JPH03274530A - Semiconductor light amplifier - Google Patents

Abstract

PURPOSE:To efficiently and uniformly excite along a light amplification layer irrespective of the size of the light amplification layer and to obtain a light amplifier of high gain by finding out a method for exciting the light amplification layer by light and providing an amplifier in a resonator of laser for excitation. CONSTITUTION:The amplifier is constituted of the light amplification layer(band gap wavelength lambdaa) 1 in a striped state constituted of non-doped semiconductor formed on one surface of a lower clad layer (lambdacl) 3 constituted of, for example, an n type semiconductor, an activation layer for excitation light (lambdap) 2 formed on both sides of the layer 1 in the extending direction of the layer 1 and an upper clad layer (lambdac2) 4 constituted of a p type semiconductor formed on the light amplification layer 1 and the activation layer for the excitation light 2. Since lambdaa>=lambdain>lambdap>lambdacl,lambdac2 is satisfied in the relation between the band gap wavelength of respective layers and the wavelength lambdain of the light made incident on the end side of the light amplification layer 1, the incident light travels in the light amplification layer 1 as the wave guide and then the light is emitted from the other end side. Thus, the voltage drop in the clad layer is reduced and a reactive current flowing in a buried layer is eliminated, so that the light amplification of high gain can be accomplished.

Description

[Detailed Description of the Invention] [Summary] Regarding an optical amplification device used as an optical repeater or a preamplifier of a photoreceiver in optical fiber communication, 1. The purpose of this invention is to provide a high gain optical amplification device. a first cladding layer (bandgap wavelength λcl) made of a type semiconductor, and a band-shaped layer made of a non-doped semiconductor formed on one surface of the first cladding layer,
Wavelength λ3. an optical amplification layer (bandgap wavelength λ1) into which light enters; and an active layer 1i for excitation light (bandgap wavelength λp) formed on the surface of the first cladding layer made of a non-doped semiconductor and exposed from the optical amplification layer. ), a second cladding layer (bandgap wavelength λcz) made of a semiconductor of the opposite conductivity type formed from the optical amplification layer to the excitation light active layer, and a second cladding layer (bandgap wavelength λcz) that is parallel to the surface of the first cladding layer and that a resonator having both end faces of the excitation-filled active layer as reflective surfaces in a direction crossing the stretching direction of the amplification layer, or a resonator comprising a diffraction grating provided in a region defined by the end faces; and a pair of electrodes provided opposite to each other with the second cladding layer in between, and λa≧λ8, 〉λ2〉λ. 1. λc2
The structure is structured so that there is a relationship between the two.

[Industrial application field]

The present invention is used to amplify weak input light such as in an optical repeater in optical fiber communication, or to amplify input light to a light receiving element such as a PIN photodiode which has a high response frequency but relatively low sensitivity. The present invention relates to a semiconductor optical amplifier.

[Prior Art] The structure of a conventional semiconductor optical amplification device is basically the same as that of a normal semiconductor laser, except that an antireflection film is provided on the open face of the arm constituting the laser resonator. Incident light from the outside enters the active layer through one of the arm openings, is amplified while traveling through the active layer, and is emitted to the outside through the other arm opening. The operation is similar to that of a semiconductor laser in that the population inversion required to make the optical amplification gain (g) g≧1 is achieved by injecting carriers into the pn junction via the active layer. .

(Problems to be Solved by the Invention) Incidentally, the optical amplification layer in the optical amplification device, that is, the optical waveguide, needs to have one fundamental waveguide mode characteristic.

In addition, in order to reduce the dependence of the amplification factor on the polarization plane of the incident light,
It is desirable that the cross section of the waveguide has in-plane symmetry such as a circle or a square. For these reasons, for example 1
．． What is the optical amplification layer in an optical amplification device used for so-called long wavelength optical fiber communication in the 3 μ-band or 1.55 μ-band?

When its refractive index is 3, it is necessary to have a cross-sectional dimension of about 0.5 μm or less. Incidentally, the cross-sectional dimensions of the active layer in a typical semiconductor laser are, for example, 2 μm wide x 0.2 μm thick, which does not meet the above-mentioned conditions for the fundamental waveguide mode.

When the injection current is increased to produce 5 population inversion for the optical amplification layer having the above dimensions.

Heat generation due to voltage drop in the cladding layer increases.

Furthermore, since the reactive current flowing through the buried layers formed on both sides of the mesa stripe including the active layer increases, heat generation further increases. As a result, the temperature of the optical amplification layer increases and the gain becomes saturated. Generally, in order to obtain a high gain, it is necessary to run a current larger than the operating current of a semiconductor laser in a semiconductor optical amplifier, and in order to put the semiconductor optical amplifier into practical use, it is necessary to solve the above heat generation problem. It was considered an urgent task.

The present invention solves the above heat generation problem by causing population inversion by optical excitation instead of current injection as in conventional semiconductor optical amplifiers, and makes it possible to provide a high-gain semiconductor optical amplifier. purpose.

[Means for Solving the Problems] The above object consists of a first cladding layer (bandgap wavelength λcl) made of a semiconductor of one conductivity type, and a non-doped semiconductor formed on one surface of the first cladding layer. A band-shaped layer with a wavelength λa on one end surface in the stretching direction.
Optical amplification layer (bandgap wavelength λa) into which the light of No. 7 enters
an active layer for excitation light (bandgap wavelength λa) formed on the surface of the first cladding layer made of a non-doped semiconductor and exposed from the optical amplification layer; A second cladding layer (bandgap wavelength λ
cm), a resonator having both end faces of the active layer for excitation light as reflective surfaces in a direction parallel to the surface of the first cladding layer and intersecting the stretching direction of the optical amplification layer; and a pair of electrodes provided opposite to each other with the second cladding layer in between, and between the wavelengths λa≧λ2. ．． A semiconductor optical amplifier device according to the present invention, characterized in that there is a relationship: 〉λa〉λCI+λc2.

Alternatively, in the above, instead of having both end faces of the excitation light active layer as reflective surfaces, a guide made of a semiconductor of one conductivity type formed between the first cladding layer and the excitation light active layer may be used. layer (bandgap wavelength λa), and a diffraction grating made of unevenness formed at a pitch of 8 in a direction intersecting the stretching direction of the optical amplification layer at the interface between the first cladding layer and the guide layer, and λa between each wavelength
≧λa□〉λa〉λ1〉λa, λc, and the pitch Δ is 8·λ2/2n, where n is the refractive index of the active layer for excitation light. This is achieved by an optical amplification device.

[For production]

FIG. 1 is an explanatory diagram of the principle of the present invention, in which FIG. 1(a) is a perspective view of a main part, and FIG. 1(b) is a sectional view taken along line XX in FIG. 1(a).

As shown in the figure, 1 is formed on the -surface of a lower cladding layer 3 (bandgap wavelength λcl) made of, for example, an n-type semiconductor. A striped optical amplification layer 1 (bandgap wavelength λa) made of a non-doped semiconductor, and an active layer 2 for excitation light (bandgap wavelength λ9) made of a non-doped semiconductor formed on both sides of the optical amplification layer 1 in the stretching direction. ), and an upper cladding layer 4 (bandgap wavelength λa2) made of, for example, a p-type semiconductor, formed on the optical amplification layer 1 and the active layer 2 for excitation light.

Between the bandgap wavelength of each layer and the wavelength λa7 of the incident light incident on one of the end faces of the optical amplification layer l, λa≧λ
Since there is a relationship such as in>λa>λa, λa, the incident light travels through the optical amplification layer 1 as a waveguide.

The light is emitted from the other end face of the light amplifying layer 1.

A current is passed across the excitation light active layer 2 between the upper cladding layer 4 and the lower cladding layer 3 to cause the excitation light active layer 2 to emit light. This light emission travels back and forth within the excitation active layer 2 and the optical amplification layer 1 using the side surfaces of the excitation light active layer 2, that is, the arm opening surfaces 5 and 6 as reflecting surfaces, and oscillates as a laser. Reference numeral 7 in FIG. 2B indicates this oscillation light, that is, excitation light for exciting the optical amplification layer 1.

The excitation light 7 causes population inversion of carriers in the optical amplification layer 1 . As a result, stimulated emission occurs due to the passage of the incident light (λa7). By making the excitation light 7 sufficiently strong,
Since stimulated emission light sufficient to compensate for absorption loss in the optical amplification layer 1 is emitted, the gain (g) of the optical amplification layer 1, that is,
The ratio Pout/Pi- of the intensity of the emitted light from the optical amplification layer 1 (P,,t) to the intensity of the incident light (Pi,) becomes greater than 1. Optical amplification is performed in this way.

As described above, in the present invention, the excitation light active layer 2 separated from the optical amplification layer 1 is provided between the two common cladding layers 3 and 4, and the laser generated in the excitation light active layer 2 is provided. Population inversion is caused by light. What is the current required to generate this laser light?

Injected over a sufficiently large area. therefore.

The voltage drop in the cladding layer is also smaller.

There is also no reactive current flowing to the buried layer, which solves the problem of heat generation, and as a result, high-gain optical amplification is possible. Since the structure includes the amplification layer 1, highly efficient optical excitation is possible.

In addition, since the optical excitation is performed evenly from the side surfaces along the stretching direction of the optical amplification layer l, the gain distribution becomes uniform. Furthermore, as shown in FIG. 1, a laser consisting of an active layer 2 for excitation light is used as a so-called DFB (Discussion Block), which uses a diffraction grating as a resonator.
In the case of the structure (ributed feed-back), the distance between the arm opening surfaces 5 and 6 of the excitation light active layer 2 can be set substantially arbitrarily, and the design regarding the dimensions and shape of the optical amplification device can be adjusted. The degree of freedom increases. In the figure, reference numeral 8 denotes a diffraction grating formed so that the pitch Δ in the direction intersecting the stretching direction of the optical amplification layer 1 is Δ8λp/2n (n is the refractive index of the active layer 2 for excitation light). In addition, a guide layer indicated by reference numeral 9 (bandgap wavelength λ
a; λ2>λa>λC1+λax) is provided to prevent recombination centers from occurring at the interface between the excitation light active layer 2 and the lower cladding layer 3 due to the formation of the nine diffraction gratings 8. ] Examples of the present invention will be described below with reference to the drawings.

In the following drawings, the same parts as those in the previously shown drawings are designated by the same reference numerals.

FIG. 3 is a detailed explanatory diagram of the present invention, in which (a) is a perspective view of a main part, and (b) is a sectional view taken along line XX in (a).

An example of a method for forming the illustrated structure will be described.

For example, on the surface of an n-type 1nP substrate (lower cladding layer 3), non-doped 1nGaAsP with a thickness of 0.5 μm is applied.
layer (bandgap wavelength λa・1.6 μm) and thickness 0
．． Two 5 μm InPn flange layers are epitaxially grown one after another. The impurity concentration of the lower cladding layer 3 is lXl018
It is about cm-3.

Next, a 5i02 layer is deposited on the 1nP buffer layer and patterned by the well-known glue sogelaf method to form a striped Si0g mask with a width of 0.5 μm. The 1nP buffer layer and InGaAsP layer exposed from this 5iOz mask are sequentially etched by a well-known method until the InP substrate is exposed.

As a result, the inP buffer layer and non-doped InG
A mesa stripe consisting of aAsP layer is formed. The 1nGaAsP layer below this mesa stripe is the optical amplification layer 1.
, and its width and thickness are both 0.5 μm.

Next, with the 5i02 mask remaining, a 1nP substrate surface exposed on both sides of the mesa stripe is coated with 1.
A 0.5 μM thick excitation light active layer 2 made of non-doped InGaAsP (band gap wavelength λa·1.3 μm) and a 0.5 μM thick P-type 1nP layer are epitaxially grown in sequence. This is +InGaAsP
This method utilizes the selectivity that InP and InP grow on the surface of the crystal but not on the surface of the insulating layer.

Next, after removing the SiO□ mask 9, the In
A p-type InP layer with a thickness of 1.5 μm is epitaxially grown on the P-type buffer layer and the P-type 1nP layer.

As a result, on the optical amplification layer 1 and the excitation light active layer 2,
Thickness 2. consisting of the 1nP buffer layer and p-type InP layer.
An upper cladding layer 4 of OVA is formed. The bandgap wavelengths λal and λa of the lower cladding layer 3 and upper cladding layer 4 made of InP layers are 0°95μ theory, and λa>λa>λ. .

The relationship λa is maintained. The impurity concentration of the upper cladding layer 4 is about IXIO''CJl-3.

Next, on the upper cladding layer 4, p” 1nGaAsP
(Impurity concentration 2XIO”cm-3) with a thickness of 0.
A contact layer 8 of 5 μm is epitaxially grown.

Then, the InP substrate constituting the lower ground layer 3 is polished to a thickness of approximately 100 μm.

Then the lower surface of the lower ground layer 3 and the contact layer 1
04, an njvJ electrode 11 and a P-side electrode 12 are formed, respectively. The length (1,) along the stretching direction of one optical amplification layer 1 is 300 mm.
am, and the width (W) is about 300 am. Note that the height (H) is approximately 100 μm.

Thereafter, an anti-reflection film 9 is formed on the end face of the folded crystal in the stretching direction of the optical amplifying layer 1, thereby completing the optical amplifying device. This antireflection film is made of well-known 5iN1, S
It consists of iOx or UOX, and the optical film thickness is λa1/
A thin film of 4 may be used. Since these are generally formed by a method such as sputtering, their respective compositions are Si3N4. The above expression was used because it is presumed that the stoichiometric composition of S+l)z or Ah(h) deviates from the stoichiometric composition.

FIG. 4 is an explanatory diagram of another embodiment of the present invention.

This is a case where a reflective film 13 with a high reflectance is formed on the arm opening surfaces 5 and 6 along the stretching direction of the optical amplification layer 1 in the folded crystal. As the reflective film 13.

For example, a thin gold film or a well-known dielectric multilayer film can be used as appropriate. By providing the high reflectance reflective film 13, the threshold current (I) of the laser consisting of the excitation light active layer 2 can be
This is reduced compared to the case where the arm opening surfaces 5 and 6 are directly used as the reflecting surfaces of the resonator.

FIG. 5 is an explanatory diagram of still another embodiment of the present invention, which has a structure in which one diffraction grating 8 is provided as a reflection surface of a laser resonator made up of an active layer 2 for excitation light.

A method for forming this structure will be explained.

A diffraction grating 8 consisting of unevenness with a pitch (A) of 0.2 μm, for example, is formed on the surface of, for example, an n-type 1nP substrate (with an impurity concentration of about IXIO cm − ), which will become the lower cladding layer 3 . As mentioned above, the above pitch is 8・λa/2n
It is. Formation of the above-mentioned diffraction grating 8.

This may be performed by a lithographic technique using a well-known dual interference exposure method.

Next, on the surface of the n-type 1nP substrate, n-type! nGaAs
A guide layer 9 with a thickness of about 0.1 μm (band gap wavelength λb・
1.1μm) and n-type 1nP (impurity concentration 5XIO
An etching stop layer 15 having a thickness of approximately 0.05 .mu.m is sequentially grown epitaxially.

Next, in the same manner as in the embodiment shown in FIG. 4, an optical amplification layer 1 made of non-doped InGaAsP and having a thickness and width of 0.5 μm (band gap wavelength λa/1.6 μs) is formed on the etching stop layer 15. + Non-doped InG
An active layer 2 for excitation light with a thickness of 0.5 μm made of aAsP (
bandgap wavelength λa・1.3 μm), p-type TnP (
An upper cladding layer 4 with a thickness of 2 μm consisting of an impurity concentration I
A hazy contact layer 10 is sequentially epitaxially grown. In addition.

Generally, a mesa stripe forming the optical amplification layer 1 is formed.

Although the individual irregularities in the diffraction grating 8 are formed parallel to the extending direction, the mesa stripes and the irregularities may be formed so as to intersect with each other to some extent. Further, the etching stop layer 15 is made of InGaAs epitaxially grown thereon.
It functions as an etching stopper when processing the P layer and the InP buffer layer grown thereon into the mesa stripes.

Next, the n-type 1nP substrate that will become the lower cladding layer 3 is polished to a thickness of approximately 100 μm, and then the n-side electrode 11 and the p(l!lii electrode) are placed on the lower surface of the lower cladding layer 3 and the upper surface of the contact layer 10, respectively. form 12.

Then, the width (-) and length (
L) An antireflection film (not shown) is formed on the open surface of the porcelain in the extending direction of the optical amplification layer 1, both having dimensions of about 300 μm, to complete the semiconductor optical amplification device. In the structure of this example as well, the bandgap wavelength of each layer is λ
a≧λ1>λa>λa>λ. 3. The relationship λcm (λa, λa are the bandgap wavelengths of the lower cladding layer 3 and the upper cladding layer 4) is maintained.

Note that in the structure of the embodiment shown in FIG. 5 as well, a reflective film 13 with a high reflectance shown in FIG. 4 is provided.

The effect of reducing dark value current (I) can be obtained.

FIG. 6 shows a case where the excitation charging active layer 2 in the structure shown in FIG. 3 or FIG. 5 is composed of a layer having a multi-quantum well structure, and is shown as a modification of the embodiment shown in FIG. 3. That is, the active layer 2 for excitation light in FIG.
) As shown in 9, for example, the bandgap wavelength λq+□1
．． 15 μI11. I of thickness Ions (0,01u*)
nGaAsP layer 21 and bandgap wavelength λB□1.4
gm and InGaAsP layers 22 having a thickness of 10 layers m are alternately laminated in 25 layers each. This results in a total layer thickness of 0.5 μm.

The emission wavelength in the multiple quantum well layer composed of InGaAsP layers 21 and 22 is determined by the InGaAsP layer 22 having a larger bandgap wavelength (λoz□1.4 am).

Therefore, λa>λa!・λa>λCI+ λc2
．． Or λa>λ. - The relationship λa>λ1>λa, λa is maintained.

By using a multiple quantum well structure as in the embodiment shown in FIG. 6 above, the threshold current (xth) for laser oscillation is lower (
1, and the optical absorption loss is low, so the excitation efficiency for the optical amplification layer is improved.

Note that the optical amplification layer and excitation light active layer 2. As a method for epitaxially growing any of the semiconductor layers constituting the upper cladding layer 4 and other layers, a well-known liquid phase epitaxy method or metal organic vapor phase epitaxy (MO-VPE) method can be appropriately used. It goes without saying that the present invention can also be applied to a structure in which the conductivity types of the semiconductor layers in each of the above embodiments are switched.

〔Effect of the invention〕

According to the present invention, a method is adopted in which the optical amplification layer is excited by light, and since the optical amplification layer is provided within the resonator of the excitation laser, high efficiency and high efficiency are achieved regardless of the dimensions of the optical amplification layer. In addition, uniform excitation is performed along the optical amplification layer, making it possible to provide a high-gain optical amplification device. Further, by using a diffraction grating in the resonator of the optically assisted laser, the degree of freedom in designing the shape and dimensions of the optical amplification device can be increased. As a result, it is possible to realize a high-gain optical amplifier suitable as a preamplifier for high-speed light receiving elements such as optical repeaters and PIN diodes in optical fiber communications.

It is.

In fig.

1 is a light amplification layer, 2 is an active layer for excitation light.

3 is the lower cladding layer; 4 is the upper cladding layer 5 and 6 are the arm openings; 7 is the excitation light.

8 is a diffraction grating, 9 is a guide layer.

10 is a contact layer, and 11 and 12 are electrodes.

13 is a high reflection film, and 15 is an engineering/inching stop layer.

21 and 22 are InGaAsP layers It is.

(shi) Fufu [B month's principle theory Sun Moon 5 Ko (Misaki/) 2) 2nd 2nd Sun Moon original theory J village theory B Moon diary (So/) 1) 1 1 3 :'FeqJqnf tl Theory B Tsukidariko (Wataru/) 2)!
-+ 3 1 Non-invention/) Jitsugi-kan II explanatory notes (various/) I) (b) Non-invention f) Tora Tsumugi 1 theory 7 Bitake pouto) 7 b notes

Claims (2)

[Claims] (1) A first cladding layer (bandgap wavelength λ_c_1) made of a semiconductor of one conductivity type, and a band-shaped layer made of a non-doped semiconductor formed on one surface of the first cladding layer, the stretching direction of which is an optical amplification layer (bandgap wavelength λ_a) on which light with a wavelength λ_i_n is incident on one end face of the optical amplification layer; and an excitation light active layer formed on the surface of the first cladding layer exposed from the optical amplification layer and made of a non-doped semiconductor. layer (bandgap wavelength λ_p
), a second cladding layer (bandgap wavelength λ_c_2) made of a semiconductor of the opposite conductivity type formed from the optical amplification layer to the excitation light active layer, and a second cladding layer (bandgap wavelength λ_c_2) parallel to the surface of the first cladding layer and A resonator having both end faces of the excitation light active layer as reflective surfaces in a direction intersecting the stretching direction of the optical amplification layer, and a resonator provided to face each other with the first cladding layer and the second cladding layer in between. and between each of the wavelengths, λ_a≧λ_i_n>λ_p>λ_c_1, λ_c_2
A semiconductor optical amplifier device characterized by having the following relationship. (2) In the semiconductor optical amplification device according to claim 1, instead of having both end surfaces of the active layer for excitation light as reflective surfaces, a layer formed between at least the first cladding layer and the active layer for excitation light is provided. A guide layer (bandgap wavelength λ_b) made of a semiconductor of one conductivity type, and unevenness formed at the interface between the first cladding layer and the guide layer with a pitch Λ in a direction intersecting the stretching direction of the optical amplification layer. and between the respective wavelengths, λ_a≧λ_i_n>λ_p>λ_b>λ_c_1, λ
_c_2, and the pitch Λ is Λ=λ_p/2n, where n is the refractive index of the active layer for excitation light.