Large pKa shifts of α-carbon acids induced by copper(II) complexes

Article abstract of DOI:10.1002/chem.200400396

A series of synthetic receptors (4-6) incorporating metal ions, specifically copper(II), were examined for their ability to enhance the acidity of active methylene compounds. The copper(II) complexes were observed to reduce the pK_a of 1,3-diket

Full text of DOI:10.1002/chem.200400396

FULL PAPER
Large pKa Shifts of a-Carbon Acids Induced by Copper(ii) Complexes
Zhenlin Zhong, Brenda J. Postnikova, Robert E. Hanes, Vincent M. Lynch, and
Eric V. Anslyn*[a]
Abstract: A series of synthetic recep-
tors (4–6) incorporating metal ions,
specifically copper(ii), were examined
for their ability to enhance the acidity
of active methylene compounds. The
copper(ii) complexes were observed to
reduce the pKa of 1,3-diketone carbon
acids in acetonitrile by as much as 12
pKa units. The relatively large pKa re-
duction achieved by the complex is at-
tributed to the electrostatic interaction
between the anionic p system of the
enolate and the copper(ii) ions. The
cage structure and hydrogen bonding
sites in receptors 4 and 5 lead to a very
modest further enhancement of the
acidity relative to that with 6. This
study provides insight into the way in
which metalloenzymes stabilize an eno-
late intermediate.
Keywords: acidity · copper com-
plexes · enzyme models · N ligands ·
receptors
Introduction
not been able to account for the 10-unit or more pKa shifts
often observed in enzyme catalysis. For example, polyaza
cleft 1 was designed to form hydrogen bonds directed
toward the oxygen lone pairs of carbon acids (Scheme 1).[5]
The catalytic function of enzymes relies on the preferential
stabilization of a transition state compared to a substrate.[1]
Preferential stabilization of an enolate intermediate results
in the depression of the pKa of a carbon acid substrate, al-
lowing deprotonation by a relatively weak base in an
enzyme active site. For example, yeast enolase utilizes an
active site lysine (pKa ~10) to catalyze the dehydration of 2-
phospho-d-glycerate (PGA) to produce phosphoenolpyru-
vate by the abstraction of a proton a to a carboxylic acid
(pKa ~32) in PGA.[2] Several enzymatic studies have fo-
cused on determining the mechanism through which enolate
stabilization is achieved in enolases,[2] racemases,[3] and aldo-
lases.[4] These enzymes share a common feature: the use of a
transition-metal ion in the active site. It has been proposed
for each of these classes of enzymes that the metal ion plays
a pivotal role in stabilizing the negative charge on the transi-
tion state and intermediate enolate during substrate depro-
tonation.
Scheme 1. Synthetic receptors 1 and 2 and their idealized binding modes
for 1,3-cyclohexanedione and 2-acetylcyclopentanone, respectively.
The use of hydrogen bonds to stabilize enolate intermedi-
ates has been proposed, but to date, organic models have
Hydrogen bonding was proposed to reduce the pKa of active
methylene compounds by withdrawing electron density
away from the enolate intermediate. A modest reduction of
1.0 unit in the pKa of 1,3-cyclohexanedione was obtained
with receptor 1. Subsequently, it was proposed that hydro-
gen bond geometry is crucial to achieving a larger reduction
in the pKa of carbon acids, as supported by studies of 4-
chlorobenzoyl-Co-A-dehalogenase. A crystal structure of
this enzyme revealed hydrogen bonds oriented towards the
p system of a thioester carbonyl enolate.[6] Host 2 was de-
[a] Dr. Z. Zhong, Dr. B. J. Postnikova, Dr. R. E. Hanes, V. M. Lynch,
Prof. Dr. E. V. Anslyn
The Department of Chemistry and Biochemistry
The University of Texas at Austin
1 University Station A5300
Austin, TX, 78712 (USA)
Fax : (+1)512-471-7791
E-mail: anslyn@ccwf.cc.utexas.edu
Chem. Eur. J. 2005, 11, 2385 – 2394
DOI: 10.1002/chem.200400396
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2385

signed to bind the enolate of 2-
acetylcyclopentanone with
amide hydrogen bonds directed
towards the p system of the
enolate because this is where
the negative charge resides in
the enolate.[7] A pKa shift of 2.9
units was obtained with recep-
tor 2 and 2-acetylcyclopenta-
none. These results suggest that
hydrogen bond geometry plays
a significant role in enolate sta-
bilization. However, traditional
hydrogen bonds alone are insuf-
ficient to decrease the pKa of
Scheme 2. Copper (ii) complexes of synthetic receptors (4–7) used to study the effects of metal coordination
on the pKa of active methylene compounds.
carbon acids in enzyme active sites. The possibility of stabili-
zation by low barrier hydrogen bonds[8] or cooperative ef-
fects between hydrogen bonds and electrostatic effects with
metal ions cannot be ruled out.[3a]
ylene compounds in the presence of copper(ii) complexes of
these receptors. The enolates of active methylene com-
pounds were used in our studies in order to compare our re-
sults with those obtained using receptors 1 and 2. Upon
comparison of the pKa values of 2-acetylcyclopentanone in
the presence of receptors 4 and 5 and with our results with
receptors 1 and 2, it was concluded that the use of metal
ions greatly enhances the acidity of carbon acids, and that
the additional hydrogen bonding sites only lead to a modest
further enhancement.
Synthetic models have been designed to explain the pKa
reduction in enzymatic systems based on the use of metal
coordination. It has been established that aldol condensa-
tions are in part catalyzed in the active site of class II aldo-
lases by a zinc ion ligated by the imidazole groups of three
histidines.[9] In these enzymes, zinc functions as a Lewis acid,
activating methylene groups a to a carbonyl towards depro-
tonation by withdrawing negative charge from the carbonyl
oxygen in the transition state. Kimura et al. proposed a
model system for class II aldolases to study how much the
acidity of a hydrogen a to a carbonyl may be increased
through coordination of the zinc(ii) center.[10] The model
system, 4-bromophenacyl-pendant zinc cyclen (3), specifical-
Results and Discussion
Design criteria: Our receptor design is a bicyclic host that
may bind an anionic guest through charge-pairing with one
or two bound divalent metal centers and hydrogen bonding,
with the two binding interactions behaving cooperatively to
induce pKa shifts. As illustrated in Scheme 2, receptors 4
and 5 incorporate amide hydrogen bond donors, as modeled
from receptor 2. Whereas receptor 2 utilizes three pyridine-
2,6-dicarboxylic acid diamide moieties, receptor 4 utilizes
two pyridine-2,6-dicarboxylic acid diamide moieties for hy-
drogen bond formation and one 2,6-bis(aminomethyl)pyri-
dine moiety for binding a metal ion inside the macrocycle.
Receptor 5 utilizes one pyridine-2,6-dicarboxylic acid di-
amide moiety for hydrogen bond formation and two 2,6-bis-
(aminomethyl)pyridine moieties for binding metal ions. The
similar macrocyclic scaffolds of our receptors allow a direct
comparison with carbon acid pKa shift results obtained by
using receptor 2. Receptor 6 was designed to bind enolates
exclusively through metal coordination to examine the
effect of the cage structure in inducing pKa shifts in active
methylene enolates, and may also allow an evaluation of co-
operative effects of metal interactions and hydrogen bond-
ing in hosts 4 and 5.
ly sought to determine how the formation of the enediolate
by a relatively weak base (glutamic carboxylate) at physio-
logical pH is possible in the active site of class II aldolases.
A reduction in the pKa of the methylene hydrogen of 10
pKa units was found. The large pKa shift is explained by the
coordination of the carbonyl to the dicationic zinc center, as
3a was isolated from solution. Such large enhancement in
the pKa of a carbon acid suggests that metal binding may be
much more influential in reducing the pKa of carbon acids
in enzyme active sites than hydrogen bonding.
The 2,6-bis(aminomethyl)pyridine or similar triaza moiet-
ies are known to bind metal ions, such as copper(i), cop-
per(ii), and zinc(ii).[11] Copper(ii) was used in our host design
because it can adopt a trigonal-bipyramidal or octahedral
geometry with triaza ligands geometrically similar to those
used in receptors 4 and 5.[12,13] This kind of CuII complexes is
We have designed a series of synthetic receptors (4–6)
(Scheme 2) and examined the carbon acidity of active meth-
2386
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 2385 – 2394

Receptor Complexes
FULL PAPER
known to form chelates with oxygen-containing ligands such
as (benzyloxy)acetaldehyde.[13] Therefore, an enolate will be
bound in the equatorial position on the metal and hence lie
parallel to the host aromatic rings (Scheme 3). This orienta-
in 10 were coupled by reductive amination with 2,6-pyridi-
nedicarboxaldehyde giving the final cagelike receptor 4 in
80% yield.
Similarly, two equivalents of di-tBoc-protected 1,3,5-tris-
(aminomethyl)-2,4,6-triethylbenzene (11) were coupled by
reaction with 2,6-pyridine dicarbonyl dichloride. Removal of
the tBoc protecting groups with trifluoroacetic acid followed
by ion exchange chromatography yields cleft 12. The synthe-
sis of cagelike receptor 5 is completed by nucleophilic sub-
stitution of the primary amines of the cleft to two equiva-
lents of 2,6-pyridinedicarboxaldehyde followed by reductive
amination in which the imine groups are converted to secon-
dary amines. The reaction is carried out in toluene at high
dilution.
Scheme 3. Schematic representation of the (A) trigonal bipyramidal or
(B) octahedral CuII center in the possible structures of the receptor–eno-
late complexes.
The acyclic ligands 2,6-bis(n-butylaminomethyl)pyridine
(6) and 2,6-bis(benzylaminomethyl)pyridine (7) were syn-
thesized by the reductive amination of 2,6-pyridinedicarbox-
aldehyde with n-butylamine or benzylamine.
tion may position an enolate bound within the macrocycle
for simultaneous hydrogen bonding and additional metal co-
ordination.
Binding studies: To initiate our studies with receptors 4–6, it
was necessary to determine the binding stoichiometry be-
tween the receptors and metal ions. Copper(ii) has been
used in similar macrocyclic receptors and the binding stoi-
chiometry determined by spectroscopic methods.[11] The
binding stoichiometry of receptors 4–6 with copper(ii) salts
was determined by the mole ratio method,[15] by monitoring
the absorbance increase of the ligand–copper(ii) complexes
by spectrophotometric titrations (Figure 1). The titration
curves clearly reveal that receptors 4 and 6 form 1:1 com-
plexes with Cu(OTf)2, whereas receptor 5 forms a 1:2 com-
plex with CuCl2. The binding stoichiometry is in accordance
with the number of 2,6-bis(aminomethyl)pyridine chelating
moieties in the receptors. The sharp turning points on the ti-
Synthesis: The syntheses of receptors 4 and 5 may be ac-
complished by joining two hexasubstituted benzene rings,
1,3,5-tris(aminomethyl)-2,4,6-triethylbenzene with 2,6-disub-
stituted pyridine linkers (Scheme 4). A hexasubstituted ben-
zene ring is used to arrange the functional groups on the 1,
3, and 5 positions to one face of the benzene ring.[14]
Mono-tBoc-protected
1,3,5-tris(aminomethyl)-2,4,6-tri-
ethylbenzene (8) was allowed to react with 1,6-pyridinedi-
carbonyl dichloride giving a double linked cleft 9. Removal
of the tBoc protecting groups in 9 with trifluoroacetic acid
in dichloromethane afforded 10. The two free amino groups
Scheme 4. Synthesis of receptors 4 and 5.
Chem. Eur. J. 2005, 11, 2385 – 2394
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2387

E. V. Anslyn et al.
Figure 1. Titrations of free hosts with Cu(OTf)2 in acetonitrile: (ꢁ) ab-
*
sorbance at 640 nm of 1.5 mm 5; ( ) absorbance at 570 nm of 2.0 mm 4;
(+) absorbance at 620 nm of 3.0 mm 6.
tration curves suggest high association constants of the re-
ceptors with copper(ii) salts. Limited by the sensitivity of ab-
sorbance spectral method and the formation of two-ligand–
one-copper complexes when less than one equivalent of cop-
per(ii) was presented, no binding constant was obtained
from the titration curves. Based on these stoichiometry stud-
ies, the copper(ii) complexes of 4–6 were made in situ by
mixing a receptor solution with one or two equivalents of a
copper(ii) salt.
The enolates of active methylene compounds 2-acetyl-
cyclopentanone (13) and 1,3-cyclopentane dione (14) were
selected for binding studies with receptors 4, 5, and 6 in
order to make a direct comparison with our previous results
for receptors 1 and 2.[5,7] In our previous studies, the active
methylene compounds were selected due to their increased
functionality, and therefore an increased number of poten-
tial binding sites, compared with monocarbonyl compounds.
Additionally, active methylene compounds have lower pKa
values than monocarbonyl compounds.
Figure 2. Absorption spectral change of a 2.0 mm solution of 4-Cu(OTf)2
(a) or 6-Cu(OTf)2 (b) in acetonitrile upon addition of up to one equiva-
lent of 13-Na+·[15]crown-5. For clarity, spectra of adding more than one
equivalent of enolate 13 were not shown because of the formation of a
1:2 complex.
The binding mode of enolates with the copper complexes
of receptors 4, 5, and 6 was studied with UV/Vis spectro-
scopic titrations (Figure 2, Figure 3, Figure 4). Although an
aqueous medium is optimal for the titrations, acetonitrile is
also an acceptable solvent because its dielectric constant
(37.5) is proposed to mimic the interior of an enzyme active
site.[5] Additionally, a direct comparison may be made with
the results of pKa shift experiments with receptors 1 and 2,
also performed in acetonitrile. UV/Vis spectral change in
the titrations of copper(ii) complexes of 4 and 6 with sodium
enolate 13 are shown in Figure 2 and Figure 3. The titration
curves in Figure 3 reveal that the copper(ii) centers in the
complexes can bind one or two molecules of 13 dependent
on the amount of the enolate added. The receptor ligands 4
or 6 did not dissociate from the copper(ii) center by addition
of the enolate, revealed by a control titration of Cu(OTf)2
with the enolate in the absence of the receptors that showed
completely different absorbance spectra from those titra-
tions in the presence of the receptors. The clear isosbestic
points in Figure 2 and sharp turning points at one equivalent
in Figure 3 suggest that the first binding constant (K1) is
Figure 3. Absorbance at 760 nm of a 2.0 mm solution of 4-Cu(OTf)2
(open circle) or 6-Cu(OTf)2 (closed circle) in acetonitrile upon addition
of 13-Na+·[15]crown-5.
much greater than the second binding constant (K2). There-
fore, when only one equivalent or less of the enolate is pres-
ent, the systems can be treated simply as 1:1 complexes. The
binding mode of the enolate of 1,3-cyclopentane dione (14)
with the copper complexes of 4 and 6 are very similar to
that of 13.
A 2:1 binding stoichiometry was found from a decrease in
absorbance of the binuclear copper(ii) complex of receptor
5 at 460 nm upon titration of sodium enolate 13 in aceto-
nitrile, as shown in Figure 4.
2388
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 2385 – 2394

Receptor Complexes
FULL PAPER
Efforts to make crystals of receptor–copper(ii)–enolate
complexes were not successful.
Deprotonation studies: A key step in enzyme catalysis with
an enolate intermediate is the deprotonation of carbon acid
substrates using peptide bases with much lower conjugate
acid pKa values than the substrate. It was shown in experi-
ments with receptors 1 and 2 that bases with lower conju-
gate acid pKa values than the active methylene compounds
1,3-cyclohexanedione and 2-acetylcyclopentanone, respec-
tively, could be used to deprotonate the carbon acids in the
presence of the receptors and induce complex formation.[5,7]
The stabilization of the enolate of the carbon acids results in
the observed pKa shifts in acetonitrile.
Figure 4. Titration curve for the 5-CuCl2 (2.09 mm) and 13- Na+
·[2.2.1]cryptand in acetonitrile.
UV spectroscopy was used with receptors 4, 5, and 6 to
determine the extent of deprotonation of 2-acetylcyclopen-
tanone and 1,3-cyclopentanedione in the presence of system-
atically selected bases. The strategy for selective deprotona-
tion of the parent active methylene compound in the pres-
ence of the receptor is to optimize the base such that the
pKa of the conjugate acid of the base is lower than that of
the carbon acid, but close to the pKa of the carbon acid re-
ceptor complex. Sterically hindered amine or pyridine bases,
whose conjugate acid pKa values in acetonitrile have been
previously reported, were used in our studies to induce com-
plex formation between our receptors and the enolates 13
and 14. Table 1 lists the enolates and organic bases (15–18)
Crystal structure: Extensive efforts were made trying to pre-
pare crystals of the copper(ii) complexes of 4–6 by slow
evaporation of their methanol, water, or dichloromethane
solutions. However, only extremely small needles that were
not suitable for X-ray analysis were obtained. To solve this
problem, 2,6-bis(benzylaminomethyl)pyridine (7) was syn-
thesized, because it is structurally more rigid than its butyl
analogue 6. Slow evaporation of a dichloromethane solution
of 7 and Cu(OTf)2 in 1:1 molar ratio gave the complex as
blue needles. X-ray analysis showed that the complex is 7-
Cu(OTf)2-H2O (Figure 5). The geometry of the copper(ii)
Table 1. pKa values of the conjugate acids of enolates and organic bases.
Compound
Number
pKa in CH3CN
pKa in H2O
13
25.4[7]
7.8[17]
14
15
19.2[5]
18.2[5]
4.5[18]
11.25[19]
16
17
14.2[20]
13.4[5]
6.81[21]
7.03[22]
Figure 5. Molecular structure of 7-Cu(OTf)2-H2O (ORTEP plot; 50%
probability ). Most hydrogen atoms have been removed for clarity.
Dashed lines are indicative of H-bonding interactions. Selected inter-
18
–
4.95[23]
ꢀ
ꢀ
ꢀ
atomic distances [ꢂ]: Cu1 N1 1.927(3), Cu1 N8 2.062(4), Cu1 N17
ꢀ
ꢀ
ꢀ
2.062(4), Cu1 O1A 2.479(4), Cu1 O1B 2.352(4), Cu1 O1W 1.945(3),
ꢀ
ꢀ
N8 O2A 3.065(6), N17 O3B 2.996(5).
used in the titrations with their conjugate acid pKa values in
both water and acetonitrile.
center is octahedral. The three nitrogen atoms of 7 and an
oxygen atom of water coordinate to copper are in an ap-
proximate plain. The two triflate anions make the other two
coordination bonds sitting on the axe. Each triflate forms a
Figure 6 illustrates a UV/Vis titration study of 2-acetyl-
cyclopentanone with organic bases in the presence of recep-
tor 6-Cu(OTf)2. Titration of a 1:1 mixture of 6-Cu(OTf)2
and 2-acetylcyclopentanone with 1,2,2,5,5-pentamethylpiper-
idine (15) resulted in a spectral change (Figure 6a) almost
identical to that of the titration of 6-Cu(OTf)2 with enolate
ꢀ
hydrogen-bond with the NH group with N H···O distance
of 3.00 to 3.07 ꢂ.
Chem. Eur. J. 2005, 11, 2385 – 2394
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2389

E. V. Anslyn et al.
12.7 units on 2-acetylcyclopen-
tanone, whereas the CuCl2 com-
plexes induce 7.0–7.6 unit shifts
on the same acid. The pKa shift
of
1,3-cyclopentanedione
(5.7 units) induced by the
Cu(OTf)2 complexes of 4 and 6
is about 7 units smaller than
that of 2-acetylcyclopentanone.
Comparable or slightly larger
pKa shifts were found for the
copper(ii) complexes of the
cagelike molecule 4 relative to
the acyclic simple molecule 6.
Binuclear complex 5–2(CuCl2)
induces a significantly bigger
pKa shift than mononuclear
complex 4-CuCl2 (10.6 versus
7.6) on 2-acetylcyclopentanone.
Hence, the two metals in 5 are
cooperative. The shifts of 5.7–
12.7 units listed in Table 2 are
comparable with the 10 unit
pKa shift found by Kimura
et al. with receptor 3, which
similarly utilizes a metal cation
to activate a carbonyl and sub-
sequently reduce the pKa of the
a methylene proton.[10]
Figure 6. UV/Vis spectra of 2.0 mm 6-Cu(OTf)2 in the presence of one equivalent of 2-acetylcyclopentanone in
acetonitrile upon titration of (a) 15, (b) 17, and (c) 18. Plot d shows the absorbance changes at 760 nm of the
~
*
*
titrations by 15 ( ), 17 ( ), and 18 ( ).
13 (Figure 3), revealing that one equivalent of 15 can nearly
quantitatively deprotonate 2-acetylcyclopentanone in the
presence of the receptor complex. Titrations with weaker
bases 2-dimethylaminopyridine (17) (Figure 6b) and 2,6-di-
tert-butylpyridine (18) (Figure 6c) also caused similar but
smaller spectral changes.
Table 2. pKa shifts of 2-acetylcyclopentone (ACP) and 1,3-cyclopentane-
dione (CP) induced by CuII complex receptors in acetonitrile at 258C.
Carbon acid Receptor
Base
(2.0 mm) ratio[a] [%]
Deprotonation pKa pKa shift[b]
(2.0 mm)
(2.0 mm)
ACP
ACP
ACP
ACP[c]
ACP
CP
4-Cu(OTf)2 17
6-Cu(OTf)2 17
68
61
61
39
45
47
48
12.7 12.7
13.0 12.4
Figure 6d further illustrates that one equivalent of 2-di-
methylaminopyridine (17) with a conjugate acid pKa of 13.4
in acetonitrile is capable of deprotonating 61% of 2-acetyl-
cyclopentanone (pKa of 25.4 in acetonitrile) to form an eno-
late complex with receptor 6-Cu(OTf)2 under the titration
conditions. The observed pKa of the substrate carbon acid
(SH) was calculated from the deprotonation ratio and the
4-CuCl2
5-CuCl2
6-CuCl2
15
17.8
7.6
[d]
16[e]
15
14.8 10.6
18.4
13.5
13.5
7.0
5.7
5.7
4-Cu(OTf)2 17
6-Cu(OTf)2 17
CP
[a] The ratio of enolate complex concentration related to the initial con-
centration of the ketone ACP or CP. [b] The pKa difference of the carbon
acid in the absence or presence of the CuII receptor. [c] 2.14 mm.
[d] 0.97 mm. [e] 2.91 mm.
BH
pKa of the base (BH+) to be 13.0 using Equation (1) (for
details, see Experimental Section).
pKa ¼ pKa ꢀlogfð½Sꢀꢁ½BHþꢁÞ=ð½SHꢁ½BꢁÞg
BH
ð1Þ
The above trends shown in Table 2 can be explained
based on the thermodynamic cycle of receptor(R)-substra-
te(SH) binding illustrated in Figure 7 in which two routes to
the bound substrate anion are shown. The first path, K1–K2,
represents deprotonation of the substrate before binding to
the host. In the second path, K4–K3, host–substrate binding
occurs prior to deprotonation. Equilibrium constant K1 de-
notes the acid dissociation constant of the free carbon acid,
whereas K3 signifies that of the substrate bound by the re-
ceptor. The ratio of the acid dissociation constants, K3/K1, is
equal to the ratio of the binding constants of the substrate
The pKa of 2-acetylcyclopentanone is therefore shifted by
12.4 by the presence of 6-Cu(OTf)2, as its normal pKa is 25.4
in acetonitrile.
Similarly, the deprotonation studies of 2-acetylcyclopenta-
none and 1,3-cyclopentanedione in the presence of our
other copper(ii) complexes of 4 and 5 were performed and
the results are listed in Table 2. The pKa shifts induced by
the complex receptors are related to the counterions and
the structures of the ligands and carbon acids. The Cu(OTf)2
complexes of 4 and 6 induce a pKa shift of about 12.4–
2390
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 2385 – 2394

Receptor Complexes
FULL PAPER
value achieved by the complex is attributed to the electro-
static or coordination interaction between the anionic p
system of the enolate and the copper(ii) center. As in the
active site of class II aldolases, electrostatic induction is
more effective in stabilizing highly reactive intermediates
and inducing pKa shifts of carbon acids than traditional hy-
drogen bonds.
Figure 7. Thermodynamic cycle for substrate (SH) association and sub-
strate deprotonation in the presence of a receptor (R).
Experimental Section
and enolate with the receptor, K2/K4. The binding constants
(K4) of 2-acetylcyclopentanone with receptors in acetonitrile
were measured by UV/Vis titrations to be 87(ꢂ23)mꢀ1 with
4-Cu(OTf)2, 19(ꢂ9)mꢀ1 with 6-Cu(OTf)2, 36(ꢂ12)mꢀ1 with
4-CuCl2, and <10mꢀ1 with 6-CuCl2. The relatively narrow
range of K4 values suggests that the pKa shifts are mainly
determined by the K2 values which are the binding constants
of the enolates with the receptors. Therefore, the large shift
in the pKa of the carbon acids in the presence of our cop-
per(ii) host complexes is attributed to strong charge-pairing
or coordination interaction between the enolate anion and
metal center. The enolate–metal interaction is in competi-
tion with the association of the counterion with the metal
complex. Because chloride has a much stronger interaction
with copper(ii) than triflate, the pKa shifts induced by the
CuCl2 complexes of 4 and 6 are around five units lower than
those induced by the Cu(OTf)2 complexes (Table 2).
General considerations: 1H and 13C NMR spectra were recorded on a
Varian Mercury 400 or Varian Unity Plus 300 spectrometer. UV/Vis
measurements were performed on a Beckman DU-70 UV/Vis spectrome-
ter. Low-resolution and high-resolution mass spectra were measured with
Finnigan TSQ70 and VG Analytical ZAB2-E instruments, respectively.
1-{[(1,1-Dimethylethoxy)carbonyl]aminomethyl}-3,5-bis(aminomethyl)-
2,4,6-triethylbenzene (8) and 1,3-bis{[(1,1-Dimethylethoxy)carbonyl]ami-
nomethyl}-5-aminomethyl-2,4,6-triethylbenzene (11) were synthesized ac-
cording to reference [16]. All other chemicals were purchased from Al-
drich or Acros and used without further purification.
1,6-Bis(n-butylaminomethyl)pyridine (6): A mixture of 2,6-pyridinedicar-
boxaldehyde (117 mg, 0.867 mmol), n-butylamine (0.188 mL, 1.91 mmol),
and 3 ꢂ molecular sieves (ca. 20 beads) in anhydrous methanol (4 mL)
was stirred at room temperature for 3 h under argon. NaBH4 (150 mg,
4.0 mmol) was added. The reaction mixture was stirred for 4 h and fil-
tered through celite. The filtrate was concentrated by evaporation and
acidified by adding 1n HCl to decompose the excess borohydride. The
mixture was adjusted to strongly basic by adding 10% aqueous NaOH
solution (ca. 5 mL), and then extracted with dichloromethane (3ꢁ
10 mL). The organic layers were combined, washed once with water
(5 mL), dried over MgSO4, evaporated under reduced pressure, and dried
in vacuum to give a pale yellow oil (0.210 g) in 97% yield. 1H NMR
(400 MHz, CDCl3, TMS): d=7.58 (t, 3J(H,H)=7.6 Hz, 1H), 7.16 (d,
3J(H,H)=7.6 Hz, 2H), 3.88 (s, 4H), 2.65 (t, 3J(H,H)=7.2 Hz, 4H), 1.93
(br s, 2H), 1.56–1.48 (m, 4H), 1.36 (p, 3J(H,H)=7.3 Hz, 4H), 0.92 ppm
(t, 3J(H,H)=7.4 Hz, 6H); 13C NMR (100 MHz, CDCl3): d=155.38,
132.67, 116.26, 51.24, 45.34, 28.26, 16.43, 9.96 ppm; HRMS (CI): m/z:
250.22807 (calcd for C15H28N3 ([M+H+]): 250.22832).
The very similar pKa shifts induced by the copper(ii) com-
plexes of the cagelike molecule 4 and the acyclic simple
ꢀ
molecule 6 suggest that O Cu coordination bonds are domi-
nant binding forces between the diketone enolates and the
receptors. No significant cooperative hydrogen-bonding or
solvophobic interaction attributed to the amide linkages and
cagelike cavity in 4 were observed under the experimental
conditions in acetonitrile. Enolates of 1,3-diketones are
well-known ligands for copper(ii). Enolate of 2-acetylcyclo-
pentanone can form a six-membered chelating ring with a
metal center leading to strong binding. The binding of the
enolate of 1,3-cyclopentanedione, however, is much weaker
because the orientation of the two oxygen atoms fixed on
the ring is not favorable for chelation. Therefore, the pKa
shifts of 1,3-cyclopentanedione induced by the copper(ii)
complex receptors are much smaller than that of 2-acetylcy-
clopentanone (Table 2). Because the binding of enolates
with receptors is in competition with the counterions of cop-
per(ii), the receptors with noncoordinating counterions such
as triflate can bind an enolate stronger and therefore induce
a bigger pKa shift than those with coordinating counterions
such as chloride.
1,6-Bis(benzylaminomethyl)pyridine (7): Compound 7 was synthesized by
using the same procedure as for 6 from 2,6-pyridinedicarboxaldehyde
(126 mg, 0.93 mmol) and benzylamine (200 mg, 1.87 mmol) as a pale
1
yellow oil (288 mg) in 98% yield. H NMR (250 MHz, CDCl3, TMS): d=
7.57 (t, 3J(H,H)=7.6 Hz, 1H), 7.36–7.23 (m, 10H), 7.16 (d, 3J(H,H)=
7.6 Hz, 2H), 3.90 (s, 4H), 3.83 (s, 4H), 2.21 ppm (br s, 2H); 13C NMR
(62.5 MHz, CDCl3): d=159.12, 140.10, 136.65, 128.26, 128.11, 126.82,
120.33, 54.36, 53.42 ppm; HRMS (CI): m/z: 318.19699 (Calcd for
C21H24N3 ([M+H+]): 318.19702).
7,23-Bis(N’-tBoc-aminomethyl)-6,8,22,24,34,36-hexaethyl-
3,11,19,27,33,35-hexaazopentacyclyo[27.3.1.15,9.113,17.121,25]hexatriaconta-
1(33),5,7,9(36),13,15,17(35),21,23,25(34),29,31-dodecaene-2,12,18,28-te-
trone (9): A solution of 1-{[(1,1-dimethylethoxy)carbonyl]aminomethyl}-
3,5-bis(aminomethyl)-2,4,6-triethylbenzene (8) (0.647 g, 1.85 mmol) and
triethylamine (2.6 mL, 19 mmol) in dry dichloromethane (150 mL) was
cooled with an ice-water bath under argon. With stirring, 1,6-pyridinedi-
carbonyl dichloride (0.378 g, 1.85 mmol) was added in one portion as a
solid. The reaction solution was stirred at 08C for 30 min and then at
room temperature for 2 h. The reaction mixture was concentrated by
evaporation under reduced pressure to about 5 mL. Ethyl acetate
(10 mL) was added to facilitate the precipitation of triethylamine hydro-
gen chloride. The precipitate was removed by filtration. The filtrate was
concentrated by evaporation and subjected to column chromatography
(silica gel, CH2Cl2/acetone 100:5 to 100:30 v/v) to give an off-white solid
in 46% yield (0.410 g). Rf =0.6 (silica gel, CH2Cl2/acetone 4:1 v/v);
1H NMR (400 MHz, CDCl3, TMS): d=8.45 (d, 3J(H,H)=8.0 Hz, 4H),
8.07 (t, 3J(H,H)=7.6 Hz, 2H), 7.22 (br s, 4H), 4.68 (d, 3J(H,H)=4.4 Hz,
Conclusion
The use of receptors 4, 5, and 6 incorporating metal ions,
specifically copper(ii), has been shown to reduce the pKa
value of 1,3-diketone carbon acids in acetonitrile by as
much as 12 pKa units. The relatively large reduction in pKa
Chem. Eur. J. 2005, 11, 2385 – 2394
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2391

E. V. Anslyn et al.
8H), 4.42 (br s, 6H), 2.75 (q, 3J(H,H)=7.5 Hz, 8H), 2.66 (q, 3J(H,H)=
7.5 Hz, 4H), 1.44 (s, 18H), 1.24–1.16 ppm (m, 18H); 13C NMR (100 MHz,
CDCl3): d=163.41, 155.25, 148.66, 144.84, 144.57, 138.98, 131.18, 125.77,
79.79, 38.39, 28.37, 23.47,22.94, 16.75, 16.05 ppm; HRMS (CI): m/z:
961.55559 (calcd for C54H73N8O8 ([M+H+]): 961.55514).
solid (3.9 g, 92%). 1H NMR (300 MHz, CDCl3, TMS): d=8.35 (d, 2H,
7.9 Hz), 7.99 (t, 1H, 7.9 Hz), 7.54 (bs, 2H), 4.63 (d, 3J(H,H)=4.9 Hz,
4H), 3.83 (s, 8H), 2.82 (q, 3J(H,H)=7.6 Hz, 4H), 2.7 (q, 3J(H,H)=
7.4 Hz, 8H), 1.86 (bs, 8H), 1.2 ppm (t, 3J(H,H)=7.4 Hz, 12H); 13C NMR
(75 MHz, CDCl3): d=163.56, 149.33, 142.22, 139.08, 131.52, 125.58, 39.61,
38.44, 23.12, 17.06, 16.85 ppm; IR (deposit from CDCl3 on NaCl): n˜ =
3294 cmꢀ1 (NH) and 1659 cmꢀ1 (C=O); HRMS (CI+): m/z: 630.449
(calcd for C37H56N7O2 ([M+H+]): 630.449).
7,23-Bis(aminomethyl)-6,8,22,24,34,36-hexaethyl-3,11,19,27,33,35-hexa-
azopentacyclyo[27.3.1.15,9.113,17.121,25]hexatriaconta-
1(33),5,7,9(36),13,15,17(35),21,23,25(34),29,31-dodecaene-2,12,18,28-te-
trone (10): A solution of 9 (385 mg, 0.401 mmol) in dichloromethane–tri-
flouroacetic acid (1:1 (v/v); 5 mL) was stirred at room temperature for
4 h and then concentrated by evaporation under reduced pressure to
remove most of the acid. The resulting oil was redissolved in dichlorome-
thane (10 mL) and washed with 20% K2CO3 aqueous solution. The aque-
ous phase was separated and extracted with dichloromethane (2ꢁ
10 mL). The organic layers were combined, dried over Na2SO4, evaporat-
ed to dryness, and dried in vacuum to give a pale pink solid (290 mg) in
95% yield. 1H NMR (400 MHz, CDCl3, TMS): d=8.44 (d, 3J(H,H)=
7.6 Hz, 4H), 8.04 (t, 3J(H,H)=7.8 Hz, 2H), 7.85 (br s, 4H), 4.65 (br s,
8H), 3.89 (s, 4H), 2.75 (q, 3J(H,H)=7.3 Hz, 8H), 2.59 (q, 3J(H,H)=
7.3 Hz, 4H), 1.80 (br s, 2 NH2 and H2O), 1.23–1.13 ppm (m, 18H);
13C NMR (62.5 MHz, CDCl3): d=163.73, 148.85, 143.84, 143.38, 138.65,
131.15, 125.59, 39.01, 38.54, 22.90, 16.56, 15.96 ppm; HRMS (CI): m/z:
761.45092 (calcd for C44H57N8O ([M+H+]): 761.45028).
2,16,18,32,45,47-Hexaethyl-5,13,21,29,34,42,44,46,48-nonaazaheptacy-
clo[15.15.11.13,31.17,11.115,19.123,27.136,40]octatetraconta-
1,3(45),7,9,11(48),15,17,19(47),23,25,27(46),31,36,38,40(44)-pentadecaene-
6,12-dione (5): To a 500-mL round-bottomed flask equipped with an ad-
dition funnel, a condenser and a Dean–Stark, trap was added 12 (0.500 g,
0.794 mmol) followed by toluene (550 mL). The solution was heated to
reflux and the azeotrope of water was received over 4 ꢂ molecular sieves
in the Dean–Stark trap. To the solution was then added a solution of 2,6-
pyridine dicarboxaldehyde (0.214 g, 1.59 mmol) in toluene (275 mL) over
6 h at room temperature. The reaction mixture was then heated to reflux
for 18 h. The remaining dialdehyde was then added over 6 h and the so-
lution again heated to reflux for 18 h. The solution was then cooled to
room temperature and the toluene removed in vacuo. To residue was dis-
solved in anhydrous methanol (100 mL) under argon. To the solution was
then added NaBH4 (365 mg, 9.65 mmol). The solution was then stirred
for three hours under argon. The reaction was quenched with water and
the methanol removed in vacuo. The residue was dissolved in 1n NaOH
(100 mL) and extracted with dichloromethane (3ꢁ100 mL). The com-
bined organic layers were washed with saturated brine, dried over
MgSO4, filtered, and the solvent was removed under vacuum. The result-
ing solid was purified by column chromatography (silica, 1% ammonia-
saturated methanol in dichloromethane) to yield 5 as a yellow solid
2,16,18,32,45,47-Hexaethyl-5,13,21,29,34,42,44,46,48-nonaazaheptacy-
clo[15.15.11.13,31.17,11.115,19.123,27.136,40]octatetraconta-
1,3(45),7,9,11(48),15,17,19(47),23,25,27(46),31,36,38,40(44)-pentadecaene-
6,12,22,28-tetrone (4): In a 250-mL flask equipped with a Dean–Stark
trap, a solution of 2,6-pyridinedicarboxaldehyde (51 mg, 0.38 mmol) and
10 (288 mg, 0.38 mmol) in benzene (150 mL) and methanol (5 mL) was
refluxed for 1 h under argon. The reaction mixture was concentrated to
about 50 mL by distillation off the solvents and then cooled to room tem-
perature. Anhydrous methanol (39 mL) and NaBH4 (140 mg, 3.7 mmol)
were added sequentially. The reaction mixture was stirred at room tem-
perature for 4 h and evaporated to dryness. The residue was partitioned
into dichloromethane (15 mL) and 1n NaOH aqueous solution (10 mL).
The aqueous phase was extracted with dichloromethane (2ꢁ10 mL). The
organic layers were combined, dried over MgSO4, evaporated to dryness.
The residue was separated by column chromatography on silica gel (di-
chloromethane/acetonitrile/methanol 100:30:(0–10) v/v) to give a while
solid (260 mg) in 80% yield. 1H NMR (400 MHz, CDCl3, TMS): d=8.49
1
3
(496 mg, 75%). H NMR (300 MHz, CDCl3, TMS): d=8.35 (d, J(H,H)=
7.4 Hz, 2H), 8.16 (bs, 2H), 8.06 (t, 3J(H,H)=7.4 Hz, 1H), 7.59 (t,
3J(H,H)=7.4 Hz, 2H), 7.08 (d, 3J(H,H)=7.4 Hz, 2H), 4.64 (d, 3J(H,H)=
4.6 Hz, 4H), 3.93 (s, 8H), 3.78–3.77 (m, 8H), 2.88–2.74 (m, 8H), 2.62–
2.55 (m, 4H), 1.82 (s, 4H), 1.17–1.08 ppm (m, 18H); 13C NMR (75 MHz,
CDCl3): d=164.4, 159.8, 150.3, 144.3, 138.8, 137.5, 135.22, 132.0, 129.3,
126.3, 120.9, 118.3, 56.2, 48.6, 38.7, 23.3, 23.2, 16.9, 16.6 ppm; HRMS
(CI+): m/z: 836.534 (calcd for C51H66N9O2: 836.533).
Titration Studies
Stock solutions of 4–6 and their copper(ii) complexes: 10.0 mm solutions
of 4–6, and 50.0 mm solutions of CuCl2 and Cu(OTf)2 in acetonitrile were
made separately in 10-mL flasks by weighing calculated amounts of the
anhydrous compounds and adding HPLC grade acetonitrile to scale.
5.0 mm stock solutions of 4-CuCl2, 4-Cu(OTf)2, 5–2(CuCl2), 6-CuCl2, and
6-Cu(OTf)2 were made in 10-mL flasks by adding 5.00 mL of the 10.0 mm
4–6 solutions, calculated amounts (1.00 or 2.00 mL) of the 50.0 mm cop-
per(ii) salt solutions, and diluted with acetonitrile to scale.
3
3
(d, J(H,H)=7.8 Hz, 4H), 8.04 (t, J(H,H)=7.8 Hz, 2H), 7.72 (br s, 4H),
7.59 (t, 3J(H,H)=7.6 Hz, 1H), 7.08 (d, 3J(H,H)=7.6 Hz, 2H), 4.75 (dd,
2J(H,H)=15.0, J(H,H)=4.7 Hz, 4H), 4.65 (dd, J(H,H)=15.0, J(H,H)=
4.0 Hz, 4H), 4.08 (s, 4H), 3.84 (s, 4H), 2.94–2.82 (m, 4H), 2.68–2.60 (m,
8H), 2.61 (br s, 2H), 1.25–1.15 ppm (m, 18H); 13C NMR (100 MHz,
CDCl3): d=163.97, 158.01, 149.00, 144.77, 144.45, 138.48, 136.50, 134.30,
130.71, 125.99, 120.26, 55.72, 47.90, 38.65, 23.04, 16.18, 15.81 pm; HRMS
(CI): m/z: 864.49280 (calcd for C51H62N9O4 ([M+H+]): 864.49248).
3
2
3
Sodium enolate solutions of 13 and 14: The sodium enolate solutions
were made freshly before each titration by mixing the ketones with one
equivalent of sodium methoxide, because they are not stable, as revealed
by development of yellow color and changes on 1H NMR spectra upon
standing. Then [15]crown-5 or [2.2.1]cryptand were added to complex the
sodium cation and increase solubility. A 0.94m sodium methoxide so-
lution was made directly from 216 mg of sodium metal and anhydrous
methanol in a 10-mL flask. The methoxide solution was transferred into
a polyethylene bottle and kept under argon. The 200 mm solutions of
Na+-13 or Na+-14 were made by mixing 0.667 mL of 300 mm 2-acetylcy-
clopentanone or 1,3-cyclopentanedione and [15]crown-5 (1:1 mol/mol) in
acetonitrile with the 0.94m NaOMe solution (0.213 mL) and diluted with
acetonitrile to 1.00 mL.
Pyridine-2,6-dicarboxylic acid bis[3,5-bis(aminomethyl)-2,4,6-triethylben-
zylamide] (12): To a round-bottomed flask equipped with an addition
funnel, condenser and
a Dean–Stark trap, was added 11 (3.04 g,
6.76 mmol) in benzene. The solution was refluxed and the water re-
moved. The benzene was then removed in vacuo and the residue dis-
solved in anhydrous THF (50 mL; previously distilled over sodium and
benzophenone ketyl) and triethylamine (3.42 g, 33.8 mmol, previously
distilled from calcium hydride). Anhydrous potassium carbonate was
then added. 2,6-Pyridine dicarbonyl dichloride (0.69 g, 3.38 mmol) was
added dropwise in anhydrous THF (25 mL). The reaction was monitored
by thin layer chromatography. The solution was vacuum filtered and the
solvent was removed under vacuum. The residue was purified by column
chromatography (silica, ethyl acetate). The Boc protecting groups were
removed in an aqueous solution of trifluoroacetic acid (100 mL; 1:1 v/v)
and dichloromethane (20 mL). The reaction mixture was stirred at room
temperature for 2 h. The solvent was removed under vacuum and the res-
idue was dissolved in a minimum amount of water. Chloride anion ex-
change was accomplished by using Amberlite IRA-400 (Cl) to yield a
white solid. Continuous extraction for the resulting solid from a 5 n so-
lution of sodium hydroxide using dichloromethane gave 12 as a foamy
General procedure for titrations: All the UV/Vis titrations were per-
formed in a 1-cm quartz cell equipped with a silicon septum. The solu-
tions to be titrated (1.00 mL) were made directly in the cell by adding
calculated amounts of stock solutions and acetonitrile solvent using Ham-
ilton gas-tight syringes. The solutions were mixed by shaking with hand.
Absorbance spectrum of the titrand was recorded. Aliquots (usually
0.1 equivalents) of titrant were added, and spectra were recorded after
each addition of the titrant. In the cases of titrations of copper(ii) com-
2392
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 2385 – 2394

Receptor Complexes
FULL PAPER
plexes with the neutral ketones (2-acetylcyclopentanone or 1,3-cyclopen-
tanedione) or their sodium enolates (13 or 14), the titrant solutions con-
tained the copper(ii) complexes at the same concentrations as the titrand
solutions, so that the copper(ii) complex concentrations kept constant
during the titrations. In the cases of the copper(ii) complexes and the ke-
tones being titrated with the organic bases (15–18), no copper(ii) com-
plexes were presented in the titrant solutions. Instead, high concentration
titrants (150 times of the titrand concentration) were used, and the ab-
sorbance spectra were corrected for the dilution caused by the addition
of titrants that was usually less than 3% of the volume.
Table 3. Crystal data and structure refinement for 7-Cu(OTf)2-H2O.
empirical formula
formula weight
temperature [K]
wavelength [ꢂ]
crystal system
space group
unit cell dimensions
a [ꢂ]
C23H25CuF6N3O7S2
697.12
153(2)
0.71073
orthorhombic
P212121
10.3304(2)
b [ꢂ]
12.4782(2)
Calculation of the pKa shifts: For the acid–base equilibrium between a
carbon acid (the substrate SH) and a base (B), the pKa of the carbon
acid can be calculated from the deprotonation ratio (R) and the pKa of
the conjugate acid of the base (pKaBH) according to Equations (2) and
(3).,
c [ꢂ]
22.0691(5)
2844.81(10)
4
1.628
1.001
1420
0.36ꢁ0.11ꢁ0.09
volume [ꢂ3]
Z
1calcd [Mgmꢀ3
]
m [mmꢀ1
F(000)
]
SH þ B Ð Sꢀ þ BHþ
ð2Þ
ð3Þ
crystal size [mm]
q range for data collection
index ranges
3.16 to 27.488
BH
pKa ¼ pKa ꢀlogfð½Sꢀꢁ½BHþꢁÞ=ð½SHꢁ½BꢁÞg
ꢀ13ꢄhꢄ13, ꢀ16ꢄkꢄ16, ꢀ28ꢄlꢄ28
reflections collected
6427
Since R=[Sꢀ]/[SH]i, we have [BH+]=[Sꢀ]=R[SH]i, [SH]=(1ꢀR)[SH]i,
and [B]=[B]iꢀR[SH]i, where [SH]i and [B]i are the initial total concen-
trations of the acid and base, Equation (3) becomes Equation (4).
independent reflections
completeness to q=27.488 [%]
absorption correction
refinement method
data/restraints/parameters
goodness-of-fit on F2
final R indices [I>2s(I)]
R indices (all data)
6427
99.8
none
full-matrix least-squares on F2
6427/0/381
BH
i
2
i
i
i
ð4Þ
pKa ¼ pKa ꢀlogfðR½SHꢁ Þ =ðð1ꢀRÞ½SHꢁ ð½Bꢁ ꢀR½SHꢁ ÞÞg
1.003
R1=0.0491, wR2=0.0928
R1=0.1004, wR2=0.1112
0.007(16)
none
0.663 and ꢀ0.615
The deprotonation ratio (R) was obtained from two titrations. First, a so-
lution of receptor-copper(ii) complex was titrated with sodium enolate
(13 or 14) to get the absorbance change caused by binding of one equiva-
lent of the enolate (Figure 2 and Figure 3, for example). Second, in the
presence of the receptor-copper(ii) complex, the carbon acid was titrated
with an organic base (Figure 6, for example) to determine the absorbance
change caused by the formation of enolate complex in the presence of
the base. Comparison of the absorbance changes in these two titrations
gave the deprotonation ratio. For example, addition of one equivalent of
enolate 13 into 2.0 mm 6-Cu(OTf)2 solution caused an absorbance change
of 0.23 at 760 nm (Figure 3), and addition of one equivalent of organic
base 17 into 2.0 mm 2-acetylcyclopentanone, which is the carbon acid of
enolate 13, and 2.0 mm 6-Cu(OTf)2 solution caused an absorbance
change of 0.14 at the same wavelength (Figure 6d). Thus, the deprotona-
tion ratio was 0.14/0.23=0.61. The pKa of 17 (pKaBH) is 13.4 in acetoni-
trile, and the initial concentrations of the carbon acid ([SH]i) and base
([B]i) were both 0.002m. Therefore, the observed pKa of 2-acetylcyclopen-
tanone was calculated according to Equation (5).
absolute structure parameter
extinction coefficient
largest diff. peak and hole
[eꢂꢀ3
]
ꢀ
0.6 ꢂ and its Uiso went negative. As a result, the O H bond lengths were
idealized at 0.8 ꢂ by moving the H atom along its O H bond vector. The
subsequent H atom positions were refined riding on the oxygen atom po-
sition with Uiso set to 1.5ꢁUeq for the oxygen atom. The absolute struc-
ture was determined by the method of Flack.[27] The Flack ꢁ parameter
ꢀ
2
refined to 0.01(2) for the configuration given. The function, ꢀw(jFo j ꢀj
2
Fc j )2, was minimized, where w=1/[(s(Fo))2 + (0.0372P)2 + (0.387P)]
2
2
and P=(jFo j + 2jFc j )/3. Rw(F2) refined to 0.111, with R(F) equal to
0.0491 and a goodness of fit, S=1.00. Definitions used for calculating
R(F), Rw(F2) and the goodness of fit, S, are given below.[28] The data were
checked for secondary extinction but no correction was necessary. Neu-
tral atom scattering factors and values used to calculate the linear ab-
sorption coefficient are from the International Tables for X-ray Crystal-
lography (1992).[29] Figure 5 was generated by using SHELXTL/PC.[30]
2
pKa ¼ 13:4ꢀlogfð0:61 ꢃ 0:002Þ =½ð1ꢀ0:61Þ ꢃ 0:002
ꢃð0:002ꢀ0:61 ꢃ 0:002Þꢁ ¼ 13:0
ð5Þ
CCDC-247492 contains the supplementary crystallographic data for this
paper. These data can be obtained free of charge via www.ccdc.cam.a-
c.uk/conts/retrieving.html (or from the Cambridge Crystallographic Data
Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44) 1223-336-
033; or e-mail: deposit@ccdc.cam.ac.uk).
X-ray structure analysis: Crystals of 7-Cu(OTf)2-H2O were prepared as
blue needles by slow evaporation of a solution of 7 (22 mg, 0.070 mmol)
and Cu(OTf)2 (25 mg, 0.070 mmol) in dichloromethane (0.5 mL). The
data crystal was cut from a long needle and had approximate dimensions
of 0.36ꢁ0.11ꢁ0.09 mm. The data were collected on a Nonius Kappa
CCD diffractometer using a graphite monochromator with MoKa radia-
tion (l=0.71073 ꢂ). A total of 309 frames of data were collected using w
scans with a scan range of 1.18 and a counting time of 181 s per frame.
The data were collected at 153 K using an Oxford Cryostream low tem-
perature device. Details of crystal data, data collection and structure re-
finement are listed in Table 3. Data reduction was performed by using
DENZO-SMN.[24] The structure was solved by direct methods using
SIR97[25] and refined by full-matrix least-squares on F2 with anisotropic
displacement parameters for the non-H atoms using SHELXL-97.[26] The
hydrogen atoms on carbon were calculated in ideal positions with iso-
tropic displacement parameters set to 1.2ꢁUeq of the attached atom
(1.5ꢁUeq for methyl hydrogen atoms). The hydrogen atoms on the water
molecule and on the nitrogen atoms were observed in a DF map and re-
fined with isotropic displacement parameters. One hydrogen atom on
Acknowledgements
We gratefully acknowledge the NIH for support of this work (GM-
65515–2).
[1] J. Kraut, Science 1988, 242, 533–540.b) J. R. Knowles, Nature 1991,
350, 121–124.
[2] a) C. Alhambra, J. Gao, J. C. Corchado, D. G. Truhlar, J. Am. Chem.
Soc. 1999, 121, 2253–2258; b) H. Liu, Y. Zhang, W. Yang, J. Am.
Chem. Soc. 2000, 122, 6560–6570.
ꢀ
O1W did not refine well, H1wb. The O H bond length refined to below
Chem. Eur. J. 2005, 11, 2385 – 2394
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2393

E. V. Anslyn et al.
[3] a) J. P. Guthrie, R. Kluger, J. Am. Chem. Soc. 1993, 115, 11569–
11572; b) M. Maurice, S. L. Bearne, Biochemistry 2000, 39, 13324–
13335; c) S. Glavas, M. E. Tanner, Biochemistry 2001, 40, 6199–
6204.
[4] a) M. K. Dreyer, G. E. Schulz, J. Mol. Biol. 1996, 259, 458–466;
b) W.-D. Fessner, A. Schneider, H. Held, G. Sinerius, C. Walter, M.
Hixon, J. V. Scholss, Angew. Chem. 1996, 108, 2366–2369; Angew.
Chem. Int. Ed. Engl. 1996, 35, 2219–2221.
[17] C. A. Blanco, C. Arroyo, J. Phys. Org. Chem. 2000, 13, 713–718.
[18] J. H. Boothe, R. G. Wilkinson, S. Kushner, J. H. Williams, J. Am.
Chem. Soc. 1953, 75, 1732–1733.
[19] H. K. Hall, J. Am. Chem. Soc. 1957, 79, 5444–5447.
[20] C. L. Hannon, D. A. Bell, A. M. Kelly-Rowley, L. A. Cabell, E. V.
Anslyn, J. Phys. Org. Chem. 1997, 10, 396–404.
[21] B. Chawla, S. K. Mehta, J. Phys. Chem. 1984, 88, 2650–2655.
[22] B. H. Aue, H. M. Webb, W. R. Davidson, P. Toure, H. P. Hop-
kins, Jr., S. P. Moulik, D. V. Jahagirdar, J. Am. Chem. Soc. 1991, 113,
1770–1780.
[5] A. M. Kelly-Rowley, V. M. Lynch, E. V. Anslyn, J. Am. Chem. Soc.
1995, 117, 3438–3447.
[6] M. M. Benning, K. L. Taylor, R.-Q. Liu, G. Yang, X. Hong, G. We-
senberg, D. Dunaway-Mariano, H. M. Holden, Biochemistry 1996,
35, 8103–8109.
[23] H. P. Hopkins, Jr., D. V. Jahagirdar, S. P. Moulik, B. H. Aue, H. M.
Webb, W. R. Davidson, M. Pedley, J. Am. Chem. Soc. 1984, 106,
4341–4348.
[7] T. S. Snowden, A. P. Bisson, E. V. Anslyn, J. Am. Chem. Soc. 1999,
121, 6324–6325.
[8] W. W. Cleland, M. M. Kreevoy, Science 1994, 264, 1887–1890.
[9] R. D. Kobes, R. T. Simpson, B. L. Vallee, W. J. Rutter, Biochemistry
1969, 8, 585–588.
[10] E. Kimura, T. Gotoh, T. Koike, M. Shiro, J. Am. Chem. Soc. 1999,
121, 1267–1274.
[11] a) P. Molenveld, J. F. J. Engbersen, D. N. Reinhoudt, Angew. Chem.
1999, 111, 3387–3390; Angew. Chem. Int. Ed. 1999, 38, 3189–3192;
b) D. A. Rockcliffe, A. E. Martell, J. H. Reibenspies, J. Chem. Soc.
Dalton Trans. 1996, 167–175.
[12] M. Komiyama, S. Kina, K. Matsumura, J. Sumaoka, S. Tobey, V. M.
Lynch, E. V. Anslyn, J. Am. Chem. Soc. 2002, 124, 13731–13736.
[13] D. A. Evans, M. C. Kozlowski, J. A. Murry, C. S. Burgey, K. R.
Campos, B. T. Connell, R. J. Staples, J. Am. Chem. Soc. 1999, 121,
669–685.
[24] Z. Otwinowski, W. Minor, Methods in Enzymology, 276: Macromo-
lecular Crystallography, part A, (Eds.: C. W. Carter, Jr., R. M.
Sweets) Academic Press 1997, pp. 307–326.
[25] A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C. Giaco-
vazzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori, R. Spagna, J.
Appl. Crystallogr. 1999, 32, 115–119.
[26] G. M. Sheldrick, Program for the Refinement of Crystal Structures,
University of Gottingen, Germany, 1997.
[27] H. D. Flack, Acta Crystallogr. Sect. A 1983, 39, 876–881.
2
2
[28] Rw(F2)={ꢀw(jFo j ꢀjFc j )2/ꢀw(jFo j)4}1/2 where
w is the weight
given each reflection. R(F)=ꢀ(jFo jꢀjFc j)/ꢀjFo j} for reflections
with Fo >4(s(Fo)). S=[ꢀw(jFo j ꢀjFc j )2/(nꢀp)]1/2, where n is the
number of reflections and p is the number of refined parameters.
[29] International Tables for X-ray Crystallography, Vol. C (Ed.: A. J. C.
Wilson), Kluwer Academic Press, Boston, 1992, Tables 4.2.6.8 and
6.1.1.4.
2
2
[14] a) D. J. Iverson, G. Hunter, J. F. Blount, J. R. Damewood, K. Mislow,
J. Am. Chem. Soc. 1981, 103, 6073–6083; b) T. D. P. Stack, S. Hou,
K. N. Raymond, J. Am. Chem. Soc. 1993, 115, 6466–6467; c) C.
Walsdorff, W. Saak, S. Pohl, J. Chem. Res. Synop. 1996, 282–283.
[15] A. S. Meyer, Jr., G. H. Ayres, J. Am. Chem. Soc. 1957, 79, 49–53.
[16] L. A. Cabell, M. D. Best, J. J. Lavigne, S. E. Schneider, D. M. Per-
reault, M.-K. Monahan, E. V. Anslyn, J. Chem. Soc. Perkin Trans. 2
2001, 315–323.
[30] G. M. Sheldrick, SHELXTL/PC (Version 5.03), Siemens Analytical
X-ray Instruments, Inc., Madison, Wisconsin, USA, 1994.
Received: April 22, 2004
Revised: October 14, 2004
Published online: December 30, 2004
2394
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 2385 – 2394