MAE was applied at different temperatures (80–120 °C), times (5–25 min), and solvent/solid ratios (300–700 mL/g), and resulting in 17 combinations of these variables for leaves and each tobacco waste type (scrap, dust, and midrib). The objective of the present study was to determine the potential of MAE from tobacco waste compared to MAE from leaves, and also to determine the influence of process parameters, and optimize the nicotine extraction process.
2.1. Bioactive Compounds Composition of Tobacco Leaves and Waste Extracts Obtained with MAE
MAE is well known for accelerating the extraction process and improving the extraction yield [
12,
14]. In this study, extraction yield was in the range from 19.40 to 69.78% for tobacco leaves, from 35.98 to 69.08% for scrap, from 29.42 to 58.23% for dust, and from 37.84 to 57.96% for midrib (
Table 1,
Table 2,
Table 3 and
Table 4). For all waste samples, the highest extraction yield was obtained at temperature of 120 °C. During MAE microwave power was dependent on temperature. From
Supplementary Materials Figure S1, it can be seen that when the temperature was set at 80 °C, the microwave power reach its maximum (700 W) first 2 min, then decreased on the value of 100 W during the extraction process while on 120 °C microwave power reach maximum (700 W) at first 4 min, then decreased on the value of 200 W during the extraction process. Thus, the higher extraction yield showed to be directly related to the higher microwave power applied.
In comparison with our previously published results, for the subcritical water extraction [
7] and two-stage extraction process (supercritical and subcritical extraction) [
25] the obtained extraction yields were slightly lower. Furthermore, in comparison with results for ultrasound-assisted extraction obtained results were slightly higher [
26]. Since, ultrasound-assisted extraction was performed on lower temperatures (up to 70 °C), higher applied temperatures could be responsible for increase of extraction yield. On the other hand, obtained results were significantly higher than those obtained with supercritical extraction [
25]. However, subcritical water extraction was conducted at significantly higher temperatures (150–250 °C) while supercritical CO
2 extraction was highly selective to non-polar compounds which are naturally present in very low concentrations in tobacco materials. Moreover, obtained yields were significantly higher in comparison with the conventional extraction process performed by Docheva et al. [
27] where for Virginia blend type lower extraction yield (34%) was reported.
The obtained results for the extracted amount of bioactive compounds from tobacco leaves and waste are summarized in
Table 1,
Table 2,
Table 3 and
Table 4. In this study nicotine was dominant compound with concentrations in range 1.512–5.480, 1.886–3.709, 2.628–4.840, and 0.867–1.783% for leaves, scrap, dust, and midrib extracts, respectively. Nicotine well known as dominant bioactive compound of tobacco, naturally present in a concentration from 0.3 to 3% referred to dry plant material [
28].
Moreover, the concentration of nicotine is highly dependent on tobacco type, variety, growing and environmental conditions [
6,
29]. In comparison to other extraction techniques, nicotine concentrations were similar to those obtained with subcritical water extraction. However, concentrations for leaves and dust were slightly higher than those obtained with subcritical water extraction. Yet, those differences should be carefully interpreted because nicotine tends to degrade under the conditions of subcritical water extraction [
7].
A recent study by Popova et al. [
30] also found that nicotine was a major compound in tobacco resinoid (extract obtained with polar solvent), where nicotine concentrations were over 3%. Additionally, a spectrophotometric assay for determination of antiradical activity was performed. Applied process parameters did not significantly influenced the antiradical activity of MAE extracts, while obtained values were similar to those obtained for tobacco waste with other extraction techniques [
7,
8,
26].
Besides nicotine (alkaloid), five phenolic compounds were detected in leaf extracts, namely, chlorogenic acid (CA), neochlorogenic acid (NCA), cryptochlorogenic acid (CCA), rutin, and nicotiflorin (
Table 1,
Table 2,
Table 3 and
Table 4). Polyphenols profile of leaves extract were in conformity with the findings from other studies [
2,
29] where other extraction techniques were applied. In comparison with leaf extracts, waste extracts were less rich in bioactive compound content implying possible losing of the bioactive compounds during tobacco processing which was also confirmed in our previous studies [
7,
26]. CA (5-
O-caffeoylquinic acid), NCA (3-
O-caffeoylquinic acid), and CCA (4-
O-caffeoylquinic acid) are members of the phenolic acid group named chlorogenic acids and represent the main phenolic acids in tobacco [
31]. CA was present in concentrations in range 0.145–1.501, 0.180–0.525, 0.237–1.135, 0.014–0.265% in leaves, scrap, dust and midrib extracts, respectively. NCA concentrations varied in range 0.054–0.409, 0.048–0.256, 0.119–0.344, 0.055–0.323% in leaves, scrap, dust and midrib extracts, respectively. CCA acid concentration were in range 0.017–0.432, 0.030–0.136, 0.012–0.249, and up to 0.444% in leaves, scrap, dust and midrib extracts, respectively. Chlorogenic acid contents were in good agreement with a study from Wang et al. [
2] where sonification was used as an extraction technique. Moreover, two dominant flavonoids were identified, rutin (quercetin 3-rutinoside) and nicotiflorin (kaempferol-3-
O-rutinoside). Both are rutinosides group members. In this study, rutin concentrations were in range 0.111–0.635, 0.130–0.631, 0.126–0.602, 0.079–0.166% in leaves, scrap, dust and midrib extracts, respectively. Nicotiflorin concentrations were up to 0.089% for leaves, and from 0.040 to 0.083% in scrap, from 0.039 to 0.099% in dust and up to 0.075% in midrib. The content of individual phenolic compounds was similar to those obtained in the study [
29] for Plovdiv variety.
2.3. Statistical Analysis
For better understanding of differences between waste and leaves, as well as influence of process condition on MAE, statistical analysis was provided (
Supplementary Materials). As
Table S1 shows, for leaves, higher yield correlates significantly with nicotiflorin and total phenols. For scrap, higher yield means less rutin, and the same negative correlations can be seen for dust. The impact of extraction yield is more evident on dust samples, where higher yield provides significantly lower amounts of nicotiflorin, nicotine, rutin and lower radical scavenging activity (DPPH). This can be again explained through correlation between temperature and microwave power, explained previously. With the increase of temperature above 100 °C, higher temperatures caused an extraction yield enhancement for all samples. On the other hand, increase of temperature means waste of energy and lower content of bioactive compounds.
Sample comparison was done between tobacco leaves (chosen as a model sample), as starting material with all other samples (tobacco waste materials). Differences in yield selected components concentrations and antiradical activity (DPPH) are shown in
Table S2. As expected, the biggest difference was found between leaves and midrib. This is probably due to physiological and chemical differences between leaf lamina and midrib previously explained by Banožić et al. [
1]. The midrib is the central vein of a leaf and has more malonic and oxalic acids, total ash, calcium, potassium and chlorine while leaf lamina has more reducing sugars, total alkaloids, nitric and citric acid, and phenols. Interestingly, yield does not significantly differ between leaves and scrap which can again be explained with materials similarity since scrap is more similar to leaf than other types of waste.
Finally, the comparison between applied process parameters was done based on the extraction method. The comparator method used to compare all others was 100 °C, 15 min and 20 mL/g of solvent/solid ratio, and central point of Box–Behnken design (BBD). Method #1 was 80 °C, 5 min and 20 mL/g of solvent/solid ratio, method #2 80 °C, 25 min and 20 mL/g of solvent/solid ratio, method #3 120 °C, 5 min and 20 mL/g of solvent/solid ratio, and method #4 120 °C, 25 min, 20 mL/g of solvent/solid ratio. The only difference was found between Comparator and Method #4 where Comparator gave higher yield (p = 0.004).
2.4. Predictive Modeling and Optimization of MAE of Bioactive Compounds from Tobacco Waste
In order to obtain the regression models, the experimental data (responses) were analyzed and fitted to various models (quadratic and cubic). Obtained models represent an empirical relationship between dependent variables (responses) and independent variables (process condition). Models are expressed by a second-order polynomial equation with interaction terms, which serves for the extraction efficiency prediction:
where X
1, X
2, and X
3 are the coded variables for temperature, time, and solvent/solid ratio, respectively.
The fitting Equations (1)–(8), which describe the dependence of extraction yield and nicotine content on the independent variables (temperature, time and solvent/solid ratio) were used for the individuation of the optimal extraction operating conditions. The optimization was carried out in two aspects; to obtain a maximum extraction yield with minimum nicotine content and secondly, to obtain a maximum extraction yield with maximum nicotine content. Optimization results are showed in
Table 5. Theoretical results were experimentally confirmed with an average relative deviation of up to 5%. This reveals that proposed models have high predictive ability for the optimization of the MAE process. In comparison with leaves, it can be seen that the obtained nicotine values in waste are significantly lower. Even the waste is directly delivered from leaves, it is subjected to various other processes after separation of leaf lamina [
1,
35]. Moreover, from
Table 5, it can be seen that the higher difference in obtained nicotine values are in midrib extracts, where, also, a higher treatment time as optimum was provided. Those results clearly demonstrate that every type of tobacco waste should be consider separately, according to its unique properties and chemical composition of each type of tobacco waste.