University of Groningen

Supramolecular chirality transferDijk, Everhardus Wilt

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2010

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Dijk, E. W. (2010). Supramolecular chirality transfer. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 21-05-2021

79

4 Kinetic Investigations of DNA-based Conjugate Additions

In this chapter, kinetic studies on DNA-based conjugate additions to α,β-unsaturated 2-acyl imidazoles and azachalcone, the synthetic applications of which have been discussed in chapter 3, are presented. From time-dependent UV-Vis measurements, enhancement of the reaction rate by DNA was observed in the Michael † addition of dimethylmalonate, whereas the addition of nitromethane is slowed down significantly by DNA. This result may be explained by a better charge stabilization of deprotonated nitromethane by DNA as compared to dimethylmalonate. The observed effect of oligonucleotide sequences on the reaction rate and enantioselectivity was much smaller than what has been found for the Diels-Alder reaction, and a clear trend for the Michael addition cannot be observed. For DNA-based asymmetric Friedel-Crafts alkylations of substituted indoles, the influence of substituents at the enones and indoles on the reaction rate was studied as well, and significant electronic and steric effects were identified. From temperature-dependent kinetics of the Friedel-Crafts reaction, it is demonstrated that the Gibbs energy of activation in the presence of DNA is lower because of an enthalpic effect, similar to that observed for the Diels-Alder reaction. A mechanistic comparison is given between the conjugate additions discussed here and the DNA-based catalytic Diels-Alder reactions described earlier by our group. Parts of this chapter will be published: E.W. Dijk, A.J. Boersma, B.L. Feringa, G. Roelfes, manuscript in preparation.

† Throughout this chapter, the term Michael addition will be used to describe the

conjugate addition of a carbon nucleophile to an α-β-unsaturated ketone. When the

conjugate addition of indoles to enones is referred to, the term Friedel-Crafts

reaction will be used. The conjugate addition of water1 to enones of the type 4.1

will not be discussed.

Chapter 4

80

4.1 Introduction In view of the successful application of DNA/Cu2+-L4 as a supramolecular catalyst for asymmetric Michael additions,2 the question arose what the effect of DNA is on the rate and enantioselectivity of this conjugate addition. In the DNA-based Diels-Alder reactions, a significant positive effect of DNA on both the reaction rate and the ee of the products had been observed by Boersma et al. 3 Therefore, a kinetic study into the DNA-based asymmetric Michael reaction (Scheme 4.1A) was initiated to shed light on some of the observations made regarding the rate of the reactions. Initial research on the Friedel-Crafts reaction, which was added to our group’s DNA-based catalysis portfolio very recently,4 showed significant rate enhancements by DNA, and therefore this reaction was subjected to an extensive kinetic analysis as well (Scheme 4.1B).

To investigate the effect of DNA on the outcome of the conjugate additions under study, various mechanistic issues will be addressed. These include questions as to

Scheme 4.1A. The Michael additions of nitromethane and dimethylmalonateto enones 4.1a and 4.2; B. The Friedel-Crafts reaction of enones 4.1 with indoles 4.4.

Cu2+-L4

N

N

Cu2+

B

N

R'

R"

DNA, Cu2+-L4Aqueous buffer

Indoles 4.4

4.1

O

RN

NCH3

N

N O

CH3 R

N

R'

R''Friedel-Crafts adducts 4.5

A O

Ar

4.1a, Ar = 2-(1-methyl)imidazolyl4.2, Ar = 2-pyridyl

O

Ar

Nu

Michael adducts 4.3HNu = CH3NO2,H2C(CO2Me)2

DNA, Cu2+-L4Aqueous buffer

Kinetic Investigations of DNA-based Conjugate Additions

81

whether a positive effect of DNA on the reaction rates is apparent, and if so, if the sequence of the DNA used has an effect on the reaction rate and ee of the product. Additionally, the high enantioselectivities achieved will be explained, considering the fact that a significant amount of unbound copper complex is present in solution under the reaction conditions.3,5 Obviously, since the DNA-based Michael and Friedel-Crafts reactions take place in water, the hydrophobic effect and hydrogen-bonding with the substrates and/or reactants may play an important role. The pH of the reaction medium is also of critical significance; the nucleophiles employed in the Michael additions react in their deprotonated form, making the outcome of the reaction susceptible to variations in the medium. Finally, comparison of kinetic information from different reaction classes, viz. Michael additions, Friedel-Crafts reactions and cycloadditions, may yield a deeper under-standing of how our DNA-metal conjugates affect these catalytic reactions.

4.2 Michael Reactions with Nitromethane and Dimethylmalonate

4.2.1 Mechanistic Information on Michael Reactions Catalyzed by Lewis Acids

Despite the potential of Lewis-acid catalyzed asymmetric Michael reactions in water, only a few successful attempts have been reported to date. The addition of β-ketoesters to methyl vinyl ketone (MVK) in water, catalyzed by Silver(I)/Tol-BINAP,6 Pd(II)/BINAP7 or Yb(III)/α-amino acids8 gave moderate to good enan-tioselectivities. Only in the latter example reported by Lindström, Wennerberg and co-workers, kinetic and mechanistic details were disclosed (Scheme 4.2).

The use of D-alanine as the ligand for ytterbium resulted in a significant rate acceleration (up to 138-fold) compared to the ligand-free reaction in a biphasic system. For the addition of 2,4-pentadione (R=CH3) to 2-cyclohexene-1-one

O

R

O O

R = CH3, OC2H5

Yb(OTf)3Amino Acid

Water, Δ

O

R

O

O4.6, R=CH3; cyclohexanone

Scheme 4.2. The ytterbium-catalyzed asymmetric Michael addition of ethyl acetoacetate or 2,4-pentanedione to MVK or 2-cyclohexen-1-one.8

Chapter 4

82

(Scheme 4.2), a complex dependence of the adduct’s ee on the pH of the medium was observed, which was attributed to a delicate balance between various events: both N-protonation of the basic ligand at low pH and a nonselective base-catalyzed pathway at higher pH are detrimental for enantioselectivity. Around the optimal pH (6.7), the coordination of the aminoacid to the metal ion is most efficient, leading to the highest chiral induction. The reaction conditions were further optimized to a temperature of 70°C, a catalyst loading of 5 mol% of Yb(OTf)3 and 12 mol% of D-alanine, giving adduct 4.6 in full conversion and an ee of 64%. The polarity and size of the amino acid ligand had a large effect on the yield and ee of 4.6, indicating that solvation effects play an important role in the enantio-differentiating step of the reaction. This conception was confirmed by the observation of an inversed temperature effect on the ee and a large contribution of the differential entropy of activation (ΔΔ‡SR-S) to the differential Gibbs energy of activation (ΔΔ‡GR-S) obtained from an Eyring plot. Using this information, optimization using a Yb(III)/ D-proline catalyst provided access to 4.6 in 79% ee and 94% yield after 21h at 90°C. Mandolini and co-workers have briefly commented on mechanistic aspects of their class of uranyl-salophen complexes that catalyze both Michael additions and Diels-Alder reactions in chloroform (Scheme 4.3).9 It was noted that the same catalyst shows very different affinities towards the enone reactants in both reactions. The Diels-Alder reaction between benzoquinone and 1,3-cyclohexadiene was catalyzed exclusively by virtue of transition-state lowering; no appreciable affinity of the complex for the reactants or the products was observed. In contrast, a significant Lewis acid/Lewis base interaction between the complex and the substrate was observed for the Michael addition of the β-ketoester to MVK.

Kinetic Investigations of DNA-based Conjugate Additions

83

4.2.2 The Effect of st-DNA on the Rate of the Michael Additions

In order to address the effects of DNA and its sequence and the pH of the medium, outlined in section 4.1, the DNA-based Michael additions (Scheme 4.4) were studied under various conditions using time dependent UV-Vis spectroscopy, in a similar fashion as described for the Diels-Alder reaction.3 In the absence and presence of DNA, first-order reactions with respect to the nucleophiles were observed using initial rate kinetics (general procedure 2) for the addition of both dimethylmalonate and nitromethane to enones 4.1a and 4.2. A large excess of Cu2+-L4 with respect to the substrate was used, so that the dissociation of the addition product from the complex could be ignored in the overall reaction rate. The fact that the reactions are first-order in nucleophile, and not pseudo-zero-order despite the vast excess applied, supports the notion that the deprotonated forms are the reactive species in the addition.

Scheme 4.3. Michael and D-A reactions, catalyzed by the same uranyl-salophen catalyst, but proceeding via completely different mechanisms of activation.

NN

OC12H25H25C12O

O O

UO2

Uranyl Catalyst

D-A reaction: Only the transition state is stabilized by the catalyst

O

O

Uranyl catalystChloroform, RT

H

H

O

O

Michael reaction: Significant catalyst-reactant interaction

O

CO2CH3

HO

CO2CH3

OO

(C2H5)3NUranyl catalystChloroform, RT

Chapter 4

84

The results of the kinetic measurements proved to be somewhat variable between different batches of DNA solution. This can be rationalized by the conception that the deprotonated nucleophiles are the reactive species. Since the nucleophiles are in equilibrium with their deprotonated forms in water (Scheme 4.5), the observed reaction rates are strongly influenced by the exact reaction conditions, including pH and pre-equilibration times of the stock solutions, especially if the equilibria depicted are slow compared to the addition reaction.

The use of more concentrated buffer solutions (100 mM instead of 20 mM), or preparing stock solutions in the buffer (rather than in acetonitrile) did not increase reproducibility in the addition of dimethylmalonate at pH 6.5.10 For this reason, apparent rate constants for the additions at pH 6.5 were determined in triplo simultaneously under exactly the same conditions, using a fixed excess of the nucleophile (1.50·103 eq. of nitromethane, 150 eq. of dimethylmalonate). The results are listed in Table 4.1.

O

Ar

4.1a, Ar = 2-(1-methyl)imidazolyl4.2, Ar = 2-pyridyl

O

Ar

Nu

4.3a, Ar=2-(1-methylimidazolyl), Nu=CH(CO2CH3)24.3b, Ar=2-(1-methylimidazolyl), Nu=CH2NO24.3c, Ar=2-pyridyl, Nu=CH(CO2CH3)24.3d, Ar=2-pyridyl, Nu=CH2NO2

HNu = CH3NO2,H2C(CO2Me)2

st-DNA, Cu2+-L420 mM MOPS, pH 6.5

Scheme 4.4. The DNA-based asymmetric Michael addition of nitromethane anddimethylmalonate to enones 4.1 and 4.2.

A

O O

OO

O O

OO-H+

+H+

O O

OO

O O

OOH

BH3C N

O

O

H2C N

O

O

H2C N

O

O

H2C N

OH

O

-H+ +H+

+H+ -H+

Scheme 4.5. Possible acid-base and tautomerization equilibria of dimethylmalonate (A) and nitromethane (B) in water.

Kinetic Investigations of DNA-based Conjugate Additions

85

Table 4.1. Rate constants for the addition of dimethylmalonate and nitromethane to enones 4.1a and 4.2.a

Entry Enone Nucleo-

phile kapp (DNA)

(M-1 s-1) kapp (no DNA)

(M-1 s-1) kDNA/knoDNA

1 4.1a nitro-

methane(1.15±0.24)·10-4 b (5.0±0.89)·10-4 b 0.23

2 dimethyl-malonate

(3.31±0.40)·10-3 b (1.03±0.09)·10-3 b 3.2

3 4.2 nitro-

methane(2.12±0.17)·10-3 b (1.02±0.01)·10-2 b 0.21

4 dimethyl-malonate

(9.36±0.25)·10-2 c (2.38±0.04)·10-2 c 3.9

a. Conditions: st-DNA (0.67 mg mL-1), Cu2+-L4 (0.15 mM), 20 mM MOPS (pH 6.5), substrate (6.0 μM), dimethylmalonate (0.9 mM) or nitromethane (9.0 mM); b. Reactions were run in triplo with different batches of st-DNA and/or Cu2+-L4 solution with a known, variable concentration of nucleophile; c. Reactions were run in triplo using the same batch of st-DNA and/or Cu2+-L4 solution with a fixed excess of the nucleophile (see text).

Interestingly, DNA retards the addition of nitromethane by a factor of almost 5. The fact that the addition reaction of nitromethane proceeds faster in the absence of DNA might help to explain the vast excess of nucleophile needed to obtain full conversion in this reaction, and the somewhat lower ee levels obtained compared to the addition of dimethylmalonate.2 The observed binding constant of the complex Cu2+-L4 to DNA (Kb = 1.12·10-4 M-1)5 leads to a significant amount (around 5%) of free complex in solution under the reaction conditions. This free Cu2+-L4 catalyzes the addition more efficiently than bound complex. Hence, a considerable blank reaction takes place in the bulk phase, leading to a lower overall enantio-selectivity. It should be noted that the conditions under which the adduct of nitro-methane to 4.1a is isolated differ significantly (temperature, relative concentrations of reactants) from those used for the determination of the rate constants. Thus the ee of typically 85% is very well possible on an analytical scale despite the considerable blank reaction that is observed in the kinetics experiments. As opposed to the Michael addition of nitromethane, the addition of dimethylmalonate to enones 4.1a and 4.2 is moderately accelerated by the presence

Chapter 4

86

of DNA (Table 4.1, entries 2 and 4). This is in line with the observation that a much smaller excess of dimethylmalonate is needed to obtain full conversion in the addition, as compared to nitromethane (see chapter 3). From the pKa values of both nucleophiles, an opposite effect would be expected; nitromethane’s pKA of ~ 10 in water11 is 3 units lower than that of dimethylmalonate (pKA around 13).12 This means that in the buffer used, at a given initial concentration of nucleophile, approximately 1000 times more of deprotonated nitromethane is present than of deprotonated dimethylmalonate. However, this is not reflected in the reaction rates of the respective addition reactions in the presence of DNA. The observation that pKa values and nucleophilicities are not correlated was made before in various solvents by the group of Mayr.13 For example, in the nucleophilic addition to the electrophilic benzhydrilium cation in methanol, the anion of dimethylmalonate is about 3 orders of magnitude more reactive than the anion of nitromethane.13c This was attributed to the conception that reactivity is controlled more by solvation than by the intrinsic properties of the anions, such as their basicities. A similar effect is observed in the relatively slow acid-base equilibria of nitroalkanes, despite their high acidities.14 A possible explanation for this may be Bernasconi’s “principle of nonperfect synchronization”.15 In the protonation at the carbon atom of deprotonated nitromethane (the first equilibrium in Scheme 4.5B), the activation barrier is increased because the change in resonance (negative charge mainly delocalized on the oxygen atoms) lags behind the proton transfer event (at the carbon atom). A parallel can be drawn to the nucleophilic attack of deprotonated nitromethane to the Michael acceptors, as compared to the attack of deprotonated dimethyl-malonate. DNA might somehow make the nonperfect synchronization larger for nitromethane than for dimethylmalonate, for example by a difference in hydrogen bonding interactions. This may explain the observed difference in kDNA/knoDNA of both nucleophiles in the Michael reaction. The question whether hydrogen bonding with DNA is indeed responsible for the observed rate effects might be answered by adding different H-bond donors, e.g. sugars, to the reaction mixture, or by kinetic experiments in D2O. Since 2H-bonds would be involved in D2O, a kinetic solvent isotope effect might occur.

Kinetic Investigations of DNA-based Conjugate Additions

87

4.2.3 The Effect of Catalyst Loading and pH on the Rate of Michael Reactions

The variation of reaction rate with the concentration of Cu2+-L4 was studied at a fixed ratio of 6.7 base pairs per complex, showing a first-order reaction with respect to Cu2+-L4 (Figure 4.1). This result indicates that the DNA-bound copper complex is involved in the rate-determining step of the reaction. Presumably, the enone coordinates to copper in a bidentate fashion and is activated for nucleophilic attack at the β-carbon.

In view of the anticipated significant effect of pH on the outcome of the Michael addition, the addition of nitromethane and dimethylmalonate to 4.1a was also studied at pH 5.5 and pH 7.5. Since more deprotonated nucleophile is present at higher pH, a rate increase with pH is expected. Indeed, the rate of the reaction was much higher and more reproducible at pH 7.5. Moreover, enantioselectivities of the reactions were comparable to those at pH 6.5 (Table 4.2, entries 5-6). At pH 5.5 (using MES buffer instead of MOPS), incomplete conversion and significantly lower enantioselectivity was observed for the addition of nitromethane to 4.1a on a preparative scale after 3d (entry 1). The addition of dimethylmalonate did show full conversion at pH 5.5 after 3d, giving 4.3a in a slightly higher ee than at pH 6.5 (entry 2). The reaction rates for the addition of both nucleophiles at pH 5.5 were

Figure 4.1. Dependence of reaction rate constants of the addition ofdimethylmalonate (squares) and nitromethane (circles) to 4.1a on the concentration of Cu2+-L4. Note the different scales on both vertical axes.

0.05 0.10 0.15 0.20 0.25 0.300.000

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0.008

k 2,ap

p (M-1 s-1

) (di

met

hylm

alona

te)

[Cu2+-L4] (mM)

0.0000

0.0005

0.0010

0.0015

0.0020

0.0025

0.0030

k 2,ap

p (M-1 s-1

) (ni

trom

etha

ne)

Chapter 4

88

very low, and no reliable data could be extracted from time-dependent UV-Vis spectroscopy measurements. Table 4.2. The effect of pH on the ee of 4.3a/b and the rate of the Michael addition of nitromethane and dimethylmalonate to 4.1a. Entry pH Nucleophile Ee

(%)a

kapp (DNA)b

(M-1 s-1) kapp (no DNA)b

(M-1 s-1)

kDNA/knoDNA

1 5.5c nitromethane 74 n.d.d n.d. 2 dimethylmalonate 95 n.d. n.d. 3 6.5 nitromethane 85 (1.15±0.24)·10-4 (5.0±0.89)·10-4 0.23 4 dimethylmalonate 92 (3.31±0.40)·10-3 (1.03±0.093)·10-3 3.2 5 7.5 nitromethane 77 (5.97±0.56)·10-4 (1.11±0.08)·10-3 0.54 6 dimethylmalonate 90 (6.73±0.32)·10-3 (1.12±0.037)·10-3e 6.0

a. Determined independently using general procedure 1; b. Determined in duplo using general procedure 2; c. MES buffer was used instead of MOPS; d. n.d. = not determined: values were too low to be accurately determined; e. Reactions were run in triplo using the same batch of or Cu2+-L4 solution with a fixed excess of the nucleophile (300 eq.). In both cases, the increase of the reaction rates with the pH of the medium is more pronounced in the presence of DNA than in its absence: the kDNA/knoDNA of the reactions is roughly 2 times larger at pH 7.5 than at pH 6.5. From the pH-dependence data, it is clear that the reactivity of the nucleophiles is not simply correlated to their pKA values. Because of the higher rates and the observed increase of kDNA/knoDNA of the addition of dimethylmalonate to 4.1a at pH 7.5, measurements of sequence-dependent reaction rates (section 4.2.4) were performed at that pH.

4.2.4 Sequence Dependence of the Michael Reaction

For the Michael reaction of enone 4.1a with dimethylmalonate, apparent rate constants and ees were determined for a selection of synthetic oligonucleotides (Table 4.3). Oligonucleotide d(TC GGG AT CCC GA)2 was chosen as the parent sequence because some information on the binding mode of a few transition metal complexes has been reported. In NMR studies by Collins et al., binding of various complexes (Pt(en)22+, Δ-Co(en)33+ and Δ-[Ru(phen)2 dpq]2+) was shown to occur either at the central AT sequences in the minor groove16 or at a GG/CC base pair

Kinetic Investigations of DNA-based Conjugate Additions

89

in the major 17 or minor 18 groove. By varying the number of consecutive deoxyguanines (3, 2 or 1) and the central AT sequence (present, absent or inversed), we observed interesting effects in the DNA-based Diels-Alder reaction,3 and in view of these results, the oligonucleotides screened there were also used for the present study.

The effect of the DNA sequence on the ee of 4.3a was studied first at an analytical scale at 5°C. The same trends were roughly observed as for the Diels-Alder reaction. The copper-oligonucleotide conjugates that induce the highest and the lowest enantioselectivities are the same for the Michael and D-A reactions. Apparently, similar requirements, e.g. the length of the G-tracts and the presence of the central AT base pairs, apply for both reaction classes. The reaction rate of the Michael reaction, however, is less susceptible to the DNA sequence than in the D-A reaction; the difference between the highest and lowest rate constant observed here is a factor of 5 at most, where an order-of-magnitude difference was found for the D-A reaction using the same sequences. As can be seen in the graphical representation of the ee vs. rates (Figure 4.2), no clear trend can be observed between the results obtained from the various Cu2+-L4/oligonucleotide conjugates. Possibly, the optimal sequence has not been identified yet, as catalysis in the presence of st-DNA proceeds up to 2.5 times faster than the fastest rate observed in the presence of the tested synthetic

Table 4.3. Sequence-dependent enantioselectivity and apparent second-order rate constants for the addition of dimethylmalonate to enone 4.1a.

Entry Sequence (5’→3’) Ee (%)a kapp (M-1 s-1)b

1 d(TC GGG AT CCC GA) 86 (2.17±0.14)·10-3

2 d(TCA GGG CCC TGA) 93 (2.60±0.11)·10-3

3 d(TC GG AA TT CC GA) 76 (2.71±0.56)·10-3

4 d(TCG CGA TCG CGA) 71 (1.24±0.06)·10-3

5 d(TCG CGT ACG CGA) 74 (1.47±0.20)·10-3

6 st-DNA 90 (6.54±0.08)·10-3

7 no DNA - (1.85±0.26)·10-3

a. Ees of 4.3a were determined in duplo (error ≤ 1%) using general procedure 1 at pH 6.5 at a 0.6 μmol scale; b. Apparent rate constants were determined using general procedure 2, pH 7.5, 300 eq. of dimethylmalonate, 18°C, in triplo.

Chapter 4

90

oligonucleotides. It is conceivable that a very active and selective chiral micro-environment exists in st-DNA that is responsible for both a high ee and rate in the Michael reaction. Two of the sequences tested exhibit addition rates that are even slightly lower than those found for reactions without DNA present. Apparently, these sequences contain an environment that decreases the activity of the catalyst or have lower binding affinity to the Cu2+-L4 complex. However, the effects of DNA sequence on the outcome of the reaction are much less pronounced than for the Diels-Alder reaction, and trends are more difficult to identify.

Kinetic Investigations of DNA-based Conjugate Additions

91

4.3 The DNA-based Friedel-Crafts Alkylation of Indoles Another classical metal-catalyzed C-C bond forming reaction that has been widely studied20 and found to be water-tolerant21 is the Friedel-Crafts reaction. The first copper-catalyzed asymmetric Friedel-Crafts alkylation of β,γ-unsaturated α-keto-esters was reported by Jørgensen,22 and Evans et al., reported the use of α,β-

70 75 80 85 90 95

1

2

3

4

5

6

7k 2,

app (1

0-3 M

-1 s-1

)

Ee (%)

st-DNA→A

B

4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0

-6.8

-6.6

-6.4

-6.2

-6.0

-5.8

-5.6

-5.4

-5.2

-5.0

-4.8

Ln (k

app /

M-1 s-1

)

ΔΔ‡G (kJ mol-1)

Figure 4.2. Graphical representation of sequence-dependent rates and ees; B. Plot of ln (kapp) against the difference in ΔΔ‡G for both enantiomers of 4.3a (B)19.

Chapter 4

92

unsaturated 2-acylimidazoles 4.1 as viable substrates in the Friedel-Crafts alkylation of electron-rich aromatic compounds, such as indoles (Scheme 4.6).23

Using chiral scandium-pybox complexes as the catalyst, highly selective reactions were achieved in organic solvents under the strict exclusion of water. Although the reaction is formally a Lewis acid-catalyzed electrophilic aromatic substitution of the indole by the enone, it may also be described as the conjugate addition of the indole to the activated enone. In an extensive study on the substrate scope and mechanism of the reaction, a mononuclear complex was identified as the active catalyst. 24

The asymmetric DNA-based Friedel-Crafts reaction employs Cu2+-L4 as the complex of choice for chirality transfer from DNA in water.4 As opposed to the conjugate addition of enolizable nucleophiles described in the previous sections, the addition of the indole to the enone does not depend on acid-base equilibria prior to attack of the nucleophile. As a result, the reactions are much less

N

N

O

CH3

R

N

R'

R''N

N O

CH3 R4.1

Lewis acid (cat.)

N

R'

R"

Scheme 4.6. The Lewis acid-catalyzed asymmetric Friedel-Crafts alkylation of indoles by enones 4.1, first reported by Evans et al.23

N

N O

CH3

R

N

R'

R"

4.5a-ga, R = C6H5, R' = H, R'' = OCH3b, R = 4-ClC6H4, R' = H, R'' = OCH3c, R = 4-BrC6H4, R' = H, R'' = OCH3d, R = 4-CH3OC6H4, R' = H, R'' = OCH3e, R = C6H5, R' = CH3, R'' = Hf, R = CH3, R' = H, R'' = OCH3g,R = C5H11, R' = H, R'' = OCH3

N

N O

CH3 R

4.1a-fa, R = C6H5b, R = 4-ClC6H4c, R = 4-BrC6H4d, R = 4-CH3OC6H4e, R = CH3f, R = C5H11

N

R'

R"

st-DNA, Cu2+-L4water, 20 mM MOPS pH 6.5

4.4 a-ba, R' = H, R'' = OCH3b, R' = CH3, R'' = H

Scheme 4.7. The DNA-based copper-catalyzed asymmetric Friedel-Crafts reaction described in this section.

Kinetic Investigations of DNA-based Conjugate Additions

93

susceptible to subtle changes in the reaction medium. The fact that indoles act as neutral nucleophiles in this reaction facilitates the addition as well, since charge-repulsion with the negatively charged phosphate backbone of DNA does not play a role. The kinetics of the reaction of 4.4a with 4.1a were briefly studied before4 and a significant rate acceleration by st-DNA was observed. In this section, a more detailed kinetic study of the DNA-based Friedel-Crafts reaction (Scheme 4.7) is presented.

4.3.1 Kinetic Study of the Friedel-Crafts Reaction: Screening of Substrates and Indoles

In view of the promising initial results for the kinetics of the Friedel-Crafts reaction obtained before, the effect of various substituents at both the enone and the indole on the rate of the reaction was investigated. The observed reaction rates were first-order in the indole, and well reproducible between different batches of DNA solutions. The results are shown in Table 4.4.

Table 4.4. Apparent second-order rate constants for the Friedel-Crafts reaction of substituted indoles to various substrates. a

Entry Enone Indole Product Ee (%)bkapp (DNA)c

(M-1 s-1) kapp (no DNA)c

(M-1 s-1) kDNA/knoDNA

1 4.1a 4.4a 4.5a 57 7.75·10-2 4.31·10-3 18 2 4.1b 4.4a 4.5b 79 0.135 6.45·10-3 21 3 4.1c 4.4a 4.5c 66 0.235 8.96·10-3 26 4 4.1d 4.4a 4.5d 69 4.62·10-2 3.73·10-3 12 5 4.1a 4.4b 4.5e 75 6.33·10-2 7.02·10-3 9.0 6 4.1ed 4.4a 4.5f 83 0.842 3.10·10-2 27 7 4.1fd 4.4a 4.5g 80 0.429 2.72·10-2 16

a. Determined at 25±0.2°C using general procedure 2; b. Determined before on a semi-preparative scale, see ref. 4; c. Values are given in M-1s-1, errors in the rate constants were ≤ 9% in all cases; d. A significant decrease in the concentration of 4.1e and 4.1f was observed in the absence of indole, indicating the conjugate addition of water as a side reaction; the setup of the reaction compensates for this.25

In all cases, DNA had a significant and positive effect on the rate of the reaction. The introduction of electron-withdrawing substituents on the aromatic enone (4-Cl, 4-Br) led to a significant rate increase, as well as a larger acceleration by DNA

Chapter 4

94

(entries 1-3). On the other hand, an electron-donating substituent (4-OCH3, entry 4) gave a slightly slower reaction and a less pronounced acceleration by DNA than for the parent enone 4.1a. The rate of alkylation of N-methylindole 4.4b by enone 4.1a (entry 5) is enhanced by DNA to a lesser extent than alkylation by 5-methoxyindole 4.4a (entry 1). For the aliphatic enones,25 an increase of the chain length resulted in a slightly lower addition rate, especially in the presence of DNA (entries 6 and 7). The DNA/Cu2+-L4 catalyst assembly might accommodate the smaller 4.1e better than 4.1f because of steric factors, leading to a faster reaction. In general, no clear trend between the DNA acceleration factor and the ee of the product is apparent between the different substrates. From the results in Table 4.4, it is appropriate to describe the DNA/Cu2+-L4 conjugate as a substrate-specific catalyst, providing either a more appropriate chiral environment or larger rate enhancements (or a combination of both) for the Friedel-Crafts reaction to occur enantioselectively. The catalyst works best for the reaction of indole 4.4a with enones 4.1a and 4.1e, giving both high rates and enantioselectivities. In order to obtain further insight in the origins of the rate enhancements described above, it would be informative to know whether the acceleration originates from a stronger binding of the substrate to the copper in the presence of DNA, or from an intrinsically faster reaction. Thus, an attempt was made to determine the association constant (KA) of enone 4.1a to Cu2+-L4 by UV-Vis titration. However, both in the presence and absence of st-DNA, no significant changes in the UV-Vis spectrum of 4.1a were observed in the range of [Cu2+-L4] = 6.4·10-5 – 3.2·10-4 M ([4.1a]=1.5-5.0 μM). The more electron-rich substrate 4.1d, which is expected to bind more strongly to copper, showed no significant spectral changes upon titration with Cu2+-L4 either. Apparently, binding of the enones to copper is not very strong. Since the value of kcat can only be obtained if KA is known, it could not be determined for the Friedel-Crafts alkylation of these enones. As the aliphatic substrates 4.1e and 4.1f undergo significant conjugate addition of water under the reaction conditions, accurate determination of their binding constants to Cu2+-L4 in water is not possible. The Friedel-Crafts reaction of azachalcone 4.2 with 1-methylindole 4.4b did not give reproducible results for both conversion and ee on an analytical scale. A possible reason for this may be the fact that the product has a strongly acidic proton at the newly formed stereogenic center. Therefore, the product racemizes relatively easily under the reaction conditions and the outcome

Kinetic Investigations of DNA-based Conjugate Additions

95

of the reaction strongly depends on reaction times and conditions. For this reason, no attempts were made to investigate the effect of DNA on the outcome of this reaction.

4.3.2 Activation Parameters of the Friedel-Crafts Reaction

For the DNA-based asymmetric Diels-Alder reaction between 4.2a and cyclopentadiene, described earlier in our group, it was shown that the observed rate enhancement is caused by a sharp decrease in activation enthalpy Δ‡H in the presence of DNA.3 This effect is partly counteracted by a less favorable entropy of activation Δ‡S. This may be caused by lowering of the transition state energy by favorable interactions between the catalyst and the transition state, as was put forward as an explanation for comparable effects in Diels-Alder reactions catalyzed by Antibody 1E9.26 From quantum mechanical calculations and docking studies, the enthalpic effect responsible for the observed rate acceleration was attributed to shape complementarity between the hydrophobic binding site in the antibody and the transition state of the reaction, together with favorable hydrogen-bonding interactions. For the copper-catalyzed Friedel-Crafts reaction between 4.1e and 4.4a, temperature-dependent rate constants in the presence and absence of st-DNA were determined and plotted in an Eyring plot as ln (kapp/T) vs. 1/T (Figure 4.3) in order to extract the activation parameters Δ‡H, Δ‡S, and the related Gibbs’ energy of activation Δ‡G.27

Chapter 4

96

A clear deviation from linearity is apparent in the Eyring plot for the reaction in the presence of DNA below 20°C. Curvature in the Eyring plot may be caused by many factors, such as temperature-dependence of the numerous binding equilibria (DNA/Cu2+-L4,28 Cu2+-L4/4.1e and possibly Cu2+-L4/4.4a) in solution. Another possibility is a negative heat capacity of activation (Δ‡Cp = ∂Δ‡H/∂T), giving rise to curved Eyring plots.29 Alternatively, a discontinuity (‘break’) in an Eyring plot may

Figure 4.3 Eyring plots for the Friedel-Crafts alkylation of indole 4.4aby enone 4.1e in the presence (A) and absence (B) of st-DNA.

A

0.00310 0.00315 0.00320 0.00325 0.00330 0.00335 0.00340 0.00345 0.00350-6.4

-6.3

-6.2

-6.1

-6.0

-5.9

-5.8

-5.7

-5.6

-5.5

Ln (k

app/T

)

1/T (K-1)

B

0.00310 0.00315 0.00320 0.00325 0.00330 0.00335 0.00340 0.00345 0.00350-9.8

-9.6

-9.4

-9.2

-9.0

-8.8

-8.6

-8.4

-8.2

-8.0

Ln(k

app/T

)

1/T (K-1)

Kinetic Investigations of DNA-based Conjugate Additions

97

be attributed to a change in mechanism at a certain temperature. This may be due to conformational changes in the catalyst, as has been proposed to occur in –among others – the cleavage of the cobalt-carbon bond in vitamin B12, catalyzed by ribonucleotide triphosphate reductase.30 However, the number of data points and the accuracy of the measured apparent rate constants and the temperature is generally not considered31 high enough to distinguish between the possible causes for the observed nonlinearity, and therefore no attempt was made in this direction. Interestingly, the copper-catalyzed Friedel-Crafts reaction in the absence of DNA did exhibit linear behaviour over the complete temperature window of 30 K (15 – 45°C, Figure 4.3B). To simplify the extraction of activation parameters from the Eyring plot in the presence of DNA, the linear part of the graph in the high temperature range (between 45°C and 20°C, the dotted line in Figure 4.3A) was considered, thus ignoring the complicating factors described above. The activation parameters thus obtained are listed in Table 4.5.

Table 4.5. Activation parameters for the Friedel-Crafts alkylation of 5-methoxyindole 4.4a by enone 4.1e.

Parameter Without st-DNAa With st-DNAb

Δ‡G (298K) 81.5 kJ mol-1 73.5 kJ mol-1 Δ‡H 37.2 kJ mol-1 12.4 kJ mol-1

T·Δ‡S(298 K) -44.3 kJ mol-1 -61.1 kJ mol-1 a. Error margins in these values are estimated ±5 kJ mol-1 and the curvature in the Eyring plot at temperatures below 20°C has been ignored, see text; b. Error margins in these values are estimated ±2 kJ mol-1.

Despite the two different substrates, reaction types and conditions, the activation parameters for the Friedel-Crafts reaction show a remarkable similarity to those determined for the Diels-Alder reaction. The observed rate acceleration by DNA is caused by a sizeable decrease in enthalpy of activation, which is partly counteracted by a less favorable entropy of activation for both reactions. These results point in the direction of nonclassical hydrophobic effects, 32 for example as a result of favorable π-π interactions between the reactants and the DNA. However, this is

Chapter 4

98

highly speculative, and further (computational) studies should be performed to account for the interesting kinetic observations made for the two reaction classes.

4.3.3 Sequence Dependence of the Friedel-Crafts Reaction

For the Friedel-Crafts alkylation of 5-methoxyindole 4.4a by the enones 4.1d (R=4-C6H4OCH3) and 4.1e (R= CH3), a remarkable effect of the DNA sequence on the ee of the reaction product has been observed. In the reaction with 4.1e, the self-complementary oligonucleotide d(TCA GGG CCC TGA)2 gave an increase of enantioselectivity from 83% (st-DNA) to 93%. In contrast, the reaction with 4.1d showed a considerably lower ee with this oligonucleotide than with st-DNA (49% and 69%, respectively). In view of the significant rate acceleration of the Friedel-Crafts reaction by DNA (section 4.3.1), sequence-dependent reaction rates were measured to determine whether the rate of reaction is correlated to the ee of the product, as was the case in the DNA-based asymmetric Diels-Alder reaction. Oligonucleotide d(TCA GAG CTC TGA)2, inducing a lower ee than st-DNA in both adducts 4.5d and 4.5f , was included in the study as well. The results are displayed in Table 4.6.

Table 4.6.Sequence-dependent enantioselectivity and apparent second-order rate constantsa in the addition of 5-methoxyindole 4.4a to enones 4.1d and 4.1e. Entry Substrate Sequence Ee (%)a kapp (M-1 s-1) b Relative ratec

1 4.1d st-DNA 69 (3.10±0.20)·10-2 1 2 d(TCA GGG CCC TGA)2 49 (1.52±0.14)·10-2 0.49 3 d(TCA GAG CTC TGA)2 32 (1.45±0.091)·10-2 0.47 4 4.1e st-DNA 83 0.63±0.036 1 5 d(TCA GGG CCC TGA)2 93 0.66±0.032 1.05 6 d(TCA GAG CTC TGA)2 65 0.31±0.015 0.49

a. Determined by Arnold Boersma under standard conditions on an analytical scale (Ref. 4); b. Determined at 18±0.2°C using general procedure 2; c. Relative rate for st-DNA = 1

The influence of the source and sequence of the DNA used on the reaction rate is minimal, despite the substantial effect on the enantioselectivity of the product.33 No significant rate increase caused by the oligonucleotide d(TCA GGG CCC TGA)2 relative to st-DNA in the addition of 5-methoxyindole 4.4a to 4.1e was observed (entries 4 and 5). The oligonucleotide d(TCA GAG CTC TGA)2, which

Kinetic Investigations of DNA-based Conjugate Additions

99

induces a lower ee in 4.5a than st-DNA, also gives a lower reaction rate, although the effect is rather small (entry 6). The same holds for the reaction between 4.1d and 4.4a, displaying a lower enantioselectivity using both oligonucleotides than using st-DNA (entries 1-3); in these cases reaction rates are a factor 2 lower compared to st-DNA.

4.4 Discussion

4.4.1 The DNA-based Asymmetric Michael vs. Friedel-Crafts Reactions

From the kinetic data presented in sections 4.2 and 4.3, distinct similarities between the Michael additions and Friedel-Crafts alkylations become apparent. Most prominent is the extent to which the reaction is accelerated by DNA: this is around an order of magnitude higher for the Friedel-Crafts reaction than for the Michael reaction with dimethylmalonate. The addition of nitromethane to enones 4.1a and 4.2 even becomes up to 5 times slower in the presence of DNA (Table 4.1). The main reason for this is probably that deprotonated nucleophiles are the reactive species in the Michael reaction, experiencing charge repulsion from the negatively charged phosphate backbone of DNA. On the other hand, the indoles reacting as neutral nucleophiles in the Friedel-Crafts reaction do not require deprotonation prior to addition to the enone, and hence do not suffer from charge repulsion by the DNA. The good ees (around 85%) obtained for the nitromethane adducts 4.3b and 4.3d are noteworthy, considering the presence of ~5% of unbound Cu2+-L4 in solution. The free complex catalyzes the addition of nitromethane up to 5 times more efficiently, although leading to racemic product. The addition event catalyzed by the Cu2+-L4/DNA conjugate leading to nonracemic product must therefore be almost completely enantioselective. The 29-fold rate difference between the additions of nitromethane and dimethylmalonate to 4.1a at pH 6.5 in the presence of st-DNA (Table 4.1, entries 1 and 2) may explain the need for a larger excess of nitromethane for the addition to occur at desirable speed on a preparative scale (see chapter 3). The reason as to why the addition rates are so different is unknown, but it might be speculated that nitromethane in its deprotonated form experiences the “principle of nonperfect synchronization”, which has been put forward as a possible cause to explain its slow protonation.15 In principle, a similar kinetic barrier may exist for the addition

Chapter 4

100

of dimethylmalonate, but DNA might have a different influence for nitromethane than for dimethylmalonate, e.g. by hydrogen bonding. To prove this conception, additional kinetic studies should be performed.

4.4.2 The DNA-based Conjugate Additions vs. Diels-Alder Reactions

The acceleration by DNA of the copper-catalyzed Michael additions to 4.2 is around one order of magnitude lower than for the Diels-Alder reaction with cyclopentadiene3 using Cu2+-L4 as the metal complex. Furthermore, the effects of the DNA sequence on the rate and enantioselectivity of the Michael additions –although significant – are much less pronounced. For the Diels-Alder reaction, it was convincingly shown that specific microenvironments exist that significantly enhance both rate and enantioselectivity, but this conclusion is not immediately obvious from the kinetic data for the conjugate additions described in section 4.2. Of the polynucleotides tested, st-DNA induces the highest addition rate, although the synthetic oligonucleotide d(TCA GGG CCC TGA)2 induces a slightly higher ee in 4.3b, albeit at a lower rate of addition (Table 4.3). It may well be that the optimal sequence, enhancing both aspects of the reaction, has not yet been identified. For the DNA-based Friedel-Crafts reactions, a different comparison may be made to the Diels-Alder reaction. For both reactions, acceleration by DNA of at least one order of magnitude has been demonstrated (Table 4.4). Most intriguingly, temperature-dependent kinetics (section 4.3.2) indicate that in both cases the rate increase is of enthalpic origin, partly counteracted by an unfavorable entropy of activation. Despite the different reaction classes, the activation parameters for both reactions are strikingly similar. For the Friedel-Crafts reaction in the presence of DNA, curvature in the Eyring plot is apparent, especially below 20°C. This may indicate a negative heat capacity of activation Δ‡Cp for the reaction or a change of reaction pathway in the temperature window studied. Alternatively, the assumptions on which the Eyring plot is based (i.e. all binding constants are invariable over the 30K temperature range) may be oversimplified and derivation of the activation parameters from the complex rate constant kapp is not justified. Unfortunately, for enones 4.1, the binding affinity for Cu2+-L4 could not be determined, hampering the deconvolution of kapp into KA and kcat.3 The nonlinearity of the Eyring plot between 15 and 45°C was not investigated further, since activation parameters could be extracted from a linear part of the

Kinetic Investigations of DNA-based Conjugate Additions

101

plot over an acceptable temperature range (20 – 45°C). The enthalpic origin of the rate enhancement by DNA indicates that classical hydrophobic effects do not constitute the main causes for the observed acceleration: a favorable entropic term would be expected in that case. Possibly, computational studies may shed light on the decrease of the activation barriers for both reactions.

4.5 Conclusions In conclusion, the kinetic studies on the DNA-based asymmetric conjugate additions to 4.1 and 4.2 described in this chapter give valuable insights into the reaction mechanisms. From the results of the Michael addition, it is clear that deprotonation and/or tautomerization equilibria are involved before the actual C-C bond forming step. These equilibria complicate the reaction kinetics dramatically, and no direct relationship between the pKa values of the nucleophiles and their reactivity is present. At higher pH, the reactions display a higher kDNA/knoDNA, at the cost of a slightly lower ee of the product. In general, the effects of the sequence of the oligonucleotide on the rate and enantioselectivity of the reaction are less pronounced than for the Diels-Alder reaction, and no clear trend is apparent. For the Friedel-Crafts reaction, the DNA-Cu2+-L4 conjugate may be described as a very selective catalyst, inducing the highest rate enhancements and enantioselectivities in the reactions between 5-methoxyindole by 4.1b and 4.1e. The observed rate increase for the copper-catalyzed asymmetric Friedel-Crafts reaction in the presence of DNA is caused by a significant lowering of the enthalpy of activation, as was observed in the Diels-Alder reaction catalyzed by the same catalyst system. Curvature in the Eyring plot obtained indicates interesting behavior of the reaction in the presence of DNA, but the experimental error in the values obtained preclude an unambiguous conclusion as to its cause.

4.6 Experimental Section For general remarks, see section 2.4. General procedure 1. Analytical-scale addition reactions to enones 4.1 and 4.2. To an ice-cold solution (prepared 24 h in advance) of DNA (final concentration 1.3 mg mL-1 or 2 mM in base pairs) in MOPS buffer (final concentration 20mM, pH 6.5 or 7.5), a solution of Cu(NO3)2-L4 in MilliQ water was added to a final concentration of 0.3 mM. The solution was equilibrated by

Chapter 4

102

gentle mixing, and the enone was added as a stock solution in acetonitrile. The nucleophile was added neat or as a stock solution in acetonitrile, and the solution was mixed by continuous inversion at 5°C. After the time indicated in the text, the reaction mixture was extracted with ethyl acetate (3x5 mL). After drying (Na2SO4) and removal of the solvent, the crude product was analyzed by 1H-NMR for conversion and chiral HPLC for ee, data are generally reproducible within 2%. General Procedure 2. Determination of rate constants using time-dependent UV-Vis spectroscopy. Rate constants for a variety of DNA-based Michael additions to enones were determined under similar conditions as described by Boersma et al.3 Freshly dialyzed st-DNA (final concentration 0.67 mg mL-1 in 20mM MOPS pH 6.5) was used, and this was pre-equilibrated at 5°C for 18h with Cu2+-L4 at a final concentration of 0.15 mM (1 complex per ~6-7 base pairs). After transfer of the solution thus obtained to a quartz cuvette, the enone substrate was added to give a final concentration of 6.0 μM (Michael additions) or 14 μM (Friedel-Crafts reactions). The appropriate amount of the nucleophile (typically to a final concentration of 0.9 mM (dimethylmalonate), 9.0 mM (nitromethane) or 0.5 – 2.0 mM (4.4)) was then added as a stock solution in acetonitrile (nitromethane and indoles 4.4) or 20 mM MOPS pH 6.5 (dimethylmalonate) and the cuvette was gently shaken. The UV absorption was monitored at 326nm (Michael additions to 4.2), 335nm (Michael additions to 4.1a) or 340 nm (Friedel-Crafts alkylations to 4.1a-f) at a temperature of 25.0±0.2°C. For the reactions with half-lives of more than approx. one hour, apparent rate constants were determined by initial-rate kinetics with the following expression:34

in which d(Aenone)/dt is the slope of the decrease of absorption in time, d the pathlength of the cuvette (1 cm), Δε the difference in extinction coefficients of the substrate and the product (determined separately) and [enone]0 is the initial substrate concentration. Apparent second-order rate constants were then deduced from the slope of a plot of these values of k1 versus the concentration of nucleophile. For faster reactions (complete conversion within several hours, viz. Friedel-Crafts alkylations of indoles 4.4 by 4.1e-f), the decrease of absorption in time (At) was curve-fitted using Grafit 3.0 (Erithacus Software Ltd., 1992) to the exponential equation At = A∞ + A0·e-k1·t, giving apparent rate constants k1 directly.

,]enone[Δ

1d

)(d

0

enone1 ⋅⋅

⋅=εdt

Ak

Kinetic Investigations of DNA-based Conjugate Additions

103

Synthesis of substrates and identification of products. Enone substrates 4.1a-f35 and 4.234 were synthesized according to known procedures. Indoles 4.4a-b were commercially available from Sigma-Aldrich and used without further purification. Analytical data of adducts 4.3a-b (chapter 3), 4.5b and 4.5d-f (refs. 4 and 35) were in accordance with literature. Addition products, for which initial-rate kinetics were used to study rate constants, were synthesized independently as a racemate, starting from 0.25 mmol of enone, 0.25 mmol of copper (II) nitrate and 5 eq. of the appropriate nucleophile in 5 mL of distilled water overnight. The adducts were purified by flash-column chromatography and their molar extinction coefficients were determined in a concentration range close to that of the catalytic reactions (typically 4 – 40 μM). 3-(5-methoxy-1H-indol-3-yl)-1-(1-methyl-1H-imidazol-2-yl)-3-phenyl-1-

propanone 4.5a was purified by flash column chromatography (SiO2, hexane / ethyl acetate 1:2, Rf 0.25), giving a light brown solid, mp 139°C (dec.); 1H NMR: δ 8.07 (br s, 1H), 7.38 (d, J=8.0 Hz, 2H), 7.20-7.26 (m, 2H), 7.11-7.20 (m, 3H), 7.09 (d, J=2.4 Hz, 1H), 6.98 (s, 1H), 6.92 (d, J=2.2 Hz, 1H), 6.78 (dd, J=8.8, 2.4 Hz,

1H), 5.01 (t, J=7.6 Hz, 1H), 4.01 (dd, J=16.3, 7.4 Hz, 1H), 3.90 (s, 3H), 3.82 (dd, J=16.4, 7.8 Hz, 1H), 3.75 (s, 3H); 13C NMR: δ 190.9 (s), 153.5 (s), 144.2 (s), 143.0 (s), 131.6 (s), 128.7 (d), 128.2 (d), 127.8 (d), 126.98 (s), 126.95 (d), 126.0 (d), 122.2 (d), 118.7 (s), 111.8 (d), 111.7 (d), 101.1 (d), 55.6 (q), 45.3 (t), 37.9 (d), 35.9 (q) MS (ESI) m/z 382 ([M+Na]+), 360([M+H]+), 236, 213; HRMS calcd for C22H22N3O2 ([M+H]+): 360.1707, found: 360.1700; chiral HPLC: Chiralpak AD, heptane/i-PrOH 80/20, flow 1.0 mL min-1, Tr 19.1, 26.9 min. 3-(4-bromophenyl)-3-(5-methoxy-1H-indol-3-yl)-1-(1-methyl-1H-imida-

zol-2-yl)-1-propanone 4.5c was purified by flash column chromatography (SiO2, hexane / ethyl acetate 1:2, Rf 0.29), giving a pale yellow oil; 1H NMR: δ 8.50 (br s, 1H), 7.55 (d, J=7.9 Hz, 1H), 7.12-7.22 (m, 3H), 7.09 (t, J=7.5 Hz, 1H), 7.04 (d, J=2.4 Hz, 1H), 6.96-7.01 (m, 3H), 6.77 (dd, J=2.4 Hz, 8.8, 1H), 5.47 (t, J=7.7 Hz,

N

N

O

CH3N

H

OCH3

N

N

O

CH3N

H

OCH3

Br

Chapter 4

104

1H), 4.18 (dd, J=8.1, 16.2 Hz, 1H), 3.90 (s, 3H), 3.76 (s, 3H), 3.56 (dd, J=6.7, 16.2 Hz, 1H); 13C NMR: δ 190.4 (s), 153.7 (s), 143.2 (s), 143.1 (s), 132.7 (d), 131.6(s), 129.5 (d), 128.9 (d), 127.7 (d), 127.2 (s), 122.4 (d), 124.1 (s), 118.0 (s), 112.3 (d), 111.7 (d), 101.3 (d), 55.7 (q), 44.4 (t), 37.2 (d), 36.1 (q); MS (ESI) m/z 462 ([M+Na]+), 460 ([M+Na]+), 440 ([M+H]+), 438 ([M+H]+), 422, 420, 316, 314, 293, 291, 122; HRMS calcd for C22H21BrN3O2 ([M+H]+): 438.0812, found: 438.0805; chiral HPLC: Chiralpak AD, heptane/i-PrOH 90/10, flow 1.0 mL min-1, Tr 37.1, 46.8 min. 3-(5-methoxy-1H-indol-3-yl)-1-(1-methyl-1H-imidazol-2-yl)-1-octanone

4.5g was purified by flash column chromatography (SiO2, hexane / ethyl acetate 2:1, Rf 0.21), giving a brown oil. 1H NMR: δ 7.83 (br s, 1H), 7.20 (d, J=9.0 Hz, 1H), 7.12 (d, 0.91, 1H), 7.05 (dd, J=2.4, 17.7 Hz, 2H), 6.96 (s, 1H), 6.81 (dd, J=2.4, 8.7 Hz, 1H), 3.85 (s, 3H), 3.82 (s, 3H), 3.63 – 3.69 (m, 1H), 3.44-3.57 (m, 2H), 1.71-1.82 (m, 2H), 1.20-1.31 (m, 6H), 0.82

(t, 7.9, 3H), 13C NMR: δ 192.3 (s), 153.5 (s), 143.2 (s), 131.4 (s), 128.6 (d), 127.3 (s), 126.7 (d), 121.8 (s), 119.1 (s), 111.7 (d), 111.6 (d), 101.0 (d), 55.8 (q), 45.6 (t), 36.0 (t), 36.0 (d), 32.2 (q), 31.9 (t), 27.1 (t), 22.5 (t), 14.0 (q); MS (ESI) m/z 376 ([M+Na]+), 354 ([M+H]+), 149; HRMS calcd for C21H28N3O2+ ([M+H]+): 354.2176, found: 354.2184; chiral HPLC: Regis (R,R)-Whelk-O 1, 0.04% diethylamine in heptane/ i-PrOH 80/20, flow 0.5 mL min-1, Tr 35.6, 42.3 min.

4.7 Notes and References

1 A.J. Boersma, D. Coquière, D. Geerdink, B.L. Feringa, G. Roelfes, manuscript in preparation. 2 D. Coquière, B.L. Feringa, G. Roelfes Angew. Chem. Int. Ed. 2007, 46, 9308 and Chapter 3 of this thesis. 3 A.J. Boersma, J.E. Klijn, B.L. Feringa, G. Roelfes J. Am. Chem. Soc. 2008, 130, 11783. 4 A.J. Boersma, B.L. Feringa, G. Roelfes Angew. Chem. Int. Ed. 2009, 48, 3346. 5 G. Roelfes, A.J. Boersma, B.L. Feringa Chem. Commun. 2006, 635. 6 S. Kobayashi, K. Kakumoto, Y. Mori, K. Manabe Isr. J. Chem. 2001, 41, 247. 7 Y. Hamashima, D. Hotta, N. Umebayashi, Y. Tsuchiya, T. Suzuki, M. Sodeoka Adv. Synth. Catal. 2005, 347, 1576.

N

N

O

CH3N

H

OCH3

Kinetic Investigations of DNA-based Conjugate Additions

105

8 (a) K. Aplander, R. Ding, U.M. Lindström, J. Wennerberg, S. Schultz Angew. Chem. Int. Ed. 2007, 46, 4543; (b) K. Aplander, R. Ding, M. Krasavin, U.M. Lindström, J. Wennerberg Eur. J. Org. Chem. 2009, 810. 9 (a) A. Dalla Cort, L. Mandolini, L. Schiaffino Chem.Commun. 2005, 3867; (b) A. Dalla Cort, L. Mandolini, L. Schiaffino J. Org. Chem. 2008, 73, 9439. 10 The solubility of nitromethane in the buffer is too low to prepare a stock solution concentrated enough to be able to use a small aliquot when adding the amount of nitromethane needed (1.0-2.0·103 eq.). 11 The pKA values reported for nitromethane range from 10 to 10.2: (a) R.P. Bell The Proton in Chemistry, 2nd ed., Chapman and Hall, London, 1973, p. 106. (b) D. Turnbull, S.H. Maron J. Am. Chem. Soc. 1943, 65, 212. (c) T. Matsui, L.G. Hepler Can. J. Chem. 1973, 51, 1941. 12 For the homologous diethylmalonate, pKa values ranging from 12.9 to 13.3 were reported: (a) A. Albert, E.P. Serjeant The Determination of Ionization Constants – A Laboratory Manual, 2nd ed., Chapman and Hall, London, 1971, p. 90; (b) R.G. Pearson, R.L. Dillon J. Am. Chem. Soc. 1953, 75, 2439. 13 (a) In water: T. Bug, H. Mayr, J. Am. Chem. Soc. 2003, 125, 12980; (b) In DMSO and in water: T. Bug, T. Lemek, H. Mayr J. Org. Chem. 2004, 69, 7565; (c) In methanol: T.B. Phan, H. Mayr Eur. J. Org. Chem. 2006, 2530. 14 (a) M.H. Davies, B.H. Robinson, J.R. Keeffe, Annu. Rep. Prog. Chem., Sect. A 1973, 70, 123; (b) ref. 11a, pp. 208-214; (c) J.R. Keeffe, J. Morey, C.A. Palmer, J.C. Lee J. Am. Chem. Soc. 1979, 101, 1295. 15 C.F. Bernasconi Acc. Chem. Res. 1992, 25, 9. 16 C.A. Franklin, J.V. Fry, J.G. Collins Inorg. Chem. 1996, 35, 7541. 17 J.V. Fry, J.G. Collins Inorg. Chem. 1997, 36, 2919. 18 J.G. Collins, A.D. Sleeman, J.R. Aldrich-Wright, I. Greguric, T.W. Hambley Inorg. Chem. 1998, 37, 3133. 19 Calculated using the expression ΔΔ‡G = -RT Ln (er), in which R is the gas constant, T the temperature and er the enantiomeric ratio. For the DNA-based Diels-Alder reaction between 4.2 and cyclopentadiene, a linear relationship between ΔΔ‡G and log kapp was observed: Ref. 3. 20 For a recent review on the Lewis-acid catalyzed asymmetric Friedel-Crafts reaction, see T.B. Poulsen, K.A. Jørgensen Chem. Rev. 2008, 108, 2903. 21 (a) K. Manabe, N. Aoyama, S. Kobayashi Adv. Synth. Catal. 2001, 343, 174; (b) W. Zhuang, K.A. Jørgensen Chem. Commun. 2002, 1336; (c) N. Azizi, F. Arynasab, M.R. Saidi Org. Biomol. Chem. 2006, 4, 4275; (d) S. Shirakawa, S. Kobayashi Org. Lett. 2006, 8, 4939.

Chapter 4

106

22 K.B. Jensen, J. Thorhauge, R.G. Hazell, K.A. Jørgensen Angew. Chem. Int. Ed. 2001, 40, 160. 23 D.A. Evans, K.R. Fandrick, H.-J. Song J. Am. Chem. Soc. 2005, 127, 8942. 24 D.A. Evans, K.R. Fandrick, H.-J. Song, K.A. Scheidt, R. Xu J. Am. Chem. Soc. 2007, 129, 10029. 25 It should be noted that these substrates undergo significant conjugate addition of water under the reaction conditions, see Ref. 1. The rate constants of the Friedel-Crafts reactions reported here have been corrected for this, using variable concentrations of indoles 4.4 and extracting kapp values from the slope of the plot of k1 vs. [4.4]; see general procedure 2. 26 J. Chen, Q. Deng, R. Wang, K.N. Houk, D. Hilvert ChemBioChem 2000, 1, 255. 27 The slope of the plot thus obtained equals -Δ‡H/R, the x-intercept equals ln(kB/h)+ Δ‡S/R, in which R is the gas constant, kB Boltzmann’s constant and h Planck’s constant. 28 Binding of the complex Cu2+-L1 to DNA has been shown to change only marginally over this temperature range, see ref. 3. 29 (a) H. Maskill The Physical Basis of Organic Chemistry, Oxford University Press: Oxford, 1989, p. 250. A negative heat capacity of activation has been observed

frequently in solvolyses of alkyl halides and sulfonates: (b) R.E. Robertson Solvolysis in Water, in Progr. Phys. Org. Chem., A. Streitwieser, Jr., R.W. Taft, Eds., Wiley Interscience: New York, 1967, Vol. 4, pp. 224-235; (c) solvolysis of alkyl nitrates: R.E. Robertson, K.M. Koshy, A. Annessa, J.N. Ong, J.M.W. Scott, M.J. Blandamer Can. J. Chem. 1982, 60, 1780; and (d) in enzymatic reactions, e.g. the oxidative deamination of L-glutamate catalyzed by L-glutamate dehydrogenase: A.H. Colen, R.T. Medary, H.F. Fisher Biopolymers 1981, 20, 879. 30 K.L. Brown, J. Li J. Am. Chem. Soc. 1998, 120, 9466. 31 Ref. 29b; pp. 216-223. 32 E.A. Meyer, R.K. Castellano, F. Diederich Angew. Chem. Int. Ed. 2003, 42, 1210. 33 The 10% ee difference between e.g. entries 4 and 5 of Table 4.6 corresponds to a difference in ΔΔ‡G for the diastereomeric activation complexes of 2.2 kJ mol-1 at 298K. This value was calculated using the expression ΔΔG‡ = - RT ln (er), in which er is the enantiomeric ratio. 34 S. Otto Ph.D. Thesis, University of Groningen, 1998, pp. 64-66. 35 M.C. Myers, A.R. Bharadwaj, B.C. Milgram, K.A. Scheidt J. Am. Chem. Soc. 2005, 127, 14675.