Adsorption involves molecules colliding at the surface of a solid and losing their incidence energy by traversing a dynamical pathway to equilibrium. The interactions responsible for energy loss generally include both chemical bond formation (chemisorption) and nonbonding interactions (physisorption). In this work, we present experiments that revealed a quantitative energy landscape and the microscopic pathways underlying a molecule’s equilibration with a surface in a prototypical system: CO adsorption on Au(111). Although the minimum energy state was physisorbed, initial capture of the gas-phase molecule, dosed with an energetic molecular beam, was into a metastable chemisorption state. Subsequent thermal decay of the chemisorbed state led molecules to the physisorption minimum. We found, through detailed balance, that thermal adsorption into both binding states was important at all temperatures.

Most experimental methods for studying the kinetics of surface reactions – for example, temperature programmed desorption (TPD), molecular beam relaxation spectrometry (MBRS) and velocity‐resolved kinetics (VRK) – employ detection schemes that require thermal desorption. However, many adsorbates – for example reaction intermediates – never leave the surface under reaction conditions. In this paper, we present a new method to measure adsorbate concentrations on catalytic surfaces and demonstrate its utility for studying thermal desorption kinetics. After a short‐pulsed molecular beam deposits CO or NH3 on Pt (111), the surface is irradiated with an ultrashort laser pulse that induces desorption. Another tightly focused ultrashort laser pulse ionizes the gas‐phase molecules by a non‐resonant multiphoton process and the ions are detected. This two‐laser signal is then recorded as a function of time after the dosing molecular beam pulse and decays exponentially. First‐order thermal desorption rate constants are obtained over a range of temperatures and found to be in good agreement with past reports. Ion detection is done mass selectively with ion‐imaging, dispersing the gas phase molecules by their velocities. Since laser‐induced desorption (LID) produces hyperthermal gas phase molecules, they can be detected with little or no background. This approach is highly surface‐specific and exhibits sensitivity below 10−4 ML coverage. Because the signals are linearly proportional to adsorbate concentration, the method can be employed at lower temperatures than VRK, whose signal is proportional to reaction rate.

We describe here the application of an inexpensive event-based/neuromorphic camera in an ion imaging experiment operated at 1 kHz detection rate to study real-time velocity-resolved kinetics of thermal desorption. Such measurements involve a single gas pulse to initiate a time-dependent desorption process and a high repetition rate laser, where each pulse of the laser is used to produce an ion image. The sequence of ion images allows the time dependence of the desorption flux to be followed in real time. In previous work where a conventional framing camera was used, the large number of megapixel-sized images required data transfer and storage rates of up to 16 GB/s. This necessitated a large onboard memory that was quickly filled and limited continuous measurement to only a few seconds. Read-out of the memory became the bottleneck to the rate of data acquisition. We show here that since most pixels in each ion image contain no data, the data rate can be dramatically reduced by using an event-based/neuromorphic camera. The data stream is thus reduced to the intensity and location information on the pixels that are lit up by each ion event together with a time-stamp indicating the arrival time of an ion at the detector. This dramatically increases the duty cycle of the method and provides insights for the execution of other high rep-rate ion imaging experiments.