´
´
K. Grudzien, W. Nogas, G. Szczepaniak et al.
Polyhedron 199 (2021) 115090
Unfortunately, when we repeated this process on a bigger scale
(48 mmol) we observed a significant drop in yield (48%), presum-
ably due to the partial decomposition of 3 during rather tedious sil-
ica-gel purification.
To solve this problem, we opted to modify the work-up proce-
dure so it would allow us to omit the potentially problematic chro-
matographic purification step. After some exploratory
experiments, we derived a new protocol that avoids silica-gel
purification. After full conversion of 2, the solvent (CH3CN) was
removed in vacuo and replaced by dichloromethane (DCM). Then
a newly developed, commercially available isocyanide metal scav-
enger SC (SnatchCat) [20] (5.8 mol%; 1.16 equiv. relative to Cu) was
added. The colour of the mixture immediately changed from blue
to green, and after a while, large amounts of green, sticky, gluti-
nous solid formed, while the solution turned clear yellow (Fig. 2).
This mixture was filtered through a short pad of neutral Al2O3
placed in a funnel and the pale-yellow coloured clear filtrate was
concentrated yielding 91% of crude aldehyde 3 (48.0 mmol scale,
Scheme 4). Such produced 3 was directly converted into imida-
zolinium chloride 4 in 73% total yield (19.0 mmol scale) via 3-step
sequence consisted of (i) condensation with N-Mes-ethylenedi-
amine; (ii) NaBH4 reduction and (iii) cyclisation using triethyl
orthoformate. No chromatography separation was performed and
salt 4 was isolated by precipitation with Et2O and filtration. Finally,
N-Boc group was removed by reaction with a dioxane solution of
HCl. The resulting product turned out to be extremely hygroscopic.
Therefore we decided to exchange Clꢁ anions with hydrophobic
PFꢁ6 by reaction with KPF6 in water, which yielded in beige, crys-
talline, non-hydrophilic solid 5 which was conveniently purified
and isolated by crystallization from hot water. The total yield of
deprotection and anion exchange steps was 75% (0.5 mmol scale;
Scheme 2).
Fig. 1. General structure of N-heterocyclic carbene (NHC, insert) and selected
precursors of chiral: a) (hydroxyalkyl)imidazolinium and imidazolium NHCs; b)
(aminoalkyl)imidazolium NHCs useful in catalysis; c) imidazolium NHC derived
from L-proline; d) closely related imidazolinium NHC selected for this study.
2. Results
Attempts to explain the Cu-scavenger mode of action.
Despite the use of SnatchCat (SC) to scavenge copper complexes
was very recently disclosed [20d], no information about the exact
nature of Cu/SC complexes was reported. We speculate that during
the oxidation step work-up the bis(isocyanide) fragment might
have formed with Cu ions a kind of coordination polymer, which
is easily removable from the reaction mixture, due to its very
low solubility and high polarity [21]. Unfortunately, because of
the physical properties of this product (green, sticky, glutinous,
insoluble solid – see: Fig. 2. picture 4. and 5.) we were not able
to characterize it and thus its structure remains unknown. To shed
some light on the possible structure of such complex we switched
to mono-isocyanide analogue [17a] (SC’, Fig. 3) in a hope that in
this case non-polymeric material will be formed that will be easier
to characterise. Copper(I) iodide was selected as a source of metal
in this model study. Fortuitously, it was relatively easy to grow a
single crystal of a complex formed when CuI was reacted with
SC’. The complex crystallizes in the trigonal R3c space group with
one third of the molecule in the asymmetric unit of the crystal lat-
tice (for bond lengths, valence and torsion angles bond lengths, see
SI). The solidstate structure, presented in Fig. 3 shows that three
molecules of isocyanide SC’ bond to the copper cation forming
(SC’)3CuI type of complex (Fig. 1). We are aware that this molecule
can be understood as only a very approximate model for the real
complexes formed between bis(isocyanide) SC and a variety of
Cu species present in solution after the Stahl oxidation. Nonethe-
less, one can assume that because Cu ion was found to bond to
more than one isocyanide group, and SnatchCat (SC) molecule
has two opposite isocyanide groups available, the complexes
formed shall be of polymeric nature ([Cu] :CN-A-NC:?[Cu] :
CN-A-NC:?)n. This can explain the formation of a sticky glutinous
insoluble mass observed by us, which was then very easy to be
separated, leading to clear solution of the aldehyde product (cf.
photographs in Fig. 2).
State of the art. The reported syntheses of related hydroxyalkyl-
NHC precursors are quite straightforward and do not employ costly
reagents (Scheme 1a) [7]. Therefore, our aim was to develop a
route leading to chiral aminoalkyl-NHC precursors that is also
easily scalable and relies on readily available substrates and
reagents. The presence of L-proline function in the target com-
pound requires usage of the suitable protecting group throughout
the synthesis (Scheme 1b).
Our synthesis. Retrosynthesis of this new family of ligands
(Scheme 1b) includes following elementary transformations of
general character: (1) removal of the N-Boc protecting group; (2)
heterocyclization; (3) imine condensation and in situ reduction;
(4) selective reduction or reduction/oxidation. This analysis was
then reduced to practice using the arbitrarily selected proline-
based NHC ligand precursor Fig. 1d). The synthesis of the ligand
precursor is presented on Scheme 2 (for detailed experimental pro-
cedures and analytical data see Supporting Information, SI). Both
N-Boc-L-proline 1 and N-Boc-L-prolinol 2 are commercially avail-
able, but due to a much higher price of 2 as compared with 1 we
decided to convert the former into the latter, which was made pos-
sible in quantitative yield using simple reducing reagent (BH3-
ꢀSMe2). Next we faced
a challenge to perform a selective
oxidation of 2. Fortunately, this particular reaction can be carried
out in aerobic conditions using Cu(I)-TEMPO-bipy-NMI system,
described by Hoover and Stahl, giving > 98% yield of aldehyde 3
on a 1.0 mmol scale after silica-gel purification [19]. We found it
a particularly interesting procedure as it uses atmospheric oxygen
as a stoichiometric oxidant and earth-abundant metal as a catalyst.
A convenient way of monitoring of the reaction progress is another
advantage: the initial solution of the reaction is dark red, which
turns blue upon completion (the change in brown to green colour
indicating a change in the resting state of the copper species) [19].
2
Grudzień, Krzysztof
