Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Powder Technology 372 (2020) 394–403

Contents lists available at ScienceDirect

Powder Technology

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

Effect of storage conditions on physicochemical and microstructural


properties of skim and whole milk powders
Arissara Phosanam a,b, Jayani Chandrapala a, Thom Huppertz c,d, Benu Adhikari a,⁎, Bogdan Zisu e
a
School of Science, RMIT University, Bundoora, Melbourne, VIC 3083, Australia
b
Faculty of Natural Resources and Agro-Industry, Kasetsart University, Chalermphakiat Sakon Nakon Province Campus, 47000, Thailand
c
FrieslandCampina, Amersfoort 3818, the Netherlands
d
Food Quality and Design Group, Wageningen University, Wageningen 6708, the Netherlands
e
Spraying System Co., Fluid air, Melbourne, VIC 3029, Australia

a r t i c l e i n f o a b s t r a c t

Article history: This study examined the effects of storage conditions on physicochemical and microstructural characteristics of
Received 18 April 2020 whole milk powder (WMP) and low-, medium- and high-heat skim milk powder (SMP). Powders were stored at
Received in revised form 3 June 2020 three temperatures (25 °C, 35 °C and 45 °C) and three humidity levels (11%, 44% and 85%) for 21 days. WMP
Accepted 7 June 2020
exhibited lower equilibrium moisture content (EMC) than SMP at these storage conditions. SMP and WMP stored
Available online 10 June 2020
at 45 °C had lower EMC and monolayer moisture content than those stored at 25 °C and 35 °C. WMP had lower Tg,
Keywords:
crystallinity and caking strength than SMP. At 85% RH, the particles became irregularly shaped, caused by lactose
Milk powders crystallization. Crystallization of lactose and caking of powders became prominent when Tg decreased below the
Storage conditions storage temperature. Crystallization of lactose in these powders increased caking strength, Maillard reaction and
Glass transition surface free fat, particularly at high RH (85%) and high temperature (45 °C) conditions.
Crystallinity © 2020 Elsevier B.V. All rights reserved.
Surface-free fat
Caking

1. Introduction also results into excess fat on the powder surface and subsequent crys-
tallization of lactose hastens the migration of fat to surface and thus con-
Milk powder is convenient to use, saves transportation costs and can tributes to caking [14].
be stored for a relatively long time (12–24 months) at ambient temper- Whole milk powder (WMP) and skim milk powder (SMP) are
ature [1–3]. Spray drying is one of the most commonly applied tech- widely used as ingredients in infant formula products, bakery products,
niques for converting fluid milk into powder. Despite its many ice cream, yoghurt, dairy beverages, confectionary products and proc-
advantages, spray drying process causes chemical and physical modifi- essed meats due to their desirable functional (e.g., solubility, emulsify-
cations in milk components that increase propensity of Maillard reac- ing, and foaming) and nutritional properties. The main compositional
tion, milk fat oxidation and crystallization of lactose [4]. The nature difference between WMP and SMP is the fat content; WMP contains
and extent of such modifications depends upon several parameters ~26–27% (w/w) fat whereas SMP contains <1.0% (w/w) fat [15]. Three
such as the composition of milk and the storage and processing condi- principal categories of SMP are distinguished: high-, medium- and
tions [5,6]. The rapid removal of water during spray drying results in low-heat SMP, depending on the undenatured whey protein content
powders with an amorphous glassy matrix [7], which tends to convert per grams of powder, known as whey protein‑nitrogen index (WPNI).
to a rubbery state when stored at high temperature and relative humid- SMPs with a WPNI value ≥6.0 mg are referred to as low-heat (LH).
ity (RH) [8,9]. Powders in the glassy state can support their own weight SMPs with WPNI values between 1.5 and 6.0 mg are identified as
due to the high viscosity (1012 Pa.s) while those in the rubbery phase medium-heat (MH), and SMPs with WPNI value ≤1.5 mg are defined
display lower viscosity (106–108 Pa.s) [10] and greater molecular mobil- as high-heat (HH). This classification system of milk powders is used
ity, which can lead to undesirable phenomena such as lactose crystalli- commercially in various food applications [16]. Spray-dried SMPs can
zation, Maillard reaction, and caking [11,12]. The existence of free fat on be produced from milk heated at high (~120 °C for 240 s), medium
surface of the powders and lactose crystallization play significant roles (~90 °C for 30 s) and low (~80 for <5 s) temperature conditions [17].
in caking [13]. Rapid moisture removal during the spray-drying process These processing conditions substantially affect the physicochemical
characteristics of milk powders. Thus, the storage condition for these
powders should be designed in such a way that the fresh-like physico-
⁎ Corresponding author. chemical properties of the milk powders are maintained as long as
E-mail address: benu.adhikari@rmit.edu.au (B. Adhikari). possible.

https://doi.org/10.1016/j.powtec.2020.06.020
0032-5910/© 2020 Elsevier B.V. All rights reserved.
A. Phosanam et al. / Powder Technology 372 (2020) 394–403 395

Despite continued research on milk powders, no reported study has probe speed with a 15 mm compression distance. The peak force was
compared the physicochemical properties of SMP based on heat class used as the cake strength index.
(high-, medium- and low-heat) and WMP differentiated by the milk
fat content when stored at very low and high RH conditions at mild tem- 2.5. Measurement of lactose crystallinity
perature. Thus, this study aimed to study effects of storage at three RHs
(11%, 44% and 85%) and three temperatures (25 °C, 35 °C and 45 °C) car- Lactose crystallinity refers to the percentage of crystallized lactose in
ried out for 21 days on the moisture sorption isotherms, Tg, caking the entire (bulk) milk powder. X-Ray diffraction (XRD) using a wide-
strength, lactose crystallization, microstructure and the browning of angle X-ray diffractometer (Bruker Axs D4 Endeavor Wide-angle XRD,
low-, medium- and high-heat SMP and WMP. Germany) with Co-Kα radiation (40 kV, 35 mA) was used to determine
th lactose crystallinity (%) of the samples. About 1.5 g sample was placed
2. Materials and methods in sample holder and was scanned at 10–30° (2-theta) with increments
of 0.02°. The Diffrac.Eva.V4.3 software (Bruker) was used to analyze the
2.1. Materials spectra and to identify the lactose peaks by matching the spectra of lac-
tose available in the chemical database.
LH-, MH- and HH-SMP and WMP were obtained from Saputo
Australia (Melbourne Victoria 3001, Australia). The declared protein 2.6. Analysis of surface morphology
and the fat contents for SMPs were 33% (w/w) and 1% (w/w), respec-
tively, whereas WMP contained 26% (w/w) fat and 24% (w/w) protein. To analyze the morphology of powder particles, milk powders were
Potassium carbonate (K2CO3), potassium acetate (CH3COOK), lith- transferred to adhesive carbon tabs mounted on SEM stubs and gold
ium chloride (LiCl), sodium chloride (NaCl), potassium iodide (KI), po- coated for 1 min under vacuum with a sputter coater (SPI-Module™
tassium chloride (KCl) and magnesium nitrate (Mg (NO3)2) were Sputter Coater, West Chester, USA). Gold coated samples were analyzed
provided by Sigma Aldrich Pty Ltd. (Castle Hill, NSW, Australia). All with a scanning electron microscope (SEM, FEI Quanta 200,
chemicals are analytical-grade chemicals and were used without any Netherlands) functioning in high vacuum mode.
further modification/purification.
2.7. Measurement of colour parameters
2.2. Determination of sorption isotherms
The colour parameter of powders was measured in CIE L ∗ and b∗ col-
our space including lightness (L*, L* = 0, black; L* = 100, white) and
Approximately 2 g milk powders were placed in a pre-weighed plas-
yellowness (b∗, positive values; blueness b∗, negative values). The CR-
tic cup and this cup was kept in desiccator containing saturated salt so-
400 chroma meter (Konica Minolta Sensing, Inc., Tokyo, Japan) cali-
lutions of LiCl, CH3COOK, K2CO3, Mg (NO3)2, KI, NaCl and KCl with
brated with standard plate (No. 12133027) was used to determine
respective RHs of 11, 23, 44, 53, 69, 75 and 85%. The desiccator was
these parameters. Illuminant and observer conditions were D65 and
placed in an incubator pre-maintained at 25, 35 and 45 °C for 21 days
2°, respectively.
to ensure that samples were equilibrated to headspace RH. The powder
samples were then analyzed for the moisture content with an infrared
2.8. Measurement of surface free fat
moisture analyzer (Ohaus Moisture Analyzer MB45). An equilibrium
moisture of powder samples for every temperature is plotted against
A gram milk powder was placed in a sheet of filter paper (Whatman
water activity (= %RH/100) for producing sorption isotherms. The
no. 4) pre-folded in a glass funnel and then the sample was washed five
Guggenheim–Anderson–de Boer (GAB) model was applied to fit exper-
times with n-hexane at time interval of 10–15 s. The filtrate containing
imental data through regression analysis to determine monolayer mois-
extracted fat was evaporated on a heating block at 60 °C in a fume hood
ture content (Mo; g water/100 g solid) [18]. The equilibrated samples
until a constant weight of extracted fat residue was achieved. The
were used to carry out all the tests described below. The Tg of the sam-
surface-free fat was calculated by Eq. (1)
ples was measured immediately (21st day).The remaining samples
were sealed in aluminum foil bags and stored at −18 °C before further B−A
analysis. Surface free fat ð%Þ ¼ x 100 ð1Þ
Weight of sample ðgÞ

2.3. Measurement of glass transition temperature where, B is the weight of sample and Petri dish (g) after drying, and A is
the weight of empty Petri dish (g).
Five to ten milligrams of powder was weighed into DSC pans (Tzero-
T180823) and sealed using a DSC's sample press (Tzero-T080723). An 2.9. Visualization of protein and fat distribution on surface of powders
empty pan was used as a control/reference. The thermal analysis of
samples was conducted in four steps: (1) equilibration at −70 °C for 1 Initially, about 1 mg of powder was placed on a glass-slide and ex-
min; (2) heating from −70 to 150 °C at 5 °C /min; (3) cooling from cess was removed by tapping. SMP and WMP were stained with one
150 to −70 °C at 10 °C /min; and (4) heating from −70 to 150 °C at 5 drop of equally mixed dye containing 0.1%(w/w) Fast Green and 0.1%
°C /min. Tg was determined as the midpoint of transition from the sec- (w/w) Nile Red dissolved in polyethylene glycol 400. A coverslip was
ond heating thermogram using Universal Analysis 2000 software (TA used to protect the final mixture before viewing with a CLSM (Nikon
Instruments, USA). ECLIPSE 90i digital microscope, Japan). A 60× water immersion objec-
tive lens was used to determine the distribution of fat and protein on
2.4. Determining caking strength the surface of powders dual-excitation at 488/633 nm following the
method of Tham et al. [19] with minor modifications.
Milk powder samples weighing 20 g were placed in plastic cups
(40.6 mm height × 33.0 mm diameter) then exposed to the tempera- 2.10. Statistical analysis
ture and RH settings described in section 2.2. The strength of cake was
analyzed by texture analyzer (TA-XT2 Plus, Stable Micro Systems Ltd., IBM's statistical software (SPSS®, version 25, IBM Corp., Melbourne,
UK) equipped with a 5 mm diameter stainless steel cylindrical probe Australia) was used to statistically analyze results. Tests were conducted
and a 284.2 N load cell. The compression was conducted at 2 mm/s in triplicate unless otherwise specified and results were reported as
396 A. Phosanam et al. / Powder Technology 372 (2020) 394–403

mean ± standard deviation. To find significant difference in samples,


one-way analysis of variance (ANOVA) was conducted. Significant
groups were identified by post-hoc comparison tests (Duncan's Multi-
ple Range Test, DMRT) at 95% confidence level (P < .05).

3. Results and discussion

3.1. Moisture sorption isotherms

The moisture sorption isotherms of LH-, MH- and HH-SMP and WMP
at tested temperatures (25, 35, and 45 °C) are shown in Fig. 1. WMP
showed low equilibrium moisture content compared to SMP at all tem-
perature and relative humidity combinations. Since fat is not a water-
absorbing constituent, the samples containing higher percentage of fat
absorbed less moisture upon equilibration [20]. However, the equilib-
rium moisture content (EMC) on solids non-fat (SNF) basis of WMP
was higher than on total solid (TS) basis and higher in WMP than in
SMPs at the higher RH studied (>53%RH). Jouppila and Roos [21] stated
that water content after 24 h (SNF basis) was higher in milk powders
containing fat than in skim milk powder at the highest RH studied
(76.4%). This higher water content correlated with the delayed lactose
crystallization as the amount of adsorbed water released by the pow-
ders containing fat decreased at the beginning of storage. The sorption
isotherms showed that the EMC increases by increasing aw at a given
temperature. This is due to an increase in vapour pressure in the head-
space with increasing relative humidity when the temperature is con-
stant [22]. Also, the EMC increased linearly at aw values up to 0.44 and
then increased non-linearly above this aw for both SMPs and WMPs.
This could be due to the presence of characteristic multilayer sorption
and capillary condensation regions in SMP and WMP as they contain
amorphous lactose and protein. At low and intermediate aw values
(multilayer sorption region), equilibrium moisture content rises line-
arly with aw, whereas with higher aw values (capillary-condensation re-
gion) a rapid improvement in equilibrium moisture content was
experienced [23].
According to McMinn and Magee [24], a decrease of EMC at higher
temperatures is attributed to higher activation of water molecules
which results into freeing water molecules from binding sites and ulti-
mately decreasing the moisture content at equilibrium. Temperature af-
fects mobility of water molecules thereby affecting the dynamic
equilibrium between adsorbents phase and vapour phase. In food sys-
tems, at a given aw, higher temperature decreases equilibrium moisture
content [25]. The monolayer moisture content (Mo) in these milk pow-
ders varied from 3.0 to 4.0 g water/100 g solid. The results showed that
the Mo of SMPs and WMP decreased by increasing temperature(Table
1). Lower Mo at higher temperatures is ascribed by reducing number
of binding sites because of chemical and physical changes instigated
by temperatures [24].

Table 1
Monolayer moisture content (Mo, g water/100 g total solid) of of LH SMP, MH SMP, HH
SMP and WMP stored at 25 °C, 35 °C and 45 °C. Mo was determined by fitting GAB
equation.

Samples Temperature (°C)

25 35 45

LH SMP 3.5 3.2 3.0


Fig. 1. Equilibrium moisture content (EMC) of low heat (LH), medium heat (MH), and high MH SMP 3.9 4.0 3.0
heat (HH) skim milk powder (SMP) and whole milk powder (WMP) stored at 25 °C, 35 °C HH SMP 3.2 3.1 3.0
and 45 °C. The solid lines are best fit lines produced by fitting GAB equation to WMP 3.2 3.7 3.0
experimental data.
A. Phosanam et al. / Powder Technology 372 (2020) 394–403 397

Table 2
Glass transition temperature (Tg) of LH SMP, MH SMP, HH SMP and WMP stored at 25 °C, 35 °C and 45 °C and equilibrated at 11% RH, 44% RH and 85% RH.

Samples Tg (midpoint), oC

LH SMP MH SMP HH SMP WMP

Control 55.0cB ± 0.0 58.8dB ± 1.6 54.5eB ± 2.3 49.0dA ± 0.6


25C 11RH 58.7cC ± 0.7 59.9cC ± 0.0 57.1eB ± 0.6 54.0eA ± 0.5
25C 44RH 13.3bB ± 1.26 13.4bB ± 0.0 15.7dB ± 1.4 9.3eA ± 0.3
25C 85RH −34.5aNS ± 4.12 −32.5aNS ± 0.0 −33.8aNS ± 2.1 −32.1bNS ± 2.0
35C 11RH 68.3dB ± 0.7 67.2dAB ± 0.0 65.9fAB ± 0.5 65.2fA ± 0.9
35C 44RH 13.3bB ± 0.2 11.1bA ± 0.0 10.6cA ± 0.6 10.8cA ± 0.4
35C 85RH −34.8aAB ± 0.4 −31.9aC ± 0.0 −33.8aB ± 0.2 −35.2aA ± 0.9
45C 11RH 70.2dC ± 0.4 68.8dB ± 0.1 68.7fB ± 0.3 66.3fA ± 0.2
45C 44RH 9.5bA ± 0.7 12.2bA ± 2.5 12.1cA ± 2.5 9.9cA ± 0.4
45C 85RH −32.1aB ± 3.6 −32.9aB ± 0.4 −28.4dA ± 0.0 −31.2bA ± 1.4

Values are presented as mean ± standard deviation (n = 2);


Values in a column with different lowercase superscript differ significantly (p < .05). Values in a row for a given parameter with different uppercase superscript
differ significantly (p < .05).

3.2. Effect of temperature and storage RH on Tg ingredients [31]. Tg of carbohydrates and proteins declines with an in-
creasing water content which acts as plasticizer and consequently in-
Tg is a critical parameter in assessing stability of stored dairy pow- creasing molecular mobility, which initiates the glass transition
ders as Tg process is closely related to their stability [26–29]. It is re- process at a lower temperature [11,32]. The Tg of various milk powders
ported that the lactose content dominates the glass transition was expressed as a function of SNF, a close similarity was observed to
behavior in milk powders [30]. Thus, we presumed that the Tg measured values for amorphous lactose [33].
in milk powder samples indicated Tg of lactose. Tg is a significant caking
indicator in milk powder with higher Tg indicating better stability and 3.3. Effect of the temperature and storage RH on crystallinity
delayed lactose crystallization [19]. Tg values for low-, medium-, high-
heat SMPs and WMP equilibrated at different temperatures RH values Lactose in spray-dried SMP and WMP is converted to the amorphous
are provided in Table 2. The Tg for all powder samples decreased with glassy state because of rapid drying. The glassy form of lactose is ther-
increasing RH irrespective of temperature indicating increased moisture modynamically unstable and highly hygroscopic [34]. A higher mobility
uptake and plasticizing effect of absorbed water. Tg values of WMPs of lactose molecules at high moisture content and temperature supports
were lower than that of SMPs at any given storage temperature and RH. its crystallization. The effect of temperature and RH on crystallinity of
Tg is considered an essential characteristic of milk powders and is LH-, MH-, and HH-SMP and WMP are shown in Fig. 2. The X-ray
highly dependent upon RH and storage temperature. Results show diffractograms of control powder had no characteristic peaks (Fig. 2 A
that Tg of the samples decreased when the headspace RH increased, and B) indicating an amorphous glassy state. Crystallinity was negligible
which is consistent with the reported effect of RH on the Tg of dairy in all powders stored at 11% RH, irrespective of the storage temperature.

Fig. 2. Diffraction patterns of control medium-heat (MH) SMP (A), control full cream milk powder (WMP) (B), MH SMP stored at 45 °C and 85% RH (C), and WMP stored at 45 °C and 85%
RH (D).
398 A. Phosanam et al. / Powder Technology 372 (2020) 394–403

caused by the increase in sorbed moisture and/or temperature. Thus,


caking of amorphous powder particles depends strongly on the sur-
rounding temperature and RH [32]. The behavior of amorphous and
crystalline particles is different when they are exposed to moist air.
When exposed to higher RH, the amorphous particles readily bridge/
stick/sinter with each other due to their hygroscopic nature while the
crystalline particles are less sensitive to variation in RH [41,42]. Caking
of amorphous solid particles can be explained by sintering mechanism
in which molecular mass from neighboring particles diffuses/transfers
towards each other to fill the void/gap between them. Sintering in
amorphous solids can only occur when the surrounding temperature
becomes higher than their Tg because molecules acquire higher ability
to flow. The sintering mechanism also explains that the free specific sur-
face energy of powder systems decreases when the gap/void between
two neighboring particle decreases; thus, the sintering or caking pro-
Fig. 3. Powder crystallinity of low-heat (LH) SMP, medium-heat (MH) SMP, high-heat
cess is thermodynamically favored. The rate formation of sinter bridges
(HH) SMP and full cream milk powder (WMP) stored at 25 °C, 35 °C and 45 °C and 11%
RH, 44% RH and 85% RH. Columns within the same temperature with different (tendency to cake) depends on the molecular mobility. The molecular
superscript differ significantly (P˂0.05). mobility of rubbery solids depends on their viscosity, which in itself de-
pends on the temperature difference between the powder and their Tg,
i.e. (T-Tg) [43,44]. Thus, the crystallization of lactose in milk powders oc-
Crystallinity was not detected in any of the powders stored at 44% RH curs when (T-Tg) is positive. The crystallization of lactose within sinter
and 25 °C. However, the crystallinity of all the powders increased signif- bridges has potential to alter the caking dynamic which occurs when
icantly at 85% RH when stored at 25 °C (Fig. 3). All powders stored at 35 amorphous lactose becomes rubbery. It is expected that when the lac-
°C and 45 °C at 44% and 85% RH showed significant (p < .05) crystallin- tose crystallizes in the sinter bridges, it will affect the caking strength
ity. The crystallinity in powders increased further when stored at 85% [43,44].
RH than 44% RH irrespective of temperature (Fig. 2 C, D and Fig. 3). The other factor that affects the moisture adsorption and subsequent
The rate of crystallization of amorphous sugars depends on moisture crystallization and caking of milk powders is their surface composition.
content and storage temperatures above Tg (i.e., T-Tg) [20]. Interest- It has been reported that up to 91.8% of the surface of whole milk pow-
ingly, WMP showed lower crystallinity compared to SMPs at 44% and der can be covered by milk fat [20]. As the surface of WMP is expected to
85% RH at 35 °C and 45 °C, which was likely due to lower lactose content be mostly covered by fat, these powders are expected to have lower lac-
in WMP. These observations agree with that of Jouppila and Roos [36], tose crystallinity and caking strength compared to SMPs due to the for-
who defined that the presence of milk fat reduced the rate of lactose mation of fewer lactose (solid) bridges due to higher surface fat.
crystallization. It is also consistently shown that milk fat (or plant oil)
overrepresents the surface composition of spray dried powders [37]. 3.5. Effect of temperature and storage RH on powder microstructure
Thus, lipid-rich surface layer of WMP is expected to dissuade the ad-
sorption of water molecules. Lactose crystallization and caking are in- SEM is a technique which is used to assess in the microstructure of
terrelated [12]. powder samples. The microstructure of medium-heat SMPs stored at
25 °C, 35 °C and 45 °C at 11%, 44% and 85% RH are shown in Fig. 5. The
particle of medium-heat SMP stored at 25 °C and 11% RH (Fig. 5, A-1)
3.4. Effect of temperaure and storage RH on cake strength appeared to have an irregular surface with both shallow and deep
folds and an absence of sharp crystals. MH-SMPs stored at 25 °C and
Spray dried milk powders are unstable because of their amorphous 44% RH (Fig. 5, A-2) showed agglomeration of rough-surfaced particles
structure and glassy state and have a tendency of sorbing water and cre-
ating caked powder which occurs when they are exposed to high RH
and temprature [35]. The effects of RH and temperature on caking
strength of powders is represented in Fig. 4. Caking was not observed
in any of powders stored at 11% RH at all storage temperatures; how-
ever, strong caking was observed in SMPs stored at 85% RH and 25 °C,
44% RH and 85% RH at 35 °C and 45 °C. WMP had a lower caking
strength compared to SMPs at 44% RH and 85% RH at all three temper-
atures, which can be attributed to their high fat content and a low lac-
tose crystallinity [20,36]. Caking occurs when amorphous glassy
powders transition to rubbery state at above their Tg, which greatly in-
creases the mobility of water and facilitates the formation of inter-
particle liquid and solid bridges. Humidity-induced caking is a common
mechanism for caking which occurs when the increased moisture con-
tent depresses the Tg below the storage temperature [37,38]. Powder
particles containing low molecular weight carbohydrates such as lac-
tose become sticky when their temperature is 10–20 °C their Tg,
known as sticky-point temperature [39,40]. Increased molecular mobil-
ity on the surface of rubbery solid particles favors cohesive interactions
among each other making powders cohesive. The cohesive interaction/
force of sticky particles increases initially as a response to increase in
Fig. 4. Caking strength of low-heat (LH), medium-heat (MH), high-heat (HH) SMPs and
temperature and/or moisture content. Thus, the strength of cohesive whole milk powder (WMP) stored at 25 °C, 35 °C and 45 °C and 11% RH, 44% RH and
stickiness, measured in this work as caking strength, initially rises to a 85% RH. Bars in the same temperature with different superscript letters differ
maximum level and then decreases due to continuous drop in viscosity significantly (P˂0.05).
A. Phosanam et al. / Powder Technology 372 (2020) 394–403 399

Fig. 5. SEM micrographs of MH SMP: (1), (2), (3) powders stored at 11%, 44%, 85% RH, respectively at 25 °C (A), 35 °C (B) and 45 °C (C) compared with control (N). Scale bar = 50 μm.

Fig. 6. SEM micrographs of WMP: (1), (2), (3) powders stored at 11%, 44%, 85% RH, respectively at 25 °C (A), 35 °C (B) and 45 °C (C) compared with control (N). Scale bar = 50 μm.
400 A. Phosanam et al. / Powder Technology 372 (2020) 394–403

and an absence of lactose crystals. In contrast, large numbers of needle- powders. As expected, the lightness values of control SMPs were slightly
like lactose crystals were found on surface of these SMPs when stored at higher than that of control WMP which were yellowish due to the
25 °C and 85% RH (Fig. 5, A-3), and 45 °C at all tested RHs (Fig. 5, C-1, C-2 higher fat content. Milk fat contains a fat-soluble beta carotene which
and C-3). imparts the yellow colour. After 21 days of storage, both the SMP and
SEM micrographs of WMP stored at 25 °C and 45 °C are shown in Fig. WMP powders developed brown colour, and this was especially pro-
6. WMP stored at 25 °C and 45 °C and 11% and 44% RH (Figs. 6, 25-A, B nounced at 85% RH at all storage temperatures. The formation of
and 45-A, B) showed a regular and smooth surface. The smooth surface brown colour is associated with Maillard reaction, which occurs be-
and absence of sharp crystals indicated that lactose remained in the tween a protein containing lysine and reducing sugar such as lactose
amorphous state. The powder particles were also clustered and agglom- [48]. As shown in Fig. 8, the lightness values of SMP and WMP powders
erated into noticeably larger structures (Fig. 6) compared to SMPs (Fig. were lower than that of control powders. On the other hand, lightness
5). Storage of WMP at 25 °C and 85% RH (Figs. 6, 25-C) formed promi- value for both SMP and WMP were lowest in samples stored at 85%
nent clusters of agglomerated particles with increasingly irregular and RH. The yellowness value is useful for monitoring the browning reac-
rougher surface indicating that crystallization of lactose has occurred. tions in milk powders [49]. Control and powders stored at 25 °C and
The agglomerated clusters were more pronounced at 45 °C and 85% 11% RH had the lowest yellowness. On the other hand, SMPs and
RH (Figs. 6, 45-C). WMP stored at 45 °C and 85% RH had the highest yellowness.
WMP had relatively smooth and uniformly-sized, and agglomerated Yellowness is the indicator of non-enzymatic browning and it further
powder particles compared to SMPs due to the high fat coverage on the increased at higher RHs (Fig. 8) due to the late-stage Maillard reaction
surface [45–47]. A high proportion of fat on the surface makes WMP during storage. The Maillard reaction between lactose and milk proteins
particles weakly sticky and facilitated their clustering through fat bind- produces lactulosyl-lysine conjugates. The extent of Maillard reaction
ing. The storage of high fat containing powder at high temperature and/ induced browning, as indicated by the change in yellowness values,
or in a humid atmosphere resulted in release of oil from the inner par- confirmed that it increases with temperature and RH of the storage
ticle to the surface, which increased surface fat coverage. It has been [50]. It has been reported that at a given RH, the browning rate increases
shown that the particle structure is altered when amorphous lactose with the increase in storage temperature [51]. Our results showed that
undertakes crystallization [20]. When this occurs, fat is excluded from powders stored at 85% RH containing a higher proportion of crystallized
the ordered structure of lactose crystals and expelled to the powder sur- lactose (Fig. 3) resulted into higher degree of Maillard reaction induced
face [12,19]. Therefore, an increase in free fat on the powder surface oc- browning as compared to those containing lower proportion of crystal-
curs as a result of lactose crystallization. lized lactose (e.g., control powders, SMP and WMP stored at 11%RH). In-
creased crystallization of lactose increased the rate of Maillard reaction
3.6. Effect of temperature and storage RH on the colour of powders induced browning due to release of water expelled during the crystalli-
zation process [52]. The higher free moisture content made the milk
Colour is an essential parameter that influences consumer accept- proteins and lactose more mobile and available for Maillard reaction.
ability. The lightness and yellowness of SMP and WMP powders stored Higher temperatures and RHs are essential preconditions that induce
under the tested conditions are presented in Figs. 7 and 8, respectively. Maillard reactions and browning in WMP and SMP. Fat in WMC can
The control milk powders were lighter and less yellow than the stored slow down the Maillard reaction by lowering the propensity to absorb

Fig. 7. Lightness (L*) of low-heat (LH), medium-heat (MH), high-heat (HH) SMPs and whole milk powder (WMP) stored at 25 °C, 35 °C and 45 °C and 11% RH, 44% RH and 85% RH. Columns
within the same sample having different superscript differ significantly (P˂0.05).
A. Phosanam et al. / Powder Technology 372 (2020) 394–403 401

Fig. 8. The yellowness (b*) of low-heat (LH), medium-heat (MH), high-heat (HH) SMPs and full cream milk powder (WMP) stored at 25 °C, 35 °C and 45 °C and 11% RH, 44% RH and 85%
RH. Columns within the same sample having different superscript differ significantly (P˂0.05).

moisture. Also, high fat content on particles' surface increases surface 3.8. Effect of temperature and storage RH on fat and protein distribution on
hydrophobicity, thus decreasing wettability of particle and limiting surface
crystallization of lactose [21].
CLSM is used to visualize the microstructure of food including
powders. The effect of storage parameters on surface composition
3.7. Effect of temperature and storage RH on surface free fat of SMP and WMP in terms of fat and protein distribution is investi-
gated using CLSM (Fig. 9). The control MH SMP powders exhibited a
Free fat in powders is the fat that is not entirely covered by protec- continuous protein phase containing a small number of very fine fat
tive matrix. It is an indicator of damage to membrane of the milk fat droplets (Fig. 9A). Similar observation was found for MH SMP pow-
globules [53]. The free fat can alter the desirable properties of dairy ders stored at 11% RH of 25 °C and 45 °C (Fig. 4B and D). The control
powder such as oxidative stability and flowability [54]. The free fat ex- and stored WMP at 25 °C and 11%RH showed a continuous protein
tracted from MH-SMP and WMP at differences storage treatments is phase containing a large number of fat droplets (Fig. 9, F and G).
presented in Table 3. The free fat of control and stored MH-SMP and The lipid components initially embedded inside particles tended
WMP at 25 °C and 11% RH (the most stable powders), and 45 °C and to migrate towards surface when stored at high temperature and
85% RH (the most unstable powders) was similar except in WMP stored high RH as observed in MH-SMP stored at 85% RH with 25 °C and
at 45 °C and 85%RH which increased significantly (P < .05). In powders 45 °C (Fig. 9C and E) and WMP stored at 85%RH of 25 °C (Fig. 9H
containing amorphous lactose, milk fat is encapsulated in a matrix com- and J) and 45 °C and 11%RH and 45 °C (Fig. 9I). A higher fat content
prised of amorphous lactose and protein. Crystallization of lactose re- observed on surface of particle after storage represents the aging-
leases encapsulated fat by damaging the membrane of the milk fat induced fat migration. Higher intensity of red colour on particle sur-
globule and also rupturing the protective matrix [55]. The free fat of face shows an increased fat coverage. This coverage of particle sur-
control and stored MH-SMP did not increase significantly (P < .05) as face by fat resulted in particle bridging (agglomeration) which is
shown in Table 3, possibly due to the very low-fat content in MH- responsible for the fat-induced caking [57]. Such increased fat local-
SMP (1%). ization on surface of SMP and WMP is also aided by lactose
crystallization.

Table 3 4. Conclusion
Surface free fat (%) of medium-heat SMP and WMP stored at 25 °C and 45 °C
and equilibrated at 11% RH 85% relative humidity.
The fat content in milk powders affects their physicochemical is con-
Treatments Samples sidered during storage. This study quantified and explained the effect of
MH SMP WMP fat content on the properties of whole milk powder (WMP, 26% fat) and
Control 0.2 ±0.0 a
1.2 ±0.1a skim milk powders (SMPs, 1% fat) stored at three temperatures (25 °C,
25C 11RH 0.4 ±0.1a 1.6 ±0.2a 35 °C and 45 °C) and RHs (11%, 44% and 85%) for 21 days. Moisture sorp-
45C 85 RH 0.7 ±0.1a 16.5 ±1.0b tion isotherms, changes in Tg, crystallinity, surface fat content and cak-
Values are mean ± standard deviation; n = 3. Values of the surface free fat ing strength were measured. SMPs and WMP stored at 45 °C
with different superscript are significantly different (P < .05). equilibrated at the lowest moisture content (EMC) compared to those
402 A. Phosanam et al. / Powder Technology 372 (2020) 394–403

Fig. 9. Confocal laser microscopic images showing fat and protein distribution on the surface of MH SMP (A-E) and WMP (F-J) stored at 25 °C and 45 °C and 11% RH and 85% RH compared
with control powder at magnification 60×. (scale bar = 10 μm). Labelled by Nile Red (red) = fat and labelled by Fast Green (green) = protein.

stored at 25 °C and 35 °C. WMP had lower EMC than SMPs at all tested [7] S. Miao, Y.H. Roos, Isothermal study of nonenzymatic browning kinetics in spray-
dried and freeze- dried systems at different relative vapor pressure environments,
temperatures and RHs. Tg of both WMP and SMP decreased with in- Innov. Food Sci. Emerg. Technol. 7 (2006) 182–194, https://doi.org/10.1016/j.ifset.
creasing storage RH. WMP had lower Tg and lower crystallinity than 2005.11.001.
SMPs in all cases due to lower lactose content in the former. The caking [8] P. Boonyai, B. Bhandari, T. Howes, Stickiness measurement techniques for food pow-
ders: a review, Powder Technol. 145 (2004) 34–46, https://doi.org/10.1016/j.
strength of SMPs were higher than that of WMP and SMP stored at low
powtec.2004.04.039.
RH (11%) showed particles with a regular-shaped surface, while those [9] E. Fernández, C. Schebor, J. Chirife, Glass transition temperature of regular and lac-
stored at high temperature or/and RHs (85% RH at 25 °C and 44% RH tose hydrolyzed milk powders, LWT-Food Sci Technol. 36 (2003) 547–551,
and 85% RH at 35 °C and 45 °C) had particles with higher surface rough- https://doi.org/10.1016/s0023-6438(03)00022-7.
[10] V. Truong, Optimization of spray drying of sugar-rich foods, in: F. Erdogdu (Ed.),
ness and irregularity due to lactose crystallization. Maillard reaction was
Optimization in Food Engineering, CRC Press 2008, pp. 452–487, https://doi.org/
more prominent in both SMP and WMP at higher storage RH and tem- 10.1201/9781420061420.
perature. Ambient temperature (25 °C) and lower RH (11% and 44%) [11] Y.H. Roos, Importance of glass transition and water activity to spray drying and sta-
were found to be suitable for storage of both WMP and SMP. The heat bility of dairy powders, Lait. 82 (2002) 475–484, https://doi.org/10.1051/lait:
2002025.
class (LH-, MH- and HH-heat) did not have significant influence on
[12] M.E. Thomas, Milk powders ageing: effect on physical and functional properties,
the physicochemical and microstructural properties of SMPs as lactose Food Sci Nutr. 44 (2004) 297–322, https://doi.org/10.1080/10408690490464041.
and fat contents determined these properties. [13] T.W.Y. Tham, Moisture sorption isotherm and caking properties of infant formulas, J.
Food Eng. 175 (2016) 117–126, https://doi.org/10.1016/j.jfoodeng.2015.12.014.
[14] N.A. Mccarthy, Effect of protein content on the physical stability and microstructure
Declaration of Competing Interest of a model infant formula, Int. Dairy J. 29 (2013) 53–59, https://doi.org/10.1016/j.
idairyj.2012.10.004.
The authors declare that they have no known competing financial [15] A. Pugliese, G. Cabassi, E. Chiavaro, Physical characterization of whole and skim
dried milk powders, J. Food Sci. Technol. 54 (2017) 3433–3442, https://doi.org/10.
interests or personal relationships that could have appeared to influ-
1007/s13197-017-2795-1.
ence the work reported in this paper. [16] V. Sikand, P.S. Tong, J. Walker, Impact of protein standardization of milk powder
with lactose or permeate on whey protein nitrogen index and heat classification,
Acknowledgement Dairy Sci. Technol. 88 (2008) 105–120, https://doi.org/10.1051/dst:2007011.
[17] G.J.O. Martin, R.P.W. Williams, D.E. Dunstan, Comparison of casein micelles in raw
and reconstituted skim milk, J. Dairy Sci. 90 (2007) 4543–4551, https://doi.org/10.
The first author acknowledges the scholarship support from The 3168/jds.2007-0166.
Royal Thai Government (Ministry of Science and Technology of Thai- [18] K. Muzaffar, P. Kumar, Moisture sorption isotherms and storage study of spray dried
land). The authors wish to thank Saputo Dairy Australia for providing tamarind pulp powder, Powder Technol. 219 (2016) 322–327, https://doi.org/10.
1016/j.powtec.2015.12.046.
powder samples.
[19] T.W.Y. Tham, X. Xu, A.T.H. Yeoh, W. Zhou, Investigation of caking by fat bridging in
aged infant formula, Food Chem. 218 (2017) 30–39, https://doi.org/10.1016/j.
References foodchem.2016.09.043.
[20] I. Murrieta-Pazos, C. Gaiani, L. Galet, B. Cuq, S. Desobry, J. Scher, Comparative study
[1] R. Hedegaard, L. Skibsted, Shelf-life of food powders, Handbook of Food Powders, of particle structure evolution during water sorption: skim and whole milk pow-
Woodhead Publishing 2013, pp. 409–434, https://doi.org/10.1533/ ders, Colloid Surf. B. 87 (2011) 1–10, https://doi.org/10.1016/j.colsurfb.2011.05.001.
9780857098672.2.409. [21] K. Jouppila, Y.H. Roos, Water sorption and time-dependent phenomena of milk
[2] E.H.-J. Kim, X.D. Chen, D. Pearce, Surface composition of industrial spray-dried milk powders, J. Dairy Sci. 77 (1994) 1798–1808, https://doi.org/10.3168/jds.s0022-
powders. 2. Effects of spray drying conditions on the surface composition, J. Food 0302(94)77121-6.
Eng. 94 (2) (2009) 169–181, https://doi.org/10.1016/j.jfoodeng.2008.10.020. [22] A. Al-Muhtaseb, W. Mcminn, T. Magee, Water sorption isotherms of starch powders:
[3] A. Sharma, A.H. Jana, R.S. Chavan, Functionality of milk powders and milk-based part 1: mathematical description of experimental data, J. Food Eng. 61 (2004)
powders for end use applications-a review, Compr. Rev. Food Sci. Food Saf. 11 297–307, https://doi.org/10.1016/S0260-8774(03)00133-X.
(2012) 518–528, https://doi.org/10.1111/j.1541-4337.2012.00199.x. [23] N.D. Mrad, C. Bonazzi, N. Boudhrioua, N. Kechaou, F. Courtois, Influence of sugar
[4] H.C. Deeth, J. Hartanto, in: A.Y. Tammime (Ed.), Chemistry of Milk - Role of Constit- composition on water sorption isotherms and on glass transition in apricots, J.
uents in Evaporation and Drying, Blackwell Publishing Ltd, Ayr, UK 2009, pp. 1–27, Food Eng. 111 (2012) 403–411, https://doi.org/10.1016/j.jfoodeng.2012.02.001.
https://doi.org/10.1002/9781444322729.ch1 , Dairy Powders and Concentrated [24] W.A.M. McMinn, T.R.A. Magee, Thermodynamic properties of moisture sorption of
Products. potato, J. Food Eng. 60 (2003) 157–165, https://doi.org/10.1016/s0260-8774(03)
[5] F. Fenaille, E. Campos-Giménez, P.A. Guy, C. Schmitt, F. Morgan, Monitoring of β-lac- 00036-0.
toglobulin dry-state glycation using various analytical techniques, Anal. Biochem. [25] J. Štencl, Water activity of skimmed milk powder in the temperature range of 20-45
320 (2003) 144–148, https://doi.org/10.1016/s0003-2697(03)00357-9. °C, Acta Vet. Brno 68 (1999) 209–215.
[6] M.K. Haque, Y.H. Roos, Water plasticization and crystallization of lactose in spray- [26] Q. Rao, T.P. Labuza, Effect of moisture content on selected physicochemical proper-
dried lactose/protein mixtures, J. Food Sci. 69 (2004) FEP23–FEP29, https://doi. ties of two commercial hen egg white powders, Food Chem. 132 (2012) 373–384,
org/10.1111/j.1365-2621.2004.tb17863.x. https://doi.org/10.1016/j.foodchem.2011.10.107.
A. Phosanam et al. / Powder Technology 372 (2020) 394–403 403

[27] A. Mizuno, M. Mitsuiki, M. Motoki, Glass transition temperature of casein as affected [43] J.J. Fitzpatrick, E. O’Callaghan, J. O’Flynn, Application of a novel cake strength tester
by transglutaminase, J. Food Sci. 64 (1999) 796–799, https://doi.org/10.1111/j.1365- for investigating caking of skim milk powder, Food Bioprod. Process. 86 (2008)
2621.1999.tb15914.x. 198–203, https://doi.org/10.1016/j.fbp.2007.10.009.
[28] N. Silalai, Y.H. Roos, Roles of water and solids composition in the control of glass [44] J.J. Fitzpatrick, N. Descamps, K. O’Meara, C. Jones, D. Walsh, M. Spitere, Comparing
transition and stickiness of milk powders, J. Food Sci. 75 (2010) E285–E296, the caking behaviours of skim milk powder, amorphous maltodextrin and crystal-
https://doi.org/10.1111/j.1750-3841.2010.01652.x. line common salt, Powder Technol. 204 (2010) 131–137, https://doi.org/10.1016/j.
[29] S. Yang, X.Y. Mao, F.F. Li, D. Zhang, X.J. Leng, F.Z. Ren, G.X. Teng, The improving effect powtec.2010.07.029.
of spray-drying encapsulation process on the bitter taste and stability of whey pro- [45] E.H.J. Kim, X.D. Chen, D. Pearce, Surface composition of industrial spray-dried milk
tein hydrolysate, Eur. Food Res. Technol. 235 (2012) 91–97. powders. 3. Changes in the surface composition during long-term storage, J. Food
[30] S.A. Hogan, D.J. O’Callaghan, Influence of milk proteins on the development of lac- Eng. (2009) 182–191, https://doi.org/10.1016/j.jfoodeng.2008.12.001.
tose-induced stickiness in dairy powders, J. Dairy Sci. 20 (2010) 212–221, https:// [46] C. Gaiani, M. Morand, C. Sanchez, E.A. Tehrany, P. Schuck, R. Jeantet, J. Scher, How
doi.org/10.1016/j.idairyj.2009.11.002. surface composition of high milk proteins powders is influenced by spray-drying
[31] P. Schuck, E. Blanchard, A. Dolivet, S. Méjean, E. Onillon, R. Jeantet, Water activity temperature, Colloids Surf. B 75 (2010) 377–384, https://doi.org/10.1016/j.
and glass transition in dairy ingredients, Lait 85 (2005) 295–304, https://doi.org/ colsurfb.2009.09.016.
10.1051/lait:2005020. [47] C. Gaiani, P. Schuck, J. Scher, J.J. Ehrhardt, E. Arab-Tehrany, M. Jacquot, S. Banon, Na-
[32] J.J. Fitzpatrick, M. Hodnett, M. Twomey, P.S.M. Cerqueira, J.O’. Flynn, Y.H. Roos, Glass tive phosphocaseinate powder during storage: lipids released onto the surface, J.
transition and the flowability and caking of powders containing amorphous lactose, Food Eng. 94 (2009) 130–134, https://doi.org/10.1016/j.jfoodeng.2009.01.038.
Powder Technol. 178 (2007) 119–128, https://doi.org/10.1016/j.powtec.2007.04. [48] H.M. Semagoto, D. Liu, K. Koboyatau, J. Hu, N. Lu, X. Liu, J.M. Regenstein, P. Zhou, Ef-
017. fects of UV induced photo-oxidation on the physicochemical properties of milk pro-
[33] T. Huppertz, I. Gazi, Lactose in dairy ingredients: effect on processing and storage tein concentrate, Food Res. Int. 62 (2014) 580–588, https://doi.org/10.1016/j.
stability, J. Dairy Sci. 99 (2016) 6842–6851, https://doi.org/10.3168/jds.2015-10033. foodres.2014.04.012.
[49] A. Ramirez-Jimenez, B. Garcia-Villanova, E.J. Guerra-Hernandez, Nonenzymatic
[34] A.K. Shrestha, T. Howes, B.P. Adhikari, B.R. Bhandari, Water sorption and glass tran-
Browning during storage of infant cereals, Cereal Chem. 81 (2004) 399–403,
sition properties of spray dried lactose hydrolysed skim milk powder, LWT Food Sci.
https://doi.org/10.1094/cchem.2004.81.3.399.
Technol. 40 (2007) 1593–1600, https://doi.org/10.1016/j.lwt.2006.11.003.
[50] G. Vuataz, The phase diagram of milk: a new tool for optimizing the drying process,
[35] N. Yazdanpanah, T.A. Langrish, Comparative study of deteriorative changes in the
Lait 82 (2002) 485–500, https://doi.org/10.1051/lait:2002026.
ageing of milk powder, J. Food Eng. 114 (2013) 14–21, https://doi.org/10.1016/j.
[51] M.K. Thomsen, L. Lauridsen, L.H. Skibsted, J. Risbo, Temperature effect on lactose
jfoodeng.2012.07.026.
crystallization, Maillard reactions, and lipid oxidation in whole milk powder, J.
[36] S. Palzer, The relation between material properties and supra-molecular structure of Agric. Food Chem. 53 (2005) 7082–7090, https://doi.org/10.1021/jf050862p.
water-soluble food solids, Trends Food Sci. Technol. 21 (2010) 12–25, https://doi. [52] E. Ferrer, A.A. Alegría, R. Farré, P. Abellán, F. Romero, Effects of thermal processing
org/10.1016/j.tifs.2009.08.005.
and storage on available lysine and furfural compounds contents of infant formulas,
[37] Y.H. Roos, Solid and liquid states of lactose, in: P.L.H. McSweeney, P.F. Fox (Eds.), Ad- J. Agric. Food Chem. 48 (2000) 1817–1822, https://doi.org/10.1021/jf991197l.
vanced Dairy Chemistry, Springer, New York 2009, pp. 17–33. [53] E.H.-J. Kim, X.D. Chen, D. Pearce, Surface characterization of four industrial spray-
[38] M. Carpin, H. Bertelsen, J.K. Bech, R. Jeantet, J. Risbo, P. Schuck, Caking of lactose: a dried dairy powders in relation to chemical composition, structure and wetting
critical review, Trends Food Sci. Technol. 53 (2016) 1–12, https://doi.org/10.1016/ property, Colloids Surf. B. 26 (2002) 197–212, https://doi.org/10.1016/S0927-7765
j.tifs.2016.04.002. (01)00334-4.
[39] B. Adhikari, T. Howes, D. Lecomte, B.R. Bhandari, A glass transition temperature ap- [54] M.L. Vignolles, R. Jeantet, C. Lopez, P. Schuck, Free fat, surface fat and dairy powders:
proach for the prediction of the surface stickiness of a drying droplet during spray interactions between process and product, Lait 87 (2007) 187–236, https://doi.org/
drying, Powder Technol. 149 (2005) 168–179, https://doi.org/10.1016/j.powtec. 10.1051/lait:2007010.
2004.11.007. [55] C.W. Park, M.A. Drake, The distribution of fat in dried dairy particles determines fla-
[40] Y. Roos, M. Karel, Phase transitions of mixtures of amorphous polysaccharides and vor release and flavor stability, J. Food Sci. 79 (2014) R452–R459, https://doi.org/10.
sugars, Biotechnol. Prog. 7 (1991) 49–53, https://doi.org/10.1021/bp00007a008. 1111/1750-3841.12396.
[41] S. Palzer, Influence of supramolecular structure and storage conditions on the caking [57] J. Burgain, J. Petit, J. Scher, R. Rasch, B. Bhandari, C. Gaiani, Surface chemistry and mi-
of powders, Fifth World Congress on Particle Technology, 2006. croscopy of food powders, Prog. Surf. Sci. 94 (2017) 409–429, https://doi.org/10.
[42] N. Descamps, S. Palzer, Modeling the Sintering of Water Soluble Amorphous Parti- 1016/j.progsurf.2017.07.002.
cles. PARTEC, 2007.

You might also like