Efficient intramolecular hydroalkoxylation of unactivated alkenols mediated by recyclable lanthanide lriflate ionic liquids: Scope and mechanism

FULL PAPER

DOI: 10.1002/chem.200902269

Efficient Intramolecular Hydroalkoxylation of Unactivated Alkenols

Mediated by Recyclable Lanthanide Triflate Ionic Liquids:

Scope and Mechanism

Alma Dzudza^[b]and Tobin J. Marks*^[a]

Abstract: Lanthanide triflate com-

plexes of the type [Ln(OTf)₃] (Ln=La,

rapid than those of internal alkenols,

which suggests modest steric demands

in the cyclic transition state. Cycliza-

tion rates of aryl-functionalized hy-

droxyalkenes are more rapid than

those of the linear alkenols, whereas

five- and five/six-membered spirobicy-

clic skeletons are also regioselectively

closed. In cyclization of primary, steri-

cally encumbered alkenols, turnover-

frequency dependence on metal-ionic

radius decreases by approximately 80-

fold on going from La³⁺(1.160 ꢀ) to

Lu³⁺(0.977 ꢀ), presumably reflecting

steric impediments along the reaction

coordinate. The overall rate law for al-

kenol hydroalkoxylation/cyclization is

vꢁk

[catalyst]¹[alkenol]¹. An observed

ACHUTGTNRENNUG CAHTUNGTRENNUGN

G

ROH/ROD kinetic isotope effect of

2.48 (9) is suggestive of a catalytic

pathway that involves kinetically signif-

icant intramolecular proton transfer.

The present activation parameters—en-

thalpy (DH^°)=18.2 (9) kcalmol^ꢀ1, en-

tropy (DS^°)=ꢀ17.0 (1.4) eu, and

energy (E_a)=18.2 (8) kcalmol^ꢀ1—sug-

gest a highly organized transition state.

Proton scavenging and coordinative

probing results suggest that the lantha-

nide triflates are not simply precursors

of free triflic acid. Based on the kinetic

and mechanistic evidence, the pro-

posed catalytic pathway invokes hy-

droxyl and olefin activation by the

electron-deficient Ln³⁺center, and in-

tramolecular H⁺transfer, followed by

alkoxide nucleophilic attack with ring

closure.

Sm, Nd, Yb, Lu) serve as effective, re-

cyclable catalysts for the rapid intramo-

lecular hydroalkoxylation (HO)/cycli-

zation of primary/secondary and ali-

phatic/aromatic hydroxyalkenes in imi-

dazolium-based

room-temperature

ionic liquids (RTILs) to yield the corre-

sponding furan, pyran, spirobicyclic

furan, spirobicyclic furan/pyran, benzo-

furan, and isochroman derivatives.

Products are straightforwardly isolated

from the catalytic solution, conversions

exhibit Markovnikov regioselectivity,

and turnover frequencies are as high as

47 h^ꢀ1at 1208C. The ring-size rate de-

pendence of the primary alkenol cycli-

zations is 5>6, consistent with a steri-

cally controlled transition state. The

hydroalkoxylation/cyclization rates of

terminal alkenols are slightly more

Keywords: alkenols · hydroalkoxy-

lation · ionic liquids · lanthanide

triflates · reaction mechanisms

Introduction

formation for accessing oxygen-containing heterocycles.^[1]

Such heterocycles are important structural components of

naturally occurring and biologically active molecules includ-

ing acetogenins, prostaglandins, polyether ionophore antibi-

otics, and macrocyclic natural products.^[2]Currently, oxygen

heterocycles are catalytically accessible by either direct

Wacker oxidative cyclization^[3]or by means of intramolecu-

lar hydroalkoxylation (HO),^[4–7]with the latter being a rela-

tively unexplored transformation mediated by a limited

number of catalysts. In principle, intramolecular hydroalkox-

ylation is a concise, elegant route that should be possible to

effect with 100% atom efficiency and with negligible waste

coproduction. Thus, HO/cyclization fulfills green chemical

requirements more satisfactorily than conventional byprod-

uct-producing substitution reactions.^[1,8]However, although

catalytic intramolecular hydroalkoxylation offers many at-

ꢀ

Catalytic regioselective O H addition to the C=C bonds of

pendant olefins is a highly desirable, atom-economical trans-

[a] Prof. T. J. Marks

Department of Chemistry and Material Science

Northwestern University

2145 Sheridan Road, Evanston IL 60208-3113 (USA)

Fax : (+1)847-491-2990

E-mail: t-marks@northwestern.edu

[b] Dr. A. Dzudza

Department of Chemistry, Northwestern University

2145 Sheridan Road, Evanston IL 60208-3113 (USA)

Supporting information for this article, including experimental details

and kinetic and mechanistic studies, is available on the WWW under

http://dx.doi.org/10.1002/chem.200902269.

Chem. Eur. J. 2010, 16, 3403 – 3422

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

3403

10 mol% [PdCl₂ACHTNUGRTENUNG(MeCN)₂] af-

fords b-alkoxyketones in high

yield [Eq. (3)].

Subsequently, the Furuka-

wa^[5f]and Oe^[5b,c]groups em-

ployed RuCl₃·nH₂O^[5f]and

[(Cp*RuCl₂)₂]^[5b,c](Cp*=pen-

tamethylcyclopentadienyl) cat-

alysts to generate benzylic

cyclic ethers by means of intra-

molecular HO [Eqs. (4) and

(5)] in varying yields.

tractions, efficient transformations remain elusive due fac-

ꢀ

tors such as the large O H bond enthalpies, the modest re-

activity of electron-rich olefins with nucleophiles, high cata-

lyst costs, and catalyst metal toxicity.^[5,6]There has been a

growing research effort to address the aforementioned

issues and to develop more efficient and selective hydroal-

koxylation catalysts for this attractive transformation.^[4–7]

For intramolecular alkenol HO/cyclizations, two possible

products, exemplified by the exo and endo ethers shown in

Equation (1), are a priori possible where either exo-trig or

endo-trig cyclization to five- or six-membered rings is al-

lowed by the Baldwin ring-closure rules.^[9]

Moreover, RuCl₃·nH₂O catalytic activity is moderately en-

hanced by addition of AgOTf and CuACTHNUTRGNEUNG(OTf)₂cocatalysts,

thereby improving the yields without compromising selecti-

Although direct catalytic intramolecular OH addition to

alkenes is attractive, such seemingly simple processes have

not generally proven efficient and frequently lack generality.

Catalysts have included Brønsted acids^[4]and transition-

metal complexes,^[5,6]which activate either the hydroxyl or

C=C double bond with varying degrees of efficiency.

DuÇach and co-workers employed catalytic amounts of trifl-

ic acid to generate Markovnikov-type cyclic ethers from un-

activated olefins [Eq. (2)].^[4b]

vity.^[5f]In an example of Pt-catalyzed intramolecular HO/

cyclization, Widenhofer et al. showed that [{PtCl₂ACHTUNGTRENNUNG(CH₂=

CH₂)}₂]/2P(4-C₆H₄CF₃)₃is an effective catalyst for cyclizing

a range of g- and d-hydroxyolefins to yield highly substitut-

ed furans and pyrans [Eq. (6)].^[5d]The selectivity of the Pt

system contrasts markedly with Pd-based catalysts, which

tend to give oxidized products by means of Wacker-type

oxypalladation/b-hydride elimination sequences.^[3]

However, several authors,^[5c,10]including Hartwig,^[4a]re-

ported this approach to be of limited synthetic utility since

controlling the acid concentration is crucial for acceptable

yields. Of the effective intermolecular HO processes, transi-

tion-metal catalysis offers the potential to achieve selective

olefin HO to Markovnikov products under mild condi-

tions.^[5,6]The first example was reported by Murahashi and

co-workers in 1989.^[12]Reaction

Recently, He et al. made an important advance towards a

general intermolecular hydroalkoxylation approach;^[5a]they

reported that combining 2 mol% Ph₃PAuCl and 2 mol%

AgOTf yields Ph₃PAuOTf, which in turn catalyzes a variety

of hydroalkoxylation processes [Eq. (7)]; both endo and exo

product formation is observed, with the latter formed in

minor quantities.^[5a]

of methyl vinyl ketones with

BnOH in the presence of

3404

www.chemeurj.org

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2010, 16, 3403 – 3422

Hydroalkoxylation of Unactivated Alkenols

FULL PAPER

A variety of metal triflates have also been shown to effec-

tively mediate intramolecular cyclization.^[6]For example,

AgOTf, often used as a cocatalyst in Ru-catalyzed intramo-

lecular HO/cyclizations of inert olefins, also mediates the

transformation under relatively mild conditions [Eq. (8)].^[6d]

Moreover, DuÇach et al., in a combined experimental and

hydroamination,^[13]hydrophosphination,^[14]hydrosilation,^[15a]

hydroboration,^[15b]and more recently, hydroalkoxylation.^[7]

Moreover, hydroelementation can be coupled with catalytic

polymerization, so that a variety of functional groups can be

introduced into otherwise inert polyolefins.^[17a,b]Lastly, indi-

ꢀ

vidual C N and C C bond-forming reactions can be tied in

theoretical study, showed AlACHTNUTRGNENG(U OTf)₃to be an efficient cata-

sequences to effect tandem processes.^[13a,b,16]

lyst for intramolecular cyclization of unactivated olefins

Recently, our group reported that homoleptic [Ln{N-

AHCTUNGTRENN(GUN SiMe₃)₂}₃] complexes are effective and selective precatalysts

[Eq. (9)].^[6b]

for intramolecular HO/cyclization of alkynyl and allenyl al-

cohols.^[7]These precursors undergo instantaneous protonoly-

sis at 258C to yield the corresponding alkoxides and free

HNACTHNUTRGNEUNG

(SiMe₃)₂(Scheme 1).^[7]Mechanistic data implicate turn-

ꢀ

over-limiting insertion of C C unsaturation into the Ln O

ꢀ

bond (step i, Scheme 1) with subsequent, rapid Ln C proto-

nolysis (step ii, Scheme 1) to release product heterocycle

and regenerate the catalyst.^[7]Although bonding energetic

considerations^[18]predict net exothermic processes for all al-

[18a,c]

ꢀ

cohols (see Scheme 1), the large Ln O bond enthalpy

Nevertheless, due to the limitations of the current synthet-

ic methodologies, for example, high catalyst cost,^[4,5]metal

toxicity, and lack of catalyst recyclability, the development

of new, more efficient catalysts and catalytic methods for

cyclic ether synthesis presents an intriguing challenge.

Lanthanides exhibit a number of distinctive as well as po-

tentially informative and useful characteristics for activating

carbon–carbon unsaturation as well as amine, phosphine,

and hydroxyl functionalities.^[7,12–16]Unique lanthanide-ion

characteristics include high electrophilicity, very large ionic

radii (which afford high coordination numbers and coordi-

native unsaturation), relatively constrained yet tunable an-

cillary ligation, and facile bond

renders the insertive step i rather endothermic for alkenols,

and likely for this reason, rapid intramolecular alkenol HO/

cyclization has not yet been observed.^[7]

In the past decade, lanthanide triflates, [LnACTHNUGTRENUNG(OTf)₃], envi-

ronmentally friendly f-block element Lewis acids, have

emerged as important reagents and catalysts.^[19–22]Kobayashi

and co-workers popularized the application of lanthanide

triflates as catalysts in Aldol reactions, Diels–Alder, retro-

and aza- Diels–Alder reactions, allylation, and so on.^[19,20,22]

Lanthanide ions have far larger ionic radii and formal coor-

dination numbers than typical transition-metal ions,^[20a–c]and

are very strong, hard Lewis acids. Nevertheless, lanthanide

activation by means of concert-

ed four-centered s-bond meta-

thesis processes, rather than by

two-electron oxidative addi-

tion/reductive elimination se-

quences.^[12]Among these cata-

lysts, homoleptic amides, for

example, [Ln{N

lanthanocenes, for example,

[Cp*₂LnE(SiMe₃)₂], are versa-

ACHTUNGTRENNUNG(SiMe₃)₂}₃], and

ACHTUNGTRENNUNG

ꢀ

tile agents for a variety of C C

ꢀ

and C heteroatom bond-form-

ing transformations, which can

be either intermolecular or in-

tramolecular, and can exhibit

very high turnover frequencies,

versatile reaction scope in

terms of heteroatom substitu-

tion pattern and ring size, as

well as high stereoselectiv-

ity.^[7,13–15]Established catalytic

transformations include, but

are not limited to, polymeri-

zation processes,^[17]and hydro-

elementation processes such as

Scheme 1. Proposed catalytic cycle for representative homoleptic lanthanide amido-catalyzed hydroalkoxyla-

tion/cyclization of alkyne- and allene-bearing alcohols (TMS=SiMe₃).

Chem. Eur. J. 2010, 16, 3403 – 3422

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

3405

A. Dzudza and T. J. Marks

triflates are remarkably stable to hydrolysis,^{[20a–c,21,22]}and the

strongly electron-withdrawing properties of the triflate

counterion suggest an excellent leaving group.^[19,21b,22]In

most cases, only substoichiometric (catalytic) quantities of

ILs are used in catalysis as the actual catalyst [Eq. (10)],

as a ligand [Eq. (11)], or as the solvent [Eq. (12); acac=ace-

tylacetonate].^[25–29]Chloroaluminate

ACTHNUTRGNEUNG(III) ionic liquids,

[C_nmim]Cl–AlCl₃(C_nmim=1-alkyl-3-methylimidazolium),

are heavily exploited and well-studied IL catalysts,^[28]with

Diels–Alder cycloaddition first demonstrated by Wolf

et al.^[28d]Here rate enhancement effects are attributed to

IL–polar substrate hydrogen-bonding.^[28d]

[LnACHTUNGTRENNUNG

(OTf)₃] effect useful catalytic turnover,^[19,21b,22]and they

can be easily recovered on reaction completion and recycled

with negligible loss of activity.^[19,21b,22]Nevertheless, conven-

tional [LnACHTUNGTRENNUNG(OTf)₃]-catalyzed processes typically require toxic,

polar, moderately coordinating

solvents such as nitromethane

and acetonitrile, which raise

significant environmental and

economic issues.^[23]

Ionic liquids (ILs) have gen-

erated increasing research in-

terest as green alternatives to

volatile organic solvents.^[24,25]

They are comprised solely of

positive

and

negatively

charged ions,^[24,25]and typically

have melting points below ap-

proximately 1008C.^[24]Several

types of ILs are free-flowing

fluids at room temperature be-

cause large asymmetric constit-

uent ions hinder close packing and thus lower the lattice

energy.^[24]These are termed room-temperature ionic liquids

(RTILs).^[24]IL physical and chemical properties depend on

both the nature of the cation and anion,^[24]with typical or-

ganic cations being bulky, asymetric imidazolium, pyridini-

um, ammonium, phosphonium, or other heteroaromatics,

with low symmetry, weak intermolecular interactions, and

delocalized/screened charge densities (Scheme 2).^[24]Popular

RTILs as ligands were first reported by Welton et al., who

successfully performed Pd-catalyzed arene–arene Suzuki

cross-coupling reactions in [C₄mim]ACHTNUGRTNEUNG

[BF₄] [Eq. (11)].^[29a]

Analysis of the solutions revealed the presence of a Pd

phosphine-imidazolidene complex.^[28b]The other major role

played by ILs in catalytic processes is as the solvent.^[25,27,28]

They have been exploited in hydrogenation by de Souza

et al.^[30a]and Chauvin et al.,^[30b]hydroformylation by Song

and Roh [Eq. (12)],^[30c]oxidation by Chauvin et al.,^[30b]Heck

reactions by Kaufmann et al.,^[30d]and in many other process-

es.^[24,25]As solvents, ILs can provide better solvation by

means of large dipolar and dispersion forces, and by means

of hydrogen bonding.^[24,25,27]IL hydrogen-bond basicity is

dominated by the anion, whereas hydrogen-bond acidity is

dominated by the cation.^[24,25,27]Moreover, ILs with aromatic

ions are capable of p–p and n–p interactions with sol-

utes.^[24,25,27]In addition to the aforementioned attractions,

ILs are thermally stable fluids, frequently allowing higher-

reaction temperatures than conventional solvents, and ease

of product separation by extraction or distillation.^[24,25,30]

Proper IL choice can afford larger turnover frequencies,

higher yields, and enhanced selectivity.^[30]

Scheme 2. Common ionic liquid cations and anions.

ꢀ

anions include Cl^ꢀ, Br^ꢀ, BF₄, PF₆, or CF₃SO₃, the proper-

ties of which, such as hydrogen-bond basicity, coordinative

ability, and hydrophobicity, are readily manipulated

(Scheme 2).^[24]In addition to high thermal stability and low

volatility, ionic liquids have dielectric constants ranging

from 10 to 28,^[26]are usually aprotic, and often afford unique

reaction efficiencies and selectivities.^[24,25]Since IL proper-

ties such as miscibility with water and other solvents, solva-

tion characteristics, polarity, viscosity, and density are tuna-

ble by anion and cation choice, they are considered to be

“designer solvents.”^[24]

For [LnACHTUNTGRNE(UNG OTf)₃]-mediated intramolecular alkenol HO/cyc-

lization, RTILs based on noncoordinating anions should en-

hance Ln³⁺Lewis acidity due to minimized solvation

(Scheme 3), and offer catalyst and IL recyclability, as well as

ease of product separation by vacuum transfer or ether ex-

traction. In a recent a communication, we briefly reported

for a limited set of substrates, that in imidazolium-based ILs,

[LnACTHNUGTRNENG(U OTf)₃]-mediated alkenol HO/cyclizations are efficient

processes for cyclic ether synthesis with Markovnikov selec-

tivity.^[31]Herein we present a detailed report on the broad

3406

www.chemeurj.org

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2010, 16, 3403 – 3422

Hydroalkoxylation of Unactivated Alkenols

FULL PAPER

1!2 catalytic conversion in

RTILs based on weakly coordi-

ꢀ

nating OTf^ꢀand NTf₂anions

proceeds with large rate en-

hancements, with approximate-

ly 70-fold increases in turnover

frequencies (Table 1, entries 6–

11). There is a general consen-

sus that RTILs, in particular

those based on imidazolium

cations, have polarities compa-

rable to those of short-chain

alcohols and coordinative ten-

Scheme 3. Enhancement of Ln³⁺Lewis acidity due to minimization of solvent coordination.

reaction scope in terms of the ring size, substitution pattern,

and selectivity of heterocycle formation, kinetics, and mech-

Table 1. Screening of [Ln

molecular hydroalkoxylation/cyclization of 1.

ACHTUNGRTENUN(NG OTf)₃] complexes and reaction media for intra-

anism of [LnACHTUNGTRENNUNG(OTf)₃]-catalyzed intramolecular HO/cycliza-

tion of unactivated alkenols in RTILs. This full account in-

cludes discussion of reaction scope, selectivity, Ln ion ef-

fects, rate law, kinetic isotope effects, activation energetics,

and proton-scavenging effects for this new catalytic process.

In doing so, we compare the results with previously charac-

terized Ln-mediated intramolecular hydroamination/cycliza-

tion and hydroalkoxylation/cyclization processes.

Entry Ln

G

M³⁺Solvent

e^[a]

N_t[h^ꢀ1

(T [8C])^[c]

]

1

La

1.172 CD₃NO₂

1.098 CD₃NO₂

1.008 CD₃NO₂

36.4 0.01 (100)^[d]

36.4 0.04 (100)^[d]

36.4 0.10 (100)^[d]

2

3

4

Sm

Yb

–

[b]

–

A

[OTf]^[e]

E

– (120)

0.67 (120)

5.37 (120)

6.37 (120)

6.01 (120)

1.18 (120)

1.07 (120)

5

–

A

[OTf]^[f]15.1

[b]

6

7

8

9

10

11

La

Yb

1.172

1.008

N

E

[b]

Results

E

T

ACHTUNGTRENNUNG

The principal goal of this contribution is to examine the

scope, selectivity, and mechanism of [LnACTHNURTGNENG(U OTf)₃]-mediated

ACHTUNGTRENNUNG

HO/cyclization of unactivated alkenols in imidazolium-

based room-temperature ILs, to define the sequence of

events that lead to the facile oxygen heterocycle formation.

This section begins with observations concerning reaction-

[a] Solvent dielectric constant at 258C. [b] Data not available in the liter-

ature. [c] Turnover frequencies determined by ¹H NMR spectroscopy

using 1,1,2,2-tetrachloroethane as internal standard. [d] Percent forma-

tion of the final product was determined by ¹H NMR spectroscopy inte-

gration versus internal standard due to low observed product formation.

medium effects on the efficiency of the [LnACTHNURTGNENG(U OTf)₃]-catalyzed

[e] [1-Ethyl-3-methylimidazolium]

ACHTUNGTRENNUNG

intramolecular HO/cyclization, followed by examination of

the broad reaction scope in terms of the ring size, substitu-

tion pattern, and selectivity of heterocycle formation. Next,

the dependence of reaction turnover frequency on Ln³⁺

ionic radius is discussed. Finally, a kinetic and mechanistic

study of the HO/cyclization process includes examination of

molecularity, activation parameters, OH/OD kinetic isotope

effects, proton trapping, coordinative probing, and the

mechanistic implications of these results.

G

R

[j] 1-Ethyl-3-methylimidazolium trifluoromethanesulfonyl amide. [i] 1-

Ethyl-3-methylimidazolium trifluoromethanesulfonyl amide.

dencies similar to CH₂Cl₂.^[24,25]Among these ILs, 1,3-dialkyl

imidazolium RTIL derivatives based on OTf^ꢀand

ꢀ

NTf₂anions are minimally coordinating,^[24,25]thus likely en-

hancing Ln³⁺Lewis acidity/unsaturation, in accord with the

increased HO/cyclization turnover frequencies (Table 1, en-

tries 6–9). The present large rate enhancements in the hy-

Reaction-medium effects on catalytic alkenol HO/cycliza-

drophilic ionic liquid,^[25][C_nmim]

ACHTUNGTRENNUNG

tion: The efficiency of the [Ln

version was first optimized as

(Table 1), and an optimum reaction protocol was identified.

Initial screening of [Ln(OTf)₃]-mediated intramolecular

R

phobic^[25][C_nmim]

ACHTUNGTRENNUNG

a

10), are reasonably attributable to solvation differences,

with further explanation deferred to the Discussion section

below.

AHCTUNGTRENNUNG

HO/cyclization of 1 in nitromethane revealed exclusive for-

mation of 2 [Eq. (13)], Table 1]; the formation of b-hydride

elimination/isomerization products was not observed. [Ln-

Scope of catalytic alkenol intramolecular hydroalkoxylation/

cyclization in [C₂mim]ACTHNUGRTENUNG[OTf]: To probe the scope of the pres-

A

ent catalytic HO/cyclization process, a variety of substrates

were examined with respect to regioselectivity, substituent

effects, and turnover frequency. It can be seen that [Ln-

version 1!2 in nitromethane (Table 1, entries 1–3), with rel-

ative ordering of catalytic activity, Yb³⁺>Sm³⁺>La³⁺. In

marked contrast to the catalytic results in nitromethane, the

AHCTUNGRTEG(NNUN OTf)₃] complexes in RTILs serve as effective catalysts for

Chem. Eur. J. 2010, 16, 3403 – 3422

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

3407

A. Dzudza and T. J. Marks

Table 3. [Ln

sterically encumbered alkenols.

ACHTUGNRTEN(UNNG OTf)₃]/ACHTUNRTEGN[GNUN C₂mim]ACTHNUGTREN[NNGU OTf]-mediated HO/cyclization of primary,

the intramolecular (HO)/cyclization of primary/secondary

and aliphatic/aromatic alkenols to yield the corresponding

furan, pyran, spirobicyclic furan, spirobicyclic furan/pyran,

benzofuran, and isochroman derivatives as summarized in

Tables 2, 3, and 4, in which N_tis the turnover frequency at

Entry

Ln³⁺

Substrate

Product

N_t[h^ꢀ1

]

(T [8C])^[a,b]

1

2

3

Yb

Sm

La

0.023 (90)

0.019 (90)

0.012 (90)

Table 2. [Ln

terminal alkenols in [C₂mim]AHCTNUGTRENNNUG

ACHTUNGTRENNUNG

4

5

6

Yb

Sm

La

0.19 (120)

0.31 (120)

1.63 (120)

Entry

Ln³⁺

Substrate

Product

N_t[h^ꢀ1

]

(T [8C])^[a,b]

7

8

9

Yb

Sm

La

7.88 (120)

0.36 (120)

0.11 (120)

1

2

3

Yb

Sm

La

4.78 (120)

0.52 (120)

0.12 (120)

10

11

12

Yb

Sm

La

5.15 (120)

0.64 (120)

0.31 (120)

4

Yb

1.92 (120)

5

6

7

Yb

Sm

La

1.89 (120)

0.62 (120)

0.36 (120)

13

14

15

Yb

Sm

La

14.89 (120)

1.74 (120)

0.49 (120)

8

9

10

Yb

Sm

La

25.14 (120)

0.75 (120)

0.21 (120)

16

17

18

Yb

Sm

La

30.15 (120)

1.22 (120)

0.48 (120)

11

12

13

Yb

Sm

La

46.97 (120)

1.37 (120)

0.57 (120)

19

20

21

Yb

Sm

La

8.62 (120)

1.04 (120)

0.29 (120)

14

15

16

Yb

Sm

La

9.52 (120)

0.24 (120)

0.09 (120)

22

23

24

Yb

Sm

La

7.61 (120)

0.85 (120)

0.19 (120)

17

18

Yb

La

3.25 (60)

6.48 (60)

[a] Turnover frequencies determined by ¹H NMR spectroscopic integra-

tion versus internal standard. [b] For more details, see the Supporting In-

formation

19

20

21

Yb

Sm

La

6.34 (120)

0.73 (120)

0.22 (120)

and for new compounds, by high-resolution MS (HRMS)

and/or elemental analysis (see the Experimental Section for

details). Preparative-scale reactions were carried out in

50 mL three-neck round-bottomed flasks to afford isolated

products in approximately 90% yield (see the Experimental

Section for details).

[a] Turnover frequencies determined by ¹H NMR spectroscopic integra-

tion versus internal standard. [b] For experimental details, see the Sup-

porting Information.

the indicated temperature. In general, alkenol cyclizations

proceed with near-quantitative conversions and reasonably

[Ln

ACHTUNGTRENNUNG

large turnover frequencies at 1 mol% [Ln

G

in [C₂mim]ACTHUNGTRENNUNG

1–24 h at 60–1208C. In the present study, reaction progress

is conveniently monitored by extracting the reaction aliquots

of a known mass at preset times, recording the ¹H NMR

spectrum, and monitoring intensity changes in the olefinic

resonances, with Cl₂CHCHCl₂used as the internal standard

(see Figure 1). Turnover frequencies were determined from

the slope of the kinetic plots of substrate: catalyst ratio

versus time. The final furan, pyran, spirobicyclic furan, spi-

robicyclic furan/pyran, benzofuran, and isochroman products

can be isolated either by simple ether extraction or by

vacuum distillation. General workup conditions for prepara-

tive-scale reactions involve vacuum distillation of the vola-

tiles from the ionic liquid and/or catalyst. In all cases, prod-

ucts were isolated and purified by flash chromatography,

and characterized by ¹H, ¹³C, and ¹⁹F NMR spectroscopy,

and reasonably high turnover frequencies at 60–1208C

(Tables 2, 3, and 4). Note that this process is effective in the

catalytic formation of five- and six-membered mono- and bi-

cyclic oxygen heterocycles. Note also that [LnACTHNUGTRENUNG(OTf)₃]-cata-

lyzed HO/cyclizations are not restricted to primary terminal

and sterically encumbered hydroxyalkenes (Tables 2 and 3),

and that secondary hydroxyalkenes also undergo cyclization

(Table 4). This methodology is additionally applicable to ar-

omatic hydroxyalkenes, as exemplified by synthesis of ben-

zofurans and isochromans (Table 4). As illustrated in

Tables 2–4), [LnACTHNUTRGNEUNG(OTf)₃]-catalyzed HO/cyclizations are effec-

tive for formation of diverse furans (Table 2, entries 1–4, 8–

13; Table 3, entries 1–3, 7–9, 13–15; Table 4, entries 7–9, 13–

15), pyrans (Table 2, entries 5–7, 14–18, Table 3, entries, 4–6;

Table 4, entries, 10–12, 16–18), spirobicyclic furans (Table 2,

3408

www.chemeurj.org

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2010, 16, 3403 – 3422

Hydroalkoxylation of Unactivated Alkenols

FULL PAPER

Table 4. [LnACHTUNGTRENNUNG

entries 17 and 18; Table 3, entries 19–21), spirobicyclic

furans/pyrans (Table 3, entries 16–18, 22–24), benzofuran

(Table 4, entries 1–3), and isochroman (Table 4, entries 4–6).

These are formed cleanly with the Markovnikov-type selec-

tivity, also observed in Ru-,^[5c]Pt-,^[5d]Au-,^[5e]and Brønsted

acid^[4b]based catalytic systems.

[C₂mim]

ACHTUNGTRENNUNG

Entry

Ln³⁺

Substrate

Product

N_t[h^ꢀ1

]

(T [8C])^[a,b]

1

2

3

Yb

Sm

La

6.37 (120)

1.58 (120)

0.82 (120)

The ring-size dependence of [LnACHTUNRGTNEUNG(OTf)₃]/RTIL-catalyzed

4

5

6

Yb

Sm

La

2.46 (120)

0.62 (120)

0.11 (120)

cyclization rates for the present primary and secondary hy-

droxyalkenes is 5>6, which is consistent with a classical,

sterically controlled ring-forming transition state (Table 2,

entry 1 vs. 5, 11 vs. 14; Table 4, entry 13 vs. 16). For five-

membered ring formation, the 3!4, 11!12, and 41!42

conversions (N_t=4.78, 46.97, and 25.32 h^ꢀ1, respectively) are

significantly more rapid than analogous six-membered mole-

cules 13!14, 15!16, and 43!44 (N_t=1.89, 9.52, and

2.75 h^ꢀ1, respectively), with N_tincreasing by approximately

three- to ten-fold. Interestingly, the ionic-radius-related N_t

variations are markedly smaller than the approximately 10³-

fold range observed in organolanthanide-mediated aminoal-

kene hydroamination/cyclization,^[13b,e]and larger than the

approximately 20-fold effect range observed in organolan-

thanide-mediated hydroxyalkyne and hydroxyallene HO/

cyclizations.^[7]

7

8

9

Yb

Sm

La

9.24 (120)

1.14 (120)

0.51 (120)

10

11

12

Yb

Sm

La

4.88 (120)

2.97 (120)

0.52 (120)

13

14

15

Yb

Sm

La

25.32 (90)

2.03 (90)

1.25 (90)

16

17

18

Yb

Sm

La

2.75 (120)

0.48 (120)

0.11 (120)

For substrates bearing geminal dimethyl and diphenyl

substituents, significant rate enhancements are attributed to

angle compression effects (“Thorpe–Ingold effect,” Table 2,

entries 8–13, 20–22).^[13b,e,32]Thus, the 9!10 and 11!12 cycli-

zations proceed more rapidly than the analogous 3!4 (N_t=

46.97 h^ꢀ1, and 25.14 vs. 4.78 h^ꢀ1, respectively) with N_tin-

creased by approximately five- to ten-fold. The six-mem-

bered ring cyclizations of primary alkenyl alcohols 7 and 13

[a] Turnover frequencies determined by ¹H NMR spectroscopic integra-

tion versus an internal standard. [b] For more details, see the Supporting

Information.

also exhibit

a

Thorpe–Ingold effect (N_t=1.89 h^ꢀ1vs.

9.52 h^ꢀ1, respectively), but with diminished cyclization rates

versus the five-membered rings. The same trend is observed

in organolanthanide-mediated aminoalkene hydroamina-

tion^[13b,e]and hydroxyalkyne HO.^[7]

The catalytic sensitivity to olefin substituent steric encum-

brance for conversions 15!16, 21!22, and 39!40 (Table 2,

entries 17 and 18; Table 3, entries 4–6, 22–24; Table 4, en-

tries 10–12) is generally consistent with the superiority of

the larger ionic radius Ln³⁺catalysts, reflected in a three-

fold increase in N_t. This argues for a slightly sterically de-

manding transition state, thus reflecting subtle changes in

the catalyst coordination environment.^[7]Interestingly, the

15!16, 21!22, and 39!40 cyclization regiochemistries are

different from that expected from conventional Ln inser-

tion/protonolysis catalytic processes^[7,13,14](see more in the

Discussion section). Furthermore, transformations 23!24

and 27!28 (Table 3, entries 7–9, 13–15) illustrate that cycli-

zations are not restricted to alkenols that bear substituents

at terminal olefin positions. The present catalytic process is

also applicable to alkenols that have substituents at internal

olefinic positions (Table 3, entries 1–6, 16–18, 19–21). Note

the 27!28 cyclization, which proceeds more rapidly than

23!24 (N_t=14.89 vs. 7.88 h^ꢀ1respectively, at 1208C). If the

transition state of this HO/cyclization process were partially

or completely carbocationic, it would enjoy greater reso-

Figure 1. Typical ¹H NMR spectroscopy step plot of the [Yb

[OTf]. The reaction progress was moni-

ACHTUNGRTENUN(NG OTf)₃]-medi-

ated 1!2 conversion in [C₂mim]

ACHTUNGTRENNUNG

tored by extracting aliquots of a known mass at preset times and collect-

ing the ¹H NMR spectrum (CD₃NO₂, 500 MHz). Reaction conditions:

[Yb

(OTf)₃]=1.0710^ꢀ2m,

[1]=1.07m,

[1,1,2,2-tetrachloroethane]=

0.11212m. A) t=0 min, B) t=5 min, C) t=10 min. D) ¹H NMR spectrum

(CD₃NO₂, 500 MHz) of clean final product 2 isolated by column chroma-

tography (silica gel). [*] [D₃]Nitromethane. [**] 1,1,2,2-Tetrachloro-

ethane. [***] Note: The aliquots varied slightly in the amounts of para-

magnetic Yb³⁺present, hence the slight discrepancies in ¹H chemical

shifts.

Chem. Eur. J. 2010, 16, 3403 – 3422

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

3409

A. Dzudza and T. J. Marks

Table 5. The effect of varying Ln³⁺ionic radius on catalytic intramolecu-

lar HO/cyclization turnover frequency.

nance stabilization by phenyl over methyl substituents, as

observed.

Note also that intramolecular hydroalkoxylation/cycliza-

tion of alkenol 25, which bears a cyclopropyl substituent at

the internal olefinic position, proceeds through cyclopropyl

ring-opening to generate the furan 26 (Table 3, entries 10–

12), thus plausibly supporting transient carbocationic species

along the reaction coordinate (see more in the Discussion

section).^[33]

Secondary alcohols results raise more intriguing questions

about the structural effects that govern cyclization rates.

HO/cyclization of aryl-functionalized secondary hydroxyal-

kenes proceeds at enhanced rates (Table 4, entries 1–6)

versus linear hydroxyalkenes (Table 2, entries 1–3, 5–7),

thereby suggesting some combination of Thorpe–Ingold/pre-

organization and electronic effects. For five-membered

rings, the 1!2 transformation is slightly more rapid than

3!4 (N_t=6.37 vs. 4.78 h^ꢀ1, respectively, at 1208C). For six-

membered ring formation, the 7!8 versus 35!36 transfor-

mations experience similar acceleration magnitudes (N_t=

2.75 vs. 1.89 h^ꢀ1, respectively, at 1208C).

Entry

Ln³⁺

Ionic radius

N_t[h^ꢀ1] (T [8C])^[a,b]

1

2

3

4

5

La

1.160

1.109

1.079

1.008

0.977

0.57 (120)

1.21 (120)

1.37 (120)

46.97 (120)

47.15 (120)

Nd

Sm

Yb

Lu

[a] Turnover frequencies measured in [C₂mim]ACTHNUTRGNE[NUG OTf] with 1 mol% cata-

lyst. [b] For details, see the Supporting Information.

of over approximately 80-fold, with HO/cyclization more

rapid for smaller ionic radius catalysts. The effects of Ln³⁺

ionic radius on the aliphatic/aromatic alkenol cyclization N_t

values generally parallel Ln contraction trends^[19](Table 5),

with the smallest, most Lewis acidic^[20a–c]Ln³⁺center in

most cases being more efficient. This pattern parallels that

of organolathanide-catalyzed aminoalkyne hydroamination/

cyclization,^[13a,f]but not that of aminoalkenes^[13b,e]or of hy-

droxyalkynes and hydroxyallenes.^[7]Further explanations are

deferred to the Discussion section.

Another instructive observation here is that [LnACHTNUTRGNEUNG(OTf)₃]

complexes also serve as efficient catalysts for the intramo-

lecular HO/bicyclization of aliphatic dihydroxydienes to

yield the corresponding bicyclic oxygen-containing frame-

works as shown in Equations (14) and (15) (Table 2, en-

tries 19–21; Table 3, entries 19–24).

Kinetic and mechanistic studies: Most kinetic data acquired

in this study were obtained using the [YbACTHNURTGNENG(U OTf)₃] catalyst.

The rationale is two-fold: 1) the small ionic radius leads to

increased cyclization rates, thus increasing efficiency and ac-

curacy in measurements on less reactive substrates, 2) the

relatively short Yb³⁺T_1e(electron spin-lattice relaxation

time) allows convenient and informative NMR spectroscopic

monitoring of reactions without excessive line broadening

induced by some other paramagnetic Ln ions. Quantitative

kinetic studies of representative cyclization 1!2 were car-

ried out in the presence of 1–10 mol% [Yb

ACHTUNGRTEN(NUNG OTf)₃] catalyst

in [C₂mim][OTf]. Kinetics were monitored from intensity

AHCTUNGTRENNUNG

changes in the substrate olefinic resonances over three or

more half-lives. The concentration of product 2 was mea-

sured from the area of the olefinic resonances (Figure 1)

standardized to the area of the CH₂peak of added 1,1,2,2,-

tetrachloroethane. The methylene signal of 1,1,2,2-tetra-

chloroethane (d=6.26 ppm) can be readily distinguished

from substrate 1 and product 2 resonances at 500 MHz.

These cyclizations were performed with a 100-fold molar

excess of substrate over the catalyst, and NMR spectroscop-

ic analysis indicates that, within the instrumental limits, only

substrate and product are present in detectable quantities at

all stages of conversion. Monitoring substrate consumption

as a function of time (Figure 2A) yields initial rate constants

(Figure 2B).

ꢀ

These results demonstrate that C O fusions can be regio-

specifically coupled in sequence to exclusively assemble bi-

cyclic rather than monocyclic products (Table 2, entries 17

and 18), especially rapid when smaller ionic radius [Yb-

ACHTUNGTRENNUNG(OTf)₃] is used. This catalytic transformation is therefore

competent to form five/five and five/six polycyclic skeletons

with a single catalyst center, and at turnover frequencies

comparable to those of the corresponding primary and sec-

ondary hydroxyalkene cyclizations (3, 21, and 23, respective-

ly) for the same catalyst, temperatures, and reaction condi-

tions (Table 2, entry 1; Table 3, entries 4 and 7).

Metal ion effects on catalytic alkenol hydroalkoxylation/cyc-

lization: For the representative cyclization 11!12

[Eq. (16)], Table 5 shows that N_tincreases substantially with

a decrease in the Ln ionic radius. Proceeding from the larg-

est eight-coordinate ionic radius, La³⁺(1.160 ꢀ),^[34]to inter-

mediate Sm³⁺(1.079 ꢀ), to the smallest six-coordinate ionic

radius, Lu³⁺(0.977 ꢀ),^[34]the N_tvariation spans a rate factor

A van ’t Hoff plot of lnk_obsversus ln[1] reveals a linear re-

lationship consistent with a first-order dependence in the

concentration of 1 (Figure 2C). The same trend is observed

for the k_obsdependence on the [Yb

(Figure 2D). Kinetic studies under the same conditions but

at different substrate 1 and [Yb(OTf)₃] catalyst concentra-

ACHTUNGRTEN(NUGN OTf)₃] concentration

AHCTUNGTRENNUNG

3410

www.chemeurj.org

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2010, 16, 3403 – 3422

Hydroalkoxylation of Unactivated Alkenols

FULL PAPER

Turnover frequencies, N_t(h^ꢀ1), were calculated from the

least-squares determined slope (m) according to Equa-

tion (21) (see the Experimental Section for details). Deter-

mination of the cyclization N_tvalues for other substrates

was carried out in the same way as for the cyclization 1!2.

Standard Eyring (Figure 3B) and Arrhenius (Figure 3C)

Figure 2. A) Concentration of starting material (1) as a function of time

Figure 3. A) Natural log of product 3 concentration as a function of time

and temperature for the hydroalkoxylation/cyclization of 3 using 1 mol%

for the hydroalkoxylation/cyclization using [Yb

[C₂mim][OTf] at 1208C. B) Determination of k_obsfrom the plot of ln[1]

versus time. C) van ’t Hoff plot of the hydroalkoxylation/cyclization of 1

using [Yb(OTf)₃] as the catalyst in [C₂mim][OTf] at 1208C. The lines are

least-squares fits to data points. D) van’t Hoff plot of the hydroalkoxyla-

tion/cyclization of 1 using [Yb(OTf)₃] as the catalyst in [C₂mim][OTf] at

1208C. The lines are least-squares fits to data points.

ACHTUNGRTEN(NUNG OTf)₃] as the catalyst in

ACHTUNGTRENNUNG

of [Yb

B) Eyring plot for the intramolecular hydroalkoxylation/cyclization of 3

using [Yb(OTf)₃] catalyst in [C₂mim][OTf]. The lines represent the least-

squares fit to the data points. C) Arrhenius plot for the intramolecular

hydroalkoxylation/cyclization of 3 using [Yb(OTf)₃] catalyst in [C₂mim]-

[OTf]. The lines represent the least-squares fit to the data points.

ACHUTGTNERN(NUG OTf)₃] catalyst in [C₂mim]HCATUNGTRENN[UGN OTf] at various temperatures.

A

ACHTUNGTRENNUNG

A

ACHTUNGTRENNUNG

A

ACHTUNGTRENNUNG

AHCTUNGTRENNUNG

tions yield linear plots of rate versus substrate concentra-

tion. By combining these results, the overall experimental

rate law can be expressed as in Equation (17). The 1!2

analyses^[35]of data obtained in variable-temperature experi-

ments yielded enthalpy (DH^°)=18.2 (9) kcalmol^ꢀ1, entropy

(DS^°)=ꢀ17.0 (1.4) eu, and energy (E_a)=18.9 (8) kcalmol^ꢀ1

activation parameters.^[36]The errors in these parameters

were computed from published error propagation formu-

las,^[35b]which were derived from the Eyring equation^[35]

using the OriginPro 7.5 program.

[YbACHTUNGTRENNUNG(OTf)₃]-mediated intramolecular HO/cyclization rate

law is overall second-order, but first-order in substrate and

catalyst.

ꢀ

Kinetic isotope effect data ( OH versus OD) were ac-

quired for the cyclization 5!6 (Table 2, entries 1 and 4) by

1

n ꢁ k½Yb^3þꢂ ½1ꢂ

ð17Þ

Chem. Eur. J. 2010, 16, 3403 – 3422

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

3411

A. Dzudza and T. J. Marks

¹H NMR spectroscopy, and yield k_H/k_D=2.48 (9), which is

2

suggestive of a primary kinetic isotope effect. The H NMR

spectrum of cyclized product 6 shows it to be monodeuterat-

ed at the 2-methyl position (see the Supporting Informa-

tion). Proton-transfer processes typically exhibit KIEs of

2.5–7.0.^[37]Although NH/ND labeling studies for analogous

organolanthanide-catalyzed alkene hydroamination process-

2

es also exhibit a primary KIE with same H regioselectivi-

ty,^[13h]a number of pathways that yield the same regioselec-

tivity can be envisioned for H⁺transfer in the present case,

ꢀ

for example, olefin insertion into the Ln O bond followed

by inter/intramolecular proton transfer, or Ln³⁺-coordinated

alkoxide nucleophilic attack on olefin, with pre-/post- inter-/

intramolecular proton transfer (see the Discussion Section).

Figure 4. The effect of varying concentration of sterically hindered base

Mechanistic investigations with proton-trapping reagents:

The pathways for hydroalkoxylation/cyclization catalyzed by

TfOH and main-group metal triflates are proposed to be

closely related,^[4a]since the reaction selectivities are simi-

on [Yb

(OTf)₃]- and TfOH-mediated 1!2 cyclization in [C₂mim]

ACHTUNGTRENNUNG[OTf].

For experimental details, see the Supporting Information. Lines through

data points are drawn to guide the eye.

lar.^[4a,5a]In principle, [Ln

(OTf)₃] complexes could serve as a

lization rates differently, the explanation of which is defer-

red to the Discussion section.

source of a protic acid such as TfOH, a known catalyst for

intramolecular hydroalkoxylation/cyclization of unactivated

alkenes.^[4b]A classic test for proton-transfer participation in-

volves monitoring the reaction rate in the presence of steri-

cally hindered amine “proton sponges,” for example, 2,6-di-

tert-butylpyridine (DTP) or N,N,N’,N’-tetramethyl-1,8-naph-

thalenediamine (TMAN), which differ in their basicities

(pK_a=3.6 and 12.1, respectively).^[38,39a,d]In the case of

Brønsted acid catalysis, the H⁺is captured by the base,

thereby resulting in a depression or suppression of the cycli-

zation rate [Eq. (18)].^[38,39]

¹⁹F NMR spectroscopic analysis of the [Yb

ACTHUNTRGNE(UGN OTf)₃]- and

TfOH-mediated 1!2 HO/cyclization reaction mixture vola-

tiles, trapped by vacuum transfer after 15 min reaction, con-

firm the presence of triflic acid (d=ꢀ77.14 ppm) only in the

case of the TfOH-mediated reaction (see the Experimental

Section). These results argue that lanthanide triflates do not

serve as free TfOH precursors, even though the proton-trap-

ping results suggest the involvement of protons in the cata-

lytic cycle. These results imply that other possible alkenol

deprotonation pathways may be responsible for the catalyst

deactivation and observed de-

crease in cyclization turnover

frequencies.

To better understand the

origin of the acidic proton

trapped in the proton sponge

experiments, additional proton

traps that have different prop-

For the present comparative studies of [Ln

(OTf)₃]- and

erties were explored. Of particular interest were non-nucle-

ophilic traps that react rapidly and irreversibly with protons.

TfOH-catalyzed (1 and 3 mol%, respectively) 1!2 HO/cyc-

ꢀ

lization, the two proton sponges employed exhibit different

Here, main group aryls of the type R_nE Ar, for example,

inhibition patterns. Kinetic results for [Yb

N

R₃-Sn-Ph and R₃-Si-Ph, are attractive candidates since they

ꢀ

TfOH-mediated cyclizations reveal that the former turns

over approximately seven-fold faster (Figure 4), and that the

two proton sponges affect the respective reaction rates dif-

exhibit substantial E aryl cleavage rates in the presence of

protons to yield readily identifiable arenes and the corre-

[38,40]

ꢀ

sponding E OTf species.

Therefore, [YbACTHUNGRETNNU(G OTf)₃]-mediat-

ferently. For [Yb

G

ed 3!4 cyclization was carried out in the presence of

TMAN concentrations marginally affect the N_tvalues; only

upon the addition of three equivalents of TMAN is cycliza-

tion completely inhibited (Figure 4). In contrast, 2,6-di-tert-

Me₃SiPh (1.0 equiv per [YbACHTGUNTRENU(NG OTf)₃]) in [C₂mim]CATHUNGTREN[NUNG OTf]. The

reaction produces only small quantities of HO/cyclization

product (ꢁ2% yield) along with benzene (5 mol%) and

1

butylpyridine addition (1.0 equiv) to the [Yb

N

Me₃SiOTf (5 mol%), as indicated by H and ¹⁹F NMR spec-

ic mixture results in complete inhibition of turnover

(Figure 4). For TfOH-mediated 1!2 HO/cyclization, incre-

mental TMAN addition (0.5–3.0 equiv) decreases the rate in

an approximately linear fashion (Figure 4). Thus, the trap-

ping experiments that utilize TMAN and DTP indicate that

troscopic analysis of the reaction mixture. Furthermore, a

stoichiometric mixture of Me₃SiPh and 3 remains unchanged

after 12 h under identical reaction conditions. These results

argue that H⁺scavenging by Me₃SiPh inhibits HO/cycliza-

tion and deactivates the catalyst stoichiometrically. These re-

sults implicate the requirement of the Ln³⁺catalyst in facili-

proton sponges affect [YbACTHNUTRGNE(UNG OTf)₃]- and TfOH-mediated cyc-

3412

www.chemeurj.org

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2010, 16, 3403 – 3422

Hydroalkoxylation of Unactivated Alkenols

FULL PAPER

tating a proton transfer, which does not appear to originate

from free triflic acid, but rather from the alkenol hydroxyl

group. Literature reports also indicate that metal triflates do

not readily eliminate HOTf in the presence of alcoholic nu-

cleophiles.^[41]

RTILs, primarily because of the possibility of fine-tuning of

the physicochemical properties through the N-alkyl substitu-

ents and anions. Imidazolium-based ionic liquids display

pronounced self-organization in the solid and liquid phases,

an extended network of cations and anions interconnected

by hydrogen bonds.^[24,25]Introduction of solute disrupts the

extended network, thereby allowing the IL to act as a hy-

drogen-bond donor (cation) or acceptor (anion), thus inter-

acting with substrates that have accepting and donating

sites, respectively. The Hughes–Ingold rules, which describe

solvent effects on reaction rates, are essentially qualitative

and rely on generalized solvent polarity concepts.^[42]For mo-

lecular solvents, polarities are commonly expressed in terms

of the dielectric constant, thus implying an electrostatic sol-

vation model. Much effort has been made to develop empir-

ical solvent-polarity scales for RTILs to aid in explaining/

predicting differences in solvation effects.^[26]For most

RTILs, the single-parameter polarity approach has not been

definitive because most RTILs appear to fall within the

same narrow range of values.^[26]Yet, as shown here, two dif-

ferent ionic liquids that have essentially the same polarity

indices produce significantly different results as solvents for

HO/cyclization. Clearly, a single polarity parameter is insuf-

ficient to describe the variations in experimental results.

In comparison to unimolecular solvents, which have a lim-

ited number and type of interactions with solutes, ionic liq-

uids, given their diverse structures and functionality, are

complex and capable of many types of solvating interactions

(e.g., dispersive, p–p, n–p, hydrogen-bonding, dipolar, ionic,

charge–charge), with many of these interactions occurring

simultaneously, thereby enhancing solvation and reaction ef-

ficiencies.^[24–26,43]Due to the inherent polarity of RTILs, re-

actions that involve charge development in the transition

state will likely experience greater transition-state stabiliza-

tion than in polar unimolecular solvents.^[43]Identifying the

types of solute–solvent interactions between RTILs and sub-

strates is useful in explaining the origin of rate enhance-

ments, which can depend upon both RTIL constituents.^[43]

Solvation studies based on linear free-energy approaches to

analyze RTIL interactions with probe solute molecules

Coordinative probing of the Yb³⁺-olefin interaction: While

kinetic studies provide the overall reaction molecularity and

1

suggest the turnover-limiting step, in situ H NMR spectro-

scopic experiments give qualitative insight into which sub-

strate and product species enter the paramagnetic Ln³⁺co-

ordination sphere. In the present study, arrayed ¹H NMR

spectroscopic experiments were carried out over the course

of the [Yb

N

1

ꢀ

OH) H resonances. Yb was chosen, since it induces iso-

tropic paramagnetic shifts accompanied by reasonably small

line broadening. In [Yb

zation, substantial upfield shifts of the olefinic and methyl-

G

ꢀ

ene (adjacent to OH) proton resonances are observed on

initiation and during the course of turnover [Eq. (19) and

Figure 5].

m

N_t

¼

ð19Þ

½catalystꢂ₀

reveal that [C_nmim]ACTHNUTRGNE[NUG OTf] exhibits greater hydrogen-bond

ꢀ

basicity than does the NTf₂analogue, thus producing stron-

ger interactions with solutes such as alkenols, and resulting

in rate accelerations for polar transition states.^[43]

Scope of catalytic intramolecular hydroalkoxylation/cycliza-

tion: A central goal of this investigation was to explore the

1

Figure 5. Overlaid 500 MHz H NMR spectra of the [YbACTHNURGTNEUNG(OTf)₃]-mediated

15!16 cyclization reaction solution, in [D₃]nitromethane at room tem-

scope of [Ln

terminal/internal primary and secondary aliphatic/aromatic

alkenols. The results in Tables 2–4 indicate that [Ln(OTf)₃]

ACHTUNGRTENUN(NG OTf)₃]-mediated intramolecular cyclization of

perature. For more details, see the Experimental Section.

ACHTUNGTRENNUNG

complexes are competent catalysts for the formation of di-

verse five- and six-membered oxygen heterocycles that have

a wide variety of alkyl and aryl substituents. The ring-size

Discussion

Reaction-medium effects on the catalytic alkenol hydroal-

koxylation/cyclizaton process: Derivatives of 1,3-dialkylimi-

dazolium cations associated with a variety of anions are

among the most widely used and investigated classes of

dependence of [LnACTHNGUTERNNU(G OTf)₃]/RTIL-catalyzed cyclization rates

for the primary alkenols is 5>6, which is consistent with a

classical, sterically controlled ring-forming transition state

(Table 2, entry 2 vs. 5, entry 6 vs. 8).^[13–15]In general, the

Chem. Eur. J. 2010, 16, 3403 – 3422

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

3413

A. Dzudza and T. J. Marks

five-membered ring transformations proceed more rapidly

than the analogous six-membered rings, with N_tincreasing

by approximately three- to ten-fold. The same trend is ob-

served in organolanthanide-mediated hydroaminations, but

with far larger rate dispersions than observed here.^[13]

In the HO/cyclizations of both primary and secondary al-

kenols, there are distinct differences in rates between alkyl

and aromatic substrates. In general, cyclization rates of aryl-

functionalized alkenols are slightly more rapid than those of

the linear hydroxyalkenes. This may reflect transition-state

configurations in which the aryl-functionalized substrates

have more accessible, preorganized structures that involve

interaction of the electrophilic Ln³⁺

rangement, followed by intramolecular H⁺transfer and alk-

oxide nucleophilic attack (or conventional HO/cycliza-

tion^[7]), the cyclization would presumably result in a spiro-

furan (Scheme 4, route A). Alternative pathway

B

(Scheme 4), in which H⁺transfer results in cyclopropyl ring

opening, involves a transitory allylcarbynyl cation, which

when quenched by an alkoxide nucleophile would result in

26, the observed final product, and consistent with cyclo-

propyl rearrangement route B. Prior studies have shown

that main group triflate-mediated methylenecyclopropane

ring opening with alcoholic nucleophiles proceed along

route B (Scheme 4).^[41]

Regarding 1,3-diol–alkenyl substrates, the present [Yb-

center with an adjacent p system (e.g.,

ACHTUNGRTEN(NNUG OTf)₃]-mediated intramolecular HO/bicyclization coupling

process is capable of regiospecific closing of five- and five/

six-membered spirobicyclic skeletons [Eqs. (14) and (15)].

A).^[7a,44]The catalytic HO/cyclization

rates exhibit appreciable sensitivity to

ꢀ

olefin substituent steric encumbrance.

In general, terminal alkenol cycliza-

tion rates (Table 2) are greater than

These results (Tables 2 and 3) illustrate that two C O bond

fusions can be coupled in sequence at a single Ln³⁺center

to exclusively assemble bicyclic products. The substrates

that were examined form Markovnikov-regiospecific spiro-

those of sterically encumbered alkenols (Tables 3 and 4).

For alkenols that bear two substituents at the olefin termi-

nus, the superiority of the larger ionic radius Ln³⁺catalysts

is evident, as reflected in an approximately nine-fold N_tin-

crease. R substituent effects on cyclization rates can be ra-

tionalized to some degree in terms of a slightly sterically de-

manding transition state in which the R substituent occupies

a less encumbered position, thus reflecting subtle changes in

the catalyst coordination environment.^[13,45]Moreover, the

regiochemistry of the 15!16, 21!22, and 39!40 cycliza-

tions is different from what is expected in conventional lan-

thanide C=C insertion/protonolysis sequences,^[7,13–15]thereby

suggesting a pathway in which the Ln³⁺center facilitates in-

tramolecular proton transfer, followed by alkoxide nucleo-

philic attack (see more below).

Intramolecular HO/cyclization of alkenol 25, which bears

a cyclopropyl substituent at the internal olefinic position,

provides information on the nature of transient species

along the reaction coordinate. Cyclopropyl substituents are

highly strained and suffer both Bayer (angular)^[33a]and

Pitzer (torsional) strain,^[33a]thereby resulting in pronounced

reactivity.^[33]Possible rearrangement routes for HO/cycliza-

tion are illustrated in Scheme 4 and take into account the

established, greatly altered reactivity when in conjugation

with an adjacent p system.^[33]In route A (Scheme 4), the

process traverses a well-documented, thermally induced vi-

nylcyclopropane unimolecular rearrangement pathway to

yield a cyclopentene, for which a concerted pathway has

been proposed.^[46]If the first step is thermally induced rear-

bicyclic furans/pyrans. The increased N_tfor the [Yb

versus [La

versus that of parent molecule 15 suggests that the interplay

of factors including sterics, electronics, and metal-coordina-

tion strength govern the rates.

ACHTUNGTRENNUNG(OTf)₃]-

(OTf)₃]-mediated HO/cyclization of 33!34

Metal-ion effects: Catalytic activity dependence on metal-

ion size has previously been observed in other Ln³⁺-mediat-

ed transformations.^[7,13,14]In the present study, N_tvariation

spans a factor of over approximately 80-fold from the larg-

est eight-coordinate ionic radius, La³⁺(1.160 ꢀ),^[34]to inter-

mediate Sm³⁺(1.079 ꢀ),^[34]to the smallest six-coordinate

ionic radius, Lu³⁺(0.977 ꢀ)^[34](Table 5). Interestingly, the

present range of N_tvariations is markedly smaller than the

approximately 10³-fold range observed in organolanthanide-

mediated aminoalkene hydroamination/cyclization,^[13e]but

larger than the approximately 20-fold range observed in or-

ganolanthanide-mediated hydroxyalkyne and hydroxyallene

HO/cyclizations,^[7]presumably reflecting significant steric

and electronic differences along the different reaction coor-

dinates.^[13e,45]In the present study, the HO/cyclization rates

of alkenols that bear two substituents at the olefin terminus

follow an opposite trend to cyclization rates of terminal al-

kenols, with larger radius lanthanides correlating with great-

er reaction rates. In the latter case, the observed behavior is

most likely due to steric hindrance of olefin approach and

activation by the catalytic center.^[13e,45]

Kinetics and mechanism of in-

tramolecular alkenol hydroal-

koxylation/cyclization: The ki-

netic results for the 1!2 con-

version were acquired by using

[Yb

pirical

[catalyst]¹[substrate]¹.

ACHTUNGRTEN(NUNG OTf)₃], and yield the em-

rate

law

nꢁk-

N

This

Scheme 4. Plausible 25!26 rearrangement hydroalkoxylation/cyclization routes A and B.

first-order dependence on sub-

3414

www.chemeurj.org

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2010, 16, 3403 – 3422

Hydroalkoxylation of Unactivated Alkenols

FULL PAPER

Table 6. Kinetic isotope effects comparison observed for intramolecular hydroamination/cyclizations and intra-

molecular hydroalkoxylation/cyclizations processes.

strate is significantly different

from the zero-order depend-

ence observed in homoleptic

lanthanide amido- or lanthano-

Entry

1^[a]

Catalyst

Substrate

Product(s)

k_H/k_D

[Yb(OTf)₃]

G

OH/OD=2.48 (9)

cene-mediated

hydroamina-

tion/cyclization and HO/cycli-

zation processes.^[7,13]The ob-

served zero-order (substrate)

dependence argues for turn-

2^[b]

3^[c]

[La{N

(SiMe₃)₂}₃]

OH/OD=0.95 (3)

NH/ND=2.7 (4)

ACHUTNGREN[NUG Cp*₂LaNACHTUNTGRNE(NUGN SiMe₃)₂]

over-limiting

olefin insertion into an Ln N/

intramolecular

4^[c]

A

E

NH/ND=5.2 (8)

ꢀ

[7,13]

ꢀ

Ln O bond (Scheme 1),

in

[OTf]. [b] Determined using the precatalyst in [D₆]benzene.^[7]

ACHTUNGTRENNUNG

accord with thermodynamic

considerations^[17]and DFT-

level quantum chemical calcu-

lations.^[45]The rate law in the

[a] Determined using the catalyst in [C₂mim]

[c] Determined using the precatalyst in [D₈]toluene.^[13h]

present case argues that either alkenol complexation is turn-

over-limiting (step i, Scheme 5), or that proton transfer

(step ii, Scheme 5) following a rapid pre-equilibrium is turn-

over-limiting (hence the observed kinetic isotope effect; see

below). Following subsequent alkoxide nucleophilic attack/

ring closure (step iii, Scheme 5), product dissociation

(step iv, Scheme 5) regenerates the [LnACTHNGURETNNU(G OTf)₃] catalyst. Ki-

ꢀ

netic isotope effect (KIE) data ( OH vs. OD) for cycliza-

tions 3!4 and 5!6 (Table 2, entries 1 and 4) yield k_H/k_D=

2.48 (9) (Table 6, entry 1). This value is consistent with

proton-transfer processes, which usually exhibit primary

KIEs of 2.5–7.0,^[37]and that it is part of the turnover-limiting

step or is intimately coupled to it. The magnitude of this

KIE is reminiscent of values observed for organolanthanide-

catalyzed aminoalkene hydroamination process,^[13h]but

stands in contrast to a nonprimary effect observed for ho-

moleptic lanthanide amido-catalyzed hydroxyalkyne and hy-

droxyallene hydroalkoxylation processes,^[7]as summarized in

2

Table 6. Moreover, H NMR spectra of cyclized product 6

show it to be monodeuterated at the 2-methyl position.

Note that NH/ND labeling studies for the analogous orga-

nolanthanide-catalyzed aminoalkene hydroamination pro-

cesses also exhibit primary KIEs with the same deuterium

regioselectivity,^[13h]however, experimental and theoretical

ꢀ

evidence here for turnover-limiting olefin insertion into Ln

N is compelling.^[13]

Variable-temperature kinetic studies of the 3!4 hydroal-

koxylation/cyclization yield activation parameters DH^°=

18.2 (9) kcalmol^ꢀ1, DS^°=ꢀ17.0 (1.4) eu, and E_a=18.9

(8) kcalmol^ꢀ1(Table 7, entry 7). Organolanthanide-mediated

intramolecular hydroamination and HO processes (Table 7,

entries 1–6) proceed with moderate activation enthalpies

and energies, but with negative DS^°values, thereby suggest-

ing highly organized transition states. The somewhat higher

DH^°and E_avalues and a negative DS^°value^[7]observed for

intramolecular HO/cyclization (Table 7, entry 6) most likely

ꢀ

reflect the large Ln O bond enthalpy and again, a highly or-

ganized/polar transition state.^[7]Intermolecular aminoalkyne

hydroamination (Table 7, entry 8) proceeds with DH^°and

E_avalues slightly larger than most of the intramolecular hy-

droamination/cyclizations but with a comparably negative

DS^°(Table 7, entry 8).^[13c,g,j]The present DH^°and E_avalues

for Table 7, entry 7 may reflect a transition state in which

significant bond formation compensates for bond breaking,

with the large negative DS^°consistent with a well-organized,

polar transition state and substantial loss of degrees of free-

dom, not unexpected for an intermolecular process.^[13c]

Scheme 5. Proposed catalytic pathways for [LnACTHNUTRGENUG(N OTf)₃]-mediated intramo-

lecular hydroalkoxylation/cyclization of A) terminal alkenols and alke-

nols that bear one substituent at the olefin terminus, and B) alkenols that

bear two substituents at the olefin terminus.

Chem. Eur. J. 2010, 16, 3403 – 3422

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

3415

A. Dzudza and T. J. Marks

Table 7. Activation parameter comparison for intra- and intermolecular hydroamination/cyclization and hydroalkoxylation/cyclization processes (Cp’ =

Cp*, Cp’’ = Me₄Cp).

Entry

Substrate

Product

Catalyst

H^°

S^°

[eu]

E_a

[kcalmol^ꢀ1

]

[kcalmol^ꢀ1

]

1^[13b]

[Me₂SiCp’’₂SmCH

(SiMe₃)₂]

17.7 (2.1)

16.9 (1.3)

24.7 (5)

16.5 (4)

18.5 (2.0)

17.6 (1.4)

2^[13f]

ACHTUTGNRNENUG[Cp’₂LaCHACHTUNTGRNE(NUGN SiMe₃)₂]

3^[13i]

4^[13g]

5^[14a]

N

G

12.7 (1.4)

10.7 (8)

27.0 (5)

27.4 (6)

25.9 (5.2)

13.4 (1.5)

11.3 (2.0)

13.0 (1.4)

12.3 (1.6)

6^[7]

[La{N

[Yb(OTf)₃]

[Me₂SiCp’’₂NdCH

(SiCH₃)₂}₃]

20.2 (1.0)

18.2 (9)

11.8 (0.3)

17.0 (1.4)

25.9 (9)

20.9 (0.3)

18.9 (8)

7^[a]

ACHTUNGTRENNUNG

8^[13c]

(SiMe₃)₂]

17.2 (1.1)

17.8 (1.8)

[a] Determined using the catalyst in [C₂mim]ACTHNUTRGENUG[N OTf].

Investigations with proton traps: Brønsted acids are known

Comparative studies of [Ln

(1 and 3 mol%, respectively) 1!2 HO/cyclization in the

presence of TMAN are informative. For [Yb(OTf)₃]-mediat-

ed 1!2 cyclization, low TMAN concentrations marginally

affect N_tvalues; only upon the addition of 3.0 equiv of

TMAN is cyclization completely inhibited (Figure 4). For

TfOH-mediated 1!2 hydroalkoxylation/cyclization, incre-

mental TMAN addition (0.5–3.0 equiv) decreases the rate

monotonically (Figure 4). These trapping experiments indi-

ACHTUNGRTEN(NUNG OTf)₃]- and TfOH-catalyzed

ꢀ

catalysts for intramolecular addition of O H bonds across a

pendant olefin.^[4]Thus, detailed characterization of the role

of electrophilic Ln metal species in this catalytic transforma-

tion must include an understanding of potential proton par-

ticipation. In principle, this issue may be difficult to probe,

since standard mechanistic tests for Brønsted acid involve-

ment include addition of hindered amines (“proton spong-

es”), which might inhibit product formation for variety of

reasons. In the present study, the proton sponges 2,6-di-tert-

butylpyridine (DTP) and 1,8-bis(dimentylamino)naphtha-

lene (TMAN), which exhibit large differences in relative

base strength (pK_a=3.6 and 12.1, respectively)^[39a,d]were

ACHTUNGTRENNUNG

cate that proton sponges affect [YbACTHNUTRGENUG(N OTf)₃]- and TfOH-

mediated cyclization rates differently, plausibly reflecting

the differing steric and basicity characteristics of the two

sponges.^[39]To better understand the origin of the proton

sponge results, non-nucleophilic proton traps that react rap-

idly and irreversibly with H⁺were also employed.^[38,40]It is

found that Me₃SiPh inhibits the HO/cyclization process and

deactivates the catalyst stoichiometrically. Furthermore, a

stoichiometric mixture of Me₃SiPh and 3 remains unchanged

after 12 h under identical conditions. The totality of the H⁺

scavenging results suggest that the Ln³⁺center facilitates in-

tramolecular H⁺transfer by withdrawing the electron densi-

ty from the alkenol hydroxyl functionality, thus rendering

ꢀ

the O H proton more acidic and susceptible to transfer

[e.g., Eq. (20)]. Analogous results have been observed for a

employed. The results reveal that the two sponges affect the

HO/cyclization rates differently. [Yb

zation of 1 in the presence of TMAN (0.5 equiv with respect

to [Ln(OTf)₃]) exhibits a sig-

ACHTUNGRTEN(NUNG OTf)₃]-mediated cycli-

Pt-mediated hydroamination process.^[39]

ACHTUNGTRENNUNG

nificantly slowed reaction rate,

whereas negligible HO/cycliza-

tion is observed in the pres-

ence of 1.0, 2.0, and 3.0 equiv

of DTP (Figure 4).

3416

www.chemeurj.org

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2010, 16, 3403 – 3422

Hydroalkoxylation of Unactivated Alkenols

FULL PAPER

In addition to these studies, ¹⁹F NMR spectroscopic analy-

sis of the volatiles vacuum transferred (with heating) from

preparative-scale [LnACHTUNTRGNEUNG(OTf)₃]- and TfOH-mediated 2-allyl

phenol HO/cyclizations reveals the presence of the TfOH

only in the case of the latter reactions (see the Experimental

upon nucleophilic attack. This pathway may involve the for-

mation of sterically hindered b-carbocationic species, with

positive charge buildup favoring the more substituted

carbon atom, which is generally the site of nucleophilic

attack.^[47]

Section). These results argue that in [LnACTHNURTGNENG(U OTf)₃]-mediated

cyclizations, free TfOH is not formed in any significant

quantities during the catalytic cycle. These proton-scaveng-

ing results indicate the H⁺transferred originates in the alke-

nol hydroxyl functionality, and that lanthanide triflates do

not serve as precursors of free TfOH. The results are also

consistent with reports that metal triflates do not form sig-

nificant quantities of TfOH in the presence of alcohols.^[41]

Although the aforementioned H⁺trapping experiments

suggest proton involvement in the catalytic cycle, other pos-

sible alkenol deprotonation pathways may be responsible

for the decreased turnover frequencies. If the role of the

electron-deficient Ln³⁺center is to activate the hydroxyl

functionality through oxygen coordinaton, H⁺abstraction

by basic amines would yield Ln-alkoxide complexes, thereby

inhibiting catalytic turnover. The plausible alternative for

Coordinative probing of the Yb³⁺-alkenol interaction: Lan-

thanides are hard Lewis acids and their interaction with li-

gands is almost exclusively electrostatic due to shielding of

the f orbitals. In contrast to p-backbonding-capable transi-

tion metals, lanthanides do not engage in extensive, perhaps

requisite, metal–olefin backbonding. However, transitory

lanthanide–olefin complexes are frequently invoked in path-

ways for insertion processes,^[7,13–15]and Ln²⁺-olefin complex-

ation has been detected spectroscopically using an unsatu-

rated, paramagnetic 4f⁷NMR spectroscopic probe.^[18a]Thus,

in the present study, the viability of Ln³⁺-olefin coordination

was also examined.

For Yb³⁺, strong spin–orbit coupling results in rapid elec-

tron spin–lattice relaxation, thus resulting in reasonably

sharp ligand NMR spectroscopic linewidths but expectation

lanthanide-mediated alkenol HO/cyclization is through

of substantial dipolar shifts.^[48]Analysis of the [Yb

ACHTUNGTRENNUNG(OTf)₃]-

[7]

ꢀ

olefin insertion into Ln O bond, as in Scheme 1. As previ-

mediated 15!16 hydroalkoxylation/cyclization reaction

ously noted, homoleptic lanthanide amido compelexes serve

as lanthanide-alkoxide precursors.^[7]However, bonding-ener-

getic estimations^[18](Scheme 1) indicate that the very large

mixture reveals a significant upfield displacement of the ole-

1

finic (Dꢁ0.12 ppm ; Figure 5a–c, j) and a-carbon H reso-

nances (Dꢁ0.62 ppm; Figure 5f) versus the 1,1,2,2-tetra-

[18a,c]

ꢀ

Ln O bond enthalpy

renders the olefin insertive step i

chloroethane internal standard. The upfield displacement of

1

(Scheme 1) significantly endothermic for olefinic alcohols,

thus explaining why intramolecular alkenol HO/cyclization

is not observed for these catalysts.^[7]These observations in

conjunction with the H⁺scavenging results provide a rea-

sonable explanation why HO/cyclization is inefficient for

lanthanide alkoxide catalysts.

the H resonances that result from Ln³⁺-alkenol functionali-

ty interactions likely arises from both dipolar (through

space) and contact (through bond) mechanisms.^[17a,48]Ln³⁺

–

ꢀ

alkoxide interactions plausibly labilize the OH proton for

intramolecular transfer to the Ln³⁺-polarized olefin (see

Scheme 5A, step ii; Scheme 5B, step ii).

The activation process for intramolecular HO/cyclization

processes depends on the nature of the catalyst center,

which can activate either the hydroxyl or C=C functionality

to varying degrees, thereby resulting in either Markovnikov

or anti-Markovnikov regioselectivity.^[4,5,7–15]The most com-

monly invoked pathway for late-transition-metal-catalyzed

nucleophilic addition to olefins that yields Markovnikov re-

gioselection products (Scheme 6) begins with formation of

metal-bound olefin, which slips from h²to h¹coordination

Mechanistic recapitulation: In formulating a HO/cyclization

scheme that accommodates the mechanistic observations, it

is useful to briefly summarize the results. 1) Ln ion-size ef-

fects on rates generally follow the Ln contraction with the

most Lewis acidic Ln³⁺centers being the most efficient.

2) Substrate rate sensitivity generally parallels other hydro-

elementation/cyclization processes. 3) Cyclizations proceed

with Markovnikov-type selectivity. 4) Kinetics are first-order

in catalyst and substrate over a wide concentration and con-

version range. 5) DH^°, DS^°, and E_avalues suggest a highly

ꢀ

organized transition state. 6) OH/ OD labeling experi-

ments establish regiospecific delivery of a single D to the

less-substituted olefinic position with a primary KIE of 2.48

(9). 7) Proton-scavenging experiments that employ nucleo-

philic and non-nucleophilic proton traps suggest participa-

tion of an acidic proton in the catalytic cycle that originates

from the hydroxyl functionality, and they show that Ln tri-

flates do not serve as precursors of free triflic acid in signifi-

cant amounts. 8) Coordinative probing experiments are indi-

cative of hydroxyl and olefin coordination to Yb³⁺

.

Observations 1 to 8 argue for irreversible Ln³⁺-facilitated

intramolecular H⁺transfer to a less-substituted olefinic po-

sition to generate a partially or fully carbocationic transition

Scheme 6. Catalytic addition of NuH to an alkene by outer-sphere nucle-

ꢀ

ophilic attack and subsequent M C bond protonolysis.

Chem. Eur. J. 2010, 16, 3403 – 3422

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

3417

A. Dzudza and T. J. Marks

Physical and analytical measurements: NMR spectra were recorded

using Varian Inova-500 (FT; 500 MHz, ¹H; 100 MHz, ¹³C), Inova-400

(FT; 400 MHz, ¹H; 100 MHz, ¹³C), Mercury-400 (FT; 400 MHz, ¹H;

state. The charge developed is then quenched by alkoxide

nucleophilic attack and ring closure to generate a Ln³⁺

cyclic ether complex. The heterocyclic product then dissoci-

ates, thereby restoring the active catalytic species (Sche-

me 5A, B).

–

2

100 MHz ¹³C; 76.7 MHz, H; 376.9 MHz, ¹⁹F), or Bruker Advance III-500

NMR (FT; 500 MHz, ¹H; 100 MHz, ¹³C) spectrometers. Chemical shifts

(d) for ¹H, ²H, and ¹³C are referenced to internal solvent resonances

1

2

unless stated otherwise. Chemical shifts (d) for H, H, and ¹³C are refer-

enced to internal solvent resonances. Chemical shifts (d) for ¹⁹F are refer-

enced with respect to CFCl₃(d=0.00 ppm) in [D₃]nitromethane. NMR

spectroscopic experiments on air-sensitive samples were conducted in

Teflon valve-sealed J-Young tubes. Mass spectral data were obtained

using a Varian 1200 Quadrupole mass spectrometer. Elemental analyses

were performed by Midwest Miscrolabs, Indianapolis, IN.

Conclusion

Lanthanide triflates are efficient catalysts for the intramo-

lecular hydroakoxylation/cyclization of unactivated primary/

secondary and aliphatic/aromatic alkenols in room tempera-

ture ionic liquids to yield the corresponding furan, pyran,

spirobicyclic furan, spirobicyclic furan/pyran, benzofuran,

and isochroman derivatives. Cyclizations of diverse alkenols

proceed with near-quantitative conversions and reasonably

large turnover frequencies. The final products can be isolat-

ed either by simple diethyl ether extraction or by vacuum

transfer, thus allowing efficient catalyst and ionic liquid re-

cycling. Overall, the present results obtained from the kinet-

ic studies (i.e., rate law, activation parameters, isotope ef-

fects) and from structural factors that affect cyclization rates

(i.e., metal-ion size, product ring size, substrate–substituent

effects) support the general scenario shown in Scheme 5.

Final product regioselectivities along with kinetic and mech-

anistic studies implicate turnover-limiting Ln³⁺–alkenol

complexation, in concert with intramolecular H⁺transfer

and alkoxide nucleophilic attack/ring closure.

Compound 33: Dimethyl allylmalonate (1.40 g, 0.00781 mol) was added

dropwise to a suspension of NaH (0.0162 mol) in THF (30 mL) at 08C.

After the addition was complete, the mixture was allowed to warm to

room temperature and stirred for 1 h. The reaction mixture was then

cooled to 08C, and 4-chloro-2-methyl-2-butene (0.0324 mol) in THF

(10 mL) was added dropwise. The reaction mixture was then allowed to

warm to room temperature and stirred for additional 16 h. The reaction

was next quenched with 1m HCl at 08C, and the organic material was ex-

tracted with diethyl ether. The combined organic extracts were washed

with 1m HCl (2ꢂ10 mL), saturated NaHCO₃(20 mL), and brine

(10 mL). The solution was dried over MgSO₄, filtered, and concentrated.

The crude product, 2-allyl-2-(3-methyl-but-2-enyl)malonic acid dimethyl

ester, was purified by column chromatography eluting with hexanes/ethyl

acetate=8:2 to yield 1.54 g (0.00636 mol, 81%) of a colorless oil. Spec-

tral data match the literature.^{[4a] 1}H NMR (CD₃NO₂, 400 MHz): d=5.74–

5.64 (m, 1H), 5.14–5.09 (m, 2H), 5.00 (t, J=15.2, 7.6 Hz, 1H), 3.60 (s,

6H), 2.58 (t, J=14.8, 7.2 Hz, 2H), 1.70 (s, 3H), 1.63 ppm (s, 3H).

2-Allyl-2-(3-methyl-but-2-enyl)malonic acid methyl ester (1.54 g,

0.00636 mol) in THF (10 mL) was added dropwise to a suspension of

LiAlH₄(0.0178 mol) in THF (30 mL) at 08C. The suspension was stirred

at room temperature for 1 h then quenched with NaOH (5 mL,

15 mol%) and H₂O (5 mL). The organic material was extracted with di-

ethyl ether and the combined organic extracts were then washed with sa-

turated NaHCO₃(2ꢂ10 mL) and brine (2ꢂ10 mL). The ether solution

was dried over MgSO₄, filtered, and concentrated. The crude product

was purified by column chromatography eluting with hexanes/ethyl ace-

tate=8:2 to yield the title compound (1.07 g, 0.00581 mol, 92%) as a vis-

cous, colorless oil. ¹H NMR (400 MHz): d=4.01–3.93 (m, 1H), 3.74–3.44

(m, 4H), 2.11–2.01 (m, 2H), 1.54–1.44 (m, 2H), 1.80 (d, J=6 Hz, 3H),

1.58 ppm (d, J=6 Hz, 3H); ¹³C NMR (100 MHz): d=134.62, 133.58,

119.14, 116.5, 66.78, 42.26, 35.42, 29.17, 24.7, 20.45 ppm; elemental analy-

sis calcd (%) for C₉H₁₆O₂: C 71.70, H 10.94; found: C 71.67, H 10.91.

Experimental Section

Materials and methods: All manipulations of reagents were carried out

with rigorous exclusion of O₂and H₂O in flame- or oven-dried Schlenk-

type glassware on a dual-manifold Schlenk-line, or interfaced to a high-

vacuum line (10^ꢀ6torr), or in a N₂-filled Vacuum Atmospheres glove box

with a high-capacity recirculator (<1 ppm O₂). Argon (Matheson, pre-

purified) was purified by passage through an MnO oxygen-removal

column and a Davison 4A molecular sieve column. [D₃]Nitromethane

(Cambridge Isotope Laboratories) used for NMR spectroscopy-scale re-

actions, was predistilled under the reduced pressure, dried over CaCl₂,

and stored over activated Davison 4A molecular sieves in vacuum-tight

storage flasks under Ar. Pentane was dried using activated alumina col-

umns.^[49]THF was distilled from Na/benzophenone ketyl. D₂O (Cam-

bridge Isotope Laboratories; all 99+ atom% D) was used as received.

Substrates 1, 3, 7, 19, and 39 were purchased from Aldrich Chemical Co.,

distilled under the reduced pressure, and dried as appropriate prior to

use. Substrates 5,^[7a]9,^[50]11,^[5d]13,^[5d]15,^[5a]17,^[51]21,^[52]23,^[53]25,^[54]27,^[55]

29,^[5d,56]31,^[51]35,^[57]37,^[58]41,^[59]and 43^[60]were prepared as reported in

the literature. All liquid substrates were transferred twice from freshly

activated molecular sieves and were freeze–pump–thaw degassed. They

were then stored in vacuum-tight storage flasks. Solid substrates were pu-

rified by column chromatography, dried under high vacuum, and stored

General catalytic reaction information: Intramolecular HO/cyclizations

of each substrate were first studied as a function of Ln for different [Ln-

AHCTUNGTRENNUNG

were carried out in ILs.

Preparation of standard solutions for typical NMR spectroscopy-scale

catalytic reactions in [D₃]nitromethane: For each data set, a separate

standard solution was prepared under an inert atmosphere in a three-

neck round-bottomed flask interfaced to the high-vacuum line. The solu-

tion consisted of [D₃]nitromethane, an internal NMR spectroscopic stan-

dard, and the appropriate alkenol (20 equiv per [LnACHTNUTRGNEUNG(OTf)₃]).

Typical NMR spectroscopy-scale catalytic reaction in [D₃]nitromethane:

In the glove box, a J-Young NMR spectroscopy tube equipped with a

Teflon valve was charged with a [LnACHTNUGRTENUNG(OTf)₃] (Ln=La, Sm, Yb) catalyst

in the glove box before use. [LnACTHNUTRGNE(UNG OTf)₃] complexes (Ln=La, Sm, Nd, Yb,

(5 mol%; 1.6ꢂ10^ꢀ5mol). On the high-vacuum line, the tube was evacuat-

ed (10^ꢀ6torr), then frozen at ꢀ788C, followed by addition of the sub-

strate standard solution (700 mL, 20 equiv per Ln³⁺, [substrate]=0.457m)

through a gas-tight syringe to the NMR spectroscopy tube under an

argon flush. The tube was then evacuated and backfilled with Ar while

frozen at ꢀ788C, and then sealed with the Teflon valve. The frozen reac-

tion mixture was maintained in a dry ice/acetone bath until the time for

NMR spectroscopic analysis, and then brought to the desired tempera-

and Lu) were prepared according to published procedures.^[61]The

¹H NMR spectroscopic integration internal standard, 1,1,2,2-tetrachloro-

ethane, was purchased from Aldrich Chemical Co., distilled under re-

duced pressure, and stored under N₂in a vacuum-tight storage flask

before use. The ¹⁹F NMR spectroscopic integration internal standard, flu-

orotrichloromethane, was purchased from Aldrich Chemical Co. (dis-

tilled, sure-seal) and used as received. RTILs were prepared and dried

according to published procedures,^[62]and were stored in the glove box.

3418

www.chemeurj.org

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2010, 16, 3403 – 3422

Hydroalkoxylation of Unactivated Alkenols

FULL PAPER

ture, and the ensuing hydroalkoxylation/cyclization reaction monitored

by H NMR spectroscopy.

1.14 ppm (d, J=6.4 Hz, 3H); ¹³C NMR (CD₃NO₂, 100 MHz): d=75.5,

74.7, 70.7, 68.7, 67.5, 43.4, 43.3, 34.2, 30.4, 20.4 ppm; HRMS (ESI): m/z:

calcd for C₁₁H₂₁O₂[M⁺]: 185.2875; found 185.2868.

1

Typical small-scale catalytic reaction in RTILs: In the glove box, a 25 mL

Kinetic studies of hydroalkoxylation/cyclization in [C₂mim]ACTHNUTRGNE[NUG OTf]: In a

three-neck round-bottomed flask equipped with a J-Young adapter with

typical experiment, a small-scale reaction was set up as described above

(see above). The reaction progress was monitored by withdrawing the ali-

quots of known mass at preset time intervals. The aliquots were diluted

with 1,1,2,2-tetrachloroethane (500 mL) in [D₃]nitromethane standard so-

lution (see above). For each aliquot, the H NMR spectrum was recorded

and integrated versus the internal standard, 1,1,2,2-tetrachloroethane. A

long pulse delay was used during data acquisition to avoid saturation.

a Teflon valve and magnetic stir bar was charged with the [LnACTHNUTRGNEUNG(OTf)₃] cat-

alyst (Ln=La, Sm, Nd, Yb, and Lu; 1 mol%; 1.5ꢂ10^ꢀ5mol), and RTIL

(8.9ꢂ10^ꢀ3mol). On the Schlenk line, the flask was evacuated (10^ꢀ3torr),

degassed, and backfilled three times with N₂while heating gently for ap-

proximately 20 min at around 408C to induce catalyst dissolution in the

RTIL. Next, the mixture was heated to the desired reaction temperature

1

(60–1208C) and alkenol (1.5ꢂ10^ꢀ3mol, 100 equiv per [Ln

ACTHNUGTRNE(UNG OTf)₃]) was in-

jected through a gas-tight syringe under a nitrogen flush. Upon comple-

tion of the reaction, the reaction mixture was cooled to room tempera-

ture, and the final product was extracted with small portions of diethyl

ether. Final product percent yield was determined directly either by

a standard solution of

1,1,2,2-tetrachloroethane in [D₃]nitromethane or after chromatographic

product isolation.

Reaction order measurements for conversion 1!2: Kinetics were moni-

tored from ¹H NMR spectroscopic intensity changes in the substrate ole-

finic resonances over three or more half-lives. The concentration of prod-

uct 2 was measured from the area of the olefinic peak, standardized to

the area of the CH₂peak of 1,1,2,2,-tetrachloroethane in solution. NMR

spectroscopic analysis of all reaction mixtures indicates that, within the

detection limits, only substrate and product are present in detectable

quantities at all stages of conversion. Monitoring substrate consumption

as a function of time yields initial rate constants [Eq. (21)]. The data

were processed with the SigmaPlot 2000 program.^[63]Turnover frequen-

cies, N_t[h^ꢀ1], were calculated for the least-squares determined slope (m)

according to Equation (19) in which [catalyst]₀is the initial concentration

of the catalyst. Determination of cyclization N_tvalues for other sub-

strates was carried out in the same way as for cyclization 1!2.

¹H NMR spectroscopic integration by using

Preparative-scale catalytic reaction for [Yb

version in [C₂mim]

[OTf]: In the glove box, 1 mol% (1.5ꢂ10^ꢀ4mol) [Yb-

(OTf)₃] and [C₂mim]

[OTf] (8.9ꢂ10^ꢀ2mol) were loaded into a 50 mL

A

ACHTUNGTRENNUNG

A

ACHTUNGTRENNUNG

three-neck round-bottomed flask equipped with a J-Young adapter with

a Teflon valve and magnetic stir bar. On the Schlenk line, the flask was

evacuated (10^ꢀ3torr), degassed, and backfilled three times with N₂while

heating gently for approximately 20 min at around 408C to induce cata-

lyst dissolution in the RTIL, until the solution became transparent. Next,

the mixture was heated to the desired reaction temperature (1208C) and

ln½substrateꢂ ¼ ꢀmt

ð21Þ

1 (1.5ꢂ10^ꢀ2mol, 100 equiv per [Yb

ACTHNUTRGNEN(UG OTf)₃]) was injected through a gas-

tight syringe under a nitrogen flush. Upon completion of the reaction,

the reaction mixture volatiles were removed by means of vacuum distilla-

tion. Volatiles removed were analyzed by ¹H and ¹⁹F NMR spectroscopy.

Final product 2 percent yield was determined by ¹H NMR spectroscopic

integration by using a standard solution of 1,1,2,2-tetrachloroethane in

[D₃]nitromethane (1.85 g, 1.38 mmol). ¹⁹F NMR spectroscopic analysis of

the reaction volatiles (CD₃NO₂, 376 MHz) removed from reaction mix-

ture by means of vacuum transfer (with heating at 1658C) revealed no

detectable triflic acid. In a different trial, a preparative-scale reaction was

run to completion. ¹⁹F NMR spectroscopic analysis of the vacuum-trans-

ferred (with heating) reaction volatiles also revealed no detectable triflic

acid.

Activation parameter determination: Eyring and Arrhenius kinetic analy-

ses^[35a,b]of data obtained in variable-temperature experiments for the

15!16 conversion (see above) were used to extract DH^°, DS^°, and E_a

parameters. The errors in these activation parameters were computed

from the published error propagation formulas,^[35b]which were derived

from the Eyring equation.

Isotopic labeling studies: In the glove box, a 25 mL three-neck round-

bottomed flask equipped with a J-Young adapter with a Teflon valve and

magnetic stir bar was charged with 1 mol% [Yb

ACHTUNGTRENNUNG

10^ꢀ5mol) and [C₂mim]

AHCTUNGTRENNUNG

flask was evacuated (10^ꢀ3torr), degassed, and backfilled three times with

N₂while heating gently for approximately 20 min at around 408C to

induce catalyst dissolution. Next, the mixture was heated to the desired

reaction temperature (1208C) and [D₁]4-penten-1-ol (1.5ꢂ10^ꢀ3mol,

Preparative-scale catalytic reaction for TfOH-mediated 1!2 conversion

[OTf] (8.9ꢂ10^ꢀ2mol) was

in [C₂mim]ACHTUNGTRENNUNG[OTf]: In the glove box, [C₂mim]ACHTUNGTRENNGUN

loaded into a 50 mL three-neck round-bottomed flask equipped with a J-

Young adapter with a Teflon valve and magnetic stir bar. On the Schlenk

line, the flask was evacuated (10^ꢀ3torr), degassed, and backfilled three

times with N₂. Next, 3 mol% (4.5ꢂ10^ꢀ4mol) TfOH was injected under

nitrogen, the mixture was heated to the desired reaction temperature

100 equiv per [YbACHTUNRGTNEUNG(OTf)₃]) was injected through a gas-tight syringe under

a nitrogen flush. Upon completion of the reaction, the final product was

removed by vacuum distillation. Final product percent yield was deter-

1

mined by H NMR spectroscopic integration by using a standard solution

of 1,1,2,2-tetrachloroethane in [D₃]nitromethane. ²H regioselectivity was

determined by integration with respect to [D₃]nitromethane by utilizing a

5.0% [D₃]nitromethane in nitromethane.

(1208C), and

1

(1.5ꢂ10^ꢀ2mol, 33.3 equiv per TfOH) was injected

through a gas-tight syringe under a nitrogen flush. Upon completion of

the reaction, the reaction mixture volatiles were removed by means of

vacuum distillation and were analyzed by ¹H and ¹⁹F NMR spectroscopy.

Final product 2 percent yield was determined by ¹H NMR spectroscopic

integration by using a standard solution of 1,1,2,2-tetrachloroethane in

[D₃]nitromethane (1.72 g, 1.26 mmol). ¹⁹F NMR spectroscopic analysis of

the reaction mixure volatiles removed by means of vacuum transfer (with

heating at 1658C) revealed the presence of triflic acid (CD₃NO₂,

376 MHz): d=ꢀ77.14 ppm (compared to a chemical shift of commercial-

ly available triflic acid).

Mechanistic investigations of [Yb

(OTf)₃]-mediated 1!2 conversion in

[C₂mim][OTf] with proton-trapping reagents

AHCTUNGTRENNUNG

With N,N,N’,N’-tetramethyl-1,8-naphthalenediamine: In the glove box,

four 25 mL three-neck round-bottomed flasks equipped with a J-Young

adapter with a Teflon valve and magnetic stir bar were charged with

1.0 mol% [Yb

(with respect to [YbACTHNUGTNRE(NUG OTf)₃]) of N,N,N’,N’-tetramethyl-1,8-naphthalenedi-

(OTf)₃] catalyst (1.5ꢂ10^ꢀ5mol), 0.5, 1.0, 2.0, 3.0 equiv

ACHTUNGTRENNUNG

amine, and [C₂mim]ACTHNUTRGNEUNG

[OTf] (8.9ꢂ10^ꢀ3mol). On the Schlenk line, the flask

was evacuated (10^ꢀ3torr), degassed, and backfilled three times with N₂

Compound 18: This compound was obtained by cyclizing 17 (1.5 mmol).

¹H NMR (CD₃NO₂, 400 MHz; rac): d=4.01–3.93 (m, 2H), 3.74–3.44 (m,

4H), 2.11–2.01, 2.11–2.01 (m, 2H), 1.54–1.44 (m, 2H), 1.80 (d, J=6.0 Hz,

3H), 1.58 ppm (d, J=6.0 Hz, 3H); ¹³C NMR (CD₃NO₂, 100 MHz): d=

77.4, 77.1, 76.5, 76.1, 51.8, 44.8, 44.6, 20.5, 20.4 ppm; HRMS (ESI): m/z:

calcd for C₉H₁₆O₂[M⁺]: 156.2256; found 156.2252.

while heating gently for approximately 20 min at around 408C to induce

catalyst dissolution in [C₂mim]

the desired reaction temperature (1208C) and 1 (1.5ꢂ10^ꢀ3mol, 100 equiv

per [Yb(OTf)₃]) was injected through a gas-tight syringe under a nitrogen

ACHTUNGTRENN[UNG OTf]. Next, the mixture was heated to

AHCTUNGTRENNUNG

flush. Reaction progress was monitored by withdrawing aliquots of

known mass at preset times. The aliquots were diluted with 1,1,2,2-tetra-

chloroethane (500 mL) in the [D₃]nitromethane standard solution (see

above). For each aliquot, the ¹H NMR spectrum was recorded and inte-

grated versus the internal standard, 1,1,2,2-tetrachloroethane. A long

pulse delay was used during data acquisition to avoid saturation.

Compound 34: This compound was obtained by cyclizing 33 (1.5 mmol).

¹H NMR (CD₃NO₂, 400 MHz): d=3.98–3.87 (m, 1H), 3.79 (d, J=8.8 Hz,

1H), 3.53–3.36 (m, 2H), 3.31 (d, J=8.8 Hz, 1H), 1.99 (dd, J=12.4, 6 Hz,

1H), 1.76 (dd, J=12.8, 6.8 Hz, 1H), 1.52–1.38 (m, 2H), 1.38 (s, 6H),

Chem. Eur. J. 2010, 16, 3403 – 3422

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

3419

A. Dzudza and T. J. Marks

With 2,6-di-tert-butylpyridine: In the glove box, four 25 mL three-neck

round-bottomed flasks equipped with a J-Young adapter with a Teflon

Acknowledgements

valve and magnetic stir bar were charged with 1.0 mol% [YbACHTUNGTRENNUNG

We thank the NSF (grant CHE-0809589) for support of this research.

alyst (1.5ꢂ10^ꢀ5mol) and [C₂mim]

ACHTUNGTRENNUNG

manifold Schlenk line, the flasks were evacuated (10^ꢀ3torr), degassed,

and backfilled three times with N₂while heating gently for approximately

20 min at around 408C to induce catalyst dissolution in the [C₂mim]-

[1] a) K. Tani, V. Kataoka in Catalytic Heterofunctionalization, Vol. 2

(Eds.: A. Togni, H. Grꢃtzmacher), Wiley-VCH, Weinheim, 2001,

pp. 171–216.

ACHTUNGTRENNUNG

[2] a) M. C. Elliott, E. Williams, J. Chem. Soc. Perkin Trans. 1 2001 ,

230 3 –2 340; b) Z.-Y. Yao, H.-P. Wu, Y.-L. Wu, J. Med. Chem. 2000 ,

43 , 248 4 –2 487; c) A. Mitchenson, A. Nadin, J. Chem. Soc. Perkin

Trans. 1 2000 , 286 2 –2 892; d) M. C. Elliott, J. Chem. Soc. Perkin

Trans. 1 2000 , 129 1 –1 318; e) T. L. B. Boivin, Tetrahedron 1987 , 43 ,

330 9 –3 362; f) P. A. Bartlett, Tetrahedron 1980 , 36 , 272.

ACHTUNGTRENNUNG

tion progress was monitored by withdrawing aliquots of known mass at a

preset time intervals. The aliquots were diluted with 1,1,2,2-tetrachloro-

ethane (500 mL) in the [D₃]nitromethane standard solution (see above).

For each aliquot, the ¹H NMR spectrum was recorded and integrated

[3] a) J. Tsuji, Palladium Reagents and Catalysts, Wiley, Chichester,

2004, pp. 27–103; b) T. Pei, X. Wang, R. A. Widenhoefer, J. Am.

Chem. Soc. 2003 , 125 , 64 8 –6 49; c) K. MuÇiz, Adv. Synth. Catal.

2004, 346, 1425–1428; d) Y. Hirai, J. Watanabe, T. Nozaki, H. Yo-

koyama, S. Yamaguchi, J. Org. Chem. 1997 , 62 , 77 6 –7 77; e) M.

Rçnn, J. E. Bꢄckvall, P. G. Andersson, Tetrahedron Lett. 1995, 36,

7749–7752.

[4] For examples of Brønsted acid mediated alkene hydroalkoxylation,

see: a) D. C. Rosenfeld, S. Shekhar, A. Tameiyama, M. Utsunomiya,

J. F. Hartwig, Org. Lett. 2006 , 8 , 417 9 –4 182; b) L. Coulombel, E.

DuÇach, Green Chem. 2004 , 6 , 49 9 –5 01; c) B. Wang, Y. Gu, L.

Yang, J. Suo, O. Kenichi, Catal. Lett. 2004 , 96 , 7 1 –7 4; d) P. J. Li-

nares-Palomino, S. Salido, J. Altarejos, A. Sꢅnchez, Tetrahedron

Lett. 2003 , 44 , 665 1 –6 655.

[5] For examples of transition-metal-mediated alkene hydroalkoxyla-

tion, see: a) C.-G. Yang, C. He, J. Am. Chem. Soc. 2005 , 127 , 696 6 –

6967; b) Y. Oe, T. Ohta, Y. Ito, Synlett 2005, 179–181; c) Y. Oe, T.

Ohta, Y. Ito, Chem. Commun. 2004 , 162 0 –1 621; d) H. Qian, X.

Han, R. A. Widenhoefer, J. Am. Chem. Soc. 2004 , 126 , 953 6 –9 537;

e) E. Marotta, E. Foresti, T. Marcelli, F. Peri, P. Righi, N. Scardovi,

G. Rosini, Org. Lett. 2002 , 4 , 445 1 –4 453; f) K. Hori, H. Kitagawa,

A. Miyoshi, T. Ohta, T. I. Furukawa, Chem. Lett. 1998 , 108 3 –1 084;

g) K. Hori, H. Kitagawa, A. Miyoshi, T. Ohta, T. I. Furukawa,

Chem. Lett. 1998 , 108 3 –1 084.

[6] For examples of metal-triflate-mediated alkene hydroalkoxylation,

see: a) K. Komeyama, T. Morimoto, Y. Nakayama, K. Takaki, Tetra-

hedron Lett. 2007 , 48 , 325 9 –3 261; b) L. Coulombel, M. Rajzmann,

J.-M. Pons, S. Olivero, E. DuÇach, Chem. Eur. J. 2006 , 12 , 635 6 –

6365; c) C.-G. Yang, N. W. Reich, C. He, Org. Lett. 2005 , 7 , 455 3 –

4556; d) C.-G. Yang, N. W. Reich, Z. Shi, C. He, Org. Lett. 2005 , 7 ,

455 3 –4 556.

versus the internal standard, 1,1,2,2-tetrachloroethane.

delay was used during data acquisition to avoid saturation.

A long pulse

Mechanistic investigation of TfOH-mediated 1!2 conversion in

[C₂mim][OTf] using proton-trapping reagents—With N,N,N’,N’-tetra-

ACHTUNGTRENNUNG

methyl-1,8-naphthalenediamine: In the glove box, a 25 mL three-neck

round-bottomed flask equipped with a J-Young adapter with a Teflon

valve and magnetic stir bar was loaded with a N,N,N’,N’-tetramethyl-1,8-

naphthalenediamine (0.5, 1.0, 2.0, 3.0 equiv with respect to TfOH), and

[C₂mim]ACHTUNGTRENNUNG

[OTf] (8.9ꢂ10^ꢀ3mol). On the dual-manifold Schlenk line, the

flask was evacuated (10^ꢀ3torr), degassed, and backfilled three times with

N₂, and TfOH (3 mol%) was added under N₂flush. Next, the mixture

was heated to 1208C, and 1 (1.5ꢂ10^ꢀ3mol, 33.3 equiv with respect to

TfOH), was injected through a gas-tight syringe under N₂flush. Reaction

progress was monitored by withdrawing aliquots of known mass at preset

time intervals. The aliquots were diluted with 1,1,2,2-tetrachloroethane

(500 mL) in the [D₃]nitromethane standard solution (see above). For each

aliquot, the ¹H NMR spectrum was recorded and integrated versus the

internal standard, 1,1,2,2-tetrachloroethane. A long pulse delay was used

during data acquisition to avoid saturation.

Mechanistic investigations of [Yb

[C₂mim][OTf] with proton-trapping reagents—With phenyltrimethylsi-

lane: In the glove box, 25 mL three-neck round-bottomed flask

equipped with a J-Young adapter having a Teflon valve and magnetic stir

bar was charged with 1.0 mol% [Yb

(OTf)₃] catalyst (1.5ꢂ10^ꢀ5mol), phe-

nyltrimethylsilane (1.0 equiv with respect to [Yb(OTf)₃]), and 8.9ꢂ

A

ACHTUNGTRENNUNG

a

AHCTUNGTRENNUNG

10^ꢀ3mol [C₂mim]

ACHTUNGTRENNUNG[OTf]. On the Schlenk line, the flask was evacuated

(10^ꢀ3torr), degassed, and backfilled three times with N₂while heating

gently for approximately 20 min at around 408C to induce catalyst disso-

lution in the [C₂mim]

reaction temperature (1208C) and 3 (1.5ꢂ10^ꢀ3mol, 100 equiv with re-

spect to [Yb(OTf)₃]) was injected through a gas-tight syringe under N₂

ACHTUNGTRENNUNG[OTf]. Next, the mixture was heated to the desired

ACHTUNGTRENNUNG

[7] For examples of lanthanide-mediated alkyne and allene hydroalkox-

ylation, see: a) S. Seo, X. Yu, T. J. Marks, J. Am. Chem. Soc. 2009 ,

131 , 26 3 –2 76; b) X. Yu, S. Seo, T. J. Marks, J. Am. Chem. Soc. 2007 ,

129 , 724 4 –7 245; c) S. Seo, T. J. Marks, Chem. Eur. J. 2010, in press

DOI: 10.1002/chem.200903027.

[8] a) F. Alonso, I. P. Beletsaya, M. Yus, Chem. Rev. 2004 , 104 , 307 9 –

3159; b) M. A. Tius in Modern Allene Chemistry, Vol. 2 (Eds.: N.

Krause, A. S. K. Hashmi), Wiley-VCH, Weinheim, 2004, pp. 834–

838; c) P. A. Bartlett in Asymmetric Synthesis, Vol. 3 (Ed.: J. D. Mor-

risson), Academic Press, New York, 1984, p. 455.

flush. Reaction progress was monitored by withdrawing aliquots of

known mass at a preset time intervals. The aliquots were diluted with

1,1,2,2-tetrachloroethane (500 mL) in [D₃]nitromethane standard solution

(see above). For each aliquot, the ¹H NMR spectrum was recorded and

integrated versus the internal standard, 1,1,2,2-tetrachloroethane. A long

pulse delay was used during data acquisition to avoid saturation. The

final reaction mixture was also analyzed by ¹⁹F NMR spectroscopy.

Coordinative probing of the Yb³⁺-olefin interaction: In the glove box, a

J-Young NMR spectroscopy tube equipped with a Teflon valve was

charged with the paramagnetic (4f¹³) [Yb

ACTHNUTRGENUN(G OTf)₃] catalyst (5 mol%, 4.5ꢂ

[9] Baldwin 3- to 7-membered ring-closure rules: a) J. E. Baldwin, J.

Chem. Soc. Chem. Commun. 1976 , 73 4 –7 35; b) J. E. Baldwin, R. C.

Thomas, L. I. Kruse, L. Silberman, J. Org. Chem. 1977 , 42 , 384 6 –

3852.

10^ꢀ5mol). On the high-vacuum line, the tube was evacuated to 10^ꢀ6torr,

then frozen at ꢀ788C, followed by addition through a gas-tight syringe of

a substrate 15 standard solution (700 mL, 0.457m, 20 equiv with respect to

[10] J. G. Taylor, H. Whithall, K. K. M. Hii, Chem. Commun. 2005 , 5103.

[11] T. Hosokawa, T. Shinohara, Y. Ooka, S.-I. Murahashi, Chem. Lett.

1989 , 200 1 –2 004.

[12] For recent organolanthanide reviews, see: a) S. B. Amin, T. J. Marks,

Angew. Chem. 2008 , 120 , 203 4 –2 054; Angew. Chem. Int. Ed. 2008 ,

47 , 200 6 –2 025; b) H. C. Aspinall, Chem. Rev. 2002 , 102 , 180 7 –1 850;

c) F. T. Edelmann, D. M. M. Freckmann, H. Schumann, Chem. Rev.

2002 , 102 , 185 1 –1 896; d) G. A. Molander, A. C. Romero, Chem.

Rev. 2002 , 102 , 216 1 –2 186; e) M. Shibasaki, N. Yoshikawa, Chem.

Rev. 2002 , 102 , 218 7 –2 210; f) F. T. Edelmann in Comprehensive Or-

[YbACHTUNGTRENNUNG(OTf)₃]) in [D₃]nitromethane under an argon flush. The tube was

then evacuated and backfilled with Ar while frozen at ꢀ788C, and then

sealed with the Teflon valve. The frozen reaction mixture was maintained

in dry ice/acetone bath until the time for NMR spectroscopic analysis.

¹H NMR spectra were recorded at room temperature.

3420

www.chemeurj.org

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2010, 16, 3403 – 3422

Hydroalkoxylation of Unactivated Alkenols

FULL PAPER

ganometallic Chemistry, Vol. 4 (Eds.: G. Wilkinson, F. G. A. Stone,

E. W. Abel), Pergamon Press, Oxford, 1995, Chapter 2; g) T. J.

Marks, R. D. Ernst in Comprehensive Organometallic Chemistry

(Eds.: G. Wilkinson, F. G. A. Stone, E. W. Abel), Pergamon Press,

Oxford, 1982, Chapter 21.

ci, P. G. Cozzi, A. Umani-Ronchi, Tetrahedron Lett. 1995 , 36 , 728 9 –

7292.

[23] C. Reichardt in Solvents and Solvent Effects in Organic Chemistry,

3rd ed., Wiley-VCH, Weinheim, 2003.

[24] a) Ionic Liquids IIIB: Fundamentals, Progress, Challenges, and Op-

portunities, ACS Symposium Series 902 (Eds.: R. D. Rogers, K. R.

Seddon), American Chemical Society, Washington DC, 2005;

b) Ionic Liquids IIIA: Fundamentals, Progress, Challenges, and Op-

portunities, ACS Symposium Series 901 (Eds.: R. D. Rogers, K. R.

Seddon), American Chemical Society, Washington DC, 2005;

c) Ionic Liquids: Industrial Applications to Green Chemistry, ACS

Symposium Series 818 (Eds.: R. D. Rogers, K. R. Seddon), Ameri-

can Chemical Society, Washington DC, 2002.

[25] a) V. I. Parvulescu, C. Hardacre, Chem. Rev. 2007, 107, 2615–2665;

b) K. Binnemans, Chem. Rev. 2007 , 107 , 259 2 –2 614; c) J. B. Harper,

M. N. Kobrak, Mini-Rev. Org. Chem. 2006 , 3 , 25 3 –2 69; d) V. A. Co-

calia, K. E. Gutowski, R. D. Rogers, Coord. Chem. Rev. 2006 , 250 ,

75 5 –7 64; e) J. Dupont, P. A. Z. Suarez, Phys. Chem. Chem. Phys.

2006 , 8 , 244 1 –2 452; f) J. Dupont, R. F. de Souze, P. A. Z. Suarez,

Chem. Rev. 2002 , 102 , 366 7 –3 692; g) R. Sheldon, Chem. Commun.

2001, 23, 2399–2407; h) P. Wasserscheid, W. Keim, Angew. Chem.

2000 , 112 , 392 6 –3 945; Angew. Chem. Int. Ed. 2000 , 39 , 377 2 –3 789;

i) T. Welton, Chem. Rev. 1999 , 99 , 207 1 –2 083.

[26] a) T. Singh, A. Kumar, J. Phys. Chem. B 2008 , 112 , 1296 8 –1 2972;

b) H. Weingꢄrtner, Phys. Chem. 2006, 220, 1395–1405; c) F. V.

Bright, G. A. Baker, J. Phys. Chem. B 2006 , 110 , 582 2 –5 823; d) T.

Kçddermann, C. Wertz, A. Heintz, R. Ludwig, Angew. Chem. 2006 ,

118 , 378 0 –3 785; Angew. Chem. Int. Ed. 2006 , 45 , 369 7 –3 702; e) C.

Wakai, A. Oleinikova, M. Ott, H. Weingꢄrtner, J. Phys. Chem. A

2005, 109, 17028–17030; f) A. Kawai, T. Hidemori, K. Shibuya,

Chem. Lett. 2004 , 33 , 146 4 –1 465.

[27] a) K.-S. Yeung, M. E. Farkas, Z. Qiu, Z. Yang, Tetrahedron Lett.

2002 , 43 , 5793; b) J. A. Boon, J. A. Levisky, J. L. Pflug, J. S. Wilkes,

J. Org. Chem. 1986 , 51 , 480.

[28] a) S. V. Dzyuba, R. A. Bartsch, Tetrahedron Lett. 2002 , 43 , 465 7 –

4659; b) P. Ludley, N. Karodia, Tetrahedron Lett. 2001 , 42 , 201 1 –

2014; c) M. J. Earle, P. B. McCormac, K. R. Seddon, Green Chem.

1999 , 1 , 2325; d) T. Fisher, A. Sethi, T. Welton, J. Wolf, Tetrahedron

Lett. 1999, 40, 793–796.

[29] a) C. J. Mathews, P. J. Smith, T. Welton, Chem. Commun. 2000 ,

124 9 –1 250; b) C. J. Mathews, P. J. Smith, T. Welton, A. J. P. White,

D. J. Williams, Organometallics 2001 , 20 , 384 8 –3 850.

[30] a) P. E. Z. Suarez, J. E. L. Dullius, S. Einloft, R. F. de Souza, J.

Dupont, Polyhedron 1996 , 15 , 121 7 –1 219; b) L. Mußmann, H. Olivi-

er, Y. Chauvin, Angew. Chem. 1995, 107, 2941–2943; Angew. Chem.

Int. Ed. Engl. 1995, 34, 2698–2700; c) C. E. Song, E. J. Roh, Chem.

Commun. 2000 , 83 7 –8 38; d) D. E. Kaufmann, M. Nouroozian, H.

Henze, Synlett 1996 , 109 1 –1 092.

[31] A. Dzudza, T. J. Marks, Org. Lett. 2009 , 11 , 152 3 –1 526.

[32] Thorpe–Ingold effect: a) S. M. Bachrach, J. Org. Chem. 2008 , 73 ,

246 6 –2 468; b) P. G. Sammes, D. J. Weller, Synthesis 1995 , 120 5 –

1222.

[33] a) G. A. Olah, P. V. Reddy, G. K. Prakash, Chem. Rev. 1992 , 92 ,

6995; b) H. N. C. Wong, M. Y. Hon, C. W. Tse, Y. C. Yip, J. Tanko,

T. Hudlicky, Chem. Rev. 1989 , 89 , 16 5 –1 98; c) M. Brookhart, W. B.

Studabaker, Chem. Rev. 1987 , 87 , 41 1 –4 32.

[34] R. D. Shannon, Acta Crystallogr. Sect. A 1976 , 32 , 75 1 –7 67.

[35] a) S. W. Benson, Thermochemical Kinetics, 2nd ed., Wiley, New

York, 1986, pp. 810; b) P. J. Robinson, J. Chem. Educ. 1978 , 55 ,

50951 0; c) P. M. Morse, M. D. Spencer, S. R. Wilson, G. S. Girolami,

Organometallics 1994 , 13 , 164 6 –1 655.

[36] Parameters in parentheses represent three values derived from the

least-squares fit.

[37] a) R. P. Bell, The Proton in Chemistry, 2nd ed., Cornell University

Press, Ithaca, New York, 1973, Chapter 12; b) L. Melander, W. H.

Saunders, Jr., Reaction Rates of Isotopic Molecules, Wiley, New

York, 1980.

[13] For lanthanide-mediated hydroamination, see: a) S. Hong, T. J.

Marks, Acc. Chem. Res. 2004 , 37 , 67 3 –6 86; b) J.-S. Ryu, T. J. Marks,

F. E. McDonald, J. Org. Chem. 2004 , 69 , 103 8 –1 052; c) J.-S. Ryu, Y.

Li, T. J. Marks, J. Am. Chem. Soc. 2003 , 125 , 1258 4 –1 2605; d) S.

Hong, T. J. Marks, J. Am. Chem. Soc. 2002 , 124 , 788 6 –7 887; e) J.-S.

Ryu, T. J. Marks, F. E. McDonald, Org. Lett. 2001 , 3 , 309 1 –3 094;

f) V. M. Arredondo, F. E. McDonald, T. J. Marks, Organometallics

1999 , 18 , 194 9 –1 960; g) Y. Li, T. J. Marks, J. Am. Chem. Soc. 1998 ,

120 , 175 7 –1 771; h) Y. Li, T. J. Marks, Organometallics 1996 , 15 ,

377 0 –3 772; i) C. M. Haar, C. L. Stern, T. J. Marks, Organometallics

1996 , 15 , 176 5 –1 775; j) Y. Li, T. J. Marks, Organometallics 1994 , 13 ,

43 9 –4 40; k) M. R. Gagnꢆ, C. L. Stern, T. J. Marks, J. Am. Chem.

Soc. 1992, 114, 275–294.

[14] For lanthanide-mediated hydrophosphination, see: a) A. M. Kawao-

ka, M. R. Douglass, T. J. Marks, Organometallics 2003 , 22 , 463 0 –

4632; b) M. R. Douglass, M. Ogasawara, S. Hong, M. V. Metz, T. J.

Marks, Organometallics 2002 , 21 , 28 3 –2 92; c) M. R. Douglass, C. L.

Stern, T. J. Marks, J. Am. Chem. Soc. 2001 , 123 , 1022 1 –1 0238.

[15] For lanthanide-mediated hydrosilylation and hyboration, see: a) P.-

F. Fu, L. Brard, Y. Li, T. J. Marks, J. Am. Chem. Soc. 1995 , 117 ,

715 7 –7 168; b) K. N. Harrison, T. J. Marks, J. Am. Chem. Soc. 1992 ,

114 , 922 0 –9 221.

[16] Y. Li, T. J. Marks, J. Am. Chem. Soc. 1998 , 120 , 175 7 –1 771.

[17] a) S. B. Amin, T. J. Marks, J. Am. Chem. Soc. 2007 , 129 , 1010 2 –

10103; b) A. M. Kawaoka, T. J. Marks, J. Am. Chem. Soc. 2005 , 127 ,

631 1 –6 324; c) V. C. Gibson, S. K. Spitzmesser, Chem. Rev. 2003 ,

103 , 28 3 –3 16; d) Y.-X. Chen, T. J. Marks, Chem. Rev. 2000, 100 (spe-

cial issue on “Frontiers in Metal-Catalyzed Polymerization”).

[18] a) S. P. Nolan, D. Stern, T. J. Marks, J. Am. Chem. Soc. 1989 , 111 ,

784 4 –7 853; b) D. F. McMillen, D. M. Golden, Annu. Rev. Phys.

Chem. 1982 , 33 , 49 3 –5 32; c) S. W. Benson, Thermochemical Kinetics,

2nd ed., Wiley, New York, 1976, Appendix Tables A10, 11, and 22;

d) M. A. Giardello, W. A. King, S. P. Nolan, M. Porchia, C. Sishta,

T. J. Marks in Energetics of Organometallic Species (Ed.: J. A. Mar-

tinho Simoes), Kluwer Academic, Dordrecht, 1992, pp. 35–54;

e) M. R. Gagnꢆ, S. P. Nolan, A. M. Seyan, D. Stern, T. J. Marks in

Metal–Metal Bonds and Clusters in Chemistry and Catalysis (Ed.:

J. P. Fackler, Jr.), Plenum Press, New York, 1990, pp. 113–125;

f) T. J. Marks in Bonding Energetics in Organometallic Compounds

(Ed.: T. J. Marks), ACS Symposium Series 428, American Chemical

Society, Washington, DC, 1990, pp. 1–18; g) S. P. Nolan, D. Stern, D.

Hedden, T. J. Marks in Bonding Energetics In Organometallic Com-

pounds (Ed.: T. J. Marks), ACS Symposium Series 428, American

Chemical Society, Washington, DC, 1990, pp. 159–174.

[19] Reviews of [LnACHTUNGTRENNUNG(OTf)₃]-mediated reactions: a) C. J. Li, L. Chen,

Chem. Soc. Rev. 2006 , 35 , 6882; b) S. Luo, L. Zhu, A. Ralukdar, G.

Zhang, X. Mi, J. P. Chen, P. G. Wang, Mini-Rev. Org. Chem. 2005 , 2 ,

17 7 –2 02; c) C. Li, Chem. Rev. 2005 , 105 , 309 5 –3 165; d) S. Kobaya-

shi, Chem. Rev. 2002, 102, 2227–2302; e) Lanthanides: Chemistry

and Use in Organic Synthesis, Series: Topics in Organometallic

Chemistry, Vol. 2 (Ed.: S. Kobayashi), Springer, Berlin, 1999, Chap-

ter 2.

[20] Lanthanide chemistry: a) T. Imamoto, Lanthanides in Organic Syn-

thesis, Academic Press, San Diego, 1994; b) S. Cotton, Lanthanide

and Actinide Chemistry, Wiley, Chichester, England, 2006; c) H. C.

Aspinall, Chemistry of f-Elements, Gordon and Breach Science Pub-

lishers, Amsterdam, Netherlands, 2001; d) Aqueous-Phase Organo-

metallic Catalysis (Eds.: B. Cornils, W. A. Herrman), Wiley, New

York, 1998.

[21] a) G. A. Molander, Chem. Rev. 1992 , 92 , 2968; b) G. Lawrence,

Chem. Rev. 1986, 86, 1733.

[22] [LnACHTUNGTRENNUNG(OTf)₃]-mediated C–C bond-forming reactions: a) S. Kobayashi,

I. Hachiya, J. Org. Chem. 1994 , 59 , 359 0 –3 596; b) J. B. F. N. Eng-

berts, J. W. Wijnen, J. Org. Chem. 1997, 62, 2039–2044; c) C. Belluc-

[38] J. L. McBee, A. T. Bell, D. T. Tilley, J. Am. Chem. Soc. 2008 , 130 ,

1656 2 –1 6571.

Chem. Eur. J. 2010, 16, 3403 – 3422

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

3421

A. Dzudza and T. J. Marks

[39] For proton traps, see: a) T. Saupe, H. A. Staab, Angew. Chem. 1988,

100, 895–909; Angew. Chem. Int. Ed. Engl. 1988, 27, 865–1008;

b) R. W. Alder, M. R. Bryce, N. C. Goode, J. Chem. Soc. Perkin

Trans. 2 1982 , 47 7 –4 83; c) A. Awwal, F. Hibbert, J. Chem. Soc.

Perkin Trans. 2 1982, 1589–1592; d) B. Kanner, H. C. Brown, J. Am.

Chem. Soc. 1966, 88, 986–992.

[40] a) M. H. Alizadeh, H. Razavi, F. F. Bamoharram, K. Daneshvar, J.

Mol. Catal. A 2003, 206, 89–93; b) A. Alvanipour, C. Eaborn,

D. R. M. Walton, J. Organomet. Chem. 1980 , 201 , 23 3 –2 47; c) R. W.

Bott, C. Eaborn, P. M. Greasley, J. Chem. Soc. 1964 , 480 4 –4 806.

[41] B. Xu, M. Shi, Org. Lett. 2002, 4, 2145–2148.

[42] a) C. K. Ingold, Structure and Mechanism in Organic Chemistry, 2nd

ed., Bell, London, 1969; b) E. D. Hughes, C. K. Ingold, J. Chem.

Soc. 1935 , 244; c) E. D. Hughes, Trans. Faraday Soc. 1941 , 37 , 60 3 –

631; d) E. D. Hughes, C. K. Ingold, Trans. Faraday Soc. 1941 , 37 ,

65 7 –6 85; e) K. A. Cooper, M. L. Dhar, E. D. Hughes, C. K. Ingold,

B. J. MacNulty, L. I. Woolf, J. Chem. Soc. 1948 , 204 3 –2 049.

i) T. E. Mꢃller, M. Beller, Chem. Rev. 1998 , 98 , 67 5 –7 04; j) J. P. Coll-

man, L. S. Hegedus, J. R. Norton, R. G. Finke in Principles and Ap-

plications of Organotransition Metal Chemistry (Ed.: A. Kelly), Uni-

versity Science Books, Mill Valley, 1987.

[48] a) I. Bertini, C. Luchinat, Coordination Chemistry Reviews: NMR of

Paramagnetic Substances, Vol. 150 (Ed.: A. B. P. Lever), Elsevier,

Netherlands, 1996, Chapter 19; b) B. R. McGarvey, Can. J. Chem.

1984 , 62 , 134 9 –1 355; c) B. R. McGarvey, J. Chem. Phys. 1976 , 65 ,

95 5 –9 61; d) W. D. Luke, A. Streitwieser, Jr., ACS Symp. Ser. 1980 ,

131 , 9 3 –1 40; e) W. D. Horrocks, Jr., NMR of Paramagnetic Mole-

cules (Eds.: G. N. La Mar, W. D. Horrocks, R. H. Holm), Academic

Press, New York, 1973, Chapter 4.

[49] A. B. Pangborn, M. A. Giardello, R. H. Grubbs, R. K. Rosen, F. J.

Timmers, Organometallics 1996 , 15 , 151 8 –1 520.

[50] A. E. Hill, D. C. Link, P. Donndelinger, J. Org. Chem. 1981 , 46 ,

117 7 –1 182.

[51] M. Bradley, P. J. May, D. C. Harrowven, Tetrahedron Lett. 2003, 44,

503–506.

[52] T. Okazaki, Y. Maeda, T. Sugai, K. Mori, Tetrahedron 1985, 41, 307–

311.

[53] P. H. Teijꢈn, L. M. Burꢈn, R. R. Clemente, R. R. Gonzꢅlez, A.

Fernꢅndez-Mateos, J. Org. Chem. 2007, 72, 9973–9982.

[54] P. S. Engel, A. M. Culotta, J. Am. Chem. Soc. 1991 , 113 , 268 6 –2 696.

[55] A. Takemiya, J. F. Hartwig, J. Am. Chem. Soc. 2006 , 128 , 604 2 –

6043.

[43] J. L. Anderson, J. Ding, T. Welton, D. W. Armstrong, J. Am. Chem.

Soc. 2002 , 124 , 1424 7 –1 4254.

[44] For Ln³⁺–arene complexes, see: a) H. Liang, Q. Shen, J. Guan, Y.

Lin, J. Organomet. Chem. 1994, 474, 113–116; b) G. B. Deacon,

Aust. J. Chem. 1990, 43, 1245–1257; c) B. Fan, Q. Shen, Y. Lin, J.

Organomet. Chem. 1989 , 376 , 6166; d) B. Fan, Q. Shen, Y. Lin, J. Or-

ganomet. Chem. 1989 , 377 , 5158; e) A. Cotton, W. Schwotzer, Orga-

nometallics 1987 , 6 , 127 5 –1 279; f) A. Cotton, W. Schwotzer, J. Am.

Chem. Soc. 1986 , 108 , 465 7 –4 568.

[56] S. M. Perala, B. P. Conrad, C. A. Hughey, J. A. Shugart, G. W.

Breton, Molecules 2001, 6, 655–662.

[57] P. Yates, T. S. Macas, Can. J. Chem. 1988, 66, 110.

[58] M. C. Marcotullio, V. Campagna, S. Sternativo, F. Constantini, M.

Curini, Synthesis 2006 , 276 0 –2 766.

[59] B. Janza, A. Studer, Org. Lett. 2006 , 8 , 187 5 –1 878.

[60] E. C. Ashby, R. N. DePriest, A. B. Goel, B. Wenderoth, T. N. Pham,

J. Org. Chem. 1984 , 49 , 354 5 –3 556.

[45] a) A. Motta, I. L. Fragala, T. J. Marks, Organometallics 2006 , 25 ,

553 3 –5 539; b) A. Motta, I. L. Fragala, T. J. Marks, Organometallics

2005 , 24 , 499 5 –5 003; c) A. Motta, G. Lanza, I. A. Fragala, T. J.

Marks, Organometallics 2004 , 23 , 409 7 –4 114; d) A. Motta, G.

Lanza, I. L. Fragalꢇ, T. J. Marks, Organometallics 2004 , 23 , 409 7 –

4104; e) Y. K. Kim, T. Livinghouse, J. E. Bercaw, Tetrahedron Lett.

2001 , 42 , 293 3 –2 935.

[46] a) R. Breslow in Molecular Rearrangements (Ed.: P. de Mayo), Inter-

science, New York, 1963, pp. 233–294; b) P. H. Mazzocchi, H. J.

Tamburin, J. Am. Chem. Soc. 1970 , 92 , 722 0 –7 221.

[61] a) A. Dzudza, T. J. Marks, J. Org. Chem. 2008 , 73 , 400 4 –4 016;

b) J. L. Pascal, M. M. Hamidi, Polyhedron 1994, 13, 1787–1792;

c) A. Abbasi, P. Lindqvist-Reis, L. Eriksson, D. Sandstrom, S. Lidin,

I. Persson, M. Sandstrom, Chem. Eur. J. 2005 , 11 , 406 5 –4 077; d) K.

Egashira, Y. Yoshimura, H. Kanno, Y. Suzuki, J. Therm. Anal. Calo-

rim. 2003 , 71 , 50 1 –5 08; e) H. Zineddine, M. Hnach, M. Hamidi, J.

Fluorine Chem. 1998, 88, 139–141; f) M. Nakayama, S. Nakamura,

N. Yanagihara, Polyhedron 1998, 17, 3625–3631.

[47] a) J. Muzart, Tetrahedron 2008, 64, 5955–6008; b) M. R. Gagnꢆ, S. J.

Lee, A. R. Chianese, Angew. Chem. 2007, 119, 4118–4136; Angew.

Chem. Int. Ed. 2007, 46, 4042–4059; c) C. Munro-Leighton, S. A.

Delp, E. D. Blue, B. T. Gunnoe, Organometallics 2007 , 26 , 148 3 –

1493; d) J. Muzart, Tetrahedron 2005, 61, 5815–5849; e) Transition

Metals for Organic Synthesis, Vols. 1,2 (Eds.: M. Beller, C. Bolm),

Wiley-VCH, Weinheim, 2004; f) H. Jiao, A. Tillack, J. Seayad, M.

Beller, Angew. Chem. 2004, 116, 3448–3479; Angew. Chem. Int. Ed.

2004, 43, 3368–3398; g) E. Hahn, Chem. Eur. J. 2004 , 10 , 588 8 –

5899; h) L. S. Hegedus in Transition Metals in the Synthesis of Com-

plex Organic Molecules, University Science Books, Mill Valley, 1999;

[62] P. Bonhꢉte, A.-P. Dias, N. Papageorgiou, M. Kalyansundaram, K.

Grꢄtzel, Inorg. Chem. 1996, 35, 1168–1178.

[63] SigmaPlot 2000 version 6.0., SPSS Inc, Chicago, IL, 2000.

Received: August 16, 2009

Published online: February 9, 2010

3422

www.chemeurj.org

ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2010, 16, 3403 – 3422

Article Doi

Efficient intramolecular hydroalkoxylation of unactivated alkenols mediated by recyclable lanthanide lriflate ionic liquids: Scope and mechanism

DOI: 10.1002/chem.200902269

Source and publish data:

Authors:

Article abstract of DOI:10.1002/chem.200902269

Full text of DOI:10.1002/chem.200902269

Products guided by the article

R&D Labs maybe for 21865-78-1

Relevant to this article

Hot Product