Ligand spheres in asymmetric hetero Diels-Alder reactions catalyzed by Cu(II) box complexes: experiment and modeling [original]

Dalton

Transactions

PAPER

Cite this: Dalton Trans., 2014, 43,

698

Received 17th June 2013,

Accepted 13th September 2013

DOI: 10.1039/c3dt51602d

www.rsc.org/dalton

Ligand spheres in asymmetric hetero Diels–Alder

reactions catalyzed by Cu(II) box complexes:

experiment and modeling†

V. Umamaheswari,

Pawel Cias,

Andreas Pöppl,

Martin Kaupp

and

Georg Gescheidt*

The stereoselective hetero Diels–Alder reaction between ethyl glyoxylate and cyclohexadiene catalyzed

by [Cu(II)t-Bu-(box)](OTf)

was investigated. The reaction was performed step-by-step and the geometry

of the Cu(II) complexes formed in the course of the catalysis was analysed by EPR spectroscopy, advanced

pulsed EPR methods (ENDOR, and HYSCORE) and DFT calculations. Our results show that one triﬂate

counterion is directly coordinated to Cu(II) during the catalytic process (axial position). This leads to

penta-coordinated Cu(II) complexes. Solvent molecules are able to alter the geometry of the Cu(II) com-

plexes although their coordination is weak. These ﬁndings provide an explanation for the solvent and

counterion eﬀects observed in many catalytic reactions.

1. Introduction

Stereoselective catalysis provides an eﬃcient access for the

construction of complex molecules containing chiral

centers.

1–4

A prominent strategy to introduce C–C bonds into

even complex molecular skeletons is the Diels–Alder reaction.

It is well known that the presence of Lewis acids activates di-

enophiles and leads to a substantial enhancement of the con-

version.

5,6

In particular, Cu(II) complexes carrying chiral

ligands have been shown to be very eﬃcient catalysts.

1,7,8

Here, C

-symmetric bis(oxazoline) (box) derivatives have been

particularly successful.

2,9–18

and a number of related mole-

cules binding to Cu(II)via two nitrogen atoms have been

developed.

19–24

However, not only the nature of the ligand but

also the solvent and counterions play a decisive role in the

eﬃciency and stereoselectivity of the catalytic reactions.

25–30

This is particularly reflected in the recent developments of

ligand design and the involvement of alternative

counterions.

31,32

Nevertheless, mechanistic details in terms of the arrange-

ment of ligands are scarce. In the models utilized for

theoretical investigations,

30,33–40

stereoselectivity is particularly

ascribed to steric eﬀects in tetracoordinated Cu(II) environ-

ments; counterion and solvent eﬀects are mostly neglected. A

recent study of Cu(II)box complexes shows that environmental

eﬀects are crucial.

EPR studies provide rather precise insights into the geo-

metry of transition-metal complexes.

42,43

For Cu(II) bis(sulfoxi-

mine) complexes, we have shown that penta coordination

occurs in the course of (hetero) Diels–Alder reactions, with

counterions and solvent molecules participating in the first

coordination sphere.

44,45

A compatible geometry was found by

X-ray analysis in the complex [Cu(II)t-Bu(box)(OH

)

](OTf)

The question is whether such an arrangement of ligands

(penta-coordination) represents a general concept within the

course of Cu(II) catalyzed Diels–Alder reactions. We therefore

investigated a hetero Diels–Alder reaction under “real con-

ditions”. This means that we utilised frequently used sub-

strates and performed the reactions under synthetic

conditions. Ethyl glyoxylate (1) serves as the dienophile which

adds to cyclohexadiene (2) with [Cu(II)t-Bu(box)](OTf)

as the

catalyst in CH

(the most commonly used solvent,

Scheme 1). The reaction is followed stepwise by EPR spec-

troscopy. The shape of the solid-state continuous-wave EPR

spectra indicates whether a paramagnetic system has a high or

low symmetry (e.g. axial symmetry). Interactions between the

Cu(II) and its nearest neighbors can be derived from anisotro-

pic hyperfine coupling constants (HFCs). In experimental EPR

spectra the corresponding splittings are often not resolved. To

obtain precise, orientation-dependent HFC values, we utilized

pulsed ENDOR (electron nuclear double resonance)

and

†Electronic supplementary information (ESI) available. See DOI:

10.1039/c3dt51602d

Institute of Physical and Theoretical Chemistry, Graz University of Technology,

Stremayrgasse 9, 8010 Graz, Austria. E-mail: [email protected];

Fax: (+43)316 873 32202

Faculty of Physics and Earthscience, University of Leipzig, Linnéstr. 5, D-04103,

Leipzig, Germany

Technische Universität Berlin, Fakultät II, Institut für Chemie, Strasse des 17,

Juni 135, 10623 Berlin, Germany

698 |Dalton Trans.,2014,43, 698–705 This journal is © The Royal Society of Chemistry 2014

Published on 13 September 2013. Downloaded by TU Berlin - Universitaetsbibl on 25/02/2016 14:01:27.

View Article Online

View Journal

| View Issue

HYSCORE (hyperfine sub-level correlation).

With the help of

theoretical DFT calculations, which also provide EPR para-

meters, these experimental parameters can be translated into

the geometry of the chemical environment around the para-

magnetic Cu(II) center.

48–52

2. Experimental

Sample preparation: the Cu(OTf)

and (S)-(−)-2,2′-isopropylidene-

bis(4-tert-butyl-2-oxazoline) ligand were dissolved in dry

in a molar ratio of 1 : 1.2 and stirred for 3 hours under

argon to form the catalyst, [Cu(II)t-Bu(box)](OTf)

in solution.

Part of this solution was transferred to the quartz X-band EPR

sample tube, degassed by three freeze–pump–thaw cycles and

sealed. This forms the catalyst sample (sample S1). Then ethyl

glyoxylate in seven fold molar excess was added to the catalyst

solution and part of this solution was transferred to another

EPR sample tube under argon, degassed and sealed. This

forms the sample S2. To the solution forming sample S2, 1,3-

cyclohexadiene was added in a four fold molar excess to that of

ethyl glyoxylate, and this solution was transferred into the EPR

sample tube under argon, degassed and sealed. This forms the

sample S3. EPR measurements: the CW and pulsed EPR

measurements were performed using X-band (ν

= 9.7 GHz)

BRUKER ELEXYS E580 and ESP 380 spectrometers. The CW

EPR spectra were recorded at 7 K using an Oxford He flow cryo-

stat. The B

modulation amplitude used was 0.4 mT, and the

modulation frequency was adjusted to ν

mod

= 100 kHz. The

microwave power used was low enough to prevent the satur-

ation of spin systems. The simulations of the CW EPR spectra

were done using the Easyspin simulation package.

Two pulse

field-swept electron spin echo (FS ESE) experiments were

carried out at 4 K using microwave pulses with pulse lengths

of 16 ns for π/2 and 32 ns for πpulses and a pulse delay of τ=

160 ns. Orientation-selective two dimensional ESEEM experi-

ments were performed at 4 K using the HYSCORE

sequence

(π/2-τ-π/2-t

-π-t

-π/2-τ-echo). Pulse lengths of 16 ns for π/2 and

32 ns for πpulses were employed and three diﬀerent pulse

delays of τ= 32, 104 and 160 ns were chosen in the 2D experi-

ments to minimize blind spots in the HYSCORE powder pat-

terns. Since the dead time of the spectrometer exceeds 32 ns

in our experimental setup, in the case of τ= 32 ns the

HYSCORE echo was detected via a remote echo sequence

(π/2-τ

-π/2-τ

-π-τ

-echo).

The pulse delays of remote echo

sequence were t

=4μs and τ

= 140 ns. A four-step phase

cycle suggested by Gemperle et al.

was used to avoid interfer-

ence with the unwanted two- and three-pulse echoes. A 170 ×

170 2D data matrix was sampled with a dwell time of 16 ns.

Baseline correction was done by subtracting a third-order poly-

nomial of the experimental data set in both time domains.

Finally, the HYSCORE spectra recorded with diﬀerent τvalues

were added and the 2D FT magnitude spectra were calculated

and presented as contour plots. The simulations of the

HYSCORE spectra were calculated in the time domain by exact

diagonalization of the spin Hamiltonian. Subsequently, the

computed time domain spectra were likewise zero filled and

transformed into the frequency domain by Fast Fourier Trans-

formation. For further information concerning the procedure

for the calculation of orientation selective HYSCORE spectra,

we refer to the paper by Pöppl et al.

Selective microwave

pulses of t

π/2

= 96 and t

= 192 ns, and a radiofrequency pulse

length of t

=8μs were used in orientation selective Davies

pulsed ENDOR experiments.

In an eﬀort to suppress both

the proton and the fluorine signals and enhance those from

the two imine nitrogen atoms (from the t-Bu-box ligand) with

substantially larger hyperfine couplings (HFC) involved in the

coordination with the Cu(II), additional hyperfine contrast

selective ENDOR experiments were done with t

π/2

= 16 and

= 32 ns.

The geometries of [Cu(II)t-Bu(box)](OTf)

and [Cu(II)t-Bu-

(box)(ethyl glyoxylate)](OTf) complexes used for the hyperfine

structure and g-tensor computations were optimized in unrest-

ricted Kohn–Sham calculations at the B3LYP

58–60

level using

the Turbomole package.

The basis sets were of polarized

triple-ζquality for all atoms (TZVP).

The same holds for the

calculations of the energetic eﬀects of S3 formation. All hyper-

fine calculations were carried out with the ORCA

program

package at the optimized geometries and using hybrid B3LYP

functional. The choice of this functional was based on pre-

vious computations which show that it is very successful in the

prediction of hyperfine coupling (HFC) and g-tensor in nitro-

gen and Cu(II) complexes.

48,64–66

Ligand atoms were treated by

Huzinaga–Kutzelnigg type basis sets BII (denoted also as

IGLO-II).

67,68

For the Cu center an accurate triply polarized

basis set CP(PPP) was employed.

This basis set is especially

flexible in the core region and is believed to provide results

close to the basis set limit for the Fermi contact interaction.

Because the spin–orbit eﬀects are known to influence the

HFC results for 3d transition metal complexes, in the case

of the copper atom the contributions of spin–orbit

coupling (SOC) to the HFC were taken into account. It was

proven that the implementation of SOC contributions into the

HFC computations results in a significant improvement of cal-

culated parameters when compared to experiment.

48,66

The

calculation of g-tensor values was carried out with the same

geometry and basis sets. A common gauge at the copper atom

was employed.

Scheme 1 A model hetero Diels–Alder reaction.

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2014 Dalton Trans.,2014,43,698–705 | 699

Published on 13 September 2013. Downloaded by TU Berlin - Universitaetsbibl on 25/02/2016 14:01:27.

View Article Online

3. Results and discussion

EPR ENDOR/HYSCORE and calculations

General –cw EPR Spectra: Frozen-solution EPR spectra

obtained at each single reaction stage (S1,S2, and S3,

Scheme 1) exhibit diﬀerences (Fig. 1). Whereas for stage S1

one EPR signal is recorded, the spectra taken after addition of

1(S2) and subsequently 2(S3) indicate an overlay of two com-

ponents (one resembling the preceding stage). The shapes of

these spectra were reproducible in several runs and are very

likely based on equilibria connecting the diﬀerent stages of

the reaction (for simulations and g-factors, see ESI†). In the

following sections we will describe and discuss the experi-

mental and calculated EPR parameters of S1–S3.

Reaction stage S1: the complex of t-Bu-box with Cu(II) tri-

flate has been described frequently since it is the basis of

several catalyses. Based on the shape of the EPR signal (Fig. 1),

the complex between Cu(OTf)

and the t-Bu-box (box) ligand

S1 reveals a pseudo axial symmetry. Corresponding HYSCORE

spectra taken at the perpendicular and the parallel region

(Fig. 2a and b, respectively) reveal the hyperfine coupling con-

stants shown in Table 1. From the above obtained constants,

substantial distances between the Cu center and some of the

Fand H nuclei (indicated in Fig. 3) could be inferred (Table 2).

To do so, a clear-cut assignment of EPR coupling constants

was accomplished. In this context it was of benefit, that,

according to the calculations, only few distinct atoms revealed

characteristic coupling constants. This was particularly the

case for protons at the 4-positions of the box ligand (H

Fig. 3) and the F atoms in the triflate counterions. This

finding essentially holds for all ligand spheres we have taken

Fig. 1 X-band CW EPR spectra obtained in the course of the reaction

sequence displayed in Scheme 1 corresponding to S1,S2,andS3. The

dotted lines point to the presence of lines stemming from the preceding

EPR spectrum in S2 and S3.

Fig. 2 Comparison of HYSCORE spectra of stage S1 and S3 at selected

orientations in the EPR spectra (a, perpendicular region of S1; b, parallel

region of S1; c, perpendicular region of S3; d, parallel region of S3).

Table 1 Experimental and calculated EPR parameters used for the

determination of the geometry of S1 (Scheme 1)

⊥

/MHzA

iso

/MHz

Exp. Calc. Exp. Calc. Exp. Calc.

Cu 37 66 464 −487

34 29 40 39 36 32

−1.18 1.19 8.75 7.17 2.13 2.75

−2.28 −1.85 5.31 5.77 0.25 0.75

−2.23 −1.04 7.16 5.63 0.9 1.42

−2.30 −1.14 2.80 1.28 −0.6 −0.12

The values for the

N hfc have an error of ±4 MHz based on contrast-

selective ENDOR, cf. ESI.

Fig. 3 Calculated geometry of [Cu(II)t-Bu(box)](OTf)

Paper Dalton Transactions

700 |Dalton Trans.,2014,43, 698–705 This journal is © The Royal Society of Chemistry 2014

Published on 13 September 2013. Downloaded by TU Berlin - Universitaetsbibl on 25/02/2016 14:01:27.

View Article Online

into account here, and substantially simplified the analysis of

the experimental data. Thus, distances could be determined

via a point-dipole approximation from the dipolar hyperfine

coupling constants (HFC) obtained from HYSCORE spectro-

scopy (see ESI†). These results, in principle, confirm the for-

merly established structure of the t-Bu-box ligand. In order to

support these spectroscopic data, however, we have computed

the geometry of S1 using the B3LYP/TZVP protocol. In line

with the X-ray structure determinations and the EPR data,

Cu(II) is surrounded by the two imino nitrogen atoms of the

box ligand and two oxygens of the OTf

−

counterions at (quasi)

equatorial positions. The geometry is displayed in Fig. 3 and

selected distances are summarized in Table 2.

To investigate whether the calculated geometries corres-

pond to those from the X-ray analysis we calculated the well-

described complex [Cubox]Cl

as a reference. As shown in the

ESI,†the calculated and the experimental geometry are vir-

tually matching. Accordingly, the computed geometries can be

regarded as very suitable predictions for the experimental

data. Moreover, in Table 3, the calculated values are compared

with experimentally determined data of related Cu(II)-box

derivatives.

10,37

The experimental and the DFT-calculated distances and

angles for this reference agree rather well (Table 3) and are in

line with X-ray structure data of related complexes.

37,70

Reaction stages S2 and S3: the EPR spectra taken for these

two stages reveal almost matching

63,65

Cu coupling constants

(S2:A

= 424, A

⊥

= 38 MHz: S3:A

= 460, A

⊥

= 38 MHz) indicat-

ing pseudo-axial symmetry. Moreover, the HYSCORE reson-

ances obtained from S2 and S3 are basically identical. It can,

therefore, be assumed that the first ligand sphere around

Cu(II) cation remains essentially unchanged when the diene is

added to S2,i.e. only the (activated) dienophile reveals direct

bonding with Cu(II) center. Therefore, we discuss the structure

of S2 and S3 jointly. For computational reasons it is more

straightforward to develop S2 first and then to add the diene,

to obtain S3. The HYSCORE spectra particularly reveal two

clearly discernible interactions: one set of

H coupling con-

stants (A

= 6.82, A

⊥

=−1.46 MHz), one of

F(A

= 1.46, A

⊥

−0.73 MHz), and of

N(A

= 48, A

⊥

= 32 MHz, also corrobo-

rated by ENDOR). These experimental data have to be com-

pared to their calculated counterparts to obtain the shapes of

the ligand spheres (see Table 4). The

N hfc of the two bis

(oxazoline) nitrogens compares very well with its calculated

counterpart (A

iso

37 MHz, vs. 41 MHz). The hfc values of the

Cu center also agree closely with the calculated ones (A

⊥

38 MHz, A

= 424 vs. 52 and −536 MHz, respectively). The

same holds for the biggest

H hfc of A

⊥

=−1.46 A

6.82 MHz

(calc.: −1.40 and 4.60 MHz, resp.) and the

F hfc of A

⊥

−0.73 A

= 1.46 MHz (calc.: −0.60 and 1.19 MHz, resp.) which

also translates into a Cu⋯F distance of 477 pm (calc.,

483 pm). A comparison with published investigations on the

popular Cu(II) bis(oxazoline)-based catalytic systems is in place

here. In most cases, the mechanism of Diels–Alder reactions

performed in the way shown in Scheme 1 is implied to

proceed via tetra coordinate 17-electron stages with the box

ligand and a substrate molecule (the dienophile) coordinated

to Cu(II). This concept was followed in computational

approaches.

33,34,37

Rarely were solvent molecules explicitly

taken into account.

Stereoselectivity has been predominately

ascribed to the steric congestion of the substituents at the 4.4′

position in the 1,3-bisoxazoline rings; moreover it has been

emphasized that the constitutional flexibility and the

dynamics of the ligands play an important role in stereo-

selectivity.

Pentacoordination was reported in the course of

catalyses performed in zeolites containing the Cu(II) box

complexes

and in enantioselective aldol reactions with

[(Cu(II)S,S)-pybox](OTf)

where the five ligand atoms origi-

nate from the three nitrogens of a pyridylbox ligand and the

two oxygens of a glyoxylate. Moreover enzymes, e.g. galactose

oxidase display penta-coordinated Cu(II) centers.

72,73

With this

background, we have calculated models for ligand spheres S2

but unlike for [Cu(II)t-Bu(box)](OTf)

, in this case, both the

counterion and solvent molecules are taken into account.

Three diﬀerent environments are constructed. They mirror the

composition of the reaction mixture and the experimentally

established finding that pentacoordination exists around Cu(II)

Table 2 Comparison of selected experimental and calculated Cu(II)⋯H

and Cu(II)⋯F distances (r)inS1. (For numbering, see Fig. 3)

Atoms r(EPRDipolexp)/pm r(calc.)/pm

Cu⋯H

316 311

Cu⋯H

280 291

Cu⋯F

294 352

Cu⋯F

360 353

Table 4 Experimental EPR parameters used for the simulation of the

EPR and the HYSCORE spectra attributed to S2 (Scheme 1) and calcu-

lated data

⊥

/MHzA

iso

/MHz

Exp. Calc. Exp. Calc. Exp. Calc.

Cu 38 52 424 −536

32 36 48 49 37 41

H−1.46 −1.4 6.82 4.6 1.3 0.5

F−0.73 −0.6 1.46 1.19 0.0 0.0

See text for the explanation of the ligand spheres.

The values for the

N hfc have an error of ±4 MHz based on contrast-selective ENDOR,

cf. ESI.

Table 3 Comparison of relevant interatomic distances of calculated

[Cu(II)t-Bu-(box)](OTf)

and [Cu(II)t-Bu(box)]Cl

complexes with reported

X-ray data

Cl Cl

OTf OTf

Cu–N (pm) 200.6 198.3 202.4 196.1

N–Cu–N (°) 90.9 90.5 90.1 93.6

Cu-ligand (pm) 224.7 223.2 195.4 195.9

Ref. 35.

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2014 Dalton Trans.,2014,43,698–705 | 701

Published on 13 September 2013. Downloaded by TU Berlin - Universitaetsbibl on 25/02/2016 14:01:27.

View Article Online

(Fig. 4). For all calculations, the ligands were placed at clearly

non-bonding distances (<500 pm away from Cu(II)) as the start-

ing point and geometry optimizations were performed without

any constraints. The first ligand sphere (LP1, Fig. 4) comprises

the t-Bu-box ligand, and ethyl glyoxylate (1) and one triflate

anion representing an 18-electron complex providing two N,

and three oxygen atoms for coordination to Cu(II). However,

geometry optimization with this combination leads to a basi-

cally tetra-coordinate ligand sphere with an essentially tetrahe-

dral arrangement consisting of the two t-Bu-box N atoms

, Fig. 5), and two O atoms provided by the OTf

−

and the

aldehyde oxygen of ethyl glyoxylate (O

, Fig. 5). The

second carbonyl oxygen belonging to the dienophile resides at

a significantly longer distance from the Cu(II) center (314 pm)

and can hardly be regarded as a coordinating ligand (O

Moreover, this structure is not consistent with the experi-

mental results, which indicate that ethyl glyoxylate is co-

ordinated equatorially through its carbonyl oxygens in a

bidentate fashion while the triflate ligand is attached to the

Cu center in the axial position through the sulfonic oxygen

atom (revealed by the experimentally established Cu(II)⋯Fdis-

tance). In LP2 (Fig. 4) one additional OTf

−

anion is added to the

components of LP1, to test the influence of charge compen-

sation by two counterions. This indeed leads to pentacoordina-

tion of Cu(II)witht-Bu-box N atoms (N

), both glyoxylate

carbonyl oxygen atoms (O

) and one triflate oxygen (O

The second OTf

−

resides at a distant position without directly

interacting with Cu(II). Again, there is a significant mismatch

between this calculated structure and the experimental data.

Only one carbonyl oxygen lies in the equatorial plane (O

while the other (O

) acts as the axial ligand. In contrast to the

findings by HYSCORE the triflate is coordinated at an equatorial

position. Finally, since it has been observed that CH

is a

particularly good solvent for these catalyses and an additional

triflate does not coordinate directly but leads to a geometry

change of the complex,

LP3 was constructed the following way:

it includes one t-Bu-box ligand, ethyl glyoxylate, one triflate

anion and, explicitly, one solvent molecule (CH

,Fig.4).The

optimized environment of LP3, shows an equatorial (distorted)

square formed by the two t-Bu-box N atoms (N

), the two

carbonyl O atoms of ethyl glyoxylate (O

). Importantly,

one oxygen atom of the triflate anion (O

)servesasanaxial

ligand at a distance of 226/223 pm (S2/S3). This arrangement is,

in fact, consistent with the experimental data. Table 5 presents

a comparison between the experimental and the calculated

interatomic distances Cu⋯F, Cu⋯H

(for numbering of the

H-atom, see Fig. 3), for LP1,LP2,andLP3) together with the ϑ

angles with respect to g

axis. Essential computed geometrical

parameters of the complexes LP1,LP2,andLP3 in Cu(II)center

proximity are listed in Table 6.

Fig. 4 Calculated geometries for ligand spheres LP1–LP3. The grey shades indicate the equatorial, square-type coordination whereas the arrows

highlight the axial oxygen atom, which for LP1 and LP2 is provided by ethyl glyoxylate 1and for LP3 by one of the triﬂate oxygens.

Fig. 5 Detailed view of the ﬁrst coordination sphere around Cu(II) calculated for LP1–LP3. Calculated by DFT at B3LYP/TZVP level. N

and N

illus-

trate two nitrogen atoms of box, O

–oxygen atom coming from the triﬂate anion. O

depict two oxygen atoms of ethyl glyoxylate. The most

relevant bond lengths and dihedrals on which the comparison with the experiment is based are listed in Table 6.

Paper Dalton Transactions

702 |Dalton Trans.,2014,43,698–705 This journal is © The Royal Society of Chemistry 2014

Published on 13 September 2013. Downloaded by TU Berlin - Universitaetsbibl on 25/02/2016 14:01:27.

View Article Online

Energetic eﬀects found with the final stage S3 of the reac-

tion was also calculated for all three structures. The calcu-

lations confirm a high feasibility of the LP3 geometry (ΔE=

−12.4 kcal mol

−1

)) and render LP2 very unfavorable (ΔE=

0.9 kcal mol

−1

). The structure LP1 is favorable from the ener-

getic point of view (ΔE=−9.3 kcal mol

−1

), however, as shown

above, the calculated EPR parameters are not in such a good

agreement with the experiments like LP3. Deeper insight into

the structures and coordination spheres could be provided by

the calculations of full reaction pathways including the compu-

tations of the transition states. In the case of calculations of

transition state structures containing counterions and solvent

molecules, however, it was not possible to eliminate imaginary

frequencies corresponding to their relative motions. These

modes obviously caused the deviation in the TS energy of the

complexes. Due to this inaccuracy this part of the study is still

in progress and will be published in the next paper. Luckily

the calculations of the EPR parameters were free of such

incorrectness.

4. Conclusions

In several investigations based on product analysis, a distorted

square-type ligand sphere formed by the two N atoms of a box-

type ligand and two O atoms of the substrate (1) has generally

been suggested.

Our results, however, show that penta-coordi-

nation at Cu(II) is potentially decisive in the course of stereo-

selective (hetero) Diels–Alder reactions. One triflate counterion

acts as an additional axially oriented ligand providing an

18-electron complex. Our calculations show that even (weakly

coordinating) solvent molecules have a subtle impact on the

geometry of the catalytically-active complex. Numerous

publications on stereoselective catalyses reveal that the

eﬃciency of these reactions substantially depends on the

parent Cu(II) salt and the solvent used. Our results lead to

the following suggestions: (i) an eﬃcient ligand providing an

additional site for axial coordination should diminish the

counterion dependence. (ii) One additional weakly binding

group should decrease the influence of the solvent. These

structural features are displayed in Fig. 6. These results could

serve as a basis for the development of eﬃcient catalytic

systems and may open an additional starting point for the

modeling of enzymatic activity.

Our results elucidate the observations from the synthesis: it

was shown that box ligands with (4,4′-sulfonamidomethyl)

substituents allowing additional coordination sites (i.e. penta-

coordination)

show higher conversion and stereoselectivity

even in the strongly coordinating solvent MeNO

Eﬃcient

penta-coordinate arrangements were shown for Cu(II)-contain-

ing tripodal bis(oxazoline) (“trisox”) derivatives

and

enzymes,

51,72,73

and recent NMR investigations show that sol-

vents and counterions substantially influence the geometry of

the transition-metal complexes in solution.

Establishing the transition state responsible for the for-

mation of the Diels–Alder adduct in the presence of solvent

and counterion molecules will be the next step of our investi-

gations. The most likely starting point will be the geometry

found for state S3, however, the complexity of such an environ-

ment around the Cu center makes these calculations highly

challenging and demanding. We are now further elaborating

this concept by experiment and, particularly, by theory.

Acknowledgements

Financial support for this work by the Austrian Science Fund

(FWF), Project no. P19711 is gratefully acknowledged.

References

1 S. Reymond and J. Cossy, Chem. Rev., 2008, 108, 5359–5406.

2 G. Desimoni, G. Faita and K. A. Jorgensen, Chem. Rev.,

2006, 106, 3561–3651.

Table 6 Essential bond lengths and dihedrals in Cu proximity for LP1,

LP2 and LP3 (Fig. 5) calculated by DFT

LP1 LP2 LP3 (S2/S3)

Dihedral (°)

–Cu–N

–O

139 175 174/174

–Cu–N

–O

76 102 166/167

–Cu–N

–O

155 144 99/99

Distance (pm)

Cu⋯O

314 233 206/205

Cu⋯O

223 202 208/210

Cu⋯O

190 202 226/223

Table 5 Calculated and experimental interatomic distances Cu⋯H

and Cu⋯F distances for LP1,LP2 and LP3. Also given: θangles of F and

H nuclei with respect to g

axis

LP1 LP2 LP3 (S2/S3) Exp.

θ(°) 40 60 52/52 51

Cu⋯H

(pm) 351 320 317/316 307

Fθ(°) 25 36 26/28 32

Cu⋯F (pm) 486 471 483/480 477

Fig. 6 Suggested ligand modiﬁcations on a box ligand leading to

higher catalytic eﬃciency.

Dalton Transactions Paper

Published on 13 September 2013. Downloaded by TU Berlin - Universitaetsbibl on 25/02/2016 14:01:27.

View Article Online

3 R. Rasappan, D. Laventine and O. Reiser, Coord. Chem.

Rev., 2008, 252, 702–714.

4 D. A. Evans, J. S. Johnson and E. J. Olhava, J. Am. Chem.

Soc., 2000, 122, 1635–1649.

5 S. Yao, M. Roberson, F. Reichel, R. G. Hazell and

K. A. Jørgensen, J. Org. Chem., 1999, 64, 6677–6687.

6 M. Johannsen, K. A. Jørgensen, X.-F. Zheng, Q.-S. Hu and

L. Pu, J. Org. Chem., 1999, 64, 299–301.

7 A. Sakakura, R. Kondo, Y. Matsumura, M. Akakura and

K. Ishihara, J. Am. Chem. Soc., 2009, 131, 17762–17764.

8 M. Yu and S. J. Danishefsky, J. Am. Chem. Soc., 2008, 130,

2783–2785.

9 A. Landa, B. Richter, R. L. Johansen, A. Minkkilae and

K. A. Jorgensen, J. Org. Chem., 2007, 72, 240–245.

10 D. A. Evans, C. S. Burgey, N. A. Paras, T. Vojkovsky and

S. W. Tregay, J. Am. Chem. Soc., 1998, 120, 5824–5825.

11 D. A. Evans, M. C. Kozlowski and J. S. Tedrow, Tetrahedron

Lett., 1996, 37, 7481–7484.

12 D. H. Ryu, G. Zhou and E. J. Corey, J. Am. Chem. Soc., 2004,

126, 4800–4802.

13 A. S. Gokhale, A. B. E. Minidis and A. Pfaltz, Tetrahedron

Lett., 1995, 36, 1831–1834.

14 A. K. Ghosh, P. Mathivanan, J. Cappiello and K. Krishnan,

Tetrahedron: Asymmetry, 1996, 7, 2165–2168.

15 A. Pfaltz, Acta Chem. Scand., 1996, 50, 189–194.

16 J. M. Takacs, D. A. Quincy, W. Shay, B. E. Jones and

C. R. Ross II, Tetrahedron: Asymmetry, 1997, 8, 3079–

3087.

17 M. Johannsen, S. Yao, A. Graven and K. A. Jorgensen, Pure

Appl. Chem., 1998, 70, 1117–1122.

18 H. Pellissier, Tetrahedron, 2009, 65, 2839–2877.

19 I. Atodiresei, I. Schiﬀers and C. Bolm, Tetrahedron: Asym-

metry, 2006, 17, 620–633.

20 C. Bolm, M. Verrucci, O. Simic, P. G. Cozzi, G. Raabe and

H. Okamura, Chem. Commun., 2003, 2826–2827.

21 C. Bolm and O. Simic, J. Am. Chem. Soc., 2001, 123, 3830–

3831.

22 K. A. Jorgensen, M. Johannsen, S. L. Yao, H. Audrain and

J. Thorhauge, Acc. Chem. Res., 1999, 32, 605–613.

23 M. Gomez, G. Muller and M. Rocamora, Coord. Chem. Rev.,

1999, 193–195, 769–835.

24 N. Hutabarat, C. Seki, T. Shimizu, M. Hirama, Y. Kohari,

H. Nakano, K. Uwai, N. Takano, E. Kwon and

H. Matsuyama, Heterocycles, 2012, 86, 203–217.

25 O. Acevedo and W. L. Jorgensen, J. Chem. Theory Comput.,

2007, 3, 1412–1419.

26 M. Johannsen and K. A. Jorgensen, J. Chem. Soc., Perkin

Trans. 2, 1997, 1183–1185.

27 M. Johannsen and K. A. Joergensen, Tetrahedron, 1996, 52,

7321–7328.

28 A. Schaetz, R. Rasappan, M. Hager, A. Gissibl and

O. Reiser, Chem.–Eur. J., 2008, 14, 7259–7265.

29 A. Macchioni, Chem. Rev., 2005, 105, 2039–2073.

30 J. M. Fraile, J. I. Garcia, M. J. Gil, V. Martinez-Merino,

J. A. Mayoral and L. Salvatella, Chem.–Eur. J., 2004, 10,

758–765.

31 A. Sakakura and K. Ishihara, Chem. Soc. Rev., 2011, 40,

163–172.

32 S. Antoniotti, V. Dalla and E. Dunach, Angew. Chem., Int.

Ed., 2010, 49, 7860–7888.

33 A. E. Hayden, J. DeChancie, A. H. George, M. Dai, M. Yu,

S. J. Danishefsky and K. N. Houk, J. Org. Chem., 2009, 74,

6770–6776.

34 J. DeChancie, O. Acevedo and J. D. Evanseck, J. Am. Chem.

Soc., 2004, 126, 6043–6047.

35 R. K. Jangid, N. Gupta, R. K. Bansal, M. von Hopﬀgarten

and G. Frenking, Tetrahedron Lett., 2011, 52, 1721–

1724.

36 J. B. Siegel, A. Zanghellini, H. M. Lovick, G. Kiss,

A. R. Lambert, J. L. S. Clair, J. L. Gallaher, D. Hilvert,

M. H. Gelb, B. L. Stoddard, K. N. Houk, F. E. Michael and

D. Baker, Science, 2010, 329, 309–313.

37 J. Thorhauge, M. Roberson, R. G. Hazell and

K. A. Joergensen, Chem.–Eur. J., 2002, 8, 1888.

38 P. Moerschel, J. Janikowski, G. Hilt and G. Frenking, J. Am.

Chem. Soc., 2008, 130, 8952–8966.

39 Theoretical calculations of metal-catalyzed cycloaddition reac-

tions, ed. K. A. Jorgensen, Wiley, New York, 2002.

40 Y.-h. Lam, P. H.-Y. Cheong, J. M. Blasco Mata, S. J. Stanway,

V. Gouverneur and K. N. Houk, J. Am. Chem. Soc., 2009,

131, 1947–1957.

41 M. E. Owen, E. Carter, G. J. Hutchings, B. D. Ward and

D. M. Murphy, Dalton Trans., 2012, 41, 11085–11092.

42 S. Van Doorslaer, I. Caretti, I. A. Fallis and D. M. Murphy,

Coord. Chem. Rev., 2009, 253, 2116–2130.

43 Y. Traa, D. M. Murphy, R. D. Farley and G. J. Hutchings,

Phys. Chem. Chem. Phys., 2001, 3, 1073–1080.

44 C. Bolm, M. Martin, G. Gescheidt, C. Palivan,

D. Neshchadin, H. Bertagnolli, M. Feth, A. Schweiger,

G. Mitrikas and J. Harmer, J. Am. Chem. Soc., 2003, 125,

6222–6227.

45 C. Bolm, M. Martin, G. Gescheidt, C. Palivan, T. Stanoeva,

H. Bertagnolli, M. Feth, A. Schweiger, G. Mitrikas and

J. Harmer, Chem.–Eur. J., 2007, 13, 1842–1850.

46 E. R. Davies, Phys. Lett. A, 1974, A47,1–2.

47 P. Hofer, A. Grupp and M. Mehring, Phys. Rev. A, 1986, 33,

3519–3522.

48 C. Remenyi, R. Reviakine, A. V. Arbuznikov, J. Vaara and

M. Kaupp, J. Phys. Chem. A, 2004, 108, 5026–5033.

49 S. Kababya, J. Nelson, C. Calle, F. Neese and D. Goldfarb,

J. Am. Chem. Soc., 2006, 128, 2017–2029.

50 C. Remenyi, R. Reviakine and M. Kaupp, J. Phys. Chem. B,

2007, 111, 8290–8304.

51 S. Vancoillie, J. Chalupsky, U. Ryde, E. I. Solomon,

K. Pierloot, F. Neese and L. Rulisek, J. Phys. Chem. B, 2010,

114, 7692–7702.

52 J. Gauss, M. Kallay and F. Neese, J. Phys. Chem. A, 2009,

113, 11541–11549.

53 S. Stoll and A. Schweiger, J. Magn. Reson., 2006, 178,

42–55.

54 P. Hoefer, A. Grupp, H. Nebenfuehr and M. Mehring,

Chem. Phys. Lett., 1986, 132, 279–282.

Paper Dalton Transactions

Published on 13 September 2013. Downloaded by TU Berlin - Universitaetsbibl on 25/02/2016 14:01:27.

View Article Online

55 P. Hofer, J. Magn. Reson., Ser. A, 1994, 111,77–86.

56 C. Gemperle, G. Aebli, A. Schweiger and R. R. Ernst,

J. Magn. Reson., 1990, 88, 241–256.

57 A. Pöppl, M. Hartmann, W. Böhlmann and R. Böttcher,

J. Phys. Chem. A, 1998, 102, 3599–3606.

58 A. D. Becke, Phys. Rev. A, 1988, 38, 3098–3100.

59 A. D. Becke, J. Chem. Phys., 1993, 98, 1372–1377.

60 A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.

61 R. Ahlrichs, M. Bar, M. Haser, H. Horn and C. Kolmel,

Chem. Phys. Lett., 1989, 162, 165–169.

62 A. Schafer, C. Huber and R. Ahlrichs, J. Chem. Phys., 1994,

100, 5829–5835.

63 ORCA –an ab initio, Density Functional and Semiempirical

Program Package, 2.5–20.

64 M. Kaupp, R. Reviakine, O. L. Malkina, A. Arbuznikov,

B. Schimmelpfennig and V. G. Malkin, J. Comput. Chem.,

2002, 23, 794–803.

65 F. Neese, J. Phys. Chem. A, 2001, 105, 4290–4299.

66 F. Neese, J. Chem. Phys., 2003, 118, 3939–3948.

67 S. Huzinaga, Approximate Atomic Functions, University of

Alberta Edmonton, Canada, 1971.

68 W. Kutzelnigg, U. Fleischer, M. Schindler, P. Diehl,

E. Fluck, H. Günther, R. Kosfeld and J. Seelig, NMR Basic

Principles and Progress, Springer-Verlag, Berlin, 1990, vol. 23.

69 F. Neese, Inorg. Chim. Acta, 2002, 337, 181–192.

70 D. A. Evans, E. J. Olhava, J. S. Johnson and J. M. Janey,

Angew. Chem., Int. Ed., 1998, 37, 3372–3375.

71 D. A. Evans, S. J. Miller, T. Lectka and P. von Matt, J. Am.

Chem. Soc., 1999, 121, 7559–7573.

72 M. Orio, O. Jarjayes, H. Kanso, C. Philouze, F. Neese and

F. Thomas, Angew. Chem., Int. Ed., 2010, 49, 4989–4992.

73 J. W. Whittaker, Chem. Rev., 2003, 103, 2347–2363.

74 A. Livieri, M. Boiocchi, G. Desimoni and G. Faita,

Chem.–Eur. J., 2011, 17, 516–520.

75 C. Foltz, M. Enders, S. Bellemin-Laponnaz, H. Wadepohl

and L. H. Gade, Chem.–Eur. J., 2007, 13, 5994–6008.

76 M. Kruck, D. C. Sauer, M. Enders, H. Wadepohl and

L. H. Gade, Dalton Trans., 2011, 40, 10406–10415.

Dalton Transactions Paper

Published on 13 September 2013. Downloaded by TU Berlin - Universitaetsbibl on 25/02/2016 14:01:27.

View Article Online