This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/authorsrights
Author's personal copy
Ultrasonics Sonochemistry 21 (2014) 951–957
Contents lists available at ScienceDirect
Ultrasonics Sonochemistry
journal homepage: www.elsevier.com/locate/ultson
Impact of ultrasound treatment on lipid oxidation of Cheddar cheese
whey
Amir Ehsan Torkamani a,b, Pablo Juliano a,⇑, Said Ajlouni b, Tanoj Kumar Singh a
a
b
CSIRO Division of Animal, Food and Health Sciences, 671 Sneydes Road, Werribee, VIC 3030, Australia
Department of Agriculture and Food Systems, Melbourne School of Land and Environment, The University of Melbourne, Parkville, VIC 3010, Australia
a r t i c l e
i n f o
Article history:
Received 25 June 2013
Received in revised form 26 November 2013
Accepted 30 November 2013
Available online 8 December 2013
Keywords:
Ultrasound
Whey
Lipid oxidation
Phospholipids
a b s t r a c t
Ultrasound (US) has been suggested for many whey processing applications. This study examined the
effects of ultrasound treatment on the oxidation of lipids in Cheddar cheese whey. Freshly pasteurized
whey (0.86 L) was ultrasonicated in a contained environment at the same range of frequencies and energies for 10 and 30 min at 37 °C. The US reactor used was characterized by measuring the generation of
free radicals in deionized water at different frequencies (20–2000 kHz) and specific energies
(8.0–390 kJ/kg). Polar lipid (PL), free and bound fatty acids and lipid oxidation derived compounds were
identified and quantified before and after US processing using high performance liquid chromatography
equipped with an evaporative light scattering detector (HPLC–ELSD), methylation followed by gas chromatography flame ionized detector (GC-FID) and solid phase micro-extraction gas chromatography mass
spectrometry (SPME-GCMS), respectively. The highest concentration of hydroxyl radical formation in the
sonicated whey was found between 400 and 1000 kHz. There were no changes in phospholipid composition after US processing at 20, 400, 1000 and 2000 kHz compared to non-sonicated samples. Lipid oxidation volatile compounds were detected in both non-sonicated and sonicated whey. Lipid oxidation was
not promoted at any tested frequency or specific energy. Free fatty acid concentration was not affected by
US treatment per se. Results revealed that US can be utilized in whey processing applications with no negative impact on whey lipid chemistry.
Ó 2013 Elsevier B.V. All rights reserved.
1. Introduction
Ultrasound (US) technology has many potential applications in
the dairy industry such as homogenization, crystallization, defoaming with low frequencies between 18 and 100 kHz, and milk fat
separation enhancement with frequencies equal or higher than
400 kHz [1,2]. While low frequency ultrasound promotes predominantly physical effects, applications utilizing high frequencies (in
the range of few hundred to thousand kHz) predominantly promote the production of free radicals, which result from different
degrees of cavity dynamics [3]. The frequency of the ultrasound
influences the total number of bubbles that may be sonochemically
active in the system [4] and will dictate the maximum size reached
by the cavity before a violent collapse [3]. As the frequency is increased, the cavitation bubbles become smaller. Therefore the
force resulting from the implosion of cavitation bubbles is decreased [4–6]. The linear resonance size equation and Blake
Threshold Radius equation specify the boundaries for active bubble
implosion [7].
⇑ Corresponding author. Tel.: +61 3 9731 3276.
E-mail address: Pablo.Juliano@csiro.au (P. Juliano).
1350-4177/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.ultsonch.2013.11.021
Studies have demonstrated that physical effects of ultrasound
at low frequencies can result in viscosity reduction and enhanced
solubility, gel strength, transparency and foaming capacity of whey
protein concentrates (WPC)[8,9]. Chandrapala et al. [10] showed
that sonication at 20 kHz of reconstituted WPC up to 60 min at
31 W, had minor effects on WPC protein structure. The authors reported higher protein denaturation enthalpy, minor secondary
structural changes and no alteration in thiol content [10]. Martini
et al. [11] studied sensory and functional attributes of whey suspensions set at different pH levels and treated at 20 kHz, 15 W
for 15 min. While emulsions generated by violent cavity collapse
were reported to be more stable in the mentioned study, sensory
quality of sonicated whey was not significantly different from control samples [11].
On the other hand, some studies have demonstrated lipid oxidation in US treated milk. For example, Reiner et al. [12] identified
volatile compound generation most likely through US (24 kHz,
2.5–20 min, nominal power 400 W) induced pyrolysis and lipid
oxidation as the main factors causing rubbery flavor in milk. That
was explained by the occurrence of cavitation, leading to local high
temperatures and pressures upon violent collapse of the cavitation
bubbles [13]. Pyrolysis reactions lead to homolytic fission of
O–H bonds and OH, H formation, resulting from intense local
Author's personal copy
952
A.E. Torkamani et al. / Ultrasonics Sonochemistry 21 (2014) 951–957
temperature and pressure conditions caused during bubble collapse [14]. Chouliara et al. [15] also reported oxidative volatile
compounds in raw, thermized and pasteurized milk at various
treatments (24 kHz, 0–16 min, nominal power 200 W). However,
the above mentioned studies on milk have only described sonochemical effects at low frequencies, and little or no published literature has described the effects of ultrasound in the megasonic
range (400–2000 kHz) in view of other potential applications.
The literature search indicated that none of the previous studies
have investigated the impact of US on whey quality. This study
explores the effect of ultrasonication on oxidation of lipids in a
range of US frequencies (20–2000 kHz) and specific energies
(8–390 kJ/kg) applied to whey.
2. Materials and methods
2.1. Sample preparation and handling
Fresh whey generated from Cheddar cheese manufacturing was
collected from cheese making trials at the Food Processing Centre
of CSIRO Animal, Food and Health Sciences (CAFHS), Werribee,
Australia. The whey was pasteurized at 73.9 °C for 15 s in a plate
heat exchanger. Pasteurized whey samples were immediately
transferred into 2 L plastic bottles (high density polyethylene,
HDPE) with polystyrene (PS) caps and were stored under frozen
conditions ( 20 °C) until treatments.
Ultrasonic Corporation, Connecticut, USA, nominal power 450 W)
was introduced into the reactor from the top and placed in the center. Rectangular plate transducers (with an active area of
100 mm 100 mm) operating at 400 kHz, 1000 kHz and
2000 kHz (Sonosys Ultraschallsysteme GmbH, Neuenbuerg, Germany, nominal power 520, 500 and 300 W, respectively) were
placed against the reactor wall to produce horizontal pressure
waves. The induced power was controlled from the power generator unit.
Ultrasonication trials were performed in a food grade environment. In addition, the ultrasonic setup was placed in a laminar flow
cabinet. Prior to each trial, frozen pasteurized liquid whey samples
were thawed at 4 °C for 24 h. The vessel was filled with liquid
whey (0.86 L) and was placed in a water bath (equipped with a
stirrer) at 37 ± 1 °C, to ensure fat globules were in liquid state. Prepared whey samples were sonicated at 20, 400, 1000 and 2000 kHz
at a range of treatment times (10 and 30 min) and various specific
power inputs. The US systems were run at 10%, 50%, 100% and also
at another point called ‘‘matched power spot (MPS)’’ (Table 1). The
MPS was established by matching the actual dissipated power in
all reactor systems. Each treatment was conducted in triplicates
and included control samples without US processing. Control and
sonicated whey were immediately transferred into freeze dryer
trays, frozen at 20 °C and then lyophilized (Virtis Genesis 35 EL,
SP Scientific, Warminster, PA, USA).
2.2. Ultrasound reactor and experimental design of whey treatment
2.3. Ultrasonic reactor vessel characterization
The US treatment reactor was made of stainless steel in a rectangular shape. The plates were designed with screw connections
instead of being welded to prevent corrosion of welded junctions,
which otherwise may transfer metal components into the whey
during US processing. The reactor was kept leak-proof by inserting
a food grade gasket on the plate intersections through the connecting points. The reactor was built in the smallest possible dimensions (250 mm 180 mm 80 mm) to fit all transducers tested
and maximize the contact areas between the transducer and whey
sample and the sound pressure in each reactor system.
The treatments were carried out by two types of ultrasound system configurations: (a) a sonotrode and (b) a plate transducer system. The 20 kHz sonotrode (Branson Digital Sonifier, Branson
The reactor was characterized by measuring power input, specific energy, acoustic intensity and free radical formation at different treatment conditions in 18 Mohm deionized water.
2.3.1. Power measurement
The total electrical power consumed by the US systems at different power input settings was measured by a powermeter (Belkin, Playa Vista, CA, USA). The power consumed by the generator
unit at standby mode was measured by keeping the transducer offline. Net power was determined by subtracting these measured
powers from the actual power levels used during treatments.
Table 1
Hydroxyl radical concentration, power and energy input into the ultrasonic reactor vessel at different frequencies (20, 400, 1000 and 2000 kHz) and power output %. Data shown
are the means ± standard deviation of 3 experiments.
Frequency (kHz)
a
Power output (%)
Sound pressure (kPa)
Net electrical power input (Watt)
Specific energy (kJ/kg)
H2O2 concentration (lM)
10 min
30 min
10 min
30 min
20
10
50
92a
100
77
100
153
160
14 ± 1
27 ± 2
70 ± 1
89 ± 2
8±1
15 ± 1
34 ± 2
36 ± 1
25 ± 2
45 ± 2
101 ± 5
108 ± 3
0.2 ± 0.02
0.5 ± 0.04
1.0 ± 0.1
1.0 ± 0.1
400
10
28.5a
50
100
100
150
182
266
31 ± 2
70 ± 2
114 ± 2
223 ± 3
15 ± 1
32 ± 2
46 ± 1
100 ± 2
45 ± 3
95 ± 5
138 ± 3
300 ± 6
3.0 ± 0.2
4.5 ± 0.3
6.5 ± 0.5
9.5 ± 0.8
8.4 ± 0.9
10.2 ± 1.0
47.8 ± 5.0
70.0 ± 7.0
1000
10
17.5a
50
100
93
155
216
302
43 ± 2
70 ± 2
179 ± 3
345 ± 3
13 ± 0.3
36 ± 1
67 ± 1
130 ± 2
38 ± 1
107 ± 3
201 ± 3
390 ± 4
0.1 ± 0.0
4.0 ± 0.3
33.0 ± 2.5
51.0 ± 4.0
1.5 ± 0.08
6.0 ± 0.3
95.0 ± 5
110.5 ± 8
2000
10
20a
50
100
92
148
187
268
42.5 ± 1
70 ± 2.2
126 ± 1
219 ± 3
13 ± 1
32 ± 2
50 ± 2
100 ± 2
38 ± 4
95 ± 4
150 ± 6
300 ± 5
Matched specific energy at each frequency.
0.5 ± 0.02
0.7 ± 0.03
1.0 ± 0.05
3.5 ± 0.2
0.3 ± 0.03
0.7 ± 0.06
1.5 ± 0.1
1.8 ± 0.2
0.7 ± 0.05
1.2 ± 0.08
2.5 ± 0.2
3.5 ± 0.3
Author's personal copy
953
A.E. Torkamani et al. / Ultrasonics Sonochemistry 21 (2014) 951–957
2.3.2. Specific energy
The methodology consisted of monitoring temperature increase
for 300 s by using a digital thermometer. Temperature increase
was interpreted into dissipated power (calorimetric power) and
specific energy.
2.3.3. Acoustic intensity
A needle hydrophone (model HNC-1000, Onda Corp., Sunnyvale,
USA) connected to an oscilloscope (TDS 2022C, Tektronix, Beaverton, USA) was used to receive, digitize and measure US waves and
calculate the sound intensity level (I). The transducer was tuned to
the power input settings described earlier (10%, MPS, 50% and
100%) to sonicate dionized water. Sound signals were monitored
as peak to peak voltage (Vp–p) and the average recording during a
one minute time range was reported. The sound pressure level
was calculated in Pa considering a reference sound intensity of
6.7 10 19 W/m2 in water.
2.3.4. Free radical characterization
A colorimetric method described by Alegria et al. [16] was used
to determine hydrogen peroxide (H2O2) concentrations in a water
matrix media. Initially, two sets of reagents (A: 0.4 M KI, 0.05 M
NaOH and 1.6 10 4 M (NH4)6Mo7O24.4H2O, and B: 0.1 M
KHC8H4O4) were prepared, 1 mL of each reagent was added to a
1-cm cuvette and allowed to react for 1 min. Subsequently, test
samples (1 mL) were added to the cuvette and the mixture was
allowed to react for a further minute. The absorbance was then
measured using a spectrophotometer set at 351 nm (Shimadzu
UV-1201V, Kyoto, Japan). The concentration of peroxide in the test
sample was calculated using the Beer–Lambert law with a molar
extinction coefficient of 26400 M 1 cm 1 and a path length of 1 cm.
2.4. Whey volatile compound profile
2.4.1. Volatile compound profile
Headspace solid phase micro-extraction (SPME) analysis was
used to analyze the volatile compounds in the Cheddar cheese
whey. For SPME analysis, an Agilent Technologies gas chromatograph-mass spectrometer (GCMS) system (GC model 6890 N, MS
model 5975 series B; Agilent Technologies Inc., Palo Alto, CA,
USA) equipped with a Combi PAL robotic auto sampler (CTC
Analytics AG, Zwingen, Switzerland) and a hot splitless injector
(temperature 260 °C, splitless time 1 min; Gerstel, Mülheim,
Germany) was used. The SPME fiber, coated with carboxen/polydimethylsiloxane (CAR/PDMS StableFlex fibre, phase thickness
85 lm, length 1 cm; Supelco, Bellefonte, Pa, USA), was preconditioned prior to analysis at 270 °C for 1.5 h. Samples (1.0 g each)
were placed in 20-mL headspace vials (Grace Davidson Discovery
Sciences, Rowville, VIC, Australia), along with 1 lL of internal standard (IS) solution (2-methyl-3-heptanone, 0.035 mg, and 2-ethyl
butanoic acid, 0.071 mg, per 10 mL of methanol). The autosampler
was operated under the following conditions: pre-incubation time
of 5 min, pre-incubation and absorption temperature of 45 °C,
absorption time of 30 min; desorption time of 8 min (purge after
1 min), and desorption temperature of 260 °C. The whey volatile
compounds were chromatographed on a VF-WAXms column
(30 m length, 0.32 mm i.d., 1.0 mm film thickness, Agilent Technologies Australia, Mulgrave, VIC, Australia) a temperature gradient:
35 °C for 5 min, raised to 225 °C at a rate of 10 °C/min and finally
held at 225 °C for 10 min. Helium was used as carrier gas at constant flow rate of 1.6 mL/min. The MSD conditions were as follows:
capillary direct interface temperature of 280 °C, ionization energy
of 70 eV, mass range of 35–300 amu; EM voltage of autotune
+200 V, with a scan rate of 5.27 sans/s.
Volatile compounds were identified by comparison of mass
spectra with the spectra in the NIST05 database (National Institute
of Standards Technology; United States of America), and linear
retention indices were determined using a set of saturated alkanes
C7 to C22 and spectra of authentic standards. Quantitative analysis
was performed using an internal standard methodology. The odor
active value (OAV) of each volatile compound was calculated by
dividing the concentration of a volatile compound by its threshold
value in aqueous media. Threshold values of aldehydes generated
during different treatments in whey are shown in Table 2.
2.5. Whey powder lipid analysis
2.5.1. Fat extraction
The lipid profile analysis consisted of characterizing PLs and
FFAs of extracted fat from freeze dried whey powders. Total fat
Table 2
Concentration of aldehydes (lg/kg) and their odour threshold value (shown in bracketsB) in whey samples at different frequencies and specific energies after 30 min of
ultrasonication (values are averages of triplicate analysis ± SD of 3 individual runs)C.
Frequency (kHz)
A
Power input (%)
Control
0
20 kHz
10
50
92
100
400 kHz
Specific energy (kJ/kg)
Hexanal (7.5)B
c
(E)-2-hexanal (40)B
d
Heptanal (31)B
e
Nonanal (3.5)B
(E)-2-heptanal (40)B
f
3.0 ± 0.3g
0
171.0 ± 19.0
24.0 ± 5.0
21.0 ± 3.0
6.0 ± 0.6
25.2 ± 1.5
45 ± 2.25
100.5 ± 4.5A
108 ± 3
126.0 ± 23.0b
110.5 ± 12.0b
122.0 ± 11.0b
128.0 ± 13.0b
16.6 ± 6.0d
17.4 ± 4.0d
6.2 ± 1.8d
6.8 ± 1.0d
8.4 ± 2.0e
8.4 ± 1.0e
14.2 ± 0.4e
13.6 ± 0.5e
4.0 ± 1.0f
2.7 ± 0.3f
1.0 ± 0.1f
1.2 ± 0.3f
1.7 ± 0.2g
1.6 ± 0.4g
0.6 ± 0.1g
0.4 ± 0.1g
10
28.5
50
100
45 ± 3
94.5 ± 4.5A
138 ± 3
300 ± 6
197.0 ± 34.0c
187.0 ± 38.0c
111.0 ± 17.0c
182.0 ± 18.0c
26.7 ± 8.0d
27.2 ± 6.0d
6.0 ± 1.2d
23.7 ± 4d
19.0 ± 4.0e
21.0 ± 5.0e
10.0 ± 1.4e
20.0 ± 2.0e
6.0 ± 1.5f
6.5 ± 1.1f
4.2 ± 0.8f
7.0 ± 0.9f
2.7 ± 0.7g
2.9 ± 0.3g
1.6 ± 0.2g
2.6 ± 0.4g
1000 kHz
10
17.5
50
100
37.5 ± 1
106.5 ± 2.5A
201 ± 3
390 ± 4
80.0 ± 8.0a
100.0 ± 19.0a
82.0 ± 6.2a
88.0 ± 8.9a
17.6 ± 0.8d
19.6 ± 3.0d
7.0 ± 0.9d
15.2 ± 2.1d
8.0 ± 1.0e
15.0 ± 5.0e
9.0 ± 1.0e
7.6 ± 0.9e
3.6 ± 0.3f
4.0 ± 0.3f
2.1 ± 0.3f
7.0 ± 1.0f
1.5 ± 0.1g
2.0 ± 0.2g
1.0 ± 0.2g
3.0 ± 0.6g
2000 kHz
10
20
50
100
37.5 ± 4
94.5 ± 4A
150 ± 6
300 ± 5
125.0 ± 9.9c
145.0 ± 18.0c
100.0 ± 3.0c
130.0 ± 32.0c
17.8 ± 3.0d
19.3 ± 3.0d
13.4 ± 2.0d
20.3 ± 3.3d
15 ± 2.0e
17.2 ± 2.0e
10.0 ± 2.3e
12.5 ± 1.4e
3.8 ± 0.3f
5.0 ± 0.5f
2.6 ± 0.2f
4.0 ± 0.6f
2.0 ± 0.2g
2.2 ± 0.2g
1.1 ± 0.1g
2.5 ± 0.2g
Matched specific energy at each frequency.
Means within the column for a given frequency followed by different superscript letters differ significantly at P = 0.05.
Concentration values presented in the brackets are lg/kg sample. OAV is calculated by dividing volatile compound concentration in samples by their threshold value
(presented in brackets).
C
The bolded values represent odour active compounds which could be sensed by human olfactory system (OAV P 1).
a–g
B
Author's personal copy
954
A.E. Torkamani et al. / Ultrasonics Sonochemistry 21 (2014) 951–957
was extracted using a modification of Folch method using an Accelerated Solvent Extractor (ASE). A total of 4 g of whey powder was
mixed and ground with diatomaceous earth. The powder was
placed in a stainless steel cell and placed on the top carousel.
The extraction solvent mixture used was chloroform and methanol
mixture (65%:35% vol%). Extraction was performed at 80 °C and
1500 psi.
2.5.2. Whey PL profile
Aliquots of 1.5 mL of chloroform/methanol mixture containing
extracted fat were transferred into HPLC vials. PL separation and
identification were performed using a modified method described
by Rombaut et al. [17]. The lipids were separated using a Dionex
HPLC system (Ultimate 3000 series, Thermofisher, CA, USA)
equipped with a quaternary HPLC pump, auto sampler, a column
heater and an evaporative light scattering detector (ELSD; ESA
Chromachem, Chelmsford, MA, USA). The HPLC system was interfaced with a workstation running Chromeleon software (version
6.8, Thermofisher, CA, USA).
The lipid extract was chromatographed on a normal phase silica
column (Prevail Silica, 150 mm length, 4.6 mm internal diameter
and 3 lm packing; Grace Davison Discovery Sciences) using a gradient system of following two solvent systems: Solvent A, chloroform/methanol/buffer (1 M formic acid, neutralized to pH 3 with
triethylamine), 87.5:12:0.5 (v/v/v) and solvent B, chloroform/
methanol/buffer, 28:60:12 (v/v/v). Phospholipid class separation
was achieved by a gradient system which involved initially running solvent A (100%) from 0 to 7 min followed by a linear gradient
of 0–100% solvent B in solvent A from 7 to 27 min. Flow rate of mobile phase and injection volume was set at 0.5 mL/min and 25 ll,
respectively. Eluted whey lipid components were detected by ELSD
(nebulizing gas, nitrogen at 1.8 bars and 40 °C; evaporation temperature, 60 °C. Phospholipid classes were identified by using pure
standards (Avanti Polar Lipids, Alabaster, AL, USA) of phosphotidyl
ethanolamine (PE), phosphotidyl inositol (PI), phosphotidyl serine
(PS), phosphotidyl choline (PC) and sphingomyelin (SM). Quantitative analysis of phospholipid classes was achieved by external calibration curve methodology.
2.5.3. Fatty acid composition analysis
The fatty acid profile was characterized using the fatty acid
methyl esterification (FAME) method described in detail by Shen
et al. [18]. Alkaline and acid FAME methods were applied to quantify bound and total fatty acid concentrations, respectively. Free
fatty acid amount was determined by subtracting bound fatty acids
from the total fatty acid value.
FAME analysis was performed by a GC-FID equipped with an
auto sampler (7890A/7693, Agilent technologies, Santa Clara CA,
United States). Aliquotes of 2 lL derivatized samples were
injected (20:1 split) on a BPX 70 column (250 lm id and 30 m long,
SGE Analytical Science, Austin, TX, USA). Helium carrier gas flow
rate was set to 1.5 mL/min. The FID was set at 250 °C and
hydrogen, air and nitrogen flow rates were set at 30, 400 and
25 mL/min, respectively. The GC column was programmed from
60 °C first increasing at a rate of 20 °C/min to 100 °C and then
increasing at rate of 10 °C/min to 180 °C which was kept at this
temperature for 10 min and then increasing at rate of 20 °C/min
to a final temperature of 220 °C. Quantitative analysis of fatty composition was achieved by internal standard (heptadecanoic acid)
methodology.
2.6. Statistical analysis and experimental design
Each set of experiments was repeated at least twice (2 processing trials) and each measurement in each trial was assessed in triplicate. Principal component analysis (PCA) was performed using XL
Stat Pro (Addinsoft, Paris, France) in order to determine any possible data clustering related to each individual ultrasound treatment
with respect to the control (non-sonicated) sample. A two-way
analysis of variance (ANOVA) using General Linear Model at a
95% level of significance (Minitab 16, Minitab Ltd., Coventry, UK)
was performed to estimate differences in concentration of FFA, polar lipids and oxidative volatile compounds between control and
sonicated samples. Due to intrinsic differences between the ultrasound reactor systems at low and high frequencies, the impact of
each system on whey parameters is considered in isolation and
comparisons of the sonicated samples are only made with the
equivalent control (non-sonicated) samples. Tukey’s honestly significant difference test was used to separate means at 95% confidence level.
3. Results and discussion
3.1. Reactor vessel characterization
As described earlier, the vessel was assembled from rectangular
stainless plates to form a rectangular box. This approach allowed
the transducers to be easily fitted inside the vessel and face reflectors with similar shape and dimensions. Power and energy values
are illustrated in Table 1.
The net electrical power at 100% power achieved by each transducer was 89 ± 2 W at 20 kHz, 223 ± 3 at 400 kHz, 345 ± 3 W at
1000 kHz, and 219 ± 3 W at 2000 kHz. Based on the calorimetric
studies, specific energy levels at full power output ranged from
36 ± 1 to 130 ± 2 kJ/kg and 108 ± 3 to 390 ± 4 kJ/kg after 10 and
30 min of sonication, respectively (Table 1). Sound pressure ranged
from 77 kPa (20 kHz and 10% power input) to the highest value of
302 kPa (1000 kHz and 100% power input) (Table 1). The pressure
measurements gave similar values when comparing 10 min and
30 min sonication but, as expected, increased when increasing
the specific energy input resulting from increased power. Conditions of maximum pressure in all systems also achieve maximum
cavitation intensity [19]. The higher frequency systems generated
smaller cavities, which have higher threshold pressures of subharmonic emission. Therefore, it is expected that the higher frequency
systems contain more stable bubbles as well as transient bubbles
that collapse less violently [19].
Cavitation yield of free radicals was assessed following the
decomposition of KI into free iodine (Weissler reaction) [20], which
indicates hydrogen peroxide formation. The method quantifies the
macroscopic manifestation (i.e., the sonochemical or cavitational
yield) of microscopic phenomena (i.e., transient collapse of the cavitation bubble) [21]. Hydroxyl radical generation in multi-bubble
systems is a function of experimental parameters such as media
volume, frequency, pressure (external and sound pressure),
induced power, sound intensity and treatment time [21,22].
Hydrogen peroxide formation after 10 and 30 min US treatment
is plotted as a function of specific energy, respectively (Table 1).
The concentration of hydrogen peroxide generated at 1000 kHz
was in general higher than in systems utilizing other frequencies.
However, sonication with 20 and 2000 kHz generated significantly
(P < 0.05) lower peroxide levels. Peroxide concentration increased,
as expected, when increasing the treatment time from 10 to
30 min. Cavitational yield of free radicals has been demonstrated
to depend on the sonication time and specific energy [3]. In fact,
cavitational yield has been shown to increase linearly with increased dissipated power [3]. That was due to inertial cavitation
activity during continuous sonication which is dependent on US
energy/intensity [23]. This observation was clearly observed in
the systems run at 400 and 1000 kHz where high specific energy
levels were used at maximum power input (Table. 1). This is the
Author's personal copy
A.E. Torkamani et al. / Ultrasonics Sonochemistry 21 (2014) 951–957
result of enhanced water vapor entrapment in the bubble due to
more intense transient cavitation of smaller more stable bubbles
[21].
Some studies have reported acoustic power thresholds (1.17
and 3.14 W/cm2 at 1.7 and 2.4 MHz, respectively) for iodine production which represents cavitation [24]. Ashokkumar et al. [5]
reported that concentration of hydroxyl radicals increased when
elevating sonication frequency from 20 to 358 kHz as well as
sonication time. Peroxide value was lower at frequencies above
358 kHz or below 20 kHz [5,25]. The peak peroxide value at
358 kHz was attributed to an increase in the number of active bubbles with reduced size with frequency, which collapse less violently than larger transient cavitation bubbles formed at 20 kHz
[5]. Mason et al.[26], however, reported the peak peroxide values
at a higher frequency ranged from 582 to 863 kHz. The decrease
in peroxide value observed at 2000 kHz (Table 1) could be attributed to less time available during the expansion phase of the bubble growth in the megasonic range [5]. Presence of peroxide in milk
has been reported to oxidize milk lipids [27].
3.2. Lipid oxidative compound profile
Lipid oxidation generates various compounds including aldehydes, alcohols, acids and ketones. The GCMS volatile profile of
the non-sonicated and sonicated whey samples was dominated
by acids, alcohols, and ketones, while aldehydes appeared in lower
amounts.
Pasteurization treatment resulted in a significant (P < 0.05) increase in aldehyde, ketones and alcohols possibly due to accelerated degradation of hydroperoxide present in whey. For example,
Fig. 1 shows the increase of hexanal and 1-Pentanol concentrations
resulting from the pasteurization and subsequent incubation at
37 °C (for 10 and 30 min) of unpasteurized whey. Fatty acids present in the whey medium can potentially react with free radicals
and then degrade through autooxidative reactions resulting in generation of aldehydes and ketones as secondary volatile products.
The mechanisms of milk fat oxidation have been thoroughly studied and described in the literature [27,28].
The PCA analysis was performed to detect trends in the total
volatile data detected by the GCMS according to specific variables
(sonication time, power input or applied acoustic frequency). However, the PCA figures (data not shown) revealed no distinct pattern
or separation of variables into clusters including the effects of sonication in systems with different frequencies. It is worth mentioning that only low amounts of aldehydes and a slight increase in
concentration will be enough to change the flavor profile of whey
955
once oxidation reactions are triggered. Therefore, Table 2 only reports aldehydes detected and quantified.
As illustrated in Table 2, aldehyde concentrations in pasteurized
whey samples sonicated in each system for 30 min were not significantly (P > 0.05) higher than their respective control (non-sonicated) sample values (the same was found after 10 min
sonication, data not shown). Moreover, no significant (P > 0.05) difference in aldehydes generated up to 30 min at a matched specific
energy of 100 kJ/kg was observed. Increasing the power input (and
the specific energy) at a selected frequency also did not result in
significant (P > 0.05) change with respect to their respective control (non-sonicated) sample at any processing time. This indicates
that whey lipid oxidation was not enhanced. In general, aldehydes
odor active values (OAVs) (calculated by dividing the concentration
of a volatile compound by its threshold value in aqueous media)
were significantly below 1 with the exception of hexanal, E-2-hexanal and nonanal (Table 2) in all frequency ranges.
As mentioned earlier, ultrasonication of water resulted in higher amounts of hydroxyl radicals with higher power input, acoustic
pressure in the system, and time, particularly at 400 kHz and
1000 kHz. Even the relatively high free radical concentration detected at these frequencies did not modulate oxidation reactions
differently from the control (non-sonicated) samples. No further
increase in oxidation compounds was found in sonicated whey
possibly due to the presence of natural antioxidants (whey proteins such as lactoferrin and amino acids), which could quench free
radicals or chelate metal ions thereby preventing oxidation [29,30].
Sulfhydryl groups present in whey proteins have anti-oxidative
properties and are therefore able to inhibit formation of thiobarbituric acid reactive substances [30].
Based on the reactor characterization, oxidation effects could
have been more prominent in the 400 kHz and 1000 kHz systems,
which achieved greater decomposition of iodine into iodide; however, little or none has been published on the oxidative effects of
high frequency ultrasound on fatty acids in general. However,
physical effects of cavitation at low frequency ultrasound (microturbulence and shockwaves) can increase the interfacial area for
reaction of free radicals with fatty acids by creating very fine fat
droplets [31,32]. Chakinala et al. [33] described the mechanism
for oxidation of salicylic acid into 2,5-dihydroxybenzoic acid,
2,3-dihydroxybenzoic acid, and catechol via hydroxylation at
20 kHz and 65 W dissipated power. They demonstrated that the
hydroxylation of salicylic acid increases with sonication time and
presence of air bubbles and solid particles (TiO2 and iron powder),
which may provide additional nuclei for generation of free radicals.
In the case of whey, solid particles could be the fines (casein particulates) remaining from cheese making. However, no evidence of
such hypothetical reaction could be collected with the methodology utilized in this study.
3.3. Polar lipid profile
Fig. 1. Lipid oxidation volatile profile in the whey: (a) raw untreated whey, (b)
pasteurized whey, (c) 10 min control (pasteurized non-sonicated) sample and (d)
30 min control (pasteurized non-sonicated) sample. Data shown are means ± standard deviation of 3 experiments.
Polar lipid concentrations in freeze-dried whey samples were
reported as percentage of total lipid (%w/w). All major phospholipids detected in milk including (phosphatidylinositol (PI), ethanolamine (PE), serin (PS), choline (PC) and sphingomyelin (SM)) in
milk were also present in dried whey powder. The polar lipid
profile of control and sonicated samples obtained at different processing stages (HTST process and incubation for 30 min) are illustrated in Table 3. The concentration of polar lipid species at all
frequencies and specific energies was relatively close to values
reported at control (non-sonicated) sample conditions (incubated
for 10 and 30 min at 37 °C). Total polar lipid concentration of the
control and the sonicated samples after 30 min ranged from
1.55 ± 0.5% to 2 ± 0.9%. Polar lipids content in freeze dried whey
samples did not significantly (P > 0.05) change at different
Author's personal copy
956
A.E. Torkamani et al. / Ultrasonics Sonochemistry 21 (2014) 951–957
Table 3
Polar lipid relative concentration after ultrasound treatment for 30 min at different frequencies and specific energy inputs in comparison to control value. Values are expressed as
g polar lipid compound on a 100 g fat basis (averages of triplicate analysis ± SD of 3 individual runs).
Frequency (kHz)
a–e
Power input (%)
Specific energy (kJ/kg)
PE
PI
a
PS
b
PC
c
0.13 ± 00
Total Pl
e
Control
0
0
0.4 ± 0.10
0.6 ± 0.20
1.83 ± 0.5
20
10
50
92
100
25.2 ± 1.5
45 ± 2.25
100.5 ± 4.5
108 ± 3
0.43 ± 0.10a
0.44 ± 0.10a
0.45 ± 0.20a
0.51 ± 0.30a
0.15 ± 0.00b
0.12 ± 0.05b
0.13 ± 0.05b
0.14 ± 0.06b
0.15 ± 0.00c
0.18 ± 0.00c
0.2 ± 0.10c
0.22 ± 0.10c
0.43 ± 0.20d
0.48 ± 0.12d
0.44 ± 0.19d
0.42 ± 0.19d
0.44 ± 0.20e
0.51 ± 0.25e
0.58 ± 0.28e
0.61 ± 0.34e
1.60 ± 0.5
1.73 ± 0.5
1.8 ± 0.9
1.9 ± 1.0
400
10
20
50
100
45 ± 3
94.5 ± 4.5
138 ± 3
300 ± 6
0.50 ± 0.20a
0.40 ± 0.20a
0.45 ± 0.20 a
0.42 ± 0.20a
0.14 ± 0.06b
0.13 ± 0.08b
0.14 ± 0.09 b
0.14 ± 0.05b
0.22 ± 0.10c
0.20 ± 0.10c
0.20 ± 0.10 c
0.21 ± 0.10c
0.46 ± 0.25d
0.47 ± 0.13d
0.44 ± 0.23 d
0.39 ± 0.12d
0.62 ± 0.39e
0.60 ± 0.20e
0.58 ± 0.24 e
0.55 ± 0.22e
1.94 ± 1.0
1.8 ± 0.7
1.81 ± 0.8
1.71 ± 0.8
1000
10
17.50
50
100
37.5 ± 1
106.5 ± 2.5
201 ± 3
390 ± 4
0.45 ± 0.20a
0.51 ± 0.20a
0.50 ± 0.20a
0.52 ± 0.30a
0.13 ± 0.0b
0.15 ± 0.07b
0.15 ± 0.07b
0.15 ± 0.05b
0.21 ± 0.10c
0.23 ± 0.10c
0.23 ± 0.10c
0.23 ± 0.10c
0.43 ± 0.10d
0.40 ± 0.20d
0.48 ± 0.12d
0.46 ± 0.17d
0.60 ± 0.23e
0.61 ± 0.20e
0.64 ± 0.32e
0.62 ± 0.38e
1.82 ± 0.6
1.9 ± 0.8
2 ± 0.9
1.98 ± 1
2000
10
22.50
50
100
37.5 ± 4
94.5 ± 4
150 ± 6
300 ± 5
0.40 ± 0.20a
0.50 ± 0.20a
0.49 ± 0.20a
0.51 ± 0.10a
0.11 ± 0.00b
0.12 ± 0.10b
0.14 ± 0.07b
0.14 ± 00b
0.21 ± 0.10c
0.23 ± 0.10c
0.23 ± 0.10c
0.23 ± 0.10c
0.45 ± 0.23d
0.46 ± 0.10d
0.47 ± 0.19d
0.43 ± 0.20d
0.56 ± 0.20e
0.57 ± 0.20e
0.63 ± 0.36e
0.65 ± 0.30e
1.73 ± 0.7
1.88 ± 0.7
1.96 ± 0.9
1.96 ± 0.7
0.5 ± 0.20
0.2 ± 00
SM
d
Means within the column followed by different superscript letters differ significantly at P = 0.05.
frequencies (20–2000 kHz) and treatment times (10–30 min) with
respect to the non-sonicated controls. This indicated that US treatments and associated local pressure did not trigger reactions
resulting in PL degradation (Table 3).
sonication; data not shown). This trend coincides with the increase
in lipid oxidation compounds reported earlier in the non-sonicated
samples. As illustrated in Table 4 higher levels of FFA were found
after 30 min of incubation in comparison to 10 min trials in the
non-sonicated samples. Oleic acid (precursor of nonanal) and linoleic acid (precursor of hexanal, heptanal and E-2-heptanal) concentrations increased from 700 ± 50 to 880 ± 64 lg/kg and 60 ± 10 to
82 ± 9 lg/kg, respectively, from 10 min up to 30 min incubation
at 37 °C without US. a- and c-linolenic (precursors of (E)-2
hexanal) increased from 5 ± 2 to 9 ± 1 lg/kg and 20 ± 4 to
40 ± 4 lg/kg, respectively, from 10 min up to 30 min incubation
at 37 °C without US.
However, no significant (P > 0.05) decrease in the unsaturated
FFA concentration measured in whey powder was detected after
sonication. Overall, no significant difference (P > 0.05) in FFA concentration (with the exception of oleic acid) was noted at each
3.4. Free fatty acid profile
The major FFA (C10:0-C20:0) detected in whey powder as a
function of treatment time, power input and specific energy
included myristic, palmitic, stearic and oleic acid. A range of odd
numbered carbon chain fatty acids (C15:0-C17:1) was also reported in the profile.
Quantitative data on FFA that are most affected by oxidative
reactions (C18:1 – C18:3) are summarized in Table 4. FFA concentration increased significantly (P < 0.05) while pasteurizing and
incubating whey samples for 10 and 30 min at 37 °C (without
Table 4
FFA profile in whey at different frequencies, specific energies and SPL after 10 and 30 min of treatment (lg/kg sample; values are averages of triplicate analysis ± SD of 3 individual
runs).
Frequency (kHz)
Power input (%)
Specific energy (kJ/kg)
100
0A
20
A
c-Linolenic acid
(C18:3n6)
100
100
100
300
880 ± 64a
60 ± 10c
8.4 ± 0.5
15 ± 0.75
33.5 ± 1.5
36 ± 1.0
25.2 ± 1.5
45 ± 2.25
100.5 ± 4.5
108 ± 3
740 ± 48a
990 ± 69a
890 ± 47a
1005 ± 86a
300
a-Linolenic acid
(C18:3n3)
300
100
20 ± 4h
40 ± 4i
300
88 ± 9d
5 ± 2f
9 ± 1g
1015 ± 86a
1190 ± 89b
1210 ± 85b
1205 ± 87b
60 ± 13c
70 ± 13c
60 ± 10c
70 ± 17c
100 ± 12d
160 ± 14d
65 ± 4c
80 ± 8c
3.5 ± 0.4f
4 ± 0.5f
3 ± 0.8f
6 ± 1.5f
7 ± 2g
6 ± 2.5g
8 ± 3.5g
10 ± 2g
25 ± 5h
27 ± 6h
31 ± 8h
32 ± 10h
40 ± 9i
65 ± 8j
38 ± 5i
37 ± 6i
400
10
28.50(a)
50
100
15 ± 1.0
31.5 ± 1.5
46 ± 1
100 ± 2.2
45 ± 3
94.5 ± 4.5
138 ± 3
300 ± 6
780 ± 34a
1314 ± 56b
950 ± 32a
980 ± 34a
1360 ± 89b
1365 ± 89b
1100 ± 78b
2100 ± 78b
75 ± 18c
60 ± 13c
60 ± 6c
65 ± 12c
260 ± 23e
120 ± 12d
160 ± 12d
230 ± 65e
6.5 ± 1f
3 ± 0.8f
3.5 ± 1.5f
5 ± 1.2f
9 ± 3g
11 ± 2g
8 ± 2g
7.5 ± 3g
30 ± 9h
40 ± 8h
33 ± 6h
35 ± 4h
35 ± 8i
45 ± 8i
60 ± 4i
90 ± 13j
1000
10
17.5(a)
50
100
12.5 ± 0.35
35.5 ± 0.75
67 ± 1
130 ± 1.75
37.5 ± 1
106.5 ± 2.5
201 ± 3
390 ± 4
635 ± 34a
900 ± 42a
1000 ± 45a
1245 ± 110b
1180 ± 78b
790 ± 14a
1045 ± 78a
1760 ± 89b
80 ± 8c
65 ± 15c
70 ± 10.0c
65 ± 12c
140 ± 14d
70 ± 8c
95 ± 9d
180 ± 22d
3 ± 0.4f
5 ± 1f
5 ± 0.5f
8 ± 1g
12 ± 2g
10 ± 3g
7 ± 2g
12 ± 3g
25 ± 6h
33 ± 8h
16 ± 1h
35 ± 8h
50 ± 8i
25 ± 4i
35 ± 5i
80 ± 12j
12.5 ± 0.75
31.5 ± 1.5
50 ± 1.75
100 ± 1.5
37.5 ± 4
94.5 ± 4
150 ± 6
300 ± 5
860 ± 89a
820 ± 78a
822 ± 79a
910 ± 78a
1050 ± 110a
1235 ± 89b
1310 ± 98b
1250 ± 45b
60 ± 13c
55 ± 11c
60 ± 9c
80 ± 12c
110 ± 15d
120 ± 19d
130 ± 12d
130 ± 12d
4 ± 1.5f
6 ± 2.5f
4 ± 0.5f
6.5 ± 2.5f
10 ± 4g
9 ± 3g
8 ± 2.5g
11 ± 3g
30 ± 8h
15 ± 1h
30 ± 8h
31 ± 7h
40 ± 4i
45 ± 6i
50 ± 8i
60 ± 12j
2000
a–j
Linolenic acid
(C18:2n6, cis)
700 ± 50aa
0
10
50
92(a)
100
300
Oleic acid
(C18:1n9, cis)
10
20(a)
50
100
Means within the column followed by different superscript letters differ significantly at P = 0.05.
The first row illustrates FFA concentration in pasteurized control samples after being incubated for 10 and 30 min.
Author's personal copy
A.E. Torkamani et al. / Ultrasonics Sonochemistry 21 (2014) 951–957
treatment time between the non-sonicated and sonicated samples (Table 4). Oleic acid concentration in samples sonicated at
31.4 kJ/kg (400 kHz), 130 kJ/kg (1000 kHz), 301 kJ/kg (400 kHz)
and 389 kJ/kg (1000 kHz) was significantly higher (P < 0.05) than
control values (incubated for 10 and 30 min). However, no significant
decrease in unsaturated FFA concentration was detected, indicating
that unsaturated fatty acids were not consumed in autoxidation
reactions at any frequency, specific energy or treatment time.
It is worth noting that a decrease in total fatty acid concentration would have been indicative of oxidation reactions leading to
the production of shorter chain aldehydes, ketones and alcohols.
However, the increase in FFA observed in both control and ultrasonic samples is mostly likely attributed to lipolysis reactions catalyzed by lipases/esterases, originating from lactic acid bacteria
used as starter culture in the manufacture of Cheddar cheese,
and agitation and temperature in the US treatment vessel may augment the same. Krukovsky and Herrington [34] indicated maximum enzyme activity was observed at 30–50 °C. Furthermore,
starter cultures might be the source of lipase enzyme in whey as
indigenous milk lipase was destroyed during pasteurization [35].
4. Conclusion
Ultrasonication as a standalone treatment did not promote lipid
oxidative reactions beyond detectable odor thresholds for volatile
compounds, even at the highest specific energy input value
(390 kJ/kg) or when the highest level of free radical formation
occurred near 1000 kHz. Variation of frequency, power input, and
processing time did not result in a significant change in concentration of oxidative volatile compounds, PL species or FFA, in whey
samples. This work suggests that US can be utilized for processing
whey in several applications such as whey powder solubility
enhancement, foaming capacity modification, viscosity reduction
of concentrates, among other listed applications, without changing
its oxidative quality.
References
[1] P. Juliano, A. Kutter, L. Cheng, P. Swiergon, R. Mawson, M. Augustin, Enhanced
creaming of milk fat globules in milk emulsions by the application of
ultrasound and detection by means of optical methods, Ultrason. Sonochem.
18 (2011) 963–973.
[2] P. Juliano, S. Temmel, M. Rout, P. Swiergon, R. Mawson, K. Knoerzer, Creaming
enhancement in a liter scale ultrasonic reactor at selected transducer
configurations and frequencies, Ultrason. Sonochem. 20 (2013) 52–62.
[3] P.R. Gogate, V.S. Sutkar, A.B. Pandit, Sonochemical reactors: important design
and scale up considerations with a special emphasis on heterogeneous
systems, Chem. Eng. J. 166 (2011) 1066–1082.
[4] T.G. Leighton, Bubble population phenomena in acoustic cavitation, Ultrason.
Sonochem. 2 (1995) 123–136.
[5] M. Ashokkumar, D. Sunartio, S. Kentish, R. Mawson, L. Simons, K. Vilkhu, C.
Versteeg, Modification of food ingredients by ultrasound to improve
functionality: a preliminary study on a model system, Innovative Food Sci.
Emerg. Technol. 9 (2008) 155–160.
[6] M. Ashokkumar, R. Bhaskaracharya, S. Kentish, J. Lee, M. Palmer, B. Zisu, The
ultrasonic processing of dairy products – an overview, Dairy Sci. Technol. 90
(2010) 147–168.
[7] K. Yasui, T. Tuziuti, J. Lee, T. Kozuka, A. Towata, The range of ambient radius for
an active bubble in sonoluminescence and sonochemical reactions, J. Chem.
Phys. 128 (2008) 184705.
[8] A.R. Jambrak, T.J. Mason, V. Lelas, Z. Herceg, I.L. Herceg, Effect of ultrasound
treatment on solubility and foaming properties of whey protein suspensions, J.
Food Eng. 86 (2008) 281–287.
957
[9] B. Zisu, J. Lee, J. Chandrapala, R. Bhaskaracharya, M. Palmer, Effect of
ultrasound on the physical and functional properties of reconstituted whey
protein powders, J. Dairy Res. 78 (2011) 226–232.
[10] J. Chandrapala, B. Zisu, M. Palmer, S. Kentish, M. Ashokkumar, Effects of
ultrasound on the thermal and structural characteristics of proteins in
reconstituted whey protein concentrate, Ultrason. Sonochem. 18 (2011)
951–957.
[11] S. Martini, M.K. Walsh, Sensory characteristics and functionality of sonicated
whey, Food Res. Int. 49 (2012) 694–701.
[12] R. Reiner, F. Noci, A. Denis, D. Morgan, J. Lyng, Characterization of volatile
compounds generated in milk by high intensity ultrasound, Int. Dairy J. 19
(2009) 269–272.
[13] K.S. Suslick, Y. Didenko, M.M. Fang, T. Hyeon, K.J. Kolbeck, Acoustic cavitation
and its chemical consequences, Philosophical transactions of the Royal Society
of London. Series A: Mathematical, physical, and engineering sciences 357
(1999) 335–353.
[14] K. Makino, M.M. Mossoba, P. Riesz, Chemical effects of ultrasound on aqueous
solutions. Formation of hydroxyl radicals and hydrogen atoms, J. Phys. Chem.
1952 (87) (1983) 1369–1377.
[15] E. Chouliara, K.G. Georgogianni, N. Kanellopoulou, M.G. Kontominas, Effect of
ultrasonication on microbiological, chemical and sensory properties of-áraw,
thermized and pasteurized milk, Int. Dairy J. 20 (2010) 307–313.
[16] A.E. Alegria, Y. Lion, T. Kondo, P. Riesz, Sonolysis of aqueous surfactant
solutions: probing the interfacial region of cavitation bubbles by spin trapping,
J. Phys. Chem. 1952 (93) (1989) 4908–4913.
[17] R. Rombaut, J.V. Camp, K. Dewettinck, Analysis of phospho- and sphingolipids
in dairy products by a new HPLC method, J. Dairy Sci. 88 (2005) 482–488.
[18] Z. Shen, C. Apriani, R. Weerakkody, L. Sanguansri, M.A. Augustin, Food matrix
effects on in vitro digestion of microencapsulated tuna oil powder, J. Agric.
Food Chem. 59 (2011) 8442–8449.
[19] V.S. Moholkar, S.P. Sable, A.B. Pandit, Mapping the cavitation intensity in an
ultrasonic bath using the acoustic emission, AIChE J. 46 (2000) 684–694.
[20] P.R. Gogate, I.Z. Shirgaonkar, M. Sivakumar, P. Senthilkumar, N.P. Vichare,
Cavitation reactors: efficiency assessment using a model reaction, AIChE J. 47
(2001) 2526–2538.
[21] T. Sivasankar, A.W. Paunikar, V.S. Moholkar, Mechanistic approach to
enhancement of the yield of a sonochemical reaction, AIChE J. 53 (2007)
1132–1143.
[22] J.P.L.T.J. Mason, Sonochemistry: Theory, Applications and uses of Ultrasound in
Chemistry, Wiley-Interscience, 1989.
[23] A. Ebrahiminia, M. Mokhtari-Dizaji, T. Toliyat, Correlation between iodide
dosimetry and terephthalic acid dosimetry to evaluate the reactive radical
production due to the acoustic cavitation activity, Ultrason. Sonochem. 20
(2013) 366–372.
[24] D.M. Kirpalani, K.J. McQuinn, Experimental quantification of cavitation yield
revisited: focus on high frequency ultrasound reactors, Ultrason. Sonochem.
13 (2006) 1–5.
[25] S. Koda, T. Kimura, T. Kondo, H. Mitome, A standard method to calibrate
sonochemical efficiency of an individual reaction system, Ultrason. Sonochem.
10 (2003) 149–156.
[26] T.J. Mason, A.J. Cobley, J.E. Graves, D. Morgan, New evidence for the inverse
dependence of mechanical and chemical effects on the frequency of
ultrasound, Ultrason. Sonochem. 18 (2011) 226–230.
[27] T.P. O’Connor, N.M. O’Brien, Lipid oxidation, in: P. McSweeney, P. Fox (Eds.),
Advanced Dairy Chemistry, 2006, pp. 567–600.
[28] E. Frankel, Free radical oxidation, in: Lipid Oxidation, Woodhead Publishing
Cambridge, Cambridge, UK, 2005, pp. 15–24.
[29] A.A. Browdy, N.D. Harris, Whey improves oxidative stability of soybean oil, J.
Food Sci. 62 (1997) 348–350.
[30] L.M. Tong, S. Sasaki, D.J. McClements, E.A. Decker, Mechanisms of the
antioxidant activity of a high molecular weight fraction of whey, J. Agric.
Food Chem. 48 (2000) 1473–1478.
[31] A. Kalva, T. Sivasankar, V.S. Moholkar, Physical mechanism of ultrasoundassisted synthesis of biodiesel, Ind. Eng. Chem. Res. 48 (2009) 534–544.
[32] P.A. Parkar, H.A. Choudhary, V.S. Moholkar, Mechanistic and kinetic
investigations in ultrasound assisted acid catalyzed biodiesel synthesis,
Chem. Eng. J. 187 (2012) 248–260.
[33] A.G. Chakinala, P.R. Gogate, A.E. Burgess, D.H. Bremner, Intensification of
hydroxyl radical production in sonochemical reactors, Ultrason. Sonochem. 14
(2007) 509–514.
[34] V.N. Krukovsky, P.F. Sharp, Effect of lipolysis on the churnability of cream
obtained from the milk of cows in advanced lactation, J. Dairy Sci. 19 (1936)
279–284.
[35] M. El Soda, E. Law, G. Tsakalidou, G. Kalantzopoulos, Lipolytic activity of cheese
related microorganisms and its impact on cheese flavour, in: C. George (Ed.),
Developments in Food Science, Elsevier, New York, 1995, pp. 1823–1847.