Measurement and optimization of organic chemical reaction yields by GC-MS with supersonic molecular beams

Article abstract of DOI:10.1016/j.tet.2012.05.031

A new type of gas chromatograph mass spectrometer (GC-MS) was used for semi-online monitoring of organic chemical reactions for their yield optimization, mechanism elucidation, and for obtaining information on the reaction products identity and purity. It was used with reaction mixtures without prior separation and purification as needed for NMR, thereby saving time and effort. Our unique GC-MS named 5975-SMB Supersonic GC-MS is based on GC interface with the MS with supersonic molecular beams (SMB) and on ionization of the sample molecules during their axial flight through an open electron ionization ion source as vibrationally cold molecules. GC-MS with SMB is demonstrated to significantly extend the range of compounds amenable for analysis, practically always giving molecular ions, enabling effective fast GC-MS analysis, and providing elemental formulas via isotope abundance analysis with unit mass resolution quadrupole MS. In addition, it uniquely provides uniform response to all compounds, a feature, which is vital for the measurement of chemical reaction yields. In this manuscript, four different organic synthetic reactions were studied and are described. Based on the collected data, we were able to better understand how the reaction conditions should be optimized in order to maximize the yields and purity of target products. Consequently, we propose that GC-MS with SMB can serve as a novel tool for the fast optimization of chemical reactions.

Full text of DOI:10.1016/j.tet.2012.05.031

Tetrahedron 68 (2012) 5793e5799
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Measurement and optimization of organic chemical reaction yields by GCeMS
with supersonic molecular beams
*
Aviv Amirav , Alexander Gordin, Youlia Hagooly, Shlomo Rozen, Bogdan Belgorodsky, Boaz Seemann,
Hanit Marom, Michael Gozin, Alexander B. Fialkov
School of Chemistry, Tel Aviv University, Tel Aviv 69978, Israel
a r t i c l e i n f o
a b s t r a c t
Article history:
A new type of gas chromatograph mass spectrometer (GCeMS) was used for semi-online monitoring of
organic chemical reactions for their yield optimization, mechanism elucidation, and for obtaining in-
formation on the reaction products identity and purity. It was used with reaction mixtures without prior
separation and purification as needed for NMR, thereby saving time and effort. Our unique GCeMS
named 5975-SMB Supersonic GCeMS is based on GC interface with the MS with supersonic molecular
beams (SMB) and on ionization of the sample molecules during their axial flight through an open
electron ionization ion source as vibrationally cold molecules. GCeMS with SMB is demonstrated to
significantly extend the range of compounds amenable for analysis, practically always giving molecular
ions, enabling effective fast GCeMS analysis, and providing elemental formulas via isotope abundance
analysis with unit mass resolution quadrupole MS. In addition, it uniquely provides uniform response to
all compounds, a feature, which is vital for the measurement of chemical reaction yields. In this man-
uscript, four different organic synthetic reactions were studied and are described. Based on the collected
data, we were able to better understand how the reaction conditions should be optimized in order to
maximize the yields and purity of target products. Consequently, we propose that GCeMS with SMB can
serve as a novel tool for the fast optimization of chemical reactions.
Received 1 January 2012
Received in revised form 27 April 2012
Accepted 8 May 2012
Available online 16 May 2012
Keywords:
Reaction yield
Reaction monitoring
Reaction optimization
GCeMS with supersonic molecular beams
Organic mass spectrometry
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
yield, products purity, availability of isomers and on the reaction
mechanism, since the already-purified compounds are analyzed
Chemical reactions, and particularly organic ones, are typically
performed in one of several well-established methods, usually on
a small scale. A typical approach involves mixing the reactants
together in an appropriate solvent, often with addition of a cata-
lyst, and allowing them to undergo a reaction, which may take
from several minutes to few days. The progress of the reaction is
either assumed or monitored most-commonly by thin layer
chromatography and by NMR, but little or no information is
obtained online on the reaction yield or its actual progress. At the
perceived end of the reaction, the products are separated and
purified by preparative chromatography, distillation, sublimation,
selective precipitation or crystallization, processes, which may
take several hours or even several days. Subsequently, the purified
products are analyzed by 1H and 13C NMR and by high resolution
mass spectrometry (often Quadrupole Time of Flight (QTOF) mass
spectrometer), typically via flow injection electrospray ionization
(FI-ESI). However, while FI-ESI-QTOF provides elemental formulas
information, it does not provide information about the synthesis
and since ESI has non-uniform, highly compound dependent
ionization yields. In contrast, GCeMS with its standard electron
ionization (EI), unlike ESI, has for volatile compounds approxi-
mately uniform, semi-quantitative ionization yield. However, it
provides molecular ions only for about 50% of the compounds
having molecular weights above 300 amu (our statistical analysis
of NIST library) and standard GCeMS is limited to volatile and
thermally stable compounds only, which severely restricts its
applicability. Furthermore, ion source degradation and peak-tail-
ing reduce the standard EI response uniformity for semi-volatile
compounds.
Thus, in order to measure and better optimize chemical re-
action yields, new methods and technologies are needed that will
enable online, or almost online, monitoring of the reacting com-
pounds and products. Such technology should be applicable to
broad range of compounds, enable sample quantitative de-
termination via uniform compound independent response, enable
positive sample identification and be compatible with sample
separation devices.
In the last two decades our group research has been focused on
the development of GCeMS with supersonic molecular beams
* Corresponding author. E-mail address: amirav@tau.ac.il (A. Amirav).
0040-4020/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2012.05.031

5794
A. Amirav et al. / Tetrahedron 68 (2012) 5793e5799
(SMB) (also named Supersonic GCeMS).1e6 Supersonic GCeMS is
based on a GC and MS interface with SMB and on the electron
ionization of vibrationally cold analytes in the SMB (cold EI) in
a fly-through ion source. This ion source is inherently inert and
further characterized by an ultra fast response time and vacuum
background filtration capability.1,7 The same ion source also offers
a mode of classical EI.8 Cold EI, as a main mode, provides trust-
worthy enhanced molecular ions combined with effective library
sample identification, that is, supplemented and complemented by
a powerful isotope abundance analysis method and software,9
which provides elemental formulas with unit resolution quadru-
pole MS data. In addition, the range of low volatility and thermally
labile compounds amenable for analysis is significantly increased
(about doubled) with Supersonic GCeMS due to the use of contact-
free fly-through ion source and the ability to lower sample elution
temperatures through the use of short GC columns with high
carrier gas flow rates.5 This compatibility with high column flow
rates further facilitates effective fast and ultra fast GCeMS ana-
lysis,1e3,10 which was also explored in its use in GCeMS of reaction
products.6
The basic requirement from a system that should measure
reaction yields is that its response will be uniform and
compound independent or the response should represent in
a known and systematic manner the compounds molecular weights
and amounts.
Electron ionization, unlike any other ionization method,
uniquely provides ionization yields that depend predominantly on
the sum of the numbers of each different atom in the ionized
analyte11e13 due to the localized nature of electroneelectron in-
teraction in the ionization process. In simple terms, the electron
ionization cross section approximately depends on the number of
electrons in the molecule, which relates to its molecular weight.
Thus, the electron ionization probability for organic compounds
approximately linearly increases with the ionized compound’s
molecular weight hence it is approximately proportional to the
sample weight. Consequently, it is well known that standard
GCeMS exhibits uniform response to all volatile compounds found
in analyzed samples.
For GCeMS analysis, that is, based on a quadrupole mass
spectrometry, the ionization yield is further reduced by the mass
analyzer and ion detector transmission, hence its response slightly
declines with mass. However, for semi-volatile and low volatility
compounds, peak-tailing and ion source degradation can signifi-
cantly reduce the observed signal, sometimes in a difficult to
predict way, thus the feature of uniform response of standard
GCeMS is eroded and often does not exist anymore. In contrast, in
GCeMS with SMB analysis, ion source-related peak-tailing and
degradation are completely eliminated, due to the use of a fly-
through ion source, thus providing practically uniform com-
pound independent response for all analytes that enter the SMB.
The implication of such uniform response is that in a given chro-
matogram, the peak area relates to the amount (weight) of the
analyte.
Fig. 1. The analysis of a mixture of n-C16H34, methyl stearate, cholesterol and n-C32H66
at about 0.25 ng on-column amount with standard GCeMS (upper trace) and
5975-SMB Supersonic GCeMS (bottom trace). 30 m, 0.25 mm ID capillary column with
0.25
m DB-5ms UI column was used with the standard GCeMS with 1.2 mL/min column
flow rate while only 5 m of that column was used with 5 mL/min column flow rate
with the Supersonic GCeMS. The GC oven program was 80 ꢀC followed by 10 ꢀC/min to
320 ꢀC for the standard 5975 GCeMS and 50 ꢀC followed by 30 ꢀC/min to 300 ꢀC for
the 5975-SMB.
EI mass spectra while for standard EI it was either weak or absent
in the case of n-C32H66
.
Thus, the uniform response of GCeMS with SMB can be as-
sumed and routinely used for semi-quantitative analysis. Further-
more, a simple chromatogram with few compounds at various
molecular weights could be utilized for obtaining more accurate
calibration of the cold EI system response for closer accounting of
small experimental mass dependent effects on the response factors
dependence on mass.
In this work we describe the use of GCeMS with SMB for the
semi-online monitoring of organic reactions (fractions are taken
during the reaction time and progress), while optimizing these
reactions yields, obtaining information on their mechanisms and
determining products purity and identity, including in some cases
information regarding existence and possible structures of formed
isomers.
In Fig. 1, the uniform response of our 5975-SMB Supersonic
GCeMS is demonstrated via the comparison of two chromato-
grams of the same test mixture that included n-C16H34, methyl
stearate, cholesterol and n-C32H66, each at 0.25 ng on-column
amount (n-C32H66 amount is somewhat lower). As clearly ob-
served, the response of standard EI is declining with the sample
volatility and mass, and already for cholesterol its relative re-
sponse has declined by more than an order of magnitude in the
total ion count scale (TIC) (factor of 10.2 in relative peak area). In
contrast, the response of 5975-SMB remained unaffected by the
sample volatility due to the use of contact-free fly-through ion
source and it is very similar to that of GC-FID. In addition, all the
four sample compounds had dominant molecular ion in their cold
2. Experimental: the Supersonic GCeMS systems
2.1. General
Two types of GCeMS with Supersonic Molecular Beams (SMB)
(Supersonic GCeMS) were used. The first system was based on the
conversion of a Varian 3800 GC plus 1200 triple quadrupole MS
(Varian Walnut Creek CA USA) into a Supersonic GCeMS as de-
scribed in details in Ref. 4. The second system was a recently de-
veloped Supersonic GCeMS based on the conversion of an Agilent
7890 GCþ5975 MSD (Agilent Technologies, Santa Clara CA USA)
into
a 5975-SMB Supersonic GCeMS (Aviv Analytical, Hod
Hasharon Israel).

A. Amirav et al. / Tetrahedron 68 (2012) 5793e5799
5795
GCeMS with SMB is based on the use of an SMB for interfacing
the GC to the MS and as a medium for electron ionization of
sample compounds.1 SMBs are characterized by intra-molecular
vibrational super-cooling, unidirectional molecular motion with
controlled hyperthermal kinetic energy (1e20 eV), mass focusing
similar to that in a jet separator, and capability of handling very
broad range of column flow rates from standard 1 mL/min (or
lower) up to 100 mL/min, without affecting the sensitivity. In
these systems the column output is mixed with helium make-up
gas (w70 mL/min typical total flow rate), and flows to the su-
personic nozzle through a heated and temperature-controlled
transfer line. The helium flow can be mixed (via the opening of
one valve) with perfluorotributylamine (PFTBA) for periodic sys-
tem tuning and calibration. The sample compounds seeded in the
helium gas expand from a 110 mm diameter supersonic nozzle into
a nozzle vacuum chamber, that is, differentially pumped by
a Varian Navigator 301 turbomolecular pump (Varian Inc., Torino
Italy) with 250 L/s pumping speed. The helium pressure at this
vacuum chamber is about 6ꢁ10ꢂ3 mbar. The supersonic expansion
vibrationally cools the sample compounds and the expanded su-
personic free-jet is skimmed by a 0.8 mm skimmer and collimated
in a second differentially pumped vacuum chamber, where an
SMB is formed. This second vacuum chamber is pumped by the
original GCeMS turbomolecular pump. The SMB, seeded with
vibrationally cold sample compounds, flies through a dual cage EI
ion source7 where these beam species are ionized by 70 eV elec-
trons with typically 8 mA emission current. The ions are focused
by an ion lens system, deflected 90ꢀ by an ion-mirror and enter the
original mass analyzer. The 90ꢀ ion-mirror is separately heated
and serves to keep the mass analyzers clean from sample induced
contaminations and for the suppression of helium meta-stable
induced mass independent noise. Data analysis is performed with
the original GCeMS data analysis software (Chemstation for the
5975-SMB) plus the Tal-Aviv Isotope Abundance Analysis (IAA)
software (Aviv Analytical, Hod Hasharon, Israel). For ultra fast GC
separation, a newly developed low thermal mass fast GC was
employed, as described in details in Ref. 10. It is based on the use
of a capillary GC column, that is, inserted into resistively heated
thin stainless steel tubing, which is mounted on the top of a Varian
3800 GC.
Fig. 2. The analysis of synthetic organic chemical reaction products (reaction mixture)
with 5975-SMB Supersonic GCeMS. C14H18N2O3S2 was the target product and it is the
main total ion chromatogram peak (upper trace) whose cold EI mass spectrum is
shown at the bottom trace. A 5 m column with 0.25 mm ID and 0.25
m RTx-5MSI film
was used with 16 mL/min helium column flow rate and 40 ꢀC/min GC temperature
programming rate.
with the fluorinating agent, bromine trifluoride, led to the for-
mation of fluorinated product N-butyl-4-nitro-N-(trifluoromethyl)
benzamide, a rare compound with a potential of antifungal prop-
erties and as useful electrolyte additives for storage batteries. The
synthesis of this ethyl butyl(4-nitrophenylcarbonyl)carbamodi-
thioate was unknown but the use of ethanethiol and coupling
reaction with 4-nitrobenzoic acid led to the desired product
(shown below in Fig. 3) that was identified by the Supersonic
GCeMS.
The ethyl butyl(4-nitrophenylcarbonyl)carbamodithioate (tar-
get compound) has a molecular ion m/z¼326.2, and isotope
abundance analysis using the Tal-Aviv IAA software showed
C14H18N2O3S2 to be number one in the list of possible elemental
formulas with a matching factor of 999 (out of 1000). The IAA
software provided further information that all first 20 hits had two
sulfur atoms, at least one oxygen atom and 14ꢃ2 carbon atoms.
This elemental formula was later further confirmed with accurate
mass ESI-QTOF and by NMR. However, unique to our approach was
the fact that the 5975-SMB Supersonic GCeMS provided valuable
information on the reaction yield, prior to sample purification,
which related to the target compound peak area ratio to all other
peaks. We measured this reaction yield as 59ꢃ4%. In addition, an
undesirable product with molecular ion m/z¼332.2 was identified
and information on the reactants stability and their residual
amounts was revealed, thereby enabling reaction conditions op-
timization prior to target compound purification.14
Several columns and variable experimental conditions were
used as described in the figure captions. In all cases short columns
were used from 1.7 to 15 m with helium column flow rates in the
5e32 mL/min range.
3. Results
In Fig. 2 typical analysis of a chemical reaction mixture is
shown, as obtained with the 5975-SMB Supersonic GCeMS. The
investigated reaction was a simple AþB/CþD type, aimed at
obtaining an intermediate product. As shown, four main GCeMS
peaks are observed and the last one that eluted at 3.7 min (Fast
GCeMS) was the synthesis target compound as initially perceived
by its molecular ion m/z¼326.2. (While the 5975 quadruple mass
spectrometer had unit resolution its mass accuracy was ꢃ 0.1 amu,
which further helped in identification). The reaction of this target
molecule (ethyl butyl(4-nitrophenylcarbonyl)carbamodithioate)
Fig. 3. Reagents and conditions: (i) EtSH, THF, rt, 20 h; (ii) DCC, DMAP, CH2Cl2, rt, 18 h.

5796
A. Amirav et al. / Tetrahedron 68 (2012) 5793e5799
Figs. 4 and 5 show the monitoring and optimization of another
peroxides, since the fluorescence spectrum of the resulting oxida-
tion product sulfone SA2, is significantly different than the spec-
trum of the starting thioether SA1.15 For oxidation of SA1 to the
corresponding sulfone SA2, MoO2Cl2(OPPh3)2 complex was used as
a catalyst and H2O2, as the re-oxidant.
organic reaction, based on analysis by the 5975-SMB Supersonic
GCeMS. In this reaction, thioether SA1 was catalytically oxidized
to a corresponding sulfone SA2. The fluorescent thioether SA1
was developed as a potential chemosensor for the detection of
The transformation from SA1 to SA2 was found to be problem-
atic and we had difficulties determining the correct yield for this
reaction. A complex mass spectrum, with little information content,
was obtained by the FI-ESI-QTOF technique for the isolated SA2
product. Moreover, based on one of the many peaks in this mass
spectrum, oxidation of thioether SA1 to a corresponding sulfoxide
could be easily assumed with a mass spectral peak presumably
assigned via its exact mass to an ion containing adduct potassium
ion in its composition. This assumption was found to be erroneous
by the 5975-SMB Supersonic GCeMS analysis. As shown in the
upper chromatogram of Fig. 5, the target sulfone SA2 emerged as
a relatively small peak at near 6 min elution time, whose area re-
lated to its yield while other non target compounds were the
majority.
Fig. 4. Synthesis of compound SA2. Reagents and conditions: (i) MoO2Cl2(OPPh3)2
catalyst, H2O2, CH3CN, 60 ꢀC, 24 h.
The reaction conditions were optimized based on Supersonic
GCeMS data, and as shown in the middle chromatogram of the
Fig. 5, compound SA2 was purified, and its cold EI mass spectrum
(bottom spectrum, Fig. 5) was dominated by its molecular ion (m/z
435.2), while isotope abundance analysis of this molecular ion
resulted in C25H25NO4S composition at the top of the list of all
possible elemental formulas. The identity of compound SA2 was
later further confirmed by FI-ESI-QTOF and 13C NMR, using chro-
matographically pure material. However, the ESI-QTOF mass
spectrum still contained several peaks and was void of any quan-
titative information. The 5975-SMB Supersonic GCeMS on the
other hand, as shown in Fig. 5, provided quantitative information
on the compound SA2 purity and uniquely indicated that the two
small peaks, eluted just after the main peak (one on the main peak
shoulder and second at w6.3 min), were assigned as isomers of this
material since they exhibited similar mass spectra with small var-
iations of the relative mass spectral fragment intensities. The ad-
ditional compound that eluted at around
7 min contained
a bromine atom in its structure that most probably originated from
an impurity in the starting SA1. Fig. 5 clearly shows how the 5975-
SMB Supersonic GCeMS analysis of a reaction mixture enabled its
optimization and resulted in quantitative information on the
product purity and prevailing isomers.
Another example of a chemical reaction yield optimization
by using the 5975-SMB Supersonic GCeMS is shown in Figs. 6
and 7. In this case, the synthetic target compound was a bio-
degradable and potentially reusable derivative of the S,S-ethyl-
enediaminedisuccinic tetraacid (N,N0-bis-n-dodecyl-S,S-EDDS X5;
Fig. 6). This chelating agent was specifically designed for im-
provement of a Sediments Remediation Phase Transition Extraction
(SR-PTE) process developed by Brauner and Ullmann for extraction
of heavy metals from contaminated soils.16 The structure of the
chelator X5 was based on S,S-ethylenediaminedisuccinic tetraacid
(S,S-EDDS, Fig. 6) and contained two dodecyl-alkyl chains to pro-
mote an extraction of heavy metal complexes of this ligand into
recyclable organic media.
Despite all our efforts, the direct alkylation of S,S-EDDS was
found to be unattainable. Thus, the alternative synthetic strategy
for preparation of the desired chelator X5 included conversion of
S,S-EDDS into a corresponding tetra-butylester X1 (C26H48N2O8; m/
z 516.3), which was subsequently subjected to bis-N,N0-alkylation,
giving the key intermediate X2 (C50H96N2O8; m/z 852.7) that was
finally hydrolyzed into the target tetraacid. Among all of the eval-
uated alcohols, used for conversion of the S,S-EDDS into the cor-
responding tetraester derivatives, utilizing n-butanol, as a reagent
and as a reaction solvent, allowed us to achieve the best esterifi-
cation yields. Moreover, we found that alkaline hydrolysis of the
Fig. 5. The analysis of synthetic organic chemical reaction products with 5975-SMB
Supersonic GCeMS. SA2 (C25H25NO4S) was the target product and it is the minor
last to elute peak in the upper mass chromatogram and main peak in the middle trace
of the finally purified product. The cold EI mass spectrum of this compound is shown at
the bottom trace. The small peak after the main peak in the middle mass chromato-
gram is of its isomer, while the last to elute peaks are of impurities. A 15 m column
with 0.32 mm ID and 0.1
m DB-1HT film was used with 32 mL/min helium column flow
rate and 30 ꢀC/min GC oven temperature programming rate.

A. Amirav et al. / Tetrahedron 68 (2012) 5793e5799
5797
Fig. 6. Synthesis of X5. Reagents and conditions: (i) SOCl2, n-BuOH, 0/60 ꢀC, 24 h; (ii) dodecylbromide, K2CO3, CH3CN, 80 ꢀC, 48 h; (iii) KOH (aq), THF, RT, 24 h.
N,N0-bis-alkylated tetra-butylester X2 provided almost quantitative
yields of the target chelator X5.
5975-SMB Supersonic GCeMS analysis showed that at optimal
conditions the yield of the X2 is 18% (middle chromatogram, Fig. 7).
After compound X2 separation and purification by liquid chroma-
tography, the reaction yield was determined gravimetrically and
was found to be 18.6%, which is in agreement with the 5975-SMB
Supersonic GCeMS results (18ꢃ3%). Thus, Fig. 7 demonstrates an
exceptional ability of the 5975-SMB Supersonic GCeMS to analyze
an extended range of compounds that are well beyond the limits of
standard GCeMS. In the 5975-SMB Supersonic GCeMS, compounds
in the N,N0-bis-alkylation reaction mixture not only eluted, but also
had no ion source peak-tailing and/or ion suppression effects, en-
abling correct determination of this reaction yield. Moreover, cold
EI mass spectrum (bottom mass spectrum, Fig. 7) of compound X2
provided a good molecular ion peak, despite the presence of two
dodecyl chains, which normally preclude the observation of any
molecular ion in standard EI, such as with large phthalate esters.
The optimization of the above presented process could not be
performed without the 5975-SMB Supersonic GCeMS analysis,
thereby clearly demonstrating that this technique can serve as
a powerful tool for reaction yield optimization.
Sometimes synthetic organic reactions can be fast, hence the
obvious next step was to explore the use of ultra fast 5975-SMB
Supersonic GCeMS for semi-online monitoring of the progress of
fast synthetic organic chemical reactions. We were involved in
exploring new ways for selective oxidation of organic compounds
at their thioester function. The progress monitoring of sulfur oxi-
dation reaction is demonstrated in Fig. 8, as probed by our low
thermal mass fast GCeMS with SMB (1200-SMB Supersonic
GCeMS). The organic chemical reaction of the thioester with meta-
chloroperbenzoic acid (m-CPBA) is shown vs the added amount of
oxidizing agent. The upper trace demonstrates the TIC of the mix-
ture from the reactor with two times initial excess of the oxidizing
agent and the MS of the last peak in the insert, which is identified as
a singly oxidized product. It shows the remaining organic reactant
plus the appearance of two isomers of the singly oxidized product.
Correspondently, the middle trace was obtained at 3.5 times initial
excess of the oxidizing agent. The mass spectrum proves that at
least two doubly oxidized product isomers are formed. Finally, the
lower trace was obtained with five times excess of the oxidizing
agent that yielded triply oxidized product, which was the target of
this synthesis. The insert shows the mass spectrum of that desirable
triply oxidized product. As shown, each analysis took only 1.5 min.
However, significant challenges in the synthesis, purification,
and analysis of the key intermediate X2 were encountered
during development of the second step of the process. Under
many reaction conditions, formation of substantial amounts of
mono-N-alkylated tetraester X3 (C36H72N2O8; m/z 648.5) and
lactam X4 (C34H62N2O7; m/z 610.5) side-products were observed
as measured with the 5975-SMB (Fig. 7 below). The latter
compound is most probably formed in reaction mixtures by
an intra-molecular amidation. Only traces of quaternary am-
monium derivatives were formed, mostly due to the steric hin-
drance around the tertiary amines of compounds X2 and X3.
These traces were easily removed by
chromatography.
a preparative liquid
In attempt to find, which conditions, including various solvents,
alkylating agents, bases, reagent concentrations and temperature
regimes, would lead to a minimum amount of side-products, all
evaluated reaction mixtures were monitored by the 5975-SMB
Supersonic GCeMS, since standard GCeMS was incapable of ana-
lyzing our relatively large and polar starting materials and products
while FI-ESI-QTOF is incapable of providing any quantitative in-
formation on the reaction progress.
For example, under many non-optimized conditions, the target
intermediate X2 was assumed to be absent by all other techniques
but could be observed at about 0.4% yield (retention time of 7.7 min
in Fig. 7), even after prolonged heating (5 days) of the reaction
mixture (upper chromatogram, Fig. 7). It should be noted that de-
spite concerns regarding our compounds stability under prolonged
heating, we still found significant amounts of unreacted starting
materials, as well as formed side-products, while no products of
decomposition could be detected.
By screening a large number of possible parameters, we were
able to optimize our reaction conditions to the lowest amount of
side-products, especially of the lactam X4 (which is a ‘dead-end’
side-product), and obtain the highest yield of the desired com-
pound X2. We found that the best reaction conditions for N,N0-bis-
alkylation of the tetraester X1 included use of n-dodecylbromide as
an alkylating agent, CH3CN as a solvent and K2CO3 as a base. A
temperature of 80 ꢀC as well as specific concentrations of the
starting materials were also very important for the N,N0-bis-alkyl-
ation reaction outcome.

5798
A. Amirav et al. / Tetrahedron 68 (2012) 5793e5799
Fig. 8. Fast GCeMS with SMB monitoring of the progress and yield of synthetic organic
chemical reaction with the 1200-SMB Supersonic GCeMS. The sample was injected
into a Varian 1177 injector with split 20 at 200 ꢀC; and 3800 standard GC oven tem-
perature at 220 ꢀC. The short (1.5 m) fast GC column flow rate was 8 mL min.
Fig. 7. The analysis of chemical mixture by the 5975-SMB Supersonic GCeMS. X5
(C50H96N2O8 (m/z¼852.8)) is the target reaction product and it is the last to elute total
ion chromatogram peak (upper and middle mass chromatograms) whose cold EI mass
spectrum is shown in the bottom trace. The three main peaks before it are of a reaction
‘dead-end’ lactam plus co-eluting partial substitution product and the two reactants as
can benefit from online or semi-online reaction monitoring for its
improved optimization. Among the currently used tools, standard
GCeMS is very limited in its applicability (range of compounds),
often does not provide molecular ions and due to ion source peak-
tailing does not provide fully uniform compound independent peak
areas. In addition, standard GCeMS analysis can take up to 40 min.
LCeMS requires lengthy LC method development time hence is not
often used and electrospray has highly compound dependent ion-
ization yield (plus sometimes ion suppression effects) hence cannot
be used for the monitoring of reaction progress or for the provision
of reaction yields. NMR is incompatible with the analysis of most
mixtures and requires deuterated solvents, hence usually requires
compound purification at the mg amount. Thus, currently TLC is
often being used with its limited information content for the
monitoring of reaction progress.
In this paper we demonstrated the use of the Supersonic
GCeMS for the monitoring and optimization of synthetic organic
chemical reactions. Our main goal was to advance a new approach
for the performance of organic synthesis that will by-pass the
lengthy steps of synthetic target compound separation and purifi-
cation for NMR analysis. According to our approach, the synthetic
mixture is analyzed semi-online for the provision of quantitative
content information for the purpose of reaction optimization, while
leaving the steps of separation and purification to the end after
reaction optimization.
indicated in the figure. A 15 m column with 0.32 mm ID and 0.1
m DB-1HT film was
used with 32 mL/min helium column flow rate and 30 ꢀC/min GC temperature pro-
gramming rate.
The performance of such fast semi-online analysis provided two
surprising results for the synthetic chemists: (A) it was found that
the reaction took about 5 min to be completed as opposed to 3e5 h
perceived rate, which led only to product degradation; (B) it was
found that the oxidizing reactant was partially decomposed during
the reaction, hence 5 equiv were needed for attaining almost full
yield of three oxygen atoms transfer. Thus, semi-online fast GCeMS
with SMB monitoring enabled careful ‘titration’ of the amount of
oxidizing reagent needed and the fine tuning of optimal yield in
a small fraction of the time and effort, that is, usually needed. We
note that these labile compounds are not amenable for either
standard GCeMS (degradation) or ESI-MS (poor ionization yield)
but are compatible as shown for analysis by the 1200-SMB Super-
sonic GCeMS.
4. Discussion and conclusions
Organic chemistry and other types of synthetic chemistry, such
as the synthesis of drugs, industrial chemicals and fine chemicals

A. Amirav et al. / Tetrahedron 68 (2012) 5793e5799
5799
The Supersonic GCeMS was demonstrated as uniquely char-
acterized by: (A) having significantly extended range of
compounds amenable for analysis due to its use of high column
flow rates, short columns and fully inert ion source; (B) having
uniform response with cold EI without any ion source peak-
tailing; (C) provision of trustworthy enhanced molecular ions
to the vast majority of all compounds; (D) being compatible
with isotope abundance analysis for the provision of elemental
formulas; (E) compatibility with fast and ultra fast GCeMS
analysis.
Another important side benefit of semi-online reaction moni-
toring is the ability to work in micro scale due to the elimination of
the need to purify mg amounts as needed for NMR. As a result,
micro chemical reaction reactors can be used for the optimization
of the reaction yield with significant time, chemicals and solvent
saving (also safer), and only at the end larger amount can be used as
needed for the final product.
Award No. US-4273-09 from BARD, the United StateseIsrael Bi-
national Agricultural Research and Development Fund, by the
James Franck Center for Laser Matter Interaction Research and by
the Infrastructure Research Program of the Ministry of Science and
Technology (MOST) of the state of Israel.
References and notes
1. Amirav, A.; Gordin, A.; Poliak, M.; Fialkov, A. B. J. Mass Spectrom. 2008, 43,
141e163.
2. Dagan, S.; Amirav, A. Int. J. Mass Spectrom. Ion Processes 1994, 133, 187e210.
3. Amirav, A.; Gordin, A.; Tzanani, N. Rapid Commun. Mass Spectrom. 2001, 15,
811e820.
4. Fialkov, A. B.; Steiner, U.; Jones, L.; Amirav, A. Int. J. Mass. Spectrom. 2006, 251,
47e58.
5. Fialkov, A. B.; Gordin, A.; Amirav, A. J. Chromatogr. A. 2003, 991, 217e240.
6. Fialkov, A. B.; Amirav, A. J. Chromatogr., A 2004, 1058, 233e242.
7. Amirav, A.; Fialkov, A. B.; Gordin, A. Rev. Sci. Instrum. 2002, 73, 2872e2876.
8. Gordin, A.; Amirav, A.; Fialkov, A. B. Rapid Commun. Mass Spectrom. 2008, 22,
2660e2666.
9. Alon, T.; Amirav, A. Rapid Commun. Mass Spectrom. 2006, 20, 2579e2588.
10. Fialkov, A. B.; Morag, M.; Amirav, A. J. Chromatogr. A. 2011, 1218, 9375e9383.
11. Electron Impact Ionization; Mark, T. D., Dunn, G. H., Eds.; Springer: New York, NY,
1985.
We feel that when combined these benefits can provide a new
approach for improved synthetic chemistry.
12. Bobeldijk, M.; van der Zande, W. J.; Kistemaker, P. G. Chem. Phys. 1994, 179,
125e130.
Acknowledgements
13. Beran, J. A.; Kevan, L. J. Phys. Chem. 1969, 73, 3866e3876.
14. Hagooly, Y.; Gatenyo, J.; Hagooly, A.; Rozen, S. J. Org. Chem. 2009, 74, 8578e8582.
15. Marom, H.; Popowski, Y.; Antonov, S.; Gozin, M. Org. Lett. 2011, 13, 5532e5535.
16. Ludmer, Z.; Golan, T.; Ermolenko, E.; Brauner, N.; Ullmann, A. Environ. Eng. Sci.
2009, 26, 419e430.
This research was supported by the Israel Science Foundation
funded by the Israel Academy of Sciences and Humanities (grant
No. 393/11). This research was also supported by a Research Grant