JPS62164019A - Optical directional coupling element - Google Patents

Abstract

PURPOSE:To improve an extinction ratio and to make size smaller and electrification lower by disposing the 2nd semiconductor optical waveguide including a waveguide layer consisting of a multiple quantum well layer to the 1st semiconductor optical waveguide in such a manner that light can be coupled between both. CONSTITUTION:This element is constituted by contg. the 1st semiconductor optical waveguide 7a for incidence of the light to be modulated, the 2nd semiconductor optical waveguide 7b contg. the waveguide layer consisting of the multiple quantum well layer 3 alternately laminated with the multiple quantum well layers consisting of the 1st semiconductor and the barrier layers consisting of the 2nd semiconductor having the forbidden band width larger than the forbidden band width of the 1st semiconductor, and electrodes to impress voltages respectively independently to the 1st and 2nd semiconductor optical waveguides 7a, 7b. A sharp exciton peak by exciton absorption is observed near the absorption end even at an ordinary temp. in the absorption spectra of the multiple quantum well layer 3 and there are Kramers-Kronig relations between the absorption and refractive index. The extinction ratio is thereby improved and the low voltage operation is made possible with the smaller size.

Description

[Detailed description of the invention] [Industrial application field] TECHNICAL FIELD The present invention relates to an optical directional coupling device, and particularly to optical communication.

In the field of optical information processing, this field relates to optical switches that open and close optical signals in optical transmission lines.

[-Conventional technology]

Semiconductor optical switches are important as optical devices for use in future optical communication systems because of their potential for ultra-high-speed switching, low-voltage operation, small size, and easy integration. Among these, there is a directional coupler type optical switch that controls the coupling state between two optical waveguides using an electric field. This will be explained here.

When light is passed through two optical waveguides that are arranged sufficiently close to each other in parallel, optical energy is exchanged between the optical waveguides due to optical coupling between the optical waveguides. When an optical waveguide is made of a material that has an electro-optic effect, when a leading waveguide is placed in an electric field, the refractive index of the optical waveguide changes due to the electric field, and the coupling state changes, so the electric field causes light to flow between the leading waveguides. It is possible to control the exchange of energy. When focusing on one leading waveguide, its output light changes due to fluctuations in the electric field. Therefore, by using this effect, it is possible to construct an optical modulator or a switch using an electric field as a control means.

One of them is the InGaAsP optical directional coupling device (IEICE technical report, Vol. 0QE84-51, 1984, No. 1).
(described on page 01°) has been reported.

An oblique view of an example of a conventional directional firing element is shown in 3L/1, and the manufacturing process of the element is as follows. n'
-1nP (Dopant is Sn, impurity concentration is 1.
2\'10 ``cm-', forbidden band width is expressed in wavelength ^6-
11.927. n' on the substrate 11 consisting of
-1nGaAsP (TP doped, 1.2 x 10
n-cladding layer 12 consisting of n-1nGaAsP (non-doped o,
95y, 10”~I x 1017cm-3, λ□
=], 2 μrn), P′ −rn
GaAsP (Zn doped 1.+1 x 1018c
m-', λ+v = 1. 1]”
- The cladding layer 14 was sequentially formed by liquid phase epitaxial growth. Each n+-crat layer 12 has a thickness of 1.5μ
m, n-waveguide layer 13 is 1.3 μm, p+-cladding layer 1
4 had a thickness of 1.4 μrn. The n-side electrode 15 is made of Au-
Ge-Ni is used, and the ρ side electrode 16 is made of Au-Zn/^U.
Striped electrodes were formed using the lift-off method. Also, the element length is 7 inches.

The interval between the two leading waveguides is 3 μm, and the width of the optical waveguide is 3 μm.

We will discuss the SWITCHing characteristics when focusing on a single optical waveguide of this device. Here, in addition to the switching phenomenon due to the electro-optic effect, the effect of light absorption due to the Franz Keldysch effect is also used. Here, an extinction ratio of 8.6 dB was obtained when the switch ink voltage was 5.

Generally, in optical directional coupling devices that exchange optical energy between two phase-synchronized optical waveguides, the energy of the light incident on one optical waveguide is completely coupled to the other optical waveguide. The length and complete coupling length LO are the propagation constants of the even mode and odd mode of the optical waveguide, respectively. ,β. Then, if the 7-element length that can be found from Lo・π,・′(β and -β(, ) is equal to the complete coupling length, one optical guide should be used to create the phase mismatch required for 100% modulation. The relationship between the refractive index change δn found from the wavepath propagation constant change Δβ and LO is expressed as LO = π, where the wave number at that time is 1 ((l), ``(ko · Sn).

, the refractive index change δn due to the electro-optic effect, which is a control means of a conventional optical directional coupling element, is as follows: Let E be the electric field strength, r be the electro-optic constant, and let n (+) be the refractive index when E=[). S
It is expressed as n = rlo 3r E/'2. Also, assuming that the electric field strength E is proportional to the applied voltage ■, the perfect coupling length L (+ and 5V
Considering the case where the voltage is applied, E=5×in' V, /'C
l11. r = 1.4 X 1O-10(+a,,”
V, n, ==31 ~ 3.5, Sn = 1.3
~1.5:<River 0-4, L0~51m. In the conventional example, the switching voltage is 5 V (upper field strength E = 4 x lθ' V / c
m), but this is because the switching state is incomplete, so the switching voltage for 100% modulation is thought to be higher. To obtain a high extinction ratio, Franz
Although the Keldysh effect is also used, the extinction ratio is still 6 dB, which is not a sufficient value.

[Problem that the invention seeks to solve]

Since the conventional optical directional coupling device described above uses the electro-optic effect, the refractive index change δn due to practical electric field strength is on the order of 10-4 at most, so the device length is as small as several dragons. There is a limit to lower voltage (< 5
There is a problem in that it is impossible to simultaneously satisfy V) and miniaturization (<lnm).An object of the present invention is to provide an optical directional element that is compact and capable of low voltage operation. .

Means for Solving Problem C 1 The optical directional coupling device of the present invention includes a first semiconductor optical waveguide for inputting modulated light, a quantum well layer made of a first semiconductor, and a quantum well layer made of the first semiconductor. a second semiconductor optical waveguide including a waveguide layer made of a multi-quantum well layer in which barrier layers made of a second semiconductor having a large forbidden width are alternately laminated;
．． It has a configuration including electrodes that independently apply voltages to the second semiconductor guided waveguides.

[Function] In the absorption spectrum of a multi-quantum well layer, a sharp exciton peak due to exciton absorption can be seen near the absorption edge even at room temperature, and there is a Kramers-Kronig relationship between absorption and refractive index, and exciton・Due to the presence of the exciton peak, the refractive index spectrum shows a large change near the exciton peak wavelength.

FIGS. 2(a) and 2(b) are characteristic diagrams showing the absorption spectrum and refractive index spectrum of the multiple quantum well layer, respectively.

Next, we will discuss absorption and refractive index when an electric field is applied to the multiple quantum well layer.

When an electric field is applied to the multiple quantum well layer, the exciton peak shifts to the longer wavelength side and its half-width widens. With the movement of the long wavelength side l\ of the exciton beak, the refractive index spectrum 1~ also moves to the long wavelength side.

FIGS. 2(c) and 2(d) are characteristic diagrams showing the absorption spectrum and refractive index spectrum when an electric field is applied to the quantum well layer, respectively.

At a wavelength λ3 on the longer wavelength side of the exciton peak wavelength in the absence of an electric field, the absorption coefficient and refractive index are significantly increased by applying an electric field.

Therefore, when the wavelength of the modulated light incident on the first semiconductor optical waveguide is slightly larger than the exciton peak wavelength of the quantum well layer (λ3 in FIG. 2), a voltage cannot be applied to the second semiconductor optical waveguide. Accordingly, the coupling state between the waveguides can be effectively controlled.

That is, when no voltage is applied to the second semiconductor optical waveguide, the light incident on the first semiconductor optical waveguide is
By phase-synchronizing the two semiconductor optical waveguides, the light is emitted from the second semiconductor optical waveguide. When an electric field is applied to the second semiconductor optical waveguide, the refractive index increases as described above, and the phase synchronization condition between the first and second semiconductor optical waveguides is broken, so that light does not reach the second semiconductor optical waveguide. It will no longer be combined.

〔Example〕

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a perspective view of one embodiment of the present invention.

This embodiment includes a first semiconductor optical waveguide 7a for inputting modulated light, a quantum well layer made of n-GaAs with a thickness of 85 mm, and n--, e o-5aao, + with a thickness of 95 mm. ;^S
barrier layers consisting of 45 layers are alternately stacked n --
A second semiconductor optical waveguide 7 including a waveguide layer made of GaAs, -' Ae g, 5Gao-;^S multiple quantum well layer 3
b, and 1st. The n-side electrode 6a and pI! apply voltage independently to the second semiconductor optical waveguides 7a and 7b, respectively. Il
It has a configuration including an electrode 6b.

n-GaAs, /': ff. , 5Gao, 5^s multi-quantum well layer 3 has a thickness of 1 on one side.
~2μm n”Aff (1,3Gan, 7^
An S cladding layer 2 is provided to form a heterojunction, and on the other side there are 0.2 μm thick n--^ffo, 3G
a(1-7^S cladding layer 4 is provided to form a heterojunction.Furthermore, n-1 e o,)(iag, 7A
On top of the layer 1, there is a layer with a thickness of 1 to 2 μm and a width of 8
jj, m, length 300) trn's mesus l-live-like p--^e,), 3Gan, 7^S Kura...I
De layer 5a.

5I') are arranged at intervals of 10 .mu.In, and p-side type i6a, 6b are provided on the surfaces of the p-^e11.3GaO,7^S cladding layers 5a, 5b, respectively. Note that 1 is an n''-1iaAs substrate and 8 is an n-side electrode.

In the vertical direction, n -GaAs, / , '+erl, ,
, 1Gan, 5^s, the difference in the refractive index between the isometric refractive index of the g-weighted well layer 3 and the Talarado layer on both sides, in the horizontal direction there is a mesa stria-like p゛-~(!n, %
G ao , 7^S Kratsud et al.

Two closely spaced, phase-matched rib-type three-dimensional waveguides are constructed due to the isometric refractive index difference caused by the provision of the waveguide 5b.

Next, one example of the manufacturing method of this example will be explained. n” = Molecular beam epitaxial on GaAs substrate 1.

By the method (MBE method), n+-^(! 1), 3
G a r+ , 7^S cladding layer 2 is formed to a thickness of 1 to 2 μm. Next, an 85-thick n-GaAs quantum well layer and a 95-thick n-GaAs quantum well layer were formed using the same <, MBE method.
-ke O,'1tia0.5^S barrier layers are grown alternately to 45 layers each to n--Ga^s/,! ! g,
, ;Ga, 1.5^S'5 quantum well layer 3 is formed.

Next, by the MBE method, a 0.2 μm thick n--Ae
+1-3 Ga +), 7^S cladding layer 4 and 1-2 μm thick p + - 42'o, 3Gao, 7A
Form a cladding layer. Next, stripe-shaped p-side electrodes a, 6b are formed selectively with ﾜu-ZnH and 7M.
is provided, and using this p-side electrode as a mask, p''-, he.
3Gan, the As-clad layer is removed by selective etching, but at this time, n-1, to! . , 3Ga(
The 1,7^S cladding layer is also removed to a depth of 11.1 μm. Finally, a u-Ge film is attached to the back side and an n-side electrode 8 is provided.

Note that not only the MBE method but also C V using organometallic compounds
The D method or the pond vapor phase growth method may be used.

Next, the operation of this embodiment will be explained. First, modulated light 9 having six wavelengths near the long wavelength side of the exciton peak wavelength is made incident on the first semiconductor optical waveguide 7a including a multi-quantum well layer. If the element length is set to an integral multiple of the complete bond length, when no reverse bias voltage is applied between the p-side electrode 6b and the n-side electrode 8, that is, n --GaAs/ken
, i+Gau, a second layer comprising a qAs multiple quantum well layer.
When the optical waveguide 7b is not affected by an electric field, the optical energy of the modulated light incident on the first semiconductor optical waveguide 7a is completely coupled to the second semiconductor optical waveguide 7b, and the optical energy from the second semiconductor optical waveguide 7b is It can be extracted as modulated light 10. When a reverse bias voltage is applied to the second semiconductor optical waveguide 7b, its n −-GaAs/Ae O,5(iafl,5
As the exciton peak of the As multi-quantum well layer moves, the average refractive index increases, and the two first... The phase synchronization condition of the second semiconductor optical waveguides 7a and 7b collapses, and the energy of the light incident on the first semiconductor optical waveguide 7a is transferred to the second semiconductor optical waveguide 7a.
is no longer coupled to the semiconductor optical waveguide 7b.

Further, since the value of the absorption coefficient in the second semiconductor optical waveguide 7b becomes extremely large as the exciton peak moves, the crosstalk from the first semiconductor optical waveguide 7a is sufficiently suppressed in the second semiconductor optical waveguide 7b. The light is extinguished, and a very large extinction ratio is obtained between the modulated light 9 and the modulated light 10 before and after voltage application. In this way, by combining a large change in refractive index and a large change in absorption coefficient due to the electric field of the multi-quantum well layer, it is possible to obtain an optical directional coupling element having an optical switch function and a large extinction ratio at a low voltage.

Further, this switching action will be explained in detail with reference to FIG. 2 regarding the wavelength of the light source, changes in refractive index, changes in absorption coefficient, element length, and operating voltage.

In the absorption spectrum of Fig. 2 (a>), the wavelength of the incident light is λ3 = l], 84 μm (stone ω = 1.47 eV)9.
In the 5As-multi-quantum well layer, if the transition wavelength between ground order electrons and heavy holes is λ1 and the exciton peak wavelength is ^2, they are λl~0, 112 μm, and λ2, respectively, which are near the short wavelength of the incident light wavelength λ3. ~0°83 μm. At wavelength λ3, this multi-quantum well layer experiences absorption λ1, but the absorption at λ1 is sufficiently small compared to the absorption due to the exciton peak. Further, in this composition, the average refractive index of the multi-quantum well layer with respect to wavelength λ3 is n1=3.48, and the refractive index of the cladding layers above and below it is 3.4. Therefore, when no electric field is applied, if the element length of this optical directional coupling element is set to the perfect coupling length, the modulated light 9 incident on the first semiconductor optical waveguide 7a will undergo a small absorption α1 within the optical waveguide. Just by receiving the light, it is coupled to the second semiconductor optical waveguide 7b and extracted as modulated light 10.

Next, a second semiconductor optical waveguide 7b including a multiple quantum well layer
We will discuss the case where a voltage is applied to.

The absorption spectrum at that time is shown in Figure 2(c). In this case, if the reverse bias voltage is about 5V, the electric field strength will be ~5 x 104V, -"cm, and the exciton peak will be λ2~11, from 83μm to ^4 ~111
15μm'\approximately 20nm, converted to energy 30.
It moves to the lower energy side by ~40meV. Therefore, the absorption coefficient α2 for the wavelength λ of the modulated light 9, ~0.84 μm
is very large compared to α, U2~】0・+cm−
It will be about 1. As the entire absorption spectrum with its excin peak moves to the longer wavelength side by applying a voltage, the refractive index spectrum also moves to the longer wavelength side, as shown in FIG. 2 (d). Therefore, the wavelength λ, = +lJ4μm
The average refractive index of the second semiconductor optical waveguide 7b including the multiple quantum well layer is n,=' when no voltage is applied.
J43, whereas when voltage is applied n
2=L55. The refractive index change due to the electric field at this time is δn-11,07, and from this change, the element length at which phase mismatching occurs completely, that is, the complete bond length, is 10 μm.
The following is true. Considering the practical element length, L o ”
The necessary refractive index change when the thickness is 300 μm is δn=],
4 x 10-3, and a change of this magnitude can be sufficiently achieved with an applied voltage of 5V or less. Therefore 300 knots
The modulated light incident on the first semiconductor optical waveguide 7a at a voltage of 5V or less with an element length of
is no longer coupled to the semiconductor optical waveguide 7b.

Furthermore, since the absorption coefficient of the second semiconductor guiding waveguide 71 is so large that almost no modulated light 10 can be extracted from the cross-over that occurs when the coupling is incomplete, a high extinction ratio is obtained. be able to.

The above explanation has been given using the example of a rib-type semiconductor optical waveguide, but it is not limited to rib-type semiconductor optical waveguides.
A buried type in which a Kurasod layer or the like is epitaxially grown may also be used. Also, although we have shown a case in which two semiconductor optical waveguides are fabricated so that they are phase synchronized, by applying an appropriate electric field to each semiconductor optical waveguide, the phase alignment can be achieved! As long as you can satisfy the tJi requirements, that's fine. Furthermore, the semiconductor optical waveguide into which the modulated light enters is made of a multi-quantum well layer.
There's no need. Furthermore, regarding semiconductor materials, GaAs
zz^12 Not limited to GaAs system, Inl', /'G
aAsP, 1nGaAs.

A material such as InAj'^S may also be used.

[Effects of the Invention] As explained above, the present invention provides a first semiconductor waveguide into which modulated light is incident, and a second semiconductor waveguide including a waveguide layer made of a multiple quantum well layer between the two. By arranging them so that they can be optically coupled, the optical directional coupling element has the effect of improving its extinction ratio, making it smaller, and using less electricity.

[Brief explanation of drawings]

FIG. 1 is a perspective view of an embodiment of the present invention, FIG. 2 (εL),
(+)) is a characteristic diagram showing the absorption spectrum I~ and refractive index spectrum of the multi-quantum well layer, respectively, and Fig. 2 (c
), (d> are characteristic diagrams showing the absorption spectrum and refractive index deviation when an electric field is applied to the quantum well layer, respectively. Figure 3 is a perspective view of an example of a conventional directional coupling element. 1-n'' IEaAs substrate, 2,...n'
-＾e1. HaGal+, 7As crusted layer, 3-n
---1iaAs −′ to en, q(ia
, 1, +; ^S-2 double Yoshiko well layer, 4-n-A
e, +, d: an, 7 clad layer, 53-, 5
1]-1) 'AI! n, tGan, 7^S Cla-'/de layer, 6a, +)l)... n-side electrode, 7a.
. . . 1st semiconductor guided wave path, 7b . . . 2nd semiconductor guided wave path, 8 .
n-side electrode, 16-n side electrode. Kaya 10 Sθ,, fp: pl-fly a, 36p, 7ΔS Kufutsu) 2 pictures included, λ1: ttuλ2 people)

Claims (1)

[Claims] A first semiconductor optical waveguide for inputting modulated light, a quantum well layer made of a first semiconductor, and a barrier layer made of a second semiconductor having a larger forbidden band width than the first semiconductor are laminated alternately. A light direction comprising: a second semiconductor optical waveguide including a waveguide layer made of a multi-quantum well layer; and electrodes that independently apply a voltage to the first and second semiconductor optical waveguides. Sexually coupled element.