Controlled light-induced release of volatile aldehydes and ketones by photofragmentation of 2-oxo-(2-phenyl)acetates

Article abstract of DOI:10.2533/chimia.2007.661

The light-induced controlled release of fragrances from photolabile 2-oxo-(2-phenyl)acetates via Norrish Type II photofragmentation was evaluated by irradiation of the precursors in different solvents and on cotton in a typical fabric softener application. The desired photooxidation was found to work efficiently in water-based systems, and it tolerates the presence of oxygen. The formation of a certain amount of alcohol besides the desired aldehyde or ketone was attributed to further reaction of the photochemically released carbonyl compound, rather than to ester hydrolysis in an aqueous environment. Schweizerische Chemische Gesellschaft.

Full text of DOI:10.2533/chimia.2007.661

PHOTOCHEMISTRY
661
doi:10.2533/chimia.2007.661
CHIMIA 2007, 61, No. 10
Chimia 61 (2007) 661–664
Schweizerische Chemische Gesellschaft
ISSN 0009–4293
©
Controlled Light-Induced Release
of Volatile Aldehydes and Ketones by
Photofragmentation of
2-Oxo-(2-phenyl)acetates
Barbara Levrand and Andreas Herrmann*
Abstract: The light-induced controlled release of fragrances from photolabile 2-oxo-(2-phenyl)acetates via Nor-
rish Type II photofragmentation was evaluated by irradiation of the precursors in different solvents and on cotton
in a typical fabric softener application. The desired photooxidation was found to work efficiently in water-based
systems, and it tolerates the presence of oxygen. The formation of a certain amount of alcohol besides the desired
aldehyde or ketone was attributed to further reaction of the photochemically released carbonyl compound, rather
than to ester hydrolysis in an aqueous environment.
Keywords: Delivery systems ⋅ Fragrances ⋅ Keto esters ⋅ Photooxidations ⋅ Triggered release
1
. Introduction
a consequence of their high volatility, fra- with a carboxylic acid in a photooxidation
grances easily evaporate and can thus only pathway, as illustrated in Scheme 1.[ As
be perceived for a limited period of time. aldehydes tend to be relatively unstable in
On the other hand, this evaporation takes various applications of functional perfum-
place from surfaces which are generally ex- ery, this class of compounds is of particular
posed to ambient daylight, and thus photo- interest for release from photolabile profra-
12]
Natural outdoor sunlight is an important
energy source for numerous biological pro-
cesses. Historically, it also served as the
reaction trigger for the development of the
first photochemical reactions. To day, with
responsive fragrance delivery systems seem grances. Both, alkyl and aryl 2-oxoacetates
to be particularly appropriate to control the were found to work efficiently for the re-
release of volatile compounds. Th e design
lease of different fragrance aldehydes and
the availability of modern photochemical
equipment, preparative syntheses are rarely
carried out by photoirradiation with natu-
ral daylight, but rather with lamps emitting
light at specific wavelengths and at a high
intensity. Up to now, photochemical pro-
cesses have mainly been used in functional
group protection for organic synthesis,
for various applications in bioorganic sys-
tems or drug discovery and, only re-
cently, they have also been applied to the
controlled release of volatiles in various
[8−11]
of suitable precursor structures, so-called ketones.
profragrances’ has become an important availability of 2-oxo-(2-phenyl)acetic acid
area of research in the flavour and fragrance (= phenylglyoxylic acid) (1) and its methyl
Given the ready commercial
‘
[1]
[6]
industry.
or ethyl esters (2 and 3), the present work
To be efficient for the release of fra-
focuses on the development of 2-oxo-(2-
[
2]
grances, photochemical delivery systems phenyl)acetates as photolabile delivery
have to respond to the relatively low light systems in functional perfumery.
intensities of ambient daylight and to work
in the presence of oxygen in polar solutions,
preferentially in water. A typical reaction 2. Results and Discussion
fulfilling these requirements is the Norrish
[
3]
[4]
[
5,6]
applications of functional perfumery. As
Ty pe II photoreaction of carbonyl deriva-
As a first step, the synthesis of 2-oxo-
[
5,7]
tives, with 2-oxoacetates (α-keto esters) (2-phenyl)acetates 4−7 (Scheme 2) was
being particularly relevant in this con- attempted by transesterification of 2 (or
[
8−11]
text.
latter release aldehydes or ketones together propanol (obtained by reduction of Lilial
In the presence of oxygen, these 3) with 3-(4-tert-butylphenyl)-2-methyl-
®
R2
O
2
R3
R2
R
R3
R
O
O
H
O
OH
O
3
R2
*
Correspondence: Dr. A. Herrmann
O
h"
O2
- CO2
O
O
HO
+
Firmenich SA
Division Recherche et Développement
B. P. 239
O
O
R1
R1
R1
R3
R1
HO
O
CH-1211 Genève 8
Tel.: +41 22 780 34 74
Fax: +41 22 780 33 34
E-Mail: andreas.herrmann@firmenich.com
Scheme 1. Mechanism for the photochemical release of aldehydes and ketones from 2-oxo-(2-
phenyl)acetates in the presence of oxygen[
12]

PHOTOCHEMISTRY
662
CHIMIA 2007, 61, No. 10
®
with LiAlH in ether), Floralol , geraniol
4
O
and α-ionol, respectively, in the presence
of a small amount of sodium methoxide
in refluxing cyclohexane. Surprisingly,
O
R2
O
O
2
2
R OH
O
R OH
O
OH
R1
NaOCH3
DCC, DMAP
under these conditions 3-(4-tert-butylphe-
4-7, 12
O
cyclohexane
CH2 Cl2
O
nyl)-2-methylpropanol and Floralol® did
1
OH
2
3
: R = CH3
: R = CH2 CH3
O
not afford the desired keto esters 4 and 5,
but a mixture of α-hydroxy esters 8 and 9
together with benzoates 10 and 11, respec-
tively (presumably as the result of a redox
process), whereas oxo(phenyl)acetates 6
and 7 were formed as expected from gera-
1
O
R2
1
R2
O
+
O
8, 9
10, 11
4, 8, 10: R2
=
5, 9, 11: R2
=
6: R2
=
niol and α-ionol (Scheme 2). Th e formation
of the α-hydroxy ester was also observed
in the reaction of hexylcinnamic alcohol
with ethyl ester 3, but this reaction pathway
7
: R2
=
12: R2
=
[
13]
was not further investigated as 2-oxo-(2-
phenyl)acetates 4, 5 and 12 (Scheme 2) can
be easily prepared from 1 using N,N’-dicy-
clohexylcarbodiimide (DCC) as described
in the literature.[
14]
Scheme 2. Preparation of 2-oxo-(2-phenyl)acetates 4−7 and 12
Th e release of the different aldehydes
Table. Amount of fragrance aldehydes and ketones released and quantity of remaining starting material
and ketones from their respective precur-
(
in brackets, rounded to ± 5%) after photoirradiation of 2-oxo-(2-phenyl)acetates in undegassed
sorswasinvestigatedinundegassedsolution
by irradiation with a xenon lamp (Heraeus
solution for 3 h. All numbers are average values of three samples.
2
[15]
Suntest CPS at 460 W/m )
or outdoor
N°
Name of
Released
Compound
Light Source
Yield of Released Aldehyde or Ketone and
Amount of Remaining Starting Material
(in brackets) [mol-%]
sunlight ( Ta ble). Ca. 8 mM solutions of the
different 2-oxo-(2-phenyl)acetates 4−7 and
2 in the indicated solvent were prepared by
adding 1 ml of a 0.01 M solution of decanol
which was used as an internal standard for
the GC analysis) to 5 ml of a 0.01 M solu-
1
Toluene
2-Propanol
Acetonitrile
(
®
4
Lilial
xenon
39
(10)
8
(35)
53
(<5)
tion of the 2-oxo-(2-phenyl)acetate. Th ese
solutions were then irradiated for 3 h in 10
sunlight
n.d.
n.d.
8
(30)
70
(5)
ml volumetric Pyrex glass flasks and anal-
ysed by GC and GC/MS as described previ-
[
9,10,16]
ously
. In each case a control experi-
Triplal®
xenon
10
(10)
4
(35)
16
(10)
ment was performed in the dark.
5
Th e data in the Ta ble show that all com-
sunlight
10
(5)
n.d.
n.d.
6
(5)
pounds afforded the desired aldehydes or
ketones after only 3 h of irradiation. Compa-
rable amounts of fragrances were released
upon irradiation with the xenon lamp or
6
trans/cis-Citral
xenon
10/10
(5)
3/3
(25)
6/7
(5)
natural outdoor sunlight. Th e lower yields
of aldehydes or ketones (and higher yields
of remaining starting material) obtained for
the irradiation in 2-propanol, as compared
to toluene or acetonitrile as solvent, suggest
that the reaction proceeds more slowly in
polar solution.
sunlight
10/9
(<5)
2/2
(25)
8/9
(5)
7
α-Ionone
xenon
13
(5)
(5)
2
(30)
32
(<5)
(<5)
sunlight
11
n.d.
n.d.
24
To evaluate the performance of the de-
livery systems under more realistic appli-
cation conditions, we performed dynamic
12
Phenylacetic
aldehyde
xenon
51
41
(10)
n.d.
n.d.
n.d.
n.d.
53
21
(<5)
(<5)
headspace analyses on small cotton sheets
that had previously been treated with a
fabric softener containing the 2-oxo-(2-
phenyl)acetate or a molar equivalent of the
aldehyde or ketone to be released as the
reference sample. In a typical experiment,
sunlight
(15)
either 0.39% of 6 or 0.14% of citral (serv- and then wrung out, while ensuring a con- tube of CaCl to remove the humidity from
2
ing as the reference) were thus weighed in- stant amount of residual water. Th e sheets
the ambient air) was continuously pumped
to 1.8 g of the fabric softener formulation, were each put into a home-made headspace across the headspace cell. After equilibrat-
respectively, and filled up to ca. 600 ml sampling cell (ca. 160 ml of volume), ther- ing the system for 15 min, the volatiles were
®
with (demineralised) tap water. Th en one
pre-washed cotton sheet (12 x 12 cm) was xenon lamp at ca. 108500 lux. A constant 5 min, and the sampling was repeated 6−8
mostatted at 25 °C and irradiated with a adsorbed onto a clean Te nax cartridge for
added to each sample and manually stirred flow of air (ca. 200 ml/min, filtered through times at constant time intervals (every 50
®
for 3 min, left standing for another 2 min activated charcoal and passed through a min). After the experiment the Te nax car-

PHOTOCHEMISTRY
663
CHIMIA 2007, 61, No. 10
from a fabric softener formulation. Th e evo-
500
400
300
200
100
0
lution of the headspace concentrations of
the photochemically released aldehyde and
the formation of the corresponding alcohol
suggest a consecutive multistep reaction as
a possible explanation for the observed da-
ta. However, additional experiments will be
required to support this hypothesis. As the
alcohol formed in the photoreaction is also
a perfumery alcohol with a similar olfactive
descriptor as the corresponding aldehyde,
the released mixture of the two compounds
is well appreciated by the perfumers.
unmodified citral
headspace
conc.
[
[ nn gg / /Ll oo ff aa ii rr ]]
citral
released from 6
geraniol
released from 6
0
100
200
300
400
4. Experimental Section
time [min]
General
Th e preparation of precursor 6 and the
Fig. Headspace concentrations of unmodified citral (---Å---, data from
measurement with NaCl) and of citral (—ò—) and geraniol (—¢—) released
from precursor 6 by irradiation with a xenon lamp (108500 lux, data from
measurement with CaCl2)
photoirradiations in solution were carried
out as described previously.[
10,16]
All new
compounds were fully characterised by
UV/Vis, IR and NMR spectroscopy as well
as by mass spectrometry.
tridges were thermally desorbed and the in the reaction, they closely resemble the
amount of trapped volatiles quantified by ones expected for stepwise, consecutive
Synthesis of
2
-Oxo-(2-phenyl)acetates under
GC analysis by calibration with an external reactions. Th e photochemical formation of
DCC Coupling Conditions
A solution of 1, DMAP (0.1 equiv.) and
standard.
citral from the precursor at constant light
intensity would give rise to an initially
Th e results obtained are depicted in the
Fig. Th e amount of citral released from
increasing headspace concentration of alcohol (1.8 equiv.) in dichloromethane (ca.
first increased and then constantly de- the aldehyde. Th e fact that a maximum is
100 ml for 45 mmol of 1) was cooled in
creased after reaching a maximum after reached, followed by a slow decrease of the an ice-bath before a solution of DCC (1.2
h, whereas the headspace concentration of aldehyde concentration and, a concurrent equiv.) in dichloromethane (ca. 40 ml for
the citral of the reference sample decreased increase of the alcohol concentration could 40 mmol of DCC) was added during 15−40
6
1
continuously. Th e headspace concentration
curve of the reference sample crosses the generated from the former under the given 10−15 min at 0 °C, then at 20 °C for sev-
suggest that the latter was (at least partially) min. Th e reaction mixture was stirred for
one of the citral released from the precur- reaction conditions. Th is could also explain
sor after ca. 150 min, thus showing the ex- the dependency of alcohol formation on the and the filtrate taken up in ether, washed
eral hours. Th e precipitate was filtered off
pected slow release effect. Similar results light intensity. Whether the formation of with water (3x), HCl (10%, 3x), and a satu-
were also obtained in a second experiment, geraniol is due to a reduction of the alde- rated solution of Na CO (3x). Th e organic
2
3
when the humidity of the air was kept at hyde in the reaction medium or the result layer was dried (Na SO ), concentrated and
2
4
about 75% (using a saturated solution of of a photochemical pathway remains unex- chromatographed (SiO , heptane/ether 8:2)
2
NaCl instead of the CaCl tube).
plained. As many other parameters such as to give the target compound.
Besides the desired citral, a certain matrix effects of the surfactant may have to
2
1
4 (91%): H-NMR (360 MHz, CDCl ):
3
amount of geraniol was also detected in the be considered, further experimental investi- 8.05−7.97 (m, 2 H); 7.71−7.62 (m, 1 H);
headspace of the irradiated sample. As 2- gations are required to support our prelimi- 7.58−7.47 (m, 2 H); 7.34−7.26 (m, 2 H);
oxo-(2-phenyl)acetates are relatively labile nary data.
to ester hydrolysis this seems unsurprising
at first view. Partial hydrolysis of the pre-
cursors could always give rise to a certain 3. Conclusion
amount of alcohol, especially if the com-
7.13−7.05 (m, 2 H); 4.29 (ABX, J = 10.7,
5.9, 1 H); 4.21 (ABX, J = 10.7, 6.3, 1 H);
2.73 (ABX, J = 13.6, 6.5, 1 H); 2.51 (ABX,
J = 13.6, 7.5, 1 H); 2.33−2.17 (m, 1 H); 1.31
1
3
(s, 9 H); 1.01 (d, J = 6.7, 3 H). C-NMR
pounds are exposed to acidic or alkaline
Due to their ease of preparation from (90.6 MHz, CDCl ): 186.44 (s); 164.08 (s);
3
aqueous media. However, it is interesting to inexpensive starting materials, 2-oxo-(2- 149.01 (s); 136.32 (s); 134.92 (d); 132.48
note that the release of the alcohol seems to phenyl)acetates were found to be very ef- (s); 130.01 (d); 128.93 (d); 128.78 (d);
depend on the light intensity of irradiation ficient delivery systems for the controlled 125.28 (d); 70.23 (t); 39.04 (t); 34.49 (d);
(
as also shown in other cases) and that no release of volatile aldehydes and ketones. 34.38 (s); 31.38 (q); 16.79 (q).
evidenceforalcoholformationwasobtained Photoirradiation with outdoor sunlight or a
1
5 (93%): H-NMR (360 MHz, CDCl ):
3
either upon irradiation of the precursors in xenon lamp in the presence of oxygen re- 8.04−7.96 (m, 2 H); 7.70−7.61 (m, 1 H);
solution or in a series of other water-based leased the desired aldehydes and ketones 7.51 (t, J = 7.7, 2 H); 5.36−5.29 (m, 1 H);
applications such as all-purpose cleaner in solution, as well as on cotton in a fabric 4.32 (dd, J = 7.7, 1.4, 2 H); 2.45−2.31 (m,
films on glass or hair conditioners on hair. softener application. In contrast to irradia- 1 H); 2.21−2.09 (m, 1 H); 2.06−1.84 (m, 2
Th e formation of alcohol thus seems to be
tions in solution or in other types of prac- H); 1.69−1.41 (m, 2 H); 1.64 (s, 3 H); 0.91
1
3
restricted to fabric softeners.
tical applications, the oxo(phenyl)acetates (d, J = 7.1, 3 H). C-NMR (90.6 MHz, CD-
If one considers the shape of the curve release a certain amount of the correspond- Cl , major isomer): 186.52 (s); 164.18 (s);
3
corresponding to the citral released from 6 ing alcohol with the desired aldehyde (or 134.90 (d); 133.06 (s); 132.48 (s); 129.97
together with that of the geraniol formed ketone) after being deposited on cotton (d); 128.92 (d); 126.50 (d); 67.98 (t); 35.91

PHOTOCHEMISTRY
664
CHIMIA 2007, 61, No. 10
(
(
d); 30.56 (d); 29.21 (t); 23.47 (q); 21.23 [3] a) ‘Dynamic Studies in Biology: Photo-
t); 15.67 (q).
triggers, Photoswitches and Caged Bi-
omolecules’, Eds. M. Goeldner, R. Gi-
vens, Wiley-VCH, Weinheim, 2005; b) G.
Mayer, A. Heckel, Angew. Chem., Int. Ed.
1
1
2 (95%): H-NMR (360 MHz, CDCl ):
3
7
4
4
.89−7.82 (m, 2 H); 7.66−7.58 (m, 1 H);
.49−7.40 (m, 2 H); 7.36−7.20 (m, 5 H);
.62 (t, J = 7.1, 2 H); 3.08 (t, J = 6.9, 2 H).
2
006, 45, 4900.
1
3
[4] G. Dormán, G. D. Prestwich, Trends Bio-
technol. 2000, 18, 64.
5] A. Herrmann, The Spectrum (Bowling
Green) 2004, 17(2), 10−13 and 19 and re-
ferences cited therein.
C-NMR (90.6 MHz, CDCl ): 186.28 (s);
3
1
63.72 (s); 136.96 (s); 134.87 (d); 132.32
[
(
1
s); 130.02 (d); 129.01 (d); 128.85 (d);
28.68 (d); 126.86 (d); 66.40 (t); 34.93 (t).
[
[
6] A. Herrmann, Angew. Chem., Int. Ed.
Synthesis of
2
007, 46, 5836.
2
-Oxo-(2-phenyl)acetates by
7] a) C. H. Bamford, R. G. W. Norrish, J.
Chem. Soc. 1935, 1504; b) J. C. Scaia-
no, E. A. Lissi, M. V. Encina, Rev. Chem.
Intermed. 1978, 2, 139 and standard text-
books of organic photochemistry.
Transesterification
A solution of 2, alcohol (1.2 equiv.)
and 1 ml of NaOCH (30% in methanol)
3
in cyclohexane (ca. 150 ml for 40 mmol of
2
) was heated under reflux for 112 h. Af-
[
8] a) S. Hu, D. C. Neckers, J. Org. Chem.
ter cooling to room temperature the reac-
tion mixture was taken up in ether, washed
with water (2x), dried (Na SO ), filtered
1
997, 62, 564; b) S. Hu, D. C. Neckers, J.
Org. Chem. 1997, 62, 6820.
2
4
[9] J. Pika, A. Herrmann, C. Vial, to Firme-
nich SA, WO 99/60990, 1999, Chem.
Abstr. 2000, 132, 15499.
[
[
[
and concentrated. Column chromatogra-
phy (SiO , heptane/ether 95:5) afforded the
2
product.
10] S. Rochat, C. Minardi, J.-Y. de Saint Lau-
mer, A. Herrmann, Helv. Chim. Acta 2000,
1
7
(76%): H-NMR (360 MHz, CDCl ):
3
7
7
5
.97 (d, J = 7.5, 2 H); 7.64 (t, J = 7.5, 1 H);
.49 (t, J = 7.7, 2 H); 5.76−5.48 (m, 3 H);
.43 (s, br., 1 H); 2.14 (d, J = 9.1, 1 H); 2.00
8
3, 1645.
11] A. Herrmann, C. Debonneville, V. Laub-
scher, L. Aymard, Flavour Fragr. J. 2000,
(
s, br., 2 H); 1.58 (s, 3 H); 1.51−1.34 (m, 1
H); 1.47 (dd, J = 6.3, 2.0, 3 H); 1.23−1.14
m, 1 H); 0.90 (s, 3 H); 0.83 (d, J = 5.9, 3 H).
1
5, 415.
12] a) S. Hu, D. C. Neckers, J. Photochem.
Photobiol. A, Chem. 1998, 118, 75; b) M.
(
1
3
C-NMR (90.6 MHz, CDCl ): 186.69 (s),
3
C. Pirrung, R. J. Te pper, J. Org. Chem.
995, 60, 2461.
1
63.48 (s), 136.17 (d); 135.97 (d); 134.78
1
(
1
d); 133.28 (s); 132.55 (s); 129.97 (d); [13] Th e synthesis of 6 by transesterification
29.80 (d); 128.85 (d); 121.53 (d); 74.09
of 2 with geraniol in the presence of di-
octyltinoxide turned out to be successful
(S. Linder, R. L. Snowden, unpublished
results).
(
(
2
2
d); 73.95 (d); 53.97 (d); 53.94 (d); 32.03
s); 31.98 (s); 31.54 (t); 31.37 (t); 27.56 (q);
7.41 (q); 26.96 (q); 26.86 (q); 23.02 (t);
2.84 (q); 22.77 (q); 20.54 (q).
[14] S. Hu, D. C. Neckers, J. Org. Chem. 1996,
6
1, 6407.
Acknowledgements
[15] Th is value corresponds to the total ener-
We thank Dr. Esther Oliveros for fruitful
discussions and Dr. Roger Snowden for his
constructive comments on the manuscript.
gy between 300 and 800 nm. Note that
the irradiation energy is dependent on
the lifetime of the lamp. We measured
an average irradiation intensity of about
Received: July 27, 2007
8
1
0000−100000 lux, which corresponds to
-2
15−150 W m . Th ese values were found
[
1] See for example:A. M. Braun, M.- T. Mau-
to be comparable to outdoor sunlight irra-
rette, E. Oliveros, ‘Photochemical Te chno-
logy’, J. Wiley & Sons, Chichester, 1991.
diation (60000−100000 lux).
[
16] B. Levrand, A. Herrmann, Photochem.
[
2] a) V. N. R. Pillai, Synthesis 1980, 1; b) C.
G. Bochet, J. Chem. Soc., Perkin Trans.
Photobiol. Sci. 2002, 1, 907.
1
2002, 125; c) A. P. Pelliccioli, J. Wirz,
Photochem. Photobiol. Sci. 2002, 1, 441.