Rate Enhancement of Acid-Catalyzed Alcohol Dehydration by Supramolecular Organic Capsules

Article abstract of DOI:10.1021/acscatal.0c03625

The presence of a (sub)nanometric environment has been demonstrated to enhance the rate of homogeneous acid-catalyzed reactions. Conventionally, such catalysts are constructed by directly embedding a catalytically active acid site into the structure of a porous material, linking the acid properties with the properties of the embedding solid. In this work, a holoenzyme mimicking approach was used, in which a self-assembled hexameric resorcinarene capsule (1.4 nm3 cavity) acts like an apoenzyme that transforms into a catalytically active site by hosting hydrated HCl as a cofactor. This capsule binds the hydronium ion together with cyclohexanol (CyOH) and catalyzes the dehydration to cyclohexene with enhanced rates that are 2 orders of magnitude higher than those in the unconstrained environment. Kinetic analysis shows that cyclohexanol dehydration proceeds via an E2 mechanism on hydrated HCl in the unconstrained environment, while the constrained environment of the active capsule induces the stabilization of an ionic intermediate, allowing an E1 mechanism to dominate. By comparison with the reaction in zeolite environments, we show that the direct chemical environment of the capsule or a zeolite wall exerts only a minor direct influence. The volume constraint increases the reaction rate by inducing a larger activation entropy, indicating the stabilization of a later transition state in the E1 mechanism, as well as a higher reaction space of the carbenium ion constituting the transition state.

Full text of DOI:10.1021/acscatal.0c03625

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
pubs.acs.org/acscatalysis
Research Article
Rate Enhancement of Acid-Catalyzed Alcohol Dehydration by
Supramolecular Organic Capsules
Wei Zhang, Guanhua Cheng, Gary L. Haller, Yue Liu,* and Johannes A. Lercher*
Cite This: ACS Catal. 2020, 10, 13371−13376
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: The presence of a (sub)nanometric environment has been
demonstrated to enhance the rate of homogeneous acid-catalyzed reactions.
Conventionally, such catalysts are constructed by directly embedding a catalytically
active acid site into the structure of a porous material, linking the acid properties
with the properties of the embedding solid. In this work, a holoenzyme mimicking
approach was used, in which a self-assembled hexameric resorcinarene capsule (1.4
nm3 cavity) acts like an apoenzyme that transforms into a catalytically active site by
hosting hydrated HCl as a cofactor. This capsule binds the hydronium ion together
with cyclohexanol (CyOH) and catalyzes the dehydration to cyclohexene with
enhanced rates that are 2 orders of magnitude higher than those in the
unconstrained environment. Kinetic analysis shows that cyclohexanol dehydration
proceeds via an E2 mechanism on hydrated HCl in the unconstrained environment, while the constrained environment of the active
capsule induces the stabilization of an ionic intermediate, allowing an E1 mechanism to dominate. By comparison with the reaction
in zeolite environments, we show that the direct chemical environment of the capsule or a zeolite wall exerts only a minor direct
influence. The volume constraint increases the reaction rate by inducing a larger activation entropy, indicating the stabilization of a
later transition state in the E1 mechanism, as well as a higher reaction space of the carbenium ion constituting the transition state.
KEYWORDS: acid catalysis, alcohol dehydration, kinetics, confinement, supramolecular capsule
containing cavities, known as molecular containers,9 show
INTRODUCTION
■
higher robustness.
Chemical reactions in (sub)nanometric dimensions have been
Molecular containers include covalent molecules such as
demonstrated to increase the reaction rate effectively
cyclodextrins and corresponding derivatives10,11 and metal−
compared to rates of catalytic transformations in unconstrained
ligand and metal-free supramolecular self-assemblies.9,12−15
environments.1−3 To realize reactions in a constrained space,
Such containers provide a rigid organic framework in a single
the conventional heterogeneous catalyst design usually embeds
molecule or molecular ensemble, which isolates a confined
catalytically active sites directly into porous materials, such as
void from bulk solution for binding substrates. At variance with
Brønsted or Lewis acid sites created by introducing Al(III),
inorganic porous catalysts possessing both a (sub)nanometric
B(III), and Sn(IV) into zeolite siliceous frameworks.4−8 This
space and a catalytically active site for a reaction, some
effective strategy encounters challenges caused by long
diffusion paths required to access the catalytically active sites.
Such limitations would be conceptually mitigated by
mimicking apoenzymes and coenzymes/cofactors that are
structurally separated but capable of combining into a
catalytically active holoenzyme. In the holoenzyme entity, the
apoenzyme part offers the constrained space for the catalytic
reaction, while the coenzyme/cofactor provides the catalyti-
cally active center. For a Brønsted acid catalyst, such catalyst
design means that the acid site itself exists as an individual
chemical entity (e.g., a hydrated acid molecule forming
hydrated hydronium ions) and combines with a porous host
to form an active catalyst. The stability of host lattices (e.g.,
molecular organic frameworks) and the hydrophobicity of the
lattice (for siliceous microporous structures) drastically limit
such an approach. Linked and supramolecular entities
molecular containers only act as hosts encapsulating external
catalytically active species and the reacting substrate into their
cavities where the reaction occurs.9,16
Self-assembled hexameric resorcinarene capsules17 are
among the promising supramolecular examples with nanosized
cavities. The resorcinarene capsule is self-assembled from six
monomers (Figure 1) together with eight water molecules,
stabilized by 60 hydrogen bonds in an apolar solvent. This
leads to a cavity of 1.4 nm3 internal volume, containing weak
Received: August 18, 2020
Revised: October 17, 2020
https://dx.doi.org/10.1021/acscatal.0c03625
© XXXX American Chemical Society
ACS Catal. 2020, 10, 13371−13376
13371

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
ACS Catalysis
pubs.acs.org/acscatalysis
Research Article
Figure 1. Schematic illustration of resorcinarene (I) self-assembling
into a capsule (II) in water-saturated chloroform and catalyzing
cyclohexanol (CyOH) dehydration in the presence of HCl.
Brønsted acid sites (pKa 5.6−5.9).17−19 Suitable sized cationic
substrates can be encapsulated in the cavity by means of
electrophilic π-interactions with the aromatic ring of
resorcinarene.16,20,21 Exchange of molecules between the
external solvent and the inside of the cage occurs via partial
dissociation of the hexameric capsule.22 Such resorcinarene
capsules have been used, for example, as enzyme mimics in
acetal hydrolysis,18 terpene cyclization,23 Friedel−Crafts
reactions,24 and other organic transformations,25−27 illustrating
substrate and regioselectivity. By hosting mineral acids such as
HCl, catalytic activity for acetal hydrolysis has been reported.16
We show here that resorcinarene capsules hosting hydrated
HCl (H+(H2O)n···Cl−) catalyze alcohol dehydration at a rate
that exceeds the rates of hydrated hydronium ions
(H+(H2O)n) in an unconstrained environment to an extent
that is even higher than the increase induced by zeolite cavities.
Different from previous reports that use the proton in the
capsule hydrogen network as the acid site, we address that the
catalytic site here is a hydrated hydronium ion that is in the
form of a Brønsted acid site in the presence of water in
particular, in the reactions generating water (i.e., alcohol
dehydration in this work). We quantitatively establish the
association of the reactant and the hydrated acid within the
capsules using NMR techniques. A kinetic model demonstrates
a similar impact of the two environments as long as the
reaction volume constraints are identical.
Figure 2. Encapsulation of cyclohexanol in the resorcinarene capsule.
RESULTS
(A) Cyclohexanol encapsulation in resorcinarene capsule II (3.3 mM)
■
1
with H NMR; encapsulation degree θ is the ratio of encapsulated
The hexameric resorcinarene capsule hosts both neutral and
cationic guests,28 having high affinity for cations (such as
aminium ions) through cation−π interactions with aromatic
rings.18,19,22,29 A fraction of cyclohexanol in contact with 3.3
mM resorcinarene capsules in water-saturated CHCl3 is,
therefore, enclosed in the formed capsule (monitored by H
NMR and diffusion ordered spectroscopy (DOSY)). Free
cyclohexanol dissolved in CHCl3 has peaks at around 1.1−2.0
cyclohexanol to the capsule. (B) Diffusion ordered spectroscopy
(DOSY) of the cyclohexcanol−capsule (II) mixture in CDCl3 (3.3
mM capsule and 19.8 mM cyclohexanol or 6.6 mM TBAB). (C)
Diffusion coefficients of different species in the capsule solution from
DOSY by NMR.
unassociated cyclohexanol, and TBAB were much higher
(Figure 2B,C). Both the new peaks with low chemical shifts
and the low diffusion coefficients determined by DOSY lead to
the conclusion that a fraction of cyclohexanol has been
successfully encapsulated. Cyclohexene in contact with the
resorcinarene capsules neither shows signals in the low
1
and 3.6 ppm in the H NMR spectrum, corresponding to the
H signals of methylene and methylidyne groups. In the
presence of resorcinarene capsules, new peaks appeared at a
chemical shift of around −1.0 to 0 (Figure 2A). These are
attributed to encapsulated and highly shielded molecules inside
the capsules.20 The intensities of the new peaks increased in
parallel with the increase of the cyclohexanol concentration
until reaching about two cyclohexanol molecules enclosed per
capsule. Encapsulated cyclohexanol has the same diffusion
coefficient (0.2 × 10−5 cm2 s−1) as that of the capsules and of
encapsulated tetrabutylammonium bromide (TBAB) (Figure
S2).16,30 The diffusion coefficients of the solvent molecules,
1
chemical shift region in H NMR (Figure S3A) nor shows a
decrease in the diffusion rate in DOSY spectra (Figures 2C and
S3B). Thus, we conclude that cyclohexene is not as well
stabilized in the capsule as cyclohexanol and tends to detach as
readily as formed.
Dehydration of cyclohexanol was investigated at 333 K in
CHCl3 using HCl or/and resorcinarene capsules as catalysts
13372
https://dx.doi.org/10.1021/acscatal.0c03625
ACS Catal. 2020, 10, 13371−13376

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
ACS Catalysis
pubs.acs.org/acscatalysis
Research Article
Figure 3. Rate of cyclohexanol dehydration catalyzed by the capsule. (A) Cyclohexanol dehydration rate on the resorcinarene capsule with different
catalysts (reaction conditions: 20 mM cyclohexanol, 3.3 mM capsule, 32 mM HCl, and 6.6 mM TBAB, 333 K.) (B) Reaction rate as a function of
cyclohexanol concentration in the presence of 32 mM HCl. (C) Reaction rate as a function of HCl concentration in the presence of 3.3 mM
cyclohexanol.
Figure 4. Reaction pathways of cyclohexanol dehydration on different catalysts. Reaction pathways of cyclohexanol catalyzed by the resorcinarene
capsule in the presence of HCl (path A), zeolite HBEA (path B), and HCl alone (path C). The blue dashed square represents the capsule.
(Figure 3A). The conversion rate of cyclohexanol was below 6
Figure 3B,C shows the dehydration rate of cyclohexanol as a
function of cyclohexanol concentration and HCl concen-
tration, respectively. The reaction rate increased with the
cyclohexanol concentration at low concentrations, while
reaching a plateau at high concentrations (Figure 3B), in a
way identical to Michaelis−Menten kinetics in enzyme
catalysis. This is concluded to be caused by an increase of
cyclohexanol in the capsules, reaching saturation at high
concentrations. A similar increase in rate was observed upon
variation of the HCl concentration (Figure 3C).
Based on the concentration dependence above, we
hypothesize the reaction pathway in the resorcinarene capsule
to follow Figure 4A. Cyclohexanol is encapsulated with an
equilibrium constant (Keon). Hydronium ions, H+(H2O)n,
formed by HCl dissociation in the presence of water are in
part also included in the resorcinarene capsule (equilibrium
constant KoH), which transforms the capsule into a catalytically
active capsule. This leads to association and protonation of the
encapsulated cyclohexanol. In the next step, the protonated
alcohol eliminates water either in an E1 or in an E2 step (with
the intrinsic rate constant k). Thus, the rate equation is
obtained (see detailed derivations in Supporting Information
(SI) “Derivations of rate equations”) as
× 10−7 mol molacid−1 s−1 in the presence of only HCl or below
2 × 10−8 mol molcapsule s−1 in the presence of only the
−1
capsule in CHCl3 at 333 K (Figure 3A). Thus, the Brønsted
acid sites (hydrogen-bonded water and the OH groups of the
capsule constituents) of the resorcinarene capsule are
concluded to be insufficiently strong to catalyze the
dehydration of cyclohexanol. When capsules and HCl were
used together, the conversion rate increased to 2.3 × 10−5 mol
molacid−1 s−1 (Figure 3A). When the capsule cavity was blocked
by tetrabutylammonium bromide (TBAB), the reaction rate of
cyclohexanol was markedly lower (Figure 3A), demonstrating
that the dehydration of cyclohexanol occurs only inside the
capsule.
Besides cyclohexene, the conversion of cyclohexanol on HCl
alone also produces chlorocyclohexane (C6H11Cl), which is
the substitution product of cyclohexanol reacting with the Cl−
ion (Figure S4). Chlorocyclohexane could also be produced via
HCl addition to cyclohexene. However, this pathway is
excluded because chlorocyclohexane was identified as a
primary product from cyclohexanol (Figure S4) and the rate
of HCl addition to cyclohexene measured separately was 2
orders of magnitude lower than the chlorocyclohexane
formation rate observed here (e.g., 2.7 × 10−7 vs 1.4 × 10−5
mol molacid s−1 at 408 K).
−1
13373
https://dx.doi.org/10.1021/acscatal.0c03625
ACS Catal. 2020, 10, 13371−13376

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
ACS Catalysis
pubs.acs.org/acscatalysis
Research Article
Keon[Cy]
KHo[H3O+]
transitional and rotational freedoms, leading to the large loss of
entropy.31 In this mechanism, the C−O and C−H bonds in
cyclohexanol are cleaved in one concerted step. When the
reaction occurs in the active capsule, the highly polar
confinement in the capsule appears to stabilize the cyclohexyl
carbenium ion, shifting the reaction mechanism from E2 to E1.
The E1 mechanism is also seen in cyclohexanol dehydration by
hydrated hydronium ions in the aqueous phase,8 which is by
far more polar than CHCl3. It should be emphasized that the
higher transition entropy suggests that the constrained
environment induces a later transition state and/or a higher
reaction volume for the elimination reaction.
Chlorocyclohexane is formed via the substitution of the OH
group in cyclohexanol by the Cl− ion (Figure 4C).
Chlorocyclohexane is not observed, and we hypothesize that
blocking this reaction route is related to the strong hydrogen
bonding of the chloride anion to the capsule framework,24,26
hindering it to approach cyclohexanol from the appropriate
orientation.
r = k[Cap]·
·
1 + Keon[Cy] 1 + KHo[H3O+]
(1)
where r is the dehydration rate, [Cap] is the concentration of
resorcinarene capsules, and [Cy] and [H3O+] are the
cyclohexanol and the hydronium ion concentrations, respec-
tively. Regression of the plots in Figure 3 using eq 1 gives the
values of Koen, KoH, and k at different temperatures (Table 1).
Table 1. Kinetic Parameters for Cyclohexanol Dehydration
Catalyzed by the Resorcinarene Capsule with HCl
temperature (K)
Koen
KHo
k (s−1
)
318
323
328
333
586
446
296
217
−60
30
16
12
6.5
1.3 × 10−5
2.8 × 10−5
5.8 × 10−5
1.4 × 10−4
ΔHo or ΔHo‡ (kJ mol−1
)
4
−86
6
138
5
ΔSo or ΔSo‡ (J mol−1 K−1
)
−134 11
−243 19
93 16
For comparison, zeolite HBEA, a microporous solid acid,
was tested in cyclohexanol dehydration. It has an identical
reacting space for each turnover of cyclohexanol, that is, 1.4
nm3 in the capsule cavity and 1.1 nm3 micropore volume per
BAS in HBEA (0.20 cm3 g−1 micropore volume normalized to
BAS concentration). Under identical conditions, it shows a
catalytic activity higher than free HCl and lower than the active
capsule (Figure 5) and a reaction order of zero for
cyclohexanol (Figure S8). The activation energy and activation
entropy for the active capsule and in HBEA are very close (145
vs 138 kJ mo−1; 93 vs 97 J mol−1 K−1, Figure 5), while largely
different from those in HCl (47 kJ mol−1, −212 J mol−1 K−1).
This indicates the same change in the mechanism and kinetics
of the reaction by confining the reacting space, irrespective of
whether an organic or an inorganic structure is constituting the
pore or cavity.
Raising the reaction temperature reduced Keon and KHo ,
indicating an exothermic process for the encapsulation of
cyclohexanol and the association of hydronium ions with the
capsule. Using the van’t Hoff relation (Figure S6A,B), the
enthalpy and entropy changes for the two steps were obtained,
and by the Eyring relation (Figure S6C), the activation
enthalpy and entropy were deduced. The enthalpies of
cyclohexanol encapsulation and hydronium ion association
were −60 and −86 kJ mol−1, respectively; the intrinsic
activation enthalpy is 138 kJ mol−1, and the activation entropy
is 93 J mol−1 K−1. Such high intrinsic activation enthalpy and
entropy indicate that the cyclohexanol dehydration follows an
E1 mechanism, which proceeds via a metastable cyclohexyl
carbenium ion intermediate (C6H11+).8,31
In contrast, the activation enthalpy of the HCl-catalyzed
dehydration in CHCl3 with water is only 47 kJ mol−1 and the
activation entropy is −212 J mol−1 K−1 (Figures 5 and S7).
Such low activation enthalpy and entropy indicate that the
dehydration by hydronium ions in CHCl3 proceeds via an E2
mechanism. In this mechanism, the hydronium ion associates
with cyclohexanol to form the transition state and loses
In spite of these analogies, there is still a 7 kJ mol−1
difference in the activation energy of the active capsule and
HBEA, which caused a different rate by a factor of 8 (at 333
K). This is speculated to be caused by the differences in
forming the confined space. With the same reacting volume,
the capsule activity offers a completely closed space in all three
dimensions, while HBEA constrains the reacting space along
the channel walls but not along the channel axial dimension.
Thus, the reacting cyclohexanol in HBEA may have higher
freedom of motion, which agrees with the slightly higher
activation entropy in HBEA than in the capsules. However, the
capsule cavity constituted by hydrogen bonds is not as rigid as
the micropore channels held by covalent bonds in HBEA, so
that it is possibly capable of flexibly readjusting the dimension
to better fit the reacting transition state and thus decrease the
activation energy.
CONCLUSIONS
■
We have shown that resorcinarene hexamer capsules in CHCl3
are capable of encapsulating cyclohexanol with a maximum
capacity of two cyclohexanol molecules per capsule. In the
presence of HCl forming hydrated hydronium ions, the
encapsulated cyclohexanol is dehydrated to cyclohexene, with
rates 2 orders of magnitude higher than in the unconstrained
solvent. The reaction of cyclohexanol in the capsule did not
lead to partially chlorinated cyclohexanol because Cl− was
stabilized by interacting with the capsule and thus became less
reactive.
Figure 5. Arrhenius plots of the cyclohexanol dehydration rate on
different catalysts. Rate on HCl is normalized to the amount of HCl;
rate on HBEA is normalized to the BAS amount in the loaded HBEA;
and rate on the active capsule is normalized to the amount of the
active capsule. (Reaction conditions: 300 mM cyclohexanol, 150 mM
HCl, and 1/40 volume of water in chloroform (orange); 20 mM
cyclohexanol, 3.3 mM capsule, and 32 mM HCl in water-saturated
chloroform (blue); and 23 mg of water-saturated HBEA, 20 mM
cyclohexanol in 4 mL of water-saturated chloroform (black)).
13374
https://dx.doi.org/10.1021/acscatal.0c03625
ACS Catal. 2020, 10, 13371−13376