Peculiarities of the photoluminescence of metastable Ga(N,As,P)/GaP quantum well structures

The metastable quaternary Ga(N,As,P) has been studied experimentally as well as theoretically as promising material for integration of optoelectronics with Si-based microelectronics. The optical studies reveal unusual peculiarities in the photoluminescence response and its temperature dependence. The disorder-induced features have been analyzed by means of Monte Carlo simulations. A two-energy-scale approach is necessary to achieve a satisfactory agreement between experiment and simulation.

Figure 1

(a) Schematic representation of the temperature-dependent PL peak position (full line) and the band-gap energy as a function of temperature (dashed line). The deviation between the PL peak energy and the band gap determines the Stokes shift which is caused by the emission from localized excitonic states. $T_1$ gives the temperature of the maximal Stokes shift. (b) Schematic behavior of the PL linewidth with temperature. $T_2$ gives the temperature of the maximal PL linewidth. The inset depicts the low-temperature PL spectra with the logarithmic slope $\beta$ shown by a dotted line.

Figure 2

Dark field (top) and bright field (bottom) TEM image $\left[g=\left(002\right)\right]$ of a cross section through the Ga(N,As,P) MQW structure. Within the primary quantum well layers a height variation in the quantum well thickness of about 1 nm at a lateral scale of about 100–200 nm can be identified.

Figure 3

(Color online) Temperature-dependent photoluminescence of the $\text{Ga}\left({\text{N}}_{0.04}{\text{As}}_{0.91}{\text{P}}_{0.05}\right)/\text{GaP}$ MQW structure, varying from 10 K (top) to 290 K (bottom).

Figure 4

(Color online) Excitation density dependence of the peak emission energy. Upon reaching a critical excitation density, the excitonic emission starts to show saturation behavior, pushing the emission energy higher with increasing excitation density.

Figure 5

(Color online) Temperature dependence of the peak emission with varying excitation densities. As the carrier concentration increases, the observed S shape of the luminescence more and more smudges until it completely disappears.

Figure 6

(Color online) Temperature-dependent electromodulated reflectance of the $\text{Ga}\left({\text{N}}_{0.04}{\text{As}}_{0.91}{\text{P}}_{0.05}\right)/\text{GaP}$ MQW structure. Apart from the energetically lowest first electron-hole transition within the quantum well, further transitions of higher order can be identified.

Figure 7

(Color online) Upper part: temperature dependence of peak emission (see Fig. 3) and absorption (by fitting the first transition obtained from the CER measurements to the standard line-shape model, see Fig. 6). Lower part: temperature dependence of the linewidth of the emission.

Figure 8

Temperature dependencies of (a) the Stokes shift and (b) FWHM of Ga(N,As,P)/GaP QWs. (●) Experimental points taken from Fig. 7. (○) Theoretical simulation using Eq. (4). (◻) Simulation using Eq. (5) with $\gamma =2/3$. (△) Simulation using Eq. (6). (c) PL spectra at low temperatures. (Full line) Experimental curve taken from Fig. 3; (dashed line) simulation with Eq. (4); (dashed-dotted line) simulation Eq. (5) with $\gamma =2/3$; and (dotted line) simulation with Eq. (6). See text for details.

Figure 9

The function $\text{exp}\left[{\left(E/E_0\right)}^{\gamma }\right]$ for various values of $E_0$ and $\gamma$. See text for details.

Figure 10

(a) Schematic representation of the QW layer as a sequence of cells, (b) schematic representation of the long-range disorder potential (solid line) and corresponding steplike function (thin line). Inset: the short-range disorder potential within a single cell.

Figure 11

Integrated PL intensity as function of temperature of a Ga(N,As,P)/GaP QW. (●) Experimental points and (○) simulated results.

Figure 12

Temperature dependencies of (a) the Stokes shift and of (b) the FWHM of Ga(N,As,P)/GaP QWs. (●) Experimental points from Fig. 7 and (open symbols) simulation using the model with two scales of disorder and different values of the transition probability ${\xi }_0$ between the cells including the case without transitions $\left({\xi }_0=0\right)$. The lines are guide for the eyes. (c) PL spectra at various temperatures simulated for ${\xi }_0=0$.

Figure 13

Temperature dependencies of (a) the Stokes shift and of (b) the FWHM of Ga(N,As,P)/GaP QWs. (●) Experimental points from Fig. 7. (○) Simulation using the theoretical model with two scales of disorder and the temperature-dependent transition probability between the cells shown in the inset of (a). (c) Corresponding PL spectra simulated for various temperatures. All lines are guide for the eye.

Figure 14

PL spectra at various temperatures simulated for Ga(N,As,P)/GaP QW simulated by means of the theoretical model with two scales of disorder as well as temperature- and energy-dependent transition probabilities between the cells. (a) Abrupt energy limits for the transition probability with the corresponding ${\xi }_0\left(T\right)$ dependence shown by filled squares in the inset of (c); (b) continuous energy-dependent transition probabilities with the corresponding ${\xi }_0\left(T\right)$ dependence shown by open squares in the inset of (c). (c) Temperature dependencies of the Stokes shift; (d) FWHM; (●) experimental points from Fig. 7. (○) abrupt energy limit for the transition probability; (◻) continuous energy-dependent transition probability; lines are just eye guides.