JPS6336643B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a duplexer using a distributed index spherical lens having a structure that corrects spherical aberration and at the same time enables close integration with other optical elements. It is.

In the fields of optical communication, optical information processing, and medicine, high-performance small lenses for collimating and condensing laser light, so-called micro-optics lenses, are indispensable. In order to meet this demand,
First, a spherical shell-shaped cladding and a rod-shaped cladding are covered with a spherical shell-shaped cladding and a rod-shaped cladding, each having a thickness corresponding to the distribution, on a spherically symmetrically distributed refractive index sphere core.
We proposed a non-uniform refractive index lens that corrects spherical aberration and at the same time enables close integration with other optical elements via a rod portion (see Japanese Patent Application No. 1983-66612 (hereinafter referred to as the prior application)). This utilizes not only the refractive index gradient of the spherical core but also the lens action of the surface of the spherical shell-like cladding, so a large numerical aperture NA of about 0.45 can be obtained, and it can be applied to pick-up lenses for optical disks. It was also suitable for but,
In order to increase the numerical aperture NA, the structure became somewhat complex, and the most difficult manufacturing task was to fabricate a spherical shell-like cladding with a uniform thickness.

By the way, if we consider a coupling lens for optical communication, that is, a lens that couples an optical fiber with other optical elements, the numerical aperture NA of the optical fiber is approximately 0.2 at most, and it is not necessarily the case that the numerical aperture NA of the optical fiber is about 0.2. There is no need to use a lens with a high numerical aperture; the spherical shell-like cladding can be removed and a simple lens with a rod-like cladding can be used instead. Furthermore, by using only the rod-shaped cladding, the degree of freedom in integrating with other optical elements is further increased.

In consideration of the above points, the numerical aperture NA is set to about 0.3, which is somewhat larger than the numerical aperture NA of the optical fiber, and the spherical transverse aberration is several μm, which is within the core diameter of a single mode optical fiber (hereinafter referred to as SMF). It enables more complete integration through close contact with other optical elements than the prior-application lens, and is therefore structurally sound, reduces loss due to boundary reflection of the medium, and is easy to manufacture. We proposed a lens with high performance and high reliability when used in optical communication, optical information processing, and laser medicine (1981 Fall 42nd
(Refer to the proceedings of the Japan Society of Applied Physics, Japan).

The configuration of the element lens is shown in FIG. 1a. In this figure, reference numeral 1 denotes a spherical core with a spherical-to-spherical refractive index distribution, and a spherical lens 3 embedded in a surrounding medium 2 with a uniform refractive index.The relationship between the refractive indexes is shown in FIG. 1b.
Note that a BSL (Buried Sphere Lens) in which a uniform refractive index sphere is embedded in another surrounding medium has been developed. Although this lens is very good in terms of integration with other optical elements, it has the drawback of extremely large spherical aberration.

Therefore, in Fig. 1a, a spherical core 1 whose refractive index decreases from the center to the periphery approximately as the square of the distance.
However, in the surrounding medium 2 having an arbitrary refractive index, spherical aberration still remains, and if the refractive index n(r) is within a certain range and the refractive index n _d selected according to the distribution is μ for the first time in the surrounding medium 2
It becomes possible to achieve a spherical transverse aberration on the order of m. Note that r ₀ is the radius of the spherical core 1, and f is the distance from the center of the spherical core 1 to the focal point, where the focal point is located at the end of the surrounding medium 2. Further, H represents the incident height, r represents the distance in the radial direction, and θ represents the exit angle.

Figure 2 shows the incident height H/r ₀ and (lateral aberration r ₀ ) x
10 ³ is a diagram showing a set of the fourth-order distribution coefficient G ₄ of the spherical core 1 and the refractive index n _d of the surrounding medium 2 as parameters. In this figure, the fourth distribution coefficient
The aberration is minimized at G ₄ =−0.16 and refractive index n _d =1.45. In addition, the curve is G ₄ = −0.32, n _d = 1.46,
The curves are for G ₄ =0, _nd = 1.44, and in both cases NA = 0.3, n(o) = 1.6, and G ₂ = -0.1. Here, as the refractive index distribution n(r) of the ball core 1,
n ² (r) is expressed by the following polynomial.

n ² (r)=n ² (o) [1+G ₂ (r/r ₀ ) ² +G ₄ (r/r
₀ ) ⁴ ]...(1) The ranges of the distribution coefficients G ₂ and G ₄ of the spherical core 1 that can reduce aberrations in this way are shown in FIG. 3. In this diagram,
The solid line is the refractive index n _d and the dashed line is (lateral aberration/r ₀ ) ×
10 ³ , the dotted line is f/r ₀ , NA=0.3, n(o)=
It was set to 1.6.

For the given distribution coefficients G ₂ and G ₄ , if we choose the refractive index n _d of the surrounding medium 2 as the parameter (solid line) that passes through the point whose coordinate is This means that the lateral aberration shown by the dashed line) can be reduced. Here, the lateral aberration is the exit angle θ
The lateral aberration between =0 and sin ^-1 NA is shown as a peak when positive and negative are balanced as shown in Fig. 2. Also, since it is written as the value of (lateral aberration/r ₀ ) × 10 ³ ,
The values shown in the figure for a lens with r ₀ =1 mm give the transverse aberration in μm.

From Figure 3, in order to realize a lens with lateral aberration <2×10 ^-3 r ₀ using a spherical core 1 with a refractive index difference of 5% (G ₂ -0.1),
The fourth-order distribution coefficient _G4 may be controlled to -0.025< _G4 <-0.005.

Now, in order to understand the difficulty of this control, we will explain the expected distribution coefficient (vertical axis) time (normalized diffusion time Dt/
Figure 4 shows the changes to ro ² ).
The initial conditions are that the internal refractive index n of the ball core 1 is 1.58,
The refractive index n of the surface of the spherical core 1 through which ions are diffused at the tip of the eye is set to a lower value of 1.5.

It is assumed that ions that lower the refractive index diffuse into the interior of the spherical core 1, and that the change in the refractive index is proportional to the ion concentration. From Figure 4, the fourth distribution coefficient
Since G ₄ changes from negative to positive over time,
Since the region of FIG. 3 where the aberration can be reduced is crossed, ion exchange can be stopped when the region is entered. The time corresponding to the range of the fourth-order distribution coefficient _G4 mentioned above corresponds to ±5% of the total time, and this control is not difficult.

Up to this point, we have considered up to the 4th distribution coefficient G ₄ of the refractive index distribution and set the 6th distribution coefficient G ₆ to 0.
It can be seen from the figure that the sixth-order distribution coefficient G ₆ remains a positive value considerably in the time range that gives the desired fourth-order distribution coefficient G ₄ . Therefore, the sixth distribution coefficient
The analysis takes into consideration up to G ₆ , and the G ₂ - _{G 4} plane in Figure 3 is
Figure 5 shows the _G4 - _G6 cross section of the three-dimensional drawing with the _G6 axis added.

In this case, it is expanded as n ² (r)=n ² o [1+G ₂ (r/r ₀ ) ² +G ₄ (r/r ₀ ) ⁴ +G ₆ (r/r ₀ ) ⁶ ].

It should be noted that the area where aberrations can be reduced (encircled by a dashed-dotted line) expands in the positive direction of the sixth-order distribution coefficient _G6 , and has the same sign as the sixth-order distribution coefficient _G6 in Figure 4. It is. In other words, it becomes very easy to control the ion exchange process. In this case, a cross section taken at G ₂ =-0.1 is shown, and the other details are the same as in FIG. 4.

Next, tolerances in manufacturing the embedded type distributed index spherical lens used in the present invention will be explained. The characteristics of the spherical lens design are that the spherical core 1
Since the refractive index n _d of the surrounding medium 2 is selected according to the distribution of n d , the tolerance of the refractive index n _d is first investigated.

FIG. 6 shows the relationship between the refractive index n _d of the surrounding medium 2 and the transverse aberration for an example of a given distribution. Note that NA=0.3, n(o)=1.6, _G2 =-0.1, _G4 =-
0.02, and G ₆ =0.05. In this case, the aberration is minimal at n _d = 1.414, but it is difficult to find a surrounding medium with the same value as this value, so it is necessary to know the tolerance depending on the intended use of the lens.

Now, if we aim to focus light on the SFM core diameter (8 μmφ), we only need to suppress the lateral aberration to 4 μm, which is half of that, and Figure 6 shows that for a lens with radius r ₀ = 1 mm of spherical core 1, The tolerance of the refractive index n _d is ±0.005. However, this value was obtained assuming that lenses with NA = 0.3 are used as many times as possible.
In actual coupling to an optical fiber (NA 0.1 to 0.2), only a portion of the lens is used, so the tolerance increases.

Another consideration is the tolerance regarding the measurement of the refractive index distribution n(r) of the ball core 1.
In the above, the lens was designed assuming that the accurate distribution was known, but in reality, there are measurement errors, so aberrations must be overestimated.

Figure 7 a and b are 2nd, 4th and 6th degree distribution coefficients G ₂ , G ₄ ,
It shows the aberrations that must be estimated as G ₆ errors. In other words, it shows the measurement accuracy of distribution coefficients required to suppress aberrations within a certain range.
However, to get a rough guide here,
Only the isoaberration lines of (lateral aberration/r ₀ )×10 ³ =4 (corresponding to lateral aberration=4 μm when r ₀ =1 mm) are shown. Here, the true value of the distribution coefficient is n(o)=1.6,
Since the n _d of the surrounding medium is determined as G ₂ = -0.1, G ₄ = -0.02, and G ₆ = 0.05, the aberration is least at the center of the shaded area in Figure 7, and the distribution coefficient deviates from there. It gets bigger and bigger. Since it is an elongated (hatched) region, it has directionality and is difficult to express numerically, but if we force it to represent the δG ₂ , δG ₄ , and δG ₆ directions in the figure, we can say that for a lens with r ₀ = 1 mm, the SMF (core diameter
To fit the aberration within the diameter of 8μmφ), 2, 4,
The deviations from the center of the sixth distribution coefficients G ₂ , G ₄ , and G ₆ are:
Required to be ±0.005, ±0.01, ±0.1 respectively,
This is the measurement accuracy that is required.

Now, the relationship between the refractive index n _d of the spherical core 1 and the surrounding medium 2 for reducing spherical aberration of the element lens has been described above.Next, an example of the configuration of an optical circuit using this lens will be shown.

Figures 8a, b, and c show a ball lens 3 for inserting another optical element between the optical fibers 4. The optical fiber 4 and the ball lens 3 are tightly integrated, and the two balls The distance between the lenses 3 can be freely selected depending on the length of the optical element to be inserted. Depending on the shape of the optical element, it can also be integrated. Since the numerical aperture NA of the ball lens 3 is about 0.3, and the numerical aperture NA of the SMF optical fiber 4 is about 0.1, there is a margin in the ball lens 3, and a large number of optical fibers 4 can be used as shown in Fig. 8a and b. can share one set of ball lenses 3. In Figure 8a, the multiplicity is about 7,
An advantage is that the light beams at intervals can be made parallel to each other.

In FIG. 8b, the multiplicity is sufficient to line up the optical fibers 4 within a range in which the light beams from the optical fibers 4 do not deviate from the two spherical cores 1. For example, the radius r ₀ = 1 mm of the spherical core 1, the distance l10r ₀ between the spherical cores 1, the numerical aperture NA of the spherical lens 3, and the optical fiber 4, respectively.
0.3 and 0.1, and assuming that the outer diameter of the optical fiber 4 is 50 μm, it becomes approximately 300. However, since the beams cross each other at the interval, their use is limited to applications that are not affected by this. In FIG. 8b, the contact surface between the optical fiber 4 and the spherical lens 3 is made into a spherical surface, so that the light beams from the optical fiber 4 that are away from the central axis also become parallel beams at intervals. Unlike the lens of the prior application, this spherical lens action is not utilized.

In Fig. 8c, the spaced portions are made to face each other at the Brillustat angle in an attempt to eliminate reflection loss even if an air layer enters. FIG. 8d is an optical switch that utilizes total reflection of the prism 5 and is configured so that the prism 5 can be moved in and out in the direction of the arrow. Also,
FIG. 8e shows a demultiplexer that integrates the diffraction grating 6 and separates light of each wavelength from the optical fiber 4 including multiple wavelengths.
Or, conversely, it can also be seen as a multiplexer that combines light of multiple wavelengths into a single optical fiber 4.

Next, an embodiment of the duplexer of the present invention using the above-mentioned distributed index lens will be described.

FIG. 9a shows interference filters 7a and 7 in the ball lens 3.
This is an embodiment in which a reflector 8 and a reflector 8 are combined. The light from the optical fiber 4 containing multiple wavelengths λ ₁ , λ ₂ , λ ₃ , λ ₄ is made into a parallel beam by the spherical core 1 and reflected by the reflector 8, and then only the light of wavelength λ ₁ is transmitted and other wavelengths are transmitted. The light of λ 1 is made incident on the interference filter 7a that reflects the light, and the transmitted light of wavelength λ ₁ is focused using a part of the spherical core 1a and is injected into the optical fiber 4a. The lights of wavelengths λ ₂ , λ ₃ , λ ₄ reflected by the interference filter 7a are reflected by the reflecting mirror 8, and then only the wavelengths λ ₂ are transmitted, and the lights of wavelengths λ ₃ , λ ₄ are transmitted.
The transmitted light of wavelength λ ₂ is focused through a different part of the spherical core 1a and is injected into the optical fiber 4b. Thereafter, the wavelengths λ ₃ and λ ₄ are also separated in the same manner.

The above-mentioned demultiplexer is located in a multiplexer when viewed in the opposite direction of the light beam. The embodiment shown in FIG. 9a has a cross section perpendicular to the plane of the paper that is four-sided columnar as shown in FIG. 9b, or the example shown in FIG. 9c.
A cylindrical type shown in Figure 9c is considered, and in Figure 9c, the same array of spherical cores (multiple arrays perpendicular to the plane of the paper) are commonly used in a radial manner to construct a multi-channel demultiplexer (combiner). Can be done. Note that the reference numeral 7 indicates interference filters generically.

FIG. 9d shows a configuration in which a single spherical core 1 constitutes a demultiplexer (multiplexer) equivalent to the embodiment shown in FIG. 9a. The cross section perpendicular to the plane of the paper can be configured as shown in FIG. 9e or f, respectively, in a cylindrical or spherical shape, and the latter can have a multi-channel configuration similar to FIG. 9c. In all of the above, the symmetry of the spherical lens (spherical core 1) and the numerical aperture NA are SMF (optical fiber 4).
They take advantage of the fact that they have more leeway than they do.

As explained in detail above, this invention uses a spherical core whose refractive index decreases approximately to the square of the distance from the center to the periphery, and a plurality of interference filters with different transmission ranges with a homogeneous refractive index. embedded in the surrounding medium of
This demultiplexer is equipped with an optical fiber that makes the laser beam enter the surface of this peripheral medium, a reflecting mirror that reflects the incident laser beam, and a plurality of optical fibers that output the demultiplexed laser beam. The element lenses have low aberrations, are close-contact type with no intervening air layer, and make good use of the spherical symmetry of the element lenses, so they are small and lightweight with low loss even when the number of demultiplexes is large. things can be provided. Therefore, it has the advantage of improving the performance and reliability of the entire device.

[Brief explanation of the drawing]

Figure 1a shows the basic configuration of the distributed index lens used in this invention, Figure 1b shows the refractive index of each part, and Figure 2 shows the relationship between the incident height and lateral aberration in Figure 1. Figure 3 is a diagram for determining the refractive index of the surrounding medium to minimize spherical aberration when the 2nd and 3rd order distribution coefficients of the spherical core are given, and also determining the residual transverse aberration and focal length. , Fig. 4 is a diagram of the distribution coefficient of the sphere center expected by ion exchange, calculated against the normalized diffusion time, and Fig. 5 shows the G ₄ −G in the range where aberrations can be reduced by considering up to the 6th distribution coefficient. Figure ₆ shows the tolerance for the selection of the refractive index of the surrounding medium; Figures 7a and b show the tolerance for the measurement of the distribution coefficient of the spherical core; Figures 8a-e Figure 9 shows examples of optical circuits.
Figures a to f are diagrams showing embodiments of the invention, respectively. In the figure, 1 is the sphere core with distributed refractive index, 2 is the surrounding medium,
3 is a ball lens, 4, 4a to 4d are optical fibers, 5
is a prism, 6 is a diffraction grating, 7, 7a to 7d are interference filters, 8 is a reflecting mirror, 11 is an element, 12 is a ball joint, and 13 is a liquid medium.

Claims (1)

[Claims] 1. A spherical core whose refractive index decreases approximately by the square of the distance from the center to the periphery, and a plurality of interference filters with different transmission ranges in a surrounding medium with a homogeneous refractive index. An optical fiber is embedded in the surrounding medium and allows the laser beam to enter the surface of the surrounding medium, a plurality of reflecting mirrors reflect the incident laser beam, and a plurality of optical fibers emit the demultiplexed laser beam. Features a distributed refractive index sphere-centered wavelength demultiplexer. However, the refractive index distribution n(r) of the spherical core is n ² (r)=n ² (o) [1+G ₂ (r/r ₀ ) ² +G ₄ (r/r ₀
) ⁴ ], the refractive index of the surrounding medium is selected to minimize the spherical aberration according to the coefficients n(o), G ₂ , and G ₄ . 2 The surrounding medium is a polygonal prism having at least one pair of parallel planes, a plurality of spherical cores are arranged in the axial direction, the opposing surfaces are set as a set, a reflecting mirror is provided on one side, an optical fiber is provided on the other side, and an interference filter is used. 2. The distributed refractive index sphere-centered wavelength demultiplexer according to claim 1, wherein said wavelength demultiplexer is provided near said plurality of sphere-core rows in parallel to a reflecting mirror. 3. The surrounding medium is a polygonal column including a cylindrical portion, a plurality of spherical cores are arranged in the axial direction, the opposing surfaces are set as a set, a reflecting mirror is provided on one side and an optical fiber is provided on the other side, and an interference filter is provided between the plurality of spherical cores. 2. The distributed refractive index sphere core type wavelength demultiplexer according to claim 1, wherein the distributed refractive index sphere core type wavelength demultiplexer is provided in parallel to the reflecting mirror near the row of the sphere cores. 4 The surrounding medium is a polygonal column including a cylindrical portion, a single spherical core is embedded in the center thereof, a plurality of interference filters are embedded along the periphery of the spherical core, and a reflecting mirror and a cylindrical portion are provided in the polygonal column. A distributed refractive index spherical core type wavelength demultiplexer according to claim 1, characterized in that an optical fiber is provided in the distributed refractive index spherical core type wavelength demultiplexer. 5 The surrounding medium is a polyhedron including a spherical part, a single spherical core is embedded in the center, a plurality of interference filters are embedded along the periphery of the spherical core, a reflecting mirror is placed in the polyhedral part, and a light beam is provided in the spherical part. A distributed refractive index spherical core type wavelength demultiplexer according to claim 1, characterized in that a fiber is provided.