Continuous Processing and Efficient in Situ Reaction Monitoring of a Hypervalent Iodine(III) Mediated Cyclopropanation Using Benchtop NMR Spectroscopy

Article abstract of DOI:10.1021/acs.oprd.6b00177

Real-time NMR spectroscopy has proven to be a rapid and an effective monitoring tool to study the hypervalent iodine(III) mediated cyclopropanation. With the ever increasing number of new synthetic methods for carbon-carbon bond formation, the NMR in situ monitoring of reactions is becoming a highly desirable enabling method. In this study, we have demonstrated the versatility of benchtop NMR using inline and online real-time monitoring methods to access mutually complementary information for process understanding, and we developed new approaches for real-time monitoring addressing challenges associated with better integration into continuous processes.

Full text of DOI:10.1021/acs.oprd.6b00177

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Full Paper
Continuous Processing and Efficient In Situ Reaction
Monitoring of a Hypervalent Iodine (III) Mediated
Cyclopropanation using Benchtop NMR Spectroscopy
Batool Ahmed-Omer, Eric Sliwinski, John Paul Cerroti, and Steven V Ley
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.6b00177 • Publication Date (Web): 10 Aug 2016
Downloaded from http://pubs.acs.org on August 16, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Organic Process Research & Development is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 39
Organic Process Research & Development
1
2
3
4
5
6
Organic Process Research & Development 2016, xx, xxx-xxx
7
8
9
Continuous Processing and Efficient in Situ Reaction
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Monitoring of a Hypervalent Iodine (III) Mediated
Cyclopropanation
Spectroscopy
using
Benchtop
NMR
Batool Ahmed-Omer,*,† Eric Sliwinski, † John P. Cerroti,fl and Steven V. Ley†
† Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW, UK.
fl Magritek GmbH, Gebäude VO (Building VO), Triwo Technopark Aachen, Philipsstrasse 8,
52068 Aachen, Germany.
1
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 2 of 39
1
2
3
4
5
6
TOC Graphic
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 3 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
ABSTRACT: Real-time NMR spectroscopy has proven to be a rapid and an effective monitoring
tool to study the hypervalent iodine (III) mediated cyclopropanation. With the ever increasing
number of new synthetic methods for carbon-carbon bond formation, the NMR in situ monitoring
of reactions is becoming a highly desirable enabling method. In this study, we have demonstrated
the versatility of benchtop NMR using inline and online real-time monitoring methods to access
mutually complementary information for process understanding, and developed new approaches
for real-time monitoring addressing challenges associated with better integration into continuous
processes.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
KEYWORDS: Benchtop NMR spectroscopy; in Situ reaction monitoring; Flow process;
Cyclopropanation; Solvent switching process; Compact Mass spectroscopy.
INTRODUCTION: The understanding, quality and control of chemical processes is of vital
importance from early discovery through to process optimization and scale-up onto full scale
manufacturing. Chemical manufacturing, in fact, dictates the need for significant resources and
effort to be invested in processes knowledge at an early stage to increase business efficiency.1
Similarly, the Quality-by-Design approach to chemical synthesis relies on the knowledge acquired
by process monitoring for control and risk minimization at source.2 This is best achieved by
monitoring the chemical processes using a basis of real-time analytics. Similarly, integrating real-
time analysis is fundamental to delivering the green chemistry agenda.3 In this respect, the
application of Process Analytical Technologies (PAT) is an established approach for compliance
across a number of industries.4 Furthermore, the reliance on appropriate technologies for efficient
data processing in real-time becomes ever more apparent. In fact, the acquisition of large data is
of little value without suitable tools to extract the crucial knowledge. Hence multivariate data
3
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 4 of 39
1
2
3
4
5
6
7
8
processing along with other statistical tools and intelligent algorithms are already being applied to
data analysis.5 On the other hand, the —Internet of Things“ is a revolution already underway, based
on a vision in which every element of a process is connected via network, allowing continuous
data exchange in real time. This entails significant revenue opportunities for research and process
manufacturing integrated with real-time analytics, an area which is expected to significantly
impact and grow over the following years. Approaches based on continuous flow process have
addressed a number of inherent synthesis challenges over the decades, owing to its well-
established benefits and amenability towards automation. Likewise, the potential for new
understanding of reaction processes can be realized when fully integrated with appropriate
analytics. Recent studies, in fact, have shown the application of flow technologies with real-time
analysis offering an insight into intelligent reaction optimization.6 In the recent years, the
improvements in analytical technologies by achieving enhanced sensitivity, improved acquisition
and sampling techniques has progressively presented a greater efficiency.Techniques such as mass
spectrometry and infra-red spectroscopy, are increasingly being applied for in situ reaction
monitoring.7 Among these methods, the inherent versatility of Nuclear Magnetic Resonance
(NMR) spectroscopy enables simultaneous structure identification and quantification in process
understanding, therefore the demand for rapid reaction monitoring using NMR spectrometers is
significantly on the increase.8 The recent development of compact NMR spectroscopy, based on
high-homogeneity permanent magnet designs, has enabled NMR to emerge as a convenient
analytical and quality control tool with improved sensitivity for process study of both batch and
flow reactions.9 In the work reported below we demonstrate the potential and reliability of
benchtop NMR as an in situ monitoring tool for continuous processing employing inline and online
methods. In the study, we have developed a continuous flow process that is fully integrated with
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 5 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
benchtop NMR technology for a hypervalent iodine (III) mediated cyclization to generate
substituted cyclopropanes. Real-time monitoring using the benchtop NMR for this study proved
to be an ideal informative analytical tool for cyclopropanation due to its distinctive spectral
features. In this application we also successfully developed an inline solvent switching device to
aid the monitoring process. Moreover, automated scripting was developed to expand on the NMR
monitoring capabilities by generating sequences of both one and two-dimensional homonuclear
proton NMR experiments. This enabled us to study the evolution of the distinctive coupling
relationships of cyclopropane protons over time when using 2D correlation spectroscopy (COSY)
and J-resolved spectroscopy (J-RES). Finally, we have also integrated compact mass spectrometry
into the inline NMR setup, demonstrating a complementary approach to multiple monitoring
technique to provide additional insights into our process understanding.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
REACTION MONITORING SYSTEM AND METHODS: (i) The Benchtop NMR System:
Reaction monitoring in this study was based on proton NMR measurements carried out using the
Spinsolve benchtop NMR spectrometer, designed by Magritek and built with a permanent magnet
based on a Halbach design, working at a frequency of 43.62 MHz. The benchtop NMR has a
vertical through-bore that runs from top to bottom of the device. This allows measurements to be
made using two types of NMR flow cells: (i) PTFE tubing inserted into a glass cell guide to center
the tubing through the NMR bore, and (ii) glass NMR flow cell tube designed to provide an
enlarged sample volume at the NMR measurement zone, thus maximizing sensitivity, whilst
minimizing volume elsewhere.10 In this work, the glass NMR flow cell was primarily used unless
otherwise stated. For continuous flow process, the benchtop NMR was integrated with the
Vapourtec R2+/R4 flow system as illustrated in Figure 1. During a typical experiment, the flow
output was streamed through the glass NMR cell. The benchtop NMR can be interfaced with the
5
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 6 of 39
1
2
3
4
5
6
7
8
Spinsolve software, where the collected data are displayed and processed using MNOVA in real-
time, including automatic spectrum baseline correction and phasing.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(ii) Reaction Monitoring Methodologies: For the purpose of our reaction monitoring studies, two
main approaches were adopted using both one and two-dimensional proton NMR experiments,
namely inline monitoring and online monitoring.11 The inline approach, was based on subjecting
the continuous flow reaction stream to a series of proton NMR measurements, with acquisitions
typically between 0.4 to 6.4 seconds, to profile the steady-state of the flow reaction and obtain
information on the flow stream composition (Figure 1). With this method, the continuous process
gives the intrinsic advantage of allowing a sequential varying of reaction conditions, better
established when dispersion of concentration gradient is minimized between reaction runs.
Figure 1. Default setup for reaction monitoring with Spinsolve benchtop NMR integrated with
Vapourtec R2+/R4 flow system (left). Glass NMR flow cell placed outside the NMR bore
(middle). Reaction monitoring configuration (right). Legend: (A) Vapourtec R2+ double pump
unit. (B) Vapourtec R4 flow reactor unit. (C) Flow stream input into the benchtop NMR. (D)
Benchtop NMR. (E) Flow stream output from the NMR tube. (F) Glass NMR flow cell. (G) Inline
reaction monitoring configuration under continuous flow conditions.
6
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 7 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
In this study, we further integrated the inline methodology with the solvent evaporation device
between the flow outlet and the benchtop NMR, in order to allow the switching from reaction to
analytical medium, hence recording spectra in a semi-deuterated media (vide infra). On the other
hand, the online monitoring of reaction approach allowed for tracking the reaction progress with
respect to changes in reactant or product concentration occurring over time. The reaction was
carried under batch using the R4 glass reactor, easily accommodated into the Vapourtec R4 unit
holder, coupled with the benchtop NMR system as illustrated in Figure S2 of supporting
information. Circulation of the reaction stream from the R4 glass reactor and through the NMR
magnet zone was achieved by using a Reglo® digital peristaltic pump, allowing measurements at
desired intervals. The online approach was found to be a convenient method for observing the
reaction progress over time, while making use of the relatively low reaction volume with the
advantage of the intrinsic reaction consistency given by the use of one single stock solution under
unperturbed state. For both inline and online methods, optimal flow rates were evaluated with
respect to the mixture‘s concentration in order to ensure effective acquisition of the magnetized
section of the flow stream to enhance the signal-to-noise ratio. Reaction times reported for all
reactions in this work take into account the time necessary for the solution transfer between
reaction chamber and NMR detection zone, given by the volume divided by flow rate. NMR
spectra timeline, however, correspond to the software measurement times, which is manually
activated with the first sample aliquot reading.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(iii) Solvent Switching System: Compared to high-field NMR, the fast automatic hardware lock
using an external lock configuration of the benchtop NMR allows the utilization of non-deuterated
solvent. However the need to carry analysis in deuterated media may still be desirable in cases
7
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 8 of 39
1
2
3
4
5
6
7
8
where the reaction‘s solvent and its satellite signals found to obscure areas of interest in the NMR
spectrum. This was the case in some of our monitoring experiments, and hence the need to integrate
a solvent switching step while enabling uninterrupted reaction-to-analysis process was realized.
Although continuous solvent removal and switching is widely used on an industrial scale as well
as on microfluidic scales, however, there is a growing demand for continuous evaporation
techniques in discovery laboratory scales to facilitate downstream processing. Recently, we
reported a convenient and efficient prototype device for evaporation, concentrating and switching
solvents in continuous flow.12 This simple and robust prototype was constructed from
commercially available parts, and developed based on exposure of a high surface spray of solution
to a desolvating gas stream, removing solvent quickly and efficiently, and compatibly with the
existing flow chemistry equipment. Key parameters affecting the solvent switching such as the
sprayer dimensions were studied and optimized. In this work, an in-house modified version of the
initial prototype solvent evaporation system was used, based on a polymeric spraying device with
a more compact assembly, minimizing the amount of components necessary for replacement and
ease of servicing. This modified solvent evaporation device consists of two blocks of chemically
resistant PTFE attached by a single threaded screw, and fitted into a standard Omnifit™ glass
column acting as the evaporation chamber, as illustrated in the supporting information section
(Figure S5). The final assembly could easily accommodate on a Vapourtec R4 heating jacket. The
switching process was configured to allow for a complete integrated operation of Flow-Evaporate-
Analyze protocol, i.e. carrying out flow reaction, then evaporate and switch to deuterated solvent,
followed by direct NMR analysis (Figure 2). During the switching operation, the continuous
reaction output was pumped into the evaporation device, while the desolvating air stream was
maintained at an optimal flow rate allowing for spray formation in the chamber. Deuterated
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
8
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 9 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
chloroform was then pumped into the device‘s chamber and the resulting diluted crude mixture
was streamed into the benchtop NMR using the peristaltic pump. Parameters for the solvent
evaporation and switching were initially optimized based on a dichloromethane (DCM) and
chloroform (CHCl3) system, mimicking the reaction medium and deuterated solvent respectively.
The optimization study was carried out according to a set of variables, namely (a) DCM and
chloroform solvent flow rates, (b) chloroform solvent time of injection, (c) gas stream flow rate
and (d) evaporation chamber temperature. The optimum set of conditions led to a satisfactory 97:3
CHCl3:DCM ratio monitored by the benchtop NMR (Table S2 and Figure S8 and S9)
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Flow-Evaporate-Analyze operation set-up.
RESULTS AND DISCUSSION
(i) Online Monitoring of Malononitrile 1 Cyclopropanation Reaction Progress: The
cyclopropanation reaction of olefins with malononitrile in the presence of PhI(OAc)2
(diacetoxyiodobenzene or DIB) gives access to a range of 1,1-dicyanocyclopropane derivatives
under mild oxidative conditions. Compared with other common routes to substituted cyclopropyl
9
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 10 of 39
1
2
3
4
5
6
7
8
substrates, this method has the advantage of not requiring the use of transition metal catalysis, thus
making the process cleaner and avoiding additional scavenging steps. To initiate the reaction
monitoring study we selected the direct preparation of 1,1-dicyanocyclopropane 3 derived from
the activated methylene malononitrile 1 and 4-tert-butylstyrene 2 (Scheme 1), based upon
literature precedents on the DIB mediated cyclopropanation first reported by Wirth et al., and later
by Zhang et al.13,14 It was observed that the optimized batch conditions of the reaction required an
inert atmosphere, and replacing molecular oxygen with DIB to obtain the cyclopropane moiety
exclusively while suppressing the hydro-cyanation of the olefin substrate.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1.Hypervalent iodine (III) mediated synthesis of 1,1 dicyanocyclopropane 3.
The initial selection of a system for real-time monitoring studies found that reacting 4-tert-
butylstyrene 2 with malononitrile 1, in excess DIB and K2CO3 in 1,2-dichloroethane (DCE) at 50
°C afforded cyclopropane 3 in good yield (82%) and over a relatively short time (2 hrs.). Having
therefore found satisfactory reaction conditions, we set out to investigate the reaction progress over
time using the online method as previously described (Figure S2). In order to prevent any potential
blockages, we segregated the heterogeneous K2CO3 from the circulating stream using a filter at
the reactor outlet. This filtration will also avoids the perturbation of the NMR field homogeneity
if a heterogeneous stream passes through the NMR detection zone. Although the benchtop NMR
exhibit an external lock feature, the reaction was nevertheless carried out in deuterated chloroform
for greater spectrum clarity. The reaction mixture was continuously circulated through the NMR
10
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 11 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
detection zone to carry out a series of 1D proton NMR measurements in intervals. Due to the
absence of heating and the need for K2CO3, we assumed that the reaction does not advance
significantly in the path between reaction bulk and NMR detection zone. Measurements were then
taken at the intervals of 14, 18, 23, 26, 30, and 47 minutes, representing a good monitoring window
to observe product formation as predominant, generating a total of six stacked spectra to profile
the reaction progress as shown in Figure 3a. With each measurement, integration values of five
selected resonances on the spectrum, were automatically derived by the MNOVA software (Figure
3b). From both representations in Figure 3, we were able to monitor the consumption of the
limiting reactant styrene 2 and the formation of the cyclopropane adduct 3. As profiled in Figure
3a, the starting material 1, 2 and DIB were predominantly present at the initial measurement at 14
min reaction time. Within 26 minutes, the majority of the styrene 2 was consumed, coinciding with
the appearance of the distinctive peaks of the vicinal CHa of cyclopropane 3 (/H 3.1-3.6) and the
geminal Hc-C-Hb (/H 2.2-2.5). Progressive disappearance of starting materials malononitrile and
DIB was also successfully detected by monitoring of peaks Pk2 and Pk5, while the chemical shift
of Pk1, on the other hand, did not change overall, due to the presence of the tert-butyl moiety on
both the styrene 2 and the adduct 3. Moreover, the progress profile of the cyclopropanation
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
reaction in Figure 3b, acquired from the H NMR measurements, gave a visual and quantitative
account of reagent consumption and product formation, with respect to selected resonances,
expressed as relative integration vs. time coordinate. The clear advantage of such type of output is
in the numerical values which can be used as triggers during computer controlled processes, for
instance, in order to suppress alternative reaction pathways or in defining optimal process times.
In the progress profile shown in Figure 3b, however, oscillation of the recorded chemical shift
intensity trends were observed. This is especially noticeable for Pk1 which corresponds to the tert-
11
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 12 of 39
1
2
3
4
5
6
7
8
butyl moiety present in both substrate 2 and product 3, as opposed to the expected constant
progress trend. This fluctuation in NMR signal intensity is likely caused by the flow turbulences
created due to the peristaltic pumping mechanism as well as the enlargement of the NMR flow cell
at the measurement zone.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(a)
(b)
Figure 3. Online monitoring of the cyclopropanation reaction of 1 with 2 using Spinsolve benchtop
1
NMR (43 MHz): (a) H NMR array plot from 14-47 min reaction time. (b) Profile of reaction
progress acquired using the 1H NMR data referenced to selected resonances.
12
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 13 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
It is important to point out that both representations shown in Figure 3, are equally important and
beneficial in their own right: the NMR spectra array, although non-numerical, can visually help
detect all events including the unexpected ones, whereas the relative integration profile highlights
the progress of a pre-selected resonances.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Furthermore, the monitoring of the cyclopropanation reaction progress of malononitrile 1 with 2
was also correlated with the two-dimensional homonuclear correlation spectroscopy experiments
(COSY). The cyclopropane structure lends itself well to this technique, due to the presence of a
number of distinctive coupling relationships between its protons. While the use of COSY is
traditionally and largely used for detailed characterization purposes, its application as a reaction
monitoring tool is becoming increasingly attractive. The traditional 2D COSY NMR has a long
acquisition duration, which considered slow for reaction monitoring purposes. With recent
advances in benchtop NMR spectroscopy, however, development of new 2D pulse sequences
technology such as Quickcosy has been realized to enable fast and efficient 2D NMR
measurements within a single scan at a medium field.15 We took advantage of the Quickcosy
experiment protocol provided by the Spinsolve benchtop NMR, one scan every 10 min, which is
more suited for monitoring the progress of cyclopropanation of malononitrile 1 with 2, to identify
the interactions between the coupled protons in the reaction mixture, thus enabling to track the
simultaneous formation and disappearance of correlations over time. The measured COSY
contour at 5, 16, 27 and 50 min reaction times, displayed a number of diagonal and off-diagonal
(cross) correlations representing coupled protons (Figure 4). At 5 min of reaction time, the
presence of unreacted styrene 2 substrate was confirmed by the coupling relationship (C1) at /H
5.0-7.2, between its geminal He-C-Hf and vicinal CHd. The presence of starting substrates
malononitrile 1 and DIB was established simply through the diagonal peaks C2 and C3, due to no
13
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 14 of 39
1
2
3
4
5
6
7
8
coupling interaction within their structures. As the reaction progressed, the correlation peaks for
all starting species progressively decreased, while the characteristic cross and diagonal correlation
of the cyclopropyl adduct 3 due to the vicinal and geminal coupling interactions (C4 and C5)
intensified by 50 min of reaction. The overall observation of reactant consumption and progressive
appearance of cyclopropane 3 using 2D Quickcosy coincided with that of the 1D 1H NMR. Used
in combination with 1D, generally 2D techniques can offer a complementary quantitative,
qualitative as well as characterization information on the cyclopropanation progress monitoring of
1.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(a)
(b)
Figure 4. (a) COSY contour array profiling the reaction progression over time of cyclopropanation
of 1 using Quickcosy. (b) Close-up region, of reaction monitored at 50 min, between /H 0.8-4.0,
highlighting the geminal and vicinal correlations.
14
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 15 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
(ii) Inline Monitoring of Malononitrile 1 Cyclopropanation under Continuous Process: A
continuous flow process for the hypervalent iodine (III) mediated preparation of 1,1-
dicyanocyclopropane 3 was developed and configured to allow for a complete continuous
operation of the Flow-Evaporate-Analyze protocol as described earlier. This approach allowed us
to successfully address one of the major challenges associated with integrating the real-time NMR
analysis with flow process. During monitoring of the cyclopropanation reaction, peak
identification and integration proved challenging due to the use of a protonated solvent, which
obscured some of the spectral features. Using the Vapourtec R2+/R4 flow system, the preparation
of the cyclopropane 3 under continuous flow conditions was carried out in dichloromethane
(DCM) by introducing a 0.30 M solution of malononitrile 1 and styrene 2 along with 0.55 M
solution of DIB, through a pre-packed column of K2CO3, over 44 min residence time (Figure 5).
The reaction stream output was then directed to the solvent evaporation device and then into the
NMR detection zone. The use of prepacked K2CO3 cartridge was found best when mixed with 1
mm sized glass beads, to avoid the agglomeration of the K2CO3 salts as the organic reaction media
was flowed through the reactor, subsequently blocking the flow stream over time. In addition, the
advantage of superheating under flow conditions allowed the use of more volatile DCM in the
system instead of the conventional DCE, therefore improving the solvent exchange process. The
inline monitoring of malononitrile 1 cyclopropanation was carried out using the Spinsolve reaction
monitoring protocol (RM), and with each measurement the integration values of selected
resonances were extracted to produce a series of trends, representing the selected resonances, to
profile the steady-state of the continuous reaction in real-time mode (Figure 6). Upon closer
examination of the steady-state profile, it was clear that a significant reaction conversion took
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
15
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 16 of 39
1
2
3
4
5
6
place as an increase in the relative integration values of the cyclopropane ring protons of product
7
8
3 were observed compared to that of the starting substrates malononitrile 1 and styrene 2.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Inline monitoring set-up of malononitrile 1 cyclopropanation under flow conditions
using the Flow-Evaporate-Analyze protocol.
Figure 6. Inline monitoring of cyclopropanation of 1 using Spinsolve benchtop NMR (43 MHz):
Steady-state profile under flow conditions acquired using the 1H NMR data referenced to selected
resonances.
16
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 17 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
(iii) Online Reaction Progress Monitoring of Ethyl-nitroacetate 4 Cyclopropanation: An
analogous approach for the online monitoring, as previously described, was extended to the
cyclopropanation of ethyl-nitroacetate 4. This hypervalent iodine (III) mediated cyclization
provides access to a range of 1-nitro-1-cyclopropyl carboxylates which are valuable synthetic
building blocks ultimately leading to cyclopropane .-amino acids.16 In order to exemplify further
the use of the benchtop NMR, we selected the preparation of cyclopropane carboxylate 5 derived
from the ethyl-nitroacetate 4 and 4-tert-butylstyrene 2 (Scheme 2). As a suitable candidate, the
cyclisation was carried out using the conditions based on the work of Charette et al., to generate
iodonium ylide in situ from the .-nitroesters which subsequently undergoes cyclopropanation with
olefins via the generated carbene in the presence of Rh(II) catalyst.17
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Hypervalent iodine (III) mediated synthesis of Ethyl-1-nitro-2(4-tert-butyl)-
cyclopropane carboxylate 5.
In initial experiments, we reacted ethyl-nitroacetate 4 in 5-fold excess of styrene 2 at an ambient
temperature in the presence of 1.1 equivalents DIB and catalytic amounts of bis-(rhodium-(., .,
.‘, .‘-tetramethyl-1,2-benzenedipropionicacid) Rh2(esp)2, providing the cyclopropane
carboxylate 5 in moderate yield (73%) with E:Z ratio of 5:1 (in substituted cyclopropane isomer E
indicates a trans relationship between the vicinal nitro and aryl group, whereas Z indicates the cis
relationship).
17
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 18 of 39
1
2
3
4
5
6
7
8
The use of styrene 2 in excess in solvent-less conditions, however, was undesirable with the
monitoring study, therefore we aimed at reducing the amount of styrene 2 to an equimolar ratio,
in the presence of an equivalent of DIB, at 50 °C, providing cyclopropane 5 in 48 % after 3 hours.
The online reaction monitoring was carried out with un-optimized conditions in CDCl3 using the
R4 glass reactor coupled with the benchtop NMR set-up shown in Figure S2. As the reaction
stream was continuously delivered into the NMR detection zone, a series of 1D proton scans were
measured at intervals between 5 min and 3 hours, generating an NMR array of nine spectra (Figure
7a). After 5 min of reaction time, the predominant species were the starting styrene 2, .-nitroester
4 and DIB substrates, and within 1 hour, the starting species appreciably decreased, while
cyclopropane 5 began to appear. Overlapped peaks of three different resonances appeared in the
region of /H 3.4-4.7, assigned for the vicinal CHa on the cyclopropane ring, the CH2 on the ethyl
moiety of 5, and the CH2 of the ethyl group of 4. Those coincided with the appearance of the
multiplet at /H 2.3-2.6 assigned to the geminal cyclopropyl protons (Hb-C-Hc). Moreover, the
cyclization process was better monitored through the evolution of the triplet at /H 0.5-1.1, assigned
to the CH3 on the ethyl moiety of 5. The formation of acetic acid was evident as the singlet at /H
2.14 appeared concurrently with the disappearance of the DIB‘s acetate ligand singlet at /H 2.10.
This was expected from the DIB due to the strongly electrophilic nature of the iodine in DIB,
leading to displacement of acetic acid and iodobenzene by-product.18 Although the singlet assigned
to the methylene moiety of 4 (/H 5.2) overlapped with the CHe of styrene 2 (/H 5.1-5.4), it was
nonetheless possible to track it‘s disappearance. The profile of the cyclopropanation progress
acquired using the 1H NMR data from the nine spectra array clearly demonstrated the progression
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
18
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 19 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
of the ethyl group on product 5 (/H 0.5-1.1) in parallel to the disappearance of the acetate ligands
of DIB, the methylene moiety of .-nitroacetate 4 and the geminal protons of styrene 2 (Figure 7b).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(a)
(b)
Figure 7. Online monitoring of cyclopropanation of 4 using Spinsolve benchtop NMR (43 MHz):
(a) 1H NMR array plot. (b) Profile of reaction progress acquired using the 1H NMR data referenced
to selected resonances (Table S3).
(iv) Inline Monitoring of Ethyl-nitroacetate 4 Cyclopropanation under Continuous Process: A
continuous flow process for the preparation of the 1-nitro-1-cyclopropylcarboxylate 5, was
developed. Initially the inline monitoring of cyclopropanation of 4 was carried out using the
Vapourtec R2+/R4 flow system by passing a 0.55 M solution of ethyl-nitroacetate 4 and styrene 2
in DCM along with 0.55 M solution of DIB and catalytic amounts of Rh2(esp)2 (0.5 mol %) through
a heated PTFE coil to provide product 5 in un-optimized 37% yield (Figure 8). The reaction out
stream was passed into the solvent evaporation device, to facilitate the exchange of DCM with a
total of 4 mL CDCl3 at 35°C. Upon NMR measurements, the generated steady-state profile for the
1
cyclopropanation of 4 in Figure 9, along with stacked H NMR data (Figure S11), demonstrated
the partial consumption of the starting species ethyl-nitroacetate 4, while the formation of
19
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 20 of 39
1
2
3
4
5
6
7
8
cyclopropane 5 was observed by the increased trend assigned for the geminal protons of
cyclopropane 5, along with its ethyl group. As the spectrum was recorded off a crude mixture, the
steady state profile appears rather noisy due to the low reaction yield. It is noteworthy that even in
non-ideal conditions, the steady state is still observed and the product can be tracked. In addition
the tracking of the relative integration for the cyclopropane vicinal proton expected at /H 3.7, was
challenging due to the overlapping with other resonances exhibiting similar chemical shifts. In
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
fact, the H NMR spectrum obtained by the inline monitoring of the cyclopropanation of 4,
displayed a number of overlapping signals. Peak overlapping is an inherent problem associated
with NMR spectroscopy especially at lower field, however, 2D NMR spectroscopy proved
particularly useful in identifying resonances. We demonstrated earlier how the online monitoring
of cyclopropanation of malononitrile 1 with Quickcosy facilitated the tracking of correlations
between coupled protons under reaction conditions. Likewise, inline monitoring using the
Quickcosy for the cyclopropanation of 4 under continuous flow enabled by the use of the Spinsolve
automated scripting, allowed more flexibility in monitoring scenarios. Hence, we developed a
LEY-2D-COSY script to run a loop sequence of 1D 1H NMR and 2D COSY, executed every 13
minutes over 3 loops (Figure S12) NMR experiment scripting thus creates tailor-made
experimental sequence automation allowing significant time saving and enabling repetitive runs
in loops.
20
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 21 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. Inline monitoring set-up of ethyl-nitroacetate 4 cyclopropanation under flow conditions
using the Flow-Evaporate-Analyze protocol.
Figure 9. Inline monitoring of cyclopropanation of 4 using Spinsolve benchtop NMR (43 MHz):
Steady-state profile under flow conditions acquired using the 1H NMR data referenced to selected
resonances.
The steady-state monitoring of cyclopropanation of 4 with the LEY-2D-COSY script produced
three COSY contours along with three 1D proton spectra, demonstrating a number of diagonal and
cross correlations (Figures S13 & S14). The presence of unreacted styrene 2 was verified through
cross correlations above and below the diagonal line of its olefinic protons at /H 5.0-7.2, while the
21
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 22 of 39
1
2
3
4
5
6
7
8
residual substrate of ester 4 appeared with a diagonal peak at /H 5.1 assigned to its methylene
moiety. Taking a closer look at the COSY horizontal region between /H 0.5-5.0 ppm (Figure 10),
the formation of cyclopropyl carboxylate 5 was established through its correlation peaks p5 and
p5‘ due to its vicinal coupling, while the CH3 and CH2 protons of cyclopropane 5 ethyl moiety
correlated at p1 and p1‘. Coupling between the CH3 and CH2 protons of the ethyl moiety of 4 was
observed by the appearance of cross peaks p3 and its pair p3‘. The overlapped peaks in the region
of /H 3.4-4.7, corresponded to the cross peaks over the COSY contour p1‘, p3‘ and p5‘, and on a
closer observation of those cross peaks the extent of overlapping was clarified to be over a portion
of 0.5 ppm wide. Similarly, the geminal protons on the cyclopropane 5 were found overlapping
with the unreacted acetate peak of DIB over a portion of 0.25 ppm wide. Reaction monitoring
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
using the 2D H NMR, in summary, allowed us to clearly distinguish and track starting species
consumption and product formation in regions of signal overlap, otherwise unsolvable by 1D 1H
NMR.
Figure 10. Inline steady-state monitoring of cyclopropanation of 4 with LEY-2D-COSY: 2D
COSY contour horizontal region between /H 0.5-5 ppm. Resonances: 1 - ethyl CH3 of 5; 2 - tert-
Butyl of 5 & 2; 3 - ethyl CH3 of 4; 4 œDIB; 5 - geminal CH of 5; 6 - vicinal CH of 5; 7 - ethyl CH2
of 5; 8-ethyl CH3 of 4. (Further information on the LEY-2D-COSY experiment are found in the
supplementary section).
22
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 23 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
An additional 2D NMR J-resolve spectroscopy (1H J-RES) experiment was used for the
deconvolution of regions of peak overlap. In J-RES the chemical shift and spin-spin coupling are
separated, therefore allowing a spin-spin coupling constant J (Hz) plot against chemical shift /H
(ppm). As a result, the signal multiplicity observed in the 1D proton NMR coalesces into a peak
in the chemical shift coordinate (horizontal); the signal multiplicity information, however, is
transferred into the vertical axis (J in Hz). Hence overlapped multiplet signals should coalescence
separately, resulting in a discrete set of broader signals, according to their specific resonance
environment. The 2D J-RES spectrum in the region of 0.5-5.0 ppm shown in Figure 11a,
corresponds to the signals of the crude flow mixture from the cyclopropanation reaction of 4, with
proton signals marked A to H. Resonance A, a triplet on the 1D NMR spectrum, was coalescing
into a singlet on the J-RES horizontal axis, with J = 7.0 Hz, assigned to the CH3 of ethyl group of
5. On the other hand, overlapped resonances at /H 3.5-4.7, appeared all resolved into three distinct
singlets of resonances (F, G, H) on the horizontal axis of the J-RES spectrum, where those marked
as G and H corresponded to the CH2 moiety of the ethyl group of 4 and 5 respectively. The
remaining resonance F on the J-RES assigned to the cyclopropane ring CHa proton, appearing as
a doublet of doublet in the 2D spectrum. From this doublet of doublet the vicinal couplings was
deduced for CHa with Jcis =10.1 Hz and Jtrans = 9.0 Hz, in line with the expectation of Jcis>Jtrans for
a cyclopropane ring due to the rigid dihedral angle. Moreover the ring rigidity and the substituent
effect on the cyclopropane 5 was found to have a positive effect on the geminal coupling between
CHb and CHc to give Jgem of 6.5 Hz, extracted from the two singlets of resonances E and E‘ (Figure
11). Furthermore, carrying the J-RES measurement of the isolated isomer E (major) and isomer Z
(minor) of 1-nitro-1-cyclopropylcarboxylate 5, we were also able to observe the differences in
chemical shifts between both isomers as demonstrated in Figure 11b and S15, respectively. The
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
23
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 24 of 39
1
2
3
4
5
6
7
8
distinction in the NMR spectra between the E and Z isomer was attributed to the orientation of the
aryl group with respect to the nitro moiety on the cyclopropane ring. The CHa signal of the E
isomer overlapped with the CH2 of the ethyl group, while both the geminal protons CHb and CHc
forms a multiple of overlapped peaks between 2.0-2.6 ppm (Figure 11b). On the other hand the
steric compulsion by the aryl group and the anisotropic effect by the nitro group on the Z isomer
of cyclopropane 5 caused the ethyl ester group to shift downfield, in comparison to the E isomer,
whereas both CHa and CHb were mildly more shielded, and the CHc (now trans to CHa) slightly
shifted to downfield. It appeared that the J-RES of the E isomer was visually similar to that of the
cyclopropanation reaction crude, confirming that the E isomer of 5 was the predominant isomer in
the reaction mixture. Hence, through this example, the J-RES demonstrated its potential in
providing an insightful feedback helping to refine and improve reaction monitoring further.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(a)
(b)
Figure 11. (a) 1H J-RES spectrum at 0.5-5.0 ppm of cyclopropanation reaction of 4 crude mixture,
with proton signals marked A to H. (b) 1H J-RES spectrum of isolated [E] 1-nitro-1-
24
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 25 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
cyclopropylcarboxylate 5. (For 1H J-RES of isomer [Z] of 5 refer to Figure S 15). Resonances: A
- ethyl CH3 of 5; B - tert-Butyl of 5; C - ethyl CH3 of 4; D - DIB; E & E‘- geminal CH of 5; F-
vicinal CH of 5; G - ethyl CH2 of 5; H - ethyl CH3 of 4.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Further optimization of the cyclopropanation reaction of ethyl-nitroacetate 4 with an equimolar
amount of styrene 2 under continuous flow was sought, following the initial 37% yield obtained
for cyclopropane 5 (E and Z combined). This moderate yield was attributed to the high dilution
conditions of the reaction, as literature methods highlight the importance of using alkene in excess,
with reactions carried out under neat conditions. Several attempts at improving the reaction
conditions were made, namely (i) increasing the reaction temperature, (ii) increasing the overall
reaction concentration and (iii) increasing the number of equivalent of DIB alone (ditto), with no
success due to low solubility of DIB. However, by flowing an equimolar solution of styrene 2 in
DCM along with Rh(II) through an Omnifit column prepacked with 2.2 equivalents of solid DIB
and 1 mm recyclable glass beads, cyclopropane 5 was afforded in 77% yield (E:Z , 4:1), in 30 min
residence time. Reaction set-up under continuous flow is illustrated in Figure 12.
Figure 12. Flow reaction setup for cyclopropanation of 4 under optimized reaction conditions.
25
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 26 of 39
1
2
3
4
5
6
7
8
(v) Multiple Monitoring Technique for Cyclopropanation of Phenyl Iodonium Ylide 6 using CMS-
NMR: A multiple monitoring technique approach, using compact mass spectrometry with benchtop
NMR (CMS-NMR), was developed in order to provide complementary analytical information for
the cyclopropanation of phenyl iodonium ylide. Phenyl iodonium ylide derivatives are widely used
in carbene-transfer reactions including cyclopropanation, presenting a comparable reactivity to
their diazo counterparts under milder conditions.19 The iodonium ylide method takes advantage of
the readily generated carbene species, induced by metal catalysis, displacing the molecular
iodobenzene which then cyclizes in the presence of an olefin. However, it has been established
that the iodonium ylide can also provide dimerization products via the competing decomposition
of carbene intermediate.20 As further discussed within, mass spectrometry provided a better insight
into the occurrence of such decomposition, compared to NMR which proved challenging with
resonance signal overlap. In addition, whilst NMR monitoring could help to monitor the
cyclopropanation of the olefin moiety, the MS could also help to identify the iodobenzene by-
product formation. A continuous process was set for the phenyl iodonium ylide 6 mediated
cyclopropanation using the Vapourtec R2+/R4 flow system configured with the Spinsolve
benchtop NMR and the Advion compact MS with Atmospheric Pressure Chemical Ionization
(APCI) detector (Figure S18). Using this combined CMS-NMR setup, the cyclopropanation of
phenyl iodonium ylide 6 with 4-tert-butylstyrene 2 was carried out under continuous flow in the
presence of Rh(II) catalyst, to afford cyclopropane-1,1-dicarboxylate 7 (Scheme 3). Ylide 6 was
prepared from the reaction of dimethylmalonate 8 with DIB under alkaline conditions in dry
acetonitrile, and obtained as an amorphous solid in 65% of crude, without further purification to
avoid decomposition of the ylide (Scheme 4).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
26
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 27 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Hypervalent iodine (III) mediated synthesis of cyclopropane-1,1-dicarboxylate 7.
Scheme 4. Synthesis of Phenyl iodonium ylide 6.
Due to the poor solubility of phenyl iodonium ylide 6 in DCM, we opted for the pre-packing into
an Omnifit glass column along with 1 mm recyclable glass beads, moreover preventing the caking
of the ylide over time. The reaction successfully took place by introducing DCM solution of
styrene 2 along with a solution of 0.1 mol % Rh2(esp)2, through the pre-packed column of ylide 6
over 33 min while cooled at 18-20 °C (Figure 13). The ylide gradually consumed and disappeared
from the pre-packed column as the styrene 2 solution is streamed through, leaving behind the glass
beads. An aliquot of the reaction outflow was periodically redirected to the MS detector zone by
mean of a Rheodyne™ MRA splitting device, while most of the reaction stream output was
enriched with deuterated chloroform by passing through the evaporator, prior to NMR analysis.
From the resultant MS measurements, the cyclopropane adduct was identified by the molecular
ion peak at m/z 291 [M+H] accompanied by its fragment peak at m/z 259, while the iodobenzene
by-product was identified at m/z 204 [M+H] (Figure S20). Cyclopropane 7 formation was also
identified using 1H NMR monitoring by the presence of the two dimethoxy moieties appeared as
27
Submitted May 19, 2016
ACS Paragon Plus Environment

Organic Process Research & Development
Page 28 of 39
1
2
3
4
5
6
7
8
two separate singlets approximately at /H 3.0 and 3.4, as opposed to the dimethoxy moieties of the
starting iodonium ylide 6 expected as a one singlet at 3.5 ppm (Figures S21 and S22).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 13. Multiple reaction monitoring set-up for the cyclopropanation of ylide 6 under flow
conditions.
Although ylide 6 demonstrated limited solubility in the reaction medium, however, leaching of the
excess ylide from the pre-packed column was detectable with the MS by its molecular ion fragment
observed at 229 m/z, while the ylide 6 molecular ion [M+H] was not observed under the APCI
conditions (Figure S24). Additionally, the competing decomposition pathway of ylide 6 via
dimerization was best established by MS rather than NMR due to potential signal overlap, with
product and ylide, in the region of /H 3.5-3.9. This decomposition was demonstrated by the
controlled reaction carried in the presence of Rh(II) catalyst where upon MS analysis the molecular
ion [M+H] was identified at m/z 261 (Figure S23). When mass monitoring of the cyclopropanation
28
Submitted May 19, 2016
ACS Paragon Plus Environment

Page 29 of 39
Organic Process Research & Development
1
2
3
4
5
6
7
8
reaction carried at 18-20 °C, on the other hand, no dimerization was observed proving the
decomposition pathway did not take place.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(vii) Scope of Hypervalent iodine (III) cyclopropanation under Continuous Flow Conditions:
Finally, the potential of the continuous process of these cyclopropanations were further explored
at preliminarily level by studying the overall profile of all discussed substrates, malononitrile 1,
ethyl-nitroester 4 and phenyl iodonium ylide 6, with a number of styrene derivatives to afford
substituted cyclopropanes in moderate to good yields (Table 1). The best cyclopropanation yields
were observed with the 4-tert-butylstyrene, while lower yields observed with the styrene
derivatives bearing electron withdrawing groups. Moreover, a couple of trends emerged with
respect to reaction parameters. Although we observed that higher reaction concentration generally
improves the productivity, however, excessive concentration can favor substrate decomposition in
the case of the malononitrile reactions. Further challenges of achieving optimal reaction
concentration in flow, were due to reagent limited solubility in the case of DIB. This problem can
be eliminated by use of polymer supported DIB reagents or by enhancing solubility with ring
functionalization such as o-alkoxyphenyliodonium derivatives.21 A similar balance is found with
reaction temperature which, while generally favoring reactivity, can also create polymerization of
some styrene substrates at higher temperatures especially when bearing an electron withdrawing
group. These preliminary studies generally demonstrate the potential of the hypervalent iodine
(III) mediated cyclopropanation continuous process.
29
Submitted May 19, 2016
ACS Paragon Plus Environment