A semiconductor laser device and a method for the production of the same

BACKGROUND OF THE INVENTION

1. Field of the invention

The invention relates to a semiconductor laser device made of AlGaAs emitting laser light from its facet, and also relates to a method of manufacturing such a semiconductor laser device.

Second Description of the Related Art

In recent years, AlGaAs-type semiconductor laser devices and other types have widely come into practical use as light sources for optical disk drive units. When semiconductor laser devices are used as a light source for write-once optical disk drive units or rewritable optical disk drive units, they must have high reliability with a high output power of 40 to 50 mW. When used as a light source for optical pumping of solid state laser devices, such as YAG lasers, an output power of 100 mW or more is required.

However, it has been reported that the reliability of semiconductor laser devices, as practicable today and as they can achieve relatively high output laser oscillation, is inevitably proportional to the fourth power of the optical output when comparing components of the same design , In other words, it is extremely difficult to increase the optical output power while maintaining high reliability.

The main reason for degrading semiconductor laser devices in high output power operation is facet degradation. This is because, due to the high light density at the laser light emission facet, this heat is locally generated. The mechanism of this heat generation will be explained with reference to Figs. 10a-10b and 11a-11b.

Figures 10a and 10b are schematic diagrams showing near-surface energy band structures resulting from a surface condition that occurs when a (110) surface of either n- or p-GaAs is readily oxidized. Both in the case of n- and p-GaAs, numerous charge carriers accumulate near the surface in order to produce a so-called "enrichment layer", which is designated by the reference number 1 in these figures.

In general, it is well known that the surface state bends the energy bands near the surface. In addition to the enhancement layer 1 shown in Figs. 10a and 10b, minority carriers may accumulate near the surface, with majority charge carriers spaced from the surface, as shown in Figs. 11a and 11b, resulting in the formation of an inversion layer 2 represents a local reversal of the type of line. Whether an enhancement layer 1 or an inversion layer 2 forms depends on the height relationship between the surface state and the Fermi level of the semiconductor. Both n- and p-GaAs form an enrichment layer.

The electrons and positive holes trapped in the surface states Es are released after a short relaxation time, and this energy is released as heat. Then, electrons and positive holes are again trapped in the surface state, which becomes an unoccupied state, and the above process is repeated, so that heat is constantly released.

As the above process repeats, the heat released from the surface state concentrates on the facets of the semiconductor, and this heat narrows the width of the forbidden band in the energy bands. Furthermore, the absorption of light increases the minority carriers, and the heat generation continues to increase beyond the surface state. This process raises the temperature of the semiconductor surface, where it can reach the melting point of the semiconductor, resulting in the destruction of the facet.

In the case of GaAs, an enrichment layer is formed, while in the case of other materials, such as AlGaAs, an inversion layer may be formed. In the latter case, majority carriers are captured in the surface state, and destruction of the facet occurs with the same process as an enhancement layer. In the case of semiconductor laser devices used in the high injection state, generation of heat resulting from the surface state becomes a more serious problem.

As a measure for preventing the facet from being affected by heat generation at the facets as described above, a structure has been proposed in which a window region is formed on the facet face. By this method, a transparent region for laser light is generated on the facet surface, whereby light absorption in the facets is eliminated and heat generation caused by light absorption is suppressed. However, the process used to fabricate a window of such structure is extremely complicated, and the difficulty of fabricating an optical waveguide in the vicinity of the facets becomes a problem.

The method proposed on pages 163 to 266 of Extended Abstracts of the 20th Conference on Solid State Devices and Materials, Tokyo (1988) attempts to improve the surface properties in an MIS structure using GaAs. In this method, an oxide film formed on a GaAs surface in air may be removed to deposit GaS instead of treating it with an aqueous solution of (NH 4) 2 S. The generation of a GaS film makes it possible to lower the surface state caused by the oxide film.

However, optical components such as semiconductor laser devices made of AlGaAs have not been attempted to improve the facets by the above-mentioned surface treatment. This is because aluminum is an extremely active material and its oxide film is stable, so that removal of the oxide film was not considered possible.

US-A-3 849 738 discloses a semiconductor laser device having a multilayered structure formed on a semiconductor substrate and having an active layer of AlGaAs for laser vibration; a facet-made ZnS layer and an Al 2 O 3 protective overcoat prepared thereon. having.

The ZnS layer and the Al 2 O 3 layer together form an antireflection coating which also provides physical protection of the facet. The refractive index and thickness of the ZnS layer are chosen to compensate for the low refractive index of the Al 2 O 3 layer.

"Extended abstracts of the 20th (1988) International Conference on Solid State Devices and Materials, pp. 263-266 addresses the problem of improving the surface properties of GaAs. There, the effect of treating the surfaces of GaAs with (NH 4) 2 S solutions is examined. It is concluded that the treatment provides for a thin (one atomic layer) coating of sulfur on the GaAs surface, resulting in a passivation of the surface. The treatment also removes oxides or as from the surface.

According to a first aspect of the invention, there is provided a semiconductor laser device configured to emit laser light from a facet and provided with:

a multilayered structure formed on a semiconductor substrate and having an active layer of AlGaAs for laser vibration; a sulfur-containing film formed on this facet; and a protective film formed on the sulfur-containing film, wherein the sulfur-containing film was prepared by treating a split or etched surface of the multi-layered structure with a sulfur-containing solution selected from undiluted (NH 4) 2. S, aqueous (NH 4) 2 S, neat (NH 4) 2 Sx, and aqueous (NH 4) 2 Sx solutions. The term undiluted solution is here to be understood as a commercial standard solution.

In a preferred embodiment, the protective film is made of an oxygen-free material.

In a preferred embodiment, the oxygen-free material is selected from the group consisting of Si 3 N 4, AlN, C, MgF 2, CaF 2, NaF, ZnS and ZnSe.

According to a second aspect of the invention there is provided a method of making a semiconductor laser device as set forth above comprising treating the cleaved or etched surface of the multilayer structure with a sulfur-containing solution to form the sulfur on the cleaved or etched surface of the multilayer structure produce film containing; and preparing the protective film on this sulfur-containing film.

In a preferred embodiment, the cleaved or etched area is treated for a time of 1.5 / x seconds or more when the sulfur concentration of the solution is x mol / l.

In a preferred embodiment, the protective film is made of an oxygen-free material.

In a preferred embodiment, the oxygen-free material is selected from the group consisting of Si 3 N 4, AlN, C, MgF 2, CaF 2, NaF, ZnS and ZnSe.

In a preferred embodiment, the protective film is made of an oxygen-free material selected from the group consisting of Si₃N₄, AlN, C, MgF₂, CaF₂, NaF, ZnS and ZnSe.

In a preferred embodiment, the facet is treated for a time of 1.5 / x seconds or more when the sulfur concentration of the solution is x mol / l.

In a preferred embodiment, the protective film is formed by electron beam evaporation at a rate of 10 Å / s or less.

Thus, the invention described herein makes possible the achievement of the following objects: (1) to provide a semiconductor laser device of AlGaAs having a structure capable of suppressing a surface state caused by an oxide film on the laser light emission facet, so that itself under high output conditions, facet destruction is not easy; (2) to provide an AlGaAs semiconductor laser device in which a sulfur-containing film is formed on the facets so that the surface state of the facets can be greatly lowered, therefore, heat generation in the laser light emitting facet to which it is subjected Surface recombination comes effectively suppressed, thereby achieving improved output performance with high reliability; (3) providing an AlGaAs semiconductor laser device in which a film of oxygen-free materials as a protective film is formed on the facets after surface treatment of the surface of the resonator facet is performed so that oxidation does not occur at the interface between the protective film and the semiconductor crystal oxygen contained in the protective film can take place, for which reason non-radiative recombination centers can be suppressed on the facet surface, which, in turn, suppresses deterioration of the facets, thereby achieving stable laser oscillation at a high value of the output power for a long period of time; and (4) to provide a method of manufacturing such a semiconductor laser device made of AlGaAs having excellent characteristics.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention may be better understood by those skilled in the art with reference to the accompanying drawings, and its numerous objects and advantages will be apparent thereto.

Fig. 1 is a perspective view showing the front facet region of a semiconductor laser device according to the present invention.

FIGS. 2a-2c are perspective views illustrating the fabrication of the semiconductor laser device of FIG. 1 illustrate.

Fig. 3 is a graph showing the injection current-light output characteristic of the semiconductor laser device of Fig. 1;

Fig. 4 is a graph showing the facet degradation output for a conventional untreated facet semiconductor laser device and various faceted semiconductor laser device devices of the present invention.

Fig. 5 is a graph showing the relationship between the Auger signal intensity for oxygen and the immersion time.

Fig. 6 is a graph showing the relationship between the immersion time until the Auger signal intensity for oxygen reaches the background level and the sulfur concentration of the solution.

Fig. 7 is a perspective view showing still another semiconductor laser device according to the present invention.

Fig. 8 is a graph showing the relationship between the signal intensity and the bonding energy of specific elements at the interface between the protective film and the semiconductor crystal in the semiconductor laser devices of Fig. 7 and a comparative example.

Figs. 9a-9c are perspective views illustrating another method of manufacturing the semiconductor laser device of Fig. 7; Figs.

Figures 10a and 10b are schematic energy band diagrams showing the formation of a carrier accumulation layer in the surface region of n- and p-GaAs, respectively.

Figures 11a and 11b are schematic energy band diagrams showing the formation of an inversion layer in the surface region of an n-type and p-type semiconductor, respectively.

Fig. 12 is a graph showing the relationship between the injection current-light output characteristic of various semiconductor laser devices in which a protective film has been formed at a predetermined rate on the laser light emitting facet.

Fig. 13 is a graph showing the relationship between the maximum light output and the generation rate of a protective film in the semiconductor laser devices of Fig. 12;

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The semiconductor laser device of this example has the same structure as a conventional semiconductor laser device emitting laser light from its facet, except that a film mainly composed of sulfur is formed on the facets and a protective film on the film consisting mainly of sulfur is trained. Therefore, the following explanation mainly covers the manufacturing process for the facets and their structure.

Fig. 1 is a perspective view showing the front facet region of the semiconductor laser device of this example. The manufacturing process for this semiconductor laser device will be explained below.

As shown in FIG. 2 a, a semiconductor substrate 11 was grown with a multilayered structure 12 including an active layer 16 of AlGaAs for laser vibration, and on the upper surface of the multilayered structure 12 and the back surface of the semiconductor substrate 11 were formed ohmic electrodes 13 or 14 produced.

Next, the thus obtained wafer was cleaved by a known cleavage method to obtain the predetermined resonator length, resulting in rods 15 shown in FIG. 2b. Each of the bars 15 has a plurality of resonator units arranged in the direction at right angles to the resonator direction.

After splitting, the bars 15 were immediately submerged in a 10% (NH 4) 2 S solution 20 in which they were left at room temperature for 3 minutes, so that the nip of each of these bars 15 was treated so that on the facet a sulfur-containing film 18 was formed.

After completion of the treatment, the bars 15 were washed with water and dried. Then, on the front facet, a reflection film 17 having a reflectivity of 4% of Al & sub2; O & sub3; and on the back facet, a reflection film having a reflectivity of 95% was formed by a multilayer structure of Al & sub5; and amorphous silicon. Thereafter, the bars 15 were again split to provide semiconductor laser devices as shown in FIG.

Fig. 3 is a graph showing the injection current-light output characteristic of the semiconductor laser device obtained in the above manner (curve A). For comparison, the curve B in this figure shows the injection current-light output characteristic of a conventional semiconductor laser device in which the reflection film is formed directly on the cleavage surface. In this conventional semiconductor laser device, degradation of the facets occurs at an optical output power of about 200 mW, deterioration of device characteristics, whereas in the semiconductor laser device of this example, an optical output power of about 600 mW can be achieved and moreover at this point none Destruction of the facets occurs.

In measurements of actual surface state densities carried out to show that the above findings are based on a lowering of the surface state caused by the oxide film, it has been clarified that the surface state density in the semiconductor laser device of this example is about 10 & sup8; to 10? cm &² · eV &¹. When a semiconductor laser device made of GaAs was treated in the same manner, the smallest value obtained was 10¹¹ to 10¹² cm &² · eV &¹, and no effective lowering of the surface state as mentioned above was recognized.

Therefore, in contrast to the previously assumed, it is considered that the large lowering effect on the surface state density is due to the presence of Al.

In the above examples, after the facet treatment, an Al 2 O 3 film was prepared by electron beam evaporation as a protective film on the laser light emitting facet. The following describes the effect that the rate of forming a protective film has on the light output characteristic of semiconductor laser devices.

First, various semiconductor laser devices in which a protective film was formed on the laser light emitting facet at a rate of 5 Å / sec, 10 Å / sec, 12 Å / sec, and 15 Å / sec were used in the same manner as in Figs made the above examples. Fig. 12 shows the injection current-light output characteristic of the semiconductor laser devices thus obtained. In this figure, the line A corresponds to the case where the rate of producing a protective film was 5 Å / sec, the line B is 10 Å / sec, the line C is 12 Å / sec and the line D is 15 Å / sec ,

Fig. 13 is a graph showing the relationship between the maximum optical output power and the production rate of a protective film in the above-mentioned semiconductor laser devices. For comparison, in this figure, a broken line represents the maximum output obtained from a conventional untreated facet semiconductor laser device.

As can be seen from Figs. 12 and 13, as the production rate of a protective film increases, the maximum optical output decreases and approaches the value of the conventional untreated facet semiconductor laser device. A possible reason is that when the intensity of the emitted electron beams is increased to increase the manufacturing rate of a protective film, a large number of high energy scattering electrons are generated, and therefore, the film containing sulfur generated on the laser light emitting facet is caused by them Streulektronen is impaired.

Fig. 13 shows that deterioration of the sulfur-containing film significantly affects the maximum optical output when the production rate of a protective film is more than 10 Å / sec. However, when the manufacturing rate of a protective film is 10 Å / sec or less, a maximum optical output power of 540mW or more can be achieved because the sulfur-containing film made on the laser light emitting facet is not significantly deteriorated.

Moreover, the same experiments as mentioned above were carried out with respect to the other facet located on the other side of the semiconductor laser device. From the results obtained, it was found that the output optical characteristics do not significantly depend on the production rate of a protective film to be formed on the other facet from which no laser light is emitted.

Therefore, in order to obtain a larger value of the maximum optical output power, it is preferable that a protective film is formed on the laser light emitting facet at a rate of 10 Å / sec or less.

Although in the above examples, an aqueous (NH 4) 2 S solution was used to treat the cleavage surface, the surface treatment may also be carried out with an undiluted instead of an aqueous solution or with a nonaqueous solution.

The following is an explanation of another example of the invention in which the surface treatment was carried out with an undiluted or an aqueous (NH 4) 2 S solution of 10% or an undiluted or an aqueous (NH 4) 2 Sx solution of 10 % was executed. Here, (NH 4) 2 S x denotes a mixture of (NH 4) 2 S and (NH 4) 2 S 2, where x, which indicates the proportion of sulfur, is a real number of 1 to 2, inclusive, is.

Fig. 4 is a graph showing the output power (mW) of semiconductor laser devices of this example in the destruction of the facet, which components undergo facet treatment using the undiluted or aqueous (NH 4) 2 S solution of Fig. 10 % or the undiluted or aqueous (NH 4) 2 Sx solution of 10%, and for semiconductor laser devices which were not facet-treated as a comparative example.

As can be seen from the graph of Fig. 4, the output powers of the semiconductor laser devices in facet destruction were faceted using an undiluted or aqueous (NH 4) 2 S solution of 10% or undiluted or 10% aqueous (NH 4) 2 Sx solution, significantly higher than those of the semiconductor laser devices of the comparative example.

The thicknesses of AlGaAs active layers in the semiconductor laser devices of this example and the semiconductor laser devices of the comparative example were larger than those of the AlGaAs active layer in the semiconductor laser device, the results of which are shown in FIG. Therefore, in the semiconductor laser devices of this example and the comparative example, the light density is higher and the output power in the destruction of the facet is lower than that of the semiconductor laser device whose test results are shown in FIG.

In this example, a facet treatment of the semiconductor laser device with a neat or an aqueous (NH 4) 2 S solution of 10% or an undiluted or an aqueous (NH 4) 2 Sx solution of 10% was carried out The same effect can be obtained when a solution of an alkali metal compound with sulfur is used instead of a (NH 4) 2 S or (NH 4) 2 Sx solution. In other words, there is no limitation on the material usable for treating the cleavage surface as long as a sulfur-containing film is formed on the facets of the semiconductor laser device.

In order to improve the output of a semiconductor laser device in destroying a facet in this way, it is important that sulfur be contained in the solution used to treat the facets.

The following is an explanation of an experiment carried out to determine the relationship between the sulfur concentration of the sulfur-containing solution and the time required to treat the facets with this solution and the results.

The experiment was carried out by immersing AlGaAs crystals in (NH 4) 2 S solutions having respective sulfur concentrations of 0.015 mol / liter, 0.15 mol / liter and 0.5 mol / liter, and then the AlGaAs crystal face was evaluated by Auger electron spectroscopy. The temperature of the solutions during the immersion process was about room temperature.

Fig. 5 is a graph showing the relationship between the Auger signal intensity of the oxygen in the AlGaAs crystal surface determined by Auger electron spectroscopy and the immersion time. In the graph, line D indicates the relationship in the case where the sulfur concentration in the solution used was 0.015 mol / liter, the line E is 0.15 mol / liter, and the line F is 0.5 mol / liter.

As indicated by the graph, regardless of the concentration used, the Auger signal intensity of oxygen becomes lower, and it approaches the signal intensity of the background as the immersion time becomes longer.

Fig. 6 is a double logarithmic graph showing the relationship (line G) between the immersion time until the Auger signal intensity of oxygen reached the background level as determined by the above experiment (time required to remove oxygen) and the sulfur concentration the solution shows. The graph also shows the relationship (dashed line) between the immersion time until the Auger signal intensity of oxygen reached the background level and the sulfur concentration of the solution when an aqueous Na 2 S solution was used as the treatment solution, which is a solution of an alkali metal compound with sulfur.

As can be seen from Fig. 6, the immersion time until the Auger signal intensity of oxygen reached the background level and the sulfur concentration of the solution were almost inversely proportional. This indicates that oxygen existing as a component of the oxide film or the like on the AlGaAs crystal surface is removed from the AlGaAs crystal surface by reaction with sulfur in the solution, as indicated in the following reaction equation I:

Oxide + Sulfur → Sulfide + Oxygen (I)

The rate of reaction to the right within the above reaction is given by the following equation II:

- d [oxide] / dt = k · [oxide] · [S] (II)

where [oxide] is the oxide concentration at the AlGaAs crystal face, [S] is the sulfur concentration of the solution, K is the rate constant, and t is the time elapsed from the start of the immersion. The left side of Equation II is the time differential of [Oxide] and it indicates the rate at a particular time when the oxide on the AlGaAs crystal face is removed therefrom.

By solving the above differential equation II, [oxide] can be represented as a function of time, as indicated by the following equation III:

[Oxide] = [oxide] o · exp (-K [S] t) (III)

wherein [oxide] & sub0; the value of [oxide] at the start of immersion is (ie, t = 0). Since the total amount of sulfur in the solution is much larger than the total amount of oxide on the AlGaAs crystal face, [S] can be handled as a constant regardless of time when the above differential equation II is solved.

From the equation III, the time tc required for [oxide] to be 1% of [oxide] & sub0; decreases according to the following equation IV:

tc = (ln100) / K · [S]) (IV)

Equation IV indicates that the time required to remove 99% of the oxygen present as a component of the oxide film or the like on the AlGaAs facet by reaction with the sulfur in the solution from that surface is inversely proportional to the sulfur concentration in the solution is. This agrees with the experimental result shown in the graph of FIG. 6.

By fitting the above equation IV to the relationship (line G) shown in the graph of Fig. 6, the following equation V is obtained.

tc = 1.5 / [S] (V)

wherein the unit for tc is seconds, the unit is for [S] moles / liter, and the unit for 1.5 units is s · moles / liter.

When a solution having a sulfur concentration of 0.015 mol / liter is used, the above equation V gives tc = 100 s. Therefore, when a solution having a sulfur concentration of 0.015 mol / liter is used, the oxide film on the AlGaAs facet facet can not be completely removed when the facets are treated for less than 100 seconds. D. That is, for removing almost all of the oxide film on the AlGaAs crystal face, the facets must be treated for a period of time equal to or longer than the time tc indicated by the above equation Vc.

Using the above test results and observations in this manner, the immersion time required to remove almost all of the oxide film on the AlGaAs crystal surface was determined.

Based on these results, by treating the facets for a suitable period of time, depending on the sulfur concentration of the solution used, insufficient removal of the oxide film on the facets due to insufficient treatment time and wasted time by treating the surfaces for an unnecessary long time could be avoided.

As indicated by the graph of Fig. 6, using Na 2 S as a facet treating solution, the oxide film on the AlGaAs crystal face was removed as well as when a (NH 4) 2 S solution was used. However, semiconductor laser devices subjected to a facet treatment using a (NH 4) 2 S or (NH 4) 2 Sx solution have shown excellent reliability as compared with semiconductor laser devices subjected to facet treatment using a Na & sub2; ; S solution have been subjected. Of sulfur-containing solutions, (NH 4) 2 S and (NH 4) 2 Sx solutions are excellent facet treating solutions from the standpoint of the reliability of the semiconductor laser device.

On the facets of a semiconductor laser device prepared by forming an Al 2 O 3 film on the facets as a protective film after the surface of the facets has been treated with a sulfur-containing solution, oxygen contained in the protective film oxidizes on the crystal surface of the semiconductor. when the semiconductor laser device is used for a long period of time, and prevents this the semiconductor laser device generates stable laser oscillation at a high output power for a long period of time. An example by which this problem is solved will be explained below.

Fig. 7 is a perspective view showing the front surface of the semiconductor laser device of this example. In the semiconductor laser device of this figure, a multilayered structure 12 including an active layer 16 of AlGaAs is formed on a semiconductor substrate 11, and on the upper surface of the multilayered structure 12 and the backside of the semiconductor substrate 11 are provided ohmic electrodes 13 and 14, respectively. The facets are covered with a protective film 22 of an oxygen-free material. Between the semiconductor crystal and the protective film 22, a sulfur-containing thin film 18 having a thickness of only a few molecules is formed.

The following is an explanation of the method used to manufacture the semiconductor laser device shown in FIG.

First, after a multilayered structure 12 having an active layer 16 of AlGaAs and other semiconductor layers was formed on the semiconductor substrate 11, ohmic electrodes 13 and 14 were fabricated on top of the multilayered structure 12 and the backside of the semiconductor substrate 11, respectively.

Next, the thus-obtained wafer was cleaved to produce bars in which a number of resonator units of predetermined resonator length were arranged in a direction perpendicular to the cleavage direction.

The bars were then immersed in a 10% aqueous (NH 4) 2 S solution to treat the surface of the resonator facets. At this time, the temperature of the aqueous solution was about room temperature. After the surface treatment, the bars were washed in purified water and then dried. This surface treatment removed the oxide film that had formed on the facet surface. At this time, a sulfur-containing thin film was formed on the facet surface from which the oxide film was removed. This film prevents direct contact between the surface of the semiconductor crystal and air, etc., thereby preventing the formation of a natural oxide film on the facets.

Thereafter, on the laser light emission side of the resonator facets, a protective film 22 made of Si.sub.3 N.sub.4 which was free of oxygen was prepared by using a plasma CVD method. The thickness of the protective film 22 was set to λ / (4n) Å, where λ is the oscillation wavelength of the laser light in Å and n is the refractive index of the protective film 22. The reflectance of the laser light emission facet thus produced was 5%. Using a plasma CVD method, a multilayer film of an Al 2 O 3 film and a silicon film was formed on the other facet of the resonator. The reflectivity of this facet was 95%. Then, semiconductor laser devices were manufactured by the usual process.

Hereinafter, the results of an experiment in which a semiconductor laser device of this example is compared with a semiconductor laser device as a comparative example in which the facet surface has been treated with an aqueous (NH 4) 2 S solution and then a protective film of Al 2 O 3; was made on the facet.

First, a comparison of both semiconductor laser devices with respect to the optical output power and the facet degradation was carried out. The results showed that the semiconductor laser device of the comparative example suffered facet destruction when operated at an optical output power of 200 mW. However, in the semiconductor laser device of this example, even when operating with an optical light output power of 400 mW, no facet destruction occurred and stable laser oscillation continued. The same result was obtained when the experiment was carried out after the semiconductor laser device of this example was left standing for a long period of time.

As explained above, in the semiconductor laser device of this example, the facets were covered with a sulfur-containing film, and on this film, a protective film was made of an oxygen-free material. Therefore, oxidation did not occur at the interface between the semiconductor crystal of the semiconductor laser device and the protective film by oxygen contained in the protective film, thereby suppressing the increase of nonradiative recombination centers and accordingly suppressing facet degradation. For this reason, the semiconductor laser device of this example can achieve stable laser oscillation at high output for a long period of time.

Next, after the semiconductor laser devices of this example and the comparative example were in air for six months, they were placed in a high vacuum chamber, the protective film was removed by argon ion sputtering, and the state of chemical bonding at the interface between the semiconductor crystal became and the protective film analyzed by photoelectron spectroscopy. The Fig. 8th shows the relationship between the signal intensity (any unit) obtained in these results and the binding energy (in ev). In Fig. 8, the signals A, B, and C for the level 32 P3 / 2 for As, Ga, and Al, respectively, were obtained at the interface of the semiconductor laser device of this example. The signals D, E, and F were obtained for the level ³P3 / 2 for As, Ga, and Al, respectively, at the interface of the semiconductor laser device of the comparative example. The signals A, B and C in FIG. 8th show no shift in binding energy due to binding to oxygen. However, signals D, E, and F are shifted along the x-axis due to bonding to oxygen. From this, it can be seen that in the semiconductor laser device of the comparative example, oxidation occurred at the interface between the semiconductor crystal and the protective film. This indicates that, in the semiconductor laser device of the comparative example, some of the oxygen contained in the protective film reached the interface between the protective film and the semiconductor crystal, where it bonded to component atoms on the surface of the semiconductor crystal. In the semiconductor laser device of this example, however, the protective film contains no oxygen, so that no oxidation of this kind occurred at the interface between the protective film and the semiconductor crystal.

In this example, an oxygen-free film as a protective film was made only on the laser light emitting facet, but the same film can be made on the other facet.

Next, an explanation will be given of another method of manufacturing the semiconductor laser device of FIG. 7.

First, a multi-layered structure 12 having an active layer of AlGaAs and other semiconductor layers was formed on the semiconductor substrate 11 by the usual method, and then, on the upper surface of the multi-layered structure 12 and the back surface of the semiconductor substrate 11, ohmic electrodes 13 and 14, respectively, were fabricated. as shown in Fig. 9a.

Then, specific portions of the semiconductor substrate 11 were etched by reactive ion milling using chlorine gas to produce mirror planes 19, producing on the substrate 11 portions 25 each having a plurality of resonator units of predetermined resonator length in the direction perpendicular to the resonator direction, as in FIG Fig. 9b shown.

The wafer thus obtained was then dipped in an aqueous (NH 4) 2 S solution 20 to treat the surfaces of the resonator facets, as shown in Fig. 9c. The temperature of the aqueous (NH 4) 2 S solution 20 was at room temperature, and the immersion time was 3 minutes. After completion of the surface treatment, the wafer was washed in pure water and then dried.

Thereafter, on both facets of the resonator, by a plasma CVD method, a protective film 22 of Si & sub3; N & sub4; made (see Fig. 7). The thickness of the protective film 22 was set to λ / (2n) Å, where λ is the oscillation wavelength of the laser light and n is the refractive index of the protective film 22. As a result, the reflectance of both resonator facets was 32%. Then, semiconductor laser devices were manufactured by the usual process.

The semiconductor laser devices in which the facets were prepared by etching in this manner provided the same results as the previous example.

In the above examples, a Si 3 N 4 film is used as the oxygen-free film, but films of other materials such as AlN, C, MgF 2, CaF 2, NaF, ZnS or ZnSe may also be used. It is also possible to use multilayer films in which films of these materials are laminated one on top of the other. Moreover, in the above examples, as a method of producing an oxygen-free protective film, a plasma CVD method is used, but other film-forming methods such as electron beam evaporation and sputtering may be used.

It should be understood that various other modifications will be apparent to and can be readily made by those skilled in the art without departing from the scope of the invention. Accordingly, the scope of the invention should be limited not by the description set forth herein but by the appended claims.