Biomedical Engineering Reference
In-Depth Information
The FWHM reduction at intermediate temperatures is well explained by thermal
escape and carrier redistribution. At low temperatures, excitons are created, con-
fined, and recombined radiatively in individual QDs. The FWHM reflects QD size
distribution. For uniform ensembles, the FWHM can be lower than 30 meV [ 42 ]
while for nonuniform ones the FWHM can be greater than 100 meV [ 43 ]. As sample
temperature increases, carriers are thermally activated out of individual QDs, into
the WL and/or the GaAs barriers where they can subsequently be re-trapped by
nearby QD. Since confined electron levels in small dots are higher than in large
dots, or the thermal energy required by the electrons to escape into the adjacent
WL is smaller for small dots than for large dots, it is more probable for carriers
to be distributed from small to large dots. As temperature increases, the dominant
emitting structure thus shifts from small to large dots, resulting in a peak energy
shift much faster than the rate due to bulk bandgap reduction with temperature. This
explains the dotted lines in Figs. 3.6 band 3.7 b.
The carrier redistribution occurs at a slightly lower temperature in the 1.8/25/1.2
QDMs (
75 K) than in the 1.8/25/1.5 QDMs (100 K). This is possibly due to the
smaller sQDs in the former because the regrown InAs thickness is 1.2 ML, lower
than 1.5 ML in the latter. The smaller sQDs emit at 1.240 eV while the bigger sQDs
emit at 1.195 eV. The smaller sQDs thus require a slightly lower thermal energy to
escape to the 1.4-eV WL previously identified in Fig. 3.5 a.
The co-existence of slow, Varshni type, and fast sigmoidal changes in the same
sample is unusual. The slow red-shift that follows Varshni's equation is usually
observed in QD ensembles which are carefully grown to achieve low FWHM,
especially if they are later to be fabricated into lasers [ 42 ]. On the other hand, the fast
red-shift is usually observed in highly inhomogeneous QD ensembles, particularly
if the FWHM is greater than 80 meV [ 39 , 40 ] and if they are later to be fabricated
into superluminescent diodes [ 44 ]. The two types of QD ensembles are achieved
with conflicting growth parameters and not usually observed on the same sample
even when clear bimodal size distribution exists. Kissel et al., for example, reported
the growth of intermediate sized QD ensembles with clear bimodal size distribution
but both exhibit essentially the same sigmoidal behavior [ 45 ].
Carrier redistribution from small to large sQDs is effective up to a certain
temperature beyond which thermal broadening will increase the FWHM and carrier
loss to NRR channels will quench the overall PL signal. The more complicated
growth procedure of lateral QDMs as compared to standard SK QDs makes it likely
that additional NRR channels/centers maybe present and limit the usefulness of
QDMs.
Using Arrhenius plots and simple, single activation energy fittings we prove that
these concerns are unwarranted as the main loss mechanism is caused by the WL, as
is the case for typical SK QDs. Figure 3.8 a, b shows the IIs vs. inverse temperature
plots for the 1.8/25/1.2 and 1.8/25/1.5 QDMs, respectively. The dashed lines are
fits to the equation I
where I stands for the integrated
intensity, I 0 is the low-temperature integrated intensity, A is the pre-exponential
factor, E A is the activation energy, k B is the Boltzmann's constant, and T is the
temperature. The best-fit values for E A in the 1.8/25/1.2 QDMs are 250 meV for
=
I 0 / [
1
+
A exp
(
E A /
k B T
)]
Search WWH ::




Custom Search