Air Diffusion of Aroma-Active Components from Crude Citrus Essential

Apr 3, 2018 - Citrus essential oils' (CEOs') quality has been enhanced by solvent extraction, aiming at both the recovery of extract phases enriched o...
0 downloads 4 Views 3MB Size
Article Cite This: Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

pubs.acs.org/IECR

Air Diffusion of Aroma-Active Components from Crude Citrus Essential Oils and Their Extract Phases Obtained by Solvent Extraction Daniel Gonçalves,*,† Patrícia Costa,*,‡ Caroline L. Bejar,§ Agathe Bocquet,§ Christianne E. C. Rodrigues,*,† and Alírio E. Rodrigues‡ †

Separation Engineering Laboratory (LES), Department of Food Engineering (ZEA-FZEA), University of São Paulo (USP), Pirassununga, São Paulo 13635-900, Brazil ‡ Laboratory of Separation and Reaction Engineering-Laboratory of Catalysis and Materials (LSRE-LCM), Faculdade de Engenharia, Universidade do Porto, 4099-002 Porto, Portugal § IUT Lyon 1, Université de Lyon, 69622 Lyon, France S Supporting Information *

ABSTRACT: Citrus essential oils’ (CEOs’) quality has been enhanced by solvent extraction, aiming at both the recovery of extract phases enriched on oxygenated compounds and maintenance of the typical citrus aroma. The air diffusion of aroma-active compounds found in crude CEOs, orange and acid lime, and their extract phases obtained by liquid−liquid extraction using hydroalcoholic solvents were thus evaluated. Experimental assays were conducted in a diffusion tube, where lower odor intensities were related to longer distances from the liquid mixture. Terpenes were mainly responsible for crude CEOs and orange extract phase aroma due to their high concentration in the vapor phase, while citral exhibited higher odor intensity in a region close to the acid lime extract phase. Finally, diffusion profiles simulated by COMSOL software fit well with those experimentally obtained, mainly for terpenes from crude CEOs. process are described as more stable to oxidation reactions8 and highly soluble in aqueous solutions,18 which facilitates their application in aqueous systems as perfumes or alcoholic beverages (when food-grade ethanol is used).11 However, it is crucial that the separation process does not affect the typical CEO aroma. Therefore, recent studies about the effect of the fractionation process on the aroma profiles of Citrus latifolia (acid lime)12 and Citrus sinensis (orange)11 have proved that it is possible to recover extract phases with sensory qualities similar to those of the crude CEOs. The aroma of a liquid mixture is perceived by the human nose when the volatile components are diffused through the

1. INTRODUCTION Citrus essential oils (CEOs) are important raw materials present in the formulation of perfumes, foods, beverages, soft drinks, cosmetics, pharmaceuticals, and so on.1−4 The volatile fraction of CEOs is mainly composed of terpenes (∼0.95 mass fraction),5 which have been reported as unstable components, easily degraded when exposed to the air, light, or heat,6 generating compounds with unpleasant odor7 that will compromise the final aroma and commercial value of the CEOs.8,9 CEOs are also composed of oxygenated compounds, which, although in lower amounts, are mainly responsible for the typical CEOs’ aroma,10 accounting for their quality and price establishment.9 Solvent extraction is a technique widely applied to separate the oxygenated fraction from the terpenes present in this kind of natural matrix9,11−14 that allows one to maintain the original CEO aroma15,16 and improve its stability.17 Extract phases resulting from the solvent extraction © XXXX American Chemical Society

Received: Revised: Accepted: Published: A

December 18, 2017 March 9, 2018 April 3, 2018 April 3, 2018 DOI: 10.1021/acs.iecr.7b05203 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

Table 1. Chemical Names, Empirical Formulas, CAS Numbers, Provenance, Suppliers, and Experimental Purities of the Standards

a

component

empirical formula

CAS number

source

supplier mass fraction

experimental puritya

water ethanol limonene β-pinene γ-terpinene α-pinene citral linalool octanal

H2O C2H6O C10H16 C10H16 C10H16 C10H16 C10H16O C10H18O C8H18O

7732-18-5 64-17-5 5989-27-5 18172-67-3 99-85-4 7785-26-4 5392-40-5 78-70-6 124-13-0

Chem-Lab, Belgium Sigma-Aldrich, U.S.A. Fluka, Switzerland Sigma-Aldrich, U.S.A. Sigma-Aldrich, U.S.A. Sigma-Aldrich, U.S.A. Sigma-Aldrich, U.S.A. Sigma-Aldrich, U.S.A.

≥0.998 0.97 ≥0.99 ≥0.97 0.98 ≥0.96 0.97 ≥0.92

0.999 0.988 0.999 0.977 0.99 0.99 0.992 0.920

Experimentally determined by GC-FID analysis without further purification, given as mass fraction.

surrounding air.19 It is a complex phenomenon dependent on several parameters, such as the volatility of the components (vapor pressure), diffusion coefficients, molecular structure, as well as the evaporation time and distance from the odor source.20−23 Additionally, the interactions between the molecules have a direct influence on the migration of the components from the liquid to the vapor phase. In the particular case of the extract phases from C. sinensis and C. latifolia, the presence of polar components (e.g., ethanol and water) tends to “push out” nonpolar components (e.g., terpenes), increasing their concentration in the vapor phase and affecting their olfactory perception, as observed by Gonçalves and co-workers.11,12 Taking into consideration all of these aspects, it would be interesting to study the aroma profile of citrus extract phases over time and distance, in order to evaluate what happens with the aroma when a consumer opens a bottle containing these mixtures. The diffusion of aroma-active components from a liquid mixture in the air and their perception by the human nose are processes that can be described by chemical engineering and psychophysics concepts, as demonstrated by Teixeira and co-workers23 through the simulation of diffusion above a liquid perfume mixture, over time and distance from the source. The air diffusion of aroma-active components of extract phases obtained from the fractionation of CEOs was thus evaluated. Extract phases from C. sinensis and C. latifolia EOs, obtained by solvent extraction with hydroalcoholic solvents, were studied simulating an alcoholic beverage with high ethanol content (over 0.4 mass fraction). The propagation of the volatiles from crude C. sinensis and C. latifolia EOs was also studied for comparison purposes. The experimental diffusion profiles were evaluated over time and distance from the liquid mixture in a diffusion tube at 298.2 K, whereas the vapor compositions were assessed by gas chromatography (GC) and the odor intensity of each compound was estimated by the Stevens’s power law concept.24 Moreover, the diffusion phenomenon of each component over the time and distance from the source was simulated by the COMSOL Multiphysics software, and the calculated results were compared to the experimental ones by their relative deviation.

the solvents. The empirical formulas, molar masses, CAS registry numbers, source, and supplier purities of each standard are displayed in Table 1. 2.2. Liquid−Liquid Equilibrium. Extract phases of both citrus species were acquired by liquid−liquid equilibrium (LLE) following the procedure reported in several studies.9,11−13,25−29 For that, each CEO was mixed with the solvent using a mass ratio of 1:1. Ethanol with a 0.4 water mass fraction was prepared for the orange system, while ethanol with a 0.5 water mass fraction was prepared for the acid lime system. Solvents were chosen based on results obtained in previous studies.11,12 The liquids were mixed in a polypropylene tube, centrifuged, and maintained at a controlled temperature of 298.2 ± 0.5 K for approximately 20 h in order to attain the thermodynamic equilibrium. After that, two phases were formed with a welldefined interface: the top terpene-rich phase (raffinate) and the bottom solvent-rich phase (extract). Finally, the extract phases were collected using syringes, and their compositions were assessed by gas chromatography (GC, Varian CP-3800, U.S.A.) and Karl Fisher titration (Metrohm, 787 KF Titrino, Switzerland) analysis. This procedure was performed in triplicate. 2.3. Chemical Analysis of the Liquid Mixtures. Identification of the main volatile components in the liquid CEOs was performed in a gas chromatograph (GC, Varian CP3800, U.S.A.) equipped with a split/splitless injector coupled to a mass spectrometer (MS, Varian SATURN 2000, U.S.A.), and DB-FFAP (Agilent, U.S.A.) capillary column (0.25 μm film thickness, 30 m × 0.25 mm i.d.). The injector temperature was set at 523 K. Samples were diluted in 1-propanol (SigmaAldrich, U.S.A.) using a mass ratio of 1:1, and 0.1 μL of the sample was injected using a split ratio of 50:1. Liquid sampling and injection were performed manually using a syringe from SGE (Australia). The carrier gas (He N60) flow rate was set at 1.56 mL/min. The oven temperature program was initially set at 373 K, then raised up to 503 K at a rate of 8 K/min, and held isothermal for 2.25 min. The mass spectra scanning range was from 80 to 500 m/z, and the source ionization energy was 70 eV. The mass spectra of the components were compared to the NIST98 Spectral Library, the mass spectral database of Flavors and Fragrances of Natural and Synthetic Compounds 2 (FFNSC2),30 and the retention times of the pure compounds injected under the same conditions. This procedure was conducted in triplicate for each crude CEO. Quantification of the identified components was performed using a GC Varian CP-3800 (U.S.A.) equipped with a split/ splitless injector, flame ionization detector (FID), and DBFFAP (Agilent, U.S.A.) (0.25 μm film thickness, 30 m × 0.25 mm i.d.) capillary column. The oven temperature program was

2. MATERIAL AND METHODS 2.1. Chemicals. The CEOs were industrially extracted from the peel of orange (Citrus sinensis (L.) Osbeck, var. Valencia, and Pera Rio) and acid lime (Citrus latifolia Tanaka) by cold pressing. The crude CEOs were kindly donated by Louis Dreyfus Company (Bebedouro/São Paulo, Brazil). Ethanol was mixed with deionized water in different mass ratios to obtain B

DOI: 10.1021/acs.iecr.7b05203 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

Figure 1. Diffusion tube. (a) Schematic representation: the arrows indicate the vapor direction; (b) photograph of the apparatus in the LSRE-LCM; (c) glass vessel with a liquid orange crude essential oil sample.

the same experimental conditions employed for the liquid samples (section 2.3). The concentration of each component in the vapor phase (Cvi , g/m3) was achieved by an external standard procedure, and the calibration curves can be accessed in the Supporting Information. The experimental assay in the diffusion tube was carried out in triplicate for each liquid sample, and the GC-FID analysis of vapor samples was carried out only once for each port and time. The uncertainty in the vapor concentration was estimated by the Type A standard31 considering the experimental results from the three diffusion assays. 2.5. Odor Intensity Estimation. The aroma diffusion profile was expressed in terms of odor intensity of the volatile aroma-active components. For that, the Cvi of each major component present in the liquid phase was converted into perceived sensations using the psychophysical Stevens power law concept, which is a nonlinear relation between the strength of a stimulus and its sensory perception.24 The odor intensity of the component i (Ψi) was calculated as expressed in eq 1.

the same as the one employed for the GC-MS analysis. The carrier gas (He N60) flow rate was set at 1.13 mL/min, and the detector temperature was maintained at 523 K. The water content in the extract phases was assessed by Karl Fisher titration (Metrohm, 787 KF Titrino, Switzerland), while the volatile components were quantified by an external standard method, and the calibration curves equations can be found in the Supporting Information. The GC-FID analysis was conducted in triplicate for each liquid sample (crude CEOs and extract phases). The uncertainties for the crude CEOs and the extract phase compositions were estimated by the Type A standard.31 2.4. Experimental Diffusion Profiles. Diffusion of the crude CEOs and their extract phases was carried out in a diffusion tube (Figure 1) at a controlled room temperature of 298.2 ± 0.5 K. The equipment was a stainless steel-made jacketed tube, 2 m long with an internal diameter of 2.41 cm. It contained five sampling ports positioned at distances of 0.13 (Port 1), 0.38 (Port 2), 0.63 (Port 3), 1.13 (Port 4), and 1.63 m (Port 5) from the liquid−vapor interface, where only Ports 1−4 were used. More specifications about the equipment can be found in Teixeira and co-workers.23 For the analysis of odor diffusion profiles, 1 mL of each liquid sample was placed in a glass vessel (Figure 1c) and connected to the bottom of the diffusion tube (Figure 1b). Vapor phases (0.2 mL) were collected from each sampling port using a gastight syringe with pressure lock (Hamilton, U.S.A.) at the first minute and then every hour during 7 h from the instant that the glass vessel containing the sample was connected to the diffusion tube. Vapor samples were immediately analyzed by GC-FID, under

⎛ Civ ⎞ni Ψi = ⎜ ⎟ ⎝ ODTi ⎠

(1)

where ODTi is the odor detection threshold of component i in the air (in g/m3), defined as the minimum concentration of the fragrance i in the air that can be detected by the human nose,32 and ni is the power exponent. 2.6. Air Diffusion Simulation. The diffusion of each component for each port (distance) over time was simulated by COMSOL Multiphysics software (COMSOL, Inc., Burlington, C

DOI: 10.1021/acs.iecr.7b05203 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

Table 2. Initial Concentration in the Liquid Mixture (C0, in g/m3) and Air Diffusion Coefficient (Di) of Each Studied Component crude essential oils component

orange

orange

1.99 × 10−2 8.08 × 10−1 3.20 × 10−3 4.21 × 10−3

1.99 1.16 5.31 1.31

× × × ×

4.50 × 10 8.39 × 10−4

10−2 10−1 10−1 10−1

Di (106 m2/s)a

acid lime −1

ethanol α-pinene β-pinene limonene γ-terpinene octanal linalool citral (neral + geranial) a

extract phases

acid lime

5.12 × 10−3

5.36 × 10−2

3.96 5.16 3.44 3.53 8.60

× × × × ×

−1

10 10−4 10−4 10−3 10−4

9.2 6.0 6.0 6.0 6.0 6.4 5.8 5.9

1.14 × 10−2

± ± ± ± ± ± ± ±

0.9 0.6 0.6 0.6 0.6 0.7 0.6 0.6

Calculated by eq 9. Standard uncertainties u for diffusion coefficient u(Di) are expressed after each mean value.

Table 3. Chemical Composition of the Liquid CEOs and Their Extract Phases, Expressed in Mass Percentage (100wi), Molar Mass (Mi), Vapor Pressure at 298.2 K (Psat i ), Odor Detection Threshold in Air (ODTi), and Power Law Exponent (ni) of Each Identified Component crude essential oils (100wi)a component (i)

orange

Solvent water ethanol Terpene Hydrocarbons α-pinene 2.37 β-pinene limonene 96.1 γ-terpinene Oxygenated Compounds octanal 0.38 octanol 0.08 nonanal 0.04 neral 0.05 geranial 0.09 linalool 0.50 α-terpineol 0.05 geraniol nerol citronellal 0.30 decanal undecanal 0.02 dodecanal 0.05 geranyl acetate

± 0.06 ± 0.1

± ± ± ± ± ± ±

0.01 0.01 0.01 0.01 0.01 0.01 0.01

acid lime

2.3 13.4 61.4 15.1

2.0 4.2 0.35 0.42 0.27 0.06

± ± ± ±

± ± ± ± ± ±

0.5 0.1 0.3 0.2

0.2 0.4 0.01 0.01 0.01 0.01

± 0.01

extract phases (100wi)a orange

acid lime

Mi (g/mol)

45.4 ± 0.3 53.6 ± 0.3

51.9 ± 0.1 46.0 ± 0.1

18.02 46.07

0.01 ± 0.01

0.06 0.02 0.41 0.10

0.61 ± 0.01

0.06 ± 0.01 0.03 ± 0.01 Tracesg 0.01 ± 0.01 0.01 ± 0.01 0.14 ± 0.01 0.02 ± 0.01

± ± ± ±

0.05 0.01 0.03 0.01

0.50 ± 0.01 0.83 ± 0.01 0.03 ± 0.01 0.06 ± 0.01 0.04 ± 0.01 Tracesg

0.02 ± 0.01 0.11 ± 0.01

± 0.01 ± 0.01

0.02 ± 0.01 Tracesg Tracesg

0.39 ± 0.01

0.01 ± 0.01

b Psat i (Pa)

3172 7901

ODTi (g/m3)c

nid

3.74 × 10−2

0.58

136.2 136.2 136.2 136.2

633.3 390.6 206.4 144.9

2.88 8.75 5.60 2.00

× × × ×

10−4 10−3e 10−4 10−3e

0.49 0.35 0.37 0.35

128.2 130.2 142.2 152.2 152.2 154.2 154.3 154.3 154.3 154.3 156.3 170.3 184.3 196.3

157.3 10.59 49.33 12.17 12.17 21.33 5.64 4.001 1.773 37.33 13.73 11.09 2.04 4.18

1.46 1.11 5.92 2.78 2.78 1.73 1.09 9.01 5.67 2.94 5.67 1.40 3.30 1.47

× × × × × × × × × × × × × ×

10−5 10−4 10−5 10−5e 10−5e 10−5 10−4 10−6 10−5 10−5f 10−5 10−4 10−5 10−2

0.33 0.29 0.34 0.32 0.32 0.35 0.35 0.36 0.35 0.36 0.39 0.39 0.35 0.35

a

Standard uncertainties u for the mass percentage u(100wi) are expressed after each mean value. bExperimental, estimated, or extrapolated values from SRC HysProp,41 NCBI PubChem,42 and Parchem43 Databases. cGeometric averages from van Gemert.44 dFrom Devos and co-workers.45 e From Gonçalves and co-workers.12 fFrom Gonçalves and co-workers.11 gComposition lower than 0.01 mass percentage.

∂Ci ∂ 2Ci = Di × ∂t ∂z 2

MA, U.S.A.). The diffusion phenomenon was governed by Fick’s second law (eq 2), assuming no mass generation or consumption and no convection at a constant temperature of 298.2 K and pressure of 1.02 × 105 Pa. The transport of diluted species module was used to represent the air diffusion phenomenon, with extremely fine mesh built in all computational domains. Once the tube radius was much smaller than its height (r ≪ z), it was assumed to have a 1D axisymmetric space dimension. Therefore, the diffusion was evaluated only over the distance from the source (z), as shown in eq 3. ∂Ci + ∇( −Di × ∇Ci) = 0 ∂t

(3)

where Ci is the concentration of component i (g/m3) and Di is the diffusion coefficient of component i in the air (m2/s), which was assumed to be constant over the time and position. The initial conditions were Vapor t = 0: Ci = Ci0 = 0

(4)

Liquid mixture

t = 0: Ci = C0 (2)

(5)

The boundary conditions were D

DOI: 10.1021/acs.iecr.7b05203 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

Figure 2. Diffusion profile of the crude orange essential oil. (a) Port 1; (b) Port 2; (c) Port 3; (d) Port 4. Experimental data: ■, α-pinene; ▲, limonene; ●, octanal; ◆, linalool. Simulated values by COMSOL Multiphysics software: , α-pinene; − − −, limonene; · · ·, octanal; · · − · · − · · −, linalool.

2.7. Statistical Analysis. The effect of the diffusion time and distance from the source on the odor intensity of each monitored component was evaluated by variance analysis with Duncan’s test35 at a significance level of P ≤ 0.05, using SAS software (version 9.2, SAS Institute Inc., U.S.A.). The statistical analysis is available in the Supporting Information.

Vapor z = 0: Ci = C0

z → ∞: Ci → 0

(6)

t → ∞: Ci → 0

(7)

Liquid mixture z = 0, t ≥ 0: Ci = C0

(8)

3. RESULTS AND DISCUSSION 3.1. Chemical Composition of the Liquid Samples. The chemical compositions of the crude CEOs and their extract phases are shown in Table 3, in mass fraction. The main volatile components identified and their composition are in agreement with data reported in previous studies for the same species.9,11,12 Both orange and acid lime EOs are rich in terpenes, with the orange EO containing the highest concentration (0.99 against 0.92 mass fractions). Although the terpene profile of the acid lime EO is quite different from the orange EO, in this last EO, terpenes are basically represented by limonene. Linalool and octanal were the main oxygenated compounds quantified in the orange EO, whereas neral and geranial were the most abundant in the acid lime EO. Relating to the extract phases, they were mostly composed of ethanol and water (more than 0.90 mass fraction). The water content in the acid lime extract phase was higher than that in the orange one because the solvent used for its fractionation was composed of higher water content (0.4 against 0.5 mass fraction). As a result of the deterpenation process, the composition of the terpenes in the extract phases was much lower than that in the crude EOs, being close to 0.006 mass fraction for the orange and acid lime extract phases. 3.2. Diffusion Profiles. 3.2.1. Crude Orange Essential Oil. The diffusion profiles of the crude orange EO are shown in Figure 2 for all of the distances from the vapor−liquid interface, where the lines correspond to the simulated diffusion profile. Although several components were identified in the liquid phase (Table 2), only limonene, α-pinene, octanal, and linalool

where C0 is the composition of component i in the liquid mixture (Table 2). The Di value of each component was calculated by eq 9 as proposed by Fuller and co-workers33 and described by Poling and co-workers.34 Di = 1.013 × 10−2 × T1.75 ×

Mi + Mair Mi × Mair

P × ( 3 Vi +

3

Vair )2

(9)

where T is the temperature (298.2 K), P is the local pressure (1.02 × 105 Pa), and Mi and Mair are the molar masses of component i and of the air (28.971 g/mol). Vair and Vi are the molecular diffusion volumes of the air (19.7 × 10−6 m3/mol) and of component i calculated by eq 10. Vi =



Ni × vi

(10)

where Ni is the number of atoms of element i (C, H, or H), and vi is the atomic volume (m3/mol) from Fuller and co-workers.33 Di values calculated for each component are presented in Table 2, with the following uncertainty. The relative deviations (δi, in %) between the experimental calc (Ψexp i ) and calculated (Ψi ) odor intensity of each component was calculated by eq 11 and can be found in the Supporting Information. δi = 100 ×

|Ψiexp − Ψicalc| Ψiexp

(11) E

DOI: 10.1021/acs.iecr.7b05203 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

Figure 3. Diffusion profile of the crude acid lime essential oil. (a) Port 1; (b) Port 2; (c) Port 3; (d) Port 4. Experimental data: ■, α-pinene; ▲, limonene; ●, β-pinene; ▼, γ-terpinene; ◆, citral. Simulated values by COMSOL Multiphysics software: , α-pinene; − − −, limonene; · · ·, βpinene; - - -, γ-terpinene; · · − · · − · · −, citral.

Figure 4. Diffusion profile of the orange extract phase. (a) Port 1; (b) Port 2; (c) Port 3; (d) Port 4. Experimental data: ■, α-pinene; ▲, limonene; ○, ethanol. Simulated values by COMSOL Multiphysics software: , α-pinene; − − −, limonene; · − · − · −, ethanol.

3) in comparison with linalool, which was only identified near the aroma source (Port 1). The main volatile components present in crude orange EO are in agreement with those from a previous study.11 The aroma of the crude EO is dominated by terpenes up to a distance of 0.63 m (Port 3), being then replaced by oxygenated compounds, namely, octanal, for longer distances (Port 4). The odor of all components is more intense close to the liquid (Port 1) and after some time (5 h), meaning that the aroma of the crude orange EO becomes attenuated when the distance from the liquid EO is longer. Close to the aroma source (Port 1,

were detected in the vapor samples collected from the diffusion tube, as can be observed in Figure 2. As a consequence of the high amount of limonene in the crude orange EO, together with its high vapor pressure (206.4 Pa), this component is present in higher amounts in the vapor phase, which contributes to high values of odor intensity. In contrast, αpinene exists at a lower composition in the crude EO, but due to its high vapor pressure and power law exponent (Table 2), this component also presents a high odor intensity value. Regarding the oxygenated compounds, octanal was identified over longer distances from the liquid mixture (Ports 1, 2, and F

DOI: 10.1021/acs.iecr.7b05203 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

Figure 5. Diffusion profile of the acid lime extract phase. (a) Port 1; (b) Port 2; (c) Port 3; (d) Port 4. Experimental data: ■, α-pinene; ▲, limonene; ●, β-pinene; ▼, γ-terpinene; ◆, citral; ○, ethanol. Simulated values by COMSOL Multiphysics software: , α-pinene; − − −, limonene; · · ·, β-pinene; - - -, γ-terpinene; · · − · · − · · −, citral; · − · − · −, ethanol.

Figure 2a), limonene and α-pinene exhibited similar behavior. In general, an increase in the odor intensities was observed over time, mainly between 1 min and 2 h of diffusion. After that, the odor intensities did not exhibit significant variation (Table S1). In short, the crude orange EO aroma will be a combination of limonene and α-pinene odor over 7 h of evaporation up to 0.63 m from the liquid. 3.2.2. Crude Acid Lime Essential Oil. The diffusion profiles of the crude acid lime EO are exhibited in Figure 3 for all distances from the vapor−liquid interface. As observed for the crude orange EO, only the main components present in the acid lime EO were detected in the vapor phase, and the results are in accordance with prior headspace study.12 Terpenes αpinene, limonene, and β-pinene were identified in all ports, whereas γ-terpinenes were not detected in the vapor samples collected at the longest distance from the liquid EO (Port 4). As observed for crude orange EO, overall, the odor intensities increased slightly over time, and after about 2 or 3 h of diffusion, their values were constant (Table S2). The oxygenated compounds neral and geranial, represented as citral, were only detected in Port 1, exhibiting almost no variation in its odor intensity over time (Table S2). Citral is presented at low concentration in the crude EO, but its odor intensity in a region close to the aroma source is similar to the one of limonene after 3 h (Figure 3a). This means that the highest contribution of citral to the final aroma of the crude EO is close to the liquid, and in the first hours after a bottle containing the EO is opened. α-Pinene dominated the odor space of the crude acid lime EO, whereas β-pinene and γterpinene exhibited a lower and equivalent contribution. As in the crude orange EO, the significant contribution of α-pinene can be explained by its high vapor pressure, low ODT, and power law exponent (Table 3). In addition, the odor intensity of the components decreased with the distance from the liquid (Table S2). Briefly, the crude

acid lime OE will smell mainly like α-pinene during 7 h of diffusion up to 0.63 m from the source. 3.2.3. Extract Phases. The diffusion profiles of the extract phases from the LLE are presented in Figure 4 for orange system and in Figure 5 for the acid lime system. It was observed that the odor intensities of the components in the extract phases were lower than those attained in the respective crude EOs, which means that the aroma of the extract phases is less powerful than that of the crude EOs. As result of its high content in the extract phases and high vapor pressure (Table 2), ethanol was identified in all ports and in both systems. However, this alcohol has a high ODT value, meaning that it will have low contribution to the final aroma of the mixture. The low contribution of ethanol to the aroma of the extracts phases was mainly observed for orange extract in Ports 1 and 2 (Figure 4a,b) and for acid lime extract at all distances (Figure 5). Supposing that the extract phase was a beverage or a perfume, it is positive that the unpleasant ethanol aroma was not identified during its consumption or use. The low odor intensity of ethanol in the orange extract phases in a region close to the liquid mixture was confirmed in the sensorial analysis reported by Gonçalves and co-workers,11 where panelists did not identify the ethanol odor in these phases. However, the ethanol odor becomes more evident at longer distances, such as 0.63 m (Port 3, Figure 4c) and 1.13 m (Port 4, Figure 4d) from the liquid. Linalool and octanal were not detected in any sample from the orange extract phase (Figure 4), probably due to their low compositions in the liquid mixture, low vapor pressures (Table 3), and higher affinity with the solvent in relation to the terpenes, 12 denoting lower release of this group of components.21 Limonene was the main contributor for the aroma of the orange extract phase in Ports 1 and 2. Therefore, when a bottle containing a mixture (such as the orange extract phase) is opened, it will smell like limonene near the liquid mixture. Although limonene and α-pinene (nonpolar comG

DOI: 10.1021/acs.iecr.7b05203 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

4 (Figure 2d and Table S1), and ethanol from the extract phases (Figures 4 and 5 and Tables S3 and S4). However, some exceptions were observed in the orange extract phase for αpinene in Port 3 (Figure 4c) and in the acid lime extract phase for γ-terpinene in Port 3 (Figure 5c) and limonene in Port 4 (Figure 5d), where the simulated results were similar to the experimental ones. In the case of ethanol, a good description of its diffusion process was observed in Ports 1 and 4 for the orange extract phase (Figure 4a,d) and in Port 3 for the acid lime extract phase (Figure 5c). Although some discrepancies between the calculated and experimental data exist, the simulation procedure using Fick’s second law demonstrated to be a simple but important tool to describe the air diffusion phenomenon of volatile components from citrus EOs.

pounds) are at low concentrations in the mixture (Table 3), the presence of water and ethanol induces an increase in the polarity of the mixture, pushing out nonpolar components to the vapor phase.11,21 Moreover, it was observed that the odor itensity of these terpenes almost did not changed over the diffusion time (Table S3). Oppositely, in the acid lime extract phase, the oxygenated compound citral exhibited the highest odor intensity in Port 1 after 3 h (Figure 5a), but it was not identified in the samples collected from other distances. Citral is reported as the main component responsible for the typical lemon/lime aroma.12,36−40 Then, it is expected that when a vessel containing acid lime extract phase is opened, it will exhibit a strong lemon odor close to the liquid mixture. Terpenes limonene, β-pinene, and γ-terpinene were detected in all ports, while α-pinene was not detected in Port 4. Analogously to the other assays, a slight increase in the odor intensities was observed over time, mainly until 3 h of propagation in air (Table S4). β-Pinene and γ-terpinene presented similar odor intensities over distance, while in Ports 2 and 3, the odor intensities of limonene and α-pinene were closer and higher than those of the other terpenes. As observed for the crude EOs, overall for both extract phases, the odor intensities decreased with the distance from the liquid source (Tables S3 and S4), meaning that their aroma became weaker for longer distances from the liquid mixture. Here, the extract phases were assumed as a final product, as a beverage, or as a perfume basis, containing the usual perfume solvent (ethanol + water) and fractionated CEO. For the obtainment of a pure fractionated essential oil, a subsequent step of solvent removal is required, which may affect the aroma profile, depending on the method employed. 3.3. Computational Simulation. The diffusion profile of each component simulated by COMSOL Multiphysics software is displayed in Figures 2−5 as lines. The relative deviation between experimental and calculated data, available in Tables S1−S4, ranged from 0.002 to 213%, with global deviation of 26.8%. In general, satisfactory results in the description of the air diffusion phenomenon of the terpenes from the crude CEOs were observed (Figures 2 and 3). However, in these same systems, the odor intensities of the oxygenated compounds linalool and citral were overestimated. Additionally, it was also observed that the simulation procedure was not able to calculate the odor intensity of several components for a short diffusion time (up to 1 h). Regarding the extract phases, it was verified that, in general, the odor intensities of terpenes were underestimated, while the calculated results were fitted well with the experimental citral diffusion from the acid lime extract phase (Figure 5a). It should be mentioned that for the simulation procedure employed in this study only the initial concentration in the liquid mixture and the diffusion coefficient of each component were considered, which were assumed as constant over the entire simulation, while thermodynamic relations such as the interaction between the molecules were not considered. Therefore, it is expected that for nonideal solutions such as the extract phases, which are rich in water, the diffusion model is unable to simulate the “pushing” effect that happens with the terpenes in these mixtures, as observed by Gonçalves and coworkers.11,12 Besides that, variations in the experimental data can be responsible for higher deviations between the experimental and calculated data, as observed for octanal from orange EO in Port

4. CONCLUSIONS The odor intensity of the components decreases as the distance from the liquid source increases. In practical terms, when a bottle containing the crude orange EO is opened, limonene and α-pinene will dominate the aroma up to 7 h and up to 1.13 m from the liquid source. In the case of the crude acid lime EO, it will smell like α-pinene; however, an a larger distance from the liquid, its odor intensity will be very low. The aroma of the extract phases is less intense in comparison with that of the crude CEOs. Ethanol did not present strong odor intensity, mainly at a distance close to the liquid mixture, which is interesting in the case of alcoholic beverages or perfumes. Limonene and α-pinene were mainly responsible for the aroma of the extract phases despite their low compositions in the liquid mixtures. In the case of the acid lime extract, at a distance close to the liquid mixture, citral dominated the odor space up to 7 h, indicating a strong lemon/lime typical odor. Simulated diffusion profiles were similar to the experimental results mainly for terpenes from the crude EOs, but overall, the simulation procedure was inefficient for calculation of the aroma diffusion profiles of the components from the studied extract phases.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.iecr.7b05203. Odor intensity of each component from the crude CEO (Tables S1 and S2) and their extract phases (Tables S3 and S4), statistical analysis of experimental data, calculated values by COMSOL Multiphysics software and relative deviation between the experimental and calculated data, calibration curve equations followed by the corresponding determination coefficients, details of the obtainment of the calibration curves used for the quantification of the components in the liquid and vapor samples, and equations for the standard uncertainties estimation (PDF).



AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (D.G.). *E-mail: [email protected] (P.C.). *E-mail: [email protected] (C.E.C.R.). ORCID

Daniel Gonçalves: 0000-0002-2581-7207 H

DOI: 10.1021/acs.iecr.7b05203 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

Essential Oils by Solvent Extraction: Phase Equilibrium at 298.2 K. J. Chem. Eng. Data 2015, 60, 37−46. (14) Lago, S.; Rodríguez, H.; Arce, A.; Soto, A. Improved Concentration of Citrus Essential Oil by Solvent Extraction with Acetate Ionic Liquids. Fluid Phase Equilib. 2014, 361, 37−44. (15) Fantin, G.; Fogagnolo, M.; Maietti, S.; Rossetti, S. Selective Removal of Monoterpenes from Bergamot Oil by Inclusion in Deoxycholic Acid. J. Agric. Food Chem. 2010, 58, 5438−5443. (16) Gironi, F.; Maschietti, M. Continuous Countercurrent Deterpenation of Lemon Essential Oil by Means of Supercritical Carbon Dioxide: Experimental Data and Process Modelling. Chem. Eng. Sci. 2008, 63, 651−661. (17) Lago, S.; Rodríguez, H.; Soto, A.; Arce, A. Deterpenation of Citrus Essential Oil by Liquid−Liquid Extraction with 1-Alkyl-3Methylimidazolium Bis(trifluoromethylsulfonyl)amide Ionic Liquids. J. Chem. Eng. Data 2011, 56, 1273−1281. (18) Koshima, C. C.; Capellini, M. C.; Geremias, I. M.; Aracava, K. K.; Gonçalves, C. B.; Rodrigues, C. E. C. Fractionation of Lemon Essential Oil by Solvent Extraction: Phase Equilibrium for Model Systems at T = 298.2K. J. Chem. Thermodyn. 2012, 54, 316−321. (19) Teixeira, M. A.; Rodríguez, O.; Mata, V. G.; Rodrigues, A. E. The Diffusion of Perfume Mixtures and the Odor Performance. Chem. Eng. Sci. 2009, 64, 2570−2589. (20) Costa, P.; Teixeira, M. A.; Lièvre, Y.; Loureiro, J. M.; Rodrigues, A. E. Modeling Fragrance Components Release from a Simplified Matrix Used in Toiletries and Household Products. Ind. Eng. Chem. Res. 2015, 54, 11720−11731. (21) Costa, P.; Teixeira, M. A.; Mestre, G.; Carneiro, L.; Loureiro, J. M.; Rodrigues, A. E. Predicting Vapor-Phase Concentrations for the Assessment of the Odor Perception of Fragrance Chemicals Diluted in Mineral Oil. Ind. Eng. Chem. Res. 2017, 56, 8767−8777. (22) Teixeira, M. A.; Rodríguez, O.; Rodrigues, S.; Martins, I.; Rodrigues, A. E. A Case Study of Product Engineering: Performance of Microencapsulated Perfumes on Textile Applications. AIChE J. 2012, 58, 1939−1950. (23) Teixeira, M. A.; Rodríguez, O.; Rodrigues, A. E. Diffusion and Performance of Fragranced Products: Prediction and Validation. AIChE J. 2013, 59, 3943−3957. (24) Stevens, S. S. On the Psychophysical Law. Psychol. Rev. 1957, 64, 153−181. (25) Capellini, M. C.; Carvalho, F. H.; Koshima, C. C.; Aracava, K. K.; Gonçalves, C. B.; Rodrigues, C. E. C. Phase Equilibrium Data for Systems Composed of Oregano Essential Oil Compounds and Hydroalcoholic Solvents at T = 298.2K. J. Chem. Thermodyn. 2015, 88, 61−71. (26) Chiyoda, C.; Capellini, M. C.; Geremias, I. M.; Carvalho, F. H.; Aracava, K. K.; Bueno, R. S.; Gonçalves, C. B.; Rodrigues, C. E. C. Deterpenation of Bergamot Essential Oil Using Liquid−Liquid Extraction: Equilibrium Data of Model Systems at 298.2 K. J. Chem. Eng. Data 2011, 56, 2362−2370. (27) Gonçalves, D.; Koshima, C. C.; Nakamoto, K. T.; Umeda, T. K.; Aracava, K. K.; Gonçalves, C. B.; Rodrigues, C. E. da C Deterpenation of Eucalyptus Essential Oil by Liquid+liquid Extraction: Phase Equilibrium and Physical Properties for Model Systems at T = 298.2K. J. Chem. Thermodyn. 2014, 69, 66−72. (28) Koshima, C. C.; Umeda, T. K.; Nakamoto, K. T.; Venâncio, L. L.; Aracava, K. K.; Rodrigues, C. E. C. (Liquid+liquid) Equilibrium for Systems Composed of Clove and Allspice Essential Oil Compounds and Hydrous Ethanol at T = 298.2K. J. Chem. Thermodyn. 2016, 95, 54−62. (29) Oliveira, C. M.; Koshima, C. C.; Capellini, M. C.; Carvalho, F. H.; Aracava, K. K.; Gonçalves, C. B.; Rodrigues, C. E. C. Liquid−liquid Equilibrium Data for the System Limonene+carvone+ethanol+water at 298.2K. Fluid Phase Equilib. 2013, 360, 233−238. (30) Mondello, L. Flavors and Fragrances of Natural and Synthetic Compounds 2; Wiley: Hoboken, NJ, 2011. (31) Taylor, B. N.; Kuyatt, C. E. Guidelines for Evaluating and Expressing the Uncertainty of NIST Measurement Results; United States

Patrícia Costa: 0000-0001-9246-5611 Christianne E. C. Rodrigues: 0000-0002-5456-9708 Alírio E. Rodrigues: 0000-0002-0715-4761 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors would like to acknowledge FAPESP (Fundaçaõ de Amparo à Pesquisa do Estado de São Paulo, Brazil, 2013/ 11150-3, 2015/06162-8) and CNPq (303797/2016-9) for financial support and the Louis Dreyfus Company for the crude orange and acid lime essential oil donation. This work was cofinanced by FCT and FEDER through COMPETE 2020 (Project UID/EQU/50020/2013 - POCI-01-0145-FEDER006984). P.C. acknowledges her postdoctoral grant from the Fundaçaõ para a Ciência e a Tecnologia (SFRH/BPD/93108/ 2013).



REFERENCES

(1) Li, L. J.; Hong, P.; Chen, F.; Sun, H.; Yang, Y. F.; Yu, X.; Huang, G. L.; Wu, L. M.; Ni, H. Characterization of the Aldehydes and Their Transformations Induced by UV Irradiation and Air Exposure of White Guanxi Honey Pummelo (Citrus Grandis (L.) Osbeck) Essential Oil. J. Agric. Food Chem. 2016, 64, 5000−5010. (2) Masson, J.; Liberto, E.; Beolor, J. C.; Brevard, H.; Bicchi, C.; Rubiolo, P. Oxygenated Heterocyclic Compounds to Differentiate Citrus Spp. Essential Oils through Metabolomic Strategies. Food Chem. 2016, 206, 223−233. (3) Silvestre, W. P.; Agostini, F.; Muniz, L. A. R.; Pauletti, G. F. Fractionating of Green Mandarin (Citrus Deliciosa Tenore) Essential Oil by Vacuum Fractional Distillation. J. Food Eng. 2016, 178, 90−94. (4) Zhang, K.; Lin, Y.; Diao, Z. J.; Zhang, W. H.; Zheng, S.-P.; Liang, S. L.; Han, S. Y. Enzymatic Process Enhances the Flavour Profile and Increases the Proportion of Esters in Citrus Essential Oils. Chem. Biodiversity 2017, 14, e1700187. (5) Raeissi, S.; Peters, C. J. Experimental Determination of HighPressure Phase Equilibria of the Ternary System Carbon Dioxide +limonene+linalool. J. Supercrit. Fluids 2005, 35, 10−17. (6) López-Muñoz, G. A.; Balderas-López, J. A. Photothermal Characterization of Citrus Essential Oils and Their Derivatives. Thermochim. Acta 2014, 579, 40−44. (7) Raeissi, S.; Peters, C. J. Liquid−vapor and Liquid−liquid−vapor Equilibria in the Ternary System Ethane+limonene+linalool. J. Supercrit. Fluids 2005, 33, 201−208. (8) Cháfer, A.; de la Torre, J.; Muñoz, R.; Burguet, M. C. Liquid− liquid Equilibria of the Mixture Linalool+ethanol+water at Different Temperatures. Fluid Phase Equilib. 2005, 238, 72−76. (9) Gonçalves, D.; Teschke, M. E. E.; Koshima, C. C.; Gonçalves, C. B.; Oliveira, A. L.; Rodrigues, C. E. C. Fractionation of Orange Essential Oil Using Liquid−liquid Extraction: Equilibrium Data for Model and Real Systems at 298.2K. Fluid Phase Equilib. 2015, 399, 87−97. (10) Strano, M. C.; Calandra, M.; Aloisi, V.; Rapisarda, P.; Strano, T.; Ruberto, G. Hot Water Dipping Treatments on Tarocco Orange Fruit and Their Effects on Peel Essential Oil. Postharvest Biol. Technol. 2014, 94, 26−34. (11) Gonçalves, D.; Costa, P.; Rodrigues, C. E. C.; Rodrigues, A. E. Effect of Citrus sinensis Essential Oil Deterpenation on the Aroma Profile of the Phases Obtained by Solvent Extraction. J. Chem. Thermodyn. 2018, 116, 166−175. (12) Gonçalves, D.; Costa, P.; Rodrigues, C. E. C.; Rodrigues, A. E. Fractionation of Acid Lime Essential Oil Using Ethanol/water Mixtures: Effect of the Process on the Aroma Profile. J. Chem. Thermodyn. 2017, 108, 118−126. (13) Koshima, C. C.; Nakamoto, K. T.; Aracava, K. K.; Oliveira, A. L.; Rodrigues, C. E. C. Fractionation of Bergamot and Lavandin Crude I

DOI: 10.1021/acs.iecr.7b05203 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research Department of Commerce Technology Administration: Gaithersburg, MD, 1994. (32) Teixeira, M. A.; Rodríguez, O.; Gomes, P.; Mata, V.; Rodrigues, A. E. Perfume Engineering; Design, Performance & Classification, 1st ed.; Butterworth-Heinemann: Oxford, U.K., 2013. (33) Fuller, E. N.; Ensley, K.; Giddings, J. C. Diffusion of Halogenated Hydrocarbons in Helium. The Effect of Structure on Collision Cross Sections. J. Phys. Chem. 1969, 73, 3679−3685. (34) Poling, B. E.; Prausnitz, J. M.; O’Connell, J. P. The Properties of Gases and Liquids, 5th ed.; McGraw Hill: New York, 2004. (35) Duncan, D. B. Multiple Range and Multiple F Tests. Biometrics 1955, 11, 1−42. (36) Başer, K. H. C.; Buchbauer, G. Handbook of Essential Oils: Science, Technology, and Applications; CRC Press: Boca Raton, FL, 2010. (37) Brechbill, G. O. A Reference Book on Fragrance Ingredients, 1st ed.; Fragrance Books Inc.: New Jersey, 2006. (38) Hausch, B. J.; Lorjaroenphon, Y.; Cadwallader, K. R. Flavor Chemistry of Lemon-Lime Carbonated Beverages. J. Agric. Food Chem. 2015, 63, 112−119. (39) Lota, M. L.; de Rocca Serra, D.; Tomi, F.; Jacquemond, C.; Casanova, J. Volatile Components of Peel and Leaf Oils of Lemon and Lime Species. J. Agric. Food Chem. 2002, 50, 796−805. (40) Perez-Cacho, P. R.; Rouseff, R. Processing and Storage Effects on Orange Juice Aroma: A Review. J. Agric. Food Chem. 2008, 56, 9785−9796. (41) HysProp Database. SRC Inc. http://esc.srcinc.com/fatepointer (accessed Dec 18, 2017). (42) PubChem Database. National Center for Biotechnology Information (NCBI), https://pubchem.ncbi.nlm.nih.gov/ (accessed Dec 18, 2017). (43) Parchem Fine & Specialty Chemicals Home Page. http://www. parchem.com/ (accessed Dec 18, 2017). (44) van Gemert, L. J. Compilations of Odour Threshold Values in Air, Water and Other Media, 2nd ed.; Oliemans Punter & Partners BV: Houten, The Netherlands, 2003. (45) Devos, M.; Rouault, J.; Laffort, P. Standardized Olfactory Power Law Exponents; Editions Universitaires-Sciences: Dijon, France, 2002.

J

DOI: 10.1021/acs.iecr.7b05203 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX