Open Access
Add to BibSonomy
Abstract
In this report, we propose GaN-based vertical cavity surface emitting lasers with a p-GaN/n-GaN/p-GaN (PNP-GaN) structured current spreading layer. The PNP-GaN current spreading layer can generate the energy band barrier in the valence band because of the modulated doping type, which is able to favor the current spreading into the aperture. By using the PNP-GaN current spreading layer, the thickness for the optically absorptive ITO current spreading layer can be reduced to decrease internal loss and then enhance the lasing power. Furthermore, we investigate the impact of the doping concentration, the thickness and the position for the inserted n-GaN layer on the lateral hole confinement capability, the lasing power, and the optimization strategy. Our investigations also report that the optimized PNP-GaN structure will suppress the thermal droop of the lasing power for our proposed VCSELs.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
GaN-based vertical cavity surface emitting lasers (VCSELs) have the characteristics of small volume, low threshold current, two-dimensional arrays and small divergence for circular beam in contrast to edge emitting laser diodes. As a result, GaN VCSELs are considered to be promising candidates for varieties of applications in portable displays, solid-state lighting, biosensors, display technologies, high-density optical storage and visible light communication [1–4]. However, at the current stage, a very important target is to increase the lasing power for GaN based VCSELs, and this is doable if a very strong current crowding effect in the cavity is enabled so that the population inversion can be obtained [5,6]. For achieving that, a very effective method is to fabricate GaN VCSELs on n-type conductive AlInN/GaN DBRs, and hence GaN VCSEL with vertical-current-injection paths is obtained [7]. Nevertheless, AlGaNIn based DBRs that maintain both high optical reflectivity and electrical conductivity are still challenging. Therefore, most GaN VCSEL designs are still based on insulating DBRs, which require lateral-current-injection paths. Then, the current injection efficiency into the aperture becomes low. Meanwhile, growing AlN/GaN DRBs with p-type conductivity is even more difficult, and insulating dielectric DBRs are also often used. Therefore, to more favor the current injection into the cavity, an aperture that is surrounded by an insulating layer is utilized. The p-type current spreading layer is then fabricated on the insulator, which further increases the difficulty of the effective population inversion in the cavity for GaN VCSELs with lateral-current-injection paths [8–10]. To facilitate the current transport into the aperture region, an indium tin oxide (ITO) current spreading layer is commonly used for GaN-based VCSELs [11–13]. To enhance the current injection efficiency into the aperture, the less electrical-resistive ITO layer is needed, which can be realized by increasing the ITO layer thickness. However, ITO has a very strong optical absorption, the numbers for which range from 1×103 m−1 to 2×106 m−1 according to different fabrication processes [14–17]. To reduce the optical absorption arising from the thick ITO layer, tunnel junctions have ever been used to make intracavity contacts [18–21]. Nevertheless, due to the lower Mg ionization efficiency, the inter-band tunneling efficiency for p+-GaN/n+-GaN tunnel junction shall be low, leading to a larger electrical resistance, and moreover, the fabrication of tunnel-junction-based intracavity contacts requires multiple epitaxial growth processes, which further makes the VCSEL fabrication process more complex. Therefore, it is necessary to find alternative VCSEL structures which can better spread the current to the cavity center even by using thin ITO layers.
2. Device structure and important parameters
In this work, we propose adopting p-GaN/n-GaN/p-GaN (PNP-GaN) current spreading layer for GaN-based VCSELs. The studied VCSELs only differ in the ITO layer and the PNP-GaN junction. The VCSELs can generate laser at the wavelength of 451 nm. They possess 15 pairs of SiO2/Ta2O5 top distributed Bragg reflectors (DBRs) and 20 pairs of AlN/GaN bottom DBRs. The cavity length is designed to be 7λ cavity. The cavity for VCSEL A consists of a 976-nm-thick n-GaN layer with the Si doping concentration of 5 × 1018 cm−3, 5 pairs of In0.21Ga0.79N (3 nm)/GaN (4 nm) multiple quantum wells (MQWs), a 20-nm-thick p-type Al0.18Ga0.82N electron blocking layer (p-EBL) with the Mg doping concentration of 1×1019 cm−3, a 175-nm-thick p-GaN layer with the Mg doping concentration of 1.2×1019 cm−3 and a 10-nm-thick ITO layer [see Fig. 1(a)]. We empirically set the ionization efficiency of Si dopants and Mg dopants to be 100% and 1% for n-GaN layer and p-GaN layer, respectively. The proposed VCSEL is identical with VCSEL A except that the thin n-GaN layer is inserted into the p-GaN layer to form the PNP-GaN junction. All the VCSELs employ 20-nm-thick SiNx layers to form the apertures for confining current and light. Note, the SiNx layer is exactly designed underneath the ITO layer to suppress the hole diffusion to the mesa edge [see Fig. 1(b1)]. The radius of the circular aperture is set to 5 µm in our devices.
Fig. 1. (a) Equivalent circuit for the conventional VCSEL, (b1) simplified equivalent circuit for the VCSEL with PNP-structure, (b2) schematic structure and schematic band diagram for the PNP-GaN junction, for which φ stands for the barrier height of the PNP-junction for holes.
First, we shall elucidate the origin for the better current spreading effect, which is favored by the (p-GaN/n-GaN/p-GaN) PNP-GaN current spreading layer. The equivalent circuit for VCSEL A is presented in Fig. 1(a), which shows the lateral current injection scheme for GaN based VCSELs. We also show the simplified equivalent circuit for a VCSEL with PNP-GaN junction in Fig. 1(b1), in which we define J1 and J2 as the current level in the two paths for VCSEL. A better current spreading shall ensure a small J1/J2. The relationship between J1 and J2 can be formulated by Eq. (1) as follows,
Crosslight PICS3D is used to probe the sensitivity of the current transport to different PNP-GaN designs, which can self-consistently solve different equations such as drift-diffusion equation, Poisson equations etc. [23]. The Auger recombination coefficient is set to be 1 × 10−43 m6/s [24]. The Shockley-Read-Hall (SRH) recombination lifetime is set to be 1 × 10−8 s [25]. The polarization level is set to 40% for computing the polarization charges for III-nitride based devices [26]. The ratio of conduction band offset/valence band offset is set to 70/30 for our InGaN/GaN MQWs [27]. The optical absorption coefficient for the ITO current spreading layer is set to 7 × 104 m−1 that is in the range of our investigation. The average optical loss due to other passive layers is set to be 1000 m−1 in our models.
3. Results and discussions
3.1 Proof of the PNP-GaN junction on current spreading improvement
To prove the impact of the PNP-GaN junction on the current spreading improvement, we design VCSEL B which is identical with VCSEL A except that VCSEL B has a 20-nm-thick n-type GaN layer with the Si doping concentration of 4 × 1017 cm−3 inserted into the p-GaN layer. To eliminate the impact of temperature on the current injection and stimulated recombination, we set the temperature to 300 K. The current density-voltage (J-V) characteristics and the lasing power density for all modes as the function of the injection current density are shown in Fig. 2(a). We can see that VCSEL B consumes more bias because the PNP-GaN junction produces RPNP. It can also be seen from Fig. 2(a) that VCSEL B exhibits a higher lasing power density than VCESL A. Moreover, the threshold current density for both VCSELs is ∼ 6 kA/cm2, which agrees well with those experimental measured numbers [28,29], and this validate the parameters for DBRs, internal loss, Auger recombination etc in our models. Figure 2(b) indicates that the wall-plug efficiency (WPE) is improved by inserting the current spreading layer. To explain the observation, the energy band for VCSEL B at the current density of 8800 A/cm2 is demonstrated in Fig. 2(c) and the barrier in the valence band for holes is observed clearly at the position of approximately 0.2 µm. Such barrier is generated by the inserted PNP junction and will induce RPNP, thus better spreading the holes to the cavity center.
Fig. 2. (a) Lasing power density and applied voltage, (b) wall-plug efficiency at different current levels for VCSELs A and B, (c) energy band of the MQW and p-type layers for VCSEL B at the current density of 8800 A/cm2.
Figure 3(a) then demonstrates the lateral distribution for holes in the quantum well closest to the p-GaN layer [i.e., the last quantum well] for VCSELs A and B. We can see that compared with VCSEL A, VCSEL B has larger hole concentration in the aperture region, which is exactly attributed to the enhanced current spreading effect. The higher hole concentration in the aperture correspondingly yields the higher net modal gain and the improved stimulated recombination rate for VCSEL B as shown in Figs. 3(b) and 3(c), respectively. As a result, the PNP junction in the p-GaN layer can make a great contribution to improve the current spreading and the lasing power for VCSELs. However, careful investigations infer that the position for the peak hole concentration and the net modal gain in Figs. 3(a) and 3(b) does not coincide that for the stimulated recombination in Fig. 3(c). Because of the resonant properties for stimulated photons, the stimulated emission is a compromised effect of net modal gain and the near field pattern. Figure 3(c) also shows the near field patterns of all optical modes for VCSELs A and B at the current density of 8800 A/cm2. We can see that the peak position for the near field patterns is not affected by the PNP-GaN junction and almost identical to that for the stimulated recombination rate. The conclusion here is also applied to other devices which will be discussed subsequently. Note, we can obtain three optical modes that are able to receive stimulated recombination and the strongest lasing power is produced by the third-order mode.
Fig. 3. (a) Lateral distribution of hole concentration, (b) lateral distribution of net modal gain, (c) lateral distribution of stimulated recombination rate and the near field patterns in the last quantum well for VCSELs A and B at the current density of 8800 A/cm2.
3.2 Impact of the doping concentration in the n-type insertion layer on the device performance
It is well known that the energy barrier for PNP-GaN junction can be influenced by the doping concentration in the n-GaN layer. Therefore, we vary the Si doping concentrations of 4 × 1017, 8 × 1017, 1.2 × 1018, 1.6 × 1018, 2 × 1018 cm−3 in the n-type insertion layer for the PNP-GaN region for VCESLs D1, D2, D3, D4 and D5, respectively. The thickness of the n-type insertion layer for VCSELs D1, D2, D3, D4 and D5 is set to 20 nm. We set the temperature to 300 K for the purpose of eliminating the impact of temperature on the current injection and stimulated recombination.
As can be seen from the J-V curves in Fig. 4(a), the forward bias increases as the increased Si doping concentration in the PNP-GaN junction for the studied VCSELs, which indicates the value of RPNP gets larger. Meanwhile, when compared with VCSEL A, the lasing power density for all VCSELs with PNP-GaN structure can be significantly enhanced due to the better current spreading. However, it can also be found that the lasing power density for all modes decreases when the Si doping concentration in the PNP-GaN junction is larger than 1.2 × 1018 cm−3 for VCSEL D4 and D5. In order to figure out the physical mechanism of the above results, we show the valence band barrier heights formed in different PNP-GaN junctions in Table 1, which are 0.2199 eV, 0.2354 eV, 0.2420 eV, 0.2460 eV, 0.2491 eV for VCSELs D1, D2, D3, D4 and D5, respectively. Thus, the increased Si doping concentration for the PNP-GaN junction increases the valence band barrier height for holes, thus leading to the increased RPNP and more applied bias. As has been predicted previously, the enhanced RPNP can decrease J1/J2, meaning the current spreading effect increases as the increased Si-doping concentration in the PNP-GaN junction. Figure 4(b) shows that VCSEL D1 presents the highest wall-plug efficiency, while VCSEL D5 shows the lowest one. The reduced WPE for VCSEL D5 is due to the much increased barrier height when the doping concentration for the n-type insertion layer is 2 × 1018 cm−3. Therefore, by analyzing Figs. 4(a) and 4(b), we summarize that the doping concentration in the n-type insertion layer shall be optimized, and in this case, we recommend VCSEL D3.
Fig. 4. (a) Lasing power density and applied voltage (inset shows the power density at the current density of 8800 A/cm2) (b) wall-plug efficiency at different current levels (inset shows the wall-plug efficiency at the current density of 8800 A/cm2) for VCSELs A, D1, D2, D3, D4 and D5.
Table 1. Valence band barrier heights (φ) of holes in the PNP-GaN junctions for VCSELs D1, D2, D3, D4 and D5
To address the underlying reason for Figs. 4(a) and 4(b), Figs. 5(a)–5(c) demonstrate the profiles for the lateral holes, the lateral net modal gain and the lateral stimulated recombination rate in the last quantum well at the current density of 8800 A/cm2 for all VCSELs. As can be seen from Fig. 5(a), a higher Si doping concentration for the inserted n-GaN layer helps to increase the hole concentration near the aperture center, which is well attributed to the enhanced current spreading effect, agreeing well with our previous analysis. However, insightful investigations into Fig. 5(a) also indicate that the adoption of PNP-GaN junction hinders the hole injection at the position between 4 µm to 5 µm in the aperture region, which further proves the current spreading effect is enabled by the PNP-GaN junction. Figures 5(b) and 5(c) demonstrate the profiles for net modal gain and stimulated radiative recombination rate, respectively, both of which show the enhancement especially near the aperture center as the Si doping concentration in the PNP-GaN junction increases. Nevertheless, the net modal gain and the stimulated radiative recombination rate for VCSEL D4 and D5 slightly decrease in the aperture periphery, which is caused by the low hole concentration therein.
Fig. 5. (a) Lateral distribution of hole concentration, (b) lateral distribution of net modal gain, (c) lateral distribution of stimulated radiative recombination rate in the last quantum well for VCSELs A, D1, D2, D3, D4 and D5. Data are computed at the current density of 8800 A/cm2.
Hence, it can be summarized here that the proper increase of the Si doping concentration in the PNP-GaN junction can promote the current spreading effect. However, a too high Si dosage level might sacrifice the hole injection especially in the aperture periphery, which also agrees with our proposal such that a big RPNP can significantly increase the current density in the aperture center.
3.3 Impact of the thickness for the n-type insertion layer on the device performance
Besides the doping concentration, the value of RPNP is also affected by the n-GaN layer thickness in the PNP-GaN junction. To address the point, VCSELs T1, T2, T3, T4 and T5 are designed, for which the Si doping concentration of 1.2 × 1018 cm−3 is set for PNP-GaN junctions with the n-GaN layer thicknesses of 20 nm, 24 nm, 28 nm, 32 nm and 36 nm, respectively. We set the temperature to 300 K for the purpose of eliminating the impact of temperature on the current injection and stimulated recombination.
The relationship among the lasing power density for all laser modes, the applied bias, the injection current density and the wall-plug efficiency for VCSELs T1, T2, T3, T4 and T5 can be found in Figs. 6(a) and 6(b). It is obvious that the forward voltage becomes increased with the increasing thickness for the n-GaN insertion layer, which is because of the increased valence band barrier height according to Table 2. Figure 6(a) also presents that VCSEL T4 produces the highest lasing power density. The lasing power density decreases if the n-GaN insertion layer is not properly designed, such that a thin n-GaN insertion layer cannot fully exhibit the current spreading effect by the PNP-GaN junction while a too thick n-GaN insertion layer thickness will block the hole injection. Figure 6(b) shows that the WPE is slightly affected by the n-GaN insertion layer thickness with the WPE for VCSEL T3 being the highest.
Fig. 6. (a) Lasing power density and applied voltage (inset shows the power density at the current density of 8800 A/cm2) (b) wall-plug efficiency at different current levels (inset shows the wall-plug efficiency at the current density of 8800 A/cm2) for VCSELs A, T1, T2, T3, T4 and T5.
Table 2. Valence band barrier heights (φ) of holes in the PNP-GaN junctions for VCSELs T1, T2, T3, T4 and T5
We then show the lateral profiles for the hole concentration in the last quantum well at the current density of 8800 A/cm2 for VCSELs A, T1, T2, T3, T4 and T5. The hole concentration profile becomes more uniform as the thickness of the insertion layer increases. However, insightful investigations into Fig. 7(a) also indicate that the adoption of more than 28 nm thick n-GaN layer in PNP-GaN junction hinders the hole injection at the position between 4 µm to 5 µm in the aperture region, which is because the holes prefer the lateral transport rather than the vertical transport when encountering a high barrier height in the valence band. The profiles for the lateral net modal gain and the stimulated radiative recombination rate at the current density of 8800 A/cm2 are presented in Figs. 7(b) and 7(c), respectively. Being consistent with the hole concentration profiles, both the net modal gain and the stimulated radiative recombination rate are enhanced in the aperture centers as the n-GaN insertion layer becomes thick. Nevertheless, as has been mentioned previously, the adopted PNP-GaN junction spreads the current by hindering the vertical hole injection, and therefore the hole concentration in the aperture peripheries becomes small especially for VCSEL T5. As a result, the lasing power density becomes even worse for VCSEL T5.
Fig. 7. (a) Lateral distribution of hole concentration, (b) lateral distribution of net modal gain, (c) lateral distribution of stimulated radiative recombination rate in the last quantum well for VCSELs A, T1, T2, T3, T4 and T5. Data are computed at the current density of 8800 A/cm2.
Therefore, it can be concluded that, for maintaining both the excellent current spreading effect and decent hole injection, a proper thickness for the n-GaN insertion layer is required for the PNP-GaN junction. A too thick n-GaN insertion layer increases the valence band barrier height and sacrifices the hole injection especially in the aperture periphery, which agrees well with our model such that a big RPNP can significantly increase the current density in the aperture center.
3.4 Impact of the position for the n-type insertion layer on the device performance
All our previously discussed VCSELs have the n-GaN layer precisely contacted with the buried insulator [see Fig. 1(b)]. However, the PNP-GaN junction has different values of resistivity in the n-GaN layer and in the p-GaN layer. Therefore, it is speculated that the position of the n-GaN layer might have impact on the current transport. Then, we define the distance between the n-GaN insertion layer and the buried insulator as p, the numbers for which are 0 nm, 10 nm, 20 nm, 30 nm and 40 nm for VCSELs P1, P2, P3, P4 and P5, respectively. The doping concentration and the thickness for the n-GaN insertion layer are set to 1.2 × 1018 cm−3 and 28 nm, respectively for the investigated VCSELs. We set the temperature to 300 K for the purpose of eliminating the impact of temperature on the current injection and stimulated recombination.
The calculated lasing power density for all lasing modes, the forward voltage and the wall-plug efficiency in terms of the injection current density are shown in Fig. 8(a). It is found that the forward voltage is less affected by the PNP-GaN junctions with different values of p. This makes sense because the architectural configuration does not remarkably change. Figure 8(a) also shows that the lasing power density increases with p and then reaches a saturation level. In addition, the WPE is more sensitive to the position of the n-GaN insertion layer, such that VCSELs P4 and P5 show the most improved WPE when compared with VCSEL A.
Fig. 8. (a) Lasing power density and applied voltage (inset shows the power density at the current density of 8800 A/cm2) (b) wall-plug efficiency at different current levels (inset shows the wall-plug efficiency at the current density of 8800 A/cm2) for VCSELs A, P1, P2, P3, P4 and P5.
For better addressing the origin in Fig. 8, we show the valence band barrier heights in different PNP-GaN junctions in Table 3. Being consistent with the current-voltage characteristics in Fig. 8, the energy barrier heights for the PNP-GaN junctions do not significantly vary. The slight variation in the current-voltage characteristics for VCSELs P1, P2, P3, P4 and P5 is likely caused by the mentioned RCL, which impacts the local carrier density, the net modal gain and the stimulated radiative recombination rate.
Table 3. Valence band barrier heights (φ) of holes in the PNP-GaN junctions for VCSELs P1, P2, P3, P4 and P5
To further prove that point, the profiles for the lateral hole concentration, the lateral gain and the lateral stimulated radiative recombination rate in the last quantum well at the current density of 8800 A/cm2 for all the studied VCSELs are shown in Figs. 9(a)–9(c). As shown in Fig. 9(a), as the value of p increases, the current spreading effect is enhanced and then stabilized, which is well attributed to the decreased RCL. Note, the increased p1-GaN layer thickness can decrease RCL because of the increased cross-sectional area for the lateral current. The decreased RCL can then reduce J1/J2 according to Eq. (1). However, further increased p will no longer enhance the current spreading. The same observations are also obtained for Figs. 9(b) and 9(c). Our results indicate that the optimized VCSEL design shall possess a none-zero p. On the other hand, the current spreading effect for the proposed VCSELs P1, P2, P3, P4 and P5 can be no more explained by using RPNP.
Fig. 9. (a) Lateral distribution of hole concentration, (b) lateral distribution of net modal gain, (c) lateral distribution of stimulated radiative recombination rate in the last quantum well for VCSELs A, P1, P2, P3, P4 and P5. Data are computed at the current density of 8800 A/cm2.
3.5 Comprehensive evaluations for different PNP-GaN junction designs
In our previous discussions, we have proven the sensitivity of the current spreading effect and the lasing power to the thickness, the doping concentration and the position for the n-GaN insertion layer, and now we shall probe if a thin ITO current spreading layer can be used for our proposed VCSELs. We select the device structures for VCSEL A without PNP-GaN junction and P4 in Fig. 9. Note, VCSEL P4 possesses the largest lasing power density in all our designs, which is selected to probe if a thin ITO layer can still maintain high lasing power density after using the PNP-GaN current spreading layer. The ITO layer thicknesses are set to 10 nm and 40 nm for VCSEL A. The ITO layer thickness is set to 10 nm for VCSEL P4. We set the temperature to 300 K for the purpose of eliminating the impact of temperature on the current injection and stimulated recombination.
The lasing power in terms of the absorption coefficient for the ITO layer at the current density of 8800 A/cm2 is shown in Fig. 10. The composition between VCSELs A with 10 nm ITO layer and 40 nm ITO layer indicates that the ITO thickness can strongly affect the lasing power density. Although a thick ITO layer can enhance the lasing power density for VCSEL A, the lasing power density is very sensitive to the ITO absorption coefficient. Note, the absorption coefficient for the ITO layer is strongly subject to different deposition conditions. Nevertheless, by using the PNP-GaN junction, a much thin ITO layer can be adopted while the enhanced lasing power can be obtained according to Fig. 10(a), such that VCSEL P4 exhibits the largest lasing power. Figure 10(b) represents that the WPE is greatly improved by inserting the PNP current spreading layer when we compare VCSEL A with 10 nm-thick ITO layer and VCSEL P4. The increased ITO thickness can reduce the on-resistance, and thus the WPE for VCSEL A with 40 nm thick ITO shows the largest WPE. However, the lasing power for it is inferior when compared with that for VCSELs P4. As a result, the advantage of PNP-GaN current spreading layer for GaN-based VCSELs is obvious and the proposed structures provide more freedom for fabricating ITO layer. We also believe that the reported VCSELs with 20 nm thick ITO in Refs. [30,31] can further boost the lasing power by using the proposed PNP-GaN current spreading layer in our work.
Fig. 10. (a) Lasing power density, (b) wall-plug efficiency for VCSEL A with 10 nm thick ITO layer, VCSEL A with 40 nm thick ITO layer and VCSEL P4 with 10 nm thick ITO layer in terms of different absorption coefficient.
3.6 Thermal effect for VCSELs with PNP current spreading layer
Our previous discussions show that a properly designed PNP-GaN structure can significantly enhance the lasing power and the WPE for the proposed VCSEL. However, our results also show that the adopted PNP-GaN can also increase the forward bias because of the additional energy barriers. Thus, it is necessary to study the impact whether the PNP-GaN junction can improve the device performance when the thermal models are considered. Table 4 presents the important thermal parameters for nitride based VCSELs.
We select VCSELs A, D1, D3, T1, T3, P1, P4 for in-depth investigations. Details regarding the doping concentration, the thickness for the n-GaN insertion layer, and the distance between the n-GaN insertion layer and the buried insulator can refer to our structural parameters in the previous discussions. The calculated lasing power in term of the injection current for VCSELs A, D1, D3, T1, T3, P1 and P4 are shown in Figs. 12(a), 12(b) and 12(c). All the figures illustrate that the increasing junction temperature cause the roll-over for the lasing power when compared to the cases at the temperature of 300 K. On the other hand, the lasing power also decreases when compared to the counterparts. The decreased lasing power is likely attributed to the increased electron leakage when the junction temperature increases. However, the proposed VCSELs D1, D3, T1, T3, P1 and P4 still possess the enhanced lasing power and WPE when compared with VCSEL A. This further manifests the advantages of the PNP-GaN junction for GaN based VCSELs from the perspective of current spreading effect.
However, according to Figs. 11(a), 11(b) and 11(c) the rollover current at which the lasing power decreases with the further increase of the injection current slightly increases for VCSELs D1, D3, T1, T3, P1 and P4 when compared with VCSEL A. We then selectively calculate and show the temperature distribution profiles for VCSELs A, D3, T3 and P4 in Figs. 12(a), 12(b), 12(c) and 12(d), respectively. The thermal mount is in contact with the GaN substrate and the temperature for the thermal mount is set to 300 K in our calculations. Our results show that VCSEL A has a very poor current spreading and hence the very high local temperature is obtained around the right mesa edge. When the PNP-GaN current spreading layer is adopted, we can see that the temperature distribution is more uniform, such that the most uniform temperature distribution can be obtained for VCSEL P4. However, as we have discussed previously, the adoption of PNP-GaN junction increases the forward bias because of the additional resistance. Hence, the overall temperature becomes high for the proposed VCSELs, e.g., VCSEL P4. We believe the rollover current can be further increased once the even better thermal management can be obtained for the proposed VCSELs.
Fig. 11. Lasing power density and WPE in terms of the injection current density for (a) VCSELs A, D1 and D3, (b) VCSELs A, T1 and T3, and (c) VCSELS A, P1 and P4.
Fig. 12. Temperature mapping profiles for (a) VCSEL A, (b) VCSEL D3, (c) VCSEL T3 and (d) VCSEL P4 at the current density of 11000 A/cm2. The p-EBL, the SiNx buried insulator and the n-GaN insertion layer in the PNP-GaN junction cannot be explicitly shown because of the very small size in the presented geometrical scale.
4. Conclusion
To summarize, we have proposed and demonstrated that the PNP-GaN junction helps to enhance the current spreading effect for InGaN/GaN VCSELs. The current turns to flow into the center for the aperture with the increase of the barrier height, which is enabled by properly increasing the vertical resistance. The valence band barrier height can be affected by the doping concentration and the thickness of the n-GaN insertion layer in the PNP-GaN junction. We also suggest a certain distance between the n-GaN insertion layer and the buried insulator, which can effectively enhance the current diffusion into the cavity center. Our results show that, the adoption of the PNP-GaN current spreading can avoid the usage of thick ITO current spreading layer, thus reducing the optical absorption and increasing the lasing power. Therefore, the proposed PNP-GaN current spreading layer is useful to make high-power GaN-based VCSELs, and the device physics reported here is very important for the laser community.
Funding
National Natural Science Foundation of China (61975051); Program for Top 100 Innovative Talents in Colleges and Universities of Hebei Province (SLRC2017032); Program for 100-Talent-Plan of Hebei Province (E2016100010); Suzhou Institute of Nano-Tech and Nano-Bionics (SINANO) Research Fund of Chinese Academy of Sciences (19ZS02); Tunghsu Group and Hebei University of Technology (HI1909); Graduate Innovation Foundation of Hebei Province (CXZZBS2020027).
Disclosures
The authors declare no conflicts of interest.
References
1. S. Izumi, N. Fuutagawa, T. Hamaguchi, M. Murayama, M. Kuramoto, and H. Narui, “Room-temperature continuous-wave operation of GaN-based vertical-cavity surface-emitting lasers fabricated using epitaxial lateral overgrowth,” Appl. Phys. Express 8(6), 062702 (2015). [CrossRef]
2. M. Kuramoto, S. Kobayashi, K. Tazawa, K. Tanaka, T. Akagi, and T. Saito, “In-phase supermode operation in GaN-based vertical-cavity surface-emitting laser,” Appl. Phys. Lett. 115(4), 041101 (2019). [CrossRef]
3. S. M. Mishkat-Ul-Masabih, A. A. Aragon, M. Monavarian, T. S. Luk, and D. F. Feezell, “Electrically injected nonpolar GaN-based VCSELs with lattice-matched nanoporous distributed Bragg reflector mirrors,” Appl. Phys. Express 12(3), 036504 (2019). [CrossRef]
4. T. Hamaguchi, M. Tanaka, J. Mitomo, H. Nakajima, M. Ito, M. Ohara, N. Kobayashi, K. Fujii, H. Watanabe, S. Satou, R. Koda, and H. Narui, “Lateral optical confinement of GaN-based VCSEL using an atomically smooth monolithic curved mirror,” Sci. Rep. 8(1), 10350 (2018). [CrossRef]
5. Y. Q. Hao, Y. Luo, Y. A. Feng, C. L. Yan, Y. J. Zhao, Y. X. Wang, X. H. Wang, Y. Qu, and G. J. Liu, “Large aperture vertical cavity surface emitting laser with distributed-ring contact,” Appl. Opt. 50(7), 1034–1037 (2011). [CrossRef]
6. V. P. Kalosha, V. A. Shchukin, N. Ledentsov, and N. N. Ledentsov, “Comprehensive Analysis of Electric Properties of Oxide-Confined Vertical-Cavity Surface-Emitting Lasers,” IEEE J. Quantum Electron. 25(6), 1–9 (2019). [CrossRef]
7. W. Muranaga, T. Akagi, R. Fuwa, S. Yoshida, J. Ogimoto, Y. Akatsuka, S. Iwayama, T. Takeuchi, S. Kamiyama, M. Iwaya, and I. Akasaki, “GaN-based vertical-cavity surfaceemitting lasers using n-type conductive AlInN/GaN bottom distributed Bragg reflectors with graded interfaces,” Jpn. J. Appl. Phys. 58(SC), SCCC01 (2019). [CrossRef]
8. Y. Y. Lai, S. C. Huang, T. L. Ho, T. C. Lu, and S. C. Wang, “Numerical analysis on current and optical confinement of III-nitride vertical-cavity surface-emitting lasers,” Opt. Express 22(8), 9789–9797 (2014). [CrossRef]
9. Y. Mei, R.-B. Xu, H. Xu, L.-Y. Ying, Z.-W. Zheng, B.-P. Zhang, M. Li, and J. Zhang, “A comparative study of thermal characteristics of GaN-based VCSELs with three different typical structures,” Semicond. Sci. Technol. 33(1), 015016 (2018). [CrossRef]
10. M. Kuramoto, S. Kobayashi, T. Akagi, K. Tazawa, K. Tanaka, T. Saito, and T. Takeuchi, “Enhancement of slope efficiency and output power in GaN-based vertical-cavity surface-emitting lasers with a SiO2-buried lateral index guide,” Appl. Phys. Lett. 112(11), 111104 (2018). [CrossRef]
11. J. T. Leonard, D. A. Cohen, B. P. Yonkee, R. M. Farrell, S. P. DenBaars, J. S. Speck, and S. Nakamura, “Smooth e-beam-deposited tin-doped indium oxide for III-nitride vertical-cavity surface-emitting laser intracavity contacts,” J. Appl. Phys. (Melville, NY, U. S.) 118(14), 145304 (2015). [CrossRef]
12. Y. Y. Lai, T. C. Chang, Y. C. Li, T. C. Lu, and S. C. Wang, “Electrically Pumped III-N Microcavity Light Emitters Incorporating an Oxide Confinement Aperture,” Nanoscale Res. Lett. 12(1), 15 (2017). [CrossRef]
13. T. C. Lu, C. C. Kao, H. C. Kuo, G. S. Huang, and S. C. Wang, “CW lasing of current injection blue GaN-based vertical cavity surface emitting laser,” Appl. Phys. Lett. 92(14), 141102 (2008). [CrossRef]
14. J. Ederth, A. HultåKer, G. A. Niklasson, P. Heszler, A. R. Vandoorn, M. J. Jongerius, D. Burgard, and C. G. Granqvist, “Thin porous indium tin oxide nanoparticle films: effects of annealing in vacuum and air,” Appl. Phys. A: Mater. Sci. Process. 81(7), 1363–1368 (2005). [CrossRef]
15. L. Y. Ying, X. L. Hu, W. J. Liu, J. Y. Zhang, B. P. Zhang, and H. C. Kuo, “Control of optical loss in GaN-based planar cavities,” Superlattices Microstruct. 88, 561–566 (2015). [CrossRef]
16. M. Cesaria, A. P. Caricato, G. Maruccio, and M. Martino, “Optical analysis of Cr-doped ITO films deposited by double-target laser ablation,” J. Lumin. 162, 155–163 (2015). [CrossRef]
17. J. T. Leonard, D. A. Cohen, B. P. Yonkee, R. M. Farrell, T. Margalith, S. Lee, S. P. DenBaars, J. S. Speck, and S. Nakamura, “Nonpolar III-nitride vertical-cavity surface-emitting lasers incorporating an ion implanted aperture,” Appl. Phys. Lett. 107(1), 011102 (2015). [CrossRef]
18. C. A. Forman, S. Lee, E. C. Young, J. A. Kearns, D. A. Cohen, J. T. Leonard, T. Margalith, S. P. DenBaars, and S. Nakamura, “Continuous-wave operation of m-plane GaN-based vertical-cavity surface-emitting lasers with a tunnel junction intracavity contact,” Appl. Phys. Lett. 112(11), 111106 (2018). [CrossRef]
19. S. Lee, C. A. Forman, C. Lee, J. Kearns, E. C. Young, J. T. Leonard, D. A. Cohen, J. S. Speck, S. Nakamura, and S. P. DenBaars, “GaN-based vertical-cavity surface-emitting lasers with tunnel junction contacts grown by metal-organic chemical vapor deposition,” Appl. Phys. Express 11(6), 062703 (2018). [CrossRef]
20. J. T. Leonard, E. C. Young, B. P. Yonkee, D. A. Cohen, T. Margalith, S. P. DenBaars, J. S. Speck, and S. Nakamura, “Demonstration of a III-nitride vertical-cavity surface-emitting laser with a III-nitride tunnel junction intracavity contact,” Appl. Phys. Lett. 107(9), 091105 (2015). [CrossRef]
21. C. A. Forman, S. G. Lee, E. C. Young, J. A. Kearns, D. A. Cohen, J. T. Leonard, T. Margalith, S. P. DenBaars, and S. Nakamura, “Continuous-wave operation of nonpolar GaN-based vertical cavity surface-emitting lasers,” Proc. SPIE 10532, 105321C (2018). [CrossRef]
22. H. Xing, D. S. Green, H. Yu, T. Mates, P. Kozodoy, S. Keller, S. P. Denbaars, and U. K. Mishra, “Memory effect and redistribution of Mg into sequentially regrown GaN layer by metalorganic chemical vapor deposition,” Jpn. J. Appl. Phys. 42(Part 1, No. 1), 50–53 (2003). [CrossRef]
23. PICS3D (Photonic Integrated Circuit Simulator in 3D) by Crosslight Software, Inc., Burnaby, Canada, 2018.
24. R. Vaxenburg, E. Lifshitz, and A. Efros, “Suppression of Auger-stimulated efficiency droop in nitride-based light emitting diodes,” Appl. Phys. Lett. 102(3), 031120 (2013). [CrossRef]
25. Q. H. Pham, J. C. Chen, and H. B. Nguyen, “Three-Dimensional Numerical Study on the Efficiency Droop in InGaN/GaN Light-Emitting Diodes,” IEEE Photonics J. 11(1), 1–17 (2019). [CrossRef]
26. Z.-H. Zhang, S. Tiam Tan, Z. Kyaw, Y. Ji, W. Liu, Z. Ju, N. Hasanov, X. Wei Sun, and H. Volkan Demir, “InGaN/GaN light-emitting diode with a polarization tunnel junction,” Appl. Phys. Lett. 102(19), 193508 (2013). [CrossRef]
27. C. Y. Lai, T. M. Hsu, W. H. Chang, K. U. Tseng, C. M. Lee, C. C. Chuo, and J. I. Chyi, “Direct measurement of piezoelectric field in In0.23Ga0.77N/GaN multiple quantum wells by electrotransmission spectroscopy,” J. Appl. Phys. (Melville, NY, U. S.) 91(1), 531–533 (2002). [CrossRef]
28. D. Kasahara, D. Morita, T. Kosugi, K. Nakagawa, J. Kawamata, Y. Higuchi, H. Matsumura, and T. Mukai, “Demonstration of Blue and Green GaN-Based Vertical-Cavity Surface-Emitting Lasers by Current Injection at Room Temperature,” Appl. Phys. Express 4(7), 072103 (2011). [CrossRef]
29. D. H. Hsieh, A. J. Tzou, T. S. Kao, F. I. Lai, D. W. Lin, B. C. Lin, T. C. Lu, W. C. Lai, C. H. Chen, and H. C. Kuo, “Improved carrier injection in GaN-based VCSEL via AlGaN/GaN multiple quantum barrier electron blocking layer,” Opt. Express 23(21), 27145 (2015). [CrossRef]
30. T. Hamaguchi, H. Nakajima, M. Tanaka, M. Ito, M. Ohara, T. Jyoukawa, N. Kobayashi, T. Matou, K. Hayashi, H. Watanabe, R. Koda, and K. Yanashima, “Sub-milliampere-threshold continuous wave operation of GaN-based vertical-cavity surface-emitting laser with lateral optical confinement by curved mirror,” Appl. Phys. Express 12(4), 044004 (2019). [CrossRef]
31. M. Kuramoto, S. Kobayashi, T. Akagi, K. Tazawa, K. Tanaka, K. Nakata, and T. Saito, “Wattclass blue vertical-cavity surface-emitting laser arrays,” Appl. Phys. Express 12(9), 091004 (2019). [CrossRef]
32. K. Mehta, Y.-S. Liu, J. Wang, H. Jeong, T. Detchprohm, R. D. Dupuis, and P. D. Yorder, “Thermal Design Considerations for III-N Vertical-Cavity Surface-Emitting Lasers Using Electro-Opto-Thermal Numerical Simulations,” IEEE J. Quantum Electron. 55(5), 1–8 (2019). [CrossRef]
33. R. P. Sarzała, P. Śpiewak, and M. Wasiak, “Influence of Resonator Length on Performance of Nitride TJ VCSEL,” IEEE J. Quantum Electron. 55(6), 1–9 (2019). [CrossRef]
34. A. Sikora, H. Ftouni, J. Richard, C. Hébert, D. Eon, F. Omnès, and O. Bourgeois, “Highly sensitive thermal conductivity measurements of suspended membranes (SiN and diamond) using a 3ω-Völklein method,” Rev. Sci. Instrum. 83(5), 054902 (2012). [CrossRef]
