Research
Molecular ions · Reactive intermediates · Low temperatures
I develop instruments to prepare, manipulate and measure molecular species. From ion deposition on surfaces to the detection of reactive intermediates and low-temperature reactions, my research addresses both the phenomena under study and the experimental conditions needed to measure them accurately.
My work combines instrument design, experimental protocol development and data analysis. Distinguishing the physical phenomenon from the response of the measuring apparatus is central to my approach.
Preparing and depositing molecular ions
At the Max Planck Institute for Solid State Research (MPI-FKF), we are developing an electrospray ion beam deposition (ESIBD) platform. The aim is to prepare molecular ions in defined solvation or complexation states, then deposit them onto a surface for single-molecule imaging by scanning probe microscopy. This will allow us to investigate how solvent molecules and metal ions affect their structure.
I contribute to the instrumental development of the setup, particularly the electrospray source, ion optics and equipment interfacing. The platform is being assembled; control over the chemical state of ions before deposition is a development goal, not yet a demonstrated capability of the new instrument.
Instrument principle and ongoing developments
The instrument is designed to transfer ions generated from solution onto a surface under ultra-high vacuum, using mass-to-charge selection and soft landing. Two developments are intended to control their chemical environment: a quadrupole trap chamber for interactions with volatile molecules, and a nano-electrospray source operating under controlled gas-phase conditions.
These developments are described in my IMSC 2026 report. For the scientific context of the project, see Kelvin Anggara’s group.
Measuring reactive intermediates
During my postdoc at KTH in Stockholm, I developed measurement tools to study dilute, highly reactive species, including peroxy radicals and Criegee intermediates. How can we measure their reactivity without confusing gas-phase reactions with losses to the walls of the apparatus? This question connects my work on mass spectrometric detection, flow reactors and data processing.
I contributed to the development and optimisation of a proton-transfer-reaction time-of-flight mass spectrometer (PTR-TOF-MS) for direct measurement of peroxy radicals. By monitoring methylperoxy and isopropylperoxy radicals in contact with different materials, we showed that their losses to stainless steel and aluminium could dominate gas-phase reactions under the conditions studied. Published in JACS Au, these results identify the role of surface redox reactions: the reactor material can change the chemistry we intend to measure.
Reactors, detection and data processing
I developed a continuously operating tubular reactor with no moving parts for quantitative measurements at room temperature and atmospheric pressure. A variant uses a buffer arm to vary the residence time through the flow rate. These devices allow us to study gas-phase reactivity and losses at interfaces.
In the surface study, mass spectra were acquired every second. Radical signal assignments were checked by adding nitric oxide (NO), which consumes the radicals. Monitoring the radicals and their products allowed us to relate their disappearance to reactions on the materials.
To process PTR-TOF-MS data, I wrote MassSpec.jl in Julia. This open-source software handles data sets on the order of 1 GB per hour of experiment. The software tools are collected on the Resources page.
I also participated in measurement campaigns on gaseous and particulate emissions from brake wear. We showed that oxidation of the gaseous emissions could form secondary particles, in addition to those emitted directly by wear (Scientific Reports, 2024). The emitted volatile organic compounds and their role in ultrafine particle nucleation are the subject of an article in ES&T Air, 2025.
Studying reactions at low temperatures
At the Institute of Physics of Rennes, my PhD work focused on CRESUSOL, an instrument for studying neutral–neutral reactions and cluster formation under conditions relevant to astrochemistry. It combines a uniform supersonic flow, produced by a de Laval nozzle, with photoionisation and mass spectrometric detection.
I worked on developing this instrument and calculating the nozzle profiles that determine the flow conditions. The instrument enabled observations of formic acid dimer formation and diacetylene formation at temperatures below 100 K (Review of Scientific Instruments, 2021).
CRESUSOL and de Laval nozzle design
CRESUSOL uses vacuum-ultraviolet synchrotron radiation to photoionise species from the flow. Photoelectron–photoion coincidence detection, combined with ion time-of-flight measurement, allows reaction and clustering products to be studied at low temperatures.
I developed a de Laval nozzle design program, published alongside its method in Physics of Fluids, 2022. The calculation combines the method of characteristics with a boundary-layer correction to produce flows with uniform velocity, temperature and density. The program is publicly available.
Distinguishing chemical decay from a density drop
Building on this work, I am critically re-examining kinetic measurements in uniform supersonic flows (CRESU). A decrease in a reactant signal does not by itself demonstrate chemical consumption: a density drop associated with gas mixing can produce a decay mistakenly attributed to reaction. My reanalysis of CRESU-SIS data identifies this problem and challenges the rate coefficients inferred from those measurements (commentary, 2023).
Observations, interpretation and scope
In the CRESU-SIS data analysed, the initial ion signal decreases as the neutral reactant concentration increases, whereas it should remain constant under the kinetic interpretation used. A drop in flow density provides an alternative explanation for this observation and the measured decays.
This critique extends beyond CRESU-SIS: the flow conditions during mixing must be checked before assigning signal decays to chemistry. It does not amount to recalculating all published rate coefficients using a single correction; the conditions and signals of each experiment need to be examined. I developed this reappraisal in my CPLT 2026 presentation.