KR100878390B1 - Method of Growing InGaN-based Mutilayer Structure by Plasma-assisted MBE and Manufacturing ?-Nitride Light Emitting Device Using the Same - Google Patents

Abstract

InGaN-based multi-layer growth method according to the present invention comprises the steps of placing a substrate in the growth chamber of the PA MBE device having a nitrogen activator and preparing for growth of the group III nitride layer; Forming an InGaN-based multilayer structure by injecting Ga flux and In flux onto the substrate using the nitrogen activator to grow InGaN-based layers having different In compositions by providing Ga flux and In flux on the substrate, In the InGaN-based multilayer structure forming step, the In composition of the InGaN-based layer is changed by changing the RF power of the nitrogen activator while maintaining a constant flow rate of the incident Ga flux and In flux and the nitrogen activator supplied to the nitrogen activator. .

Nitride, Light Emitting Diode, MBE, RF Power

Description

Method of Growing InGaN-based Mutilayer Structure by Plasma-assisted MBE and Manufacturing III-Nitride Light Emitting Device Using the Same}

The present invention relates to an InGaN-based multi-layer growth method and a method for manufacturing a group III nitride light emitting device using the same, in particular, InGaN / GaN quantum well structure using a plasma molecular beam epitaxy (PA MBE) The present invention relates to a method for growing an InGaN-based multilayer structure quickly and reproducibly.

Currently, Group III nitride light emitting devices have been successfully applied to white light emission, full color display, automobile lighting, and the like. For mass production of Group III nitride light emitting devices operating in the blue-green and near-ultraviolet spectral region, metal organic vapor phase epitaxy (MOCVD) is mainly used.

Another epitaxial technique for growing a group III nitride light emitting device structure is a molecular beam epitaxy (MBE) using two different nitrogen activation treatment methods. The first of these uses high temperature cracking of ammonia in the substrate, similar to MOCVD, while the second involves the activation of molecular nitrogen in the plasma generated in a specially designed remote plasma source (nitrogen activator). Technology (ie PA MBE).

Common advantages of the two types of MBE are relatively low material consumption and low epitaxial growth temperature over MOCVD techniques. In addition to these economic advantages, the extremely low gas pressure in the growth reactor eliminates parasitic gas reactions, while the low growth temperature results in a Group III nitride compound (In _x Ga _1-x ) having a low dislocation temperature. It enables the growth of N, x having a high x value up to 1 and makes use of substrates with low decomposition temperatures (substrate of GaAs, ZnO, etc. as well as plastics or organic materials).

Important advantages of PA MBE are: (i) it can generate a flux of activated nitrogen (N *) (F _{N *} ) regardless of substrate temperature (Ts), and (ii) hydrogen in the epitaxial layer growth environment. And no ammonia. In addition to environmental safety, PA MBE makes it possible to eliminate the need for post-growth treatment (which is necessary for MOCVD) to activate Mg dopants. In addition, in PA MBE, even simpler design Knudsen cells and conventional UHV pumps can be used.

Over the years, insufficient nitrogen activation efficiency and low maximum growth rate (˜0.1 μm / h) with respect to PA MBE have been major barriers to PA MBE. In the late 1990s, however, this problem was successfully overcome by the development of a new type of RF inductively-coupled plasma nitrogen activator. Currently, most series of nitrogen activators supply F _{N * up} to a value corresponding to a maximum growth rate above 1 μm / h.

MBE growth of Group III nitride compounds under different stoichiometric conditions, ranging from nitrogen rich conditions (V / III> 1) to Group III source-rich conditions (V / III <1), has actually been realized. The former (nitrogen-rich condition) PA MBE is characterized by the formation of a nanocolumnar-shaped layer surface with a coarse layer surface, whereas the latter (a condition rich in group 3 sources) provides a flat surface. . Furthermore, Heying et al., In "Journal of Applied Physics" vol. 88, pp. 1855-1860 (2000), describe a technical window (ie, to obtain a flat surface at an atomic level without metallic microdroplets). , Range of group III flux change for a particular substrate temperature. This possibility of growing Group III nitride based compounds in Group 3 rich (N-limiting) stoichiometric conditions is a unique feature of PA MBE, which allows it to be distinguished from other technologies.

Indium (In) incorporation kinetics during the PA MBE process to form In _x Ga _{1 - x} N layers has been studied by several research groups. A pioneering paper by T. Botcher et al., Applied Physics Letters vol.73 pp. 3232-3234 (1994), reports experimental results on In incorporation at various In / Ga plus ratios. According to this, it is shown that the maximum possible value of In content (X _max ) is determined by the excess of N * flux over the Ga flux value (X _max = 1-F _Ga / F _{N *} ). Thus, in order to obtain a high In content, the Ga flux (F _Ga ) must be lower than the N * flux (F _{N *} ). It was also shown that the In adhesion coefficient α _In = 0.16 at Ts = 650 ° C., where α _In is the incorporation ratio for the provided In atoms. The adhesion coefficient α _Ga for Ga is 1 at this temperature. GaN has a higher binding energy compared to InN, which means that firstly all available Ga is incorporated and secondly In until it reaches stoichiometry. This idea has been identified and developed in C. Adelmann et al. Applied Physics Letters vol. 75 pp.3518-3520 (1999), which also demonstrates the negligible re-evaporation of In at Ts = 600 ° C. (Ie α _In = 1). According to Applied Physics Letters vol. 75 pp. 2280-2282 (1999) by MLO'Steen et al., The value of x decreases by more than one order with increasing Ts over a narrow temperature range of 590 to 670 ° C. One strong dependency was observed. Thus, in this paper, the importance of the re-evaporation process within the Ts range of interest (> 600 ° C.) was argued.

Some researchers have attempted to develop phenomenological models describing In incorporation. First, in a publication by H. Chen in the MRS Internet Journal Nitride Semiconductor Research vol. 6, 11 (2001), the authors of the article describe the incorporation of In into the InGaN alloys with (0001) and (000-1) polarities. Their experimental data were used to describe the dependence of In incorporation on the Ts and III / N ratios. They found that there were two different regions of stoichiometric conditions-the nitrogen-rich region and the group-III-rich region. In the former (nitrogen-rich region), when F _Ga + α _In F _In <F _{N *} (wherein F _In represents In flux), In incorporation is mainly determined by Ts and depends on the III / N ratio At constant In / Ga ratio, it shows a functional relationship that increases as the total Group III flux increases. In this case, x = α _{I n} F _In / (F _Ga + α _In F _In ) and α _In = constant.

_{In the} Group III rich region, defined as F _Ga + α _In F _In > F _{N *} , the authors demonstrated the existence of a stoichiometric limit, ie by observing that x decreases while the total Group III flux is increasing. , In incorporation (as shown first by Botcher et al.) Was determined by x = 1-F _Ga / F _{N *} . The authors suggest that if there is too much In, excess In will evaporate or form droplets. Unfortunately, these authors did not quantitatively determine the conditions for the realization of each of these processes.

Another theoretical approach describing In incorporation in In _x Ga _{1 - x} N was developed by DF Storm and published in the Journal of Applied Physics vol.89 pp.2452-2457 (2001), which is rich in nitrogen. It is suitable when growth proceeds only under conditions, i.e. when F _Ga <F _{N *} and the total metal flux F _Ga + F _In is not less than or too large than F _{N *} .

Note that in all the above studies, to change x, the F _{N *} flux remained constant and only the Group 3 flux and Ts changed.

The growth of InGaN-based quantum well (QW) heterostructures has been carefully studied by a research group led by K. Ploog and J. Speck. This is reflected in a paper published by O. Brandt in "Applied Physics Letters" vol.83 pp. 90-92 (2003), in which the authors control stoichiometric conditions to grow InGaN-based quantum well heterostructures. It is suggested to use. It was also shown that nitrogen enrichment conditions are more desirable for growth of the quantum well layer, while group III enrichment conditions should be used during barrier layer growth. In addition, the authors suggested using a relatively high growth temperature (> 600 ° C) to suppress surface segregation due to In evaporation from InGaN growth layers.

Growth by controlling the stoichiometry at high substrate temperature Ts provides a clear InGaN / GaN interface and suppresses surface segregation, resulting in an increase in the radioactive recombination efficiency of the QW structure. This was also confirmed by high resolution transmission electron microscopy (HRTEM), X-ray diffraction analysis (XRD) and low-temperature PL spectra studies. The minimum half width (FWHM) value of 95 meV for the PL band of 2.766 eV was measured at low temperature (5K). Moreover, using this method, LEDs with a wavelength of 480 nm with an output of 0.87 mW at a current of 20 mA can be manufactured, which is described by P. Waltrei et al. "Applied Physics Letters" vol. 84 pp.2748-2750 (2004). ) Was reported.

In order to control the stoichiometric conditions, K. Ploog et al. Changed the nitrogen flow rate from 1 sccm to 2 sccm, and J. Speck et al. Had a pair of Ga sources (twin Ga) with a 2-fold difference in Ga flux. source). The main disadvantages of the proposed stoichiometric condition control method are the complexity of the MBE equipment when using a pair of Ga sources (cells) and the rapid composition change due to the delay of the evaporation Knudsen cell. The main disadvantage in the case of the nitrogen flow regulation is the nonlinear dependence of the output activated nitrogen beam flux F _{N *} on the mass flow rate.

The present invention is to solve the above problems, an object of the present invention is to change the In composition quickly and reproducibly and to eliminate the complexity of the control of the Group 3 source flux or nitrogen flow rate, InGaN-based InGaN-based It is to provide a multi-layer growth method.

Another object of the present invention is to provide a method for manufacturing a high quality Group III nitride light emitting device using the InGaN-based multilayer structure growth method.

In order to achieve the above technical problem, the InGaN-based multilayer structure growth method of the present invention,

Placing a substrate in a growth chamber of a PA MBE device with a nitrogen activator and preparing for growth of a Group III nitride layer;

Forming an InGaN-based multilayer structure by injecting Ga flux and In flux onto the substrate using the nitrogen activator to grow InGaN-based layers having different In compositions by providing Ga flux and In flux on the substrate,

In the InGaN-based multilayer structure forming step, the In composition of the InGaN-based layer is changed by changing the RF power of the nitrogen activator while maintaining a constant flow rate of the incident Ga flux and In flux and the nitrogen activator supplied to the nitrogen activator. Let's do it.

According to the present invention, the In composition of the InGaN-based layer increases as the RF power of the nitrogen activator increases in the InGaN-based multilayer structure forming step. The activated nitrogen flux F _{N *} is linearly dependent on the RF power at a constant nitrogen flow rate. Thus, a change in RF power can linearly change F _{N *} over a wide range of N-limit growth rates. The range depends on (i) the output aperture design of the RF plasma cell, (ii) the total pumping rate in the growth chamber, (i) the nitrogen flow rate, and (i) the range of applied RF power, You can go beyond that.

According to a preferred embodiment, the substrate temperature is kept constant in the InGaN-based multilayer structure forming step. The method may further include growing a group III nitride buffer layer, such as a GaN buffer layer, on the substrate between the group III nitride layer growth preparation step and the InGaN-based multilayer structure forming step.

Preferably, in the InGaN-based multilayer structure forming step, the Ga flux (F _Ga ) and In flux (F _In ) which are constantly incident are selected to satisfy the following requirements: F _Ga + F _In > F _{N *} and F _Ga <F _{N *} (F _{N *} is an activated nitrogen flux). In this case, F _Ga limits the minimum growth rate of the In _x Ga _{1 - x} N layer and F _{N *} limits the maximum growth rate. Ga flux and In flux are simultaneously supplied to the growth surface of the InGaN based multilayer structure.

Preferably, the temperature Ts of the substrate is selected to provide a higher flux of In that evaporates from the In melt than the (incident) In flux F _In . Under conditions where the growth conditions used do not completely inhibit In incorporation into the InGaN-based layer (F _Ga ≥ F _{N *} ), the substrate temperature (Ts) selected as described above forms the formation of In droplets on the growth surface. To prevent it.

The substrate temperature Ts is the main factor governing the In incorporation efficiency α _In into the InGaN-based layer. Here In mixing efficiency (α _In ) is defined as the ratio of the incorporated In flux to the incident In flux. Increasing the substrate temperature (Ts) in the range of 590 to 650 ° C drastically decreases α _In , while increasing the F _{N *} / (F _Ga + F _In ) ratio increases α _In , although the effect is much smaller. Let's do it. Preferably, the substrate temperature Ts is kept constant during the growth of the entire InGaN based multilayer structure.

Under the conditions determined as described above, the composition of the In _x Ga _1-x N layer including the InGaN-based multilayer structure was varied from a minimum of x = 0 at a wide range (F _{N *} / F _Ga = 1, F _{N *} / (α _In F _In + F _Ga ) ≥1 for x = α _In F _In / (α _In F _In + F _Ga ) or F _{N *} / (α _In F _In + F _Ga ) <1 x = 1-F _Ga / F _{N *} range). In, Ga and N * fluxes are simultaneously supplied to the growth surface of each layer. The change in x (In composition) between the layers is controlled by quickly adjusting the RF power of the nitrogen activator only during growth interruptions at the layer interface.

The F _{N *} / (α _In F _In + F _Ga ) ratio also dominates the surface stoichiometry of the InGaN based layer. Thus, the method proposed in the present invention is based on the stoichiometry of the InGaN-based layer according to the exact deviation from N * / III * flux ratio = 1 (ie, F _{N *} / (α _In F _In + F _Ga ) = 1). It allows for changing theoretical conditions. For example, the barrier layer may be grown in group III rich conditions, while the N rich condition may be employed for the quantum well layer to improve the luminescence properties.

Embodiments of the present invention provide an InGaN / GaN single quantum well (SQW) structure, an InGaN / GaN multiple quantum well (MQW) structure-more broadly In _x Ga _{1 - x} N / In _y Ga _{1 - y} N (x> A method of growing an InGaN-based multi-layer heterostructure of another In composition profile, including the SQW structure or MQW structure of y) and InGaN / GaN heterostructure with controllable In distribution slope, and the like.

According to one embodiment of the invention, the first RF power of the nitrogen activator to form an In _x Ga _{1 - x} N / In _y Ga _{1 - y} N (x> y) single quantum well (SQW) structure forming a _y N barrier _layer, the active nitrogen flux in the in _y Ga ₁ was fed to the substrate; And forming a _x N well layer, _said first larger second while supplying active nitrogen flux in the RF power to the substrate on which the In _y Ga ₁ than the RF _{power-on y} N barrier layer, In _x Ga ₁ . In this case, the In _y Ga _{1 - y} N barrier layer may be GaN (or GaN: In), and the In _x Ga _{1 - x} N well layer may be InGaN.

According to another embodiment of the present invention, at a first RF power of the nitrogen activator to form an In _x Ga _{1 - x} N / In _y Ga _{1 - y} N (x> y) multi-quantum well (MQW) structure in _y Ga ₁ while supplying active nitrogen flux to the _substrate to form a _y N barrier layer; The first larger second while supplying active nitrogen flux in the RF power to the substrate on which the In _y Ga ₁ than the RF _power-y N barrier layer on the Ga _x In _1-x N to form a well layer; And alternately repeating the barrier layer forming step and the well layer forming step. In this case, the In _y Ga _{1 - y} N barrier layer may be GaN (or GaN: In), and the In _x Ga _{1 - x} N well layer may be InGaN. At least one of the first RF power and the second RF power may be changed. Accordingly, the In composition of the barrier layer or the well layer may be changed.

In order to achieve another object of the present invention, the method for producing a group III nitride light emitting device according to the present invention,

Placing a substrate in a growth chamber of a PA MBE device with a nitrogen activator and preparing for growth of a group III nitride layer;

Forming a first conductive group III nitride layer on the substrate;

Forming an active layer having an In _x Ga _{1 - x} N / In _y Ga _{1 - y} N (x> y) heterostructure on the first conductive group III nitride layer; And

Forming a second conductive group III nitride layer on the active layer,

In the forming of the active layer, by providing the activated nitrogen flux on the substrate using the nitrogen activator to form a GaGa and In flux on the substrate to form an InGaN-based multilayer structure of InGaN-based layers having different In composition ,

When the InGaN-based multilayer structure is formed, the In composition of the InGaN-based layer is changed by changing the RF power of the nitrogen activator while maintaining a constant flow rate of the incident Ga flux and In flux and the nitrogen activator supplied to the nitrogen activator.

According to an embodiment of the present invention, the InGaN-based multilayer structure may be an In _x Ga _{1 - x} N / In _y Ga _{1- y} N (x> y) single quantum well structure, in particular, an InGaN / GaN single quantum well structure Can be.

According to another embodiment of the present invention, the InGaN-based multilayer structure may be an In _x Ga _{1 -x} N / In _y Ga _1-y N (x> y) multi quantum well structure, in particular an InGaN / GaN multi quantum well structure. It may be a structure.

In the present specification, "GaN: In" contains a small amount of In in GaN (so that it is only impurity doping, that is, it is hard to see In composition), and has a composition substantially the same as "GaN". Thus, in this specification, "GaN: In" may be used synonymously with "GaN".

According to the present invention, the In composition can be changed quickly and reproducibly during growth of the InGaN-based multilayer structure. In addition, in order to change the In composition, the complexity of adjusting the flux of the group III source such as Ga or In can be eliminated, and growth of an InGaN-based multilayer structure having another In composition can be more easily realized. Accordingly, a high quality single quantum well structure or a multi quantum well structure having a clearer InGaN / GaN interface can be more easily manufactured. In addition, there is no need to use a pair of (2) Ga sources, which simplifies the MBE equipment used.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In particular, the following describes the present invention by illustrating the PA MBE growth of the bulk InGaN-based multilayer structure and single quantum well structure. However, embodiments of the present invention may be modified in various other forms, and the scope of the present invention is not limited to the embodiments described below. Embodiments of the present invention are provided to more completely explain the present invention to those skilled in the art. Shapes and sizes of the elements in the drawings may be exaggerated for more clear description, elements denoted by the same reference numerals in the drawings are the same element.

1 is a schematic cross-sectional view of a test sample structure grown to confirm the possibility of the present invention to control In incorporation by changing RF power of a nitrogen activator. All structures (see FIGS. 1A and 1B) were grown on the (0001) plane of sapphire substrate 101 using PA MBE equipment (using Riber's Compact 21T). The backside of the sapphire substrate 101 was previously coated with a 0.2 μm thick Ti layer 110 to provide radiative heating of the transparent sapphire substrate 101. This substrate 101 was annealed for 3 hours at a substrate temperature Ts of 800 ° C. in an isolated chamber after the oil component on the surface was removed in an organic solvent to evaporate most of the remaining contaminants. After transferring the substrate 101 to the growth chamber, the substrate was finally annealed for one hour at Ts = 800 ° C. The substrate temperature (Ts) was then lowered to 700 ° C. and the sapphire substrate (101) at 115 W RF power and 5 sccm N ₂ flow rate under activated nitrogen flux generated by a nitrogen activator (using Oxford Applied Research's HD-25 equipment). ) The surface was nitrided for 10 minutes. This nitrogen activator was equipped with an output aperture having 185 holes each having a diameter of 0.3 mu m. Through this nitriding treatment, an initial stage for growing the GaN layer was performed.

On the nitrided sapphire substrate 101 surface, the growth of the GaN buffer layer 103 was started at Ts = 750 ° C. at a slight Ga-rich stoichiometry at a growth rate of 0.4 μm / h without a separate nucleation layer. When the GaN buffer layer 103 was grown, the RF activator and the N ₂ flow rate of the nitrogen activator were 145 W and 5 sccm, respectively. The Ga flux supplied by the double-zone effusion cell is 0.65 μm / h (growth rate unit is used to define the flux—this value is at the given growth rate without atomic re-evaporation from the surface. Applicable). The dependence of Ga flux and In flux on cell temperature is shown in the graphs of FIGS. 2A and 2B, respectively. The GaN buffer layer 103 is about 0.8 mu m thick. After buffer layer growth was complete, the Ga and N * shutters were closed and the Ts and Ga fluxes were reduced to 637 ° C. and 0.22 μm / h (T _Ga = 828 ° C.), respectively.

An important point of the method proposed in the present invention is to select the level of RF power during different stages in quantum well (QW) growth. A study of the characteristics of the high brightness mode of the N plasma in the nitrogen activator finds a linear dependence of the intensity of the glowing plasma (I _g ) on the RF power of the nitrogen activator as shown in FIG. 3A. It was. On the other hand, it was found that the strength (I _g ) is nonlinearly dependent on the N ₂ flow rate as shown in FIG. 3B. From these measurements, RF power was chosen as a parameter for the adjustment of the N * flux. As shown in the graph of FIG. 4, the linear dependence of F _{N *} on the RF power was confirmed from the test measurement result of the dependence of the maximum GaN growth rate on the RF power.

Using this data, a working range of change of F _{N *} from 110 _W to 170 _W was selected to grow the barrier layer and the quantum well layer, respectively. The former RF power value 110W corresponds to a maximum GaN growth rate of 0.22 μm / h (see FIG. 4). Considering the complete incorporation of Ga flux at Ts (<650 ° C.) used, this means the realization of stoichiometric conditions of F _{N *} / F _Ga = 1: 1 during barrier layer growth. The upper limit 170W of the RF power corresponds to a maximum GaN growth rate of 0.64 μm / h, and implements a growth condition of F _{N *} / F _Ga > 1.

With the adjustment of the RF power described above, the temperature of the In cell was set to 870 ° C., which corresponds to a maximum InN growth rate of 0.91 μm / h (see FIG. 2). Through preliminary experiments by the inventors, it was found that the In incorporation efficiency (α _In ) was strongly influenced by the substrate temperature as shown in the graph of FIG. 5.

FIG. 6 illustrates a technique for growing an InGaN / GaN: In single quantum well (SQW) structure, and illustrates a shutter sequence and an RF power change. Two important features should be emphasized in this growth sequence:

First, the In cell is open simultaneously with the Ga cell during barrier layer and capping layer formation. At selected values of F _Ga and F _{N *} , set the N * / III flux ratio to be F _{N *} / F _Ga = 1: 1. At this time, In incorporation into the GaN layer is limited to a level of several percent. The relatively high Ts (637 ° C.) results in the re-evaporation of the entire In flux from the growth surface, which affects the surface reaction process of Ga atoms through the segregation effect, thus improving the quality of GaN: In. This was confirmed not only by observing in situ the relatively clear and striped RHEED pattern seen during barrier layer growth, but also by using a SEM as shown in FIG. This was confirmed by observation with ex situ.

Second, in order to obtain a clear interface for the quantum wells, a short period of time (approximately 20 seconds) of growth interruption was applied during the change in raising and lowering the RF power of the nitrogen activator (before and after the growth of the quantum well layer). All cell shutters, including the nitrogen shutters, remain closed during the growth interruption. Growth interruption after GaN: In barrier layer growth also allows the redeposition of residual In monolayer segregated on GaN: In.

After reaching the upper limit level of the RF power, the Ga and In shutters are opened for quantum well layer growth. At this stage, the effective Group 3 flux was smaller than F _{N *} as F _Ga + α _In F _In = 0.22 + 0.91 × 0.13 = 0.34 μm / h (see FIG. 5 for α _In ). That is, F _{N *} / (F _Ga + α _In F _In )> 1. This means that the In incorporation is limited only by Ts and the nominal thickness of the well well layer is 3 nm during 35 seconds of well well growth. In this case, the average In content may be calculated as x = F _In α _In / (F _Ga + α _In F _In ) = 0.35. Note that this x value calculation is accurate for the bulk layer but can vary by about 10% for thin film layers such as quantum well layers. This is due to strain effects as well as any transients in the cell and substrate temperatures after growth stops.

After the quantum well layer is complete, close the Ga, In and N cell shutters again and reduce the RF power to 110W, then grow the GaN: In capping layer to 110nm thickness for 30 minutes at the same Ts = 637 ° C. .

According to the embodiment described above, the desired In composition profile can be obtained quickly and reproducibly only by adjusting the RF power without changing process parameters such as In flux, Ga flux, and nitrogen flow rate. That is, since only the RF power is modulated to change the In composition, there is no time delay due to Ga or In injection cell control as in the prior art. Therefore, fast In composition change (and thus a distinct InGaN / GaN interface implementation) can be achieved. In addition, due to the apparent linear dependence of F _{N *} on RF power, the reproducibility of changes in In composition is excellent.

Photoluminescence (PL) spectra for single quantum well (SQW) structures at room temperature are characterized by a 480 nm line with an FWHM of 150 meV, as shown in FIG. 8. Practically, no PL emission is observed even at low temperatures from the GaN: N barrier layer (see FIG. 9), indicating a high quality capping layer that allows full collection of unbalanced carriers in the quantum well layer.

In order to identify the main points of the proposed method of growing the InGaN / GaN: In quantum well structure, two reference bulk layers were added to the growth conditions of the single well well layer (InGaN) and barrier layer (GaN: In). Accordingly. The PL spectrum at 20K corresponding to this bulk InGaN is shown in FIG. The PL spectrum of the 200 nm-InGaN layer exhibits a red shift of about 520 nm with respect to SQW PL (480 nm) compared to the 3 nm-InGaN / GaN single well well structure itself, and the PL intensity is about twice as high. The spectral position of the PL maximum value of this bulk InGaN layer fits well with the experimental data available in the literature. On the other hand, the PL spectrum of the GaN: In layer shows a negligible In incorporation (a few percent) in the barrier layer (see FIG. 9).

InGaN quantum well layer, as shown in Figure 7c and 7d shows the surface state of the nano-pillar from this can confirm the N-rich growth conditions. On the other hand, the low temperature GaN: In capping layer exhibits a much flatter surface condition with larger size grains with flat tops (see FIGS. 7E and 7F).

All experimental results presented above confirm the applicability of the proposed technique to the growth of InGaN-based multilayer structures (including quantum well structures). By the technique proposed by the inventor, it is possible to easily grow an InGaN-based multilayer structure having In content and stoichiometric change along the growth axis. The grown In _x Ga _{1 - x} N / In _y Ga _{1 - y} N (x> y) single quantum well (SQW) structure exhibits a clear room temperature luminescence in the green-red spectral range.

The InGaN-based multilayer structure growth method according to the above embodiment can be easily applied to form a high quality multi quantum well structure having a distinct interface. That is, the InGaN / GaN multi-quantum well structure may be obtained by repeatedly performing the growth of the barrier layer / quantum well layer of InGaN / GaN by the modulation of the RF power described above.

10 illustrates a method of growing an InGaN / GaN single quantum well structure (broadly In _x Ga _{1 - x} N / In _y Ga _{1 - y} N (x> y) single quantum well structure) according to an embodiment of the present invention. Sectional drawing which shows the process of manufacturing a group 3 nitride light emitting element using this.

First, a substrate 201 such as sapphire is placed in a growth chamber of a PA MBA device equipped with a nitrogen activator, and then annealing and / or nitriding are performed to prepare a group III nitride semiconductor growth. Thereafter, a GaN buffer layer 203 is formed on the substrate 201 as described above (see Fig. 10 (a)).

Next, the first cladding layer 205 is formed by growing the first conductive group III group nitride semiconductor layer on the GaN buffer layer 203 (see FIG. 10 (b)). For example, an n-type AlGaN layer may be grown using a PA MBE process to obtain the first clad layer 205.

Next, as described above, the active layer 207 is obtained by growing the InGaN / GaN single quantum well structure by adjusting the RF power of the nitrogen activator (see FIG. 10 (c)). For example, the active layer of the InGaN / GaN single quantum well structure on the first cladding layer 205 using the growth conditions mentioned with reference to FIG. 1 (substrate temperature of 637 ° C., constant Ga flux and In flux, constant nitrogen flow rate). 207 can be formed. In order to control the In composition in the active layer 207, a first RF power (eg, 110W) is applied to the nitrogen activator when the barrier layer is formed, and a second higher RF power (eg, 170W) when the quantum well layer is formed. Can be applied.

Thereafter, the second cladding layer 209 is formed by growing a second conductive group III nitride layer on the active layer 207 (see Fig. 10 (d)). For example, a p-type AlGaN layer may be grown using a PA MBE process to obtain a second clad layer 209.

Next, as shown in FIG. 10E, mesa etching is performed to expose a portion of the first cladding layer 105, and the exposed top surface and the second cladding layer 209 of the first cladding layer 205. The n-side electrode 211 and the p-side electrode 212 are formed on the upper surface thereof. As a result, a group III nitride light emitting device according to one embodiment is obtained.

In the above embodiment, the InGaN / GaN single quantum well structure is grown to obtain an active layer, but the present invention is not limited thereto. More broadly, it is possible to grow In _x Ga _{1 - x} N / In _y Ga _{1 - y} N (x> y) single quantum well structures with different In compositions by adjusting the RF power.

11 is a cross-sectional view for explaining a step of manufacturing a group III nitride light emitting device according to another embodiment of the present invention. In this embodiment, a multi-quantum well structure is grown to obtain an active layer. Also in this embodiment, during the formation of the active layer (when growing a multi-quantum well structure), other process parameters (substrate temperature, Ga flux, In flux, nitrogen flow rate, etc.) other than RF power are kept constant and only the RF power is adjusted to In composition. Get a change.

First, as shown in Fig. 11A, a GaN buffer layer 203 and a first cladding layer 205 such as n-type AlGaN are formed on a substrate 201 such as sapphire. Thereafter, an active layer 270 of InGaN / GaN multi-quantum well structure is formed on the first cladding layer 205, and a second cladding layer 209 such as p-type AlGaN is formed thereon (FIG. 11 ( b)). After the semiconductor layer stacking is completed, a mesa etching process and a process of forming electrodes 211 and 212 are performed as shown in FIG. 11C.

The active layer 270 may have a stacked structure in which the GaN quantum barrier layer 270a and the InGaN quantum well layer 270b are alternately repeated. In order to obtain such an InGaN / GaN heterostructure, the growth method described above with reference to FIG. 1 is used. By adjusting RF power while maintaining other process parameters such as substrate temperature, Ga, In flux, and mass flow rate constant, rapid and sharp changes in the In composition in the active layer can be obtained. For example, a relatively low first RF power (e.g., 110W) and a higher second RF power (e.g., with Ga flux, In flux, and N ₂ flow rate are maintained constant (substrate temperature is maintained at 637 ° C). , 170W) are alternately applied to obtain an InGaN / GaN multi quantum well structure set (more broadly, In _x Ga _{1 - x} N / In _y Ga _{1 - y} N (x> y) multi quantum well structure). Can be. In this case, a barrier layer is formed when the first RF power is applied, and a well layer is formed when the second power is applied.

The present invention is not limited by the above-described embodiment and the accompanying drawings, but is intended to be limited by the appended claims, and various forms of substitution, modification, and within the scope not departing from the technical spirit of the present invention described in the claims. It will be apparent to those skilled in the art that changes are possible.

1 shows an InGaN / GaN heterostructure obtained by a growth method according to the present invention, and FIG. 1 (a) is a cross-sectional view showing a hetero structure having a single quantum well (SQW) InGaN, and FIG. It is sectional drawing which shows the heterostructure which has a bulk InGaN layer.

FIG. 2A is a graph showing the dependency of group 3 flux on Ga ejection cell temperature, and FIG. 2B is a graph showing the dependency of group 3 flux on In ejection cell temperature.

FIG. 3 (a) is a graph showing the intensity vs. the RF power of the plasma formed inside the nitrogen activator, and FIG. 3 (b) shows the strength vs. the N ₂ mass flow rate of the plasma formed inside the nitrogen activator. It is a graph.

4 is a graph showing the maximum growth rate of the GaN layer vs. the RF power of the nitrogen activator at a constant N ₂ flow rate of 5 sccm.

FIG. 5 is a graph showing In incorporation efficiency (α _In ) into the InGaN layer as a function of substrate temperature T _S and flux ratio (N * / III = F _{N *} / (F _Ga + F _In )).

6 is a diagram illustrating a growth shutter sequence for growing an InGaN / GaN single quantum well structure according to one embodiment of the present invention.

7 is a cross-sectional view and a plan view SEM pictures of the InGaN / GaN structure according to the present invention,

7A and 7B are cross-sectional and planar SEM photographs of a single quantum well InGaN / GaN structure,

7 (c) and 7 (d) are cross-sectional and planar SEM images of an InGaN / GaN structure having bulk InGaN formed on a GaN buffer layer under growth conditions of the single quantum well.

7 (e) and 7 (f) are cross-sectional and plan view SEM photographs of the GaN: In capping layer, respectively.

8 is a graph showing the room temperature emission spectrum (excited by He-Cd laser (200 nm)) of a single quantum well structure grown according to the present invention.

FIG. 9A is a graph showing low-temperature emission spectra of an InGaN / GaN single quantum well and a bulk InGaN layer (200 nm thick) formed on a GaN buffer layer according to growth conditions of the single quantum well, and FIG. ) Is a graph showing the low-temperature emission spectrum of the GaN: In cap layer.

FIG. 10 is a cross-sectional view illustrating a process of manufacturing a group III nitride light emitting device according to one embodiment of the present invention. FIG.

FIG. 11 is a cross-sectional view illustrating a process of manufacturing a group III nitride light emitting device according to another embodiment of the present invention. FIG.

<Description of the symbols for the main parts of the drawings>

101: sapphire substrate 103: GaN buffer layer

104: bulk In _x Ga _{1 - x} N layer 105: GaN: In barrier layer

107: InGaN well layer 109: GaN: In capping layer

110: Ti layer

Claims (16)

Placing a substrate in a growth chamber of a PA MBE device with a nitrogen activator and preparing for growth of a group III nitride layer; Forming an InGaN-based multilayer structure by injecting Ga flux and In flux onto the substrate using the nitrogen activator to grow InGaN-based layers having different In compositions by providing Ga flux and In flux on the substrate, In the InGaN-based multilayer structure forming step, the In composition of the InGaN-based layer is changed by changing the RF power of the nitrogen activator while maintaining a constant flow rate of the supplied Ga flux and In flux and the nitrogen activator. InGaN-based multilayer structure growth method characterized in that. The method of claim 1, InGaN-based multilayer structure growth method characterized in that the In composition of the InGaN-based layer is higher as the RF power of the nitrogen activator in the step of forming the InGaN-based multilayer structure. The method of claim 1, InGaN-based multilayer structure growth method characterized in that the substrate temperature is kept constant in the step of forming the InGaN-based multilayer structure. The method of claim 1, And growing a Group III nitride buffer layer on the substrate between the growth preparation step of the Group III nitride layer and the step of forming the InGaN-based multilayer structure. The method of claim 1, In the InGaN-based multilayer structure forming step, the Ga flux (F _Ga ) and In flux (F _In ) that is constantly incident is selected to satisfy the following conditions: InGaN-based multilayer structure growth method: F _Ga + F _In > F _{N *} and F _Ga <F _{N *} , where F _{N *} represents the activated nitrogen flux. The method of claim 1, In the InGaN-based multi-layer structure forming step, the temperature of the substrate, InGaN-based multi-layer growth method characterized in that it is selected to provide a higher flux of In evaporating from the In melt than the incident In flux. The method of claim 1, In the step of forming the InGaN-based multi-layer structure, the growth method of the InGaN-based multilayer structure characterized in that the growth is stopped at the interface of the grown InGaN-based layer, and the RF power of the nitrogen activator is adjusted during the growth stop. The method of claim 1, The InGaN-based multilayer structure forming step includes growing an In _x Ga _{1 - x} N / In _y Ga _{1 - y} N (x> y) single quantum well structure. The method of claim 1, The InGaN-based multilayer structure forming step comprises the step of growing an InGaN / GaN single quantum well structure. The method of claim 1, The InGaN-based multilayer structure forming step includes growing a multi-quantum well structure of In _x Ga _{1 - x} N / In _y Ga _{1 - y} N (x> y). . The method of claim 1, The InGaN-based multilayer structure forming step comprises the step of growing an InGaN / GaN multi-quantum well structure. The method of claim 1, The InGaN-based multilayer structure forming step, Forming _y N barrier _layer, the nitrogen activator of claim 1 while supplying active nitrogen flux in the RF power to the substrate of the In _y Ga _1; And The first and the activated nitrogen flux in the larger claim 2 RF power than the RF power supplied to the substrate on which the In _y Ga _{1 -} forming a _x N well layer _{- y} N barrier In _x Ga ₁ on the layer InGaN-based multilayer structure growth method characterized in that. The method of claim 1, Forming _y N barrier layer _- In _y Ga ₁ while supplying active nitrogen flux to the substrate 1 at the RF power of the nitrogen activator; Forming an _x N well layer of the first larger and second supplying active nitrogen flux in the RF power to the substrate on which the In _y Ga ₁ than the _RF-power-y N barrier layer on the In _x Ga _1; And InGaN-based multi-layer growth method comprising the step of repeatedly performing the barrier layer forming step and the well layer forming step. Placing a substrate in a growth chamber of a PA MBE device with a nitrogen activator and preparing for growth of a group III nitride layer; Forming a first conductive group III nitride layer on the substrate; Forming an active layer having an In _x Ga _{1 - x} N / In _y Ga _{1 - y} N (x> y) heterostructure on the first conductive group III nitride layer; And Forming a second conductive group III nitride layer on the active layer, In the forming of the active layer, by providing the activated nitrogen flux on the substrate using the nitrogen activator to form a GaGa and In flux on the substrate to form an InGaN-based multilayer structure of InGaN-based layers having different In composition , When forming the InGaN-based multilayer structure, changing the In composition of the InGaN-based layer by changing the RF power of the nitrogen activator while maintaining a constant flow rate of the supplied Ga flux and In flux and the nitrogen activator. A method for manufacturing a group III nitride light emitting device. The method of claim 14, The InGaN-based multilayer structure is a group III nitride light emitting device manufacturing method characterized in that it is formed to have an In _x Ga _{1 - x} N / In _y Ga _{1 - y} N (x> y) single quantum well structure. The method of claim 14, The InGaN-based multilayer structure is a group III nitride light emitting device manufacturing method characterized in that it is formed to have an In _x Ga _{1 - x} N / In _y Ga _{1 - y} N (x> y) multi-quantum well structure.

Images