-m,30 fs.

oo+00
0 I990 Pcrgnmon Pra plc

TETRAHEDRON REPORT NUMBER 269

ACETIC FORMIC ANHYDRIDE

A REVIEW

PAOLOSTRAZZOLINI,ANGELOG. GILJMANMI*and SABINACAUCI

Department of Chemistry, University of Udine, Vialc Unghcria 43.1-33100 Udinc, Italy

(Received in UK 23 February 1989)

c0NTEml-s

Introduction ............................. 1081

I. Physicu-Chemical Properties and Structural Aspects ............... 1082
2. Spuztroscopic Studies ......................... 1085
2.1. Infrared spectroscopy ....................... 1085
2.2. Nuclear magnetic resonance ..................... 1086
2.3. Other. ............................ 1088
3. Kinetic Work. ........................... 1088
4. Irradiation Chemistry and Mass Spectrometry ................. 1089
5. Preparations. ........................... 1089
6. Chemical Prop&es ......................... 1091
6.1. O-Formylations ......................... 1091
6.1.1. Phenolsandalcohols ..................... 1091
6.1.2. Hydroxyacids. ....................... 1095
6. I .3. Tetracyclines ..................... 1095
6.1.4. Ribonucleosida .. : : .................... 1096
6.1.5. Hydroxy polymeric systems ................... 1097
6.1.6. N-Oxides ......................... 1097
6.2. N-Fonnylations ......................... 1097
6.2.1. Amins and polyamines .................... 1097
6.2.2.Aminoacid.s ........................ 1104
6.2.3. Amino alcohols and amino phenols. ................ 1107
6.2.4. Sulfonamides ........................ 1108
6.2.5. Phosphonamines ....................... 1108
6.3. S-Formylations ......................... II08
6.4. P-Formylations ......................... II08
6.5. C-Fonnylations ......................... 1109
6.6. Cyclixation reactions. ....................... III0
6.7. Formylation of metal complexes .................... Ill2
6.8. Miscellaneous reactions. ...................... Ill3
6.9. Halo- and thioderivatives ...................... 1113
7. Dangerous Properties and Toxicology. ................... 1113
8. Miscellaneous Remarks ........................ 1114

INTRODUCTION
The goal of this Report is that of collecting all available published data on the properties, reactivity
and applications of acetic formic anhydride (AFA, 1). The Report is exhaustive and contains
literature coverage through June 1988. A recent review by 01ah96 deals less specifically with
formylating agents.
1081
1082 P. %RAZZoLlNl er a!.

1. PHYSICOCHEMICAL PROPERTIES AND STRUCIURAL ASPECTS

Physical data on 1 and other formic anhydrides are very scant. No accurate data about vapour
pressure and heats of vaporization are available.
The simplest member of the formic anhydride series, formic anhydride, is a liquid [reported
boiling points are : 2627C at 3600 Pa (a fraction with ca 90% mol/mol titre) and 27C at 3733
Pa (91/ purity)Z4]. It is known to decompose to carbon monoxide and formic acid at room
temperature with a time constant of 1 h.82 Even trap to trap distillation at ca - 10C under reduced
pressure gave mixtures of formic anhydride and formic acid.
Acetic formic anhydride (l), is a colorless lachrymatory liquid, b.p. 29C (2267 Pa) (Table 1).
The melting point of AFA as well as its free energy of formation, the free energy of decomposition

Table I. Some reported boiling points of AFA

T(C) P @a) Refs

27-B 1333 35
25.~26.5 1333 55
21-28 1333 80
29. 2267 I7
32 2267 48
3S36 2666 122
35 3200 II2
losl20* 101325 I7

*From distillation of formic acid-acetic anhydride

mixture (FAM, 2, see Section 5).

to acetic acid and carbon monoxide and its spontaneous decomposition rate constant are not
known. In fact, 1 is somewhat unstable and decomposes slowly at room temperature, but much
faster above 60C. A sealed sample was reported to have exploded after two weeks because of CO
evolution. For this reason AFA should be stored at 4C in an unsealed standard round bottom
flask fitted with a polyethylene stopper.

CH,CO&CHO- A CH#XOH + CO
1

The decomposition of 1 is accelerated by sodium acetate lo or pyridine.7*90*o Decomposition

depends upon the solvent : it is very fast in toluene, carbon tetrachloride, nitrobenzene and benzal-
dehyde, moderate in benzene and acetone, slow in ethanol, I-pentanol and isobutanol and im-
perceptible in carbon disulfide and ether.' lo This reaction is catalyzed by some strong tertiary
nitrogen bases such as N,N-dimethylbenzeneamine, brucine, strychnine and nicotine whereas the
action of quinine, morphine, quinoline and papaverine is very weak. It was found that neither
primary nor secondary bases acted similarly and other compounds with less basic character did not
cause the liberation of carbon monoxide.
A possible mechanism for the decomposition by tertiary bases requires an attack by the base
on the formyl carbonyl of 1 with formation of the positively charged formyl base 3 and the counterion
acetate; analogously to the similar reaction between pyridine and acetic anhydride. Now, if the
formylated base 3 is primary or secondary, it will recover its lone pair simply by transferring its
proton to another molecule of base or acetate leading to substitution, whereas, if there are no
hydrogen atoms available on the nitrogen, it will recover its neutrality by transferring the formyl
proton with evolution of CO.
 Acetic formic anhydride 1083

R
I
ACOH + :N-Cl+0

R R
I I w R
//O
Ad-c - :N--It"- AcO + RLNI-CHO
I I 1.1
R II. R

ko- + R*kH + CO
1 3 I
R

The acids (sulfuric, nitric, hydrofluoric, hydrochloric and, to a lesser extent, phosphoric) also
catalyze this reaction. Hydrofluoric acid and 1 at 0C and atmospheric pressure produced a mixture
of formyl fluoride, acetyl fluoride and the corresponding acids. The reaction, depending on relatively
minor variations of conditions, may yield equimolar amounts of the formyl halides or almost solely
formyl fluoride (61%).95 On the other hand concentrated acetic acid did not cause evolution of
CO,O therefore, in the preparation of 1 from acetic anhydride and concentrated formic acid, it is
reasonable to conclude that the observed partial decomposition of 2 is due to formic acid. Olah and
Dunne observed94 some evolution of CO upon protonation of 1 in superacid medium at low
temperature (- 80C) : an excess of FS03H-SbF&302 solution cleaved the protonated anhydride
(4) to protonated formic (5) and acetic (6) acids plus the acetyl cation (7). Fonnyl cation (8) was
not observed, because it decomposed as soon as it was formed. The cleavage of the mixed anhydride
under these conditions was rationalized in terms of a further protonation on the ether oxygen (the

CN
3--C=O+
/ 1 .OtR 4O*H + t
0 * CH3-C + H-C t CH3- c= 0 t H-C= 0
\ ,n.H OH OH [ I
,c=o
H

4 6 5 7 8

remaining lone pair of the carbonyl oxygens was already coordinated in the first protonation, which
occurs even at low concentrations of superacid medium) : the charge-charge repulsion destabilizes
this diprotonated species causing the cleavage. By treating AFA with an equimolar amount of
superacid solution only the monoprotonated species (4) was obtained : it underwent a rapid inter-
and intra-molecular proton exchange, as shown by the PMR spectroscopy (Fig. 1).94

m -.I
16 15 10 9 3 2 PPa
Fig. I. H-NMR spectrum of protonsted AFA in superacid medium.*
1084 P. SmZWLlNl el d.

Acetic formic anhydride was found to be a non planar molecule, by gas phase electron diffraction
studies.46 The formyl and the acetyl fragments are twisted out of the C-O-C plane by 21 and
46, respectively, giving a dihedral angle of 126 between the two moieties. The formyl C-H and
the acetyl c---O bonds are nearly parallel (Fig. 2) making an asymmetric structure like formic

Bond lengths 6) Bond angles (deg.)

in AFA. in AFA.

Fig. 2. Structure and structural parameters of aatic formic anhydride. M

anhydride, but differing from that of acetic anhydride, which has a symmetrical structure. Further
torsional vibrations around C-O are wider in the molecule of 1. The origin of these conformational
differences is not clear; resonance and a more favourable intramolecular carbonyl dipole-dipole
interaction might be held responsible for this difference. The resonance energy of 1 has not been
determined whereas that of acetic anhydride is reported to be about 30 kcal mole-. There is no
evidence of any intramolecular 0. . . H bond and the calculated 0 * *. H distance is 2.7 A, which
nearly equals the sum of the Van der Waals radii of hydrogen and oxygen atoms.45 By and large
this molecule shows a stiffer conformation than that of acetic anhydride (Table 2).

Table 2. Structural parameters* for acetic, acetic formic and

formic anhydrides

Aatic
Acetic formic Formic

Bond kngths (A)

1.405(l) 1.397(l) 1.384(3)
1.183(l) 1.195(l) 1.188(2)
C-H - 1.10 1.10

Bond angles (dcgrax)

(C_ 115.8(8) 115.3(S) llS(2)
(o--c=o) 121.7(2) 122.8(4) -
(crc=oY - 118.3(5) 123.4(5)
119(13) 119.1(9)
I=z 108.3(2)
- 108.3

Torsional angle around C-O bond (degrees)

folmyl c-o - 25(m) 17(6)
26(9)
aatyl C-O 48.5(8) 39(24) -

lEsthataI standard deviations in parentheses.

Aatylmoiety.
bFormyl moiety.

The sigmatropic isomerization (migration of the acetyl group from one oxygen to the other) of
1 was theoretically studied by the MINDO/3 method. The calculated enthalpy of activation of the
process is 200.9 kJ mole- and lies at the limits of the scale for tautomeric reactions.*
 Acetic formic anhydride 1085

2. SPECI-ROSCOPIC mDIFS

2.1. Infrared spectroscopy

An accurate determination of infrared absorption bands of 1 and their assignments was per-
formed by Vledder et al. in 1970 on pure mixed anhydride samples. The normal coordinate
calculations were carried out with Wilsons GF formalism (Table 3).

Table 3. Vibrational spectroscopic study of 1 in a Ccl, solution

Frequency Raman
bands IR bands
@n- ) @n-v Assignments

1789 1794 stretching Ck=O formyl

1765 1775 stretching CL0 acetyl
1432 1427 CH, asym stretching oop
1373 1373 CH, sym stretching oop
I200 II94 CH, asymbending ir,
II84 1183 CH Jsyh bendingip-
1046 1038 C-H bendina
928 925 CH flocking ocbp
648 635. COO formyl bending
557 556, OCC bending

l Pure liquid, measured in a 0.012 mm KBr cell.

ip : in plane ; oop : out of plane.

A comparison of the spectra of 1 and its mono and trideutero analogues in the region between
3200 and 2000 cm- led to the conclusion that the formyl C-H stretching frequency nearly fitted
with the methyl mode at 2990 cm- , the formyl stretching was found at 2256 cm- in monodeutero
1. The remaining absorptions are probably due to overtones and combination bands. The region
between 2000 and 1500 cn- is the most characteristic : the carbonyl shows two peaks (1784 and
1779 cn- ) with an unusually small splitting of about 20 cn- considerably less than the value of
about 60 cm. found for symmetrical anhydrides. * This splitting increases to about 35 cn- in the
monoformyl and fully deuterated forms. From the differences noted in the spectrum the authors
concluded that the 1775 cm- band is mainly due to stretching vibration contributions of the formyl
carbonyl whereas the lower frequency was assigned to an out of plane (oop) mode of vibration. The
two deformation frequencies of the methyl group are to be found at about 1378 (symmetric) and
1450 cm- (antisymmetric) ; the antisymmetric should be doubly degenerate. As a consequence the
two bands at 1432 and 1373 cm- were readily assigned to these vibrations; the doubly degenerate
band could not be resolved. The formyl C-H deformation could also fall in this region (precisely
at 1370 cm ). The situation in the next region (1250-1750 cn- ) is complicated because large
interactions are known to occur between formyl and acetyl C-O and C-C bonds : all compounds
(deuterated as well as undeuterated) showed two bands at about I 190 and 900 cm- ; both were
relatively insensitive to deuteration. They were assigned to acetyl C-O, formyl C-O and C-C
stretching respectively. The 1038 cn- band was attributed to the methyl rocking mode by com-
parison with the bands of mono and trideutero compounds. Finally the 557 cn- band should
mainly be due to the C-H oop bending of the formyl group. The Raman spectrum of 1 was also
recorded (Table 3) and no significant differences were observed. The force constants associated with
C-H and CL0 stretching and bending are all quite normal, a fact implying the absence of any
intramolecular hydrogen bond.45
The IR spectrum of equilibrated FAM at 45C in CHCI, (Fig. 3) was recorded by Stevens and
Van Es in 1964.* It is possible to note the above mentioned splitting in the carbonyl region.
1086 P. %RAZZOLINI et d,

100 x
90
80
70
60
M
40
30
20
10

600 1000 1200 1400 1600 1600 2000

-1
cm

Fig. 3. Infrared spectrum of FAM, equilibrated at 45C for 1 h (chloroform).

Regarding the pure AFA, prepared from ketene and formic acid, a spectrum of the reaction mixture
was recorded in CHC13:** a comparison with the spectra of acetic anhydride, formic acid and
acetic acid shows that significant amounts of these compounds are not present in the product. Its
bands are consistent with Vledders spectrum of pure 1.14 We report in Fig. 4 a spectrum recorded
on a neat distilled sample of AFA, prepared according to Krimen,* in the range 200 to 4000
cm- between KBr windows.

100

.
01 --I
4000 3000 2000 1500 1000 560 c.

Fig. 4. The infrared spectrum of AFA, neat, through KBr windows.

2.2. Nuclear magnetic resonance

H-NMR measurements were performed by Krimenao on 1 (prepared according to his own
method) using tetramethylsilane (TMS) as an internal standard; he observed a singlet at S = 5.25
ppm assigned to the acetyl protons and another singlet at 6 = 9.05 ppm assigned to the formyl
proton (neat product). When the product was not pure the following extra peaks could be observed :
6 = 2.05 ppm (CH,COOH), 6 = 2.20 ppm [(CH 3CO)20], 6 = 2.68 ppm (CH,COCl), 6 = 8.05 ppm
(HCOOH) and 6 = 8.85 ppm [(HCO),O]. We report two spectra recorded by Stevens and Van
Es : 22 one of FAM, prepared from acetic anhydride and formic acid, and one of AFA, prepared
from ketene and formic acid. The former (Fig. 5) contirmed the presence of the mixed anhydride
and that of acetic anhydride, formic acid, acetic acid and in addition a peak showed at 6 = 8.77
 Acetic formic anhydride 1087

12 10 8 6 4 2 0
ppm
Fig. 5. H-NMR spectrum of FAM.

ppm (neat reaction mixture, recorded at regular intervals) that could be assigned to the formic
anhydride protons. The latter (recorded on the neat distilled product, Fig. 6) exhibited two peaks
the area ratio of which (1: 3) was in accordance with the structure of 1. This product seemed to be
contaminated only by traces of acetic anhydride and by another substance that was proved to
be formic anhydride. Both spectra are consistent with the values of chemical shifts reported by
Krimen.

I
A . /

12 10 f3 6 4 2 0 ppm

Fig. 6. H-NMR spaztnun of 4FA. I2

La Palme found two singlets at 6 = 8.9 ppm (formyl proton) and at 6 = 2.2 ppm (acetyl protons)
for 1 prepared by ozonolysis of vinyl acetate.8 Olah recorded the PMR spectrum of 1 in superacid
medium at low temperature (Fig. 1) : it showed the deshielded singlet at 6 = 15.74 ppm of the
[=0-H+] proton and the alkyl protons chemical shifts almost unchanged; a singlet at 6 = 9.01
ppm for the formyl proton and the singlet at 6 = 2.21 ppm for the acetyl ones.94 No C-NMR
studies have appeared in the literature to date. We report the C-NMR spectrum of 1 (neat),
prepared following Krimens procedure, Soshowing peaks at 21.07 (CH ,), 157.92 &HO) and 169.27
(CH,CO) ppm (Fig. 7) which can be easily distinguished from those of possibly accompanying
impurities (CH,COOH: 20.52, 178.33; Ac,O: 21.99, 167.48; and HCOOH: 167.09 ppm).
1088 P. tbtMZOLtNI et 01.

169.27

2& l& i ppm

Fig. 7. C-NMR spectrum of AFA. I Recorded as a neat compound; external standard D20, internal
standard TMS. on a Bruker WF%OSY spectrometer working at 80 MHz.

2.3. Other
No structural microwave, X-ray diffraction, neutron diffraction studies were reported. Similarly,
nq UV studies have been carried out.
Theoretical studies (SCF-LCAO-MO-INDO calculations) by Curvale indicate a preferential
almost planar conformation for 1. PES and UPS data are not available.

3. KIMIC WORK

The rate of hydrolysis of AFA in acetone-water mixture and its decomposition into acetic acid
and carbon monoxide in toluene, were studied by Gold and Jefferson.55 The hydrolysis in 80%
acetone-water was 100 times faster than that of acetic anhydride under the same conditions. In the
case of 1 the electronic and steric effects should reinforce each other and therefore 1 should be more
reactive with water and with other nucleophiles. The reaction undergoes catalysis by pyridine. The
uncatalyzed hydrolysis strictly obeys a first order law.
The decomposition reaction of 1 is catalyzed by amines (pyridine, 2,&dimethylpyridine, 2- and
Cpicoline) : it is a first order reaction with respect to 1. The velocity of this reaction was determined
by measuring the rate of evolution of CO from the solution. The order of catalytic efficiency of
pyridine and its homologues differs from the order of their basic dissociation constants or from the
sequence of their catalytic efficiencies for reactions initiated by a rate determining proton transfer
from the substrate to the catalyst. Because of the close similarity of the results for the hydrolysis
and the decomposition, the authors suggested that the catalytic action was similar in the two cases
and that it consists in an association between the amine and the anhydride molecule. The simplest
mechanism in accordance with these observations involves a rate-determining acyhum transfer from
the anhydride to the base followed by breakdown of the resulting complex as shown below. The
second reaction may be the unimolecular decomposition of [BadHO] + or may involve a more
complicated sequence. However, the observed reaction order requires that these cannot be kinetically
significant. The absence of any detectable formation of CO during the catalyzed hydrolysis of 1
indicates that, in 80% acetonewater, the hydrolysis reaction of the mixed anhydride is much more
 Acetic formic anhydride 1089

rapid than its decomposition. No kinetic work appeared to date in the literature concerning either
the rate of formation of 1 under different conditions or the velocity of reactions under catalytic
conditions and with different substrates.
CH,-CO-O-CHO+ B:- [B-CHO] + + AcO - - BH+ +CO+AcO-

4. IRRADIATION CHEMISTRY AND MASS SPECTROMETRY

There is no report about the photochemical behaviour of 1. The mass spectrum of 1 is shown
in Fig. 8.

100 --
13

45

I__
I

I---
60

73 87
J d
) 40 50
56
60 70 60 1 m/z
Fig. 8. Mass spectrum of AFA (70 eV).

5. PREPARATIONS

Acetic formic anhydride was synthesized for the first time from anhydrous acetic anhydride and
formic acid by Behal. This reaction resulted in an equilibrium mixture and the presence of
the mixed anhydride in Behals reaction mixture was determined by means of H-NMR and IR
spectroscopy by Stevens and Van Es. 22The reaction mixture, obtained by heating the reagents for
 Acetic formic anhydride 1091

Many authors have prepared the mixed anhydride 1 simply by equilibrating variable amounts
of formic acid (usually in large excess) with acetic anhydride. This mixture, which we call Formic
acid Acetic anhydride Mixture (FAM, 2). is used as a formylating agent.

6. CHEMICAL PROPERTIES

The reactions involving AFA and FAM are mostly nucleophilic attacks at the formyl carbon
atom by 0, N, S, P and C nucleophiles with the generation of the corresponding formate esters,
formamides, monothioformates, formyl phosphonates and C-formyl derivatives. There are a few
interesting reactions involving organometallic compounds.

6. I .O-Formylations
This reaction has been studied extensively on alcohols, phenols, carbohydrates, hydroxy acids
and polymeric compounds. Formylation of an aldehyde oxygen (9) was reported to give the
corresponding enol formate (yield : 3%) in one single configuration.

6. I. I. Phenols and alcohols. Phenols (Table 5) were found to react very slowly with an equivalent
of FAM at room temperature. After IO days, 30% of the phenol had not reacted. After treatment
with cold 5% aqueous NaOH followed by fractional distillation, the reaction mixture from phenol
and 2 (2 equiv.. room temp., 23 days) yielded a product consisting of phenyl formate (92%). phenyl
acetate (7%) and unreacted phenol (1%). * At higher temperatures (7oC), this reaction loses
selectivity yielding formate (40%). acetate (20%) and phenol (40%) after I2 h.34 Selectivity of

Table 5. Reactionmixture composition (molar percentage) between

phenols and a double amount of FAM at room temperature, after 50
days

Fonnate AOZtate Unnactcd phenol

Phenols (mole %) (mole %) (mole %)

Phenol 92 7 1
2-Methylphenol 94 3 3
3-Methylphenol 91 5 4
4-Methylphenol 93 4 3
3,CDimethylphenol 94 3 3
2.5-Dimethylphenol 96 2 2
2,&Dimethylphenol 93 trace 7

this reaction can be improved by carrying out the esterification in a solution of benzene or carbon
tetrachloride. This procedure gives the formates (80-8S/,) containing only negligible amounts of
the acetates (< I %) and unreacted phenols after distillation under reduced pressure (Table 6). In
the case of the slowly reacting di-orrhosubstituted phenols (2.6dimethylphenol) only traces of
acetates could be detected in the undistilled formate. Stevens and Van Es also prepared Cnitrophenyl
formate using pyridine and Ccl4 as a mixed solvent. Using 2 they obtained a 70% yield.
1092 P. %x~zz.OLlNl el al.

Table 6. Formates from phenols and FAM in benzene

Reaction Yield of
time distilled formate
Phenols (days) (%)

Phenol 30 85
2-Methylphenol 60 80
3-Methylphenol 60 80
CMethylphenol 60 80
3,CDimethylphenol 40 85
2,5-Dimethylphenol 60 80
2,kDimethylphenol 80 80

Although the esterification with FAM is a good method to prepare pure aryl formates in high
yields, it is rather time consuming. Therefore, Stevens and Van Es carried out the esterification with
pure AFA, but the formatelacetate ratio obtained (85 : 15) was lower than that (93 : 7) found when
using twice the equivalent of 2. Addition of tertiary bases and sodium for-mate catalyzed the
esterification with both esterifying agents (2 in slight excess and 1, 20C 24-48 hours) when
practically only formate was formed (yield 99%). The procedure is quite general (Table 7) and
affords an easy and rapid method of synthesis of pure aryl formates in high yields.42 A catalysis
by 4-methylbenzenesulfonic acid @-toluenesulfonic acid) was noted, but the complete reversion of
the reaction outcome was observed: phenvl acetate was the only product of the action of 2 on
phenol (yield 99O/,). The kinetically most active species might be a protonated form of AFA such
as 10.

,0-c
/
Q\,t PhOH )
AFA + HA _ PhOAc + HCOOH + HA
I
H...A

Complications may arise using phenols containing electron withdrawing substituents when their
formates fail to crystallize during the esterification. The formate first formed would eventually
undergo a transesterification and/or decomposition with acetic acid and/or acetic anhydride. On
the other hand a fortunate circumstance is that this kind of aryl formates (Table 7) often crystallized

Table 7. Preparation of aryl formates

By FAM By AFA
Yield Yield (%)
Aryl group (%.) Aryl group no NaHCO, with NaHCO,

2+Butyl4methylphenyl 83 Phenyl 69
6+Butyl-3-methylphenyl 86 2-Methylphenyl 82
a
4-r-Butyl-2-methylphenyl 78 3-Methylphenyl 79
2Xhlorophenyl 77 CMethylphenyl 73 87
CChlorophenyl 74 2-Chlorophenyl D 72
2-Bromophenyl 78 QChlorophenyl L1 79
CBromophenyl 76 2-Isopropylphenyl 65 70
3-Nitrophenyl 79 4-l-Butylphenyl 0 93
4-Nitrophenyl 82 2,CDimethylphenyl D 68
CMethoxypheoyl 74 bA=tylphenyl 87 a
2,4,6_TrichiorophenyI 80 3-Nitrophenyl . 82
2,4,6_Tribromophenyl 78 CNitrophenyl 82
3-Methoxycarbonylphtyl a 8;
4-Methoxycarbonylphenyl . 70

aNot reported
 Aceticformicanhydride 1093

from the reaction mixture, thus preventing similar occurrences. Good catalysts for the reaction
with electron withdrawing substituents seemed to be 2,6-dimethylpyridine (2,6-lutidine), 2,4,6-
trimethylpyridine (2,4,6_collidine) and sodium formate. Another drawback may be due to the
decomposition of formate esters to phenols, during work up of the original reaction mixture, by
distillation at higher temperatures.42 Another and faster variation of the original Stevens and Van
Es method requires the use of pure AFA (2 eq.) and two equivalents of sodium hydrogen carbonate
at room temperature for several hours (Table 7).12 The insoluble salts were removed by filtration
and the solution concentrated under reduced pressure. The pure esters were obtained by fractional
distillation in uucuo or by recrystallization. a- and /I-Naphthols were formylated under these
conditions. 2-f-Butyl- and 4-methyl-2,6-di-r-butyl-phenol did not react because of the steric hin-
drance whereas 2-nitro-, 2-acetyl- and 2carbomethoxy-phenol are equally inhibited because their
hydroxyl groups are hydrogen bonded to the orrho-substituents.20 In a patent dealing with skin-
tanning compositions, 5,6-dihydroxyindole was described to react with AFA in ether (2OC, over-
night) to yield both possible 0-monoformyl derivatives. 84 Isoflavones 11, where the substituents R
represent several aromatic and heteroaromatic groups, were obtained by cyclization of ketones 12
with 1, catalyzed by sodium formate or triethylamine, in a high yield reaction, where the initial step
is expected to be a regioselective 0-formylation. lo2 5,7-Dihydroxyisoflavones 13 with several ring

liOWR - q$R
R 0 if 0

12 R.H 11 R.H

13 R.OH

substituents were synthesized in quantitative yield .I 3-Aryl- or 3-heteroaryl-chromones were

obtained by reaction of some a-substituted 2-hydroxyacetophenones with AFA in the presence of
a tertiary alkylamine. For other examples of cyclizations : see Section 6.6.
Van Es and Stevens14 carried out analogous studies on aliphatic alcohols. The esterification of
tertiary alcohols by FAM yielded the formate esters exclusively, whereas primary and secondary
alcohols produced mixtures containing appreciable amounts of acetate. The velocity of esterification
depends upon the alcohol and decreases from primary to tertiary (Tables 8 and 9).12 The formate

Table 8. Tertiary formate esters from tertiary alcohols and

FAM

Reaction time Yield*

Formate (h) W)

r-Bury1 48 85
2.2~Dimcthylpropyl 48 92
3-Methyl-3-pentyl 168 90
3-Ethyl-3-pcntyl 240 94
2-Methyl-2-hexyl 96 90
l No alcohol could be detected in the reaction mixtures

after the times indicated ; only negligible amounts of acetate

(< I %) were found. The yields refer to distillal products.
FAM was used in 50% excess.

may be difficult to purify by distillation due to the presence of the acetate or unreacted alcohol.4
Temperature changes (IO-50C) affected only the reaction rate. The influence of different molar
1094 P. SnuzzDuNI er al.

Table 9. EsteriIication of akohols with FAM (molar ratio I : 1.I) at room temperature,
by Stevens and Van Es=

Product ratios
Original reaction mixture Isolated products
AlCOhOl formate acetate fomlate acetate

Methyl 82 I8 87 I3
Ethyl 86 I4 92 8
n-Propyl 89 II 87 I3
n-Butyl IO 90 IO
Isobutyl z IO 90 IO
AllYl 87 I3 0 D
97 3 0 0
&nzyl
Isopropyl % 4 96 4
set-Butyl % 4 96 4
Tertiary alcohols* (t 0 199 <I

Not reported.
b r-Butyl (separated yield : 85%). I-amyl (92%), 3-methyl-3-pcotyl (90%), 3-ethyl-3-
pentyl (94%). 3-methyl-2-hexyl (90%).

ratio in order to displace the equilibrium towards mixed anhydride formation led to the exclusive
formation of formates in the reactions of I-propanol and set-butyl alcohol. The most important
single influence appeared to be that of the catalyst. Tertiary nitrogen bases and sodium formate
increase the selectivity in the esterification of primary and secondary alcohols at 20C. The reaction
is accelerated in the order pyridine > imidazole > sodium formate. The use of acid catalysts (4
methylbenzenesulfonic acid) does not lead to good results yielding considerable amounts of acetate.
The authors could not find useful catalysts in the case of tertiary alcohols.4
Glucose monoacetonide was dissolved into AFA at 40C (reaction time: 24 h) to yield the
tetracyclic orthoformate 14. Racemic 1,2,3,4_tetrahydroxybutane slowly dissolved in 1 from which
the corresponding tetraformate crystallized out (60%).48 In the course of the synthesis of the

antibiotic A-32390-A, poly-0-formylation of sugar hydroxyl groups by AFA, catalyzed by pyri-

dineIs was employed. Substrates with the hydroxyl groups masked by acetalization with ben-
zaldehyde were similarly formylated by 1 in formic acid using sulfuric acid as catalyst.
Some nitro-alcohols react smoothly with AFA to yield their formic esters in absolute ether at
about 50C (Table 10). 69 Higher reaction temperatures or addition of sulfuric acid in catalytic
amounts caused the transfer of the acetyl moiety to the substrate, as in the case of the reaction with
phenols.24 It can be seen from Table 10 that the yields of formates are not high except in the case
of 2-nitro-I-butanol. The presence of the acetate is due to the formation of acetic acid during the
esterification procedure and to the consequent equilibrium yielding acetic anhydride. Stevens and
Van Es prepared Cnitrophenyl formate from Cnitrophenol (1 mole) with a slight excess of FAM
(1.5 mole) in presence of pyridine at room temperature. They obtained the formate (70%) after two
recrystallizations from carbon tetrachloride. 24
 Acetic formic anhydride 1095

Table 10. Esterification of qitro-alcohols with AFAb9

Temperat= Tii Yield

Nitro-alcohol solvent (0 Win) PrOdUCt (%)

2-Methyl-2-nitro-I-propanol ether r&lx 480 formate 61

2-Methyl-2-nitro-I-propanol b X%60 30 formate 44
2-Methyl-2-nitro-I-propanol b reflux 60 acetate V-80
2-Methyl-Znitro- I -propanol C refiux 60 acetate 7S80
2-Nitro- 1-butanol d 1 d formate 88
2-Nitro-I-butanol ether retlux 480 formate 48
2-Methyl-2-nitro-l.3-propanediol ether reflux 480 diformate 53
2-Methyl-2-nitro-l,3-propanediol b , , diacctate 89

aCharacterized by conversion to 4-bromophenacyl derivative of the hydrolysis product.

bNo solvent.
No solvmt, one drop of concentrated sulfuric acid added.
Not reported.
Exothermic, instant, vigorous reaction even on coolingwith an ice-bath.

In the essential oils technology a procedure for the determination of the free alcohol hydroxy
function consisted in the reaction of the mixture under analysis with FAM (72 h, room temperature),
acid-base extraction of the reaction mixture and final saponification with 0.5 N alcoholic sodium
hydroxide of the separated formates. The free formate anion formed was then titrated suitably.4*60**6
6.1.2. Hydroxy acids. Gibberellin A3 (15), with AFA (&2C, in pyridine, 24 h) yields

9
/ 15 R-H. R=H

co 16 R&HO. R.tt

RO --OR R=CHO, R=CIiO

17
II
m a
CH
3 cOott CHZ

an interesting 3-0-monoformyl derivative (85%, 16) and 0-diformyl compound (13%, 17).
17-L-Hydroxyoctadecanoic acid was refluxed with FAM in formic acid giving the intermediate
formyl derivative which was then transformed into the acid chloride with oxalyl chloride (62%).
6.1.3. Tefracyclines. A solution of AFA in acetic acid (acetoformic reagent) reacts with tetra-
cycline 18 in cold pyridine to form 0 , ,-formyltetracycline (19, 61%). In boiling toluene the

R R R N(CH3t2 NKH3)2
\/ 1 H t I

011 0 OH 0 OH 011 OH

16 R.CH 3, RcOH. R4. R=OH 20 21

19 R=CH3. RzOll. R.H. R=OCttO

22 R=tt, R=H. R=lt * R-Ott

-

23 R-H. R=lt. R4. R=OCtKJ

24 R&It 3, R=Olt. R=OH. R=Ott

formate 19 is transformed into 4a,l2a-anhydrotetracycline (20) by pyrolytic &-elimination of

formic acid. 19 Also undergoes catalytic hydrogenolysis giving 12a-deoxytetracycline (21, 42%).
1096 P. .%RAZWLlNl el al.

6-Demethyl-6-deoxytetracycline (22) similarly produces 0 , h-formyl-6-demethyl-6-deoxytetra-

cycline (23). 5-Hydroxytetracycline (24) undergoes a more complex reaction with AFA in pyridine
producing a diformate of undetermined structure.**
6.1.4. Ribonucleosides. Reaction of the appropriate anhydride and methoxyalkylidene derivatives
of ribonucleosides (25) in the synthesis of oligoribonucleotides by the phosphotriester approach
afforded the 5-O-acyl-2,3-methoxyalkylidene derivative 26 which was then submitted to acidic
hydrolysis to regenerate the 2- and 3-hydroxy groups. FAM and 2,3-benzytideneuridine (27)
yielded 3-0-benzoyl-5-0-formyluridine (28) after acidic hydrolysis,49 whereas the product of reac-
tion between 2,3-O-methoxymethylideneuridine (29) and 2 was the diforrnyl derivative 3,5-di-O-

ROH2C 9 9 ROH* c B
0

w w
0 0 RCO 0 OH

R
Y OCH
3

25 R=H; 26 R=acyl 26 R=HCO, R*=Ph, B=uracil

27 R=H, R=Ph. B=urecil 30 R=HCO. R=H, B=uracil

29 R=H, R=ll, B=urecil

31 R=H, R=CH30CH2, 32 R=HCO, R=CH30CH ,

2

a B=uracil
b B=6-N-(4-methoxybenzoyljedenine
C B=2-N-benzoylguanlne

formyluridine (30) (68%).* 2,3-O-(Dimethoxyethylidene)uridine (31a), 6-N-(4_methoxybenzoyl)-

2,3-0-(dimethoxyethylidene)adenosine (31b) and 2-N-benzoyl-2,3-O-(dimethyoxyethylidene)-
guanosine (31~) were treated with 2 at low temperature in pyridine to give, after treatment with
95% formic acid, the corresponding 5-0-formyl-3-0-(methoxyacetyl) derivatives 32a (44%),
32b (60%) and 32e (77%). lo4 The selective sensitivity of formate esters towards methanolysis in
comparison with acetates made formylation with AFA a method of choice for the protection of the
hydroxyl functionalities of the sugar moiety of some nucleosides such as uridine (33), 6-azauridine

bH Oil OH OH

33 34 x=0

35 X=NH
 Acetic formic anhydride 1097

(34) and 6-azacytidine (35). Poly-0-formylation occurred when there were free hydroxyl groups.
No N-formyl derivative was isolated after attempted formylation of 6-azacytidine.j
6.15 Hydroxy polymeric systems. FAM has also been used in the preparation of formyl
derivatives of hydroxy polymeric systems. Poly(Cformyloxy)ethenylbenzene (poly-p-formyloxy-
styrene) can be obtained by either of two different routes : polymerization of Cformyloxyethenyl-
benzene monomer or by chemical modification of poly(4-hydroxy)ethenylbenzene chain. Both of
these routes involve the formylation of a phenolic substrate by 2. The latter route is somewhat slower
than the corresponding formylation of the lower molecular weight compound. 4-Methylphenol (p-
cresol) gave ptolyl formate (70%). 3
6.1.6. N-Oxides. The peculiarity of the formyl group circumvents the Polonovski reaction lo3and
other rearrangements. to produce clean and quantitative deoxygenations of N-oxides of tertiary

I?

/\ + aC-OOCCH 3-
CH3COOH + CO
2
R R H
amines using AFA, via the assumed intermediacy of 0-formyl derivatives. Sulfoxides were found
to be unreactive toward 1. 32

6.2. N-Formylarions
6.2.1. Amines and polyamines. Ammonia in dry ether reacted with AFA at 0C yielding the
corresponding formamide. No mention was made of any polyformylation products such as the
diformamide (36) or triformamide (37), which are known compounds..6.

CHO c HO
H-N
/ OK-N
/
\ \
CHO cno

36 37

An efficient and general one pot procedure for the N-monomethylation of primary amines calls
for the formylation of the amino group with excess FAM under mild conditions. The yields are
excellent even for weakly basic and sterically hindered aromatic amines, like Cnitrobenzeneamine,

4-amino-3-chloro-N.N-dimethylbenzenesulfonamide and 2,6-diisopropylbenzeneamine (Table I I).

N-Diformylation was not observed, a fact which is important when the formamides are to be
reduced to N-methylamines.* n-Octadecylamine gave only N-monoformylation at OC in pentane-
ether. whereas some 10% N.N-diformylation product was obtained when 2 was used without solvent
under reflux.
N-Formylation by FAM forms the basis of an analytical method for tertiary amines in presence
of primary and secondary amines.
A large number of N-formylations have been reported on aliphatic and aromatic amines.
2-Amino-45methylenedioxybenzaldehyde was N-formylated to 38 with AFA in dichlorometh-
ane (I h, room temp., 62% yield). FAM was successfully used (I h. 60C) in N-formylating
2-(methylamino)benzaldehyde (39, 40%) and its 2,4_dinitrophenylhydrazone to give 40 (92%).4
The heterocyclic compound 41 (R=H) was formylated with AFA in the patented preparation of the
1098 P. %RAZZOLINI Cl 01.

Table 1I. Formylation of primary amines by FAM according to Krisbnamurthy

Temperature Time Yield

Primary amine ((7 (h) WI

Baucncamine -20 0.25 100

2,6_Dimethylbenzeneamine -20 0.25 loo
2,rXXisopropylbenzeneamine -20 0.25 99
2-Aminophenol -20 0.25 loo
CMethylbenzeneamine -20 0.25 99
CBromobenzeneamine -20 0.25 99
4-lodobenzeneamine -20 0.25 loo
2,6Dichlorobenzeneamine -20 0.25 loo
CNitrobenzeneamine 0 0.25 loo
2-Cbloro4nilrobeneamine 25 3.00 loo
Ethyl Caminobenzoate -20 0.25 97
4-Amino-3chloro-N,Ndimeth lylbesznesulfonamide 25 2.00 99
Benzylamine -20 0.25 99
3-Aminopyridine -20 0.25 99

o@cc &EHO (y&

0
38 39 x=0 41 R=H

40 X=N-NH-(2,4-dinitmphenyl) 42 R=CHO

N-formyl derivative 42 (R=CHO), an insecticide. 64A number of N-unsubstituted oxazolidines and

thiazolidines (43, X=0 or S) were N-formylated in excellent yields with 1: the stereochemical
selectivity was found to be about 3 : 1 (cis, 44/rrans, 49.
C0,R2 C02R 2 CqR2
f-f AFA r-i f+?
X NH ) X NCOH + X NdOH

Y Y
E t,N
Y Rl
0 %
Rl Rl
43 x = 0,s 44 45

Configuration was preserved when one enantiomer of I-phenylethylamine was formylated by

FAM. 2 in the presence of sodium acetate gave a quantitative yield of the N-formyl derivative
by reaction with N-[a-(3,5-dibenzyloxyphenyl)-8-(3,4,5-trmethoxyphenyl)]ethylaminoacetaldehyde
diethylacetal (46) at room temperature. The steroidal amine 3a-amino-(25R)-5a-spirost-9( 1I)-

Ph C $0 OEt
OEt

P hCH20
0 \
NH

RtlNu
4
Ar H
47 t3.H
46 46 R = CHO
 ACHIC
formic anhydride 1099

ene (47) was formylated to 48 with AFA. I4 2-Deoxy-2-amino-/?-D-glucose 1,3,4,6_tetraacetate

(49r) was N-formylated to 49b (94%). The formylatio n of 2,3,4,6-tetra-O-benzoyl+D-gluco-
pyranosylamine (508) by 1 at room temperature (6 h) in ethyl acetate yielded the corresponding N-
formyl derivative 5Ob quantitatively.9 Both ( f )- and meso-l,2-diamino-I ,Zdiphenylethane were

49a R=OAc. R=Nll2. R=Ac

49b RzOAc, R=NIICIIO, R=Ac 51 II.11

500 R=NII2. R.OOCPh . RC-COPh 52 R=CHO

5Ob R-NIICIlO. R=COCPh, R=COPh

bis N-formylated by 1 in 60% yield.14 AFA was also successfully used in the N-performylation of
aminomethyl-polystyrene (51) to 52 in dichloromethane in the course of the synthesis of iso-
cyanomethyl-polystyrene resins. The reaction was performed at room temperature for 72 h with
quantitative yields. I9 The weakly basic 2-aminobenzonitrile was formylated by FAM (0. I6 h, 60C.
77%).7 Analogously, 2,6-dichloro+nitroaniline was formylated (64%) with 2. A very bulky
group in o-position did not exclude the formation of the formamide 53 (98%). The aromatic
amine 55 gave its N-formamide (56) with AFA.92

CH NHR
3

NllR CH3

53 R=CHO 55 R=H

54 R=H 56 R-CIIO

An extensive literature exists for heterocyclic compounds. Side chain N-formylations and ring
N-formylations have been reported. 2-Amino-3-picoline reacted with FAM giving N-formylpicoline
(82%). 29 3-Aminothiophene, 3-benzylaminothiophene, 3-methylaminothiophene, 3,4_diamino-
thiophene and 3+di(benzylamino)thiophene were N-formylated by 2 in excellent yields.97 2-Acyl-
3-amino-thiophenes and -selenophenes (57) were N-fonnylated with AFA,3 whereas a ring sub-
stituted 2-thienylamine (58) could be N-formylated to 59 (83%) by FAM.

NH2 CN

1/1
I I
X
CCR NIIR

57 58 R-H
x4, se; R=H, CH 59 R&HO
3

I-Benzyl-6-aminoindazole (60) in DMF was formylated by AFA producing the 6-formamido

derivative (61).6 2-Aminobenzothiazole (62) in anhydrous ether was formylated by 1 giving 63
(70%).28
1100 P. .%RMZDLINI el al.

I
CH2Ph

60 t?.II 62 R-H

61 R.cilo 63 R=CttO

An uncommon instance of either formylation or acetylation which depends upon the substrate
is shown by a 2-aminothiazole derivative and its N-oxide. When exposed to an excess of FAM, 2-
amino4ethoxycarbonylmethylthiazole (64) undergoes formylation producing the N-formyl deriva-
tive 65 (91%) whereas the corresponding N-oxide 66 in dry chloroform at - 10C yields the

RNT-- N
co Et
2
RtiNr- co2m
0

64 R:tI 66 R:II

65 R.CIIO 67 R=Ac

N-acetyl derivative 67 (65%) exclusively.Y9 FAM (83C. 4 h) reacted with 5-aminothiazole-4-

carboxamide (68) to yield the corresponding 5-formamido derivative (69, 95h).29 In the absence
of solvent the reaction between 5-amino-l,2,4-dithiazole-3-thione (70) and AFA gave 5-N-formyl-

CONlIZ RIIN

NltR

68 R-II 70 W:ll 73 R=II

69 R=CtIO 71 R=CIIO 74 R.CHO

72 R=Ac

amino-l ,2,4-dithiazole-3-thione (71, 85%). In pyridine this reaction yielded a mixture of the 5-N-
acetyl (72, 80%) and the 5-N-formyl (71, 6%) derivatives.
2-Methyl-l,3,4-thiadiazolin-5-thione (73) was formylated by FAM to yield 74 (75%).lS6 2 In
excess was used in the formylation of a number of heterocyclic amines (Table l2).6R

Table 12. Formylation of heterocyclic amines with FAM at room

temperatureb

Yield
Amine Product (Oh)

2-Aminoquinoxaline 2-Qumoxalinoformemide 91
2-Aminothiazole 2-Thiazolylformamide 64
2-Amino-5-nitrothiazole 2-(5-Nitrothiazolyl)formamide 84
2-Aminobendmidazole 2-Eknzimidazolyfonnamide 83
2-Aminobenzothiazole 2-Bcnzothiazolylformamide 94
 Acetic formic anhydride 1101

N-(7-Chloro-5-phenyl-3H-1 ,4-benzodiazepin-2-yI)-N-methylhydroxylamine (75) incorporates

both acyl residues of FAM (16 h, room temp., 32% yield) at two different sites yielding the N-
formyl-3-acetyl derivative 76 (32%). The sites of acylation were confirmed by performing the
reaction on 3-acetoxy-7-chloro-2-methyhunino-5-phenyl-3H- 1,Cbenzodiazepine (77) which pro-

5 'N-R

75 R=OH, R:H

76 R=CHO, R=OAc

77 R.H. R=OAc 79 R-H

78 R:CIIO. R=OAc 81 R=CHO

duced the N-formyl derivative 78.OvN-(2-Methylaminobenzoyl)tryptamine (79) yielded N-[2-formyl-

N-mcthylamino)benzoyI]tryptamine (81, 85%) and 13h,l4-dihydrorutaecarpine (80) yielded 14-
formyl-136,14_dihydrorutaecarpine (82, 96%) when treated with AFA at room temperature. The
product 83 undergoes ring closure followed by N-formylation giving 82 (76%). by treatment with

80 R=ll
84 RX110

1 at 2560C. It can be seen that the less basic heteroaromatic nitrogen is prevented _. from being
formylated, although traces of the ring formylated derivative 84 could be detected.
Some indole derivatives (85, 86, 87 and 88) were treated with AFA or FAM, in the dark.
C-Formylation in the 3-position occurred, whenever this position was free (8586 and 87), but the
primary reaction product then condensed with the original substrate yielding products of the type

m
:I IR N
I
R
R
\
I
R
/3

85 RLII . R=II. R.ll 89 R=Il. R=ll

86 RXII 3, R=Il. R=II 90 R.CH3. R=ll

87 R=ll. R=Cll3, R.H 91 R=II. RXll3

88 R=II. R=ll, R=CII

3
96 RXIIO. Rzll, R.ClI
3
1102 P. %RAZZOLlNl 6-1 ai.

92 R&HO 94 x=1. Y=2

93 R=H 95 x-2. Y=l

89, 90 and 91. In the case of indole (85) dimerization and N-formylation gave 92. Product 92 was
also obtained by direct N-formylation with 1 of 2-(3-indolyl)indoline (93, 80%). If this reaction is
performed at higher temperature (SOC), products of mono- and di-N-formylation of 89 (94 and
95) were also observed. 3-Methylindole (88) underwent N-formylation giving 96 (28%).*
Ethyl 2-ethoxycarbonyl4piperidinepropionate (97) in ether was formylated giving ethyl 2-
ethoxycarbonyl-I-formyl-4-piperidinepropionate (98, 79%) by FAM. The bicyclic amine 99 (2-
benzoyl- 1,2,3,4,5,6-hexahydro- 1.Smethanopyrido[2,3-clazocine) gave the N-formyl derivative 100

CH2CH2C02Et

I
CO2Et
R H3C CH3
it

97 R-H 99 R=H 101 R-H

98 RXIIO loo R=CHO 102 R&HO

(58%). 4-(Diphenylmethyl)piperidine was formylated by 2 at 25C producing I-formyl4(diphenyl-

methyl)piperidine (71%). I4 Formylation of piperidine4carboxylic acid was successfully achieved
(76%) by 2 without any solvent.* This experiment shows that a free carboxyl function does not
interfere with the N-formylation reaction. A regiospecific formylation by FAM took place with
2,2,5,5-tetramethyl4thioxoimidazoline (101) leading to I-formyl-2,2,5,5-tetramethyl4thioxo-
imidazoline (102, 86%).* 7-(2-Chloro-6-fluorobenzyl)amino[ 1,2,5]thiadiazolo[3,4-d]-pyrimidine
(103) reacted with 2 yielding its 7-N-formyl derivative (104,91h).65

.CHO

CHO
103 R.H

104 R.CIIO 105 106

The tetracyclic structure of hexamethylenetetramine (105) was cleaved by FAM yielding the
compound 106 (68%). Similar behaviour was exhibited with AFA by diazamantanone 107 yielding
the N,N-diformyldiazabicyclononanone 108 (42%). Nitrotriazaadamantane 109 gave the
nitrotriazabicyclononanes 110 and 111. *
 Aoztic fomic anhydride 1103

110 R=CHO

107 108 t 09 111 R=Ac

2-Furylamines were N-formylated with FAM (60C 0.5 h, 58-71%) and this was followed
by subsequent reactions of the initial formamides. Reaction of some 2-aminopyrroles gave the
corresponding formamides, which might react further giving cyclic products.
Reductive 1,4_diformylation of pyrazine (112) to 1,4-diformyl-l+dihydropyrazine (113, 28%)
was accomplished by the combined use of AFA and zinc powder. The mechanism of the reaction
was shown to be rather peculiar and different from the electrochemical reductiondiacylation of
112. Zinc metal reacts first with 1 yielding the corresponding radical anion (not directly observed),
which eventually transfers one electron to pyrazine. The newly formed radical anion (114) then
reacts with 1 to yield I-formyl-l,4-dihydropyrazin-4-yl radical (115), which will be reduced to the
corresponding anion 116 by zinc and finally formylated to the final product 113.

112 114 115 116 113

Formylation of all the ammo groups took place in the reaction between AFA and some
polyamines. cis,cis-1,3,5Triaminocyclohexane, 1,2-bis(2-aminophenoxy)ethane and 1,2-bis(2-
amine-4-ferr-butylphenoxy)ethane8 yielded the corresponding N-performylated derivatives respec-
tively in 65%, 99% and 87% yield.
As it was suggested from the reactions described above of some amino derivatives of indole
compounds, the heterocyclic nitrogen is much less reactive to formylation. The amino group of
methyl (Z)-cr-amino-6-methylindole-3-acrylate (117) was exclusively formylated by 1 to 118 (82%)

c$q&=LezcH3 q-JJIYnR

A
II9 R.H. R=H

117 tt=H 120 R&HO. R.41

118 R-CHO 121 R=CHO. R=CHO

at room temperature in tetrahydrofuran.66 Selective formylation at the strongest basic centre of

4-amino-5-aminomethyI-3-benzyl-l,2,3-triazole (119) was observed affording 4-amino-3-benzyl-5-
formamidomethyl-1,2,3-triazole (120,80%) with an excess of AFA in dry pyridine at 22C. The same
compound exhibited different behaviour yielding the diformyl derivative 3-benzyl-4-formamido-5-
1104 P. SrrUmLlNl et al.

formamidomethyl-1,2,3-triazole (121,60%) when added to 1 at 20C4 The last step in the patented
preparation 98 of an antiallergic drug is the formylation of 5-anthranoylaminotetrazole (122) with 1
giving 123. Three 6,7_disubstituted I-alkyl-3,4-dihydroisoquinolines 124 reacted in their tautomeric
form (enamine) with 1 to yield Z and E formamides 12S9 Another instance of imine-enamine

@;;;<;q 2
Q.cHo
FA_
122 R=H

123 RXHO 124 R=H. CH 125

3

isomerisation under the conditions of N-formylation by AFA is given by the preparation of the
enamide 126 (99O/,) from the 3,4-dihydro-6,7-dimethoxy-1-(2,4,5-trimethoxybenzyl)isoquinoline
(127).62 In the one pot reaction of ketoximes with titanium(IV) acetate and 1 in DMF, the final
outcome is the interception of the corresponding N-formylenamine 128, which is stable in two

127 126 128 (E- and Z-l

rotameric forms under these conditions. Nine successful examples of this reaction were offered with
yields from good to excellent. 3
The first synthesis of a stable salt of the type HCOXY , where X is an onium ligand, was
recently reported. The reaction product between a trimethylsilyl halide or trifluoroacetate and 4-
dimethylaminopyridine was put to react with AFA to yield (70-90%) N-formyl-4_dimethylamino-
pyridinium salts. Traces of the pyridine caused rupiddecarbonylation whereas the ion only slowly
decomposed to CO under the same conditions. This points to a dual mechanism for the reaction.
6.2.2. Amino acids. Amino acids of widely different nature have been N-formylated either with
AFA or FAM. N-Formylation of amino acids has been achieved using 2 in good yield, without
racemization of the parent amino acids. However, this procedure has the drawback that an excess
of reagents (acetic anhydride and formic acid) is necessary because the real formylating agent seems
to be the mixed anhydride and its concentration is determined by the equilibrium constant.507*49
Muramatsu therefore suggested the use of pure 1 in the formylation and this method required only
a two or three molar excess of formylating agent. Even when acetic acid was used as the solvent,
formylation occurred and acetylation did not. Amino acids, as well as the alanine ethyl ester, were
formylated by this method with retention of configuration in good yields (7@-90%, see Table 13).90
The N-formylation of glutamic acid (129) was performed by several authors.2J.37.90.27 The
procedure described by Borek and Waelsch for L-glutamic acid gave 72% yield. In the same paper
there is the analogue preparation of N-formyl-rglutamine (130.62%) and of N-formyl+glutamic
acid 5-benzyl ester (131, 80%).25 Glycine was quickly formylated by AFA at OC (1 min, 94%
yield). 26 Racemic 6chlorotryptophan (132) was formylated by 1 in tetrahydrofuran to 133 (100%
 Acetic formic anhydride 1105

Table 13. Formylation of ammo acids with AFA in excess

Time Yield Molar excess

Amino acid (min) Solvent (%) of 1 (%)

Glycine 120 formic acid 85 200

Glycine 120 acetic acid 76 200
Glycine 1 no solvent 94 254
L1
ot_-Alanine formic acid 94 200
L-Valine 0 277
formic acid 84
L-Leucine 0
formic acid 70 200
L-Methionine 0
formic acid 77 200
L-Phenylalanine 0 83 88
formic acid
t_-Tyrosine 120 formic acid 84 200
Y 27
L-Glutamic acid formic acid 35*
Y 271
/?-Alanine formic acid 79
DL-Alanine ethyl ester 180 formic acid 85 93
t_-Alanine methyl ester 180 chloroform 73 (r
d 0
DL-['NjAlanine methyl ester chloroform 9

Not reported.
*Better yields are reported in literature by the usual FAM method.~
(With equimolecular amount of triethylamine.
Overnight.

yield, 20 min at 20C). loo Amino acids containing hydroxyl groups such as tyrosine (yield not
reported) 49 and 3,4-diacetyloxyphenylalanine (86~)23 under mild conditions were formylated by
FAM. Basic amino acids required a special procedure for the selective formylation of one of the
two amino groups: however this method did not use 2. 93 In two other reports D- or L-lysine
hydrochloride was treated with FAM or AFA to yield the n-formyl derivative. The regioselectivity
is remarkable. Sodium salts of carboxylic acids function as catalysts (yields : 92%). 07.08

129 R-OH. R=H

130 R=NH2, R=CHO 132 R-H

131 R=OCH2Ph, R=CHO 133 R.CIIO

Amino acids were N-formylated using a large excess of FAM at 0C. then at 50C during 15 min
under argon at 4C. N-Formylleucine was prepared (60%),15 DL-cystine was N,Ndiformylated by
a modified procedure (55%, 36 69%) which amounts to the in situ formation of formic acetic
anhydride : the product was recrystallized from water without undergoing hydrolysis. An improved
procedure (87%) used treatment of cystine only with formic acid. Treatment of DL-N,N-diformyl-
cystine with Zn and formic acid afforded DL-N-formylcysteine (40h).50
N-Formylhydroxyaminoacetic acid (hadacidin) was obtained from hydroxyaminoacetic acid and
FAM at room temperature (no yields are reported).* In a different context another hydroxylamine,
namely 4-(4chlorophenoxy)phenylhydroxylamine, was formylated with 2.*
The N-tosyl group could be quantitatively removed from N-benzoyl-L-hystidine (134) with
AFA and a catalytic amount of pyridine in 3 h to give 135.This reaction is believed to involve the
1106 P. STrUzzOLlxl el al.

UK , OHC, R
N
NI R $5 13 R -
135
(7 N (7 (7
N
:0,t p-blyl) dCcp~~Oz~p-tOlyll + ACO-sO*(p-tolyl)

134 RXII CHNHCOPh 136

21
COOH

formylation of the other ring nitrogen and the subsequent elimination of acetic tosyl anhydride
(136).* The fonnyl group has been used extensively as a blocking group in the synthesis of
peptides. *49The formyl protective group can be selectively removed without cleavage of peptide
linkages by acid hydrolysis with a slight excess of 0.5 N hydrochloric acid in methanol either at
room temperature for 48 h or by refluxing for 1 h. These peptides retain a high optical purity
(85-90%) because the formyl group can be introduced without racemization of the parent amino
acid. The N-formyl group is relatively stable towards basic hydrolysis. Saponification of esters of
amino acids in aqueous dioxane gave the corresponding N-formylamino acids in yields consistently
better than 85%. This provides a choice between extending the peptide chain either at the amino or
at the carboxyl ends.
The acylation of kynurenine (137), a compound formally exhibiting an aliphatic and an aromatic
primary amino group, using FAM showed a peculiar dependence upon reaction conditions and the
molar ratio HCOOH/Ac,O. When an equimolecular amount of reagents was used and acetic
anhydride was added to a solution of 137 in formic acid the selective acetylation of the aromatic
amino group (138) occurred without affecting the aliphatic group. In contrast using twice the
amount of acetic anhydride then diacylation occurred and the resulting product contained one acetyl
and one formyl group (139). The aromatic amino group was acetylated whereas the aliphatic group
was formylated.32 Formylation of phenacylglycine, (2-nitrophenacyl)glycine, tryptophan and
phenacylalanine did occur under the same conditions (I-ICOOH/A~O 1: I), whereas 2aminobenzoic
acid underwent mainly acetylation. N-Acetylation in the above reactions was rationalized as being
due to a more rapid reaction of acetic anhydride with the aromatic amino groups than with formic
acid. If acetic formic anhydride is allowed to form, by adding the substrate to the reaction mixture
after a long enough period, then only formylation occurred. Using these modified conditions N-
formyl-DL-kynurenine (140,90%) was obtained. Further additions of FAM gave the N,N-diformyl-
kynurenine (141). 32Low yields of ethyl 2-cyano-2-formamidoacetate (142, 38%) were obtained by

OOEt
COC2(iHNHR
COOH CH-CN
I
NHR NHR

137 R=H. R=II 140 R=H, R=CHO 142 R=CHO

138 R=ll. R=Ac 141 R=CHO, R.CHO 143 R-H

139 RXHO. R:Ac

the formylation of ethyl 2-amino-2cyanoacetate (143) by AFA.67 Some a-aminomethyl esters were
N-formylated in CHCI, with triethylamine and a stoichiometric amount of 1 (3 h, room temp.).
Typically, alanine methyl ester was formylated (93%).73
Formylation of penicillins by 1 has been achieved. All the following substituted penicillanic (144
and 145) and penilloic acids (146) gave products of N-formylation from low (- 70C) temperature
reactions : sodium (6R)-6-formamidopenicillanate (147), benzyl (6S)-6-formamidopenicillanate
 Acetic fornk anhydride 1107

RHNCtl2

II

144 R=H, R=SiKH313 I47 R=CHO. R=Na 146 R:H

145 R-H. R= Ctl2Ph I48 R=CHO. R= CH2Ph 149 R=CHO

(148, 71%) and (2R,4S)-5,5-dimethyl-2-formamidomethylthiazolidine~~r~xylic acid (149,

4O%).O
6.2.3. Amino alcohols and amino phenols. The amino group is usually by far more reactive than
the hydroxyl group towards anhydrides. It is therefore usually possible to obtain the formamide
from an amino alcohol.
A comparative study of the N-formylation of some sympathomimetic amines, 150, 151,152,153,
154and 155 (hydroxyamines), using formic acid, formamide and FAM was accomplished. Complex
mixtures were obtained in all cases, but the N-formyl derivatives could be separated.44 The results
are said to be best in the case of 2. 151 Apparently gave the N.C-diformyl derivative 156 on the
basis of weak evidences. This requires confirmation.

150 151 152 153 154 155 156

R (X3 H M3 CWH3b2 Ct13 H

R at 3 Cli aI H
3 3 CH 3 H
R CM (3H H ON Ii H

R H H H Cl 0CH3 H

R H H CH H H OH

Formylation of hepta(N-methylamino)-fi-cyclodextrin (157) or hepta(N-ethylamino)-Bcyclo-

dextrin (158) with FAM (at 0C for two hours and then at room temperature for a further two
hours) afforded the hepta(N-methylformamide) (159) and the hepta(N-ethylformamide) (160)
respectively. The formyl group can be removed adding NaOH (pH = 12) : no yields are reported.26

157 R=CH3. R=H

158 R=C2H5, R=H 161 R=ll 162 R=H, R-H

159 R=CH3. R=CHO 163 R=CIlO 164 R=CIIO. R.II

160 R=C2115. R=CHO 165 R&HO. R=CIIO

1108 P. %RAZZOLINI er al.

Different results were obtained in the treatment of the two closely related 2-aminobenzyl alcohols
161 and 16274 with AFA. The former, treated with I in hexaneether at 0C (1.5 h), gave the
corresponding formamide (163. 90%). The latter was treated with 1 in the presence of pyridine at
0C (I 8 h) : the reaction mixture contained products of N-formylation (164) and N,O-difotmylation
(165). These results are not easily rationalized and an even more confusing pattern arises from
experiments with salicylamide (166), where the nature of the substrate and the reaction conditions
may lead to cyclization and the incorporation of three acyl groups from AFA. 34The heat sensitive
0-formylsalicylamide (167) may be obtained by the action of 1, using pyridine as a catalyst (53%
yield). By simply raising the reaction temperature to 20C (using sodium acetate or very small
amounts of pyridine) the product outcome changed to 2-formyloxy-3-formyl-2,3-dihydro- 1,3-
benzoxazin-4-one (168, 41%). In the same conditions, but with larger amounts of pyridine, the 2-
acetyloxy derivative (169) was isolated (62%). Analogously, the 2-acetyloxy-3-acetyl derivative
(170, 37%) was obtained starting from N-acetylsalicylamide (171) : a collateral product formed

166 R=H. R=H

167 R.H, R=CHO 168 RXHO. R=CHO 174 R-H

171 R=Ac, R=H 169 RXHO. R=Ac 175 R-CHO

172 R=Ac. R=Ac 170 R=Ac. R=Ac

173 R=CHO, R=H

in equal amount was N,O-diacetylsalicylamide (172). Compound 167 rearranged quickly to N-

formylsalicylamide (173) which, in turn, could be transformed to 168 by AFA and pyridine below
5C (75% yield).34
6.2.4. Sulfonamides. It was found that AFA in acetic acid and chloroform at room temperature
was successful (yields ranging from 71 to 88%) in N-formylating N-phenyl aromatic and aliphatic
sulfonamides. However, this reaction surprisingly failed with N-methyl derivatives.33
6.2.5. Phosphonamines. The secondary amino group of a phosphonamine 174 was formylated at
0C (95 h) to 175 in the last step of a patented preparation of a pesticide.42

6.3. S-Formylations
Few results have been reported on the S-formylation of thiols. One is perhaps of lesser applicative
interest because formyl fluoride was the formylating agent.95 S-formylation was observed when
thiophenol or methanethiol were treated with excess FAM followed by basic work up and dis-
tillation : no yields were reported. *

6.4. P- Formylations
The sole example of formylation of organophosphorous compounds is reported by Vasella.
Dimethylphosphite (176) and sodium hydride reacted with AFA in dry ether at - 10C to 20C for

0\\ ,OC 0~ ,OCH

N&i, 3 AFA
- P 3
-AcONs CHQ CIIO
412 CH 0 e Na@
3

176 178 177

 Acetic formic anhydride II09

45 min, yielding distilled formylphosphonate dimethyl ester (177, 47%). The strong base NaH
extracted the hydrogen as a proton from 176 to provide the active nucleophile 178.43

6.5. C-Formylations
Only a few cases of C-formylations have been reported. Indole derivatives react with AFA and
FAM undergoing C-formylation in position-3 as described in Section 6.2.1 .20
The enamine (179) is acylated at C2 by AFA (180, 58%) : another product is the 3-formylamino

H ,CHO OIICHN, H
ZN\c=c -
zN,
c=c + EC
H3C CONHCH H3C CONHCH H3C CONHCII
3 3 3

179 180 181

derivative (181. 19%). 1 was found to be effective in the C-formylation of a-benzoylamino

acids in the presence of TEA and 4-dimethylaminopyridine. The reaction is assumed to follow the
mechanism reported below :

Coo CHO 0
I tl 0 I F
R-CH - 2_ R-X-COOH -CO?,R-C
I I I
NHCOPh NtlCOPh NHCOPh

Ph Ph

R = CII C*l15. Kll3pIctl*

3

The reaction of AFA and aromatic Grignard reagents3* (182) was of particular interest because
it gave only carbonyl products (183) without alcohol formation. The aldehyde was one of the two

Ar
Ctyo-c\
/ + Ar-Hg-X _) CH3COcLC-H -AcOMgX
c
)
\o A-C+
0
[ &X 1

possible carbonyl reaction products in every instance. Success in the aldehyde synthesis appeared
to depend principally on two factors : (i) preferential addition of the Grignard reagent to the formyl
group rather than to the acetyl group and (ii) a good degree of stability of the resulting complex
(184). The reaction was performed at low temperatures (- 70C). Aldehydes predominated decidedly
over ketones when diethyl ether was used as solvent (see Table 14). This predominance may be
attributed to the electronic and steric differences between a hydrogen atom and a methyl group
which favoured nucleophilic attack by R - on the formyl carbon. The aldehyde/ketone ratio dropped
sharply when tetrahydrofuran replaced diethyl ether, but without significant change in total quantity
of carbonyl products. When aliphatic Grignard reagents were used there was substantial alcohol
formation.3
C-Formylations by AFA in THF with N-butyllithium at low temperatures (- 10C then cooled
to - 78C) has been reported for a number of cyclic esters. Excellent yields of rx-formyl lactones
(18-a) were obtained from 1854b. For other cases, as for the substrates 189b and 19Ob, lower
yields were ascribed to an insolubility of the enolates. This drawback was circumvented in part by
1110 P. ~AZZOLINI et at.

Table 14. Reaction between aromatic Grignard reagents and pure AFA

Products molar ratios

Molar ratio Temperatux Aldehyde Ketone
AFA : RMgBr Solvent (Cl (%I W)
Pllcayl l:I ether -70 82.9 17.1
2-Methylphenyl I:1 ether -70 97. I 2.9
3-Methylphcnyl I:1 ether -70 97.1 2.9
CMethylphenyl I:1 ether -70 97.7 2.3
3-Chlorophenyl I:1 ether -70 100.0 0.0
Phenyl I:1 THF -70 53.5 46.5
Phenyl 2:1 ether -70 85.2 14.8
CMethylphenyl 2:1 ether -70 97.8 2.2
Phenyl I:1 ether 0 84.0 16.0

co
R
GR

OJ
+

20 *0
0 0
ll?ia,b lt%o,b 187a.b 186e.b

GR GR

a R = CHOH
m 0 +I m 0 *0
b R-H2
189a.b 190e,b

reacting the lithium enolate of Mb, but the yield was not satisfactory and 19Ob did not react.63
Formylation of 1-(l-methylcyclopentyl)-2-(triphenylphosphoranyliden)ethanone (191) in toluene

I$ CH=N H213-C_C9%
o//- 2
o/ ai
3

191 R=H

192 R-CHO 193 194

by 1 yielded the aldehyde (192, 40%).79 Carbonyl-C-diazoacetaldehyde (193) was obtained from
C-formyl-AFA and diazomethane in unreported yields. IS8 Analogously, 2diazopropanol (194)
was obtained in 56% separated yield by reacting diazoethane at - 30C with l.59

6.6. Cyclization reactions

Substituted phenols (195, see also Section 6.6.1.) were cyclized at room temperature using AFA
to chromones (196) according to the general reaction :

1% 196
 Acttic formic anhydride 1111

Aroyl chromones, O6 3-nitrochromones, 3,6_dinitrochromones, 3-acetylchromones, 3-benzoyl-

chromones, 3-methylsulfinylchromones and 3-methylsulfonylchromones were obtained in fairly
good yields. . 6 Quinazolinocarboline alkaloids, such as rutaecarpine, were synthesized by cycliz-
ation with 1, as already discussed in Section 6.2.1 .2 5-Amino- 1,2,3-triazole-4carboxamidine (197)
reacted with labeled Cr,,rmY- AFA at 70C giving 8-azaadenine-2-14C (198, 77%). The 14C label
was incorporated into the 8-azapurine by ring closure of 197. 89This ring closure was also performed

NH
I 2
HNSc

197 fr=n 196 R-II

199 R=/+-D-Rlbofuranosyl 200 R=p-D-Rlbofuranosyl

on 5-amino- 1-/?-D-ribofuranosyl- 1,2,3-triazole-4carboxamidine (199) with acetic [ C]-formic anhy-

dride, prepared according to Krimen general procedure, yielding 9-(@-ribofuranosyl)-8-aza-
adenine-2-14C (200, 81%). I9 Ring closure with concomitant N-formylation occurred when two
suitably acylated thioamides (201) were treated with FAM to yield 5-formamidothiazole-4carbox-

CONH2
,NH-CH
R-C+ C-NH2 R-($ONH2 a;, CH3v$
0 8

201 R=H. CH 202 203 204

3

amides (202).The 2-aminothiophenol203 undergoes cyclization to 6-methoxybenzothiazole (204,

49%) by reaction with AFA at 5C in water.40 2-(Acylmethylthio)naphtho[ 1,Zd]imidazoles (205)
could be N(3)-formylated with FAM in excellent yields to 206: these products and similar N(3)-

205 R=H

206 R-CHO R=CH 3. Ph. aryt 208 R-H. CH3 ZH5

207 R=Ac, C2115C0

acyl derivatives (207) underwent cyclization when refluxed with sodium acetate and FAM in a I : 2
molar ratio to the corresponding naphtho [1,2: 4,5]imidazo[2,1-h]-thiazoles (208, 67-97h).76 AFA
and ethyl 3-hydroxy- I -methyl-j&oxo- 1H-indole-2-propanoate (209) yielded both the cyclization
product (210,40%) and 211 (27%). Cyclizations involving salicyclic amide were discussed under
Section 5.2.3. 34
1112 P. Smzzm.1~1 el al.

209 210 211

6.7. Formylarion of metal complexes

AFA was used as a formylating agent of organometallic complexes of transition metals. Acyl
complexes are well known in the synthesis of organotransition metal derivatives whereas the cor-
responding formyl compounds are rare. The first kinetically stable neutral formyl complex was
isolated and characterized by Collman : the formyl tetracarbonylferrate ion (212) was formed by
the reaction of sodium tetracarbonylferrate (213) with 1 in THF at room temperature (25C) under
argon. The reaction of acetic [Cl-formic anhydride with sodium pentacarbonylmanganate (214)
proceeds rapidly at OC to give 3CO-substituted pentacarbonylmanganese hydride (215) as the
main reaction product and the decacarbonyldimanganese 216 as another product.4s An unstable
formyl complex (217) was the intermediate on the way to 215. By using labeled 1 the authors
demonstrated that the reaction pathway involved the effective nucleophilic attack by 214 generating
the unstable formyl complex 217. This, eventually, underwent decarbonylation giving the
corresponding pentacarbonyl hydride (215) which finally gave the neutral dimer 216.45 Quantitative
mass spectrometric analysis of the reaction products with 3Cro,,,,y,-AFAand the efficiency of C-
incorporation gave results consistent with this pathway. I3

(Na)[Fe(CO),]- 2 (Na+)[HOC-Fe(CO),]-
- AcONa
213 212

(Na)[Mn(CO),]2~~[HO*C-Mn(CO),]~HO*C-Mn(CO)4~Mn2(*CO)2(CO),
I
214 217 215 216

Oxidative addition of AFA was examined further in the transition metal series rhodium,
ruthenium, iridium, palladium and platinum triphenylphosphine complexes in benzene at room
temperature. Although the formyl complexes may be formed as unstable intermediates, no neutral
formyl complex could be isolated and metal carbonyl hydrides or their dimers were formed.
Platinum and palladium complexes did not undergo oxidative formylation whereas Rh(PPh3)3CI,
Ru(NO)(PPhJ2Cl with an equimolar amount of 1 yielded the monocarbonyl complexes : Rh(C0)
(PPh3)2C1 and Ru(CO)(NO)-(PPh3)2CI.33
Two equivalents of AFA at 25C added rapidly to the square-planar eras-[IrCI(CO)LJ (218.
L-PPh3 or PMe,Ph) leading to the cis-dihydrido-frans-fertiary-phosphino complex IrCIH2(CO)L2.
The formyl metal product resulting from simple oxidative addition of 1 was not observed. Initial

oc
L

L7 II

L
Cl
l AFA _ ocjy?_
,Jc,+c3_
co

218 219 220

 Acetic formic anhydride 1113

interaction involved addition of 1 to the rrans-iridium complex 218 giving a formyl acetato irid-
ium(Il1) complex (219). This species then underwent a rapid internal rearrangement eliminating
acetic acid and producing the dicarbonyl species 220. 39

6.8. Miscellaneous reactions

AFA reacts with anhydrous HF to yield formyl fluoride (68% separated yield) accompanied by
variable amounts of acetyl fluoride.95 Olah and Kuhn found that only acetyl derivatives were
the outcome of the reaction between 1 and aromatic substrates in the presence of dialuminium
hexachloride : the observation was not accompanied by experimental details. Evolution of carbon
monoxide accompanied the reaction. Later, Edwards and Sibille confirmed the absence of alde-
hydes in Friedel-Craft reactions between 1 and aromatic as well as heteroaromatic substrates. These
results may be rationalized in terms of a preliminary complexing of the more electron rich carbonyl
oxygen, resulting in the activation of this group in the acyl transfer.

6.9. Halo- and thio-derivatives

A survey of the published literature did not reveal any thio- or halo-derivative of AFA. It was
found that an equimolecular mixture of anhydrous formic acid and trifluoroacetic anhydride acted
both as a N-formylating and trifluoroacetylating agent (Table 15). * The results may be interpreted
only on the basis of the formation of the mixed anhydride. The electron withdrawing properties of
the CF, group provide an important influence upon reactivity. This anhydride was found to be
thermally labile (loss of CO).

Table 15. Reaction of amines with a trifiuoroacetic anhydridcformic acid mixturea.

Products (rel %) :
Amine Excess reagent* formamide trifluoroacetamide

3-Fluorobenzeneamine acylating mixt. I 99

Benzeneamine acylating mixt. 100 0
Benzeneamine amine 100 0
Benzylamine acylating mixt. 8 92
Benzylamine amine 3 97
Cyclohexylamine acylating mixt. 9 91
Cyclohexylamine amine II 89
N-Methylcyclohexylamine acylating mixt. 100 0
N-Methylcyclohexylamine amine 100 0
r-Butylamine amine 100 0

Formic acid (98%) was added dropwise to one equivalent of freshly distilled, acid free,
trifluoroacetic anhydride under stirring in inert atmosphere at - 20C. The obtained mixture
was used 30 min at - 20C after completion of the mixing of the reagents. It was advisable
to dilute the anhydride with an inert dry solvent like chloroform or toluene to moderate the
strongly exothermal reaction.
The acylating mixture at - 20C was added very rapidly under stirring to a solution of
the amine in chloroform or toluene.
The relative yield is approximately evaluated by first separating the two amides by GC
and then evaluating peak area ratios.

7. DANGEROUS PROPERTIES AND TOXICOLOGY

The adverse effects on human health caused by some kind of contacts with AFA (and any formic
mixed anhydride) were not investigated directly, but could be inferred. The compound is more
irritating than acetic anhydride and being more volatile the chances of inhalation are stronger. Eye
and skin protection are therefore mandatory. Decomposition to a gaseous poisonous product,
carbon monoxide. is to be foreseen ; therefore, tightly stoppered containers should be avoided.
1114 P. S~~Z~LlsI el al.

8 MISCELLANEOUS REMARK!3

The knowledge of this simple compound seems to be rather scant. The pure compound is stable
thermally both in the condensed and the gas phase.43*46q Even traces of acids and bases can start
irreversible changes. One has not to forget that glass itself is not inert and distillation apparatus
of another nature may be more suitable.
The reactivity pattern has not been adequately studied. The few studies available were not always
carried out with the now contemporary and accurate analytical techniques and, more importantly,
with AFA of precisely defined purity. Apart from our data (see Section 4), not a single batch of 1
has ever been analyzed by GC. Acetylation vs formylation and acylation regioselectivity, when
observed with close to stoichiometric ratios of 1 vs substrate and after long times, do not cast light
on the intrinsic reactivity pattern of I, if one keeps the reactions of Scheme 1 in consideration. It is
likely that as the concentrations of AcG or FoO- will increase, some Ac,O or For0 might be
formed, which will interfere with the observation, which, therefore, must be performed either at
very short times or with a very large excess of pure 1. If catalysts are used, the study must foresee
the observation of the nature and stability of the adduct AFA-catalyst.

k
FoO-,1
1 + FoO- _- FoZO + AcO-
-k
FOO_,l
FOX-Y + AcO-

k
Ace-,l
I + Ace- d- AcZO t FoO-
-4 XwYFo + AcO-
-k
ho-) I

AcX-Y + FoO-

X-Y + FozO

XuYAc + FoO- b X-YFo

2 *y

X-Y + AcZO
9 AcX-Y
%? 2 .r
X-YAc

Scheme 1.AC- vs Fo-ation and X vs Y rcgiosektivity.

Among the aprotic syntheses, the use of substrates with newly designed leaving groups may be
attractive : one of these, applied to the synthesis of a number of mixed formic anhydrides, employed
as a reactant the 0-acyl derivatives of polymeric pyridine oxide.46
Among few other formic mixed anhydrides, mention could be made of benzoic formic
anhydride, which is thermally stable as to be distilled at 69.OC (80 Pa). Its reactivity was tested
against aniline which gave formanilide in unreported yield.43 The synthesis of this anhydride may
be improved and adapted to some derivative yielding a solid mixed formic anhydride.
In 1973 a polymeric material I6 was introduced (pop corn polystyrene) with carboxylic functions
which can be converted to COCI by the action of oxalyl dichloride ; subsequent reaction with benzoic
 Acetic formic anhydride 1115

acid produced the mixed anhydride. Perhaps such a procedure could be applied to formic acid (or
some of its metallic salts) to produce a formylating agent of good stability and highly selective
reactivity especially when a too reactive nucleophile might be (in part) unable to discriminate
between the two carbonyls of 1.14
None of the reactions involving AFA reported to date were either shown or believed to involve
radical mechanisms, except, perhaps, for one case. 57

Acknow/edgemenf.s-A. G. Giumanini was the recipient of grants for University research (1987. 1988) from the Italian
Ministry of Education. Sabina Cauci rcccivcd a postdoctoral scholarship from Assindustria-Chimica del Friuli S.p.A. (1985
1987). Part of the financial support came also from a CNR grant to AGG (no 84/02150.03 and 86/01649.03). Assistance has
been provided by Dr A. Luchctta during her scholarship sponsored by Asschimici-Chimica dcl Friuli S.p.A. (19861987).
Our thanks arc also due to Dr S. Bono.

REFERENCES
Adachi, J.; Nomura, K.; Shiraki, K.; Mitsuhashi, K. C/rem. Phurm. Bull. 1974, 22,658.
r Agadzhanyan. T. E.; Arutyunyan, G. L.; Minasyan, G. G. ; Movscsyan, R. A. Arm. Khim. Zh. 1983,36.669.
Ah-kow. G.; Paulmicr. C.; Pastour, P. C. R. Acad. Sci.. Ser. C 1974,278, 1513.
Albert. A. J. Chem. Sot. Perkin Trans. 11973, 1634.
Allen&n, E. ; Bcryl, V. Chem. Ber. 1%7.100, 3551.
6 Allenstein, E. ; Bcryl. V. ; Eitcl. W. Chem. Ber. 1969,102.4089.
Ando, W.; Igarashi, Y. ; Huang, L. Chem. L~II. 1987,7, 1361.
* Angelici, R. J. ; Quick, M. H. ; Kraus, G. A. ; Plummcr, D. T. Inorg. Chem. 19%2,2I,2178.
Atancs. N.; Guitian, E.; Saa. C.; Castedo. L.; Saa,J. M. Tetrahedron LPII. 1987,28.817.
I0 Baltzcr, B. ; Lund, F. ; Rastrup-Andcrscn. N. J. Pharm. Sci. 1979.68, 1207.
Barta-&alai, G. ; Fetter, J. ; Lcmpcrt, K. ; Msller, J. Acra Chim. Acad. Sci. Hung. 1930, 104, 253.
Bartmcss. J. E. ; Hays, R. L. ; Caldwell. G. J. Am. Chem. Sot. 1981, 103, 1338.
Barton, D. H. R. ; Bowlcs, T. ; Husincc, S. ; Forbes, J. E. ; Llobera, A. ; Porter, A. E. A. ; Zard, S. Z. Tetrahedron Leu.
M&I, 29, 3343.
Barton, D. H. R.; Bringmann, G.; Lamottc, G.; Motherwell. W. B.; Motherwell, R. S. H.; Porter, A. E. A. J. Chem.
SW. Perkin Trans. I 1980, 2657.
Bcckct, G. J. P.; Ellis. G. P. Terrahedron Larr. 1976, 17, 719.
I6 Bcckct. G. J. P.; Ellis, G. P. ; Trindadc. M. 1. U. J. Chem. Res. (S) 1978.47.
Bchal, M. A. Ann. de Chim. er de Phys. 1900,20,41 I.
* Bellamy. L. J. ; Connclly, B. R. ; Philpotts, A. R. ; Williams, R. L. Z. Elekrrochem. MO, 64, 563.
Bennett. L. L.. Jr.; Allan. P. W. Cuncer Res. 1976.36, 3917.
20 Bergman, J. J. Heterocycl. Gem. lW1.8, 329.
2 Bergman, J. : Bergman. S. J. Org. Chem. 1985,50, 1246.
I2 Blackwood, R. K. ; Rcnnhard, H. H. ; Stephens, C. R. J. Am. Chem. Sot. 1!&0,82.5194.
2 Bodor. N., Sloan, K. B. ; Higuchi, T. ; Sasahara. K. J. Med. Gem. 1977.20, 1435.
Boogaard. A. ; Gcisc. H. J. ; MijlholT, F. C. J. Mol. S~rucr. 1972,13, 53.
2 Borck, B. A.; Waclsch, H. J. Biol. Chem. 1953.205.459.
Brcslow, R.; Czarniccki, M. F.; Emctt. J.; Hamaguchi. H. J. Am. Chem. Sot. 1980,102.762.
Bushcy, D. F. : Hoover, F. C. J. Org. Chem. 1980,45,4198.
* Claude, S. ; Tabacchi, R. ; Due, L. ; Fuchs, R. ; Booscn, K. J. Helu. Chim. Acra. 1!%0,63,682.
29 Clcmo. G. R.; Swan, G. A. J. Chem. Sot. l!M,I48,603.
Collman, J. P.; Winter, S. R. J. Am. Chem. Sot. 1973,95.4089.
Curvalc. R. A.; Yamin, L. J. ; Ponce, C. A. ; Vet?. F. T. An. Quim.. Ser. A 1982,78,255.
Dalglicsh. C. E. J. Chem. Sot. 1952,IS& 137.
Doyle, G. J. Orqanomef. Chem. 1982,224,355.
Du&ssc, J. Buli Sot. Chim. Fr. 1945,12. 918.
Dunbar. R. E.: Garvcn. F. C. J. Am. Chem. Sot. 1955. 77.4161.
6 DuVigncaud. V. ; Dorfmann. R. ; Loring. H. S. J. Biol: Gem. 1932,98, 577.
Du Vigncaud, V.; Patterson. W. 1. J. Biol. Chem. 1935, 109.97.
* Edwards. W. R.. Jr.; Kammann, K. P.. Jr. J. Org. Chem. 1964.29.913.
Edwards, W. R., Jr. : Sibillc. E. C. J. 0r.q. Chem. l!X3,28,674.
Eltsov. A. V.; Kukushkin. V. Yu.; Bvkova, L. M. Zh. Obshch. Khim. 1981.51.2116.
Evans. D. J. ; Eastwood. F. W. Ausr. j. Chem. 1974.27. 537.
I2 Fahmv, M. A. H. Eur. Par. ADD/. EP 215.509 (Cl.CO7F9144). , 25 Mar 1987.
Fanta: G. F. J. 0r.q. Chem. 1961.29.981:
Fersht. A. R.; Jcncks. W. P. J. Am. Chem. Sot. 1970.92. 5432.
Fiato, R. A.; Vidal, J. L. ; Pructt, R. L. J. Orgatwmef. Chem. 1979, 72. C4.
Fife, W. K.; Zhang. Z. 1. Org. Chem. 1986,51, 3744.
Fracheboud, M. G.; Shimomura, 0.; Hill, R. K. ; Johnson, F. H. Tefrahedron L.err. 1969.45.3951.
1116 P. Snuzzot.tn et al.

Freudenberg, K. ; Jakob, W. Chem. Ber. 1%7,80,325.

9 Fromageof H. P. M. ; GrifFin,B. E. ; Reese, C. B. ; Sulston. J. E. Tetrahedron 1967.23, 2315.
Fruton, J. S. ; Clarke, H. T. J. Biol. C/rem. 1934,106.667.
Gilbert, E. E.; Leccacorvi. J. R. ; Warman, M. ACS Symp. Ser. 22 1976, [lnd. Lab. Nitrations, Symp., (1975)], 327.
Giumanini, A. G. ; Verardo. G. Zh. Org. Khim. 1989, 25, 650.
Glamkowski, E. J.; Chiang, Y. J. Hererocycl. Chem. 1%7,24. 733.
Glichitch, L. S. Bull. Sot. Chim. Fr. 1923.33. 1284.
Gold. V.; Jefferson, E. J. J. C/tern. Sot. 1953, 156. 1416.
6 Gorog, S. ; Szepesi. G. Freseniuc Z. Anal. Chem. 1970,251. 303.
Gottlieb, R. ; Plleiderer, W. Lie&s Ann. Chem. 1981, 1451.
* Griffin, B. E.; Jarman, M.; Reese, C. B.; SuIston. J. E. Tetrahedron 1%7,23,2301.
Grundmann. C.; F&on, M. B. Chem. Ber. 1964.97.566.
Guenther, E. The Essential Oilr, Van Nostrand : New York ; Vol. I, p. 276, 1943.
6 Hannig. E. ; Kollmorgen, C. ; Dressel, M. Pharmuzie lW4. 29. 685.
Hara, H. ; Hosaka, M.; Hoshino, 0.; Umezawa, B. J. Chem. Sot. Perkin Trans. f1980, 1169.
6J Harmon, A. D.; Hutchinson. C. R. J. Org. Chem. 1975.40, 3474.
Harris, M. Eur. Par. Appl. EP 117.577 (Cl. C07D279/06). 05 Sept 1984.
61 Hartman, G. D.; Bilfar, S. E.; Wcinstock, L. M.; Tull. R. J. Org. Chem. 1978.43. 960.
LbHengartner, U. : Valentine, D.. Jr. ; Johnson, K. K. : Larscheid, M. E. ; Pigott, F. ; Scheidl, F.; Scott.J. W. ; Sun, R. C. ;
Townsend. J. M.: Williams, T. H. J. Ore. Chem. 1979.44.3741.
67Holtwick. J. B. ; Golankiewicz, B.; Holmes. B. N.; Leonard, N. J. J. Org. Chem. 1979.44, 3835.
Hulfman. C. W. J. Org. Chem. 1958,23,727.
69Hurd, C. D. ; Drake, S. S. ; Fancher, 0. J. Am. Chem. Sot. 1946.68.789.
Hurd, C. D.; Roe, A. S. J. Am. Chem. Sot. 1939.61.3355.
Johannsen. F. ; Jorgensen, A.; Pedersen. E. B. Chem. Ser. 1986. 26, 347.
72 Kaaka. E. A. ; Gitterman. C. 0.; Dulaney. E. L.; Folkers, K. Biochem. 1962, I, 340.
Karim, A. ; Mortreux, A. ; Petit. F.; Buono. G.; Peiffer, G.; Siv, C. 1. Organomef. Chem. 1936.317.93.
Kashman. Y. : Edwards, J. A. J. Ora. Chem. 1978 43, 1538.
Katritzky; A. R. ; Lagowski, J. N. chpmirrry oj he;erocyclic N-oxides; Academic Press: New York p. 279 and 362, 1971.
Khilya, V. P.; Pivovamtko, V. G.; Gorbdenko. N. V. Ofkrytiya, lzobrer. M7,32.90.
Kikukawa, K.; Kono. K.; Nagira. K.; Wada. F.; Matsuda, T. J. Org. Chem. 1981.46.4413.
I Knysh. E. G. ; Krasovskii, A. N.; Kochergin. P. M. Khim. Gererotsikl. Soedin. 1972, 30.
Koller, M.; Karpf. M.; Dreiding, A. S. Helu. Chim. Acfo 1983.66, 2760.
Krimen, L. I. Org. Synrh. 1970.50. I.
Krishnamurthy, S. Terrahedron L&r. 198223. 33 15.
* Ktihne, H.; Ha, T.-K.; Meyer, R.; Gtinthard, Hs. H. J. Mol. Specfr. 1979, 77, 251.
Kumano, T.; Yoshioka, T.; Uematsu, T. Drug. Metub. Dispos. 1986, 14,487.
Lang, G. : Richard. H. : Leduc,M. ; Junino, A. Eur. Pal. Appl. EP 239,826 (Cl. A6IK7/42). 07 Ott 1987.
OrLa palme, R.; Borschberg. H. J. ; Saucy, P. ; Deslongchamps, P. Can. J. Chem. 1979,57,3272.
*6 Mehlenbacher, V. C., in Organic analyses Determination of hydroxyl groups in essential oils with acetoformic acid;
Interscience Publishers Inc. : New York; Vol. I. p. 37, 1953.
* Michelin, R. A.; Angelici, R. J. Inorg. Chem. 1980, 19, 3853.
* Michelin, R. A.; Facchin, G.; Braga, D. ; Sabatini, P. Organomelallics 1986,5, 2265.
x9 Montgomery, J. A. ; Thomas, H. J. J. L.ubeNed Compd Radiopharm. 1978, 15 (suppl. vol.), 727.
9o Muramatsu, 1. ; Murakami, M.; Yoneda, T.; Hagitani, A. Bull. Chem. Sot. Japan 1965,38.244.
9 Nolte, R. J. M. ; Van Zomeren, J. A. J. ; Zwikker, J. W. J. Org. Chem.1!?7&43, 1972.
92Nyhammar, T. ; Grivas, S. ; Olsson, K. ; JHgerstad, M. Mufaf. Res. 19%. 174, 5.
Okawa. K.; Hase, S. Bull. Chem. Sot. Japan 1963.36.754.
Olah, G. A.; Dunne, K. ; MO, Y. K. ; Szilagyi, P. J. Am. Chem. Sot. 1972,94,4200.
9 Olah, G. A.; Kuhn, S. J. J. Am. Chem. Sot. 1960,82,2380.
96 Olah, G. A. ; Ohannesian, L.; Arvanaghi, M. Chem. Reu. 1%7,87,671.
Outurquin, F.; Lerouge, P.; Paulmier. C. Bull. Sot. Chim. Fr. 1986. 259.
Peet. N. P.; Sunder. S. Eur. Par. Appl. EP 132.788 (Cl. C07D257/06), I3 Feb 1985.
Perrone, E.; Alpegiani, M. ; Giudici, F. ; Bedeschi, A.; Pellizzato. R. ; Nannini, G. J. Heferocycl. Chem. 1984, 21, 1097.
looPerry, C. W.; Brossi, A. Synrhesti 1977, 79,492.
lo Pivovarenko. V. G. ; Khilya, V. P. Dopov. A&ad. Nauk U&r. RSR, Ser. 8: Geol., Khim. Biol. Nauki 1985.44.
lo2Pivovarenko V. G.. Khilya V. P.; Babichev, F. S. Dopov. Akad. Nauk Ukr. RSR. Ser. B: Geol.. Khim. Biol. Nauki
1985.56.
Chim. Fr. 1927,41, I 190.
lo Polonovski, M. ; Polonovski. M. Bull. Sot.
lo4 Reese. C. B.: Stewart. J. C. M.: van Boom, J. H.: de Leeuw, H. P. M. ; Nagel, J.; de Rooy. J. F. M. J. Chem. Sot.
Perki Tram: I 1975,934.
lo Russel. G. A. : Mikol. J. G. Mechanism u/ molecular m&arions, _- _ Ed., Interscience: New York ; Vol.
B. S. Thyagarajan,
I. p. 176, I%8
lo6Sammes, P. G. ; Wallace, T. W. J. Chem. Sot. Perkin Trans. I1975, 1845.
Sate. H.; Imamura, S. Jpn. Kokai Tokkyo Koho JP 61.148.155 [86,148.155] (Cl. CU7Cl25/065), 05 Jul 1986.
Sate. H.; Imamura, S. Jpn. Kokai Tokkyo Koho JP 61.140.552 [86.140,552] (Cl. CO7C103/048), 27 Jun 1986.
 Acetic formic anhydndc 1117

lop Schecker, H. G.; Zinner, G. Arch. Pharm. I%O, 313,926.

Schietz, E. R. J. Am. Chem. Sot. 1923.45.455.
Schijf, R. ; Scheeren. J. W. ; Van Es. A. ; Stevens. W. Rec. Trm. Chim. l%S. 84. 594.
Schijf. R. ; Stevens, W. Rec. Troo. Chim. 1966.85.627.
Schoening. R. C. ; Vtdal. J. L.: Fiato. R. A. J. Mol. Coral. 1981. 13.83.
Scott, M. K. ; Jacoby, H. I.; Mills, J. E. ; Bonfilio. A. C.; Rasmussen, C. R. J. Med. Chem. 1X3,26,535.
Serebrvakov. E. P. Izv. Akad. Nouk SSSR. Ser. Khim. 1980.2596.
Shambhu, M. B. ; Digmis, G. A. Tetrahedron Lpft. 1973, 14. 1627.
"Sheehan,J. C. ; Yang. D. D. H. J. Am. Chem. Sot. 1958.80. 1154.
Simkin, B. Ya. ; Kletskii. M. E.; Minyaev. R. M.; Minkin. V. 1. Zh. Org. Khim. 1983, 19, 3.
Skoma. G.; Stemmer, R.; Ugi, I. C&m. Ber. 1978. Ill, 806.
Sofuku. S. : Muramatsu. I. : Haaitani. A. Bull. Chem. Sot. Juoon 1967.40. 2942.
I Steinmetz. W. E. J. Am. &em. &cc. i973,95, 2777. .
I Stevens, W. ; Van Es. A. Rec. Two. Chim. 1%4,83,863.
Stevens. W.; Van Es. A. Rec. Trov. Chim. 1964.83, 1287.
I Stevens, W. ; Van Es, A. Rec. Troo. Chim. 1964.83. 1294.
zStrupcz.ewski. J. T.; Allen, R. C. ; Gardner, B. A. : S&mid, B. L.; Stache, V. ; Glamkowski, E. J.; Jones, M. C.;
Ellis, D. B.; Huger. F. P.; Dunn, R. W. 1. Med. Chem. 1985.28, 761.
rbSullivan. D. F.; Scopes. D. I. C. ; Kluge. A. F.; Edwards, J. A. J. Org. Chem. 1976.41, I 112.
Tabor, H.; Mehler, A. H. J. Biol. Chem. 1%4,2/O. 559.
Takahashi, T.; Huokami, S.; Kato, K.; Nagata. M. J. Org. Chem. 1X3.48,2914.
Tamura. Y.; Mtyamoto, T.; Shimooka. K.; Masui. T. Chem. Phorm. Bull. 1971,19, 119.
Terashima, S. ; Takashima. T. ; Sato, T. ; Yamada. S. C/tern. Phorm. Bull. 1973.21. 1135.
I Tesster T Frechet, J. M. J.; Ito, H.; Willson, C. G. Polym. Prepr. (Am. Chem. Sot., Dir. Polym. Chem.) 1984.25, 313.
r Tokttoh. N.; Okazaki. R. Chem. Larr. 1985, 1517.
I Treppendahl. S.: Jakobsen. P. Acre Chem. Scond.. Ser. B 1978.32,697.
Treppendahl. S., Jakobsen. P. Acra Chem. Scond.. Ser. B 1983.37.953.
"Tulloch. A. P. Chem. Phvs. Lwids 1971.6. 235.
Unangst. P. C.; Brown, k. E.J. Helero&. Chem 1984.21. 283.
I Unpublished results from our laboratory.
I* Van dcr Eijk. J. M. ; Nolte. R. J. M.; Zwikker. J. W. J Org. Chem. 1980.45. 547.
W Van Doorn. J. A. ; Masters. C.; Van der Woude. C. 1. Organome;. Chem. 1977,141.231.
Van Dort. M.; Ncubtg. R.; CounselI. R. E. J. Med. Chem. 1981.30, 1241.
I Van Es A. Stevens, W. Rec. Tra-. Chum. 1%6,84.704.
Van Es: A. I Stevens, W. Rec. Trot?. Chim. 1965.84, 1247.
I Vasella. A.; VoetTray. R. Heh. Chim. ACIO 1982, 65. 1953.
Vilvala. R. Acre Phorm. Fenn. 1978.87. 85.
I Vledder. H. J.; MijlhotT, F. C.; Leyte, J. C. J. Mol. Strucf. 1971, 10. 57.
I Vledder . H. J. . Mttlhoff. F. C.: Van Well. F. P. : Dofferhoff. G. M. T.: Levte. J. C. J. Mol. Sfrucl. 1971. 9. 25.
I Vlietstra. E. J.; N&e, R. J. M. ; Zwikkerl J. W.: Drenth, W.; Jansen,R. H. A. M. Reel.: J. R. Nerh. bh&. Sot. 1982,
101. 183.
* Vlietstra, E. J. : Zwikker. J. W.; Nolte. R. J. M.; Drenth. W. Reel. J. R. Nerh. Chem. Sot. 1982, IOI. 460.
I Waley. S. G. Chem. Ind. 1953.27. 107.
We& L. B. ; Koeppe. R. E. II Inr. J. Peppr. Protein Rex 1985.26. 305.
I Weiss. R: . Roth. R. J. Gem. Sot.. Chem. Commun. 1987. 317.
I Whcland. G. W. : Re.~ononcc in organic chemrsrry. Wiley : New York ; p. 100. 1955.
I Wobtg. D.; Losch. R.: Goerdeler. J. Lrcbigs Ann. Chem. 1973, 1416.
I Wust. H. H.; Bardenhaaen. J.: Schiillkopf. U. Laebus Ann. Chem. 1985. 1825.
I Yamada. K. ; Takeda. M. : lyoh, N. ; Ohisuka, H.; Tsunashima. A.; Iwakuma. T. Chem. Phorm. Bull. 1982, 30. 3197.
Yazawa, H. ; Goto, S. Telrohedron Lar. 1985.26. 3703.
Zavyalov. S. I.: Zavozm. A. G. lx. Akad. Nook SSSR. Ser. Khim. 1979.2834
Zeller. K. P. Terruhedron L.err. 1977, 18. 707.
Zeller, K. P. Lwbigs Ann. Chem. 1979, 2036.
a Zemlicka. J.; Beranek. J. ; Smrt, J. Co//. Czech. Chem. Commw. 1962. 27, 2784.

NOTE ADDED IN PROOF

The followrng expertments wtth AFA appeared after the completton of the manuscript (Refs through Chemical Abstracts
July-December 1988).
1118 P. STRA7.7.OLINIer a/.

Hambley. T. W.: Harsanyi. M. C.: Norris. R. K. J. Or.9. C/rem. 198B,j3. 3104.

Lang. G. ; Richard, H. ; Leduc. M. ; Junino. A. Ger. Ofirr. DE 3,707.088 (Cl. CO9B7100). IOSep 1987 and Be/g. BE 1.000.139
(Cl. A6lK). 26 Apr 1988.

Sauter, F.; Jordis. U.: Rudolf. M.; Wieser, J.; Baumann. K. Ger. Ofin. DE 3.721,745 (Cl. CO7D487;08). 14 Jan 1988.

C3 C3
(3 bfA
3 I - JC
H3Cn h/ n
2C R
tno

Zclenin. K. N.; Malov. M. Y.; Z&ova. 1. V.; Terentev. P. B.; Kalandarishvili. A. G. Khim. Geterorsikl. Sodin. 1987, 1210.

Barnes. K. F. ; Browning, 1. R.; Clark, N. G. ; Clark, R. J. fe.v/ic. Sri. 1988. 23, 65.

P PH3P.#A I"'
a%!n 0 It1
U)T.0.5-l" vo-tn 0 Mmt3
n- ,,.,.A
u 57
R"0 OR' R"0 01'

Hiebl. J. ; Zbiral. E. Liebigs Ann. Chem. 1988. 765

Gollnick. K. ; Koegler. S. Tetrahedron LPII. 1988, .?Y, 1007.