Kinetics of an air- and water-stable ruthenium(IV) catalyst for the deprotection of allyl alcohol in water

Article abstract of DOI:10.1021/om100892v

Kynurenic allyl ester reacts readily with [CpRu(NCCH₃) ₃]PF₆ in dry acetone to yield the Ru^IV allyl. The complex is an effective catalyst for the hydrolysis of methyl allyl carbonate in aqueous solution in air yet decomposes rapidly in aerobic solutions of methanol. Kinetic studies are consistent with rate-limiting attack of water (or methanol) on the Ru^IV allyl as a key step. A polystyrene- supported version of the catalyst was active for the catalytic hydrolysis of methylallyl carbonate, but kinetic studies indicate that the Ru leaches from polystyrene upon repeated catalytic cycles.

Full text of DOI:10.1021/om100892v

Organometallics 2010, 29, 6051–6056 6051
DOI: 10.1021/om100892v
Kinetics of an Air- and Water-Stable Ruthenium(IV) Catalyst for
the Deprotection of Allyl Alcohol in Water
Matthew K. Kiesewetter and Robert M. Waymouth*
Department of Chemistry, Stanford University, Stanford, California 94305, United States
Received September 15, 2010
Kynurenic allyl ester reacts readily with [CpRu(NCCH3)3]PF6 in dry acetone to yield the RuIV
allyl. The complex is an effective catalyst for the hydrolysis of methyl allyl carbonate in aqueous
solution in air yet decomposes rapidly in aerobic solutions of methanol. Kinetic studies are consistent
with rate-limiting attack of water (or methanol) on the RuIV allyl as a key step. A polystyrene-
supported version of the catalyst was active for the catalytic hydrolysis of methylallyl carbonate,
but kinetic studies indicate that the Ru leaches from polystyrene upon repeated catalytic cycles.
Scheme 1. Synthesis of Kitamura’s Catalyst, 1
Introduction
The rich reaction chemistry of transition metal allyl com-
plexes has spawned a diverse array of catalytic reactions for
selective bond forming and cleaving reactions.1-4 Catalytic
nucleophilic allylic substitution is a powerful synthetic
method for the selective formation of carbon-carbon and
carbon-heteroatom bonds.1-3 Allylic ethers, esters, carbo-
nates, and carbamates are also versatile protecting groups
due to their high reactivity with many transition metals.1,5-10
Several metals are known to cleave allyl esters/ethers
in protic10-12 or aprotic solvents,6,7,13 and Ru complexes,
including commercially available [CpRu(CH3CN)3]PF6,14,15
are particularly attractive due to their tolerance to many
functional groups.16-21 Kitamura reported Ru complexes
derived from [CpRu(CH3CN)3]PF6 and quinaldic acid that
are particularly effective for deallylation reactions that op-
erate in protic solvents and feature an internal base,5,13,22-24
which obviates the need for additives.10-12 Complex 1
exhibits a wide substrate tolerance in alcoholic solvents13
and has been shown to reversibly allylate a variety biological
molecules, including peptides and nucleic acids.5,23,25 As part
of a program on the synthesis of water-soluble bioactive
polymers,26 we sought a facile method for the deprotection of
allylic carbonates or carbamates in aqueous solution. Here-
in, we report our investigations on the kinetics and mecha-
nism of the hydrolysis of allyl carbonates in water in the
presence of both homogeneous and supported analogues of
complex 1. We also compare the rates of hydrolysis and
methanolysis of allyl methyl carbonate in both aqueous and
methanol solutions.
*To whom correspondence should be addressed. E-mail: waymouth@
stanford.edu.
(1) Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96, 395–422.
(2) Trost, B. M.; Crawley, M. L. Chem. Rev. 2003, 103, 2921–2944.
(3) Tietze, L. F.; Ila, H.; Bell, H. P. Chem. Rev. 2004, 104, 3453–3516.
(4) Consiglio, G.; Waymouth, R. M. Chem. Rev. 1989, 89, 257–276.
(5) Tanaka, S.; Saburi, H.; Ishibashi, Y.; Kitamura, M. Org. Lett.
2004, 6, 1873–1875.
(6) Greene, T. W.; Wuts, P. G. M., Eds. Protective Groups in Organic
Synthesis, 3rd ed.; John Wiley & Sons: New York, 2000.
(7) Tsuji, J.; Mandai, T. Synthesis 1996, 1–24.
(8) Trost, B. M.; Fullerton, T. J. J. Am. Chem. Soc. 1973, 95, 292–294.
(9) Tsuji, J.; Takahashi, H.; Morikawa, M. Tetrahedron Lett. 1965, 6,
4387–4388.
(10) Guibe, F. Tetrahedron 1998, 54, 2967–3042.
(11) Genet, J. P.; Savignac, M. J. Organomet. Chem. 1999, 576, 305–
317.
(12) Genet, J. P.; Blart, E.; Savignac, M.; Lemeune, S.; Lemaireaudoire,
S.; Paris, J. M.; Bernard, J. M. Tetrahedron 1994, 50, 497–503.
(13) Tanaka, S.; Hajime, S.; Murase, T.; Ishibashi, Y.; Kitamura, M.
J. Organomet. Chem. 2007, 692, 295–298.
(14) Trost, B. M.; Older, C. M. Organometallics 2002, 21, 2544–2546.
(15) Trost, B. M.; Fraisse, P. L.; Ball, Z. T. Angew. Chem., Int. Ed.
2002, 41, 1059–1061.
(16) Morisaki, Y.; Kondo, T.; Mitsudo, T.-a. Organometallics 1999,
18, 4742–4746.
(19) Achard, M.; Derrien, N.; Zhang, H.-J.; Demerseman, B.;
Bruneau, C. Org. Lett. 2008, 11, 185–188.
(20) Gruber, S.; Zaitsev, A. B.; Woirle, M.; Pregosin, P. S.; Veiros,
ꢀ
^
L. F. Organometallics 2008, 27, 3796–3805.
(21) Bruneau, C.; Renaud, J.-L.; Demerseman, B. Chem.—Eur. J.
2006, 12, 5178–5187.
(22) Saburi, H.; Tanaka, S.; Kitamura, M. Angew. Chem., Int. Ed.
2005, 44, 1730–1732.
(23) Tanaka, S.; Saburi, H.; Murase, T.; Yoshimura, M.; Kitamura,
M. J. Org. Chem. 2006, 71, 4682–4684.
(24) Tanake, S.; Saburi, H.; Kitamura, M. Adv. Synth. Catal. 2006,
348, 375–378.
(25) Tanaka, S.; Hirakawa, T.; Oishi, K.; Hayakawa, Y.; Kitamura,
(17) Hermatschweiler, R.; Ferandez, I.; Breher, F.; Pregosin, P. S.;
Veiros, L. F.; Calhorda, M. J. Angew. Chem., Int. Ed. 2005, 44, 4397–
4400.
M. Tetrahedron Lett. 2007, 48, 7320–7322.
(18) Fernandez, I.; Hermatschweiler, R.; Breher, F.; Pregosin, P. S.;
Veiros, L. F.; Calhorda, M. J. Angew. Chem., Int. Ed. 2006, 45, 6386–
6391.
(26) Cooley, C. B.; Trantow, B. M.; Nederberg, F.; Kiesewetter,
M. K.; Hedrick, J. L.; Waymouth, R. M.; Wender, P. A. J. Am. Chem.
Soc. 2009, 131, 16401–16403.
r
2010 American Chemical Society
Published on Web 10/27/2010
pubs.acs.org/Organometallics

6052 Organometallics, Vol. 29, No. 22, 2010
Kiesewetter and Waymouth
Scheme 2. Synthesis of the Ru Complex 4
Results and Discussion
where kobs = k0[4] = (1.25 ( 0.05) ꢁ 10-3 M h-1 (20 °C),
Figure 2.
Quinaldic acid and pyridine carboxylic acids have proven
to be effective ligands for Ru-catalyzed allylation reactions
and allyl alcohol/carbonate deprotection.5,13,22-24 We
sought a readily functionalized derivative of quinaldic acid
that might enable facile anchoring of Ru complexes to
inorganic27 or organic supports. Kynurenic acid (4-hydro-
xyquinoline-2-carboxylicacid), ametaboliteof tryptophan,28,29
provided an attractive and readily available synthon. The
kynurenic acid allyl ester can be readily alkylated with a
variety of functional groups to generate a series of substi-
tuted kynurenic acid derivatives. The kynurenic allyl ester
2 reacts with [CpRu(NCCH3)3]PF6 in dry acetone to yield the
orange-yellow RuIV allyl 4, which precipitates from concen-
trated solution (Scheme 2).22,24,30
The mechanism proposed for the deprotection of allyl
carbonates in alcohol solvents was proposed to involve the
nucleophilic attack of the alcohol on the RuIV allyl species as
a key step.5,13,23 An analogous mechanism for the aqueous
hydrolysis of allyl methyl carbonate is shown in Scheme 3.
This mechanism assumes that attack of water on the RuIV
allyl generates allyl alcohol and a solvated RuII species,
which subsequently reacts with allyl methyl carbonate to
regenerate the RuIV allyl species 4.
A rate law for the mechanism in Scheme 3 can be derived if
a low, steady-state concentration of the RuII intermediate
and the reversible formation of 4 from allyl alcohol is
assumed (eqs 3 and 4).22,30
The Ru allyl complex 1, derived from quinaldic acid, has
been shown to be an active catalyst for the deallylation of
allyl carbonates13 and carbamates23 in methanolic solutions
under argon (eq 1). For these reactions, it was proposed that
the RuIV allyl species is the resting state and that attack of
methanol on the Ru allyl is rate-limiting.5,13,22
To assess the activity of the analogous Ru allyl complex 4
in aqueous solution, we investigated the kinetics of deal-
lylation of allyl methyl carbonate in D2O, eq 1.
With these assumptions, the rate of disappearance of allyl
methyl carbonate can be derived (eq 5):
In the absence of Ru, allyl methyl carbonate is stable in
water for more than a week at room temperature. In the
presence of the Ru complex 4 (0.6-10 mol %), allyl methyl
carbonate hydrolyzes to allyl alcohol, methanol, and carbon
dioxide. In the solid state, complex 4 is stable in air for
months without the loss of activity. Kinetic studies carried
out in air at room temperature revealed that the hydrolysis of
allyl methyl carbonate was zero order in carbonate, Figure 1,
and first order in Ru to give a rate law:
d½Cꢀ
dt
k1k2½4ꢀ½H2Oꢀ½Cꢀ
k- 1½Aꢀ þ k2½Cꢀ
R ¼ -
¼
ð5Þ
where [C] = concentration of allyl methyl carbonate, [A] =
concentration of allyl alcohol, and [4] = concentration of
complex 4. This rate equation would be consistent with the
experimental rate law under conditions where the rate of
formation of 4 from the reaction of allyl methyl carbonate
with RuII is faster than the formation of 4 from the reaction
of RuII with allyl alcohol22,30 (k2[C] . k-1[A]), for which
- d½carbonateꢀ=dt ¼ R ¼ kobs
ð2Þ
(27) Hirakawa, T.; Tanaka, S.; Usuki, N.; Kanzaki, H.; Kishimoto,
M.; Kitamura, M. Eur. J. Org. Chem. 2009, 789–792.
(28) Erhardt, S.; Schwieler, L.; Nilsson, L.; Linderholm, K.; Engberg,
Rate ¼ k1½H2Oꢀ½4ꢀ, where k0 ¼ k1½H2Oꢀ
ð6Þ
G. Physiol. Behav. 2007, 92, 203–209.
These results imply that the attack of water on the RuIV
allyl complex 4 is rate limiting, even in aqueous solution
([H2O] ≈ 55.5 M; [D2O] ≈ 55.5 M, 20 °C), and that the RuIV
allyl complex 4 is the resting state during the catalytic
(29) Barth, M. C.; Ahluwalia, N.; Anderson, T. J. T.; Hardy, G. J.;
Sinha, S.; Alvarez-Cardona, J. A.; Pruitt, I. E.; Rhee, E. P.; Colvin,
R. A.; Gerszten, R. E. J. Biol. Chem. 2009, 284, 19189–19195.
(30) Tanaka, S. J.; Saburi, H.; Hirakawa, T.; Seki, T.; Kitamura, M.
Chem. Lett. 2009, 38, 188–189.

Article
Organometallics, Vol. 29, No. 22, 2010 6053
Section).35 Under these conditions, the equilibrium constant
(eq 3) was determined to be Keq = [RuII][A]/[4][D2O] =
0.0075 ( 0.0008 with k1= 0.036 ( 0.011 M-1 h-1 and k-1
=
4.80 ( 1.53 M-1 h-1
.
For comparison, the kinetics of the catalytic methanolysis
of allyl methyl carbonate with 4 were examined in CD3OD.
When carried out in air at room temperature, the methanoly-
sis of allyl methyl carbonate with 4 proceeds only to 40%
conversion under these conditions. Kinetic studies revealed
that the evolution of [allyl methyl carbonate] with time
is linear to approximately 25% conversion; at longer times,
the rate decays rapidly. Similar rates and conversions are
observed in the presence and absence of 5 equiv of allyl
methyl ether, suggesting that product inhibition by allyl
methyl ether is not the cause of decaying rate. However, if
the methanolysis of allyl methyl carbonate with 4 is carried
out under argon, the reaction proceeds to full conversion in
less than 90 min with a rate that is zero order in allyl
Figure 1. Plot of allyl methyl carbonate [C] vs time. Conditions:
[4]0 = 0.0005 M in D2O with (a) [C] = 0.105 M (triangles); (b) [C] =
0.054 M (squares); (c) [C] = 0.025 M (diamonds); (d) [C] =
0.013 M (circles).
carbonate, kobs(methanol) = (6.6 ( 0.2) ꢁ 10-2 M h-1
.
While we have not unambiguously determined the cause
of the low conversions observed for the methanolysis in air,
we propose that oxidative decomposition of the catalyst is
a likely cause, due to the higher conversions observed under
argon. We propose that different behaviors observed in
water and methanol are due to the higher solubility of oxygen
in methanol relative to water: χO2/water (298 K, 1 atm) =
2.3 ꢁ 10-5; χO2/CH3OH (298 K, 1 atm) = 4.2 ꢁ 10-4 36-38
.
As
the RuIV compounds 4 and 1 are shelf stable for months in
air, it is likely that oxidation of the RuII species is competitive
with catalysis in methanol, whereas in water the lower
solubility of O2 leads to longer catalysts lifetimes. If borne
out in further studies, this hypothesis highlights another
potential advantage of water as reaction solvent.
Figure 2. Plot of kobs (M h-1) vs Ru concentration [4]0. Con-
ditions: [allyl methyl carbonate]0 = 0.10 M in D2O.
A comparison of the rates of hydrolysis of allyl methyl
carbonate in water (under air) versus the methanolysis
(under argon) reveals that the deallylation of allyl methyl
carbonate is considerably faster in methanol. If we assume
that k0=k1[ROH] (for R=H or CH3) and that the concen-
tration of [D2O] = 55.5 M and [CD3OH] = 24.6 M, then
estimates for k1 reveal that the rate constant for nucleophilic
attack by methanol is approximately 60 times that of water
3
reaction (i.e., [Ru]T ≈ [4]0) under these conditions. Analysis
of the kinetics of the catalytic hydrolysis of allyl methyl
carbonate reaction in a 1:1 mixture of H2O/D2O solvent
revealed a kinetic isotope effect of kH/kD = 2.5, which is
consistent with proposals by Kitamura5,22,30 that nucleo-
philic attack on the RuIV allyl is coupled to deprotonation of
the nucleophile (H2O in the present case, ROH in previous
examples), presumably by the Ru carboxylate (Scheme 3).
The rate law derived in eq 5 assumes the reversible forma-
tion of 4 from allyl alcohol and the RuII, but an equally valid
rate law could be derived without this assumption (i.e.,
k-1[A] = 0). As CpRu complexes ligated by pyridine car-
boxylic acids or quinaldic acids have been shown to catalyze
allylic substitutions of allyl alcohols,22,30-34 we sought in-
dependent kinetic evidence for the equilibrium represented in
eq 3. To this end, we treated the Ru complex 4 with D2O in
acetone-d6 and measured both the equilibrium constant and
the rate of approach to equilibrium at 20 °C (see Experimental
(k1(methanol)=2.68 ( 0.10 M-1 h-1 39
; k1(water) = 0.045 (
0.002 M-1 h-1). These estimates imply that methanol is
considerably more reactive than water toward the RuIV allyl.
Further insights were obtained from measurements of the
equilibration of the RuIV allyl 4 and methanol in acetone.
The slow approach to equilibrium of the Ru complex
4 treated with methanol in acetone-d6, eq 7, was examined
at 20 °C.35 Under these conditions, the equilibrium con-
stant (eq 7) was determined to be Keq(methanol) = [RuII]-
[AE]/[4][CH3OH] = 0.0828 ( 0.0083 with k10 = 0.86 ( 0.26
M-1 h-1 and k-10= 10.4 ( 3.3 M-1 h-1 (where [AE] =
concentration of allyl methyl ether). These results confirm
the higher rate of attack of methanol on the RuIV allyl 4. In
either methanol or water, 4 is the predominant species at
(31) Tanaka, S.; Seki, T.; Kitamura, M. Angew. Chem., Int. Ed. 2009,
48, 8948–8951.
(32) Sundararaju, B.; Achard, M.; Demerseman, B.; Toupet, L.;
Sharma, G. V. M.; Bruneau, C. Angew. Chem., Int. Ed. 2010, 49,
2782–2785.
(36) Fischer, K.; Wilken, M. J. Chem. Thermodyn. 2001, 33, 1285–
1308.
(37) Battino, R.; Rettich, T. R.; Tominaga, T. J. Phys. Chem. Ref.
Data 1983, 12, 163–178.
(33) Gruber, S.; Pregosin, P. S. Adv. Synth. Catal. 2009, 351, 3235–
3242.
(34) Zaitsev, A. B.; Caldwell, H. F.; Pregosin, P. S.; Veiros, L. F.
Chem.—Eur. J. 2009, 15, 6468–6477.
(35) Espenson, J. H. Chemical Kinetics and Reaction Mechanisms, 2nd
ed.; McGraw-Hill: New York, 2002.
(38) Steinhoff, B. A.; Stahl, S. S. J. Am. Chem. Soc. 2006, 128, 4348–
4355.
(39) This corresponds to a turnover frequency of approximately
1 min-1; for comparison, water-soluble Pd catalysts catalyze the depro-
tection of allyl alkyl carbonateswith Et2NH with turnover frequencies of
approximately 2- 10 min-1 (ref 12).

6054 Organometallics, Vol. 29, No. 22, 2010
Kiesewetter and Waymouth
Scheme 3. Proposed Mechanism for Catalytic Hydrolysis of Allyl Methyl Carbonate in Water by RuIV-Allyl Complex 4a
a PF6 anion omitted for clarity.
Scheme 4. Synthesis of the Polystyrene-Supported Ru Complex 5
equilibrium, but the equilibrium constant is higher in the case
of methanol than for water in acetone solution.
The amino-functionalized ligand 3 was coupled to a
chloromethyl polystyrene resin (1% cross-linked, 200-400 mesh,
3.5-4.5 mmol Cl-/g) to generate the polystyrene-supported
ligand PS-3. Microanalysis of PS-3 provided an estimate of
1.164 mmol/g for the amount of ligand loaded onto the poly-
styrene. Treatment of PS-3 with excess [CpRu(CH3CN)3]-
PF6 for 1.5 h in acetone generated the polystyrene-supported
complex 5, Scheme 4.
The supported catalyst 5 is competent for the hydrolysis of
allyl methyl carbonate in aqueous solution. Conducting the
hydrolysis of allyl methyl carbonate under identical condi-
tions with the supported Ru complex 5 and homogeneous
analogue 4 allowed for the direct comparison of the rates,
enabling an estimate of the Ru loading on the PS bead.47,48
When 0.9 mg of 5 was used for the hydrolysis of allyl methyl
carbonate in 0.32 mL of D2O, the rate constant, kobs, from
the zero-order plot of [allyl methyl carbonate] versus time of
this reaction was compared to the kobs versus [4] plot
(Figure 2) to determine an effective concentration of ruthe-
nium, [Ru]eff = 1.847 ( 0.047 mM, and a catalyst loading of
0.66 ( 0.15 mmol/g of active RuIV on the 5 bead. Combus-
tion analysis of 5 provided an independent estimate of the Ru
content in 5, [Ru] = 0.48 ( 0.04 mmol/g, which is within
error of the loading determined by kinetic analysis.
Studies of the catalytic hydrolysis of allyl methyl carbon-
ate suggest 4 is air/water compatible. To facilitate removal of
the catalyst from the products, we generated a solid-sup-
ported version27 of the Ru catalyst. Solid-supported cata-
lysts facilitate the removal and recycling of metal residues,40
and several heterogeneous catalysts for (de)allylation have
been reported.41-44 Indeed, among other solid-supported allyla-
tion catalysts,44-46 a heterogeneous version of 1 supported
on silica was developed by Kitamura and co-workers.27 The
supported catalyst provided comparable yields and rates to
the homogeneous analogues27 and exhibited no leaching
from the solid support after nine reaction cycles in methanol
solvent under argon, but only one instance of water being
used as a cosolvent was reported.27
After 4 h of reaction time, the supported catalyst was
isolated from the reaction solution by filtering, washed with
allyl alcohol and acetone, removed of volatiles under high
vacuum, and subjected to identical catalytic conditions. For
the second run, the rate was lower and the calculated catalyst
loading on 5 was 0.53 ( 0.06 mmol/g. This process was
(40) Bergbreiter, D. E.; Tian, J. H.; Hongfa, C. Chem. Rev. 2009, 109,
530–582.
(41) Miura, H.; Wada, K.; Hosokawa, S.; Sai, M.; Kondo, T.; Inoue,
M. Chem. Commun. 2009, 4112–4114.
(42) Li, G.; Zhao, G. Org. Lett. 2006, 8, 633–636.
(43) McNamara, C. A.; Dixon, M. J.; Bradley, M. Chem. Rev. 2002,
102, 3275–3300.
(44) Uozumi, Y.; Shibatomi, K. J. Am. Chem. Soc. 2001, 123, 2919–
2920.
(45) Uozumi, Y.; Danjo, H.; Hayashi, T. J. Org. Chem. 1999, 64,
3384–3388.
(46) Akiyama, R.; Kobayashi, S. Angew. Chem., Int. Ed. 2001, 40,
34693471.
(47) Demerseman, B.; Renaud, J. L.; Toupet, L.; Hubert, C.; Bruneau,
C. Eur. J. Inorg. Chem. 2006, 1371–1380.
(48) Phan, T. S.; Sluys, M. V. D.; Jones, C. M. Adv. Synth. Catal.
2006, 348, 609–679.

Article
Organometallics, Vol. 29, No. 22, 2010 6055
presence of air, catalyst 4 exhibits increased stability in water
versus methanol, although methanol is a better nucleophile
toward 4. A kynurenic acid ligand bearing a pendant amine
allows for the attachment of the catalyst to a solid PS
support. However, catalyst 5, a heterogeneous version of 4,
was observed to leach from the polystyrene support upon
several catalyst recycling experiments, suggesting that the
stability of supported versions of these homogeneous cata-
lysts depends sensitively on the nature of the support and
reaction conditions.
Experimental Section
Figure 3. Plots of [allyl methyl carbonate] vs hours for the
deallylation of allyl methyl carbonate in D2O catalyzed by 5,
recycled over successive runs.
General Considerations. All materials were purchased from
Aldrich and used as received unless stated otherwise. Tris-
(acetonitrile)cyclopentadienylruthenium(II) hexafluorophos-
phate was purchased from Strem and used as received. All
preparations of ruthenium(IV) compounds were carried out in
an inert atmosphere using standard glovebox or Schlenk tech-
niques but stored in atmosphere in closed vials. Combustion
analyses were conducted on a Perkin-Elmer TGA 7 thermo-
gravimetric analyzer.
Synthesis of Kynurenic Acid Allyl Ester. Allyl bromide (1.4
mL, 16.6 mmol) was added in a single aliquot to a vigorously
stirred mixture of kynurenic acid (2.0 g, 10.6 mmol) and KHCO3
(1.7 g, 17.4 mmol) in 60 mL of DMF. The single-neck, round-
bottom flask was equipped with reflux condenser and heated to
40 °C for 15 h. The reaction was quenched with water (∼60 mL)
and extracted three times with ethyl acetate. Combined organics
were washed with 5% aqueous NaCl, dried with MgSO4,
filtered, and removed of volatiles. Material was taken on with-
repeated, and on the third run the rate was even lower,
yielding an estimate of 0.31 ( 0.03 mmol/g, Figure 3.
The lower rates observed on subsequent runs imply that Ru is
leaching from the polystyrene support.49 As a test for leaching,
the supernatant from the washings was assessed for catalytic
activity;48 these experiments revealed that the supernatant
was catalytically active albeit at a much slower rate than 5
(approximately 20 times slower). We could not detect any free
kynurenic acid ligand in the supernatant by 1H NMR or ESI-
MS, implying that, in water, Ru is extracted off of the bead,
presumably by solvolysis of the kynurenic ligand, leaving PS-3
intact. While similar leaching phenomena have been observed
with other PS-supported Ru catalysts,50,51 this behavior is in
contrast to that reported by Kitamura for a silica-supported
version of 1, which did not exhibit loss of catalytic activity over
nine recycling cycles under an argon atmosphere.27 To test
whether the presence of air might contribute to the higher
leaching, we repeated the catalyst recycling experiments with 5
in water under argon. For these experiments, leaching was
attenuated but still evident with each catalyst cycle, giving
catalyst loadings of 0.84 ( 0.08,52 0.66 ( 0.15, and 0.31 (
0.03 mmol/g over three successive runs.
The higher degree of leaching observed for the polystyrene
catalyst 5 in water, relative to the silica-supported analogue,27
could be due to the nature of the solvent (water versus
methanol), the nature of the modified ligands (kynurenic acid
versus amide-substituted pyridine carboxylic acid), or the sup-
port itself. Arenes are known to displace coordinated ligands on
RuII species.53 The phenyl-rich environment of the PS bead
could labilize the Ru complexes; further studies are ongoing to
test these hypotheses.
1
out further purification. Yield: 1.575 g (65%). H NMR (300
MHz, CDCl3): δ 8.38 (1H, d, 8.37 Hz); 8.05 (1H, bs); 7.70 (1H, t,
7.75); 7.49 (1H, d, 8.40); 7.41 (1H, t, 7.75); 7.05 (1H, s); 6.4
(1H, m); 5.44 (2H, m); 4.93 (2H, d, 4.93). 13C NMR (125 MHz,
CDCl3): δ 162.9; 136.6; 130.6; 133.4; 103.8; 126.7; 126.5; 126.4;
124.8; 120.2; 118.4; 112.0; 67.8.
Synthesis of 2 (R=CH3). Methyl iodide (0.155 mL, 2.5 mmol)
was added in a single aliquot to a vigorously stirred mixture
of kynurenic acid allyl ester (0.505 g, 2.2 mmol) and Cs2CO3
(0.803 g, 2.5 mmol) in 130 mL of acetone. The single-neck,
round-bottom flask was fitted with a reflux condenser and
heated to reflux for 15 h. The reaction was quenched with brine
(∼70 mL) and extracted three times with ethyl acetate. The
combined organics were dried with MgSO4, filtered, andremoved
of volatiles. Material was taken on without further purification.
Yield: 0.301 g (56%). 1H NMR (300 MHz, acetone-d6): δ 8.22
(1H, d, 8.27 Hz); 8.12 (1H, d, 8.56 Hz); 7.84 (1H, t, 7.66 Hz); 7.68
(1H, t, 7.66 Hz); 7.59 (1H, s); 6.17 (1H, m); 5.44 (2H, m); 4.95
(2H, d, 5.83 Hz); 4.20 (3H, s). 13C NMR (125 MHz, acetone-d6):
δ 163.4; 149.9; 148.7; 132.9; 130.7; 130.2; 127.8; 122.2; 121.9;
118.2; 118.0; 100.3; 66.1; 56.1.
Conclusions
Synthesis of 4. 2 (0.21 g, 0.86 mmol) and [CpRu(NCCH3)3]PF6
(0.316 g, 0.73 mmol) were added to a flame-dried Schlenk flask with
a stir bar and removed from the glovebox to a Schlenk line. Under
N2, 10 mL of extra-dry acetone (Aldrich) was added via syringe.
The red solution produced a yellow precipitate over 25 min. The
stirring was stopped and, under N2 flow, the supernatant was
removed from the precipitate with a pipet and discarded. The
precipitate was washed with approximately 1 mL of extra-dry
acetone, and the precipitate was removed of volatiles under high
Cyclopentadienyl ruthenium complexes of a modified natural
product, kynurenic acid, were found to be an effective catalyst
for the deallylation of carbonates in water. A variety of
substituted kynurenic acid ligands are available using simple
procedures and commercially available materials. In the
(49) Catalyst decomposition could also contribute to the lower rates
upon recycling, but the activity of the supernatant implies that leaching
of Ru occurs.
1
vacuum, yielding pure 4: 0.218 g (54%). H NMR (300 MHz,
acetone-d6):δ 8.43 (1H, d, 8.74 Hz); 8.10 (2H, m); 7.91 (1H, t, 7.12);
7.59 (1H, s); 4.87 (2H, m); 4.70 (1H, d, 10.46 Hz); 4.45 (1H, m); 4.40
(1H, m); 4.35 (1H, s). Anal. Calcd for C19H18F6NO3PRu: 41.16 C,
3.27 H, 2.53 N. Found: 41.26 C, 3.08 H, 2.53 N.
(50) Nguyen, S. T.; Grubbs, R. H. J. Organomet. Chem. 1995, 497,
195–200.
(51) Nieczypor, P.; Buchowicz, W.; Meester, W. J. N.; Rutjes, F. P. J.
T.; Mol, J. C. Tetrahedron Lett. 2001, 42, 7103.
(52) The initial Ru loading estimated from the rates is higher under
argon atmosphere than under air.
Synthesis of 3. 2-Bromoethylamine HBr (0.21 g, 1.02 mmol)
was added in a single aliquot to a vigorously stirred mixture of
kynurenic acid allyl ester (0.22 g, 0.87 mmol) and Cs2CO3 (0.8 g,
3
(53) Fairchild, R. M.; Holman, K. T. Organometallics 2008, 27, 1823–
1833.

6056 Organometallics, Vol. 29, No. 22, 2010
Kiesewetter and Waymouth
2.5 mmol) in 65 mL of acetone. The single-neck, round-bottom
flask was fitted with a reflux condenser and heated to reflux for
15 h. The reaction was quenched with brine (∼70 mL) and
extracted three times with ethyl acetate. The combined organics
were dried with MgSO4, filtered, removed of volatiles, and taken
into acetone (material is self-labile when neat). Material was
taken on without further purification. Yield: 0.120 g (46%).
1H NMR (300 MHz, acetone-d6): δ 8.30 (1H, d, 8.4 Hz); 7.86
(1H, d, 8.4Hz);7.53(1H, t, 7.8Hz);7.27(1H, t, 7.8Hz); 7.06(1H,
s); 6.04 (1H, m); 5.41 (1H, d, 17.4 Hz); 5.22 (1H, d, 10.7); 4.79
(2H, d, 5.6); 4.36 (2H, t, 7.8); 3.59 (2H, t, 7.8 Hz); 3.48 (2H, bs).
13C NMR (125 MHz, acetone-d6): δ 163.0; 132.3; 132.1; 126.8;
125.3; 124.2; 120.9; 118.5; 110.3; 66.8; 64.6; 54.9; 53.0; 40.6. IR
(diamond anvil cell) cm-1: 1708 (s, CdO); 1607 (m, 1°NH); 1564
(m, C-C aromatic); 1095 (s, C-O ester); 760 (s, 1°NH).
Synthesis of PS-3. 3 (1.23 g, 4.5 mmol) and Merrifield’s
peptide resin (1% cross-linked, 200-400 mesh, 0.132 g, 0.49
mmol) were added to a round-bottom flask with a stir bar.
Acetone (100 mL) and Cs2CO3 (0.169 g, 0.52 mmol) were added,
and the reaction mixture was stirred for 19 h. The crude reaction
mixture was filtered, and the filtrate was washed with copious
quantities of acetone. IR (diamond anvil cell) cm-1: 1708
(s, CdO); 1564 (m, C-C aromatic); 1095 (s, C-O ester); 760
(s, 2°NH) (see Supporting Information).
Synthesis of 5. PS-3 (0.141 g, 0.52 mmol) and [CpRu-
(NCCH3)3]PF6 (0.5363 g, 1.24 mmol) were added to a flame-
dried Schlenk flask with a stir bar and removed from the glovebox
to a Schlenk line. Under N2, 15 mL of extra-dry acetone (Aldrich)
was added via syringe, and the reaction mixture was stirred for
1.5 h. The stirring was stopped and, under N2 flow, the supernatant
was removed from the precipitate with a pipet and discarded. The
precipitate was washed with copious extra-dry acetone and re-
moved of volatiles under high vacuum. IR (diamond anvil cell)
cm-1: 1587 (m); 1334 (m); 833 (s) (see Supporting Information).
Combustion analysis of 5: Samples were loaded onto the analytical
balance of the TGA under air flow and subjected to an annealing
program: 1 min hold at 50 °C, temperature ramp to 100 °C, 2 min
hold at 100 °C, temperature ramp to 900 °C, 1 min hold at 900 °C,
temperature ramp to 50 °C; all temperature ramps were 50 °C/min.
Initial sample masses were taken after the 100 °C temperature hold
and final masses at the end of the program. A sample of Merrifield’s
peptide resin subjected to the annealing program resulted in com-
plete loss of mass. After subjecting 4 (1.2567 mg, 0.00227 mmol) to
the annealing program, a black residue remained (0.3341 mg) that
provided a molecular weight of the fully oxidized material (0.3341
mg/0.00227 mmol =147.2 g/mol; RuO2 is 133 g/mol). In a repre-
sentative annealing, 5 (1.7490 mg) was reduced to a black residue
(0.1202 mg), which was taken to be 147.3 g/mol in molecular weight,
yielding a Ru loading on 5 of 0.467 mmol/g. Triplicate measure-
ments yield: 0.478 ( 0.034 mmol/g.
Kinetic Analysis to Give the Ru Loading on 5. In a typical
experiment, 5 (0.9 mg) was loaded directly into an NMR tube
along with D2O (0.32 mL) and allyl methyl carbonate (3.6 μL,
0.032 mmol), the tube was shaken to mix, and reaction progress
was monitored by 1H NMR against an internal standard
(acetone). From a (zero-order) plot of [allyl methyl carbonate]
versus hours (Figures 1), the slope was taken to be kobs, and this
value was divided by k0 (the slope of the kobs versus [Ru] plot
generated from the solution catalysis, Figure 2) to determine
a [Ru] in solution, 1.847 mM. Accounting for the solution
volume and mass of 5 in the given experiment gives the loading
of 5, 0.657 mmol/g of active RuIV on 5. The error of Ru loading
was propagated from the rate constant errors.
Equilibration of 4 and D2O in Acetone. Into an NMR tube
under argon was added 4 (2.6 mg, 0.0047 mmol), 1.0 mL of
acetone-d6, and D2O (8.6 μL, 0.4773 mmol), in order. (For
methanol the amounts were 4 (2.7 mg, 0.0049 mmol), 0.5 mL of
acetone-d6, and CD3OD (2.0 μL, 0.049 mmol) added in the given
order.) The tube was shaken to mix, and reaction progress was
monitored by 1H NMR (see Supporting Information). Concen-
trations of allyl alcohol and 4 were determined by integration
against an internal standard. The concentration of RuII was
determined from mass balance: [RuII] = [4]O - [4]. The reported
rate and equilibrium constant errors were propagated from the
error of the NMR integrations, which were taken to be (10%.
Typical Kinetic Experiment. 4 (1.8 mg, 0.0032 mmol) was
stirred with 1.1 mL of D2O (or other solvent) to yield a 2.95 mM
stock solution. In an NMR tube, 0.22 mL of stock solution was
diluted with 0.43 mL of D2O to give a final concentration of 4
(0.65 μmol, 0.001 M), to which was added allyl methyl carbonate
(7.35 μL, 0.065 mmol, 0.100 M) via syringe. The tube was
capped and shaken to mix. The reaction was monitored by
1H NMR (see Supporting Information), and concentrations
were determined versus an internal standard (acetone). The
reported rate constant errors were propagated from the error
of the NMR integrations, which were taken to be (10%. When
reactions were conducted under argon, the solvent was sparged
with a submerged jet of gas for 2 min, and the reaction vessel was
purged out before and after reagent addition.
Acknowledgment. We gratefully acknowledge support
from the NSF Center on Polymeric Interfaces and Macro-
molecular Assemblies (CPIMA: NSF-DMR-0213618)
and NSF-GOALI Grant NSF-CHE-0957386.
Supporting Information Available: 1H NMR and IR spectra
and relevant thermodynamic and kinetic data. This material is
available free of charge via the Internet at http://pubs.acs.org.