Open Access
Add to BibSonomy
Abstract
Within this work, we studied InGaN QWs with nominally 17% InN mole fraction grown within an 80 × 80 μm area with local misorientation angle change from 0.3° to 3.2°. We observed a significant improvement of the photoluminescence intensity for the area with misorientation above 1.5°, which we attribute to the quenching of nonradiative recombination processes. From the structural point of view, the increase of the misorientation angle above 1.5° is accompanied by the improvement of the morphology of the sample and quality of the quantum wells observed through atomic force microscopy and transmission electron microscopy. We show that the structural and emission qualities in high-InN- mole fraction layers can be improved just by increasing the misorientation angle of the substrate and that the improved qualities are preserved even for large misorientation angles.
© 2021 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Nitride material system is extremely important in the field of the optoelectronic emitters. It allows, in theory, to produce quantum wells (QWs) emitting in the whole visible spectral region and beyond, and at the same time keeping a direct bandgap of the semiconductor material. In practice, the expansion of the device emission range from the initial blue-violet [1–6] region was one of the most important challenges in the development of the nitride semiconductor technology. This includes moving to both: shorter wavelengths, with current records in the UVC region [7–11], and longer wavelengths spanning through green [12–17] even to red [18–24] emission region. The fabrication of the efficient III-nitride optoelectronic devices with high InN mole fraction in the active region is a demanding challenge, due to the mismatch of the lattice constant between the binary nitride materials – GaN, InN and AlN [25,26] – and also often used foreign substrates. With the increasing difference in the composition of the ternary materials used to build heterostructures, a significant strain appears in the system, leading to a deterioration of the material quality and possibly phase separation in the InGaN layers [27]. Additionally, because III-nitrides crystallize in the wurtzite structure, the strain leads to the appearance of piezoelectric fields in layers grown in (0001) c-plane direction [28]. As a consequence, the light emission from the QWs is subject to Quantum Confined Stark Effect (QCSE) [29,30] which reduces the radiative recombination rate and increases the emission wavelength. Another aspect which makes the growth of III-N structures difficult is the significant difference in the optimal growth conditions of the different binary materials. Although high quality green devices have been already reported and are commercialized, the optimal way of growing high InN mole fraction layers is still not widely known and the studies are still continued.
Many groups argue, that a good solution for fabrication of green QWs is the usage of semipolar and nonpolar planes [31–33]. The change of the growth direction not only reduces the electric fields present in the structure, but also can improve In incorporation in the InGaN layers thanks to the strain reduction [34]. The latter applies e.g. to (20-21) and (10-10) planes [35]. However, currently the semipolar and nonpolar substrates are still expensive and are not easily available. The importance of the detailed structure of the sample surface in an atomic level and its influence on the growth process can be observed also on GaN crystals with a misoriented surface (surface plane is misoriented with respect to the atomic planes by a certain angle, called misorientation, miscut or off-cut angle). The changes of morphology and In incorporation into InGaN layers grown on misoriented substrates were observed for + c-plane (Ga-polar) [36,37], m-plane [38,39], semipolar plane [40] and N-polar plane [41,42] substrates. It was even shown that a “double miscut” may be preferred over a standard miscut approach towards one of the main crystallographic directions [43]. In case of growth of InGaN layers on c-plane GaN, which is the most popular substrate, the misorientation angle influence was reported by Tian et al. [44]. The authors show that the growth quality and sample morphology is worsened when the growth temperature is reduced down to 688°C, but the good morphology can be restored by increasing the substrate misorientation angle, from 0.2° to 0.48°. A more recent work [45] shows a similar result also for 20% InN mole fraction InGaN layers but in wider range of misorientation – up to almost 1°.
In this work, we intend to gain a deeper insight into the above phenomenon by looking how structural and optical quality can be improved by growth on substrate with substantially increased off-cut angle. To enable an easy comparison, we fabricated a wide range of misorientation angles on one substrate within an 80 × 80 μm area. We use a typical structure optimized for blue wavelength but we reduced the QWs growth temperature to 700°C. We report a significant increase of the emission intensity of QWs with nominally 17% of InN mole fraction observed for higher local misorientation of the substrate – from 1.5° to 3.2°. This modification is coexisting with improved morphology of the sample, as well as improved structure of the QWs.
2. Sample fabrication and experimental methods
2.1 Fabrication of the sample
Within this experiment, we decided to study InGaN QWs with high InN mole fraction grown on a substrate with varying misorientation angle. To do this, we patterned a GaN substrate to prepare an 80 × 80 μm square with the local misorientation value ranging from 0° to 4°. The modification of the surface shape was obtained thanks to the multilevel photolithography process of a positive photoresist, which resulted in a 3D structure after the resist development. Next, the sample with the structuralized photoresist was dry-etched in an Inductively Coupled Plasma Reactive Ion Etching system which transferred the shape onto the bulk GaN substrate [46]. It needs to be emphasized that thanks to a small size of the test structure, the local conditions of the layer growth at different studied points can be safely assumed to be the same (e.g. temperature of the substrate or mixture of gasses above the crystal surface).
Next the patterned sample was used for MOVPE growth in an Aixtron closed-coupled-showerhead reactor. The intended structure of the epitaxial layers, presented in Table 1, consists of two staggered InGaN QWs separated by GaN barriers. The staggered QWs were used in this study because of the expectation that they will reduce the QCSE and enhance the spontaneous emission rate as well as the optical gain in high InN mole fraction QWs [47,48]. Below the active layers, we incorporated layers mimicking the bottom waveguide and cladding layer of laser or superluminescent diodes. The design used in this experiment is based on a structure optimized for emission in the blue region (450-460 nm), but the temperature of the growth of the deep part of the QW was reduced to 700°C in order to increase incorporation of indium atoms.
Table 1. Intended structure of the studied sample.
2.2 Experimental methods
For studying the detailed shape of the fabricated sample we used a confocal microscope (3D laser profiler) Keyence VK 9510 equipped with a 50x objective giving a lateral resolution below 0.3 μm. The nominal vertical resolution was 0.01 μm, however this value was increased to about 0.07 μm due to the vibrations in the system.
The microphotoluminescence mapping of our sample was performed using a system described in [46]. The excitation light-source was a continuously operated laser diode emitting at 375 nm and the diameter of the excitation spot on the sample was 2.5 μm (obtained through a 0.42 NA objective). The sample was placed on a motorized stage and the photoluminescence spectra were studied by a 560 mm long spectrometer equipped with a 200 grooves/mm grating (blazed angle of 430 nm). We examined the sample under low and high excitation power density of around 0.12 kW/cm2 and 12 kW/cm2. This system was used for the room temperature measurements presented in Section 4 as well as low temperature measurement shown in Fig. 5.
The micro time-resolved photoluminescence (μTRPL) was performed on a system equipped with a picosecond pulsed Ti:Sapphire laser [46]. For excitation we used a second harmonic at the wavelength of 375 nm with the frequency of 4 MHz and the peak excitation power density of around 5.7 MW/cm2. We used a 0.42 NA objective for focusing the exciting light into a spot with 2.5 μm diameter. The decay of the system response was around 1.2 ns.
The sample was also studied by Atomic Force Microscopy (AFM) using an MFP-3D Origin, Asylum Research system. Measurements were performed with an OLYMPUS AC160TS-R3 cantilever with the nominal tip radius of 7 nm.
Scanning Transmission Electron Microscopy (STEM) was performed using a JEOL JEM-F200 system equipped with an Energy-Dispersive X-ray spectroscopy (EDX) unit (JEOL JED-2300T). The TEM specimens were fabricated by focused ion beam etching. The acceleration voltage for the STEM observation was 200 keV.
The supplementary photoluminescence measurements presented in Section 7 were performed using a system which allows CW excitation by both 375 nm and 320 nm lasers with a spot size having diameter of around 50 μm. In this system we gathered data at room temperature and cryogenic temperature of 20 K.
3. Mapping of the local misorientation angle
After the epitaxial growth, the sample was first studied by a confocal laser microscope in order to check the final shape of the surface, Fig. 1 (a). Based on the obtained data, the local misorientation after the growth was estimated, Fig. 1 (b). The calculation included the initial value and direction of the substrate obtained by X-ray diffraction before substrate patterning and epitaxy – 0.04° in the x direction and 0.68° in the y direction. We were able to obtain a wide range of misorientation from around 0.3° to above 3.2°. The corner corresponding to the highest intended misorientation shows a plateau, which was also observed on a similar sample in previous experiments [46,49]. It most probably results from the expansion of the c-plane on the tip of the 3D structure during the epitaxial growth. In Fig. 1 (b) the black markers show the positions of the measurements described in the further part of this work, while the gray crosses indicate additional studied points presented in Supplement 1. The local misorientation angle value of points 1, 3, 5 is also written explicitly in Table 2. The points 1, 3, 5 were initially chosen searching for a dataset with uniform step of off-cut angle. Additional studied points (1’, 5’) along the white, dashed line were added later due to the requirement of the sample preparation for Scanning Transmission Electron Microscopy study.
Fig. 1. (a) Three-dimensional structure of the studied sample measured by a laser microscope and (b) the estimated map of the local misorientation of the sample. The small hills observed in (a) and white squares in (b) correspond to SiN markers on the sample. The black markers represent positions at which the point measurements were done (together with assigned numbers). Gray crosses correspond to additional point measurements included in Supplement 1.
Table 2. Misorientation of the studied point regions, as presented in Fig. 1 (b).
4. Micro photoluminescence study
Next, the structure was studied by microphotoluminescence mapping under room temperature using a high resolution system described in Section 2.2. The maps of the peak emission wavelength presented in Fig. 2 (a) and (b) show an emission wavelength shift of 10 nm to 15 nm. This shift is smaller than in case of our previous experiments, performed for samples with lower InN mole fraction in QWs, and also smaller than the full width at half maximum (FWHM) of the spectra (above 32 nm up to even 50 nm; comment in the next paragraph). We also notice that under low excitation, the low misorientation area shows a spotted wavelength distribution. This suggests the presence of a structure having characteristic dimensions smaller than the resolution of the imaging method. Interestingly, under low excitation the area with longest measured emission wavelength does not correspond to the lowest misorientation, but has a shape of a band corresponding to a narrow range of misorientation around 1.3°-1.5°, which is shown as dashed lines in Fig. 2 (a) and (c). The area with higher misorientation than the band shows a continuous wavelength landscape. Furthermore, it was observed that the peak intensity of photoluminescence, Fig. 2 (c) and (d), strongly increases for the area of higher misorientation with the border roughly following the mentioned long-emission-wavelength band in Fig. 2 (a). This might suggest a change of the growth mechanism with the local misorientation. The improved uniformity of the peak wavelength and peak intensity maps measured under higher excitation, Fig. 2 (b) and (d), is probably a result of the filling of the potential minima by the excited carriers.
Fig. 2. Comparison of the results of the PL mapping of the studied sample. The maps show the central emission wavelength (a,b) and peak intensity (c,d). The measurements were done under low excitation of around 0.12 kW/cm2 for (a,c) and under higher excitation of around 12 kW/cm2 for (b,d). The dashed lines mark the band with the long peak wavelength from (a).
We would like to point out, that the shift of emission spectrum with misorientation in the studied sample is different from what was observed in our previous experiments [46,49] - for samples with smaller InN mole fraction (and also higher growth temperature). A comparison is presented in Fig. 3. In case of the sample with lower InN mole fraction in QWs, the shift of the spectral maximum with misorientation is significantly larger than the FWHM of the spectrum. As such, the change of the misorientation causes a significant modification of the emission spectrum. Contrary to this, in case of the sample studied in this work, the spectra are quite broad and the shift of maximum is smaller than the FWHM. We also see that the shape of the PL spectra changes with the excitation power density. The comparison of the spectra suggests that in this study the InN mole fraction in QWs is weakly influenced by the local misorientation angle. At least, the effect is rather subtle, if compared with our previous experiment with the sample with lower InN mole fraction.
Fig. 3. Comparison of the PL spectra acquired for the studied sample under low (a) and high (b) excitation, measured in points 1,3,5. The dashed lines depict the PL spectra measured for a similar sample, but with lower nominal InN mole fraction in the QWs. The values of local misorientation angle corresponding to each measurement point are written in the picture.
5. Time-resolved μPL measurement
We continued the study by μTRPL under room temperature in chosen points of the sample. The sample was locally excited with the peak excitation power density of around 5.7 MW/cm2. According to our estimations, the resulting initial carrier density within the μTRPL study – 1.7 · 1018 cm-3 – is similar to the excited carrier density in the CW-μPL experiment obtained for 12 kW/cm2 excitation – 1.8 · 1018 cm-3. The estimation of carrier density under pulsed excitation was done assuming that the number of created carriers is equal to the number of absorbed photons (derived from the excitation pulse energy density) [50]. We assumed the absorption coefficient in the quantum wells of 105 cm-1 and sum of well width of 3.8 nm.
As shown in Fig. 4 (a), in the low misorientation region – point 1 – we observed a very fast emission decay with the central emission wavelength of around 465 nm. But in region with high misorientation, point 5 presented in Fig. 4(b), the spectra consist of two contributions: one similar to the Fig. 4 (a) case and the other positioned at longer wavelengths (around 480 nm) and characterised by much longer time decay. The source of the 480 nm wavelength emission is not obvious. As it starts dominating over 5 ns after reaching the maximal intensity, it should be correlated with emission at lower carrier density – e.g. Figure 3 (a). Photoluminescence decays measured in other areas of the sample are presented in Supplement 1 in Fig. S2. The changes of the shape of the TRPL spectra with time are also presented in Fig. 4 (c) which shows TRPL spectra integrated over 1.7 ns windows at different time delays. We can observe an initial domination of the short wavelength spectral component or coexistence of two peaks and a transition to the domination of the long wavelength component at later time. To observe the two components in more details, we fitted the spectra by two gaussian distributions and estimated the position of the maxima of both contributions – which are shown in Fig. 4 (d). We observed that the long wavelength contribution shows a bigger peak wavelength shift with time, which reflects the shift with carrier concentration in the QWs. The observed blue-shift with increasing carrier concentration can be related either to band-filling effect or screening of polarization field and QCSE. Such a difference in behaviour might reflect the presence of two types of light emitting areas in the active region. Next, we will concentrate next on analysing the short-wavelength range of TRPL spectra.
Fig. 4. Photoluminescence decay measured for two points: 1 (a) and 5 (b). (c) Time evolution of the spectra from (b) integrated over 1.7 ns time windows. The middle time of each window is written in the graph. (d) Wavelength shift with time of the two components obtained from the decay shown in (b) by fitting of two gaussian functions to the integrated spectrum for a given time window. The time values on x axis are in fact middle time of each window.
The faster TRPL decay observed for point 1 when compared with point 5, Fig. 4 (a) and (b), suggests a more effective recombination process. The difference can be observed more clearly in Fig. 5 (a), which shows the intensity decays integrated over a spectral range from 455 nm to 465 nm, for both points. This wavelength range was chosen as it corresponds to a considerable intensity of the signal while still having higher energy than the slow-decaying contribution around 480 nm. Figure 5 (a) shows around two times faster intensity decay, with lifetimes, estimated as 1/e intensity drop, equal to 0.40 ns for point 1 and 0.81 ns for point 5. The increased recombination rate should not be correlated with the radiative recombination as in Fig. 2 (c) and (d) we see a significantly lower PL peak intensity of point 1 when compared with point 5. We believe that the initial decay is dominated by the non-radiative recombination, which agrees with the PL results. This understanding is supported by additional PL maps of the studied sample presented in Fig. 5 (b) and (c), which were obtained in the same system introduced earlier and under room temperature and 4 K. The excitation power density was around 12 kW/cm2. For low temperatures (4 K), the peak intensity in the area of point 1 shows around 0.5 intensity of the normalized map, while under room temperature (293 K) the intensity drops to around 0.1. This difference is most probably related to the activation of non-radiative recombination processes with temperature. Additionally, we studied the decay in point 1 for several wavelength windows and we observed a consecutive increase of the lifetime with the increasing central wavelength of the spectral integration window. This shows that carriers characterized by higher energies are more easily captured by the non-radiative recombination centers present in our sample. Based on the 293K and 4K μPL maps, we also estimated the room temperature IQE (assuming that under 4K the IQE of the structure is 100%). The IQE value changes with the off-cut angle, in particular for points 1, 3 and 5 the IQE was estimated as 9%, 40% and 35%, respectively. The high IQE obtained for point 3 is related to the expansion of the high intensity area observed in Fig. 5 (a) and (b).
Fig. 5. (a) Comparison of the TRPL decays measured at 293 K in points 1 and 5, shown in Fig. 4, and integrated over a wavelength range from 455 nm to 465 nm. The slower recombination observed for point 5 is most probably related to suppressed non-radiative recombination processes. This is supported by the comparison of PL maps measured under 4 K (b) and 293 K (c), showing an increase of the relative peak intensity in low miscut area at low temperature.
6. Atomic force microscopy
The morphology of the top of the sample was studied through Atomic Force Microscopy (AFM) in the selected areas, Fig. 6. We observe that for the area with low misorientation, which is characterised by low emission intensity, the surface is rough with many V-pits. The monolayer steps often coalesce into step bunches. The situation is changed for the area with higher misorientation. It was not possible to distinguish the monolayer steps due to the geometric restrictions of the AFM tip. We used an cantilever with the tip radius of 7 nm, while the distance between the monolayer step edges for points 1, 3, 5, and 5’ (estimated from off-cut shown in Table 2) is 37.1, 10.5, 5.9 and 6.6 nm, respectively. But still, in case of point 5, we observed a smooth morphology suggesting an improved structural quality, while in point 5’ we noticed the appearance of V-pits. In general, the distribution of V-pits seems to be random and, what is most important, the areas between them are characterised by smooth surface with small roughness. This improvement correlates with the higher peak intensity of the PL signal from previous measurements. The scan measured for point 5’ is characterised by highest V-pit density when compared with other AFM scans made for regions of off-cut angle above 1.5° (more results are presented in Supplement 1).
Fig. 6. AFM images obtained for the different studied areas on the sample. The difference in roughness for different misorientation regions can be observed. A full set of AFM scans is presented in the Supplement 1. The RMS roughness values of a),b),c),d) are 0.402, 0.947, 0.183, and 0.639 nm, respectively.
7. Transmission electron microscopy
In the last step of the experiment, the sample was studied by Scanning Transmission Electron Microscopy (STEM). From the studied sample, by the use of Focused Ion Beam (FIB), a filament was cut out along the line connecting two points: 1 and through 5, as it is presented in Fig. 1 (b) and Fig. 7 (a). After the filament was cut, the STEM images were taken in 8 points, as marked in Fig. 7 (a). We collected both Bright Field and High-Angle Annular Dark-Field (HAADF) images at the same positions. In each point, two neighbouring images of lower magnification were made for the purpose of monitoring the QW shape in a wider range. Also, one image with the highest magnification was taken for examination by energy-dispersive X-ray spectroscopy (EDX). In this paper, we present the STEM images of three chosen points with the most pronounced changes. Additionally, all the acquired data is available in the Supplement 1.
Fig. 7. (a) The SEM image of the sample during preparation for the STEM studies. The STEM HAADF images were obtained in all the marked points. The orange stars are points presented in (b), (c), (d), while images obtained for green crosses are presented in the Supplement 1. Pairs of images were taken in the HAADF mode, next the image data was stitched, cropped and presented in colour scale instead of grayscale, to emphasize the change in the quantum well shape and uniformity with misorientation. In the used color scale, dark blue represents dark areas of the dark field image (smaller atomic number) while red represents brightest areas of the dark-field image (larger atomic number).
Each two lower-magnification images were exported as data files, stitched together, cropped and plotted in colour scale. Figure 7 presents the STEM images for three chosen points on the sample. The colour scale allows to clearly observe the differences between the studied areas. Unfortunately, for all cases, the top quantum well is noticeably thinner to what was intended, suggesting possibility of thermal decomposition after growth. In case of the low misorientation, both QWs are not continuous. Small islands with contrast suggesting higher InN mole fraction and areas with the InN mole fraction similar to the waveguide composition can be seen. For the medium misorientation, point 3, the QWs seem to be more uniform with less pronounced gaps in the bottom QW. The image obtained for the highest misorientation shows that the bottom QW is quite uniform and continuous. The top QW, although still thin, shows less pronounced gaps. The pronounced roughness observed for point 1 may promote defect or impurity incorporation during the epitaxial growth. This would lead to increased amount of non-radiative recombination centres agreeing with the observations from μTRPL measurements. It should be noted that the dependence of impurity incorporation on substrate misorientation angle can be observed also without pronounced change in the surface morphology. For example, in the work of Jiang et al. [51] the changes of carbon incorporation into p-type GaN are reported.
The reason for the observed change in the growth quality of the QWs at various misorientations can be related to the dependence of surface supersaturation (during growth) on the substrate misorientation. Similar experiments were performed by Bryan et al. [52] for AlN layers and Liu et al. [45] for 4-6 nm InGaN layers (not staggered). Both groups showed a significant change in the layer morphology from a 2D island growth to a step-flow growth, or even formation of step bunches, with increasing misorientation angle. This transition is explained in accordance to the Burton-Cabrera-Frank theory [53]. During growth, the maximal surface supersaturation σs,max at a monolayer step depends on the vapour supersaturation σ, terrace width λ0 and surface diffusion length λs, as follows:
If the σs,max exceeds a critical value σs*, island nucleation occurs on the terraces which marks the onset of the morphological deterioration. At the same time, the terrace width λ0 increases when misorientation decreases and λs depends on the growth temperature. Figure 8 presents a calculation illustrating Eq. (1), based on the parameters of our sample and the growth conditions. We see that for large terrace widths (low misorientation) the maximal surface supersaturation increases. In this experiment, the threshold value of misorientation between two growth modes is around 1.5°. The corresponding value of the σs* was marked by the horizontal yellow line.
Fig. 8. Simulation of the ratio of maximal surface supersaturation at a monolayer step and vapour supersaturation. The terrace width is estimated as monolayer step height, 0.259 nm, divided by tangent of the misorientation. Horizontal axes in (a) and (b) present a corresponding data range.
It should be also emphasized, that the growth of InGaN layers is always accompanied by the roughening of the monolayer step-edges [37,45]. In reality λ0 is not a single and well-defined width, but rather a quite broad range of values. Nevertheless, the model presents a valuable insight into the growth process.
During the high magnification imaging of the regions of interest, we collected the EDX signal in order to estimate the local InN mole fraction. The results are presented in Fig. 9. The values were measured in the locations characterized by the highest InN mole fraction within the studied InGaN layers in the area visible in a single image. If several points were accepted, the average value was shown in the diagram. It should be noted that although the electron beam was focused to 0.16 nm, irradiated electrons were spread on its way through the TEM specimen. In case of our TEM specimen with a typical thickness of 100 nm, the spread is estimated to be 5.2 nm [54].
Fig. 9. (a) InN mole fraction obtained through EDX measurement during the STEM study. (b) Example of the studied points in high-magnification HAADF image obtained for point 1’. We show an unmodified dark-field grayscale contrast. (c) Comparison of the dependence between InN mole fraction and the off-cut angle reported in the literature with results from (a) obtained for the bottom QW (deep region). Values from this work were shifted vertically by 3% to compensate for the underestimation related to the EDX experimental method (to give a more straightforward comparison with more precise literature data obtained by XRD). The value of 3% was chosen to match InN mole fraction for low off-cut angles with the intended 17%.
A general trend of decreasing of InN mole fraction with misorientation is observed in Fig. 9, which agrees with the expectation based on other reports [36,37,45,46,49,55]. In our previous experiment [49] we observed a larger change in InN mole fraction, ranging from 9% to 18% in an even smaller misorientation angle range: 0.5°-2.6°. In case of the current sample, with misorientation angle ranging from 0.3° to 3.2°, the composition of QWs changes by around 3% so a smaller value. One conclusion is that the process of epitaxial growth had significantly different conditions when compared to the previous sample (mainly lowered temperature of the QW growth), and that change weakened the dependence of In incorporation on the misorientation angle of the substrate. The mechanism of off-cut angle dependent In adatom incorporation can be related to the growth temperature, which is demonstrated e.g. in Ref. [36] for a sample having around 15% of InN mole fraction for 0.4°. Please see Fig. 9(c) where rectangle symbols represent 780°C and circles 820°C. Moreover, Liu and co-authors [45] presented samples with large and almost constant InN mole fraction – 21-22% – in a range of off-cut angle between 0.24° and 0.97° for QWs grown at low temperature of 730°C.
Additionally, it can be noticed that the InN mole fraction profile measured by EDX does not reflect the non-monotonic peak wavelength distribution presented in Fig. 2 (a). This is because the peak wavelength is also influenced by the size of the area generating light. In the areas of low misorientation angle the QW is not continuous, forming quantum dots or dashes of varying sizes both in terms of the in-plane directions and the thickness. Secondly, due to the small change in the InN mole fraction, the increase of intensity and improvement of material quality observed for high misorientation region probably should not be explained through the easier fabrication of material with lower InN mole fraction. It also needs to be reminded that the InN mole fraction obtained by EDX is underestimated for most layers due to the spread of the electrons. Therefore, in reality, the InN mole fraction in the InGaN layers should be higher than the EDX results. This effect explains, at least partially, the difference in the composition between the top and bottom QW – the top QW which is thinner, shows around 2.5% lower composition. The effect is also at least partially responsible for the observed reduction of the InN mole fraction in QWs, with respect to the intended structure.
The TEM results show a pronounced difference in the width between the two QWs, which may raise concerns in relation to the long wavelength peak observed in Fig. 4. In order to prove that the emission observed in the μPL and μTRPL experiments comes from the bottom quantum well (less decomposed), we performed additional PL measurements on a different system than introduced in Section 4. We used a PL system which allows CW excitation by 375 nm and 320 nm lasers with a spot size having diameter of around 50 μm. We examined a reference 80 × 80 μm region of the same sample, characterized by nominally uniform misorientation angle of around 3°. The PL spectra were taken for different excitation power densities. In Fig. 10 we compare spectra obtained with excitation by 375 nm laser at room temperature, analogic condition to what is shown in Fig. 3(a), and by 320 nm laser at room and cryogenic temperatures (20 K). The excitation power density was around 0.2 kW/cm2 for 375 nm laser and 0.3 kW/cm2 for 320 nm laser. In case of the 375 nm excited spectrum, we observe a similar shape to the result presented in Fig. 3(a). In case of the 320 excited spectra, we see the appearance of an additional peak at the short wavelength (high energy) side. It is better observed in the low temperature spectrum. We attribute this peak to the emission from the top QW which is decomposed and thin. The usage of higher excitation energy allowed to generate carriers in the GaN barriers. These carriers were captured in the area of the top QW allowing to observe its emission. With this result, we believe it can be assumed that the PL and TRPL spectra obtained within the previous measurements (Fig. 2, Fig. 3, Fig. 4) are dominated by the signal from the bottom QW.
Fig. 10. Comparison of PL spectra obtained in an area of the same sample having a uniform misorientation angle of around 3°. The excitation by higher energy reveals an additional emission peak around 445 nm which we attribute to the top QW. The excitation power density was 0.2 kW/cm2 for 375 nm and 0.3 kW/cm2 for 320 nm laser.
8. Conclusions
In summary, we presented a comparison of the structural and emission properties of InGaN QWs with nominally 17% InN mole fraction grown on bulk GaN in an area with changing misorientation in a range from 0.3° to 3.2°. We observe that the structure of the active region fabricated under the same growth conditions varies significantly with substrate misorientation. In a typical range of values, below 1.5°, we observed low emission intensity together with a non-uniform emission wavelength map suggesting poor structural quality of the QWs. The higher misorientation area, around 2° and above, showed higher emission intensity and better uniformity. The time-resolved photoluminescence measurements suggest that the high misorientation region is characterised by lower rate of non-radiative recombination processes. This conclusion is supported by the low-temperature PL mapping. AFM scans confirmed an improvement of the surface morphology for increased misorientation. This change is also accompanied by the better QW continuity and quality, as it was demonstrated in STEM images. The observed differences in structural quality can be attributed to the modification of surface supersaturation during the MOVPE growth by local substrate misorientation. Our studies confirm that quality of the growth of high-InN-mole fraction layers can be improved just by increasing the misorientation angle of the substrate and that the improved quality is preserved even for large misorientation angle – up to 3°.
Funding
Japan Society for the Promotion of Science (JP16H06426, JP20H05622, JP21H04661); Fundacja na rzecz Nauki Polskiej (TEAM TECH/2017-4/24).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
Supplemental document
See Supplement 1 for supporting content.
References
1. H. Amano, M. Kito, K. Hiramatsu, and I. Akasaki, “P-Type conduction in Mg-doped GaN treated with low-energy electron beam irradiation (LEEBI),” Jpn. J. Appl. Phys. 28(Part 2, No. 12), L2112–L2114 (1989). [CrossRef]
2. S. Nakamura, T. Mukai, M. Senoh, and N. Iwasa, “Thermal annealing effects on P-Type Mg-doped GaN films,” Jpn. J. Appl. Phys. 31(Part Part 2, No. 2B), L139–L142 (1992). [CrossRef]
3. S. Nakamura, M. Senoh, S. Nagahama, N. Iwasa, T. Yamada, T. Matsushita, H. Kiyoku, and Y. Sugimoto, “InGaN-based multi-quantum-well-structure laser diodes,” Jpn. J. Appl. Phys. 35(Part 2, No. 1B), L74–L76 (1996). [CrossRef]
4. Y. Narukawa, J. Narita, T. Sakamoto, K. Deguchi, T. Yamada, and T. Mukai, “Ultra-high efficiency white light emitting diodes,” Jpn. J. Appl. Phys. 45(No. 41), L1084–L1086 (2006). [CrossRef]
5. N. F. Gardner, G. O. Müller, Y. C. Shen, G. Chen, S. Watanabe, W. Götz, and M. R. Krames, “Blue-emitting InGaN–GaN double-heterostructure light-emitting diodes reaching maximum quantum efficiency above 200A∕cm2,” Appl. Phys. Lett. 91(24), 243506 (2007). [CrossRef]
6. A. Laubsch, M. Sabathil, J. Baur, M. Peter, and B. Hahn, “High-Power and high-efficiency InGaN-based light emitters,” IEEE Trans. Electron Devices 57(1), 79–87 (2010). [CrossRef]
7. H. Hirayama, N. Maeda, S. Fujikawa, S. Toyoda, and N. Kamata, “Recent progress and future prospects of AlGaN-based high-efficiency deep-ultraviolet light-emitting diodes,” Jpn. J. Appl. Phys. 53(10), 100209 (2014). [CrossRef]
8. S. M. Islam, K. Lee, J. Verma, V. Protasenko, S. Rouvimov, S. Bharadwaj, H. Grace Xing, and D. Jena, “MBE-grown 232–270 nm deep-UV LEDs using monolayer thin binary GaN/AlN quantum heterostructures,” Appl. Phys. Lett. 110(4), 041108 (2017). [CrossRef]
9. Z. Zhang, M. Kushimoto, T. Sakai, N. Sugiyama, L. J. Schowalter, C. Sasaoka, and H. Amano, “A 271.8 nm deep-ultraviolet laser diode for room temperature operation,” Appl. Phys. Express 12(12), 124003 (2019). [CrossRef]
10. M. Kneissl, T.-Y. Seong, J. Han, and H. Amano, “The emergence and prospects of deep-ultraviolet light-emitting diode technologies,” Nat. Photonics 13(4), 233–244 (2019). [CrossRef]
11. H. Amano, R. Collazo, C.D. Santi, S. Einfeldt, M. Funato, J. Glaab, S. Hagedorn, A. Hirano, H. Hirayama, R. Ishii, Y. Kashima, Y. Kawakami, R. Kirste, M. Kneissl, R. Martin, F. Mehnke, M. Meneghini, A. Ougazzaden, P.J. Parbrook, S. Rajan, P. Reddy, F. Römer, J. Ruschel, B. Sarkar, F. Scholz, L.J. Schowalter, P. Shields, Z. Sitar, L. Sulmoni, T. Wang, T. Wernicke, M. Weyers, B. Witzigmann, Y.-R. Wu, T. Wunderer, and Y. Zhang, ”The 2020 UV emitter roadmap,“ J. Phys. D: Appl. Phys. 53(50), 503001 (2020). [CrossRef]
12. S. Nakamura, M. Senoh, N. Iwasa, S. Nagahama, T. Yamada, and T. Mukai, “Superbright green InGaN single-quantum-well-structure light-emitting diodes,” Jpn. J. Appl. Phys. 34(Part 2, No. 10B), L1332–L1335 (1995). [CrossRef]
13. T. Miyoshi, S. Masui, T. Okada, T. Yanamoto, T. Kozaki, S. Nagahama, and T. Mukai, “510–515 nm InGaN-based green laser diodes on c -plane GaN substrate,” Appl. Phys. Express 2, 062201 (2009). [CrossRef]
14. Y. Jiang, Y. Li, Y. Li, Z. Deng, T. Lu, Z. Ma, P. Zuo, L. Dai, L. Wang, H. Jia, W. Wang, J. Zhou, W. Liu, and H. Chen, “Realization of high-luminous-efficiency InGaN light-emitting diodes in the “green gap” range,” Sci Rep 5(1), 10883 (2015). [CrossRef]
15. S. Lutgen, A. Avramescu, T. Lermer, D. Queren, J. Müller, G. Bruederl, and U. Strauss, “True green InGaN laser diodes,” phys. stat. sol. (a) 207(6), 1318–1322 (2010). [CrossRef]
16. J. W. Raring, M. C. Schmidt, C. Poblenz, Y.-C. Chang, M. J. Mondry, B. Li, J. Iveland, B. Walters, M. R. Krames, R. Craig, P. Rudy, J. S. Speck, S. P. DenBaars, and S. Nakamura, “High-efficiency blue and true-green-emitting laser diodes based on non- c -plane oriented GaN substrates,” Appl. Phys. Express 3(11), 112101 (2010). [CrossRef]
17. S. Saito, R. Hashimoto, J. Hwang, and S. Nunoue, “InGaN light-emitting diodes on c -face sapphire substrates in green gap spectral range,” Appl. Phys. Express 6(11), 111004 (2013). [CrossRef]
18. H. Sekiguchi, K. Kishino, and A. Kikuchi, “Emission color control from blue to red with nanocolumn diameter of InGaN/GaN nanocolumn arrays grown on same substrate,” Appl. Phys. Lett. 96(23), 231104 (2010). [CrossRef]
19. J.-I. Hwang, R. Hashimoto, S. Saito, and S. Nunoue, “Development of InGaN-based red LED grown on (0001) polar surface,” Appl. Phys. Express 7(7), 071003 (2014). [CrossRef]
20. T. Frost, A. Hazari, A. Aiello, M. Z. Baten, L. Yan, J. Mirecki-Millunchick, and P. Bhattacharya, “High performance red-emitting multiple layer InGaN/GaN quantum dot lasers,” Jpn. J. Appl. Phys. 55(3), 032101 (2016). [CrossRef]
21. F. Jiang, J. Zhang, L. Xu, J. Ding, G. Wang, X. Wu, X. Wang, C. Mo, Z. Quan, X. Guo, C. Zheng, S. Pan, and J. Liu, “Efficient InGaN-based yellow-light-emitting diodes,” Photon. Res. 7(2), 144 (2019). [CrossRef]
22. T. Ozaki, M. Funato, and Y. Kawakami, “Red-emitting In x Ga 1− x N/In y Ga 1− y N quantum wells grown on lattice-matched In y Ga 1− y N/ScAlMgO 4 (0001) templates,” Appl. Phys. Express 12(1), 011007 (2019). [CrossRef]
23. A. Dussaigne, F. Barbier, B. Damilano, S. Chenot, A. Grenier, A. M. Papon, B. Samuel, B. Ben Bakir, D. Vaufrey, J. C. Pillet, A. Gasse, O. Ledoux, M. Rozhavskaya, and D. Sotta, “Full InGaN red light emitting diodes,” J. Appl. Phys. 128(13), 135704 (2020). [CrossRef]
24. S. Zhang, J. Zhang, J. Gao, X. Wang, C. Zheng, M. Zhang, X. Wu, L. Xu, J. Ding, Z. Quan, and F. Jiang, “Efficient emission of InGaN-based light-emitting diodes: toward orange and red,” Photon. Res. 8(11), 1671 (2020). [CrossRef]
25. I. Vurgaftman, J. R. Meyer, and L. R. Ram-Mohan, “Band parameters for III–V compound semiconductors and their alloys,” J. Appl. Phys. 89(11), 5815–5875 (2001). [CrossRef]
26. A. V. Zinovchuk and E. A. Sevost’yanov, “Optimized valence force field model for the lattice properties of non-ideal III-nitride wurtzite materials,” Phys. B 574, 411682 (2019). [CrossRef]
27. A. I. Duff, L. Lymperakis, and J. Neugebauer, “Understanding and controlling indium incorporation and surface segregation on In x Ga 1 − x N surfaces: An ab initio approach,” Phys. Rev. B 89(8), 085307 (2014). [CrossRef]
28. V. Fiorentini, F. Bernardini, F. Della Sala, A. Di Carlo, and P. Lugli, “Effects of macroscopic polarization in III-V nitride multiple quantum wells,” Phys. Rev. B 60(12), 8849–8858 (1999). [CrossRef]
29. D. A. B. Miller, D. S. Chemla, T. C. Damen, A. C. Gossard, W. Wiegmann, T. H. Wood, and C. A. Burrus, “Band-edge electroabsorption in quantum well structures: the quantum-confined stark effect,” Phys. Rev. Lett. 53(22), 2173–2176 (1984). [CrossRef]
30. Jae-Hyun Ryou, P. D. Yoder, Jianping Liu, Z. Lochner, Hyunsoo Kim, Suk Choi, Hee Jin Kim, and R. D. Dupuis, “Control of quantum-confined Stark effect in InGaN-Based quantum wells,” IEEE J. Select. Topics Quantum Electron. 15(4), 1080–1091 (2009). [CrossRef]
31. M. Funato, M. Ueda, Y. Kawakami, Y. Narukawa, T. Kosugi, M. Takahashi, and T. Mukai, “Blue, green, and amber InGaN/GaN light-emitting diodes on semipolar {11-22} GaN bulk substrates,” Jpn. J. Appl. Phys. 45(No. 26), L659–L662 (2006). [CrossRef]
32. J. E. Northrup, “GaN and InGaN(1122) surfaces: Group-III adlayers and indium incorporation,” Appl. Phys. Lett. 95(13), 133107 (2009). [CrossRef]
33. Y. Zhao, Q. Yan, C.-Y. Huang, S.-C. Huang, P. Shan Hsu, S. Tanaka, C.-C. Pan, Y. Kawaguchi, K. Fujito, C. G. Van de Walle, J. S. Speck, S. P. DenBaars, S. Nakamura, and D. Feezell, “Indium incorporation and emission properties of nonpolar and semipolar InGaN quantum wells,” Appl. Phys. Lett. 100(20), 201108 (2012). [CrossRef]
34. M. V. Durnev, A. V. Omelchenko, E. V. Yakovlev, I. Yu. Evstratov, and S. Yu. Karpov, “Indium incorporation and optical transitions in InGaN bulk materials and quantum wells with arbitrary polarity,” Appl. Phys. Lett. 97(5), 051904 (2010). [CrossRef]
35. D. A. Browne, E. C. Young, J. R. Lang, C. A. Hurni, and J. S. Speck, “Indium and impurity incorporation in InGaN films on polar, nonpolar, and semipolar GaN orientations grown by ammonia molecular beam epitaxy,” Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films 30(4), 041513 (2012). [CrossRef]
36. M. Sarzynski, M. Leszczynski, M. Krysko, J. Z. Domagala, R. Czernecki, and T. Suski, “Influence of GaN substrate off-cut on properties of InGaN and AlGaN layers,” Cryst. Res. Technol. 47(3), 321–328 (2012). [CrossRef]
37. M. Leszczynski, R. Czernecki, S. Krukowski, M. Krysko, G. Targowski, P. Prystawko, J. Plesiewicz, P. Perlin, and T. Suski, “Indium incorporation into InGaN and InAlN layers grown by metalorganic vapor phase epitaxy,” J. Cryst. Growth 318(1), 496–499 (2011). [CrossRef]
38. R. M. Farrell, D. A. Haeger, X. Chen, M. Iza, A. Hirai, K. M. Kelchner, K. Fujito, A. Chakraborty, S. Keller, S. P. DenBaars, J. S. Speck, and S. Nakamura, “Effect of carrier gas and substrate misorientation on the structural and optical properties of m-plane InGaN/GaN light-emitting diodes,” J. Cryst. Growth 313(1), 1–7 (2010). [CrossRef]
39. K. M. Kelchner, L. Y. Kuritzky, K. Fujito, S. Nakamura, S. P. DenBaars, and J. S. Speck, “Emission characteristics of single InGaN quantum wells on misoriented nonpolar m-plane bulk GaN substrates,” J. Cryst. Growth 382, 80–86 (2013). [CrossRef]
40. M. Sarzyński, T. Suski, R. Czernecki, E. Grzanka, Ł. Marona, A. Khachapuridze, P. Dróżdż, K. Pieniak, J. Z. Domagała, M. Leszczyński, and P. Perlin, “Properties of InGaN/GaN multiquantum wells grown on semipolar (20-21) substrates with different miscuts,” J. Cryst. Growth 423, 28–33 (2015). [CrossRef]
41. J. Du, S. Xu, Z. Lin, Y. Zhang, J. Zhang, Y. Zhao, R. Peng, X. Fan, M. Niu, J. Huang, and Y. Hao, “Improvement of crystalline quality of N-polar green InGaN/GaN multiple quantum wells on vicinal substrate,” Mater. Sci. Semicond. Process. 96, 167–172 (2019). [CrossRef]
42. S. Keller, C. S. Suh, N. A. Fichtenbaum, M. Furukawa, R. Chu, Z. Chen, K. Vijayraghavan, S. Rajan, S. P. DenBaars, J. S. Speck, and U. K. Mishra, “Influence of the substrate misorientation on the properties of N-polar InGaN/GaN and AlGaN/GaN heterostructures,” J. Appl. Phys. 104(9), 093510 (2008). [CrossRef]
43. L. Y. Kuritzky, D. J. Myers, J. Nedy, K. M. Kelchner, S. Nakamura, S. P. DenBaars, C. Weisbuch, and J. S. Speck, “Electroluminescence characteristics of blue InGaN quantum wells on m -plane GaN “double miscut” substrates,” Appl. Phys. Express 8(6), 061002 (2015). [CrossRef]
44. A. Tian, J. Liu, L. Zhang, L. Jiang, M. Ikeda, S. Zhang, D. Li, P. Wen, Y. Cheng, X. Fan, and H. Yang, “Significant increase of quantum efficiency of green InGaN quantum well by realizing step-flow growth,” Appl. Phys. Lett. 111(11), 112102 (2017). [CrossRef]
45. Z. Liu, S. Nitta, Y. Robin, M. Kushimoto, M. Deki, Y. Honda, M. Pristovsek, and H. Amano, “Morphological study of InGaN on GaN substrate by supersaturation,” J. Cryst. Growth 508, 58–65 (2019). [CrossRef]
46. A. Kafar, R. Ishii, K. Gibasiewicz, Y. Matsuda, S. Stanczyk, D. Schiavon, S. Grzanka, M. Tano, A. Sakaki, T. Suski, P. Perlin, M. Funato, and Y. Kawakami, “Above 25 nm emission wavelength shift in blue-violet InGaN quantum wells induced by GaN substrate misorientation profiling: towards broad-band superluminescent diodes,” Opt. Express 28(15), 22524 (2020). [CrossRef]
47. H. P. Zhao, V. Dierolf, G. Y. Liu, G. S. Huang, X.-H. Li, J. D. Poplawsky, R. A. Arif, S. Tafon Penn, and N. Tansu, “Design and characteristics of staggered InGaN quantum-well light-emitting diodes in the green spectral regime,” IET Optoelectron. 3(6), 283–295 (2009). [CrossRef]
48. H. Zhao, G. Liu, J. Zhang, J. D. Poplawsky, V. Dierolf, and N. Tansu, “Approaches for high internal quantum efficiency green InGaN light-emitting diodes with large overlap quantum wells,” Opt. Express 19(S4), A991 (2011). [CrossRef]
49. A. Kafar, A. Sakaki, R. Ishii, S. Stanczyk, K. Gibasiewicz, Y. Matsuda, D. Schiavon, S. Grzanka, T. Suski, P. Perlin, M. Funato, and Y. Kawakami, “Influence of substrate misorientation on the emission and waveguiding properties of a blue (In, Al, Ga) N laser-like structure studied by synchrotron radiation microbeam X-ray diffraction,” Photon. Res. 9(3), 299 (2021). [CrossRef]
50. G. M. Christian, S. Schulz, M. J. Kappers, C. J. Humphreys, R. A. Oliver, and P. Dawson, “Recombination from polar InGaN/GaN quantum well structures at high excitation carrier densities,” Phys. Rev. B 98(15), 155301 (2018). [CrossRef]
51. L. Jiang, J. Liu, A. Tian, X. Ren, S. Huang, W. Zhou, L. Zhang, D. Li, S. Zhang, M. Ikeda, and H. Yang, “Influence of substrate misorientation on carbon impurity incorporation and electrical properties of p-GaN grown by metalorganic chemical vapor deposition,” Appl. Phys. Express 12(5), 055503 (2019). [CrossRef]
52. I. Bryan, Z. Bryan, S. Mita, A. Rice, J. Tweedie, R. Collazo, and Z. Sitar, “Surface kinetics in AlN growth: a universal model for the control of surface morphology in III-nitrides,” J. Cryst. Growth 438, 81–89 (2016). [CrossRef]
53. W. K. Burton, N. Cabrera, and F. C. Frank, “The growth of crystals and the equilibrium structure of their surfaces,” Phil. Trans. R. Soc. Lond. A 243(866), 299–358 (1951). [CrossRef]
54. D. B. Williams and C. B. Carter, Transmission Electron Microscopy (Plenum Press, 1996), Chap. 36. [CrossRef]
55. P. A. Dróżdż, M. Sarzyński, J. Z. Domagała, E. Grzanka, S. Grzanka, R. Czernecki, Ł. Marona, K. P. Korona, and T. Suski, “Monolithic cyan − violet InGaN/GaN LED array,” Phys. Status Solidi A 214(8), 1600815 (2017). [CrossRef]
