Henini M. Handbook of Self Assembled Semiconductor Nanostructures for Novel devices in Photonics and Electronics

Подождите немного. Документ загружается.

Advanced Growth Techniques of InAs-system Quantum Dots for

Integrated Nanophotonic Circuits

Kiyoshi Asakawa ,

Nobuhiko Ozaki

Shunsuke Ohkouchi,

Yoshimasa Sugimoto,

and Naoki Ikeda

Center for Tsukuba Advanced Research Alliance (TARA), University of Tsukuba,

1-1-1 Tennoudai, Tsukuba, Ibaraki 305-8577, Japan;

Nano Electronics Research Laboratories, NEC Corporation, 34 Miyukigaoka,

Tsukuba, Ibaraki 305-8501, Japan;

National Institute for Materials Science (NIMS), 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan

17.1 Introduction

For the last two decades, semiconductor quantum dots (QDs) have been intensively investigated

from the viewpoints of both band engineering and growth technology. As a result, they have

provided a great deal of attractive electronic/optoelectronic devices thanks to their high density-

of-states speciﬁ c to the low-dimensional structures. Some of the results involve QD-based laser

diodes [1] and optical ampliﬁ ers [2] in the commercial base, while others have been exploited

intensively for advanced electronic/optoelectronic devices.

As a topic of growth technology, site-controlled quantum dots (SCQDs) by means of pat-

terned substrates or nano-probe-assisted technologies have attracted much attention for creation

of functional QD structures such as single QD and arrayed QDs, as well as achievement of QDs

with high uniformity and high density. At the ﬁ rst stage for creating such an SCQD, we proposed

an idea of an in situ STM (scanning tunnelling microscope)-probe-assisted MBE technique and

demonstrated two-dimensionally arrayed InAs SCQDs with 50  100 nm pitches [3] . At the sec-

ond stage, we have developed an in situ AFM (atomic force microscope) probe with a specially

designed cantilever capable of a practical throughput (1  10 msec/dot, for example). By using

the new probe, we have reproducibly fabricated uniform In (indium) nanodots in the selected

area ﬁ rst. Since the AFM chamber is connected in vacuum to an MBE chamber, In dots could be

directly converted to InAs QDs by subsequent arsenic-ﬂ ux irradiation [4] . For implementation of

a stacked array of uniform and high-density QDs, combination of the SCQD and subsequent S–K

mode QD growth was successfully demonstrated.

As a new topic of application, on the other hand, we proposed a unique research scheme by com-

bining QDs with another nanostructure, i.e. a photonic crystal (PC), to provide key photonic devices

for future advanced telecommunication systems, as shown in Fig. 17.1 [5] . In the left-hand wing is

an ultra-fast digital photonic network of the future, where an ultra-small and ultra-fast symmetri-

cal Mach–Zehnder (SMZ)-type [6] all-optical switch (PC-SMZ) has been developed so far in the ﬁ rst

phase by using GaAs-based two-dimensional PC slab waveguides embedded with InAs-based QDs

( Fig. 17.1 ) [7] . In the second phase, the PC-SMZ is now evolving into a new functional key device, i.e.

an ultra-fast all-optical ﬂ ip-ﬂ op (PC-FF) device that is essential for the digital photonic network [5] .

CHAPTER 17

CH017-I046325.indd 529CH017-I046325.indd 529 6/25/2008 3:17:18 PM6/25/2008 3:17:18 PM

530 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics

As long as QD technology is concerned all through these two phases, a selective-area-growth

(SAG) of QDs has been shown to be an important subject for exhibiting an optical non-linear effect

in the selective optical non-linear arms particularly for the PC-SMZ and PC-FF. For this purpose,

a metal-mask (MM)/molecular beam epitaxy (MBE) method of InAs QDs has been developed [8] .

In the right-hand wing in Fig. 17.1 , on the other hand, another category of a future quantum

information system is described, where a single photon qubit composed of a single QD embed-

ded in a PC-based high- Q cavity is situated as another important product of the PC/QD-combined

nano-photonic structure [9–11] . The AFM-probe-assisted SCQD growth method mentioned above,

renamed recently as a nano-jet probe (NJP) method, has been developed for this purpose because of

its potential to position a single QD at the centre of the point defect PC high- Q cavity [12–14] .

Advanced telecommunication

Quantum information

Digital photonic network

Phase-2: All-optical flip-flop

Large ONL/phase shift

Wide/flat band

Phase-1: All-optical switch

Single photon qubit

Quantum entanglement

Site-controlled single dot

High-Q cavity

PC/QD

Nano-photonics

technologies

Matrix switch

PC-SMZ

~500 μm

Figure 17.1 Schematic research scenario of PC/QD combined nano-photonic structures for advance telecommuni-

cation systems.

In this chapter, two types of unique QD growth techniques have been reviewed. The MM

method for InAs QDs in the PC-based all-optical switching devices is shown in section 17.2 and

the advanced techniques for the growth of InAs SCQDs by using the NJP method is described in

section 17.3.

17.2 Selective-area-growth of InAs quantum dots using the metal-mask/MBE method

17.2.1 Introduction

A selective-area-growth (SAG) technique for self-assembled QDs is useful for various applications

of the QDs. In the case of our proposed all-optical devices, PC-SMZ [7] and PC-FF [5] , the SAG

technique is intensively required for their effective operations. For instance, although the embed-

ded QDs in the PC waveguides (WGs) of PC-SMZ act as a phase shifter due to their optical non-

linearity, the QDs simultaneously absorb the propagation light and result in a low transmittance

in the PC-WG. Therefore, the QD growth in limited regions in the WGs is desirable rather than an

entire growth on a substrate. Furthermore, the PC-FF, which composes two PC-SMZs, requires

SAG of two kinds of QDs having different absorption wavelengths for embedding the QDs in each

optical non-linearity arm in the PC-SMZs.

In this section, we describe a metal mask (MM)/MBE method [8] for SAG of self-assembled

InAs QDs. First, we introduce the MM/MBE method and characterizations of the SA-grown QDs

by atomic-force-microscopy (AFM) and photoluminescence (PL) measurements, including two-

dimensional (2D) PL intensity mapping. Successful SAG of InAs QDs was clearly demonstrated,

while the QD density and homogeneity were almost equal to those of the QDs grown without

the MM. Second, we describe technologies for controlling the PL wavelength of InAs QDs by

CH017-I046325.indd 530CH017-I046325.indd 530 6/25/2008 3:17:18 PM6/25/2008 3:17:18 PM

Advanced Growth Techniques of InAs-system Quantum Dots for Integrated Nanophotonic Circuits 531

using a strain-reducing-layer (SRL) [15] inserted on the QDs. Sequential variation of the PL peak

wavelength as a function of SRL thickness was attained by the SRL insertion method. Finally,

we report on fabrications of a PC-WG with SA-grown QDs. For avoiding reﬂ ection due to a step

at a boundary of the selective growth area, we reduce the step height with a developed growth

sequence. Clear transmission spectra and pulse intensities/phase modulations in the PC-WG due

to the optical non-linearity of the QD are conﬁ rmed.

17.2.2 Experimental apparatus and procedures for MM/MBE method

Figure 17.2a shows an image of the prepared MM attached on a holder. The tantalum (Ta) MM is

employed to cover undesired growth areas during conventional MBE growth. The MM holder can

be mounted on and off in an ultra-high vacuum (UHV) chamber without exposing the sample to

the atmosphere.

Large window for

RHEED observation

Selective growth area

(4 mm ⴛ 500 ␮m)

PC-SMZ

(b)

Metal mask E-beam exit

E-beam

entrance

Metal mask holder

(a)

E-beam

GaAs substrate

Sample holder

Molecular beams

Figure 17.2 (a) Photograph of a metal-mask (MM) holder. (b) Schematic of patterns on the MM designed for

PC-SMZ fabrications. (c) Cross-sectional illustration of the MM and e-beam incident for RHEED observations

during the MBE growth.

CH017-I046325.indd 531CH017-I046325.indd 531 6/25/2008 3:17:19 PM6/25/2008 3:17:19 PM

532 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics

As illustrated in Fig. 17.2b , the MM has a large window (6 mm  6 mm) at the centre and sev-

eral small windows of 4 mm  0.5–1.0 mm around it. The small windows are prepared for SAG

of QDs, suited for the PC-SMZ. On the other hand, the large window is used for RHEED (reﬂ ection

high-energy electron-diffraction) observations and temperature measurements of the sample

surface by using a pyrometer during the MBE growth. An electron beam for RHEED observations

goes through the tunnels formed on the frame of the MM holder and is diffracted at the surface

where molecular beams are irradiated through the large open window during MBE growth, as

shown in Fig. 17.2c . These in situ observations enable us to optimize growth conditions for QDs.

We have performed the SAG of the QD using two methods, i.e. “ mask ” and “ cap/anneal ” meth-

ods [16] , as schematically shown in Fig. 17.3 . In the mask method, QDs are grown when the

MM covers the substrate, namely the SAG of QDs is directly operated with the MM. In the cap/

anneal method, on the other hand, the MM is used for selective growth of capping layers of QDs

grown on the entire substrate. After the entire QDs ’ growth, a selective area GaAs layer is grown

by using the MM for capping selective area QDs, followed by an annealing to evaporate the rest of

the QDs.

Molecular beams

Annealed

Mask

Cap/anneal

Capped

Figure 17.3 Two SAG sequences: “ mask ” and “ cap/anneal ” methods.

In both methods, sequences before QD growth are identical. At ﬁ rst, a multilayer structure of

a 50 nm thick GaAs layer/30 nm thick Al

0.3

0.7

As layer/GaAs buffer layer was grown on the

entire GaAs substrate without the MM at approximately 540˚C. In the mask method, the MM

holder was then mounted after the sample was cooled down below 200˚C. Following this, the

sample was gradually heated up to the InAs QD growth temperature, typically 470–490˚C, and

2.6 monolayers (ML) of InAs were deposited at a rate of 0.2 ML/s for the QD growth. During this

process, we conﬁ rmed that the RHEED patterns varied from streaky to spotty patterns, as shown

in the photographs in Fig. 17.4a and b, respectively. A chevron pattern originated from the QD

facet, as shown in Fig. 17.4b , indicating 3D QD growth.

After the sample was slightly cooled down by approximately 40˚C to avoid the degradation of

the QD, a 3 nm thick In

0.2

0.8

As layer/3 nm thick GaAs cap layer was sequentially grown. In

the cap/anneal method, on the other hand, the InAs QDs were grown entirely on the GaAs layer

with a 2.6 monolayer (ML) InAs deposition at a rate of 0.2 ML/s without the MM. Then, the MM

was mounted on the sample after cooling down below 200˚C and the sample temperature was

CH017-I046325.indd 532CH017-I046325.indd 532 6/25/2008 3:17:19 PM6/25/2008 3:17:19 PM

Advanced Growth Techniques of InAs-system Quantum Dots for Integrated Nanophotonic Circuits 533

gradually raised up to typically 430–450˚C. Selective capping layers of 3 nm thick In

0.2

0.8

and 20 nm thick GaAs were grown at the temperature via the MM. After removing the MM, the

temperature of the sample was gradually raised up to the QD growth temperature, for instance

480˚C, and was kept for 3 min to evaporate the QDs on regions uncovered by the capping layer.

SAG of QDs was completed in this way. After SAG of QDs using both methods, a multilayer struc-

ture of 50 nm thick GaAs capping layer/30 nm thick Al

0.3

0.7

As layer/50 nm thick GaAs layer

was grown on the entire substrate. The Al

0.3

0.7

As layers were inserted below and above the QD

layer as an energy barrier against free carriers in the PL measurement of the QD. As mentioned

later, the capping layer of 3 nm thick In

0.2

0.8

As on the QD is employed for reducing the strain

in grown QDs and a red shift of the PL from QDs to around 1300 nm in wavelength.

17.2.3 Structural and optical properties of SA-grown Qds

We characterized the SA-grown QDs by AFM observations and PL measurements. Afterwards,

we showed results from samples grown by the mask method on behalf of the two SAG methods.

Figure 17.5 shows typical AFM images of regions masked and unmasked by the MM. The

image of an unmasked region shows formation of InAs QDs whose density is around 4  1 0

 2

, whereas no QD was formed on the masked regions of the surface. Such high density is

comparable to that of QDs conventionally grown without the MM. The mean lateral size and

height of the QD was approximately 40 nm and 5 nm, respectively.

Figure 17.4 RHEED patterns obtained (a) before and (b) after growth of QDs with the MM.

(a)

(b)

Masked region

(a)

Unmasked region

100nm

(b)

Figure 17.5 AFM images of masked and unmasked regions with the MM. SA-grown QDs with a density of

4  10

2

were observed on the unmasked regions.

CH017-I046325.indd 533CH017-I046325.indd 533 6/25/2008 3:17:19 PM6/25/2008 3:17:19 PM

534 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics

For investigating the optical quality of SA-grown QDs, PL measurements were obtained at

room temperature (RT). An He–Ne laser ( λ  633 nm), focused to a diameter of several microns,

was used for QD excitation with a power of 80 μ W. A stepping-motor attached to the sample

stage was utilized for positioning the excitation area and 2D mapping of the PL intensity.

PL spectra and 2D intensity mapping from the unmasked and masked regions are shown in

Fig. 17.6 . The spectra shown with black and grey lines are measured from the unmasked regions

which are indicated in the PL intensity mapping by arrows. High contrast in the mapping shows

high PL intensity attributed to the SA-grown QDs. The PL peak and the FWHM are 1295 nm

(0.957 eV) and approximately 32 meV, respectively, which are almost the same as those of the

QDs grown without the MM. This small FWHM value obtained even at RT indicates the high

homogeneity of the SA-grown QDs comparable to that of QDs grown without the MM.

1100 1200 1300 1400

1.1 1.05 1 0.95 0.9

Photon energy (eV)

PL intensity (arb. unit)

Wavelen

th (nm)

021 3 4 (mm)

2.5E3

5.0E3

1.0E4

3.3mm

250om

7.5E3

3.3 mm

250μm

Figure 17.6 PL spectra and intensity mapping from SA-grown sample at room temperature.

0 4 8 12 16(mm)

2.5E3

5.0E3

7.5E3

1.0E4

Figure 17.7 PL intensity mapping image from the entire region under the MM. Successful SAG corresponding to

the MM pattern was conﬁ rmed.

Moreover, these PL spectra measured at two different regions show almost identical intensity and

linewidth, suggesting that such high density and homogeneous QDs were grown on the selective

areas of the substrate independent of the position. On the other hand, no peak was found in the

PL spectrum from the masked region, as shown by a dashed line. Figure 17.7 shows a PL intensity

mapping of the entire masked and unmasked areas. Through these PL intensity mapping images,

the successful SAG of the entire region underneath the MM was clearly exhibited.

CH017-I046325.indd 534CH017-I046325.indd 534 6/25/2008 3:17:21 PM6/25/2008 3:17:21 PM

Advanced Growth Techniques of InAs-system Quantum Dots for Integrated Nanophotonic Circuits 535

Each QD-grown area decreased to approximately 0.25 m m  3.3 mm SAG area obtained from

a window size of 0.5 mm  4 mm. This shrinkage of the SAG area is considered to be due to ther-

mal evaporation of QDs near the edge of the windows by the heated MM.

17.2.4 PL wavelength control of InAs QDs with a strain-reducing-layer

As previously mentioned, successful SAG of QDs was attained by the MM/MBE method. However,

when applying the MM/MBE method to realize the PC-FF [5] , controlling absorption wavelength

of SA-grown QDs is also required. For that purpose, we inserted a strain-reducing-layer (SRL)

between QDs and spacers. This method is operated by covering grown QDs with an In

0.2

0.8

layer to reduce strain in the QDs, and it results in a red shift of the PL emission peak of QDs [15] .

We have investigated the effectiveness of the SRL for controlling the absorption wavelength of

QDs. Figure 17.8a summarizes the PL spectra from strain-reduced QDs with different SRL thick-

nesses ranging from 0 to 6 nm. The PL spectra exhibit the PL peak wavelength that shifted from

1240 nm to 1320 nm and with the FWHM that was almost maintained at approximately 30 meV,

as shown in Fig. 17.8b .

1.05 1 0.95

1200 1250 1300 1350

PL intensity (arb. units)

Wavelength (nm)

SRL thickness: 0nm

2nm 3nm 4nm

6nm

Photon energy (eV)

0123456

1220

1240

1260

1280

1300

1320

1340

SRL thickness (nm)

PL peak wavelength (nm)

FWHM (meV)

(a)

(b)

Figure 17.8 Effectiveness of a strain-reducing-layer (SRL) for tuning the PL peak wavelength. (a) Summary of

PL spectra from QDs as a function of SRL thickness. (b) PL peak wavelength and FWHM variations against the SRL

thickness.

This result clearly indicates that the SRL can control the PL wavelength of QDs without degra-

dation of the QD optical quality. Thus, combination of the MM method and the insertion of the

SRL has a potential to develop the PC-FF, which requires SAG of QDs with different absorption

wavelengths at different areas.

17.2.5 Selective area-grown QD embedded in PC-WG

Here, we report on an application of the MM/MBE method to fabricate the SA grown QDs embed-

ded in PC-WGs. As previously mentioned, we developed the SAG technique for our proposed

PC-based devices. Considering an application of the SAG to PC-WGs, the step height of the SAG

area should be suppressed as much as possible since the step causes undesired reﬂ ections of the

propagating light and narrowing the bandwidth of the transmitted light. Thus, we attempted to

minimize a step height of the SAG with optimizing the growth sequences. As a result, the SAG

method enables us to minimize the step height within 6 nm at the SAG region against the core

thickness 250 nm, as schematically shown in Fig. 17.9 . Also, we succeeded in characterizations

of the PC-WG with QDs partially embedded, i.e. the transmittance of the continuous wave (CW)

light and the intensity and phase modulations of light pulses due to the optical non-linear phe-

nomenon induced by the embedded QDs.

CH017-I046325.indd 535CH017-I046325.indd 535 6/25/2008 3:17:23 PM6/25/2008 3:17:23 PM

536 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics

We prepared a GaAs-based straight PC-WG composed of an air bridge-type 2DPC slab and sin-

gle InAs QDs layer embedded in a selective area of the WG. Figure 17.10 shows a structure of the

MBE-grown sample for fabricating the PC-WG with SA-grown QDs. A 250 nm thickness GaAs

core layer with QDs on a 2 μ m thickness sacriﬁ cial clad layer was grown on a GaAs substrate. To

minimize the step height of the selective area, the MM was used for only QD layers and capping

layers of 6 nm in thickness.

InAs QDs

Core ~250nm

Step height ~6nm

SAG area

Figure 17.9 Schematic of PC-WG embedding SA-grown QDs. The ratio of a step height at SAG area to a core

thickness is below 2.4%, which should be negligible for pulse propagations.

GaAs 90 nm

0.3

0.7

As 10 nm

GaAs 20 nm

GaAs (001)

GaAs buffer

0.3

0.7

As 10 nm

GaAs 90 nm

GaAs 20 nm

0.6

0.4

As 2000 nm

GaAs: 3 nm

0.2

0.8

As: 3 nm

InAs-QD: 2.6 ML

With MM

Step height 6nm

Core layer

~250 nm

Sacrificial layer

for air-bridge

~2 μm

Figure 17.10 MBE-grown sample structure for PC-WG with SA-grown QDs.

The MBE-grown sample was then transferred to electron beam lithography and dry etching

processes for fabricating the air bridge-type 2DPC slab [17] . The PC is constructed with holes of

hexagonal arrays. In the PC, a missing line of holes is prepared as a straight WG whose input and

output ports involve solid immersion lens and taper structures [18] resulting in a high coupling

efﬁ ciency. By marking the position of SAG in advance, one can set the SA-grown QDs within the

WG. The lattice constant and radius of air holes are designed to include the centre wavelength of

the SA-grown QDs, around 1300 nm, within the transmission band of the WG.

Figure 17.11a shows an optical microscope image of a fabricated sample. The area surrounded

by the dashed lines is embedded with the QDs. A PL intensity mapping obtained from the same

area in Fig. 17.11a exhibits clear PL emission of 1.29 μ m in wavelength from QDs embedded

in a selective area, as shown in Fig. 17.11b . These images indicate successful fabrication of the

PC-WG, embedding the SA grown QDs.

We then characterized the fabricated PC-WGs by CW transmission and pump/probe pulse

measurements. The transmission spectra were observed with CW white light as a light source

CH017-I046325.indd 536CH017-I046325.indd 536 6/25/2008 3:17:24 PM6/25/2008 3:17:24 PM

Advanced Growth Techniques of InAs-system Quantum Dots for Integrated Nanophotonic Circuits 537

and a conventional measurement set-up. Figure 17.12b shows the transmission spectra obtained

from the PC-WG. The transmission band through the WG mode is seen from 1270 nm to

1330 nm indicated by a black arrow. This value almost corresponds to the transmission band of

the straight WG without QDs, therefore, the step height of the SAG area, below 6 nm, should not

inﬂ uence the transmission. An absorption attributed to the QDs is observed around 1290 nm, as

depicted by a dashed line, corresponding to the inverse of the PL spectrum shown in Fig. 17.12a .

From the above results, we conﬁ rmed a successful fabrication of the PC-WG embedding the

SA-grown QDs; the transmission band was set to include the QD absorption centre wavelength,

while no inﬂ uence of the step height was observed.

0.3

0.5

0.8

PL intensity (a.u.)

500 μm

(a) (b)

SA-grown QD area

Figure 17.11 (a) Optical microscope image of a fabricated sample with PC/QD straight WGs. (b) PL intensity

mapping from the sample. PL intensity of 1290 nm in peak wavelength due to the SA-grown QDs is clearly shown.

Figure 17.12 (a) PL spectrum from the SA-grown QDs at RT. (b) The transmission spectrum obtained from the

PC/QD WG.

PL intensity (arb. unit)

Peak : 1292 nm

FWHM: 29.3 meV

@RT

1200 1250 1300 1350 1400

50

40

30

Intensity (dB)

Wavelength (nm)

(a)

(b)

pump

probe

PC-WG

Absorption

by QDs

CH017-I046325.indd 537CH017-I046325.indd 537 6/25/2008 3:17:25 PM6/25/2008 3:17:25 PM

538 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics

Then, we measured the non-linear phenomena by the QD, amplitude and phase shift of signal

pulses (SPs) induced by control pulses (CPs), which are essential for the operation of PC-SMZ.

The optical non-linearity-induced modulations were measured by utilizing a two-colour pump

and probe set-up, as shown in Fig. 17.13 [7] .

Probe

AOM1

AOM2

OPO (probe)

Delay

Sample

Reference

Pump

OPO (pump)

Delay

PIN PD

Figure 17.13 Schematic of the experimental apparatus for two-colour pump/probe measurements.

The SPs and CPs were collinearly coupled into the PC-WG. By considering the transmission

spectrum shown in Fig. 17.12b , we set the CPs ’ centre wavelength to the centre of the QDs ’

absorption wavelength, 1290 nm, and set the SPs ’ centre wavelength to a longer wavelength,

1310 nm. The ONL-induced amplitude and phase shifts of the SPs were measured by lock-in-

based heterodyne detection. Figure 17.14 shows typical non-linearity-induced amplitude and

phase modulations of SP against time delay between the CP and the SP. The shift of the ampli-

tude and the phase of SPs due to the CPs were clearly observed. The energies of the SP and the

CP in the WG were estimated to be 33.8 fJ/pulse and 338 fJ/pulse, respectively. The net energy is

estimated from the coupling efﬁ ciency into the PC-WG.

100 0 100 200 300 400 500

0.5

0.0

0.5

1.0

ΔAmp. (a. u.)

CP/SP delay (ps)

100 0 100 200 300 400 500

20

Phase shift (deg)

CP/SP delay (ps)

Figure 17.14 The ONL-induced amplitude and phase shifts of SPs against time delay between CPs and SPs.

In this case, we have obtained the phase shift value around 45°. This value is consistent with

that previously obtained from the sample with entire-grown three-layer-stacked QDs, e.g. 130°

[7] , considering that this sample includes a single layer of QDs. However, for PC-SMZ operations,

which requires the SP phase shift of at least 180°, further improvement of the non-linearity efﬁ -

ciency is necessary, e.g. by QD stacking growth or utilizing the slow light in the hetero V

(group

velocity) PC-WG [19] .

17.2.6 Summary

We have succeeded in the SAG of InAs QDs by the developed MM/MBE method. The SA-grown

QDs, which exhibit comparable structural and optical qualities with conventional QDs grown

CH017-I046325.indd 538CH017-I046325.indd 538 6/25/2008 3:17:26 PM6/25/2008 3:17:26 PM