Nanocellulose as an additive in foodstuff

Innventia Report No.: 403

Nanocellulose as an additive in
foodstuff

Göran Ström, Camilla Öhgren and Mikael Ankerfors

Innventia Report No.: 403

June 2013
 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403

Acknowledgements
The authors wish to thank RISE Research Institutes of Sweden for financial support.
 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403
1

Table of contents
Page

1 Summary ................................................................................................................. 2

Sammanfattning.............................................................................................................. 2

2 Introduction .............................................................................................................. 2

3 Materials and methods ............................................................................................ 3

3.1 Preparation of MFC ...................................................................................................... 3
3.2 Characterization of structure of MFC gels using microscopy ...................................... 5
3.3 MFC in food emulsions ................................................................................................ 5
3.4 MFC in food foams ....................................................................................................... 6
3.5 MFC as an additive in bread ........................................................................................ 6
3.6 MFC as an additive in hamburger ................................................................................ 8

4 Results and discussion ............................................................................................ 9

4.1 Structure of MFC gels as characterized by microscopy .............................................. 9
4.2 Impact of MFC on food emulsions ............................................................................. 13
4.3 Impact of MFC on food foams.................................................................................... 16
4.4 Impact of MFC on properties of baked bread ............................................................ 19
4.5 4.5 Impact of MFC water retention during frying of hamburger ................................. 21

5 Conclusions ........................................................................................................... 23

6 References ............................................................................................................ 24

7 Innventia Database information ............................................................................. 25

 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403
2

1 Summary
The impact of microfibrillated (MFC) as an additive in food stuff has been studied in a
cooperation between the Swedish Institute for Food and Biotechnology (SIK) and
Innventia AB. The work included microscopy studies of MFC, the effect of MFC on
stability of oil in water emulsions and foams containing high amounts of dissolved
sugar. Also studied was the impact of MFC as an additive in bread and hamburger.
The work showed that MFC has a strong potential to stabilize oil in water emulsions
and foams. Very stable foams were obtained at low additions of MFC. An addition of
MFC in dough gave the bread better appearance like higher volume and more even
form. The bread also became smoother. As an additive in hamburger MFC gave no off-
flavour and the same texture and mouthfeel as hamburger without MFC. Moreover,
hamburger with MFC could hold more water during frying without negative side
effects.

Sammanfattning
Mikrofibrillerad cellulose (MFC) har utvärderats som tillsats i livsmedel i ett samarbete
mellan Institutet för Livsmedel och Bioteknik (SIK) och Innventia AB. Arbetet
omfattade mikroskopistudier av MFC, inverkan av MFC på stabiliteten av olja i
vattenemulsioner och skum innehållande höga halter av löst socker. Vidare studerades
inverkan av MFC som tillsatsmedel i bröd och hamburgare.
Arbetet visade att MFC har en god förmåga att stabilisera olja i vatten emulsioner och
skum. Mycket stabila skum erhölls vid låga tillsatser av MFC. En tillsats av MFC till
deg resulterade i att brödet fick högre volym, jämnare form och blev slätare. En tillsats
av MFC till hamburgare gav ingen bismak och samma textur och munkänsla erhölls
som för hamburgare utan MFC. Vidare kunde hamburgare med MFC behålla vattnet i
större utsträckning vid stekning än hamburgare utan MFC.

2 Introduction
Microfibrillated cellulose (MFC) also referred to as nanofibrillated cellulose (NFC) or
just nanocellulose is microfibrills released from the cell wall of cellulose, see figure 1.
The material can be obtained after thorough homogenization using a high pressure
homogenizer (Pääkkö et al. 2007; Wågberg et al. 2008; Aulin et al. 2011). The
microfibrills have a diameter in the nano size range and a length in the micro size range.
It is obtained as a highly viscous gel with a solid content of a few percent. This material
was pioneered by Turback and co-workers (Turback et al. 1983) about 30 years ago and
the area was recently reviewed (Klemm et al. 2012).
The process to prepare MFC has been developed during the last decade. One of the
most important discoveries is the fact the various pre-treatments of the pulp may
drastically reduce the energy consumption during the homogenization step (Ankerfors
2012), and also generate a more uniform material. An example of a MFC gel is shown
in Figure 2.
 Nanoce
ellulose as an additive in foodstuff
Innvenntia Report Noo.: 403
3

Figuree 1. Microfibrills in the cellulosee cell wall. (Couurtesy Figuure 2. MFC gel with a solid conntent of 2%. (C
Courtesy
Geoff Daniels, SLU, Sweden).
S Innvventia AB, Sweeden) Pre-treatmment according to
genneration 2 (see below).
MFCC has the pootential of becoming
b ann important componentt in producttion of papeer and
boardd. It also has
h the pottential of bbeing incorpporated in the bio-bassed materiaals of
tomoorrow.
The MMFC produuced by Turrback withouut pre-treatm ment was in
nvestigated as an additive in
food and severaal patents were
w filed at that tim
me. One of those (Turbbak et al. 1984)
descrribes the positive
p imppact of MF FC on the stability of suspensioons, oil-in-water
emullsions, dressing mixturres, meat eemulsions forfo production of sausaages and dessert
toppiings made from whipp ped foams. When MFC C was usedd in hamburrgers they found
f
that a hamburger containin ng MFC loost less watter on fryinng, was juiccier and a better
b
taste. However,, due to th he high ennergy consu umption and cost of pproducing MFC
withoout pre-treaatment, the product
p wass never com
mmercializedd as a food aadditive.
The ssituation todday is comp pletely diffeerent since the
t pre-treatments makke the produ uction
of M
MFC cost eff fficient. Thee work in thhe present report
r is a short surveyy of the potential
of M
MFC in variious food applications
a like emulssions, foamms, bread annd hamburg ger. It
was shown the MFC has interesting
i ppotentials to stabilize food emulssions and fooams.
The product alsso had stron ng potentiaal to increasse the moissture of ham mburger wiithout
deterriorating tasste, texture and
a mouthffeel.

3 Materia
als and methods
m

3.1 Preparattion of MFC

Geneeration 1 microfibrilllated cellullose was produced
p using
u a sooftwood sullphite
dissoolving pulpp (Dissolvinng Plus, D Domsjö Faabriker AB, Sweden).. The cell wall
delammination waas carried out by treatinng the pulp in five step
ps with enzyymatic treattment
beforre homogennization (Pääkkö et al. 22007):
 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403
4

1. First, a 4% cellulose suspension was mechanically refined using a Voith refiner

with 90 kWh/tonne at a specific edge load of 1.5 Ws/m to 50 SR.
2. Next, the enzyme was added. The pulp was dispersed in 2.5 L of a phosphate
buffer (pH 7, final pulp concentration 4% w/w) with 0.17 µL mono-component
endoglucanase per gram fibre (5 ECU/µL) and was incubated at 50 C for 2 h.
The samples were mixed manually every 30 min.
3. Then, the samples were washed with deionized water and the monocomponent
endoglucanase was denaturated at 80 C for 30 min. At the end, the pulp sample
was washed with deionized water.
4. The pre-refined and enzyme-treated pulp was refined once again with the Voith
refiner, this time, to 90 SR (average refining energy 152 kWh/tonne, specific
edge load 1.5 Ws/m).
5. Finally the pulp was passed through 3 large and 5 small slits one time in a high-
pressure homogenizer (Microfludizer M-110EH, Microfluidics, USA).
In Innventia’s terminology, this MFC corresponds to generation 1. The total charge
density of this MFC is normally 0.04-0.05 meq/g as determined by conductometric
titration (Katz et al. 1984).
Generation 2. In order to prepare the anionic MFC, a carboxymethylation pre-treatment
method was used (Walecka 1956a; Walecka 1956b). In this procedure 110 grams of
fibres (Dissolving Plus, Domsjö Fabriker AB, Sweden) were pre-treated per batch. The
procedure can be divided into five steps (Wågberg 2008):
1. Firstly, the never dried fibres were dispersed in deionised water at 10000
revolutions using an ordinary laboratory pulper. This was conducted in smaller
batches of 30 g of fibres in two litres of deionised water.
2. The fibres were then liquid exchanged to ethanol by washing the fibres in
one litre of ethanol four times with a filtering step in between.
3. The fibres were then impregnated for 30 minutes with a solution of
monochloroacetic acid amounts in 500 mL of iso-propanol. The fibres were then
added in portions to a solution of 16.2 g of NaOH in 500 mL methanol and
mixed with two litres of iso-propanol that had been heated just below its boiling
temperature in a five litre reaction vessel fitted with a condenser. The
carboxymethylation reaction was allowed to continue for one hour.
4. Following the carboxymethylation step, the fibres were filtrated and washed in
three steps. Firstly, the fibres were washed with 20 L of deionised water.
Secondly, they were washed with two litres of acetic acid (0.1 M) and finally
with ten litres of deionised water. The fibres were then impregnated with a two
litre NaHCO3-solution (4% w/w solution) for 60 minutes in order to convert the
carboxyl groups to their sodium form. Then the fibres were washed with 15
litres of deionised water and drained on a Büchner funnel.
5. Finally, the pulp was passed 3 large and 5 small slits one time in a high-pressure
homogenizer (Microfludizer M-110EH, Microfluidics, USA) in the same fashion
as was described by (Pääkkö et al. 2007) for enzymatically treated pulps.
 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403
5

In Innventia’s terminology, this MFC corresponds to generation 2. The total charge

density of this MFC is normally 0.5-0.6 meq/g which corresponds to a degree of
substation of roughly 0.1%.
Seven different samples of MFC were used in the studies, see Table 1.
Table 1. MFC samples used in the present work. DS are given for MFC generation 2 and it refers to the degree of
substitution of the pulp after the carboxymethylation.
Sample MFC type Dry substance Comment Used in
(%)
1 Gen 2 (lab) 1.8 CLSM, TEM
2 Gen 1 (lab) 1.0 CLSM
3 Gen 1 2.4 CLSM
(pilot)
4 Gen 1 1.0 Sample 3 pressed to CLSM, TEM
(pilot) 30% dryness and then
dispersed to 1%
5 Gen 2 (lab) 2.25 Hamburger,
bread
6 Gen 1 (lab) 2.1 Food emulsions
7 Gen 2 (lab) 1.95 Foams

3.2 Characterization of structure of MFC gels using microscopy

CLSM. The samples 1-4 were analysed under the confocal laser scanning microscope,
CLSM, Leica TCS SP5 II (Heidelberg, Germany). The staining, akriflavine or
acridineorange dissolved in water was dried on a cover glass. The samples were places
in the cavity of the objects slide, and the stained cover glass was sealed on top. The light
source was an Argon laser with an emission maximum at 488nm and the signal emitted
in the wave length interval 500-580 nm was recorded. A 10x air objective and a 63x
water objective was used. Computer zooming was done at 1x, 5x and 10x and images
were recorded with formats of 1024x1024 pixels.
TEM. A small amount of sample 1 and 4 was placed in a gold cup and cryofixed in
liquid propane. The frozen samples were fractured in vacuum and etched before coating.
After sublimation of the water, replicas were formed by rotary shadowing of
platinum/carbon on top. The replicas were cleaned in solvent before examination.
Micrographs were taken in a TEM, LEO 906e, (LEO Electron Microscopy Ltd,
Germany) at an accelerating voltage of 80 kV.

3.3 MFC in food emulsions

Emulsion preparation. Sample 6 of MFC was mixed with a stick mixer, Ultra Turrax
T25 (JANKE & KUNKEL IKA® -labortechnik, Germany) together with rapeseed oil
(ICA) and water in four different combinations with a an MFC content between 0.5-1%
and 10-50% of rapeseed oil for two minutes. The emulsions formed were stored in
refrigerator for three days where after the stability was documented before and after
centrifugation (3000 rpm for 30 min).
Emulsions containing 0.5-1% MFC (sample 6) and 10-20% rapeseed oil was made in a
high pressure homogenizer Panda 2k (Niro Soavi, Italy) at two different pressures, 100
 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403
6

and 1000 bar. Thereafter the stability was checked after storage in refrigerator in three
days before and after centrifugation.
CLSM. The high pressure homogenized emulsions were studied in confocal laser
scanning microscope, CLSM, Leica TCS SP5 II (Germany). Akriflavine and Nile red,
staining the MFC and the fat droplets respectively, was dissolved in water, mixed and
dried on a cover glass. The emulsions were places in the cavity of the objects slide, and
the stained cover glass was sealed on top. The light source was an Argon laser and a
HeNe-laser with an emission maximum at 488 nm and 594 nm, respectively. The
signals were emitted in the wave length interval of 500-570 nm and 605-670 nm. A 20x
water objective was used. Computer zooming was done at 1x, 2x and 4x and images
were recorded with formats of 1024x1024 pixels.

3.4 MFC in food foams

Foams were prepared from solutions containing a high amount of sugar, MFC
generation 2 (sample 7) emulsifier (Colco 2282-00 alpha-gel, delivered by Aromatic
AB) and water. The composition of the liquids used is given in Table 2. The foams were
generated from a 200 g solution using an OBH kitchen hand mixer at highest speed for
90 seconds.
Table 2. Composition of the liquids used for foam generation.
Sample Sugar Emulsifier MFC Water Volume of 200 g
(wt.-%) (wt.-%) (wt.-%) (wt.-%) solution (mL)
A 48 1.0 0 51.0 168
B 48 1.0 0.2 50.8 168
C 42 1.0 0 57,0 185
D 42 1.0 0.4 56.6 185
E 48 0.5 0 51.5 168
F 48 0.5 0.1 51.4 168
G 48 0.5 0.2 51.3 168

The foams were poured into measuring cylinders. The initial foam volume was
determined and is given in Table 2. The drainage of liquid out from the foam column
was read from the bottom of the measuring cylinder at different time intervals.
The density gradient of the foam column formed from sample D was measured 52 hours
after foam generation by sucking out 50 mL foam and take the weight of the rest of the
foam and the cylinder, and subtract it from the weight before the foam was sucked out
from the cylinder.

3.5 MFC as an additive in bread

Bread was produced with and without MFC (sample 5) and with and without different
enzymes. The recipe is shown in Table 3.
 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403
7

Table 3. Recipe of bread.

Ingredients Relative amount (%) Amount (g)
Flour 100 182,73 g
MFC (sample 5) 55.6 (1.35) 101.52 (2.54 )
Extra water 7.11 13
Yeast 4.16 7.62
Sugar 2.78 5.07
Salt 1.67 3.03
Emulsifier 0.5 0.9
Enzymes 0-100 ppm
 
Mixing of the ingredients was performed in a Kitchen Aid (St. Joseph, Michigan, USA)
model KSM90. The mixing was carried out as 2:30 min at medium speed and 2:30 min
at the highest speed. Thereafter, the dough was removed, briefly kneaded on a lightly
floured table and shaped into a round loaf. Covered with kitchen cloth, the dough was
left to rest 10 min on the table. After resting, the dough was divided into pieces of 45 g.
Pieces were then kneaded, hand shaped into round balls and let to rest 5 min. Using a
special board, pieces of dough were then moulded until forming a perfectly even round
ball. After that, pieces of dough were placed on grills covered by baking paper and
fairly shared into two batches: one for freezing and the second one immediately baked
and analyzed.
Pieces of dough for freezing were transferred in a freezing room to be frozen at -20 °C.
Two hours later, frozen dough pieces were transferred in plastic bags and stored in the
same freezing room at -20 °C. After 7 days freezing storage pieces of dough were
withdrawn from the freezing room, put on trays with baking paper and covered with a
plastic sheet. When placed on tray, dough pieces were thawed 2 hours in a thawing
chamber calibrated at 25 °C and 65% relative humidity (RH). Thereafter the plastic
cover was removed and the samples were transferred in a proofer from the brand Sveba
Dahlen, model Fermatic. The proofing lasted 36 minutes at 38 °C and 80% RH.
Then the trays were placed for 10 minutes in an oven from Sveba Dahlen, model S8,
previously set at 210 °C and 20% RH. Afterwards trays were put on a table and let
cooling one hour at room temperature. Then the buns were transferred into plastic bags.
The experimental design is shown in Figure 3. Totally 12 different types of buns were
analyzed; with or without MFC, no enzyme or α-amylase or hemicellulase and freeze
stored or not (indicated by the ´prefix).
 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403
8

Emulsifier MFC Enzyme Sample

2.0.0
No 2.0.0´
2.0.1
No (0)  ‐amylase 2.0.1´
Hemicellulas 2.0.2
2.0.2´
E472 (2)
2.1.0
No 2.1.0´
Yes (1) ‐amylase 2.1.1
2.1.1´
Hemicellulas 2.1.2
2.1.2´

Figure 3. Experimental design for evaluating MFC as additive in bread.

A macroscopic investigation was performed on the buns by taking photos of the cross
section and the surface of the buns using a Nikon Digital Camera D90 (Nikon
Corporation, Japan) equipped with a lens Micro-Nikkor 55 mm 1:2.8. The aim was to
observe crust aspect, bread shape and crumb aspects. The slices cut for texture analyzes
were also photographed with the same camera, to observe the crumb aspect and control
its homogeneity.
The texture was anlaysed by measuring the compressive stress using a modified version
of the AACC method 74-09 (AACC, 2001) using an Instron Universal Testing Machine
5542 (Instron, USA) two hours after baking. For each sample three breads were used for
texture measurements. Two vertical slices of 2 cm thickness were cut out from each
bread. A cylindrical metal probe of 20 mm diameter was pushed into the crumb in the
middle of the slices with a constant speed of 1.7 mm/s. The compressive stress at 25%
compression was used as a measure of the bread firmness. The measurements were
performed 2 h after baking.

3.6 MFC as an additive in hamburger

Hamburgers were performed with and without different additives; water, MFC (sample
5) and potato starch. All hamburgers contained 100 g of minced meat, and water, MFC
and potato starch were added in different amounts. MFC was added to hamburgers in
form of a 2.25% gel. The additives were gently kneaded in the minced meat. The
hamburgers were formed to a circle with a diameter of 10 cm before frying. The
hamburgers were weighted before and after frying and the juiciness and taste was
judged after cooling.
 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403
9

4 Results and discussion

4.1 Structure of MFC gels as characterized by microscopy

CLSM. The microstructure of the four gel types of MFC was studied in CLSM using
florescent staining. MFC was shown to be distributed differently in the different
samples dependent on generations and manufacturing, see Figure 4. Sample 1,
generation 2 made in lab scale is shown in Figure 4a-c; sample 2, generation 1 made in
lab scale in Figure 4d-f; sample 3, generation 1 made in pilot scale in Figure 4g-I and
sample 4, generation1 made in pilot scale (i.e. sample 3) but pressed and re-dispersed in
Figure 4j-l.
Sample 1, which is a generation 2 sample made in lab scale, contain some very large
fibres with a thickness of about 50 m and more than 100 m in length, in contrast to
the samples containing generation 1 that contain no such large fibre pieces. However,
there are few thin fibres with a thickness of about 0.5 m and length around 100 m in
sample 1, which there is a lot of in sample 2-4, compare Figure 4c, f, i and j.
In sample 2 there is few large fibres compared to the other samples containing
generation 1, sample 3 and 4, see Figure 4d, g and j. Sample 3 and 4 show very similar
microstructure at both low and high magnifications in CLSM. However, sample 4
which is the pressed and re-dispersed pilot made generation 1, appears to be somewhat
more inhomogeneous at low compared to sample 3, which is the same material but
before pressing and re-dispersion.
 Nanoce
ellulose as an additive in foodstuff
Innvenntia Report Noo.: 403
10

Figuree 4. CLSM imagges of MFC sam mples. Fig. a, b and c are imagges of sample 1 (gen. 2) Fig. dd, e and f are im
mages
of sammple 2 (lab made gen. 1). Fig. g, mages of sample 3 (pilot made
g h and i are im e gen. 1). Fig. j,, k and l are imaages
made from sample 4 (sample 3, presssed and re-disspersed). The MFC n bright and thee scale bar is 1000 m
M is shown in
(a, d, g and j), 25 m
m (b, e and h and k) and 5 m ((c, f, I and l).
 Nanoce
ellulose as an additive in foodstuff
Innvenntia Report Noo.: 403
11

TEM M. In order tot get informmation on a higher mag gnification and the fibrrillar structuure of
the MMFC:s two of the sam mples (samplle 1 and 4) were freezze-etched annd studied under u
the TTEM. Figuree 3 shows the CLSM im mages (Figure 5a and d) of the tw wo samples at the
higheest magnificcation possible using C CLSM and the images of the sam me microstru ucture
showwn by usingg TEM on frreeze-etchedd samples at a a much higher magnnification (F Figure
5b-c,, e-f). The microstructu
m ure that wass not possibble to resolv
ve in CLSM M is now posssible
to reesolve in TE EM. Sample 1, containning MFC generation 2, is built up by a fib brillar
struccture with a thickness ofo the fibrilss of 20-30nm
m, which co orrespond w well to whatt have
beenn seen previiously (Wåg gberg 20088; Siro 2010 0) see Figurre 3b-c. Thhe black spo ots in
Figurre 3c are prrobably the cross sectioons of the fibrils.
f The fibrillar stru
ructure in saample
4, coontaining a generation
g 1 MFC, is nnot that easyy to analyzee in TEM aas sample 1 since
o thin fibrees (around 00.5 m in th
it conntain a lot of hickness wh hich was ressolved in CL LSM)
whicch will be veery large in TEM. Theyy are appeaaring in the TEM T imagees as large layers
l
abovve each otheer, see Figure 5e. Betw ween these layers
l the fiine fibrillar structure can be
seen,, see Figurre 3f. Thee fibrillar sstructure iss more fine-meched iin generatiion 1
comppared to genneration 2.

Figuree 5. CLSM and TEM

T images off MFC samples 1(a-c) and 4(d-f). The MFC iss shown in brighht in the CLSM
images and in grey inn the TEM imagges. The scale bbar is 5µm (a, d),
d 200 nm (b, e,)
e and 50 nm (c
(c, f).
 Nanoce
ellulose as an additive in foodstuff
Innvenntia Report Noo.: 403
12

The ffew thin fibbres which were

w resolvved using CL LSM in the size of 0.5 m thickneess in
sampple 1, was also
a found in
i TEM. Fiigure 6a sho ows an imaage of the fi
fibre and fib
brillar
microostructure in TEM on n an overaall level off structure and a CLSSM image on a
correesponding magnificatio
m on of samplle 1. To thee right in th
he TEM-imaage in Figu ure 6a
anothher type of structure iss revealed w
which probably is und delaminated pulp fibre.. This
piecee of fibre is built upp by denseely packed fibrils sho ow in Figur
ure 6b at higher
h
magnnification. As
A a comparison, the fibrillar strructure wheen not packked in a fib bre is
show
wn in Figuree 6c.

Figuree 6. TEM imagees of MFC sampple 1. Overview

w (a), focusing in a fibre (b) and in the fibrillarr part. The scalee bar is
5 µm ((a) and 100 nm
m (b, c,).
 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403
13

Characterization of MFC using microscopy techniques shows that merging of fibrils

into large fibres with different thickness varies between the different generations and to
some extent between lab-scale and pilot-scale manufactured MFC. Generation 2 contain
the largest fibres (50 m in thickness and >100 m in length). Generation 1 contain a
lot of thin fibres (0.5 m in thickness and 100 m in length) which not is found to a
large extent in generation 2. On a high magnification in TEM both types are
characterized by thin fibrils with a thickness of 20-30 m. The fibres are built up of
fibrils densely packed targeted in the same direction.

4.2 Impact of MFC on food emulsions

A first test of the potential for MFC (sample 6) as an oil-in-water emulsion stabilizer
was made with a simple stick mixer with mixtures of 0.5-1% MFC, 10-50% rape seed
oil and water. The results from the stability studies of the emulsions are shown in Table
3. The emulsions containing the highest amount of MFC and oil was the most stable
ones after three days storage. After centrifugation it was possible to separate both an oil
and a water phase from the emulsion, except in emulsion 2 with 50% oil that was very
stable.
Homogenizing in Panda was thereafter performed at two different pressures for
emulsion 1, 3 and 4. (Emulsion 2 was already stable by simple mixing with Ultra
Turrax). This homogenizing showed that it was possible to make emulsion in all the
combinations tested. Figure 7 shows the phase separation of the emulsions after three
days of storage and following centrifugation. Only some of the water phase was
possible to press out of the emulsion after centrifugation. The higher the amount of
MFC and oil the less water is possible to press out of the emulsion.
Table 4. MFC as stabilizing agent in emulsions.
Emulsion MFC Oil Mixing in Centrifugation Pressure in Homogenizing Centrifugation
sample 6 (%) Ultra Turrax after Ultra homogenizer, and 3 days after
(%) and 3 days Turrax and 3 (bar) storage homogenizing and
storage days storage 3 days storage

1a 1 20 stable separates in 3 100 stable separates in 2

phases phases
1b 1 20 stable separates in 3 1000 stable separates in 2
phases phases
2 1 50 stable separates in 2
phases
3a 0.5 20 separates separates in 3 100 stable separates in 2
phases phases
3b 0.5 20 separates separates in 3 1000 stable separates in 2
phases phases
4a 0.5 10 separates separates in 3 100 stable separates in 2
phases phases
4b 0.5 10 separates separates in 3 1000 stable separates in 2
phases phases
 Nanoce
ellulose as an additive in foodstuff
Innvenntia Report Noo.: 403
14

Figuree 7. Emulsion coontaining MFC, rape seed oil aand water after storage and ce
entrifugation.
The eemulsifiers containing 1% MFC w were firmer,, with a connsistency ressembling off sour
creamm, than the emulsions containing 0.5% MFC that instead d resembledd of soured milk.
A haarder homoggenizing at a higher ppressure gav ve also a soomewhat firrmer consisstency
comppared to thee lower pressure. Imagges on the emulsions
e are shown inn Figure 8 and 9
togetther with thhe CLSM-im mages on thhe microstru ucture of th
he emulsionns. In the CLSM
C
imagges the oil droplets
d are shown in ggreen colour and the MFC M in red. The oil dro oplets
are iincorporatedd in a netwwork of MF FC fibres making
m a steric stabilizzation of th
he oil
dropllets. As seeen in the im
mages the oiil droplet siize and distrributions vaary dependeent of
oil ccontent andd pressure. The higheer pressuree the smalller dropletss and the more
homoogeneous siize distributtion of dropplets. At low w pressure there
t are booth large dro
oplets
and very small ones wherreas at highh pressure the larger ones are noon-existing. The
dropllet size is also
a somewh hat smaller at 10% oill content compared to 220% oil content,
but thhe MFC and the oil dro oplets are ddistributed in
n aggregatees at the low
wer concentrration
of oiil. The MFC C is more in ncorporatedd between thet oil droplets/well diistributed arround
the ooil droplets when
w the hiigher pressuure was used d, see Figurre 9.
 Nanoce
ellulose as an additive in foodstuff
Innvenntia Report Noo.: 403
15

Figuree 8. CLSM imagges of emulsions containing M FC (sample 6). The MFC is sh

hown in red andd the oil in green.
The sccale bar is 50 m.


Figuree 9. CLSM imagges of emulsions containing M FC (sample 6). The MFC is sh

hown in red andd the oil in green.
The sccale bar is 10 m.

 Nanoce
ellulose as an additive in foodstuff
Innvenntia Report Noo.: 403
16

To suummaries; MFC has a very goodd potential as a stabilizin

ng agent forr emulsions. It is
possiible to makke very firrm emulsioons by incrreasing the content off MFC, oil and
increeasing the pressure at homogenizin
h ng. The firm
m emulsionss is characteerized by sm
maller
oil ddroplets withh a smaller size distribbution (incrreased homoogenizing ppressure), a more
well--distributed MFC arou und the sepparate oil droplets
d (higher MFC content), anda a
moree homogeneeous distrib bution of thhe oil drop plets and th
he MFC inn the bulk phase
p
(incrreased oil coontent) com
mpared to em mulsions witth a more liquid texturee.

4.3 Impact of o MFC on food foam ms

Foamms generateed from liquuid fluids coontaining suugar and em
mulsifier diissolved in water
with a minor adddition of MFC
M dispersiion were geenerated witth a kitchenn hand mixeer and
poureed into meaasuring cyliinders. The experimenttal set up iss shown in FFigure 10 which
w
also shows the drained
d liqu
uid 23 and 447 hours afteer foam gen
neration.

Figuree 10. Experimenntal set up for studies of foam stability. The situations 23 and
d 47 hours afterr foam generatiion are
shownn. The liquid column beneath thhe foam is obseerved and measured.
The ffoams weree characteriized with reespect to:
 Initial fooam volumee.
 Time to drained liqu
uid became visible (T1).
)
 Time to 30 vol.-% ofo the liquidd had becomme visible beeneath the ffoam colum
mn
(T2). Thuus, roughly 50 mL of thhe liquid haad drained out
o from thee foam.
No fo
foam could beb generateed without eemulsifier. The
T initial foam
f volum
me increasedd with
increeasing amouunt of emu ulsifier but decreased with increaasing conteent of sugar and
MFC C (see Tablee 5).
 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403
17

Table 5. Composition and foam characteristics of the investigated samples.

Sample Emulsifier MFC Sugar Initial foam T1 T2
(wt.-%) (wt.-%) (wt.-%) volume (mL) (h) (h)
A 1 0 48 540 ~3 ~5
B 1 0.2 48 450 24 58
C 1 0 42 620 2 3,5
D 1 0.4 42 370 >52 >52
E 0.5 0 48 440 0,1 0,3
F 0.5 0.1 48 380 0,2 1,2
G 0.5 0.2 48 350 0,3 2,0

160 E F G 160 E F G

Volume of drained liquid, ml
140
Volume of drained liquid, ml

140
120 120
100 100
80 80
60 60
40 40
20 20
0 0
0 10 20 30 40 50 60 0 1 2 3 4 5 6 7
Time, h Time, h

Figure 11. Foam stability given as volume of the liquid column beneath the foam column as a function of time after foam
generation for samples E, F and G. The right diagram is an enlargement of the time interval up to 7 hours.
Foams generated from sample E, F and G had low stability as can be seen in Figure 11
These samples contained only 0.5% emulsifier and 0, 0.1 and 0.2% MFC, respectively.
Even though the foam stability was low, MFC had a significant positive impact on the
foam stability
When the content of emulsifier was increased to 1% the initial foam volume increased
significantly and very high foam stability was obtained in the presence of MFC. The
stability parameters T1 and T2 increased roughly 10 times when 0.2% MFC was present
(B) compared to the reference without MFC (A). When the content of MFC was raised
to 0.4%, the foam showed no tendency to drain liquid within the 52 hours the
experiment lasted. Instead of extending the time for the foam stability characterization,
this foam was subjected to a determination of its density gradient (see Figure 13). The
foam column was divided into 7 volume units, each with a volume of 50 mL. The
bottom volume unit had a slightly higher density than the other units, which were very
constant in density.
 Nanoce
ellulose as an additive in foodstuff
Innvenntia Report Noo.: 403
18

160 A B C D
Volume of drained liquid, ml

140
120
100
80
60
40
20
0
0 0
20 40 60 80 100
Time, h

Figuree 12. Foam stabbility given as voolume of the liqquid column benneath Figurre 13. Density ggradient of the foam
f
the foaam as a function of time after foam
f generationn for samples A,
A B, C colum
mn of sample C C, 52 hours afteer foam
and D. geneeration.
MFCC also appeaared to redu uce foam brrakeage at the
t top of the foam. TThis phenom menon
was observed affter two day ys for foamms without MFC.
M No sign
s of this phenomenaa was
obserrved withinn 3 days for the system containing 0.2% MFC
C as can be sseen in Figu
ure 14
and 115

Figuree 14. Observatioons of foam Figure 115. Foam collappse did not occur with sample B which contained
collapsse on top of thee foam column of
o 0.2% M FC.
samplee C, 23, 27 andd 71 hours afterr
foam ggeneration.
 Nanoce
ellulose as an additive in foodstuff
Innvenntia Report Noo.: 403
19

4.4 Impact of o MFC on propertie es of baked d bread

The effect of adddition of MFC
M (samplle 5) in com
mbinations with
w two diifferent enzzymes
on w
wheat bread was studied d on bread bbaked from fresh dough
h and doughh that was freeze
f
storeed for 7 days.
The quality of the
t bread was
w evaluateed by measu uring the texxture by coompression stress
in the Instron teesting machhine and by taking photos of the appearance
a of the bunss. The
comppression strress of breadd baked froom fresh do ough contain ning MFC w was significcantly
softeer than breaad without MFC, see figure 8 an nd comparee the blues stacks sho owing
breadd with no additives,
a “n
no add” witth bread con ntaining MF FC, “MFC””. The effecct was
evenn more proonounced in n presence of the en nzymes -aamylase andd hemicellu ulase.
(Com mpare the redr and grreen stacks showing breads b con
ntaining enzzymes with h and
withoout MFC inn Figure 16..) However,, the effect of softening g the bread was absentt after
freezze storage of the doughh, see the staacks with prrefix ´in Fig
gure 16.

Figuree 16. Compresssion stress at 255% stress for brreads containing different addiitives; no additivves, MFC, enzyymes
(αa=α--amylase, he=hhemicellulase) Prefix
P ´means ffreeze stored saamples.
In Fiigures 17 and
a 18 the surface andd cross secctions of thee buns are shown for buns
baked from fressh dough (FFigure 17) aand buns baaked from freeze
f storeed dough (F
Figure
18). T
The buns coontaining MFC
M resulte d generally in bread with a more ssmooth crusst and
evenn form and also
a a higherr bun, whicch is shown in the imag
ges in Figurees 17 and 18.
To suummarize; MFC has a clear favouurable effecct on bread baked from m fresh douggh by
makiing the bunns softer and giving thhem a betterr appearancce in form oof smooth crust,
high volume andd even form m. Howeverr, the effect of MFC on n freeze stoored dough needs
n
to bee further evaaluated.
 Nanoce
ellulose as an additive in foodstuff
Innvenntia Report Noo.: 403
20

Figuree 17. Buns bakeed from fresh doough containingg different addittives; no additivves, MFC, enzyymes (αa=α-am
mylase,
he=heemicellulase) Prrefix ´means freeeze stored sam
mples.
 Nanoce
ellulose as an additive in foodstuff
Innvenntia Report Noo.: 403
21

Figuree 18. Buns bakeed from freeze stored

s dough coontaining differeent additives; no additives, MFFC, enzymes
(αa=α--amylase, he=hhemicellulase) Prefix
P ´means ffreeze stored saamples.

4.5 4.5 Im
mpact of MFC
M water retention during fry ying of haamburger
MFC C as a mooisture reten ntion agentt in hambu urgers wass studied bby adding MFC
(sam
mple 5) to minced
m meaat before fryying. Sincee MFC is in n form of a gel with a dry
substtance of 2.225% the haamburgers w will be dilu
uted with a lot of wateer when MF FC is
added. A hambuurger contaaining 100gg meat was diluted witth 0, 10, 166 and 27% water
respeectively whhen 0, 0.24, 0.47 and 00.63% MFC C was added d to the meeat. The ressult in
weigght loss is seen in Figure 19. IIrrespectivee of additio onal amounnt of MFC C and
attenndant water the hambu urgers lost 330±1% wateer during frrying, but ssince more water
was present in the hambu urgers contaaining MFC C the end weight
w becoome higher after
MFC C addition. The end weight
w was increased byb 10, 23 anda 41% affter the diffferent
addittions of waater, which is a substaantial increease. The highest
h wateer addition gave
indeeed a more juuice but also
o a more waatery taste and
a lost tastte of hambuurger. The lo owest
addittion of watter (10%) gave a moore juice hamburger
h compared
c tto the refeerence
withoout any adddition of water or MFC C, but otherwwise it was similar in taaste comparred to
the hhamburger containing 100% meaat. Photos of o these tw wo types off hamburgerrs are
showwn in Figuree 20. The haamburger coontaining ad ddition of MFC
M and waater is someewhat
largeer than the one
o without.
 Nanoce
ellulose as an additive in foodstuff
Innvenntia Report Noo.: 403
22

  160

140
startv ikt
Start  weight
w
End w
weight
slutvik
kt
120 ‐29%
‐29
9%
weight (g)

100
‐31%
‐
‐30%
80

60

40

20

0
ref2
100% meat MFFC 0,3%
MFC  0.24% MFC 0,5%
MFC 0.47% MFC 0
0.63%
MFC 0
0,9%

10% water 16% water 27% w

water

Figuree 19. Weight of hamburgers beefore and after ffrying containing different amo
ount of MFC andd water.

Figuree 20. A hamburgger containing 100%

1 a a hamburgeer containing 0.24%
meat beffore and after frrying (a and c) and
MFC aand 10% water before and afteer frying (b and d).
Potatto starch was
w used as a referencee to the MF FC. High am mount of ppotato starch
h was
requiired to loweer the waterr lost to thee same levell as for MFC. Additionn of 1.4% potato
p
starch was requuired to receeive the sam me water lo
ost as for 0.63%
0 MFC C, see Figurre 21.
This high amouunt of potato starch ggave the haamburger a bad consisstency and taste.
Loweer addition of potato sttarch (0.75% %) gave no o texture off the hambu
o bad taste or urger,
but nnot either anny effect on the water hholding capaacity.
 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403
23

To summarize; MFC (i) gave no off-flavor, (ii) gave the same texture and mouthfeel as
the hamburger without additions, (iii) hold more water without side effects as watery
taste, (iv) is easy to mix with the meat, but (v) due to its low dry weight it is difficult to
add the right amount of MFC without adding to much water.

  40

35
35% 35%

30 29% 27%
Loss of water (%)

25

20

15

10

0
ref 4 0.63% MFC
MFC 0,9% 0.75% pot. St.
Potatis stärkelse 1% 1.4% pot. St.
Potatis stärkelse 2%
25% water 27% water 24% water 26% water

Figure 21. The loss of water during frying of hamburgers containing different amounts of additives (water, MFC and
potato starch).

5 Conclusions
Characterization of MFC using microscopy techniques shows that merging of fibrils
into large fibres with different thickness varies between the different generations and to
some extent between lab-scale and pilot-scale manufactured MFC. Generation 2 contain
the largest fibres (50 m in thickness and >100m in length). Generation 1 contain a lot
of thin fibres (0.5 m in thickness and 100m in length) which not is found to a large
extent in generation 2. On a high magnification in TEM both types are characterized by
thin fibrils with a thickness of 20-30 m. The fibrillar structure is more fine-meched in
generation1 compared to generation 2. The fibres are built up of fibrils densely packed
targeted in the same direction.
MFC has a very good potential as stabilizing agent for emulsions. It is possible to make
very firm emulsions by increasing the content of MFC, oil and increasing the pressure at
homogenizing. The firm emulsions is characterized by smaller oil droplets with a
smaller size distribution (increased homogenizing pressure), a more well-distributed
MFC around the separate oil droplets (higher MFC content), and a more homogeneous
distribution of the oil droplets and the MFC in the bulk phase (increased oil content)
compared to emulsions with a more liquid texture.
MFC also has a very good potential as stabilizing agent for food foams already at low
additions (0.2%). Higher additions gave extremely stable foams.
 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403
24

MFC has a clear favorable effect on bread baked from fresh dough by making the buns
softer and giving them a better appearance in form of smooth crust, high volume and
even form. However, the effect of MFC on freeze stored dough needs to be further
evaluated.
As an additive in hamburgers MFC gave no off-flavor, gave the same texture and
mouthfeel as the hamburger without additions, hold more water without side effects as
watery taste, is easy to mix with the meat, but due to its low dry weight it is difficult to
add the right amount of MFC without adding to much water.

6 References
Ankerfors, M. (1012)
Microfibrillated cellulose: Energy-efficient preparation techniques and key properties.
Licentiate thesis, KTH Royal Institute of Technology, School of Chemical Science and
Engineering, Department of Fibre and Polymer Technology, Division of Fibre
Technology, TRITA-CHE-Report 2012:38, ISSN 1654-1081, ISBN 978-91-7501-464-
7.
Aulin, C., Johansson, E., Wågberg, L. and Lindström, T. (2010)
Adsorption behavior and structural properties of microfibrillated cellulose-based
multilayers.
Biomacromolecules 11: 872-882.
Pääkkö, M., Ankerfors, M., Kosonen, H., Nykänen, A., Ahola, S., Östberg, M.,
Ruokolainen, J., Laine, J., Larsson, P. T., Ikkala, O. and Lindström, T. (2007)
Enzymatic hydrolosis combined with mechanical shearing and high-pressure
homogenization for nanoscale cellulose fibrils and strong gels
Biomacromolecules 8: 1934-1941.
Siró, I. and Plackett, D. (2010)
Microfibrillated cellulose and new nanocellulose materials: a review.
Cellulose 17:459-494.
Turbak, A. F., Snyder, F. W. and Sandberg K. R. (1984)
Suspensions containing microfibrillated cellulose
US Patent 4,487,634 Dec 11, 1984
Wågberg, L., Decher, G., Norgren, M., Lindfors, T., Ankerfors, M. and
Axnäs, K. (2008)
The build-up of polyelectrolyte multilayers of microfibrillated cellulose and cationic
polyelectrolytes.
Langmuir 24:784-795
 Nanocellulose as an additive in foodstuff
Innventia Report No.: 403
25

7 Innventia Database information

Title
Nanocellulose as an additive in foodstuff

Author
Göran Ström, Camilla Öhgren and Mikael Ankerfors

Abstract
The impact of microfibrillated (MFC) as an additive in food stuff has been studied in a
cooperation between the Swedish Institute for Food and Biotechnology (SIK) and
Innventia AB. The work included microscopy studies of MFC, the effect of MFC on
stability of oil in water emulsions and foams containing high amounts of dissolved
sugar. Also studied was the impact of MFC as an additive in bread and hamburger.
The work showed that MFC has a strong potential to stabilize oil in water emulsions
and foams. Very stable foams were obtained at low additions of MFC. An addition of
MFC in dough gave the bread better appearance like higher volume and more even
form. The bread also became smoother. As an additive in hamburger MFC gave no off-
flavour and the same texture and mouthfeel as hamburger without MFC. Moreover,
hamburger with MFC could hold more water during frying without negative side
effects.

Keywords
Microfibrillated cellulose, food, foam, emulsion, microscopy

Classification
1320, 1153

Type of publication
Innventia report

Report number
403

Publication year
April 2013

Language
English
INNVENTIA AB is a world leader in research and development relating to pulp,
paper, graphic media, packaging and biorefining. Our unique ability to translate
research into innovative products and processes generates enhanced value for our
industry partners. We call our approach boosting business with science. Innventia is
based in Stockholm, Bäckhammar and in Norway and the U.K. through our
subsidiaries PFI and Edge respectively.

INNVENTIA AB Tel +46 8 676 7000

Drottning Kristinas väg 61, BOX 5604 Fax +46 8 411 55 18 info@innventia.com
SE-114 86 Stockholm Sweden VATno 556603110901 www.innventia.com