Food Control 78 (2017) 144e149

Contents lists available at ScienceDirect

Food Control
journal homepage: www.elsevier.com/locate/foodcont

Gum Arabic authentication and mixture quantification by near

infrared spectroscopy
Yongjiang Dong a, 1, Klavs Martin Sørensen b, *, 1, Sailing He a, Søren Balling Engelsen b
a
Centre for Optical and Electromagnetic Research, Zhejiang Provincial Key Laboratory for Sensing Technologies, State Key Laboratory Modern Optical
Instrumentation, JORCEP, Zhejiang University, Hangzhou 310058, PR China
b
Spectroscopy & Chemometrics, Department of Food Science, University of Copenhagen, Rolighedsvej 26, 1958 Frederiksberg, Denmark

a r t i c l e i n f o a b s t r a c t

Article history: A rapid and reliable method is developed for Gum Arabic authentication based on Near Infrared (NIR)
Received 15 September 2016 spectroscopy and chemometric methods. On a large industrial collection of authentic gum Arabics, the
Received in revised form two major Acacia gum species, Acacia senegal and Acacia seyal could be assigned perfectly by the NIR
18 January 2017
spectroscopic method. In addition, a partial least squares (PLS) regression model is calibrated to predict
Accepted 1 February 2017
Available online 3 February 2017
the blending percentage of the two pure gum types, producing an accuracy, root mean square error of
cross validation (RMSECV) of 2.8%. Sampling of the Gum Arabic ‘tears’ is discussed, and it was determined
that subsamples from three ‘tears’ is required for a representative result. It is concluded that NIR
Keywords:
Gum Arabic
spectroscopy is a very powerful and reliable method for authenticity testing of Gum Arabic species.
Near infrared (NIR) spectroscopy © 2017 Elsevier Ltd. All rights reserved.
Sampling of ‘tears’
Acacia senegal
Acacia seyal
Principal component analysis (PCA)
Partial least squares (PLS) regression

1. Introduction
Gum Arabic is one of the most ancient and popular natural gums
and can be traced back to 2650 BCE (A. M. H. Abuarra, 2014). Ac-
cording to the definition of the Joint Expert Committee for Food
Additives (JECFA), gum Arabic are dried exudate lumps, the so-
called tears, obtained from the stem and branches of Acacia trees
(Phillips, Ogasawara, & Ushida, 2008). There are more than 1000
species of the genus Acacia gums (A. Abuarra, Hashim, Bauk,
Kandaiya, & Tousi, 2014), but only two of the Acacia species are
significant for commercial purposes: Acacia senegal (L.), which is
considered to be the best in quality due to a low quantity of tannins
(Egadu, Mucunguzi, & Obua, 2007) and comprises the majority of
global trade, and Acacia seyal (Del.), which produces a lower grade
of gum (Ibrahim, Osman, & Hassan, 2013). Acacia trees are abun-
dant in central Sudan, central and West Africa, tropical and semi-
tropical areas of the world (Hadi, Elderbi, & Mohamed, 2010;
* Corresponding author.
E-mail address: kms@food.ku.dk (K.M. Sørensen).
1
Yongjiang Dong and Klavs Martin Sørensen has contributed equally to this
work.
Wyasu & Okereke, 2012). Sudan is the leading producer of Acacia
gums worldwide, followed by Nigeria, Chad, Mali and Senegal
(Vanloot, Dupuy, Guiliano, & Artaud, 2012). A. Senegal and A. seyal
exudates exist as a multifunctional hydrocolloid with a highly
branched, neutral or slightly acidic, arabino-galactan-protein
complex containing calcium, magnesium, and potassium (Renard,
Lavenant-Gourgeon, Ralet, & Sanchez, 2006). The molecular
structure of the carbohydrate moieties of both gum Arabics are
constituted of chains of 1,3-linked b-D-galactopyranosyl units with
sidechains including primarily b-D-glucuronopyranose, a-L-arabi-
nofuranose, a-L-rhamnopyranose, and 4-O-methyl-b-D-glucur-
onopyranose (Li, Gan, Liu, Lin, & Huang, 2014). Gum Arabic has
wide applications, mainly in the food industry due to its emulsi-
fying, stabilizing, thickening, and binding properties, which makes
it useful for applications such as confectionery products (Verbeken,
Dierckx, & Dewettinck, 2003), beverages (Buffo, Reineccius, &
Oehlert, 2001; Mirhosseini, Tan, Hamid, & Yusof, 2008), fruits
(Maqbool, Ali, Alderson, Zahid, & Siddiqui, 2011), and as a micro-
encapsulating agent (Butstraen & Salaun, 2014). In addition, gum
Arabic is also widely used for industrial purposes as in traditional
lithography, printing, textiles (Patel & Goyal, 2015), and pharma-
ceutical applications (Ali, Ziada, & Blunden, 2009).
Authentication of food ingredients is an important issue for

http://dx.doi.org/10.1016/j.foodcont.2017.02.002
0956-7135/© 2017 Elsevier Ltd. All rights reserved.
 Y. Dong et al. / Food Control 78 (2017) 144e149
industrial quality control and international trading (Sørensen,
Khakimov, & Engelsen, 2016), in particular in cases such as the
two Acacia gum species which can be traded in a freeze-dried or
powdered form. Conventional methods to characterize (Gashua,
Williams, Yadav, & Baldwin, 2015) and to classify (Jurasek, Varga,
& Phillips, 1995) the two types of gums were based on analyzing
different physical and chemical parameters, such as intrinsic vis-
cosity, total nitrogen, sugar composition, and total glucuronic acid.
Using optical rotation is also an effective approach to classify A.
senegal and A. seyal (Biswas & Phillips, 2003), however, this uni-
variate method can relatively easily be “cheated” by adding sugars.
Compared to these complex, time-consuming and less-sensitive
techniques, spectroscopic techniques are more rapid, effective
and easily implemented, inspiring their development for gum
authentication applications. Nuclear Magnetic Resonance (NMR) is
an effective analytical technology to identify A. senegal and A. seyal
(Nie et al., 2013b, 2013a), but rather inaccessible for industrial high
throughput screening purposes. Infrared (IR) spectroscopy and
Raman spectroscopy, are also very useful as more convenient and
common methods. In combination with chemometrics, IR was used
to identify gum from other organic binding media such as waxes
and resins (Sarmiento et al., 2011), and to classify different gums
€
(Prado, Kim, Ozen, & Mauer, 2005). Vanloot et al. has demonstrated
the use of IR spectroscopy to classify A. senegal and A. seyal (Vanloot
et al., 2012). Also, Raman spectroscopy has shown promising
capability for identifying Arabic gum from other binder materials
(Pallipurath, Skelton, Ricciardi, Bucklow, & Elliott, 2013) and for
differentiating between gums of different origin (Srivastava,
Wolfgang, & Rodriguez, 2016).
Near infrared (NIR) spectroscopy, which has the advantage of
being robust, cost-effective and requiring less sample preparation,
has a long record of being used widely for industrial food analysis
applications (Luypaert, Massart, & Vander Heyden, 2007; Massie &
Norris, 1975; van den Berg, Lyndgaard, Sørensen, & Engelsen, 2013;
Williams & Norris, 1987) and should also be promising for gum
characterization. Indeed, the near infrared technique has been used
to study bean gum mixed with other biopolymers (Lundin, Stenlo € f,
& Hermansson, 1998), to determine the molecular weight of xan-
than gum (Song et al., 2015), and to identify the gum Arabic as a
paint binder (Ricciardi et al., 2012). Thus, the feasibility of using NIR
spectroscopy for quality control of Acacia gum has been demon-
strated, however, to our knowledge, no previous attempt has been
made to discriminate between the two gum Arabic species using
NIR spectroscopy. Quantification of gum mixtures is another
interesting issue for gum characterization and authentication, as
powdered gum Arabic may be easily adulterated by the addition of
cheaper and more inferior gums. Prado et al. accomplished the
quantitative analysis of individual “gum in gum” mixtures using IR
spectroscopy and partial least squares regression (Prado et al.,
2005). Vanloot et al. made a basic test on the adulteration of A.
senegal and A. seyal using IR spectroscopy (Vanloot et al., 2012).
Also, to our knowledge, the present study is the first to use NIR
spectroscopy to study the two Acacia gums mixed with each other
at different concentrations.
The primary aim of the present work is to investigate the
effectiveness and reliability of the NIRS technique for the authen-
tication of two Acacia gum exudates using a large industrial sample
collection. Gum mixtures are used for calibration model building in
order to predict the two Acacia gums.
2. Materials and methods
2.1. Gum sample preparation & measurements
A total of 26 Acacia gum samples were selected from a large
145
industrial collection provided by Toms Group A/S (Ballerup,
Denmark). The selected collection is composed of A. senegal sam-
ples (n ¼ 19) and A. seyal samples (n ¼ 7) originating from various
provenances. Only samples with unambiguous identification were
used in the following experiments.
Experiment 1 (see Fig. 1): for each of the 26 selected Gum Arabic
samples ten subsamples were prepared by randomly selecting
approximately 15 g of the non-uniform gum Arabic tears and then
ground into a fine powder with a coffee grinder mill. This experi-
mental design produced a total of 260 separate powder samples. All
samples were processed, labelled and stored in airtight containers
before any measurements.
Near-infrared spectra of the 260 samples were recorded in a
random order. Measurements were conducted using a QFA Flex
Fourier transform spectrometer (Q-Interline A/S, Roskilde,
Denmark) equipped with a reflectance kinetic powder sampler that
rotates a vial with the gum Arabic powder over the instrument
window. Spectra were recorded in the range from 1100 nm to
2500 nm using an InGa detector with a 16 cm1 resolution and 512
scans. The spectra were converted to log(1/R) units using a PFTE-
filled vial as a background reference. The background reference
was measured every hour of the experiment using the same mea-
surement conditions.
Experiment 2: In this experiment three samples of each of the A.
senegal and A. seyal types were chosen randomly (see Fig. 1). Pool
samples were produced for each of these selections with an equal
amount of the 10 available subsamples (from experiment 1). The
pool samples were then paired, again randomly, producing three
pairs with each one A. senegal and one A. seyal pool sample. Each of
the pairs was mixed into 11 new samples, with a blend of 0%e100%
A. senegal in A. seyal in 10% increments. This procedure resulted in
33 mixtures that were then measured using NIR spectroscopy as
described above.
3. Theory
In experiment 1, principal component analysis (PCA) (S. Wold,
Esbensen, & Geladi, 1987) was used to gauge if it was possible to
distinguish between the two species. PCA is the most widely
applied linear projection method for unsupervised exploratory
multivariate data analysis. It is often used in data analysis for class
detection as its ability to visualize the similarities and differences
between the spectra including outlier detection. The basic objective
of PCA is to extract lower dimensional orthogonal variables, prin-
cipal components or latent variables, from a large number of
original correlated variables. These components describe, in
decreasing order, the maximum variance of original data. PCA de-
composes the original data X (the NIR spectra) as
X ¼ TP T þ E (1)
where T, the scores matrix, represents the information of the pro-
jected data, and the loadings matrix P corresponds to the variables
contribution. E contains the residual variation (non-systematic in-
formation) not captured by the model.
The partial least squares (PLS) regression used in experiments 2
is a powerful and arguably the most widely used method for su-
pervised quantitative analysis (Haaland & Thomas, 1988). PLS
builds up a linear regression mode between predictor matrix X and
target matrix Y. PLS not only finds a smaller subspace explaining the
maximum amount of variation in X as PCA does, but also ensure the
coordinates of the new subspace that describe the response vector
Y well. Thus, PLS decomposes both matrices into several latent
variables, explaining maximum variations of X and Y. In PLS, y is
correlated with X as (Svante Wold, Sjo €stro
€ m, & Eriksson, 2001)
146 Y. Dong et al. / Food Control 78 (2017) 144e149
Fig. 1. Experimental design.
X ¼ TP T þ E∧y ¼ UQ T þ f ; U ¼ BT
where T and P are the respective score matrices of X and y, P and Q
the loadings and weights of X and y, respectively, and E and f the
respective residuals, and B the relation between the two pro-
jections. In this work, y is simply a row vector with the mixing
levels of the 33 samples.
3.1. Preprocessing and validation of models
In order to determine which is the most effective for the elim-
ination of the scatter from the raw spectra of the gum powders,
different preprocessing methods of the NIR spectra (Rinnan, van
den Berg, & Engelsen, 2009) were investigated, including MSC
(multiplicative scatter correction) (Geladi, MacDougall, & Martens,
1985), EMSC (extended multiplicative scatter correction) (Martens
& Stark, 1991), using the mean spectra as reference spectra with no
filtering options, and second derivatives (Savitzky Golay, 2. degree
polynomial, window size 5) (Savitzky & Golay, 1964). All the pre-
processed spectra were mean-centered before modelling with
either PCA or PLS.
Models were made using Matlab R2016b (Mathworks Inc,
Natick, MS, U.S.A.) using the PLS Toolbox version 8.2 (Eigenvector
Research Inc., Manson, WA, USA).
The sample set for experiment 2 has a high number of non-
independent samples, in that 11 species of the same physical
sample can be present in the calibration at once. To remedy this, a
cross validation scheme was selected where one whole concen-
tration series spanning 0%e100%, e.g. the 11 samples, was left out at
a time, in a leave-one-series-out validation scheme. Furthermore,
when the PLS model developed in experiment 2 is re-applied to
predict concentration in the experiment 1 data set, all 66 samples
originating from the 6 physical samples used for calibration is
excluded so that the prediction is independent of the calibration.
(2)
3.2. Model optimization
The developed PLS model was further optimized using an in-
terval PLS (iPLS) variable selection scheme (Nørgaard et al., 2000).
In iPLS, PLS models are fitted on subintervals of the full spectra,
which is then used to identify only those regions in the spectra that
contributes to the regression. The result of the iPLS model is a
subset of the original full spectrum variables, and a final PLS model
build on this selection is often less complex and therefore more
robust. Although several configurations of the iPLS algorithm (not
shown) has been tested, the optimum result was found using the
iPLS algorithm in forward mode with a window width of 15 variable
points.
4. Results and discussion
The mean NIR spectra, including their individual differences, of
the A. senegal and A. seyal samples in the region from 1100 to
2500 nm are shown along with their confidence intervals in Fig. 2.
From the figure, it is clear that the spectra of the two Acacia gums
are quite similar and difficult to distinguish from visual inspection.
Assignments of important absorption bands of the near infrared
spectra of gum Arabic are listed in Table 1.
Fig. 3 shows the PCA scores plot of all the 260 Acacia gum
subsamples and the 33 mixed gum samples. In each case, the PCs
showing the best discrimination ability were selected. As shown in
Fig. 3(a) the Acacia gum species cannot be distinguished by their
spectra without preprocessing. However, as Fig. 3(b) shows, EMSC
preprocessing of the spectra allows for an almost perfect separation
of the two gum species. In fact, when considering the extend of the
two groups in the direction of separation as indicated with the
arrow in Fig. 3(b), the two species each only spans approximately
 Y. Dong et al. / Food Control 78 (2017) 144e149 147

Fig. 2. (A) mean NIR spectra of the Acacia gums, Senegal (red line) and Seyal (green
line). (B) The difference spectrum of the two mean spectra. (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of
this article.)

25% of the total variance, providing a comfortable margin of sepa-

ration. The scores of the gum mixture samples are all located be-
tween the two pure gum samples, and show a clear trend with the
concentration from 0% to 100%, as depicted by the three lines in
Fig. 3(b). In general, EMSC preprocessing was found to yield the
best PCA and PLS results and was therefore used for all the sub-
sequent experiments.
A PLS model calibrated on the 66 gum mixture samples pro-
duces a model with r ¼ 0.68 and RMSECV ¼ 9.2%, using 5 compo-
nents. Variable selection using an iPLS with a window width of 15
Fig. 3. PCA score plot of pure Acacia senegal (red dots) and Acacia seyal (green dots)
spectral data points selected 4 unique sub bands centered at
samples, including three groups of gum mixtures (square dots), with Acacia senegal
1186 nm, 1509 nm, 1647 nm and 1686 nm respectively (see concentration increasing from 0% to 100%, blue lines indicate the gradually-changing
Fig. 4(b)), which allows for an optimal fit with r ¼ 0.98 and order. (a) PCA of the raw near infrared spectra, (b) PCA of the near infrared spectra
RMSECV ¼ 5.26% (Fig. 4(a)) using 2 principal components, a sig- after EMSC preprocessing. The black dashed arrow indicates the direction of separation
and increasing A. senegal concentration. (For interpretation of the references to colour
nificant improvement in accuracy and robustness. Of the selected
in this figure legend, the reader is referred to the web version of this article.)
bands three corresponds to the first overtone area of N-H stretches
(1509 nm) and C-H stretches (1647 nm and 1686 nm, the latter
presumably first overtone of aromatic CH stretch from aromatic
amino acids) whereas the fourth band (1186 nm) is found in the information stems both largely from the peptide moieties and
second overtone aliphatic C-H stretches. Apparently, the model possibly also from a little from the carbohydrate moieties (aliphatic
CH stretching). This discrimination pattern between A. senegal and
A. seyal, that rely primarily on the proteinaceous part of the gum
arabic's, is quite surprising, since the gum Arabic varieties are
Table 1
NIR absorption bands of gum Arabic and their tentative assignments.
normally differentiated per their detailed carbohydrate composi-
tion(Nie et al., 2013a; Vanloot et al., 2012).
Characteristic Vibrational mode Structure The PLS model built on the three gum mixture series were
wavelength [nm]
subsequently used on the 260 pure Acacia gum samples, shown in
1210 C-H stretch second galactose, arabinose Fig. 5 to illustrate how efficient NIRS can be used for authentication
overtone of CH2
of the gum species. The application of the previously developed
1360 C-H stretch first Rhamnose
overtone þ CH deformation model to predict the A. seyal concentration shows that an ideal
of CH3 separation can be achieved for the two species by thresholding the
1450 O-H stretch first overtone H2O prediction at a limit between the two group centers. Equally
C]O third overtone glucorunic acid interesting is the fact that using a prediction of only one single sub-
1580 O-H stretch first overtone galactose/galactan
1770 C-H stretch first overtone of galactose, arabinose
sample for each physical sample will be sufficient to achieve the
CH2 ideal separation e there is no need for sampling multiple sub-
1935 O-H stretch þ O-H H2O samples for each physical sample. Note that the A. seyal samples
deformation have smaller standard deviations compared to A. senegal. This may
2100 O-H deformation þ C-O arabino-galactan
be due to the fact that the A. seyal ‘tears’ were consistently smaller
stretching
2323 CH stretch þ deformation of arabino-galactan and thus a greater number of tears were needed to prepare the
CH2 same weight of sample powder, which in turn may have led to more
2490 C-H stretch and C-C stretch arabino-galactan uniform subsamples.
148 Y. Dong et al. / Food Control 78 (2017) 144e149
Fig. 4. (A) Regression model of concentration (cross validated) of Acacia senegal in the
Acacia senegal/Acacia seyal mixture. The colored dots each represents a cross-validated
prediction for one of three concentration series (see text) for the two-component PLS
model (red line) yielding r ¼ 0.98, RMSECV ¼ 5.26%. The green dotted line is the ideal
1:1 relationship. (B) Mean absorbance spectra of the 33 Acacia seyal/Acacia Senegal
mixtures (blue line), with an overlay of the iPLS selected bands (red lines). (For
interpretation of the references to colour in this figure legend, the reader is referred to
the web version of this article.)
5. Conclusions
A comprehensive authentication method for gum Arabic using
near infrared spectroscopy was developed. The classification of A.
senegal and A. seyal has been demonstrated to be totally discrimi-
nated without any ambiguity in a PCA model using EMSC pre-
processed spectra. Besides this qualitative identification, the
quantitative analysis of the NIR technique combing with PLS for
gum mixtures is also quite promising. The predictive model has
been reduced in complexity by selecting relevant sub-spectra using
iPLS variable selection, and shows an excellent result with low
RMSECV of 5.26% (and r ¼ 0.98) using just two components.
Moreover, by using the gum mixture PLS regression model to
predict all the Acacia gums, the authentication ability of the NIR
technique has been demonstrated, with an ideal separation of the
two gum-species. Surprisingly, it is demonstrated that the perfect
NIRS authentication method rely on spectral information which
primarily stems from the proteinaceous moiety of the gum arabic's.
This study illustrates that NIRS can be used as a simple, quick,
and reliable analytical technique for quality control and quantita-
tive analysis in the gum industries, with almost no sample prepa-
ration time.
Fig. 5. Predicted concentrations (cross validated) of 20 gums (each with ten sub-
samples) based on the PLS gum mixture model. The red and green crosses indicate
Acacia senegal and seyal subsamples, respectively. The mean values and standard
deviations of each gum sample are shown by blue squares (mean) and error bars (±1
standard deviation around mean). The dashed lined shows the ideal separation
threshold between the two groups. (For interpretation of the references to colour in
this figure legend, the reader is referred to the web version of this article.)
Funding
This work was supported by a XCHINAX PhD grant to Yongjiang
Dong.
Acknowledgements
The authors wish to thank Q-interline (www.qinterline.dk) who
supplied the spectroscopic equipment used, MSc student Niels
Sommer for working with different aspect of Gum Arabic quality
and to Jørgen Schou for enthusiastic discussions and for making the
gum Arabic collection available for us.
References
Abuarra, A. M. H. (2014). Fabrication and characterization of gum arabic bonded
Rhizophora spp. Particleboard Head Phantom For Pet/Ct Applications. Universiti
Sains Malaysia.
Abuarra, A., Hashim, R., Bauk, S., Kandaiya, S., & Tousi, E. T. (2014). Fabrication and
characterization of gum Arabic bonded Rhizophora spp. particleboards. Mate-
rials & Design, 60, 108e115.
Ali, B. H., Ziada, A., & Blunden, G. (2009). Biological effects of gum arabic: A review
of some recent research. Food and Chemical Toxicology, 47(1), 1e8.
van den Berg, F., Lyndgaard, C. B., Sørensen, K. M., & Engelsen, S. B. (2013). Process
analytical technology in the food industry. Trends in Food Science & Technology,
31(1), 27e35.
Biswas, B., & Phillips, G. O. (2003). Computation of specific optical rotation from
carbohydrate composition of exudate gums Acacia Senegal and Acacia seyal.
Food Hydrocolloids, 17(2), 177e189.
Buffo, R., Reineccius, G., & Oehlert, G. (2001). Factors affecting the emulsifying and
rheological properties of gum acacia in beverage emulsions. Food Hydrocolloids,
15(1), 53e66.
Butstraen, C., & Salaun, F. (2014). Preparation of microcapsules by complex coac-
ervation of gum Arabic and chitosan. Carbohydrate Polymers, 99, 608e616.
Egadu, S. P., Mucunguzi, P., & Obua, J. (2007). Uses of tree species producing gum
arabic in Karamoja, Uganda. African Journal of Ecology, 45(s1), 17e21.
Gashua, I. B., Williams, P. A., Yadav, M. P., & Baldwin, T. C. (2015). Characterisation
and molecular association of Nigerian and Sudanese Acacia gum exudates. Food
Hydrocolloids, 51, 405e413.
Geladi, P., MacDougall, D., & Martens, H. (1985). Linearization and scatter-correction
for near-infrared reflectance spectra of meat. Applied Spectroscopy, 39(3),
491e500.
Haaland, D. M., & Thomas, E. V. (1988). Partial least-squares methods for spectral
analyses. 1. Relation to other quantitative calibration methods and the extrac-
tion of qualitative information. Analytical Chemistry, 60(11), 1193e1202.
Hadi, A., Elderbi, M. A., & Mohamed, A. (2010). Effect of gum Arabic on coagulation
 Y. Dong et al. / Food Control 78 (2017) 144e149
system of albino rats. International Journal of PharmTech Research, 2(3),
1762e1766.
Ibrahim, O. B., Osman, M. E., & Hassan, E. A. (2013). Characterization and simple
fractionation of Acacia Senegal. Journal of Chemica Acta, 2(1), 11e17.
Jurasek, P., Varga, S., & Phillips, G. (1995). Classification of natural gums. VII. Re-
lationships between the series Vulgares (Acacia Senegal) and Gummiferae
(Acacia seyal). Food Hydrocolloids, 9(1), 17e34.
Li, Y., Gan, W., Liu, X., Lin, T., & Huang, B. (2014). Dispersion mechanisms of Arabic
gum in the preparation of ultrafine silver powder. Korean Journal of Chemical
Engineering, 31(8), 1490e1495.
Lundin, L., Stenlo €f, B., & Hermansson, A.-M. (1998). NIR spectra in relation to
viscoelastic properties of mixtures of Na-k-Carrageenan, locust bean gum and
casein. Food Hydrocolloids, 12(2), 189e193.
Luypaert, J., Massart, D., & Vander Heyden, Y. (2007). Near-infrared spectroscopy
applications in pharmaceutical analysis. Talanta, 72(3), 865e883.
Maqbool, M., Ali, A., Alderson, P. G., Zahid, N., & Siddiqui, Y. (2011). Effect of a novel
edible composite coating based on gum arabic and chitosan on biochemical and
physiological responses of banana fruits during cold storage. The Journal of
Agricultural and Food Chemistry, 59(10), 5474e5482.
Martens, H., & Stark, E. (1991). Extended multiplicative signal correction and
spectral interference subtraction: New preprocessing methods for near infrared
spectroscopy. Journal of Pharmaceutical and Biomedical Analysis, 9(8), 625e635.
Massie, D., & Norris, K. (1975). A high-intensity spectrophotometer inferfaced with
a computer for food quality measurement. Transactions of the American Society
of Agritultural Engineers, 18(1), 173e176.
Mirhosseini, H., Tan, C. P., Hamid, N. S. A., & Yusof, S. (2008). Effect of Arabic gum,
xanthan gum and orange oil contents on z-potential, conductivity, stability, size
index and pH of orange beverage emulsion. Colloids and Surfaces A: Physico-
chemical and Engineering Aspects, 315(1e3), 47e56.
Nie, S.-P., Wang, C., Cui, S. W., Wang, Q., Xie, M.-Y., & Phillips, G. O. (2013a). The core
carbohydrate structure of Acacia seyal var. seyal (Gum arabic). Food Hydrocol-
loids, 32(2), 221e227.
Nie, S.-P., Wang, C., Cui, S. W., Wang, Q., Xie, M.-Y., & Phillips, G. O. (2013b). A further
amendment to the classical core structure of gum arabic (Acacia Senegal). Food
Hydrocolloids, 31(1), 42e48.
Nørgaard, L., Saudland, A., Wagner, J., Nielsen, J. P., Munck, L., & Engelsen, S. B.
(2000). Interval partial least-squares regression (iPLS): A comparative chemo-
metric study with an example from near-infrared spectroscopy. Applied Spec-
troscopy, 54(3), 413e419.
Pallipurath, A., Skelton, J., Ricciardi, P., Bucklow, S., & Elliott, S. (2013). Multivariate
analysis of combined Raman and fibre-optic reflectance spectra for the iden-
tification of binder materials in simulated medieval paints. Journal of Raman
Spectroscopy, 44(6), 866e874.
Patel, S., & Goyal, A. (2015). Applications of natural polymer gum arabic: A review.
International Journal of Food Properties, 18(5), 986e998.
Phillips, G. O., Ogasawara, T., & Ushida, K. (2008). The regulatory and scientific
149
approach to defining gum arabic (Acacia Senegal and Acacia seyal) as a dietary
fibre. Food Hydrocolloids, 22(1), 24e35.
€
Prado, B. M., Kim, S., Ozen, B. F., & Mauer, L. J. (2005). Differentiation of carbohydrate
gums and mixtures using Fourier transform infrared spectroscopy and che-
mometrics. Journal of Agricultural and Food Chemistry, 53(8), 2823e2829.
Renard, D., Lavenant-Gourgeon, L., Ralet, M.-C., & Sanchez, C. (2006). Acacia s enegal
gum: Continuum of molecular species differing by their protein to sugar ratio,
molecular weight, and charges. Biomacromolecules, 7(9), 2637e2649.
Ricciardi, P., Delaney, J. K., Facini, M., Zeibel, J. G., Picollo, M., Lomax, S., et al. (2012).
Near infrared reflectance imaging spectroscopy to map paint binders in Situ on
illuminated manuscripts. Angewandte Chemie International Edition, 51(23),
5607e5610.
Rinnan, Å., van den Berg, F., & Engelsen, S. (2009). Review of the most common pre-
processing techniques for near-infrared spectra. TRAC Trends in Analytical
Chemistry, 28(10), 1201e1222.
Sarmiento, A., Perez-Alonso, M., Olivares, M., Castro, K., Martinez-Arkarazo, I.,
Fernandez, L. A., et al. (2011). Classification and identification of organic binding
media in artworks by means of Fourier transform infrared spectroscopy and
principal component analysis. Analytical and Bioanalytical Chemistry, 399(10),
3601e3611.
Savitzky, A., & Golay, M. J. E. (1964). Smoothing and differentiation of data by
simplified least squares procedures. Analytical Chemistry, 36(8), 1627e1639.
Song, Z., Ling, P., Zang, H., Li, L., Wang, J., Jin, Y., et al. (2015). Development, vali-
dation and influence factor analysis of a near-infrared method for the molecular
weight determination of xanthan gum. Carbohydrate Polymers, 115, 582e588.
Sørensen, K. M., Khakimov, B., & Engelsen, S. B. (2016). The use of rapid spectro-
scopic screening methods to detect adulteration of food raw materials and
ingredients. Current Opinion in Food Science, 10, 45e51.
Srivastava, H. K., Wolfgang, S., & Rodriguez, J. D. (2016). Rapid screening of guar gum
using portable Raman spectral identification methods. Journal of Pharmaceutical
and Biomedical Analysis, 118, 387e392.
Vanloot, P., Dupuy, N., Guiliano, M., & Artaud, J. (2012). Characterisation and
authentication of A. Senegal and A. seyal exudates by infrared spectroscopy and
chemometrics. Food Chemistry, 135(4), 2554e2560.
Verbeken, D., Dierckx, S., & Dewettinck, K. (2003). Exudate gums: Occurrence,
production, and applications. Applied Microbiology and Biotechnology, 63(1),
10e21.
Williams, P., & Norris, K. (1987). Near-infrared technology in the agricultural and food
industries. American Association of Cereal Chemists, Inc.
Wold, S., Esbensen, K., & Geladi, P. (1987). Principal component analysis. Chemo-
metrics and Intelligent Laboratory Systems, 2(1e3), 37e52.
Wold, S., Sjo € stro
€ m, M., & Eriksson, L. (2001). PLS-regression: A basic tool of che-
mometrics. Chemometrics and Intelligent Laboratory Systems, 58(2), 109e130.
Wyasu, G., & Okereke, N. (2012). Improving the film forming ability of gum Arabic.
Journal of Natural Product and Plant Resources, 2, 314e317.