Rapid and mild synthesis of Au-NHC complexes in a simple two-phase flow reactor

Article abstract of DOI:10.1039/d1dt01357b

We describe a simple two-phase flow reactor which allows for the rapid synthesis of several Au(i)-NHC complexes in high yields (>88%), under mild conditions, and with minimal workup. Translation of the standard weak base method to a two-phase flow reaction prevents the common problem of decomposition to Au(0). The reaction can be scaled up more than ten-fold without loss in conversion efficiency. An optional second stage allows for direct synthesis of Au(iii)-NHC complexes, without isolation of the Au(i)-NHC intermediate, with a two-step isolated yield of 82%.

Full text of DOI:10.1039/d1dt01357b

Volume 50
Number 23
21 June 2021
Pages 7857-8260
Dalton
Transactions
An international journal of inorganic chemistry
rsc.li/dalton
ISSN 1477-9226
PAPER
Helgi Freyr Jónsson, Andrew J. Harvie et al.
Rapid and mild synthesis of Au–NHC complexes in a simple
two-phase flow reactor
Anniversary
Volume

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Rapid and mild synthesis of Au–NHC complexes in
a simple two-phase flow reactor†
Cite this: Dalton Trans., 2021, 50,
7969
Helgi Freyr Jónsson, * Anne Fiksdahl and Andrew J. Harvie
*
We describe a simple two-phase flow reactor which allows for the rapid synthesis of several Au(I)–NHC
complexes in high yields (>88%), under mild conditions, and with minimal workup. Translation of the stan-
dard weak base method to a two-phase flow reaction prevents the common problem of decomposition
to Au(0). The reaction can be scaled up more than ten-fold without loss in conversion efficiency.
An optional second stage allows for direct synthesis of Au(III)–NHC complexes, without isolation of the
Au(I)–NHC intermediate, with a two-step isolated yield of 82%.
Received 23rd April 2021,
Accepted 28th May 2021
DOI: 10.1039/d1dt01357b
rsc.li/dalton
route, as well as the light sensitivity of Ag(I)–NHC complexes
are significant downsides.
Introduction
N-Heterocyclic carbenes (NHCs) are one of the most widely
The weak base route has in recent years shown to be an
used classes of ligands for organometallic complexes. The easy and efficient alternative to synthesise Au(I)–NHC com-
strength of the metal–carbene bond as well as the ease of plexes. The use of weak, inexpensive bases such as K2CO3,
modification of both the backbone as well as the N-wingtip NEt3 and NaOAc, alongside DCM or acetone as the most
substituents make this ligand class especially attractive.1,2 common (single-phase) solvents, all under ambient atmos-
NHC complexes of gold have been shown to catalyse a wide pheric conditions makes this route attractive.8,9
variety of reactions, and Au(I)–NHC complexes especially are
Mechanistic studies indicate that the reaction proceeds via
important in gold catalysis.3,4 Au(III) has found considerably an aurate intermediate 2, which is formed rapidly by displace-
less use than Au(I)–NHC complexes, despite being readily avail- ment of Me2S by the chloride anion upon mixing of the imid-
able from the Au(I) precursors by a simple oxidation. This is azolium chloride 1 and the Au(I) precursor. These aurate salts
most likely due to Au(III)’s propensity towards reduction to are generally stable, and selected examples have been isolated
Au(I) or Au(0).5 Both Au(I)– and Au(III)–NHC complexes have and fully characterised.10,11 The Au(I)–NHC complex 3 for-
also shown promise as anticancer agents and their potential mation is then postulated to go through a concerted metalla-
application in medicinal chemistry has been widely studied.6
tion-deprotonation process, at least when NEt3 is used as the
There are three main methods to synthesise Au(I)–NHC base. Spectroscopic evidence indicates that the free carbene is
complexes; the free carbene route (Scheme 1a), the transmetal- not generated during the reaction.12 As these reactions are
lation route (Scheme 1b), and the weak base route typically performed in a single solvent phase over an insoluble
(Scheme 1c).7 All the routes start with the imidazolium chlor- base, workup involves at least one filtration step. Additionally,
ide salt 1 and a gold precursor with a labile ligand, such as the long reaction times which are often required result in the
Me2SAuCl. Both the free carbene route and the transmetalla- reduction of some of the Au(I) precursor to Au(0), potentially
tion route suffer from several disadvantages. In the free harming reaction yields and complicating workup.
carbene route, generation of the carbene requires a strong
base, such as KHMDS, n-BuLi, or NaH. Additionally, the
instability of the free carbene makes rigorously inert con-
ditions a necessity, usually requiring the use of a glovebox.
The transmetallation route, on the other hand, requires the
synthesis of either the Ag(I)– or Cu(I)–NHC complex ahead of
time. The higher cost and lower atom economy of this reaction
Department of Chemistry, NTNU, NO-7491 Trondheim, Norway.
E-mail: andrew.j.harvie@ntnu.no, helgi.f.jonsson@ntnu.no
†Electronic supplementary information (ESI) available: NMR spectra, experi-
mental data and schematics of flow components. See DOI: 10.1039/d1dt01357b
Scheme 1 Synthetic routes towards Au(I)–NHC complexes.
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 7969–7975 | 7969

View Article Online
Paper
Dalton Transactions
We have recently reported a study on alcohol functionalised pumps, and a reactor can be immediately reused after flushing
Au(I/III)–NHC complexes, which were synthesised via a modi- with a suitable solvent.
fied, two-phase version of the weak base route.13 Compared to
With these advantages in mind, we explored the synthesis
a typical reaction performed in a single DCM phase, we found of several Au(I)––NHC complexes in a simple two-phase flow
that performing the reaction in a two-phase DCM : H2O system reactor, with the aim of improving the total reaction time and
sped up the reaction significantly; full conversion was usually simplifying workup. This is, to the best of our knowledge, the
observed in under 2 hours, compared to only 35% after first example of a segmented flow reactor being used for the
12 hours in the single-phase system. However, in that batch synthesis of metal-NHC complexes.
process, we still noted a significant presence of Au(0) contami-
nation, requiring filtration of the product through a silica plug
to acquire the pure Au(I)–NHC complexes.
Results and discussion
Droplet-based microfluidic flow reactors offer several advan-
A schematic of our liquid–liquid flow reactor is shown in
Fig. 1. Matched syringes containing an aqueous base solution
(S1) and a DCM solution of the relevant NHC-aurate 2a–g (S2)
are connected to the inputs of a y-shaped mixer (Y1, see
Fig. S1 in ESI†), and mounted within a dual-channel syringe
pump. The co-injection of immiscible liquids at a stable flow
rate into Y1 generates a stable, uniform segmented flow, which
is directed into a 5 m length of coiled tubing (Zone A) where
the aurate intermediate in the DCM reacts with the aqueous
base, forming the desired Au(I)–NHC complex 3a–g. The
tubing in Zone A is held at 35 °C by submersion in a tempera-
ture-controlled water bath. For all experiments listed here, the
flow rate of each solution was held at 200 µl min−1, yielding a
total flow rate in Zone A of 400 µl min−1. The internal diameter
of the tubing in Zone A was 1 mm, which corresponds to
a total internal volume of 3.925 ml. At a total flow rate of
400 µl min−1, the total transit time within Zone A is 9.8 min.
The segmented flow exits Zone A into a previously-described
liquid–liquid phase separator PS (see Fig. S2 in ESI†), which
physically separates the two phases and stops the reaction.16
The DCM phase, which contains the newly formed Au(I)–NHC
complex, is collected for analysis by 1H NMR, whereas the
tages over batch reactors for performing multiphase reactions.
In two-phase chemistry, reaction rates are often limited by the
interfacial area (per unit volume) between the two phases. By
splitting the phases into microlitre-scale droplets in intimate
contact, this interfacial area-to-volume ratio is typically much
greater than in an equivalent bulk reaction, greatly increasing
the amount of phase mixing. Additionally, as each droplet is
generated identically (at least for stable flows), mixing and
heating are highly reproducible, avoiding much of the batch-to-
batch variation typical of batch processes.14 This property of flow
reactors is also beneficial during scale-up; in batch processes,
scale-up requires use of larger reaction volumes, reducing the
contacting surface area between phases, as well as larger reac-
tion vessels which exhibit different heat transfer characteristics.
In a flow reactor, however, the reaction can simply be run for
longer. For larger scale-ups, the reaction stream can simply be
split into multiple parallel streams, each with the same flow con-
ditions as the reaction optimised at the smaller scale. For two-
phase batch reactions, phases are generally manually separated
using a separation funnel or the like, whereas flow reactors can
make use of simple in-line separators that perform this step
without requiring human input.15,16 This reduces the amount of
workup that must be performed manually, but is also particu-
larly useful for multi-step reactions where new reagents need to
be added to the reaction volume downstream. Two-phase flow
reactions can also be performed in very simple flexible tubes.
This is not only cost effective, but is convenient for temperature
control, as the reaction temperature can be controlled by sub-
merging the tube in a standard water bath.17
Simoens et al. have recently reported a methodology using
a continuous flow solid-phase reactor for the synthesis of
Au(I)–, Cu(I)– and Pd(II)–NHC complexes using the common
ligand IPr.18 They achieved impressive yields of up to 93% for
the Au(I)–IPr complex, using short (2 min) residence times. Fig. 1 Schematic of flow reactor used for synthesis of Au(I)–NHC com-
plexes. An aqueous base solution and a solution of an imidazolium
aurate salt 2 in DCM are injected at equal flow rates into a y-shaped
mixer (Y1) using a matched pair of syringes (S1 and S2). The mixer gener-
ates a 1 : 1 segmented flow of the aqueous and DCM solutions, which
However, solid phase flow reactors suffer from several disad-
vantages when compared to liquid–liquid reactors. Solid-phase
reactors generally require the solid phase to be replenished
between each experimental run, which can be laborious.
enters a 5 m length of coiled tubing (Zone A) where the aurate inter-
Additionally, solid-phase reactors are susceptible to poisoning mediate 2a–g in the DCM reacts with the aqueous base, forming the
corresponding Au(I)–NHC complex 3a–g. The segmented flow then
enters a liquid–liquid phase separator (PS). The aqueous phase contain-
ing partially depleted base exits the separator at the through-port T,
then passes through a micro-metering needle valve NV, before being
of the packed solid phase, which can greatly reduce the reactor
performance. This is a particular issue for gold chemistry, due
to gold’s propensity for forming nanoparticles.18 Liquid–liquid
reactors suffer from no such issues: preparation of the reactor
discarded as waste. The DCM phase containing the Au(I)–NHC complex
generally consists of simply loading reactant solutions into is collected at the side port S.
7970 | Dalton Trans., 2021, 50, 7969–7975
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
Table 1 Synthesis of Au(I)–NHC complexes 3a–g as shown in
Scheme 3a
Entry
Preligand
Conversion to Au(I)–NHC 3b
Isolated yield
1
2
3
4
5
6
1a
1b
1c
1d
1e
1f
100%
100%
100%
100%
100%
98%
90%
88%
88%
94%
—
Scheme 2 Imidazolium-based salts 1a–g used in this study.
c
—
7
8
9
10
11
1g
1g
1g
1a
1a
38%
—
76%d
93%d,e
88%g
100%h
—
21%f
—
aqueous phase containing the partially depleted base is dis-
carded as waste. Prior to 1H NMR analysis, the solvent was
simply removed under reduced pressure, and the crude
product redissolved in deuterated chloroform.
95%
a 6 ml of 0.012 M DCM solutions of aurates 2a–g were reacted with
6 ml of 0.25 M aqueous solution of K2CO3 at 35 °C in the flow reactor
described in Fig. 1. b Conversion determined by 1H NMR spectroscopy.
c Fast decomposition to elemental gold. d CHCl3 was used as the
solvent and the reaction temperature was increased to 53 °C. e The flow
rate was halved to 100 μl min−1 per syringe leading to a residence time
of 20 minutes. f 93% purity according to 1H NMR. g 0.25 M NaOAc solu-
tion was used as base. h The reaction was performed at 0.548 g scale.
A wide range of imidazolium-based salts 1a–g (Scheme 2)
were chosen for this study. These include the two most
common NHC ligands IPr and IMes (1a,b) as well as their satu-
rated counterparts SIPr and SIMes (1c,d). To demonstrate our
method on functionalised ligands we also included the pre-
viously reported alcohol functionalised imidazolinium salt 1e.
Au(I)–NHC complexes with less hindered ligands are challen-
ging to synthesise with the weak base method,19 so to test the
scope of this method we chose N,N-dialkylimidazolium chlor-
ide 1f. Additionally, to test if the efficacy of our method
extends to a benzimidazolium motif, we also selected N,N-di-
alkylbenzimidazolium chloride 1g, which has previously only
been coordinated to gold via the free carbene route.20
The azolium salts 1a–g were mixed with equimolar
amounts of Me2SAuCl in DCM and sonicated to form 0.012 M
solutions of the corresponding aurate salts 2a–g. The aurate
solutions were then reacted under the segmented flow con-
ditions described above with a 0.25 M aqueous K2CO3 solution
acting as base (Scheme 3). The results from the synthesis of
the Au(I) complexes 3a–g are shown in Table 1.
For imidazoli(ni)um salts 1a–e, with bulky aryl sidegroups,
our conditions were extremely effective, giving full conversion
to gold complexes 3a–e (Table 1, entries 1–5) and excellent
yields ranging from 88% for 1c and 1d, to 98% for 1a (com-
pared with 93% reported for the same complex as synthesised
with the solid-phase flow reactor of Simoens et al.). It is worth
mentioning that the only workup required was the evaporation
of the organic phase under reduced pressure. This method is
therefore well suited for the coordination of both imidazolium
and imidazolinium chlorides, the latter of which have been
reported to require longer reaction times.10
showing decomposition to elemental gold. An exception is a
method reported by Johnson et al. which uses NBu4(acac) as the
base.19 To test our method on this type of substrate, we
attempted the coordination of N,N-dialkylimidazolium chloride
1f. Unfortunately, this led to almost instantaneous formation of
elemental gold (see Fig. S3†) and no complex was obtained
(Table 1, entry 6). This method is therefore clearly not suited for
less sterically hindered imidazolium salts, similar to most ver-
sions of the weak base route. Finally, N,N-dialkylbenzimidazo-
lium chloride 1g was found to give only 38% conversion to
Au(I)–NHC complex 3g (Table 1, entry 7). We attempted to
improve the conversion by changing the solvent to CHCl3,
which allowed for an increase in temperature to 53 °C (higher
temperatures caused bubbles to form inside the tube). This led
to an improvement in conversion to 76% (Table 1, entry 8),
using the same residence time of 10 minutes. A further increase
in conversion to 93% was observed by halving the flow rate and
thus doubling the residence time to 20 minutes. However, at
increased temperatures we observed significant decomposition
of the gold complex which led to a greatly reduced yield of only
21% of the 93% pure Au(I)–NHC complex 3g (Table 1, entry 9).
NaOAc has also been reported to be an alternative to K2CO3
for this reaction.21 We therefore attempted to synthesise
complex 3a by using a 0.25 M aqueous solution of NaOAc
instead (Table 1, entry 10). This led to a reduced 88% conver-
sion to Au(I)–NHC complex 3a, compared to 100% for K2CO3.
As mentioned above, one advantage of performing reac-
tions in flow over conventional batch reactions is that scale up
can be achieved simply by running the reactor for longer
periods. To test this for our flow reactor we repeated the syn-
thesis of Au(I)–NHC complex 3a at 12.5× scale (see
Experimental), using identical reaction parameters. As
expected, analysis of the collected solution showed 100% con-
version to 3a, and 0.521 g of pure Au(I)–NHC complex 3a was
isolated in 95% yield (Table 1, entry 11).
Au(I)–NHC complexes with sterically unencumbered ligands
are challenging to synthesise via the weak base route, usually
Scheme 3 Synthesis of Au(I)–NHC complexes 3a–g in the segmented
flow reactor described in Fig. 1.
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 7969–7975 | 7971

View Article Online
Paper
Dalton Transactions
Scheme 4 Batch synthesis of Au(I)–NHC complex 3a in (a) single-
phase (b) and (c) two-phase solvent systems.
Fig. 2 Schematic of second stage of the flow reactor used for synthesis
of the Au(III)–NHC complex 4a. The first stage of the reactor consists of
the single-step reactor used for synthesising the Au(I)–NHC complexes
3a–g (see Fig. 1 and S4†). The DCM stream emitted from S, containing
the Au(I)–NHC complex 3a is directed into one input of a second
y-shaped mixer Y2. A DCM solution of the oxidant PhICl2 is injected into
the other input port from a third syringe (S3), forming a continuous
flow which enters a second 5 m length of tubing (Zone B). The oxidised
Au(III)–NHC complex 4a is collected at the output of Zone B.
To compare our flow reaction with the conventional batch
reaction we conducted the reaction of imidazolium chloride 1a
with Me2SAuCl in both the commonly used single-phase
system, as well as our two-phase reaction system (Scheme 4).
Performing the reaction in just DCM, with 20 eq. of K2CO3
as base, led to extremely poor conversion. Only 5% conversion
(Scheme 4a) was observed after 10 min at 35 °C (cf. 100% in
the two-phase flow reaction). Meanwhile, the two-phase batch
reaction also gave full conversion to Au(I)–NHC complex 3a
under otherwise identical reaction conditions (Scheme 4b) but
resulted in a reduced isolated yield of 87% (cf. 98% for the
two-phase flow reaction). The two-phase batch reaction at
room temperature is also effective, giving 98% conversion
(Scheme 4c). The reason for this dramatic increase in reaction
rate with the two-phase system is likely due to more efficient
mixing between the aurate salt 2a and the base in the two-
phase system compared to the heterogeneous reaction.
Comparing the two-phase flow and batch reactions, the batch
reaction resulted in greater decomposition of the Au(I)–NHC
complex; within minutes the aqueous phase had turned a dark
grey and under evaporation of the separated organic phase the
resulting solid was grey, requiring a silica plug filtration to
acquire the pure gold complex. In the flow reaction, however,
both the phases remained clear and colourless throughout.
Evaporation of the organic phase resulted immediately in a
white solid product which was shown to be pure Au(I)–NHC
mixer Y2 to the separator’s side channel S. A third syringe (S3)
containing a DCM solution of the oxidant PhICl2 was con-
nected to the second input port of Y2. The output port of Y2
was connected to a second 5 m length of tubing (Zone B),
which was kept at 35 °C by submersion in the same water bath
as the tubing in Zone A. The flow rate from S3 was set at 100 µl
min−1, generating a continuous flow in Zone B with a flow rate
of 300 µl min−1, corresponding to a residence time of
13.1 min. Here, the Au(I)–NHC complex 3a emitted from the
separator at S is oxidised by PhICl2 to form the Au(III)–NHC
complex 4a, which is collected after passing through Zone
B. The reaction is shown in Scheme 5.
Performing the reaction with 1.2 equiv. PhICl2 (Scheme 5a)
led to poor conversion (9%) to the Au(III) complex 4a, as deter-
mined by 1H NMR. However, increasing the amount of oxidant
to 5 equiv. (Scheme 5b) gave a 100% conversion and gave
Au(III)–NHC complex 4a in 82% isolated yield for the two-step
reaction. This two-stage reactor therefore greatly simplifies the
synthesis of 4a from the starting imidazolium salt 1a – once
the reactor is started, the Au(III)–NHC complex 4a can be col-
lected after around 25 minutes, without isolation or purifi-
cation of any intermediates. The only workup required is
washing of the acquired solid product 4a with pentane and fil-
tration through a short silica plug.
1
complex 3a, according to H NMR spectroscopy. Whilst we do
not fully understand the mechanism behind this improve-
ment, as the decomposition of Au(I) to Au(0) in such syntheses
is poorly understood, we suggest that the more uniform and
efficient phase mixing offered by the microfluidic flow reactor
is likely to be a factor. The advantage that this brings in the
reduction of necessary workup, as well as the inherent scalabil-
ity of the flow reaction, is likely to be useful in environments
where Au(I)–NHC complexes are synthesised regularly.
Au(III)–NHC complexes are readily available from Au(I)–NHC
complexes via oxidation by PhICl2.22 We envisioned an easy
route towards Au(III)–NHC complexes by adding a second reac-
tion step to our flow apparatus, oxidising the formed Au(I)–
NHC complex without isolation. A schematic of the second
reaction stage is shown in Fig. 2, and a schematic of the full
two-stage reactor is shown in Fig. S4.† The single-step reactor
used for synthesis of the Au(I)–NHC complexes 3a–g was
Scheme 5 Synthesis of Au(III)–NHC complex 4a directly from imidazo-
extended by connecting an input port of a second y-shaped lium salt 1a in the two-stage flow reactor described in Fig. 2.
7972 | Dalton Trans., 2021, 50, 7969–7975
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
General procedure for the flow synthesis of Au(I)–NHC
complexes
Conclusions
In summary, we have presented a simple and efficient route to
Au(I)–NHC complexes from imidazolium salts by use of a two-
phase version of the weak base route under segmented flow
conditions. The flow reactor consists of a small number of low-
cost parts, including a syringe pump, mixer, separator block,
and a few low-cost fittings, making it simple to construct
without significant outlay. The method allows the synthesis of
the most common Au(I)–NHC complexes under mild con-
ditions and short reaction times. The important benefits of
the flow conditions are reduced decomposition to elemental
gold, simple scalability, and minimal workup, with only evap-
oration needed to acquire the 100% pure Au(I)–NHC com-
plexes. This method is excellent for both imidazolium and
imidazolinium chlorides with bulky groups, but is less
effective for N,N-dialkylbenzimidazolium chlorides and
ineffective for less hindered N,N-dialkylimidazolium chlorides.
We have also shown that performing the coordination in a
two-phase H2O : DCM system greatly increases the reaction rate
for the bulk reaction (performed outside of segmented flow
conditions).
Prior to each experimental run, the reactor was flushed with
clean distilled water and DCM to remove any traces of products
from previous runs. The water bath used to control the temp-
eratures of tubing zones A and B was set to 35 °C, and the
temperature was allowed to stabilise. 0.012 M solutions of
aurate salts 2a–g were formed by dissolving 0.0706 mmol of
the NHC-preligands 1a–g and Me2SAuCl in 6 ml of DCM and
sonicating for 5 minutes. The aurate solution, as well as 6 ml
of 0.25 M aqueous K2CO3 solution were loaded into syringes
S1 and S2 respectively, and the pumps were started. Collection
of reaction products was started after one full volume of stable
segmented flow had passed through the reactor, and contin-
ued until the delivery syringes were nearly empty. The syringes
were then filled with blank solvent, and the pumps restarted
until the dead volume of the reactor had been fully flushed
into the collection flasks. Evaporation of the collected organic
phase under reduced pressure afforded the pure Au(I)–NHC
complexes 3a–e as white solids.
Au(I)–NHC complex 3a
The flow reaction was expanded to include a second, oxi-
dation step. This allowed for the synthesis of a Au(III)–NHC
complex directly from the parent imidazolium salt, without
isolation or purification of the intermediate Au(I) complex.
This proves to be a rapid and user-friendly approach towards
Au(III)–NHC complexes, making research into their properties
more accessible.
We believe this method provides simple access to Au–NHC
complexes of importance to both catalysis and medicinal
chemistry, and could potentially be applied to the synthesis of
NHC complexes of other metals.
According to the general procedure for the flow synthesis of
Au(I)–NHC complexes, NHC-preligand 1a (30.0 mg,
0.0706 mmol) and Me2SAuCl (20.8 mg, 0.0706 mmol) gave Au
(I)–NHC complex 3a as a white solid in 98% yield (43.1 mg,
0.0692 mmol). 1H NMR (400 MHz, CDCl3) d(ppm) 7.50 (t,
J = 7.8 Hz, 2H), 7.29 (d, J = 7.8 Hz, 4H), 7.17 (s, 2H), 2.56 (sept,
J = 7.0 Hz, 4H), 1.34 (d, J = 7.0 Hz, 12H), 1.22 (d, J = 7.0 Hz,
12H). 13C NMR (150 MHz, CDCl3) d(ppm) 175.5 (Au–Ccarbene),
145.7, 134.1, 130.9, 124.4, 123.2, 28.9, 24.6, 24.2. NMR data
correspond to previously reported data.23
Au(I)–NHC complex 3b
According to the general procedure for the flow synthesis of Au
(I)–NHC complexes, NHC-preligand 1b (24.1 mg, 0.0706 mmol)
and Me2SAuCl (20.8 mg, 0.0706 mmol) gave Au(I)–NHC
Experimental
General
complex 3b as
a white solid in 88% yield (33.5 mg,
Commercial grade reagents were used without any additional
purification. 1H NMR spectra were recorded using a Bruker
Avance DPX 400 MHz spectrometer in CDCl3 as the solvent.
Chemical shifts are reported in ppm (δ) downfield from tetra-
methylsilane (TMS) as an internal standard. 1 mm inner dia-
meter, 2 mm outer diameter PTFE tubing (228–4118,VWR) was
used for all external tubing. Porous PTFE (Aeos ePTFE) with an
inner diameter of 1.8 mm, an outer diameter of 2.5 mm and
0.0621 mmol). 1H NMR (400 MHz, CDCl3) δ(ppm) 7.09 (s, 2H),
6.99 (s, 4H), 2.34 (s, 6H), 2.10 (s, 12H). 13C NMR (150 MHz,
CDCl3) δ(ppm) 173.4 (Au–Ccarbene), 139.9, 134.8, 134.75, 129.6,
122.3, 21.3, 17.9. NMR data correspond to previously reported
data.23
Au(I)–NHC complex 3c
an internodal distance of 15–25 μm was used inside the According to the general procedure for the flow synthesis of
liquid–liquid phase separator. All connections were made with Au(I)–NHC complexes, NHC-preligand 1c (30.2 mg, 0.0706 mmol)
standard nut-and-ferrule type connectors. The back-pressure at and Me2SAuCl (20.8 mg, 0.0706 mmol) gave Au(I)–NHC
the separatoŕs through-outlet T was controlled using a PEEK complex 3c as
a white solid in 90% yield (39.3 mg,
1
micro-metering needle valve (Rheodyne XP-230). Diagrams 0.0635 mmol). H NMR (600 MHz, CDCl3) δ(ppm) 7.41 (t, J =
detailing designs of and connections to the phase separator 7.8 Hz, 2H), 7.23 (d, J = 7.8 Hz, 4H), 4.04 (s, 4H), 3.05 (sept, J =
and the y-mixers are shown in Fig. S1 and S2 (see ESI†). All 6.9 Hz, 4H), 1.41 (d, J = 6.8 Hz, 12H), 1.33 (d, J = 6.9 Hz, 12H).
liquids were pumped from 10 ml syringes (Hamilton 13C NMR (150 MHz, CDCl3) δ(ppm) 196.2 (Au–Ccarbene), 146.7,
GASTIGHT) mounted within dual-channel syringe pumps 134.2, 130.2, 124.8, 53.6, 29.1, 25.3, 24.3. NMR data correspond
(Harvard PHD2000).
to previously reported data.23
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 7969–7975 | 7973

View Article Online
Paper
Dalton Transactions
Au(I)–NHC complex 3d
stirred for 5 minutes at 35 °C to form a 0.012 M solution of the
NHC-aurate 2a. K2CO3 (195 mg, 0.141 mmol) was then added.
The resulting suspension stirred for 10 minutes at 35 °C after
which it was filtered through Celite. The solvent was evapor-
ated under reduced pressure and the conversion to Au(I)–NHC
complex 3a was determined by 1H NMR spectroscopy.
According to the general procedure for the flow synthesis of Au(I)–
NHC complexes, NHC-preligand 1d (24.2 mg, 0.0706 mmol)
and Me2SAuCl (20.8 mg, 0.0706 mmol) gave Au(I)–NHC complex
3d as a white solid in 88% yield (33.5 mg, 0.0621 mmol).
1H NMR (400 MHz, CDCl3) δ(ppm) 6.94 (s, 4H), 3.98 (s, 4H),
2.31 (s, 12H), 2.29 (s, 6H). 13C NMR (100 MHz, CDCl3) δ(ppm)
195.0 (Au–Ccarbene), 139.0, 135.6, 134.7, 129.9, 50.8, 21.2, 18.1.
NMR data correspond to previously reported data.23
Two-phase batch synthesis of Au(I)–NHC complex 3a
NHC-preligand (1a, 30.0 mg, 0.0706 mmol) and Me2SAuCl
(20.8 mg, 0.0706 mmol) were dissolved in 6 ml DCM and stirred
for 5 minutes at 35 °C to form a 0.012 M solution of the NHC-
aurate 2a. 6 ml of 0.25 M aqueous K2CO3 solution was added,
and the mixture was stirred vigorously for 10 minutes. After
completion of the reaction, 10 ml of DCM and 10 ml of brine
were added, and the phases separated. The organic phase was
dried over anhydrous Na2SO4 and evaporated under reduced
pressure. The resulting mixture was filtered through a short
silica plug with EtOAc as the eluent to filter off the formed Au
(0). Evaporation yielded the pure Au(I)–NHC complex 3a as a
white solid in 87% yield (37.9 mg, 0.0614 mmol).
Au(I)–NHC complex 3e
According to the general procedure for the flow synthesis of
Au(I)–NHC complexes, NHC-preligand 1e (24.4 mg, 0.0706 mmol)
and Me2SAuCl (20.8 mg, 0.0706 mmol) gave Au(I)–NHC
complex 3e as
a white solid in 94% yield (35.9 mg,
0.0664 mmol). 1H NMR (400 MHz, CDCl3) δ(ppm) 7.36–7.43
(m, 5H), 6.91 (s, 2H), 5.93 (dd, J = 8.6, 5.4 Hz, 1H), 4.23–4.33
(m, 2H), 3.88–3.95 (m, 1H), 3.77–3.84 (m, 1H), 3.65–3.72 (m,
1H), 3.48–3.55 (m, 1H), 2.30 (dd, J = 7.2, 5.6 Hz, 1H, OH), 2.29
(s, 3H), 2.23 (s, 3H), 2.20 (s, 3H). 13C NMR (150 MHz, CDCl3)
δ(ppm) 194.0 (Au–Ccarbene), 138.9, 136.1, 135.6, 135.4, 135.1,
129.8, 129.7, 129.2, 128.8, 127.5, 64.1, 61.2, 50.6, 44.9, 21.2,
18.3, 18.2. NMR data correspond to previously reported data.13
Two-step flow synthesis of Au(III)–NHC complex 4a
The single-stage flow reactor was extended via addition of the
hardware for the second oxidation step, see Fig. 2 and Fig. S4.†
NHC-preligand 1a (60.0 mg, 0.141 mmol) and Me2SAuCl
(41.6 mg, 0.141 mmol) were dissolved in 12 ml of DCM and
sonicated for 5 minutes to form a 0.012 M solution of aurate
2a. The aurate solution, as well as 12 ml of 0.25 M aqueous
K2CO3 solution were loaded into their respective syringes and
the flow reactor was operated as described above. PhICl2
(194.0 mg, 0.706 mmol) was dissolved in 6 ml of DCM to form
a 0.12 M solution and loaded into syringe S3. The pumps for
syringes S1 and S2 were started first, and the pump for syringe
S3 was started after around 10 minutes ( just before the Au(I)–
NHC complex 3a had reached Y2). The organic phase contain-
ing the newly formed Au(III)–NHC complex 4a was collected at
the reactor outlet, evaporated under reduced pressure and
redissolved in a minimal amount of DCM. Pentane was added
to induce precipitation and the resulting solid was washed
thoroughly with pentane. The washed solid was then filtered
through a short silica plug with EtOAc as the eluent to filter off
remaining PhICl2. Evaporation under reduced pressure gave
Au(III)–NHC complex 4a as a yellow solid in 82% yield
(80.1 mg, 0.116 mmol). 1H NMR (400 MHz, CDCl3) δ(ppm)
7.56 (t, J = 8.0 Hz, 2H), 7.36 (d, J = 8.0 Hz, 4H), 7.34 (s, 2H),
2.85 (sept, J = 6.7 Hz, 4H), 1.40 (d, J = 6.8 Hz, 12H), 1.13 (d, J =
6.8 Hz, 12H). 13C NMR (100 MHz, CDCl3) δ(ppm) 146.2 (Au–
Au(I)–NHC complex 3g
NHC-preligand 1g (18.8 mg, 0.0706 mmol) and Me2SAuCl
(20.8 mg, 0.0706 mmol) were dissolved in 6 ml of CHCl3 and
sonicated for 5 minutes to form a 0.012 M solution of aurate
intermediate 2g. The aurate solution and 6 ml of a 0.25 M
solution of aqueous K2CO3 were then loaded into their respect-
ive syringes. The flow reactor was run at 53 °C with a flow rate
of 100 μl min−1 (residence time approx. 20 min) and the
organic phase was collected. Evaporation under reduced
pressure gave Au(I)–NHC complex 3g in 93% purity as a yellow
1
solid in 21% yield (7.3 mg, 0.0148 mmol). H NMR (400 MHz,
CDCl3) δ(ppm) 7.89–7.92 (m, 2H), 7.30–7.32 (m, 2H), 2.18 (s,
18H). 13C NMR (150 MHz, CDCl3) 177.0 (Au–Ccarbene), 134.0,
123.0, 116.2, 61.8, 32.9. NMR data correspond to previously
reported data.20
Scaled up flow synthesis of Au(I)–NHC complex 3a
NHC-preligand 1a (375 mg, 0.882 mmol) and Me2SAuCl
(260 mg, 0.882 mmol) were dissolved in 75 ml of DCM and
sonicated for 5 minutes to form a 0.012 M solution of aurate
2a. The aurate solution and 75 ml of a 0.25 M aqueous K2CO3
were then loaded into their respective syringes. The flow reac-
tion was then performed as described in the general pro-
cedure. The collected organic phase was dried over anhydrous
Na2SO4, filtered and dried under reduced pressure to give
Au(I)–NHC complex 3a as a white solid in 95% yield (521 mg,
0.839 mmol).
C
carbene), 132.3, 132.0, 126.4, 124.9, 29.2, 26.7, 22.9. NMR data
correspond to previously reported data.24
Author contributions
Single-phase batch synthesis of Au(I)–NHC complex 3a
HFJ and AJH conceived the study, performed experiments, and
NHC preligand 1a (30.0 mg, 0.0706 mmol) and Me2SAuCl wrote the manuscript. HFJ analysed data. AF assisted with con-
(20.8 mg, 0.0706 mmol) were dissolved in 6 ml DCM and ceptualisation and reviewed the manuscript.
7974 | Dalton Trans., 2021, 50, 7969–7975
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
C. S. J. Cazin and S. P. Nolan, Chem. – Eur. J., 2020, 26,
4515–4519.
Conflicts of interest
There are no conflicts of interest to declare.
13 H. F. Jónsson, A. Orthaber and A. Fiksdahl, Dalton Trans.,
2021, 50, 5128–5138.
14 M. B. Plutschack, B. Pieber, K. Gilmore and
P. H. Seeberger, Chem. Rev., 2017, 117, 11796–11893.
15 A. Adamo, P. L. Heider, N. Weeranoppanant and
K. F. Jensen, Ind. Eng. Chem. Res., 2013, 52, 10802–
10808.
References
1 M. N. Hopkinson, C. Richter, M. Schedler and F. Glorius,
Nature, 2014, 510, 485–496.
2 S. Díez-González, N. Marion and S. P. Nolan, Chem. Rev.,
2009, 109, 3612–3676.
16 A. J. Harvie, J. O. Herrington and J. C. DeMello, React.
Chem. Eng., 2019, 4, 1579–1588.
3 N. Marion and S. P. Nolan, Chem. Soc. Rev., 2008, 37, 1776–
1782.
17 J. H. Bannock, S. H. Krishnadasan, M. Heeney and J. C. de
Mello, Mater. Horiz., 2014, 1, 373–378.
4 M. Michalak and W. Kośnik, Catalysts, 2019, 9, 890.
5 L. Rocchigiani and M. Bochmann, Chem. Rev., DOI:
10.1021/acs.chemrev.0c00552.
18 A. Simoens, T. Scattolin, T. Cauwenbergh, G. Pisanò,
C. S. J. Cazin, C. V. Stevens and S. P. Nolan, Chem. – Eur. J.,
2021, 27, 5653–5657.
6 M. Mora, M. C. Gimeno and R. Visbal, Chem. Soc. Rev.,
2019, 48, 447–462.
19 A. Johnson and M. C. Gimeno, Chem. Commun., 2016, 52,
9664–9667.
7 T. Scattolin and S. P. Nolan, Trends Chem., 2020, 2, 721–736.
8 E. A. Martynova, N. V. Tzouras, G. Pisanò, C. S. J. Cazin and
S. P. Nolan, Chem. Commun., 2021, 57, 3836–3856.
9 D. J. Nelson, Eur. J. Inorg. Chem., 2015, 2015, 2012–2027.
10 A. Collado, A. Gómez-Suárez, A. R. Martin, A. M. Z. Slawin
and S. P. Nolan, Chem. Commun., 2013, 49, 5541–5543.
11 R. Visbal, A. Laguna and M. C. Gimeno, Chem. Commun.,
2013, 49, 5642–5644.
20 S. Gonell, M. Poyatos and E. Peris, Angew. Chem., Int. Ed.,
2013, 52, 7009–7013.
21 T. Scattolin, N. V. Tzouras, L. Falivene, L. Cavallo and
S. P. Nolan, Dalton Trans., 2020, 49, 9694–9700.
22 P. de Frémont, R. Singh, E. D. Stevens, J. L. Petersen and
S. P. Nolan, Organometallics, 2007, 26, 1376–1385.
23 P. de Frémont, N. M. Scott, E. D. Stevens and S. P. Nolan,
Organometallics, 2005, 24, 2411–2418.
12 N. V. Tzouras, F. Nahra, L. Falivene, L. Cavallo, M. Saab,
K. Van Hecke, A. Collado, C. J. Collett, A. D. Smith,
24 S. Gaillard, A. M. Z. Slawin, A. T. Bonura, E. D. Stevens and
S. P. Nolan, Organometallics, 2010, 29, 394–402.
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 7969–7975 | 7975