JP2005352445A - Optical kerr switch and nonlinear optical fiber - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an optical Kerr switch with which the control of the disturbance fluctuation of a nonlinear optical fiber is easy. <P>SOLUTION: The optical Kerr switch makes at least linearly polarized light signal light incident on one end of a nonlinear optical fiber from the one end of which the linearly polarized light signal and linearly polarized light control light can be made incident, guides the light emitted from the other one end of the optical fiber to an analyzer and transmits or shuts off or shuts off or transmits the signal light according to whether the control light is made incident simultaneously with the signal light or not. The nonlinear coefficient to the light of the wavelength 1,550 nm of the nonlinear optical fiber is ≥470 W<SP>-1</SP>km<SP>-1</SP>. The glass transition temperature of the core glass thereof is ≥300°C. The peak power of the control light thereof is ≤500 mW. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an optical fiber suitable for nonlinear optical elements such as an optical Kerr switch and an optical Kerr switch.

The OTDM system has been vigorously developed to construct a photonic network using an OTDM (optical time division multiplexing) system that utilizes the ultra-high speed of optical devices.
In the OTDM system, ultrafast optical switching is essential, and an optical Kerr switch is expected as an optical switch suitable for such ultrafast optical switching.

An optical Kerr switch causes linearly polarized control light to enter a nonlinear optical medium such as a nonlinear optical fiber to induce anisotropy in the refractive index of the same medium. The plane is rotated so that it is parallel or orthogonal to the polarization plane of the analyzer, and this polarization plane rotation signal light is guided to the analyzer to be transmitted or reflected (blocked).
As the nonlinear optical fiber, a silica-based glass fiber in which fluorine or the like is added to the clad and GeO ₂ or the like is added to the core, and the mode field diameter is about 4 μm and the third-order nonlinear coefficient (γ) is 15 W ⁻¹ km ^−1. Is known (see Patent Document 1).

JP 2004-38016 A (Page 5)

When the length of the nonlinear optical fiber is L and the control light peak power in the nonlinear optical fiber is P, when the angle formed by both polarization planes of the signal light and the control light simultaneously incident on the optical fiber is π / 4, That is, when the rotation angle of the polarization plane of the signal light in the optical Kerr switch becomes maximum with respect to the same P and L, the following equation is established with φ _max as the phase difference at the _maximum .
φ _max = (4/3) × γ × P × L.

If a general-purpose laser diode (LD) is used as the control light source, P is at most 500 mW, and if the silica glass fiber is used as the nonlinear optical fiber, γ is 15 W ⁻¹ km ⁻¹ . In this case, when φ _{max is set} to π, that is, by setting the angle formed by both the polarization planes to π / 4, the rotation angle of the signal light polarization plane is maximized with respect to the same P and L. If the phase difference is to be π, L must be about 310 m.

However, for such extremely long glass fibers, disturbances such as arrival time fluctuations (jitter) due to non-uniformity in the longitudinal direction of the fiber, temperature fluctuations that cause polarization mode dispersion, and vibrations are suppressed to a satisfactory level. Was not easy.
An object of the present invention is to provide an optical Kerr switch that can solve such problems and a nonlinear optical fiber suitable for such an optical Kerr switch.

In the present invention, at least linearly polarized signal light is incident on one end of a nonlinear optical fiber capable of receiving linearly polarized signal light and linearly polarized control light from one end thereof, and light emitted from the other end of the optical fiber is used as an analyzer. An optical Kerr switch that transmits or blocks or blocks or transmits signal light depending on whether or not control light is incident simultaneously with signal light, and has a nonlinear coefficient of 470 W ⁻¹ for light having a wavelength of 1550 nm of a nonlinear optical fiber. An optical car switch that is km ⁻¹ or more is provided.
In addition, a nonlinear optical fiber (hereinafter referred to as a first optical fiber) having a nonlinear coefficient of 1100 W ⁻¹ km ⁻¹ or more for light having a wavelength of 1550 nm is provided.

In addition, the core is expressed by mol% based on the following oxide, Bi ₂ O ₃ 60 to 70%, B ₂ O ₃ 10 to 15.5%, Ga ₂ O ₃ 6.5 to 12.5%, In ₂ O _{3 to} 10%, ZnO 1 to 10%, MgO + CaO + SrO + BaO 0.5 to 7%, CeO ₂ 0 to 2%, and the cladding is represented by mol% based on the following oxide basis, Bi _{2 O 3 50~60%, B 2 O} 3 18~28%, Ga 2 O 3 4~15%, In 2 O 3 1~10%, ZnO 1~10%, MgO + CaO + SrO + BaO 1~8%, Li 2 O + Na 2 Provided is a nonlinear optical fiber (hereinafter referred to as a second optical fiber) which is a glass consisting essentially of O + K ₂ O 0.1-2% and CeO ₂ 0-2%.

The optical switching function is less affected by disturbances such as temperature fluctuations and vibrations than the conventionally known optical Kerr switch, and therefore the disturbance is not effective even if the disturbances impede the optical switching function in the prior art. It is possible to obtain an optical Kerr switch that can easily suppress the influence of disturbance to a problem-free level.
In addition, an optical Kerr switch in which the optical switching function is hardly inhibited by jitter can be obtained.
Further, an optical Kerr switch capable of optical switching can be obtained even when the P is 500 mW.
In addition, an optical Kerr switch that has the effects described above and can be fusion-spliced with a single mode silica-based optical fiber can be obtained.
Further, a nonlinear optical fiber suitable for realizing an optical Kerr switch having these effects can be obtained.

FIG. 1 is a diagram for explaining an optical Kerr switch according to the present invention. Hereinafter, the present invention will be described with reference to the drawings, but the present invention is not limited thereto.
The signal light is a linearly polarized pulse, which enters from the left side of the figure, enters the left end of the nonlinear optical fiber via the signal light polarization controller, and the signal light emitted from the right end of the optical fiber enters the analyzer.
If the plane of polarization of the light entering the analyzer is parallel to the analyzer transmission axis, the signal passes through the filter after passing through the analyzer and enters the next system (not shown). If they are orthogonal, the signal light cannot be transmitted through the analyzer and is blocked.

  The control light emitted from the control light source is linearly polarized light, and its polarization plane is formed at an angle of π / 4 with the signal light polarization plane through the control light polarization controller, and then the nonlinear optical fiber together with the signal light. Enter the left end of. The control light emitted from the right end of the optical fiber that has passed through the analyzer reaches the filter and is blocked there.

When the signal light enters the left end of the nonlinear optical fiber together with the control light, the polarization plane of the signal light rotates as it travels through the optical fiber, and when it exits from the right end of the optical fiber, the rotation angle becomes π / 2. Is done.
When the control light is not emitted from the control light source, the polarization plane of the signal light travels through the nonlinear optical fiber without rotating.

Therefore, if the polarization plane of the signal light entering the left end of the nonlinear optical fiber is perpendicular to the analyzer transmission axis, the signal light will pass the analyzer when the signal light and the control light enter the left end of the nonlinear optical fiber at the same time. When the signal light enters the left end of the nonlinear optical fiber in the absence of control light, the signal light cannot pass through the analyzer and is blocked and the optical Kerr switch is turned off.
If the polarization plane of the signal light entering the left end of the nonlinear optical fiber is parallel to the analyzer transmission axis, the signal light cannot be transmitted through the analyzer and blocked if the signal light and the control light enter the left end of the nonlinear optical fiber simultaneously. Then, the optical Kerr switch is turned off. When the signal light enters the left end of the nonlinear optical fiber in the absence of control light, the signal light passes through the analyzer and the optical Kerr switch is turned on.

The bit rate of the optical car switch of the present invention is usually 10 to 400 Gb / s, typically 80 to 400 Gb / s.
The wavelength of signal light that can be processed by the optical Kerr switch of the present invention is typically 1500 to 1650 nm.
When processing WDM (wavelength division multiplexing) signal light, for example, AWG (demultiplexer) is used to separate the signal light for each wavelength, and for each of the separated one wavelength signal light, the optical car switch of the present invention is used. What is necessary is just to process. That is, the WDM signal light can be processed by arranging a plurality of optical car switches of the present invention.
The signal light pulse frequency is usually 10 to 400 GHz, typically 80 to 400 GHz.

  Unlike the wavelength of the control light and the wavelength of the signal light, the difference between the two is preferably 2 to 20 nm. If it is less than 2 nm, the extinction ratio of the filter becomes a problem and the crosstalk performance may be lowered. If it exceeds 20 nm, the dispersion of the nonlinear optical fiber is not zero, so that the nonlinear effect is weakened, and it may be difficult to sufficiently increase the phase rotation. More preferably, it is 10 nm or less.

The control light pulse frequency is usually 10 to 400 GHz, typically 80 to 400 GHz.
The shape of the control light pulse is preferably Sech type or Gaussian type, but may be other than that, for example, rectangular.
The duty ratio (on / off ratio) of the control light pulse is typically 1: 1 to 1: 4.

The control light peak power (P) is preferably 10 W or less. Above 10 W, the average power of the control light becomes large, and the connection point through which the control light passes, for example, the connection point between the silica-based optical fiber and the analyzer, and the nonlinear optical fiber are usually connected to the silica-based optical fiber. In some cases, a fiber fuse is formed at the connection point of the two fibers in the case, and as it progresses, damage may spread to the light source, the analyzer, and the like. More preferably, it is 5 W or less.
Problems such as excessive power consumption and disturbed pulse waveforms, large cooling units and amplifying units are required, and a large space is required, and it is difficult to use general-purpose light sources. When it is desired to avoid it, P is preferably 500 mW or less.

  Further, P is preferably 10 mW or more. If it is less than 10 mW, the length (L) of the nonlinear optical fiber must be increased, and it is difficult to suppress the disturbance, or the jitter may increase. More preferably, it is 50 mW or more, particularly preferably 100 mW or more, and typically 200 mW or more.

An example of the control light source is an ultrafast fiber laser. In this case, the control light is usually obtained by compressing the laser output light with a pulse compressor, and then converting the light amplified using an EDFA into linearly polarized light with a polarizer.
In some cases, it is preferable to use a high-power LD that can obtain control light having a desired P without using an optical amplifier such as an EDFA as the control light source.

  Both the signal light and control light polarization controllers are not essential, but at least the signal light polarization in order to control the polarization plane of the signal light or the control light so that the angle formed by both polarization planes is π / 4. It is preferable to install the controller upstream from the point where the signal light and the control light are combined, and it is more preferable to install the polarization controller for the control light as well.

  The analyzer preferably has an extinction ratio of -40 dB or more in the signal light wavelength region. If it is less than −40 dB, the crosstalk becomes large and the signal quality may be deteriorated. More preferably, it is −50 dB or more.

The filter is not essential, but you want to remove the control light from the signal light and control light that has passed through the analyzer to make only the signal light, or to suppress the crosstalk of the optical car switch and improve the signal quality. Is preferably installed.
The extinction ratio in the signal light wavelength region of the filter is preferably -60 dB or more. If it is less than −60 dB, the crosstalk becomes large and the signal quality may be deteriorated.
An example of the filter is a dielectric multilayer bandpass filter.

  The nonlinear optical fiber is preferably an optical fiber whose core and cladding are both glass. Without such an optical fiber, there is a possibility that high-efficiency connection with a silica-based optical fiber, which is a general-purpose optical transmission fiber, may not be easy.

  The glass transition point (Tg) of the core glass is preferably 300 ° C. or higher. If it is less than 300 ° C., fusion splicing with the silica-based optical fiber may be difficult. More preferably, it is 350 ° C. or higher.

  Further, it is preferable that neither the glass of the core nor the clad contains Pb, As, Sb, Se, Tl or Cd.

The length (L) of the nonlinear optical fiber is preferably 10 m or less. In such a case, since it is about 1/30 of the length (about 310 m) when a conventional nonlinear optical fiber is used, the disturbance can be remarkably reduced or jitter can be reduced. L is preferably 5 m or less, more preferably 3 m or less, and particularly preferably 1 m or less.
In order to improve the operational stability of the optical Kerr switch, the product of P and L is preferably 5 W · m or less. More preferably, it is 4 W · m or less.

If γ of the nonlinear optical fiber is less than 470 W ⁻¹ km ⁻¹ , it becomes difficult to set L to 10 m or less under the condition that P is 500 mW or less. γ is preferably ^625W -1 ^{miles -1} or more, more preferably ^1000W -1 ^{miles -1} or more, and particularly preferably ^{1260W -1 km} -1 or more, most preferably ^{1300 W} -1 ^{miles -1} or more. Typically a 1100~5000W ^-1 km ^-1.

Examples of the nonlinear fiber include the second optical fiber having γ of 470 W ⁻¹ km ⁻¹ or more, or the first optical fiber.

In the first optical fiber of the present invention, γ is 1100 W ⁻¹ km ⁻¹ or more. Therefore, when used in the optical Kerr switch of the present invention, the L can be set to 5 m or less. Preferably at ^{1300 W} -1 ^{miles -1} or more.
The core and clad of the first optical fiber of the present invention are typically made of glass when high efficiency connection with a silica-based optical fiber that is a general-purpose optical transmission fiber is desired.
In this typical embodiment, the Tg of the core glass is preferably 300 ° C. or higher. If it is less than 300 ° C., fusion splicing with the silica-based optical fiber may be difficult. More preferably, it is 350 ° C. or higher.

Next, the second optical fiber of the present invention will be described.
γ is typically 470 W ⁻¹ km ⁻¹ or more, more typically 625 W ⁻¹ km ⁻¹ , and is suitable for the nonlinear optical fiber of the optical Kerr switch of the present invention. γ is preferably ^1100W -1 ^{miles -1} or more, more preferably ^{1260W -1 km} -1 or more, and particularly preferably ^{1300 W} -1 ^{miles -1} or more, whereas, typically below ^{5000 W} -1 ^{miles -1} is there.
Both Tg of the core glass and the clad glass are typically 350 ° C. or higher.

Next, the core glass of the second optical fiber of the present invention will be described.
It is preferable that the nonlinear refractive index (n ′) of the core glass with respect to light having a wavelength of 1550 nm is 4 × 10 ⁻¹⁹ m ² W ⁻¹ or more. If it is less than 4 × 10 ⁻¹⁹ m ² W ^−1, it may be difficult to obtain large γ. More preferably, it is 8 × 10 ⁻¹⁹ m ² W ⁻¹ or more, and particularly preferably 1 × 10 ⁻¹⁸ m ² W ⁻¹ or more.

  The difference ΔT (= Tx−Tg) between the crystallization start temperature (Tx) and Tg of the core glass is preferably 80 ° C. or higher. If ΔT is less than 80 ° C., the thermal stability becomes too low, and heat treatment tends to cause crystallization, or the heat-treated glass has more crystals, which may increase propagation loss when light propagates through the glass. is there. ΔT is more preferably 100 ° C. or higher.

Tx is measured as follows. That is, glass powder is pulverized to produce a glass powder having a particle size of 74 to 106 μm, and a differential heat curve is measured in an air atmosphere using this as a sample. The rate of temperature rise is 10 ° C./min, and the temperature is raised to a lower temperature of (Tg + 300) ° C. or 1000 ° C.
When an exothermic peak accompanying crystal precipitation is observed in the obtained differential heat curve, the temperature at which the exothermic peak rises is defined as Tx. When the exothermic peak accompanying crystal precipitation is not observed in the differential heat curve, Tx is set to the highest temperature rise, that is, (Tg + 300) ° C. or 1000 ° C. As a differential thermal analyzer for measuring the differential thermal curve, for example, THERMOFLEX DTA8121 manufactured by Rigaku Corporation can be mentioned. In this case, the sample mass is preferably 1.0 g.

  Since the core glass of this preferred embodiment is excellent in thermal stability, when manufacturing the second optical fiber of the present invention, the allowable fluctuation range of the drawing temperature and the drawing speed can be increased. As a result, the chromatic dispersion distribution in the longitudinal direction of the optical fiber can be reduced.

Next, the component of the core glass of the 2nd optical fiber of this invention is demonstrated using mol% display content.
Bi ₂ O ₃ is a component that increases n and is essential. If it is less than 60%, n becomes small. Preferably it is 63% or more. If it exceeds 70%, vitrification becomes difficult, ΔT becomes small, or Tg becomes low, and it becomes difficult to reduce the difference from Tg of the clad glass. Preferably it is 68% or less.
B ₂ O ₃ is a network former and is essential. If it is less than 10%, vitrification becomes difficult. Preferably it is 12% or more. If it exceeds 15.5%, n becomes small. Preferably it is 14.5% or less.

Ga ₂ O ₃ is a component that increases ΔT and is essential. If Ga ₂ O ₃ is less than 6.5%, the thermal stability is lowered. Preferably it is 8% or more. If it exceeds 12.5%, n tends to be small. Preferably it is 11% or less.

In ₂ O ₃ is a component that increases n and increases ΔT, and is essential. If In ₂ O ₃ is less than 1%, the thermal stability is lowered, or n is reduced. Preferably it is 3% or more, More preferably, it is 5% or more. If it exceeds 10%, vitrification becomes difficult. Preferably it is 8% or less.

  ZnO is a component that increases n or facilitates vitrification, and is essential. If it is less than 1%, n becomes small or vitrification becomes difficult. Preferably it is 2% or more. If it exceeds 10%, crystals are more likely to precipitate when the glass is melted, and the transmittance of the glass may be reduced. Preferably it is 6% or less.

  MgO, CaO, SrO and BaO are components that facilitate vitrification, and must contain at least one of them. If the total of these contents is less than 0.5%, vitrification becomes difficult. Preferably it is 1% or more. If it exceeds 7%, crystals are liable to precipitate on melting, or n tends to be small. Preferably it is 4% or less. Of these, MgO is preferably contained.

CeO ₂ is not essential, but it may be contained up to 2% in order to prevent Bi ₂ O ₃ in the glass composition from being reduced during glass melting and precipitating as metal bismuth to lower the transparency of the glass. . If CeO ₂ exceeds 2%, vitrification becomes difficult, or yellow or orange coloring becomes strong, and the transmittance of the glass may decrease. Preferably it is 0.5% or less. When CeO ₂ is contained, its content is preferably 0.01% or more, more preferably 0.05% or more, and particularly preferably 0.1% or more.
In addition, when it is desired to avoid a decrease in the transmittance of the glass due to the coloring, the CeO ₂ content is preferably less than 0.15%, and more preferably substantially free of CeO ₂ .

This core glass consists essentially of the above components, but may contain other components as long as the object of the present invention is not impaired. In that case, the content of the other components is preferably 5% or less in total.
Note that SiO ₂ , Al ₂ O ₃ , GeO ₂ , SnO ₂ , TeO ₂ , TiO ₂ , ZrO ₂ , Ta ₂ O ₃ , Y ₂ O _3, and WO ₃ are used to suppress crystal precipitation during glass melting. It is preferable not to contain any of them.
When it contains an alkali metal oxide, the total content is preferably 1% or less.
When P ₂ O ₅ is contained, the content is preferably 1% or less.

Next, the clad glass of the second optical fiber of the present invention will be described.
ΔT of the clad glass is preferably 80 ° C. or more for the same reason as that of the core glass. More preferably, it is 100 ° C. or higher.

Hereinafter, the components of the clad glass of the second optical fiber of the present invention will be described using the mol% display content.
Bi ₂ O ₃ is a component that keeps n of the clad glass in an appropriate range with respect to n of the core glass, and is essential. If it is less than 50%, the difference between n and the core glass n becomes too large to satisfy the single mode condition. Preferably it is 53% or more. If it exceeds 60%, the difference may be too small, and the effective area described later may increase. Preferably it is 57% or less.
B ₂ O ₃ is a network former and is essential. If it is less than 18%, vitrification may be difficult. Preferably it is 20% or more, and preferably 25% or less.

Ga ₂ O ₃ is a component that increases ΔT and is essential. If Ga ₂ O ₃ is less than 4%, the thermal stability may be lowered. Preferably it is 7% or more. If it exceeds 15%, vitrification may be difficult. Preferably it is 12% or less.
In ₂ O ₃ is a component that increases n and increases ΔT, and is essential. If In ₂ O ₃ is less than 1%, the thermal stability may be decreased, or n may be decreased. Preferably it is 2% or more. If it exceeds 10%, vitrification may become difficult, or Tg may increase, making it difficult to reduce the difference from Tg of the core glass. Preferably it is 6% or less.

  ZnO is a component that lowers Tg or facilitates vitrification and is essential. If it is less than 1%, Tg becomes high, and it may be difficult to reduce the difference from Tg of the core glass, or vitrification may be difficult. Preferably it is 2% or more. If it exceeds 10%, crystals are more likely to precipitate when the glass is melted, and the transmittance of the glass may be reduced. Preferably it is 8% or less.

  MgO, CaO, SrO, and BaO are components that lower Tg or facilitate vitrification, and must contain at least one of them. If the total of these contents is less than 1%, Tg becomes high, and it is difficult to reduce the difference from Tg of the core glass, or vitrification may be difficult. Preferably it is 2% or more. If it exceeds 8%, there is a possibility that crystals are likely to precipitate when the glass is melted. Preferably it is 5% or less. Of these, BaO is preferably contained.

Li ₂ O, Na ₂ O and K ₂ O are components that lower Tg or increase α, and must contain at least one of them. If the total of these contents is less than 0.1%, Tg becomes high, and it may be difficult to reduce the difference from Tg of the core glass. Preferably it is 0.5% or more. If it exceeds 2%, vitrification may be difficult. Preferably it is 1.5% or less. Of these, Na ₂ O is preferably contained.

CeO ₂ is not essential, but may be contained up to 2% in order to prevent Bi ₂ O ₃ in the glass composition from being reduced during glass melting to precipitate as metal bismuth and lower the transparency of the glass. . If CeO ₂ exceeds 2%, vitrification becomes difficult, or yellow or orange coloring becomes strong, and the transmittance of the glass may decrease. Preferably it is 0.5% or less.
When CeO ₂ is contained, its content is preferably 0.01% or more, more preferably 0.05% or more, and particularly preferably 0.1% or more.
In addition, when it is desired to avoid a decrease in the transmittance of the glass due to the coloring, the CeO ₂ content is preferably less than 0.15%, and more preferably substantially free of CeO ₂ .

This clad glass consists essentially of the above components, but may contain other components as long as the object of the present invention is not impaired. In that case, the content of the other components is preferably 5% or less in total.
Note that SiO ₂ , Al ₂ O ₃ , GeO ₂ , SnO ₂ , TeO ₂ , TiO ₂ , ZrO ₂ , Ta ₂ O ₃ , Y ₂ O _3, and WO ₃ are used to suppress crystal precipitation during glass melting. It is preferable not to contain any of them.
When P ₂ O ₅ is contained, the content is preferably 1% or less.

It is preferable that neither the core glass nor the clad glass of the second optical fiber of the present invention contains Pb, As, Sb, Se, Tl or Cd.
The second optical fiber of the present invention is manufactured, for example, as follows.
Predetermined raw materials are mixed and placed in a gold crucible, platinum crucible, alumina crucible, quartz crucible or iridium crucible, and melted at 800 to 1300 ° C. in an air atmosphere to prepare a core glass melt and a clad glass melt. The obtained melt is cast into a predetermined mold to prepare a preform in which the core glass and the clad glass are combined. The preform is stretched to produce an optical fiber.

In the optical fiber of the present invention, the refractive index (n _co ) of the core for light having a wavelength of 1550 nm is 2.1 or more, and the value obtained by subtracting the refractive index (n _cl ) of the cladding for the light from n _co is 0.06 or more. It is preferable.
If n _co is less than 2.1, γ may be small. More preferably, it is 2.2 or more.
If (n _co -n _cl ) is less than 0.06, the effective area (A) may be too large. More preferably, it is 0.08 or more.
It is preferable that the relationship of the following equation is established between n _cl and n _co .
0.0005 ≦ (n _co −n _cl ) / n _co ≦ 0.2.

  The core diameter and the cladding diameter of the optical fiber of the present invention are typically 0.5 to 3 μm and 50 to 500 μm, respectively. When the optical fiber of the present invention is used for the nonlinear optical fiber of the optical Kerr switch of the present invention, the core diameter is typically 2 μm or less.

  The mode field diameter (D) for light having a wavelength of 1550 nm of the optical fiber of the present invention is typically 1.5 to 3 μm when it is used for the nonlinear optical fiber of the optical Kerr switch of the present invention. If it is less than 1.5 μm, the connection loss when fused with a single-mode silica-based optical fiber may not be reduced. If it exceeds 3 μm, A may increase and γ may decrease.

The A is preferably 10 μm ² or less. If A exceeds 10 μm ² , large γ may be difficult to obtain, and more preferably 5 μm ² or less. A is preferably 2 μm ² or more. If A is less than 2 μm ² , connection loss may increase when connecting to a single mode fiber made of quartz glass. More preferably, it is 3 μm ² or more.

In the case where both the core and the clad of the first optical fiber of the present invention are made of glass or in the second optical fiber of the present invention, the absolute value of the difference ΔTg obtained by subtracting the Tg of the clad glass from the Tg of the core glass | ΔTg | is preferably 15 ° C. or less. If | ΔTg | exceeds 15 ° C., molding during fiber production may be difficult. | ΔTg | is more preferably 10 ° C. or less, and particularly preferably 5 ° C. or less.
Further, the Tg of the core glass is preferably equal to or higher than the Tg of the clad glass, that is, ΔTg ≧ 0 ° C. If ΔTg is less than 0 ° C., the drawing temperature becomes high, and crystal precipitation may occur in the core, or the risk thereof increases. More preferably, it is 2 degreeC or more, Most preferably, it is 5 degreeC or more. If it exceeds 10 ° C., the difference in viscosity-temperature characteristics between the core glass and the clad glass becomes large, and there is a possibility that the fiber cannot be easily produced. More preferably, ΔTg is 2 to 10 ° C.

The value Δα obtained by subtracting α of the cladding glass from the average linear expansion coefficient (α) of the core glass at 50 to 300 ° C. is preferably −3 × 10 ⁻⁷ to 15 × 10 ⁻⁷ / ° C. If Δα is outside this range, problems such as thermal cracking may occur during preform production. More preferably, it is 0-10 * 10 < ^-7 > / degreeC, Most preferably, it is 0-5 * 10 < ^-7 > / degreeC.

  As a preferred embodiment of the optical fiber of the present invention, the portion where the diameter of the nonlinear optical fiber exceeds (D + 1 μm) is Fe, Ni, Co, Mn, Cr, V, Cu, Pr, Nd, Sm, Eu, Er, Tb, A light-absorbing region containing one or more light-absorbing elements selected from the group consisting of Dy, Ho, Tm and Yb, the radial width of the region is at least 5 μm, and the diameter of the nonlinear optical fiber is Examples of the portion that is (D + 1 μm) or less include those that do not substantially contain a light absorbing element. This preferred embodiment (hereinafter referred to as this embodiment) will be described below by taking as an example the case of applying to the second optical fiber of the present invention. Note that D is a mode field diameter for light having a wavelength of 1550 nm, and the radial width is, for example, a half of the difference between the diameters of the two circles in a region sandwiched between two concentric circles. May be described.

This aspect is suitable when it is desired to reduce the variation in connection loss when the optical fiber of the present invention is connected to another optical fiber by power alignment fusion. That is, the existence of a cladding mode is considered as a major factor in the variation in the connection loss. In this embodiment, the cladding mode absorption is achieved in the light absorption region existing in the portion where the diameter exceeds (D + 1 μm). Yes.
The width of the light absorption region is set to 5 μm or more so that the absorption in the cladding mode does not become insufficient. Typically, it is 10 μm or more.

Further, the increase in transmission loss is prevented by making the portion whose diameter is (D + 1 μm) or less substantially not contain a light absorbing element. Usually, the portion where the diameter of the optical fiber is (D + 3 μm) or less is substantially free of a light absorbing element.
The diameter of the region which does not substantially contain the light absorbing element is usually (D + 50 μm) or less. If it exceeds (D + 50 μm), the cladding mode absorption may be insufficient. When D is 5 μm or less, the diameter is typically 45 μm or less.

A typical cross section of the clad has a concentric layer structure in the radial direction of the optical fiber, and a two-layer structure is typical. The core side (inner side) layer of the two-layer structure is a layer not containing a light absorbing element, and the outer layer is a light absorbing region.
The inner layer must cover the cladding in the region where the diameter of the optical fiber is (D + 1 μm) or less. The width is typically 2 μm or more when the difference between the core diameter and D is less than 0.5 μm.

Co or Er is typical as the light absorbing element.
The content of the light absorbing element in the light absorbing region should be appropriately selected according to the purpose of use of the optical fiber of the present invention, and this selection method will be described below using the case where the light absorbing element is Co or Er as an example. I will explain it.

When the light absorption region contains only Co as a light absorption element, the attenuation effect on the light having a wavelength of 1550 nm by the region was examined. The attenuation rate α ′ (unit: dB) and the mass of Co in the region The following equation is established between the percentage display content m (unit: ppm) and the optical fiber length z (unit: m), and a in the following equation is 1.6 × 10 ^−1. I found.
α ′ = a × m × z
The same was recognized when Er was contained instead of Co, and a according to Er was 3.9 × 10 ⁻³ .
Α ′ is preferably 20 dB or more, and more preferably 40 dB or more.

When the length of the optical fiber is 2 m and only Co is contained, the m is 125 ppm or more, and when only Er is contained, the α is 40 dB or more by setting m to 5060 ppm or more. It becomes possible.
On the other hand, the light absorbing element content in the light absorbing region is preferably 50000 ppm or less. If it exceeds 50,000 ppm, the glass may become unstable. More preferably, it is 10,000 ppm or less.

Since the clad of this embodiment has a constant composition in the radial direction, its refractive index is usually not constant in the same direction. Therefore, the n _cl is the refractive index of the inner layer when the clad cross section has a two-layer structure, and the refractive index of the outer layer is usually different from n _cl .

The raw materials were prepared and mixed so as to have a composition represented by mol% in the columns from Bi ₂ O ₃ to CeO _{2 in the} table, to prepare 250 g of the prepared raw material. This prepared raw material is put into a platinum crucible and melted by holding at 1100 ° C. for 2 hours in the air atmosphere, and the resulting molten glass is poured out into a plate shape, and then continuously at a temperature indicated in ° C. in the column of Tg for 4 hours. After holding, it was gradually cooled to room temperature. Examples 1 to 4 are glasses suitable for the core glass of the second optical fiber of the present invention, and Examples 5 and 6 are glasses suitable for the clad glass of the optical fiber. A glass of Example 6Co having a composition in which the glass of Example 6 was used as a mother glass and 150 ppm of Co was added thereto in terms of parts by mass was produced in the same manner.

For the glasses of Examples 1 to 6, Tg (unit: ° C.) and ΔT (unit: ° C.) were determined by differential thermal analysis, and the average linear expansion coefficient α (unit: × 10 ⁻⁷ / ° C.) at 50 to 300 ° C. was determined as the thermal machine. The density d (unit: g / cm ³ ) was measured by an analyzer (TMA) using a specific gravity measuring device SGM-300 manufactured by Shimadzu Corporation.
Further, after processing the obtained glass into a plate shape having a thickness of 2 mm and a size of 20 mm × 20 mm, both surfaces thereof are mirror-polished to obtain a sample plate, and a refractive index n with respect to light having a wavelength of 1550 nm is used as a model 2010 prism coupler manufactured by Metricon. Measured.
The above measurement results are shown in the table ("-" in the table indicates that measurement was not performed).
Further, Tg and n of the glass of Example 6Co were measured in the same manner, and were 366 ° C. and 2.127, respectively.

An optical fiber 1 having the glass of Example 2 as the core glass and the glass of Example 5 as the clad glass was produced as follows. The optical fiber 1 is the second optical fiber of the present invention.
First, the glass of Example 2 was melted and poured into a SUS310S tea tube mold (cylindrical mold having a bottom surface) having an inner diameter of 10 mm and a height of 180 mm, and slowly cooled to obtain a rod-shaped glass. The melting was performed in the same manner as described above.
Next, this rod-shaped glass was redrawn at 430 ° C. to obtain a rod glass having a diameter of 5 mm.

On the other hand, four glass tubes made of the glass of Example 5 having an outer diameter of 15 mm, an inner diameter of 7 mm, and a height of 150 mm were produced.
The rod glass was put into one of the glass tubes, and the glass tube and the rod glass were simultaneously redrawn at 440 ° C. while keeping the inside of the glass tube with a vacuum pump to produce a glass rod having a diameter of 5 mm.
This glass rod was put into one of the remaining three glass tubes and redrawn in the same manner to produce a glass rod having a diameter of 5 mm.

This was repeated two more times using the remaining two glass tubes to produce a preform having a core diameter of 0.06 mm and an outer diameter of 5 mm.
This preform was drawn at 445 ° C. to produce an optical fiber 1 having a core diameter of 1.72 μm and a fiber diameter of 125 μm.

Since the D of the optical fiber 1 is 2.0 μm, the effective area A is 3.3 μm ² .

  The γ of the optical fiber 1 was measured by four-wave mixing. That is, light having a wavelength of 1550 nm is used as pump light, and signal light having wavelengths away from the pump light wavelength in increments of 1559.5 nm, 1549 nm, 1548.5 nm and 0.5 nm is simultaneously transmitted through the coupler to an optical fiber (length 2 m). The output was observed with an optical spectrum analyzer. At this time, the ratio r between the idler light and the signal light was measured.

When the value of γ was estimated from the following relational expression between r and γ, γ was 1330 W ⁻¹ km ⁻¹ . In the following relational expression, P _av is the average pump power passing through the optical fiber, and z is the length of the optical fiber.
r = (γP _av z) ² .
The nonlinear refractive index n ′ at the wavelength of 1550 nm of the optical fiber 1 is estimated to be 1.1 × 10 ⁻¹⁸ m ² W ⁻¹ from the values of A and γ.

An optical fiber 2 in which the glass of Example 2 is a core glass, the glass of Example 6 is a clad inner layer glass, the glass of Example 6Co is a clad outer layer glass, and the clad is a two-layer structure is as follows. Produced. The optical fiber 2 is the optical fiber of the present aspect.
First, a rod-like glass made of the glass of Example 2 was produced in the same manner as in the production of the optical fiber 1, and this rod-like glass was redrawn at 430 ° C. to obtain a rod glass having a diameter of 3 mm.
On the other hand, two glass tubes made of the glass of Example 6 having an outer diameter of 15 mm, an inner diameter of 4 mm, and a height of 150 mm, and one glass tube of the same shape made of the glass of Example 6Co were prepared.

The rod glass was put into one of the glass tubes made of the glass of Example 6, and the glass tube and the rod glass were simultaneously redrawn at 440 ° C. while keeping the inside of the glass tube in a vacuum with a vacuum pump. A glass rod was produced.
This glass rod was put into the remaining one of the glass tubes made of the glass of Example 6, and similarly redrawn to produce a glass rod having a diameter of 3.5 mm.

This glass rod having a diameter of 3.5 mm was put into a glass tube made of the glass of Example 6Co and redrawed at 440 ° C. while maintaining a vacuum in the same manner as described above. The core diameter was 0.059 mm and the outer diameter was 5. A preform of 1 mm was produced.
This preform was drawn at 440 ° C. to produce an optical fiber 2 having a core diameter of 1.64 μm and a fiber diameter of 125 μm.
The outer diameter of the inner layer of the clad of the optical fiber 2, that is, the diameter of the boundary between the inner layer and the outer layer is 29.4 μm, the outer layer is a light absorption region, and its width is 47.8 μm.

The D and A of the optical fiber 2 are 1.97 μm and 3.04 μm ² , respectively.
Further, the γ and n 'optical fiber 1 when the similar manner, respectively ^1100W -1 ^miles -1 was estimated for, was ^{8.9 × 10 -18 m 2 W -1} .

Optical fibers 1 and 2 having a length of 4 m were prepared. Light having a wavelength of 1550 nm and a power of −30 dBm (= 0.001 mW) was input thereto, and the output was monitored with a power meter. After that, 50 cm was cut for each optical fiber, the length was shortened, and the output was monitored in the same manner until the fiber length reached 50 cm. The output was plotted against each fiber length, and propagation loss was estimated from the slope (cutback method). As a result, the propagation losses of the optical fibers 1 and 2 were 1.9 dB / m and 0.8 dB / m, respectively.
The propagation loss of the optical fiber 2 is smaller than that of the optical fiber 1 because the ΔTg in the former is 3 ° C. whereas the ΔTg in the latter is 0 ° C. As a result, the drawing temperature in the former is reduced to that in the latter. This is considered to be because 440 ° C., which is 5 ° C. lower than the temperature, was achieved, and crystal precipitation in the core was suppressed.

In the measurement by the four-wave mixing, the optical fibers 1 and 2 are fusion-spliced with a high NA silica-based optical fiber (UHNA4 manufactured by Nuferan) each having an NA of 0.35, and the UHNA4 is standard single-mode silica-based light. This was performed using a fusion spliced fiber.
The fusion splicing of the optical fibers 1 and 2 and the UHNA 4 was performed by the power alignment method, but the dispersion of the connection loss was also measured by performing fusion splicing for each fiber 20 times. The results are shown in Table 2, and it can be seen that the variation in connection loss is significantly reduced in the optical fiber 2 in which the clad has a two-layer structure and the outer layer is a light absorption region.

  The optical switch of the present invention may be used for, for example, OTDM type ultra-high speed optical switching. Further, the optical fiber of the present invention may be used for the optical switch of the present invention, for example.

The figure for demonstrating the optical car switch of this invention.

Claims (18)

At least linearly polarized signal light is incident on the one end of the nonlinear optical fiber that can receive linearly polarized signal light and linearly polarized control light from one end thereof, and light emitted from the other end of the optical fiber is guided to the analyzer to control light. Is an optical Kerr switch that transmits or blocks or blocks or transmits signal light according to whether or not it is incident simultaneously with the signal light, and has a nonlinear coefficient of 470 W ⁻¹ km ⁻¹ or more for light having a wavelength of 1550 nm of the nonlinear optical fiber. Is an optical car switch.   The optical Kerr switch according to claim 1, wherein the length of the nonlinear optical fiber is 10 m or less.   The optical car switch according to claim 1 or 2, wherein the peak power of the control light is 500 mW or less.   4. The optical Kerr switch according to claim 1, wherein the product of the peak power of the control light and the length of the nonlinear optical fiber is 5 W · m or less.   The optical Kerr switch according to any one of claims 1 to 4, wherein both the core and the cladding of the nonlinear optical fiber are glass.   The optical car switch according to claim 5, wherein the glass transition point of the core glass is 300 ° C or higher. 7. The optical Kerr switch according to claim 5, wherein the core of the nonlinear optical fiber is expressed in terms of mol% based on the following oxide, Bi ₂ O ₃ 60 to 70%, B ₂ O ₃ 10 to 15.5%. , Ga ₂ O ₃ 6.5 to 12.5%, In ₂ O ₃ 1 to 10%, ZnO 1 to 10%, MgO + CaO + SrO + BaO 0.5 to 7%, CeO ₂ 0 to 2%. And the clad of the optical fiber is expressed in terms of mol% based on the following oxide, Bi ₂ O ₃ 50-60%, B ₂ O ₃ 18-28%, Ga ₂ O ₃ 4-15%, In ₂ O _{3 1~10%, ZnO 1~10%, MgO} + CaO + SrO + BaO 1~8%, Li 2 O + Na 2 O + K 2 O 0.1~2%, optical Kerr switch is a glass consisting essentially of CeO 2 0~2%,.   A portion where the diameter of the nonlinear optical fiber is greater than (D + 1 μm) is Fe, Ni, Co, Mn, Cr, V, Cu, Pr, Nd, Sm, Having a light absorbing region containing one or more light absorbing elements selected from the group consisting of Eu, Er, Tb, Dy, Ho, Tm and Yb, the radial width of the region being at least 5 μm, 8. The optical Kerr switch according to claim 5, 6 or 7, wherein a portion of the nonlinear optical fiber having a diameter of (D + 1 μm) or less contains substantially no light absorbing element.   9. The optical car switch according to claim 8, wherein the light absorbing element contained in the light absorbing region is Co, and the content expressed in mass parts per million of the Co in the region is 125 to 50000 ppm. Nonlinear optical fiber nonlinear coefficient for light having a wavelength 1550nm is ^1100W -1 ^{miles -1} or more.   The nonlinear optical fiber according to claim 10, wherein both the core and the clad are glass.   The nonlinear optical fiber according to claim 11, wherein the glass transition point of the core glass is 300 ° C. or higher. The core is expressed in terms of mol% based on the following oxide, Bi ₂ O ₃ 60 to 70%, B ₂ O ₃ 10 to 15.5%, Ga ₂ O ₃ 6.5 to 12.5%, In ₂ O ₃ 1 10%, ZnO 1 to 10%, MgO + CaO + SrO + BaO 0.5 to 7%, CeO ₂ 0 to 2%, and the cladding is represented by mol% on the basis of the following oxide, Bi ₂ O _{3 50~60%, B 2 O 3 18~28} %, Ga 2 O 3 4~15%, In 2 O 3 1~10%, 1~10% ZnO, MgO + CaO + SrO + BaO 1~8%, Li 2 O + Na 2 O + K 2 A nonlinear optical fiber which is a glass consisting essentially of O 0.1-2% and CeO ₂ 0-2%. The nonlinear optical fiber according to claim 13, wherein a nonlinear coefficient for light having a wavelength of 1550 nm is 1100 W ⁻¹ km ⁻¹ or more. The refractive index of the core for light having a wavelength of 1550 nm is 2.1 or more, and the value obtained by subtracting the refractive index n _{cl of the} cladding for the light from the refractive index of the core is 0.06 or more. The nonlinear optical fiber described.   The nonlinear optical fiber according to claim 10, wherein the core diameter is 2 μm or less.   The nonlinear optical fiber according to any one of claims 10 to 16, wherein a portion where the diameter of the nonlinear optical fiber exceeds (D + 1 μm) is Fe, Ni, Co, where D is a mode field diameter for light having a wavelength of 1550 nm. , Mn, Cr, V, Cu, Pr, Nd, Sm, Eu, Er, Tb, Dy, Ho, Tm, and a light absorbing region containing one or more light absorbing elements selected from the group consisting of Yb A portion in which the width of the region in the radial direction is at least 5 μm or more and the diameter of the nonlinear optical fiber is (D + 1 μm) or less substantially does not contain a light absorbing element. The nonlinear optical fiber according to claim 17, wherein the light-absorbing element contained in the light-absorbing region is Co, and the Co content in parts per million by mass in the region is 125 to 50000 ppm.

Images