How could one practice chemistry and carry out any chemical reactions in a way that it is absolutely safe? Yes, this is a rhetorical question, but the problem is not an entirely artificial one. We surmised that the solution to the problem is to carry out reaction at such a small scale, that the amount of the products cannot harm anybody. Hypothetically, when we react one molecule of A with one molecule of B to yield one molecule of product C, can the single C molecule harm us? The answer is a resounding NO! C would have to be a super-efficient catalyst! Sadly, we do not know ANY catalyst that effective. It appears that if we carry out reactions at a scale of few hundred molecules, even the very dangerous products could be safe. In a way, this is similar to natural radioactivity background: Radioactive species are everywhere around us, but at the normal levels are not harmful. Even the most toxic compound on earth, botulinum toxin, needs roughly 8x1011 molecules to kill a 100kg professor!
Well, then. We decided to develop a method that would allow running reactions on a scale of 500-1000 molecules. This could be easily done using so-called high-dilution method when extremely small amounts of reagents are reacted in large volume (i.e., at very low concentration). The problem here is that reaction times in high-dilution methods are extremely long (proportionally to the dilution). Second approach is to use very small reaction vessels. Such super-small reaction vessels are difficult to fabricate, and they are very expensive. This problem has been partly addressed by nano/microfluidic reactors, in which tiny droplets of immiscible liquids were carried through the channels (see the figure below, Left), mixed, and their solutes reacted (Nature 2006, 442, 368-373. doi:10.1038/nature05058). Next figure actually shows two water droplets separated in oil-like liquid can coalesce in a nano/microfabricated channel (Left), droplets comprising each two micro-droplets of different dyes, and finally a microfluidic super-reactor chip. However, in the microfluidic setups, the flow and droplets must be synchronized to meet, and the reactors with products collected in a receiver reservoir, where the amounts of products accumulates to exceed the safe amount of products. Also, there is the issue of nano/microfluidic chip fabrication, pumps to bring in the liquids, etc. All in all, it may be possible, but faaaar from easy-to-do. Thus, we felt that the nano/microfluidic approach does not provide the solution to the problem we wanted to explore. After all, we are organic chemists, and at the end of the day we want reactions and products!
Our approach, on the other hand, is simple. The following figure shows the principle of the nanofiber reactors. The nanofibers are doped with reagents, but are not hollow (i.e., the reagents do not flow through the fiber). The red object at the nanofiber junction is the idealized representation of the attoliter reactor (a). The SEM images show the fibers overlaid (b) before they are softened. After the softening (c and d) the fibers fuse and flatten a little. This is when the reagent molecules start to migrate through the atto-space of the junction (atto-reactor).
When the nanofibers are softened, the reagent molecules diffuse and percolate through the interstitial space (between the polymer chains), meet and react. Because seeing is believing, we show a reaction between two non-fluorescent starting materials which form a fluorescent product in the attoreactor. The figure bellow shows three microscope images: Brightfield image (A) where you cannot see the fluorescence; Darkfield image with only UV-light excitation, which shows only the fluorescence from the reactor (turquoise-colored dot in the center)(D), and the combined Brightfield + UV excitation, which shows that the fluorescence indeed comes from the attoreactor and nowhere else (see the figure below).
Higher density fiber mats can be fabricated if a larger number of reactors is required (Right) and products harvested for traditional analysis (e.g., MS, NMR). Furthermore, we can also print larger (tens-of-micrometers) fiber-like "lines", that can be probed directly by MALDI-TOF HR-MS from the single reactor. Again, brightfield image (A), UV excitation (D), mixed excitation (C), and black-white image with contrast (C). In this case, the reaction is a Diels Alder addition of acetylene to aryl-substituted cyclopentadienone to form substituted terphenylene derivative (see the figure below).
Needless to say, this method is general and is not limited to fluorescent products (their formation is just so nicely observed by fluorescence confocal microscope.
We are now actively pursuing this avenue of research to explore what kind of reactions can be carried out this way and study the activation and kinetic parameters of these new reactions.
For more information, see Nat. Chem. 2009, 1, 80-86. Angew. Chem. Int. Ed. 2012, 51, 2345-2348. This research has also been highlighted in The New York Times and Nature.