Stabilized hemiacetal complexes as precursors for the controlled release of bioactive volatile alcohols

Article abstract of DOI:10.1002/cbdv.201100383

Hemiacetals of pyridine-2-carbaldehyde derivatives and volatile alcohols can be stabilized in organic solution in the presence of protons or different metal cations. Despite the inherent instability of hemiacetals in H ₂O, stabilizing them with zinc(II) triflate and adding them to a cationic surfactant formulation resulted in the slow release of the alcohol from cotton surfaces being treated with the hemiacetal complex. Stabilized hemiacetals might thus be suitable delivery systems of bioactive volatiles by rapid hydrolysis in H₂O-based media.

Full text of DOI:10.1002/cbdv.201100383

CHEMISTRY & BIODIVERSITY – Vol. 9 (2012)
689
Stabilized Hemiacetal Complexes as Precursors for the Controlled Release of
Bioactive Volatile Alcohols
by Barbara Buchsa), Wolfgang Fiebera), Dusˇan Drahonˇovskyb), Jean-Marie Lehnb), and
´
Andreas Herrmann*a)
a
´
) Firmenich SA, Division Recherche et Developpement, 1 Route des Jeunes, B.P. 239,
`
CH-1211 Geneve 8 (e-mail: andreas.herrmann@firmenich.com)
b
´
´
) ISIS, Universite de Strasbourg, 8 Allee Gaspard Monge, B.P. 70028, F-67083 Strasbourg Cedex
Hemiacetals of pyridine-2-carbaldehyde derivatives and volatile alcohols can be stabilized in organic
solution in the presence of protons or different metal cations. Despite the inherent instability of
hemiacetals in H2O, stabilizing them with zinc(II) triflate and adding them to a cationic surfactant
formulation resulted in the slow release of the alcohol from cotton surfaces being treated with the
hemiacetal complex. Stabilized hemiacetals might thus be suitable delivery systems of bioactive volatiles
by rapid hydrolysis in H2O-based media.
Introduction. – Bioactive volatile compounds play an important role as flavors and
fragrances in our everyday life [1]. As a consequence of their high vapor pressures,
volatiles rapidly evaporate from various surfaces, which limits their duration of
perception. The development of fragrance precursors, so-called profragrances, which
release the volatile compound by covalent-bond cleavage under mild environmental
conditions, has recently been described as an interesting possibility to increase the long-
lasting character of fragrance performance in consumer products [2].
Acetals or ketals, sometimes as a mixture with hemiacetals, have been obtained by
the reaction of alcohols with carbonyl derivatives and investigated as hydrolytically
cleavable profragrances to control the release of alcohols and/or carbonyl compounds
[2][3]. Generally, acetals or ketals (in particular, cyclic ones) are relatively stable
under very mild reaction conditions and are, therefore, only of limited interest for the
release of volatiles in practical applications. The hydrolysis of acetals and ketals is
generally acid catalyzed [4] and proceeds via the formation of an unstable hemiacetal,
which is formed in small amounts and is thus usually difficult to isolate (Scheme, a)
[5][6]. On the other hand, if hemiacetals could be generated in sufficient quantities by
suitable stabilization in the reaction medium, their inherent instability should allow fast
release kinetics, resulting in the rapid evaporation of the corresponding volatile
compounds. Some acyclic hemiacetals have been generated or isolated under various
conditions: for example, from electron-withdrawing aldehydes and small alcohols [7],
in the solid state by melting a mixture of fatty alcohols and aldehydes [8] or by
benefiting from particularly strong intramolecular H-bonding [9], in ionic and
molecular liquids [10], as methyl hemiacetals of some tripeptide aldehydes [11], as
polymer-bound systems in a sterically demanding environment [12], or as a hydrolyti-
ꢀ 2012 Verlag Helvetica Chimica Acta AG, Zꢁrich

690
CHEMISTRY & BIODIVERSITY – Vol. 9 (2012)
cally cleavable moiety for polymer degradation [13]. They have also been stabilized as
inclusion complexes by surrounding them with suitably designed cavitands [14].
Hemiacetals formed by reaction of pyridine-2-carbonyl derivatives [15] with different
alcohols have successfully been stabilized by protonation or by metal cation
coordination with the electron-donating N-atom of the pyridine moiety and the OH
group of the hemiacetal (Scheme,b) [10][16][17]. They may thus also be implemented
in the context of dynamic covalent chemistry [17].
Scheme. Reversible Formation of Hemiacetals and Acetals (a) and of Stabilized Hemiacetal Metal Cation
Complexes (b)
Stabilized hemiacetals of pyridine-2-carbaldehyde (1; Fig. 1) or pyridine-2,6-
dicarbaldehyde (2) with various primary, secondary, or tertiary alcohols were obtained
in equilibrium with the corresponding free aldehyde and alcohols in yields of up to
80%, whereas, in the absence of a stabilizing proton (e.g., from CF3COOH (TFA)) or
metal cation (such as Zn2þ or Pb2þ ), considerable amounts of hemiacetals were formed
only when a large excess of alcohol was used [17]. The optimum ratio between the
ligand and the stabilizing cation was determined to be 2 :1 (for Zn2þ ) or 1:1 (for Pb2þ ).
As expected, primary alcohols were found to be more reactive than secondary or
tertiary alcohols, whereas phenols did not react at all [17].
Fig. 1. Carbonyl derivatives allowing the preparation of
stabilized hemiacetals
Encouraged by these findings, we were interested to see whether this concept might
apply to the release of fragrance alcohols under practical application conditions in
perfumery1). With lead complexes being undesirable as constituents in consumer
articles, we focused our interest on the study of zinc complexes.
1
)
Parts of this publication are the subject of a patent application [18].

CHEMISTRY & BIODIVERSITY – Vol. 9 (2012)
691
Results and Discussion. – Formation of Stabilized Hemiacetals in Solution. The
1
formation of stabilized hemiacetals was investigated by H-NMR spectroscopy with
commercially available aldehydes 1 and 2, and a series of primary fragrance alcohols,
notably (R)-3,7-dimethyloct-6-en-1-ol (citronellol; A), (E)-3,7-dimethylocta-2,6-dien-
1-ol (geraniol; B), 2-phenylethanol (C), and (Z)-hex-3-en-1-ol (D). In general, the
aldehyde concentration was fixed at 0.15 mol lꢀ1, while the concentrations of the
alcohol and the stabilizing proton or metal cation were adjusted with respect to the
concentration of the aldehyde. In a typical experiment, 0.9 mmol of the aldehyde, a
mol-equiv. of fragrance alcohol, and 0.5 mol-equiv. of stabilizing salt were weighed into
MeCN and pipetted into an NMR tube. After mixing the compounds, the samples were
left to stand overnight and analyzed the next day. The 1H-NMR spectra were acquired
with 320 scans, and quantitative data were calculated by integrating the signals of the
H-atoms at the C-atom adjacent to the O-atom of the alcohol moiety for both the
hemiacetal and the free alcohol in the compound mixture. The peaks were in general
well separated; overlapping peaks were observed in only a few cases.
In a first set of experiments (carried out in CDCl3), we determined the amount of
non-stabilized hemiacetal formed by reaction of aldehydes 1 and 2, with alcohols A and
B, respectively, in the absence of stabilizing protons or metal cations. The data obtained
for the different product distributions are compiled in Table 1.
As expected with equimolar amounts of alcohol, less than 10% of hemiacetal was
obtained with aldehyde 1 (Table 1, Entries 1 and 2). In the case of aldehyde 2, ca. 20–
30% of mono-hemiacetal, but no bis-hemiacetal, was formed (Entries 27 and 28).
Adding 3 mol-equiv. of TFA (with respect to the aldehyde) considerably increased the
amount of hemiacetal detected in the mixture (Entries 3–6). In the case of geraniol
(B), the solvent does not seem to play a major role for the formation of the hemiacetal.
Comparable results were obtained with CDCl3 or CD3CN, whereas, in the case of
citronellol (A) as the alcohol, about twice as much hemiacetal was formed when
CD3CN was used as the solvent.
As strongly acidic conditions are rarely encountered in practical applications, the
stabilization of the hemiacetals by protic acids is only of limited interest. In our further
investigations, we thus focused on the use of metal cations to stabilize the hemiacetals
of fragrance alcohols. Tests were carried out with different fragrance alcohols by using
0.5 equiv. of zinc triflate (Zn(CF3SO3)2, Zn(OTf)2), or zinc chloride (ZnCl2) as
stabilizing salts in CD3CN (Entries 7–14). Our data indicated that the nature of the
anion seems to be important for the stabilization of the hemiacetal complexes. When
ZnCl2 was used, only about half as much of the hemiacetal complexes were formed
as compared with the measurements carried out with Zn(OTf)2 as the stabilizing
salt.
Higher amounts of hemiacetals were also formed by using dicarbaldehyde 2 for the
formation of the complexes. In the presence of Zn(OTf)2, 77% of hemiacetal was
formed with citronellol (A) in CD3CN, while the formation of the corresponding bis-
hemiacetal was not observed (Entry 29). 2-Oxoacetic acid also formed stabilized
hemiacetals under the present conditions. Use of citronellol (A) as the fragrance
alcohol and Zn(OTf)2 as the stabilizing salt resulted in the formation of 66% of
hemiacetal in CD3CN. Adding 3 equiv. of TFA instead of 0.5 equiv. of Zn(OTf)2
resulted in the formation of gels.

692
CHEMISTRY & BIODIVERSITY – Vol. 9 (2012)
Table 1. Amounts of Free Alcohol and Corresponding Hemiacetal Determined by 1H-NMR in CDCl3 or CD3CN
for an Equilibrated Mixture of Fragrance Alcohols A–D, and Aldehydes 1 and 2 in the Presence or Absence of
TFA or Zn2þ Salts as Stabilizing Agents
Entry Aldehyde (solvent)
Equiv. of
alcohol
Equiv. of TFA or
stabilizing salt
Amount of Amount of
remaining mono-hemi- bis-hemi-
alcohol [%] acetal acetal
formed [%] formed [%]
Amount of
1
2
3
4
5
6
7
8
9
1 (CDCl3)
1 (CDCl3)
1 (CD3CN)
1 (CD3CN)
A: 1.0 equiv.
B: 1.0 equiv.
A: 1.0 equiv. TFA: 3.0 equiv.
B: 1.0 equiv. TFA: 3.0 equiv.
A: 1.0 equiv. TFA: 3.0 equiv.
B: 1.0 equiv. TFA: 3.0 equiv.
A: 1.0 equiv. Zn(OTf)2: 0.5 equiv. 61
B: 1.0 equiv. Zn(OTf)2: 0.5 equiv. 60
C: 1.0 equiv. Zn(OTf)2: 0.5 equiv. 49
D: 1.0 equiv. Zn(OTf)2: 0.5 equiv. 64
A: 1.0 equiv. ZnCl2: 0.5 equiv.
B: 1.0 equiv. ZnCl2: 0.5 equiv.
C: 1.0 equiv. ZnCl2: 0.5 equiv.
D: 1.0 equiv. ZnCl2: 0.5 equiv.
–
–
97
94
75
43
44
42
3
6
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
0
0
0
25
57
56
58
39
40
51
36
20
21
25
18
3
33
43
48
51
37
28
40
37
34
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
1 (CD3CN)
1 (CD3CN)
80
79
75
82
B: 1.0 equiv. Zn(OTf)2: 0.0 equiv. 97
Zn(OTf)2: 0.2 equiv. 67
Zn(OTf)2: 0.3 equiv. 57
Zn(OTf)2: 0.5 equiv. 52
Zn(OTf)2: 0.7 equiv. 49
Zn(OTf)2: 1.0 equiv. 63
1 (CD3COCD3)
A: 1.0 equiv. Zn(OTf)2: 0.5 equiv. 72
B: 1.0 equiv. Zn(OTf)2: 0.5 equiv. 60
C: 1.0 equiv. Zn(OTf)2: 0.5 equiv. 63
D: 1.0 equiv. Zn(OTf)2: 0.5 equiv. 66
1 (CD3CN/D2O 2 : 1) A: 1.0 equiv. Zn(OTf)2: 0.5 equiv. 97
B: 1.0 equiv. Zn(OTf)2: 0.5 equiv. 95
2 (CDCl3)
5
A: 2.0 equiv.
B: 2.0 equiv.
A: 2.0 equiv. Zn(OTf)2: 0.5 equiv. 23
B: 1.0 equiv. Zn(OTf)2: 0.0 equiv. 87
Zn(OTf)2: 0.2 equiv. 58
–
–
81
74
19
26
77
13a)
42a)
52a)
63a)
66a)
80a)
7
2 (CD3CN)
2 (CD3CN)
Zn(OTf)2: 0.3 equiv. 48
Zn(OTf)2: 0.5 equiv. 37
Zn(OTf)2: 0.7 equiv. 34
Zn(OTf)2: 1.0 equiv. 20
2 (CD3CN/D2O 2 : 1) A: 2.0 equiv. Zn(OTf)2: 0.5 equiv. 93
0
a) Sum of mono- and bis-hemiacetals.
To determine the ideal stoichiometric relationship for hemiacetal formation, we
varied the amount of cation added to the solution between 0 and 1 mol-equiv. Data
obtained for the reaction between aldehydes 1 and 2 with geraniol (B) are collected in
Table 1 (Entries 15–20 and 30–35). As shown in [17], the optimum ligand/Zn2þ ratio

CHEMISTRY & BIODIVERSITY – Vol. 9 (2012)
693
for monoaldehyde 1 is ca. 2 :1 (which corresponds to 0.5 equiv. of Zn(OTf)2). Addition
of more Zn2þ to the solution reduced the amount of hemiacetal formed. The structures
of zinc complexes 3–6 obtained by reaction of aldehyde 1 with fragrance alcohols A–D
in the presence of Zn(OTf)2 should thus correspond to those illustrated in Fig. 2. In the
case of dicarbaldehyde 2, the 1H-NMR signals of the mono- and bis-hemiacetals could
not be distinguished. Nevertheless, a constant increase of hemiacetal formation was
observed with an increasing amount of Zn2þ salt added.
Fig. 2. Proposed structures of hemiacetals 3–6 obtained by equilibration of 1 with fragrance aldehydes A
– D, respectively, in the presence of Zn(OTf)2 as the stabilizing salt

694
CHEMISTRY & BIODIVERSITY – Vol. 9 (2012)
The structures of the hemiacetal complexes 3–6 were analyzed by 1H- and
1
13C-NMR spectroscopy. Fig. 3 shows the enlarged aromatic region of the 2D H,13C-
HSQC spectrum of complex 3. Two sets of data can be distinguished, notably the peaks
of the unreacted aldehyde 1 and those corresponding to the hemiacetal formed by
reaction of 1 with citronellol (A) in the Zn2þ complex. A striking feature of the
13C-NMR signals of the hemiacetal complex is that a significant line broadening was
observed for the doublets of the C-atoms in ortho- and para-position to the N-atom
(which almost disappeared in the baseline of the spectrum). Furthermore, a general
broadening of the signals corresponding to the aromatic H-atoms of the pyridine
moiety was noticed, while the signals of the free aldehyde remained narrow and well-
resolved. This particular line-broadening pattern, which was attributed to the chemical
exchange of the hemiacetal in the Zn2þ complex, turned out to be a general feature that
was observed for all hemiacetal complexes studied in this work. The NMR spectra
recorded for the same hemiacetals in the absence of Zn2þ did not reveal any line-
broadening effects (data not shown). Therefore, the strong line broadening indicates
the successful complexation of the hemiacetals to Zn2þ and thus confirms the structure
assignment of complexes 3–6.
1
Fig. 3. Enlarged aromatic region of the H,13C-HSQC spectrum of stabilized hemiacetal 3 together with
unreacted aldehyde 1 and citronellol (A)

CHEMISTRY & BIODIVERSITY – Vol. 9 (2012)
695
The solubility of the stabilizing salt turned out to be one of the most important
prerequisites for the successful generation of the hemiacetal complexes. Other organic
solvents that allow dissolution of a series of metal salts are dimethyl sulfoxide (DMSO)
or acetone. Acetone, in particular, is interesting, as this solvent is compatible with the
targeted perfumery applications. While DMSO turned out to be unsuitable for the
formation of the hemiacetal complexes, Zn(OTf)2 was successfully dissolved in acetone
(after sonication). Table 1 (Entries 21–24) lists the amounts of free alcohol and
stabilized hemiacetal complexes obtained from the reaction of aldehyde 1 with
fragrance alcohols A–D in deuterated acetone. The yields of hemiacetals formed are
slightly lower in acetone than in CD3CN, but still in a range that should be interesting
for practical application2). With Cu(OTf)2 as the stabilizing salt, we obtained ca. 15%
of hemiacetal in a mixture of aldehyde 1 and geraniol (B).
Because H2O, as well as alcohols such as MeOH or EtOH, competes with the
fragrance alcohols to hydrate aldehydes, a decrease of the amount of stabilized
hemiacetals in favor of the hydrated aldehyde was expected in aqueous solution. To
obtain information about the efficiency of hemiacetal formation in an aqueous
environment, we performed measurements in CD3CN/D2O 2 :1. For the analysis, the
aldehyde and the fragrance alcohol were dissolved in CD3CN, and the stabilizing
Zn(OTf)2 in D2O. Pipetting equivalent amounts of each solution into an NMR tube
resulted in the final mixture of CD3CN/D2O 2 :1. As in previous experiments, the final
aldehyde concentration was kept at 0.15 mol lꢀ1; 0.5 mol-equiv. of Zn(OTf)2 were used
with respect to the aldehyde concentration. The data in Table 1 (Entries 25, 26, and 36)
indicated that the formation of hemiacetals was reduced in the presence of H2O. The
percentages of hemiacetals varied between 3 and 5% in the case of aldehyde 1
(Entries 25 and 26), amounts comparable to those measured in the absence of the
stabilizing cations (Entries 1 and 2). In the presence of H2O, the amount of hemiacetal
formed from dialdehyde 2 and citronellol (Entry 36) dropped to ca. 1/3 of that in the
non-stabilized mixture studied previously (Entries 27 and 28).
With the formation of the hemiacetals proceeding quite rapidly [17], the presence
of H2O should hydrolyze the stabilized complexes instantaneously. As the next step, we
investigated the release of the volatile fragrance alcohol from hemiacetal 5 under more
realistic conditions, in particular in the presence of H2O.
Release of Alcohols from Stabilized Hemiacetals in H2O. H2O is the most common
solvent in perfumery, with many cleaning and conditioning agents being based on
aqueous formulations. The performance of functional perfumery formulations, such as
bodycare or household products, is often judged on the freshness and long-lasting
character of the fragrance perception. The cleaning and softening of textiles represents
a typical example of an application requiring an improved duration of fragrance release
over several days. Fabric detergents are conceived to remove hydrophobic substances
from the fabric surface or to prevent their deposition. It is, therefore, particularly
challenging to deposit large amounts of fragrances during fabric cleaning. Therefore,
fragrances are often deposited during the fabric conditioning (softening) step following
the washing process.
2
)
A weak competition of acetone in forming the stabilized hemiacetal complex cannot be ruled out.
However, this possibility has not been examined in further detail within this work.

696
CHEMISTRY & BIODIVERSITY – Vol. 9 (2012)
In the present work, we investigated the performance of the stabilized hemiacetals
described earlier in improving the long-lasting property of fragrance perception in
fabric cleaning and conditioning. The evaporation of the fragrances released from the
stabilized hemiacetals on dry cotton was analyzed by dynamic headspace analysis [19]
after the washing or conditioning process. Dynamic headspace analysis allows
quantification of the amount of fragrance above the cotton surface without requiring
complicated sample preparation. The desired increased duration of fragrance
perception is achieved if the samples containing the stabilized hemiacetal give rise to
higher headspace concentrations than that of a reference sample containing an
equimolar amount of the corresponding unmodified fragrance alcohol.
Solid powder detergents are interesting supports for the stabilized hemiacetal
complex, as the dry powder should allow storage of the complex without major
decomposition due to hydrolysis. To simulate a powder-detergent washing process, we
prepared solutions of stabilized hemiacetal complex 5 and a reference sample of
unmodified 2-phenylethanol (C) in MeCN, and left them standing for 3 d. The solutions
were then added to a commercially available non-perfumed powder detergent, which
was then placed into two stainless steel containers. After addition of tap water, two
cotton sheets were added to each of the containers and the containers rotated in a water
bath to simulate a machine washing process. The cotton sheets were then rinsed by
agitating them manually in cold tap water before line-drying them overnight. Two of
the sheets (one with the hemiacetal complex and one without) were analyzed the next
day.
The dry cotton sheets were then put into a dynamic headspace sampling cell (ca.
160 ml) and exposed to a constant airflow [20]. After equilibrating the system for a few
minutes, the volatiles were adsorbed on a clean poly(2,6-diphenyl-p-phenylene oxide)
(Tenaxꢂ) cartridge at constant time intervals. The cartridges were then thermally
desorbed, and headspace concentrations (in ng lꢀ1) were obtained by external standard
calibrations of the corresponding fragrance alcohol. The data in Table 2 and Fig. 4
illustrate that a slight long-lasting effect of fragrance release was achieved after 1 d.
At the beginning of the measurements (for ca. 150 min), the headspace concen-
trations increased before reaching a plateau or slightly decreasing again. This initial rise
Table 2. Average Dynamic Headspace Concentrations and Standard Deviations (in parentheses) of 2-
Phenylethanol (C) Measured on Dry Cotton after Washing with a Powder Detergent Containing the
Hemiacetal Stabilized with Zn(OTf)2 (5) or an Equimolar Amount of Unmodified 2-Phenylethanol
(reference)
Time [min]
1 day
1 day
Reference [ng lꢀ1
]
From hemiacetal stabilized with Zn(OTf)2 (5) [ng lꢀ1
]
45
105
165
225
285
345
405
465
0.8 (ꢁ 0.5)
1.1 (ꢁ 1.0)
1.2 (ꢁ 1.1)
1.0 (ꢁ 0.7)
0.9 (ꢁ 0.6)
0.9 (ꢁ 0.6)
0.8 (ꢁ 0.5)
0.8 (ꢁ 0.5)
0.9 (ꢁ 0.3)
1.4 ( ꢁ 0.8)
1.5 (ꢁ 0.8)
1.5 (ꢁ 0.7)
1.6 (ꢁ 0.6)
1.6 (ꢁ 0.5)
1.7 (ꢁ 0.4)
1.8 (ꢁ 0.4)

CHEMISTRY & BIODIVERSITY – Vol. 9 (2012)
697
Fig. 4. Comparison of the headspace concentrations of
2-phenylethanol released from stabilized hemiacetal
*
*
complex 5 (— —) and free 2-phenylethanol (— —)
measured in a powder detergent application on dry
cotton after drying for 1 day
was generally observed for this type of measurement and attributed to the conditioning
of the line-dried cotton sheet in the headspace sampling cell [20]. Furthermore, because
the headspace measurements were carried out under realistic everyday conditions, the
measured concentrations showed large variations, probably due to the lack of control of
various parameters (such as temperature, humidity, and convection) during the line
drying of the cotton sheets. Nevertheless, relative values obtained by comparison to the
respective reference samples were found to be quite reproducible.
In an additional experiment, we analyzed the performance of the stabilized
hemiacetal with respect to a reference sample in a surfactant-based aqueous
formulation such as a fabric softener. Cationic surfactants, such as quaternized
triethanolamine esters of fatty acids (TEA-esterquats) [21], have fabric-softening
properties and are efficiently deposited onto cotton from aqueous media. Fragrances,
as well as fragrance precursors, are expected to be incorporated into the surfactant
structure and, thus, to be carried to the cotton surface [22].
For the measurements, a solution of hemiacetal 5 (stabilized with Zn(OTf)2) was
prepared in MeCN and left standing for 3 d before being added to a TEA-esterquat
formulation. The mixture was then diluted with demineralized H2O, and a cotton sheet
was added, which was manually stirred for 3 min and left standing for 2 min. The sheet
was then wrung out and line-dried overnight. Another cotton sheet prepared in the
same way was left drying for 3 d. The dry cotton sheets were then put into a dynamic
headspace sampling cell as described in [20]. After equilibrating, the volatiles were
adsorbed on a clean Tenaxꢂ cartridge at constant time intervals. The cartridges were
then desorbed, and headspace concentrations were obtained by external standard
calibrations. Reference samples containing equimolar amounts of 2-phenylethanol
were prepared and analyzed in the same way. The data are summarized in Table 3 and
illustrated in Fig. 5.
After drying for 1 d, almost identical headspace concentrations were measured for
the free fragrance alcohol and the alcohol released from stabilized hemiacetal 5
(Fig. 5). This indicated that, indeed, the complex rapidly exchanged the fragrance
alcohol against H2O to give free 2-phenylethanol, which then evaporated from the
cotton surface. The fact that the headspace concentrations of 2-phenylethanol released
from 5 slightly increased from ca. 95 ng lꢀ1 after 1 d (Fig. 5,a, solid line) to ca. 110 ng lꢀ1
after 3 d (Fig. 5,b, solid line), while those of the corresponding free alcohol in the
reference sample decreased from ca. 95 ng lꢀ1 after 1 d (Fig. 5,a, dotted line) to ca.

698
CHEMISTRY & BIODIVERSITY – Vol. 9 (2012)
Table 3. Average Dynamic Headspace Concentrations and Standard Deviations (in parentheses) of 2-
Phenylethanol (C) Measured on Dry Cotton after Washing with a Cationic Surfactant Formulation
Containing the Hemiacetal Stabilized with Zn(OTf)2 (5) or an Equimolar Amount of Unmodified 2-
Phenylethanol (reference)
Time [min] 1 day
Reference [ng lꢀ1
3 days
Reference [ng lꢀ1
]
From hemiacetal
stabilized with
Zn(OTf)2 (5) [ng lꢀ1
]
From hemiacetal
stabilized with
Zn(OTf)2 (5) [ng lꢀ1
]
]
30
90
14.4 (ꢁ 9.0)
33.9 (ꢁ 7.7)
61.0 (ꢁ 27.0)
76.1 (ꢁ 38.8)
90.9 (ꢁ 32.2)
92.8 (ꢁ 38.3)
95.6 (ꢁ 41.1)
94.6 (ꢁ 42.1)
11.5 (ꢁ 0.2)
44.8 (ꢁ 25.3)
78.1 (ꢁ 24.3)
87.4 (ꢁ 26.9)
94.0 (ꢁ 28.8)
95.5 (ꢁ 33.2)
96.7 (ꢁ 31.7)
96.3 (ꢁ 34.7)
12.4 (ꢁ 13.8)
19.8 (ꢁ 18.0)
33.0 (ꢁ 32.1)
62.4 (ꢁ 35.6)
70.8 (ꢁ 25.7)
70.7 (ꢁ 21.5)
77.8 (ꢁ 16.1)
77.4 (ꢁ 11.4)
15.5 (ꢁ 12.7)
44.4 (ꢁ 13.7)
86.9 (ꢁ 7.1)
150
210
270
330
390
450
104.7 (ꢁ 14.6)
108.8 (ꢁ 9.1)
109.2 (ꢁ 13.0)
112.3 (ꢁ 14.2)
112.1 (ꢁ 15.0)
Fig. 5. Comparison of the headspace concentrations of 2-phenylethanol released from stabilized
*
*
hemiacetal complex 5 (— —) and free 2-phenylethanol (— —) measured in a fabric softener application
on dry cotton after drying for 1 day (a) and 3 days (b)
75 ng lꢀ1 after 3 d (Fig. 5,b, dotted line), demonstrated the desired slow release effect
of the stabilized hemiacetal with respect to the reference despite the expected rapid
hydrolysis in the presence of H2O (see also Table 3).
Conclusions. – Hemiacetals of pyridine-2-carbaldehyde derivatives and a series of
volatile alcohols can be stabilized in organic solution in the presence of protons or
different metal cations such as Zn2þ. In MeCN, up to 50% of stabilized hemiacetals
were formed in solution. Acetone can be used as an alternative solvent to prepare the
hemiacetal complexes. Zn(OTf)2 was found to be a versatile stabilizing salt, being quite
soluble in a series of organic solvents.
The performance of the stabilized hemiacetals as delivery systems for the controlled
release of volatile fragrance alcohols was investigated in fabric cleaning and
conditioning applications. Dynamic headspace analysis on dry cotton showed that a
slight long-lasting effect of fragrance release was achieved from a powder detergent

CHEMISTRY & BIODIVERSITY – Vol. 9 (2012)
699
washing process after 1 d. Despite the presence of H2O in the surfactant formulation,
an even more pronounced slow-release effect was achieved in a fabric conditioning
process after 3 d.
Besides applications in functional perfumery, stabilized hemiacetals might also find
use in the screening of pharmaceutically active compounds and receptors. Nevertheless,
on the basis of the current data, it is probably too early to evaluate the full potential of
stabilized hemiacetals as delivery systems for the controlled release of bioactive
compounds.
Experimental Part
General. Commercially available reagents and solvents were used without further purification.
Demineralized H2O was obtained from a Millipore Synergy 185 H2O purifier. 1H- and 13C-NMR spectra
were recorded at 258 on a Bruker 400 MHz DPX spectrometer, d [ppm] downfield from Me4Si as internal
1
standard, J in Hz. Standard pulse sequences and parameters were used for one-dimensional H- and
13C-NMR spectra and for 2D, gradient-selected COSY, NOESY, 1H-,13C-HSQC, and 1H-,13C-HMBC
spectra.
NMR Analysis. In a typical experiment, 0.9 mmol of the aldehyde, 1 mol-equiv. of fragrance alcohol,
and 0.5 mol-equiv. of stabilizing salt were weighed into 2 ml of MeCN. Then, 0.3 ml of each soln. were
pipetted into an NMR tube and left standing overnight. 1H-NMR Spectra were recorded with 320 scans,
and the amount of hemiacetal was determined by integrating the signals of the H-atoms at the C-atom
adjacent to the O-atom of the alcohol moiety for both the hemiacetal and the free alcohol in the
compound mixture. The following signals of the CH2OH and CH2OR groups (in CD3CN and in the
presence of Zn(SO3CF3)2) of the different alcohols and their respective hemiacetals were considered for
the quantification of the compounds in different mixtures: citronellol (A) 3.5 (m) and 3.7 (m) ppm;
geraniol (B) 4.0 (d, CHOH) and 4.2 (d, CHOR) ppm; 2-phenylethanol (C) 3.7 (t) and 3.9 (m) ppm; hex-
3-en-1-ol (D) 3.5 (t) and 3.7 (m) ppm. The formation of hemiacetal complexes 3–6 in CD3CN was also
verified by 13C-NMR spectroscopy (2000 scans).
Complex 3: Zn(OTf)2 (0.5 equiv.), 1 (1 equiv.), and A (1 equiv.). 1H-NMR (400 MHz, CD3CN): 8.60
(br. s, 1 H); 8.22 (t, J¼7.6, 1 H); 7.80 (d, J¼8.1, 1 H); 7.73 (d, J¼6.3, 1 H); 5.99 (s, 1 H); 5.11 (m, 1 H);
3.72 (m, 2 H); 1.97 (m, 2 H); 1.65 (s, 3 H); 1.63 (m, 1 H); 1.58 (s, 3 H); 1.52 (m, 1 H); 1.43–1.10 (m, 3 H);
0.87 (d, J¼6.7, 3 H). 13C-NMR (100.6 MHz, CD3CN): 147.3 (br. d); 147.0 (s); 143.0 (br. d); 131.9 (s);
127.2 (d); 125.8 (d); 125.7 (d); 121.6 (s, 1J (13C,19F)¼320); 93.6 (d); 93.5 (d); 67.8 (t); 67.7 (t); 37.8 (t); 37.7
(t); 37.1 (t); 37.0 (t); 30.3 (d); 30.1 (d); 26.1 (t); 25.9 (q); 19.7 (q); 17.7 (q).
Complex 4: Zn(OTf)2 (0.5 equiv.), 1 (1 equiv.), and B (1 equiv.). 1H-NMR (400 MHz, CD3CN): 8.60
(br. s, 1 H); 8.23 (t, J¼7.6, 1 H); 7.78 (d, J¼7.9, 1 H); 7.27 (t, J¼6.3, 1 H); 6.00 (s, 1 H); 5.34 (t, J¼6.6,
1 H); 5.11 (m, 1 H); 4.23 (d, J¼6.9, 2 H); 2.09 (m, 2 H); 2.00 (m, 2 H); 1.65 (s, 3 H); 1.64 (s, 3 H); 1.59 (s,
3 H). 13C-NMR (100.6 MHz, CD3CN): 147.3 (br. d); 147.3 (s); 143.1 (br. d); 138.9 (s); 132.7 (s); 127.1 (d);
125.7 (d); 124.9 (d); 121.7 (s, 1J (13C,19F)¼320); 120.4 (d); 92.8 (d); 65.7 (t); 40.2 (t); 27.1 (t); 25.9 (q); 17.8
(q); 16.7 (q).
Complex 5: Zn(OTf)2 (0.5 equiv.), 1 (1 equiv.), and C (1 equiv.). 1H-NMR (400 MHz, CD3CN): 8.57
(br. s, 1 H); 8.20 (m, 1 H); 7.72 (t, J¼6.1, 1 H); 7.64 (d, J¼7.3, 1 H); 7.28 (m, 2 H); 7.24–7.19 (m, 3 H);
5.98 (s, 1 H); 3.90 (m, 2 H); 2.91 (t, J¼6.8, 2 H). 13C-NMR (100.6 MHz, CD3CN): 147.2 (s); 147.0 (br. d);
143.3 (br. d); 139.7 (s); 130.0 (d); 129.4 (d); 127.4 (d); 127.3 (d); 125.7 (d); 121.7 (s, 1J (13C,19F)¼320); 93.5
(d); 70.0 (t); 36.5 (t).
Complex 6: Zn(OTf)2 (0.5 equiv.), 1 (1 equiv.), and D (1 equiv.). 1H-NMR (400 MHz, CD3CN): 8.61
(br. s, 1 H); 8.25 (m, 1 H); 7.83 (d, J¼7.8, 1 H); 7.75 (t, J¼6.1, 1 H); 6.00 (s, 1 H); 5.47 (m, 1 H); 5.33 (m,
1 H); 3.69 (m, 2 H); 2.35 (td, J¼6.8, 6.8, 2 H); 2.04 (qd, J¼7.5, 7.5, 2 H); 0.94 (t, J¼7.5, 3 H). 13C-NMR
(100.6 MHz, CD3CN): 147.2 (s); 147.2 (br. d); 143.0 (br. d); 134.9 (d); 127.3 (d); 125.8 (d); 125.7 (d); 121.7
1
(s, J (13C,19F)¼320); 93.5 (d); 69.2 (t); 28.2 (t); 21.3 (t); 14.6 (q).
Dynamic Headspace Analysis on Cotton (from a Powder Detergent). Stabilized hemiacetal 5 was
prepared by mixing 1.2 ml of a soln. containing 96.3 mg of 1 in 2 ml of MeCN, 1.2 ml of a soln. containing

700
CHEMISTRY & BIODIVERSITY – Vol. 9 (2012)
109.8 mg of C in 2 ml of MeCN, and 164.0 mg of Zn(OTf)2 as the stabilizing salt in 2 ml of MeCN.
Similarly, a reference sample was prepared by adding 2.4 ml of MeCN to 1.2 ml of the soln. of C. The
solns. were left standing at r.t. for 3 d.
Then, 1.8 g of a non-perfumed, commercially available powder detergent (Via Professional Sensitive
from Unilever), 2 ml of the stabilized hemiacetal soln., 400 ml of tap water, and two cotton sheets
(EMPA cotton test cloth Nr. 221; pre-washed with an unperfumed detergent powder and cut to ca. 12ꢂ
12 cm) were added to a stainless steel container of a Linitestꢀ washing-machine simulator (Heraeus).
This procedure was repeated with a second container, with the corresponding reference sample replacing
the hemiacetal soln. The containers were fixed inside the machine and rotated at 458C for 20 min. The
cotton sheets were rinsed by agitating them manually in 600 ml of tap water for 2 min and line-dried
overnight. Two of the sheets (one with the hemiacetal complex and one without) were analyzed the next
day. For the analysis, the sheets were each put into a headspace sampling cell (160 ml) thermostatted at
258, and exposed to a constant airflow of 200 ml minꢀ1. The air was filtered through active charcoal and
aspirated through a sat. soln. of NaCl (to ensure a constant humidity of the air of ca. 75%). After
equilibrating for 15 min, the volatiles were adsorbed during 30 min on a clean Tenaxꢀ cartridge and
30 min on a waste cartridge, which was discarded after the measurements. The sampling was repeated
seven times every hour. The cartridges were thermally desorbed on a Perkin-Elmer TurboMatrix 350
desorber coupled to a Perkin-Elmer AutoSystem XL gas chromatograph equipped with a J&W Scientific
DB1 cap. column (30 m, i.d. 0.45 mm, film 0.42 mm) and a Perkin-Elmer TurboMass Upgrade MS
detector. The volatiles were analyzed in the single-ion monitoring (SIM) mode using a two-step temp.
gradient starting from 608 to 1108 at 28 minꢀ1 and then going to 2608 at 458 minꢀ1. Headspace
concentrations (in ng lꢀ1) were obtained by external standard calibration of the corresponding fragrance
alcohol using EtOH solns. of five different concentrations. Each calibration soln. (0.1 ml) was injected
onto Tenaxꢀ cartridges, which were immediately desorbed under the same conditions as those resulting
from the headspace sampling. Average values were obtained from three measurements.
Dynamic Headspace Analysis on Cotton (from a Fabric Softener). Stabilized hemiacetal 5 was
prepared by mixing 0.6 ml of a soln. containing 96.5 mg of pyridine-2-carbaldehyde (1) in 2 ml of MeCN,
0.6 ml of a soln. containing 109.7 mg of C in 2 ml of MeCN, and 0.6 ml of a soln. containing 164.0 mg of
Zn(OTf)2 as the stabilizing salt in 2 ml of MeCN. Similarly, a reference sample was prepared by adding
1.2 ml of MeCN to 0.6 ml of the soln. of C. The solutions were left standing at r.t. for 3 d.
Four beakers, each containing 1.8 g of a fabric softener formulation consisting of Stepantexꢀ VL 90 A
(16.5% (w/w)), CaCl2 (10% aq. soln., 0.6% (w/w)), and demineralized H2O (82.9% (w/w)) were
prepared. Then, 1 ml of the stabilized hemiacetal soln. was pipetted into each of the first two beakers and
1 ml of the reference soln. of C into each of the two other beakers. Then, 600 ml of tap water and one
cotton sheet were added to each beaker. The sheets were agitated manually for 3 min, left standing for
2 min, wrung out by hand, weighed to obtain a constant quantity of residual H2O, and line-dried
overnight. Two of the sheets (one with the hemiacetal and one without) were analyzed the next day, the
other two after 3 d. For the analysis, the sheets were each put into a headspace sampling cell and analyzed
as described in [20]. Headspace concentrations (in ng lꢀ1) were obtained by external standard calibration
as described earlier. Average values were obtained from two measurements.
REFERENCES
[1] E. Breitmaier, ꢃTerpenes – Flavors, Fragrances, Pharmaca, Pheromonesꢄ, Wiley-VCH, Weinheim,
2006; ꢃFlavours and Fragrances – Chemistry, Bioprocessing and Sustainabilityꢄ, Ed. R. G. Berger,
Springer Verlag, Berlin, 2007; ꢃThe Chemistry and Biology of Volatilesꢄ, Ed. A. Herrmann, John
Wiley & Sons, Chichester, 2010.
[2] A. Herrmann, Angew. Chem., Int. Ed. 2007, 46, 5836, and refs. cit. therein; A. Herrmann, in ꢃThe
Chemistry and Biology of Volatilesꢄ, Ed. A. Herrmann, John Wiley & Sons, Chichester, 2010,
pp. 333–362.
[3] H. Kamogawa, Y. Haramoto, T. Nakazawa, H. Sugiura, M. Nanasawa, Bull. Chem. Soc. Jpn. 1981,
54, 1577; G. D. Yadav, A. A. Pujari, Can. J. Chem. Eng. 1999, 77, 489; H. Morinaga, H. Morikawa, Y.

CHEMISTRY & BIODIVERSITY – Vol. 9 (2012)
701
Wang, A. Sudo, T. Endo, Macromolecules 2009, 42, 2229; Y. Wang, H. Morinaga, A. Sudo, T. Endo,
J. Polym. Sci., Part A: Polym. Chem. 2010, 48, 3816.
[4] E. H. Cordes, H. G. Bull, Chem. Rev. 1974, 74, 581.
[5] S. C. Coote, L. H. S. Smith, D. J. Procter, Sci. Synth., Knowl. Updates 2010, 3, 417.
[6] A. Mꢁller, Helv. Chim. Acta 1936, 19, 225; J. M. Bell, D. G. Kubler, P. Sartwell, R. G. Zepp, J. Org.
´
Chem. 1965, 30, 4284; P. Le Henaff, Bull. Soc. Chim. Fr. 1968, 4687; R. B. Jensen, E. C. Munksgaard,
Acta Chem. Scand. 1969, 23, 79; R. Fuchs, T. M. Young, R. F. Rodewald, Can. J. Chem. 1973, 51,
4122; B. Capon, Pure Appl. Chem. 1977, 49, 1001; J. L. Jensen, P. A. Lenz, J. Am. Chem. Soc. 1978,
100, 1291; T. J. Przystas, T. H. Fife, J. Am. Chem. Soc. 1981, 103, 4884; D. Penn, D. P. N. Satchell, J.
Chem. Soc., Chem. Commun. 1982, 54; Y. Chiang, A. J. Kresge, J. Org. Chem. 1985, 50, 5038; P. E.
Sørensen, W. P. Jencks, J. Am. Chem. Soc. 1987, 109, 4675; R. A. McClelland, K. M. Engell, T. S.
Larsen, P. E. Sørensen, J. Chem. Soc., Perkin Trans. 2 1994, 2199; A. L. Balashov, S. M. Danov, V. L.
Krasnov, A. Y. Chernov, A. I. Kvasheninnikov, Russ. J. Gen. Chem. 1998, 68, 1099; K. M. Engell,
R. A. McClelland, P. E. Sørensen, Can. J. Chem. 1999, 77, 978.
[7] M. Hashimoto, T. Isono, K. Mano, Ber. Bunsen-Ges. Phys. Chem. 1994, 98, 793; M. Matsui, K.
Yamada, K. Funabiki, Tetrahedron 2005, 61, 4671.
[8] J. L. E. Erickson, C. R. Campbell Jr., J. Am. Chem. Soc. 1954, 76, 4472; V. Mahadevan, Lipids 1970,
5, 563.
[9] A. Solchinger, K. Wurst, H. Kopacka, B. Bildstein, Cryst. Growth Des. 2007, 7, 2380.
[10] M. D. Soutullo, R. A. OꢄBrien, K. E. Gaines, J. H. Davies Jr., Chem. Commun. 2009, 2529.
[11] J. F. Miller, A. Spaltenstein, Tetrahedron Lett. 1996, 37, 2521.
[12] H. Kamogawa, S. Okabe, M. Nanasawa, Bull. Chem. Soc. Jpn. 1976, 49, 1917; H. Kamogawa, Y.
Haramoto, M. Nanasawa, Bull. Chem. Soc. Jpn. 1979, 52, 846; P. A. Berger, J. R. Garbow, A. M.
DasGupta, E. E. Remsen, Macromolecules 1997, 30, 5178.
[13] B. Reid, S. Tzeng, A. Warren, K. Kozielski, J. Elisseeff, Macromolecules 2010, 43, 9588.
[14] R. J. Hooley, P. Restorp, T. Iwasawa, J. Rebek Jr., J. Am. Chem. Soc. 2007, 129, 15639.
[15] P. Gianni, E. Matteoli, Gazz. Chim. Ital. 1975, 105, 125; S. Huang, A. K. Miller, W. Wu, Tetrahedron
Lett. 2009, 50, 6584.
[16] L. You, E. V. Anslyn, Org. Lett. 2009, 11, 5126; L. You, S. R. Long, V. M. Lynch, E. V. Anslyn, Chem.
Eur. J. 2011, 17, 11017; L. You, J. S. Berman, E. V. Anslyn, Nat. Chem. 2011, 3, 943.
ˇ
´
[17] D. Drahonovsky, J.-M. Lehn, J. Org. Chem. 2009, 74, 8428.
ˇ
´
´
[18] J.-M. Lehn, D. Drahonovsky, A. Herrmann, to Firmenich SA, Universite Louis Pasteur and CNRS,
WO 2010/020954, 2010; Chem. Abstr. 2010, 152, 295943.
[19] ꢃHeadspace Analysis of Foods and Flavors: Theory and Practiceꢄ, Eds. R. L. Rouseff, K. R.
Cadwallader, Kluwer Academic/Plenum Publishers, New York, 2001.
[20] B. Levrand, W. Fieber, J.-M. Lehn, A. Herrmann, Helv. Chim. Acta 2007, 90, 2281; B. Buchs, G.
Godin, A. Trachsel, J.-Y. de Saint Laumer, J.-M. Lehn, A. Herrmann, Eur. J. Org. Chem. 2011, 681.
[21] M. I. Levinson, J. Surfactants Deterg. 1999, 2, 223; F. E. Friedli, H. J. Koehle, M. Fender, M. Watts,
R. Keys, P. Frank, C. J. Toney, M. Doerr, J. Surfactants Deterg. 2002, 5, 211; S. Mishra, V. K. Tyagi, J.
Oleo Sci. 2007, 56, 269; M. Bahmaei, F. Badiee, H. Kasehgari, J. Surfactants Deterg. 2011, 14, 173.
[22] S. D. Escher, E. Oliveros, J. Am. Oil Chem. Soc. 1994, 71, 31; T. Stora, S. Escher, A. Morris, Chimia
2001, 55, 406; S. K. Obendorf, H. Liu, K. Tan, M. J. Leonard, T. J. Young, M. J. Incorvia, J.
Surfactants Deterg. 2009, 12, 43.
Received November 17, 2011