IDENTIFICATION AND ANALYSIS OF PIPERITONE IN RED WINES

Reviewed by Pijar Beyna Fatamorgana

College of Food Science and Technology
Huazhong Agricultural University

Abstract

The present study concerns the search for new aroma compounds associated with the flavor of
aged and prematurely aged red wines. Sensory descriptive analysis associated with gas
chromatography–olfactometry was first performed to find specific odoriferous zones. One of the
zones, reminiscent of mint, was found in red wines with a prune flavor. Thanks to several
analytical approaches including preparative gas chromatography,p-menth-1-en-3-one
(piperitone), has been identified in red wines. The olfactory detection threshold of this terpene
reminiscent of mint was 0.9lg/L in model hydroalcoholic solution and 70lg/L in red wines.
Chiral multidimensional GC–MS was used to show that piperitone was present mainly in the (R)
form in red wines. The concentration of piperitone in aged red wines ranged from a few ng/L to
435 ng/L, but the level was not linked to the age of the wine nor to its premature evolution.

Keywords: Preparative gas chromatography, Piperitone, Red wine, Flavor.

1. Introduction
Aroma is a key attribute for professionals and consumers and is therefore one of the major
traits driving the intrinsic quality of wine. In recent years, research in enology and in wine flavor
chemistry has made it possible to identify and quantify hundreds of volatile compounds
including terpenes, C13-norisoprenoids, thiols, carbonyls, pyrazines and benzene derivatives.
Wine flavor, resulting in the combination of volatile compounds found in grapes, produced
during fermentations and also aging (González-Barreiro, Rial-Otero, Cancho-Grande, & Simal-
Gándara, 2015; RibéreauGayon, 2006), cannot be fully described without an understanding of
the role played by its individual molecular components, their concentrations, odor thresholds and
interactions with other compounds (Ferreira, Ortín, Escudero, López, & Cacho, 2002; Pineau,
Barbe, Van Leeuwen, & Dubourdieu, 2007).
Searching for unknown odorant compounds in wine to assess their contribution to the global
aroma of a sample is tricky because of the need to obtain a well-resolved chromatographic peak
in monodimensional gas chromatography coupled to mass spectrometry (GC–MS) while
analyzing complex matrices. To overcome this drawback, multidimensional separation is used to
increase the peak capacity of the separation system and obtain a better resolution of
chromatographic peaks (Picard et al., 2016; Pons, Lavigne, Frérot, Darriet, & Dubourdieu, 2008).
Basically, GC can be coupled in an offline and an online manner to an interface. Many studies
have been performed in recent years to investigate the value of online systems using ‘‘heart
cutting” to solve tricky chromatographic issues (Chin & Marriott, 2015). Concerning offline
approaches, this technique corresponds to the manual or automated collection of an effluent from
a column prior to manual reinjection to a second column (Danilo Sciarrone, Pantò, Ragonese,
Dugo, & Mondello, 2015). This analytical approach is used to isolate large amounts of material
(through multiple injections of samples) that have already been subjected to a first-dimension
GC separation. The main parameters affecting the quality of the purification of volatile
compounds produced by this technique have already been described by Mandalakis and
Gustafsson (2003).
Less common and less convenient than the online technique, offline GC coupling can be
used to purify and concentrate samples. It has been successfully used to characterize volatile
compounds (Rühle et al., 2009) and identify trace volatile compounds (Ochiai & Sasamoto, 2011;
Pantò et al., 2015; Sciarrone et al., 2012). As mentioned in our first work performed in 2006
dealing with the identification of aroma associated with the premature evolution of wines (Pons,
2006), GC coupled to olfactometry (GC–O revealed an interesting odorant zone (OZ)
reminiscent of mint in several aged red wines. The goal of this study was to identify the volatile
compounds associated with this OZ.

2. Materials and Methods

2.1. Chemicals and reference compounds.
Dichloromethane (Chromasolv grade) and silica gel were from Sigma–Aldrich (St Quentin
Fallavier, France). Anhydrous sodiumsulfate was supplied by Prolabo (France). Trisma base
(99%) and L(+) tartaric acid (99.5%) were from Fluka (France). Ethanol (Lichrosolv grade) was
from Merck (France). Pentane (Normapur quality) from Prolabo (France) was distilled to
improve its quality. The reference compound piperitone (90%) was from ABCR Gmbh & Co. K
(Germany) and from TCI (Belgium).
2.2. Wine samples
Red wines used in this study were from Bordeaux (vintages from 1974 to 2015), produced
by standard winemaking procedures with or without wood contact. They were analyzed in 2006
and in 2015.
2.3. Wine extraction
2.3.1. Procedure 1 (identification procedure)
Red wine (0.5 L) was extracted with pentane/dichloromethane (1/1, v/v;
3 160 mL) dried over Na2SO4 and concentrated to 50 mL by gentle
evaporation of the solvent with a Rotavapor system. The organic phase was treated three times
with an aqueous Trizma buffer (0.1 M; pH 10, 50 mL). Then, the organic extract containing
neutral and basic compounds was dried over Na2SO4 and concentrated to 1 mL under nitrogen
flow.
2.3.2. Procedure 2 (assay procedure)
Red wines (100 mL) were spiked with 100lL 3-octanol (100 mg/L, EtOH) and extracted
three times with pentane/dichloromethane (1/1, v/v; 310 mL). The organic layer was dried over
Na2SO4 and then concentrated to 0.5 mL under nitrogen flow (100 mL/min).
2.4. Partial purification of crude red wine extract on silica gel chromatography
 The concentrated pentane/dichloromethane (1/1) extract (procedure 1) was further purified
by adsorption chromatography on silica gel (Aldrich, 70–230 mesh, 60 Å). Silica (15 g) was
activated at 1200C overnight, then cooled in a dehumidifier, deactivated with 5 mL water and
treated with pentane (100 mL) and placed in a controlled temperature (T=100C) glass flash
chromatography column (10 mm x 100 mm). The extract was then applied on the top of the
column and fractionation was performed by an eluotropic series of solvents (100 mL) under 0.5
bar of nitrogen. Pentane/dichloromethane ratios were distributed as follows: (100/0) (I), (90/10)
(II), (66/33) (III), (10/90) (IV), and (0/100) (V). Each fraction was concentrated under vacuum
until 20 mL and under nitrogen flow to 500lL, then analyzed by GC.
2.5. GC–O–MS apparatus and piperitone quantification
The organic extract obtained from procedure 2 was analyzed on a Star 3400CX gas
chromatograph fitted to a Saturn 2000 electronic Ion Trap Mass Spectrometer from Varian
(France). The column was coated with polyethylene glycol, BP-20 from SGE (France), 50 m x
0.22 mm; film thickness: 0.25lm. The end of the column was connected to a split flow connected
to two deactivated silica columns of the same dimension, one to the mass spectrometer and the
other to the olfactometer (ODO-I). For quantification experiments, the column connected to the
olfactometer was plugged. The carrier gas was He with a flow rate of 1.0 mL/min. Oven
temperature was initially set at 450C for 1 min, then raised to 2300C at 30C/min and maintained
at this temperature for a further 30 min. Trap temperature was maintained at 1700C. Injection
(2lL) was in splitless mode (closure time: 0.75 min). Ionsm/z59 and 110 (SIS mode) were used
to quantify 3-octanol (internal standard) and piperitone, respectively. The linearity was validated
between 20 and 1000 ng/L with a correlation coefficient (R2) of 0.995 in red wine. The LOD was
10 ng/L.

3. Result and Discussion

Identification of volatile compounds extracted from a complex matrix such as wine requires
the use of a combination of separative techniques. Off-line multidimensional gas
chromatography, such as sample analysis by preparative gas chromatography coupled to
olfactometry (P-GC–O) followed by GC–MS analysis was applied to purify red wine extract in
order to identify volatile compounds associated with an OZ reminiscent of mint (OZ1) detected
in aromagrams of prematured aged red wines (Pons, 2006).
3.1. Identification of OZ 1
500 mL of red wine were extracted first with pentane/dichloromethane (1/1). The crude
organic extract was washed with a trizma base solution (pH 10) to remove acids. After that step,
the extract was further purified on silica gel. Each purification step was validated by GC–O to be
sure that the odorant zone reminiscent of mint was retained. Under these experimental conditions,
the minty OZ was obtained only with the pentane–dichloromethane (1/1) fraction. Finally, this
fraction was purified by P-GC–O.
To be sure to collect systematically the compound(s) associated to OZ reminiscent of mint in
the vessel during the 50 injections, it was decided to select a large 1.5 min window around the
retention time of the OZ as determined by P-GC–O. Of course, the drawback of this approach
was the risk of collecting a large number of peaks that potentially overlapped the peak(s) of
interest during monodimensional GC–MS analysis, as already mentioned by Sciarrone et al.
(2012). Once the sample was injected 50 times, the vessel was washed and the solvent
concentrated. The purified sample was analyzed by GC–O and GC–MS.
Table 1 shows an aromagram corresponding to the analysis of purified red wine samples.
The minty zone (OZ1) was detected at the same retention time as another aroma reminiscent of
fruit pit. Moreover, many odorant zones were also detected despite the large purification protocol
applied to the crude wine extract. Thanks to this approach, the analysis of the purified extract by
GC–O–MS revealed a chromatographic peak associated with the OZ1 (Fig. 1) reminiscent of
mint. The corresponding mass spectrum is presented in Fig. 2. Although the fragmentation
spectrum presented some impurities, comparison with the mass spectrum found in a commercial
NIST (National Institute of Standards) database allowed us to identify this peak atp-menth-1-en-
3-one, also called piperitone. This compound has the following retention indices LRI 1730 on
polar BP20 (50 m, 0.22 mm, 0.22lm) column and LRI 1214 on a nonpolar SPB1 (30 m, 0.25 mm,
0.25lm) column. These retention indices were similar to those indicated in the literature: LRI
nonpolar 1230 (Dob & Chelghoum, 2006). Identification was validated by co-injection of the
pure commercial product with the purified red wine organic extract. Moreover, GC–O analysis
of a diluted organic solution of piperitone (1lg/mL, CH2Cl2) also confirmed the identification of
this compound associated with the minty OZ1. Therefore, the presence of piperitone, an
oxygenated terpene chiral reminiscent of an intense minty and camphor flavor, was evidenced
for the first time in premature aged red wine. It is interesting to note the similarity between the
LRI values of this compound which is reminiscent of minty flavor and the odor of another ketone,
3-methyl-2,4-nonanedione (MND), which was identified several years ago in prematured aged
red wines (Pons et al., 2008). The latter compound which is reminiscent of fruit pit and dried
fruits has also been described as modulating the minty nuance of red wines according to its
concentration (Pons, Lavigne, Darriet, & Dubourdieu, 2013). It cannot be ruled out that MND,
which has been described as having also a minty odor and oil.
3.2. Contribution of piperitone to wine flavor
3.2.1. Distribution of piperitone enantiomers in wines
Using chiral MDGC–MS, it was possible to separate the two piperitone enantiomers (Rs=
1.65) in different products such as oil from natural extract and wine. Since we did not possess the
two enantiomers in the laboratory, the two peaks obtained by MDGC–MS corresponding to the
two enantiomers were assigned by comparing the distribution of a pure commercial piperitone
and a commercialM. piperitaoil in our chromatographic conditions (Table 2) to those described
in the literature (Ravid et al., 1994). Thanks to this approach, it was observed that piperitone was
found mainly in the (R) form in red wines (Table 2). Moreover, the enantiomer distribution of
the commercial piperitone used for the sensory test is similar to that found in wine. The use of
commercial piperitone for determining its detection threshold in wines was thus considered for
studying the sensory contribution of this compound in wine.
3.2.2. Detection threshold determination
Pure piperitone has a strong minty camphor-like odor. Using a large panel made up of 60
participants, the olfactory detection threshold was determined to be 0.9lg/L in a model wine
solution. In a young red wine without mint nuances its detection threshold reaches 70lg/L (Table
3). This result was validated in another young red Bordeaux wine without mint nuances (Dth
71.5lg/L). These values were quite different than those estimated by Picard et al. (2016)close to 1
and 2lg/L in a red wine but the methodology was not exactly the same. Moreover, according
toPelosi and Viti (1978)piperitone and compounds with a similar structure such asL-carvone
presented high level of anosmia in humans.

4. Conclusion
This work reports the identification of piperitone, reminiscent of mint, in red wines marked
by dried fruit flavors (prune flavor). Moreover, this terpene was also assayed in several young
and aged wines. A traditional approach consisting of determining the organoleptic contribution
of a volatile compound thanks to the knowledge of its detection threshold in wine model solution
and in wine, as well as its concentration in wines was undertaken. We demonstrated that the
concentration of piperitone in red wines was not linked with the intensity of dried fruit flavors.
Finally, at the concentration assayed in the wines analyzed, the direct contribution of piperitone
to the flavor of young or old red wines was not established.
References
1. Chin, S.-T., & Marriott, P. J. (2015). Review of the role and methodology of high resolution
approaches in aroma analysis.Analytica Chimica Acta, 854, 1–12.
2. Ochiai, N., & Sasamoto, K. (2011). Selectable one-dimensional or two-dimensional gas
chromatography–olfactometry/mass spectrometry with preparative fraction collection for
analysis of ultra-trace amounts of odor compounds. Journal of Chromatography A, 1218(21),
3180–3185.
3. Pantò, S., Sciarrone, D., Maimone, M., Ragonese, C., Giofrè, S., Donato, P., ... Farnetti, S.
(2015). Performance evaluation of a versatile multidimensional chromatographic preparative
system based on three-dimensional gas chromatography and liquid chromatography–two-
dimensional gas chromatography for the collection of volatile constituents. Journal of
Chromatography A, 1417, 96–103.
4. Pons, A., Lavigne, V., Darriet, P., & Dubourdieu, D. (2013). Role of 3-methyl-2,4-
nonanedione in the flavor of aged red wines.Journal of Agriculture and Food Chemistry,
61(30), 7373–7380.
5. Sciarrone, D., Pantò, S., Ragonese, C., Dugo, P., & Mondello, L. (2015). Evolution and
status of preparative gas chromatography as a green sample-preparation technique.Trends in
Analytical Chemistry, 71, 65–73.