the reflective layer being reflective to light incident from the direction of the rear surface of the active semiconductor layer.

10. The RF device of any preceding claim, wherein the active semiconductor layer comprises a region which is doped with deep-level impurities, the doped region encompassing a volume of the active semiconductor layer beneath the active region of the RF transmission line.

11. The RF device of claim 10, wherein the doped region is non-photoconductive.

12. The RF device of any preceding claim, wherein the active region of the RF transmission line is shaped so as to substantially conform to an area corresponding to the area of the active semiconductor layer that is illuminated by the light source when in use.

13. The RF device of claim 12, wherein the active region of the RF transmission line is configured to be substantially circular in shape.

14. The RF device of any preceding claim, wherein the active region of the RF transmission line comprises a meandered RF transmission line structure comprising: a meandered signal line; and a front ground plane separated from and surrounding the meandered signal line by a pair of meandered gaps.

15. The RF device of any of claims 1 to 13, wherein the active region of the RF transmission line comprises an interdigitated-gap-loaded RF transmission line structure comprising a first part and a second part, the first part and second part being separated by at least one interdigitated gap.

16. A radio frequency, RF, system comprising: an RF device according to any of claims 1 to 15; a light source configured to illuminate the active semiconductor layer; a controller configured to selectively activate the light source such that the RF transmission line of the RF device is selectively switchable between the first state and the second state.

17. The RF system of claim 16, wherein the light source is configured to illuminate the active semiconductor layer from the rear surface of the active semiconductor layer.

18. The RF system of claim 16 or 17, wherein the light source is monochromatic and configured to have a photon energy at least equal to the bandgap of the semiconductor material.

19. The RF system of any of claims 16 to 18, wherein the light source comprises: a light emitting diode, LED; or a laser.

20. A method of manufacturing a radio frequency, RF, device, the method comprising: bonding a semiconductor wafer to a substrate; thinning the semiconductor wafer such that the thickness of the semiconductor wafer is less than 100 microns;forming an RF transmission line on a surface of the thinned semiconductor wafer.

21. The method of claim 20, wherein bonding the semiconductor wafer to the substrate comprises: surface-activated bonding, the bonded substrate forming part of the final RF device; or the use of a dissolvable adhesive.

22. The method of claim 20 or 21 , wherein forming an RF transmission line comprises: depositing a conducting layer on the surface; and using photolithography to pattern the conducting layer.

23. The method of any of claims 20 to 22, further comprising modifying at least one surface of the semiconductor wafer so as to produce a passivation layer.

24. The method of any of claims 20 to 23, further comprising selectively doping a region of the semiconductor wafer to contain deep -level impurities, said doping comprising implanting the semiconductor wafer via an ion beam.

25. The method of any of claims 20 to 24, wherein the method is used to manufacture an RF device according to any of claims 1 to 15.

Images