Organic Letters
Letter
conversion or lead to oligomerization into undefined product
mixtures. However, monothiol 3 was efficiently dimerized via
the literature-known mild oxidant DMSO (Scheme 1d).21 The
formation of disulfide 4 was confirmed by HRMS and NMR
NMR measurements have been established as a reliable tool
for the size determination of supramolecular capsules.22 In
order to verify the disulfide formation for 4, CDCl3/DMSO
1:1, a polar solvent mixture that prevents self-assembly via
hydrogen bonds, was chosen. A smaller diffusion value D was
observed for 4 as compared to its precursor 3 and the
evidence for the formation of dimeric species. The studies
concerning the disulfide formation indicated that the
investigated thiol macrocycles require irreversible oxidative
conditions for disulfide formation, which prevented the
formation of disulfide bridged capsules via DCvC.
Figure 3. (a) Stability of capsules I−IV toward the protic solvent
MeOD as characterized by the change in the diffusion value D upon
titration. (b) HCl-cocatalyzed supramolecular terpene cyclization of
nerol and geraniol. (c) Nerol conversion by I and III and controls
blocked by TBAB.
Next, the self-assembling properties of the only defined
disulfide product, disulfide 4, were investigated by means of
DOSY-NMR measurements in the apolar solvent CDCl3. The
diffusion value observed indicates the formation of a trimeric
capsule IV of similar dimensions as the systems I−III (Figure
terpene cyclizations on its own. Next, we investigated potential
differences in the terpene cyclization with HCl as a cocatalyst
in direct comparison to I (Figure 3c). A very similar product
distribution for the cyclization of nerol and geraniol was
proceeded slower and showed a more significant background
reaction when the capsules were blocked by the strongly
binding alkylammonium guest TBAB. The slower reaction and
the stronger background conversion might be a result of the
decreased stability of III compared to I (Figure 3a).
With capsules III and IV at hand, they were investigated
concerning their guest uptake capabilities. Both assemblies
encapsulate alkyl ammonium salts such as TBAB (SI, Chapter
4). This was expected as both assemblies possess similar
hydrogen bond networks as I, capable of anion stabilization.23
However, two observation came as a surprise. First, both III
and IV showed a fast guest exchange for the alkyl ammonium
1
salt on the H NMR time scale, and second, III was able to
encapsulate fullerenes is unexpected since I, II, and IV are
incapable of functioning as hosts for this type of guest (SI,
capsule has been reported to encapsulate both alkylammonium
salts and fullerenes before.13b,24 The fast guest exchange on the
1H NMR time scale for alkyl ammonium salts is likely a result
of the decreased stability of capsules III and IV as compared to
I and II. To quantify the stability of the assemblies and gain
insight into the strength of their hydrogen bond network,
DOSY-NMR titrations with MeOD were performed.25 These
studies demonstrated that both III and IV are considerably less
stable than their relatives I and II, as they disassemble at much
attribute these findings to the presence of the sterically
demanding sulfur atoms disrupting the hydrogen bond
network in both III and IV. According to computational
analysis, the bulky thiols of assembly III do not participate in
the hydrogen bonding, and the disulfide bridges present in IV
add additional constraints to the supramolecular structure (SI,
Finally, we tried to answer the third question posed in the
introduction: Is the increased acidity of a thiol-bearing capsule
3b) without the need for HCl as cocatalyst? Capsule III is the
only suitable assembly identified in this study to answer this
question, as assembly IV does not contain free thiols. Test
reactions with nerol and geraniol demonstrated no significant
conversion, while geranyl acetate suffered from a nucleophilic
attack by 3’s thiol, leading to alkylation and subsequent
that the increased acidity of III is not sufficient to catalyze
In summary, we have demonstrated the applicability of two
distinct methods to install thiols on the resorcin[4]arene
framework. These protocols allowed us to synthesize three
novel thiol analogs of resorcin[4]arene and pyrogallol[4]arene,
investigate the influence of free thiols on their properties, and
compare them to RS and PG. We found that (1) Thiol-
derivatives of RS can self-assemble into a hexameric capsule as
long as the amount of thiols is kept to a minimum. (2)
Disulfide formation with thiol-derivatized resorcin[4]arenes is
possible but requires irreversible oxidative conditions. Thus, it
is limited to monothiol macrocycles in order to prevent
oligomerization, and is not suited for DCvC. The dimeric
disulfide 4 obtained under mild oxidative conditions from 3
forms a trimeric capsule IV of similar size as capsules I−III, as
confirmed by DOSY-NMR studies. (3) The increased acidity
introduced by the six thiols of III is not sufficient to catalyze
terpene cyclizations on its own. Additionally, it was found that
capsule III displays unusual guest uptake properties by being
able to encapsulate both alkylammonium salts and C60-
fullerenes. Overall, these studies clarified the consequences of
introducing thiols into hydrogen bond-based molecular
capsules and contributed to the understanding of the
catalytically active resorcin[4]arene capsule that has attracted
growing interest over recent years.
ASSOCIATED CONTENT
* Supporting Information
■
sı
The Supporting Information is available free of charge at
Experimental details, computational models, and NMR
spectra and HRMS data of new compounds (PDF)
C
Org. Lett. XXXX, XXX, XXX−XXX