Open Access
Add to BibSonomy
Abstract
The influence of the nonradiative recombination in a multiple quantum well of GaN-based blue laser diodes (LDs) has been are studied experimentally and theoretically by analyzing the optical and electrical properties of LDs with various thickness and indium content of quantum wells (QWs). It is found that when keeping the LD emission wavelength nearly unchanged, the LD device performance with thinner QW and higher indium content of InGaN QWs is much better than the LD with thicker QW and lower indium content, having smaller threshold current density, higher output optical power and larger slope efficiency. Typically, the threshold current density is as low as 0.69 kA/cm2, and the corresponding threshold current is only 250 mA. The lifetime is more than 10,000 hours at a fixed injection current of 1.2 A under a room-temperature continuous-wave operation. Characteristics of photoluminescence (PL) microscopy images, temperature dependent PL spectra, time-resolved PL and electroluminescence spectra demonstrate that a reduction of the nonradiative recombination centers and an improvement of homogeneity in QWs are the main reason for the performance improvement of GaN-based LD using thinner QW layers with a higher indium content in a certain range. Moreover, theoretical calculation results demonstrate that using a thinner quantum well is also helpful for improving the device performance if the change of alloy material quality is considered during the calculation.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
Corrections
26 September 2022: A typographical correction was made to the author listing.
1. Introduction
GaN-based blue laser diodes (LDs) have undergone impressive development in the last decade for the critical applications in material processing [1,2], wireless communication [3,4], laser display [5,6] and quantum technology [7,8]. In material processing applications of blue LDs, a better device performance of laser source makes manufacture of the process equipment easier and it is beneficial to reduce the size of the processing system. A high device performance is also helpful to realize the optical communication and laser display with high quality and low power consumption. Thus, a further improvement of laser source is urgent and meaningful for pushing the applications of GaN-based laser source in a new high. Actually, in the last thirty decades, researchers have made much effort to improve device performance and great advances have been achieved since Nichia Ltd. reported the first GaN-based blue-violet LD working under pulsed current injection in 1996 [9–11]. Some researchers focused on the investigations of the physical mechanism such as the lasing process [12], the dynamic characteristics of devices [13], and the spectral-temporal dynamics [14]. Researchers have also developed new epitaxial growth methods to improve the material quality, including homogeneity control of the wafer surface [15], the low temperature p-type doping technology [16] and using GaN plasmonic substrate to reduce the thickness of AlGaN cladding layer [17]. Researchers also have proposed new LD structures to improve performance, using a lattice matched AlInN optical blocking layer to improve optical beam quality [18] or using a thick InGaN waveguide to eliminate the optical mode leakage [19]. Moreover, redesign the LD structures has been employed to improve the LD performance by adopting the thickness and indium content of the quantum wells, modifying the quantum barriers, electron blocking layer and waveguide layers [20–24].
Actually, the multiple quantum well (MQW) is the active region of GaN-based LDs whose property is crucial for the device performance. Therefore, its investigation is a hot topic. A lot of research on MQW has been conducted to reveal the effects of non-uniform quantum well on threshold and linewidth [25], how to suppress the V-pit formation in MQW to improve device performance [26], to study the influence of InGaN interlayer on the strain relaxation of MQW in laser diode structures [27]. However, it is worth noting that the discussion of the width of InGaN quantum well (QW) on the device performance is controversial so far. On one side, it is reported that a wide InGaN QW can be more effective as the active region of laser diodes than the thinner ones as usually used, attributing to a higher optical gain and higher quantum efficiency in wider InGaN QWs [28–30]. On the other side, it is also reported that a thinner InGaN QW is beneficial to improve the LD’s performance, offering a higher TE gain due to the enhanced transverse electric field [31] or obtaining a higher characteristic temperature because of the dominated exciton localization effect [32], and to reach higher output optical power due to a smaller polarized electric field [33]. Thus, a study is needed to further investigate the physical mechanisms about how InGaN QW layer affects the device performance of GaN-based LDs, especially taking the same lasing wavelength as the premise. Therefore, in this work, GaN-based LDs with various quantum well thickness are designed and fabricated. Experimental and theoretical results demonstrate that using thinner quantum wells in a certain range is beneficial to suppress the nonradiative recombination in MQW and is helpful to concentrate the optical field distribution. Thus, device performance can be improved significantly, and a laser diode with low threshold current density of 0.69 kA/cm2 is obtained.
2. Fabrication and theoretical calculation of GaN-based LDs
2.1. Epitaxial growth and device fabrication
Three GaN-based blue LDs, named as LD1, LD2 and LD3, are grown in an AIXTRON 3 × 2 in. close-coupled showerhead reactor on c-plane GaN substrates. The schematic device structure of LD1, LD2 and LD3 is shown in Fig. 1, consisting of an lower cladding layer, a lower waveguide layer, a multiple quantum well with two periods of quantum wells, an upper waveguide layer, an electron blocking layer, an upper cladding layer, and a p-GaN layer, successively. TMGa/TEGa, TMAl, TMIn, NH3, Cp2Mg and SiH4 are used as the sources of Ga, Al, In, N, Mg and Si sources, respectively. After epitaxial growth, LD chips with 30-µm-wide ridge stripe and 1200-µm-long cavity are fabricated with the help of lithography, dry etching, cleaving etc. The reflectivity of front and rear cleaved cavities facets after coating is about 10% and 90%, respectively. In addition, n-type and p-type metal electrodes are formed with the help of Ti/Pt/Au and Pd/Pt/Au, respectively.
It should be mentioned that only the thickness and indium content of QW layers are different for LD1, LD2 and LD3. The parameters are adjusted by modifying the flow rates of TMGa and TMIn during the QW layers growth in the condition that after such modification, the lasing wavelength of LD1, LD2 and LD3 is kept nearly the same. This condition guarantees that the discussions on the influence of quantum well structure on the device performance are comparable and valuable. Additionally, three MQW samples, names as MQW1, MQW2 and MQW3, are also prepared to measure the structural parameters of quantum wells in LD1, LD2 and LD3, and which are also used to further investigate the physical mechanism of how InGaN QW affects the LD’s performance. Figure 2 shows the schematic structure of MQW samples. Actually, the epitaxial conditions of QWs for MQW1, MQW2 and MQW3 are the same as the QWs of LD1, LD2 and LD3, respectively.
The structural parameters of quantum wells in LD1, LD2 and LD3 are obtained by X-ray diffraction (XRD) rocking curve measurements by using MQW samples. The details on the XRD measurements are given in section 2.2 “Measurements”. Figure 3 shows the XRD ω−2θ rocking curves at the (002) reflection of MQW1, MQW2 and MQW3, and the QW structural parameters are obtained according to the fitted lines by Global Fit, and the results are shown in Table 1. The thickness of QW for LD1, LD2 and LD3 is 4.4 nm, 3.8 nm and 3.1 nm, respectively, and the indium content of QW is 9.0%, 9.5% and 10.2%, respectively. Their quantum barrier layers are close to each other in the structural parameters, where the thickness and indium content is around 10.5 nm and 2%, respectively. It can be seen that for LD1, LD2 and LD3, the thickness of QW layers decreases sequentially, and the indium content increases in turn.
Fig. 3. Schematic structure of MQW samples. The (002) ω−2θ rocking curves of MQW1, MQW2 and MQW3. The dark and red lines are the measurement data and fitting data, respectively.
Table 1. Structural parameters of MQW samples according to the ω−2θ rocking curves.
2.2 Measurements
For LD1, LD2 and LD3, the optical output power, optical spectra of stimulated emission and photoluminescence microscopies were measured. Output optical power was measured under direct current mode using a Si-based photodetector and a Keithley source meter. The lasing spectra were obtained using a Keithley source meter and a spectrometer (Ocean Optics HR2000). PL microscopy was measured with the help of a confocal optical system (Nikon A1) which is equipped with a 405 nm-laser excitation source. Moreover, for MQW1, MQW2 and MQW3, the XRD rocking curves, PL microscopic images, electroluminescence spectra, temperature dependent PL spectra (TDPL) and time resolved PL (TRPL) spectra were measured. XRD rocking curves of MQW samples are measured (Rigaku SmartLab) using the Cu Kα1 radiation of 0.154 nm which is equipped with a 2D detector and a 2-bounch monochromator of Ge crystal. TDPL spectra were measured with the excitation source of 325-nm He-Cd laser, and the temperature control was realized through a closed-cycle refrigerator of CTI Cryogenics. Electroluminescence spectra of MQW samples were also measured on wafer using a spectrometer (Ocean Optics HR2000) under direct current injection. In addition, TRPL measurements of MQW1 and MQW3 were conducted through a time-correlated single photon counting equipment with a 405 nm-laser excitation source, and the pulse repetition frequency was setup to 10 MHz and the corresponding excitation cycle is 100ns during the measurement.
2.3 Design and simulation calculation of LDs with different QWs
Six GaN-based blue LDs, named as LDI, LDII, LDIII, LDIV, LDV and LDI-1 are designed as shown in Table 2 and then calculated by simulation through LASTIP (Crosslight Software Inc.). This software is helpful to analyze the electrical and optical characteristics of LDs thermionically. The device structures of LDI-LDV were designed based on the device structures of LD1-3 as shown in Fig. 1. Only the thickness and indium content of QW are different among LDI-LDV, and more structural parameters of QW for LDI-LDV can be found in Table 2. Device structure of LDI-1 is the same as the LDI, and only the carrier life is different during calculating. The thickness of QW for LDI-LDV is 4.5 nm, 4.0 nm, 3.5 nm, 3.0 nm and 2.5 nm, respectively, decreasing in turn one by one. The indium content of QW is 11.5%, 11.8%, 12.3%, 12.9% and 13.8%, respectively, increasing in turn. During calculation, the screening factor is set to 25%, and the front and rear reflectivity are set to be 20% and 90%, respectively. In addition, the refractive indexes of AlGaN and InGaN alloys are obtained according to the method proposed by Piprek and Laws et al [34,35]. Table 2 also shows the lasing wavelength of LDI-LDV which is close to each other, i.e., all around 445 nm. A close lasing wavelength is a crucial precondition for theoretical discussions on the influence of quantum wells on the LD’s performance.
Table 2. Structural parameters of InGaN quantum well in LDI-LDV for calculation with LASTIP and their lasing wavelengths.
3. Results and discussion
The optical spectra of stimulated emission and the optical output power of LD chips LD1, LD2 and LD3 are measured under forward direct current at room temperature which are shown in Fig. 4. The optical spectra of stimulated emission show that the lasing wavelengths of LD1-LD3 are close to each other, i.e., approximately 445.9 nm, 445.4 nm and 444.8 nm, respectively. The curves of the optical output power versus forward direct current indicate that the threshold current of LD1, LD2 and LD3 is approximately 1387 mA, 556 mA and 250 mA, respectively, and the corresponding threshold current densities are approximately 3.85 kA/cm2, 1.54 kA/cm2 and 0.69 kA/cm2, respectively. The optical output power of LD1, LD2 and LD3 reaches to 1.09 W, 2.26 W and 2.89 W respectively when the direct current is 2.5 A. In addition, the slope efficiency is calculated by liner fitting of the optical output power-current curves after lasering. It is 0.94 W/A, 1.16 W/A and 1.39 W/A for LD1, LD2 and LD3, respectively, showing an increase in turn. It indicates that device performance is improved when the thickness of QW layer decreases from 4.4 nm to 3.1 nm, and the threshold current, optical power, and slope efficiency have an improvement of 82.1%, 100.7% and 47.8%, respectively. This experimental result demonstrates that LD with a relative thinner QW with higher indium content is beneficial to improve the device performance of GaN-based blue LDs.
Fig. 4. (a) Optical spectra of stimulated emission of LDI and LDII under 1.8 A current injection at room-temperature. (b) Optical output power versus current (P-I) curves of LD1, LD2 and LD3 at room temperature under continuous-wave operation.
For light emitting devices, the defects in InGaN quantum well active region normally play a role of nonradiative recombination centers, and have a negative effect on the luminescence characteristics. The PL microscopic images of LD1, LD2 and LD3 are measured to check the defect’s distribution, in which the dark regions present concentrated nonradiative recombination centers. In the first row of Fig. 5, it can be seen that both LD1 and LD2 have many dark regions, whereas there are much fewer dark regions in LD3. Meanwhile, the size of dark regions of LD1, LD2 and LD3 decreases in turn. Actually, the crystalline quality is often getting worse when the thickness increases, and the poor material quality of InGaN is attributed to the big miscibility gap, the low growth temperature and the large lattice mismatch between GaN and InN [36–38]. The density of nonradiative recombination centers in LD3 seems much less than those in LD1 and LD2 according to the results of photoluminescence microscopy. Therefore, at least in a certain range, QWs which are relatively thin are beneficial to reduce the nonradiative recombination centers, and that is why device performance of LD3 is much better than those of LD1 and LD2. In addition, to make the difference of the dark region clearer between LD1 and LD2, an enlarged view of the PL microscopic images are shown in the second row of Fig. 5 which is in double size of the central area of Fig. 5. It is obvious that LD2 has less dark regions than LD1.
Fig. 5. Photoluminescence microscopic images of LD1, LD2 and LD3. Pictures in the last row are the enlarged views of LD1 and LD2 for the central area.
Actually, it is known that the epitaxial growth of thick p-AlGaN upper cladding layer under high temperature may also introduce dark regions, because the quantum well layers in LD samples may suffer a thermal degradation and then the nonradiative recombination centers may be enhanced. Therefore, in order to further check the origin of the dark regions, the photoluminescence microscopy of MQW samples is measured, in which no p-AlGaN cladding layers are grown. The result is shown in Fig. 6. It can be seen that there are many dark regions in both MQW1 and MQW2, whereas there is almost no dark region in MQW3. This result demonstrates that the nonradiative recombination centers can be suppressed using thinner QW layers with higher indium content. Meanwhile, the trend of dark regions in MQW samples is consistence with that in corresponding LD samples, indicating that the nonradiative recombination centers is indeed originated from the quantum well layers.
Furthermore, TRPL spectra of MQW1 and MQW3 are measured to check the carrier life time. The decay curves are shown in Fig. 7. It can be seen that the intensity of both MQW1 and MQW3 presents both a fast decay and a linearly decay, where the vertical ordinate is set to be exponential. The lifetime can be obtained by a linear fit in the fast decay region which is the absolute value of the slope for the fitted line. For MQW1 and MQW3, the lifetime is 2.8 ns and 23.3 ns, respectively. Actually, less nonradiative recombination centers will present a longer carrier life. Therefore, the result of TRPL spectra demonstrates that the density of nonradiative recombination centers in MQW3 should be less than that in MQW1, and this result is consistent with the discussions on the photoluminescence microscopies of MQW and LD samples.
Fig. 7. Time-resolved photoluminescence of MQW1 and MQW3. The red lines are the fitted lines in the region of fast decay.
The TDPL spectra of MQW samples are also measured to further investigate the influence of quantum well on the luminescence properties. In Fig. 8, the intensity of TDPL spectra is normalized to the peak intensity of GaN. Three distinct emission bands are observed in the TDPL spectra, marked as UVL1, UVL2 and BL, respectively, whose peak wavelength increases in turn. UVL1 and BL is originated from the GaN layer and the MQW active region. It is interesting to note that there is another emission band, i.e., UV2, which is located on the high-energy side of the BL band. Normally, it is reported that there is an emission band located on the low-energy side of GaN emission for the p-type GaN layer, and it is originated from the donor-acceptor pair recombination. However, in this work, the p-type GaN layer in MQW1, MQW2 and MQW3 are the same, indicating that the emission intensity of UVL2 band should be close to each other for them. However, it is noted that the UVL2 band has an irregular shape, and the emission intensity decreases in turn for MQW1, MQW2 and MQW3. Thus, it is suggested that the UVL2 band may originate from the MQW active region instead of from the p-type GaN layer. Thus, a stronger emission intensity of UVL2 band indicates a worse homogeneity of MQW. Therefore, TDPL spectra shows that the homogeneity of MQW3 is the best, and it should be another reason for the better performance of LD3.
Fig. 8. TDPL spectra of MQW samples excited by 325-nm He-Cd laser and measured in a temperature range of 30–300 K.
Moreover, EL spectra are measured to further check the homogeneity of MQW samples. Figure 9(a) illustrates the EL spectra of the MQW samples when the injection current is 50 mA. It can be seen that the emission intensity of MQW1, MQW2 and MQW3 increases in turn, indicating a highest luminescence efficiency of MQW3. EL spectra under injection current of 1 mA, 10 mA, 20 mA and 40 mA are also measured, from which the FWHM of EL spectral peak versus injection current is plotted and shown in Fig. 9(b). The values of current-dependent EL spectral FWHM of MQW3 are much smaller than those of both MQW1 and MQW2, indicating that MQW3 has more homogeneous quantum wells compared to MQW1 and MQW2. It is well studied that the fluctuations of Indium content and thickness of InGaN quantum wells result in localization states in MQW and carrier’s recombination through localization states, and thus narrower EL spectral FWHM presents more homogeneous quantum wells [39,40]. Therefore, in this study a smallest value of EL spectral FWHM of MQW3 indicates that the homogeneity of MQW3 is the best, and this should be another reason for the best device performance of LD3.
Fig. 9. (a) EL spectra of MQW1, MQW2 and MQW3 under direct injection current of 50 mA. (b) EL spectral FWHM of MQW1, MQW2 and MQW3 under different direct injection current.
It is also noted that the current-dependent curve of EL spectral FWHM of MQW1 is different compared to those of MQW2 and MQW3. For MQW1, EL spectral FWHM drops sharply first and then rise slightly with increasing injection current, but the EL spectral FWHM of MQW2 and MQW3 increases monotonically. Actually, EL spectral FWHM is a result of joint actions. On one side, carrier injection will weaken the band tilt of quantum wells, then energy separation between quantum energy levels will decrease, and the EL spectral FWHM will decrease [41,42]. On the other side, more localized states will participate in carrier recombination with the increase of injection current, and then the EL spectral FWHM will increase [43,44]. In this work, a sharp decrease is observed only in MQW1 when the injection current is lower than 30 mA, indicating that there is a stronger polarization shielding effect in MQW1 and thus fewer carriers recombine through local states. Considering on the poor material quality of MQW1 with more non-radiative recombination centers, it suggests that a large number of injected carriers is captured by non-radiative recombination centers instead of by the recombination in local states, and thus the EL spectral FWHM decreases when the injection current is lower than 30 mA. For MQW2 and MQW3, the material quality of quantum wells is better compared to MQW1, and most of carrier recombination happens in local states instead of capturing by non-radiative recombination centers, thus the EL spectral FWHM increases monotonically. This result is consistent with the discussion on the photoluminescence microscopy images and TDPL spectra of MQW samples.
The aging experiment for LD2 and LD3 has been taken to checked the lifetime which is performed at room temperature under a fixed current injection of 1.2 A. Figure 10 shows the relative optical output power as the function of aging time which is normalized to 100% by the optical power around 15 hours and 56 hours for LD2 and LD3, respectively. Actually, the threshold current of LD1 (1.387 A) is much larger than the aging current(1.2A), thus there is no lifetime measurement. It can be seen that the optical power of LD2 decreases sharply at about 100 hours and the lifetime is much less than that of LD3. The large threshold current and low wall-plug efficiency due to worse material quality of MQW in LD2 will lead to more serious problem of heat buildup, and thus the cavity surface damage is more likely to happen and result in a short lifetime of LD2. In addition, it can be seen that the optical power of LD3 decreases fast when the aging time is less than 0.6 hour, following an increase when the aging time is less than 56 hours, and then decreases slightly when the aging time is more than 56 hours. The relative optical output power is about 95% when the aging time is about 1200 hours. Thus, the aging time will be larger than 11550 hours when the relative optical output power reaches to the end-of-life criterion of 50% at this rate of decay. Therefore, the lifetime of LD3 is estimated more than 10000 hours at a fixed injection current of 1.2 A under a room-temperature continuous-wave operation. It indicates that such a long lifetime of LD3 gives a clue that using a relative thinner quantum well and higher indium content of InGaN QWs may be beneficial to well device performance.
Fig. 10. Relative optical output power for devices tested under a fixed injection current for LD2 and LD3 at room temperature.
Moreover, simulation calculation of LDs with different quantum wells are taken to further investigate the influence of quantum wells on the device performance. Figure 11(a) presents the threshold current and optical output power of five LDs, i.e., LDI, LDII, LDIII, LDIV and LDV. The threshold current of LDI-LDV is 128.8 mA, 131.1 mA, 139.1 mA, 149.1 mA and 164.2 mA, respectively. The optical power of LDI-LDV at 1.2 A is 1775.7 mW, 1819.1 mW, 1841.5 mW, 1834.2 mW and 1809.1 mW, respectively. It can be seen that the threshold current increases when the thickness of quantum well layers decreases. The reason for the increase of threshold current can be found based on the analysis on the optical confinement factor as shown in Fig. 12(a). It can be seen that the optical confinement factor decrease when the thickness of QW decreases indicating a less ration of optical field in thin MQW active region and a smaller material gain. It is speculated that a thin QW means a small active region and a low material gain, and thus a high threshold current. On the other side, it is noted that the optical power under 1.2 A increases first and then decreases. Why these two parameters have such kind of thickness dependence will be discussed and analyzed later. Fig11 (b) shows that the slope efficiency of LDI-LDV is 1.65 W/A, 1.69 W/A, 1.72 W/A, 1.73 W/A and 1.74 W/A, respectively, indicating an increase with the decrease of thickness of quantum well. This calculation result of slope efficiency is consistent with the experimental result, that relative thinner quantum wells with higher indium content are beneficial to improve the slope efficiency of GaN-based LDs.
Fig. 11. (a)Threshold current and optical output power, and (b) the slope efficiency of LDI-LDV calculated by LASTIP. All lines are a guide to the eye.
Fig. 12. (a) Optical confinement factor and optical loss of LDI-LDV. (b) FWHM of the optical field distribution curve along the growth direction of LDI-LDV. All the lines are a guide to the eye.
The optical field of LDI-LDV is analyzed to check the reason for the improvement of slope efficiency. In Fig. 12(a) shows that the optical confinement factor of LDI-LDV is 2.31%, 2.06%, 1.79%, 1.54% and 1.29%, respectively, and the total optical loss of LDI-LDV is 5.374 cm−1, 5.362 cm−1, 5.355 cm−1, 5.321 cm−1 and 5.241 cm−1, respectively. It can be seen that optical confinement factor and total optical loss decrease in turn when the thickness of QW decreases from 4.5 nm to 2.5 nm. Thus, the reduction of optical loss should be a reason for the increase of slope efficiency of LDI-LDV. Actually, the OCF is defined as the ratio of optical field located in the QW layers. On one side, the active region will decrease when the thickness of QW decreases, and the OCF will deceases. On the other side, the full width at half maximum of the optical field distribution curve decreases indicating a more concentrated optical field due to a larger refractive index of QW with higher indium content, and then optical field distributed in QW layers may also decrease. Therefore, combined action results in a decrease of OCF in this study for LD1I-LDV. Moreover, the change of full width at half maximum of the optical field distribution curve will give a clue for the reason of the reduction of optical loss. As shown in Fig. 12(b), the value of FWHM of the optical field distribution curve decreases from 340.4 nm to 325.8 nm when the thickness of quantum well layer decreases from 4.5 nm to 2.5 nm. A smaller FWHM of the optical field distribution presents a more concentrated optical field. Actually, the indium content of quantum well of LDI-LDV increases in turn, indicating that the refractive index of quantum well increases, thus optical field can be confined better. For the more concentrated optical field, there is less proportion of optical field distributed in p-type area, which leads to a smaller optical loss. This result indicates that using a thinner quantum well with higher indium content is beneficial to obtain a well concentrated optical field and a reduced the optical loss, and that is the important reason for the improvement of slope efficiency.
It is interesting to note that the threshold current of LDI-LDV increases in turn with decreasing quantum well thickness. Actually, the reduction of luminescence efficiency caused by poor material quality has not been taken into account during theoretical calculation of LDI-LDV. It is guessed that the threshold current of LD in simulation with thicker quantum well layers will be worse if considering the possible change of material quality for QWs with higher indium composition of the InGaN alloy. According to this idea, another device, named as LDI-1, is also designed and calculated, whose structure parameters are the same as LDI except for the lower carrier life time. During calculation, carrier life time of LDI-1 is set to be a tenth compared with LDI. Figure 13 shows the calculation result of LDI and LDI-1. The threshold current, optical output power under 1.2 A and slope efficiency of LDI-1 is 176.4 mA, 839.7 mW and 0.82 W/A, respectively. These values are much worse than those of LDI-LDV. This result demonstrates that using thinner quantum wells with higher indium content is indeed only in a certain range helpful to improve the device performance if the change of alloy material quality is considered during the calculation.
4. Conclusion
GaN-based blue LDs with varied quantum well thickness and nearly unchanged emission wavelength are investigated experimentally and theoretically. It is found that device performance can be improved using thinner quantum wells with higher indium content in a certain range, which may be attributed to the two factors. One factor is that luminescence efficiency of MQW with thinner quantum wells is improved because of reduction of the nonradiative recombination and improvement of the quantum well’s homogeneity. The other reason is that the optical loss can be reduced due to a better concentrated optical field distributed in LD when using a relative thinner quantum well with higher indium content. Finally, a laser diode is made which threshold current density is lower to 0.69 kA/cm2, the slope efficiency reaches to 1.39 W/A. A long lifetime of more than 10000 hours has been also reached at a fixed injection current of 1.2 A under a room-temperature continuous-wave operation.
Funding
Beijing Municipal Science and Technology Commission , Administrative Commission of Zhongguancun Science Park (Z211100004821001); National Natural Science Foundation of China (62034008, 62074142, 62074140, 61974162,61904172, 61874175, 62127807, U21B2061); Jiangsu Provincial Key Research and Development Program(BE2021008-1); Beijing Nova Program (202093); Strategic Priority Research Program of Chinese Academy of Sciences (XDB43030101); Youth Innovation Promotion Association of the Chinese Academy of Sciences (2019115).
Disclosures
The authors declare that there are no conflicts of interest related to this article.
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.
References
1. Y. Ishige, H. Hashimoto, N. Hayamizu, N. Matsumoto, F. Nishino, M. Kaneko, H. Nasu, G. Gajdátsy, A. Cserteg, K. l. Jahn, T. Hirao, H. Matsuo, and R. Konishi, “Blue laser-assisted kW-class CW NIR fiber laser system for high quality copper welding,” Proc. SPIE 11668, 116680M (2021). [CrossRef]
2. H. Z. Wang, Y. Kawahito, R. Yoshida, Y. Nakashima, and K. Shiokawa, “Development of a high-power blue laser (445 nm) for material processing,” Opt. Lett. 42(12), 2251–2254 (2017). [CrossRef]
3. P. J. Qiu, G. G. Cui, Z. Y. Qian, S. J. Zhu, X. Y. Shan, Z. Q. Zhao, X. L. Zhou, X. G. Cui, and P. F. Tian, “4.0 Gbps visible light communication in a foggy environment based on a blue laser diode,” Opt. Express 29(9), 14163–14173 (2021). [CrossRef]
4. J. R. D. Retamal, H. M. Oubei, B. Janjua, Y. C. Chi, H. Y. Wang, C. T. Tsai, T. K. Ng, D. H. Hsieh, H. H. Kuo, M. S. Alouini, J. H. He, G. R. Lin, and B. S. Ooi, “4-Gbit/s visible light communication link based on 16-QAM OFDM transmission over remote phosphor-film converted white light by using blue laser diode,” Opt. Express 23(26), 33656–33666 (2015). [CrossRef]
5. Y. Nakatsu, Y. Nagao, K. Kozuru, T. Hirao, E. Okahisa, S. Masui, T. Yanamoto, and S. Nagahama, “High efficiency blue and green laser diodes for laser displays,” Proc. SPIE 10918, 109181D (2019). [CrossRef]
6. Y. Hirano, Y. Tani, M. Ishida, S. Miyamoto, and A. Ariyoshi, “Red, green and blue laser diodes for display applications,” SID Symposium Digest of Technical Papers 52(1), 673–676 (2021). [CrossRef]
7. A. D. Ludlow, M. M. Boyd, J. Ye, E. Peik, and P O Schmidt, “Optical atomic clocks,” Rev. Mod. Phys. 87(2), 637–701 (2015). [CrossRef]
8. J.N. Tinsley, S. Bandarupally, J.P. Penttinen, S. Manzoor, S. Ranta, L. Salvi, M. Guina, and N. Poli, “Watt-level blue light for precision spectroscopy, laser cooling and trapping of strontium and cadmium atoms,” Opt. Express 29(16), 25462–25476 (2021). [CrossRef]
9. S. Nakamura, S. Masayuki, N. Shin-ichi, I. Naruhito, Y. Takao, M. Toshio, K. Hiroyuki, and S. Yasunobu, “InGaN-based multi-quantum-well-structure laser diodes,” J. Appl. Phys. 35(1B), L74–L76 (1996). [CrossRef]
10. S. Nakamura, S. Masayuki, N. Shin-ichi, I. Naruhito, Y. Takao, M. Toshio, K. Hiroyuki, S. Yasunobu, K. Tokuya, U. Hitoshi, S. Masahiko, and C. Kazuyuki, “Violet InGaN/GaN/AlGaN-based laser diodes with an output power of 420 mW,” J. Appl. Phys. 37(6A), L627–L629 (1998). [CrossRef]
11. I. Akasaki, S. Sota, H. Sakai, T. Tanaka, M. Koike, and H. Amano, “Shortest wavelength semiconductor laser diode,” Electron. Lett. 32(12), 1105–1106 (1996). [CrossRef]
12. A. Hangleiter, “Gain coefficient and lasing threshold in GaN-based lasers,” EMIS Datarev. Ser. 23, 603 (1999).
13. S. Gwyn, S. Watson, T. Slight, M. Knapp, S. Viola, P. Ivanov, W.K. Zhang, A. Yadav, E. Rafailov, M. Haji, K. E Doherty, S. Stanczyk, S. Grzanka, P. Perlin, S. P Najda, M. Leszczyski, and A.E. Kell, “Dynamic device characteristics and linewidth measurement of InGaN/GaN laser diodes,” IEEE Photonics. J. 13(1), 1–10 (2021). [CrossRef]
14. L. Uhlig, M. Wachs, D.J. Kunzmann, and U.T. Schwarz, “Spectral-temporal dynamics of (Al,In)GaN laser diode,” Opt. Express 28(2), 1771 (2020). [CrossRef]
15. V. Hoffmann, A. Knauer, C. Brunner, S. Einfeldt, M. Weyers, G. Tränkle, K. Haberland, J. T. Zettler, and M. Kneissl, “Uniformity of the wafer surface temperature during MOVPE growth of GaN-based laser diode structures on GaN and sapphire substrate,” J. Cryst. Growth. 315(1), 5–9 (2011). [CrossRef]
16. M. Malinverni, J.-M. Lamy, D. Martin, E. Feltin, J. Dorsaz, A. Castiglia, M. Rossetti, M. Duelk, C. Velez, and N. Grandjean, “Low temperature p-type doping of (Al)GaN layers using ammonia molecular beam epitaxy for InGaN laser diodes,” Appl. Phys. Lett. 105(24), 241103 (2014). [CrossRef]
17. S. Stańczyk, T. Czyszanowski, A. Kafar, R. Czernecki, G. Targowski, M. Leszczynski, T. Suski, R. Kucharski, and P. Perlin, “InGaN laser diodes with reduced AlGaN cladding thickness fabricated on GaN plasmonic substrate,” Appl. Phys. Lett. 102(15), 151102 (2013). [CrossRef]
18. A. Castiglia, J. F. Carlin, E. Feltin, G. Cosendey, J. Dorsaz, and N. Grandjean, “Emission characteristics of GaN-based blue lasers including a lattice matched Al0.83In0.17N optical blocking layer for improved optical beam quality,” Appl. Phys. Lett. 97(11), 111104 (2010). [CrossRef]
19. G. Muziol, H. Turski, M. Siekacz, S. Grzanka, P. Perlin, and C. Skierbiszewski, “Elimination of leakage of optical modes to GaN substrate in nitride laser diodes using a thick InGaN waveguide,” Appl. Phys. Express 9(9), 092103 (2016). [CrossRef]
20. G. Alahyarizadeh, M. Amirhoseiny, and Z. Hassan, “Effect of QW thickness and numbers on performance characteristics of deep violet InGaN MQW lasers,” Int. J. Mod. Phys. B 29(13), 1550081 (2015). [CrossRef]
21. N. A. Shapiro, P. Perlin, C. Kisielowski, L. S. Mattos, J. W. Yang, and E. R. Weber, “The effects of indium concentration and well-thickness on the mechanisms of radiative recombination in InxGa1−xN quantum wells,” MRS Internet J. Nitride Semicond. Res. 5(1), e1 (2000). [CrossRef]
22. M. Smith, K.C. Zeng, J.Y. Lin, H.X. Jiang, A. Salvador, G. Popovici, H. Tang, W. Kim, H. Morkoc, A. Khan, and Q. Chen, “Effects of well thickness on the light emission in InGaN/GaN and GaN/AlGaN multiple quantum wells,” in Digest of the IEEE/LEOS Summer Topical Meeting: Vertical-Cavity Lasers/Technologies for a Global Information Infrastructure/WDM Components Technology/Advanced Semiconductor Lasers and Application, 1997, pp. 31–32.
23. G. Alahyarizadeh, Z. Hassan, S. M. Thahab, F. K. Yam, and A. J. Ghazaic, “Numerical study of performance characteristics of deep violet InGaN DQW laser diodes with AlInGaN quaternary multi quantum barrier electron blocking layer,” Optik 124(24), 6765–6768 (2013). [CrossRef]
24. G. Alahyarizadeh, Z. Hassan, S. M. Thahab, F. K. Yam, and A. J. Ghazaic, “Performance characteristics of deep violet InGaN DQW laser diodes with InGaN/GaN superlattice waveguide layers,” Optik 125(1), 341–344 (2014). [CrossRef]
25. J. Jeschke, U. Zeimer, L. Redaelli, S. Einfeldt, M. Kneissl, and M. Weyers, “Effect of quantum well non-uniformities on lasing threshold, linewidth, and lateral near field filamentation in violet (Al,In)GaN laser diodes,” Appl. Phys. Lett. 105(17), 173501 (2014). [CrossRef]
26. J. Yang, D. G. Zhao, D. S. Jiang, P. Chen, J. J. Zhu, Z. S. Liu, F. Liang, S. T. Liu, and Y. Xing, “Suppression the formation of V-pits in InGaN/ GaN multi-quantum well growth and its effect on the performance of GaN based laser diodes,” J. Alloy. Compd. 822, 153571 (2020). [CrossRef]
27. L. Liu, L. Wang, D. Li, N. Y. Liu, L. Li, W. Y. Cao, W. Yang, C. H. Wan, W. H. Chen, W. M. Du, X. D. Hu, and Z. C. Feng, “Influence of indium composition in the prestrained InGaN interlayer on the strain relaxation of InGaN/GaN multiple quantum wells in laser diode structures,” J. Appl. Phys. 109(7), 073106 (2011). [CrossRef]
28. G. Muziol, M. Hajdel, M. Siekacz, K. Szkudlarek, S. Stanczyk, H. Turski, and C. Skierbiszewski, “Optical properties of III-nitride laser diodes with wide InGaN quantum wells,” Appl. Phys. Express 12(7), 072003 (2019). [CrossRef]
29. M. Gladysiewicz, C. Skierbiszewski, and R. Kudrawiec, “Material gain in polar GaInN and AlGaN quantum wells: How to overcome the ‘dead’ width for light emitters in these QW systems?” IEEE J. Sel. Top. Quantum Electron. 28(1), 1–9 (2022). [CrossRef]
30. G. Muziol, M. Hajdel, M. Siekacz, H. Turski, K. Pieniak, A. Bercha, W. Trzeciakowski, R. Kudrawiec, T. Suski, and C. Skierbiszewski, “III-nitride optoelectronic devices containing wide quantum wells-unexpectedly efficient light sources,” J. J. Appl. Phys. 61(SA), SA0801 (2022). [CrossRef]
31. Y. C. Yeo, T. C. Chong, M. F. Li, and W. J. Fan, “Analysis of optical gain and threshold current density of wurtzite InGaN/GaN/AlGaN quantum well lasers,” J. Appl. Phys. 84(4), 1813–1819 (1998). [CrossRef]
32. S.M. Thahab, H.A. Hassan, and Z. Hassan, “Effect of varying quantum well thickness on the performance of InGaN/GaN single quantum well laser diode,” in AIP Conference Proceedings1017(1), 149 (2008).
33. M. Zhou and D.G. Zhao, “Barrier and well thickness designing of InGaN/GaN multiple quantum well for better performances of GaN based laser diode,” Acta Phys. Sin. 65(7), 077802 (2016). [CrossRef]
34. J. Piprek, T. Peng, G. Qui, and J. 0. Olowolafe, “Energy Gap Bowing and Refractive Index Spectrum of AlInN and AlGaInN,” in Proceedings of the IEEE Twenty-Fourth International Symposium on Compound Semiconductors, 1997, pp. 227–230.
35. G.M. Laws, E.C. Larkins, I. Harrison, C. Molloy, and D. Somerford, “Improved refractive index formulas for the AlxGa1−xN and InyGa1−yN alloys,” J. Appl. Phys. 89(2), 1108–1115 (2001). [CrossRef]
36. C. Bazioti, E. Papadomanolaki, T. Kehagias, T. Walther, J. Smalc-Koziorowska, E. Pavlidou, P. Komninou, T. Karakostas, E. Iliopoulos, and G. P. Dimitrakopulos, “Defects, strain relaxation, and compositional grading in high indium content InGaN epilayers grown by molecular beam epitaxy,” J. Appl. Phys. 118(15), 155301 (2015). [CrossRef]
37. N. A. El-Masry, E. L. Piner, S. X. Liu, and S. M. Bedair, “Phase separation in InGaN grown by metalorganic chemical vapor deposition,” Appl. Phys. Lett. 72(1), 40–42 (1998). [CrossRef]
38. B. N. Pantha, J. Li, J. Y. Lin, and H. X. Jiang, “Evolution of phase separation in In-rich InGaN alloys,” Appl. Phys. Lett. 96(23), 232105 (2010). [CrossRef]
39. L. J. Brillson, T. M. Levin, G. H. Jessen, and F. A. Ponce, “Localized states at InGaN/GaN quantum well interfaces,” Appl. Phys. Lett. 75(24), 3835–3837 (1999). [CrossRef]
40. S. H. Feng, S. C. Cheng, Y. Y. Chung, C. C. Yang, Y. S. Lin, C. Hsu, K. J. Ma, and J. I. Chyi, “Impact of localized states on the recombination dynamics in InGaN/GaN quantum well structures,” J. Appl. Phys. 92(8), 4441–4448 (2002). [CrossRef]
41. J. Binder, K. P. Korona, A. Wysmołek, M. Kaminska, K. Kohler, L. Kirste, O. Ambacher, M. Zaja, and R. Dwilinski, “Dynamics of thermalization in GaInN/GaN quantum wells grown on ammonothermal GaN,” J. Appl. Phys. 114(22), 223504 (2013). [CrossRef]
42. S. Grzanka, G. Franssen, G. Targowski, R. Czernecki, A. Khachapuridze, I. Makarowa, R. Wisniewska, P. Mensz, P. Perlin, and T. Suski, “Carrier recombination mechanisms in nitride single quantum well light-emitting diodes revealed by photo- and electroluminescence,” J. Appl. Phys. 104(9), 094504 (2008). [CrossRef]
43. T. Mukai, M. Yamada, and S. Nakamura, “Current and Temperature Dependences of Electroluminescence of InGaN-Based UV/Blue/Green Light-Emitting Diodes,” J. J. Appl. Phys. 37(11B), L1358–L1361 (1998). [CrossRef]
44. S. Cmchibu, T. Deguchi, T. Sota, K. Wada, and S. Nakamura, “Localized Excitons in InGaN,” in MRS Online Proceedings Library482, 648–659 (1997).
