Journal of the American Chemical Society
Communication
In another set of experiments, we screened microdroplets of
different solvents to examine their carbocation-capturing
efficiencies. We selected seven different types of electrospray-
friendly solvent systems25 (Table S2) for capturing four typical
carbocations (1, 2, 4, and 6 in Figure 2), and the results are
presented as histograms in Figure 4a−d. The data indicated the
The electrospray voltage is known to contribute charges in
the microdroplet,26 and that can also contribute to increased
focusing of charged microdroplets into the MS inlet. To
investigate the effect of DESI spray voltage, we tuned the
voltage from 0 to +5 kV for the detection of the above
carbocations (Figure 4e−h). We observed that the carbocation
detection efficiency improved on increasing the voltage of the
water spray. However, the opposite effect was observed for
carbocations wherever detected by the acetonitrile spray
(Figures 4g,h), indicating better carbocation detection with
no voltage applied to the acetonitrile spray. This result has also
been separately shown in Figure S7, which could be attributed
to the high volatility of the acetonitrile and its rapid droplet
All the above observations collectively suggested that water
microdroplets outperformed the organic microdroplets con-
sidered in this study for the capture and detection of several
short-lived and elusive carbocation intermediates. One may
wonder why the intermediate carbocations survived in the
aqueous microdroplet although they are highly susceptible to
be attacked by the nucleophilic water forming alcohol. It
appears that this microdroplet phenomenon is strikingly
different from that of the corresponding bulk phase.7,27−29
Indeed, this observation can be attributed to the emerging field
of microdroplet chemistry demonstrating the unusual process
that could occur in a tiny droplet.9−11,13,27,30−33
Although we are unable to conclusively establish the
mechanism of the carbocation interception at present, we
rationalize a viable mechanism for the same (Figure 4i) based
on what has been presented. An electrospray microdroplet,
produced under a positive potential, is highly acidic because of
the accumulation of protons by the solvent oxidation.25,26
Unlike a normal Brønsted acid, these protons lack their
counteranions (conjugate base) in the droplet. Moreover,
these protons are likely distributed on or nearer to the
microdroplet surface with equidistant spacing to minimize the
potential energy.34,35 This unique polar environment of the
air−liquid interface possibly facilitates the carbocation capture.
When the charged microdroplet impacts the reaction aliquot
on the microscope glass slide, it causes the splashing of
secondary microdroplets encapsulating the chemical and
reactive intermediate species from the reaction aliquot. A
charged species can be destabilized inside the charged
microdroplet by a huge Coulomb force of repulsion imparted
by the surface protons.31 Therefore, the carbocation (R+) can
exist as an intimate or tight ion pair in association with a
counteranion X− (e.g., the leaving group or other anions)
inside the charged microdroplet, or R+ can also exist
preferentially on the charged surface to minimize the Coulomb
force of repulsion. The existence of intimate ion pairs inside
charged microdroplets was also proposed before by Fenn.34
The mechanism of DESI suggests that secondary micro-
droplets evolve through repeated solvent evaporation and the
Coulomb fission assisted by the flow of a sheath gas and the
heated MS inlet capillary (Figure 1).36 When the intimate ion
pair (R+X−) in the vanishing charged droplet encounters the
surface by its Brownian dynamics, the attractive interaction
between the counteranion (X−) and surface protons might
weaken the interaction between R+ and X− (Figure 4i). The
thermal activation (kT) and Coulomb repulsion may
subsequently provide sufficient energy to desorb the intrinsi-
cally charged carbocation species (R+) from the highly charged
Figure 4. Screening of spray solvents and voltages to elucidate the
mechanism of carbocation interception. Four typical species, e.g.,
tertiary (1), secondary (2), nonclassical (4), and benzylic (6)
carbocations, were considered in the assessment of DESI spray
solvents (a−d) and voltages (e−h). Ion intensities are presented as
mean
SD from the triplicate analysis. An ion evaporation
mechanism is illustrated in the lower panel (i) to demonstrate the
feasibility of intercepting short-lived carbocations using positively
charged aqueous microdroplets.
superior performance of microdroplets composed of water
compared to that of organic or binary aqueous−organic
solvents in intercepting the carbocation species from an
ongoing reaction. For example, the highly unstable tert-butyl
carbocation was only detected by water spray, while other
solvent sprays failed (Figure 4a) to do so. The tuning of
solvent flow rate (5−50 μL/min) and sheath gas flow (100−
170 psi back pressure) could not establish the better
performance of organic microdroplets than that of aqueous
microdroplets (data not shown). This phenomenon can be
ascribed by the high polarity of water and its microdroplet
surface, which is discussed later. Relatively less polar solvents
(methanol, acetonitrile, and their binary mixtures) either could
not intercept (Figure 4a,b) or poorly intercepted (Figure 4c,d)
the intermediate carbocation species depending on the nature
and stability of the carbocation. For example, carbocation
intermediates 4 and 6 are somewhat better stabilized by
resonance when compared with 1 and 2 and thereby detected
with less polar solvents, albeit poorly, as mentioned above.
2461
J. Am. Chem. Soc. 2021, 143, 2459−2463