ORGANIC

LETTERS

Resonance Energy Transfer Approach 2009

Vol. 11, No. 13
and a New Ratiometric Probe for Hg2+ 2740-2743
in Aqueous Media and Living Organism
Moorthy Suresh, Sandhya Mishra, Sanjiv Kumar Mishra, E. Suresh,
Amal K. Mandal, Anupama Shrivastav, and Amitava Das*
Central Salt and Marine Chemicals Research Institute (CSIR), BhaVnagar 364002, Gujarat, India
amitaVa@[Link]

Received April 14, 2009

ABSTRACT

Resonance energy transfer from dansyl to the rhodamine moiety in a newly synthesized chemosensor L2 has been utilized successfully for
detection of Hg2+ in aqueous solution and living cells such as Pseudomonas putida.

A fluorescent chemosensor capable of sensing a specific Usually, there are three different photoinduced processes that
analyte has potential application in chemistry and biology,1 are involved in the signaling or response phenomena of
as this generally allows detection of an analyte present in luminescence based chemosensors, namely, PET (photoin-
ultratrace quantity. Such detection of heavy transition metal duced electron transfer),3 PCT (photoinduced charge trans-
ions is of paramount interest due to the high toxicity of these fer),4 and RET (resonance energy transfer).5 RET is a
metal ions toward human health and the environment.2
(3) (a) Kim, J. S.; Noh, K. H.; Lee, S. H.; Kim, S. K.; Kim, S. K.;
Yoon, J. J. Org. Chem. 2003, 68, 597–600. (b) Kim, J. S.; Shon, O. J.;
(1) (a) Nolan, E. M.; Lippard, S. J. Chem. ReV. 2008, 108, 3443–3480. Rim, J. A.; Kim, S. K.; Yoon, J. J. Org. Chem. 2002, 67, 2348–2351. (c)
(b) de Silva, A. P.; Gunaratne, H. Q. N.; Gunnlaugsson, T.; Huxley, A. J. M.; Aoki, L.; Sakaki, T.; Shinkai, S. J. Chem. Soc., Chem. Commun. 1992,
McCoy, C. P.; Rademacher, J. T.; Rice, T. E. Chem. ReV. 1997, 97, 1515– 730–732. (d) Jin, T.; Ichikawa, K.; Koyama, T. J. Chem. Soc., [Link]-
1566. (c) Callen, J. F.; de Silva, A. P.; Magri, D. C. Tetrahedron 2005, 61, mun. 1992, 499–501. (e) Ji, H.-F.; Brown, G. M.; Dabestani, R. Chem.
8551–8588. (d) Bell, T. W.; Hext, N. M. Chem. Soc. ReV. 2004, 33, 589– Commun. 1999, 609–610.
598. (e) Rurack, K.; Resch-Genger, U. Chem. Soc. ReV. 2002, 31, 116– (4) Leray, I.; Lefevre, J. P.; Delouis, J. F.; Delaire, J.; Valeur, B.
127. (f) Haugland, R. P. Handbook of Fluorescent Probes and Research [Link]. J. 2001, 7, 4590–4598.
Porducts, 9th ed.; Molecular Probes: Eugene, OR, 2002; Chapter 20. (5) (a) Lee, M. H.; Quang, D. T.; Jung, H. S.; Yoon, J.; Lee, C.-H.;
(2) (a) Suresh, M.; Ghosh, A.; Das, A. Chem. Commun. 2008, 3906– Kim, J. S. J. Org. Chem. 2007, 72, 4242–4245. (b) Ono, A.; Togashi, H.
3908. (b) Yoon, S.; Albers, A. E.; Wong, A. P.; Chang, C. J. J. Am. Chem. Angew. Chem., Int. Ed 2004, 43, 4300–4302. (c) Serin, J. M.; Brousmiche,
Soc. 2005, 127, 16030–16031. (c) Yoon, S.; Miller, E. W.; He, Q.; Do, D. W.; Frechet, J. M. J. J. Am. Chem. Soc. 2002, 124, 11848–11849. (d)
P. K.; Chang, C. J. Angew. Chem., Int. Ed. 2007, 46, 6658–6651. (d) Zhang, Arduini, M.; Felluga, F.; Mancin, F.; Rossi, P.; Tecilla, P.; Tonellato, U.;
G.; Zhang, D.; Yin, S.; Yang, X.; Shuai, Z.; Zhu, D. Chem. Commun. 2005, Valentinuzzi, N. Chem. Commun. 2003, 1606–1607. (e) Zhang, H.;
2161–2163. (e) Nolan, E. M.; Lippard, S. J. J. Am. Chem. Soc. 2003, 125, Rudkevich, D. M. Chem. Commun. 2007, 1238–1239. (f) Hecht, S.;
14270–14271. (f) Descalzo, A.; Martinez-Manez, R.; Radeglia, R.; Rurack, Vladimirov, N.; Fre’chet, J. M. J. J. Am. Chem. Soc. 2001, 123, 18–25. (g)
K.; Soto, J. J. Am. Chem. Soc. 2003, 125, 3418–3419. (g) Hennrich, G.; Godwin, H. A.; Berg, J. M. J. Am. Chem. Soc. 1996, 118, 6514–6515. (h)
Walther, W.; Resch-Genger, U.; Sonnenschein, H. Inorg. Chem. 2001, 40, Albers, A. E.; Okreglak, V. S.; Chang, C. J. J. Am. Chem. Soc. 2006, 128,
641–644. (h) Hennrich, G.; Sonnenschein, H.; Resch-Genger, U. J. Am. 9640–9641. (i) Lee, M. H.; Kim, H. J.; Yoon, S.; Park, N.; Kim, J. S. Org.
Chem. Soc. 1999, 121, 5073–5074. (i) Dujols, V.; Ford, F.; Czarnik, A. W. Lett. 2008, 10, 213–216. (j) Bolletta, F.; Costa, I.; Fabbrizzi, L.; Licchelli,
J. Am. Chem. Soc. 1997, 119, 7386–7387. (j) Kwon, J. Y.; Jang, Y. J.; M.; Montalti, M.; Pallavicini, P.; Prodi, L.; Zaccheroni, N. J. Chem. Soc.,
Lee, Y. J.; Kim, K. M.; Seo, M. S.; Nam, W.; Yoon, J. J. Am. Chem. Soc. Dalton Trans. 1999, 1381–1385. (k) Coskun, A.; Akkaya, A. U. J. Am.
2005, 127, 10107–10111. (k) Suresh, M.; Mishra, S. K.; Mishra, S.; Das, Chem. Soc. 2005, 127, 10464–10465. (l) Zhu, Z.; Yu, M.; Yang, H.; Huang,
A. Chem. Commun. 2009, 2496–2498. K.; Li, F.; Yi, T.; Huang, C. Chem. Commun. 2008, 3387–3389.

10.1021/ol900810q CCC: $40.75  2009 American Chemical Society

Published on Web 06/09/2009
nonradiative energy transfer process in which the excitation UV-vis spectra recorded for L1 (CH3CN-H2O, 1:1,
energy of the donor is transferred to the nearby acceptor via v/v) shows an absorption maxima at 309 and 232 nm with
long-range dipole-dipole interaction and/or short-range two distinct shoulders at 340 and 260 nm. Absorption
multipolar interaction. Donor and acceptor units with ap- bands/shoulders were predominantly due to intraligand
preciable spectral overlap between the donor emission and π-π* charge transfer transitions. Binding ability of L1
acceptor absorption spectra constitute an appropriate RET toward various metal ions (Li+, Na+, K+, Cs+, Sr2+, Mg2+,
pair. For commonly used luminescence-based probe mol- Ca2+, Cr3+, Fe2+, Co2+, Ni2+, Cu2+, Zn2+, Cd2+, Hg2+, and
ecules, quantitative measurements are possible if a linear Pb2+) was checked. Spectral studies, using a solution of
relationship exists between luminescence intensity and the L1 (CH3CN-H2O, 1:1, v/v), revealed a significant change
fluorophore concentration. However, this may not be known in electronic spectral pattern only when Cu2+ or Hg2+ was
with sufficient accuracy in the case of aggregation or added: a new absorption band around 530 nm was
photobleaching. For biological applications RET-based probe developed with detectable change in solution color (Figure
molecules are generally more useful than single dye-based 1a and Supporting Information). While an enhancement
probes, as the RET-based process is independent of the
concentration of a single fluorescent dye and one can quantify
the analyte concentration by using the ratio of intensities of
the well resolved fluorescence peaks with reasonable intensi-
ties at two different wavelengths for analyte-free and analyte-
bound probe.6 However, despite many advantages, examples
of RET-based off-on fluorogenic sensors for Hg2+ in
aqueous solution is not common in the literature.7
Herein, we report a new chemosensor (L2), which was
obtained following the synthetic methodology shown in
Scheme 1. A newly synthesized thiophene derivative of
Figure 1. (a) Absorption spectra of L1 (20 µM) in H2O-CH3CN
(1:1, v/v) with varying [Hg2+] (0-600 µM). Inset: Job’s plot that
indicates the 1:1 stoichiometry. (b) Fluorescence spectra of L1 (10
Scheme 1. Synthetic Route of L1 and L2 µM) in the presence of varying [Hg2+] (λext ) 500 nm).

in fluorescence intensity of L1 (Φ ) 0.006; λext ) 500

nm) at ∼557 nm was observed on addition of either of
these two metal ions, this enhancement was more signifi-
cant (105-fold) for Hg2+ (Φ ) 0.62) as compared to that
for Cu2+ (12-fold, Φ ) 0.07) (Figure 1b and Supporting
Information). Binding constants for two respective metal
ions with L1 were evaluated and found to be (KHg2+ )
(1.45 ( 0.1) × 104 and KCu2+ ) (3.1 ( 0.2) × 103 M-1
at 25 °C), and in both cases 1:1 complex formation was
evident. A similar formation constant value for these two
respective metal ions was reported earlier for related
rhodamine derivatives.9 Appearance of a new absorption
spectral band at around 530 nm and an enhancement in
rhodamine (L1) was used as an intermediate. Both L1 and fluorescence intensity at around 555 nm on binding to
L2 were characterized using various analytical, spectroscopic Hg2+/Cu2+ suggest opening of the spirolactam ring in L1
techniques (see Supporting Information), which agreed well on metal ion coordination.10 Spectral studies using a
with the proposed structures for L1 and L2. Molecular freshly prepared solution of dansyl phenyl amide (DNPA),
structure for L1 was also confirmed by single crystal X-ray
analysis.8 Ability of the receptor L2 to bind specifically to (8) Crystal data for the compound: CCDC No. 695747; molecular
Hg2+ in mixed aqueous media with associated fluorescence formula C31H30N4O2S, M ) 522.65, crystal size 0.34 × 0.25 × 0.20 mm3,
triclinic, space group P-1 with a ) 9.657(5) Å, b ) 11.328(5) Å, c
on response has given this molecule an edge over some of )12.770(6) Å, R ) 93.435(8)°,  ) 109.501(8)°, γ ) 92.979(9)°, V )
the earlier reported RET-based receptor molecules.5k,l,7a,c 1310.7(11) Å3, Z ) 2, Dcalcd ) 1.324 g/cm3, T ) 293(2) K, F(000) ) 590,
absorption coefficient ) 0.160 mm-1, λ ) 0.71073 Å, 10230 reflections
were collected, 5091 observed reflections, R(int) ) 0.0349, no. of parameters
(6) (a) Royzen, M.; Dai, Z.; Canary, J. W. J. Am. Chem. Soc. 2005, ) 347, R1 ) 0.0930, wR2 ) 0.2000 with (I g 2σ(I)), goodness of fit on
127, 1612–1613. (b) Ajayaghosh, A.; Carol, P.; Sreejith, S. J. Am. Chem. F2 ) 1.174. Largest difference peak and hole: 0.337 and -0.473 eÅ-3,
Soc. 2005, 127, 14962–14963. (c) Kiyose, K.; Kojima, H.; Urano, Y.; respectively.
Nagano, T. J. Am. Chem. Soc. 2006, 128, 6548–6549. (d) Banthia, S.; (9) (a) Suresh, M.; Shrivastav, A.; Mishra, S.; Suresh, E.; Das, A. Org.
Samanta, A. J. Phys. Chem B 2006, 110, 6437–6440. (e) Lakowicz, J. R. Lett. 2008, 10, 3013–3016. (b) Rurack, K.; Kollmannsberger, M.; Resch-
Principles of Fluorescence Spectroscopy, 3rd ed.; Springer: New York, Genger, U.; Duab, J. J. Am. Chem. Soc. 2000, 122, 968–969.
2008. (10) (a) Zheng, H.; Qian, Z.-H.; Xu, L.; Yuan, F.-F.; Lan, L.-D.; Xu,
(7) (a) White, B. R.; Liljestrand, H. M.; Holcombe, J. A. Analyst 2008, J.-G. Org. Lett. 2006, 8, 859–861. (b) Yang, H.; Zhou, Z.; Huang, K.; Yu,
133, 65–70. (b) Zhang, X.; Xiao, Y.; Qian, X. Angew. Chem., Int. Ed. 2008, M.; Li, F.; Yi, T.; Huang, C. Org. Lett. 2007, 9, 4729–4732. (c) Shai, W.;
47, 1–6. (c) Coskun, A.; Akkaya, A. U. J. Am. Chem. Soc. 2006, 128, Ma, H. Chem. Commun 2008, 1856–1858. (d) Wu, D.; Huang, W.; Duan,
14474–14475. C.; Lin, Z.; Meng, Q. Inorg. Chem. 2007, 46, 1538–1540.

Org. Lett., Vol. 11, No. 13, 2009 2741

synthesized following a known procedure,11 revealed that the xanthene fragment (minor fraction that may exist in
its emission spectra had a significant overlap (inset, Figure equilibrium along with the spirolactam form),10a and a longer
2a) with the absorption spectra of the Hg2+-bound and major component was assigned to the dansyl unit present
in L2. Thus, comparison of the decay time constants for the
DNPA and L2 revealed the predominant decay of the excited
singlet state based on the dansyl moiety. Further, absence
of any xanthene-center based emission at ∼560 nm and
absorption band at ∼530 nm signified its spirolactam
structure. Preliminary studies reveal that when a solution of
Hg2+ or Cu2+ (CH3CN-H2O, 1:1, v/v) was added to a
solution of L2 (CH3CN-H2O, 1:1, v/v), a distinct color
change could be noticed by the naked eye (inset, Figure 2b).
Electronic spectral studies reveal that a new spectral band
appeared at around 530 nm on addition of the respective
metal ion (10 molar equiv) solution to that of L2 (see
Figure 2. Absorption spectra of L2 (1.0 × 10-5 M) in H2O-CH3CN Supporting Information), an observation similar to that for
(1:1, v/v) in the presence of varying (a) [Hg2+] (0-7 × 10-4 M) Hg2+·L1. The absorption band for L2 in the presence of either
(inset: spectral overlap between emission spectrum of the donor Hg2+ or Cu2+ was dominated by the new charge transfer
and absorption spectrum of the acceptor) and (b) [Cu2+] (0-6.14 band. Intensity of the new absorption at 530 nm was more
× 10-4 M). Inset: visible color changes of L2. intense for Hg2+ as compared to that for Cu2+. The
association constant for respective complexes Hg2+·L2 and
Cu2+·L2 was evaluated from systematic spectrophotometric
xanthene moiety of L1 (Hg2+·L1), and this led us to explore titrations (Figure 2) using the Benasi-Hildebrand equation
the possibility of using dansyl functionality in combination and was found to be (3.9 ( 0.1) × 104 M-1 (KHg2+·L2) and
with L1 for ratiometric sensing of Hg2+. (6.8 ( 0.2) × 103 M-1 (KCu2+·L2) at 25 °C. Stoichiometry for
Electronic spectra recorded for L2 (CH3CN-H2O, 1:1, v/v) the complexes formed for both metal ions were evaluated
was basically dominated by the absorption bands that belong on the basis of the Job’s plot and was found to be 1:1.
to the L1 moiety, while a shoulder at 340 nm became more Reversible binding of L2 with Hg2+ and Cu2+ was also
prominent presumably due to contribution from the dansyl examined. Addition of 10 equiv of EDTA2- to a mixture of
moiety in L2. On excitation of the CH3CN-H2O solution of M2+·L2 (5.0 × 10-5 M, M2+ is Hg2+/Cu2+) results in
L2 at 340 nm, a relatively weak and very broad emission bleaching of the absorption band at 530 nm, which signifies
band (λems ) 500 nm) was observed (Figure 3), which could the regeneration of the spirolactam structure (see Supporting
Information).
Emission spectra recorded for L2 (λext ) 340 nm, dansyl
moiety absorbs predominantly at this wavelength) in the
presence of Hg2+ and Cu2+ showed respective emission bands
at 555 and 545 nm. Further, L2 was found to be almost
nonluminescent when excited at 500 nm (where the xanthene
moiety absorbs predominantly). However, a significant
increase in emission intensity at 555 nm was observed when
a similar experiment was repeated in the presence of Hg2+/
Cu2+ (5 molar equiv). Quantum yield for L2, Hg2+·L2, and
Cu2+·L2 was found to be 0.009, 0.35, and 0.16, respectively,
for λext ) 500 nm. More interestingly, appreciable enhance-
Figure 3. Fluorescence spectra of L2 (1.0 × 10-5 M) in ment in emission intensity at 555 nm was also registered
H2O-CH3CN (1:1, v/v) with varying (a) [Hg2+] (0-7.0 × 10-4
M) and (b) [Cu2+] (0–6.14 × 10-4 M) using λext of 340 nm. following excitation at 340 nm, a wavelength where dansyl
units absorb predominantly. This tends to demonstrate the
RET process in the presence of only the Hg2+ ion (Figure
3). Appearance of the absorption band at 530 nm for Hg2+·L2
be assigned to the dansyl unit based emission. This was /Cu2+·L2 and the emission band at ∼555 nm suggested the
further confirmed by time correlated single photon counting opening of the spirolactam ring and generation of the
(TCSPC) studies using a 340 nm nano LED as an excitation delocalized xanthene moiety.10 Figure 2a (inset) clearly
source. Emission decay traces (λmon ) 539 nm) for L2 could shows that absorption spectra of the related xanthene moiety
be best fitted with a biexponential function with time (Hg2+·L1) and emission spectra of the dansyl unit has an
constants (τ1 ) 3.9 ( 0.2 ns (14.3%) and (τ2 ) 17.2 ( 0.2 appreciable overlap, which makes nonradiative transfer of
ns) (85.7%), (2 ) 1.16), whereas for the DNPA this was excitation energy between donor dansyl to acceptor xanthene
found to be a single exponential with τ being 13.6 ns (2 ) moiety feasible. Systematic fluorescence spectral titration of
1.71). The shorter and minor component was assigned to L2 (CH3CN-H2O, 1:1, v/v) on excitation at 340 nm in the
presence of varying [Hg2+] revealed a gradual decrease in
(11) Cardona, C. M.; Alvarez, J.; Kaifer, A. E.; McCarley, T. D.; Pandey,
S.; Baker, G. A.; Bonzagni, N. J.; Bright, F. V. J. Am. Chem. Soc. 2000, the dansyl unit based emission at 483 nm along with a
122, 6139–6144. concomitant increase in the new emission band at 555 nm
2742 Org. Lett., Vol. 11, No. 13, 2009
with a well-defined isoemissive point at 539 nm (Figure 3). (Pseudomonas putida) were studied by confocal laser
These binding constant values were also evaluated from the microscopy (Olympus 1 × 81 with FV1000 confocal laser
luminescence titration studies using λext of 340 and 555 nm microscope) before (control) and after exposing these
as the monitoring wavelengths (λmon). Respective binding microbes to an aqueous solution of Hg2+. Then, subsequently
constants for the two metal ions thus obtained agreed well the control and cells exposed to Hg2+ solution were further
(KHg2+·L2 ) (5.0 ( 0.2) × 104 M-1 and KCu2+·L2 ) (7.9 ( exposed to the aqueous solution of the reagent L2. Pseudomo-
0.1) × 103 M-1 at 25 °C) with those obtained from absorption nas putida is known to adsorb Hg2+ ion14,15 and appeared
spectral studies. Further, spectral studies revealed that a lower either colorless or nonfluorescent (Figure 4). However, cells
detection limit for Hg2+ was 0.1 ppm (for signal-to-noise
ratio of 3:1), and thus L2 could be used as a sensitive and
selective chemosensor for Hg2+ using the RET process.
The ratio of acceptor-to-donor emission (λext ) 340 nm)
intensity (λ555/λ483), in the absence and presence of varying
[Hg2+], varied from 0.83 to 22.46, and this 27-fold enhance-
ment was attributed to the RET process. Optical spectral
studies revealed that L2 could also bind Cu2+; however, no
significant enhancement in emission intensity at 545 nm was
observed. The quenching of the fluorescence of the xanthene
moiety in L2 by Cu2+ could be explained on the basis of the
well-known paramagnetic effect of the d9 Cu(II) system.12
To probe the singlet-singlet resonance energy transfer
from the dansyl unit to the xanthene moiety in Hg2+·L2, time-
resolved fluorescence decay studies were also undertaken Figure 4. Confocal laser microscope images of (a) blank cells of
using the TCSPC technique. A solution of L2 (CH3CN-H2O, Pseudomonas putida; (b) and (c) confocal image (λmoni ) 560 nm)
1:1, v/v) in the absence and presence of 1 molar equiv of and bright field image for the cells, exposed to Hg2+ and (10 µM)
Hg2+ was used for following excitation with a 340 nm nano and then to L2 (20 µM), respectively; (d) confocal image of the
LED source. Fluorescence decay was monitored at 539 nm, cells exposed to Hg2+ and L2 and the fluorescence intensity
monitored at 490 nm with λexc ) 405 nm.
the iso-emissive point for two fluorophores (dansyl unit and
Hg2+·L1). The emission decay curve for Hg2+·L2 could be
best fitted to a biexponential decay function with τ1 ) 4.0
( 0.4 ns (89%) and τ2 ) 15.0 ( 0.5 ns (11%) (2 ) 1.14). exposed to Hg2+ (10 µM) and then stained with L2 (20 µM,
A larger component with a lifetime of 4 ns was attributed to in 1:1 v/v water-ethanol) at 25 °C appeared as a pink-red
the xanthene unit in Hg2+·L2, while the smaller and slower color when observed under the optical microscope and red
component was assigned for the dansyl unit. Thus, TCSPC fluorescent when viewed through a confocal microscope
studies reveal that in the presence of Hg2+, the situation is using 405 nm excitation source (Figure 4).
completely different and deactivation of the excited state Appearance of the red fluorescence confirmed the forma-
occurs predominantly through the excited state that belongs tion of Hg2+·L2 with Hg2+ that was adsorbed within the
to the xanthene moiety in Hg2+·L2 rather than the dansyl bacterial cells. Thus, the reagent L2 could be used as an
unit in L2, although the excitation wavelength of 340 nm, optical and fluorescent staining agent for detection of the
specific for the dansyl unit, was used as the excitation source. Hg2+ uptake in bacteria.
Thus, results of the time-resolved emission studies also Thus we could demonstrate that L2 could be used as a
corroborate the RET process. staining agent for detection of Hg2+ uptake in bacteria and
The singlet-singlet excitation energy-transfer efficiency can be used in combination with dansyl moiety as ratiometric
(ΦET) and rate constant for the energy-transfer process (kET) sensitizer for the Hg2+ ion in aqueous solution. As desired
between donor and acceptor were evaluated from steady- for the development of chemosensors that show ratiometric
state and time-resolved fluorescence data (see Supporting RET response to a specific and biologically important metal
Information, eqs 1 and 2).5a,13 The ΦET for the present study ion, L2 has a large crosssection for energy absorption,
was found to be 83%, and while kET was found to be 2.84 × solubility in water, and high quantum yield.
108 s-1. The Förster critical distance (R0) was calculated (see
Supporting Information) using eq 1 and was found to be 74.3 Acknowledgment. DST and CSIR have supported this
Å. R0 is the distance at which 50% energy transfer takes work. M.S., A.K.M., S.K.M., and A.S. acknowledge CSIR
place between donor and acceptor. for research fellowship.
Supporting Information Available: Synthetic details,
R0 ) 9.79 × 103[(J)Q(n-4)(κ2)]1/6 (1) characterization of L1 and L2, and selected spectroscopic data
for L1 and L2. This material is available free of charge via
the Internet at [Link]
To explore the possibility of using L2 as a probable reagent
for bioimaging application, Gram-negative bacterial cells OL900810Q

(12) McClure, D. S. J. Phys. Chem. 1952, 20, 682–686. (14) Beveridge, T. J. Biotechnol. Bioeng. 1986, 16, 127–139.
(13) Adronov, A.; Gilat, S. L.; Frechet, J. M. J.; Ohta, K.; Neuwahl, (15) Ledin, M.; Pedersen, K.; Allard, B. Water, Air, Soil Pollut. 1997,
F. V. R.; Fleming, G. R. J. Am. Chem. Soc. 2000, 122, 1175–1185. 93, 367–381.

Org. Lett., Vol. 11, No. 13, 2009 2743