Thin layer photonic integrated circuit based optical signal manipulators

Abstract

Integrated optical intensity or phase modulators capable of very low modulation voltage, broad modulation bandwidth, low optical power loss for device insertion, and very small device size are of interest. Such modulators can be of electro-optic or electro-absorption type made of an appropriate electro-optic or electro-absorption material in particular or referred to as an active material in general. An efficient optical waveguide structure for achieving high overlapping between the optical beam mode and the active electro-active region leads to reduced modulation voltage. In an embodiment, ultra-low modulation voltage, high-frequency response, and very compact device size are enabled by a semiconductor modulator device structure, together with an active semiconductor material that is an electro-optic or electro-absorption material, that are appropriately doped with carriers to substantially lower the modulator voltage and still maintain the high frequency response. In another embodiment, an efficient optical coupling structure further enables low optical loss. Various embodiments combined enable the modulator to reach lower voltage, higher frequency, low optical loss, and more compact size than previously possible in the prior arts.

Claims

1. A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate, comprising: an input connecting waveguide core deposed on the substrate connecting an energy of an optical beam to and from an electro-active layer, the optical beam having one or more optical wavelengths around an operating optical wavelength .sub.op; the input connecting waveguide core becomes an input tapering waveguide core and enters and extends below an electro-active layer, wherein the optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer, and the optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core extends below the electro-active layer; and a refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of a material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer, wherein the electro-active layer is either part of or in spatial proximity to an electro-active waveguide core.

2. The device as claimed in claim 1, wherein the electro-active waveguide core and an electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that a refractive index contrast of the waveguide core layer with both a top and a bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region.

3. The device as claimed in claim 1, wherein the electro-active waveguide core and electro-active waveguide cladding structure is in an very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.5, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region.

4. The device as claimed in claim 1, wherein the electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co).

5. The device as claimed in claim 1, wherein the electro-active waveguide core thickness d.sub.CORE is in the ultra-thin region or very-thin region such that d.sub.CORE<(.sub.op/n.sub.Co).

6. The device as claimed in claim 1, wherein the electro-active layer has a low-refractive-index Ohmic transparent conductor (LRI-OTC) layer electrically connected from the top to the electro-active layer, wherein the LRI-OTC forms part of the top electro-active waveguide cladding.

7. The device as claimed in claim 1, further comprising A structure electrically connected to the electro-active layer comprises at least a first PN junction in which a first P-layer with P-dopant is vertically connected (vertical means in a direction perpendicular to the substrate plane; horizontal means in a direction parallel to the substrate plane) to a first N-layer with N-dopant, or a PqN junction in which a first P-layer with P-dopant is connected to a first q-layer with either N or P dopant or that is undoped (i.e. being an Intrinsic semiconductor material) and the q-layer is further connected to a first N-layer with N-dopant.

8. The device as claimed in claim 7, wherein a voltage is applied across the first P-layer of this first PN junction, and the first N-layer of the first PN junction to result in an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer.

9. The device as claimed in claim 7, wherein a voltage is applied across the second N-layer of the second PN junction, and the first N-layer of the first PN junction to result in an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer.

10. The device as claimed in claim 1, further comprising a structure electrically connected to the electro-active layer comprises at least a first NqN junction in which a first N-layer with N-dopant is vertically connected (vertical means in a direction perpendicular to the substrate plane; horizontal means in a direction parallel to the substrate plane) to a first q-layer with either N or P dopant or that is undoped (i.e. being an Intrinsic semiconductor material) and the first q-layer is further connected to a second N-layer with N-dopant.

11. The device as claimed in claim 10, wherein a voltage is applied across the first N-layer and the second N-layer of this first NqN junction to result in an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer.

12. A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate, comprising: an input connecting waveguide core deposed on the substrate connecting an energy of an optical beam to and from an electro-active layer, the optical beam having one or more optical wavelengths around an operating optical wavelength .sub.op; the input connecting waveguide core becomes an input tapering waveguide core and enters and extends below an electro-active layer, wherein the optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer, and the optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core extends below the electro-active layer and the refractive index of the tapering waveguide core is given by n.sub.ITWCo-z1; a refractive index n.sub.EC or the optical gain/absorption coefficient n.sub.EC of at least part of a material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer, wherein the electro-active layer is either part of or in spatial proximity to an electro-active waveguide core; and the width w.sub.ITWCo-z1 of the input tapering waveguide core after penetrating below the electro-active layer is reduced from a value approximately equal to or larger than half the wavelength in the material .sub.op/(2n.sub.ITWCo-z1) to a value smaller than half the wavelength in the material .sub.op/(2n.sub.ITWCo-z1), so that w.sub.ITWCo-z1<.sub.op/(2n.sub.ITWCo-z1) at some point under the electro-active layer, wherein the electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.5, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region.

13. The device as claimed in claim 12, wherein the electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co).

14. The device as claimed in claim 12, further comprising a structure electrically connected to the electro-active layer comprises at least a first PN junction in which a first P-layer with P-dopant is vertically connected (vertical means in a direction perpendicular to the substrate plane; horizontal means in a direction parallel to the substrate plane) to a first N-layer with N-dopant, or a PqN junction in which a first P-layer with P-dopant is connected to a first q-layer with either N or P dopant or that is undoped (i.e. being an Intrinsic semiconductor material) and the q-layer is further connected to a first N-layer with N-dopant.

15. The device as claimed in claim 14, further comprising a second P-layer with P-dopant of a second PN junction, wherein second P-layer is electrically connected to a second N-layer with N-dopant of this second PN junction, wherein a voltage is applied across the second N-layer of the second PN junction, and the first N-layer of the first PN junction to result in an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer.

16. The device as claimed in claim 14, wherein at least one of the first P-layer, first N-layer, or the first q-layer contains at least one quantum well, wherein the doping density at the quantum well is in the highly-doped, medium-highly-doped, very-highly-doped, or ultra-highly-doped regime with a dopant density higher than about 210.sup.17/cm.sup.3 with either N doping or P doping.

17. The device as claimed in claim 14, wherein at least one of the first P-layer, first N-layer, or the first q-layer contains at least one quantum well, wherein the doping density at the quantum well is in the medium-highly-doped, very-highly-doped, or ultra-highly-doped regime with a dopant density higher than about 510.sup.17/cm.sup.3 with either N doping or P doping.

18. The device as claimed in claim 14, wherein at least one of the first P-layer, first N-layer, or the first q-layer contains at least one quantum well, wherein the doping density at the quantum well is in the very-highly-doped, or ultra-highly-doped regime with a dopant density higher than about 1.510.sup.18/cm.sup.3 with either N doping or P doping.

19. A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate, comprising: an input connecting waveguide core deposed on the substrate connecting an energy of an optical beam to and from an electro-active layer, the optical beam having one or more optical wavelengths around an operating optical wavelength .sub.op; the input connecting waveguide core becomes an input tapering waveguide core and enters and extends below an electro-active layer, wherein the optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer, and the optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core extends below the electro-active layer; and a refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of a material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer, wherein the electro-active layer is either part of or in spatial proximity to an electro-active waveguide core, wherein an electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both a top and a bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.5, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding, and n.sub.Co is the averaged material refractive index of the electro-active waveguide core.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The preferred embodiments of the invention will hereinafter be described in conjunction with the appended drawings provided to illustrate and not to limit the invention, wherein like designations denote like elements, and in which:

(2) FIG. 1a is a diagram illustrating the case for which the electron carriers fill up the conduction band, leading to a shift in the absorption energy from close to the bandgap energy Eg to larger than the bandgap energy Eg+E (or in wavelength );

(3) FIG. 1b is a diagram showing a change in the absorption energy edge from absorption curve .sub.Eg() to .sub.Eg+E() leads to a change in the refractive index of the material n().

(4) FIG. 2 is a diagram showing an electro-optic (EO) modulator of prior art. In particular a modulator made of III-V compound semiconductor material.

(5) FIG. 3 is a diagram illustrating the cross-section of the PIN modulator structure.

(6) FIG. 4a is a diagram showing the location of the modulator input beam coupler structure (IBCS), the Active Layer structure (ALS), and the modulator output beam coupler structure (OBCS)

(7) FIG. 4b is an exemplary schematic for a cross-section of the input waveguide region. The region occupied by the optical beam (OB) is shown as a shaded region.

(8) FIG. 4c is an exemplary schematic for a cross-section of IBCS region or OBCS region. The region occupied by the optical beam (OB) is shown as a shaded region.

(9) FIG. 4d is an exemplary schematic for a cross-section of ALS region showing the active material (ACM) layer, the electro-active layer (ECL), and the waveguiding layers. The region occupied by the optical beam (OB) is shown as a shaded region.

(10) FIG. 5a is a diagram showing a pair of capacitively-loaded traveling wave electrodes.

(11) FIG. 5b is a diagram showing the cross-section d-d in FIG. 5a, illustrating that the bottom parts of the two phase modulators are connected.

(12) FIG. 6a is a diagram showing equivalent lumped-element circuit of a pair of capacitively-loaded traveling wave electrodes (CL-TWE) powering the two phase modulators along the two arm of the Mach Zehnder Interferometer.

(13) FIG. 6b is a diagram showing the top view of the CL-TWE lines for the traveling wave electrodes.

(14) FIG. 7a/b is a diagram illustrating how the beam from the SOI waveguide is coupled into the thin-film modulator structure using two tapers

(15) FIG. 8a is a diagram showing the cross-section for the side-conduction-layer (SCL) case.

(16) FIG. 8b is a diagram showing the cross-section for the Ohmic Transparent Conductor (OmTC) case.

(17) FIG. 9 is a diagram showing the details structures for an exemplary NIN modulator.

(18) FIG. 10 is a diagram showing the details structures for an exemplary NPNN modulator.

(19) FIG. 11 is a diagram showing the modulation BW for the NIN case with 2-m-wide active region so w.sub.CAP=2 m.

(20) FIG. 12 shows the modulation bandwidth for the NPNN case with 2-m-wide active region so w.sub.CAP=2 m.

(21) FIG. 13 is a diagram illustrating a general geometry of the EO Modulator of the present invention. FIG. 13b is a semi-transparent illustration of FIG. 13a.

(22) FIG. 14 shows a diagram illustrating that material surrounding the input connecting waveguide core.

(23) FIG. 15 shows a diagram illustrating the input beam coupler structure (IBCS). FIG. 15a show the input beam coupler structure (IBCS) comprises at least a tapering waveguide section generally tapering from wide to narrow. Optionally, the active layer structure ALS on top of the input tapering waveguide section can also be tapering in the form of an up taper (tapering from narrow to wide in the direction toward the active layer structure ALS). See for example FIG. 15b.

(24) FIG. 16 shows a diagram illustrating a few exemplary cases for the top metal contact pad placements.

(25) FIG. 17 shows a diagram illustrating a particular exemplary embodiment of a electro-active waveguiding core structure EWCoS 22600.

(26) FIG. 18 shows a diagram illustrating the output connecting waveguide structure.

(27) FIG. 19 shows a diagram illustrating the division of the output connecting waveguide cladding into four different regions.

(28) FIG. 20 shows a diagram illustrating the output beam coupler structure (OBCS)

(29) FIG. 21 shows a diagram illustrating the electro-active layer (ECL).

(30) FIG. 22a shows a diagram illustrating the structure around the PqN. NqN, or PN junction.

(31) FIG. 22b shows a diagram illustrating the structure around the PqN, NqN, or PN junction with PN changing PN junction.

(32) FIG. 23 shows a diagram illustrating an alternative structure involved having the metal Ohmic contact on the side away from the center region of layer TVSCOC 21700.

(33) FIG. 24 shows a diagram illustrating an alternative structure involved having the top lateral conduction geometry with metal contact but also a lowloss dielectric material as layer 21810.

(34) FIG. 25 shows a diagram illustrating an alternative structure for which the metal can even go on top of the top lowloss dielectric material region TDMR 21810 to make this top lateral conduction structure mechanically robust.

(35) FIG. 26 shows a diagram illustrating an alternative structure with low capacitance but also low optical loss for example by using the top vertical/side conduction and Ohmic contact layer TVSCOC 21700 to confine the mode laterally.

(36) Skilled artisans will appreciate that the elements in the figures are illustrated for simplicity and clarity, and have not necessarily been drawn to scale. For example, the dimensions of some of the elements in the figures may be exaggerated, relative to the other elements, to help in improving understanding of the embodiments of the present invention.

DETAILED DESCRIPTION

(37) Motivations of the Present Invention

(38) There are various needs for ultra-low-RF-power ultra-wide-RF-bandwidth low-optical-loss high-optical-power modulators for various applications. Certain exemplary modulators employing exemplary embodiments of the present invention are capable of either Ultra Low Voltage, Ultra-High Modulation Bandwidth, Low Optical Loss, or High Optical Power, or a plurality of the above. In addition, they are generally ultra-compact, can be integrated with semiconductor laser, and can be made based on mass-producible silicon-photonic platform with EPIC (electronic-photonic integrated circuit) capability enabling future expansions to integrate with RF circuits or other photonic devices on chip.

(39) Needs for Compact Wide-Bandwidth Low-Power Low-Loss Modulators

(40) New applications in communications and sensing require transmission of high-frequency electronic signals. Transmission of ultra-fast digital data over optical fiber system is also important for next generation data centers. In order to address such needs, optical phase or intensity modulators that are capable of low switching voltage (lower than 0.5 to 1V), broad RF bandwidth (BW) (higher than 20 GHz; preferably over 100 GHz), low device optical power loss (preferably <6 dB), and capable of withstanding optical powers of a few hundred milliWatts would be desirable. Prior arts in EO modulators are not able to realize such modulators. For example, the commonly available Lithium Niobate modulators in the market can go up to 40-100 GHz but cannot reach low enough modulation voltage of below 1V. Polymer modulators could give broader bandwidth of over 100 GHz but still have high switching voltages. Also they cannot withstand high optical powers. Semiconductor based modulators have various advantages in terms of their smaller size and shorter physical length but they are not yet able to give modulation voltage lower than 1V, broad modulation bandwidth higher than 20 GHz, and low device optical throughput loss smaller than 6 dB concurrently.

(41) Exemplary embodiments of the present invention described below will utilize a few key factors combined to fully address the abovementioned problems resulting in low modulation voltage and wide modulation bandwidth concurrently, In some preferred embodiments, the resulting modulator also has short physical device length and low optical loss.

(42) Broad Overview of the Present Invention

(43) An optical modulator device can be divided into a few key components composing the modulator. An input optical beam must be channeled to the modulator's active material medium (ACM) efficiently without too much loss of the beam's optical power. In conventional EO modulator, this is just done by joining input waveguide to the modulator waveguide. In the EO modulator of the present invention, due to the small optical mode in the modulator, in one exemplary embodiment, this input structure is an integral part of the modulator of the present invention. We call this the modulator input beam coupler structure (IBCS). A diagram illustrating such IBCS is shown in FIG. 4. An exemplary schematic for a cross-section of the input waveguide region is shown by FIG. 4b. An exemplary schematic for a cross-section of such IBCS region is shown by FIG. 4c. The region occupied by the optical beam (OB) in FIG. 4b and FIG. 4c is shown as a shaded region.

(44) The modulator input beam coupler structure brings the optical beam from a waveguide into the modulator's main waveguide that contains the active electro-optic (EO) material or electro-absorption (EA) material. An active EO (or EA) material is a material whose refractive index (or absorption or gain) can be altered by an applied electric field or an electrical current. Such EO or EA material will be called collectively as active material or medium (ACM). The active medium is typically embedded as part of the active layer stricture (ALS) of the modulator. An exemplary schematic for such ALS is shown by FIG. 4d. The active medium (ACM) plus the immediately connected structure next to it for applying the required electric field or introducing the required current is called the electro-active layer (ECL). An exemplary schematic for such ECL is shown by FIG. 4d.

(45) In addition, a more extensive electrical conduction structure is integrated with the electro-active layer and a waveguiding structure so that an electric field or electrical current can be brought into the electro-active layer encompassing the active material, and at the same time the waveguiding structure will guide an optical beam so that part of its beam power overlaps with the EO/EA material. This enables the optical beam to experience the change in phase shift induced by a change in the refractive index in the EO material layer or a change in the optical absorption (or in some cases optical gain) in the EA material layer.

(46) A most commonly used structure immediately connected to and next to the active material for applying the required electric field or introducing the required current is a PIN structure, meaning that the electrical conduction to the active material is with a P-doped semiconductor followed by an intrinsic (I) semiconductor, and then followed by another N-doped semiconductor. An exemplary schematic for such PIN structure is shown by FIG. 4d. The active medium is typically in the I layer but can also be in any of the P or N layer, including the transition region between the I and any of the P and N layer, or in plurality of these layers. The active medium can be the layer itself, or a quantum-well structure, or other active-medium structure embedded in the layer. Most commonly, the active material is quantum wells in the intrinsic layer. Most EO and EA modulators with such PIN structure and quantum wells in the intrinsic layer as the electro-active layer operates under a reverse bias to such a PIN structure (with negative voltage applied to P side and positive voltage applied to the N side). Reverse bias will not bring in much current to the active material but will bring in a strong electric field to the active material. The electric field acting on the quantum well then changes its refractive index or absorption at the beam's optical wavelength. In some situation, a modulator can operate with gain and a forward bias can be applied to such PIN structure (with positive voltage applied to P side and negative voltage applied to the N side). Forward bias brings a strong injection current to the active material. The electrical current exciting the carriers in the quantum well then changes its refractive index or gain (or absorption) at the beam's optical wavelength. Typically, modulator based on forward bias has a slow turning off characteristics due to the slow nanosecond decay of excited carriers in the active medium after the voltage is turned off. Hence, typically the reverse bias case will give rise to much faster modulation speed than the forward bias case.

(47) For modulators with, for example PIN junction or the like, the Electro-Active layer (ECL) will be the EO/EA material and the immediate P and N doped regions or the like. An exemplary schematic for such ECL with a PIN structure and the EO/EA active material (ACM) layer is shown by FIG. 4d.

(48) The entire larger structure is called the active layer structure (or ALS) below. In short, the entire structure comprising: (1) the waveguiding layers: (2) the electro-Active layers; and (3) the other electrical conduction layers, is called the active layer structure (ALS) of the modulator in the present invention. An exemplary schematic for such ALS is shown by FIG. 4d. In many situations, part of the Electro-Active layer or part of the other electrical conduction layers also serve the double function as part of the waveguiding layers. Thus, each these layers often by itself, serves as multiple-function layer.

(49) The optical waveguide in the ALS is called active waveguide, so as to distinguish it from the input and output waveguides that have no active EO/EA material. In another exemplary embodiment, the present invention is concerning on the specific structure of the active layer structure independent of the input and output mode coupling structures. The location of such ALS layer is shown in FIG. 4.

(50) At the output, we have a modulator output beam coupler structure (OBCS) that couples the optical beam efficiently from the modulator active layer structure into a primarily passive output optical waveguide. Passive in this context means the waveguide acts primarily to transmit the optical beam energy. Primarily passive means it can also be active (e.g. with optical gain, absorption, or modulation) but for the purpose of this invention, the passive beam transmission function is the function utilized. In as yet another exemplary embodiment, this output structure is an integral part of the modulator of the present invention. The location of such OBCS is shown in FIG. 4. The cross section of OBCS is similar to that of IBCS, An exemplary schematic for a cross-section of such OBCS region is shown by FIG. 4c. The region occupied by the optical beam (OB) is shown as a shaded region.

(51) For the purpose of illustration but not limitation, it is useful to provide an overview of an exemplary modulator device employing the present invention. The Active-Layer Structure in an exemplary EO modulator of the present invention could make use of up to six main key elements, namely: (1) The use of an efficient coupling waveguide platform (EC-WG); (2) Low-Optical-Loss Ohmic Contact (LOL-OC), such as the use of transparent conductor and side conduction geometry, (3) Low-Optical-Loss and High Electrical Conductivity Waveguide Structure (LOL-HEC-WS) such as the use of PN-changing PN junction or PN tunnel junction to reduce the region with P-doping, (4) High-Response Active Material, such as material that has high EO or EA response under applied voltage, an electric current, or either injection or depletion of carriers (HR-AM); in an exemplary embodiment of the present invention, this is achieved with appropriately high carrier doping in quantum wells, and (5) Highly-Confined Thin-Film Electro-Active Waveguide (TF-ECW), so as to increase the amount of overlapping between the optical mode energy and the active material.

(52) To summarize these few advantages, the modulators of the present invention encompass one or more of the above five main key elements, including the advantages in Beam input/output Coupling Waveguide, Ohmic Contact, Waveguide Conductive Structure, Active Layer, and Strongly Confined Thin-Film Active Waveguide. For general references, optoelectronic or photonic device structures that take advantage of a few of the above five main factors will be generally referred to as WOCAT device structures. While the WOCAT structure here is applied to optical phase and intensity modulators, it has applications beyond optical modulators such as applications to optical amplifier, photodetector, laser, light-emitting device, optical switching and logic device, and optical signal processing device.

(53) As an exemplary embodiment, for the purpose of illustration and not limitation, the EO Modulators of the present invention are capable of achieving the significant advantages of ultra-low-voltage (<0.5V typical), broad RF bandwidth (20-100 GHz), low optical loss (<6 dB), short device length (<1 mm), or high optical power (>100 mW), or a plurality of the above. As an exemplary illustration but not limitation, such a modulator could make use of one or more of the following few key factors in its structure.

(54) Key Factor I: Low Voltage Via Strong Mode Confinement

(55) For the purpose of discussion and not limitation, for an operating optical wavelength of 1550 nm, an exemplary approach will be based on InP/InAlGaAs material system (called simply as InP/III-V). When used as 1550 nm EO modulation material, the InP/III-V material system will involve quantum wells (QWs) at 1300-1400 nm wavelength range that is close enough to the 1550 nm operation wavelength to give high EO phase shift, but is still far away from 1550 nm so that optical carrier excitation will be low enabling high optical power (>100 mW). This will result in a high maximum optical power or high MP value (used in the Modulator Figure of Merits: MFOM). When used as 1550 nm EA modulation material, the InP/III-V material system will involve quantum wells (QWs) at 1400-1550 nm wavelength range that is close enough to the 1550 nm operation wavelength to give high electro-absorption, and in some cases, optical gain (both absorption and gain can be used to modulate the intensity of an optical beam). However in that case, the MP value will be smaller due to the optical beam energy being absorbed and saturating the absorbing quantum well, thereby reducing the amount of optical modulation due to electro-absorption, when the optical beam power is high.

(56) In order to achieve very low modulation voltage, one way is to make the optical mode confinement a lot tighter so as to drastically reduce the effective distance between the voltage-applying electrodes. This will enable the same electric field strength to be achieved with proportionally lower applied voltage (assuming the mode-medium overlapping factor is near 100% and cannot be increased further). If the mode-medium overlapping factor is small, then a stronger mode confinement giving a small mode size will increase the overlapping, which will also increase the modulation phase shift or absorption even with the same electric field strength. In either case, the voltage is lower due either to higher electric field or higher mode-medium overlap.

(57) For a conventional semiconductor EO or EA modulator, the vertical mode size (FWHM) is about 0.5 m to 1 m. With high-refractive-index-contract material using semiconductor with high refractive index (n3 to 4) as the waveguide core surrounded by air, dielectric material, or polymer with low refractive index of n1 to 2 as waveguide cladding, it is possible to reduce the vertical mode size by about 2.5 times to 10 times to 0.1-0.2 m (at =1550 nm). For example, using the high refractive index of III-V semiconductor with n3.5 as the waveguide core and air as cladding will result in a single-mode strongly confined waveguide physical height of about 0.2 m, given by d.sub.SM/(2n)=1500 nm/(2*3.5)=0.214 m. This gives a vertical mode size of 0.1 m, which is a 5 to 10 times reduction in mode size or 5 to 10 times increase in the mode-medium overlapping factor. Suppose other factors remain the same, the modulation voltage is inversely proportional to the mode-medium overlapping factor. Thus, the 5 to 10 times higher mode-medium overlapping factor will reduce the modulation voltage by 5 to 10 times, and the 5V modulation voltage of a typical conventional modulator structure can be reduced to below 1 volt with use of the modulator structure of the present invention that has a strong mode confinement.

(58) For the purpose of illustrations and not limitations, unless otherwise stated, all the dimensional numbers such as mode size and structural sizes given in this invention assume that the operating optical wavelength is around the wavelength of 1550 nm. As is well known to those skilled in the arts, all these dimensions will scale proportionally when operating at other wavelength so that if the operating wavelength is at around 750 nm, all the physical dimensions will be about half of that given for the 1550 nm wavelength case. This invention is applicable to all other wavelengths and is not limited to the exemplary operating wavelength of 1550 nm.

(59) Key Factor II: Efficient Coupling into Strongly-Confined Waveguide

(60) In the present invention, the vertical mode confinement is reduced to <0.2-0.3 m. A challenge is how to achieve efficient optical beam coupling to the sub-micrometer waveguide. We will solve this problem using tapered waveguide that can be fabricated on a substrate and the ALS thin film on top of it can be attached via wafer bonding method or other methods after the waveguide is fabricated. Such tapered waveguide coupling structure can achieve over 90-95% optical power coupling efficiency.

(61) Key Factor I: High Modulation Bandwidth Via Low Ohmic Contact Resistance & Low-Optical-Loss High-Conductivity Waveguide Structure; and High Modulation Bandwidth Via Enhanced EO Response

(62) To enable electrical current injection and voltage conduction into the modulator device with strong optical mode confinement, as an exemplary embodiment in the present invention, a transparent conducting (TC) material that has low refractive index and yet can achieve excellent Ohmic contact with N-doped InP semiconductor material with very low contact resistance is used. We call these Ohmic Transparent Conducting (OmTC) materials. These TC materials are typically metal oxides (In.sub.2O.sub.3, ZnO, InSnO, CdO, ZnInSnO, InGaO, etc) or doped metal oxides (e.g. the above listed metal oxides doped with magnesium Mg or zinc Zn etc), the most familiar one is ITO (Indium Tin Oxide; InSnO) used widely in LCD display. They are called transparent conducting oxides (TCO). For example, with appropriate processes, it is possible to achieve good Ohmic contact between In.sub.2O.sub.3 or CdO and N-doped InP. We will call these Ohmic TCO (OmTCO). OmTCO will enable robust electrical structures to be realized that also has high mode confinement.

(63) Alternatively, we can use a side conduction layer (SCL) to bring the voltage into the top layer for the low voltage modulators. The side conduction geometry enables the waveguiding layer to be thin giving high mode-medium overlapping factor and yet maintaining low optical loss as the optical beam energy will not touch the optically lossy metal that is already moved to the side. Often the top cladding in such thin waveguide structure can be made to be either air or some low-refractive-index dielectric material.

(64) Both OmTCO or SCL can be used for the top contact. When the structure is thin, comparing to SCL the use of OmTCO for top contact has an advantage in terms of ease in fabrication and also potentially better device performances as the metal contact area can be larger.

(65) However, solving electrical conduction is only half of the matter. The modulation speed of a conventional semiconductor EO or EA modulator is about 20-40 GHz. It is desirable to reach 100 GHz with voltage <1V. How can one do so? It turns out that the high contact resistance for the p-doped material with metal, that is the high P-Ohmic contact resistance with metal, is a main problem that limits higher modulation frequency. P-Ohmic contact typically has 10 times higher resistance than N-Ohmic contact (with their respective appropriate Ohmic-contact metals that can give reasonably low contact resistance).

(66) We note that at the same dopant density, the P-doped cladding layer typically also has about 10 times higher resistance than if it is N-doped. While the cladding resistance is typically smaller than the P-Ohmic contact resistance especially since the active-layer structure is thin, the P-doped cladding can cause radio-frequency (RF) loss when a high-frequency modulating voltage pulse (or electrical signal) propagates along the modulator structure. In terms of free-carrier optical absorption, at the same dopant density, the P-doped cladding layer is also about 10 times higher than N-doped. If one reduces the P-doped cladding resistance by increasing the carrier doping density, one will also increase the optical loss, making it hard to achieve the long modulator length needed for low voltage, giving a trade-off between low voltage (need low doping density) and high frequency (need high doping density).

(67) To gain higher electrical modulating frequency response for the modulator, it is important to reduce the P-Ohmic contact and cladding resistance by changing them to N-contact and N-doped cladding instead. A few exemplary structures can do so. These structures are broadly classified as alternative Electro-Active Layer Structure A, B, C, and D discussed below.

(68) Alternative Electro-Active Layer Structure A: NIN Structure

(69) Besides the usual PIN structure noted above that can be used as the electro-active layer structure in the ALS, there are other alternative electro-active layer in the ALS structures that may have certain advantages. First is the use of an NIN electro-active layer in the ALS structure, meaning that the electrical conduction to the active medium is with an N-doped semiconductor followed by an intrinsic (I) semiconductor, and then followed by another N-doped semiconductor. The active medium is typically in the I layer but can also be in any of the two N layers, including the transition region between the I and any of the two N layers, or in plurality of these layers. The active medium can be the layer itself, or a quantum-well structure, or other active-medium structure embedded in the electro-active layer.

(70) In the situation whereby the modulator devices in the present invention requires largely only electric field to be applied to the active medium to affect the refractive index or optical absorption (or optical gain) of the active medium, it is appropriate to use such a NIN structure as the active layer.

(71) Comparing to the use of the conventional PIN structure, such NIN structure will reduce the Ohmic contact resistance by 10 times as both sides of the metal contacts will be contacting to N-doped layers only. Note that in NIN, sometime a thin P-doped layer is introduced so that it forms NPIN, where the P-layer helps to block the electric current, the NIN here broadly includes NPIN. For NPIN case, a positive voltage applied to the N layer of the NP side will cause reverse bias at the NP junction and hence cut off any current flow (PIN side becomes forward bias). A positive voltage applied to the N layer of the PIN side will cause reverse bias at the PIN junction and hence also cut off any current flow. This reverse bias to the PIN junction case is normally preferred as it will mimic the revered bias to the conventional PIN structure case more closely with voltage drop mainly across the PIN part of the structure (instead of the NP part of the structure).

(72) The NIN structure also reduces the RF-loss due to P-doped cladding by 5 to 10 times. This enables the maximum length of the modulator limited by RF-loss cutoff to reach a longer length of over 1 cm (for 40 GHz bandwidth). RF-loss is sensitive to the capacitance loading of the modulator. The capacitance is mainly determined by the lateral width of the junction capacitance region, called modulator-capacitor lateral width (labeled as w.sub.CAP or w.sub.EC) below. Lateral is in a direction perpendicular to the direction of optical beam propagation and parallel to the substrate surface.

(73) For a modulator with a typical 200 nm thick active region that defines the effective separation of the junction capacitor plates, the required modulator capacitor lateral width has to be w.sub.EC<2-m in order to reach 40 GHz BW for a 1 cm-long modulator. At 100 GHz, this RF-loss limited modulator length could be around 3-5 mm (for the same 2-m modulator capacitor width).

(74) The optical-absorption-loss limited length (at 2 dB attenuation so as to keep the total loss less than 6 dB) due to N-doped claddings in NIN can be made to reach longer than 0.5-1 cm so that the length will not be limited by optical loss (because it has lower optical loss than P-doped). If we put in a QW structure that increases the nonlinear response by about 3 times from just the use of LEO, then the modulator length for such NIN structure can be 1.5 to 3 mm for achieving 1V. If we then also reduce the modulator capacitor lateral width to 1 m or less (to increase the frequency response), combined with the shorter length for 1V, 100 GHz and 1V can then be reached concurrently.

(75) Both these alternative NIN and NPIN (and the conventional PIN) structures are good for modulator devices that operate under reverse bias, and are alternative exemplary embodiments for the modulators in the present invention. Other alternative variations include reverse biased NPIN, NPIN, NPIN, NPIN, PIN, PIN, PIN, PIN, or PIN structures; or PNIP, PNIP, PNIP, PNIP, PNIP or PIP, PIP, PIP, NIN, NIN structures; where quantum wells are placed in the P, N, and I layers (those layers in inverted commas X).

(76) Alternative Electro-Active Layer Structure B: PNN Structure

(77) The PIN (or NIN, NPIN) structures, while attractive in some ways, are suffering from their relatively long device length of over 1 mm, which will become challenging when one tries to reach very high frequency such as 100 GHz as in that case the electrodes' RF velocity matching with the optical velocity will have to be good, which is potentially possible to engineer but will make manufacturing more complex and costly. It would be desirable to reduce the modulator length to a much shorter length. It is also desirable to achieve a lower voltage of <0.5V. It turns out that it is possible to do so with use of the PNN structure described below.

(78) Another exemplary embodiment of the active-layer structure in the modulator structure of the present invention makes use of a novel PNN structure with or without QWs. Again strongly confined waveguide is used to reduce the voltage by for example 5 times.

(79) The use of PNN structure involves doped layers with or without QWs. Reverse bias is applied to the PNN junction (with positive voltage to the N side). The center N layer is doped with or without QWs. The doping enables carrier band-filling and plasma effects to be used to increase the phase shift. It enables BF+PL and also QCSE effects to be used beside the LEO effect. When properly designed, this enables the nonlinear EO response to be many times higher than that with the use of LEO plus the typical QW QCSE. For EA modulation, this enables the EA response to be many times higher than that with the use of the typical QW QCSE.

(80) Basically as is known to those skilled in the art, there is a carrier depletion layer at the PN junction (the carrier depletion is in a direction perpendicular to the PN junction creating a small carrier-free region D resulting in PDN in terms of carrier density distribution where D is depleted of carriers). Under reverse bias, the width of the depletion region (D) will increase. This means the carrier band-filing in the D region is modified, resulting in a change in the refractive index (at long wavelength of 1550 nm) due to BF effect. If quantum wells are placed in the D region, they will enhance the refractive index change due to BF. If quantum wells are used, there will be a change in the optical absorption also (at energy above the quantum wells' lowest energy level) due to band filling as well. When operating as EO modulator (i.e. based on the refractive index change), the operating wavelength will be at an energy way below the quantum wells' lowest energy level so the optical absorption or optical absorption modulation will be minimal, and the main optical modulation effect will be due to refractive index modulation.

(81) Further optimizing the design of the QW structure and doping can push the electro-optic or electro-absorption response even higher, but a price to pay for the optimized PNN structure is higher optical loss, which restricts the length of the modulator to be shorter than 1.0 mm. The higher EO or EA response, however, more than made up for the shorter length. A good design we have computed for the PNN+QW structure for EO response gives 1 Volt at 0.2 mm length and 50 GHz bandwidth (at modulator-capacitor width of w.sub.EC=0.7 m). The length is shorter than the 50 GHz RF wavelength, which is around 2 mm. As a result, it reduces the RF-optical velocity matching requirement and makes it easier to fabricate and easier to reach 50 GHz. It also opens up various exciting possibilities. For example, 0.25V is also achievable when a longer device length of less than 1.0 mm is used.

(82) This PNN modulator will be another exemplary embodiment of the active-layer structure for the modulators of the present invention. Other alternative variations include reverse biased PPN, structure (typically N doped is preferred for the middle layer for achieving lower optical loss but P doped is also possible). Other variations include PNN or PPN in which quantum wells are placed in both the P and N layers.

(83) Alternative Electro-Active Layer Structure C: NPNN Structure

(84) The short device length advantage of the PNN structures can be combined with the lower optical loss and lower electrical contact resistance advantage of NPIN (or NIN) structures to give both low optical loss and high modulation, by utilizing a NPNN structure described below. The NPIN (or NIN) structure only requires N Ohmic contacts as the two other layers contacting the metal contacts are both N-doped. As no P Ohmic contact is needed, it will reduce the total electrical series resistance of the modulator structure substantially, leading to higher frequency response. The material layers with P-dopant will also be thinner or fewer, which will also reduce the free-carrier induced optical absorption loss and further reduce the electrical series resistance as well (P-doped layer typically has 10 times higher optical absorption and 10 times higher electrical resistance than N-doped layer with the same carrier doping density).

(85) Another exemplary embodiment of the active-layer structure in the modulator structure of the present invention makes use of a novel NPNN structure with or without QWs. Again strongly confined waveguide is used to reduce the voltage by for example 5 times.

(86) The quantum wells, if used, are typically in the N layer but can also be in any of the N or P layers, including the transition region between the N and any of the two P or N layer, or in plurality of these layers. The active medium can be one or more of the layers (N, or P, or N layer), or the quantum-well structure, or other active-medium structure embedded in the electro-active layer, or a combination of effects from the quantum wells and one or more of the layers (N, or P, or N layer).

(87) The use of NPNN structure involves doped layers with or without QWs. Reverse bias is normally applied to the PN junction but with both sides being N contacts (with positive voltage on the N side of the PN junction). Thus the NP junction is forward biased. The center N layer has doped layers with or without QWs. The doping enables carrier band-filling and plasma effects to be used to increase the phase shift. It enables BF+PL and also QCSE effects to be used beside the LEO effect. When properly designed, this enables the nonlinear EO response to be many times higher than that with the use of LEO plus the typical QW QCSE. For EA modulation, this enables the EA response to be many times higher than that with the use of the typical QW QCSE.

(88) Further optimizing the design of the QW structure and doping can push the electro-optic or electro-absorption response even higher, but a price to pay for the optimized NPNN structure is higher optical loss, which restricts the length of the modulator to be shorter than 1.0 mm. The higher EO or EA response, however, more than made up for the shorter length. A good design we have computed for the NPNN+QW structure for EO response gives 1 Volt at 0.2 nm length and 100 GHz bandwidth (at modulator-capacitor lateral width of w.sub.EC=0.7 m). The length is shorter than the 100 GHz RF wavelength, which is around 1 mm. As a result, it reduces the RF-optical velocity matching requirement and makes it easier to fabricate and easier to reach 100 GHz. It also opens up various exciting possibilities. For example, 0.25V is also achievable when a longer device length of less than 1.0 mm is used. In addition, it can achieve lower optical loss and lower electrical contact resistance than the PNN structure.

(89) This NPNN modulator will be another exemplary embodiment of the active-layer structure for the modulators of the present invention. The NIN is simpler but the NPNN has better low-voltage and high-frequency performances and also almost as low an optical loss and as low the electrical contact resistance as the NIN structure. Other alternative variations include reverse biased NPPN, structure (typically N doped for the X layer is preferred for achieving lower optical loss but P doped is also possible).

(90) Still other alternative variations include NPNN, NPNN for which if QWs are used, they will be in both the P and N layers. For NPNN case, under reverse bias, the depletion width between the PN junction will widen and will sweep out carriers from the quantum wells in both the P side and the N side, resulting in twice the refractive index change than if the quantum wells are only in the N layer such as in a NPNN structure.

(91) Other alternative variations include reverse biased NPPN; or PNIP, PNPP, PNPP, PNPP, or PNNP structures; where quantum wells are placed in the P, N, and I layers (those layers in inverted commas X).

(92) Alternative Electro-Active Layer Structure D: Forward Biased NN(+)P(+)PIN Structure

(93) In the forward biased case, the low-optical-loss and low-electrical-resistance advantages of NIN structure may be achieved, by utilizing a NN(+)P(+)PIN structure described below.

(94) Another exemplary embodiment of the active-layer structure in the modulator structure of the present invention makes use of a novel NN(+)P(+)PIN structure. Again strongly confined waveguide is used to reduce the voltage by for example 5 times.

(95) The active medium is typically in the I layer but can also be in any of the N or P layer, including the transition region between the I and any of the N or P layer, in the other doped N(+) or P(+) or N or P layer, or in plurality of these layers. The active medium can be the layer itself, or a quantum-well structure, or other active-medium structure embedded in the electro-active layer.

(96) Forward bias is normally applied to the PIN junction (with positive voltage on the P side of the PIN junction). In that case, the N(+)P(+) junction is formally under reverse biased, which normally would not have much current flow. However, as is known to those skilled in the art, when the N(+) and the P(+) layers are highly doped (typically at a doping density of higher than about 110.sup.18/cm.sup.3 and preferably higher than 110.sup.19/cm.sup.3 for both the N and P material, the carrier can actually tunnel through under the reverse bias, resulting in current flow through the N(+)P(+) junction, into the PIN junction area that is forward biased. In that case, the N(+)P(+) junction is normally referred to as a carrier tunneling junction (or simply as tunnel junction). Such tunnel junction can be very thin with the N(+) and P(+) layer only tens of nanometers in thickness each. The net result is the changing of the P Ohmic contact at layer P to N Ohmic contact at layer N of the NN(+) side. As noted above, N Ohmic contact generally has a much lower (10 times lower) contact resistance than P Ohmic contact. The use of such pair of N(+)P(+) tunnel junction layers thus enables one to have N Ohmic contacts on both sides of the device. This structure also works if only electric field is wanted at the active medium (i.e. with PIN junction under reverse bias).

(97) This NN(+)P(+)PIN structure will be another exemplary embodiment of the active-layer structure for the devices of the present invention. Other alternative variations include NN(+)P(+)IN, N(+)P(+)PIN, N(+)P(+)IN structures and the likes or with some doping in the active-medium layer typically in the intrinsic layer resulting in NN(+)P(+)PNN, NN(+)P(+)PNN, NN(+)P(+)NN, N(+)P(+)PNN, N(+)P(+)PNN, N(+)P(+)NN, NN(+)P(+)PPN, NN(+)P(+)PN, N(+)P(+)PPN, or N(+)P(+)PN structures; or PIN, PP(+)N(+)NIP, PP(+)N(+)NPP, PP(+)N(+)NPP, PP(+)N(+)PP, P(+)N(+)NPP, P(+)N(+)NPP, P(+)N(+)PP, PP(+)N(+)NNP, PP(+)N(+)NP, P(+)N(+)NNP, or P(+)N(+)NP structures; where quantum wells are placed in the P, N, and I layers (those layers in inverted commas X). They are good for devices that operate under forward bias for the PIN (or P(+)IN or PIN(+) or P(+)IN(+)) junction. They are also good for devices that operate under reverse bias for the PIN (or P(+)IN or PIN(+) or P(+)IN(+)) junction, and are alternative exemplary embodiments for the active photonic devices in the present invention.

(98) When the context is clear below, we will drop the inverted commas in N or P designations in the electro-active layer structures above. The above examples are for the purpose of illustrations and not limitations. For example, the various doped structures may also be joint one on top of another forming a cascaded structure. Those skilled in the art will know other obvious variations that are variations of the above examples of the various doped structures with or without the use of quantum wells.

(99) Slow-Wave Electrode Structure for Velocity+Impedance Matching

(100) In certain applications such as traveling-wave modulator, a travelling-wave RF transmission line electrode structure should be fabricated along the device waveguide. Such traveling-wave RF transmission line electrode structure is often needed in order to achieve high-frequency response of 10-100 Gb/s or higher for the modulator. Below describe such a travelling wave electrode and their optimization to match the velocity of propagation of the optical beam and the RF wave. In such travelling wave electrode, it is often advantages to engineer the electrode impedance to be around the standard impedance of 50 or some other preferred value depending on the application.

(101) As the RF dielectric constant in III-V semiconductors is close to their dielectric constant at optical frequency, and the RF wave in the case of semiconductor modulators tends to have electric field fringing to the surrounding materials with lower dielectric constant, the RF wave tends to propagate at a faster velocity than the optical wave. This can be managed by using an adjustable slow-wave capacitively-loaded traveling wave (CL-TWE) RF transmission-line structure as shown in FIG. 5a. The price to pay is a longer length, trading off mainly optical loss (not much of RF loss) but it allows impedance matching.

(102) It turns out that the slow-wave structure also enables freedom to engineer concurrent impedance matching to 50 as there is freedom to choose its filling factor F that will change its effective inductance-length product L and capacitance per unit length C. It usually ended up with slightly larger voltage-length product than if the impedance is allowed to be lower than 50, resulting in longer length for the same modulation voltage. Most of the structures above have plenty of rooms to absorb the longer length. Hence, velocity matching and 50 can be engineered. However, velocity matching is less important when the modulator is shorter than 0.2 mm as the RF wavelength at 100 GHz is about 1 mm.

(103) Exemplary Device for the Modulators of the Present Invention

(104) The exemplary device below illustrates a particular exemplary embodiment of the EO modulator of the present invention, including the travelling-wave electrode, the Mach Zehnder Interferometer, and push pull geometry. The general scheme for the modulator is illustrated in FIG. 4. In one particular exemplary embodiment and realization of such modulator, the said modulator can be fabricated on a silicon photonics platform for operation at the fiber-optic communication wavelength of around 1550 nm. The modulator is fabricated on a silicon-on-insulator (SOI) wafer for which a thin layer (about 300 nm thick) of silicon is pre-bonded onto a thermal silicon oxide layer with a thickness of 1-2 micrometers grown on top of a bottom silicon substrate. The 300 nm thick top-layer silicon on SOI wafer serves as an on-chip optical waveguide, which can be fabricated into the channel waveguide shown in FIG. 4b, for which the optical beam is propagated in the channel waveguide as shown in FIG. 4b. In this particular embodiment, the channel waveguide core is made of high-refractive-index silicon (with a refractive index n.sub.core3.6). The waveguide core is surrounded by silicon dioxide or air with much lower refractive index of n.sub.clad1 to 1.5, that act as waveguide cladding. The use of silicon as waveguide enables electronic integrated circuits to be fabricated on the same chip. Such waveguides are well known to those skilled in the art and are often referred to as silicon photonics platform. If active photonic devices such as modulators are then fabricated, the entire platform is some time referred to as Electronic-Photonic Integrated Circuits (EPICs).

(105) For the modulator shown in FIG. 5a, a pair of travelling-wave RF transmission line electrodes is used to propagate the RF electrical signal along the EO modulator. These electrodes are designed so that the RF wave propagates at a velocity close to the optical wave velocity in the modulator's optical waveguide. This is referred to as velocity matched transmission line. As is known to those skilled in the art, velocity matching is desirable when the modulator length is long comparing to the RF wavelength so that maximum modulation can be achieved in the optical beam.

(106) In this example for the case of an EO modulator geometry, an optical beam is further split into two propagating arm forming a Mach Zehnder Optical Interferometer (MZI). Each arm has an EO phase modulator. As is known to those skilled in the art, it is typically arranged so that the RF wave will cause positive phase shift in one of the arm and negative phase shift in the other arm, resulting in what is known as push-pull configuration for the modulator. As is known to those skilled in the art, a number of configurations can be used beside the push pull configuration. Thus such RF electrode and MZI geometry are shown only for the purpose of illustration and not limitation.

(107) An embodiment of the present invention is focused on the realization of an EO phase modulator that can be used as shown in a MZI geometry or as an isolated phase modulator for various applications as is well known to those skilled in the art. The RF electrodes illustrate an exemplary embodiment in a particular application of the modulators of the present invention.

(108) For this particular exemplary embodiment, the RF electrodes are a pair of capacitively-loaded traveling wave electrodes (CL-TWE). Each arm of the CL-TWE is electrically in contact with electrode traveling along an optical waveguide based EO optical phase modulator. The EO optical phase modulator is reverse biased by a DC applied voltage. Thus there are two phase modulators, one power by each arm. The two optical waveguiding arms then form a Mach Zehnder optical Interferometer (MZI) geometry for an input optical beam. The input optical beam is split with half the power going to each arm. At the output, the phase modulated beams from the two phase modulator outputs are then combined again giving the optical output for the MZI. As is well known to those skilled on the art, the MZI enables the phase modulation to be converted to optical intensity modulation at the beam combined optical output of the MZI. In the case of a push-pull geometry, as is known to those skilled in the arm, the two arms will receive opposite optical phase shift, enabling double the intensity modulation under the same applied voltage to the two phase modulators, one on each arm of the MZI.

(109) In order to realize a push-pull geometry, in terms of powering the two arms of the Mach Zehnder Interferometer (MZI), the illustrated scheme follows that of the PIN modulator. As shown in FIG. 5a, RF wave is launched into the CL-TWE from one end of the CL-TWE. The segmented periodic pairs of T-shaped electrodes bring the voltage from the main slotted electrodes to the modulator device. Each pair is separated by a periodic length l.sub.p=l.sub.m+l.sub.g and each of the T-shape segment has a length l.sub.m. The fill factor F is defined as F=l.sub.m/l.sub.p (see FIG. 5a), where A/B denotes division of A over B. Typically l.sub.p=50-100 m. Along the gap l.sub.g=l.sub.pl.sub.m, the modulator section has no applied voltage.

(110) In this particular exemplary embodiment, the bottom parts of the two phase modulators are connected. This is shown in FIG. 5b showing the cross-section f-f in FIG. 5a. Thus, any voltage Vs applied is split +Vs/2, and Vs/2 on the two arms forming a push-pull pair of phase modulators. Unlike the conventional push-pull modulators, there is no net voltage reduction by half as the line voltage Vs is twice the voltage applied to each modulator Vs/2. However the junction capacitances (C.sub.j in FIG. 5a) of the two phase modulators are connected in series, reducing the total capacitance or increasing the frequency, so there is a gain over the conventional push-pull EO modulator in broader frequency response (at the expense of higher voltage).

(111) The equivalent lumped-element circuit of such a CL-TWE powering the two phase modulators is shown in FIG. 6a, in which the dotted line labeled as Transmission Line represents the basic unloaded transmission line model for the CL-TWE. The solid line part labeled as Device Load takes into account RF propagation loss of all the modulator structural elements including: (1) Metal Ohmic Contact Resistance and Capacitance Z.sub.m; (2) Resistance and Capacitance Z.sub.d of the Doped Waveguide Layers including those in the TCO materials (if TCO is used); (3) Modulator active-region Capacitance C.sub.j; (4) Both Transverse and Longitudinal Current losses in the Semiconductor Structure are included. The model gives us the frequency bandwidth limited by the RF losses (for long modulator) or by the RC time constant (for short modulator), and the transmission line impedance. It also gives us the CL-TWE filling factor F needed to achieve velocity matching.

(112) FIG. 6b shows the top view of the CL-TWE lines for the traveling wave electrodes. The voltage V.sub.b in FIG. 6b connecting between the bottom common contact and the top of one of the modulators giving V.sub.b1=V.sub.b2=V.sub.b (when Vs=0) is used to reverse bias the modulators. DC offset in Vs gives differential biasing that can be used to tune the operating point of the Mach Zehnder optical interferometer by tuning the differential phase shift imposed by the two phase modulators along its two arms.

(113) As is known to those skilled in the art, there are many other electrode structures that can be used. The above illustrates one embodiment of a traveling wave structure that can slow down the propagation velocity of the RF wave so as to achieve better velocity matching with the optical beam in the modulator (it is some time referred to as slow-wave RF traveling-wave structure). Such velocity matching will help to achieve higher frequency response as is known to those skilled in the art. The above exemplary embodiment on the traveling-wave electrodes is shown for the purpose of illustrations and not limitation.

(114) In terms of the optical beam, it enters the silicon waveguide from an optical fiber. There are many ways to couple optical beam from an optical fiber to silicon waveguide as is known to those skilled in the art such as via an integrated mode size transformer on silicon called Super-High-Numerical Aperture Graded Refractive Index (SuperGRIN) lens, which will efficiently couple the beam power from the optical fiber to the 300 nm thick silicon waveguide on a SOI substrate (SOI waveguide). Alternative fiber to silicon waveguide couplers such as tapering down waveguide or surface grating can also be used as is well known to those skilled in the art.

(115) The beam from the SOI waveguide is then coupled into the thin-film modulator structure as shown by FIG. 7a/b using two tapers (one on the active material layer (e.g. InP/III-V), one on the primarily passive waveguide layer (e.g. Si)see the top view). Due to the thin-thickness of the InP/III-V modulator's active layer structure (only 300-400 nm thick) for operation at 1550 nm wavelength range, this taper section can be very short (like 5-30 m depending on the thickness) and over 95% of the energy can be transferred into the thin film modulator device. The output back into the optical fiber goes through a reverse process via two output tapers (one on the active material layer (e.g. InP/III-V), one on the primarily passive waveguide layer (e.g. Si)see the top view) to bring the optical beam to the primarily passive waveguide layer (e.g. Si) and then to an output fiber-coupling lens.

(116) The general structure for the NIN and NPNN modulators are also similar. What differentiate them are the detailed layer structures. There are two versions of the general structure, one uses side-conduction layer (SCL), another Ohmic Transparent Conductor (OmTC).

(117) SCL Case.

(118) In a particular exemplary embodiment for application to 1550 nm wavelength range, the cross-section for the SCL case is shown in FIG. 8a. At the bottom, the modulator has a highly doped lower layer that is about 100 nm thick for this particular exemplary embodiment. This layer is used to conduct the voltage side way and is called bottom side conduction layer (BSCL). As noted above, in the push-pull modulator circuit scheme for the particular exemplary embodiment, this layer is connected to the phase modulator at the other arm of the MZI.

(119) Above this BSCL is an active EC layer structure that will be different for NIN and NPNN. It is then followed by a 100 nm thick top layer. This top layer is largely N-doped and is used to conduct voltage side way and is called the top side conduction layer or TSCL. On top of this at the middle is deposited with 300 nm-thick low-refractive-index SiO.sub.2. Both sides of TSCL are deposited with metal. The center SiO.sub.2 layer prevents optical energy in the waveguide from touching the optically lossy metal.

(120) The center active EC layer has to be narrow in width to make the device capacitance small as its width w.sub.EC will define the modulator-capacitor width w.sub.CAP=w.sub.EC. To enable fabrication, one way is to use a material for this active EC layer that can be chemically selectively etched sideway without etching the material that formed the BSCL or the TSCL layers. Other fabrication schemes may be used as long as the small width for w.sub.EC is achieved (e.g. by etching the required width for w.sub.EC, then depositing the material required for forming the TSCL layer). In order to achieve high frequency response, an exemplary modulator structure requires w.sub.EC to be around 2 m=2000 nm or as narrow as 0.7 m=700 nm. While it may seem to be small (making it challenging to fabricate), it is still larger than the typical thickness of this thin-film modulator structure with a thickness of 300-400 nm. Thus, the width w.sub.EC still has a low aspect ratios (<1:3) with the other nearby structural parameters. Low aspect ratios make it not too difficult to fabricate. It can be done by careful control of the etching. Both sides of the metal contact will be around 2 m=2000 nm in order to have a large enough Ohmic contact area with metal so as to have small enough metal contact resistance.

(121) OmTC Case.

(122) In a particular exemplary embodiment for application to 1550 nm wavelength range, the cross-section for the OmTC case is shown in FIG. 8b. In the case of a OmTC based structure, there will be no TSCL. Thus, in this particular exemplary embodiment, the top layer thickness can be reduced by about 60 nm, leaving 40 nm for Ohmic contact with TC (Transparent Conductor) material in general or TCO (Transparent Conducting Oxide) material in particular. This 60 nm thickness can be added to the bottom BSCL making it 160 nm thick resulting in lower resistance and more robust structure. The top layer contacting the TC is N-doped for achieving good Ohmic contact. In this particular exemplary embodiment, the TC is In.sub.2O.sub.3. As the TC generally has low refractive index (e.g. n1.7), the optical field energy will decay rapidly in the TC and will not touch the metal much. As TC can be deposited after the dry etching to form the width W.sub.EC needed for the electro-active (EC) layer, no side etching of the center electro-active layer (ECL) is needed. We just need to etch vertically for the require thickness and then planarize with polymer, then deposit with TC. Hence, this OmTC case will enable a structure that is more robust, with lower resistance, and easier to fabricate than the SCL case.

(123) Detailed Structures for NIN and NPNN Modulators

(124) NIN Structures.

(125) In a particular exemplary embodiment for application to 1550 nm wavelength range, the detail structures for a NIN modulators are shown in FIG. 9. In this particular exemplary embodiment, the structure has a highly-doped bottom InP N-layer about 120 nm thick (this becomes the BSCL) followed by a 195 nm thick intrinsic layer made of AlGaInAs QW structure (with .sub.QW1350 nm), and then by another highly-doped top InP N-layer about 125 nm thick for side contact with metal. This gives a total thickness of 440 nm. For this structure, the active EO region width w.sub.EC or the modulator-capacitor width w.sub.CAP defined by it (w.sub.CAP=w.sub.EC) is chosen to be 2,000 nm (2 m). This width also gives the lateral optical mode confinement as shown by FIG. 10b. In FIG. 10a, we show the mode of a conventional modulator in comparison. The NIN mode in FIG. 10b is 4 times smaller in the vertical direction, giving around 4 times larger mode-medium overlapping factor.

(126) The computed modulator length for V=1V and F=1 is L.sub.MOD=2 mm.

(127) Table 1 show the material layer structure for this NIN structure based modulators with side metal contact for its top contact, spelling out the thicknesses and bandgap energies of the compound semiconductor material in each layer with the various doping density and strain (with InP as the substrate).

(128) TABLE-US-00001 TABLE 1 Layer Layer Number Thickness NIN CASE Metal Doping 1 120 nm InP (Bottom Layer-just n = 1 above the substrate) 10{circumflex over ()}19 2 5 nm AlGaInAs 1.3 um I 3 5 nm barrier AlGaInAs/1.1 um/0.8% I tensile strained 4 2 7 nm AlGaInAs/1.1 um/0.8% I barrier inside tensile strained 5 3 6.5 nm Well AlGaInAs/1.55 um/0.9% I (PL = 1350 nm) compressive strained 6 5 nm barrier AlGaInAs/1.1 um/0.8% I tensile strained 7 60 nm AlGaInAs 1.3 um I 8 125 nm InP (Top Layer) n = 1 10{circumflex over ()}19 Total 440 nm

(129) NPNN Structure.

(130) In a particular exemplary embodiment for application to 1550 nm wavelength range, as shown in FIG. 10, a NPNN structure has a highly-doped bottom InP N-layer about 160 nm thick, followed by a 115 nm thick active Electro-Active layer (ECL) made of AlGaInAs QW structure, then by a thin layer of InP P-doped layer (around 25 nm), then by a highly-doped top InP N-layer about 80 nm thick for top contact to Om-TCO. This gives a total thickness of 380 nm. The Electro-Active region width w.sub.EC is chosen to be 2,000 nm (2 m).

(131) TABLE-US-00002 TABLE 2 Layer Layer Number Thickness NPNN TCO CASE Doping 1 160 nm InP (Bottom layer-just n = 1 above the substrate) 10{circumflex over ()}19 2 10 nm AlGaInAs 1.3 um n = 4 10{circumflex over ()}17 3 4 nm barrier AlGaInAs/1.1 um/0.8% n = 4 tensile strained 10{circumflex over ()}17 4 2 7 nm AlGaInAs/1.1 um/0.8% n = 4 barrier inside tensile strained 10{circumflex over ()}17 5 3 6.5 nm Well AlGaInAs/1.55 um/0.9% n = 4 (PL = 1350 nm) compressive strained 10{circumflex over ()}17 6 4 nm barrier AlGaInAs/1.1 um/0.8% n = 4 tensile strained 10{circumflex over ()}17 7 63 nm AlGaInAs 1.3 um n = 4 10{circumflex over ()}17 8 25 nm InP p = 1 10{circumflex over ()}18 9 80 nm InP (Top layer) n= 1 10{circumflex over ()}19 Total 380 nm

(132) As shown in FIG. 10, comparing to NIN case, the main difference for this particular exemplary embodiment of a NPNN case is that the active EC layer has QWs that are also doped with N-type carriers and it is next to a bottom p-doped layer. Under a reverse bias, the depletion width between the p-doped layer and the N-doped QW region is enlarged with voltage. The voltage drop across this depletion region then imposes a constant electric field on the QW. The increase in phase shift is because more QWs are applied with the electric field when the voltage increases. The doped carriers also cause phase shift due to band-filling (BF) and Plasma (PL) effects. When optimized, the BF+PL+LEO+QCSE can lead to 3-5 times larger phase shift than LEO+QCSE. The thinner structure of 380 nm comparing to NIN of 440 nm also reduces the modulator length by another 20%. The computed modulator length for V=1V is L.sub.MOD=0.4 mm.

(133) Table 2 shows the material layer structure for the NPNN structure based modulators with Om-TCO contact for its top contact, spelling out the thicknesses and bandgap energies of the compound semiconductor material in each layer with the various doping density and strain (with InP as the substrate).

(134) The Role of F for Velocity & Impedance Matching

(135) In a particular exemplary embodiment for application to 1550 nm wavelength range, the lumped-element model shown in FIG. 6a is used to estimate the modulator device performances. We use a Table to summarize the different cases. We will illustrate some plots using a NIN case with SCL, and NPNN case with OmTC.

(136) NIN CASE with SCL.

(137) In a particular exemplary embodiment, the modulation BW for the NIN case with 2-m-wide active region so w.sub.CAP=2 m is shown in FIG. 11.

(138) A calculation shows that for the NIN structure, in order to reach 1V modulation voltage, the modulator length shall be 2 mm. FIG. 11a shows the net frequency response limited by RF loss (in dB/mm). Taking the 3.24 dB/mm loss line in FIG. 11a, one finds a frequency response bandwidth of 60 GHz for the NIN structure with side metal conduction. These are for w.sub.CAP=2 m. FIG. 11b shows the impedance for different F values as a function of the frequency. FIG. 11c shows the propagating refractive index of the RF wave along the RF transmission line for different F values as a function of the frequency illustrating F0 makes the RF refractive index close to 3.5, which is nearly matching the propagating refractive index of the optical beam in the modulator (typically around n3 to 3.5), which is required for velocity matching.

(139) NPNN Case with OmTC.

(140) In a particular exemplary embodiment, the modulation bandwidth for the NPNN case with 2-m-wide active region so w.sub.CAP=2 m (this width determines the Modulator Capacitance C.sub.j) is shown in FIG. 12.

(141) A calculation shows that for the NPNN structure, in order to reach 1V modulation voltage, the modulator length shall be 0.4 mm. FIG. 12a shows the net frequency response limited by RF loss (in dB/mm). Taking the 34.2 dB/mm loss line in FIG. 12a, one finds a frequency response bandwidth of 100 GHz for the NPNN structure with Ohmic TCO as top contact. These are for active-region width of 2 m. FIG. 12b shows the impedance for different F values as a function of the frequency. FIG. 12c shows the propagating refractive index of the RF wave along the RF transmission line for different F values as a function of the frequency illustrating F0.4 makes the RF refractive index close to 3.5, which is nearly matching the propagating refractive index of the optical beam in the modulator (typically around n3 to 3.5), which is required for velocity matching.

(142) More Detailed Descriptions of the Various Embodiments of the Present Invention

(143) A schematics showing the general geometry of the Active Photonic Devices of the present invention is shown in FIG. 13. FIGS. 13a and 13b show a Active Photonic Device 20000 with a modulating section or referred to below as Active-Layer Structure ALS 22500 section, coupled to an input connecting waveguide core ICWCo 22200. FIG. 13b is a semi-transparent illustration of FIG. 13a. The input connecting waveguide core ICWCo 22200 is fabricated on an input connecting-waveguide bottom cladding material ICWBCd 22200B disposed on a substrate SUB 21100.

(144) In one exemplary embodiment shown in FIG. 14, illustrating an exemplary embodiment of the cross section at location A-A of FIG. 13, to the top of the input connecting waveguide core is further deposed with input connecting waveguide top cladding material ICWTCd 22200T. Bottom refers to direction closer to the substrate and top refers to direction away from the substrate. In FIG. 14, to the left of the input connecting waveguide core is also deposed with input connecting waveguide left cladding material ICWLCd 22200L and to the right of the waveguide core is disposed with input connecting waveguide right cladding material ICWRCd 22200R. Right or left is defined by taking the direction of beam propagation as the front direction and right or left means relative to this front direction. The division of the cladding into four different material regions is for the purpose of discussion and not limitation as these can be all the same materials or there can be more than 4 regions forming plurality of different regions as long as these regions act as waveguide cladding materials with effective refractive indices smaller than the effective refractive index of the waveguide core ICWCo 22200 so that the light beam power is confined mainly in the region of the waveguide core ICWCo 22200, as is well known to those skilled in the art. The cladding material is a general designation that can include air or vacuum or any transparent dielectric material as the material.

(145) Input Connecting Waveguide Region

(146) The input connecting waveguide core ICWCo 22200 is made up of a material or mixture of materials with an averaged material refractive index n.sub.ICWCo 22200n, has a thickness d.sub.ICWCo 22200d, and width W.sub.ICWCo 22200w. Let the refractive index of the bottom input connecting-waveguide bottom cladding material be n.sub.ICWBCd 22200Bn. Let the refractive index of the top cladding material ICWTCd 22200T be n.sub.ICWTCd 22200Tn, the refractive index of the left cladding material ICWLCd 22200L be n.sub.ICWLCd 22200Ln, and the refractive index of the right cladding material ICWRCd 22200R be n.sub.ICWRCd 22200Rn. The waveguide core 22200 and the claddings 22200T, 22200B, 22200R, 22200L, together forms input connecting waveguide ICWG 22200WG.

(147) The vertical confinement of the optical beam is due to the refractive-index difference between the top and bottom waveguide claddings and the waveguide core and the claddings generally have lower refractive indices than that of the waveguide core so that n.sub.ICWTCd<n.sub.ICWCo and n.sub.ICWBCd<n.sub.ICWCo. The horizontal confinement of the optical beam is due to the refractive-index difference between the left and right waveguide claddings and the waveguide core and the claddings generally have lower refractive indices than that of the waveguide core so that n.sub.ICWRCd<n.sub.ICWCo and n.sub.ICWLCd<n.sub.ICWCo. The vertical direction is the direction perpendicular to the substrate plane and the horizontal direction is the direction parallel to the substrate plane.

(148) The above illustration of an exemplary embodiment of input connecting waveguide ICWG 22200WG, showing the waveguide cladding can be divided into different material regions (in the above case with four main material regions), is for the purpose of illustration and not limitation. As is known to those skilled in the art, the waveguide cladding can be made up of one single material or plurality of material regions, as long as the refractive indices of most of the cladding material regions is lower than the refractive index n.sub.ICWCo of the waveguide core. This is also generally applicable to the other waveguide cladding situations below for other optical waveguides described in the present invention.

(149) Definition of Refractive Index Contrast and Cladding Refractive Index Averaging

(150) An important quantity in terms of waveguide mode confinement is the refractive index contrast between the averaged refractive index of the waveguide core and its immediate surrounding cladding materials called the refractive-index difference n.sub.Rd defined by n.sub.Rd.sup.2=n.sub.Co.sup.2n.sub.Cd.sup.2), where n.sub.Co is the refractive index of the waveguide core (e.g. n.sub.Co=n.sub.ICWCo) and n.sub.Cd is the refractive index of the waveguide cladding (e.g. n.sub.Cd=n.sub.ICWBCd or n.sub.ICWTCd or n.sub.ICWRCd or n.sub.ICWLCd) or an averaged of them thereof given by n.sub.aICWCd 22200aCdn.

(151) The refractive-index averaging is more accurately done as averaged of its square values which are their dielectric constant =n.sup.2. This is because dielectric constants which describe the dipole strengths add linearly with each other as is known to those skilled in the art. Thus n.sub.aICWCd.sup.2 for example can be computed by weighting the refractive index square in each of the different cladding regions by the fraction of beam energy in each of the cladding regions. Hence:
n.sub.aICWCd.sup.2=(n.sub.ICWBCd.sup.2A.sub.ICWBCd+n.sub.ICWTCd.sup.2A.sub.ICWTCd+n.sub.ICWRCd.sup.2A.sub.ICWRCd+n.sub.ICWLCd.sup.2A.sub.ICWLCd)/(A.sub.ICWBCd+A.sub.ICWTCd+A.sub.ICWRCd+A.sub.ICWLCd),(13)

(152) where in Eq. (13), A.sub.ICWBCd is some effective cross-sectional weighting for the optical power in the bottom cladding material (e.g. given by the percentage of the total beam power), A.sub.ICWTCd is some effective cross-sectional weighting for the optical power in the top cladding material, A.sub.ICWRCd is some effective cross-sectional weighting for the optical power in the right cladding material, A.sub.ICWLCd is some effective cross-sectional weighting for the optical power in the left cladding material. A.sub.ICWBCd, A.sub.ICWTCd, A.sub.ICWRCd, and A.sub.ICWLCd are called the effective beam power distribution areas in the respective regions of the waveguide cladding materials. Each of these cross-sectional weighting has a value proportional to the fractional optical power (beam power integrated over the beam cross-sectional area of that region) in that region of the material for the guided optical beam or is given by the integration over the beam energy density (energy per unit volume) over the volume of that region of the material assuming the volume is taken over a short propagation length. These are some definitions of the effective cross-sectional weighting labeled with prescript A. Many other equivalent but approximate definitions of the effective cross-sectional weighting A can be used. Note n.sub.aICWCd.sup.2(n.sub.ICWBCd.sup.2+n.sub.ICWTCd.sup.2+n.sub.ICWRCd.sup.2+n.sub.ICWLCd.sup.2)/4, if these weightings are about equal.

(153) Likewise the waveguide core can also generally be made up of one or plurality of materials, and n.sub.ICWCo=n.sub.aICWCo can also be an averaged refractive index of the m number of materials with slightly different refractive indices n.sub.ICWCo1, n.sub.ICWCo2, n.sub.ICWCo3 . . . n.sub.ICWCom, that made up the waveguide core materials where
n.sub.aICWCo.sup.2=(n.sub.ICWCo1.sup.2A.sub.ICWCo1+n.sub.ICWCo2.sup.2A.sub.ICWCo2+n.sub.ICWCo3.sup.2A.sub.ICWCo3+ . . . +n.sub.ICWCom.sup.2A.sub.ICWCom)/(A.sub.ICWCo1+A.sub.ICWCo2+A.sub.ICWCo3+ . . . +A.sub.ICWCom),(14)

(154) In Eq. (14), each of the A.sub.ICWCo1, . . . , A.sub.ICWCom is some effective cross-sectional weighting A.sub.ICWCoj for the optical power in core material with refractive index n.sub.ICWCoj, where j is one of 1, . . . , m. A.sub.ICWCo1+A.sub.ICWCo2+A.sub.ICWCo3+ . . . +A.sub.ICWCom are called the effective beam power distribution areas in the respective regions of the waveguide core materials.

(155) Input Optical Beam

(156) As shown in FIG. 13, an input optical beam IBM 22140 is launched into input connecting waveguide core ICWCo 22200. As the cladding regions of input connecting waveguide ICWG 22200WG or in fact of any waveguide are typically ill-defined (the beam power can go into various depths and directions into the cladding) as is known to those skilled in the art, an input optical beam IBM 22140 launched into waveguide ICWG 22200WG is taken to mean it is launched with its power centered essentially at the waveguide core ICWCo 22200, so the reference to the beam being in waveguide 22200WG or waveguide core 22200 will generally be used interchangeably below. As is known to those skilled in the art, the input optical beam IBM 22140 will propagate in waveguide 22200 with a propagating refractive index n.sub.IBM 22140n. This propagating refractive index n.sub.IBM 22140n generally has a value smaller than the material refractive index n.sub.ICWCo 22200n of the waveguide core ICWCo 22200 so that n.sub.IBM<n.sub.ICWCo as is known to those skilled in the art. Also, the propagating refractive index n.sub.IBM 22140n generally has a value larger than the material refractive index n.sub.ICWQCd 22200Qn of the waveguide cladding ICWQCd 22200Q where Q=T, B, L, or R, so that for most of the Q, n.sub.IBM>n.sub.ICWQCd to enable waveguiding, as is known to those skilled in the art. The optical beam has an optical power given by P.sub.IBM 22140P, electric field polarization given by E.sub.IBM 22140E, and beam effective area given by A.sub.IBM 22140A.

(157) In this invention, the propagating optical beam is generally assumed to have a spread of optical wavelength centered at an operating wavelength .sub.IBM 22140L. For illustration and not limitation, the optical beam may be in the form of a train of optical pulses to transmit digital information. The optical beam may also be made up of light wave of one or plurality of (N) different frequency channels (.sub.IBM1, .sub.IBM2, .sub.IBM3, . . . , .sub.IBMN) where N is an integer. When the optical beam is made up of plurality of frequency channels, the optical transmission system or device is generally known as a wavelength division multiplexing (WDM) optical system or device. Generally, the optical beam is made up of beam of light with a spectral width around the center operating wavelength .sub.IBM.

(158) Input Beam Coupler Structure (IBCS) Region

(159) FIG. 15a show the input beam coupler structure (IBCS) comprises at least a tapering waveguide section (preferably tapering from wide to narrow but can also maintain the same width or taper from narrow to wide as will be elaborated below) connected to the input waveguide. Optionally, the active layer structure ALS on top of the input tapering waveguide section can also be tapering in the form of an up taper (preferably tapering from narrow to wide in the direction toward the active layer structure ALS, but can also maintain the same width or taper from wide to narrow). See for example FIG. 15b.

(160) Specifically, the input optical beam IBM 22140 enters from input connecting waveguide core ICWCo 22200 into an input connecting-waveguide taper section with an input tapering waveguide core ITWCo 22300 parameterized by a location z1 (FIG. 15a), ITWCo-z1 22300z1, in which the width of a tapering waveguide core w.sub.ITWCo-z1 22300w-z1 at distance z1, measured from the beginning point of the taper, is changed from its input value at z1=0 w.sub.ITWCo-z1=0 22300w-z1=0 of w.sub.ITWCo-z1=0=V.sub.ITWCo 22200 to another width (that can be the same width) at z1>0 w.sub.ITWCo-z1>0 22300w-z1>0. The thickness of the tapering waveguide core is d.sub.ITWCo-z1 22300d-z1 and its refractive index is n.sub.ITWCo-z1 22300n-z1. Typically, though not always, d.sub.ITWCo-z1 22300d-z1 and n.sub.ITWCo-z1 22300n-z1 are constant value in z1 so we can drop the z1 designation with d.sub.ITWCo-z1=d.sub.ITWCo 22300d and n.sub.ITWCo-z1=n.sub.ITWCo 22300n. In a preferred embodiment, d.sub.ITWCo-z1 is about the same value as d.sub.ICWCo. The tapering may be such that w.sub.ITWCo-z1 is a linear function of z1 or quadratic function of z1 (i.e. depending on z1.sup.2), but can also be of any curvilinear function of z1. Let g.sub.ITWCo 22300g denotes the total length of this tapering waveguide.

(161) The end of the taper at z1=g.sub.ITWCo 22300g at which the width of the waveguide core is w.sub.ITWCo-g 22300w-g is connected to an input supporting structure ISTR 21200. While illustrated as a line that is continuation of the connecting waveguide material with a narrow width and air or other low refractive index materials surrounding its side, the supporting structure can be random dots or any shape of small amount of any materials that have an effective refractive index or small averaged refractive index (e.g. as defined by Eq. (13)) within the layer extended in the horizontal direction, given by an effective layer averaged refractive index n.sub.laISTR 21200nla. In the case it acts as the bottom waveguide cladding, n.sub.laISTR has a value lower than the refractive index of the waveguide core n.sub.WCo 22600Con in the electro-active waveguiding core structure EWCoS 22600 defined below. The input supporting structure ISTR 21200 may continue to guide wave or just acts as a supporting structure, depending on application scenarios.

(162) In an exemplary embodiment, the input supporting structure ISTR 21200 is a narrow line. In that particular case, we can describe it as having a width w.sub.ISTR 21200w, thickness d.sub.ISTR 21200d, and length g.sub.ISTR 21200g. The length g.sub.STR 21200g may be zero. In that case, input supporting structure ISTR 21200 does not exist (the thin ALS film can still be supported in some way such as by its corners or sides, but not directly below). In a preferred embodiment, d.sub.ISTR is about the same value as d.sub.ICWCo.

(163) Along the taper in region outside the ALS region, the vertical confinement of the optical beam is due to the refractive-index difference between the waveguide core and top and bottom tapering waveguide claddings at the location z1 defined above: ITWTCd-z1 22300T-z1 (refractive index n.sub.ITWTCd-z1 22300Tn-z1) and ITWBCd-z1 22300B-z1 (refractive index n.sub.ITWTCd-z1 22300Bn-z1) and the waveguide core and the claddings have lower refractive indices than that of the waveguide core so that the refractive index n.sub.ITWTCd-z1<n.sub.ITWCo-z1 and n.sub.ITWBCd-z1<n.sub.ITWCo-z1. The horizontal confinement of the optical beam is due to the refractive-index difference between the left and right waveguide claddings at z1: ITWLCd-z1 22300L-z1 (refractive index n.sub.ITWLCd-z1 22300Ln-z1) and ITWRCd-z1 22300R-z1 (refractive index n.sub.ITWRCd-z1 22300Rn-z1), and the waveguide claddings have lower refractive indices than that of the waveguide core so that n.sub.ICWRCd-z1<n.sub.ICWCo-z1 and n.sub.ICWLCd-z1<n.sub.ICWCo-z1. The vertical direction is the direction perpendicular to the substrate plane and the horizontal direction is the direction parallel to the substrate plane. Again, there can be one or plurality of cladding material regions, and the four cladding regions are mentioned for the purpose of illustration and not limitation.

(164) In an exemplary embodiment, n.sub.ITWTCd-z1=n.sub.ITWBCd-z1=n.sub.ITWLCd-z1=n.sub.ITWRCd-z1=n.sub.ICWTCd, and n.sub.ICWTCd=n.sub.ICWBCd=n.sub.ICWLCd=n.sub.ICWRCd so all the cladding indices in the tapering regions and the input connecting waveguide regions are all approximately equal. For example, these cladding regions can be filled with silicon dioxide materials with refractive index of n1.45. The refractive index of the waveguide core n.sub.ITWCo-z1 22300n-z1 can be silicon so that n.sub.ITWCo=n.sub.ICWCo3.6, where n.sub.ICWCo 22200n is the refractive index of the waveguide core for the input connecting waveguide.

(165) On top of the input tapering waveguide core ITWCo 22300 starting at z1=z1ALS 22300z1ALS, is laid with an active layer structure ALS 22500. Typically z1ALS is before g.sub.ITWCo 22300g so that 0<z1ALS<g.sub.ITWCo. The active layer structure starting at z1ALS can also have an up-taper with width tapering from narrow to wide in the direction toward the ALS structure. The various embodiments of this active layer structure ALS 22500 will be described in more detail below.

(166) Alignment Insensitive Input Beam Coupler Structure (AI-IBCS)

(167) (1) Broadened input region, preferably narrower then mode 3.

(168) (2) Top taper first

(169) (3) Close to equal width

(170) (4) Narrow the lower one down

(171) (5) Narrow the top one down

(172) (6) Zigzag situation

(173) Active Layer Structure-Beam Transport into the Structure

(174) Bottom Side Conduction and Ohmic Contact Layer

(175) FIGS. 16a, 16b, and 16c show exemplary embodiments of the Active Layer Structure ALS 22500, which are exemplary cross-sections at location B-B of FIG. 15a or FIG. 15b. In ALS 22500, at least a bottom side conduction and Ohmic contact layer BSCOC 21300 is disposed somewhere above part of input supporting structure ISTR 21200. The layer BSCOC 21300 with thickness d.sub.BSC 21300d, refractive index n.sub.BSC 21300n, a total width w.sub.BSC 21300w serves to conduct electrical current and voltage from a contact region to an electro-active layer (ECL).

(176) Bottom Interspaced Material Layer

(177) There can be other bottom interspaced material layer BIM 21250 between layer BSCOC 21300 and ISTR 21200, with thickness d.sub.BIM 21250d, a total width w.sub.BIM 21250w, and refractive index n.sub.BIM 21250n. This layer may be electrically conducting or electrical insulating. The value of d.sub.BIM may take on zero thickness, in that case the bottom interspaced material layer BIM 21250 does not exist. The existence of a bottom interspaced material layer BIM 21250 is thus optional.

(178) Bottom Metal Contact Pads

(179) On top and to the left side of the bottom side conduction and Ohmic contact layer BSCOC 21300 is deposed of at least a first bottom left metal contact pad FBLM 21900L with thickness d.sub.FBLM 21900Ld, width w.sub.FBLM 21900Lw, and length g.sub.FBLM 21900Lg.

(180) On top and to the right side of the bottom side conduction and Ohmic contact layer BSCOC 21300 is deposed of at least a first bottom right metal contact pad FBRM 21900R with thickness d.sub.FBRM 21900Rd, width w.sub.FBRM 21900Rw, and length g.sub.FBRM 21900Rw. In an exemplary embodiment, only either the first bottom left or the first bottom right metal contact pad is present. In another exemplary embodiment, plurality of such bottom metal contact pads is present. The exact location of these metal contact pads can be in many other locations beside the left or right location shown as long as the metal contact pads are in electrical contact with the bottom side conduction and Ohmic contact layer BSCOC 21300.

(181) Bottom Metal Electrodes

(182) On top of the first bottom left metal contact pad FBLM 21900L is a first bottom left metal electrode FBLME 21120L. On top of the first bottom right metal contact pad FBRM 21900R is a first bottom right metal electrode FBRME 21120R. In an exemplary embodiment, only either the first bottom left or the first bottom right metal electrode is present. In another exemplary embodiment, plurality of such bottom metal electrodes is present. The exact location of these bottom metal electrodes can be in many other locations beside the left or right location shown as long as the bottom metal electrodes are in electrical contact with the respective bottom metal contact pads.

(183) Bottom Interspaced Dielectric Current Conduction Layer

(184) On top of the center region of the layer BSCOC 21300 (i.e. region above or near supporting structure 21200) is deposed of a bottom interspaced dielectric current conduction layer BIDC 21350 with thickness d.sub.BIDC 21350d, layer width w.sub.BIDC 21350w, and an averaged refractive index n.sub.BIDC 21350n. The layer width w.sub.BIDC is the dimension of that layer in a horizontal direction perpendicular to the direction of the optical beam propagation. This layer is optional in that when thickness d.sub.BIDC 21350d is zero, this layer does not exist.

(185) Bottom Vertical Current Conduction Layer

(186) On top of the bottom interspaced dielectric current conduction layer BIDC 21350 is deposed of a bottom vertical current conduction layer BVC 21400 with thickness d.sub.BVC 21400d, layer width w.sub.BVC 21400w, and an averaged refractive index n.sub.BVC 21400n. The layer width w.sub.BVC is the dimension of that layer in a horizontal direction perpendicular to the direction of the optical beam propagation.

(187) Electro-Active Layer

(188) On top of the bottom vertical current conduction layer BVC 21400 is deposed of an electro-active layer EC 21500 with thickness d.sub.EC 21500d, width w.sub.EC 21500w, an averaged refractive index of the entire layer given by n.sub.EC 21500n, and an averaged absorption coefficient of the entire layer given by .sub.EC 21500a. The refractive index averaging is given in a similar way as illustrated by Eq. (13). The refractive index n.sub.EC or the optical absorption coefficient .sub.EC (.sub.EC>0 means optical absorption and .sub.EC<0 means optical gain) describing the fraction of energy absorbed per unit beam propagation length of the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The guided optical beam in the electro-active layer BEC 21140 in this electro-active waveguiding core structure EWCoS 22600 has a propagating refractive index n.sub.BEC 21140n. While in a preferred embodiment described below, the EC layer is made of semiconductor, it can also be any other active material according to various embodiments of the current invention. For example, it can be ferroelectric electro-optic material (e.g. LiNbO.sub.3 or BaTiO.sub.3) or organic electro-optic or electro-absorption material, whose refractive index or optical intensity gain and absorption coefficient .sub.EC (.sub.EC<0 means optical gain and .sub.EC>0 means optical absorption) can be altered under an applied electric field or electric current or optical excitation beam as is well known to those skilled in the art.

(189) In the case of an electro-optic modulator, a small averaged increment or decrement in the averaged refractive index of the electro-active layer EC 21500 is denoted as dn.sub.EC 21500dn so that its new average refractive index becomes n.sub.EC(new)=n.sub.EC+dn.sub.EC will cause a change in the propagating refractive index n.sub.BEC 21140n by dn.sub.BEC 21140dn from n.sub.BEC to n.sub.BEC(new)=n.sub.BEC+dn.sub.BEC due to the overlapping of the optical beam energy with the material regions in which dn is non-zero.

(190) In the case of an optical amplifier or an electro-absorption modulator, a small averaged increment or decrement in the averaged optical intensity absorption/gain coefficient of the electro-active layer EC 21500 is denoted as d.sub.EC 21500da so that its new average optical intensity loss/gain coefficient becomes .sub.EC(new)=.sub.EC+d.sub.EC will cause a change in the absorption/gain coefficient .sub.BEC 21140a of the optical beam by d.sub.BEC 21140da from .sub.BEC to .sub.BEC(new)=.sub.BEC+d.sub.BEC due to the overlapping of the optical beam energy with the material regions in which d is non-zero.

(191) Top Vertical Current Conduction Layer

(192) On top of the electro-active layer EC 21500 is deposed of a top vertical current conduction layer TVC 21600 with thickness d.sub.TVC 21600d, width W.sub.TVC 21600w, and an averaged refractive index n.sub.TVC 21600n

(193) Top Interspaced Dielectric Current Conduction Layer

(194) On top of the top vertical current conduction layer TVC 21600 is deposed of a top interspaced dielectric conduction layer TIDC 21650 with thickness d.sub.TIDC 21650d, width w.sub.TIDC 21650w, and an averaged refractive index n.sub.TIDC 21650n. This layer is optional in that when thickness d.sub.TIDC 21650d is zero, this layer does not exist.

(195) Top Vertical/Side Conduction and Ohmic Contact Layer

(196) On top of the top interspaced dielectric conduction layer TIDC 21650 is deposed of a top vertical/side conduction and Ohmic contact layer TVSCOC 21700 with thickness d.sub.TVSC 21700d, width w.sub.TVSC 21700w, and an averaged refractive index n.sub.TVSC 21700n.

(197) Top Metal Contact Pads

(198) In one embodiment (FIG. 16a), on top of the top vertical/side conduction and Ohmic contact layer TVSCOC 21700 is deposed of a first top right metal contact pad FTRM 21800R with thickness d.sub.FTRM 21800Rd, width w.sub.FTRM 21800Rw, and length g.sub.FTRM 21800Rg. FTRM 21800R is typically to the right end on the top of TVSCOC 21700. In another embodiment (FIG. 16b), On top of the top vertical/side conduction and Ohmic contact layer TVSCOC 21700 is deposed of a first top left metal contact pad FTLM 21800L with thickness d.sub.FTLM 21800Ld, width w.sub.FTLM 21800Lw, and length g.sub.FTLM 21800Lg. FTLM 21800L is typically to the left end on the top of TVSCOC 21700. In as yet another embodiment (FIG. 16C), on top of the top vertical/side conduction and Ohmic contact layer TVSCOC 21700 is deposed of a first top middle metal contact pad FTMM 21800M with thickness d.sub.FTMM 21800Md, width w.sub.FTMM 21800Mw, and length g.sub.FTMM 21800Mg. FTMM 21800M is typically at the middle part on the top of TVSCOC 21700. Middle part is the portion closest to the beam BEC 21140. This case shown by FIG. 16C is particularly applicable when the top vertical/side conduction and Ohmic contact layer TVSCOC 21700 is an Ohmic Transparent Conductor (OTC) or Low-Refractive-Index Ohmic Transparent Conductor (LRI-OTC).

(199) In an exemplary embodiment, only either the first top left, first top middle, or the first top right metal contact pad is present. In another exemplary embodiment, plurality of such top metal contact pads are present. The exact location of these top metal contact pads can be in many other locations beside the left or right location shown as long as the top metal contact pads are in electrical contact with the top vertical/side conduction and Ohmic contact layer TVSCOC 21700.

(200) Top Metal Electrodes

(201) On top of the first top left metal contact pad FTLM 21800L is a first top left metal electrode FTLME 21130L. On top of the first top middle metal contact pad FTMM 21800M is a first top middle metal electrode FTMME 21130M. On top of the first top right metal contact pad FTRM 21800R is a first top right metal electrode FTRME 21130R. In an exemplary embodiment, only either the first top left, first top middle, or the first top right metal electrode are present. In another exemplary embodiment, plurality of such top metal electrodes is present. The exact locations of these top metal electrodes can be in many other locations beside the left or right location shown as long as the top metal electrodes are in electrical contact with the respective top metal contact pads.

(202) Electro-Active Waveguiding Core Structure and Central Waveguide Core Layer

(203) A layer or several layers that are in spatial proximity to the electro-active layer EC 21500 form an electro-active waveguiding core structure EWCoS 22600 at least a portion of it contains a central waveguide core layer WCo 22600Co. For the purpose of illustration and not limitation, a particular exemplary embodiment of an electro-active waveguiding core structure EWCoS 22600 is formed by the bottom vertical current conduction layer BVC 21400, the top vertical current conduction layer TVC 21600, and the electro-active layer EC 21500 as shown by FIG. 17. However, this electro-active waveguiding core structure can also be formed by any other layer or collection of layers as long as it is in spatial proximity to the electro-active layer so that the optical beam guided by it will have a reasonable amount of optical energy in the electro-active layer. The central waveguide core layer WCo 22600Co has an averaged refractive index n.sub.WCo 22600Con higher than the refractive index of most of the materials surrounding it. As is known to those skilled in the art, a waveguide core only needs its refractive index to be generally higher than the refractive indices of most of its surrounding materials in order to confine and guide optical beam (e.g. one example of such waveguides is commonly known as rib waveguide or ridge waveguide). For illustration and not limitation, in an exemplary embodiment, the central waveguide core layer WCo 22600Co is the electro-active layer EC 21500. As is known to those skilled in the art, the central waveguide core layer WCo 22600Co can also be in part of layer 21400 or layer 21600.

(204) The central waveguide core layer WCo 22600Co has an averaged refractive index n.sub.WCo 22600Con higher than the refractive indices of most its surrounding and confines optical beam energy of beam BEC 21140, called the beam electro-active or beam EC, in the vertical and horizontal directions so that the peak of the beam intensity is within or near the central waveguide core layer 22600Co, and the optical beam is said to be a guided optical beam. The guided optical beam BEC 21140 in this electro-active waveguiding core structure EWCoS 22600 has a propagating refractive index n.sub.BEC 21140n that is smaller than the material refractive index of the central waveguide core layer n.sub.WCo 22600Con so that n.sub.BEC<n.sub.WCo. This criterion can be taken as the definition of the material region that made up the waveguide core (i.e. it is the region in which the material refractive index is higher than the beam propagating refractive index n.sub.BEC).

(205) Electro-Active Waveguide Core and Cladding Regions for Beam EC

(206) As is known to those skilled in the art, the entire electro-active waveguide core region for beam EC is the material region occupied by the beam EC, BEC 21140, in which the refractive index of the material is generally higher than n.sub.BEC 21140n. As is also known to those skilled in the art, the electro-active waveguide cladding regions for beam EC are the material regions occupied by the beam in which the refractive index of the material is generally lower than n.sub.BEC. For the purpose of discussion, one may take the electro-active waveguiding core structure EWCoS 22600 mentioned above as defined by this electro-optic waveguide core region.

(207) Thus, the electro-active layer 22500 may be or may not be part of the waveguide core region for beam guided in the EC-layer BEC 21140 as long as the electro-active layer 22500 is in spatial proximity to the waveguide core region for EC-layer beam BEC 21140 so that a reasonable amount of the beam's optical energy is in the electro-active layer. Even if the electro-active layer 22500 is part of the waveguide core region, it is not necessarily the entire waveguide core region for EC-layer beam BEC 21140.

(208) Beam Transport to Electro-Active Waveguide Core Structure

(209) Most of the input optical beam energy of input beam IBM 22140 is transported from input tapering waveguide core ITWCo 22300 to the electro-active waveguide core structure EWCoS 22600, through the input tapering waveguide region between z1=z1ALS 22300z1ALS and z1=g.sub.ITWCo 22300g, where the tapering waveguide core width w.sub.ITWCo-z1 22300w-z1 varies to a value of w.sub.ITWCo-g at z1=g.sub.ITWCo 22300g from its value at z1=z1ALS 22300z1ALS (it can be the same value as, smaller than, or larger than its value at z1=z1ALS 22300z1ALS). In a preferred embodiment, for the purpose of illustration and not limitation, this is enabled by reducing the tapering waveguide core width from a value approximately equal to or larger than half the optical wavelength in the waveguide core given by .sub.bm/(2*n.sub.ITWCo), to well below half the optical wavelength in the waveguide core given by .sub.bm/(2*n.sub.ITWCo) so that W.sub.ITWCo-g<<.sub.bm/(2*n.sub.ITWCo), where * is numerical multiplication. More exactly, it is reduced from a width that is a width that enables the optical energy to be well confined in the waveguide core ITWCo 22300 just before it enters the ALS 22500 to a width (after it enter the ALS 22500) such that the optical energy is no longer well confined in the waveguide core ITWCo 22300 after it enters the ALS 22500 (the width for no longer well-confined is defined by the beam confinement after the taper waveguide core enters ALS 22500). Well confined means over 50% of the beam energy is in the waveguide core ITWCo 22300. Depending on the application situation, this can mean a smaller width (e.g. if the refractive index of the EC layer is approximately equal to or lower than the refractive index of the input tapering waveguide). It can also maintain the same width or even go to a larger width (e.g. if the refractive index of the EC layer is higher than the refractive index of the input tapering waveguide).

(210) After the energy is transported to electro-active waveguide core structure EWCoS 22600 that contains the electro-active layer EC 21500, the optical beam is denoted as optical beam in the electro-active region or EC layer beam, BEC 21140.

(211) Output Connecting Waveguide

(212) Output connecting waveguide core OCWCo 28200. The output connecting waveguide core OCWCo 28200 is fabricated on an output connecting-waveguide bottom cladding material OCWBCd 28200B disposed on a substrate SUB 21100 (FIG. 18).

(213) In one exemplary embodiment shown in FIG. 19, illustrating an exemplary embodiment of the cross section at location C-C of FIG. 18, to the top of the output connecting waveguide core is further deposed with output connecting waveguide top cladding material OCWTCd 28200T (FIG. 19). Bottom refers to direction closer to the substrate and top refers to direction away from the substrate. To the left of the input connecting waveguide core is deposed with output connecting waveguide left cladding material OCWLCd 28200L and to the right of the waveguide core is disposed with output connecting waveguide right cladding material OCWRCd 28200R. Right or left is taking the direction of beam propagation as the front direction and right or left means relative to this front direction. The division of the cladding into four different material regions as illustrated by FIG. 19 is for the purpose of discussion and not limitation as these can be all the same materials or there can be more than 4 material regions forming plurality of different material regions as long as these regions act as waveguide cladding materials with effective refractive index generally smaller than the effective refractive index of the waveguide core OCWCo 28200 so that the light beam power is confined near the region of the waveguide core OCWCo 28200, as is well known to those skilled in the art. The cladding material is a general designation that can include air or vacuum or transparent dielectric material as the material.

(214) Output Connecting Waveguide Region

(215) The output connecting waveguide core OCWCo 28200 is made up of a material or mixture of materials with an averaged material refractive index n.sub.OCWCo 28200n, has a thickness d.sub.OCWCo 28200d, and width W.sub.OCWCo 28200w. Let the refractive index of the bottom input connecting-waveguide bottom cladding material be n.sub.OCWBCd 28200Bn. Let the refractive index of the top cladding material OCWTCd 28200T be n.sub.OCWTCd 28200Tn, the refractive index of the left cladding material OCWLCd 28200L be n.sub.OCWLCd 28200Ln, and the refractive index of the right cladding material OCWRCd 28200R be n.sub.OCWRCd 28200Rn. The waveguide core 28200 and the claddings 28200T, 28200B, 28200R, 28200L, together forms output connecting waveguide OCWG 28200WG.

(216) The vertical confinement of the optical beam, called output optical beam OBM 28140, in the output connecting waveguide is due to the refractive-index difference between the top and bottom waveguide claddings and the waveguide core and the claddings generally have lower refractive indices than that of the waveguide core so that n.sub.OCWTCd<n.sub.OCWCo and n.sub.OCWBCd<n.sub.OCWCo. The horizontal confinement of the optical beam is due to the refractive-index difference between the left and right waveguide claddings and the waveguide claddings generally have lower refractive indices than that of the waveguide core so that n.sub.OCWRCd<n.sub.OCWCo and n.sub.OCWLCd<n.sub.OCWCo. The vertical direction is the direction perpendicular to the substrate plane and the horizontal direction is the direction parallel to the substrate plane. The output beam has a propagating refractive index given by n.sub.OBM 28140n.

(217) An important quantity in terms of waveguide mode confinement is the refractive index contrast between the averaged refractive index of the waveguide core and its immediate surrounding cladding materials called the refractive-index difference n.sub.Rd defined by n.sub.Rd.sup.2=n.sub.Co.sup.2n.sub.Cd.sup.2), where n.sub.Co is the refractive index of the waveguide core (e.g. n.sub.Co=n.sub.OCWCo) and n.sub.Cd is the refractive index of the waveguide cladding (e.g. n.sub.Cd=n.sub.OCWBCd or n.sub.OCWTCd or n.sub.OCWRCd or n.sub.OCWLCd) or an averaged of them thereof given by n.sub.aOCWCd 28200aCdn. The refractive-index averaging is more accurately done as averaged of its square values which are their dielectric constant =n.sup.2, as illustrated by Eq. (13).

(218) Likewise the waveguide core can also be made up of one or plurality of materials, and n.sub.OCWCo=n.sub.aOCWCo can also be an averaged refractive index of the m materials with slightly different refractive indices n.sub.OCWCo1, n.sub.OCWCo2, n.sub.OCWCo3 . . . n.sub.OCWCo m, that made up of the waveguide core materials.

(219) Output Beam Coupler Structure (OBCS) Region

(220) FIG. 20 shows the output beam coupler structure (OBCS) comprises at least a tapering waveguide section (preferably tapering from wide to narrow but can also maintain the same width or taper from narrow to wide as will be elaborated below) connected to the input waveguide. Optionally, the active layer structure ALS on top of the output tapering waveguide section can also be tapering in the form of an up taper (preferably tapering from narrow to wide in the direction toward the active layer structure ALS, but can also maintain the same width or taper from wide to narrow). See for example FIG. 20b.

(221) Specifically, the energy of the electro-active beam BEC 21140 in the electro-active waveguiding core structure EWCoS 22600 is coupled efficiently to the output optical beam IBM 28140 energy in the output connecting waveguide core OCWCo 28200 via propagating through an output connecting-waveguide taper region. The output connecting-waveguide taper region has a output tapering waveguide core OTWCo 28300. The output tapering waveguide core at a location z2 is denoted by OTWCo-z2 28300z2 (FIG. 19a), at which the width of the tapering waveguide core is denoted by w.sub.OTWCo-z2 28300w-z2. The distance or location parameter z2 is given by the distance measured from the beginning point of the taper that is at a point along the output connecting waveguide core OCWCo 28200, typically outside (or away from) the ALS region. It is changed from its input value at z2=0 w.sub.OTWCo-z2=0 28300w-z2=0 of w.sub.OTWCo-z2=0=w.sub.OCWCo 28200 to another width (that can be the same width) at z2>0 w.sub.OTWCo-z2>0 28300w-z2>0. The thickness of the tapering waveguide core is d.sub.OTWCo-z1 28300d-z2 and its refractive index is n.sub.OTWCo-z1 28300n-z2. Typically, though not always, d.sub.OTWCo-z2 28300d-z2 and n.sub.OTWCo-z2 28300n-z2 are constant value in z2 so we can drop the z2 designation with d.sub.OTWCo-z1=d.sub.OTWCo-z2 28300d and n.sub.OTWCo-z2=n.sub.OTWCo 28300n. In a preferred embodiment, d.sub.OTWCo-z2 is about the same value as d.sub.OCWCo. The tapering may be such that w.sub.OTWCo-z2 is a linear function of z2 or quadratic function of z2 (i.e. depending on z2.sup.2), but can also be of any curvilinear function of z2. Let g.sub.OTWCo denote the total length of this tapering waveguide.

(222) The end of the taper at z2=g.sub.OTWCo 28300g, typically inside (or toward) the ALS region at which the width of the waveguide core is w.sub.OTWCo-g 28300w-g, is connected to output supporting structure OSTR 29200 that may be a continuation from and in some way physically connected to the input supporting structure ISTR 21200 or may be independent of it. While illustrated as a line that is continuation of the connecting waveguide material with a narrow width and air or other low refractive index materials surrounding its side, the output supporting structure can be random dots or any shape of small amount of any materials that have a low effective refractive index or small averaged refractive index (e.g. as defined by Eq. (13)) as is known to those skilled in the art, comparing to the refractive index of the waveguide core n.sub.WCo 22600Con in the electro-active waveguiding core structure EWCoS 22600, resulting in an effective averaged refractive index n.sub.aOSTR 29200na for this entire layer of supporting structure. The output supporting structure OSTR 29200 may continue to guide wave or just acts as a supporting structure, depending on application scenarios.

(223) In an exemplary embodiment, the output supporting structure OSTR 29200 is a narrow line. In that particular case, we can describe it as having a width w.sub.OSTR 29200w, thickness d.sub.OSTR 29200d, and length g.sub.OSTR 29200g. The length g.sub.OSTR 29200g may be zero. In that case, output supporting structure OSTR 29200 does not exist (the thin ALS film can still be supported by its corners or sides, but not directly below). In a preferred embodiment, d.sub.OSTR is about the same value as d.sub.OCWCo. At some point the output supporting structure OSTR 29200 merges with the input supporting structure ISTR 21200 and thus ISTR 21200 and OSTR 29200 may be used interchangeably.

(224) In region outside the ALS region, the vertical confinement of the optical beam along the taper is due to the refractive-index difference between the tapering waveguide core and the top and bottom tapering waveguide claddings at the location z defined above: OTWTCd-z2 28300T-z2 (refractive index n.sub.OTWTCd-z1 28300Tn-z2) and OTWBCd-z1 28300B-z2 (refractive index n.sub.OTWBCd-z1 28300Bn-z2) and the waveguide core and the claddings generally have lower refractive indices than that of the waveguide core so that the refractive index n.sub.OTWTCd-z2<n.sub.OTWCo-z2 and n.sub.OTWBCd-z2<n.sub.OTWCo-z2. The horizontal confinement of the optical beam is due to the refractive-index difference between the left and right waveguide claddings at z2: OTWLCd-z2 28300L-z2 (refractive index n.sub.OTWLCd-z2 28300Ln-z2) and OTWRCd-z2 28300R-z2 (refractive index n.sub.OTWRCd-z2 28300Rn-z2), and the waveguide claddings generally have lower refractive indices than that of the waveguide core so that n.sub.OCWRCd-z2<n.sub.OCWCo-z2 and n.sub.OCWLCd-z2<n.sub.OCWCo-z2. The vertical direction is the direction perpendicular to the substrate plane and the horizontal direction is the direction parallel to the substrate plane. Again, there can be one or a plurality of cladding material regions and the four cladding material regions are mentioned for the purpose of illustration and not limitation.

(225) In an exemplary embodiment, n.sub.OTWTCd-z2=n.sub.OTWBCd-z2=n.sub.OTWLCd-z2=n.sub.OTWRCd-z2=n.sub.OCWTCd, and n.sub.OCWTCd=n.sub.OCWBCd=n.sub.OCWLCd=n.sub.OCWRCd so all the cladding indices in the tapering regions and the input connecting waveguide regions are all approximately equal. For example, these cladding regions can be filled with silicon dioxide materials with refractive index of n1.45. The refractive index of the waveguide core n.sub.OTWCo-z2 28300n-z2 can be silicon so that n.sub.ITWCo=n.sub.OCWCo3.6, where n.sub.OCWCo 28200n is the refractive index of the waveguide core for the input connecting waveguide.

(226) On top of the tapering waveguide core OTWCo 28300 starting at z2=z2ALS 28300z2ALS, is laid with an active layer structure ALS 22500. Typically z2ALS is before g.sub.OTWCo so that 0<z2ALS<g.sub.OTWCo. An active layer is a material layer that can give optical gain or optical absorption or change in the refractive index. The various embodiments of this active layer structure ALS 22500 have already been described above.

(227) Active Layer Structure-Beam Transport from the Structure to Output

(228) Most of the output optical beam energy of beam OBM 28140 is transported to output tapering waveguide core OTWCo 28300 from the electro-active waveguiding core structure EWCoS 22600, through the output tapering waveguide region that typically lies inside the ALS region, between z2=g.sub.OTWCo 28300g and z2=z2ALS 28300z2ALS, where the tapering waveguide core width w.sub.OTWCo-z1 28300w-z2 varies from a value of w.sub.OTWCo-g at z2=g.sub.OTWCo 28300g to a value of w.sub.OTWCo-z2ALS at z2=z2ALS 28300z2ALS (it can be the same value as, smaller than, or larger than its value at z2=z2ALS 28300z2ALS). In a preferred embodiment, for the purpose of illustration and not limitation, this is enabled by changing the tapering waveguide core width at z2=z2ALS 28300z2ALS from a value approximately equal to or larger than half the optical wavelength in the waveguide core given by .sub.bm/(2*n.sub.OTWCo), to well below half the optical wavelength in the waveguide core at z2=g.sub.OTWCo 28300g given by .sub.bm/(2*n.sub.OTWCo) so that W.sub.OTWCo-g<<.sub.bm/(2*n.sub.OTWCo), where * is number multiplication. More exactly, it is increased from a narrow width (in a region inside the ALS 22500) for which the optical energy is not well confined in the waveguide core OTWCo 28300 (the width for not well-confined is defined by the beam confinement in the waveguide core OTWCo 28300) to a wider width that enables the optical energy to be well confined in the waveguide core OTWCo 28300 just around when it exits the ALS 22500 region.

(229) Well confined means over 50% of the beam energy is in the waveguide core OTWCo 28300. Depending on the application situation, this can mean a smaller width (e.g. if the refractive index of the EC layer is approximately equal to or larger than the refractive index of the input tapering waveguide). It can also maintain the same width or even go to a larger width (e.g. if the refractive index of the EC layer is higher than the refractive index of the input tapering waveguide).

(230) After the energy is transported from the electro-active waveguiding core structure EWCoS 22600 that contains the electro-active layer EC 21500 down to the output taper at z2=Z2ALS and further propagated to the taper starting location at z2=0 where the taper core width is w.sub.OTWCo-z2=0 28300w-z2=0 and w.sub.OTWCo-z2=0=w.sub.OCWCo 28200, the optical beam is denoted as output optical beam or beam OBM 28140. Note that at z2=0, the output tapering waveguide core OTWCo-z2 28300z2 is joined to output connecting waveguide core OCWCo 28200.

(231) Length of Active Layer Structure

(232) The active layer structure ALS runs a length from the input tapering waveguide core ITWCo 22300 at z1=z1ALS to the output tapering waveguide core OTWCo 28300 at z2=z2ALS. Along the ALS structure, the distance from z1=z1ALS is parameterized as coordinate z. Location z thus measures a specific location along the length of the active layer structure ALS 22500. The total length of ALS 22500 from z1=z1ALS to z2=z2ALS is called the structure length of the modulator SL.sub.mod 22550. Coordinate z ends at z2=z2ALS at which z=SL.sub.mod.

(233) Along z, the various widths and thicknesses of each of the layers in the ALS may vary and do not necessarily have to stay constant. As is known to those skilled in the art, such variation in widths and thicknesses will not affect the general performance of the modulator. In addition, there may be more or fewer layers in the ALS other than specified as long as the functionalities of those layers specified are equivalently performed by the additional or fewer layers. As is known to those skilled in the art, such variations will not affect the general performance of the modulator. Hence, the various ALS structural variations as described above are for the purpose of illustration and not limitation.

(234) Active Layer Structure-Electro Active Layer

(235) The active material ACM 21500M in EC layer 21500 can be any active material as is known to those skilled in the art in which an applied electric field will change its refractive index or optical absorption or optical gain in at least a portion of the material. The entire structure described above can be used with any active material in layer 21500. While we illustrate a particular semiconductor active material below, it is only one of the many possibilities, and is to illustrate a particular preferred embodiment of the intensity or phase modulator in the present invention. They are not meant to limit the scope of the invention.

(236) Semiconductor EC Material Layer

(237) As noted, the electro-active layer EC need not be made of semiconductor materials. As an exemplary embodiment, in the case for which the electro-active layer is made of semiconductor based material as shown in FIG. 21, the structure in the electro-active layer could comprise a PN junction 21500PN in which a first P-layer 21500LP.sub.1 with P-type carrier dopant or called P-dopant (i.e. the resulting carriers from the dopant atoms are holes) and dopant density given by P.sub.1, 21500P.sub.1 is vertically physically connected (vertical means in a direction perpendicular to the substrate plane: horizontal means in a direction parallel to the substrate plane) to a first N-layer 21500LN.sub.1 with N-type carrier dopant or called N-dopant (i.e. the resulting carriers from the dopant atoms are electrons) and dopant density given by N.sub.1 21500N.sub.1. Depending on the application situations, the electro-active layer may be the entire PN structure itself or may be part of the PN structure or may be just electrically connected to the PN structure.

(238) Alternatively, the electro-active layer structure could also comprise a PqN junction 21500PqN in which a first P-layer 21500LP.sub.1 with P-dopant and dopant density given by P.sub.1 21500P.sub.1 is connected to a middle q-layer with either N dopant, P dopant, or being intrinsic I (i.e. commonly means with very low dopant or no dopant or being an Intrinsic semiconductor material) labeled as 21500MLqm (e.g. it will be labeled as 21500MLI.sub.m if it is intrinsic (i.e. undoped or being an intrinsic semiconductor material), 21500MLN.sub.m if it is N doped, and 21500MLP.sub.m if it is P-doped; m is an integer to sub-label the layer number and dopant density given by Mq.sub.m, 21500Mq.sub.m (e.g. it will be labeled as 21500MI.sub.1 if it is intrinsic I.sub.1, 21500MN.sub.1 if it is N.sub.1 doped, and 21500MP.sub.1 if it is P.sub.1-doped), and the middle q-layer is further connected to a first N-layer 21500LN.sub.1 with N.sub.1-dopant and dopant density given by N1 21500N.sub.1. This middle q-layer may be made up of plurality of one or more doped layers 21500MLq1, 21500MLq2, . . . 21500MLqT, where T is an integer specifying the number of layers. Depending on the application situations, the electro-active layer may be the entire PqN structure itself or may be part of the PN structure or may be just electrically connected to the PqN structure.

(239) Further alternatively, the electro-active layer structure could comprise a NqN junction 21500NqN in which a first N-layer 21500LN.sub.1 with N-dopant and dopant density given by N.sub.1 21500N.sub.1 is connected to a middle q-layer with either N dopant, P dopant, or being intrinsic I (i.e. commonly means with very low dopant or no dopant or being an intrinsic semiconductor material) labeled as 21500MLqm (e.g. it will be labeled as 21500MLI.sub.m if it is intrinsic (i.e. undoped or being an Intrinsic semiconductor material), 21500MLN.sub.m if it is N doped, and 21500MLP.sub.m if it is P-doped: m is an integer to sub-label the layer number) and dopant density given by Mq.sub.m, 21500Mq.sub.m (e.g. it will be labeled as 21500MI.sub.1 if it is intrinsic I.sub.1, 21500MN.sub.1 if it is N.sub.1 doped, and 21500MP.sub.1 if it is P.sub.1-doped), and the middle q-layer is further connected to a second N-layer 21500LN.sub.2 with N.sub.2-dopant and dopant density given by N.sub.2 21500N.sub.2. This middle q-layer may be made up of plurality of one or more doped layers 21500MLq1, 21500MLq2, . . . 21500MLqT, where T is an integer specifying the number of layers. Depending on the application situations, the electro-active layer may be the entire NqN structure itself or may be part of the NqN structure or may be just electrically connected to the NqN structure.

(240) Further alternatively, the electro-active layer structure could comprise a XqY junction 21500NqN in which a first X-layer 21500LX.sub.1 is connected to a middle q-layer with either N dopant, P dopant, or being intrinsic I (i.e. commonly means with very low dopant or no dopant or being an intrinsic semiconductor material) labeled as 21500MLqm (e.g. it will be labeled as 21500MLI.sub.m if it is intrinsic (i.e. undoped or being an Intrinsic semiconductor material), 21500MLN.sub.m if it is N doped, and 21500MLP.sub.m if it is P-doped; m is an integer to sub-label the layer number) and dopant density given by Mq.sub.m, 21500Mq.sub.m (e.g. it will be labeled as 21500MI.sub.1 if it is intrinsic I.sub.1, 21500MN.sub.1 if it is N.sub.1 doped, and 21500MP.sub.1 if it is P.sub.1-doped), and the middle q-layer is further connected to a second Y-layer 21500LY.sub.1. This middle q-layer may be made up of plurality of one or more doped layers 21500MLq1, 21500MLq2, . . . 21500MLqT, where T is an integer specifying the number of layers. Depending on the application situations, the electro-active layer may be the entire XqY structure itself or may be part of the XqY structure or may be just electrically connected to the XqY structure. In the above, X.sub.1 and Y.sub.1, each may either be N-doped, P-doped, or being an intrinsic I semiconductor, and X.sub.1 and Y.sub.1 can be doped differently with different dopant type.

(241) The P and N dopants may have spatially varying profiles in terms of their doping density (number of dopant carriers per unit volume) and the profiles may vary from one application to another. While there are various mode of operation for the active material, a commonly used mode is to apply reverse bias voltage V.sub.R 21500VR across the abovementioned PN or PqN layers (with negative voltage on the P side and positive voltage on the N side), so that an electric field E.sub.EC is generated to go across part of the EC layer 21500.

(242) Depending on the application, for the abovementioned PN or PqN or NqN or XqY structure, the N.sub.1 doped layer may be above or below the P.sub.1-doped layer (above means further away from the substrate and below means closer to the substrate). The EC layer 21500 may have quantum wells in the structure, typically in the q layer or close to the PN junction. At least one of the first P-layer, first N-layer, or the middle q-layer contains at least one quantum well. One or more quantum wells can also be in both the first P-layer and first N-layer or in all the three layers: first P-layer, first N-layer, and middle q-layer or just in the middle q-layer. The quantum wells can be strained, unstrained, or double-well or multiple-well quantum wells as is known to those skilled in the art. It can also have no quantum well.

(243) As an exemplary illustration, without quantum wells and without carrier doping in the q layer, in the case of EO modulation, the main electro-optic phase shift will be due mainly to linear electro-optic (LEO) effect. If q has carrier doping (N or P) then it will add plasma (PL) and bandfilling (BF) effect. If q layer has quantum wells, then quantum-confined stark effect (QCSE) will be added to enhance the EO phase shift. If the PqN is forward bias, then a lot of carriers will be injected into the q layer, causing refractive index change due to carrier injections or depletions. This may give significant phase shifts in the electro active (EC) material layer just due to PL and BF effects. However such modulator will be slow as removing the carrier is a slow process, typically at nano second speed or slower (e.g. <1 GHz). In order to go to high modulation frequency (e.g. >1 GHz), typically revised biased is applied. In that case the electric field in the q layer will cause carrier depletion which will also give rise to PL or BF effects, and the electric field will cause LEO effect (with or without quantum wells) and QCSE also (with use of quantum wells). The PL, BF, QCSE can also cause the absorption coefficient .sub.EC to change (.sub.EC>0 gives optical absorption) resulting in electro-absorption modulation, depending on the operating wavelength. As is known to those skilled in the art, for electro-absorption modulation, the operating optical wavelength is typically at relatively close to the band edge (edge of the material or quantum-well bandgap). For electro-optic modulation, the operating optical wavelength is typically at relatively far from the band edge (edge of the material or quantum-well bandgap). With carrier injection, it may also give population inversion resulting in optical gain (.sub.EC<0 gives optical gain) and hence giving rise to optical gain induced optical intensity or phase modulation. These effects due to change in .sub.EC will result in electro-absorption or gain based modulators, instead of electro-optic modulators, and are a particular embodiment of the present invention.

(244) The averaged incremental change in the refractive index dn.sub.EC 21500dn or change in the optical intensity loss/gain coefficient d.sub.EC 21500da of at least part of the material in the semiconductor electro-active layer 21500 can be caused by an applied electric field E.sub.EC 21500E, an electric current C.sub.EC 21500C, or either injection or depletion of carriers in the electro-active layer 21500 (note dn.sub.EC 21500dn is not the same as n.sub.EC 21500n, which is the averaged refractive index of the entire EC layer 21500 when there is no field).

(245) Voltage and Current Conduction to the Electro-Active Layer

(246) As shown in FIG. 21, the semiconductor electro-active layer EC 21500 is electrically connected to the top vertical current conduction layer TVC 21600 on the top side and is electrically connected to the bottom vertical current conduction layer BVC 21400 on the bottom side. Bottom side is the side nearer to the substrate 21100. Two materials are electrically connected if an electric current can be passed between the two materials with a total electrical resistance times area lower than about 10 -cm.sup.2 so that for 10,000 m.sup.2 area, the total contact resistance is less than about 100,000=100 k (10,000 m.sup.2 area is the area for a relatively large 5 mm-long, 2 m-wide modulator device).

(247) The Use of Forward-Biased PN Junction or Tunnel PN Junction to Reduce P-Dopant Optical Loss

(248) As an illustration but not limitation, for the EC 21500 layer, if its N layer is below the P layer, then the bottom vertical current conduction layer BVC 21400 can be (not always) an N-doped semiconductor material and the top vertical current conduction layer TVC 21600 can be (not always) a P-doped semiconductor material to enable easy current conduction without significant voltage dropped. The problem is that P-doped material has much higher (typically 10 times) electrical resistance and optical absorption than that of N-doped material at the same dopant density.

(249) As will be noted below, this can be addressed as with use of a forward-bias PN junction, it is possible to make electrical connection to region of opposite dopant type without significant voltage dropped and that could have certain advantages. We will refer to this as a PN-changing PN junction (labeled as PNCPN junction) to distinguish it from the PN junction inside EC layer 21500. Such PNCPN junction will conduct current or voltage when it is forward biased. Note that if such a PN-changing PN junction has highly doped P and N layers, it can also be conducting electricity even when it is under a reverse bias. In that case we called it a tunnel junction as the current conduction is depending of some type of carrier tunneling across the reverse-biased junction, as is well known to those skilled in the art. Then in that case, it can be used for when the PN junction inside EC layer 21500 PN junction is either reverse or forward biased. Thus, when we call it PN-changing PN junction, it will be generally referred to when the PN layer involved is either forward bias and conducting current or when it is highly doped and reverse biased but acts as a current-conducting tunneling junction.

(250) For example, as shown in FIG. 22a, suppose in the abovementioned PN or PqN or XqY structure, the N doped layer is below the P-doped layer, then in that case the top vertical current conduction layer TVC 21600 can also be an N.sub.2-doped (subscript 2 is just to label doping in this layer) semiconductor material in contact with the top layer of EC layer 21500 that is P.sub.1 doped (called layer 21500LP.sub.1), resulting in a PNCPN junction between layer 21600 and 21500 labeled as 21600-21500PNCPN.

(251) Alternatively, as shown in FIG. 22b, layer TVC 21600 may start with a P-doped layer (called layer 21600LP.sub.2) with a dopant density of P.sub.2, 21600P.sub.2, connecting to the top P-doped layer of layer EC 21500 with dopant density P.sub.1, and the P-doped layer 21600LP.sub.2 is also connected to another layer that is N doped (called layer 21600LN.sub.2) with dopant density N.sub.2, 21600N.sub.2, resulting in a PNCPN junction in layer 21600 labeled as 21600PNCPN. Then an applied voltage with negative voltage applied to the N.sub.2-doped layer TVC 21600 will give a forward bias across the N.sub.2P.sub.2 junction and transmit the voltage to the top P.sub.1 doped layer of EC layer 21500 that forms part of the P.sub.1q N.sub.1 structure, resulting in reverse bias across the P.sub.1q N.sub.1 junction in EC layer 21500. In that case, the N.sub.2P.sub.2 layers act as a PN-changing PN junction. Likewise if the P.sub.1q N.sub.1 junction on EC 21500 layer has P.sub.1 at the bottom connecting to an N.sub.2-doped layer BVC 21400, then one has a PNCPN junction between 21400 and 21500 labeled as 21500-21400PNCPN junction forming the structure N.sub.2P.sub.1q N.sub.1 (see FIG. 22a); or one can have a PNCPN junction inside 21400 labeled as 21400PNCPN junction forming the structure. N.sub.1 P.sub.2P.sub.1q N.sub.1 (see FIG. 22b).

(252) The reason to effectively change the P-doped to N-doped layer via such P-N changing PN junction is because an N-doped layer typically can be doped to have a much lower electrical resistance than P-doped semiconductor material for two reasons: (1) the dopant density for N dopant typically can be higher than that of P dopant; (2) even at the same dopant density, the electrical conductivity of N doped material can typically be higher than that of P doped material by about 10 times. Note that, as is also well known to those skilled in the art, N-doped semiconductor material also typically has a much lower optical absorption than P-doped semiconductor material even if the N-type material is doped to the same electrical resistance as a P-type material (typically can be about 10 times lower in optical absorption).

(253) This enables the use of highly N-doped layer with low electrical resistance for layer 21300, 21350, and 21400 from the bottom half up and 21700, 21650, and 21600 from the top half down, thereby substantially lowering the series electrical resistance of the modulator structure. Low series electrical resistance will give high modulation frequency.

(254) The example above is for the purpose of illustration and not limitation. For example, the PqN or PN junction in the electro-active layer EC 21500 may have P-doped side at the top, instead of the bottom, and a PN-changing PN junction is used so that the top layers can become N-doped materials. There are thus various variations in the use of the PN-changing PN junction as shall be obvious to those skilled in the art.

(255) Top Vertical/Side Conduction and Ohmic Contact Layer

(256) The top vertical/side conduction and Ohmic contact layer TVSCOC 21700 with thickness d.sub.TVSC 21700d and width w.sub.TVSC 21700w is also electrically connected at its bottom to the top vertical current conduction layer TVC 21600 through the top interspaced dielectric conduction layer TIDC 21650, and at its top to the top left/middle/right metal contact pad FT(L/M/R)M 21800(L/M/R) or any top metal contact pad FTXM 21800X (X refers to any of the plurality of top metal contact pads).

(257) Upper and Lower Waveguide Claddings of Active-Layer Structure

(258) In one embodiment, the top vertical/side conduction and Ohmic contact layer TVSCOC 21700, with an averaged refractive index n.sub.TVSC 21700n, forms part of an top electro-active waveguide cladding 22600TCd for which n.sub.TVSC 21700n is smaller than the refractive index n.sub.WCo 22600Con of the central waveguide core 22600Co. In one exemplary embodiment, TVSCOC 21700 is a low-refractive-index Ohmic transparent conductor (LRI-OTC) (see illustration in FIG. 16)

(259) The top interspaced dielectric conduction layer TIDC 21650 with an averaged refractive index n.sub.TIDC 21650n, in another exemplary embodiment also forms part of a top electro-active waveguide cladding 22600TCd for which n.sub.TIDC 21650n is smaller than the refractive index n.sub.WCo 22600Con of the central waveguide core 22600Co.

(260) In as yet another embodiment, the top electro-active waveguide cladding is formed by an air or dielectric region (e.g. the dielectric region TDMR 21810 in FIG. 24 described later below) above layer TVSCOC 21700.

(261) In one embodiment, part of a bottom electro-active waveguide cladding 22600BCd, in the case where the width w.sub.ISTR of the input support structure ISTR 21200 is narrow, may be made up of the input support structure ISTR 21200 below the bottom vertical current conduction layer BVC 21400 plus the cladding materials to its left, and right as follows: The input support structure ISTR 21200 is made up of a material or mixture of materials with a material refractive index n.sub.ISTR 21200n.

(262) In a preferred embodiment, for the purpose of illustration and not limitation, typically the input connecting-waveguide core ICWCo 22200, the input tapering waveguide core ITWCo 22300, and the input support structure ISTR 21200 all have a similar bottom, left and right cladding materials, though they can also have different cladding materials. For the input supporting structure ISTR 21200, let the refractive index of the left cladding material ISTRLCd 21200L be n.sub.ISTRLCd 21200Ln, and the refractive index of the right cladding material ISTRRCd 21200R be n.sub.ISTRRCd 21200Rn. The supporting structure 21200, the left claddings 21200L, and right cladding 21200R, together forms a material region with an effective layer averaged refractive index n.sub.laISTR 21200nla that is a weighted average of n.sub.ISTR 21200n, n.sub.ISTRLCd 21200Ln, and n.sub.ISTRRCd 21200Rn, similar to the computation of averaged refractive index given by equation Eq. (13). The weighting for the averaging is depending on the distribution of the beam energy for guided beam BEC 21140 inside these material regions similar to that given by Eq. (13). The layer averaged material refractive index n.sub.laISTR 21200nla experienced by the guided beam BEC 21140 in regions 21200, 21200L, 21200R, is typically smaller than the refractive index n.sub.WCo 22600Con of the central waveguide core 22600Co. In that case, they form part of the bottom electro-active waveguide cladding 22600BCd.

(263) However, in the case where the width W.sub.ISTR of the input supporting structure ISTR 21200 is relatively wide, part of a bottom electro-active waveguide cladding 22600BCd will be made up mainly of the bottom cladding ISTRBCd 21200B below the input support structure ISTR 21200 with an averaged refractive index n.sub.ISTRBCd 21200Bn, which as part of the embodiment would be filled with materials with n.sub.ISTRBCd smaller than the refractive index n.sub.WCo 22600Con of the central waveguide core 22600Co. In that case, substantial optical energy can be in the input supporting structure STR 21200, and the input supporting structure STR 21200 may become part of the waveguide core for beam BEC 21140.

(264) In another embodiment, part of a bottom electro-active waveguide cladding 22600BCd may also be made up of the bottom interspaced material layer BIM 21250 (if it exists) with refractive index n.sub.BIM 21250n. In one exemplary embodiment, BIM 21250 is a low-refractive-index Ohmic transparent conductor (LRI-OTC).

(265) In as yet another embodiment, part of a bottom electro-active waveguide cladding 22600BCd may also be made up of the bottom interspaced dielectric current conduction layer BIDC 21350 with refractive index n.sub.BIDC 21350n.

(266) In as yet another embodiment, part of a bottom electro-active waveguide cladding 22600BCd may also be made up of the bottom side conduction and Ohmic contact layer BSCOC 21300 with refractive index n.sub.BSCOC 21300n.

(267) In as yet another embodiment, part of a bottom electro-active waveguide cladding 22600BCd may also be made up of the bottom side conduction and Ohmic contact layer ISTRBCd 21200B with refractive index n.sub.ISTRBCd 21200Bn.

(268) As to which layer shall be considered as the waveguide cladding is that in the waveguide cladding material region (a material region that surround the waveguide core), the energy density of the guided mode shall decay largely exponentially in a direction away from the waveguide core, as is known to those skilled in the art. This waveguide cladding region may be made of a single layer or spot (i.e. small cluster) of material or a collection of multiple layers or spots (i.e. small clusters) of connected materials (including air as a material). A spot is a three-dimensional cluster of material volume. The waveguide cladding refractive index n.sub.Cd (e.g. as use in the in the next section) shall be taken as the averaged refractive index of this collection of multiple layers/spots of cladding materials that can have one layer/spot or plurality of layer/spots. The waveguide core shall be taken as the material region close to the center energy portion of the optical beam in which the material refractive index n.sub.MAT is larger than or equal to the beam propagating refractive index n.sub.BEC and the waveguide core refractive index n.sub.Co (e.g. as use in the in the next section) shall be taken as the averaged refractive index of the entire core material region (which again can be composed of layers or spots of materials). For the purpose of illustration and not limitation, it is useful to divide the cladding regions to be the top waveguide cladding situated above the waveguide core, the bottom waveguide cladding situated below the waveguide core, the left waveguide cladding situated to the left of the waveguide core, and the right waveguide cladding situated to the right of the waveguide core.

(269) High Refractive Index Contrast and Mode-Medium Overlap

(270) For the purpose of definition, it is useful to define a refractive index contrast parameter as described below. If a waveguide core refractive index is n.sub.Co and the waveguide cladding (as defined by the exponential energy decay above) immediately adjacent to the waveguide core has a refractive index n.sub.Cd, then we can define a waveguide core-to-cladding refractive index difference square to be n.sub.rd.sup.2=(n.sub.Co.sup.2n.sub.Cd.sup.2) and a refractive index contrast ratio to be: R.sub.cts=n.sub.rd.sup.2/(n.sub.Co.sup.2+n.sub.Cd.sup.2). For the purpose of definition and not limitation, we define very-strongly wave guiding regime to be when R.sub.cts>0.5 or R.sub.cts=0.5. It is also useful to define the medium-strongly wave guiding regime to be when 0.5>R.sub.cts>0.2 or R.sub.cts=0.2, weakly guiding regime to be when 0.2>R.sub.cts>0.02 or R.sub.cts=0.02 and the very-weakly guiding regime to be when 0.02>R.sub.cts.

(271) In a preferred embodiment, the electro-active waveguiding core structure EWCoS 22600 is in the very-strongly guiding or medium-strongly guiding regime at least in the vertical direction (direction perpendicular to the substrate) in which the refractive index contrast of the waveguide core layer with the top and bottom cladding immediately adjacent to the electro-active waveguide core given by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) is larger than or equal to about 0.2 or is larger than or equal to about 0.5, where n.sub.Cd is the refractive index of the top or bottom cladding. In the case of waveguiding core structure EWCoS 22600, n.sub.Co=n.sub.WCo where n.sub.WCo 22600Con is the averaged refractive index of the central waveguide core layer WCo 22600Co, and n.sub.Cd is either n.sub.BIM, n.sub.BIDC, n.sub.laISTR or n.sub.ISTRBCd depending on which one(s) is(are) the bottom cladding(s).

(272) This strong waveguiding in the vertical direction will enable much higher mode confinement that will push higher fraction of the beam energy into the electro-active layer. As a result, the phase shift in the guided beam BEC 21140 given by a change in n.sub.BEC 21140n under an applied voltage will be larger. This will result in substantially lower switching voltage. In order to increase this mode-medium overlapping factor, it is useful to reduce the total thickness of the electro-active waveguide core.

(273) More precisely, it is useful to define the thickness d.sub.CORE of the electro-active waveguide core as the distance between a first top boundary and a first bottom boundary. The first top boundary is the boundary between the waveguide core and the top cladding immediately adjacent to the waveguide core, and the first bottom boundary is the boundary between the waveguide core and the bottom cladding immediately adjacent to the waveguide core. If d.sub.CORE is smaller than (.sub.bm/(2*n.sub.Co)), the waveguide core is said to be in ultra-thin regime. If d.sub.CORE is smaller than or equal to (.sub.bm/n.sub.Co) and larger (.sub.bm/(2*n.sub.Co)), then the waveguide core is said to be in very-thin regime. If d.sub.CORE is smaller than or equal to (1.5*.sub.bm/n.sub.Co) and larger than (.sub.bm/n.sub.Co), the waveguide core is said to be in medium-thin regime. If d.sub.CORE is smaller than (3*.sub.bm/n.sub.Co) and larger than (1.5*.sub.bm/n.sub.Co), the waveguide core is said to be in the thin regime. If d.sub.CORE is larger than (3*.sub.bm/n.sub.Co), the waveguide core is said to be in the thick regime.

(274) In a preferred embodiment for the modulator of the present invention, in order to achieve additional enhanced performances such as very low modulation voltage, the electro-active waveguiding core structure EWCoS 22600 shall be in the very-strongly guiding regime, and d.sub.CORE shall either be in the ultra-thin regime or very-thin regime.

(275) In a preferred embodiment for the modulator of the present invention, in order to achieve additional enhanced performances such as low modulation voltage, the electro-active waveguiding core structure EWCoS 22600 shall be in the medium-strongly guiding or very-strongly guiding regime, and d.sub.CORE shall either be in the ultra-thin regime, very-thin regime, or medium-thin regime.

(276) In another preferred embodiment for the modulator of the present invention, the electro-active waveguiding core structure EWCoS 22600 shall be in the weakly guiding regime, and d.sub.CORE shall either be in the ultra-thin, very-thin, medium-thin, or thin regime.

(277) For example, if .sub.bm=1550 nm, n.sub.EC=3.0, then if d.sub.CORE is smaller than or equal to (.sub.bm/n.sub.Co)=517 nm, it is in the very-thin regime, and if n.sub.Cd=1.5, it also has R.sub.cts>0.5 and hence is in the very-strongly wave guiding regime as well, which will satisfy the co-requirements. Both requirements have to be satisfied in order to draw an exemplary benefit of the present invention such as to enhance the low voltage performance of the modulator.

(278) It is useful to define the electro-active field overlapping factor more precisely. Let the electric field distribution of the guided mode of optical beam BEC 21140 be given by E.sub.OPT(x,y) 21140E, where E.sub.OPT is the electric field strength, and x and y are the coordinates in the cross-sectional area of the beam. The mode m is typically the fundamental guided mode with a single intensity peak at the beam center region. Let n 21140dn be the change in the optical propagating refractive index experienced by the beam under an applied electric field E.sub.EC(x,y) 21500E (for the case of constant field E.sub.EC(x,y) V.sub.EC/D.sub.EC, where V.sub.EC 21500VEC is the applied voltage and D.sub.EC 21500DEC is the effective physical distance for which the voltage V.sub.EC is applied across) that again has a value profile depending on the x-y cross-sectional coordinates. For the case of refractive index modulation, then the phase shift 21140Ph experienced by an optical beam propagating through the modulator under an applied voltage V(zo,t) 21500Vzo, where zo is the propagating distance along the modulation, is given by:

(279) $=\left(2/{}_0\right)n,n={}_0^{L}V\left(\mathrm{zo},t\right)d\mathrm{zo},=\frac{1}{V_{\mathrm{EO}}}\frac{\left(1/2\right)n_{\mathrm{eff}}^3{r}_{\mathrm{EO}}\left(x,y\right)E_{\mathrm{EO}}\left(x,y\right){.Math.E_{\mathrm{OPT}}\left(x,y\right).Math.}^{2}dxdy}{{.Math.E_{\mathrm{OPT}}\left(x,y\right).Math.}^{2}dxdy},$
where V.sub.EC is the applied RF voltage that gives rise to E.sub.EC(x,y), r.sub.EO(x,y), 21500rEO is an effective electro-optic coefficient describing how much the material's refractive index is changed under an applied field, n.sub.eff is the effective refractive index of the propagating optical mode (same as n.sub.BEC). The quantity 21500ROMOF is thus called the RF-field, optical mode, and active medium overlapping factor (also simply called the mode-medium overlapping factor). It is independent on V.sub.EC as E.sub.EC(x,y) is proportional to V.sub.EC. The voltage V(z,t) is the actual applied voltage that may change with propagation distance z and time t.

(280) The voltage to the entire modulator V.sub.MOD is approximately given by V.sub.EC assuming the voltage drop between the electro-active layers and the top or bottom electrode is small compared to V.sub.EC.

(281) While the above mode-medium overlapping factor is illustrated for the case of an electro-optic (EO) modulator, there are various other definitions of mode-medium overlapping factor more suitable for other applications such as for the case of electro-absorption (EA) modulator or the case involving optical absorption and gain medium, as is well known skilled in the art. These other definitions of mode-medium overlapping factor shall be used when appropriate and the specific definition of mode-medium overlapping factor is not meant to limit the present invention.

(282) Low-Refractive-Index Ohmic-Transparent-Conductor & Metal Contact Case

(283) The difficulty of obtaining high mode-medium overlapping or tight mode confinement is that the optical mode will inevitably touch the metal contact pad if the top vertical/side conduction and Ohmic contact layer TVSCOC is the usual doped semiconductor. One way to solve this problem in the present embodiment is to utilize Low-Refractive-Index Ohmic Transparent Conductor (LRI-OTC) that is electrically conductive but having a low refractive index (the low-refractive-index criterion is defined in the section on High-refractive-index-contrast mode confining structure in electro-active region). In that case, the mode can be strongly confined in the electro-active waveguide core structure EWCS 22600 region and will rapidly decay in the top electro-active waveguide cladding 22600UC. Layer 22600UC is basically the top vertical/side conduction and Ohmic contact layer TVSCOC 21700. This would be possible only if n.sub.TVSC is small compared to the refractive index n.sub.WCo 22600Con of the central waveguide core layer WCo 22600Co. However, it is important that layer TVSCOC 21700, now made of transparent conductor, shall have Ohmic-like contact with the top vertical conduction layer TVC 21600 that may be semiconductor.

(284) As an embodiment, layer TVSCOC 21700 is a low-refractive-index transparent conductor with Ohmic contact capability with layer TVC 21600. In that case, the material for layer TVSCOC 21700 will be called Low-Refractive-Index Ohmic Transparent Conductor (LRI-OTC). Ohmic transparent conductor differs from just transparent conductor as they have to have low electrical contact resistance with the next conduction layer in contact with it to pass current down to the electro-active layer without causing high voltage across the contact surface. For the purpose of illustration and not limitation, the next conduction layer is typically N-doped or P-doped semiconductor. Preferably, it shall also have low low electrical contact resistance with appropriate metal electrode. Materials for LRI-OTC include but are not limited to transparent conducting oxide (TCO) materials such as In.sub.2O.sub.3 (or various Indium Oxides), ZnO (Zinc Oxides), ITO (Indium Tin Oxides), GITO (Gallium Indium Tin Oxides), Gallium Indium Oxide (GIO), ZITO (Zinc Indium Tin Oxides), CdO (Cadmium Oxides), or materials containing any one or more than one of these oxides.

(285) In as yet another embodiment, layer TVSCOC 21700 can also be a low-refractive-index transparent conductor. In that case, the material for layer TVSCOC 21700 will be called Low-Refractive-Index Transparent Conductor (LRI-TC). Transparent conductor (TC) differs from Ohmic transparent conductor (OTC) as they do not need to have low electrical contact resistance with the next conduction layer in contact with it. For example, one can have modulators in which layer TVSCOC 21700 directly applies the electric field to the active material without further conducting the voltage down to the next layer. In that case, the other layers such as layer TVC 21600 may be undoped or an intrinsic semiconductor that does not conduct electric voltage or current.

(286) In another embodiment, in order to achieve high frequency response, it is also desirable that the Ohmic contact between layer TVC 21600 and TVSCOC 21700 has low Ohmic contact resistance. Ohmic-like contact means the relation between the voltage and the current is largely linear. For the purpose of illustration and not limitation, low Ohmic contact resistance between any two materials A and B generally means the voltage-over-current ratio for a current going between material A and material B is not substantially worse than the total sum of other electrical resistances that will affect the frequency response of the modulator.

(287) In one exemplary embodiment, on top of layer TVSCOC 21700 with LRI-OTC material is metal pad that gives good metal Ohmic contact with the LRI-OTC material used. In an embodiment, the metal pad is the first top middle metal contact pad FTMM 21800M with thickness d.sub.FTMM 21800Md, width w.sub.FTMM 21800Mw, and length g.sub.FTMM 21800Mg.

(288) This case is referred to as top LRI-OTC-metal contact case.

(289) Side-Conduction and Metal Contact Case

(290) Other alternative contacts include having layer TVSCOC 21700 to extend side way and have metal Ohmic contact on the side away from the center region of layer TVSCOC 21700 as shown in FIG. 23. In that case, Ohmic contact can be achieved from the metal to layer TVSCOC 21700 and the optical mode will not touch the metal, which can keep the optical mode to be low loss. This is referred to as top side-conduction-metal contact case.

(291) Still another alternative structure involved having the top lateral conduction geometry with metal contact but also a top lowloss dielectric material region TDMR 21810 as shown in FIG. 24. This dielectric material can be chosen to have low refractive index and hence acting as a top electro-active waveguide cladding. Alternatively, the metal can even go on top of TDMR 21810 to make this top lateral conduction structure mechanically robust as shown in FIG. 25. Thus, there are various ways to realize what we refer to as the top lateral conduction geometry with metal contact.

(292) Still another alternative structure involved having the bottom LRI-OTC in that the bottom interspaced material layer BIM 21250 between bottom side-conduction and Ohmic-contact layer BSCOC 21300, and ISTR 21200, with thickness d.sub.BIM 21250d is made of LRI-OTC. This enables a thicker layer for conducting current and voltage from the bottom electrode(s) with electrode(s) either at the bottom of the LRI-TOC layer or on top of the LRI-TOC layer (e.g. using top via hole on one or both sides of the ALS to contact the LRI-TOC layer through layer 21300) or directly on top of the bottom side-conduction and Ohmic contact layer BSCOC 21300.

(293) Lateral Optical Mode Confinement

(294) Note that in another embodiment, the width w.sub.TVSC 21700w of layer 21700 can act on the optical mode of guided beam BEC 21140 laterally so as to confine the optical mode. Such lateral mode confinement is called rib waveguide structure, which is known to have low optical loss. Thus, in an embodiment, layer 21700 also forms a rib-waveguide structure.

(295) It must be understood that there are various ways to confine the optical mode laterally, including a small lateral width w.sub.EC 21500w for a small vertical portion of the electro-active layer 21500, which can also confine the optical mode laterally, called lateral mode confinement structure. Thus, in another embodiment, the electro-active layer is a lateral mode confinement structure.

(296) Similarly, a small lateral width w.sub.SVC 21600w or w.sub.FVC 21400w for a small vertical portion of layer 21600 or 21400 can also confine the optical mode laterally. Thus in another embodiment, the bottom vertical conduction layer BVC 21400 is a lateral mode confinement structure. In as yet another embodiment, the top vertical conduction layer TVC 21600 is a lateral mode confinement structure.

(297) In as yet another embodiment, the bottom interspaced dielectric current conduction layer BIDC 21350 with thickness d.sub.BIDC 21350d, layer width w.sub.BIDC 21350w, and an averaged refractive index n.sub.BIDC 21350n is a lateral mode confinement structure.

(298) In as yet another embodiment, the bottom interspaced dielectric current conduction layer BIM 21250 with thickness d.sub.BIM 21250d, layer width was 21250w, and an averaged refractive index n.sub.BIM 21250n is a lateral mode confinement structure.

(299) In as yet another embodiment, the input supporting structure ISTR 21200 with thickness d.sub.ISTR 21200d, layer width w.sub.ISTR 21200w, and a refractive index n.sub.ISTR 21200n is a lateral mode confinement structure.

(300) In as yet another embodiment, the output supporting structure OSTR 29200 with thickness d.sub.OSTR 29200d, layer width w.sub.OSTR 29200w, and a refractive index n.sub.OSTR 29200n is a lateral mode confinement structure.

(301) In as yet another embodiment, the top lowloss dielectric material region TDMR 21810 with thickness d.sub.TDMR 21810d, layer width w.sub.TDMR 21810w, and an averaged refractive index n.sub.TDMR 21810n is a lateral mode confinement structure.

(302) In as yet another embodiment, the top interspaced dielectric conduction layer TIDC 21650 with thickness d.sub.TIDC 21650d, layer width w.sub.TUDC 21650w, and an averaged refractive index n.sub.TIDC 21650n is a lateral mode confinement structure.

(303) Reducing Modulator Junction Capacitance

(304) In a preferred embodiment, the small lateral width w.sub.EC 21500W for a small vertical portion of the electro-active layer EC 21500 acts as a lateral mode confinement structure, but at the same time also reduces the capacitance between the top and bottom electric-field applying junction in the modulator structure. This is because capacitance is proportional to the plate area and the lateral width w.sub.EC 21500W will define the effective capacitance plate area across the PN (or PqN) layer in layer 21500, with P side serving as one capacitance plate and N side serving as another capacitance plate, spaced by the carrier depletion width between the P and N doped material regions, as is known to those skilled in the art. Reducing the modulator junction capacitance can increase the modulator frequency response. This will be referred to as capacitance reduction via EC-layer width reduction. This can also be applied to either layer 21400 or layer 21600 if a PN junction responsible for part of the total device capacitance is in layer 21400 or 21600. In that case, the width of either layer 21400 or 21600 or both shall be carefully chosen to reduce the total device capacitance.

(305) Separate Lateral Mode Confinement and Modulator Junction Capacitance Reduction

(306) Note that it is possible to implement this capacitance reduction via EC-layer width reduction and still use the narrowed width of other layers to confine the optical mode laterally if the narrowed width of other layers is comparable to or smaller than this EC-layer width. This may have certain advantage by having low capacitance but also by using the top vertical/side conduction and Ohmic contact layer TVSCOC 21700** (or any other layer between this layer and the layer ISTR 21200 including layer ISTR 21200 itself, except the EC-layer) to confine the mode laterally. This case is shown in FIG. 26. (**such as when layer 21700 is an Ohmic Transparent Conductor case; for side metal contact case, it will be difficult to use layer 21700 to confine the optical beam mode. In that case, one can alternatively use the top interspaced dielectric conduction layer TIDC 21650 to confine the optical mode laterally)

(307) The Use of Highly Doped Quantum Wells for Lower Modulation Voltage

(308) This structure enables the modulation voltage to be drastically reduced using high carrier doping. While both N and P doping can be used, for the purpose of illustration and not limitation, the preferred embodiment is the use of N-doping in the active electro-active region as P doping will cause higher optical absorption loss than N doping at the same dopant density. The higher the doping density, the smaller the carrier depletion width at the PN junction and the larger the PN-junction capacitance. For a conventional modulator, the doping density is limited to N=10.sup.17/cm.sup.3 as otherwise the high junction capacitance will begin to severely limit the frequency bandwidth of the modulator. In the applications below, the doping is made into the EC layer that may or may not have quantum wells present. The presence of quantum wells may enhance the refractive index change due to change in carrier band-filling in the quantum wells under an applied voltage. However, the absence of quantum wells will also work in that refractive index will also be changed due to change in carrier band-filling under an applied voltage. Thus, when quantum wells are mentioned, it is for the purpose of illustration and not limitation. The presence of quantum wells also enables refractive index change due to quantum confined Stark effects as noted above, which can further increase the change in the refractive index under an applied voltage. The quantum wells can be strained, unstrained, double-well, or multi-well quantum wells as is known to those skilled in the art.

(309) In the present invention, in one application area, the EC layer has region with high-level doped carrier density with P-type or N-type doping and a doping density at or higher than 210.sup.17/cm.sup.3 and lower than 510.sup.17/cm.sup.3 primarily but not exclusively for low modulation voltage V.sub.MOD 20000V or low modulation RF power P.sub.MOD 20000P, and low-loss high-frequency modulator applications. In an exemplary embodiment, for illustration and not limitation, low means V.sub.MOD<2 Volt or P.sub.MOD<80 mW. Typically V.sub.MOD and P.sub.MOD are approximately related by the R.sub.LOAD 20000R transmission line resistance or load resistance: P.sub.MOD=V.sub.MOD.sup.2/R.sub.LOAD. In an exemplary embodiment, R.sub.LOAD=50 Ohms. This is referred to as having the quantum wells in highly-doped regime.

(310) In another application area, the EC layer has region with medium-high-level doped carrier density with P-type or N-type doping and a doping density at or higher than 510.sup.17/cm.sup.3 and lower than 1.510.sup.18/cm.sup.3 primarily but not exclusively for medium-low modulation voltage V.sub.MOD or medium-low modulation RF power P.sub.MOD, and low-loss high-frequency modulator applications. In an exemplary embodiment, for illustration and not limitation, medium-low means V.sub.MOD<1 Volt or P.sub.MOD<20 mW. This is referred to as having the quantum wells in medium-highly-doped regime.

(311) In as yet another application area, the EC layer has region with very-high-level doped carrier density with P-type or N-type doping and a doping density at or higher than 1.510.sup.18/cm.sup.3 and lower than 510.sup.18/cm.sup.3 primarily but not exclusively for very-low modulation voltage V.sub.MOD or very-low modulation RF power P.sub.MOD, and low-loss high-frequency modulator applications. In an exemplary embodiment, for illustration and not limitation, very-low means V.sub.MOD<0.6 Volt or P.sub.RF<7 mW. This is referred to as having the quantum wells in very-highly-doped regime.

(312) In as yet another application area, the EC layer has region with ultra-high-level doped carrier density with P-type or N-type doping and a doping density at or higher than 510.sup.18/cm.sup.3 primarily but not exclusively for ultra-low modulation voltage V.sub.MOD or ultra-low modulation RF power P.sub.MOD, and low-loss high-frequency modulator applications. In an exemplary embodiment, for illustration and not limitation, ultra-low means V.sub.MOD<0.2 Volt or P.sub.MOD<0.8 mW. This is referred to as having the quantum wells in ultra-highly-doped regime.

(313) The quantum wells can be strained, unstrained, double-well quantum wells, or multi-well as is known to those skilled in the art.

(314) While the preferred embodiments and advantages of the invention have been illustrated and described, it will be clear that the invention is not limited to these embodiments and advantages only. Numerous modifications, changes, variations, substitutions and equivalents will be apparent to those skilled in the art without departing from the spirit and scope of the invention as described.

(315) To further illustrate the present invention, for the purpose of illustration and not limitation, we describe a few exemplary devices below.

(316) A First Exemplary Device of Electro-Optic Modulator with Side Conduction Geometry

(317) A preferred embodiment of an exemplary device is Modulator Device 20000 with the following specifications referred to as a first exemplary device of electro-optic modulator with side-conduction geometry:

(318) Substrate SUB 21100 is silicon wafer substrate with a thickness of about 0.3 mm. Input connecting waveguide core ICWCo 22200 is made of silicon for which its averaged material refractive index n.sub.ICWCo 22200n is around n.sub.ICWCo=3.6, thickness d.sub.ICWCo 22200d is d.sub.ICWCo=250 nm, and width W.sub.ICWCo 22200w is W.sub.ICWCo=400 nm.

(319) Input connecting-waveguide bottom cladding material ICWBCd 22200B is silicon dioxide (SiO.sub.2), for which its refractive Index n.sub.ICWBCd 22200Bn is n.sub.ICWBCd=1.45.

(320) Input connecting waveguide top cladding material ICWTCd 22200T is silicon dioxide (SiO.sub.2) for which its refractive Index n.sub.ICWTCd 22200Tn is 1.45.

(321) Input connecting waveguide left cladding material ICWLCd 22200L is silicon dioxide (SiO.sub.2), for which its refractive Index n.sub.ICWLCd 22200Ln is 1.45

(322) Input connecting waveguide right cladding material ICWRCd 22200R is silicon dioxide (SiO.sub.2), for which its refractive Index n.sub.ICWRCd 22200Rn is 1.45

(323) The above form an input connecting waveguide ICWG 22200WG. The core-cladding refractive-index difference n.sub.Rd defined by n.sub.Rd.sup.2=(n.sub.Co.sup.2n.sub.Cd.sup.2) for waveguide ICWG 22200WG is n.sub.Rd.sup.2=(3.6.sup.21.45.sup.2)=10.86 with n.sub.Co=3.6 and n.sub.Cd=1.45. Its averaged Cladding Refractive Index is given by n.sub.aICWCd=(n.sub.ICWBCd.sup.2A.sub.ICWBCd+n.sub.ICWTCd.sup.2A.sub.ICWTCd+n.sub.ICWRCd.sup.2 A.sub.ICWRCd+n.sub.ICWLCd.sup.2 A.sub.ICWLCd)/(A.sub.ICWBCd+A.sub.ICWTCd+A.sub.ICWRCd+A.sub.ICWLCd).sup.0.5=1.45. Its averaged Core Refractive Index is given by n.sub.aCo=(n.sub.Co1.sup.2A.sub.Co1+n.sub.Co2.sup.2A.sub.Co2+n.sub.Co3.sup.2 A.sub.Co3+ . . . +n.sub.Com.sup.2 A.sub.Com)/(A.sub.Co1+A.sub.Co2+A.sub.Co3+ . . . +A.sub.Com).sup.0.5=3.6.

(324) The input optical beam IBM 22140 has propagating refractive index n.sub.IBM 22140n, for which n.sub.IBM is approximately 2.8 with optical power P.sub.bm 22140P approximately 1 mW, electric field polarization E.sub.bm 22140E to be in the horizontal direction parallel to the substrate surface. It has a beam effective area A.sub.bm 22140A of A.sub.bm=0.04 m.sup.2 and an optical wavelength centered at .sub.bm 22140L with .sub.bm=1550 nm with plurality of (N) frequency channels .sub.bm1=1548 nm, .sub.bm2=1549 nm, .sub.bm3=1550 nm, .sub.bm4=1551 nm, and .sub.bm3=1552 nm centered at .sub.bm=1550 nm.

(325) Input Beam Coupler Structure (IBCS) Region

(326) The input tapering waveguide core ITWCo 223000 is made of silicon. Its width at a location z1, ITWCo-z1 22300z1 is denoted as width w.sub.ITWCo-z1 22300w-z1. This width is tapered from width at z1=0 w.sub.ITWCo-z1=0 22300w-z1=0 that has a value of w.sub.ITWCo-z1=0=400 nanometers (nm) to a width at z1>0 w.sub.ITWCo-z1>0 22300w-z1>0 that is narrower than 400 nm in a linear fashion.

(327) The thickness of the tapering waveguide core d.sub.ITWCo-z1 22300d-z1 made of silicon is d.sub.ITWCo-z1=250 nm with a refractive index n.sub.ITWCo-z1 22300n-z that is n.sub.ITWCo-z1=3.6.

(328) The total length of tapering waveguide g.sub.ITWCo 22300g is g.sub.ITWCo=20 micrometers (m). The width of the waveguide core at the end of the tapering at z1=g.sub.ITWCo is w.sub.ITWCo-g 22300w-g with w.sub.ITWCo-g=50 nm.

(329) Input supporting structure ISTR 21200 has width w.sub.ISTR 21200w with w.sub.ISTR=50 nm and thickness d.sub.ISTR 21200d with d.sub.ISTR=250 nm and length g.sub.ISTR 21200g with g.sub.ISTR=20 micrometers. It has an effective layer averaged refractive index n.sub.laISTR 21200nla with n.sub.lnISTR<2.5.

(330) Left cladding material ISTRLCd 21200L is air and has a refractive index n.sub.ISTRLCd 21200Ln given by n.sub.ISTRLCd=1, and Right cladding material ISTRRCd 21200R is air and has a refractive index n.sub.ISTRRCd 21200Rn given by n.sub.ISTRRCd=1. Its bottom cladding ISTRBCd 21200B is silicon dioxide (this is part of a Burried-Oxide BOX layer in a typical Silicon-On-Insulator SOI wafer) with averaged refractive index n.sub.ISTRBCd 21200Bn of n.sub.ISTRBCd=1.45.

(331) The top cladding ITWTCd-z1 22300T-z1 before going into the ALS region is silicon dioxide (SiO.sub.2) has refractive index n.sub.ITWTCd-z1 22300Tn-z1 with n.sub.ITWTCd-z1=1.45.

(332) The bottom cladding ITWBCd-z1 22300B-z1 before going into the ALS region is silicon dioxide (SiO.sub.2) has refractive index n.sub.ITWTCd-z1 22300Bn-z1 with n.sub.ITWTCd-z1=1.45.

(333) The left cladding ITWLCd-z1 22300L-z1 before going into the ALS region is silicon dioxide (SiO.sub.2) has refractive index n.sub.ITWLCd-z1 22300Ln-z1 with n.sub.ITWLCd-z1=1.45.

(334) The right cladding ITWRCd-z1 22300R-z1 before going into the ALS region is silicon dioxide (SiO.sub.2) has refractive index n.sub.ITWRCd-z1 22300Rn-z1 with n.sub.ITWRCd-z1=1.45.

(335) In this exemplary embodiment, n.sub.ITWTCd-z1=n.sub.ITWBCd-z1=n.sub.ITWLCd-z1=n.sub.ITWRCd-z1=n.sub.ICWTCd, and n.sub.ICWTCd=n.sub.ICWBCd=n.sub.ICWLCd=n.sub.ICWRCd. Input tapering waveguide core ITWCo 22300 starting at z1=z1ALS 22300z1ALS, where z1ALS=10 micrometers, is laid with an active layer structure ALS 22500. 0<z1ALS<g.sub.ITWCo.

(336) Active Layer Structure-Beam Transport into the Structure

(337) The active layer structure ALS 22500 is shown by the Table 3-1 below:

(338) TABLE-US-00003 TABLE 3-1 ALS 22500 Layer Doping/ Layer Number Thickness NPNN TCO CASE (1/cm.sup.3) BIM 100 nm In.sub.2O.sub.3 (21250) BSCOC 1 100 nm InGaAsP 1.3 um N = 1 (21300) (Bottom layer-just 10.sup.19 above the substrate) BIDC 2 40 nm InP N = 1 (21350LN) 10.sup.19 BVC 3 20 nm InGaAsP 1.3 um N = 1 (21400) 10.sup.19 EC 4 10 nm AlGaInAs 1.3 um N.sub.1 = 1 (21500LN.sub.1) 10.sup.19 EC 5 4 nm barrier AlGaInAs/1.1 um/0.8% MN.sub.1 = 4 (21500MLN.sub.1) tensile strained 10.sup.17 EC 6 2 7 nm AlGaInAs/1.1 um/0.8% MN.sub.2 = 4 (21500MLN.sub.2) barrier inside tensile strained 10.sup.17 EC 7 3 6.5 nm AlGaInAs/1.55 um/0.9% MN.sub.3 = 4 (21500MLN.sub.3) Well (PL = compressive strained 10.sup.17 1350 nm) EC 8 4 nm barrier AlGaInAs/1.1 um/0.8% MN.sub.4 = 4 (21500MLN.sub.4) tensile strained 10.sup.17 EC 9 43 nm AlGaInAs 1.3 um MN.sub.5 = 4 (21500MLN.sub.5) 10.sup.17 EC 10 20 nm AlGaInAs 1.3 um P.sub.1 = 1 (21500LP.sub.1) 10.sup.18 TVC 11 25 nm InGaAsP 1.3 um P.sub.2 = 1 (21600P.sub.2) 10.sup.18 TVC 12 20 nm InGaAsP 1.3 um N.sub.2 = 1 (21600N.sub.2) 10.sup.19 TIDC 13 20 nm InP N = 1 (21650) 10.sup.19 TVSCOC 14 40 nm InGaAsP (Top layer) N = 1 (21700) 10.sup.19 Total 380 nm

(339) In the table, the materials are unstrained (with InP as the substrate) if not specified as strained. The wavelength specified will be the material bandgap wavelength of the quaternary material involved (proper choice of the material composition is needed to achieve the required material bandgap and strain when grown on InP substrate).

(340) Bottom Side Conduction and Ohmic Contact Layer

(341) The active layer structure ALS 22500 has a bottom side conduction and Ohmic contact layer BSCOC 21300 that is InGaAsP layer given by layer 1 in Table 3-1 with thickness d.sub.BSC 21300d, where d.sub.BSC=100 nm and width w.sub.BSC 21300w, where w.sub.BSC is approximately 54 micrometers along most of the length of the ALS. Its refractive index n.sub.BSC 21300n is n.sub.BSC=3.4.

(342) Bottom Interspaced Material Layer

(343) The bottom interspaced material layer BIM 21250 is made of a Low-Refractive-Index Ohmic Transparent Conducting (LRI-OTC) material composed of Indium oxide (In.sub.2O.sub.3) with thickness d.sub.BIM 21250d equals to d.sub.BIM=100 nm, width w.sub.BIM 21250w equals to w.sub.BIM=54 micrometers, and average refractive index n.sub.BIM 21250n equals to n.sub.BIM=1.7.

(344) Bottom Metal Contact Pads

(345) The first bottom left metal contact pad FBLM 21900L is a multi-layer metal made up of (17 nm Au followed by 17 nm Ge followed by 17 nm Au followed by 17 nm Ni followed by 1000 nm Au) deposited on top of the top surface of n-doped layer 21300 given by layer 1 in Table 3-1. The total thickness of the metal contact pad is d.sub.FBLM 21900Ld, with d.sub.FBLM=1068 nm, and width w.sub.FBLM 21900Lw, where w.sub.FBLM is approximately 20 micrometers. The length of the metal contact pad g.sub.FBLM 21900Lg is approximately 500 micrometers.

(346) The first bottom right metal contact pad FBRM 21900R is multilayer metal made up of (17 nm Au followed by 17 nm Ge followed by 17 nm Au followed by 17 nm Ni followed by 1000 nm Au) deposited on top of the top surface of n-doped layer 21300 given by layer 1 in Table 3-1. The total thickness of the metal contact pad is d.sub.FBRM 21900Rd, with d.sub.FBRM=1068 nm, and width w.sub.FBRM 21900Rw, where w.sub.FBRM is approximately 20 micrometers. The length of the metal contact pad g.sub.FBRM 21900Rg is approximately 500 micrometers.

(347) Bottom Metal Electrodes

(348) On top of the first bottom left metal contact pad FBLM 21900L is deposited the first bottom left metal electrode FBLME 21120L which is gold of thickness of approximately 2 micrometer thick.

(349) On top of the first bottom right metal contact pad FBRM 21900R is deposited the first bottom right metal electrode FBRME 21120R which is gold of thickness of approximately 2 micrometer thick.

(350) Bottom Interspaced Dielectric Current Conduction Layer

(351) Bottom interspaced dielectric current conduction layer BIDC 21350 is a n-doped InP given by layer 2 in Table 3-1 with thickness d.sub.BIDC 21350d equals to d.sub.BIDC=40 nm, width w.sub.BIDC 21350w equals to w.sub.BIDC=54 micrometers, and average refractive index n.sub.BIDC 21350n equals to about n.sub.BIDC=3.0.

(352) Bottom Vertical Current Conduction Layer

(353) Bottom vertical current conduction layer BVC 21400 is n-doped InGaAsP given by layer 3 in Table 3-1 with thickness d.sub.BVC 21400d equals to d.sub.BVC=20 nm, width w.sub.BVC 21400w equals to w.sub.BVC=2 micrometers, and an averaged refractive index n.sub.BVC 21400n equals to n.sub.BVC=3.4.

(354) Electro-Active Layer

(355) Electro-active layer EC 21500 is made up of layers 4, 5, 6, 7, 8, 9, 10 in Table 3-1 with an averaged refractive index of the entire layer given by n.sub.EC 21500n with n.sub.EC equals to approximately n.sub.EC=3.4. Under an applied electric field, there will be a change in averaged refractive index dn.sub.EC 21500dn. The average refractive index becomes n.sub.EC(new)=n.sub.EC+dn.sub.EC.

(356) The total thickness d.sub.EC 21500d of this Electro-active layer is d.sub.EC=114.5 nm. Its width w.sub.EC 21500w is equal to w.sub.EC=2 micrometers.

(357) The electro-active layer has a PqN junction at layer 4 to 10 for which layer 4 is layer 21500LN.sub.1 that is N-doped with a dopant density of 21500N.sub.1=110.sup.19/cm.sup.3 and layer 10 is layer 21500LP.sub.1 that is P-doped with a dopant density of 21500P.sub.1=110.sup.18/cm.sup.3

(358) The intermediate layers 21500MLN.sub.m are all N-doped with a dopant density of 21500MN.sub.m=410.sup.17/cm.sup.3.

(359) The applied field E.sub.EC 21500E (which may cause a current C.sub.EC 21500C to flow) is across the entire electro-active layer with a negative voltage applied to the top and positive voltage applied to the bottom of this entire electro-active layer known to those skilled in the art as revered bias (with respect to the PN junction in the electro-active layer) of voltage V.sub.R 21500VR so the applied electro-active V.sub.EC 21500VEC is V.sub.R.

(360) The voltage applied to the electrodes of the modulator V.sub.MOD 20000V is approximately given by V.sub.EC.

(361) Top Vertical Current Conduction Layer

(362) Top vertical current conduction layer TVC 21600 is given by layer 11 and 12 in Table 3-1 made up of InGaAsP layer that is composed of 25 nm-thick layer 21600LP.sub.2 that is P-doped with dopant density 21600P.sub.2=110.sup.18/cm.sup.3, followed by 20 nm-thick N-doped InGaAsP layer 21600LN.sub.2 with dopant density 21600N.sub.2=110.sup.19/cm.sup.3. The total thickness for TVC 21600 is d.sub.TVC 21600d with d.sub.TVC=45 nm. Its width is W.sub.TVC 21600w equals to W.sub.TVC=2 micrometers, and its averaged refractive index is n.sub.TVC 21600n equals to n.sub.TVC=3.4. This N.sub.2P.sub.2 junction forms a forward-Biased PN Junction (or Tunnel PN Junction). It forms a PN-changing PN junction (called PNCPN junction) 21600PNCPN.

(363) Top Interspaced Dielectric Current Conduction Layer

(364) Top interspaced dielectric conduction layer TIDC 21650 is N-doped InP layer given by layer 13 in Table 3-1 with thickness d.sub.TIDC 21650d equals to d.sub.TIDC=20 nm, width w.sub.TIDC 21650w equals to w.sub.TIDC=8 micrometers, and averaged refractive index n.sub.TIDC 21650n equals to n.sub.TIDC=3.0.

(365) Top Vertical/Side Conduction and Ohmic Contact Layer

(366) Top vertical/side conduction and Ohmic contact layer TVSCOC 21700 is made up of InGaAsP given by layer 14 in Table 3-1 with thickness d.sub.TVSC 21700d equals to d.sub.TVSC=40 nm, width w.sub.TVSC 21700w equals to w.sub.TVSC=8 micrometers, and an averaged refractive index n.sub.TVSC 21700n equals to n.sub.TVSC=3.4.

(367) Top Metal Contact Pads

(368) The first top left metal contact pad FTLM 21800L is multilayer metal made up of (17 nm Au followed by 17 nm Ge followed by 17 nm Au followed by 17 nm Ni followed by 1000 nm Au) deposited on top of the top surface of n-doped layer 21700 given by layer 14 in Table 3-1. The total thickness of the metal contact pad is d.sub.FTLM 21800Ld, with d.sub.FTLM=1068 nm, and width W.sub.FTLM 21800Lw, where w.sub.FTLM is approximately 3 micrometers. The length of the metal contact pad g.sub.FTLM 21800Lg is approximately 500 micrometers.

(369) The first top right metal contact pad FTRM 21800R is multilayer metal made up of (17 nm Au followed by 17 nm Ge followed by 17 nm Au followed by 17 nm Ni followed by 1000 nm Au) deposited on top of the top surface of n-doped layer 21700 given by layer 14 in Table 3-1. The total thickness of the metal contact pad is d.sub.FTRM 21800Rd, with d.sub.FTRM=1068 nm, and width w.sub.FTRM 21800Rw, where w.sub.FTRM is approximately 3 micrometers. The length of the metal contact pad g.sub.FTRM 21800Rg is approximately 500 micrometers.

(370) There is no top middle metal contact pad FTMM 21800M.

(371) Top Metal Electrodes

(372) On top of the first top left metal contact pad FTLM 21800L is deposited the first top left metal electrode FTLME 21130L which is gold of thickness of approximately 2 micrometer thick.

(373) On top of the first top right metal contact pad FTRM 21800R is deposited the first top right metal electrode FTRME 21130R which is gold of thickness of approximately 2 micrometer thick.

(374) Beam Transport to Electro-Active Waveguiding Core Structure

(375) Input tapering waveguide region between z1=z1ALS 22300z1ALS and z1=g.sub.ITWCo 22300g, Tapering waveguide core width w.sub.ITWCo-z 22300w varies down to a smaller value of w.sub.ITWCo-g=50 nm at z1=g.sub.ITWCo 22300g from its vale at z1=z1ALS 22300z1ALS. Clearly W.sub.ITWCo-g<<.sub.bm/(2*n.sub.ITWCo), with .sub.bm=1550 nm and n.sub.ITWCo=3.6, where * is number multiplication.

(376) Output Connecting Waveguide

(377) Output connecting waveguide core OCWCo 28200 has averaged Refractive Index n.sub.OCWCo=n.sub.aOCWCo=3.6, thickness d.sub.OCWCo 28200d is d.sub.OCWCo=250 nm, and width W.sub.OCWCo 28200w is W.sub.OCWCo=400 nm.

(378) Output connecting waveguide OCWG 28200WG has Output connecting-waveguide bottom cladding material OCWBCd 28200B that is silicon dioxide (SiO.sub.2) for which the refractive index n.sub.OCWBCd 28200Bn is n.sub.OCWBCd=1.45.

(379) Output connecting waveguide top cladding material OCWTCd 28200T is silicon dioxide for which the refractive index n.sub.OCWTCd 28200Tn is n.sub.OCWTCd=1.45.

(380) Output connecting waveguide left cladding material OCWLCd 28200L is silicon dioxide for which the refractive index n.sub.OCWLCd 28200Ln is n.sub.OCWLCd=1.45.

(381) Output connecting waveguide right cladding material OCWRCd 28200R is silicon dioxide for which the refractive index n.sub.OCWRCd 28200Rn is n.sub.OCWRCd=1.45.

(382) The resulted averaged cladding refractive Index n.sub.aOCWCd 28200aCdn is n.sub.aOCWCd=1.45.

(383) Output Optical Beam OBM 28140

(384) Output Beam Coupler Structure (OBCS) Region

(385) Output tapering waveguide core OTWCo 28300 is made of silicon. Its width at a location z2 OTWCo-z2 is denoted as width w.sub.OTWCo-z2 28300w-z2. This width is tapered from width at z2=0 w.sub.OTWCo-z2=0 28300w-z2=0 that has a value of w.sub.OTWCo-z2=0=400 nm to a width at z2>0 w.sub.OTWCo-z2>0 28300w-z2>0 that is narrower than 400 nm in a linear fashion.

(386) The thickness of the tapering waveguide core d.sub.OTWCo-z2 28300d-z2 made of silicon is d.sub.OTWCo-z2=250 nm with a refractive index n.sub.OTWCo-z2 28300n-z2 that is n.sub.OTWCo-z2=3.6.

(387) The total length of tapering waveguide g.sub.OTWCo 28300g is g.sub.OTWCo=20 micrometers (m). The width of the waveguide core at the end of the tapering at z2=g.sub.OTWCo is w.sub.OTWCo-g 28300w-g with w.sub.OTWCo-g=50 nm.

(388) Output supporting structure OSTR 29200 has width w.sub.OSTR 29200w with w.sub.OSTR=50 nm and thickness d.sub.OSTR 29200d with d.sub.OSTR=250 nm and length g.sub.OSTR 29200g with g.sub.OSTR=20 micrometers. It has an effective layer averaged refractive index n.sub.laOSTR 29200nla with n.sub.laOSTR<2.5.

(389) The top cladding OTWTCd-z2 28300T-z2 is silicon dioxide (SiO.sub.2) has refractive index n.sub.OTWTCd-z2 28300Tn-z2 with n.sub.OTWTCd-z2=1.45 before going into the ALS region.

(390) The bottom cladding OTWBCd-z2 28300B-z2 is silicon dioxide (SiO.sub.2) has refractive index n.sub.OTWBCd-z2 28300Bn-z2 with n.sub.OTWBCd-z2=1.45.

(391) The left cladding OTWLCd-z2 28300L-z2 is silicon dioxide (SiO.sub.2) has refractive index n.sub.OTWLCd-z2 28300Ln-z2 with n.sub.OTWLCd-z2=1.45.

(392) The right cladding OTWRCd-z2 28300R-z2 is silicon dioxide (SiO.sub.2) has refractive index n.sub.OTWRCd-z2 28300Rn-z2 with n.sub.OTWRCd-z2=1.45.

(393) In this exemplary embodiment, n.sub.OTWTCd-z2=n.sub.OTWBCd-z2=n.sub.OTWLCd-z2=n.sub.OTWRCd-z2=n.sub.OCWTCd, and n.sub.OCWTCd=n.sub.OCWBCd=n.sub.OCWLCd=n.sub.OCWRCd

(394) Output tapering waveguide core OTWCo 28300 starting at z2=z2ALS 28300z2ALS, is laid with an active layer structure ALS 22500. 0<z2ALS<g.sub.OTWCo.

(395) Most of the output optical beam energy of beam OBM 28140 is transported to output tapering waveguide core OTWCo 28300 from the electro-active waveguiding core structure EWCoS 22600, through the output tapering waveguide region between z2=z2ALS 28300z2ALS and z2=g.sub.OTWCo 28300g, where the output tapering waveguide core width w.sub.OTWCo-z2 28300w-z2 varies down to a smaller value of W.sub.OTWCo-g at z2=g.sub.ITWCo 28300g from its vale at z2=z2ALS, 28300z2ALS. The tapering waveguide core width is reduced to well below half the optical wavelength in the waveguide core given by .sub.bm/(2n.sub.OTWCo) so that w.sub.OTWCo-g<<.sub.bm/(2n.sub.OTWCo). After the energy transported from the electro-active waveguiding core structure EWCoS 22600 that contains the electro-active layer EC 21500 down to the output taper at z2=0 where the taper core width is w.sub.OTWCo-z2=0 28300w0 and w.sub.OTWCo-z2=0=w.sub.OCWCo 28200, the optical beam is denoted as output optical beam or beam OBM 28140.

(396) Length of Active Layer Structure

(397) The length of the active layer structure SL.sub.mod 22550 is approximately 500 micrometers.

(398) High Refractive Index Contrast and Mode Overlapping

(399) For the bottom cladding:

(400) Waveguide core refractive index is n.sub.co=3.6

(401) Waveguide bottom cladding is n.sub.BCd=1.45 (given by layer ISTRBC with n.sub.ISTRBCd=1.45)

(402) Waveguide core-to-cladding refractive index difference square to be n.sub.rd.sup.2=(n.sub.co.sup.2n.sub.BCd.sup.2)=10.86.

(403) Refractive index contrast ratio to be: R.sub.cts=n.sub.rd.sup.2/(n.sub.co.sup.2+n.sub.BCd.sup.2)=0.7, which is in the very-strongly guiding regime.

(404) For the top cladding:

(405) Waveguide core refractive index is n.sub.co=3.6

(406) Waveguide bottom cladding is n.sub.TCd=1 (given by material above TVSCOC layer which is air with n=1)

(407) Waveguide core-to-cladding refractive index difference square to be n.sub.rd.sup.2=(n.sub.co.sup.2n.sub.TCd.sup.2)=11.96.

(408) Refractive index contrast ratio to be: R.sub.cts=n.sub.rd.sup.2/(n.sub.co.sup.2+n.sub.TCd.sup.2)=0.86, which is in the very-strongly guiding regime.

(409) A Second Exemplary Device of Electro-Absorption Modulator with Transparent Conductor Geometry

(410) A preferred embodiment of an exemplary device is Modulator Device 20000 with the following specifications referred to as a second exemplary device of electro-absorption modulator with Ohmic transparent conductor geometry: The main difference between this and the First Exemplary Device is in Table 3-2, in which the active layer is designed for electro-absorption modulation. Also, there are no left and right top metal contact pads, only middle metal contact pad.

(411) Substrate SUB 21100 is silicon wafer substrate with a thickness of about 0.3 mm.

(412) Input connecting waveguide core ICWCo 22200 is made of silicon for which its averaged material refractive index n.sub.ICWCo 22200n is around n.sub.ICWCo=3.6, thickness d.sub.ICWCo 22200d is d.sub.ICWCo=250 nm, and width W.sub.ICWCo 22200w is W.sub.ICWCo=400 nm.

(413) Input connecting-waveguide bottom cladding material ICWBCd 22200B is silicon dioxide (SiO.sub.2), for which its refractive Index n.sub.ICWBCd 22200Bn is n.sub.ICWBCd=1.45.

(414) Input connecting waveguide top cladding material ICWTCd 22200T is silicon dioxide (SiO.sub.2) for which its refractive Index n.sub.ICWTCd 22200Tn is 1.45.

(415) Input connecting waveguide left cladding material ICWLCd 22200L is silicon dioxide (SiO.sub.2), for which its refractive Index n.sub.ICWLCd 22200Ln is 1.45

(416) Input connecting waveguide right cladding material ICWRCd 22200R is silicon dioxide (SiO.sub.2), for which its refractive Index n.sub.ICWRCd 22200Rn is 1.45

(417) The above form an input connecting waveguide ICWG 22200WG. The core-cladding refractive-index difference n.sub.Rd defined by n.sub.Rd.sup.2=(n.sub.Co.sup.2n.sub.Cd.sup.2) for waveguide ICWG 22200WG is n.sub.Rd.sup.2=(3.6.sup.21.45.sup.2)=10.86 with n.sub.Co=3.6 and n.sub.Cd=1.45. Its averaged Cladding Refractive Index is given by n.sub.aICWCd=(n.sub.ICWBCd.sup.2A.sub.ICWBCd+n.sub.ICWTCd.sup.2A.sub.ICWTCd+n.sub.ICWRCd.sup.2 A.sub.ICWRCd+n.sub.ICWLCd.sup.2 A.sub.ICWLCd)/(A.sub.ICWBCd+A.sub.ICWTCd+A.sub.ICWRCd+A.sub.ICWLCd).sup.0.5=1.45. Its averaged Core Refractive Index is given by n.sub.aCo=(n.sub.Co1.sup.2A.sub.Co1+n.sub.Co2.sup.2A.sub.Co2+n.sub.Co3.sup.2 A.sub.Co3+ . . . +n.sub.Com.sup.2 A.sub.Com)/(A.sub.Co1+A.sub.Co2+A.sub.Co3+ . . . +A.sub.Com).sup.0.5=3.6.

(418) The input optical beam IBM 22140 has propagating refractive index n.sub.IBM 22140n, for which n.sub.IBM is approximately 2.8 with optical power P.sub.bm 22140P approximately 1 mW, electric field polarization E.sub.bm 22140E to be in the horizontal direction parallel to the substrate surface. It has a beam effective area A.sub.bm 22140A of A.sub.bm=0.04 m.sup.2 and an optical wavelength centered at .sub.bm 22140L with .sub.bm=1550 nm with plurality of (N) frequency channels .sub.bm1=1548 nm, .sub.bm2=1549 nm, .sub.bm3=1550 nm, .sub.bm4=1551 nm, and .sub.bm3=1552 nm centered at .sub.bm=1550 nm.

(419) Input Beam Coupler Structure (IBCS) Region

(420) The input tapering waveguide core ITWCo 223000 is made of silicon. Its width at a location z1. ITWCo-z1 22300z1 is denoted as width w.sub.ITWCo-z1 22300w-z1. This width is tapered from width at z1=0 w.sub.ITWCo-z1=0 22300w-z1=0 that has a value of w.sub.ITWCo-z1=0=400 nanometers (nm) to a width at z1>0 w.sub.ITWCo-z1>0 22300w-z1>0 that is narrower than 400 nm in a linear fashion.

(421) The thickness of the tapering waveguide core d.sub.ITWCo-z1 22300d-z1 made of silicon is d.sub.ITWCo-z1=250 nm with a refractive index n.sub.ITWCo-z1 22300n-z that is n.sub.ITWCo-z1=3.6.

(422) The total length of tapering waveguide g.sub.ITWCo 22300g is g.sub.ITWCo=20 micrometers (m). The width of the waveguide core at the end of the tapering at z1=g.sub.ITWCo is w.sub.ITWCo-g 22300w-g with w.sub.ITWCo-g=50 nm.

(423) Input supporting structure ISTR 21200 has width w.sub.ISTR 21200w with w.sub.ISTR=50 nm and thickness d.sub.ISTR 21200d with d.sub.ISTR=250 nm and length g.sub.ISTR 21200g with g.sub.ISTR=20 micrometers. It has an effective layer averaged refractive index n.sub.laISTR 21200nla with n.sub.laISTR<2.5.

(424) Left cladding material ISTRLCd 21200L is air and has a refractive index n.sub.ISTRLCd 21200Ln given by n.sub.ISTRLCd=1, and Right cladding material ISTRRCd 21200R is air and has a refractive index n.sub.ISTRRCd 21200Rn given by n.sub.ISTRRCd=1. Its bottom cladding ISTRBCd 21200B is silicon dioxide (this is part of a Burried-Oxide BOX layer in a typical Silicon-On-Insulator SOI wafer) with averaged refractive index n.sub.ISTRBCd 21200Bn of n.sub.ISTRBCd=1.45.

(425) The top cladding ITWTCd-z1 22300T-z1 before going into the ALS region is silicon dioxide (SiO.sub.2) has refractive index n.sub.ITWTCd-z1 22300Tn-z1 with n.sub.ITWTCd-z1=1.45.

(426) The bottom cladding ITWBCd-z1 22300B-z1 before going into the ALS region is silicon dioxide (SiO.sub.2) has refractive index n.sub.ITWBCd-z1 22300Bn-z1 with n.sub.ITWBCd-z1=1.45.

(427) The left cladding ITWLCd-z1 22300L-z1 before going into the ALS region is silicon dioxide (SiO.sub.2) has refractive index n.sub.ITWLCd-z1 22300Ln-z1 with n.sub.ITWLCd-z1=1.45.

(428) The right cladding ITWRCd-z1 22300R-z1 before going into the ALS region is silicon dioxide (SiO.sub.2) has refractive index n.sub.ITWRCd-z1 22300Rn-z1 with n.sub.ITWRCd-z1=1.45.

(429) In this exemplary embodiment, n.sub.ITWRCd-z1=n.sub.ITWBCd-z1=n.sub.ITWLCd-z1=n.sub.ITWRCd-z1=n.sub.ICWTCd, and n.sub.ICWTCd=n.sub.ICWBCd=n.sub.ICWLCd=n.sub.ICWRCd. Input tapering waveguide core ITWCo 22300 starting at z1=z1ALS 22300z1ALS, where z1ALS=10 micrometers, is laid with an active layer structure ALS 22500. 0<z1ALS<g.sub.ITWCo.

(430) Active Layer Structure-Beam Transport into the Structure

(431) The active layer structure ALS 22500 is shown by the Table 3-2 below:

(432) TABLE-US-00004 TABLE 3-2 ALS 22500 Layer Doping/ Layer Number Thickness NPNN TCO CASE (1/cm.sup.3) BIM 100 nm In.sub.2O.sub.3 (21250) BSCOC 1 100 nm InGaAsP 1.3 um N = 1 (21300) (Bottom layer-just 10.sup.19 above the substrate) BIDC 2 40 nm InP N = 1 (21350LN) 10.sup.19 BVC 3 20 nm InGaAsP 1.3 um N = 1 (21400) 10.sup.19 EC 4 10 nm AlGaInAs 1.3 um N.sub.1 = 1 (21500LN.sub.1) 10.sup.19 EC 5 11 nm barrier AlGaInAs/1.19 um/0.3% MI.sub.1 (21500MLI.sub.1) tensile strained EC 6 3 7 nm AlGaInAs/1.19 um/0.3% MI.sub.2 (21500MLI.sub.2) barrier inside tensile strained EC 7 4 11 nm AlGaInAs/1.55 um/0.31% MI.sub.3 (21500MLI.sub.3) Well (PL = compressive strained 1500 nm) EC 8 11 nm barrier AlGaInAs/1.19 um/0.3% MI.sub.4 (21500MLI.sub.4) tensile strained EC 9 43 nm AlGaInAs 1.3 um MI.sub.5 (21500MLI.sub.5) EC 10 20 nm AlGaInAs 1.3 um P.sub.1 = 1 (21500LP.sub.1) 10.sup.18 TVC 11 25 nm InGaAsP 1.3 um P.sub.2 = 1 (21600P.sub.2) 10.sup.18 TVC 12 20 nm InGaAsP 1.3 um N.sub.2 = 1 (21600N.sub.2) 10.sup.19 TIDC 13 20 nm InP N = 1 (21650) 10.sup.19 TVSCOC 14 240 nm In.sub.2O.sub.3 (Top layer) (21700) Total 625 nm

(433) In the table, the materials are unstrained (with InP as the substrate) if not specified as strained. The wavelength specified will be the material bandgap wavelength of the quaternary material involved (proper choice of the material composition is needed to achieve the required material bandgap and strain when grown on InP substrate).

(434) Bottom Side Conduction and Ohmic Contact Layer

(435) The active layer structure ALS 22500 has a bottom side conduction and Ohmic contact layer BSCOC 21300 that is InGaAsP layer given by layer 1 in Table 3-2 with thickness d.sub.BSC 21300d, where d.sub.BSC=100 nm and width w.sub.BSC 21300w, where w.sub.BSC is approximately 54 micrometers along most of the length of the ALS. Its refractive index n.sub.BSC 21300n is n.sub.BSC=3.4.

(436) Bottom Interspaced Material Layer

(437) The bottom interspaced material layer BIM 21250 is made of a Low-Refractive-Index Ohmic Transparent Conducting (LRI-OTC) material composed of Indium oxide (In.sub.2O.sub.3) with thickness d.sub.BIM 21250d equals to d.sub.BIM=100 nm, width w.sub.BIM 21250w equals to w.sub.BIM=54 micrometers, and average refractive index n.sub.BIM 21250n equals to n.sub.BIM=1.7.

(438) Bottom Metal Contact Pads

(439) The first bottom left metal contact pad FBLM 21900L is a multi-layer metal made up of (17 nm Au followed by 17 nm Ge followed by 17 nm Au followed by 17 nm Ni followed by 1000 nm Au) deposited on top of the top surface of n-doped layer 21300 given by layer 1 in Table 3-2. The total thickness of the metal contact pad is d.sub.FBLM 21900Ld, with d.sub.FBLM=1068 nm, and width w.sub.FBLM 21900Lw, where w.sub.FBLM is approximately 20 micrometers. The length of the metal contact pad g.sub.FBLM 21900Lg is approximately 500 micrometers.

(440) The first bottom right metal contact pad FBRM 21900R is multilayer metal made up of (17 nm Au followed by 17 nm Ge followed by 17 nm Au followed by 17 nm Ni followed by 1000 nm Au) deposited on top of the top surface of n-doped layer 21300 given by layer 1 in Table 3-2. The total thickness of the metal contact pad is d.sub.FBRM 21900Rd, with d.sub.FBRM=1068 nm, and width w.sub.FBRM 21900Rw, where w.sub.FBRM is approximately 20 micrometers. The length of the metal contact pad g.sub.FBRM 21900Rg is approximately 500 micrometers.

(441) Bottom Metal Electrodes

(442) On top of the first bottom left metal contact pad FBLM 21900L is deposited the first bottom left metal electrode FBLME 21120L which is gold of thickness of approximately 2 micrometer thick.

(443) On top of the first bottom right metal contact pad FBRM 21900R is deposited the first bottom right metal electrode FBRME 21120R which is gold of thickness of approximately 2 micrometer thick.

(444) Bottom Interspaced Dielectric Current Conduction Layer

(445) Bottom interspaced dielectric current conduction layer BIDC 21350 is a n-doped InP given by layer 2 in Table 3-2 with thickness d.sub.BIDC 21350d equals to d.sub.BIDC=40 nm, width w.sub.BIDC 21350w equals to w.sub.BIDC=54 micrometers, and average refractive index n.sub.BIDC 21350n equals to about n.sub.BIDC=3.0.

(446) Bottom Vertical Current Conduction Layer

(447) Bottom vertical current conduction layer BVC 21400 is n-doped InGaAsP given by layer 3 in Table 3-2 with thickness d.sub.BVC 21400d equals to d.sub.BVC=20 nm, width w.sub.BVC 21400w equals to w.sub.BVC=2 micrometers, and an averaged refractive index n.sub.BVC 21400n equals to n.sub.BVC=3.4.

(448) Electro-Active Layer

(449) Electro-active layer EC 21500 is made up of layers 4, 5, 6, 7, 8, 9, 10 in Table 3-1 with an averaged refractive index of the entire layer given by n.sub.EC 21500n with n.sub.EC equals to approximately n.sub.EC=3.4. Under an applied electric field, there will be a change in averaged refractive index dn.sub.EC 21500dn. The average refractive index becomes n.sub.EC(new)=n.sub.EC+dn.sub.EC.

(450) The total thickness d.sub.EC 21500d of this Electro-active layer is d.sub.EC=160 nm. Its width w.sub.EC 21500w is equal to w.sub.EC=2 micrometers.

(451) The electro-active layer has a PqN junction at layer 4 to 10 for which layer 4 is layer 21500LN.sub.1 that is N-doped with a dopant density of 21500N.sub.1=110.sup.19/cm.sup.3 and layer 10 is layer 21500LP.sub.1 that is P-doped with a dopant density of 21500P.sub.1=110.sup.18/cm.sup.3

(452) The intermediate layers 21500MLN.sub.m are all undoped (intrinsic).

(453) The applied field E.sub.EC 21500E (which may cause a current C.sub.EC 21500C to flow) is across the entire electro-active layer with a negative voltage applied to the top and positive voltage applied to the bottom of this entire electro-active layer known to those skilled in the art as revered bias (with respect to the PN junction in the electro-active layer) of voltage V.sub.R 21500VR so the applied electro-active V.sub.EC 21500VEC is V.sub.R.

(454) The voltage applied to the electrodes of the modulator V.sub.MOD 20000V is approximately given by V.sub.EC.

(455) Top Vertical Current Conduction Layer

(456) Top vertical current conduction layer TVC 21600 is given by layer 11 and 12 in Table 3-2 made up of InGaAsP layer that is composed of 25 nm-thick layer 21600LP.sub.2 that is P-doped with dopant density 21600P.sub.2=110.sup.18/cm.sup.3, followed by 20 nm-thick N-doped InGaAsP layer 21600LN.sub.2 with dopant density 21600N.sub.2=110.sup.19/cm.sup.3. The total thickness for TVC 21600 is d.sub.TVC 21600d with d.sub.TVC=45 nm. Its width is W.sub.TVC 21600w equals to W.sub.TVC=2 micrometers, and its averaged refractive index is n.sub.TVC 21600n equals to n.sub.TVC=3.4. This N.sub.2P.sub.2 junction forms a forward-Biased PN Junction (or Tunnel PN Junction). It forms a PN-changing PN junction (called PNCPN junction) 21600PNCPN.

(457) Top Interspaced Dielectric Current Conduction Layer

(458) Top interspaced dielectric conduction layer TIDC 21650 is N-doped InP layer given by layer 13 in Table 3-2 with thickness d.sub.TIDC 21650d equals to d.sub.TIDC=20 nm, width w.sub.TIDC 21650w equals to w.sub.TIDC=2 micrometers, and averaged refractive index n.sub.TIDC 21650n equals to n.sub.TIDC=3.0.

(459) Top Vertical/Side Conduction and Ohmic Contact Layer

(460) Top vertical/side conduction and Ohmic contact layer TVSCOC 21700 is made up of Low-Refractive-Index Ohmic Transparent Conductor (LRI-OTC) (In.sub.2O.sub.3) given by layer 14 in Table 3-2 with thickness d.sub.TVSC 21700d equals to d.sub.TVSC=240 nm, width w.sub.TVSC 21700w equals to w.sub.TVSC=2 micrometers, and an averaged refractive index n.sub.TVSC 21700n equals to n.sub.TVSC=1.7.

(461) Top Metal Contact Pads

(462) The first top middle metal contact pad FTMM 21800M is multilayer metal made up of (17 nm Au followed by 17 nm Ge followed by 17 nm Au followed by 17 nm Ni followed by 1000 nm Au) deposited on top of the top surface of n-doped layer 21700 given by layer 14 in Table 3-2. The total thickness of the metal contact pad is d.sub.FTMM 21800Md, with d.sub.FTMM=1068 nm, and width w.sub.FTMM 21800Mw, where w.sub.FTMM is approximately 2 micrometers. The length of the metal contact pad g.sub.FTMM 21800Mg is approximately 500 micrometers.

(463) There is no top left or right metal contact pad FTLM 21800L or FTRM 21800R.

(464) Top Metal Electrodes

(465) On top of the first top middle metal contact pad FTMM 21800M is deposited the first top middle metal electrode FTMME 21130M which is gold of thickness of approximately 2 micrometer thick.

(466) Beam Transport to Electro-Active Waveguiding Core Structure

(467) Input tapering waveguide region between z1=z1ALS 22300z1ALS and z1=g.sub.ITWCo 22300g, Tapering waveguide core width w.sub.ITWCo-z 22300w varies down to a smaller value of w.sub.ITWCo-g=50 nm at z1=g.sub.ITWCo 22300g from its vale at z1=z1ALS 22300z1ALS. Clearly W.sub.ITWCo-g<<.sub.bm/(2*n.sub.ITWCo), with .sub.bm=1550 nm and n.sub.ITWCo=3.6, where * is number multiplication.

(468) Output Connecting Waveguide

(469) Output connecting waveguide core OCWCo 28200 has averaged Refractive Index n.sub.OCWCo=n.sub.aOCWCo=3.6, thickness d.sub.OCWCo 28200d is d.sub.OCWCo=250 nm, and width W.sub.OCWCo 28200w is W.sub.OCWCo=400 nm.

(470) Output connecting waveguide OCWG 28200WG has Output connecting-waveguide bottom cladding material OCWBCd 28200B that is silicon dioxide (SiO.sub.2) for which the refractive index n.sub.OCWBCd 28200Bn is n.sub.OCWBCd=1.45.

(471) Output connecting waveguide top cladding material OCWTCd 28200T is silicon dioxide foe which the refractive index n.sub.OCWTCd 28200Tn is n.sub.OCWTCd=1.45.

(472) Output connecting waveguide left cladding material OCWLCd 28200L is silicon dioxide for which the refractive index n.sub.OCWLCd 28200Ln is n.sub.OCWLCd=1.45.

(473) Output connecting waveguide right cladding material OCWRCd 28200R is silicon dioxide for which the refractive index n.sub.OCWRCd 28200Rn is n.sub.OCWRCd=1.45.

(474) The resulted averaged cladding refractive Index n.sub.aOCWCd 28200aCdn is n.sub.aOCWCd=1.45.

(475) Output Optical Beam OBM 28140

(476) Output Beam Coupler Structure (OBCS) Region

(477) Output tapering waveguide core OTWCo 28300 is made of silicon. Its width at a location z2 OTWCo-z2 is denoted as width w.sub.OTWCo-z2 28300w-z2. This width is tapered from width at z2=0 w.sub.OTWCo-z2=0 28300w-z2=0 that has a value of w.sub.OTWCo-z2=0=400 nm to a width at z2>0 w.sub.OTWCo-z2>0 28300w-z2>0 that is narrower than 400 nm in a linear fashion.

(478) The thickness of the tapering waveguide core d.sub.OTWCo-z2 28300d-z2 made of silicon is d.sub.OTWCo-z2=250 nm with a refractive index n.sub.OTWCo-z2 28300n-z2 that is n.sub.OTWCo-z2=3.6.

(479) The total length of tapering waveguide g.sub.OTWCo 28300g is g.sub.OTWCo=20 micrometers (m). The width of the waveguide core at the end of the tapering at z2=g.sub.OTWCo is w.sub.OTWCo-g 28300w-g with w.sub.OTWCo-g=50 nm.

(480) Output supporting structure OSTR 29200 has width w.sub.OSTR 29200w with w.sub.OSTR=50 nm and thickness d.sub.OSTR 29200d with d.sub.OSTR=250 nm and length g.sub.OSTR 29200g with g.sub.OSTR=20 micrometers. It has an effective layer averaged refractive index n.sub.laOSTR 29200nla with n.sub.laOSTR<2.5.

(481) The top cladding OTWTCd-z2 28300T-z2 is silicon dioxide (SiO.sub.2) has refractive index n.sub.OTWTCd-z2 28300Tn-z2 with n.sub.OTWTCd-z2=1.45 before going into the ALS region.

(482) The bottom cladding OTWBCd-z2 28300B-z2 is silicon dioxide (SiO.sub.2) has refractive index n.sub.OTWBCd-z2 28300Bn-z2 with n.sub.OTWBCd-z2=1.45.

(483) The left cladding OTWLCd-z2 28300L-z2 is silicon dioxide (SiO.sub.2) has refractive index n.sub.OTWLCd-z2 28300Ln-z2 with n.sub.OTWLCd-z2=1.45.

(484) The right cladding OTWRCd-z2 28300R-z2 is silicon dioxide (SiO.sub.2) has refractive index n.sub.OTWRCd-z2 28300Rn-z2 with n.sub.OTWRCd-z2=1.45.

(485) In this exemplary embodiment, n.sub.OTWTCd-z2=n.sub.OTWBCd-z2=n.sub.OTWLCd-z2=n.sub.OTWRCd-z2=n.sub.OCWTCd, and n.sub.OCWTCd=n.sub.OCWBCd=n.sub.OCWLCd=n.sub.OCWRCd

(486) Output tapering waveguide core OTWCo 28300 starting at z2=z2ALS 28300z2ALS, is laid with an active layer structure ALS 22500. 0<z2ALS<g.sub.OTWCo.

(487) Most of the output optical beam energy of beam OBM 28140 is transported to output tapering waveguide core OTWCo 28300 from the electro-active waveguiding core structure EWCoS 22600, through the output tapering waveguide region between z2=z2ALS 28300z2ALS and z2=g.sub.OTWCo 28300g, where the output tapering waveguide core width w.sub.OTWCo-z2 28300w-z2 varies down to a smaller value of w.sub.OTWCo-g at z2=g.sub.ITWCo 28300g from its vale at z2=z2ALS, 28300z2ALS. The tapering waveguide core width is reduced to well below half the optical wavelength in the waveguide core given by .sub.bm/(2n.sub.OTWCo) so that w.sub.OTWCo-g<<.sub.bm/(2n.sub.OTWCo). After the energy transported from the electro-active waveguiding core structure EWCoS 22600 that contains the electro-active layer EC 21500 down to the output taper at z2=0 where the taper core width is w.sub.OTWCo-z2=0 28300w0 and w.sub.OTWCo-z2=0=w.sub.OCWCo 28200, the optical beam is denoted as output optical beam or beam OBM 28140.

(488) Length of Active Layer Structure

(489) The length of the active layer structure SL.sub.mod 22550 is approximately 500 micrometers.

(490) High Refractive Index Contrast and Mode Overlapping

(491) For the bottom cladding:

(492) Waveguide core refractive index is n.sub.BCo=3.6

(493) Waveguide bottom cladding is n.sub.BCd=1.45 (given by layer ISTRBC with n.sub.ISTRBCd=1.45)

(494) Waveguide core-to-cladding refractive index difference square to be n.sub.rd.sup.2=(n.sub.co.sup.2n.sub.BCd.sup.2)=10.86.

(495) Refractive index contrast ratio to be: R.sub.cts=n.sub.rd.sup.2/(n.sub.co.sup.2+n.sub.BCd.sup.2)=0.7, which is in the very-strongly guiding regime.

(496) For the top cladding:

(497) Waveguide core refractive index is n.sub.co=3.6

(498) Waveguide bottom cladding is n.sub.TCd=1.7 (given by TVSCOC layer which is In.sub.2O.sub.3 with n.sub.TVSCOC=1.7)

(499) Waveguide core-to-cladding refractive index difference square to be n.sub.rd.sup.2=(n.sub.co.sup.2n.sub.TCd.sup.2)=10.

(500) Refractive index contrast ratio to be: R.sub.cts=n.sub.rd.sup.2/(n.sub.co.sup.2+n.sub.TCd.sup.2)=0.64, which is in the very-strongly guiding regime.

(501) A Third Exemplary Device of Electro-Optic Modulator with Transparent Conductor Geometry

(502) A preferred embodiment of an exemplary device is Modulator Device 20000 with the following specifications referred to as a third exemplary device of electro-optic modulator with Ohmic transparent conductor geometry: The main difference between this and the Second Exemplary Device is in Table 3-3, in which the active layer is designed for electro-optic modulation.

(503) Substrate SUB 21100 is silicon wafer substrate with a thickness of about 0.3 mm.

(504) Input connecting waveguide core ICWCo 22200 is made of silicon for which its averaged material refractive index n.sub.ICWCo 22200n is around n.sub.ICWCo=3.6, thickness d.sub.ICWCo 22200d is d.sub.ICWCo=250 nm, and width W.sub.ICWCo 22200w is W.sub.ICWCo=400 nm.

(505) Input connecting-waveguide bottom cladding material ICWBCd 22200B is silicon dioxide (SiO.sub.2), for which its refractive Index n.sub.ICWBCd 22200Bn is n.sub.ICWBCd=1.45.

(506) Input connecting waveguide top cladding material ICWTCd 22200T is silicon dioxide (SiO.sub.2) for which its refractive Index n.sub.ICWTCd 22200Tn is 1.45.

(507) Input connecting waveguide left cladding material ICWLCd 22200L is silicon dioxide (SiO.sub.2), for which its refractive Index n.sub.ICWLCd 22200Ln is 1.45

(508) Input connecting waveguide right cladding material ICWRCd 22200R is silicon dioxide (SiO.sub.2), for which its refractive Index n.sub.ICWRCd 22200Rn is 1.45

(509) The above form an input connecting waveguide ICWG 22200WG. The core-cladding refractive-index difference n.sub.Rd defined by n.sub.Rd.sup.2=(n.sub.Co.sup.2n.sub.Cd.sup.2) for waveguide ICWG 22200WG is n.sub.Rd.sup.2=(3.6.sup.21.45.sup.2)=10.86 with n.sub.Co=3.6 and n.sub.Cd=1.45. Its averaged Cladding Refractive Index is given by n.sub.aICWCd=(n.sub.ICWBCd.sup.2A.sub.ICWBCd+n.sub.ICWTCd.sup.2A.sub.ICWTCd+n.sub.ICWRCd.sup.2 A.sub.ICWRCd+n.sub.ICWLCd.sup.2 A.sub.ICWLCd)/(A.sub.ICWBCd+A.sub.ICWTCd+A.sub.ICWRCd+A.sub.ICWLCd).sup.0.5=1.45. Its averaged Core Refractive Index is given by n.sub.aCo=(n.sub.Co1.sup.2A.sub.Co1+n.sub.Co2.sup.2A.sub.Co2+n.sub.Co3.sup.2 A.sub.Co3+ . . . +n.sub.Com.sup.2 A.sub.Com)/(A.sub.Co1+A.sub.Co2+A.sub.Co3+ . . . +A.sub.Com).sup.0.5=3.6.

(510) The input optical beam IBM 22140 has propagating refractive index n.sub.IBM 22140n, for which n.sub.IBM is approximately 2.8 with optical power P.sub.bm 22140P approximately 1 mW, electric field polarization E.sub.bm 22140E to be in the horizontal direction parallel to the substrate surface. It has a beam effective area A.sub.bm 22140A of A.sub.bm=0.04 m.sup.2 and an optical wavelength centered at .sub.bm 22140L with .sub.bm=1550 nm with plurality of (N) frequency channels .sub.bm1=1548 nm, .sub.bm2=1549 nm, .sub.bm3=1550 nm, .sub.bm4=1551 nm, and .sub.bm3=1552 nm centered at .sub.bm=1550 nm.

(511) Input Beam Coupler Structure (IBCS) Region

(512) The input tapering waveguide core ITWCo 223000 is made of silicon. Its width at a location z1, ITWCo-z1 22300z1 is denoted as width w.sub.ITWCo-z1 22300w-z1. This width is tapered from width at z1=0 w.sub.ITWCo-z1=0 22300w-z1=0 that has a value of w.sub.ITWCo-z1=0=400 nanometers (nm) to a width at z1>0 w.sub.ITWCo-z1>0 22300w-z1>0 that is narrower than 400 nm in a linear fashion.

(513) The thickness of the tapering waveguide core d.sub.ITWCo-z1 22300d-z1 made of silicon is d.sub.ITWCo-z1=250 nm with a refractive index n.sub.ITWCo-z1 22300n-z that is n.sub.ITWCo-z1=3.6.

(514) The total length of tapering waveguide g.sub.ITWCo 22300g is g.sub.ITWCo=20 micrometers (m). The width of the waveguide core at the end of the tapering at z1=g.sub.ITWCo is w.sub.ITWCo-g 22300w-g with w.sub.ITWCo-g=50 nm.

(515) Input supporting structure ISTR 21200 has width w.sub.ISTR 21200w with w.sub.ISTR=50 nm and thickness d.sub.ISTR 21200d with d.sub.ISTR=250 nm and length g.sub.ISTR 21200g with g.sub.ISTR=20 micrometers. It has an effective layer averaged refractive index n.sub.laISTR 21200nla with n.sub.laISTR<2.5.

(516) Left cladding material ISTRLCd 21200L is air and has a refractive index n.sub.ISTRLCd 21200Ln given by n.sub.ISTRLCd=1, and Right cladding material ISTRRCd 21200R is air and has a refractive index n.sub.ISTRRCd 21200Rn given by n.sub.ISTRRCd=1. Its bottom cladding ISTRBCd 21200B is silicon dioxide (this is part of a Burried-Oxide BOX layer in a typical Silicon-On-Insulator SOI wafer) with averaged refractive index n.sub.ISTRBCd 21200Bn of n.sub.ISTRBCd=1.45.

(517) The top cladding ITWTCd-z1 22300T-z1 before going into the ALS region is silicon dioxide (SiO.sub.2) has refractive index n.sub.ITWTCd-z1 22300Tn-z1 with n.sub.ITWTCd-z1=1.45.

(518) The bottom cladding ITWBCd-z1 22300B-z1 before going into the ALS region is silicon dioxide (SiO.sub.2) has refractive index n.sub.ITWBCd-z1 22300Bn-z1 with n.sub.ITWBCd-z1=1.45.

(519) The left cladding ITWLCd-z1 22300L-z1 before going into the ALS region is silicon dioxide (SiO.sub.2) has refractive index n.sub.ITWLCd-z1 22300Ln-z1 with n.sub.ITWLCd-z1=1.45.

(520) The right cladding ITWRCd-z1 22300R-z1 before going into the ALS region is silicon dioxide (SiO.sub.2) has refractive index n.sub.ITWRCd-z1 22300Rn-z1 with n.sub.ITWRCd-z1=1.45.

(521) In this exemplary embodiment, n.sub.ITWTCd-z1=n.sub.ITWBCd-z1=n.sub.ITWLCd-z1=n.sub.ITWRCd-z1=n.sub.ICWTCd, and n.sub.ICWTCd=n.sub.ICWBCd=n.sub.ICWLCd=n.sub.ICWRCd.mInput tapering waveguide core ITWCo 22300 starting at z1=z1ALS 22300z1ALS, where z1ALS=10 micrometers, is laid with an active layer structure ALS 22500. 0<z1ALS<g.sub.ITWCo.

(522) Active Layer Structure-Beam Transport into the Structure

(523) The active layer structure ALS 22500 is shown by the Table 3-3 below:

(524) TABLE-US-00005 TABLE 3-3 ALS 22500 Layer Doping/ Layer Number Thickness NPNN TCO CASE (1/cm.sup.3) BIM 100 nm In.sub.2O.sub.3 (21250) BSCOC 1 100 nm InGaAsP 1.3 um N = 1 (21300) (Bottom layer-just 10.sup.19 above the substrate) BIDC 2 40 nm InP N = 1 (21350LN) 10.sup.19 BVC 3 20 nm InGaAsP 1.3 um N = 1 (21400) 10.sup.19 EC 4 10 nm AlGaInAs 1.3 um N.sub.1 = 1 (21500LN.sub.1) 10.sup.19 EC 5 4 nm barrier AlGaInAs/1.1 um/0.8% MN.sub.1 = 4 (21500MLN.sub.1) tensile strained 10.sup.17 EC 6 2 7 nm AlGaInAs/1.1 um/0.8% MN.sub.2 = 4 (21500MLN.sub.2) barrier inside tensile strained 10.sup.17 EC 7 3 6.5 nm AlGaInAs/1.55 um/0.9% MN.sub.3 = 4 (21500MLN.sub.3) Well (PL = compressive strained 10.sup.17 1350 nm) EC 8 4 nm barrier AlGaInAs/1.1 um/0.8% MN.sub.4 = 4 (21500MLN.sub.4) tensile strained 10.sup.17 EC 9 43 nm AlGaInAs 1.3 um MN.sub.5 = 4 (21500MLN.sub.5) 10.sup.17 EC 10 20 nm AlGaInAs 1.3 um P.sub.1 = 1 (21500LP.sub.1) 10.sup.18 TVC 11 25 nm InGaAsP 1.3 um P.sub.2 = 1 (21600P.sub.2) 10.sup.18 TVC 12 20 nm InGaAsP 1.3 um N.sub.2 = 1 (21600N.sub.2) 10.sup.19 TIDC 13 20 nm InP N = 1 (21650) 10.sup.19 TVSCOC 14 240 nm In.sub.2O.sub.3 (Top layer) N = 1 (21700) 10.sup.19 Total 580 nm
In the table, the materials are unstrained (with InP as the substrate) if not specified as strained. The wavelength specified will be the material bandgap wavelength of the quaternary material involved (proper choice of the material composition is needed to achieve the required material bandgap and strain when grown on InP substrate).
Bottom Side Conduction and Ohmic Contact Layer

(525) The active layer structure ALS 22500 has a bottom side conduction and Ohmic contact layer BSCOC 21300 that is InGaAsP layer given by layer 1 in Table 3-3 with thickness d.sub.BSC 21300d, where d.sub.BSC=100 nm and width w.sub.BSC 21300w, where w.sub.BSC is approximately 54 micrometers along most of the length of the ALS. Its refractive index n.sub.BSC 21300n is n.sub.BSC=3.4.

(526) Bottom Interspaced Material Layer

(527) The bottom interspaced material layer BIM 21250 is made of a Low-Refractive-Index Ohmic Transparent Conducting (LRI-OTC) material composed of Indium oxide (In.sub.2O.sub.3) with thickness d.sub.BIM 21250d equals to d.sub.BIM=100 nm, width w.sub.BIM 21250w equals to w.sub.BIM=54 micrometers, and average refractive index n.sub.BIM 21250n equals to n.sub.BIM=1.7.

(528) Bottom Metal Contact Pads

(529) The first bottom left metal contact pad FBLM 21900L is a multi-layer metal made up of (17 nm Au followed by 17 nm Ge followed by 17 nm Au followed by 17 nm Ni followed by 1000 nm Au) deposited on top of the top surface of n-doped layer 21300 given by layer 1 in Table 3-3. The total thickness of the metal contact pad is d.sub.FBLM 21900Ld, with d.sub.FBLM=1068 nm, and width w.sub.FBLM 21900Lw, where w.sub.FBLM is approximately 20 micrometers. The length of the metal contact pad g.sub.FBLM 21900Lg is approximately 500 micrometers.

(530) The first bottom right metal contact pad FBRM 21900R is multilayer metal made up of (17 nm Au followed by 17 nm Ge followed by 17 nm Au followed by 17 nm Ni followed by 100 nm Au) deposited on top of the top surface of n-doped layer 21300 given by layer 1 in Table 3-3. The total thickness of the metal contact pad is d.sub.FBRM 21900Rd, with d.sub.FBRM=1068 nm, and width w.sub.FBRM 21900Rw, where w.sub.FBRM is approximately 20 micrometers. The length of the metal contact pad g.sub.FBRM 21900Rg is approximately 500 micrometers.

(531) Bottom Metal Electrodes

(532) On top of the first bottom left metal contact pad FBLM 21900L is deposited the first bottom left metal electrode FBLME 21120L which is gold of thickness of approximately 2 micrometer thick.

(533) On top of the first bottom right metal contact pad FBRM 21900R is deposited the first bottom right metal electrode FBRME 21120R which is gold of thickness of approximately 2 micrometer thick.

(534) Bottom Interspaced Dielectric Current Conduction Layer

(535) Bottom interspaced dielectric current conduction layer BIDC 21350 is a n-doped InP given by layer 2 in Table 3-3 with thickness d.sub.BIDC 21350d equals to d.sub.BIDC=40 nm, width w.sub.BIDC 21350w equals to w.sub.BIDC=54 micrometers, and average refractive index n.sub.BIDC 21350n equals to about n.sub.BIDC=3.0.

(536) Bottom Vertical Current Conduction Layer

(537) Bottom vertical current conduction layer BVC 21400 is n-doped InGaAsP given by layer 3 in Table 3-3 with thickness d.sub.BVC 21400d equals to d.sub.BVC=20 nm, width w.sub.BVC 21400w equals to w.sub.BVC=2 micrometers, and an averaged refractive index nave 21400n equals to n.sub.BVC=3.4.

(538) Electro-Active Layer

(539) Electro-active layer EC 21500 is made up of layers 4, 5, 6, 7, 8, 9, 10 in Table 3-1 with an averaged refractive index of the entire layer given by n.sub.EC 21500n with n.sub.EC equals to approximately n.sub.EC=3.4. Under an applied electric field, there will be a change in averaged refractive index dn.sub.EC 21500dn. The average refractive index becomes n.sub.EC(new)=n.sub.EC+dn.sub.EC.

(540) The total thickness d.sub.EC 21500d of this Electro-active layer is d.sub.EC=160 nm. Its width w.sub.EC 21500w is equal to w.sub.EC=2 micrometers.

(541) The electro-active layer has a PqN junction at layer 4 to 10 for which layer 4 is layer 21500LN.sub.1 that is N-doped with a dopant density of 21500N.sub.1=110.sup.19/cm.sup.3 and layer 10 is layer 21500LP.sub.1 that is P-doped with a dopant density of 21500P.sub.1=110.sup.18/cm.sup.3

(542) The intermediate layers 21500MLN.sub.m are all N-doped with a dopant density of 21500MN.sub.m=410.sup.17/cm.sup.3.

(543) The applied field E.sub.EC 21500E (which may cause a current C.sub.EC 21500C to flow) is across the entire electro-active layer with a negative voltage applied to the top and positive voltage applied to the bottom of this entire electro-active layer known to those skilled in the art as revered bias (with respect to the PN junction in the electro-active layer) of voltage V.sub.R 21500VR so the applied electro-active V.sub.EC 21500VEC is V.sub.R.

(544) The voltage applied to the electrodes of the modulator V.sub.MOD 20000V is approximately given by V.sub.EC.

(545) Top Vertical Current Conduction Layer

(546) Top vertical current conduction layer TVC 21600 is given by layer 11 and 12 in Table 3-3 made up of InGaAsP layer that is composed of 25 nm-thick layer 21600LP.sub.2 that is P-doped with dopant density 21600P.sub.2=110.sup.18/cm.sup.3, followed by 20 nm-thick N-doped InGaAsP layer 21600LN.sub.2 with dopant density 21600N.sub.2=110.sup.19/cm.sup.3. The total thickness for TVC 21600 is d.sub.TVC 21600d with d.sub.TVC=45 nm. Its width is W.sub.TVC 21600w equals to W.sub.TVC=2 micrometers, and its averaged refractive index is n.sub.TVC 21600n equals to n.sub.TVC=3.4. This N.sub.2P.sub.2 junction forms a forward-Biased PN Junction (or Tunnel PN Junction). It forms a PN-changing PN junction (called PNCPN junction) 21600PNCPN.

(547) Top Interspaced Dielectric Current Conduction Layer

(548) Top interspaced dielectric conduction layer TIDC 21650 is N-doped InP layer given by layer 13 in Table 3-3 with thickness d.sub.TIDC 21650d equals to d.sub.TIDC=20 nm, width w.sub.TIDC 21650w equals to w.sub.TIDC=2 micrometers, and averaged refractive index n.sub.TIDC 21650n equals to n.sub.TIDC=3.0.

(549) Top Vertical/Side Conduction and Ohmic Contact Layer

(550) Top vertical/side conduction and Ohmic contact layer TVSCOC 21700 is made up of Low-Refractive-Index Ohmic Transparent Conductor (LRI-OTC) (In.sub.2O.sub.3) given by layer 14 in Table 3-3 with thickness d.sub.TVSC 21700d equals to d.sub.TVSC=240 nm, width w.sub.TVSC 21700w equals to w.sub.TVSC=2 micrometers, and an averaged refractive index n.sub.TVSC 21700n equals to n.sub.TVSC=1.7.

(551) Top Metal Contact Pads

(552) The first top middle metal contact pad FTMM 21800M is multilayer metal made up of (17 nm Au followed by 17 nm Ge followed by 17 nm Au followed by 17 nm Ni followed by 1000 nm Au) deposited on top of the top surface of n-doped layer 21700 given by layer 14 in Table 3-3. The total thickness of the metal contact pad is d.sub.FTMM 21800Md, with d.sub.FTMM=1068 nm and width w.sub.FTMM 21800Mw, where w.sub.FTMM is approximately 2 micrometers. The length of the metal contact pad g.sub.FTMM 21800Mg is approximately 500 micrometers.

(553) There is no top left or right metal contact pad FTLM 21800L or FTRM 21800R.

(554) Top Metal Electrodes

(555) On top of the first top middle metal contact pad FTMM 21800M is deposited the first top middle metal electrode FTMME 21130M which is gold of thickness of approximately 2 micrometer thick.

(556) Beam Transport to Electro-Active Waveguiding Core Structure

(557) Input tapering waveguide region between z1=z1ALS 22300z1ALS and z1=g.sub.ITWCo 22300g, Tapering waveguide core width w.sub.ITWCo-z1 22300w varies down to a smaller value of w.sub.ITWCo-g=50 nm at z1=g.sub.ITWCo 22300g from its vale at z1=z1ALS 22300z1ALS.

(558) Clearly W.sub.ITWCo-g<<.sub.bm(2*n.sub.ITWCo), with .sub.bm=1550 nm and n.sub.ITWCo=3.6, where * is number multiplication.

(559) Output Connecting Waveguide

(560) Output connecting waveguide core OCWCo 28200 has averaged Refractive Index n.sub.OCWCo=n.sub.aOCWCo=3.6, thickness d.sub.OCWCo 28200d is d.sub.OCWCo=250 nm, and width W.sub.OCWCo 28200w is W.sub.OCWCo=400 nm.

(561) Output connecting waveguide OCWG 28200WG has Output connecting-waveguide bottom cladding material OCWBCd 28200B that is silicon dioxide (SiO.sub.2) for which the refractive index n.sub.OCWBCd 28200Bn is n.sub.OCWBCd=1.45.

(562) Output connecting waveguide top cladding material OCWTCd 28200T is silicon dioxide foe which the refractive index n.sub.OCWTCd 28200Tn is n.sub.OCWTCd=1.45.

(563) Output connecting waveguide left cladding material OCWLCd 28200L is silicon dioxide for which the refractive index n.sub.OCWLCd 28200Ln is n.sub.OCWLCd=1.45.

(564) Output connecting waveguide right cladding material OCWRCd 28200R is silicon dioxide for which the refractive index n.sub.OCWRCd 28200Rn is n.sub.OCWRCd=1.45.

(565) The resulted averaged cladding refractive Index n.sub.aOCWCd 28200aCdn is n.sub.aOCWCd=1.45.

(566) Output Optical Beam OBM 28140

(567) Output Beam Coupler Structure (OBCS) Region

(568) Output tapering waveguide core OTWCo 28300 is made of silicon. Its width at a location z2 OTWCo-z2 is denoted as width w.sub.OTWCo-z2 28300w-z2. This width is tapered from width at z2=0 w.sub.OTWCo-z2=0 28300w-z2=0 that has a value of w.sub.OTWCo-z2=0=400 nm to a width at z2>0 w.sub.OTWCo-z2>0 28300w-z2>0 that is narrower than 400 nm in a linear fashion.

(569) The thickness of the tapering waveguide core d.sub.OTWCo-z2 28300d-z2 made of silicon is d.sub.OTWCo-z2=250 nm with a refractive index n.sub.OTWCo-z2 28300n-z2 that is n.sub.OTWCo-z2=3.6.

(570) The total length of tapering waveguide g.sub.OTWCo 28300g is g.sub.OTWCo=20 micrometers (m). The width of the waveguide core at the end of the tapering at z2=g.sub.OTWCo is w.sub.OTWCo-g 28300w-g with w.sub.OTWCo-g=50 nm.

(571) Output supporting structure OSTR 29200 has width w.sub.OSTR 29200w with w.sub.OSTR=50 nm and thickness d.sub.OSTR 29200d with d.sub.OSTR=250 nm and length g.sub.OSTR 29200g with g.sub.OSTR=20 micrometers. It has an effective layer averaged refractive index n.sub.laOSTR 29200nla with n.sub.laOSTR<2.5.

(572) The top cladding OTWTCd-z2 28300T-z2 is silicon dioxide (SiO.sub.2) has refractive index n.sub.OTWTCd-z2 28300Tn-z2 with n.sub.OTWTCd-z2=1.45 before going into the ALS region.

(573) The bottom cladding OTWBCd-z2 28300B-z2 is silicon dioxide (SiO.sub.2) has refractive index n.sub.OTWBCd-z2 28300Bn-z2 with n.sub.OTWBCd-z2=1.45.

(574) The left cladding OTWLCd-z2 28300L-z2 is silicon dioxide (SiO.sub.2) has refractive index n.sub.OTWLCd-z2 28300Ln-z2 with n.sub.OTWLCd-z2=1.45.

(575) The right cladding OTWRCd-z2 28300R-z2 is silicon dioxide (SiO.sub.2) has refractive index n.sub.OTWRCd-z2 28300Rn-z2 with n.sub.OTWRCd-z2=1.45.

(576) In this exemplary embodiment, n.sub.OTWTCd-z2=n.sub.OTWBCd-z2=n.sub.OTWLCd-z2=n.sub.OTWRCd-z2=n.sub.OCWTCd, and n.sub.OCWTCd=n.sub.OCWBCd=n.sub.OCWLCd=n.sub.OCWRCd.

(577) Output tapering waveguide core OTWCo 28300 starting at z2=z2ALS 28300z2ALS, is laid with an active layer structure ALS 22500. 0<z2ALS<g.sub.OTWCo.

(578) Most of the output optical beam energy of beam OBM 28140 is transported to output tapering waveguide core OTWCo 28300 from the electro-active waveguiding core structure EWCoS 22600, through the output tapering waveguide region between z2=z2ALS 28300z2ALS and z2=g.sub.OTWCo 28300g, where the output tapering waveguide core width w.sub.OTWCo-z2 28300w-z2 varies down to a smaller value of w.sub.OTWCo-g at z2=g.sub.ITWCo 28300g from its value at z2=z2ALS, 28300z2ALS. The tapering waveguide core width is reduced to well below half the optical wavelength in the waveguide core given by .sub.bm/(2n.sub.OTWCo) so that w.sub.OTWCo-g<<.sub.bm/(2n.sub.OTWCo). After the energy transported from the electro-active waveguiding core structure EWCoS 22600 that contains the electro-active layer EC 21500 down to the output taper at z2=0 where the taper core width is w.sub.OTWCo-z2=0 28300w0 and w.sub.OTWCo-z2=0=w.sub.OTWCo 28200, the optical beam is denoted as output optical beam or beam OBM 28140.

(579) Length of Active Layer Structure

(580) The length of the active layer structure SL.sub.mod 22550 is approximately 500 micrometers.

(581) High Refractive Index Contrast and Mode Overlapping

(582) For the bottom cladding:

(583) Waveguide core refractive index is n.sub.co=3.6

(584) Waveguide bottom cladding is n.sub.BCd=1.45 (given by layer ISTRBC with n.sub.ISTRBCd=1.45)

(585) Waveguide core-to-cladding refractive index difference square to be n.sub.rd.sup.2=(n.sub.co.sup.2n.sub.BCd.sup.2)=10.86.

(586) Refractive index contrast ratio to be: R.sub.cts=n.sub.rd.sup.2/(n.sub.co.sup.2+n.sub.BCd.sup.2)=0.7, which is in the very-strongly guiding regime.

(587) For the top cladding:

(588) Waveguide core refractive index is n.sub.co=3.6

(589) Waveguide bottom cladding is n.sub.TCd=1.7 (given by TVSCOC layer which is In.sub.2O.sub.3 with n.sub.TVSCOC=1.7)

(590) Waveguide core-to-cladding refractive index difference square to be n.sub.rd.sup.2=(n.sub.co.sup.2n.sub.TCd.sup.2)=10.

(591) Refractive index contrast ratio to be: R.sub.cts=n.sub.rd.sup.2/(n.sub.co.sup.2+n.sub.TCd.sup.2)=0.64, which is in the very-strongly guiding regime.

(592) Final Summary

(593) The main parts of the embodiments can be summarized as follows: 1. A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co). 2. A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co). The electro-active layer has a low-refractive-index Ohmic transparent conductor (LRI-OTC) layer electrically connected from the top to the electro-active layer. The LRI-OTC forms part of the top electro-active waveguide cladding. 3. A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin region or very-thin such that d.sub.CORE<(.sub.op/n.sub.Co). 4. A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in an very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.5, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin or very-thin region such that d.sub.CORE<(.sub.op/n.sub.Co).

(594) (PqN Case) 5. A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co). A structure electrically connected to the electro-active layer comprises at least a first PN junction in which a first P-layer with P-dopant is vertically connected (vertical means in a direction perpendicular to the substrate plane: horizontal means in a direction parallel to the substrate plane) to a first N-layer with N-dopant, or a PqN junction in which a first P-layer with P-dopant is connected to a first q-layer with either N or P dopant or that is undoped (i.e. being an Intrinsic semiconductor material) and the q-layer is further connected to a first N-layer with N-dopant. A voltage is applied across the first P-layer of this first PN junction, and the first N-layer of the first PN junction to result in an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. 6. (PqN case plus Tunnel) A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co). A structure electrically connected to the electro-active layer comprises at least a first PN junction in which a first P-layer with P-dopant is vertically connected (vertical means in a direction perpendicular to the substrate plane; horizontal means in a direction parallel to the substrate plane) to a first N-layer with N-dopant, or a PqN junction in which a first P-layer with P-dopant is connected to a first q-layer with either N or P dopant or that is undoped (i.e. being an Intrinsic semiconductor material) and the q-layer is further connected to a first N-layer with N-dopant. The first P-layer is electrically connected to a second P-layer with P-dopant of a second PN junction, referred to as the PN-changing PN junction (PNCPN). This second P-layer is electrically connected to a second N-layer with N-dopant of this second PN junction. A voltage is applied across the second N-layer of the second PN junction, and the first N-layer of the first PN junction to result in an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. 7. (NqN case) A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co). A structure electrically connected to the electro-active layer comprises at least a first NqN junction in which a first N-layer with N-dopant is vertically connected (vertical means in a direction perpendicular to the substrate plane; horizontal means in a direction parallel to the substrate plane) to a first q-layer with either N or P dopant or that is undoped (i.e. being an Intrinsic semiconductor material) and the first q-layer is further connected to a second N-layer with N-dopant. A voltage is applied across the first N-layer and the second N-layer of this first NqN junction to result in an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. 8. (PqN case plus Tunnel plus TCO) A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co). A structure electrically connected to the electro-active layer comprises at least a first PN junction in which a first P-layer with P-dopant is vertically connected (vertical means in a direction perpendicular to the substrate plane; horizontal means in a direction parallel to the substrate plane) to a first N-layer with N-dopant, or a PqN junction in which a first P-layer with P-dopant is connected to a first q-layer with either N or P dopant or that is undoped (i.e. being an Intrinsic semiconductor material) and the q-layer is further connected to a first N-layer with N-dopant. The first P-layer is electrically connected to a second P-layer with P-dopant of a second PN junction, referred to as the PN-changing PN junction (PNCPN). This second P-layer is electrically connected to a second N-layer with N-dopant of this second PN junction. A voltage is applied across the second N-layer of the second PN junction, and the first N-layer of the first PN junction to result in an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer has a low-refractive-index Ohmic transparent conductor (LRI-OTC) layer electrically connected from the top to the electro-active layer. The LRI-OTC forms part of the top electro-active waveguide cladding. 9. (PqN case plus Tunnel plus taper WG) A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index of the tapering waveguide core is given by n.sub.ITWCo-z1. The width w.sub.ITWCo-z1 of the input tapering waveguide core after penetrating below the electro-active layer is reduced from a value approximately equal to or larger than half the wavelength in the material .sub.op/(2n.sub.ITWCo-z1) to a value smaller than half the wavelength in the material .sub.op/(2n n.sub.ITWCo-z1), so that w.sub.ITWCo-z1<.sub.op/(2n n.sub.ITWCo-z1) at some point under the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co). A structure electrically connected to the electro-active layer comprises at least a first PN junction in which a first P-layer with P-dopant is vertically connected (vertical means in a direction perpendicular to the substrate plane; horizontal means in a direction parallel to the substrate plane) to a first N-layer with N-dopant, or a PqN junction in which a first P-layer with P-dopant is connected to a first q-layer with either N or P dopant or that is undoped (i.e. being an Intrinsic semiconductor material) and the q-layer is further connected to a first N-layer with N-dopant. The first P-layer is electrically connected to a second P-layer with P-dopant of a second PN junction, referred to as the PN-changing PN junction (PNCPN). This second P-layer is electrically connected to a second N-layer with N-dopant of this second PN junction. A voltage is applied across the second N-layer of the second PN junction, and the first N-layer of the first PN junction to result in an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. 10. (PqN case plus Tunnel plus taper WG plus QW) A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index of the tapering waveguide core is given by n.sub.ITWCo-z1. The width w.sub.ITWCo-z1 of the input tapering waveguide core after penetrating below the electro-active layer is reduced from a value approximately equal to or larger than half the wavelength in the material .sub.op/(2n.sub.ITWCo-z1) to a value smaller than half the wavelength in the material .sub.op/(2n n.sub.ITWCo-z1), so that w.sub.ITWCo-z1<.sub.op/(2n n.sub.ITWCo-z1) at some point under the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co). A structure electrically connected to the electro-active layer comprises at least a first PN junction in which a first P-layer with P-dopant is vertically connected (vertical means in a direction perpendicular to the substrate plane; horizontal means in a direction parallel to the substrate plane) to a first N-layer with N-dopant, or a PqN junction in which a first P-layer with P-dopant is connected to a first q-layer with either N or P dopant or that is undoped (i.e. being an Intrinsic semiconductor material) and the q-layer is further connected to a first N-layer with N-dopant. The first P-layer is electrically connected to a second P-layer with P-dopant of a second PN junction, referred to as the PN-changing PN junction (PNCPN). This second P-layer is electrically connected to a second N-layer with N-dopant of this second PN junction. A voltage is applied across the second N-layer of the second PN junction, and the first N-layer of the first PN junction to result in an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. 11. (PqN case plus Tunnel plus taper WG plus doped QW) A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index of the tapering waveguide core is given by n.sub.ITWCo-z1. The width w.sub.ITWCo-z1 of the input tapering waveguide core after penetrating below the electro-active layer is reduced from a value approximately equal to or larger than half the wavelength in the material .sub.op/(2n.sub.ITWCo-z1) to a value smaller than half the wavelength in the material .sub.op/(2n n.sub.ITWCo-z1), so that w.sub.ITWCo-z1<.sub.op/(2n n.sub.ITWCo-z1) at some point under the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co). A structure electrically connected to the electro-active layer comprises at least a first PN junction in which a first P-layer with P-dopant is vertically connected (vertical means in a direction perpendicular to the substrate plane; horizontal means in a direction parallel to the substrate plane) to a first N-layer with N-dopant, or a PqN junction in which a first P-layer with P-dopant is connected to a first q-layer with either N or P dopant or that is undoped (i.e. being an Intrinsic semiconductor material) and the q-layer is further connected to a first N-layer with N-dopant. At least one of the first P-layer, first N-layer, or the first q-layer contains at least one quantum well. The doping density at the quantum well is in the highly-doped, medium-highly-doped, very-highly-doped, or ultra-highly-doped regime with a dopant density higher than about 210.sup.17/cm.sup.3 with either N doping or P doping. The first P-layer is electrically connected to a second P-layer with P-dopant of a second PN junction, referred to as the PN-changing PN junction (PNCPN). This second P-layer is electrically connected to a second N-layer with N-dopant of this second PN junction. A voltage is applied across the second N-layer of the second PN junction, and the first N-layer of the first PN junction to result in an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. 12. (PqN case plus Tunnel plus taper WG plus very highly doped QW) A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index of the tapering waveguide core is given by n.sub.ITWCo-z1. The width w.sub.ITWCo-z1 of the input tapering waveguide core after penetrating below the electro-active layer is reduced from a value approximately equal to or larger than half the wavelength in the material .sub.op/(2n.sub.ITWCo-z1) to a value smaller than half the wavelength in the material .sub.op/(2n n.sub.ITWCo-z1), so that w.sub.ITWCo-z1<.sub.op/(2n n.sub.ITWCo-z1) at some point under the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co). A structure electrically connected to the electro-active layer comprises at least a first PN junction in which a first P-layer with P-dopant is vertically connected (vertical means in a direction perpendicular to the substrate plane; horizontal means in a direction parallel to the substrate plane) to a first N-layer with N-dopant, or a PqN junction in which a first P-layer with P-dopant is connected to a first q-layer with either N or P dopant or that is undoped (i.e. being an Intrinsic semiconductor material) and the q-layer is further connected to a first N-layer with N-dopant. At least one of the first P-layer, first N-layer, or the first q-layer contains at least one quantum well. The doping density at the quantum well is in the medium-highly-doped, very-highly-doped, or ultra-highly-doped regime with a dopant density higher than about 510.sup.17/cm.sup.3 with either N doping or P doping. The first P-layer is electrically connected to a second P-layer with P-dopant of a second PN junction, referred to as the PN-changing PN junction (PNCPN). This second P-layer is electrically connected to a second N-layer with N-dopant of this second PN junction. A voltage is applied across the second N-layer of the second PN junction, and the first N-layer of the first PN junction to result in an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. 13. (PqN case plus Tunnel plus taper WG plus ultra-highly doped QW) A lowloss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index of the tapering waveguide core is given by n.sub.ITWCo-z1. The width w.sub.ITWCo-z1 of the input tapering waveguide core after penetrating below the electro-active layer is reduced from a value approximately equal to or larger than half the wavelength in the material .sub.op/(2n.sub.ITWCo-z1) to a value smaller than half the wavelength in the material .sub.op/(2n n.sub.ITWCo-z1), so that w.sub.ITWCo-z1<.sub.op/(2n n.sub.ITWCo-z1) at some point under the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co). A structure electrically connected to the electro-active layer comprises at least a first PN junction in which a first P-layer with P-dopant is vertically connected (vertical means in a direction perpendicular to the substrate plane; horizontal means in a direction parallel to the substrate plane) to a first N-layer with N-dopant, or a PqN junction in which a first P-layer with P-dopant is connected to a first q-layer with either N or P dopant or that is undoped (i.e. being an Intrinsic semiconductor material) and the q-layer is further connected to a first N-layer with N-dopant. At least one of the first P-layer, first N-layer, or the first q-layer contains at least one quantum well. The doping density at the quantum well is in the very-highly-doped, or ultra-highly-doped regime with a dopant density higher than about 1.510.sup.18/cm.sup.3 with either N doping or P doping. The first P-layer is electrically connected to a second P-layer with P-dopant of a second PN junction, referred to as the PN-changing PN junction (PNCPN). This second P-layer is electrically connected to a second N-layer with N-dopant of this second PN junction. A voltage is applied across the second N-layer of the second PN junction, and the first N-layer of the first PN junction to result in an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. 14. (PqN case plus Tunnel plus taper WG plus ultra-highly doped QW plus top side conduction) A low loss-low-voltage-high-frequency optical phase or intensity modulator device deposed on a substrate. The device has at least an input connecting waveguide core deposed on the substrate connecting the energy of an optical beam to and from an electro-active layer. The optical beam has one or more optical wavelengths around an operating optical wavelength .sub.op. The input connecting waveguide core becomes an input tapering waveguide core and enters below an electro-active layer. The optical beam energy is well-confined in the input tapering waveguide core before the tapering waveguide core enters below the electro-active layer. The optical beam energy is no longer well-confined in the input tapering waveguide core at some point after the tapering waveguide core enters below the electro-active layer. The refractive index of the tapering waveguide core is given by n.sub.ITWCo-z1. The width w.sub.ITWCo-z1 of the input tapering waveguide core after penetrating below the electro-active layer is reduced from a value approximately equal to or larger than half the wavelength in the material .sub.op/(2n.sub.ITWCo-z1) to a value smaller than half the wavelength in the material .sub.op/(2n n.sub.ITWCo-z1), so that w.sub.ITWCo-z1<.sub.op/(2n n.sub.ITWCo-z1) at some point under the electro-active layer. The refractive index n.sub.EC or the optical gain/absorption coefficient .sub.EC of at least part of the material in the electro-active layer can be altered by an applied electric field, an electric current, or either injection or depletion of carriers in the electro-active layer. The electro-active layer is either part of or in spatial proximity to an electro-active waveguide core. The electro-active waveguide core and electro-active waveguide cladding structure is in a medium-strongly guiding or very-strongly guiding regime such that the refractive index contrast of the waveguide core layer with both the top and the bottom waveguide cladding defined by: R.sub.cts=(n.sub.Co.sup.2n.sub.Cd.sup.2)/(n.sub.Co.sup.2+n.sub.Cd.sup.2) are both larger than about 0.2, where n.sub.Cd is the averaged material refractive index of either the top or the bottom waveguide cladding region, and n.sub.Co is the averaged material refractive index of the waveguide core region. The electro-active waveguide core thickness d.sub.CORE is in the ultra-thin, very-thin, medium-thin, or thin region such that d.sub.CORE<(2*.sub.op/n.sub.Co).

(595) A structure electrically connected to the electro-active layer comprises at least a first PN junction in which a first P-layer with P-dopant is vertically connected (vertical means in a direction perpendicular to the substrate plane; horizontal means in a direction parallel to the substrate plane) to a first N-layer with N-dopant, or a PqN junction in which a first P-layer with P-dopant is connected to a first q-layer with either N or P dopant or that is undoped (i.e. being an Intrinsic semiconductor material) and the q-layer is further connected to a first N-layer with N-dopant.

(596) At least one of the first P-layer, first N-layer, or the first q-layer contains at least one quantum well. The doping density at the quantum well is in the very-highly-doped, or ultra-highly-doped regime with a dopant density higher than about 1.510.sup.18/cm.sup.3 with either N doping or P doping.

(597) The first P-layer is electrically connected to a second P-layer with P-dopant of a second PN junction, referred to as the PN-changing PN junction (PNCPN). This second P-layer is electrically connected to a second N-layer with N-dopant of this second PN junction.