Nonresonant tunneling in single asymmetric pairs of vertically stacked $\mathrm{In}\mathrm{P}$ quantum dots

Single pairs of vertically stacked asymmetric pairs of InP quantum dots embedded in GaInP barriers have been investigated as a function of interdot spacer thickness. Time integrated and time-resolved photoluminescence measurements have been performed, with the former showing a change in the intensity ratio between the two dots and the latter an increasing difference in the photoluminescence decay time of the two dots when reducing the spacer thickness. Hence, we suggest transitions from vanishing tunnel coupling to electron tunneling and, finally, to electron and hole tunneling for decreasing barrier widths. The different times are estimated from the measurement data, and the changes are described by a rate equation model. The results clearly show the nonresonant character of the tunneling process as a result of the different ground state energies (approximately $40\mathrm{meV}$) of the unequally sized dots.

Figure 1

(Color online) Schematic of the sample structure (left) and the resulting band structure (right). Carriers have the possibility to tunnel for small enough barriers by emitting a phonon.

Figure 2

(Color online) Power dependent PL spectra of the AQDP for (a) a $20\mathrm{nm}$ spacer with no significant interdot tunneling, (b) a $10\mathrm{nm}$ spacer with significant interdot electron tunneling, and (c) a $2.5\mathrm{nm}$ spacer with significant interdot electron and hole tunneling.

Figure 3

(Color online) (a) Change of the PL intensity ratio $I_{\mathrm{SD}}∕I_{\mathrm{LD}}$ with excitation power density for a $20\mathrm{nm}$ spacer, a $10\mathrm{nm}$ spacer, and a $2.5\mathrm{nm}$ spacer. (b) The results of the rate equation model [Eq. (1)] for different tunneling times. These tunneling times were chosen to qualitatively show the influence of different tunneling times on the time integrated PL.

Figure 4

(Color online) Time-resolved PL measurements of the LD and the SD for (a) a $20\mathrm{nm}$ spacer, (b) a $10\mathrm{nm}$ spacer, and (c) a $2.5\mathrm{nm}$ spacer. (d) Results of the rate equation model [Eq. (1)] for different tunneling times.

Figure 5

(Color online) Estimated interdot transfer times (symbols) obtained from fittting the time-resoved PL data that were recorded for various barrier widths using a monoexponential function (solid line). Insets: Relation between barrier width and spacer thickness for both large barriers and small barriers.