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 549

large potential strain ﬁ eld on the GaAs covered layer and creates a few nucleation sites. This con-

ﬁ guration can be easily changed by varying the thickness of the covered layer and the amount of

InAs for the second layer QDs; in this way, the density of the second layer QDs can be controlled.

These results indicate that selective growth of InAs QDs can be achieved by using the In nano-

dots produced by the NJP method.

200 nm pitch (2.5  10



)

In nano-dot

array

(NJP QDs)

4 μm

Figure 17.29 AFM image of a fabricated In nano-dot array.

1 μm

Figure 17.30 AFM image of the stacked high-density QD array.

17.3.4 Summary

We have demonstrated the use of a new nanoprobe-assisted technology that forms InAs QDs on

a GaAs substrate by a microfabrication technique (NJP) employing a specially designed canti-

lever. Using the NJP, uniform In nano-dot arrays were formed in a reproducible manner on an

MBE-grown surface by applying electric pulses between the tip and the sample. These In nano-

dot arrays were converted directly into InAs QD arrays by subsequent annealing with irradiation

of arsenic ﬂ ux. Furthermore, we have demonstrated the selective area growth of high-density

InAs QDs by using site-controlled InAs QDs that were formed in the desired regions as templates.

The NJP method enables the formation of the required number of InAs QDs in a desired region. In

other words, the NJP method offers perfect selectivity in forming InAs QDs. An example of the inte-

gration of QDs into photonic devices is a photonic crystal-based ultra-fast all-optical switch which

CH017-I046325.indd 549CH017-I046325.indd 549 6/25/2008 3:17:33 PM6/25/2008 3:17:33 PM

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

we are currently developing [5] . These devices use QDs as a non-linear optical material to switch the

optical signal pulses. For these devices, QDs should be formed only in the photonic crystal straight

waveguide area. The NJP method will enable the formation of the required number of QDs in a

desired region with both high uniformity and high density. Furthermore, we can achieve the required

volume of QDs for optical switching by combining this site-controlled QD formation technique with

the subsequent stacking of spatially ordered InAs QD arrays using MBE growth [31] . In addition

to our proposed PC-based devices [5] , the NJP method has applications to other optical switching

devices, including future high-performance functional devices such as regular arrays of quantum

bits and single-photon emitters for quantum computers and quantum communications [10, 11] .

References

1. Y. Arakawa and H. Sakaki , Multidimensional quantum well laser and temperature dependence of

its threshold current , Appl. Phys. Lett. 40 , 939 – 941 ( 1982 ) .

2. M. Sugawara , N. Hatori , T. Akiyama , Y. Nakata , and G. Ishikawa , Quantum-dot semiconductor

optical ampliﬁ ers for high bit-rate signal processing over 40 Gbit/s , Jpn. J. Appl. Phys. 40 , L488 –

L491 ( 2001 ) .

3. S. Kohmoto , H. Nakamura , T. Ishikawa , and K. Asakawa , Site-controlled self-organization of indi-

vidual InAs quantum dots by scanning tunneling probe-assisted nano-lithograph , Appl. Phys.

Lett. 75 , 3488 – 3490 ( 1999 ) .

4. S. Ohkouchi , Y. Nakamura , H. Nakamura , and K. Asakawa , InAs nano-dot array formation using

nano-jet probe for photonics applications , Jpn. J. Appl. Phys. 44 , 5777 – 5780 ( 2005 ) .

5. K. Asakawa , Y. Sugimoto , Y. Watanabe , N. Ozaki , A. Mizutani , Y. Takata , Y. Kitagawa , H. Ishikawa ,

N. Ikeda , K. Awazu , X. Wang , A. Watanabe , S. Nakamura , S. Ohkouchi , K. Inoue , M. Kristensen ,

O. Sigmund , P.I. Borel , and R. Baets , Photonic crystal and quantum dot technologies for all-optical

switch and logic device , New J. Phys. 8 , 208 ( 2006 ) .

6. K. Tajima , All-optical switch with switch-off time unrestricted by carrier lifetime , Jpn. J. Appl.

Phys. 32 , L1746 – L1749 ( 1993 ) .

7. H. Nakamura , Y. Sugimoto , K. Kanamoto , N. Ikeda , Y. Tanaka , Y. Nakamura , S. Ohkouchi ,

Y. Watanabe , K. Inoue , H. Ishikawa , and K. Asakawa , Ultra-fast photonic crystal/quantum dot

all-optical switch for future photonic network, Opt. Express 12 , 6606 – 6614 ( 2004 ) .

8. N. Ozaki , Y. Takata , S. Ohkouchi , Y. Sugimoto , Y. Nakamura , N. Ikeda , and K. Asakawa , Selective

area growth of InAs quantum dots with a metal mask towards optical integrated circuit devices ,

J. Cryst. Growth 301 –302 , 771 – 775 ( 2007 ) .

9. T. Yoshie , A. Scherer , J. Hendrickson , G. Khitrova , H.M. Gibbs , G. Rupper , C. Ell , O.B. Schchekin ,

and D.G. Deppe , Vacuum Rabi splitting with a single quantum dot in a photonic crystal nanocav-

i t y , Nature 432 , 200 – 203 ( 2004 ) .

10. E. Waks , K. Inoue , C. Santori , D. Fattal , J. Vuckovic , G.S. Solomon , and Y. Yamamoto , Secure com-

munication: quantum cryptography with a photon turnstile , Nature 420 , 762 ( 2002 ) .

11. A. Imamoglu , D.D. Awschalom , G. Burkard , D.P. DiVincenzo , D. Loss , M. Sherwin , and A. Small ,

Quantum information processing using quantum dot spins and cavity QED , Phys. Rev. Lett. 83 ,

4204 – 4207 ( 1999 ) .

12. S. Ohkouchi , Y. Nakamura , H. Nakamura , and K. Asakawa , Indium nano-dot arrays formed by

ﬁ eld-induced deposition with a nano-jet probe for site-controlled InAs/GaAs quantum dots , Thin

Solid Films 464– 465 , 233 – 236 ( 2004 ) .

13. Y. Nakamura , N. Ikeda , S. Ohkouchi , Y. Sugimoto , H. Nakamura , and K. Asakawa , Regular array

of InGaAs quantum dots with 100-nm-periodicity formed on patterned GaAs substrates , Physica

E 21 , 551 – 554 ( 2004 ) .

14. Y. Nakamura , N. Ikeda , S. Ohkouchi , Y. Sugimoto , H. Nakamura , and K. Asakawa , Two-dimensional

InGaAs quantum-dot arrays with periods of 70–100 nm on artiﬁ cially prepared nanoholes , Jpn.

J. Appl. Phys. 43 , L362 – L364 ( 2004 ) .

15. K. Nishi , H. Saito , S. Sugou , and J.S. Lee , A narrow photoluminescence linewidth of 21 meV

at 1.35 μ m from strain-reduced InAs quantum dots covered by In

0.2

0.8

As grown on GaAs

substrates, Appl. Phys. Lett. 74 , 1111 – 1113 ( 1999 ) .

CH017-I046325.indd 550CH017-I046325.indd 550 6/25/2008 3:17:34 PM6/25/2008 3:17:34 PM

Advanced Growth Techniques of InAs-system Quantum Dots for Integrated Nanophotonic Circuits 551

16. S. Ohkouchi , Y. Nakamura , H. Nakamura , N. Ikeda , Y. Sugimoto , and K. Asakawa , Selective

growth of high quality InAs quantum dots in narrow regions using in situ mask , J. Cryst. Growth

293 , 57 – 61 ( 2006 ) .

17. Y. Sugimoto , N. Ikeda , N. Carlsson , K. Asakawa , N. Kawai , and K. Inoue , Fabrication and charac-

terization of different types of two-dimensional AlGaAs photonic crystal slabs , J. Appl. Phys. 91 ,

922 – 929 ( 2002 ) .

18. N. Ikeda, H. Kawashima, Y. Sugimoto, T. Hasama, K. Asakawa, and H. Ishikawa, Coupling char-

acteristic of micro planar lens for 2D photonic crystal waveguides, 19th International Conference

on InP and related materials (IPRM) (2007).

19. N. Ozaki , Y. Kitagawa , Y. Takata , N. Ikeda , Y. Watanabe , A. Mizutani , Y. Sugimoto , and

K. Asakawa , High transmission recovery of slow light in a photonic crystal waveguide using a

hetero group-velocity waveguide , Opt. Exp. 15 , 7974 – 7983 ( 2007 ) .

20. C. Lee and A.L. Barabasi , Spatial ordering of islands grown on patterned surfaces , Appl. Phys. Lett.

73 , 2651 – 2653 ( 1998 ) .

21. S. Ohkouchi , Y. Nakamura , H. Nakamura , and K. Asakawa , Nano-probe-assisted technology of

indium-nano-dot formation for site-controlled InAs/GaAs quantum dots , Physica E 21 , 597 – 600

( 2004 ) .

22. S. Ohkouchi , Y. Sugimoto , N. Ozaki , H. Ishikawa , and K. Asakawa , Selective growth of InAs quan-

tum dots using In nano-dot arrays formed by nano-jet probe method , J. Cryst. Growth 301 –302 ,

726 – 730 ( 2007 ) .

23. H.J. Mamin , P.H. Guethner , and D. Ruger , Atomic emission from a gold scanning-tunneling-micro-

scope tip , Phys. Rev. Lett. 65 , 2418 – 2421 ( 1990 ) .

24. N. Aoki , K. Fukuhara , T. Kikutani , A. Oki , H. Hori , and S. Yamada , Fabrication of buried metal dot

structure in split-gate wire by scanning tunneling microscope , Jpn. J. Appl. Phys. 35 , 3738 – 3742

( 1996 ) .

25. K. Higa and T. Asano , Joule heating of ﬁ eld emitter tip fabricated on glass substrate , Jpn. J. Appl.

Phys. 43 , 2749 – 2750 ( 2004 ) .

26. G. Taylor , Electrically driven jets , Proc. Roy. Soc. Lond. A 313 , 453 – 475 ( 1969 ) .

27. R.P.A. Hartman , D.J. Brunner , D.M.A. Camelot , J.C.M. Marijnissen , and B. Scarlett ,

Electrohydrodynamic atomization in the cone-jet mode physical modeling of the liquid cone and

jet , J. Aerosol Sci. 30 , 823 – 849 ( 1999 ) .

28. T. Chikyow and N. Koguchi , MBE growth method for pyramid-shaped GaAs micro crystals on

ZnSe(001) surface using Ga droplets , Jpn. J. Appl. Phys. 29 , L2093 – L2095 ( 1990 ) .

29. R.S. Wagner and W.C. Ellis , Vapor-liquid-solid mechanism of single crystal growth , Appl. Phys.

Lett. 4 , 89 – 90 ( 1964 ) .

30. N. Koguchi , K. Ishige , and S. Takahashi , New selective molecular-beam epitaxial growth method

for direct formation of GaAs quantum dots , J. Vac. Sci & Technol. B

11 , 787 – 790 ( 1993 ) .

31. Y. Nakamura , N. Ikeda , S. Ohkouchi , Y. Sugimoto , H. Nakamura , and K. Asakawa , Two-

dimensional InGaAs quantum-dot arrays with periods of 70–100 nm on artiﬁ cially prepared

nanoholes , Jpn. J. Appl. Phys. 43 , L362 – L364 ( 2004 ) .

CH017-I046325.indd 551CH017-I046325.indd 551 6/25/2008 3:17:35 PM6/25/2008 3:17:35 PM

Nanostructured Solar Cells

S.G. Bailey ,

Seth Hubbard,

and R.P. Raffaelle

Space Environment and Experiments Branch, NASA Glenn Research Center,

MS 302-1, Cleveland, OH 44135;

Rochester Institute of Technology, Rochester, NY 14623

18.1 Introduction

There has been considerable investigation regarding the potential for the use of nanomateri-

als and nanostructures to increase the efﬁ ciency of photovoltaic devices [1] . Examples of such

investigation include the use of materials and/or combinations of materials that provide either

one-dimensional, two-dimensional, or even three-dimensional quantum conﬁ nement (i.e. quan-

tum wells [2] ; quantum wires [3] ; or quantum dots [4,5] , respectively). Quantum conﬁ nement

refers to the fact that a dimension or dimensions associated with a given material are smaller

than the Bohr exciton radius for that material. This results in the possible electronic states asso-

ciated within the material to deviate from the ordinary band theory of solids and begin to show

discrete-like optoelectronic behaviour more familiar with atoms or molecules. The term nano-

material would generally be reserved for isolated materials that have at least some degree of

quantum conﬁ nement (e.g. colloidal suspensions of semiconducting nanocrystals, fullerenes,

single- or multi-wall carbon nanotubes). A nanostructure would commonly refer to a composite

material in which the boundaries of one material by another results in the dimensions of the ﬁ rst

material to be on the nanoscale. A prime example of this would be an extremely thin (i.e. a few

nanometres) epitaxially grown layer of a narrow band semiconductor within a wider band gap

host material. This resulting “ nanostructured ” material would be a quantum well if the thick-

ness of this narrow band gap layer is of a sufﬁ ciently small dimension, as mentioned above.

Nanostructures can also be comprised of arrays of individual nanomaterials. Semiconducting

quantum dots (QDs) can be combined in a three-dimensional array, often through the use of self-

ordering. The discrete-like energy levels of the quantum dots will combine and form bands of

allowed energy states in an analogous way in which atomic energy levels combine to produce the

energy bands in conventional solids. In this case, it is easy to see why quantum dots are some-

times referred to as “ super atoms ” . By changing the quantum dot size and the spacing between

the quantum dots, the optoelectronic bandstructure will change in ways similar to the way a

semiconductor bandstructure changes with compositional doping. The shift in the effective band

gap will vary inversely with the particle size of the dot [6] .

Quantum wells or quantum wires can also be arrayed. The use of many closely spaced quan-

tum wells, or a multiple quantum well, is a very well-known structure within the optoelectronic

industry. This has been a prime means of “ band gap engineering ” semiconducting materials. By

changing the quantum well width, spacing of the wells (i.e. the thickness of the barriers between

the wells), and the composition well and/or barrier material, one can effectively “ tune ” the band

gap to values that cannot be achieved using merely compositional changes alone.

CHAPTER 18

CH018-I046325.indd 552CH018-I046325.indd 552 6/27/2008 5:20:08 PM6/27/2008 5:20:08 PM

Nanostructured Solar Cells 553

The role of a nanomaterial or nanostructure in a given photovoltaic solar cell design can vary

dramatically. In some cases, the goal may be simply to provide the means to disassociate exci-

tons throughout a bulk material as in the use of colloidal quantum dots in organic or polymeric

solar cells. In other cases, the use of a nanostructure is motivated by the desire to band gap engi-

neer a device in a similar approach to what is done more conventionally through compositional

changes in a device, albeit in regimes where compositional solutions may not be readily available.

A more ambitious use of nanostructures for photovoltaics lies in the desire to exploit underlying

quantum conﬁ nement concepts to fundamentally change the way we can exploit photoconver-

sion for electrical energy production. Examples of quantum mechanical phenomena offered by

nanostructures that could be exploited to exceed the standard Shockley–Quiesser limit for a con-

ventional solar cell [7] (i.e. the maximum thermodynamic efﬁ ciency for the conversion by a sin-

gle band gap cell of unconcentrated solar irradiance into electrical free energy of approximately

31%) are hot carrier effects, multiple exciton generation, phonon bottlenecking, and up- and

down-conversion. Hot-carrier solar cells [8] attempt to use nanostructures to obviate the normal

thermalization processes associated with conventional solar cells, or in other words, extract the

carriers before they relax down to the conduction band edge of the host semiconductor. Solar

cells attempting to exploit the multiple exciton generation processes demonstrated by colloidal

by quantum dots hope to produce multiple electron–hole pairs per photon through this inverse

Auger impact ionization [9] . Finally, the use of quantum dot arrays may allow approaches that

expand the spectral response of conventional solar cells by shifting incoming photon energy from

inaccessible regions or regions of the spectrum that are inefﬁ ciently converted to more favoura-

ble energy ranges, and potentially provide the means for multiband intermediate band cells [10] .

This chapter discusses a few of the recent approaches to produce nanostructured solar cells.

18.2 Quantum dot solar cells

The term quantum dot solar cells is somewhat ambiguous as it is colloquially used to refer to two

general types of solar cells in which both the role of the quantum dots, the materials and synthe-

sis methods used to create them, and the types of cells in which they are used, are drastically dif-

ferent. These two classes of cells are those which use colloidal quantum dots in a polymer matrix

and their crystalline counterparts which utilize epitaxially grown, normally Stranski–Krastanov

(SK) mode, quantum dots (QDs). The motivation for the former class is due to the rapid increase

in device efﬁ ciencies in polymer solar cells with the addition of nanomaterials, such as CdSe

quantum dots and more recently fullerenes. This work has led to the type of solar cell called a

distributed heterojunction solar cell. The latter class of quantum dot solar cells are those involv-

ing the epitaxially grown quantum dots, usually comprised of III–V materials in the form of

quantum dot arrays (QDAs). Much of the motivation for the research in this area was provided

by the theoretical work on intermediate band solar cells (IBSCs). It has been proposed that an

IBSC could be achieved via a quantum dot array embedded into a suitable host solar cell struc-

ture. It is still somewhat of an open debate whether or not the IBSC band solar cell as proposed in

the aforementioned theoretical treatments is theoretically achievable using a QDA.

18.2.1 Intermediated band solar cells

Theoretical results of Luque and Marti have shown that a photovoltaic device with a single inter-

mediate band of states resulting from the introduction of quantum dots offers a potential efﬁ -

ciency of 63.2% [11] . This was extended to two intermediate bands and a limiting efﬁ ciency of

71.7% was calculated [9] . The enhanced efﬁ ciency results from converting photons of energy

less than the band gap of the cell by an intermediate band. The intermediate band provides a

mechanism for low-energy photons to excite carriers across the energy gap by a two-photon

process. In principle, the use of a QDA to achieve this intermediate band may have some distinct

advantages to that of a multiple quantum well (MQW), such as in directionality of the illumina-

tion that can result in the ability to optically excite carriers from the intermediate band into the

CH018-I046325.indd 553CH018-I046325.indd 553 6/27/2008 5:20:09 PM6/27/2008 5:20:09 PM

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

host semiconductor conduction band and in terms of the practical range of tunability offered by

a QDA over that of an MQW.

Quantum dots offer the potential to control the intermediate band energies since the indi-

vidual quantum energy levels associated with isolated quantum dots is a function of their size,

separation from one another, and material composition. Placing the appropriate quantum dot

material of the necessary size into an organized matrix within an ordinary p-i-n structure solar

cell should result in the formation of accessible energy levels within what would normally be the

forbidden band of the device. A simple schematic of one such proposed intermediate band solar

cell structure is shown in Fig. 18.1 .

Grid fineer

p-InGaP window

p-GaAs emitter

Back contact

n-InGaP window

n⬘-GaAs buffer

n⬘-GaAs substrate

n⬘-GaAs base

InAs QDA InGaAs

Figure 18.1 Intermediate band cell structure.

The electronic wavefunctions associated with the discrete electronic states of the individual

quantum dots in the QDA array will overlap creating “ mini-bands ” in the host semiconductors ’

band gap. The composition and related properties (i.e. bulk band gap, electron afﬁ nity, effective

masses, etc.), size, symmetry, and spacing of the quantum dots need to be chosen to produce

mini-bands which are appropriately spaced within the band gap of the host material (see Fig.

18.2 ). The majority of the work to date on this type of QDA for use in a solar cell has been on the

development of InAs QDAs in GaAs. Fortunately, the strain associated with the SK growth of QDs

acts to self-align the QD into vertical columns to produce the desired arrays as each additional

layer is added in this system. More recently, it has been proposed that QDAs of Si could be used in

a similar fashion in an amorphous silicon solar cell [12] .

A major challenge in designing solar cells which can take maximum advantage of a QDA is the

balance between balancing absorption and transport. A QDA with many layers could improve the

subhost band gap absorption at the cost of the efﬁ cient extraction photo-generated carriers from

the quantum-conﬁ ned region [13] due to the diminished electric ﬁ eld across the QDA within the

intrinsic region of the device and increased probability of recombination and radiative losses.

Transport of carriers generated in bulk would presumably also suffer as the QDA is widened. On

the other hand, a QDA that is too thin would provide little beneﬁ t from photoconversion of the

subhost band gap illumination to offset the increased defect density that would be likely to result

from its inclusion. Recent modeling of an IBSC based on quantum dot supracrystals was under-

taken to engineer optimum array dimensions necessary for maximum solar cell efﬁ ciency [14] .

CH018-I046325.indd 554CH018-I046325.indd 554 6/27/2008 5:20:09 PM6/27/2008 5:20:09 PM

Nanostructured Solar Cells 555

Using the material system of an InAsN/GaAsSb quantum dot supralattice, it was shown that

the power conversion efﬁ ciency of the IBSC was maximized by a quantum dot size of approxi-

mately 4.7 nm. It was noted that solar cells based on these quantum dots may also show

enhanced radiation hardness [15] and improved collection efﬁ ciency [16, 17] .

Transport through the QDA can be enhanced by reducing the spacing between the QD layers.

This, of course, will change the absorption spectrum and reduce the 3D quantum conﬁ nement.

Taken to the extreme, this type of a cell would approach what would be in a sense a quantum

wire solar cell. The closely spaced QDs would form nanoscale wires across the intrinsic region of

the device (see Fig. 18.3 ) . In such a device, the presence of a one-dimensional electron gas across

the junction region will result in a fast collection of the carriers photo-generated in the quantum

wires region, and the solar cell output will thus fully beneﬁ t from improved absorption properties

associated with the presence of quantum conﬁ ned states. In practice, arrays of vertical quantum

wires can be formed by a self-organization process occurring during the deposition of ultra-short

period stacks of highly strained/strain balanced ﬁ lms of III–V semiconductors (e.g InAs/AlAs on

InP, InGaAs/InGaP on GaAs, etc.). One could also envision potentially producing solar cells with

such structures utilizing other materials through electron lithography (wire widths in excess of

20–30 nm).

p-type

region

Insulating

region

n-type

region

Quantum dot

mini-bands

Vac

Figure 18.2 Idealized energy band diagram of an intermediate band solar cell.

⫹

-emitter

n-Base

i-region

(Quantum wires)

Z(001)

Figure 18.3 Schematic representation of the proposed vertical quantum wire device.

Modeling of a quantum wire structure was undertaken in which the dots were assumed to

stack vertically in the cell [18] . For an indium phosphide emitter and base with indium arsi-

nide wires, there was a peak efﬁ ciency with quantum wire length. It has been suggested that the

quantum wire solar cell, while circumventing carrier collection issues (when compared to quan-

tum dot and quantum well systems), may also provide a signiﬁ cantly larger density of states and

consequently enhanced photo-absorption.

CH018-I046325.indd 555CH018-I046325.indd 555 6/27/2008 5:20:10 PM6/27/2008 5:20:10 PM

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

18.3 Quantum dot growth

18.3.1 Stranksi–Kranstanow quantum dots

The ideal QD behaves much like the idealized box in the famous quantum mechanical of the par-

ticle-in-a-box problem. This QD is a structure capable of conﬁ ning carriers (electrons and holes)

in three-dimensions, allowing zero dimensions (0D) in their degrees of freedom, and creating

atom-like levels with discrete (Dirac delta) densities of states. This is why quantum dots can also

be visualized and described as “ artiﬁ cial atoms ” . In SK grown QDs, this structure is achieved by

having a very small volume of a low band gap material embedded in larger band gap material(s),

thus creating the potential barrier for quantum conﬁ nement of the carriers. Until recently, such

three-dimensional conﬁ nement in semiconductor structures existed mostly in theoretical treat-

ments. In addition to SK growth, other approaches to manufacture quantum dots made use of

photolithography and etching. While providing good control of the order and offering design

ﬂ exibility; these types of processing affected the interfaces, introducing impurities and surface

damage that limited the optical performance of these quantum structures.

The SK growth mode for III–V QDs takes advantage of a strain-induced transformation that

happens naturally in the initial stages of growth for lattice-mismatched materials. The growth

usually starts layer by layer, and after a certain critical thickness is reached, the structure spon-

taneously forms nanometre-size islands (SK mode [19] ) that show good size uniformity and large

surface densities. In this process, the growth is interrupted [20] immediately after formation

of the islands, but before the islands reach a size for which strain relaxation and misﬁ t disloca-

tions occur. This spontaneous island formation during growth precludes the interface quality

problems often associated with ex situ processed quantum structures of low dimensionality.

The mechanisms of SK strain-induced QD growth in molecular beam epitaxy (MBE) are fairly

straightforward. However, in the case of organometallic vapour phase epitaxy (OMVPE) to initi-

ate the SK growth mode, growth temperatures must be reduced to well below the values typically

associated with mass-transport limits [21] . The result is that QD nucleation, growth, and uni-

formity becomes much more sensitive to OMVPE growth parameters, such as growth tempera-

ture, growth rate and V/III ratio. However, for commercial production of solar cells, the OMVPE

has deﬁ nite advantages in terms of production rate, scalability, and cost to its MBE counterpart.

The growth of SK mode InAs QDs on GaAs by organometallic vapour phase epitaxy (OMVPE)

or metal organic chemical vapour deposition (MOCVD) commonly use metallorganics, such

as trimethyl gallim (TMGa) and trimethyl indium (TMIn) as precursor materials, along with

hydrides phosphine (PH

), arsine (AsH

), and 1% AsH

in hydrogen. The typical growth tem-

peratures are between 500 and 700

C and a pressure of around 600 torr. An example of how

OMVPE growth conditions effect InAs quantum dots on GaAs can be seen in Fig. 18.4 .

(a) (b)

0.25 0.50 0.75 1.00

μm μm

0.25

0.15 nm

0.0 nm

0.30 nm

0.50

0.75

1.00

0.25

0.50

0.75

1.00

Figure 18.4 AFM micrographs near wafer center for OMVPE-grown SK mode InAs QDs grown on GaAs with

V–III ratios of (a) 58 and (b) 12 [20] .

CH018-I046325.indd 556CH018-I046325.indd 556 6/27/2008 5:20:11 PM6/27/2008 5:20:11 PM

Nanostructured Solar Cells 557

20 nm

0.1

100

1000

10000

15.0

Omega - two theta (deg)

Count rate (a.u.)

GaN

InN

500

400

300

100 nm

200

20 nm

15.5

16.0

16.5 17.0 17.5 18.0

Figure 18.5 AFM image of InN quantum dots grown on a GaN buffer layer. The presence of InN material has been

veriﬁ ed by X-ray diffraction spectrum shown on the inset [21].

75 sec

400 sec

Figure 18.6 Height of the InN islands as a function of growth time for a series of samples deposited at a

temperature set point of 720 °C. Insets show the AFM image from 5 mm ⫻ 5 mm scans of samples grown for 75

seconds and 400 seconds. The lines through the data points are provided as a guide to the eye [21].

There are other examples of quantum dot formation by OMVPE. A recent study of InN dots syn-

thesized on GaN templates has been reported [22] . Using sapphire substrates, a 1.5 micron thick

GaN template was grown using a standard two-step process prior to the deposition of InN. The lat-

tice mismatch between InN and GaN is quite large ( ⬎ 10%), causing to proceed in a three-dimen-

sional mode. Indeed, AFM images revealed the formation of discrete InN islands on the GaN surface,

as illustrated in Fig. 18.5 and similar to reports from other groups [23, 24] . X-ray diffraction meas-

urements conﬁ rmed the deposition of InN in the sample set grown on GaN templates (see Fig. 18.5

inset). The strong peak near 17.3 ° seen in the Fig. 18.5 inset corresponds to diffraction from the

GaN lattice in the buffer layer, while the weaker peak near 15.8 ° is in the range expected for epitax-

ial InN. The intensity of the InN-related peak generally increases with growth time, and both the

position and full width-at-half maximum exhibit a slight dependence on the growth conditions.

Figure 18.6 demonstrates the evolution in the height of the InN islands with the growth

time under ﬁ xed deposition conditions. The InN islands do not form until approximately 15 to

20 seconds of growth time, and they are still discrete after 900 seconds of growth. The delayed

CH018-I046325.indd 557CH018-I046325.indd 557 6/27/2008 5:20:12 PM6/27/2008 5:20:12 PM

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

onset of three-dimensional islands is consistent with a Stranski–Krastanov growth mode for the

deposition of InN on a GaN template. The height of the InN islands increases rapidly from 7 nm

to 50 nm as the growth time increases from 25 to 225 seconds, and then begins to saturate,

approaching 80 nm as the growth time increases further from 300 to 900 seconds. At ﬁ rst, the

density of the InN islands remains constant near 5 ⫻ 1 0

⫺ 2

, and the distribution of island

sizes is relatively narrow, as illustrated by the AFM image of the sample grown for 75 seconds

(see inset in Fig. 18.5) . However, a second group of much smaller islands appears in structures

with a growth time of 300 seconds or greater, resulting in a total InN island density approaching

1 ⫻ 1 0

⫺ 2

. The bimodal distribution of InN islands is evident in the AFM image of the 400

second sample also seen in the inset in Fig. 18.5 . The appearance of the second group of InN

islands near 300 seconds in growth time corresponds to the saturation in the height of the initial

set of InN islands.

Photoluminescence emissions from both dots and wetting layer are observed in other lattice-

mismatched III–V material systems, most notably InAs on GaAs [25] . Non-radiative recom-

bination at surface states is well known to degrade the photoluminescence of many III–V

semiconductors, and most PL studies of InAs dots and wetting layers on GaAs employ capped

structure and/or low-temperature measurements to enhance emission intensities. However,

nitride materials can behave quite differently. For example, room temperature photoluminescence

emission is observed in thin, uncapped GaN ﬁ lms grown on wider energy gap AlGaN layers [26] .

While surface states may be less severe in nitride materials, it has been observed that the buffer

layer quality has a strong impact on the intensity of PL emissions from InN dots.

Quantum dots of Si have also been formed by alternating layers of silicon dioxide followed by

a silicon-rich layer of the same material and then heat treating [27] . On heating, the surface

energy minimization favoured the precipitation of silicon into spherical quantum dots. These

quantum dots could be inserted into an amorphous silicon solar cell. A proposed all-silicon tan-

dem cell has been proposed.

18.4 Organic quantum dot solar cells

Improvement in polymer solar cell performance has long suffered from several shortcomings (i.e.

low electron mobility, absorption characteristics that were not well suited to the solar spectrum,

environmental degradation, etc.). However, the lure of potential low cost and large area manu-

facturing of thin ﬁ lm ﬂ exible solar cells continued to attract researchers [28] . A main fundamen-

tal limitation associated with the use of organic thin ﬁ lms for PV is that photonic absorption in

these materials produces bound state excitons. Dissociation of these charge pairs requires a sub-

stantial potential difference across a polymer–metal or polymer–semiconductor junction, pro-

vided the excitons are near the interface. Thus, polymer solar efﬁ ciencies improved dramatically

with the introduction of the use of nanomaterial/conjugated polymer complexes or what are

now commonly referred to as bulk heterojunction solar cells [29] . This refers to the fact that the

materials that are responsible for both the donors and acceptors are intimately mixed through-

out the absorber layer of the device.

A suitable choice of QD, or alternatively a fullerene, can provide the necessary electron-accept-

ing impurity in a polymer matrix. Thus, under illumination a preferential transfer of electrons

from the excitons to the acceptors leaves behind holes. This process is known as photo-induced

charge transfer. The carriers, once liberated from one another, are now free to be transported

through the conjugated polymer. All that remains is to incorporate these composites or blends

into a suitable device structure.

A variety of acceptor materials have been introduced into conjugated polymers to produce

photovoltaic devices (i.e. buckminster fullerenes [30] , CdSe quantum dots, CdSe nano rods [31] ,

and single-wall carbon nanotubes [32] (SWCNT)) [33] . The devices are produced by placing the

doped polymeric ﬁ lms between a transparent conductive oxide, typically indium tin oxide (ITO),

top contact and a metallic back contact. There has also been recent work on thin ﬁ lm polymeric

solar cells which incorporate quantum dots as well as quantum dot/SWCNT complexes [34] .

CH018-I046325.indd 558CH018-I046325.indd 558 6/27/2008 5:20:14 PM6/27/2008 5:20:14 PM