JPS63276004A - Grating coupler - Google Patents

Abstract

PURPOSE:To enlarge an allowable range against a variation of an incident angle and wavelength, by forming a grating which is formed by synthesizing plural grating vectors whose periods are different in an opening of prescribed width and has a refractive index distribution, against a waveguide layer. CONSTITUTION:On a glass substrate 11 whose refractive index is (ns), a glass layer whose refractive index is (nc) (nc>ns) is formed by a spattering method, and it becomes a waveguide layer 12. Subsequently, an electron beam resist 14 (refractive index nG) which can control height of unevenness by a dope quantity of an electron beam is applied, a grating is brought to picture drawing by the electron beam, and a relief type grating 15 in which the electron beam resist 14 is allowed to have a periodic characteristic of a refractive index is formed. That is, it is formed as the grating 15 constituted so that a cross sectional shape of the electron beam resist 4 loaded on the waveguide layer 12 becomes an uneven shape corresponding to a prescribed periodic function. In such a way, even if a variation of an incident angle and a wavelength variation are generated, the waveguide can be executed in a state that the efficiency is scarcely varied.

Description

DETAILED DESCRIPTION OF THE INVENTION Technical Field The present invention relates to a grating coupler used as an optical waveguide device such as an optical IC 1 optical integrated device, an optical integrated pickup, an optical communication device, or a sensor IC.

Prior Art The development of optoelectronics-related technology in recent years has been remarkable, and integration is progressing.

For example, in an optical pickup such as a magneto-optical disk, an optical waveguide device is used for integration in order to achieve high-speed access through reduction in size and weight. Here, in such a guided waveguide device, a waveguide layer is formed on a substrate, and in order to propagate and couple incident light to the waveguide layer, for example, a grating (
A diffraction grating) is formed and the waves are guided through this grating. Optical coupling using such a grating is described, for example, in ``Light Wave Electron Optics'' published by Corona.
(written by Tsugube Koyama and Hiroshi Nishihara), pages 98 to 108, etc. Such a grating coupler is required to have little variation in the waveguide efficiency of incident light in the waveguide layer. In particular, in optical pickup, optical communication, etc., the incident light contains information.
It is important that efficiency fluctuations be small.

However, this type of grating is generally configured as a grating with a single periodic structure, and as the aperture width increases, the width of the grating vector becomes narrower, which affects the coupling conditions of incident light into the waveguide layer. becomes severe, and the amount of guided light tends to fluctuate. That is, when the wavelength λ of the incident light changes or the incident angle θ changes, the coupling efficiency changes. As a result, when considering practical use as a waveguide device, there are disadvantages such as positional accuracy that strictly controls the angle of incidence, and the difficulty of using a semiconductor laser that is prone to wavelength fluctuation as a light source. There is. Furthermore, when assuming an optical pickup for a magneto-optical disk, it would be convenient if both TE mode and TM mode waves could be coupled and propagated in the waveguide layer when creating a waveguide device structure. Mode waves have only a slight difference in effective refractive index and cannot be simultaneously propagated into the waveguide layer.

By the way, there are gratings called chirp gratings in which the period gradually changes depending on the location (for example, "Optical Integrated Circuits" published by Ohm Co., Ltd. (written by Hiroshi Nishihara et al.)).
(see pages 273 and 274 of the above), even if the wavelength or angle of incidence of incident light changes, it can be coupled and propagated into the waveguide layer. However, this makes coupling possible by changing the incident position, and propagation coupling is not possible if the incident position of the incident light on the waveguide layer is fixed. In other words, when considered within a predetermined limited opening width, it can be said to be a single periodic structure.

Purpose The present invention has been made in view of the above points, and even if the incident light entering the same aperture area has a wavelength variation, an incident angle variation, etc., or even if there is a variation in the wavelength or incident angle of the incident light, such as TE mode light or TM mode light. An object of the present invention is to obtain a grating coupler that can efficiently couple incident light to a waveguide layer even if the incident light has different modes.

Structure In order to achieve the above object, the present invention forms, in a waveguide layer, a grating having a refractive index distribution formed by synthesizing a plurality of grating vectors with different periods within an aperture of a predetermined width. It is characterized by achieving a broadband grating vector over the entire grating and increasing the tolerance range for variations in incident angle and wavelength.

Hereinafter, one embodiment of the present invention will be described based on the drawings and according to the following items.

A. Analysis of conventional grating B1 Principle and theoretical analysis of grating of the present invention C0 Configuration example C-1 Configuration example 1...Surface relief loaded type configuration C-2 Configuration example 2...Surface relief integrated type configuration C- 3 Configuration example 3・
...Refractive index distribution type configuration C-4 and others, Design example D-1 Design example 1... Margin for incident angle variation D-2 Design example 2... Margin for wavelength variation D-3 Design example 3...
- Analysis of TE and TM mode optical propagation A and conventional grating coupler First, before explaining the present invention, a conventionally known grating coupler with a single period and a finite aperture will be analyzed. That is, as shown in FIG. 13, the propagation constant for guided light in a certain mode is σ=
Consider the leading waveguide 1 where σx (−nE·2π/λ) (a value unique to the waveguide). This optical waveguide 1 has a refractive index of n
A waveguide layer (core layer) 3 having a refractive index nc (where nc>ns) is formed on a substrate 2 made of glass, pyrex, or the like. nQ is the refractive index of the outside, which is a lossless medium. A grating (periodic structure) 4 formed with a single periodic refractive index distribution is formed in a part of the waveguide layer 3.
is formed with the width of the opening. The effective refractive index nH of the waveguide layer 3 is determined by the refractive index nc of the material of the waveguide layer 3, the thickness of the waveguide layer 3, and the refractive index of the cladding layer sandwiching the waveguide layer 3.

Therefore, at an incident angle θ from the outside, the propagation constant ρ(lipl
=p=2π/λ+ n @ =1 + ρX=ρ'
Consider a case where the grating 4 portion is irradiated with plane wave light of 5lno). Here, if the grating vector (grating vector) for grating 4 is K (where l K l = K = K x), then in the direction of pitch 8 of grating 4, = 2π/Δ
・・・・・・・・・・・・・・・・・・・・・
・・・・・・・・・(1) There is a relationship. Now, assuming that the propagation direction of light in the leading waveguide l is the X direction and the thickness direction of the optical waveguide l is two directions, the refractive index distribution in the grating 4 portion is as follows: n(x)=ncoNsr+n, ・cos(K− x) --
”12). However, n C0N
It is assumed that 8T is a constant value and n coNsT> n +. From equation (2), if we extract only the refractive index modulation, we get n, (x)=n, -cos(K-x) --...
(3) is obtained.

Here, if the aperture in the X direction of the grating 4 is divided centrally symmetrically into positive and negative, and x : (-D/2.D/2), then the spectral distribution of the grating vector will be the spatial frequency (wave number By performing Fourier transformation on the vector) as kx, it is expressed as 5(kx) by the following equation.

jkzx S(kx) = J"n, -cos(Kx)e
dxjkxx = f n, ・cos(Kx)e dxn2 = ”(sine(p-(kx 10K))+5ine(”
(kz-K)))・・・・・・・・・・・・・・・・・・
It is given by the sink function (4). Here, the sink function 5inc means 5inc
It is defined by the formula x = sinx / x. This equation (4) is schematically illustrated in FIG. 14. In other words, the spectral distribution of the lattice vector 5(kx
) becomes a spectrum with a peak value at kX=±K.

As a result, a lattice vector having a width of ±2π/D for kx = ±K (ie, the width between the zero cross points closest to the axis) is designed.

However, the periodic structure of the grating 4 is sinusoidal, and when considering the bandwidth of the grating vector, the peak value of the sink function is 1, and the width between the zero crossing points closest to the axis is Δx.
=2 π=4 π/D rad/n@, it is appropriate to consider the half-width, which is half the peak value, as the bandwidth. In other words, according to the number of sinks, the half width Δ that replaces the zero cross point width Δx=2π is 3.79099 r
ad=1.20671πrad, 4π/D r
The half width Δ instead of ad/nm is 2.41342
yt/Drad/nm.

Generally, the diffraction conditions for coupling the incident light to the guided light by the grating 4 having such a structure are as follows: a=p+qK
・・・・・・・・・・・・・・・・・・・・・・・・
・・・・・・(5) (However, q”Ot ±1.±2.
) is given by

Here, for the guided light, as shown in FIG.
Since only the directional component needs to be considered, the relationship between the incident light ρ and the guided light σ (=σX) is σX=ρx1q K
・・・・・・・・・・・・・・・・・・・・・・・・
...(6) (however, q=0.±1.±2.....) can be rewritten as follows. If q is selected that satisfies this conditional expression (6), there will be as many conditions as possible for the incidence. Now, if q=l, equation (6) is σX=ρx+K...
. . . It is shown as (7). This can also be understood from FIG. 15.

Here, the X component of the incident light ρ is ρX;ρ・sine, and the guided light σ is given only by the X direction component, so σ
Since it can be expressed as X=σ, 2π.

σX engineering□S1nθ10K・・・・・・・・・・・・
...(8) Denoted as λ.

Here, if K has a width as expressed by equation (4), that is, a bandwidth, under the condition that K x OX, then (6
Since ρz (=psinθ) that satisfies the condition of the equation ), even if ρ or θ of the incident light changes slightly, a portion that satisfies the propagation constant OX will be guided.

In other words, if the value of the lattice vector is formed to have a width, the incident angle θ and the wavelength λ of the incident light will correspond to this.
OX light can be propagated into the waveguide layer 3 even if the condition deviates somewhat from the designed value.

However, in reality, when the aperture width of grating 4 is determined to be a certain value, the bandwidth of the grating vector is also fixed to a certain width, so a grating vector with an arbitrary bandwidth wider than this can be obtained. It is not possible. Therefore, when there is a fluctuation in the incident light, the waveguide conditions for the waveguide light become severe.

B0 Principle and theoretical analysis of the grating of the present invention In order to improve the drawback of the limited grating vector bandwidth, the present invention provides grating vectors with different periods within the same grating aperture set to a predetermined width. By synthesizing grating vectors to form a grating, a plurality of different grating vector components are included, thereby attempting to widen the grating vector bandwidth.

First, consider simple composition of multiple lattice vectors. This corresponds to the case where 1 is combined with N lattice vectors with different periods for equation (1), and the refractive index distribution at that time is based on equation (2), n(x) = nc'0NsT +In, 1 ・cos
It is expressed by the formula (Kt·x) −(9)i=1. According to equation (9), the modulation part of the refractive index is n, (x) = Σn, 1- cos (K1-x)
−・・−・・−・(10) i=1 It is shown by the formula. Assuming that the aperture width is constant for all i, when this equation (10) is Fourier transformed, the following equation is obtained.

・・・・・・・・・・・・・・・・・・・・・・・・(
11) That is, 1 ''' l, L '''+ N, qt
= Or ±1°±2. ...Under the conditions, OX':pX1 + (1 in 11
In the formula °゛°°°°゛°゛° (12), each i
If even one of the conditions is satisfied, light with a propagation constant σX will be guided through the waveguide layer. The basic idea is to synthesize a plurality of such lattice vectors with different periods.

Next, we will consider widening the band into lattice vectors, that is, means of synthesizing them. This is a study on an optimal design method for making the value of the grating's spectral distribution 5 (kx) have a flat characteristic in a certain region.

In general, according to the sampling theorem of signal theory, any function f(
x) sink function 5incx = sinx/X
The optimal expression method is 5i as shown in Figure 10.
By moving ncx by π, △ f(x)=Σf (x-mπ) 5ine (x-mπ
)...--...--(13)△. Here, f(x) in equation (13) means an approximate equation of f(x).

Using the sampling theorem using such a sink function, we can approximate the case where the spectral distribution 5(kx) has a box-shaped band characteristic like a rectangular pulse. It can be considered separately.

In both cases, the central lattice vector is ±.

First, in the case of composition of an even number of sink functions.

For example, as shown in FIG.
) ))Ko··················(
14). On the other hand, in the case of combining an odd number of sink functions, for example, as shown in FIG.
・・・・・・・・・・・・・・・(15)

Therefore, by inverse Fourier transforming these equations (14) and (15), the refractive index distribution equation n(x) becomes (14)
Equation (15) becomes equation (17), and so on.

n(x)=Σn, cos [(K+-5(2m+1
)l ・x) −・・・・(16) As a result, in the case of composition of an even number of sink functions, (1
From equation 6), for the central lattice vector ±K, ±2πN
/D ・・・・・・・・・(18
) almost flat spectral distribution in the band 5(kx)
give. Similarly, in the case of combining an odd number of sink functions, as can be seen from equation (17), the central lattice vector ±
For K, ±π(2N+1)/D...
...(19) gives a substantially flat spectral distribution 5(kx).

As described above, according to the grating coupler configuration of the present invention which has a refractive index distribution by combining a plurality of grating vectors with different periods, these (18) (19) ), it is possible to widen the band of grating vectors as a grating. As a result, even if the incident light on the grating is incident on the grating at a deviation from the reference angle of incidence, the wave can be guided into the waveguide layer with constant efficiency. This means that the precision of the incident angle and the like can be relaxed in optical pickup, etc., and has the potential to make it easier to put the guided waveguide device into practical use. Furthermore, even if the wavelength of the light source of the incident light changes somewhat, waveguide coupling to the waveguide layer is possible. This means that even semiconductor lasers whose wavelengths tend to fluctuate due to thermal effects can be used, and in two ways, it is possible to put waveguide devices, including light sources, into practical use at low cost. Become. Furthermore, even TE mode and TM mode waves having a slightly different effective refractive index can be guided and coupled into the waveguide layer at the same time at the same radiation angle. This makes it possible to detect the polarization state, etc., thereby increasing its practicality as an optical pickup for magneto-optical disks. Furthermore, from a developmental perspective,
It is also possible to simultaneously inject and guide light of two or more wavelengths, and it can be used as a wavelength-multiplexed waveguide device.

C0 configuration example Based on the principles and theory described above, the actual grating should be formed with a refractive index distribution that satisfies the conditional expression after inverse Fourier transform shown in equation (16) or (17). This can be achieved by

C-1 Configuration Example 1...Surface relief loading type configuration FIG. 1 shows an example of the configuration. This is configured as a leading waveguide 10 having a loaded grating coupler having a refractive index distribution formed by combining grating vectors with two types of periods, for example, as a plurality of different periods. First, on the glass substrate 11 having a refractive index ns, a refractive index nc (nc>n
s) A glass layer made of Corning #7059 or the like is formed with a thickness of 0.76 μm by sputtering method,
This will be referred to as the waveguide layer 12. At this time, as shown in the figure, a buffer layer 13 such as SiO□ is formed between the glass substrate 11 and the waveguide layer 12.
(refractive index nB) may be formed. Then, a photosensitive agent, ie, an electron beam resist 14 (refractive index na) whose height of unevenness can be controlled by the amount of electron beam doping is applied. Therefore, a grating is drawn on such an electron beam resist 14 using an electron beam, and this electron beam resist 14 is
For example, a relief type grating 15 having a refractive index distribution, that is, a periodic characteristic of the refractive index by changing the cross-sectional shape according to equation (16), is formed. That is, the grating 15 is formed such that the cross-sectional shape of the electron beam resist 14 loaded on the waveguide layer 12 has an uneven shape according to a predetermined periodic function. The opening width of this grating 15 is D. or,
It is desirable that the height of the unevenness of the grating 15 is equal to or less than the wavelength.

C-2 Configuration Example 2: Surface relief integrated configuration FIG. 2 shows an example of the configuration. Basically, the case is the same as the case of FIG. 1 described above, but whereas in FIG. In the waveguide 20, a grating 25 is formed directly on the surface of the waveguide layer 22.

That is, in order to draw the grating 25 on the glass substrate 21 with an electron beam, a waveguide layer 22 (having a refractive index of nc and nc>n3) which also functions as a photosensitive agent is formed by coating. Then, for such a surface of the waveguide layer 22,
The grating 25 is drawn with an electron beam so as to have a periodic characteristic of the refractive index according to the equation (16) or the equation (17). Thereafter, development is performed to obtain a leading waveguide 2o provided with a surface relief type grating 25 having an uneven cross-sectional shape as shown in FIG. The grating 25 in this case is also formed with a predetermined opening width. Here, according to equation (16) shown in the principle, etc., the modulation amount of the refractive index distribution is calculated by the Habo grating 2 if it is a planar grating.
It is proportional to the height of the unevenness of No. 5.

C-3 Configuration Example 3...Refractive index distribution type configuration A grating 35 with a refractive index distribution is applied to the waveguide layer itself with periodic characteristics of the refractive index according to periodic function characteristics such as equation (16) or (17). FIG. 3 shows an example of the structure of the waveguide 30. FIG. First, a waveguide layer 32 having a high refractive index nc having a relationship nc)n3 is formed on a glass substrate 31 having a refractive index n3. The waveguide layer 32 is made of a material whose refractive index changes depending on the doping amount of the electron beam to form a refractive index grating 35, for example, As25 with n c = 2.47. Formed by etc. Also in this case, the opening width of the grating 35 is D.

C-4 Others In the above-mentioned configuration examples 1 to 3, the gratings 15, 25, 35 are formed by the doping amount of the electron beam, but other methods such as maskless deposition method using a focused ion beam may also be used. It can also be formed by ion milling method. In addition, in the case of a surface relief type with unevenness, even if the pattern is once formed on a mask and then processed onto the leading wave layer, it may be carried out, for example, by the so-called 2P method using an ultraviolet curable resin. obtain.

D. Design Example In order to clarify the effects of the grating structure of this example, numerical values based on a design example of a practical grating structure are shown.

D-1 Design Example 1... Allowance for Incident Angle Variation Even if the incident angle of the incident light on the grating varies, the grating vectors with five different periods can be guided into the waveguide layer. An example of grating design by combining i=o, ±1°±2) will be described with reference to FIG. FIG. 4 schematically shows a waveguide 40 in which a waveguide layer 42 is formed on a substrate 41, and a grating 45 is formed on this waveguide layer 42. Here, since the number of composite lattice vectors is 5, which is an odd number, the design is basically based on equation (17) obtained by inverse Fourier transform. First, - Reference angle of incidence θ = 15° - Incident light wavelength (center wavelength) λ. = 632.8 nm・Grating aperture width D = 1 mm = 10@nm・
Effective refractive index of the waveguide layer 42 nH = 1.5160 (value for center wavelength λ = 632.8 nm) - Modulation amount of refractive index for each period (amplitude value) n = 0.01 - Interval of each period component 2 π/D = 6.28X 10-”rad/nm・
Propagation constant of incident light p = 2yt/λ = 9.9292 X 10-' ra
d/nm x-direction component of propagation constant of incident light px = psin15' = 2.5699X10-'ra
d/nm・Propagation constant of guided light σ=σX=2πnH/λ = 1.50526 X 10−” rad/nm・1
Spectral half width of periodic component (see Figure 5) ΔK =
1.2067 yt/D = 3.79X 10-
'rad/nm・Lattice vector K (15°) at incident angle 0 = 15° = OK -p X = 1.24827
X 10-"rad/nm and substitute it into the refractive index modulation after inverse Fourier transform in equation (17). In this equation,
Since the number of cycles is 5, N=2. This results in (
17) The formula is n (x) = 0.01 X (cos(
1,24701x 10-” x )+cos(1,2
4764X 10-” x )+cos(1,2482
7X 10-” x ) 10 cos(1,24890X
10”” x )+cos(1,24953X 10-
” x )).

The range of incidence angles 01 to θ2 at this time is defined as K(
λ. ), the lattice vector has approximately 1,24664 x 10-”≦=1.24991
X 10-” rad/nm. By finding the angle of incidence at both ends of K with such a bandwidth,
The range of incident angles θ, ˜θ2 that can be guided by the grating 45, which is the width of the aperture, is determined.

Here, the angle θ can be obtained as ρ sin θ=σx−.

,', θ, = 14.91' ≦θ≦θ, =
15.09', and has an incident angle width of 0.18a.

This means that when the reference incident angle θ is set to 15°, the grating 45 can guide the incident light to the waveguide layer 42 even if it fluctuates by ±0°09°. Especially the first
As can be understood from the sampling theorem shown in Figure 2, for example, a band with flat characteristics is obtained by sampling in a box shape by combining five lattice vectors. Even if there is, the guided light in the waveguide layer 45 is coupled with a substantially constant amount of light.

Incidentally, in the case of a conventional single-period grating structure with aperture width D = 1 mm, only the half-width at half maximum value ΔK = 1.2067π/D of one grating vector acts, so K(15') ± (1,2067π/D), so 14.98°≦θ≦15.02° and 0.
It has a margin of only 0.04°. On the other hand, according to this embodiment, there is a margin of 0.18° as described above, and the band can be made about 4.5 times wider. Also, in the conventional structure, when calculated in the same way, the opening width D = 0.0
When it is monovalent, 12.74°≦θ≦17.28° and the width is 4.54°, and when the opening width D = 0.1mm, it is 14.77”≦θ≦15.22°, which is 0.4
Although it has a width of 5°, such an aperture width setting is too small to be practical as a grating. On the other hand, in the conventional structure, when the opening width D=2 mm, 14
．． 99°≦θ≦15.01″′ and 0.02′ t
, there is no width, and if the opening width D = 5 mm, then 1
4.997°≦O≦15.003°, and there is no longer any angular width.

D-2 Design example 2: margin for wavelength fluctuation Even if the wavelength of the light incident on the grating 45 fluctuates, the waveguide layer 4
An example of a grating design by combining 1 (i=±1.±2) with, for example, four types of grating vectors with different periods so that the wave can be guided in 2 will be explained with reference to FIGS. 6 and 7. . First, - Incident angle θ = 15° (= constant) - Incident light wavelength (center wavelength) λ. = 790 nm・Grating aperture width D=1 mm=10@nm・Effective refractive index of waveguide layer 42 nE=2.02 (center wavelength λ,
= value for 632.8 nm) - Amount of modulation of refractive index (amplitude value) n for each period, = 0.01 - Interval of grating vector components in each period 2 yt / D = 6.28 X 10-" ra
d/nm・Propagation constant of incident light p=2rc/λ=7.9534X10-'rad/nm
・X-direction component of propagation constant of incident light p x = p 5in15° = 2.0585 X 1
O-3 rad/nm・Propagation constant of guided light (center wavelength λ.

) σ=σX=2πnE/λ = 1.60659 X 10-” rad/nm・1
Spectral half width of periodic component (see Figure 5) ΔK =
1.2067 π/D = 3.79X 10-”r
ad/r+m・Lattice vector K (15°) at incident angle θ=15°= OX −p x= 1.40074X
10-" rad/nm and substitute it into the refractive index modulation after inverse Fourier transform in equation (16). In this equation, the number of periods is 4, so N=2. As a result, (16
) formula is n (x )=0.01 X (cos(1,3
9980X 10" x ) + cos(1,4004
2X 10-” x ) + cos(1,40105X
10-” x ) + cos(1,40168x 1
0−”x)].

Referring to FIG. 7, the half-width at half maximum ΔW of the spectrum of the grating vector of such a composite grating is determined as = 1.322 x 10-' rad/nm. Therefore, the spectral bandwidth of the grating vector is: K(λ.)±1.322X10-' by 1,'', 1.39942X10-''≦≦1.402
06 x 10-2 rad/nm.

Here, according to the diffraction conditional expression, psin15°=σX−
Then, from ρ=2π/λ, the range of incident wavelengths λ, ˜λ2 that can be guided by the grating 45, which corresponds to the aperture width, is determined using the following formula. That is, 785.0≦λ≦795.0 nm, and a wavelength fluctuation width of 10.0 nm can be tolerated. In other words, under the condition that the center wavelength is set to 790 nm, even if the wavelength varies by ±5 nm, the grating 45 can guide the incident light into the waveguide layer 42 with a substantially constant amount of light.

By the way, in the case of a conventional single-period grating structure with aperture width D=1 mm, only the half-width at half maximum value Δ of the spectral distribution due to one grating vector acts, and K(λ.)± (1,2067π/D), so the lattice vector has 1, and 40036≦≦1.40112×1O−2
rad/nm. Therefore, 788.6≦λ≦791.5 nm, the bandwidth is less than 2.9 nm L, and waveguide becomes difficult if the wavelength changes. On the other hand, according to this embodiment, the bandwidth is 10 nm, which is about 3 nm compared to the conventional one.
A twice the wavelength variation tolerance band is secured.

D-3 Design Example 3... TE mode light and TM mode light (both fundamental mode lights) incident on the TE and TM mode light propagation grating 45 at the same incident angle are simultaneously guided into the waveguide layer 42. An example of a grating design that can be corrugated will be described with reference to FIGS. 8 and 9. First, - Incident angle θ = 15° (= constant) - Incident light wavelength λ. = 632.8 nm (= constant) Grating aperture width D = 1 mm = 10' nm Effective refractive index of waveguide layer 42 TE mode light n ETE = 1.5160 TM mode light nETM = 1.5125 Refraction for each period Amount of rate modulation (amplitude width) n, = 0.01 ・Propagation constant of incident light p = 2yr/λ = 9.9292 x 10-rad/
nm・X direction component of propagation constant of incident light p X= p 5in15°= 2.56986 X
10-rad/nm. In addition, the propagation constant σ of guided light is different for TE waves and TM waves, and each OTM = 27rxnE" = 1..5017.9X
10-" rad/nmλ.

Therefore, the grating vector KTE+KTM that can guide the incident TE wave and TM wave is KTE=(7TE p
)(=1.24828X10-2rad/nmKTM=
aTH-p z = 1.24480X 10-”ra
d/nm. That is, as shown in FIG. 9, there is a vector shift Δ between the TE wave and the TM wave.
/nm.

On the other hand, the width of the grating vector of the grating 45 due to the aperture width is the half-width at half maximum value ΔW, and as in the case of the design example described above, ΔW = 3. ,
．． 79×10-' rad/nm. The value of this half-maximum half-width value ΔW is the half-maximum deviation amount Δ/ between the TE and 'TM waves.
2; This value is considerably smaller than 1.768 x 10-' rad/nm.

This can be easily understood from FIG.

Then, when the modulation component n(x) for the refractive index distribution of the grating 45 is calculated based on equation (16), (KT
E and KTM, N=1), n(x)=0.
0IX (cos(1,24828X 10-”x)+co
s(1,24480X 10-" x )).

Therefore, the bandwidth when converted to the incident angle of the lattice vector of each of the TE and TM waves is 1.24790X1
0″″≦KTE≦1, 24866×10−”r
ad/nm1.24442 x 10-”≦KTM≦1
．． 24518X 10-" rad/nm. Here, since the vector is = σ - ρ sin θ, using
TE+07M is 14.98°≦e TE+e TM≦15.02
° (0.045° at half width), the same person angle of incidence θ = 1
It is possible to design a grating 45 that can guide both TE mode waves and TM mode waves incident at an angle of 5°. In other words, the lattice vector KTEI KT as described above
TE by forming a grating 45 having a refractive index distribution obtained by synthesizing M.

Simultaneous waveguiding of both TM mode waves becomes possible.

Incidentally, in the case of a conventional grating having an aperture width D=1 mm and a single periodic structure, simultaneous waveguiding could not be achieved when TE and TM waves were incident at the same angle.

Effects As described above, in the present invention, a grating having a refractive index distribution formed by combining a plurality of grating vectors with different periods within an aperture of a predetermined width is formed in a waveguide layer.
The grating vector becomes broadband, and the incidence and waveguiding conditions can be relaxed. Therefore, even if there are variations in the incident angle or wavelength of incident light entering the same aperture, there is little variation in efficiency. It is now possible to guide waves in the same way, and practical use can be achieved by relaxing the accuracy of installation conditions, etc. On the other hand, even if the mode lights have a slightly different effective refractive index, such as TE mode wave and TM mode wave, the same It can guide waves simultaneously at the incident angle, and can be used for detecting the state of polarization, for example, in optical pickup for magneto-optical disks.Furthermore, it is also possible to simultaneously input light of two or more wavelengths. It is what it is.

[Brief explanation of drawings]

FIG. 1 is a schematic side view showing configuration example 1 of the present invention, FIG. 2 is a schematic side view showing configuration example 2 of the present invention, FIG. 3 is a schematic side view showing configuration example 3 of the present invention, and FIG. The figure shows design example 1 of the present invention.
5 is an explanatory diagram of the spectral distribution for the design, and FIG. 6 is a design example 2 of the present invention.
7 is an explanatory diagram of the spectral distribution for the design, and FIG. 8 is the design example 3 of the present invention.
Fig. 9 is an explanatory diagram of the spectral distribution for its design, Fig. 10 is a characteristic diagram for explaining the present invention in detail, and Fig. 11 is a spectrum when an even number is combined. An explanatory diagram of the distribution, Fig. 12 is an explanatory diagram of the spectral distribution when odd numbers are combined, Fig. 13 is a schematic side view showing a conventional example of a single periodic structure, and Fig. 14 is an explanation of the spectral distribution for its analysis. FIG. 15 is a vector diagram. 12... Waveguide layer, 15... Grating, 22...
... Waveguide layer, 25... Grating, 32... Waveguide layer, 35... Grating, 42... Waveguide layer,
45... Grating, D... Opening width, K...
Lattice vector applicant Rico Co., Ltd.
Akira Aiki, 'N
j,,,'3 d figure λ

Claims (1)

[Claims] A grating coupler characterized in that a grating having a refractive index distribution formed by combining a plurality of grating vectors with different periods within an aperture of a predetermined width is formed in a waveguide layer.