![Page 1: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/1.jpg)
Optimization of caramel flavour generation upon extrusion cooking
13th Weurman Flavour Research Symposium
Zaragoza, September 27-30, 2011
generation upon extrusion cooking
Tomas Davidek, Silke Illmann, Andreas Rytz, Hélène Chanvrier, Greet Vandeputte, Heike
Schuchmann, Josef Kerler and Imre Blank
Nestle PTC Orbe
![Page 2: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/2.jpg)
Flavour is one of the major drivers of consumer preference
However, there are many other factors of similar importance
Colour, structure, texture, Nutritive value, safety, Convenience, packaging, …
2
Flavouring can be achieved by
- adding compounded and/orreaction flavours
- generating flavours upon food processing, e.g. extrusion
Louis-Leopold Boilly, Les Cinq Sens (The Five Senses), 1823.
![Page 3: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/3.jpg)
Extrusion with direct expansionProduct structure created by extrusion
Flour Mix (~10% H2O)
Water, Oil, ...
The cereal is expanded to its
3
SteamExpansion
10%
~15–20%
~5-10%
Moisture:
110–180°C80–170 bars
Average residence time: 20–30 seconds
The cereal is expanded to its final size at the extruder outlet
Downstream processing
![Page 4: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/4.jpg)
Flavour generation upon extrusion
Extrusion Heat load (T, t) Screw speed, SME Moisture Number of barrels Slurry vs. dry addition…
Recipe Ingredients Specific precursors Concentration, ratio Catalyst pH…
4
……
Manifold interactions, Systematic & multidisciplinary Effects difficult to predict approach, Experimental design
O
OHO
CH3CH3
Favourable conditions
• low moisture• high pH• T > 120°C
4-Hydroxy-2,5-dimethyl-3(2H)-furanone (Furaneol) as a target compound generated from sugars
![Page 5: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/5.jpg)
Objective
Study the effect of extrusion parameters and recipe composition on furaneol formation from rhamnose and lysine
NH2 OH
O
NH2 O
OHO
CH3CH3
+ Recipe & Process
parameters
5
Identify recipe and processing conditions during extrusion of rice flour favouring the formation of caramel flavour while considering physico-chemical properties of the final poduct
Optimised recipe and processing conditions
![Page 6: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/6.jpg)
Experimental setup: Key product attributes affected by recipe and extrusion parameters
Recipe parameters • pH
• Ratio Rha/Lys
• Phosphate
Extrusion parameters• Moisture levels
• Screw speed
• Temperature
• Residence time
Mixing
Extrusion 500400300Screw speed (rpm)
150135120Temp. (°C)
0.1340.035PO4 (mol/kg)
232017Moisture (%)
75Barrel length
3:13:0Rha:Lys
7.76.4pH
SME
6
• Residence time
• Slurry vs. dry
Drying
Milling
Final product
Product characterisation
• Texture
• Crispness
• Colour
• Flavour
• Acrylamide
• Starch degradation
• Granulometry
• Viscosity
• Sensory
Fractional factorial design:- 32 instead of 576 trialsModelling furaneol yield & rhamnose degradation by estimating - all main effects and - two-factor interactions of 3 recipe & 5 process parameters
SlurryDryAddition
![Page 7: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/7.jpg)
Holistic approach: Analytical characterisation of samples from experimental design
List of analysis:• Residual rhamnose (HPLC/DAD)• Furaneol (GC/MS)• Acrylamide (HPLC/MS)• Chemical composition (NIR)• Moisture (MNR)• Molecular weight profile of starch (GFAS = Gel Filtration Analysis of Starch)
Flavour Recipe/process parameters Other key product attributes
7
• Molecular weight profile of starch (GFAS = Gel Filtration Analysis of Starch) • Molecular weight profile of debranched starch (SEC = Size Exclusion
Chromatography) • Molecular weight profile of proteins (HPSEC = High Performance Size
Exclusion Chromatography) • Colour• Viscosity (rheometer, RVA profile)• Granulometry (laser diffraction)• Texture: X-ray tomography
• Porosity (cells size & cell walls distributions, degree of anisotropy)• Crispiness (average drop off, force Max)
• Texture: Spoon test• Sensory evaluation
![Page 8: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/8.jpg)
• Trained panel (14 panelists)
• Product– Extruded powder (85%)
– Sugar (15%)
• Reconstitution– 12.4g product
– 100mL milk at 70°C
Sensory assessment Reconstituted products from experimental design
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6
tEasySwallowaLump
tSmooth
BurntNutty
aDarkCaramelToasted
Range covered by A01-A32 (vs. REF)
Text
ure
& A
pp
eara
nce
Range of sensory attributes
Wide diversity of sensory perception
8
– 100mL milk at 70°C
• 32 Products evaluated vs. reference
• Identification of statistically relevant trends
Parameters units Low Medium High
pH 6.4 7.7Rha:Lys 3:0 3:1Phosphate mol/kg 0.035 0.134Moisture % 17 20 23Screw speed rpm 300 400 500T °C 120 135 150length barrel short longAddition dry slurry
ToastedProcessy
OffOverall
acidWholeGrain
Mushroom
astringentbitter
aftertaste
t1Thick
tThicktFluffy
t1WettabilitytSemolina
CookedMilkyRice
Vanillasweet
Fla
vor
Text
ure
& A
pp
eara
nce
![Page 9: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/9.jpg)
16
20
Fu
ran
eo
l dry
/ m
ol%
Correlation matrixNo correlation between Furaneol level and SME
32 samples produced in experimental design
Resulting in products with sufficiently different properties
9
0
4
8
12
20 30 40 50 60 70 80 90 100
SME measured / Wh/kg
Fu
ran
eo
l dry
/ m
ol%
![Page 10: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/10.jpg)
Correlation matrixFuraneol vs SME
ab
C
Furaneol ext
Furaneol dry
12
16
20
Fur
ane
ol d
ry (
mo
l%)
Furaneol is positively correlated with the degradation of rhamnose and colour development
10
SME measured
SME calculated
Visco5min
ViscoMax
ViscoMaxTime
Visco5/3min
L
h
Rha extRha dry
0
4
8
12
20 30 40 50 60 70 80 90 100
SME measured (Wh/kg)
Fur
ane
ol d
ry (
mo
l%)
Acrylamide
SME is negatively correlated with viscosity
and colour development
![Page 11: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/11.jpg)
Furaneol formation Free AA, moisture, T and phosphate are most critical
-6 -4 -2 0 2 4 6
pH 6.4pH 7.7
Rha:Lys 3:0Rha:Lys 3:1
Phosphate 0.035 mol/kgPhosphate 0.134 mol/kg
Furaneol ext / mol% Furaneol dry / mol%
Conversion of rhamnose to furaneol can be modulated through changes of extrusion and/or recipe
11
17% H2O20% H2O23% H2O
300 1/min400 1/min500 1/min
120 °C-Long120 °C135 °C150 °C
DrySlurry
changes of extrusion and/or recipe parameters
![Page 12: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/12.jpg)
Acrylamide Temperature is most critical
-75 -50 -25 0 25 50 75
pH 6.4pH 7.7
Rha:Lys 3:0Rha:Lys 3:1
Phosphate 0.035 mol/kgPhosphate 0.134 mol/kg
Acrylamide
-6 -4 -2 0 2 4 6
pH 6.4pH 7.7
Rha:Lys 3:0Rha:Lys 3:1
Phosphate 0.035 mol/kgPhosphate 0.134 mol/kg
Furaneol ext / mol% Furaneol dry / mol%
12
17% H2O20% H2O23% H2O
300 1/min400 1/min500 1/min
120 °C-Long120 °C135 °C150 °C
DrySlurry
17% H2O20% H2O23% H2O
300 1/min400 1/min500 1/min
120 °C-Long120 °C135 °C150 °C
DrySlurry
![Page 13: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/13.jpg)
Colour generation
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6
tEasySwallowaLump
tSmooth
Range covered by A01-A32 (vs. REF)
Colour range
Colour developmentModulate colour in product through recipe
Browning reactions
pH 6pH 7
Colour (C)
* High pH
pH 6Rha/Lys 3:0
pH 7Rha/Lys 3:1
13
tSmooth
BurntNutty
aDarkCaramelToasted
ProcessyOff
Overallacid
WholeGrainMushroomastringent
bitteraftertaste
t1ThicktThicktFluffy
t1WettabilitytSemolina
CookedMilkyRice
Vanillasweet
Sen
sory
im
pac
t
Colour range
-6 -4 -2 0 2 4 6
pH 7
Rha:Lys 3:0Rha:Lys 3:1
Phosphate lowPhosphate high
17% H2O20% H2O23% H2O
300rpm400rpm500rpm
120°C-Long120°C135°C150°C
DrySlurry
** Lysine
* High pH
Key
par
amet
ers
Light Dark
![Page 14: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/14.jpg)
SME Temperature, moisture, screw speed are most critical
-15 -10 -5 0 5 10 15
pH 6.4pH 7.7
Rha:Lys 3:0Rha:Lys 3:1
Phosphate 0.035 mol/kgPhosphate 0.134 mol/kg
SME measured / Wh/kg SME calculated / Wh/kg
No effect of recipe parameters on SME
14
17% H2O20% H2O23% H2O
300 1/min400 1/min500 1/min
120 °C-Long120 °C135 °C150 °C
DrySlurry
Higher SME is achieved with low moisture, high screw speed, low temperature and long extruder.
![Page 15: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/15.jpg)
Structure and Texture Moisture and temperature are most critical
-1500 -1000 -500 0 500 1000 1500
pH 6.4pH 7.7
Rha:Lys 3:0Rha:Lys 3:1
Phosphate 0.035 mol/kgPhosphate 0.134 mol/kg
Visco5min / mPas ViscoMax / mPas
-15 -10 -5 0 5 10 15
Cell walls / µm
Structure = f (SME)Texture = f (SME)
ProcessProcess
15
Phosphate 0.134 mol/kg
17% H2O20% H2O23% H2O
300 1/min400 1/min500 1/min
120 °C-Long120 °C135 °C150 °C
DrySlurry
![Page 16: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/16.jpg)
StructureWide variety of structures obtained
Rha:Lys 1:0Rha:Lys 3:1
0.035 mol/kg PO40.134 mol/kg PO4
17% H2O20% H2O
Cell walls (um)
600
800
1000
1200M
ean
cel
l siz
e (u
m)
16
-20 -10 0 10 20
20% H2O23% H2O
300 1/min400 1/min500 1/min
120 °C-Long120 °C135 °C150 °C
DrySlurry
Main effect
0
200
400
600
70 75 80 85 90
Porosity (%)
Mea
n c
ell s
ize
(um
)
12 mm
![Page 17: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/17.jpg)
Long ChainDP~50
Short ChainDP~25
Rice Flour
Moisture % 8-13
Starch %d.b. 88-92
Total fibers %d.b. 0.3-1Proteins %d.b. 6-9Fat %d.b. 0.5-1Ash %d.b. 0.3-1
A28
40
50
60
70
80
90
100
Car
bo
hyd
rate
s (g
/100
g)
Total Amylopectin Intermediate1
Intermediate2 Amylose
A04
0
10
20
30
40
50
60
70
80
90
100
1000 10000 100000 1000000 10000000 100000000
Molecular weight
Car
bo
hyd
rate
s (g
/100
g)
Total Amylopectin Intermediate1
Intermediate2 Amylose
Amylopectin
High MW
Am
ylo
pec
tin
, h
igh
MW
In
term
edia
te d
epo
lym
eriz
ed
Lo
w M
W In
termed
iate gen
erated, am
ylose co
nstan
t
SME 92 SME 35
Texture: Starch degradation Depolymerisation of amylopectin increases with SME
17
A17
0
10
20
30
40
50
60
70
80
90
100
1000 10000 100000 1000000 10000000 100000000
Molecular weight
Car
bo
hyd
rate
s (g
/100
g)
Total Amylopectin Intermediate1
Intermediate2 Amylose
A31
0
10
20
30
40
50
60
70
80
90
100
1000 10000 100000 1000000 10000000 100000000
Molecular weight
Car
bo
hyd
rate
s (g
/100
g)
Total Amylopectin Intermediate1
Intermediate2 Amylose
0
10
20
30
40
1000 10000 100000 1000000 10000000 100000000
Molecular weight
Car
bo
hyd
rate
s (g
/100
g)
Am
ylo
pec
tin
, h
igh
MW
In
term
edia
te d
epo
lym
eriz
ed
Lo
w M
W In
termed
iate gen
erated, am
ylose co
nstan
t
Low MW
Amylose
SME 92 SME 35
![Page 18: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/18.jpg)
A28
40
50
60
70
80
90
100
Car
bo
hyd
rate
s (g
/100
g)
Total Amylopectin Intermediate1
Intermediate2 Amylose
A04
0
10
20
30
40
50
60
70
80
90
100
1000 10000 100000 1000000 10000000 100000000
Molecular weight
Car
bo
hyd
rate
s (g
/100
g)
Total Amylopectin Intermediate1
Intermediate2 Amylose
Amylopectin
High MW
Am
ylo
pec
tin
, h
igh
MW
In
term
edia
te d
epo
lym
eriz
ed
Lo
w M
W In
termed
iate gen
erated, am
ylose co
nstan
t
SME 92 SME 35
0
20
40
60
80
100
0 50 100SME
Sta
rch
(%
)
Amylopectin
High MWintermediate
Low MW intermediate
Starch degradation MW = f (SME)
0
1000
2000
3000
4000
5000
6000
0 2 4 6 8 10 12V
isco
sity
(m
Pa.
s)Time (min)
SM
E 9
2S
ME
35RVA 50°C, 50rpm
21g+170mL milk
Viscosity = f (MW)
18
A17
0
10
20
30
40
50
60
70
80
90
100
1000 10000 100000 1000000 10000000 100000000
Molecular weight
Car
bo
hyd
rate
s (g
/100
g)
Total Amylopectin Intermediate1
Intermediate2 Amylose
A31
0
10
20
30
40
50
60
70
80
90
100
1000 10000 100000 1000000 10000000 100000000
Molecular weight
Car
bo
hyd
rate
s (g
/100
g)
Total Amylopectin Intermediate1
Intermediate2 Amylose
0
10
20
30
40
1000 10000 100000 1000000 10000000 100000000
Molecular weight
Car
bo
hyd
rate
s (g
/100
g)
Am
ylo
pec
tin
, h
igh
MW
In
term
edia
te d
epo
lym
eriz
ed
Lo
w M
W In
termed
iate gen
erated, am
ylose co
nstan
t
Low MW
Amylose
SME 92 SME 35
High H2OHigh T
Low H2OLow T
![Page 19: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/19.jpg)
Effect of recipe/extrusion parameters Options for flavour optimization
Rha:Lys 3:1
Phosphate high
17% H2O
500rpm
150°C
1
2
3
4
%F
ura
ne
ol-
dry
Temperature affects
Free amino acid and phosphate affect furaneol, but not SME
19
pH 6pH 7
Rha:Lys 3:0
Phosphate low
20% H2O
23% H2O
300rpm
400rpm
500rpm
120°C-Long
120°C135°C Dry
Slurry
-4
-3
-2
-1
0
-20 -15 -10 -5 0 5 10 15 20
SME-measured
%F
ura
ne
ol-
dry
Temperature affects both furaneol and SME150°C → more viscous
Moisture affects both furaneol and SME17% → less viscous
![Page 20: Optimization of caramel flavour generation ... - Imre Blank · generation upon extrusion cooking 13th Weurman Flavour Research Symposium Zaragoza, September 27-30, 2011 Tomas Davidek,](https://reader031.vdocuments.us/reader031/viewer/2022011915/5fc94a4c6ff4323eac4006b3/html5/thumbnails/20.jpg)
Conclusions
Critical parameters for generation of furaneol from rhamnose during extrusion:
• Furaneol generation is affected both by recipe parameters (Lys, PO4) and extrusion parameters (T, moisture)
• Temperature and moisture affect both furaneol formation and texture (→ SME)
• Recipe parameters permit furaneol modulation w/o impact on texture
• Lower pH will limit colour and acrylamide development w/o affecting furaneol formation
20
formation
• Addition of amino acid will enhance furaneol while limiting acrylamide formation
Full factorial design