*3.3. Plant Samples – Germination*

The seeds of *Trifolium pratense* and *Medicago sativa* were purchased from Agrosem, Targu Mures, Romania. Approximately 50 g of seed of each plant species were washed in a plastic container with 20 ◦C sterile distilled water for 30 min and then transferred to a growth chamber (automat sprout germinator, Biovita model GE-1, Cluj-Napoca, Romania) controlled at 25 ◦C and 80% humidity in the dark condition. Seed hydration was automatically controlled as follows. The seeds were soaked for 7 h. Then they were irrigated with water every 5 min for 7 h, and then they were irrigated for 1 min every 4 h. For both plant species, we obtained sprouts at 24 h, 48 h, 72 h, 96 h, and 120 h. The entire sprouts were dried for 4.5 h at 40 ◦ C in a fruit dryer (Zilan model- ZLN-9645l). Average humidity loss was 90% for alfalfa and 89% for red clover. After drying, the twins were kept in closed containers, away from light. Before analysis, aliquots of 5 g of each sample were ground.

#### *3.4. Extraction*

Three extraction procedures were compared in the present work, maceration, ultrasound assisted extraction (UAE), and microwave assisted extraction (MAE). The extracts were prepared in the same condition for both plant species.

A tincture was prepared according to EU Pharmacopeia using 70% ethanol (1:10 g DW/ g). The solvent was initially heated at 60 ◦C. The maceration continued for 10 days in dark conditions.

An ultrasonic assisted extraction (UAE) method optimised by response surface methodology (RSM) [45] was adapted for the present study. The extraction conditions were a ratio of liquid to solid of 1:10 g DW/g 70% ethanol, 60 ◦C, 60 min, at 60 kHz.

A microwave assisted extraction (MAE) procedure was adapted after Zhang, 2008 [26]. An amount of 0.5 g of each sprout sample were extracted with 15 mL and 50% ethanol (1:25 g DW/g) at 50 ◦C (10 min. gradient with 5 min. maintaining), and microwave power at 300 W.

For all procedures, the extracts were filtered through Whatman No. 2 filter paper and a 0.20 nm Millipore MF syringe filter. Dilution 1:3 with water:methanol (80:20) before instrumental analysis was completed.

#### *3.5. Instrumentation*

#### 3.5.1. LC Parameters

A Thermo Scientific Dionex Ultimate 3000 Series RS pump coupled with a Thermo Scientific Dionex Ultimate 3000 Series TCC-3000RS column compartments and a Thermo Fisher Scientific Ultimate 3000 Series WPS-3000RS autosampler controlled by Chromeleon 7.2 Software (Thermo Fisher Scientific, Waltham, MA and Dionex Softron GMbH Part of Thermo Fisher Scientific, Germany) were used for analysis.

The application of a 35-min gradient over an ultra-performance Accucore U-HPLC Column C18 (150 × 2.1 mm, 2.6 μm), (Thermo Scientific) was applied. The column temperature was set at 40 ◦C. The mobile phase consisted of: eluent A, ultrapure water containing 500 μL/L formic acid (pH 2.5), and eluent B, methanol. The step gradient was as follows: 0–1 min 100% A, 1–10 min linear increase to 30% B, 10–26 linear increased to 100% B and held for 4.0 min, 30–32.5 decreasing to 0% B. The initial conditions were obtained again at the 35th min with an equilibration time of 2.5 min. The run was performed at 0.4 mL/min for a total of 35 min.

#### 3.5.2. MS Parameters and Data Processing

A HESI (Heated Electrospray) ion source was used for the ionization. The HESI parameters were optimized as follows. Nitrogen as sheath and auxiliary gas flow rate was set at 8 and respectively 6 units. The source heater temperature was set at 300 °C. The capillary temperature was set at 300 ◦C. The aux gas heater temperature was set at 300 ◦C. The electrospray voltage was 2800 V. The S lens RF level was 50.

Detection of the compounds was performed using an Q-Exactive mass spectrometer. Full scan data in negative mode was acquired at a resolving power of 70,000 FWHM at *m*/*z* 200. For the compounds of interest, a scan range of *m*/*z* 100–1000 Da was chosen. The automatic gain control (AGC) was set at 3e6 and the injection time was set to 200 ms. The scan rate was set at 2 scan/sec. External calibration was performed by calibration solution in a positive and a negative mode.

A total of six scan events were combined including one full scan event with mentioned parameters and five MS-MS events. In the MS2 scan events, the precursor ion ranges were *m*/*z* 95–205, 195–305, 295–405, 395–505, and 500–10005, which were consecutively selected, fragmented in an higher-energy collisional dissociation cell HCD, and measured in five separate Orbitrap scans at a resolving power of 35,000 FWHM. The fragmentation events were performed at 30, 60, and 80 NCE (normalised collision energy). The C-trap parameters for all scan events were the following: Automatic Gain Control (AGC) target 1e6 and the injection time of 100 ms.

Data were evaluated by the Quan/Qual Browser Xcalibur 2.3 (Thermo Fisher). The mass tolerance window was set to 5 ppm for the two analysis modes. For the MS/MS analysis, detection of at least two fragment ions with the appropriate ion-ratio was performed by comparing the reference standards.

For those compounds without available references, the most reasonable molecular formula with a lower mass error was sought in the chemical Chemspider database (www.chemspider.com). Considering that the flavones, isoflavones, and phenolic acids had the same skeleton, the fragment ions from MS-MS analysis were used to further confirm the chemical structure with the aid of NORMAN MassBank (https: //massbank.eu/MassBank/), mzCloudeTM Advanced Mass Spectral Database (https://www.mzcloud.org/), and PubChem (https://pubchem.ncbi.nlm.nih.gov/). ACDLabs MS Fragmenter 2019.2.1 software was used to generate a fragmentation pattern of the identified compounds for a comparison analysis.
