Chemical reactions occur in our everyday life. Our basic knowledge of chemistry tells us that temperature, concentration, surface area, and catalysts can speed up or slow down chemical reactions. For example, a carton of milk would spoil much faster if kept out of a refrigerator. In the ultracold regime, typically less than several millionths of a kelvin above the absolute zero (-459 °F or -273 °C), the description of chemical reactions is quite different. Scattering particles are represented as waves, and depending on their quantum states, the reaction rate may not depend on their temperature at all. More surprisingly, chemical reactions in the ultracold regime can be controlled using resonances. You may have seen numerous fallen bridges and demolished buildings by mechanical resonances, showing how powerful such effects can be. Similarly, the effect of collisional resonances can yield dramatic changes in the rate of chemical reactions between ultracold molecules, in degrees much greater than what we would normally expect in our daily lives. 

Macroscopically, chemical reactions seem like a magical process in which reactant particles jump into the state of the final product. However, in a microscopic view, there exists an intermediate step where the reactant particles collide and form a collisional complex which can be represented as a single bound state. Feshbach resonance, a type of magnetically tuned collisional resonance, occurs when the energy of the collisional complex is equal to the energy of the two colliding particles. It can immensely change interactions between the reactant particles from weak to strong and/or repulsive to attractive. Since its first observation in sodium atoms (Na) near absolute zero temperature in 1998, Feshbach resonance has become an essential tool in ultracold atomic experiments. As a result of collective efforts to create and manipulate ultracold molecules, Feshbach resonances were recently observed in the two systems of ultracold atom and properly bound molecule mixtures (Na + NaK in 2018 and Na + NaLi in 2021). It has been shown that these atom-molecule collisions with low reactivity can have a much faster chemical reaction through Feshbach resonance. The chemical reaction rate in Na + NaLi has been reported to be enhanced by more than a factor of 10 at a certain magnetic field.

While its effect in atom-molecule reactions is observed, there has been an open question of whether the Feshbach resonance could also be applied to control the rate of molecule-molecule reactions. Many theorists rather predicted negatively, postulating that collisional resonances may not be observed between tightly bound molecules, mainly because of two reasons. First, the collisional resonances may occur at multiple magnetic field values within a small range. Consequently, the resonances are likely to be unresolvable. Second, resonances can be more pronounced when a longer time is spent in an intermediate collisional complex state during a chemical reaction process. However, there exist many mechanisms which can destroy the intermediate collisional complexes. Such mechanisms are likely to broaden the resonant features making it more difficult to identify resonance from the background signal.

Experimentalists always seek empirical evidence of the theory. As such, we searched for Feshbach resonances in collisions between NaLi molecules. With various cooling techniques to prepare ultracold gases of Na and Li atoms and using these atoms as building blocks, we assembled a gas of magnetic NaLi molecules in their ground state at 1.8 micro-Kelvin (0.0000018 K). We searched for collisional resonances by sweeping a large range of magnetic field values and identified a single Feshbach resonance centered at 334.92 Gauss (G). The resonance that we observed can increase the reaction rate by more than a factor of 100 while being extremely narrow (~25 milli-Gauss). The observation of a single Feshbach resonance that is not broadened at all contradicts the current molecular collision theories. 

Our observation of the unexpected resonance calls for further studies on the properties of collisional complexes. While the resonance raises many questions, the observation is particularly interesting in two regards. First, the pronounced Feshbach resonance provides strong evidence for a stable, long-lived collision complex, which is unexpected in a molecular system of high reactivity such as NaLi. Is a long-lived state coexisting with unstable states a common feature of molecular systems? Second, the resonance is observed at the magnetic field where another internal quantum state of a NaLi molecule becomes energetically close to the reactant quantum state. It is possible that there are many more resonances existing at other magnetic field values but are only detected by a mechanism that involves another state that has the same energy. Our result suggests a new type of resonance that could be ubiquitous in ultracold molecular physics, offering a powerful new mechanism for controlling ultracold chemical reactions.