Food Chemistry 166 (2015) 316–323

Contents lists available at ScienceDirect

Food Chemistry

journal homepage: www.elsevier .com/locate / foodchem

Simultaneous extraction of oil- and water-soluble phase from sunflowerseeds with subcritical water

http://dx.doi.org/10.1016/j.foodchem.2014.06.0250308-8146/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +386 2 22 94 463; fax: +386 2 2527 774.E-mail address: mojca.skerget@um.si (M. Škerget).

Matej Ravber, Zeljko Knez, Mojca Škerget ⇑University of Maribor, Faculty of Chemistry and Chemical Engineering, Smetanova 17, SI-2000 Maribor, Slovenia

a r t i c l e i n f o

Article history:Received 14 March 2014Received in revised form 2 June 2014Accepted 5 June 2014Available online 14 June 2014

Keywords:Sunflower oilSubcritical water extractionKineticsHydrothermal degradationHydrolysis

a b s t r a c t

In this study, the subcritical water extraction is proposed as an alternative and greener processingmethod for simultaneous removal of oil- and water-soluble phase from sunflower seeds. Extraction kinet-ics were studied at different temperatures and material/solvent ratios in a batch extractor. Degree ofhydrothermal degradation of oils was observed by analysing amount of formed free fatty acids and theirantioxidant capacities. Results were compared to oils obtained by conventional methods. Water solubleextracts were analysed for total proteins, carbohydrates and phenolics and some single products ofhydrothermal degradation.

Highest amount of oil was obtained at 130 �C at a material/solvent ratio of 1/20 g/mL after 30 min ofextraction. For all obtained oils minimal degree of hydrothermal degradation could be identified. Highantioxidant capacities of oil samples could be observed. Water soluble extracts were degraded at temper-atures P100 �C, producing various products of hydrothermal degradation.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

The aqueous oil extraction (AOE) is an alternative and greenerprocessing method for removal of oil from oilseeds. It offers manyadvantages compared to conventional organic solvent extractions(Rosenthal, Pyle, & Niranjan, 1996), one of them being the uniqueability of simultaneous extraction of oil- and water-soluble (usu-ally protein and carbohydrate) phase. In the past, AOE has success-fully been implemented on different types of oil sources e.g.peanuts (Cater, Rhee, Hagenmaier, & Mattil, 1974), coconuts(Cater et al., 1974), soybeans (Rosenthal, Pyle, & Niranjan, 1998),rapeseeds (Embong & Jelen, 1977), sunflower seeds (Hagenmaier,1974) etc. with reported oil and protein yields ranging up to 93%and 92% of total obtainable yields, respectively. Mostly, these stud-ies have been performed at temperatures lower than 100 �C, at dif-ferent oilseed/water ratios (up to 1/20 g/mL) and with appliedchanges in pH of medium (acidic and basic).

Although the obtained yields for the above mentioned sourcesseem potentially attractive, for commercial use and large indus-trial-scale production, there are still many things to be considered.Long required agitation times, formation of stable emulsions withrigorous mixing of media during extraction and difficult separationof phases, as well as required changes of pH of the media (addition

of acids or bases) have discouraged the further process design anddevelopment in the past (Rosenthal et al., 1996).

One solution that could potentially improve the above men-tioned difficulties is an increase of applied extraction temperature.It is a well known fact that generally by increasing the temperatureextraction rates can be increased to some extent. Until today AOEhas only been implemented at extraction temperatures lower thanthe boiling point of water. Only a few studies have been performedat temperatures higher than the boiling point of water, i.e. in thesubcritical region (Ndlela, Moura, Olson, & Johnson, 2012;Pourali, Salak Asghari, & Yoshida, 2009).

Subcritical water (SubCW) is a term commonly used for waterheated under pressure from its atmospheric boiling temperature(100 �C) to its supercritical point (374 �C). At these conditions thethermal motion of water molecules increases, markedly changingits properties. Unlike ambient water, the highly hydrogen-bondedstructure at subcritical conditions slowly starts to dissipate, result-ing in a decrease of permittivity (polarity), increase of diffusionrate and a decrease in viscosity and surface tension (Smith, 2002).

Another interesting property of SubCW is its increasing self-ionization at increasing temperature, meaning that water at theseconditions becomes more acidic, giving it a more hydrolythicnature.

Extractions applying SubCW differ quite significantly from con-ventional extraction methods. Firstly, they are known to be veryfast (Aliakbarian, Fathi, Perego, & Dehghani, 2012; Carr,Mammucari, & Foster, 2011; Singh & Saldaña, 2011), due to the

M. Ravber et al. / Food Chemistry 166 (2015) 316–323 317

mentioned changes in the physical properties. Also the decrease inpolarity gives SubCW a tendency for dissolving less polar com-pounds (Carr et al., 2011). Secondly, the mentioned hydrolythicnature of SubCW means that applying this extraction medium fornatural materials will result not only in water soluble extractsbut also in hydrolysed products of the extract (Fernández-Ponce,Casas, Mantell, Rodríguez, & Martínez de la Ossa, 2012; Ruen-ngam,Quitain, Tanaka, Sasaki, & Goto, 2012). Also the insoluble cellmaterial, normally comprised of numerous complex polymericstructures (proteins, polysaccharides etc.), can be simultaneouslyhydrolysed during extraction, producing various water solubleproducts (amino acids, sugars) thus increasing the overall extrac-tion yield. The destruction of the complex structures would resultin formation of less stable emulsions mentioned earlier, sincenormally these structures are the main cause for their stability(Sanguansri & Ann Augustin, 2010).

Furthermore, the increased acidity of the medium does notresult only in hydrolythic reactions but also in other hydrothermalreactions characteristic for SubCW, such as dehydration and decar-boxylation (Pavlovic, Knez, & Škerget, 2013). This means that theobtained hydrolysed products can react even further with thewater molecules resulting in a variety of other products of hydro-thermal degradation e.g. furfurals from carbohydrates. Some ofthese formed products can have a health diminishing effect, whenpresent in large concentration. Their effect on human healthshould therefore be studied prior to consumption of food product.

When applying the AOE at subcritical conditions (subcriticalwater extraction) the possibility of triglyceride hydrolysis to freefatty acids and glycerol also arises. Such a result would cause adecrease in oil quality, since more intensive down-stream process-ing (refining) would be required for removal of these compounds.Also, another factor which diminishes oil quality is the hydrother-mal degradation of naturally present vitamins and antioxidants e.g.phytosterols and tocopherols.

In this work, we propose the subcritical water extraction as afeasible processing method for removal of oil- and water-solublephase from sunflower seeds. Extraction kinetics of both phaseswere investigated at different extraction temperatures and mate-rial to water ratios and the obtained oil yields of the subcriticalwater extraction (SubWE) were compared to those obtained withthe standard Soxhlet extraction procedure. Degree of hydrothermaldegradation of oil extracts (OE) was checked, by determining (1)the free fatty acid composition and (2) the total antioxidant capac-ity of lipid soluble compounds and results were compared to theoil obtained with the Soxhlet extraction procedure. Degree ofhydrothermal degradation of water soluble extracts (WSE) waschecked by analyzing the content of (1) proteins, (2) carbohydratesand (3) phenolic compounds. WSE were also analysed for productsof hydrothermal degradation and the total antioxidant capacities ofwater soluble compounds.

2. Materials and methods

2.1. Chemicals and reagents

All reagents, standards and solvents were of analytical grade.Coomassie brilliant blue, albumin bovine serum (P96%), phenol,glucose (P98%) and sodium carbonate were purchased fromSigma–Aldrich (Slovenia). Oleic acid (P98%), linoleic acid (P98%),stearic acid (P98.5%), palmitic acid (P97%) and gallic acid(P98%) were purchased from Fluka (Germany). Sulfuric acid,Folin–Ciocalteu phenol reagent, acetic acid and ethanol were pur-chased from Merck (Germany). 5-Hydroxymethyl furfural(P98%), caffeic acid (P99%) and chlorogenic acid (P98%) werepurchased from Acros Organics (Belgium). Phosphoric acid waspurchased from Kemika (Croatia), hexane was purchased from

Carlo Erba (Italy) and methanol was purchased from J.T. Baker(Netherlands). Photochem� reagents and standards were purchasedfrom Chemass (Slovenia).

Helium 6.0 and nitrogen 5.0 were supplied from Messer(Slovenia).

2.2. Preparation of material

Dehulled sunflower seeds (Natura, Slovenia) were used for thisresearch work. The seeds were ground in a grinder (Bosch,Germany) prior to every experiment. Material humidity wasmeasured by a thermo-balance (Mettler-Toledo, Switzerland) andit was (3.68 ± 0.08)%. Total lipid content in seeds was determinedby the AOAC official method 948.22 (Venkatachalam & Sathe,2006) and it was (51.02 ± 1.95)%.

2.3. Extraction procedures

2.3.1. Subcritical water extractionFor extraction of grinded sunflower seeds with subcritical water

a 60 mL cylindrical stainless steel high-pressure batch extractor(Autoclave Engineers, USA) was used. Temperature regulationwas performed with a heating cable and stirring of the extractionmedia was performed by using a magnetic stirrer. The filled extrac-tor was purged three times with inert nitrogen gas to remove pres-ent atmospheric oxygen, which could cause oxidation of oil duringextraction. The applied extraction pressure for all extractions wasequal to 30 bar and was held constant throughout the extraction.

Kinetics of extractions were studied at four different tempera-tures (Te), namely 60 �C, 100 �C, 130 �C and 160 �C for extractiontimes (te) ranging from 5 min to 120 min at a material to solventratio (M/S) of 1/20 g/mL. The extraction kinetics at 130 �C were alsoobserved at the M/S ratio of 1/10 g/mL and 1/30 g/mL. The extrac-tion suspension was filtered and the obtained liquid extract alongwith 50 mL of hexane was introduced into a separation funnelwhich was then shaken rigorously for 3 min. The formed emulsionwas centrifuged (Eppendorf, Germany) at 11,000 rpm for 2 min inorder to separate the two phases, both of which were then col-lected and evaporated until dryness. Both oil (OE) and water solu-ble extract (WSE) samples were stored at �20 �C until further use.

2.3.2. Conventional extraction with the Soxhlet apparatusApproximately 10 g of ground sunflower seeds were placed in a

thimble, which was inserted into a Soxhlet apparatus andextracted with 200 mL of hexane (M/S = 1/20 g/mL). The extractionwas performed at normal boiling point for 4 h and afterwards thesolvent was evaporated until dryness. Obtained oil was stored at�20 �C until further use.

Kinetics of Soxhlet extraction were studied by collecting 1 mL ofsample solution after every solvent cycle (approximately every30 min) for 4 h. The solvent was evaporated from the collectedsample and the mass of the remaining oil was determined.

2.4. Analysis of extracts

2.4.1. Analysis of free fatty acidsThe extracted crude oil was analysed for the content of free

fatty acids (FFAs) by gas chromatography (Kotnik, Škerget, &Knez, 2006). The analyses were performed for linoleic, oleic, stearicand palmitic acids. The 6890 HP model (USA) consisted of flameionization detector (FID) with temperature set at 300 �C and capil-lary column (HP-FFAP 30 m � 0.25 mm � 0.25 lm). The oventime–temperature profile was as follows: 120 �C (1 min), 25 �Cper min to 180 �C (1 min), 5 �C per min to 220 �C (10 min), 5 �Cper min to 230 �C (30 min). The carrier gas was helium with totalflow through the column 64.0 mL/min. The samples were analysed

318 M. Ravber et al. / Food Chemistry 166 (2015) 316–323

in hexane and the quantification of single FFA was done fromcalibration curves obtained from standards. The ranges of the cal-ibration curves for linoleic, oleic, stearic and palmitic acids were61.2–611.7 lg/mL, 38.2–519.7 lg/mL, 45.1–451.1 lg/mL and85.4–854.6 lg/mL, respectively.

2.4.2. Antioxidant capacities of oil extracts and water soluble extractsAntioxidant capacities of lipid-soluble compounds (ACL) and

water soluble compounds (ACW) were determined with the photo-chemiluminescence (Photochem�, Analytik Jena AG, Germany)instrument. Solutions and reagents were prepared according tothe Photochem protocols for ACL (Analytik Jena AG., 2008) andACW (Analytik Jena AG., 2004). The antioxidant capacities of oilextracts (OE) and water soluble extracts (WSE) were expressed inmmol/L of Trolox equivalents and in mmol/g of Ascorbic Acidequivalents, respectively. Trolox and Ascorbic acid standards wereused to obtain the calibration curves.

2.4.3. Total protein contentTotal proteins concentration in water-soluble sunflower seed

extracts was determined with the Bradford colorimetric method(Bradford, 1976). Briefly, to a 1 mL aliquot of a protein solution1 mL of prepared Bradford reagent (100 mg of Coomassie Bluewas mixed in a 1000 mL glass flask with 50 mL of 95% ethanoland 100 mL 85% (v/v) phosphoric acid and diluted with deionisedwater) was mixed in a test tube and vortexed for 30 s before theabsorbance of the solution was measured at 595 nm on a UV–Visspectrophotometer (Varian, USA). A reference solution was pre-pared in identical manner as explained above, except that the1 mL aliquot of protein solution was replaced by deionised water.The quantification was done based on a calibration curve obtainedwith BSA (Albumin Bovine Serum). Total proteins content in watersoluble extracts (wBSA) was expressed in mg of BSA per g of extract(mg BSA/g ext.).

2.4.4. Total carbohydratesTotal carbohydrates concentration in water-soluble sunflower

seed extracts was determined with the phenol/sulfuric acid color-imetric method (DuBois, Gilles, Hamilton, Rebers, & Smith, 1956).Briefly, to a 2 mL aliquot of a carbohydrate solution 1 mL of 5%aqueous solution of phenol was mixed in a test tube. Subsequently,5 mL of concentrated sulfuric acid was added rapidly to the mix-ture. Test tubes were shaken in an ultrasonic bath for 10 min andthen left to stand at room temperature for 20 min for colour devel-opment. The absorbance of the acquired solution was measured at490 nm on a UV–Vis spectrophotometer. A reference solution wasprepared in identical manner as explained above, except that the2 mL aliquot of carbohydrate solution was replaced by deionisedwater. The quantification was done based on a calibration curveobtained with glucose. Total carbohydrates in water solubleextracts (wGLU) were expressed in mg of glucose per g of extract(mg GLU/g ext.).

2.4.5. Total phenolic compoundsTotal phenolic compounds in water-soluble sunflower seed

extracts were determined according to the Folin–Ciocalteu colori-metric method (Škerget et al., 2005). Briefly, to a 0.5 mL aliquotof water-soluble sample solution 2.5 mL of Folin–Ciocalteu reagent(diluted ten times with water) and 2 mL of 75 g/L Na2CO3 wereadded. The temperature of the solution was then elevated in awater bath at 50 �C for 5 min. After cooling at room temperaturefor 30 min, the absorbance of the solution was measured at760 nm by an UV–Vis spectrophotometer. A reference solutionwas prepared in identical manner as explained above, except thatthe 0.5 mL aliquot of extract solution was replaced by deionisedwater. The quantification was done by calibration curve prepared

with gallic acid. Total phenolics in water soluble extracts (wGA)were expressed in mg of gallic acid per g of extract (mg GA/g ext.).

2.4.6. Chromatographic analysis of WSEA high performance liquid chromatography (HPLC) method was

developed for determination of products of hydrothermal degrada-tion present in the water soluble sunflower seed extracts, namely5-hydroxymethyl furfural (5-HMF), 3-O-caffeoylquinic acid (chlor-ogenic acid) and its hydrolytic derivative caffeic acid. Chromato-graphic analysis was performed on a Varian system, equippedwith a Prostar 210 binary pump, a column heater, a Prostar 410autosampler and a Prostar 310 variable wavelength detector(VWD), connected to a ChemStation software. The separation wasachieved on chromatography column Agilent Zorbax SB-C18150 � 4.6 mm with 5 lm particle size at 25 �C and a flow rate of1 mL/min. The mobile phase consisted of two solvents, (A) 2% ace-tic acid in water, and (B) methanol. The gradient of the solventswas: 0 min 15% B, 20 min 35% B, 22 min 15% B. Detection of thecompounds was performed at 280 nm and the quantification wasdone using calibration curves. Amount of a compound wasexpressed in mg of compound per g of extract (mg/g ext.).

2.5. Statistical analysis

Experimental results were expressed as means ± standard devi-ation (SD) of two parallel experiments (n = 2). Each data point rep-resents the average of at least three measurements and the relativestandard deviation between measurements was 1%.

3. Results and discussion

3.1. Extraction kinetics of oil

The modern solvent-based oil extraction process usually con-sists of extraction by successive countercurrent extractions withhexane of the previously mechanically ruptured oleaginous mate-rial (Rosenthal et al., 1996). Due to many concerns regarding foodsafety when applying hexane in food processing technologies itsuse decreased drastically over the years. Hence it is important, thatfor newly developed oil-processing technologies the use of hexaneis minimised or possibly even completely avoided. Although in thisstudy the use of hexane was not completely avoided due to theneed of high purity oil for analysis, for large scale application ofSubWE the separation of the water and oil phase would probablybe sufficient just by using centrifugation (Rosenthal et al., 1996).

The subcritical water extraction (SubWE) kinetics of sunfloweroil and the comparison to extraction kinetic curve obtained bySoxhlet extraction are presented in Fig. 1a. Results show that forthe SubWE the highest oil yields (gOE) are achieved at an extractiontemperature (Te) of 130 �C and a material to solvent ratio (M/S) of1/20 g/mL, with a maximum yield of 44.3 ± 0.3% after 30 min ofextraction, giving comparable results to those obtained after 4 hby the Soxhlet (46.2 ± 0.7%). At 160 �C a similar maximum(43.9 ± 0.3%) is achieved for SubWE but after 2 h of extraction. Atother applied temperatures and M/S ratios lower yields areobtained. Increasing temperature hence improves the extractionyield of SubWE as was predicted. It can be also observed that atall conditions investigated, except at 160 �C, the extraction ratesare much higher compared to Soxhlet extraction. The highest oilyields are obtained in te 6 30 min, while in the case of Soxhletextraction the time needed is 4 h.

Although it would be expected that increasing temperaturewould have a direct positive effect on extraction yield due tohigher solubility of non-polar phase, the slower extraction kineticsat 160 �C can be explained with lower solubility of protein phase at

0 30 60 90 120 150 180 210 240

05

101520253035404550

0 15 30 45 60 75 90 105 120

te,Soxhlet [min]

OE

[wt.

%]

te, SubWE [min]

(a)

60 °C 100 °C 130 °C 1/10 g/mL130 °C 1/20 g/mL 130 °C 1/30 g/mL 160 °CSoxhlet

0

5

10

15

20

25

30

0 15 30 45 60 75 90 105 120

WSE

[wt.

%]

te [min]

(b)

60 °C 100 °C 130 °C 1/10 g/mL130 °C 1/20 g/mL 130 °C 1/30 g/mL 160 °C

Fig. 1. Extraction kinetics of oil extract (OE) obtained using subcritical water at different extraction conditions and comparison of extraction kinetics of OE using the Soxhletextraction (a) and extraction kinetics of water soluble extract (WSE) obtained using subcritical water at different extraction conditions (b).

M. Ravber et al. / Food Chemistry 166 (2015) 316–323 319

the applied temperature. In the detailed review by Rosenthal et al.(1996) it was reported that the amount of oil obtained from oilseeds using water as extraction medium depends mostly fromthe amount of cotyledon cell wall rupture which is done by eitherflaking or grounding of seeds. Cotyledon cells present in sunflower(and many other) seeds contain most of the oil and protein phasepresent within the seed, of which the protein phase enclaves theoil. Rupturing the protein structure or its removal by extractionis therefore essential for high extraction yields of oil, since oil isthen allowed to diffuse into the extraction medium. At 160 �Cthe dielectric constant of SubCW is approximately 42 and at130 �C it is approximately 49, which seems to decrease the abilityof the medium to extract the protein phase, consequently notallowing more oil to be released. Based on these observations, fur-ther increase in Te could therefore have a negative effect on extrac-tion kinetics. On the other hand, higher temperature would resultin more protein structure rupture, which in this case would havethe same positive effect on the oil yield but at the same time couldcause more structural damage to the other components in thematerial.

From Fig. 1a it can be observed that besides Te, M/S ratio alsohas a significant effect on total extractable oil. Comparing thekinetic curves obtained at 130 �C and at different M/S ratios, thehighest yield of oil is obtained at M/S = 1/20 g/mL. It seems thatat M/S = 1/10 g/mL not enough water is provided for the removalof the protein phase, which consequently resulted in lower oilyields. Interestingly, at M/S = 1/30 g/mL a decrease of total oil yieldcan be observed, although a higher seeds to water ratio was pro-vided, than the optimal M/S ratio of 1/20 g/mL.

Material to solvent ratio is an important process parameter inthe SubWE. Higher M/S ratios require more water to be com-pressed and heated-up, which consequently increase the processescosts significantly. It is therefore of great importance that the M/Sratio is as small as possible but at the same time should be highenough to provide the highest possible extraction yield.

3.2. Extraction kinetics of water soluble extract

The extraction kinetics of water soluble extracts (WSE) usingthe subcritical water extraction are shown in Fig. 1b. Looking atthe kinetic curves we can presume that hydrolysis of the sunflowerseed material occurs at temperatures higher than 100 �C, sinceyield of WSE keep increasing over longer periods of time. The high-est WSE extraction rate is obtained at 160 �C, where the maximalyield of 30.0 ± 0.8% is reached after 60 min of extraction. Compar-ing the two lower temperature curves (60 �C and 100 �C), it can be

observed that higher yields of WSE are achieved at 60 �C. This isprobably due to higher solubility of WSE at lower temperature.

M/S ratio seems to have a lower impact on extraction yield ofWSE than Te, since identical curves were obtained at 130 �C forall M/S, with the only exception for the M/S = 1/10 g/mL curve,which has a bit lower values of extraction yield throughoutthe whole process. It seems that for the applied M/S range, higherM/S increases the overall yield of WSE, but only by moderateamounts.

Increase of WSE yield at Te P 130 �C can be explained by twooccurring hydrolysis reactions: (1) hydrolysis of protein (produc-ing water-soluble peptides and amino acids) or (2) hydrolysis ofcarbohydrates (producing water-soluble oligosaccharides and sim-ple sugars). Both of these reactions can contribute to the totalamount of WSE produced, but it is possible that one contributesmore than the other, since hydrolysis of different types of bondsis preferred at different conditions. For instance, in the articlereported by Rogalinski, Liu, Albrecht, and Brunner (2008), whichstudied the hydrolysis kinetics of cellulose, starch and BSA in Sub-CW from 210 �C to 310 �C, it was proven that the two carbohydratebiopolymers were much more susceptible towards hydrolytic deg-radation at the studied reaction conditions than the protein one.Peptide bonds present within the protein structure are thereforemuch more stable than glycosidic bonds present within the carbo-hydrates. Also, as already mentioned, other products may havebeen formed during extraction.

3.3. Content of free fatty acids in oil samples

Degree of hydrothermal degradation of oil samples obtained bySubWE and Soxhlet extraction was checked by analyzing their con-tent of free fatty acids (FFAs). Four characteristic FFAs, typicallyfound in sunflower oil, namely linoleic, oleic, stearic and palmiticacid, were observed. Results are presented in Table 1.

It can be seen that in the obtained samples of sunflower oil oleicand linoleic acids are present in free form. Stearic and palmiticacids were not detected in the free form. Oil samples obtained bySubWE contain small amounts of total FFAs. Smallest amountscan be observed at 60 �C (<2.6 wt.%), followed by 100 �C (<3.8wt.%), 130 �C (<4.4 wt.%) and 160 �C (<5.1 wt.%). Oil samplesobtained by Soxhlet extraction contained less than 3.5 wt.% of totalFFAs. SubWE therefore produces higher amounts of FFAs than theSoxhlet extraction at temperatures higher than 60 �C. Generallycrude sunflower oil contains approximately 0.5 or more % of totalFFAs (Campbell, 1983) but according to the American Oils and FatsAssociation (Grompone, 2011) a maximum of 2% of total FFA is

Table 1Amount of free fatty acids (FFAs) in oil extracts obtained using subcritical water at different extraction conditions and Soxhlet extraction.

Te (�C) te (min) M/S (g/mL) FFAs mg/g oil Te (�C) te (min) M/S (g/mL) FFAs mg/g oil

Linoleic acid Oleic acid Linoleic acid Oleic acid

60 5 1/20 13.17 ± 0.19 12.30 ± 0.08 130 5 1/20 19.93 ± 0.28 18.14 ± 0.2310 1/20 13.34 ± 0.09 12.57 ± 0.04 10 1/20 20.17 ± 0.07 18.92 ± 0.0330 1/20 12.77 ± 0.07 11.99 ± 0.08 30 1/20 21.14 ± 0.12 19.85 ± 0.0660 1/20 12.97 ± 0.05 12.17 ± 0.07 60 1/20 21.60 ± 0.09 20.29 ± 0.13120 1/20 13.15 ± 0.07 12.29 ± 0.03 120 1/20 22.47 ± 0.18 21.08 ± 0.12

100 5 1/20 18.95 ± 0.07 17.82 ± 0.05 130 5 1/30 20.68 ± 0.22 19.47 ± 0.0510 1/20 19.07 ± 0.13 17.64 ± 0.06 10 1/30 21.14 ± 0.10 19.84 ± 0.0630 1/20 18.81 ± 0.07 17.71 ± 0.03 30 1/30 22.44 ± 0.18 21.22 ± 0.1460 1/20 18.61 ± 0.30 17.39 ± 0.11 60 1/30 23.21 ± 0.08 21.90 ± 0.05120 1/20 19.43 ± 0.12 18.35 ± 0.08 120 1/30 22.95 ± 0.05 21.69 ± 0.10

130 5 1/10 18.92 ± 0.12 17.59 ± 0.17 160 5 1/20 23.53 ± 0.22 22.08 ± 0.1310 1/10 19.59 ± 0.09 18.47 ± 0.05 10 1/20 24.19 ± 0.21 22.77 ± 0.0330 1/10 20.47 ± 0.12 19.25 ± 0.09 30 1/20 24.37 ± 0.17 22.76 ± 0.1160 1/10 21.35 ± 0.08 20.11 ± 0.10 60 1/20 25.26 ± 0.22 23.72 ± 0.10120 1/10 22.09 ± 0.15 20.81 ± 0.09 120 1/20 25.82 ± 0.19 24.26 ± 0.19

Soxhlet 240 1/20 17.23 ± 0.26 16.08 ± 0.35

320 M. Ravber et al. / Food Chemistry 166 (2015) 316–323

specified. Higher values of FFA result in rancidity of the oil andshould therefore be removed (refined) prior to consumption. Fromthe obtained results it can hence be concluded that further refiningis mandatory for all oils obtained by SubWE and Soxhlet extractionobtained oils.

For oil samples obtained at Te 6 100 �C no correlation of FFAyield with te can be identified, since composition of FFAs doesnot change over time and remains constant after first 5 min ofextraction. Only at temperatures higher than 100 �C a mildincrease of FFAs over time can be noticed and even at 160 �C prac-tically negligible increase of FFAs can be observed. Nevertheless asmall increase of overall FFAs is present at Te P 130 �C, althoughthe overall content of FFAs in the samples does not change morethan for 2.5 wt.% throughout the whole temperature range. Sincenaturally fatty acids in sunflower oil are mostly bonded as triglyc-erides, their low presence in the obtained OEs is not unexpected,however due to the hydrolythic nature of SubCW a concern of tri-glyceride stability in this medium exists. Although hydrolysis ofester bonds in triglycerides with SubCW is possible (Alenezi,Leeke, Santos, & Khan, 2009; Milliren, Wissinger, Gottumukala, &Schall, 2013), looking at the results it seems that polar–non-polarinteractions are still too repulsive at the temperatures applied inthis study, causing hydrophobicity of the compounds, thereforenot allowing the hydrolysis reaction to take place. At much highertemperatures (>250 �C) solubility of fats increases to such a level,that hydrolysis can occur at much higher intensities.

When regarding the stability of FFA in SubCW, the studyreported by Shin, Ryu, Park, & Bae (2012) claims that FFA are stablein SubCW at temperatures below 300 �C, i.e. no products of

0

50

100

150

200

250

300

0 15 30 45 60 75 90 105 120

wBS

A [m

g BS

A/g

ext.]

te [min]

(a)

60 °C 100 °C 130 °C 1/10 g/mL130 °C 1/20 g/mL 130 °C 1/30 g/mL 160 °C

0

50

100

150

200

0 15 30 45 60 75 90 105 120

wG

LU[m

g G

LU/g

ext

.]

te [min]

(b)

60 °C 100 °C 130 °C 1/10 g/mL130 °C 1/20 g/mL 130 °C 1/30 g/mL 160 °C

Fig. 2. Total proteins (wBSA) (a) and carbohydrates (wGLU) (b) in water soluble e

hydrothermal degradation are formed below this temperature. Wecan therefore safely assume that the quality of the oil was pre-served and the triglycerides were not hydrolysed.

3.4. Total proteins and carbohydrates in water soluble extracts

In this study, the determination of total proteins present in WSEwas performed using the Bradford colorimetric method. In the pastit has been proven that the interference of carbohydrates presentin the sample matrix significantly deviates the actual absorbanceof the proteins, resulting in less accurate results if not primarilyremoved (Banik, Pal, Ghorai, Chowdhury, & Khowala, 2009). Sincethe primary aim of these preliminary studies was to observe thehydrothermal degradation of proteins in SubCW and not quantifythe exact amount of proteins in extract, carbohydrates present inWSE were not primarily precipitated and were analysed directlyusing the above mentioned method.

In contrast to the minimally hydrolysed triglycerides, total pro-teins (Fig. 2a) and carbohydrates (Fig. 2b) present in WSE show amuch more different course.

From Fig. 2a it can be observed that after 5 min of extractionagain the lower temperature kinetic curves (60 �C and 100 �C) showno dependence on extraction time. The highest amounts of watersoluble proteins are obtained at 60 �C equaling to approximately30 wt.%. At temperatures higher than 100 �C overall protein yieldin WSE decreases significantly with increasing Te and at 160 �Cthe concentration of proteins (wBSA) even starts to decrease withte. The decrease of overall wBSA with increasing Te is probably a con-sequence of lower solubility of protein at higher temperatures,

xtracts (WSE) obtained using subcritical water at different extraction conditions.

M. Ravber et al. / Food Chemistry 166 (2015) 316–323 321

while the gradual decrease with extraction time at 160 �C is possi-bly due to hydrolysis of proteins. M/S ratio seems to have practicallyno influence on wBSA.

For total carbohydrates a quite similar pattern can be observed.Again, for the lower temperature curves (60 �C and 100 �C), noinfluence of te can be observed, since carbohydrate concentration(wGLU) remains constant after 5 min of extraction. Oppositely ahigher yield of total carbohydrates can be observed at Te P 100 �C,which indicates a higher carbohydrate solubility at higher temper-atures. At 130 �C wGLU starts to decrease with te and the decrease isthe highest at 160 �C. This decrease of wGLU not only indicates adecrease of carbohydrates in the sample but also a decrease ofpresent furfurals. Furfurals are namely the main constituentsformed in the phenol/sulfuric acid colorimetric method appliedfor determination of total carbohydrates. Therefore it can be pre-sumed that products other than furfurals have formed duringextraction.

Since the carbohydrate profile of sunflower kernels consistsmainly from hexose derivatives (glucose, sucrose, cellulose etc.)(Grompone, 2011) we can assume that after dehydration of thesugars mostly 5-hydroxymethyl furfural (5-HMF) is formed(Hayes, Fitzpatrick, Hayes, & Ross, 2008). Afterwards, further dehy-dration of formed 5-HMF is possible in a second step, which pro-duces levulinic acid and formic acid. Due to the fact that at130 �C and 160 �C the content of carbohydrates in extractsdecreases with te it can be assumed that 5-HMF has been furtherdegraded to organic acids.

When comparing the degree of hydrothermal degradation ofproteins and carbohydrates (Fig. 2a and b) it can be observed thatcarbohydrates are more susceptible to SubCW induced reactionsthan proteins in the applied temperature range. As already men-tioned in Subsection 3.3, this difference could be a consequenceof unequal solubilities of the two materials. Furthermore, the reac-tivity could also be dependent on the ability of the water moleculesto reach the reaction site i.e. on degree of crystallinity of the twomaterials etc. Based on these observations, it can be concluded thathydrothermal degradation reactions of SubCW can indeed have amore selective behaviour towards different types of chemical com-pounds as was assumed.

3.5. ACL of oil samples and ACW of water soluble samples

Antioxidant capacities of lipid soluble compounds in oil sam-ples detected by the ACL photochemiluminescence method arereported in Table 2. The obtained results show a similar course

Table 2The antioxidant capacity of lipid soluble compounds (ACL) in oil extracts obtained using suTrolox equivalents and the antioxidant capacity of water soluble compounds (ACW) in wateAscorbic Acid equivalents.

Te (�C) te

(min)M/S(g/mL)

ACL Trolox eq.(mmol/L)

ACW Ascorbic Acid eq.(mmol/g)

60 5 1/20 1.08 ± 0.04 1.48 ± 0.1310 1/20 1.21 ± 0.09 1.36 ± 0.0130 1/20 1.24 ± 0.03 1.52 ± 0.0560 1/20 1.49 ± 0.08 1.51 ± 0.01120 1/20 1.88 ± 0.05 1.32 ± 0.07

100 5 1/20 2.72 ± 0.16 3.33 ± 0.0710 1/20 2.59 ± 0.19 3.47 ± 0.0330 1/20 2.64 ± 0.15 3.37 ± 0.0660 1/20 2.64 ± 0.09 3.51 ± 0.11120 1/20 2.49 ± 0.10 3.19 ± 0.09

130 5 1/10 3.10 ± 0.14 1.85 ± 0.1310 1/10 3.49 ± 0.31 1.30 ± 0.0230 1/10 5.04 ± 0.36 0.80 ± 0.0460 1/10 6.24 ± 0.30 0.79 ± 0.07120 1/10 7.79 ± 0.24 0.80 ± 0.01

Soxhlet 240 1/20 1.37 ± 0.06

as the results presented in Fig. 1b. The amount of ACL Trolox equiv-alents (cTRX) increases at 130 �C with te for the whole extractiontime range whereas at 160 �C it only increases until te = 60 min,then slowly starts to decrease, which is probably a consequenceof hydrothermal degradation of antioxidants. At 60 �C and 100 �CcTRX stays constant throughout the whole te range after 5 min ofextraction. Only little increase of cTRX over time can be noticedfor Te = 60 �C. M/S ratio seems to have no effect on cTRX at theapplied Te. Compared to the cTRX of the oil obtained by Soxhletextraction (1.37 ± 0.06 mmol/L) significantly higher cTRX can beobserved for most oil samples extracted with SubCW, with a max-imum ACL value of 8.81 ± 0.04 mmol/L obtained at 160 �C after60 min of extraction.

One explanation possible for this increase of ACL is the hydro-thermal reactions of phenolic compounds present in sunflowerseeds (Karamac, Kosinska, Estrella, Hernández, & Dueñas, 2012).These compounds although generally more soluble in polar sol-vents due to their natural occurency as glycosides or quinic acidesters (chlorogenic acid), hydrolyse under the conditions of SubCWand consequently free phenolic compounds (caffeic acid, ferulicacid etc.) are formed, which generally have low solubility in waterat ambient conditions (Mota, Queimada, Pinho, & Macedo, 2008).These compounds could therefore have been extracted with hex-ane into the oil extract (OE), consequently increasing the overallACL of the obtained oil samples. Another possible explanation forthe cTRX increase is the hydrolysis of fat-soluble glycosides (ster-ols). They are generally present in natural sources (Phillips,Ruggio, & Ashraf-Khorassani, 2005) and their free forms usuallyhave higher antioxidant capacities.

When looking at the ACW of the water soluble extracts (Table 2)it can be observed that compared to the cTRX of the oil samples, theamount of ACW Ascorbic Acid equivalents (cAA) in WSE show analmost complementary relation. This relationship confirms theexplanation of hydrolysis of phenolic glycosides and quinic acidesters and the extraction of the obtained aglycones by hexane sta-ted above. Results show that at 130 �C cAA decreases with te. At160 �C, primarily cAA also decreases with extraction time but after30 min of extraction it suddenly starts to increase, which could bea consequence of hydrothermal degradation of antioxidants that at130 �C are stable enough to remain unchanged in the appliedextraction time period but start to decompose at 160 �C afterlonger exposure times. It is well possible that these newly formedantioxidants have better solubility in the aqueous phase, thereforecould not have been extracted to the oil phase in the phase separa-tion step.

bcritical water at different extraction conditions and Soxhlet extraction procedure inr soluble extracts obtained using subcritical water at different extraction conditions in

Te

(�C)te

(min)M/S(g/mL)

ACL Trolox eq.(mmol/L)

ACW Ascorbic Acid eq.(mmol/g)

130 5 1/20 2.70 ± 0.14 2.22 ± 0.0510 1/20 3.30 ± 0.13 1.48 ± 0.1330 1/20 4.65 ± 0.05 1.27 ± 0.0660 1/20 6.26 ± 0.33 1.12 ± 0.01120 1/20 7.44 ± 0.11 0.60 ± 0.01

130 5 1/30 3.24 ± 0.06 2.00 ± 0.1610 1/30 3.76 ± 0.10 1.58 ± 0.1430 1/30 4.81 ± 0.18 1.18 ± 0.0160 1/30 6.44 ± 0.09 0.65 ± 0.01120 1/30 7.64 ± 0.16 0.42 ± 0.00

160 5 1/20 4.22 ± 0.17 0.76 ± 0.0110 1/20 5.98 ± 0.02 0.66 ± 0.0330 1/20 8.38 ± 0.04 0.53 ± 0.0660 1/20 8.81 ± 0.04 0.89 ± 0.02120 1/20 7.82 ± 0.28 1.51 ± 0.05

322 M. Ravber et al. / Food Chemistry 166 (2015) 316–323

The effect of M/S ratio on the amount of cAA seems to be verymodest in the studied range. The lowest decrease can be observedfor M/S = 1/10 g/mL. At lower temperatures cAA does not changeover te. The highest amount of cAA is observed at Te = 100 �C andis approximately 3.51 ± 0.11 mmol/g. The higher values of ACWat lower temperatures are probably due to the presence of non-hydrolysed glycosides and quinic acid esters in the samples. Also,these compounds are generally more soluble in water at ambientconditions.

3.6. Total phenolics of water soluble samples

Total phenolic compounds detected by the Folin–Ciocalteu col-orimetric method are presented in Table 3. As the antioxidantcapacity of WSE it can be observed that total phenolics (wGA) fol-low a similar trend. Results show that at 60 �C and 100 �C wGA

remains constant for both temperatures over the whole extractiontime period. Also, at these temperatures, the highest amounts ofwGA can be observed, being higher than 115 mg/g extract for thesamples obtained at 100 �C and higher than 99 mg/g extract forthe samples obtained at 60 �C. At 130 �C the content of total phen-olics similarly as ACW decreases with te for all M/S ratios, whereasat 160 �C a decrease of wGA can be observed only until 60 min ofextraction. Afterwards, an increase can be observed, whichindicates, as mentioned earlier in Subsection 3.5, that other typesof phenolics (antioxidants) could have been formed due to hydro-thermal degradation.

3.7. HPLC analysis of WSE

Degree of hydrothermal degradation in WSE was assessed byanalyzing the composition of two phenolic compounds typicallypresent in sunflower seeds, namely chlorogenic acid (CHA) and

Table 3Amount of chlorogenic acid (wCHA), caffeic acid (wCA), total phenolics (wGA) and 5-HMF (w5

extraction conditions.

Te (�C) te (min) M/S (g/mL) wCHA (mg/g ext.)

60 5 1/20 80.32 ± 0.0910 1/20 78.97 ± 0.2930 1/20 79.09 ± 0.3160 1/20 80.60 ± 0.09120 1/20 79.03 ± 0.78

100 5 1/20 82.07 ± 0.8110 1/20 66.46 ± 0.2230 1/20 49.22 ± 0.3660 1/20 39.51 ± 0.12120 1/20 34.16 ± 0.48

130 5 1/10 55.72 ± 0.0510 1/10 36.61 ± 0.0630 1/10 25.73 ± 0.0560 1/10 22.82 ± 0.30120 1/10 19.08 ± 0.35

130 5 1/20 56.63 ± 0.0510 1/20 37.03 ± 0.0630 1/20 26.12 ± 0.1360 1/20 22.82 ± 0.32120 1/20 19.44 ± 0.36

130 5 1/30 56.97 ± 0.0710 1/30 36.97 ± 0.0330 1/30 26.09 ± 0.1560 1/30 22.74 ± 0.24120 1/30 19.42 ± 0.38

160 5 1/20 27.63 ± 0.6610 1/20 21.21 ± 0.0330 1/20 14.04 ± 0.0760 1/20 10.73 ± 0.04120 1/20 7.97 ± 0.04

a Not detected.

one of its hydrolythic derivatives – caffeic acid (CA). Also, anotheraspect of hydrothermal degradation in WSE was studied, namelythe decomposition (dehydration) of sugars to 5-HMF by measuringamount of formed 5-HMF during extraction.

Amounts of chlorogenic acid (wCHA), caffeic acid (wCA) and5-hydroxymethyl furfural (w5-HMF) present in the WSE are pre-sented in Table 3. Compared to the total phenolics content, it canbe observed that CHA is the main phenolic compound in WSE,however its content quickly starts to decrease with te at tempera-tures P100 �C. It seems that the ester bond present in CHA (unlikethe ones present in the triglycerides) is not very stable in SubCW atthese conditions. At the same time the increase of CA with te can beobserved, what indicates that the derivative of CHA (CA) is formedduring extraction. Highest decomposition rate of CHA can beobserved at 160 �C, with the lowest obtained yield of wCHA afterte = 5 min, which is 27.63 ± 0.66 mg/g ext. Although we wouldexpect CA to have a similar rate of formation as the decompositionrate of CHA, we can observe that this is not the case. More CHA isdecomposed with te than CA is formed. Although furtherhydrothermal degradation of CA is likely to occur at 160 �C, a morefeasible explanation of this anomality is the previously mentionedextraction of phenolics (CA) to OE in the phase separation step.CHA could probably not be extracted with non-polar phase (hex-ane) since its composition in WSE at 60 �C does not change ovet te.

From Table 3 it can be also observed that 5-HMF is formed dur-ing extraction at the studied extraction conditions. It seems thatformation of 5-HMF occurs at Te P 100 �C, whereas at 60 �C no5-HMF is formed. Interestingly, the highest amounts of 5-HMF areobtained at 100 �C, whereas at higher temperatures lower amountsare obtained. Also, at 130 �C and 160 �C a decrease of 5-HMF withte can be observed. At 130 �C the decrease of 5-HMF is visible after30 min of extraction for all M/S ratios and at 160 �C the decreasecan already be observed after 5 min of extraction. This decrease

-HMF) in water soluble extracts obtained with subcritical water extraction at different

wCA (mg/g ext.) wGA (mg/g ext.) w5-HMF (mg/g ext.)

NDa 100.25 ± 0.02 NDND 100.66 ± 0.15 NDND 100.29 ± 0.31 NDND 99.47 ± 0.08 NDND 103.29 ± 0.15 ND0.72 ± 0.02 115.57 ± 0.03 1.22 ± 0.010.87 ± 0.02 115.84 ± 0.05 2.21 ± 0.011.96 ± 0.01 115.62 ± 0.05 5.91 ± 0.013.27 ± 0.03 117.19 ± 0.30 8.05 ± 0.034.61 ± 0.02 119.74 ± 0.09 8.42 ± 0.051.93 ± 0.03 102.91 ± 0.02 5.30 ± 0.012.59 ± 0.02 101.71 ± 0.02 5.55 ± 0.014.43 ± 0.04 95.13 ± 0.04 5.77 ± 0.045.11 ± 0.11 89.05 ± 0.10 5.21 ± 0.044.76 ± 0.08 79.22 ± 0.12 4.32 ± 0.061.93 ± 0.03 97.26 ± 0.01 4.99 ± 0.012.62 ± 0.02 86.54 ± 0.01 5.67 ± 0.014.60 ± 0.01 76.88 ± 0.09 5.88 ± 0.035.31 ± 0.05 75.68 ± 0.27 5.28 ± 0.044.95 ± 0.06 79.21 ± 0.03 4.49 ± 0.051.85 ± 0.02 102.59 ± 0.45 4.65 ± 0.012.46 ± 0.01 99.64 ± 0.12 5.38 ± 0.064.47 ± 0.03 94.31 ± 0.39 5.53 ± 0.065.10 ± 0.08 86.30 ± 0.09 4.97 ± 0.034.76 ± 0.11 78.93 ± 0.08 4.18 ± 0.024.16 ± 0.18 97.26 ± 0.03 5.48 ± 0.164.04 ± 0.00 86.54 ± 0.05 4.42 ± 0.002.18 ± 0.01 76.88 ± 0.06 2.94 ± 0.020.91 ± 0.00 75.68 ± 0.19 2.18 ± 0.010.38 ± 0.01 79.21 ± 0.10 1.51 ± 0.05

M. Ravber et al. / Food Chemistry 166 (2015) 316–323 323

in concentration could indicate the previously mentioned furtherhydrothermal degradation of 5-HMF to other products, such asorganic acids (levulinic acid and formic acid).

4. Conclusions

Subcritical water extraction was applied as an alternative andgreener method for simultaneous removal of oil and water solubleextract (WSE) from sunflower kernels. Extraction kinetics anddegree of hydrothermal degradation was studied. Results showthat optimal extraction yield of oil was obtained at 130 �C andM/S = 1:20 g/mL after 30 min. Low presence of oil hydrolysis couldbe observed. WSE was found to be more susceptible towardshydrothermal degradation of which carbohydrates showed thelowest stability. Hydrolysis of ester and glycoside bonded antioxi-dants occurred, which produced oil with much higher antioxidantcapacities than oil extracted using the Soxhlet method.

Acknowledgements

Authors are grateful to the Slovenian Ministry of High Educa-tion, Science and Technology for the financial support of this work.This paper was produced within the framework of the operationentitled ‘‘Centre of Open innovation and ResEarch of Universityof Maribor (CORE@UM)’’.

References

Alenezi, R., Leeke, G. A., Santos, R. C. D., & Khan, A. R. (2009). Hydrolysis kinetics ofsunflower oil under subcritical water conditions. Chemical Engineering Researchand Design, 87(6), 867–873.

Aliakbarian, B., Fathi, A., Perego, P., & Dehghani, F. (2012). Extraction of antioxidantsfrom winery wastes using subcritical water. The Journal of Supercritical Fluids,65, 18–24.

Analytik Jena AG. (2004). Determination of water-(ACW) and lipid-(ACL) solubleantioxidative capacity in plant extract.

Analytik Jena AG. (2008). Determination of lipid (ACL) soluble antioxidativecapacity in macadamia nut oil.

Banik, S. P., Pal, S., Ghorai, S., Chowdhury, S., & Khowala, S. (2009). Interference ofsugars in the Coomassie Blue G dye binding assay of proteins. AnalyticalBiochemistry, 386(1), 113–115.

Bradford, M. M. (1976). A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein-dye binding.Analytical Biochemistry, 72(1–2), 248–254.

Campbell, E. J. (1983). Sunflower oil. Journal of the American Oil Chemists’ Society,60(2), 387–392.

Carr, A. G., Mammucari, R., & Foster, N. R. (2011). A review of subcritical water as asolvent and its utilisation for the processing of hydrophobic organiccompounds. Chemical Engineering Journal, 172(1), 1–17.

Cater, C., Rhee, K., Hagenmaier, R., & Mattil, K. (1974). Aqueous extraction – analternative oilseed milling process. Journal of the American Oil Chemists’ Society,51(4), 137–141.

DuBois, M., Gilles, K. A., Hamilton, J. K., Rebers, P. A., & Smith, F. (1956). Colorimetricmethod for determination of sugars and related substances. AnalyticalChemistry, 28(3), 350–356.

Embong, M. B., & Jelen, P. (1977). Technical feasibility of aqueous extraction ofrapeseed oil-a laboratory study. Canadian Institute of Food Science andTechnology Journal, 10(4), 239–243.

Fernández-Ponce, M. T., Casas, L., Mantell, C., Rodríguez, M., & Martínez de la Ossa,E. (2012). Extraction of antioxidant compounds from different varieties ofMangifera indica leaves using green technologies. The Journal of SupercriticalFluids, 72, 168–175.

Grompone, M. A. (2011). Sunflower Oil. In Vegetable Oils in Food Technology, (pp.137–167). Wiley-Blackwell.

Hagenmaier, F. (1974). Aqueous processing of full-fat sunflower seeds: Yieldsof oil and protein. Journal of the American Oil Chemists’ Society, 51(10),470–471.

Hayes, D. J., Fitzpatrick, S., Hayes, M. H. B., & Ross, J. R. H. (2008). The Biofine Process– Production of Levulinic Acid, Furfural, and Formic Acid from LignocellulosicFeedstocks. In B. Kamm, P. R. Gruber, & M. Kamm (Eds.). Biorefineries-IndustrialProcesses and Products (Vol. 1, pp. 139–164). Weinheim: Wiley-VCH VerlagGmbH.

Karamac, M., Kosinska, A., Estrella, I., Hernández, T., & Dueñas, M. (2012).Antioxidant activity of phenolic compounds identified in sunflower seeds.European Food Research and Technology, 235(2), 221–230.

Kotnik, P., Škerget, M., & Knez, Z. (2006). Kinetics of supercritical carbon dioxideextraction of borage and evening primrose seed oil. European Journal of LipidScience and Technology, 108(7), 569–576.

Milliren, A. L., Wissinger, J. C., Gottumukala, V., & Schall, C. A. (2013). Kinetics ofsoybean oil hydrolysis in subcritical water. Fuel, 108, 277–281.

Mota, F. L., Queimada, A. J., Pinho, S. P., & Macedo, E. A. (2008). Aqueous solubility ofsome natural phenolic compounds. Industrial & Engineering Chemistry Research,47(15), 5182–5189.

Ndlela, S. C., Moura, J. M. L. N., Olson, N. K., & Johnson, L. A. (2012). Aqueousextraction of oil and protein from soybeans with subcritical water. Journal of theAmerican Oil Chemists’ Society, 89(6), 1145–1153.

Pavlovic, I., Knez, Z., & Škerget, M. (2013). Hydrothermal reactions of agriculturaland food processing wastes in sub- and supercritical water: a review offundamentals, mechanisms, and state of research. Journal of Agricultural andFood Chemistry, 61(34), 8003–8025.

Phillips, K. M., Ruggio, D. M., & Ashraf-Khorassani, M. (2005). Analysis of sterylglucosides in foods and dietary supplements by solid-phase extraction and gaschromatography. Journal of Food Lipids, 12(2), 124–140.

Pourali, O., Salak Asghari, F., & Yoshida, H. (2009). Simultaneous rice bran oilstabilization and extraction using sub-critical water medium. Journal of FoodEngineering, 95(3), 510–516.

Rogalinski, T., Liu, K., Albrecht, T., & Brunner, G. (2008). Hydrolysis kinetics ofbiopolymers in subcritical water. The Journal of Supercritical Fluids, 46(3),335–341.

Rosenthal, A., Pyle, D. L., & Niranjan, K. (1996). Aqueous and enzymatic processes foredible oil extraction. Enzyme and Microbial Technology, 19(6), 402–420.

Rosenthal, A., Pyle, D. L., & Niranjan, K. (1998). Simultaneous aqueous extraction ofoil and protein from soybean: mechanisms for process design. Food andBioproducts Processing, 76(4), 224–230.

Ruen-ngam, D., Quitain, A. T., Tanaka, M., Sasaki, M., & Goto, M. (2012). Reactionkinetics of hydrothermal hydrolysis of hesperidin into more valuablecompounds under supercritical carbon dioxide conditions. The Journal ofSupercritical Fluids, 66, 215–220.

Sanguansri, L., & Ann Augustin, M. (2010). Microencapsulation in Functional FoodProduct Development. In J. Smith & E. Charter (Eds.). Functional Food ProductDevelopment (Vol. 1, pp. 1–23). Oxford: Wiley-Blackwell.

Shin, H.-Y., Ryu, J.-H., Park, S.-Y., & Bae, S.-Y. (2012). Thermal stability of fattyacids in subcritical water. Journal of Analytical and Applied Pyrolysis, 98,250–253.

Singh, P. P., & Saldaña, M. D. A. (2011). Subcritical water extraction of phenoliccompounds from potato peel. Food Research International, 44(8), 2452–2458.

Smith, R. M. (2002). Extractions with superheated water. Journal of ChromatographyA, 975(1), 31–46.

Škerget, M., Kotnik, P., Hadolin, M., Hraš, A. R., Simonic, M., & Knez, Z. (2005).Phenols, proanthocyanidins, flavones and flavonols in some plant materials andtheir antioxidant activities. Food Chemistry, 89(2), 191–198.

Venkatachalam, M., & Sathe, S. K. (2006). Chemical composition of selectededible nut seeds. Journal of Agricultural and Food Chemistry, 54(13),4705–4714.