JP4806114B2 - Optical wavelength conversion element, laser light generation apparatus and optical information processing apparatus using the same - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a wavelength conversion element used in the field of optical information processing or optical applied measurement using a coherent light source, a coherent light generation apparatus using the same, and an optical information processing apparatus using the coherent light generation apparatus.
[0002]
[Prior art]
Optical wavelength conversion elements using the nonlinear optical effect are used in many fields because the wavelength used for laser light sources can be expanded by converting the wavelength of light by wavelength conversion. For example, in the wavelength conversion using the second harmonic, the laser light is wavelength-converted to the second harmonic having a half wavelength, thereby realizing a short wavelength light that has been difficult in the past. Furthermore, when parametric oscillation is used, light having different wavelengths can be continuously generated from a light source having a single wavelength, and a wavelength variable light source can be realized. In addition, if the sum frequency is used, two lights having different wavelengths can be converted into light having a third wavelength.
[0003]
In wavelength conversion of light using such a nonlinear optical effect, it is necessary to satisfy a phase matching condition between a fundamental wave before conversion and a harmonic after conversion. For this purpose, for example, the birefringence method that uses the birefringence of the crystal to align the propagation speed in the crystal between the fundamental wave and the harmonic, and the pseudo phase that uses a nonlinear grating to achieve phase matching. There is a matching method.
[0004]
However, in practice, since the tolerance of the wavelength that satisfies these phase matching conditions is extremely narrow, it is necessary to control the wavelength of the fundamental wave with very high accuracy, and it is difficult to stabilize the output.
[0005]
Therefore, studies have been made to increase the stability of optical wavelength conversion by expanding these wavelength tolerances. FIG. 19 shows a configuration diagram of a conventional optical wavelength conversion element for the purpose of expanding the wavelength tolerance (see Japanese Patent Application No. 3-16198). Hereinafter, generation of the second harmonic P2 having a wavelength of 0.42 μm with respect to the fundamental wave P1 having a wavelength of 0.84 μm will be described in detail with reference to FIG.
[0006]
In the configuration of FIG. 19, LiNbO_ThreeAn optical waveguide 6102 is formed on the substrate 6101, and a layer 6103 (polarization inversion layer) whose polarization is periodically inverted is formed in the optical waveguide 6102. By compensating for mismatch in propagation constant between the fundamental wave P1 and the generated harmonic wave P2 by the periodic structure of the domain-inverted layer 6103, the second harmonic wave P2 can be generated with high efficiency.
[0007]
As described above, the optical wavelength conversion element that performs wavelength conversion by the periodically domain-inverted layer 6103 has high conversion efficiency, but has a very narrow phase matching wavelength tolerance that enables wavelength conversion. Therefore, in the configuration of FIG. 19, the wavelength tolerance of the optical wavelength conversion element is increased by partially changing the propagation constant of the optical waveguide 6102. When the propagation constant of the optical waveguide 6102 is changed, the phase matching wavelength in the optical waveguide 6102 changes. The phase matching condition is a condition that enables wavelength conversion in the wavelength conversion element, and the wavelength of incident light that satisfies this condition is called a phase matching wavelength. Therefore, when the width of the optical waveguide 6102 is partially changed to the regions A, B, C, and D, the phase matching wavelength varies depending on the width of the optical waveguide 6102 in each region. For this reason, even if the wavelength of the incident light changes, the phase matching condition is satisfied in any of the regions A to D having different optical waveguide widths, so that the phase matching wavelength of the entire element increases. As a result, the wavelength tolerance of the optical wavelength conversion element increases, and a stable wavelength conversion element can be manufactured. The phase matching condition between the regions A to D can be realized by changing the depth of the optical waveguide 6102 in each region A to D or the period of the domain-inverted layer 6103 between the regions A to D. In these cases as well, an optical wavelength conversion element having a large wavelength tolerance can be obtained.
[0008]
Furthermore, a configuration in which a periodic domain-inverted structure and a phase control unit are combined has also been reported (Japanese Patent Application No. 4-070726). FIG. 20 shows a configuration of a conventional optical wavelength conversion element that realizes an increase in tolerance by such a method.
[0009]
The optical wavelength conversion element illustrated in FIG. 20 includes a plurality of polarization inversion regions 6105 and a phase control unit 6106 formed between the polarization inversion regions 6105 on the nonlinear optical crystal 6101. By utilizing the difference in the phase matching conditions in each polarization inversion region 6105 to increase the phase matching wavelength tolerance, and adjusting the phase mismatch generated between the polarization inversion regions 6105 by the phase control unit 6106, The output fluctuation of the harmonic P2 is reduced with respect to the wavelength fluctuation of the fundamental wave P1.
[0010]
Furthermore, by increasing the number of domain-inverted regions 6105, the phase matching wavelength tolerance can be expanded over a wider wavelength range. For example, when the domain-inverted region 6105 has a three-divided structure (n = 3) and a four-divided structure (n = 4), a phase matching characteristic showing a relationship between a fundamental wave wavelength and a second harmonic (SHG) output ( The tuning curves are shown in FIGS. 21 (a) and 21 (b), respectively. This shows that the wavelength tolerance can be greatly expanded by increasing the number of divisions.
[0011]
In addition, attempts have been reported to increase the phase matching wavelength tolerance by modulating the periodic structure of polarization inversion.
[0012]
For example, Sugawara et al., IEEE Journal of Quantum Electronics, vol. 26, pp. 1265 to 1276 and 1990, a method for expanding the tolerance of the phase matching wavelength by changing the periodic structure of polarization inversion to a chirp shape has been reported. Specifically, here, a method of expanding the phase matching wavelength tolerance by a linear chirp structure in which the period of polarization inversion is increased in proportion to the distance has been proposed. In this case, the phase matching curve can be greatly increased by the polarization inversion structure in which the phase shift changes linearly.
[0013]
[Problems to be solved by the invention]
The problem of the method of expanding the phase matching wavelength tolerance in the conventional optical wavelength conversion element using the polarization inversion structure is that the phase matching characteristic, that is, the shape of the SHG output characteristic when the phase matching wavelength is tuned, fluctuates in the vicinity of the peak. It is a point to do.
[0014]
As described above, in an optical wavelength conversion element based on a domain-inverted layer, the wavelength tolerance of the wavelength conversion element conversion element can be increased by dividing the element into two or more regions and changing the phase matching conditions between the regions. In the method of increasing, since the phase matching wavelength in each region is different, the second harmonic is generated in a wide wavelength range. However, since the second harmonics generated in each region interfere with each other, as shown in FIGS. 21 (a) and (b), the output fluctuation of the harmonics with respect to the wavelength fluctuation of the fundamental light is near the output peak of SHG. large.
[0015]
In addition, although the phase matching wavelength can be significantly increased by making the polarization inversion period a linear chirp structure, a large ripple is generated near the peak of the tuning curve even with this method.
[0016]
As described above, the conventional method cannot obtain a flat output characteristic in the vicinity of the peak of the tuning curve. For this reason, the harmonic output changes with respect to a slight fluctuation of the fundamental wave wavelength within the phase matching wavelength tolerance, and a stable output cannot be obtained.
[0017]
Furthermore, by providing a phase control unit between the domain-inverted regions, interference between harmonics generated between the domain-inverted regions is reduced, and conventional light that aims to reduce harmonic output fluctuations with respect to fundamental wavelength fluctuations. In the wavelength conversion element, the harmonic output still has a fluctuation of 10% or more. For this reason, it is possible to realize flatness near the peak value necessary to realize the stabilization of the output. ,Have difficulty.
[0018]
Furthermore, in the conventional optical wavelength conversion element, it is easy to expand the wavelength tolerance over a relatively wide range, but the conversion efficiency of the wavelength conversion element is drastically reduced. For example, when crystals having the same length are used, the conversion efficiency is reduced to 1/9 or more. That is, with the conventional technology, it is difficult to increase the wavelength tolerance while maintaining highly efficient wavelength conversion characteristics.
[0019]
The present invention has been made in order to solve the above-mentioned problems. The purpose of the present invention is to (1) maintain the phase matching characteristic (peak flat characteristic) having a wide flatness in the vicinity of the peak while maintaining the wavelength tolerance. Realizing expansion and providing a light wavelength conversion element having stable wavelength conversion characteristics, (2) By configuring a coherent light generation device using the light wavelength conversion element and the semiconductor laser as described above, Providing a coherent light generation device that stabilizes oscillation wavelength fluctuation in a semiconductor laser and has stable output characteristics; and (3) providing an optical information processing device using the coherent light generation device as described above. With the goal.
[0020]
[Means for Solving the Problems]
The optical wavelength conversion element of the present invention includes a nonlinear optical crystal and a periodic domain-inverted structure formed in the nonlinear optical crystal, and the domain-inverted structure has a single period Λ.₀And a chirp period portion having a gradually changing period, whereby the above object is achieved.
[0021]
For example, the single period portion is located substantially near the center of the domain-inverted structure, and the chirp period is located near both ends of the domain-inverted structure.
[0022]
Another optical wavelength conversion element of the present invention comprises a nonlinear optical crystal and a periodic domain-inverted structure formed in the nonlinear optical crystal, and the period of the domain-inverted structure is Λ_-m, Λ_-(m-1), ..., Λ_-2, Λ_-1, Λ₀, Λ₁, Λ₂, ..., Λ_m-1, Λ_mThe polarization inversion period has a phase mismatch amount distribution f (z), and the distribution f (z) is
f (i * Λ₀) = (Λ₁+ Λ₂+ ... + Λ_i) -I * Λ₀,as well as
f (−i * Λ₀) = (Λ_-1+ Λ_-2+ ... + Λ_-i) -I * Λ₀,
However, i = 1, 2, 3,...
In the vicinity of z = 0, f (z) = 0, and the distribution f (z) is expressed as f (i * Λ₀) = − F (−i * Λ₀And the second order differential coefficient increases toward the end of the nonlinear optical crystal in at least a part of the distribution f (z), whereby the above-mentioned object is achieved. The
[0023]
In one embodiment, when the total length of the domain-inverted structure is L, the distribution f (z) is
f (z) = a · | sin (bz) |^m            z> 0
f (z) = − a · | sin (bz) |^m          z <0
And satisfies the relationship 2 <m <6 and b · L / 2 <0.5 · π.
[0024]
In another embodiment, the distribution f (z) is
f (z) = a · | z |^m                        z> 0
f (z) = − a · | z |^m                      z <0
And satisfies the relationship 2 <m <4.
[0025]
Still another optical wavelength conversion element of the present invention includes a plurality of nonlinear optical crystals having the same polarization inversion structure, and a phase control unit disposed between the nonlinear optical crystals, the phase control unit comprising: The nonlinear optical crystal is composed of a domain-inverted structure having a different period, and thereby the above-mentioned object is achieved.
[0026]
Preferably, when the total length of the domain-inverted structure is L, the absolute value | f (L / 2) / Λ of a value obtained by standardizing the phase mismatch amount at both ends of the optical wavelength conversion element₀| Has a value of 0.4 to 1.
[0027]
In the above configuration, the fundamental wave is converted into a harmonic in the nonlinear optical crystal, and the propagation loss of the fundamental can be approximately half of the propagation loss of the harmonic.
[0028]
According to another aspect of the present invention, coherent light comprising the optical wavelength conversion element of the present invention as described above and a laser light source, and the light of the laser light source being wavelength-converted by the optical wavelength conversion element. A generator is provided.
[0029]
For example, the laser light source is a semiconductor laser having a wavelength variable function.
[0030]
For example, the semiconductor laser is superimposed on a high frequency.
[0031]
Preferably, the wavelength tolerance of the optical wavelength conversion element is wider than the longitudinal mode interval of the semiconductor laser.
[0032]
According to another aspect of the present invention, the coherent light generation device of the present invention as described above and a condensing optical system are provided, and the coherent light emitted from the coherent light generation device is transmitted by the condensing optical system. A focused optical information processing apparatus is provided.
[0033]
DETAILED DESCRIPTION OF THE INVENTION
The present invention relates to an element structure for the purpose of stabilizing an output in an optical wavelength conversion element using a nonlinear optical effect, and a coherent light source and an optical information processing apparatus using the element structure. Specifically, a new method for realizing phase matching characteristics having a wide wavelength tolerance and a flat peak by using a special structure for the periodic structure of polarization inversion constituting the optical wavelength conversion element. provide.
[0034]
More specifically, the following four items:
・ Nonlinear optical element with polarization inversion structure combining single periodic structure and linear chirp structure
.Nonlinear optical element when the polarization inversion periodic structure is modulated using the function f (z)
A structure that can always realize stable output characteristics in a coherent light source composed of a semiconductor laser and an optical wavelength conversion element
・ Structure of optical information processing device with stable characteristics by coherent light generator with stable output and condensing optical system
Will be described using embodiments.
[0035]
(About the principle of pseudo phase matching)
First, optical wavelength conversion and phase matching in the quasi phase matching type SHG element will be described.
[0036]
The wavelength conversion of light using the second-order nonlinear optical effect includes methods such as second harmonic generation, parametric generation, sum frequency generation, difference frequency generation, and the like. The wavelength conversion of the light is performed by the interaction of the three lights having the wavelength λ3. At this time, the wavelength of light is
1 / λ1 = 1 / λ2 + 1 / λ3 (1)
Must satisfy the relationship.
[0037]
For example, the second harmonic generation is in the case of λ2 = λ3. At this time, the expression (1) is
1 / λ1 = 2 / λ2 (2)
Thus, light having a half wavelength λ2 with respect to the wavelength λ1 of the fundamental wave is generated.
[0038]
Parametric generation is a phenomenon in which light of wavelengths λ2 and λ3 is generated with respect to the fundamental wave of wavelength λ1. A phenomenon in which light of wavelength λ1 is generated with respect to light of wavelength λ2 and light of wavelength λ3 is a sum frequency, while light having wavelength λ3 is generated from light of wavelength λ1 and light of wavelength λ2. Is the difference frequency generation.
[0039]
In these wavelength conversions, what is necessary for realizing highly efficient wavelength conversion is the establishment of the phase matching condition. The phase matching means that the phase relationship of the wavelengths λ1, λ2, and λ3 of the three lights related to wavelength conversion is matched in a propagation medium (for example, a nonlinear optical crystal, an optical waveguide, etc.)
N1 · 2π / λ1 = N2 · 2π / λ2 + N3 · 2π / λ3 (3)
Satisfy the relationship. Here, N1, N2, and N3 are refractive indexes (effective refractive indexes) of light actually sensed by light of wavelengths λ1, λ2, and λ3, and are different in the case of an optical waveguide and in the case of a bulk crystal. It depends on the polarization direction.
[0040]
The phase relationship between the three wavelengths in the second-order nonlinear optical effect is as follows:
A · exp {i (−N1 · 2π / λ1 + N2 · 2π / λ2
+ N3 · 2π / λ3) z} (4)
The light conversion efficiency with respect to the distance is proportional to the square of the definite integral value with respect to the distance z in Equation (4). For example, the expression (4) where the condition of the expression (3) is not satisfied is
A · exp {i · β · z} (5)
The integrated value becomes a vibration function and does not exceed the value of A / β.
[0041]
On the other hand, when the condition of Expression (3) is satisfied, Expression (4) is
A (6)
The integrated value is A · z, which increases in proportion to the distance. When this state satisfies the phase matching condition, the conversion efficiency increases in proportion to the square of the distance, enabling highly efficient conversion.
[0042]
However, in general, all substances have chromatic dispersion, and the refractive index varies depending on the wavelength of light. Therefore, it is difficult to satisfy the expressions (1) and (2) at the same time. A method for solving this is quasi-phase matching using periodic inversion of nonlinear polarization. In the quasi-phase matching, by providing a nonlinear grating structure with a period Λ, it is possible to add a propagation vector due to the grating to the relationship of the propagation constant in Expression (3). That is, when a domain-inverted structure having a period Λ is formed in a nonlinear optical crystal, the phase relationship of Equation (4) is
A ′ · exp {i (−N1 · 2π / λ1 + N2 · 2π / λ2
+ N3 · 2π / λ3-q · 2π / Λ) z} (7)
However, q = 1, 2, 3,...
When q = 1, it is called a primary periodic structure, and when q = 2, it is called a secondary periodic structure. That is, the phase matching condition can be established by canceling out the phase mismatch amount (−N1 · 2π / λ1 + N2 · 2π / λ2 + N3 · 2π / λ3) by the nonlinear grating structure.
[0043]
In the case of the primary periodic structure, the polarization inversion period for satisfying the quasi-phase matching condition is as follows:
-N1 · 2π / λ1 + N2 · 2π / λ2 + N3 · 2π / λ3
= 2π / Λ (8)
Can be derived. For example, in the case of second harmonic generation, from equation (8)
Λ = λ1 / 2 (N2-N1) (9)
Is obtained.
[0044]
(About phase mismatch)
The inventors of the present application have found that the phase matching wavelength tolerance can be efficiently controlled by controlling the phase matching mismatch state in the periodically poled structure. That is, as shown in FIG._-m, Λ_-(m-1), ..., Λ_-2, Λ_-1, Λ₀, Λ₁, Λ₂, ..., Λ_m-1, Λ_mThe domain-inverted region 601 having a domain-inverted period represented by:
L = Λ_-m+ Λ_-(m-1)+ ... + Λ_-2+ Λ_-1+ Λ₀+ Λ₁+ Λ₂+ ... + Λ_m-1+ Λ_m
And Further, f (z) is assumed as a distribution function of the phase mismatch amount,
f (i * Λ₀) = (Λ₁+ Λ₂+ ... + Λ_i) -I * Λ₀,as well as
f (−i * Λ₀) = (Λ_-1+ Λ_-2+ ... + Λ_-i) -I * Λ₀,
However, i = 1, 2, 3,...
And When the phase mismatch amount f (z) is normalized,
f (z) / Λ₀                                        (10)
It is represented by The following phase mismatch amounts are the above normalized values.
[0045]
Period Λ of domain-inverted structure₀The phase matching condition is expressed by equation (9), and the phase matching condition is satisfied for the fundamental wavelength λ0 that satisfies this condition.
[0046]
Next, generation of the phase mismatch amount in the conventional domain-inverted structure will be described.
[0047]
For a fundamental wave with a center wavelength of λ0, the period of the domain-inverted structure is Λ₀With reference to FIGS. 2A and 2B, description will be made of how the phase mismatching amount is generated when shifted from the above.
[0048]
For example, in the case of a periodic structure that changes in a chirp shape as shown in the prior art, FIG. 2A shows a change with a distance in the z direction of the polarization inversion period. Since the period changes linearly, the phase mismatch amount becomes an integral of a linear function and is expressed by a quadratic function, and changes as shown in FIG.
[0049]
In addition, with respect to the divided periodic structure shown in the prior art, as a structure realizing phase matching wavelength tolerance by a three-divided structure, a polarization inversion structure having different periods Λ1, Λ2, and Λ3 and phase control units δ1 and δ2 FIG. 2 (c) shows the change of the polarization inversion period with the distance in the z direction, and FIG. 2 (d) shows the relationship of the phase mismatch amount. The center period is Λ2, and the amount of phase mismatch with respect to the center period is illustrated. In this case, the phase mismatch amount is the sum of the phase mismatch amount generated in the phase controller and the phase mismatch amount generated in the polarization inversion structure having a different period. Further, when a phase shift portion is formed in a periodic domain-inverted structure, phase mismatch can be similarly expressed.
[0050]
On the other hand, in the structure given only the phase shift, the polarization inversion period is constant, so that the polarization inversion period is constant regardless of the distance z as shown in FIG. However, the phase mismatch amount is generated in the phase controller as shown in FIG.
[0051]
(Pseudo phase matching issues)
Here, the tolerance of the phase matching wavelength, which is a problem in the optical wavelength conversion element using pseudo phase matching, will be described.
[0052]
In wavelength conversion by quasi phase matching, phase matching conditions can be established in a pseudo manner by nonlinear grating, and high conversion efficiency can be realized in proportion to the square of the element length. However, the width of the wavelength tolerance that satisfies the phase matching condition decreases in inverse proportion to the action length of the grating. For example, LiNbO with an element length of 10 mm_ThreeIn the quasi-phase matching type SHG element using the above, when the light of wavelength 850 nm is converted into the second harmonic of wavelength 425 nm, for example, the period of polarization inversion is about 3.2 μm. At this time, the wavelength tolerance of the fundamental wave for satisfying the quasi-phase matching condition is 0.1 nm or less at the full width at half maximum. This value is a very severe value when performing stable wavelength conversion, and there is a problem that the output becomes unstable due to environmental changes such as ambient temperature.
[0053]
Several specific embodiments of the present invention will be described below with reference to the accompanying drawings.
[0054]
(Embodiment 1)
Here, the structure of an optical wavelength conversion element having a wide wavelength tolerance will be described. Specifically, the wavelength tolerance of the optical wavelength conversion element is expanded by controlling the distribution of the phase mismatch amount.
[0055]
The inventor of the present application has found a new method for expanding the tolerance of the phase matching wavelength of the optical wavelength conversion element. As shown in FIG. 1A, the optical wavelength conversion element of the present invention expands the phase matching wavelength tolerance by partially changing the period of the periodic domain-inverted region 601. More specifically, as shown in FIG. 1B, phase matching is achieved by combining a single periodic part 603 having a single-period polarization reversal structure and a chirp periodic part 602 having a chirped periodic structure. Increase wavelength tolerance. As a result, it is possible to suppress a change in the output of the harmonic with respect to the wavelength variation of the fundamental wave to an extremely small level while having a wide wavelength tolerance. That is, a peak flat tuning curve is realized. Furthermore, it is possible to minimize a decrease in wavelength conversion efficiency due to an increase in wavelength tolerance and simultaneously achieve an increase in wavelength tolerance and high efficiency characteristics. Further, the degree of freedom in designing the phase matching characteristic is increased, and the enlargement ratio of the wavelength tolerance can be freely designed.
[0056]
Conventionally, it has been shown that phase matching wavelength tolerance can be expanded with several structures. These are, for example, 1) a divided periodic structure (a structure in which polarization inversion structures having different periods are combined) and 2) a chirped periodic structure. However, with the conventional method, the wavelength tolerance can be greatly expanded, but since a large ripple is generated near the peak of the phase matching curve, it is difficult to keep the SHG output stable.
[0057]
On the other hand, in the present invention, by adding a special device to the domain-inverted structure, it is possible to increase the wavelength tolerance and realize a peak-flat phase matching characteristic. That is, the structural difference from the conventional optical wavelength conversion element is that the phase matching characteristic can be designed by controlling the distribution by paying attention to the phase mismatch amount.
[0058]
FIGS. 4A and 4B show the relationship between the polarization inversion period and the phase mismatch amount in the divided periodic structure and the distance in the z direction, respectively. As can be seen from the diagram of the polarization inversion period in FIG. 4A, the two-part divided periodic structure is configured by combining two different periodic structures with respect to the center of the polarization inversion structure. In this case, the occurrence of the phase mismatch amount is represented by a combination of a linear function and a phase shift, as shown in FIG. The phase matching characteristic at this time has a large ripple as shown in FIG.
[0059]
On the other hand, the relationship between the polarization inversion period and the phase mismatch amount in the conventional chirp structure and the distance in the z direction is shown in FIGS. At this time, the phase mismatch amount is generated according to a quadratic function. The phase matching characteristics at this time are shown in FIG. Compared to FIG. 4C, the ripple near the peak in FIG. 5C is considerably reduced. From this, the inventors of the present application speculate that the magnitude of the ripple in the vicinity of the peak of the phase matching characteristic is reduced by generating the phase mismatch amount from the linear function to the quadratic function. did.
[0060]
Therefore, the inventors of the present application have studied a structure in which a phase mismatch amount is generated (distributed) according to several function systems.
[0061]
  First, a structure that generates a phase mismatch amount according to a higher-order power function was studied. As the domain-inverted structure, the period Λ₀Assuming a centrosymmetric structure around the center, the phase matching characteristics were calculated. That is,
f (z) = a · | z |^m          z> 0
f (z) = − a · | z |^m        z <0
And f (z) is designed as a power function of an oddly symmetric function of the domain-inverted periodic structure, and a phase matching curve (tuning curve, that is, phase matching characteristic) having a shape closest to the peak flat is Several values of the order m of the function were obtained. The result is shown in FIG.
[0062]
In FIG. 6, (1) is a = Λ₀× 1.2 × 10^-Four, M = 1, (2) is a = Λ₀× 2.5 × 10^-8, M = 2, (3) is a = Λ₀× 6.3 × 10^-12, M = 3, and (4) a = Λ₀× 3.2 × 10^-15, M = 4, the results are shown respectively.
[0063]
Specifically, in FIG. 6, curve (1) corresponds to the case of a two-divided periodic structure, and is a case where two domain-inverted structures having different periods are combined. A curve (2) corresponds to the case of the linear chirp structure, and shows a case where the polarization inversion period changes linearly. As can be seen from FIG. 6, the curves (1) and (2) both vary greatly in the vicinity of the peak of the phase matching curve, and flat peak characteristics cannot be obtained. On the other hand, in the curve (3), a peak flat characteristic was obtained. Furthermore, when m = 4 (4), the maximum value decreases.
[0064]
From the above results, it has been clarified that in the case of the power function, a peak flat phase matching characteristic can be obtained when the order m is 2 <m <4. In particular, in the vicinity of m = 3, a very flat phase matching characteristic was obtained as shown by curve (3) in FIG. That is, it became clear that peak flat phase matching characteristics can be realized by increasing the value of the order m as compared with the conventional linear chirp structure (in the case of m = 2).
[0065]
Further, it has been found that in order to realize better peak flat characteristics, it is necessary to provide some restrictions on the function f (z) that the distribution of the phase mismatch amount follows.
[0066]
The function f (z) is preferably an odd function having the origin at the center of the domain-inverted structure. This is a structure in which the phase matching characteristic is a symmetric structure and the efficiency is highest. However, when the optical propagation loss and conversion efficiency increase, the harmonic conversion efficiency generated from the domain-inverted structure due to the phenomenon that the fundamental wave is converted into a harmonic and its intensity decreases (pump depletion). In some cases, the center of the domain-inverted structure may be slightly shifted.
[0067]
In addition, FIG. 7 shows a normalized distribution of phase mismatch amounts (specified by equation (10)) when peak flat phase matching characteristics are obtained. Curves (1), (2), (3), and (4) in FIG. 7 correspond to curves (1), (2), (3), and (4) in FIG. 6, respectively.
[0068]
The values of the phase mismatch amounts at both ends of the domain-inverted structure are concentrated in the vicinity of 0.5 to 1, as can be seen from FIG. Although it takes a value close to 2 only in the case of the curve (4), as can be seen from FIG. 6, the phase matching curve in this case takes a value considerably different from the peak flat shape.
[0069]
Thus, regardless of the order m of the function of f (z), when the phase mismatch amount takes a value in the vicinity of 0.5 to 1 at both ends of the domain-inverted structure, the phase matching characteristic has a peak flat shape. It was confirmed to approach. In other ranges, the ripple was too large, and good characteristics could not be obtained.
[0070]
  Further, when the occurrence of the phase mismatch amount shown in FIG. 7 is observed, not only a high-order function but also other function systems can be used as the function f (z) for realizing the peak flat phase matching characteristics. It was speculated. Therefore, an attempt was made to use a trigonometric function as f (z). Specifically, the period Λ₀Is the central period,
f (z) = a · | sin (bz) |^m          z> 0
f (z) = − a · | sin (bz) |^m        z <0
As calculated.
[0071]
First, when the value of b is examined, when the value of b increases, f (z) vibrates because it is a sin function. At this time, the conversion efficiency was greatly reduced. Only when f (z) is an increasing or decreasing function, a peak flat and highly efficient phase matching characteristic was obtained. Therefore, it has become clear that b · (L / 2) <0.5π must be satisfied with respect to the distance L / 2 from the center to the end of the domain-inverted structure.
[0072]
Next, in order to examine the order m, the relationship between the value of m and the phase matching characteristics was calculated as in the case of the polynomial. The results are shown in FIGS. 8A (a) to 8D (b). Specifically, the phase matching curve in the case of m = 2 and the distribution of the normalized phase mismatch amount at that time are shown in FIGS. 8A and 8B as the phase matching curve in the case of m = 3. 8B and FIGS. 8B and 8B show the distribution of the normalized phase mismatch amount in FIG. 8B and the distribution of the normalized phase mismatch amount when m = 4. 8C (a) and (b), the phase matching curve in the case of m = 5 and the distribution of the normalized phase mismatch amount at that time are shown in FIGS. 8D (a) and (b), respectively. Show.
[0073]
Referring to FIGS. 8A (a), 8B (a), 8C (a), and 8D (a), when m = 2 shown in FIG. 8A (a), a ripple occurs in the peak flat portion. However, in the case of m = 3 to 5 shown in FIG. 8B (a), FIG. 8C (a), and FIG. 8D (a), substantially peak flat characteristics were obtained. For example, in the case of m = 3 shown in FIG. 8B (a), almost peak flat characteristics are obtained in curves (3), (4), and (5). The phase mismatch amount at this time is It was about 0.6 to 0.7. Further, in the case of m = 4 shown in FIG. 8C (a), substantially peak flat characteristics were obtained in the curves (3), (4), and (5). The amount of phase mismatch at this time was about 0.6 to 0.7. When m = 5 shown in FIG. 8D (a), substantially peak flat characteristics were obtained in the curves (2), (3), (4), and (5). The amount of phase mismatch at this time was in the vicinity of 0.5 to 0.8.
[0074]
Further, in the case of m = 3 to 5 in which peak flat characteristics are obtained, the values of a and b used in the calculation are shown in each corresponding drawing. From these results, it can be seen that peak flat characteristics can be obtained only when the value of b is set in the range of 0.00005 to 0.0003.
[0075]
On the other hand, although not shown in the figure, when m = 6, the value of the peak value was significantly reduced.
[0076]
Therefore, it was found that peak flat characteristics can be obtained when the order m is in the range of 2 <m <6. However, the value of m is a real number.
[0077]
Further, regarding the distribution of the phase mismatch amount, referring to FIGS. 8A (b), 8B (b), 8C (b), and 8D (b), the normalized phase mismatch at both ends of the domain-inverted structure. It can be seen that the value of the matching amount takes a value of 0.5 to 1. When the normalized value of the phase mismatch amount is 1 or more, the conversion efficiency significantly decreases. When the value is 0.5 or less, the ripple of the phase matching curve increases.
[0078]
Thus, in order to obtain a peak flat phase matching curve, the value of the phase mismatch amount at both ends of the periodic structure must be limited to a range of 0.5 to 1.
[0079]
Similarly, when the power function of the tan (z) function is used as the function f (z) to be followed by the distribution of the phase mismatch amount, the peak flat phase matching characteristic is obtained.
[0080]
As a result, a peak flat characteristic can be realized by using a power function larger than the second order or a trigonometric function (sin function or tan function) as the function f (z) to be followed by the distribution of the phase mismatch amount. It was. Further, a combination of a power function and a trigonometric function, for example, f (x) = axⁿ* | Sin (bx) |^mOr f (x) = axⁿ* | Tan (bx) |^mHowever, peak-flat phase matching characteristics were realized. At this time, peak flat characteristics were obtained in the range of 2 <n * m <6.
[0081]
Next, the phase mismatch amount distribution capable of realizing the peak flat phase matching characteristic was analyzed from FIG. 7, FIG. 8A (b), FIG. 8B (b), FIG. 8C (b), and FIG. 8D (b). .
[0082]
Thus, in order to realize a phase mismatch distribution that realizes a peak-flat phase matching characteristic, a substantially single period Λ matching the center wavelength of the phase matching wavelength is provided at the center of the domain-inverted structure.₀It can be seen that there is a need for a structure in which there is a portion of a single periodic structure having, and the phase mismatch amount gradually increases toward both ends of the domain-inverted periodic structure. Further, in any of the illustrated phase mismatch amount distributions, f (z) has a portion where the slope increases. Further analysis shows that both the power function and the sin function require a value of 2 or more as the order m, and the second-order differential coefficient of f (z) has a portion that increases toward both ends of the domain-inverted structure. Need to be. Further, the normalized value | f (z) / Λ of the phase mismatch amount at both ends of the domain-inverted structure₀The value of | is in the range of 0.5 to 1 in any case. If a phase mismatch distribution satisfying these conditions is given, the phase matching characteristics can be designed in a peak flat shape.
[0083]
Next, a method for realizing the characteristics of the phase mismatch distribution shown in FIG. 7 with a simpler configuration was examined.
[0084]
In order to realize a peak-flat phase matching characteristic, the distribution of the phase mismatch amount shown in FIG. 7 may be given to the domain-inverted periodic structure. That is, the same phase matching characteristics can be obtained by approximately giving the distribution of the phase mismatching amount described above with reference to FIGS. 7 and 8A (a) to 8D (b) to the polarization inversion periodic structure. It is done. 7 and 8A (a) to 8D (b), if the phase mismatch amount distribution realizing the peak flat phase matching characteristics is approximately expressed, the central period Λ is near the center of the domain-inverted structure.₀And a structure in which the amount of phase mismatch increases in the vicinity of both ends of the domain-inverted structure.
[0085]
Therefore, phase mismatch amount distribution characteristics as shown in FIGS. 9A and 9B are set. As the domain-inverted structure, as shown in FIG.₀The polarization inversion period changes in a chirp shape toward the end. As the phase mismatch distribution, as shown in FIG. 10B, the phase mismatch is zero because of the single period near the center, and the phase mismatch increases depending on the distance at both ends. It has a quantity distribution. The slope of the increase in the amount of phase mismatch is also set so as to increase toward both ends. With such a combination of a single periodic structure and a chirp periodic structure, phase matching characteristics substantially equal to those in FIG. 6 can be realized.
[0086]
Furthermore, other structures approximating the phase mismatch distribution shown in FIG. 7 were also examined.
[0087]
That is, the center period is Λ₀And the period Λ toward one end of the domain-inverted structure₀+ Α1 and Λ₀+ Α2 polarization inversion region, period Λ toward the opposite end₀-Α1 and Λ₀A polarization inversion region of -α2 is provided (where α2> α1). This is a five-part polarization reversal structure in which a phase mismatch distribution that changes according to a power function is simplified to a structure in which the period changes between two regions. In this structure, the polarization inversion period changes in two steps from the center toward the end, and the amount of change increases as the distance from the end increases. This is a structure that approximates a power function, and the phase mismatch increases toward the end of the periodic structure.
[0088]
Actually, simulation is performed on the phase mismatch amount distribution having the grating length (distance) dependency as shown in FIGS. 10 (a-2), (b-2), (c-2), and (d-2). The phase matching characteristics in each case were determined, and the characteristics shown in FIGS. 10A-1, B-1, C-1, and D-1 were obtained. Comparing the characteristics depicted in these figures, it is possible to realize a structure that exhibits characteristics close to the peak flat by changing the dependency of the phase mismatch amount on the grating length (distance), thereby allowing the phase matching wavelength tolerance. It can be seen that it can be designed freely. It can also be seen that the tolerance increases as the amount of phase mismatch at both ends of the grating increases. When the phase mismatch amount at both ends of the grating is in the range of 0.5 to 1, phase-matching characteristics having a substantially peak flatness can be realized.
[0089]
The difference between the above structure and the phase shift structure is that the phase matching characteristic can be expanded over a wider range. In the case of a phase shift structure in which a phase shift is inserted into a domain-inverted structure having an equal period, the extended range of wavelength tolerance for a phase matching wavelength of 0.85 μm is a structure in which the phase matching wavelength region is divided into three parts of the same size In this case, the thickness is about 0.1 to 0.13 nm. On the other hand, in the structure of the present invention, the wavelength tolerance can be expanded to 0.2 nm or more.
[0090]
If the number of divisions is further increased, more peak flat characteristics can be realized. For example, in the case of a seven-divided structure, a phase error having a grating length (distance) dependency as shown in FIGS. 11 (a-2), (b-2), (c-2), and (d-2), respectively. When the phase matching characteristics in each case are obtained by performing a simulation on the matching amount distribution, the characteristics shown in FIGS. 11A-1, 11B-1, 11C-1, and 11D-1 are obtained. Obtained. From this, it can be seen that even if the number of divisions is increased, a phase matching characteristic close to a peak flat can be realized, and the allowable wavelength width can be designed freely.
[0091]
As can be seen from the above, the structure that realizes the wavelength tolerance expansion that realizes the peak-flat phase matching characteristic has a single period portion that satisfies the phase mismatch condition at the center, and the phase mismatch toward both ends. It is necessary for the structure to increase the matching amount and to increase the slope of the increase in the phase mismatch amount toward both ends. That is, the domain-inverted structure has a single-period structure at the center, and on both sides thereof, a chirp structure in which the domain-inverted period Λ decreases toward one end of domain-inverted and the other end of domain-inverted. It has been found that the wavelength tolerance of the phase matching characteristic can be expanded in a peak flat shape when it has a chirp structure in which the polarization inversion period Λ increases.
[0092]
In the present embodiment, the shape of the light wavelength conversion element is not described, but the present invention can be applied to any structure of a bulk type or an optical waveguide type. When the present invention is applied to an optical waveguide type structure, optical wavelength conversion between guided lights enhances light confinement and realizes a long interaction length, thus enabling highly efficient wavelength conversion. It is effective. In addition, when the present invention is applied to a bulk type structure, the conversion efficiency can be greatly improved by installing the crystal in the resonator structure.
[0093]
Further, as the periodic domain-inverted structure, LiNbO_ThreeLiTaO_ThreeIn KTP, it is possible to form a domain-inverted structure with a short period, and high efficiency conversion can be achieved as an optical wavelength conversion element.
[0094]
(Embodiment 2)
Here, a method for realizing the phase shift polarization inversion structure with another structure will be described.
[0095]
There is a phase shift polarization reversal structure as a method of expanding the wavelength tolerance by the polarization reversal structure. In this method, a phase shift unit (phase control unit) is inserted between polarization inversion structures having the same period to expand the phase matching wavelength tolerance. In this structure, a phase mismatch amount is generated in the inserted phase shift unit.
[0096]
On the other hand, in this embodiment, the occurrence of the same phase mismatch is realized by inserting a periodic structure having a different polarization inversion period. Specifically, as shown in FIG. 12A, the polarization inversion period is Λ instead of the conventional phase shift unit.₀Between the domain-inverted regions 613, the domain-inverted regions 614 and 615 having different periods, specifically, the period Λ₀Polarization inversion region 614 giving a phase shift of + δ at + α and period Λ₀A domain-inverted region 615 that gives a phase shift of −δ at −α is inserted as phase control units 614 and 615. The phase mismatch amount in this case is illustrated in FIG. Thus, the phase shift is formed with a distribution in the phase control units 614 and 615. However, the phase matching characteristics are substantially equal when the phase shift δ is given by the conventional phase shift unit. When the fundamental wave 606 is incident on such a configuration, a harmonic 607 is obtained.
[0097]
In the configuration of the present embodiment, the tolerance expansion ratio of the peak flat phase matching wavelength characteristic can be increased as compared with the conventional phase shift structure. In the case of the conventional phase shift structure, the expansion range of the wavelength tolerance with respect to the phase matching wavelength of 0.85 μm is 0.1 to 0.13 nm in the case of the structure in which the phase matching wavelength region is divided into three equal parts. Degree. On the other hand, in the structure of the present invention, the wavelength tolerance can be expanded to 0.18 nm or more. As described above, the design value of the enlargement ratio of the wavelength tolerance increases, so that an optical wavelength conversion element having more stable output characteristics can be manufactured.
[0098]
(Embodiment 3)
Here, a structure capable of always realizing stable output characteristics in a coherent light generator (also referred to as a coherent light source) including a semiconductor laser and an optical wavelength conversion element will be described.
[0099]
Specifically, in this embodiment, in the case where a coherent light source is configured by a semiconductor laser and an optical wavelength conversion element, in order to stabilize the harmonic output emitted from the optical wavelength conversion element, the optical wavelength conversion element that is indispensable The wavelength tolerance and tuning curve characteristics are clarified.
[0100]
FIG. 13 shows the structure of the coherent light source according to the present embodiment. This configuration includes an optical wavelength conversion element 621 and a semiconductor laser 622, and the optical wavelength conversion element 621 has an incident part 623 and an emission part 624. The semiconductor laser 622 has a function capable of varying the emission wavelength, and the output is stabilized by matching the emission wavelength to a wavelength that satisfies the phase matching condition of the optical wavelength conversion element 621.
[0101]
In such a configuration, the characteristics necessary for stabilizing the output of the harmonics emitted from the light wavelength conversion element 621 were examined.
[0102]
The oscillation wavelength of the semiconductor laser can be varied by grating feedback or optical feedback using a wavelength filter. It is also possible to vary the oscillation wavelength of the semiconductor laser by integrating the DBR grating in a part of the optical waveguide in the semiconductor laser and changing the reflection wavelength of the DBR grating by utilizing temperature and plasma effect. . However, since the semiconductor laser oscillates only at the longitudinal mode interval due to the resonator structure of the semiconductor laser, the oscillation wavelength is variable only at the jumping oscillation wavelength. For example, the oscillation wavelength can be controlled at intervals of about 0.1 nm.
[0103]
On the other hand, the optical wavelength conversion element usually has a characteristic with a very narrow wavelength tolerance. For example, LiNbO with an element length of 10 mm_ThreeThe waveguide type optical wavelength conversion element manufactured in (1) has a full width at half maximum of 0.1 nm or less. FIGS. 14A and 14B schematically show the relationship between the phase matching characteristics of the optical wavelength conversion element at this time and the oscillation wavelength of the semiconductor laser. 14A and 14B, reference numeral 625 denotes an oscillation mode (longitudinal mode) of the semiconductor laser, and reference numeral 626 denotes a tuning curve of the optical wavelength conversion element.
[0104]
In FIG. 14A, the peak of the tuning curve 626 and the oscillation mode 625 of the semiconductor laser overlap. In this case, the maximum harmonic output can be obtained by adjusting the longitudinal mode 625 of the semiconductor laser to the peak position of the tuning curve 626. However, as shown in FIG. 14B, when the peak of the tuning curve 626 is located near the center of the longitudinal mode 625 of the semiconductor laser, the harmonic output is maximized even if the oscillation wavelength of the semiconductor laser is adjusted. Can not do. Therefore, in such a coherent light source, even if the wavelength of the semiconductor laser is adjusted, a stable harmonic output cannot be obtained.
[0105]
In order to solve this, the flat part of the tuning curve of the optical wavelength conversion element is required to be wider than at least the interval of the longitudinal mode 625 of the semiconductor laser. When the flat portion of the tuning curve is wider than the interval of the longitudinal mode 625 of the semiconductor laser, that is, in the case of the tuning curve 636 shown in FIGS. 15A and 15B, the peak of the tuning curve 636 and the longitudinal mode of the semiconductor laser. Regardless of the positional relationship with 625, the maximum output of the harmonic can be obtained. Thus, the output can be stabilized by changing the wavelength of the semiconductor laser.
[0106]
Next, an attempt was made to modulate the output from the semiconductor laser and to modulate the harmonic output emitted from the coherent light source. As a result, the problem of the chirping of the oscillation wavelength of the semiconductor laser became clear. The chirping of the semiconductor laser is a phenomenon in which when the output of the semiconductor laser is modulated, the temperature of the active layer of the semiconductor laser changes in proportion to the output intensity, thereby changing the oscillation wavelength of the semiconductor laser.
[0107]
For example, as shown in FIG. 16A, when the output of the semiconductor laser is changed from a constant state A to a state B where the output is modulated at a specific frequency, the average light intensity differs between the state A and the state B. As a result, the temperature of the semiconductor laser active layer differs between states A and B, and the oscillation wavelength changes on the order of several tens of μs at the moment when the state A changes to state B. When the harmonic output from the coherent light source at this time is monitored, the output gradually changes as shown in FIG.
[0108]
On the other hand, when the modulation characteristics of the semiconductor laser were measured using an optical wavelength conversion element having a wavelength tolerance wider than the longitudinal mode interval of the semiconductor laser as shown in FIG. As shown in c), a stable modulation characteristic was obtained. Further analysis of this characteristic has found that the chirping wavelength of the oscillation wavelength of the semiconductor laser changes at a maximum by the longitudinal mode interval of the semiconductor laser. That is, even if the oscillation wavelength of the semiconductor laser is stabilized by optical feedback, the wavelength of the oscillation mode varies due to the change in the temperature of the active layer of the semiconductor laser. For this reason, the oscillation wavelength varies by the longitudinal mode interval at the maximum. In order to stabilize this, it has been found that it is important that the tuning curve of the optical wavelength conversion element has a flat portion near the peak and the flat portion is larger than the longitudinal mode interval of the semiconductor laser.
[0109]
It should be noted here that the wavelength tolerance and the conversion efficiency in the optical wavelength conversion element are in a trade-off relationship. That is, as the wavelength tolerance increases, the conversion efficiency of the optical wavelength conversion element decreases. Therefore, it is necessary to increase the wavelength tolerance of the optical wavelength conversion element to the minimum necessary.
[0110]
As a result, to stabilize the coherent light source by controlling the wavelength of the semiconductor laser, the tuning curve of the optical wavelength conversion element must be flat near the peak, and the flat portion must be larger than the longitudinal mode interval of the semiconductor laser. There is. The flatness of the tuning curve depends on the range of output fluctuation required for the coherent light source. In an ordinary laser light source, an output fluctuation of about 5% is allowed. In such a case, the flatness of the tuning curve in the wavelength range corresponding to the longitudinal mode interval of the semiconductor laser may be 5% or less. On the other hand, a flatter tuning curve is required for applications that require more stringent characteristics.
[0111]
Based on the above examination results, a coherent light source was configured by the optical wavelength conversion element and the semiconductor laser shown in the first embodiment, and an output stabilization experiment was performed. However, when the wavelength of the semiconductor laser is adjusted to match the phase matching wavelength of the optical wavelength conversion element and the harmonic output is stabilized, the output greatly fluctuates and the designed stabilization may not be obtained. It became clear that there was. Therefore, in order to further examine the cause of the harmonic output fluctuation, as shown in FIG. 17, the wavelength of the fundamental wave was continuously changed, and the change in the harmonic output was observed.
[0112]
As a result, it became clear that the harmonic output fluctuated finely within the tolerance range of the optical wavelength conversion element. When this cause was examined, it was found that the fundamental wave and the harmonic wave were Fresnel-reflected at the incident part end face and the emission part end face of the optical wavelength conversion element, and this light interfered with each other, thereby destabilizing the harmonic output. It was revealed. That is, the end surface reflected light at the incident part and the output part interfere with each other, and the optical path length of the optical wavelength conversion element changes due to the change in wavelength, thereby changing the degree of interference of the end surface reflected light and making the output unstable. ing. Furthermore, similar output fluctuations were observed when the fundamental wave output was modulated to modulate the harmonic output.
[0113]
As described above, when the semiconductor laser is modulated, the wavelength of the semiconductor laser changes, so that it is difficult to stabilize the output even if an optical wavelength conversion element with an increased wavelength tolerance is used, and the noise level of the harmonics is greatly increased. . In order to solve this problem, as a method for preventing interference due to end face reflection, an antireflection film for harmonics and fundamental waves is formed on at least one of the incident part and the emission part. By depositing an antireflection film on the end face, Fresnel reflection can be prevented, and the interference effect due to end face reflection can be reduced. As a result, a coherent light source having very stable output characteristics can be realized.
[0114]
Furthermore, it is desirable to form an antireflection film for the fundamental wave in both the incident part and the emission part. In the semiconductor laser, the light emitted to the outside returns to the active layer again, thereby causing problems such as an increase in noise and fluctuation in output. In order to prevent this, it is desirable to form an antireflection film for the fundamental wave in both the emitting part and the incident part. On the other hand, the antireflection film for harmonics may be either the incident part or the emission part, but is preferably formed on the emission part. This is because by forming a harmonic antireflection film at the emission part, loss due to Fresnel reflection on the end face can be prevented, and higher output harmonics can be obtained.
[0115]
Therefore, as a desirable configuration, an antireflection film having an antireflection effect for both the fundamental wave and the harmonic wave is formed on the emission part, and an antireflection film for the fundamental wave is formed on the incident part.
[0116]
As another method for solving the problem that the harmonic output becomes unstable due to the interference effect of the end face reflection at the entrance and exit, there is a method of superposing a high frequency on the drive current of the semiconductor laser.
[0117]
The cause of interference between the end face reflection light is that the light may interfere with each other due to the high coherence of the light. Therefore, in order to overcome this point, it is conceivable to reduce the coherence and reduce the degree of interference. This is a method of reducing the coherence by modulating the drive current of the semiconductor laser at a high frequency, and the modulation is applied at a high frequency of several hundred MHz or more. At this time, the oscillation wavelength of the semiconductor laser spreads from a single mode to a multimode. Such a multimode can be prevented by feeding back strong optical feedback into the active layer using a DBR grating or the like.
[0118]
In this way, by superimposing a high frequency on the semiconductor laser, it is possible to reduce end face interference and to ensure the stability of the harmonic output.
[0119]
Furthermore, the high frequency superposition has a feature of greatly increasing the efficiency of the optical wavelength conversion element. The efficiency of the optical wavelength conversion element increases in proportion to the power of the fundamental wave. When a semiconductor laser is driven with high frequency superposition, pulse train oscillation with a high spire value occurs, and the value of each peak increases from several times to 10 times or more. Therefore, wavelength conversion of high-intensity pulsed light is performed, and the conversion efficiency is greatly increased. In the experiment, a conversion efficiency improvement of 2-3 times or more was observed. Also from this point, it is effective to use a semiconductor laser with high frequency superimposed.
[0120]
Furthermore, even when the width of the oscillation wavelength of the semiconductor laser is slightly widened due to the high frequency superposition, the optical wavelength conversion element with an expanded wavelength tolerance can realize a wavelength tolerance wider than the oscillation wavelength of the semiconductor laser. All the light can be efficiently wavelength-converted, and the light wavelength can be converted with high efficiency.
[0121]
(Embodiment 4)
Here, a coherent light generator using an optical wavelength conversion element will be described.
[0122]
With the configuration of the optical wavelength conversion element according to the above-described embodiment, a highly efficient and stable optical wavelength conversion element can be realized. Therefore, an attempt was made to produce a short wavelength light source as a coherent light generator using the present optical wavelength conversion element. This short wavelength light source is composed of a semiconductor laser having a wavelength of 850 nm band, a condensing optical system, and an optical wavelength conversion element, and the light emitted from the semiconductor laser is converted into a waveguide short-circuit of the optical wavelength conversion element by the condensing optical system. Focus on the surface and excite the guided mode. Wavelength-converted SHG light is emitted from the other end face of the optical wavelength conversion element.
[0123]
Since an optical wavelength conversion element with high conversion efficiency is realized by the present invention, the short wavelength light source (coherent light generator) of the present embodiment having the above-described configuration uses a semiconductor laser with an output of about 100 mW and a blue wavelength of 10 mW. SHG light was obtained. In addition, the wavelength conversion element used has an increased wavelength tolerance and has a flat tuning characteristic, so that a stable output characteristic can be obtained against wavelength fluctuations. As a result, the output fluctuation can be suppressed to 5% or less.
[0124]
The wavelength in the 400 nm band is desired in a wide range of application fields such as special measurement fields such as printing plate making, bioengineering, and fluorescence spectral characteristics, and optical disk fields. The short wavelength light source using the light wavelength conversion element of the present invention can be put into practical use in these application fields from both aspects of output characteristics and operational stability.
[0125]
In the present embodiment, the light of the semiconductor laser is coupled to the optical waveguide using a condensing optical system, but it is also possible to directly couple the semiconductor laser and the optical waveguide. Specifically, when an optical waveguide for TE mode propagation is used, the electric field distribution in the optical waveguide can be made equal to the waveguide mode of the semiconductor laser, so that coupling can be performed with high efficiency without a condensing lens. In the experiment, it was confirmed that direct coupling was possible with a coupling efficiency of 80%, and that coupling characteristics almost equivalent to lens coupling were obtained. If direct coupling is used, a small and low-cost light source can be realized, which is promising.
[0126]
Furthermore, even if parametric oscillation is used, the variable wavelength region of the wavelength tunable laser can be expanded.
[0127]
Parametric oscillation is possible by using an optical wavelength conversion element having a periodic domain-inverted structure and a laser light source. According to parametric oscillation, when a fundamental wave having a wavelength λ3 is input, it is possible to generate a signal light having a wavelength λ2 and an idler light having a wavelength λ1 that satisfy the relationship 1 / λ3 = 1 / λ1 + 1 / λ2. As a result, it is possible to output light having a wavelength that satisfies the above-described conditions using the fundamental wave having the wavelength λ3 while changing the wavelength, and a wavelength-variable laser light source can be realized.
[0128]
In such a configuration that enables parametric oscillation, if the optical wavelength conversion element of the present invention is used, an optical wavelength conversion element having a wide wavelength tolerance can be realized, and a stable output can be obtained.
[0129]
Furthermore, the expansion of the wavelength variable range, which has been a problem with the conventional parametric oscillation, can be realized.
[0130]
When parametric oscillation is performed using a domain-inverted structure with a period Λ, light having a wavelength λ1 and light having a wavelength λ2 satisfying the relationship Λ = 2mπ / (β3-β1-β2) can be generated. However, in the prior art, since the wavelength tolerance that satisfies the Λ condition is narrow, there is a problem that the wavelength condition that satisfies the generation condition in the same periodic structure is narrow, and the wavelength variable range is extremely narrow. On the other hand, when the optical wavelength conversion element of the present invention is used, the phase matching wavelength tolerance can be increased with a peak flat tuning curve. This increases the tolerance for wavelength fluctuations of the fundamental wave, but in the case of parametric oscillation, the wavelength tolerance for the signal light and idler light can also be increased. Therefore, the variable wavelength range of the output wavelength, which was difficult with the conventional optical wavelength conversion element, can be greatly expanded.
[0131]
Furthermore, since it has a peak-flat tuning curve, the oscillation wavelength can be varied while keeping the output intensity substantially constant.
[0132]
(Embodiment 5)
Here, an optical information processing apparatus configured according to the present invention will be described.
[0133]
FIG. 18 shows the configuration of the optical information processing apparatus of the present invention. In FIG. 18, a beam of 10 mW output from the coherent light generator 640 having the characteristics described in Embodiment 6 is transmitted through the beam splitter 641 and irradiated onto the optical disk 643 that is an information reproducing medium by the lens 642. . On the contrary, the reflected light from the optical disk 643 is collimated by the lens 642, reflected by the beam splitter 641, and a signal is read by the photodetector 644. Furthermore, information can be written on the optical disk 643 by intensity-modulating the output of the coherent light generator 640.
[0134]
According to the present invention, since the tolerance of the optical wavelength conversion element constituting the coherent light generation device 640 is expanded, the output can be stabilized, and the output fluctuation is 5% or less even with respect to an external temperature change. Can be suppressed.
[0135]
Furthermore, since high-output blue light can be generated, not only reading but also information can be written to the optical disc 643 as described above. In addition, since a semiconductor laser is used as a fundamental wave light source, it becomes very small, so that it can be used for a small consumer optical disk reading / recording apparatus.
[0136]
Although writing to the optical disk 643 requires modulation of output, the optical information processing apparatus of the present invention modulates output from the coherent light generator 640 by modulating the output intensity of the semiconductor laser. When the wavelength of the semiconductor laser is modulated, the oscillation wavelength fluctuates. However, as described above, since the optical wavelength conversion element has a flat peak phase matching characteristic, instability of the harmonic output due to the modulation of the semiconductor laser does not occur. As a result, stable modulation output characteristics can be obtained, and low noise characteristics can be realized.
[0137]
Furthermore, the aspect ratio of the output beam can be optimized by optimizing the optical waveguide width of the optical wavelength conversion element. For example, by providing a waveguide structure having a high refractive index layer narrower than the width of the optical waveguide on the optical waveguide, the aspect ratio of the outgoing beam can be made close to 1: 1. As a result, it is possible to improve the light collecting characteristic of the optical pickup without using a beam shaping prism or the like, thereby realizing high transmission efficiency, excellent light collecting characteristic, and low price. Furthermore, the noise of scattered light generated during beam shaping can be reduced, and the pickup can be simplified.
[0138]
【The invention's effect】
As described above, according to the present invention, a peak-flat phase matching characteristic can be realized by using a polarization inversion structure including a single period structure and a chirp period structure in the optical wavelength conversion element. Further, a new method for designing the phase matching characteristics of the optical wavelength conversion element using the distribution of the phase mismatch amount is provided. As a result, the degree of freedom in designing the phase matching wavelength tolerance in the optical wavelength conversion element has been greatly increased, and the wavelength tolerance has been increased by 1.5 times or more than the conventional one. With this configuration, an optical wavelength conversion element having stable output characteristics can be realized.
[0139]
Further, output stabilization is realized in a coherent light generator (coherent light source) configured by an optical wavelength conversion element and a semiconductor laser. Specifically, the optical wavelength conversion element included in the coherent light generator has a wavelength tolerance wider than the longitudinal mode interval of the semiconductor laser, and has a characteristic that the tuning curve is flat within the wavelength tolerance. It becomes possible to always stabilize the wavelength of the semiconductor laser within the wavelength tolerance. As a result, it is possible to realize a coherent light source having stable output characteristics by suppressing output fluctuation of coherent light, and its practical effect is great.
[Brief description of the drawings]
1A is a configuration diagram of a polarization inversion structure in an optical wavelength conversion element of the present invention, and FIG. 1B is a diagram showing a distribution of polarization inversion periods in the structure of FIG.
2A is a diagram showing a distribution of polarization inversion periods in a linear chirp polarization inversion structure, and FIG. 2B is a diagram showing a distribution of phase mismatch amounts in a linear chirp polarization inversion structure; (c) is a figure showing the distribution of the polarization inversion period in a division | segmentation periodic polarization inversion structure, (d) is a figure showing the distribution of the phase mismatching amount in a division | segmentation period polarization inversion structure.
3A is a diagram showing a distribution of polarization inversion periods of a domain-inverted structure having a phase shift, and FIG. 3B is a diagram showing a distribution of phase mismatch amounts of the domain-inverted structure having a phase shift. FIG.
FIG. 4A is a diagram illustrating a distribution of polarization inversion periods of a periodically poled structure divided into two, and FIG. 4B is a diagram illustrating a distribution of phase mismatch amounts of the periodically poled structure divided into two. (C) is a figure showing the phase matching characteristic (tuning curve) obtained.
FIG. 5A is a diagram showing a distribution of polarization inversion periods of a split period polarization inversion structure with a linear chirp period, and FIG. 5B is a distribution of phase mismatch amounts of a polarization inversion structure with a linear chirp period. (C) is a figure showing the phase matching characteristic (tuning curve) obtained.
6 is a diagram showing the phase matching characteristics when the function f (z) followed by the phase mismatch distribution is a power function in the domain-inverted structure of the present invention. (1) When the order m = 1, 2) When the order m = 2, (3) When the order m = 3, (4) When the order m = 4.
7 is a diagram showing a phase mismatch amount distribution when a function f (z) followed by the phase mismatch amount distribution is a power function in the domain-inverted structure according to the present invention. (1) When the order m = 1 (2) When the order m = 2, (3) When the order m = 3, (4) When the order m = 4.
FIG. 8A is a diagram showing a phase matching characteristic when a function f (z) followed by a phase mismatch amount distribution is a trigonometric function (order m = 2) in the domain-inverted structure of the present invention; (B) is a figure showing the phase mismatching amount distribution in that case.
FIG. 8B is a diagram showing a phase matching characteristic when a function f (z) followed by a phase mismatch amount distribution is a trigonometric function (order m = 3) in the domain-inverted structure of the present invention; (B) is a figure showing the phase mismatching amount distribution in that case.
FIG. 8C is a diagram showing phase matching characteristics when the function f (z) followed by the phase mismatch distribution is a trigonometric function (order m = 4) in the domain-inverted structure of the present invention; (B) is a figure showing the phase mismatching amount distribution in that case.
FIG. 8D is a diagram showing phase matching characteristics when the function f (z) that the phase mismatch distribution follows is a trigonometric function (order m = 5) in the domain-inverted structure of the present invention; (B) is a figure showing the phase mismatching amount distribution in that case.
9A is a diagram showing phase matching characteristics in the optical wavelength conversion element of the present invention, and FIG. 9B is a diagram showing a phase mismatch amount distribution in that case.
10 (a-1), (b-1), (c-1), and (d-1) have phase mismatch distributions (a-2), (b-2), and (c−), respectively. It is a figure which shows the phase matching characteristic obtained in the case of (2) and (d-2) (5-part dividing structure).
11 (a-1), (b-1), (c-1), and (d-1) have phase mismatch distributions (a-2), (b-2), and (c−), respectively. It is a figure which shows the phase matching characteristic obtained in the case of 2) and (d-2) (7 division structure).
12A is a diagram showing phase matching characteristics obtained in a polarization inversion structure having a phase control unit of the present invention, and FIG. 12B is a diagram showing a phase mismatch distribution in that case. .
FIG. 13 is a structural diagram of a coherent light generator according to the present invention.
14A is a diagram showing the relationship between the oscillation wavelength of the semiconductor laser and the phase matching characteristics when the maximum value of the tuning curve of the optical wavelength conversion element matches the oscillation wavelength of the semiconductor laser. FIG. (B) is a figure showing the relationship between the oscillation wavelength of a semiconductor laser and a phase matching characteristic when the oscillation wavelength deviates from the vicinity of the maximum value of the tuning curve.
FIG. 15A is a diagram illustrating an oscillation wavelength and phase of a semiconductor laser when the maximum value of the tuning curve of the optical wavelength conversion element matches the oscillation wavelength of the semiconductor laser in the coherent light generation device of the present invention. It is a figure showing the relationship of a matching characteristic, (b) is the relationship between the oscillation wavelength of a semiconductor laser, and a phase matching characteristic when the oscillation wavelength shift | deviates from the maximum value of a tuning curve in the coherent light generator of this invention. FIG.
FIG. 16A is a diagram showing a modulation output of a fundamental wave in a coherent light generator, and FIG. 16B is a diagram showing an output fluctuation of a coherent light generator using a conventional optical wavelength conversion element. (C) is a figure which shows the output fluctuation | variation of the coherent light generator using the optical wavelength conversion element of this invention.
FIG. 17 is a diagram illustrating output characteristics when the fundamental wavelength of the coherent light generator is changed.
FIG. 18 is a configuration diagram of an optical information processing apparatus according to the present invention.
FIG. 19 is a configuration diagram of a conventional optical wavelength conversion element.
FIG. 20 is a configuration diagram of a conventional optical wavelength conversion element.
FIG. 21A is a diagram showing phase matching characteristics in a conventional two-part structure optical wavelength conversion element, and FIG. 21B is a diagram showing phase matching characteristics in a conventional three-part construction optical wavelength conversion element. is there.
[Explanation of symbols]
601 Domain inversion
602 Chirp period
603 Single period part
606 fundamental wave
607 harmonic
613 Polarization inversion region
614 Phase control unit
615 Phase control unit
621 Optical wavelength conversion element
622 Semiconductor laser
623 Incident part
624 emitting part
625 Semiconductor laser oscillation mode (longitudinal mode)
626 Tuning curve of optical wavelength converter
636 Tuning curve of optical wavelength conversion device according to the present invention
640 coherent light generator
641 Beam splitter
642 lens
643 optical disc
644 photodetector
6101 LiNbO_ThreeSubstrate (nonlinear optical crystal)
6102 Optical waveguide
6103 Polarization inversion layer
6105 Polarization inversion region
6106 Phase controller

Claims (1)

A nonlinear optical crystal;
An optical wavelength conversion device comprising a periodic domain-inverted structure formed in the nonlinear optical crystal,
該分pole inversion structure, a single period portion having a single period lambda _0, the polarization poling period towards the one end of the inverted structure lambda is lambda _-1, and Λ _-2 ··· Λ _-m A first chirp period portion that gradually decreases, and a second chirp period portion in which the polarization inversion period Λ gradually increases as Λ _1, Λ ₂ ... Λ _m toward the other end of the polarization inversion structure, Have
The first chirp period portion and the second chirp period portion are arranged at both ends of the single period portion,
The length of the periodic domain-inverted structure is L,
The distance from the center of the periodic domain-inverted structure is z,
Add the difference between Λ _i (i ≦ m), which is between the center of the domain-inverted structure and the distance z, and the period Λ ₀ of the single-period part. F (z) = (Λ ₋₁ + Λ ₋₂ ... + Λ _−i ) −iΛ ₀ z <0
The difference between Λ _i (i ≦ m) that is the period of the second chirp period portion and the period Λ ₀ of the single period portion that exists between the center of the domain-inverted structure and the distance z is Addition formula f (z) = (Λ ₁ + Λ ₂ ... + Λ _i ) −iΛ ₀ z> 0
, The standard mismatch values | f (L / 2) / Λ ₀ | and | f (−L / 2) / Λ ₀ | of the phase inversion amounts at both ends of the domain-inverted structure are in the range of 0.5 to 1. An optical wavelength conversion element.

Images