Optimising the Design of Spaceborne Time-of-Flight Mass Spectrometers with Particle Swarm Algorithms

A time-of-flight mass spectrometer (TOF-MS) separates charged particles by their mass-per-charge ratio on the basis of their transit times through an electric field. For accurate mass determination, ions must be guided from the source inlet, through the ion-optical system, and onto the detector as a beam with a controlled, narrow opening angle. This requires careful dimensioning of the electrodes and computation of the applied electric fields to control beam deflection and achieve both spatial and temporal focusing across the entire detector plane. The primary goal of this work is to develop an efficient optimisation framework for TOF-MS ion-optical designs that addresses performance trade-offs and computational challenges. For spaceborne instrumentation, high sensitivity and resolution must be balanced against size and mass constraints, making their design and optimisation particularly challenging.

The ion-optical design process involves trading off numerous interdependent parameters without an analytical solution. Mathematically, this problem can be interpreted as a derivative-free constrained optimisation problem with high dimensionality. Bieler et al. (2011) successfully used an adaptive particle swarm algorithm (APSA) to optimise voltages and electric fields for several existing ion-optical systems [1], including the Reflectron TOF (RTOF) mass spectrometer flown on the Rosetta mission of the European Space Agency. However, their approach focused on optimising voltages for predefined ion-optical geometries, without addressing the simultaneous optimisation of geometry and voltages during the early design phase. This limitation restricts the flexibility of the optimisation process and may lead to suboptimal ion focusing.

This work fills that gap by applying a particle swarm algorithm during the early design phase of a novel TOF-MS instrument. Using the SIMION® ion and electron optics simulator [2] at its base, this approach simultaneously optimises both the electrode geometries and the applied voltages, resulting in more precise control over the electric field profiles. Additionally, parallel computation techniques are implemented at thread and process levels to efficiently manage a large number of degrees of freedom, reducing computation time by allowing multiple independent particle swarms to explore the solution space concurrently. This approach provides a scalable framework for designing more precise and computationally efficient spaceborne TOF-MS instruments, contributing to the development of next-generation instruments for planetary exploration and scientific research.

 

[1] A. Bieler, K. Altwegg, L. Hofer et al., ‘Optimization of mass spectrometers using the adaptive particle swarm algorithm,’ Journal of Mass Spectrometry, vol. 46, no. 11, pp. 1143–1151, 2011, issn: 1096-9888. doi: 10.1002/jms.2001.

[2] D. A. Dahl, ‘Simion for the personal computer in reflection,’ International Journal of Mass Spectrometry, Volume 200: The state of the field as we move into a new millenium, vol. 200, no. 1, pp. 3–25, 25th Dec. 2000, issn: 1387-3806. doi: 10.1016/S1387-3806(00)00305-5.