Bioresource Technology 98 (2007) 3506–3512

Processing of Rosa rubiginosa: Extraction of oiland antioxidant substances

Daniel Franco, Manuel Pinelo *, Jorge Sineiro, Marıa Jose Nunez

Escuela Tecnica Superior de Ingenierıa, Universidad de Santiago de Compostela, Lope Gomez de Marzoa sn, 15782 Santiago de Compostela, Spain

Received 13 March 2006; accepted 11 November 2006Available online 3 January 2007

Abstract

In this work, a study about the effect of various operational conditions on the quantity of oil and soluble solids capable of beingextracted from rosa mosqueta rosehip seeds is undertaken. Both the particle sizes assayed (60.6 mm, 0.6–1 mm, and 1–2 mm) andthe solvent-to-solid ratios (15:1, 25:1, and 50:1) showed a remarkable influence on the extraction efficiency. Extracted substancesobtained by using the minor particle size or the maximum solvent-to-solid ratio doubled, at least, those attained by working underany other conditions. A major weight of kinetics upon equilibrium factors can be inferred from the short extraction times and high effec-tive diffusivity values (being the lower one 1.97 · 10�11 m2 s�1) assessed for any condition.

The antioxidant power of extracts was evaluated by ability to scavenge the DPPH radical. Results noteworthy depended on the sol-vent used to extract; whilst an �80% DPPH inhibition percentage was reached in ethanol extracts, values of 52.2% or 41% were found inmethanol and aqueous extracts, respectively. Even so, antioxidant capacity of Rosa rubiginosa extracts was much higher than thatreported for other agricultural matrixes.� 2006 Elsevier Ltd. All rights reserved.

Keywords: Rosa mosqueta (Rosa off. rubiginosa L.); Extraction process; Oil; Effective diffusivity; Antioxidant capacity

1. Introduction

Rosa mosqueta (common name for Rosa off. rubiginosa L.)is a wild shrub growing in some specific areas of CentralEurope and the Andean region. Bio-Bio river closeness inChile can be in fact considered one of the zones where thisspecie is most widely spread. Despite fairly scarce literaturecan be found about composition and characterization ofchemicals occurring in this oilseed, beneficial-health andcosmetic properties of its extracts have been transmittedand tapped by natives for centuries.

These properties could be in part attributed to the con-siderable vitamin C content detected in the fruit (400 mg/100 g), being 10-fold higher than in orange juice (Galeb

0960-8524/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.biortech.2006.11.012

* Corresponding author. Tel.: +34 981 563100x16785; fax: +34 981528050.

E-mail address: mpinelo@usc.es (M. Pinelo).

and Kiger, 1973). Moreover, oil and fibre fractions, as wellas other non-oily extractables were found to contain sub-stances exerting well-known therapeutic and pharma-cological effects. In the latter, phenolic compounds withantioxidant properties could certainly have a role in theproper beneficial mechanisms (Moure et al., 2000).

Oil fraction has been found to posses a high content oftrans-retinoic acid, having effective influence on anti-psori-ases, healing and anti-keratosis processes. For this reason,creams containing a rosa mosqueta oil basis are employedas a skin regenerating treatment in scars and burns (Vallad-ares et al., 1986). Likewise, triglycerides with high molecu-lar weight (deduced from a high saponification index) inthe same order as olive, sesame, corn, cotton, soy and sun-flower oils, the presence of arachidonic acid (essential forthe defence mechanism of prostaglandin synthesis andother basic processes for regeneration of tissues) as wellas unsaturated fatty acids like oleic, linoleic and linolenichave been reported (Rodrıguez and Soto, 1987; Badolato

D. Franco et al. / Bioresource Technology 98 (2007) 3506–3512 3507

et al., 1993). These applications have a great importancefor agricultural industry, since this oily fraction is findingincreasing applications as an active part for cosmetic andpharmaceutical compositions.

No less important health implications have beenassigned to the consumption of substances contained inthe non-oily fraction. By comparing with other similarvegetal products, phenolic compounds seem to be the res-ponsible for the antioxidant properties of rosa mosquetaextracts (Oomah et al., 1995; Lehtinen and Laakso,1998). Abundant studies support the role of phenolics onhuman health by lowering the incidence of coronary heartdisease and preventing thrombotic and atherogenic pro-cesses (Aruoma, 1994; Ames et al., 1993), as well as actingas antiviral agents against some diseases, such as diarrhea,arthritis, influenza and poliomyelitis (Serkedjieva and Hay,1998; Parmar et al., 1996). Apart from this, their part aspotential preservatives on lipid-containing foods has foundapplication in several commercial products like raw androast ham, hamburgers or fish (Valenzuela et al., 1992;Osada et al., 2001).

On the grounds of the applications previously broughtup, a maximization of the yields in these two fractions(oil and non-oily fractions) obtained from rosa mosquetawould be desirable. It is known that efficiency of an extrac-tion process is usually affected by some critical variableslike temperature, solvent employed, solid matrix structure,particle size, etc. This ensures that each vegetal matrix–sol-vent extraction system behaves in a particular way andmust be studied separately. To this effect, a study aboutthe influence of some variables involved in the extractionof both fractions by using different extraction methods willbe undertaken in this work. Mass transfer mechanismsexplaining such differences will be analyzed and yieldsbelonging to the non-oily antioxidant fraction and the oilfraction will be compared. In the former, besides, an eval-uation of the antioxidant quality of the extract by ability toscavenge the DPPH radical will be also carried out.

2. Methods

2.1. Materials

Rosa mosqueta seeds were supplied by the School ofBiochemical Engineering (Valparaıso, Chile). Due to theirconsiderable hardness, the seeds were ground in an agatemill. Effect of abrasion removes the most external coatingsof the material – where a lower oil amount is found. In thisway, a higher relative quantity of oil is detected since onlyinternal coatings are considered. Powdered samples weresieved to select particles in three different ranges: 1–2 mm, 0.6–1 mm and 60.6 mm and stored at 4 �C untiluse. The solvents used to achieve extraction of the differentfractions – 96% ethanol and petroleum ether – as well asacetone and methanol used for antioxidant power determi-nation were purchased to Drogas Vigo, S.L. DPPH, BHAand BHT were from Sigma.

2.2. Determination of extractable substances

2.2.1. Soxhlet extraction

The maximum weight of total extracted substances(including oil and non-oily fraction) was assessed after eth-anol extraction in Soxhlet during 8 h. After evaporatingsolvent in a Buchi Rotavapor R-114, values were deter-mined by increasing weight of flasks. Use of petroleumether allowed a selective assessment of oil. All the weightresults were expressed as the percentage of the initialamount of employed sample in a dry basis. Relevanthumidity values were assessed by Karl–Fisher technique(AOAC 934.20).

2.2.2. Orbital shaker extraction

Samples and ethanol were disposed in capped flasks(liquid-to-solvent ratios 15:1, 25:1, and 50:1) and main-tained at 150 rpm during 100 h in a ‘New Brunswick’ orbi-tal incubator shaker G24. Once more, petroleum ether wasused to selectively extract the occurring oil. The differenceswill therefore indicate the percentage non-lyposolubleextracts.

Afterwards, kinetics of oil extraction were followed.Samples previously dried up to 63% humidity wereinserted in capped flask containing 100 mL of 96% ethanol.A 50 �C temperature and 150 rpm were maintained whilesamples were extracted every 30 min up to 3 h.

2.3. Mathematical tools

2.3.1. Extraction kinetics modeling

The data corresponding to the overall amount of totalsolids remaining inside the vegetal matrix at the differentoperation times were fitted to the equation of Othmerand Jaatinen (1959):

C ¼ mt�b ð1Þbeing C the quantity of non-extracted solute referred to thetotal solid weight, g/g, t the extraction time, s and m and bthe fitting parameters. This equation was commonly em-ployed to model the kinetic extraction data in other vegetalmatrixes, providing important clues regarding ‘facility’ ofextraction in such matters.

2.3.2. Assessment of effective diffusivities

The result of extraction is a function of how fast thecompound is dissolved and the equilibrium in the liquidis reached. Four mass transfer steps are involved in the passof solute from the solid to the bulk of the solvent. In orderto simplify the data processing when solute diffusivities invegetal matrixes is undertaken, working in such a way thatthe rate limiting step is the diffusion of the dissolved solutewithin the solid into the solvent is advisable (Gertenbach,2001). The latter approach can be employed when negligi-ble external resistance to mass transfer is supposed. Thishypothesis is performed when working with very dilutedseed extracts. The rate of extraction therefore increases

3508 D. Franco et al. / Bioresource Technology 98 (2007) 3506–3512

with a larger concentration gradient (Pifferi and Vaccari,1983; Landbo and Meyer, 2001). Thus, assuming nochange of the effective diffusivity with solute concentration,the rate of diffusion of the limiting step can be describedby Fick’s second law:

oCot¼ rð�DeffrCÞ ð2Þ

where C is the concentration of the solute, t is time, s, andDeff is the effective diffusion coefficient or diffusivity, m2 s�1.

Considering that particles have spherical shape andconcentration differences will be only relevant in radialdirection,

oCot¼ Deff

o2C

or2ð3Þ

where r is the radio of diffusion, m. Crank (1975) assessedsecond Fick’s law analytical solutions for different geome-tries. Bargaining for all the particularities previously as-sumed (spherical geometry, negligible external resistance,and only radial direction flow), the following equationcan be applied to 60.6 rose seed particles:

E ¼ C � Ce

Ci � Ce

¼ C1 � Ct

C1

¼ 6

p2

X1n¼1

1

n2exp �n2 p2Deff t

r2

� �ð4Þ

where E is the non-extracted fraction at time t, C is theextractable content of solid at time t, Ce is the extractablecontent of solid at equilibrium, Ci initial extractable con-tent of solid, Ct is the extractable content in the solventat time t, and C1 extractable content of solvent at equilib-rium. Dependence of the effective diffusivity with tempera-ture can be described by Arrhenius equation:

Deff ¼ D0 exp � E0

RT

� �ð5Þ

where D0 is a constant, T is temperature (K), E0 is theactivation energy and R is the universal gas constant. Sinceseries converge quickly, the first terms are enough forassessments. Only the first three terms will be used in thiscase, though according to previous reports, the only firstterm of the series solution can usually be used with littleerror for extraction from plant matrixes where externalresistance is negligible (Schwartzberg, 1975):

E ¼ 6

p2expð�atÞ þ 1

4expð�4atÞ þ 1

9expð�9atÞ

� �;

where a ¼ p2

r2Deff ð6Þ

Taking the first term of the equation and applyinglogarithms:

E ¼ 6

p2ðexpð�atÞÞ ) ln E ¼ ln

6

p2� at ¼ �0:49� at

ð7Þ

2.4. Determination of antioxidant capacity

Ground seeds were previously hexane-defatted at roomtemperature and solvents with different polarity – metha-nol, ethanol, acetone, and water at pH 3–4, were evaluatedas phenolic extractants. Extraction was carried out at roomtemperature except for acidified water, where an enhance-ment of phenol solubility was attempted at 40 �C. Asolvent-to-solid ratio 15:1 was used in order to obtainrelatively concentrated extracts, which were lyophilizedand stored at 4 �C.

A DPPH radical-scavenging assay was employed asdescribed by Von Gadow et al. (1997) to determine thehydrogen-donating ability of the re-dissolved extract. Avolume of 1850 lL of 6.1 · 10�5 M DPPH� methanol solu-tion was used. The reaction was started by the addition of150 lL of sample. The bleaching of DPPH� was followed at515 nm (Shimadzu UV-160A) at 25 �C for 16 min. Theinhibition percentage (IP) of the DPPH� radical was calcu-lated as follows:

IP ¼ ðabsorbancet¼0 min � absorbancet¼16 minÞðabsorbancet¼0 minÞ

� 100 ð8Þ

2.5. Statistical analysis

The results reported in this work are the averages of atleast three measurements, and the coefficients of variation,expressed as the percentage ratio between standard devia-tions (SD) and the mean values, were found to be <10 inall cases. Significant variables were calculated, subjectingresults to a linear regression, using SPSS statistical pro-gram version 10.0 (SPSS Inc., Chicago, IL). Only variableswith a confidence level superior to 95% (p < 0.05) were con-sidered as significant.

3. Results and discussion

3.1. Extraction of oil and non-oily fractions

In Fig. 1a, the percentages of extractable solids as wellas oil (g extracted substances/100 g sample in dry basis)obtained by Soxhlet extraction with ethanol and petroleumether respectively are plotted. The higher percentages ofextracted substances were attained when 60.6 mm particlesize were used, with values �3-fold higher than thosedetected for particle sizes 1–2 mm. It seems obvious thatdecreasing particle sizes promote a major solid transferfrom rose seeds, since surface contact between two phasesinvolved increases to a large extent. This effect was alsoconfirmed by Meyer (2002) and Cacace and Mazza(2003), reporting an enhancement of extraction efficiencywhen subjected to extraction smaller particles of grapepomace and milled berries, respectively. Smaller particlesize reduces the diffusion distance of the solute within thesolid, thus increasing the extraction rate. The solute there-fore takes a shorter time to reach the surface. Fig. 1b shows

0

5

10

15

20

25

30

35

0.6 mm 0.6-1 mm 1-2 mm

% E

xtra

cted

su

bst

ance

s extractables oil

0

5

10

15

20

25

30

35

0 1 2 3 4 5 6

Time (h)

% E

xtra

cted

su

bst

ance

s

≥

Fig. 1. Percentage of extractable substances obtained by ethanol extraction from rosa mosqueta seeds in Soxhlet: (a) results for different particle sizeranges and (b) kinetics of extraction.

D. Franco et al. / Bioresource Technology 98 (2007) 3506–3512 3509

the percentage of extracted substances at different times ofthe extraction process. As can be observed, no appreciablechanges were found after 1–2 h. This confirms no furthermass transfer after this time and noticeably reduces theeffective duration of the process. The required time to solvethe whole solute into the bulk of the solvent indicates howretained is the former one inside the structure of the plantmatrix. Although no important impediments seem to occurin this case, other reported examples show the differentnature that solids subjected to extraction can have. Indeed,Pinelo et al. (2004a,b, 2005) needed 8 h to completelyexhaust the extractable solids contained in grape byprod-ucts using comparable extraction conditions.

Total percentages found by shaker extraction are shownin Fig. 2a. Since higher values of extractables were previ-ously obtained using 60.6 particles, only this particle sizewas employed in this case. Extraction was carried out atthree different solvent-to-solid ratios during 100 h. The bestresults were detected for a solvent-to-solid ratio 50:1 (36%oil and 40% total solids), more than 4-times higher thanthose reached at 15:1 (7% oil, 9% total solids). Higher sol-vent-to-solid ratios promoted an increasing concentrationgradient between solvent and solid to extract. As a conse-quence, a major mass transfer between both phases wasobserved. Comparing the results taken from the twoextraction procedures, higher quantities of extractableswere detected in shaker. Differences could be mainlyexplained in basis of the stirring effects. In fact, agitationfavors the convective movement in the bulk of the solventand thus compensates the gradual decrease in gradientcaused by increasing solute concentration.

3.2. Kinetic of oil extraction

Fig. 2b shows the evolution of extracted oil with extrac-tion time, whilst mathematical fitting of Othmer and Jaati-nen (1959) is plotted in Fig. 2c. As can be observed, analmost complete exhaustion of the solid to extract is

accomplished for a solvent-to-solid ratio 50:1. By contrast,almost 20% and 40% of soluble solids were remained in thevegetal matrix when working with 15:1 and 25:1 ratios. Nofurther oil extraction after �30 min can be deduced fromthe data, confirming the fluent solute transfer to the bulkof the solvent also under these conditions. A good model-ling of data was achieved for all the solvent-to-solid ratiosassayed (Table 1).

Values of effective diffusivity for the three solvent-to-solid ratios shown in Table 2 were determined from linearregression of straight lines plotted in Fig. 3 (resulting fromEq. (7)). The higher values of effective diffusivity enhancedconcomitantly with increasing values of solvent-to-solidratio (Fig. 2a). As it was previously indicated, a major con-centration gradient is derived from the utilization of highersolvent quantities. As a result, an expectable intensificationof mass transfer from the vegetable matrix to solvent isproduced. Values of effective diffusivity are known tostrongly depend on the particular vegetal matrix–solventsystem analysed. Reported data bore out that their valuescan largely vary as a function of the employed material.Sineiro et al. (1996) reported diffusivity values ranged from22.1 · 10�14 and 61.26 · 10�14 m2 s�1 when extracted sun-flower press cake (cake flake thickness was 0.6 mm), whilstvalues as high as 2.079 · 10�11 m2 s�1 were found byCacace et al. (2003) during an aqueous-ethanol extractionof 6 mm-diameter milled berries. Notwithstanding theseconsiderations, values of effective diffusivity detected in thiscase are higher than others reported for similar agriculturalmatters.

3.3. Antioxidant capacity

Very few works regarding the antioxidant properties ofroses or other flower extracts can be found in literature. Inparticular, findings about rosa mosqueta are focussed onthe presence of carotenoids and vitamin C, whose antioxi-dant properties have been widely studied (Hornero-Mendez

0

5

10

15

20

25

30

0 30 60 90 120

Time (min)

% E

xtra

cted

su

bst

ance

s

50:1 25:1 15:1

0

0.2

0.4

0.6

0.8

1

0 50 100

Time (h)

g n

on

-ext

ract

ed o

il/ g

iner

t so

lid

05

101520

253035

4045

15:1 25:1 50:1Solvent-to-solid ratio

% T

ota

l so

lids

Fig. 2. Percentage of total solids obtained by ethanol extraction fromRosa mosqueta seeds in orbital shaker (150 rpm, 50 �C). (a) Results fordifferent solvent-to-solid ratios. (b) Kinetic of total solids extracted atdifferent extraction times. (c) Kinetic data fitting to Othmer–Jaatinenequation.

Table 1Values resulting from fitting kinetic data of shaker extraction to Othmer–Jaatinen equation

Solvent-to-solid ratio m b R2

15:1 0.060 0.182 0.98025:1 0.241 0.151 0.97150:1 0.363 0.202 0.990

Table 2Values resulting from fitting kinetic data of shaker extraction to Eq. (7)

Solvent-to-solid ratio a ¼ Deffp2

r2 Deff · 10�11 (m2 s�1) R2 cte

15:1 0.0004 0.61 0.980 0.14025:1 0.0021 3.19 0.999 0.40050:1 0.0046 6.99 0.997 0.944

Relevant effective diffusivity values are also indicated.

-1.6-1.4-1.2

-1-0.8-0.6-0.4-0.2

030 60 90 120

Time (min)

lnE

50:1 25:1 15:1

Fig. 3. Plot according to Eq. (7).

Table 3Antioxidant capacity values of rosa mosqueta extracts obtained by usingsolvents with different polarity

Solvent Polarity DPPH inhibition percentage

Ethanol 5.2 80.5 ± 3.24Acetone 5.4 76.3 ± 3.78Methanol 6.6 52.2 ± 4.05Acidified water 9.0 41.0 ± 3.33BHAa – 67.3 ± 2.22BHTa – 39.4 ± 1.98

a Concentration corresponding to 0.5 mol antioxidant/mol DPPH.

3510 D. Franco et al. / Bioresource Technology 98 (2007) 3506–3512

and Mıguez-Mosquera, 2000; Gomez et al., 1993). Table 3shows the antioxidant activity values of seed extractsobtained by using different solvents. Antioxidant activitywas undoubtedly affected by the solvent. In particular, aspolarity was increased, the same amount of extract exerted

a lower antioxidant capacity. It is popularly known thefacility of polar and non polar solvents to dissolve polarand non polar molecules, respectively. Thus, the higherantioxidant activity detected in extracts from low polaritysolvents could be probably due to a major contain ofphenolic antioxidants with low values of polarity in rosamoschata. Furthermore, these differences could be in partattributed to the influence of solvent on the phenolicmolecules enclosed in extracts, affecting their antioxidantproperties. This hypothesis is in agreement with studiescarried out by Valgimigli et al. (1995), emphasising thathydrogen bonding in solvents may induce dramatic changesin the H-atom donor activities of phenolic antioxidants. Inthis way, Pedrielli et al. (2001) observed that antioxidantactivity of the same flavonoid was much higher in the non-hydrogen-bonding solvent, chlorobenzene, than in the‘water-like’ solvent, tert-butyl alcohol. Likewise, previousfindings demonstrated that the antioxidant activity ofquercetin in ethanol solution was near twice higher thanthat observed in the organic solvent Triton X-100 (whichpresents an important polar group) or in an aqueoussolution (Van der Berg et al., 1999; Pinelo et al., 2004a).

D. Franco et al. / Bioresource Technology 98 (2007) 3506–3512 3511

Higher DPPH inhibition percentages found in theseextracts in comparison to those detected for other agricul-tural products, suggest that rosa moschata is a good sourceof antioxidant compounds. Moure et al. (2000), forinstance, reported a 13.2% DPPH inhibition for ethanolextracts of Gevuina avellana, and a maximum of26.00 ± 0.68% was attained in ethanol extracts of pinesawdust (Pinelo et al., 2004b).

4. Conclusions

The quantity of oil and total solids capable of beingextracted from rosa moschata seeds is noteworthy affectedby conditions to which the particular extraction processused is carried out. In this work, the positive effect ofdecreasing the particle size or enhancing the solvent-to-solid ratio is reported. In the former case, the soluble solidspercentage extracted in Soxhlet by using 60.6 mm particlesizes (�27%) doubled that obtained when size ranges of0.6–1 mm or 1–2 mm were used. In the same way, �1.5-fold higher quantity of extractables was detected whensolvent-to-solid ratio passed from 25:1 to 50:1.

The low extraction times needed to achieve the nearly allsoluble solids from seed matrix as well as the high values ofeffective diffusivity demonstrated the low capacity of rosamoschata structure to retain the soluble solids, noticeablyminor to other vegetal matters.

The high values of the DPPH inhibition percentage ofextracts confirm rosa moschata as a good and cheap sourceof antioxidant substances. In fact, despite the remarkableinfluence of the solvent used to extraction, the antioxidantcapacity values detected are superior to other similar agri-culture products. This work pretends to be a first approachto the study of the antioxidant properties of the bush. Bothchromatographic techniques revealing the species responsi-ble for such properties and possible fractionation ofextracts could be the following step to be taken in furtherstudies.

References

Ames, B.N., Shigena, M.K., Hagwn, T.M., 1993. Oxidants, antioxidants,and the degenerative diseases of aging. Proc. Natl. Acad. Sci. USA 90,7915–7922.

Aruoma, O.I., 1994. Nutrition and health aspects of free radicals andantioxidants. Food Chem. Toxicol. 32, 671–683.

Badolato, E., Pimentel, S., Tabares, M., 1993. Oleo natural de Rosamosqueta e cosmeticos contendo esse oleo: verificasao de sua qualid-ade por cromatografia en fase gaseosa. Aerosol y Cosmeticos 14, 42–47.

Cacace, J.E., Mazza, G., 2003. Mass transfer process during extraction ofphenolic compounds from milled berries. J. Food Eng. 59, 379–389.

Crank, J., 1975. The Mathematics of Diffusion, second ed. ClarendonPress, Oxford.

Galeb, P., Kiger, F., 1973. Elaboracion de jugo s concentrados, conge-lados y refrigerados de maranja (Citrus sinensis O.) de las variedadeschilena, Washington y tardıa de Valencia. Tesis de grado, Fac.Agronomıa, Universidad de Chile.

Gertenbach, D.D., 2001. Solid–liquid extraction technologies for manu-facturing nutraceuticals from botanicals. In: Shi, J., Mazza, G., Le

Maguer, M. (Eds.), Functional Foods: Biochemical and ProcessingAspects. CRC Press Inc., Boca Raton, FL, pp. 331–366.

Gomez, R.G., Malec, L.S., Vigo, M.S., 1993. Composicion quımica defruto maduro de Rosa mosqueta. Anales de la Asociacion de Quımicade Argentina 81, 9–14.

Hornero-Mendez, D., Mıguez-Mosquera, M.I., 2000. Carotenoid pig-ments in rosa mosqueta hips, an alternative carotenoid source forfoods. J. Agric. Food Chem. 48, 825–828.

Landbo, A.K., Meyer, A.S., 2001. Enzyme-assisted extraction of antiox-idantive phenols from black currant juice press residues (Ribes nigrum).J. Agric. Food Chem. 49, 3169–3177.

Lehtinen, P., Laakso, S., 1998. Effect of extraction conditions on therecovery an d potency of antioxidants in oat fibre. J. Agric. FoodChem. 46, 4842–4845.

Meyer, A.S., 2002. Enhanced extraction of antioxidant phenols from wineand juice press residues via enzymatic polysaccaride hydrolysis. FruitProc. 1, 29–33.

Moure, A., Franco, D., Sineiro, J., Domınguez, H., Nunez, M.J., Lema,J.M., 2000. Evaluation of extracts from Gevuina avellana hulls asantioxidants. J. Agric. Food Chem. 48, 3890–3897.

Oomah, B.D., Kenaschuk, E.O., Mazza, G., 1995. Phenolic acids inflaxseed. J. Agric. Food Chem. 43, 2016–2019.

Osada, K., Hoshina, S., Nakamura, S., Sugano, M., 2001. Cholesteroloxidation in meat products and its regulation by supplementation ofsodium nitrite and apple polyphenol before processing. J. Agric. FoodChem. 48, 3823–3829.

Othmer, D.F., Jaatinen, W.A., 1959. Extraction of soybeans. Ind. Eng.Chem. 51, 543–546.

Parmar, V.S., Bisht, K.S., Jain, R., Singh, S., Sharma, S.K., Gupta, S.,Malhotr a, S., Tyagi, O.D., Vardhan, A., 1996. Synthesis, antimicro-bial and antiviral activities of novel polyphenolic compounds. IndianJ. Chem. 35, 220–232.

Pedrielli, P., Pedulli, G.F., Skibsted, L.H., 2001. Antioxidant mechanismof flavo noids. Solvent effect on rate constant for chain-breakingreaction of quercetin and epicatechin in autoxidation of methyllinoleate. J. Agric. Food Chem. 6, 3034–3040.

Pifferi, P.G., Vaccari, A., 1983. The anthocyanins of sunflower II. A studyof the extraction process. J. Food Technol. 18, 629–638.

Pinelo, M., Manzocco, L., Nunez, M.J., Nicoli, M.C., 2004a. Solventeffect on quercetin antioxidant capacity. Food Chem. 88, 201–207.

Pinelo, M., Rubilar, M., Sineiro, J., Nunez, M.J., 2004b. Extraction of antioxidant phenolics from almond hulls (Prunus amygdalus) and pinesawdust (Pinus pinaster). Food Chem. 85, 267–273.

Pinelo, M., del Fabbro, P., Manzocco, L., Nunez, M.J., Nicoli, M.C.,2005. Optimization of continuous extraction from Vitis vinifera

byproducts. Food Chem. 92, 109–117.Rodrıguez, A., Soto, G., 1987. Caracterizacion del aceite crudo de semilla

de Rosa mosqueta. Grasas aceites 1, 20–22.Schwartzberg, H.G., 1975. Mathematical analysis of solubilization.

Kinetics and diffusion in foods. J. Food Sci. 40, 211–213.Serkedjieva, J., Hay, A., 1998. In vitro anti-influenza virus activity of a

plant preparation from Geranium sanguineum L. Antivir. Res. 37, 121–130.

Sineiro, J., Domınguez, H., Nunez, M.J., Lema, J.M., 1996. Etanolextraction of polyphenols in an immersion extractor. Effect of pulsingflow. J. Am. Oil. Chem. Soc. 73, 1121–1125.

Valenzuela, A., Nieto, S., Cassels, B.K., Speisky, H., 1992. Inhibitoryeffect of boldine on fish oil oxidation. J. Am. Oil Chem. Soc. 68, 935–937.

Valgimigli, L., Banks, J.T., Ingold, K.U., Lusztyk, J., 1995. Kineticsolvent effect on hydroxylic hydrogen atom abstractions are indepen-dent of the nature of the abstracting radicals. Two extreme tests usingvitamin E and phenol. J. Am. Oil Chem. Soc. 117, 9966–9971.

Valladares, J., Palma, M., Sandoval, C., Carvajal, F., 1986. Crema deaceite de semilla de mosqueta (Rosa aff. Rubiginosa L.) II parte:Estudio de las propiedades fısico-quımicas, de estabilidad, eficaciacosmetica y aplicacion sistematica en clınica. Anales de la RealAcademia de Farmacia 51, 597–612.

3512 D. Franco et al. / Bioresource Technology 98 (2007) 3506–3512

Van der Berg, R., Haenen, G.R.R.M., Van der Berg, M., Bast, A., 1999.Applicability of an improved Trolox equivalent antioxidant capacity(TEAC) assay for evaluation of antioxidant capacity measurements ofmixtures. Food Chem. 66, 511–517.

Von Gadow, A., Joubert, E., Hansmann, C.F., 1997. Comparison of theantioxidant activity of aspalathin with that of other plant phenols ofrooibos tea (Aspalathus linearis), alpha-tocopherol, BHT and BHA. J.Agric. Food Chem. 45, 532–638.