Characterization of Electrospray Drop Size Distributions by Mobility-Classified Mass Spectrometry: Implications for Ion Clustering in Solution and Ion Formation Pathways

Uncovering the Pathway of Serine Octamer Magic Number Cluster Formation during Electrospray Ionization: Experiments and Simulations

Electrospray ionization (ESI) of serine (Ser) solution generates Ser8H+ as an abundant magic number cluster. ESI clustering of most other solutes yields nonspecific stoichiometries. It is unclear why Ser8H+ dominates in the case of Ser, and how Ser8H+ forms during ESI. Even the location of Ser8H+ formation is contentious (in solution, in ESI droplets, or elsewhere). Here we unravel key aspects of the l-Ser8H+ formation pathway. Harsh ion sampling conditions promote the collision-induced dissociation (CID) of regular ESI analytes. Unexpectedly, Ser8H+ was seemingly resistant against CID during ion sampling, despite its extremely low tandem mass spectrometry (MS/MS) stability. This unusual behavior reveals that Ser8H+ forms during ion sampling. We propose the following pathway: (1) Nonspecific Ser clusters are released when ESI droplets evaporate to dryness. These initial clusters cover a wide size range, from a few Ser to hundreds or thousands of monomers. (2) The clusters undergo dissociation during ion sampling, mostly via successive loss of neutral monomers. For any source activation voltage, there is a subpopulation of clusters for which this CID cascade tends to terminate at the octamer level, culminating in Ser8H+-dominated product distributions. Mobile proton molecular dynamics simulations were used to model the entire pathway. Ser8H+ structures formed in these simulations were consistent with ion mobility experiments. The most compact structures resembled the model of [Scutelnic, V. J. Am. Chem. Soc. 2018, 140, 7554-7560], with numerous intermolecular salt bridges and H-bonds. Our findings illustrate how the interplay of association and dissociation reactions across phase boundaries can culminate in magic number clusters.

High-Speed Schlieren Imaging of Vapor Formation in Electrospray Plume

Previous mechanistic descriptions of electrosprays mostly focused on the dynamics of Taylor cones, initial droplets, and progeny droplets. However, vapor formation during droplet desolvation in an electrospray plume has not been discussed to a great extent. Here, we implement a double-pass on-axis schlieren high-speed imaging system to observe generation and propagation of vapors in an offline electrospray source under different conditions. Switching between turbulent and laminar vapor flow was observed for all of the scanned conditions, which may be attributed to randomly occurring disturbances in the sample flow inside the electrospray emitter. Calculation of mean vapor flow velocity and analysis of vapor flow patterns were performed using in-house developed image processing programs. Experiments performed at different electrospray voltages (0-6 kV), solvent flow rates (100-600 μL min-1), and methanol concentrations (50-100%), indicate only a weak dependency between electrospray voltage and mean vapor velocity, implying that the vapor is mostly neutral; thus, the vapor is not accelerated by electric field. On the other hand, electrospraying solutions of analytes (with mass 151 Da or 12 kDa) did not remarkably increase the overall vapor flow velocity. The source of vapor's velocity is attributed to the inertia of the electrospray droplets. Although there are some differences between a modern electrospray ionization (ESI) setup and the setup used in our experiment (e.g., using a higher flow rate and larger emitter), we believe the findings of our study can be projected to a modern ESI setup.

Are Most Micrometer Droplets (>10 μm) Wasted in Electrospray Ionization? An Insight from Real-Time High-Speed Imaging

Electrospray ionization (ESI) is one of the main techniques used in mass spectrometry (MS) of nonvolatile compounds. ESI is a disordered process, in which a large number of polydisperse droplets are projected from a fluctuating Taylor cone and jet protruding ESI emitter. Here, we disclose a system for sectioning electrospray plumes to discrete packets with millisecond and submillisecond lifetime, which are introduced to the MS orifice, one at a time. A high-speed camera was triggered at 10,000 frames per second to capture consecutive images of the electrospray packets transmitted to the mass spectrometer. We further correlated the high-speed images of electrospray packets with MS signals of a test analyte (acetaminophen). Following computational treatment of the images, we determined the number of droplet observations (<300), average diameter of droplets (∼10-20 μm), and average volume of droplets (few tens of picoliters) in the individual electrospray packets. The result shows that most micrometer droplets (>10 μm) do not have any significant contribution to the MS signals. This finding is in agreement with the prior conjecture that most of the MS signals are mainly attributed to nanodroplets. Based on this finding, one can deduce that only a small number of the initial microdroplets effectively carry analyte molecules that undergo ionization. We discuss that, in future, one may propose a way to "recharge" the emitted initial micrometer droplets to increase the efficiency of conventional ESI setups.

Single-Molecule Pycnometry and Shape Analysis of Ions in the Gas Phase

The analysis of ions and clusters by mobility-classified mass spectrometry provides information on the mobility of analytes in the drift gas and the analyte mass. Mass equivalent and mobility equivalent diameters of globular analytes, such as ions, poly(ethylene glycol) (PEG), and ionic liquid nanodroplets, can be correlated with good accuracy by the Stokes-Millikan mobility model. A prerequisite to such an analysis is, however, the assumption of a globular analyte shape, which then allows determination of material density for globular ions. We show that the analyte density can be evaluated with high precision, independent of any assumptions on the analyte shape, by careful analysis of analyte-PEG-cluster ions following the concept of classical pycnometry. In particular, the analyte is entrapped in a globular PEG-analyte droplet. Based on the now independently derived mobility diameter and volume equivalent diameter, it is possible to attribute two parameters, size and shape, to the analyte molecule. We demonstrate the approach for lysozyme, cyano-cobalamin (vitamin B12), and glucose, which cover two orders of magnitude in analyte mass (180···14 300 Da). The derived densities for these analytes are highly accurate, i.e., they deviate less than 1% from literature values. Our method can be applied to newly synthesized molecules, supramolecular assemblies, isolated biomolecules, and molecular clusters, where only minor amounts of materials are available. The obtained shape parameters of lysozyme and cyano-cobalamin agree well with the expected molecular shapes. Data evaluation relies only on locations of the species in the mass-mobility plane and is in principle independent of any mobility theory. Our approach is thus robust with respect to experimental uncertainties and produces identical results irrespective of the type of mobility classification and drift gas.

Efficient quenching sheds light on early stages of gold nanoparticle formation

The formation mechanism of plasmonic gold nanoparticles (Au NPs) by fast NaBH₄ induced reduction of the precursors is still under debate. In this work we introduce a simple method to access intermediate species of Au NPs by quenching the solid formation process at desired time periods. In this way, we take advantage of the covalent binding of glutathione on Au NPs to stop their growth. By applying a plethora of precise particle characterization techniques, we shed new light on the early stages of particle formation. The results of in situ UV/vis measurements, ex situ sedimentation coefficient analysis by analytical ultracentrifugation, size exclusion high performance liquid chromatography, electrospray ionization mass spectrometry supported by mobility classification and scanning transmission electron microscopy suggest an initial rapid formation of small non-plasmonic Au clusters with Au₁₀ as the main species followed by their growth to plasmonic Au NPs by agglomeration. The fast reduction of gold salts by NaBH₄ depends on mixing which is hard to control during the scale-up of batch processes. Thus, we transferred the Au NP synthesis to a continuous flow process with improved mixing. We observed that the mean volume particle sizes and the width of the particle size distribution decrease with increasing flow rate and thus higher energy input. Mixing- and reaction-controlled regimes are identified.