ELECTRONIC NOSES APPLICATIONS AND TECHNOLOGIES
Discrimination of different valerian types with an electronic nose
Baby, R. E1.; Cabezas, M2; Kutschker, A3.; Messina, V2.; Walsöe de Reca, N.E4*
CONICET-UTN (FRBA)1,
CITEFA2, UNPSJB3 and CITEFA-CONICET4
CINSO(Centro de INvestigaciones en SOlidos) CITEFA-CONICET
Juan Bautista de La Salle 4397, Villa Martelli (B1603ALO) Buenos Aires,
Fax: (54) (11) 4709 8241, E-mail: walsoe@citefa.gov.ar
*
Received November 05, 2004. In final form May 23, 2005
Abstract
Odors of two Valerian species (Valeriana carnosa Sm. and Valeriana clarionifolia
Phil.), grown in the Nordwest zone of Chubut Province (Argentina) were distinguished
with an electronic nose (EN) MOSES II (Modular Sensor System) and both valerian
varieties were chemically analyzed with TLC (Thin Layer Chromatography). The
two valerian species were gathered between March 2000 and November 2001. Specimen
of Valeriana clarionifolia Phil. were gathered from two different regions of
the same Chubut zone: wood-steppe and steppe. EN and TLC data were compared
with those of Valeriana officinalis coming from India. The odors of valerian
species were clearly discriminated with the EN enabling to separate not only
the different varieties but also the regions from where they were cropped. The
statistical Principal Component Analysis (PCA) method was used to evaluate the
EN data. TLC analysis was performed to identify the active compounds of valerians.
In a previous work [1] the authors determined that different species were separated
because they exhibited a different concentration of active compounds (like the
valerenic acid and valepotriates)whose concentration is closely related to sedative
properties of valerians.
Resumen
Los olores de dos especies de valerianas (Valeriana carnosa Sm. and Valeriana
clarionifolia Phil.) que crecen en la zona noroeste de la provincia de Chubut
(Argentina) fueron discriminados con una nariz electrónica (NE) MOSES
II (Modular Sensor System) y ambas especies fueron analizadas químicamente
con TLC (Thin Layer Chromatography) para identificar sus componentes activos.
Las dos variedades fueron recogidas entre marzo de 2000 y noviembre de 2001.
Las muestras de Valeriana clarionifolia Phil. fueron recogidas, a su vez, en
dos regiones diferentes de la misma zona de Chubut: bosque-estepa ecotono y
estepa. Los resultados hallados fueron comparados con los obtenidos con muestras
de Valeriana officinalis proveniente de la India. Los olores de los diferentes
tipos de valerianas fueron claramente discriminados por la NE permitiendo no
sólo separar las diferentes variedades sino también los diferentes
orígenes de los cuales provenían. El método estadístico
de Análisis de Componentes Principales (ACP) fue utilizado para evaluar
los datos de la NE. El análisis con TLC fue realizado para identificar
los componentes activos de las valerianas. En un trabajo previo [1] los autores
determinaron que la separación de las diferentes especies se producía
porque éstas presentaban una concentración distinta de los componentes
activos (entre ellos: ácido valerénico y valepotriatos) la cual
está estrechamente relacionada con la capacidad sedativa de las valerianas.
Introduction
In the last decades, a renewed interest on the active components of phitodrugs
has arisen generating the necessity of a major control of vegetal raw materials
to assure a constant and high quality of final product. Valerians are among
plants exhibiting numerous traditional uses in medicine. In Argentine Patagonia
grow nineteen native species, mainly in the Andean region though several varieties
grow even on the steppe zone.
Valerians L. [2] exhibits several species which are well known by their medicinal
value, the Valeriana officinalis L. among them. Active components of valerians
include valepotriates (valtrate, isovaltrate, dihydrovaltrate and acevaltrate)
and volatile oils of monotherpenes and sesquitherpenes (elemol, bornyl acetate,
bornyl-isovalerenate, isovalerate, valerenal, valerenona) and valerenic, hydrovalerenic
and acetoxivalerenic acids [1, 3]. The Valeriana carnosa Sm., Valeriana clarionifolia
Phil. and the Valeriana Laphatifolia Vahl. are considered among the natives
species of Patagonia, which were already used by mapuches communities in the
last three centuries [4].
With regards to valerians applications: they are widely used at the present
in phitotherapy, in allopathic and homeopathic medicine as sedative infusions
and as restaurative of nervous system sicknesses. It was demostrated an anxiolytic
effect in human experiments. It differs from barbiturics because it neither
interacts with alcohol nor produces hang-over when waking up in the morning.
It is also a weakly hypotensor and antispasmodic (intestinal cramps, digestive
troubles, muscles cramps related to stresses, etc.) and exhibits diuretic properties
[5-8].
In a previous work [1] performed with the MOSES II electronic nose (EN) was
demonstrated that it constitutes a valuable tool to separate different species
of medicinal plants through odors discrimination (valerians among them). In
the last twenty years, electronic noses appeared enabling to discriminate odors
of gas, solid or liquid specimen, performing a fast, continuous and non destructive
analysis.
An electronic nose is defined as a sensors array in which each sensor results
partially specific to detect a wide number of odorous molecules. The EN is
provided with a software system to perform the pattern recognition. Consequently,
this device constitutes a real “olfactory robot” which measurement
system could be compared with the human olfactory system. The MOSES II EN [9]
contains two modules of gas sensors, one of them composed of eight quartz microbalance
sensors (QMS). This type of sensors consist of vibrating quartz crystals covered
with polymeric selective coatings on which gases are adsorbed. The initial
vibration frequency (n0) of crystals
decreases according to the mass increase due to the gases adsorption [10]
and the difference between
n0 and the final frequency (nf
) results proportional to the adsorbed gas concentration. The other module
of sensors (SnO2) is composed by eight pure and doped semiconductive
SnO2 sensors. Doping with different elements increases SnO2
selectivity for different gases. The SnO2 surface conductivity changes
as the semiconductor adsorbs oxidizing or reductive gases [11]. The adopted
configuration results very flexible for general purposes and convenient for
a wide range of applications.The advantages of EN in comparison with the human
olfactory system include: higher
objectivity since it depends neither on people´s
condition nor on detection capacity, invariable response with time (as it
occurs with the sensorial or receptors saturation) assuring, consequently,
the success of routine analysis,
Experimental
Procedure
Specimen preparation: EN specimen of valerians from Patagonia were taken from
pieces of rhizomes or roots of valerians by milling them in a commercial coffee
grinder. After homogeinization, six samples of 1g were weighed and placed in
10 ml vials which were hermetically closed with Teflon septa and Aluminium
strap. Samples were incubated at 70°C for 60 min with an interval of 30
min between samples. The flow of the carrier gas transporting the specimen
from the headspace to the sensors camera was 20ml.min-1. In another experiment, four
samples like the above described were evaluated together with other four Valeriana
officinalis sample. EN data were analyzed with the PCA method.
In order to obtain the chromatographic profiles, each extract was prepared as
follows: 0.5g of grinded material was extracted with 5 ml of CH2Cl2 at
40°C for 5 min, filtered and the material remaining in filter was rinsed
with 2 ml of CH2Cl2. Both extracts were evaporated to
dryness under N2 flow. The residue was redissolved in CH3COOC2H5
to be used for the chromatographic analysis. Samples were seeded on Merck-G
60 F254 Silicagel layer. A mixture of (3:1) toluene:ethylacetate was used as
a mobile phase and the development was performed with a mixture of (2:8) Hydrochloric
acid:Acetic acid enabling to identify: valtrate, acevaltrate e isovaleroxi-hydroxi-dihydrovaltrate
(IVHD), all them in blue color while the other identified compound: dihydrovaltrate
turned to brown color.
Results
and Discussion
Figure 1 shows the PCA plot, as obtained with the MOSES II EN, of the two Argentine
valerian species (Patagonia): V. carnosa y V. clarionifolia. The last
specie presents two forms arisen from two different zones: wood-steppe isotonus
(clarionifolia 1) and steppe (clorionifolia 2), respectively. EN measurements
were taken with both types of sensors sets. A clear discrimination of the three
types of samples was obtained, the major part of information being concentrated
on the Principal Component 1 (PC1: 98%). Odor of V. carnosa is different from
odors of V. clarionifolia, the odors of both types of this specie are
similar as they grow close to each other.
Figures 2 and 3 are the PCA plots corresponding
to the same specimen of Figure 1, measurements taken with
the QMB sensors set (Figure 2) and with the SnO2
sensors (Figure 3), respectively. The two sets give similar
informations, the QMB sensors giving the best discrimination.
EN measurements with the three above considered specimens and the V. officinalis
(from India) are shown in Figure 4 (with both sensors sets),
Figure 5 (with QMB sensors) and Figure 6
(with SnO2 sensors). The use of both sensor sets enables to discriminate
the four types of specimen but, the data of the two types of V.clarionifolia
(coming from diferent regions of the same zone) appeared nearer than before
(Figure 1). Odors of V. carnosa and V. officinalis
seem to be similar in plot of Figure 4 (major information
on PC1: 84.5%). If only QMB sensors set is used for EN measurements (Figure
5) it appears more or less the same configuration than in Figure
4 but, odors of V. carnosa and V. officinalis seem better discriminated
(PC1: 88.2%). Figure 6 (EN measurments with SnO2
sensors) gives once again the optimal discrimination since odors of V. carnosa
and V. officinales are well separated between them and with regards
to V. clarionifolia. Odors of both types of V. clarionifolia,
are also discriminated but they smell similar. Besides the best discimination
is given by QMB sensors.
Figure 1: PCA plot of Patagonic valerian species. MOSES II EN measurements with both sensors (QMB + SnO2) sets.
Figure 2: PCA plot of Patagonic valerian species. MOSES II EN measurements with only one sensors (QMB) set.
The TLC analysis is shown in Figure 7, appearing from left to right the following information: V. officinalis L.(1), V. clarionifolia Phil. (2) and V. carnosa (3) together with dihydrovaltrate (4), iso-valeroxi-hydrovaltrate (5), valtrate (6) and iso-valtrate (7), the four last data corresponding to the pure active compounds of Valerians taken as references. The analysis of data shows that the V. officinalis (from India) contains the 4, 5, 6 and 7 compounds; the V. clarionifolia contains 5, 6 and 7 compounds in minor concentration while the compound 4 is absent. Similar data were found for the V. carnosa. In a previous work [1] the authors found that the valepotriates concentration in different origen V. officinalis was always ~1.2%. This concentration assures good therapeutic properties. Higher valepotriates concentration (>2.8%) were found in the V. walichii specie. It is known that higher concentrations of valepotriates (>3%) are able to produce cytotoxic and mutagenic effects [5].
Figure 3: PCA plot of Patagonic valerian species. MOSES II EN measurements with only one sensors (SnO2) set.
Figure 4: PCA plot of all the analyzed valerian species. MOSES II EN measurements with both sensors (QMB + SnO2) sets.
Figure 5:PCA plot of all the analyzed valerian species. MOSES II EN measurements with only one sensors (QMB) set.
Figure 6: PCA plot of all the analyzed valerian species. MOSES II EN measurements with only one sensors (SnO2) set.
Conclusions
The electronic nose MOSES II proved to be a fast and convenient tool to separate
different species of valerians or valerians from the same specie but grown in
different regions of the same zone, discriminating them by their odor. In spite
of successful separations for both cases (Patagonian species or Patagonian +
Indian species) employing both sensor sets together (QMB + SnO2)
or only one of them each time, it was observed a better separation when QMB
sensors were used,
The chromatographic profiles, as obtained with TLC technique enabled to prove
that Patagonian species contained three of the four active components: valtrate,
iso-valtrate and iso-valeoxihydrovaltrate in lower concentration than those
of V.officinalis (from India) which contains also another active component:
dihydrovaltrate,
Figure
7: Chromatographic (TLC) profiles corresponding to: 1. V. officinalis L.,
2. V. clarionifolia Phil., 3. V. carnosa Sm., 4. Dihydrovaltrate, 5. Iso-valeroxi-hydroxi-dihydrovaltrate,
6. Valtrate and 7. Iso-valtrate.
Acknowledgement
The authors are indebted to the ANPCyT by Grant (PICT 8688) given to N.E.W.R,
which enabled this work to be accomplished.
References
[1] Baby R. E.; Cabezas M.; Castro E.; Filip R.; Walsöe de Reca N. E., “Quality control of medicinal plants with an electronic nose”, Sensors & Actuators- B (Chemical), 2005, 106(1), 24-28.
[2] Bos, R.; Woerdenbag H. J.; Van Putten F. M. S.; Hendriks H.; Scheffer J.C., Planta Medica, 1998, 64, 2, 143-147.
[3] Wagner, H.; Bladt S., “Plant Drug Analysis”, Springer Verlag, 2nd. Edition, 1995, pp. 341-347.
[4] Correa, M. “Flora Patagónica”, Vol.VIII, Part VI, INTA Scientific Collection, Bs. As., Argentina, 1999.
[5] Leathwood P. D.; Chauffard F., Journal of Psychiatric Research, (1982-1983), 17, 2, 115-122.
[6] Leathwood P. D.; Chauffard F., Planta Medica, 1985, 2, 144-148.
[7] Leathwood P. D.; Chauffard F.; Heck F.; Muñoz-Box R., Pharmacology, Biochemistry & Behavior, 1982, 17, 1, 65-71.
[8] Lindahl L.; Lindwall M., Pharmacology, Biochemistry & Behavior, 1989, 32, 4, 1065-1066.
[9] Mitrovic I.; N. E. Walsöe de Reca, Industria y Química, 1998, 332, 33-37.
[10] King W. Piezoelectric Sorption Detectors, 1964, 36, 1735-1739.
[11] Kohl D. Sensors & Actuators- B (Chemical), 1989, 18, 71-113.