Process optimization for the supercritical carbondioxide extraction of lycopene from ripe grapefruit (Citrus paradisi) endocarp

In this study, an underutilized citrus family fruit named grapefruit was explored for the extraction of lycopene using supercritical carbon dioxide (CO2) extraction technique. An experimental design was developed using response surface methodology to investigate the effect of supercritical carbon dioxide (CO2) operating parameter viz., pressure, temperature, CO2 flow rate, and extraction time on the extraction yield of lycopene yield from grapefruit. A total of 30 sets of experiments were conducted with six central points. The statistical model indicated that extraction pressure and extraction time individually, and their interaction, significantly affected the lycopene yield. The central composite design showed that the polynomial regression models developed were in agreement with the experimental results, with R2 of 0.9885. The optimum conditions for extraction of lycopene from grapefruit were 305 bar pressure, 35 g/min CO2 flow rate, 135 min of extraction time, and 70 °C temperature.

www.nature.com/scientificreports/ as it causes a strong bitter taste, which results in very limited or no utilization 12,13 . Very few attempts have been made for the value addition to grapefruit, such as solvent extraction as well as ultrasound assisted extraction of lycopene 14,15 , and also by using petroleum ether as extraction solvent 16 for successful extraction of this bioactive pigment called lycopene. Apart from this, oil extraction and isolation of polysaccharides has also been carried out from grapefruit 17 . However, it is necessary to carry out a sustainable and food-grade extraction of lycopene from grapefruit to commercialize as an antioxidant and natural colorant into the food system. The extraction of lycopene from fruits and vegetables is done using a solvent extraction method, which is the most prevalent and commercial method used in the natural colorant industries. Using standard nonpolar solvents like hexane, tetrahydrofuran, and chloroform, is non-sustainable and unsuitable for human consumption 15,18 . The solvent extraction process has several disadvantages: toxicity, disposal of solvent, a large quantity of organic solvent needed for extraction, high extraction temperatures, prolonged extraction time, less extraction efficiency, and dilution of the extract 19,20 . Moreover, higher extraction temperatures and prolonged extraction time lead to significant lycopene degradation, resulting in the isomerization of stable trans-isomer to a relatively unstable cis-isomer 21,22 .
These setbacks of the solvent extraction process necessitate the development of safer and cleaner methods of extraction of lycopene. One such technique is the supercritical fluid extraction (SFE), which utilizes supercritical fluids (the solvent) to separate the desired component (extractant) from the intricate food matrix 19,23 . The extraction of biotic compounds using a supercritical fluid emerges to be a green technology and produces the purest, cleanest, and safest product among all known extraction techniques available today. Although costlier, it is preferred because of its higher extraction efficiency and absence of residual solvent in the extracted material. Besides, it allows working at moderate temperatures, which invariably reduces degradation caused by thermal effect. Light and oxygen's negligible presence significantly enhances stability and prevents oxidation reactions 19,20 . It is also non-toxic and smoothly gets separated from the extract. Carbon dioxide is generally the preferred solvent in this technique because it converts into the supercritical stage at 31 °C under a pressure of 74 bars 19,23 . Several extraction parameters such as extraction temperature, pressure, time, CO 2 flow rate, co-solvent ratio (i.e., ethanol), among others, can be controlled to enhance the extraction efficiency and retain the quality of extracts maximally. So, it is essential to understand that the extraction conditions that maximally influence polyphenolics compounds' stability and oxidative degradation 19,24 .
Therefore, this study was investigated to understand the effect of the supercritical extraction parameter such as temperature, pressure, time, and CO 2 flow rate on the extraction yield of lycopene from the freeze-dried grapefruit powder. The central composite rotatable design (CCRD) of response surface methodology (RSM) was used to design the experiment. The extraction yield of the lycopene was measured on supercritical fluid chromatography.

Materials and methods
Materials and reagents. The ripe grapefruit was procured from the Division of Fruits and Horticultural Technology, IARI, New Delhi, India. It was harvested at a stage when the fruit exhibited 2/3rd yellow color. Lycopene was purchased from the Sigma Aldrich (St. Louis, MO, USA). Liquid CO 2 with the purity of over 99.5% was purchased from the Amit Labs, New Delhi, India. Other solvents used in quantifying extracted lycopene were purchased from Merck Life Science Pvt. Ltd. Mumbai, India.
Supercritical carbon dioxide extraction. The grapefruits were peeled, diced into small pieces, and lyophilized at − 52 °C for four days under dark condition. The lyophilized grapefruits were ground into powder using hammer mill and passed through a 250 µm size SI sieve to restrict the particle size. The powder was packed in LDPE zip lock bag and stored at − 20 °C for supercritical fluid extraction.
A 100 g of the powdered grapefruit was loaded into the extraction vessel of the supercritical carbon dioxide extraction system, Thar Technologies, USA (Model No. 7100). The operating parameters i.e. extraction temperature, pressure, CO 2 , and time were controlled using the operating software SuperChrom SFC Suite, Thar Technologies, USA (v5.9 version). ethanol @ 5% along with supercritical CO 2 @ 95% was used as extraction solvent. Ethanol was used as a co-solvent with CO 2 to enhance extraction yield 23,25 . A schematic of the supercritical carbon dioxide extraction of lycopene is given in Fig. 1. The extracted lycopene was collected in amber colored bottle and stored at − 20 °C before further quantitative study.
Quantification of lycopene. The lycopene's extraction yield was measured on supercritical fluid chromatography as per the protocol suggested by Runco et al. 26 , with some modification. The extracted lycopene was vacuum dried and dissolved into hexane. Dissolved lycopene was filtered through 0.22 µm of pore size PVDF membrane (Millipore Ltd., Germany), and 1 mL was loaded into a 1.5 mL amber-colored vial. Quantification was performed on UPC 2 Acquity System (Waters Technologies, USA) equipped with a BEH 2-EP column (2.1 × 150 mm × 5 µm) and a photodiode array detector. The absorbance was recorded at 452 nm. The mobile phase consisted of CO 2 and methanol of 85:15 (v/v) run under isocratic conditions. The chromatographs of the lycopene and n-hexane are given in Fig. 2.
Experimental design and statistical analysis. The experiment was designed using a 2 4  www.nature.com/scientificreports/  www.nature.com/scientificreports/ where Y is the response (yield), β 0 is the constant coefficient, β i , β j and β ij are the linear, quadratic, and interaction coefficients respectively, and x i , x j are the coded values of independent variables. The design point for the different process variables was obtained using the PROC RSREG procedure of the SAS software (version 9.3, SAS Institute, Cary, NC, USA) with no coding option. It was used for analysis by Response Surface Methodology (RSM) of the experimental data obtained. Five replicates at the center of the design were used to estimate a pure error sum of squares.

Results and discussion
Recovery of lycopene from grapefruit at different experimental runs varied between 13.4% and 77.2% (Table 1). The recovery percentages were calculated by evaluating the lycopene content in the fruit powder before extraction, recovered lycopene using SFE and the lycopene content in the residual powder. Analysis of the data indicated that the lycopene yield from grapefruit could be adequately described using a second order quadratic model as follows: where Y is the lycopene recovery (%), X 1 , X 2 , X 3 and X 4 represents pressure, flow rate, time and temperature respectively.
Validity of the statistical model is generally adjudged by the lack of fit to check adequacy of the model (Table 2(A)). Significance of the model and a non-significance of the lack of fit at five percent level of significance (1) www.nature.com/scientificreports/ indicated that the developed model for lycopene yield prediction from grapefruit was a good fit. An R 2 value of 0.99 reinforces the same. ANOVA (Table 2(B)) of the effect of linear, quadratic and interaction effects of the variables indicated that the pressure (X1) and time (X3) individually as well as their interaction effect significantly affected the lycopene yield. However, in its quadratic term, pressure and temperature were found to have significant effect on lycopene yield. All other linear, quadratic and interaction effects were found to be non-significant at 5% level of confidence (P < 0.05).
Effect of independent variables on lycopene extraction was done by analysing the response surface graphs plotted between any two independent variables keeping the third at the central point ( Fig. 3a-f).

Effect of independent variables of lycopene extraction. Pressure and time. With an increase in
pressure from 150 to 300 bars, it has been observed that lycopene yield increased at a given flow rate of the supercritical CO 2 and further reduced with increase in pressure up to 450 bars. The rate of increase however gradually reduces with the increase in the flow rate (Fig. 3a). Besides, lycopene yield increased with increase in the flow rate of supercritical CO 2 from 15 to 45 g/min and almost got stabilized with further increase in flow rate up to 55 g/min. This increase may be attributed to the fact that the increased pressure enhanced solvent density of the supercritical CO 2 resulting in enhanced solubility of lycopene as it has been reported by some researchers too 25,31 . Nevertheless, beyond a certain limit, pressure plays a negative role by decreasing the diffusion ability of the solvent because of the enhanced compaction of the samples at higher pressure leading to channelling of the supercritical CO 2 around it rather than diffusing through it 25,30,33 .
An increase in pressure decreases lycopene yield significantly at lower extraction time up to about 100 min. The trend, however, reverses to a significant increase with the increase in pressure.The increase of lycopene yield with pressure could have been due to increased solvent density of the supercritical CO 2 resulting in enhanced solubility of lycopene 25,31 . The response surfaces indicated the significant (p < 0.05) effect of both pressure and time individually as well as in conjugation over lycopene yield, which reinforces the significance of these factors found in ANOVA ( Table 2). The time of extraction is an indicative measure of the availability of the quantity of supercritical CO 2 for the extraction process. The completion of the extraction is adversely affected if the available supercritical CO 2 is a limiting factor. Nevertheless, beyond a certain point, any increase in extraction time could have a detrimental effect because of the overpowering of other factors such as temperature 24,27,28 . There is a significant (p < 0.05) role in both positive and negative terms in the extraction of lycopene from the matrix. The positive effect involves enhancing the extraction rate by solvent densification, whereas the negative effect of sample compaction is higher. Hence, an appropriate balance between the two parameters is of utmost importance for maximizing lycopene yields 23,25 .
Temperature. Temperature plays a vital role in any extraction process involving supercritical CO 2 . Increase in lycopene yield is observed with the increase in extraction temperature upto80°C, irrespective of pressure and Table 2. Analysis of variance (ANOVA) (A) response surface quadratic model for the lycopene yield from grapefruit and (B) Independent variables and their interactions for lycopene yield from grapefruit. www.nature.com/scientificreports/ time of extraction. This is because higher temperatures are known to enhance the solubility of solute, thereby increasing mass transfer of solute in the matrix [29][30][31] . However, increase in extraction temperature beyond 80 °C causes a significant reduction in the lycopene yieldfrom around 80% to about 30%. This can be explained by the loss of balance between the supercritical CO 2 density and solute vapour pressure 19,23,31 and lycopene degradation because of isomerization 24,27 . www.nature.com/scientificreports/ Flow rate. An increase in the flow rate of the supercritical CO 2 from 15 to 45 g/min was found to increase lycopene yield in the entire range of temperatures considered in the study. This can be attributed to enhanced dissolution rate of lycopene into the supercritical CO 2 because of universal concentration gradient phenomenon 24,25 . However, a further increase in the CO 2 flow rate beyond 45 g/min significantly (p > 0.05) reduced the lycopene extraction yield. The reduction in extraction yields may have been because of reduced interaction between the solute (grapefruit powder) and solvent (ethanol + supercritical CO 2 ) 28,30 . The reduced interaction is because of the solvent quickly passing around attributed to the solvent the sample matrix at a higher flow rate rather than defusing through it. This description is strengthened by the surface plot between extraction and the flow rate of supercritical CO 2 (Fig. 3). With the increase in extraction time and the flow rate of supercritical CO 2 , it was observed that the lycopene yields significantly increased. This increase is primarily because the solvent's improved dwell time with the solutes triggering the solvent's penetration into the grapefruit sample matrix which enables maximization of extraction of lycopene 32,33 . Taking into account the interaction effect of the independent parameters, up to an extraction time of 100 min, the lycopene yield decreases with increase in pressure. The trend however, reverses to a significant increase with the increase in pressures (Fig. 3b). The surface in fact clearly indicates the significant effect of both pressure and time individually as well as in conjugation over lycopene yieldwhich reinforces the significance of these factors found in ANOVA ( Table 2). The time of extraction is an indicative measure of the amount of supercritical CO 2 available for the extraction process. In case the available supercritical CO 2 is a limiting factor, the completeness of extraction is adversely affected. However, increase in time period beyond a point where the available supercritical CO 2 suffices the completeness of extraction, could lead to detrimental effect because of other controlling parameters like temperature 24,27,28 . Pressure level as described earlier could either play a significant role in aiding extraction by solvent densification or limit it because of sample compaction. Hence, an appropriate balance between the two is essential for maximization of lycopene yields.
Optimization of extraction condition. Second order response surface was fitted to the responses (lycopene extraction) obtained in all the thirty experimental runs and stationary point was obtained through canonical analysis using SAS software. The nature of stationary point was evaluated and found to be the point of maxima and therefore considered the optima. At the optimum conditions i.e. a combination of 305 bars of pressure, 70 °C temperature, flow rate of 35 g/min of supercritical CO 2 , and 135 min. of extraction time, 93% extraction efficiency of lycopene was achieved. In order to validate the same five additional experimental runs were conducted using the optimum conditions. The experimental yield obtained was 91.03 ± 1.86%indicating a good agreement between the predicted and observed extraction yields. This clearly established the authenticity of the statistical model in predicting lycopene extraction yield at any given experimental combination of independent variables considered for the study.

Conclusion
The results indicated the possibility of successfully achievingupto 93% extraction of lycopene from the grapefruit matrix using the optimum combinations of independent variables considered for the study. Statistical analysis revealed a non-significant effect of temperature, carbon dioxide (CO 2 ), and their interaction on the extraction yield of the lycopene. Whereas, pressure, extraction time, and their interaction, had a significant (p < 0.05) effect on lycopene extraction. Second-order polynomial showed high regression coefficients and a well-fitted model to the obtained experimental data.