Frequency chasing of individual megadalton ions in an Orbitrap analyser improves precision of analysis in single-molecule mass spectrometry

The study addresses how individual megadalton-scale ions behave in an Orbitrap mass analyser during charge-detection mass spectrometry (CDMS), with the goal of improving sensitivity, resolving power, and precision for large, heterogeneous biomolecular assemblies measured under native conditions. Traditional mass analysers lose resolving power at high m/z and suffer from ion loss due to collisions, metastable decay, and space-charge effects. Single-particle approaches like CDMS and nanoelectromechanical systems MS circumvent charge-state assignment challenges but have limitations. Building on recent demonstrations of Orbitrap-based CDMS, the authors investigate why and how very large ions show unexpected stability during long transient recordings, quantify collision effects, and develop a frequency-chasing strategy to correct frequency drifts from desolvation and rare charge-loss events. The overarching aim is to enable longer transients and improved mass/charge precision for applications including complex virus particles and gene delivery vectors.

Prior work established native MS for large assemblies and single-particle techniques such as CDMS and nanoelectromechanical systems MS for highly heterogeneous samples up to 100 MDa. Two groups demonstrated Orbitrap-based CDMS by exploiting the linear relation between ion charge and image current/peak height, enabling single-ion analysis on Orbitrap instruments. In contrast, ensemble knowledge of Orbitrap ion behaviour largely stems from smaller/denatured ions, for which single collisions at low m/z often cause fragmentation and ion loss, necessitating ultra-high vacuum. Frequency drift mitigation methods for FT-ICR (e.g., frequency chasing, quadrature heterodyning) exist but operate under different field and energy regimes and are not directly transferrable to Orbitrap analyses of high-mass native ions. Previous reports also describe limits to native MS resolution due to overlapping isotopes and adducts, and theory/observations on Orbitrap dynamics, space charge, and enhanced Fourier transform (eFT) processing.

Samples: Megadalton particles included hepatitis B virus (HBV) capsids (T=3 and T=4, ~3 and ~4 MDa), Flock House virus (FHV, ~9.4 MDa), engineered IgG1-RGY oligomers (mono- to hexamers: ~150–900 kDa), and the ~3 MDa AaLS-neg nanocage. Samples were buffer-exchanged into 150 mM ammonium acetate (pH 7.5). HBV capsids were formed by diluting dimers into ammonium acetate. Instrumentation: Q Exactive UHMR Orbitrap mass spectrometer operated for high m/z transmission with native MS source/interface settings. Neutral gas in the collision cell was xenon or nitrogen; collisional activation in the HCD cell aided desolvation. Ion transmission was attenuated to achieve predominantly single-ion events. Transient durations ranged from 128 ms up to 4,096 ms. Data acquisition and processing: Transients were saved and processed offline. Time-domain signals were four-times zero-padded and converted to magnitude-mode FFT spectra using instrument calibration for m/z conversion. Peak centroids were determined via three-point least-squares parabolic fits. To study temporal behaviour, transients were segmented into overlapping windows to perform segmented Fourier transforms. Ions were traced across segments by nearest-neighbour centroid matching in m/z, applying filters on intensity and m/z standard deviations to avoid misassignment. Frequency chasing: Segment-wise centroid positions were used to track frequency drift for individual ions. Using conventional charge assignments (via m/z binning), frequency drifts were converted into neutral mass losses (Da) per segment and cumulatively. Peak-splitting artefacts from eFT processing (due to frequency drift across FT bins) were mitigated by averaging segment intensities at near-constant local frequency, enabling drift-corrected reconstruction of single-ion intensities and m/z. Modeling ion behaviour: The observed mass (frequency) drift was modeled as a combination of a linear term (constant-rate desolvation from collisions during Orbitrap detection, proportional to travelled distance/pressure) and an exponential term (initial activation during injection from the C-trap into the Orbitrap, decaying over time). Parameters scaled linearly with measured pressures. Ion path length, collision cross-section (CCS), mean free path, and expected collision counts were estimated using trap geometry, oscillation frequencies, ideal gas law relations, and CCS approximations based on particle radii from structural models. Optimization experiments: Effects of reducing Orbitrap pressure, increasing initial desolvation/activation, shortening transient times, and applying frequency-chasing drift correction were evaluated for their impact on the fraction of usable single-ion events (reduction of split peaks), mass/charge resolution, and signal-to-noise. Charge calibration: Under acquisition conditions, magnitude FT single-ion intensities were converted to charge by a proportionality factor (175).

- Stability plateau for megadalton ions: Ions >1 MDa display remarkable stability in the Orbitrap, surviving multisecond transients while traversing multi-kilometre paths. For HBV (~4 MDa) and FHV (~9.4 MDa), mass resolution increased linearly with transient length from ~3,000 at 128 ms to ~100,000 at 4,096 ms, enabling sub-10 ppm precision (e.g., <40–90 Da at multi-MDa masses). - Collision tolerance: Despite an estimated ~30 collisions during a ~4 s transient (xenon background), high-mass ions persist, with signal-to-noise scaling as the square root of transient duration as expected for stable detection. In contrast to small/denatured ions, collisions predominantly cause gradual solvent loss rather than fragmentation. - Size-dependent survival: In IgG1-RGY oligomers (150–900 kDa) and AaLS-neg (~3 MDa), survival over ~1 s improved markedly with increasing mass and lower charge states, reflecting favourable scaling of centre-of-mass collision energy and surface energy deposition with mass/CCS. Larger particles exhibited minimal decay despite multiple collisions. - Three behaviour classes in eFT spectra: (1) Stable, symmetric peaks (majority) with no frequency shift; (2) Gradually desolvating ions showing frequency drift to lower m/z (leading to eFT peak splitting/satellites at lower m/z); (3) Rare quantized frequency jumps to higher m/z indicating single charge-loss events (−1 z). Such events enabled direct mass/charge determination (example: z=144.07±0.16; mass=3,073,275±29 Da at m/z=21,343.1±0.2). - Dual-source activation model: Segment-wise analysis showed solvent loss decreases over time within transients. A composite model (linear term from ongoing Orbitrap collisions + exponential term from initial injection activation in the C-trap region at ~10 mbar and ~1 kV acceleration) fit the data with r^2≈0.998–0.9998, and both terms scaled linearly with pressure. - Performance gains via optimization and drift correction: Reducing pressure, improving initial desolvation, and shortening transients each reduced eFT peak splitting and increased usable ion fractions (e.g., sampling improvements of ~23-fold by pressure optimization, ~7-fold by better desolvation, ~13-fold by shorter transients). Frequency-chasing drift correction recovered split peaks and, overall, improved effective ion sampling by ~23-fold and approximately doubled mass precision and resolution in practice. - Charge resolution limitations and scaling: Charge resolving power improved approximately with the square root of transient time up to ~2,048 ms, with electronic preamplifier noise (amplified by Orbitrap capacitance) as the dominant limitation. Deviations at long transients were attributed to electronic ringing/background and voltage drifts, and ion-behaviour artefacts (e.g., non-circular orbits). - First experimental observation of radial motion: Long, high-S/N transients revealed modulation sidebands corresponding to radial oscillation frequencies, with mirrored features on either side of the axial peak. Both stable and unstable radial motions were observed, validating theoretical descriptions and implicating radial dynamics in intensity variability.

The findings reveal an unexpected plateau of stability for megadalton ions in Orbitrap CDMS: multiple collisions over seconds do not eject these ions but instead induce gradual desolvation, preserving coherent motion and enabling unprecedented resolving power at ultra-high m/z. This contrasts with small/denatured ions where single collisions can cause fragmentation and loss. Modeling identifies two activation sources—initial injection and ongoing collisions—governing solvent loss and frequency drift. By segmenting transients and chasing frequencies, frequency drifts and rare charge-loss events can be identified and corrected, dramatically increasing the fraction of usable single-ion events and enhancing precision/resolution. Observation of radial frequency modulations provides new mechanistic insight into ion dynamics and offers avenues to correct intensity variations due to non-circular orbits. These improvements have broad implications for single-particle analysis of heterogeneous macromolecular assemblies (e.g., viruses and gene therapy vectors), enabling Orbitrap-based CDMS on widely available platforms with higher practical performance and throughput.

This work introduces a frequency-chasing strategy for Orbitrap-based CDMS that characterizes and corrects frequency drifts of individual megadalton ions caused by desolvation and rare charge stripping. High-mass ions exhibit a stability plateau, enabling multisecond trapping and effective resolution >100,000 at very high m/z. Combining optimized pressure and desolvation with drift correction improves effective ion sampling by ~23-fold and approximately doubles mass precision and resolution. The first experimental detection of radial ion motion in an Orbitrap further expands understanding of single-ion behaviour and suggests future corrections for intensity variability. Looking ahead, extending transient durations beyond current 4 s electronics limits could approach single-elementary-charge resolution (projected near 16–32 s), unlocking even higher fidelity single-particle mass and charge measurements for complex biological and nanomaterial systems.

- Current instrumentation limited to ~4 s transients on the Q Exactive UHMR platform, constraining ultimate charge resolution achievable without hardware upgrades. - Charge resolving power is primarily limited by electronic noise (preamplifier transistors and Orbitrap capacitance), with additional deviations at long transients due to electronic ringing, background processes, and voltage drifts. - Performance and ion survival depend sensitively on pressure and desolvation/activation settings; trade-offs exist between transmission/desolvation and Orbitrap pressure that can promote frequency drift and eFT peak splitting. - Ion behaviour artefacts such as non-circular orbits and a range of accepted kinetic energies can affect induced image currents and intensity-based charge estimates. - Some lower-mass or higher-charge ions (e.g., smaller proteins) still exhibit reduced survival over long transients compared to megadalton ions.