This is an author-deposited version published in: Eprints ID: 6245

Transcription

1 Open Archive Toulouse Archive Ouverte (OATAO) OATAO is an open access repository that collects the work of Toulouse researchers and makes it freely available over the web where possible. This is an author-deposited version published in: Eprints ID: 6245 To cite this document: Gerbaud, Vincent and Teles dos Santos, Moises and Carrillo Le Roux, Galo Vegetable oil modeling applied for the formulation of structured lipids. (22) In: Journées chevreul 22 : Chimie du végétal et lipochimie, 5-6 Jun 22, Maison Alfort, France. Any correspondence concerning this service should be sent to the repository administrator: staff-oatao@inp-toulouse.fr

2 Vegetable oil modelingapplied for the formulation of structured lipids Vincent Gerbaud Laboratoire de Génie Chimique, Toulouse Moises Teles Dos Santos, Galo Carillo Le Roux Universidad de Sao Paulo (Brasil)

3 Outline - Motivations - SFC : Solid fat content : the core property - Vegetable Oil modeling Analysis Development - Results Binary TAGs mixture Four TAGs mixture Vegetable oil: 7 TAGs and more.. Oil blends, interesterified or not Structured lipids application - Conclusions 2

4 - Growing need for TAG based products = more systematic tools for their design Motivations Solvents Waxes Vegetable Oil Our model concerns food & industrial uses Cutting fluids Cosmetics Lubricants ROSILLO-CALLE, F. et al. A global overview of vegetable oils with reference to biodiesel. Report for the IEA Bioenergy Task 4 (29) Foods 3

5 Solid Fat Content: the core property - Many evidences that end-user attributes are correlated with Solid Fat Content Flöter. The role of physical properties data in product development. Eur. J. Lipid Sci. Technol. 29,,

6 Solid Fat Content: the core property End-user spreadability stability texture shine Flavor Attributes (end-user perception) Macroscopic properties Physical Chemical Microstructure Storage Processing SFC Polymorphism The molecular scale Composition FA TAGs Veg.Oil / Fat Natural sources 5

7 Vegetable Oil modeling - Datas are not uniform and consistent across the scales. - Predictive models did not exist so far. Data spreadability stability texture shine Flavor Attributes (end-user perception) Data (SP, T m ) Macroscopic properties Physical Chemical (IV) QSPR correlations Storage Microstructure Processing Data (NMR) SFC Polymorphism Data Data (FAME) Composition FA TAGs Veg.Oil / Fat Data Data Natural sources 6

8 - A predictive model for the missing links: SFC vs Veg. Oil composition By using FA distribution or TAGs composition Vegetable Oil modeling? SFC vs Veg. Oil composition SFC Polymorphism Composition FA TAGs Veg.Oil / Fat 7

9 Vegetable Oil modeling - Model analysis SFC calculation solid liquid equilibrium (SLE) Minimizing the gibbs free energy G A very complex mixture pb Veg. Oils blends m Veg. Oils > 95% of n TAGs combination of 3 FA Each scale give rise to phase diagrams, SFC, DSC Polymorphism affects the SLE SFC Polymorphism Composition FA TAGs Veg.Oil / Fat Liq. - The final model must be predictive to help oil blend design Solids TAG SLE 8

10 The SLE model - Model INPUT: Veg. Oil considered as a mixture of n TAGs + LIQUID α β β LIQUID LIQUID + α β β α β β + +. Getting TAGs composition? (rare) from TAGs analysis (frequent) from FA distribution from which we generate a probable TAGs distribution - Model OUTPUT: The TAGs distribution among the phases liquid and solids vs temperature Enable to compute SFC and DSC curves 9

11 - Polymorphic solid(s) liquid equilibrium The SLE model min G( n) with G( n) = nc np i= j= n j i g j i ( n) g g solid ( j) liquid = RT x = RT nc nc i= i= - Results: ( x solid ( j) i liquid i H m, i R ln x liquid i solid ( j) ) T solid ( j) m, i + ln T Predicted for each phase H m & Tm = f ( polymorhic, FAtype, FA position) Solid Fat Content obtained directly from n TAG phase j solid ( j) solid ( j) ( γ x ) i i Function of TAGs composition (equilibrium) DSC curves from: C app p = C p E G + T + np nc j= i= Hm j i n i j T

12 7 SLE Results: PPP/OOO vs PPP/CCC - Molecular size greatly affects crystal lattice and the SLE! Phase diagram Molar fraction (OOO) P: palmitic acid (C6:) / C: capric acid (C:) / O: oleic acid (C8:) Cp [kj/mol.k] DSC Curve OOO PPP + CCC PPP solid molar fraction solid molar fraction SFC SFC CCC (C:) PPP (C6:) OOO (C8:) PPP (C6:)

13 25% OOO 25% CCC 25% PPP 25% SSS mixture - Capturing the complexity of TAG behavior in mixture SFC (mass) SFC Cp (kj/k) 5 5 DSC Curve OOO + CCC PPP SSS mols on solid state/total of mols PPP SSS CCC OOO mols on liquid state/total of mols PPP SSS CCC OOO OOO and CCC melt together over the same T range PPP and SSS melt as pure compounds P: palmitic acid (C6:) / S: estearic acid (C8:) / C: capric acid (C:) / O: oleic acid (C8:) 2

14 - Palm oil: 7 TAGs. SFC (mass fraction) SFC Vs Temperature Cp [kj/mol.k] Veg. Oil: Palm oil Simulated DSC Melting Point correctly calculated Good SFC agreement, considering that we used two incoherent data sources Exp. Data: TAG composition from Sambanthamurthi R et al. (2) Exp. Data: mean SFC from Lin, S.W. (22) 3

15 Veg. Oil: Palm oil - DSC: temperature profile and main phase transitions We solve S- L Equilibrium = we get equilibrium DSC = DSC with infinitely slow heating rate.6.5 Experimental ( ºC/min).4 Experimental (5 ºC/min) Simulated.3.2 PamPamPam Temperature ºC C/min T7 = 4,5 C 4

16 - Canola Oil (3%) Fully Hydrogenated Palm Oil Stearin (7%) No reaction: 96 TAGs After interesterification: 62 TAGs - Analysis: Oil blends and interesterified blends Both physical & interesterified blends predictions agree with experimental data Predicts the trend at lower temperature Reaction makes peaks different and broader SFC (mass fraction) Cp [kj/mol.k] Before reaction After reaction

17 Ternary Veg. Oil blend - Influence of interesterification SFC is higher when SFO fraction is low Interesterification greatly reduces the SFC Palm Oil (PO) Sunflower Oil (SFO) Palm Kernel Oil (PKO) (SFO) Solid Fat Content (SFO) Solid Fat Content (PKO) Not interesterified PREDICTED in 2 hrs (PO) (PKO) T = ºC Interesterified (PO) 6

18 Structured Lipids Application - Designing functional foods to get nutraceutical benefits ex. DHA enhanced products sn2 Medium FA TAG TAG2 TAG3 TAG4 sn caprylic caprylic caprylic caprylic PUFA Medium FA sn2 EPA DHA γ-linoleic AA Omega-3 Omega-3 Omega-6 Omega-6 sn3 caprylic caprylic caprylic caprylic EPA C2:5 Arachidonic Acid C2:4 γ-linoleic C8:3 DHA C22:6 7

19 Structured Lipids Application - High melting SL affects palm oil initial melting T pure palm oil SFC (mass fraction) palm oil + SL (3%): C8-EPA-C8 palm oil + SL2 (3%): C8-DHA-C8 palm oil + SL3 (3%): C8-g-linolênico-C8 palm oil + SL4 (3%): C8-AA-C Teles dos Santos, Gerbaud, Le Roux. Phase Equilibrium and Optimization Tools: Application for Enhanced Structured Lipids for Foods. J Am Oil Chem Soc (2) 88:

20 Structured Lipids Application - The effect on SFC depends on the T range Palm oil + (Caprylic EPA Caprylic) SFC (mass fraction) pure palm oil palm oil + SL 5% palm oil + SL % palm oil + SL 2% palm oil + SL 3% palm oil + SL 4% palm oil + SL 5% Teles dos Santos, Gerbaud, Le Roux. Phase Equilibrium and Optimization Tools: Application for Enhanced Structured Lipids for Foods. J Am Oil Chem Soc (2) 88:

21 - Computer aided design of oil blends Conclusions Finding the blend that matches targeted SFC and DSC Objective: to anticipate the most promising lab experiments n to - A knowledge-based predictive model Solid (multi) liquid equilibrium based model Uses FA distribution or TAGs composition data FULLY PREDICTIVE model for SFC & equilibrium DSC curves Accuracy is good We need coherent TAG & SFC data to prove that it could be even better Valid for any TAG mixture: binary, ternary, vegetable oil, oil blend, interesterified oils blend Not valid for binary FA mixture Teles dos Santos, Gerbaud, Le Roux. Phase Equilibrium and Optimization Tools: Application for Enhanced Structured Lipids for Foods. J Am Oil Chem Soc (2) 88:

22 - Composition in each phase SLE Results: 4 TAGs mixture Solid phase is gradually enriched with SSS; OOO fraction gradually decreases in liquid phase mols on solid state/total of mols PPP SSS CCC OOO mols on liquid state/total of mols PPP SSS CCC OOO mols on solid/total of mols on solid PPP SSS CCC OOO mols on liquid/total of mols on liquid PPP SSS CCC OOO

23 eq - DSC theoretical prediction Vs Experimental Data Palm Oil C/min T7 = 4,5 C A FA-based model could not capture that C.P. Tan, Y.B. Che Man. Differential scanning calorimetric analysis of palm oil, palm oil based products and coconut oil: effects of scanning rate variation. Food Chemistry 76 (22)

24 eq - DSC theoretical prediction Vs Experimental Data Transition Temperatures (ºC) Exp. Calc. Exp. Calc. Exp. Calc. Exp. Calc. Exp. Calc. Peanut Oil -5,7 nc* -29, ,57 -,83 2 8,26 7 Grapeseed Oil -39, , , , Exp Data.: C.P. Tan and Y.B. Che Man. Differential Scanning Calorimetric Analysis of Edible Oils: Comparison of Thermal Properties and Chemical Composition. JAOCS, Vol. 77, no. 2 (2) Calc.: calculated by VODesign *nc: no convergence 23

25 Single VO: Cocoa Butter - Cocoa butter: 2 TAGs. High accuracy for lower temperatures Shorter melting range corrected detected time elapsed: 26s.9 SFC Vs Temperature PREDICTED 9 Simulated DSC PREDICTED SFC (mass fraction) Cp [kj/mol.k] Exp. Data: WON, K. W. THERMODYNAMIC MODEL OF LIQUID-SOLID EQUILIBRIA FOR NATURAL FATS AND OILS. Fluid 24 Phase Equilibria [S.I.], v. 82, p , 993.

26 Unconventional oils - Milkfat corn oil (8-2) interesterified : 45 TAGs Simulated and experimental results in agreement (T > 2 ºC) Time: 5 min 37 s SFC (mass fraction) Cp [kj/mol.k] (a) (b) MilkCorn_82.txt 25 Exp. Data: RODRIGUES, J. N.; GIOIELLI, L. A. Food Research International, Journées v. 36, Chevreul n. 2, p , 22 : chimie 23. du végétal et lipochimie

27 Unconventional oils - Theoretical prediction for value-added vegetable oil Grapeseed Oil: TAGs SFC (mass fraction) Cp [kj/mol.k] (a) (b) 26

28 Influence of Temperature: ternary blend not interesterified PO SFO PKO.8.45 C C C 25 C C 35 C PREDICTED 27

29 Influence of Temperature: ternary blend interesterified.5 PO SFO PKO C.9 C 5 C C C 35 C.6 PREDICTED 28