Electro-optic device

Abstract

An electro-optic device includes a first semiconductor layer including the rib-type waveguide, which includes a rib part and a first slab part, which extends in a first direction from the rib part; a dielectric layer, which is formed on the rib part; a second semiconductor layer, which extends in a second direction, which is opposite to the first direction, from an upper surface of the dielectric layer; a first high-concentration impurity region, which is formed in the first semiconductor layer to be in contact with the first slab part on the first direction side; and a second high-concentration impurity region, which is formed in a region of the second semiconductor layer on the second direction side. The second high-concentration impurity region is formed in a region other than a region overlapping the first semiconductor layer in a lamination direction.

Claims

1. An electro-optic device, comprising: a first semiconductor layer including a rib-type waveguide, which includes a rib part and a first slab part, which extends in a first direction from the rib part, the rib part having a first height, the first slab part having a second height lower than the first height; a dielectric layer, which is formed on the rib part; a second semiconductor layer, which extends in a second direction, which is opposite to the first direction, from an upper surface of the dielectric layer; a first high-concentration impurity region, which is formed in the first semiconductor layer to be in contact with the first slab part on the first direction side; and a second high-concentration impurity region, which is formed in a region of the second semiconductor layer on the second direction side other than a region overlapping the first semiconductor layer in a lamination direction wherein the first semiconductor layer further includes a second slab part, which extends in the second direction from the rib part, the second slab part having the second height, and any high-concentration impurity region being not in contact with the second slab part.

2. An electro-optic device according to claim 1, wherein the second slab part has a length in the second direction of 1 m or less.

3. An electro-optic device according to claim 1, wherein the first semiconductor layer is of a first conductivity type, and the second semiconductor layer is of a second conductivity type.

4. An electro-optic device according to claim 1, wherein the first semiconductor layer includes a first electrical contact portion which has a height that is equal to the first height of the rib part.

5. An electro-optic device according to claim 1, wherein the rib part has one of a Si.sub.1-xGe.sub.x layer and a strained Si.sub.1-xGe.sub.x layer formed to be at least partially buried therein.

6. An electro-optic device according to claim 5, wherein the Si.sub.1-xGe.sub.x layer comprises one of a laminated layer of a plurality of Si.sub.1-xGe.sub.x layers having different compositions and a composition-modulated Si.sub.1-xGe.sub.x layer having a composition changed in a thickness direction.

7. An electro-optic device according to claim 5, wherein a surface of the Si.sub.1-xGe.sub.x layer is subjected to one of oxidization and nitriding to form at least a part of the dielectric layer.

8. An electro-optic device according to claim 1, wherein the dielectric layer is formed of one of: one kind selected from the group consisting of silicon oxide, silicon nitride, hafnium oxide, zirconium oxide, and a rare earth oxide; and one of an alloy and a laminated film made of at least two kinds thereof.

9. An electro-optic device according to claim 1, wherein the first semiconductor layer and the second semiconductor layer each comprise at least one layer selected from the group consisting of polysilicon, amorphous silicon, strained silicon, single crystal silicon, and Si.sub.1-xGe.sub.x.

Description

BRIEF DESCRIPTION OF THE DRAWING

(1) FIG. 1 is a cross-sectional view for illustrating the structure of an electro-optic device having the PIN structure according to Non Patent Document 1.

(2) FIG. 2 is a cross-sectional view for illustrating the structure of an electro-optic device having the SIS structure according to Patent Document 1.

(3) FIG. 3 is a cross-sectional view for illustrating the structure of an electro-optic device according to Patent Document 3.

(4) FIG. 4 is a cross-sectional view for illustrating the structure of an electro-optic device according to a first embodiment of this invention.

(5) FIG. 5 is a cross-sectional view for illustrating the structure of an electro-optic device according to a second embodiment of this invention.

(6) FIG. 6 is a cross-sectional view for illustrating the structure of an electro-optic device according to a third embodiment of this invention.

(7) FIG. 7 is a cross-sectional view for illustrating the structure of an electro-optic device according to a fourth embodiment of this invention.

(8) FIG. 8 is a cross-sectional view for illustrating the structure of an electro-optic device according to a fifth embodiment of this invention.

(9) FIG. 9 is a cross-sectional view for illustrating the structure of an electro-optic device according to a sixth embodiment of this invention.

(10) FIG. 10 is a cross-sectional view for illustrating the structure of an electro-optic device according to a seventh embodiment of this invention.

(11) FIG. 11 is a cross-sectional view for illustrating the structure of an electro-optic device according to an eighth embodiment of this invention.

(12) FIG. 12A, FIG. 12B, and FIG. 12C are cross-sectional views for illustrating a process of producing the electro-optic device according to the third embodiment of this invention in order of steps.

(13) FIG. 13A, FIG. 13B, and FIG. 13C are cross-sectional views for illustrating steps following the steps illustrated in FIG. 12A to FIG. 12C.

(14) FIG. 14A, FIG. 14B, FIG. 14C, and FIG. 14D are cross-sectional views for illustrating steps following the steps illustrated in FIG. 13A to FIG. 13C.

(15) FIG. 15 is a graph for showing dependence of optical modulation characteristics of the electro-optic device according to this invention and the related electro-optic device on a frequency.

(16) FIG. 16 is a plan view for illustrating one configuration example of an optical intensity modulator of a Mach-Zehnder interferometer type, to which the electro-optic device according to this invention is applied.

MODES FOR EMBODYING THE INVENTION

(17) Now, embodiments of this invention are described in detail.

(18) An electro-optic device according to embodiments of this invention is specifically a silicon-based electro-optic modulator configured to convert a high-speed electrical signal, which is required in information processing and communication fields, into an optical signal at high speed. In order to achieve a sufficiently high-speed operation, this electro-optic device uses the capacitor structure made of silicon-insulator-silicon formed on a silicon-on-insulator (SOI) substrate.

(19) An electro-optic device according to a first embodiment of this invention has the structure in which a first semiconductor layer (first silicon semiconductor layer), which is doped to exhibit a first conductivity type, and at least a part of a second semiconductor layer (second silicon semiconductor layer), which is doped to exhibit a second conductivity type, are laminated. The first silicon semiconductor layer is processed into a rib waveguide shape, and the second silicon semiconductor layer is laminated at least on the rib part of the first silicon semiconductor layer. Between the first silicon semiconductor layer and the second silicon semiconductor layer that are laminated, a relatively thin dielectric layer is formed to form a so-called semiconductor-insulator-semiconductor (SIS) type junction. The phrase relatively thin dielectric layer means a dielectric layer of about 10 nm or less.

(20) In the SIS type junction, electrical signals from electrical terminals coupled to the first and second silicon semiconductor layers accumulate, remove, or invert free carriers on both sides of the relatively thin dielectric layer. In other words, the electro-optic device according to the first embodiment is an electro-optic device using modulation of a free carrier concentration sensed by an electric field of an optical signal passing through the rib part of the first silicon semiconductor layer. A rib waveguide shape is designed so that a region having a peak intensity of an optical signal field is arranged in a region in which the free carriers are accumulated, removed, or inverted on both sides of the dielectric layer.

(21) As described above, the first silicon semiconductor layer is processed into the rib waveguide shape. The rib waveguide shape is formed by the rib part and a slab part adjacent thereto. Then, the rib part serves as a rib waveguide. The first silicon semiconductor layer further includes a first electrical contact portion that is adjacent to the slab part, and is doped at a high concentration.

(22) Moreover, the above-mentioned dielectric layer is formed in an upper region of the rib part of a rib-type waveguide formed of the first silicon semiconductor layer. The second silicon semiconductor layer is laminated on the dielectric layer. The second silicon semiconductor layer has a region that extends laterally more than the dielectric layer. A part of the region that extends laterally is doped at a high concentration to form a second electrical contact portion.

(23) Further, the first electrical contact portion and the second electrical contact portion are formed respectively on the left and right with respect to a light propagation direction in the rib-type waveguide.

(24) In the first embodiment, the slab parts in the rib waveguide shape may be provided on both sides of the rib part. However, it is desired that a length (length in the direction of extending laterally from the rib part) of one slab part, that is, a slab part located on the side opposite to the first electrical contact portion be 1 m or less. Through reduction of the length of the slab part, the spread of electric field distribution at the time when a voltage is applied to the first silicon semiconductor layer and the second silicon semiconductor layer can be suppressed. As a result, a silicon-based electro-optic device that is operable with a lower voltage can be realized.

(25) Moreover, in the first embodiment, a height of the first electrical contact portion may be set to be equal to a height of the rib part of the rib waveguide shape. As a result, an electrode extraction resistance for connection from an optical phase modulation portion having the rib waveguide shape to an electrode through the slab part can be reduced, and easiness of production and production margin can be improved.

(26) Further, in the first embodiment, a Si.sub.1-xGe.sub.x layer formed of at least one layer exhibiting the first conductivity type may be formed to be at least partially buried in an upper region of the rib part of the first silicon semiconductor layer. As a result, the carrier plasma effect is improved, and the above-mentioned silicon-based electro-optic device with a smaller size and a lower voltage can be realized.

(27) Further, in the first embodiment, the dielectric layer may be formed of one kind selected from the group consisting of silicon oxide, silicon nitride, hafnium oxide, zirconium oxide, and a rare earth oxide, or an alloy or a laminated film made of at least two kinds thereof.

(28) In addition, in the first embodiment, each of the first and the second silicon semiconductor layers may be a layer formed of at least one layer selected from the group consisting of polycrystaline silicon, amorphous silicon, strained silicon, single crystal silicon, and Si.sub.1-xGe.sub.x.

(29) Further, in the first embodiment, the electrical terminals may be arranged to reduce an optical signal loss, and to provide a low series resistance.

(30) Now, before describing the particular exemplary structure of a silicon-based electro-optic device according to another embodiment of this invention, an overview of a modulation mechanism in silicon is described. Some electro-optic devices according to embodiments of this invention to be described later relate to the modulation structure, but the silicon-based electro-optic device according to this invention uses an electro-optic effect (free carrier plasma effect) described below.

(31) As described above, in pure silicon, no or very small, if at all, electro-optic effect occurs. Therefore, in a silicon-based optical modulator, the free carrier plasma effect and the thermo-optic effect are effective in an optical modulation operation. For a high-speed operation (Gb/sec or more), which is an object of the embodiments of this invention, the free carrier plasma effect is especially effective. The free carrier plasma effect is described as a first-order approximation value of the following relational expressions.

(32) $\begin{array}{cc}n=-\frac{e^2{}^2}{8{}^3{c}^3{\mathrm{.Math.}}_{0}n}\left(\frac{N_e}{m_e}+\frac{N_h}{m_h}\right)& \left[\mathrm{Expression}1\right]\\ k=-\frac{e^3{}^2}{8{}^3{c}^3{\mathrm{.Math.}}_{0}n}\left(\frac{N_e}{m_e^2{}_e}+\frac{N_h}{m_h^2{}_h}\right)& \left[\mathrm{Expression}2\right]\end{array}$
In the expressions 1 and 2, n and k represent a real part and an imaginary part of a change in refractive index of a silicon layer. Moreover, e represents an electric charge, represents an optical wavelength, .sub.0 represents a dielectric constant in vacuum, and n represents a refractive index of intrinsic semiconductor silicon. Moreover, m.sub.e represents an effective mass of electron carriers, and m.sub.h represents an effective mass of hole carriers. Moreover, .sub.c represents a mobility of the electron carriers, and .sub.h represents a mobility of the hole carriers. Further, N.sub.e represents a change in concentration of the electron carriers, and N.sub.h represents a change in concentration of the hole carriers.

(33) From the above-mentioned expressions, it can be understood that through reduction of the effective masses of the electron and hole carriers, that is, through application of the Si.sub.1-xGe.sub.x layer or the like, a larger change in refractive index can be obtained.

(34) However, the application of the Si.sub.1-xGe.sub.x layer increases an optical absorption loss. In other words, modulation efficiency and an optical loss are in a tradeoff relationship. In order to alleviate the tradeoff problem, a plurality of Si.sub.1-xGe.sub.x layers having different compositions may be laminated on the silicon layer, and a Si.sub.1-xGe.sub.x layer having a large Ge content may be arranged in an optical core region. Alternatively, the Si.sub.1-xGe.sub.x layer may be composition-modulated, and a Si.sub.1-xGe.sub.x layer having a large Ge content may be arranged in the optical core region. As a result, the tradeoff problem between the modulation efficiency and the optical loss can be alleviated.

(35) Moreover, experimental evaluations of the electro-optic effect in silicon have been performed, and it has been found that a change in refractive index with respect to a carrier density at wavelengths of 1,310 nm and 1,550 nm, which are used in an optical communication system, is a good match with the Drude expression. In addition, in an electro-optic modulator using the finding, an amount of phase change is defined by the following expression.

(36) $\begin{array}{cc}=\frac{2}{}{n}_{\mathrm{eff}}L& \left[\mathrm{Expression}3\right]\end{array}$
In the expression 3, L represents a length of an active layer along a light propagation direction of the electro-optic modulator.

(37) It can be said that the amount of phase change in the electro-optic device according to the embodiments of this invention is relatively large considering an amount of light absorption thereof. Further, the electro-optic device according to the embodiments to be described below basically exhibits features of a phase modulator.

(38) Now, embodiments of this invention are described in more detail with reference to the accompanying drawings. As described below, the electro-optic device according to each of the embodiments is an electro-optic phase modulator, in which silicon semiconductor-dielectric layer-silicon semiconductor (SIS) junction is formed on the SOI substrate, and which uses the free carrier plasma effect.

(39) As described above, the electro-optic devices having the PIN structure and the SIS structure illustrated in FIG. 1 and FIG. 2 have disadvantages in that there is a small overlap between an optical field and a region in which the carrier density is modulated, and that sizes of the electro-optic devices are increased. Moreover, with the electro-optic device having the SIS structure illustrated in FIG. 3, electrode layers respectively connected to the first silicon semiconductor layer of the first conductivity type and the second silicon semiconductor layer of the second conductivity type need to be arranged at positions spaced apart from each other so that the electrode layers do not interfere with each other. Therefore, there is a problem in that an extraction resistance from the first silicon semiconductor layer to the electrode layer is high. Meanwhile, in order to at least partially solve this problem, when the electrode layer connected to the second silicon semiconductor layer is brought closer to the rib-type waveguide, there arises a problem in that the optical loss is increased.

(40) In order to improve at least some of the problems (disadvantages and challenges) inherent in those electro-optic devices illustrated in FIG. 1 to FIG. 3, the electro-optic device according to the first embodiment of this invention adopts the structure illustrated in FIG. 4.

(41) The structure illustrated in FIG. 4 includes the structure in which a first silicon semiconductor layer 9, which is doped to exhibit a first conductivity type, and at least a part of a second silicon semiconductor layer 10, which is doped to exhibit a second conductivity type, are laminated. Then, between those laminated semiconductor layers, a relatively thin dielectric layer (insulating film) 12 is formed to form semiconductor-insulator-semiconductor (SIS) type junction. In this SIS type structure portion, electrical signals from electrode wirings 7-1 and 7-2, which are electrical terminals coupled to the first and second silicon semiconductor layers 9 and 10, accumulate, remove, or invert the free carriers on both sides of the dielectric layer 12. As a result, a free carrier concentration felt by an electric field of an optical signal passing through the SIS type structure portion is modulated.

(42) The first silicon semiconductor layer 9 is processed into the rib waveguide shape. In other words, the first silicon semiconductor layer 9 includes a rib part 141, and first and second slab parts 142 and 143 extending on both sides of the rib part 141. The rib part 141 extends in the front-and-back direction of FIG. 4, and serves as an optical waveguide (rib-type waveguide). The first slab part 142 extends (or stretches) in the left direction (first direction) of FIG. 4, and the second slab part 143 extends (stretches) in the right direction (second direction) of FIG. 4. The second slab part 143 may be very short as compared to the first slab part 142 with respect to the left-and-right direction of FIG. 4, or no second slab part 143 may be provided.

(43) The first silicon semiconductor layer 9 further includes a p-type region 4, which is doped at a high concentration, adjacent to the first slab part 142. A part of the p-type region 4 serves as a first electrical contact portion 6-1.

(44) The dielectric layer 12 is formed on an upper surface of the rib part 141 formed of the first silicon semiconductor layer 9. The second silicon semiconductor layer 10 is laminated to extend laterally on and beyond an upper surface of the dielectric layer 12. The second silicon semiconductor layer 10 is formed to extend (stretch) mainly in the second direction. A portion (region toward the second direction) of the second silicon semiconductor layer 10 includes an n-type region 11, which is doped with impurities at a high concentration. A part of the n-type region 11 serves as a second electrical contact portion 6-2.

(45) As can be readily understood from FIG. 4, the first electrical contact portion 6-1 and the second electrical contact portion 6-2 are formed respectively on the left and the right with respect to the light propagation direction (approximately in the center of FIG. 4 in the front-and-back direction of the drawing sheet) in the rib-type waveguide. In other words, the first electrical contact portion 6-1 is located in the first direction with respect to the rib-type waveguide, and the second electrical contact portion 6-2 is located in the second direction with respect to the rib-type waveguide. Moreover, the second electrical contact portion 6-2 (n-type region 11) is formed in a region other than a region overlapping the first silicon semiconductor layer 9 in a lamination direction. In other words, the second electrical contact portion 6-2 (n-type region 11) is formed above the region in which the first silicon semiconductor layer 9 is not formed.

(46) In this structure, the first silicon semiconductor layer 9 does not exist below the n-type region 11 in the lamination direction. Therefore, even when annealing is performed at a high temperature in forming the n-type region 11, there is no or small effect on the first silicon semiconductor layer 9 due to undesirable diffusion of the impurities. It can be said at least that there is substantially no effect on the p-type region 4. Therefore, in forming the n-type region 11, annealing can be performed at a sufficiently high temperature. As a result, a defect in crystal lattice that is generated in the n-type region 11 accompanying the introduction of the impurities can be sufficiently repaired. In addition, an electrical resistance (electrode extraction resistance) of a power supply path including the p-type region 4 is not substantially increased. As a result, satisfactory characteristics as the electro-optic device can be obtained.

(47) Moreover, in this structure, the first electrical contact portion 6-1 and the second electrical contact portion 6-2 are arranged on the left and the right of the rib-type waveguide, respectively. Therefore, with this structure, as compared to the electro-optic device having the SIS type junction structure illustrated in FIG. 3, the first electrical contact portion 6-1 may be brought remarkably closer to the rib-type waveguide. As a result, the electrode extraction resistance can be reduced, and the high-speed operation can be achieved.

(48) FIG. 5 is a view for illustrating the structure of an electro-optic device according to a second embodiment of this invention. Though it is hard to see in FIG. 5, in this electro-optic device, a length of the second slab part 143 in the rib waveguide shape is set to 1 m or less. There may be no second slab part 143, but in order to make an intensity distribution (cross-sectional shape) of light passing through the rib-type waveguide closer to left and right symmetry, a length of about 1 m is required. Meanwhile, the existence of the slab part 143 spreads an electric field generated by a voltage applied to the first silicon semiconductor layer 9 to deteriorate modulation characteristics. Therefore, as a compromise for the tradeoff, it is preferred to set the length of the slab part 143 to 1 m or less. In this manner, coupling with optical fibers or the like to an input side and an output side of the electro-optic device may be performed satisfactorily, and the spread of the electric field distribution by the applied voltage may be suppressed. As a result, the electro-optic device that is operable with a lower voltage can be realized.

(49) FIG. 6 is a view for illustrating the structure of an electro-optic device according to a fourth third embodiment of this invention. In this electro-optic device, in the first silicon semiconductor layer 9, a part (including the electrical contact portion 6-1) of the p-type region 4 has a thickness that is equal to that of the rib part 141. In other words, the first electrical contact portion 6-1 has a height that is equal to that of the rib part 141 in the rib waveguide shape. As a result, the electrode wiring 7-1 may be shortened to reduce the electrode extraction resistance, that is, a series resistance component, to thereby reduce the RC time constant. Therefore, this electro-optic device is capable of operation at higher speed while maintaining high optical modulation efficiency as compared to this electro-optic device of FIG. 5.

(50) Further, a thickness of the second silicon semiconductor layer 10 may be adjusted to improve a shape and width of an overlapping range between the optical field and the region in which the carrier density is modulated. As a result, the electro-optic device may be downsized.

(51) When a region in which a doping density is increased and the optical field overlap, the optical absorption loss is increased. In all of the above-mentioned embodiments, the optical waveguide having the rib (or ridge) shape is adopted, and the region in which the doping density is increased is provided in a region that is in contact with the first slab part 142. As a result, the electro-optic device that has a small optical loss and a small RC time constant, and operates at high speed can be obtained. Moreover, with this structure, excitation of higher-order light propagation mode, which occurs when the first electrical contact portion 6-1 having the height that is equal to that of the rib part is brought closer to the rib part 141 side, may be suppressed.

(52) FIG. 7 and FIG. 8 are views for illustrating the structures of electro-optic devices according to fourth and fifth embodiments of this invention, respectively. The electro-optic device illustrated in FIG. 7 includes a Si.sub.1-xGe.sub.x (x=0.01 to 0.9) layer 13 in an upper region of the rib part 141. Similarly, the electro-optic device illustrated in FIG. 8 includes a Si.sub.1-xGe.sub.x laminated layer 14, in which two or more Si.sub.1-xGe.sub.x (x=0.01 to 0.9) layers (having different composition ratios) are laminated, in an upper region of the rib part 141. The Si.sub.1-xGe.sub.x layer 13 or the Si.sub.1-xGe.sub.x laminated layer 14 is formed to be at least partially buried in the rib part 141.

(53) According to the fourth and fifth embodiments, provision of the Si.sub.1-xGe.sub.x layer 13 or the Si.sub.1-xGe.sub.x laminated layer 14 may further increase the modulation efficiency. In order to increase the modulation efficiency, it is desired that, with respect to a thickness W of the semiconductor layers in which the free carriers are accumulated, removed, or inverted on both sides of the dielectric layer 12, each of the Si.sub.1-xGe.sub.x (x=0.01 to 0.9) layer 13 and the Si.sub.1-xGe.sub.x laminated layer 14 have a thickness of 2 W or less. The modulation efficiency is improved when the thickness of the Si.sub.1-xGe.sub.x layer 13 or the Si.sub.1-xGe.sub.x laminated layer 14 is 2 W or more, but it is desired that the thickness be 2 W or less for higher effectiveness.

(54) FIG. 9 is a view for illustrating the structure of an electro-optic device according to a sixth embodiment of this invention. In the sixth embodiment, a composition-modulated Si.sub.1-xGe.sub.x (x=0.01 to 0.9) layer 15, in which a composition is changed (which is composition-modulated) in a thickness direction, is formed in an upper region of the rib part 141. This composition-modulated Si.sub.1-xGe.sub.x layer 15 is composite-modulated in the thickness direction so that a larger Ge content is located on the lower layer side. With arrangement of the layer having a large Ge content in the optical core region, the optical modulation efficiency may be improved, and an increase in optical loss may be reduced.

(55) FIG. 10 is a view for illustrating the structure of an electro-optic device according to a seventh embodiment of this invention. In the seventh embodiment, a strained Si.sub.1-xGe.sub.x (x=0.01 to 0.9) layer 16, to which lattice strain is introduced, is formed in an upper region of the rib part 141. With the introduction of the lattice strain to the strained Si.sub.1-xGe.sub.x layer 16, effective masses of the carriers are reduced, and higher optical modulation efficiency can be obtained.

(56) FIG. 11 is a view for illustrating the structure of an electro-optic device according to an eighth embodiment of this invention. In the eighth embodiment, an oxidized or nitrided Si.sub.1-xGe.sub.x layer 17 is formed on an upper surface of the Si.sub.1-xGe.sub.x (x=0.01 to 0.9) layer 13. This oxidized or nitrided Si.sub.1-xGe.sub.x layer 17 is formed by, after forming the Si.sub.1-xGe.sub.x layer 13, oxidizing or nitriding a surface of the Si.sub.1-xGe.sub.x layer 13. This oxidized or nitrided Si.sub.1-xGe.sub.x layer 17 may be configured to form a part or all of the dielectric layer 12. As a result, a defect density at an interface between the Si.sub.1-xGe.sub.x (x=0.01 to 0.9) layer 13 and the dielectric layer 12 is reduced, and more stable optical phase modulation operation is obtained.

(57) A maximum thickness W of a depletion layer generated in the electro-optic device according to each of the above-mentioned embodiments is given by the following mathematical expression in a thermal equilibrium state.

(58) $\begin{array}{cc}W=2\sqrt{\frac{{}_s\mathrm{kT}\ln \left(\frac{N_c}{n_i}\right)}{e^2{N}_c}}& \left[\mathrm{Expression}4\right]\end{array}$
In the expression 4, .sub.s represents a dielectric constant of a semiconductor layer, k represents a Boltzmann constant, N.sub.c represents a carrier density, n.sub.i represents an intrinsic carrier concentration, and e represents an amount of electric charges. For example, when N.sub.c is 1,017/cm.sup.3, the maximum thickness of the depletion layer is about 0.1 m. Then, with an increase in carrier density, the thickness of the depletion layer, that is, the thickness of the region in which the carrier density is modulated is reduced.

(59) In each of the above-mentioned embodiments, each of the first and second silicon semiconductor layers 9 and 10 is formed of at least one layer selected from the group consisting of polysilicon, amorphous silicon, strained silicon, single crystal silicon, and Si.sub.1-xGe.sub.x.

(60) Next, referring to FIG. 12 to FIG. 14, a method of producing the electro-optic device according to the third embodiment illustrated in FIG. 6 is described.

(61) To describe briefly, in this method, the first silicon semiconductor layer exhibiting the first conductivity type is first processed into the rib waveguide shape. Next, a region adjacent to a slab part of the rib waveguide shape of the first silicon semiconductor layer is doped at a high concentration to form the first electrical contact portion. Next, the relatively thin dielectric layer is formed on the upper surface of the rib part of the rib-type waveguide. Thereafter, the second silicon semiconductor layer that is doped to exhibit the second conductivity type is laminated. Then, a part of the second silicon semiconductor layer is doped at a high concentration to form the second electrical contact portion. Now, details are described.

(62) FIG. 12A is a cross-sectional view of an SOI substrate used to form the electro-optic device. This SOI substrate is formed by laminating a Si layer (9) of from about 100 nm to about 1,000 nm on a buried oxide layer 2, which is formed to be buried in the support substrate 3. In order to reduce the optical loss, a thickness of the buried oxide layer may be 1,000 nm or more. The Si layer on this buried oxide layer 2 may be doped to exhibit the first conductivity type in advance for use as the first silicon semiconductor layer 9. Alternatively, after a surface layer of the Si layer is doped with impurities (phosphorus (P) or boron (B)) by ion implantation or other such methods, heat treatment may be performed so that the Si layer becomes the first silicon semiconductor layer 9 exhibiting the first conductivity type.

(63) Next, as illustrated in FIG. 12B, the laminate structure of an oxide mask 18 and a SiN.sub.x hard mask layer 19 is formed on an upper surface of the first silicon semiconductor layer 9. Then, the oxide mask 18 and the SiN.sub.x hard mask layer 19 are processed to form a mask for forming the rib waveguide shape in the first silicon semiconductor layer 9. In other words, the oxide mask 18 and the SiN.sub.x hard mask layer 19 are patterned into a predetermined shape using UV lithography, a dry etching method, or other such methods.

(64) Next, as illustrated in FIG. 12C, using the oxide mask 18 and the SiN.sub.x hard mask layer 19 as a mask, the first silicon semiconductor layer 9 is patterned to form the rib waveguide shape. As a result, the rib part 141 and the slab parts 142 and 143 are formed. In the third embodiment, at this time, a portion to be the p-type region 4 is left without being etched. Thereafter, an unnecessary region (portion on the right side of FIG. 12C) of the first silicon semiconductor layer 9 is removed to form the second slab part 143. This can be realized by etching using another mask (not shown) or other such methods.

(65) Next, as illustrated in FIG. 13A, a region adjacent to the first slab part 142 of the first silicon semiconductor layer 9 is doped with impurities (for example, B) at a high concentration using the ion implantation method or other such methods to form the p-type region 4. This region includes a region having a height that is equal to that of the rib part 141.

(66) Next, as illustrated in FIG. 13B, an oxide cladding 8 is laminated, and an upper surface of the oxide cladding 8 is flattened by the chemical mechanical polishing (CMP) method.

(67) Next, as illustrated in FIG. 13C, the SiN.sub.x hard mask layer 19 and the oxide mask 18 are removed by treatment with heat phosphoric acid and diluted hydrofluoric acid or other such methods to expose an upper surface of the rib part 141. Subsequently, the relatively thin dielectric layer 12 of from about 5 mm to about 10 nm is formed on the exposed upper surface of the rib part.

(68) Next, as illustrated in FIG. 14A, an n-doped polysilicon layer is laminated as the second silicon semiconductor layer 10. Then, the second silicon semiconductor layer 10 is patterned by the dry etching method or other such methods into a shape having a width that is required for forming the second electrical contact portion 6-2.

(69) Next, as illustrated in FIG. 14B, a part of the second silicon semiconductor layer 10 is doped with impurities at a high concentration using an ion implantation method or other such methods to form the n-type region 11, a part of which is to form the second electrical contact portion 6-2.

(70) Next, as illustrated in FIG. 14C, the oxide cladding 8 is further laminated to about 1 m. Then, contact holes 20 passing through the oxide cladding 8 to reach the p-type region 4 and the n-type region 11, respectively, are formed by the dry etching method or other such methods. Portions of the p-type region 4 and the n-type region 11 that are exposed in the contact holes 20 become the first and second electrical contact portions 6-1 and 6-2, respectively.

(71) Next, as illustrated in FIG. 14D, a metal layer, which is made of a laminated film of Ti/TiN/Al(Cu), Ti/TiN/W, or the like to fill the contact holes 20, is formed by a spattering method, a CVD method, or other such methods. Then, the formed metal layer is patterned by reactive etching to form the electrode wirings 7-1 and 7-2.

(72) Thereafter, wiring layers (not shown), which are respectively connected to the electrode wirings 7-1 and 7-2, are formed. Those wiring layers are used for connection with a drive circuit (not shown).

(73) As described above, the electro-optic device is completed.

(74) FIG. 15 is a graph for showing frequency characteristics 151 of the optical modulation efficiency (normalized responsivity) of the electro-optic device according to the third embodiment. For comparison, in FIG. 15, frequency characteristics 152 of the electro-optic device of FIG. 3 are also shown. With the device according to the third embodiment, the first and second electrical contact portions 6-1 and 6-2 can be formed to have large film thicknesses as compared to the device of FIG. 3. In this manner, a series resistance connected to each of the first and second silicon semiconductor layers 9 and 10 can be reduced. As a result, the RC time constant of the device is reduced, and operation in a frequency band of about 30 GHz can be achieved.

(75) In addition to the above, in order to improve the frequency characteristics, the mobilities and lifetime of the carriers are very important. In particular, the mobilities of the carriers in the polysilicon layer may become a problem in high-speed operation. Therefore, when a polysilicon layer is used as the second silicon semiconductor layer 10, it is effective to increase a particle diameter by recrystallization by annealing, to thereby improve carrier mobilities. Alternatively, it is effective to form the second silicon semiconductor layer using an epitaxial lateral overgrowth (ELO) method or other such methods, to thereby improve crystal quality.

(76) FIG. 16 is a view for illustrating one configuration example of an optical intensity modulator of a Mach-Zehnder interferometer type, to which the electro-optic device according to any one of the embodiments of this invention is applied. In this example, as each of a first arm 21 and a second arm 22, which are arranged in parallel, the electro-optic device according to any one of the embodiments of this invention is used. Moreover, this optical intensity modulator includes three driving electrode pads 23, which are arranged along the first arm 21 and the second arm 22.

(77) The center electrode pad of the three electrode pads 23 is connected in common to one of the electrode wirings 7-1 and 7-2 of the electro-optic devices forming the first arm 21 and the second arm 22. The remaining two electrode pads are connected to others of the electrode wirings 7-1 and 7-2 of the electro-optic devices forming the first arm 21 and the second arm 22, respectively.

(78) Further, this optical intensity modulator includes an optical divider 24 coupled to the input side of the first arm 21 and the second arm 22, and an optical coupler 25 coupled to the output side of the first arm 21 and the second arm 22. The optical divider 24 divides an input optical signal into two optical signals so that the optical signals have equal powers, and outputs the two optical signals obtained by the division to the first arm 21 and the second arm 22. The first arm 21 and the second arm 22 perform phase modulation of the optical signals passing therethrough depending on a voltage applied to the electrode pads 23. The optical coupler 25 couples the optical signals that have undergone different phase modulation in the first arm 21 and the second arm 22. In other words, the optical coupler 25 causes phase interference between the optical signals that have passed through the first arm 21 and the second arm 22, to thereby convert the optical signals into a light intensity modulation signal.

(79) For example, it is assumed that a negative voltage is applied to the center electrode pad of the three electrode pads 23, and that a positive voltage and a negative voltage are applied to one and another one of the remaining two electrode pads, respectively. In this case, for example, the carriers are accumulated on both sides of the dielectric layer 12 of the electro-optic device forming the first arm 21 (carrier accumulation mode), and carriers are removed on both sides of the dielectric layer 12 forming the second arm 22 (carrier removal mode). In the carrier accumulation mode, a refractive index sensed by the optical signal field in the electro-optic device becomes smaller. Meanwhile, in the carrier removal (depletion) mode, the refractive index sensed by the optical signal field becomes larger. As a result, a phase difference between the optical signals passing through the arms 21 and 22 is maximized. Moreover, when a negative voltage is applied to all the electrode pads 23, a phase difference between the optical signals passing through the arms 21 and 22 is minimized (ideally becomes zero). The optical coupler 25 multiplexes the optical signals passing through the arms 21 and 22 on the output side. As a result, a light intensity is minimized when the phase difference is maximized, and the light intensity is maximized when the phase difference is minimized. In this manner, the optical intensity modulator achieves light intensity modulation. With the optical intensity modulator using the electro-optic device according to the third embodiment, it has been confirmed that an optical signal of 40 Gbps or more can be transmitted.

(80) The electro-optic device according to this invention is also applicable to, for example, an optical modulator, in which a plurality of optical intensity modulators of the Mach-Zehnder interferometer type are connected in parallel or series to achieve a high transfer rate, or to a matrix optical switch.

(81) This invention has been described above in detail with the use of some embodiments. However, this invention is not limited to the above-mentioned embodiments, and various modifications and alterations may be made thereto.

(82) This application claims priority from Japanese Patent Application No. 2015-071022, filed on Mar. 31, 2015, the entire disclosure of which is incorporated herein by reference.

REFERENCE SIGNS LIST

(83) 1 intrinsic semiconductor silicon layer

(84) 2 buried oxide film

(85) 3 support substrate

(86) 4 p-type region (p+-doped semiconductor silicon layer)

(87) 5 n-type region (n+-doped semiconductor silicon layer)

(88) 6-1 first electrical contact portion

(89) 6-2 second electrical contact portion

(90) 7-1, 7-2 electrode wiring (metal electrode)

(91) 8 oxide cladding

(92) 9 first silicon semiconductor layer (p-doped semiconductor silicon)

(93) 10 second silicon semiconductor layer (n-doped polysilicon)

(94) 11 n-type region (n+-doped polysilicon layer)

(95) 12 dielectric layer

(96) 13 Si.sub.1-xGe.sub.x (x=0.01 to 0.9) layer

(97) 14 Si.sub.1-xGe.sub.x laminated layer

(98) 15 composition-modulated Si.sub.1-xGe.sub.x (x=0.01 to 0.9) layer

(99) 16 strained Si.sub.1-xGe.sub.x (x=0.01 to 0.9) layer

(100) 17 oxidized or nitrided Si.sub.1-xGe.sub.x (x=0.01 to 0.9) layer

(101) 18 oxide mask

(102) 19 SiN.sub.x hard mask layer

(103) 20 contact hole

(104) 21 first arm

(105) 22 second arm

(106) 23 driving electrode pad

(107) 24 optical divider

(108) 25 optical coupler

(109) 101 intrinsic semiconductor region

(110) 102, 103 slab region

(111) 105 body region

(112) 106 (intrinsic semiconductor) silicon layer

(113) 107 gate region

(114) 108 polysilicon layer

(115) 110 rib-type waveguide

(116) 111 slab part

(117) 141 rib part

(118) 142 first slab part

(119) 143 second slab part

(120) 151 frequency characteristics of electro-optic device according to embodiments of this invention

(121) 152 frequency characteristics of related electro-optic device