ELECTROCHEMICAL REACTIONS DURING OHMIC HEATING DISSERTATION Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University By Chaminda Padmal Samaranayake, B.Sc. Honors ∗ ∗ ∗ ∗ ∗ The Ohio State University 2003 Dissertation Committee: Professor Sudhir K. Sastry, Adviser Approved by Professor Q. Howard Zhang Professor David B. Min Adviser Professor Russ C. Hille Food Science and Nutrition Graduate Program
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
ELECTROCHEMICAL REACTIONS DURING OHMIC HEATING
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the Graduate
School of The Ohio State University
By
Chaminda Padmal Samaranayake, B.Sc. Honors
∗ ∗ ∗ ∗ ∗
The Ohio State University 2003
Dissertation Committee:
Professor Sudhir K. Sastry, Adviser Approved by Professor Q. Howard Zhang
Professor David B. Min Adviser Professor Russ C. Hille Food Science and Nutrition Graduate Program
ABSTRACT Electrochemical reactions, chemical reactions at electrode/solution interfaces
induced by current, are undesirable during ohmic heating. These reactions may be
avoidable or suppressible through an understanding of electrochemical behavior of ohmic
heaters. Though numerous studies have dealt with the applications of ohmic heating, little
is known regarding electrochemical aspects.
Electrochemical behavior of four types of electrode materials: titanium, stainless
steel, platinized-titanium, and graphite, was studied at (initial) pH 3.5, 5.0, and 6.5 using
60 Hz sinusoidal alternating current. Concentrations of metal ions and elemental carbon
migrated into the heating media were determined by inductively coupled plasma (ICP) –
mass, and -emission spectrometers. Hydrogen gas accumulation in the headspace of the
ohmic heater, and pH changes of the heating media were also measured. Stainless steel
was found to be the most electrochemically active electrode material, whereas platinized-
titanium exhibited relatively inert electrochemical behavior at all the pH values. The
potential use of platinized-titanium electrodes for ohmic heating operations was further
demonstrated on a pilot scale.
Effects of frequency, pulse width, and delay time of a pulsed ohmic heating
technique on electrochemical reactions were studied, in comparison with conventional
iii
(60 Hz, sine wave) ohmic heating. Analyses of electrode corrosion, hydrogen generation,
and pH measurements suggest that the pulsed ohmic heating is capable of significantly
reducing the electrochemical reactions of titanium, stainless steel, and platinized-titanium
electrodes. The delay time was found to be a critical factor.
Electrochemical and secondary chemical reactions during 60 Hz ohmic heating of
ascorbic acid in citrate-phosphate buffer with stainless steel electrodes were characterized
by a number of analytical methods. Electrode corrosion showed marked effects on the
heating buffer medium forming metal-phosphates and metal-citrate complexes. Effects of
reactions on pH, buffer capacity, and ascorbic acid degradation are discussed.
Free radical generation was investigated by spin trapping with 5,5-dimethyl-1-
pyrroline N-oxide (DMPO), and employing electron spin resonance (ESR) spectroscopy.
The frequency range of 1 – 8 kHz is recommended to suppress free radical generation
with platinized-titanium electrodes. Ohmic heating operated at 60 Hz (sine wave) and 10
kHz (pulses) indicated the generation of •OH radicals.
iv
Dedicated to all who assisted my education prior to and during my study at The Ohio State University.
v
ACKNOWLEDGMENTS
I would like to thank my adviser, Dr. Sudhir Sastry, for intellectual support,
guidance, and financial aid offered to me through out my degree program. I wish to thank
Drs. Howard Zhang, David Min, Russ Hille, and Richard McCreery for kindly serving on
my dissertation and candidacy examination committees.
I gratefully acknowledge the collaboration of Dr. Russ Hille, and Craig Hemann
(Department of Molecular and Cellular Biochemistry) in the free radical study (chapter
5). I am also grateful for the hospitality received from Hille’s lab.
A special thanks goes to Brian Heskitt for making all my experimental setups, and
for the technical assistance provided to me through out this research. I thank Dr. John
Olesik, director of microscopic and chemical analysis research center at OSU, for
providing ICP-MS, ICP-OES, SEM, and SEM-EDX analytical services. I appreciate the
assistance received from Dr. Johnie Brown, the former associate director of campus
chemical instrument center-mass spectrometry laboratory at OSU, with GC-MS and ESI-
MS analyses.
Finally, I wish to thank Drs. Salengke and Ilkay Sensoy, the former members of our
ohmic heating group, and Pisit Wongsa-ngasri for their wide range of support during my
study at OSU. My appreciation is extended to Karthik Vembu who helped me in cleaning
vi
up viruses and restoring the programs in my computer, and also to Rakhith U.C. for his
contribution to some digital images.
vii
VITA
August 26, 1970 …………………..Born – Chilaw, Sri Lanka
1991 – 1995 ……………………….B.Sc. (Chemistry) Special Degree, with First Class Honors pass (subsidiary subject: Mathematics) University of Sri-Jayawardenepura, Sri Lanka 1995 ……………………………….Demonstrator (Inorganic Chemistry), Department of Chemistry, University of Sri-Jayawardenepura, Sri Lanka 1996 – 1998 ……………………….Assistant lecturer (Phys. and Environ. Chemistry), Department of Chemistry, University of Kelaniya, Sri Lanka 1998 – present …………………….Graduate Research Associate, The Ohio State University
PUBLICATIONS
Research Publications
1. Assiray A., Sastry S.K., and Samaranayake C. (2003); Degradation kinetics of ascorbic acid during ohmic heating with stainless steel electrodes, Journal of applied electrochemistry, 33 (2), pp. 187-196. 2. Edirisinghe E.M.R.K.B., Samaranayake C.P., Bamunuarchchi A., Walpola S., and De Alwis A.A.P. (1997); Nutrient retention in ohmic heating; 7 th International Congress on Engineer and Food (ICEF – 7), Brighton U.K., SA 43-46. 3. Samaranayake C.P., De Alwis A.A.P., and Bamunuarchchi A. (1996); Peroxide formation in ohmic heating of meats; Ceylon Journal of Science (Physical Sciences), 3(1), pp. 30-35.
viii
Published Abstracts
1. Samaranayake C.P. and Sastry S.K. (2003); Electrochemical corrosion of platinized-titanium electrodes during ohmic heating; Institute of Food Technologists annual meeting, Chicago, IL. 2. Samaranayake C.P. and Sastry S.K. (2002); Electrochemical reactions during ohmic heating; Institute of Food Technologists annual meeting, Anaheim, CA.
FIELDS OF STUDY
Major Field: Food Science and Technology
ix
TABLE OF CONTENTS
Page Abstract …………………………………………………………………………………. ii Dedication …………………………………………………………….………………… iv Acknowledgments …………………………………………….………………………… v Vita …………………………………………………………………………………….. vii List of Tables ……………………………………………………………………………xii List of Figures ………………………………………………………………...……….. xiv Chapters: 1. Introduction ………………………………………………….……………………. 1
1.1 The ohmic heating process ……………………………….…………………….. 1 1.2 Electrochemical reactions ……………………………………………………… 2
References ………….……………………………………………………………… 68 4. Electrochemical reactions during 60 Hz ohmic heating of ascorbic acid in buffer medium with stainless steel electrodes ………………..……………… 89
List of References ………………..…………………………………………………... 145
xii
LIST OF TABLES Table Page 2.1 Comparison of corrosion rates (in ppb per KJ) of the electrode materials with respect to the migrations of their major (surface) elements at different pH values ……………………………………………………………………..……... 41 2.2 pH changes of the heating media observed with stainless steel electrodes at different pH values ………………………………………………………….… 41 2.3 Pt and Ti concentrations (in parts per trillions) of the ohmically heated heating medium in the pilot scale study …………………………………….…… 42 2.4 Comparison of estimated metal intakes via consumption of an 8 oz ohmically heated meal with the published upper-level daily dietary exposure limits for adult consumers. The estimation is based on unit conversions: 1 ppt = 1 picogram/ g; 8 oz = 227 g; 1 picogram = 10-12 g = 10-6 µg ……………….…… 42 3.1 Frequency-pulse width combinations with required initial electrical conductivities, and cooling water temperatures for stainless steel electrodes ………………………………………………….……… 69 3.2 Frequency-pulse width combinations with required initial electrical conductivities, and cooling water temperatures for titanium electrodes ……………………………………………………….………. 70 3.3 Frequency-pulse width combinations with required initial electrical conductivities, and cooling water temperatures for platinized-titanium electrodes ………………………………………….………… 71 3.4 Frequency-pulse width combinations with required initial electrical conductivities, and cooling water temperatures for graphite electrodes ………………………………………………………..……… 72 4.1 Ohmic heating conditions ……………………………………...……………….. 109
xiii
4.2 Some indicators of the electrochemical processes at different power densities and NaCl concentrations ……………………………………………… 110 4.3 Chemical compositions (as weight %) of electrode deposits at different power densities and NaCl concentrations. The values are means of five replicates (n=5) with respective standard deviations in parentheses …………… 111 4.4 Minimum migratory metal ion concentration [Mn+] needed to precipitate some metal-phosphates and metal-hydroxides in the presence of the same [Na2HPO4] as in the citrate-phosphate buffer system at pH 3.5 …………….….. 112 4.5 The effect of AA-induced Fenton’s reaction on buffer pH, in comparison with the observed pH changes at different power densities and NaCl concentrations ……………………………………………………………..……. 113 4.6 The spectral maxima (λmax) and the respective absorptivity coefficients of 1:1 Fe(III)-citrate, and the ohmically heated heating medium at 1.5 Wcm-3 (1.0% NaCl) ………………………………………..……………….. 114 4.7 GC-MS characteristics and % losses of the buffer components ………..………. 114 5.1 Selected ohmic heating conditions to study free radical generation. See figure 5.3 for typical time-temperature history for all these ohmic heating conditions ……………………………………………………….. 133
xiv
LIST OF FIGURES Figure Page 1.1 The concept of ohmic heating …………………………………………………… 15
1.2 A simplified electrical equivalent circuit of the interface during the application of AC; Cd: electrical double layer capacitor, Rct: charge-transfer resistance, Rs: electrolyte resistance ………………………………………………………… 15 1.3 Mechanisms of generating free radicals by electrolysis and electrode corrosion. . 16 2.1 The laboratory scale ohmic heater ………………………………………………. 43 2.2 Schematic diagram of the laboratory scale experimental setup …………………. 44 2.3 Typical time vs. temperature curve for all electrodes at all pH values during ohmic heating ………………………………………………………………….… 45 2.4 Typical time vs. current curve for all electrodes at all pH values during ohmic heating ……………………………………………………………….…… 45 2.5 Typical SEM micrograph of titanium electrodes …………...…………………… 46 2.6 Typical SEM micrograph of stainless steel electrodes ………………………….. 47 2.7 Typical SEM micrograph of platinized-titanium electrodes ……………….……. 48 2.8 Typical SEM micrograph of graphite electrodes ……………...………………… 49 2.9 Hydrogen generation with titanium electrodes during ohmic heating …….…….. 50 2.10 Hydrogen generation with stainless steel electrodes during ohmic heating …….. 50 2.11 Identified graphite corrosion products by GC-MS analysis; (a) 2-hydroxy, propanoic acid (lactic acid) (MW: 90), (b) 2-hydroxy, 4- methyl, pentanoic acid (MW: 132) ………………………………….…………………… 51
xv
2.12 Positive ion ESI-MS spectra of the heating medium before and after ohmic heating ………………………………………………………….………… 52 3.1 The ohmic heater used for both pulsed and conventional ohmic heating experiments ……………….………………………………..……………………. 73 3.2 Schematic diagram of the experimental setup used for pulsed ohmic heating ….. 74 3.3 Typical time vs. temperature curve for all the ohmic heating experiments …...… 75 3.4 Schematic diagram of the centering of bipolar pulses within the period to study the effects of frequency and pulse width …….……………………….…… 76 3.5 Typical pulse waveforms at 10 kHz for different pulse widths. The top and the bottom waves in each diagram represent the current and the voltage, respectively ……………….………………………………………… 77 3.6 Typical pulse waveforms at 4 kHz for different pulse widths. The top and the bottom waves in each diagram represent the current and the voltage, respectively …….…………………………………………………… 78 3.7 Schematic diagram of the centering of bipolar pulses within the period to demonstrate the effect of delay time ……….……………………………….…… 79 3.8 Typical pulse waveforms for different delay times. The top and the bottom waves in each diagram represent the current and the voltage, respectively ….…………………………………………..………….. 80 3.9 The corrosion rates (in ppb per KJ) of stainless steel electrodes with respect to Fe at different frequencies and pulse widths. Number shown in parenthesis is the corresponding frequency (kHz) for each pulse width (µs). The corresponding duty cycle is indicated as a percentage. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, b, c, d, and e denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates ……………………...….. 81 3.10 The corrosion rates (in ppb per KJ) of stainless steel electrodes with respect to Cr at different frequencies and pulse widths. Number shown in parenthesis is the corresponding frequency (kHz) for each pulse width (µs). The corresponding duty cycle is indicated as a percentage. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, b, c, and d denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates ………………….………. 82
xvi
3.11 The corrosion rates (in ppb per KJ) of titanium electrodes with respect to Ti at different frequencies and pulse widths. Number shown in parenthesis is the corresponding frequency (kHz) for each pulse width (µs). The corresponding duty cycle is indicated as a percentage. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, b, c, d, and e denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates …………………..……… 83 3.12 The corrosion rates (in ppb per KJ) of platinized-titanium electrodes with respect to Pt at different frequencies and pulse widths. Number shown in parenthesis is the corresponding frequency (kHz) for each pulse width (µs). The corresponding duty cycle is indicated as a percentage. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, b, c, and d denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates ………………….………. 84 3.13 The corrosion rates (in ppb per KJ) of platinized-titanium electrodes with respect to Ti at different frequencies and pulse widths. The presence of asterisk (*) indicates corrosion rate either < 0.001 ppb/ KJ or undetectable. Number shown in parenthesis is the corresponding frequency (kHz) for each pulse width (µs). The corresponding duty cycle is indicated as a percentage. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, and b denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates ……………… 85 3.14 The corrosion rates (in ppm per KJ) of graphite electrodes with respect to elemental carbon at different frequencies and pulse widths. The presence of asterisk (*) indicates an undetectable corrosion rate. Number shown in parenthesis is the corresponding frequency (kHz) for each pulse width (µs). The corresponding duty cycle is indicated as a percentage. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, and b denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates ……………… 86 3.15 The corrosion rates (in ppb per KJ) with respect to Fe for different delay times. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, b, and c denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates …………………………………………………………………… 87 3.16 The corrosion rates (in ppb per KJ) with respect to Cr for different delay times. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, b, c, and d denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates …………………………………………………………………… 88 4.1 Buffer capacities (in µmol pH-1ml-1) at 0.25% (w/v) NaCl. [1] and [2] correspond to the amounts of metal ions migrated at 0.5 Wcm-3 and 0.75 Wcm-3, respectively ………………………………………………………. 115
xvii
4.2 Buffer capacities (in µmol pH-1ml-1) at 0.50% (w/v) NaCl ……………………. 116 4.3 Buffer capacities (in µmol pH-1ml-1) at 1.0% (w/v) NaCl ………….………….. 117 4.4 The buffer solution before being subjected to ohmic heating (a), and the ohmically heated medium at 1.5 Wcm-3 (1.0% NaCl) (b) …………………. 118 4.5 UV-Visible absorption spectra of 1:1 Fe(III)-citrate( ), and the ohmically heated medium ( ) at 1.5 Wcm-3 (1.0% NaCl) …………...… 119 4.6 Total ion chromatograms of TBDMS derivatized solutions of the unheated buffer (A), and the ohmically heated medium at 1.5 Wcm-3 (1.0% NaCl) (B). The CA and P peaks represent citric acid and phosphate, respectively ……..… 120 5.1 Schematic diagram of the pressurized ohmic heater …………….…………….. 134 5.2 Schematic diagram of the experimental setup used for the ohmic heating ……. 135 5.3 Typical time-temperature histories for ohmic and conventional heating ……… 136 5.4 Schematic diagram of the centering of bipolar pulses within the period (T) at each frequency (f). The positive and negative pulses having the same pulse width (tp) were equally spaced by adjusting the delay time (td) as T = 2 (tp + td) ………………………………………………………………… 137 5.5 The ESR spectrum of the DMPO-OH reference. This signal represents spin concentration of 0.63 µM …………………………………………………. 138 5.6 Typical ESR spectra of ohmic and conventional heating experiments, in comparison with the ESR spectrum of DMPO-OH reference. The signals at 60 Hz (sine wave) and 10 kHz correspond to average spin concentrations of 0.14 and 0.11 µM, respectively …………………………..…………………. 139 5.7 Chemistry of •OH and O2
•− trapping by DMPO in the presence and absence of ethyl alcohol ………………………………………………………………… 140 5.8 Comparison of typical ESR spectra of the ohmic heating experiments carried out at 60 Hz (sine wave) and 10 kHz in the presence (2%, v/v) and absence of ethyl alcohol …………………………………………………… 141
1
CHAPTER 1
INTRODUCTION
In recent decades, technologies utilizing electrical energy directly into food
processing have attracted renewed interest in the food industry. Some of those are now
being used on a commercial scale for processing of a broad range of food products.
Research in this area will provide the food processor with the opportunity to produce new
and value-added food products with enhanced quality attributes preferred by consumers.
Since heating is one of the traditional and still widely popular treatments applied to food
both in the industry and the home, electroheating technologies have gained increasing
industry interest and attention. Ohmic heating, a well-known electroheating technique,
has extensively developed during the past two decades; and today it is in commercial
scale operation for processing of a number of food products, especially those containing
particulates.
1.1 The ohmic heating process
As shown in figure 1.1, the concept of ohmic heating is quite simple. The passage of
electric current through an electrically conductive food material obeys Ohm’s law (V =
IR); and heat generation due to the electrical resistance of the food, is given by:
2
Pheat = I2R (1.1)
The design of ohmic heaters is governed by the electrical conductivity of the food. Since
most food materials contain a considerable amount of free water with dissolved ionic
species, the conductivity is high enough for a heating effect to occur.
Applications of ohmic heating in the food industry emerged in the 1930’s as a
pasteurization process for milk (Getchell, 1935; Moses, 1938). Then, the technique has
been applied to blanching of vegetables (Mizrahi et al., 1975), thawing of frozen foods
(Naveh et al., 1983); and recently, pasteurization and sterilization of liquid and
particulate food products that can be packed under aseptic conditions (Parrott, 1992;
Zoltai et al., 1996). In addition, a large number of potential future applications exist for
ohmic heating, including its use in evaporation, dehydration, fermentation, and extraction
(Butz et al., 2002).
1.2 Electrochemical reactions
1.2.1 Electrode/solution interface
To understand the electrochemical behavior of ohmic heaters, it is necessary to
identify the characteristics of the electrode/solution interface. As described in interfacial
electrochemistry, the electrode/solution interface is analogous to a parallel combination
of a resistor and a capacitor (Rubinstein, 1995). A simplified electrical equivalent circuit
of the interface during the application of alternating current (AC) is shown in figure 1.2.
In reality, the so-called electrical double layer capacitor (Cd) can hold only a limited
number of charges. Once it is fully charged or ‘saturated’, it becomes a ‘leaky’ capacitor,
3
and charge-transfer occurs between the plates of the capacitor generating faradaic current,
However, most electrode/solution interfaces also exhibit a potential range (1 – 2 V
wide at most) where no faradaic reactions can take place. If the potential difference
across the double layer is maintained within its faradaic reaction-free potential window,
electrons from the electrode cannot be transferred to the electrolyte and nor can ions from
the electrolyte react at the electrode. The only phenomenon occurring at the electrode is a
periodic change in charge density on both sides of the interface. Under such
circumstances the current flowing through the interface becomes purely capacitive.
1.2.2 Electrochemical reactions induced by alternating currents
In AC circuits, both current and voltage oscillate as a wave at a certain frequency.
When AC is applied to an electrolytic cell, the double layer capacitors of the
electrode/solution interfaces charge and discharge periodically. If the frequency of the
AC wave is low, the capacitors can be fully charged during the rising part of the wave
turning on electrochemical reactions. Those reactions involve simultaneous cathodic (i.e.
reduction) and anodic (i.e. oxidation) half-reactions; and the overall reactions produce
periodic concentration changes of redox species at the electrode surfaces. The extents of
those chemical changes primarily depend upon the frequency of the applied AC signal
and the chemistry of the electrolytic cell.
Electrochemical phenomena induced by AC were first reported in the early
nineteenth century, and it was a common difficulty encountered in measuring
conductivity of electrolytes. Shaw (1950) has reported that when an alternating current is
4
applied to an electrolytic cell, the cell shows both dissipative and reactive characteristics.
Bentley et al. (1957) observed corrosion of stainless steel, platinum, and gold electrodes
when low-frequency (50 Hz) alternating currents were passed through concentrated acids;
and this corrosive effect was not evident at frequencies greater than a few kHz. They
further encountered distorted waveforms of voltage across the test cells at the frequency
of 50 Hz; and they attempted to correlate this waveform distortion with corrosion. Most
of the previous investigations under AC induced corrosion have been briefly reviewed by
Kulman (1961). He implied that AC induced electrolysis is closely associated with the
corrosion of electrodes. Then, Venkatesh et al. (1979) have reviewed some of the
fundamentals of AC induced electrochemical processes. Their discussion indicates that
when a sinusoidal alternating electric field is applied to an electrolytic cell, a direct
current (DC) or a ‘faradaic rectification current’ is generated at each of the electrode
surface; and this DC component of the current is related to the amplitude of the applied
voltage signal. Another discussion and a review of AC induced anodic and cathodic
reactions, and the effect of frequency have been given by Venkatachalam et al.(1981).
Recently, Lalvani et al. (1994, 1996) and Bosch et al. (1998) carried out some theoretical
studies for predicting AC induced corrosion; and they also reported that the corrosion
behavior strongly depends on the amplitude of applied voltage signal.
1.2.3 Electrochemical reactions during ohmic heating
Typically, ohmic heaters are powered by low-frequency (50 - 60 Hz) AC coming
from the public utility supply, because that mainly minimizes the cost and power supply
complexity. Under such alternating frequencies, a part of the current passing through the
5
electrode/solution interfaces causes electrochemical processes. Although a number of
studies have addressed the basic engineering and heat transfer aspects of ohmic heating
over the years, surprisingly limited attention was paid to electrochemical processes.
The possibility of electrochemical reactions occurring at electrode/solution interfaces
during ohmic heating has been described by Stirling (1987). He demonstrated the
selection of safe maximum current density that minimizes the faradaic current to less than
0.1% of the total current, using a platinized-titanium/ saturated NaCl ohmic cell.
Palaniappan et al. (1991), Uemura et al. (1994), Assiry (1996), Reznick (1996), Wu et al.
(1998), and Assiry et al. (2003) observed apparent electrolysis of the heating medium and
electrode corrosion during ohmic heating. Some of these authors reported that those
electrochemical effects diminish with increasing frequency. A broad discussion of
fundamental electrochemistry related to ohmic heating has been given by Amatore et al.
(1998); and the use of high alternating frequencies was suggested to inhibit adverse
electrochemical effects. Tzedakis et al.(1999) has recently examined the electrochemical
behavior of platinum and platinized-titanium electrodes for ohmic sterilization of some
commercial food products. Their results indicate that, at the frequency of 50 Hz, only
platinized-titanium would be capable of suppressing the electrochemical behavior.
1.3 Effects of electrochemical reactions on the ohmic heating process
Food materials are inherently a complex mixture of several different chemical
compounds. During ohmic heating, various electrochemical reactions can potentially
occur. In addition, some of the products of those electrochemical reactions may initiate a
number of secondary chemical reactions. Although it is not possible to examine the
6
effects of all those electrochemical and chemical processes, the following more
frequently encountered electrochemical reactions and their effects on the ohmic heating
process cannot be overlooked.
1.3.1 Electrode corrosion
In ohmic heating, electrodes are necessary to convey the current to the food material
to be heated. During heating, electrode corrosion occurs mainly via electrodissolution
induced by the low-frequency AC. For metallic electrodes (M), a generalized anodic half-
reaction for the electrode corrosion can be written as follows.
M (solid) ⇔ M n+ (aqueous) + n e, (where n = 1,2,3….) (1.2)
The metal ions (M n+) migrated into the heating medium are basically contaminants, and
may have some toxic potential. However, on the other hand, the electrode corrosion
might represent an opportunity to introduce essential minerals into the processed foods.
Since food systems are generally rich in ligands, the migrated transition metal ions
can form various coordination complexes. Those metal complexes typically have
characteristic colors; and therefore, they may involve alteration of color of the processed
foods. It is also known that, some transition metal ions have catalytic effects for certain
food reactions, such as lipid oxidation. Therefore, the electrode corrosion may have an
impact on flavor quality of the processed food products.
1.3.2 (Partial) Electrolysis
Most food formulations subjected to ohmic heating contain more than 50% water.
During ohmic heating, the low-frequency AC induces electrolysis of the water generating
7
H2 and O2 gases at the electrode/solution interfaces. The corresponding anodic and
cathodic half-reactions, and the overall electrolysis reaction are as follows.
Anodic half-reaction:
2H2O (liquid) ⇔ O2 (g) + 4H+(aqueous) + 4e (1.3)
Cathodic half-reactions:
2H+(aqueous) + 2e ⇔ H2 (g) (1.4)
2H2O (liquid) + 2e ⇔ H2 (g) + 2OH−(aqueous) (1.5)
The overall reaction:
2H2O(liquid) ⇔ 2H2 (g) + O2 (g) (1.6)
Molecular oxygen generated by electrolysis can oxidize almost all the oxidizable
food components, particularly lipids and vitamins like ascorbic acid (Vitamin C). The
molecular oxygen also involves electrode corrosion, or formation of insulating species on
the electrode surfaces partially passivating the electrodes (Tzedakis et al., 1999). Because
of the high flammability and explosive nature, uncontrolled liberation of hydrogen gas
might pose safety concerns in large-scale continuous ohmic heating practice.
However, the liberation of gas bubbles at the electrode/solution interfaces does not
necessarily indicate the overall electrolysis reaction (equation 1.6). Sometimes, the
anodic half-reaction for electrode corrosion (equation 1.2) may be accompanied with one
of the cathodic half-reactions for electrolysis (equation 1.4 or 1.5) resulting in electrode
corrosion with H2 (g) generation. In addition, since some electrode materials show low
overpotential for Cl2 (g) liberation than oxidation of water to O2 (g), one of the cathodic
half-reactions for electrolysis (equation 1.4 or1.5) may be also coupled with the following
8
anodic half-reaction (equation 1.7), especially when there is a significant amount of
chloride ions in the heating medium, resulting in H2 (g) and Cl2 (g) generation.
2Cl −(aqueous) ⇔ Cl2 (g) + 2e (1.7)
1.3.3 Generation of free radicals
Electron transfer associated with electrochemical reactions at the electrode/solution
interfaces leads to generation of radical species (Schafer, 2001). However, such radical
generation specifically during ohmic heating has not yet been reported. Konya (1979)
described the formations of hydroxyl (•OH) and hydroperoxyl (•OOH) radicals, and
hydrogen peroxide (H2O2) in oxygen evolution (equation 1.3) during the electrolysis of
water. The cathodic half-reactions of hydrogen generation (equations 1.4 and 1.5) are
also mediated via hydride radicals (H•) (Sawyer, 2003). Some (hypothetical) mechanisms
of generating free radicals by electrolysis and electrode corrosion are illustrated in figure
1.3.
Since electrolysis and corrosion reactions occur in the microenvironments of the
electrodes, the radical species formed under ohmic heating conditions might be H•, and
oxygen-containing free radicals, such as •OH, •OOH, and superoxide anion radicals
(O2•−), as well as the molecules like H2O2 and singlet oxygen (1O2). These reactive
oxygen species can aggressively attack food components, in particular lipids, vitamins,
and amino acids causing oxidative degradation of those nutrients. However, on the other
hand, since the above reactive oxygen species can function as bactericides, the generation
of free radicals might improve the sterilization efficiency of ohmic heating.
9
1.3.4 Loss of energy
In ohmic heating, the current is exclusively for the purpose of heating, and no
electrochemical phenomena are desirable. However, at any time, total current (I total)
across an electrode/solution interface is given by equation 1.8 (Tzedakis et al., 1999).
I total = I c + I f (1.8)
The current that involves heat generation by passing through the food material is
obviously the capacitive current (I c). The faradaic current (I f) is associated with
electrode/solution interfaces, and causes electrochemical reactions. Therefore, I f can be
regarded as a ‘stray’ current, and it is essentially a loss of useful electrical energy.
Tzedakis et al. (1999) reported that the ratio of I f / I c could even be 20- 40 %.
1.4 Research objectives
The overall objective of this research was to acquire a better understanding of
electrochemical behavior of ohmic heaters. The following were the specific objectives of
the investigations.
1. To investigate the behavior of some electrodes at different pH values during ohmic
heating;
2. To test the feasibility of using pulse inputs to minimize electrochemical processes;
3. To characterize electrochemical processes during ohmic heating of ascorbic acid;
4. To investigate free radical generation during ohmic heating.
Various analytical techniques were used to characterize and quantify the
electrochemical, and subsequent chemical reactions. The metal ions migrated into the
heating medium were measured by state-of-the-art Inductively Coupled Plasma (ICP)
10
spectrometers. Results from this research contribute to the production of safe and high
quality ohmically processed food products, and to smooth and efficient ohmic heating
practice.
11
SYMBOLS
AC alternating current
DC direct current
e electron
I current (A)
Pheat amount of heat liberation (W)
R electric resistance (Ω)
V voltage (V)
12
REFERENCES
Amatore C., Berthou M., and Hebert S.(1998); Fundamental principles of electrochemical ohmic heating of solutions; Journal of Electroanalytical Chemistry; 457, pp. 191-203. Assiry A.M (1996); Effect of ohmic heating on the degradation kinetics of ascorbic acid; Ph.D thesis, The Ohio State University. Assiry A., Sastry S.K., and Samaranayake C. (2003); Degradation kinetics of ascorbic acid during ohmic heating with stainless steel electrodes, Journal of applied electrochemistry, 33 (2), pp. 187-196. Bently R., and Prentice T.R. (1957); The alternating current electrolysis of concentrated acids; J. Appl. Chem.; 7 (November), pp. 619-626. Bosch R.W., and Bogaerts W.F.(1998); A theoretical study of AC-induced corrosion considering diffusion phenomena; Corrosion Science, 40(2/3), pp.323-336. Butz P., and Tauscher B. (2002); Emerging technologies: chemical aspects; Food Research International, 35, pp.279-284. Getchell B.E.(1935); Electric pasteurization of milk; Agr. Eng., 16(10), pp.408-410.
Konya J. (1979); Notes on the “non-faraday” electrolysis of water; Journal of electrochemical society: electrochemical science and technology, 126(1), pp. 54-56. Kulman F.E. (1961); Effects of alternating currents in causing corrosion; Corrosion; 17(3), pp. 34-35. Lalvani S.B. and Lin X.A. (1994); A theoretical approach for predicting AC-induced corrosion; Corrosion Science, 36(6), pp. 1039-1046. Lalvani S.B. and Lin X.A. (1996); A revised model for predicting corrosion of materials induced by alternating voltages; Corrosion Science, 38(10), pp. 1709-1719. Mizrahi S., Kopelman I.J., and Perlman. J.(1975); Blanching by electroconductive heating; J. Food Technology, 10, pp. 281-288. Moses B.D. (1938); Electric pasteurization of milk; Agr. Eng., 19(12), pp.525-526. Naveh D., Kopelman I.J., and Mizrahi S. (1983); Electroconductive thawing by liquid contact; J. Food Technology, 18, pp. 171-176.
13
Palaniappan S., and Sastry S. (1991); Electrical conductivity of selected juices: Influences of temperature, solid content, applied voltage, and practical size; Journal of Food Process Engineering, 14, pp. 247-260. Parrott D.L.(1992); Use of ohmic heating for aseptic processing of food particulates; J. Food Technology, 12, pp. 68-72. Reznick D.(1996); ohmic heating of fluid foods; J. Food Technology; 5, pp. 250-251. Rubinstein I. (1995); Physical electrochemistry: principles, methods, and applications, Rubinstein I. (Ed.), Chapter 1: Fundamentals of physical electrochemistry; Marcel Dekker, Inc., New York, pp. 1-4. Sawyer D.T. (2003); Electrochemical transformations of metals, metal compounds, and metal complexes: invariably (ligand/ solvent)-centered; Journal of molecular catalysis A: Chemical, 194, pp. 53-67. Schafer H.J.(2001); Organic electrochemistry (Fourth edition), Lund H. and Hammerich O. (Ed.), Chapter 4: Comparison between electrochemical reactions and chemical oxidations and reductions; Marcel Dekker, Inc., New York, pp. 207-221 Shaw M. and Remick A.E.(1950); Studies on alternating current electrolysis; J. of Electrochemical Society; 97(10), pp. 324-334. Stirling R. (1987); ohmic heating - a new process for the food industry; Power Engineering Journal; 1(6), pp. 365-371. Tzedakis T., Basseguy R., and Comtat M.(1999); Voltammetric and coulometric techniques to estimate the electrochemical reaction rate during ohmic sterilization; Journal of Applied Electrochemistry; 29(7), pp. 821- 828. Uemura K., Noguchi A., Park S.J., and Kim D.U. (1994); ohmic heating of food materials- Effect of frequency on the heating rate of fish protein; Developments in Food Engineering - Proceedings of the 6 th International Congress on Engineering and Food; Blackie Academic & Professional Press, London, pp. 310-312. Venkatachalam S. and Mehendale S.G.(1981); Electrodissolution and corrosion of metals by alternating currents; Journal of Electrochemical Society, India; 30-3, pp. 231-237. Venkatesh S. and Chin D. (1979); The alternating current electrode processes; Israel Journal of Chemistry, 18, pp.56-64.
14
Wu H., Kolbe E., Flugstad B., Park J.W., and Yongsawatdigul J.(1998); Electrical properties of fish mince during multi-frequency ohmic heating; Journal of Food Science; 63(6), pp. 1028-1032. Zoltai P. and Swearingen P. (1996); Product development considerations for ohmic processing; J. Food Technology, 5, pp. 263-266.
15
I Electrodes
Alternating ~ Food V R current power supply
Electrical analogue ohmic heating
Figure 1.1: The concept of ohmic heating
Cd
Rs
Rct
Figure 1.2: A simplified electrical equivalent circuit of the interface during the application of AC; Cd: electrical double layer capacitor, Rct: charge-transfer resistance, Rs: electrolyte resistance.
16
Figure 1.3: Mechanisms of generating free radicals by electrolysis and electrode corrosion.
Mn+(aqueous)
3O2 (g) (Triplet dioxygen)
Electrolysis
Electrode corrosion
H• H2 (g)
H2O2
H+ + e
•OOH •OH
e
OH −
O2•− H+
1O2 (g) (Singlet oxygen)
Fenton’s ReactionOCl −
Cl − + H2O
H+ H2O2
Fe3+
Cl2 (g)
Cl•
H2O
Chlorine generation H+ (or H2O)
M2+(aqueous)
Fe2+
O2•−
OH −
17
CHAPTER 2
ELECTROCHEMICAL BEHAVIOR OF VARIOUS ELECTRODE MATERIALS DURING OHMIC HEATING AT pH 3.5, 5.0, AND 6.5
ABSTRACT
Undesirable electrochemical phenomena at electrode/solution interfaces during
ohmic heating can be avoided or effectively inhibited by choosing an appropriate
electrode material. We attempted to understand the electrochemical behavior of four
types of electrode materials: titanium, stainless steel, platinized-titanium, and graphite at
pH 3.5, 5.0, and 6.5. The electrodes were comparatively examined using 60 Hz sinusoidal
alternating current at a RMS voltage of 110 V. Analyses of surface morphologies of the
electrode surfaces, electrode corrosion, hydrogen gas generation, and pH change of the
heating medium were performed. The results highlight the relatively inert
electrochemical behavior of platinized-titanium electrodes at all the pH values. Pilot scale
study at 39.8 kW further demonstrates the potential use of platinized-titanium electrodes
for ohmic heating with commonly available low-frequency alternating currents. The
amounts of migrated Pt and Ti due to electrode corrosion were well below dietary
exposure limits of those elements.
18
INTRODUCTION Electrodes in ohmic heating can be regarded as a ‘junction’ between a solid-state
conductor (i.e. current feeder) and a liquid-state conductor (i.e. heating medium). They
play a vital role by conveying the current uniformly into the heating medium. Various
materials, so far, have been used as electrodes in different ohmic heating studies and
applications. Those materials include platinized-titanium (Stirling, 1987; Tzedakis et al.,
1999), platinum (Tzedakis et al., 1999), titanium (Assiry, 1996), aluminum (Mizrahi et al,
1975; Uemura et al., 1994), carbon/graphite (Gatchell, 1935; Moses, 1938),
Platinized-titanium electrodes Platinization of titanium electrodes has been a popular choice because of the high
cost of pure platinum electrodes for industrial processes (Indira et al., 1968; Iniesta et al.,
1999). On the other hand, platinization is also an effective method of passivating titanium
(James et al., 1976). The platinized- titanium electrodes exhibited significantly lower (p ≤
0.05) corrosion rates compared to those of the stainless steel and graphite electrodes at all
the pH values. Further, there were no signs of hydrogen or any other gas evolution at the
electrode/solution interfaces, and also no detectable pH change of the heating medium at
any pH. The rich double layer capacitance, as indicated by the SEM analysis, would be
the major reason for this superior electrode performance. Tzedakis et al. (1999) have
already demonstrated the superiority of platinized-titanium electrodes over platinum
electrodes for ohmic sterilization of food products with low-frequency alternating
33
currents. Iniesta et al. (1999) reported that the large surface area of platinized-titanium
electrodes also affects adsorption controlled surface processes, such as hydrogen
adsorption-desorption, and the surface oxidation.
Graphite electrodes Graphite, one of the allotropic forms of carbon, has been used as an electrode
material in electrochemical applications for a long time. Although there are various types
of commercially available graphitic carbons, polycrystalline graphite (PCG) is the
material most often referred to as ‘Graphite’ (McCreery, 1999). In spite of the very rich
double layer capacitance as indicated by SEM analysis, the corrosion rates of graphite
electrodes at all the pH values were significantly greater (p ≤ 0.05) than those of titanium
and platinized-titanium electrodes. However, as in the case of platinized-titanium
electrodes, there were no signs of gas evolution at the electrode/solution interfaces, and
also no detectable pH change of the heating medium at any pH. The above
electrochemical behavior can be explained by means of the chemical structure of
graphite, as follows.
Graphite consists of sp2 hybridized carbon atoms arranged as parallel sheets of
hexagonal rings. Since sp2 hybridized carbon is capable of forming covalent bonds and
has a propensity towards adsorption of a broad range of substances, graphite electrode
surfaces usually contain various functional groups and oxides (McCreery, 1999).
Therefore, the migration of surface functional groups and oxides as organic compounds
was anticipated during ohmic heating. In the analysis of those migratory corrosion
products, the heating medium exhibited about – 0.03 pH units change indicating the
34
migration of acidic organic compounds. The observed corrosion rate was 6.1 ± 0.1 ppb
KJ-1. Figure 2.11 shows the identified organic compounds by GC-MS analysis. The
quasimolecular ion [M+H]+ peaks observed at m/z 91 and m/z 133 in the ESI-MS
spectrum of the heating medium after the ohmic heating treatment (see figure 2.12)
correspond well with the identified polar organic compounds by the GC-MS. In addition,
the peaks observed at m/z 109 and m/z 123 may represent [M+H]+ ions for ortho- or
para-quinones (C6H4O2) and their corresponding methylated counterparts, respectively.
Such types of quinones are reported to be present on most carbon surfaces (McCreery,
1999; Tarasevich et al., 1987). It is clearly seen that the organic compounds migrated into
the heating medium always contain more than one carbon atom per molecule. Therefore,
migration of even a few molecules results in intense corrosion rates if the corrosion is
measured as elemental carbon.
Since sp2 hybridized carbon has a high affinity towards oxygen (McCreery, 1999),
surface oxides and oxygen-containing functional groups could be formed by reacting
with atmospheric oxygen even before using the graphite as electrodes, and also due to the
electrochemical reactions during ohmic heating. Soffer et al. (1972) suggested the
following anodic and cathodic half-reactions related to the electrolysis of water creating
oxides and functional groups on graphite electrode surfaces.
Anodic half-reactions:
C + H2O (liquid) ⇔ C − O + 2H +(aqueous) + 2e (2.14)
C + H2O (liquid) ⇔ C − OH + H +(aqueous) + e (2.15)
35
Cathodic half-reaction:
C + H2O (liquid) + e ⇔ C − H + OH−(aqueous) (2.16)
where ‘C’ represents sp2 hybridized carbon on the graphite surface. During ohmic
heating, once a set of carbon atoms leaves the graphite surface as organic compounds a
new set of sp2 hybridized carbon atoms in the graphite structure is exposed to the heating
medium, and keep undergoing the corrosion process. The migration of compounds into
the heating medium could be due to thermal, electric field, and pH (of the heating
medium) effects. The significantly high (p ≤ 0.05) corrosion rate observed at pH 3.5 may
be due to acid catalyzed hydrolyses of ester and ether linkages on the graphite electrode
surfaces facilitating the migration of functional groups as compounds than at the other pH
values.
The equations 2.14 - 2.16 basically indicate the adsorption of electrolysis products on
the electrodes causing oxidation and reduction of the surfaces, ultimately creating some
functional groups. Soffer et al. (1972) also reported the adsorption of chloride ions on the
graphite electrodes according to the following anodic half-reaction.
C + Cl −(aqueous) ⇔ C− Cl + e (2.17)
Such adsorption processes as well as very rich double layer capacitance apparently
inhibited the gas evolution at the electrode/solution interfaces. Although there were no
detectable pH changes of the heating media in the presence of citric acid and sodium
bicarbonate, pH change of the electrolyte due to positive and negative charging of the
electrodes is considered to be a unique property of high surface area graphite electrodes
(Soffer et al., 1972). In general, graphite electrodes are also considered to have a wider
faradaic reaction-free potential window compared to that of the metallic electrodes
36
(McCreery, 1999). However, ordinary faradaic processes, such as generation of hydrogen
and oxygen due to the electrolysis of water without adsorption on the surfaces, can also
take place irrespective of the chemical nature of the surface groups (Soffer et al., 1972).
Pilot scale study of electrode corrosion In the laboratory scale studies, platinized-titanium exhibited the best electrode
performance out of the four electrodes tested. Therefore, it was subjected to further
investigation on a pilot scale for electrode corrosion. Table 2.3 shows Pt and Ti
concentrations of the ohmically heated medium after running through the heat exchanger
when the ohmic heater was at the steady state. The concentrations of Pt and Ti in the
blank (i.e. the heating medium in the feed tank before being subjected to ohmic heating)
were too low for reliable measurements. Since platinized-titanium exhibited the highest
corrosion at pH 3.5 in the laboratory scale studies, the values shown in table 2.3 would be
the ‘worst-case’ concentrations. Using those concentrations, intakes of the metal
contaminants were evaluated with respect to a typical meal of 8 oz (227 g) comparing
with recently published upper-level daily dietary exposure limits for adult consumers (see
table 2.4). It can be seen that the estimated metal intakes via consumption of an
ohmically heated meal of 8 oz are far below the published upper-level daily dietary
exposure limits. Therefore, ohmic heating may be performed in pilot scale without
significant electrode corrosion using platinized-titanium electrodes; and the migrations of
Pt and Ti may result in concentrations that are far below the published dietary exposure
limits.
37
CONCLUSIONS
Using the alternating frequency of 60 Hz, we demonstrated that electrochemical
behavior of an electrode material is unique to the material itself. Although, in general, the
large microscopic surface area can suppress the electrochemical processes, the type and
extent of electrochemical reactions are determined by the chemical nature of the electrode
surface, as well as the pH of the heating medium. All the electrode materials exhibited
intense electrode corrosion at pH 3.5 compared to that of the other pH values. Although
the titanium electrodes were having a relatively high corrosion resistance, apparent
electrolysis was seen at all the pH values during ohmic heating. Stainless steel was found
to be the most electrochemically active electrode material during ohmic heating at all the
pH values. It was proven that, the intense corrosion of graphite electrodes was due to the
migration of surface functional groups and oxides as organic compounds during ohmic
heating; and the pH of the heating medium seemed to facilitate such migrations. Because
of the relatively inert electrochemical behavior, platinized-titanium would be the
electrode material-of-choice for ohmic heating at all the pH values. The potential use of
platinized-titanium electrodes for ohmic heating operations was further demonstrated in
pilot scale at 39.8 kW; and the concentrations of migrated Pt and Ti were far below the
published dietary exposure limits.
38
SYMBOLS
e electron
Irms RMS current (A)
MW molecular weight
m/z mass to charge ratio
Pinput power input (W)
ppb parts per billion
ppm parts per million
ppt parts per trillion
RMS root-mean-square
Vrms RMS voltage (V)
39
REFERENCES
Amatore C., Berthou M., and Hebert S.(1998); Fundamental principles of electrochemical ohmic heating of solutions; Journal of Electroanalytical Chemistry; 457, pp. 191-203. Assiry A.M (1996); Effect of ohmic heating on the degradation kinetics of ascorbic acid; PhD thesis, The Ohio State University. Assiry A., Sastry S.K., and Samaranayake C. (2003); Degradation kinetics of ascorbic acid during ohmic heating with stainless steel electrodes, Journal of applied electrochemistry, 33 (2), pp. 187-196. Getchell B.E.(1935); Electric pasteurization of milk; Agr. Eng., 16(10), pp.408-410. Indira K.S, Sampath S., and Doss K.S.G.(1968); Recent trends of platinized titanium as node material in electrochemical industries; Chemical processing & Engineering (February), pp. 35-37. Iniesta J., Gonzalez-Garcia J., Fernandez J., Montiel V., and Aldaz A. (1999); On the Voltammetric behavior of platinized titanium surface with respect to the specific hydrogen and anion adsorption and charge transfer processes; Journal of materials chemistry, 9 (12), pp. 3141-3145. James W.J., and Straumanis M.E.(1976); Encyclopedia of electrochemistry of the elements, Bard A.J. (Ed.), Vol.(V), Chapter V-7 : Titanium; Marcel Dekker Inc., New York, pp. 305 – 386. McCreery R.L.(1999); Interfacial Electrochemistry: Theory, Experiment, and Applications, Wieckowski A. (Ed.), Chapter 35 : Electrochemical properties of carbon surfaces; Marcel Dekker Inc., New York, pp. 631 – 646. Mizrahi S., Kopelman I.J., and Perlman. J.(1975); Blanching by electroconductive heating; J. Food Technology, 10, pp. 281-288. Moses B.D. (1938); Electric pasteurization of milk; Agr. Eng., 19(12), pp.525-526. Official methods of analysis of AOAC International – 17th edition (2000); Vol.1 (Agricultural Chemicals, Contaminants, Drugs), Ch.9: metals and other elements at trace levels in foods, pp. 46-59.
40
Palaniappan S., and Sastry S. (1991); Electrical conductivity of selected juices: Influences of temperature, solid content, applied voltage, and practical size; Journal of Food Process Engineering, 14, pp. 247-260. Redmond J.D. (1996); Marks’ Standard handbook for mechanical engineers (Tenth Edition); McGraw – Hill Companies, Inc., pp. 6(32) – 6(33). Reilly C.; Metal contamination of food, second edition (1991); Elsevier Science Publishers Ltd., New York; pp. 14-15 & 235-237. Soffer A., and Folman M. (1972); The electrical double layer of high surface porous carbon electrodes; Journal of Electroanalytical Chemistry, 38, pp.25 – 43. Stirling R. (1987); Ohmic heating - a new process for the food industry; Power Engineering Journal; 1(6), pp. 365-371. Tarasevich M.R., Bogdanovskaya V.A., and Zagudaeva N.M.(1987); Redox reactions of quinones on carbon materials; Journal of Electroanalytical Chemistry, 223, pp.161-169. Tomat R., and Rigo A. (1979); Oxidation of polymethylated benzenes promoted by •OH radicals; Journal of Applied Electrochemistry, (9), pp. 301-305. Tzedakis T., Basseguy R., and Comtat M. (1999); Voltammetric and coulometric techniques to estimate the electrochemical reaction rate during ohmic sterilization; Journal of Applied Electrochemistry; 29(7), pp. 821- 828. Uemura K., Noguchi A., Park S.J., and Kim D.U. (1994); ohmic heating of food materials- Effect of frequency on the heating rate of fish protein; Developments in Food Engineering - Proceedings of the 6 th International Congress on Engineering and Food; Blackie Academic & Professional Press, London, pp. 310-312. Venkatachalam S. and Mehendale S.G.(1981); Electrodissolution and corrosion of metals by alternating currents; Journal of Electrochemical Society, India; 30-3, pp. 231-237. Wu H., Kolbe E., Flugstad B., Park J.W., and Yongsawatdigul J.(1998); Electrical properties of fish mince during multi-frequency ohmic heating; Journal of Food Science; 63(6), pp. 1028-1032. Ysart G., Miller P., Crews H., Robb P., Baxter M., De L’Argy, Lofthouse S., Sargent C., and Harrison N. (1999); Dietary exposure estimates of 30 elements from the UK total diet study; Food Additives and Contaminants, Vol.16 (9), pp. 391-403.
41
Corrosion rate* (ppb/ KJ)
Electrode type pH = 3.5 pH = 5.0 pH = 6.5
Titanium(Ti) 0.26 a (0.21) 0.03 a (0.01) 0.05 a (0.03)
Stainless steel (Fe) 14.20 b (1.95) 8.33 c (0.30) 11.43 b,e (1.51)
Platinized-titanium (Pt) 0.25 a (0.10) 0.07 a (0.04) 0.05 a (0.02)
Graphite (C) 26.6 d (2.2) 7.2 c (0.0) 8.4 c,e (1.1)
* Numbers in parentheses represent standard deviations of the means (n=3). Different superscript letters with the means denote significant differences (p ≤ 0.05). Table 2.1: Comparison of corrosion rates (in ppb per KJ) of the electrode materials with respect to the migrations of their major (surface) elements at different pH values.
pH = 3.5
pH = 5.0
pH = 6.5
∆ pH*
+ 0.04 a (0.01)
+ 0.18 b (0.03)
+ 0.10 c (0.01)
* Numbers in parentheses represent standard deviations of the means (n=3). Different superscript letters with the means denote significant differences (p ≤ 0.05). Positive signs indicate increase of pH. Table 2.2: pH changes of the heating media observed with stainless steel electrodes at different pH values.
42
Element
Concentration* / ppt
Pt 61.6 a (10.3)
Ti 69.2 a (14.6)
* Numbers in parentheses represent standard deviations of the means (n=6). The means with the same superscript letter are not significantly different (p > 0.05). Table 2.3: Pt and Ti concentrations (in parts per trillions) of the ohmically heated heating medium in the pilot scale study.
Element
Estimated intake via 8 oz
meal (µg)
Published upper-level daily
dietary exposure limits (µg/ day)
Pt 0.014 0.3*
Ti 0.016 600**
* Ysart et al. (1999) ** Reilly (1991) Table 2.4: Comparison of estimated metal intakes via consumption of an 8 oz ohmically heated meal with the published upper-level daily dietary exposure limits for adult consumers. The estimation is based on unit conversions: 1 ppt = 1 picogram/ g; 8 oz = 227 g; 1 picogram = 10-12 g = 10-6 µg.
6 – Removable lid: the electrodes, hydrogen sensor, and thermocouple, are attached. It was tightly clamped to the cell body during ohmic heating. Figure 2.1: The laboratory scale ohmic heater.
1
2
3
4
5
6
44
Figure 2.2: Schematic diagram of the laboratory scale experimental setup.
60 Hz/ 0-110 V ~ Public utility Supply
Isolationmodule
Hydrogen gas meter 0 – 250 ppm
Data logger
Microcomputer
V
AVariac
V = voltage transducer A = Current transducer
Thermocouple
ohmic heater
45
0
10
20
30
40
50
0 100 200 300time/ seconds
Tem
p./ 0 C
Figure 2.3: Typical time vs. temperature curve for all electrodes at all pH values during ohmic heating.
0.00
0.50
1.00
1.50
2.00
0 100 200 300
time/ seconds
Cur
rent
/ A
Figure 2.4: Typical time vs. current curve for all electrodes at all pH values during ohmic heating.
46
Figure 2.5: Typical SEM micrograph of titanium electrodes.
47
Figure 2.6: Typical SEM micrograph of stainless steel electrodes.
48
Figure 2.7: Typical SEM micrograph of platinized-titanium electrodes.
49
Figure 2.8: Typical SEM micrograph of graphite electrodes.
50
0
50
100
150
200
250
0 100 200 300time/ seconds
[H2]ppm
pH = 3.5pH = 5.0pH = 6.5
Figure 2.9: Hydrogen generation with titanium electrodes during ohmic heating.
0
50
100
150
200
250
0 100 200 300time/ seconds
[H2]ppm
pH = 3.5pH = 5.0pH = 6.5
Figure 2.10: Hydrogen generation with stainless steel electrodes during ohmic heating.
monohydrate (Aldrich, WI); and trace metal grade concentrated nitric acid (Fisher
Scientific, PA) were purchased from the suppliers. Demineralized double distilled water
(Resistivity: 3 megohm; pH: 5.5) was obtained from the Reagent Laboratory Store, at
The Ohio State University.
Experimental setup
The laboratory scale ohmic heater shown in figure 3.1 was used for both pulsed and
conventional (60 Hz, sine wave) ohmic heating experiments. The ohmic heater was
connected to an IGBT power supply that was capable of delivering bipolar potential
pulses, as shown in figure 3.2. The IGBT power supply had a fixed peak voltage (Vp) of
170 V with switching frequency up to 10 kHz. The same experimental setup attached to a
variac and powered by the public utility supply, instead of the IGBT power supply, was
used for the conventional ohmic heating experiments. Stainless steel (316), titanium,
platinized-titanium, and graphite were used as electrodes in all the experiments, except in
the study on delay-time effects, where only stainless steel was used. All electrodes were
rectangular (7.5 cm × 5.2 cm) with a slight curvature (radius ~ 4.5 cm) (see figure 3.1),
and had the same geometric dimensions. The effective geometric surface area of each
56
electrode involved in ohmic heating was constant at 21.5 cm2 with a volume of 250.00 (±
0.12) ml of heating medium.
An external cooling system was operated by a Haake F3 Fisons thermostatic water
bath having inflow and outflow attached to the ohmic heater. The cooling was required
for precise matching of time-temperature history of each experiment (see figure 3.3).
Since reaction rates depend upon temperature, performing the experiments under equal
temperature conditions was considered necessary to eliminate temperature as a variable.
The cooling also allowed for prolonged heating exposure at relatively low temperatures
minimizing safety risks while increasing the extent of electrochemical reactions, and
facilitating detection.
Heating media
Experiments were performed at an initial pH of 3.5 (at 25 °C) using freshly prepared
aqueous heating media. This particular pH value was specifically chosen, since it was
determined from our previous studies (chapter 2) to be the pH of the worst-case scenario
for all electrode materials with respect to corrosion. The desired pH value of the heating
media was achieved by citric acid. Initial electrical conductivity was adjusted by NaCl,
and was varied for different experiments (described below) to maintain the same time-
temperature history. The above components used for the pH and conductivity adjustments
are common ingredients in food formulations. The heating media were not buffered to
determine if pH changes were caused by electrochemical reactions.
57
Experimental procedure
Effects of frequency and pulse width:
To study the effects of frequency and pulse width, two switching frequencies, 10 and
4 kHz, were chosen. These were considered as representing upper and lower frequency
ranges for minimized electrode corrosion, based on the findings of Wu et al. (1998) who
reported drastically reduced corrosion of stainless steel electrodes at frequencies > 5 kHz.
Figure 3.4 shows the centering of bipolar pulses within the period (T) at a given
frequency (f). Both positive and negative pulses of the bipolar pulse inputs had the same
pulse width (tp), and were equally spaced by adjusting the delay time (td) according to the
following relationship.
T = 2 (tp + td) (3.1)
The values for pulse width were arbitrarily chosen allowing at least 15 µs delay time.
From preliminary experiments, we found that about a 10 – 15 µs delay time was
necessary to prevent hydrogen generation, and to yield symmetric positive and negative
pulses. Since varying the frequencies and pulse widths essentially varies the duty cycle
(θ) and thereby power input (Pinput) (see equations 3.2 and 3.3), the uses of different
initial electrical conductivities and cooling water temperatures were necessitated to
maintain the same time-temperature history. Under these circumstances, all data used for
comparison were normalized on the basis of unit energy input.
θ = 2tp / T (3.2)
Pinput = VpIpθ (3.3)
A volume of 250.00 (± 0.12) ml was subjected to ohmic heating in each experimental
run. Tables 3.1 – 3.4 show the frequency-pulse width combinations with required initial
58
electrical conductivities, and cooling water temperatures for various electrode materials
used in this study. Duration of heating was kept constant with all the experiments for the
purpose of comparison (184 seconds). The shapes of corresponding pulse waveforms for
voltage and current are shown in figures 3.5 and 3.6. The conventional ohmic heating
was carried out with all electrodes at a RMS voltage of 110 V, which is the common
single-phase public utility supply voltage in the US.
The ohmic heating experiments of each electrode material were completely
randomized with respect to order of experimentation. The electrodes were thoroughly
rinsed using demineralized double distilled water before each run. All experiments were
triplicated; and adherent films formed on the stainless steel and titanium electrode
surfaces during ohmic heating were removed by brushing and cleaning after each three
replicates. Analyses of electrode corrosion, hydrogen generation, and pH measurements
were performed according to the procedures described below.
Effect of delay time:
Preliminary experiments were conducted at the frequencies and pulse widths
described in tables 3.1 – 3.4 using all electrode materials, in parallel with the above
experiments. Figure 3.7 shows the centering of bipolar pulses within the period (T) at a
given frequency (f) obeying the following relationship.
T = 2tp + td1 + td2 (3.4)
The delay time denoted by td1 was varied from 0 to td1 = td2 (i.e. until pulses were equally
spaced). These experiments indicated hydrogen generation at all the frequencies and
pulse widths, and with all electrode materials, when td1 = 0 µs. Interestingly, unlike in
59
conventional ohmic heating, the gas evolution was seen only at the electrode to which the
neutral wire was attached. This hydrogen generation was more pronounced at 10 kHz
with shorter pulse widths (where heating media contained higher NaCl concentrations).
We further noticed that the hydrogen generation was gradually diminished with
increasing td1, and completely disappeared when td1 > 10 µs.
In order to illustrate the effect of delay time, a heating procedure with analysis of
electrode corrosion was carried out at the frequency of 10 kHz and the pulse width of
25.0 µs using stainless steel electrodes only. Stainless steel was specifically chosen, since
it exhibits pronounced electrochemical behavior compared to that of the other electrode
materials. The selection of frequency, 10 kHz, was simply to illustrate the enhanced
hydrogen generation, especially when td1 = 0 µs. However, shorter pulse widths (less than
25.0 µs) were avoided, because of more pronounced hydrogen generation that raised
safety concerns. The delay time, td1, was varied as 0.0, 5.0, 10.0, 15.0, 20.0, 25.0 µs. The
shapes of corresponding pulse waveforms for voltage and current are shown in figure 3.8.
All these experiments were randomized and triplicated as before. The results were also
compared with those of conventional ohmic heating.
Analysis of electrode corrosion Concentrations of Fe and Cr (from the stainless steel electrodes); Ti (from the
titanium electrodes); Pt and Ti (from the platinized-titanium electrodes); and elemental
carbon (from the graphite electrodes) migrated into the heating media were taken as
measures of electrode corrosion. In each experimental run, once the ohmic heating was
completed, a 25.00 (± 0.03) ml sample was pipetted out after removing the electrodes,
60
and thoroughly mixing the fluid. A 25.00 (± 0.03) ml sample of the respective unheated
heating medium was used as a method blank. All the samples were collected into
polypropylene sample bottles, and then stabilized by adding concentrated nitric acid (5%,
v/v). Quantitative analyses of the metal ions were performed by a Perkin-Elmer Sciex
ELAN 6100 DRC inductively coupled plasma - mass spectrometer (ICP-MS) (AOAC,
2000). The elemental carbon concentrations were determined by a Perkin-Elmer Optima
3000 DV inductively coupled plasma - optical emission spectrometer (ICP-OES)
monitoring the emission spectra near 193.03 nm.
Hydrogen generation and pH measurements A series U hydrogen detector (CEA Instruments, Inc., NJ) was used to measure
headspace hydrogen gas generated during ohmic heating. The pH of the medium before
and after the ohmic heating treatment was measured by a Cole-Parmer 59003 Benchtop
pH meter (resolution: 0.01 pH) at 25 °C.
Data analysis
Total energy input in the pulsed ohmic heating experiments was determined by
integrating the power input (equation 3.3) vs. time curve for each experimental run. The
total energy input was also calculated for conventional ohmic heating using its power
input (Pinput = Vrms Irms) vs. time curve. Concentration of metal migration normalized per
unit energy input was defined as ‘corrosion rate’. The corrosion rates were calculated for
each migratory element. Descriptive statistics including means and standard deviations
were calculated for the quantitative measurements. The corrosion rates with respect to
61
each migratory element were individually analyzed using one-factor analysis of variance
to determine if: (1) the frequency-pulse width combinations of pulsed ohmic heating
together with conventional ohmic heating; and (2) the delay times of pulsed ohmic
heating together with conventional ohmic heating; had significant effects on corrosion
rate. Tukey’s specific comparison test determined which particular means were
significantly different. Significance of differences was defined as p≤ 0.05. SPSS 11.5 for
windows (SPSS Inc., 2002) statistical software package was used for the statistical
analyses.
RESULTS AND DISCUSSION
Effects of frequency and pulse width
Stainless steel electrodes:
Figures 3.9 and 3.10 show the corrosion rates with respect to Fe and Cr, major
elements of stainless steel, at different frequencies and pulse widths, in comparison with
those of conventional ohmic heating. It can be seen that the corrosion rates become
enhanced at 4 kHz, compared to those at 10 kHz. It is also seen that significantly (p ≤
0.05) reduced corrosion rates can be achieved for the same duty cycle when a higher
frequency and a shorter pulse width are used. The figures further demonstrate that pulsed
ohmic heating is capable of significantly (p ≤ 0.05) reducing corrosion rates, compared
to conventional ohmic heating. In all pulsed ohmic heating experiments, there were no
signs of hydrogen or any other gas evolution at the electrode/solution interfaces, and also
no detectable pH change of the heating medium at any of the frequencies and pulse
widths. However, with conventional ohmic heating, 6 (± 2) ppm hydrogen gas
62
accumulation in the headspace, and + 0.04 pH change of the heating medium were
observed.
Titanium electrodes:
The corrosion rates of titanium electrodes become enhanced at 10 kHz longer pulse
widths, and at 4 kHz shorter pulse widths (see figure 3.11). Therefore, higher duty cycles
with reduced corrosion rates can be achieved using 4 kHz with longer pulse widths. The
figure further indicates significantly (p ≤ 0.05) reduced corrosion rates of pulsed ohmic
heating, compared with conventional ohmic heating. There were no signs of gas
evolution at the electrode/solution interfaces, and also no detectable pH changes of the
heating media in both pulsed and conventional ohmic heating experiments.
Platinized-titanium electrodes:
Figures 3.12 and 3.13 show the corrosion rates with respect to Pt and Ti at different
frequencies and pulse widths, in comparison with those of conventional ohmic heating.
As can be seen, the corrosion rates become greatly enhanced at 10 kHz with increasing
pulse width, whereas 4 kHz yields greatly reduced corrosion rates with all the pulse
widths. Therefore, higher duty cycles with reduced corrosion rates can be achieved using
lower frequencies and longer pulse widths in pulsed ohmic heating with platinized-
titanium electrodes. It is also seen that pulsed ohmic heating significantly (p ≤ 0.05)
reduces the corrosion rates with respect to Pt, compared to conventional ohmic heating.
However, conventional ohmic heating yields a reduced corrosion rate with respect to Ti,
compared to pulsed ohmic heating at 10 kHz. Both pulsed and conventional ohmic
heating experiments did not indicate any signs of gas evolution at the electrode/solution
interfaces; and also there were no detectable pH changes of the heating media.
63
Graphite electrodes:
Pulsed ohmic heating yields reduced corrosion rates at 10 kHz with shorter pulse
widths, and at 4 kHz with longer pulse widths, in the case of graphite electrodes (see
figure 3.14). However, it is seen that pulsed ohmic heating can enhance the corrosion
rate, compared to conventional ohmic heating. As in the cases of titanium and platinized-
titanium electrodes, we neither observed any signs of gas evolution, nor any detectable
pH changes of the heating media in both pulsed and conventional ohmic heating
experiments.
Effect of delay time
Figures 3.15 and 3.16 represent variations of the corrosion rates (with respect to Fe
and Cr) with delay time, in comparison with the corrosion rates for conventional ohmic
heating. It is evident that delay time has a significant effect on corrosion rate. When there
was no delay time (i.e. td1 = 0.0 µs), we observed 71 (± 25) ppm hydrogen gas
(selectively generated at the electrode to which the neutral wire was attached)
accumulation in the headspace, and a + 0.33 (± 0.02) pH change of the heating medium,
in addition to the significantly (p ≤ 0.05) greater corrosion rates. These observations can
be explained by means of the corresponding pulse waveforms as follows.
The shapes of the corresponding pulse waveforms shown in figure 3.8 indicate
markedly incomplete discharge of the double layers after each positive pulse, when there
was no delay time (see ‘Delay: 0 µs’). As a result, with continued pulsation, there was
likely to be negative charge (e.g. Cl − and citrate ions) accumulation in the vicinity of the
electrode to which the hot wire was attached, and, simultaneously, a positive charge (e.g.
64
H+ and Na+) accumulation at the electrode to which the neutral wire was attached. The
latter phenomenon led to selective hydrogen generation at that particular electrode,
together with a positive pH change (i.e. an increase of pH) due to the loss of H+ ions in
the heating medium (see equation 3.5). Then, in order to maintain electrical neutrality,
the electrode to which the hot wire was attached had to liberate (positive) metal ions into
the heating medium resulting in enhanced corrosion rates (see equation 3.6).
Cathodic half-reaction: 2H+(aqueous) + 2e ⇔ H2 (g) (3.5)
Anodic half-reaction: M (solid) ⇔ M2+(aqueous) + 2e, (where, M = Fe, Cr, Ni, Mo) (3.6)
The overall reaction:
M (solid) + 2H+(aqueous) ⇔ M 2+
(aqueous) + H2 (g) (3.7)
With increasing delay time, the corrosion rates were drastically reduced, and became
insignificant after 10 µs (see figures 3.15 and 3.16). We further noticed no signs of gas
evolution at any of the electrode/solution interfaces when td1 > 10 µs. Also, the heating
media did not indicate a detectable pH change when delay time was ≥ 5 µs. These
observations also relate to the shapes of pulse waveforms shown in figure 3.8, which
indicates gradual completion of discharge of the double layers yielding more symmetric
positive and negative pulses, with increasing delay time. It can be seen that the symmetry
of positive and negative pulses remains almost unchanged after the delay time of 10 µs.
The above results demonstrate the importance of allowing enough delay time for
discharge of the double layers after each pulse input. The results further indicate that
pulsed ohmic heating with insufficient delay times can be worse than conventional ohmic
heating. The symmetry of positive and negative pulses of the pulse waveforms may be
used as a reliable indicator to determine the sufficiency of delay time in pulsed ohmic
65
heating. On the other hand, the delay time requirement in pulsed ohmic heating limits the
accomplishment of higher duty cycles, especially at higher frequencies. For instance,
when allowing a delay time of 15 µs, the maximum duty cycle at 10 kHz is 70%,
compared with the 88% maximum duty cycle at 4 kHz.
66
CONCLUSIONS
Electrochemical reactions during ohmic heating with stainless steel, titanium, and
platinized-titanium electrodes can be significantly (p ≤ 0.05) reduced, in some cases, to
undetectable levels by use of IGBT pulse inputs. For stainless steel electrodes, pulsed
ohmic heating at higher frequencies and shorter pulse widths yields the lowest rates of
electrochemical reactions. However, pulsed ohmic heating at lower frequencies and
longer pulse widths is more effective in suppressing the electrochemical reactions of
titanium and platinized-titanium electrodes, while achieving higher duty cycles. In
general, pulsed ohmic heating is not capable of suppressing the electrochemical reactions
of graphite electrodes. Delay time was found to be a critical factor in pulsed ohmic
heating. The sufficiency of a given delay time is dependent on the symmetry of positive
and negative pulses of the pulse waveforms.
67
SYMBOLS
e electron
f frequency (Hz)
Ip peak current (A)
Irms RMS current (A)
Pinput power input (W)
ppb parts per billion
ppm parts per million
RMS root-mean-square
T period (µs)
td, td1, td2 delay times (µs)
tp pulse width (µs)
Vp peak voltage (V)
Vrms RMS voltage (V)
Greek letters
θ duty cycle
68
REFERENCES
Amatore C., Berthou M., and Hebert S.(1998); Fundamental principles of electrochemical ohmic heating of solutions; Journal of Electroanalytical Chemistry; 457, pp. 191-203. Baliga B. J.; Power semiconductor devices for variable frequency devices; IEEE Technology update series: Power electronics technology and applications II (selected conference papers), Lee F.C. ed. (1998), pp. 50-60. Grant D. (1996); Power semiconductor devices – continuous development; Microelectronics Journal, 27, pp. 161-176 Wu H., Kolbe E., Flugstad B., Park J.W., and Yongsawatdigul J.(1998); Electrical properties of fish mince during multi-frequency ohmic heating; Journal of Food Science; 63(6), pp. 1028-1032.
69
Frequency / Hz
Pulse width (µs), or RMS voltage (V)
Initial electrical conductivity/
mS cm-1 (at 25 0 C)
Cooling water temp. / (± 1) 0 C
10.0
3.75
31
15.0
2.97
28
25.0
2.23
26
10 000
35.0
1.90
23
30.0
3.34
31
62.5
2.28
27
75.0
2.04
25
100.0
1.76
20
4 000
110.0
1.64
19
60 (sine wave)
110 (Vrms)
2.59
29
Table 3.1: Frequency-pulse width combinations with required initial electrical conductivities, and cooling water temperatures for stainless steel electrodes.
70
Frequency / Hz
Pulse width (µs), or RMS voltage (V)
Initial electrical conductivity/
mS cm-1 (at 25 0 C)
Cooling water temp. / (± 1) 0 C
10.0
3.80
30
15.0
2.99
28
20.0
2.56
27
10 000
25.0
2.21
25
30.0
3.30
31
50.0
2.58
29
75.0
2.05
24
4 000
100.0
1.78
21
60 (sine wave)
110 (Vrms)
2.58
28
Table 3.2: Frequency-pulse width combinations with required initial electrical conductivities, and cooling water temperatures for titanium electrodes.
71
Frequency / Hz
Pulse width (µs), or RMS voltage (V)
Initial electrical conductivity/
mS cm-1 (at 25 0 C)
Cooling water temp. / (± 1) 0 C
10.0
3.76
29
15.0
2.94
27
20.0
2.51
25
10 000
25.0
2.17
23
30.0
3.24
28
50.0
2.53
26
75.0
2.02
24
4 000
100.0
1.76
21
60 (sine wave)
110 (Vrms)
2.53
26
Table 3.3: Frequency-pulse width combinations with required initial electrical conductivities, and cooling water temperatures for platinized-titanium electrodes.
72
Frequency / Hz
Pulse width (µs), or RMS voltage (V)
Initial electrical conductivity/
mS cm-1 (at 25 0 C)
Cooling water temp. / (± 1) 0 C
10.0
3.04
30
15.0
2.48
28
20.0
2.10
26
10 000
25.0
1.86
24
30.0
2.85
31
50.0
2.16
28
75.0
1.71
24
4 000
100.0
1.46
21
60 (sine wave)
110 (Vrms)
2.19
29
Table 3.4: Frequency-pulse width combinations with required initial electrical conductivities, and cooling water temperatures for graphite electrodes.
6 – Removable lid: the electrodes, hydrogen sensor, and thermocouple, are attached. It was tightly clamped to the cell body during ohmic heating 7 – Attachments of differential voltage probe
Figure 3.1: The ohmic heater used for both pulsed and conventional ohmic heating experiments.
1
2
3
4
5
6
7 7
74
Figure 3.2: Schematic diagram of the experimental setup used for pulsed ohmic heating.
V = Differential voltage probeA = Current monitor
Isolation module
Hydrogen gas meter 0 – 250 ppm
ohmic heater
~
Oscilloscope
Data logger
Microcomputer
V
A Digital multimeter
Rectifier
+
_
N
~
~
~ IGBT
Pulse generator
IGBT Power supply
60 H
z, 3
pha
se
Pow
er su
pply
75
0
10
20
30
40
50
0 50 100 150time/ seconds
Tem
p./ 0 C
Figure 3.3: Typical time vs. temperature curve for all the ohmic heating experiments.
76
Figure 3.4: Schematic diagram of the centering of bipolar pulses within the period to study the effects of frequency and pulse width.
td td/2 td/2
T = 1/ f
tp
tp
77
Figure 3.5: Typical pulse waveforms at 10 kHz for different pulse widths. The top and the bottom waves in each diagram represent the current and the voltage, respectively
10 µs 15 µs
25 µs
35 µs
20 µs
78
Figure 3.6: Typical pulse waveforms at 4 kHz for different pulse widths. The top and the bottom waves in each diagram represent the current and the voltage, respectively.
30 µs
62.5 µs 75 µs
100 µs 110 µs
50 µs
79
Figure 3.7: Schematic diagram of the centering of bipolar pulses within the period to demonstrate the effect of delay time.
td1 td2/2 td2/2
T = 1/ f
tp
tp
80
Figure 3.8: Typical pulse waveforms for different delay times. The top and the bottom waves in each diagram represent the current and the voltage, respectively.
Delay: 0 µs Delay: 5 µs
Delay: 10 µs Delay: 15 µs
Delay: 20 µs Delay: 25 µs
81
0.00
2.50
5.00
7.50
10.00
12.50
10(10)20%
15(10)30%
25(10)50%
35(10)70%
30(4)
24%
62.5(4)
50%
75(4)
60%
100(4)
80%
110(4)
88%
Con
[Fe]
in p
pb/ K
J
Figure 3.9: The corrosion rates (in ppb per KJ) of stainless steel electrodes with respect to Fe at different frequencies and pulse widths. Number shown in parenthesis is the corresponding frequency (kHz) for each pulse width (µs). The corresponding duty cycle is indicated as a percentage. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, b, c, d, and e denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates.
a
ab
aba
d
cd
bc bcab
e
82
0.00
0.30
0.60
0.90
1.20
1.50
1.80
10(10)20%
15(10)30%
25(10)50%
35(10)70%
30(4)
24%
62.5(4)
50%
75(4)
60%
100(4)
80%
110(4)
88%
Con
[Cr]
in p
pb/ K
J
Figure 3.10: The corrosion rates (in ppb per KJ) of stainless steel electrodes with respect to Cr at different frequencies and pulse widths. Number shown in parenthesis is the corresponding frequency (kHz) for each pulse width (µs). The corresponding duty cycle is indicated as a percentage. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, b, c, and d denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates.
a a
a
ab
c
bc
abc abcab
d
83
0.00
0.05
0.10
0.15
0.20
10(10)20%
15(10)30%
20(10)40%
25(10)50%
30(4)
24%
50(4)
40%
75(4)
60%
100(4)
80%
Con
[Ti]
in p
pb/ K
J
Figure 3.11: The corrosion rates (in ppb per KJ) of titanium electrodes with respect to Ti at different frequencies and pulse widths. Number shown in parenthesis is the corresponding frequency (kHz) for each pulse width (µs). The corresponding duty cycle is indicated as a percentage. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, b, c, d, and e denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates.
ab
bcdd
a
cd
abc
a abc
e
84
0.000
0.025
0.050
0.075
0.100
10(10)20%
15(10)30%
20(10)40%
25(10)50%
30(4)
24%
50(4)
40%
75(4)
60%
100(4)
80%
Con
[Pt]
in p
pb/ K
J
Figure 3.12: The corrosion rates (in ppb per KJ) of platinized-titanium electrodes with respect to Pt at different frequencies and pulse widths. Number shown in parenthesis is the corresponding frequency (kHz) for each pulse width (µs). The corresponding duty cycle is indicated as a percentage. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, b, c, and d denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates.
ab ab
ab
c
bab
ab a
d
85
0.000
0.025
0.050
0.075
0.100
10(10)20%
15(10)30%
20(10)40%
25(10)50%
30(4)
24%
50(4)
40%
75(4)
60%
100(4)
80%
Con
[Ti]
in p
pb/ K
J
Figure 3.13: The corrosion rates (in ppb per KJ) of platinized-titanium electrodes with respect to Ti at different frequencies and pulse widths. The presence of asterisk (*) indicates corrosion rate either < 0.001 ppb/ KJ or undetectable. Number shown in parenthesis is the corresponding frequency (kHz) for each pulse width (µs). The corresponding duty cycle is indicated as a percentage. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, and b denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates.
ab
ab
a
a
b b b b b
* * *
86
0.00
0.05
0.10
0.15
0.20
10(10)20%
15(10)30%
20(10)40%
25(10)50%
30(4)
24%
50(4)
40%
75(4)
60%
100(4)
80%
Con
[C] i
n pp
m/ K
J
Figure 3.14: The corrosion rates (in ppm per KJ) of graphite electrodes with respect to elemental carbon at different frequencies and pulse widths. The presence of asterisk (*) indicates an undetectable corrosion rate. Number shown in parenthesis is the corresponding frequency (kHz) for each pulse width (µs). The corresponding duty cycle is indicated as a percentage. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, and b denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates.
*a a
abab b
ab ab
ab
ab
87
0.0
50.0
100.0
150.0
0 5 10 15 20 25 ConDelay time/ micro-seconds
[Fe]
in p
pb/ K
J
Figure 3.15: The corrosion rates (in ppb per KJ) with respect to Fe for different delay times. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, b, and c denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates.
a
cb b
c c c
88
0.0
15.0
30.0
0 5 10 15 20 25 Con
Delay time/ micro-seconds
[Cr]
in p
pb/ K
J
Figure 3.16: The corrosion rates (in ppb per KJ) with respect to Cr for different delay times. ‘Con’ represents the corrosion rate for conventional ohmic heating. Different letters: a, b, c, and d denote significant differences (p ≤ 0.05) of mean (n = 3) corrosion rates.
c
a
b dc c c
89
CHAPTER 4
ELECTROCHEMICAL REACTIONS DURING 60 Hz OHMIC HEATING OF ASCORBIC ACID IN BUFFER MEDIUM WITH STAINLESS STEEL
ELECTRODES
ABSTRACT
This study was aimed at understanding electrochemical and secondary chemical
reactions during ohmic heating of ascorbic acid in a 3.5 pH buffer medium. Ohmic
heating experiments were performed at different power densities and NaCl concentrations
using 60 Hz sinusoidal alternating current and stainless steel electrodes. A number of
reactions seem to occur in the cell during ohmic heating. Electrode corrosion is shown to
have marked effects on the buffer medium as well as ascorbic acid degradation. The uses
of more inert electrode materials, and/or high frequency alternating currents are
suggested to minimize electrocatalytic effects on ascorbic acid.
90
INTRODUCTION
Ascorbic acid (AA), also known as vitamin C, has been the subject of numerous
investigations in many scientific disciplines, including food science, medicine, and
biochemistry. In recent years, AA has gained a renewed interest as a nutraceutical since it
possesses antioxidant properties providing potential health benefits. AA is considered to
be one of the most heat sensitive nutrients in foods, and its degradation has been reported
to vary with pH, oxygen, enzymes, metal catalysts, initial concentration, and light
(Assiry, 1996). This inherent instability of AA is a major concern in thermal food
processing. Although a number of studies have examined AA degradation under
conventional heat treatments, a little information is available related to ohmic heating.
Assiry (1996) studied degradation kinetics of AA under ohmic heating conditions
with stainless steel electrodes, and compared it with conventional heating. The results
indicate that, at pH 3.5, although kinetics of AA degradation can be described adequately
by a first order model for both conventional and ohmic heating, a number of
electrochemical as well as secondary chemical reactions appear to have some effects on
the kinetic parameters. Edirisinghe et al. (1997) found a greater loss of AA in ohmic
heating with carbon electrodes, compared to that in conventional heating. Electric field
interactions and electrode effects were suggested as being possible explanations for this
enhanced AA loss. In contrast, Lima et al. (1999) found that the electric field had no
significant effects on AA degradation.
Assiry et al. (2003) noted the influence of reactions at electrode/solution interfaces
on degradation of AA in buffer medium during 60 Hz ohmic heating with stainless steel
electrodes. Their study however does not include a detailed delineation of possible
91
reactions, and suggests a follow-up study to characterize and quantify specific reactions.
Therefore, we attempted to fill this gap by studying electrochemical and secondary
chemical reactions revisiting Assiry et al’s (2003) ohmic heating conditions. Our present
study is a comprehensive approach to understanding the reaction kinetics of AA in the
buffer medium during 60 Hz ohmic heating with stainless steel electrodes.
As can be seen, AA degradation heavily depends upon voltage (Vrms), current (Irms), and
NaCl concentration. In the present study, the dependence of these three variables (Vrms
and Irms were combined as power density) on the electrochemical processes, particularly
electrode corrosion, hydrogen generation, and pH changes, was observed throughout.
Therefore, AA degradation during ohmic heating is logically associated with the
electrochemical processes.
The metal ions migrated into the heating media catalyze oxidative degradation of AA
(Assiry, 1996), particularly via the AA-induced Fenton’s reaction. The metal migration
also causes pH changes, which in turn affect AA degradation (Assiry, 1996). The
electrolytic generation of oxygen (equation 4.3) further promotes the oxidative
degradation of AA. It is known that the primary oxidation product of AA, DHAA which
104
is as biologically active as AA, can be irreversibly hydrolyzed and degraded to over 50
different compounds ( Deutsch, 1998; Niemela, 1987).
In the present study, we identified some reactions responsible for the observations
reported by Assiry et al. (2003). These reactions affect the citrate-phosphate buffer
system as well as AA degradation. With such reactions, the differences in degradation
rate between conventional and ohmic heating observed by Assiry et al. (2003) are hardly
surprising. We would like to note that the system we studied is a model only, and is not a
representation of a real ohmic heating system designed for food processing, which may
consist of different electrode materials, such as platinized-titanium. Moreover, real foods
have not shown a greater susceptibility to AA degradation as in this buffer system (Lima
et al., 1999). Still, our study explains the chemistry underlying the kinetics of AA
degradation, and may helpful to understand the electrocatalytic effects on real food
systems. We further note that the unsuitability of citrate-phosphate buffers and stainless
steel electrodes for this type of studies. The uses of more inert electrode materials, such
as platinized-titanium (Tzedakis et al., 1999), and/or high frequency alternating currents
(Amatore et al., 1998) would be viable options.
105
CONCLUSIONS
Some electrochemical and secondary chemical reactions during ohmic heating of AA
in a 3.5 pH citrate-phosphate buffer medium have been successfully identified. Electrode
corrosion forms electrode deposits, also causing metal ion migration into the heating
medium. The electrode deposits are believed to consist mainly of CrPO4 (solid) and
FePO4(solid) that are formed simply by precipitation of some migrated Cr(III) and Fe(III)
ions on the same electrode surfaces as the insoluble phosphates. The migrated metal ions
are involved in changing the buffer pH, and affect AA degradation. The presence of 1:1
Fe(III)-citrate complex was identified in the heated medium at 1.5 Wcm-3 (1.0% NaCl).
GC-MS analysis indicates complete destruction of the buffer system during this particular
ohmic heating treatment. The results of this study highlight the needs of using more inert
electrode materials, and/or high frequency alternating currents to minimize
electrocatalytic effects.
106
SYMBOLS
AA ascorbic acid
DHAA dehydroascorbic acid
Irms RMS current (A)
m/z mass to charge ratio
ppm parts per million
RMS root-mean-square
Vrms RMS voltage (V)
Greek letters β buffer capacity in the alkaline direction (µmol pH−1 ml−1)
λmax spectral maxima (nm)
107
REFERENCES
Abrahamson H.B., Rezvani A.B.; and Brushmiller J.G.(1994); Photochemical and spectroscopic studies of complexes of iron (III) with citric acid and other carboxylic acids; Inorganica Chemica Acta, 226, pp 117-127. Amatore C., Berthou M., and Hebert S.(1998); Fundamental principles of electrochemical ohmic heating of solutions; Journal of Electroanalytical Chemistry; 457, pp. 191-203. Assiry A.M (1996); Effect of ohmic heating on the degradation kinetics of ascorbic acid; PhD thesis, The Ohio State University. Assiry A., Sastry S.K., and Samaranayake C. (2003); Degradation kinetics of ascorbic acid during ohmic heating with stainless steel electrodes, Journal of applied electrochemistry, 33 (2), pp. 187-196. Davis J.R. (Ed.) (1994); ASM Specialty Handbook (Stainless Steel); ASM International, The Materials Information Society, pp. 3-5. Dean J.A. (Ed) (1992); Lange’s Handbook of Chemistry, 14th Edition, McGraw-Hill Inc, pp. 8.6-8.11, 8.83-8.103, and 8.109-8.112. Deutsch J.C., Santhosh-Kumar C.R., Hassell K.L., and Kolhouse J.F. (1994); Variation in Ascorbic Acid oxidation routs in H2O2 and cupric ion solution as determined by GC/MS; Analytical Chemistry,66, pp. 345-350. Deutsch J.C. (1998); Spontaneous hydrolysis and dehydration of dehydroascorbic acid in aqueous solution; Analytical Biochemistry, 260, pp. 223-229. Edirisinghe E.M.R.K.B., Samaranayake C.P., Bamunuarchchi A., Walpola S., and De Alwis A.A.P. (1997); Nutrient retention in ohmic heating; 7 th International Congress on Engineer and Food (ICEF – 7), Brighton U.K., SA 43-46. Elving P.J., Markowitz J.M., and Rosenthal I.(1956); Preparation of buffer systems of constant ionic strength; Analytical Chemistry, 28(7), pp. 1179-1180. Hao X., Wei Y., and Zhang S.(2001); Synthesis, crystal structure and magnetic property of a binuclear iron(III) citrate complex; Transition Metal Chemistry, 26, 384-387. Harris D.C.(1999); Quantitative Chemical Analysis - fifth edition, W.H. Freeman and Company, New York, Ap: 12-14.
108
Lima M., Heskitt B.F., Burianek L.L., Nokes S.E., and Sastry S.K.(1999); Ascorbic acid degradation kinetics during conventional and ohmic heating; Journal of Food Processing Preservation, 23, pp. 421-434.
Manzurola E., Apelblat A., Markovits G., and Levy O. (1989); Mixed-metal hydroxycarboxylic acid complexes; J. Chem. Soc., Faraday Trans.1, 85(2), pp.373-379.
Matzapetakis M., Raptopoulou C.P., Tsohos A., Papaefthymiou V., Moon N., and Salifoglou A.(1998); Synthesis, spectroscopic and structural characterization of the first mononuclear, water soluble Iron-Citrate complex, (NH4)5Fe(C6H4O7)2 . 2H2O; J. Am. Chem. Soc., 120, pp. 13266-13267. Niemela K. (1987); Oxidative and non-oxidative alkali-catalyzed degradation of L-ascorbic acid; Journal of Chromatography, 399, pp. 235-243. Ohie T., Fu X., Iga M., Kimura M., and Yamaguchi S. (2000); Gas chromatography-mass spectrometry with tert.-butyldimethylsilyl derivatization: use of the simplified sample preparations and the automated data system to screen for organic acidemias; Journal of Chromatography B, 746, pp. 63-73. Still E.R., and Wikberg P. (1980); Solution studies of systems with polynuclear complex formation. 2. The nickel (II) citrate system; Inorganica Chimica Acta, 46, 153-155. Tomat R., and Rigo A. (1979); Oxidation of polymethylated benzenes promoted by •OH radicals; Journal of Applied Electrochemistry, (9), pp. 301-305. Tzedakis T., Basseguy R., and Comtat M. (1999); Voltammetric and coulometric techniques to estimate the electrochemical reaction rate during ohmic sterilization; Journal of Applied Electrochemistry; 29(7), pp. 821- 828. Zhao M.J., and Jung L.(1995); Kinetics of the competitive degradation of deoxyribose and other molecules by hydroxyl radicals produced by the Fenton reaction in the presence of ascorbic acid; Free Radical Research,23(3),pp.229-243.
109
Parameter
Value
NaCl concentration of the buffer a (w/v %)
0.25
0.50
1.0
Power density b (W/ cm3)
0.5
0.75
0.75
1.5
Corresponding voltage (Vrms) (at 60 Hz) / (± 1) V
33.0
40.3
33.9
41.1
Corresponding current (Irms) (at 60 Hz) / (± 1) A
3.0
3.7
4.4
7.3
Isothermal temperature / °C
80 ± 2
Duration of heating/ minutes
60
a McIlvaine type citrate-phosphate buffer system. 0.1313% (w/v) citric acid monohydrate, and 0.2390% (w/v) Na2HPO4 .12 H2O form the buffer having pH 3.5 ± 0.25 (at 25 °C) depending on the NaCl concentration. b Volume = 200.0 (± 0.80) cm3 Table 4.1: Ohmic heating conditions.
Approximate time to accumulate 250 ppm H2 gas * / min
19 a
12 b
4 c
2 c
% Electrical conductivity change * (at 25 °C)
+ 4.3 a
+ 6.4 a
+ 7.1 a
+ 6.1 a
% pH change * (at 25 °C)
+ 1.3 a,b
0 a
+ 4.8 b
+ 119.8 c
* Mean values (n=2) in the same row with different superscript letters are significantly different (p ≤ 0.05). Table 4.2: Some indicators of the electrochemical processes at different power densities and NaCl concentrations.
111
NaCl concentration of the buffer (w/v %)
0.25
0.50
1.0
Power density (W/ cm3)
0.5
0.75
0.75
1.5
Fe
15.19 (2.81)
12.90 (1.96)
18.99 (1.95)
23.26 (8.30)
Cr
21.63 (3.85)
22.71 (3.69)
27.59 (3.58)
30.78 (11.42)
Ni
1.96 (0.29)
1.94 (0.25)
2.61 (0.32)
4.13 (0.85)
Mo
1.70 (0.38)
1.80 (0.12)
1.54 (0.26)
2.62 (1.17)
P
7.95 (0.73)
8.49 (0.50)
7.63 (0.71)
2.36 (0.97)
Si
0.14 (0.07)
0.16 (0.03)
0.14 (0.04)
0.40 (0.10)
Cl
1.34 (0.18)
1.29 (0.40)
1.45 (0.29)
2.15 (0.98)
Na
1.76 (0.36)
2.12 (0.26)
1.70 (0.37)
1.07 (0.23)
O
45.05 (1.99)
48.58 (6.51)
38.70 (4.69)
33.36 (6.59)
Table 4.3: Chemical compositions (as weight %) of electrode deposits at different power densities and NaCl concentrations. The values are means of five replicates (n=5) with respective standard deviations in parentheses.
112
Compound
Solubility product b
(Ksp)
Minimum [Mn+] / moles dm-3
(where n = 2,3)
FePO4 (solid)
1.3 × 10 −22
4.87 × 10 −12
c Fe3(PO4)2 (solid)
1 × 10 −36
1.12 × 10 −5
d CrPO4 (solid)
2.4 × 10 −23
9 × 10 −13
Metal-Phosphate a
Ni3(PO4)2 (solid)
5.0 × 10 −31
8.88 × 10 −4
Fe(OH)3 (solid)
4 × 10 −38
1.26 × 10 −6
Fe(OH)2 (solid)
8.0 × 10 −16
7.94 × 105
Cr(OH)3 (solid)
6.3 × 10 −31
19.95
Cr(OH)2 (solid)
2 × 10 −16
2.0 × 105
Metal-Hydroxide
Ni(OH)2 (solid)
2.0 × 10 −15
2.0 × 106
a [Na2HPO4] = 0.0067 M; and HPO42− ⇔ H+ + PO4
3 −, where Ka = 1.26 × 10 −12 (Dean, 1992). Ka b Ksp at 18 – 25 °C (Dean, 1992). c Fe3(PO4)2 .8H2O (Harris, 1999). d CrPO4 .4H2O (green). Table 4.4: Minimum migratory metal ion concentration [Mn+] needed to precipitate some metal-phosphates and metal-hydroxides in the presence of the same [Na2HPO4] as in the citrate-phosphate buffer system at pH 3.5.
113
NaCl concentration of the buffer (w/v %)
Power density (W/ cm3)
Observed % pH change* in ohmic heating
% pH change* (Buffer + AA + Fe3+)
0.5
+ 1.3 a,b
- 1.0 a
0.25
0.75
0 a
- 2.1 a
0.50
0.75
+ 4.8 b
+ 1.6 b
1.0
1.5
+ 119.8 c
+ 25.4 c
* All the pH measurements were carried out at 25 °C. Mean values (n=2) in the same column with different superscript letters are significantly different (p ≤ 0.05). Table 4.5: The effect of AA-induced Fenton’s reaction on buffer pH, in comparison with the observed pH changes at different power densities and NaCl concentrations.
114
Sample
λmax / nm
Absorptivity coefficient
1:1 Fe(III)-citrate 221.00 ( cmvwE 0.1
)/%(025.0 ) 73.08 ml g-1. cm-1
The ohmically heated heating medium
219.50 ( cmvvE 0.1
)/%(5 ) 0.35 cm-1
Table 4.6: The spectral maxima (λmax) and the respective absorptivity coefficients of 1:1 Fe(III)-citrate, and the ohmically heated heating medium at 1.5 Wcm-3 (1.0% NaCl).
Buffer component
Retention time/ min
[M]+ (TBDMS)
(m/z)
[M-57]+
(m/z)
% Loss
Citric acid (CA)
18.8
648
591
99.8
Phosphate (P)
11.2
440
383
99.9
Table 4.7: GC-MS characteristics and % losses of the buffer components.
115
0.00
1.00
2.00
3.00
4.00
5.00
3.5 4.5 5.5 6.5 7.5 8.5pH
Buf
fer c
apac
ity
Buffer
Buffer
Buffer + metal ions [1]
Buffer + metal ions [2]
Figure 4.1: Buffer capacities (in µmol pH-1ml-1) at 0.25% (w/v) NaCl. [1] and [2] correspond to the amounts of metal ions migrated at 0.5 Wcm-3 and 0.75 Wcm-3, respectively.
at 25 °C
at 80 °C
at 80 °C
at 80 °C
116
0.00
1.00
2.00
3.00
4.00
5.00
3.5 4.5 5.5 6.5 7.5 8.5pH
Buf
fer c
apac
ity
Buffer
Buffer
Buffer + metal ions
Figure 4.2: Buffer capacities (in µmol pH-1ml-1) at 0.50% (w/v) NaCl.
at 25 °C
at 80 °C
at 80 °C
117
0.00
1.00
2.00
3.00
4.00
5.00
3.5 4.5 5.5 6.5 7.5 8.5pH
Buf
fer c
apac
ity
Buffer
Buffer
Buffer + metal ions
Figure 4.3: Buffer capacities (in µmol pH-1ml-1) at 1.0% (w/v) NaCl.
at 25 °C
at 80 °C
at 80 °C
118
Figure 4.4: The buffer solution before being subjected to ohmic heating (a), and the ohmically heated medium at 1.5 Wcm-3 (1.0% NaCl) (b).
119
Figure 4.5: UV-Visible absorption spectra of 1:1 Fe(III)-citrate( ), and the ohmically heated medium ( ) at 1.5 Wcm-3 (1.0% NaCl).
A
120
Figure 4.6: Total ion chromatograms of TBDMS derivatized solutions of the unheated buffer (A), and the ohmically heated medium at 1.5 Wcm-3 (1.0% NaCl) (B). The CA and P peaks represent citric acid and phosphate, respectively.
6 8 10 12 14 16 18 20 22 24 26 28
Time (min)
0
25
50 75
100
Rel
ativ
e A
bund
ance
CA
P
6 8 10 12 14 16 18 20 22 24 26 28
Time (min)
0
25 50 75
100
Rel
ativ
e A
bund
ance
CAP
A
B
121
CHAPTER 5
INVESTIGATION OF FREE RADICAL GENERATION DURING OHMIC HEATING
ABSTRACT
Free radical generation in food processing is generally a concern. Research in this
area would aid to avoid or inhibit radical generating events during food processing. There
have been no studies reported about free radical generation during ohmic heating.
However, electrochemical phenomena at the electrode/solution interfaces may be
involved in generating radicals. This study investigates free radical generation during
ohmic heating at different frequencies. Ohmic heating experiments were carried out with
platinized-titanium electrodes using an aqueous heating medium containing 5,5-dimethyl-
1-pyrroline N-oxide (DMPO) as a spin trapping agent. Electron spin resonance
spectroscopy (ESR) was used for detection of radicals. No radical generation was evident
with pulsed ohmic heating operated at 1, 4, and 8 kHz under these experimental
conditions suggesting an operational frequency range effective in suppressing free radical
generation. However, results indicate generation of •OH radicals with conventional low-
frequency (60 Hz, sine wave) ohmic heating, and also with pulsed ohmic heating
operated at 10 kHz.
122
INTRODUCTION
Free radicals are by definition chemical species containing unpaired electrons. In
biological systems, free radicals are thought to play a major role in many oxidative
processes within cells, and have been implicated in a number of human diseases as well
as aging (Reiter et al., and Wickens, 2001). Generally, free radicals can be generated in
both chemical and biological systems by multiple pathways. In electrochemistry,
heterogeneous electron transfer associated with electrochemical reactions at the
electrode/solution interfaces is known to generate radical species (Schafer, 2001; Sawyer,
2003). This electrochemical generation of radicals has practical importance in the areas
of organic electrosynthesis (Schafer, 2001) and water treatment (Malik et al., 2001; Sun
et al., 1997). Some electrotechnologies used in food processing, such as high voltage arc
discharge (U.S.FDA, 2000) and pulsed electric discharge (Anpilov et al., 2002), are
considered to generate radicals and reactive molecular species, which in turn affect
microbial inactivation. Since there are redox type reactions at the electrode/solution
interfaces, the possibility of generating radicals during ohmic heating cannot be ruled out.
Tzedakis et al.(1999) have already briefly discussed this possibility implying the
formation of hydroxyl (•OH) and superoxide anion (O2•−) radicals during ohmic heating.
However, no conclusive evidence of radical generation during ohmic heating has so far
been reported.
As generally known, free radicals are short-lived and highly reactive. They can
readily react with various food components including lipids, vitamins, and amino
acids/proteins causing damaging effects on these nutrients, as in the cases of biological
systems (Reiter et al., Wickens, and Hawkins et al., 2001). In particular, oxygen-
123
containing free radicals cause oxidation of these food components, and also consume
antioxidants present in food formulations. Therefore, in addition to the nutritional losses,
the oxidation of foods leads to produce undesirable flavor, toxic, and color compounds,
which make foods less acceptable or unacceptable to consumers (Min et al., 2002).
The objective of this study was to investigate free radical generation during ohmic
heating. More specifically, since electrochemical reactions diminish with increasing
frequency (Wu et al., 1998; Uemura et al., 1994), we studied free radical generation at a
range of frequencies. A comparison was also made with conventional heating. Electron
Spin Resonance (ESR) spectroscopy with a spin trapping technique was employed for the
detection of radicals. The results will provide basic understanding of free radical
Pritsos et al., 1985; Sargent et al., 1976). Therefore, we used this analytical method to
acquire precise information of our experimental conditions. ESR spectra of all the
126
samples were collected on a Bruker ESP 300 spectrometer equipped with an ER 035M
NMR gaussmeter and Hewlett-Packard 5352B microwave frequency counter, using a
standard flat cell. The spectrometer was operated at 9.77 GHz microwave frequency, 10
mW microwave power, 100 kHz field modulation, and 5 G modulation amplitude.
Quantification of approximate amounts of free radicals was obtained by
measurement of the ESR spectrum of a relatively stable flavosemiquinone radical
prepared by flavoquinone/ flavodoxin radical-generating system. The spectrum was
double integrated and compared with the double integrals of the ESR spectra of the
samples that exhibited characteristic ESR signals.
DMPO-OH reference:
The generation of DMPO-OH radical adducts either via trapping of •OH radicals, or
as the decay product of DMPO-OOH (unstable) radical adducts formed by trapping of
O2•− radicals is well-known (Shi et al., 1998; Pritsos et al., 1985; Finkelstein et al., 1980).
We obtained DMPO-OH radical adducts by the latter method using xanthine/ xanthine
oxidase as a source of O2•− (Shi et al., 1998) and treating with DMPO. The ESR spectrum
of this particular radical adduct was recorded at the same spectrometer settings as
described above, and was used as the reference. Figure 5.5 shows the characteristic
quartet (1:2:2:1) spectrum of the DMPO-OH radical adducts with hyperfine splittings of
aN = aH = 14.9 G.
127
RESULTS AND DISSCUSSION
As can be seen from figure 5.6, pulsed ohmic heating operated at 1, 4, and 8 kHz did
not indicate any radical generation. However, a DMPO-OH signal appeared, when the
frequency was raised to 10 kHz. The ohmic heating carried out at 60 Hz (sine wave) also
exhibited this characteristic signal. As previously described, trapping of both •OH and
O2•− radicals by DMPO yields DMPO-OH radical adducts. In the presence of ethyl
alcohol, an efficient •OH scavenger, trapping of •OH radicals (i.e. generation of •OH
radicals during ohmic heating) can be verified, because of the formation of DMPO-
CH(OH)CH3 radical adducts (see figure 5.7), simultaneously inhibiting the characteristic
DMPO-OH signal (Pritsos et al., 1985; Finkelstein et al., 1980). Figure 5.8 shows the
effect of ethyl alcohol on the ESR spectra of the ohmic heating experiments carried out at
60 Hz (sine wave) and 10 kHz. It is clearly seen that ethyl alcohol caused almost
complete disappearance of the characteristic DMPO-OH signal at both frequencies,
strongly implying the generation of •OH radicals. The carbon-centered DMPO-
CH(OH)CH3 radical adduct formed in the presence of ethyl alcohol, is known to exhibit
an ESR signal composed of six identical lines. However, such an ESR signal was not
evident, possibly because of the very weak signal intensities. The ESR spectra shown in
figures 5.6 and 5.8 were reproducible at least twice in each case.
The conventional heating procedure did not indicate any free radical generation (see
figure 5.6). The ESR spectrum of the method blank ensured that there were no interfering
signals. Further, the ESR spectra of both the conventional heating and the method blank,
together with the ohmic heating in the presence of ethyl alcohol (which is a competitive
inhibitor) clearly demonstrated that the observed DMPO-OH signals at 60 Hz (sine wave)
128
and 10 kHz were not simply due to artifacts caused by heat or nucleophilic addition of
water to DMPO described by Robert et al., 2002 and Makino et al., 1990. In our previous
laboratory scale studies (chapter 3), we observed enhanced corrosion of platinized-
titanium electrodes during ohmic heating at these frequencies. The migrated Pt and Ti at
the above frequencies could function in a Fenton-like reaction generating •OH radicals.
The •OH radical is highly electrophilic (Hawkins et al. and Reiter et al, 2001), and
therefore can aggressively attack electron-rich molecules (i.e. virtually all food
components) causing their oxidation. The best protection against the •OH is considered to
be the prevention of its formation (Reiter et al, 2001). Therefore, ohmic heating would be
better performed at the frequencies (1 – 8 kHz) where no radical generation was detected.
In this study, although radical generation was detected at 60 Hz (sine wave) and 10 kHz
(pulses), our results may not imply the occurrence of radical generation in pilot scale. Our
pilot scale study of electrode corrosion (chapter 2) indicated extremely low Pt and Ti
migrations (ppt levels at 39.8 kW), which may not allow occurring a Fenton-like reaction
that generate •OH radicals. Moreover, since food systems are inherently complex and
consist of natural antioxidants, such as tocopherols, and other phenolics and
polyphenolics, some amount of electrochemically generated radicals can be tolerated
without undergoing significant changes during long storage periods.
129
CONCLUSIONS
Free radical generation during ohmic heating can be suppressed by using IGBT pulse
inputs. The operational frequency, however, needs to be 1 ≤ f < 10 kHz with platinized-
titanium electrodes. Ohmic heating operated at 60 Hz (sine wave) and 10 kHz (IGBT
pulses) indicated the generation of •OH radicals.
130
SYMBOLS
aN, aH hyperfine splitting constants (G)
f frequency (Hz)
ppt parts per trillion
RMS root-mean-square
T period (µs)
tp pulse width (µs)
td delay time (µs)
131
REFERENCES
Anpilov A.M., Barkhudarov E.M., Christofi N., Kop’ev V.A., Kossyi I.A., Taktakishvili M.I., and Zadiraka Y. (2002), Pulsed high voltage electric discharge disinfection of microbially contaminated liquids; Letters in applied microbiology, 35, pp. 90-94. Finkelstein E., Rosen G.M., and Rauckman E.J. (1980); Spin trapping of superoxide and hydroxyl radical: practical aspects; Archives of biochemistry and biophysics, 200(1), pp. 1-16. Hanaoka K. (2001); Antioxidant effects of reduced water produced by electrolysis of sodium chloride solutions; Journal of applied electrochemistry, 31, pp. 1307-1313. Hawkins C.L., and Davies M.J. (2001); Generation and propagation of radical reactions on proteins; Biochimica et biophysica acta, 1504, pp. 196-219. Makino K., Hagiwara T., Hagi A., Nishi M., and Murakami A. (1990); Cautionary note for DMPO spin trapping in the presence of iron ion; Biochemical and biophysical research communications, 172 (3), pp. 1073-1080. Malik M.A., Ghaffar A., and Malik S.A.(2001); Water purification by electrical discharges; Plasma sources science and technology, 10, pp.82-91. Min D.B., and Boff J.M. (2002); Chemistry and reaction of singlet oxygen in foods; Comprehensive Reviews in Food Science and Food Safety – Institute of Food Technologists, 1, pp. 58-72. Pritsos C.A., Constantinides P.P., Tritton T.R., Heimbrook D.C., and Sartorelli A.C. (1985); Use of high-performance liquid chromatography to detect hydroxyl and superoxide radicals generated from mitomycin C; Analytical Biochemistry, 150, pp. 294-299. Reiter R.J., Tan D., Manchester L.C., and Qi W. (2001); Biochemical reactivity of melatonin with reactive oxygen and nitrogen species; Cell biochemistry and biophysics, 34, pp. 237-256. Robert R., Barbati S., Ricq N., and Ambrosio M. (2002); Intermediates in wet oxidation of cellulose: identification of hydroxyl radical and characterization of hydrogen peroxide; Water research, 36, pp. 4821-4829. Sargent F.P., and Gardy E.M.(1976); Spin trapping of radicals formed during radiolysis of aqueous solutions. Direct electron spin resonance observations; Canadian journal of chemistry, 54, pp. 275-279.
132
Schafer H.J.(2001); Organic electrochemistry (Fourth edition), Lund H. and Hammerich O. (Ed.), Chapter 4: Comparison between electrochemical reactions and chemical oxidations and reductions; Marcel Dekker, Inc., New York, pp. 207-221 Shi X., Leonard S.S., Liu K.J., Zang L., Gannett P.M., Rojanasakul Y., Castranova V., and Vallyathan V. (1998); Cr(III)-mediated hydroxyl radical generation via Haber-Weiss cycle; Journal of inorganic biochemistry, 69, pp. 263-268. Sun B., Sato M., and Clements J.S.(1997); Optical study of active species produced by a pulsed streamer corona discharge in water; Journal of electrostatics, 39, pp. 189-202. Tzedakis T., Basseguy R., and Comtat M.(1999); Voltammetric and coulometric techniques to estimate the electrochemical reaction rate during ohmic sterilization; J. of Applied Electrochemistry; 29(7), pp. 821- 828. Uemura K., Noguchi A., Park S.J., and Kim D.U. (1994); ohmic heating of food materials- Effect of frequency on the heating rate of fish protein; Developments in Food Engineering - Proceedings of the 6 th International Congress on Engineering and Food; Blackie Academic & Professional Press, London, pp. 310-312. U.S. FDA- Center for food safety and applied nutrition (June 2, 2000); Kinetics of microbial inactivation for alternative food processing technologies – High voltage arc discharge. Wickens A.P.(2001); Ageing and the free radical theory; Respiration physiology, 128, pp. 379-391. Wu H., Kolbe E., Flugstad B., Park J.W., and Yongsawatdigul J.(1998); Electrical properties of fish mince during multi-frequency ohmic heating; Journal of Food Science; 63(6), pp. 1028-1032.
133
Frequency / Hz
RMS voltage (V) or pulse width (µs)
60 *(sine wave)
110 V (RMS)
1 000
226
4 000
57
8 000
27
10 000 *
22
* Ohmic heating experiments were also carried out with the heating medium containing 2% (v/v) ethyl alcohol. Lowering of electrical conductivity due to the presence of ethyl alcohol was compensated by using 115 V (RMS) and 25 µs pulse width. Table 5.1: Selected ohmic heating conditions to study free radical generation. See figure 5.3 for typical time-temperature history for all these ohmic heating conditions.
6 - Removable lid: thermocouple and airflow tubing are attached. It was also clamped to the cell body with bolts during ohmic heating 7 – Inlet airflow P – Air pressure gauge
Figure 5.1: Schematic diagram of the pressurized ohmic heater.
P
1
22
3
4
5
76
135
Figure 5.2: Schematic diagram of the experimental setup used for the ohmic heating.
Isolation module
Pressurized ohmic heater
~
Oscilloscope
Data logger
Microcomputer
V
A Digital multimeter
V = Differential voltage probe A = Current monitor
~ 60 Hz Public utility supply/ IGBT Power supply
136
0
20
40
60
80
100
120
0 10 20 30 40time/ seconds
Tem
p./ 0
C
Ohmic
Conventional
Figure 5.3: Typical time-temperature histories for ohmic and conventional heating.
137
Figure 5.4: Schematic diagram of the centering of bipolar pulses within the period (T) at each frequency (f). The positive and negative pulses having the same pulse width (tp) were equally spaced by adjusting the delay time (td) as T = 2 (tp + td).
td td/2 td/2
T = 1/ f
tp
tp
138
-80000
-40000
0
40000
80000
3405 3435 3465 3495 3525 3555Gauss
Inte
nsity
Figure 5.5: The ESR spectrum of the DMPO-OH reference. This signal represents spin concentration of 0.63 µM.
139
3405 3435 3465 3495 3525 3555
Gauss
Figure 5.6: Typical ESR spectra of ohmic and conventional heating experiments, in comparison with the ESR spectrum of DMPO-OH reference. The signals at 60 Hz (sine wave) and 10 kHz correspond to average spin concentrations of 0.14 and 0.11 µM, respectively.
DMPO-OH
60 Hz (sine wave)
10 kHz
8 kHz
4 kHz
1 kHz
Conventional
Blank
140
Figure 5.7: Chemistry of •OH and O2
•− trapping by DMPO in the presence and absence of ethyl alcohol.
N
O
+
−
N
O
OOH•
OO− (H+)
N
O
O H•
N
O
CH CH3
OH
•
•OH
CH3CH2OH
O2•−
•CH(OH)CH3 (α-hydroxyethyl radical)
•OH
DMPO
DMPO-OOH
DMPO-OH DMPO-CH(OH)CH3
141
3405 3435 3465 3495 3525 3555
Gauss
Figure 5.8: Comparison of typical ESR spectra of the ohmic heating experiments carried out at 60 Hz (sine wave) and 10 kHz in the presence (2%, v/v) and absence of ethyl alcohol.
60 Hz (sine wave)
10 kHz
With alcohol
Without alcohol
Without alcohol
With alcohol
142
CHAPTER 6
CONCLUSIONS
1. With 60 Hz (sine wave) ohmic heating, corrosion of all the electrode materials
(titanium, stainless steel, platinized-titanium, and graphite) is enhanced at pH 3.5
compared to that at pH 5.0 and 6.5.
2. Stainless steel was found to be the most electrochemically active electrode material
during ohmic heating.
3. Corrosion of graphite electrodes yields soluble organic compounds due to the
migration of surface functional groups and oxides during ohmic heating.
4. Among the materials tested in our study, platinized-titanium can be considered as
the electrode material-of-choice for ohmic heating with commonly available
low-frequency (60 Hz, sine wave) alternating currents.
143
5. Pulsed ohmic heating significantly inhibits the electrochemical reactions of
stainless steel, titanium, and platinized-titanium electrodes, in comparison to
conventional 60 Hz ohmic heating.
6. Pulsed ohmic heating at higher frequencies and shorter pulse widths yields the
lowest rates of electrochemical reactions of stainless steel electrodes.
7. Pulsed ohmic heating at lower frequencies and longer pulse widths is more effective
in suppressing the electrochemical reactions of titanium and platinized-titanium
electrodes, while achieving higher duty cycles.
8. In general, pulsed ohmic heating is unable to suppress the electrochemical
reactions of graphite electrodes.
9. Delay time (off-time between adjacent pulses) was found to be a critical factor in
pulsed ohmic heating.
10. Ohmic heating (60 Hz, sine wave) of ascorbic acid in citrate-phosphate buffer with
stainless steel electrodes showed the formations of metal-phosphates and
metal-citrate complexes, also indicating electrocatalytic effects on ascorbic acid
degradation.
144
11. Free radical generation during ohmic heating with platinized-titanium electrodes
can be effectively suppressed by using IGBT pulse inputs at the frequency range of
1 – 8 kHz.
12. Ohmic heating with platinized-titanium electrodes operated at 60 Hz (sine wave)
and 10 kHz (pulses) indicated the generation of •OH radicals.
145
LIST OF REFERENCES Aboul-Enein H.Y., Kladna A., Kruk I., and Lichszteld K. (2003); Effect of psoralens on Fenton-like reaction generating reactive oxygen species; Biopolymers, 72(1), pp. 59-68. Abrahamson H.B., Rezvani A.B.; and Brushmiller J.G.(1994); Photochemical and spectroscopic studies of complexes of iron (III) with citric acid and other carboxylic acids; Inorganica Chemica Acta, 226, pp 117-127. Amatore C., Berthou M., and Hebert S.(1998); Fundamental principles of electrochemical ohmic heating of solutions; Journal of Electroanalytical Chemistry; 457, pp. 191-203. Anpilov A.M., Barkhudarov E.M., Christofi N., Kop’ev V.A., Kossyi I.A., Taktakishvili M.I., and Zadiraka Y. (2002), Pulsed high voltage electric discharge disinfection of microbially contaminated liquids; Letters in applied microbiology, 35, pp. 90-94. Assiry A.M (1996); Effect of ohmic heating on the degradation kinetics of ascorbic acid; Ph.D thesis, The Ohio State University. Assiry A., Sastry S.K., and Samaranayake C. (2003); Degradation kinetics of ascorbic acid during ohmic heating with stainless steel electrodes, Journal of applied electrochemistry, 33 (2), pp. 187-196. Baliga B. J.; Power semiconductor devices for variable frequency devices; IEEE Technology update series: Power electronics technology and applications II (selected conference papers), Lee F.C. ed. (1998), pp. 50-60. Bently R., and Prentice T.R. (1957); The alternating current electrolysis of concentrated acids; J. Appl. Chem.; 7 (November), pp. 619-626. Bosch R.W., and Bogaerts W.F.(1998); A theoretical study of AC-induced corrosion considering diffusion phenomena; Corrosion Science, 40(2/3), pp.323-336. Butz P., and Tauscher B. (2002); Emerging technologies: chemical aspects; Food Research International, 35, pp.279-284.
146
Davis J.R. (Ed.) (1994); ASM Specialty Handbook (Stainless Steel); ASM International, The Materials Information Society, pp. 3-5. Dean J.A. (Ed) (1992); Lange’s Handbook of Chemistry, 14th Edition, McGraw-Hill Inc, pp. 8.6-8.11, 8.83-8.103, and 8.109-8.112. Deutsch J.C., Santhosh-Kumar C.R., Hassell K.L., and Kolhouse J.F. (1994); Variation in Ascorbic Acid oxidation routs in H2O2 and cupric ion solution as determined by GC/MS; Analytical Chemistry,66, pp. 345-350. Deutsch J.C. (1998); Spontaneous hydrolysis and dehydration of dehydroascorbic acid in aqueous solution; Analytical Biochemistry, 260, pp. 223-229. Dodge C.J., and Francis A.J. (1997); Biotransformation of binary and ternary citric acid complexes of iron and uranium; Environmental Science & Technology, 31(11), pp. 3062-3067. Donnet J.B. (1968); The chemical reactivity of carbons; Carbon, 6, pp. 161-176. Edirisinghe E.M.R.K.B., Samaranayake C.P., Bamunuarchchi A., Walpola S., and De Alwis A.A.P. (1997); Nutrient retention in ohmic heating; 7 th International Congress on Engineer and Food (ICEF – 7), Brighton U.K., SA 43-46. Elving P.J., Markowitz J.M., and Rosenthal I.(1956); Preparation of buffer systems of constant ionic strength; Analytical Chemistry, 28(7), pp. 1179-1180. Ershler B. (1947); Investigation of electrode reactions by the method of charging-curves and with the aid of alternating currents; Discussions of the Faraday Society, 1, pp. 269-277. Evans C.A.(1979); Spin trapping; Aldrichimica Acta, 12(2), pp. 23-30. Finkelstein E., Rosen G.M., Rauckman E.J., and Paxton J. (1979); Spin trapping of superoxide; Molecular Pharmacology, 16, pp. 676-685. Finkelstein E., Rosen G.M., and Rauckman E.J. (1980); Spin trapping of superoxide and hydroxyl radical: practical aspects; Archives of biochemistry and biophysics, 200(1), pp. 1-16. Floyed T.L. (1995); Instructor’s edition for electric circuits fundamentals and electronics fundamentals: circuits, devices, and applications (Third edition), Chapters 8 and 15; Prentice Hall, Inc., Columbus, Ohio, pp. 317- 320 and 626-661. Getchell B.E.(1935); Electric pasteurization of milk; Agr. Eng., 16(10), pp.408-410.
147
Gilbert T.W., Newman L., and Klotz P. (1968); Mixed-metal hydroxycarboxylic acid complexes, The chromium (iii) inhibition of the solvent extraction of indium (iii); Analytical Chemistry, 40(14), pp. 2123-2130. Golub D., Oren Y., and Soffer A. (1987); The electrochemical double layer of carbon and graphite electrodes, Part IV: Dependence of carbon electrode dimensions and electrical capacity on pH; Journal of Electroanalytical Chemistry, 227, pp. 41-53. Grant D. (1996); Power semiconductor devices – continuous development; Microelectronics Journal, 27, pp. 161-176 Hanaoka K. (2001); Antioxidant effects of reduced water produced by electrolysis of sodium chloride solutions; Journal of applied electrochemistry, 31, pp. 1307-1313. Hanna P.M., Chamulitrat W., Mason R.P. (1992); When are metal ion-dependent hydroxyl and alkoxyl radical adducts of 5,5-dimethyl-1-pyrroline N-oxide artifacts?; Archives of biochemistry and biophysics, 296(2), pp. 640-644. Hao X., Wei Y., and Zhang S.(2001); Synthesis, crystal structure and magnetic property of a binuclear iron(III) citrate complex; Transition Metal Chemistry, 26, 384-387. Harbour J.R., Chow V., and Bolton J.R. (1974); An electron spin resonance study of the spin adducts of OH and HO2 radicals with nitrones in the ultraviolet photolysis of aqueous hydrogen peroxide solutions; Can. J. Chem., 52, pp. 3549-3553. Harris D.C.(1999); Quantitative Chemical Analysis - fifth edition, W.H. Freeman and Company, New York, Ap: 12-14. Hawkins C.L., and Davies M.J. (2001); Generation and propagation of radical reactions on proteins; Biochimica et biophysica acta, 1504, pp. 196-219. Horlick G. (1992); Plasma-source mass spectrometry for elemental analysis; Spectroscopy, 7(1), pp. 22-27. Hu C. and Liu K. (1999); Voltammetric investigation of platinum oxides. I. Effects of ageing on their formation/ reduction behavior as well as catalytic activities for methanol oxidation; Electrochemica Acta, 44, pp. 2727-2738. Indira K.S, Sampath S., and Doss K.S.G.(1968); Recent trends of platinized titanium as node material in electrochemical industries; Chemical processing & Engineering (February), pp. 35-37.
148
Iniesta J., Gonzalez-Garcia J., Fernandez J., Montiel V., and Aldaz A. (1999); On the Voltammetric behavior of platinized titanium surface with respect to the specific hydrogen and anion adsorption and charge transfer processes; Journal of materials chemistry, 9 (12), pp. 3141-3145. James W.J., and Straumanis M.E.(1976); Encyclopedia of electrochemistry of the elements, Bard A.J. (Ed.), Vol.(V), Chapter V-7 : Titanium; Marcel Dekker Inc., New York, pp. 305 – 386. Koike M. and Fujii H. (2001); The corrosion resistance of pure titanium in organic acids; Biomaterials, (22), pp. 2931 – 2936. Konya J. (1979); Notes on the “non-faraday” electrolysis of water; Journal of electrochemical society: electrochemical science and technology, 126(1), pp. 54-56. Kremer M.L. (1999); Mechanism of the Fenton reaction. Evidence for a new intermediate; Phys. Chem. Chem. Phys., 1, pp. 3595-3605. Kulman F.E. (1961); Effects of alternating currents in causing corrosion; Corrosion; 17(3), pp. 34-35. Lalvani S.B. and Lin X.A. (1994); A theoretical approach for predicting AC-induced corrosion; Corrosion Science, 36(6), pp. 1039-1046. Lalvani S.B. and Lin X.A. (1996); A revised model for predicting corrosion of materials induced by alternating voltages; Corrosion Science, 38(10), pp. 1709-1719. Lima M., Heskitt B.F., Burianek L.L., Nokes S.E., and Sastry S.K.(1999); Ascorbic acid degradation kinetics during conventional and ohmic heating; Journal of Food Processing Preservation, 23, pp. 421-434. Lindeburg M.R. (1998); Engineer-in-training reference manual, Chapter 48: Alternating current circuits; 8th edition, Professional publications, Inc., Belmont, CA, pp. 48-4 – 48-5. Makino K., Hagiwara T., Hagi A., Nishi M., and Murakami A. (1990); Cautionary note for DMPO spin trapping in the presence of iron ion; Biochemical and biophysical research communications, 172 (3), pp. 1073-1080. Malik M.A., Ghaffar A., and Malik S.A.(2001); Water purification by electrical discharges; Plasma sources science and technology, 10, pp.82-91. Manzurola E., Apelblat A., Markovits G., and Levy O. (1989); Mixed-metal hydroxycarboxylic acid complexes; J. Chem. Soc., Faraday Trans.1, 85(2), pp.373-379.
149
Matzapetakis M., Raptopoulou C.P., Tsohos A., Papaefthymiou V., Moon N., and Salifoglou A.(1998); Synthesis, spectroscopic and structural characterization of the first mononuclear, water soluble Iron-Citrate complex, (NH4)5Fe(C6H4O7)2 . 2H2O; J. Am. Chem. Soc., 120, pp. 13266-13267. McCreery R.L.(1999); Interfacial Electrochemistry: Theory, Experiment, and Applications, Wieckowski A. (Ed.), Chapter 35 : Electrochemical properties of carbon surfaces; Marcel Dekker Inc., New York, pp. 631 – 646. Min D.B., and Boff J.M. (2002); Chemistry and reaction of singlet oxygen in foods; Comprehensive Reviews in Food Science and Food Safety – Institute of Food Technologists, 1, pp. 58-72. Mizrahi S., Kopelman I.J., and Perlman. J.(1975); Blanching by electroconductive heating; J. Food Technology, 10, pp. 281-288. Moses B.D. (1938); Electric pasteurization of milk; Agr. Eng., 19(12), pp.525-526. Naveh D., Kopelman I.J., and Mizrahi S. (1983); Electroconductive thawing by liquid contact; J. Food Technology, 18, pp. 171-176. Niemela K. (1987); Oxidative and non-oxidative alkali-catalyzed degradation of L-ascorbic acid; Journal of Chromatography, 399, pp. 235-243. Official methods of analysis of AOAC International – 17th edition (2000); Vol.1 (Agricultural Chemicals, Contaminants, Drugs), Ch.9: metals and other elements at trace levels in foods, pp. 46-59. Ohie T., Fu X., Iga M., Kimura M., and Yamaguchi S. (2000); Gas chromatography-mass spectrometry with tert.-butyldimethylsilyl derivatization: use of the simplified sample preparations and the automated data system to screen for organic acidemias; Journal of Chromatography B, 746, pp. 63-73. Pai S., Tsau Y., and Yang T. (2001); pH and buffering capacity problems involved in the determination of ammonia in saline water using the indophenol blue Spectrophotometric method; Analytica Chimica Acta, 434, pp. 209-216. Palaniappan S., and Sastry S. (1991); Electrical conductivity of selected juices: Influences of temperature, solid content, applied voltage, and practical size; Journal of Food Process Engineering, 14, pp. 247-260. Parrott D.L.(1992); Use of ohmic heating for aseptic processing of food particulates; J. Food Technology, 12, pp. 68-72.
150
Pritsos C.A., Constantinides P.P., Tritton T.R., Heimbrook D.C., and Sartorelli A.C. (1985); Use of high-performance liquid chromatography to detect hydroxyl and superoxide radicals generated from mitomycin C; Analytical Biochemistry, 150, pp. 294-299. Rabah M.A., Nassif N., and Azim A.A.A. (1991); Electrochemical wear of graphite anodes during electrolysis of brine; Carbon, 29(2), pp. 165-171. Ramshaw R.S. (1993); Power electronics semiconductor switches, Chapters 1 and 7; Chapman & Hall, Inc., New York, pp. 1-23 and 294-319. Redmond J.D. (1996); Marks’ Standard handbook for mechanical engineers (Tenth Edition); McGraw – Hill Companies, Inc., pp. 6(32) – 6(33). Reilly C.; Metal contamination of food, second edition (1991); Elsevier Science Publishers Ltd., New York; pp. 14-15 & 235-237. Reiter R.J., Tan D., Manchester L.C., and Qi W. (2001); Biochemical reactivity of melatonin with reactive oxygen and nitrogen species; Cell biochemistry and biophysics, 34, pp. 237-256. Reznick D.(1996); ohmic heating of fluid foods; J. Food Technology; 5, pp. 250-251. Robert R., Barbati S., Ricq N., and Ambrosio M. (2002); Intermediates in wet oxidation of cellulose: identification of hydroxyl radical and characterization of hydrogen peroxide; Water research, 36, pp. 4821-4829. Rubinstein I. (1995); Physical electrochemistry: principles, methods, and applications, Rubinstein I. (Ed.), Chapter 1: Fundamentals of physical electrochemistry; Marcel Dekker, Inc., New York, pp. 1-4. Sargent F.P., and Gardy E.M.(1976); Spin trapping of radicals formed during radiolysis of aqueous solutions. Direct electron spin resonance observations; Canadian journal of chemistry, 54, pp. 275-279. Sawyer D.T. (2003); Electrochemical transformations of metals, metal compounds, and metal complexes: invariably (ligand/ solvent)-centered; Journal of molecular catalysis A: Chemical, 194, pp. 53-67. Schafer H.J.(2001); Organic electrochemistry (Fourth edition), Lund H. and Hammerich O. (Ed.), Chapter 4: Comparison between electrochemical reactions and chemical oxidations and reductions; Marcel Dekker, Inc., New York, pp. 207-221
151
Scott V.D., Love G., and Reed S.J.B. (1995); Quantitative electron-probe microanalysis (second edition), Chapters 4 and 6; Ellis Horwood Ltd., New York, pp. 61-75 and 93-109. Shaw M. and Remick A.E.(1950); Studies on alternating current electrolysis; J. of Electrochemical Society; 97(10), pp. 324-334. Shi X., Leonard S.S., Liu K.J., Zang L., Gannett P.M., Rojanasakul Y., Castranova V., and Vallyathan V. (1998); Cr(III)-mediated hydroxyl radical generation via Haber-Weiss cycle; Journal of inorganic biochemistry, 69, pp. 263-268. Shriver D.F. and Atkins P.W. (1999); Inorganic Chemistry (third edition), Chapter 6: Oxidation and reduction; W.H. Freeman and Company, New York, pp. 177-210. Soffer A., and Folman M. (1972); The electrical double layer of high surface porous carbon electrodes; Journal of Electroanalytical Chemistry, 38, pp.25 – 43. Speiser C., Baumann T., and Niessner R. (2001); Characterization of municipal solid waste incineration (MSWI) bottom ash by scanning electron microscopy and quantitative energy dispersive X-ray microanalysis (SEM/EDX); Fresenius J. Anal. Chem., 370, pp. 752-759. Still E.R., and Wikberg P. (1980); Solution studies of systems with polynuclear complex formation. 2. The nickel (II) citrate system; Inorganica Chimica Acta, 46, 153-155. Stirling R. (1987); ohmic heating - a new process for the food industry; Power Engineering Journal; 1(6), pp. 365-371. Stott F.H.(1984); Some recent experience with scanning electron microscope in corrosion research; Journal of Microscopy, 133 (2), pp. 191-204. Sun B., Sato M., and Clements J.S.(1997); Optical study of active species produced by a pulsed streamer corona discharge in water; Journal of electrostatics, 39, pp. 189-202. Tanner S.D., Baranov V.I., and Volkopf U. (2000); A dynamic reaction cell for inductively coupled plasma mass spectrometry (ICP-DRC-MS), Part III: Optimization and analytical performance; J. Anal. At. Spectrom., 15, pp. 1261-1269. Tarasevich M.R., Bogdanovskaya V.A., and Zagudaeva N.M.(1987); Redox reactions of quinones on carbon materials; Journal of Electroanalytical Chemistry, 223, pp.161-169. Thomas M.J.K. (1996); Ultraviolet and visible spectroscopy, Ando D.J. (Ed.), Chapter 5: Qualitative analysis and structural relationships; second edition, John Wiley & Sons, Ltd., England, pp. 163-187.
152
Thomson B.A., Iribarne J.V., and Dziedzic P.J. (1982); Liquid ion evaporation/ mass spectrometry/ mass spectrometry for the detection of polar and labile molecules; Analytical Chemistry, 54(13), pp.2219-2224. Tomat R., and Rigo A. (1979); Oxidation of polymethylated benzenes promoted by •OH radicals; Journal of Applied Electrochemistry, (9), pp. 301-305. Tomov T. and Tsoneva I. (2000); Are the stainless steel electrodes inert?; Bioelectrochemistry, 51, pp. 207-209. Tzedakis T., Basseguy R., and Comtat M.(1999); Voltammetric and coulometric techniques to estimate the electrochemical reaction rate during ohmic sterilization; Journal of Applied Electrochemistry; 29(7), pp. 821- 828. Uemura K., Noguchi A., Park S.J., and Kim D.U. (1994); ohmic heating of food materials- Effect of frequency on the heating rate of fish protein; Developments in Food Engineering - Proceedings of the 6 th International Congress on Engineering and Food; Blackie Academic & Professional Press, London, pp. 310-312. UK patent applications GB 2067390 A and 2068200 A (1981); Apparatus for heating electrically conductive flowable media. U.S. FDA- Center for food safety and applied nutrition (June 2, 2000); Kinetics of microbial inactivation for alternative food processing technologies – High voltage arc discharge. Van den Broeke L.T., Graslund A., Nilsson J.L.G., Wahlberg J.E., Scheynius A., and Karlberg A.T. (1998); Free radicals as potential mediators of metal-allergy: Ni2+- and Co2+ -mediated free radical generation; European Journal of Pharmaceutical Sciences, 6, pp. 279-286. Venkatachalam S. and Mehendale S.G.(1981); Electrodissolution and corrosion of metals by alternating currents; Journal of Electrochemical Society, India; 30-3, pp. 231-237. Venkatesh S. and Chin D. (1979); The alternating current electrode processes; Israel Journal of Chemistry, 18, pp.56-64. Voyksner R.D. (1992); Elecrospray LC/MS – can it be used to determine lower molecular weight molecules?; Nature, 356, pp. 86-87. Wickens A.P.(2001); Ageing and the free radical theory; Respiration physiology, 128, pp. 379-391.
153
Wu H., Kolbe E., Flugstad B., Park J.W., and Yongsawatdigul J.(1998); Electrical properties of fish mince during multi-frequency ohmic heating; Journal of Food Science; 63(6), pp. 1028-1032. Ysart G., Miller P., Crews H., Robb P., Baxter M., De L’Argy, Lofthouse S., Sargent C., and Harrison N. (1999); Dietary exposure estimates of 30 elements from the UK total diet study; Food Additives and Contaminants, Vol.16 (9), pp. 391-403. Zhao M.J., and Jung L.(1995); Kinetics of the competitive degradation of deoxyribose and other molecules by hydroxyl radicals produced by the Fenton reaction in the presence of ascorbic acid; Free Radical Research,23(3),pp.229-243. Zoltai P. and Swearingen P. (1996); Product development considerations for ohmic processing; J. Food Technology, 5, pp. 263-266.