INTRODUCTION:

Essential oils can be extracted by hydrodistillation conventional methods or by hydrodistillation assisted by different other techniques (microwaves, ultrasound ...), to improve the quantity and quality of the extracted oil. Other more innovative techniques such as supercritical CO₂ extraction are also used, but a question arises: extracts obtained by these methods can still be called essential oils?

Extraction by hydrodistillation, with its various variants, remains the most widely used method, especially on a domestic and semi-industrial scale. There are generally 3 techniques for extracting essential oils from aromatic plants with water (hydro-distillation): (i) if the plant material is completely bathed in the distillation water, it is distillation with water, often called hydrodistillation Sensus strictus (HD); (ii) if the water vapor produced under pressure outside the still, or in situ with a separation grid between the plant material and the distillation water, passes through the plant material, it is vapor distillation or vapo-distillation (VD ) and (ii) if the plant material is partially bathed in the distillation water, with a coexistence of water in the liquid and vapor states in the still, it is the vapor-hydrodistillation (VHD) (Milojevi et al., 2013).

A real demand exists for optimized hydrodistllation technology on this scale in emerging countries and in particular in rural areas, as part of the fight against poverty.

So, the extraction by hydrodistillation of two species of lemongrass was modeled to support the project for the essential oil production from cymbopogon  species in the Central African region (Citratus, flexuosus, Baou et al., 2020).

Cymbopogon citratus, which is the traditional lemongrass used as infusion drink is one of the characteristic plants of home gardens. The offshoots of these gardens were used to set up the experimental plantation. Cymbopogon flexuosus offshoots, acclimatized in Touloungou in the province of Tandjilé in Chad came from the Congo Brazzaville. With its high biomass production, it can be an alternative to the production of lemongrass essence, like in Congo. There is already some modeling works on the extraction of "lemongrass" in the literature:  in Brazil (Desai et al., 2014), in Vietnam (Dao et al. 2020.), in India (Koul et al., 2004) and in Nigeria (Amenaghawon et al., 2014). It is not possible to compare experimental data nor to transpose the results obtained in such different contexts on 3 continents.

This study concerns the modeling of the extraction kinetics, as basis for optimizing yields and ensuring quality and stability of essential oils from two species of lemongrass: Cymbopogon flexuosus and Cymbopogon citratus.

MATERIALS AND METHODS:

1. Touloungou experimental site:

The study was carried out in Touloungou the province of Tandjilé (located 394km south of N’Djaména, on the banks of the Logone river) The soil is predominantly hydromorph, silty and clayey-silty-sandy. Soil pH varies with the area and its physical environment. The department of Tandjilé-Est is located in the Sudanian zone. The climate is subtropical with two main seasons in the year: a dry season (from November to April) and a rainy season (from May to October). It is among the hottest locality in the country with an average temperature of 31°C. The rain intensity depends on climatic conditions and geographic formation.

The Cymbopogon flexuosus species was established on an experimental site on the banks of the Logone river in Touloungou, in September 2019 (Figure 1a). The Cymbopogon flexuosus offshoots came from the experimental field of Nkama (“Plateau des Cataractes”, Louingui district in Congo Brazzaville) in March 2018. The species was acclimatized on an experimental field in the National Institute of Agronomic Sciences and Agri-Food Technologies of Lai (INSATAL), before being re-planted on the experimental site of Touloungou. The aerial part was harvested on December 22, 2021 and then dried for 5 days in the shade.

The Cymbopogon citratus species was established at the same time on the same site from offshoots acquired locally from farmers (Figure 1c). It was harvested at the same time and was also dried for 5 days in the shade before distillation.

Extraction of essential oils:

The pilot scale extraction was carried out in a 200 L local distiller in which the plant material was bathed in water. The essential oil hydrodistillation was carried out for seven hours (7h). The distillate was collected in a 500mL separating funnel from which the essential oil was separated by gravity. The essential oil was collected, and dried over magnesium sulfate. The essential oil yield in % is expressed as a volume/mass ratio (V_EO (mL in)/m (g of dried vegetable plant))

Y(%)= [V_EO (mL)/m_sample(g)]100

Extraction kinetics of the essential oil:

For kinetic studies, three distillation operations were followed for each species to evaluate the quantity of extracted essential oil, every 30 minutes from 0 to 120 minutes and every hour from 120 to 420 minutes.

Kinetic modeling of the extraction:

Interpretation of the results of the extraction of essential oil was first carried out with the formalism of first order  chemical kinetics.

With this formalism, the variation of the concentration C was proportional to extraction time t:

(dC/dt) = kt, with

·       the kinetic constant k expressed in t^-1;

·       the half-life time corresponding to the half-way point of the process which is given by:

t_1/2= (ln2) / k = 0.693/k;

the rate of progress of the process noted: (C₀ – C)/C₀, with C₀ the initial concentration (t = 0)

Table 1: Data of the hydrodistillation kinetics of C. flexuosus and C. citratus leaves

*Cymbopogon flexuosus*
t(min)	30	60	90	120	180	240	300	360	420
1/t	0,033	0,017	0,011	0,008	0,006	0,004	0,003	0,0027	0,0023
V_HE [mL)	12,5	33	44,5	62,5	89	102	115,5	122,5	128,5
Y_t (%)	0,083	0,22	0,30	0,45	0,60	0,68	0,77	0,82	0,86
y = Y_t/Y_∞,	0,096	0,255	0,348	0,523	0,697	0,790	0,895	0,953	1
1/Y_t	12,048	4,545	3,333	2,325	1,666	1,470	1,298	1,219	1,162
t/ Y_t	361,44	272,72	300	266,66	300	352,94	389,61	439,02	488,37
ln (1/ (1-y))	0,1007	0,294	0,427	0,740	1,193	1,560	2,253	3,057	-
Cymbopogon citratus
t (min)	30	60	90	120	180	240	300	360	420
1/t	0,033	0,017	0,011	0,008	0,006	0,004	0,003	0,0027	0,0023
V_HE [mL)	25	61,5	87	114	135	150,5	180	190,5	194,5
Y_t (%)	0,167	0,41	0,58	0,76	0,90	1,03	1,20	1,27	1,29
y = Y_t/Y_∞,	0,129	0,317	0,449	0,589	0,697	0,798	0,930	0,984	1
1/Y_t	5,988	2,439	1,724	1,315	1,111	0,970	0,833	0,787	0,775
t/ Y_t	179,64	146,34	155,17	157,89	200,00	233,00	250,00	283,46	325,58
ln (1/ (1-y))	0,138	0,381	0,595	0,889	1,193	1,599	2,659	4,135	-

The transposition of the first-order chemical kinetics to essential oil extraction was carried out by Mejri et al. (2014), Amenaghawon et al. (2014), Desai et al., 2014: (i) by assimilating the essential oil content (V in mL) of essential oil extracted from 100 g of plant material) to the concentration C at time t and C₀ at initial time t=0. (ii) by defining, V₀, as the quantity of essential oil contained in 100 g of plant material at the initial time (t = 0); V, the amount of essential oil extracted from 100 g of plant material at time t; (V₀-V) then becomes the amount of essential oil remaining in 100 g of plant material at time t. By setting the rate of evolution of the extraction y = (V₀-V)/V₀, which, in formal kinetics, corresponds to the state of progress of the chemical reaction, after integration, one obtains, the equation:

ln (1/(1-y)) = kt

When the extraction follows these kinetics, its representative curve is a straight line passing through the origin and with a slope k, the kinetic constant.

A review of the literature carried out by Méziane et al. (2019) indicates that more than 80% of kinetic studies on essential oil extraction lead to pseudo-first order kinetics.

The most commonly used model fitted by the first-order kinetics (Sovova and Aleksovski, 2006) assums that during the extraction, a fraction f of cells are broken during the operations prior to extraction and a (1-f) fraction remains intact in a two-step extraction: (i) a rapid washing step which corresponds to the extraction of free essential oil from broken cells and (ii) a second much slower step, depending on the intra-particulate diffusion of the essential oil before the extraction into the cells which have remained intact. This translates mathematically to:

q/q_∞ = 1 - f exp(-k₁t) - (1-f) exp (-k₂t)

This model was verified by Molojevic who introduced approximations leading to the following two borderline cases (Milojevic et al., 2008):

(i) instant washing followed by slow diffusion

(k₁   ∞)

q/q∞ = 1- (1-f) exp(-k₂t)

(ii) absence of the washing step (f = 0)

q/q∞ = 1 - exp(-k₂t)

This last assumption leads to a kinetics of pseudo-first order which admits as linear form:

ln (1/(1-y)) = kt

with y = q/q∞

Peleg's model (1988) was also used to formalize the extraction of a metabolite from a plant matrix. It is formally a kinetics of pseudo-second order :

dCt/dt = k (C_S-Ct)²

k = kinetic constant of order 2 (L g^-1 min^-1)

C_S = extraction capacity (g L^-1); it corresponds to the saturation concentration (t_∞): C_∞

C_t= concentration of the oil in the solution at time t (g L^-1),

t: time (min)

After integration, we get:

C_t = (C_S²kt)/(1 + C_Skt)

which is reorganized into:

t/C_t= 1 / kC_S² + t/C_S

This expression is identical to that of the linearized form of the empirical Peleg model proposed to explain the asymptotic behavior of the evolution of several natural phenomena including the sorption or desorption of certain constituents by plant matrices.

Peleg postulates an evolution of the phenomenon which follows a type law:

C_t = C₀ ± t/(k₁ + K₂t)

with: ±: sorption, adsorption (+) and desorption, (-); C_t: EO concentration or mass extracted at time t (m_t); C_∞: concentration or mass of EO extracted at t_∞ (m_∞); C₀ = 0: the mass of essential oil extracted at t = 0; k₁: kinetic extraction constant of first order, constant extraction capacity K₂ linked to the equilibrium at the end of the process (Shafaeï et al., 2016). If we set: y = (C_t - C₀)/C_∞ –C₀) we deduce y = C_t/C_∞ = m_t/m_∞ and finally we can write the starting equation in its following linearized form:

t/C_t = k₁ + K₂t

We finally define:

k₁ = 1/kC_S² and K₂ = 1/C_S

Experimental data show that 1/C_t = f(1/t) gives a straight line, which by extrapolation to t = ∞ (1/t = 0) leads with more precision C_∞.

The Michaelis-Menton model was validated by the same equation:

Y_t = Y_maxt/(K_m + t) and 1/Y_t = (K_m/Y_max) (1/t) + (1/Y_max)

with Y_t: yield at time t ; Y_max, the yield at time t_∞and Y_max/K_m: slope of the line (Mejri et al., 2014).

All these models were tested for essential oil extraction from lemongrass acclimatized in Chad.

RESULTS AND DISCUSSION:

Characterization of the extraction curves:

Figure 2 shows the essential oil extraction curves of C. flexuosus (CF) and C. citratus (CC).

Figure 2: Essential oil extraction curves of C. flexuosus (CF) and C. citratus(CC)

The Cymbopogon flexuosus essential oil yields 0.86% from the dried plant material and that of Cymbopogon citratus 1.29% in the same experimental conditions. These curves recall the metabolite extraction curves from plant matrices, in particular : essential oils (Dao et al., 2020; Chun et al., 2014), polyphenols (Bucic-Kojic et al., 2007), moisture (Peleg, 1988). One notes the two steps of the process characterized by different extraction rates (fast and slow) with an asymptotic or hyperbolic end depending on the explanatory model used. The slope break was around 200 min. The very slow ending of the process can be seen for a sufficient long period of extraction time. Unfortunately, the prolongation of the extraction time, which consumes energy, does not bring a significant gain in terms of the yield of essential oil and therefore has no practical interest, hence the interest of a good evaluation of the end of extraction.

Modeling of the extraction:

The extraction kinetics were followed for 420 minutes, the results obtained are shown in Table 1, as well as all the data necessary for the tests of the various models studied.

First order model test:

With Y_∞ = 0.86% (in first approximation), the test of the first order kinetic model leads to  straight lines (figure 2).

Y_t= 0.0087 t - 0.2903 (R² = 0.9816) for C flexuosus;

Y_t = 0.0109 t - 0.4294 (R² = 0.922) for C citratus.

leading respectively to the following pseudo-first order extraction kinetic constants:

k_flexuosus = 0.0087 min^-1 and k_citratus = 0.0109 min^-1

Figure 3: Test curves for the extraction of essential oils from C flexuosus and C citratus (pseudo- first order kinetics)

These values should be compared with those met in the literature which were done under varied operating conditions for the other cymbopogons. Amenaghawon et al. (2014) obtains k = 0.045 min^-1 (R² =), for the steam distillation of 100 g lemongrass (Cymbopogon ssp.) That they consider to be the pseudo-first  order kinetic constant corresponding to step vaporization of essential oils in the plant matrix during the extraction process; similarly Tang et al., (2017) fitted this model with a coefficient of determination R² = 0.9426 and a kinetic constant k = 0.024 min^-1 for the steam distillation of 9.5 kg of lemongrass, corresponding to the vaporization of the essential oil in the plant matrix; Desai et al. obtain for Cymbopogon flexuosus k = 0.025 and 0.199 min^-1. With cymbopogons acclimatized in Congo-Brazzaville and in by hydrodistillation, Baou (2021) obtains k = 0.0206   min^-1 for C. flexuosus flexuosus and k = 0.035 min^-1 for C. nardus. Table 2 summarizes the values of the constants for the extraction of essential oils from cymbopogons reported in the literature as well as the conditions of their production.

These rate constants have the same order of magnitude whatever the species and the operating conditions (k = 0.013-0,199 min^-1) with very good coefficients of determination (R² = 0.94-0.99). The values obtained for the extraction on an artisanal pilot scale in hydrodistillation of Cymbopogon flexuosus and Cymbopogon citratus acclimatized in Chad (k = 0.0084 min^-1 and k = 0.0109 min^-1), were also in the same order of magnitude. They are nevertheless a little lower.

Peleg model test:

The experimental data which also fitted Peleg's model (Figure 3) lead to the following results: kinetic constant k₁= 217.94 % min^-1 and 108.94 % min^-1 and extraction capacity K₂= 0.6022%^-1 and 0.4971%^-1with very good values of the coefficient of determination (R² = 0.9885 and 0.9384).

Figure 4: Validation of Peleg's model for hydrodistillation of essential oils of C. flexuosus and C. citratus

The hydrodistillation rate constants k = K₂/k₁ = 0.04563 min^-1 and 0.0028 min^-1. These values is approximately three times higher than that obtained with the Molojevic model (k = 0.0084 min^-1).

Table 2: Pseudo first order rate constants of the extraction essential oils of the few species of cymbopogons in different extraction conditions.

Species	k(min^-1)*	R²	Conditions	References
Lemongrass (Cymbopogon ssp.)	na	-	Steam distillation (100 g)	Amenaghawon et al. (2014)
lemongrass,	0.024	0,9426	Steam distillation (9.5 kg)	Tang et al.,(2017)
Cymbopogon flexuosus	0.199 0.025	-	Hydrodistillation Hydrodistillation	Desai and Parikh, 2014 Desai and Parikh, 2015
Cymbopogon winterianus	0.0470	0.99	Steam distillation	Lutfi et al. 2016
Cymbopogon citratus	0.0129	0.97	Steam distillation	Koul et al. 2004
Cymbopogon schoenanthus	0.0263	0.95	NA	Bellik et al., 2019
Cymbopogon flexuosus	0.0206	0.974	Hydrodistillation	Baou, 2021
Cymbopogon nardus	0.035	0.976	Hydrodistillation	Baou, 2021
Cymbopogon citratus	0.0739-0.0901		Hydrodistillation	Milojevic et al., 2018 Silou et al., 2004

* Pseudo first order rate constants; na : non available

Michaelis-Menton (Monod) model:

Figure 4 gives the validation straight lines of the Michaelis-Menton model, leading to following equations:

Figure 5: Validation straight lines of the Michaelis-Menton model for essential oil hydro-distillation of C. flexuosus and C. citratus

1/Y_t = 237 (1/t) + 0.5275 (R² = 0.9862)

 for C. flexuosus;

1/Y_t = 113.12 (1/t) + 0.4807 (R² = 0.9952)

for C. citratus.

Y_t= Y_maxt/(K_m+ t) and 1/Y_t = (K_m/Y_max) (1/t) + (1/Ymax)

The maximum extraction yields Y_max = 1.90% for C. flexuosus and 2.1% for C. citratus which were twice higher than experimental ones (0.89% and 1.26%) and a kinetic constant whose physical meaning has never been given in the literature K_m/Y_max = 450.30 min %^-1for C. flexuosus and K_m/Y_max = 113.12 min %^-1, for C. citratus. However the model is validated with very good coefficients of determination: R² = 0.9862 and 0.9952.

CONCLUSION:

The species Cymbopogon flexuosus flexuosus from Congo Brazzaville were perfectly adapted to Chad, despite climatic differences. It has retained its high biomass productivity and its low extraction yield in essential oil (0.86% DM). This lower essential oil content than that of local Cymbopogon citratus (1.26%) is largely balanced by a biomass production 5 times higher. The extraction kinetics fitted the three different models (Milojevic, Peleg and Michelis-Menton), each providing part of the overall, necessarily complex, explanation of the extraction process. This mechanism, already demonstrated on samples from the Congo, seems to be general to the cymbopogon genus and independently of the conditions of extraction by hydrodistillation, as can be seen by compiling data from the literature. For the species studied (flexuosus, winterianus, nardus, schoenanthus, citratus) the pseudo first order kinetic constant for hydrodistillation k varies 0.013-0,199 min^-1. The pseudo-second ordre (Peleg lead to k = 0.04563 min^-1, the pseudo first order kinetic constant for hydrodistillation. The maximum yields obtained were 1.7% by the Peleg model and 2.1% by the Michaelis-Menton model. The results obtained are consistent with each other.

ACKNOWLEDGMENT:

Authors thank, for their scientific and logistic support, following institutions: Institut National des Sciences Agronomiques et des Technologies Agroalimentaires de Lai (INSATAL) Tchad; Faculté des Sciences et Technique (UMNG) Brazzaville Congo; Ecole Supérieure de Technologie des Cataractes (EPrES) Brazzaville Congo

CONFLICT OF INTEREST:

No conflicts of interest.

REFERENCES:

1. Amenaghawon NA, Okhueleigbe KE, Ogbeide SE and Okieimen CO, 2014. Modelling the kinetics of steam distillation of essential oils from lemon grass (Cymbopogon spp) International Journal of Applied Science and Engineering 12 (2) : 107-115.

2. Baou AC, Manissa H Silou T and Chalchat JC, 2020. Pilot scale modeling of biomass and essential oil production from Cymbopogon nardus (L.) Rendle, acclimatized on the “plateau des Cataractes” in Congo-Brazzaville. J. Mater. Environ. Sci., 11(8):1315-1331.

3. Baou AC, 2021. Modélisation d’une unité semi-industrielle de production des huiles essentielles. Cymbopogon nardus acclimatées sur le plateau des Cataractes. Thèse de Doctorat d’Université, Université Marien Ngouabi, Brazzaville Congo.

4. Bellik, F.-Z., Benkaci-Ali, F., Alsafra, Z., Eppe, G., Tata, S., Sabaou, N., Zidani, R., 2019. Chemical composition, kinetic study and antimicrobial activity of essential oils from Cymbopogon schoenanthus L. Spreng extracted by conventional and microwave-assisted techniques using cryogenic grinding. Industrial Crops & Products, 139, 111505

5. Bucic-Kojic A, Mirela ST, Mate B and Darko V, 2007. Study of solid-liquid extraction kinetics of total polyphenols from grape seeds, Journal of Food Engineering 8, 236-242.

6. Chan CH, Yusoff R, Ngoh GC, 2014. Modeling and kinetics study of conventional and assisted batch solvent extraction, Chemical Engineering Research and Design, 92:1169-1186,

7. Dao TP, Do HT, Le QK, Phap NVG, Bach LG, Muoi NV and Cang MH, 2020. Kinetic Studies on Extraction of Essential Oil from Lemongrass Leaves (Cymbopogon citratus) by Steam Distillation Industrial. Scale Asian Journal of Chemistry 32 (6) :1399-1403.

8. Desai M A, Parikh J, De AK, 2014. Modelling and optimization studies on extraction of lemongrass oil from Cymbopogon flexuosus (Steud,) Wats, Chemical Engineering Research and Design 92 : 793-803.

9. Desai M A, Parikh J, De AK, 2015Extraction of essential oil from lemongrass using microwave radiation: optimization comparative kinetics and biological studies. ACS Sustainable Chemistry Engineering 3(3):421-431.

10. Koul VK, Gandotra BM, Koul S, Ghosh S, Tikoo CL, Gupta AK, 2004. Steam distillation of lemon grass (Cymbopogon spp,), Indian Journal of Chemical Technology 11: 135-139.

11. Lutfi N.A.F, Atan M.F, Rahman N.A, Salleh S.F, Wahab N.A, 2016. Study on Mathematical Model In Simulating Cymbopogon Winterianus Essential Oil Extraction By Steam Distillation. Journal of Applied Science, Process Engineering, 3(1): 1-16.

12. Meziane IAA, Bali N, Belblidia NB, Abatzoglou N, Benyoussef E-Hadi, 2019. The first-order model in the simulation of essential oil extraction kinetics, Journal of Applied Research on Medicinal and Aromatic Plants, doi: https://doi.org/10.1016/j.jarmap.2019.100226
Milojevic SZ, Stojanovi TD, Palic R., Miodrag L., Lazic M., Vlada B. Veljkovic V., 2008. Kinetics of distillation of essential oil from comminuted ripe juniper (Juniperus communis L.) berries, Biochemical Engineering Journal, 39: 547–553.

13. Milojevic SZ, Radosavljevic DB, Pavicevic VP, Pejanov S, Veljkovic VB, 2013. Modeling the kinetics of essential oils from plant materials. Hem. Ind.; 843-850.

14. Mjeri J, Chakroun I, Abderraba M, Mejri M, 2014. Study of hydro-distillation process of Ruta chalpensis L essential Oil, Research Journal of Agriculture and Environmental Mangement 3 (4): 311-518.

15. Peleg M, 1988. An Empirical model for description of moisture sorption curves, Journal of Food Science, 53 (4): 1216- 1219.

16. Shafaeï SM, Masoumi AA, Roshan H, 2016. Analysis of water adsorption of bean etchickpea during soaking using Peleg model, Journal of the Saudi Society of Agriculture Sciences 15: 135 – 144

17. Silou T, Malanda M., Loubaki L., 2004. Optimisation de l’huile essentielles de Cymbopogon citratus grace à un plan factoriel 2³. J. Food Eng. 219-223.

18. Sovová H, Aleksovski SA., 2006. Mathematical model for hydrodistillation of essential oils. Flavour and Fragrance Journal, 21(6): 881-889.

Received on 17.02.2022 Modified on 09.03.2022

Asian J. Research Chem. 2022; 15(3):228-234.

DOI: 10.52711/0974-4150.2022.00041