Class-model is constructed based on data from samples of authentic products only. Class-models are applied for adulteration detection, and their ability to do that is assessed by testing the model against randomly selected authentic samples mixed with adulterant(s). However, due to natural within-class variance, the choice of samples used for blend preparation influences the distribution of adulterated samples in the parameters space. Therefore, it is difficult to assess the actual ability of the class-model to detect adulterated samples. The present study addressed this issue upon the example of honeybush and rooibos teas adulteration detection based on their elemental profiles. An approach based on simulated adulterated samples using pure authentic samples is presented to achieve the most representative set of adulterated samples. Evaluation of the class-models’ ability to detect adulteration was more reliable when based on simulated data than for experimental data, e.g., class-model constructed for rooibos recognised all actual rooibos samples experimentally mixed with ≥20% honeybush as adulterated but in case of the simulated adulterated samples, more than 90% of rooibos samples adulterated with 20% honeybush were misclassified as pure rooibos samples. The detection of adulterated rooibos samples was improved by 78% using the Partial Least Squares (PLS) regression model.

Ultra-high temperature oxidation (UHTO) was explored as an alternative to the standard oxidation of honeybush tea, aiming to shorten the processing time and enhance the fruity and sweet-associated sensory characteristics. Different UHTO temperature × time regimes were applied to Cyclopia subternata, C. genistoides, C. intermedia and C. longifolia. Descriptive sensory analysis showed that UHTO treatment, regardless of species, increased the intensities of fruity and sweet-associated aromas while decreasing the intensities of floral aromas. The optimal UHTO regime for each species was subsequently applied to a larger number of plant material batches (n = 10 per species) to introduce a wider range of intra-species variation and confirm broad trends for the sensory profile, as well as for the colour and phenolic composition, of the infusions over species. The effect of treatment conditions on the retention of the phenolic compounds varied, depending on the compound and species. The substantial decrease in the mangiferin and isomangiferin concentrations of C. genistoides infusions, which was not observed for the other species, was notable. The study highlights the potential of UHTO to reduce oxidation time while enhancing the fruity and sweet-associated aroma attributes of honeybush tea, changes that may be beneficial in attracting new consumers and expanding the honeybush tea market.

Class- and discriminant modelling are two types of classification tools applied to construct models used for prediction of belongingness of samples to the classes studied. Although class- and discriminant methods have similar goals, the areas of their applicability are different. A class-model is constructed individually for each of the classes studied, based on the similarities among samples from the same class. A classical discriminant model is constructed based on differences among classes studied, and a new sample is always assigned to one of these classes. This characteristic differentiates the discriminant approach from class-modelling, where a new sample can be assigned to none of the classes studied. Due to this property, the class-modelling approach is widely applied for authentication purposes. Yet, if more classes similar to one another are authenticated, individual class-models constructed for these classes can lead to poor classification results. In such cases, the discriminant model can provide a better classification outcome since it takes advantage of differences between the classes. However, the discriminant approach alone is inappropriate for authentication, thus a possible solution is to use a method that benefits from both class- and discriminant modelling. In this study, several methods that combine class-modelling with a discriminant approach were tested, i.e., two-step approach, two soft discriminant methods based on PLS-DA (Partial Least Squares Discriminant Analysis), and ROC (Receiver Operating Characteristics) curve-based SIMCA (Soft Independent Modelling of Class Analogy). The methods were compared with respect to their pros, cons and the scope of their applicability for the example of authentication of three Cyclopia species. They are used for the production of honeybush tea (protected in the European Union as a Geographical Indication (GI)). Moreover, several different authentication scenarios were considered to test the methods thoroughly. It was revealed that the two-step approach, soft discriminant method by Calvini et al. and ROC curve-based SIMCA led to the most efficient models.

The elemental analysis of tea is essential since it is a commonly consumed beverage around the world. In this paper, non-destructive energy dispersive X-ray fluorescence spectrometry (EDXRF) using a standardless method based on fundamental parameters was applied in honeybush and rooibos analysis. Tea samples were measured directly in the form of loose powders. The observed element concentrations (mg kg(-1)) were as follows: Ca (1260-3990), Cl (914-7710), K (2270-7190), Mg (374-2510), S (474-1320), P (183 1210), Si (303 1580), Al (92-765), Fe (67-514), Mn (26-163), Cr (1.8-20), Cu (2.2-11.5), Ni (1.1-5), Rb (0.8-8), Sr (5.4-24.5), Ti (4.1-39), and Zn (4.2-18.6). What is more, the determined concentrations (mg kg(-1)) of selected elements in rooibos and honeybush, are statistically different: Cl (2200-7710 rooibos and 913.7-2160 honeybush); Mg (1232-3465 rooibos and 535.9-1030 honeybush); P (388.0-1566 rooibos and 244.0-403.6 honeybush); Br (12.1-52.8 rooibos and 1.6-5.3 honeybush). Multivariate analysis of variance (ANOVA) was successfully applied to the data, showing statistical significance of the concentration differences for all the elements in both types of tea. The developed method provided good precision (RSD < 6%) with an accuracy of more than 90 %, and LOD similar to 0.5 mg kg(-1-) for trace elements. The method was validated using suitable certified reference materials of tea (CRM).

Honeybush tea infusions, especially those prepared from Cyclopia genistoides, can be unacceptably bitter, given the association of this herbal tea with sweet taste. Infusions prepared from both 'fermented' (high-temperature oxidised) Cyclopia longifolia and C. genistoides contain high levels of the bitter xanthone, mangiferin, however, C. longifolia is generally less bitter than C. genistoides. The effect of phenolic changes during fermentation on the bitterness of infusions was determined for both species. Bitterness was reduced by 34-68% for C. genistoides and 55-86% for C. longifolia. Fermentation affected the phenolic composition of the two species differently, notably the ratio of mangiferin to isomangiferin remained higher for C. genistoides. Bitterness prediction models for two data sets of infusions (126 fermented and green plant material; 122 fermented plant material) including variable selection based on phenolic content gave good performance (RMSECV < 5). Nine compounds, depending on the data set, were determined to be important for the models, including xanthones, benzophenones, flavones and flavanones. Mangiferin, two tetrahydroxyxanthone-di-O,C-hexose isomers, 3-beta-D glucopyranosyliriflophenone, vicenin-2 and scolymoside were common to both models. Future application of the model would be to screen large numbers of genotypes for the breeding of selections with low bitterness potential based on phenolic composition.

Honeybush is an indigenous herbal tea highly valued for its aroma, flavour and medicinal properties. It is protected as Geographical Indication (GI) since it is produced from a number of Cyclopia species that are endemic to South Africa. Most commonly used for honeybush tea production are C. intermedia, C. subternata and C. genistoides, differing slightly, but distinctly in flavour. Demand for species-specific honeybush tea instead of mixtures have increased, meriting a strategy for authentication of C. intermedia, C. subternata and C. genistoides. Samples of these three species were analysed, using hyperspectral imaging (HSI) in the near-infrared spectral range. The data were pre-processed and used for class-modelling, a general approach well suited for authentication purposes. Unfortunately, since the HSI data of Cyclopia species studied are very similar, the classification results obtained with individual class-models are unsatisfactory, e.g., class-models constructed for C. genistoides and C. subternata yielded correct classification rate (CCR) values of 76.4 and 83.1%, respectively. On the other hand, discriminant modelling, which is another type of classification technique, led to good classification outcomes (CCR 98.9%). However, the classical discriminant model cannot be applied for authentication purposes since it always assigns a new sample to one of the classes studied, even if in reality, it belongs to none of them. Counterfeits or non-representative samples would be incorrectly assigned by the discriminant model to one of the authentic classes. Therefore, in this study, a two-step authentication of overlapping classes is proposed, which combines the advantages of class-modelling and discriminant methods. When applied to the authentication of Cyclopia species studied, the two-step approach yielded a CCR of 97.4%, which is a significant improvement compared to results obtained with the individual class-models. The proposed approach is general and can be applied when classes studied are very similar, and individual class-models lead to unsatisfactory results.

In this paper, the determination of ultratrace heavy metal ions was developed by combining a preconcentration method using graphene oxide/carbon nanotubes membranes (GO/CNTs) and total-reflection X-ray fluorescence spectrometry (TXRF). Due to the excellent adsorptive properties of the GO, the foregoing membranes are suitable for effective simultaneous sorption of Co(II), Ni(II), Cu(II), Zn(II), Cd(II), and Pb(II) from aqueous solutions. In this method, the aqueous solution is passed through the GO/CNTs membrane. The analytes are eluted from the GO and afterward transferred onto a siliconized quartz reflector for further TXRF analysis using W and Mo target X-ray tubes. The maximum recoveries for all the elements were obtained at pH 5; thus it was chosen for all further experiments. The face centered central composite design was performed to study the influence of the flow-rate and volume of the solution on the recovery of the determined metal ions. Recovery values higher than 96% for all studied metals allow performing efficient preconcentration with an enrichment factor of 133, achieving the limits of detection (LODs) in the range of 0.08-0.21 ng mL(-1) for W target X-ray tube with a measurement time of 2000 s, and much lower LODs for Mo target X-ray tube: 0.001-0.002 ng mL(-1) with the exception for Cd (0.11 ng mL(-1)) with a very short measurement time of 600 s. Certified reference materials of spring water and seawater were examined to verify the reliability of the method. The evaluated procedure does not require toxic reagents or organic solvents, thus minimizes the portion of the sample for TXRF measurement, and stands in good accordance with green analytical chemistry basic principles.

Class-modelling methods are applied to construct a mathematical model based on the similarities among samples belonging to the same category, i.e., the target class. This model is used to study the belongingness of a new sample to the class for which the model was constructed. If the sample is recognised as not belonging to the target class, it is considered as an outlier. Therefore, the class-modelling techniques are widely used for food or drug authentication and confirmation of the product origin, in order to detect samples of poor quality or potential counterfeits. Structure of the target class data might suggest which of the available class-modelling approaches is well suited for model construction. Data structure can generally be described as normal, or heterogeneous. Normal structure exhibits Gaussian distribution, whereas heterogeneous data deviates from normal distribution, e.g., the objects can have multimodal distributions, create subgroups, and form complex shapes in the feature space. Class-modelling of the normally structured data can directly be performed on an original dataset, whereas heterogeneous datasets are usually subjected to kernel transformation prior to modelling. In this study, several datasets of various structures are analysed with different class-modelling methods to test their scope of applicability and the pros and cons from the practical point of view. It is revealed that in most cases, the Support Vector Domain Description (SVDD) leads to the best classification results. However, it is also discussed and illustrated that SVDD applied to a multimodal data can lead to sub-optimal model. In that case, the Potential Functions Method (PFM) is recommended for model construction. The PFM model is based on local densities of samples in the multivariate space and in that way, it tends to be more accurate than SVDD model. Additionally, some practical remarks are given on optimisation of the Gaussian kernel width.

Spectrometric and analytical techniques in general collect multivariate signals from chemical or biological materials by means of a specific measurement instrumentation, usually in order to characterize or classify them through the estimation of one of several compounds of interest. However, measurement conditions might induce various additive (baseline) or multiplicative effects on the collected signals, which may jeopardize the accuracy and generalizability of estimation models. A common way of dealing with such issues is signal normalization and in particular, when the baseline is constant, the standard normal variate (SNV) transform. Despite its efficiency, SNV has important drawbacks, in terms of physical interpretation and robustness of estimation models, because all the variables are equally considered, independently on what their actual relationship with the response(s) of interest is. In the present study, a novel algorithm is proposed, named variable sorting for normalization (VSN). This algorithm automatically produces, for a given set of multivariate signals, a weighting function favoring signal variables that are only impacted by additive and multiplicative effects, and not by the response(s) of interest. When introduced in SNV preprocessing, this weighting function significantly improves signal shape and model interpretation. Moreover, VSN can be successfully used not only for constant but also with more complex baselines, such as polynomial ones. Together with the description of the theory behind VSN, its application on various synthetic multivariate data, as well as on real SWIR spectral data, is presented and discussed.

The Class Modelling (CM) approaches like Soft Independent Modelling of Class Analogy (SIMCA) aim at developing a mathematical model for determination of belongingness of new samples to the studied classes. The main feature of CM is that for each target class an individual model is constructed. CM is widely exploited, e.g., in the food and drug quality testing and authenticity or origin verification. It is well known that the most critical stage in construction of a class model is optimization of its parameters. There exist two basic strategies for optimization of class model, i.e., the "compliant" strategy where the target and nontarget class samples are required in the model optimization process, and the "rigorous" strategy where only the target class samples are used. Since the nontarget class samples are usually available, the compliant scenario is more often explored. In the present study, four different resampling methods for optimization of the SIMCA model (applied in both, a compliant and a rigorous fashion) are thoroughly compared. Each method is tested in combination with two distinct decision threshold estimation criteria: i) an a priori fixing it based on a desired statistical significance level and ii) optimizing it through appropriate data-driven procedures. For the sake of a comprehensive assessment of the studied strategies, several real-world datasets are exploited and final results are post-processed by means of ANalysis Of VAriance (ANOVA). The study reveals that both, a compliant approach with an optimized decision threshold and a rigorous approach with a fixed decision threshold can yield satisfactory classification outcomes, no matter which resampling technique is used. Finally, it is shown how unrepresentativeness of the nontarget classes can lead to the biased classification models when a compliant optimization is carried out. Therefore, a rigorous optimization can be considered as a safer option for the SIMCA model parameter tuning.

In the present article, the theory of ANOVA-target projection (ANOVA-TP), a method aimed at analyzing multivariate data coming from designed experiments is presented. ANOVA-TP, similarly to ANOVA-principal component analysis (ANOVA-PCA), ANOVA-simultaneous component analysis (ASCA) and other related methods, starts from a multiple ANOVA decomposition of the multivariate experimental data matrix. Partial least squares analysis is then used to relate suitably built effect submatrices to the design terms, and postprocessing by target projection constitutes the basis for interpretation. The proposed approach is illustrated by different examples, showing the versatility of the method in dealing with crossed-factors full factorial designs, repeated-measurement designs and longitudinal studies.

Mounting evidence of the ability of aspalathin to target underlying metabolic dysfunction relevant to the development or progression of obesity and type 2 diabetes created a market for green rooibos extract as a functional food ingredient. Aspalathin is the obvious choice as a chemical marker for extract standardisation and quality control, however, often the concentration of a single constituent of a complex mixture such as a plant extract is not directly related to its bio-capacity, i.e. the level of in vitro bioactivity effected in a cell system at a fixed concentration. Three solvents (hot water and two EtOH-water mixtures), previously shown to produce bioactive green rooibos extracts, were selected for extraction of different batches of rooibos plant material (n = 10). Bio-capacity of the extracts, tested at 10 μg ml-1, was evaluated in terms of glucose uptake by C2C12 and C3A cells and lipid accumulation in 3T3-L1 cells. The different solvents and inter-batch plant variation delivered extracts ranging in aspalathin content from 54.1 to 213.8 g kg-1. The extracts were further characterised in terms of other major flavonoids (n = 10) and an enolic phenylpyruvic acid glucoside, using HPLC-DAD. The 80% EtOH-water extracts, with the highest mean aspalathin content (170.9 g kg-1), had the highest mean bio-capacity in the respective assays. Despite this, no significant (P ≥ 0.05) correlation existed between aspalathin content and bio-capacity, while the orientin, isoorientin and vitexin content correlated moderately (r ≥ 0.487; P < 0.05) with increased glucose uptake by C2C12 cells. Various multivariate analysis methods were then applied with Evolution Program-Partial Least Squares (EP-PLS) resulting in models with the best predictive power. These EP-PLS models, based on all quantified compounds, predicted the bio-capacity of the extracts for the respective cell types with RMSECV values ≤ 11.5, confirming that a complement of compounds, and not aspalathin content alone, is needed to predict the in vitro bio-capacity of green rooibos extracts. Additionally, the composition of hot water infusions of different production batches of green rooibos (n = 29) at 'cup-of-tea' equivalence was determined to relate dietary supplementation with the extract to intake in the form of herbal tea.

Rooibos (Aspalathus linearis) and honeybush (Cyclopia species) are popular indigenous herbal teas originating from South Africa. Both are enjoyed for their taste and aroma and more importantly, valued for their medicinal properties such as antioxidant, anti-diabetic, anti-inflammatory or immunomodulatory activity. In the European Union, rooibos and honeybush are protected as Geographical Indications. The Geographical Indication refers to products with unique characteristics that are related to their geographical origin. The authentication of products labelled as GIs is regarded as an issue of food quality and safety. Routine quality control procedures of GIs products prevent their fraud and counterfeiting on the market. However, techniques to determine adulteration or mislabelling of rooibos and honeybush do not exist yet. Therefore, in this study, the authentication of rooibos and honeybush based on their elemental composition was investigated. The methodology presented in this study combines energy-dispersive X-ray fluorescence spectrometry (EDXRF) for elemental analysis and a one-class classification approach. Elemental composition of plant material highly depends on soil on which the plant has been grown, but also on some other factors such as, e.g., atmospheric pollution, or plant metabolism. Determination of the elemental composition of the samples by EDXRF is non-destructive and does not require any complex sample preparation. One-class classification methods are well suited for authentication and origin verification problems. Based on the EDXRF data of samples from studied teas, individual class models were constructed for rooibos and honeybush. Several linear and nonlinear classification techniques were tested in order to find the model that handles the authentication task the best. For honeybush tea, the best classification results were obtained with the use of a nonlinear method based on Potential Functions. For rooibos tea, the highest classification outcomes were obtained by nonlinear One Class Partial Least Squares (OC-PLS) and the machine learning technique, Support Vector Domain Description (SVDD). The methodology implemented in the present study has the potential to be successfully applied for routine authentication of honeybush and rooibos teas.

Increasing demand for honeybush tea (Cyclopia spp.) and the need for industry expansion have created interest in non-utilised species such as Cyclopia pubescens Eckl. & Zeyh. Very limited information is available on the phenolic composition of this species. A reversed phase core-shell biphenyl column was used to develop and validate a quantitative HPLC diode-array detection method for separation of the major phenolic compounds in C. pubescens. Eight phenolic compounds were identified and a further six tentatively identified by comparison of retention time, UV-vis and high resolution mass spectrometric characteristics with those of authentic reference standards and literature, respectively. Genotypic variation in the phenolic composition of C. pubescens was determined by analysing the leaves and stems of seedlings (n = 17) in a field gene bank. The xanthone and benzophenone present in the highest levels in the leaves were mangiferin and 3-beta-D-glucopyranosyl-4-beta-D-glucopyranosyloxyiriflophenone, respectively. The leaves contained higher quantities of all compounds, except hesperidin, the major compound in the stems, and a second hesperetin glycoside. Statistical analysis included hierarchical clustering to determine the degree of dissimilarity between genotypes, which provided valuable information for future breeding programs. Comparison of the dendrograms for leaves and stems indicated different clustering patterns for genotypes.

Instrumental signals of samples cannot be compared and/or analysed directly if their concentrations are unknown. Differences in overall concentration need to be removed at the data normalization step. The choice of normalization method has a profound effect on the final results of data analysis, and especially on biomarker identification. One of the possible approaches to deal with the 'size effect' is to work with size-irrelevant (log) ratios instead of the original variables. In the presented study, the performance of log-ratio methods, namely pairwise log-ratio (plr) and centered log-ratio (clr), is discussed for real and simulated data sets with different characteristics. It was found that the clr method can lead to distribution of local differences along an entire signal and as such, it should be avoided in all studies aiming to identify biomarkers.

Phenolic compounds of Aspalathus linearis (rooibos) are susceptible to oxidation during “fermentation”, a process characterized by the formation of a red-brown leaf color. The role of enzymes in this process is not yet understood. An experiment with dried green rooibos plant material pre-treated at 170 °C for 30 min to denature and “inactivate” endogenous enzymes was conducted to confirm the role of oxidative enzymes. The phenolic composition of “enzyme inactivated” plant material was not significantly (p ≥ .05) affected by simulated fermentation, compared to control samples, as determined using piece-wise multivariate analysis of variance for successive time intervals. This proves that rooibos enzymes participate in the oxidation of phenolic compounds during fermentation of the plant material. A kinetic modeling approach was subsequently used to establish reaction kinetic parameters for selected rooibos phenolic compounds. The degradation of aspalathin and nothofagin and formation of eriodictyol glucosides during simulated fermentation at four temperatures from 37 to 50 °C were best described by the fractional conversion model based on first-order kinetics (r2 > 0.98), which allows for non-zero equilibrium concentrations. The extent of degradation for other compounds was too low to enable kinetic modeling. Reaction rates for the degradation/formation of phenolic compounds during fermentation followed the Arrhenius law. Less phenolic degradation (higher equilibrium concentration), but a higher reaction rate constant, was observed at higher temperatures, which could possibly be attributed to inactivation of enzymes.