HomeAbout Reactive ReportsRecent reports (archives)HumorUseful linksSearch
Interview by David Bradley ISSUE #59
October 2006
Amilra Prasanna "AP" de Silva

Amilra Prasanna "AP" de Silva was born on April 29, 1952, in Colombo, Sri Lanka, and obtained his PhD in Organic Photochemistry at Queen's University of Belfast, in 1980, having graduated from the University of Sri Lanka in 1975. He has been Professor of Chemistry, Queen's University of Belfast, since 1997, and his fascinating research into logical molecules has been featured on several occasions in Reactive Reports. Here we learn a little more about the man behind those glowing reports.

Questions

What inspired you to study chemistry and how did you end up in the field of molecular logic that you have pioneered?

Amilra Prasanna "AP" de SilvaI was lucky to have a brilliant chemistry teacher, Errol Fernando, for A/L at school in Sri Lanka. I wasn't the only one to forego the usual first-choice of engineering/medical school and to dive into chemistry which underpins so much of our lives and environments. I was equally lucky to publish, with Nimal Gunaratne and Colin McCoy, the first experimental cases of molecular logic in the primary literature. This grew from molecular sensors and switches, for which we had established a general design principle based on fluorescence and PET (photoinduced electron transfer) in the early 1980s. However this step required a clear impetus, which at least partly came from Satish Namasivayam (now the Director of the Arthur C Clarke Centre for Advanced Technologies, Sri Lanka). He taught me how to build a digital clock, among other marvels of digital electronic gates. These hands-on experiences led me to some computer programming courses at the University of Colombo. From then on, the hard/software of computing ceased to be foreign to this chemist.

What chemistry are you working on right now?

One line is to integrate sensor technology with logic capability within single molecules. We published the first of such 'lab-on-a-molecule' schemes earlier this year (J Am Chem Soc, 2006, 128, 4950; http://dx.doi.org/10.1021/ja058295+). The idea here is that the workload of medical doctors can be reduced by flagging up possible positives for a particular disease by checking for logical combinations of elevated or depressed levels of several parameters in blood for instance. Another line is to address the oft-raised question, 'Can molecular computer elements do anything useful that is not possible with silicon?'.

Tell us briefly about your ultimate aim in using small molecules for logical functions?

The ultimate aim of most scientific efforts must surely be to become useful to others. I share this belief. It is not enough to be satisfied with the demonstration of molecular logic operations, even though it was considered to be nearly-impossible by many serious observers for a long time.

What are the advantages of your approach over that being taken by the supramolecular chemists?

Actually, I would consider myself a supramolecular chemist in several respects. So perhaps there is no comparison to be made. For instance, our approach to molecular logic relies on chemical inputs to molecular devices, which translates to a chemical species binding reversibly to a suitable receptor within a supermolecule. Of course, we also require photochemical phenomena such as fluorescence quenching to complete the task. However, it is true that there are imaginative approaches involving supramolecular phenomena alone. Enzyme cascades come to mind, for example.

Your compounds have already been incorporated into sensor technology, how long do you think it will be before we see them as components of a molecular computer?

You raise a crucial point here. Our collaboration with Roche/OSMETECH led to the market-leading OPTI point-of-care blood electrolyte analyzer (see it at www.osmetech.com) with sales of $40m for the sensor component alone so far. This demonstrated that a straightforward design such as the fluorescent PET sensor/switch system could succeed in the marketplace with sufficient commercial will. Perhaps more importantly, it showed that a bit of sensor science could be socially, and even personally, relevant. What can be of more social and personal relevance than a loved one being treated at an intensive care unit or in an ambulance?

Molecular logic gates have not made this transition as yet. However, I am delighted to say that we recently have a demonstrator, which now needs the driver of commercial will. But this is for others, and not for me to provide. I must also quickly add that this is not the direct answer to the question. Our demonstrator is for a wide-scope application of molecular computer elements (rather than molecular computers as complex as those available with semiconductor technology). Nevertheless, this involves molecular computing at an elementary level which is sufficient for the application.

What obstacles remain before that will be achieved?

Silicon-based electronic logic gates are true components of computers since the latter are made up of connected arrays of the former. On the other hand, all currently available molecular-scale logic gates are less easy to connect since their inputs and outputs are not perfectly compatible. In fact, this incompatibility has prevented feedback so that molecular-scale logic behavior was demonstrable! Nevertheless, imaginative schemes have allowed the functional integration of multiple (about 20) logic gates within single molecules. So, small arrays of molecular-scale logic gates are all we have at the present time. However, this is not an excuse to allow molecular logic to continue without proving its usefulness. The elementary molecular computing provided by molecular-scale logic gates is good enough to provide at least one wide-scope application.

What will such a device be able to do that is not possible with silicon?

Silicon-based computing has limitations to the spaces in which they can operate, especially if remote communication (e.g., with radio waves) is involved. The antennas and inductors required in such cases are quite hard to miniaturize. For instance, radiofrequency identification (RFID) chips are currently limited to 1mm x 1mm dimensions. Then, the identification of objects smaller than 1mm in dimension in large populations is currently a difficult problem. The seriousness of the problem is clear when we note that biological cells and the polymer beads in combinatorial chemistry fit this description.

We have now invented molecular computational identification (MCID) tags which are molecular in size (ca. 1 nm) and can be covalently attached to polymer beads (Nature Materials, http://reactivereports.com/58/58_1.html). Their identity can be read remotely by their logic type (singly or in a parallel array), input chemical combination, switching threshold besides their excitation and emission color. A simple 'wash & watch' procedure allows parallel identification with a fluorescence microscope. So here is a molecular-scale logic device performing a useful application that is barred to silicon technology. Molecular computing may be small-scale in terms of its complexity, but it is powerful enough to go and serve on small objects right here, right now.

Speaking of computers, what role does computational chemistry itself play in your research?

Computational chemistry continues to help in the design of fluorescent PET systems. The judgment of the viability of electron transfer requires the evaluation of redox potentials which can be correlated with frontier orbital energies. The latter are directly obtainable by calculation according to programs of increasing sophistication. However, the experimental measurement of redox potentials is a straightforward alternative in many laboratories. Perhaps more importantly, the electron density distribution of the occupied frontier orbitals is quite informative for PET designers.

How is the internet changing the way experimental chemists, such as yourself work?

The internet has brought most of the published work of other laboratories to our desk or laptop, so that no effort need be wasted in re-inventing wheels. This has freed up time to contemplate and mull over ideas. The easy communication with friends and co-workers around the world is an important aspect too.

How important do you think systems such as e-Molecules (http://www.emolecules.com/) (formerly Chmoogle), PubChem (http://pubchem.ncbi.nlm.nih.gov/), Chemspy (http://www.chemspy.com), and other online molecular search tools are?

These services have sped up the search for preparative procedures to no end. Hitting on a good synthetic procedure reduces the pain of a project on molecular switch/sensor/logic so that the evaluation phase can be reached in reasonable time and the true benefits can be reaped.

What are your views on the growing rift in scientific publishing as journal subscription costs?

The fact that working scientists write, referee, edit, and read scientific papers makes one wonder what publishers (especially the for-profit companies) do to justify the rising prices. Open access is indeed an ideal in many ways, but has become realistic because of electronic publishing. Nevertheless, the status quo will not go away tomorrow.

Where are you publishing currently and will that change as open access journals become more prominent?

You hit the right button there - 'prominent'. The current OA offerings for chemical science don't have the impact of the conventional journals from the major publishing houses for getting a paper read and cited. I cannot forget that a citation is tangible evidence of the existence of a community in a given field, something most human enterprises strive for. Therefore we aim for the high-impact journals when publishing and we have been fortunate in recent years to have work appearing in Nature Materials, JACS, and Chemical Communications. If OA journals can be equally prominent in getting exposure for our work, then there is no reason not to go there.

What has been your most successful research paper in terms of citations and follow-on work by others?

This has to be Dr. Nimal Gunaratne's, Dr. Colin McCoy's, and my 1993 Nature paper (1993, 364, 42; http://dx.doi.org/10.1038/364042a0), which introduced the field of molecular-scale logic and computation as an experimentally robust science. Though its citations are not huge (about 250 so far), it certainly launched the field to which a steadily-increasing number of imaginative scientists have been drawn. The achievements of this community include the generalization of molecular logic and computation to designs other than PET, the co-option of biomolecules such as oligonucleotides and polypeptides, and the realization of small-scale logic integration, reconfigurability, numeracy, and gaming capability.

How can we ensure that chemistry remains at the top of the scientific tree?

This is an important issue. We must strive to demonstrate that chemistry is a creative science in its own right. We must also show, in engaging ways, that chemistry is around and within all of us. No chemistry, no us. Perhaps some of chemistry's commercial success has been seen as a weakness. The policies and practices of large chemical corporations run by accountants and managers have led to socially unacceptable situations. Then who takes the rap in the media? Not the accountants and managers but chemistry itself. Of course, the former are quick to claim credit when the opportunity arises. The media must take some responsibility for allowing itself to be manipulated in this way.

http://www.ch.qub.ac.uk/staff/desilva/