Inside R&D Alert

SINGLE-MOLECULE WIRE MEASURABLY CONNECTED

One of the most frustrating problems in creating nanoelectronics has been an inability to connect the molecular electronic components to anything else.Not only does this preclude making a useful electronic device,but it makes research exceedingly difficult - you don't know what your molecular electronics device is doing,how much it is doing,or - really - if it is doing anything at all.

But now we can begin to get serious about molecular electronics because a team of physicists and chemists at Arizona State University has made a good enough connection to a single- molecule wire to make accurate electrical measurements.

Two problems have held back molecular electronics,Arizona State University (ASU)chemist Devens Gust tells us. "The first has been in making robust,reproducible electrical connections to both ends of molecules.After this has been achieved,the next problem is knowing how many molecules there actually are between the electrical contacts."

Agrees ASU physicist Stuart Lindsay, "The results of studying the conductivity of molecules attached to wires have been all over the map.For example,according to published articles,DNA has been found to be everything from an insulator to a superconductor.The problem has been that no one has been able to reliably connect a single molecule."

The ASU interdisciplinary team has found a way to,in effect,solder the components together. They created a way to make through-bond electrical contacts with single molecules,then achieved reproducible measurements of the molecules' conductivity..

They attached,through chemical bonds,long,octanethiol insulator molecules to a uniform atomic layer of gold atoms.This gave them a fur-like coating of aligned molecules.Taking a few of the insulators away with a solvent,they replaced them with molecules of 1,8-octanedithiol.This is a similar molecule but it is capable of bonding with gold at both ends and acting as a molecular wire.

Then they added 2-nm gold particles to the solvent where they bonded to the free ends of the 1,8-octanedithiol molecules.This gave them a bonded metallic contact at either end of the conducting molecules.A gold-coated conducting AFM (atomic force microscope)probe was run across the surface and conductivity was measured when it made contact with the gold particles. Electrical measurements were made on more than 4000 gold particles,and they found that virtually all of the measurements fell into one of five groups (five distinct conductivity curves).

Gust tells us, "This answers the basic question of how you know when you are measuring just one molecule." The fundamental curve represents conduction by a single molecule of octanedithiol attached to the two gold contacts.If more than one molecule is bound,each additional molecule increases the current capacity by the single unit amount of current that can be carried by one molecule.When the probe encounters octanethiol insulator molecules - which cannot bond with a gold particle - a much higher electrical resistance is recorded.

The important thing: "The experimental results closely agree with theoretical quantum mechanical calculations for the conductivity of these molecules,and this gives us confidence that current theories can provide useful guidance for future experiments," Gust stresses.. "The molecule becomes a much better conductor when it is soldered into the circuit by the bonds to gold at each end.This suggests how we can wire single-molecule components into a molecular circuit board and lays some important groundwork for doing practical molecular electronics."

A lot of work has already been done on single-molecule rectification,nanotube transistors,and negative differential resistance from small collections of molecules.In none of this research have the experimenters been able to measure and predict electron transport with any confidence.Nor could they see how to reliably make a molecular wire to carry signals from one molecular circuit element to another.

The key requirement is to be able to measure the conductivity of a single molecule.You must connect a macroscopic current source and volt meter to each end of a single molecule.This means that what molecular electronics is all about is contacts.They should be ohmic so any nonlinearity in the conductivity of the wire can be correctly attributed and studied.But they should also be low in resistance to ensure that the properties measured are those of the molecule and not the contact interface.And it must all be done in a medium several orders of magnitude more insulating than the molecule.

To make good electrical contact between a molecule and a conducting substrate,you need a chemical bond,usually binding sulfur or selenium to gold or silver in what has been called a "molecular alligator clip." The ASU researchers have made such a clip..Their through-bond electrical contacts to molecules let them quantize current-voltage curves as integer multiples of one fundamental curve and thus identify single-molecule contacts.The resistance of a single octanedithiol molecule was found to be 900,plus or minus 50 megohms.Nonbonded contacts were at least four orders of magnitude more resistive,less reproducible,and had a different voltage dependence.

Three things stand out in their work,notes Washington State University chemist K.W.Hipps:1) statistical significance - where other studies rely on a handful of replications,the ASU team has more than 4000 separate measurements;2)the computed current-voltage curves agree with experiments to within a factor of six,better than anyone has achieved before;and 3)the method can be extended to other nanoparticles and to functional groups other than thiol.

EACH MOLECULE IS A SEPARATE MAGNET

One molecule,one magnet - that's about as dense as a memory storage device can get.There is a material in which each independent molecule possesses the ability to function as a magnetizable magnet below a critical temperature.Intrinsic intramolecular properties - a large spin ground state and a large and negative (easy-axis-type)magnetoanisotropy give the material this property.More familiar magnetics depend on intermolecular interactions and long-range ordering.

Realization that such a thing could be done is recent - only about five years old - and one of the main teams chasing the idea,George Christou at Indiana University (now at University of Florida) and David Hendrickson at University of California - San Diego,has now shown how to vary the basic manganese material used.They have demonstrated that carboxylate ligands can be replaced by other organic ligands to expand the crucial manganese-12 family of single molecule magnets (SMMs).

This is a major expansion.Replacing carboxylate ligands with diphenylphosphinate groups makes a species unique in displaying three distinct Jahn-Teller isomers,two of which cocrystallize in the same crystal.

Much of the SMM research effort has gone into systematic variation of the carboxylate R group and solvate molecules of crystallization and that led,in 1999,to discovery of Jahn-Teller isomerism.Differing relative orientations of the Jahn-Teller elongation axes of the eight manganese centers result in significantly different magnetic behavior for the different Jahn- Teller isomers.You get different rates of magnetization relaxation. Hendrickson explains the Christou-Hendrickson team approach to SMM research to Inside R&D: "The basic requirements for a molecule to function as an SMM are that it have a ground state with a relatively high spin (I.e.,a large number of unpaired electron spins)and appreciable magnetoanisotropy.

"There are three main areas of interest in these SMMs.First a SMM is the penultimate in small magnetic memory devices.Not only is each SMM of small size,but these molecular species can be modified much more easily than classic nanomagnetic materials. "Second,the quantum tunneling of the direction of magnetization for a nanomagnet can for the first time be established with SMMs.Up to this time it has been controversial as to whether the magnetic moment of a nanomagnet could exhibit quantum tunneling. "Third,very recent theoretical papers have been published to show that SMMs may be used for quantum computation."

The Christou-Hendrickson team's immediate research goals include the following: "A)design an SMM with a higher blocking temperature (I.e.,temperature below which a nanomagnetic can be magnetized);B)understand the mechanism of tunneling of magnetization; C)prepare SMMs with a metal other than manganese and iron;D)synthesize and study 2D arrays of SMMs on surfaces;and F)examine SMMs as a medium for quantum computation." Hendrickson cautions us: "Research into SMMs is in an early stage.It is difficult to predict what major technical problems need to be solved before we have a commercial memory device." Moving the concept on,they have now developed the first significantly altered derivative.They incorporated non-carboxylate organic ligands and got ligand-induced core distortions.They saw multiple Jahn-Teller isomerism,emphasizing the small energy differences involved.One of their complexes is magnetochemically similar to its 16-carboxylte parent.They even saw some quantum tunneling of the magnetization,showing up as steps in the hysteresis loops.They have the prototype of a major new thrust in the SMM field.

Hendrickson and Christou (along with three others)have received a grant from National Science Foundation to pursue their SSM work and are collaborating with several physics groups around the world.They would also like to establish interactions with one or more companies.Although they have focused on basic research,they are interested in expanding into applications (but have not thought in terms of patents).

TRENDS AND ANALYSES:SEARCH TOOLS GIVE WEB BACK TO SCIENCE

It may startle you to remember that the World Wide Web was not invented for amazon.com and pop-up ads,but was begun as an information-sharing project at CERN,the European particle physics lab.It has grown in importance to scientists since then,but not in scientist friendliness. The pop ups,noise,and spam have been taking over.

Watch a small band of companies and academic researchers working to grab the Web back by creating a new breed of search engine.The first search engine generation was based on classic information retrieval.A key word or phrase sends the software scurrying for matching words in documents.The more times a word appears,the higher the document ranks in results.

This is crude for scientific Web searches.Ranking by hits does not indicate how important, authoritative,or useful the pages are.Assumption was that users would look through 10 pages of results to find what they want.They don't:They only look at the first page,so ranking is everything.Yahoo and some other services tried to get around this by using human analysts to construct Web directories that retained only the most useful or authoritative Web sites. Metaengines,Web sites that queried dozens of search engines,also helped.

In the late 1990s,a second generation of tools showed up with software that performs link analysis.They not only digested the content of pages but also figured out where the pages point and what pages point at them.Google is the commercial pioneer here. Now it's time for a third generation.Search engines being designed will figure out the intent behind the query.Looking at patterns of searches and incorporating machine intelligence,the software will try to anticipate what a user really wants.Other search engines are presenting results in a file cabinet stack of categorized folders.

The ramp-up in search engine power will benefit scientists.Scirus,a specialized search engine for scientists,taps into cofounder Elsevier's proprietary journal content while searching the Web for the same key words.It is designed to filter search results to present matches only from Web pages with scientific content,based on attributes such as domain (dot.edu would be better than dot.com).It can do automatic categorization to estimate whether something is scientific content or not.It also searches content in PDF files,widely used in research.

The danger:It will yield too much proprietary subscriber-only content,tending to promote subscriptions to Elsevier journals.Scirus says it is not stacking the ranking its way and is,in fact,inviting other publishers (including the Los Alamos physics preprint server)to join.

MAKE CHEAP SOLAR CELLS WORK BETTER

If you've been following solar cell research,you know that the photonics lab of Michael Graetzel in Switzerland has been developing an interesting concept in low-cost solar cells over the past decade.Some Graetzel cells,we are told by the lab's Jessica Krueger,are already being made and sold by small companies.Their solar cells based on dye-sensitized mesoporous films of titanium dioxide are low-cost alternatives to inorganic semiconductor devices.They were getting conversion efficiencies up to 10%with such films when used in conjunction with liquid electrolytes,but not doing nearly as well when they went to a more useful solid state. Now they have found a way to considerably improve this efficiency.In 1998,they developed a dye-sensitized solar cell based on a heterojunction between the dye-sensitized titanium dioxide and the organic hole conducting material spiro-MeOTAD.The performance of this has just been considerably improved by controlling charge recombination across the interface of the heterojunction.

Interfacial charge recombination is the main loss mechanism in solid-state dye-sensitized solar cells.A number of strategies have been tried in efforts to overcome the recombination problem, including modification of surface states with organic molecules and altering the dipole field at the interface between the titanium dioxide and the hole conductor.

The latest effort in the Graetzel lab blends the hole conductor matrix with a combination of 4- tert-butylpyridine (t BP)and lithium ions.In a sandwich-type cell such as they are working with, the working electrode is comprised of a conducting SnO2:F (subscript)and a compact titanium dioxide layer deposited by spray pyrolysis.The compact titanium dioxide layer avoids a direct contact between the hole conductor and the tin oxide layer which would short circuit the cell.

They deposited a 2.5-micrometer-thick mesoporous film of titanium dioxide by screen printing onto this compact layer and dye coated the titanium dioxide electrodes.The hole conductor matrix was applied by spin coating a chlorobenzene solution of the hole conductor.Partial oxidation of spiro-MeOTAD controlled the dopant level and increased the conductivity of the hole-conducting layer.Lithium ions were added and a gold layer was thermally evaporated on top of the hole conductor film as counter electrode.

The result:They got open circuit voltages over 900 mV and short circuit currents up to 5.1 mA, yielding an overall efficiency of 2.56%at AM 1.5 illumination.This is a dramatic improvement. Unlike previous results for the electrochemical cell,the effects of lithium and tBP are not purely opposing in terms of their influence upon the charge recombination losses.In fact,both tBP and lithium inhibit the interface charge carrier recombination.

Energetics at the interface are different in the solid-state solar cell than they are in the liquid junction cell,where the electrolyte can completely propagate into the porous titanium dioxide and the space charge layer is confined to a very thin layer at the interface.There charge transport is due mainly to diffusion.

But in the solid state,a space charge layer can be formed due to photoinduced charge injection, which might enhance the recombination between injected electrons and the hole conductor cation at the heterojunction.

Lithium ions specifically adsorb to the titanium dioxide surface where they alter the space charge formed at the interface by electron injection into the titanium dioxide. "To my opinion,the device is not at all ready for a commercial application," Krueger tells us.. "The sealing issue is less problematic for the solid-state version of the Graetzel cell.Some of the physical properties of the charge transport material are still causing problems,so that an improvement in device efficiency can be obtained by further optimization of the hole conducting material." When it's ready it will still be cheaper: "Due to lower material processing costs the production of the Graetzel cell should be cheaper."

ORGANIC MEMORY MEANS CHEAP MEMORY

A lot of labs have been trying to develop an organic-based digital memory.It would be cheaper. But no one could come up with a way to make an organic device that has two stable states to encode the 0s and 1s.Now it looks as though someone has.UCLA materials chemist Yang Yang, working with Liping Ma and Jerry Liu,have developed a device that works much like flash memory,one of the more expensive silicon-based technologies,but desirable because it retains its data even when the power is turned off.If you could make an organic version,it would potentially be much cheaper because it could be far simpler to produce.

The Yang group,following a hunch by postdoctoral assistant Liping Ma,followed the compact flash device lead,in which small strips of metal act as batteries to store electronic charges.Ma sandwiched a thin metal layer between layers of conducting organic molecules.It works extremely well.

A bit of data is written onto the memory by applying a potential of 3 V between a pair of electrodes bracketing the organic-metal-organic sandwich.Applying the small voltage makes the material between the electrodes more conductive,and it stays that way even when the voltage is turned off.This high-conductivity state gives you a digital 1.The bit is read by applying a second voltage (1 V).This causes the highly conductive material to produce a rush of electrons between the electrodes,signaling that the device is in the 1 state.You get the cell back to its original low-conductivity state - a 0 - by applying a third voltage of -0.5 V.

People seeing new tricks with conductive polymers,things such as light-emitting displays and flexible circuits,are accustomed to saying, "Fine - how long will the thing keep working?" Durability has been a conducting polymer problem.

But Yang's team has run their memory device through its write-read-erase cycle some 1 million times without seeing any signs of degradation.

This amazes other conducting polymer researchers,but a bigger mystery is how the thing works at all.No one is yet clear on what causes the material in the device to change its conductivity when different voltages are applied.

Earlier devices that seemed to have the same properties did not work out as practical machines. Metal from the electrodes was migrating into the organic layers,and this was behind some of the effect.The devices worked for a short while but were difficult to reproduce and stopped working when the small metal fragments broke down.

Yang's group was particularly interested in whether similar metal reactions were behind the properties in their device.They replaced its aluminum metal layer with layers of less-reactive silver,copper,or gold,and all worked similarly to the aluminum.

However they work,organic memories could move into applications quickly.You can manufacture the devices simply by evaporating different materials through a mask in a vacuum.Yang tells Inside R&D that the major manufacturing problem will be patterning of fine metal lines on an organic compound.The "traditional photolithographic process is not applicable." This could mean that "the density might not be as high as we want to go up to."

We have learned that UCLA has granted an exclusive license to an unnamed Boston-based startup to commercialize the technology.It intends to direct it toward an ultracheap flash memory-based computer that turns on instantly without the long boot-ups ordinary computers need to reload their working memories with data.How soon? "At least a few years."