Journal of Food Engineering

Journl of Food Engineering 97 () 533 53 Contents lists vilble t SieneDiret Journl of Food Engineering journl homepge: www.elsevier.om/lote/jfoodeng Kneding olive pste from unripe Piul fruits: I. Effet on oil proess yield Mri P. Aguiler, Gbriel Beltrn, *, Sebstián Snhez-Villslrs b, Mrino Ued, Antonio Jimenez Centro Vent del Llno, IFAPA, Junt de Andluí, P.O. Box 5, 3 Mengibr, Jen, Spin b Deprtmento de Ingenieri Quími, Ambientl y de los Mteriles, Universidd de Jen, Cmpus Ls Lgunills s/n, 37 Jen, Spin rtile info bstrt Artile history: Reeived My 9 Reeived in revised form 7 Otober 9 Aepted 7 November 9 Avilble online November 9 Keywords: Virgin olive oil Extrtion proess Oil yield Kneding onditions Mironised nturl tl Unripe olives Olive growers strt erlier to hrvesting in order to obtin high qulity oils, however unripe olives show different hrteristis ffeting tehnologil nd rheologil properties. This work is imed to study the effet of mlxtion onditions on proess oil yield for unripe Piul olives. Experiments were rried out using Piul fruits hrvested t erly dtes for three onseutive rops sesons. The mlxtion onditions ssyed using ftoril design were: time (, 9 min), temperture (, 3, C) nd mironised tl ddition (, %). Olive pste from unripe Piul fruit re lssified s diffiult pstes. The min ftor ffeting oil yield ws the rop yer, while for eh rop yer the effet of eh ftor vried onsiderbly. Results indited s inresing time nd temperture gve lower pome oil losses lthough its effet vried on rop yer. Mironised Nturl Tl ddition lowered pome oil ontent on dry weight. Therefore for erly hrvesting dtes knowing olive hrteristis on line is essentil to dpt the kneding onditions. Ó 9 Elsevier Ltd. All rights reserved.. Introdution Virgin olive oil extrtion is rried out by using only physil methods. Around % of oil loted in the vuoles of olive pulp ells n be esily extrted, the rest n be found dispersed in the ytoplsm s mirogels nd thus, is diffiult to extrt. In order to disrupt the interfil films the olive pste is mixed slowly modifying the dispersion degree of oil droplets nd promoting the olesene phenomen. Colesene inreses the mount of free olive oil beuse of breking of oil/wter emulsion formed during the rushing opertion. The oil droplets inrese their size nd when they hieve dimeter greter thn 3 lm, the oil n be extrted esily (Mrtinez et l., 957). Effiieny of kneding opertion depends on olive pste rheologil hrteristis nd mlxtion vribles suh s time, temperture nd tehnologil odjuvnt ddition. Olive pste kneding onditions ffet the proess yield, improving when kneding time ws prolonged from 5 to 75 min lthough kneding time exessively longer (5 min) did not give signifint better results (Hermoso et l., 99). Di Giovhino (99) desribed s olive pste kneding time higher thn 5 min is needed to hve good oil yields. Olive pste temperture dereses oil visosity, improving the olesene phenomen nd the proess yield (Hermoso et l., 99; Di Giovhino et l., b). Extrtion t C produes pome with lower oil * Corresponding uthor. Address: Centro Vent del Llno, IFAPA, P.O. Box 5, 3 Mengibr, Jen, Spin. Tel.: +3 953337; fx: +3 95333. E-mil ddress: gbriel.beltrn@juntdendlui.es (G. Beltrn). ontent thn those obtined t 3 C, lthough high temperture n ffet negtively to oil qulity nd its sensory hrteristis (Di Giovhino, 99; Hermoso et l., 99). Mironised Nturl Tl (MNT) is used s tehnologil odjuvnt beuse of its orgnophylli properties reduing the oil/wter emulsions nd liberting oil for its extrtion. Its use is reommended for the so-lled diffiult pstes tht show emulsions nd low proess yields (Mrtinez et l., 975; Muñoz et l., 979; Di Giovhino, 9). Experiments t industril sle proposed the intermedite ontiner of the pste kneder s the optiml lotion for tl ddition sine lower oil losses were observed. Different MNT doses hve been tested, obtining the best oil yields for %. When % of tl ws dded non-signifint differenes were observed lthough for diffiult pstes, higher MNT mounts my be needed (Hermoso et l., 99). Most of these works were rried out just for single dte nd rop yer nd they did not onsider the effet of hrvesting dte nd rop yer for speifi hrvest seson. Furthermore some results re diffiult to trnslte to other ultivrs sine extrtbility nd pste rheologil hrteristis vry mong ultivrs (Beltrn et l., 3). Piul is onsidered s the most importnt ultivr grown in Andlusi, Spin, EU nd the World. In spite of its importne there re sre dt bout its behviour during oil extrtion nd how to proess it. Among its ultivr hrteristis, Piul olive fruit flls on ground when ripens beuse of its low fruit retention fore (Brrno et l., 5). It ours erly, giving low qulity oils from ground-piked fruits. Beuse of tht, my be interesting to ntiipte the hrvesting period in order to void oil qulity losses. For -77/$ - see front mtter Ó 9 Elsevier Ltd. All rights reserved. doi:./j.jfoodeng.9..3

53 M.P. Aguiler et l. / Journl of Food Engineering 97 () 533 53 the lst yers, in order to improve oil qulity, sensory hrteristis nd nutritionl omposition more oil mills re extrting olive oil t erlier hrvesting dtes nd lower temperture, however no dt re vilble bout the optiml kneding onditions using this kind of fruit to obtin rentl oil yields. The im of this work ws to study how the mlxtion onditions ffet the oil proess yield when erly hrvested fruit were proessed.. Mterils nd methods.. Plnt mteril This work hs been rried out t the beginning of norml hrvesting period for three onseutive rop yers /, / nd /3. Around kg of Piul olive fruits were piked from mture trees grown under irrigtion in the experimentl olive orhrd of IFAPA in Centro Vent del Llno in Mengibr, Jén. For eh yer, trees were seleted with rop lod index 3. Olives were hrvested by mehnil shoker nd trnsported immeditely to the pilot plnt for proessing. The hrvesting dtes nd fruit ripeness index for eh rop yer re shown in Tble... Proessing nd extrtion onditions The olives were lef-removl nd not wshed, n homogeneous lot of kg of olives ws stored in divided hopper with ommon exit to improve the fruit homogeneity; bthes of 5 kg of olives were proessed in eh ssy. Olives were rushed in metlli hmmer mill with sieve size mm for ll the experiments, filling eh bth mixer ontiner. The IFAPA pilot plnt ws equipped with thermobeter formed by three ontiners (Pierlisi, Spin), two three phses horizontl entrifuge Pierlisi SC-9 (working t two phses wy) with theoril proessing pity of 5, kg/dy nd vertil entrifuge for oil lening Pierlisi P5. The proess ws ontrolled nd monitored by Proioleo utomtion softwre (Prois, Seville). The olive pste ws loded t 75 kg/h in the horizontl entrifuge, being ontrolled by the utomtion system. In the experiments three kneding tempertures were ssyed (, 3 nd C) for two mlxtion times ( nd 9 min). Tble Hrvesting dte nd ripening index for Piul olive fruit used in the mlxtion experiments rried out during the rop yers /, / nd /3. Crop yer Hrvesting dte Ripening index / /9/. / //.3 /3 /7/. Tble Ftoril design of olive pste mlxtion experiments. t (min) T ( C) MNT (%) 3 9 3 Mironised nturl tl (MNT) ws dded t %, when used, by n utomti MNT dispenser previously librted. The experimentl design is reported in Tble..3. Smpling Olive smples were olleted, by triplite, t the beginning of eh experiment of time nd temperture being ommon for both MNT doses ssys. Pome smples were tken from the horizontl entrifuge outlet for the lst 5 min of eh ssy in order to urte the stbilistion of the denter performne. Replites were tken t 5 min intervls... Anlytil determintions Fruit ripening index ws mesured ording to the method the method proposed by the Estion de Oliviultur y Elioteni (Beltrn et l., ). Crushed olives nd pomes (3 g) were weighted nd dried until onstnt weight in n oven t 5 C, the moisture ws expressed s perent. Oil ontent ws nlysed by Nuler Mgneti Resonne (NMR) ft nlyser mq NMR Anlyser (Bruker, Spin) nd were expressed s perent on fresh nd dry weight bsis. The equipment ws librted periodilly, for both fruit nd pome, by Soxhlet nlyses using n-hexne s extrting solvent..5. Sttistil nlysis ANOVA nlysis for ftoril design ws performed using Sttistix. Anlytil Softwre in order to determine the effet of mlxtion time, temperture nd mironised nturl tl on pome oil ontent on dry weight. Tukey s test ws pplied to determine differenes between mens. Dt for the tretment C min nd % of MNT of rop yer / were missing. 3. Results nd disussion All the oils fitted within extr virgin tegory s stted by Europen Union Regultion 5/9 (99). Qulity prmeters nlysed were below the upper limits estblished for this tegory, the rnges observed for them were: Aidity (..3%), Peroxide vlue (. 9. meqo kg ), K 3 (..7) nd K 7 (..). In Tble 3 re reported the men hrteristis of olive fruits used in the experiments, s n be observed fruit hrteristis vried signifintly for eh rop yer ffeting extrtbility nd pste rheologil properties. Fruits hrvested during / nd /3 rop yers showed similr vlues for dry extrt (%) nd moisture (%), lthough for the lst yer lower oil ontent ws registered probbly beuse lipogenesis did not finish yet. Nevertheless the fruit with more importnt differenes were observed in the rop yer / sine they showed very high humidity (5.35%), high oil ontent (on dry weight) nd very low level of dry extrt. Beuse of this speil hrteristis, greter emulsions my be formed during rushing step giving diffiult pstes. Oil yield is usully expressed s pome oil ontent on fresh weight (POFW), the min oil mill by-produt, lthough when expressed on dry weight bsis is not ffeted by the wter ddition in the proess. Therefore results nd disussion will be bsed on pome oil ontent on dry weight (PODW). In Tble re reported the pome oil ontent on dry weight for this work. Considering the three rop yers together, the ANOVA nlysis (Tble 5) showed s the rop yer ws the most importnt ftor ffeting the oil yield proess (.%) wheres the mlxtion vribles explined round % of the observed vribility. As shown in Tble, oil yields were better for the rop yer / while the

M.P. Aguiler et l. / Journl of Food Engineering 97 () 533 53 535 Tble 3 Chrteristis of olive fruits used in the mlxtion experiments rried out during the rop yers /, / nd /3. Crop yer t (min) T ( C) MNT (%) Moisture (%) Oil/FW (%) b Dry extrt (%) / 55.9..5 3.7 / 3 5.. 3.7 7.7 / 5.3..9 7. / 9 53.95.7 5. 5.3 / 9 3 5.7.73 5.5. / 9 5.79.9.5.9 / 55.5 9... / 3 57.9 9.9 7.. / 57.3.7..9 / 9 57. 9..5 3. / 9 3 5.9.99.77.5 / 9 53.3.57..9 /3 5.33. 3. 3.5 /3 3 53.7 9...9 /3 5.9.35 39.3.7 /3 9 5.. 3. 9.3 /3 9 3 53..33 3.9 5.99 /3 9 5. 9.55 3. 5.5 b Oil ontent on fresh weight. Oil ontent on dry weight. Tble Effet of mlxtion onditions on pome oil ontent on dry weight (%) for three onseutive rop yers: /, / nd /3. T ( C) t (min) MNT (%) Crop yer Men / / /3 7.39 ±.. ±.3.33 ±.3.79 ±. 7.7 ±. 7.35 ±..5 ±.5 9.3 ±.3.9 ±.5 7.75 ±.53.53 ±.33 5.95 ±.7.7 ±.. ±.53 7. ±. 3. ±.33. ±.5. ±.77. ±.75 5.7 ±.9.95 ±.9 7.7 ±.57. ±.33 9 5.3 ±.. ±. 7. ±.9 7.5 ±.9 5.9 ±.. ±.3 7. ±. 7.75 ±.5 5.9 ±..9 ±.3 7.9 ±.. ±.5 5.3 ±.9.5 ±.7 7.9 ±.3 7.9 ±.5 9 5.99 ±. 9.5 ±. 7.3 ±.3 7.5 ±.5 5.97 ±.. ±..3 ±.3 7. ±. Men vlues ± SD (n = 3). Tble 5 Prtil men squres from nlysis of vrine for the effet of mlxtion onditions on pome oil ontent on dry weight (%) t erly hrvesting dte for the rop yers: /, / nd /3. Soure DF MS b SST Totl 7 Crop yer (C) 393..3. Mironised tl (MNT).535.9. Temperture (T).3.7. Time (t).5.. C MNT.3.9.359 C T.5.7. C t.3..73 MNT T...9 MNT t.9.. T t.95.. C MNT T.7.5. C MNT t...957 C T t 5.99 3.. MNT T t.3.. C MNT T t...7 Error 7 3.9 3. b Degrees of freedom. Men squres. Prtil men squre for the effet, expressed s perentge, of the totl orreted sum of squres. P worst were registered for /. This different extrtion effiieny my be explined beuse of the lrge differenes in fruit moisture reorded for the three rop yers (Tble 3). In generl, olive pstes showing moisture higher thn 5% re lssified s diffiult pstes (Di Giovhino, 99). Furthermore olive pste from rop yer / my be lssified s very diffiult pste, sine it showed the highest moisture (5.3%), high oil ontent nd very low dry extrt perent. During its proessing the oil lost ws very high (%) nd mlxtion onditions did not improve signifintly the oil yield. Therefore nd beuse of the signifint effet of rop yer, seprted ANOVA nlysis for eh one ws neessry to get more robust informtion (Tble ). Mlxtion temperture ffeted the proess oil yield lthough its effet vried signifintly depending on the fruit hrteristis for eh yer (Fig. ). The low oil yield improvement obtined when temperture ws inresed is due to the diffiult pstes produed from these olive fruits. For rop yer / only for C signifint redution in the oil losses ws observed wheres for / rop yer temperture ws the min ftor ffeting the PODW; non-signifint effet ws observed for /3 rop yer. Inresing mlxtion temperture redued the pome oil ontent on dry weight for ll the rop yers studied, obtining the lower vlues nd thus, the highest yields, kneding t C. But when similr temperture inrese ws performed we did not observed

53 M.P. Aguiler et l. / Journl of Food Engineering 97 () 533 53 Tble Prtil men squres from nlysis of vrine for the effet of mlxtion onditions on pome oil ontent on dry weight (%) t erly hrvesting dte for the rop yers: /, / nd /3. b Crop yer DF Totl MNT T t T t MNT T MNT t Error 35 / MS b.5.59 **.79 ***.9 *** 3.93 ***..35.3 SST 3. 53.5.7.7. 7. / MS 35..7 * 3.37 ** 3.555 *.9 ***.35.353.5 SST 7.59.9 9.97 7.5.7.99 33. /3 MS.99.979 **..75 ***.5.577.9.9 SST.3.9 3.7.5 9.7.7.7 Degrees of freedom. Men squres. Prtil men squre for the effet, expressed s perentge, of the totl orreted sum of squres. * Signifine level p:.5. ** Signifine level p:.. *** Signifine level p:.. / / /3 Men Hrvesting dtes ºC 3 ºC ºC Fig.. Effet of mlxtion temperture on pome oil ontent on dry weight for three onseutive rop yers: /, / nd /3. Different letters men signifint differenes t *p:.5, **p:. nd ***p:.. ** / / /3 Men Crop yer the sme oil yield redution. For the first rop yer higher temperture thn C improve the proess yield signifintly by.7%, min 9 min * * ** Fig.. Effet of mlxtion time on pome oil ontent on dry weight for three onseutive rop yers: /, / nd /3. Different letters men signifint differenes t *p:.5, **p:. nd ***p:.. however for the seond one only inresing mlxtion temperture to C let to redue the pome oil losses. Although, in generl, derese of PODW ws observed for higher mlxtion temperture, ttention should be pid into the vlues obtined for C sine only for one rop yer oil yield were worse thn 3 C nd thus, it my be interesting to proess t low temperture to obtin Virgin olive oils with singulr hrteristis. Longer mlxtion times gve signifint lower PODW (Fig. ), s desribed previously (Hermoso et l., 99; Di Giovhino et l., ), sine inresing kneding time let to redue the emulsion formed during rushing opertion, inresing the free oil in the olive pste. This vrible hieved higher importne for the rop yer /3 (Tble ) lthough ws signifint for ll the rop yers studied. Prolonging mlxtion from to 9 min gve similr redution in PODW for the three rop yers: round 7%, independently of the fruit hrteristis. Although both, time nd temperture showed signifint effet on oil yield by themselves, their intertion showed highly signifint effet for the rop yers / nd / (Tble ). In Fig. 3 intertion between time nd temperture is represented for eh rop yer. For / signifint derese in PODW ws observed only when kneding t C for 9 min. However for the rop yer /, the signifine of this intertion ws higher s the temperture rose: mlxing t 3 C for 9 min lowered the PODW by 9.7% wheres this redution ws.5% t C. The importnt effet of this intertion my be explined beuse olive pste during / n be lssified s very diffiult pste (moisture higher thn 5%) nd thus both higher temperture nd longer time re needed to redue the mirogels nd semiolloidl strutures desribed for these olive pstes. The third vrible studied ws the ddition of mironised nturl tl (MNT) s tehnologil odjuvnt. MNT lowered signifintly (Fig. ) the pome oil ontent on dry weight (p <.5) s desribed for % MNT ddition by Hermoso et l. (99), this redution vried depending on the fruit hrteristis for eh rop yer. MNT ws more effiient for rop yer /3 reduing the oil lost by.% wheres for / tht redution ws lower sine eptble oil yields were obtined for ll the tretments. Although rop yer / hd very diffiult pstes, MNT redued the oil losses by.%, the high oil losses indited tht higher MNT should hve been dded. Proessing unripe Piul fruits required dpted proessing onditions sine beuse of their own hrteristis ll the olive pstes n be onsidered s diffiult even very diffiult pstes. In generl eptble oil yields were obtined under the mlxtion onditions ssyed sine vlues of pome oil losses were lower thn those desribed previously for two phses system (Di Giovhino et l., ; Inrejos-Gri et l., 9). In generl,

M.P. Aguiler et l. / Journl of Food Engineering 97 () 533 53 537 * min 9 min / * MNT % MNT % * ** ºC 3 ºC ºC / / /3 Men min Crop yer 9 min * / Fig.. Effet of Mironised Nturl Tl (MNT) ddition on pome oil ontent on dry weight for three onseutive rop yers: /, / nd /3. Different letters men signifint differenes t *p:.5, **p:. nd ***p:.. ** by.3%. Considering the worst pste rheologil onditions, moisture 5%, for eh temperture the oil yield ws improved by both longer mlxtion time nd MNT ddition. The different mlxtion onditions required for eh fruit hrteristis, mkes neessry the development of potent tools for olive fruit hrteristion on line in order to dpt the proess to the olives. Aknowledgements ºC 3 ºC ºC min 9 min ºC 3 ºC ºC temperture let to improve their oil yields just for C thn n hve negtive effet on oil qulity. Longer mlxtion redued by 7% the pome oil ontent while MNT ddition (%) lowered it /3 Fig. 3. Intertion between mlxtion time nd temperture for pome oil ontent on dry weight from erly hrvesting fruits for three onseutive rop yers: /, / nd /3. Different letters men signifint differenes t *p:.5, **p:. nd ***p:.. This work ws supported by the Projet CAO-9 from Progrm de Mejor de l Clidd del Aeite de Oliv, Junt de Andluí (FEDER-FEOGA) nd Projet FEDER-INIA RTA-- C3- nd is prt of the PhD Thesis of Dr. Mrí de l Pz Aguiler Herrer. The uthors grteful knowledged to Mr. Jun Torres the oil mill mnger by its helpful during the experiments. Referenes Brrno, D., Trujillo, I., Rllo, L., 5. Libro I. Eliogrfí hispni. In: Rllo, L., Brrno, D., Cbllero, J.M., Del Rio, C., Mrtin, A., Tous, J., Trujillo, I. (Eds.), Vrieddes de olivo en Espñ. Junt de Andlui-MAPA-Ediiones MundiPrens, Mdrid, pp. 5 3. Beltrn, G., Ued, M., Jiménez, A., Aguiler, M.P., 3. Olive oil extrtbility index s prmeter for olive ultivr hrteristion. Journl of the Siene of Food nd Agriulture 3, 53 5. Beltrn, G., Ued, M., Hermoso, M., Fris, L.,. Mdurion. In: Brrno, D., Fernndez-Esobr, R., Rllo, L. (Eds.), El ultivo del olivo. Junt de Andlui- Ediiones MundiPrens, Mdrid, pp. 5 3. Di Giovhino, L., 9. L impiego del tlo mironizzto nell estrzione dell olio dlle olive on il sistem ontinuo. Rivist Itlin delle Sostnze Grsse 5, 3. Di Giovhino, L., 99. L estrzione dell olio on l entrifugzione dirett delle pste di olive. Not I: influenz dell grmolzione. Rivist Itlin delle Sostnze Grsse, 3 3. Di Giovhino, L., Constntini, N., Ferrnte, M.L., Serrioo, A.,. Influene of mlxtion time of olive pste on oil extrtion yields nd hemil nd orgnolepti hrteristis of virgin olive oil obtined by entrifugl denter t wter sving. Grss y Aeites 53, 79. Di Giovhino, L., Sestili, S., Di Vinenzo, D., b. Influene of olive proessing on virgin olive oil qulity. Europen Journl of Lipid Siene nd Tehnology, 57. Europen Union Commission Regultion 5/9, 99. Chrteristis of Olive nd Olive Pome Oils nd Their Anlytil Methods, vol. L. Offie of Joint Commission Europen Communities, pp.. Hermoso, M., Ued, M., Gri-Ortiz, A., Morles, J., Frís, L., Fernndez, A., 99. Elborión de eite de oliv de lidd. Obtenión por el sistem de dos fses. Junt de Andlui, Seville.

53 M.P. Aguiler et l. / Journl of Food Engineering 97 () 533 53 Inrejos-Gri, A.M., Gomez-Rio, A., Slvdor, M.D., Fregpne, G., 9. Influene of mlxtion onditions on virgin olive oil yield, overll qulity nd omposition. Europen Food Reserh nd Tehnology, 7 77. Mrtinez, J.M., Gomez, C., Jner, C., 957. Estudios fisio-químios sobre ls psts de eitun molids. IV. Ls gots de eite. Grss y Aeites,. Mrtinez, J.M., Muñoz, E., Alb, J., Lnzon, A., 975. Informe sobre l utilizión del nlizdor de rendimientos Abenor. Grss y Aeites, 379 35. Muñoz, E., Alb, J., Lnzon, A., 979. Informe de l Almzr Experimentl del Instituto de l Grs orrespondiente l mpñ 97 79. Grss y Aeites 3, 5.