Food and Humanity 2 (2024) 100194

Contents lists available at ScienceDirect

Food and Humanity

journal homepage: www.editorialmanager.com/foohum/journal_overview.html

Application of UV–vis spectrophotometry and chemometrics to investigate

adulteration by glucose syrup in Brazilian polyfloral honey
Aline Nunes a, *, Gadiel Zilto Azevedo a, Beatriz Rocha dos Santos a, Giuseppina Pace
Pereira Lima b, Sidnei Moura c, Marcelo Maraschin a
a
Plant Morphogenesis and Biochemistry Laboratory, Federal University of Santa Catarina, 1346, Admar Gonzaga Road, 88034-000 Florianópolis, Santa Catarina,
Brazil
b
UNESP, São Paulo State University, Institute of Biosciences, Botucatu, São Paulo, Brazil
c
LBIOP – Laboratory of Biotechnology of Natural and Synthetics Products, Technology Department, Biotechnology Institute, University of Caxias do Sul, Caxias do Sul,
Brazil

A R T I C L E I N F O A B S T R A C T

Keywords: The quality control of food in an easy, fast, and low-cost way has been increasingly sought by industry, and
Analytical methods methods based on spectroscopy have stood out in recent years. Honey is among the most adulterated foods
Food adulteration worldwide, and fraud detection is difficult because it is a chemically complex matrix. In this sense, this study
Glucose syrup
aimed to develop a simple, fast, and cheap method based on UV–vis spectrophotometry coupled to chemometrics
Sugar adulteration
for discriminating pure Brazilian polyfloral honey from adulterated ones. For that, 139 real samples of polyfloral
honey produced in southern Brazil (Santa Catarina state) were analyzed by UV–vis scanning spectrophotometry
(λ 200–800 nm), and a set of 42 samples (i.e., 30%) were intentionally adulterated with glucose syrup (10–60%,
w/w). Three wavelength windows were shown to be relevant in the UV–vis spectra (e.g., 260–500 nm, 260–360
nm, and 280–300 nm), and these data sets were further analyzed by principal component analysis (PCA). Pure
and adulterated honey samples were better discriminated by computing the 280–300 nm spectral window data
(PC1 and PC2 = 98% total variance of the data set). To validate the method, 59 samples produced in south­
eastern Brazil (São Paulo state) were included in the descriptive model, which allowed identifying pure and
adulterated honey again (PC1 + PC2 = 97% of the data set’s total variance). Thus, these findings support that
UV–vis coupled with PCA proved to be a simple and efficient method to discriminate pure honey from adul­
terated ones, providing a fast and inexpensive protocol to monitor honey quality.

1. Introduction the case of imports, as many countries, despite adopting the Interna­
tional Honey Commission (EU)’s recommendations, have established
Honey is highlighted as a complete food due to its chemical divergent reference limits for the physicochemical traits of honey.
composition, containing carbohydrates, organic acids, proteins, en­ Therefore, it appears essential to standardize current legislation and to
zymes, minerals, vitamins, and other functional compounds (Ali et al., develop new, reliable, faster, cheaper, and environmentally friendly
2020). These components are responsible for its nutraceutical and methods for detecting adulteration (Thrasyvoulou et al., 2018).
therapeutic properties, including antidiabetic, antimicrobial, and anti­ In this sense, over the past years spectrometric (nuclear magnetic
cancer activities. Because of these significant biological effects, honey resonance – NMR; and mass spectrometry – MS) and chromatographic
ranks among the most consumed foods worldwide (Cianciosi et al., (liquid – LC and gas – GC) analytical methods have been improved to
2018). detect honey adulteration. However, such techniques demand expensive
For economic reasons, honey has continuously ranked among the top equipment, in addition to the need for recurring maintenance and hiring
three adulterated foods, following milk and olive oil (García, 2018). of specialized people to perform the analysis. Besides, samples also need
Tampering methods have improved, making fraud identification more to undergo specific preparations (Gerhardt et al., 2018; Se et al., 2019),
challenging (Fakhlaei et al., 2020). This issue is even more pertinent in what is time consuming and, in some cases, the use of chemicals is

* Corresponding author.
E-mail address: [email protected] (A. Nunes).

https://doi.org/10.1016/j.foohum.2023.12.002
Received 3 October 2023; Received in revised form 30 November 2023; Accepted 6 December 2023
Available online 9 December 2023
2949-8244/© 2023 Elsevier B.V. All rights reserved.
A. Nunes et al. Food and Humanity 2 (2024) 100194

requested. crystals and homogenization of the honey constituents. Furthermore,

In turn, ultraviolet and visible (UV–vis) spectroscopy is a fast, easy, 0.6 mg honey was weighed, added of 10 mL distilled water, following
reliable, and low-cost technique almost unexplored for the analysis and magnetic stirring (5 min) and centrifuging (2000 rpm, 5 min). The
detection of honey adulterations so far. In fact, few studies have reported UV–vis scanning spectra (n = 3/sample) were recorded over the ultra­
such methods for identifying corn syrup, agave syrup, and cane molasses violet (200–380 nm) and visible (380–800 nm) spectral windows,
as adulterants in honey (Guellis et al., 2020; Souza et al., 2021; Valinger through a Biochrom Libra S22 UV–vis spectrophotometer (Biochrom,
et al., 2021; Dimakopoulou-Papazoglou et al., 2023). Thus, aiming to Brazil) with a 2 nm/data point resolution (Nunes et al., 2023). The
contribute to the advancement of quality control protocols, this study spectra were obtained using the Biochrom software (Version 2.10.0.0).
aimed to develop a simple, fast, and cheap method based on UV–vis
spectrophotometry coupled to chemometrics for discriminating pure
Brazilian polyfloral honey from adulterated ones. 2.3. Honey adulteration tests by UV–vis spectrophotometry

2. Material and methods For further adulteration tests, 42 pure honey samples were randomly
chosen and added of glucose syrup according to their harvest seasons, as
2.1. Sample collection follows (2018–2019 harvest - 20 samples, 2019–2020 harvest - 15
samples, 2020–2021 harvest - 7 samples), comprising 30% of the sam­
One hundred and thirty-nine samples of polyfloral honey were ples from each harvest. Glucose syrup was added to the pure honey
collected during the 2018–2019 (n = 66 samples) and 2019–2020 (n =
51 samples) harvest seasons. Additionally, another 22 samples were Table 1
further collected in the 2020–2021 production season from different Concentrations of pure honey and glucose syrup used in the adulteration test.
locations in southern Brazil (Santa Catarina state – Fig. 1; Supplemen­ Glucose concentration (%) Pure honey (mg) Glucose syrup (mg)
tary material, Table S1).
10 0.54 0.06
20 0.48 0.12
30 0.42 0.18
2.2. Analysis of pure honeys by UV–vis spectrophotometry 40 0.36 0.24
50 0.30 0.30
In preparation for analysis, all individual samples were heated in a 60 0.24 0.36
water bath (15 min, 40 ◦ C) for complete solubilization of the sugar

Fig. 1. Location of honey samples collection in the state of Santa Catarina, southern Brazil.
Source: Prepared by the author (2023)

2
A. Nunes et al. Food and Humanity 2 (2024) 100194

samples in concentrations ranging from 10% to 60% (w/w) as shown in for all analyzes (Nunes et al., 2022; Nunes et al., 2023).
Table 1, following homogenization. The same UV–vis scanning protocol For purposes of validation of the PCA-descriptive model built, fifty-
(item 2.2.) applied to profile the pure honey was adopted for the adul­ nine honey samples collected from different locations in southeastern
terated samples (Nunes et al., 2023). All the honey samples were Brazil (São Paulo state – 2020–2021 harvest; Supplementary material,
analyzed in triplicate (n = 3), totaling 756 analyzes. Table S2) were added to the data set, considering the absorbance signals
recorded over the 280–300 nm spectral window. By doing so, it was
intended to verify if the developed model allowed one to discriminate
2.4. Statistical and chemometric analysis pure and adulterated honey samples produced in other geographic re­
gions that firstly used to discover underlying patterns in samples, i.e.,
The UV–vis spectra recorded with support of the Biochrom software glucose syrup adulteration. After the results obtained by PCA, linear
were exported as an.xlsx-format data set to be further analyzed in the discriminant analysis (LDA) was performed in the Past software (Version
Microsoft Excel® environment. The average UV–vis spectrum for each 4.03) as a classification model.
sample (n = 3) was calculated and further used for chemometric anal­
ysis. For that, the UV–vis data set firstly collected swept the 200–800 nm 3. Results
spectral window, but since no relevant signals were found after 500 nm
wavelength, only the data collected in the region between 260 nm and The visual analysis of the UV–vis spectra revealed a significant
500 nm was further analyzed. Thus, principal components analysis discrepancy on absorbance intensities between the pure honey samples
(PCA) was applied to the UV–vis data set considering different spectral and the adulterated ones (Fig. 2). Indeed, the adulterated samples
regions, as follows: 260–500 nm (whole spectrum – UV and visible exhibited prominent absorbances across the entire spectrum compared
absorbance regions), 260–360 nm (UV absorbances), and 280–300 nm to pure honey, irrespective the amount of glucose syrup added to that
(phenolic and flavonoid compounds absorbance). Importantly, other food (i.e., 10–60%) for adulteration. Among the pure honeys, a very
spectral windows were also analyzed, however, no satisfactory results similar profile was observed, with only a difference being in absorbance.
regarding the discrimination of pure and adulterated honey samples The UV–vis spectral profiles of pure honey have been previously
were found (data not shown). PCA was performed by importing the.xlsx reported and revealed a similar pattern of absorbances, even with
data set into the Unscrambler® X software (Version 10.4), following the distinct sample preparation protocols and different geographical and
pretreatment of the spectral data with the Savitzky-Golay filter botanical origins of the samples. Nonetheless, visual analysis of the
(smoothing – 3 points) to increase the signal-to-noise ratio without honey’s spectral profiles is usually not reliable enough to discriminating
altering the signals of the spectra. Subsequently, principal components between pure honey from fake ones. To overcome this limitation,
were calculated using the singular value decomposition (SVD) algorithm

Fig. 2. UV–vis spectroscopic profiles of pure honeys samples and those adulterated with glucose syrup at six concentrations (10–60%). A) 139 samples of pure honey
and 256 samples of adulterated honey (all harvests, λ = 260–500 nm); B) 66 samples of pure honey and 20 samples of adulterated honey from the 2018–2019 harvest
(λ = 280–310 nm); C) 51 samples of pure honey and 15 samples of adulterated honey from the 2019–2020 harvest (λ = 280–310 nm); D) 22 samples of pure honey
and 7 samples of adulterated honey from the 2020–2021 harvest (λ = 280–310 nm).

3
A. Nunes et al. Food and Humanity 2 (2024) 100194

statistical and computational tools based on multivariate statistical phenolic and flavonoid compounds. By doing so, it was possible to
techniques and machine learning have been developed and applied to improve the discrimination of samples, by grouping most of the adul­
spectroscopic and chromatographic data sets. These tools help extract terated and pure honey in the right (PC1 +) and left (PC1-) quadrants,
relevant information that allows identifying forgery in honey. This way, respectively (Fig. 4). In fact, this model showed to be improved in
in a second experimental approach descriptive models were built aiming respect to the previous ones, as PC1 and PC2 accounted for 98% of total
to discriminate pure honey samples from the adulterated ones, by variance. From all tested samples, only 1.2% (n = 3) samples containing
applying principal component analysis to the UV–vis data set. Experi­ 10% glucose syrup were found in the left, as 3 pure honey samples (~
mentally, it was noted that absorbance signals at wavelengths < 260 nm 2.2%) remained in the right quadrant (Fig. 4).
were quite noisy, being suppressed of the data set. In the same line, from To complement the study and to validate the descriptive model built,
500 nm onward, no relevant spectral signals were found. After this data 59 honey samples produced and collected in the State of São Paulo
preprocessing step, PCA was performed by computing absorbances (southeastern Brazil) were tested alongside the previous sample groups
detected in both UV and visible spectral regions, i.e., from 260 to (i.e., honey adulterated with glucose syrup and pure honey). The
500 nm. Secondly, a narrower window of wavelengths was considered, calculation of the principal components PC1 and PC2 allowed for the
focusing on UV absorbances (260–360 nm). Finally, absorption signals accumulation of 98% of the total variance in the dataset when analyzing
associated with phenolic and flavonoid compounds (280–300 nm) were the absorbances recorded at 280–300 nm wavelengths. Furthermore,
also utilized for calculating of the principal components of the the descriptive model indicated only two samples from São Paulo and
descriptive models built. The choice of this spectral region was based on five samples from Santa Catarina in the adulteration quadrants (Fig. 5).
the absorption signals associated with phenolic and flavonoid com­ After verifying the groupings in the PCA, the linear discriminant
pounds, along with their contents, which were determined in a previous analysis (LDA) was performed, where 99.8% of the variance was verified
study by our research group. on axis 1% and 0.20% on axis 2. 96.43% of the adulterated honey
The results found after calculating of the principal components from samples were grouped in the right quadrants, contrary to what occurred
the spectroscopic dataset within the 260–500 nm spectral window are for pure honeys from SC and SP. Two SP-honeys were found in the
presented in Fig. 3. The samples are predominantly distributed along the quadrants referring to adulterated honeys, which demonstrates that
PC1 axis, which explaind 83% of the variability in the dataset. Addi­ these must have undergone adulteration with glucose syrup. In loadings,
tionally, a distinct separation between pure honey and adulterated it is observed that from 280 to 300 nm, there is a greater correlation with
samples can be observed along the PC2 axis thataccounted for 13% of adulterated honeys, i.e., these have higher absorbance when fraudulent
the dataset’s variability. (Fig. 6).
Although PC1 and PC2 together accounted for a total variance of Thus, the UV–vis/PCA protocol described herein marks the initiation
96% in the spectral data within the 260–500 nm region, a second of a project to support the creation of a database for the State of Santa
calculation of the principal components was performed to achieve a Catarina, which could subsequently be applied to other Brazilian regions
more effective separation of pure honey from adulterated ones Thus, by for the easy, fast, and secure detection of honey adulteration with
applying PCA to the absorbance data detected over the UV-region, i.e., glucose syrup. Inspection and research bodies in the state will have
260–360 nm, the total variance explained by PC1 and PC2 was not access to the dataset provided in this study, enabling the analysis and
increased (PC1 + PC2 = 96%) compared to the previous descriptive classification of new honey samples and adding robustness to the
model. established protocol.
Finally, a third descriptive model was built by calculating the prin­ Considering the different types of adulterants used in honey, there is
cipal components of the UV–vis data set encompassed in the a need to test the protocol for other adulterant products, such as sugar
280–300 nm spectral window, typically associated to the absorbances of syrup and starch. This approach will enable the creation of an even more

Fig. 3. Score plot of the principal component analysis of the UV–vis (260–500 nm) dataset of pure and adulterated honey samples collected in southern Brazil (Santa
Catarina state), over the 2018–2019, 2019–2020, and 2020–2021 harvest seasons.

4
A. Nunes et al. Food and Humanity 2 (2024) 100194

Fig. 4. Score plot of the principal component analysis of the UV–vis (280–300 nm) dataset of pure and adulterated honey samples collected in southern Brazil (Santa
Catarina state), over the 2018–2019, 2019–2020, and 2020–2021 harvest seasons.

Fig. 5. Principal component scores plot of the UV–vis (280–300 nm) dataset of pure and adulterated honey samples collected in southern Brazil (Santa Catarina
state), over the 2018–2019, 2019–2020, and 2020–2021 harvest seasons.
* SP – state of São Paulo, southeastern Brazil; SC – state of Santa Catarina, south Brazil.

robust database for analysis. Similarly, the method developed needs to control, such as no sample preparation steps allowing direct food anal­
be tested regarding its robustness on honey from other countries. To ysis, fast spectral reading, and often without the use of chemical prod­
achieve this, it will be essential to consider glucose syrup sold on the ucts, being therefore ecologically friendly (Aslam et al., 2023; Haque
domestic markets abroad, validating the technique with scenarios that et al., 2021; Valinger et al., 2021). Coupled to spectroscopic analysis,
could potentially occur in the beekeeping sector worldwide. which might give rise thousands of data from chemically complex
matrices such as foods, chemometrics could be helpful to recognize
4. Discussion chemical patterns and interactions between compounds and/or other
constituents present in food samples (Roberts et al., 2018; Yu et al.,
Spectroscopic techniques have been highlighted by their features 2018). Such an approach has allowed to expand the amount of useful
and advantages over other analytical techniques for food quality information present in a certain spectroscopic data set, which would be

5
A. Nunes et al. Food and Humanity 2 (2024) 100194

Fig. 6. Linear discriminant analysis of the UV–vis (280–300 nm) dataset of pure and adulterated honey samples collected in southern Brazil (Santa Catarina state),
over the 2018–2019, 2019–2020, and 2020–2021 harvest seasons.
* SP – state of São Paulo, southeastern Brazil; SC – state of Santa Catarina, south Brazil.

more hardly recovered without the use of machine learning tools typi­ as adulterant in honey, proving that PCA was efficient, with PC1 and
cally adopted in chemometrics. PC2 explaining 100% of the total variance of the data set. Dimakopou­
Among spectroscopic techniques, near-infrared (NIR) has been most lou-Papazoglou et al. (2023) demonstrated the applicability of UV–vis to
used for the analysis of adulteration in honey (e.g., Ferreiro-González determine adulteration by the addition of syrups and colorants in honey
et al., 2018; Aliaño-González et al., 2019; Huang et al., 2020), with few collected in different Mediterranean countries. The authors point out
studies using UV–vis having been reported, although the technique has that using the spectral range of 220–550 nm and chemometrics (e.g.,
several advantages for food analysis, such as simplicity, versatility, and PCA, PLS, and DD-SIMCA) it was possible to find the good predictive
economy (Kharbach et al., 2023). capacity of UV–vis spectroscopy for determining fraud in honey.
In this line, we have recently reported a metabolomic study on the Studies have also delved into machine learning classifiers based on
discrimination of Brazilian polyfloral honey according to their UV–vis spectra. Mitra et al. (2023) point to the efficiency of the neural
geographic and floral origins, by coupling three spectroscopic tech­ network and random forest to distinguish adulterated honey from un­
niques, e.g., UV–vis, NIR, and NMR analyzes to chemometrics (i.e., PCA adulterated one. In addition to this study, Razavi and Kenari (2023)
and PLS). In that study, it was shown that UV–vis and NMR spectros­ usedUV–vis coupled to machine learning algorithm and described that
copies were the best approaches for honey discrimination, as UV–vis support vector regression (SVR) presented better calibration than partial
allowed detecting absorbance signals that may be associated to com­ least squares regression (PLSR) to detect adulteration in honey. These
pounds added to honey, eventually suggesting adulteration (Nunes studies have demonstrated the applicability of the UV–vis technique
et al., 2022). Thus, the findings herein described are part and a based on different chemometric approaches.
continuation of a broad project of typification and characterization of In our study, we have proposed the use of PCA for building
honey produced in southern Brazil, from where some honeys, awarded descriptive models of samples as a dimensionality reduction technique.
globally due to their superior quality, came. It is important to point out that reduction in data dimensionality pro­
Similar to the results obtained by our research group in recent years, vided by PCA is essential for machine learning application, especially
a few works have reported the use of UV–vis spectroscopy associated when high-throughput analytical techniques are adopted, as usually
with PCA to detect adulteration in honey. For instance, Valinger et al. found in metabolomic analysis as herein described. In addition to
(2021) used of UV–vis-NIR to detect honey adulteration with corn syrup. facilitate the understanding of variables effects, PCA allows for
The authors UV–vis scanned (325–900 nm) 135 samples of pure and improving the results of the classifier with an unsupervised method
adulterated (10%− 90% corn syrup) acacia honey collected in 2018. PCA (Jiménez-Carvelo et al., 2019; Musulin & Basualdo, 2012). In this study,
of the UV–vis data set showed a total variance of 94.5% (PC1 and PC2), UV–vis scanning spectrophotometry and PCA (98% total variance)
with separation of the samples accordingly, proving the potential of that assessment were considered, as they are easily reproducible for honey
analytical technique coupled to PCA to detect adulteration in honey. analysis in a fast, environmentally friendly, and robust way, being
In a similar way, Souza et al. (2021) showed the discrimination of applicable to samples from any geographic region and country.
pure and adulterated honey with corn syrup, agave syrup, and sugarcane The food quality control is essential for consumer’s safety issues, for
molasses using UV–vis (320–800 nm) and the Data-Driven Soft Inde­ instance, demanding appropriate analytical tools, as well as data pro­
pendent Modeling of Class Analogy (DD-SIMCA). The correct classifi­ cessing. In Brazil, cases of honey adulteration have been reported by
cation of the entire sample set was verified, reaching 100% sensitivity Carvalho et al. (2020); Franz et al. (2018), and Tibola et al. (2018), in
and specificity. Finally, Guellis et al. (2020) have used voltammetry the past years. However, there is no national or regional databases
(Cu/CuO electrode) and UV–vis spectrophotometry to detect corn syrup available for traceability and monitoring of frauds in honey. Likewise,

6
A. Nunes et al. Food and Humanity 2 (2024) 100194

the ongoing Brazilian legislation is based on physicochemical parame­ Appendix A. Supporting information
ters dating from the year 2000 (Brazil, 2000), which demonstrates the
need to improve the techniques considered in assessing the quality of Supplementary data associated with this article can be found in the
honey for human consumption. online version at doi:10.1016/j.foohum.2023.12.002.

5. Conclusion References

The UV–vis technique coupled with PCA proved to be an efficient Ali, A., Sajood, S., Taban, Q., Mumtaz, P. T., Rather, M. A., Paray, B. A., & Ganie, S. A.
(2020). Honey as component of diet: importance and scope. In Therapeutic Applications
approach to discriminate pure honey from adulterated ones with glucose of Honey and its Phytochemicals (pp) (pp. 215–236). Singapore: Springer (pp).
syrup. From the PCA, 98% of the total variance was verified in the Aliaño-González, M. J., Ferreiro-González, M., Espada-Bellido, E., Palma, M., &
proposed model using wavelengths of 280–300 nm. Validation of the Barbero, G. F. (2019). A screening method based on Visible-NIR spectroscopy for the
identification and quantification of different adulterants in high-quality honey.
method with LDA reached a variance of 99.8%, which demonstrates the Talanta, 203, 235–241. https://doi.org/10.1016/j.talanta.2019.05.067
reliability of this technique for detecting adulteration in honey. The Aslam, R., Sharma, S. R., Kaur, J., Panayampadan, A. S., & Dar, O. I. (2023). A systematic
protocol described involves a simple sample preparation step, without account of food adulteration and recent trends in the non-destructive analysis of food
fraud detection. Journal of Food Measurement & Characterization, 17(3), 3094–3114.
the use of hazardous chemicals, with fast scanning, and simple data https://doi.org/10.1007/s11694-023-01846-3
processing. Therefore, it is recommended that the method be used in the Brazil. 2000. Ministério da Agricultura, Pecuária e Abastecimento. Instrução Normativa
analysis of honey, helping to determine the quality of this bee product. 11, de 20 de outubro de 2000. Regulamento Técnico de identidade e qualidade do
mel. Available at: 〈http://www.cidasc.sc.gov.br/inspecao/files/2012/08/IN-11-de
As it is a fast and less expensive technology for monitoring fraud in
-2000.pdf〉.
honey, the UV–vis technique and chemometrics protocol developed can Carvalho, R. de A., Ribeiro, A. C., Lima, C. M., Mariz, W. P. da S., Silva, L. S., Silva, A. M.
be essential to guarantee the quality of that food by the beekeeping da, Guimarães, D. R., & Trombete, F. M. (2020). Assessment of adulteration and
sector. mycoflora identification of honey samples marketed in the metropolitan region of
Belo Horizonte, Brazil. Research Society and Development, 9(7), Article e440974246.
Finally, the need to create a state or national database to monitor the https://doi.org/10.33448/rsd-v9i7.4246
quality of this product is emphasized. In addition to creating the pro­ Cianciosi, D., Forbes-Hernández, T., Afrin, S., Gasparrini, M., Reboredo-Rodriguez, P.,
posed database, it should also be increasingly improved by continuously Manna, P., Zhang, J., Bravo Lamas, L., Martínez Flórez, S., Agudo Toyos, P.,
Quiles, J., Giampieri, F., & Battino, M. (2018). Phenolic compounds in honey and
adding new UV–vis honey spectra from both pure and adulterated their associated health benefits: A review. Molecules, 23(9), 2322. https://doi.org/
samples, even with new adulterants that not glucose syrup. 10.3390/molecules23092322
Dimakopoulou-Papazoglou, D., Ploskas, N., Serrano, S., Silva, C. S., Valdramidis, V.,
Koutsoumanis, K., & Katsanidis, E. (2023). Application of UV–Vis spectroscopy for
Funding the detection of adulteration in Mediterranean honeys. European Food Research and
Technology, 249(12), 3043–3053. https://doi.org/10.1007/s00217-023-04347-1
To the Coordenação de Aperfeiçoamento de Pessoal de Nível Supe­ Fakhlaei, R., Selamat, J., Khatib, A., Razis, A. F. A., Sukor, R., Ahmad, S., & Babadi, A. A.
(2020). The toxic impact of honey adulteration: A review. Foods, 9(11), 1538.
rior (CAPES) for granting the scholarship to A. Nunes. The research https://doi.org/10.3390/foods9111538
fellowship from CNPq (process nº 304657/2019–0) on behalf of M. Ferreiro-González, M., Espada-Bellido, E., Guillén-Cueto, L., Palma, M., Barroso, C. G., &
Maraschin is acknowledged, as well as, the IC/CNPq-UFSC scholarship Barbero, G. F. (2018). Rapid quantification of honey adulteration by visible-near
infrared spectroscopy combined with chemometrics. Talanta, 188, 288–292. https://
awarded to B.R. Santos and G.Z. Azevedo. Support from the São Paulo
doi.org/10.1016/j.talanta.2018.05.095
Research Foundation (FAPESP – grant 2019/27227-1) and the Foun­ Franz, G. M., Galbiati, C., Araujo, E., & Almeida-Muradian, L. B. de (2018). Antibacterial
dation for Research Support of Santa Catarina (FAPESC – grant activity and physicochemical characteristics of honey from Mato Grosso’s Amazon,
2020TR1452). Cerrado and Pantanal, Brazil. Revista Ibero-Americana Délelő tt Ciências Ambientais, 9
(1), 26–42. https://doi.org/10.6008/spc2179-6858.2018.001.0003
García, N. L. (2018). The current situation on the international honey market. Bee World,
CRediT authorship contribution statement 95(3), 89–94. https://doi.org/10.1080/0005772x.2018.1483814
Gerhardt, N., Birkenmeier, M., Schwolow, S., Rohn, S., & Weller, P. (2018). Volatile-
compound fingerprinting by headspace-gas-chromatography ion-mobility
Lima Giuseppina Pace Pereira: Funding acquisition, Supervision, spectrometry (HS-GC-IMS) as a benchtop alternative to 1H NMR profiling for
Writing – review & editing. Moura Sidnei: Funding acquisition, Su­ assessment of the authenticity of honey. Analytical Chemistry, 90(3), 1777–1785.
pervision, Writing – review & editing. Maraschin Marcelo: Conceptu­ https://doi.org/10.1021/acs.analchem.7b03748
Guellis, C., Valério, D. C., Bessegato, G. G., Boroski, M., Dragunski, J. C., & Lindino, C. A.
alization, Funding acquisition, Project administration, Resources, (2020). Non-targeted method to detect honey adulteration: Combination of
Supervision, Writing – review & editing. Nunes Aline: Conceptualiza­ electrochemical and spectrophotometric responses with principal component
tion, Data curation, Formal analysis, Methodology, Writing – original analysis, 103466 Journal of Food Composition and Analysis, 89, Article 103466.
https://doi.org/10.1016/j.jfca.2020.103466.
draft, Writing – review & editing. Azevedo Gadiel Zilto: Conceptuali­ Haque, F., Bubli, S. Y., & Khan, M. S. (2021). UV–Vis spectroscopy for food analysis. In
zation, Formal analysis, Methodology, Writing – review & editing. Techniques to Measure Food Safety and Quality, 169–193 (pp).
Santos Beatriz Rocha dos: Conceptualization, Formal analysis, Meth­ Huang, F., Song, H., Guo, L., Guang, P., Yang, X., Li, L., … Yang, M. (2020). Detection of
adulteration in Chinese honey using NIR and ATR-FTIR spectral data fusion.
odology, Writing – review & editing.
Spectrochimica Acta. Part A, Molecular and Biomolecular Spectroscopy, 235(118297),
118297. https://doi.org/10.1016/j.saa.2020.118297
Declaration of Competing Interest Jiménez-Carvelo, A. M., González-Casado, A., Bagur-González, M. G., & Cuadros-
Rodríguez, L. (2019). Alternative data mining/machine learning methods for the
analytical evaluation of food quality and authenticity – A review. Food Research
The authors declare that they have no known competing financial International, 122, 25–39. https://doi.org/10.1016/j.foodres.2019.03.063
interests or personal relationships that could have appeared to influence Kharbach, M., Alaoui Mansouri, M., Taabouz, M., & Yu, H. (2023). Current application of
the work reported in this paper. advancing spectroscopy techniques in food analysis: Data handling with
chemometric approaches. Foods (Basel, Switzerland), 12(14), 2753. https://doi.org/
10.3390/foods12142753
Acknowledgments Mitra, P. K., Karmakar, R., Nandi, R., & Gupta, S. (2023). Low-cost rapid workflow for
honey adulteration detection by UV–Vis spectroscopy in combination with factorial
design, response surface methodology and supervised machine learning classifiers,
To the beekeepers who provided the honey samples for analysis and 101327 Bioresource Technology Reports, 21, Article 101327. https://doi.org/
to the extension workers of the Agricultural Research and Rural Exten­ 10.1016/j.biteb.2022.101327.
sion Company of Santa Catarina (EPAGRI, in Portuguese) for carrying Musulin, E., & Basualdo, M. (2012). Improved principal component analysis models for
fault detection using delay adjustment (pp). In PEM Fuel Cells with Bio-Ethanol
out the sample collection. Processor Systems (pp. 431–451). London: Springer (pp).
Nunes, A., Zilto Azevedo, G., Rocha dos Santos, B., Vanz Borges, C., Pace Pereira
Lima, G., Conte Crocoli, L., Moura, S., & Maraschin, M. (2022). Characterization of
Brazilian floral honey produced in the states of Santa Catarina and São Paulo

7
A. Nunes et al. Food and Humanity 2 (2024) 100194

through ultraviolet–visible (UV–vis), near-infrared (NIR), and nuclear magnetic Souza, R. R., Fernandes, D. D., de, S., & Diniz, P. H. G. D. (2021). Honey authentication in
resonance (NMR) spectroscopy, 111913 Food Research International, 162, Article terms of its adulteration with sugar syrups using UV–Vis spectroscopy and one-class
111913. https://doi.org/10.1016/j.foodres.2022.111913. classifiers, 130467 Food Chemistry, 365, Article 130467. https://doi.org/10.1016/j.
Nunes, A., Azevedo, G. Z., Santos, B. R. dos, Liz, M. S. M. de, Schneider, F. S. de S., foodchem.2021.130467.
Rodrigues, E. R. de O., Moura, S., & Maraschin, M. (2023). A guide for quality Thrasyvoulou, A., Tananaki, C., Goras, G., Karazafiris, E., Dimou, M., Liolios, V.,
control of honey: Application of UV–vis scanning spectrophotometry and NIR Kanelis, D., & Gounari, S. (2018). Legislation of honey criteria and standards. Journal
spectroscopy for determination of chemical profiles of floral honey produced in of Apicultural Research, 57(1), 88–96. https://doi.org/10.1080/
southern Brazil. Food and Humanity, 1, 1423–1435. https://doi.org/10.1016/j. 00218839.2017.1411181
foohum.2023.10.010 Tibola, C. S., da Silva, S. A., Dossa, A. A., & Patrício, D. I. (2018). Economically
Razavi, R., & Kenari, R. E. (2023). Ultraviolet–visible spectroscopy combined with motivated food fraud and adulteration in Brazil: Incidents and alternatives to
machine learning as a rapid detection method to the predict adulteration of honey. minimize occurrence. Journal of Food Science, 83(8), 2028–2038. https://doi.org/
Heliyon, 9(10), Article e20973. https://doi.org/10.1016/j.heliyon.2023.e20973 10.1111/1750-3841.14279
Roberts, J., Power, A., Chapman, J., Chandra, S., & Cozzolino, D. (2018). The use of UV- Valinger, D., Longin, L., Grbeš, F., Benković, M., Jurina, T., Gajdoš Kljusurić, J., &
Vis spectroscopy in bioprocess and fermentation monitoring. Fermentation, 4(1), 18. Jurinjak Tušek, A. (2021). Detection of honey adulteration – The potential of UV–VIS
https://doi.org/10.3390/fermentation4010018 and NIR spectroscopy coupled with multivariate analysis, 111316 Food Science and
Se, K. W., Wahab, R. A., Syed Yaacob, S. N., & Ghoshal, S. K. (2019). Detection Technology, 145, Article 111316. https://doi.org/10.1016/j.lwt.2021.111316.
techniques for adulterants in honey: Challenges and recent trends. Journal of Food Yu, J., Wang, H., Zhan, J., & Huang, W. (2018). Review of recent UV–Vis and infrared
Composition and Analysis, 80, 16–32. https://doi.org/10.1016/j.jfca.2019.04.001 spectroscopy researches on wine detection and discrimination. Applied Spectroscopy
Reviews, 53(1), 65–86. https://doi.org/10.1080/05704928.2017.1352511