effect of whole milk concentrate carbonation on functional ...380454/uq380454_oa.pdf · 20...
TRANSCRIPT
Accepted Manuscript
Effect of whole milk concentrate carbonation on functional, physicochemical andstructural properties of the resultant spray dried powder during storage
Leni Kosasih, Bhesh Bhandari, Sangeeta Prakash, Nidhi Bansal, Claire Gaiani
PII: S0260-8774(16)30037-1
DOI: 10.1016/j.jfoodeng.2016.02.005
Reference: JFOE 8473
To appear in: Journal of Food Engineering
Received Date: 24 November 2015
Revised Date: 26 January 2016
Accepted Date: 5 February 2016
Please cite this article as: Kosasih, L., Bhandari, B., Prakash, S., Bansal, N., Gaiani, C., Effect ofwhole milk concentrate carbonation on functional, physicochemical and structural properties of theresultant spray dried powder during storage, Journal of Food Engineering (2016), doi: 10.1016/j.jfoodeng.2016.02.005.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service toour customers we are providing this early version of the manuscript. The manuscript will undergocopyediting, typesetting, and review of the resulting proof before it is published in its final form. Pleasenote that during the production process errors may be discovered which could affect the content, and alllegal disclaimers that apply to the journal pertain.
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
1
Effect of whole milk concentrate carbonation on functional, physicochemical and 1
structural properties of the resultant spray dried powder during storage 2
Leni Kosasih1, Bhesh Bhandari1,*, Sangeeta Prakash1, Nidhi Bansal1 & Claire Gaiani1, 2,* 3
1 The University of Queensland, School of Agricultural and Food Science, St. Lucia, Qld. 4072, 4
Australia. 5
2 Université de Lorraine, LIBio, 2 avenue de la Forêt de Haye, TSA 40602, 54518 Vandoeuvre-6
lès-Nancy, France. 7
* Corresponding authors 8
Tél. : +33(0)3 83 59 60 73 - Fax : +33(0)3 83 59 57 72 9
[email protected] and/or [email protected] 10
11
Abstract 12
The effect of carbonation (1000 and 2000 ppm) on whole milk concentrate and the resultant 13
spray dried whole milk powder (WMP) was investigated in this research. Carbonation was 14
found to produce WMP with reduced surface fat content, dispersibility, solubility and true 15
density, and increased occluded air content. During accelerated storage at 37 °C for 18 weeks, 16
the surface coverage of fat on powder particles increased from 51, 29 and 8 to 94, 88 and 69 % 17
(WMP without treatment and treated with 1000 ppm and 2000 ppm CO2 respectively) due to 18
the release and spreading of encapsulated fat. In addition to the release of fat onto the surface, 19
microscopy observations showed the migration of free fat into the powder particle vacuoles. 20
Meanwhile, dispersibility and solubility of the powders decreased during storage for 18 weeks. 21
These results suggest that carbonation may result in powders with better shelf life due to the 22
reduced surface fat content. Improvements in functional properties was not observed, possibly 23
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
2
due to the fine size of the powders (<15 µm) that may have masked the positive effect of 24
carbonation. 25
26
Keywords 27
Whole milk powder ; storage ; CO2 treatment ; milk fat ; powder surface. 28
29
1. Introduction 30
Milk powder is produced from milk concentrate to prolong its shelf life. It may be consumed as a 31
fresh milk substitute and also used as a food ingredient. There are several types of milk powder. 32
Among them, two common types are skim milk powder (SMP) and whole milk powder (WMP). 33
The main difference between these powders is the fat content, in which SMP contains very little 34
or no fat (<1.5 % w/w), whereas WMP contains about 26 % fat (Kim et al., 2002; Murrieta Pazos 35
et al., 2012). The removal or presence of fat causes distinct functionality changes between SMP 36
and WMP. It also affects their shelf life and quality, especially during storage. 37
The presence of fat causes WMP to undergo several changes associated with quality 38
deterioration during storage. These include lactose crystallization, which involves puncturing of 39
fat globule membranes and the generation of capillary interstices network that stresses and 40
causes fat droplets disruption and eventual migration towards the surface of powder particles 41
when fat is under a melted form (Thomas et al., 2004). Free fats are also susceptible to 42
oxidation and produce volatile compounds, such as aldehydes, ketones and lactones, which are 43
responsible for the development of off-flavour and off-odour in milk powders (Li et al., 2012). In 44
addition, the presence of fats on the surface provides hydrophobic layers that cause milk 45
powder to become less flowable and soluble in water (Bhandari, 2013; Kim et al., 2009a). 46
Surface fat may also form weak bridges between powder particles and promote agglomeration 47
and caking, thus reducing powder’s functional properties (Kim et al., 2009a; Nijdam and 48
Langrish, 2006; Ye et al., 2007). Additionally, lipase enzyme which is active even at low water 49
activity (0.1-0.6), reacts with lipid and releasing free fatty acids (Thomas et al., 2004). These 50
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
3
factors cause free fats to accumulate during storage and accelerate powder deterioration. 51
During storage, both SMP and WMP particles may collapse and shrink due to the release of 52
entrapped air and cause the particle volume and surface area to diminish (Thomas et al., 2004). 53
Consequently, powder density will increase and powder rehydration properties will deteriorate 54
because there is less contact area for water interaction and limited access for water penetration 55
(Thomas et al., 2004). 56
The addition of CO2 in dairy products such as raw and pasteurized milks, cheese and fermented 57
milk products has been investigated in the last decade for improvement in their shelf life, 58
quality and yield (Hotchkiss et al., 2006). Most of these studies mainly focus on the effect of CO2 59
acidification in skim milk concentrates. Few researchers have also studied its effect on the 60
resulting powders (Marella et al., 2015). It was demonstrated that carbonation modified micelle 61
structure and mineral contents of milk that led to improved functional properties. CO2 addition 62
increased milk acidity, which causes release of calcium phosphate, which in turn destabilizes 63
and releases casein micelles (Akissi-Kouame et al., 2009; Raouche et al., 2008; Raouche et al., 64
2007). This induced reorganization of micelle structure and modification of its surface activity, 65
improves renneting properties (Akissi-Kouame et al., 2009; Guillaume et al., 2004a; Guillaume 66
et al., 2004b; Klandar et al., 2009). In addition, increased amount of soluble calcium and 67
phosphate ions in the serum phase of milk protein concentrate (MPC) helps to produce high 68
protein MPC powder with reduced ionic calcium content, that contributes to reduced amount of 69
solublity loss during storage (Marella et al., 2015). 70
On the other hand, there has been a lack of research related to the application of CO2 in whole 71
milk concentrate. In fact the effect of carbonation on the functionality of WMP has never been 72
studied before. The effect of CO2 on fat is also poorly reported. The current commercial 73
application of CO2 for milk powders is modified atmosphere packaging, which is defined as the 74
replacement of air surrounding the product at the headspace of the packaging with CO2 or a 75
mixture of CO2 and N2 gases (Hotchkiss et al., 2006; Singh et al., 2012). This method effectively 76
improves WMP shelf life during storage mainly due to retardation of fat oxidation. However, its 77
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
4
protective effect is terminated once the package is opened and WMP is exposed to oxygen in the 78
atmosphere. CO2 is more soluble in hydrophobic materials, such as lipids, because it is non-polar 79
and has a dipole moment of zero (Arul et al., 1994; Ma and Barbano, 2003). Therefore, the 80
presence of fat in whole milk concentrate may demonstrate different effect of CO2 from that 81
seen in skim milk concentrate. It is expected that CO2 will dissolve in milk fat and provide 82
protection against oxidation and improve functional properties and shelf life of the resulting 83
powders. 84
The overall objective of this study is to investigate the effect of carbonation of whole milk 85
concentrate, at 1000 and 2000 ppm CO2 concentrations, towards the functional, 86
physicochemical and structural properties of the resulting spray dried powder during an 87
accelerated storage at 37 °C. 88
89
2. Materials and methods 90
2.1. Materials 91
WMP for preparing the concentrates were purchased in 25 kg bag from Total Foodtec Pty Ltd. 92
(Brisbane, Australia). Carbonation was accomplished by addition of a known amount of solid 93
CO2, also known as dry ice. 94
2.1.1. Preparation of concentrates 95
Whole milk concentrates (WMC) were prepared at 25 % solids content (w/v) by dissolving 25 g 96
of the powder in 100 mL of Milli-Q (deionized) water at 25 °C with constant stirring at a high 97
speed overhead stirrer for 1 hour. For each experiment, 3 litres of concentrate was prepared. 98
2.1.2. Carbonation of concentrates 99
For carbonation, 3 litre of the concentrate was poured into an 11 litre stainless steel keg 100
equipped with a manometer. Adequate amounts of dry ice was added into the keg allowing a 101
theoretical CO2 content of 1000 and 2000 ppm in the concentrate (Lee, 2014). The kegs were 102
then stored overnight at 4 °C. 103
2.1.3. Spray drying of the carbonated concentrates to powder and storage 104
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
5
A single-stage Anhydro Lab S1 spray dryer (Copenhagen, Denmark) was used to dry the 105
carbonated concentrates (without a decarbonation step). The spray dryer was fitted with a 106
pneumatic nozzle, supplied compressed air (640 kPa), and operated at 170 °C and 85 °C inlet 107
and outlet air temperatures, respectively. The spray dried samples were collected in zipped 108
aluminium bags and incubated at 37 °C for accelerated storage (aw = 0.2), along with the 109
commercial WMP (which was used to make the whole milk concentrate for carbonation and 110
eventual spray drying). 111
2.2. Powder chemical analysis 112
2.2.1. CO2 concentration in milk powder 113
The procedure as described by Jakobsen and Grete (2005) was used for the determination of 114
residual CO2 content in milk powder with some modification. A system was prepared by joining 115
two 100 mL Buchner flasks through the arms by a neoprene tubing of minimal length. The flasks 116
were closed by neoprene plugs. Then, the joints around the neoprene tubing and plugs were 117
applied with petroleum jelly and wrapped with parafilm to prevent gas leaks. One flask 118
contained 20 mL of standard 0.1M Ba(OH)2 solution and another contained 5 g milk powder and 119
40 mL of 0.5M H2SO4 solution. H2SO4, added to the powder in the flasks released CO2 from the 120
milk powder, which was absorbed in the standard Ba(OH)2 solution, producing BaCO3 121
precipitates. After 20 hours, the residual quantity of Ba(OH)2 was titrated against a standard 122
0.1M HCl solution using 1 % phenolphthalein indicator. The absorbed amount of CO2 was 123
calculated as CO2 per g of powder. 124
2.2.2. Moisture content and water activity 125
A Sartorius MA35 Infrared Moisture Analyzer (Sartorius, Goettingen, Germany) was used to 126
determine the moisture content of 5 g milk powder (heated at 105oC until the powder reached 127
constant weight). An AquaLab Series 3TE Water Activity meter (Decagon Devices Inc., Pullman, 128
USA) was used to determine the water activity of milk powders at 25 °C. 129
2.3. Functional properties of the powders 130
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
6
The solubility (Standardization, 2005), dispersibility (Standardization, 2014) and wettability 131
(Standardization, 2014) of the powders was determined as per the International Organization 132
for Standardization (ISO, Geneva) standards with slight modifications (due to the limited 133
quantity of powder). Nevertheless, similar ratios between powder and water were maintained. 134
2.4. Powder physical properties 135
Loose and tapped bulk density of powder are defined as the weight of powder divided by the 136
volume it occupies when loosely poured into a container and after being tapped 100 times, 137
respectively (Niro, 2006a). The weight and volume of the powders were recorded after being 138
poured into a 100 mL measuring cylinder and tapped 100 times against a table from 139
approximately 10 cm height. 140
True density of powder is defined as the mass of particles per unit volume (Niro, 2006a). A 141
Quantachrome Multipycnometer (Quantachrome Instruments, Florida, USA) was used to 142
determine the true density of milk powders. The pycnometer was operated using nitrogen gas 143
at 1.2 kPa. 144
Occluded and interstitial air are defined as the difference between the volume of particles at a 145
given mass and the volume of the same mass of air-free solids and of powders tapped 100 times, 146
respectively. The occluded and interstitial air contents of milk powder were calculated using the 147
formulas described by GEA Niro (Niro, 2006b). 148
2.5. Fat analysis 149
2.5.1. Free fat extraction 150
Free fat extraction from milk powder followed the procedures described elsewhere (Kim et al., 151
2002; Vignolles et al., 2007) with some modifications. Milk powder (2 g) was weighed and 152
mixed with 50 mL petroleum spirit for 5 minutes. The solvent was separated by filtration using 153
11 µm pore size Whatman filter paper into a round-bottom flask. The powders on the filtrate 154
paper were dried and kept for encapsulated fat analysis. The solvent in the flask was totally 155
evaporated, then the solvent-free flask was dried in the oven. Free fat percentage is the ratio 156
between the weight of extracted fat and the powder. 157
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
7
2.5.2. Encapsulated fat extraction 158
The encapsulated fat in the milk powder was extracted following the procedures described by 159
other authors (Kim et al., 2002; Vignolles et al., 2007). Milk powder recovered after free fat 160
extraction was weighed, then warm water was added. The warm mixture was vortexed to 161
completely rehydrate the powder and release the encapsulated fat. A solvent mixture made of n-162
hexane and 2-propanol (3:1 ratio v/v) was added and vortexed to extract the fat. The solution 163
was then centrifuged (1000 x g for 15 minutes) and the organic phase was filtered into a round-164
bottom flask. The aqueous phase was re-extracted with the solvent mixture and the collected 165
organic phase was totally evaporated. Then, the solvent-free flask was dried in the oven. 166
Encapsulated fat percentage is the ratio between the weight of extracted fat and the powder. 167
2.5.3. Total fat extraction 168
Total fat was extracted from 2 g of milk powder following the same procedure used for the 169
extraction of encapsulated fat described in section 2.6.2. 170
2.6. Structural and surface properties of the powders 171
2.6.1. Confocal Laser Scanning Microscopy 172
Milk powders were analyzed by CLSM using a Zeiss LSM 700 confocal microscope (Carl Ziess 173
Ltd. New South Wales 2113, Australia). Both the dyes, nile red and rhodamine B obtained from 174
Sigma Aldrish, Australia were used at a concentration of 0.1 g.L-1 in PEG 200, to label fat and 175
proteins, respectively (Auty et al., 2001). A ratio of 1/100 (dye/powder) was used to stain the 176
powders for 10 minutes before imaging. Observations were done with a 63x oil-immersion 177
objective. An argon laser was used to excite nile red and rhodamine B at wavelengths of 488 and 178
555 nm, respectively. Each micrograph is a representative of at least 10 images of each sample. 179
2.6.2. Scanning Electron Microscopy 180
A JEOL JSM-6460LA (JEOL Ltd., Tokyo, Japan) with a tungsten filament electron gun was used to 181
characterize the powder surface at 5 kV. Powders were subsequently mounted onto SEM stubs 182
by placing or sputtering them on a carbon double-sided adhesive tape. Excess particles were 183
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
8
removed with gentle tapping. Coating was done with platinum (Q150T Turbo-Pumped Sputter 184
Coater, ProSciTech Pty Ltd, Queensland, Australia) for 2 minutes (~ 10 nm thick). 185
2.6.3. XPS 186
Surface elemental composition of WMP (5-6 nm surface depth) was measured by X-ray 187
Photoelectron Spectroscopy (XPS). Spectra were obtained with a KRATOS Axis Ultra X-ray 188
photoelectron spectrometer (Kratos Analytical, Manchester, UK) equipped with a 189
monochromated Al Kα X-ray (hν = 1486.6 eV) operated at 150 W. Spectra were collected at 190
normal take-off angle (90°), and the analysis area was 700 × 300 μm2. Data analysis was done 191
using Casa software. 192
The relative atomic percentages of the elements (C, O, N, Ca and P) were used (via a matrix 193
formula) to determine the relative amounts of protein, fat, lactose and mineral on the surface of 194
the powders. The method is based on elemental ratios of the pure components in the sample 195
determined by XPS (Gaiani et al., 2006; Kim et al., 2002; Nikolova et al., 2015). The matrix 196
assumes that this ratio is a linear combination of elemental ratios of the pure components that 197
constitute the sample. In this work, the following real values were used: for lactose (C=61.6; 198
O=38.4; N=0); for proteins (C=68.2; O=18.5; N=13.3); for fat (C=87.0; O=12.3 and N=0.7) (Gaiani 199
et al., 2006). 200
2.7. Statistical analysis 201
Statistical analyzes (ANOVA and regression) were performed using Minitab 17 software 202
(Minitab Pty Ltd., Sydney, Australia) with 95 % level of confidence. 203
204
3. Results and discussion 205
3.1. Evolution of structural properties of WMP during storage and with CO2 treatment 206
3.1.1 SEM observations 207
The surface structures of industrial powders (Figure 1), untreated (Figure 2) and treated 208
(2000 ppm only) (Figure 3) spray dried powders were analyzed with SEM at two 209
magnifications. Comparisons between these images, clearly shows that the commercial powders 210
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
9
were much larger in size than the spray dried powders. It was visualised that the commercial 211
WMP have a relatively smooth surface, similar to that observed by Kim et al., (2002) and 212
Murrieta Pazos et al., (2012). The higher magnification images (Figures 1 - A2) have also 213
showed some pores present in the commercial WMP structure. However, for both of the spray 214
dried powders in this research, cracks or pores were not visible due to the relatively fine size. In 215
turn, many small particles were observed being attached to the bigger particles, which 216
presented a few dents and folds on the surface. 217
In addition, the outer structure of industrial WMP seemed to change quickly over time. As seen 218
in Figures 1 - C1 and C2, after 18 weeks of storage, the surface of industrial WMP was not as 219
smooth as the fresh powders with some cracks and pores clearly visible on the surface of aged 220
industrial powders. However, no significant changes were visible in the untreated and treated 221
WMPs during storage. 222
3.1.2 Confocal Laser Scanning Microscopy observations 223
After fat and protein labelling, the distribution of fat and inner structures of the powders were 224
analyzed with CLSM. For the fresh commercial WMP, small fat globules were seen uniformly 225
distributed in the particles (Figure 1 - A3) while with fresh spray dried WMP, small and large 226
fat globules were irregularly distributed (Figures 2 and 3 - A3). The particles are also seen 227
surrounded with red protein layers, while surface fat layers (quantified later by XPS) were not 228
visible due to the low CLSM resolution (Vignolles et al., 2007). Nevertheless, greater amount of 229
vacuoles were observed in the spray dried powder particles than the commercial powders. The 230
treated WMP particles (2000 ppm CO2) were noted to have larger vacuoles than the untreated 231
WMP. 232
Some significant modifications in the internal structure of powder particles were observed 233
during storage. Larger globules and irregular distribution of fat was visible in the aged 234
industrial powders. Moreover, free fat was seen exuding out of the particles in 6 and 18 weeks 235
of accelerated storage (Figures 1 - B3 and C3) and created interparticular bridges on the 236
surface, which has been demonstrated to cause caking (Kim et al., 2009a; Nijdam and Langrish, 237
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
10
2006; Ye et al., 2007). These microscopy observations were consistent with those reported 238
elsewhere (Kim et al., 2002). Several authors have also agreed that an increase in temperature 239
will cause the melting of some fats and consequently increase the formation of liquid bridges 240
that are detrimental to the powder functional and physical properties (Buma, 1971; Fitzpatrick 241
et al., 2004; Thomas et al., 2004). Additionally, this release of fat might have caused the 242
formation of uneven thickness of fat layers on the surface of industrial WMP as seen in Figure 1 243
- C2. On the other hand, small fat globules were observed in some of the vacuoles of aged WMP 244
particles without treatment (Figure 2 - B3 and C3). Meanwhile, large fat globules were clearly 245
seen deposited in the vacuoles of aged WMP produced from concentrates treated with 2000 246
ppm CO2 (Figure 3 – B3 and C3). 247
Twelve z-stack CLSM images were also obtained for better visualisation of the internal 248
distribution of fat in each powder after 6 weeks of storage (Figure 4). From these images, the 249
commercial WMP was observed to contain free fat at the surface and in the pores, while very 250
few or no vacuoles were present. In comparison to the spray dried WMP in this work, the lack of 251
vacuoles in the commercial WMP have caused fats to migrate towards the surface. 252
3.2. Effect of CO2 treatment on powder surface composition during storage 253
3.2.1 Effect of carbonation on fresh WMP 254
CO2 treatment of milk concentrates was found to affect the surface atomic composition of the 255
resulting powders analyzed by XPS. As described in Table 1, carbonation has significantly 256
increased the amount of surface nitrogen and minerals. Moreover, increasing the concentration 257
of CO2 from 1000 to 2000 ppm have further increased the amount of surface nitrogen of the 258
resulting powders (4.82 to 5.89 %). The increase in nitrogen content suggested an increase in 259
protein surface content as confirmed in Table 2. Meanwhile, significant reduction of carbon 260
content (75.51 to 65.59 %) and increased phosphorus content (0.13 to 0.41 %) were observed 261
on the surface of the powders treated with 2000 ppm CO2 only. 262
These surface atomic compositions obtained from XPS analyzes were converted into lactose, 263
protein, fat and mineral percentages using a matrix formula developed by other researchers 264
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
11
(Gaiani et al., 2006; Kim et al., 2002; Nikolova et al., 2015). As presented in Table 2, carbonation 265
was found to significantly reduce the amount of surface fat and increase lactose, protein and 266
minerals. Moreover, increasing the amount of CO2 treatment to 2000 ppm have further reduced 267
the surface fat content and increased protein content of the resulting powders. 268
In addition, fat was found to dominate the surfaces of both commercial and untreated spray 269
dried WMP, similar to those reported by others (Kim et al., 2002, 2009a, b). However, some of 270
the published works have reported surface fat coverage as high as 98%, whereas the values 271
obtained in our study are lower (58.1 % for commercial WMP and 51.5 % for untreated WMP). 272
It was also reported that WMP surface was covered with 2% lactose and negligible amount of 273
protein (Kim et al., 2002, 2009b), while the surfaces of commercial and untreated WMP in this 274
study were found to be covered with 26.5 and 29.7 % lactose, and 15.4 and 18.3 % protein, 275
respectively. These differences may be attributed to dissimilar spray drying processes, or 276
homogenization procedures (Kim et al., 2009a). Additionally, the matrix formula used in this 277
study was more accurate, because a real matrix formula was applied (with pure milk 278
compounds analyzed by XPS) instead of a theoretical formula (calculated from the chemical 279
formula of the milk compounds) which systematically overestimates the surface fat (Gaiani et 280
al., 2006). In comparison, the surfaces of both treated WMP were dominated by lactose and 281
protein, not fat, which is similar to the surface composition of SMP reported elsewhere (Gaiani 282
et al., 2006; Kim et al., 2002; Murrieta Pazos et al., 2012; Nijdam and Langrish, 2006; Vignolles 283
et al., 2007). These results were supported by SEM images, which showed a relatively smooth 284
surface of industrial powders due to the presence of surface fat (Figure 1 - A1 and A2), and a 285
more wrinkled surface of the treated powders due to the formation of lactose-protein matrix 286
and lower fat content (Figures 3 - A1 and A2) (Gaiani et al., 2006; Kim et al., 2002; Murrieta 287
Pazos et al., 2012; Nijdam and Langrish, 2006; Vignolles et al., 2007). 288
3.2.2 Effect of carbonation on WMP during storage 289
The surface elemental composition of the powders changed during storage. Over time, 290
significant increase in carbon content and decrease in oxygen and nitrogen contents were 291
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
12
observed on the surface of all powders (Table 1). The decreasing trend of oxygen content 292
during storage is inconsistent with others who have reported that the amount of oxygen on 293
WMP surface increased after 2 days of storage at 40 °C as oxidation of surface fat resulted in the 294
uptake of oxygen (Kim et al., 2002). Meanwhile, no calcium or phosphorus were detected on the 295
surface of the commercial powders during storage, which is in agreement with others (Gaiani et 296
al., 2006; Murrieta Pazos et al., 2012). In comparison, the amount of surface calcium of all of the 297
spray dried powders were observed to decrease significantly over time, while the surface 298
phosphorus content was significantly reduced for the 2000 ppm CO2-treated powders only. The 299
reduction in surface calcium and phosphorus content suggests a decrease in protein content on 300
the surface of the powder as confirmed in Table 2. 301
Carbonation was also found to have effect on the surface composition of WMP during storage. 302
For all powders, surface fat content significantly increased over storage, while the amount of 303
lactose and protein significantly decreased (Table 2). In comparison to the fresh powders, after 304
two weeks of storage, the surfaces of both treated WMP were dominated with fat rather than 305
lactose and protein. The domination of fat on the surface significantly increased over time and 306
may be due to some fractions of fat being melted at high storage temperature, causing it to 307
present in a mobile fluid form that allows migration towards the surface and spread (Murrieta 308
Pazos et al., 2012; Nijdam and Langrish, 2006). Despite the increasing amount of surface fat 309
during storage, after 18 weeks of storage, the surface of CO2 treated powders were still covered 310
with significantly lower fat content (88.4 % for 1000 ppm and 69.1 % for 2000 ppm CO2) than 311
the untreated (94.2 %) and commercial powders (95.3 %). The high amount of surface fat in 312
commercial WMP (increased from 58.1 to 95.3 %) is also supported by SEM observations which 313
demonstrated a different surface after 18 weeks of storage (Figure 1 – A2 and C2). 314
3.3. Effect of CO2 treatment on free and encapsulated fat fractions during storage 315
3.3.1 Effect of carbonation on fresh powders 316
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
13
Free fat quantification by solvent extraction (Table 3) involves those present in pores, cracks 317
and on the surface of the powder, while XPS analysis (Table 2) only measures the amount of fat 318
on the surface. Thus, differences are expected between the two analyzes. 319
Carbonation of whole milk concentrates was found to have insignificant effects (P>0.05) on the 320
amount of free fat, with those in the commercial powders was significantly lower (P<0.05) than 321
the spray dried powders treated with 2000 ppm CO2 (Table 3). Since carbonation at 4 °C was 322
found to induce fat coalescence (Kosasih et al., 2015), this result may be attributed to the larger 323
fat droplets in milk concentrates which has been reported to produce powders with greater 324
amount of free fat (Vignolles et al., 2007; Ye et al., 2007). The increase of free fat content due to 325
carbonation is also certainly linked to the surface area of the powders. As the treated WMP 326
particles are smaller, free fat extraction will be more important (Buma, 1971). Additionally, no 327
significant differences (P>0.05) were observed in the encapsulated fat contents between treated 328
and untreated powders. On the other hand, the sum of free and encapsulated fat fractions were 329
not exactly equal to the total fat content. This difference was also observed by others (Kim et al., 330
2009a, b; Murrieta Pazos et al., 2012) and may be due to losses during the extraction as some fat 331
may remain on the filter paper and glassware. 332
3.3.2 Effect of carbonation on WMP during storage 333
The spray dried powders in this study were produced from milk concentrates prepared by 334
reconstitution of the commercial WMP. Thus, it can be assumed that the commercial and spray 335
dried powders will have the same amount of total fat, which were found to range between 30-336
33 % (Table 3). The total fat content was also assumed to be the same during storage, but the 337
amount of free and encapsulated fats may vary. 338
The free fat fraction of all powders did not change significantly during storage. However, after 339
storage of two weeks and beyond, the free fat content of the spray dried powders (treated and 340
untreated) were significantly higher than the commercial powders. Again, this difference may 341
be due to the smaller size of the spray dried WMP particles and their tendency to disintegrate 342
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
14
during storage which caused them to have larger surface area (Murrieta Pazos et al., 2012; 343
Tamsma, 1959). Thus, at the same length of time, more fat may be extracted. 344
Meanwhile, the encapsulated fat content of all powders was found to significantly decrease after 345
18 weeks of storage. The reduced amount of encapsulated fat in the commercial powders may 346
be caused by the migration of free fat towards the surface (Thomas et al., 2004), which was 347
clearly visible through CLSM images (Figures 1 and 4) and XPS results. This movement was 348
expected to increase the amount of free fat (solvent extraction) and surface fat (XPS). However, 349
significant changes were only observed for surface fat content. This could be due to the 350
presence of cracks in industrial powders that may interfere with solvent penetration during 351
extraction (Tamsma, 1959; Vignolles et al., 2007). On the other hand, the decreasing trend of 352
encapsulated fat content for the spray dried WMP may be caused by the migration of fat 353
towards the internal air spaces. Microscopy observations have suggested that other than the 354
surface, fat also migrated into the air/vacuole spaces in the powder structure (Figure 4). 355
Moreover, others (Kim et al., 2009b; Nijdam and Langrish, 2006) have suggested that the 356
release of encapsulated fat, which was observed to be time-dependent, might be due to a change 357
in molecular arrangement and/or redistribution of components promoted by concentration 358
gradient that is caused by temperature fluctuation. 359
3.4. Evolution of physical properties of the powders during storage and with CO2 360
treatment 361
3.4.1 Bulk density and interstitial air content 362
Carbonation of milk concentrates had no effects on bulk density and interstitial air content of 363
the resulting powders during storage. Meanwhile, all of the spray dried powders were found to 364
have significantly lower bulk density and higher interstitial air content than the industrial WMP. 365
Commercial production of WMP involved agglomeration and/or lecithination, whereas the 366
powders in this study were produced via single-stage spray drying. This resulted in very fine 367
powders with increased cohesion (Fitzpatrick et al., 2004), thus causing it to form agglomerates 368
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
15
with wide particle size distribution, as seen in Figures 2 and 3. Consequently, interstitial air 369
content increased, leading to reduced compactness. 370
During storage, bulk density of the commercial powders did not change, while the interstitial air 371
content significantly increased after 18 weeks. For the untreated WMP, bulk density and 372
interstitial air content were found significantly increased and decreased, respectively, over time. 373
Meanwhile, for both treated WMP, the results were inconclusive as the values fluctuate with 374
time. This may be attributed to the small particle size and its tendency to form various sizes of 375
agglomerates. 376
3.4.2 True density and occluded air content 377
True density is influenced by the amount of air entrapped in the particles. In this study, no 378
significant differences were observed for WMP treated with 1000 ppm CO2, whereas those 379
treated with 2000 ppm CO2 were found to contain significantly higher amount of occluded air. 380
Concurrently, the 2000 ppm powders have a lower true density as air entrapped in the particle 381
is inversely correlated to the true density. As described in Table 4, occluded air content of the 382
treated powders (2000 ppm) was about 8 times greater than the untreated powders and 18 383
times greater than the commercial powders. This result was supported by the CLSM images as 384
seen in Figures 1, 2 and 3. According to Lee, (2014) and Skanderby et al., (2009), the presence 385
of air in milk concentrates is responsible for the achievement of internal porosity in the 386
resulting powders. Therefore, high amounts of occluded air in the treated powders were 387
expected as the carbonated milk concentrates in this study were not degassed prior to spray 388
drying. 389
During storage, the occluded air content of the treated powders decreased. However, these 390
reductions were not statistically significant. Nevertheless, the release of entrapped air indicated 391
that the powder particles collapsed (Aguilar and Ziegler, 1994; Thomas et al., 2004), and 392
resulted in increased bulk density. Economically, manufacturers are interested in high bulk 393
density to reduce the transport volume and save packing materials and storage capacity 394
(Skanderby et al., 2009). However, particle collapse also caused reduction of surface area and 395
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
16
consequently decreased reconstitutional properties due to limited contact area for water 396
interaction, which are not desirable for consumers (Aguilar and Ziegler, 1994; Thomas et al., 397
2004). 398
3.5. Evolution of functional properties of the powder during storage and with CO2 399
treatment 400
3.5.1 Effect of carbonation on fresh powders 401
The ability of milk powder to instantly dissolve in water is affected by several factors, including 402
wettability, dispersibility and solubility. Wettability is the ability of powder to absorb water on 403
the surface and to penetrate the surface of still water (Sharma et al., 2012). Dispersibility is a 404
measure of how easily lumps or agglomerates break into individual particles in water, and 405
solubility measures the amount of powder that can be brought into solution or stable 406
suspension (Sharma et al., 2012). 407
In this experiment, all WMPs did not wet within 5 minutes (results not shown). These results 408
were expected as the spray dried powders were not agglomerated and/or lecithinated as in 409
commercial WMP production. Meanwhile, the incapability of industrial WMP to wet within 5 410
minutes may be attributed to the presence of surface free fats (Table 2 and 3), which render 411
the powder surface hydrophobic (Bhandari, 2013; Kim et al., 2009a, b). 412
Carbonation at 2000 ppm was found to significantly (P<0.05) reduce the dispersibility and 413
solubility of the resulting powders (Figures 5A and 5B). This result contradicts the XPS 414
analysis that showed significant reduction in surface fat content and increase in surface protein 415
and lactose contents of the treated powders (Table 2). Lactose and protein are soluble in water, 416
therefore the treated powders should have improved reconstitutional properties. This adverse 417
effect may be attributed to the very fine size and increase in interstitial air content of the treated 418
powder (Table 4). It has been reported that when milk powder is added to water, capillary 419
forces attract water molecules to move toward the powder particles located on and above the 420
surface to replace interstitial air, which is often incomplete due to insufficient amount of 421
penetrating water (Skanderby et al., 2009). As a result, air bubbles are present between wetted 422
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
17
particles, creating lumps that are wet and swollen outside and dry inside, and are difficult to 423
dissolve, even with strong agitation (Skanderby et al., 2009). Additionally, the dispersibility of 424
all of the spray dried powders was lower than the commercial powders. These results were 425
expected because as mentioned before, the spray dried powders in this study were not 426
agglomerated after spray drying. Thus, they have less desirable reconstitutional properties than 427
the industrial WMP. 428
3.5.2 Effect of carbonation of concentrates on WMP during storage 429
During storage, the dispersibility of industrial powders significantly (P<0.05) increased, 430
whereas the dispersibility of all of the spray dried powders decreased (about 10 %) (Figure 431
5A). These results were expected, although the increasing trend of dispersibility for industrial 432
powder is unusual. High amounts of free fat coverage on the surface of milk powder have been 433
associated with reduced dispersibility (Tamsma, 1959; Thomas et al., 2004). Thus, dispersibility 434
of industrial WMP should decrease over time. One possible explanation is that the migration of 435
free fats toward the surface of particles caused the industrial powders to cake, therefore 436
producing large agglomerates that are more dispersible. Additionally, this contradicting results 437
may be due to poor reproducibility of these analyses as already demonstrated by others (Gaiani 438
et al., 2006). 439
Similarly, the solubility of all powders were found to decrease significantly (P<0.05) after 18 440
weeks of accelerated storage. As shown in Figure 5B, industrial WMP has the least solubility 441
loss over storage (from 98.5 to 96.4 %), whereas WMP treated with 2000 ppm CO2 has the 442
greatest amount of loss (from 87.9 to 71.8 %). Meanwhile, at the end of storage, the solubility of 443
WMP treated with 1000 ppm CO2 (88.3 %) was greater than the untreated spray dried WMP 444
(83.8 %). This decreasing trend of solubility was expected as the amount of surface fat 445
increased during storage. Moreover, after 18 weeks storage, the XPS results showed that the 446
treated powders have significantly lower amount of surface fat than the commercial and 447
untreated WMP (Table 2), thus they were expected to have lesser amount of solubility loss. This 448
expectation was observed between the untreated and 1000 ppm CO2-treated WMP. Meanwhile, 449
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
18
the greater extent of solubility loss for the 2000 ppm CO2-treated WMP as compared to the 450
untreated and commercial powders may be attributed to the physical properties (Table 4). It 451
may also be attributed to the amount of denatured protein during drying and storage (Thomas 452
et al., 2004). According to (Thomas et al., 2004), during spray drying, water molecules bound to 453
milk proteins are replaced by lactose, which do not impact the functionality when it is in 454
amorphous form. However, crystallization of lactose releases water molecules, causing 455
destabilization and modification of milk protein structures and resulting in enhanced non-456
covalent hydrophobic interactions to form high molecular weight protein aggregates that are 457
insoluble in water (Thomas et al., 2004). Since, the powders in this experiment were produced 458
by reconstituting a commercial WMP, it can be assumed that these powders were subjected to 459
thermal processing twice as much, which would increase the amount of denatured proteins as 460
compared to the industrial WMP. 461
462
4. Conclusions 463
In this study, carbonation of whole milk concentrates was found to produce WMP with reduced 464
surface fat content, dispersibility, solubility and true density, and increased occluded air 465
content. Moreover, encapsulated fat content was found to decrease over the storage period, 466
while the amount of surface fat increased. In addition to the release of fat onto the surface, 467
Confocal Laser Scanning Microscopy observations have shown the migration of free fat into the 468
vacuoles. Meanwhile, dispersibility and solubility of the powders decrease during storage. 469
The decrease of surface fat content with CO2 treatment supported the hypothesis because this 470
suggests that there are less fat exposure to air, hence less chances of fat oxidation to occur, 471
resulting in powders with better shelf life, which needs to be investigated further. The surface 472
composition of the treated powders has also shown potential improvements in functional 473
properties. However, this expectation was not met due to the spray dried particles produced by 474
a laboratory scale dryer which is very small in size. Therefore, additional investigations should 475
be performed by producing the treated WMP via two-stage spray drying, as well as to analyze 476
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
19
peroxide and/other values to determine the fat oxidation in order to confirm these hyphotesis 477
further. 478
479
Acknowledgements 480
The authors would like to thank the European grant (Milk PEPPER n°621727, International 481
Outgoing Fellowship grant) for their financial support towards this project. The authors 482
acknowledge the facilities, and scientific and technical assistance provided by the School of 483
Agriculture and Food Sciences (SAFS) and the Australian Microscopy & Microanalysis Research 484
Facility at the Centre for Microscopy and Microanalysis (CMM) at The University of Queensland. 485
References 486
Aguilar, C.A., Ziegler, G.R., (1994). Physical and microscopic characterization of dry whole milk 487
with altered lactose content. 2. Effect of lactose crystallization. Journal of Dairy Science 488
77(77), 1198-1204. 489
Akissi-Kouame, F., Klandar, A.H., Marchesseau, S., Lagaude, A., (2009). Effect of a CO2-490
acidification cycle on physicochemical and acid gelation properties of skim milk 491
reconstituted with added pectin. International Dairy Journal 19, 393-398. 492
Arul, J., Tardiff, R., Boudreau, A., McGinnis, D.S., Lencki, R.W., (1994). Solubility of milk fat 493
triglycerides in supercritical carbon dioxide. Food Research International 27, 459-467. 494
Auty, M.A.E., Twomey, M., Guinee, T.P., Mulvihill, D.M., (2001). Development and application of 495
confocal scanning laser microscopy methods for studying the distribution of fat and 496
protein in selected dairy products. Journal of Dairy Research 68, 417-427. 497
Bhandari, B., (2013). Introduction to food powders, in: Bhandari, B., Bansal, N., Zhang, M., 498
Schuck, P. (Eds.), Handbook of food powders processes and properties. Woodhead 499
Publishing, Cambridge, pp. 1-25. 500
Buma, T.J., (1971). Free fat in spray-dried whole milk 5. Cohesion. Determination, influence of 501
particle size, moisture content and freefat content. Netherlands Milk and Dairy Journal 25, 502
107-122. 503
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
20
Fitzpatrick, J.J., Iqbal, T., Delaney, C., Twomey, T., Keogh, M.K., (2004). Effect of powder 504
properties and storage conditions on the flowability of milk powders with different fat 505
contents. Journal of Food Engineering 64, 435-444. 506
Gaiani, C., Ehrhardt, J.J., Scher, J., Hardy, J., Desobry, S., Banon, S., (2006). Surface composition of 507
dairy powders observed by X-ray photoelectron spectroscopy and effects on their 508
rehydration properties. Colloids and Surfaces B-Biointerfaces 49(1), 71-78. 509
Guillaume, C., Gastaldi, E., Cuq, J.L., Marchesseau, S., (2004a). Effect of pH on rennet clotting 510
properties of CO2-acidified skim milk. International Dairy Journal 14, 437-443. 511
Guillaume, C., Jimenez, L., Cuq, J.L., Marchesseau, S., (2004b). An original pH-revesible treatment 512
of milk to improve rennet gelation. International Dairy Journal 14, 305-311. 513
Hotchkiss, J.H., Werner, B.G., Lee, E.Y.C., (2006). Addition of Carbon Dioxide to Dairy Products to 514
Improve Quality: A Comprehensive Review. Comprehensive Reviews in Food Science and 515
Food Safety 5(4), 158-168. 516
Jakobsen, M., Grete, B., (2005). Solubility of carbon dioxide in far and muscle tissue. Journal of 517
Muscle Foods 17(1), 9-19. 518
Kim, E.H.J., Chen, X.D., Pearce, D., (2002). Surface characterization of four industrial spray-dried 519
dairy powders in relation to chemical composition, structure and wetting property. 520
Colloids and Surfaces B: Biointerfaces 26, 197-212. 521
Kim, E.H.J., Chen, X.D., Pearce, D., (2009a). Surface composition of industrial spray-dried milk 522
powders. 1. Development of surface composition during manufacture. Journal of Food 523
Engineering 94, 163-168. 524
Kim, E.H.J., Chen, X.D., Pearce, D., (2009b). Surface composition of industrial spray-dried milk 525
powders. 3. Changes in the surface composition during long-term storage. Journal of Food 526
Engineering 94, 182-191. 527
Klandar, A.H., Chevalier-Lucia, D., Lagaude, A., (2009). Effects of reversed CO2 acidification 528
cycles, calcium supplementation, pH adjustment and chilled storage on physico-chemical 529
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
21
and rennet coagulation properties of reconstituted low- and medium-heat skim milk 530
powders. Food Hydrocolloids 23, 806-815. 531
Kosasih, L., Bhandari, B., Prakash, S., Bansal, N., Gaiani, C., (2016). Physical and functional 532
properties of whole milk powders prepared from concentrate partially acidified with CO2 533
at two temperatures. International Dairy Journal, in press. 534
Lee, E.Y.C., (2014). The effects of carbon dioxide on the physico-chemical and microbiological 535
properties of skim milk and whey protein concentrates, PhD thesis, School of Agriculture 536
and Food Sciences. The University of Queensland, Brisbane, Australia p. 174. 537
Li, Y., Zhang, L., Wang, W., (2012). Formation of aldehyde and ketone compounds during 538
production and storage of milk powder. Molecules 17, 9900-9911. 539
Ma, Y., Barbano, M., (2003). Effect of temperature of CO2 injection on the pH and freezing point 540
of milks and creams. Journal of Dairy Science 86, 1578-1789. 541
Marella, C., Salunke, P., Biswas, A.C., Kommineni, A., Metzger, L.E., (2015). Manufacture of 542
modified milk protein concentrate utilizing injection of carbon dioxide. Journal of Dairy 543
Science 98(6), 3577-3589. 544
Murrieta Pazos, I., Gaiani, C., Galet, L., Scher, J., (2012). Composition gradient from surface to 545
core in dairy powders: Agglomeration effect. Food Hydrocolloids 26(1), 149-158. 546
Nijdam, J.J., Langrish, T.A.G., (2006). The effect of surface composition on the functional 547
properties of milk powders. Journal of Food Engineering 77, 919-925. 548
Nikolova, Y., Petit, J., Sanders, C., Gianfrancesco, A., Scher, J., Gaiani, C., (2015). Toward a better 549
determination of dairy powders surface composition through XPS matrices development. 550
Colloids and Surfaces B-Biointerfaces 125, 12-20. 551
Niro, G., (2006a). GEA Niro analytical method A-2a-Powder bulk density, Soeborg, Denmark. 552
Niro, G., (2006b). GEA Niro analytical method A-11-a Particle density occluded air and 553
interstitial air (air pycnometer method), Soeborg, Denmark. 554
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
22
Raouche, S., Dobenesque, M., Bot, A., Cuq, J.L., Marchesseau, S., (2008). Stability of casein 555
micelles subjected to CO2 reversible acidification: impact of carbonation temperature and 556
chilled storage time. International Dairy Journal 18, 221-227. 557
Raouche, S., Dobenesque, M., Bot, A., Lagaude, A., Cuq, J.L., Marchesseau, S., (2007). Stability of 558
casein micelle subjected to reversible CO2 acidification: impact of holding time and chilled 559
storage. International Dairy Journal 17, 873-880. 560
Sharma, A., Jana, A.H., Chavan, R.S., (2012). Functionality of Milk Powders and Milk-Based 561
Powders for End Use Applications - A Review. Comprehensive Reviews in Food Science 562
and Food Safety 11(5), 518-528. 563
Singh, P., Wani, A.A., Karim, A.A., Langowski, H.C., (2012). The use of carbon dioxide in the 564
processing and packaging of milk and dairy products: a review. International Journal of 565
Dairy Technology 65(2), 161-177. 566
Skanderby, M., Westergaard, V., Partridge, A., Muir, D.D., (2009). Dried Milk Products, in: 567
Tamime, A.Y. (Ed.), Dairy Powders and Concentrated products. Blackwell Publishing Ltd., 568
UK, pp. 180-234. 569
Standardization, I.O.f., (2005). Dried milk and dried milk products, Determination of insolubility 570
index. ISO, Geneva. 571
Standardization, I.O.f., (2014). Dried milk and dried milk products, Determination of the 572
dispersibility and wettability. ISO, Geneva. 573
Tamsma, A., (1959). Free fat in foam-dried whole milk. Journal of Dairy Science 42(2), 240-250. 574
Thomas, M.E.C., Scher, J., Desobry-Banon, S., Desobry, S., (2004). Milk powders ageing: effect on 575
physical and functional properties. Critical Reviews in Food Science and Nutrition 44, 576
297-322. 577
Vignolles, M.-L., Jeantet, R., Lopez, C., Schuck, P., (2007). Free fat, surface fat and dairy powders: 578
interactions between process and product. A review. Le Lait 87(3), 187-236. 579
Ye, A., Anema, S.G., Singh, H., (2007). Behaviour of homogenized fat globules during the spray 580
drying of whole milk. International Dairy Journal 17, 374-382. 581
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Figure 1. Industrial powder images observed by SEM and CLSM during storage.
A: fresh powder; B: 6 weeks storage; C: 18 weeks storage. 1: SEM (x 1000); 2: SEM (x 3000); 3:
CLSM with fat (green) and proteins (red) labelling (100 x 100 µm).
Figure 2. Spray-dried WMP untreated with CO2 (0 ppm) images observed by SEM and CLSM
during storage.
A: fresh powder; B: 6 weeks storage; C: 18 weeks storage. 1: SEM (x 1000); 2: SEM (x 3000); 3:
CLSM with fat (green) and proteins (red) labelling (40 x 40 µm).
Figure 3. Spray-dried WMP treated with CO2 (2000 ppm) images observed by SEM and CLSM
during storage.
A: fresh powder; B: 6 weeks storage; C: 18 weeks storage. 1: SEM (x 1000); 2: SEM (x 3000); 3:
CLSM with fat (green) and proteins (red) labelling (40 x 40 µm).
Figure 4. Twelve z-stack images (each 2 µm depth) of commercial powder (C), 0 ppm powder
(0) and 2000 ppm powder (2) after 6 weeks storage obtained by CLSM after fat (green) and
protein (red) labelling.
Grey arrows are surface fat (for commercial powder) and surface protein layer (for 0 and 2000
ppm powder); white arrows are fat deposits in the vacuoles and blue arrows are encapsulated fat.
Figure 5. Evolution of dispersibility (A) and solubility (B) for commercial WMP and spray dried
WMP (untreated and treated with CO2 at 1000 and 2000 ppm) during an accelerated storage at
37 °C for 18 weeks.
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Table 1. Surface elemental composition in carbon (C), oxygen (O), nitrogen (N), calcium (Ca) and phosphorus (P) of the powders obtained by XPS
(mean of two independent analysis).
Tim
e
(we
eks)
Commercial powder Powder without CO2 treatment Powder with 1000 ppm CO2
treatment
Powder with 2000 ppm CO2
treatment
C O N Ca P C O N Ca P C O N Ca P C 0 N Ca P
0
77.3
1 ±
0.11
20.1
5 ±
0.34
2.45
±
0.06
0.05
±
0.00
0.03
±
0.00
75.5
1 ±
3.39
21.1
1 ±
2.50
2.79
±
0.20
0.45a
±
0.06
0.13
±
0.06
69.9
1 ±
2.96
23.1
2 ±
2.67
4.82a
,b ±
0.07
0.90a
,b ±
0.03
0.26a
±
0.03
65.6
9a,b ±
0.11
27.16
±
0.07
5.89a
,b,c ±
0.30
0.85a
,b ±
0.07
0.41a
,b,c ±
0.01
2
80.1
7 ±
1.48
18.0
0 ±
0.33
1.84d
±
0.07
0.00
±
0.00
0.00
±
0.00
79.4
7 ±
3.01
18.0
3 ±
0.67
2.04d
±
0.10
0.31a
,d ±
0.03
0.16a
±
0.01
77.0
3d ±
0.76
18.3
4d ±
0.04
3.96a
,b,d ±
0.07
0.41a
,b,d ±
0.01
0.27a
,b ±
0.00
74.9
0d ±
0.30
20.07a,b,c,d ±
0.10
4.00a
,b,d ±
0.17
0.71a
,b,c ±
0.01
0.32a
,b,c,d ±
0.00
6
86.1
5d ±
1.78
13.0
9d ±
1.44
0.76d
±
0.07
0.00
±
0.00
0.00
±
0.00
85.9
6d ±
2.03
12.1
7d ±
0.21
1.57a
,d ±
0.29
0.20d
±
0.00
0.10a
±
0.00
84.9
4d ±
1.16
13.1
2d ±
0.16
1.55a
,d ±
0.08
0.34a
,d ±
0.06
0.04d
±
0.03
79.5
1a,b,d
±
0.03
17.84a,b,c,d ±
0.11
2.02a
,d ±
0.06
0.51a
,b,d ±
0.08
0.12a
,d ±
0.03
18
86.1
7d ±
2.91
12.9
3d ±
1.39
0.90d
±
0.03
0.00
±
0.00
0.00
±
0.00
85.6
5d ±
1.90
12.5
7d ±
0.62
1.40a
,d ±
0.07
0.28a
,d ±
0.01
0.10a
±
0.01
84.9
6d ±
0.62
13.2
0d ±
0.07
2.10a
,b,d ±
0.11
0.35a
,d ±
0.07
0.29a
,b ±
0.04
79.5
7d ±
0.34
17.74a,b,c,d ±
0.74
2.08a
,b,d ±
0.03
0.49a
,b,d ±
0.00
0.12a
,c,d ±
0.00 a significant from commercial powder b significant from untreated powder (0 ppm) c significant from treated powder (1000 ppm) d significant during storage (regression analyses used T0 as a baseline)
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Table 2. Powder surface composition in lactose (Lac), proteins (Prot), fat and minerals (Min) calculated from XPS data (Table 1) using a matrix
formula (mean of two independent analysis).
Time
(weeks)
Commercial powder Powder without CO2 treatment Powder with 1000 ppm CO2
treatment Powder with 2000 ppm CO2 treatment
Lac Prot Fat Min Lac Prot Fat Min Lac Prot Fat Min Lac Prot Fat Min
0 26.5
± 1.2
15.4 ±
0.4
58.1 ±
0.4
0.1 ±
0.0
29.7 ±
6.4
18.3 ±
1.6
51.5 ±
1.8
0.6a ±
0.1
34.3 ±
7.4
34.8a,b
± 0.6
28.9a,b
± 2.4
1.2a,b
±
0.1
47.2a ±
0.5
43.9a,b,c
± 2.3
7.8a,b,c
±
1.3
1.3a,b
±
0.1
2 19.2
± 1.3
10.1d
± 0.7
70.5 ±
3.4
0.0 ±
0.0
19.4d
± 0.6
11.7d
± 0.6
68.5d
± 2.5
0.5a ±
0.0
17.1d
± 0.4
26.9a,b,d
± 0.5
55.4a,b,d
± 0.7
0.7a,b,d
± 0.0
23.6a,b,c,
d ± 0.3
27.9a,b,d
± 1.8
47.5a,b,d
± 0.8
1.0a,b,c,d
± 0.0
6 3.4
d±
4.7
0.7d ±
0.9
95.9d
± 5.8
0.0 ±
0.0
0.0d ±
0.0
7.0a,d
± 2.1
94.8d
± 1.3
0.3a,d
± 0.0
1.7d ±
0.1
6.9a,d
±
0.6
91.1d ±
1.1
0.4a,d
±
0.1
18.8a,b,c,
d ± 0.3
11.6a,d
± 0.5
69.0a,b,c,
d ± 0.6
0.6a,b,d
± 0.1
18 3.1
d ±
4.4
1.7d ±
0.6
95.3d
± 6.6
0.0 ±
0.0
0.3d ±
0.4
5.6a,d
± 0.5
94.2d
± 1.1
0.4a,d
± 0.0
0.7d ±
0.3
11.1a,b,d
± 0.8
88.4d ±
0.3
0.6a,b,d
± 0.1
18.3a,b,c,
d ± 2.3
12.0a,b,d
± 0.3
69.1a,b,c,
d ± 1.4
0.6a,d
±
0.0 a significant from commercial powder b significant from untreated powder (0 ppm) c significant from treated powder (1000 ppm) d significant during storage (regression analyses used T0 as a baseline)
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Table 3. Fat fraction evolution of commercial WMP (C), spray-dried WMP untreated (0), spray-dried WMP treated with CO2 at 1000 ppm (1) and
spray-dried WMP treated with CO2 at 2000 ppm (2) during 18 weeks storage at 37 °C.
Time
(weeks)
Free fat (g/100g) Encapsulated fat (g/100g) Average total fat (g/100g)
C 0 1 2 C 0 1 2 C 0 1 2
0 1.08 ± 0.03 2.45 ±
0.90
3.18 ±
0.67 3.89a ±
0.93
29.64 ±
0.76
28.93 ±
0.53
28.95 ±
2.00
27.99 ±
2.41
33.55 ±
4.58
32.16 ±
4.12
30.31 ±
2.31
32.49 ±
5.06
2 1.02 ± 0.19 3.62a ±
0.03
4.15a ±
0.31
4.42a,b ±
0.05
6 1.04 ± 0.21
3.59a ±
0.12
4.01a ±
0.32
4.03a ±
0.21
12 0.93 ± 0.10 -* 3.29a ±
0.24
2.73a ±
0.20
18 0.98 ± 0.03 3.21a ±
0.58
2.99a ±
0.21
2.60a ±
0.41
25.22d ±
0.12
18.36d ±
3.99
17.55d ±
2.85
18.55d ±
2.42 a significant from commercial powder (C) b significant from untreated powder (0 ppm) c significant from treated powder (1000 ppm) d significant during storage (regression analyses used T0 as a baseline)
*0 ppm powders were not analysed at T12w due to limited quantity
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Table 4. Physical properties evolution of commercial WMP (C), spray-dried WMP untreated (0), spray-dried WMP treated with CO2 at 1000 ppm (1)
and spray-dried WMP treated with CO2 at 2000 ppm (2) during 18 weeks storage at 37 °C.
Time
(wee
ks)
Loose bulk density**
(g/mL)
Tapped bulk density**
(g/mL) True density (g/mL)
Occluded air content
(mL/100g)
Interstitial air content
(mL/100g)
C 0 1 2 C 0 1 2 C 0 1 2 C 0 1 2 C 0 1 2
0
0.46
±
0.00
0.29a ±
0.02
0.28a ±
0.04
0.28a ±
0.04
0.63
±
0.00
0.48a ±
0.01
0.46a ±
0.04
0.45a ±
0.05
1.25
±
0.00
1.18
±
0.02
1.06
±
0.10
1.01a,b ±
0.09
0.00
±
0.03
2.01
±
1.32
14.3
6 ±
9.48
17.67a,b ±
10.45
78.6
9 ±
0.03
123.5
6a ±
5.06
125.7
7a ±
11.35
125.8
1a ±
19.27
2
0.46
±
0.00
0.31a ±
0.00
0.29a,b ±
0.00
0.32a,b,c
±
0.00
0.65
±
0.00
0.47a ±
0.00
0.45a,b ±
0.00
0.45a,b ±
0.00
1.24
±
0.00
1.18
±
0.00
1.10a ±
0.06
1.07a ±
0.00
0.00
±
0.05
1.64
±
0.22
10.0
3a ±
4.84
11.23a ±
6.63
72.6
4d ±
0.05
128.2
6a ±
0.22
131.2
7a ±
4.84
127.7
4a ±
0.22
6
0.47
±
0.00
0.33a,d ±
0.00
0.31a,b ±
0.00
0.33a,b,c
±
0.00
0.65
±
0.00
0.53a,d ±
0.00
0.52a,b,d
±
0.00
0.51a,b,c
±
0.00
1.24
±
0.00
1.19a ±
0.00
1.14a,b ±
0.00
1.08a,b,c
±
0.01
0.00
±
0.06
0.84a ±
0.18
6.83a
,b ±
0.30
9.84a,
b,c ±
5.43
73.2
8d ±
0.06
105.8
0a,d ±
0.18
103.6
8a,b,d
± 0.30
102.2
6a,b,c ±
0.51
12
0.51
±
0.00
-*
0.34a,d ±
0.00
0.33a,c ±
0.00
0.64
±
0.00
-*
0.60a,d ±
0.00
0.55a,c,d
±
0.00
1.24
±
0.00
-*
1.14a ±
0.00
1.04a,c ±
0.00
0.00
±
0.02
-* 6.77a
±
0.09
13.76a,c ±
7.59
76.1
8d ±
0.02
-*
78.84a
,d ±
0.09
85.51a
,c,d ±
0.21
18
0.47
±
0.00
0.35
a,d ±
0.00
0.31
a,b ±
0.00
0.33
a,b,c
±
0.00
0.62
±
0.00
0.53
a,d ±
0.00
0.44
a,b ±
0.00
0.44
a,b,c
±
0.00
1.24
±
0.01
1.19
a ±
0.01
1.14
a,b ±
0.00
1.04
a,b,c
±
0.00
0.00
±
0.34
1.17
a ±
0.71
6.96
a,b ±
0.23
13.92
a,b,c ±
6.86
80.7
1d ±
0.34
106.0
8 a,d ±
0.71
137.9
2 a,b ±
0.23
131.6
3 a,b,c
± 0.42
a significant from commercial powder (C) b significant from untreated powder (0 ppm) c significant from treated powder (1000 ppm) d significant during storage (regression analyses used T0 as a baseline)
*0 ppm powders were not analysed at T12w due to limited quantity.
**analyses were not replicated, except for T0.
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPTHighlights :
• Carbonation was found to produce powders with reduced surface fat content and increased occluded air content.
• During storage, the surface coverage of fat on powder particles increased due to the release and spreading of encapsulated fat.
• Improvement of powder’s functional properties with CO2 was not observed.