High Activity and Efficient Turnover by a Simple, Self-Assembled "artificial Diels-Alderase"

Article abstract of DOI:10.1021/jacs.7b12146

The Diels-Alder (DA) reaction is a cornerstone of synthesis, yet Nature does not use catalysts for intermolecular [4+2] cycloadditions. Attempts to create artificial "Diels-Alderases" have also met with limited success, plagued by product inhibition. Using a simple Pd₂L₄ capsule we now show DA catalysis that combines efficient turnover alongside enzyme-like hallmarks. This includes excellent activity (k_cat/k_uncat > 10³), selective transition-state stabilization comparable to the most proficient DA catalytic antibodies, and control over regio- and chemoselectivity that would otherwise be difficult to achieve using small-molecule catalysts. Unlike other catalytic approaches that use synthetic capsules, this method is not defined by entropic effects; instead multiple H-bonding interactions modulate reactivity, reminiscent of enzymatic action.

Full text of DOI:10.1021/jacs.7b12146

Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES
Article
High Activity and Efficient Turnover by a Simple,
Self-Assembled “Artificial Diels–Alderase”
Vicente Martí-Centelles, Andrew L. Lawrence, and Paul J. Lusby
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b12146 • Publication Date (Web): 06 Feb 2018
Downloaded from http://pubs.acs.org on February 6, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 8
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
High Activity and Efficient Turnover by a Simple, Self-
Assembled “Artificial Diels–Alderase”
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Vicente MartíꢀCentelles, Andrew L. Lawrence and Paul J. Lusby*
EaStCHEM School of Chemistry, University of Edinburgh, Joseph Black Building, David Brewster Road, Edinburgh, Scotꢀ
land, UK. EH9 3FJ. Eꢀmail: Paul.Lusby@ed.ac.uk.
ABSTRACT: The Diels–Alder (DA) reaction is a cornerstone of synthesis, yet Nature does not use catalysts for intermolecular
[4+2] cycloadditions. Attempts to create artificial “Diels–Alderases” have also met with limited success, plagued by product inhibiꢀ
tion. Using a simple Pd2L4 capsule we now show DA catalysis that combines efficient turnover alongside enzymeꢀlike hallꢀmarks.
This includes excellent activity (kcat/kuncat > 103), selective transitionꢀstate (TS) stabilization comparable to the most proficient DA
catalytic antibodies, and control over regioꢀ and chemoꢀselectivity that would otherwise be difficult to achieve using smallꢀ
molecule catalysts. Unlike other catalytic approaches that use synthetic capsules, this method is not defined by entropic effects;
instead multiple Hꢀbonding interactions modulate reactivity, reminiscent of enzymatic action.
that could provide direct interactions with either reactant(s) or
transition state (TS), in the same way an enzyme would use a
Introduction
Creating artificial catalysts that rival the levels of activity
and selectivity exhibited by enzymes remains challenging. A
bioꢀmimetic approach defined by using a microenvironment
superficially reminiscent of an enzyme active site is oneꢀway
reꢀsearchers have chosen to tackle this challenge.1 In this reꢀ
gard, selfꢀassembled systems are appealing because the conꢀ
struction of chemical microenvironments requires a macromoꢀ
lecular design, most easily realized using reversible, weak
interactions to bring together smaller components in a threeꢀ
dimensional array.2
collection of polar amino acid residues. In contrast, we envisꢀ
aged a new approach in which multiple nonꢀcovalent interacꢀ
tions would modulate reactivity (Figure 1b). This exploits a
novel guestꢀbinding approach recently described with the simꢀ
ple Pd2L4 capsule, C-1,15a wherein pockets of electron defiꢀ
cient Hꢀatoms (Figure 1c, shown in blue) form interactions
with complementary guests such as quinones.15b The increased
reactivity of the dienophile would therefore arise due to a lowꢀ
ering of the guest LUMO energy—akin to Hꢀbond organocaꢀ
talysis.16,17 An internalized and activated enone would elimiꢀ
nate the need to coꢀbind the diene, thereby decreasing product
inhibition effects, whilst still giving the capsule microenviꢀ
ronment opportunity to influence both reactivity and selectiviꢀ
ty.
A common theme to emerge from investigations with variꢀ
ous “supramolecular” catalyst systems is the use of entropic
mechanisms to underpin reactivity. Using multiꢀsubstrate enꢀ
capsulation to achieve high effective concentration of species
is an established concept, first exemplified by Breslow using
cyꢀclodextrins.3 Later, Kelly,4 Rebek5 and Sanders6 all showed
that bimolecular reactions could be similarly accelerated using
synthetic covalent receptors capable of binding multiple reacꢀ
tants. However, these and more recent studies with the selfꢀ
assembled systems of Rebek7, Fujita8 and others9 have reꢀ
vealed that catalysis is frequently halted before even one turnꢀ
over is complete due to tight product binding. This has been
particularly comꢀmon with cycloaddition reactions (Figure
1a), where the 100% atom efficiency reinforces product inhiꢀ
bition, although rare exceptions are known.10 As a result, the
recent trend in capsule catalysis has been to select reactions,
such as hydrolysis,11 ring openings12 and sigmaꢀtropic rearꢀ
rangements,13 on the basis that the products show the same or
preferably weaker affinity than the substrates. The avoidance
of using capsules to catalyze annulation reactions exposes a
weakness in current methods.
Results and Discussion
Initial tests of this catalytic strategy focused on the DA reacꢀ
tion of benzoquinone, 1, with a few small dienes (isoprene, 2;
cyclopentadiene 3; cyclohexadiene, 4). However, the results
using 20 mol% C-1 were not encouraging, showing no disꢀ
cernible acceleration compared to the control reactions (Figꢀ
ures 2a and S1; black squares uncatalyzed, red triangles with
C-1). Considering possible capsule modifications, we envisꢀ
aged that the endoꢀpyridyl variant,15c C-2 (Figure 1c), should
retain similar LUMOꢀlowering properties. However, it was
also reasoned that the lone pair could destabilize substrate
binding by replacing ArH···π with less favorable n···π interacꢀ
tions. As thermodynamic stabilization of reactants has a detꢀ
rimental effect on the catalytic activation barrier (ꢁG‡cat), we
were hopeful that C-2 could show increased reactivity. We
were thus pleased to observe that 20 mol% C-2 gave a proꢀ
nounced increase in the rate of production of the cycloadducts
5–7 (Figures 2a and S1; green triangles), and that each cataꢀ
lyzed reaction proceeded smoothly to completion. While the
stark difference in reactivity from almost identical capsules
highlights the subtleties of developing enzymeꢀlike catalysts,
the lack of activity from C-1 is also a useful control. Firstly, it
likely eliminates catalysis from residual Lewis acidic Pd2+
In addition to entropic effects, including constrictive bindꢀ
ing13 and increased effective concentration through ionꢀ
pairing,12b coulombic stabilization of charged intermediates by
cationic, anionic and even neutral capsules has also been inꢀ
voked to explain rate enhancement.1e,14 Nonetheless, the caviꢀ
ties of most synthetic capsules lack polar functional groups
ACS Paragon Plus Environment

Journal of the American Chemical Society
ions. It also hints that there is no contribution from the interacꢀ tion of 1 with the outer orthoꢀpyridyl Hꢀ
Page 2 of 8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. (a) Previous intermolecular DA capsule catalysis has utilized coꢀencapsulation. (b) Our approach utilizes internal dienophile
polarization. (c) Pd2L4 capsules, C-1 and C-2, used in this study.
bond donor sites. The confirmation that DA activity is deꢀ
pendent on quinoneꢀencapsulation was obtained by adding
excess competitive inhibitor, anthraquinone, which almost
completely stops any acceleration (Figures 2a and S1, yellow
circles).
As expected, C-2 binds substrate 1 less well (KA = 1100
M−1) in comparison to C-1 (KA = 8000 M−1). However, subꢀ
strate stabilization effects—wherein C-1 “deactivates” 1 by
1.2 kcal mol−1 in comparison to C-2 (Figure 3a)—only partly
explain the difference in DA activity. Specifically, C-2 apꢀ
pears much better suited to TS recognition,18 as evidenced by
the significantly tighter binding of conformationally locked
adduct 6 (KA = 11000 M−1), which is a close structural mimic
of the TS, and only binds weakly to C-1 (KA = 550 M−1).
Combining more effective TS stabilization—1.8 kcal mol−1
based on TS mimic 6—with less favorable substrate binding
(1.2 kcal mol−1) would give an estimate of 3 kcal mol−1 differꢀ
ence in catalytic activation barriers (ꢁG‡cat) for C-1 and C-2
1
based on measured thermodynamic parameters alone. The H
NMR titration data for 6 and C-2 (Figure S36) also showed
that several of the more Hꢀbond acidic atoms of 6 exhibit large
downfield shifts (ꢁδ > 1 ppm). This is consistent with the forꢀ
mation of favorable n···CH capsuleꢀproduct, and by extension,
capsuleꢀTS interactions, thus providing a plausible structural
rationale for the difference in reactivity between such similar
capsules (Figure 3a).
To calculate the C-2 rate enhancements (kcat/kuncat), we have
used a nonꢀlinear analysis that involves fitting total dienophile
consumption and/or product formation to a simulation based
on a simple kinetic model (see SI). This considers the contriꢀ
bution from the uncatalyzed reaction (kuncat), the rate of ingress
and egress of substrate and product, as well as the 2nd order
reaction of [1⊂C-2] with dienes 2–4 (kcat). Utilizing the subꢀ
strate and product KA values as ratios of inꢀout rates, and asꢀ
suming these are significantly quicker than cycloaddition
(based on the timeꢀaveraged NMR spectra), yielded kcat values
that are listed in Table S1. Additionally, kcat/kuncat for the reacꢀ
tion of quinone 1 with dienes 2ꢀ4 are graphically represented
in Figure 4a. The kcat, kuncat and KA
values have also
(quinone)
been used to calculate the TS stabilization energies (ꢁGTS;
Figure 3a, Table S2), which relates to the TS association conꢀ
stant (KA TS; Figure 4), sometimes referred to as catalytic proꢀ
ficiency (KA(quinone)·kcat/kuncat). This analysis reveals that the
binding strength of the TS mimic 6 significantly underestiꢀ
mates the acceleration afforded to the reaction of 1 and 3 by
C-2 (see above), as the energy barrier is lowered by more than
2 kcal mol−1 compared to ꢁGproduct (Figure 3a). It is even more
apparent that C-2 selectively stabilizes an early TS with the
DA reaction of acyclic diene 2 (Figure 3b). Here, the signifiꢀ
cant TS stabilization is maintained even though the flexibility
Figure 2. (a) Representative DA kinetic data. While C-2 gives a
marked acceleration (green triangles), structurally similar C-1
(red triangles) gives no enhancement over the uncatalyzed reacꢀ
tion (black squares). C-2 catalysis is retarded with anthraquinone
(yellow circles) showing dienophile encapsulation necessary. (b)
C-2 (10 mol%) still accelerates DA of 1 and 2 with up to 100 eq.
of product 5 added, showing 103 turnovers are feasible.
ACS Paragon Plus Environment

Page 3 of 8
Journal of the American Chemical Society
of product 5 allow conformational reꢀorganization resulting in
weak adduct binding (KA = 500 M−1).19 With this reaction also
fulfilling an ideal catalytic scenario, we have probed how
many cycles are feasible.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. “Diels–Alderase” activity rationalized using substrate and TS stabilization effects. (a) DA activity correlates with TS mimic 6
binding affinities; stabilizing interactions with the endoꢀpyridyl groups of C-2 are key. These lone pairs also destabilize substrate binding
(n···π). (b) Similar acceleration with acyclic diene 2 to give weak binding adduct 5 shows that C-2 stabilizes an “early” TS. Substrate /
product / TS stabilization energies calculated from experimentally determined KA, kcat and kuncat values.
This was done using conditions that give negligible backꢀ
ground reaction (Figure 2b; 3 mM 1, 30 mM 2; blue squares)
so that the substrate consumption: C-2 loading ratio approxiꢀ
mates to turnover number (TON). Under these conditions, 10
mol% C-2 (0.3 mM) still gave full conversion in 48 hours
(Figure 2b, green triangles: TON ≈ 10). However, reducing
capsule concentration further gave an impracticably slow reacꢀ
tion. To avoid increasing the rate of the bulkꢀphase DA reacꢀ
tion, we therefore charged preꢀprepared 5 to the start of each
DA reaction to simulate low C-2 loading. Significantly, when
10 eq. and even 100 eq. of 5 were added, consumption rates of
1 still proved superior relative to the uncatalyzed process
(Figure 2b, dark green diamonds / circles). This latter experiꢀ
ment replicates the last 1% of a 0.1 mol% capsuleꢀmediated
reaction, demonstrating that TONs of 1000 are feasible. It is
also worth noting that 6 is likely the worseꢀcase scenario for
this catalytic model in terms of product inhibition (similar but
less pronounced effects are also seen with the other conꢀ
strained adduct 7) because of the structural similarity with the
TS. Even in this case, the 20 mol% catalyzed 1 + 3 → 6 still
goes smoothly to completion (Figure S1). It is likely that the
sudden drop in activity that has characterized many other capꢀ
suleꢀpromoted cycloadditions is avoided because of the comꢀ
bination of good activity20 and rapid substrateꢀproduct exꢀ
change. Indeed, simulation of this catalytic model at 20 mol%
Figure 5. (a) The catalytic model under study. (b) Simulation of
loading shows clearly that higher activity promotes more tolꢀ
the cage catalyzed Diels–Alder reaction as a function activity
erance towards higher binding products (Figure 5).
(kcat/kuncat = 10 (red); 100 (turquoise); 750 (purple); 1500 (green))
and product binding strength (KP Ass =100, 103, 5×103 and 104 M–1.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. kcat/kuncat and KA TS values for a series of DA reactions where either the diene or dienophile is systematically varied. (a)
The reduction in activity for bulky dienes 9ꢀ11 provides clear evidence for the necessary ingress of this component. (bꢀc) With
small, acyclic dienes goodꢀexcellent activity is observed for a range of quinone dienophiles. (d) With increasing quinone size, reꢀ
duced activity (kcat/kuncat) does not correlate with poor TS stabilization, rather stronger substrate binding. KA TS calculated from exꢀ
perimental KA, kcat and kuncat values.
The influence the microenvironment has upon reactivity
became apparent when the substrate scope was expanded
(Figure 4). Initially focusing on diene variation (Figure 4a),
comparing 2 to isomeric 1,3ꢀpentadiene, 8, a moreꢀthan douꢀ
bling in kcat/kuncat to almost 103 was observed, a significant
difference for such a subtle substrate change. To demonstrate
the necessity of diene ingress, larger substrates were also
screened. We were initially surprised to see that some activity
is retained with dienes 9 and 10, however, the complete lack
of acceleration with reactive 11 provides strong evidence that
catalytic action involves cycloaddition of the cavityꢀbound
quinone. Turning to dienophile variation, isomeric dienes 2
and 8 were selected for further exploration (Figures 4b–c).
These experiments revealed that catalysis is tolerant to modest
increases in quinone size, wherein a single small substituent
(e.g. 12 and 13) makes little difference and even with naphꢀ
thoquinone, 14, reasonable activity is retained. These experiꢀ
ments also showed that 8 is a universally better diene with
several dienophiles, the reactions with 1, 12 and 13 all giving
excellent kcat/kuncat, with values ranging from 750ꢀ1400. It is
worth noting that the smallest acceleration of these three reacꢀ
tions (8 + 12) counterintuitively possess the highest TS stabiliꢀ
zation, with a KA TS of 3.2 × 106 M−1. Rather, lower acceleraꢀ
tion is due to the greater affinity of substrate 12 (KA = 4300
M–1), which offsets the better TS stabilization. Similar effects
are observed when the size of quinone is increased further
(Figure 4d). For instance, 2,5ꢀdimethylꢀbenzoquinone, 15,
exhibits a much more modest kcat/kuncat of 14. While it could
have been assumed that this poorer catalytic effect was due to
the restricted environment disfavoring tertiary bond formation,
this is in fact not the case. Rather, the TS stabilization is greatꢀ
er than that afforded to the reaction of unsubstituted 1 with the
same diene (i.e., 1 + 3 → 6). Instead, the relatively poor
kcat/kuncat is a consequence of the unusually high quinone KA
value of 15 (43000 M−1), further exemplifying the subtle interꢀ
play between reactant and TS stabilization effects.
ence between TS and reactant stabilization energies (ꢁꢁG).21
By this measure, the catalytic antibody 1E9 that catalyzes the
reaction of maleimide and thiophene dioxide substrates is conꢀ
sidered one of the most efficient nonꢀcovalent DA catalysts,
with ꢁꢁG = −4.0 kcal mol−1. Significantly, the ꢁꢁG for the
reaction of 8 and 13 catalyzed by C-2 is −4.3 kcal mol−1, with
several other examples possessing values of –3.5 kcal mol−1 or
better. Careful consideration should be applied when comparꢀ
ing such different systems, from solvent effects through to the
disparity between catalysts that preꢀbind one or both DA reacꢀ
tions partners. For example, the greater reduction in ꢁG‡ in
cat
the case of C-2 can be partly explained by a catalytic model
that only involves dienophile stabilization, unlike 1E9 that
binds and therefore lowers the energy of both reactants. Conꢀ
versely, the entropic contribution to ꢁG‡ for C-2 will be
cat
greater than with systems that bind both diene and dienophile.
Nonetheless, it is clear that C-2 is very proficient at catalyzing
the DA reactions of quinones with small dienes. We have also
measured the thermodynamic and kinetic parameters associatꢀ
ed with DA catalysis by a prototypical Hꢀbond catalyst, the
trifluoromethylaryl substituted thiourea first reported by
Schreiner (see SI).16 This is a useful comparison because of
the similarity in the bimolecular catalytic step (kcat). The data
we have collected for the thioureaꢀcatalyzed DA reaction of
methyl vinyl ketone (KA = 12 M–1) with 3 shows a relatively
modest kcat/kuncat of 41. As could be expected from the signifiꢀ
cantly fewer nonꢀcovalent interactions involved, the stabilizaꢀ
tion afforded both the reactant and the TS state (−1.5 and −3.6
kcal mol−1, respectively) are much weaker in the case of the
small molecule Hꢀbond donor system. Consequently, such
catalysts often only function at relatively high substrate conꢀ
centrations (typically 100 mM dienophile, with excess diene).
In contrast, C-2 accelerates DA reactions under much more
dilute conditions (2.5 mM dienophile), also minimizing conꢀ
tributions from any background process.
While Sanders6 and Fujita10b have shown that synthetic hosts
can affect DA exo/endoꢀ and regioꢀselectivity, respectively,
neither example exhibit turnover. We were therefore keen to
demonstrate specificity bias under subꢀstoichiometric condiꢀ
tions. Where diasteroselectivity can be readily determined
It has been revealing to compare the efficiency of C-2 to
other nonꢀcovalent DA catalysts. While our kcat/kuncat values are
not directly comparable to many other “artificial Diels–
Alderases” because of the discrepancy in units, a more univerꢀ
sal analysis has been described by Houk that uses the differꢀ
ACS Paragon Plus Environment

Page 5 of 8
Journal of the American Chemical Society
lyzed and catalyzed reactions are 100% endo selective.22
(i.e., all reactions of dienes 3, 4 and 8), in all cases the uncataꢀ
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Capsule catalyzed modulation of selectivity. Product distributions for catalyzed (20 mol%) and uncatalyzed reactions determined
by 1H NMR integration. (aꢀb) Regiochemical switching of remote substituents (Cl, Me). (c) C-2 completely alters the chemoselectivity of
bulky diene 11 and the small but intrinsically less reactive 8. (d) While smallꢀmolecule catalysis (BF3·OEt2) promotes conventional selecꢀ
tivity of most reactive substrates (1 and 11), C-2 enhances a different reaction. Note: 21 is not a guest for C-2. Excess equimolar dienes
used for C-2 catalyzed / uncatalyzed chemoselective reactions to drive to completion with respect to quinone starting materials. *Solvent
spinning sideband.
It is perhaps not unsurprising that C-2 does not perturb the
intrinsic endo preference, as the exo TS is less compact and
therefore likely to be less well accommodated by the cavity of
C-2. However, while C-2 does not modulate diasteroselectiviꢀ
ty with these substrates, we do observe that it can alter regioꢀ,
siteꢀ and chemoꢀselectivity under true catalytic conditions (20
mol%). With the reaction of 13 and 8, the uncatalyzed reaction
produces isomers 17a-c (Figure 6a), however, 17c is comꢀ
pletely suppressed in the presence of C-2, possibly due to the
steric confinement preventing CꢀC bond formation between
adjacent methyl and chloroꢀsubstituents. The loss of 17c is
also accompanied by an increase in the proportion of the other
less favored species, 17b. While the precise reasons for this
switch are probably complex and multiꢀtiered, the trend apꢀ
pears to be that C-2 favors an increased proportion of products
with more “transoid” arrangement of substituents. This is
even more noticeable with the reaction of 2 and the same quiꢀ
none 13, which produces a larger regioselective switch under
catalytic conditions (Figure 6b). Here, the intrinsically less
favored “para” isomer 18b is favored by 3:1 over “meta” 18a
in the presence of C-2. This is a significant change considerꢀ
ing the remoteness of the substituents that define this regioꢀ
chemistry. Finally, we also envisaged that the microenvironꢀ
ment could be used to catalytically influence chemoselectivity.
When reactive but bulky diene 11 is competed with smaller
but less reactive 8 for cycloaddition with 1 under catalystꢀfree
conditions, almost complete selectivity (>95%) for anthracene
adduct 19 (Figure 6c) was observed. However, 20 mol% C-2
gives almost complete reversal of this selectivity, producing
20 in 90% yield. This is not solely a result of C-2 protecting
the dienophile from reaction with the most reactive diene,23
but additionally accelerating the lessꢀfavored bulkꢀphase proꢀ
cess thus showing that the capsule can reverse the intrinsic
reactivity of different dienes. We can also add another level of
complexity to this reaction by including a second dienophile,
tꢀbutylbenzoquinone 21 (Figure 6d), which is not a guest for
C-2. When these substrates are reacted in a stoichiometric
ratio under nonꢀcatalytic conditions, this bulkꢀphase process is
very slow. However, in the presence of a conventional Lewisꢀ
acid catalyst, BF3·OEt2, all four possible products are obꢀ
served, with a significant excess of 19 and 22. This bias is a
consequence of preferred cycloaddition between most reactive
diene and dienophile pair (11 and 1), which then leaves a
higher concentration of least reactive substrates (21 and 8).
When BF3·OEt2 is changed for 20 mol% C-2, the intrinsically
lessꢀfavored 20 and 23 are produced with greater selectivity
(>90%), showing the capsule is able to selectively accelerate
one of the least favored out of the four possible reactions. The
amplification of a single reaction from a collection of subꢀ
strates that otherwise show similar reactivity is a frequent
hallmark of biological processes, from biosynthesis though to
signaling mechanisms, and almost impossible to achieve using
conventional catalytic approaches. The credentials of this curꢀ
rent system thus pave the way to various opportunities, from
complex synthetic cascades though to molecular sensing.
Conclusions
While artificial receptors that coꢀbind diene and dienophile
substrates frequently show increased DA reactivity, this “bioꢀ
inspired” approach is invariably hindered by severe product
inhibition. In contrast, we have shown that a simple quinoneꢀ
binding Pd2L4 capsule system can promote enclosed catalysis
with efficient turnover by negating the need to formally coꢀ
bind the diene component. Instead, increased reactivity is proꢀ
vided by a convergent array of weak Hꢀbond donor and accepꢀ
tor groups, which stabilize the cycloaddition pathway. The
influence the microenvironment has on catalysis is profound;
together, the collection of interactions lowers the free energy
of activation to an extent only previously observed in catalytic
antibodies, and more recently in naturally occurring intramoꢀ
ACS Paragon Plus Environment

Journal of the American Chemical Society
lecular Diels–Alderases.19 We have also shown that the enꢀ
closed environment is able to influence regioꢀ and chemoꢀ
selectivity in a way that simple small molecules systems could
not. We envisage that the transition from entropic to enthalpic
capsule catalytic models will be widely applicable to a range
of different reaction types.
Page 6 of 8
uncatalyzed reactions, quinone (20 ꢂL of a 62.5 mM stock
solution in CD2Cl2), and the internal standard
tetrakis(trimethylsilyl)silane (10 ꢂL of a 15.2 mM stock soluꢀ
tion in CD2Cl2). For the reactions with the competitive inhibiꢀ
tor, solid anthraquinone S13 was added to the NMR tube (5.2
mg, 25 ꢂmol). The Diels–Alder reaction was then started by
the addition of the corresponding diene (20 ꢂL of a stock soluꢀ
tion in CD2Cl2, 5–100 equivalents depending on diene reacꢀ
1
2
3
4
5
6
7
8
Experimental Methods
1
tivity). H NMR spectra were recorded at regular intervals
General. All reagents and solvents were purchased from
Alfa Aesar, VWR or Sigma Aldrich and used without further
purification, except 1, which was recrystallized form hot
CH2Cl2/pet ether 60ꢀ80 (1:5) and 14, which was recrystallized
form hot CH2Cl2/pet ether 60ꢀ80 (1:3). Cage C-1 was prepared
using the previously reported method.15b
until sufficient data was collected to determine the kinetic
parameters. In all cases the NMR reactions were kept at 298
K. Kinetic NMR data was processed using the MestreNova 11
software and the concentration of all chemical species were
determined for each reaction time. All reactions were perꢀ
formed at least twice and a representative example is reported
in the supporting information.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
All H, 13C and 19F NMR spectra were recorded on either a
500 MHz Bruker AV III equipped with a DCH cryoꢀprobe
(Ava500), a 500 MHz Bruker AV IIIHD equipped with a
Prodigy cryoꢀprobe (Pro500), a 600 MHz Bruker AV IIIHD
equipped with a TCI cryoꢀprobe (Ava600) or a 400 MHz
Bruker AV III equipped with BBFO+ probe (Ava400) at a
constant temperature of 300 K. Chemical shifts are reported in
parts per million. Coupling constants (J) are reported uncorꢀ
rected in hertz (Hz). Apparent multiplicities are reported using
the following standard abbreviations: m = multiplet, q = quarꢀ
tet, t = triplet, d = doublet, s = singlet, bs = broad singlet. All
analysis was performed with MestReNova, version 11. For full
assignment(s), see the Supporting Information. These were
made using a combination of COSY, NOESY, HSQC and
HMBC NMR spectra. MS was performed on a Synapt G2
(Waters, Manchester, UK) mass spectrometer, using a direct
infusion electrospray ionization source (ESI), controlled using
Masslynx v4.1 software.
Experimental kinetic constants for the catalyzed reaction
(kcat) were obtained by fitting to the simulated kinetic model
using the LevenbergꢀMarquardt Nonlinear LeastꢀSquares Alꢀ
gorithm25 implemented in the R software26 and the RStudio27
software interface. Fittings were carried out simultaneously to
total product and quinone concentration (due to fast exchange
of these guests on the NMR timescale) in order to both miniꢀ
mize the mathematical fitting errors and also ensure that the
data fit to the kinetic model. For more information see the
supporting Information. The association constants of the quiꢀ
nones were determined by 1H NMR titrations while the associꢀ
ation constants of the Diels–Alder adducts (KP Ass) were deterꢀ
mined either by similar titrations or by estimation from fitting
the kinetic data (see Supporting Information). Initial rates
(Vmax) of the catalyzed and uncatalyzed reactions were calcuꢀ
lated from the slope obtained by linear fitting to the initial data
points of the reaction ([quinone]total vs. time).The associated
Gibbs energy barrier (ꢁG‡) for each the kinetic constant (k)
and additional parameters determined from the kinetic conꢀ
stants were calculated as described in the supporting inforꢀ
mation.
Synthesis of C-2. The cage C-2·4OTf was prepared accordꢀ
ing to a literature method.24 To a solution of C-2·(OTf)4 (147
mg, 75.9 ꢂmol, 1 equiv) in CH2Cl2 (40 mL) was added NaBꢀ
ArF (269 mg, 307 ꢂmol, 4 equiv), which was then sonicated
for 5 min. The reaction mixture was filtered and the undisꢀ
solved material washed with further CH2Cl2 (150 mL). The
solvent was removed under reduced pressure to obtain C-2
(cage·(BArF)4) as an off white solid (333 mg, 69.4 ꢂmol, 91%
yield). 1H NMR (500 MHz, CD2Cl2): δ 8.88 (s, 8H), 8.57 (d, J
= 5.5 Hz, 8H), 8.08 (dt, J = 8.1, 1.5 Hz, 8H), 7.77–7.72 (m,
4H), 7.72 – 7.67 (m, 32H, HBArF), 7.58–7.52 (m, 12H), 7.52
ppm (s, 16H, HBArF). For a full assignment, see the SI. 13C
NMR (126 MHz, CD2Cl2): δ 162.29 (q, JC–F = 50.0 Hz, BArF),
152.74, 149.12, 145.52, 142.16, 138.34, 135.34 (m, BArF),
130.00–128.27 (m, BArF), 129.58, 129.13, 125.99, 125.10 (q,
JC–F = 272.5 Hz, BArF), 118.27–117.86 (m, BArF), 96.70,
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website. This includes synthetic details and
characterization, binding constant and kinetic experiments
AUTHOR INFORMATION
Corresponding Author
Paul.Lusby@ed.ac.uk
1
81.52 ppm. 19F NMR (471 MHz, CD2Cl2): δ –62.75 ppm. H
DOSY NMR (500 MHz, CD2Cl2): 4.32 × 10ꢀ10 m2/s, hydrodyꢀ
namic radius = 12.1 Å. ESI TOF HRMS m/z: Found
3922.3789 [M–BArF]+, calculated for [C172H80B3F72N12Pd2]+
3927.3928. Found 1532.1631 [M–2BArFꢀ]2+, calculated for
[C140H68B2F48N12Pd2]2+ 1532.1548. Found 733.7532 [M–
3BArF]3+, calculated for [C108H56BF24N12Pd2]3+ 733.7559.
Paul Lusby ORCID ID: https://orcid.org/0000ꢀ0001ꢀ8418ꢀ
5687
Vicente MartíꢀCentelles ORCID ID: https://orcid.org/0000ꢀ
0002ꢀ9142ꢀ9392
Found
[C76H44N12Pd2]4+ 334.2977.
Experimental determination of kinetic constants
334.2973
[M–4BArF]4+,
calculated
for
Andrew L. Lawrence ORCID ID: http://orcid.org/0000ꢀ
0002ꢀ9573ꢀ5637
All kinetic constants were determined the following general
procedure. To an NMR tube was introduced either a solution
containing the cage (C-1 or C-2) compound (450 ꢂL of a stock
0.56 mM solution in CD2Cl2) or just CD2Cl2 (450 ꢂL) for the
Notes
The authors declare no competing financial interests.
ACS Paragon Plus Environment

Page 7 of 8
Journal of the American Chemical Society
(17) For a Hꢀbond cavitand that activates partially encapsulated
maleimide dienophiles, see Hooley, R. J.; Rebek Jr., J. Org. Biomol.
Chem. 2007, 5, 3631.
(18) (a) Pauling, L. Nature 1948, 161, 707. For an example of reꢀ
cent supramolecular TS stabilization of DA reactions using anionꢀπ
interactions, see (b) Liu, L.; Cotelle, Y.; Bornhof, A.ꢀB.; Besnard, C.;
Sakai, N.; Matile, S. Angew. Chem. Int. Ed. 2017, 56, 13066.
ACKNOWLEDGMENT
1
2
3
4
This work was supported by the Leverhulme Trust (RPGꢀ
2015ꢀ232).
ABBREVIATIONS
5
6
7
8
DA, DielsꢀAlder; LUMO, lowest unoccupied molecular orbital;
NMR, Nuclear Magnetic Resonance; TON, turnover number, TS,
transition state.
(19) It is likely that similar product conformational changes allow
naturallyꢀoccurring Diels–Alderase enzymes to turnover. For a recent
example, see Byrne, M. J.; Lees, N. R.; Han, L.ꢀC.; van der
Kamp,M. W.; Mulholland, A. J.; Stach, J. E. M.; Willis, C. L.; Race,
P. R. J. Am. Chem. Soc. 2016, 138, 6095.
(20) Palma, A.; Artelsmair, M.; Wu, G.; Lu, X.; Barrow, S. J.; Udꢀ
din, N.; Rosta, E.; Masson, E.; Scherman, O. A. Angew. Chem. Int.
Ed. 2017, 56, 15688.
(21) Kim, S. P.; Leach, A. G.; Houk, K. N. J. Org. Chem. 2002,
67, 4250.
(22) In the case of the catalyzed reaction of 1 and 8, we observe the
formation of a single diastereomer (see SI). This strongly suggests
that C-2 is catalyzing a concerted [4+2] cycloaddition and not a stepꢀ
wise reaction.
(23) Smulders, M. M. J.; Nitschke, J. R. Chem. Sci. 2012, 3, 785.
(24) M. Lewis, J. E.; Crowley, J. D. Supramol. Chem. 2014, 26,
173–181.
(25) Elzhov, T. V.; Mullen, K. M.; Spiess, A.ꢀN.; Bolker, B.
minpack.lm: R Interface to the Levenberg-Marquardt Nonlinear
Least-Squares Algorithm Found in MINPACK, Plus Support for
9
REFERENCES
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(1) (a) Michael, J. W.; Ulmann, P. A.; Mirkin, C A. Angew. Chem.
Int. Ed. 2011, 50, 114. (b) Slagt, V. F.; Reek, J. N. H.; Kamer, P. C.
J.; van Leeuwen, P. W. N. M. Angew. Chem. Int. Ed. 2001, 40, 4271.
(c) Lee, S. J.; Cho, S.ꢀH.; Mulfort, K. L.; Tiede, D.M.; Hupp, J. T.;
Nguyen, S. T. J. Am. Chem. Soc. 2008, 130, 16828. (d) Kaphan, D.
M.; Levin, M. D.; Bergman, R. G.; Raymond, K. N.; Toste, F. D.
Science 2015, 350, 1235. (e) Zhang, Q.; Tiefenbacher, K. Nat. Chem.
2015, 7, 197. (f) Wang, Q.ꢀQ.; Gonell, S.; Leenders, S. H. A. M.;
Dürr, M.; IvanovićꢀBurmazović, I.; Reek, J. N. H. Nat. Chem. 2016,
8, 225. (g) Omagari, T.; Suzuki, A.; Akita, M.; Yoshizawa, M. J. Am.
Chem. Soc. 2016, 138, 499.
(2) (a) Yoshizawa, M.; Klosterman, J. K.; Fujita, M. Angew.
Chem. Int. Ed. 2009, 48, 3418. (b) Chakrabarty, R.; Mukherjee, P. S.;
Stang, P. J. Chem. Rev. 2011, 111, 6810. (c) Ward, M. D.; Raithby, P.
R. Chem. Soc. Rev. 2013, 42, 1619. (d) Cook, T. R.; Stang, P. J.
Chem. Rev. 2015, 115, 7001; (e) Brown, C. J.; Toste, F. D.; Bergman,
R. G.; Raymond, K. N., Chem. Rev., 2015, 115, 3012. (f) Zarra, S.;
Wood, D. M.; Roberts, D. A.; Nitschke, J. R. Chem. Soc.
Rev. 2015, 44, 419. (g) Otte, M. ACS Catalysis 2016, 6, 6491. (h)
Bloch, W. M.; Clever, G. H. Chem. Commun., 2017, 53, 8506.
(3) Rideout, D. C.; Breslow, R. J. Am. Chem. Soc. 1980, 102, 7817.
(4) Kelly, T. R.; Zhao, C.; Bridger, G. J. J. Am. Chem. Soc. 1989,
111, 3744.
Bounds.
R
package version 1.2ꢀ0, 2015. https://CRAN.Rꢀ
project.org/package=minpack.lm (accessed January 23, 2018).
(26) R Core Team. R: A language and environment for statistical
computing. R Foundation for Statistical Computing: Vienna, Austria,
2015. https://www.Rꢀproject.org/ (accessed January 23, 2018).
(27) Studio Team. RStudio: Integrated Development for R, version
0.99.892. RStudio, Inc.: Boston, MA, 2015. http://www.rstudio.com/
(accessed January 23, 2018).
(5) Tjivikua, T.; Ballester, P.; Rebek Jr., J. J. Am. Chem. Soc. 1990,
112, 1249.
(6) Waiter, C. J.; Anderson, H. L.; Sanders, J. K. M. J. Chem. Soc.,
Chem. Commun.1993, 458.
(7) (a) Kang, J.; Rebek Jr., J. Nature 1997, 385, 50.; (b) Chen, J.;
Rebek Jr, J. Org. Lett. 2002, 4, 327.
(8) (a) Yoshizawa, M.; Takeyama, Y.; Okano, T.; Fujita, M. J. Am.
Chem. Soc. 2003, 125, 3243; (b) Nishioka, Y.; Yamaguchi, T.; Yoshiꢀ
zawa, M.; Fujita, M. J. Am. Chem. Soc. 2007, 129, 7000. (c) Murase,
T.; Horiuchi, S.; Fujita, M. J. Am. Chem. Soc. 2010, 132, 2866.
(9) Parthasarathy, A.; Kaanumalle, L. S.; Ramamurthy, V. Org.
Lett. 2007, 9, 5059.
(10) (a) Kang, J.; Santamaria, J.; Hilmersson, G.; Rebek Jr.; J. Am.
Chem. Soc. 1998, 120, 7389. (b) Yoshizawa, M.; Tamura, M.; Fujita,
M. Science 2006, 312, 251.
(11) (a) Pluth, M. D.; Bergman, R. G.; Raymond, K. N. Angew.
Chem. Int. Ed. 2007, 46, 8587. (b) Bolliger, J. L.; Belenguer, A. M.;
Nitschke, J. R. Angew. Chem. Int. Ed. 2013, 52, 7958.
(12) (a) Salles, A. G.; Zarra, S.; Turner, R. M.; Nitschke, J. R. J.
Am. Chem. Soc. 2013, 135, 19143. (b) Cullen, W.; Misuraca, M. C.;
Hunter, C. A.; Williams, N. H.; M. D. Ward, Nat. Chem. 2016, 8, 231.
(13) (a) Fiedler, D.; Bergman, R. G.; Raymond, K. N. Angew.
Chem. Int. Ed. 2004, 43, 6748.
(14) (a) Pluth, M. D.; Bergman, R. G.; Raymond, K. N. Science
2007, 316, 85. (b) Hastings, C. J.; Pluth, M. D.; Bergman, R. G.;
Raymond, K. N. J. Am. Chem. Soc. 2010, 132, 6938. (c) Murase, T.;
Nishijima, Y.; Fujita, M. J. Am. Chem. Soc. 2011,134, 162.
(15) (a) Liao, P.; Langloss, B. W.; Johnson, A. M.; Knudsen, E. R.;
Tham, F. S.; Julian, R. R.; Hooley, R. J. Chem. Commun. 2010, 46,
4932. (b) August, D. P.; Nichol, G. S.; Lusby, P. J. Angew. Chem. Int.
Ed. 2016, 55, 15022. (c) Lewis, J. E. M.; Gavey, E. L.; Cameron, S.
A.; Crowley, J. D. Chem. Sci. 2012, 3, 778.
(16) Wittkopp, A.; Schreiner, P. R. Chem. Eur. J. 2003, 9, 407.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment