**1. Introduction**

Black locust (*Robinia pseudoacacia*) is a deciduous tree originating from the Southeastern region of North America. Today, it can be found in all regions of the world with a moderate climate [1]. In Serbia, it is located in Vojvodina where it was used for bonding the "living" sand, as well as in the plains and valleys of northern Serbia. Black locust grows up to 20 m in height and 90 cm in tree diameter. Initially, its plantation was for ornamental purposes, but there are studies that mention the importance of a mixed plantation of this species, because it is a nitrogen fixing species that promotes a growing facilitation effect of other trees [2]. It has a smooth bark, thorny branches, and a rare crochet. The flowers are a cluster with characteristically white color, pleasant and intense aroma, and sweet taste.

Black locust flowers are a source of vitamin C (40 mg 100 g<sup>−</sup>1), resin, tannins [3], essential oil, monoterpenes (cis-β-ocimene ~26.6%, (E)-α-bergamotene ~8.9%, pinens, β-pinens, limonene), diterpene, triterpene, terpene alcohols (linalool ~33.1%), and small amount of poisonous robinin (kaempferol 3,7-di-O-glucoside) that completely disappears after thermal treatment and drying [4–7]. The content of unsaturated fatty acids in the flowers is about seven times higher than the content that of saturated fatty acids [8]. Secundiflorol, mucronulatol, isomucronulatol, and isovestitol were identified in ethanol extracts of the whole plant [9]. The presence of flavonoid glycosides, such as flavonol

3,7-di-O-glycosides, were confirmed in the black locust flower [6]. Stefova et al. [10] reported that the quercetin content is low while the kaempferol content is high in the flower extract. Sarikurkcu et al. [8] extracted the bioactive compounds from black locust flowers by the Soxhlet extractor using ethyl acetate, acetone, and methanol, and prepared the aqueous extract by maceration. The total phenolic content (TPC) and total flavonoid content (TFC) were higher in the acetone and methanol extracts compared to the aqueous extract. In order to investigate the impact of drying methods on the antioxidant properties and TPC, the ethanol extracts were prepared from the flowers previously treated by sun drying, hot-air drying, freeze-drying, and microwave-vacuum drying [11]. The highest TPC was determined in the samples treated using the freeze-drying method.

Medical use of black locust is limited only to the flower, because the other parts, especially bark, are toxic due to the high concentration of robinin [6]. It has long been used as a folk drug for alleviating the colds and coughs, stomach cramps, rheumatic pains, migraine, fever, and skin diseases [11,12]. These flowers reduce the level of cholesterol, and have diuretic and laxative effects. Phenolic compounds from the black locust flower have antioxidant [8,11,13], antimicrobial [13–15], and anticancer properties [10]. The techniques used for the extraction of phenolic compounds from black locust flowers are summarized in Table 1.

**Table 1.** The experimental conditions for extraction of phenolic compounds from black locust flowers previously reported in the literature.


Because of the relatively long extraction time, large solvent consumption, higher cost of equipment, and degradation of compounds, the reflux, maceration, and Soxhlet extractions tend to be replaced by more environment-friendly extraction techniques. Ultrasound-assisted extraction (UAE) has become a good alternative method for extracting phenolic compounds since it can offer a high reproducibility for shorter extraction times and reduced consumption of the solvent [17,18]. The UAE is commonly carried out at lower temperatures so that the thermal degradation of bioactive compounds in the extract can be prevented.

TPC is related to the antioxidants, mainly phenolics in this case, in the extracts. The extracts' phenolic composition is conditioned by liquid–solid ratio, extraction time, extraction temperature, particle size, and any other extraction variables [19]. The eco-friendly and nontoxic organic solvents are recommended by the US Food and Drug Administration for the extraction of bioactive compounds from plant materials [20]. Water and ethanol were proved as the extraction solvents of choice to reach good yields of phenolic compounds from plant materials [21,22] due to their nontoxicity. Having that in mind, the ethanolic solutions were used to extract the phenolic compounds from black locust flowers.

The one-variable-at-a-time (OVAT) approach analyzes the impact of one factor on the defined response, while all other factors are constant. In this way, the grea<sup>t</sup> number of experimental runs is necessary for optimizing the observed process. The interactions between the factors cannot be analyzed using the OVAT approach. The application of OVAT enables to obtain the local optimal extraction conditions, which do not correspond to the real (global) optimal conditions. In order to overcome these problems, the response surface methodology (RSM) is commonly used. This methodology predicts the system response, analyzes the interactions between the factors, defines the relationship between the response and factors, and optimizes the extraction conditions with a limited number of experiments [23,24]. Yang et al. [25] optimized the extraction parameters (extraction time, solvent concentration, and liquid-to-solid ratio) of UAE in terms of common bean (*Phaseolus vulgaris* L.) antioxidants using the two-level factorial design. Izadiyan and Hemmateenejad [26] performed multiresponse optimization of the factors (extraction temperature, solvent concentration, and extraction time) a ffecting the UAE of Iranian *Ocimum basilicum* using a central composite design (CCD). Yin et al. [27] applied the CCD to optimize the factors (extraction temperature, extraction time, and powder dosage) of the UAE of natural anthocyanin from purple sweet potato for silk fabric dyeing. Živkovi´c et al. [28] determined the optimal conditions for the UAE of phenolic compounds from pomegranate peel using CCD, and investigated the relationship between extraction time, solvent concentration, liquid-to-solid ratio, and extraction temperature.

Since the use of RSM has not been reported ye<sup>t</sup> for modeling the UAE of phenolic compounds from black locust flowers, the aim of this study was to generate a polynomial equation that describes the extraction process. The impact of ethanol concentration, extraction temperature, and extraction time were estimated using the CCD. The optimal extraction conditions were optimized using a numerical optimization method. The identification and quantification of phenolic compounds were carried out using high-performance liquid chromatography (HPLC). The TPC and antioxidant activity of extract obtained using UAE were compared to maceration and Soxhlet extraction. The structural changes of plant materials after using these extraction techniques were observed by scanning electron microscopy (SEM).

#### **2. Materials and Methods**
