High-Throughput Mass Spectrometry Screening Platform for Discovering New Chemical Reactions under Uncatalyzed, Solvent-Free Experimental Conditions

Article abstract of DOI:10.1021/acs.analchem.0c02960

A gas-phase high-throughput reaction screening platform was developed for the first time to study chemical structures of closely related functional groups and for the discovery of novel organic reaction pathways. Experiments were performed using the contained atmospheric pressure chemical ionization (APCI) source that enabled nonthermal, nonequilibrium plasma chemistry to be monitored by mass spectrometry (MS) in real time. This contained-APCI MS platform allowed an array of reagents to be tested, resulting in the studies of multiple gas-phase reactions in parallel. By exposing headspace vapor of the selected reagents to corona discharge, solvent-free Borsche-Drecsel cyclization reaction, Katritzky chemistry, and Paal-Knorr pyrrole synthesis were examined in the gas phase, outside the high vacuum environment of the mass spectrometer. A new radical-mediated hydrazine coupling reaction was also discovered, which provided a selective pathway to synthesize secondary amines without using a catalyst. The mechanisms of these atmospheric pressure gas-phase reactions were explored through the direct capture of intermediates and via comparison with the corresponding bulk solution and droplet-phase reactions.

Full text of DOI:10.1021/acs.analchem.0c02960

pubs.acs.org/ac
Article
High-Throughput Mass Spectrometry Screening Platform for
Discovering New Chemical Reactions under Uncatalyzed, Solvent-
Free Experimental Conditions
Dmytro S. Kulyk, Enoch Amoah, and Abraham K. Badu-Tawiah*
Cite This: https://dx.doi.org/10.1021/acs.analchem.0c02960
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: A gas-phase high-throughput reaction screening platform was
developed for the first time to study chemical structures of closely related
functional groups and for the discovery of novel organic reaction pathways.
Experiments were performed using the contained atmospheric pressure chemical
ionization (APCI) source that enabled nonthermal, nonequilibrium plasma
chemistry to be monitored by mass spectrometry (MS) in real time. This
contained-APCI MS platform allowed an array of reagents to be tested, resulting
in the studies of multiple gas-phase reactions in parallel. By exposing headspace
vapor of the selected reagents to corona discharge, solvent-free Borsche−Drecsel
cyclization reaction, Katritzky chemistry, and Paal−Knorr pyrrole synthesis were
examined in the gas phase, outside the high vacuum environment of the mass
spectrometer. A new radical-mediated hydrazine coupling reaction was also
discovered, which provided a selective pathway to synthesize secondary amines
without using a catalyst. The mechanisms of these atmospheric pressure gas-phase
reactions were explored through the direct capture of intermediates and via comparison with the corresponding bulk solution and
droplet-phase reactions.
confinement of reactants in smaller volumes, leading to
concentration, surface, and pH effects that have been found
to collectively increase reaction rates.6−9 The purpose of DESI
high-throughput screening has remained the rapid optimiza-
tion of solution-phase reaction conditions; it is rarely used for
high-throughput discovery of new reactions. The gas-phase
HTE platform described in this study is not only capable of
reproducing known solution-phase reactions but it also allows
new organic reaction discovery. In general, the complete
absence of solvent makes gas-phase reactions very efficient and
fast. Despite this promise, gas-phase reactions are not a
common tool in most analytical laboratories. This is because
gas-phase ion/molecule and ion/ion reactions require
substantial modification to the mass spectrometer, which
limits their widespread use.
INTRODUCTION
■
High-throughput experimentation (HTE) enables a large
number of experiments to be conducted in parallel through
systems automation. The process is very well developed for
bioassays1−3 but less so for chemical reaction screening. We
report here the first gas-phase high-throughput experiment for
chemical reaction screening under ambient conditions.
Biochemical experiments are often performed in an aqueous
medium and at (or near) room temperature so HTE
implementation is straightforward. Chemical reactions, on
the other hand, are carried out in different solvents and in a
broad temperature range. The use of volatile organic solvents
brings further complications in terms of material compatibility
and evaporative loss, all making it even more challenging to
automate the HTE process for chemical reactions. Therefore,
the ability to perform chemical reactions in a solvent-free
environment, at room temperature, and pressure may represent
a unique opportunity to increase the speed with which new
chemical reactions can be screened.
Like solution-phase experiments, reported gas-phase HTE
mimics have also focused on the development of standard
workflows for targeted proteomic assays.10 For example, the
most widely used gas-phase reaction is electron transfer
Recently, the use of desorption electrospray ionization
(DESI) in HTE has alleviated some of the difficulties
associated with engineering microarrays for chemical reaction
screening through the integration of sample introduction,
reaction, and analysis of products into a single mass
spectrometry (MS) experimental step.4,5 The presence of
evaporating charged microdroplets in DESI allows the
Received: July 11, 2020
Accepted: October 21, 2020
https://dx.doi.org/10.1021/acs.analchem.0c02960
© XXXX American Chemical Society
Anal. Chem. XXXX, XXX, XXX−XXX
A

Analytical Chemistry
pubs.acs.org/ac
Article
Figure 1. Schematic of the contained-APCI MS screening platform. Containers (A, B, and C) can be filled (<100 μL) with different reagent
combinations (A and C) and analyte (B), and robotically or manually exposed to corona discharge (red thunder) by sliding plates. Headspace
vapor or electrostatically attracted particles of reagents react with each other in the gas-phase upon plasma initiation through the application of high
direct current (DC) voltage (4−6 kV) to the stainless steel needle. Detection of reaction products is conducted by mass spectrometry in real time.
Insert (i) illustrates plasma-based ion/molecule reaction, leading to analyte (M) ionization to give [M + H]+ ions. Insert (ii) is the zoomed-in
region of the PTFE container relative the mass spectrometer, showing distances between MS inlet, microvials containing samples, and APCI needle.
dissociation, which enables high-throughput liquid chromato-
graphic tandem MS (MS/MS) analysis for rapid top-down
protein identification using commercially available ion trap
mass spectrometers.11−13 The fundamental experiments in this
field include the concentration of ion current through ion/ion
reactions (in a so-called ion packing process) to improve
analytical sensitivity.14 Ozone-induced dissociation is another
relevant gas-phase reaction that has enabled positional isomers
and the stereochemistry of unsaturated fatty acids and lipids to
be determined in a high-throughput fashion.14−16 Gas-phase
ion/molecule reactions are also extremely powerful in
providing detailed structural information for unknown
compounds, including the type and number of different
functionalities present in the compound.17 Notice that all these
gas-phase reactions are performed inside the high vacuum
environment of the mass spectrometer, something that limits
access to analyte, and thus reducing the overall throughput of
the experiment. Also, determination of the chemical structure
has been the ultimate objective for these conventional gas-
phase ion/molecule and ion/ion reactions. The proposed gas-
phase HTE platform is unique in that it operates at
atmospheric pressure, outside the high-vacuum environment
of the mass spectrometer, and it is capable not only for
structural determination but also for the discovery of new
organic reaction pathways. Most importantly, the platform is
applicable for screening both volatile and nonvolatile analytes
without heating and with no need for an external catalyst,
which results in gas-phase organic reactions in a solvent-free
environment.
there are approximately 1060 drug-like pharmaceutically
perspective molecules employed in drug discovery,18 the
speed and ease of operation of reaction screening becomes
critical parameters in HTE. MS-based analytical platforms
show great potential in this respect,19−23 but almost all
reported methods require sample preparation and optimization
of several solution-phase parameters (e.g., solvent, pH, and
temperature), which significantly increases the analysis time.
(3) To facilitate smaller sample consumption (<nmol/
min)24,25 during HTE, this objective is achieved via the use
of headspace vapors or electrostatically attracted particles from
volatile and nonvolatile reagents, respectively; (4) to enable
highly efficient and sensitive detection for a variety of analyte
types (polar, nonpolar, volatile, and nonvolatile), which can be
uniquely suited for the detection of reaction intermediates with
different chemical structures.
There is a substantial literature on the use of nonthermal
plasmas generated by electrical discharge at atmospheric
pressure in chemical synthesis.26−28 This research area is
particularly attractive because plasmas can be powered by
intermittent electricity generated from renewable sources such
as solar and wind, thus incorporating valuable chemical
synthesis into a renewable carbon cycle that can reduce
dependence on fossil fuels. An important feature about plasma
chemistry is that large quanta of energy can be transferred to a
reactant in a single excitation event, making it possible to
access higher excited states. Aspects of our previous research
have focused on the use of photons to generate excited-state
species for reactions in microdroplets,29,30 but this requires
photosensitizers whose redox potential match that of the
analyte(s) of interest. This limits the general utility of
associated analytical photoreaction screening methods. Since
electron collisions in plasma do not depend on chromophoric
groups or sensitizers, we anticipated that a screening platform
The motivation for creating atmospheric pressure MS-based
gas-phase HTE platform is fourfold: (1) possible application in
drug design and discovery to facilitate rapid prediction of
molecular structure and physicochemical properties; (2) to
enable direct analysis without sample preparation. Given that
B
https://dx.doi.org/10.1021/acs.analchem.0c02960
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
pubs.acs.org/ac
Article
based on plasma should make it possible to excite all kinds of
molecules.
technique is safe to operate, minimally destructive, and enables
the sample to be reanalyzed by different analytical methods.1−3
Headspace vapors and particles of nonvolatile solid reagent are
supplied through the outlet toward the inlet of the mass
spectrometer. Upon the application of high DC voltage (4−6
kV) to the stainless steel APCI needle, invisible corona
discharge (no sparking at optimized distances used) is induced
in the vicinity of the outlet, allowing brief exposure of the gas-
phase analyte to the plasma and initiating instantaneous
solvent-free chemical reactions and ionization at ambient
pressure and temperature. Sliding plates in their turn allow an
array of reagents to be introduced rapidly for high-throughput
reaction monitoring. The ability of the contained-APCI
platform to ionize a broad range of reagents with markedly
different physicochemical properties (volatility, polarity,
basicity, and acidity, etc.) makes this device particularly
effective for studying various gas-phase reactions using mass
spectrometry. The characterization of the ionized reaction
products and intermediates was achieved in real time using
both tandem MS and exact mass measurements.
Mass Spectrometry. Reaction systems were analyzed by a
Thermo Fisher Scientific Velos Pro LTQ and LTQ Orbitrap
(for high-resolution data) mass spectrometers (San Jose, CA,
USA). Unless otherwise stated, MS parameters employed were
as follows: 400 °C capillary temperature, 4−6 kV spray voltage,
5 mm distance between the tip of the APCI needle and
microvial containing the analyte, 5 mm distance from the tip of
APCI needle to MS inlet, 3 microscans, 100 ms ion injection
time, and 60% S-lens voltage for Velos Pro LTQ and 55 V tube
lens voltage for LTQ Orbitrap. Spectra were recorded for at
least 30 s, yielding an average of 300 individual scans. Thermo
Fisher Scientific Xcalibur 2.2 SP1 software was utilized for MS
data collecting and processing. Unless otherwise mentioned,
tandem MS with collision-induced dissociation (CID) was
performed for analyte identification. Thirty percent (manu-
facturer’s unit) and 1.5 Th (mass/ charge units) for isolation
window of normalized collision energy were selected for the
CID tests.
Reactions in Bulk and Droplet Media. All reaction
mixtures presented here for solution and droplet chemistry
were prepared in acetonitrile at 100 μM concentration (unless
otherwise stated) for each reagent. Similar mass spectra were
obtained for reaction mixtures prepared at higher concen-
trations (up to 500 μM) and in different solvents such as
methanol, methanol−water, and ethyl acetate, as well as in
their acidified solutions with formic acid. Concentrations
higher than 500 μM were not employed to prevent
contaminations of MS instruments.
Materials and Reagents. Acetonylacetone (97%), ben-
zaldehyde (99%), n-butylamine (99.5%), cyclohexanone
(99%), ethanolamine (98%), phenylhydrazine (97%), 2,4,6-
triphenylpyrylium tetrafluoroborate (98%), and screw-top glass
vials with 9 mm screw-thread neck and 100 μL glass microvial/
insert were all purchased from Sigma-Aldrich (St. Louis, MO,
USA). Aniline (99.8%) was provided by Acros Organics (Geel,
Belgium). Chlorophenylhydrazine was obtained from Santa
Cruz Biotechnology (Dallas, TX, USA). Bromophenylhydra-
zine (98%) was supplied by Syntonix Pharmaceuticals
(Waltham, MA, USA). Nitrophenylhydrazine was bought
from Chem Service (West Chester, PA, USA). Pentylhydrazine
(95%) was provided by Combi-Blocks (San Diego, CA, United
States). 2-Butanone was acquired from Fisher Scientific
(Pittsburgh, PA, USA).
Therefore, we created a novel contained-atmospheric
pressure chemical ionization (contained-APCI, Figure 1)
source for high-throughput gas-phase reaction screening
using nonequilibrium plasma chemistry. The contained-APCI
platform is capable of screening up to two chemical reactions
in parallel for the array of reagents (e.g., monitoring of reaction
between reagents A and B vs reagents B and C, where B is the
analyte). The array of reagents to be reacted against the analyte
B is delivered by a sliding plate. To minimize secondary effects
from reactive oxygen species generated from atmospheric
gases, we designed the contained-APCI screening platform to
have a short path length (<5 mm), leading to short residence
time (<5 s) of reagents in the plasma. This was achieved by
introducing reactants in the form of headspace vapors (or
particles released into air via electrostatic attraction due the
proximal DC voltage)24 in the open laboratory environment.
To increase ease of operation, the platform was further
simplified by not using external gases for plasma generation.
Plasma species generated with the assistance of N2 gas are
known to be more energetic and reactive compared with O2.31
This implies that although the terminal active reagent ions in
APCI typically involve protonated water clusters
H+(H2O)n,32,33 analyte’s internal energy deposition can be
controlled by the type of nebulizer gas used. Hence, by
employing ambient air as discharge gas, we expect to further
decrease secondary effects.
We applied the optimized contained-APCI platform to study
gas-phase Borsche−Drecsel cyclization reaction, Katritzky
chemistry, and Paal−Knorr pyrrole synthesis, all in a solvent-
free environment. Through the careful selection of these
known reactions, we showcased for the first time the ability to
differentiate closely related functional groups (e.g., amines vs
hydrazine) through gas-phase ambient reactions. Screening of
multiple reagents (5 total) against a selected 2-butanone
reactant and subsequent real-time product detection was
accomplished in less than 60 s. The contained-APCI screening
platform also enabled the discovery of a new radical mediated
gas-phase hydrazine coupling reaction, which afforded a facile
pathway to selectively synthesize secondary amines. Finally,
mechanistic insights were sought through the direct capture of
intermediates and comparison of selected gas-phase reactions
to the corresponding bulk solution and droplet-phase
reactions.
EXPERIMENTAL SECTION
■
Contained-APCI Platform. In its full operational form, the
contained-APCI reaction screening platform (Figure 1)
comprised of a PTFE container with a closed top and open
bottom (4 cm L × 2 cm W × 0.5 cm H) in which three
independent 100 μL glass microvials A, B, and C can be
inserted for sample introduction. Reagents in the two of the
glass microvials (A and C) can be interchanged via a sliding
plate containing other microvials, allowing a series of reactions
to be monitored simultaneously. The only outlet (ID, 5 mm)
for all three inputs is located at the top of the PTFE container.
The stainless steel APCI needle is placed adjacent to this single
outlet, allowing gas-phase reaction and ionization to be
initiated upon the application of DC voltage. Because the
condensed-phase sample (100 μL) is contained in a microvial
and analyzed through a noncontact means (i.e., there is no
direct contact between the sample and APCI needle), our
C
https://dx.doi.org/10.1021/acs.analchem.0c02960
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
pubs.acs.org/ac
Article
Although this species is not directly involved in the B-D
cyclization reaction, its presence indicates high energy
collisions (energy ≥2.6 eV for N−N bonds in hydrazines)38
in the corona discharge upon the application of ≥4 kV DC
voltage). No ion signal corresponding to reactants and
products was observed in the absence of applied DC voltage.
To evaluate solvent effects, the same reaction was conducted in
solution-phase (reagent concentrations ≤500 μM), which was
subsequently compared to the reactivity of charged micro-
droplets derived from electrospray ionization. All experiments
were performed without the addition of external acid. As
expected, cyclization product 4 was not formed, even at higher
MS inlet temperatures (Figure S2). Acidification also did not
provide any improvements in the charged microdroplet
reaction condition under the concentrations tested (data is
not shown); this is perhaps due to the need to properly tune
pH and solvent properties. This interpretation is consistent
with a recent study, which required 1 M methanolic HCl to be
added to the reaction mixture (in the presence of 0.1 M
phenylhydrazine) before droplet reactions39 were initiated in
the electrospray process.
Scheme 1. Uncatalyzed Gas-Phase B-D Cyclization Reaction
between Cyclohexanone (1) and Phenylhydrazine (2)
Figure 2. Positive ion mode mass spectrum of the B-D cyclization
showing (a) full MS analysis and (b) MS2 product ion scan at (i) m/z
172 and MS3 spectrum of m/z 172 derived from the MS2 of (ii) m/z
189. *[2 + H−NH2]·+ species at m/z 93.
Solvent-Free Katritzky Chemistry. Gas-phase version of
Katritzky chemistry40−42 involving pyrylium cation (5) and
ethanolamine (6) was also investigated in the absence of heat
and solvent (Scheme 2). In this case, the type of product
RESULTS AND DISCUSSION
■
Gas-Phase Borsche−Drecsel Cyclization. Motivated by
our desire to differentiate closely related organic compounds
through the high-throughput gas-phase reaction screening, we
started our investigation with Borsche−Drecsel (B-D)
cyclization. The conventional solution-phase B-D cyclization
reaction (Scheme 1) is achieved by heating the reaction
mixture (i.e., hydrazine and ketone) in the presence of
concentrated hydrochloric acid.34,35 In our experiment,
headspace vapors of cyclohexanone 1 (MW, 98) and
phenylhydrazine 2 (MW, 108) were simply brought in close
proximity and ionized by the contained-APCI source. The
results of this experiment are shown in Figure 2a, where the
expected cyclization product 4 was observed at m/z 172
immediately upon the application of DC voltage. The
protonated form of the initial condensation product, cyclo-
hexanone phenylhydrazone 3, was also observed at m/z 189,
alongside the unreacted reagents 1 and 2 at m/z 99 and 109,
respectively. The identity of the cyclization product 4 was
confirmed by MS/MS experiment (Figure 2b-i), which showed
fragment ions at m/z 144 and 130 via the loss of ethylene and
propene, respectively. These fragment ions are consistent with
the fragmentation pattern observed for protonated tetrahy-
drocabazole (4) generated in the gas phase (under vacuum)
through multiple stage (MS3) fragmentation of m/z 189 to
give m/z 172 (Figure 2b-ii). The correspondent MS2
fragmentation pattern of m/z 189 is provided in the
Supporting Information (Figure S1). Moreover, the identities
of reaction products and intermediates were further proved by
high-resolution (Orbitrap MS) experiments that provided −4.0
and −4.5 ppm mass accuracy for compounds 3 and 4,
respectively. The successful implementation of gas-phase B-D
cyclization under the contained-APCI experimental condition
indicates that the ionizing particles are highly acidic in nature,
suggesting the presence of low molecular weight protonated
water clusters [H+(H2O)n; n = 1, or 2].36,37 The fact that
expected products are detected without heating also suggests
that reactants are activated by the collision events in the
plasma. The energy deposited is sufficient to break the N−N
bond in the hydrazine, resulting in the release of amino phenyl
radical species, which was detected at m/z 93 (Figure 2a).
Scheme 2. Solvent-free Katritzky Chemistry Involving 2,4,6-
Triphenylpyrylium Cation (5) and Ethanolamine (6) under
Contained-APCI Reaction Condition
formed was strongly dependent on the magnitude of the
applied DC voltage. For example, only the pyrylium cation
(nonvolatile solid sample, estimated vapor pressure, VP ≈ 1.9
× 10−7 kPa)43 was detected in high abundance at m/z 309
(Figure 3b) when no DC voltage was applied, indicating gas-
phase reaction between 5 and 6 did not occur. An obvious new
peak appeared at m/z 308 (8) upon the application of 4 kV
DC potential (Figure 3c); this marked the onset of micro-
plasma formation (see insert, Figure 3c for the dependence of
m/z 308 signal on applied voltage). We confirmed that 8 is not
produced by hydride elimination from 5 since no such peak
was observed in the absence of 6. Instead, the conversion of
the pyrylium cationinto the corresponding pyridinium cation
occurred via reaction with ethanolamine vapor to give 7, which
spontaneously decomposed by losing ethylene oxide (MW, 44
Da) to yield 8. This spontaneous decomposition is not
surprising given the reactive nature of the contained-APCI
source. In this case, the presence of the nonmobile permanent
charge on the quaternary ammonium nitrogen atom in 7 drives
the loss of neutral ethylene oxide.
To further explore the full spectrum of chemistry occurring
in this gas-phase Katritzky transamination reaction, we
increased the applied voltage from 4 to 6 kV, expecting a
concomitant increase in the concentration of active species
(e.g., electron density) present in the corona discharge. As
anticipated, two additional reaction products were observed at
m/z 327 and 388 under this intense plasma condition (Figure
3d), which provided valuable insight into the reaction
D
https://dx.doi.org/10.1021/acs.analchem.0c02960
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
pubs.acs.org/ac
Article
mol) in a gas-phase reaction with 2,4,6-triphenylpyrylium
cation at 6 kV. Despite the marked difference in vapor pressure
(>100× in favor of n-butylamine), we observed significant
decrease in reaction efficiency when n-butylamine vapor was
reacted with the pyrylium cation under this intense plasma
reaction condition. In particular, while pyrylium to pyridinium
conversion still occurred in real time (Scheme S1, Figure S5),
detectable ion signals related to the pseudobase intermediate,
10, [e.g., m/z 325 (M−H)+, 327 (M + H)+, and 400 (M + n-
butylamine + H)+] were not observed. This result clearly
demonstrates that the contained-APCI reaction screening
platform is sensitive to subtle changes in a chemical
environment and that its operation is not solely governed by
activation via electron collisions. In this case, the opening of
the ring in the pyrylium cation to form the di-ketone
pseudobase intermediate is found to be affected by the
presence of different primary amines, an effect that is also
observed in solution phase47 and thus confirming distinct
mechanisms can be activated under different reaction
conditions.
Gas-Phase Paal−Knorr Pyrrole Synthesis. The realiza-
tion that primary amines can react favorably with the di-ketone
functionality (e.g., as in the pseudobase intermediate 10) in the
gas-phase led us to further investigate solvent-free Paal−Knorr
pyrrole synthesis (Scheme 3 and Figure S6) using the new
contained-APCI platform. The Paal−Knorr pyrrole synthesis is
a very valuable synthetic method for obtaining substituted
furans and pyrroles, which are common structural components
of many natural products. The importance of this reaction is
exemplified by the numerous methodologies developed to
improve yield through the use of greener reaction conditions,
including the use of micellar systems in water,48 ionic liquids,49
microwave heating,50 and ultrasound.51 We envisioned that
charged microdroplets derived from electrospray52,53 may also
represent another green media for the Paal−Knorr pyrrole
synthesis but all attempts failed to yield expected reaction
products under this confined volume droplet condition
(Scheme S2 and Figure S7). We then performed the reaction
under solvent-free conditions using headspace vapors of
acetonylacetone (11) and n-butylamine (12) in the presence
of corona discharge. We observed high abundance of the
Schiff’s base 13 at m/z 170 and the final cyclized pyrrole
product 14 (m/z 152) immediately after applying 6 kV of DC
voltage (Figure 4a). Here too, the final product structure was
confirmed by MS/MS (Figure 4b-i and Figure S6), high-
resolution Orbitrap analysis (mass tolerance 0.4 ppm), and via
the analysis of standard 1-butyl-2,5-dimethyl-1H-pyrrole
compound, which showed the identical MS/MS fragmentation
pattern (Figure 4b-ii). The fact that two molecules of n-
butylamine do not attack acetonylacetone to react with the
individual ketone groups separately indicates that the
cyclization step following the formation of the Schiff’s base is
fast. Similar product distribution was recorded when ethanol-
amine was used to react with acetonylacetone in gas phase
(Figures S7 and S8). Here, we observed improved reaction
efficiency in accordance with increased gas-phase basicity for
ethanolamine compared with n-butylamine. The Schiff’s base
(m/z 158) as well as the cyclized pyrrole (m/z 140) reaction
products were detected at 3.6 and 3.9 ppm mass accuracies,
respectively, and further characterized by MS/MS. It is
important to note that stable cyclization products were
detected in the gas-phase Paal−Knorr pyrrole synthesis
without any decomposition. This is contrary to the
Figure 3. (a) Schematic showing the proposed pathway for gas phase
Katritzky reaction involving 2,4,6-triphenylpyrylium cation and
ethanolamine. Positive-ion mass spectra recorded at (b) 0 kV, (c) 4
kV, and (d) 6 kV. Insert in (c) shows extracted ion chronogram at m/
z 308 (final product 8), indicating dependence of reaction on applied
DC voltage. Inserts in (b, d) show MS2 spectra of reaction products
m/z 309 and m/z 388, respectively. *[10−H]+ species at m/z 325.
mechanism. The reaction between 5 and an amine is known to
proceed via two stable intermediates: (i) the pseudobase
intermediate (10, Figure 3a) is formed from a direct
interaction between water and 5. Subsequent reaction between
10 and the amine proceeds to give the final pyridinium product
7 after losing two water molecules, and (ii) 2H-pyran
intermediate (9), which is formed through a direct reaction
between 5 and 6. The conspicuous absence of the stable 2H-
pyran intermediate at m/z 370 (Figure 3d) rules out this
pathway, which is consistent with previously reported solvent-
limited Katritzky chemistry.44,45 Instead, reaction pathway
involving the pseudobase intermediate is clearly evident in the
detection of ion signals at m/z 327 and 388, which represent
the protonated and ethanolamine adducts of 10, respectively.
Multiple-stage tandem MS experiments confirmed that these
two intermediates are related in that the adduct at m/z 388
dissociated to give m/z 327 (insert, Figure 3d), which in turn
produced fragmentation pattern in MS3 that was similar to the
MS2 product ion spectrum generated from the pseudobase 10
(Figure S3). The identity of all observed reaction products was
confirmed by high-sresolution data (mass accuracy: 1.3 ppm
for 8, 0.04 ppm for 9, 1.9 ppm for 10, and 1.9 ppm for [10 +
6] adduct). For the purposes of comparison, we performed
bulk-solution Katritzky reaction, which showed the existence of
the 2H-pyran intermediate 9 and final product 7, within <3
min of reaction time using excess ethanolamine (Figure S4).
The pseudobase intermediate 10 and the decomposition
product 8 were completely absent in the solution-phase
reaction, again confirming a different pathways for the gas-
phase atmospheric pressure Katritzky chemistry.
pH and amine basicity are known to influence the efficiency
of bulk solution-phase Katritzky reactions.46 To investigate if
similar effect will exist in the gas-phase, we replaced
ethanolamine (VP, 64 Pa; gas-phase basicity, 896.8 kJ/mol)
with n-butylamine (VP 9 kPa; gas-phase basicity, 886.6 kJ/
E
https://dx.doi.org/10.1021/acs.analchem.0c02960
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
pubs.acs.org/ac
Article
corresponding gas-phase version of Katritzky reaction where
the presence of the permanent positive charge on the cyclized
pyridinium cation facilitated product decomposition. This
insight reaffirms the contained-APCI MS reaction screening
platform is sensitive to the chemical structure, and that the
observed product distribution is not a mere result of harsh
ionization conditions. For example, when phenylhydrazine
vapor (as opposed to primary amine) was exposed to
acetonylacetone, the reaction is stopped at Schiff’s base
formation without subsequent cyclization (Scheme S3 and
Figure S9). Just like the charged micropdroplet environment
failed to yield expected products in the absence of catalyst,
bulk solution-phase reaction also did not afford detectable
cyclization products for all reactants tested.
reaction between the two radical intermediates. Figure 5b
shows a comparison of the MS/MS product ion spectrum for
synthesized diphenylamine product (18a, Figure 5b-i) to the
MS/MS spectrum derived from a standard diphenylamine
compound (Figure 5b-ii). As can be observed, the two spectra
are identical proving the structure of 18a. The identity of the
gas-phase product 18a from phenylhydrazine (15a) was
confirmed by exact mass measurements where we recorded
0.02 ppm mass accuracy. Aside from phenylhydrazine, other
aromatic hydrazines have shown the same type of radical
chemistry when exposed to plasma under the contained-APCI
reaction condition, even when the polarity of the discharge
voltage was changed from positive to negative.
Scheme 4. Uncatalyzed Conversion of aromatic
phenylhydrazine derives in radical-mediated plasma
chemistry
Scheme 3. Paal−Knorr Pyrrole Synthesis from
Acetonylacetone (11) and n-Butylamine (12) Using the
Gas-Phase Contained-APCI Reaction Screening Platform
a
a
Note that R = H (15a, 18a), R = Cl (15b, 18b), R = Br (15c, 18c),
and R = NO2 (15d, 18d)
Figure 4. (a) Positive ion mode full mass spectrum for Paal−Knorr
reaction recorded after exposing 11 and 12 headspace vapors to
corona discharge. (b) Comparison of MS2 spectra of gas-phase
reaction product 14 produced in (a) at (i) m/z 152 with MS/MS of
(ii) standard 1-butyl-2,5-dimethyl-1H-pyrrole.
Figure 5. (a) Positive ion mode mass spectrum recorded after
exposure of phenylhydrazine headspace vapor to corona discharge.
(b) CID product ion mass spectra for (i) atmospheric pressure
coupling product 18a in (a) and (ii) standard diphenylamine at m/z
170.
Gas-Phase Radical-Mediated Coupling of Hydrazines
to Give Secondary Amines. During the screening process,
we observed a unique radical chemistry involving hydrazines
that occurred only in the absence of a second reactant. Free
radical formation from hydrazines is not uncommon,54,55 it can
result from both one- and two-electron oxidations reactions.
However, the subsequent coupling chemistry that we have
observed (Scheme 4) yielding secondary amines has not been
reported before. This uncatalyzed plasma-mediated amine
synthesis via the formation of free radicals from hydrazine is
applicable to both aromatic and aliphatic compounds,
indicating that it could represent a facile route to synthesize
secondary amines, which are extremely important pharmaco-
phores in numerous physiological activities.56 In a typical
experiment, headspace vapor of different hydrazines (aliphatic
and aromatic) were exposed to corona discharge at 6 kV. All
hydrazines followed the mechanism illustrated in Scheme 4, in
which the N−N and C−N bonds were cleaved homolytically
to form R−NH· (16) and R· (17) radicals, respectively, which
subsequently reacted to give the final coupling product (18).
An example is shown in Figure 5a demonstrating the formation
of protonated radical 16 (m/z 93) and protonated final
product 18a (diphenylamine, m/z 170) due to coupling
For instance, Figure S10a shows the negative-ion mode mass
spectrum recorded when p-chlorophenylhydrazine (MW, 143
Da) was exposed to corona discharge generated by negative
DC voltage (−5 kV). The spectrum shows the presence of
radical anions M·− and (M−NH2)·− at m/z 143 and 127,
respectively, alongside the expected coupling product as
deprotonated (M−H)− and hydrated (M−H + H2O)− species
at m/z 237 and 255, respectively. The insert in Figure S10a
shows the isotopic distribution of the hydrated product ion,
which matches expected pattern for a compound containing
two chlorine atoms. Figure S10b represents the positive ion
mode mass spectrum recorded after exposure of pentylhy-
drazine vapor to corona discharge, which showed the
production of a coupled product at m/z 158 in high abundance
from this aliphatic hydrazine. An unexpected NH adduct (m/z
173, Figure S10b) was also observed. Other hydrazines tested
include p-bromophenylhydrazine and p-nitrophenylhydrazine;
F
https://dx.doi.org/10.1021/acs.analchem.0c02960
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
pubs.acs.org/ac
Article
Figure 6. (a) Total ion chromatogram, TIC, and (b−f) extracted ion chronograms (XIC) of high-throughput screening involving reaction of 2-
butanone with (b) butylamine (product m/z 128), (c) phenylhydrazine (product m/z 163), (d) ethanolamine (product m/z 116), (e)
pentylhydrazine (product m/z 157), and (f) aniline (product m/z 148). Reaction time was kept at 5 s per sample, followed by another 5 s waiting
time to limit carryover issues.
these were analyzed in positive and negative ion modes and
showed the expected coupling products (Figures S11−S13).
High-Throughput Experimentation. Thus far, we have
discussed the gas-phase versions of B-D cyclization, Katritzky
chemistry, Paal−Knorr pyrrole synthesis, and the novel radical
mediated hydrazine coupling reaction and observed the
expected solution-phase reaction product without the use of
a catalyst. As already alluded to, the underlying objective for
studying these reactions is the creation of high-throughput gas-
phase screening platform for the differentiation of closely
related chemical structures. Here, we summarize our findings
based on reactions discussed above. Take primary amines and
hydrazines as an example. These two very similar functional
groups can be easily distinguished simply by exposing the
headspace vapors to corona discharge in the contained-APCI
MS platform. Hydrazines undergo coupling reactions via
radical-mediated mechanism and culminate in the formation of
secondary amine, while primary amines form N-alkylation
tertiary amine product via carbocation formation (Scheme S4
and Figure S14). Both intermediates (radical vs carbocation)
can be detected in real time as well as the distinct products.
When reacted with monoketones (cyclohexanone and
acetone), primary amines form only the corresponding Schiff’s
base (Schemes S5−S7and Figures S15−S17), but hydrazines
afford both the condensation and cyclization products, always
spaced by 17 Da mass difference, indicating the loss of
ammonia from the Schiff’s base to give the final cyclized
product (Figure 2a and Figure S18,S19, and Scheme S8). On
the other hand, when exposed to headspace vapors of di-
ketone functionality (e.g., acetonylacetone), primary amines
cyclized through two sequential losses of water molecules
(Figure 4) whereas hydrazines did not (Figure S9). Clearly,
these gas-phase reactions also enable aliphatic mono-ketones
and di-ketones to be distinguished through reaction with
primary amines and hydrazines. Aromatic and aliphatic
hydrazines can be distinguished via gas-phase B-D cyclization;
aromatic hydrazines cyclized when exposed to vapors of
monoketones but aliphatic hydrazine did not. Hydrazines did
not react with the pyrylium cation in gas-phase Katritzky
chemistry, providing a means to differentiate aliphatic amines
from aliphatic hydrazines although the two compounds react
with the carbonyl functional group to form the corresponding
imine.
To demonstrate the high-throughput capabilities of this new
contained-APCI MS screening platform, we tested the
reactivity of five different compounds (n-butylamine, phenyl-
hydrazine, ethanolamine, pentylhydrazine, and aniline),
separately toward 2-butanone vapor in real time. Exposure
time for each reagent was kept at 5 s, followed by another 5 s
delay time yielding a total of 10 s interval between reactants,
which was found optimal to limit carryover effect. No
significant carryover effect was observed for all 13 chemical
reagents tested. The noncontact nature of the contained-APCI
platform also aids in limiting contamination.24 Therefore, the
reactivity of all five reagents could be screened in under 60 s.
The results of this experiment are summarized in Figure 6 and
Figure S20, which show clean product formation for each
reactant without interference from previously analyzed
reagents. While this experiment attempts to differentiate
amines from hydrazines using their reaction with 2-butanone,
it can be observed that the majority of the reactants form
similar product, making functional group identification
challenging. This issue can be addressed through the
implementation of other reactions in parallel. Here is where
the high-throughput experimentation capabilities of the
contained-APCI MS platform can be realized. In this respect,
the experimental setup described in Figure 1 having three
inputs is not intended for three component reaction screening.
Instead, the three inputs are proposed to allow a given analyte
(reagent B, Figure 1) to be interrogated by two different
reagents (A and C) in parallel. For example, both aniline and
phenylhydrazine reagents react individually with 2-butanone to
give the corresponding Schiff’s base via the loss of a water
molecule. By replacing the 2-butanone analyte with pyrylium
cation, only the amine is expected to react to produce the
corresponding pyridinium cation as demonstrated in Figure
S21. By combining this high-throughput experimentation
procedure with tandem MS, it should be possible to obtain
complete structural information in a matter of seconds,
something that could be useful for structural discriminations
G
https://dx.doi.org/10.1021/acs.analchem.0c02960
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
pubs.acs.org/ac
Article
in situations where high-resolution instruments are not
available. The process can be accomplished manually or via a
robotic arm. Analytically, the ability to perform this experiment
manually will be advantageous in field applications (i.e., on-site
analysis) for complex mixture analysis, where the front end
reactions can produce a shift in mass for the analyte, thereby
providing more confidence for MS/MS experiments conducted
without prior separation.
43210, United States; orcid.org/0000-0001-8642-3431;
Email: badu-tawiah.1@osu.edu
Authors
Dmytro S. Kulyk − Department of Chemistry and
Biochemistry, The Ohio State University, Columbus, Ohio
43210, United States
Enoch Amoah − Department of Chemistry and Biochemistry,
The Ohio State University, Columbus, Ohio 43210, United
States
CONCLUSIONS
■
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.analchem.0c02960
In summary, we have introduced a new contained-APCI MS
reaction screening platform to monitor gas-phase reactions at
atmospheric pressure in real time using headspace vapors. We
demonstrated that solution-phase reaction products expected
from B-D cyclization, Katritzky chemistry, and Paal−Knorr
pyrrole synthesis could be formed in the gas-phase without the
use of a catalyst. This suggested that the nonequilibrium
plasma chemistry employed in the contained-APCI platform is
chemically selective, enabling closely related functional groups
(e.g., amines and hydrazine) to be distinguished in a high-
throughput fashion. The contained-APCI MS screening
platform also enabled the discovery of a novel radical-mediate
hydrazine coupling reaction in the gas-phase for the synthesis
of aliphatic and aromatic secondary amines. The absence of
solvent and effective electron collisions in the ambient corona
discharge offered increased reaction efficiency for the gas-
phase, compared with the corresponding reactions performed
under bulk solution-phase and charged microdroplet reaction
conditions. The simplicity, low sample consumption, and the
speed of this high-throughput experimentation involving gas-
phase reaction screening, occurring without heat, solvent,
catalyst, and nebulizing gas, and the possibility of studying the
reactivity of nonvolatile organic compounds under solvent-free
environment give these findings potential to have significant
impact in drug discovery as well as in new reaction discovery.
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by National Science Foundation
(award no. CHE-1900271) and the U.S. Department of
Energy, Office of Science, Office of Basic Energy Sciences,
Condensed Phase and Interfacial Molecular Science (award no.
DE-SC0016044).
REFERENCES
■
(1) Inglese, J.; Johnson, R. L.; Simeonov, A.; Xia, M.; Zheng, W.;
Austin, C. P.; Auld, D. S. Nat. Chem. Biol. 2007, 3, 466−479.
(2) Montgomery, J. Science 2011, 333, 1387−1388.
(3) Sun, S.; Kennedy, R. T. Anal. Chem. 2014, 86, 9309−9314.
(4) Wleklinski, M.; Loren, B. P.; Ferreira, C. R.; Jaman, Z.;
Avramova, L.; Sobreira, T. J. P.; Thompson, D. H.; Cooks, R. G.
Chem. Sci. 2018, 9, 1647−1653.
(5) Loren, B. P.; Ewan, H. S.; Avramova, L.; Ferreira, C. R.; Sobreira,
T. J. P.; Yammine, K.; Liao, H.; Cooks, R. G.; Thompson, D. H. Sci.
Rep. 2019, 9, 14745.
(6) Girod, M.; Moyano, E.; Campbell, D. I.; Cooks, R. G. Chem. Sci.
2011, 2, 501−510.
(7) Banerjee, S.; Zare, R. N. Angew. Chem., Int. Ed. 2015, 54, 14795−
14799.
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.analchem.0c02960.
Tandem MS analysis of product 3 from B-D cyclization
reaction; B-D cyclization reaction for bulk, droplet, and
gas phases; fragmentation of the pseudobase, pyrylium/
pseudobase adduct, and pyridinium cation for Katritzky
reaction; Katritzky reaction, and kinetics in bulk
conditions; reaction of 2,4,6-triphenylpyrylium cation
and n-butylamine; tandem MS fragmentation patterns
for Paal−Knorr synthesis; analysis of acetonylacetone/
ethanolamine reaction; interaction between acetonyla-
cetone and phenylhydrazine; radical coupling of
hydrazines; N-alkylation of amines; interaction between
cyclohexanone and aniline; reaction of cyclohexanone
with ethanolamine; interaction of 2-butanone with
ethanolamine; analysis of 2-butanone/phenylhydrazine
reaction mixture; high-throughput experimentation; and
references (PDF)
(8) Badu-Tawiah, A. K.; Li, A.; Jjunju, F. P. M.; Cooks, R. G. Angew.
Chem., Int. Ed. 2012, 51, 9417−9421.
(9) Lee, J. K.; Samanta, D.; Nam, H. G.; Zare, R. N. Nat. Commun.
2018, 9, 1562.
(10) Picotti, P.; Aebersold, R. Nat. Methods 2012, 9, 555−566.
(11) Zhao, Y.; Riley, N. M.; Sun, L.; Hebert, A. S.; Yan, X.;
Westphall, M. S.; Rush, M. J. P.; Zhu, G.; Champion, M. M.; Medie,
F. M.; DiGiuseppe Champion, P. A.; Coon, J. J.; Dovichi, N. J. Anal.
Chem. 2015, 87, 5422−5429.
(12) Fornelli, L.; Durbin, K. R.; Fellers, R. T.; Early, B. P.; Greer, J.
B.; LeDuc, R. D.; Compton, P. D.; Kelleher, N. L. J. Proteome Res.
2017, 16, 609−618.
(13) Park, J.; Piehowski, P. D.; Wilkins, C.; Zhou, M.; Mendoza, J.;
Fujimoto, G. M.; Gibbons, B. C.; Shaw, J. B.; Shen, Y.; Shukla, A. K.;
́
̆
́
Moore, R. J.; Liu, T.; Petyuk, V. A.; Tolic, N.; Pasa-Tolic, L.; Smith,
R. D.; Payne, S. H.; Kim, S. Nat. Methods 2017, 14, 909−914.
(14) Foreman, D. J.; McLuckey, S. A. Anal. Chem. 2020, 92, 252−
266.
(15) Poad, B. L.; Pham, H. T.; Thomas, M. C.; Nealon, J. R.;
Campbell, J. L.; Mitchell, T. W.; Blanksby, S. J. J. Am. Soc. Mass
Spectrom. 2010, 21, 1989−1999.
(16) Marshall, D. L.; Pham, H. T.; Bhujel, M.; Chin, J. S. R.; Yew, J.
Y.; Mori, K.; Mitchell, T. W.; Blanksby, S. J. Anal. Chem. 2016, 88,
2685−2692.
AUTHOR INFORMATION
■
Corresponding Author
Abraham K. Badu-Tawiah − Department of Chemistry and
Biochemistry, The Ohio State University, Columbus, Ohio
H
https://dx.doi.org/10.1021/acs.analchem.0c02960
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
pubs.acs.org/ac
Article
(17) Sheng, H.; Tang, W.; Yerabolu, R.; Max, J.; Kotha, R. R.;
(52) Kulyk, D. S.; Miller, C. F.; Badu-Tawiah, A. K. Anal. Chem.
2015, 87, 10988−10994.
̈
Riedeman, J. S.; Nash, J. J.; Zhang, M.; Kenttamaa, H. I. J. Org. Chem.
(53) Miller, C. F.; Kulyk, D. S.; Kim, J. W.; Badu-Tawiah, A. K.
Analyst 2017, 142, 2152−2160.
2016, 81, 575−586.
(18) Cernak, T. Chem 2016, 1, 6−9.
(54) Kalyanaraman, B.; Sinha, B. K. Environ. Health Perspect. 1985,
64, 179−184.
́
(19) Jeffries-El, M.; Sauve, G.; McCullough, R. D. Macromolecules
2005, 38, 10346−10352.
(55) Goodwin, D. C.; Aust, S. D.; Grover, T. A. Chem. Res. Toxicol.
1996, 9, 1333−1339.
(20) Xie, Y.; He, L.-F.; Lin, S.-C.; Su, H.-F.; Xie, S.-Y.; Huang, R.-B.;
Zheng, L.-S. J. Am. Soc. Mass Spectrom. 2011, 20, 2087−2092.
(21) Papanastasiou, M.; McMahon, A. W.; Allen, N. S.; Johnson, B.
W.; Keck-Antoine, K.; Santos, L.; Neumann, M. G. Int. J. Mass
Spectrom. 2008, 275, 45−54.
(56) Salvatore, R. N.; Yoon, C. H.; Jung, K. W. Tetrahedron 2001,
57, 7785−7811.
(22) Cabrera-Pardo, J. R.; Chai, D. I.; Liu, S.; Mrksich, M.; Kozmin,
S. A. Nat. Chem. 2013, 5, 423−427.
(23) Zhang, W.; Huang, G. Rapid Commun. Mass Spectrom. 2015,
29, 1947−1953.
(24) Kulyk, D. S.; Sahraeian, T.; Wan, Q.; Badu-Tawiah, A. K. Anal.
Chem. 2019, 91, 6790−6799.
(25) Kulyk, D. S.; Swiner, D. J.; Sahraeian, T.; Badu-Tawiah, A. K.
Anal. Chem. 2019, 91, 11562−11568.
(26) Kortshagen, U. R.; Sankaran, R. M.; Pereira, R. N.; Girshick, S.
L.; Wu, J. J.; Aydil, E. S. Chem. Rev. 2016, 116, 11061−11127.
(27) Wang, A.; Qin, M.; Guan, J.; Wang, L.; Guo, H.; Li, X.; Wang,
Y.; Prins, R.; Hu, Y. Angew. Chem., Int. Ed. 2008, 47, 6052−6054.
(28) Yi, Y.; Zhou, J.; Guo, H.; Zhao, J.; Su, J.; Wang, L.; Wang, X.;
Gong, W. Angew. Chem., Int. Ed. 2013, 52, 8446−8449.
(29) Chen, S.; Wan, Q.; Badu-Tawiah, A. K. Angew. Chem., Int. Ed.
2016, 128, 9491−9495.
(30) Jayaraj, S.; Badu-Tawiah, A. K. Sci. Rep. 2019, 9, 11280.
(31) Jin, C.; Viidanoja, J.; Li, M.; Zhang, Y.; Ikonen, E.; Root, A.;
̈
Romanczyk, M.; Manheim, J.; Dziekonski, E.; Kenttamaa, H. I. Anal.
Chem. 2016, 88, 10592−10598.
(32) Ding, X.; Duan, Y. Mass. Spectrom. Rev. 2015, 34, 449−473.
(33) Smoluch, M.; Mielczarek, P.; Silberring, J. Mass. Spectrom. Rev.
2016, 35, 22−34.
(34) Drechsel, E. J. Prakt. Chem. 1858, 38, 69.
(35) Borsche, W. Justus Liebigs Ann. Chem. 1908, 359, 49−80.
(36) Sekimoto, K.; Takayama, M. J. Inst. Electrostat. Jpn. 2009, 33,
38−42.
(37) Sekimoto, K.; Takayama, M. Eur. Phys. J. D 2010, 60, 589−599.
(38) Szwarc, M. Proc. R. Soc., A 1949, 198, 267−284.
(39) Marsh, B. M.; Iyer, K.; Cooks, R. G. J. Am. Soc. Mass Spectrom.
2019, 30, 2022−2030.
(40) Katritzky, A. R.; Mokrosz, J. L.; De Rose, M. J. Chem. Soc.
Perkin Trans. 2 1984, 5, 849−855.
(41) Katritzky, A. R.; Marson, C. M. Angew. Chem., Int. Ed. 1984, 23,
420−429.
(42) Badu-Tawiah, A. K.; Wu, C.; Cooks, R. G. Anal. Chem. 2011,
83, 2648−2654.
(43) Software for estimating vapor pressures (EPI Suite - Estimation
Program Interface v4.11) was provided by United States Environmental
Protection Agency. https://www.epa.gov/tsca-screening-tools/
download-epi-suitetm-estimation-program-interface-v411. Accessed
February 5, 2020.
(44) Badu-Tawiah, A. K.; Cyriac, J.; Cooks, R. G. J. Am. Soc. Mass
Spectrom. 2012, 23, 842−849.
(45) Sahraeian, T.; Kulyk, D. S.; Badu-Tawiah, A. K. Langmuir 2019,
35, 14451−14457.
(46) Yan, X.; Augusti, R.; Li, X.; Cooks, R. G. ChemPlusChem. 2013,
78, 1142−1148.
(47) Katritzky, A. R. Tetrahedron 1979, 36, 679−699.
(48) Veisi, H.; Azadbakht, R.; Ezadifar, M.; Hemmati, S. J.
Heterocycl. Chem. 2013, 50, E241−E246.
(49) Wang, B.; Gu, Y.; Luo, C.; Yang, T.; Yang, L.; Suo, J.
Tetrahedron Lett. 2004, 45, 3417−3419.
̈ ̈
̧ an, A.; Kulkarni, A.; Torok, B. Green Chem. 2012, 14, 17−
(50) Dast
37.
(51) Satyanarayana, V. S. V.; Sivakumar, A. Ultrason. Sonochem.
2011, 18, 917−922.
I
https://dx.doi.org/10.1021/acs.analchem.0c02960
Anal. Chem. XXXX, XXX, XXX−XXX