SEMICONDUCTOR NANOCRYSTALS

Louis Brus
Chemistry Department and Columbia Radiation Laboratory
Columbia University
New York, NY 10027

Scientific Drivers, Technology, and Critical Questions

Electrical Circuits for Computers

The silicon microelectronics industry is the technological driver of modern society. The industry is built upon two major inventions: the solid state transistor and the photolithographic integrated circuit. The present technology is robust and advancing rapidly, as shown by the amazing improvement in and decrease in cost of simple PCs. Much of this is due to continued miniaturization in both magnetic memory and Si device size. This engineering improvement will continue to an ultimate Si feature size of perhaps 3-6 nm, an order of magnitude smaller in size than the smallest production devices at present. At even smaller sizes, several factors slow progress: basic statistical fluctuations in the power necessary for reliable driving of one device by the next and in the performance of individual transistors, coupled with the optical problem of photolithography on such a small scale. It is likely to be only several decades into the future when Si circuits ultimately reach a practical size limit. However, in approaching the 3-6 nm size, manufacturing and design must evolve, as semiconductors in this size regime are not true bulk materials. Nanocrystal research will thus guide the "ultimate" design principles of photolithographic Si circuits.

Going much smaller towards single-electron and small nanocrystal ("molecule") devices will require an absolute revolution in all aspects of the industry. Such hypothetical circuits do not appear to violate the fundamental laws of physics and would offer huge performance advances. Today, with slow scanned probe techniques, we can move single atoms and assemble molecular-level devices; however, this is not a realistic manufacturing process. At present we are far from understanding how to carry out a practical engineered "chemical self-assembly" of diverse components to create a circuit approaching the complexity and especially the reliability of modern photolithographic circuits. Just as the photolithographic Si circuit of 1997 could not be imagined in the vacuum tube era of 1937, today we can not envision the circuit of 2057. Major, unexpected inventions and truly novel scientific insights are necessary; it is not simply a question of increasing money for development. The engineering of such circuits would represent a merging of synthetic chemistry and industrial technology. Indeed, academic chemistry is evolving towards fundamental study of very large molecules, self-assembly, and macroscopic matter. Solid state physics is evolving towards new materials with molecular level structure and very large unit cells. The nanometer range is being approached from both ends: top down and bottom up.

In this context, nanocrystals are basic building blocks. Even the most elementary questions are not quantitatively understood: How does charge transfer on and off nanocrystals? What is the influence of the environment? How can the internal electron-hole kinetics and the charge transfer be controlled by surface passivation, with both inorganic and organic species? Can we entirely eliminate surface states? Can we find ways to make crystalline nanocrystals of uniform shape, size, and surface stoichiometry? Can we find ways to assemble nanocrystals of different materials and electronegativities? What role does the difference in dielectric constants play? In a material with high internal interface area, can long-term stability while carrying current approach the stability of bulk semiconductors?

Solar Cells, Displays, and New Materials

In the next century, the question of global warming due to carbon dioxide in the atmosphere is likely to become critical as increased industrialization in the third world causes accelerated warming. Direct conversion of sunlight into electricity would require large-area, inexpensive, environmentally benign solar cell designs that could be integrated into building construction. The O'Regan-Gratzel liquid junction, thin film solar cell is composed of intimate, percolating networks of liquid electrolyte and partially fused titanium dioxide 10 nm particles (O'Regan and Gratzel 1991). Photoexcited charge separation occurs at the interface between the two phases. This cell represents a new design in solar cells and is very efficient, despite the fact it is made from inexpensive, poor electrical quality oxide semiconductor nanocrystals presently manufactured in huge quantities for the ceramic and paint industries. Can a liquid junction device be as inherently stable as an all solid state device? How do we understand charge transport through such as device when the diffusing carrier interacts with both the electrolyte and titanium dioxide? This device offers a glimpse of what novel designs might be possible; can we learn from this device to optimize other applications?

Nanocrystals have higher electron affinities than organic polymers such as PPV, which are used as hole conductors in organic displays. Moreover, nanocrystal affinities are size-tunable, and when used as luminescence centers for electron-hole recombination in electroluminescent displays, the emission color is also size-tunable. Thus, in the construction of inexpensive flat panel displays, layers of neat passivated nanocrystals make good electron transport layers, and the resulting hybrid organic/nanocrystal device can emit a range of colors (Dabbousi et al. 1995; Colvin, Schlamp, and Alivisatos 1994). Can we learn to optimize transport, and can we achieve a fundamental understanding of junctions between nanocrystal layers and organic layers?

Monodisperse nanocrystals can self-assemble into three-dimensional "supercrystals" that structurally resemble naturally occurring gem opals. The optical and electrical properties of such organized nanocrystal solids could be tuned through nanocrystal size, through the inter-nanocrystal coupling as controlled by surface passivation, and by doping via material in the interstitial spaces. Can we learn to make a wide variety of such materials?

Present Scientific Understanding and Synthetic Abilities

Work over the past decade has lead to an understanding of the evolution from molecules to bulk solids, at least in outline form, for individual nanocrystals, as shown in Fig. 4.5 (Brus et al. 1995).


Fig. 4.5. Schematic size regimes for semiconductor nanocrystals.

In spectroscopic properties, there are three physical regimes: molecular, quantum dot (or nanocrystal), and polariton. Small clusters of some tens of atoms are molecules -- the bulk unit cell is not present. In the nanocrystal regime, the unit cell is present, yet the electronic states are strongly quantum-confined in three dimensions. The excited electronic states and the band gap variation with size can be quantitatively calculated with advanced theory. In the polariton regime, the interaction with the electromagnetic field becomes strong and can be described by Mie theory.

In kinetic properties, there are two limiting regimes: unimolecular kinetics in small nanocrystals, and many-body kinetics in large nanocrystals. In small nanocrystals, the electrical field of a moving charge exits the nanocrystal and interacts with the local environment: this is essentially a molecular property (Brus 1996). In the absence of a moving charge, nanocrystals are intrinsic dielectric particles. (In bulk semiconductor transport, the physics is entirely internal; the fields of moving carriers terminate on ionized donor atoms.) Also, monolayer surface passivation with a higher band gap material can eliminate most nanocrystal surface states, while still allowing efficient electron transfer on and off the particle. In the CdSe system, organometallic synthesis has achieved the ability to make nearly uniform nanocrystals (Murray, Norris, and Bawendi 1993). An equivalently good synthesis for Si nanocrystals does not exist.

The ability to do single nanocrystal luminescence spectroscopy should accelerate our efforts to rigorously understand the physics (Nirmal et al. 1996). Nanocrystals are highly polarizable molecules, some thousands of atoms in size; in the electronic properties, every atom counts. Real samples have a wide distribution of shapes and surface structures; thus, the ability to examine properties one nanocrystal at a time is quite useful. Figure 4.6 shows room temperature luminescence from single CdSe nanocrystals. The luminescence is intermittent; the nanocrystal appears to "blink" on and off on a time scale of minutes. Such blinking is incoherent from one nanocrystal to the next and was averaged out in prior hole-burning and size-selective photoluminescence ensemble studies. The blinking is slowed down by surface passivation with ZnS and has been tentatively attributed to photoionization. The single molecule method is quite powerful for such complex systems.



Fig.4.6. Scanning luminescence image of single CdSe nanocrystals (top); time trace of single nanocrystal luminescence under quasi-cw excitation (bottom) (Nirmal et al. 1996).

In the past five years, we have begun to understand the size dependence of indirect gap materials such as Si. The remarkably high photoluminescence quantum yield of Si nanocrystals and porous Si thin films can be traced to the size dependence of Auger nonradiative recombination. Small silicon nanocrystals remain indirect gap-like in all essential properties. They do not become partially direct gap-like, as was initially conjectured when the quantum properties of porous Si films were first discovered.

Just recently, systematic study of solid-solid phase transitions in nanocrystals has lead to the realization that nanocrystals in six-coordinate, dense structural phases, which are only theromodyanmically stable at high pressure, may be metastable at STP (Brus, Harkless, and Stillinger 1996). These phases have completely different optical and electrical properties than the normal sp3 hybridized phases. This opens up the possibility of an expanded inventory of possible nanocrystal materials in the future; to achieve this, novel synthetic methods are required (Brus 1997).

References

Brus, L.E. 1996. Model for carrier dynamics and photoluminescence quenching in wet and dry porous silicon thin films. Phys. Rev. B53: 4649-56.

_____. 1997. Metastable dense semiconductor phases. Science 276: 373-4.

Brus, L.E., J.A.W. Harkless, and F.H. Stillinger. 1996. Theoretical metastability of semiconductor crystallites in high-pressure phases, with application to beta-tin structure silicon. J. Am. Chem. Soc. 118: 4834-38.

Brus, L.E., P.F. Szajowski, W.L. Wilson, T.D. Harris, S. Schuppler, and P.H. Citrin. 1995. Electronic Spectroscopy and Photophysics of Si Nanocrystals: Relationship to Bulk c-Si and Porous Si. J. Am. Chem. Soc. 117: 2915-22.

Colvin, V., M. Schlamp, and P. Alivisatos. 1994. Light-emitting diodes made with cadmium selenide nanocrystals. Nature 370: 6488-6491.

Dabbousi, B., M. Bawendi, O. Onitsuka, and M. Rubner. 1995. Electroluminescence from CdSe quantum-dot/polymer composites. Appl. Phys. Lett. 66: 1316-1319.

Murray, C.B., D.J. Norris, and M.G. Bawendi. 1993. Synthesis and characterization of nearly monodisperse CdE (E=S, Se, Te) semiconductor nanocrystals. J. Am. Chem. Soc. 115: 8706-15.

Nirmal, M., B.O. Dabbousi, M.G. Bawendi, J.J. Macklin, J.K. Trautman, T.D. Harris, and L.E. Brus. Fluorescence intermittency in single cadmium selenide nanocrystals. Nature 383: 802-804.

O'Regan, B., and M. Gratzel, 1991. A low-cost, high efficiency solar cell based upon dye-sensitized colloidal TiO2 films. Nature 353: 737-740.

[Previous Section][Top of Report][Send Your Comments][WTEC Welcome Page][Next Section]

Published: January 1998; WTEC Hyper-Librarian