Limits...
Luminescence studies on green emitting InGaN/GaN MQWs implanted with nitrogen.

Sousa MA, Esteves TC, Sedrine NB, Rodrigues J, Lourenço MB, Redondo-Cubero A, Alves E, O'Donnell KP, Bockowski M, Wetzel C, Correia MR, Lorenz K, Monteiro T - Sci Rep (2015)

Bottom Line: The as-grown and as-implanted samples were found to exhibit a single green emission band attributed to localized excitons in the QW, although the N implantation leads to a strong reduction of the PL intensity.The green band was found to be surprisingly stable on annealing up to 1400°C.This band is more intense for the implanted sample, suggesting that defects generated by N implantation, likely related to the diffusion/segregation of indium (In), have been optically activated by the thermal treatment.

View Article: PubMed Central - PubMed

Affiliation: Departamento de Física e I3N, Universidade de Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugal.

ABSTRACT
We studied the optical properties of metalorganic chemical vapour deposited (MOCVD) InGaN/GaN multiple quantum wells (MQW) subjected to nitrogen (N) implantation and post-growth annealing treatments. The optical characterization was carried out by means of temperature and excitation density-dependent steady state photoluminescence (PL) spectroscopy, supplemented by room temperature PL excitation (PLE) and PL lifetime (PLL) measurements. The as-grown and as-implanted samples were found to exhibit a single green emission band attributed to localized excitons in the QW, although the N implantation leads to a strong reduction of the PL intensity. The green band was found to be surprisingly stable on annealing up to 1400°C. A broad blue band dominates the low temperature PL after thermal annealing in both samples. This band is more intense for the implanted sample, suggesting that defects generated by N implantation, likely related to the diffusion/segregation of indium (In), have been optically activated by the thermal treatment.

No MeSH data available.


Related in: MedlinePlus

Integrated PL intensity dependence on the excitation intensity for the green band in the #as-grown sample (a), and green (b) and blue (c) in the post growth treated samples.Full lines correspond to the best fits of the experimental data according with a power law dependence (IαPm).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4389667&req=5

f4: Integrated PL intensity dependence on the excitation intensity for the green band in the #as-grown sample (a), and green (b) and blue (c) in the post growth treated samples.Full lines correspond to the best fits of the experimental data according with a power law dependence (IαPm).

Mentions: The appearance after HTHP annealing of a blue band, actually the dominant recombination in the #as-imp-HTHP sample, (Figure 3c)), deserves to be explored more deeply. Blue bands have been extensively studied in unintentionally and intentionally doped GaN20. In the case of non-intentionally doped GaN layers, such as those involved in our MQW samples, a 2.9 eV blue luminescence has been reported202122. This luminescence behaves like the one of a donor-acceptor pair (DAP) at low temperatures, transforming to a free-to-bound (e-A) recombination at temperatures above ~100 K21. Moreover the emission band exhibits a full width at half maximum (FWHM) of ~400 meV, decay times in the microsecond range, and a luminescence thermal quenching for temperatures above 200 K with an activation energy of 380 meV2122. As shown in Figure 3, the BB identified in the present MQW structures has a narrower FWHM, a thermal quenching described by two relatively small activation energies (Table 2) and shows a RT lifetime shorter than 1 ns, since no measurable signal could be collected by using the current experimental set-up. These evidences clearly indicate that the BB in the studied samples behaves very differently from that previously reported in undoped GaN layers202122 both as a function of temperature (Figure 3) and excitation density (Figure 4). Furthermore, the PLE monitored at the BB band have shown unequivocally the changes on the onset of the absorption, well below the GaN near band edge, which is much more consistent with the hypothesis to relate the observed BB emission to defects in the InGaN active layers or InGaN/GaN interface regions, specifically InN-poor regions as identified by TEM after HTHP annealing (to be published elsewhere). In order to identify a recombination model for the luminescence bands in our samples, further PL studies were realized as a function of excitation density, as presented in Figure 4. No power-dependent shift of either band maximum was observed, discounting any recombination model involving DAP transitions. As a function of the excitation density, the PL intensity can be well fitted to a power law over three decades of excitation density, IαPm with an exponent close to unity (Figures 4 (a), (b) and (c)). T. Schmidt et al. reported a PL power dependence well described by an exponent m between 1 < m < 2 for free and bound excitons23. A deviation of the power law was observed for higher decades, namely when saturation effects occur23. As shown in Figure 4 (a), at high excitation densities, the green band intensity for the #as-grown sample saturates as a consequence of the limiting radiative decay process. On the contrary, the same GB power dependence analysis for the #as-imp and #as-grown-HTHP samples do not exhibit such saturation. Furthermore, a similar linear behavior was found for the blue bands for both the #as-grown-HTHP and the #as-imp-HTHP samples, supporting a recombination model where the defects generated by post-growth treatments, such as those related with In redistribution24, are able to localize excitons.


Luminescence studies on green emitting InGaN/GaN MQWs implanted with nitrogen.

Sousa MA, Esteves TC, Sedrine NB, Rodrigues J, Lourenço MB, Redondo-Cubero A, Alves E, O'Donnell KP, Bockowski M, Wetzel C, Correia MR, Lorenz K, Monteiro T - Sci Rep (2015)

Integrated PL intensity dependence on the excitation intensity for the green band in the #as-grown sample (a), and green (b) and blue (c) in the post growth treated samples.Full lines correspond to the best fits of the experimental data according with a power law dependence (IαPm).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4389667&req=5

f4: Integrated PL intensity dependence on the excitation intensity for the green band in the #as-grown sample (a), and green (b) and blue (c) in the post growth treated samples.Full lines correspond to the best fits of the experimental data according with a power law dependence (IαPm).
Mentions: The appearance after HTHP annealing of a blue band, actually the dominant recombination in the #as-imp-HTHP sample, (Figure 3c)), deserves to be explored more deeply. Blue bands have been extensively studied in unintentionally and intentionally doped GaN20. In the case of non-intentionally doped GaN layers, such as those involved in our MQW samples, a 2.9 eV blue luminescence has been reported202122. This luminescence behaves like the one of a donor-acceptor pair (DAP) at low temperatures, transforming to a free-to-bound (e-A) recombination at temperatures above ~100 K21. Moreover the emission band exhibits a full width at half maximum (FWHM) of ~400 meV, decay times in the microsecond range, and a luminescence thermal quenching for temperatures above 200 K with an activation energy of 380 meV2122. As shown in Figure 3, the BB identified in the present MQW structures has a narrower FWHM, a thermal quenching described by two relatively small activation energies (Table 2) and shows a RT lifetime shorter than 1 ns, since no measurable signal could be collected by using the current experimental set-up. These evidences clearly indicate that the BB in the studied samples behaves very differently from that previously reported in undoped GaN layers202122 both as a function of temperature (Figure 3) and excitation density (Figure 4). Furthermore, the PLE monitored at the BB band have shown unequivocally the changes on the onset of the absorption, well below the GaN near band edge, which is much more consistent with the hypothesis to relate the observed BB emission to defects in the InGaN active layers or InGaN/GaN interface regions, specifically InN-poor regions as identified by TEM after HTHP annealing (to be published elsewhere). In order to identify a recombination model for the luminescence bands in our samples, further PL studies were realized as a function of excitation density, as presented in Figure 4. No power-dependent shift of either band maximum was observed, discounting any recombination model involving DAP transitions. As a function of the excitation density, the PL intensity can be well fitted to a power law over three decades of excitation density, IαPm with an exponent close to unity (Figures 4 (a), (b) and (c)). T. Schmidt et al. reported a PL power dependence well described by an exponent m between 1 < m < 2 for free and bound excitons23. A deviation of the power law was observed for higher decades, namely when saturation effects occur23. As shown in Figure 4 (a), at high excitation densities, the green band intensity for the #as-grown sample saturates as a consequence of the limiting radiative decay process. On the contrary, the same GB power dependence analysis for the #as-imp and #as-grown-HTHP samples do not exhibit such saturation. Furthermore, a similar linear behavior was found for the blue bands for both the #as-grown-HTHP and the #as-imp-HTHP samples, supporting a recombination model where the defects generated by post-growth treatments, such as those related with In redistribution24, are able to localize excitons.

Bottom Line: The as-grown and as-implanted samples were found to exhibit a single green emission band attributed to localized excitons in the QW, although the N implantation leads to a strong reduction of the PL intensity.The green band was found to be surprisingly stable on annealing up to 1400°C.This band is more intense for the implanted sample, suggesting that defects generated by N implantation, likely related to the diffusion/segregation of indium (In), have been optically activated by the thermal treatment.

View Article: PubMed Central - PubMed

Affiliation: Departamento de Física e I3N, Universidade de Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugal.

ABSTRACT
We studied the optical properties of metalorganic chemical vapour deposited (MOCVD) InGaN/GaN multiple quantum wells (MQW) subjected to nitrogen (N) implantation and post-growth annealing treatments. The optical characterization was carried out by means of temperature and excitation density-dependent steady state photoluminescence (PL) spectroscopy, supplemented by room temperature PL excitation (PLE) and PL lifetime (PLL) measurements. The as-grown and as-implanted samples were found to exhibit a single green emission band attributed to localized excitons in the QW, although the N implantation leads to a strong reduction of the PL intensity. The green band was found to be surprisingly stable on annealing up to 1400°C. A broad blue band dominates the low temperature PL after thermal annealing in both samples. This band is more intense for the implanted sample, suggesting that defects generated by N implantation, likely related to the diffusion/segregation of indium (In), have been optically activated by the thermal treatment.

No MeSH data available.


Related in: MedlinePlus