HYBRID III-V / SILICON DEVICE, DIRECTIONAL COUPLER AND METHOD FOR TUNING A DIRECTIONAL COUPLER

Abstract

A hybrid III-V / silicon device comprising: a first silicon layer disposed over a buried oxide layer, BOX layer (832), the first silicon layer comprising: a first doped region with a first trench (813), a second doped region with a second trench (814), a first gap region (836) disposed between the first doped region and the second doped region; a first oxide layer (815) disposed over the first doped region and the second doped region; a first mesa (821) disposed on the first doped region of the first silicon layer, the first mesa comprising: a first Group III-V layer (807) disposed over the first oxide layer; a first optically active region (825) disposed over the first group III-V layer; and a second group III-V layer (827) disposed over the optically active region; and a second mesa (822) disposed on the second doped region of the first silicon layer, the second mesa comprising: a third Group III-V layer (808) disposed over the first oxide layer; a second optically active region (826) disposed over the third group III-V layer; and a fourth group III-V layer (828) disposed over the second optically active region.

Description

background

[0001] Parity-time symmetric (PT-symmetric) directional couplers require a balance between the gain and loss of the upper and lower waveguides to achieve efficient optical switching. Simultaneous tuning of the gain and loss elements results in a refractive index difference, which can affect the power cancellation of the optical switch. By adding an independent hybrid metal-semiconductor capacitor (MOSCAP), we have an additional control for tuning the refractive index (optical phase) and achieve improved power cancellation.

[0002] In current configurations, there is no way to control gain / loss and phase difference between the two branches independently.

[0003] WO 2017 / 123 245 A1 relates to a hybrid multilayer device and methods for producing a hybrid multilayer device.

[0004] The present invention relates to a hybrid III-V / silicon device according to independent claim 1, a directional coupler according to claim 12, and a method for tuning a directional coupler. Embodiments are subject to the respective dependent claims. Brief description of the drawings

[0005] The present disclosure will be described in detail according to one or more different embodiments with reference to the following figures. The figures are for illustrative purposes only and show merely typical or exemplary embodiments. Fig. 1 shows an example of a directional coupler. The Fig. 2A-2C show the propagation of optical waves in exemplary directional couplers. The Fig. 3A-3C show the transmissions of an exemplary parity-time symmetric (PT-symmetric) directional coupler at different refractive index differences between a gain waveguide and a loss waveguide. The Fig. Figures 4A to 4D show example thermal maps illustrating the transmission intensity matrix for an example PT-symmetric directional coupler with a difference in refractive index between the gain and loss waveguides. The Fig. Figures 5A to 5D show example thermal maps illustrating the transmission intensity matrix for another example of a PT-symmetric directional coupler with a different difference in refractive index between the gain and loss waveguides. The Fig. 6A and Fig. 6B show an example of a metal oxide semiconductor (MOS) optical modulator that can be implemented in the technology disclosed herein. The Fig. 7A and Fig. 7B show simulated relationships between the power supplied to the optical modulator of the Fig. 6A and Fig. 6B applied voltage for different dielectric thicknesses plotted against a change in modal refractive index and a change in FCA. Fig. Figure 8A shows a top view of an exemplary directional coupler according to implementations disclosed herein. Fig. Figure 8B shows a cross-sectional view of an exemplary directional coupler along line BB' of Fig. 8A. Fig. Figure 8C shows a cross-sectional view of another example of a directional coupler along the line BB' of Fig. 8A. Fig. Figure 9 shows the migration of the charge carrier concentration within the example directional coupler of Fig. 8A-8C according to an implementation disclosed herein. Fig. 10 is an exemplary flowchart illustrating an exemplary method for tuning a directional coupler according to the implementations disclosed herein. Fig. 11 is an example of a computer system that may be used to implement various features of the various optical devices of the present disclosure.

[0006] The figures are not exhaustive and do not limit the present disclosure to the precise form disclosed. Detailed description

[0007] The examples presented here relate to devices and methods for fabricating parity-time symmetric (PT-symmetric) directional couplers. PT-symmetric directional couplers require a balanced gain-loss ratio between the waveguides of the directional coupler to enable efficient optical switching. Accordingly, the optical losses of a first waveguide can be modulated separately from the optical gain of a second waveguide to allow tuning of the gain-loss ratio. The implementations described here also enable modulation of the optical gain of the second waveguide.

[0008] However, tuning the optical gain and / or optical loss can lead to variations in the refractive indices of the waveguides, which can degrade the performance of the directional coupler. Therefore, the implementations included here integrate hybrid metal-oxide-semiconductor capacitors (MOSCAPs) into one or both waveguides, allowing separate and simultaneous tuning of the refractive index. By adjusting the gain-to-loss ratio and the refractive indices, the implementations described here provide PT-symmetric directional couplers that can be used for efficient optical switching.

[0009] Examples of hybrid MOSCAPs used by the implementations disclosed herein may be formed by a thin dielectric layer between two semiconductor layers of different materials, each of which is formed in a waveguide of the implementations disclosed herein. In one example, the dielectric layer is an interfacial oxide layer between a layer of a III-V compound semiconductor material (referred to herein as a Group III-V material) and a silicon (Si) layer or other Group IV compound semiconductor layer (referred to herein as a Group IV material), such as germanium, silicon carbide, silicon germanium, etc. A bias voltage applied to the MOSCAP provides phase matching through plasma dispersion or carrier accumulation effects, depending on the polarity of the bias voltage.For example, the effect of carrier accumulation (at positive or forward bias) can change the phase and decrease the effective refractive index of the waveguide, while the effect of plasma dispersion (at negative or reverse bias) can increase the effective refractive index. Thus, controlling the bias voltage applied to the MOSCAP according to the implementations disclosed herein enables changing the refractive index of a corresponding waveguide. The hybrid MOSCAPs provided as examples use standard Group III-V materials and can be fabricated together with components for modulating gain or loss.

[0010] The realization of controllable and manufacturable PT-symmetric directional couplers with integrated hybrid MOSCAPs, as described here, can significantly reduce the footprint of large-scale optical neural networks, optical accelerators for machine learning, programmable photonic networks, and RF photonics. Directional couplers, including the implementations described here, can be used, for example, in photonic integration circuits (PIC) for dense wavelength division multiplexing (DWDM). DWDM is an optical multiplexing technology used to increase bandwidth in existing fiber optic networks. The DWDM-PIC architecture comprises photonic devices fully integrated on a single manufacturing platform.Many devices on the PIC can be realized with higher performance and smaller footprints using implementations disclosed herein, which enable improved optical switching and power splitting. For example, some conventional directional couplers that use thermal phase shifters for full power transfer can be more than 75 µm long, while directional couplers that do not use thermal phase shifters can be on the order of millimeters long. The devices and methods disclosed herein enable ultrashort optical power splitters and switches compared to conventional directional couplers. For example, the implementations described herein provide PT-symmetric directional couplers that can manipulate light at the nanoscale and include subwavelength components.As an illustrative example, the implementations disclosed herein may provide directional couplers that may have a length on the order of 25 µm or less, depending on the achieved gain-to-loss ratio. This significantly reduces the space required for the implementations described herein compared to conventional directional couplers, while enabling efficient switching and power sharing enabled by PT-balanced directional couplers.

[0011] Furthermore, the fabrication of the implementations disclosed herein is fully compatible with existing and future PIC designs without additional processes. For example, the MOSCAPs used in the directional couplers disclosed herein can be formed using manufacturing techniques and processes that are integrated with the manufacturing processes of the various components that make up PICs (e.g., lasers, modulators, detectors, etc.). Therefore, the implementations described herein need not complicate the fabrication and / or assembly of PICs and other systems using the devices disclosed herein.

[0012] As mentioned above, the operation of PT-balanced directional couplers requires a fine balance of the gain-to-loss ratio within the directional coupler. Current systems do not provide a way to adjust the optical loss to tune the gain-to-loss ratio. One example of a current system is amplification by a gallium arsenide (GaAs) or aluminum gallium arsenide (AlGaAs) material with a fixed loss generated by chromium (Cr) metal absorption. Another example system uses a lithium niobate (LiNO3) platform and optical pumping via an amplitude mask. These existing systems can provide amplification, but with a fixed loss, and they do not provide a way to modulate the optical loss.Furthermore, as previously mentioned, changes in the gain-to-loss ratio result in changes in the refractive index, which degrades system performance. Existing directional couplers do not provide a way to tune the refractive index independently of gain or loss. The implementations described here use MOSCAPs as phase shifters, which provide a lower modulation efficiency V. π L than conventional phase shifts, e.g. V π L < 0.3 V per cm.

[0013] Accordingly, the implementations disclosed herein provide separately tunable gain and loss, while simultaneously allowing phase and refractive index tuning. By using the implementations disclosed herein, high-performance PT-symmetric directional couplers can be achieved on silicon photonics platforms, which can provide ultrashort optical switches and power dividers compared to conventional directional couplers. As a result, the footprint of large-scale systems can be reduced. Furthermore, refractive index tuning by hybrid MOSCAPs as disclosed herein provides extremely low power consumption compared to other phase shifters such as microheaters, e.g., V π L < 0.3 V per cm, while the power consumption is in the order of nanowatts.

[0014] Furthermore, the present implementations enable the identity matrix transformations important for neural networks. Identity matrix transformations require lossless components. The implementations presented here can utilize optical amplification to compensate for optical losses within a system, thereby achieving the lossless components required to realize identity matrix transformations. Accordingly, the implementations described here are applicable and can be implemented as part of neural network systems.

[0015] It should be noted that the terms "optimize," "optimal," and the like, as used herein, may be used to make or achieve performance as effective or perfect as possible. However, as one skilled in the art reading this document will recognize, perfection cannot always be achieved. Accordingly, these terms may also mean to make or achieve performance as good or effective as possible or practical under the circumstances, or to cause or achieve better performance than that achievable with other settings or parameters.

[0016] Fig. 1 shows an example of a directional coupler 100. The directional coupler 100 is an example of a 2x2 optical directional coupler having a first waveguide 110 and a second waveguide 120. The first waveguide 110 includes an input port 112 and an output port 114. The second waveguide 120 includes an input port 122 and an output port 124. The first waveguide 110 defines a first leg of the directional coupler 100 and the second waveguide 120 defines a second leg of the directional coupler 100. The directional coupler 100 can be used as an optical switch or other optical device.

[0017] The directional coupler 100 comprises a coupling region 102 with a coupling length L cWithin the coupling region, light propagating in one waveguide can be branched into the other waveguide (e.g., a portion of the light propagating in one waveguide is transmitted into the other waveguide). The branched portion of the light can then propagate in the other waveguide and be output via the output port of the other waveguide. Furthermore, the portion of light remaining in one waveguide can be output via the output port of one waveguide. Furthermore, the light output from either output port may depend on propagation losses, such as excess loss (e.g., ratio of total input power to total output power); polarization-dependent loss (e.g., ratio of maximum and minimum transmissions due to polarization states in couplers); optical return loss / directivity (e.g.,the fraction of input light exiting through an input port rather than the intended output port); insertion loss (e.g., ratio of input power to output power at one of the coupler's output ports; in the case of a 2x2 coupler, insertion loss can be determined for each output port).

[0018] The amount of light branched into the other waveguide can depend on the coupling ratio between the two waveguides (e.g., 50:50, 90:10, or any coupling ratio desired for a particular application). The coupling ratio is the ratio of the optical power of each output port to the sum of the total power of both output ports as a function of wavelength. The coupling ratio can be calculated from the measured insertion loss (e.g., the loss of signal power resulting from the insertion of a device into a transmission line or optical fiber, expressed in decibels (dB)).

[0019] For example, first light 116 is fed into the input port 112 of the first waveguide 110 and propagates toward the coupling region 102. Within the coupling region 102, at least a portion 118 of the first light 116 is transmitted to the second waveguide 120 based on the coupling ratio of the directional coupler 100. The transmitted portion 118 is then coupled via the output port 124 to a transmitted light intensity matrix T 12 output as described below. The remaining part 119 of the first light 116 propagating in the first waveguide 110 is applied via the output port 114 to a light intensity matrix T 11 output. The output power at each output terminal 114 and 124 may be based on one or more losses, as described above.

[0020] In another example, the second light 126 is fed into the input port 122 of the second waveguide 120 and propagates toward the coupling region 102. Within the coupling region 102, at least a portion 128 of the second light 126 is transmitted into the first waveguide 110 based on the coupling ratio of the directional coupler 100. The transmitted portion 128 is then coupled via the output 114 to a transmitted light intensity matrix T 21 The remaining portion 129 of the first light 116 propagating in the first waveguide 110 is combined via the output 124 with a transmitted light intensity matrix T 22output. In one case, the second light may be input to the directional coupler 100 instead of the first light 116. In another case, the second light 126 may be input to the directional coupler 100 simultaneously with the first light 116 or in alternating light pulses. In this case, the output port 114 may output the remaining portion 119 of the first light 116 and the transmitted portion 128 of the second light 126, while the output port 124 outputs the transmitted portion 118 of the first light 116 and the remaining portion of the second light 129.

[0021] For example, a directional coupler can be implemented as a PT-symmetric directional coupler based on a gain-to-loss relationship between the waveguides. The development of recent nanofabrication technologies has led to the emergence of artificial structures such as PT-symmetric devices that exploit PT symmetry. PT-symmetric devices are capable of manipulating light at the nanoscale and can include sub-wavelength components that narrow the miniaturization gap between photonics and electronics. While propagation losses, such as those mentioned above, are generally considered detrimental in optical systems, propagation losses can play a positive role in optical devices, e.g., in the provision of PT-symmetric directional couplers.

[0022] PT-symmetric directional couplers can operate as efficient directional couplers (e.g., optical switches) if some loss is introduced into the system. With reference to Fig. 1, the directional coupler 100 can be implemented as a PT-symmetric directional coupler by introducing an optical gain into the first waveguide 110 (referred to herein as gain waveguide) and an optical loss (e.g., propagation loss) into the second waveguide 120 (referred to herein as loss waveguide). Fig. 6A-10 show examples of directional couplers for introducing optical gain and optical loss into coupled waveguides.

[0023] By varying the gain and loss levels in each waveguide relative to each other (e.g., varying the gain-to-loss ratio within the coupling region 102), a PT-symmetric directional coupler can be obtained, where the output signal of the directional coupler can be abruptly modulated based on small changes in the gain-to-loss ratio. This property arises from the eigenvalues ​​of the two waveguides coinciding when the gain-to-loss ratio reaches a critical threshold (referred to here as the exception point). The exception point has a singular character, providing abrupt changes in the eigenvalues ​​from the real to the imaginary axis, which respond to small changes in the gain-to-loss ratio.Thus, a large differential gain can be achieved with a modest variation in gain and / or loss, changing the gain-to-loss ratio near the point of exception. For example, PT-balanced directional couplers can dynamically switch the light output at output 114 from a first light 116 injected into the first waveguide 110 to a second light 126 injected into the second waveguide 122 by slightly modulating the gain and / or loss to tune the gain-to-loss ratio between the waveguides in the coupling region 102. The intensity of the light output from such a PT-balanced directional coupler can be the same as, or approximately the same as, the intensity at the input, regardless of the light output from the system.

[0024] The Fig. 2A to 2C show, for example, the propagation of optical waves in a conventional directional coupler ( Fig. 2A) and a PT-symmetric directional coupler, where the gain-loss ratio is below the exception point ( Fig. 2B) and equal to or above the exception point ( Fig. 2C). Fig. Figures 2A-2C illustrate how the light input into a directional coupler is transmitted under different operating conditions. Fig. 2A-2C show the directional couplers 200a-200c. Each directional coupler 200a-200c (collectively referred to herein as directional coupler 200) is substantially identical and has a first waveguide 201 and a second waveguide 203, except that the gain-loss ratio between the Fig. 2A-2C is weakened near the exception point.

[0025] With the conventional directional coupler 200a in Fig. 2A, propagation losses are minimized, and no gain is added to the system. In this case, light input to a first waveguide 201 (as shown by arrow 202) is transmitted to the second waveguide 203, for example, as described above. The light transmitted from the first waveguide 201 to the second waveguide 203 is reciprocal (i.e., there is path dependence), so the input light is output from both waveguides. Similarly, light input to the second waveguide (as shown by arrow 204) is transmitted to the first waveguide 201 in the reverse manner.

[0026] In the PT-symmetric directional coupler 200b of Fig. 2B, losses and gains are added to the system so that the gain-to-loss ratio is below the exception point. In the directional coupler 200b, the first waveguide 201 adds gain to the light propagating therein, and the second waveguide 203 introduces a loss. In this case, the light coupled into the first waveguide 201 (as shown by arrow 202) is transferred to the second waveguide 203, for example, as described above. In this case, the output power depends on the previous path. The power at the output of each waveguide therefore depends on which waveguide the light was input to.

[0027] In the PT-symmetric directional coupler 200c from Fig. 2C, loss and gain are added to the system so that the ratio of gain to loss is at or above the exception point. As in Fig. 2B, the first waveguide 201 of the directional coupler 200b introduces gain and the second waveguide 203 introduces loss. In this case, the interaction no longer occurs, and the output power can be fully switched regardless of the coupling length and is independent of the previously taken path (i.e., regardless of which waveguide the light is input to, the output power in these examples always comes from waveguide 201). For example, the light of arrow 202, which is coupled into the first waveguide 201, is not transmitted to the second waveguide 203, while the light of arrow 204, which is coupled into the second waveguide 203, is fully transmitted to the first waveguide 201 (e.g., by switching). In this way, the output power can be substantially fully switched between the light input to the respective waveguide (e.g.,approximately 100% of the input light is output from the gain waveguide). As used in conjunction with . Fig. 3A, for example, under certain conditions, an output on the loss waveguide can be as low as -50 dB, which can be essentially close to zero, meaning that essentially all the power is output from the gain waveguide.

[0028] PT-symmetric directional couplers can dynamically switch light from one waveguide to another, delivering essentially the same output intensity as that injected into an input waveguide. The light output of such a PT-symmetric directional coupler can be determined by a transmission matrix for a given coupling length L c which is defined as follows: π / 2κ. The transfer matrix M(z) with the coefficients m ij of such a directional coupler is specified as follows: M(z)=(cos(Ωz)−iδΩsin(Ωz)iκΩsin(Ωz)iκΩsin(Ωz)cos(Ωz)+iδΩsin(Ωz))exp(g1−χ22z) where z is the propagation distance, κ is the coupling coefficient, g 1 is the gain introduced by a first waveguide (e.g., waveguide 110), χ 2 (which is also known as -g 2 can be called) is the propagation loss introduced by a second waveguide (e.g. waveguide 120), and Ω is a variable representative of a detuning of the coupling coefficient, given as: Ω=δ2+κ2 δ is the detuning of the propagation constants of the waveguides (β j ) in the presence of the combined gain / loss resulting from: δ=β1+ig1−(β2−iχ2)2=β1−β22+iΔim where β 1 is the propagation constant of the first waveguide, β 2 is the propagation constant of the second waveguide and Δ imis the combined imaginary detuning (e.g. |gain| + |loss|). Δ im is specified as: Δim=g1+χ22

[0029] From equations 1-4, the transmitted light intensity matrix T ij at the output of the j-th waveguide for light injected through the i-th waveguide, where z is the coupling length L c which corresponds to a complete transition from one waveguide to the other for δ=0 (Ω = κ), can be given as follows: Tij=|Mij|2

[0030] Thus, with reference to Fig. 1 using the above equations, the matrix of the transmitted light intensity for the output light section 118 (e.g., first light 116 input to the first waveguide 110 and transmitted to and output from the second waveguide 120) is T 12can be calculated. Likewise, the matrix of the transmitted light intensity for the output light section 119 (e.g., first light 116 input to and output from the first waveguide 110) can be calculated as T, first light 116 input to and output from the first waveguide 110) can be calculated as T 11 be calculated; the transmitted light intensity matrix for the output light section 128 (e.g., second light 126 input to the second waveguide 120 and transmitted to and output from the first waveguide 110) can be calculated as T 21 be calculated; and the transmitted light intensity matrix for the output light portion 129 (e.g., second light 126 input to and output from the second waveguide 120) can be calculated as T 22 be calculated.

[0031] As mentioned above, the operation of PT-symmetric directional couplers requires balancing the gain and loss of the respective waveguides to achieve a gain-to-loss ratio that meets or exceeds the exception point. When these conditions are met, PT-symmetric directional couplers can realize efficient optical switching (e.g., optical switching with shorter device lengths than conventional directional couplers). However, the simultaneous tuning of the gain and loss of the respective waveguides results in a refractive index difference. Tuning the gain and / or loss is related, for example, to the Kramer-Kronig relationship and the linewidth enhancement factor, which are well known in the art. In addition, tuning the gain also generates heat, which changes the refractive index within the waveguide.

[0032] Fig. 3A-3C show the transmissions of an example of a PT-symmetric directional coupler at different refractive index differences between a gain waveguide and a loss waveguide. Fig. 3A-3C are simulations in which light is injected into a gain waveguide and the power of the light emitted by the gain waveguide (e.g., the matrix T calculated according to equations 1-5 above) is 11 ) and the power of the light emitted by the loss waveguide (e.g. the matrix T 12 ) as a function of the gain added by the gain waveguide, keeping the loss added by the loss waveguide constant. In each of the Fig. 3A-3C certain parameters were kept constant, e.g. the coupling length (L c ) 25 µm, the light wavelength (λ) 1.31 µm, the actual length of the gain / loss section L is 2.1 times L c(It should be noted that the labeling of L and L c in Fig. 1 is not drawn to scale and is for illustrative purposes only), and the introduced loss was -10 cm -1 The difference in refractive index (Δn) between the gain waveguide and the loss waveguide was 10e -6 , 10e -4 or 10e -3 for Fig. 3A, Fig. 3B and Fig. 3C.

[0033] The Fig. Figures 3A-3C show the power cancellations and exception points as a function of the gain-loss ratio for each difference in refractive index (Δn). Fig. 3A shows the power cancellation 302 at a gain of approximately 870 cm -1 for the transmitted light intensity matrix T 11 , the power cancellation 304 at an amplification of approx. 399 cm -1 for the transmitted light intensity matrix T 12 and the exception point 310 with a reinforcement of approx. 1125 cm-1 , wherein the light is input into an amplifying waveguide (e.g., waveguide 110) of a directional coupler. Fig. 3B shows the power cancellation 312 at a gain of approximately 870 cm -1 for the transmitted light intensity matrix T 11 , the power cancellation 314 at an amplification of approx. 399 cm -1 for the transmitted light intensity matrix T 12 and the exception point 320 with a reinforcement of approx. 1125 cm -1 , where light is input into an amplifying waveguide (e.g., waveguide 110) of a directional coupler. As can be seen from Fig. 3A and Fig. 3B, a difference in the refractive index of 10e -6 a power cancellation of more than negative 40 dB (e.g. almost negative 50 dB for T 11 and almost negative 70 dB for T 12 However, with increasing difference in refractive index, the power extinction decreases, as in Fig. 3B, where the power cancellations 312 and 314 are just above -10 dB and about -25 dB, respectively. If the difference in refractive index is too large (e.g., 10e -3 in Fig. 3C), the power cancellation and the exception point are no longer present.

[0034] The Fig. Figures 4A to 4D show example thermal maps showing the transmission intensity matrix (T ij ) for a PT-symmetric directional coupler where the difference in refractive index between the gain and loss waveguide is 10e -6 The diagrams T 11 and T 12 in Fig. 3A can each be a horizontal section from T11 in Fig. 4A and T12 in Fig. 4B, which was recorded along a single length. Fig. 5A-5D show example thermal images showing the transmission intensity matrix (T ij) for another example of a PT-symmetric directional coupler where the difference in refractive index between the gain and loss waveguides is 10e -3 The transmission intensity matrices (T ij ) are shown as heatmaps, where the color in each heatmap corresponds to a transmission intensity value in dB, as shown in the legend, plotted as length L versus gain. The dotted line in each heatmap corresponds to a transmission intensity of 0 dB. Fig. 4A and Fig. 5A correspond to the transmission intensity matrix T 11 ; Fig. 4B and Fig. 5B correspond to the transmission intensity matrix T 12 ; Fig. 4C and Fig. 5C correspond to the transmission intensity matrix T 21 ; Fig. 4C and Fig. 5C correspond to the transmission intensity matrix T 22. The transmission intensity matrix of each thermal image was fitted with a coupling length (L c ) of 25 µm, a light wavelength (λ) of 1.31 µm and a loss of -10 cm -1 calculated.

[0035] The Fig. 4A-4D illustrate that the optical circuit can be achieved independently of the actual length of the gain / loss section by adjusting the gain-loss ratio. As shown in Fig. 4A to 4D, for example, with an increase in gain and a constant loss of -10 cm -1 the transmitted intensity increases regardless of the focusing length. However, if the difference in the refractive index becomes too large (e.g. 10e -3 ), the power cancellation deteriorates, as in Fig. 5A-5D. Accordingly, as mentioned above, the PT symmetry deteriorates as the difference in refractive index between the first and second waveguides increases, thereby degrading the performance of the directional coupler.

[0036] Accordingly, implementations of the technology disclosed herein provide directional couplers that can be tuned to achieve PT symmetry. For example, the gain and / or loss introduced into the waveguides can be modulated to tune the gain-to-loss ratio near the exception point for PT symmetry. For example, a first mesa having a first optically active region can be provided with respect to a first waveguide. A bias voltage can be applied to the first mesa to cause optical gain or loss in the first waveguide based on changes in the carrier concentrations within the first mesa. By controlling the bias voltage applied to the first mesa, the gain or loss can be varied to optimize the gain-to-loss ratio with respect to the exception point.Whether gain or loss is induced may depend on the polarity of the bias voltage (e.g., forward or reverse bias). Additionally or alternatively, a second mesa having a second optically active region may be provided with respect to a second waveguide. A separate bias voltage may be applied to the second mesa to induce optical loss or gain in the second waveguide. Thus, the optical loss induced in the second waveguide can be varied by controlling the bias voltage applied to the second mesa. In an example including both the first and second mesa, controlling the respective applied bias voltage allows modulation of the optical gain and loss in each individual waveguide and fine-tuning of the gain-to-loss ratio.

[0037] In various implementations, either alone or in combination with gain-to-loss ratio tuning, phase modulation can be used to tune the refractive index between the waveguides. For example, according to various examples of the disclosed technology, a hybrid MOSCAP can be provided with respect to a waveguide, which can be biased to modulate the phase of the optical mode of a corresponding waveguide. Phase modulation corresponds to an effective change in the refractive index of the waveguide. Thus, the effective refractive index of the waveguide can be modulated by applying a bias voltage to the hybrid MOSCAP, enabling tuning of the refractive index difference between the waveguides.For example, the bias voltage applied to the MOSCAP ensures phase tuning of the optical mode through plasma dispersion or charge carrier accumulation effects, which lead to phase changes. By tuning the refractive index difference, the examples presented here can maintain a power cancellation point to achieve PT symmetry and realize efficient optical switching. This means that for lengths shorter than those possible with conventional directional couplers, the optical switching depends only on the gain-to-loss ratio. An example.

[0038] The Fig. 6A and Fig. 6B show an example of a hybrid MOS optical modulator 600 that can be implemented in the technology disclosed herein. Fig. 6A is a plan view of the optical modulator 600 and Fig. 6B is a cross-sectional view of the optical modulator 600 along a Fig. Line A-A' shown in Figure 6A.

[0039] The optical modulator 600 includes an optical waveguide 602, a cathode 604 comprising a first material and formed within the optical waveguide 602, and an anode 606 comprising a second material different from the first material and formed within the optical waveguide 602. The anode is adjacent to the cathode. A capacitor is defined between the anode and the cathode.

[0040] In some examples, a substrate 601 comprises oxide grown on an underlying layer 608. A silicon layer 610 is deposited on the substrate 601. A trench 612 separates the optical modulator 600 into two sections 614 and 616. The first section 614 includes the anode 606. The optical waveguide 602 is formed in the anode 606. The cathode 604 is integrated into the second part 616. In various embodiments, the cathode 604 includes a layer of a group III-V material as the first material. A MOS capacitor (MOSCAP) 624 is located between the cathode 604 and the anode 606.

[0041] A dielectric 618 is formed between the cathode 604 and the anode 606. The dielectric 618 may be an electrically insulating material formed between the cathode 604 and the anode 606 of the MOSCAP 624, and polarization of the dielectric 618 by an applied electric field may increase the surface charge of the MOSCAP 624 for a given electric field strength. The dielectric 618 may be composed of native oxides of the cathode or the anode, or both, or of external dielectric materials such as high-k dielectrics or polymers, which may be formed by deposition, oxidation, wafer bonding, or other dielectric coating processes.

[0042] The cathode 604 may comprise negatively doped Group III-V material, and the anode 606 may comprise positively doped silicon, other Group IV material, or any semiconductor material that is electrically conductive and optically transparent. A cathode electrode 620 is disposed on the cathode 604, and an anode electrode 622 is disposed on the anode 606. When a voltage is applied between the electrodes, charge carrier accumulation, depletion, or inversion may occur around the dielectric 618. Because the capacitor region overlaps with the optical waveguide, a change in the charge carrier concentration can lead to changes in the modal refractive index of the waveguide and propagation losses. By biasing the voltage applied between the electrodes, the refractive index can be appropriately modulated, causing optical intensity modulation, phase shift modulation, and attenuation.

[0043] The light can be input to the optical modulator 600 via the waveguide 602. As the light propagates through the waveguide 602, it can be modulated, attenuated, and phase shifted based on changes in the modal refractive index of the waveguide induced by the bias voltage of the MOSCAP 624. The modulated light is then output from the optical modulator 600. The modal refractive index of the waveguide can be changed, for example, by accumulation, depletion, or inversion of charge carriers when a voltage is applied to the electrodes.

[0044] Fig. 6B, for example, includes a DC power source 626. The DC power source 626 serves as a signal source and has a negative terminal connected to the cathode electrode 620 and a positive terminal connected to the anode electrode 622. This results in a migration of negative charges from the cathode 604 to a side of the optical waveguide 602 adjacent to the cathode 604 and a migration of positive charges ("holes") from the anode 606 to an opposite side of the waveguide 602 (also referred to herein as accumulation mode). In other examples, the polarity of the DC power source 626 can be reversed. Reversing the polarity of the DC power source 626 causes a migration of negative charges from the waveguide 602 to the cathode electrode 620 and a migration of holes from the waveguide 602 to the anode electrode 622 (also referred to herein as depletion mode).

[0045] The MOSCAP 624 forms at the interface between the Group III-V material of the cathode 604 and the underlying capacitor portion of intrinsic silicon or other Group IV material of the anode 606. A thin layer of silicon and Group III-V oxides (e.g., the dielectric 618) naturally forms at this interface and serves as the dielectric for the capacitor. In some examples, this thin layer has a thickness on the nanoscale, e.g., a few nanometers thick. In some examples, no steps need to be taken to promote the formation of the dielectric 618. In other examples, the formation of the dielectric 618 can be stimulated, e.g., by increasing the temperature, by exposing the materials to an oxygen-rich atmosphere, or by another suitable technique.

[0046] In some examples, the Group III-V material may include gallium arsenide (GaAs). In other examples, the Group III-V material may include indium phosphide (InP) or other compounds of indium, gallium, phosphorus, and arsenic. Generally, the cathode and anode are formed from different materials, which may include II-VI semiconductor compounds or other materials. Metals may also be used.

[0047] In some examples, the Group III-V oxides that form the dielectric 618 may be aluminum oxide (Al 2 O 3 ), hafnium oxide (HfO 2 ) or vanadium dioxide (VO 2 ). In other examples, the Group III-V oxides may include, but are not limited to, the compounds listed in Table 1 below (note that Si has a gap of 1.1 eV): Table 1 K Spalt (eV) CB-Offset (eV) Not. 2 3.9 9 3.2 Yes 3 N 4 7 5.3 2.4 <h2 style=";text-align:left;direction:ltr">Al<h2 style=";text-align:left;direction:ltr"> 2 <h2 style=";text-align:left;direction:ltr"> O<h2 style=";text-align:left;direction:ltr"> 3 9 8.8 2.8 (nicht ALD) The 2 The 5 22 4.4 0.35 TiO 2 80 3.5 0 SrTiO 3 2000 3.2 0 ZrO 2 25 5.8 1.5 HfO 2 25 5.8 1.4 HfSiO 4 11 6.5 1.8 to 2 A 3 30 6 2.3 AND 2 EITHER 3 15 6 2.3 a-LaAlO 3 30 5.6 1.8

[0048] As previously mentioned, the MOSCAP 624 is formed within the optical waveguide 602 such that charge carriers accumulating / decreasing on both sides of the capacitor dielectric cause a change in the refractive index of the optical waveguide and a change in the waveguide attenuation (e.g., loss or attenuation of the transmitted signal power in the waveform).

[0049] The MOSCAP 624 can operate in accumulation, depletion, or inversion mode (e.g., accumulation of electrons at the dielectric layer in addition to the presence of holes). As described above, a DC voltage can be applied between an anode 606 and a cathode 604, accumulating, depleting, or inverting a thin layer of charge on both sides of the dielectric layer 618. The resulting change in the density of free charge carriers causes a plasma or charge carrier accumulation effect, which causes a change in the phase of the optical mode of the optical waveguide 602. This phase change is manifested by a change in the effective refractive index of the optical mode (Δn eff ). The extent of change or modulation of the effective refractive index (Δn eff ) and the associated change in optical losses (Δα) can be described as follows: Δneff=−q2λ028π2c2nε0(ΔNemce*+ΔNhmch*) Δα=−q3λ024π2c3nε0(ΔNemce*μe+ΔNhmch*2μh)

[0050] Where q is the electric charge applied to the cathode 604 and the anode 606, c is the speed of light in vacuum, ε 0 the dielectric constant of free space and n the refractive index of the material, ΔN represents a change in the charge carrier density, so that ΔN e represents the change in the charge carrier density with respect to electrons, that ΔN h represents the change in charge carrier density with respect to holes, m* is the relative effective mass of electrons (m* ce ) and holes (m* ch ), µ h represents the mobility of the holes, µ e represents the mobility of electrons and λ 0 is the wavelength of free space.

[0051] An optical phase shift (Δφ) at the end of the capacitor depends on the magnitude of the voltage-induced Δn eff, the device length L and the optical wavelength λ. In this example, the optical phase shift can be written as Δφ=2π Δn eff Lλ can be calculated. Thus, the optical phase of the light in the optical waveguide 602 can be calculated based on the voltage-induced Δn eff shifted and vice versa. In various examples, the waveguide loss in silicon and Group III-V materials can change simultaneously with the change in carrier density, and controlling the change in waveguide loss can be used as an optical attenuator. For example, changes in waveguide loss can be controlled based on the change in carrier density, which can lead to attenuation of the waveguide losses. The attenuated waveguide losses can be used to modulate a signal.

[0052] The Fig. 7A and Fig. 7B show simulated relationships between the signal applied to the optical modulator 600 of the Fig. 6A and Fig. 6B applied voltage for different slit thicknesses plotted against the change in refractive index (Δn eff ) of the optical mode of the waveguide (left axis) and the change in losses due to free carrier absorption (FCA) (right axis). In the simulations used to create Fig. 7A and Fig. 7B, the cathode 604 was made of n-doped gallium arsenide (GaAS) with a doping of 3×10 18 cm -2 and the anode 606 made of p-doped silicon with a doping of 5×10 16 cm -2 The dielectric 618 in the simulation of Fig. 7A was made of aluminum oxide (Al 2 O 3 ) and the dielectric 618 in the simulation of Fig. 7B from hafnium oxide (HfO 2). The thickness of the gap between the anode 606 and the cathode 604 (e.g., the thickness of the dielectric 618 formed therebetween) was varied in each simulation in 5nm steps from 5nm to 25nm, as shown in legend 705 of Fig. 7A and legend 715 of Fig. 7B. These are only example configurations that are used to generate the Fig. 7A and Fig. 7B were used. Other configurations are conceivable within the scope of the implementations disclosed here.

[0053] As indicated by the pointer 701 in Fig. 7A and the pointer 711 in Fig. 7B, the solid lines show the change in the refractive index (Δn eff) for each gap thickness as a function of the bias voltage applied to the optical modulator 600. Similarly, pointers 703 and 713 indicate that the dashed lines represent the change in FCA loss for each gap thickness as a function of the bias voltage applied to the optical modulator 600. As shown in the Fig. 7A and Fig. 7B, the amount of refractive index change depends on the type of dielectric 618 and the gap thickness, and in particular, the change in refractive index decreases with increasing gap thickness. Furthermore, the Fig. 7A and Fig. 7B that the FCA loss depends on the type of dielectric 618 and the gap thickness, and in particular, the FCA loss decreases with increasing gap thickness.

[0054] The Fig. 8A-8C show examples of directional couplers according to the implementations disclosed herein. Fig. Figure 8A shows a top view of the directional coupler 800. Fig. Figure 8B shows a cross-sectional view of an exemplary directional coupler 800a along line BB' within a coupling range of Fig. 8A. Fig. Figure 8C shows a cross-sectional view of another example directional coupler 800b along line BB' of Fig. 8A. The directional couplers 800a and 800b of the Fig. 8B and Fig. 8C are substantially identical to the directional coupler 800, except that the directional coupler 800a has an isolation region 838 implemented as an ion region 838a, and the directional coupler 800b has an isolation region 838 implemented as a gap 838b, for example, an air gap. Accordingly, the following disclosure applies equally to each of the Fig. 8A-8C, unless otherwise stated here.

[0055] In various examples, the directional coupler 800 includes a buried oxide layer (BOX) 832 grown on a substrate 834, e.g., a silicon dioxide layer. The directional coupler 800 also includes a first portion 801 (also referred to herein as a first leg) and a second portion 802 (also referred to herein as a second leg) formed on the BOX layer 832. The first portion 801 is spaced from the second portion 820 by a gap 836 (e.g., an air gap). The first portion 801 includes a first optical waveguide 803, and the second portion 802 includes a second optical waveguide 804. Each optical waveguide 803 and 804 has one or more optical modes, and an example of the optical energy of the one or more modes is shown as regions 805 and 806, respectively. In various embodiments, the first and second waveguides 803 and 804 are single-mode waveguides.

[0056] The first optical waveguide 803 and the second optical waveguide 804 may be provided to the first waveguide 110 and the second waveguide 120, respectively, of Fig. 1. For example, the first optical waveguide 803 includes an input port and an output port, and the second optical waveguide 804 includes an input port and an output port. Light input to one optical waveguide can be branched into the other optical waveguide via the gap 836, as described above in connection with the Fig. 1-4C. In the Fig. In the example illustrated in Figure 8A, light is input to an input port of the second optical waveguide 804 and output from the output ports of both optical waveguides 803 and 804. Light input to one waveguide (e.g., one of the optical waveguides 803 and / or 804) may be transmitted to the other waveguide (e.g., one of the optical waveguides 804 and / or 803) via the gap 836 and the BOX layer 832 and output from the directional coupler 800 for downstream use. The BOX layer 832 may be provided to confine the optical mode in the vertical direction (e.g., into the layers provided on the BOX layer 832). Controlling the width of the gap 836 may change the coupling ratio between the first and second optical waveguides 803 and 804 (e.g., a larger width results in a smaller coupling ratio).

[0057] The first section 801 also includes a first mesa 821 and a first MOSCAP 811. The first mesa 821 serves to introduce optical gain or optical loss into the first optical waveguide 803 by applying a bias voltage between the contact electrode 823 (herein referred to as electrode 823) and the contact electrode 817 (herein referred to as electrode 817). Whether optical gain or optical loss occurs depends on the polarity of the bias voltage; for example, a forward bias may be used for gain and a reverse bias may be used for loss. The first MOSCAP 811 is configured to change the refractive index of the first optical waveguide 803 based on a bias voltage between the contact electrode 819 (herein referred to as electrode 819) and the electrode 817.As previously mentioned, for example, a bias applied to MOSCAP 811 provides phase tuning through plasma dispersion or charge accumulation effects, depending on the polarity of the bias. A charge accumulation effect via a positive or forward bias changes the phase and decreases the effective refractive index of the first waveguide 803, while a plasma dispersion effect via a negative or reverse bias increases the effective refractive index. Further details on the components of the first mesa 821 and the first MOSCAP 811 are explained below.

[0058] The second section 802 also includes a second mesa 822 and a second MOSCAP 812. The second mesa 822 serves to introduce optical loss or optical gain to the second optical waveguide 804 by applying a bias voltage (e.g., reverse bias or forward bias) between the contact electrode 824 (herein referred to as electrode 824) and the contact electrode 818 (herein referred to as electrode 818). Similar to the first MOSCAP 811, the second MOSCAP 812 is configured to change the refractive index of the second optical waveguide 804 based on a bias voltage between the contact electrode 820 (herein referred to as electrode 820) and the electrode 818. Further details on the components of the second mesa 822 and the second MOSCAP 812 are described below.

[0059] Accordingly, the directional coupler 800 provides for modulating the gain and / or loss simultaneously with tuning a refractive index difference between the first and second optical waveguides 803 and 804 to optimize the gain-to-loss ratio and power cancellation and to obtain a PT-symmetric directional coupler. For example, one or more bias voltages may be applied to the first mesa 821 and / or the second mesa 821 to individually tune the optical gain and / or loss, thereby optimizing the gain-to-loss ratio with respect to the exception point. For example, a bias voltage may be applied to the first mesa 821 to modulate the gain or loss depending on the polarity of the bias voltage (e.g., forward or reverse bias).Separately, either simultaneously, subsequently, or before the biasing of the first mesa 821, a bias voltage may be applied to the second mesa 822 to modulate the loss or gain depending on the polarity of the bias voltage (e.g., forward or reverse bias).

[0060] In some examples, a bias voltage may be applied to the first and / or second MOSCAPs 811 and 812 simultaneously with the bias voltage of the first and / or second mesas 821 and 822 to tune a refractive index difference between the optical waveguides 803 and 804. For example, applying a voltage to the first or second MOSCAPs 811 and 822 causes a change in the refractive index of the respective optical waveguides 803 and 804 (e.g., according to Equations 6 and 7 above). The refractive index change in one or both of the first and second optical waveguides 803 and 804 can be used to tune the refractive index difference to maximize power cancellation while optimizing the gain-to-loss ratio.

[0061] The information provided here with reference to the Fig. While the examples illustrated in Figures 8A-8C provide both gain and loss tuning, the technology disclosed herein is not so limited. For example, instead of the first and second mesa 821 and 822, implementations including one of the mesa 821 and 822 may be provided. For example, in one implementation, the first mesa 821 may be provided and the second mesa 822 may not be included. For example, as described above, the first mesa 821 may be biased to modulate the gain while keeping the loss constant. Another example is that the second mesa 822 may be present and the first mesa 821 may not be included. In this case, the first mesa 821 may be biased to modulate the loss while keeping the gain constant.

[0062] Likewise, even if this is done with reference to the Fig. 8A and Fig. 8B provides two MOSCAPs for independently modulating the refractive index of each waveguide, the technology disclosed herein is not limited thereto. For example, instead of the first and second MOSCAPs 811 and 812, implementations including either of the MOSCAPs 811 and 812 may be provided. For example, in one implementation, the first MOSCAP 811 may be provided and the second MOSCAP 812 may not be included. Thus, as described above, the first MOSCAP 811 may be biased to modulate the refractive index of the first optical waveguide 803 while keeping that of the second optical waveguide 804 constant. Another example is that a second MOSCAP 812 is provided and the first MOSCAP 811 is not included.In this case, the second MOSCAP 812 may be biased so that the refractive index of the second optical waveguide 804 is modulated while that of the first optical waveguide 803 is kept constant.

[0063] The first section 801, which will now be discussed in more detail, includes a cathode 807 comprising a first material, a portion of which is formed in the first optical waveguide 803, and an anode 809 comprising a second material different from the first material and also formed in the first optical waveguide 803. In some embodiments, the cathode 807 tapers toward the optical waveguide 803 at the input and output of the optical waveguide to enable an optical transition from the optical waveguide to the cathode 807, such as tapers 840 and 842 in Fig. 8A. The anode 809 is adjacent to the cathode 807 within the first optical waveguide 803. A first capacitor 811 is defined between the anode 809 and the cathode 807. In various embodiments, the cathode 807 has a stepped cross-section to reduce the presence of the first material in areas not required to perform the functions and operations of the first portion 801.

[0064] The anode 809 is formed on the BOX layer 832, and the cathode 807 is formed above the anode 809 opposite the BOX layer 832. The anode 809 includes a trench 813 formed therein, and the cathode 807 spans the trench 813. The trench 813 may be provided to confine the optical mode in the horizontal direction. In various embodiments, the cathode 807 includes a layer of Group III-V material as the first material, such as indium phosphide (InP), gallium arsenide (GaAs), or other compounds of indium, gallium, phosphorus, and arsenic. The cathode 807 may be formed, for example, but not exclusively, by deposition, wafer bonding, monolithic growth, or other manufacturing techniques. The anode 809 may contain silicon or another Group IV material as a second material, e.g., germanium, silicon carbide, silicon germanium, and so on.The capacitor 811 may be the first MOSCAP located between the cathode 807 and the anode 809.

[0065] A dielectric 815 (also referred to herein as an interface layer) is formed between the cathode 807 and the anode 809 and in the first optical waveguide 803. The dielectric 815 may consist of native oxides of the cathode or the anode, or both, or of external dielectric materials such as high-k dielectrics or polymers, which may be formed by deposition, oxidation, wafer bonding, or other dielectric coating processes. The dielectric 815 may, for example, be similar to the dielectric 618 in the Fig. 6A and Fig. 6B and can be aluminum oxide (Al 2 O 3 ), hafnium oxide (HfO 2 ), vanadium dioxide (VO 2 ) or one of the compounds listed in Table 1.

[0066] The first MOSCAP 811 forms at the boundary between the first material of the cathode 807 and the underlying first optical waveguide 803 made of semiconductor material (e.g., silicon or another Group IV material). At this boundary, a thin layer of silicon and III-V oxides forms, which serves as a dielectric for the capacitor (e.g., dielectric 815). In some examples, the dielectric layer 815 may have a thickness in the nanoscale, e.g., the dielectric layer 815 may be several nanometers thick. As shown in the Fig. 7A-7B, the thickness of the dielectric layer 815 may be, for example, between 5 nm and 25 nm. As described in connection with the Fig. 7A and Fig. 7B, a smaller thickness results in larger changes in the effective refractive index for a lower applied bias voltage; however, a smaller thickness also results in a larger FCA loss. Therefore, some implementations use a dielectric with a small thickness (e.g., on the order of 5 nm) and account for the FCA loss through gain / loss modulation across the first mesa 821. In some examples, no steps need to be taken to promote the formation of the dielectric 815. In other examples, the formation of the dielectric 815 may be encouraged, for example, by increasing the temperature, exposing the materials to an oxygen-rich atmosphere, or by another suitable technique.

[0067] Electrode 817 is arranged on cathode 807 and electrode 819 on anode 809. When a bias voltage is applied to electrodes 817 and 819, depending on the polarity of the bias voltage (e.g., forward or reverse bias), accumulation or depletion of charge carriers around dielectric 815 may occur, as described in more detail below, leading to charge carrier or plasma dispersion effects. Since the capacitor region overlaps with the optical waveguide, a change in the charge carrier concentration can lead to changes in the optical mode phase, the effective refractive index, and the propagation loss, as explained in Equations 6 and 7. By biasing the voltage applied between electrodes 817 and 819, the refractive index can be modulated accordingly. For example, a bias voltage (e.g.,A DC voltage (DC) is applied between anode 809 and cathode 807, causing a thin charge layer to build up or break down on both sides of the dielectric layer 815. The resulting change in the density of free charge carriers causes a plasma or charge carrier accumulation effect, which causes a change in the phase of the optical mode of the first optical waveguide 803. This phase change is expressed by a change in the effective refractive index of the optical mode (Δn. eff ). The extent of change or modulation of the effective refractive index (Δn eff ) and the associated change in optical losses (Δα) can be described with equations 6 and 7 above.

[0068] The cathode 807 may comprise negatively doped material (e.g., an n-doped semiconductor layer comprising silicon or another Group IV material), and the anode 809 may comprise a first portion 809a comprising positively doped material (e.g., a p-doped semiconductor layer). To enhance the change in effective refractive index and the optical phase shift induced in the first optical waveguide 803, the anode 809 comprises a second portion 809b formed on the BOX layer 832 adjacent to the first portion 809a opposite the first optical waveguide 803. The electrode 819 may be disposed on the second portion 809b.

[0069] In various implementations, the cathode 807 comprises a heavily negatively doped material, the first portion 809a comprises a positively doped material, and the second portion 809a comprises a heavily positively doped material. The terms "heavily doped" and "lightly doped" as used herein refer to the doping concentrations with respect to a doped material. That is, a heavily doped material has a doping concentration (e.g., charge carrier concentration, either holes or electrons, depending on whether the doping is p-type or n-type) that is greater than that of a doped material, while a lightly doped material has a doping concentration that is less than that of a doped material. In various examples disclosed herein, light doping may refer to a case where the doping concentration is on the order of one atom per 100 million atoms of the material.Heavily doped can refer to a case where the dopant concentration is much higher, on the order of one atom per ten thousand atoms of the material. In the example shown, the cathode 807 may be a Group III-V material with an n-type doping concentration of about 1e. 18 cm -3 or more. In another example, the cathode 807 may be set to a value between about 1e 18 cm -3 and about 1e 19 cm -3 In one example, the first portion 809a may be silicon or another Group IV material with a p-type doping concentration of about 1e 16 cm -3 and about 5e 18 cm -3 In some examples, the second portion 809b may comprise silicon or another Group IV material that is heavily positively doped, for example, to more than about 5e 18 cm -3 .

[0070] The first section 801 also includes the first mesa 821, as described above, which is configured to provide optical gain and / or optical loss. Optical gain may be achieved, for example, by the light emission generated in the first mesa 821, which generates light that can be absorbed in the optical energy region 805 and injected into the first optical waveguide 803. The light injected into the first optical waveguide 803 acts as optical gain induced in the first optical waveguide 803. The light emission may be provided, for example, by a light-emitting diode, a laser diode, or the like.

[0071] Alternatively, the optical loss can be achieved by light absorption within the first mesa 821 by absorbing optical energy from the optical energy region 805, thereby absorbing the light propagating in the first optical waveguide 803. A reverse bias on an active region causes electron-hole band bending, resulting in absorption. The light absorption in the second optical waveguide 804 corresponds to the propagation loss that occurs in the second optical waveguide 804. The light absorption can be achieved, for example, by biasing an optically active medium.

[0072] To achieve modulation of the optical gain or optical loss, the first mesa 821 includes an optically active region 825 (or laser gain material in the case of a laser diode) formed on the cathode 807. The optically active region 825 may comprise, for example, quantum dot (QD), quantum well (QW), quantum well (QD) structures, or any other structure capable of generating an inversion of the charge carrier population for optical amplification as an optical gain medium. A semiconductor layer 827 is formed on the optically active region 825. The semiconductor layer 827 may be made of positively doped group III-V material, such as GaAs or the like. In one illustrative example, the semiconductor layer 827 may be coated with p-type dopants at a concentration of more than about 5e 18 cm -3 be endowed.

[0073] Electrode 823 may be formed on semiconductor layer 827 opposite optically active region 825. In one example, where a forward bias is applied between electrodes 823 and 817, the change in carrier concentration due to accumulation results in stimulated emission in optically active material 825, thereby generating light. The generated light passes through the layers and is absorbed in optical energy region 805, and is received and propagated in first optical waveguide 803. In another example, in response to a reverse bias applied between electrodes 823 and 817, electron-hole band bending occurs, causing absorption, whereby the light propagating in first optical waveguide 803 is absorbed, resulting in propagation loss.

[0074] Accordingly, by biasing the voltage applied between electrodes 823 and 817, optical gain or loss can be added to the first optical waveguide 803 via the first mesa 821. The amount of optical gain or loss can be modulated based on changing the magnitude of the bias voltage applied to the first mesa 821, which can be tuned with respect to the loss or gain in the second section 802 to achieve the exception point. Furthermore, the optical gain-to-loss ratio can be tuned simultaneously with tuning the refractive index difference between the first and second sections 801 and 802 by controlling the bias voltage applied to electrodes 823 and 817 and the bias voltage applied to electrodes 819 and 817.Thus, while a change in gain can also cause a change in refractive index (and vice versa), both the refractive index and the gain-to-loss ratio can be tuned to ensure that power cancellation is maintained and that the gain-to-loss ratio reaches or exceeds the exception point.

[0075] The first mesa 821 also includes an ion region 829 within the semiconductor layer 827. The ion region 829 provides efficient electrical charge carrier confinement within the semiconductor layer 827. In the case of induced optical amplification, for example, the ion region 829 limits charge carrier migration to a region that vertically overlaps with the first optical waveguide 803. In this way, the ion region 829 ensures that the optical amplification is generated in the direction of the first optical waveguide 803 and in the horizontal direction, so that the optical amplification is not supplied to the first optical waveguide 803. The ion region 829 may be formed by an ion implantation process to implant ions into a portion of the semiconductor layer 827. The ions may be hydrogen ions (H + ) or any other ions desired for a particular application.

[0076] Now to the second section 802 in detail. The second section 802 can be constructed substantially similarly to the first section 801, having a mirrored structure. The second section 802 includes, for example, a cathode 808 comprising a first material, a portion of which is formed in the second optical waveguide 804. The second section 802 also includes an anode 810 made of a second material different from the first material and also formed in the second optical waveguide 804. In some embodiments, the cathode 808 tapers at the entrance and exit of the waveguide toward the waveguide 804 to enable an optical transition from the waveguide to the cathode 808, such as tapers 844 and 846 in Fig. 8A. The anode 810 is adjacent to the cathode 808 within the second optical waveguide 804. A second capacitor 812 is defined between the anode 810 and the cathode 808. In various embodiments, the cathode 808 has a stepped cross-section to reduce the presence of the first material in areas not required to perform the functions and operations of the second portion 802.

[0077] The anode 810 is formed on the BOX layer 832, and the cathode 808 is formed above the anode 810 opposite the BOX layer 832. The anode 810 includes a trench 814 formed therein, and the cathode 808 spans the trench 814. The trench 814 may confine the optical mode in the horizontal direction. In various embodiments, the cathode 808 includes a layer of a Group III-V material as the first material. The cathode 808 may be formed, for example, but not exclusively, by deposition, wafer bonding, monolithic growth, or other manufacturing techniques. The anode 810 may include silicon or another Group IV material as the second material. The capacitor 812 may be the second MOSCAP located between the cathode 808 and the anode 810.

[0078] A dielectric 816 (also referred to herein as an interface layer) is formed between the cathode 808 and the anode 810 and in the second optical waveguide 804. The dielectric 816 may consist of native oxides of the cathode or the anode, or both, or of external dielectric materials such as high-k dielectrics or polymers, which may be formed by deposition, oxidation, wafer bonding, or other dielectric coating processes. The dielectric 816 may, for example, be similar to the dielectric 618 in the Fig. 6A and Fig. 6B and can be aluminum oxide (Al 2 O 3 ), hafnium oxide (HfO 2 ), vanadium dioxide (VO 2 ) or one of the compounds listed in Table 1.

[0079] The second MOSCAP 812 forms at the boundary between the first material of the cathode 808 and the underlying second optical waveguide 804. A thin layer of silicon and III-V oxides forms at this boundary and serves as a dielectric for the capacitor (e.g., dielectric 816). In some examples, the dielectric layer 816 may have a thickness in the nanoscale, e.g., the dielectric layer 816 may be several nanometers thick. As shown in the Fig. 7A-7B, the thickness of the dielectric layer 816 may be, for example, between 5 nm and 25 nm. As described in connection with the Fig. 7A and Fig. 7B, a smaller thickness results in larger changes in the effective refractive index for a lower applied bias voltage; however, a smaller thickness also results in a larger FCA loss. Therefore, some implementations use a dielectric with a small thickness (e.g., on the order of 5 nm) and account for the FCA loss through gain / loss modulation across the second mesa 822. In some examples, no steps need to be taken to promote the formation of the dielectric 816. In other examples, the formation of the dielectric 816 may be encouraged, for example, by increasing the temperature, by exposing the materials to an oxygen-rich atmosphere, or by another suitable technique.

[0080] The electrode 818 is arranged on the cathode 808 and the electrode 820 on the anode 810. When a bias voltage is applied to the electrodes 818 and 820, depending on the polarity of the bias voltage (e.g., forward or reverse voltage), an accumulation or depletion of charge carriers around the dielectric 816 may occur, for example, similarly to that described above in connection with the first MOSCAP 811. Changes in the charge carrier concentration in the overlap of the first MOSCAP 811 and the second optical waveguide 804 may lead to changes in the waveguide refractive index and propagation loss, as explained, for example, in connection with equations 6 and 7, which may be used to modulate the refractive index of the second optical waveguide 804. For example, a bias voltage (e.g.,A DC voltage (DC) is applied between anode 810 and cathode 808, causing a thin charge layer to build up or deplete on both sides of the dielectric layer 816. The change in the density of the free charge carriers leads to a change in the effective refractive index of the optical mode (Δn. eff ), as shown in equations 6 and 7 above.

[0081] The cathode 808 may comprise negatively doped material (e.g., an n-doped semiconductor layer comprising silicon or another Group IV material), and the anode 810 may comprise a first portion 810a comprising positively doped material (e.g., a p-doped semiconductor layer). To enhance the change in effective refractive index and the optical phase shift induced in the second optical waveguide 804, the anode 810 includes a second portion 810b formed on the BOX layer 832 adjacent to the first portion 810a opposite the second optical waveguide 804. The electrode 820 may be disposed on the second portion 810b.

[0082] In various embodiments, the cathode 808 comprises a heavily negatively doped material, the first portion 810a comprises a positively doped material, and the second portion 810a comprises a heavily positively doped material. In the example shown, the cathode 808 may comprise a Group III-V material with an n-type doping concentration of about 1e 18 cm -3 or more. In another example, the cathode 808 may be set to a value between about 1e 18 cm -3 and about 1e 19 cm -3 In one example, the first portion 810a may be silicon or another Group IV material with a p-type doping concentration of about 1e 18 cm -3 and about 5e 18 cm -3 In some examples, the second portion 810b may be made of silicon or another Group IV material that is heavily positively doped, e.g., to more than about 5e 18 cm -3 .

[0083] The second section 802 also includes the second mesa 822, as described above, which is configured to induce optical loss and / or optical gain, similarly as described in connection with the first mesa 821. To achieve modulation of the optical gain or optical loss, the second mesa 821 includes an optically active region 826 (or laser gain material in the case of a laser diode) formed on the cathode 808. The optically active region 826 may, for example, include QD or QW as the optical gain medium. A semiconductor layer 828 is formed on the optically active region 826. The semiconductor layer 828 may comprise positively doped group III-V material, e.g., GaAs or the like. In one illustrative example, the semiconductor layer 828 may be doped with p-type dopants at a concentration of more than about 5e 18 cm -3 be endowed.

[0084] Electrode 824 may be formed on semiconductor layer 828 opposite optically active region 826. In one example, where a forward bias is applied between electrodes 824 and 828, the change in charge carrier concentration due to accumulation results in stimulated emission in optically active material 826, thereby generating light. The generated light passes through the layers and is absorbed in optical energy region 806, and is received and propagated in second optical waveguide 804. In another example, a change in charge carrier concentration in response to a reverse bias applied between electrodes 824 and 828 results in energy absorption, causing light propagating in second optical waveguide 804 to be absorbed and cause propagation loss.

[0085] Accordingly, by biasing the voltage applied between electrodes 824 and 818, optical gain or loss can be added to the second optical waveguide 804 via the second mesa 822. The amount of optical gain or loss can be modulated based on changing the magnitude of the bias applied to the second mesa 822, which can be tuned with respect to the loss or gain in the second section 801 to achieve the exception point. Furthermore, the ratio of optical gain to loss can be tuned simultaneously with tuning the refractive index difference between the first and second sections 801 and 802 by controlling the bias applied to electrodes 824 and 818 and the bias applied to electrodes 820 and 818.Thus, while a change in gain can also cause a change in refractive index (and vice versa), both the refractive index and the gain-to-loss ratio can be tuned to ensure that power cancellation is maintained and that the gain-to-loss ratio reaches or exceeds the exception point.

[0086] The second mesa 822 also includes an ion region 830 within the semiconductor layer 828. The ion region 830 provides efficient electrical charge carrier confinement within the semiconductor layer 828. In the case of induced optical gain, for example, the ion region 830 limits charge carrier migration to a region that vertically overlaps with the second optical waveguide 804. In this way, the ion region 830 ensures that the optical gain is generated toward the second optical waveguide 804 and in a horizontal direction, so that the optical gain would not leak into the second optical waveguide 804. The ion region 830, similar to the ion region 829, can be formed by an ion implementation process to implant ions into a portion of the semiconductor layer 828. The ions can be hydrogen ions (H +) or any other ions desired for a particular application.

[0087] In various examples, the first and second sections 801 and 802 may be formed simultaneously. For example, the anode 810 and the anode 809 may be formed as a single anode layer by, but are not limited to, deposition, wafer bonding, monolithic growth, or other manufacturing techniques. Then, the gap 836 may be formed by a dry or wet chemical etching process, creating separate anodes 809 and 810 from the anode layer. Simultaneously or in a different step, the trenches 814 and 813 may be etched from the anode layer. Next, a cathode layer of Group III-V material may be formed on the anodes 809 and 810, for example, but not limited to, deposition, wafer bonding, monolithic growth, or other manufacturing techniques as a single cathode layer. An optically active layer may be formed on the cathode layer, e.g.by wafer bonding a donor wafer that already contains the active region. The active regions on the donor wafer are created by metal-organic chemical vapor deposition or molecular beam epitaxy. Subsequently, an additional layer of group III-V material can be formed on the optically active layer, for example, but not exclusively, by deposition, wafer bonding, monolithic growth, or other manufacturing techniques.

[0088] Next, an isolation region 838 may be formed in the cathode layer, the optically active layer, and the other layer of group III-V material to form the first and second mesas 821 and 822 by electrically isolating the first and second mesas 821 and 822 from each other. In the example of Fig. 8B, the isolation region 838 may be an ion region 838a formed by an ion implementation process similar to the process used to form the ion regions 829 and 830 to implant ions between the mesas 821 and 822. The ions may be hydrogen ions (H + ) or any other ions desired for a particular application. As another example, as shown in Fig. 8C, the isolation region 838 may be formed by an etching process to remove material from the cathode layer, the optically active layer, and the other layer of Group III-V material to form a gap or space 838b between the first and second mesas 821 and 822. In this case, the gap 836 may be formed as part of the formation of the isolation region 838; e.g., the formation of the gap 838b may include the formation of the gap 836. In both examples, the gap 838b or the ion region 838a electrically isolates the first and second mesas 821 and 822 and simultaneously forms separate cathodes 808 and 807, optically active regions 826 and 825, and the semiconductor layers 828 and 827. Then, the electrodes 820, 818, 824, 823, 817, and 819 can be formed on the respective layers, for example, by deposition, wafer bonding, monolithic growth, or other manufacturing techniques, as shown in the Fig. 8B and Fig. 8C.

[0089] In various embodiments, as in Fig. 8B and Fig. 8C, dielectrics 815 and 816 may span gap 836. For example, dielectrics 815 and 816 may be a single layer spanning first and second portions 801 and 802, as shown here. The single dielectric layer may be disposed in gap 836 and cover the sides of anodes 810a and 808b.

[0090] While certain materials are described herein as negatively or positively doped, implementations are not limited thereto, and the polarity doping can be reversed. While in the example above, cathode 807 was described as negatively doped and anode 809 and semiconductor layer 827 were described as positively doped, the polarity of each layer can be switched so that cathode 807 is positively doped and anode 809 and semiconductor layer 827 are negatively doped. Similarly, the polarity of cathode 808, anode 810, and semiconductor layer 828 can be switched so that cathode 808 is positively doped and anode 810 and semiconductor layer 828 are negatively doped.

[0091] Fig. Figure 9 illustrates the migration of charge carrier concentration in an example directional coupler according to an implementation disclosed herein. Fig. 9 shows the directional coupler 800 of the Fig. 8A-8C, where charge carrier concentrations and current sources are superimposed.

[0092] In the Fig. In the example shown in Figure 9, multiple current sources are provided. For example, in the first section 801, a first source 910 is electrically coupled between electrodes 817 and 819, and a second source 920 is electrically coupled between electrodes 817 and 823, such that the first and second sources 910 and 920 have a common ground. In the second section 802, a third source 930 is electrically coupled between electrodes 818 and 820, and a fourth source 940 is electrically coupled between electrodes 818 and 824, such that the third and fourth sources 930 and 940 have a common ground. The current sources 910 and 920 may be DC sources whose negative terminals are connected to electrode 817 and whose positive terminals are connected to electrodes 819 and 823, respectively. The current sources 930 and 940 may also be DC sources whose positive terminals are connected to the electrode 818 and whose negative terminals are connected to the electrodes 820 and 822, respectively.824. Each current source 910-940 can be controlled by a control device (e.g. in the form of a . Fig. 11) to apply a bias voltage between the respective electrodes. That is, a first bias voltage may be applied between electrodes 817 and 819, a second bias voltage between electrodes 817 and 823, a third bias voltage between electrodes 818 and 820, and a fourth bias voltage between electrodes 818 and 824. Each of the first through fourth bias voltages may be applied simultaneously or selectively, as desired. Each bias voltage may be independently controlled across the respective electrodes to modulate each connected component; e.g., each bias voltage may be applied simultaneously or sequentially.

[0093] When the first bias voltage from the first source 910 is applied as a forward bias between the electrodes 817 and 819, negative charges and holes accumulate around the dielectric 815, and the first MOSCAP 811 of the section 801 operates in accumulation mode. As shown in Fig. 9, for example, negative charges migrate from the cathode 807 to the optical energy region 805 (which contains the first optical waveguide 803, as shown in Fig. 8B and Fig. 8C) to accumulate on a top surface of the dielectric 815, and holes (positive charges) migrate from the anode 809 to the first optical waveguide 803 to accumulate on a bottom surface of the dielectric 815. The accumulation of charges and the change in the carrier concentration at the dielectric 815 result in a carrier accumulation effect, which changes the refractive index of the first optical waveguide 803 and the propagation losses, as described above.

[0094] When the third bias voltage from the third source 930 is applied as a reverse bias voltage between the electrodes 818 and 820, the second MOSCAP 812 operates in depletion mode and the concentration of free carriers around the dielectric 815 moves away from the dielectric layer 816 in the optical energy region 806. That is, free carriers (represented as circles with a "+" for positive free carriers or holes and a "-" for negative free carriers or electrons) migrate away from the optical energy region 806, leaving behind a higher proportion of fixed carriers (represented as boxes with a "+" for positive fixed carriers or holes and a "-" for negative fixed carriers or electrons). As shown in Fig. 9, for example, negative charges migrate from the optical energy region 806 to the electrode 818 to reduce the concentration of free carriers and increase the concentration of fixed carriers at the top surface of the dielectric 815. Similarly, holes migrate from the optical energy region 806 to the electrode 820 to reduce the concentration of free carriers and increase the concentration of fixed carriers at a bottom surface of the dielectric 816. The depletion of charge and the change in the carrier concentration at the dielectric 816 results in a plasma dispersion effect that changes the refractive index of the second optical waveguide 804 (e.g., within the optical energy region 806, as shown in the Fig. 8B and Fig. 8C) and changes the propagation loss as described above.

[0095] Accordingly, selective control of the first source 910 and / or the third current source 930 enables modulation of the refractive index of the optical waveguides 803 and / or 804. As described above, by modulating the refractive index of the optical waveguides 803 and / or 804, a refractive index difference between them can be varied to tune the difference as desired. Tuning the refractive index difference, for example, allows for accounting for the attenuation of the gain / loss in the systems and maintaining the power cancellation for the realization of PT symmetry to obtain a PT-symmetric directional coupler.

[0096] When the second bias voltage from the second source 920 is applied as a forward bias between the electrodes 817 and 823, the first mesa 821 serves to introduce optical gain, and control of the second bias voltage enables gain tuning across the first mesa 821. As an illustrative example, negative charges and holes accumulate in the optically active region 825 of the first mesa 821, which serve as a pump source and can cause optical amplification. For example, negative charges migrate from the cathode 807 into the optically active region 825, and holes migrate through the semiconductor layer 827 into the optically active region 825. The accumulation of charges and holes in the optically active region 825 provides energy transition states for generating stimulated emissions, resulting in optical gain in the optical energy range 805, from which light can be emitted.The emitted light propagates through the cathode 807 into the first optical waveguide 803 in the range of optical energy 805. Thus, the optical gain introduced into the first optical waveguide 803 can be modulated by selectively controlling the second bias voltage. Such modulation can be used to tune the gain-to-loss ratio of the directional coupler 800 to locate the exception point and obtain a PT-symmetric directional coupler.

[0097] When the fourth bias voltage is applied by the fourth source 940 as a reverse bias between the electrodes 818 and 824, the second mesa 822 induces optical loss, and controlling the fourth bias voltage enables loss tuning across the second mesa 822. An illustrative example: Free carriers (e.g., negative charges and holes) are removed from the optically active region 826 of the second mesa 821, resulting in an increase in the fixed carrier concentration in the optically active region 826, which bends the electron-hole band diagrams, resulting in a decrease in the band gap energy and thus absorbing higher energy light. This absorption of higher energy light leads to optical loss. Thus, when the fourth bias voltage is applied, higher energy light propagating in the waveguide 804 is absorbed by the cathode 808.The optical loss introduced into the second optical waveguide 804 can thus be modulated by selectively controlling the fourth bias voltage. Such modulation can be used to tune the gain-to-loss ratio of the directional coupler 800 to locate the exception point and obtain a PT-symmetric directional coupler.

[0098] By controlled preload of the directional coupler 800, as described above in conjunction with Fig. 9, the refractive index difference can be tuned to maximize the power cancellations at the outputs of the first and second waveguides 803 and 804, while tuning the gain-loss ratio to enable optical switching. For example, with reference to Fig. 3A-3C, the first and second MOSCAPS 811 and 812 are biased to adjust the directional coupler 800 to the conditions of Fig. 3A. At the same time, the first and second mesas 821 and 822 can be adjusted to allow switching between the inputs. An illustrative example: First light is fed into the first waveguide 803 and second light into the second waveguide 804, and the refractive index difference is so large that the conditions of Fig. 3A in the directional coupler 800. If one wants to switch from the first light from the first waveguide 803 to the second light, the first and second mesas 821 and 822 can be set to a gain-loss ratio of 870 cm -1 be coordinated as in Fig. 3A. At this point, the intensity matrix of the transmitted light is T 11 -50 dB and the output power is essentially equal to all the second light injected into the second waveguide 804. Alternatively, the gain-loss ratio can be set to 499 cm -1 be set as in Fig. 3A, when one wants to switch from the second light output of the first waveguide 804 to second light. At this point, the matrix of the transmitted light intensity T 12 -70 dB, and the output power is substantially equal to all of the first light input to the first waveguide 803.

[0099] Fig. 10 is an example flowchart illustrating an example process for tuning a directional coupler according to the implementations disclosed herein. In this example, the process may be performed by one or more controllers, embodied, for example, as one or more computer systems 1100 of Fig. 11, which are described below. The flowchart of Fig. 10 may be stored as a set of instructions in a machine-readable storage medium that, when executed by a processor, cause the processor to execute the flowchart. The process may be performed to tune a directional coupler, such as the directional coupler 800 used in conjunction with the Fig. 8A-9.

[0100] In block 1002, a first bias voltage is applied to a first MOSCAP. The first MOSCAP may, for example, be the first MOSCAP 811 formed in the first optical waveguide 803. As described above, the first MOSCAP 811 may be formed between the anode 809 and the cathode 807, between which a dielectric 815 is located. The first bias voltage may be supplied by a power source, such as the first source 910 in Fig. 9, can be created.

[0101] In block 1004, a first refractive index of the first waveguide is modulated by adjusting the first bias voltage. As described above, the first refractive index changes based on the carrier concentration in the first MOSCAP as a function of the first bias voltage. That is, the first bias voltage can be adjusted, for example, via one or more control devices, and the carrier concentration in the dielectric layer 815 of the first MOSCAP 811 changes depending on the settings. Based on the changes in the carrier concentration, the refractive index of the waveguide is changed as described in Equations 6 and 7 above.

[0102] In block 1006, a second bias voltage is applied to a first mesa on the first MOSCAP. The first mesa may, for example, be the first mesa 821 disposed on the first MOSCAP 811. As described above, the first mesa 821 includes an optically active region 825 that can be biased to produce optical gain and / or loss depending on the polarity of the bias voltage. The second bias voltage may be supplied by a power source, such as the second source 920 in Fig. 9, can be created.

[0103] In block 1008, the optical gain or loss induced in the first waveguide by the first mesa is modulated by adjusting the second bias voltage. As described above, the gain or loss changes based on the carrier concentration in the optically active region 825 of the first mesa 821 as a function of the second bias voltage. That is, the second bias voltage can be adjusted, for example, via the one or more control devices, and in response to the adjustments, the carrier concentration in the optically active region 825 changes. Based on the changes in the carrier concentration, the gain or loss (depending on the polarity of the bias voltage) is changed as described above.

[0104] In block 1018, a gain-to-loss ratio and a refractive index difference between the first optical waveguide and a second optical waveguide are adjusted based on the modulation of at least the first refractive index in block 1004 and the modulation of the optical gain or loss in block 1008. For example, the modulation of the first refractive index may be used to adjust the refractive index difference between the first optical waveguide 803 and the second optical waveguide 804 to maintain power cancellation. Simultaneously, or sequentially before or after, the optical gain or loss may be modulated to adjust the gain-to-loss ratio between the first and second optical waveguides 803 and 804 to find or exceed the exception point.By tuning the gain-loss ratio together with the refractive index difference, a PT-symmetric directional coupler can be obtained, for example, in block 1020, where the gain-loss ratio is tuned to be equal to or exceed the exception point, while the refractive index difference is tuned to maintain power cancellation.

[0105] Optionally, in block 1010, a third bias voltage may be applied to a second MOSCAP. The second MOSCAP may, for example, be the second MOSCAP 812 formed in the second optical waveguide 804. As described above, the second MOSCAP 812 may be formed between anode 810 and cathode 808 with a dielectric 816 therebetween. The third bias voltage may be supplied by a power source, such as the third source 930 in Fig. 9, can be created.

[0106] Optionally, in block 1012, a second refractive index of the second waveguide can be modulated by adjusting the third bias voltage. As described above, the second refractive index changes based on the carrier concentration in the second MOSCAP as a function of the third bias voltage. That is, the third bias voltage can be adjusted, for example, via one or more control devices, and the carrier concentration in the dielectric layer 816 of the second MOSCAP 812 changes depending on the settings. Based on the changes in the carrier concentration, the refractive index of the second waveguide is changed as described in Equations 6 and 7 above.

[0107] Optionally, in block 1014, a fourth bias voltage may be applied to a second mesa disposed on the second MOSCAP. The second mesa may, for example, be the second mesa 822 disposed on the second MOSCAP 812 (and thus on the second optical waveguide 804). As described above, the second mesa 822 includes an optically active region 826 that can be biased to produce optical gain and / or loss depending on the polarity of the bias voltage. The fourth bias voltage may be provided by a current source, such as the fourth source 940 of Fig. 9, can be created.

[0108] Optionally, in block 1016, the optical gain or loss induced in the second waveguide by the second mesa is modulated by adjusting the fourth bias voltage. As described above, the gain or loss changes based on the carrier concentration in the optically active region 826 of the second mesa 822 as a function of the fourth bias voltage. That is, the fourth bias voltage can be adjusted, for example, via the one or more control devices, and the carrier concentration in the optically active region 826 changes as a function of the settings. Based on the changes in the carrier concentration, the gain or loss (depending on the polarity of the bias voltage) is changed as described above.

[0109] The operations of block 1018 may include the modulations of blocks 1016 and 1012, if such processes are performed. For example, modulation of the first and second refractive indices may be used to tune the refractive index difference between the first and second optical waveguides in block 1018. Furthermore, modulation of the gain and loss across the first and second mesas 821 and 822 enables tuning of the gain-to-loss ratio in block 1018.

[0110] Fig. 11 shows a block diagram of an example computer system 1100 in which various embodiments described herein may be implemented. Computer system 1100 includes a bus 1102 or other communication mechanism for conveying information, and one or more hardware processors 1104 connected to bus 1102 for processing information. Hardware processor(s) 1104 may be, for example, one or more general-purpose microprocessors. As described above, computer system 1100 may be implemented to control the power sources described above (e.g., one or more of power sources 910, 920, 930, and / or 940).

[0111] Computer system 1100 also includes main memory 1106, such as random access memory (RAM), a cache, and / or other dynamic storage devices, connected to bus 1102 for storing information and instructions to be executed by processor 1104. Main memory 1106 may also be used to store temporary variables or other intermediate information during the execution of instructions to be executed by processor 1104. When such instructions are stored in storage media accessible to processor 1104, computer system 1100 becomes a special-purpose machine adapted to perform the operations specified in the instructions.

[0112] Computer system 1100 also includes a read-only memory (ROM) 1108 or other static storage device connected to bus 1102 to store static information and instructions for processor 1104. A storage device 1110, such as a magnetic disk, an optical disk, or a USB stick (flash drive), etc., is provided and connected to bus 1102 to store information and instructions.

[0113] Computer system 1100 may be connected via bus 1102 to a display 1112, e.g., a liquid crystal display (LCD) (or a touch screen), for displaying information to a computer user. An input device 1114, including alphanumeric and other keys, is coupled to bus 1102 to communicate information and command selections to processor 1104. Another type of user input device is cursor control 1116, e.g., a mouse, trackball, or cursor direction keys for communicating direction information and command selections to processor 1104 and controlling cursor movement on display 1112. In some embodiments, the same direction information and command selections as with cursor control may be implemented via receiving touches on a touchscreen without a cursor.

[0114] Computer system 1100 may include a user interface module for implementing a graphical user interface, which may be stored on a mass storage device as executable software code executed by the computing device(s). This and other modules may include, for example, components such as software components, object-oriented software components, class components and task components, processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables.

[0115] In general, the word "component," "engine," "system," "database," "data store," and the like, as used herein, may refer to logic embodied in hardware or firmware, or to a collection of software instructions that may have entry and exit points and may be written in a programming language such as Java, C, or C++. A software component may be compiled and linked into an executable program, installed in a dynamic link library, or written in an interpreted programming language such as BASIC, Perl, or Python. It is understood that software components may be callable by other components or by themselves, and / or may be called in response to detected events or interrupts. Software components configured to run on computing devices may be embodied on a computer-readable medium, such asa compact disc, digital video disc, flash drive, magnetic disk, or other tangible medium, or as a digital download (and may be originally stored in a compressed or installable format that must be installed, decompressed, or decrypted before execution). Such software code may be stored partially or entirely in a memory of the executing computing device so that it can be executed by the computing device. Software instructions may be embedded in firmware, such as an EPROM. In addition, the hardware components may consist of interconnected logic units, such as gates and flip-flops, and / or programmable units, such as programmable gate arrays or processors.

[0116] Computer system 1100 may implement the techniques described herein using custom hard-wired logic, one or more ASICs or FPGAs, firmware, and / or program logic that, in combination with the computer system, causes or programs computer system 1100 to be a special-purpose machine. According to one embodiment, the techniques described herein are performed by computer system 1100 in response to processor(s) 1104 executing one or more sequences of one or more instructions contained in main memory 1106. Such instructions may be read into main memory 1106 from another storage medium, such as storage device 1110. Execution of the instruction sequences contained in main memory 1106 causes processor(s) 1104 to perform the process steps described herein.In alternative embodiments, hard-wired circuits may be used instead of or in combination with software instructions.

[0117] The term "non-transitory media" and similar terms as used herein refer to any media that stores data and / or instructions that cause a machine to operate in a particular manner. Such non-transitory media may include non-volatile media and / or volatile media. Non-volatile media includes, for example, optical or magnetic disks, such as storage device 1110. Volatile media includes dynamic memory, such as main memory 1106. Common forms of non-transitory media include, for example, floppy disks, flexible disks, hard disks, solid-state drives, magnetic tape or other magnetic data storage media, CD-ROMs, other optical data storage media, physical media with hole patterns, RAM, PROM and EPROM, FLASH EPROM, NVRAM, other memory chips or cartridges, and networked versions thereof.

[0118] Non-transitory media are distinct from transmission media but can be used in conjunction with them. Transmission media are involved in the transfer of information between non-transitory media. Examples of transmission media include coaxial cable, copper wire, and fiber optic cable, including the wires that make up bus 1102. Transmission media can also take the form of sound or light waves, such as those generated in radio and infrared data communications.

[0119] Computer system 1100 also includes a communications interface connected to bus 1102. Communications interface 1118 establishes a two-way data communications connection to one or more network connections connected to one or more local area networks. For example, communications interface 1118 may be an Integrated Services Digital Network (ISDN) card, a cable modem, a satellite modem, or a modem for establishing a data communications connection to a corresponding type of telephone line. As another example, network interface 1118 may be a Local Area Network (LAN) card for establishing a data communications connection to a compatible LAN (or a WAN component for communicating with a WAN). Wireless connections may also be implemented.In each of these implementations, the communication interface 1118 sends and receives electrical, electromagnetic, or optical signals that carry digital data streams containing various types of information. The communication interface 1118 may be communicatively coupled to one or more power sources (e.g., one or more of the power sources 450, 460, 550, 560, 650, and 660 of the . Fig. 4-6).

[0120] A network connection typically enables data communication across one or more networks to other data devices. For example, a network connection may connect across a local area network to a host computer or data devices operated by an Internet service provider (ISP). The ISP, in turn, provides data communication services over the worldwide packet data communications network, now commonly referred to as the "Internet." Both the local area network and the Internet use electrical, electromagnetic, or optical signals that carry digital data streams. The signals in the various networks and the signals on the network connection and across the communications interface 1118 that carry the digital data to and from the computer system 1100 are examples of transmission media.

[0121] Computer system 1100 can send messages and receive data, including program code, over the network(s), the network connection, and the communications interface 1118. In the Internet example, a server could transmit requested code for an application program over the Internet, the ISP, the local network, and the communications interface 1118.

[0122] The received code may be executed by processor 1104 when received and / or stored in storage device 1110 or other non-volatile memory for later execution. Accordingly, computer system 1100 may be configured to control the bias voltage applied between the contact electrodes by sending instructions to one or more power sources via communication interface(s) 1118 according to the instructions stored in main memory 1106 and / or ROM 1108.

[0123] Each of the processes, methods, and algorithms described in the preceding sections may be embodied in, and fully or partially automated by, code components executed by one or more computer systems or computer processors comprising computer hardware. The one or more computer systems or computer processors may also operate to support the performance of the corresponding operations in a cloud computing environment or as software as a service (SaaS). The processes and algorithms may be partially or fully implemented in application-specific circuitry. The various features and methods described above may be used independently of one another or combined in various ways.Various combinations and sub-combinations are intended to be within the scope of this disclosure, and certain method or process blocks may be omitted in some implementations. The methods and processes described herein are also not limited to any particular sequence, and the associated blocks or states may be performed in other suitable sequences, in parallel, or otherwise. Blocks or states may be added to or removed from the disclosed examples. The execution of certain operations or processes may be distributed among computer systems or computer processors located not only on a single machine, but distributed across a number of machines.

[0124] As used herein, a circuit may be implemented in any form of hardware, software, or a combination thereof. For example, one or more processors, controllers, ASICs, PLAs, PALs, CPLDs, FPGAs, logic components, software routines, or other mechanisms may be implemented to form a circuit. In implementation, the various circuits described herein may be implemented as discrete circuits, or the described functions and features may be distributed, in part or in whole, among one or more circuits.Although various features or functional elements are individually described or claimed as separate circuits, these features and functions may be shared by one or more common circuits, and such description is not intended to assume or imply that separate circuits are required to implement these features or functions. If a circuit is implemented in whole or in part with software, that software may be implemented to operate with a computer or processing system capable of performing the functionality described with respect to it, such as computer system 1100.

[0125] As used herein, the term "or" can be interpreted both inclusively and exclusively. Furthermore, descriptions of resources, acts, or structures in the singular should not be construed to exclude the plural. Conditional expressions such as "may" or "could," unless expressly stated otherwise or understood by context, are generally intended to convey that certain embodiments include certain features, elements, and / or steps, while other embodiments do not.

[0126] Unless expressly stated otherwise, the terms and expressions used in this document, as well as variations thereof, are not to be interpreted as limiting, but as open-ended. Adjectives such as "conventional," "traditional," "normal," "standard," "known," and terms of similar import are not to be construed as limiting the subject matter described to a particular period of time or to a subject matter available at a particular time, but should be understood to include conventional, traditional, normal, or standard technologies that may be available or known now or at any time in the future.The presence of broader words and phrases such as “one or more,” “at least,” “but not limited to,” or similar phrases in some cases should not be construed as meaning that the narrower case is intended or required in the absence of such broader phrases.

Claims

[1] A hybrid III-V / silicon device comprising: a first silicon layer disposed over a buried oxide layer, BOX layer (832), the first silicon layer comprising: a first doped region with a first trench (813), a second doped region with a second trench (814), a first gap region (836) disposed between the first doped region and the second doped region; a first oxide layer (815) disposed over the first doped region and the second doped region; a first mesa (821) disposed on the first doped region of the first silicon layer, the first mesa comprising: a first Group III-V layer (807) disposed over the first oxide layer; a first optically active region (825) disposed over the first group III-V layer; and a second group III-V layer (827) disposed over the optically active region; and a second mesa (822) disposed on the second doped region of the first silicon layer, the second mesa comprising: a third Group III-V layer (808) disposed over the first oxide layer; a second optically active region (826) disposed over the third group III-V layer; and a fourth group III-V layer (828) disposed over the second optically active region. [2] A hybrid III-V / silicon device according to claim 1, wherein the first oxide layer comprises at least one of HfO 2 , Al 2 O 3 and VO 2 includes. [3] The hybrid III-V / silicon device of claim 1, wherein the first and second optically active regions are either a quantum dot layer, a quantum well layer, or a quantum well layer. [4] The hybrid III-V / silicon device of claim 1, wherein the first, second, third and fourth group III-V layers comprise gallium arsenide and / or indium phosphide. [5] A hybrid III-V / silicon device according to claim 1, further comprising: a first waveguide (803) disposed in the first doped region between the first trench and the gap region; and a second waveguide (804) disposed in the second doped region between the second trench and the gap region. [6] A hybrid III-V / silicon device according to claim 1, wherein the second group III-V layer comprises a first ion region (829) and the fourth group III-V layer comprises a second ion region (830). [7] A hybrid III-V / silicon device according to claim 1, wherein the gap region is arranged between the first and second mesa, and an isolation region (838) is disposed on the first gap region between the first and second mesa. [8] A hybrid III-V / silicon hybrid device according to claim 7, wherein the isolation region comprises a third ion region (838a). [9] The hybrid III-V / silicon device of claim 7, wherein the isolation region comprises a second gap region (838b). [10] The hybrid III-V / silicon device of claim 1, wherein the first and second doped regions are doped with a first polarity and the first, second, third and fourth group III-V layers are doped with a second polarity different from the first polarity. [11] The hybrid III-V / silicon device of claim 10, wherein the first polarity is p-type and the second polarity is n-type. [12] A directional coupler (800) comprising: a buried oxide layer, BOX layer (832); a first metal oxide semiconductor MOS capacitor (811) formed on the BOX layer, the first MOS capacitor comprising: a first optical waveguide (803); a first cathode (807) comprising a first material and formed in the first optical waveguide; a first anode (809) formed in the first optical waveguide, the first anode comprising a second material different from the first material; and a first dielectric (815) disposed between the first cathode and the first anode, the first dielectric comprising an oxide of the first material and an oxide of the second material, the first MOS capacitor being defined between the first anode and the first cathode; and a first mesa (821) disposed on the first MOS capacitor, the first mesa comprising: the first anode; a first optically active region (825) disposed on the first anode; and a first semiconductor layer (827) comprising the second material and disposed on the optically active region; and a second optical waveguide (804) disposed on the BOX layer adjacent to the first optical waveguide and having a first gap (836) between the first and second optical waveguides. [13] Directional coupler according to claim 12, further comprising: a second MOS capacitor (812) formed on the BOX layer, the second MOS capacitor comprising: the second optical waveguide; a second cathode (810) comprising the first material and formed in the second optical waveguide; a second anode (808) formed in the second optical waveguide, the second anode comprising the second material; and a second dielectric (816) disposed between the second cathode and the second anode, the second dielectric comprising the oxide, the second MOS capacitor being defined between the second anode and the second cathode. [14] Directional coupler according to claim 12, further comprising: a second mesa (822) disposed on the second waveguide, the second mesa comprising: the second anode formed in the second optical waveguide, the second anode comprising the second material; a second optically active region (826) disposed on the first anode; and a second semiconductor layer (828) comprising the second material and disposed on the optically active region. [15] A method for tuning a directional coupler, the method comprising: Applying a first bias voltage to a first metal oxide semiconductor MOS capacitor (811), the first MOS capacitor being formed in a first optical waveguide (803); Modulating a first refractive index of the first optical waveguide by adjusting the first bias voltage, wherein the first refractive index changes based on the carrier concentration in the first MOS capacitor in response to the first bias voltage; Applying a second bias voltage to a first mesa (821) disposed on the first MOS capacitor, the first mesa comprising a first optically active region (825); Modulating the optical gain or optical loss introduced into the first optical waveguide by the first mesa by adjusting the second bias voltage, wherein the optical gain or optical loss changes based on the carrier concentration in the first optically active region in response to the second bias voltage; and Tuning the gain-loss ratio and the refractive index difference between the first optical waveguide and a second optical waveguide (804) based on the modulation of the first refractive index and the optical gain or the optical loss. [16] The method of claim 15, wherein: the first and second bias voltages are forward voltages and optical gain is introduced through the first mesa into the first optical waveguide based on the adjustment of the second bias voltage. [17] The method of claim 15, further comprising: Applying a third bias voltage to a second metal oxide semiconductor MOS capacitor (812), the second MOS capacitor being formed in the second optical waveguide; and Modulating a second refractive index of the second optical waveguide by adjusting the third bias voltage, wherein the second refractive index changes based on the carrier concentration in the second MOS capacitor in response to the third bias voltage. [18] The method of claim 15, further comprising: Applying a fourth bias voltage to a second mesa (822) disposed on the second optical waveguide, the second mesa comprising a second optically active region (826); and Modulating the optical gain or optical loss introduced into the second optical waveguide by the second mesa by adjusting the fourth bias voltage, wherein the optical gain or optical loss changes based on the carrier concentration in the second optically active region in response to the fourth bias voltage. [19] The method of claim 18, wherein: the fourth bias voltage is a reverse bias voltage and optical loss through the second mesa into the second optical waveguide based on the adjustment of the fourth bias voltage. [20] The method of claim 15, further comprising: Achieving a parity-time symmetric directional coupler based on tuning the gain-loss ratio and the refractive index difference between the first optical waveguide and the second optical waveguide.

Images