Organic Process Research & Development
bis-NHC complex was produced in this reaction, in contrast to
the analogous reaction with copper electrodes. Au(I)−NHC
complexes are often isostructural with Cu(I)−NHC com-
plexes, but in this case the larger size of the Au(I) cation may
provide reduced steric hindrance, resulting in the formation of
24,25
the bis-NHC Au(I) complex.
When imidazolium salt L3
was subjected to identical reaction conditions using Au
electrodes, a conversion of 63% to the mono-NHC Au(I)
complex was observed, and no bis-NHC Au(I) complex was
observed. This complex was also isolated in 77% yield with 2.0
electron equivalents (Au3). Iron foil was a suitable electrode
material for the electrosynthesis of Fe complexes of different
oxidation states. When imidazolium salt L6 was subjected to
the standard reaction conditions, 67% conversion to the Fe(II)
complex Fe6 was observed. However, when salen ligand L7
was used in place of the imidazolium salt and the reaction was
carried out in air, full conversion to the Fe(III) complex Fe7
was obtained with an isolated yield of 99%. This reaction was
performed using 2.4 electron equivalents, as the ligand
Figure 3. Parallel screening reactions were performed to optimize an
oxidative coupling of amines and thiols.
10,26
precursor had to be reduced twice.
This highlights the
versatility of the electrochemical procedure to allow access to
different metal oxidation states.
To expand the applicability of the bottle reactor, its utility in
synthetic organic electrochemistry was examined. Specifically, a
derived from the smaller interelectrode gap in this setup
(5 mm) compared with that used by Noel and co-workers (7
̈
26
mm). Decreasing the potential further to 3.2 V resulted in an
increase in the yield to 51% (Table 1, entry 3). The
galvanostatic mode was found to be more efficient, as a yield
of 58% was obtained when a constant current of 20 mA was
applied (Table 1, entry 4). This yield is comparable to that
sulfonamide synthesis recently reported by Noel and co-
̈
26
workers (Table 1) was investigated. This reaction involves
Table 1. Rapid Screening and Optimization of an Oxidative
Coupling of Thiols and Amines Using a Parallel Plate
reported by Noel and co-workers. Finally, because of the
̈
success of the galvanostatic mode, a further experiment was
performed. A constant current of 10 mA was applied for
double the time (48 h) to deliver the same number of electron
equivalents at a lower potential, but this delivered a decreased
yield of 40% (Table 1, entry 5). The successful and fast
optimization of this synthetic organic electrochemical reaction
by performing six reactions in parallel showcases the bottle
reactor’s versatility and highlights the requirement for well-
characterized experimental setups to enable interlaboratory
reproducibility.
E/
appl,cell
average
(%)
entry
mode
I
yield 1 (%) yield 2 (%)
1
2
3
4
5
potentiostatic
potentiostatic
potentiostatic
galvanostatic
3.7 V
3.4 V
3.2 V
20 mA
10 mA
<2
−
37
−
39
51
58
a
a
a
a
41
Cyclic Voltammetry. In addition to a synthetic setup, the
bottle reactor can be used to perform cyclic voltammetry and
other analytical experiments. When the reactor was connected
to an external potentiostat and a reference electrode was used,
high-quality voltammograms were obtained. This was achieved
by connecting a glassy carbon disk working electrode and
platinum wire counter electrode as they would be for synthetic
electrochemistry in the bottle reactor. A silver wire, acting as a
pseudoreference electrode, was connected using a standard
a
a
b
a
48
56
54 (48 )
60 (58 )
−
a
b
a
,
a c
galvanostatic
40
−
a
1
Determined by H NMR spectroscopy with the aid of an internal
standard. Isolated yield. 48 h reaction time.
b
c
the oxidative coupling of amines and thiols through an
electrochemical mechanism, with H2 gas being the only
byproduct. To begin the investigation, the reported batch
conditions were repeated. Unsurprisingly, the results obtained
1
23
/4
‑28 fitting through one of the ports. Cyclic voltammo-
grams of ligand precursors L1−L5 using ferrocene as an
internal standard were obtained in this manner (Figure 4). All
of the voltammograms showed an irreversible reduction
between −1.9 and −2.5 V vs SCE, corresponding to reduction
of the imidazolium salt (Figure 4). Also, L1, L3, and L4
were different from those obtained by Noel and co-workers
̈
because of the different experimental setup, with the most
significant difference being the interelectrode gap and lower
electrolyte concentration. Utilizing the reaction conditions
−
described by Noel
̈
and co-workers involving a potentiostatic
displayed peaks corresponding to the oxidation of Cl and
−
27
mode with the potential set at 3.7 V resulted in no observed
product in our setup (Table 1, entry 1). Fortunately, the
reaction was quickly optimized by employing a parallel plate
and performing a set of three reaction conditions in duplicate
Br , and additional peaks were observed for L4 that were
28
potentially due to oxidation of the allyl groups.
CONCLUSIONS
■
(i.e., six reactions were performed simultaneously) (Figure 3).
A new flexible and standardized electrochemical batch reactor
enabling both organic and inorganic transformations and
analytical electrochemistry has been designed, developed, and
tested. This has enabled the synthesis of Cu(I)−, Au(I)−, and
Fe(II)−NHC complexes as well as an Fe(III)−salen complex.
Both galvanostatic and potentiostatic modes with a range of
currents and potentials were screened (Table 1). The yield
dramatically increased to 39% when the potential was lowered
to 3.4 V (Table 1, entry 2). The increase in yield is potentially
D
Org. Process Res. Dev. XXXX, XXX, XXX−XXX