Anal Bioanal Chem (2011) 401:1475–1484
DOI 10.1007/s00216-011-4946-y
REVIEW
Technical solutions for analysis of grape juice,
must, and wine: the role of infrared
spectroscopy and chemometrics
D. Cozzolino & W. Cynkar & N. Shah & P. Smith
Received: 26 January 2011 / Revised: 15 March 2011 / Accepted: 22 March 2011 / Published online: 5 April 2011
# Springer-Verlag 2011
Abstract Information about constituents of grape juice,
must, and wine can be used for management and decision
support systems in order to improve, monitor, and adapt
grape and wine production to new challenges. Numerous
sensors that gather this information are either currently
available or in development. Nevertheless there is still a
need to adapt these sensors to special requirements, for
example robustness, calibration and maintenance, operating
costs, duration, sensitivity, and specificity to a particular
application. The sensors commonly used by the wine
industry are those that are based on mid-infrared (MIR),
near-infrared (NIR), visible (VIS) and ultraviolet (UV)
spectroscopy. This article reviews some recent technical
solutions for analysis of juice, must and wine based on the
combination of infrared spectroscopy and chemometrics.
Keywords Vibrational spectroscopy . Wine . Must .
Near-infrared . Mid-infrared . Authentication . Composition
Introduction
Public interest in food quality and methods of production
has increased significantly in recent decades, due in part to
changes in eating habits, consumer behaviour, and the
increased industrialization and globalization of food-supply
Published in the special issue Wine Analysis with Guest Editor Isabelle
Pianet.
D. Cozzolino (*) : W. Cynkar : N. Shah : P. Smith
The Australian Wine Research Institute,
Waite Road, Urrbrae, P.O. Box 197,
Glen Osmond, SA 5064, Australia
e-mail: Daniel.Cozzolino@awri.com.au
chains [1]. Demand for high levels of quality and safety in
food production obviously requires high standards in
quality assurance and process control; satisfying this
demand in turn requires appropriate analytical tools for
food analysis both during and after production. Desirable
features of such tools include speed, ease-of-use, minimal
or no sample preparation, and the avoidance of sample
destruction. These features are characteristic of a range of
spectroscopic methods including mid-infrared (MIR) and
near-infrared (NIR) spectroscopy [1, 2].
Food products are mainly composed of water, carbohy-
drates, proteins, fats, and other constituents that are present
at low (mg/100 g) concentrations, for example vitamins and
minerals. All of these components may contribute to the
shape of the absorbance spectrum obtained in the MIR and
NIR regions, although, in practice, the major components
(water, carbohydrates, proteins, fats) dominate because
constituents present at concentrations below approximately
0.1% w/w are difficult to detect in water-rich systems [1, 2].
Food heterogeneity results in considerable spectral com-
plexity and conventional approaches to the use of spectra,
for example, quantitative predictions of major compounds
may not be applied [1, 2].
A major advance in the application of spectroscopic
techniques to analysis of food and other commodities has
been in the application of powerful mathematical techni-
ques known collectively as chemometrics [1–4]. Such data
analysis methods enable the extraction of valuable infor-
mation from large and complex data sets and now underpin
the application of infrared (IR) spectroscopy in many
analytical fields [1–4].
Given the complex nature of grape juice, must, and
wine, there are many advantages of developing instrumen-
tal methods to describe, quantify and monitor their
composition. However, to be of practical use by the wine
1476
industry, instrumental methods must be objective, cost-
effective, and provide rapid, reproducible results, with
continuous operation. Existing analytical methods for
measurement of the composition and quality of wine and
other alcoholic beverages are not adequate for the demands
of production in a global market, because of their cost and
slow turnaround time. Factors such as promptness and low
cost of analysis, minimal sample preparation, and environ-
mentally friendly methods are of paramount importance in
the modern and sustainable wine industry [5–7].
In the last 40 years NIR spectroscopy became one of the
most attractive and used methods of analysis which enables
simultaneous, rapid, and non-destructive determination of
major components in many agriculture-related products and
plant materials. Recently, new applications involving the
determination of other minor compounds in plant materials
have been also reported. NIR analysis is characterized by
low molar absorptivities and scattering, which leaves nearly
effortless evaluation of pure materials [5–7].
Spectral “signatures” in the MIR region result from the
fundamental stretching, bending, and rotating vibrations of
the sample molecules, whereas NIR spectra result from
complex overtones and high-frequency combinations at
shorter wavelengths. Spectral peaks in the MIR frequencies
are often sharper and better resolved than in the NIR
domain, the higher overtones (i.e. 1st through 6th) of the
O–H (oxygen–hydrogen), N–H (nitrogen–hydrogen), C–H
(carbon–hydrogen), and S–H (sulfur–hydrogen) bands from
the MIR wavelengths are still observed in the NIR region,
although much weaker than the fundamental frequencies in
the MIR [5–7].
Information about constituents of juice, must and wine
can be used for management and decision support systems
in order to improve, monitor and adapt grape and wine-
production systems to new challenges. Objective quality
measures will enable vineyard managers to target required
quality levels and will allow rewards for quality, in terms of
quality-related grape payment systems where large areas of
new plantings coming on stream will apply a correction to
the fruit supply and demand situation, placing further
urgency on the requirement to determine quality levels.
The use of NIR spectroscopy in the wine industry dates
back to some early work by Kaffka and Norris [5]. Their
preliminary work was performed on a relatively small
number of test samples prepared by standard addition of
some of the main components of interest (viz. ethanol,
fructose, and tartaric acid) to a red and a white wine. It is
well recognized that ethanol in alcoholic beverages has a
strong NIR absorbance signal, usually second only to that
of water. However, both accuracy and robustness of
calibrations can be limited by matrix variations, particularly
variations in sugar concentration. Much of the NIR work in
the wine industry has concentrated on ethanol analysis.
D. Cozzolino et al.
There are now a number of dedicated NIR-based alcohol
analyzers, and NIR spectroscopy has become one of the
many routine methods for analysis for alcohol [5].
This article reviews some of the recent technical
solutions for analysis of grape juice, must, and wine based
on the combination of IR spectroscopy and chemometrics.
Juice compositional analysis
Manley and collaborators attempted NIR calibrations for
free amino nitrogen (FAN) in grape must and assessment of
malolactic fermentation status of wines [8]. Although these
calibrations could not accurately quantify the concentra-
tions of the compounds of interest (malic acid, lactic acid,
FAN), a soft independent modelling by class analogy
(SIMCA) routine could distinguish between groups of
high, medium, and low concentration [8].
The use of an attenuated total-reflectance (ATR) cell
attached to a MIR spectrophotometer to measure juice
compositional data was reported by several authors [9, 10].
Analytical values of juice ammonia, yeast assimilable
nitrogen (YAN), and FAN ranged from 0–150 mg L−1,
0–300 mg L−1, and 3–384 mg L−1, respectively [9]. The
coefficient of determination (R2) and the standard error in
cross validation (SECV) reported were 0.79 and
28.3 mg L−1 for FAN, 0.92 and 14.8 mg L−1 for ammonia,
and 0.80 and 36.9 mg L−1 for YAN, respectively. The
standard error of prediction (SEP) was 36.7 mg L−1 for
FAN, 17.2 mg L−1 for ammonia, and 42.4 mg L−1 for YAN,
respectively [9]. The results from this study showed that
ATR-MIR spectroscopy is capable of both measuring and
monitoring grape juice properties in an industrial situa-
tion and will have considerable advantages compared
with using traditional transmission cells in MIR, because
the sample presentation is simple, requiring only 100 μL
applied on to the ATR crystal [9]. Furthermore, unlike
transmission, the ATR does not require a short-path length
cell, with inherent problems such as interference fringes
and cell cleaning [9, 10].
Preys and co-workers found that application of chemo-
metric methods, for example orthogonal projection and
experimental design, to the NIR spectra of must can correct
for scattering in a turbid media, such as grape must,
enabling the use of this technique for fermentation
monitoring in a commercial winery [11].
The use of FTMIR for quantification of total soluble
solids, pH, and titratable acidity in grape must has also
been reported [12] (Table 1). The direct comparison of
NIR and MIR was reported as a rapid method to monitor
glucose, fructose, ethanol, glycerol, total phenolics, total
anthocyanins, and total flavonoids in red wine during
fermentation [13].
Technical solutions for analysis of grape juice, must, and wine 1477

Table 1 Novel and recent appli-

cations of IR in grape juice, Sample Application Method IR or NIR range SEP Ref.
must, and wine analysis
Wine Ethyl octanoate (mg L−1) FTNIR 1200–10000 cm−1 0.07 [24]
Ethyl 2-phenylacetate (mg L−1) 0.002
Diethyl succinate (mg L−1) 0.80
Diethyl glutarate (mg L−1) 0.003
2-Phenylethanol mg L−1 4.08
Octanoic acid (mg L−1) 0.28
Hexanoic acid (mg L−1) 0.15
FTNIR, Fourier-transform near Wine Glucose and fructose (g L−1) ATR-MIR [20]
infrared; FTMIR, Fourier- Alcohol (%)
transform mid infrared; Must TSS (°Brix) FTMIR 930–5011 cm−1 0.38 [12]
ATR-MIR, attenuated total- pH 0.05
reflectance mid infrared;
VIS–NIR, visible near infrared; TA (g L−1) 0.49
SEP, standard error of predic- Grape juice YAN ATR-MIR [9]
tion; TSS, total soluble solids; Grape juice TSS (°Brix) VIS–NIR 400–1100 nm 0.35 [51]
TA, titratable acidity; YAN, pH 0.048
yeast assimilable nitrogen

Wine compositional analysis
IR spectroscopy has been used to measure several wine
compositional properties, for example alcohol content, pH,
volatile acidity, organic acids, malic, tartaric, and lactic
acids, reducing sugars, and sulfur dioxide in red, rose, and
white wines [5, 14–17]. Although NIR spectroscopy has
already found widespread applications in quality control
and process analysis in the grape and wine industry,
because NIR absorptions reflects overtone and combination
bands, NIR spectra are much less distinct than MIR [5, 17].
Hence, the increasing use of MIR in wine analysis was of
special interest because of the presence of sharp and
specific absorption bands of some wine constituents [5,
17]. Dedicated FTMIR instruments (e.g. WineScan,
GrapeScan, Microdom) are now available and are used
extensively in routine analysis of grape juice and wine
samples by the wine industry worldwide [5, 17]. Wine
compositional properties, for example alcohol content,
volatile acidity, pH, tartaric acid, lactic acid, glucose plus
fructose, polysaccharides, volatile compounds, polyphenols,
and glycerol, have been proposed and implemented for
routine analysis using FTMIR instruments by the industry
[14–19]. Figures 1 and 2 show and compare PLS calibration
statistics based on spectra recorded using NIR and MIR
spectroscopy to predict the ethanol content of a set of red
and white wine samples.
In the last two years new and non-traditional applications
of IR have been reported by several authors. The
development of sampling accessories attached to a wide
range of IR spectrophotometers, for example ATR cells, has
led to major improvements in routine IR analysis, by
simplifying sample handling and avoiding measurement
problems often found using transmission cells [9, 20]. In
commercial FTMIR spectrophotometers commonly used in
wine routine analysis, samples are analyzed through a
short-path-length transmission cell. The advantage of the
transmission cell is that it enables very accurate and
reproducible spectroscopic measurements and is easily
temperature controlled [5, 9]. On the other hand, the use
of transmission cells has several drawbacks such as filling
and cleaning the cell, variation of sample path length
because of window wear, and turbidity of the sample,
among others [9, 20]. Therefore the alternative use of ATR
cells combined with MIR makes this an attractive means of
sample presentation for wine compositional analysis.
Volatile chemical compounds responsible for the aroma
of wine are derived from a number of different biochemical
and chemical pathways [3]. These chemical compounds are
formed during grape berry metabolism, crushing of the
berries, fermentation processes (i.e. yeast and malolactic
bacteria), and also from the ageing and storage of wine [3,
21]. Not surprisingly, there are a large number of chemical
classes of compounds found in wine which are present at
varying concentrations (ng L−1 to mg L−1), have differing
potencies, and have a broad range of volatilities [21].
The volatile composition profiles of South African
young cultivar wines were reported by Louw and co-
workers [22], who combined chemometrics with Fourier
transform mid-infrared (FTMIR) to identify compositional
trends and to distinguish differences arising from by
cultivars or wine origin. Reference data were generated
using gas chromatography (GC) and analysis of variance
(ANOVA); principal-components analysis (PCA) and linear
discriminant analysis (LDA) were used for discrimination
[22]. Significant differences were found in the volatile
composition of the cultivar wines, with marked similarities
in the composition of Pinotage wines and white wines,
specifically for 2-phenylethanol, butyric acid, ethyl acetate,
isoamyl acetate, isoamyl alcohol, and isobutyric acid [22].
1478 D. Cozzolino et al.

Fig. 1 Partial least-squares cross validation for measurement of ethanol using near-infrared spectroscopy

The ability of NIR spectroscopy to determine oak types were analysed and calibration models were developed
volatile compounds and levels of ethylphenols in aged red by applying partial least-squares (PLS) regression to GC
wines was also reported [23]. For this purpose 510 wines reference data and NIR spectra. Results from this study
aged with different storage times and in different oak barrel showed that the calibration statistics were very good for all

Fig. 2 Partial least-squares cross validation for measurement of ethanol using mid-infrared spectroscopy
Technical solutions for analysis of grape juice, must, and wine
the compounds analysed (R2 >0.86) both in wines aged in
American and French oak barrels, and in “reserva” and
“gran reserva” wines [23] (Table 1).
Several wines (n=240) belonging to different geographic
zones and elaborated with one or two varieties were
analysed and quantified by means of GC. Spectra obtained
by NIR were correlated with these values by use of PLS
regression [24]. The authors concluded that NIR spectros-
copy can be used as an easy and rapid tool to determine
fermentative volatile compounds in aged red wines [24]
(Table 1).
Volatile chemical compounds present in commercial
Riesling wines were measured by GC coupled with selected
ion monitoring mass spectrometry and by NIR spectrosco-
py [21]. PLS regression models with full cross validation
were developed yielding R2 and SECV of 0.74 (SECV:
313.6 μg L−1) for esters, 0.90 (SECV: 20.9 μg L−1) for
monoterpenes, and 0.80 (SECV: 1658 μg L−1) for short-
chain fatty acids [21].
Fourier transform infrared (FTMIR) spectroscopy was
used for prediction of red wine total antioxidant capacity
(TAC). Full cross-validated PLS predicted TAC values
revealed good correlation (r=0.85) and a slope of 0.74;
the prediction error provided by the PLS model was
consistent with the uncertainty derived from the reference
method [25].
The ability of an electronic tongue based on FTMIR
spectroscopy as a gustative sensor was assessed by
emulating the responses of a tasting panel for the gustative
mouthfeel “tannin amount” [26]. The FTMIR spectra were
modelled against the sensory responses evaluated in 37 red
wines using PLS regression models. In order to find the
wavenumbers most correlated with the sensorial attribute
and thus providing the best predictive models, six
different variable selection techniques were also tested.
The iterative predictor weighting (IPW-PLS) technique
gave the best results (root mean square error of cross
validation 0.13) [26].
In another experiment a set of 39 single cultivar Pinotage
wines comprising 13 samples with medium to high
bitterness were analyzed using FTMIR (e.g. WineScan)
and high-performance liquid chromatography (HPLC) [27].
A trained sensory panel assessed the bitterness intensity of
15 wines, 13 of which had a bitter taste of medium to high
intensity. Calibration models for classification of wine
samples according to the perception of the bitter taste and
quantification of the bitterness intensity were calculated by
partial least-squares discriminant analysis (PLS-DA)
regression. Both spectral and chemical analysis data
could discriminate between bitter and control wines with
the correct classification rates of 94% and 91%, respec-
tively [27]. Prediction of the bitterness intensity was also
possible using FTMIR data [27].
1479
Several enological properties (colour, free and total SO2,
total aldehyde, and total phenol) were measured and
correlated with oxygen transmission rate (OTR) and ATR-
MIR spectral data using PLS regression [28]. The ATR-
MIR spectra of the wine samples analyzed were able to
predict colour (absorbance at 420 nm), free SO2, total SO2,
total phenol, total aldehyde, and storage time [28].
Monitoring wine fermentation
Optimized process control is essential to address safety
rules and maintain a commercially viable and sustainable
wine industry. This implies, among other aspects, rapid
assessment of the chemical and physical properties of raw
materials (e.g. grape and juice), process streams (e.g.
ferments), and end products (e.g. wine) [5, 29]. Currently,
many sensors are used in the agro-food industries to assess,
monitor, and control composition. These sensors can be
divided into physical sensors to monitor temperature,
pressure, flow rate, and humidity or chemical sensors that
are commonly used to assess conductivity, pH, and sugars
[5, 29, 30].
Wine fermentation is a complex process in which grape
juice is transformed by microbial action into a high-value
product, wine [5, 30, 31]. Control of the wine fermentation
process is a very important step during wine production in
order to accurately and rapidly monitor and control both
substrate conversion (e.g. sugars to ethanol, malic acid to
lactic acid) and product quality [5, 30, 31].
Several studies report the use of both NIR and MIR
spectroscopy to monitor wine fermentation such as the
use of telecommunications-grade fibre optics to measure
ethanol in wines of different grape varieties, the
measurement of sugar and ethanol using different
fermenter sampling systems and temperature equilibra-
tion, comparison of a filter-based NIR instrument with an
FTNIR scanning instrument to measure sugars and
ethanol, and the use of ATR-MIR to monitor wine
fermentation [5, 30, 31].
In recent years, research has been conducted in order to
assess the potential of VIS–NIR spectroscopy to predict
the concentration and monitor the extraction and
evolution of phenolic compounds during red wine
fermentation in real time [5, 30, 31]. Results showed
that VIS–NIR spectroscopy could predict the concentra-
tion of major anthocyanins, for example malvidin-3-
glucoside (R2 =0.91 and SECV= 28.0 mg L−1), pigmented
polymers (R2 =0.87 and SECV=5.9 mg L−1), and tannins
(R 2 = 0.83 and SECV = 131.1 mg L −1 ), in Cabernet
Sauvignon and Shiraz wines during fermentation [5].
The use of FTMIR and fibre optics to monitor large-scale
wine fermentations has also been reported [32–34].
1480
The prediction of density and sugars during white wine
fermentation was evaluated using a miniature fibre optic
NIR (800–1050 nm) instrument attached to a 50 mm path
length cell [35]. Both PLS and multiple linear regression
(MLR) were used to measure reducing sugars (glucose and
fructose) with high accuracy [35].
Micro-fermentation trials conducted during the 2008
vintage harvest in the Valtellina (Northern Italy) viticultural
area were monitored using both NIR and MIR [13].
Samples were also analysed by using chemical methods to
evaluate sugars (glucose and fructose), alcohols (ethanol
and glycerol), and phenolic compounds (total phenolics,
total anthocyanins and total flavonoids). The pretreated
spectral data were processed by applying PCA, PLS
regression, and LDA to spectral data to classify and predict
sugar content, ethanol, glycerol, and phenolic compounds
simultaneously during the fermentation stage from initial to
final phase [13]. LDA results, characterized by a high
percentage of correct classification (87% and 100% as
average value in prediction for NIR and MIR spectroscopy,
respectively), showed that samples belonging to a particular
fermentation step could be correctly classified. Good
calibration models for prediction of the main compositional
changes during alcoholic fermentation were obtained with
both NIR and MIR, suggesting that either instrument could
be used to evaluate online and simultaneously these
compounds in red wine [13].
Compositional changes occurring during a commercial
red wine fermentation were effectively monitored using
FTMIR spectroscopy and modelled with the aid of two-
dimensional correlation techniques (2D). The application of
2D spectroscopy showed that the reaction rates for the
conversion of fructose and glucose to alcohol were
different, with the latter being more rapid [36, 37].
Wine authenticity
Food adulteration has been practiced since biblical times
but has become more sophisticated in the recent past. Foods
or ingredients most likely to be targets for adulteration
include those which are of high-value, or are subject to the
unpredictable changes of weather during their growth or
harvesting. The practice of adulteration arises for two main
reasons: first it can be profitable, and, second, adulterants
can be easily mixed and are subsequently difficult to detect
[5, 38, 39].
Verification of authenticity of wine products has become
a potential application of IR spectroscopy methods.
Adulteration can take many forms, including the addition
of sugars, acids, volatile compounds, over-dilution of
concentrate, addition of juices/wines of other fruits, use of
concentrate in a “fresh” product, and use of low quality
D. Cozzolino et al.
product recovered from what are normally waste products
[5, 38, 39].
Detection of adulteration by traditional analytical meth-
ods can be very complex and is often inconclusive. An
authentic food is one which is what it purports to be, in
other words one which conforms to the description
provided by the producer or processor [5, 38, 39].
Important aspects of this description may relate to the
process history of a product (e.g. fresh meat as opposed to
frozen meat) or its geographic origin (e.g. Italian olive oil
produced only from olives grown in Italy) [5, 38, 39].
Although there are numerous reports of the use of IR
spectroscopy to authenticate foods, until recently not many
published works were found in the literature on the use of
IR spectroscopy to authenticate grape or wine samples. The
use of VIS–NIR was investigated to discriminate Riesling
and unwooded Chardonnay wines, with accuracy of up to
95% [40]. Both VIS and NIR spectroscopy combined with
multivariate analysis were used to classify the geographical
origin of commercial Tempranillo wines from Australia and
Spain [41]. Wines (n=63) were scanned in the VIS and
NIR regions (400–2500 nm) in a monochromator instru-
ment in transmission. PCA, PLS-DA, and LDA based on
PCA scores were used to classify Tempranillo wines
according to their geographical origin [41]. PLS-DA
models correctly classified 100% and 84.7% of the
Australian and Spanish Tempranillo wine samples, respec-
tively. Additionally, LDA calibration models correctly
classified 72% of the Australian wines and 85% of the
Spanish wines. These results demonstrated the potential use
of VIS and NIR spectroscopy, combined with chemo-
metrics as a rapid method for classifying Tempranillo wines
according to their geographical origin [41]. Figure 3 shows
an example of the PCA scores (PC1 vs PC2) derived from
the MIR spectra of white and red wine samples. A
separation between red and white wine samples can be
observed.
The combination of FTMIR with 2D IR correlation
spectroscopy was used to analyze the main components of
red wine samples with different sugar contents and
volatilization residues of dry red wine from different
manufactures [42]. The second derivative IR spectra of
red wine show the typical peaks of alcohol, whereas the
spectra of sweet wine are composed of the peaks of both
alcohol and sugar, and the contribution of sugar is enhanced
with increasing sugar content [42]. Using PCA as pre-
processing method, dry and sweet wine samples can be
classified correctly. Analysis of the MIR spectra of the dry
red wine samples from five different manufacturers indi-
cates that glycerol, carboxylic acids, or esters can be
identified to discriminate the different manufacturers and
evaluate the quality of wine samples [42]. The results
suggested that IR spectroscopy combined with 2D IR
Technical solutions for analysis of grape juice, must, and wine
Fig. 3 Score plot of the first two principal components used to
discriminate wine samples analysed by mid-infrared spectroscopy (red
label = red wines; black label = white wines)
correlation spectroscopy is an effective method for analysis
and classification of red wine samples [42] (Table 2).
Several instrumental techniques, including NIR and UV–
visible spectroscopy, headspace-mass-based artificial nose
and voltammetric artificial tongue, combined with chemo-
metric pattern recognition techniques, were applied and
compared in addressing authentication of wine samples
from the same Italian oenological region, according to
grape variety [43]. In the study, 59 certified samples
belonging to the Barbera d’Alba and Dolcetto d’Alba
appellations and collected from the same vintage (2007)
were analyzed [43]. The instrumental responses, after
proper data pre-processing, were used as fingerprints of
the characteristics of the samples [43]. The results from this
study showed that both NIR and UV–visible have the best
classification and predictive ability (range 91.7% to 100%)
compared with the artificial nose and tongue [43].
Table 2 Application of infrared (IR) and chemometric methods for wine authenticity and traceability
Application
Italian red wines
Greek red wines
Australian and New Zealand Sauvignon Blanc wines
Organic and non-organic wines
Dry and sweet wines
MIR, mid-infrared; NIR, near-infrared; ATR-MIR, attenuated total-reflectance mid-infrared; PCA, principal-components analysis; LDA, linear
discriminant analysis; PLS-DA, discriminant partial least-squares
1481
The discrimination of Greek wines of three different
varieties (Agiorgitiko, Xinomavro, and Merlot) using ATR-
MIR was reported [44]. The authors reported that the MIR
region between 1800 and 900 cm−1 was used to fingerprint
wines on the basis of grape variety [43]. The authentication
of Australian organic and non organic wines using MIR
spectra [45] and Sauvignon Blanc wines from Australia
and New Zealand using NIR and MIR was also reported
[46].
The use of 2D correlation spectroscopy and MIR was
used to discriminate between dry and sweet red wine
samples [47]. The results showed that 2D correlation is
useful for quantitative evaluation of the similarity among
the wine samples and their corresponding 2D correlation
spectra [47].
Single-bounce micro ATR-MIR combined with multi-
variate and artificial neural networks (ANN) data analysis
has been used to determine the adulteration of selected red
wines with industrial grade glycerol [48]. The results from
this study on the classification of glycerol adulterants in the
different brands of red wine samples analyzed resulted in
classification accuracy in the range between 94% and 98%
[48]. The authors also reported that ATR-MIR could
determine the presence of industrial glycerol in red wine
at a detection limit of 1% [48].
In-bottle measurement
An ideal method for determination of the chemical
composition of wine in a routine manufacturing schedule
should be non-invasive, non-destructive, and rapid to
ensure timely processing of the food being analyzed [5,
49]. Classical univariate spectroscopy methods in the ultra
violet (UV) and VIS regions of the electromagnetic
spectrum require physical separation of the constituent of
interest from the matrix, usually by dissolution in a solvent.
NIR spectroscopy combined with chemometrics offers the
advantages of simplicity of sample presentation, the
Method Chemometric method Ref.
NIR, VIS, and UV transmission PCA, LDA [43]
ATR-MIR Spectrum matching [44]
NIR transmission PCA, PLS-DA, SIMCA [46]
MIR transmission PCA, LDA [45]
MIR transmission PCA, 2D IR [42]
1482
speed of collection of the information (spectra) and low
cost [5, 49].
In NIR short wavelengths (700–1600 nm) the spectrum
has less intense peaks and can tolerate longer path lengths
(5–30 mm) but yields poor spectral resolution, whereas
peaks in longer wavelength regions (>1500 nm) are more
intense and require short optical path length but provide
better spectral resolution [50].
Glass is transparent to NIR radiation—a property that
has enabled the pharmaceutical industry to use NIR to
analyse liquids through the bottle in order to determine
moisture in lyophilized sucrose through intact glass vials
and for real-time monitoring in solid-phase synthesis of a
resin-bound alcohol. It has also been used to analyze Sufu,
beer, and whisky, non-destructively, in their point of sale
containers [5, 49]. Combining both the VIS and NIR
regions in one instrument provides a vast improvement in
instrumental, sampling, and analytical efficiency [5, 49].
New commercially available spectrophotometers (e.g.
diode-array instruments) with simple sample presentation
extend this potential for at/on-line applications. Although
these results are promising, the authors believe that this
type of instrument or sample presentation can be used only
as an indicative rather than a quantitative analytical tool to
monitor a process, because of limitations in analytical
accuracy. It is expected that future development of such
applications will provide the wine industry with a very fast
and non-destructive method to monitor composition or
changes and to detect unwanted problems (e.g. oxidation)
in bottled wine before retail sale, and provide a rapid means
of qualitative rather than quantitative analysis [5, 49].
Critical aspects and limitations of IR methods
As a fast and easy-to-operate technique, IR spectroscopy
has already gained wide industrial acceptance for routine
wine analysis. Considering the continuing improvements in
hardware and software design, and the analytical require-
ments of real-time or multi-variable analysis by modern
industry, it is expected that in the near future IR
spectroscopy will progressively become a routine method
for process monitoring and process control. Several critical
aspects and limitations, for example instrument availability,
type of application, understanding of the technology might,
however, be acting as barriers for adoption of IR technol-
ogies by the wine industry.
There are currently over 65 companies that manufacture
IR spectrophotometer instruments in a variety of formats
including bench top, on-line, and portable instruments.
Selection of a suitable IR spectrophotometer instrument for
purchase is important. For example, in considering
purchase of a spectrophotometer, it is necessary to define
D. Cozzolino et al.
some fundamental specifications, for example wavelength
(wavenumber) scanning range, wavelength data point
interval (number of absorption data points), noise,
stability, and measurement time and the type of sample
to be analysed or the application.
However, one of the main factors that determine the type
of instrument to be selected is the type of the measurement
to be conducted. This will be highly dependant on the
nature of the sample matrix (e.g. grape juice, must, ferment,
or wine), and the concentration and nature of the chemical
constituents that we want to measure.
Instruments for industrial applications should provide
a simple operational method to collect, store, and retrieve
data, and instrument-control software that performs
diagnostic checks.
One of the most important critical aspects of the
development of IR spectroscopy is the need for appropriate
training, for example routine analyses can be performed by
analysts with a high school education, although knowledge
of the chemistry of the sample material will be useful.
However, calibration development is, by far, the critical
aspect that requires a high level of expertise, particularly in
chemometrics or multivariate data analysis in order to make
an application successful.
Together these technologies will change the efficiency of
laboratories and the range of data they can produce. There
is, however, a need to change the way of transfer and
communication of the information generated by this
methods. These new technologies will require a new
approach to analysis and a change of attitude by viticultu-
rists, winemakers, and others in the industry. This shift
needs to move away from the classical idea of wine
chemical data being discreet numbers reflecting what has
happened, to a model in which it can be seen as a dynamic
picture in which change itself is vital information which
triggers a process (E. Wilkes, personal communication).
Conclusions
During the past 30 years an increasing number of
applications have been investigated and used in viticultural
and oenological studies. However, the use of IR spectros-
copy in the alcoholic beverage industry is still in its infancy.
From analysis of the scattered information available it
seems that IR spectroscopy has been applied in different
steps during the production of wine.
The potential of vibrational spectroscopy (MIR and NIR)
has been investigated for many applications in the grape
and wine industry. Perhaps the most comprehensively
examined area is that of wine composition, where MIR
and NIR methods have been used for analysis at all
stages throughout the winemaking process, from charac-
Technical solutions for analysis of grape juice, must, and wine
terization of grapes and juice to finished products and
has even been extended to process-control applications.
Monitoring of the progress of fermentations has also
received some attention through the use of NIR and
ATR-MIR spectroscopy to predict ethanol and sugar
concentration during winemaking.
The greatest effect of vibrational spectroscopy in the
wine industry so far has been its use for measurement of the
alcohol content of wines, with commercially available
“turn-key” NIR instruments in widespread use in the
industry worldwide. The potential of calibrations for other
properties of wine, for example organic acid and sugar
content, and other compositional data has also been
investigated with varying degrees of success. More recent-
ly, attention has focused on the potential of NIR and MIR
methods to predict other properties of wine, for example
volatiles and phenolics, and for authenticity issues.
The advantages of ability to predict multiple properties and
the speed of analysis mean that vibrational spectroscopy has a
bright future in the grape and wine industry. The future
development of such applications will provide the wine
industry with a very fast and non-destructive method to
monitor composition or changes and to detect unwanted
problems in bottled wine (e.g. oxidation) before retail sale,
and provide a rapid means of qualitative rather than quantitative
analysis. The potential savings, reduction in time and cost of
analysis, the environmentally friendly nature of the technology,
have positioned rapid instrumental techniques as extremely
attractive in the field of analysis of alcoholic beverages.
Undoubtedly IR spectroscopy has become a relevant
technique for the grape and wine industry that will provide
an inexpensive and rapid method of analysis. Further, the
technology offers the exciting prospect of potentially
enabling the development of small-scale, inexpensive,
portable hand-held instruments, which would be of a great
benefit to the whole supply chain of the industry. As the
technology of IR spectroscopic instrumentation and chemo-
metrics advances further, the resulting spin-offs may further
assist the industry in its quest to define and objectively
measure grape and wine quality, and to assure consumers of
the quality of the final product to be enjoyed.
Acknowledgements This project is supported by Australia’s grape-
growers and winemakers through their investment body the Grape and
Wine Research and Development Corporation, with matching funds
from the Australian government. The work was conducted by The
Australian Wine Research Institute.
References
1. Karoui R, Downey G, Blecker Ch (2010) Chem Rev 110:6144–
6168
1483
2. Woodcock T, Downey G, O’Donnell CP (2008) J Near Infrared
Spectrosc 16:1
3. Smyth HE (2005) The Compositional basis of the aroma of
Riesling and unwooded Chardonnay wine, PhD Thesis, Univer-
sity of Adelaide, Australia
4. Martens M (1999) Food Qual Pref 10:233–244
5. Cozzolino D, Cynkar W, Janik L, Dambergs RG, Gishen M
(2006) J Near Infrared Spectrosc 14:279–289
6. McClure WF (2003) J Near Infrared Spectrosc 11:487–518
7. Nicolai BM, Beullens K, Bobelyn E, Peirs A, Saeys W, Theron
KI, Lammertyn J (2007) Post Harvest Biol Tech 46:99–118
8. Manley M, van Zyl A, Wolf EEH (2001) S Afr J Oenol Vitic
22:93–100
9. Shah N, Cynkar W, Smith P, Cozzolino D (2010) J Agric Food
Chem 58:3279–3283
10. Cocciardi RA, Ismail AA, Sedman J (2005) J Agric Food Chem
53:2803–2809
11. Preys S, Roger JM, Boulet JC (2008) Chemometric Intell Lab Sys
91:28–33
12. Swanepoel M, du Toit M, Niewoudt HH (2007) S Afr J Oenol
Vitic 28:140–148
13. Di Egidio V, Sinelli N, Giovanelli G, Moles A, Casiraghi E (2010)
Eur Food Res Technol
14. Urbano-Cuadrado M, Luque de Castro MD, Perez-Juan PM,
Garcia-Olmo J, Gomez-Nieto MA (2004) Anal Chim Acta
527:81–88
15. Urbano-Cuadrado M, Luque de Castro MD, Pérez-Juan PM,
Gómez-Nieto MA (2005) Talanta 66:218–224
16. Guggenbichter W, Huck ChW, Kobler A, Popp M, Bonn GK
(2006) J Food Agric Environ 4:98–106
17. Bauer R, Nieuwoudt HH, Bauer FF, Kossmann J, Koch KR,
Esbensen KH (2008) Anal Chem 50:1371–1379
18. Boulet JC, Williams P, Doco T (2007) Carbohydr Polym 69:79–
85
19. Boulet JC, Doco T, Roger JM (2007) Chemometr Intell Lab Sys
87:295–302
20. Cozzolino D, Shah N, Cynkar W, Smith P (2011) Food Res Int
44:181–186
21. Smyth HE, Cozzolino D, Cynkar W, Dambergs RG, Sefton M,
Gishen M (2008) Anal Bioanal Chem 390:1911–1916
22. Louw L, Roux K, Tredoux A, Tomic O, Naes T, Nieuwoudt HH,
van Rensburg P (2009) J Agric Food Chem 57:2623–2632
23. Garde-Cerdan T, Lorenzo C, Alonso GL, Salinas MR (2010) Food
Chem 119:823–828
24. Lorenzo C, Garde-Cerdan T, Pedroza MA, Alonso GL, Salinas
MR (2009) Food Res Int 42:1281–1286
25. Versari A, Parpinello GP, Scazzina F, Del Rio D (2010) Food
Control 21:786–789
26. Vera L, Acena L, Boque R, Guasch J, Mestres M, Busto O (2010)
Anal Bioanal Chem 397
27. Rudnitskaya A, Nieuwoudt HH, Muller N, Legin A, du Toit M,
Bauer FF (2010) Anal Bioanal Chem 397:3051–3060
28. Fu Y, Lim LT, McNicholas PD (2009) J Food Sci 74:C608–C618
29. Sablayrolles JM (2009) Food Res Int 42:418–424
30. Cozzolino D, Parker M, Dambergs RG, Herderich M, Gishen M
(2006) Biotech Bioeng 95:1101–1107
31. Gishen M, Dambergs RG, Cozzolino D (2005) Aust J Grape Wine
Res 11:296–305
32. Zeaiter M, Roger JM, Bellon-Maurel V (2006) Chemometr Intell
Lab Sys 80:227–235
33. Urtubia A, Pérez-Correa JR, Meurens M, Agosin E (2004) Talanta
64:778–784
34. Uturbia A, Pérez-Correa JR, Soto A, Pszczolkowski P (2007)
Food Control 18:1512–1517
35. Fernandez-Novales J, Lope MI, Sanchez MT, Morales J,
Gonzalez-Caballero V (2009) Food Res Int 42:285–291
1484
36. Adams MJ, Wynne L, Clark S, Barnett NW (2007) Vib Spectrosc
394
37. Wynne L, Clark S, Adams MJ, Barnett NW (2007) Vib Spectrosc
44:394–400
38. Arvantoyannis I, Katsota MN, Psarra P, Soufleros E, Kallinthraka
S (1999) Trends Food Sci Technol 10:321–336
39. Cordella Ch, Moussa I, Martel A-C, Sbirrazzuoli N, Lizzani-
Cuvelier L (2002) J Agric Food Chem 50:1751–1764
40. Cozzolino D, Smyth HE, Gishen M (2003) J Agric Food Chem
51:7703–7708
41. Liu L, Cozzolino D, Cynkar WU, Gishen M, Colby CB (2006) J
Agric Food Chem 54:6754–6759
42. Zhang Y-L, Chen J-B, Lei Y, Zhou Q, Sun S-Q, Noda I (2010) J
Molecular Structure 974:144–150
43. Casale M, Oliveri P, Armanino C, Lanteri S, Forina M (2010)
Anal Chim Acta 668:143–148
D. Cozzolino et al.
44. Tarantilis PA, Troianou VE, Pappas CS, Kotseridis YS, Polissiou
MG (2008) Food Chem 111:192–196
45. Cozzolino D, Holdstock M, Cynkar WU, Dambergs RG, Smith P
(2009) Food Chem 116:761–765
46. Cozzolino D, Shah N, Cynkar W, Smith P (2011) Food Chem
126:673–678
47. Chen JB, Zhou Q, Noda I, Sun SQ (2009) Appl Spectrosc 63:920–925
48. Dixit V, Tewara JC, Cho B-K, Irudayaraj JMK (2005) Appl
Spectrosc 59:1553–1560
49. Cozzolino D, Kwiatkowski MJ, Waters EJ, Gishen M (2007) Anal
Bioanal Chem 387:2289–2295
50. Murray I, Cowe I (2004) In: Roberts CA, Workman J, Reeves JB
(eds) American Society of Agronomy, Crop Science Society of
America. Soil Science Society of America, Madison
51. Wu D, He Y, Nie P, Cao F, Bao Y (2010) Anal Chim Acta
659:229–237