Mild and selective reduction of aldehydes utilising sodium dithionite under flow conditions

Article abstract of DOI:10.3762/bjoc.14.129

We recently reported a novel hybrid batch-flow synthesis of the antipsychotic drug clozapine in which the reduction of a nitroaryl group is described under flow conditions using sodium dithionite. We now report the expansion of this method to include the reduction of aldehydes. The method developed affords yields which are comparable to those under batch conditions, has a reduced reaction time and improved space-time productivity. Furthermore, the approach allows the selective reduction of aldehydes in the presence of ketones and has been demonstrated as a continuous process.

Full text of DOI:10.3762/bjoc.14.129

Mild and selective reduction of aldehydes utilising sodium
dithionite under flow conditions
Nicole C. Neyt and Darren L. Riley*
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2018, 14, 1529–1536.
Department of Chemistry, University of Pretoria, Pretoria 0002, South
Africa
doi:10.3762/bjoc.14.129
Received: 05 March 2018
Accepted: 04 June 2018
Published: 22 June 2018
Email:
Darren L. Riley* - darren.riley@up.ac.za
*
Corresponding author
Associate Editor: I. R. Baxendale
Keywords:
© 2018 Neyt and Riley; licensee Beilstein-Institut.
License and terms: see end of document.
aldehyde reduction; flow chemistry; selective reduction; sodium
dithionite
Abstract
We recently reported a novel hybrid batch–flow synthesis of the antipsychotic drug clozapine in which the reduction of a nitroaryl
group is described under flow conditions using sodium dithionite. We now report the expansion of this method to include the reduc-
tion of aldehydes. The method developed affords yields which are comparable to those under batch conditions, has a reduced reac-
tion time and improved space-time productivity. Furthermore, the approach allows the selective reduction of aldehydes in the pres-
ence of ketones and has been demonstrated as a continuous process.
Introduction
Flow chemistry and continuous processing has been acknowl- These are all beneficial in moving towards more efficient and
edged by the American Chemical Society (ACS), the Green sustainable techniques in chemical processing [3,4,6,7].
Chemistry Institute (GCI) and several global pharmaceutical
companies as one of the primary areas for research and develop- The reduction of carbonyl groups are a standard type of trans-
ment for chemical manufacturing [1-4]. For the past decade formation in organic synthesis, however, to date under flow
flow chemistry and the application of flow devices has been conditions reductions have mostly been limited to soluble
gaining acceptance in laboratories because of its ease of use, reducing agents like DIBAL [7-10] and superhydride [11]
safety and control [3-6]. Continuous flow technologies are gen- which require special handling or expensive solid-supported
erally more effective than traditional batch processes with key borohydride species [12]. Recently, Seeberger and co-workers
advantages including intensified heat and mass transfer, inline demonstrated a more cost-effective sodium borohydride-medi-
reaction monitoring, higher mass throughput, safer control of ated flow reduction utilizing solid mixes of sodium borohy-
hazardous chemicals increasing lab safety and direct scalability. dride, lithium chloride and celite [12], and the Ley group were
1529

Beilstein J. Org. Chem. 2018, 14, 1529–1536.
able to demonstrate a green transfer hydrogenation of ketones [17], benzil groups [19], nitroarenes and nitroalkanes in the
under flow using catalytic lithium tert-butoxide in isopropanol presence of dialkyl viologen electron transfer catalysts [20,21]
[
13].
and immobilized nitroarene’s under phase transfer conditions
22,23].
[
We recently published a batch–flow hybrid synthesis of the
antipsychotic drug clozapine in which we demonstrated a flow- In this publication we report the efficient reduction of alde-
based reduction of a nitro group utilising sodium dithionite as a hydes under flow conditions utilising sodium dithionite under
reductant [14]. We further hypothesized that the development of basic conditions and the expansion of the approach to allow the
a flow protocol for the reduction of carbonyl groups would also selective reduction of aldehydes in the presence of ketones.
be possible using sodium dithionite (Scheme 1) [15,16].
Results and Discussion
Batch-based reduction of aldehydes and
Sodium dithionite can act as a powerful reducing agent [16], is
mild, easy to use, and it is compatible with protic solvents like ketones
water and isopropanol. It has previously been used to reduce a The reduction of simple aldehydes and ketones (Table 1,
range of different organic functional groups including alde- Scheme 1) were envisaged and successfully demonstrated util-
hydes and ketones [17,18], pyridinium ions to afford piperidines ising sodium dithionite under standard batch conditions by
Scheme 1: Sodium dithionite-mediated reductions under basic conditions.
Table 1: Reduction of aldehydes and ketones under batch and flow conditions.
entry
R1
R2
yielda,b
productivityc
flow productivity/
batch productivity
batcha
flowb
batch
flow
(g·L−1·h−1)
(g·L−1·h−1)
1
1
1
.1
.2
.3
H
H
H
92%
80%
89%
92%
81%
73%
0.96
4.27
4.24
4.32
4.4
5.0
4.6
0.84
0.93
0.87
1.4
H
83%
77%
4.55
5.2
1530

Beilstein J. Org. Chem. 2018, 14, 1529–1536.
Table 1: Reduction of aldehydes and ketones under batch and flow conditions. (continued)
1.5
1.6
1.7
H
H
H
98%
69%
85%
70%
80%
88%
1.02
0.72
0.88
4.30
2.01
7.05
4.2
2.8
8.0
1
.8
.9
H
H
92%
79%
80%
91%
0.96
0.83
3.75
1.45
3.9
1.7
1
1
.10
.11
H
H
65%
52%
71%
68%
0.68
0.55
3.54
2.97
5.2
5.4
1
1.12
1.13
1.14
Me
Me
Me
58%
60%
50%
11%d
4%d
0.61
0.63
0.52
0.72
0.23
1.2
0.4
0.1
<1%d
<0.08
1
.15
.16
cyclohexanone
84%
49%
50%d
<1%d
0.88
0.51
2.16
2.4
0.1
1
Et
<0.05
1.17
Me
73%
29%d
0.76
1.51
2.0
aBatch conditions: aldehyde or ketone (1 equiv), Na2S2O4 (4.5 equiv), NaHCO3/IPA (1:1), Δ, 12 h. bFlow conditions: for aldehydes, 0.165 M aldehyde
1 equiv) in IPA/H2O (1:1), 0.75 M Na2S2O4 (4.5 equiv) in NaHCO3/IPA (1:1), 0.250 mL·min−1 (64 min residence time for entries 1.1–1.5, 1.7, 1.8,
(
1
0
.10, and 1.11) or 0.1 mL·min−1 (160 min residence time for entries 1.6 and 1.9), 110 °C; for ketones, 0.165 M ketone (1 equiv) in IPA/H2O (1:1),
.75 M Na2S2O4 (4.5 equiv) in NaHCO3/IPA (1:1), 0.200 mL·min−1 (80 min residence time). cProductivity (space-time) = grams of product produced
per L reactor volume per hour. Batch reactions performed on 1 g scale, reactor volume = 81 mL, reaction time = 12 h, flow reactions performed on
.65 mmol scale (64 min for entries 1.1–1.5, 1.7, 1.8, 1.10, and 1.11) or 0.99 mmol scale (160 min for entries 1.6 and 1.9), reactor volume = 16 mL
2 mL chip + 14 mL coil), reaction time = 2.4 h (64 min residence) or 4.67 h (160 min residence) [24]. dConversions estimated from 1H NMR.
1
(
refluxing for 12 hours in the presence of sodium dithionite and HPLC pumps, a 2 mL mixing chip connected in series to a
sodium bicarbonate (1 M) in an isopropanol/water mixture. 14 mL HT PTFE coil mounted on a heating block and a back
Several primary and secondary alcohols (Table 1) were pre- pressure regulator fitted at the output flow stream. Reagents
pared in yields of 52–98% and 49–73%, respectively.
were introduced through two 10 mL injection loops (Figure 2).
Flow-based reduction of aldehydes and
ketones
Optimisation under flow conditions
The flow reduction was initially optimised for the conversion of
In developing a flow protocol a Uniqsis FlowSyn Stainless benzaldehyde to benzyl alcohol. The process required
Steel Flow reactor with a Multi X automated sampler was 4.5 equivalents of sodium dithionite and was optimised in terms
utilized. The reactor set-up (Figure 1) involved the use of two of flow rate and temperature using the reactor set-up shown in
1531

Beilstein J. Org. Chem. 2018, 14, 1529–1536.
dence time of 64 minutes in the mixing chip and heated coil
reactor (Table 2, Figure 3). In order to better compare the flow
and batch processes the reduction of benzaldehyde was repeated
in a seal-tube vessel superheated to 110 °C with a reaction time
of 64 min to match that of the optimised flow process. In this
instance a 43% conversion to benzyl alcohol was observed, in-
dicating that under flow conditions, the simple superheating of
the solvent was only partly responsible for the increased rate of
reaction and that in all likelihood the improved mixing of the
reagent streams also plays an important role.
Further reduction of the residence time while maintaining quan-
titative conversion was envisaged by increasing the reaction
temperature beyond 110 °C, however, the deposition of a black
residue within the reactor eventually led to blockages and
reactor fouling. The residue is possibly arising from the decom-
position of sodium thiosulfate which occurs when heating in
Figure 1: Uniqsis FlowSyn Stainless Steel Flow reactor.
Figure 2. Under optimised conditions 0.75 M sodium dithionite aqueous solutions above 90 °C. The precipitate was not ob-
in isopropanol/water/NaHCO3 [1 M] (1:1:2) was mixed with a served in the reactor at temperatures up to 110 °C and as such
0
.165 M stock solution of benzaldehyde in isopropanol/water/ the flow process was limited to a maximum operating tempera-
NaHCO3 [1 M] (1:1:2) in a 2 mL mixing chip at ambient tem- ture of 110 °C.
perature. Thereafter, the solution was superheated to 110 °C
while being passed through a 14 mL HT PTFE coil affording The optimised conditions were translated to the reduction of the
near quantitative conversion (92% isolated yield) with a resi- remaining aldehydes (Table 1). Under these conditions, alde-
Figure 2: Flow reactor configuration for the reduction of aldehydes and ketones.
Table 2: Optimization of the reduction of benzaldehyde under flow conditions.
entry
residence time (min)
flow rate (mL·min–1)
temp (°C)
conversion
yield
A
B
C
D
E
F
64
64
64
53
59
64
0.25
0.25
0.25
0.30
0.27
0.25
50
70
51%
70%
89%
90%
92%
99%
–
–
90
–
110
110
110
–
–
92%
1532

Beilstein J. Org. Chem. 2018, 14, 1529–1536.
Figure 3: NMR spectra showing the optimisation of the dithionite reduction for the reduction of benzaldehyde.
hyde reductions gave moderate to good yields (68–92%) which In all cases the yields were considered to be too low to be of use
were comparable to those observed in batch in a significantly even though in certain examples (Table 1, entries 1.15 and 1.17)
reduced reaction residence time (64 min vs 12 h). When the productivity was arguably moderately better (2.4 and
expressed in terms of reactor space-time productivity (grams of 2.0 times, respectively) than that of the analogous batch pro-
product produced per litre of reactor volume per hour) the flow cesses.
based processes displayed productivities which ranged from 3.9
Selective reduction of aldehydes in the pres-
to 8 times that observed under batch conditions for the runs
with a residence time of 64 min (Table 1, entries 1.1–1.5, 1.7, ence of ketones
.8, 1.11, and 1.12). It should be noted that the batch-based pro- The striking difference in relative reactivity between the ke-
1
cesses were not stringently optimised in terms of reaction resi- tones and the aldehydes on the flow system allowed us to
dence time with reactions being stopped and worked-up after demonstrate selective reductions of aldehydes in the presence of
1
2 h, however, at this point in time only the reduction of various ketones by simple selection of an appropriate flow rate.
benzaldehyde and 4-chlorobenzaldehyde (Table 1, entries 1.1 This selectivity was shown by the reduction of benzaldehyde in
and 1.5) showed complete consumption of starting material. In the presence of various ketones at equal concentrations
the case of a few selected examples (Table 1, entries 1.6 and (Table 3). In all cases benzaldehyde was efficiently reduced
1
.9) a longer residence time of 160 min was investigated, how- (71–91% conversion as determined by 1H NMR) and the ke-
ever, the decrease in reactor productivity on increasing the reac- tones remained largely unreduced with only acetophenone (9%)
tion residence time meant that the advantage afforded by the and 4-chloroacetophenone (8%) affording conversions above
flow approach decreased with productivity differences of 1% (Table 3, entries 3.1 and 3.2).
<
1.9 times relative to the corresponding batch processes.
The reaction was further tested by the reduction of 3-acetyl-
When the approach was adopted for the reduction of ketones, benzaldehyde which contained an aldehyde and ketone func-
surprisingly low conversions of <50% were observed when tionality on the same molecular scaffold. In this instance an
residence times ranged from 80 to 160 min, with only the aldehyde conversion of 85% was observed with <1% reduction
reduction of cyclohexanone affording a moderate yield of 50%. of the ketone (Table 3, entry 3.8, Figure 4).
1533

Beilstein J. Org. Chem. 2018, 14, 1529–1536.
Table 3: Selective reduction of benzaldehyde in the presence of various ketones.
entry
ketonea
aldehyde reduction
ketone reduction
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
acetophenone
4-chloroacetophenone
4-methylacetophenone
2-hydroxyacetophenone
4-aminoacetophenone
cyclohexanone
80%
91%
72%
72%
85%
71%
72%
85%
≤9%
8%
≤1%
≤1%
≤1%
≤1%
≤1%
≤1%
3-heptanone
3-acetylbenzaldehyde
aStock solution of 0.2 M concentration relative to both benzaldehyde and ketone in IPA/H2O/NaHCO3 (1:1:2) [1 M], 0.9 M sodium dithionite (4.5 equiv)
in IPA/H2O/NaHCO3 (1:1:2) [1 M], flow rate 0.250 mL·min−1 (64 min residence in mixing chip and coil reactor), 110 °C; b0.165 M of substrate in IPA/
H2O/NaHCO3 (1:1:2) [1 M], 0.75 M sodium dithionite (4.5 equiv) in IPA/H2O/NaHCO3 (1:1:2) [1 M], flow rate 0.250 mL·min−1, 110 °C.
Figure 4: Selective reduction of an aldehyde in the presence of a ketone.
Demonstration of a continuous process
in Figure 1 by removing the mixing chip which previously acted
Finally, we were interested in converting the approach to a con- as a trap for precipitates and diluting the sodium dithionite stock
tinuously run process, however, a concern at this stage was the solution from 0.75 M to 0.5 M (Figure 5). As a safety precau-
formation of the aforementioned precipitate and although this tion a stream of aqueous sodium hydroxide was connected to
was negligible at 110 °C, we felt it could potentially lead to the two reagent streams via selector valves which could be used
blockages and reactor fouling over extended reaction times. In to flush the reactor if blockages occurred. Finally the injection
consideration of this we modified the reactor set-up illustrated loops were exchanged for reagent reservoirs and the 14 mL
Figure 5: Flow reactor set-up for the continuous reduction of aldehydes.
1534

Beilstein J. Org. Chem. 2018, 14, 1529–1536.
2
.
.
Wiles, C.; Watts, P. Green Chem. 2014, 16, 55–62.
coil reactor was exchanged for a 29 mL coil reactor (Figure 5).
The process was tested with a 64 min residence time
doi:10.1039/C3GC41797B
3
Plutschack, M. B.; Pieber, B.; Gilmore, K.; Seeberger, P. H.
Chem. Rev. 2017, 117, 11796–11893.
doi:10.1021/acs.chemrev.7b00183
(
0.45 mL·min−1 flow rate) and allowed to run continuously for
5.3 hours during which time no significant precipitate forma-
tion was noted and no reactor flushing was required. A total of
.7 g of benzaldehyde was reduced affording 6.99 g of benzyl-
5
4. Kirschning, A. Beilstein J. Org. Chem. 2011, 7, 1046–1047.
8
doi:10.3762/bjoc.7.119
5
.
Ley, S. V.; Fitzpatrick, D. E.; Ingham, R. J.; Myers, R. M.
Angew. Chem., Int. Ed. 2015, 54, 3449–3464.
doi:10.1002/anie.201410744
alcohol (79% yield after chromatographic purification) equating
to a productivity of 4.36 g·L−1·h−1 which was comparable to
that of 4.27 g·L−1·h−1 observed for the reduction of benzalde-
hyde using the setup shown in Figure 2 (Table 1, entry 1.1).
6.
7.
8.
9.
Vaccaro, L.; Lanari, D.; Marrocchi, A.; Strappaveccia, G. Green Chem.
2014, 16, 3680–3704. doi:10.1039/C4GC00410H
Webb, D.; Jamison, T. F. Org. Lett. 2012, 14, 2465–2467.
doi:10.1021/ol300722e
Conclusion
Webb, D.; Jamison, T. F. Org. Lett. 2012, 14, 568–571.
doi:10.1021/ol2031872
In conclusion, we have developed a simple transition-metal-free
continuous flow method for the reduction of aldehydes in
aqueous media utilising sodium dithionite which does not
generate or use molecular hydrogen. The process affords
comparable yields to those obtained under batch conditions but
in reduced reaction residence time (64 min vs 12 h) and im-
proved (>3.9 times) space-time productivity. The process shows
high relative selectivity for the reduction of aldehydes over ke-
tones and through the appropriate selection of flow rate selec-
tive reductions of aldehydes in the presence of ketones can be
realised. Finally the process can be run continuously with
minimal loss in reactor productivity.
Newton, S.; Carter, C. F.; Pearson, C. M.; de C. Alves, L.; Lange, H.;
Thansandote, P.; Ley, S. V. Angew. Chem., Int. Ed. 2014, 53,
4915–4920. doi:10.1002/anie.201402056
10.Fukuyama, T.; Chiba, H.; Kuroda, H.; Takigawa, T.; Kayano, A.;
Tagami, K. Org. Process Res. Dev. 2016, 20, 503–509.
doi:10.1021/acs.oprd.5b00353
1
1
1
1
1
1
1.Fan, X.; Sans, V.; Yaseneva, P.; Plaza, D. D.; Williams, J.; Lapkin, A.
Org. Process Res. Dev. 2012, 16, 1039–1042. doi:10.1021/op200373m
2.Gilmore, K.; Vukelić, S.; McQuade, D. T.; Koksch, B.; Seeberger, P. H.
Org. Process Res. Dev. 2014, 18, 1771–1776. doi:10.1021/op500310s
3.Caldarelli, M.; Baxendale, I. R.; Ley, S. V. Green Chem. 2000, 2,
43–46. doi:10.1039/b000816h
4.Neyt, N. C.; Riley, D. L. React. Chem. Eng. 2018, 3, 17–24.
doi:10.1039/C7RE00146K
5.Makarov, S. V.; Silaghi-Dumitrescu, R. J. Sulfur Chem. 2013, 34,
Supporting Information
444–449. doi:10.1080/17415993.2012.749878
6.Makarov, S. V.; Harváth, A. K.; Silaghi-Dumitrescu, R.; Gao, Q. Sodium
Dithionite, Rongalite and Thiurea Oxides; World Scientific Publishing
Co, 2016. doi:10.1142/q0028
Supporting Information File 1
Experimental part.
[
https://www.beilstein-journals.org/bjoc/content/
1
1
1
2
2
7.Minato, H.; Fujie, S.; Okuma, K.; Kobayashi, M. Chem. Lett. 1977, 6,
supplementary/1860-5397-14-129-S1.pdf]
1091–1094. doi:10.1246/cl.1977.1091
8.De Vries, J. G.; Kellogg, R. M. J. Org. Chem. 1980, 45, 4126–4129.
doi:10.1021/jo01309a011
9.Heilmann, S. M.; Rasmussen, J. K.; Smith, H. K., II. J. Org. Chem.
Acknowledgements
1
983, 48, 987–992. doi:10.1021/jo00155a011
0.Park, K. K.; Oh, C. H.; Joung, W. K. Tetrahedron Lett. 1993, 34,
445–7446. doi:10.1016/S0040-4039(00)60148-X
This work was supported by the National Research Foundation
of South Africa (grant number 87893) and the University of
Pretoria (University, Science Faculty Research Councils and
Research and Development Program), and Pelchem Pty Ltd.
Opinions expressed in this publication and the conclusions
arrived at, are those of the authors, and are not necessarily attri-
buted to the NRF. The authors would like to gratefully
acknowledge Eric Palmer and Mamoalosi Selepe for NMR
spectroscopy services and Uniqsis Pty Ltd for flow equipment.
7
1.Park, K. K.; Oh, C. H.; Sim, W.-J. J. Org. Chem. 1995, 60, 6202–6204.
doi:10.1021/jo00124a043
22.Kaplánek, R.; Krchňák, V. Tetrahedron Lett. 2013, 54, 2600–2603.
doi:10.1016/j.tetlet.2013.03.010
2
3.Scheuerman, R. A.; Tumelty, D. Tetrahedron Lett. 2000, 41,
531–6535. doi:10.1016/S0040-4039(00)00959-X
6
2
4.Space-time calculations are based-upon the actual reaction volume
and not the entire volume of the flow reactor or batch reactor.
ORCID® iDs
Darren L. Riley - https://orcid.org/0000-0003-1328-640X
References
1.
Wiles, C.; Watts, P. Green Chem. 2012, 14, 38–54.
doi:10.1039/C1GC16022B
1535

Beilstein J. Org. Chem. 2018, 14, 1529–1536.
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which
permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
(https://www.beilstein-journals.org/bjoc)
The definitive version of this article is the electronic one
which can be found at:
doi:10.3762/bjoc.14.129
1536