JPS6019675B2 - Gallium phosphide green light emitting device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a semiconductor light emitting device, and particularly relates to a semiconductor light emitting device using gallium phosphide (
Cap) relates to a green light emitting device using a crystal.

Recently, semiconductor light emitting devices, such as light emitting devices using semiconductor crystals such as GAP or gallium phosphide (GaSaP), have been widely used for various displays.

Among these, for example, Gap light emitting elements emit light from green to red,
The aAsP light-emitting element emits yellow to red light. Among these types of light emitting elements, green light emitting elements or yellow Ga
In AsP light emitting devices, nitrogen (N) is used as the luminescent center impurity.
Dope atoms. Among such nitrogen (N)-doped light emitting devices, a green GaP light emitting device is obtained as follows. That is, n-type Ga is grown on an n-type Gap substrate by liquid phase epitaxial growth or vapor phase epitaxial growth.
A p-layer is formed, and a p-type GaP layer is formed thereon by the above-mentioned growth method or diffusion method. Nitrogen (N) atoms, which are luminescent centers, are added to the n-type Gap layer and the p-type Gap layer. For example, lEEE Tramactionson Elect, published in July 1977.
ronDevices, VoIED-24 Shame. 9th of 7
According to pages 51 to 955, nitrogen (N) atoms are added using ammonia (NH3) or gallium nitride (GaN). Further, according to the right column of page 954 of this document, the nitrogen (N) atomic concentration (NT) and donor concentration (N.) added to the n-type GaP growth layer on the n-type Gap substrate
It has been stated that there is a correlation between Therefore, from this literature, in order to improve the luminous efficiency (RIG),
It may be possible to reduce No and increase NT near the pn junction. However, Fig. 9 on the mother page of this literature. Inferring from the experimental diagram of 6, NT is 1
It can be seen that when ×1 and 8/Iso or more, the lifetime (7c) of a small number of carriers decreases to less than 150 (nsec) and the luminous efficiency (RIG) decreases.

In the case of a green light-emitting element, in order to improve the luminous efficiency, it is necessary to increase the lifetime of minority carriers in the light-emitting region (n-type GaP layer near the p-n junction).
It is necessary to improve the concentration of luminescent centers (N atoms) in the luminescent region. For example, Ap issued in March 1973
pl. Phys. Lett. Vol, 22, port. 5, pages 22 to 229 (especially the '1} formula on page 229). And according to this document, n
It is specified that the minority carrier lifetime is at most 100 (female ec) when the type Cap layer No. is approximately 1 × 1 and 7 / earth, and N' at this time is 1 × 1 and 9 / earth. (However, J.ElecUonNEt. published in 1972.
Vol. 1, pages 39 to 53, it is clearly stated that the correct value of NT is 1/4 of the above NT. However, even in this document, the lifetime of minority carriers is improved and the n
No specific means for improving the NT of the GaP layer is specified. However, in order to improve the lifetime of minority carriers in the n-type GaP layer, it is sufficient to reduce No. as shown in Fischer on page 228. It can be seen from 2. But N. is saturated below 1x1bihada-3. On the other hand 1
J. issued to King 973. ElectrMhem. So
c. , Vol. 122, M. 3, 40th sword page - 412th
On the page, if the temperature at which the n-type gap layer is epitaxially grown is lowered and the p-n junction formation temperature is lowered to 350°C,
It is specified that the lifetime of minority carriers will be as high as 200 (ship ec).

And the NT at this time is 2 x 18/clump. N. is 7.6×
1 and 6/ground (and 6 x 1 and 6/swim), which is considered to satisfy the various conditions described above to some extent. Therefore, the luminous efficiency of the diode produced by this type of manufacturing method is 25A.
/The high value of the current flowing through the earth is a very high value of 0.35% (with mold). In addition, there is a document that conducts experiments and calculations on how much nitrogen atoms enter the GaP growth layer depending on the growth temperature using the same liquid phase growth method as in the above-mentioned document, for example, J. E1ecV
ochemSoc. , Vol. 119, M. 1st of 12
Pages 780 to 1782, especially the Fig. on page 1782. 2
is clearly stated.

Fig. on page 1782 of this document. According to 2, GaP
When the growth temperature is 960 qo, the limit of N' is about 2×18 degrees of ripeness, and it increases as the temperature rises. Considering the above explanation, in order to improve the luminous efficiency, the number of the n-type Gap layer should be low, for example, about 6 x 1 and 6/Sakaki, and the NT should be about 2 x 1 and 8/lump, and the number of minority carriers should be low. It can be inferred that it is preferable to set the lifetime to 200 (ship ec).

Based on this assumption, for example, the first part of JomM1ofCr$taIGrowth27 issued to 197 Buddha
Pages 83 to 192 In particular, considering the description on page 191 that the luminous efficiency is 0.33% (with mold) at a current of 7 A/ground, the N of the n-type GaP layer.

It can be inferred that the number of carriers is low, there are many NTs, and the lifetime of minority carriers is long. For example, the first part of this document
Fig. on page 89. From 6, n-type Ga on the n-type Gap substrate
When growing the P layer by liquid phase epitaxial growth, the Ga solution is heated for 2 days.
S and ammonia (NH3) are added from the gas phase, the S is cut off midway through purging at the 97-0 position to eliminate the S in the Ga solution, and Zn and NH3 are added into the Ga solution to Since the method uses liquid phase epitaxial growth of the type GaP layer, it can be inferred that the lifetime of minority carriers becomes longer and the luminous efficiency becomes higher.In addition, in Fig. 7 on page 189 of the same document, The graph shows the change in N with respect to the thickness of the layer, but this is simply a change in No due to S annihilation when purging is performed at a temperature of about 970°C. This is not a way to grow.By the way, the G obtained by this method
The ap green light-emitting element also shows the highest luminous efficiency under certain manufacturing conditions as in the above-mentioned literature, and the average luminous efficiency is 0.15% (with mold) when the current is 2 insects/ground.
It is the rank.

Therefore, the present inventors conducted research based on the above-mentioned literature, and found that they were able to construct a structure in which No at the p-n junction could be lowered and more precisely controlled, and the lifetime of minority carriers in the light-emitting region was increased. It has been found that by configuring the structure so that the NT can be highly controlled in the state, it is possible to obtain an average luminous efficiency of about 0.4% or more (with a mold, 2 small/clump).

The present invention was made in view of the above experimental facts, and provides a GaP green light emitting device that can obtain an average luminous efficiency of about 0.4% or more (with mold, 2 cm/ground).

Hereinafter, one embodiment of the present invention will be described with reference to the drawings.

First, using a liquid phase epitaxial growth apparatus as shown in FIG.
A Gap type Gap layer is formed and a p type GaP layer is formed thereon to obtain a Gap green light emitting device.

By configuring the n-type Gap layer so that No changes stepwise in this way, a luminous efficiency higher than that described in the above-mentioned literature can be obtained. This is because the luminescent center concentration (NT) of the n-type GaP layer, which is the luminescent region, is as high as approximately 2 x 18-3, and the minority carrier lifetime is extremely long, averaging over 33 seconds. by. This will be explained in detail below. First, the growth apparatus shown in FIG. 1 has a growth boat 12 placed inside a quartz reactor 11, and two heating devices 13a and 13b provided outside the reactor 11. Reference numeral 12 denotes a slider 15 having a recess 14a for accommodating the n-type Gap substrate 14, and a solution 1.
6 and a solution storage boat 18 having a small hole 17 for doping impurities, and lids 14b and 16b made of, for example, quartz are placed on the n-type Gap substrate and the solution 16. is provided. For example, zinc (Zn
) is provided with an impurity evaporation line 9. Further, on both sides of the reactor 11, a sekiguchi 11a for supplying gas,
11, a sekiguchi 11b for discharging gas is installed. With such a growth apparatus, n is grown on an n-type Gap substrate.
When forming a type Gap layer and a p-type Gap layer, FIG.
, b, and c by driving the growth boat 12. First, as shown in FIG. 2a, a sulfur (S)-doped n-type G
The aP substrate 24 is installed, 5 tA of Ga is stored in the solution storage tank of the solution storage boat 28, and the gas supply V inlet 1 is opened from 2 days to 2 days.
Growth boat 1 is operated by operating the heating device while gas is flowing in.
2 to 1010 qo to prepare a Ga solution 26 which is Gap unsaturated and does not contain donor impurities (natural donor impurities such as silicon are included). In this way, 18 minutes after reaching 1010qo, slider 2
5 is moved, and the Gap board 24 is moved as shown in FIG. 2b.
With part of the solution 26 placed on top, it is moved to a portion having a large number of small holes 27. At this time, the depth of the recess 24a of the slider 25 is set so that the thickness of the solution 26 on the Ga substrate 24 is, for example, 1.5 ribs. This state is maintained for 10 minutes, for example, so that the surface of the Gap substrate 24 dissolves into the solution 26, and then the cooling rate is maintained at a constant rate, for example, i. 5q
Cooling is performed at a rate of 0/min to a predetermined temperature, for example, 960 pm. As a result of this cooling, an n-type Gap layer not containing nitrogen as shown in FIG. 4, which will be described later, grows on the n-type GaG substrate 24 by about 20 peaks. No in the n-type GaP layer grown to this state is n-type Ga
N of P substrate. A little less than. For example, as shown in FIG. 4, the number of the n-type gap layer is 1.8×1 and 7.

As will be explained later, the surface of the quartz reaction tube that constitutes the growth reactor is annated with hydrogen gas, so a large amount of Si is mixed into the solution, and the concentration of Si is higher than that of the substrate. Sometimes it happens. Subsequently, after the temperature reaches 960 oo, the temperature is kept constant for a predetermined period of time, for example, 60 minutes, and at the same time as the holding starts, argon (A) gas containing ammonia (N&) is introduced from the gas mochi joint port 11a. By doing this, as will be explained later, the inflowed ammonia is
GaP partially grown through many small holes in the state shown in
It reacts with the gallium solution 26 on the substrate, nitrogen atoms are added to the saturated state, and a portion of the gallium solution 26 reacts with silicon (Si), which comes from quartz in the furnace, for example, to form Si3.
Make N4. Also, S in the solution 26 is partially evaporated during this holding time. Then, after 60 minutes, add the solution again, e.g.
Cool down to 900 qo at a cooling rate of 0/min. By this cooling, an n-type GaP layer is grown on the partially grown n-type Cap layer. This state is shown in Figure 4, where the entire n-layer has 4
This is a state of 0rm growth. Furthermore, this grown n-type G
The aP layer becomes in a state with a very large amount of nitrogen, and
For example, as is clear from Figure 4, there are very few Nos. 1.
3 x 1 to 6 pieces/approximately. Subsequently, after reaching 900°C, the temperature is kept constant for a predetermined period of time, for example, 3 minutes, and at the same time, the impurity evaporation source 19 made of, for example, Zn shown in FIG.
The heating device 13b is operated to raise the temperature to, for example, 560:00, and is kept at that temperature. When kept warm in this way, Zn, which has a high vapor pressure, evaporates, and the gas supply port 1 shown in Figure 1 evaporates.
The gas from 1'a enters through a large number of small holes 7 into the solution 26 on the substrate on which the n-type Gap layer is grown. After this, the solution is again cooled to 800° C. at a cooling rate of 1.5 qo/min. In this way, a p-type Gap layer doped with nitrogen at an amount of 2×1 to 8 μm/m is grown on the substrate on which the n-type GaP layer is grown. After this, the heating devices 13a and 13b are turned off (not shown) and allowed to cool naturally. By the way, the temperature programs of the growth boat 12 and the impurity evaporation source 19 described above are distributions as shown in FIGS. 3a and 3b, as is clear from the above-mentioned points. The donor concentration distribution of the n-type GaP layer on the n-type Gap substrate thus obtained decreased stepwise in the growth direction, as shown by the solid line in FIG.

This was briefly explained in the above examples, but when growing the n-type Gap layer on the substrate side (hereinafter referred to as the ln-th layer), the G
GaP in ap unsaturated and donor impurity-free solution
Because the process is performed by melting the surface of the substrate, the number is 1.8 x 1 and 7/1 and 7/mass order, as is clear from Fig. 4.

During the growth of this ln-th layer, the surface of the quartz reaction tube constituting the growth furnace is reduced by hydrogen gas, so a large amount of Si is mixed into the solution, and this is reflected in the growth of the ln-th layer. The main donor impurity is Si (No is the above value).
On the other hand, the p-type Gap layer side of the n-type Gap layer grown in the latter half (
Since the process is carried out in a gas atmosphere containing ammonia, a large amount of nitrogen is added to the gallium solution, and a part of the nitrogen and Si in the solution form a stable compound. It is thought that the single Si in the layer decreases, and the No in the second layer decreases. Therefore, the donor concentration in the first layer is mainly determined by the donor impurity (S in this example) dissolved from the substrate, and as is clear from FIG. It is thought that it will be about an order of magnitude lower. When No in the ln-th layer and No in the sha layer change in a stepwise manner (decreasing in the growth direction) in this way, a step-like donor distribution between the ln-th layer and the ln-th layer and the first layer will occur, as will be explained later. The structure removes crystal defects from the substrate and improves crystallinity in the first layer. In other words, even when nitrogen is added at a high concentration, the lifetime of minority carriers in the second n-layer can be extended, which improves luminous efficiency. You will be able to improve your On the other hand, as is clear from the above example, n on the substrate
The NT of the type gap layer, that is, the ln-th layer and the second layer is N.

The relationship is the opposite. This is when growing the n-type Gap layer.
This is because ammonia is added midway through, but at the same time as ammonia is added to the solution, it is held in an Ar atmosphere for 60 minutes, so even if some of the ammonia forms a compound with Si in the solution, more ammonia will be added. is added,
Nitrogen atoms enter the first layer up to the solid solubility limit concentration at 960 qo. Therefore, a large amount of nitrogen enters the vicinity of the pn junction, which becomes a major factor in improving luminous efficiency. Next, the effects of the present invention based on the above embodiment will be explained with reference to FIGS. 5 to 13. First, in order to improve the luminous efficiency of the GaP green light emitting device, as specified in the conventional example of the present invention, N of the n-type GaP layer must be increased.

The objective is to lower the NT, increase the NT, and improve the lifetime of minority carriers. Therefore, when the n-type GaP layer on the substrate has a two-layer structure as in the embodiment of the present invention, the lifetime of minority carriers for No in the first layer, which contributes most to light emission, was measured, as shown in Figure 5a. It turned out to be true. That is, as is clear from FIG. 5a, when the number of the first layer is 5×1 and 6/clump or less, the lifetime of the minority carrier is 330 (ship ec) or more, and for example, 1973 as described in the conventional example. Appl.Wang published in March. p
h$.戊TT. , Vol. 22, M. 5, page 228 (
The maximum 100 (ship ec
) (Fig. 5b), J. El
ect-rochem. Soc. , Vol. 122,M
．． The actual value of approximately 200 (nsec) (Fig. 5c) described on page 410 of No. 3 (hereinafter referred to as Document-B) is larger than the x mark. In addition, in the case where an n-type GaP layer is formed on an n-type Gap substrate by vapor phase growth, for example, 19
Jomhada l of Electro published in 1973
nicMoterial, Vol. 2, M. 1 of 13
According to pages 7 to 159 (hereinafter referred to as Document-C), when nitrogen is included, the lifetime of minority carriers is 100~
It is described that the value is as low as 150 (mec). The reason why the lifetime of minority carriers in the first fox layer according to the present invention is long is because the n-type GaP layer has a two-layer structure, and the non-emissive centers (
The propagation of various dislocations or point defects such as residual impurities caused by imperfections in the crystal to the optical layer is caused by crystal discontinuities such as stress caused by rapid concentration fluctuations in the ln layer and the phantom layer. This is thought to be due to the fact that it is cut off or absorbed depending on the gender, and the amount decreases. Therefore, for comparison, the n-type Gap
When the No distribution in the layer is monotonically decreased in the growth direction, p-
The lifetime of minority carriers near the n-junction (N, = ~
2×1 and 8/ground), it is found that it is much shorter than the present invention, as shown in FIG. 5d. This is n-type Ga
This is considered to be because there is no rapid change in No in the P layer, so the non-emissive center propagated from the substrate continues up to the vicinity of the pn junction. Also, based on FIG. 5, NT for the first layer
and the product of minority carrier lifetime (conventionally, N
, and the lifetime of minority carriers are directly related to the luminous efficiency), and the results are shown in FIG. 6a.

As is clear from this Figure 6a, the number of the first layer is 5×1
and 8/T, the product of the lifetime of NT and the minority carrier is 60 or more, and Document-A (Figure 6b), Document-B (Figure 6c), or the n-type GaP layer. It can be seen that the value is much higher than that in which No monotonically decreases in the growth direction (Fig. 6d). The reason for this is thought to be that, in addition to the above explanation (explanation of FIG. 5), the NT in the second layer can be suppressed with high precision and at a high level. Since the NT in the first layer has been mentioned here, the emission wavelength associated with the NT in the present invention is illustrated in FIG. 7. Furthermore, based on FIGS. 5 and 6, N of the second layer.

The luminous efficiency was determined as shown in FIG. 8. In Fig. 8, the luminous efficiency is when the current is 2/ground, when NT is about 2 x 1 and 3/ground, and when the No of the ln layer is 1 to 5 x 1 and 7, and Epoxy (epoxy)
o Box) Values at the time of molding, and for reference, the state of change in the lifetime of minority carriers in Figure 5 is also illustrated. As is clear from this figure 8, the number of the phantom layer is 1 to 5×1
It becomes a green light-emitting element with high efficiency, with an average efficiency of 0.4% or more when 2×l and 6/block, and a maximum of about 0.7% when 2×l and 6/block, for example. This value has never been seen before and is thought to be the highest value in the world. For example, 1 was considered to have the highest luminous efficiency until now.
Published in 974 by JoumaloofChistaIGmwt
H27, page 191 (hereinafter referred to as Document-D), some green light-emitting elements become 0.7% when a current of 10 blood/ground flows (in the present invention, the current flow/swimming is reduced to 0.7%). 0 when flowing.
8% or more), compared to 10%
This is an increase above. This is illustrated in comparison with FIG. In addition, in this figure 9, 'Small present invention', 'Machine literature 1D
belongs to. The reason why the luminous efficiency improves in this way is that the change in No in the n-type gap layer becomes step-like, the lifetime of the minority carriers in the second n-layer is long, and the nitrogen concentration is high only in the phantom layer. It is thought that it is in the concentration.
Similarly to Figure 8, the lifetime and luminous efficiency of the minority carriers in the phantom layer with respect to the ND in the ln layer were investigated.
The result is as shown in Figure 10.

In addition, in this FIG. 10 as well, the luminous efficiency is when a current of 2 h/ground flows, and NT is about 2×1 p8/ground, and N of the second layer.
This is the value when changing 1 to 5 x 1 and 6/block, and when epoxy molded. As is clear from Fig. 10, when the first layer is 1 to 5 x 1 and 6/ground, even if the ND of the lnth layer changes to 1 to 5 x 1 and 7/ground, the average luminous efficiency is is 0.4% or more, and the lifetime of the minority carrier is also 330 (ship ec) or more. Moreover, in FIG. 10, it can be seen that there is a correlation between the lnth layer and the first layer. That is, if either the ln-th layer or the second layer changes, it becomes impossible to obtain a luminous efficiency of 0.4% or more on average. These luminous efficiencies are also likely influenced by the thickness of the first and second layers. For example, FIG. 11 shows the relationship between the luminous efficiency and the thickness of the first layer, and FIG. 12 shows the relationship between the luminous efficiency and the thickness of the lnth layer. From these two figures, it can be seen that when the thickness of the first layer is 10 m to 35 m and the thickness of the ln layer is 10 m or more, a luminous efficiency of 0.4% or more is exhibited. The reason for this is that, first, when the ln-th layer has a thickness of, for example, 10 rm or less, the crystal defect concentration propagated from the interface with the substrate is incompletely reduced, which adversely affects the crystal characteristics of the second layer.

Note that if it is too thick, the growth time will be long, making it unsuitable for mass production. In addition, when the first shoulder is 10 mm or less, one reason is that the emission region is less than the diffusion length of minority carriers, and another reason is that non-emission defects deposited at the interface of the first layer of the ln layer. This is not preferable because the effect tends to reach the light emitting region. Furthermore, if the first layer is thicker than 35 rms, the unnecessary region containing a high concentration of luminescent centers will increase, and the influence of the reabsorption effect will impair the device characteristics. Furthermore, when the correlation between the No. of the ln-th layer and the No. of the first layer was investigated, it was found that the relationship is as shown in FIG. 13.

That is, from this FIG. 13, the number of the first layer and the number of the lnth layer are
If the ratio is 1:3 to 20, the luminous efficiency is 0.4
It turns out that you can get more than %. For example, N between the lnth layer and the desk layer
. When the difference in the amount of light is almost eliminated, the non-emissive centers from the substrate propagate directly to the second layer, reducing the luminous efficiency. Furthermore, if the difference between the lnth layer and the phantom layer is too large, new distortion will occur in that portion. As is clear from the above description, N in the ln-th layer on the n-type GaP substrate.

1 to 5 x 1 and 7/mass, and the No. of the first layer to 1 to 5 x 1 and B
/constituted in a lump, and constituted so that the second layer contains nitrogen,
By forming a p-type GaP layer on top of this, the average
．． 4% or more of green light emitting elements are exhibited. Moreover, by configuring the numbers of the lnth layer and the second layer in a stepwise manner as described above, the yield can be improved. Note that the measurement of the lifetime of NT and minority carriers shown above is based on the EEE Transactions on
Electron DEvices, VoIED-24
inch. 7, pages 951 to 955.

In addition, No shown above refers to the net donor concentration, that is, No.
-N^, where N^ is the net acceptor concentration, that is, N^-N
.

It is. Furthermore, in the above example, sulfur (S) was used as the donor impurity of the n-type Gap substrate, but tellurium (T
e) Or it may be selenium (Se), p-type GaP
The acceptor impurity of the layer is not limited to Zn, but may be Cd.

Furthermore, the numerical values explained in the above embodiments are not limited thereto and can be changed in various ways.

[Brief explanation of the drawing]

FIG. 1 is a cross-sectional view schematically showing a liquid phase epitaxial growth apparatus used in the method of one embodiment of the present invention, and FIGS. 2 a to c are cross-sectional views showing how the growth boat of the apparatus shown in FIG. 1 is driven. , 3rd
Figures a and b are curve diagrams showing the temperature profiles of the growth boat and impurity evaporation source of the apparatus shown in Figure 1, and Figure 4 is a diagram showing the impurity profile of the light emitting device obtained by the method of the embodiment of the present invention. FIG. 5 shows the relationship between the lifetime of minority carriers and the donor concentration (No) of the n-layer. /Curve diagram showing the relationship between the lifetime of minority carriers and the number of the fox layer in the case of carp; A curve diagram showing the relationship of the minority carrier lifetime to ND, c is a curve diagram showing the relationship of the minority carrier lifetime to No of the n layer when the NT of the n layer is 2 × 1 pi 8 in the conventional literature 1B. In the curve diagram shown, d is the N of the n layer when N of the n layer is 2×1 and 8/ground in the conventional document-C. FIG. 6 is a curve diagram showing the relationship between the lifetime of the minority carrier and the lifetime of the minority carrier for No in the n layer. A curve diagram showing the mirror relationship between NT and the lifetime of minority carriers for the No of the layer, b is a curve diagram showing the relationship of the product of NT and the lifetime of minority carriers for the No of the n layer in conventional literature-A. , c is the conventional literature −
N in the n-layer of B. A curve diagram showing the relationship between the product of NT and the lifetime of minority carriers, d is N of the n-layer of conventional literature-C. Figure 7 is a curve showing the relationship between the product of NT and the lifetime of minority carriers, Figure 7 is a curve diagram showing the E-total ratio vs. wavelength, and Figure 8 is the relationship between the lifetime of minority carriers and the luminous efficiency vs. No of the first layer. FIG. 9 is a curve diagram showing the relationship between luminous efficiency and current density. FIG. 10 is a curve diagram showing the relationship between luminous efficiency and current density. Figure 11 is a curve diagram showing the relationship between the lifetime of minority carriers in the first layer and luminous efficiency, Figure 11 is a curve diagram showing the relationship between luminous efficiency and the thickness of the first layer, and Figure 12 is the thickness of the screen layer. FIG. 13 is a curve diagram showing the relationship between luminous efficiency and the ratio of No in the ln layer to No in the phantom layer. In FIG. 1 and FIG. 2, 11 is a reactor, 12
is a growth boat, 13a and 13b are heating devices, 14 and 24 are n-type GaP substrates, 14a and 24a are recesses for accommodating the substrates, 14b, 24b, I6b and 26b are lids,
15 and 25 are sliders, 16 and 26 are solutions, 17
and 27 are small holes, 18 and 28 are solution storage boats, 19
is an impurity evaporation source, 11a and 11'a are closed openings for supplying gas, and 11a is an opening □ for discharging gas. Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13

Claims (1)

[Claims] 1. In a green light-emitting element in which an n-type gallium phosphide layer and a p-type gallium phosphide layer are sequentially provided on an n-type gallium phosphide substrate, the net value of the n-type gallium phosphide layer on the substrate side The donor concentration was set to 1 to 5 x 10^1^7/cm, and
The net donor concentration on the type gallium phosphide layer side is 1 to 5×1.
The thickness of the gallium phosphide layer is 10 μm on the substrate side.
m or more, the thickness on the p-type gallium phosphide layer side is 10 to 35μ
A gallium phosphide green light-emitting device characterized by having a structure of m. 2. The net donor concentration (N_D_1) on the substrate side of the n-type gallium phosphide layer is configured to be 3 to 20 times the net donor concentration (N_D_2) on the p-type gallium phosphide layer side. A gallium phosphide green light-emitting device according to claim 1. 3. Claim 1, characterized in that only the n-type gallium phosphide layer on the p-type gallium phosphide layer side of the n-type gallium phosphide layer contains nitrogen.
The gallium phosphide green light-emitting device described in 2.