Coupling Continuous Flow Microreactors to MicroNIR Spectroscopy: Ultracompact Device for Facile In-Line Reaction Monitoring

Article abstract of DOI:10.1021/acs.oprd.8b00060

In this study, we applied a portable near-infrared spectrophotometer (MicroNIR) for in-line monitoring of the synthesis of 5-hydroxymethylfurfural (5-HMF) in a continuous flow microreactor. Under the best reaction conditions using isopropyl alcohol/dimethyl sulfoxide as the reaction solvent and a fixed-bed reactor packed with solid acid catalyst, total conversion of d-fructose was observed, and 5-HMF was obtained in 95% yield with a residence time of just 11.2 min. Principal component analyses and construction of multivariate control charts based on Hotelling's T² and Q residuals were also performed and proved the excellent response of the compact MicroNIR device for in-line monitoring of 5-HMF production without variation in the yield over 8 h/day during 5 days. Our results demonstrate the great potential for the application of this compact device in the monitoring of laboratory-scale reactions, which can be extended to industrial scales.

Full text of DOI:10.1021/acs.oprd.8b00060

Subscriber access provided by Universiteit Utrecht
Full Paper
Coupling Continuous Flow Microreactors to MicroNIR Spectroscopy:
Ultra-Compact Device for the Facile In-line Reaction Monitoring.
Renan Galaverna, Rafael Luis Ribessi, Jarbas José Rodrigues Rohwedder, and Julio Cezar Pastre
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.8b00060 • Publication Date (Web): 12 Jun 2018
Downloaded from http://pubs.acs.org on June 13, 2018
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 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 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.
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 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Coupling Continuous Flow Microreactors to
MicroNIR Spectroscopy: UltraꢀCompact Device for
the Facile Inꢀline Reaction Monitoring
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
Renan Galaverna,# Rafael L. Ribessi,# Jarbas J. R. Rohwedder* and Julio C. Pastre*
Institute of Chemistry, University of Campinas ꢀ UNICAMP, PO Box 6154 ꢀ Zip Code 13083ꢀ
970, Campinas, SP, Brazil.# These authors contributed equaly to this work.
Correspondence to Julio Cezar Pastre, Institute of Chemistry, University of Campinas ꢀ
UNICAMP, PO Box 6154 ꢀ Zip Code 13083ꢀ970, Campinas, SP, Brazil. Eꢀmail:
juliopastre@iqm.unicamp.br. Phone: +55 (19) 3521 31 43.
KEYWORDS. Coupling, MicroNIR, Flow Chemistry, 5ꢀHMF, Inꢀline monitoring
ABSTRACT. In this study, we applied a portable nearꢀinfrared spectrophotometer (microNIR)
for the inꢀline monitoring of the 5ꢀhydroxymethylfurfural (5ꢀHMF) synthesis in a continuous
flow microreactor. Under the best reaction condition using isopropyl alcohol/DMSO as reaction
solvent and a fixedꢀbed reactor packed with solid acid catalyst, total conversion of Dꢀfructose
was observed and 5ꢀHMF was obtained in 95% yield in just 11.2 minutes of residence time.
Principal Component Analyses (PCA) and multivariate control chart based on Hotelling’s T2 and
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 32
1
2
3
4
5
6
7
8
9
Qꢀresidual were also performed and proved the excellent response of the compact microNIR
device for the inꢀline monitoring of 5ꢀHMF production, without variation in the yield over 8
hours per day, during 5 days. Our results demonstrate the great potential for the application of
this compact device in the monitoring of laboratory scale reactions, which can be extended to
industrial scales.
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
INTRODUCTION
The synthesis of bioꢀbased chemicals under continuous flow conditions has increased over the
last years.1 This growing motivation follows the recent development of continuous flow
chemistry as an enabling technology, which brings several advantages compared to the
traditional synthesis in batch for both organic and inorganic syntheses.2 Regarding the synthesis
of bioꢀbased materials, an impressive number of valuable compounds have been successfully
obtained from lignocellulosic biomass.3 In particular, molecules such as 5ꢀ
hydroxymethylfurfural (5ꢀHMF) has been extensively investigated in view of its broad
applications as building block to achieve feedstocks for bulk chemicals, polymers, solvents, and
fuels.4 5ꢀHMF is usually synthesized under acidic conditions from C6 carbohydrates, including
fructose, glucose, chitosan, cellulose, sucrose, starch, agarose, inulin, and raw biomass.4 Among
the sugars cited, fructose has been the main choice since it provides better conversions and yields
through a rapid and direct dehydration reaction.5
Continuous processing is widely used for the manufacturing of petrochemicals and bulky
chemicals. In the last decade, the use of continuous flow microreactors in the lab scale has
significantly increased and numerous reviews have been reported, highlighting the benefits of
this technology.2 Reactions in continuous flow regime allow rapid and easy optimization of some
key reaction parameters by simply changing the flow rate to introduce different stoichiometric
ACS Paragon Plus Environment
2

Page 3 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
amounts of reagents along with a precise control of the temperature/pressure and residence time.
Bearing this in mind, the coupling of continuous flow microreactors to analytical tools is highly
desirable since this would allow a constant monitoring and instantaneous detection of the
reaction species which consist in reactants, transient intermediates and final products.6 This
approach also allows automation by software control as well as selfꢀregulation of the reaction
process.7 Indeed, several analytical techniques such as FTIR,8 Raman,9 UVꢀVis,10 HPLC,11 GC,12
MS,13 and NMR14 have been successfully employed for a series of continuous flow reactions for
inꢀ, onꢀ and offꢀline monitoring.
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
FTIR spectroscopy coupled to flow microreactors is one the most used technique for inꢀline
reaction monitoring.8a,b This can be rationalized in view of some commercially available ATRꢀ
based probes, which offer easy coupling to flow microreactors and easy operation. For instance,
Nigel and coꢀworkers demonstrated several examples of reaction monitoring using FTIR, such as
fluorination, oxazole formation, hydrogenation, Curtius rearrangement, and azide formation.8a
Other recent studies include oxidation of allylic alcohols and trifluoromethylation of
heteroarenes,8c thermolysis,8d aminocarbonylation8e and amineꢀredox reaction under flow
regime.8f
A widely used and complementary alternative to FTIR is Raman spectroscopy. Although
FTIR provides highly specific and instantaneous information about reaction partners (reagents,
intermediates and products), Raman spectroscopy presents some advantages over FTIR. For
example, water cause strong interference and reduces the sensitivity of FTIR for reactions
performed in aqueous media. Intense overtones and combination bands also contribute to a
difficult spectrum interpretation. In contrast, water interference, intense overtones and
combination bands are factors that do not impose complications in the Raman analysis.15 Thus,
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 32
1
2
3
4
5
6
7
8
9
the benefits of Raman spectroscopy coupled to flow microreactors for inꢀline monitoring have
been successfully demonstrated recently. Some examples comprise oxidations catalyzed by tetraꢀ
Nꢀpropylammonium perruthenate,9a Michael additions,9b kinetic studies,9c and other relevant
applications.9dꢀh
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
Regarding the synthesis of inorganic materials using microreactors in continuous flow regime,
UVꢀVis spectroscopy offers an excellent and accessible tool for the inꢀline monitoring since
many types of inorganic materials (complexes, clusters, nanoparticles, etc) show absorption that
involves transitions between filled and unfilled dꢀorbitals of the metal ion in the UVꢀVis region.
This attractive coupling has been demonstrated for several inorganic synthesis such as
fluorescent nanoparticles,10a gold nanoparticles,10b and transition metal clusters such as Periodic
Mesoporous Organosilicas (POMs).10c Photochemical reactions were also investigated by the use
of UVꢀVis devices coupled to flow microreactors.10d
A more precise approach for inꢀline monitoring can be offered by the coupling of high
sensitive techniques to continuous flow microreactors. For example, HPLC, GC, and MS provide
an ideal ambient for inꢀline quantitative monitoring. Unlike the FTIR, UVꢀVis, and Raman
techniques, some crucial care must be taken for these approaches. For instance, HPLC, GC, and
MS require indispensable split to introduce only a small amount of the reaction crude for inꢀline
monitoring to avoid saturation of the detector. Bearing this in mind, accurate quantitative and
qualitative approach for inꢀline monitoring has been demonstrated in the literature using HPLC,11
GC,12 and MS techniques.13
NMR is another attractive technique explored for inꢀline monitoring coupled to flow
microreactors for several organic reactions once it is a powerful and useful tool for structural
elucidation.14aꢀj For example, Cronin and coꢀworkers using a benchtop NMR spectrometer (43
ACS Paragon Plus Environment
4

Page 5 of 32
Organic Process Research & Development
1
2
3
4
19
MHz) performed F NMR, COSY, and HSQC experiments to follow electrophilic fluorinations
5
6
in continuous flow regime.14a
7
8
9
It is worth mentioning that combining flow chemistry and analytical tools is also important
for the development of process analytical technology (PAT). PAT consists in the use of different
analytical techniques, usually associated with multivariate statistical data analysis, in order to
obtain a greater understanding, continuous monitoring, and improvement of the process.16 Some
high performance analytical techniques, such as MS and NMR, are suitable for PAT, but the high
cost for acquisition, implementation, and maintenance are some drawbacks for their routine use.
Thus, techniques that deliver high analytical performance (precision, accuracy, and robustness)
at lower costs are more attractive for PAT.16 In this context, nearꢀinfrared spectroscopy (NIRS)
shows ideal features for the inꢀline monitoring of different processes when associated to
chemometric tools.17 NIRS analysis requires little or no sample preparation, it has short response
time and great response, and it is a nonꢀdestructive method. In addition, the instrumentation used
in NIR spectrophotometers is robust, suffering less damage from environmental conditions, thus
being particularly suitable for industrial environments.17
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
Recently, fast technological developments have provided great improvements in analytical
instrumentation. A trend that is observed in the market is the miniaturization of various
analytical tools,18 including NIRS.19 NIR ultraꢀcompact devices, also known as microNIR, unlike
traditional benchtop devices (FTꢀNIR) have no moving parts due to the monochromator used as a
thin film FabryꢀPerot cavity or wedge interference filter type. Taking into account their low
weight (< 100 g), they can be classified as handꢀheld devices and recent studies have reported
their potential application for the quantification of drugs,20a monitoring the quality of fruits,20b
and quantification of gases.20c
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 32
1
2
3
4
5
6
7
8
9
Based on the techniques described herein, there is a lack of studies showing the coupling of
continuous flow microreactors to microNIR spectroscopy for inꢀline reaction monitoring. In that
way, we demonstrate herein the application of a handꢀheld NIR spectrophotometer coupled to a
continuous flow microreactor for the inꢀline monitoring of a bioꢀbased process previously
investigated by our group.21 For this purpose, the dehydration reaction of Dꢀfructose into 5ꢀHMF
was evaluated and, considering the fundamentals of green chemistry, a heterogeneous catalytic
system was chosen along with an alcoholꢀmediated reaction. To this end, fructose dehydration
reaction was performed using solid acidꢀcatalyst Amberlystꢀ15 and isopropanol/DMSO (15%
v/v) as solvent system. Chemometric methods were also applied and demonstrated process
robustness, reproducibility and signal stability, confirming that microNIR devices are a
promising tool for inꢀline monitoring of the reactions performed in continuous flow regime
(Scheme 1).
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. Schematic assembly of microNIR coupled to flow for inꢀline monitoring of the
dehydration reaction of DꢀFructose into 5ꢀHMF: A) bottle with reagent, DꢀFructose solution; B)
piston pump; C) glass column filled with Amberlystꢀ15; D) temperature controller; E) tungsten
lamp; F) flow cuvette; G) microNIR Spectrophotometer and; H) collection flask for 5ꢀHMF.
ACS Paragon Plus Environment
6

Page 7 of 32
Organic Process Research & Development
1
2
3
4
EXPERIMENTAL SECTION
5
6
7
8
Materials and Methods: Starting materials and reagents were obtained from commercial
9
sources and used as received unless otherwise specified. Organic solutions were concentrated
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
under reduced pressure on an IKA rotary evaporator RVꢀ10 Control. H NMR experiment was
recorded on Bruker 250 MHz. The flow reactions were performed in an integrated continuous
flow reactor FlowSyn system (UNIQSIS). NIR spectra were recorded on microNIR (Viavi,
USA) spectrophotometer.
General procedure for continuous flow reactions: The continuous flow reactions were
performed in a FlowSyn system (from UNIQSIS) using the following conditions: In a roundꢀ
bottomed flask of 1.5 L containing a magnetic stirrer were added 45 g of Dꢀfructose in 1 L of a iꢀ
PrOH/DMSO solution (15 % v/v) and heated at 60 °C to completely dissolve the Dꢀfructose. An
OmniFit® glass column (10 mm i.d. × 50.0 mm length) was packed with Amberlystꢀ15 (2.0 g,
void volume of 2.8 mL). A 5 bar backꢀpressure regulator was placed after the reactor to maintain
the system pressurized. The flow rate was varied from 4.0 to 0.25 mL.minꢀ1 at 110 °C. Before the
direct pumping of the Dꢀfructose solution through the system, the packed column was heated up
and flushed with iꢀPrOH/DMSO solution (15 % v/v) for 30 min. Before collecting the NIR
spectra and/or taking samples for offꢀline analysis, a residence time of three times was run for
each flow rate to ensure steadyꢀstate operation. After achieving steadyꢀstate, samples of 10 mL
1
were collected and analyzed by quantitative H NMR analysis using 1,3,5ꢀtrimethoxybenzene as
internal standard to determine the chemical yield for 5ꢀHMF product.
Continuous flow microreactor coupled to the microNIR spectrophotometer for in-line
monitoring: Inꢀline monitoring reaction of Dꢀfructose into 5ꢀHMF by microNIR was performed
according with the assembly shown in Figure 1. A tungsten lamp (50 W, Spectral Products,
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 32
1
2
3
4
5
6
7
8
9
USA) controlled by a power supply was used as radiation source (Figure 1E). The lamp was
turned on at least 30 minutes before the start of the experiments. Two flow cuvettes (Hellma,
Germany) were used to monitor the reaction and both have 3 mm of optical path and 40 mm2 of
window (Figure 1F). One of them was connected to the outlet of the glass column, while the
other was kept empty for obtaining the backgrounds spectra. Lastly, a microNIR
spectrophotometer (Viavi, USA), operating in the spectral range of 1150 to 2150 nm, 15 ms
integration time and average of 50 scans by spectrum was used to monitor the reaction in which
each spectrum counted 127 variables (Figure 1G). A mean spectrum of three scans was obtained
for each sample.
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
Data processing: All chemometric treatments and models were performed using the
Unscrambler X software version 10.3 (Camo, Norway). For the construction of all chemometric
models (PLS and PCA’s), NIR spectra were preprocessed with the SavitzkyꢀGolay first
derivative (7ꢀpoint window and 2nd order polynomial) and mean centered.
The Partial Least Squares regression (PLS) model was constructed with NIR spectra from the
different reaction yields and the reference data obtained by qHNMR (reference method). Two
sets of data were used, each obtained on different days. For the calibration set, 27 spectra were
used and for the prediction set, 9 spectra were used.
Principal Components Analyzes (PCA) was used as tool for the exploratory analyses and to
monitor the conversion of Dꢀfructose into 5ꢀHMF under continuous flow conditions. Firstly, it
was constructed a PCA model to evaluate all the date set. For this purpose, spectra from both
maximum reaction yield (95%, 0.25 mL.minꢀ1) as well as those from different reaction yields (4
up to 0.5 mL minꢀ1) were used and resulted in 74 spectra. Next, a second PCA model was
constructed to obtain the limit values of the multivariate control charts: Qꢀresidual and Hotlling’s
ACS Paragon Plus Environment
8

Page 9 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
T2, both with 95% confidence using two principal components. In order to calculate the limits of
control charts, ten spectra were used from the samples with maximum yield (95%, 0.25 mL. minꢀ
1), which was previously confirmed with the reference method (qHNMR). Finally, to monitor the
reaction, a new PCA model was constructed using 54 spectra obtained from the reaction
monitoring at 0.25 mL.minꢀ1 by 8 hours per day during 5 days. Note that the measures were
performed every 45 min.
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
RESULTS AND DISCUSSION
To date, there are few studies on the use of analytical tools for onꢀ or inꢀline monitoring of the
dehydration reaction of C6ꢀsugars into 5ꢀHMF. It is worth mentioning that this approach would
be crucial for the scale up of this process towards industrial applications, which is still pending.
Toward this end, we recently evaluated the use of FTIR spectroscopy for the inꢀline monitoring
of the 5ꢀHMF production21 and Dai and coꢀworkers investigated the influence of sulfonated
resins on the stability of 5ꢀHMF by recording the in-situ ATRꢀIR spectra.22
In view of our interest in the preparation of 5ꢀHMF, and considering the potential of reaction
monitoring techniques for continuous flow process development, we decided to investigate the
use of a handꢀheld microNIR device to monitor the dehydration reaction of Dꢀfructose into 5ꢀ
HMF using the conditions developed previously by our group (Scheme 2).21
Scheme 2. Schematic representation of the dehydration of Dꢀfructose (1) to afford 5ꢀHMF (2).
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 32
1
2
3
4
5
6
7
8
9
Firstly, some preliminary experiments were performed to find out which range of wavelengths
would be suitable for subsequent inꢀline monitoring for the dehydration reaction of Dꢀfructose
into 5ꢀHMF using the microNIR device. For this purpose, some offꢀline measures were
performed with the starting Dꢀfructose solution (0.25 mol Lꢀ1 ) and with the reaction mixtures
that represent 50% and 95% yield for 5ꢀHMF, based on our previous work.21,23 It is important to
mention that the use of samples obtained from the reaction stream (and not a solution of standard
5ꢀHMF obtained from commercial sources) was detrimental, since water is obtained in this
process as a coꢀproduct (3 molecules of water per molecule of fructose) and also accounts for the
variation in the NIR spectrum. The spectra obtained for each solution was overlapped and shown
in Figure 1.
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 1. Spectra obtained for the offꢀline analysis using the microNIR device for Dꢀfructose
solution (black line), 5ꢀHMF solution for 50% yield (red line), and 5ꢀHMF solution for 95%
yield (blue line).
As can be seen in Figure 1, there is a significant spectral variation in the range from 1870 to
2010 nm within the region of NIR spectrum shown here, which could be associated with the
ACS Paragon Plus Environment
10

Page 11 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
conversion of Dꢀfructose into 5ꢀHMF. The greater and lower absorption intensity were correlated
to the HMF solution with higher yield (95%) and Dꢀfructose solution (black line), respectively.
The variation in the spectrum can be attributed to the overlapping of absorption bands of the
species formed in the reaction medium, which are three molecules of water and one molecule of
5ꢀHMF for each Dꢀfructose molecule. For instance, water shows strong absorption in the range
1850ꢀ2000 nm, which corresponds to its combination band.24 In addition, another important
contribution comes from the aldehyde functional group in the 5ꢀHMF product, which displays
second overtone of the carbonyl stretching in the same region.24 From these observations, we
concluded that the band centered in 1925 nm was formed by contributions of the both product
and coꢀproduct. Note that the water also has a band with maximum in 1400 nm, which
corresponds to the first overtone of the OH bond, however, due to the hydrogen bonds with the
solvents, this band are less intense and not as visible as the band in the 1870ꢀ2010 nm range.25
After these initial offꢀline measures, we confirmed that microNIR was suitable for the
proposed study. In that way, we performed an inꢀline experiment to assess the response of the
microNIR device to different reaction conditions (i.e. different flow rates). Thus, the microNIR
spectrometer was placed after the fixedꢀbed reactor filled with Amberlystꢀ15 and the reaction
stream was passed through the flow cell for inꢀline monitoring (Figures S1-S5). The flow rate
ranged from 4.0 to 0.25 mL.minꢀ1 and the results are presented in Table 1.
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
Table 1. Dehydration of Dꢀfructose into 5ꢀHMF under different flow rates.[a]
Flow rate
Residence time
(min)
5-HMF
Entry
Yield (%)[b]
(mL.min-1)
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 32
1
2
3
4
5
6
7
8
9
1
2
4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.75
0.50
0.25
0.70
0.80
0.93
1.12
1.40
1.90
2.80
3.73
5.60
11.2
17.95
22.45
26.10
35.40
48.70
61.35
72.50
84.23
86.60
95.10
3
4
5
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
6
7
8
9
10[c]
[a]Flow Conditions: 45 g of Dꢀfructose in 1 L of a iꢀPrOH/DMSO solution (15 % v/v)
using a column packed with Amberlystꢀ15 at 110 °C and a BPR of 5 bar. [b]The
1
yields were determined by H NMR analysis using 1,3,5ꢀtrimethoxybenzene as
internal standard. [c] Complete conversion for Dꢀfructose was observed by H NMR
1
analysis of the reaction mixture using D2O.
From the Table 1, it is clear that higher flow rates decrease the yield for 5ꢀHMF (entries 1, 2
and 3). In sharp contrast, lower flow rates led to the opposite reaction outcome, affording
superior 5ꢀHMF yields due to the higher contact time between Dꢀfructose and the catalyst
(entries 8, 9 and 10). As expected according to our previous study,21 in the optimum reaction
condition, 5ꢀHMF was obtained in 95% yield in just 11.2 minutes of residence time at a flow rate
of 0.25 mL.minꢀ1 with total conversion of Dꢀfructose (entry 10). The amount of DMSO in iꢀ
PrOH, catalyst loading, flow rate and temperature were not further refined to improve the HMF
yield/selectivity once these attempts were already evaluated previously, showing that this result
cannot be improved beyond 95%.21 For example, a longer residence time could increase the
formation of the main byproduct (iꢀpropoxymethylfurfural) and the resin Amberlystꢀ15 is not
stable above 120 °C, as well as higher amounts of DMSO are not attractive for a sustainable
process.27
For each flow rate presented in the Table 1, NIR spectra were also inꢀline recorded and
1
aliquots of the output stream were collected and analyzed by quantitative H NMR analysis
ACS Paragon Plus Environment
12

Page 13 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
(qHNMR). Figure 2 shows the overlapped spectra and the excellent response of microNIR to the
10 different flow rates. The 1H NMR spectra obtained for each flow rate (4 to 0.25 mL.minꢀ1) is
also provided on the ESI (Figure S6). The advantage of using qHNMR as reference method is
that there is no requirement to construct an analytical curve for the development of the analytical
method, which makes the developed method simpler compared to chromatographic techniques,
for example. In addition, the quantification of the chemical yield could also be done by an
indirect method, for example, by the quantification of water by Karl Fisher titration. (remover?)
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. Raw microNIR spectra obtained for the dehydration reaction of Dꢀfructose into 5ꢀ
HMF under 10 different flow rates.
From these results, it was possible to construct a multivariate calibration model (PLS model),
in which yield values obtained with the NMR analyses were used as reference values for the NIR
spectra calibration (Table 2). PLS model plot, which relates the reference values obtained by
qHNMR analyses with the predicted values provided by NIR spectroscopy analyses, is also
provided on the ESI (Figure S7).
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 32
1
2
3
4
5
6
7
8
9
Table 2. Root mean square errors and determination of coefficients for PLS model.
Calibration
Validation
Prevision
PLS Model
RMSE
R2
1.71[a]
1.94[a]
2.14[a]
0.9958
0.9950
0.9938
[a] RMSEs in % of yields.
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
As can be seen in Table 2, the model presented values for the determination of coefficients
(R2) higher than 0.99 and the values obtained for root mean squared errors (RMSEs) were close
to 2% yield, which is in agreement with the accuracy usually obtained by the reference method
(qHNMR).28 Both values show a high adjustment between the reference method and the
secondary method (microNIR). Therefore, we show that there is a correlation between these
methods and the yields of the reaction, consequently, can be monitored by NIR spectroscopy.
Thus, the PLS model could be used to verify when the reaction process achieves the optimum
condition as well as when a process deviation occurs in the continuous flow regime. However, in
the present study, we chose to use PCA model by the construction of control charts to perform
the inꢀline reaction monitoring, since this approach is easier to be implemented in reaction
systems, with a minor necessity of constant recalibrations.
Although the microNIR device showed excellent response by the application of different flow
rates and offering opportunities to construct the PLS model, all these experiments were
performed during a short period (same day) and the evaluation of reproducibility of the device
cannot be drawn under these circumstances. In this context, to demonstrate the reproducibility of
the microNIR device, the catalyst stability/recyclability, and also the process robustness, a longꢀ
term experiment was conducted over a period of 5 days (8 hours per day) using the same column
filled with Amberlystꢀ15. At this part, the optimum reaction condition (0.25 mL.minꢀ1) for 5ꢀ
ACS Paragon Plus Environment
14

Page 15 of 32
Organic Process Research & Development
1
2
3
4
HMF synthesis was used and continuously monitored in an experiment of 40 hours. The NIR
5
6
spectra were collected every 45 minutes (Figure 3).
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 3. A) raw microNIR spectra obtained for the variation of the flow rate (4.0 to 0.25
mL.minꢀ1); B) raw microNIR spectra obtained during the process monitoring over 40 hours (at
0.25 mL.minꢀ1 flow rate); C) first derivative of the spectra obtained for the variation of the flow
rate; D) first derivative of the spectra obtained during the process monitoring over 40 hours.
Raw microNIR spectra are presented in the Figure 3A and 3B. In Figure 3A is shown the
overlapping of raw spectra obtained in different days of measurements varying the flow rate
from 4.0 to 0.25 mL.minꢀ1. In Figure 3B is shown the overlapping of all raw spectra obtained
during the 40 hours of inꢀline monitoring at flow rate of 0.25 mL.minꢀ1. Note that the both raw
spectra present an offset in its baseline and this behavior is typical for single beam equipment,
due to variations in the temperature of the radiation source and the detector.17 Thus, the spectral
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 32
1
2
3
4
5
6
7
8
9
baseline offset was corrected with the application of the first derivative and the corrected spectra
can be seen in the Figure 3C and 3D. Figure 3C confirms the good reproducibility of the
microNIR device since there is only a variation in the spectral range of 1875ꢀ2010 nm, which
corresponds to different chemical yields due to the different residence times at which the Dꢀ
fructose solution was exposed to the catalyst. Figure 3D shows the spectra of the reaction
monitoring after correction of the baseline offset and no significant variation in the region from
1870 to 2010 nm was observed, which indicates that the reaction system remained stable during
this period (40 h).
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
Next, Principal Component Analysis (PCA)29,30 was used as a tool for exploratory analysis.
The PCA reduces the information contained in all wavelengths for only a few principal
components (PC). The correlated information from different wavelengths is clustered into the
same PC. Furthermore, the first PC contains correlated information of the wavelength with the
greatest spectral variance, and this variance explained by each PC decreases with the increase in
the amounts of PCs until all variance is explained. The original information from the dataset is
then projected onto the space of the principal components.29,30 The Figure 4A and 4B show the
scores and loadings for the PCA model constructed using both date set as explained in the
experimental section. Figure 4C and 4D show the scores and loadings for the PCA model
constructed using only the date from the inꢀline reaction monitoring at 0.25 mL.minꢀ1 by 40
hours.
ACS Paragon Plus Environment
16

Page 17 of 32
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 4. A) scores and; B) loadings obtained for the PCA of all sets of samples (flow rates from
4.0 to 0.25 mL.minꢀ1); C) scores and; (D) loadings obtained for the PCA of samples from the
reaction monitoring at 0.25 mL.minꢀ1 over 40 hours).
The Figure 4A shows that the scores are distributed along PCꢀ1 according to the
concentration of 5ꢀHMF (or its chemical yield) obtained from different flow rates. In addition,
looking at the loadings of PCꢀ1 in the Figure 4B it becomes clear that the wavelengths which
have greater relevance for this component are precisely those with greater variation in the 1870ꢀ
2010 nm range. Furthermore, PCꢀ2 present small random variations of the spectrum, i.e. it
carries part of the noise information. These two PCs in combination explain 96% of the variance
of the dataset. The Figure 4C shows the scores for measures of the reaction monitoring, which
exhibit a random distribution of the samples in PCꢀ1 and PCꢀ2, each one have 65% and 25% and
explain 90% of the variance of the dataset. The loadings shown in Figure 4D also display only
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 32
1
2
3
4
5
6
7
8
9
information relating to the spectral fluctuations, without having a large spectral variance. The
results presented for this PCA model are expected, because the spectral variations in this dataset
are caused from the instrumental noises, having little influence from reaction yields fluctuation.17
Although the PCA model developed describes well our data, only a visual evaluation of
scores and loadings cannot be considered to assess the stability of the reaction system. One of the
methods used to address the presence of abnormal variations in a reaction process is based in the
control chart construction. Bearing this in mind, and after the reduction of the variables using
PCA, we constructed a multivariate control chart using the Hotelling’s T2 and Qꢀresidual to
verify catalyst stability over the 40 hours of inꢀline reaction monitoring.31ꢀ33 To this end, a PCA
model was constructed to obtain the limit values of the Hotelling’s T2 and Qꢀresidual with 10
samples, which were analyzed by the qHNMR method. The limit values obtained and the
measures taken during the inꢀline monitoring are shown in Figure 5A e 5B. This was done to
ensure that the limits obtained to each control chart were calculated with spectra of samples that
present the maximum reaction yield.
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
ACS Paragon Plus Environment
18

Page 19 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
Figure 5. Multivariate control chart building with two principal components based on A)
Hotelling’s T2; and B) Qꢀresidual for monitoring of the reaction over 40 hours (2400 min). The
red dots in both charts show the intentional deviation in the process.
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
Hotelling’s T2 and Qꢀresidual values are applied to detect anomalous samples when PCA is
used for exploratory data analysis.31ꢀ33 In the monitoring of a process, these parameters inform
when a sample is outside of the control limits. Hotelling’s T2 values are calculated with the
information from the first PCs and represent how far a sample is from the center of the PCA
model. In this context, the control chart is sensitive to samples that have different absorbance
intensities in the same wavelengths, with greater variance, used in the construction of the PCA
model.31ꢀ33 Qꢀresidual values are a complementary tool since they are calculated from
information not included in the PCA model. Hence, its control chart is useful to detect samples
that have different spectral profiles from those which were used in the PCA model construction.
Therefore, these values can inform when the reaction progress is out of control.31ꢀ33 For example,
if at any given time a measure has a value greater than the limit for Hotelling’s T2 (red line in
Figure 5A), this indicates that there was a change in the reaction yield and, possibly, there was a
variation in the flow rate or temperature. On the other hand, when a measure presents a value
greater than the limit for Qꢀresidual, this indicates that there was some type of contamination in
the system or the formation of a byꢀproduct occurred. In that way, when the measure has both
values greater than the limits, a combination of these effects may be occurring, which indicates
that there has a process deviation in the reaction system.
Therefore, to verify the reliability of the control chart presented herein, we simulated a
process deviation by changing the reaction yield from 95% (0.25 mL.minꢀ1) to 84% (0.75
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 32
1
2
3
4
5
6
7
8
9
mL.minꢀ1). Note in the Figure 5 (highlighted in the red dots ) that after 1500 min of monitoring
at 0.25 mL.minꢀ1, we disturbed the system by changing the flow rate to 0.75 mL.minꢀ1 and the
control chart using Hotelling’s T2 alerted us about the process deviation. As described before, a
deviation on the yield only can be detected by the Hotelling’s T2 chart. Indeed, disregarding the
simulation, all measurements are below the limit value in both control charts. This indicates that
the system remained stable during the 40 hours of inꢀline monitoring by the compact microNIR
device using the same column filled with Amberlystꢀ15 at flow rate of 0.25 mL.minꢀ1. Over this
time, 27 g of DꢀFructose were processed and produced 19 g of 5ꢀHMF in 95% yield, which
correspond to a Dꢀfructose/catalyst ratio of 6 mol%.
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
Conclusions
In conclusion, microNIR spectroscopy was successfully applied for the inꢀline monitoring of
the conversion of Dꢀfructose into 5ꢀHMF in a continuous flow fixedꢀbed microreactor. Our work
also demonstrates that portable NIR spectrophotometers can be used as an alternative analytical
tool to benchtop equipment with highꢀcost and normally used for inꢀline reaction monitoring
studies.7ꢀ14 MicroNIR device give us valuable insights on the events occurring within the process
by the construction of Hotelling’s T2 and Qꢀresidual control charts, such as the high stability of
the Amberlystꢀ15, which was used by 8 hours a day, 5 days a week for 5ꢀHMF synthesis. These
results also demonstrate the feasibility of applying a continuous flow system to produce 5ꢀHMF
at the laboratory scale, which could be expanded for future industrial application. Finally, our
findings confirm this lowꢀcost and portable device as a promising tool for inꢀline reaction
monitoring in continuous flow regime, which can be extended to other types of reaction process.
ACS Paragon Plus Environment
20

Page 21 of 32
Organic Process Research & Development
1
2
3
4
Acknowledgments
5
6
7
8
9
The authors gratefully acknowledge financial support from the São Paulo Research Foundation –
FAPESP (J.C.P., awards No. 2014/26378ꢀ2 and 2014/25770ꢀ6), CNPq (J.C.P., award No.
453862/2014ꢀ4), and Instituto Nacional de Ciências e Tecnologias Analíticas Avançadas
(INCTAA – FAPESP award No. 2008/57808ꢀ1 and CNPq award No. 573894/2008ꢀ6). We also
thank CAPES (R.G.) for fellowship. We are also thankful to the Obesity and Comorbidities
Research Center – OCRC (FAPESP, award No. 2013/07607ꢀ8) for providing us with the
UNIQSIS flow system used in this work.
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
Supporting Information Available. H NMR spectra for different flow rates, PLS plot
and pictures of the setꢀup showing the microNIR coupled to the flow system are presented.
Notes and References
1. (a) T. M. Kohl, B. Bizet, P. Kevan, C. Sellwood, J. Tsanaktsidisa and C. H. Hornunga.
Efficient synthesis of 5ꢀ(chloromethyl)furfural (CMF) from high fructose corn syrup (HFCS)
using continuous flow processing. React. Chem. Eng., 2017, 2, 541. (b) C. Schotten, D. Plaza,
S. Manzini, S. P. Nolan, S. V. Ley, D. L. Browne and A. Lapkin. Continuous Flow Metathesis
for Direct Valorization of Food Waste: An Example of Cocoa Butter Triglyceride. ACS Sustain
Chem Eng., 2015, 7, 1453. (c) J. M. Tukacs, R. V. Jones, F. Darvas, G. Dibó, G, Lezsák and L.
T. Mika. Synthesis of γꢀvalerolactone using a continuousꢀflow reactor. RSC Adv., 2013, 3,
16283. (d) F. Ferlin, S. Santoro, L. Ackermann and L. Vaccaro. Heterogeneous C–H
ACS Paragon Plus Environment
21

Organic Process Research & Development
Page 22 of 32
1
2
3
4
5
6
7
8
9
alkenylations in continuousꢀflow: oxidative palladiumꢀcatalysis in a biomassꢀderived reaction
medium. Green Chem., 2017, 19, 2510. (e) T. M. Kohl, B. Bizet, P. Kevan, C. Sellwood, J.
Tsanaktsidis and C. H. Hornung. Efficient synthesis of 5ꢀ(chloromethyl)furfural (CMF) from
high fructose corn syrup (HFCS) using continuous flow processing. React. Chem. Eng., 2017,
2, 541. (f) GꢀY. Jeong, A. K. Singh, S. Sharma, K. W. Gyak, R. A. Maurya and DꢀP. Kim.
Oneꢀflow syntheses of diverse heterocyclic furan chemicals directly from fructose via tandem
transformation platform. NPG Asia Materials., 2015, 7, e173. (g) J. M. Bermudez, J. A.
Menéndez, A. A. Romero, E. Serrano, J. GarciaꢀMartinez and R. Luque. Continuous flow
nanocatalysis: reaction pathways in the conversion of levulinic acid to valuable chemicals.
Green Chem., 2013, 15, 2786. (h) M. H. Tucker, A. J. Crisci, B. N. Wigington, N. Phadke, R.
Alamillo, J. Zhang, S. L. Scott and J. A. Dumesic. AcidꢀFunctionalized SBAꢀ15ꢀType Periodic
Mesoporous Organosilicas and Their Use in the Continuous Production of 5ꢀ
Hydroxymethylfurfural. ACS Catal., 2012, 9, 1865.
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. (a) S. G. Newman and K. F. Jensen. The role of flow in green chemistry and engineering.
Green Chem., 2013, 15, 1456. (b) K. S. Elvira, X. C. Solvas, R. C. R. Wootton and A. J.
deMello. The past, present and potential for microfluidic reactor technology in chemical
synthesis. Nat Chem., 2013, 5, 905. (c) D. E. Fitzpatrick, C. Battilocchio and S. V. Ley.
Enabling Technologies for the Future of Chemical Synthesis. ACS Central Science., 2016, 2,
131. (d) T. D. McQuade and P. H. Seeberger. Applying Flow Chemistry: Methods, Materials,
and Multistep Synthesis. J. Org. Chem. 2013., 78, 6384.
3. (a) H. I. Furkan and B. C. Remzi. Lignocellulosic biomass: a sustainable platform for the
production of bioꢀbased chemicals and polymers. Polym. Chem., 2015, 6, 4497. (b) R. A.
Sheldon. Green and sustainable manufacture of chemicals from biomass: state of the art.
ACS Paragon Plus Environment
22

Page 23 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Green Chem., 2014, 16, 950. (c) J. Becker, A. Lange, J. Fabarius and C. Wittmann. Top
value platform chemicals: bioꢀbased production of organic acids. Curr Opin Biotechnol.,
2015, 36, 168. (d) J. Haveren, E. L. Scott and J. Sanders. Bulk chemicals from biomass.
Biofuel Bioprod Bior., 2008, 2, 41.
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. (a) S. Basudeb, M. and AꢀO Mahdi. Advances in 5ꢀhydroxymethylfurfural production from
biomass in biphasic solvents. Green Chem., 2014, 16, 24. (b) P. T. Siew, Y. Guangshun and
Z. Yugen. Hydroxymethylfurfural production from bioresources: past, present and future.
Green Chem., 2014, 16, 2015. (c) R. C. Benjamin, E. T. Rodrigo, G. K. Kurtis and T. R.
Ronald. Biomass to Furanics: Renewable Routes to Chemicals and Fuels. ACS Sustainable
Chem. Eng., 2015, 3, 2591. (d) W. Tianfu, W. N. Michael and H. S. Brent. Catalytic
dehydration of C6 carbohydrates for the production of hydroxymethylfurfural (HMF) as a
versatile platform chemical. Green Chem., 2014, 16, 548.
5. R. Noma, K. Nakajima, K. Kamata, M. Kitan, S. Hayashi and M. Hara. Formation of 5ꢀ
(Hydroxymethyl)furfural by Stepwise Dehydration over TiO2 with WaterꢀTolerant Lewis
Acid Sites. J. Phys. Chem. C., 2015, 119, 17117.
6. (a) V. Sans and L. Cronin. Towards dialꢀaꢀmolecule by integrating continuous flow, analytics
and selfꢀoptimisation. Chem. Soc. Rev., 2016, 45, 2032. (b) D. C. Fabry, E. Sugiono and M.
Rueping. Online monitoring and analysis for autonomous continuous flow selfꢀoptimizing
reactor systems. React. Chem. Eng., 2016, 1, 129.
7. (a) A. Lapkin and D. J. C. Constable. Green Chemistry Metrics: Measuring and Monitoring
Sustainable Processes, WileyꢀBlackwell, Chichester, 2008, 344 pages. (b) D. E. Fitzpatrick,
C. Battilocchio, S. V. Ley. Org. Process. Res. Dev., 2016, 20, 386. (c) J. S. Moore and K. F.
Jensen. Angew Chem Int Ed Engl., 2014, 53, 470. (d) J. S. Moore and K. F. Jensen. Org.
ACS Paragon Plus Environment
23

Organic Process Research & Development
Page 24 of 32
1
2
3
4
5
6
Process Res. Dev., 2012, 16, 1409. (e) D. Helmdach, P. Yaseneva, P. K. Heer, A. M.
Schweidtmann and A. A. Lapkin. ChemSusChem., 2017, 10, 3632.
7
8
9
8. (a) C. F. Carter, H. Lange, I. R. Baxendale, S. V. Ley, J. Goode, N. Gaunt and B. Wittkamp.
ReactIR Flow Cell: A New Analytical Tool for Continuous Flow Chemical Processing. Org.
Process Res. Dev. 2010., 14, 393. (b) A. Perro, G. Lebourdon, S. Henry, S. Lecomte, L.
Servanta and S. Marre. Combining microfluidics and FTꢀIR spectroscopy: towards spatially
resolved information on chemical processes. React. Chem. Eng., 2016, 1, 577. (c) A. G.
O'Brien, O. R. Luca, P. S. Baran and D. G. Blackmond. In Situ FTIR Spectroscopic
Monitoring of Electrochemically Controlled Organic Reactions in a Recycle Reactor. React.
Chem. Eng., 2016, 1, 90. (d) T. Durand, C. Henry, D. Bolien, D. C. Harrowven, S.
Bloodworth, X. Franck and R. J. Whitby. Thermolysis of 1,3ꢀdioxinꢀ4ꢀones: fast generation
of kinetic data using inꢀline analysis under flow. React. Chem. Eng., 2016, 1, 82. (e) J. S.
Moore, C. D. Smithac and K. F. Jensen. Kinetics analysis and automated online screening of
aminocarbonylation of aryl halides in flow. React. Chem. Eng., 2016, 1, 272. (f) JꢀS Poh, D.
L. Browne and S. V. Ley. A Multistep Continuous Flow Synthesis Machine for the
Preparation of Pyrazoles via a MetalꢀFree AmineꢀRedox Process. React. Chem. Eng., 2016,
1, 101.
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
9. (a) S. A. Leung, R. F. Winkle, R. C. Wootton and A. J. deMello. A method for rapid reaction
optimisation in continuousꢀflow microfluidic reactors using online Raman spectroscopic
detection. Analyst., 2005, 130, 46. (b) S. Schwolow, F. Braun, M. Radle, N. Kockmann and
T. Roder. Fast and Efficient Acquisition of Kinetic Data in Microreactors Using InꢀLine
Raman Analysis. Org. Process Res. Dev., 2015, 19, 1286. (c) S. Mozharov, A. Nordon, D.
Littlejohn, C. Wiles, P. Watts, P. Dallin and J. M. Girkin. Improved Method for Kinetic
ACS Paragon Plus Environment
24

Page 25 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Studies in Microreactors Using Flow Manipulation and Noninvasive Raman Spectrometry. J.
Am. Chem. Soc., 2011, 133, 3601. (d) T. A. Hamlin and N. E Leadbeater. Raman
spectroscopy as a tool for monitoring mesoscale continuousꢀflow organic synthesis:
Equipment interface and assessment in four medicinallyꢀrelevant reactions. Beilstein J. Org.
Chem., 2013, 9, 1843. (e) E. Cao, M. Sankar, S. Firth, K. F. Lam, D. Bethell, D. K. Knight,
G. J. Hutchings, P. F. McMillan and A. Gavriilidis. Reaction and Raman spectroscopic
studies of alcohol oxidation on gold–palladium catalysts in microstructured reactors. Chem.
Eng. J., 2011, 167, 734. (f) M. Lee, JꢀP. Lee, H. Rhee, J. Choo, C. G. Yong and E. K. Lee.
Applicability of laserꢀinduced Raman microscopy for in situ monitoring of imine formation
in a glass microfluidic chip. J Raman Spectrosc., 2003, 34, 737. (g) L. Guerrini and D.
Graham. Molecularlyꢀmediated assemblies of plasmonic nanoparticles for SurfaceꢀEnhanced
Raman Spectroscopy applications. Chem. Soc. Rev., 2012, 41, 7085. (h) M. P. Cecchini, J.
Hong, C. Lim, J. Choo, T. Albrecht, A. J. deMello and J. B. Edel. Ultrafast Surface Enhanced
Resonance Raman Scattering Detection in DropletꢀBased Microfluidic Systems. Anal.
Chem., 2011, 83, 3076. (i) Pinho B. and Hartman. R. L. Microfluidics with in situ Raman
spectroscopy for the characterization of nonꢀpolar/aqueous interfaces. React. Chem. Eng.,
2017, 2, 189.
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
10. (a) S. Krishnadasan, R. J. C. Brown, A. J. deMello and J. C. Mello. Intelligent routes to the
controlled synthesis of nanoparticles. Lab Chip., 2007, 7, 1434. (b) V. Sans, S. Glatzel, F. J.
Douglas, D. A. Maclaren, A. Lapkin and L. Cronin. Nonꢀequilibrium dynamic control of
gold nanoparticle and hyperꢀbranched nanogold assemblies. Chem. Sci., 2014, 5, 1153. (c) P.
J. Kitson, M. H. Rosnes, V. Sans, V. Dragone and L. Cronin. Configurable 3DꢀPrinted
millifluidic and microfluidic 'lab on a chip' reactionware devices. Lab Chip., 2012, 12, 3267.
ACS Paragon Plus Environment
25

Organic Process Research & Development
Page 26 of 32
1
2
3
4
5
6
(d) H. Lu, M. A. Schmidt and K. F. Jensen. Photochemical reactions and onꢀline UV
detection in microfabricated reactors. Lab Chip., 2001, 1, 22.
7
8
9
11. (a) J. P. McMullen and K. F. Jensen. An Automated Microfluidic System for Online
Optimization in Chemical Synthesis. Org. Process Res. Dev., 2010, 14, 1169. (b) J. P.
McMullen, M.T. Stone, S. L. Buchwald and K. F. Jensen. An Integrated Microreactor
System for SelfꢀOptimization of a Heck Reaction: From Microꢀ to Mesoscale Flow Systems.
Angew. Chem., Int. Ed., 2010, 49, 7076. (c) B. J. Reizman and K. F. Jensen. An Automated
ContinuousꢀFlow Platform for the Estimation of Multistep Reaction Kinetics. Org. Process
Res. Dev., 2012, 16, 1770. (d) B. J. Reizman, YꢀM. Wang, S. L. Buchwald and K. F. Jensen.
Suzuki–Miyaura crossꢀcoupling optimization enabled by automated feedback. React. Chem.
Eng., 2016, 1, 658. (e) T. C. Malig, J. D. B. Koenig, H. Situ, N. K. Chehal, P. G. Hultin and
J. E. Hein. Realꢀtime HPLCꢀMS reaction progress monitoring using an automated analytical
platform. React. Chem. Eng., 2017, 2, 309. (f) N. Holmes, G. R. Akien, A. J. Blacker, R. L.
Woodward, R. E. Meadows and R. A. Bourne. Selfꢀoptimisation of the final stage in the
synthesis of EGFR kinase inhibitor AZD9291 using an automated flow reactor. React. Chem.
Eng., 2016, 1, 366.
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
12. (a) A. J. Parrott, R. A. Bourne, G. R. Akien, D. J. Irvine and M. Poliakoff. SelfꢀOptimizing
Continuous Reactions in Supercritical Carbon Dioxide. Angew. Chem. Int. Ed., 2011, 50,
3788. (b) R. A. Bourne, R. A. Skilton, A. J. Parrott, D. J. Irvine and M. Poliakoff, P.
Adaptive Process Optimization for Continuous Methylation of Alcohols in Supercritical
Carbon Dioxide. Org. Process Res. Dev., 2011, 15, 932. (c) A. McNally, C. K. Prier, D. W.
C. MacMillan. Discovery of an αꢀamino CꢀH arylation reaction using the strategy of
accelerated serendipity. Science., 2011, 334, 1114.
ACS Paragon Plus Environment
26

Page 27 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
13. (a) J. J. Haven, J. Vandenbergh and T. Junkers. Watching polymers grow: real time
monitoring of polymerizations via an onꢀline ESIꢀMS/microreactor coupling. Chem.
Commun., 2015, 51, 4611. (b) N. Holmes, G. R. Akien, R. J. D. Savage, C. Stanetty, I. R.
Baxendale, A. J. Blacker, B. A. Taylor, R. L. Woodward, R. E. Meadowse and R. A. Bourne.
Online quantitative mass spectrometry for the rapid adaptive optimisation of automated flow
reactors. React. Chem. Eng., 2016, 1, 96. (c) T. W. T. Bristow, A. D. Ray, A. O’Kearneyꢀ
McMullan, L. Lim, B. McCullough, A. Zammataro. Utility of Ion Mobility Mass
Spectrometry for DrugꢀtoꢀAntibody Ratio Measurements in AntibodyꢀDrug Conjugates. J.
Am. Soc. Mass Spectrom., 2014, 25, 1794. (d) D. L. Browne, S. Wright, B.J. Deadman, S.
Dunnage, I. R. Baxendale, R. M. Turner and S. V. Ley. Continuous flow reaction monitoring
using an onꢀline miniature mass spectrometer. Rapid Commun. Mass Spectrom., 2012, 26,
1999. (e) J. S. Mathieson, M. H. Rosnes, V. Sans, P. J. Kitson and L. Cronin. Continuous
parallel ESIꢀMS analysis of reactions carried out in a bespoke 3D printed device. Beilstein J.
Nanotechnol., 2013, 4, 285. (f) M. R. Kristyn, Z. Xing, X. L. Shelly, H. G. Gilles, C.
Guilong, Z. Zhongli, F. S. William and H. H. Herbert. Real Time Pharmaceutical Reaction
Monitoring by Electrospray Ion MobilityꢀMass Spectrometry. Int. J. Mass Spectrom., 2013,
336, 27.
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
14. (a) V. Sans, L. Porwol, V. Dragone and L. Cronin. A self optimizing synthetic organic
reactor system using realꢀtime inꢀline NMR spectroscopy. Chem. Sci., 2015, 6, 1258. (b) E.
Danieli, J. Perlo, A. L. L. Duchateau, G. K. M. Verzijl, V. M. Litvinov, B. Blumich, F.
Casanova. OnꢀLine Monitoring of Chemical Reactions by using BenchꢀTop Nuclear
Magnetic Resonance Spectroscopy. ChemPhysChem., 2014, 15, 3060. (c) T. H. Rehm, C.
Hofmann, D. Reinhard, HꢀJ. Kost, P. Löb, M. Besold, K. Welzel, J. Barten, A. Didenko, D.
ACS Paragon Plus Environment
27

Organic Process Research & Development
Page 28 of 32
1
2
3
4
5
6
7
8
9
V. Sevenard, B. Lix, A. R. Hillson and S. D. Riegel. Continuousꢀflow synthesis of fluorineꢀ
containing fine chemicals with integrated benchtop NMR analysis. React. Chem. Eng., 2017,
2, 315. (d) M. V. Gomez, H. H. J. Verputten, A. DiazꢀOrtiz, A. Moreno, A. de la Hoz and A.
H. Velders. Onꢀline monitoring of a microwaveꢀassisted chemical reaction by nanolitre
NMRꢀspectroscopy. Chem. Commun., 2010, 46, 4514. (e) J. Bart, A. J. Kolkman, A. J Oꢀde.
Vries, K. Koch, P. J. Nieuwland, H. Janssen, J. van Bentum, K. A. M. Ampt, F. P. J. T.
Rutjes, S. S. Wijmenga, H. Gardeniers and A. P. M. Kentgens. A Microfluidic Highꢀ
Resolution NMR Flow Probe. J. Am. Chem. Soc., 2009, 131, 5014. (f) F. Dalitz, M. Cudaj,
M. Maiwald and G. Guthausen. Process and reaction monitoring by lowꢀfield NMR
spectroscopy. Prog. Nucl. Magn. Reson. Spectrosc., 2012, 60, 52. (g) A. Nordon, A. Diezꢀ
Lazaro, C. W. L. Wong, C. A. McGill, D. Littlejohn, M. Weerasinghe, D. A. Mamman, M.
L. Hitchman and J. Wilkie. Consideration of some sampling problems in the onꢀline analysis
of batch processes by lowꢀfield NMR spectrometry. Analyst., 2008, 133, 339. (h) E. Danieli,
J. Perlo, B. Blumich and F. Casanova. Small magnets for portable NMR spectrometers.
Angew. Chem., Int. Ed., 2010, 49, 4133. (i) J. Y. Buser and A. D. McFarland. Reaction
characterization by flow NMR: quantitation and monitoring of dissolved H2 via flow NMR at
high pressure. Chem. Commun., 2014, 50, 4234. (j) M. V. Gomez, A. M. Rodriguez, A. de la
Hoz, F. JimenezMarquez, R. M. Fratila, P. A. Barneveld and A. H. Velders. Determination of
Kinetic Parameters within a Single Nonisothermal OnꢀFlow Experiment by Nanoliter NMR
Spectroscopy. Anal. Chem., 2015, 87, 10547. (k) S. S. Zalesskiy, E. Danieli, B. Blümich and
V. P. Ananikov. Miniaturization of NMR Systems: Desktop Spectrometers, Microcoil
Spectroscopy, and “NMR on a Chip” for Chemistry, Biochemistry, and Industry. Chem. Rev.,
2014, 114, 5641.
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
ACS Paragon Plus Environment
28

Page 29 of 32
Organic Process Research & Development
1
2
3
4
5
15. I. Lewis, H. Edwards. Handbook of Raman Spectroscopy. 1st Ed, Boca Raton: CRC Press,
2001, 1072 pages.
6
7
8
9
16. (a) A. Chanda, A. M. Daly, D. A. Foley, M. A. LaPack, S. Mukherjee, J. D. Orr, G. L.
Reid, D. R. Thompson and H. W. Ward. Industry Perspectives on Process Analytical
Technology: Tools and Applications in API Development. Org. Process Res. Dev., 2015, 19,
63. (b) N. N. Misra, Carl Sullivan and P. J. Cullen. Process Analytical Technology (PAT)
and Multivariate Methods for Downstream Processes. Current Biochemical Engineering,
2015, 2, 4. (c) S. Challa and R. Potumarthi. Chemometricsꢀbased process analytical
technology (PAT) tools: applications and adaptation in pharmaceutical and
biopharmaceutical industries. Appl Biochem Biotechnol., 2013, 169, 66.
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
17. C. Pasquini. Near Infrared Spectroscopy: fundamentals, practical aspects and analytical
applications. Braz. Chem. Soc., 2003, 14, 198.
18. A. Gałuszka, Z. M. Migaszewski, J. Namieśnik. Moving your laboratories to the fieldꢀꢀ
Advantages and limitations of the use of field portable instruments in environmental sample
analysis. Environ. Res., 2015, 140, 593.
19. L. Sun, C. Hsiung, C. G. Pederson, P. Zou, V. Smith, M. Gunten and N. A. O’Brien.
Qualitative and Quantitative Pharmaceutical Analysis with a Novel Handheld Miniature
nearꢀInfrared Spectrometer. J. Near Infrared Spectrosc., 2013, 21, 445.
20. (a) O. Kolomiets, U. Hoffmann, P. Geladi and H. W. Siesler. Quantitative determination of
pharmaceutical drug formulations by nearꢀinfrared spectroscopic imaging. Appl Spectrosc.,
2008, 62, 1200. (b) C. A. dos Santos, M. Lopo, R. N. Páscoa and J. A. Lopes. A review on
the applications of portable nearꢀinfrared spectrometers in the agroꢀfood industryAppl
Spectrosc., 2013, 67, 1215. (c) R. L. Ribessi, T. A. Neves, J. J. R. Rohwedder, C. Pasquini,
ACS Paragon Plus Environment
29