Toward a clean alternative to Friedel-Crafts acylation: In situ formation, observation, and reaction of an acyl bis(trifluoroacetyl)phosphate and related structures

Article abstract of DOI:10.1021/jo981264v

Reaction of acyl trifluoroacetates with phosphoric acid in the presence of trifluoroacetic anhydride (TFAA) leads to the ready formation of acyl bis(trifluoroacetyl)phosphates, which are powerful acylating agents. Formation of these species and the subsequent acylation reaction are carried out, without added solvent, in a single in situ reaction process. In this reaction system, anisole is rapidly acylated at ambient temperature using a variety of carboxylic acids giving the para isomer exclusively. TFAA acts as an activating agent and can be recovered from the reaction system as trifluoroacetic acid (TFA) and converted back to TFAA using a dehydrating agent, while phosphoric acid behaves as a covalent catalyst in the process. This reaction system has many features which are required elements of a clean alternative to the Friedel-Crafts process.

Full text of DOI:10.1021/jo981264v

8946
J . Org. Chem. 1998, 63, 8946-8951
Tow a r d a Clea n Alter n a tive to F r ied el-Cr a fts Acyla tion : In Situ
F or m a tion , Obser va tion , a n d Rea ction of a n Acyl
Bis(tr iflu or oa cetyl)p h osp h a te a n d Rela ted Str u ctu r es
Timothy P. Smyth* and Brian W. Corby
Department of Chemical and Environmental Sciences, University of Limerick,
National Technological Park, County Limerick, Ireland
Received J une 29, 1998
Reaction of acyl trifluoroacetates with phosphoric acid in the presence of trifluoroacetic anhydride
(TFAA) leads to the ready formation of acyl bis(trifluoroacetyl)phosphates, which are powerful
acylating agents. Formation of these species and the subsequent acylation reaction are carried
out, without added solvent, in a single in situ reaction process. In this reaction system, anisole is
rapidly acylated at ambient temperature using a variety of carboxylic acids giving the para isomer
exclusively. TFAA acts as an activating agent and can be recovered from the reaction system as
trifluoroacetic acid (TFA) and converted back to TFAA using a dehydrating agent, while phosphoric
acid behaves as a covalent catalyst in the process. This reaction system has many features which
are required elements of a clean alternative to the Friedel-Crafts process.
Sch em e 1
In tr od u ction
Friedel-Crafts (FC) acylation has found widespread
and successful application in industry for over a century.1
In the current drive toward less wasteful and more
environmentally friendly processes, where the emphasis
is on atom efficiency and recyclability, it has many
shortcomings. It is instructive to examine briefly the
various stages of this reaction (Scheme 1) in order to
identify the origin of these disadvantages and to attempt
to make a rational approach to the development of a
viable, more benign alternative. Activation of a carboxy-
lic acid is achieved in two distinct stages in FC acylation.
Conversion of the carboxylic acid to an acid chloride
provides partial activation through covalent bond forma-
tion. Complexation of this with a strong Lewis acid
provides further essential activation. For reasons of
solubility, this last step is most efficiently carried out in
dichloromethane. Strong Lewis acids, such as AlCl3, also
complex, however, to a very significant extent with the
carbonyl group of the product.2 Because of this, more
than a stoichiometric amount of Lewis acid is frequently
used and hydrolysis of the AlCl3 is required to liberate
the product. The atom inefficient and hence wasteful
aspect of FC acylation is due to the loss of both the
“catalyst” (AlCl3) and the activating agent (SOCl2); both
chlorine atoms of this latter are ultimately lost as HCl.
The requirement of using a strong Lewis acid, in more
than stoichiometric amount, in dichloromethane, is due
in essence to the low degree of activation attained in the
covalent bond formation step. This imposes the need for
significant additional activation, which necessitates use
of a strong Lewis acid.
complex formation, if required, should then be achievable
with a mild Lewis acid, and as such, this should not
complex significantly with the product and so its action
should be catalytic in nature. Ideally, the formation of
the activated covalent complex should occur in situ in a
facile reaction starting from a carboxylic acid. Further-
more, there should be no specific solvent requirements;
the spent activating agent should be fully recoverable and
recylable in high yield, while the acylation reaction itself
should occur at moderate temperatures in high yield and
with high selectivity. The challenge is to achieve this
using reagents/catalysts that are not unduly hazardous
and are readily recyclable.
Acyl trifluoromethanesulfonates (acyl triflates) go some
way to meeting these requirements. As the conjugate
base of a superacid,3 the triflate moiety provides signifi-
cant activation of the acyl carbonyl group, and acyl
triflates have been found to effect acylation of benzene
(90%, 5 h) and of chlorobenzene (67%, 5 h), when used
in stoichiometric amounts at moderate temperatures (60
°C) without any added Lewis acids.4 Their preparation,
as reported to date, however, started with acid chlorides
and required the use of triflic acid, a hazardous material
and one not easily recoverable (bp 162 °C). Nafion-H
One approach to a potential alternative process would
be to achieve a much greater degree of activation at the
covalent bond formation stage. Further activation through
(3) (a) A value of -14 has been determined for Ho (in essence an
apparent pKa) of triflic acid. See: Farcasiu, D.; Miller, G. J . Phys. Org.
Chem. 1989, 2, 425. (b) See also: Olah, G. A.; Prakash, G. K. S.;
Sommer, J . Superacids; Wiley-Interscience: New York, 1985.
(4) (a) Effenberger, F.; Epple, G. Angew. Chem., Int. Ed. Engl. 1972,
11, 299. (b) Effenberger, F.; Eberhard, J . K.; Maier, A. H. J . Am. Chem.
Soc. 1996, 118, 12572.
(1) Olah, G. A. Friedel-Crafts Chemistry; J ohn Wiley & Sons: New
York, 1973.
(2) Ashfort, R.; Desmurs, J .-R. In The Roots of Organic Development;
Desmurs, J .-R., Ratton, S., Eds.; Elsevier: Amsterdam, 1996; p 3 and
references therein.
10.1021/jo981264v CCC: $15.00 © 1998 American Chemical Society
Published on Web 11/10/1998

Clean Alternative to Friedel-Crafts Acylation
J . Org. Chem., Vol. 63, No. 24, 1998 8947
Sch em e 2
mediated acylation with carboxylic acids at reaction rates
that were convenient to monitor using NMR spectroscopy.
Reaction progress was monitored by adding 1 drop of the
(neat) reaction mixture into CDCl3 (0.5 mL) and recording
1
the H NMR spectrum; the dilution process was found
to quench the reaction very effectively. The initial results
provided an exemplary illustration of the value of this
acylation process and unambiguous proof of the strong
catalytic effect of phosphoric acid. Some of these results
are presented here and are discussed in terms of the
reactions outlined in Scheme 3, which shows the various
acylated phosphoric acid structures that can occur on
reaction of acid anhydrides with H3PO4. The chemical
shift of H-R (i.e., R to the acyl carbonyl group) in the
various structures is quoted below as a key identifier of
reaction progress.
provides triflic acid in an immobilized form5 and, as such,
furnishes a potential solution both to the hazardous
nature of triflic acid and to its recovery and reuse. So
far, the use of Nafion-H has met with limited success in
aromatic acylation reactions; the heterogeneity of the
reaction system may be a restricting factor. Olah and
co-workers6 found that it worked well (85-96% of acy-
lated product) only when a reactive acid chloride (p-
nitrobenzoyl chloride) was heated to reflux (usually >100
°C) in an excess of various aromatic hydrocarbons; a
drawback was the occurrence of a distribution of isomers
(o, 16-22%; m, 1-3.5%; p, 68-74%) the formation of
which may have been due to the high temperatures used.
More recently, Yamato and co-workers7 found that a
limited number of intramolecular acylations worked well
(>90%, 0.5 h, 80 °C) with Nafion-H and acid chlorides;
use of the corresponding carboxylic acids was consider-
ably less efficient.
We recently reported on the successful use of an acyl
trifluoroacetate, formed in situ from a carboxylic acid and
trifluoroacetic anhydride (TFAA), as an acylating agent
in an industrially based synthesis of a key tamoxifen
intermediate.8 It was noted that the in situ reaction of
phosphoric acid with the acyl trifluoroacetate resulted
in an entity with enhanced acylation potential and that
acylation occurred exclusively in the para position at a
reaction temperature of approximately 60 °C. In addi-
tion, we demonstrated that the spent TFAA could be
recovered as trifluoroacetic acid (TFA) and readily con-
verted back to TFAA using a dehydrating agent.9 Prod-
uct throughput per batch was very high, and further-
more, reaction calorimetry indicated that the process was
suitable for scale-up. On the basis of these findings and
observations, we felt that TFAA/H3PO4-mediated acyla-
tion warranted detailed evaluation as a viable, clean
alternative to FC acylation (Scheme 2). We have carried
out a mechanistic study that has provided an incisive
picture on the unique role of H3PO4 as a covalent catalyst
in this reaction. We report here on the mechanistic work
and on the scope of this acylation process.
Reaction of TFAA (2 equiv) with 2-phenylbutanoic acid
(1a ) (1 equiv) (H-R, 3.45 ppm) led to the rapid formation
of the trifluoroacetate 2a (H-R, 3.66 ppm). Addition of
anisole (1 equiv) resulted in the quantitative formation
of the acylated product 3a (H-R, 4.48 ppm), with a
reaction half-life at 10 °C of approximately 2 h (Figure
1c). The splitting pattern of the aromatic hydrogens of
the anisole moiety of 3a (Figure 1d) indicated a para-
substituted structure exclusively, and the presence of a
clean singlet for the methoxy group was further proof for
the formation of a single isomeric product.11 In a repeti-
tion of this reaction, 85% phosphoric acid (0.1 equiv) was
added after the formation of 2a . Addition of anisole (1
equiv) again led to the quantitative formation of the same
acylated product 3a , but, significantly, the half-life at 10
°C was now less than 3 min. When 0.01 equiv of H3PO4
was used, the half-life at 10 °C was 30 min, clearly
indicating that the concentration of the active acylating
agent was dependent on the concentration of H3PO4.12
By carrying out the acetylation of anisole using acetic
acid with H3PO4 (0.1 equiv) and with either TFAA or
Ac2O as the added anhydride (2 equiv in each case), we
were able to confirm the key role of TFAA in forming the
active acylating agent. Using TFAA, the yield of acety-
lated product 3b was 68% after 1 h at 10 °C, while the
yield was less than 25% after 24 h at 25 °C using Ac2O.
These observations provided an unequivocal illustration
of the key role of both H3PO4 and TFAA in this acylation
process. Questions were still unanswered, however, as
to the precise identity of the active acylating agent.
Chemical logic would dictate that acyl bis(trifluoroacetyl)-
phosphate (6) should have the most polarized acyl
carbonyl group of the phosphate structures shown in
Scheme 3 and hence should be the most active acylating
agent. It is relevant to note that acyl dichlorophosphoric
acids, RC(O)OP(O)Cl2, are known to be reactive acylating
agents,13 and, given that the inductive effect of OC(O)-
CF3 (σm, 0.56) is larger than that of Cl (σm, 0.37),14 it is
logical that acyl bis(trifluoroacetyl)phosphates should
Resu lts a n d Discu ssion
In our preliminary mechanistic work, we used anisole
as the aromatic substrate, as it underwent TFAA/H3PO4-
(10) The unusual chemical shift for this peak is due to frequency
folding as a result of the narrow (1000 Hz) sweep width used. See:
Gu¨nther, H. NMR Spectroscopy, 2nd ed.; J ohn Wiley & Sons: Chich-
ester, 1995; pp 255-256.
(11) This regiospecificity has also been reported by others. See:
Ranu, C.; Ghosh, K.; J ana, U. J . Org. Chem. 1996, 61, 9546.
(12) The second equivalent of TFAA served to react with the water
content of the H3PO4 when this was present and maintained essentially
constant the volume of the reaction mixture between these runs and
that carried out without H3PO4.
(5) Nafion is the trade name of Du Pont for perfluorinated sulfonic
acid polymer, which is available in a variety of physical forms. See:
Aldrichimica Acta 1986, 19 (3), 76.
(6) Olah, G. A.; Malhortra, R.; Narang. S. C.; Olah, J . A. Synthesis
1978, 672.
(7) Yamato, T.; Hideshima, C.; Prakesh, G. K. S.; Olah, G. A. J . Org.
Chem. 1991, 56, 3955.
(8) Smyth, T. P.; Corby, B. W. Org. Process Res. Dev. 1997, 1, 264.
(9) TFAA is produced commercially by dehydration of TFA. Some
(13) Effenberger, F.; Konig, G.; Klenk, H. Angew. Chem., Int. Ed.
Engl. 1978, 17, 695.
(14) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 185.
processes use SO3 as the dehydrating agent giving H2SO4 as
coproduct. On a laboratory scale, P2O5 was more convenient to use.
a

8948 J . Org. Chem., Vol. 63, No. 24, 1998
Smyth and Corby
Sch em e 3
have superior acylation potential. A σm value of 1.00 was
evaluated for P(O)(OC(O)CF3)2 from the correlation shown
in Figure 2 (R2 ) 0.84; data in Table 1), and from this, a
value of 0.71 was estimated for OP(O)(OC(O)CF3)2 by
subtracting 0.29, which is the difference in the experi-
mentally determined σm values of P(O)(C3F7)2 and OP-
(O)(C3F7)2.14,15
during the reaction.8 We found that addition of PA to
the reaction solutions used here similarly prevented
precipitate formation, and this stratagem facilitated
spectroscopic study of the species occurring in these
reaction systems (the acylation of PA itself did not
interfere as at room temperature this process was quite
slow).8 The spectrum resulting from the reaction of
TFAA and 85% H3PO4 (4:1 equiv) followed by the addition
of PA (1 equiv) is shown in Figure 3c; the presence of
tris(trifluoroacetyl)phosphate (12) was clear, while the
presence of the other large peak was interpreted as
corresponding to ion pairs of PA with phosphoric acid
species. Formation of tight ion pairs of such acids with
PA was viewed as being at least partly responsible for
the lack of formation of a precipitate. The 19F chemical
shift values of 10-12 were also recorded (Table 2).19 The
mono-, bis-, and trisacetyl phosphates 4b, 7b, and 8b,
respectively were formed by reacting phosphoric acid and
acetic anhydride (a precipitate was not observed in this
instance, a finding which was consistent with the con-
siderably poorer leaving group ability of acetate compared
to trifluoroacetate).
In an effort to directly observe and characterize the
active acylating agent in this reaction system, we used
19F and 31P NMR to study the reaction patterns shown
in Scheme 3; key characterization data are given in Table
2. Species 10-12 were formed by reacting TFAA with
85% phosphoric acid. The 31P NMR spectrum obtained
shortly after mixing these materials in a ratio of 3:1 is
shown in Figure 3a. The spectrum changed with time,
and the formation of a precipitate occurred after 10-20
min;18 the spectrum of the resulting supernatant is shown
in Figure 3b. The assignment of the chemical shifts to
the mono-, bis-, and tris(trifluoracetyl)phosphates 10, 11,
and 12, respectively, was supported by observing the
disappearance of the peak assigned to 10 and an increase
in those assigned to 11 and 12 on addition of a further
equivalent of TFAA, whereas addition of D2O resulted
in the rapid disappearance of all these peaks and the
formation of phosphoric acid. In our previous work on
the acylation of N,N-dimethyl-2-phenoxyethylamine (PA),
the formation of a precipitate did not occur at any stage
Having thus established a set of reference 31P and 19F
NMR chemical shift data, we proceeded to study solutions
in which the formation of the acyl bis(trifluoroacetyl)-
phosphate 6a could occur. The 31P NMR spectrum of the
solution obtained from the reaction of 1a , TFAA, 85% H3-
PO4, and PA (1:4:1:1 equiv) is shown in Figure 4a. The
appearance of the peak at -18.70 ppm was significant
as this was not observed in the reaction mixture of the
above components when 1a was omitted (Figure 3c).
Addition of anisole (1 equiv) resulted in rapid decrease
in this peak (Figure 4, a f b f c). This observation was
paralleled in the 19F spectrum of the same solution by
the decrease in the peak at -76.79 ppm (Figure 4, a′ f
b′); the concomitant acylation of anisole was confirmed
(15) Yagupol’skii, L. M.; Pavlenko, N. V.; Ignat’ev, N. V.; Matyush-
echeva, G. I.; Sementii, V. Ya. Zh. Obsh. Khim. 1984, 54, 297EE.
(16) Gu¨nther, H. NMR SpectroscopysAn Introduction; J ohn Wiley
& Sons: Chichester, 1980; Chapter 10.
(17) Tebby, J . C., Ed. CRC Handbook of Phosphorous-31 Nuclear
Magnetic Resonance Data; CRC Press: Boca Raton, 1991; Chapter 1.
(18) The peaks assigned to 11 and 12 increased in intensity, while
the broad peaks became broader at the onset of precipitate formation.
The precipitate was considered to arise from pyro- and polyphosphate-
type materials. Displacement of trifluoroacetate from 10-12 by the
hydroxy group of some other phosphoric acid structure can lead to
pyrophosphate formation. This type of reaction has been used to form
specific pyrophosphates. See: Corby, N. S.; Kenner, G. W.; Todd, A.
R. J . Chem. Soc. 1952, 1234.
(19) The spectra are included in Supporting Information. The ion
pair TFA- PA+ was observed in these solutions; assignment of this
peak was confirmed by reaction of PA with TFA separately.

Clean Alternative to Friedel-Crafts Acylation
J . Org. Chem., Vol. 63, No. 24, 1998 8949
anisole.21 Attributing the 31P chemical shift at -18.70
ppm to a species such as 6a was reasonable, as this value
was bracketed by those of tris(trifluoroacetyl)phosphate
(12) (-24.8) and trisacetyl phosphate (8b) (-17.7) and
was similar to that observed for bis(trifluoroacetyl)-
phosphoric acid (11) (-18.57 ppm) (Table 2). On the
basis of the chemical shift value alone, we could not
distinguish between 6a and 9a as the observed species;
however, the former ought to be the more predominant
species present given that the mixed anhydride 2a and
TFAA were in a ratio of 1:3 at the outset (although 4
equiv of TFAA were added at the outset 1 equiv was
consumed by the water content of 85% H3PO4). The 19F
peak at -76.79 ppm must also be assigned to the active
acylating species, putatively 6a , and this assignment was
consistent with the general pattern of chemical shifts
observed for species such as 11 and 12, although the
small span of the 19F NMR chemical shifts observed for
these structures made absolute assignment difficult.
Overall, the foregoing 31P and 19F peaks (Figure 4, spectra
a and a′) were attributed to the active acylating agent in
the system and they unambiguously showed that this
species contained a trifluoroacetyl moiety and a phos-
phoric acid moiety (and ipso facto an acyl entity), which
concurred with our earlier observations on the key role
of TFAA and of H3PO4.
We then focused on determining the range of aromatic
structures that could be readily acylated in this reaction
system (without PA) with a variety of carboxylic acids
including benzoic acid.22 The nature of both the aromatic
substrate and the carboxylic acid played a major role in
the acylation reaction (Table 3). Anisole was very readily
acylated by a variety of carboxylic acids giving a quan-
titative yield of the para isomer, while 2-phenylbutanoic
acid was the best carboxylic acid in terms of acylating a
wide range of aromatic structures. One parameter that
was varied was the number of equivalents of TFAA and
H3PO4 used per equivalent of RCO2H. A reaction system
that could form only a low concentration of 6, i.e., with
H3PO4 (0.1 equiv), was adequate for rapid acylation of
an activated substrate such as anisole with most car-
boxylic acids. To acylate a less activated substrate such
as toluene at a reasonable rate, a higher concentration
of 6 was necessary. This was acheived by the use of
TFAA/H3PO4 (4:1 equiv). Under these conditions, a
precipitate formed in every case, indicating that 6 could
not have been present at its maximum stoichiometric
concentration with respect to the concentration of H3PO4
added at the outset. Formation of the precipitate was
viewed as a key factor in defining the present limitations
with nonactivated aromatic substrates such as benzene;
addition of triethylamine to prevent formation of the
precipitate, in lieu of PA (which was acylated at a
comparable rate to toluene), did not provide a solution.
The exclusive formation of the para isomer of the
product 3 is not likely to have resulted from product
stability as, even with quite a bulky group such as
2-phenylbutanoyl, the carbonyl group positions the bulky
moiety some distance out from the substituents on the
aromatic ring. The fact that the reactions were under
kinetic control was substantiated by the results obtained
F igu r e 1. 1H NMR spectra of (a) 2-phenylbutanoic acid (1a ),
(b) reaction sample 15 min after addition of TFAA (2 equiv)
to 1a , (c) this reaction sample 2 h after addition of ansiole (1
equiv), and (d) reaction sample 15 h after addition of ansiole.
The reaction mixture was maintained at 10 °C, and in each
case, a sample (1 drop) of the neat reaction mixture was added
directly to CDCl3 (0.5 mL). The labeled peak (*) was exchange-
able on addition of D2O and was attributed to the acidic
hydrogen of TFA.10
F igu r e 2. Correlation of the Hammet σm values of selected
groups X and the corresponding groups P(O)X2.
by 1H NMR.20 Further addition of 2a to the reaction
solution at this point led to the reappearance of the
aforementioned 19F and 31P peaks, which once more
readily disappeared on addition of another equivalent of
(21) The full sequence of 31P and 19F NMR spectra showing this
double acylation cycle is included in Supporting Information.
(22) In our previous work (ref 8) we indicated that benzoylation did
not work. We wish to clarify here that benzoylation of anisole does
occur, albeit somewhat slowly, with benzoyl trifluoroacetate.
(20) The acylation of anisole was slower in the presence of PA than
in the absence of PA.

8950 J . Org. Chem., Vol. 63, No. 24, 1998
Ta ble 1. Ha m m ett σm Va lu esa for Gr ou p s Sh ow n in F igu r e 2
Smyth and Corby
X
C3F7
Cl
F
OMe
OEt
OnBu
OnPr
C6H5
Me
Et
σm of X
σm of P(O)X2
0.44
0.95
0.37
0.78
0.34
0.81
0.12
0.42
0.10
0.55
0.10
0.41
0.10
0.38
0.06
0.38
-0.07
0.43
-0.07
0.37
a
Data from ref 14.
Ta ble 2. Ch em ica l Sh ifts (p p m ) of Key Str u ctu r es
1H (H-R)
19Fa
31Pb
1a ; 2a ; 3a ; 4a
1b; 2b; 3b
3.45; 3.66; 4.48; 3.55
2.10; 2.40; 2.55
2.20; c; c
f; -77.10; f; f
f; -77.10; f
4b; 7b; 8b
+1.28; -7.67; -17.70
TFA; TFAA; TFA-PA+
10; 11; 12
6a
-77.00; -76.05;d -76.57
-76.88;e -76.86;e -76.60
-76.79
-6.13; -18.57; -24.8
-18.70
c
a
b
TFA was used as the internal reference (δ ) -77.00 with respect to CFCl3).16 Phosphoric acid was used as the reference, and peaks
upfield of this are reported as negative values.17 c Not readily discernible. The chemical shift of TFAA was observed to vary slightly
with respect to TFA depending on the composition of the reaction mixture; addition of extra TFAA made assignment unambiguous.
e These may correspond to the ion pairs of 10 and 11 with PA, respectively. f No 19F resonance present for these structures.
d
with o- and p-xylene. The half-life for acylation of
o-xylene with 2-phenylbutanoic acid was 20 min at 25
°C, giving the para-acylated product 3e, while that for
p-xylene was 120 min at 25 °C, giving an ortho-acylated
product 3e′′; the latter was estimated23 to be the more
stable product on the basis of ∆Ho . It is probable that
f
the regiospecificity was determined by the differing
stability of the ortho and para addition intermediates or
transition states leading to these, as here the intact
acylating agent must interact with the aromatic sub-
strate. This would also imply that free acylium ions were
not involved, as then the difference in stability of the
ortho and para intermediates, or transition states leading
to these, should mirror the pattern of product stability
shown above.
We examined the effect of BF3 etherate as a homoge-
neous catalyst in the reaction system; it did not, however,
have any beneficial effect with activated or with the
nonactivated aromatic substrates. Use of AlCl3, BiCl3,
24
or Bi2O3 as a heterogeneous catalyst was similarly
without useful effect. A good deal of the corresponding
acid chloride resulted from treatment of 2a with AlCl3;
a similar result was observed in the reaction system
involving 6a .
Con clu sion s
The TFAA/H3PO4-mediated acylation system is clearly
a practical, atom efficient alternative to FC acylation
suitable for the production of a variety of fine chemical
intermediates8 and also for the bulk production of some
simple acylated aromatics. The process allows for the
in situ assembly and reaction of a highly active acylating
species, an acyl bis(trifluoroacetyl)phosphate, starting
from a carboxylic acid. There are limitations with
nonactivated substrates, but there is scope for further
development.
F igu r e 3. The 31P NMR spectrum of (a) the neat solution
obtained 5 min after addition of TFAA to 85% H3PO4 (3:1
equiv), (b) the supernatant obtained from this solution after
15 min, and (c) the neat solution obtained after addition of
PA to a mixture of TFAA and 85% H3PO4 (4:1 equiv) (no
precipitate formed here). The labeled (*) peak was attributed
to ion pair(s) of PA with phosphoric acid based species.
Exp er im en ta l Section
Gen er a l Acyla tion P r oced u r e. TFAA (5.20 mL, 36.7
mmol) was added directly to the appropriate carboxylic acid
(9.2 mmol). The solution was cooled to below 10 °C and 85%
phosphoric acid (1.06 g, 9.2 mmol) was added with stirring.
(23) Ampac 5.0; Semichem: Shawnee, KS, 1994.

Clean Alternative to Friedel-Crafts Acylation
J . Org. Chem., Vol. 63, No. 24, 1998 8951
Ta ble 3. Gen er a l Resu lts for TF AA/H3P O4-Med ia ted
Acyla tion of X-Ar -H (1 equ iv/equ iv of RCO2H)
R
X
Ph(Et)CH
Me
4/1
10, 5 m
100
3b
4/1
(Me)3C
Ph
labels
MeO
2/0.1
10, 20 m
100
2/0.1
10, 4 h
100
3c
4/1
2/0.1
60, 4 h
100
n/ma
T,b tc
Yd
3a
3d
Le
1,2-diMe 4/1
4/1
n/ma
T,b tc
Yd
25, 3 h
25, 24 h 60, 24 h 60, 24 h
100
3e
4/1
70
3f
4/1
0
3 g
4/1
100
3h
4/1
Le
Me
n/ma
T,b tc
Yd
60, 2.5 h
100
3i
60, 24 h 60, 24 h 60, 96 h
∼5
0
3k
60
3l
3j
Le
n/m ) equiv of TFAA/H3PO4. Temperature (°C). c Reaction
a
b
time. Yield (%) from 1H NMR data. e Product label.
d
added once dissolution of the 85% H3PO4 was complete. The
31P NMR spectra of the neat solution was recorded as indicated
above, while the 19F (84.25 MHz) NMR spectrum was recorded
in a 5 mm tube of solutions made by adding one drop of the
reaction mixture to CDCl3 (0.5 mL).
P r od u ct Ch a r a cter iza tion Da ta . 1H NMR (CDCl3, 90
MHz). 1-[4-[methoxy]phenyl]-2-phenyl-1-butanone (3a ):
δ
0.90 (t, J ) 7.5 Hz, 3H), 1.70-2.30 (m, 2H), 3.85 (s, 1H), 4.48
(t, J ) 7.5 Hz, 1H), 6.90 (d, J ) 10.5 Hz, 2H), 7.29 (br s, 5H),
8.0 (d, J ) 10.5 Hz, 2H) (see Figure 1d). 4-Methoxy acetophe-
none (3b): δ 2.68 (s, 3H), 3.84 (s, 3H), 7.00 (d, J ) 10.5 Hz,
2H), 8.10 (d, J ) 10.5 Hz, 2H).25 1-[4-[Methoxy]phenyl]-2,2-
dimethyl-1-propanone (3c): δ 1.45 (s, 9H), 3.90 (s, 3H), 6.95
(d, J ) 10.5 Hz, 2H), 7.87 (d, J ) 10.5 Hz, 2H). 4-Methoxy
benzophenone (3d ): δ 3.95 (s, 3H), 7.00 (d, J ) 10.5 Hz, 2H),
7.50-7.91(m, 7H).25 1-[3,4-[Dimethyl]phenyl]-2-phenyl-1-bu-
tanone (3e): δ 0.90 (t, J ) 7.5 Hz, 3H), 1.7-2.3 (m, 2H), 2.3
(s, 6H), 4.52 (t, J ) 7.5 Hz, 1H), 7.17 (d, J ) 10.5 Hz, 1H),
7.30 (s, 5H), 7.72 (d, J ) 10.5 Hz, 1H), 7.77 (s, 1H). 1-[2,5-
[Dimethyl]phenyl]-2-phenyl-1-butanone (3e′′): δ 0.95 (t, J )
7.5 Hz, 3H), 1.70-2.30 (m, 2H), 2.3 (s, 6H), 4.35 (t, J ) 7.5
Hz, 1H), 7.00-7.50 (m, 8H). 3,4-Dimethylacetophenone (3f):
δ 2.37 (d, 6H), 2.70 (s, 3H), 7.26 (d, J ) 10.5 Hz, 1H), 7.75 (d,
J ) 10.5 Hz, 1H), 7.80 (s, 1H).25 3,4-Dimethylbenzophenone
(3h ): δ 2.36 (s, 3H), 2.40 (s, 3H), 7.25-7.85 (m, 8H).25 1-[4-
[Methyl]phenyl]-2-phenyl-1-butanone (3i): δ 0.90 (t, J ) 7.5
Hz, 3H), 1.70-2.30 (m, 2H), 2.35 (s, 3H), 4.51 (t, J ) 7.5 Hz,
1H), 7.10-7.50 (m, 7H), 7.89 (d, J ) 10.5 Hz, 2H). 4-Methyl
benzophenone (3l): δ 2.46 (s, 3H), 7.20-7.90 (m, 9H).25
F igu r e 4. The 31P NMR spectrum of (a) the neat solution
obtained from the combination of 1a , TFAA, 85% H3PO4, and
PA (1:4:1:1 equiv) (see Figure 3 for assignment of the labeled
(*) peak). Anisole (1 equiv) was then added. Spectrum b is that
obtained over a full 10 min accumulation time, and (c) is that
obtained after a further 5 min. The 19F NMR spectra of (a′)
the initial solution above and (b′) the same solution 10 min
after addition of anisole are also shown; these latter spectra
were taken of reaction samples added to CDCl3, and the
acquisition time was quite short.
After complete dissolution of the phosphoric acid, the aromatic
substrate (9.2 mmol) was added and the reaction mixture was
stirred at the required temperature; where TFAA (4 equiv)
was used, heating to reflux maintained a solution temperature
of approx 60 °C. Reaction progress was monitored by adding
a small aliquot (1 drop) of the neat reaction mixture, taken at
various time intervals, directly to CDCl3 (0.5 mL) and record-
ing the 1H NMR spectrum (data below). The ratio of TFAA
and H3PO4 was varied as indicated in the text and in Table 3;
the time required for complete dissolution of small quantities
(0.1 equiv) of 85% H3PO4 in TFAA (on the scale detailed above)
was approximately 20 min. The isolation procedure involving
the recovery of TFA has previously been described.8 For small-
scale work, product isolation is best achieved by quenching
the reaction mixture in aqueous base and extracting with an
organic solvent.
Ack n ow led gm en t. B.W.C. acknowledges Klinge
Pharma, Killorglin, County Kerry and the Irish Ameri-
can Partnership for financial support. We are also
grateful to Mr. Max Stern and, in particular, Dr. Bob
Khan of Klinge Pharma for their support throughout
this work.
Su p p or tin g In for m a tion Ava ila ble: The following spec-
tra are available: 19F NMR of solutions containing 10-12; 31
P
and 19F NMR showing the formation and disappearance of 6a
in a double cycle of acylation of anisole (2 pages). This
material is contained in libraries on microfiche, immediately
follows this article in the microfilm version of the journal, and
can be ordered from the ACS; see any current masthead page
for ordering information.
P r ep a r a tion of Sa m p les for Sp ectr oscop ic Exa m in a -
tion . Reaction mixtures were prepared as above using TFAA/
H3PO4 in a ratio of 3:1 (equiv). The 31P (36.23 MHz) NMR
spectra of the neat solution was recorded using a 10 mm tube
with an insert that contained 85% H3PO4 in D2O to provide a
reference and lock signal. To a freshly made solution, N,N-
dimethyl-2-phenoxyethylamine (PA) (1.52 g, 9.2 mmol) was
J O981264V
(24) Desmurs, J .-R.; Labrouillere, M.; Dubac, J .; Laporterie, A.;
Gaspard, H.; Metz, F. In The Roots of Organic Development; Desmurs,
J .-R., Ratton, S., Eds.; Elsevier: Amsterdam, 1996; p 15.
(25) Pouchert, C. J . The Aldrich Library of NMR Spectra, 2nd ed.;
Aldrich Chemical Co., Inc.: Milwaukee, 1983; Vols. 1 and 2.