**2. Materials and methods**

#### **2.1. Enzyme and chemicals**

Immobilized lipase B from *Candida antarctica* (Novozym 435, 7000 PLU/mg: propyl laurate units synthesized per gram of catalyst) was purchased from Novozymes AS, Bagsvaerd, Denmark. *tert*‐amyl alcohol was supplied by Merck (France). Rutin was furnished by Sigma‐Aldrich (France). Hexadecanedioic acid (Cathay Biotechnology, China) was used as acyl donor. The molecular sieves 4Å, used to dry solvents and reactions media, were provided by Acros organis (France).

#### **2.2. Rutin ester synthesis**

are described in the literature. One of them is the acylation reaction, with the aim to enhance both the solubility in hydrophobic media and the biological activities of flavonoids [13]. This reaction can be carried out via enzymatic or chemical route. Enzymatic reactions are preferred to chemical ones due to their regio‐selectivity toward polyhydroxylated compounds like rutin [14, 15]. Rutin is a flavonol glycoside and one of the most studied flavonoids in the literature. Several papers dealt with improvement of the solubility and the biological activities of rutin by enzymatic acyla‐ tion under a wide range of operating conditions. These studies showed that the performance (conversion yield, regioselectivity, etc.) of this reaction was affected by several factors (reaction media, solubility and nature of the substrates, operating conditions, enzyme concentration, etc.). Lue et al. [16] reported that the use of ionic liquids as acylation medium enhances the solubility of substrates and gives high conversion yields of rutin. Zheng et al. [17] observed better rutin conversion rates when acylation was conducted under ultrasound radiations. This improve‐ ment was attributed by these authors to the enhancement of lipase activity. Water content of the medium has a strong effect on acylation reaction. It can affect both the solubility of the substrates and the activity of the biocatalyst. In the case of rutin, Ardhaoui et al. [18] reported that the high‐

est rutin conversion yield was reached with water content less than 200 ppm (76%).

conversion rates, and the regioselectivity of the reaction were quantified and discussed.

Immobilized lipase B from *Candida antarctica* (Novozym 435, 7000 PLU/mg: propyl laurate units synthesized per gram of catalyst) was purchased from Novozymes AS, Bagsvaerd, Denmark. *tert*‐amyl alcohol was supplied by Merck (France). Rutin was furnished by Sigma‐Aldrich (France). Hexadecanedioic acid (Cathay Biotechnology, China) was used as acyl donor. The molecular sieves 4Å, used to dry solvents and reactions media, were provided by Acros organis (France).

**2. Materials and methods**

114 Flavonoids - From Biosynthesis to Human Health

**2.1. Enzyme and chemicals**

The most used enzyme for rutin acylation reactions is lipase B from *Candida antarctica*. This enzyme has shown high performances in terms of conversion yield and enantioselectivity [19]. The chain length or the nature of the acyl donor can also affect the conversion yield of rutin and the regioselectivity of the reaction [16, 18, 20, 21]. Conversion yields arise from 48 to 87% with the length of carbon chain [18]. Dicarboxylic acids, another class of fatty acids, can be used for the acylation of flavonoids. These compounds have two advantages compared to monocarboxylic acids. In one hand, they exhibit more flexibility due to their second carboxyl group. In the second hand, according to their structure, they have bacteriostatic and bactericidal properties against a variety of aerobic and anaerobic bacteria [22]. However, few data are available concerning their use as acyl donors with flavonoids. Only Theodosiou et al. [23] and Ardhaoui et al. [18] reported the enzymatic modification of flavonoids with diacids without any optimization of operating conditions of this reaction. The aim of the present work was to study the enzymatic synthesis of rutin hexadecanedioate by Novozym 435 in organic medium. The effect of several factors such as drying techniques, temperature, rutin concentration, molar ratio rutin/hexadecanedioic acid, con‐ centration of biocatalyst, and its reuse were investigated. The behavior of initial rate, productivity, The enzymatic syntheses of rutin hexadecanedioate were carried out in a stirring batch reactor (250 ml) from "Pilotes Systèmes" (France) or Wheaton® reactors (USA). Agitation speed was varied from 300 to 500 rpm. Rutin (65–196 mM) was dissolved in 250 ml of dried *tert*‐amyl alco‐ hol at different temperatures (60, 80, and 90°C). The hexadecanedioic acid concentration was adjusted to obtain a di‐acid/rutin molar ratio in the range of 0.05–20 in the solution. Esterification reactions were started by the addition of Novozym 435 (10, 30, and 50 g/L). The reaction was stopped after 50–72 h by removing the biocatalyst. Blank samples, containing all components except the enzyme preparation, were carried out in tandem with the enzymatic trials.

Different drying techniques were used: (i) drying with molecular sieves added to the bulk medium or (ii) introduced in an external loop. In this last case, solvent and water were evapo‐ rated under vacuum, crossing in a fist part molecular sieves as vapor phase then as liquid phase (**Figure 1a**). Another configuration of external loop consists in the water removal only in liquid phase (**Figure 1b**).

Water content was 800–1000, 400–550, and 200–300 ppm, respectively with molecular sieves added to the bulk medium, the use of drying in vapor and liquid phases and drying only through liquid phase.

#### **2.3. Analytical procedure**

#### *2.3.1. Karl Fisher analysis*

The water content of the reaction medium was determined by a coulometric Karl Fisher appa‐ ratus (KF 737II coulometer) Metrohm (France). The reagent was Hydranal‐Coulomat AG‐H (Riedel‐de‐Haën, France).

#### *2.3.2. High‐performance liquid chromatography analysis*

The substrate and product concentrations were determined by high‐pressure liquid chroma‐ tography in external calibration. Analysis were carried out at 55°C in a system (Alliance 2690 Waters) composed of a column (Symmetry® C18, 4.6 × 250 mm, 5 µm, Waters, France), a UV detector (250 and 350 nm, Waters 2487, France) and a ELSD (Evaporative Light Scattering detector, Altech 2000, France). The various compounds were separated using water (0.1% acetic acid)/methanol (0.1% acetic acid) solutions: 0 min (70/30), 5 min (0/100), 10 min (0/100), 12 min (70/30), 15 min (70/30).

#### *2.3.3. Purification and determination of the chemical structure of rutin esters*

Rutin esters were purified by liquid–liquid extraction. The residual flavonoid was removed at 60°C under agitation during 45 min with a water/heptane solution (2/3, v/v), while, the flavonoid esters were separated from the acyl donor by using acetonitrile (50 mL) at 40°C dur‐ ing 20 min of agitation. The flavonoid esters solution was concentrated by solvent evaporation under reduced pressure for injection in a preparative HPLC (Waters, France). A column RP18

**Figure 1.** Drying techniques used to remove water from the reaction media. (a) Drying the media by using molecular sieves in the outer loop of the reactor in liquid phase. (b) Drying the media by using molecular sieves in the outer loop of the reactor in vapor and liquid phase.

(30 × 100 mm, 5 µm, Waters XTerra®, France) and a UV detector (350 nm, Waters 2487, France) were used for separate and analyze esters. A gradient of water and acetonitrile with 0.1% ace‐ tic acid at a flow rate of 18 mL/min was applied: 0 min (70/30), 10 min (20/80), 12 min (20/80), 13 min (70/30), 15 min (70/30). The medium was diluted 2.5 times in the starting phase (water with 0.1% acetic acid/acetonitrile with 0.1% acetic acid, 70/30) and 850 μl of this solution were injected for each batch. Eighteen batches were produced. Rutin hexadecanedioate and dirutin hexadecanedioate are obtained with a purity ≥95%. This purification method was adapted from Ardhaoui et al. [18].

The chemical structures of the purified dirutin hexadecanedioate and rutin hexadecanedio‐ ate were determined by 1H NMR and 13C NMR in DMSO‐d6 using a Brücker AM 400 at 400 MHz and at 100 MHz, respectively.

#### **2.4. Determination of conversion rate, initial rate, productivity, and selectivity**

#### *2.4.1. Conversion rate*

The conversion rate of rutin and acid was calculated from concentrations given by HPLC analysis.

$$\text{Conversion (\%)} \newline = \text{([S]i - [S]f) \* 100/[S]i} \newline \tag{1}$$

[S]i: initial substrate concentration (mmol/L)

[S]f: final substrate concentration (mmol/L)

#### *2.4.2. Initial rate of monorutin hexadecanedioate formation*

The initial rate was calculated during the first three hours of the synthesis reaction of rutin hexadecanedioate by taking the slope of the kinetic linear portion.

#### *2.4.3. Productivity*

The productivity is given by the following expression:

 Productivity (g/(L/h)) = Mass of ester formed divided by the working volume of the reactor and by the duration of the reaction (2)

#### *2.4.4. Selectivity*

(30 × 100 mm, 5 µm, Waters XTerra®, France) and a UV detector (350 nm, Waters 2487, France) were used for separate and analyze esters. A gradient of water and acetonitrile with 0.1% ace‐ tic acid at a flow rate of 18 mL/min was applied: 0 min (70/30), 10 min (20/80), 12 min (20/80), 13 min (70/30), 15 min (70/30). The medium was diluted 2.5 times in the starting phase (water with 0.1% acetic acid/acetonitrile with 0.1% acetic acid, 70/30) and 850 μl of this solution were injected for each batch. Eighteen batches were produced. Rutin hexadecanedioate and dirutin hexadecanedioate are obtained with a purity ≥95%. This purification method was adapted

**Figure 1.** Drying techniques used to remove water from the reaction media. (a) Drying the media by using molecular sieves in the outer loop of the reactor in liquid phase. (b) Drying the media by using molecular sieves in the outer loop

from Ardhaoui et al. [18].

of the reactor in vapor and liquid phase.

116 Flavonoids - From Biosynthesis to Human Health

The selectivity is given by the following expression:

```
 Selectivity (%) = [monorutin] t *100/([monorutin] t + [dirutin] t) (3)
```
Where, [monorutin] t and [dirutin] t are molar concentrations at t (time) of mono and dirutin esters.
