A double conformationally restricted dynamic supramolecular system for the substrate-selective epoxidation of olefins - A comparative study on the influence of preorganization

CHEMCATCHEM

FULL PAPERS

DOI: 10.1002/cctc.201402726

A Double Conformationally Restricted Dynamic

Supramolecular System for the Substrate-Selective

Epoxidation of Olefins—A Comparative Study on the

Influence of Preorganization

Emil Lindbꢀck, Sawsen Cherraben, Jean-Patrick Francoꢁa, Esmaeil Sheibani,

Bartosz Lukowski, Agnieszka ProÇ, Hassan Norouzi-Arasi, Kristoffer Mꢂnsson,

Pawel Bujalowski, Anna Cederbalk, Thanh Huong Pham, Torbjçrn Wixe, Sami Dawaigher,

and Kenneth Wꢀrnmark*^[a]

Dedicated to Professor Christina Moberg on the occasion of her retirement and Professor Torbjçrn Frejd on the occasion of his 70th

birthday

A double conformationally restricted kinetically labile supra-

molecular catalytic system, the third generation, was designed

and synthesized. We investigated the substrate selectivity of

this system by performing competitive pairwise epoxidations

of pyridyl- and phenyl-appended olefins. We compared the ob-

tained substrate selectivities to previous less preorganized

generations of this system. Five different substrate pairs were

investigated, and the present double conformationally restrict-

ed system showed higher normalized substrate selectivities

(pyridyl versus phenyl) for two of the substrate pairs than the

previous less conformationally restricted generations. As for

the preorganization of the components of the system, the cat-

alyst, and the receptor part, it was shown that for each sub-

strate pair there was one generation that was better than the

other to generate substrate-selective catalysis.

Introduction

Despite the many successful examples of manmade homoge-

neous catalysts,^[1]for example, in regioselective,^[2]chemoselec-

tive, and enantioselective^[3]transformations, very few address

substrate selectivity.^[4]One reason for this might be that those

catalysts are designed and synthesized with the aim to cata-

lyze a specific transformation for as broad a spectrum of sub-

strates as possible.^[4]Substrate-selective catalysts on the other

hand are designed with the goal to display reactivity with only

one substrate in a mixture containing several similar sub-

strates. Another reason could be the hitherto lack of applica-

tions for substrate-selective catalysis.^[4]However, one well-exe-

cuted application of substrate-selective reactions is in the elu-

cidation of reaction mechanisms for which the reaction is run

in a competitive mode with carefully selected substrates to ex-

tract information about the mechanism.^[4]

Although heterogeneous catalysis has appeared as a promis-

ing way to obtain substrate selectivity,^[5]the mechanisms of

heterogeneous catalysis are difficult to study,^[6]which leads to

difficulties in designing substrate-selective heterogeneous cat-

alysts. In contrast, homogeneous catalysts are more amenable

to design than heterogeneous ones owing to the higher mech-

anistic understanding of the former.^[6]The masters of catalysis,

the enzymes, are homogeneous catalysts that are able to swift-

ly transform one specific substrate in a mixture with others

into one product with high enantio-, regio-, and chemoselec-

tivity. Inspired by the efficiency of enzymes in these processes,

supramolecular catalytic systems have been developed.^[7]How-

ever, many suffer from poor turnover numbers,^[7]and Sanders

hypothesized that the fear of unfavorable entropy has brought

supramolecular chemists in the direction of rigid systems that

are too preorganized,^[8]which leads to, among other things,

mismatched transition states (TSs) and to product inhibition.

To avoid these limitations on supramolecular catalysis, we

wanted to use a kinetically dynamic catalytic cavity to achieve

discrimination between substrates in homogeneous catalysis.

Such an approach resembles the nonrigid catalytic cavities of

enzymes, for which, for instance, product inhibition is limited

and the catalytic cavity often self-assembled.

[a] Dr. E. Lindbꢀck, S. Cherraben, J.-P. Francoꢁa, Dr. E. Sheibani, B. Lukowski,

A. ProÇ, Dr. H. Norouzi-Arasi, K. Mꢂnsson, P. Bujalowski, A. Cederbalk,

T. H. Pham, Dr. T. Wixe, S. Dawaigher, Prof. K. Wꢀrnmark

Centre for Analysis and Synthesis,

Department of Chemistry, Lund University

P.O Box 121, S-22100 Lund (Sweden)

Fax: (+46)46-2228209

In the present study, we investigated how the degree of pre-

organization of the components of a dynamic supramolecular

catalytic system influences the substrate selectivity in the ep-

E-mail: Kenneth.warnmark@chem.lu.se

Supporting Information for this article is available on the WWW under

http://dx.doi.org/10.1002/cctc.201402726.

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&

1

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

tween the catalyst and receptor parts to linearly adjust accord-

ingly owing to the dynamic nature of the hydrogen bond. This

makes the epoxidation reaction the ideal reaction to be stud-

ied by our supramolecular catalytic cavity. Second, epoxides

are important bulk chemicals that can be converted into differ-

ent fine chemicals in a specific way by using one single syn-

thetic transformation.^[11]Third, olefins are obtained industrially

as a mixture of different homologues in the cracking of oil. To

be able to selectively epoxidize one of these homologues, in-

stead of performing expensive and difficult separation of the

very similar olefin homologues before the epoxidation, would

be an important step toward the more sustainable production

of epoxides.^[4]

There are only a few examples of homogeneous substrate-

selective catalysts that have been evaluated on the basis of

their performance in competitive substrate-selective transfor-

mations.^[4]However, for supramolecular catalysts, the epoxida-

tion of alkenes constitutes the most commonly investigated

type of reaction.^[4]Hence, for use in epoxidation reactions,

metalloporphyrins that discriminate substrates on the basis of

their geometrical shape have been synthesized; the porphyrins

have been decorated with large sterically encumbered

groups^[12]or have been part of a well-defined molecular

basket.^[13]More specific hosts such as metal-ion binding li-

gands^[14a]and cyclodextrins^[14b]have also been attached to

metalloporphyrins. In addition, hydrogen-bonding recognition

Figure 1. First generation of the supramolecular catalytic system, 1+2. The

supramolecular system is represented as so-designed substrate-selective

cyclic heterodimer 3.

oxidation of olefins. This was done by comparing the per-

formance of the present, third-generation system (having the

largest number of preorganized components) to the

second generation (having an intermediate number

of preorganized components) and the first genera-

tion (having the lowest number of preorganized

components).

We took the first steps in this direction some years

ago by designing and synthesizing catalytic Mn^III–

salen complex 1 (Figure 1) on the basis of the Jacob-

sen–Katsuki (J-K) catalyst^[9]and a receptor, Zn^II–por-

phyrin 2 (Figure 1), both designed to aggregate to

a resulting cyclic heterodimer, supramolecular cavity

3 (Figure 1), constituting the first-generation system.

The catalyst and receptor parts each contain the self-

complementary and kinetically labile 2-pyridone

motif, designed and geometrically positioned to pro-

mote the assembly of desired kinetically labile sub-

strate-selective catalytic cavity 3 (Figure 1) at the ex-

pense of other self-assembled supramolecular species

in solution,^[10]according to the model 1+2Ð3 (sche-

matically depicted in Scheme 1, cyclic heterodimer,

gen 1).

To evaluate the system, the pairwise (1:1) competi-

tive substrate-selective epoxidation of olefins was in-

vestigated for which the two substrates in each pair

differed by only one atom (N or C). The epoxidation

reaction was chosen, first, because there is a linear

expansion–contraction in size upon going from sub-

strate to product in that one oxygen atom is added

Scheme 1. Schematic representation of the three generations of the dynamic supra-

molecular system and suggested substrate-selective supramolecular catalytic species in

solution with a symbolic substrate inserted. Note: the distance between the catalyst and

during the reaction (Scheme 2). This expansion is in

the same direction as that given by the two hydro-

gen-bonding systems, which allows the distance be- receptor parts is different between the opened and closed heterodimers.

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&

2

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

Scheme 2. Proposed linear expansion and contraction of the catalyst–recep-

tor cavity of the dynamic supramolecular system suggested to promote the

different stages of the epoxidation of the pyridyl-appended olefin. This is

possible because of the directionality of the hydrogen bonds attached to

each unit (see Figure 1 and Scheme 1). The cartoon represents the dynamics

of the so-designed substrate-selective cyclic heterodimer 3 (Figure 1).

elements have been attached to porphyrin-^[15]and metallosa-

len-based^[16]epoxidation catalysts. Finally, discrimination of

substrates on the basis of the effective electron density at the

catalytic center has also been observed.^[17]However, no supra-

molecular catalyst for substrate-selective epoxidations, except

ours, contains a catalyst part connected to a receptor part by

hydrogen bonding and designed to generate a catalytic cavity.

In fact, no such system has been developed at all and our first-

generation system has for that reason been recognized as

a new principle for the construction of supramolecular cata-

lysts.^[7b]For discrimination between substrates, we used the

well-known interaction between pyridines and Zn^II–porphy-

rins.^[18]Hence, to investigate the substrate selectivity of our

supramolecular system, we incorporated one pyridyl moiety in

one substrate and one phenyl moiety in the other, together

constituting a substrate pair. Several substrates were construct-

ed in silico, and the resulting TS structure in the epoxidation

by using cavity 3 (Figure 1) containing the Mn=O reactive

species was modeled by molecular mechanics (Figure 2).^[19]It

was found that (Z)-pyridyl-appended styrene 4a (Scheme 3)

Scheme 3. Competitive pairwise (1:1) epoxidation reactions in the styrene

and stilbene series used to evaluate the substrate selectivity of the dynamic

supramolecular catalytic system.

made the least strained TS. Hence, compound 4a became the

pyridyl-appended alkene of choice to try in the substrate-selec-

tive epoxidation by using first-generation catalytic system

1+2.

Indeed, a normalized^[20]substrate-selectivity of 1.7 was ob-

served in the pairwise (1:1) competitive epoxidation of (Z)-pyr-

idyl-appended styrene 4a over phenyl-appended 5a

(Scheme 3; Table 1, entry 11) by using system 1 (5 mm)+2

(15 mm),^[21,19]the highest for the first-generation system. This

was rewarding, as according to the modeling, 4a was the sub-

strate that gave the least strained TS (see above). It was until

then assumed that the model, 1+2Ð3 (schematically depict-

ed in Scheme 1, gen 1), was valid, in which 3 constitutes a sub-

strate-selective catalytic cavity. Owing to the hydrogen-bond-

ing connection between the catalyst and receptor parts, the ki-

netic lability of the resulting catalytic cavity was supposed to

generate a low-strain TS for the reaction and thus to give rise

to a high turnover frequency in the catalytic cycle. The sub-

strate selectivity was envisaged to arise from coordination of

pyridyl-appended substrates to the endo side of receptor unit

2 in 3 to give rise to higher effective concentrations of pyridyl-

appended substrates in the vicinity of 1 in 3 relative to the ef-

fective concentrations produced by substrates lacking a recog-

nition element. The substrates and the products were expect-

ed to essentially display the same affinities with receptor unit

2, which would possibly lead to product inhibition of the cata-

lytic system. However, the reaction proceeded with good con-

version,^[19]and thus the similarities in affinities were rational-

ized to be overridden by the larger size of the epoxide prod-

ucts relative to the size of the alkene substrates. This is sup-

posed to result in less-strong hydrogen bonding in the cata-

Figure 2. Molecular mechanics 3D representation of the TS in the epoxida-

tion of receptor-bound substrate 4a inside hydrogen-bonded cavity 3.^[19]

Long alkyl-chain substituents are removed for clarity and ease of modeling.

Zero-order bonds are colored green. This picture is also a model for the

closed heterodimer (Scheme 1) that is assumed to be substrate selective.

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&

3

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

Table 1. Substrate selectivities obtained in the competitive pairwise (1:1)

epoxidation of alkenes in the styrene and stilbene series catalyzed by the

dynamic supramolecular systems.^[a]

Catalyst

Selectivity^[b,c](inherent or normalized/real)^[20]

Entry (Concentration [mm]) 4a/5a 8a/9a 10a/

25a/

27a/

11 a

26a

11 a

1

2

6 (5)

28 (5)^[19]

0.91

1.1

0.41

1.1

0.27

1.2

1.27

1.2

0.44

1.3

3

4

5^[d]

6^[d]

7

6 (5)+12 (5)

6 (5)+12 (12.5)

6 (5)

6 (5)+12 (5)

6 (5)+2 (5)^[23]

6 (5)+2 (15)^[23]

6 (5)+2 (5)^[23]

1 (5)+2 (5)^[19]

1 (5)+2 (15)^[23]

2.0/1.8 2.1/0.9 3.0/0.8 1.6/2.0

2.9/2.7 3.2/1.3 2.4^[e]/0.6 2.1/2.6

1.6/0.7

2.2/1.0

0.51

1.6/0.8

1.2/0.55

1.5/0.65

1.8/1.6 2.1/0.8 2.2/0.6 2.0/2.5

1.9/1.7 3.4/1.4 3.4/0.9 2.9/3.7

8

9^[d]

10

11

1.3/–

1.7/–

1.5/–

1.6/–

1.5/– 0.7/– 1.7/–

1.7/– 0.6/– 2.4/–

1.3/–

1.2/–

1.4/–

12^[d]1 (5)+2 (5)^[21]

[a] General procedure: Catalyst part (3 mmol), receptor part (3, 7.5, or

9 mmol), substrate pairs (30 mmol each), PhIO (24 mmol), internal standard

benzyl benzoate (15 mmol), CH₂Cl₂(0.6 or 6 mL), RT. [b] See the Experi-

mental Section and Ref. [20] for definitions. [c] Each experiment was re-

peated twice except for the experiment involving only the catalyst part,

which was run only once. The selectivity was determined by GC analysis

(double injections). The average result is reported. Accuracy=better

thanꢀ0.25. [d] 4-Ethylpyridine (90 mL, 0.72 mmol) was added. [e] This

result was confirmed by repeating the reaction one more time according

to [c].

Figure 3. Second generation of the supramolecular catalytic system, 6+2.

The supramolecular system is represented as so-designed substrate-selective

cyclic heterodimer 7.

lyst–receptor product complex than in the corresponding cata-

lyst–receptor substrate complex, which makes diffusion of the

product from the catalytic cavity easier in the former case than

in the latter; therefore, the catalytic cycle is pushed forward.

The low but still significant substrate selectivity suggested

a more complex system than the simple one represented by

the 1+2Ð3 model (Figure 1). Analysis of the supramolecular

system by NMR titrations followed by analysis of the existing

equilibria and their quantification in terms of association con-

stants showed that components 1 and 2 in CDCl₃were not

only in equilibrium with the catalytic cavity, cyclic heterodimer

3, but also with linear homo-1_nand 2_n, noncyclic hetero-oligo-

mers 1_n2_m, (Scheme 1) and finally a trimer consisting of 2,

cyclo-2₃(Figure 6).^[22,19]In fact, the degree of polymerization

was low, so most of the oligomers existed as dimers.^[22,19]An-

other analysis suggested that the large molar fraction of heter-

odimers could, in a conformation as a closed heterodimer,

c-1·2, act as a substrate-selective species.^[19]On the basis of the

equilibrium analysis, we argued that to increase the substrate

selectivity, the most straightforward action would be to in-

crease the concentration of cyclic heterodimer 3. By forcing

components 1 and 2 into a cisoid conformation, a higher ratio

of the substrate-selective cavity, the cyclic heterodimer, relative

to that of other supramolecular species should be formed

within the equilibrium system (Scheme 1).

(schematically depicted in Scheme 1, gen 2), relative to that of

previous system 1+2. It was also assumed that the strap

would somewhat hamper unselective catalysis from taking

place on the exo side of 7. To further assure this point, a pyri-

dine N-oxide moiety was incorporated into the strap. Previous-

ly, a pyridine N-oxide had been strapped to a Mn^III–salen com-

plex and its coordination to Mn^IIIwas demonstrated by X-ray

diffraction analysis of a crystal of the compound.^[24]At the

same time, it was proven that the presence of the pyridine N-

oxide moiety in the strap did not affect the reactivity or the se-

lectivity of the epoxidation reaction. In our case, we simply

wanted the pyridine N-oxide strap to block the manganese ion

from making a Mn=O species having the oxygen atom residing

on the exo side of the catalyst–receptor complex, which would

lead to unselective epoxidations. Rewardingly, in line with our

expectations, this second-generation catalytic system, 6+2,

furnished higher substrate selectivity than the corresponding

first-generation catalytic system, 1+2; it reached a normalized

substrate selectivity^[20]of 3.4 for the competitive pairwise (1:1)

epoxidation in the stilbene series of both (Z)-pyridyl 8a/(Z)-

phenyl 9a and (E)-pyridyl 10a/(E)-phenyl 11 a (Scheme 3;

Table 1, entry 8).^[23]Those were the highest substrate selectivi-

ties in the entire investigation of the second-generation

system.

On the basis of a simple equilibrium analysis and to obtain

an even higher ratio of the substrate-selective cyclic heterodi-

mer, the next clear action is to also force the receptor part into

a cisoid conformation by installing a strap on receptor part 2

(see above), which would lead to strapped receptor part 12.

Hence, herein we present the synthesis of strapped receptor

Following this line, we managed to install a strap in catalytic

part 1 to obtain catalytic part 6 (Figure 3).^[23]Assumingly, this

would thus lead to a higher molar ratio of the so-designed

substrate-selective cavity (see above), compound 7, in the

second-generation system (Figure 3), according to 6+2Ð7

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&

4

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

Synthesis of receptor part 12

The synthesis of the receptor part, strapped Zn^II–porphyrin 12,

commenced by the synthesis of dipyrromethane 14

(Scheme 4), which was needed in the porphyrin condensation

step. The synthesis of 14 started from known 2,6-dimethyl-4-

decyloxybenzaldehyde (15).^[19]Hence, compound 15 was con-

densed with p-toluenesulfonamide (pTsNH₂) to furnish tosyli-

mine 16 in 70% yield. Dipyrromethane 14 was obtained in

17% yield by using the general methodology developed by Te-

melli and Unaleroglu,^[25]which involved condensation of 16

with pyrrole in the presence of Cu(OTf)₂Tf=CF₃SO₂and appli-

cation of the workup and purification procedures that we de-

veloped specifically for 14. The overall sequence not only re-

sulted in an overall higher yield of 14, but also allowed the

synthesis of 14 on a larger scale and with a simpler purification

procedure than its original synthesis.^[19]

The synthesis of required dialdehyde 17 for the condensa-

tion with dipyrromethane 14 to yield the porphyrin product

started from commercially available 2,4-dihydroxybenzalde-

hyde (18, Scheme 5). Treatment of 18 with 1-bromoicosane by

using conditions similar to those previously reported^[26]fur-

nished expected, less sterically encumbered regioisomer 19 in

63% yield. Strapping of 19 was performed by treating 19 with

1,11-dibromoundecane in the presence of potassium carbonate

to give 20 in 95% yield. In the following step, compound 20

was subjected to N-iodosuccinimide (NIS) together with tri-

fluoroacetic acid (TFA) to yield bisiodinated product 21 exclu-

sively as the desired regioisomer in 82% yield. Compound 21

was used as an electrophile in a palladium-catalyzed double

Stille cross-coupling reaction^[27]with stannate 22^[19]as the nu-

cleophile to furnish double-coupling product 17 in 29–42%

yield. The yield of 17 varied presumably as a result of the pres-

ence of variable trace amounts of water in the reaction mix-

ture. Support for this scenario was the identification of prod-

ucts from the competing iodine–hydrogen exchange reaction

taking place on 21, as moniodo-21 and 20 were identified by

ESI-MS. In fact, the highest yield was obtained if 21 and 22

were suspended in toluene and the resulting suspension was

evaporated to dryness prior to the reaction to remove as

much trace amounts of water as possible.

Figure 4. Third generation of the supramolecular catalytic system, 6+12.

The supramolecular system is represented as so-designed substrate-selective

heterodimer 13.

unit 12 to give the third-generation catalytic system, 6+12

(Figure 4), which is expected to result in a higher concentra-

tion of substrate-selective cavity 13 (Figure 4), according to

6+12Ð13 (schematically depicted in Scheme 1, gen 3), rela-

tive to that of the catalytic cavities of the earlier generations,

cavities 3 and 7 (Figures 1 and 3).

Results and Discussion

Synthesis

Synthesis of catalyst part 6

In the porphyrin condensation, aldehyde 17 was treated

with dipyrromethane 14 in the presence of TFA, and the inter-

mediate was oxidized with 2,3-dichloro-5,6-dicyano-1,4-benzo-

quinone (DDQ) to yield porphyrin 23 in 16–22% yield

(Scheme 6). Zn^IIwas inserted

The synthesis of catalyst part 6 is described in our previous

paper on the second-generation supramolecular catalytic

system.^[23]

into 23 to furnish Zn^II–porphyrin

24 in 98% yield by using a stan-

dard protocol. Finally, palladium-

mediated bis-de-O-benzylation

of 24 under atmospheric pres-

sure of H₂gave target com-

pound 12 in 76% yield after fast

chromatography to avoid de-

composition of the product.

Noteworthy, the proton resonan-

Scheme 4. Synthesis of dipyrromethane 14 required for the porphyrin condensation reaction.

ces of the methyl groups of the

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&

5

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

lowing a general procedure^[30]

and was characterized in our

previous paper.^[23]

Methodology

The catalysis

See the Experimental Section for

details. To evaluate the substrate

selectivity of a catalyst, the most

correct method is to employ the

substrates in competitive experi-

ments.^[4]In our case, we wanted

to evaluate the effect on the

substrate selectivity of having

a pyridine auxiliary attached to

an olefin in comparison to

having a benzene auxiliary at-

tached to an olefin in the epoxi-

dation reaction. Discrimination

between five olefinic substrate

pairs (pyridyl- versus phenyl-ap-

pended) by catalytic system 6+

Scheme 5. Synthesis of aldehyde 17 required for the porphyrin condensation. Cy=cyclohexyl, dba=dibenzyli-

deneacetone.

12 on the basis of the interaction with the Zn^II

moiety of the system was investigated in competitive

pairwise (1:1) reactions. During the initial time of a re-

action, the consumption of the substrate is linear

with time. Thus, performing the competitive reac-

tions in this linear region allows for the use of the

ratio of the consumption of each substrate to be

equivalent to the relative rate (the initial-rate

method),^[31]and thus, the substrate selectivity can be

obtained directly. In our previous studies, we ob-

served that a good compromise between reaction

time and initial rate conditions in our systems was to

terminate the reaction at approximately 40% conver-

sion of each substrate.^[23]Thus, 0.40 equivalents of

the oxidant, PhIO, was added to each reaction, and

to be able to calculate the consumption of each sub-

strate, an internal standard (IS), benzyl benzoate, was

Scheme 6. Final steps in the synthesis of receptor part 12: porphyrin condensation, zinc

insertion, and deprotection.

aromatic rings at the meso position of 12 appeared at two dif-

ferent chemical shifts, which indicated that 12 was forced into

a cisoid conformation at least on the timescale of the H NMR

added to each reaction mixture. It turned out that 40% con-

version was reached after 24 h in almost all cases, and at that

time the reaction was terminated by transferring the crude

mixture to a silica column. Then, the organic starting materials

and products were eluted together with the internal standard.

The resulting sample was evaporated to dryness.

1

spectrometer. For 2 on the other hand, the corresponding

proton resonances of the methyl groups appeared at the same

chemical shift, indicating that they are identical.^[19]Another

piece of evidence for the cis conformation of 12 is that the

strap experiences ring current from the porphyrin ring, which

places the resonances of its methylene protons at negative

Analyzing the substrate consumption and product formation

1

chemical shifts in the H NMR spectrum.

See the Experimental Section for details. The consumption of

the starting material was determined by gas chromatography

(GC): a small amount of the dry sample was dissolved in dieth-

yl ether, in which the Zn^II–porphyrin receptor part was insolu-

ble; then, the solution was filtered through a microfilter and

injected into the GC. The rest of the sample was dissolved in

CDCl₃and analyzed by ¹H NMR spectroscopy [except for

Synthesis of the substrates

The substrates were obtained either from commercial sources

(for 9a, 10a, 11 a, and 27a) or were synthesized (for 4a,^[19]

5a,^[19]8a,^[28]and 26a).^[29]Substrate 25a was synthesized fol-

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&

6

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

ized substrate selectivity was obtained as the quota

between the real and the inherent substrate selectivi-

ties, thus taking into account the substrate selectivity

between the two substrates that the catalyst part

itself has induced.

Table 2. Epoxide formation in CH₂Cl₂at RT.^[a]

Entry Catalyst

(concentration

Substrate pair Total epoxide Epoxide

Pyridyl

epoxide

pyridyl/phenyl^[c]Z/E

product^[b]

[%]

selectivity

[mm])

1

2

3

4

6 (5)

6 (5)+12 (5)

6 (5)

8a/9a

27a/11 a

4a/5a

67

92

85

79

0.77

1:0.7

1.60/1.23

n.d.^[d]

0.53

1:0.6

n.d.^[d]

Catalysis

6 (5)

[a] Each experiment was repeated twice except for the experiment involving only the

catalyst part, which was run only once. The selectivity was determined by ¹H NMR

spectroscopy. [b] The molar ratio of the total amount of epoxide products from both

(Z)- and (E)-epoxides of both the pyridyl and phenyl substrates divided by the total

amount of substrates consumed and multiplied by 100. The amount of epoxide was

determined at approximately the same total conversion of the substrates, 40%.

[c] Based on inherent or normalized/real values,^[39]see text. [d] Overlapping peaks pre-

cluded determination of epoxide distribution.

Substrate selectivity

The same substrate pairs (Scheme 3) that were inves-

tigated with the first generation, 1+2 (Figure 1), and

the second generation, 6+2 (Figure 3), of the supra-

molecular catalytic system were investigated with the

third-generation catalytic system, 6+12 (Figure 4),

involving substrates from both the styrene and the

stilbene series and of both Z and E configuration. For

(Z)-stilbene analogue 8a (Scheme 3) that was analyzed by

¹H NMR spectroscopy in [D₆]DMSO as a result of E/Z isomeriza-

tion caused by residual HCl in CDCl₃]. Owing to overlapping

signals, the substrate consumption could not be analyzed by

¹H NMR spectroscopy except in one case (see below). For the

same reason, the ratio of epoxide product versus other oxida-

solubility reasons, the highest concentration of strapped recep-

tor 12 was 12.5 mm compared to 15 mm for unstrapped recep-

tor 2.

In general, the first-generation system is not discussed be-

cause it generated much lower substrate selectivities than the

second- and third-generation systems.

1

tion products could only be determined by H NMR spectros-

copy in a few cases (Table 2). For the analysis of substrate con-

Catalyst part alone

1

sumption by GC and/or H NMR spectroscopy, a chromatogram

and a spectrum, respectively, was recorded before and after

the reaction. In the chromatogram and the spectrum, respec-

tively, the area of each peak/signal of the substrate was nor-

malized with the area of the internal standard, which allowed

the correct determination of substrate consumption by sub-

tracting the normalized area of each of the two substrates in

each reaction before and after the reaction.

First, the inherent reactivity difference between the substrates

with catalyst part 6 itself was determined. As can be seen in

Table 1, entry 1, the inherent substrate selectivity varied in

such a way that the closer the pyridyl-appended group was to

the alkene double bond in the substrate and the more pyridyl

groups attached, the lower the reactivity. Thus, the stilbene

series of substrates (Scheme 3) demonstrated the lowest reac-

tivity as seen in the lowest inherent selectivity values. On the

basis of this observation one might expect that the substrate

having the highest electron density on the double bond

would react the fastest with the electrophilic Mn=O species of

the catalyst part of the supramolecular system. In fact, the J-K

epoxidation partly involves a radical mechanism that results in

a CꢁC bond in the TS, and this explains the formation of both

(E)- and (Z)-epoxides from an alkene starting material with

a specific stereochemistry.^[32]However, the major mechanism

involves concerted addition of the double bond to the Mn=O

species,^[33]which leads to an epoxide product with conserved

stereochemistry. By running the competitive pairwise (1:1) re-

actions with the use of a congener of the J-K catalyst, com-

pound 28^[19](Figure 5), as the catalyst, inherent substrate selec-

tivities very close to 1 were observed (Table 1, entry 2); this

demonstrates that the electron density on the double bond of

the substrate had in reality very little influence on the inherent

substrate selectivity. Thus, one possible conclusion about the

inherent substrate selectivity of catalyst part 6 alone, which ex-

hibits much more pronounced inherent selectivities than 28 in

the competitive pairwise (1:1) epoxidations of the investigated

substrate pairs, is that 6 forms aggregates with pyridyl-ap-

pended substrates by (2-quinolidone)NꢁH···N(pyridine) hydro-

gen bonding. These aggregates might be able to exercise sub-

Validation of the GC methodology

For one substrate, (Z)-stilbene analogue 8a, its consumption

could be determined by both GC and ¹H NMR spectroscopy

(see the Supporting Information), as one proton resonance did

not overlap with any other signal. The difference in the so-de-

termined conversion of 8a was less than 3% between the two

methodologies, which thus validated the use of the GC meth-

odology.

Determination of the substrate selectivity

First, the inherent substrate selectivity^[20]was determined by

running the reaction with only the catalyst part. The inherent

substrate selectivity is a measure of the rate difference in the

epoxidation of the pairwise substrates caused by the differ-

ence in the electronic and steric interactions of each substrate

with the catalyst part alone in the TS. Then, the reaction was

run with both the catalyst part and the receptor part to obtain

the real substrate selectivity.^[20]However, a more accurate mea-

sure of the ability of the dynamic catalytic system to discrimi-

nate between the two substrates owing to the presence of the

receptor is the normalized substrate selectivity.^[20]The normal-

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&

7

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

found to be 2.1–3.2 (mean 2.5) by using the third-generation

dynamic catalytic system, 6+12; 1.5–3.4 (mean 2.6) by using

the second-generation system, 6+2; and 0.6–2.4 (mean 1.5)

by using the first-generation system, 1+2, for which the cata-

lyst concentration was 5 mm and the receptor concentration

was 12.5 mm for 12 and 15 mm for 6 and 2, respectively

(Table 1, entries, 4, 8, and 11). Thus, taking our first-generation

system as a reference point, it is clear that it is more important

for our dynamic supramolecular catalytic system to have the

catalyst part preorganized than the receptor part to obtain

high substrate selectivity. This can be explained by considering

that the strap on the catalytic part, in addition to preorganiz-

ing the catalyst part into a cisoid conformation to form a cyclic

cavity, also blocks the exo side of the self-assembled cavities

from participating in catalysis, and this hampers unselective

catalysis.

Figure 5. Congener 28 of the Jacobsen–Katsuki catalyst employed as a refer-

ence compound.

strate selectivity by being more reactive with one substrate

than the other. Notably, the influence of the geometry of the

double bond does not have such a large impact on the inher-

ent substrate selectivity of 6, as the substrate pairs of the

same chemical composition but of different geometrical iso-

mers, 4a/5a and 25a/26a (Scheme 3; Table 1, entry 1) on the

one hand and 8a/9a and 27a/11 a (Table 1, entry 1) on the

other hand, show very similar inherent substrate selectivities.

Owing to the complex nature of the system with many dif-

ferent supramolecular aggregates present, as seen in the equi-

librium analysis of 1+2,^[22]and for which the structure of each

species is not well defined, all the rationalizations must be

treated with caution. Nevertheless one attempt to explain the

observed substrate selectivities of the catalytic part alone will

be given: one conclusion about the aggregation of the first

generation of the supramolecular catalytic system is that

a large amount of its components exists as monomers and

linear dimers at the concentration at which the catalytic reac-

tions are run, for example at 5 mm for 6,^[22,19]At the same

time, there is a large excess amount of substrate relative to

that of the catalyst, and consequently a pyridyl-appended sub-

strate is always reversibly coordinated to the 2-quinolidone

The analyses below are based on the assumption that both

systems 6+12 and 6+2 have the same type of supramolec-

ular components as system 1+2. For system 1+2, the molar

ratio of noncyclic heterodimer, closed c-1·2 and linear l-1·2

(Scheme 1), increases upon the addition of an excess amount

of the receptor part on the basis of the equilibrium analysis of

system 1+2, whereas the molar ratio of cyclic heterodimer 3

is constant.^[22,19]At this stage, this is the best assumption that

can be done. Thus, all the conclusions must be treated with

caution. We can also assume that because of the double strap-

ping there is a higher molar ratio of the cyclic heterodimer in

system 6+12 than in either 6+2 or 1+2. Another assump-

tion is that c-1·2 is more substrate selective than l-1·2

(Scheme 1).

Stilbene series

We started the investigation of the substrate selectivity of the

moiety on each arm of catalyst 6 through (2-quinolidone)Nꢁ third-generation catalytic system, 6+12, by studying the com-

H···N(pyridine) hydrogen bonding between the pyridine unit of

the substrate and one of the 2-quinolidone units of catalyst

part 6. Consequently, the pyridyl-appended substrate mole-

cules will be somewhat delayed in their itinerary to the catalyt-

ic center of 6 relative to the phenyl-appended substrates.

Thus, the phenyl-appended substrates will react faster, and this

will result in a substrate selectivity (pyridyl- versus phenyl-ap-

pended substrates) of <1. The one exception is substrate 25a,

the most elongated of the substrates investigated, which re-

acted faster than phenyl congener 26a (Table 1, entry 1). This

result might be explained by the possibility that if 26a hydro-

gen bonds reversibly to one of the 2-quinolidone moieties of

catalyst 6 through (2-quinolidone)NꢁH···N(pyridine) hydrogen

bonding between the pyridine and the 2-quinolidone moieties,

the double bond of 26a is placed more or less in the vicinity

of the catalytic center of 6 relative to the vicinity of the other

substrates and the epoxidation thus occurs more swiftly.

petitive pairwise (1:1) epoxidation between the same two sub-

strate pairs that gave the best result with the second-genera-

tion catalytic system, 6+2, that is, (Z)-pyridyl 8a/(Z)-phenyl 9a

and (E)-pyridyl 10a/(E)-phenyl 11 a, both pairs in the stilbene

series of substrates (Scheme 3).

For (Z)-stilbene pair 8a/9a, the normalized substrate selec-

tivity was the same as that for systems 6+12 and 6+2, both

at equal ratios of the catalyst and receptor parts, namely, 2

(Table 1, entry 3 vs. 7), and with an excess amount of the re-

ceptor part, namely, 3 (Table 1, entry 4 vs. 8). Given that the

substrate selectivity increases in the same way for the two sys-

tems and that the absolute values are the same, this indicates

that for substrate pair 8a/9a, the substrate-selective catalytic

species is the same, that is, c-6·12 and c-6·2, respectively (see

above). As an approximation, the equilibrium analysis of

system 1+2 was used to rationalize this result.^[22]This analysis

showed that the molar ratio of cyclic cavity 3 was not affected

by the increase in the concentration of the receptor part but

that the molar ratio of the catalyst part bound to the receptor

part as the next neighbor in hetero-oligomers, was doubled.^[19]

At the same time, the degree of polymerization was low.^[22]

This indicates that the substrate-selective catalyst for 8a/9a

Catalyst part+receptor part

The normalized substrate selectivities (pyridyl- vs. phenyl-ap-

pended olefins) of the five substrate pairs in Table 1 were

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&

8

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

was the closed-heterodimer, c-6·12 in the third-generation

system and c-6·2 in the second-generation system (Scheme 1),

both of which generated the same substrate selectivity. It is

also likely that the substrate selectivity induced by c-6·12 and

c-6·2 is similar owing to their similar structures.

and/or there is also a contribution from cavity 13 in the third-

generation system.

Styrene series

For (E)-stilbene pair 10a/11 a, 11 a containing two pyridyl

groups, the normalized substrate selectivity showed abnormal

behavior in that higher normalized substrate selectivity was

obtained at a lower receptor concentration for 6+12 (Table 1,

entry 3 vs. 4), whereas system 6+2 displayed the common

higher normalized substrate selectivity at a higher concentra-

tion of receptor relative to that at a lower concentration

(Table 1, entry 8 versus 7). The abnormal result might be ex-

plained by considering that with an excess amount of receptor

part 12, strapped in a cisoid conformation, 12 is perfectly pre-

organized to form a strong aggregating cyclic trimer, cyclo-

(12)₃(Figure 6). According to the equilibrium analysis of

system 1+2,^[22]cyclic trimer cyclo-2₃is one of the supramolec-

ular aggregates in solution. Accordingly, there is also some

cyclo-2₃in excess amount of 2 also in system 6+2, but be-

cause 2 is not as preorganized as 12, it exists to a much small-

er extent than cyclo-(12)₃in system 6+12 (Figure 6). Now,

The normalized substrate selectivity of the third-generation

catalytic system, 6+12, was investigated in the competitive

pairwise (1:1) epoxidation of the styrene series of substrates.

For (Z)-styrene pair 4a/5a, the normalized substrate selectivity

was about the same as that for system 6+2 if equal amounts

of the catalyst part and receptor part were employed (Table 1,

entry 3 vs. 7), but the selectivity increased considerably for

system 6+12, relative to that for system 6+2, when the re-

ceptor part was in excess amount, from 2.0 to 2.9 for the

former system (Table 1, entries 3 and 4), but almost nothing for

the latter system, which stayed at 1.8–1.9 (Table 1, entries 7

and 8). This indicates that the substrate-selective catalyst for

4a/5a is closed-heterodimer c-6·12 in the third-generation

system on the basis of the same reasoning as that for sub-

strate pair 27a/11 a (see above). However, according to the

same analysis, also the second-generation system should result

in a higher substrate selectivity when the concentration of the

receptor part is increased. Thus, one conclusion could be that

there is a high content of c-6·2 in the system, but the TS is not

optimal for 4a. An alternative conclusion could be that the

substrate-selective catalyst is instead l-6·2, which generates

low selectivity owing to the large distance between the cata-

lyst and receptor parts (Scheme 1).

For the second substrate pair in the styrene series, E sub-

strates 25a/26a, the normalized substrate selectivity decreased

with higher preorganization of the catalytic system, for both

equal amounts of the catalyst part and receptor part as well as

for an excess amount of the receptor part. However, in both

systems 6+12 and 6+2 it increased when the receptor part

was in excess amount. This points toward that cavity c-6·12 in

the third-generation system and cavity c-6·2 in the second-

generation system are the substrate-selective catalysts, and the

resulting TSs are a little bit different between the cavities of

both generations; this explains the difference in substrate se-

lectivity. Alternatively, the substrate-selective species for this

substrate pair could be linear heterodimer l-6·12 for the third-

generation system and l-6·2 for the second-generation system.

This result seems plausible, as both 25a and 26b are the larg-

est substrates in this series, and as such, they could experience

difficulties in entering the proposed catalytic cavity c-6·12 in

the third-generation system and c-6·2 in the second-genera-

tion system.

Figure 6. Proposed residing state of (E)-dipyridylstilbene congener 10a

bonded inside cyclo-(12)₃in system 6+12 with an excess amount of 12.

substrate 10a might fit well into the cavity of cyclo-(12)₃, re-

sulting in complex cyclo-(12)₃·10a, which then might consti-

tute a resting state for 10a, and this would retard the rate at

which molecules of 10a can reach any of the substrate-selec-

tive catalytic species; this paves the way for the epoxidation of

phenyl-appended substrates 11 a (Scheme 3).

The third stilbene pair, E substrates 27a/11 a, showed

normal behavior in that the normalized substrate selectivity in-

creased upon going from system 6+2 to system 6+12

(Table 1, entry 3 vs. 7 and entry 4 vs. 8); the selectivity in-

creased from 1.6 to 2.2 if an excess amount of the receptor

part was present in system 6+12. This substrate selectivity is

higher than that for the earlier generations of the system and

increases as we have observed with an excess amount of the

receptor part. This indicates that the substrate-selective cata-

lyst for 27a/11 a is closed-heterodimer c-6·12 in the third-gen-

eration system and c-6·2 in the second-generation system

(Scheme 1) on the basis of the same reasoning as that for sub-

strate pair 8a/9a (see above) with one exception: the sub-

strate selectivities of c-6·12 and c-6·2 are somewhat different

Further analysis of the catalytic experiments

In general, owing to the complex nature of the system with

many different supramolecular aggregates present, as seen in

the equilibrium analysis of system 1+2,^[22,19]and for which in

addition the structure of each supramolecular species in the

system is not well defined, all rationalizations must be treated

with caution. However, now in the third-generation system,

6+12, for which both the catalyst and the receptor parts are

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&

9

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

forced into a cisoid conformation, postulated substrate-selec-

tive catalytic cavities 13 (Figure 4 and Scheme 1) and c-6·12

should be the major species in solution. Thanks to molecular

modeling (Figure 2), we have a good indication that 13 should

exert substrate selectivity in the epoxidations of at least some

of our designed substrates such as 4a versus 5a (Figure 2).

Moreover, compound 13 is the only substrate-selective species

for which we have some idea of its structure and conformation

and hence becomes the basis for the reasoning. In addition, it

seems logical that the conformation of substrate-selective

closed heterodimer c-6·12 should be similar to that of 13

(Scheme 1) but with only one pyridone hydrogen-bonding

motif in operation.

ation system, complete analysis of the system regarding all

possible equilibria in each of these systems must be conduct-

ed, similar to the one that was executed for the first-genera-

tion system.^[22,19]Of special importance in this context will be

to verify that in the third- and second-generation systems

there is a substantial amount of the cyclic catalytic cavities

present, as assumed in the preliminary conclusions in the pres-

ent work. However, it seems as though having a large amount

of the catalytic cavities present does not lead to high substrate

selectivity for all the investigated substrate pairs. However, the

catalytic cavities in the complicated equilibrium mixture are

the only species for which we can make some predictions

about the structure and thus a prediction about the TS of the

epoxidation reaction (Figure 2). Hence, continuing our efforts

to increase the amount of the cyclic heterodimer and the

closed heterodimer in a future fourth system is highly desira-

ble. A clear structural change would be to lock the free rota-

tion between the pyridone unit and the attached phenylene

unit of porphyrin receptor 12 (Figure 4). That would lead to

the loss of a smaller amount of conformational entropy upon

aggregation with the catalyst part and stronger bonding and

thus more of the desired cavity species. To evaluate if the

closed heterodimer has a strong influence on the substrate se-

lectivity, the catalyst and receptor parts should be modified to

contain only one hydrogen-bonding moiety per part, which

would thus ensure the formation of the closed heterodimer as

the only substrate-selective cavity compound.

Rewardingly, the substrate selectivity for the epoxidation of

(Z)-styrene pair 4a/5a, for which the first-generation dynamic

supramolecular system was designed with substrate-selective

catalytic cavity 3 as the proposed unique supramolecular spe-

cies in solution (Figure 2), was higher than that for correspond-

ing E substrates 25a/26a for system 6+12 (Table 1, entry 3

and entry 4), for which c-6·12 (Scheme 1), close in conforma-

tion to 13 (Figure 4 and Scheme 1), were the suggested sub-

strate-selective species. Simply changing the position of the

phenyl group of 4a to the “E position” (=compound 25a)

(using Figure 2 as a starting point) leads to a case in which the

phenyl group clashed with the interior of catalytic cavity c-

6·12, thus explaining the lower substrate selectivity of 25a/

26a compared to 4a/5a. Interestingly, in system 6+2, the

above-mentioned selectivity was reversed, in that substrate se-

lectivity for 25a/26a was higher than for 4a/5a in system 6+

12 (Table 1, entry 7 vs. 3 and entry 8 vs. 4), which indicates

that there are small differences in the TSs of the epoxidation

reaction involving the cavity compounds of both systems as

catalysts or that the linear heterodimers are the substrate-se-

lective catalysts (see above).

Real substrate selectivities

For practical uses of substrate-selective catalysts, it is more rel-

evant to look at the real substrate selectivity.^[20]In the third-

generation system, 6+12, the real substrate selectivity for pyr-

idyl-appended substrates was ꢂ1 in four of five cases com-

pared to three of five cases for the second-generation system,

6+2 (Table 1, entry 4 vs. 8). Notably, the obtained highest real

substrate selectivity was for substrate pair styrene 25a/26a,

which reached a real substrate selectivity of 3.6 for system 6+

2. However, one should remember that the inherent substrate

selectivity was positive in this case, 1.27 (Table 1, entry 1),

unique to the series of substrates in the studies, thus we start-

ed from a case in which the substrate selectivity was already in

favor of the pyridyl-appended substrate.

For the two substrate pairs in the stilbene series, for which

the pyridyl-appended substrate contains one pyridyl group,

the result was that the Z substrate pair, 8a/9a, showed

a higher substrate selectivity than the E substrate pair, 27a/

11 a, by using system 6+12 (Table 1, entries 3 and 4). The

same trend was also observed for system 6+2; however, it

was not as pronounced, which supports the view that there

are fewer substrate-selective species in the latter system than

in the former system and that the observed substrate selectivi-

ty arises from the closed heterodimer. Finally, the third-genera-

tion system was also better than the previous generations in

generating good substrate selectivity for 4a and a geometrical-

ly very different substrate, 27a (in competition with 11 a).

For the substrate pairs 8a/9a, 10a/11 a, and 25a/26a, the

resulting TSs in the third-generation system was not optimal

to obtain good substrate selectivity relative to that obtained

for the earlier generations, for which, presumably, the more

conformationally flexible closed heterodimer, c-6·2, was able to

generate higher substrate selectivities for these substrates

than c-6·12.

Inhibition studies

Adding a large amount of 4-ethylpyridine to the pairwise ep-

oxidation reaction of 27a versus 11 a by using system 6+12

did not affect the normalized substrate selectivity, as was the

case for systems 6+2 and 1+2 (Table 1, entries 6, 9, and 12),

for which the substrate selectivity decreased. This observation

might be explained by the fact that this large excess amount

of Lewis base will occupy all of the available Zn^IIsites in the re-

ceptor part of systems 6+2 and 1+2, which would make the

pyridyl-appended substrates competing with 4-ethylpyridine

for the binding sites. However, if both the catalytic part and

the receptor part are strapped, a high amount of cavities of

To make stronger statements about the origin of the sub-

strate selectivity of the second- and third-generation systems

and to suggest improvements in the design for a fourth-gener-

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&10

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

c-6·12 and 13 are formed that hampers binding of 4-ethylpyri-

dine on the exo side of 12 in c-6·12 and 13, which makes it

easier for the pyridyl-appended substrate to coordinate on the

endo side because no 4-ethylpyridine, residing on the exo side,

needs to be substituted to reach the square pyramidal coordi-

nation mode between the Zn^II–porphyrin and the pyridyl-ap-

pended substrate. Specifically, this result is in agreement with

the proposition above that for substrate pair 27a/11 a in the

third-generation system, the substrate selective species is

c-6·12.

Product selectivity

Although the study of the product selectivity^[39]was not the

major issue in this investigation, some information can be ob-

1

tained from the H NMR spectra originally intended to be used

to determine the conversions of the substrates (see the Experi-

mental Section and the Supporting Information), but overlap-

ping proton resonances in the spectra made it difficult to use

¹H NMR spectroscopy for this purpose. From the studies of the

J-K epoxidation of olefins, it is known that besides epoxides,

other oxidation products are also formed such as aldehydes

and alcohols.^[40]For systems 1+2 and 6+2, a rough estima-

tion of the amount of epoxide formed was 70%.^[19,21,23]Owing

to the difficult assignment and overlapping proton resonances

Turnover number (TON)

1

in the H NMR spectra, we were only able to determine the

Despite the success of the J-K epoxidation, the overall utility of

this catalyst is limited by its facile deactivation,^[34]which is

caused by the formation of Mn^IV–m-oxo dimers^[35]and irreversi-

ble ligand oxidation.^[36]Thus, it is of interest to find out how

the addition of receptor 12 to catalyst 6 affects the TON. Thus,

first catalyst part 6 (5 mm), pyridyl-appended substrate 27a

(20 equiv.), and PhIO (1.5 equiv.) were stirred at room tempera-

ture for 1 week in CH₂Cl₂to ensure that the reaction had

reached completion (see the Experimental Section for details).

After 1 week, the TON was determined to 9. Repeating the re-

action in the presence of receptor part 12 (5 mm) increased

the TON to 13. Thus, 12 provided some shelter for catalyst part

6. The observed low TON for system 6+12 should be regard-

ed in perspective that substrate 27a is a large substrate that is

sterically encumbered and possesses a coordinating functional-

ity. A similar approach was already taken by Nguyen and

Hupp; they added Zn^II–TPP (TPP=tetraphenylporphyrin) to

a double 4-pyridyl-appended Mn^III–salen complex, which re-

sulted in a rather open Mn catalyst with two big side walls.

The TON increased from 51 to 92 by adding Zn^II–TTP and by

using the standard compound, styrene, as the substrate.^[34]In

comparison, the TON in the epoxidation of styrene with PhIO

in CH₂Cl₂by using the catalyst part of the first-generation

system, compound 1 (1 mm), and an equimolar amount of pyr-

idine N-oxide was determined to be 149.^[19]By adding a four-

fold excess amount of receptor 2 to the original mixture of

1 and pyridine N-oxide, reaching the composition of the first

generation of our supramolecular catalytic system, the TON

was raised to 162.^[19]A further comparison was provided by

a congener of J-K catalyst, compound 28, and the TON was de-

termined to be 37 in the epoxidation of styrene^[34]but in the

absence of pyridine N-oxide. The reason for the lower TON in

the case of 6+12 might be that E alkenes (27a is an E alkene)

are epoxidized more slowly than Z alkenes,^[37]and during this

time the catalyst is deactivated, which results in a small

amount of consumed E alkene. Nevertheless, for both catalysts

6 and 1, the addition of a “cap”, receptor parts 12 and 2, re-

spectively, protected the catalytic Mn site from forming catalyt-

ically inactive Mn^IV–m-oxo dimers as was, for instance, demon-

strated by Hupp in epoxidations of alkenes by using self-as-

sembled capped Zn^II–porphyrin–Re molecular squares as cata-

lysts.^[38]

total formation of epoxide and the distribution, pyridyl/phenyl

and (Z)-pyridyl/(E)-pyridyl, respectively, for a few substrate pairs

when 6+12 was used as the catalytic system (Table 2, see the

Experimental Section and the Supporting Information for de-

tails). Fortunately, in one case, a comparison between the ep-

oxidation involving only the catalyst part 6, and the combined

catalyst–receptor system 6+12, could be made, namely, for

the Z substrates in the stilbene series, 8a/9a. For this system,

as seen in Table 2, entries 1 and 2, the total amount of epoxide

relative to that of the other products was substantially in-

creased in the system containing both the catalyst part and

the receptor part, 6+12, in comparison to the case in which 6

alone was employed in CH₂Cl₂(5 mm of each components);

this indicates that with the receptor present, the catalytic

system is more specific to the formation of epoxide than to

the formation of other products or that the system consisting

of a catalyst part and a receptor part simply experiences

a longer lifetime so that more epoxide is formed. The system

was not analyzed for other products so we do not know which

case is the correct one. On the basis of this one unique obser-

vation, it is of course impossible to make generalizations for

the other substrates. As seen in the same entries, a radical

mechanism is most probably involved to a large extent on the

basis of the high degree of formation of both the Z and E dia-

stereoisomers of the pyridyl-appended epoxide, (Z)-8b/(E)-

8b(27b), by employing stilbene 8a (Z diastereoisomer) as the

starting material. The observed Z/E ratio is not affected by the

presence or absence of the receptor part, as seen in Table 2,

entries 1 and 2. The other entries in Table 2 are included just

to report the full study by including the cases for which prod-

uct-selectivity values were possible to extract by analyzing the

reaction mixtures of the competitive epoxidations in Scheme 3

1

by H NMR spectroscopy.

Conclusions

The purpose of the present work was to increase the substrate

selectivity of a dynamic supramolecular catalytic system for the

epoxidation of pyridyl- versus phenyl-appended substrates by

increasing the preorganization of the catalyst part and recep-

tor part by having each part into a cisoid conformation in an

attempt to increase the amount of the proposed substrate-se-

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&

11

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

tion was used. PhIO was prepared by the hydrolysis of commercial-

ly available diacetoxyiodobenzene following a literature proce-

dure.^[41]The glassware used for anhydrous conditions was dried in

an oven for 24 h at 1408C. Column chromatography was per-

formed with Acros Organics (40–60 mm, 60 A) silica. TLC analyses

(Merck 60 F254 sheets) were visualized under UV light (254 or

366 nm). Catalytic reaction mixtures were filtered through an Acro-

disc CR 13 mm syringe filter with a 0.2 mm polytetrafluoroethylene

membrane. GC was performed with a Hewlett Packard (5890 A) in-

strument (Agilent Technologies, Santa Clara, CA, USA). The follow-

ing protocol was used: VA-1 column, 30 mꢅ0.25 mmꢅ0.25 mm

(Varian Chromatography System, Walnut Creek, CA, USA), helium as

carrier gas, flow gradient was from 0.5 to 3 mLmin^ꢁ1with a ramp

of 0.1 mLmin^ꢁ1, oven the temperature from 140 to 1858C with

lective cavity. The present system, the third generation, was

compared to the two previous less preorganized generations

of the catalytic system.

Specifically, the third-generation system showed results that

might be explained by the formation of a high amount of the

substrate-selective cavities, that is, designed cyclic heterodimer

13 and, maybe above all, closed heterodimer c-6·12

(Scheme 1). Thus, the pyridyl-appended substrate that was de-

signed to best fit the catalytic cavity of the designed catalytic

cyclic heterodimers, (Z)-styrene 4a, gave a higher substrate se-

lectivity over phenyl-appended congener 5a than the earlier

less preorganized systems. This was also true for geometrically

very different 27a (in competition with 11 a). However, for the

three other substrates, the resulting transition states of each of

the reactions catalyzed by the cyclic heterodimers of the third-

generation system were not optimal to obtain good substrate

selectivity relative to that obtained with earlier generations.

Thus, the overall conclusion about the substrate selectivity of

the dynamic supramolecular system with respect to the stud-

ied epoxidation reaction and the given substrate pairs is that

each substrate pair needs a specific level of preorganization of

the components to give optimal substrate selectivity. However,

the results show that to obtain good substrate selectivity it is

better to have the catalyst part preorganized than the receptor

part. This can be explained by considering that the strap, in

addition to preorganizing the catalyst part into a cisoid confor-

mation to form cyclic cavities, also blocks the exo side of the

self-assembled cavities from participating in catalysis, which

thus hampers unselective catalysis.

a ramp of 28Cmin^ꢁ1, then to 2208C with a ramp of 108Cmin^ꢁ1

,

then hold for 6 min. NMR spectra were recorded with a Bruker

DRX400 NMR spectrometer; ¹H NMR spectra were recorded at

400 MHz and ¹³C NMR spectra were recorded at 100 MHz in CDCl₃

or [D₆]DMSO. Chemical shifts are reported in ppm relative to an in-

ternal standard of residual chloroform (d=7.26 ppm for ¹H NMR;

d=77.16 ppm for ¹³C NMR) and residual DMSO (d=2.50 ppm for

¹H NMR). The NMR assignment of the synthesized molecules was

performed with the aid of coupling constants and 2D correlation

NMR experiments. HRMS was performed with a Waters micromass

Q-Tof instrument. A. Kolbe, Mikroanalytisches Laboratorium, Ger-

many, performed the elemental analyses.

Syntheses

4-Decyloxy-2,6-dimethylbenzaldehyde N-tosyl imine (16): p-Tolu-

enesulfonamide (30.9 g, 0.180 mol) and p-toluenesulfonic acid

(1.5 mmol) were added to a solution of 15^[19](8.7 g, 0.030 mol) in

toluene (290 mL). The mixture was heated at reflux under Dean–

Stark conditions overnight. The mixture was cooled to RT before

water was added. The formed precipitate was filtered off, and the

aqueous phase was extracted with toluene. The combined organic

layer was washed with water and brine, dried with magnesium sul-

fate, and evaporated under reduced pressure. Heptane was added

to the crude product, which was collected as an orange oil, and

the crystallization of 16 started immediately at RT. The mixture was

placed in the fridge for a few hours and was then collected on

a filter to afford the title compound (9.6 g 70%) as white crystals.

M.p. 42.1–42.98C; ¹H NMR (400 MHz, CDCl₃): d=9.41 (s, 1H), 7.90

To obtain firmer statements about the origin of the sub-

strate selectivity of the second- and third-generation systems

and to suggest improvements in the design for a fourth-gener-

ation system, complete analysis of the system regarding all

possible equilibria in each of these systems has to be conduct-

ed, just like the one that was executed for the first-generation

system.^[19,22]Such analyses are currently underway in our labo-

ratory.

Nevertheless, we showed that the present dynamic supra-

molecular catalytic systems can generate substrate selectivity

and turnover numbers in an important organic transformation

involving structurally advanced substrate pairs. Our results

might have bearing on the construction of other supramolec-

ular catalysts and is a contribution to the ongoing discussion

about the need for preorganization in supramolecular catalytic

systems.^[8]

3

(d, J_H,H=8.40 Hz, 2H), 7.34 (d, J_H,H=8.00 Hz, 2H), 6.63 (s, 2H), 4.01

3

(t, J_H,H=6.80 Hz, 2H), 2.62 (s, 6H), 2.44 (s, 3H), 1.82–1.75 (m, 2H),

1.44–1.25 (m, 14H), 0.91–0.87 ppm (m, 3H); ¹³C NMR (100 MHz,

CDCl₃) 168.03, 163.20, 146.24, 143.89, 136.51, 129.94, 129.65,

127.65, 127.33, 121.55, 115.67, 68.16, 31.89, 29.54, 29.31, 29.04,

26.08, 25.93, 22.84, 22.68, 22.50, 21.62, 14.12 ppm; HRMS (FAB):

m/z: calcd. 444.2494 [M+1]⁺; found: 444.2564; elemental analysis

calcd (%) for C₂₆H₃₇NO₃S (443.64): C 70.39, H 8.41, N 3.16; found: C

70.68, H 8.06, N 3.08.

Experimental Section

5-(4-Decyloxy-2,6-dimethylphenyl)dipyrromethane (14): Cu(OTf)₂

(0.904 g, 2.50 mmol) was added to

a solution of 16 (9.5 g,

General

20.6 mmol) in freshly distilled pyrrole (34 mL, 490 mmol). The solu-

tion turned black immediately. The mixture was stirred at RT over-

night. The solution was filtered through a pad of silica, eluting

with EtOAc. Pyrrole was removed by bulb-to-bulb distillation to

collect a black solid crude product. The crude product was purified

by chromatography (PE to PE/EtOAc=9:1; h=10 cm, d=5 cm). A

final crystallization (PE/EtOAc=19:1) at ꢁ208C furnished the title

compound (1.4 g, 17%) as an off-white solid. R_f=0.59 (PE/EtOAc=

8:2). The spectroscopic data of the title compound were in agree-

All chemicals were used as received, unless otherwise stated. CuI

was recrystallized from saturated aqueous potassium iodide and

then dried for 1 week under vacuum (0.4 mbar) at 808C prior to

use. CH₂Cl₂was distilled from CaH₂and stored over 4 ꢄ molecular

sieves. Pyrrole was distilled from CaH₂at reduced pressure. Dry 1,4-

dioxane was purchased from Acros Organics. Toluene was dried by

distillation over sodium/benzophenone ketyl prior to use. Acetone

was dried over Na₂SO₄. For petroleum ether (PE), the 40–608C frac-

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&12

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

ment with the published ones;^[19]elemental analysis calcd (%) for

C₂₇H₃₈N₂O (406.60): C 79.76, H 9.42, N 6.89; found: C 79.66, H 9.09,

N 6.87.

2H), 4.07–4.04 (m, 8H), 1.88–1.82 (m, 8H), 1.57–1.25 (m, 82H),

0.87 ppm (t, ³J_H,H=6.6 Hz, 6H); ¹³C NMR (100 MHz, CDCl₃): d=

187.23, 163.87, 163.49, 138.97, 120.39, 96.35, 76.33, 32.07, 29.85,

29.81, 29.72, 29.68, 29.59, 29.51, 29.40, 29.11, 29.01, 26.15, 22.84,

14.28 ppm; HRMS (ESI): m/z: calcd for C₆₅H₁₁₀I₂O₆: 1241.6470

[M+H]⁺; found: 1241.6537; elemental analysis calcd (%) for

C₆₅H₁₁₀I₂O₆(1241.38): C 62.89, H 8.93; found: C 62.66, H 9.03.

2-Hydroxy-4-(icosyloxy)benzaldehyde

(19):

K₂CO₃

(2.8 g,

20 mmol) and KI (323 mg, 2.0 mmol) were added to a solution of

18 (2.7 g, 20 mmol) and 1-bromoicosane (7.2 g, 20 mmol) in dry

acetone under an atmosphere of N₂. The resulting mixture was

heated at reflux for 40 h. After this time, the mixture was filtered

while warm through a filter paper, and the collected filtrate was

concentrated under reduced pressure. CH₂Cl₂(100 mL) and aque-

ous HCl (1m, 100 mL) were added. The phases were separated,

and the aqueous layer was extracted with CH₂Cl₂(100 mL). The

combined organic layer was concentrated to dryness under re-

duced pressure, and the furnished concentrate was purified by

column chromatography (CHCl₃/PE=1:8 to 1:7 to 1:4; h=15 cm,

d=5 cm) to furnish the title compound (5.3 g, 63%) as a white

6,6’-[Undecane-1,11-diylbis(oxy)]bis{3-[2-(benzyloxy)pyridin-4-

yl]-4-(icosyloxy)benzaldehyde} (17): Diodo compound 21 (1.0 g,

0.81 mmol) and stannate 22^[19](1.5 g, 3.2 mmol) were suspended in

PhMe (30 mL), and then the mixture was evaporated to dryness.

The residue was dried under vacuum (0.4 mbar) overnight. CsF

(539 mg, 3.55 mmol), CuI (307 mg, 1.61 mmol), and dry 1,4-diox-

ane/PhMe (2:1, 48 mL) were added to the solid residue, and the re-

sulting suspension was degassed and a N₂atmosphere was intro-

duced. Then, PCy₃(452 mg, 1.61 mmol) and Pd₂(dba)₃(147.3 mg,

0.161 mmol) were added under a N₂atmosphere. The mixture was

heated to 1008C for 72 h under a N₂atmosphere. The mixture was

allowed to reach room temperature, and the solvent was removed

under reduced pressure. The furnished crude product was purified

by column chromatography (Et₂O/PE=1:4 to 2:5 to 3:7; h=12 cm,

d=5 cm) to give the title compound (420 mg, 42%) as a white

solid. R_f=0.26 (CH₂Cl₂/Et₂O=39:1); ¹H NMR (400 MHz, CDCl₃): d=

1

solid. R_f=0.35 (EtOAc/hexane=1:3); H NMR (400 MHz, CDCl₃): d=

11.48 (s, 1H), 9.70 (s, 1H), 7.41 (d, ³J_H,H=8.8 Hz, 1H), 6.52 (dd,

3

4

⁴J_H,H=2.3 Hz, J_H,H=8.8 Hz, 1H), 6.41 (d, J_H,H=2.3 Hz, 1H), 4.00 (t,

³J_H,H=6.6 Hz, 2H), 1.82–1.75 (m, 2H), 1.48–1.40 (m, 2H), 1.33–1.25

(m, 32H), 0.88 ppm (t, J_H,H=6.7 Hz, 3H); ¹³C NMR (100 MHz, CDCl₃):

3

d=194.41, 166.60, 164.67, 135.31, 115.13, 108.92, 101.17, 68.74,

32.07, 29.85, 29.82, 29.81, 29.79, 29.72, 29.68, 29.51, 29.45, 29.06,

26.06, 22.84, 14.26 ppm; HRMS (ESI): m/z: calcd for C₂₇H₄₆O₃:

441.3345 [M+Na]⁺; found: 441.3334; elemental analysis calcd (%)

for C₂₇H₄₆O₃(418.65): C 77.46, H 11.07; found: C 77.41, H 11.05.

3

10.37 (s, 2H), 8.16 (d, J_H,H=5.5 Hz, 2H), 7.89 (s, 2H), 7.49–7.47 (m,

4H), 7.39–7.36 (m, 4H), 7.33–7.29 (m, 2H), 7.07 (dd, ⁴J_H,H=1.4 Hz,

³J_H,H=5.4 Hz, 2H), 7.00 (brs, 2H), 6.49 (s, 2H), 5.42 (s, 4H), 4.12 (t,

³J_H,H=6.2 Hz, 4H), 4.06 (t, ³J_H,H=6.6 Hz, 4H), 1.91–1.85 (m, 4H),

1.82–1.75 (m, 4H), 1.54–1.21 (m, 82H), 0.88 ppm (t, ³J_H,H=6.7 Hz,

6H); ¹³C NMR (100 MHz, CDCl₃): d=188.02, 163.96, 163.82, 162.66,

147.89, 146.18, 137.63, 130.54, 128.48, 127.84, 127.76, 120.90,

118.51, 118.06, 111.39, 96.17, 69.01, 68.89, 67.62, 32.03, 29.81, 29.76,

29.69, 29.62, 29.60, 29.52, 29.47, 29.40, 29.35, 29.14, 28.95, 26.17,

22.79, 14.23 ppm; HRMS (ESI): m/z: calcd for C₈₉H₁₃₀N₂O₈:

1355.9905 [M+H]⁺; found: 1355.9934; elemental analysis calcd (%)

for C₈₉H₁₃₀N₂O₈(1356.00): C 78.83, H 9.66, N 2.07; found: C 78.84, H

9.65, N 2.06.

2,2’-[Undecane-1,11-diylbis(oxy)]bis[4-(icosyloxy)benzaldehyde]

(20): 1,11-Dibromoundecane (1.32 mL, 5.65 mmol) was added to

a

suspension of 19 (4.73 g, 11.3 mmol) and K₂CO₃(3.9 g,

11.3 mmol) in anhydrous DMF (300 mL) at RT under an atmosphere

of N₂. After the addition, the temperature was increased to 808C,

and the mixture was stirred for 10 h. Then, the mixture was al-

lowed to reach RT before H₂O (400 mL) and CH₂Cl₂(400 mL) were

added. The two phases were separated, and the aqueous layer was

extracted with CH₂Cl₂(2ꢅ400 mL). The combined organic layer

was washed with brine (2ꢅ400 mL), dried (MgSO₄), filtered, and

concentrated under reduced pressure. Purification of the residue

by column chromatography (CHCl₃/PE=1:3 to 1:1 to 3:1 to 1:0;

h=14 cm, d=5 cm) gave the title compound (5.3 g, 95%) as

Porphyrin 23: A solution of TFA (32 mg, 281 mmol) in CH₂Cl₂

(0.2 mL) was added to a solution of 17 (95.2 mg, 70.2 mmol) and di-

pyrromethane 14 (56.9 mg, 140 mmol) in CH₂Cl₂(28 mL) under a N₂

atmosphere at room temperature in a 50 mL round-bottomed flask

covered with aluminum foil. The mixture was stirred for 55 min.

After this time, DDQ (47.8 mg, 211 mmol) was added, and the mix-

ture was stirred for an additional hour. Then, the solvent was re-

moved under reduced pressure, and the residue was purified by

silica column chromatography (Et₂O/hexane=1:19 to 3:17; h=

13 cm, d=2 cm) to furnish the title compound (32.9 mg, 22%) as

1

a white solid. R_f=0.45 (EtOAc/PE=1:9); H NMR (400 MHz, CDCl₃):

3

4

d=10.32 (s, 2H), 7.78 (d, J_H,H=8.7 Hz, 2H), 6.50 (dd, J_H,H=2.1 Hz,

³J_H,H=8.7 Hz, 2H), 6.42 (d, ⁴J_H,H=2.1 Hz, 2H), 4.04–3.98 (m, 8H),

1.86–1.72 (m, 8H), 1.51–1.20 (m, 82H), 0.87 ppm (t, ³J_H,H=7.0 Hz,

6H); ¹³C NMR (100 MHz, CDCl₃): d=188.49, 165.90, 163.49, 130.27,

119.02, 106.31, 99.04, 68.57, 68.53, 32.05, 32.01, 29.83, 29.79, 29.72,

29.68, 29.63, 29.58, 29.50, 29.48, 29.42, 29.22, 29.14, 26.18, 26.09,

22.82, 22.82, 14.25 ppm; HRMS (ESI): m/z: calcd for C₆₅H₁₁₂O₆:

989.8537 [M+H]⁺; found: 989.8538; elemental analysis calcd (%)

for C₆₅H₁₁₂O₆(989.58): C 78.89, H 11.41; found: C 78.91, H 11.39.

1

a red solid. R_f=0.37 (Et₂O/hexane=1:3); H NMR (400 MHz, CDCl₃):

d=8.80 (d, ³J_H,H=4.8 Hz, 4H), 8.65 (d, ³J_H,H=4.8 Hz, 4H), 8.18 (d,

3

³J_H,H=5.6 Hz, 2H), 8.14 (s, 2H), 7.45–7.43 (m, 4H), 7.39 (d, J_H,H

=

6,6’-[Undecane-1,11-diylbis(oxy)]bis[4-(icosyloxy)-3-iodobenzal-

dehyde] (21): TFA (1.11 mL, 14.5 mmol) was added dropwise to

a solution of 20 (4.77 g, 4.82 mmol) and NIS (2.43 g, 10.6 mmol) in

CH₂Cl₂(175 mL) under an atmosphere of N₂at RT. The resulting

mixture was stirred for 20 h. After this time, the reaction was

quenched with aqueous Na₂S₂O₃(10%w/v, 150 mL). The phases

were separated, and the aqueous layer was extracted with CH₂Cl₂

(2ꢅ150 mL). The combined organic phase was dried (MgSO₄), fil-

tered, and concentrated to dryness under reduced pressure. The

obtained residue was purified by silica column chromatography

(CHCl₃/PE=1:1 to 3:1; h=8 cm, d=10 cm) to furnish the title com-

pound (4.9 g, 82%) as a white solid. R_f=0.53 (Et₂O/PE=3:7);

¹H NMR (400 MHz, CDCl₃): d=10.21 (s, 2H), 8.20 (s, 2H), 6.34 (s,

5.6 Hz, 2H), 7.34–7.24 (m, 8H), 7.01 (s, 4H), 6.97 (s, 2H), 5.40 (s,

4H), 4.28 (t, ³J_H,H=6.4 Hz, 4H), 4.23 (t, ³J_H,H=6.3 Hz, 4H), 3.88 (t,

³J_H,H=5.0 Hz, 4H), 1.99–1.85 (m, 8H), 1.88 (s, 3H), 1.85 (s, 3H),

1.65–1.27 (m, 102H), 0.94–0.87 (m, 16H), ꢁ0.25 to ꢁ0.31 (m, 4H),

ꢁ1.14 to ꢁ1.22 (m, 4H), ꢁ1.84 to ꢁ1.91 (m, 4H), ꢁ2.52 (brs, 2H),

ꢁ2.72 to ꢁ2.76 ppm (m, 2H); ¹³C NMR (100 MHz, CDCl₃): d=

163.94, 160.91, 158.96, 157.82, 149.52, 145.92, 141.12, 140.76,

137.56, 136.67, 134.15, 128.52, 127.97, 127.82, 125.17, 119.51,

118.67, 117.43, 114.70, 112.95, 112.76, 111.64, 99.33, 70.77, 69.14,

68.20, 67.87, 32.08, 31.11, 29.89–29.80 (overlapping methylene

carbon resonances), 26.77 26.60, 26.49, 26.43, 25.51, 24.87, 22.88,

22.85, 22.25, 22.05, 14.31, 14.28 ppm; HRMS (ESI): m/z: calcd for

C₁₄₃H₁₉₆N₆O₈: 1064.2652 [M+2H]²⁺; found: 1064.2638; elemental

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&13

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

analysis calcd (%) for C₁₄₃H₁₉₆N₆O₈(2127.13): C 80.74, H 9.29, N

3.95; found: C 80.54, H 9.41, N 3.99.

stock solution. Then, stock solution (250 mL) and CH₂Cl₂(350 mL)

were added to each vial by means of a 250 mL Hamilton syringe.

An aliquot (250 mL) of the stock solution was kept as a blank for

the NMR and GC analyses. After the addition, the vials were sub-

jected to ultrasonication until a clear homogeneous solution was

obtained. Another set of five 2 mL glass vials were loaded with

PhIO (24 mmol). The solutions were transferred to the vials contain-

ing PhIO by means of a cannula. The obtained mixtures were

stirred for 24 h at RT. After this time, each reaction mixture was

loaded on a silica pad (h=2.5 cm, d=0.5 cm) by means of a Pas-

teur pipette and eluted with EtOAc (50 mL) with the exception of

substrate pair 10a/11 a, which was eluted with EtOAc/MeOH (5:1,

50 mL). After removal of the solvent under reduced pressure, the

concentrate was dissolved in Et₂O (3 mL), in which the catalyst was

insoluble, and filtered through a syringe filter to remove the resid-

ual catalyst. Then, an aliquot was taken for GC analysis and the re-

maining Et₂O was removed under reduced pressure. The so-ob-

tained residue was dissolved in CDCl₃(0.7 mL) for analysis by NMR

spectroscopy with the exception of substrate pair 8a/9a, which

was dissolved in [D₆]DMSO (0.7 mL) prior to analysis by NMR spec-

troscopy. To determine the substrate selectivity, GC analysis was

performed. Each injection for each sample (both for the blank and

the mixture after reaction) was repeated twice. t_R=11.0 (9a), 13.6

(8a), 16.8 (11 a), 18.4 (5a), 18.5 (IS), 19.7 (27a), 21.1 (26a), 22.2

(4a), 22.4 (10a), 25.0 min (25a).

Zn^II–porphyrin 24: A solution containing 23 (28 mg, 13.2 mmol)

and Zn(OAc)₂·2H₂O (17.3 mg, mmol) in CHCl₃/MeOH (19:1, 30 mL)

under a N₂atmosphere was stirred at 648C for 4 h. The mixture

was allowed to reach RT before the solvent was removed under re-

duced pressure. The residue was purified by silica column chroma-

tography (Et₂O/PE=1:3) to furnish the title compound (28.3 mg,

98%) as

a

red solid. R_f=0.39 (Et₂O/hexane=1:3); ¹H NMR

3

(400 MHz, CDCl₃): d=8.90 (d, ³J_H,H=4.7 Hz, 4H), 8.75 (d, J_H,H

=

4.7 Hz, 4H), 8.17 (s, 2H), 8.16 (d, ³J_H,H=5.4 Hz, 2H), 7.43–7.39 (m,

6H), 7.32–7.22 (m, 8H), 7.00 (s, 4H), 6.96 (s, 2H), 5.39 (s, 4H), 4.28

3

(t, J_H,H=5.3 Hz, 4H), 4.23 (t, J_H,H=5.6 Hz, 4H), 3.88 (t, J_H,H=5.5 Hz,

4H), 2.00–1.93 (m, 8H), 1.84 (s, 12H), 1.66–1.26 (m, 96H), 0.93–0.86

(m, 16H), ꢁ0.27 to ꢁ0.34 (m, 4H), ꢁ1.17 to ꢁ1.24 (m, 4H), ꢁ1.80

to ꢁ1.88 (m, 4H), ꢁ2.77 to ꢁ2.80 ppm (m, 2H); ¹³C NMR (100 MHz,

CDCl₃): d=164.07, 160.83, 158.76, 157.65, 150.78, 150.21, 149.28,

146.05, 140.95, 140.60, 137.70, 136.61, 134.73, 132.07, 130.70,

128.49, 127.91, 127.75, 125.82, 119,39, 118.64, 118.50, 115.77,

112.87, 112.68, 111.56, 99.23, 70.70, 69.12, 68.18, 67.65, 32.08, 30.11,

29.89–29.52 (overlapping methylene carbon resonances), 28.00,

26.78, 26.72, 26.50, 26.44, 26.00, 24.85, 22.88, 22.85, 22.23, 22.09,

14.31, 14.28 ppm; HRMS (ESI+): m/z: calcd for C₁₄₃H₁₉₄N₆O₈Zn:

1094.7198 [M+2H]²⁺; found: 1094.7191.

Zn^II–porphyrin 12, the receptor part: Pd/C (70 mg, 10%w/v) was

added to a degassed solution of 24 (48 mg, 21.9 mmol) and AcOH

(50 mL) in EtOAc (60 mL) under an atmosphere of N₂. The mixture

was degassed, and a N₂atmosphere was introduced. The mixture

was degassed one more time, and an atmospheric pressure of H₂

was introduced by using a balloon. After 12 h, the solvent was re-

moved under reduced pressure, and the residue was purified by

very fast (ꢃ20 min) silica column chromatography (CH₂Cl₂/

MeOH=39:1 to 19:1; h=12 cm, d=2 cm) to furnish the title com-

pound (33.3 mg, 76%) as a pink solid. The title compound consti-

tuted the last pink band that eluted from the column. R_f=0.25

Determination of the substrate selectivity by GC

The conversion (C) of each substrate was calculated from the chro-

matograms by normalizing (N) the substrate (S) area (A) with the

area of the IS both before the reaction (BR), NA_S,BR=A_S,BR/A_IS,BR, and

after the reaction (AR), NA_S,AR=A_S,AR/A_IS,AR, and then subtracting the

normalized substrate area after the reaction from the normalized

substrate area before the reaction, and finally dividing the differ-

ence by the substrate area before the reaction, thus C_S=

(NA_S,BRꢁNA_S,AR)/NA_S,BR. The inherent substrate selectivity (ISS) and

real substrate selectivity (RS), obtained in the absence and in the

presence of receptor part, respectively, were calculated by dividing

the conversion of the pyridine substrate (N) with the conversion of

the competing substrate lacking a nitrogen atom (C); thus ISS and

RS=C_s,N/C_S,C. The normalized substrate selectivity (NS) was finally

obtained by dividing RS by ISS (NS=RS/ISS). Further details for the

calculations of substrate selectivity are presented in the Supporting

Information.

1

(CH₂Cl₂/MeOH=19:1); H NMR (400 MHz, c=4.15 mgmL^ꢁ1in CDCl₃

containing 2.4%v/v [D₅]pyridine): d=8.74 (d, ³J_H,H=4.7 Hz, 4H),

3

8.66 (d, J_H,H=4.7 Hz, 4H), 7.99 (brs, 2H), 7.24–7.21 (m, 4H), 6.94–

3

6.93 (m, 6H), 6.88 (s, 2H), 6.75–6.73 (m, 2H), 4.25 (t, J_H,H=6.2 Hz,

3

4H), 4.16 (t, J_H,H=6.6 Hz, 4H), 3.82 (t, J_H,H=4.5 Hz, 4H), 1.96–1.90

(m, 8H), 1.76 (s, 3H), 1.71 (s, 3H), 1.59–1.23 (m, 100), 0.87–0.82 (m,

16H), ꢁ0.22 to ꢁ0.29 (m, 4H), ꢁ1.08 to ꢁ1.15 (m, 4H), ꢁ1.88 to

ꢁ1.99 (m, 4H), ꢁ2.48 to ꢁ2.57 ppm (m, 2H), the NH peak of the

pyridone moieties appeared at 11.89 ppm at c=6 mgmL^ꢁ1in

CDCl₃containing 6%v/v of [D₅]pyridine; HRMS (ESI): m/z: calcd for

C₁₂₉H₁₈₂N₆O₈Zn: 2008.3389 [M+H]⁺, found: 2008.3418; elemental

analysis calcd (%) for C₁₂₉H₁₈₂N₆O₈Zn (2010.27): C 77.07, H 9.13, N

4.18; found: C 77.23, H 9.19, N 4.06.

1

Determination of epoxide formation by H NMR spectroscopy

1

A H NMR spectrum was recorded before the reaction of the same

stock solution used for the reaction. After the reaction, another

¹H NMR spectrum was recorded. The signal for the benzylic pro-

tons of the IS at d=5.39 ppm was integrated to 1 in the spectra

recorded both before and after reaction. Estimation of the amount

of consumed substrate (S) that had been converted into epoxide

(E) was obtained from the formula: A_E,AR/(A_S,BRꢁA_S,AR). The following

signals of substrates were integrated before and after reaction: for

8a: d=6.90 (d, ³J_H,H=12.6 Hz, 1H), 6.66 ppm (d, ³J_H,H=12.6 Hz,

1H); for 9a: d=6.67 ppm (s, 2H); for 11 a: d=7.14 ppm (s, 2H);

Catalytic experiments

General procedure

Typically, five reactions were performed in parallel for each sub-

strate pair. Thus, Mn^IIcatalyst 6 (3 mmol) was weighed into five

2 mL glass vials. Two of the glass vials were charged with Zn^II–por-

phyrin 12 (3 mmol), two other glass vials were charged with 12

(7.5 mmol), and the last glass vial was not charged with any

amount of 12. After the addition, a 2 mL stock solution in CH₂Cl₂

was prepared containing two substrates (0.24 mmol of each) and

benzyl benzoate (internal standard, 0.12 mmol). For the inhibition

experiment, 4-ethylpyridine (0.72 mmol) was also added to the

3

for 27a: d=7.04 ppm (d, J_H,H=16.3 Hz, 1H); for 4a: d=5.55 ppm

(dt, ³J_H,H=11.6 Hz, ³J_H,H=7.3 Hz, 1H); for 5a: d=5.72 ppm (dt,

3

³J_H,H=11.7 Hz, J_H,H=7.1 Hz, 1H). The following signals of epoxides

3

were integrated after the reaction: for (Z)-8b: d=4.59 (d, J_H,H

=

4.6 Hz, 1H), 4.54 ppm (d, ³J_H,H=4.6 Hz, 1H); for (E)-8b (10b): d=

4.22 ppm (d, J_H,H=1.9 Hz, 1H); for (Z)-9b: d=4.48 ppm (s, 2H); for

3

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&14

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

M. Pruski, V. S. Y. Lin, J. Am. Chem. Soc. 2004 , 126 , 101 0 –1 011; e) A.

Anan, K. K. Sharma, T. Asefa, J. Mol. Catal. A 2008 , 288 , 1 –1 3.

[6] B. Cornils, W. A. Herrmann in Applied Homogeneous Catalysis with Or-

ganometallic Compounds, Vol. 1 (Eds. B. Cornelis, W. A. Herrmann),

Wiley-VCH, Weinheim, 1996, p. 5.

[7] a) R. Breslow, Artificial Enzymes, Wiley-VCH, Weinheim, 2005; b) P. Balles-

ter, A. Vidal-Ferran in Supramolecular Catalysis (Ed.: P. W. N. M van Leeu-

wen), Wiley-VCH. Weinheim, 2008; c) J. W. Steed, J. L. Atwood, Supra-

molecular Chemistry, Wiley-VCH, Weinheim, 2009.

(E)-9b (11b): d=4.11 (s, 2H) (the signals of (Z)-9b and (E)-9b

might be the opposite), 4.16 ppm (d, ³J_H,H=1.9 Hz, 1H); for 11 b

and 27b: overlap of a multiplet at d=3.90–3.87 ppm; for 4b: d=

3

4.01 ppm (d, ³J_H,H=4.2 Hz, 1H); for 5b: d=3.26 ppm (dt, J_H,H

=

3

6.3 Hz, J_H,H=4.3 Hz, 1H). Owing to the fact that the areas of differ-

ent proton resonances were used to determine the amount of con-

sumed substrate and produced epoxide, there is an uncertainty in

the values, as each proton resonance has its own relaxation time,

which thus causes different values of the integrated area depend-

ing on which resonance is integrated.

[8] J. K. M. Sanders, Chem. Eur. J. 1998 , 4 , 1378.

[9] a) W. Zhang, J. L. Loebach, S. R. Wilson, E. N. Jacobsen, J. Am. Chem. Soc.

1990 , 112 , 2801; b) R. Irie, K. Noda, Y. Ito, N. Matsumoto, T. Katsuki, Tet-

rahedron Lett. 1990 , 31 , 7345.

[10] This construction principle has previously been applied to the cyclic di-

merization of two identical rods: Y. Ducharme, J. D. Wuest, J. Org. Chem.

1988 , 53 , 5787.

[11] a) J. Gorzynski Smith, Synthesis 1984 , 629; b) A. S. Rao, S. K. Paknikar,

J. G. Kirtane, Tetrahedron 1983 , 39 , 2323.

[12] For leading references, see: a) P. Bhyrappa, J. K. Young, J. S. Moore, K. S.

Suslick, J. Am. Chem. Soc. 1996 , 118 , 5708; b) P. Bhyrappa, J. K. Young,

J. S. Moore, K. S. Suslick, J. Mol. Catal. A 1996 , 113 , 109; c) J. P. Collman,

J. I. Brauman, B. Meunier, S. A. Raybuck, T. Kodadek, Proc. Natl. Acad. Sci.

USA 1984 , 81 , 3245.

TON experiment of 6 and 6+12 by using 27a as substrate

Two 2 mL glass vials were charged with 6 (1.99 mmol each) and 12

(1.99 mmol each). In parallel, two other 2 mL glass vials were

charged with only 6 (1.99 mmol each). A stock solution (1.5 mL) in

CH₂Cl₂was prepared containing 27a (0.24 mmol) and the IS

(0.12 mmol). Stock solution (250 mL) was added to the vials con-

taining 6 and 6+12, and the mixtures were then diluted with

CH₂Cl₂(150 mL) by means of a Hamilton syringe. After the addition,

the vials were subjected to ultrasonication until a clear homogene-

ous solution was obtained. The solutions were transferred by

means of cannula to four other glass vials each loaded with PhIO

(59.7 mmol). The obtained mixtures were stirred for 7 days at RT.

Then, the workup was performed in an identical way as that de-

scribed under the “General procedure”. The conversion of 27a was

calculated in the same way as that described in the section “Deter-

mination of the substrate selectivity by GC”. To obtain the TON,

the observed conversion of 27a was multiplied by the molar

amount of 27a before the reaction and then divided by the molar

amount of 6 present in each mixture. The two vials loaded with 6

gave TON=8.1 and 10.3. The vials loaded with 6+12 gave TON=

12.8 and 13.8 (see the Supporting Information for details).

[13] For leading references, see: a) J. P. Collman, J. I. Brauman, J. P. Fitzgerald,

P. D. Hampton, Y. Naruta, T. Michida, Bull. Chem. Soc. Jpn. 1988 , 61 , 47;

ˇ

´

b) B. Y. Wang, T. Zujovic, D. A. Turner, C. M. Hadad, J. D. Badjic, J. Org.

Chem. 2012 , 77 , 2675.

[14] a) R. Breslow, A. B. Brown, R. D. McCullough, P. W. White, J. Am. Chem.

Soc. 1989 , 111 , 4517; b) R. Breslow, X. J. Zhang, R. Xu, M. Maletic, R.

Merger, J. Am. Chem. Soc. 1996 , 118 , 11678.

[15] a) P. Fackler, C. Berthold, F. Voss, T. Bach, J. Am. Chem. Soc. 2010 , 132 ,

15911; b) P. Fackler, S. M. Huber, T. Bach, J. Am. Chem. Soc. 2012 , 134 ,

12869.

[16] P. A. Ulmann, A. B. Braunschweig, O.-S. Lee, M. J. Wiester, G. C. Schatz,

C. A. Mirkin, Chem. Commun. 2009 , 5121.

[17] a) J. P. Collman, L. Zeng, R. A. Decreau, Chem. Commun. 2003 , 2974;

b) W. Dai, J. Li, G. S. Li, H. Yang, L. Y. Wang, S. Gao, Org. Lett. 2013 , 15 ,

4138.

[18] G. M. Mamardashvili, O. M. Kulikova, Russ. J. Coord. Chem. 2006, 32, 756.

[19] S. Jꢆnsson, F. G. J. Odille, P. O. Norrby, K. Wꢀrnmark, Org. Biomol. Chem.

2006, 4, 1927.

[20] All the substrate selectivities reported in the previous papers concern-

ing our supramolecular catalytic systems are normalized, compensating

the real reactivity for the inherent reactivity of each olefin employed as

substrate: see the Experimental Section for details.

[21] S. Jꢆnsson, F. G. J. Odille, P. O. Norrby, K. Wꢀrnmark, Chem. Commun.

2005, 549.

[22] F. G. J. Odille, S. Jonsson, S. Stjernqvist, T. Rydꢇn, K. Wꢀrnmark, Chem.

Eur. J. 2007 , 13 , 9617.

Acknowledgements

We thank the Swedish Research Council, the Royal Physiographic

Society (Lund), and the Crafoord Foundation for financial sup-

port. H.N.A. and E.L. thank the Crafoord Foundation for postdoc-

toral fellowships, and E.S is grateful to the Carl Trygger Founda-

tion for a postdoctoral fellowship. Mr. Clas Wesꢃn and Prof. Ola

Wendt, Lund University, are acknowledged for valuable discus-

sions.

[23] E. Sheibani, K. Wꢀrnmark, Org. Biomol. Chem. 2012 , 10 , 2059.

[24] N. S. Finney, P. J. Pospisil, S. Chang, M. Palucki, R. G. Konsler, K. B.

Hansen, E. N. Jacobsen, Angew. Chem. Int. Ed. Engl. 1997 , 36 , 172 0 –

1723; Angew. Chem. 1997 , 109 , 179 8 –1 801.

[25] B. Temelli, C. Unaleroglu, Tetrahedron 2006 , 62 , 10130.

[26] G. Consiglio, S. Failla, P. Finocchiaro, I. P. Oliveri, S. Di Bella, Inorg. Chem.

2012 , 51 , 8409.

Keywords: epoxidation · supramolecular catalysis · hydrogen

bonds · self-assembly · substrate selectivity

[27] D. Milstein, J. K. Stille, J. Am. Chem. Soc. 1978 , 100 , 3636.

[28] Y. Iseki, E. Watanabe, A. Mori, S. Inoue, J. Am. Chem. Soc. 1993 , 115 ,

7313.

[29] G. E. Keck, K. A. Savin, M. A. Weglarz, J. Org. Chem. 1995 , 60 , 3194.

[30] J. L. Gage, H. A. Kirst, D. O’Neil, B. A. David, C. K. Smith II., S. A. Naylor,

Bioorg. Med. Chem. 2003 , 11 , 4083.

[1] a) P. W. N. M. Van Leeuwen, Homogeneous catalysis understanding the

art, Kluwer academic publishers, London, 2004; b) P. W. N. M. Van Leeu-

wen, J. C. Chadwick, Homogeneous catalysts, Wiley-VCH, Weinheim,

2011.

[2] S. Saito, Y. Yamamoto, Chem. Rev. 2000 , 100 , 290 1 –2 915.

[3] J. P. Collman, X. M. Zhang, V. J. Lee, E. S. Uffelman, J. I. Brauman, Science

1993 , 261 , 140 4 –1 411.

[31] J. H. Espenson, Chemical Kinetics and Reaction Mechanisms, 2nd ed.,

McGraw-Hill, New York, 1995.

[4] For a review on substrate-selective catalysis, see: E. Lindbꢀck, S. Da-

waigher, K. Wꢀrnmark, Chem. Eur. J. 2014 , 20 , 1343 2 –1 3481.

[5] a) B. Z. Zhan, M. A. White, T. K. Sham, J. A. Pincock, R. J. Doucet, K. V. R.

Rao, K. N. Robertson, T. S. Cameron, J. Am. Chem. Soc. 2003 , 125 , 219 5 –

2199; b) H. Yang, Y. Chong, X. Li, H. Ge, W. Fan, J. Wang, J. Mater. Chem.

2012 , 22 , 906 9 –9 076; c) C. Zapilko, Y. Liang, W. Nerdal, R. Anwander,

Chem. Eur. J. 2007 , 13 , 316 9 –3 17 6; d) S. Huh, H. T. Chen, J. W. Wiench,

[32] a) T. Katsuki in Catalytic Asymmetric Synthesis, 2^nded. (Ed.: I. Ojima,

Wiley-VCH, New York, 2000, ch. 6B; b) E. N. Jacobsen, M. H. Wu in Com-

prehensive Asymmetric Catalysis (Eds.: E. N. Jacobsen, A. Pfaltz and H. Ya-

mamoto), Springer, 2006, ch. 18.2.

[33] a) C. Linde, M. Arnold, P.-O. Norrby, B. ꢄkermark, Angew. Chem. Int. Ed.

Engl. 1997 , 36 , 1723; Angew. Chem. 1997 , 109 , 180 2; b) C. Linde, B.

ꢄkermark, P.-O. Norrby, M. Svensson, J. Am. Chem. Soc. 1999 , 121 , 5083;

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&15

&

ÞÞ

These are not the final page numbers!

CHEMCATCHEM

FULL PAPERS

www.chemcatchem.org

c) C. Linde, N. Koliai, P.-O. Norrby, B. ꢄkermark, Chem. Eur. J. 2002 , 8 ,

2568; d) P. Fristrup, B. B. Dideriksen, D. Tanner, P.-O. Norrby, J. Am. Chem.

Soc. 2005 , 127 , 13672.

[39] Product selectivity is defined analogously to substrate selectivity.

[40] For one of the few reports on oxidation products other than epoxides

in Jacobsen–Katsuki epoxidations, see: A. Mꢇou, M.-A. Garcia, P. Brun, J.

Mol. Catal. A 1999, 138, 15.

[41] H. Saltzman, J. G. Sharefkin in Organic Syntheses Collective, Vol. V (Eds.:

H. E. Baumgarten, et al.), Wiley, New York, 1973, p. 658.

[34] G. A. Morris, S. T. Nguyen, J. T. Hupp, J. Mol. Catal. A 2001 , 174 , 15.

[35] K. Srinivasan, P. Michaud, J. K. Kochi, J. Am. Chem. Soc. 1986 , 108 , 2309.

[36] C. K. Chang, M.-S. Kuo, J. Am. Chem. Soc. 1979 , 101 , 3413.

[37] E. N. Jacobsen, W. Zhang, A. R. Muci, J. R. Ecker, L. Deng, J. Am. Chem.

Soc. 1991 , 113 , 7063.

[38] a) M. L. Merlau, M. D. P. Mejia, S. T. Nguyen, J. T. Hupp, Angew. Chem. Int.

Ed. 2001 , 40 , 423 9 –4 24 2; Angew. Chem. 2001 , 113 , 436 9 –4 372; b) R. V.

Slone, J. T. Hupp, Inorg. Chem. 1997 , 36 , 542 2 –5 423.

Received: September 10, 2014

Published online on && &&, 0000

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&16

&

ÞÞ

These are not the final page numbers!

FULL PAPERS

To be or not to be preorganized… A

third generation of a kinetically labile

supramolecular catalytic system was

synthesized. Its substrate selectivity in

the epoxidation of pyridyl- versus

phenyl-appended olefins was compared

to previous less preorganized systems

in the series. The third generation

showed higher substrate selectivities in

two out of five cases.

E. Lindbꢀck, S. Cherraben, J.-P. Francoꢁa,

E. Sheibani, B. Lukowski, A. ProÇ,

H. Norouzi-Arasi, K. Mꢂnsson,

P. Bujalowski, A. Cederbalk, T. H. Pham,

T. Wixe, S. Dawaigher, K. Wꢀrnmark*

&& – &&

A Double Conformationally Restricted

Dynamic Supramolecular System for

the Substrate-Selective Epoxidation of

Olefins—A Comparative Study on the

Influence of Preorganization

ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemCatChem 0000, 00, 1 – 17

&17

&

ÞÞ

These are not the final page numbers!

Article Doi

A double conformationally restricted dynamic supramolecular system for the substrate-selective epoxidation of olefins - A comparative study on the influence of preorganization

DOI: 10.1002/cctc.201402726

Source and publish data:

Authors:

Article abstract of DOI:10.1002/cctc.201402726

Full text of DOI:10.1002/cctc.201402726

Products guided by the article

R&D Labs maybe for 847946-94-5

Relevant to this article

Hot Product