Effects of various bases on acid-catalyzed amination of 2-chloro-5-ethylpyrimidine: Synthesis of PPARpan agonist GW693085

Article abstract of DOI:10.1021/jo100562s

Figure presented A unique buffering effect of various bases, i-Pr ₂NEt and CaCO₃ in particular, was observed for the acid-catalyzed chloro displacement of 2-chloro-5-ethylpyrimidine with a 2-methyl-2-phenylpropanamine. The use of the

Full text of DOI:10.1021/jo100562s

pubs.acs.org/joc
Effects of Various Bases on Acid-Catalyzed
Amination of 2-Chloro-5-ethylpyrimidine:
Synthesis of PPARpan Agonist GW693085
Bobby N. Glover, Lynda A. Jones, Byron, S. Johnson,
Alan Millar, Martin H. Osterhout, and Shiping Xie*
Chemical Development, GlaxoSmithKline,
Research Triangle Park, North Carolina 27709
Development of 1 as a drug candidate for type 2 diabetes
required a supply of multikilogram quantities for preclinical
and clinical investigation. The target compound is a 5-ethyl-
2-pyrimidinamine with neopentyl and benzyl substituents on
the exocyclic nitrogen. Herein, we report process research
leading to a highly efficient and robust synthesis, incorpor-
ating excellent selectivity in sequential N- and O-alkylations.
Screening of various types of bases in the amination of
2-chloro-5-ethylpyrimidine revealed the unique buffering
effect of calcium carbonate and i-Pr2NEt (Hunig’s base) in
the acid-catalyzed amination.
shiping.x.xie@gsk.com
Received March 24, 2010
SCHEME 1
A unique buffering effect of various bases, i-Pr2NEt and
CaCO3 in particular, was observed for the acid-catalyzed
chloro displacement of 2-chloro-5-ethylpyrimidine with a
2-methyl-2-phenylpropanamine. The use of the carefully
chosen bases was essential for the progression of the
chloro displacement as well as the stability of the product
in the presence of HCl formed. Research work leading to
an efficient synthesis of PPARpan agonist GW693085
is described, featuring highly selective sequential N- and
O-alkylations.
Initially, the tertiary amino skeleton of 1 was constructed
as shown in Scheme 1. 2-Methyl-2-(4-(methoxyphenyl)-1-pro-
panamine (2) underwent reductive amination with 4-(trifluoro-
methoxy)benzaldehyde to provide the secondary amine (3a).3
A high-temperature coupling between free base 3a and
2-chloro-5-ethylpyrimidine in a sealed tube at 210 °C in
the presence of i-Pr2NEt afforded 62% yield of tertiary
amine 4. Other attempts including microwave irradation4
and some known palladium-catalyzed amination condi-
tions (Pd2(dba)3, BINAP, NaO-t-Bu, toluene)5 gave no
desired product at all. The lack of reactivity was attributed
to significant steric hindrance in 3a due to the N-substi-
tuents including the neopentyl moiety.
Peroxisome proliferator-activated receptors (PPARs) are
members of the steroid/thyroid/retinoid nuclear receptor super-
family. The three subtypes (R, γ, and δ) constitute a subfamily,
each having a distinct distribution in the body. PPARs have
been shown to be important in energy balance through their
participation in regulating the proteins involved in carbo-
hydrate and lipoprotein metabolism.1 Recently, selective PPAR
agonists were dosed in combination to show that there could be
synergy in lowering glucose and triglycerides.2 GW693085 (1) is
a PPARpan agonist, combining the activity of the three known
subtypes of the PPAR agonists into a single molecule.2e
Due to safety and scalability concerns,6 our efforts shifted
to identifying alternative routes. Amine 2 and its highly
(1) For reviews on modulators of PPARs and their mechanism and
actions, see: (a) Cho, M. C.; Lee, K.; Paik, S. G.; Yoon, D. Y. PPAR Res.
2008, 679137. (b) Chang, F.; Jaber, L. A.; Berlie, H. D.; O’Connell, M. B.
Ann. Pharmacother. 2007, 41, 973. (c) Rubenstrunk, A.; Hanf, R.; Hum,
D. W.; Fruchart, J. C.; Staels, B. Biochim. Biophys. Acta 2007, 1065. (d)
Sternbach, D. D. Ann. Rev. Med. Chem. 2003, 38, 71.
(2) (a) Feldman, P. L.; Lamber, M. H.; Henke, B. R. Curr. Top. Med.
Chem. 2008, 8, 728. (b) Franklin, M. C.; Lewis, M. C.; Wilson, J. G.; Brown,
J. G.; Strole, C. A.; Oliver, W. Jr.; Winegar, D. A. Diabetes 2002, 51 (Suppl. 2),
556-P. (c) Lewis, M. C.; Winegar, D. A.; Bodkin, N. L.; Hansen, B. C.; Oliver, W.,
Jr. Diabetes 2002, 51 (Suppl. 2), 566-P. (d) Lewis, M. C.; Wilson, J. G.; Franklin,
M. C.; Brown, J. G.; Strole, C. A.; Oliver, W., Jr.; Winegar, D. A. Diabetes 2002,
51 (Suppl. 2), 567-P. (e) Rodolfo, C.; Henke, B. R.; Lambert, M. H., III; Liu,
G. K.; Smith, J. S. PCT Int. Appl., WO/2003074495, 2003.
(3) Primary amine 2 was prepared from the nitrile precursor 2-(4-
methoxyphenyl)-2-methylpropionitrile by reduction with LiAlH4. For a
preparation of the nitrile from nucleophilic aromatic substitution (SNAr)
of a corresponding fluoroarene and a secondary nitrile, see: Caron, S.;
Vazquez, E; Wojcik, J. M. J. Am. Chem. Soc. 2000, 122, 712.
(4) For a report on nucleophilic substitution reactions of 2-halopyrimidyl
and pyrazyl halides under focused microwave irradiation, see: Cherng, Y.-J.
Tetrahedron 2002, 581, 887.
(5) (a) Wolfe, J. P.; Wagaw, S.; Buchwald, L. S. J. Am. Chem. Soc. 1996,
118, 7215. (b) Paul, F.; Patt, J.; Hartwig, J. F. J. Am. Chem. Soc. 1994, 116,
5969.
(6) Differential scanning calorimetry (DSC) indicated that 2-chloro-5-
ethylpyrimidine decomposes at 210-260 °C and releases 1.1 kJ/g of energy.
3904 J. Org. Chem. 2010, 75, 3904–3907
Published on Web 04/29/2010
DOI: 10.1021/jo100562s
r
2010 American Chemical Society

Glover et al.
JOCNote
FIGURE 1. Effects of various bases on the reaction of 5 and 2-chloro-5-ethylpyrimidine.
SCHEME 2
SCHEME 3
crystalline hydrochloride salt 5 were found to be much more
reactive than the secondary amine (3a) in reaction with
2-chloro-5-ethylpyrimidine (Scheme 2). The 2-aminopyrimi-
dine product (6) was highly crystalline, yet several signi-
ficant observations were made. The reaction was found to
be quite sensitive to the conditions used. Although it is
known that amination of 2-chloropyrimidine is acid cata-
lyzed,7 we found that the reaction of 5 with 2-chloro-5-ethyl-
pyrimidine was accompanied by significant acid-induced
decomposition including cleavage of the methyl ether and
breakdown of the pyrimidine ring unless the reaction was
mediated by a base.
A variety of common bases were screened for the reaction
with results shown in Figure 1. The conversion of 5 to 6 was
measured by the area under curve (AUC) of product 6 on
reversed-phase HPLC. There was only about 10% of pro-
duct without an added base. Significant amount of decom-
position to more polar impurities related to demethylation
from 5 and 6 was observed under the high temperature
needed for the reaction. On the other hand, strong bases
such as NaOH, Cs2CO3, K2CO3, and pyridine inhibited the
reaction with AUC of the product below 10%. Powdered
CaCO3 (particle size 8.5-10 μm) and i-Pr2NEt were by far
the best basic mediators in the screening with 76% and 87%
product, respectively (Figure 1).
The role of the base as an acid scavenger became prominent only
as more hydrogen chloride was generated from the chloro
displacement. This reduced the levels of byproducts induced
by decomposition in the presence of a high concentration of acid.
NaOAc and N,N-dimethylaniline, with 58% and 26% product,
respectively, were better than strong bases for the reaction but
did not provide enough product for isolation. Unlike the reac-
tion with the heterogeneous and mildly basic CaCO3, reaction in
the presence of heterogeneous but slightly acidic CaCl2 produced
less than 20% product. The reactions with CaCO3 and i-Pr2NEt
base were subsequently optimized to provide 70-80% isolated
yield of crystalline product 6 in scaleup. No optimization on the
particle size of CaCO3 was carried out.
Our initial approach to convert the 2-aminopyrimidine (6)
to the drug candidate (1) is shown in Scheme 3. Alkylation
with (4-trifluoromethoxy)benzyl bromide gave tertiary
amine 4. It was essential to preform the amide anion from
6 with a strong base for the alkylation to occur at the targeted
exocylic nitrogen. Although the amide anion could be
formed with many strong bases, NaH gave the cleanest alkyla-
tion. Use of a variety of weaker bases including tertiary amines
(Hunig’s base, Et3N), pyridine, 1,8-diazabicyclo[5.4.0]undec-
7-ene (DBU), K2CO3, and CaCO3 in various solvents led
We postulate that the heterogeneous nature of the base, in
the case with CaCO3, and the low miscibility of the base with
ethylene glycol solvent, in the case of Hunig’s base, allowed
the acid-catalyzed amination of the chloropyrimidine to
proceed while still being able to function as a base to
neutralize the excess acid generated from the reaction itself.
(7) Gunzenhauser, S.; Balli, H. Helv. Chim. Acta 1988, 71, 33.
J. Org. Chem. Vol. 75, No. 11, 2010 3905

Glover et al.
JOCNote
SCHEME 4
from methanol as a highly crystalline white solid. Sur-
prisingly, this reaction was completely O-selective in the
presence of the 2-aminopyrimidine moiety. In contrast, when
phenol 8 was used under Bargellini reaction conditions with
chloretone,10 the reaction produced small quantities of
N-alkylation as well as N,O-bisalkylation byproducts. This
should not be surprising in light of application of Bargellini
reactions in synthesis of piperazinones by Lai11a,b and
Rychnovsky.11c They have demonstrated that even a tertiary
amino group of R-amino alcohols would preferably react on
the nitrogen. Mechanistically, it was not clear if the 2-bro-
moisobutyric reagent worked through a R-lactone inter-
mediate. Online monitoring with a React-IR did not detect
any absorbance that could be attributable to a R-lactone
intermediate. Anchimeric assistance or neighboring group
participation for the bromo displacement appears to be the
most probable explanation for the fast and clean reaction.
In another example of N- vs O-selectivity, the 2-alkylami-
nopyrimidine (9) was selectively N-alkylated with (4-tri-
fluoromethoxy)benzyl bromide in the presence of NaH to
give an 86% yield of crystalline GW693085 (1). There was no
alkylation on the carboxylic oxygen. In a testimony to the
robustness of this alkylation, use of 2 equiv of the alkylating
reagent and 4 equiv of the base still produced no benzyl ester.
Any excess amount of (4-trifluoromethoxy)benzyl bromide,
a strong alkylator and therefore a potential human carcino-
gen, was quenched by addition of sodium methoxide to the
reaction mixture.
to recovered 6 with mild heating. When heated to over
100 °C in the presence of those weak bases, the reaction
gave rise to only a trace of desired tertiary amine product
4, along with a byproduct (30-60% by HPLC) which was
identified as structure i on the basis of analysis with
HPLC-MS fragementation.8 Although cleavage of the
methyl ether from 4 appeared straightforward with typi-
cal conditions such as BBr3 in CH2Cl2,9 the reaction was
complicated due to competitive loss of the N-benzyl
moiety from both 4 and 7, producing side products 6
and 8. The penultimate phenolic intermediate (7) could
only be isolated by silica gel chromatography. The iso-
butyryl moiety was then installed through a Bargellini
reaction with chloretone in the presence of NaOH.10 However,
as much as 30% of mesityl oxide byproduct derived from the
base-catalyzed acetone condensation was formed during the
reaction. An alternative synthetic sequence was desired with
a penultimate crystalline intermediate and without the side
products from debenzylation.
Further process research led to a new and much more
desirable route starting with 6 shown in Scheme 4. The new
synthesis benefits from the elimination of the competi-
tive debenzylation, usage of an alternative reagent to introduce
the isobutyryl headgroup that avoids the formation of the
mesityl oxide, and assurance of a crystalline penultimate inter-
mediate prior to the final formation of the active pharmaceu-
tical ingredient (API).
In conclusion, an efficient and robust synthesis of PPARpan
agonist GW693085 has been developed. Key features include
excellent nitrogen/oxygen chemoselectivity in the alkylation of
two major intermediates, crystalline intermediates, and the
brevity of the synthesis. The use of a heterogeneous base such
as CaCO3 in acid-catalyzed alkylamination of 2-chloro-5-
ethylpyrimidine was noteworthy. A strong base or a base
miscible with the solvent used in the reaction would prevent
the chloro displacement. On the other hand, there was signifi-
cant decomposition of the 2-aminopyridmidine product in the
absence of a base as a buffer.
Experimental Section
2-Methyl-2-(4-methyloxyphenyl)-1-propanamine Hydrochloride
(5). To a solution of 1.00 kg (6.79 mol) of (4-methoxyphenyl)-
acetonitrile in 4.6 L of THF in a 30 L reactor was added 13.9 L
(27.8 mol) of lithium diisopropylamide (2 M in heptane/THF/
ethylbenzene) with cooling. The addition rate was such that the
internal temperature was maintained below 25 °C. After being
stirred at 25 °C for 1 h, the mixture was cooled to 5 °C and treated
with 1.6 L (16.9 mol) of dimethyl sulfate as a solution in 5.1 L of
THF at below 20 °C. The reaction was cooled to 5 °C and
quenched with a mixture of 2.4 L (28.2 mol) of concentrated
HCl and 1.8 L of water. The layers were separated, and the organic
layer was successively washed with 7.4 L of water and 7.4 L of
brine. The organic layer was concentrated under vacuum to 3 L.
After being heated to 45 °C, the concentrated nitrile intermediate
was treated with 6.6 L (6.60 mol) of 1 M LiAlH4 in THF with
temperature maintained below 50 °C. After being stirred for 2 h,
the reaction was quenched by a mixture of 0.2 L of water and
0.6 L of THF followed by 0.2 L of 15% NaOH and 0.6 L of
water. The mixture was stirred for 1 h and filtered. The filtering
Demethylation from 6 by BBr3 afforded phenolic inter-
mediate 8 which was used for the next step with no need for
purification. Incorporation of the isobutyryl moiety was
accomplished in 92% yield by reaction of 2-bromo-2-methyl-
propionic acid with 8 in a basified methyl ethyl ketone
(MEK) solution. Penultimate intermediate 9 was isolated
(8) The formation of this byproduct could be explained by the initial
alkylation by the benzyl bromide through the endocyclic pyrimidine nitro-
gen. The resultant iminium would undergo a sigmatropic rearrangement to
give i. See the Supporting Information for LC-MS analysis and the
proposed pathway for the formation of i.
(9) McOmie, J. F. W.; West, D. E. Organic Syntheses; Wiley: New York,
1973; Collect. Vol. V, p 412
(10) (a) Bargellini, G. Gazz. Chim. Ital. 1906, 36, 329. (b) Weizmann, C.;
Sulzbacher, M.; Bergmann, E. J. Am. Chem. Soc. 1948, 70, 1153.
(11) (a) Lai, J. T. J. Org. Chem. 1980, 45, 754. (b) Lai, J. T. Synthesis 1984,
122. (c) Rychnovsky, S. D.; Beauchamp, T.; Vaidyanathan, R.; Kwan, T.
J. Org. Chem. 1998, 63, 6363.
3906 J. Org. Chem. Vol. 75, No. 11, 2010

Glover et al.
JOCNote
cake was washed with 3 L of THF. The combined filtrates
containing free base 3 were concentrated under vacuum to
about 2 L and diluted with 1.9 L of ethanol and 20 L of toluene.
The brown solution was then treated with 0.60 L (7.20 mol) of
concentrated HCl. After removal of about 1 L of the solvents,
the mixture was cooled to 5 °C over 2 h and stirred for 1 h. The
mixture was filtered, washed with 2.0 L of toluene, and dried at
65 °C to afford 1.10 kg (94%) of 5 as an off-white crystalline
solid: 1H NMR (300 MHz, CD3OD) δ 1.43 (s, 6H), 3.15 (s, 2H),
3.81 (s, 3H), 6.97 (d, J = 8.6 Hz, 2H), 7.38 (d, J = 8.6 Hz, 2H).
temperature, quenched with 200 mL of water, and stirred for
15 min. The layers were separated, and the aqueous layer was
extracted with 150 mL of MEK. The combined organic layers were
acidified (pH 2.5 for aqueous layer) with 27 mL (405 mmol) of 86%
H3PO4 as a solution in 130 mL of water. The organic layer was
washed with 150 mL of brine and concentrated under vacuum to
about 125 mL. The resultant white slurry was treated with 40 mL of
MeOH, stirred at -10 °C for 4 h, and filtered. The filtering cake was
washed with 30 mL of MeOH and dried at 60 °C to provide 46.3 g
(79%) of the first crop of 9 as a white crystalline solid. The filtrate
was treated with about 50 mL (50 mmol) of 1 N NaOH to raise the
pH to 6. After addition of 100 mL of EtOAc, the organic layer was
washed with brine, concentrated to 60 mL, and stirred with ice
cooling for 5 h. The resultant slurry was filtered, washed with 5 mL
of MeOH, and dried at 60 °C to afford 7.88 g (13%) of the second
crop of 9 as a white crystalline solid. The overall yield was 54.2 g
(92%) of 9: mp 142.1 °C; 1H NMR (400 MHz, CDCl3) δ 1.13 (t,
J = 7.0 Hz, 3H), 1.37 (s, 6H), 1.52 (s, 3H), 2.41 (q, J = 7.0 Hz, 2H),
3.62 (d, J = 8.0 Hz, 2H), 6.83 (d, J = 8.0 Hz, 2H), 6.97 (br. s, 1H),
7.26 (d, J = 8.0 Hz, 2H), 8.00 (s, 2H); 13C NMR (100 Hz, CDCl3)
δ178.9, 160.0, 153.8, 140.2, 127.1, 124.3, 119.3, 79.3, 53.1, 38.7, 27.2,
25.6, 22.7, 15.4. Anal. Calcd for C20H27N3O3: C, 67.20; H, 7.61; N,
11.76. Found: C 67.00; H, 7.64; N, 11.70.
Anal. Calcd for C11H17NO HCl: C, 61.25; H, 8.41; N, 6.49; Cl,
16.43. Found: C, 61.20; H, 8.44; N, 6.39; Cl, 16.66.
3
5-Ethyl-N-[2-methyl-2-[4-(methyloxy)phenyl]propyl]-2-pyrimidin-
amine (6). To a mixture of 5.64 g (26.1 mmol) of HCl salt 5 and
3.3 mL (27.4 mmol) of 2-chloro-5-ethylpyrimidine in 20 mL of
ethylene glycol was added 5.23 g (52.3 mmol) of CaCO3 (particle
size 8.5-10 μm). The addition was in three portions over 5 min
to facilitate smooth stirring. The heterogeneous mixture was
heated at 130 °C for 26 h. The mixture was cooled to ambient
temperature and filtered through a short plug of Celite 545. The
filtering cake was washed with 40 mL of EtOAc. The combined
filtrates were diluted with 40 mL of water, and the aqueous layer
was extracted with 2 ꢀ 40 mL of EtOAc. The combined organic
layers were successively washed with 40 mL of water and 40 mL
of brine. The organic layer was concentrated under vacuum to
near dryness and diluted with 10 mL of i-PrOH. After being
stirred with ice cooling for 3 h, the solids were filtered, washed
with 6 mL of heptane, and dried at 60 °C to provide 5.61 g (75%)
of 6 as a white crystalline solid: mp 77.5 °C; 1H NMR (400 MHz,
CDCl3) δ 1.16 (t, J = 7.0 Hz, 3H), 1.35 (s, 6H), 2.43 (q, J = 7.0
Hz, 2H), 3.59 (d, J = 7.0 Hz, 2H), 3.78 (s, 3H), 4.70 (s, 1H), 6.86
2-[[4-[2-[(5-Ethyl-2-pyrimidinyl)[[4-[(trifluoromethyl)oxy]phenyl]-
methyl]amino]-1,1-dimethylethyl]phenyl]oxy]-2-methylpropanoic
Acid (GW693085, 1). A slurry of 7.16 g (179 mmol) of NaH
(60% dispersion in mineral oil) in 40 mL of THF was heated to
40 °C, followed by addition of 20.0 g (56.0 mmol) of 9 as a
solution in 160 mL of THF over 20 min. The exotherm from the
addition brought the reaction to about 44 °C. After the addi-
tion, the yellow slurry was heated to 65 °C and stirred for 50 min.
The mixture was then treated with 16.1 mL (101 mmol) of
4-(trifluoromethoxy)benzyl bromide over 10 min. The resultant
light yellow mixture was heated to 65 °C and stirred for 40 min.
After being cooled to ambient temperature, the mixture was
treated with 51.4 mL of 25% NaOMe in MeOH over 10 min and
stirred for 40 min to quench the excess alkylating reagent (to
undetectable by HPLC at UV 220 nm). The mixture was diluted
with 100 mL of water, and the aqueous layer was extracted with
100 mL of EtOAc. The combined organic layers were treated
with a mixture of 4.5 mL of 87% H3PO4 and 100 mL of water.
The organic layer was washed with 100 mL of brine and
concentrated under vacuum to about 60 mL. The partially
crystallized mixture was diluted with 10 mL of EtOAc and
30 mL of heptane to facilitate the stirring. After being stirred
at -5 °C for 2 h, the mixture was filtered, washed with 40 mL of
heptane, and dried at 70 °C to give 25.7 g (86%) of 1 as a white
(d, J = 8.0 Hz, 2H), 7.30 (d, J = 8.0 Hz, 2H), 8.07 (s, 2H); 13
C
NMR (100 Hz, CDCl3) δ 161.7, 158.1, 157.5, 138.8, 127.3, 125.2,
114.0, 55.5, 53.3, 38.4, 27.2, 23.0, 15.8. Anal. Calcd for C17H23-
N3O: C, 71.55; H, 8.12; N, 14.72. Found: C 71.85; H, 8.17;
N, 14.52.
4-[2-[(5-Ethyl-2-pyrimidinyl)amino]-1,1-dimethylethyl]phenol (8).
To a solution of 50.0 g (175 mmol) of 6 in 500 mL of CH2Cl2 was
added 49.7 mL (526 mmol) of BBr3 over 15 min with ice cooling.
The resultant white slurry was stirred at 5 °C for 2 h. The reaction
was quenched by transferring to 605 mL of 20% aqueous
K2CO3 solution over 30 min with ice-cooling. The layers were
separated, and the aqueous layer was extracted with 300 mL
of CH2Cl2. The combined organic layers were washed with
300 mL of brine and concentrated under vacuum to provide
62.3 g (quantitative) of 8 as a thick oil. The crude was used for
the next step without further purification: 1H NMR (400 MHz,
CDCl3) δ1.16 (t, J = 7.0 Hz, 3H), 1.34 (s, 6H), 2.45 (q, J = 7.0 Hz,
2H), 3.58 (d, J = 7.0 Hz, 2H), 4.93 (s, 1H), 6.66 (d, J = 8.0 Hz,
2H), 7.16 (d, J = 8.0 Hz, 2H), 8.11 (s, 2H); 13C NMR (100 Hz,
CDCl3) δ 160.8, 157.5, 155.4, 137.0, 127.3, 125.3, 115.6, 53.4, 38.3,
27.3, 22.9, 15.6. An analytical sample as light yellow oil was
obtained by chromatography on silica gel (30-70% EtOAc in
hexanes). Anal. Calcd for C16H21N3O: C, 70.82; H, 7.80; N, 15.49.
Found: C 70.80; H, 7.84; N, 15.41.
2-[[4-[2-[(5-Ethyl-2-pyrimidinyl)amino]-1,1-dimethylethyl]phenyl]-
oxy]-2-methylpropanoic Acid (9). A solution of 58.3 g (175 mmol) of
crude 8 in 500 mL of 2-butanone was prepared at 43 °C, followed by
addition of 51.6 g (1.29 mol) of NaOH (20-40 mesh) and 3.8 mL
(211 mmol) of water. The mixture was stirred at 43 °C for 3 h. The
heating was removed, and the mixture was treated with 64.6 g
(387 mmol) of 2-bromo-2-methylpropionic acid as a solution in
200 mL of MEK over 20 min. After being heated back to 43 °C and
stirred for 1.5 h, the reaction mixture was cooled to ambient
1
crystalline solid: mp 112.3 °C; H NMR (400 MHz, CDCl3)
δ 1.19 (t, J = 8.0 Hz, 3H), 1.34 (s, 6H), 1.61 (s, 6H), 2.46 (q, J =
8.0 Hz, 2H), 3.80 (s, 2H), 4.25 (s, 2H), 6.88 (d, J = 8.0 Hz, 2H),
6.91 (d, J = 8.0 Hz, 2H), 7.01 (d, J = 8.0 Hz, 2H), 7.25 (d, J =
8.0 Hz, 2H), 8.17 (s, 2H), 10.95 (br. s, 1H); 13C NMR (100 Hz,
CDCl3) δ 177.9, 157.2, 153.0, 148.1, 143.7, 137.8, 128.4, 127.3,
124.6, 121.9, 120.7, 80.0, 58.3, 50.2, 40.2, 27.0, 25.4, 22.9, 15.6;
HRMS (ESIþ) m/z calcd for C28H33F3N3O4 (MHþ) 532.2418,
found 532.2406. Anal. Calcd for C28H32F3N3O4: C, 63.27; H,
6.07; N, 7.91. Found: C, 63.37; H, 6.16; N, 7.92.
Supporting Information Available: Experimental proce-
dures for compounds 3b, 4, and 7, additional procedure for
compound 1, LC-MS analysis, the proposed pathway for the
formation of i, and NMR spectra for all new compounds (1-9).
This material is available free of charge via the Internet at
http://pubs.acs.org.
J. Org. Chem. Vol. 75, No. 11, 2010 3907