doi: 10.1111/j.1471-0307.2010.00613.

ORIGINAL
RESEARCH Effects of different calcium salts on properties of milk
related to heat stability
ESEREOSA D OMOARUKHE, 1 NATTIRA ON-NOM, 2 ALISTAIR S
GRANDISON 2 and MICHAEL J LEWIS 2 *
1
Esereosa D Omoarukhe, Quality Auditor, Launceston, Cornwall PL15 7AF, and 2Department of Food and Nutritional
Sciences, University of Reading, Whiteknights, PO Box 226, Reading RG6 6AP, UK

Insoluble calcium salts were added to milk to increase total calcium by 30 mM, without changing proper-
ties influencing heat stability, such as pH and ionic calcium. There were no major signs of instability asso-
ciated with coagulation, sediment formation or fouling when subjected to ultra high temperature (UHT)
and in-container sterilisation. The buffering capacity was also unaltered. On the other hand, addition of
soluble calcium salts reduced pH, increased ionic calcium and caused coagulation to occur. Calcium
chloride showed the largest destabilising effect, followed by calcium lactate and calcium gluconate. Milk
became unstable to UHT processing at lower calcium additions compared to in-container sterilisation.
Keywords Calcium salts, Heat stability, Ionic calcium, pH.

calcium addition and removal (Geerts et al. 1983;

INTRODUCTION
Tsioulpas et al., 2007).
Milk is an important source of dietary calcium, The literature relating to heat stability of milk is
with estimates from 40% up to 74% of calcium vast; most of it is devoted to measuring heat coagu-
coming from dairy foods. People who avoid dairy lation time or heat coagulation temperature. How-
foods usually have an inadequate dietary intake, ever, the correlation between heat coagulation time
unless they have an adequate intake of other high and stability of milk toward commercial sterilisa-
calcium dietary foods (Wark and Nowson 2003). tion processes is often poor (O’Connell and Fox
At the normal pH of milk, about two thirds of its 2003). It appears that pilot plant or laboratory
calcium is bound in the micelle, predominantly in methods which simulate commercial sterilisation
the form of calcium phosphate nanoclusters (Wal- conditions are better and more practical predictors
stra and Jenness 1984; Holt 2004). About one third of stability of milk, although studies in this area
is soluble calcium, of which between 5 and 10% is have been much fewer. Most papers examine how
present as ionic calcium. Partitioning of calcium heat coagulation time is influenced by adjusting
and other minerals at different pH values has been pH. There is much less information for establishing
the subject of many investigations (Holt et al. whether milk, at its natural pH, will be stable to
1981; Holt 2004): calcium in the soluble phase UHT processing and in-container sterilisation.
increases as milk is reduced from its natural pH Horne and Muir (1990) stated that traditional tests
towards 4.6. Partitioning at different temperatures for alcohol and heat stability proved to be unreli-
has been subject to much less investigation able indicators of stability defects associated with
(Pouliot et al. 1989a,b,c), with soluble calcium season, diet and stage of lactation, and in particular
decreasing as temperature increases. of the suitability of milk for processing into evapo-
Ionic calcium is considered to be an important rated or condensed products.
determinant of casein micelle stability. It has been Calcium supplementation of milk is challenging
well established that there are natural variations in because the soluble phase of milk is considered to
ionic calcium and pH (White and Davies 1958a; be saturated with calcium at its normal pH, making
Chavez et al. 2004; Tsioulpas et al. 2007a,b) and further addition difficult. Calcium chloride addition
that ionic calcium is an important determinant of will tend to reduce pH and increase ionic calcium
ethanol stability (White and Davies 1958b; Lin (Philippe et al. 2003; Boumpa et al. 2008), which
et al. 2006; Tsioulpas et al. 2007a). Overall, the in turn will shift the heat coagulation–pH profile in
*Author for correlation between pH and ionic calcium for dif- a more alkaline direction (Sievanen et al. 2008).
correspondence. E-mail:
ferent milk samples is poor. However, for an indi- Calcium gluconate and calcium lactate have been
[email protected]
vidual milk sample, there is a good relationship less studied, but have been shown to have similar
 2010 Society of between pH and ionic calcium, if that milk is sub- effects (Singh et al. 2007). Calcium carbonate has
Dairy Technology jected to pH change (by whatever means) or to been shown to be an effective method of calcium

504 Vol 63, No 4 November 2010 International Journal of Dairy Technology

Vol 63, No 4 November 2010

supplementation in tablet form. However, it is Table 2, which were measured directly at the high
insoluble and its properties have been less investi- temperatures.
gated in milk and it was rejected for milk supple- Freezing point depression (FPD) of milk and
mentation by Singh et al. (2007). dialysates was measured using an Advance milk
Ethanol stability is taken as a good indicator of cryoscope 4L2 (Advance Instruments, Inc., New
milk stability toward UHT processing, (Shew Jersey, USA). Results are expressed in terms of
1981; Horne 2003). Further evidence for this is thousandths of a degree (C), a reading of 460 indi-
provided by Boumpa et al. (2008). However, cating a freezing point of )0.460 C.
there is less evidence for this during in-container Ethanol stability of milk samples were measured
sterilisation. When milk is heated its pH will fall, by mixing equal volumes (2 mL) of milk samples
but on cooling it is reversible following pasteuri- and ethanol solution (70 and 80% v ⁄ v) in a test
sation and UHT treatment, but not after sterilisa- tube. All measurements were done at room temper-
tion. It is proposed that an important determinant ature (20 ± 0.5 C). Results were recorded as
of heat stability might be the pH and ionic cal- being stable or unstable.
cium values at the high temperatures achieved Free ionic calcium was measured by the Ciba
during heat treatment. Overall, the vast majority Corning 634 Analyser. The instrument was cali-
of papers that deal with heat stability, using heat brated before each use. A calibration curve was
coagulation times, do not consider the changes in prepared by using six solutions of known Ca++
pH that take place on heating milk to 140 C or concentration (0.5, 1.0, 1.5, 2.0, 2.5 and 3.0 mM).
what those pH values might be. The same can A calibration curve was produced by plotting the
also be said for changes in ionic calcium. On- log of Ca2 + ion concentration in mM and the read-
Nom et al. (2010) have recently measured pH and ing of the potential difference given by the elec-
ionic calcium in milk up to 120 C and showed trode (mV), (Lin et al. 2006).
that both pH and ionic calcium fell substantially Total divalent cations, ie calcium plus magne-
as temperature increased. sium, were measured by the EDTA titration. This
The aim of this study is to evaluate the effects of is a rapid method for measuring total calcium and
different calcium salts on pH and ionic calcium magnesium in milk (Davies and White 1962; Wal-
and on various casein micelle stability parameters, stra and Jenness 1984). The procedure has more
namely ethanol stability, and heat stability toward recently been described by On-Nom et al. (2010).
in-container and UHT sterilisation. Ionic calcium Buffering capacity was measured for milk with
and pH of calcium supplemented milk samples added calcium salts, by measuring pH changes
have also been measured at 100 C, using dialysis. caused by incremental additions of 1 M HCl to
150 mL of milk, and also by addition of a fixed
weight (0.4%) of glucono delta lactone (GDL).
M AT E R I A L S A N D M E T H O D S
Milk samples UHT processing
Milk samples were collected from bulk raw milk A tubular UHT plant ((UHTAC Cooktube généra-
or pasteurised semi skimmed milk from Dairy tion 2, France) was used. Heating was obtained in
Crest, Surrey. two stages, using two oil baths. The first oil bath
temperature was set at 100 C, giving a preheat
Calcium addition temperature of about 90 C and the second oil bath
Six calcium salts [calcium chloride (CCl), calcium temperature was set at 150 C. The pump speed
carbonate (CCarb), calcium phosphate (CPh), cal- was constant (24 machine units) for all the experi-
cium citrate (CCit), calcium lactate (CLac) and cal- ments. Operating conditions for the UHT plant
cium gluconate (CGlu), AnalaR from BDH)] were were 140 C and an average residence time of
added to milk to increase total calcium by 30 mM. 5.6 s. The flow rate of the plant was checked regu-
These samples were well stirred and left for at least larly by measuring product collected in a measur-
1 h to ensure equilibration. ing cylinder for 1 min. The product was cooled to
In subsequent experiments, CGlu and CLac 20 C. UHT milk samples were collected in
were then added over progressively narrower con- labelled sterile Sterilin containers and stored in the
centration ranges (0–10 mM and 3–6 mM). refrigerator (5 C). Sediment formation was mea-
sured by centrifuging the sample at 2760 g and
Physicochemical analysis determining the dry weight of the sediment
The pH of milk, dialysate and permeate samples (Boumpa et al. 2008).
were measured by a HI8424 Digital pH meter. The
calibration of the pH probe is using standard solu- In-container sterilisation
tion of pH 4 and 7 which is obtained from VWR Milk samples were poured into metal cans and
International Ltd, Poole, UK. Most readings were sealed using the vacuum sealer. Sterilisation condi-
taken at 20 C, except for those presented in tions were 115 C for 15 min; the come-up time

 2010 Society of Dairy Technology 505

 Vol 63, No 4 November 2010

was about 5 min. The corresponding pressure for 80% ethanol. In contrast, addition of CCI, CLac
the required temperature was 0.7 bar (gauge). At and CGlu reduced pH, increased ionic calcium
the end of the process, the steam valve was closed substantially and increased FPD. Ethanol stability
and compressed air and cooling water were turned was also reduced to below 70%.
on with the aim of maintaining pressure then Using the EDTA titration method, it was not
allowing it to fall slowly. Cans were left to cool in possible to detect changes in total divalent cations
the retort for about 5 min and then removed and due to CCI and CPh additions. In contrast, more
stored in the refrigerator (5 C). clearly proportional changes due to CCarb, CLac
and CGlu additions were detected, but values mea-
Dialysis sured for each were slightly different. For example
Dialysis tubing was Visking tubing (size 9 ID the titration value for CGlu was greater than for
28.6 mm) supplied by Medicell International Ltd, CLac. Results for CCit were also interesting, with
London, UK. Dialysis was performed at 20 C for the added calcium being only partially detected.
24 h and at 100 C for 1 h, using methods These observations were reproducible for different
described by On-Nom et al. (2010). Dialysis bags batches of milk.
were removed as quickly as possible while the
samples were still at high temperature. Dialysates Buffering capacity of calcium salts
were measured once they had cooled to 20 C. Buffering capacity was assessed by measuring the
change in pH brought about by fixed amounts of
HCl or GDL. From Tables 2 and 3, it can be seen
R E S U LT S
that addition of 30 mM CCarb, CCit and CPh did
Addition of 30 mM of six different calcium not change the pH response to acid addition, or to
salts GDL addition, compared to the control. Buffering
The results for addition of 30 mM calcium to milk capacities were also determined for the three solu-
using the different salts followed two distinct pat- ble calcium salts, but the results are not so easy to
terns (Table 1). Adding CCarb, CPh and CCit interpret as they are not measured over the same
caused no noticeable changes in pH, ionic calcium pH range. It was found that there were no signifi-
and FPD. The milk samples remained stable in cant differences in the amounts of alkali required

Table 1 Addition of 30 mM of different calcium salts to milk (replicated for three different milk samples*)

Freezing point depression Ionic calcium (mM) Total divalent cations

Milk sample pH (mean ± SD) (fpd) (mºC) (mean ± sd) (mM) (mean ± SD)
Control 6.73 ± 0.04a 0.514 ± 0.00a 1.37 ± 0.22a 34.60 ± 0.69a
Calcium chloride† 6.09 ± 0.01b 0.638 ± 0.00b 9.86 ± 1.46b 63.03 ± 0.91b
Calcium lactate† 6.31 ± 0.03c 0.608 ± 0.00c 5.69 ± 0.39c 59.10 ± 0.70c
Calcium gluconate† 6.34 ± 0.04c 0.609 ± 0.02c 6.95 ± 0.70d 71.40 ± 2.13d
Calcium carbonate 6.73 ± 0.04a 0.510 ± 0.00a 1.35 ± 0.24a 34.67 ± 1.04a
Calcium citrate 6.73 ± 0.04a 0.514 ± 0.00a 1.37 ± 0.24a 47.50 ± 3.53e
Calcium phosphate 6.74 ± 0.04a 0.513 ± 0.00a 1.36 ± 0.22a 34.90 ± 1.05a
Mean ± SD in the same column followed by the same letter are not significantly different (P > 0.05).
*Original ionic calcium values ranged between 1.20 and 1.62 mM.
†
These samples were unstable in 70% ethanol.

Table 2 Measurement of buffering capacity by addition of 1 M HCl to 150 mL of calcium supplemented (30 mM) milk
samples (replicated for three different samples of milk)

1 M HCl conc. (mL)

Sample 0.0 0.5 1.0 1.5 2.0 2.5
a a a a a
Control 6.71 ± 0.04 6.44 ± 0.05 6.18 ± 0.07 5.95 ± 0.07 5.74 ± 0.07 5.57 ± 0.07a
Calcium chloride 6.02 ± 0.07b 5.79 ± 0.08b 5.56 ± 0.09b 5.39 ± 0.09b 5.27 ± 0.10b 5.16 ± 0.10b
Calcium lactate 6.29 ± 0.08c 6.04 ± 0.06c 5.81 ± 0.06c 5.62 ± 0.08c 5.50 ± 0.07c 5.38 ± 0.08c
Calcium gluconate 6.35 ± 0.09c 6.10 ± 0.05c 5.83 ± 0.03c 5.68 ± 0.02c 5.52 ± 0.02c 5.41 ± 0.03c
Calcium carbonate 6.72 ± 0.06a 6.46 ± 0.06a 6.21 ± 0.06a 5.96 ± 0.05a 5.72 ± 0.04a 5.55 ± 0.04a
Calcium citrate 6.74 ± 0.07a 6.45 ± 0.05a 6.20 ± 0.02a 5.95 ± 0.02a 5.73 ± 0.04a 5.56 ± 0.02a
Calcium phosphate 6.76 ± 0.08a 6.48 ± 0.08a 6.24 ± 0.10a 6.02 ± 0.11a 5.78 ± 0.11a 5.61 ± 0.12a
Mean ± SD in the same column followed by the same letter are not significantly different (P > 0.05).

506  2010 Society of Dairy Technology

Vol 63, No 4 November 2010

of calcium supplemented milk samples to 70%.

Table 3 Measurement of the buffering capacity of
For example, 4.5% TSC (153 mM) was required
calcium supplemented (30 mM) milk samples by addition
of 0.4% GDL (repeated for three batches of milk)
for CCl, which is well in excess of levels usually
used. Amounts for DSHP were even higher in per-
Milk sample pH (0.4% GDL addition) centage terms and resulted in destabilisation of the
Control 5.75 ± 0.06a casein micelle, presumably because of calcium
Calcium chloride 5.38 ± 0.12b chelation and consequential removal of calcium
Calcium lactate 5.57 ± 0.07c and phosphate from the micelle.
Calcium gluconate 5.60 ± 0.04c
Calcium carbonate 5.75 ± 0.06a Heat treatment of milk
Calcium citrate 5.73 ± 0.04a When milk samples containing the six calcium
Calcium phosphate 5.77 ± 0.09a salts (30 mM) were subjected to in-container steril-
Mean ± SD in the same column followed by the same isation, those containing CCl, CLac and CGlu
letter are not significantly different (P > 0.05). showed poor heat stability, producing a firm coagu-
lum. Addition of these salts decreased pH and
increased ionic calcium, both of which would con-
to restore pH, following acid addition (data not pre- tribute to the reduction in heat stability (Table 5).
sented). Also, all milk samples which were stable In contrast, milk samples with CCarb, CPh and
in 70% ethanol prior to acidification became unsta- CCit were stable, although small amounts of sedi-
ble to it after restoration of their pH. Ionic calcium ment were produced. As mentioned, addition of
values were also higher in the pH restored milks, CCarb and CPh caused no change in milk pH,
showing that the process was not reversible. For ionic calcium and freezing point depression. CCit
example it increased from 1.2 mM to 1.52 mM for addition resulted in slightly higher ionic calcium.
the control and from between 1.47 mM to In-container sterilisation brought about an irrevers-
1.62 mM for these insoluble calcium salts. ible decrease in pH of all samples.
Data are also presented for direct measurement Milk samples containing 30 mM of CCarb, CPh
of pH at different temperatures (Table 4). The and CCit were subjected to UHT treatment at
results showed that when milk temperature is 140 C for 5.6 s. In all cases, this caused no
increased, there is a considerable reduction in pH. changes in pH and slightly reduced ionic calcium.
This has been reported in the literature by Walstra No sediment was detected by taste in any of these
and Jenness (1984), Ma and Barbano (2003) and UHT samples and in the relatively short UHT pro-
most recently by On-Nom et al. (2010). There was cessing run, the UHT temperature remained con-
no significant difference in pH ⁄ temperature rela- stant, suggesting that there were no signs of
tionship for the different insoluble calcium salts. fouling. Milk samples containing 30 mM added
Measuring temperature dependence could be CCl, CGlu and CLac were not subjected to UHT
regarded as an alternative method of measuring the treatment. This was because previous experience
buffering capacity of milk. It is proposed that buf- Boumpa et al. (2008) had shown that much lower
fering capacity is of interest because a milk having levels than this produce excessive fouling and sedi-
a higher buffering capacity might also show higher ment formation. All these milks were also unstable
heat stability, since its pH would not fall so much in 70% ethanol.
during the heating process. Overall, these results have shown that CCarb
Data was also collected for the amounts of triso- and CPh are insoluble in milk, since their addition
dium citrate (TSC) and disodium hydrogen phos- does not affect pH, ionic calcium and freezing
phate (DSHP) required to restore ethanol stability point depression. CCit is also close in behaviour to

Table 4 pH of milk with added calcium salts (30 mM) at different temperatures (repeated for three batches of milk)

Temperature (ºC)
Sample 20 40 60 80
Control 6.70 ± 0.05a 6.53 ± 0.08a 6.31 ± 0.13a 6.11 ± 0.17a
Calcium chloride 6.00 ± 0.12b 5.69 ± 0.15b 5.32 ± 0.20b 5.02 ± 0.25b
Calcium lactate 6.25 ± 0.11c 5.91 ± 0.12c 5.55 ± 0.16b 5.21 ± 0.21b
Calcium gluconate 6.34 ± 0.17c 5.95 ± 0.13c 5.54 ± 0.14b 5.24 ± 0.28b
Calcium carbonate 6.68 ± 0.67a 6.51 ± 0.06a 6.28 ± 0.10a 6.06 ± 0.14a
Calcium citrate 6.70 ± 0.08a 6.50 ± 0.08a 6.27 ± 0.11a 6.03 ± 0.13a
Calcium phosphate 6.72 ± 0.08a 6.51 ± 0.06a 6.28 ± 0.10a 6.05 ± 0.14a
Mean ± SD in the same column followed by the same letter are not significantly different (P > 0.05).

 2010 Society of Dairy Technology 507

 Vol 63, No 4 November 2010

Table 5 Changes in pH, ionic calcium and fpd following in-container sterilisation and UHT processing (one trial only)

Before heating UHT treatment In-container sterilisation

2 + 2 +
Sample pH Ca (mM) fpd (mC) pH Ca (mM) fpd (mC) pH Ca2 +
(mM) fpd (mC)
Control 6.74 1.68 516 6.71 1.57 512 6.57 1.47 515
Calcium chloride 6.11 nm 643 ns ns ns 5.87 nd nd
Calcium lactate 6.37 6.87 606 ns ns ns 6.13 nd nd
Calcium gluconate 6.63 6.63 545 ns ns ns 6.33 nd nd
Calcium carbonate 6.76 1.68 514 6.73 1.57 512 6.57 1.44 515
Calcium citrate 6.77 1.68 514 6.69 1.68 512 6.53 1.54 517
Calcium phosphate 6.76 1.67 517 6.75 1.53 512 6.56 1.44 514
All readings are the average of three measurements; standard deviations are very small. ns – no UHT samples were produced as ethanol stability was
below 70%; nd – not determined as sample had coagulated; nm: not measured.

these two salts. In contrast, the other calcium salts with low ethanol stability had previously blocked
are more soluble and reduce pH and increase ionic the tubes in the high temperature section and hold-
calcium, which resulted in poor heat stability. ing tube of the UHT plant.
It was also observed that UHT conditions used in CGlu and CLac addition were investigated over
this experiment caused very little change in pH of a narrower range (0–10 mM). Milk samples were
milk, whereas in-container sterilisation resulted in a designated as being stable in ethanol solutions (70
slightly lower pH. Although this has been reported and 80%), as being stable or unstable to UHT and
to be the case for in-container sterilised milk, there in-container sterilisation, determined by an increase
is little documented information about changes in the sediment level or observed signs of coagula-
of pH resulting from UHT treatment of milk. tion. The results for CGlu and CLac additions are
presented in terms of these stability characteristics
Dialysis of milk samples (Figure 1a–c). Increasing both salts led to a reduc-
Milk containing 30 mM of all the investigated cal- tion in pH and an increase in ionic calcium. For
cium salts were subjected to the dialysis procedure, each stability parameter, a point was reached when
at 20 C for 24 h and 100 C for 60 min. Milk the milk shifted from being stable to unstable for
containing CCI, CLac and CGlu coagulated during ethanol and both heat stability parameters. It is
this procedure at 100 C, whereas the other sam- worth mentioning that similar shifts in pH and
ples remained stable. The composition of dialy- ionic calcium would occur as raw milk becomes
sates taken at these two temperatures is shown in more acidic due to microbial activity.
Table 6. It was observed that both pH and ionic It was observed that samples were more stable
calcium were significantly lower for samples dialy- to in-container sterilisation than to UHT process-
sed at 100 C, compared to 20 C. On-Nom et al. ing. Stated another way, it required a greater
(2010) have shown that the pH and ionic calcium addition of CGlu and CLac to cause instability
values for dialysates taken at high temperature are during in-container sterilisation compared to
themselves not temperature dependent. Thus even UHT processing. This was a consistent finding
though pH and ionic calcium are measured at throughout these studies. Also, in molar terms,
20 C, they are similar to values in milk at 100 C. milk could tolerate a greater amount of calcium
Data in Table 6 also contrast the behaviour of solu- in its lactate form compared to its gluconate
ble and insoluble calcium salts and shows that the form, before becoming unstable to heat. Further
milk containing soluble salts have a much lower batches of milk were subjected to similar treat-
pH and a much higher ionic calcium than soluble ments over narrower ranges of CGlu and CLac
salts, both at 20 C and 100 C. Also there are addition. The milk used in these trials showed
considerable differences between the soluble cal- differences in their ionic calcium and pH values.
cium salts in terms of their pH and ionic calcium The transition from stability to instability for
when added at equal calcium molarities. each of the parameters was determined and
expressed as a range for each instability parame-
Addition of calcium gluconate and calcium ter (Table 7). For pH and ionic calcium, the two
lactate to milk values represent the range over which the prod-
Additions of 30 mM of CGlu and CLac resulted in uct changed from being stable to unstable. It is
a reduction in pH and increase in ionic calcium not easy to define the boundaries between stabil-
(Table 1). Samples were unstable in 70% ethanol ity and instability: this is not surprising, consid-
and they produced a thick coagulum when steri- ering the range of milk compositions.
lised at 115 C for 15 min. As mentioned, they Two interesting observations were that ionic cal-
were not subjected to UHT treatment, as products cium levels required for producing milk which was

508  2010 Society of Dairy Technology

 Vol 63, No 4 November 2010

Table 6 Measurement of pH and ionic calcium in dialysates collected at 20C and 100C (one trial only)

Sample pH20 pH100 Ca2 +

20 (mM) Ca2 +
100 (mM)
Control 6.73 6.41 1.19 0.26
Calcium chloride 6.13 5.80 14.8 11.2
Calcium lactate 6.26 5.88 9.79 7.43
Calcium gluconate 6.48 5.91 8.10 5.43
Calcium carbonate 6.76 6.41 1.08 0.27
Calcium citrate 6.73 6.38 1.07 0.41
Calcium phosphate 6.77 6.42 1.11 0.28
*All measurements were taken at 20C.

(a) Stability in 70% ethanol

that these levels were higher than those required to
3 cause stability to UHT processing.
Ionic calcium concentration (mM)

2.8 CG stable
2.6 CG unstable
CL stable
2.4 DISCUSSION
CL unstable
2.2
2 Overall, statistical analysis of the results presented
1.8 in Tables 1–4 show three distinct types of behav-
1.6
iour: that observed by the control and the three
1.4
1.2 insoluble salts; that observed by CLac and CGlu;
1 and that observed by CCl. There is a shortage of
6.55 6.6 6.65 6.7 6.75 6.8
pH information comparing the effects of adding insol-
(b) uble and soluble calcium salts to milk on its heat
3 Stability during UHT processing stability. One apparent benefit of adding insoluble
Ionic calcium concentration (mM)

2.8
salts is that that they do not change pH or ionic cal-
2.6 CG stable
2.4 CG unstable
cium, in contrast to soluble salt addition. One
2.2 CL stable potential drawback arising from their low solubility
CL unstable
2 is a detrimental effect on the mouthfeel, in terms of
1.8 grittiness, and that they may not stay suspended in
1.6
the milk. This was not observed from our results.
1.4
1.2
Thus, although no adverse short-term effects were
1 noticed in UHT milk supplemented with these
6.55 6.6 6.65 6.7 6.75 6.8 salts, longer term effects during storage still need
pH
(c) further investigation. However, their use in pasteur-
3 Sability during in-container sterilisation ised products should not pose a problem because
Ionic calcium concentration (mM)

such milks are consumed within two weeks of pro-

CG stable
2.5 CG unstable
cessing. Their apparent success in terms of not
CL stable impairing heat stability arises from their low solu-
2
CL unstable
bility. There is no obvious reason why they should
not be fully utilised as calcium supplements, as
1.5 CCarb is a popular calcium source (Wark and
Nowson 2003). CCarb, CPh and CCit do not
1
appear to alter the buffering capacity (Table 2).
6.55 6.6 6.65 6.7 6.75 6.8 The average reduction in pH on addition of 2 mL
pH
HCl was 0.97 for the control, compared to 1.00,
Figure 1 Showing changes in pH and ionic calcium resulting from addition of increasing
1.01 and 0.97 for CCarb, CCit and CPh respec-
levels of calcium gluconate and calcium lactate. (a) Stability in 70% ethanol; (b) stability
tively. Factors affecting buffering capacity of milk
during UHT processing; (c) stability during in-container sterilisation.
have been discussed by Walstra and Jenness
(1984) and Salaun et al. (2005).
unstable to UHT was similar to that required for There is little published information about ionic
ethanol stability of 70% and that many milk sam- calcium measurements in milk at high temperatures.
ples which were showed instability toward UHT Results obtained using ultrafiltration and dialysis
processing were stable to in-container sterilisation. suggest that ionic calcium decreases as temperature
From the work of On-Nom et al.(2010), it was increases and that the relationship is linear (On-
observed that 5.4 mM addition of CCl would cause Nom et al. 2010). It was found that changes in pH
destabilisation during in-container sterilisation, but are reversible for milk and for UF permeate

 2010 Society of Dairy Technology 509

 Vol 63, No 4 November 2010

Table 7 pH and ionic calcium values in the transition* region between stable and unstable products (four separate samples of milk were evaluated)

Ethanol (70%) stability Sterilisation heat stability UHT heat stability

2 + 2 + 2 +
Sample Initial pH Initial Ca (mM) pH Ca (mM) pH Ca (mM) pH Ca2 +
(mM)
1 6.80* 1.84* 6.73–6.69 1.97–2.39* 6.72–6.66 2.21–2.39 6.76–6.69 1.97–2.32
2 6.76 1.26 6.73–6.70 1.47–1.87 6.70–6.63 1.87–2.31 6.72–6.65 1.52–1.99
3 6.73 1.33 6.65–6.62 1.73–1.87 6.63–6.61 1.73–1.80 6.65–6.61 1.79–1.87
4 6.72 1.62 6.63–6.58 2.04–2.28 6.58–6.56 2.36–2.40 6.64–6.60 1.98–2.25
*The first of the two values represents the highest value at which the product was stable, whereas the second of the two represents the lowest value that it
was unstable (for respective parameters).
UHT milk samples were deemed to be unstable when the sediment exceeded 1% (dwb), where it was also detectable by tasting, as a gritty mouthfeel.
In-container sterilised products were deemed to be unstable when coagulation could by visibly observed.

collected at 20 C, but not for those collected at further processing or not. Good quality of milk
80C. Results from our studies show that ionic cal- should show no flocculation in 80% ethanol solu-
cium at 100 C is less than one quarter of its value tion and milk should be stable in 74% ethanol to
at 20 C. It is known that pH of milk decreases as be stable to UHT processing (Shew et al. 1981).
its temperature increases. For pasteurisation and
UHT treatment of milk, this is fully reversible on
REFERENCES
cooling, whereas this is not the case for in-container
sterilisation, with the pH of the sterilised milk being Boumpa T, Tsioulpas A, Grandison A and Lewis M (2008)
about 0.2 pH units lower than the original milk. Effects of phosphate and citrates on sediment formation
What is less well understood is how milk composi- in UHT goat’s milk. Journal of Dairy Research 75 160–
tion affects the change in both pH and ionic calcium 166.
on heat treatment and the result of this change on Browning E, Lewis M and MacDougall D (2001) Predicting
safety and quality parameters for UHT-processed milks.
the heat stability of that milk. It was interesting that
International Journal of Dairy Technology 54 111–120.
the average change in pH, for data presented in
Chavez M S, Livia M N, Taverna M A and Cuatrin A (2004)
Table 4, over the range temperature range 20 C to
Bovine milk composition parameters affecting the ethanol
80 C was )0.0097 C)1 for the control, compared stability. Journal of Dairy Research 71 201–206.
to )0.0103, )0.0111 and )0.011 C)1 for CCarb, Davies D T and White J C D (1962) The determination of cal-
CCit and CPh respectively. This shows that addition cium and magnesium in milk and milk difusate. Journal
of insoluble calcium salts results in only a very of Dairy Research 29 285–296.
small change in the relationship between pH and Geerts J, Bekhof J and Scherjon J (1983) Determination of
temperature. This is worthy of further investigation. calcium ion activities in milk with an ion selective elec-
It is known that there are variations in ionic cal- trode. Netherlands Milk and Dairy Journal 37 197–211.
cium concentration and pH in raw cow’s milk and Grimley H, Grandison A and Lewis M (2009) Changes in
even in bulk milks collected for processing (White milk composition and processing properties during the
and Davies 1958a; b; Chavez et al. 2004; Grimley spring flush period. Dairy Science and Technology 89
et al. 2009). Milk with a high ionic calcium concen- 405–416.
tration and lower pH will show more susceptibility Holt C (2004) An equilibrium thermodynamic model of the
sequestration of calcium phosphate by casein micelles and
to coagulation in general. Such milk may best be
its application to the calculation of the partition of salts in
suited for production of cheese or other fermented
milk. European Biophysics Journal 33 421–434.
products; its RCT is likely to be shorter (White and
Holt C, Dalgleish D G and Jenness R (1981) Calculation of
Davies 1958b; Tsioulpas et al. 2007), and its curd the ion equilibria in milk diffusate and comparison with
firmness may be stronger and it may even require experiment. Analytical Biochemistry 113 154–163.
less chymosin for coagulation. However, it is also Horne D (2003) Ethanol stability. In Advanced Dairy Chemis-
likely to be much more susceptible to heat induced try, Vol. 1, Proteins, 3rd edn, pp 975–1000. Fox P F, McS-
coagulation; and would be best avoided in UHT weeney P L H, eds. New York: Kluwer Academic ⁄ Plenum
processes and in production of heat stable milk Publishers.
powder. Experimental evidence has been provided Horne D S and Muir D D (1990) Alcohol and heat stability of
in this paper to support this viewpoint. milk protein. Journal of Dairy Science 73 3613–3626.
Milk may have a low ethanol stability due to Lin M-J, Lewis M J and Grandison A S (2006) Measurement
poor microbial quality, in which case it may be of ionic calcium in milk. International Journal of Dairy
accompanied by a low pH, or due to imbalances in Technology 59 192–199.
its salt balance, which also leads to poor heat sta- Ma Y and Barbano D M (2003) Milk pH as a function of CO2
concentration, temperature, and pressure in a heat exchan-
bility (Horne 2003). The ethanol stability test is
ger. Journal of Dairy Science 86 3822–3830.
used to examine that milk has suitable quality for

510  2010 Society of Dairy Technology

Vol 63, No 4 November 2010

O’Connell J E and Fox P F (2003) Heat-induced coagulation unconcentrated and concentrated bovine milk. Interna-
of milk. In Advanced Dairy Chemistry, Volume 1, Pro- tional Journal of Dairy Technology 61 151–155.
teins, 3rd edn, pp 879–945, Fox P F, McSweeney P L H Singh G, Arora S, Sharma G S, Sindhu J S, Kansal V K and
eds. New York, USA: Kluwer Academic ⁄ Plenum Pub- Sangwan R B (2007) Heat stability and calcium bioavail-
lishers. ability of calcium fortified milk. LWT – Food Science and
On-Nom N, Grandison A S and Lewis M J (2010) Measure- Technology 40 625–631.
ment of ionic calcium, pH and soluble divalent cations in Tsioulpas A, Lewis M and Grandison A (2007a) Effect of
milk at high temperature. Journal of Dairy Science 93 minerals on casein micelle stability of cows’ milk. Journal
515–523. of Dairy Research 74 167–173.
Philippe M, Gaucheron F, Le Graet Y, Michel F and Garem A Tsioulpas A, Grandison A S and Lewis M J (2007b) Changes
(2003) Physicochemical characterisation of calcium sup- in physical properties of bovine milk from the colostrum
plemented skim milk. Lait, 83 45–59. period to early lactation. Journal of Dairy Science 90
Pouliot Y, Boulet M and Paquin P (1989a) Observations on 5012–5017.
the heat induced salt balance changes in milk 1 effect of Walstra P and Jenness R (1984) Dairy Chemistry and Physics.
heating time between 4 and 90 C. Journal of Dairy New York: John Wiley and Sons.
Research 56 185–192. Wark J D and Nowson C (2003) Calcium supplementation:
Pouliot Y, Boulet M and Paquin P (1989b) An experimental the bare bones. Australian Prescriber 26 126–127.
technique for the study of milk salt balance. Journal of White J and Davies D (1958a) The relation between the
Dairy Science 72 36–40. chemical composition of milk and the stability of the
Pouliot Y, Boulet M and Paquin P (1989c) Observations on caseinate complex. Journal of Dairy Research 25 236–
the heat induced salt balance changes in milk. Journal of 255.
Dairy Research 56 193–199. White J and Davies D (1958b) The relation between the
Salaun F, Mietton B and Gaucheron F (2005) Buffereing chemical composition of milk and the stability of the case-
capacity of dairy products. International Dairy Journal 15 inate complex, II coagulation by ethanol. Journal of Dairy
95–109. Research 25 255–266.
Shew D (1981) International Dairy Federation. Bulletin, Williams R, D’Ath L and Augustin M (2005) Production of
New monograph on UHT milk, document 133, Brus- calcium-fortified milk powders using soluble calcium
sels, 115 salts. Lait 85 369–381.
Sievanen K, Huppertz T, Kelly A L and Fox P F (2008) Influ-
ence of added calcium chloride on the heat stability of

 2010 Society of Dairy Technology 511