Flow synthesis of organic azides and the multistep synthesis of imines and amines using a new monolithic triphenylphosphine reagent

Article abstract of DOI:10.1039/c0ob00813c

Here we describe general flow processes for the synthesis of alkyl and aryl azides, and the development of a new monolithic triphenylphosphine reagent, which provides a convenient format for the use of this versatile reagent in flow. The utility of these new tools was demonstrated by their application to a flow Staudinger aza-Wittig reaction sequence. Finally, a multistep aza-Wittig, reduction and purification flow process was designed, allowing access to amine products in an automated fashion.

Full text of DOI:10.1039/c0ob00813c

View Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 1927
www.rsc.org/obc
PAPER
Flow synthesis of organic azides and the multistep synthesis of imines and
amines using a new monolithic triphenylphosphine reagent†
Catherine J. Smith, Christopher D. Smith, Nikzad Nikbin, Steven V. Ley and Ian R. Baxendale
Received 1st October 2010, Accepted 22nd November 2010
DOI: 10.1039/c0ob00813c
Here we describe general flow processes for the synthesis of alkyl and aryl azides, and the development
of a new monolithic triphenylphosphine reagent, which provides a convenient format for the use of this
versatile reagent in flow. The utility of these new tools was demonstrated by their application to a flow
Staudinger aza-Wittig reaction sequence. Finally, a multistep aza-Wittig, reduction and purification
flow process was designed, allowing access to amine products in an automated fashion.
reactor only briefly reach steady state. These conditions make
Introduction
operations such as the introduction of a third reaction component
Recently, there has been considerable interest in the utilisation of
or later stage combination of two intermediates much more
flow based techniques for laboratory scale chemical synthesis.1
technically challenging.
The general aim has been the design of flow processes that
To date, the most common solution to this ‘third stream
permit efficient assembly of small molecules in useful quantities
problem’ in segmented flow processes has been to use the third
for biological evaluation, and that also allow facile scale-up of
stream component in excess, necessitating substantial in-line or
resulting hit compounds.2
off-line purification.6,7 This strategy may be acceptable in some
cases, but leaves significant room for improvement in terms of
both process safety and efficiency, particularly when the third
stream reagent is toxic or includes precious material. However,
until the technology to effect precise monitoring and control over
flow reactions on laboratory scale is widely available,8 this problem
is best solved using very specific chemical strategies. The challenges
of segmented flow processes render the use of supported reagents
and techniques such as catch-and-release and catch-react-and-
release vital components of the flow chemist’s toolbox.9
These flow procedures are defined by high standards of safety,
efficiency, reproducibility and product purity. Further benefits
can be realised when flow processing is conducted with the use
of solid supported reagents3 that aid work-up/purification and
make telescoped multistep sequences more readily achievable.4
With advances in commercially available benchtop flow chemistry
platforms and associated technologies for reaction monitoring,
flow processing has become an increasingly viable tool for all
synthesis scales; not only for industrial production processes but
also for small research laboratories. Chemists can now consider
designing highly automated multistep processes, including the use
of in-line purification strategies to deliver high quality products in
a safe and efficient manner.5
While the use of flow processing for laboratory scale chemical
synthesis offers some advantages, it also entails its own unique set
of challenges. For the generation of a collection of compounds in
flow, syntheses are often carried out in a series of short reaction
plugs. This processing method, whereby reaction plugs are passed
through the flow system in a constant stream of solvent, is
termed segmented flow. The continuous flow processing generally
employed for larger scale syntheses operates under steady state
conditions. Working in segmented flow, however, means that over
the course of each reaction, conditions at a given point in the
One such approach is the Staudinger aza-Wittig reaction, which
has been previously employed as a key bond forming reaction
in our flow synthesis of the natural product oxomaritidine.10
Using supported triphenylphosphine, the Staudinger aza-Wittig
reaction is a particularly powerful method for the formation of
imines10,11 within a multistep segmented flow process. Trapping
an in situ generated azide as the iminophosphorane on the solid
support allows facile purification of this intermediate, provides
an opportunity for solvent switching if desired, and enables
subsequent reaction with an in situ generated aldehyde with
controlled stoichiometry. This entire process can be conducted
without any manual handling or isolation of potentially hazardous
or sensitive reaction components.
Despite the strategic value of this transformation, it has not
found wider application in our other synthetic programmes
due to the nature of the polymer-supported reagents involved.
Commercially available sources of azide ion exchange resin and
polymer-supported triphenylphosphine are supplied as low cross-
linked microbead formats, which can be difficult to employ in a
flow environment owing to their solvent-swelling characteristics
Innovative Technology Centre, Department of Chemistry, University of
Cambridge, Lensfield Road, Cambridge, UK CB2 1EW. E-mail: theitc@
ch.cam.ac.uk
† Electronic supplementary information (ESI) available. CCDC reference
numbers 791578, 791581. For ESI and crystallographic data in CIF or
other electronic format see DOI: 10.1039/c0ob00813c
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1927–1937 | 1927
©

View Online
and the resulting high pressure. An additional restriction to their
wide-scale adoption is the prohibitive financial cost of these
commercial materials.
and rapid conversion to the desired azide products. Fully or
partially depleted monoliths could be regenerated, and were found
to function with the same performance for multiple cycles of
regeneration and use. Alkyl bromides (1 M in MeCN–THF 1 : 1)
were pumped at a flow rate of 0.5 mL min-1 through an azide ion-
exchange monolith heated to 80 ◦C, generating the corresponding
alkyl azides in high yields and purities following only evaporation
of solvent. These azides could either be isolated as final products
or carried on without isolation as intermediates in a multistep flow
sequence as described below.
Our group and others have been interested in expanding
the range of solid-supported reagents available, and have been
investigating the benefits of new polymeric support materials
for application in flow synthesis.12 Monolithic columns are an
interesting and effective format for supported reagents, as they
offer a more convenient physical form for handling and improved
flow characteristics, whilst retaining the benefits of traditional
beads.13 Of further importance is the ability to rapidly synthesise
bespoke monolithic reagents without any specialist equipment
or techniques such as those required for the manufacture of
traditional bead type reagents. This is an especially attractive
concept for the preparation of customised alternatives to expensive
and/or non-reusable supported reagents, and allows tailoring of
the monolith to suit a particular application. We have recently
reported the development of an azide ion-exchange monolith,
which was successfully used to generate acyl azides for use in
the Curtius rearrangement.14 Here we report the development
of a new monolithic triphenylphosphine reagent and improved
processes for the flow synthesis of alkyl and aryl azides. With these
superior reagents and processes in hand, we return to the study of
the Staudinger aza-Wittig reaction, and demonstrate its utility for
compound library synthesis and finally for its application in the
multistep automated synthesis of amines under flow conditions.
Flow synthesis of aryl azides
The utility of the Staudinger aza-Wittig reaction as a general tool
for application in multistep segmented flow processes would be
greatly enhanced by expanding the substrate scope of our flow
procedure to include aryl azides. Aryl azides are useful compounds
with a wide variety of applications in chemistry, biology and
materials science.16 Given the importance of these compounds
and the hazards associated with their generation and handling,
a general flow process for the synthesis of aryl azides would find
widespread application in a variety of fields. Until very recently,7
however, there was no protocol reported for the synthesis of aryl
azides in flow. Therefore, building on our previous experience10,17
and initial investigations in this area,18 we set out to develop a
flow process for the synthesis of aryl azides, and to demonstrate
its utility in multistep segmented flow sequences.
We chose to use the mild and efficient batch conditions
developed by Moses and coworkers19 as a convenient starting
point for the development of a flow process. Simply carrying out
these same reactions in flow would eliminate the need to cool (and
even allow safe heating of) these reactions, and would address
the issues of handling dangerous materials and gas evolution on
scale. However, in our view, the original batch conditions still
required further optimisation to provide an ideal flow process
for application to chemical library and multistep syntheses. Our
principle goals were to avoid using any excess of potentially
hazardous components, in this case, toxic trimethylsilyl azide, and
to develop one set of general conditions that could be applied
to a wide range of substrates. Furthermore, it was of critical
importance to include in-line purification strategies (two different
strategies were employed, as discussed in this and the following
publication20) to ensure that no hazardous starting materials or
by-products were carried through into later steps.
Results and discussion
All flow reactions and monolith preparations described in this
work were conducted using the commercially available Vapourtec
R2+/R4 flow system and automated sequences controlled by Flow
Commander software.15
Flow synthesis of alkyl azides
During our flow synthesis of oxomaritidine,10 we reported the
generation of alkyl azides from alkyl bromides using commercially
available azide ion exchange resin, a costly reagent with some
undesirable characteristics for use in flow. However, this process
can be greatly improved using the azide ion exchange monolith
recently developed by our group,14 as shown in Scheme 1.
Initial optimisation indicated that it was possible to carry
out these reactions with only one equivalent each of tert-butyl
nitrite and trimethylsilyl azide. Indeed, Moses and co-workers
subsequently reported these reactions with 1 : 1 : 1 stoichiometry,21
but noted that reaction times required under these conditions
varied from 2 min for 4-nitroaniline to 16 h for aniline. By carrying
out these reactions in flow, however, we were able to heat reactions
in order to effect good conversions for all substrates tested in a
standard reaction time of 50 min.
Using the general flow conditions shown in Scheme 2, we were
able to obtain high conversions (86–95%) to the desired azide
products from aniline starting materials with a wide range of
electronic and steric demands (Table 1). For each reaction the
azides were obtained in very good purities, with the only detectable
contaminants being traces of unreacted starting materials, which
Scheme 1 Flow synthesis of alkyl azides.
The monolithic reagent was used in excess (13 mmol supported
azide used for up to 10 mmol of reaction) to ensure complete
1928 | Org. Biomol. Chem., 2011, 9, 1927–1937
This journal is
The Royal Society of Chemistry 2011
©

View Online
Scheme 2 Flow synthesis of aryl azides.
Table 1 Aryl azides generated in flow
[including position of the labelled nitrogen] was confirmed by two-
dimensional NMR studies, as well as by comparison to previously
reported IR and NMR data for (1-15N)-azidobenzene.24 In a
second experiment, we performed the same reaction beginning
with a 1 : 1 mixture of d5-aniline and (15N)aniline. Analysis of the
reaction output identified d5-azidobenzene and (15N)azidobenzene
as the only products, indicating that no crossover between the
two labelled substrates had occurred. From these experiments, we
conclude that under our reaction conditions, aryl azide formation
proceeds without cleavage of the aniline C–N bond, in agreement
with the pentazene mechanism.
Entry
R1
R2
R3
R4
R5
Conversiona
1
2
3
4
5
6
7
8
H
H
H
H
H
I
iPr
COMe
Cl
CN
H
H
H
H
H
H
H
H
H
H
H
C
H
H
H
H
H
H
H
H
H
H
H
H
H
CF3
H
H
H
H
H
H
H
H
H
H
H
H
H
94%
94%
95%
93%
86%
86%
Me
OMe
NO2
CO2Me
H
H
H
Cl
Br
H
H
93% (72%)b
89% (75%)b
91%
9
10
11
12
13
87% (88%)b
92%
CH
H
H
OMe
CF3
91%
89%
Generation of monolithic triphenylphosphine reagent
H
Triphenylphosphine is a commonly used laboratory reagent with a
wide range of applications, but is notorious for causing purification
difficulties, usually requiring expensive and time-consuming col-
umn chromatography to remove triphenylphosphine and its oxide
by-product from the final product. Thus, it is highly desirable to
use a supported version of triphenylphosphine, which can be easily
removed at the end of the reaction. The current commercially
available supported triphenylphosphine species are non-optimal,
both in terms of handling in a flow system and their inherent
cost. Here we report the development of a new monolithic
triphenylphosphine reagent and its application to promote the
Staudinger reaction in flow.
a Reactions were carried out as shown in Scheme 2, on a 1 mmol scale.
Conversions are based on the ratio of azide product to aniline starting
material as determined by 1H NMR. b Isolated yields after column
chromatography. (Low yields for Entries 7 and 8 may be due to volatility
of these products.)
could be removed prior to further reaction either by a catch-react-
and-release strategy as described below, or by in-line scavenging
as discussed in the following publication.20 Furthermore, these
reactions could be safely scaled with no modification to the
procedure; for the work discussed here azides were generated on
up to 10 mmol scale.
For monolith polymerisation, the monomer, cross-linker, poro-
gen and initiator were homogenised and decanted into a glass
column, which was then sealed at both ends and heated using the
Vapourtec R4 multi-channel convection heater. Following poly-
merisation, the monolith could be connected to the flow system
for washing and subsequent reactions, simply by exchanging the
sealed ends of the column with standard tubing connectors.
For the work described here, monoliths were formed using
a polymerisation mixture containing (weight% of total): 31.7%
styrenediphenylphosphine (SDP) monomer, 24.9% divinylben-
zene cross-linker, 42.7% 1-docecanol porogen, and 0.7% 4,4¢-
azobis(4-cyanovaleric acid) (ACVA) initiator. In practice, SDP
Mechanistic investigation
A recent publication7 has suggested that, by analogy to the
iododeamination of anilines with tert-butyl nitrite,22 the gen-
eration of aryl azides under similar conditions also proceeds
via an aryl radical intermediate. However, the mechanism of
aryl azide formation from diazonium salts and azide ions has
been extensively studied, and it is generally agreed that this
reaction progresses via pentazene and pentazole intermediates
formed by attack of the azide species on the terminal nitrogen
of the diazonium.16,23 Under our reaction conditions, it is not
necessarily clear which of these mechanistic pathways would
be favoured. While these reactions do involve heating anilines
with tert-butyl nitrite, in the presence of trimethylsilyl azide we
expected that azide formation would proceed via the precedented
pentazene/pentazole mechanism, rather than the proposed radical
pathway. Therefore we carried out two reactions involving labelled
substrates in order to determine whether the aniline C–N bond is
cleaved under our reaction conditions.
◦
was dissolved in 1-docecanol with gentle heating (>50 C), then
divinylbenzene was added, followed by ACVA, and the resulting
mixture was loaded into a glass column. The polymerisation
process was carried out by heating the sealed column at 90 ◦C in the
R4 convective heating unit for 20 h. Following the polymerisation
procedure, the monolith was washed with dry THF at 60 ◦C to
elute the porogen and any residual non-polymeric material; this
yielded a rigid white monolith that completely filled the glass
column as shown in Fig. 1.
In our first experiment, we subjected (15N)aniline to the general
conditions shown in Scheme 2. This resulted in the formation of
(15N)azidobenzene as a single product. The identity of this material
The monoliths used for this work were formed in 7 cm
length ¥ 10 mm i.d. columns, resulting in monoliths of 3 g
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1927–1937 | 1929
©

View Online
monolith at a flow rate of 0.15–0.2 mL min-1. As shown in
Scheme 4, azide starting materials could be isolated following
batch or flow synthesis and introduced via the sample loop of the
R2+ unit (Method A), however, a far superior method in terms of
both safety and efficiency was to utilise azides generated in flow
without isolation.
Indeed, both alkyl (Scheme 4, Method B) and aryl (Scheme 4,
Method C) azides could be generated and carried on directly into
the Staudinger reaction within the flow system simply by incorpo-
rating a triphenylphosphine monolith prior to the exit of the flow
system. This required no modification to the procedure for the
synthesis of aryl azides described above, although alkyl azides were
generated at a reduced flow rate of 0.2 mL min-1 to ensure adequate
reaction time for the subsequent Staudinger reaction. Following
the Staudinger reaction, the desired iminophosphorane intermedi-
ates were immobilised on solid support, allowing any excess azide
or other impurities to be washed to waste (Scheme 3, Step 2). This
simple purification step yielded the desired iminophosphoranes
as pure and stable products on the monolithic support, which
could be used immediately for further reactions, or stored in sealed
columns on the benchtop for several weeks before use.
With a clean iminophosphorane monolith in hand, the aza-
Wittig reaction was carried out by heating the monolith to 120 ◦C
and passing through a solution of aldehyde or ketone (0.1–1.0
M in THF or MeCN–THF 1 : 1) at a flow rate of 0.1 mL min-1
(Scheme 3, Step 3). It was found that each iminophosphorane
monolith could be used for multiple reactions with different
substrates without cross-contamination, simply by washing the
monolith with the reaction solvent between runs. Furthermore,
we found that repeated heating and cooling cycles did not affect
the stability of the monoliths or purity of the resulting products.
In order to achieve high conversions of the desired imine
products, the iminophosphorane monoliths needed to be used in
an excess stoichiometry. Elemental analysis indicated an average
loading of 5.1 mmol phosphorous per monolith. However, the
Fig. 1 Monolithic triphenylphosphine reagent, with and without glass
column housing.
average dry weight. Elemental analysis indicated an average
loading of 1.7 mmol phosphorous per gram of monolith. Loading
of phosphorous was found to be consistent from monolith to
monolith, and analysis of samples taken from various points
along the length and radius of a single monolith revealed even
distribution of phosphorous within the monolith.
An important practical consideration for the application of
a fixed bed reactor in flow is the pressure drop created by
the static bed. In testing, our triphenylphosphine monoliths
demonstrated a near-linear correlation of flow rate and pressure,
giving consistently low backpressures of only 2–4 bar (THF at
a flow rate of 1.0 mL min-1) when connected in-line. This low
pressure drop made them ideal for use as in-line reactor cartridges,
either on their own or in combination with additional cartridges.
Staudinger aza-Wittig Reaction in Flow
With these improved reagents and processes in hand, we turned
our attention to their application in the Staudinger aza-Wittig
reaction. The overall reaction sequence is shown in Scheme 3.
First, the Staudinger reaction was carried out by heating the
triphenylphosphine monolith to 60 ◦C using the R4 convection
heater and passing a solution of organic azide through the
Scheme 3 Overall reaction sequence for Staudinger aza-Wittig reactions in flow. Step 1: Azides were loaded onto the triphenylphosphine monolith via
the Staudinger reaction (see Scheme 4 for detail). Step 2: Washing with the appropriate solvent allowed removal of any impurities to provide a clean
iminophosphorane intermediate on support. Step 3: Aza-Wittig reactions were carried out by passing a solution of aldehyde or ketone through the
iminophosphorane monolith. Multiple sequential aza-Wittig reactions could be carried out using the same iminophosphorane monolith by washing with
the appropriate solvent after each reaction.
1930 | Org. Biomol. Chem., 2011, 9, 1927–1937
This journal is
The Royal Society of Chemistry 2011
©

View Online
Scheme 4 Staudinger reaction in flow. Method A: Isolated azides could be introduced via a sample loop and pumped through a triphenylphosphine
monolith to effect the Staudinger reaction. Method B: Alkyl azides were generated from the corresponding alkyl bromides and carried on to Staudinger
reaction without isolation. Method C: Aryl azides were generated in situ from the corresponding anilines and carried on directly to Staudinger reaction.
elemental loading of phosphorous may not be identical to the
functional loading due to the presence of triphenylphosphine
oxide or triphenylphosphine not physically accessible for reaction.
Consequently, the substrates shown in Scheme 5 were used to ex-
plore the functional loading of these monoliths and the practicali-
ties of achieving high conversion to the desired imines using these
stoichiometric supported reagents. First, the Staudinger reaction
was carried out (Scheme 4, Method C) using an excess (10 mmol)
of 1-azido-4-methoxybenzene to ensure maximal loading of
the desired iminophosphorane 3. Next, the iminophosphorane
monolith was reacted with 10 mmol of aldehyde 4, recycling
the reaction stream through the monolith for 20 h as shown in
Scheme 5. This yielded 4.14 mmol of the imine product 5 (as
a mixture with the unreacted aldehyde starting material, but in
otherwise excellent purity). Elemental analysis of the depleted
monolith indicated no residual nitrogen content, signifying that
all of the iminophosphorane intermediate had been consumed
in the aza-Wittig reaction. For this combination of aza-Wittig
partners, the active loading of the iminophosphorane monolith 3
was therefore 4.14 mmol.
In order to carry out multiple sequential aza-Wittig reactions,
it is more practical to conduct each transformation in a single
pass through the reactor. The aza-Wittig reaction shown in
Scheme 5 was repeated in a single pass fashion at a flow rate
of 0.1 mL min-1. As expected, conversion steadily decreased as
material was processed,25 but reaction with the first 2 mmol of
substrate could be carried out with good (≥80%) conversion. Based
on these findings, the aza-Wittig reactions shown in Table 2 were
carried out using each iminophosphorane monolith to react with
a maximum of 2 mmol of carbonyl partners (each reaction was
run on 0.5–1 mmol scale).
The products were either directly isolated as imines following
only evaporation of solvent, or reduced in batch by treatment with
sodium borohydride, followed by purification of the desired amine
via a catch-and-release protocol using polymer-supported sulfonic
acid (QP-SA). In each case, the imines or resulting amines were
obtained in high purities and generally high yields as shown in
Table 2.
However, the Staudinger aza-Wittig reactions involving 1-
azido-4-nitrobenzene (Table 2, Entries 24–26) proceeded with
little or no conversion to the desired imines. Examination of
the output stream from the Staudinger reaction step indi-
cated successful reaction of 1-azido-4-nitrobenzene, suggesting
that these poor conversions were due to the low reactivity of
the resulting iminophosphorane intermediate in the aza-Wittig
step.
The aza-Wittig reactions shown in Entries 17 and 20, on the
other hand, proceeded with complete conversion. The resulting
imines were isolated directly from the flow aza-Wittig process
in excellent purity following only evaporation of solvent. These
products were found to be free from cross-contamination and
Scheme 5 Aza-Wittig reaction with recycling.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1927–1937 | 1931
©

View Online
Table 2 Staudinger aza-Wittig products
Entry
1
Azide
Aza-Wittig partner
Isolated product
Yielda
93%
2
3
95%
72%
4
81%
5
6
81%
73%
7
78%
8
98%
80%
80%
9
10
11
12
91%
Quant.
1932 | Org. Biomol. Chem., 2011, 9, 1927–1937
This journal is
The Royal Society of Chemistry 2011
©

View Online
Table 2 (Contd.)
Entry
13
Azide
Aza-Wittig partner
Isolated product
Yielda
88%
14
15
97%
95%
16
62%
17
18
19
Quant.
94%
95% conversion
3 mmol scaleb
20
21
Quant.
65%
22
Quant.
8 : 2 mixture E : Z
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1927–1937 | 1933
©

View Online
Table 2 (Contd.)
Entry
23
Azide
Aza-Wittig partner
Isolated product
SM recovered
Yielda
No reaction
24
SM recovered
SM recovered
No reactionc
25
26
No reactionc
51% conversionc
a Aza-Wittig reactions were carried out on 0.5–1.0 mmol scale based on the starting aldehyde or ketone. Products shown as imines were isolated directly
from the flow aza-Wittig reaction following only evaporation of solvent. For products shown as amines, the yield reported is the overall yield for the two
step process of imine synthesis in flow followed by off-line reduction and purification by catch-and-release in batch. b 3 mmol of aldehyde starting material
1
was recycled through the monolith for 4.5 h, resulting in 95% conversion to the desired imine as determined by H NMR. c Examination of the waste
stream from the Staudinger reaction step indicated successful loading of the desired iminophosphorane, however this intermediate proved unreactive
with a number of aza-Wittig partners
≥95% pure by 1H NMR, and confirmed by elemental analysis to
be pure and free from phosphorous.
iminophosphorane monolith to carry out the aza-Wittig reaction.
Upon exiting the iminophosphorane monolith, the flow stream
was joined by a second stream of TFE at 0.1 mL min-1.
The resulting solution was pumped at a combined flow rate of
0.2 mL min-1 through the borohydride monolith, which was heated
at 70 ◦C in the R4 convective heater. The presence of the more
acidic TFE solvent was found to be very beneficial to the rate of the
reduction step and the overall purity of the final products. Upon
exiting the borohydride monolith, the solution passed through a
column of QP-SA, trapping the desired amine product onto the
solid phase and allowing any alcohol or other impurities to be
washed to waste (Scheme 6, Step 1).
The clean amines were then released (Scheme 6, Step 2) using
a solution of ammonia in methanol (2.0 M) at a flow rate of
0.5 mL min-1, facilitating isolation of the desired products in good
yields and purities following only evaporation of solvent as shown
in Table 3.
An additional benefit of this in-line purification process is the
option to make further use of the iminophosphorane monolith.
By tolerating lower yields based on the aldehyde partner, this
multistep sequence makes it possible to reclaim more of the
iminophosphorane intermediate as clean final product. Despite
reduced conversion to the imine intermediate as the iminophos-
phorane partner is consumed, the in-line reduction and purifica-
tion protocol still gives access to amine products in consistently
high purity. This is demonstrated by entries 4a–c in Table 3, which
were carried out sequentially using the same iminophosphorane
monolith: the yields decrease as the iminophosphorane monolith is
depleted, but amine products are nevertheless obtained in excellent
final purity.
It was also found that larger scale aza-Wittig reactions could
successfully be carried out with good conversion by recycling the
reaction mixture to provide longer reaction time. The aza-Wittig
reaction shown in Scheme 5 was repeated on a 3 mmol scale,
recycling the reaction solution through the monolith for 4.5 h. As
shown in Table 2, Entry 19, this resulted in 95% conversion to the
desired imine as determined by 1H NMR.
Automated multistep flow process for the generation of secondary
amines
The multistep flow process described above gives access to imine
products in high yields and purities, but imines are sensitive
compounds, which are often prepared as synthetic intermediates
for further reactions. It was therefore desirable to incorporate
additional steps into the flow sequence, in order to use the imine
intermediates in subsequent reactions without isolation. We chose
to reduce the imines to the corresponding amines using a mono-
lithic borohydride reagent (prepared by the same ion exchange
procedure as for the azide monolith14). This reduction step and a
subsequent in-line purification via catch-and-release were incor-
porated into the flow process following the aza-Wittig reaction.
Beginning with a pre-loaded iminophosphorane monolith (pre-
pared by the flow process shown in Scheme 4), the full aza-
Wittig, reduction and purification sequence shown in Scheme 6
was automated using Flow Commander software. In Step 1,
aldehyde solutions (0.5 mmol, 0.5 M in THF) were injected into
the flow system and pumped at 0.1 mL min-1 through a pre-loaded
1934 | Org. Biomol. Chem., 2011, 9, 1927–1937
This journal is
The Royal Society of Chemistry 2011
©

View Online
Table 3 Amines synthesised using the automated aza-Wittig, reduction and purification sequence shown in Scheme 6
Entry
1
Azide
Aza-Wittig partner
Isolated product
Yielda
79%
2
3
77%
74%
4ab
4bb
4cb
88%
78%
68%
a Reactions were carried out on 0.5 mmol scale based on the starting aldehyde. b These reactions were carried out sequentially using the same
iminophosphorane monolith.
Scheme 6 Automated flow process for aza-Wittig reaction, imine reduction and purification. Step 1: Aldehyde partners were introduced via a sample
loop, pumped through the iminophosphorane monolith (pre-loaded in a separate flow step as shown in Scheme 4) to effect aza-Wittig reaction, and the
resulting imine intermediates reduced by passage through the borohydride monolith. Resulting amine products were trapped onto QP-SA, allowing any
impurities to be washed to waste. Step 2: Amine products were released by washing with ammonia in methanol.
The monolithic triphenylphosphine reagent described herein
provides a new, improved and cost-effective format for this
Conclusions
The generation and handling of organic azides normally involves
a number of safety concerns, particularly on scale, which can
limit the use of these important compounds. The flow procedures
described here provide flexible working methods for the safe
generation of a range of aryl and alkyl azides in high conversions.
These compounds can either be isolated as final products or carried
on directly to further transformations without isolation within a
contained flow system.
versatile reagent. Monolithic triphenylphosphine allows the facile
use of triphenylphosphine in flow, to provide high purity products
free from contamination with phosphine by-products without
requiring any manual purification steps.
Finally, the application of these new reagents and techniques
to the Staudinger aza-Wittig reaction enabled the synthesis of
imine or amine products in good yields and purities from either
alkyl bromide or aniline starting materials. This multistep flow
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1927–1937 | 1935
©

View Online
process, involving up to four reactions and two purification steps,
requires no manual handling of the azide intermediates and no
additional purification of the desired products. Using Vapourtec
Flow Commander software, this sequence was successfully run in
a highly automated fashion, and has the potential for application
to fully automated library synthesis of imines and amines in flow.
and R. Wittenberg, Angew. Chem., Int. Ed., 2001, 40, 650–679; S. V.
Ley, I. R. Baxendale, R. N. Bream, P. S. Jackson, A. G. Leach, D. A.
Longbottom, M. Nesi, J. S. Scott, R. I. Storer and S. J. Taylor, J. Chem.
Soc., Perkin Trans. 1, 2000, 3815–4195; D. C. Sherrington, Chem.
Commun., 1998, 2275–2286; S. J. Shuttleworth, P. M. Allin and K.
Sharma, Synthesis, 1997, 1217–1239.
4 M. Brasholz, B. A. Johnson, J. M. Macdonald, A. Polyzos, J.
Tsanaktsidis, S. Saubern, A. B. Holmes and J. H. Ryan, Tetrahedron,
2010, 66, 6445–6449; M. Grafton, A. C. Mansfield and M. J. Fray,
Tetrahedron Lett., 2010, 51, 1026–1029; L. Tamborini, P. Conti, A.
Pinto and C. D. Micheli, Tetrahedron: Asymmetry, 2010, 21, 222–225;
A. R. Bogdan, S. L. Poe, D. C. Kubis, S. J. Broadwater and D. T.
McQuade, Angew. Chem., Int. Ed., 2009, 48, 8547–8550; B. Ahmed-
Omer, J. C. Brandt and T. Wirth, Org. Biomol. Chem., 2007, 5, 733–
740; H. Sahoo, J. Kralj and K. Jensen, Angew. Chem., Int. Ed., 2007,
46, 5704–5708; M. Baumann, I. R. Baxendale, S. V. Ley, C. D. Smith
and G. K. Tranmer, Org. Lett., 2006, 8, 5231–5234.
5 J. P. McMullen and K. F. Jensen, Annu. Rev. Anal. Chem., 2010, 3, 19–
42; J. P. McMullen, M. T. Stone, S. L. Buchwald and K. F. Jensen,
Angew. Chem., Int. Ed., 2010, 49, 7076–7080; S. V. Ley and I. R.
Baxendale, Chimia, 2008, 62, 162–168.
6 For illustrative examples see: I. R. Baxendale, S. Schou, J. Sedelmeier
and S. V. Ley, Chem.–Eur. J., 2010, 16, 89–94; T. P. Petersen, A. Ritze´n
and T. Ulven, Org. Lett., 2009, 11, 5134–5137; For a more detailed
discussion see: M. D. Hopkin, Ph.D. Thesis, University of Cambridge,
2010.
Acknowledgements
We gratefully acknowledge Hokko Chemical Industry Co., Ltd.
for their kind donation of styryl diphenylphosphine, Vapourtec for
assistance in setting up the reactors, the Paul Mellon Fellowship
(C. J. S.), Syngenta (C. D. S.), the EPSRC (N. N.), the Royal Society
(I. R. B.) and the BP 1702 Fellowship (S. V. L.) for funding. We
also wish to thank J. E. Davies for crystal structure determination
and the EPSRC for a financial contribution toward the purchase of
the X-ray diffractometer. We are grateful to P. C. Skelton, P. Grice
and the Department of Chemistry NMR service for assistance
with the mechanistic studies, and to the Department of Chemistry
Photography Services for help with graphics.
7 F. Stazi, D. Cancogni, L. Turco, P. Westerduin and S. Bacchi,
Tetrahedron Lett., 2010, 51, 5385–5387.
8 Development of technology for the precise control of multistep
segmented flow processes is currently underway in our laboratory, and
will be reported in due course: H. Lange, C. F. Carter, M. D. Hopkin,
A. Burke, J. G. Goode, I. R. Baxendale and S. V. Ley, Chem. Sci., 2011,
DOI: 10.1039/C0SC00603C.
9 M. D. Hopkin, I. R. Baxendale and S. V. Ley, Chem. Commun., 2010, 46,
2450–2452; Z. Qian, I. R. Baxendale and S. V. Ley, Synlett, 2010, 505–
508; A. Palmieri, S. V. Ley, A. Polyzos, M. Ladlow and I. R. Baxendale,
Beilstein J. Org. Chem., 2009, 5, No. 23; C. M. Griffiths-Jones, M. D.
Hopkin, D. Jo¨nsson, S. V. Ley, D. J. Tapolczay, E. Vickerstaffe and M.
Ladlow, J. Comb. Chem., 2007, 9, 422–430; I. R. Baxendale, S. V. Ley,
C. D. Smith and G. K. Tranmer, Chem. Commun., 2006, 4835–4837;
I. R. Baxendale and S. V. Ley, Curr. Org. Chem., 2005, 9, 1521–1534;
Y. Suzuki and M. Kodomari, Chem. Lett., 1998, 11, 1091–1092; B. J.
Cohen, M. A. Kraus and A. Patchornik, J. Am. Chem. Soc., 1977, 99,
4165–4167.
Notes and references
1 L. Malet-Sanz, J. Madrzak, S. V. Ley and I. R. Baxendale, Org. Biomol.
Chem., 2010, 8, 5324–5332; M. O’Brien, I. R. Baxendale and S. V.
Ley, Org. Lett., 2010, 12, 1596–1598; X. Y. Mak, P. Laurino and P. H.
Seeberger, Beilstein J. Org. Chem., 2009, 5, No. 19; T. Fukuyama, M. T.
Rahman, M. Sato and I. Ryu, Synlett, 2008, 151–163; F. Benito-Lo´pez,
R. J. M. Egberink, D. N. Reinhoudt and W. Verboom, Tetrahedron,
2008, 64, 10023–10040; C. Wiles and P. Watts, Eur. J. Org. Chem.,
2008, 1655–1671; J. Yoshida, A. Nagaki and T. Yamada, Chem.–Eur. J.,
2008, 14, 7450–7459; New Avenues to Efficient Chemical Synthesis:
Emerging Technologies, ed. P. H. Seeberger, and T. Blume, Springer,
Berlin, 2007; I. R. Baxendale, J. J. Hayward and S. V. Ley, Comb.
Chem. High Throughput Screening, 2007, 10, 802–836; T. N. Glasnov
and C. O. Kappe, Macromol. Rapid Commun., 2007, 28, 395–410; I. R.
Baxendale and M. R. Pitts, Chim. Oggi, 2006, 24, 41–45; A. Kirschning,
W. Solodenko and K. Mennecke, Chem.–Eur. J., 2006, 12, 5972–5990;
K. Ja¨hnisch, V. Hessel, H. Lo¨we and M. Baerns, Angew. Chem., Int.
Ed., 2004, 43, 406–446; P. Hodge, Curr. Opin. Chem. Biol., 2003, 7,
362–373.
2 C. H. Hornung, B. Hallmark, M. Baumann, I. R. Baxendale, S. V.
Ley, P. Hester, P. Clayton and M. R. Mackley, Ind. Eng. Chem. Res.,
2010, 49, 4576–4582; S. Hu¨bner, U. Bentrup, U. Budde, K. Lovis, T.
Dietrich, A. Freitag, L. Ku¨pper and K. Ja¨hnisch, Org. Process Res.
Dev., 2009, 13, 952–960; A. A. Kulkarni, V. S. Kalyani, R. A. Joshi and
R. R. Joshi, Org. Process Res. Dev., 2009, 13, 999–1002; A. Bogdan and
D. T. McQuade, Beilstein J. Org. Chem., 2009, 5, No. 17; P. Styring and
A. I. R. Parracho, Beilstein J. Org. Chem., 2009, 5, No. 29; U. Kunz
and T. Turek, Beilstein J. Org. Chem., 2009, 5, No. 70; T. Gustafsson,
F. Ponten and P. H. Seeberger, Chem. Commun., 2008, 1100–1102;
C. H. Hornung, M. R. Mackley, I. R. Baxendale and S. V. Ley, Org.
Process Res. Dev., 2007, 11, 399–405; N. Steinfeldt, R. Abdallah, U.
Dingerdissen and K. Ja¨hnisch, Org. Process Res. Dev., 2007, 11, 1025–
1031; I. R. Baxendale, C. M. Griffiths-Jones, S. V. Ley and G. K.
Tranmer, Chem.–Eur. J., 2006, 12, 4407–4416; K. Ja¨hnisch, V. Hessel,
H. Lo¨we and M. Baerns, Angew. Chem., Int. Ed., 2004, 43, 406–446;
A. J. deMello and R. Wootton, Lab Chip, 2002, 2, 7N–13N; S. H.
DeWitt, Curr. Opin. Chem. Biol., 1999, 3, 350–356.
10 I. R. Baxendale, J. Deeley, C. M. Griffiths-Jones, S. V. Ley, S. Saaby
and G. K. Tranmer, Chem. Commun., 2006, 2566–2568.
11 H. Mahdavi and J. Amani, Tetrahedron Lett., 2009, 50, 5923–5926;
C. Gil and S. Bra¨se, Chem.–Eur. J., 2005, 11, 2680–2688; A. Kamal,
K. L. Reddy, V. Devaiah and N. Shankaraiah, Synlett, 2004, 2533–
2536; B. J. Neubert and B. B. Snider, Org. Lett., 2003, 5, 765–768; A.
Grieder and A. W. Thomas, Synthesis, 2003, 1707–1711; S. Barthe´le´my,
S. Schneider and W. Bannwarth, Tetrahedron Lett., 2002, 43, 807–810;
A. B. Charette, A. A. Boezio and M. K. Janes, Org. Lett., 2000, 2,
3777–3779; K. Hemming, M. J. Bevan, C. Loukou, S. D. Patel and D.
Renaudeau, Synlett, 2000, 1565–1568.
12 C. H. Hornung, B. Hallmark, M. R. Mackley, I. R. Baxendale and
S. V. Ley, Adv. Synth. Catal., 2010, 352, 1736–1745; N. T. S. Phan,
J. Khan and P. Styring, Tetrahedron, 2005, 61, 12065–12073; J. Siu,
I. R. Baxendale, R. A. Lewthwaite and S. V. Ley, Org. Biomol. Chem.,
2005, 3, 3140–3160; M. I. Burguete, E. Garc´ıa-Verdugo, M. J. Vicent,
S. V. Luis, H. Pennemann, N. Graf, von Keyserling and J. Martens,
Org. Lett., 2002, 4, 3947–3950; A. M. Hafez, A. E. Taggi, T. Dudding
and T. Lectka, J. Am. Chem. Soc., 2001, 123, 10853–10859; S. Laue,
L. Greiner, J. Wo¨ltinger and A. Liese, Adv. Synth. Catal., 2001, 343,
711–720.
13 N. Nikbin, M. Ladlow and S. V. Ley, Org. Process Res. Dev., 2007, 11,
458–462; F. Svec and C. G. Huber, Anal. Chem., 2006, 78, 2100–2107;
B. Altava, M. I. Burguete, E. Garcia-Verdugo, S. V. Luis and M. J.
Vicent, Green Chem., 2006, 8, 717–726; U. Kunz, H. Scho¨nfeld, A.
Kirschning and W. Solodenko, J. Chromatogr., A, 2003, 1006, 241–249;
W. Solodenko, U. Kunz, G. Jas and A. Kirschning, Bioorg. Med. Chem.
Lett., 2002, 12, 1833–1835; A. Kirschning, C. Altwicker, G. Dra¨ger, J.
Harders, N. Hoffmann, U. Hoffmann, H. Scho¨nfeld, W. Solodenko and
U. Kunz, Angew. Chem., Int. Ed., 2001, 40, 3995–3998; J. A. Tripp, J. A.
Stein, F. Svec and J. M. J. Fre´chet, Org. Lett., 2000, 2, 195–198; E. C.
3 J. Rolland, X. C. Cambeiro, C. Rodr´ıguez-Escrich and M. A. Perica`s,
Beilstein J. Org. Chem., 2009, 5, No. 56; P. H. Seeberger, Nat. Chem.,
2009, 1, 258–260; A. Chighine, G. Sechi and M. Bradley, Drug Discovery
Today, 2007, 12, 459–464; A. Michrowska, K. Mennecke, U. Kunz, A.
Kirschning and K. Grela, J. Am. Chem. Soc., 2006, 128, 13261–13267;
A. Kirschning, and G. Jas, in Immobilized Catalysts, ed. A. Kirschning,
Springer, Berlin, 2004, vol. 242, pp. 209–239; S. Bhattacharyya, Curr.
Opin. Drug Discovery Dev., 2004, 7, 752–764; G. Jas and A. Kirschning,
Chem.–Eur. J., 2003, 9, 5708–5723; S. V. Ley and I. R. Baxendale, Nat.
Rev. Drug Discovery, 2002, 1, 573–586; A. Kirschning, H. Monenschein
1936 | Org. Biomol. Chem., 2011, 9, 1927–1937
This journal is
The Royal Society of Chemistry 2011
©

View Online
Peters, F. Svec and J. M. J. Fre´chet, Adv. Mater., 1999, 11, 1169–1181;
C. Viklund, F. Svec, J. M. J. Fre´chet and K. Irgum, Chem. Mater., 1996,
8, 744–750; F. Svec and J. M. J. Fre´chet, Chem. Mater., 1995, 7, 707–
715; Q. C. Wang, F. Svec and J. M. J. Fre´chet, Anal. Chem., 1995, 67,
670–674; F. Svec and J. M. J. Fre´chet, Anal. Chem., 1992, 64, 820–822.
14 M. Baumann, I. R. Baxendale, S. V. Ley, N. Nikbin and C. D. Smith,
Org. Biomol. Chem., 2008, 6, 1587–1593.
18 C. J. Smith, I. R. Baxendale, and S. V. Ley, ‘Forbidden’ Reactions in
Flow: Azide Chemistry (PC-096), presented at the Eleventh Interna-
tional Kyoto Conference on New Aspects of Organic Chemistry, Kyoto,
November, 2009.
19 K. Barral, A. D. Moorhouse and J. E. Moses, Org. Lett., 2007, 9,
1809–1811.
20 C. J. Smith, N. Nikbin, S. V. Ley, H. Lange and I. R. Baxendale, Org.
Biomol. Chem., 2011, DOI: 10.1039/c0ob00813c.
21 A. D. Moorhouse and J. E. Moses, Synlett, 2008, 2089–
2092.
15 Vapourtec R Series systems and Flow CommanderTM software are
available from VapourtecLtd, Park Farm Business Centre, Fornham
St Genevieve, Bury St Edmunds, Suffolk, IP28 6TS, UK. Website,:
http://www.vapourtec.co.uk.
22 L. Malet-Sanz, J. Madrzak, R. S. Holvey and T. Underwood, Tetrahe-
dron Lett., 2009, 50, 7263–7267 and references therein.
23 M. E. C. Biffin, J. Miller, and D. B. Paul, in The Chemistry of the
Azido Group, ed. S. Patai, John Wiley & Sons, Ltd., London, 1971, pp.
156–165.
16 S. Bra¨se, C. Gil, K. Knepper and V. Zimmermann, Angew. Chem., Int.
Ed., 2005, 44, 5188–5240.
17 I. R. Baxendale, S. V. Ley, A. C. Mansfield and C. D. Smith, Angew.
Chem., Int. Ed., 2009, 48, 4017–4021; C. D. Smith, I. R. Baxendale,
G. K. Tranmer, M. Baumann, S. C. Smith, R. A. Lewthwaite and
S. V. Ley, Org. Biomol. Chem., 2007, 5, 1562–1568; C. D. Smith, I. R.
Baxendale, S. Lanners, J. J. Hayward, S. C. Smith and S. V. Ley, Org.
Biomol. Chem., 2007, 5, 1559–1561.
ˇ
24 L. Lesˇeticky´, R. Barth, I. Neˆmec, M. St´ıcha and I. Tisˇlerova´,
Czech. J. Phys., 2003, 53, A777–A782.
25 A graph of total cumulative conversion versus total material processed
is included in the Supporting Information†.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1927–1937 | 1937
©