Colloids and Surfaces B: Biointerfaces 75 (2010) 377384

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces

journal homepage: www.elsevier.com/locate/colsurfb

How surface composition of high milk proteins powders is inuenced by

spray-drying temperature
C. Gaiani a, , M. Morand a , C. Sanchez a , E. Arab Tehrany a , M. Jacquot a , P. Schuck b , R. Jeantet b , J. Scher a
a
b

LIBio, Nancy Universit, Laboratoire dIngnierie des Biomolcules, 2 avenue de la Fort de Haye, B.P. 172, 54505 Vandoeuvre Les Nancy Cedex, France
INRA, Agrocampus Ouest, UMR 1253, Science et technologie du lait et de luf, F 35042 Rennes, France

a r t i c l e

i n f o

Article history:
Received 22 May 2009
Received in revised form
15 September 2009
Accepted 15 September 2009
Available online 20 September 2009
Keywords:
Proteins
Dairy powders
Drying
Surface composition
XPS

a b s t r a c t
High milk proteins powders are common ingredients in many food products. The surface composition of
these powders is expected to play an essential role during their storage, handling and/or nal application.
Therefore, an eventual control of the surface composition by modifying the spray-drying temperature
could be very useful in the improvement of powder quality and the development of new applications.
For this purpose, the inuence of ve spray-drying temperatures upon the surface composition of the
powders was investigated by X-ray photoelectron spectroscopy. The major milk proteins were studied:
native micellar casein and native whey, both more or less enriched in lactose.
The results show a surface enrichment in lipids for all the powders and in proteins for many powders. Whatever the drying temperature, lipids and proteins are preferentially located near the surface
whereas lactose is found in the core. This surface enrichment is also highly affected by the spray-drying
temperature. More lipids, more proteins and less lactose are systematically observed at the surface of
powders spray-dried at lower outlet air temperatures. The nature of proteins is also found essential;
surface enrichment in lipids being much stronger for whey proteins containing powders than for casein
containing powders. Additionally, we found a direct correlation between the lipids surface concentration
and the wetting ability for the 25 powders studied.
2009 Elsevier B.V. All rights reserved.

1. Introduction
Spray-drying is a commonly used technique in the food industry
and more precisely in the dairy sector. By decreasing water content, this process provides several advantages as a minimization
of microbial deterioration of the spray-dried products, a reduction
of lipids oxidation [1] and, ideally, a preservation of the original
emulsion structure [2]. It ensures also a reduction of storage and
transport cost and an easier handling of the material [3]. Spraydrying of milk concentrates involves a quick removal of water,
leading to the formation of a dry matrix containing proteins, lactose,
lipids and minerals. New processing technologies have emerged
over the past 10 years allowing the development of novel ingredients with unique functional properties. Among these, high milk
proteins powders are an example [4]. Native whey isolate (NWI)
extracted directly from milk (i.e. not whey), presents a concentration in amino acids exceeding that of conventional whey proteins.

Abbreviations: Lac, lactose; NMC, native micellar casein; NWI, native whey isolate; PCA, principal component analysis; Tout , outlet air temperature; XPS, X-ray
photoelectron spectroscopy.
Corresponding author. Tel.: +33 3 83 59 58 77; fax: +33 3 83 59 57 72.
E-mail address: claire.gaiani@ensaia.inpl-nancy.fr (C. Gaiani).
0927-7765/$ see front matter 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.colsurfb.2009.09.016

The production process of NWI is carried out at a low temperature

with membrane technologies (microltration and ultraltration),
and as such preserves the native structure of the protein. Native
micellar casein (NMC), which is obtained by tangential membrane
microltration of skimmed milk, is also an attractive material due
to its high protein content and can be used as a relevant model
of milk micelles. Therefore, high milk proteins powders are useful
ingredients for food products such as nonfat yogurt, ice cream and
cheese.
Several important technological properties of these powders
depend on particle-water interactions (wettability, rehydration,
. . .) [5,6], particleparticle interactions (owability, stickiness, . . .)
[7] and particleair interactions (oxidation, . . .) [1]. These interactions are in turn inuenced by morphological and physical
properties [8], as well as by the chemical bulk and surface composition [9,10] of powders particles. Better understanding of the surface
composition, especially of lipids, is therefore needed to predict and
appreciate some functional properties of dairy powders. Key functional properties of milk powders being largely affected by surface
composition, it is fundamental to understand how the major milk
components are distributed within the particle. Among the different techniques used for characterizing the surface and/or quantied
the lipids, scanning electron microscopy (SEM) and solvent extraction have traditionally been used. However the distinction between

378

C. Gaiani et al. / Colloids and Surfaces B: Biointerfaces 75 (2010) 377384

the various components and their quantication was impossible by

SEM and this technique was principally used to visualize the surface topography of powder particles [10,11]. The amount of surface
lipids could be quantied by solvent extraction, but it was demonstrated that extractable lipids obtained originates from the powder
surface and unfortunately also from the bulk of the particle [12,13].
Despite its crucial importance, dairy powder surface composition
understanding is still very limited.
Within the last 10 years, X-ray photoelectron spectroscopy
(XPS) has been successfully applied for investigating the surface
composition of dairy powders especially for low fat powders content [9,14,15]. From the C, O and N percentages, surface contents in
proteins, lactose and lipids were deduced. XPS was already used to
characterize powder surface in relation with functional properties
such as wetting [9,10,15] and owability [7]. The technique was
also used to study powder surface modications during storage
[1416]. Many comparisons between bulk and surface demonstrated the over-representation of lipids and proteins at the surface
whereas lactose and minerals were found mostly encapsulated in
the core [9,10,14]. On the best knowledge of the authors, only one
study concerns powder surface investigation by XPS in relation
with the spray-drying temperature [17]. Nevertheless, this study
was limited to the examination of only two spray-drying temperatures and different powders than those studied in this work were
analyzed (skim and whole milk powders).
The aim of the present work was to consider the inuence of
variable spray-drying outlet air temperatures, ranging from 70 to
150 C, upon the surface composition of high milk proteins powders. For this purpose, surface composition was investigated by
X-ray photoelectron spectroscopy (XPS) and related to powders
wetting properties.
2. Materials and methods
2.1. Materials
Native micellar casein powder was a king gift from UMR STLO
(Rennes, France). NMC was obtained from tangential membrane
microltration (0.1 m) of skimmed milk followed by purication
through water dialtration. Native whey isolate powder (Prolacta
90) was kindly provided by Lactalis (Laval, France). NWI was
obtained by membrane tangential ultraltration and dialtration
of microltrate collected during NMC production. Lactose powder
was obtained by Armor Protines (Loudac, France). All the powders are commercial products that has been freshly manufactured
and packed.
Casein and whey proteins from bovine milk (Sigma, France) and
anhydrous lactose (Fluka) were used as reference powders to carry
out XPS calculations on pure grade powders. Lipids were obtained
from NMC and NWI powders after a Folch extraction.
2.2. Laboratory-scale production of powders
The powders (NMC, NWI and lactose) were rst rehydrated to
a solids content of 15% (w/w) the day before spray-drying and
were left under stirring overnight at 20 C in order to obtain a total
rehydration of micellar casein. Sodium azide was also added to
avoid microbial development. Different mixed dispersions (500 ml)
were then prepared by combining NMC, NWI and/or lactose solutions. The blends (NMC/NWI/Lac) used in this work were: 100/0/0
(NMC), 0/100/0 (NWI), 70/0/30 (NMC + Lac), 0/70/30 (NWI + Lac)
and 80/20/0 (NMC + NWI). Finally, each dispersion was spray-dried
using a Mini Spray Drier Buchi B290 (Buchi SARL, Rungis, France)
at the following outlet air temperatures (Tout ): 70, 80, 110, 130 and
150 C. Tout cannot be controlled directly, but depends on the inlet

air temperature and the liquid feed rate. Consequently, Tout was
xed after exploratory trials between 70 and 150 C (80 C being
the outlet air temperature used by our industrial partners).
2.3. Physicochemical characterization of powders
2.3.1. Chemical analysis
The water content of powders was determined by weight loss
after drying 1 g of powder at 105 C for 5 h [18]. The total nitrogen was determined by Kjeldahl [18]. Lactose was determined by
enzymatic method using an Enzytec Lactose/D-Galactose kit. Lipids
were determined according to the Folch method [19]. Ashes were
measured after incineration at 550 C during 5 h [18]. All the experiments were done in triplicate.
2.3.2. Degree of whey protein denaturation
Denaturated whey proteins were precipitated at pH 4.6. The
precipitate was removed by centrifugation (30 min 1000 g) and
the supernatant was analyzed for the protein content according
to the Kjeldahl method. The degree of denaturation was deduced
from Kjeldahl analyses before and after precipitation (mean of three
analyses).
2.3.3. Wetting properties
The wetting properties of powders were determined by static
wetting as described by the FIL method [20] with some modication. The powder (1 g) was poured into 10 ml water. The time
required for all the particles to be submersed is recorded and called
the wetting time. Measurements were done in triplicate.
2.3.4. Physical properties
The particle size distributions were determined by static light
scattering (Mastersizer S, Malvern Instruments Ltd., Malvern, UK)
with a 5 mW HeNe laser operating at a wavelength of 632.8 nm
with a 300F lens. The size distributions were determined using a dry
powder feeder attachment and the standard optical model presentation for particles dispersed in air was used. The results obtained
are average diameters calculated from Mie theory (mean of three
analyses). The criterion selected was the d(50) that means that 50%
of the particles have a diameter lower than this criterion.
2.4. XPS analyses
2.4.1. Equipment
The XPS analyses were carried out with a Kratos Axis
Ultra spectrometer (Kratos Analytical, Manchester, UK) using a
monochromatic Al K source. The powder samples were spread
to the sample holder using a double side conductive adhesive
tape and then degassed overnight prior analyses. All spectra were
recorded at 90 take-off angle, the analyzed area being about
700 m 300 m. Survey spectra were recorded with 1.0 eV step
and 160 eV analyzer pass energy and the high resolution regions
with 0.05 eV step and 20 eV pass energy. In both cases the hybrid
lens mode was employed.
During the data acquisition the Kratos charge neutralizer system was used on all specimens with the following settings: lament
current 1.6 A, charge balance 2.4 V, lament bias 1.0 V and magnetic
lens trim coil 0.375 A. The CC (CH) carbon was set to 284.60 eV
and therefore used as an internal energy reference. With these
parameters we could obtain C1s signal with sharp, symmetric components with a FWHM of 1.2 eV.
Spectra were analyzed using the Vision software from Kratos
(Vision 2.2.2). A Shirley base line was used for the subtraction of
the background and Gaussian/Lorentzian (70/30%) peaks were used
for spectral decomposition. Quantication was performed using

C. Gaiani et al. / Colloids and Surfaces B: Biointerfaces 75 (2010) 377384

Table 1
Relative elemental composition of the reference samples measured by XPS (excluding inorganic ions).
Initial powder

Relative atomic concentration (mol%)

C

Lactose
Mean valuea
Theoreticalb
Errorc

61.6 2.5
52.2
15.2

38.4 2.8
47.8
24.5

Casein
Mean valuea
Theoreticalb
Errorc

68.6 0.3
65.0
5.2

17.6 0.8
19.0
7.9

13.8 0.1
16.0
15.9

Whey proteins
Mean valuea
Theoreticalb
Errorc

68.0 0.5
65.0
4.4

18.0 0.9
19.0
5.5

14.1 0.2
16.0
13.5

Lipids (mainly phospholipids)

Mean valuea
80.3 0.6
82.0
Theoreticalb
2.1
Errorc

18.4 0.7
16.0
13.0

1.3 0.2
2.0
53.8

Mean value of three independent measurements S.D.

Calculated on the basis of the chemical formula.
Differences between values (%) = [(mean value theoretically calculated
value)/(mean value)] 100.
a

b
c

the photoemission cross-sections and the transmission coefcients

given in the Vision package.
2.4.2. Application of XPS to the surface content of dairy powders
The foundation of the technique and its application to dairy
powders has been described elsewhere [9,10,14]. Briey, the relative atomic concentration of carbon, oxygen and nitrogen in the
surface layer (10 nm) of the powder was quantied and used in a
matrix formula related to the surface content of the different compounds making up the sample (i.e. lactose, proteins and lipids).
Ca2p , S2p , and P2p have been also detected at levels lower than 1% for
all of the powders. Consequently, these elements were neglected
in the quantication procedure:
I C = P I CP +  La I CLa +  Li I CLi
O

(1)
(2)

I N = P I NP +  La I NLa +  Li I NLi

(3)

100 = P +  La +  Li

(4)

+ La I

OLa



OLi

I = P I

OP

+ Li I

In proteins, according to the chemical stoechiometry, P = (mol

of C in proteins + mol of O in proteins + mol of N in proteins)/(mol
of C + mol of O + mol of N); the same equations hold for lipids and
lactose. In the following (Eqs. (1)(4)), this ratio will be dened as
= surface content. IC , IO , IN were the mole fractions of carbon, oxygen, and nitrogen in the sample surface. These values were obtained
from the areas of the C1s , O1s and N1s XPS peaks. I CP , I CLa , I CLi were
the mole fractions of carbon in protein, lactose and lipids; I OP , I OLa ,
I OLi were the mole fractions of oxygen in protein, lactose and lipids
and I NP , I NLa , I NLi were the mole fractions of nitrogen in protein, lactose and lipids. Experimental values obtained from the reference
powders (pure casein, whey proteins, lactose and lipids) have been
used to carry out these calculations (Table 1).
By solving the matrix, (P), ( La), ( Li), respectively the protein, lactose and lipids surface contents were determined. For
the direct comparison with the bulk composition of powders,
the relative surface content calculated must be converted into
mass concentration. However, as already noted by some authors
[14,17,21] the proportions of the sums of C + O + N molar concentrations due to lactose, protein and lipid are very close (within 1%)
to the relative mass concentrations of these compounds.

379

2.5. FTIR
2.5.1. Equipment
FTIR in transmission mode was performed on NWI powders in
order to evaluate whey denaturation and aggregation during spraydrying (70, 80, 110, 130 and 150 C). A Tensor 27 mid-FTIR Bruker
spectrometer (Bruker, Karlsruhe, Germany) equipped with an optical cell set with a KBr beamsplitter and a DTGS detector was used.
Scanning rate was 20 kHz and 256 scans were used for both reference and sample between 4000 and 850 cm1 at 2 cm1 resolution.
Samples of powders were mixed with KBr (potassium bromide) by
grinding 25 mg samples and 475 mg KBr. The powder mixture was
then crushed in a mechanical die press to form a translucent pellet
through which the beam of the spectrometer can pass.
The reference was rst recorded at 20 C on air and then the
powder pellet was analyzed. All treatments excepted principal
component analysis (PCA) were carried out using OPUS software
(Bruker, Karlsruhe, Germany).
2.5.2. Data treatments
Raw absorbance spectra were smoothed using a 13 points
Savitsky-Golay smoothing function and cut between 1720 and
1580 cm1 . Elastic baseline correction using 200 points was applied
to spectra. Spectra were then centered and normalized using OPUS
software. The Unscrambler v7.6 software (CAMO ASA, Oslo, Norway) was used to calculate the PCA on centered and normalized
FTIR spectra.
3. Results
3.1. Powders characterization
3.1.1. Chemical and size characterizations
The chemical composition of the powders is reported in Table 2.
For NMC powders, lactose and lipids traces are present. The amount
of minerals is high and found in the ash fraction. In comparison,
NWI powders present fewer minerals. Lactose and lipids traces are
also found in these powders. The moisture of all the powders is
globally constant at 4%. The chemical composition agreed reasonably well with the different mixes realized (Section 2.2). Indeed,
the experimental ratio protein/lactose is found close to 70/30 for
NMC + Lac and NWI + Lac powders. The average size of powder particles is also determined after spray-drying (Table 2). No signicant
differences are found between the powders, the d50 being around
8 m whatever the spray-drying temperature.
3.1.2. Whey proteins denaturation and aggregation
With increasing Tout , the only signicant difference between
powders (identical bulk composition) is the percentage of protein
denaturation. This percentage clearly increases with Tout (Table 2).
For NWI, the denaturation increases from 2.8% and 19.5% between
70 and 150 C. Lactose addition to NWI (NWI + Lac) leads to less
denaturation (from 1.7% to 8.7%). In addition, principal components analysis (PCA) was used to study the correlation between
the FTIR spectra (amine I) and the different outlet air temperatures
for NWI powders. PCA are vectors described by linear combinations
of the absorbancies measured in the infrared spectrum related to
amide I band. Principal components 1 (PC 1) and 2 (PC 2) account
for 99% and 1% respectively of spectral changes observed. Consequently, only the rst component is of interest. Its loading plot
presented in Fig. 1 shows that bands at 1618 and 1705 cm1 are
opposed at the region near 1654 cm1 . These three regions are easily identied as the most important absorption bands affected by
Tout increased. A simultaneous increase in the adsorption band at
1618 and 1705 cm1 reveals the formation of intermolecular sheet structures and intramolecular -strand or turns as the outlet

380

C. Gaiani et al. / Colloids and Surfaces B: Biointerfaces 75 (2010) 377384

Table 2
Characterization of the powders spray-dried (mean of triplicate analysis S.D.).
Outlet air temperature ( C)

Powder

Chemical composition (%)

Denaturation (%)

d50 (m)

Wetting time (s)

Moisture

Ash

Lipids

Proteins

4.7
4.9
3.9
4.7
3.8

0.2
0.0
0.1
0.3
0.1

7.5
2.3
6.4
1.9
7.1

0.0
0.2
0.4
0.5
0.1

0.3 0.1
0.4 0.2
TRACES
0.5 0.1
TRACES

87.3
91.9
63.1
64.4
88.8

1.2
0.8
0.8
1.3
1.5

0.2
0.5
26.6
28.5
0.3

0.0
0.1
0.5
0.4
0.0

2.8 0.9

1.7 0.4

7.5
8.3
7.9
8.5
7.9

0.1
0.2
0.0
0.1
0.0

932
1498
543
1085
1001

53
35
39
39
43

80

NMC
NWI
NMC + Lac
NWI + Lac
NMC + NWI

5.0
4.8
3.8
3.8
5.9

0.2
0.4
0.3
0.0
0.2

7.7
2.1
6.9
1.6
7.4

0.6
0.2
0.2
0.5
0.3

0.8 0.0
0.3 0.4
0.7 0.1
TRACES
TRACES

86.4
92.2
62.0
65.7
86.1

1.1
0.7
0.5
0.3
1.1

0.1
0.6
26.6
28.9
0.6

0.0
0.0
0.1
0.0
0.0

4.4 1.1

2.5 0.1

6.9
8.6
7.2
9.1
7.4

0.1
0.1
0.0
0.1
0.2

631
1290
431
991
882

31
34
36
27
21

110

NMC
NWI
NMC + Lac
NWI + Lac
NMC + NWI

4.9
4.8
5.0
4.1
3.4

0.0
0.0
0.7
0.2
0.5

7.6
2.2
5.6
1.3
6.9

0.1
0.0
0.0
0.0
0.2

0.5 0.2
TRACES
TRACES
TRACES
TRACES

86.7
92.2
62.7
66.1
89.1

1.4
1.5
1.8
1.9
0.7

0.1
0.8
26.7
28.5
0.6

0.1
0.0
0.3
0.1
0.0

9.2 0.2

7.4 0.4

8.6
8.8
6.0
7.1
7.4

0.1
0.0
0.1
0.1
0.2

642
1120
205
666
811

21
37
27
15
27

130

NMC
NWI
NMC + Lac
NWI + Lac
NMC + NWI

3.3
3.5
3.8
3.3
5.7

0.3
0.8
0.2
0.1
0.2

8.2
2.3
7.1
1.4
7.0

0.4
0.0
0.1
0.1
0.5

0.5 0.1
0.6 0.4
TRACES
0.5 0.1
TRACES

87.6
92.3
63.2
65.6
86.8

0.8
0.9
1.9
1.7
1.5

0.4
1.3
27.9
29.3
0.5

0.1

0.3 12.2 0.1

0.4

0.2 6.5 0.4

0.0

8.6
8.6
6.4
8.7
8.3

0.2
0.2
0.1
0.2
0.2

639
1179
202
557
758

12
27
12
15
16

150

NMC
NWI
NMC + Lac
NWI + Lac
NMC + NWI

5.0
4.9
4.8
4.3
3.9

0.5
0.4
0.3
0.1
0.3

7.9
2.4
6.3
1.3
7.1

0.2
0.1
0.0
0.1
0.3

0.3 0.0
0.4 0.1
TRACES
TRACES
0.6 0.1

86.5
91.6
63.3
67.8
88.7

1.4
1.5
0.5
0.6
0.7

0.3
0.7
25.6
26.6
0.5

0.0

0.1 19.5 0.3

0.3

0.4 8.7 0.6

0.2

8.0
8.7
8.1
8.1
8.3

0.2
0.2
0.1
0.1
0.0

623
1131
201
499
628

16
8
15
12
13

70

NMC
NWI
NMC + Lac
NWI + Lac
NMC + NWI

Lactose

Lac: lactose; NMC: native micellar casein; NWI: native whey isolate.

air temperature rises. Concurrently, a decrease in bands characteristic of -helix/unordered structures is noticed (1654 cm1 ). This
is the spectral signature of proteins aggregation [22,23].

3.1.3. Elemental composition of the reference powders

The elemental composition of the reference powders was measured on pure components (whey proteins, casein, lactose and
lipids) and used to calculate powder surface composition (Table 1).
The elemental composition of lipids, whey and casein agreed reasonably well with the values calculated from the chemical formula
and from literature. However, for lactose, the atomic concentration
ratio was not in full agreement with the expected stoichiometric
values.

Fig. 1. Loading plot of rst PC obtained after principal component analysis of FTIR
spectra of NWI powders spray-dried at 70, 80, 110, 130 and 150 C. Each value is the
average of 3 experiments.

3.2. Powders surface composition

3.2.1. Bulk/surface differences
The differences observed between bulk and surface composition
are in the same direction whatever the spray-drying temperature
used (Table 3). Consequently, as industrial dairy powders are generally spray-dried at 80 C, comparison between bulk and surface is
commented in details only for powders spray-dried at this temperature. As expected, the principal constituent found at the surface
of NMC powder is protein (99.4%). Lactose traces (0.6%) are also
present. When 30% of lactose is added to the bulk (NMC + Lac), the
surface appears very different. Lactose is supposed to be located in
the core of the powder (only 16% at the surface instead of 30% in the
bulk) and lipids are found over-represented at the surface (around
5%). For NWI powders, the percentage of lipids at the surface is
very high (more than 30%). When lactose is added (NWI + Lac), the
surface enrichment in lipids is important (21%) and lactose is less
present at the surface (only 12.7%). For powders containing NMC
and NWI (80/20), proteins are the main component at the surface
(96%), lipids are also found over-represented at the surface (4%) and
lactose is lacking. A differentiation between casein and whey at the
surface using the XPS technique was impossible; their atomic concentration (in C, O and N) being too close. To summarize, a surface
enrichment in lipids and in proteins is observed whatever powder
formulation and the spray-drying temperature. Furthermore, the
surface enrichment in lipids is much stronger for NWI containing
powders than for NMC containing powders.
3.2.2. Effect of drying temperature
The surface composition of casein containing powders is only
slightly modied by the drying temperature. When the outlet
air temperature increased (from 70 to 150 C), the percentage of
lipids decreased from 5.3% to 0% and from 8.9% to 0% for NMC
and NMC + Lac, respectively (Fig. 2). While the lipids concentration decreases slightly for casein containing powders, the surface
ratio of lactose to protein increases slightly from 0.18 to 0.25 for

C. Gaiani et al. / Colloids and Surfaces B: Biointerfaces 75 (2010) 377384

381

Table 3
Powder composition on the dry matter basis and surface composition (assumed that the powders are composed of the three main compounds, i.e. protein, lactose and lipids)
(mean of duplicate analysis).
Outlet air temperature ( C)

Powder

Bulk composition (mass%)

Surface composition (proportion of

C + O + N molar concentration %a )

Protein

Lactose

Lipids

Protein

Lactose

Lipids

70

NMC
NWI
NMC + Lac
NWI + Lac
NMC + NWI

99.4
99.0
70.3
69.0
99.7

0.2
0.6
29.7
30.5
0.3

0.4
0.4

0.5

93.9
66.1
76.8
62.9
89.6

0.8
0.1
14.3
9.5
0.3

5.3
33.8
8.9
27.6
10.1

80

NMC
NWI
NMC + Lac
NWI + Lac
NMC + NWI

99.0
99.0
69.4
69.5
99.4

0.1
0.6
29.8
30.5
0.6

0.9
0.4
0.8

99.4
69.8
79.3
66.3
95.6

0.6
0.0
16.0
12.7
0.0

0.0
30.2
4.7
21.0
4.4

110

NMC
NWI
NMC + Lac
NWI + Lac
NMC + NWI

99.3
99.1
70.1
69.9
99.3

0.1
0.9
29.9
30.1
0.7

0.6

97.8
88.6
83.5
81.1
94.3

2.2
0.0
16.5
18.8
0.2

0.0
11.4
0.0
0.1
5.5

130

NMC
NWI
NMC + Lac
NWI + Lac
NMC + NWI

99.0
98.0
69.4
68.8
99.4

0.5
1.4
30.6
30.7
0.6

0.5
0.6

0.5

98.5
88.1
80.3
79.2
92.9

1.5
0.0
19.7
20.8
0.0

0.0
11.9
0.0
0.0
7.1

150

NMC
NWI
NMC + Lac
NWI + Lac
NMC + NWI

99.3
98.8
71.2
71.8
98.8

0.4
0.7
28.8
28.2
0.6

0.3
0.5

0.6

96.5
85.4
80.3
77.0
99.2

3.5
4.0
19.7
19.2
0.8

0.0
10.6
0.0
3.8
0.0

Lac: lactose, NMC: native micellar casein, NWI: native whey isolate.
a
Close to the mass concentration of the compound (%).

3.3. Wetting properties

a bulk ratio of 0.43 (Fig. 3). For whey proteins containing powders, Tout modied strongly the percentage of lipids at the surface.
Indeed, lipids surface content decreased from 33.8% to 10.6% and
from 27.6% to 3.8% for NWI and NWI + Lac powders, respectively
(Fig. 2). While the lipids surface concentration decreases strongly as
the outlet temperature rises, the surface of ratio lactose to protein
increases from 0.15 to 0.25 (Fig. 3), for a bulk ratio of 0.43. Powders containing a mix of casein and whey (80/20) are also affected
by Tout . For these powders, lipids surface content decreased from
10.1% to 0% (Fig. 2).
In conclusion, a lower drying temperature favors the appearance
of proteins over lactose at the surface of the particles and enhances
the surface enrichment in lipids.

The wetting properties of the powders are reported in Table 2.

All the powders studied are completely wetted in less than
1500 s. However, important differences are observed between the
powders. As expected, powders containing lactose presented the
shortest wetting times. By its hygroscopic nature, addition of lactose before spray-drying could greatly improve water transfer
particularly if the lactose is located at the surface. For example, NMC
powder spray-dried at 80 C is wetted in 630 s whereas the wetting time of NMC + Lac powder spray-dried in the same condition
is signicantly shortened around 230 s. The tendency is exactly the
same for NWI powder (1290 s) in comparison with NWI + Lac pow-

Fig. 2. Change in lipids surface composition (%) of the powders with varying outlet
air temperature (from 70 to 150 C) (mean of duplicate analysis). Lac: lactose; NMC:
native micellar casein; NWI: native whey isolate.

Fig. 3. Surface lactose/surface protein ratio obtained for powders containing lactose
with varying outlet air temperature (from 70 to 150 C) (mean of duplicate analysis).
Lac: lactose; NMC: native micellar casein; NWI: native whey isolate.

382

C. Gaiani et al. / Colloids and Surfaces B: Biointerfaces 75 (2010) 377384

disk is not perfectly covered, some regions could be exposed to the

X-ray photons. For example, comparative determinations for pure
lactose show that the relative atomic concentration of the sample
prepared by dusting could deviate from the theoretical concentration particularly for the C1s as observed with the lactose reference
[21].

Fig. 4. Wettability of the powders spray-dried at various outlet air temperatures

(mean of three analyses) in relation with lipids surface composition (mean of two
independent analyses). Lac: lactose; NMC: native micellar casein; NWI: native whey
isolate.

der (691 s). Important variations are also observed with the nature
of the protein wetted. NMC powders present systematically shorter
wetting times than NWI powders. A correlation to evaluate the relation between wetting properties and lipids surface is presented in
Fig. 4. For the 25 powders studied, a direct correlation is found
between the wetting time and the surface concentration. Higher
correlation coefcients are obtained for subset of samples which
contain NMC and have the same formulation: r2 = 0.99; 0.98 and
0.99 for NMC, NMC + Lac and NMC + NWI, respectively. For whey
powders the correlation coefcients are lower, respectively 0.85
and 0.89 for NWI and NWI + Lac. To summarize, powder particles
become wetted more rapidly at high spray-dried temperatures (less
surface lipids).
4. Discussion
4.1. Powders surface composition
4.1.1. Evaluation of the matrix formula obtained from reference
powders (Table 1)
The elemental composition in proteins (casein and whey
proteins references) agreed reasonably well with the values calculated from the chemical formula and obtained in other studies
[9,10,14,21]. The differences observed between theoretical and
experimental values may be due to some impurities in reference
samples [24]. Residual lipids used as references were extracted
from NMC and NWI powders. In agreement with others studies
[15,25], lipids in these high proteins powders are found to be mainly
phospholipids. Comparison between theoretical and experimental
values for lipids agreed well. The O percentage was slightly higher
for experimental values. Lipids could be slightly oxidized during
preparation, resulting in an increase in oxygen content compared
to the theoretical value. For lactose, the atomic concentration ratios
were not in total agreement with the expected stoichiometric values and the difference between values was high (15.2% for C and
24.5% for O). The C peak intensity was more important and may be
due to carbon contaminant and impurities collected from the atmosphere during the sample transfer into the XPS chamber or during
product manufacturing [24]. The differences observed may be also
explained by the sample preparation that occurs before XPS analysis. Three methods are commonly used to x the powder sample
to the adhesive tape [21]: dusting (used in this study), pressing and
packing. The sample mounting technique may have an effect on
the surface composition obtained. With the dust technique, if the

4.1.2. Surface composition

The surface composition found for casein containing powders
was in agreement with previous study [9]. These authors analyzed
by XPS blends of NMC and NMC + Lac spray-dried in an industrial
atomizer instead of a laboratory one in this study. The results were
comparable. The surface of whey powders was found largely covered by lipids (30% at 80 C). For the same powder, another study
[10] established a lipid surface content higher (53%). The differences could be explained through the bulk composition of the
powders. Residual lipids were found around 6% instead of 0.4% in
our study. However, it is well known that small variations in the
average bulk lipid concentration could results in very large modications in the lipids surface content, particularly at low lipids
content [26,27]. Depending on proteins nature (whey proteins or
casein), different results were found on lipids surface enrichment.
For the same blends (prot/lac: 100/0 and 70/30), casein presents
lower lipids surface enrichment than whey proteins. This was consistent with the results of earlier investigations [1,28,29] which
found caseins more surface-active than whey proteins in the sense
that they give a lower surface tension at air-water interface. Thus,
we could assume that casein powder would give a higher surface
content than whey protein powder.

4.2. Relation between spray-drying temperature and wetting

properties
For all formulations, lipids surface content was more important
at low Tout and could explain the poorer wetting times found. A
correlation was established between the wetting properties and
lipids surface content. This correlation was linear (0.85 < r2 < 0.99)
for subsets of samples which have the same formulation (Fig. 4).
Surface compositions containing hydrophobic components (lipids)
are known to give poor wetting properties whereas surface with
hygroscopic components (lactose, minerals) gives good wetting
properties [29]. In agreement with our results (Table 2), it was
previously shown [9,29] that increase bulk lactose improved the
wetting properties. The wetting time was also improved when the
surface ratio of lactose to protein is augmented (increasing Tout ).
The presence of denatured proteins could be also detrimental to
wetting properties [30,31]. It was not the case in this study, the best
wetting time being found for powders spray-dried at high temperature. It therefore seems that the lipids surface content is the more
important together with the ratio, lactose to protein; denaturation
of whey protein being secondary.
Nevertheless, presence of lipids, proteins and/or lactose at the
surface may not be the only foundation to explain the wetting properties of the powder. External factors are also inuencing dairy
powder rehydration; the principals are temperature, concentration, and stirring conditions [32]. In this study, these factors were
kept constant for all formulations. The wetting ability is also highly
connected to the physical structure of the particle (size, density,
specic surface, porosity). As the spray-drying conditions were
different for the powders, we could suppose some differences in
porosity or specic surface area. Indeed, powders spray-dried at
high temperatures could present more cracks and pores [33]. However, particle size (Table 1) and surface observed by MEB (data not
shown) were comparable.

C. Gaiani et al. / Colloids and Surfaces B: Biointerfaces 75 (2010) 377384

4.3. Inuence of spray-drying temperature upon surface

composition
4.3.1. Lipids
The most remarkable change in surface composition with
spray-drying temperature is observed in lipids surface content. It
signicantly decreased for all the powders studied when the outlet
air temperature increased. These observations were in agreement
with another study which investigated only two spray-drying temperatures (Tout = 85 and 105 C) [17]. In addition, greater changes in
the lipid surface composition were reported for skim milk powders
and low feed solids content (10%), compared to high feed solids
content (30%). In the present study, the feed solid content can be
considered as low (15%) and may explain the important surface
modications observed. For whole milk powder, the drying temperature did not appear to greatly affect the surface composition
as all the surface was covered by lipids [17]. In the literature and
to the best knowledge of the authors, inuence of drying temperature on surface composition was not yet investigated on high milk
proteins powders.
4.3.2. Lactose to protein ratio
Higher drying temperature favors the appearance of lactose
over protein at the surface of the particle (Fig. 3). Nevertheless,
lactose/protein ratio was only slightly modied for casein containing powders whereas the ratio was considerably increased for
whey proteins powders. This difference may be explained by proteins denaturation. Heating casein micelles below 140 C has little
effect on their stability whereas the same heat treatment induces
important changes in the structure of whey proteins. It was found
that increasing Tout induced whey proteins denaturation (Table 2)
and aggregation (Fig. 1). The sensitivity of the amide I vibration
to secondary structure make it possible to study proteins folding, unfolding, aggregation, etc. [23]. It has been proposed that
the heat treated proteins are less surface-active and competes less
favorably for the interface than the untreated protein [30]. Proteins hydrophobicity and protein molecular mass are both affecting
the interface. Increasing the outlet air temperature could result to
increase hydrophobicity of whey proteins (as a result of denaturation) and to increase its molecular mass (as a result of aggregation)
both inuencing interface formation.
4.3.3. Surface formation mechanisms
Depending on the authors, two scenarios are generally mentioned about the surface formation of the drying droplet. The rst
one (scenario 1) was based on surface-active components adsorption on air/liquid interface of the spray droplet [14,34]. According to
this scenario, protein adsorbs preferentially to the air/liquid interface and appears on the powder surface after spray-drying whereas,
fat droplets are largely found inside the particles [14,29]. The second hypothesis (scenario 2) is based on solid/solutes segregation
[35,36]. During spray-drying, diffusion of water towards the surface is accompanied by an opposite diffusion of solids toward the
center of the particle. As a consequence, solutes with a high molecular weight will accumulate more slowly than smaller molecules.
Over time, the outer surface of the droplet would then become
enriched in the larger molecules. As lipids and proteins present a
lower diffusivity than lactose and inorganic ions, the surface would
be enriched in these components [10,36].
It is likely that a combination between these two scenarios could
be possible in this work. During the rst step (atomization), migration of surface-active components toward the surface could occur,
then during the last step (spray-drying) a diffusion of solutes from
the surface to the interior of the spray droplets could follow. These
hypotheses assumed that there is enough time for the components
to migrate. This is probably not the case in this study. According

383

to our results, surface-active components are found at the surface

(proteins and phospholipids) at low Tout . As the spray-drying temperature increased, the surface may solidify and immobilizes the
components quickly. Consequently, surface-active molecules could
not have enough time to migrate at the surface; the surface of the
nal particle being closer to the bulk composition. By increasing
outlet air temperature over 110 C, the diffusion of components
during spray-drying seems to be interrupted before the system has
reached an equilibrium.
5. Conclusion
It was shown that surface composition modication (more
or less surface lipids and/or lactose) was possible on high milk
proteins powders by modifying the spray-drying outlet air temperature. This study could help in the future to formulate powders with
controlled surface composition which in turn will allow a better
control of powder functionalities.
However, one point needs more investigation. Very little is
known about dynamics of the formation of the drying droplets. In
the future, knowledge on this point will be required to understand
and control mechanisms behind surface formation.
Acknowledgements
We thank J. Lambert, LCPME Engineer CNRS (Nancy), for providing XPS analysis and J.J. Ehrhardt for scientic assistance. The
authors thank also ARILAIT RECHERCHE (Paris) for nancial support
and its scientic committee is gratefully acknowledged.
References
[1] M.K. Keogh, B.T. OKennedy, J. Kelly, M.A. Auty, P.M. Kelly, A. Fureby, A.M. Haahr,
Stability to oxidation of spray-dried sh oil powder microencapsulated using
milk ingredients, Food Chemistry and Toxicology 66 (2) (2001) 217224.
[2] A. Millqvist-Fureby, Characterisation of spray-dried emulsions with mixed fat
phases, Colloids and Surfaces B: Biointerfaces 31 (2003) 6579.
[3] A. Gharsallaoui, G. Roudaut, O. Chambin, A. Voilley, R. Saurel, Applications of
spray-drying in microencapsulation of food ingredients: an overview, Food
Research International 40 (9) (2007) 11071121.
[4] V.V. Mistry, Manufacture and application of high milk protein powder, Le Lait
82 (2002) 515522.
[5] C. Gaiani, S. Banon, J. Scher, P. Schuck, J. Hardy, Use of a turbidity sensor to characterize micellar casein powder rehydration: inuence of some technological
effects, Journal of Dairy Science 88 (2005) 27002706.
[6] C. Gaiani, J. Scher, P. Schuck, J. Hardy, S. Desobry, S. Banon, The dissolution
behaviour of native phosphocaseinate as a function of concentration and temperature using a rheological approach, International Dairy Journal 16 (12)
(2006) 14271434.
[7] E.H.J. Kim, D.C. Xiao, D. Pearce, Effect of surface composition on the owability
of industrial spray-dried dairy powders, Colloids and Surfaces B: Biointerfaces
46 (3) (2005) 182187.
[8] T.J. Buma, Free fat in spray-dried whole milk. 3. Particle size. Its estimation,
inuence of processing parameters and its relation to free fat, Netherland Milk
Dairy Journal 25 (1971) 5372.
[9] C. Gaiani, J.J. Ehrhardt, J. Scher, J. Hardy, S. Desobry, S. Banon, Surface composition of dairy powders observed by X-ray photoelectron spectroscopy and
effects on their rehydration properties, Colloids and Surfaces B: Biointerfaces
49 (1) (2006) 7178.
[10] E.H.J. Kim, X.D. Chen, D. Pearce, Surface characterization of four industrial spray
dried dairy powders in relation to chemical composition, structure and wetting
properties, Colloids and Surfaces B: Biointerfaces 26 (3) (2002) 197212.
[11] V.V. Mistry, H.N. Hassan, D.J. Robinson, Effect of lactose and protein on the
microstructure of dried milk, Food Structure 11 (1992) 7382.
[12] T.J. Buma, Free fat in spray-dried whole milk. 2. An evaluation of methods for
the determination of free-fat content, Netherland Milk Dairy Journal 25 (1971)
4252.
[13] M.L. Vignolle, C. Lopez, R. Jeantet, C. Lopez, P. Schuck, Free fat, surface fat and
dairy powders: interactions between process and product. A review, Le Lait 87
(2007) 187236.
[14] P. Fldt, Surface composition of spray dried emulsions, PhD Thesis, Lund University Department of Food Engineering, 1995.
[15] C. Gaiani, J. Scher, J.J. Ehrhardt, M. Linder, P. Schuck, S. Desobry, S. Banon, Relationships between dairy powder surface composition and wetting properties
during storage: importance of residual lipids, Journal of Agricultural and Food
Chemistry 55 (16) (2007) 65616567.

384

C. Gaiani et al. / Colloids and Surfaces B: Biointerfaces 75 (2010) 377384

[16] C. Gaiani, P. Schuck, J. Scher, J.J. Ehrhardt, E. Arab-Tehrany, M. Jacquot, S. Banon,

Native phosphocaseinate powder during storage: lipids released onto the surface, Journal of Food Engineering 94 (2) (2009) 130134.
[17] E.H.J. Kim, D.C. Xiao, D. Pearce, Surface composition of industrial spray-dried
milk powders. 2. Effect of spray drying conditions on the surface composition,
Journal of Food Engineering 94 (2) (2009) 169181.
[18] ADPI (American Dairy Product Institute, E., IL), Standards for Grades of Dry
Milks Including Methods of Analysis, Bulletin 916 (2002).
[19] J. Folch, M. Lees, G.H. Sloane-Stanley, A simple method for the isolation and
purication of total lipids from animal tissues, The Journal of Biological Chemistry 226 (1957) 497509.
[20] FIL (Fdration Internationale Laitire, Brussels, Belgium), Dtermination de la
dispersibilit et de la mouillabilit, No: V04-364 (1985).
[21] E.H.J. Kim, Surface composition of industrial spray dried dairy powders and
its formation mechanisms, PhD thesis, Department of Chemical and Materials
Engineering, the University of Auckland, 2008.
[22] A. Kher, P. Udabage, I. McKinnon, D. McNaughton, M.A. Augustin, FTIR
investigation of spray-dried milk protein concentrate powders, Vibrational
Spectroscopy 44 (2007) 375381.
[23] A. Barth, A review: infrared spectroscopy of proteins, Biochimica et Biophysica
Acta 1767 (2007) 10731101.
[24] P.A. Gerin, P.B. Dengis, P.G. Rouxhet, Performance of XPS analysis of model
biochemical compounds, Journal of Chemical Physics 92 (1995) 10431065.
[25] F. Sanchez-Juanes, J.M. Alonso, L. Zancada, P. Hueso, Distribution and fatty
acid content of phospholipids from bovine milk and bovine milk fat globule
membranes, International Dairy Journal 19 (5) (2009) 273278.
[26] J.J. Nijdam, T.A.G. Langrish, The effect of surface composition on the functional
properties of milk powder, Journal of Food Engineering 77 (2006) 919925.

[27] A.K. Shrestha, T. Howes, B.P. Adhikari, B.J. Wood, B.R. Bhandari, Effect of protein
concentration on the surface composition, water sorption and glass transition
temperature of spray-dried skim milk powders, Food Chemistry 104 (2007)
14361444.
[28] C. Vega, Y.H. Ross, Invited review: spray-dried dairy and dairy-like emulsions
compositional considerations, Journal of Dairy Science 89 (2006) 383401.
[29] P. Fldt, B. Bergenstahl, The surface composition of spray dried protein/lactose
powders, Colloids and Surfaces A: Physicochemical and Engineering Aspects
90 (1994) 183190.
[30] A. Millqvist-Fureby, U. Elofsson, B. Bergenstahl, Surface composition of spraydried milk protein stabilized emulsions in relation to pre-heat treatment of
proteins, Colloids and Surfaces B: Biointerfaces 21 (2001) 4758.
[31] E.L. Sliwinski, B.W.M. Lavrijsen, J.M. Vollenbroek, H.J. van der Stege, M.A.J.S. van
Boekel, J.T.M. Wouters, Effects of spray drying on physicochemical properties
of milk protein-stabilised emulsions, Colloids and Surfaces B: Biointerfaces 31
(2003) 219229.
[32] B. Freudig, S. Hogekamp, H. Schubert, Dispersion of powders in liquid in a stirred
vessel, Chemical Engineering and Processing 38 (1999) 525532.
[33] T.J. Buma, Free fat in spray-dried whole milk. 8. The relation between free fat
content and particle porosity of spray-dried whole milk, Netherlands Milk Dairy
Journal 25 (1979) 123144.
[34] J.J. Nijdam, T.A.G. Langrish, An investigation of milk powders produced by a
laboratory-scale spray dryer, Drying Technology 23 (5) (2005) 10431056.
[35] E.H.J. Kim, D.C. Xiao, D. Pearce, On the mechanisms of surface formation and the
surface compositions of industrial milk powders, Drying Technology 21 (2003)
265278.
[36] S. Kentish, M. Davidson, H. Hassan, C. Bloore, Milk skin formation during drying,
Chemical Engineering Science 60 (3) (2005) 635646.