JP2007066947A - Spatial light transmission semiconductor laser device and spatial light transmission system - Google Patents

Abstract

A semiconductor laser device for spatial light transmission capable of high-speed modulation with high output and low power consumption is provided.
An active layer, a guide layer sandwiching the active layer, a semiconductor laser element 220 having a cladding layer sandwiching the active layer and the guide layer formed on a GaAs substrate, and a mold resin portion (223) covering the semiconductor laser element 220 , 224). The active layer of the semiconductor laser element 220 has a quantum well structure composed of a well layer made of InGaAs and a barrier layer sandwiching the well layer. The semiconductor laser element 220 has an oscillation wavelength at 25 ° C. of 868 nm (within a range of 857 nm to 871 nm) in a chip state before resin molding.
[Selection] Figure 4

Description

  The present invention relates to a semiconductor laser device for spatial light transmission and a spatial light transmission system, and more specifically, includes a semiconductor laser device for spatial light transmission used for optical transmission using infrared rays, and a semiconductor laser device for spatial light transmission. The present invention relates to a spatial light transmission system.

  Spatial light transmission using infrared rays typified by the IrDA (Infrared Data Association) standard is now widely used. Data such as an address book, a schedule table, and photographic data can be easily transmitted between a portable terminal and a personal computer each having an optical communication function conforming to the IrDA standard without connecting electric wiring. From such convenience, spatial light transmission using infrared rays is expected to become more widespread in the future.

  In recent years, with an increase in data capacity, the transmission rate has been increasing from 1 Mbps to 4 Mbps and 16 Mbps. In the future, the data capacity will further increase to high-quality photographic data and moving image data, and accordingly, the demand for higher transmission speed is also increasing.

  In general, when the transmission speed is increased, the transmission distance is shortened. However, there is an increasing demand for transmitting a long transmission distance at a higher transmission speed.

  Currently, in a spatial light transmission system compliant with the IrDA standard or the like, a light emitting diode (LED) is used as a light source, and an inexpensive Si photo is used as a light receiving element in order to construct the spatial light transmission system at a low cost. A diode (Photo Diode; PD) is used. On the other hand, as the light source, as the transmission speed increases and the transmission distance increases, it is necessary to increase the light output of the LED used.

  For example, Japanese Patent Laid-Open No. 2001-308376 (Patent Document 1) discloses a high-power infrared LED, and FIG. 13 shows the structure of the high-power infrared LED. As shown in FIG. 13, the high-power infrared LED includes a first n-AlGaAs layer 5, an n-AlGaAs cladding layer 6, a p-AlGaAs active layer 7, and a p-layer on an n-GaAs substrate (not shown). After sequentially laminating the -AlGaAs cladding layer 8 and the first p-AlGaAs layer 9 by the liquid phase epitaxial growth method, the n-GaAs substrate is selectively removed to produce an LED wafer. In this LED, the output is increased by removing the GaAs substrate so that the emitted light is not absorbed by the substrate.

  In addition to this, as an example of increasing the output by inserting a reflective layer between the substrate and the clad layer, JP-A-8-288547 (Patent Document 2) discloses MOCVD (Metal Organic Chemical Vapor). A high-power LED having a crystal grown by a deposition (organic metal vapor phase epitaxy) method is disclosed.

  As described above, higher output of the LED has been achieved, but it is considered that the upper limit of the modulation speed of the LED is about several tens of MHz, and the LED cannot cope with the modulation at higher speed than that. Yes. In addition, in the LED, if the light output is increased for high speed and long distance, a very large driving current is required, and the power consumption becomes enormous, and the reliability decreases accordingly. is there.

  On the other hand, an attempt to use a semiconductor laser element (LD) as a light source for spatial light transmission has recently started. When a semiconductor laser element is used as a light source for spatial light transmission, high-speed modulation is possible, and power consumption for obtaining the same light output can be suppressed to be lower than that of an LED.

  However, when emitting light directly into space using a semiconductor laser element, the feature of the semiconductor laser element is that the spatial coherence is high, and when the laser light enters the eye, the light is condensed on the retina, The light density on the retina becomes very high and there is a risk of eye damage. For this reason, when emitting the light of the semiconductor laser element to the space, install a diffusion plate at the location where the laser beam radiates, or take measures such as molding the semiconductor laser element with a resin mixed with a diffusing agent, It is necessary to reduce the spatial coherency of laser light, increase the image of light on the human retina, and lower the light density on the retina (hereinafter referred to as `` safe for eyes ''). In this sense, it is called “Eye Safe”).

Semiconductor laser devices have begun to be used as light sources for spatial light transmission, but little consideration has been given to what kind of semiconductor laser device is suitable as a light source for spatial light transmission.
JP 2001-308376 A JP-A-8-288547

  SUMMARY OF THE INVENTION An object of the present invention is to provide a semiconductor laser device for spatial light transmission capable of high-speed modulation with high output and low power consumption.

  Furthermore, another object of the present invention is to provide a spatial light transmission system using the above-mentioned semiconductor laser device for spatial light transmission.

In order to achieve the above object, a semiconductor laser device for spatial light transmission according to the present invention comprises:
A semiconductor laser device in which an active layer, a guide layer sandwiching the active layer, and a clad layer sandwiching the active layer and the guide layer are formed on a GaAs substrate;
A mold resin portion covering at least a part of the semiconductor laser element,
The active layer of the semiconductor laser element has a quantum well structure composed of a well layer of InGaAs and a barrier layer sandwiching the well layer,
The semiconductor laser element is characterized in that an oscillation wavelength at 25 ° C. is not less than 857 nm and not more than 871 nm in a chip state before resin molding.

  According to the semiconductor laser device for spatial light transmission configured as described above, the active layer of the semiconductor laser element has a quantum well structure including a well layer made of InGaAs and a barrier layer sandwiching the well layer, thereby achieving high output. Therefore, it is possible to realize a semiconductor laser device for spatial light transmission capable of high-speed modulation with low power consumption. In addition, the semiconductor laser element has an oscillation wavelength at 25 ° C. of not less than 857 nm and not more than 871 nm in a chip state before resin molding. ), It is possible to oscillate at 850 nm or more necessary for securing eye-safe property and 900 nm or less necessary for using inexpensive SiPD, and appropriately comply with the IrDA standard. Therefore, it is possible to realize a semiconductor laser device for spatial light transmission having no practical problem in the range of −25 ° C. to + 85 ° C.

  In one embodiment, the semiconductor laser device for spatial light transmission has an oscillation wavelength of 850 nm or more and 900 nm or less at a use environment temperature in a state where the semiconductor laser element is resin-molded.

  According to the semiconductor laser device for spatial light transmission of the above-described embodiment, the semiconductor laser element can perform high-speed and high-speed modulation within an oscillation wavelength range of 850 nm to 900 nm in a resin-molded state. Therefore, a semiconductor laser device for spatial light transmission that is optimal as a light source for the IrDA standard can be realized.

In one embodiment of the semiconductor laser device for spatial light transmission, the well layer is In _1-x1 Ga _x1 As (0.85 ≦ x1 <1), and the well width d in the quantum well structure is 4 nm ≦ d. It is preferable that ≦ 8 nm.

  According to the semiconductor laser device for spatial light transmission of the above embodiment, when the composition x1 of the well layer Ga is 0.85 ≦ x1 <1, the fluctuation of the oscillation wavelength with respect to the temperature change can be reduced, and the oscillation wavelength is stabilized. To do. In addition, by setting the well width d in the quantum well structure to 4 nm ≦ d ≦ 8 nm, the threshold current is decreased, the efficiency is increased, the output can be increased, and the light emission lifetime τ is decreased to perform high-speed modulation. It becomes possible and noise is further reduced. Therefore, when the Ga composition x1 of the well layer satisfies the conditions of 0.85 ≦ x1 <1 and the well width d of 4 ≦ d ≦ 8 nm, the threshold current, efficiency, light emission lifetime τ, noise, characteristic temperature To of the semiconductor laser device are satisfied. , All of the reliability can be satisfied.

In one embodiment of the semiconductor laser device for spatial light transmission, the well layer is In _1-x1 Ga _x1 As (0.9 ≦ x1 ≦ 0.96), and the well width d in the quantum well structure is 4 nm. It is preferable that ≦ d ≦ 6 nm.

  According to the semiconductor laser device for spatial light transmission of the above embodiment, the composition x1 of the well layer Ga is 0.9 ≦ x1 ≦ 0.96, so that the fluctuation of the oscillation wavelength with respect to the temperature change can be reduced, and the oscillation wavelength is stabilized. In addition, when the well width d in the quantum well structure is 4 nm ≦ d ≦ 6 nm, the threshold current is reduced, the efficiency is increased, the output can be increased, the emission lifetime τ can be reduced, and the high-speed modulation can be achieved. It becomes possible and noise is further reduced. Therefore, when the Ga composition x1 of the well layer satisfies the conditions of 0.9 ≦ x1 ≦ 0.96 and the well width d of 4 ≦ d ≦ 6 nm, the threshold current, efficiency, light emission lifetime τ, noise, characteristics of the semiconductor laser device are satisfied. All of the temperature To and the reliability can be satisfied at a higher level.

  Moreover, in the semiconductor laser device for spatial light transmission according to one embodiment, the number of wells in the quantum well structure is preferably 4 or less.

  According to the semiconductor laser device for spatial light transmission of the above embodiment, when the number of wells in the quantum well structure is increased, the characteristic temperature To increases and the characteristics stabilize with respect to the temperature. Can be reduced, and high-speed modulation becomes possible. Therefore, in order to satisfy the characteristic temperature To and the light emission lifetime τ which are in a trade-off relationship, the number of wells is preferably 4 or less.

Moreover, the semiconductor laser device for spatial light transmission of one embodiment is
The barrier layer is Ga _x2 In _1-x2 As _1-y P _y ,
The Ga composition x2 and P composition y of the barrier layer are
y ≦ −2.17x2 + 2.5
y ≧ −1.64 × 2 + 1.63
y ≦ −0.27x2 + 0.88
x2 ≧ 0.65
It satisfies the following conditions.

  According to the semiconductor laser device for spatial light transmission of the above embodiment, a GaInAsP barrier layer effective for the InGaAs well layer can be provided, high output, low power consumption, high reliability, and high-speed modulation are possible. .

In one embodiment of the semiconductor laser device for spatial light transmission, the barrier layer may be Al _x3 Ga _1-x3 As (0.1 ≦ x3 ≦ 0.3).

  According to the semiconductor laser device for spatial light transmission of the above embodiment, an AlGaAs barrier layer effective for forming a quantum well structure can be provided for the above InGaAs well layer, and high output, low power consumption, and high speed can be provided. Modulation is possible. Here, the barrier layer refers to a layer that forms a quantum well structure and sandwiches the well layer, and is distinguished from a guide layer described later.

Also, if the semiconductor laser element is resin-molded and used as a spatial light transmission laser module, heat dissipation is poor and wavelength fluctuations increase, so it is necessary to consider a wide range of temperature fluctuations, compared to ordinary semiconductor laser elements, The wavelength and characteristics are more stable with respect to temperature, and it is necessary to ensure reliability even at high temperature operation.
In addition, when an AlGaAs barrier layer is used, it is more preferable that the Al composition x3 of the barrier layer is smaller within the range of 0.1 ≦ x3 ≦ 0.3 (the larger the Ga composition), the lower the operating voltage Vop becomes stable.

  In addition, when the AlGaAs barrier layer is used, the characteristic temperature To is improved as compared with the case where the GaInAsP barrier layer is used. This is presumably because the difference in the conduction band between the well layer and the barrier layer can be increased by using the AlGaAs barrier layer. In particular, the effect is large at Al composition of 0.15 or more.

In one embodiment of the semiconductor laser device for spatial light transmission, the guide layer may be Al _x4 Ga _1-x4 As (x4 ≧ 0.3).

  According to the semiconductor laser device for spatial light transmission of the above-described embodiment, when AlGaAs is used as the guide layer, the efficiency is improved and the operating voltage Vop is also lowered to increase the output. Further, the AlGaAs guide layer improves the characteristic temperature To, operates stably at a high temperature, and has an effect of improving the characteristic temperature To especially when the Al composition of the AlGaAs guide layer is 0.3 or more.

  A spatial light transmission system according to the present invention includes any one of the above-described semiconductor laser devices for spatial light transmission.

  According to the spatial light transmission system configured as described above, high-speed spatial light transmission exceeding several tens to a hundred Mbps can be realized by providing the semiconductor laser device for spatial light transmission. Therefore, high-speed wireless data transmission can be performed between each device by mounting the above-mentioned semiconductor laser device for spatial light transmission in various devices such as portable terminals, personal servers, personal computers, and printers. it can. Conventionally, when transmission is performed with a semiconductor laser device conforming to the IrDA standard, since the data capacity of high-quality image data, moving image data, etc. is large, it takes 10 seconds to several tens of seconds or more, and the usability is not necessarily good. . On the other hand, in the spatial light transmission system of the present invention, by using the above-described semiconductor laser device for spatial light transmission, data can be transmitted at high speed, so that data can be sent instantaneously, Data can be easily transmitted at high frequency, and convenience is greatly improved.

  As is clear from the above, according to the semiconductor laser device for spatial light transmission of the present invention, it is possible to increase the output as compared with the LED, and high-speed transmission exceeding several tens of MHz and high-speed transmission over a longer distance are possible. It is possible to provide a light source for spatial light transmission with low power consumption. Further, for example, the oscillation wavelength in the use environment temperature range of −25 ° C. to + 85 ° C. can be set to 850 nm or more and 900 nm or less. A photodiode can be used.

  In addition, according to the spatial light transmission system of the present invention, by using the semiconductor laser device for spatial light transmission, high-speed transmission and high-speed spatial light transmission over a long distance can be achieved with low power consumption. A system can be realized.

  Before describing the embodiments of the semiconductor laser device for spatial light transmission and the spatial light transmission system according to the present invention, the present invention will be described based on the results of experiments conducted by the present inventors.

In order to use a semiconductor laser element for spatial light transmission, it is first necessary to consider an optimum oscillation wavelength. In order to perform high-speed spatial light transmission, the following (1) to (3) are required as semiconductor laser elements.
(1) High output (ie low threshold current, high efficiency, high reliability)
(2) Light emission lifetime τ is small
(3) Low noise
(4) Low threshold current, high efficiency, low operating voltage
(5) Stable characteristics such as wavelength λ, threshold current Ith, etc. in the operating environment temperature range

[About oscillation wavelength]
When a semiconductor laser element is used as a light source for spatial light transmission instead of a conventional LED, it is necessary to ensure safety when laser light enters the eyes. The shorter the wavelength of the laser light, the smaller the size of the light spot focused on the eyeball, that is, the light density increases, and the damage to the eyes increases. Therefore, from the viewpoint of eye-safety, it is desirable that the oscillation wavelength of the semiconductor laser element is longer than 850 nm.

  In addition, as a detector for spatial light transmission, it is desirable to use an inexpensive Si photodiode conventionally used in order to reduce the cost of the system. Since the sensitivity of this Si photodiode begins to decrease as the wavelength increases from around 900 nm, it is desirable that the oscillation wavelength of the semiconductor laser element be 900 nm or less within the operating environment temperature range. Even when an InP photodiode is used as the detector, the sensitivity decreases as the wavelength increases from around 900 nm. Therefore, the oscillation wavelength of the semiconductor laser element must be 900 nm or less within the operating environment temperature range.

  As the semiconductor laser device for spatial light transmission, the present inventor relates to the use in a temperature range of, for example, −25 ° C. to 85 ° C. (referred to as the use environment temperature in distinction from the temperature of the semiconductor laser element itself). A ~ 900 nm band semiconductor laser device was fabricated, and the dependence of the oscillation wavelength on the operating environment temperature was examined by experiment.

FIG. 10 shows a graph showing the dependence of the oscillation wavelength on the operating environment temperature. In this figure, the horizontal axis represents temperature [° C.], and the vertical axis represents oscillation wavelength [nm]. As a result of the experiment, it was found that the oscillation wavelength was increased by 5 to 15 nm by molding the semiconductor laser element. In the figure, black squares ◆ and black circles ● are data points without molding, and white squares ◇ and white circles ○ are obtained by molding them, respectively. In order for the oscillation wavelength of the semiconductor laser element after molding to be 850 nm to 900 nm in the temperature range of −25 ° C. to + 85 ° C., the semiconductor laser element is manufactured in consideration of oscillation wavelength fluctuations before and after molding.
Specifically, the oscillation wavelength of the unmolded semiconductor laser element indicated by the black circle ● at −25 ° C. is set to 845 nm, and the oscillation wavelength of the semiconductor laser element indicated by the black square ◆ at + 85 ° C. is It is set to 885 nm. Thus, the oscillation wavelength of the semiconductor laser element after molding indicated by a white circle ◯ at −25 ° C. is 850 nm, and the oscillation wavelength of the semiconductor laser element indicated by a white square ◇ at + 85 ° C. is 900 nm.
Note that the variation in black circles ● and white circles at + 85 ° C. and the variation in black squares ◆ and white squares ◇ at −25 ° C. are +85 depending on the well / barrier configuration (well thickness, barrier composition, well composition, etc.). It shows that the fluctuation range of the wavelength changes as the temperature changes even at the same wavelength at ℃ and -25 ℃.

  As is apparent from FIG. 10, when the oscillation wavelength rises by 1 ° C., the wavelength of about 0.24 nm to 0.29 nm becomes longer, and the difference between the wavelength at 85 ° C. and the wavelength at −25 ° C. is about 30 nm. .

  Further, the inventor molds a semiconductor laser element with a resin mixed with a diffusing agent to make it eye-safe, produces an eye-safe semiconductor laser device in which a mold size, a molding material, and a molding method are changed, and uses an oscillation wavelength. The dependence on temperature was investigated experimentally. As a result, heat dissipation becomes worse by molding, and depending on the mold size, molding material, and molding method, the wavelength at the time of driving the semiconductor laser element may be 5 nm to 15 nm longer than that when not molding as described above. I understood. For example, if the size of the module is changed, the heat dissipation becomes worse if the size is small, so that the wavelength fluctuation increases. On the other hand, if the module is large, the heat dissipation is relatively good and the wavelength fluctuation is also small.

  As described above, in the eye-safe semiconductor laser device, it is necessary to consider the wavelength variation due to the usage pattern in addition to the wavelength variation due to the use environment temperature, and the state of being molded by the module to be used, including the wavelength variation in manufacturing. Therefore, the oscillation wavelength needs to be 850 nm to 900 nm. In order to cope with various applications of semiconductor laser devices for spatial light transmission with the same semiconductor laser element chip, it is necessary to consider a wide range of temperature fluctuations. Therefore, it is necessary to stabilize the oscillation wavelength and characteristics with respect to temperature and to ensure reliability even at high temperature operation.

  Assuming that −25 ° C. to 85 ° C. is assumed as the above operating environment temperature range, the fluctuation of the oscillation wavelength due to the temperature change of the semiconductor laser element itself is about 30 nm, and the wavelength becomes longer by 5 to 15 nm by molding. In the molded semiconductor laser device, it is necessary to consider the wavelength variation of 40 nm. While the optimal oscillation wavelength range is 850 nm to 900 nm, there is only a margin for wavelength fluctuation of 10 nm. Therefore, unless the oscillation wavelength is determined strictly in a semiconductor laser element in a state where it is not molded, there is a possibility that it is outside the optimum oscillation wavelength range when molded.

  Therefore, in order to make the oscillation wavelength of the semiconductor laser element in the molded state 850 nm to 900 nm, when considering how to do it in the original unmolded state, the operating environment temperature is in the range of -25 ° C to 85 ° C. In order to obtain an oscillation wavelength of 850 nm to 900 nm, the wavelength may be 850 nm or more at −25 ° C. and 900 nm or less at 85 ° C. When the temperature of the semiconductor laser element rises by 1 ° C., the wavelength of about 0.24 nm to 0.29 nm becomes longer. Therefore, in order to reduce the wavelength to 900 nm or less at 85 ° C., the oscillation wavelength is 886 nm (= 900 nm−0 at room temperature (25 ° C.)). What is necessary is just to be below .24 nm / ° C. × (85-25) ° C.) to about 883 nm (= 900 nm−0.29 nm / ° C. × (85-25) ° C.). In order to obtain 850 nm or more at −25 ° C., the oscillation wavelength is about 862 nm (= 850 nm + 0.24 nm / ° C. × (25 − (− 25)) ° C.) to 864 nm (= 850 nm + 0.29 nm / at room temperature (25 ° C.). ° C x (25-(-25)) ° C) or more.

  Furthermore, considering that the wavelength of the semiconductor laser element is increased by 5 nm to 15 nm by resin molding, in order to reduce the wavelength to 900 nm or less at 85 ° C., the oscillation wavelength is about 871 nm at room temperature (25 ° C.) without molding. (= 886 nm−15 nm) to 868 nm (= 883 nm−15 nm) or less, and to be 850 nm or more at −25 ° C. Become.

  From the above, it is necessary as a semiconductor laser device for spatial light transmission in the operating environment temperature range of −25 ° C. to 85 ° C. in the molded state by setting the thickness to 857 nm to 871 nm, preferably 859 nm to 868 nm at 25 ° C. in the unmolded state. An oscillation wavelength of 850 nm to 900 nm can be obtained.

  In addition, although the use environment temperature was set to -25 ° C to 85 ° C here, the range of -25 ° C to 85 ° C is practically sufficient, and if the use environment temperature range is narrow, the wavelength range is widened. There is no problem as long as it is at least within the range of the oscillation wavelength.

  The oscillation wavelength of the semiconductor laser element is determined by the band gap energy Eg of the well layer, the well width, and the height of the barrier. What kind of well layer is suitable for use as a semiconductor laser device for spatial light transmission? As a result of studies by the present inventors, when an InGaAs layer is used as a quantum well layer as a semiconductor laser device for spatial light transmission, the fluctuation of the oscillation wavelength with respect to temperature decreases as the Ga composition of the well layer decreases and the well width decreases. I found it small.

  This is probably because InAs is less temperature dependent on the oscillation wavelength than GaAs, the temperature dependence of the oscillation wavelength is smaller as the mixed crystal ratio of the well is closer to InAs. That is, as the Ga composition of the InGaAs well layer is smaller, the temperature dependence of the oscillation wavelength is smaller. Accordingly, the thinner the well width or the higher the barrier, the smaller the Ga composition of the InGaAs well layer for the same wavelength, and therefore the temperature dependence of the oscillation wavelength becomes smaller. From the viewpoint of temperature dependence of the oscillation wavelength, it is desirable that the well layer has a small Ga composition and a small well width.

[Characteristic temperature To]
In addition, in the semiconductor laser device for spatial light transmission, the temperature rise of the chip of the semiconductor laser element is taken into consideration, and the wavelength needs to be stabilized within the use range, and the element characteristics need to be stabilized. The characteristic temperature To is used to express the stability of the element characteristic with respect to the temperature T, and the threshold current Ith is
Ith ∝ exp (T / To)
It is represented by The larger the characteristic temperature To, the more stable the element characteristics against temperature changes. In the semiconductor laser device for spatial light transmission, it is necessary to operate in a wide temperature range as described above, and it is necessary to increase the characteristic temperature To.

  According to experiments conducted by the present inventors, in the case of a semiconductor laser device for spatial light transmission using an InGaAs layer as a quantum well layer, the characteristic temperature To increases as the well width increases or the number of wells increases. The characteristics were stable with respect to temperature. This is presumably because the characteristic temperature To increased because the amount of carriers overflowing from the well layer decreased by increasing the volume of the well layer. From the viewpoint of the characteristic temperature To, it is desirable to increase the well width and the number of wells. However, the characteristic temperature To can be increased by the barrier layer and the guide layer. I need to think about it.

[About high output]
Further, in order to perform high-speed transmission, it is necessary to increase the output of the light source. In order to achieve a high output in a semiconductor laser element, it is necessary to emit more light with less carrier injection and to reduce the absorption of oscillation light. That is, it is desired that the threshold current is low and the efficiency is high. Further, when the amount of injected carriers and the absorption of oscillation light increase, the chip temperature of the semiconductor laser element rises, causing deterioration and making it difficult to increase the output. That is, high reliability is required for the semiconductor laser element.

  The quantum well structure is suitable as the active layer of the semiconductor laser device for spatial light transmission. When the InGaAs layer is used as the quantum well layer, the thinner the well width, the more efficiently the light is converted with less carriers due to the quantum effect. As a result, the threshold current is lowered and the efficiency is improved. However, if the well width is 4 nm or less, gain saturation occurs, so it is desirable that the well width be 4 nm or more.

  If the well width is reduced, the bandgap energy Eg of the well layer needs to be reduced in order to obtain the same oscillation wavelength, but the In composition of the well layer is increased in order to reduce the bandgap energy Eg of the well layer. As the amount of distortion increases, the amount of strain increases, the effective mass of the carrier decreases, and the threshold current decreases, thereby improving the efficiency.

  Therefore, from the viewpoint of higher output, it is desirable to reduce the Ga composition of the InGaAs well layer and reduce the well width.

[About reliability]
The crystallinity becomes important when a semiconductor is crystal-grown, but when impurities such as crystal defects and oxygen are taken into the crystal, these become non-radiative recombination centers.

  If many non-radiative recombination centers are present in the active layer, carriers are consumed as non-radiative recombination in addition to radiative recombination that contributes to oscillation, so that the carriers injected into the semiconductor laser element are converted into light. The efficiency is reduced and a high light output cannot be produced. In addition, these crystal defects and impurities such as oxygen become an absorber for laser light, and if there are many of these absorbers, the absorption of the laser light distributed in the semiconductor laser element increases, and the emitted light output is increased. Further decrease. In addition, in radiative recombination, carrier energy is converted into light energy, whereas non-radiative recombination is converted into thermal energy. Therefore, the temperature of the semiconductor laser device rises due to non-radiative recombination. . In addition, crystal defects that absorb the laser light grow and non-radiative recombination centers increase one after another, causing a problem that the semiconductor laser element deteriorates.

  As described above, when there are non-radiative recombination centers, crystal defects serving as light absorbers, impurities such as oxygen, and the like in the active layer, the active layer itself deteriorates.

  In the semiconductor laser element, a mirror for oscillation is formed by cleavage, and the laser end face is oxidized by being exposed to air at the time of cleavage, and the oxide formed on this end face also becomes an absorber. When such an absorber exists on the emission end face of the semiconductor laser element, the laser-oscillated light is absorbed and the temperature rises. As the temperature rises, the band gap of the active layer near the emission end face becomes smaller, the amount of absorption at the end face portion further increases, and the temperature rises further. Positive feedback that works, and end face destruction (COD) will eventually occur.

  However, the reliability of the semiconductor laser device is improved when the well width of the quantum well active layer is increased or the number of wells is increased. This is because the light intensity (light density) per unit volume decreases as the volume of the well layer increases, so even if there is an absorber with the same density in the well layer or at the emission end face, the absorber This is considered to be because the intensity of light received is reduced and the amount of absorption decreases, resulting in improved reliability.

  Also, according to experiments conducted by the present inventors, high reliability was maintained in the semiconductor laser device for spatial light transmission even when the well width was thin. This is because, when an InGaAs layer is used as a quantum well layer, it does not contain active Al, so that impurities such as oxygen are not easily taken into the well layer. It is considered that the reliability is improved because it is necessary to reduce the Ga composition of the layer (increase the In composition) and increase the In of the well layer to suppress the propagation of crystal defects in the active layer.

[About luminous lifetime τ]
In spatial light transmission, a 1/0 signal is transmitted by turning a light source on / off (or high output / low output). Here, when the light source is turned off from on, the light intensity attenuates over a period of time. This is called the light emission lifetime τ. This attenuation of light can be thought to be due to the recombination of the carriers injected into the active layer over time.

  Carrier recombination includes luminescent recombination contributed by laser oscillation and non-radiative recombination recombining via a non-radiative recombination center existing in the active layer. Considering the light attenuation, it can be said that as the number of non-radiative recombination centers increases, the amount of recombination increases accordingly, so that the light emission lifetime τ is small (the light attenuation is large).

  As described above, from the viewpoint of crystal growth (crystallinity) and LD characteristics, impurities such as non-radiative recombination centers, crystal defects and oxygen absorbers must be reduced as much as possible. In other words, it is generally desirable to increase the light emission lifetime τ as much as possible in manufacturing the semiconductor laser device.

  On the other hand, considering spatial light transmission, the signal becomes 0 only after the light source is turned off after the light source is turned off. Therefore, if the light emission lifetime τ is long, the time until turning it on next must be long. And 0 cannot be separated, causing a communication error. Therefore, considering the modulation at higher speed, the light emission lifetime τ needs to be as small as possible.

  Thus, the characteristics of the semiconductor laser device and the high-speed modulation are in a trade-off relationship with respect to the light emission lifetime τ. Therefore, when a semiconductor laser element is used for spatial light transmission, it is necessary to reduce the emission lifetime τ as much as possible within the range where the characteristics and reliability of the semiconductor laser element are sufficiently maintained.

  Therefore, when the present inventor examined the dependence of the emission lifetime τ on the active layer thickness by experiment, it was found that the emission lifetime τ depends on the thickness of the active layer, and the thickness of the active layer is reduced. It was found that the light emission lifetime τ can be reduced. That is, the quantum well active layer has a shorter emission lifetime τ than the bulk active layer, and the thinner the quantum well active layer, the smaller the emission lifetime τ, enabling high-speed modulation and spatial light. It was found to be suitable as a semiconductor laser device for transmission. In addition, by making the active layer a quantum well structure, the efficiency of carrier-to-light conversion is improved by the quantum effect compared to the bulk active layer, and device characteristics with low threshold, high efficiency, and high reliability are obtained. At the same time, the light emission lifetime τ can be reduced. The thinner the well width, the greater the effect, and the smaller the number of wells, the smaller the emission lifetime τ, and the number of wells is preferably 4 or less.

  The reason why the light emission lifetime τ can be shortened while maintaining such reliability is considered to be that the light emission recombination is efficiently performed with a small amount of injected carriers with little light emission recombination and light absorption.

[Relative noise intensity]
When spatial light transmission is performed at high speed, a communication error may occur unless the signal to noise ratio (SN ratio) is sufficiently large. Causes of noise include noise generated in the usage environment such as the influence of background light, and noise of the semiconductor laser element itself as a light source. When using a semiconductor laser element for spatial light transmission, it is necessary to keep the noise of the semiconductor laser element itself low.

  Therefore, the present inventor manufactured a plurality of semiconductor laser elements by changing the well width and measured the relative noise intensity (RIN: Relative Intensity Noise). As a result, the noise increased when the well width was 4 nm or less and 8 nm or more. It was found that it is not suitable as a light source for spatial light transmission.

  This is presumably because when the well width is 4 nm or less, the sharpness of the interface between the well layer and the barrier layer is greatly affected, and noise is increased. At 8 nm or more, it is considered that a plurality of quantum levels approached within the band, causing noise.

  In addition, when the well width was reduced or the barrier layer was increased to increase the In composition of the well layer, noise was reduced. This is presumably because the insertion of In separates the heavy hole band and the light hole band in the valence band of the well layer and reduces interference between the bands.

[Composition of well layer, well width, number of wells]
From the above, it is preferable to use a quantum well structure in which the InGaAs layer is a well layer as the active layer of the semiconductor laser device for spatial light transmission. The well width of the InGaAs well layer, the composition of the well layer, and the number of wells are complex intertwined with threshold current, efficiency, characteristic temperature To, light emission lifetime τ, reliability, noise, and wavelength variation with respect to temperature. When it is used as a light source, it is very difficult to balance the advantages and disadvantages while considering each effect.

  For example, regarding the well width, the threshold current, the efficiency, the light emission lifetime τ, the characteristic temperature To, and the reliability are in a trade-off relationship, and further, the noise and the composition of the well layer are involved.

  As the well width is reduced, the threshold current decreases, the efficiency increases, and high output is possible. In addition, the light emission lifetime τ can be shortened, and high-speed modulation is possible. However, the characteristic temperature To decreases and the reliability also decreases.

  On the other hand, when the well width is increased, the characteristic temperature To can be increased and the reliability is improved. However, the threshold current increases, the efficiency decreases, and the light emission lifetime τ increases.

  Further, the noise becomes low when the well width is 4 nm to 8 nm, and becomes large even if it is made thinner or thicker.

  Next, looking at the composition of the well layer, if the In composition of the well layer is increased, the threshold current decreases due to the effect of strain, the efficiency increases, and high output can be achieved. Noise is also reduced. Further, by introducing In, dislocation propagation is suppressed and reliability is improved. However, as In is increased, the band gap energy Eg becomes smaller and the wavelength becomes longer. To shorten the wavelength, the well width has to be made thinner. When the thickness becomes thinner than 4 nm, the threshold current is reversed. Increases, decreases efficiency, increases noise, and decreases reliability.

  Considering these, the combination of the well width and the well layer composition is a semiconductor laser only when the well width d satisfies the condition of 4 ≦ d ≦ 8 nm and the well layer Ga composition x satisfies 0.85 ≦ x <1. It is possible to provide a semiconductor laser device for spatial light transmission capable of satisfying all of the threshold current, efficiency, light emission lifetime τ, noise, characteristic temperature To, and reliability of the element and capable of high-speed and high-speed modulation.

  Further, in the range where the well width d is 4 ≦ d ≦ 8 nm and the Ga composition x of the well layer is 0.85 ≦ x <1, the smaller the well composition, the smaller the well composition, the smaller the variation in wavelength. be able to.

[About further limitation of well layer composition and well width]
As the well width is reduced, the threshold current decreases, the efficiency increases, and high output is possible. In addition, the light emission lifetime τ can be shortened and high-speed modulation can be achieved, and the effect is particularly prominent when the well width is 8 nm or less. More preferably, the effect is large when the well width is 6 nm or less.

  Further, the noise becomes low when the well width is 4 nm to 8 nm, and becomes large even if it is made thinner or thicker. More preferably, the noise is most reduced when the well width is 4 nm or more and 6 nm or less.

  On the other hand, when the well width is increased, the characteristic temperature To can be increased and the reliability is improved. If InGaAs is used as the well layer, reliability can be ensured even if the well width is reduced, and a decrease in the characteristic temperature To can be covered by devising the barrier layer and guide layer. It is better to make the well width thinner. However, when the well width becomes thinner than 4 nm, the threshold current increases, efficiency decreases, noise increases, and reliability also decreases. Therefore, the well width is preferably 4 nm or more and 6 nm or less.

  When the In composition of the well layer is increased, the threshold current is reduced due to the strain effect, the efficiency is increased, and the output can be increased. Noise is also reduced. Further, by introducing In, dislocation propagation is suppressed and reliability is improved. Also, the smaller the well layer Ga composition and the thinner well width, the smaller the variation in wavelength. These effects were particularly observed when the In composition was 0.04 or more (Ga composition 0.96 or less).

  However, in order to increase the In composition in the range of the well width of 4 nm to 6 nm, it is necessary to increase the barrier in order to match the wavelength. In this case, the operating voltage increases. Therefore, the In composition is preferably 0.1 or less (Ga composition 0.9 or more) in order to suppress an increase in operating voltage.

  Considering these, the Ga composition of the well layer is more preferably 0.9 or more and 0.96 or less.

  From the above, considering the combination of the well width and the composition of the well layer, the well width d is 4 ≦ d ≦ 6 nm, and the Ga composition x of the well layer is 0.9 ≦ x ≦ 0.96. It is possible to provide a semiconductor laser device for spatial light transmission that satisfies all of the wavelength, the light emission lifetime τ, the noise, the characteristic temperature To, and the reliability, and is capable of high-speed and high-speed modulation.

[About the number of wells]
In addition, the inventor made semiconductor laser devices by changing the number of wells, and conducted an experiment to compare the threshold current, efficiency, characteristic temperature To, reliability, light emission lifetime τ, and noise. As a result, it was found that increasing the number of wells improves the characteristic temperature To and improves the reliability. However, when the number of wells was 5 or more, the threshold current increased, the efficiency decreased, and the noise increased. On the other hand, if the number of wells is reduced, the light emission lifetime τ can be shortened. Considering these comprehensively, the number of wells 1 to 4 is most appropriate.

[Barrier layer of GaInAsP]
In addition, as a result of an experiment conducted by the present inventor (shown in FIG. 11), high reliability could be secured by using GaInAsP as the barrier layer. This is because active Al is not included in the quantum well structure serving as the active layer, so that non-radiative recombination centers and oxygen as absorbers are reduced in the active layer and at the end face, resulting in improved reliability. it is conceivable that.

  In applications that require smaller size and higher output, it is preferable to use a highly reliable GaInAsP barrier layer.

  Further, it was found that when the GaInAsP barrier layer was used, the smaller the Ga composition of the barrier layer, the smaller the variation in wavelength with respect to temperature.

FIG. 11 shows the composition plane of Ga _x3 In _1-x3 As _1-y P _y of the above experimental results. In FIG. 11, the horizontal axis represents the Group III Ga composition, and the vertical axis represents the Group V P composition. The lower right corner of FIG. 11 represents GaAs, and the upper right corner represents GaP. Hereinafter, the composition of the GaInAsP layer will be described with reference to FIG.

[About strain]
The amount of strain ε between GaAs and GaInAsP serving as a substrate is defined by the following equation.
Strain ε = (a (GaInAsP) −a (GaAs)) / a (GaAs)
Where a (GaInAsP) is the lattice constant of GaInAsP and a (GaAs) is the lattice constant of GaAs.

  When a layer having a lattice constant different from that of the substrate is grown, dislocation occurs and the crystallinity is remarkably lowered when the layer exceeds the critical film thickness. Below the critical film thickness, growth is performed so that the lattice constant in the lateral direction matches the lattice constant of the substrate, and growth can be performed without generating dislocations while strain is present in the crystal.

  When a crystal having a composition having a larger lattice constant than GaAs is grown on a GaAs substrate, the crystal is compressed and grown so that the lattice constant in the lateral direction coincides with GaAs. The amount of compressive strain is represented by plus. On the other hand, when a crystal having a smaller lattice constant than GaAs is grown, the crystal grows by being pulled in the lateral direction. The strain amount of this tensile strain is represented by minus.

  In the composition plane of FIG. 11, equal strain lines are drawn by connecting compositions having the same strain amount. A straight line extending from GaAs is a line lattice-matched with GaAs, and an isostriction line is drawn in parallel with the line.

  In GaInAsP, the direction in which the Ga composition is large and the direction in which the P composition is large is the tensile strain, and the direction in which the Ga composition is small and the direction in which the P composition is small is the compressive strain.

  In the active layer of the present invention, since both the well layer (compressive strain) and the barrier layer (compressive strain or tensile strain) can be strained, the total strain amount of the whole quantum well portion (well / barrier) to which strain is applied Need to think about. It is necessary to consider the composition and thickness of the barrier layer so that the total film thickness of the quantum well portion does not exceed the critical film thickness at the total strain. When growing beyond the critical film thickness, dislocation occurs, crystallinity is extremely deteriorated, and device characteristics are remarkably deteriorated.

[About the band gap energy of GaInAsP]
In the composition plane of FIG. 11, iso-Eg lines are drawn by connecting compositions having the same band gap energy Eg. As for the band gap energy Eg, Eg increases in the direction of increasing Ga composition and in the direction of increasing P composition, and Eg decreases in the direction of decreasing Ga composition and in the direction of decreasing P composition.

  In order to confine carriers in the quantum well, it is necessary to provide a band gap energy difference ΔEg between the well layer and the barrier layer. More specifically, it is necessary to provide an energy difference ΔEc between the conduction band Ec and the energy difference ΔEv between the valence band Ev and the well layer. Further, when considering the composition of GaInAsP having the same band gap energy Eg, the energy levels of Ec and Ev are different even if the band gap energy Eg is the same. For semiconductor laser elements, it is desirable to increase ΔEg, particularly ΔEc.

  Thus, by increasing the band gap energy Eg of the barrier layer and increasing ΔEg, ΔEc, carriers (electrons) injected into the well layer are less likely to leak into the barrier layer, and the characteristic temperature To can be increased. . That is, by increasing the Ga composition of the GaInAsP barrier layer and increasing the P composition, the characteristic temperature To increases. However, when the Ga composition and the P composition are increased, the lattice strain increases and there is a concern that the crystallinity is lowered. Considering the same strain amount, if the composition is changed in the direction in which the Ga composition is small and the P composition is large, the band gap energy Eg of the barrier layer is increased and ΔEg can be increased.

  On the other hand, when the composition is changed in the direction in which the Ga composition is small and the P composition is large at the same band gap energy Eg, ΔEc is not changed so much, but ΔEv is considered to increase. If ΔEv becomes too large, holes injected from the p-cladding layer side will be a barrier band of this barrier layer, making it difficult for holes to be injected into the well layer, reducing the efficiency of the semiconductor laser device, It is conceivable that the resistance increases and the operating voltage increases. Further, when the number of wells is increased, the injection of holes into each well layer is likely to be non-uniform, and the efficiency of the semiconductor laser device may be further reduced. On the other hand, if the composition is changed in the direction of increasing the Ga composition and decreasing the P composition, the amount of strain increases.

[Spinodal decomposition temperature]
In addition, the GaInAsP material has a region generally called a miscibility gap (a composition range in which a mixed crystal is not formed by causing phase separation when different crystals are mixed in a compound semiconductor). When performing growth, the mixed crystal becomes unstable and becomes difficult to control as a composition. As a parameter representing the instability of the mixed crystal, there is a spinodal decomposition temperature. When the temperature is lower than the spinodal decomposition temperature, the mixed crystal becomes unstable. In FIG. 11, the spinodal curve is drawn on the GaAsP composition plane. On the composition plane of FIG. 11, the closer to the center of the composition, the higher the spinodal decomposition temperature, and the mixed crystal instability increases. From the viewpoint of the miscibility gap, crystal growth is easier when the composition plane is off the center.

  On the composition plane, the closer the composition is to the center, the more difficult it is to control the composition, the crystallinity is reduced, and the flatness of the surface is lost. If a quantum well cannot be obtained and it is used as a semiconductor laser device for spatial light transmission, the threshold current and the efficiency are remarkably lowered, and the reliability and noise are increased.

  Considering the same strain amount or the same band gap energy Eg, the mixed crystal tends to be unstable in the direction where the Ga composition is small and the P composition is large, and the crystallinity tends to be lowered. Accordingly, a larger Ga composition and a smaller P composition are preferable in terms of crystallinity.

[Composition range of GaAsP barrier layer]
As described above, when GaInAsP is used for the barrier layer of the semiconductor laser device for spatial light transmission, the lattice distortion, band gap energy Eg, ΔEc, ΔEv, and mixed crystal instability (spinodal decomposition) are intricately intertwined. It is very difficult to put on.

  FIG. 12 shows a composition plan view of the GaInAsP barrier layer, and the composition range as the GaInAsP barrier layer will be described with reference to FIG.

The present inventor fabricated a semiconductor laser device for spatial light transmission, and investigated the relationship between the Ga composition x and P composition y of the Ga _x In _1-x As _1-y P _y barrier layer and the element characteristics by experiments. .

As a result of the experiment, when the Ga composition and the P composition are increased, the band gap energy Eg of the GaInAsP barrier layer can be increased, but the tensile strain amount is increased.
y> -2.17x + 2.5
The threshold current, efficiency, characteristic temperature To, noise, and reliability were significantly deteriorated in the region represented by. This is presumably because the total thickness of the active layer exceeded the critical thickness. Therefore,
y ≦ −2.17x + 2.5
It is necessary to use the composition in the region represented by In order to ensure good device characteristics and reliability,
y ≦ −2.05x + 2.34
It is desirable that the composition region represented by More preferably,
y ≦ −2.1x + 2.29
Is more desirable.

When the Ga composition and the P composition are decreased, the band gap energy Eg of the GaInAsP barrier layer is decreased, and the difference in the band gap energy Eg from the well layer cannot be obtained. Also,
y <−1.64x + 1.63
In the composition region represented by ## EQU2 ## the characteristics of the semiconductor laser element such as the threshold current, efficiency, and noise were significantly deteriorated. This is presumably because electrons are no longer confined in the well layer because the difference ΔEc between the well layer and the barrier layer cannot be obtained. Therefore,
y ≧ −1.64x + 1.63
It is necessary to use a composition region represented by In order to ensure good device characteristics,
y ≧ −1.64x + 1.7
It is preferable to use a composition region represented by: More preferably,
y ≧ −1.54x + 1.73
It is most desirable to use a composition region represented by:

In the composition region, when the Ga composition is decreased and the P composition is increased,
y> −0.27x + 0.88
In the composition region represented by the formula, the efficiency was lowered and the operating voltage was raised. This is presumably because the energy difference ΔEv between the valence bands of the well layer and the barrier layer became too large, so that the injected holes were not efficiently injected into the well layer. Therefore,
y ≦ −0.27x + 0.88
It is preferable to use in the area | region represented by these. More preferably,
y ≦ −0.28x + 0.83
It is most desirable to use a composition region represented by:

  Further, in the region of x <0.65, the threshold current, efficiency, noise, and reliability were significantly deteriorated. This is presumably because the mixed crystal became unstable due to spinodal decomposition, and the crystallinity was significantly reduced. Therefore, it is necessary to use a composition region where x ≧ 0.65. It is preferable to use a composition region where x ≧ 0.7.

  In the composition region where the Ga composition is large and the P composition is small, the composition range can be used. However, since the band gap energy Eg becomes small, the energy difference ΔEg with the well layer cannot be made large. It is desirable that x ≦ 0.95.

[Luminescence lifetime τ and reliability]
According to the experiments by the present inventors, it was found that when AlGaAs is used for the barrier layer, the emission lifetime τ decreases as the Al composition of the barrier layer is increased. However, in this case, as the Al composition is increased, the light emission lifetime τ is decreased, but the device characteristics are deteriorated and the reliability is also deteriorated. If the is driven, sufficient reliability cannot be obtained.

  When AlGaAs is used for the barrier layer, Al is an active material and has a very strong bond with oxygen, so that oxygen is easily taken into the barrier layer. Increasing the Al composition increases the oxygen uptake. That is, oxygen is taken into the interface between the well layer and the barrier layer, and becomes a level of non-radiative recombination of carriers confined in the well layer, and the emission lifetime τ is reduced, but the threshold current, efficiency, etc. The device characteristics deteriorate. In addition, oxygen in the barrier layer constituting the active region with high light density is taken in, and this becomes an absorber of laser light, and further, the characteristics of the semiconductor laser device are deteriorated and the reliability is also deteriorated. .

  In addition, the larger the Al composition of the barrier layer, the easier it is to bond with oxygen even at the emission end face, so that COD is more likely to occur.

  Therefore, when AlGaAs is used as the barrier layer, by setting the Al composition to 0.3 or less, a small light emission lifetime τ can be realized while maintaining device characteristics and sufficient reliability.

[About operating voltage Vop]
Further, according to the experiment by the present inventor, it has been found that the operating voltage Vop decreases as the Al composition of the AlGaAs barrier layer is lowered.

  The injected carrier feels that the hetero interface between the well layer and the barrier layer is a barrier, and the smaller the Al composition, the smaller the barrier at this hetero interface, and the easier the carriers are injected into the well layer, thereby reducing the operating voltage Vop. It is guessed. Therefore, from the viewpoint of the operating voltage Vop, it is desirable that the Al composition of the barrier layer be as small as possible.

[Noise]
Further, according to the experiment by the present inventor, when AlGaAs was used for the barrier layer, the noise increased as the Al composition was increased. This is thought to be due to the fact that by increasing the composition of Al, the oxygen taken into the interface between the well layer and the barrier layer increases, the LD characteristics deteriorate, and this causes noise. Therefore, the Al composition of the AlGaAs barrier layer is preferably set to 0.3 or less. In addition, the noise was reduced when the AlGaAs barrier layer was used as compared with the GaInAsP barrier layer. On the other hand, when the Al composition is decreased, the noise is reduced until the Al composition is 0.1, but when the Al composition is smaller than 0.1, the noise is increased. This is probably because the difference in the band gap energy Eg between the well layer and the barrier layer was reduced, and a plurality of quantum levels approached within the band, causing noise.

[Composition range of AlGaAs barrier layer]
The Al composition value of the AlGaAs barrier layer is preferably lowered as much as possible from the viewpoint of reliability and operating voltage Vop. However, if the Al composition is smaller than 0.1, ΔEg between the well layer and the barrier layer is decreased, the threshold current Ith is increased, the efficiency is decreased, and the noise is increased. Therefore, the Al composition is 0.1. It is necessary to be above.

  On the other hand, when the Al composition of the AlGaAs barrier layer is increased, the light emission lifetime τ can be reduced, the characteristic temperature To can be increased, and the wavelength variation due to temperature is small. Further, the noise becomes low when the Al composition is 0.1 or more and 0.3 or less, and the noise becomes large even if it is higher or lower than that.

  From the above, considering the LD characteristics, reliability, operating voltage, wavelength variation, light emission lifetime τ, noise, and characteristic temperature To, the Al composition of the AlGaAs barrier layer is 0.1 or more and 0.3 or less. Is preferred.

[About the effect of the guide layer]
A smaller band gap energy Eg of the barrier layer is preferable because the operating voltage Vop is lowered, but ΔEg and ΔEc with respect to the well layer are reduced, and leakage of carriers from the well layer is increased. Therefore, when the barrier layers on both sides of the quantum well layer are made thick so as to serve also as the guide layer, the confinement of carriers becomes small and the characteristic temperature To decreases. Therefore, the barrier layers on both sides of the quantum well layer are thinned to about 5 nm to 30 nm, respectively, and the guide layer for adjusting the light confinement is provided separately from the barrier layer, so that the LD characteristics such as threshold current Ith and efficiency, It is possible to achieve both a low operating voltage Vop and a high characteristic temperature To.

  As a guide layer, AlGaAs and GaInP were compared. When GaInP is used, the efficiency decreases, the device resistance increases, a large current and voltage are required to produce the same light output, the power consumption increases, the temperature of the device increases, and the reliability increases. The problem of decreasing was found. On the other hand, when AlGaAs is used as the guide layer, the efficiency is improved, the operating voltage Vop is lowered, and the output can be increased. When GaInP is used, the efficiency is lowered and the operating voltage Vop is increased because the valence band discontinuity with the active layer becomes large and becomes a large barrier against holes, so that holes are not easily injected into the well layer. It is thought that it became.

  The characteristic temperature To is improved by the AlGaAs guide layer, and the stable operation is performed at a high temperature. Compared with GaInP, AlGaAs is used as a guide layer, which increases the difference in the conduction band from the active layer, and carriers overflowing from the well layer to the barrier layer do not leak into the guide layer and enter the quantum well. It is thought to stay.

  Further, when the Al composition of the AlGaAs guide layer is 0.3 or more, it is particularly effective for improving the characteristic temperature To.

  Next, a semiconductor laser device for spatial light transmission and a spatial light transmission system according to the present invention will be described in detail with reference to the illustrated embodiments.

[First Embodiment]
FIG. 1 shows a sectional view of a semiconductor laser element of a semiconductor laser device for spatial light transmission according to a first embodiment of the present invention. The semiconductor laser device for spatial light transmission according to the first embodiment is an example in which a GaInAsP layer is used as a barrier layer of a semiconductor laser element. Here, “n−” and “p−” indicate n-type and p-type doping, respectively, and “un−” indicates that they are not doped.

In FIG. 1, 101 is an n-GaAs substrate, 102 is an n-Al _0.5 Ga _0.5 As cladding layer (1.8 μm), 103 is an n-Al _0.4 Ga _0.6 As guide layer ( _0.5 μm). 1 μm), 104 a, 104 b, 104 c are barrier layers, 105 a, 105 b are well layers, 106 is a p-Al _0.4 Ga _0.6 As guide layer (0.1 μm), 107 is p-Al _0.5 Ga _{0. .5} As first cladding layer (0.15 [mu] m), 108 is p-GaAs etching stop layer (4 nm), 109 is p-Al _0.5 Ga _0.5 As second cladding layer (1.28), 110 is n-Al _0.7 Ga _0.3 As block layer (0.6 μm), 111 is an n-GaAs block layer (0.3 μm), 112 is a p-GaAs cap layer (about 4 μm), 113a and 113b are electrodes is there. An active layer is composed of un-In _0.2 Ga _0.8 As _0.5 P _0.5 barrier layers 104a, 104b, 104c and un-In _0.06 Ga _0.94 As well layers 105a, 105b. Yes. The well layers 105a and 105b have a DQW (Double Quantum Well) structure with a thickness of 5 nm. The barrier layers 104a and 104c each have a thickness of 20 nm, and the barrier layer 104b has a thickness of 5 nm.

  Next, a method for manufacturing the semiconductor laser element will be described with reference to FIGS. 2A to 2F.

First, as shown in FIG. 2A, an n-Al _0.5 Ga _0.5 As cladding layer 102 (1.8 μm) and an n-Al _0.4 Ga _0.6 As guide layer are formed on an n-GaAs substrate 101. 103 (0.1 μm) and un-In _0.2 Ga _0.8 As _0.5 P _0.5 barrier layers 104a, 104b, 104c (104a, 104c are 20 nm, 104b is 5 nm), un-In _0.065 DQW active layer composed of Ga _0.935 As well layers 105a and 105b (5 nm each), p-Al _0.4 Ga _0.6 As guide layer 106 (0.1 μm), p-Al _0.5 Ga _0.5 As first cladding layer 107 (0.15 μm), p-GaAs etching stop layer 108 (4 nm), p-Al _0.5 Ga _0.5 As second cladding layer 109 (1.28 μm), p-GaAs cap layer 112a (0.75 μm) is sequentially deposited by MOCVD (Metal Organic Chemical Vapor Deposition) method. Formed by first crystal growth.

A 6 μm wide resist mask (not shown) is formed thereon by a normal photolithography technique, and as shown in FIG. 2B, the p-GaAs cap layer 112a and p-Al ₀ except for the region where the mask is formed. _.5 Ga _0.5 As The second cladding layer 109 is removed by etching to form a striped ridge. At this time, etching is stopped at the p-GaAs etching stop layer 108 using an etchant having high selectivity in the p-Al _0.5 Ga _0.5 As second cladding layer 109 and the p-GaAs etching stop layer 108. Thus, the ridge width is controlled. In the first embodiment, the width of the bottom of the ridge is adjusted to be 2.7 μm.

Next, as shown in FIG. 2C, after removing the resist on the ridge, the n-Al _0.7 Ga _0.3 As block layer 110 (0.6 μm) and the n-GaAs block layer 111 (0.3 μm) are used. The p-GaAs cap layer 112b (1.05 μm) is sequentially formed by the second crystal growth by the MOCVD method. At this time, the growth proceeds on both sides of the ridge and on the top of the ridge, and a growth film is formed in the shape of a mountain-shaped projection on the top of the ridge.

  A resist is applied onto the entire surface of the resist, exposed to the entire surface, and adjusted for development time, so that the resist is removed until a chevron-shaped protrusion on the ridge appears. The resist on the ridge is removed, and the resist outside the ridge remains. By etching in this state, as shown in FIG. 2D, the second growth layer formed on the ridge and a part (about 0.05 μm) of the p-GaAs cap layer 112a which is the first growth layer are removed. Remove.

Then, as shown in FIG. 2E, the resist is removed, and a p-GaAs cap layer 112c (4 μm) is formed by MOCVD. At this time, the impurity was doped at a density of 1 × 10 ²⁰ cm ⁻³ so that only the uppermost layer of 0.3 μm can be in contact with the electrode.

  Thereafter, as shown in FIG. 2F, the wafer is polished to a thickness of 100 μm, and an AuZn / Au electrode 113a is formed on the p side and an AuGe / Ni / Au electrode 113b is formed on the n side by vapor deposition.

After forming the electrodes 113a and 113b, the Si film and the Al ₂ are cleaved with a length of 500 μm in the direction perpendicular to the stripe serving as the resonator, and the reflectance is 5% and 90% on both end faces, respectively. An O ₃ film is formed by vapor deposition.

  Finally, each stripe is divided and mounted on the stem to complete the semiconductor laser device.

  When a pulse current having a duty of 50% was applied to the semiconductor laser device thus fabricated, it oscillated at an oscillation wavelength of 869 nm (25 ° C.). The oscillation threshold current was 19 mA and the efficiency was 1 W / A. Further, as the applied pulse current was increased, it oscillated without causing COD (Catastrophic Optical Damage) up to an output of 300 mW. The characteristic temperature To = 130K, the light emission lifetime τ = 8.0 nsec, and the relative noise intensity RIN = −120 dB / Hz, and the characteristic temperature was high, the light emission lifetime τ was small, and good characteristics with low noise were obtained.

  The oscillation wavelength of this semiconductor laser element was 856 nm at −25 ° C. and 885 nm at 85 ° C.

  By using the semiconductor laser element as a light source for spatial light transmission, it is possible to provide a light source with high output and low power consumption compared with the case of using an LED, and high reliability capable of high-speed modulation. A semiconductor laser device for spatial light transmission can be provided.

  FIG. 3 shows a laser module as an example of a semiconductor laser device for spatial light transmission in which this semiconductor laser element is molded to 2 mmφ with a resin mixed with a scattering agent to realize eye-safety.

In FIG. 3, 120 is the semiconductor laser element, 122 is a Si photodiode, 123 is an epoxy mold resin as an example of the mold resin portion, 123a and 123b are lenses formed of the mold resin, and 124 is a silicon precoat. Resin, 125 is a SiO ₂ diffusing agent, 126 is a circuit board, and 127 is an IC (integrated circuit) constituting a laser driving circuit and a preamplifier for receiving signal amplification. In addition, it is made eye-safe by mixing a diffusing agent 125 made of SiO ₂ resin into the lens 123a of the LD emitting portion of the mold resin. The silicon-based precoat resin 124 is in a jelly state at the time of curing, and has a role of protecting the chip of the semiconductor laser element 120 from stress strain, and ensures the reliability of the semiconductor laser element 120.

  The circuit board 126 is provided with a recess 126a whose side surface is inclined at an angle of 45 °, and the semiconductor laser element 120 is mounted on the bottom of the recess 126a. The light emitted from the semiconductor laser element 120 is reflected by the side surface of the recess 26a, shaped by the lens 123a, and emitted from the module. The signal light entering the laser module is collected by the lens 123b and then received by the Si photodiode 122.

  Next, a method for manufacturing the laser module will be described. First, the semiconductor laser element 120, the Si photodiode 122, and the IC 127 are mounted on the circuit board 126, and wire bonding is performed. The semiconductor laser element 120 is mounted on a recess 126a provided on the circuit board 126, and an appropriate amount of a silicon-based precoat resin 124 is dropped on the recess 126a to coat the entire semiconductor laser element 120. Then, after the precoat resin 124 is cured in a jelly shape, the mold resin 123 is poured and cured using a mold so that the shape of the lenses 123a and 123b is transferred, and the circuit board is integrated with the mold resin 123. 126 is taken out to complete the laser module.

  When a pulse current having a duty of 50% was applied to the laser module thus fabricated, the laser module oscillated at an oscillation wavelength of 884 nm (25 ° C.). The oscillation wavelength was 871 nm at −25 ° C. and 900 nm at 85 ° C.

  Note that the angle of the side surface of the recess 126a of the circuit board 126, the material of the mold resin, the material of the precoat resin, the material of the diffusing agent, and the like are not limited to this, and may be various within the scope of the present invention. Of course, changes can be made.

  In the first embodiment, the oscillation wavelength can be set to 850 nm to 900 nm within the use environment temperature range of −25 ° C. to + 85 ° C., and safety for the eyes is improved, and at the same time, an inexpensive Si photodiode is used. The module cost can be reduced. Further, by setting the oscillation wavelength to 850 nm or more and 900 nm or less in a molded state depending on the module to be used, it becomes possible to make the most of the performance of this semiconductor laser device for spatial light transmission.

  The temperature dependence of the oscillation wavelength of the semiconductor laser device investigated by the present inventor is about 0.24 nm to 0.29 nm when the temperature dependence of the oscillation wavelength shown in FIG. The wavelength becomes longer, and the difference between the wavelength at 85 ° C. and the wavelength at −25 ° C. is about 30 nm.

  Therefore, when the semiconductor laser element before molding has an oscillation wavelength at room temperature (25 ° C.) of about 885 nm or less in a chip state, it can be 900 nm or less at 85 ° C. during mold mounting.

  Furthermore, for eye-safe, when a semiconductor laser element is molded with a resin mixed with a diffusing agent and used as a laser module for spatial light transmission, for example, in the laser module of the first embodiment, the wavelength of 15 nm varies depending on the presence or absence of molding. To do. Further, when experiments were conducted with various mold sizes, mold materials, mold shapes, and molding methods, the wavelength during driving of the semiconductor laser element was increased by 5 nm to 15 nm. Thus, in the eye-safe laser module, the wavelength varies greatly depending on the usage pattern.

  Assuming that the operating environment temperature range is −25 ° C. to + 85 ° C., the wavelength variation due to the temperature change of the semiconductor laser element itself is about 30 nm, and the wavelength becomes longer by 5 nm to 15 nm by molding. In the apparatus, a total of 40 nm wavelength of wavelength variation 30 nm due to temperature and wavelength variation 10 nm due to mold varies. That is, while the optimum wavelength range is 850 nm to 900 nm, there is only a margin for wavelength variation of 10 nm. If the active layer structure is determined so that the oscillation wavelength is 850 nm to 900 nm without being molded, there is a possibility that it is out of the optimum wavelength range when molded.

  Therefore, it is most preferable to set it to 900 nm or less at 85 ° C. and 850 nm or more at −25 ° C. after molding.

  In other words, considering that the wavelength varies from 0.24 nm to 0.29 nm per 1 ° C., the oscillation wavelength of the semiconductor laser element should be 857 nm to 871 nm at room temperature (25 ° C.) in the chip state of the semiconductor laser element before molding. However, as an eye-safe semiconductor laser device for spatial light transmission, the performance can be most exhibited while satisfying the IrDA standard.

  As described above, when the semiconductor laser element is in a chip state, the oscillation wavelength at 25 ° C. is set to 857 nm or more and 871 nm or less, thereby realizing an eye-safe semiconductor laser device that oscillates at 850 nm or more and 900 nm or less within the operating environment temperature range. It becomes possible.

  As described above, in the semiconductor laser device for spatial light transmission, the fluctuation of the oscillation wavelength with respect to the temperature can be reduced as the Ga composition of the well layer is smaller and the well width is thinner.

  Further, when a GaInAsP barrier is used as the barrier layer, the wavelength fluctuation can be reduced as the Ga composition of the barrier layer is smaller.

  As another means for realizing eye-safety, it is conceivable to use a diffusion plate. However, when a diffusion plate is used, the size of the module becomes large, which is not preferable for use in a portable terminal. On the other hand, as shown in the first embodiment, by using a mold type laser module, it is possible to reduce the size and mount it on a small device such as a portable terminal.

  In addition, when molding with a resin mixed with a scattering agent, the degree of heat dissipation varies depending on the size of the module, and the amount of wavelength variation also varies. That is, a small module is preferable for a portable terminal. However, if the size is small, heat dissipation is poor and wavelength fluctuations are large. On the other hand, if a large module can be used, such as for home servers, increasing the size of the module improves heat dissipation and reduces wavelength fluctuations. In this case, higher output and higher speed can be achieved.

  In this way, by setting the oscillation wavelength to 850 nm or more and 900 nm or less in a molded state depending on the module to be used, it becomes possible to make the best use of the performance of the semiconductor laser device for spatial light transmission.

  Further, by adopting a quantum well structure using InGaAs as a well layer, it is possible to provide a semiconductor laser device capable of high-speed modulation with high output, low power consumption and high reliability.

  More specifically, the following was found from the experimental results in which the well width, the composition of the well layers, and the number of wells were changed.

  By setting the well layer to 8 nm or less, light is efficiently converted with a small number of carriers due to the quantum effect, and the efficiency is increased and the output can be increased by decreasing the threshold current.

  As the In composition of the well layer increases, the amount of strain increases, the effective mass of carriers decreases, the threshold current decreases, the efficiency improves, and higher output can be achieved.

  As described above, the threshold current is decreased, the efficiency is increased, and the reliability is ensured, so that the output can be increased and the high-speed modulation can be performed. When the well width is increased and the number of wells is increased, the characteristic temperature To increases and the characteristics are stabilized with respect to the temperature.

  In addition, reliability was improved by adding In to the well layer. This is probably because dislocation propagation (crystal defect growth) in the well layer was suppressed by adding In.

  In addition, increasing the well width or increasing the number of wells further improved the reliability. This is because the light intensity (light density) per unit volume decreases as the volume of the well layer increases, so even if there is an absorber with the same density in the well layer or at the emission end face, the absorber Since the intensity of light that is felt becomes smaller and the amount of absorption decreases, it is considered that the reliability is improved as a result.

  Next, an experiment on the dependence of the emission lifetime τ on the thickness of the active layer conducted by the present inventor shows that it depends on the thickness of the active layer, and the emission lifetime increases as the thickness of the active layer decreases. τ can be reduced. That is, the quantum well active layer can be modulated at a higher speed than the bulk active layer, and the thinner the well width, the more suitable as the semiconductor laser device for spatial light transmission. In addition, by making the active layer a quantum well structure, the efficiency of carrier-to-light conversion is improved by the quantum effect compared to the bulk active layer, and device characteristics with low threshold, high efficiency, and high reliability are obtained. At the same time, the light emission lifetime τ can be reduced. Desirably, the effect is particularly great when the well width is 8 nm or less. In addition, the smaller the number of wells, the shorter the light emission lifetime τ, and the effect is particularly great when the number of wells is 4 or less.

  The reason why the light emission lifetime τ can be shortened while maintaining reliability is considered to be that the light emission recombination was efficiently performed with a small amount of injected carriers with little light emission recombination and light absorption.

  Further, from the experiment on the dependency of the light emission lifetime τ on the thickness of the active layer, noise becomes large when the well width is 4 nm or less and 8 nm or more, and it is not suitable as a communication light source. This is considered to be because the sharpness of the interface between the well layer and the barrier layer is greatly influenced by 4 nm or less, and the noise is increased. At 8 nm or more, it is considered that a plurality of quantum levels approached within the band, causing noise.

  Therefore, by setting the well width to 4 nm or more and 8 nm or less, the noise of the semiconductor laser element can be suppressed to be sufficiently small for use in high-speed spatial light transmission.

  Further, when the In composition of the well layer is increased, noise is reduced. This is presumably because the insertion of In separates the heavy hole band and the light hole band in the valence band of the well layer and reduces interference between the bands.

  As the well width is reduced, the threshold current decreases, the efficiency increases, and high output is possible. In addition, the light emission lifetime τ can be shortened and high-speed modulation becomes possible, and the effect is particularly remarkable at 8 nm or less. More preferably, the effect is large when the thickness is 6 nm or less.

  On the other hand, when the well width is increased, the characteristic temperature To can be increased and the reliability is improved.

  Noise becomes low at 4 nm to 8 nm, and becomes large even if it is made thinner or thicker. More preferably, noise is reduced by setting the thickness to 4 nm or more and 6 nm or less.

  In addition, when the In composition is increased, the threshold current is reduced, the efficiency is improved and the output can be increased, the noise is reduced, the reliability is improved, and the In composition is 0.04 or more (Ga composition 0). .96 or less), these effects are particularly remarkable.

  However, as In is increased, the band gap energy Eg becomes smaller and the wavelength becomes longer. To shorten the wavelength, the well width must be made thinner. Become. In order to obtain a desired wavelength with a well width of 4 nm or more, the Ga composition must be 0.85 or more. Further, when the Ga composition is made smaller than 0.9, the operating voltage increases. It is desired that the Ga composition of the well layer is 0.85 or more, more preferably 0.9 or more.

  Considering the combination of the well width and the composition of the well layer, the well width d is 4 ≦ d ≦ 8 nm, the Ga composition x of the well layer is 0.85 ≦ x <1, and more preferably, the well width d is 4 ≦ d ≦ 4. By satisfying the Ga composition x of 6 nm and the well layer of 0.9 ≦ x ≦ 0.96, all of the LD characteristics, wavelength, emission lifetime τ, noise, characteristic temperature To, and reliability are satisfied, and high output is achieved. A semiconductor laser for spatial light transmission capable of high-speed modulation can be provided.

  As the number of wells increases, the characteristic temperature To improves. In addition, reliability is improved. However, noise increases and the light emission lifetime τ also increases. Considering the threshold current and efficiency, the number of wells 1 to 4 is most appropriate.

As is clear from the experimental results of comparison of characteristics by changing the Ga composition x and P composition y of the Ga _x In _1-x As _1-y P _y barrier layer of the quantum well active layer, the barrier layer is made of GaInAsP. As a result, higher reliability can be ensured than when an AlGaAs barrier layer is used. This is because active Al is not included in the quantum well structure serving as the active layer, so that non-radiative recombination centers and oxygen serving as absorbers are reduced in the active layer and at the end face, resulting in improved reliability. Conceivable. Even in the AlGaAs barrier layer, sufficient reliability can be ensured if the Al composition is 0.3 or less.

  For applications that require smaller size or higher output, it is preferable to use a highly reliable GaInAsP barrier layer.

  Further, when a GaInAsP barrier is used, the smaller the Ga composition of the barrier layer, the smaller the wavelength variation with respect to temperature.

By increasing the Ga composition x and the P composition y of the GaInAsP barrier, the band gap energy Eg is increased and the characteristic temperature To can be increased. However,
y> -2.17x + 2.5
In the region represented by the above, device characteristics and reliability are remarkably deteriorated. Therefore,
y ≦ −2.17x + 2.5
It is necessary to set the composition in the region represented by In order to ensure good device characteristics and reliability,
y ≦ −2.05x + 2.34
It is desirable that the composition region represented by More preferably,
y ≦ −2.1x + 2.29
Is more desirable.

When the Ga composition and the P composition are decreased, the band gap energy Eg of the GaInAsP barrier layer is decreased, and the difference in the band gap energy Eg from the well layer cannot be obtained. Also,
y <−1.64x + 1.63
In the composition region represented by the above, the characteristics of the semiconductor laser device are remarkably deteriorated. Therefore,
y ≧ −1.64x + 1.63
It is necessary to use a composition region represented by In order to ensure good device characteristics,
It is preferable to use a composition region represented by y ≧ −1.64x + 1.7. More preferably,
y ≧ −1.54x + 1.73
It is most desirable to use a composition region represented by:

In the composition region, when the Ga composition x is decreased and the P composition y is increased,
y> −0.27x + 0.88
In the composition region represented by the formula, the efficiency decreases and the element resistance increases. Therefore,
y ≦ −0.27x + 0.88
It is preferable to use in the area | region represented by these. More preferably,
y ≦ −0.28x + 0.83
It is most desirable to use a composition region represented by:

  Further, in the region where the Ga composition x is less than 0.65, the device characteristics and reliability are remarkably deteriorated. Therefore, it is necessary to use a composition region where x ≧ 0.65. It is preferable to use a composition region where x ≧ 0.7.

Comparative experiment of semiconductor laser devices with different compositions of n-Al _0.4 Ga _0.6 As guide layer 103 and p-Al _0.4 Ga _0.6 As guide layer 106 sandwiching the quantum well active layer, and GaInP From the comparison experiment of the semiconductor laser device using as a guide layer, the following became clear. That is, if a GaInP layer is used for the guide layer, the efficiency decreases, the resistance increases, a large current and voltage are required to produce the same optical output, power consumption increases, and the temperature of the element increases. , Reliability decreases. On the other hand, when AlGaAs is used as the guide layer, the efficiency is increased, the element resistance is lowered, the output can be increased, and the power consumption can be reduced.

  In addition, by using the AlGaAs guide layer, the characteristic temperature To is improved and the stable operation is performed at a high temperature.

  The AlGaAs guide layer is particularly effective when the Al composition is 0.3 or more. Increasing the Al composition facilitates the bonding of oxygen, but the influence of the guide layer is small because the light distribution is smaller than that of the active layer.

[Second Embodiment]
FIG. 4 is a perspective view of a laser module as an example of a semiconductor laser device for spatial light transmission according to the second embodiment of the present invention. The semiconductor laser element of this laser module is a semiconductor laser element manufactured by the same manufacturing method as the semiconductor laser element of the first embodiment shown in FIGS. 2A to 2F, and the semiconductor laser element of the first embodiment is The difference is that the barrier layer of the quantum well active layer is an AlGaAs layer.

The semiconductor laser element of the laser module according to the second embodiment includes un-Al _0.15 Ga _0.85 As barrier layers 104a, 104b, 104c and un-In _0.072 Ga _0.928 As well layers 105a, 105b. Constitutes the active layer. The well layers 105a and 105b have a DQW structure with a thickness of 4.6 nm, the barrier layers 104a and 104c each have a thickness of 21.5 nm, and the barrier layer 104b has a thickness of 7.9 nm.

  When a pulse current with a duty of 50% was applied to the semiconductor laser device thus fabricated, it oscillated at an oscillation wavelength of 868 nm (25 ° C.). Further, this semiconductor laser element had an oscillation wavelength at -25 ° C. of 855 nm and an oscillation wavelength at 85 ° C. of 885 nm.

  In this semiconductor laser device, the oscillation threshold current was 20 mA, and the efficiency was 1.1 W / A. As the current increased, the output oscillated without COD up to an output of 280 mW. The characteristic temperature To = 180K, the light emission lifetime τ = 3.5 nsec, the relative noise intensity RIN = −135 dB / Hz, the characteristic temperature To was high, the light emission lifetime τ was small, and good characteristics with low noise were obtained.

  As shown in FIG. 4, the semiconductor laser element is molded to 7.5 mmφ with a resin mixed with a scattering agent to produce an eye-safe laser module. In FIG. 4, 220 is the semiconductor laser element, 221 is a heat sink, 222 is a Si photodiode, 223 is a mold resin mixed with a diffusing agent (also serves as a lens), 224 is a precoat resin, and 225 is a diffusing agent. is there. The mold resin 223 and the precoat resin 224 constitute a mold resin portion.

  As described above, the semiconductor laser element 220 is coated with reflective films having a reflectance of 5% and 90%, respectively, on both end faces that are cleaved, but the end face on which the 5% reflective film is formed faces upward in the drawing. The end face on which 90% of the reflective film is formed is mounted on the heat sink 221 so that it faces downward in the figure. The light of the semiconductor laser element 220 is emitted from the reflective film side having a reflectance of 5%, shaped by the lens function of the mold resin 223, and emitted from the module. In addition, the Si photodiode 222 is installed in the laser module of the second embodiment, and signal light entering the module is collected by the lens function of the mold resin 223 and received by the Si photodiode 222. Further, the mold resin 223 is mixed with a diffusing agent 225 to make it eye-safe.

  The precoat resin 224 uses a silicon-based resin that becomes jelly when cured, and has an effect of protecting the chip of the semiconductor laser element 220 from stress strain, and ensures the reliability of the semiconductor laser element.

  Next, a method for manufacturing the laser module according to the second embodiment will be described. First, the semiconductor laser element 220 and the Si photodiode 222 are mounted on the stem, and wire bonding is performed. The semiconductor laser element 220 is mounted on the heat sink 221. Then, an appropriate amount of silicon-based precoat resin 224 is dropped and coated so that the entire semiconductor laser element 220 is covered. After the precoat resin 224 is cured into a jelly shape, the mold resin 223 is poured and cured using a mold so that the lens shape is transferred, and the stem integrated with the mold resin 223 is taken out to complete the laser module. Let

  When a pulse current having a duty of 50% was applied to the eye-safe laser module, it oscillated at an oscillation wavelength of 872 nm (25 ° C.). Moreover, it was 859 nm at -25 degreeC and 889 nm at 85 degreeC.

  In the second embodiment, the diameter of the mold resin 203 is larger than that of the first embodiment, and the wavelength variation due to the temperature with or without the mold is 6 nm. The chip state of the semiconductor laser device before molding is 868 nm at 25 ° C., which is the same as that of the first embodiment, but the oscillation wavelength is 850 nm to 900 nm in the range of −25 ° C. to 85 ° C. even after molding. can do. Note that the temperature rise due to the mold does not change whether GaInAsP or AlGaAs is used for the barrier layer.

  Further, with respect to the composition of the well layer, the well width, and the number of wells, the same effect as that of the GaInAsP barrier described in the first embodiment can be obtained.

  From the experimental results of changing the composition of the barrier layer, when AlGaAs is used for the barrier layer, the emission lifetime τ is further reduced as the Al composition of the barrier layer is increased. However, in this case, as the Al composition is increased, the light emission lifetime τ is reduced, but the reliability is also deteriorated. When the Al composition exceeds 0.3, sufficient reliability is obtained when the semiconductor laser device is driven at a high output. Cannot be obtained. Al is an active material and has a very strong bond with oxygen. If the Al composition is increased, oxygen uptake also increases. When AlGaAs is used for the barrier layer, oxygen is taken into the barrier layer, that is, oxygen is taken into the interface between the well layer and the barrier layer, and the non-radiative recombination of carriers confined in the well layer The emission lifetime τ is reduced. Further, oxygen in the barrier layer constituting the active region having a high light density is taken in, and this becomes an absorber of laser light, which is considered to deteriorate the characteristics and reliability of the semiconductor laser element.

  Further, as the Al composition of the barrier layer is larger, oxygen is more easily bonded at the end face, so that COD is more likely to occur.

  Further, when AlGaAs is used as a barrier layer, the light emission lifetime τ can be reduced while maintaining sufficient reliability for driving a semiconductor laser device at high output by setting the Al composition to 0.3 or less. Is possible.

  Further, from the experimental results of changing the composition of the barrier layer, the operating voltage Vop is lowered when the Al composition of the AlGaAs barrier layer is lowered.

  The injected carrier feels that the hetero interface between the well layer and the barrier layer is a barrier, and the smaller the Al composition, the smaller the barrier at this hetero interface, and the easier the carriers are injected into the well layer, thereby reducing the operating voltage Vop. It is guessed. Therefore, from the viewpoint of the operating voltage Vop, it is desirable that the Al composition of the barrier layer be as small as possible.

  However, if the Al composition of the barrier layer is made smaller than 0.1, ΔEg with the well layer becomes small, and characteristics such as threshold current, efficiency and characteristic temperature To deteriorate, so that the Al composition needs to be 0.1 or more. .

  Further, from the experimental results of changing the composition of the barrier layer, the characteristic temperature To is improved when the AlGaAs barrier is used than when the GaInAsP barrier is used. This is presumably because the difference (ΔEc) between the conduction layers of the well layer and the barrier layer can be increased by using the AlGaAs barrier. In particular, the effect is large at Al composition of 0.15 or more.

  Further, when the Al composition of the AlGaAs barrier layer is increased (the Ga composition is decreased), the wavelength fluctuation due to temperature can be reduced.

  As a result of the experiment in which the composition of the barrier layer was changed, when AlGaAs was used for the barrier layer, the noise increased when the composition of Al was increased. This is thought to be due to the fact that by increasing the composition of Al, the oxygen taken into the interface between the well layer and the barrier layer increases, the LD characteristics deteriorate, and this causes noise. The Al composition of the AlGaAs barrier layer is preferably 0.3 or less. In addition, when the Al composition was 0.3 or less, the noise was smaller than that of the GaInAsP barrier layer. On the other hand, when the Al composition is decreased, the noise is reduced until the Al composition is 0.1, but when the Al composition is smaller than 0.1, the noise is increased. This is probably because the difference in the band gap energy Eg between the well layer and the barrier layer was reduced, and a plurality of quantum levels approached within the band, causing noise.

  Therefore, when AlGaAs is used for the barrier layer, the noise can be reduced by setting the Al composition to 0.1 or more and 0.3 or less, and the noise can be reduced by reducing the Al composition within this range. .

  From the above, considering the LD characteristics, wavelength fluctuation, light emission lifetime τ, noise, characteristic temperature To, and reliability, the Al composition of the AlGaAs barrier layer is preferably 0.1 or more and 0.3 or less.

[Third Embodiment]
FIG. 5 shows a sectional view of a semiconductor laser element of a semiconductor laser device for spatial light transmission according to a third embodiment of the present invention.

In FIG. 5, 301 is an n-GaAs substrate, 302 is an n-Al _0.5 Ga _0.5 As cladding layer, 303 is an n-Al _0.4 Ga _0.6 As guide layer, 304a and 304b are barrier layers, 305 is a well layer of a DQW structure, 306 is a p-Al _0.4 Ga _0.6 As guide layer, 307 is a p-Al _0.5 Ga _0.5 As first cladding layer, 314 is a p-low doped Ga _{0 .84} In _0.16 As _0.4 P _0.6 layer, 309 is a p-Al _0.5 Ga _0.5 As second cladding layer, 312 is a p-GaAs cap layer, and 313 a and 313 b are electrodes.

  Next, a method for manufacturing the semiconductor laser element will be described with reference to FIGS. 6A to 6C.

On the n-GaAs substrate 301, as shown in FIG. 6A, an n-Al _0.5 Ga _0.5 As cladding layer 302 (1.8 μm), an n-Al _0.4 Ga _0.6 As guide layer 303 ( 0.1 μm), Ga _0.75 n _0.25 As _0.4 P _0.6 barrier layers 304 a and 304 b (20 nm each) and Ga _0.925 In _0.075 As well layer 305 (4.6 nm). SQW active layer, p-Al _0.4 Ga _0.6 As guide layer 306 (0.1 μm), p-Al _0.5 Ga _0.5 As first cladding layer 307 (0.15 μm), p-Ga _{0 .84} In _0.16 As _0.4 P _0.6 layer 314 (15 nm), p-Al _0.5 Ga _0.5 As second cladding layer 309 (1.28 μm), p-GaAs cap layer 312 (0 .5 μm) are sequentially grown by MOCVD. The p-GaAs cap layer 312 is doped with an impurity density of 1 × 10 ²⁰ cm ⁻³ so that only the uppermost layer of 0.3 μm can be easily contacted with the electrode. The barrier layers 304a and 304b constituting the quantum well active layer are not limited to this composition, and a GaInAsP layer or an AlGaAs layer having the composition range described in the first and second embodiments can be used.

A resist mask having a width of 6 μm is formed thereon by a normal photolithography technique, and as shown in FIG. 6B, the p-GaAs cap layer 312 and the p-Al _0.5 Ga ₀ except for the region where the resist mask is formed. _.5 As the second cladding layer 309 is removed by etching to form a striped ridge. At this time, an etchant having high selectivity is used for the p-Al _0.5 Ga _0.5 As second cladding layer 309 and the p-Ga _0.84 In _0.16 As _0.4 P _0.6 layer 314. , P-Ga _0.84 In _0.16 As _0.4 P _0.6 layer 314 is controlled to stop etching and control the ridge width. In the third embodiment, the width of the bottom of the ridge is adjusted to be 2.7 μm.

  After removing the resist on the ridge, the wafer is polished to a thickness of 100 μm, and as shown in FIG. 6C, Ti / Pt / Au electrode 313a is deposited on the p side and AuGe / Ni / Au electrode 313b is deposited on the n side by vapor deposition. Form.

Then, the film is cleaved to a length of 500 μm in a direction perpendicular to the stripe that becomes the resonator, and an Si film and an Al ₂ O ₃ film are vapor-deposited on both end faces thereof so that the reflectance becomes 5% and 90%, respectively. Form.

  Finally, each stripe is divided and mounted on the stem to complete the semiconductor laser device.

  The oscillation wavelength of this semiconductor laser element was 856 nm at −25 ° C., 868 nm at 25 ° C., and 884 nm at 85 ° C.

  Further, as in the laser module shown in FIG. 3, when the resin was molded to 2 mmφ, the oscillation wavelength was 871 nm at −25 ° C., 883 ° C. at 25 ° C., and 899 nm at 85 ° C.

  Also in the third embodiment, by setting 868 nm at 25 ° C. in the chip state of the semiconductor laser element before molding, an oscillation wavelength of 850 nm or more and 900 nm or less is possible within the operating environment temperature range when the mold is mounted. A laser module capable of high-speed modulation with low output and low power consumption can be provided.

  In the third embodiment, current confinement to the ridge portion is realized by using a Schottky junction between the p-side electrode 313a and the p-GaInAsP layer 314. Therefore, compared with 1st Embodiment, a manufacturing process is simplified and manufacturing cost can be reduced significantly. As a result, it is possible to provide a light source for spatial light transmission capable of high-speed modulation with high output and low power consumption, even though the cost is comparable to that of conventional LEDs.

[Fourth Embodiment]
FIG. 7 shows a cross-sectional view of the semiconductor laser element of the semiconductor laser device for spatial light transmission according to the fourth embodiment.

In FIG. 7, 401 is an n-GaAs substrate, 402 is an n-Al _0.5 Ga _0.5 As cladding layer, 403 is an n-Al _0.4 Ga _0.6 As guide layer, 404a, 404b, 404c, 404d. Is un-In _0.1 Ga _0.9 As _0.65 P _0.35 barrier layer, 405a, 405b and 405c are Ga _0.935 In _0.065 As well layers, and 406 is p-Al _0.4 Ga _{0. .6} As guide layer, 407 is p-Al _0.5 Ga _0.5 As first cladding layer, 408 is p-Ga As etching stop layer, 410 is n-Al _0.6 Ga _0.4 As block layer, 411 Is an n-GaAs block layer.

  Next, a method for manufacturing the semiconductor laser element will be described with reference to FIGS. 8A to 8D.

As shown in FIG. 8A, an n-Al _0.5 Ga _0.5 As cladding layer 402 (1.8 μm), an n-Al _0.4 Ga _0.6 As guide layer 403 (on the n-GaAs substrate 401). 0.1 μm), un-In _0.1 Ga 0.9 As _0.65 P _0.35 barrier layers 404 a, 404 b, 404 c and 404 d (404 a and 404 d are 20 nm, 404 b and 404 c are 5 nm) and Ga _0.935 In _{0 0.065} As well layers 405a, 405b and 405c (each 4.6 nm), TQW active layer, p-Al _0.4 Ga _0.6 As guide layer 406 (0.1 μm), p-Al _0.5 Ga _{0 0.5} As first cladding layer 407 (0.15 μm), p-GaAs etching stop layer 408 (4 nm), n-Al _0.6 Ga _0.4 As block layer 410 (0.8 μm), n-GaAs block layer 411 (0.01 μm) is sequentially formed by the first crystal growth by MOCVD. The barrier layers 404a, 404b, 404c, and 404d constituting the quantum well active layer are not limited to this composition, and the GaInAsP layer or the AlGaAs layer having the composition range described in the first and second embodiments is used. be able to.

On top of this, a resist mask is formed by a normal photolithography technique leaving a 5 μm wide stripe, and as shown in FIG. 8B, the n-Al _0.6 Ga _0.4 As block layer 410 and the n-GaAs in the stripe portion are formed. The block layer 411 is removed by etching to form stripe-shaped grooves. At this time, by using a highly selective etchant in the n-Al _0.6 Ga _0.4 As blocking layer 410 and the p-GaAs etching stop layer 408, etched in p-GaAs etching stop layer 408 so as to stop The stripe groove width is controlled. In the fourth embodiment, the width of the bottom of the groove is adjusted to be 3.0 μm.

After removing the resist on the wafer surface, a p-Al _0.5 Ga _0.5 As second cladding layer 409 (1.28 μm) and a p-GaAs cap layer 412 (4.75 μm) are formed as shown in FIG. 8C. Sequentially formed by second crystal growth by MOCVD. The p-GaAs cap layer 412 is doped with an impurity density of 1 × 10 ²⁰ cm ⁻³ so that only the uppermost layer of 0.3 μm can be easily contacted with the electrode.

  The wafer is polished to a thickness of 100 μm, and as shown in FIG. 8D, a Ti / Pt / Au electrode 413a is formed on the p side and an AuGe / Ni / Au electrode 413b is formed on the n side by vapor deposition.

Then, the Si film and the Al ₂ O ₃ film are formed by vapor deposition so as to be cleaved with a length of 500 μm in a direction perpendicular to the stripe serving as the resonator, and to have a reflectance of 5% and 90%, respectively.

  Finally, each stripe is divided and mounted on the stem to complete the semiconductor laser device.

  The oscillation wavelength of the semiconductor laser element was 855 nm at −25 ° C., 867 nm at 25 ° C., and 884 nm at 85 ° C.

  Further, similarly to the laser module shown in FIG. 3, when the semiconductor laser element was molded to 2 mmφ with resin, the oscillation wavelength was 870 nm at −25 ° C., 882 ° C. at 25 ° C., and 899 nm at 85 ° C.

  Also in the fourth embodiment, by setting 867 nm at 25 ° C. in the chip state of the semiconductor laser element before molding, an oscillation wavelength of 850 nm or more and 900 nm or less is possible within the operating environment temperature range when mounting the mold. A laser module capable of high-speed modulation with low output and low power consumption can be provided.

  In the fourth embodiment, the manufacturing process is simplified and the manufacturing cost can be reduced as compared with the first embodiment. As a result, it is possible to provide a spatial light transmission light source capable of high-speed modulation with high output and low power consumption as compared with an LED while suppressing cost.

  The semiconductor laser device for spatial light transmission according to the present invention is not limited to the laser modules of the first to fourth embodiments. It goes without saying that various changes can be made without departing from the gist of the present invention, such as the optical confinement structure of the semiconductor laser device, the composition and thickness of the cladding layer, block layer, and the like.

[Fifth Embodiment]
FIG. 9 is a perspective view showing the configuration of the spatial light transmission system according to the fifth embodiment of the present invention. The spatial light transmission system according to the fifth embodiment is an example of spatial light transmission of 100 Mbps. By using the semiconductor laser device for spatial light transmission according to the first to fourth embodiments, it is possible to provide a light source that has high output, low power consumption, and high tracking speed compared to the case of using LEDs. Thus, high-speed spatial light transmission of 100 Mbps can be realized.

  In the fifth embodiment, a mobile terminal 508 such as a mobile phone, a personal server 509, a television 500, a personal computer 501, and a printer 502 can each perform wireless high-speed data transmission. The portable terminal 508 is equipped with the 2 mm mold laser module shown in FIG. The personal server 509, personal computer 501, printer 502, and television 500 are also equipped with the laser module shown in FIG.

  The portable terminal 508 is equipped with a high-pixel digital camera, and can capture high-quality images and moving images. Since these data have a large data capacity, transmission using conventional IrDA takes 10 seconds to several tens of seconds and is not always convenient. On the other hand, in the fifth embodiment, by using the semiconductor laser device of the first to fourth embodiments, data can be transmitted at high speed, so that data can be sent instantaneously, Its convenience has improved dramatically. For example, the application range is very wide, such as instantaneously transmitting high-quality image data to the printer 502, outputting a photograph, or instantaneously transmitting moving image data to the personal server 509 and displaying it on the television 500.

  Moreover, since the capacity of the portable terminal 508 becomes full when the number of data increases, it is very convenient to be able to transmit data frequently and easily to store data in the personal server 509 or the personal computer 501. It is. The fifth embodiment is an easy-to-use system that is also suitable in this respect because data can be transmitted instantaneously because high-speed transmission is possible.

  In addition, the data stored in the personal server 509 can be easily transferred to a portable terminal and viewed on the go or sent to the personal computer 501.

  Note that the spatial light transmission system of the present invention is not limited to the illustrated embodiment described above. Of course, various modifications can be made without departing from the scope of the present invention.

FIG. 1 is a cross-sectional view of a semiconductor laser element of a semiconductor laser device for spatial light transmission according to a first embodiment of the present invention. FIG. 2A is a schematic diagram for explaining a manufacturing process of the semiconductor laser element. FIG. 2B is a schematic diagram of a manufacturing process subsequent to FIG. 2A. FIG. 2C is a schematic diagram of a manufacturing process subsequent to FIG. 2B. FIG. 2D is a schematic diagram of a manufacturing process subsequent to FIG. 2C. FIG. 2E is a schematic view of a manufacturing process subsequent to FIG. 2D. FIG. 2F is a schematic diagram of a manufacturing process subsequent to FIG. 2E. FIG. 3 is a perspective view showing a laser module as an example of the semiconductor laser device for spatial light transmission. FIG. 4 is a perspective view of a laser module as an example of a semiconductor laser device for spatial light transmission according to a second embodiment of the present invention. FIG. 5 is a sectional view of a semiconductor laser element of a semiconductor laser device for spatial light transmission according to a third embodiment of the present invention. FIG. 6A is a schematic diagram for explaining a manufacturing process of the semiconductor laser element. FIG. 6B is a schematic view of a manufacturing process subsequent to FIG. 6A. FIG. 6C is a schematic diagram of a manufacturing process subsequent to FIG. 6B. FIG. 7 is a sectional view of a semiconductor laser element of a semiconductor laser device for spatial light transmission according to a fourth embodiment of the present invention. FIG. 8A is a schematic view for explaining a manufacturing process of the semiconductor laser element. FIG. 8B is a schematic view of the manufacturing process following FIG. 8A. FIG. 8C is a schematic diagram of a manufacturing process subsequent to FIG. 8B. FIG. 8D is a schematic diagram of a manufacturing process subsequent to FIG. 8C. FIG. 9 is a perspective view showing the configuration of the spatial light transmission system according to the fifth embodiment of the present invention. FIG. 10 is a diagram showing the temperature dependence of the oscillation wavelength of the semiconductor laser device for spatial light transmission. FIG. 11 is a diagram showing a composition plane for explaining the composition of the GaInAsP barrier layer. FIG. 12 is a composition plan view for explaining the composition range of the GaInAsP barrier layer. FIG. 13 is a cross-sectional view of a conventional LED device for spatial light transmission.

Explanation of symbols

101, 301, 401 ... n-GaAs substrate 102, 302, 402 ... n-AlGaAs cladding layer 103, 303, 403 ... n-AlGaAs guide layer 104a, 104b, 104c, 304a, 304b, 404a, 404b, 404c, 404d ... Undoped barrier layer 105a, 105b, 305, 405a, 405b, 405c ... undoped quantum well layer 106, 306, 406 ... p-AlGaAs guide layer 107,306,407 ... p-AlGaAs first cladding layer 108,408 ... p-GaAs Etching stop layer 109,309 ... p-AlGaAs second cladding layer 110,410 ... n-AlGaAs block layer 111,411 ... n-GaAs block layer 112,312,412 ... p-GaAs cap layer 112a ... p-GaAs cap layer 112b ... p-GaAs cap layer 112c ... p-GaAs cap layer 113a, 313a, 41 3a ... p-side electrode 113b, 313b, 413b ... n-side electrode 114,314 ... p-GaInAsP barrier forming layer 115 ... first n-AlGaAs layer 116 ... n-AlGaAs cladding layer 117 ... p-AlGaAs active layer 118 ... p -AlGaAs cladding layer 119 ... first p-AlGaAs layer 120,220 ... semiconductor laser chip 221 ... heat sink 122,222 ... photodiode 123,223 ... mold resin 123a, 123b ... lens 124,224 ... precoat resin 125,225 ... Diffusing agent 126 ... Circuit board 127 ... IC
500 ... TV 501 ... Personal computer 502 ... Printer 508 ... Mobile terminal 509 ... Personal server

Claims (9)

A semiconductor laser device in which an active layer, a guide layer sandwiching the active layer, and a clad layer sandwiching the active layer and the guide layer are formed on a GaAs substrate;
A mold resin portion covering at least a part of the semiconductor laser element,
The active layer of the semiconductor laser element has a quantum well structure composed of a well layer of InGaAs and a barrier layer sandwiching the well layer,
The semiconductor laser device is a semiconductor laser device for spatial light transmission, wherein an oscillation wavelength at 25 ° C. is not less than 857 nm and not more than 871 nm in a chip state before resin molding. The semiconductor laser device for spatial light transmission according to claim 1,
The semiconductor laser device is a semiconductor laser device for spatial light transmission, wherein an oscillation wavelength at a use environment temperature is 850 nm or more and 900 nm or less in a resin molded state. The semiconductor laser device for spatial light transmission according to claim 1,
The well layer is In _1-x1 Ga _x1 As (0.85 ≦ x1 <1);
A semiconductor laser device for spatial light transmission, wherein a well width d in the quantum well structure is 4 nm ≦ d ≦ 8 nm. The semiconductor laser device for spatial light transmission according to claim 1,
The well layer is In _1-x1 Ga _x1 As (0.9 ≦ x1 ≦ 0.96);
A semiconductor laser device for spatial light transmission, wherein a well width d in the quantum well structure is 4 nm ≦ d ≦ 6 nm. The semiconductor laser device for spatial light transmission according to claim 1,
A semiconductor laser device for spatial light transmission, wherein the number of wells in the quantum well structure is 4 or less. The semiconductor laser device for spatial light transmission according to claim 1,
The barrier layer is Ga _x2 In _1-x2 As _1-y P _y ,
The Ga composition x2 and P composition y of the barrier layer are
y ≦ −2.17x2 + 2.5
y ≧ −1.64 × 2 + 1.63
y ≦ −0.27x2 + 0.88
x2 ≧ 0.65
A semiconductor laser device for spatial light transmission characterized by satisfying the following condition. The semiconductor laser device for spatial light transmission according to claim 1,
A semiconductor laser device for spatial light transmission, wherein the barrier layer is Al _x3 Ga _1-x3 As (0.1 ≦ x3 ≦ 0.3). The semiconductor laser device for spatial light transmission according to claim 1,
2. A semiconductor laser device for spatial light transmission, wherein the guide layer is Al _x4 Ga _1-x4 As (x4 ≧ 0.3). A spatial light transmission system comprising the semiconductor laser device for spatial light transmission according to any one of claims 1 to 8.

Images