Dynamics of Single InGaN Quantum DotsR.A. Taylor *, J.W. Robinson , J.H. Rice , A. Jarjour , J.D. Smith , R.A. Oliver , G.A.D. Briggs , M.J. Kappers ,C.J. Humphreys , Y. Arakawa

1 1 1 1 1 2 2 3 3 4

1

2

3

4

Clarendon Laboratory, Department of Physics, University of Oxford, Parks Road, Oxford Ox1 3PU, U.K.

Department of Materials, University of Oxford, Parks Road, Oxford, OX1 3PH, U.K.

Department of Materials, University of Cambridge, Pembroke Street, Cambridge CB2 3QZ, U.K.

Institute of Industrial Science, University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8505, Japan

Figure 5.(a) Time-resolved PL from a single InGaN QD

at 4.2 K (Tokyo sample), trace recorded at 2.863 eV

corresponding to the dominant dot in (c), together

with the wetting layer emission at 2.873 eV (curve

with faster decay). (b) Corrected QD decay trace

from the InGaN QD following subtraction of wetting

layer. (c) Time-integrated photoluminescence from

InGaN QDs at 4.2 K at an excitation power of 70 W.�

Figure 1. (a) and (b) - AFM scans of uncapped dot samples grown in

Cambridge. (c) Opitcal microscope image of Al mask on dot sample.

(d) Dimensions of typical dot from Cambridge sample. (e) and (f) -

Schematic of growth layers in both the Cambridge and Tokyo sample.

Figure 3. Experimental arrangement for time-resolved

and time-integrated single dot spectroscopy. The excitation

source is a frequency-tripled Ti:sapphire laser, giving 100fs

pulses at 266 nm. The microscope objective is a 36x

catadioptric unit. The time-correlated photon counting unit

has a time-resolution of 150 ps.

Plate 1. Experimental layout showing the two spectrographs

used for detection. In the foreground the piezo mounted

catadioptric objective can be seen. The microscope cryostat

is mounted on an xy stage below the objective.

Experiment

Figure 4. (a) Time-resolved PL from a single InGaN QDat 4.2 K (Cambridge sample). PL trace recorded at2.824 eV corresponding to the dominant dot in (c),together with the wetting layer emission at 2.817 eV(curve with faster decay). (b) Corrected QD decaytrace from the InGaN QD following subtraction ofwetting layer. (c) Time-integrated photoluminescencefrom InGaN QDs at 4.2 K at an excitation power of

275 W, indicating the dot used for the time-resolvedmeasurements.

�

xford

hysics

2.80 2.85 2.90 2.95 3.00

2.35

0.97

1.392.19

1.02

PL

Inte

nsity

(au)

Energy (eV)

0 10 20 30 40 501.6

1.8

2.0

2.2

2.4

2.6

2.8

2.80 2.81 2.82 2.83 2.84

(b)

Lifetim

e(n

s)

Temp (K)

(a)

6050

4030

20

16

12

8

4 K

PL

Inte

nsity

(au)

Energy (eV)

Time-resolved spectroscopy:time-correlated singlephoton counting

Time-integratedspectroscopy: micro-photoluminescence

Cube beamsplitter

Microscope objective

Cryostat

Monochromator

Ti:Al O

laser2 3

Frequencytripler

Monochromator

CCD

PMT

Quantum dot dimensions

20 nm

4 nm14 nm

1500nm

GaN Capping Layer

InGaN QD

GaN Buffer Layer

Al O Substrate2 3

Tokyo Sample

7 nm

2500nm

GaN Capping Layer

InGaN QD

GaN Buffer Layer

Al O Substrate2 3

Cambridge Sample

2 m hole�

(a) (b) (c)

(d) (e) (f)

1 m x AFM scan� 1 m� 0.4 m m� �x 0.4 AFM scan

2.70 2.75 2.80 2.85 2.90 2.95

PL

Inte

nsity

(au)

Energy(eV)

0 2 4 6 80 2 4 6 8 10

(b)

T1= 2.59 ns

PL

Inte

nsity

(au)

Time (ns)

(a)

PL

Inte

nsity

(au)

Time (ns)

(c)

0 1 2 3 4 5 0 1 2 3 4 5

(b)

PL

Inte

nsity

(au)

Time (ns)

(a)

T1= 1.07 ns

PL

Inte

nsity

(au)

Time (ns)

2.75 2.80 2.85 2.90 2.95 3.00

PL

Inte

nsity

(au)

Energy(eV)

(c)

Growth of InGaN epilayer Formation of In-richregions by spinodaldecomposition

In on surface

Decomposition of In-rich regions in N

atmosphere - In and N2

2

Formation of nanoscopicmetallic droplets

Re-nitridation of dropletsin initial stage of cappinglayer growth

Completion of cappinglayer growth; someinter-diffusion of Ga

Possible dot formation mechanism by modified droplet epitaxy (Oxford sample)

NH3 TMG

Figure 7. Micro-PL spectra from several single dots at 4.2 K.

The measured decay times in nanoseconds for some individual

dots are shown above each dot.

Figure 6. (a) Time integrated PL spectra for a single InGaN

QD from 4.2 K to 60 K (Cambridge sample). (b) Variation in

T with temperature for the single QD shown in (a).1

Figure 2. Diagram showing the suggested stages involved in the growth of the

Cambridge sample.

Y. Arakawa, IEEE J. Selected topics in quantum electronics 8 (2002) 823R.A. Oliver, G.A.D. Briggs, M.J. Kappers, C.J. Humphreys, J.H. Rice, J.D. Smith,and R.A. Taylor, Appl. Phys. Lett, Accepted for publication, July 2003F. Widmann, B. Daudin, G. Feuillet, Y. Samson, J. L. Rouvière, and N. Pelekanos,J. Appl. Phys. 83 (1998) 7618 K. Tachibana, T. Someya, and Y. Arakawa, Appl.Phys. Lett. 74 (1998) 383O. Moriwaki, T. Someya, K. Tachibana, S. Ishida, and Y. Arakawa, Appl. Phys.Lett. 76 (2000) 2361I.L. Krestnikov, N.N. Ledentsov, A. Hoffmann, D. Bimberg, A.V. Sakharov,W.V. Lundin, A.F. Tsatsul'nikov, A.S. Usikov, Z.I. Alferov, Y.G. Musikhin,D. Gerthsen. Phys. Rev. B. 66 (2002) 155310J. Simon, E. Martinez-Guerrero, C. Adelmann, G. Mula, B. Daudin, G. Feuillet,H. Mariette, and N.T. Pelekanos. Phys. Stat. Sol. B, 224 (2001) 13A. Morel, P. Lefebvre, T. Taliercio, T. Bretagnon, B. Gil, N. Grandjean, B. Damilano,and J. Massies, Physica E 17 (2003) 64K. Asaoka, Y. Ohno, S. Kishimoto, and T. Mizutani, Jap. J. Appl. Phys.Part 1 38 (1999) 546 T. Takagahara, Phys. Rev. Lett. 71 (1993) 3577

References

The photoluminescence decay of single InGaN QDs has been measured between 4.2 K and 60 K using a time-correlated photon

counting system in conjunction with a conventional micro-photoluminescence setup. Two different samples, grown in

Cambridge and in Japan, were studied. The recombination is characterized by a single exponential decay, in contrast to the

non-exponential decay from the wetting layer. We have shown that for ensemble measurements a non-exponential signal may

arise from the combination of the QD and wetting layer luminescence. Our results suggest that different growth methods

give rise to fairly similar luminescence properties. Increasing the temperature leads to a reduction in the PL decay time of a single

InGaN QD over the range measured.

This work was supported in part by the Foresight LINK Award , funded by DTI/EPSRC

(GR/R66029/01) and Hitachi Europe Ltd., and by Hewlett-Packard Laboratories. We would like to thank Shazia Yasin for her

assistance in masking the samples.

Nanoelectronics at the Quantum Edge