Controlled release of perfumery alcohols by alkaline hydrolysis of 2-formyl- and 2-acetylbenzoates and their corresponding phthalides

Article abstract of DOI:10.1039/b008082i

The controlled release of perfumery alcohols by alkaline hydrolysis of 2-formyl- and 2-acetylbenzoates and their corresponding phthalides were investigated. The rate of benzoate hydrolysis depended on the structure of alcohol released. It was suggested th

Full text of DOI:10.1039/b008082i

View Article Online / Journal Homepage / Table of Contents for this issue
Controlled release of perfumery alcohols by alkaline hydrolysis of
2-formyl- and 2-acetylbenzoates and their corresponding phthalides
Paul Enggist, Sabine Rochat and Andreas Herrmann*
2
Firmenich SA, Division Recherche et Développement, B.P. 239, CH-1211 Genève 8, Switzerland
Received (in Cambridge, UK) 6th October 2000, Accepted 26th February 2001
First published as an Advance Article on the web 8th March 2001
The 2-formyl- and 2-acetylbenzoates of fragrance alcohols
and their corresponding pseudo-esters (phthalides) are
suitable precursors for the controlled release of fragrances
by alkaline hydrolysis; the rate of benzoate hydrolysis
depends on the structure of the alcohol released.
To increase the performance of consumer products, and,
in particular to prolong the perception of volatile molecules
such as flavours or fragrances, the development of suitable
delivery systems for functional perfumery applications such as
shampoos, creams, soaps, fabric softeners or detergent powders
is becoming an increasingly important research area in the
flavour and fragrance industry.1 Most of the delivery systems
described so far are based on the release of active compounds
from an encapsulating matrix, by diffusion out of the matrix or
by destruction of the capsules. An alternative to encapsulation
consists of the preparation of suitable precursors that release
the active material by a chemical reaction in the targeted
application.2 We herein describe the preparation of a series of
2-formyl- and 2-acetylbenzoates, as well as their cyclic pseudo-
esters (phthalides), for the controlled release of perfumery
alcohols in alkaline media. The release mechanism of the
benzoates in basic solution is based on the hydration of the
carbonyl group followed by intramolecular nucleophilic attack
(neighbouring group participation)3 to form a lactone (which
then opens to the carboxylate of the corresponding acid) and
the desired perfumery alcohol, with the hydroxide ion addition4
or the cyclisation5 being the rate determining steps. In the case
of the pseudo-esters, the rate determining step is the addition of
hydroxide to the ester carbonyl group.6,7 Rapid ring opening
then gives the anion of the corresponding carboxylic acid
together with the fragrance alcohol (Scheme 1).
Several attempts to synthesise 2-substituted benzoates
from the corresponding benzoic acids, via their acid chlorides8
and a perfumery alcohol, were unsuccessful, presumably due
to ring–chain tautomerism of the carboxylic acid.9 Methyl
2-acetylbenzoate (1a) (Scheme 1) was successfully obtained
by reaction of 2-acetylbenzoic acid with diazomethane.10
Transesterification with geraniol using sodium methoxide in
cyclohexanol resulted in a ca. 1 : 1 mixture of benzoate 2a and
pseudo-ester 2b. Formyl- and acetylbenzoates 2a–7a (Scheme 1)
were finally synthesised from their corresponding acids using
1,3-dicyclohexylcarbodiimide (DCC) with 4-dimethylamino-
pyridine (DMAP) in dichloromethane.†11,12 Whereas phthal-
Scheme 1
ides 3b–5b were obtained as side products in the DCC coupling
reaction of 2-formylbenzoic acid, the formation of 3-methyl-
phthalides in the reaction with 2-acetylbenzoic acid was not
observed. All compounds were stable and could be isolated in
their pure state. No evidence for an equilibration between the
corresponding ring–chain tautomers was obtained under the
conditions described in this study. Interestingly, the reported
synthesis of 7a via the acid catalysed esterification of 2-acetyl-
benzoic acid with (Ϫ)-menthol using HCl,13 could not be
reproduced: following the published procedure we obtained
pseudo-ester 7b as a diastereoisomeric mixture.
The kinetics for the alkaline hydrolysis of the methyl or
phenyl esters of 2-formyl- or 2-acetylbenzoates4,5,12,14,15 and
their pseudo-esters6,15 have been investigated by different
research groups. In these studies the dependence of the rate of
hydrolysis with respect to substitution on either the carbonyl
function or the aromatic ring was analysed in detail.5,6,12 How-
ever, the influence of the leaving alcohol on the rate constants
of the hydrolysis, which is crucial for the targeted application in
functional perfumery, has not been studied extensively. Phenyl
2-acetylbenzoates were reported to hydrolyse ca. 1.3 times faster
438
J. Chem. Soc., Perkin Trans. 2, 2001, 438–440
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b008082i

View Article Online
Table 1 Measured rate constants k0 for the alkaline hydrolysis of
2-formyl- and 2-acetylbenzoates 1a–7a and their pseudo-esters 3b–7b
in water–acetonitrile 2 : 1 at 20 ЊC. All numbers are average values of
at least two experiments
hydrolysis of the benzoates, but not in the case of the phthalide
derivatives. Both the benzoates 1a–7a and phthalides 2b–7b
were found to liberate the desired perfumery alcohols, and can
thus be used as delivery systems for the controlled release of
fragrances in functional perfumery.17 Due to their broad range
of different rate constants, mixtures of benzoates and phthal-
ides such as 3a and 3b may well be efficient for providing a long
lasting effect of odour perception in perfumery applications;
this aspect is currently under investigation.
Compound
k0 × 105/sϪ1
pH
Compound
k0 × 105/sϪ1
1a
2a
12.7
3.98
38.2
108
10.47
10.47
11.54
10.47
11.54
10.47
11.54
10.47
11.54
10.47
11.54
10.47
11.54
3a
4a
5a
6a
7a
3b
4b
5b
0.839
7.17
0.940
9.79
0.865
6.62
Acknowledgements
315
We are grateful to Drs Eric Frérot, Sina Escher, and Roger
Snowden who contributed to this project in numerous
discussions. We thank one of the referees for very constructive
comments, which helped us to improve the manuscript.
30.0
2.50
25.6
7b
0.153
2.89
Notes and references
† Representative experimental procedure. A solution of 7.50 g (50.0
mmol) of 2-formylbenzoic acid, 4.89 g (40.0 mmol) of DMAP and
15.42 g (100.0 mmol) of geraniol in CH2Cl2 (75 ml) was cooled in an
ice-bath prior to the addition, over 15 min, of a solution of 11.37 g
(55.0 mmol) of DCC in CH2Cl2 (25 ml). The reaction mixture was
stirred for 15 min at 0 ЊC, then at 20 ЊC for 48 h. The precipitate formed
in the reaction was filtered off and the filtrate washed with HCl (10%,
2×), a saturated solution of Na2CO3 (2×) and water (2×, pH ≈ 7). The
organic layer was dried (Na2SO4), concentrated and chromatographed
(SiO2, heptane–ether 8 : 2) to give 2.55 g (22%) of 3a and 4.36 g
(38%) of 3b. Full spectroscopic data have been obtained for all new
compounds.
than the corresponding methyl esters,12 and the influence of
the alcohol released in the hydrolysis of the pseudo-esters of
2-benzoylbenzoates was found to be poor.7 In this work, the
rate constants of precursors 1–7 were determined in buffered
solutions of water–acetonitrile 2 : 1 at 20 ЊC. The reaction
solutions were injected at constant time intervals into a high
performance liquid chromatography (HPLC) apparatus and
eluted on a reversed phase column with water–acetonitrile.‡
Due to the fact that the hydroxide concentration was held
constant by the buffer, the second order rate expression
k2[OHϪ][precursor] can be reduced to the first order expression
k0[precursor].5 Plotting the logarithm of the surface area
quotients (At/A0) of the benzoates or phthalides against time
gives a straight line with good correlation coefficients (>0.99)
for all the measurements, thus justifying the general assumption
of first order kinetics. The rates of hydrolysis were found
to slightly deviate from proportionality to the hydroxide ion
concentration under the conditions described in this work.
However, for our purposes, the direct comparison of the
measured rate constants (k0) at the same pH was sufficient. The
results obtained for the release of different perfumery alcohols
are summarised in Table 1. UV spectra measured after complete
transformation showed that the reaction proceeds completely
to the acid in the open form, and the formation of the
ring–chain tautomeric phthalide 8 as reaction intermediate was
not observed. Direct comparison of the rate constants (k0)
measured for methyl 2-acetylbenzoate (1a) in this study with
those reported in water–1,4-dioxane5 shows that our reaction
rates (determined by HPLC as well as by UV spectroscopy) are
ca. ten times slower than the literature values.16 The hydrolysis
of citronellyl 2-formylbenzoate (5a) is about ten times faster
than that of the corresponding 2-acetyl derivative 6a, whereas
geranyl 2-formylbenzoate (3a) hydrolyses ca. 25 times faster
than its 2-acetyl analogue 2a. Interestingly the rate constants
are also very strongly dependent on the structure of the
alcohol released during the hydrolysis. Comparison of menthyl
2-acetylbenzoate (7a), which liberates a secondary alcohol,
with primary alcohol derivatives 6a or 2a, shows that, in the
last two cases, the rate of ester hydrolysis increases roughly
by factors of 10 and 15, respectively. Moreover, benzoates
of allylic primary alcohols (geraniol) are hydrolysed twice
as fast as their homologous saturated alcohols (citronellol).
This is not the case for phthalides 3b, 4b and 5b, where all the
measured rate constants are of the same order of magnitude.
Doubling the buffer concentration at the same pH did not
influence the rate of hydrolysis of the benzoates within the
experimental error. In the case of the phthalides, however, a
slight decrease of the rate constants (by ca. 20% for 5b) was
observed at higher buffer concentrations, a fact that indicates
the participation of the buffer components in the reaction
pathway.
‡ Procedure for the hydrolysis experiments. All samples were thermo-
statted at 20 ЊC. The pH values of the buffer solutions in water–
acetonitrile 2 : 1 were measured (on
a Mettler Toledo MP220
instrument with an InLab 410 Ag/AgCl glass electrode) to be
10.47 0.01 (borate) or 11.54 0.02 (phosphate–bicarbonate). Com-
pounds 1–7 (35 to 45 mg) were dissolved in 25 ml of acetonitrile and
0.2 ml of this solution were added to 1.0 ml of a buffer solution
in water–acetonitrile 4 : 1. The mixture was immediately injected
in a HPLC apparatus (t = 0), eluted at 1 ml minϪ1 on a reversed phase
column with a mixture of water–acetonitrile containing 0.1% of TFA
and analysed at λ = 254 nm. Most of the chromatograms were recorded
on a Macherey-Nagel, Nucleosil 100-5 C18 column (250 × 4 mm id)
using a water–acetonitrile gradient (70 : 30 to 20 : 80 over 20 min) and
re-injected (20 µl) every 35 or 70 min (20 times). The kinetics of
compounds 2a and 6a (at pH 11.54) as well as 3a and 4a were measured
on a Merck Chromolith SpeedROD RP-C18e column (50 × 4.6 mm id)
by isocratic elution with water–acetonitrile 3 : 7 (2a, 3a and 6a) or
2 : 3 (4a). The samples (10 µl) were re-injected every 5.3 or 4.3 min,
respectively (12–20 times). At = surface area at time t, A0 = surface area
at t = 0.
1 K. Rogers, Cosmet. Toiletries, 1999, 114, 53; L. Brannon-Peppas,
ACS Symp. Ser., 1993, 520, 42.
2 S. Rochat, C. Minardi, J.-Y. de Saint Laumer and A. Herrmann,
Helv. Chim. Acta, 2000, 83, 1645.
3 K. Bowden, Adv. Phys. Org. Chem., 1993, 28, 171 and references
cited therein.
4 K. Bowden and G. R. Taylor, J. Chem. Soc. B, 1971, 149.
5 M. S. Newman and A. L. Leegwater, J. Am. Chem. Soc., 1968, 90,
4410.
6 F. Anvia, K. Bowden, F. A. El Kaissi and V. Saez, J. Chem. Soc.,
Perkin Trans. 2, 1990, 1809.
7 M. V. Bhatt, K. S. Rao and G. V. Rao, J. Org. Chem., 1977, 42,
2697.
8 See for example: J. O. Halford and B. Weissmann, J. Org. Chem.,
1952, 17, 1646; M. Renson, Bull. Soc. Chim. Belg., 1961, 70,
77.
9 P. R. Jones, Chem. Rev., 1963, 63, 461; J. Finkelstein, T. Williams,
V. Toome and S. Traiman, J. Org. Chem., 1967, 32, 3229; K. Bowden
and G. R. Taylor, J. Chem. Soc. B, 1971, 1390; K. Bowden and
G. R. Taylor, J. Chem. Soc. B, 1971, 1395.
10 E. E. Smissman, J. P. Li and Z. H. Israili, J. Org. Chem., 1968, 33,
4231; P. M. Pojer, E. Ritchie and W. C. Taylor, Aust. J. Chem., 1968,
21, 1375.
11 B. Neises and W. Steglich, Org. Synth., 1990, Coll. Vol. VII, 93.
12 F. Anvia and K. Bowden, J. Chem. Soc., Perkin Trans. 2, 1990,
1805.
Our results suggest that the structure of the leaving alcohol
is important in the rate determining step of the alkaline
13 H. G. Rule and J. Smith, J. Chem. Soc., 1926, 553.
J. Chem. Soc., Perkin Trans. 2, 2001, 438–440
439

View Article Online
14 M. L. Bender and M. S. Silver, J. Am. Chem. Soc., 1962, 84,
4589.
consider the reported autoprotolysis constant pkw = 15.05 for water–
acetonitrile 2 : 1 (at 25 ЊC, see: U. Mandal, S. Bhattacharya and
K. K. Kundu, Indian J. Chem., Sect. A, 1985, 24, 191).
15 M. L. Bender, J. A. Reinstein, M. S. Silver and R. Mikulak,
J. Am. Chem. Soc., 1965, 87, 4545.
17 Parts of this publication are the subject of a patent application:
E. Frérot, A. Herrmann, J.-Y. Billard de Saint Laumer and
O. Gräther, to Firmenich SA, PCT Int. Patent Appl., WO 00/58260,
2000; Chem. Abstr., 2000, 133, 266604.
16 The k2 value of 4.83 for 1a (obtained for constant hydroxide
)
w
concentration by dividing k0 by 10(pH ؊ pk ), however, is in the
same order of magnitude as the literature value (5.05, ref. 5) if we
440
J. Chem. Soc., Perkin Trans. 2, 2001, 438–440