JPS6387782A - Semiconductor raman laser - Google Patents

Abstract

PURPOSE:To obtain a long life and compact semiconductor Roman laser by using an implantation type semiconductor laser as an excitation light source. CONSTITUTION:In the case of GaP Roman laser 1, an incident light 4 emitting by a YAG laser comes at a weight of 1MW-2MW and a beam diameter of around 1 mmphi is that of a typical instance. Accordingly, once a semiconductor laser beam is used as the incident light 4 and when a pulse input power is 1 W, a beam having 1 mum enables such a laser to obtain inputs having an almost equal density and to carry out Roman oscillations. The semiconductor laser 2 and semiconductor Roman laser 1 are so integrated on one semiconductor substrate that it is unnecessary for them to have a specific external optical system such as a microscope and the like. Yet, a compact and long lived laser can be achieved.

Description

[Detailed description of the invention] The present invention relates to semiconductor Raman lasers.

An oscillator in the infrared/far-infrared region that uses phonons in a semiconductor can be created as a small semiconductor oscillation source by exciting the phonons using an electrical excitation method, as proposed by the inventor in Patent No. 844479. can be obtained.

An electrical excitation method that does not require a light source is the most ideal, but an optical excitation method is also simple if a suitable excitation light source is available, and has the advantage of being able to change the wavelength of the output depending on the angle of incidence of the light. The present inventors also proposed this in Japanese Patent Application Laid-Open No. 49-5292. That is, it is stated that by selecting a semiconductor having a forbidden band width close to the photon energy of the incident laser beam as a Raman active material, a practical semiconductor Raman laser with a low required input laser beam input value can be obtained. However, gas lasers and solid lasers such as Co2 lasers, He-Ne lasers, and ruby lasers are used as lasers for incident light sources. Since these lasers use a 7 liter tube, they have a short lifespan and are large in size.

The present invention aims to provide an optically pumped semiconductor Raman laser without such drawbacks, and is characterized by using an injection type semiconductor laser as a pumping light source. As is well known, injection type semiconductor lasers have a significantly longer lifespan and are smaller in size than the lasers mentioned above.

By combining or integrating a semiconductor Raman laser with an injection type semiconductor laser, a compact near-infrared and far-infrared light source can be obtained.

Since there are many types of semiconductors, various wavelengths can be selected.

When the photon energy 1ω1 of the incident light is slightly smaller than the constraint width E, a resonance effect occurs as illustrated in FIG. 1, and the Raman amplification factor becomes significantly large. This effect was described in Japanese Patent Application Laid-Open No. 49-5292, but will be discussed in more detail here.

In FIG. 1, 1 is the conduction band and 2 is the 111i electron band.

Raman scattering is cut off due to the resonance effect with the upper level 1, so if 1ωi is brought closer to the forbidden band width E,), the resonance effect becomes larger. However, when 1ωI is larger than E&, the absorption coefficient is as high as 10 to 10 C11 due to absorption between bands, so the loss is large and Raman oscillation is difficult.

As for the semiconductor Raman laser body, the m-v emission compounds GaP and QaAs have a particularly large Raman scattering cross section for near-infrared incident light of about 1 μIIl, which is about 1000 times larger than that of CdS. . Since GaP is an indirect transition semiconductor (it has lower absorption than 3aΔS), it is more desirable.

In a GaP Raman laser, the incident light as shown in Figure 2 is IMW to 2 MW when it is a YAG laser, and the beam diameter is about 1u + φ, which is a p4 type example, so when using a semiconductor laser light as the incident light, it is pulsed. If the input power is 1W, if the beam diameter can be reduced to 1μ amount φ, it is possible to obtain an input with almost the same density, and Raman oscillation is possible. Since a semiconductor laser and a semiconductor Raman laser can be integrated on one semiconductor substrate, it is possible to excite the Raman laser with the above-mentioned incident power density without using a special external optical system such as a microscope.

The use of a l1l-V group mixed crystal as a semiconductor laser has the advantage that C1 can be changed appropriately and the resonance conditions described above can be approached. FIG. 3 shows an example of an integrated Raman laser using a semiconductor laser as an incident light source, in which a is a cross-sectional view and b is a top view. For example, the substrate crystal 6 is Gap or Qa As, and the substrate crystal 2 is Ga AfAS-based or L1 made by heteroepitaxial growth.
These are semiconductor lasers such as 111Qa761S type, summer n Ga PAs type, etc. 7 is a light guide path for incident light, 8 is a light guide path for near-infrared output, and 5 is a light guide path for far-infrared light. If the semiconductor is sufficiently transparent to far-infrared light, the semiconductor itself may be used. If it is not sufficiently transparent, cut out the part 5 using ion milling, sputter etching, plasma etching, etc. and use the space as a light guide, or fill the space with a far-infrared transparent material such as mylar.

The phonon frequency is ω, (LO phonon) ω□.

(To phonon), and the Raman output frequency is ω. Then ωj−ω. -C8 is equal to the number of captured phonons;
There are cases of polariton mode in which phonons and far-infrared electromagnetic waves are combined, and various modes can be oscillated depending on the crystal orientation. For example, the direction of the end face of GaP is <10
When it is close to the <110> direction, a polariton mode oscillates due to forward scattering, and when it is close to the <110> direction, a polariton mode oscillates, and a TO phonon oscillates due to backscatter. When the polariton mode oscillates, basically the incident light intensity or output ω. The far-infrared light output intensity is determined by the ratio of ω, /ω, compared to the intensity of ω.

In other words, near-infrared light ω. and far-infrared light ω can be obtained. Furthermore, by changing θ, the frequency ω3 of the far-infrared light can be continuously changed as shown in FIG.

The light guide path 7.8 for near-infrared light may be made of the substrate material itself, but in order to prevent the spread of power and to input or extract it at high density, a layer with a slightly higher refractive index is grown by heteroepitaxial growth.
If a so-called semiconductor focusing light guide path is used, the threshold value of the incident light power will be lowered. Of course, the semiconductor laser 2 and the light guide path 7, and the Raman oscillation section 1 and the light guide path 8 may be configured as twin guides. In the illustrated example, the Raman part is made of the same material as the substrate, but of course it may be grown by heteroepitaxial growth. Reflective films may be deposited on Ml and M2, but the deposition depth may be uneven, or
If vertical deposition is difficult, use a grating structure as shown in Figure 4 and select the thickness t1 of the notches and the period λ appropriately.
. The reflectance can be maximized. Adjustment may be made by placing a filling material with a refractive index n in the groove.

The angle θ formed by the plane of incidence is the necessary θ since θ1n necessary to obtain the output of the desired wavelength ω8 can be found from Fig. 10.
What is necessary is to determine θ such that in. It is preferable to select θ2 so that it is perpendicular to the output direction of ω8. The direction of ω is ω1, ω. , the wave number of ω8 is 1, kOlq should satisfy the phase matching condition by forming a triangle as shown in Fig. 5, and q can be obtained from Fig. 10, etc., and 1, k
O is ω1, ω. refractive index n1no to ki-ni
ω; Determined from the relationship: /C, ko-noω, /C.

In all of the above, a resonator is constructed to increase the reflectance for near-infrared light ωI, and oscillation is caused by Raman material 1.
If the absorption coefficient of is smaller for ω3 than for ω1, a resonator can be constructed for ω3. The advantage of this is that when a resonator is constructed using a grating, the wavelength of ω6 is 10 to 100 μm, so the grating can be easily manufactured with extremely high processing accuracy. An example is shown in FIG. The semiconductor laser section in the figure is not shown because it is the same as in FIG. 3. The grating in Figures 4 and 6 is D.
A corrugated or embedded type like an FB laser may be used, but as the length of the grating portion increases, the absorption loss increases, so it is desirable to make the cuts deep as shown in the figure to reduce the number of repetitions.

Next, as the Raman part 1, there is not only a --like single crystal layer, but also a pn junction depletion layer or a pin junction depletion layer as shown in Figure 7.
The use of layers reduces absorption loss due to free carriers, increases the incident wave electromagnetic field strength due to the waveguide effect, and lowers the Raman oscillation threshold. Furthermore, when a reverse voltage is applied to the junction to create an avalanche, high-speed electrons collide with optical phonons, helping to excite the phonons. Especially the 6th
When a resonator is configured and oscillated for ω8 as shown in the figure, this direction is effective for lowering vAIll.

In the case of the integrated type, in the above example, the angle of incidence with respect to the arm portion is determined by the angle of the notch, etc., but FIG. 8 shows an example in which the wavelength can be made variable by changing the angle of incidence. 8th
In the figure, surface wave ultrasound is excited by a pair of electrodes 10 and 11, and by changing the frequency, the diffraction angle is changed and the incident angle θ
Control in. 12 is a deposited piezoelectric material. In FIG. 9, a pn junction having a triangular prism shape when viewed from the surface is formed in the middle of the light guide path 7, and changing the thickness of the depletion layer effectively changes the prism refractive index and changes the direction in which the light travels.

[Brief explanation of the drawing]

Fig. 1 is an explanatory diagram of creepman scattering, Fig. 2 is a semiconductor Raman laser using a YAG laser as an excitation light source, and Fig. 3 is an embodiment of the present invention, showing an integrated Raman laser using a semiconductor laser as an incident light source. In one example, <a) is a cross-sectional view, (b)
is a plan view, FIG. 4 is another example of the integrated Raman laser of the present invention, (a) is a cross-sectional view, (b) is a plan view, FIG. 5 is an explanatory diagram of the present invention, and FIG. Other embodiments of the invention,
FIG. 7 is an explanatory diagram of the Flyman oscillator, FIGS. 8 and 9 are still other embodiments of the present invention, and FIG. 10 is a diagram showing the dispersion relationship of the polariton mode of GaP.

Claims (2)

[Claims] (1) A semiconductor Raman oscillation unit formed on the same semiconductor substrate, an injection type semiconductor laser that excites the semiconductor Raman oscillation unit, and a guide that makes the output light of the injection type semiconductor laser enter the semiconductor Raman oscillation unit at high density. A semiconductor Raman laser consisting of an optical path. (2) The semiconductor Raman oscillator is selected from GaP or GaAs, and the injection type semiconductor laser is made of GaAl.
The semiconductor Raman laser according to claim 1, wherein the semiconductor Raman laser is selected from As, InGaAs, and InGaPAs.