JP2016224315A - Method for driving waveguide type optical element, and waveguide type optical element used in drive method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To effectively prevent the acceleration of a drift phenomenon caused by a high voltage applied to a bias electrode or discharge breakdown of the bias electrode in a waveguide type optical element having the bias electrode.SOLUTION: The waveguide type optical element comprises: a substrate 100 having an electro-optical effect; two optical waveguides 104, 106 formed on the substrate; control electrodes 152, 154, 156 for applying electric fields of mutually opposite directions along the orientation of spontaneous polarization of the substrate to each of the two optical waveguides, and generating a difference in refractive index between the two optical waveguides. The placement of the control electrodes and/or the applied voltage to the control electrodes is (are) set so that the electric field intensity of an electric field that is in an opposite direction to the orientation of spontaneous polarization made to occur in the substrate by the control electrodes is smaller than the electric field intensity of an electric field that is in the same direction as the orientation of spontaneous polarization.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a method for driving a waveguide-type optical element including an optical waveguide and an electrode for controlling a light wave propagating through the optical waveguide, and more particularly, to a semiconductor device including a bias electrode for compensating for a so-called drift voltage. The present invention relates to a method for driving the bias electrode of a waveguide type optical element, and a waveguide type optical element used in the driving method.

  In recent years, in the fields of optical communication and optical measurement, a waveguide type optical element such as an optical modulator in which an optical waveguide is formed on a substrate having an electro-optic effect is often used. A waveguide type optical element generally includes an electrode for controlling a light wave propagating in the optical waveguide together with the optical waveguide.

As such a waveguide type optical element, for example, a Mach-Zehnder type optical modulator using a ferroelectric crystal lithium niobate (LiNbO ₃ ) (also referred to as “LN”) as a substrate is widely used. The Mach-Zehnder type optical modulator includes an incident waveguide for introducing light from the outside, a branching unit for propagating the light introduced by the incident waveguide in two paths, and a branching unit after the branching unit. Mach-Zehnder type optical waveguide composed of two parallel waveguides for propagating each of the generated light and an output waveguide for combining the light propagated through the two parallel waveguides and outputting them to the outside Is provided. The Mach-Zehnder type optical modulator includes an electrode for changing and controlling the phase of the light wave propagating in the parallel waveguide by applying the voltage and utilizing the electro-optic effect. In general, the electrodes include an RF (high frequency) signal electrode (hereinafter referred to as an “RF electrode”) formed on or near the parallel waveguide, and a ground electrode disposed apart from the RF electrode. It consists of

  In a Mach-Zehnder type optical modulator using LN as a substrate, a bias voltage necessary for obtaining a desired optical output characteristic is shifted due to a DC drift phenomenon or a temperature drift phenomenon. For example, light output from a Mach-Zehnder type modulator is used. Distortion or the like occurs in the modulation waveform, and a change in modulation characteristics (for example, deterioration in waveform quality) may occur. The main cause of DC drift is that space charges (carriers) in the substrate of the Mach-Zehnder optical modulator or in the film body between the electrode and the substrate move in a direction that cancels the external electric field for adjusting the bias. As the number of carriers increases, the effect of canceling the external electric field is high, and the DC drift phenomenon tends to be firm. The main cause of the temperature drift is a change in the external electric field due to electric charges (pyroelectricity) generated by the temperature change, a deformation of the Mach-Zehnder interferometer due to the distortion of the substrate due to the temperature change, and a change in the optical path difference.

  In order to prevent the modulation characteristic from changing due to such a drift phenomenon, in addition to the RF electrode and the ground electrode for applying a high-frequency signal voltage, a bias electrode is formed along the parallel waveguide, and the bias state A method is known in which a bias voltage shift amount (drift voltage) due to the drift phenomenon is continuously compensated by applying an appropriate voltage to the bias electrode while monitoring the above (Patent Document 1).

  That is, the bias voltage shift amount is adjusted by applying a voltage to the bias electrode to generate different electric fields in the two parallel waveguides and generating a difference in refractive index between the two parallel waveguides. It is possible to drive the modulator using both the RF electrode and the bias electrode and also using the bias tee circuit. It is also possible to secure the length of the action part of the bias electrode of 40 mm or more. In the case of this configuration, there is an advantage that a small electric field is required to generate a desired refractive index difference between the parallel waveguides. However, it is desirable to separately provide an RF signal and a bias signal in order to prevent deterioration of the optical signal quality when the signal mark rate changes and to reduce the influence of external high frequency noise on the optical signal quality.

  Currently, the upper limit of the footprint of the optical modulator is defined by the industry standards of optical transceivers and privately agreed specifications, and reduction of the power for driving the modulator at a high frequency is always required. In this case, for example, the RF electrode and the bias electrode are individually formed along a parallel waveguide having a length limited by the size of the waveguide type optical element. In a Mach-Zehnder type optical modulator, as a method for reducing power consumption, that is, reducing the half-wave voltage (Vπ), the basic solution is to increase the RF electrode length. The length may be shortened. In such a case, the electric field required to generate the desired refractive index difference between the parallel waveguides is increased, and thus the voltage applied to the bias electrode is also increased. In some cases, the action length of the bias electrode may be 10 mm or less, and in some cases, 5 mm or less. The electric field required for adjusting the bias is many times larger than when the RF electrode and the bias electrode are shared.

  The gap between the bias electrodes can be made smaller than the gap between the RF electrodes, and thus it is possible to avoid an increase in the required applied voltage for bias adjustment. In the case of an optical modulator using LN as a substrate, the electrode spacing varies depending on the presence or absence of a buffer layer, the crystal orientation of the substrate, the shape of the optical waveguide, the driving method (differential, one-side driving), and the like. In the RF electrode, reduction of high-frequency signal loss is given priority, and it is necessary to use a coplanar line having a large electrode interval. Generally, electrodes having an electrode interval of 20 μm or more are used. On the other hand, the bias electrode does not need to consider the attenuation of the high-frequency signal. The distance between the bias electrodes is designed for avoiding an increase in light loss, for manufacturing reasons, for the upper limit of the output of the bias circuit, etc., but in the case of a modulator with a buffer layer, 2 μm to 15 μm, no buffer layer In the case of a modulator, the distance between the electrodes can be reduced to 10 μm to 20 μm, and the required voltage can be reduced. However, the electric field required to generate the desired refractive index difference between the parallel waveguides has not been reduced.

In the LN modulator, a considerably large electric field is required to adjust the bias with a bias electrode having a length of 5 mm for the action portion. For example, each sub-Mach-Zehnder modulator of the QPSK modulator uses the difference in effective length between two optical waveguides forming a Mach-Zehnder interferometer to adjust the bias to a position called a null bias point. Must be. The necessary refractive index difference Δn (= 1.55 μm / 5 mm / 2) in the 1.55 μm wavelength band is 1.55 × 10 ⁻⁴ . When the bias state is adjusted by differential control, it is only necessary to cause each waveguide to have a refractive index change of half the amount, that is, ± 0.775 × 10 ⁻⁴ . In order to generate the refractive index difference with a modulator using an LN crystal (Pockels constant r ₃₃ = 30.8 × 10 ⁻¹² m / V, extraordinary refractive index n _e ≈2.2), The magnitude of the effective electric field (E _eff = 0.775 × 10 ⁻⁴ / r ₃₃ / (n _e ) ³ ) needs to be about 0.5 V / μm. This electric field is not so much that the components of the LN modulator will be destroyed immediately, but is very close to an electric field in which the phenomenon of destruction and deterioration of the members of the modulator becomes remarkable. In addition, the LN modulator has an essential problem due to a material called DC drift and a device structure. The bias voltage does not work with time, and the voltage necessary to maintain a desired bias state is not obtained. There is a universal phenomenon that grows with time.

  As a result, under use conditions where the electric field (voltage between electrodes / distance between electrodes) exceeds 1 V / μm, carriers (charges) are injected from the bias electrode to the LN substrate, and the drift phenomenon is accelerated. The occurrence of corona discharge or spark discharge between the bias electrodes may cause a phenomenon that the bias electrodes are destroyed. When an insulator (dielectric) is present between the bias electrodes of the Mach-Zehnder modulator, a corona discharge or a spark discharge forms a dendritic or needle-like discharge path along the surface of the insulator. Such a discharge is called a creeping discharge, but there is a case where there is no alteration of the insulator surface during the discharge (generally called a flashover), and dielectric breakdown occurs in a part of the surface layer portion of the dielectric. Generation of cracks, changes in polarization state, changes in electrical characteristics (electrical resistance, increase in low-frequency dielectric constant), etc., along the discharge path, such as In, Sn contained in metal or solder of electrode materials Residual monomers and solvents generated as outgases from metals with high vapor pressure, adhesives, conductive resins, sealants, solder lux, antistatic agents, plasticizers, flame retardants, antioxidants and their reactants In some cases, metallic and organic deposits and deposits are baked, and the discharge path is fixed and grows (called tracking). When the bias electrode is short, tracking is likely to occur because a high voltage is always applied to the bias electrode.

JP-A-5-224163

  From the above background, in a waveguide type optical element having a bias electrode, it is possible to effectively prevent acceleration of a drift phenomenon that occurs when a high voltage is applied to the bias electrode and discharge breakdown of the bias electrode. Realization of a configuration that can be achieved is desired.

One aspect of the present invention includes a substrate having an electro-optic effect, two optical waveguides formed on the substrate, and the two optical waveguides in directions opposite to each other along the spontaneous polarization direction of the substrate. A control electrode for applying an electric field to generate a refractive index difference between the two optical waveguides, and the control electrode includes a reference electrode disposed between the two optical waveguides, Each of the two optical waveguides with respect to the reference electrode, and two working electrodes arranged in parallel or substantially in parallel with the reference electrode, and a distance between the reference electrode and the one working electrode is The waveguide-type optical element is disposed so as to be larger than the distance between the reference electrode and the other working electrode.
Another aspect of the present invention is directed to a substrate having an electro-optic effect, two optical waveguides formed on the substrate, and the two optical waveguides in directions opposite to each other along a spontaneous polarization direction of the substrate. A control electrode for applying an electric field to generate a refractive index difference between the two optical waveguides, and the control electrode includes a reference electrode disposed between the two optical waveguides, Each of the two optical waveguides sandwiched between the two optical waveguides and arranged in parallel or substantially in parallel with the reference electrode, and between the one operating electrode and the substrate A waveguide-type optical element in which a non-conductive intermediate layer is disposed and no non-conductive intermediate layer is disposed between the other working electrode and the substrate.
Another aspect of the present invention is directed to a substrate having an electro-optic effect, two optical waveguides formed on the substrate, and the two optical waveguides in directions opposite to each other along a spontaneous polarization direction of the substrate. A control electrode for applying an electric field to generate a refractive index difference between the two optical waveguides, and the control electrode includes a reference electrode disposed between the two optical waveguides, And two working electrodes arranged in parallel or substantially in parallel with the reference electrode with the two optical waveguides interposed therebetween, and the two working electrodes and the substrate are non-conductive A waveguide type in which an intermediate layer is disposed, and the intermediate layer between one working electrode and the substrate is thicker than the intermediate layer between another working electrode and the substrate It is an optical element.
According to another aspect of the present invention, the substrate is made of lithium niobate, the two optical waveguides are two parallel waveguides constituting a Mach-Zehnder type optical waveguide, and the control electrode compensates for a drift voltage. This is a bias electrode.
Still another aspect of the present invention is the above-described method for driving a waveguide optical device, wherein a first voltage is applied between the one working electrode and the reference electrode, A step of generating an electric field in a direction opposite to the spontaneous polarization orientation in the substrate between the working electrode and the reference electrode; and applying a second voltage between the other working electrode and the reference electrode Generating an electric field in the same direction as the spontaneous polarization direction in the substrate between the other working electrode and the reference electrode, and the first and second voltages are: The electric field strength of the electric field in the opposite direction to the spontaneous polarization azimuth generated in the substrate is set to be smaller than the electric field strength of the electric field in the same direction as the spontaneous polarization azimuth.

It is a figure explaining the structure and drive method of the waveguide type optical element which concern on the 1st Embodiment of this invention. It is a figure explaining the structure and drive method of the waveguide type optical element which concern on the 1st Embodiment of this invention. It is a figure explaining the structure and drive method of the waveguide type optical element which concern on the 1st Embodiment of this invention. FIG. 4 is an AA cross-sectional view of the waveguide type optical element shown in FIG. 3.

  Embodiments of the present invention will be described below with reference to the drawings. The embodiment shown below is an optical modulator composed of a Mach-Zehnder type optical waveguide. However, the waveguide type optical element according to the present invention is not limited to this, and an electric field in a certain direction is applied to each of two optical waveguides. Then, a light composed of a Mach-Zehnder type optical waveguide, a directional coupler type optical waveguide, a Y-branch type optical waveguide, and other types of optical waveguides each having an electrode for generating a refractive index difference between the two optical waveguides. The present invention can be generally applied to a waveguide type optical element having a modulator, an optical switch, and other functions.

[First Embodiment]
First, a first embodiment of the present invention will be described.
FIG. 1 is a diagram for explaining the configuration and driving method of a waveguide type optical element according to the first embodiment of the present invention.
The waveguide optical device 10 is a Mach-Zehnder optical modulator in which a Mach-Zehnder (MZ) optical waveguide 102 is formed on a substrate 100.

  The substrate 100 is a substrate made of lithium niobate (LN), which is an electro-optic material, and is, for example, an X-cut LN substrate. The MZ type optical waveguide 102 has parallel waveguides 104 and 106 configured so that light is propagated in a direction orthogonal to the spontaneous polarization direction of the LN crystal constituting the substrate 100. In the present embodiment, it is assumed that the spontaneous polarization azimuth is directed upward in FIG. In a region sandwiched between the parallel waveguides 104 and 106 on the substrate 100, a radio frequency (RF) electrode 108 is disposed in parallel or substantially in parallel with the parallel waveguides 104 and 106. The ground electrodes 110 and 112 are arranged at positions facing the RF electrode 108 with being sandwiched therebetween. A high-frequency signal for modulating the intensity of light propagating through the MZ type optical waveguide 102 is applied between the RF electrode 108 and the ground electrodes 110 and 112, for example.

  On the substrate 100, a bias electrode 150 is formed as a control electrode for controlling the refractive index difference between the parallel waveguides 104 and 106 by applying an electric field to each of the two parallel waveguides 104 and 106. Yes. The bias electrode 150 includes a reference electrode 152 and operation electrodes 154 and 156 to which a positive voltage or a negative voltage with respect to the potential of the reference electrode 152 is applied. The reference electrode 152 is aligned with the RF electrode 108 in the length direction of the parallel waveguides 104 and 106 in a region between the parallel waveguides 104 and 106 on the substrate 100 and in parallel with the parallel waveguides 104 and 106. Has been placed. Further, the working electrodes 154 and 156 are spaced apart from the reference electrode 152 by the same or substantially the same distance (separation distance) so as to face the reference electrode 152 with the parallel waveguides 104 and 106 interposed therebetween, respectively. , 106 are arranged in parallel or substantially in parallel.

The output of the control device 160 is connected to the reference electrode 152 and the working electrodes 154 and 156 that constitute the bias electrode 150, and the control device 160 causes the potentials V _a0 of the reference electrode 152 and the working electrodes 154 and 156 to be connected. , V _a1 and V _a2 are controlled. The control device 160 controls the potentials V _a0 , V _a1 , and V _a2 of the reference electrode 152 and the working electrodes 154 and 156 based on the input control signal. Thereby, an electric field is applied to the parallel waveguide 106 by the potential difference V _a1 −V _a0 between the reference electrode 152 and the working electrode 154, and the parallel waveguide 104 is caused by the potential difference V _a2 −V _a0 between the reference electrode 152 and the working electrode 156. When an electric field is applied, a refractive index difference is generated between the parallel waveguides 104 and 106, and the drift phenomenon in the light modulation operation of the MZ type optical waveguide 102 by the RF electrode 108 is compensated.

In particular, in the waveguide type optical element 10 according to the present embodiment, the potential difference V _a1 −V _a0 is negative, and the spontaneous polarization orientation of the crystal is opposite between one working electrode 154 (or 156) and the reference electrode 152. An electric field in the direction (negative direction electric field) is generated, the electric potential difference V _a2 −V _a0 is positive, and an electric field in the same direction as the spontaneous polarization orientation of the crystal between the other working electrode 156 (or 154) and the reference electrode 152. The potential V applied to the reference electrode 152 and the working electrodes 154 and 156 by the control device 160 so as to generate (positive electric field) and the electric field strength of the negative electric field is smaller than the electric field strength of the positive electric field. _a0 , V _a1 and V _a2 are controlled.

In the present embodiment, the distances between the reference electrode 152 and the working electrodes 154 and 156 are the same, so the electric field strength that satisfies the above condition is, for example, | V _a1 −V _a0 | <| V _a2 −V _{Let a0} |. V _a0 may be set to, for example, a ground potential or a potential close to the ground potential.

  According to the knowledge of the inventors of the present invention, in an optical waveguide to which a bias electric field opposite in direction to the spontaneous polarization orientation of the crystal is applied (hereinafter referred to as a negative potential difference), the bias electric field in the same direction as the spontaneous polarization orientation of the crystal. The DC drift phenomenon is more likely to proceed than the optical waveguide to which is applied (that is, the increase in voltage shift is likely to proceed). This DC drift is considered to occur because carriers are likely to be injected from an electrode formed on the surface of the LN substrate into a localized site caused by an LN crystal defect or the like inside the LN substrate.

  In the carrier, electrons (negative charges) are more likely to be injected than holes (positive charges), and there is a dependency on the base material of the electrodes. Electrons are compared with holes in the LN substrate. Because of its high mobility, carrier injection occurs in an optical waveguide driven with a negative potential difference, and the electric charge in the LN substrate increases to easily cancel the external electric field due to the bias electrode. It is thought that the DC drift phenomenon has become easier to proceed. The inventors consider this phenomenon as one of the precursory phenomena of thermal destruction or electron avalanche, and presume that this phenomenon occurs due to the polarity.

  Although the cause of the polarity is not yet clarified, it is considered that it may be related to the coercive electric field, which is an electric field necessary for the polarization operation of the LN crystal, and the presence of microdomains in the crystal. Electron injection and hole injection are unlikely to occur in the case of a highly insulating material such as LN, and as far as the inventor knows, there is no report of confirming the generation of harmonic melt composition used in Mach-Zehnder type optical modulators in LN. . However, regions doped with impurities to form microdomain walls and optical waveguides are crystal defects, which become carrier injection sites and conduction paths, and leakage current flows preferentially to generate Joule heat. Can be a place. Electron avalanche is a phenomenon that occurs when electrons injected from the electrode repeatedly collide proliferatively in a dielectric material under a high electric field, but the above-mentioned defect location is a location where the preferential injection and collision of electrons repeat. obtain.

  The coercive electric field, which is the polarization reversal threshold value of the LN of harmonic melt composition used in the Mach-Zehnder type optical modulator, is about 21 kV / mm at room temperature, but the stoichiometric composition of LN in which the defect sites of Li are filled with Li. It is known that the crystal drops to about 2 kV / mm. It is known that Ti, which is an impurity, is doped to form an optical waveguide. However, Ti is known to enter a defect portion of Li, and the coercive electric field is considered to be low like LN crystals of stoichiometric composition. . Here, when the separation distance between the reference electrode 152 and the working electrodes 154 and 156 is 14 μm, for example, an electric field is applied between the reference electrode 152 and the working electrode 154 to which 28 V is applied in the direction opposite to the spontaneous polarization direction of the LN substrate. The intensity reaches its coercive field, the microdomain grows and the domain wall increases, and also induces the generation of new microdomains. Even if the electric field strength does not reach the coercive electric field, if it is close to the coercive electric field, microdomain growth and domain wall increase occur.

  On the other hand, when applied in the same direction as the spontaneous polarization orientation of the LN substrate, the microdomain domain wall shrinks and disappears, so that the defect portion does not increase. That is, when applied in the direction opposite to the spontaneous polarization direction of the LN substrate, the number of the above-described defect points increases, and both thermal breakdown and electron avalanche are likely to occur. According to the knowledge of the authors, in the Mach-Zehnder type optical modulator having no buffer layer between the control electrode and the LN substrate, the electric field strength at which the polarity dependence of the drift phenomenon considered to be the above-mentioned estimation phenomenon occurs is Although it depends on the structure of the Mach-Zehnder type modulator, particularly the base material of the control electrode and the composition of the buffer layer, it is about 1 kV / mm, and the phenomenon was confirmed more clearly at 2 kV / mm. This phenomenon was also clearly confirmed in a Mach-Zehnder type optical modulator in which an LN substrate was processed to 10 microns or less. This is because the surface of the processed crystal is where the arrangement of atoms is interrupted, which is the most significant crystal defect, and due to the processing, many crystal defects such as dislocations and sliding surfaces are also formed inside the crystal substrate. I think it depends. In addition, such a thin crystal substrate with many crystal defects has a high free energy state. Compared with a crystal substrate with few defects, the formation and movement of a domain-inverted wall, which is one form of crystal defects, is energetically This is considered to be easy.

  Note that even when the electric field is further increased to 4 kV / mm or more, which is twice as much, the above asymmetry with respect to the DC drift becomes more remarkable, but dielectric breakdown that is probably caused by thermal breakdown or electron avalanche is not observed. High insulation is maintained, and the leakage current flowing between the electrodes and the multiplication of electrons due to the avalanche effect do not lead to more energy than destroying the crystal. However, various phenomena that appear to be caused by the discharge between the electrodes become significant.

  In recent years, a Mach-Zehnder type optical modulator in which a part of a substrate is inverted in polarization has been proposed. However, in the region where the polarization inversion process is performed on the original substrate, the polarity of the above phenomenon is exhibited. The reverse is true. In the region where the polarization inversion process is performed, in the optical waveguide to which the bias electric field having the same direction as the spontaneous polarization direction of the crystal is applied, the DC drift is larger than that in the optical waveguide to which the bias electric field having the opposite direction to the spontaneous polarization direction of the crystal is applied. The phenomenon is easy to progress. It is necessary to consider the strength of the electric field in the positive and negative directions depending on the history of the polarization inversion process.

  Also, in general, corona discharge generated between electrodes has a discharge occurrence threshold of discharge from an object to which a negative voltage with respect to the ground potential is applied, compared to discharge from an object to which a positive voltage with respect to the ground potential is applied. Is known to be lower. In a Mach-Zehnder type optical modulator for communication, the inside of a housing is generally replaced with nitrogen gas, and the negative corona discharge threshold of nitrogen is 3.8 kV / mm. Actually, it seems that even a lower electric field is generated along the surface of the substrate (surface discharge). The discharge generated by this level of electric field is at most about μA, which is smaller than the leak current that constantly flows through the LN substrate.

  The fluctuation of the bias state due to the discharge is extremely small and does not affect the maintenance of digital communication quality. That is, as long as corona discharge and creeping discharge are in the flashover mode, there is substantially no problem. However, it is generated as outgas from metals with high vapor pressure such as In, Sn, etc. contained in the metal solder of the electrode material inside the housing of the Mach-Zehnder type optical modulator, adhesives, conductive resins, sealants, solderlux, etc. Organic deposits and deposits such as residual monomers and solvents, antistatic agents, plasticizers, flame retardants, antioxidants and their reactants are baked in, causing the discharge path to be fixed and grow. .

  If the voltage is further increased from the state of corona discharge, or the distance between the conductor's protruding portion and another conductor is decreased, a spark discharge in which the protruding portion of the conductor and the other conductor are intermittently connected by sparks will occur. Occur. Although the current of the spark discharge is not so large, it is an intermittent and unstable discharge. In the bias electrode of the Mach-Zehnder type modulator, a spark discharge may occur before reaching the complete tracking state. When the energized portion is destroyed by the spark discharge, the bias state changes and the signal quality deteriorates. The influence is particularly remarkable in the configuration in which the electrode is formed on the electro-optic substrate without using the buffer layer or the charge dispersion film. Tracking may progress without clear spark discharge or with repeated spark discharge. If the energized part is opened and the energization amount changes abruptly due to a spark discharge or a fuse phenomenon in a state where tracking is advanced, the bias state fluctuates greatly and signal quality deteriorates.

  In addition, if there is a residual gas such as halogen in the housing, or a metal with a high vapor pressure such as Sn or In or an alloy thereof, the deposition of a metal with a high vapor pressure between the electrodes becomes noticeable, the occurrence of a tracking state, The destruction of the energized part by the spark discharge is repeated.

  Considering the threshold value for the start of corona discharge, setting the electrode spacing and setting the voltage applied between the electrodes is one of the countermeasures, but to reduce the influence of bias fluctuation due to spark discharge or fuse phenomenon It is sufficient to reduce the resistance between the electrodes and increase the leak current that always flows. In the case where the electrode is formed on the electro-optic substrate without using the buffer layer or the charge dispersion film, if a semiconductive thin film is provided in a part between the bias electrodes and the interelectrode resistance is 1 MΩ or less, spark discharge or The effect of bias point fluctuation due to the fuse phenomenon is virtually invisible.

  When electrodes are formed via a buffer layer or a charge dispersion film, the interelectrode resistance is generally 1 MΩ or less, and there is no need to form a semiconductive film between the electrodes. Thereby, even if the electric field between the electrodes is set to 4 kV / mm or more, it is possible to avoid the influence of the fluctuation of the bias state due to the occurrence of tracking or the opening of the tracking portion. Regarding the electric field strength between the bias electrodes, the electrode interval and the applied voltage may be set in consideration of the asymmetry of the DC drift.

  However, it is necessary to sufficiently consider the threshold value for occurrence due to electromigration between electrodes. Electromigration is a phenomenon that strongly depends on the presence of the electrode material and halogen in the housing. When gold (Au) is used for the electrode, tracking due to electromigration is slight even at 10 kV / mm. In the case of an electrode using copper (Cu) or silver (Ag), electromigration occurs remarkably even at about 5 kV / mm. Appropriate measures are necessary.

  In addition, the variation among individual Mach-Zehnder type optical modulators is large, but when the applied electric field exceeds 10 kV / mm, the occurrence of dielectric breakdown of the LN crystal is observed. Also, DC drift is greatly accelerated in an optical waveguide to which a bias electric field having the same direction as the spontaneous polarization orientation of the crystal is applied. In other words, this technique can be said to be an effective technique for reducing the DC bias until the applied electric field is about 1 kV / mm to 10 kV / mm.

  In the waveguide type optical element 10 according to the present embodiment, as described above, an electric field generated in the substrate 100 by the working electrode 154 (or 156) to which a negative voltage with respect to the potential of the reference electrode 152 is applied and the reference electrode 152. The reference electrode 152 and the working electrode 154 have a strength smaller than the electric field strength generated in the substrate 100 by the working electrode 156 (or 154) to which a positive voltage with respect to the potential of the reference electrode 152 is applied and the reference electrode 152. Since the potentials Va0, Va1, Va2 of 156 are controlled, the carrier (electron) injection into the substrate 100 is suppressed to suppress an increase in DC drift, and the operation is driven by a negative voltage with respect to the potential of the reference electrode 152. As a whole, drift between the parallel waveguides 104 and 106 is suppressed while suppressing discharge between the electrode (for example, the working electrode 154) and the reference electrode 152. It is possible to ensure the refractive index difference required compensation.

  In the present embodiment, a non-conductive intermediate layer (for example, a buffer layer such as SiO2) is not formed between the substrate 100 and the bias electrode 150, but according to the knowledge of the inventors of the present invention, Even when such a non-conductive intermediate layer (for example, a buffer layer using a material having a lower dielectric constant than LN such as SiO 2) is formed, the optical waveguide driven with a negative voltage is driven with a positive voltage. The DC drift phenomenon is more likely to proceed than the optical waveguide that is formed (that is, the increase in voltage shift is likely to proceed). This DC drift is considered to facilitate the injection of carriers from the electrode when oxygen vacancies or metal impurities are present inside the buffer layer. As the carriers, electrons (negative charges) having a higher mobility than holes (positive charges). ) Is more likely to be injected, so the DC drift phenomenon is likely to proceed in an optical waveguide driven with a negative potential difference.

  Therefore, the driving method of the bias electrode 150 by the control device 160 in the present embodiment suppresses increase in drift and inter-electrode discharge even when there is a buffer layer between the LN substrate 100 and the bias electrode 150. As a whole, an effect of generating a difference in refractive index necessary for drift compensation in the parallel waveguides 104 and 106 can be achieved.

  In the present embodiment, the substrate 100 is described as an X-cut LN substrate. However, a Z-cut LN substrate may be used. In this case, the working electrodes 154 and 156 are formed immediately above the parallel waveguides 104 and 106 via the buffer layer (corresponding to the non-conductive intermediate layer) made of, for example, SiO 2 described above, and the reference electrode 152 is arranged in parallel or substantially parallel to the working electrodes 154 and 156 by a predetermined distance.

[Second Embodiment]
Next, a second embodiment of the present invention will be described.
FIG. 2 is a diagram for explaining the configuration and driving method of a waveguide type optical device according to the second embodiment of the present invention. 2, the same reference numerals as those in FIG. 1 are used for the same components as those of the waveguide type optical element 10 according to the first embodiment shown in FIG. 1, and according to the above-described first embodiment. The description of the waveguide type optical element 10 is incorporated.

  The waveguide type optical element 20 has the same configuration as the waveguide type optical element 10 according to the first embodiment, but includes a bias electrode 250 instead of the bias electrode 150. The bias electrode 250 has the same configuration as the bias electrode 150, but includes a reference electrode 252 and working electrodes 254 and 256 instead of the reference electrode 152 and the working electrodes 154 and 156. The working electrodes 254 and 256 are formed in parallel with the parallel waveguides 104 and 106 so as to face the reference electrode 252 with the parallel waveguides 104 and 106 interposed therebetween, respectively. However, the distance between the working electrode 254 and the reference electrode 252 is configured to be larger than the distance between the working electrode 256 and the reference electrode 252.

  For this reason, in the waveguide type optical element 20 of the present embodiment, even if the potentials Vb0, Vb1, and Vb2 of the reference electrode 252 and the operation electrodes 254 and 256 controlled by the control device 260 are Vb2 = Vb1 <Vb0, the reference The electric field strength of the negative direction electric field generated in the substrate 100 by the electrode 252 and the working electrode 254 (that is, the electric field in the direction opposite to the spontaneous polarization direction of the substrate 100) is set in the substrate 100 by the working electrode 256 and the reference electrode 252. Can be made smaller than the electric field strength of the positive direction electric field (that is, the electric field in the same direction as the spontaneous polarization direction of the substrate 100). Such a configuration can simplify the configuration of the control device 260 because Vb1 and Vb2 can be generated with the same potential, particularly when Vb0 is set to the ground potential. Further, Vb1 and Vb2 may be set to the ground potential, and Vb0 may be set to a potential of 0 V or higher.

[Third Embodiment]
Next, a third embodiment of the present invention will be described.
FIG. 3 is a diagram for explaining the configuration and driving method of a waveguide type optical element according to the third embodiment of the present invention. In FIG. 3 and FIG. 4 to be described later, the same reference numerals as those in FIG. 1 are used for the same components as those of the waveguide type optical element 10 according to the first embodiment shown in FIG. The description of the waveguide type optical element 10 according to the embodiment is incorporated.

  The waveguide type optical element 30 has the same configuration as the waveguide type optical element 10 according to the first embodiment, but includes a bias electrode 350 instead of the bias electrode 150. The bias electrode 250 has the same configuration as the bias electrode 150, but includes a reference electrode 352 and working electrodes 354 and 356 instead of the reference electrode 152 and the working electrodes 154 and 156.

  4 is a cross-sectional view of the waveguide type optical element 30 shown in FIG. The reference electrode 352 and the working electrodes 354 and 356 have the same configuration as the reference electrode 152 and the working electrodes 154 and 156, respectively, but a non-conductive intermediate layer 370 is formed between the working electrode 354 and the substrate 100. Is different.

  For this reason, in the waveguide type optical element 30 of the present embodiment, even if the potentials Vc0, Vc1, and Vc2 of the reference electrode 352 and the working electrodes 354 and 356 controlled by the control device 360 are Vc2 = Vc1 <Vc0, the reference The electric field strength of the negative direction electric field generated in the substrate 100 by the electrode 352 and the working electrode 354 (that is, the electric field in the direction opposite to the spontaneous polarization direction of the substrate 100) is Can be made smaller than the electric field strength of the positive direction electric field (that is, the electric field in the same direction as the spontaneous polarization direction of the substrate 100). In such a configuration, in particular, if Vc0 is set to the ground potential, Vc1 and Vc2 can be generated as the same potential, so that the configuration of the control device 360 can be simplified. Further, Vc1 and Vc2 may be set to the ground potential, and Vc0 may be set to a potential of 0 V or higher.

  In this embodiment, the intermediate layer 370 is not formed between the reference electrode 352 and the working electrode 356 and the substrate 100. However, the present invention is not limited to this, and the non-conductive intermediate layer 370 is used as the reference electrode. 352 and the working electrode 356 and the LN substrate 100 are formed, and the thickness of the intermediate layer 370 formed between the working electrode 354 and the substrate 100 is set to be the reference electrode 352 and the working electrode 356 and the substrate 100. By making it thicker than the thickness of the intermediate layer 370 formed between the two, an effect similar to the above can be obtained.

  As described above, the waveguide type optical elements (10, etc.) according to the first to third embodiments described above apply electric fields in opposite directions to the two parallel waveguides (104, 106), respectively. A bias electrode (150, etc.) for generating a refractive index difference between the two parallel waveguides, and the bias electrode includes a reference electrode (152, etc.) to which a reference potential is applied, and the reference electrode Two working electrodes (154, 156, etc.) to which a positive voltage or a negative voltage with respect to the potential is applied, and the crystal constituting the substrate (100) between one working electrode (154, etc.) and the reference electrode An electric field opposite to the spontaneous polarization orientation (negative electric field) is generated, and an electric field (positive electric field) in the same direction as the spontaneous polarization orientation of the crystal is generated between the other working electrode (156, etc.) and the reference electrode, And, the electric field strength of the negative direction electric field is As it is smaller than the electric field strength in the direction the electric field, the potential applied to the reference electrode and two working electrodes is controlled.

  As a result, the waveguide type optical element (such as 10) suppresses carrier (electron) injection into the substrate (100) or the intermediate layer (370) to suppress an increase in drift, and the reference electrode (such as 152). As a whole, the difference in refractive index required for drift compensation between the parallel waveguides (104, 106) is suppressed while suppressing the discharge between the working electrode (154, etc.) to which a negative voltage is applied and the reference electrode. Can be secured.

  In the above-described embodiment, for the sake of simplicity of explanation and illustration, the substrate 100 has been described as being configured using an LN X plate and a Y plate. However, the waveguide optical element according to the present invention is the substrate 100. An LN-Z plate, a substrate having another orientation, or a substrate having a concavo-convex structure can be used. That is, an electric field (negative electric field) opposite to the spontaneous polarization orientation of the crystal constituting the substrate 100 is generated between one working electrode and the reference electrode, and the substrate 100 is placed between the other working electrode and the reference electrode. An electric field (positive electric field) having the same direction as the spontaneous polarization orientation of the crystal constituting the crystal is generated, and the electric field strength of the negative electric field is smaller than the electric field strength of the positive electric field. By controlling the potential applied to the working electrode, the drift voltage can be reduced. Needless to say, when a bias electric field is applied only to one of the parallel optical waveguides, it is advantageous to apply the electric field in the same direction as the spontaneous polarization direction. In addition, an example in which the RF electrode and the bias electrode are separately provided has been shown, but even in the configuration in which the RF electrode and the bias electrode are shared, the same phenomenon occurs when the electric field between the electrodes is high, and has been described above. Needless to say, the configuration is effective as a countermeasure.

  10, 20, 30 ... waveguide type optical element, 100 ... substrate, 102 ... MZ type optical waveguide, 104, 106 ... parallel waveguide, 108 ... RF electrode, 110, 112 ..Ground electrode, 150, 250, 350 ... Bias electrode, 152,252,352 ... Reference electrode, 154,156,254,256,354,356 ... Operating electrode, 160,260,360 ..Control device, 370 ... intermediate layer.

Claims (5)

A substrate having an electro-optic effect;
Two optical waveguides formed on the substrate;
A control electrode for applying an electric field in a direction opposite to each other along the spontaneous polarization direction of the substrate to the two optical waveguides to generate a refractive index difference between the two optical waveguides;
With
The control electrode includes a reference electrode disposed between the two optical waveguides, and two operations disposed in parallel or substantially in parallel with the reference electrode with the two optical waveguides interposed therebetween. An electrode, and
The distance between the reference electrode and one working electrode is arranged to be larger than the distance between the reference electrode and the other working electrode.
Waveguide type optical element. A substrate having an electro-optic effect;
Two optical waveguides formed on the substrate;
A control electrode for applying an electric field in a direction opposite to each other along the spontaneous polarization direction of the substrate to the two optical waveguides to generate a refractive index difference between the two optical waveguides;
With
The control electrode includes a reference electrode disposed between the two optical waveguides, and two operations disposed in parallel or substantially in parallel with the reference electrode with the two optical waveguides interposed therebetween. An electrode, and
A non-conductive intermediate layer is disposed between one working electrode and the substrate;
No non-conductive intermediate layer is disposed between the other working electrode and the substrate,
Waveguide type optical element. A substrate having an electro-optic effect;
Two optical waveguides formed on the substrate;
A control electrode for applying an electric field in a direction opposite to each other along the spontaneous polarization direction of the substrate to the two optical waveguides to generate a refractive index difference between the two optical waveguides;
With
The control electrode includes a reference electrode disposed between the two optical waveguides, and two operations disposed in parallel or substantially in parallel with the reference electrode with the two optical waveguides interposed therebetween. An electrode, and
A non-conductive intermediate layer is disposed between the two working electrodes and the substrate; and
The intermediate layer between one working electrode and the substrate is thicker than the intermediate layer between another working electrode and the substrate;
Waveguide type optical element. The substrate comprises lithium niobate;
The two optical waveguides are two parallel waveguides constituting a Mach-Zehnder type optical waveguide,
The control electrode is a bias electrode for compensating for a drift phenomenon.
The waveguide type optical device according to any one of claims 1 to 3. A method for driving a waveguide-type optical element according to any one of claims 1 to 4,
A first voltage is applied between the one working electrode and the reference electrode, and an electric field is applied in a direction opposite to the spontaneous polarization orientation in the substrate between the one working electrode and the reference electrode. A step of generating
A second voltage is applied between the other working electrode and the reference electrode, and an electric field is applied in the same direction as the spontaneous polarization azimuth in the substrate between the other working electrode and the reference electrode. A process of generating;
Have
The first and second voltages are set such that the electric field strength of the electric field in the direction opposite to the spontaneous polarization direction generated in the substrate is smaller than the electric field strength of the electric field in the same direction as the spontaneous polarization direction. The way.

Images