Optical modulators

Abstract

An optoelectronic device. The optoelectronic device operable to provide a PAM-N modulated output, the device comprising: M optical modulators, M being an integer greater than 1, the M optical modulators being arranged in a cascade, the device being configured to operate in N distinct transmittance states, as a PAM-N modulator, wherein, in each transmittance state of the N distinct transmittance states, each of the M optical modulators has applied to it a respective control voltage equal to one of: a first voltage or a second voltage. One or more of the modulators may include a substrate; a crystalline cladding layer, on top of the substrate; and an optically active region, above the crystalline cladding layer. The crystalline cladding layer may have a refractive index which is less than a refractive index of the optically active region.

Claims

1. An optoelectronic device operable to provide a PAM-N modulated output, the device comprising: M optical modulators, M being an integer greater than 1, the M optical modulators being arranged in a cascade, the device being configured to operate in N distinct transmittance states, as a PAM-N modulator, wherein, in each transmittance state of the N distinct transmittance states, each of the M optical modulators has applied to it a respective control voltage equal to one of: a first voltage or a second voltage, wherein one of the M optical modulators comprises: a substrate; a crystalline cladding layer being crystalline through its entire thickness, on top of the substrate; and an optically active region, above the crystalline cladding layer, wherein the crystalline cladding layer has a refractive index which is less than a refractive index of the optically active region, wherein the optoelectronic device is on a silicon on insulator (SOI) wafer comprising the substrate, and wherein the optically active region and the substrate are part of a single crystal.

2. The device of claim 1, wherein M=N−1.

3. The device of claim 1, wherein: N=4; M=3; the first voltage is between 0 V and 0.2 V; and the second voltage is between 1.8 V and 2.0 V.

4. The device of claim 1, wherein an optical modulator of the M optical modulators is an electro-absorption modulator.

5. The optoelectronic device of claim 1, further comprising an intermediate waveguide, disposed between a first optical modulator of the M optical modulators and a second optical modulator of the M optical modulators, and operable to convey electromagnetic waves from the first optical modulator to the second optical modulator.

6. The optoelectronic device of claim 5, wherein a first interface between the first optical modulator and the intermediate waveguide is at a first angle relative to a guiding direction of the intermediate waveguide, and a second interface between the second optical modulator and the intermediate waveguide is at an opposite angle to the first angle, wherein the first angle is not a right angle and the opposite angle is not a right angle.

7. The optoelectronic device of claim 1, wherein each of the M optical modulators comprises an optically active region of a ridge waveguide.

8. The optoelectronic device of claim 1, wherein: the optoelectronic device has a first transmittance in a first transmittance state of the N distinct transmittance states; the optoelectronic device has a second transmittance in a second transmittance state of the N distinct transmittance states; and the first transmittance is one half of the second transmittance.

9. The optoelectronic device of claim 1, further comprising a heater, such that the M optical modulators are tuneable with respect to wavelength.

10. The optoelectronic device of claim 1, wherein: in a first transmittance state of the N distinct transmittance states, all of the M optical modulators have applied to them a control voltage equal to the first voltage; and in a second transmittance state of the N distinct transmittance states, all of the M optical modulators have applied to them a control voltage equal to the second voltage.

11. The optoelectronic device of claim 1, wherein M=2.

12. The optoelectronic device of claim 1, wherein the refractive index of the crystalline cladding layer is at most 0.95 times the refractive index of the optically active region.

13. An optoelectronic device operable to provide a PAM-N modulated output, the device comprising: M optical modulators, M being an integer greater than 1, the M optical modulators being arranged in a cascade, the device being configured to operate in N distinct transmittance states, as a PAM-N modulator, wherein, in each transmittance state of the N distinct transmittance states, each of the M optical modulators has applied to it a respective control voltage equal to one of k different control voltages, wherein k is an integer greater than 1 and less than N, and wherein one of the M optical modulators comprises: a substrate; a crystalline cladding layer being crystalline through its entire thickness, on top of the substrate; and an optically active region, above the crystalline cladding layer, wherein the crystalline cladding layer has a refractive index which is less than a refractive index of the optically active region, wherein the optoelectronic device is on a silicon on insulator (SOI) wafer comprising the substrate, and wherein the optically active region and the substrate are part of a single crystal.

14. The optoelectronic device of claim 13, wherein the M optical modulators are electro-absorption modulators.

15. The optoelectronic device of claim 13, further comprising an intermediate waveguide, disposed between a first optical modulator of the M optical modulators and a second optical modulator of the M optical modulators and operable to convey electromagnetic waves from the first optical modulator to the second optical modulator.

16. The optoelectronic device of claim 15, wherein a first interface between the first optical modulator and the intermediate waveguide is at a first angle relative to a guiding direction of the intermediate waveguide, and a second interface between the second optical modulator and the intermediate waveguide is at an opposite angle to the first angle, wherein the first angle is not a right angle and the opposite angle is not a right angle.

17. The optoelectronic device of claim 13, wherein each of a first optical modulator of the M optical modulators and a second optical modulator of the M optical modulators comprises an optically active region of a ridge waveguide.

18. The optoelectronic device of claim 13, wherein: the optoelectronic device has a first transmittance in a first transmittance state of the N distinct transmittance states; the optoelectronic device has a second transmittance in a second transmittance state of the N distinct transmittance states; and the first transmittance is one half of the second transmittance.

19. The optoelectronic device of claim 13, further comprising a heater, such that an optical modulator of the M optical modulators is tuneable with respect to wavelength.

20. The optoelectronic device of claim 13, wherein: in a first transmittance state of the N distinct transmittance states, all of the M optical modulators have applied to them a control voltage equal to a first voltage; and in a second transmittance state of the N distinct transmittance states, all of the M optical modulators have applied to them a control voltage equal to a second voltage.

21. The optoelectronic device of claim 13, wherein M is greater than or equal to Log.sub.2(N).

22. The optoelectronic device of claim 13, wherein the refractive index of the crystalline cladding layer is at most 0.95 times the refractive index of the optically active region.

23. An optoelectronic device operable to provide a PAM-N modulated output, the device comprising: M optical modulators, M being an integer greater than 1, the M optical modulators being arranged in a cascade, the device being configured to operate in N distinct transmittance states, as a PAM-N modulator, wherein, in each transmittance state of the N distinct transmittance states, each of the M optical modulators has applied to it a respective control voltage equal to one of k different control voltages, wherein k is an integer greater than 1 and less than N, and wherein one of the M optical modulators comprises: a substrate; a crystalline cladding layer being crystalline through its entire thickness, on top of the substrate; and an optically active region, above the crystalline cladding layer, wherein the crystalline cladding layer has a refractive index which is less than a refractive index of the optically active region, and wherein: N=4; M=2; and k=3.

24. The optoelectronic device of claim 23, wherein: in a first transmittance state of the 4 distinct transmittance states: a first optical modulator of the 2 optical modulators has a transmittance that is a maximum transmittance of the first optical modulator; and a second optical modulator of the 2 optical modulators has a transmittance that is a maximum transmittance of the second optical modulator; in a second transmittance state of the 4 distinct transmittance states: the first optical modulator has a first transmittance, the first transmittance being less than the maximum transmittance of the first optical modulator; and the second optical modulator has the maximum transmittance of the second optical modulator; in a third transmittance state of the 4 distinct transmittance states: the first optical modulator has the maximum transmittance of the first optical modulator; and the second optical modulator has a second transmittance, the second transmittance being less than the first transmittance; and in a fourth transmittance state of the 4 distinct transmittance states: the first optical modulator has a third transmittance, the third transmittance being less than the first transmittance; and the second optical modulator has the second transmittance.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Embodiments of the invention will now be described by way of example with reference to the accompanying drawings in which:

(2) FIG. 1 is an eye diagram showing the modulation states of a PAM-4 modulated signal;

(3) FIG. 2 is a schematic view of an optoelectronic device capable of PAM-4 modulation;

(4) FIG. 3 is a detailed schematic of the device of FIG. 2;

(5) FIGS. 4A and 4B schematically show variant waveguide ridge structures;

(6) FIG. 5 schematically shows a further variant waveguide ridge structures;

(7) FIG. 6 shows a variant waveguide ridge structure;

(8) FIG. 7 shows an example of an input or output waveguide;

(9) FIG. 8 shows a variant waveguide ridge structure;

(10) FIG. 9 shows a further variant waveguide ridge structure;

(11) FIG. 10 shows a further variant waveguide ridge structure;

(12) FIG. 11 shows a variant optoelectronic device capable of PAM-4 modulation;

(13) FIGS. 12A and 12B show further variant optoelectronic devices;

(14) FIGS. 13A and 13B show further variant optoelectronic devices;

(15) FIGS. 14A and 14B show further variant optoelectronic devices;

(16) FIG. 15 shows a further variant optoelectronic device;

(17) FIG. 16 shows an optoelectronic device according to FIG. 2 including a heater;

(18) FIGS. 17A and 17B show simulation results for a first optoelectronic device;

(19) FIGS. 18A and 18B show simulation results for a second optoelectronic device;

(20) FIG. 19 shows a variant optoelectronic device;

(21) FIG. 20 shows a further variant optoelectronic device; and

(22) FIG. 21 shows a further variant optoelectronic device.

DETAILED DESCRIPTION

(23) FIG. 2 shows, schematically, an optoelectronic device capable of PAM-4 modulation. Electromagnetic waves 101 may be presented to an input waveguide 102 at a first end of the device. The waves are guided from the input waveguide, through a tapered region 103 and into the first optical modulator 104. as the optical modulator may be an electro-optical modulator (EOM) in that it may utilize an electro-optical effect (for example, the Franz-Keldysh effect if the electro-optical modulator was an electro-absorption modulator). The first optical modulator has an associated length L.sub.1 105 along which the waves are guided, as well as an associated waveguide width W.sub.1 106 which is substantially perpendicular to the length. The first optical modulator is operable, via electrodes 107, to operate in various transmittance states, by applying to this and each optical modulator a respective control voltage selected from three different control voltages (e.g., 0 V, 1.4 V, and 2 V, as shown in the list below). For example, the optical modulator may be operable in either a maximum transmittance state or in a state with less than maximum transmittance, depending on the desired output of the device. In the examples where the optical modulator is an electro-absorption modulator, the optical modulator may be operated in either (i) a transmittance state of maximum transmittance or (ii) a transmittance state of less than maximum transmittance. By switching between two or more such transmittance states, the optical modulator may modulate the light propagating through the device, where “modulating” the light propagating through the device means imposing a time-varying attenuation on light transmitted through the device. The electrodes may be operable in three states for the first optical modulator, or in two states (for the second optical modulator), each state corresponding to a control voltage. For example, the optical modulators may have applied to them the control voltages listed below, to produce the four PAM-4 levels: Level 3: ER4=0 dB: V1=0V, V2=0V Level 2: ER3=1.05 dB: V1=1.4V, V2=0V Level 1: ER2=2.44 dB: V1=0V, V2=2V Level 0: ER1=4.5 dB: V1=2V, V2=2V.
In the list above, ER (e.g., ER1, ER2, etc.) is the respective extinction ratio of the optoelectronic device (and the transmittance is the opposite, in dB, e.g., an extinction ratio of 2.44 dB corresponds to a transmittance of −2.44 dB). In an alternate embodiment, Level 2 is produced using V1=1V and V2=1V. Each optical modulator may have maximum transmittance when the applied control voltage is 0V, and reduced transmittance for higher control voltages. As used herein, an “optical modulator” is an optical element with electrically controlled optical transmittance, such as an electro-absorption modulator (EAM) as discussed above.

(24) The waves are then guided into a first end of an intermediate waveguide 108, which has a taper so as to match the waveguide width of a second optical modulator 109 disposed at the opposing end. The waveguide width may affect the degree to which light is attenuated by any given optical modulator. The intermediate waveguide 108 can be formed of either undoped Si or undoped SiGe and may have a length of between 0.5 μm and 10 μm or 1 μm and 2 μm. The second optical modulator has an associated length L.sub.2 110 along which the waves are guided, as well as an associated waveguide width W.sub.2 111 which is substantially perpendicular to the length. The length and waveguide width of the second optical modulator is different to the length and waveguide width of the first optical modulator. In this example, this allows the first and second optical modulators to operate in different transmittance states as the length and width of the optical modulators is a factor in determining the degree to which electromagnetic waves are attenuated by the optical modulator. As with the first optical modulator, the second optical modulator is controllable by an electrode 112. Table 1 below show indicative values for the length and waveguide widths of the optical modulator as well as the length of the intermediate waveguide:

(25) TABLE-US-00001 TABLE 1 Parameter Base design value Range L.sub.1 15-20 [μm] 5-50 [μm] L.sub.2 35-50 [μm] 5-80 [μm] d.sub.1 1-2 [μm] 0.5-10 [μm] W.sub.1 650-750 [nm] 450-1100 [nm] W.sub.2 700-800 [nm] 450-1100 [nm]

(26) The waves are then guided into an output waveguide 113, and exit the device. In the lower-right corner of FIG. 2 an example 115 of transmittance states is shown indicating the optical power output corresponding to various transmittance states of the device. Level 4 corresponds to maximum transmittance when both optical modulators have zero control voltages (i.e. 0 V and 0 V) applied to them. Level 3 and 2 correspond to the cases where a non-zero control voltage is applied to only one optical modulator while the other has 0 V applied to it, and finally level 1 corresponds to both the first and second optical modulators having non-zero control voltage applied to them. By design all of the 4 optical output power levels are equally spaced.

(27) FIG. 3 shows in more detail the device 200 shown in FIG. 2. In this example, the input waveguide 201 is formed of Si and connects via tapered region 202 with a waveguide ridge comprising SiGe having width w.sub.1 209. The waveguide ridge could instead comprise any one or more of Ge, SiGeSn, or GeSn. The waveguide ridge has a lightly doped region of a first species 204, and adjacent to this, in a slab portion, is a heavily doped region 205 of the same species. Opposing this in the width direction, and again present in the ridge waveguide, is a lightly doped region of a second species 206 which is adjacent to, in the slab portion, a heavily doped region 207 of the same second species. The heavily doped regions 205, 207 are connected to respective electrodes 210a and 210b. The lightly doped regions have a length L.sub.1 208 which corresponds to the length of the first optical modulator.

(28) Adjacent to a distal end of both of the first optical modulator's lightly doped regions is an intermediate waveguide 211 which can be formed of either undoped SiGe or undoped Si. As the second optical modulator 213 has a different waveguide width 219 to the first optical modulator 203, the intermediate waveguide acts to taper from the first width to the second width. Therefore, the waves are guided from the first optical modulator, through the intermediate waveguide and into the second optical modulator.

(29) The second optical modulator 203, like the first optical modulator has lightly and heavily doped regions. Innermost, and within a waveguide ridge of the second optical modulator, is a region 214 lightly doped with a first species and a region 216 lightly doped with a second species. These regions oppose each other in the width direction. Adjacent to each lightly doped region, and within a slab, are respective heavily doped regions 215 and 217. The region 215 is heavily doped with the same species of dopant as lightly doped region 214, and region 217 is heavily doped with the same species of dopant as lightly doped region 216.

(30) At a distal end of the second optical modulator 203 to the intermediate waveguide 211 is an output waveguide 222. The output waveguide has a tapered region 221, which tapers from the waveguide width w.sub.2 219 of the second optical modulator to a second width.

(31) FIGS. 4A and 4B show variant waveguide ridge structures that could be used in either optical modulator. Each is formed of a Si slab which provides a slab layer. In the example shown in FIG. 4A the Si slab has heavily doped regions 302 and 303 which are doped with a different polarity dopant e.g. N++ and P++ or P++ and N++ respectively. The slab has been etched to provide a ridge portion of Si, and the heavily doped regions partially extend along this ridge into regions 307 and 308. On top of the ridge portion of Si is a SiGe (or any other optically active material) ridge 304, which has been lightly doped in regions 305 and 306. The dopant in region 305 has the same polarity as the dopant in region 302, but its concentration is reduced in comparison. Similarly, the dopant in region 306 has the same polarity as the dopant in region 303 but its concentration is also reduced in comparison.

(32) FIG. 4A is annotated with various dimensions. These include h.sub.wg—the total height of the ridge waveguide; h.sub.2—the height of the slab; h.sub.3—the height along which the heavily doped regions extend; d.sub.n—the depth into the sidewall of the ridge that the first dopant penetrates; d.sub.p—the depth into the sidewall ridge that the second dopant penetrates; d.sub.np—the depth into the sidewall that the heavily doped first dopant penetrates; and d.sub.pp the depth into the sidewall that the heavily doped second doped penetrates. h.sub.wg may have a value of around 3 μm (+/−5%).

(33) The tolerances of the values are shown in table 2 below, and table 3 indicates example dopant ranges for example dopants:

(34) TABLE-US-00002 TABLE 2 Dimension Tolerance [nm] h.sub.wg 100-800  h.sub.2 100-400  h.sub.3  0-400 d.sub.np 50-300 d.sub.pp 50-300 d.sub.p 50-300 d.sub.n 50-300

(35) TABLE-US-00003 TABLE 3 Doping type Doping range [cm.sup.−3] N 1 × 10.sup.15 to 1 × 10.sup.20 P 1 × 10.sup.15 to 1 × 10.sup.20 N++ 1 × 10.sup.18 to 1 × 10.sup.20 P++ 1 × 10.sup.18 to 1 × 10.sup.20

(36) A variant ridge waveguide structure is shown in FIG. 4B. In this example, the Si slab layer 301 has a SiGe layer 304 immediately on top. The SiGe layer is doped to produce the heavily doped regions 302 303 as well as the lightly doped regions 305 306. The lightly doped regions are, again, present in a ridge 304 of the ridge waveguide structure. In this example, the heavily doped regions do not extend up the ridge of the waveguide. Of course, the features described in relation to the first variant (FIG. 4A) are applicable where appropriate to the second variant. For example, a SiGe layer 304 could be provided atop the Si slab layer 301 in the first variant, where the heavily doped regions partially extend up the side of the ridge 304 of the waveguide.

(37) FIG. 5 shows a further variant ridge waveguide structure. FIG. 5 is an example where the heavily doped regions (N++ and P++) are provided in the Si slab layer, and do not extend into the SiGe ridge.

(38) FIG. 6 shows a variant ridge waveguide structure. This structure shares a number of features with previously discussed ridge waveguide structures, and so like reference numerals are used for like parts. The structure in FIG. 6 differs from that in FIG. 4A in that a Si, SiGe, or SiGeSn layer 601 is provided between the ridge 304 and each of the lightly doped regions 305 and 306. This layer 601 should have a lower refractive index with respect to the material making up the ridge 304. The distances d.sub.np2 and d.sub.pp2 indicate, respectively, the depth into the slab 301 that the heavily doped regions are doped. The tolerance on this measurement should be between 50 and 300 nm.

(39) FIG. 7 shows an example of an input or output Si passive waveguide 201 or 222, The input and output waveguides respectively comprise a slab portion 801 having height h.sub.1 and a ridge portion 802, the ridge and slab having a combined height of h.sub.wg. h.sub.1 should have a tolerance of between 100-800 nm, and h.sub.wg should have a value of around 3 μm (+ or −5%).

(40) FIG. 8 shows a variant ridge waveguide structure. Here, instead of the doped regions extending vertically up sidewalls of the ridge waveguide, they are disposed in upper and lower doped regions of the waveguide ridge. As can be seen in FIG. 8, an N doped region extends across the top of the central portion of the waveguide, and is connected to an N+ doped region which is in turn connected to an electrode contact. Similarly, a P doped region extends below the central portion of the waveguide, and is connected to a P+ doped region which is in turn connected to a separate electrode contact. The lower doped region exhibits a multilayer structure. The multilayer structure comprises a first doped region (i.e. a first layer) formed from an implanted doped portion of the silicon-on-insulator (SOI) located directly below the optically active region (OAR). A second doped zone (i.e. a second layer) is formed by implanting dopants into the OAR itself at a region of the OAR located directly above the first doped zone. The second doped zone has a dopant concentration greater than that of the first doped zone. The lower surface of the second doped zone forms the interface between the first doped zone and the second doped zone. The upper surface of the second doped zone forms the contact surface for a corresponding electrode.

(41) As discussed above, the first doped zone of the lower doped region is P doped, and the second doped zone of the lower doped region is P+ doped (where P+ denotes a P doped region with a greater concentration of P dopants). The upper doped region contains an upper doped region in the form of an N doped region which comprises: an upper N doped waveguide region extending across the upper surface of the OAR waveguide; a lateral N doped region which extends outwards away from the waveguide; and a connecting N doped region which extends vertically along a side of the waveguide to connect the upper N doped waveguide region with the upper lateral N doped region. The connecting N doped region, the upper later N doped region, and the upper N doped waveguide region form a single contiguous N doped region. The OAR comprises the waveguide ridge and slab regions at either side of the waveguide so that the OAR has an inverted T-shape cross section (the cross section taken transverse to the longitudinal axis of the waveguide). The P+, N, and N+ doped regions are all located within the OAR material, whilst the N region extends along the top and the side of the waveguide ridge as well as the slab, the N+ and P+ regions are only found within the slab sections of the OAR, either side of the waveguide ridge.

(42) In other embodiments (not shown) the P and N doped regions are reversed so that the lower doped region contains an N doped zone and an N+ doped zone so that the upper doped region is P doped and P+ doped.

(43) In this embodiment, an extra step of etching a region of the OAR (e.g. SiGe) has occurred before that region is implanted to form a P+ doped region. This etching process creates a P+ region of the OAR which has a reduced height as compared to the slab within which it is located.

(44) By etching the slab region of the OAR before the P+ doping takes place, it is easier to ensure that the P and P+ doped regions are connected; that is to say that the P+ dopant region (the second zone of the multilayer lower doped portion) reaches through the thickness of the slab from the contact surface at the top surface to the P doped region at the bottom surface. The thickness of the second zone of the multilayer lower doped portion is 0 μm-0.2 μm. Where the thickness has a value of 0 μm, this should be understood to mean that the P+ dopant region is completely inside of the P region.

(45) In FIG. 9, an annealing process has taken place which causes the P doped region to diffuse into the central portion of the waveguide from below by a distance of between 50-200 nm. This “migrated” area caused by the dopant diffusion can be beneficial in reducing the series resistance and increasing the bandwidth of the optical modulator. The diffusion can occur in any of the embodiments described herein.

(46) FIG. 10, in contrast to FIG. 9, shows a device where the P doped region has diffused from a side of the central portion of the waveguide horizontally into the central portion of the OAR waveguide. Furthermore, this device is based on an 800 nm cross-section, whereas some previous examples are based on a 3 μm cross-section. Both forms of devices are suitable for implementing the PAM-N modulation scheme as discussed.

(47) In some embodiments, one or more of the modulators of the optoelectronic device are fabricated according to a method disclosed in U.S. patent application Ser. No. 15/700,053 and U.S. patent application Ser. No. 15/700,055, both of which are incorporated herein by reference. The optically active region of such a modulator may be part of a waveguide on a crystalline cladding layer (instead of, e.g., the BOX (buried oxide) cladding layer shown in FIGS. 8-10), and the optically active region above the crystalline cladding layer may be the modulating component of the modulator.

(48) FIG. 11 shows a variant optoelectronic device comprising three optical modulators, a first 701, second 702, and third 703. Each of the optical modulators has a different length as can be seen. In this example L.sub.1>L.sub.2>L.sub.3 (where corresponds to the n.sup.th optical modulator). As will be understood, the length of an optical modulator correlates with the magnitude of the modulation it will impart on electromagnetic waves being guided through it. Therefore, in this example, the first optical modulator will impart a greater magnitude of modulation to the electromagnetic waves than both the second or third.

(49) Thus, in some embodiments, pairs of optical modulators can be operated substantially simultaneously in order to provide further tuning between the modulation outputs of the device, to give an additional degree of freedom for the device driver to drive PAM-4 modulation. In other embodiments, the three levels can be controlled independently in such a manner as to avoid relying upon the product of two optical modulator extinction settings (e.g., to produce PAM-4, three optical modulators may be used, and in each of the four transmittance states corresponding to the four levels of PAM-4, at least two of the three optical modulators may have 0 V applied to them as the control voltage). The following table, Table 4, gives an example of where the three levels can be controlled independently:

(50) TABLE-US-00004 TABLE 4 PAM-4 1.sup.st Optical 2.sup.nd Optical 3.sup.rd Optical Level Modulator Modulator Modulator 3 0 0 0 2 1 0 0 1 0 1 0 0 0 0 1

(51) The following table, Table 5, gives an example where three optical modulators can be operated to produce PAM-4 modulated outputs (where 1 indicates that the optical modulator is being used to attenuate the electromagnetic waves, and 0 indicates that it is not). In one embodiment, each value of 0 listed below an optical modulator in Table 4 corresponds to applying a respective control voltage equal to a first voltage (e.g., 0 V) to the optical modulator, and each value of 1 corresponds to applying a respective control voltage equal to a second voltage (e.g., 2 V) to the optical modulator. A cascade of three modulators configured to operate in this manner may be driven by a particularly simple drive circuit having three outputs and including, for example, (i) a first set of three transistors, each of which, when turned on, connects a respective output, of the three outputs, to ground (i.e., 0 V), and (ii) a second set of three transistors, each of which, when turned on, connects a respective output, of the three outputs, to a power supply voltage (e.g., 2 V).

(52) TABLE-US-00005 TABLE 5 PAM-4 1.sup.st Optical 2.sup.nd Optical 3.sup.rd Optical Level Modulator Modulator Modulator 3 0 0 0 2 1 0 0 1 1 1 0 0 1 1 1

(53) A further example is shown in Table 6, where the third optical modulator is operated simultaneously with the second optical modulator (in one embodiment, for example, V.sub.1=0 V, V.sub.2=1.4 V, and V.sub.3=2):

(54) TABLE-US-00006 TABLE 6 PAM-4 1.sup.st Optical 2.sup.nd Optical 3.sup.rd Optical Level Modulator Modulator Modulator 3 V.sub.1 V.sub.1 V.sub.1 2 V.sub.2 V.sub.1 V.sub.1 1 V.sub.1 V.sub.3 V.sub.3 0 V.sub.3 V.sub.3 V.sub.3

(55) Alternatively, the third optical modulator could be operated independently of the first and second optical modulator i.e. it could be used to generate a PAM level of modulation without necessarily being used in conjunction with another optical modulator. Therefore, a PAM-8 level modulation scheme would be possible. In some embodiments another modulation scheme, referred to herein as “N-level modulation”, may be implemented using Log.sub.2 (N) modulators. Unlike PAM-N modulation (in which the N levels are equally spaced), N-level modulation may have optical power levels that are not equally spaced (they may be logarithmically spaced, for example). 8-level modulation, for example, may be generated using three cascaded modulators, each operated in one of two respective transmittance states, as shown in Table 7:

(56) TABLE-US-00007 TABLE 7 PAM-8 1.sup.st Optical 2.sup.nd Optical 3.sup.rd Optical Level Modulator Modulator Modulator 7 0 0 0 6 1 0 0 5 0 1 0 4 0 0 1 3 1 1 0 2 1 0 1 1 0 1 1 0 1 1 1

(57) As will be recognised, an N-level modulation scheme is possible by providing at least M optical modulators where M=Log.sub.2 (N). Said another way, by providing M optical modulators, an N-level modulation scheme is possible with 2.sup.M levels.

(58) In the example shown in FIG. 11, the three optical modulators 701, 702, and 703 are provided within a single SiGe ridge waveguide. This can reduce the number of interfaces where the electromagnetic waves must move from a material of one refractive index to another (this having associated reflection losses). However, this is not necessary. As shown in FIGS. 12A and 12B, it is possible to provide Si waveguides between each of the optical modulators. This may aid manufacturability and have improved performance since intermediate Si waveguides will have less loss.

(59) For example, as is shown in FIGS. 13A and 13B, the lengths and/or widths of the optical modulators within a device according to embodiments of the invention may differ markedly. In the example in FIG. 13A, the length L.sub.1 and width W.sub.1 of the first optical modulator is much reduced in comparison to the length L.sub.2 and width W.sub.2 of the second optical modulator. However, the length and width can be varied in order to tune the optical modulators in accordance with a particular device requirement. In the example in FIG. 13B, the length L.sub.1 of the first optical modulator is markedly greater than the length L.sub.2 of the second optical modulator, however the first optical modulator's width W.sub.1 is much reduced in comparison to the width W.sub.2 of the second optical modulator.

(60) The optical modulators may be arranged in order of active length, the longest being first—i.e. exposed first to the input light. This applies to all of the PAM-N modulators of embodiments of the invention.

(61) Another feature to note in FIGS. 13A and 13B is that the intermediate waveguides are (i) formed of bulk Si, and (ii) taper from a first width corresponding to the first optical modulator to an output width (the width of the input and output waveguides) and then from the output width to a second width corresponding to the second optical modulator. FIGS. 14A and 14B illustrate similar principles to FIGS. 14A and 14B, however the intermediate waveguides (between the first and second optical modulators) are formed of undoped SiGe and taper directly from the first width to the second width.

(62) The above figures illustrate that the particular parameters of any given optical modulator may vary (or indeed be tuned) whilst still allowing the optoelectronic device comprising said optical modulators to perform PAM-4 modulation.

(63) Another variant configuration is illustrated in FIG. 15. Previously hereto, each optical modulator has been separated by an electrical insulator for example undoped SiGe or bulk Si. However, it is possible to dope the regions of the ridge waveguide such that dopants of opposing polarity are immediately adjacent. For example, as shown in FIG. 15, a heavily doped region 205 of a first species of dopant is immediately adjacent to a heavily doped region 217 of a second species of dopant. The first and second species of dopant are of opposing polarities for example N++ and P++ or P++ and N++ respectively. This is also true for the lightly doped regions 204 and 216.

(64) By doing so, it is possible to reduce the overall length of the optoelectronic device (and so may also reduce the transmission losses associated with the device) whilst retaining the first and second optical modulators which facilitate PAM-4 modulation.

(65) FIG. 16 illustrates a variant optoelectronic device 100 which further includes a heater 900 disposed near to the first and second optical modulators. The heater can be used to tune the optical modulators (and therefore the optoelectronic device) with respect to wavelength. The heater can be formed form either doped SiGe, Si, or a metal. The heater is connected to electrodes which allow it to be controlled by the device.

(66) FIGS. 17A and 17B show, respectively, plots of extinction ratio (measured in dB) and Δα/α against wavelength for a simulated optoelectronic device. α is the absorption coefficient of the absorbing material (e.g. Ge, SiGe, SiGeSn, or GeSn), Δα represents the change in this absorption coefficient as a function of wavelength from when non-zero voltage is applied to the optical modulator to when zero voltage is applied, and Δα/α is the normalised change in absorption coefficient as a function of wavelength, and the optoelectronic device has the following parameters: L.sub.1=15 μm, L.sub.2=40 μm, W.sub.1=800 nm, and W.sub.2=700 nm. In FIG. 17A, the dotted lines 402, 404, and 406 correspond to minimum target extinction ratios for the PAM-4 levels 1, 2, and 3 respectively. The solid lines 401, 403, and 405 correspond with the extinction ratio of the optoelectronic device when operated in the respective PAM-4 levels: 1, 2, and 3. The solid line 407 indicates the insertion loss in dB associated with the device.

(67) In FIG. 17B, the solid lines 401, 403 and 405 indicate the normalised change in α.

(68) FIGS. 18A and 18B show plots similar to FIGS. 17A and 17B but for a variant optoelectronic device. Here the device had the following parameters: L.sub.1=15 μm, L.sub.2=45 μm, and W.sub.1=W.sub.2=700 nm. Also, indicated in FIG. 18A is the optimum working wavelength bandwidth 408 which is between 1526 nm and 1559 nm.

(69) These plots show that an optoelectronic device formed in accordance with embodiments of the present invention is operable to provide PAM-4 modulated outputs.

(70) FIG. 20 shows a variant device which differs from previous devices by having angled interfaces. The devices share a number of features with previous devices, and so like features are indicated by like reference numerals. The input waveguide 104 can be described as guiding light along a central axis 1601. The axis is at an angle ϕ.sub.1 from a line 1602 which is parallel with a longitudinal direction of the device. Therefore, the input waveguide includes an angled interface 2101. The interface 2101 itself may be at an angle α1 relative to the line 1602 which is not equal to 90°. Similarly, an interface 2104 between the second optical modulator 213 and the output waveguide 222 may be at an angle β2 which is not equal to 90°. As with the input waveguide 104, the output waveguide 222 may also guide light along an axis 1603, and this axis 1603 may be at an angle β2 to the line 1602.

(71) FIGS. 21 and 22 show variant devices which, like FIGS. 19 and 20, differs from the previous devices by having angled interfaces. The devices share a number of features with the previous devices, and so like features are indicated by like reference numerals. Taking the device shown in FIG. 21 first, the input waveguide 201 can be described as guiding light along a central axis 1601. The axis is at an angle ϕ.sub.1 from a line 1602 which is parallel with a guiding direction of the first and second optical modulators. Therefore, the input waveguide includes an angled interface 2101.

(72) In the example shown in FIG. 21, there is an intermediate waveguide 211 which is formed of a different material to the first 203 and second 213 optical modulator. Therefore, there is a change in refractive index. In order to reduce back reflections caused by this change in refractive index: the intermediate waveguide 211 may extend at an angle γ.sub.1 relative to the longitudinal direction 1602. Finally, the output waveguide 222 extends at an angle γ.sub.2 relative to the longitudinal direction 1602 (which is parallel to the guiding directions in the first and second optical modulators). The angle γ.sub.2 can be chosen to help mitigate back reflections which may occur as light moves from the second optical modulator 213 into the output waveguide 222.

(73) Generally, the angles obey the following: −14°<ϕ.sub.1, ϕ.sub.2<14° −14°<γ.sub.1, γ.sub.2<14°

(74) In some cases, α.sub.1=β.sub.1, α.sub.2=β.sub.2, |γ.sub.1|=|γ.sub.2| and |ϕ.sub.1|=|ϕ.sub.2|.

(75) The interface 2101 between the input waveguide 201 and the first optical modulator 203 may be at an angle not equal to 90° relative to the guiding direction of the first optical modulator. As shown in FIG. 21, the interface 2101 may be at an angle α.sub.1 which may be less than 90° relative to the guiding direction of the first optical modulator. Similarly, the interface 2102 between the first optical modulator 203 and the intermediate waveguide 211 may be at an angle not equal to 90° relative to the guiding direction of the first optical modulator. As shown in FIG. 21, the interface 2102 may be at an angle β.sub.1 which may be less than 90° relative to the guiding direction of the first optical modulator.

(76) The intermediate waveguide 211 may project from the first optical modulator 203 at an angle γ.sub.1 relative to the guiding direction of the first optical modulator. This means the intermediate waveguide 211 may guide light at an angle γ.sub.1 relative to the guiding direction of the first optical modulator. The guiding direction of the intermediate waveguide 211 may also lie at an angle ϕ.sub.2 to the guiding direction of the second optical modulator 213.

(77) The interface 2103 between the intermediate waveguide 211 and the second optical modulator 213 may be at an angle which is not equal to 90° relative to the guiding direction of the first optical modulator. As shown in FIG. 21, the interface 2103 may be at an angle α.sub.2 relative to the guiding direction of the first optical modulator which may be greater than 90°. Similarly, the interface 2104 between the second optical modulator and the output waveguide 222 may be at an angle not equal to 90° relative to the guiding direction of the first optical modulator. As shown in FIG. 21, the interface 2104 may be at an angle β.sub.2 relative to the guiding direction of the first optical modulator which may be greater than 90°.

(78) The output waveguide 222 may project from the second optical modulator 213 at an angle γ.sub.2 relative to the guiding direction of the first optical modulator.

(79) In a typical example of the device shown in FIG. 21: α.sub.1=80°; β.sub.1=80°; ϕ.sub.1=3°; γ.sub.1=3°; α.sub.2=100°=90+(90−α.sub.1); β.sub.2=100°=α.sub.2; ϕ.sub.2=−3°=−ϕ.sub.1; and γ.sub.2=−3°=−γ.sub.1.

(80) In general however:

(81) TABLE-US-00008 α.sub.1 = 89° to 50° α.sub.2 = 91° to 130° β.sub.1 = 89° to 50° β.sub.2 = 91° to 130° ϕ.sub.1 = 0.3° to 14° ϕ.sub.2 = −0.3° to −14° γ.sub.1 = 0.3° to 14° γ.sub.2 = −0.3° to −14°

(82) and, in some embodiments:

(83) TABLE-US-00009 α.sub.1 = 89° to 80° α.sub.2 = 91° to 100° β.sub.1 = 89° to 80° β.sub.2 = 91° to 100° ϕ.sub.1 = 0.3° to 3° ϕ.sub.2 = −0.3° to −3° γ.sub.1 = 0.3° to 3° γ.sub.2 = −0.3° to −3°

(84) The configuration of the device shown in FIG. 22 is the same as that shown in FIG. 21 insofar as the input waveguide and first optical modulator are concerned. However, the intermediate waveguide 2201 of the device shown in FIG. 22 is not linear as shown previously. Instead, the intermediate waveguide bends as the second optical modulator 213 has been rotated by 180° relative to the second optical modulator shown in FIG. 21. Therefore light guided by the intermediate waveguide 2201 of FIG. 22 follows substantially curved path relative to the intermediate waveguide 211 of FIG. 21.

(85) In a typical example of the device shown in FIG. 22: α.sub.1=80°; β.sub.1=80°; ϕ.sub.1=3°; γ.sub.1=3°; α.sub.2=80°=α.sub.2; β.sub.2=80°=α.sub.2; ϕ.sub.2=3°=ϕ.sub.1; and γ.sub.2=3°=γ.sub.1.

(86) In general however:

(87) TABLE-US-00010 α.sub.1 = 89° to 50° α.sub.2 = 89° to 50° β.sub.1 = 89° to 50° β.sub.2 = 89° to 50° ϕ.sub.1 = 0.3° to 14° ϕ.sub.2 = 0.3° to 14° γ.sub.1 = 0.3° to 14° γ.sub.2 = 0.3° to 14°
and, in some embodiments:

(88) TABLE-US-00011 α.sub.1 = 89° to 80° α.sub.2 = 89° to 80° β.sub.1 = 89° to 80° β.sub.2 = 89° to 80° ϕ.sub.1 = 0.3° to 3° ϕ.sub.2 = 0.3° to 3° γ.sub.1 = 0.3° to 3° γ.sub.2 = 0.3° to 3°

(89) The angled interfaces in the example shown in FIG. 22 may be used instead of those shown in FIG. 21, as the angled interfaces 2102 and 2103 are non-parallel to each other and so may suppress resonator or Fabry-Perot cavity effects within the intermediate waveguide.

(90) While the invention has been described in conjunction with the exemplary embodiments described above, many equivalent modifications and variations will be apparent to those skilled in the art when given this disclosure. Accordingly, the exemplary embodiments of the invention set forth above are considered to be illustrative and not limiting. Various changes to the described embodiments may be made without departing from the spirit and scope of the invention.

(91) All references referred to above are hereby incorporated by reference.

LIST OF FEATURES

(92) 100 Optoelectronic device 101 Input light 102, 201 Input waveguide 103, 202 Tapered region of input waveguide 104, 203 First optical modulator 105, 208 Length—L.sub.1—of first optical modulator 106, 209 Waveguide width—W.sub.1—of first optical modulator 107, 112 Electrodes 210a, 210b Electrodes 220a, 220b Electrodes 108, 211 Intermediate waveguide 109, 213 Second optical modulator 110, 218 Length—L.sub.2—of the second optical modulator 111, 219 Waveguide width—W.sub.2—of the second optical modulator 113, 222 Output waveguide 114 Modulated light 115 Example levels of optical modulator 221 Tapered region of output waveguide 204, 214 Lightly doped region (first species) 205, 215 Heavily doped region (first species) 206, 216 Lightly doped region (second species) 207, 217 Heavily doped region (second species) 301 Si slab 302 Heavily doped region (first species) 303 Heavily doped region (second species) 304 SiGe ridge waveguide 305 Lightly doped region (first species) 306 Lightly doped region (second species 307 Heavily doped region in ridge (first species) 308 Heavily doped region in ridge (second species) 401, 403, 405 First, second, and third optical modulator extinction ratios 402, 404, 406 Minimum target optical modulator extinction ratio 407 Insertion loss 701 First optical modulator 702 Second optical modulator 703 Third optical modulator 900 Heater 1601 Guiding direction of input waveguide 1602 Longitudinal axis of the device/guiding direction of the electro-absorption modulator 2101 Input waveguide-optical modulator interface 2102 Optical modulator-intermediate waveguide interface 2103 Intermediate waveguide-optical modulator interface 2104 Optical modulator-output waveguide interface 2201 Curved intermediate waveguide