The comparative studies of a conventional AlGaN LED structure and single Si-doped n-Al0.80Ga0.20N TJ based LED structure at emission wavelength of 254 nm were investigated by using simulation models, shown in Figure. 1a-b. The TJ layer`s thickness is optimized by calculating the performance of TJ-based LED for various thicknesses. The optical emitted power is calculated for both conventional and proposed TJ LEDs at 200 A/cm2, shown in Figure. 2a. The conventional LED structure has an emitted power of 39.7 W/cm2 which is improved by 34% to 53 W/cm2 by using a 5 nm-thin single TJ layer. The IQE of TJ-based LED is improved up to 62% by using 5 nm-thin TJ, when compared to the conventional LED, shown in Figure. 2b. The use of 5nm-thin n-Al0.8Ga0.2N TJ offers higher IQE, low efficiency droop and high emission power in DUV LED, when compared to the conventional DUV LED or DUV LED with thicker TJ, shown in Figure. 2a-b.
To understand the underlying physics behind the improved performance of TJ-based LED we calculate the band alignment (conduction band, valance band) for conventional LED and 5nm-thin TJ-based LED, shown in Figure. 3a-b [39]. From Figure. 3a we can see that the reference device has conduction band barrier height Φe of 862 meV and valance band barrier height Φh of 360 meV, while the hole gain energy of 169 meV by passing from HSL to last-QB (LQB). In the case of TJ-based LED the Φe is 889 meV, Φh is 361 meV and hole gain energy ΔΦ is 193 meV. These calculated energy band values conclude that the TJ layer does not much affect the values of Φe and Φh but increases the hole injection toward the MQW, as a result the hole concentration inside the active region of TJ-based LED was highly enhanced. The increment in hole energy is not much obvious for the reported improvement but there is a quantum tunnelling probability of hole which is the main contributor of hole improvement in the active region. Hence it shows that TJ have not higher impact on band alignment but the improved performance is mainly contributed by the quantum tunnelling phenomena using 5nm-thin TJ structure. It can be explained by employing the physical relation given in Eq. (1), when the width of TJ reduces then the tunnelling probability may increase, as shown in Figure. 3a-b. In the case of 5 nm-thin TJ the tunnelling width reduces from 71 nm to 24 nm. The narrow is the width of the TJ the higher is the tunnelling probability; which can be represented in the following Eq. (1) [40]:
$${P}_{b}\cong exp\left({\int }_{0}^{{w}_{d}}\sqrt{\frac{2{m}_{t}^{*}{E}_{g}x}{{h}^{2}{w}_{d}}}dx\right)$$
1
Where \({P}_{b}\) is the tunnelling probability, \({w}_{d}\) represent the width of tunnelling layer, \({E}_{g}\) is the energy bandgap and \({m}_{t}^{*}\) is the carrier effective mass. The 5 nm-thin TJ offer higher \({P}_{b}\) for hole flow, which increases the hole concentration in the active region. As the hole concentration increases within MQW, the parasitic recombination rate will decrease due to the depletion of holes outside the MQW [41]. Conversely, the reduced parasitic recombination rate degrades the electron leakage from MQW as of the reduced availability of holes. In both cases, the hole injection efficiency effect electron and hole concentration within the MQW. The band energy is calculated for various thicknesses of the TJ layer and listed in Table. I. Herein, it can be summarized that as the thickness of the AlGaN TJ layer decreases, the hole injection toward the active region increases by obtaining more energy and higher tunnelling probability. Consequently, it can be deduced that high emitted power and IQE are attributed to the 5nm-thin single AlGaN -based TJ in DUV LED.
Table I: Conduction band barrier height (Φe,) valance band barrier height (Φh) and hole gain energy (ΔΦ) for conventional LED, TJ-based LED with 5, 20 and 50 nm thickness.
Energy
|
Reference LED
|
50 nm TJ layer LED
|
20 nm TJ layer LED
|
5 nm TJ layer LED
|
Φe (meV)
|
862
|
867
|
872
|
889
|
Φh (meV)
|
360
|
361
|
361
|
361
|
ΔΦ (meV)
|
169
|
173
|
180
|
193
|
To further clarify the effectiveness of single AlGaN-TJ based DUV LED, we examine the electron and hole concentration in the active region and compared with the conventional LED, shown in Figure. 4a-b. The electron concentration is slightly decreased in the last QW of the TJ-based LED because of n-doped AlGaN which helps to further block the electron toward the EBL and reduce the electron overflow from the active region as can be seen in the hetero-interface of LQB and EBL in Figure. 4a. The hole concentration scenario is different for the studied LED structures because of variation in hole energy and higher tunnelling phenomena as speculated in Figure. 4b. The hole concentration is almost double in each QW of TJ-based LED, when compared to the conventional LED device, which reveals the superiority of the TJ LED structure. This can further be elaborated in the inset plot of Figure. 4b where we compare the hole concentration of both LEDs in the region of EBL, HSL, and TJ layer, respectively. It can be seen in the inset of Figure. 4b that the hole accumulation is increased at the interface of EBL/HSL because of higher tunnelling probability of hole by TJ layer. This will improve the overall tunnelling probability of holes toward the MQW and hence a higher hole concentration will be available in the active region for better radiative recombination phenomena [42]. However, in the case of conventional LED, the hole accumulation at the interface of EBL and p-AlGaN HSL is lower and it generates the hole depletion region, which adversely affects the hole concentration in the active region. Consequently, a low hole concentration will be available in the active region for relatively low radiative recombination phenomena and the performances of the conventional LED is deteriorated.
The higher hole concentration will strongly affect the radiative recombination rate in the MQWs of TJ-based LED, which can be seen in Figure. 5a. Although the studied LEDs structure has the same concentration of electrons in the MQW except slight variation in the last QW, however due to the effective hole tunnelling phenomena, the TJ LED has a significantly higher radiative recombination rate as compared to conventional LED, shown in Fig. 5(a). In addition, the spontaneous recombination rate for TJ-based LED is higher as compared to reference LED, shown in Figure. 5b, which clearly shows the effectiveness of the 5nm-thin Si-doped n-AlGaN TJ layer. Moreover, it can be depicted from the spontaneous recombination rate that the AlGaN TJ layer only improved the recombination rate by providing a higher hole concentration without affecting the main emission peak of LED. Based on the simulation results, it can be concluded that thin TJ layer improves the hole injection efficiency toward the MQW which leads to a higher radiative recombination rate and spontaneous rate, when compared to the conventional LED. On the other hand, if we use a thick n-AlGaN TJ layer the hole injection toward the MQW is not much obvious as one can find via thin TJ layer and as a result the radiative recombination rate and spontaneous rate are deteriorated. This is quite promising single TJ-based DUV LED structure for the epi-growers of UV emitters using LP-MOVPE and MBE [43].
Research and development of efficient AlGaN-based DUV LED at 254 nm emission wavelength is the safe and effective replacement of toxic Hg DUV lamp. Carrier confinement and transport issues in the MQWs are quite critical [44, 45]. To resolve these issues of carrier confinement and transport, we provided a short roadmap for experimental efforts to realize internal quantum efficiency (IQE) beyond 70% in the AlGaN-TJ base DUV LEDs. In this regard, the proposed n-AlGaN single TJ-based LED will open a better way for the realization of highly efficient AlGaN-based DUV LEDs for fast disinfection of surfaces, air, food and water. It is believed that this finding will help in the improvement of EQE and WPE of the experimental AlGaN-based DUV LEDs.