The aim of this project is the investigation of a wafer-scale, single step, parallel process for electrical interfacing to chemically synthesised nano-scale components, by exploiting thermodynamically driven self-assembly techniques on patterned silicon substrates. The proposed system offers a natural bolt-on technology for the fabrication of hybrid CMOS/intra-molecular electronic circuits. The project will attempt to address the industrial requirements of future nano-scale IC manufacturing, namely, perfectly reproducible molecular scale minimum feature sizes that can be produced both quickly and cheaply via an alternative "bottom up" approach. The aim of this project is the investigation of a wafer-scale, single step, parallel process for electrical interfacing to chemically synthesised nano-scale components, by exploiting thermodynamically driven self-assembly techniques on patterned silicon substrates. The proposed system offers a natural bolt-on technology for the fabrication of hybrid CMOS/intra-molecular electronic circuits. The project will attempt to address the industrial requirements of future nano-scale IC manufacturing, namely, perfectly reproducible molecular scale minimum feature sizes that can be produced both quickly and cheaply via an alternative "bottom up" approach.
(1) The development of selective etching techniques on Silicon to form nano-gaps between macroscopic electrodes.
(2) The synthesis and electrical characterisation of both conjugated organic molecules and nano-crystals with integral, tailored end groups that anchor to the semiconductor electrodes such that the nano-scale components can self-assemble and bridge between the selectively etched nano-gaps.
(3) The combination of nano-scale elements and selectively etched nano-gaps to form functional molecular-scale elements that can directly interface with conventional CMOS devices.
(4) The demonstration of a hybrid MOSFET/molecular electronic memory device.
DESCRIPTION OF WORK
Present lithographically defined nano-scale devices are exponentially sensitive to atomic layer fluctuations, resulting in device specific variations that are unacceptable for manufacture. Molecular nano-electronics has attracted considerable attention because it represents the ultimate in dimensionally scaled systems. Furthermore, molar quantities of identical devices are routinely available via chemical synthesis.
An additional incentive is the potential to utilise thermodynamically driven self-assembly o the nano-scale components. This approach eliminates any critical dimension control problems whilst forming ultra dense IC arrays. Such a scheme however, requires nano-scale electrodes and interconnects. It is this contacting problem, which has so far defeated the exploitation of molecular nano-electronics, that will be addressed in this project. The contacting scheme will incorporate a large built-in engineering redundancy, allowing a certain degree of fault tolerance to be exploited in the of circuit architectures.
The proposed system allows direct integration and interconnection of chemically synthesised functional nano-scale components with conventional CMOS devices, offering a bolt-on technology for the fabrication of hybrid MOSFET/intra-molecular circuits.
1. Development of selectively-etched nanogaps on silicon wafers. The SiO2 oxide in a conventional MOSFET structure is etched away, leaving a gap between highly doped silicon electrodes. The size of the gap (~5 nm) can be tailored to the length of the molecule or diameter of the nanocrystal. We have observed reproducible negative differential resistance (NDR) when CdSe nanocrystals are incorporated into such gaps. We have developed a technique of forming a silicide on the nanogap surfaces, and have shown that molecules with thiol end groups will self-assemble on such surfaces. We have also used the selectively etched gap as a template for the formation of a gap between metal electrodes. These two techniques offer the prospect of easy attachment of molecules to a silicon device.
2. Fabrication of gold nanogaps using shadow evaporation. The edge of one layer of gold shadows the region next to it from gold atoms evaporated at an angle, forming a gap of a controllable size, usually between 2 and 6 nm. CdSe nanocrystals with dithiol linker molecules, or short conjugated molecules with thiol end groups, were self-assembled in this gap. The I-V characteristics of some of these samples showed a staircase structure (at 4K), indicative of the quantised energy levels or of Coulomb blockade of single-electron tunnelling.
3. Synthesis of functional conjugated molecules with thiol end groups. We have developed the synthesis of new “long-chain” organic molecules with lengths of between 3 and 5 nm that can span nanogaps. These molecules have been designed to make them soluble in most organic solvents so that they are easy to deposit onto the surfaces of supports and into nanogaps. The electronic properties of the “long-chain” molecules have been altered by altering the organic functional groups along the chains, in one of two ways:
(i) by introducing donor-acceptor groups on either side of the molecules which introduces a polarity perpendicular to the chain direction, or
(ii) by using steric control to introduce a twist into the molecules along the chain, and by altering the angle of the twist altering the electron flow along the chains.
In the latter stages of the project, transition-metal-containing units have been introduced into the chains at regular intervals. The introduction of the metals means that it is easier to grow longer chains, and that there is potentially greater control over the electronic properties of the chains because the redox properties of the metals can be tailored to enhance or decrease the electron flow.
4. Physical and electrical characterisation on surfaces using a low-temperature UHV STM. Selected molecules and nanocrystals have been investigated by scanning tunnelling spectroscopy with regard to their electrical properties and their suitability for an application in nanoelectronic devices. I-V characteristics collected at different molecules or nanocrystals have been measured and interpreted. The I-V characteristics on semiconductor substrates are found to be strongly dependent on whether the substrate is doped heavily p- or n-type.
5. Modelling of energy levels of molecules and nanocrystals with anchors. We have identified from electronic structure calculations molecular features required to promote an NDR signal in molecular junctions; according to these calculations, resonant tunnelling processes can be generated in conjugated wires by introducing a twist or a saturated spacer along the molecular backbone. This is further supported by sophisticated I-V simulations, thus pointing to the fact that simple molecular modelling can prove very useful to design wires with specific functionalities and electrical characteristics for use in molecular junctions. The calculations have also rationalized a large number of recent experimental data, thus providing a unified picture of the NDR behaviour in molecular wires. We have also developed a new modelling program (TranSIESTA) to calculate transport through molecules attached to realistic substrates. Arrays or individual molecules can be handled. This has produced important insights into the causes of the switching in molecules consisting of chains of one or more benzene rings.
6. Modelling of TSRAM architecture with nanoscale elements. Computer simulations have been used to determine the performance required for molecular electronic devices to be useful in computer circuits. One of the possible applications of molecular electronics in combination with silicon technology is for the refresh of DRAM cells. The simulations show that no molecules with resonant electrical properties published so far in the literature have suitable properties for this particular application. Thus the aim should be to find molecules, which show NDR at lower applied voltages and lower current levels than the molecules published so far.
Funding SchemeCSC - Cost-sharing contracts
412 96 Goeteborg
2800 Kgs. Lyngby