The Role of Physical Properties Data in Product Development

Transcription

1 The Role of Physical Properties Data in Product Development From Molecules to Market Gent June 18 th and 19 th 2008 Eckhard Flöter Unilever R&D Vlaardingen

2 Physical Properties Data Melting point, heat of fusion, solubility Crystallization behaviour characeristics Kinetics Polymorphic form Habit, seize and shape Mixing behaviour, molecular interactions Purity Others: specific heat, density, diffusivity,... We have little and no overall compilation

3 Intro Less answers, more questions Current reality of product development Rules of thumb versus first principles Do we understand what we do? Stimulate to think about the way to bridge the gap

4 Role of Fat Crystals emulsion stabilisation oil binding bulk hardness 60% fat spread liquid margarine

5 Hirachy of Product Structure Building the perfect sponge Not marketing induced consumer perception macroscopic properties storage & cycling microstructure processing SFC polymorphism composition

6 The old story of the SFC, first approximation spreadability 75 fridge kitchen stability solid fat content (%) butter wrapper margarine soft tub margarine oral perception temperature ( C) Figure 4

7 Linear Programming, old but still most practical Applicability in field of operation, meaningless but advanced rule of thumb

8 Equilibrium Modelling Advanced approach presented by Elias today Physical properties of pure TAGs components (dh fus and T m ) Non-ideality mixed solid phases via standard Gex -model choice of model and derivation of interaction parameters large number of species, necessity for lumping (oil and gas industry) assumption of ideally mixed liquid phase ln x γ liq i,eq sj sj i x i = ΔH R sj fus,i 1 T 1 T i sj Minimisation of the total Gibbs free energy, => SFC + composition

9 DSC Simple pure systems already give us a headache

10 Need of comprehensive good input Initial Group Contribution Method ΔH = H + H + f int gly H mep ΔS = S + S + f int gly S mep T m = ΔH ΔS f f H int H gly H mep lateral interaction energy glycerol energy methyl end plane (MEP) energy Triple CLS Double CLS

11 Arbitrary Choices, any better than LP? 1. The lateral interaction energy q r p q I qr I qp A) Double CLS H = qp H min( p, q ) min( q r p r r p, q ) p H int = H qp + H qr I qr I qp B) Triple CLS p r q p r Mixed FA layer Mono FA layer p r q p r r * Same lateral interaction energy * Different MEP * Differentiation of the two layers

12 Analysis of limited available data Rules of thumb Structure Prediction Tree for Calculations Saturated TAGs symmetric p q r asymmetric p p+n p p p p+n p q r Segregation of shortest chain (min (p,q,r)) Triple when max-min 8-6 n 4 double 1,3 2 * H p+n p gly d 1,3 mep d sym= f(n) n 6 triple 1,3 H pp +H p+n p+n gly tr 1,3 mep tr n -8 triple 1,3 H pp +H p+n p+n gly tr 1,3 mep tr -4 n 6 double 1,3 H pp +H pp+n gly d 1,3 mep d asym= l(n) n 8 triple 1,2 H pp +H p+n p+n gly tr 1,2 mep tr n -6 triple 1,2 H pp +H p+n p+n gly tr 1,2 mep tr

13 Significant Improvement for dh, ds and T With good results ΔHβ (kj/mole) predicted values new GC method FBFS TAG based predictions experimental values

14 Good results, with doubts Description of Polymorphic Transition kinetics of milk fat Correlation of transition times of milk fat containing compositions -set of 6 milk fats & fractions and vegatable oil, 150 blends -isothermal stagnant crystallisation at 5 temperatures -identification of transition time as maximum rate dsfc/dt parity plot tau (15C) predicted transition time 1/s 3000 But statistical contribution of similar fractions differs!! experimental transition time 1/s

15 Rules of thumb often, often not good enough Plenty of examples of surprises Milk fat often successfully treated as ternary system High melting Middle melting Low melting fraction HOWEVER

16 Crystallisation in Batch Solid fat content as a function of time during crystallisation of milk fat olein at θ c =18 C. 10 reduced growth rate Process time (h)

17 Crystallisation in Batch Microscopic pictures after 17h (left) and 22h (right) of milk fat olein crystallisation at θ c =18 C (see ). Image width 745μmC before plateau after plateau occurrence of new nuclei no growth of initial crystals

18 Isothermal Crystallisation under Shear milk fat olein at a constant temperature of 18 C viscosity as a function of time (shear rate 1000rpm). melt inserted into rheometer 50 C (Haake RN3.1, concentric cylinder H1) Increased crystal content η Temperature equilibration 6% of crystals present Time (s)

19 Isothermal Crystallisation under Shear Isothermal crystallisation at different temperatures value (6%) in agreement with batch crystalliser plateau Viscosity η (mpas) C 17 C 8.5% 18 C solid fat content at first constant viscosity plateau independent of crystallisation temperature 5.8% 5.6% 6.2% Time t (s)

20 X-ray diffraction at intermediate and final plateau Both plateaus same polymorph β nucleation of the separate second group of TAG s independent (induction time and dsfc) Not shown second step is no polymorphic transition second plateau 18 C first plateau 18 C molten 25 C Intensity *Theta Both in the β' form (4.24 Å and 3.80 Å)

21 Conclusions and Recommendations Stepwise crystallisation process is no polymorphic transition Solid immiscibility of TAG groups of milk fat olein identified Separate primary nucleation steps in various milk fat fractions Phenomenon is more pronounced in low-melting fractions. Viscosity good measure for nucleation process Even though milk fat is so complicated that it almost behaves as an ideal mixture it remains much more complicated if we take a closer look

22 Emulsion stability, break up other ingredients emulsion stabilisation oil binding bulk hardness 60% fat spread liquid margarine

23 Solubility and dissolving Simple first approximation - knowledge of basic physical properties heat of fusion, melting point - assume natural mixed system behaves as pure component ln x γ liq i,eq sj sj i x i = ΔH R sj fus,i 1 T 1 T i sj OR a few simple measurements

24 Sterols - Basic structure / Properties β-sitosterol OH β-sitosterol ester O O

25 Adding sterols Storage at 5C, 3 days

26 Characterization of sterol C 12 -SE at 80 C on cooling Linkam Optical Microscopy - C 12 sterol ester

27 C 12 -SE at 61 C on cooling

28 C 12 -SE at 55 C on cooling

29 C 12 -SE at 55 C after 20 minutes

30 Conclusions Sterol Esters Structuring/gelling combinations can be found Mechanism not yet understood Sterol ester characterization of pure SE Complex crystallization behaviour Real systems highly mixed, less problematic

31 The role of emulsifiers and how much we do not know Mainly HLB value as initial yard stick Dissolution behaviour Monoglycerides in miix with lecithins Kinetics of dissolution Competition at the interface emulsifier/emulsifier Emulsifier and Pickering stabilization through solid fat

32 Hirachy of Product Structure Building the perfect sponge Not marketing induced consumer perception macroscopic properties storage & cycling microstructure processing SFC polymorphism composition

33 Finally Products going to market are complex systems full of surprises The link between academic systems and these real systems is far from trivial However Study of understandable systems helps us to interpret and sharpen our insights Needed to formulate hypothesis explaining observations Guidance to more sound quick and dirty improvement strategies supported by rules of thumb and experience