In recent years, there has been a marked increase in the development of new foods that are organoleptically appealing and offer health benefits to consumers. Many chronic diseases are associated with an imbalance in oxidative stress [1–3]. Besides, the incorporation of antioxidants into mass-market products such as biscuits could be an interesting alternative to increase their intake and protect against cell damage caused by exposure to free radicals. Biscuits are ready-to-eat foods that are widely accepted across all age groups. Also, biscuits have a long shelf life due to their low moisture content [4].

Plants contain bioactive compounds, including phenolics, which are the main contributors to their antioxidant activity (AA). The main use of antioxidants in foods is the prevention of the off-flavours due to lipid oxidation [5]. Moreover, the addition of a wide variety of plant materials in biscuits in order to make them healthier has been explored [4]. Rosemary, thyme, cloves, cinnamon, and oregano are examples of natural aromatic plants and spices that have been used in bakery products because they provide some essential nutrients and give excellent flavour and aroma [6]. Among the millions of known plant species worldwide, only a small number of them are widely used as food sources, while more than a hundred wild edible plant species remain neglected [7]. Some edible leaves and flowers have been used as part of a meal or in infusions such as tea [8, 9].

Different parts of plants can be used to extract antioxidants, and their antioxidant power depends on both the quality of the original source and the technologies used to extract them [10]. Traditional extraction methods include maceration, percolation, and the Soxhlet method with different solvents [11, 12], while other authors reported the use of infusion as an extraction method [9, 13].

In a previous study, the total phenolic content (TPC) of 0.5% aqueous extracts, prepared from dried and ground samples of flowers and leaves collected in green places of La Plata, was analyzed [14]. Their values ranged from 6 to 50 mg gallic acid equivalents (GAE)/g for flower extracts, and from 1.8 to 15.9 mg GAE/g for leaf extracts.

Pelargonium is one of the seven genera of the Geraniaceae family, and includes almost 280 species [15]. Pelargonium × hortorum is the geranium species most widely cultivated. It is a perennial herbaceous plant with flowers almost year-round [16].

Hibiscus rosasinensis Linn (Malvaceae) is an ornamental and popular medicinal plant native to China, but is widely disseminated in tropical and sub-tropical regions. The bioactive attributes of extracts from Hibiscus rosa-sinensis are numerous [17]. Callistemon citrinus is an evergreen shrub originating from Australia. It is widely cultivated as an ornamental garden plant. Its lemon-scented flower spikes are present almost all year round on the plant, but mainly in November and December in the southern hemisphere [18].

Among the analyzed leaf extracts, the highest values were recorded in Aloysia citrodora (lemon verbena) and Bauhinia forficata (cow’s foot) extracts; both extracts showed values of 15 mg GAE/g dry sample with no significant difference between them, but higher than the other leaf extracts. The genus Bauhinia (Leguminosae) has been used in traditional medicines around the world. Seven species have been recorded as having pharmacological properties, with the most widely consumed and studied being B. forficata [19]. The most widely consumed and studied of these is B. forficata. Native to South America, B. forficata contains high levels of phenolic compounds. These are especially found in its leaves. The main compound is kaempferitrin [20]. Aloysia citrodora is a well-known aromatic species that is rich in antioxidant compounds, which originated in South America but, nowadays, grows across a wide geographical area [21]. Lemon verbena leaves have been included in the Argentine Pharmacopoeia as an herbal medicine since the 5th edition (1966), and in the European Pharmacopoeia since the 6th edition (2008) [22]. The main component of the non-volatile fraction is verbascoside, which is responsible for its various bioactivities [22].

In the present work, red geranium flower and lemon verbena leaf extracts were selected to be incorporated into a biscuit formulation, considering that these plants are more readily available throughout the year.

Therefore, the aim of this work was to evaluate the texture, colour, and antioxidant capacity of biscuits made with red geranium flower and lemon verbena leaf extracts. In this context, the study focused on assessing the feasibility of incorporating these extracts into a bakery model system and their contribution to antioxidant-related parameters [TPC and ferric reducing antioxidant power (FRAP)]. These results may serve as a starting point for future studies evaluating potential physiological effects associated with functional foods.

The wide range of colours found in flowers and leaves is generated by the quantity and quality of various pigments, such as carotenoids and anthocyanins [26]. Anthocyanins (polar) give the characteristic red, purple, or blue colours. Carotenoids (nonpolar) give colours ranging from red to yellow [27], and chlorophyll (nonpolar) gives green tones.

LE exhibited a relatively light greenish-yellow colour, GE displayed a darker, more reddish tone. These values reflect the distinct pigmentation profiles of the two plant materials, likely attributable to their differing compositions of flavonoids and other phenolic compounds. Also, phytochemicals such as phenolic acids, flavonoids, carotenoids, and tocopherols have antioxidant properties. The markedly higher TPC observed in GE is consistent with its high antioxidant capacity [14].

Organic solvents or mixtures of organic solvents are used in many studies to analyze the antioxidant content. However, their toxicity makes them unsuitable for use in the pharmaceutical and food industries [28]. For this reason, in the present study, warm water was used to extract bioactive compounds from samples. Besides, plant extracts made with water are more nutritionally relevant than extracts prepared with organic solvents [29]. Considering this extraction approach, the effect of natural extracts incorporation on bakery products could be further evaluated.

Incorporation of extracts and baking conditions were significant factors in the colour of the samples (P < 0.05). The reddish colour of the GE persisted in both 10% and 30% dough and biscuits. In samples made with LE, a brownish tinge was only noticeable in dough with 30% LE, but not in the other samples (Figure 2). According to Mokrzycki and Tatol [30], an observer may be able to discriminate the colour difference between two objects if it is greater than 5. Taking this into account, the control and 10% LE samples would be indistinguishable (Table 1). However, the rest of the samples differ noticeably in colour from the control sample, as indicated by a ΔE greater than 10. Browning index (BI) represents the purity of the brown colour and is reported as an important parameter in processes that involve enzymatic or non-enzymatic browning [31]. The control biscuits and LE ground biscuits baked at 160°C showed significantly higher BI values than the ground biscuits baked at 130°C (P < 0.05, Table 1). Among the samples with GE, significant differences were found only in the biscuits with 30% GE, and values of BI were higher in the biscuits baked at 130°C than those baked at 160°C (P < 0.05, Table 1).

Minor differences in total colour difference (ΔE) were observed in samples made with 10% LE. On the other hand, all samples with 30% LE showed significant differences from the control sample (Table 1).

Moreover, dough samples with GE had a notably lower pH compared to the other samples (P < 0.05). Despite the equal liquid-to-dry ratio in all doughs, substantial variations in moisture content were observed. The dough with 10% LE had the lowest moisture content, while the dough with 10% GE had the highest. Those differences may be due to different compositions of soluble substances extracted from both plant materials. There were no significant differences in cohesiveness or elasticity of the dough samples (P > 0.05). But the control dough and the dough with 10% LE had the highest values for firmness, consistency, and adhesiveness, while the lowest values were found in both samples with GE (P < 0.05) (Table 2). The low pH of GE could explain its softer dough texture. In agreement, other authors have suggested that acidic conditions enhance the electrostatic repulsion of proteins, which could reduce their interaction with water and starch granules, resulting in a softer dough [32].

Baking conditions strongly influenced biscuit moisture, a_w, and dimensional characteristics. The values of weight loss were higher at a lower temperature. According to Zhang et al. [32], the increase in weight loss indicates a weaker and more open gluten structure, whereas decreases could be due to the formation of a stronger and more compact gluten network. At low temperatures, the dough rises more in height than in width, resulting in a lower spread factor. Additionally, in control and samples with LE, the biscuits had a higher volume and a_w, and retained more moisture when baked at a lower temperature (130°C, 30 min) compared to those baked at a higher temperature (160°C for 22 min).

However, the opposite was observed in samples with GE: greater moisture and a_w values were found for biscuits baked at a high temperature. This may be because shorter cooking times result in more water being retained due to the higher soluble solids content (2.2°Brix). The control biscuits and the biscuits with 30% LE had a crisper texture at 160°C, with higher Y values and lower ε values being observed than at 130°C. The texture of the biscuits with 10% LE was similar to the control samples at 130°C, but at 160°C, the control samples had a higher Y (a crisper texture) than 10% LE. Biscuits with 10% GE had lower ε values (less rubbery texture) than the control samples at 130°C and lower Y values than the control sample at 160°C.

In addition to these textural differences, the antioxidant properties of the samples were also evaluated. The FRAP method has been used to assess the reducing antioxidants in food matrices, because it is fast and relatively simple to perform [33]. The highest AP values (5.9, 5.5, and 5.6 μmol Fe⁺²/g for dough, biscuits baked at 130°C, and 160°C) were observed in 30% Pelargonium × hortorum red flower extract samples (30% GE), and these values differ significantly from those of the remaining samples (P < 0.05). The lowest AP values were observed in control samples (2.1, 1.8, and 2.6 μmol Fe⁺²/g for dough, biscuits baked at 130 and 160°C, respectively). Moreover, 10% Pelargonium × hortorum red flower extract samples (10% GE) showed similar values to those with 30% Aloysia citrodora leaf extract samples (30% LE). There were no significant differences between the 10% LE samples (dough, biscuits baked at 130°C or baked at 160°C) and the control samples (P > 0.05).

The results shown in Figure 3B indicate that dough samples exhibited higher TPC than the corresponding baked biscuit samples. This difference is likely attributable to the presence of phenolic compounds in the raw ingredients of the dough. However, during the heating process, some of these compounds may be lost due to their thermolabile nature, particularly when high baking temperatures are applied (160°C). In agreement with this observation, no significant differences in TPC were detected among the formulations when biscuits were prepared at 160°C (P > 0.05). In contrast, at 130°C, phenolic compounds from the GE remained relatively stable when a high extract proportion (30%) was used. This finding suggests that phenolics from GE exhibit greater thermal stability than those derived from LE. In this regard, baking at 130°C would be a suitable alternative for the elaboration of GE cookies, as it would allow for adequate AA and preserve phenolic compounds to a greater extent than baking at 160°C. These results are in accordance with Blanch and del Castillo [34] and Žilić et al. [35], who found higher TPC values when baking at low temperatures than at high temperatures.

Because of the extraction procedure used in the present study, the antioxidants obtained were likely phenolic compounds and flavonoids present in Pelargonium × hortorum red flower and Aloysia citrodora leaf extracts. According to a revision made by Celi et al. [36] about different Pelargonium species, flavonoids were the most common compounds (with quercetin being one of the principal components), followed by phenolic acids and coumarins. Abdel Bar et al. [37] evaluated different biomolecules present in Pelargonium × hortorum, finding concentrations between 125–214 µg/mL of gallic acid. Pelargonium × hortorum also contains several anthocyanins, including peonidin-3,5-O-diglucoside [3,750 µg Ideain equivalent (IE)/g], malvidin-3,5-O-diglucoside (1,803 µg IE/g), pelargonidin glucoside derivative (620 µg IE/g), and cyanidin-3,5-diglucoside (473 µg IE/g), among others, with total anthocyanin content around 7,600 µg IE/g) [38]. Other authors also found rosmarinic acid [217 μg/g fresh weight (f.w.)] and caffeic acid (133 μg/g f.w.) on flowers of Pelargonium × hortorum [39]. On the other hand, the main compounds found in lemon verbena are flavones, iridoids, and phenylpropanoids [40]. Bilia et al. [41] reported that the leaves contain phenylpropanoid concentrations ranging from 20 to 150 mg/g dry weight, of which approximately 97% corresponds to verbascoside. Verbascoside is a hydrophilic compound and has antioxidant, anti-inflammatory, and antineoplastic properties, as well as wound-healing and neuroprotective effects [42]. More recently, Athanasiadis et al. [21] reported that the TPC of aqueous extracts of lemon verbena leaves ranged from 90 to 113 mg GAE/g, whereas ethanol extracts contained between 40 and 100 mg GAE/g, depending on the extraction conditions and techniques used. In addition, phenolic acids (vanillic acid, p-coumaric acid, and ferulic acid) ranged from 0.5 to 8.94 mg/g, while flavonoids (rutin, quercetin 3-O-galactoside, verbascoside, narirutin, and kaempferol 3-O-glucoside) ranged from 22.77 to 163.38 mg/g.