Ultrasound, Cavitation and Cleaning

Benjamin Paul Wilson B.Sc. Hons (Wales)

A thesis submitted in fulfilment of the requirements of the degree of Master of Philosophy in the University of Wales Department of Materials Engineering University of Wales, Swansea September 1997

ii

Disclaimer

Declaration: This work has not previously been accepted in substance for any degree and is not being currently submitted in candidature for any degree.

Signed: _____________

Date:____________

Statement: This thesis is the result of my own investigations, except where otherwise stated. Other sources are acknowledged by footnotes giving explicit references.

Signed: _____________

Date:____________

Statement: I hereby give consent for my thesis, if accepted, to be available for photocopying and interlibrary loan and for the title and summary to be made available to outside organisations.

Signed: _____________

Date:____________

iii

Aim of the Thesis

This thesis describes work initially aimed at elucidating the interaction of power ultrasound and electrolytic current in the removal of oxide scale from metal (steel) surfaces. The descaling of metals is currently carried out using an acid pickling process and electrolytic descaling in neutral salt would be an environmentally friendly alternative. As such the work has three parts:

I.

A systematic study of ultrasound propagation and ultrasonically induced cavitation in aqueous solution proximal to an ultrasound transducer. This work was undertaken by using the sonogenerated chemiluminescence (SCL) of luminol to produce images of ultrasonically generated cavitation fields. The background to ultrasound and cavitation in aqueous systems is introduced in sections 1.1 and 1.4. the phenomenon of luminol SCL is introduced in section 3.4

II.

A systematic study of the influence of ultrasound intensity and electrolytic current density on the rate of oxide scale removal from a steel surface. The theory of electrochemical rate processes is introduced in section 1.5. The nature and causation of oxide heat scales is introduced briefly in section 1.3.

III.

A study of the influence of ultrasound on the rate of cathodic hydrogen evolution and cathodic oxygen reduction at a titanium sonotrode (i.e. an electrode that is also an ultrasonic transducer.) This last piece of work was undertaken in order to determine how, and to what extent, ultrasound might affect the rate of interfacial electron transfer processes occurring at a metal solution interface.

The original intention was to produce a body of knowledge, which would be of use in the design of equipment for the rapid, environmentally friendly, descaling of metal surfaces. However, in the course of the above work information came to light, which may be of more general significance. For this reason an effort was made to investigate the propagation of ultrasound in cavitation fields and the operation of a "sonotrode at as fundamental level as time allowed.

iv

Acknowledgements
I would like to thank Dr. Neil McMurray for his supervision and for being a source of inspiration, encouragement and astonishment (in varying degrees) throughout the duration of this project. Thanks must also go to Duncan MacDonald for allowing me the opportunity to participate in this research and all at Maysonic Ultrasonics Ltd. for providing technical assistance throughout the year. Personal thanks to Sue, Fiona, Jafar, Ahmed, Siva and Dave for all the help, drama, comedy, and for all those hours youve had to endure the less than pleasant hullabaloo that has been emanating from my little box of tricks in the corner of the lab. Thanks once again to Justin for putting up with lack of computer literacy your assistance is always much appreciated! Finally a big diolch yn fawr for my good friend Sarah, for all the hours of fun and entertainment that we have both experienced during the course of our respective write ups.

Ultrasound, Cavitation and Cleaning

Aim of the Thesis....................................................................................... iii

Acknowledgements ................................................................................... iv

Chapter 1 Introduction ............................................................................... 2

1.1 Ultrasound and Ultrasonic Mechanisms .................................................................................... 2 1.1.1 Background ......................................................................................................................... 2 1.1.2 History of Ultrasound.......................................................................................................... 2 1.2 Applications of Ultrasound ........................................................................................................ 5 1.2.1 High Frequency (or Diagnostic) Ultrasound ....................................................................... 5 1.2.2 Power or (Low Frequency) Ultrasound............................................................................... 5 1.2.3 Ultrasonic Welding ............................................................................................................. 5 1.2.4 Biological Uses of Ultrasound ............................................................................................ 6 1.2.5 Ultrasound in Medical Applications ................................................................................... 6 1.2.6 Engineering and Ultrasound, ............................................................................................... 7 1.2.7 Dentistry.............................................................................................................................. 8 1.2.8 Ultrasonic Cleaning............................................................................................................. 8 1.3 Heat Scale Formation and Composition..................................................................................... 9 1.3.1 Introduction......................................................................................................................... 9 1.3.2 The Classical Three Layer Scale Formation ....................................................................... 9 1.4 Ultrasonic Effects on Aqueous Media ..................................................................................... 11 1.4.1 Introduction....................................................................................................................... 11 1.4.2 Acoustic Cavitation and Streaming................................................................................... 13 1.4.3 Factors Affecting Cavitation ............................................................................................. 16 1.5 Electrochemistry ...................................................................................................................... 20 1.5.1 Electrochemical Reactions ................................................................................................ 20 1.5.2 The Electrical Double Layer Hypothesis .......................................................................... 27 1.6 Ultrasound in Electrochemistry................................................................................................ 29 1.6.1 Introduction....................................................................................................................... 29 1.6.2 Mass Transport in Electrochemistry ................................................................................. 31 1.6.3 Mass Transport Boundary Layer (Nernst Diffusion Layer) .............................................. 31

Chapter 2 Experimental Set-up and Protocol ........................................ 34

2.1 Materials .................................................................................................................................. 35 2.1.1 Chemicals.......................................................................................................................... 35 2.2 Methods ................................................................................................................................... 36 2.2.1 Ultrasound Probe............................................................................................................... 36 2.2.2 Photomultiplier Tube ........................................................................................................ 36 2.2.3 Low Light Camera ............................................................................................................ 36 2.2.4 Galvanostat ....................................................................................................................... 36 2.2.5 Function Generator ........................................................................................................... 38 2.2.6 Potentiostat........................................................................................................................ 38 2.3 Calibration of the Ultrasound Probe ........................................................................................ 38 2.3.1 Calorimetry ....................................................................................................................... 38 2.3.2 Equipment Utilised in the Calibration of the Ultrasound Probe........................................ 39 2.3.3 Method .............................................................................................................................. 39 2.3.4 Results............................................................................................................................... 39

vi

2.4 Luminol Sonochemilumunescence Experiments ..................................................................... 42 2.4.1 Experimental Set-up.......................................................................................................... 42 2.5 Wire Cleaning Experiments ..................................................................................................... 43 2.5.1 Wire Samples .................................................................................................................... 43 2.5.2 Sample Preparation ........................................................................................................... 43 2.5.3 Electrolyte Preparation...................................................................................................... 47 2.5.4 Electrolytic Current........................................................................................................... 47 2.5.5 Current Density ................................................................................................................. 47 2.6 Sonotrode Experiments ............................................................................................................ 48

Chapter 3 The Visualisation of Ultrasonically Induced Cavitation with Luminol .................................................................................................. 50

3.1.1 Introduction....................................................................................................................... 51 3.2 Materials .................................................................................................................................. 51 3.2.1 Chemicals.......................................................................................................................... 51 3.2.2 Experimental Equipment................................................................................................... 51 3.3 Experimental Details................................................................................................................ 52 3.3.1 Kinetic Measurements....................................................................................................... 52 3.3.2 Image Capture and Analysis ............................................................................................. 53 3.4 Results and Discussion............................................................................................................. 53 3.4.1 Kinetic Investigations ....................................................................................................... 53 3.4.2 Mechanism: sonochemical generation of OH and O2- .................................................... 58 3.4.3 SCL Image Analysis ......................................................................................................... 64 3.4.4 Acoustic Attenuation in Cavitating Water. ....................................................................... 73 3.4.5 Conclusions....................................................................................................................... 75

Chapter 4 Determination of the Effect of Ultrasound Intensity and Proximity on Wire Cleaning Kinetics ...................................................... 76
4.1 Introduction.............................................................................................................................. 77 4.2 Experimental Details................................................................................................................ 77 4.2.1 Samples ............................................................................................................................. 77 4.2.2 Method .............................................................................................................................. 77 4.2.3 Ultrasonic Configuration................................................................................................... 78 4.2.4 Measurement of Surface Cleaning. ................................................................................... 78 4.3 Results and Discussion............................................................................................................. 81 4.3.1 Influence of Ultrasound Power and Transducer Surface Distance.................................... 81 4.3.2 Ultrasound Shadowing. ..................................................................................................... 84

Chapter 5 Hydrogen Evolution at the Titanium Sonotrode................... 89

5.1 Introduction.............................................................................................................................. 90 5.2 Experimental Details................................................................................................................ 91 5.2.1 Methodology ..................................................................................................................... 91 5.3 Results and Discussion............................................................................................................. 92

vii

Appendices ............................................................................................... 98

Paper: Hydrogen evolution and oxygen reduction at a titanium sonotrode Chem. Commun. (1998) ........................................................ 99

References .............................................................................................. 100

Chapter 1

Introduction

1.1 Ultrasound and Ultrasonic Mechanisms

1.1.1 Background Sound comprises of a series of waves transmitted through a medium - solid, liquid or gas which possesses elastic properties allowing the periodic displacement of molecules from the mean position. Sound waves are longitudinal waves and comprise of a series of alternating compressions and rarefractions, along the axis of propagation, as shown in Figure 1.1. In effect, the movement/vibration of the sound source is communicated to a layer of molecules of the medium, which in turn transmits the motion to an adjoining layer before returning to an approximately unperturbed position. This is contrast to electromagnetic waves (e.g. light, radio waves, x-rays)1 which are transverse waves oscillating at right angles to the direction of propagation. The pitch of sound depends solely on its frequency i.e. the higher the pitch the higher the frequency. For humans, the range at which sound waves are audible is around 16 to 17,000 Hz (vibratory waves/cycles per second). Ultrasound is a term that is used to describe sound waves that possess a frequency in excess of that discernible by the human ear (>17kHz). The upper level of the ultrasound is fairly indistinct but is usually taken to be in the region of 5 MHz for gases and 500 MHz for liquids and solids.

1.1.2 History of Ultrasound The beginnings of modern day ultrasound technology can be traced back as far as the 1800s with the discovery of ultrasonic generation and detection techniques. In 1847 Joule
2

revealed the phenomenon of Magnestriction which

3,4

involves the modification of a magnetic materials dimensions whilst exposed to a magnetic field. The Curie Brothers followed this, in 1880 with their observation

of the Piezoelectric Effect and its inverse. Piezoelectricity involves the production of electrical charge on certain crystalline surfaces when they are put under tension or pressure and is a method of ultrasound detection. The inverse of this effect uses alternated potential, which, when applied to the crystal, will be converted from electrical energy to mechanical (sound) energy akin to a loudspeaker. If the alternating potential is of a high enough frequency then ultrasound is generated. Galton
5

discovered ultrasounds first practical use in 1883. His specially

designed whistle transducer had a resonance cavity which was adjustable and

allowed sound of a known frequency to be generated. The so-called Silent Dog Whistle was originally used to investigate the audible range for humans and animals. As its name suggests it is still in use today for certain applications. Commercial exploitation of ultrasound didnt occur until 1917,when Langevin developed an ultrasonic echo sounding technique for water depth estimation. The impetus for the invention came out of a competition to detect icebergs in the open sea and thus prevent another Titanic disaster. This early Echosounder method utilised a simple pulse of ultrasound from the ships keel to the sea bottom that reflected it back to a detector. From Equation 1 for seawater:

Distance Travelled = 1 time velocity of sound in water 2 Equation 1 The depth could be gauged and any foreign object between the bottom of the sea and the ship appeared as an echo in advance of that from the bottom. This system was the forerunner of what is known as SONAR (SOund, Navigation And Ranging) today.5,6 Sir John Thornycroft and Sidney Barnaby were the first to observe the phenomenon of cavitation
7

in 1895. Problems with HMS Daring (a newly built

destroyer) led to discovery that the propeller blades were incorrectly aligned causing the water structure to be torn apart by the mechanical action and the induction of cavitation bubbles. The periodic rarefractions produced by intense Power ultrasound fields in water can also produce cavitation 5. Since the 1940s, an increased understanding of ultrasound and its associated cavitation phenomena has led to significant developments in the application of power ultrasound to chemical processes Sonochemistry. Modifications to chemical reactions by ultrasound are caused by the occurrence of cavitation.

Wavelength

Axis of propagation

Compression

Rarefaction

Figure 1.1: An illustration of the periodic compressions and rarefractions present in a sound wave along the axis of propagation.

Frequencies (Hz)

101

102

103

104

105

106

107

Grasshopper (7 kHz) Bumble bee (150 Hz)

Dolphin whistle (120 kHz)

Bat navigation signals (70 kHz)

Range of Human Hearing (16 Hz 17 kHz) Low Frequency Ultrasound (20 kHz 100 kHz) High Frequency Ultrasound (1MHz 10 MHz)

Figure 1.2: A diagram of frequency range.

1.2 Applications of Ultrasound

Practically, ultrasound can be divided into two main areas, High Frequency or Diagnostic ultrasound and Low Frequency or Power ultrasound.

1.2.1 High Frequency (or Diagnostic) Ultrasound This utilises ultrasound within the 2-10 MHz range and the effects of the medium on the wave (attenuation). Typical uses of high frequency ultrasound include non-destructive materials testing, SONAR and medical scanning of foetuses. 1.2.2 Power or (Low Frequency) Ultrasound The second area is known, most commonly, as Power Ultrasound and involves the region between 20 and 100 kHz in frequency. These are high-energy waves and are used for a variety of applications that include the welding of plastics, cleaning and modifications to chemical reactivity. (See Figure 1.2)

1.2.3 Ultrasonic Welding Ultrasonic welding is one of the major uses of power ultrasound in industry today. It has the advantages of not involving long heating and cooling cycles in comparison to more traditional methods. Welding of plastics in this manner also provides a weld with a high joint strength that is equivalent to between 90-98% of the normal material strength.8 Application of such technology to is not just limited to plastics, it has also been used weld snap on lids to paint tins and weld aluminium, which is extremely difficult by more conventional means due to its hard oxide layer. Welding metal
5

ultrasonically, by producing lateral vibratory movement,

leads to the oxide layer being broke up and adsorbed into the metal surround the weld. As with the plastic, this forms a high strength weld by preventing the formation of brittle inter-metallic compounds. Such flexibility also allows for the precision welding of delicate components e.g. musical instruments instead of using the more common hard soldering method.

1.2.4 Biological Uses of Ultrasound The main use of ultrasound within the field of biology is for the disruption of cell walls to release their contents (particularly genetic material) for further examination. Outer cell wall collapse is achieved using a 20 kHz probe to produce cavitation. The cavitation acts in machine gun fashion, with bubbles driven at very high speed from the tip of the probe into the cell wall, eventually leading to penetration and subsequent disruption. This method reduces the cells to their components with limited denaturation of the cellular material, e.g. macromolecular protein and nucleic acids, if appropriate measures to limit the bulk heating effects caused by cavitational collapse are used, i.e. keeping the sample cool during sonication.

1.2.5 Ultrasound in Medical Applications (i) Therapeutic applications

Power ultrasound found its first medical use as an alternative to massage in the 1930s. The mechanical movement of the tissues by the waves of ultrasound was found to mimic the rubbing movements of a masseur and as a consequence improve circulation and muscle physiology. Today, the ultrasound can be applied direct to the skin of the afflicted area, caused by e.g. sporting injuries, strains etc., by the use of flat, earthed, quartz crystals and is a common treatment in the physiotherapists armoury. Another, more recent, use of power ultrasound has been the removal of kidney stones. By using the mechanical effect of the ultrasound the stone can be annihilated and excreted via the patients urine, without the need for invasive techniques. Ultrasonic baths have also been used for sterilisation of medical instruments e.g. scalpels, forceps etc. (see section 1.2.8)

(ii)

Diagnostic Uses

This utilises high frequency (diagnostic) ultrasound to provide a noninvasive technique for human body scanning. It finds particular in foetal imaging9 and continuous/non-hazardous visualisation of surgical instruments during in-vitro procedures. The basis for what are termed Percussion Techniques stem from an 18th century method of tapping on the skin of a patient and listening to the resonant note produced from either the either air or fluid areas of the body. Diagnosis is then

based on the pitch and quality of the resonant note. Modern day percussion involves high frequency pulses of acoustic ultrasound (3-10 MHz) of short duration. Backscattering of the pulse from the boundaries in the tissues leads to a series of echoes, which can be detected at the skin.

1.2.6 Engineering and Ultrasound

(i)

Ultrasonic Machining (USM) With the increasing use of hard, brittle materials e.g. carbides, stainless

steel, glass, ceramics, an alternative to the more usual methods of machining had to be developed.10 Ultrasonic machining usually involves a 20 kHz cutting tool in an abrasive slurry (containing alumina, silicon or boron carbide) and is shaped so as to produce the required profile in the workpiece. An alternative is to use ultrasound to augment traditional machining methods, which gives an increased efficiency.

(ii)

Abrasive Jet Machining (AJM) Abrasive Jet Machining is similar to conventional sandblasting with the

exception that is more controllable and uses a finer abrasive. Such advantages mean it can be used to clean, deburr and cut a variety of hard and brittle materials mica, germanium, glass, ceramics 5.

(iii)

Drilling and Cutting Use of ultrasound by the aerospace industry includes the use of

ultrasonicated tungsten carbide cutting blades for chiselling complex carbon fibre shapes prior to sealing with epoxy resin. This allows continuous production of a quality which would be unattainable by more established methods. Ultrasound is also used to drill carbide turbine blades in aero-engines due to their fragile nature.

(iv)

Metal Tube Drawing Application of ultrasound to the cold drawing of metal tubing has led to

significant improvements. Ultrasonication of the die with 20 kHz vibrations leads to a reduction in draw pressure and draw times, and gives a better finish to the final product. Similar results have been found for the ultrasonic drawing of wire.

1.2.7 Dentistry As with engineering applications (see section 1.2.6) ultrasound is used in the dentists surgery for a whole host of tasks like cleaning, descaling, polishing, drilling and root canal treatments. Cleaning involves an instrument resonating at 25 kHz that ultrasonicates a spray of sodium carbonate, water and air. Projection of this spray on to a tooth surface gives a cleaning/polishing action, which is much gentler for tooth enamel than the previous use of pumice stone grinding. The use of a diamond-tipped drill attachment and alumina slurry allows for the drilling of tooth enamel. The same instrument, with the addition of a file, can be used for root canal treatment that is self-cleaning due to effect of the ultrasonic vibrations transmitted through the irrigation fluid.

1.2.8 Ultrasonic Cleaning Since the 1950s power ultrasound has been utilised for surface cleaning, particularly in industry, in areas as diverse as the electronics, optics and medicines. The usefulness of ultrasound for cleaning depends greatly on the nature of the material to be cleaned. Sound absorbing materials like rubber suffer from mediocre cleaning quality, whereas the method is highly effective for sound-refracting items made of glass, metal and plastic. Known in some quarters as Brushless Scrubbing,11 ultrasonic cleaning achieves its effects via cavitation bubbles which slowly erode the insoluble surface contaminants. A majority of the bubbles formed by cavitation are transient in nature, but a significant number can remain in the solution on a semi-permanent basis and oscillate for multiple acoustic cycles.12 Hence, hard surface contaminant removal can be achieved by the combination of the transient cavitations, which crack the contaminant layers, and the stable bubbles which can lift surface contamination by forming in the cracks between the coating and the surface. Another phenomenon, which occurs during ultrasonication of a liquid medium is microstreaming, whereby liquid is rapidly convected away from the ultrasound transducer surface and through the bulk medium. This convection further enhances the process of cleaning by accelerating the dissolution of contaminants and constantly supplying a fresh solution to the surface of the item to be cleaned. For over forty years the advantages of ultrasonic cleaning have been exploited in a whole host of areas. The technique allows significant time saving and

more importantly allows unusually formed articles with indents and holes to be cleaned with similar efficiencies. As recently as 1980, Geckle13 stated that ultrasonic cleaning proves faster than any conventional cleaning method for the removal of soil and contaminants. C.T. Walker and R. Walker14 provided further evidence for this when they demonstrated that, ultrasonication of metal items in an alkali solution led to a reduction in cleaning times by up to a factor of 1500 when compared to that achieved in a stirred solution.

1.3 Heat Scale Formation and Composition

1.3.1 Introduction While the oxygenation of steel in air is kinetically slow at room temperature, under conditions of intense heat, the increase in rate is such that oxidation products rapidly accumulate on the metal. When, after a relative short period of time, this surface contamination has grown into a hard and thick oxide layer, it is referred to as heat scale. While consisting predominately of an iron oxide matrix, a heat scale may contain some of the alloying elements in the steel as silicone and chromium, which also have a great affinity for oxygen.15 Consequently, determining the scale composition of highly alloyed steels can become an extremely complex problem. The following section will limit itself to a general outline of the heat scale formation of low carbon steels, which, in essence, mechanistically follow the high temperature oxidation of pure iron.

1.3.2 The Classical Three Layer Scale Formation The oxygen partial pressure (PO2) and temperature of the surrounding atmosphere both strongly influence the type of oxide formed on exposed iron surfaces. For example, a classical three layer scale, such as the one shown in Figure 1.3 is formed when oxygen when the oxygen partial pressure of the furnace gas is high (such as for air) and the heating temperature is above 570C.16 Under such conditions, a compact adherent scale is formed consisting of three distinct layers, each representing a different oxide phase. Their sequential appearance in the scale is dictated by the relative availability of oxygen. At the oxide/gas boundary where oxygen is abundant, a layer of hematite (Fe2O3) develops, as it is thermodynamically the most stable oxide of iron at high PO2. In the middle region

10

Fe2O3 Fe3O4 FeO

Steel

Figure 1.3: Classical three layer scale (not to scale) consisting of wrsite (FeO), magnetite (Fe3O4) and hematite (Fe2O3).

11

of the scale where the oxygen availability is moderate, the formation of magnetite (Fe3O4) dominates. Finally, at the metal/oxide interface, a layer of wrsite since it is the most stable iron oxide at low PO2. This innermost layer is normally represented as the stoichiometric compound FeO but in reality, is a grossly non-stoichiometric phase. This is due to its high iron deficiency and, in consequence, is more accurately described as Fe1-yO, where y is a measure of the concentration of the vacancies,17 typically 0.95.18 Diffraction studies have shown that the complex clusters of crystals, which often exist in wrsite19,20 comprise of vacant octahedral sites (normal iron vacancies) as well as tetrahedral sites occupied by iron ions (i.e. interstitial iron ions).19 the chemical formulae of magnetite and hematite, Fe3O4 and Fe2O3 respectively, represent their ideal composition but both tend to be nonstoichiometric. However, such deviations are significantly smaller (particularly for hematite) than is the case for FeO. Wrsite is known to exhibit the rock salt crystal structure, magnetite exhibits a mixed Fe2+ and Fe3+ spinel structure, and hematite is known to have a corundum structure.21 The source of such heat scales on the surface of wires comes from the various stages of manufacture that are used to produce the final product, which include wire drawing, annealing, patenting, hardening and tempering.

1.4 Ultrasonic Effects on Aqueous Media

1.4.1 Introduction In general, most cleaning applications involve the use of power ultrasound (within the region of 20-100 kHz, see section 1.2.2) within an aqueous medium. The sound energy transmitted into the solution at these lower frequencies is significantly greater than is the case for high frequency ultrasound giving rise to the rapid formation of cavitation bubbles, which are responsible for the scrubbing action of the ultrasound.22 Application of the longitudinal ultrasonic waves (see section 1.1.1) to a solution leads to the creation of physical compressions and rarefractions along the axis of propagation. (See Figure 1.1) The speed of this propagation has been reported to be as much as 1505ms-1 in degassed water with an equilibrium temperature of 31C23. The passage of sound waves through the liquid causes the propagation of periodic pressure and velocity variations as the water molecules oscillate around equilibrium positions. These variations are clearly illustrated by

12

PRESSURE

VELOCITY

Figure 1.4: An illustration of the phase differences between sound wave pressure and velocity. (Reproduced from Ref.24).

Figure 1.5: Illustration of the standing wave pattern set up due to interference between incident and reflected sound waves. (Reproduced from Ref.25)

13

(Figure 1.4) and show the relative velocity and pressure changes in successive liquid layers. Both velocity and pressure variations occur along the axis of wave propagation, but they are out of phase by 90 with the pressure lagging behind velocity. When these propagating (travelling) sound waves reach an interface liquid/air or liquid/solid they are reflected. This reflection leads to the formation of a Standing Wave Pattern caused by the interference between the incident and reflected waves (see Figure 1.5). Areas where pressure excersion is at a minimum are termed Pressure Nodes and can be observed at three key areas:

(I) (II) (III)

Transducer/liquid interface The half wavelength point (1/2) The liquid/air boundary.

Conversely areas of maximum pressure (Pressure Antinodes) occur at the 1/4 and 3/4 points. For travelling waves velocity nodes are out of phase by 90 (1/4) with pressure nodes in a standing wave pattern.

1.4.2 Acoustic Cavitation and Streaming Cavitation occurs when bubbles cavities filled with gas or vapour develop within the body of a liquid.26 A good example of this phenomenon is the effervescence of supersaturated gas containing liquids. As previously mentioned (in section 1.1.2), cavitation was first alluded to by Sir John Thornycroft during investigations into HMS Darings propeller corrosion. This was followed by Langevins
27

observation of acoustic cavitation in liquids due to ultrasound. By

1936, the first observations of the presence of cavitation in degassed liquids (at room temperature and pressure) were reported by Sllner.28 The late 1930s saw the interest in cavitation increase as numerous researchers including Harvey 16, Boyce
16 29

and Kornfeld/Suvorov

began to investigate various facets of the phenomenon

and increase understanding of the mechanisms of its inception.

(i)

Acoustic cavitation The formation of cavities in response to an alternating acoustic pressure is

known as Acoustic Cavitation. This can be generalised to include all observable

14

activity i.e. bubble formation, motion and lifetime caused by the acoustic field. Henceforth cavitation will be taken to mean acoustic cavitation. Applying ultrasound of a high intensity to aqueous media leads to the formation of alternating cycles of compressions and rarefractions (see section 1.4.1). Passage of these sonic-waves through the liquid establishes a sinusoidal wave of varying acoustic pressure (PA): Pa = PA sin 2fT Equation 2

Where PA is oscillating acoustic pressure amplitude, Frequency, T time. This Applied Acoustic Pressure is superimposed on the pre-existing, ambient Hydrostatic Pressure (PH). Pressure reaches a maximum during the positive half cycle when the molecules in the solution reach a maximum compression. Therefore maximum liquid pressure (PL) is defined as: PL = P H + P a Equation 3

Conversely the negative half cycle occurs when the molecules reach their most spaced, or rarefied state, leading to the minimum liquid pressure: PL = P H - Pa Equation 4

During the negative half-cycle the liquid pressure (PL) can actually become negative i.e. when Pa > PH. Should PL exceed the intermolecular binding forces within the fluid the molecules are torn apart and cavitation is induced. An ideal cavity is one that is purely a void and collapses almost instantaneously during the following compression cycle. This collapse releases huge

15

quantities of energy that can generate temperatures as high as 5000K, pressures in the region of 1000 atmospheres
30

, weak light emission or sonoluminesence

31

arising from the collapse of cavities often accompanies sonication of water at 25C. Such violent implosions are known as Transient Cavitations32 and are generally formed by lower frequency ultrasound with intensities in excess of 10Wcm-2.19 The existence of transient cavitation bubbles is short-lived (usually less than one cycle) and typically involves expansion to a radius that is two times greater than the initial radius of the nucleus, before ending in the violent collapse. This implosion is considered to be so extreme as there is so little time from initiation to collapse for gas or vapour to diffuse into the bubble and cushion the impact of the imploding liquid surfaces. Transient cavitation is not the only type of cavitation that can be observed. Stable Cavitations last for many acoustic cycles and oscillate in a non-linear fashion around an equilibrium size. It is thought that such micro-bubbles owe their stable nature to the presence of gas and/or vapour within the bubble preventing the implosion. Thus the lifetime of cavitation bubbles is dependent on whether there is sufficient time for gas and vapour diffusion into the cavity during the rarefraction period. If there is, this allows for the formation of stable cavitation, if there isnt the bubble is mainly a void (or highly rarefied solvent vapour) tending to favour a more transient nature to the cavitation.

(ii)

Acoustic Streaming In addition to the production of cavitation, application of ultrasound to

liquid media creates a continuous displacement of particles around their equilibrium positions in what is termed Acoustic Streaming. This displacement is also sinusoidal in nature (like the variation in acoustic pressure) and is given by: Y = YA sin2fT Equation 5 Where Y is particle displacement, YA particle displacement amplitude, Frequency,

16

T time.

Production of a continuous movement of the particles within the medium produces the effect of displacing loose contaminants from a contaminated surface and constantly supplying fresh solution to the surface during ultrasonic cleaning. (See section 1.2.8)

1.4.3 Factors Affecting Cavitation A number of experimental parameters can influence the extent and type of cavitation formed within a liquid medium by the application of ultrasound. This is of prime importance as the production of cavities is key to the ability of solution to provide ultrasonic cleaning via the scrubbing action.

(i)

Effect of Temperature Increasing temperature causes a reduction in the intensity of ultrasound

required for the formation of cavitation within an ultrasonicated medium. This is probably due to the lowering of the surface tension and/or viscosity of the liquid, either of which would lead to a reduction in the cohesive forces within the fluid and the energy required to tear it apart. However, near a liquids boiling point, the high temperature causes a significant increase of vapour pressure. So, whilst high temperature favours the nucleation of cavities the presence of liquid vapour within the bubble leads to a cushioning of the implosion during the compression cycle. Hence stable cavitation is favoured and the impact of ultrasound is reduced, as the shock wave released on cavity implosion becomes less intense. The effect of temperature on cavitation in tap water is clearly illustrated in Figure 1.6. Increasing temperature to 55C leads to an analogous increase in the intensity of cavitation. Further increases in temperature towards waters boiling point sees a reduction in cavitation intensity observed due to the increase in vapour pressure and its associated cushioning effect. Also noteworthy is the hysteresis that occurs when the water is heated to boiling and then cooled back down to the 50C 60C optimum. The upper curve in Figure 1.6 shows that the intensity of cavitation is increased when compared to the initial optimum, probably due to the degassing of solution caused by the boiling of solution.

17

Figure 1.6: The effect of temperature on cavitation in tap water and its hysteresis effect (Reproduced from Ref. 33)

Figure 1.7: Effect of ultrasonic frequency on the cavitation threshold (Reproduced from Ref.34)

18

(ii)

Dissolved Gas and Particulate Matter A majority of liquids are heterogeneous in nature and contain a certain

amount of dissolved gas or gas bubbles. These gas bubbles can have a positive effect on the frequency of cavitation when contained in an ultrasonicated liquid as they can act as nuclei for the growth of cavitation bubbles. (See Figure 1.7) Unfortunately if there is a large amount of dissolved gas present it can also have a detrimental effect on the cleaning effects of cavitation as the cushioning of shock wave intensity provided by such relatively large quantities of gas far outweighs the increase in cavitation frequency. An analogy often used to explain the cushioning effect of dissolved gas is that of the fizzy drink bottle. When the bottle is sealed it is done so under pressure, so as to maintain the drinks fizziness by super saturating it with gas. On opening the bottles pressure is suddenly reduced, leading to an immediate release of dissolved gas that floats to the surface and diffuses into the surrounding atmosphere. A similar thing happens as the pressure is quickly reduced during the rarefraction period of an ultrasound cycle. Here the reduction in pressure causes dissolved gas to be drawn towards and into a cavitation bubble. In the succeeding compression cycles this gas softens, and can even prevent, cavity collapse. Prevention of collapse inevitably leads to an increase in gas attracted into the bubble on subsequent rarefractions causing it to grow, until it floats to the surface and discharges into the atmosphere. Particulate matter can also lower the cavitation threshold by acting as sources of trapped vapour gas nuclei.

(iii)

Ultrasound Intensity (Irradiation Power) Increasing the intensity intensifies the vibration amplitude. As a

consequence, collapse pressure rises causing faster and more violent transient cavitation implosion. However there is a limit to the intensity of irradiation that can be applied to a system due to practical and engineering considerations. At sufficiently high acoustic intensities a Decoupling Phenomenon occurs leading to a loss of power being transferred into the medium from the source. This decoupling is due to the source of ultrasound being unable to remain in contact with the liquid medium for the complete cycle.

19

Another consequence of high power ultrasound on a liquid medium is the increase in the number of cavitations per unit volume. A large concentration of cavities can cause a significant amount of conglomeration leading to the formation of larger, more stable bubbles
26

. Such bubble clusters also have the effect of

dampening the sound energy as it passes through the solution causing a decrease in ultrasound impact and its application to surface cleaning for example. From the point of view of transducer design, an increase in irradiation intensity requires a greater dimensional change in the transducer material. Such changes can increase the strain on these components causing them to break and hence reduce the equipments lifetime. Compromise is therefore necessary to give both optimum performance and longevity.

(iv)

Ultrasound Frequency To produce gas or vapour filled voids by completely rupturing a liquid

requires a finite time. High frequency sound waves can suffer problems as the time needed to create a bubble is longer than the time available during the rarefraction cycle. For example at 20 kHz the rarefraction cycle lasts 25s (=) with

maximum negative pressure at 12.5s, whereas at 20 MHz the rarefraction cycle is 0.025s.35 Consequently as frequency is increased, cavitation bubble formation becomes more and more difficult. This problem can be overcome by increasing ultrasound intensity, which increases acoustic pressure amplitudes and in turn gives a more powerful rarefraction cycle than can overcome the liquids intermolecular forces. Figure 1.7 demonstrates this quite clearly, illustrating the variation of threshold intensity with ultrasonic frequency. As can be seen on the plot, there is a significant rise in the ultrasound intensity required to produce cavitation above ~100 kHz. Operating transducers at this and higher frequencies with sufficient intensities is extremely difficult. With this in mind, the typical range used for a power ultrasound application e.g. cleaning, is normally between 20-50 kHz (anything lower maybe audible which can produce discomfort in the ear of the user.)

20

1.5 Electrochemistry
1.5.1 Electrochemical Reactions Electrochemical reactions are concerned with the reduction and oxidation (so called 'REDOX') processes that occur heterogeneously at a surface which is electrically conducting i.e. an electrode. Reduction processes or 'Cathodic

Reactions' are so called because they occur at the cathode and have the general formula: ne - + Ox n + Red Equation 6

Where Red and Ox are the reduced and oxidised forms of the Redox Couple. Such cathodic reduction reactions include metal deposition from an ionic solution 36

M n+

( aq )

+ ne - M (s) Equation 7

And the cathodic evolution of hydrogen gas: 2H + ( aq ) + 2e - H2 (g) Equation 8

The opposites of such reactions are the oxidation processes or 'Anodic Reactions that occur at the anode. The general form for these anodic oxidation reactions is: Red Ox n + + ne Equation 9

Commonly encountered anodic reactions include the solvation of metal ions:

21

M (s) M n +

(aq)

+ ne Equation 10

E.g.

Fe (s) + Fe 2 +

(aq)

+ 2e Equation 11

And anodic evolution of oxygen from water: 2H2O 4H + + O2 + 4e Equation 12

It is worthy of note that both the anodic and cathodic processes can occur at the same electrode surface simultaneously. Each redox couple has its own, unique electrode potential (Eeq) where there is no net oxidation or reduction occurring i.e. the couple is in equilibrium (N.B. Cathodic and anodic rates are non-zero here, a finite exchange velocity is associated with the process of equilibrium.) When such conditions are reached the value of Eeq can be determined by the Nernst Equation37: RT [Ox ] ln nF [Red ] Equation 13

Eeq = E

Where E is the standard electrode potential of a couple R is the universal gas constant F is the Faraday constant N is the number of electrons exchanged T is the temperature [Ox] concentration of the oxidised form of the couple [Red] concentration of the reduced form of the couple

22

Application of an external current produces an overpotential, i.e. causes the electrode potential to be displaced from Eeq resulting in the net oxidation or reduction of the redox couple. When this occurs electrical currents are related to observed chemical rates by Faradays law, i.e.: iA = nF (anodic rate) Equation 14 (Anodic currents are positive by convention) iC = - nF (cathodic rate) Equation 15
(Cathodic

currents are negative by convention)

And
i = iA + iC

Equation 16

Where iA is the anodic current component iC is the cathodic current component i the external circuit current

The rates (and therefore current contributions) of electrochemical reactions depend, exponentially on the potential of the electrode and are expressed in the following form: For an anodic reaction38 (1 - )nF iA = io exp (E - Eeq) RT Equation 17

23

Figure 1.8: Anodic and cathodic Tafel Plots (Reproduced from Ref.39)

Diffusion Limited (Diffusion Controlled)

Potential

Tafel Region (Potential Controlled)

Log Current
Figure 1.9: Illustration of the Tafel plot diffusion limited current plateau

24

A cathodic reaction - nF iC = - io exp (E - Eeq) RT Equation 18 where io is the exchange current of the redox equilibrium (A measurement of the exchange velocity) is the 'Asymmetry Factor' of the redox reaction When the applied overpotential | E Eeq| is large (>75mV), then reverse currents become insignificant i.e. either iA or iC predominates within i leading to straight line plot of E verses Log i. These plots are known as 'Tafel Plots' and can be observed experimentally. (Figure 1.9) The current increases exponentially with potential until it eventually becomes limited by diffusion processes, leading to a plateau on the Tafel plot (Figure 1.8.) Another factor that can have a bearing on the appearance of the Tafel plot for anodic metal dissolution is metal surface passivation. To simplify the understanding of the metal redox system, the most likely anodic reactions for a given metal are plotted verses pH on what is termed a 'Pourbaix Diagram'. (See Figure 1.10.) Such a diagram for a metal indicates the following zones:

(i)

Corrosion (Active)
M/M n+

This is where anodic metal dissolution occurs, E > E

leading to the

production of water soluble species (ions) and corrosion of the metal (where E is the standard electrode potential for the metal.)

(ii)

Immunity
M/Mn+

Here E < E thermodynamically.

and anodic metal dissolution is impossible,

25

Figure 1.10: Pourbaix diagram showing the equilibrium potential-pH for the FEH2O system, illustrating the areas of immunity, corrosion and passivation. m = the equilibrium of H2O/O2 at Po2 = 1, I = H2O/H2 equilibrium at pH 2. (Reproduced from Ref.38)

26

Transpassive Region Ec

Passive Region

E
Epp EF

Active or Tafel Region

Log i

Figure 1.11: Diagram to illustrate the active-passive regions of a Tafel plot

27

(iii)

Passivation
M/Mn+

E > E

and although dissolution is thermodynamically possible the

reaction leads to water insoluble products (usually and an oxide or hydroxide species.) Deposition of such a non-conductive substance prevents further dissolution by blocking the electrode surface resulting in a kinetic inhibition. When the metal surface becomes passive the Tafel plot displays a rapid reduction in current. (See Figure 1.11) The current increases with electrode potential up to the 'Passivation Potential' (Epp). Once the electrode attains the 'Breakdown Potential' (EC) the passive film decays and current again begins to increase. Such an electrode is termed 'Transpassive' 40 Reduction of the electrode potential sees a similar current-potential relationship with the exception that the passive-active transition is observed at the 'Flade Potential' (Ef). The difference between the Flade Potential and the Passivation Potential (Ef < Epp) is probably due to potential drops that occur as the pre-passive film forms at the active regions upper end and local pH changes associated with anodic current.40 The shape of a Tafel plot for a metal and it's associated values of EC, Epp and Ef depends on the sort of electrolyte used. Some electrolyte anions may cause the complexation of the oxidised metal species produced by the anodic reaction, leading to them becoming soluble and hence hindering the formation of a passive film at the metal surface. Likewise these anions can breakdown a pre-formed passive film. This property is termed 'Aggressiveness' i.e. the ability of an anion to complex and solubilise anodic reaction products. Hence an aggressive anion is one that readily forms soluble complexes and reduces a metal passivity. In contrast a non-aggressive anion is one that either doesn't form a complex with the products of the anodic reaction or reacts to form insoluble complexes, causing the metal surface to remain or become passive. Halide ions are an example of a highly aggressive anion, sulphate to is also termed aggressive though is quite mild when compared to chloride.

1.5.2 The Electrical Double Layer Hypothesis The interface between an electrode and electrolyte is analogous to that of a circuit, (Figure 1.12) with a 'Charge Transfer' resistance (associated with the

28

Rct Cdl

Figure 1.12: The electrode/electrolyte interface is analogous to an electrical circuit in which Rct is a charge transfer resistance and Cdl is the double layer capacitance.

Current

Potential

Time Figure 1.13: Depiction of the electrode potentialcurrent lag caused by double layer capacitance charging

29

electrochemical reaction) Rct and a double layer capacitance Cdl.. Values of Rct are potential dependent due to the exponential relationship of the electrochemical reaction to the electrode potential (see section 1.5.1.) Cdl values are also known to vary with potential.41 However this does not prevent a simple equivalent circuit (shown in Figure 1.12) from being used to explain, qualitatively, pulsed or transient electrochemical current phenomenon. Application of pulsed or stepwise current to an electrode causes a lag between electrode potential and the transient current due to double layer capacitance charging 88 illustrated by Figure 1.13. Also impedance (Z) of the double layer varies with f :

Z = 1 fCdl 2 Equation 19

Whereas values of Rct are frequency independent. Hence low frequencies of AC current produce electrochemical reactions as they tend to pass 'Faradically' through Rct, alternatively high frequency AC produces no electrochemical reaction as it passes through Cdl 'Non-Faradically'.

1.6 Ultrasound in Electrochemistry

1.6.1 Introduction Sonoelectrochemistry
5, 42

involves the coupling of electrochemical systems

and power ultrasound together to produce new processes caused by the influence of ultrasound on reaction kinetics. This phenomenon can be compared to other synergistic approaches in which two independent sources of activation energy are joined together. Such techniques usually result in new methodology for the study of each activation source separately and the detection of new reactions caused by socalled Dual Activation43. Ultrasound may merely initiate processes by activation but also by influences mass transport. Direct exposure of homogenous and heterogeneous chemical reactions to intense ultrasound has a profound effect on reactivity 11,44. Ultrasound has also been

30

usefully applied to organic synthesises that involve electrochemistry,45,46 electroanalytical problems, degradation/mineralisation of toxic materials and wastewater treatment 47. The application of ultrasound, particularly in electroplating has led to significant improvements in terms of higher density/quality metal surface film build-up. Similarly, ultrasound has been shown to produce better quality conducting polymer films when these are generated by electrodepostition. 48 Recent studies involving ultrasound in conjunction with analytical electrochemical techniques found that the application of ultrasound to be most useful under certain conditions. Experiments made by Bard
49

utilised High Speed

50

Coulometry with ultrasound induced mass transport and Dewald and Peterson
51

found that using pulsed ultrasound led to a new form of Hydrodynamic Modulation Voltametry. Compton has been particularly active in this field having been
52

involved in a number of studies including determining the effect of ultrasound on current/voltage characteristics and developing a form of Anodic Stripping
53 54

Voltametry with an added ultrasonic component

. Brett et al

have also shown

the usefulness of in-situ ultrasonicated anodic stripping voltametry and the cleaning effect provided by ultrasound during voltametric nucleic acid detection. Studying the effects of ultrasound on an electrochemical system is not without its difficulties. Ultrasound is known to influence a large number of physical and chemical processes
55

and its application to an electrode surface must be

carefully controlled. In an ideal situation the sound intensity should be known and be able to consistently applied. The effect of ultrasound in electrochemistry can be divided into two areas:

(i)

Homogeneous Effects

Such effects can be induced by cavitation, which is coupled to the pressure changes within the solution (see section 1.4.2.) The collapse of cavitation bubbles, and the associated high transient pressures and temperatures
56, 57

may cause

homolytic cleavage of chemical bonds in compounds present within, or close to the bubble. Such cleavage gives rise to the formation of free radicals e.g. hydroxyl radicals 58, halogen radicals 59 and others 60 depending on the solution composition. Confirmation of such radical formation has been obtained from radical trapping/spectroscopic detection experiments.61

31

If a solvent medium contains dissolved macromolecules, ultrasound can mechanically rupture bonds within these macromolecules by the inducing strong shear forces in solution. Such effects have resulted in the application of ultrasound to limit molecular weight, and hence control material properties during polymerisation processes.

(ii)

Heterogeneous Effects

The most dramatic effects caused by ultrasound on heterogeneous processes are the significant change in mass transport at phase boundaries and solid surface erosion.62 Erosion is caused by asymmetric cavitation near the solid/solution interface
63

and is more prominent on materials such as lead and copper materials,

which have relatively low hardness. Mass transport plays a substantial role in electrochemical as well general heterogeneous reactions. A great number of heterogeneous/electrochemical processes are controlled by mass transport and the variation in transported material to and from interfaces can cause a change in the reaction pathway involved.
64

Electrochemical mass transport depends on the mass transport boundary layer thickness, , (see below) and control of this thickness can be used to vary the nature of the chemical process.

1.6.2 Mass Transport in Electrochemistry Transfer of charge at an electrode surface is accompanied by ion transport within the electrolyte. The transport of uncharged species is dependent on convection and diffusion. However, charged species are influenced by the effect of electric fields and for such species charge migration also contributions to mass transport. Electrochemical processes often owe their limitation to the rate at which reactant is transported to the surface of the electrode. This limited mass transport rate can be increased in a variety of ways including an increase in solution temperature, reactant concentration or fluid agitation.65

1.6.3 Mass Transport Boundary Layer (Nernst Diffusion Layer) The complex nature of the mass transport boundary layer means that simplifications are required. One of these simplifications is to assume that there is

32

Cbulk

Concentration

Extrapolation of initial concentration Actual concentration

Csurface

Distance from electrode

Figure 1.14: Nernst diffusion layer model. The solid line represents the actual concentration profile and the dashed line from Csurface is the extrapolation of the initial slope

33

the presence of a laminar sub-layer close to the electrode surface that possesses a linear concentration gradient. This is equivalent to the assumption that near the electrode mass transport occurs exclusively by diffusion across a constant diffusion barrier. The concentration gradient of the electroactive species the boundary layer (known as the Nernst Diffusion Layer see figure sad) determines the flux density and hence current at the electrode surface. The thickness of the Nernst layer () can be used as a measure of resistance to mass transport. For a given electrode is hydrodynamically determined i.e. the greater the fluid agitation, the thinner the boundary layer and easier the resultant mass transport. Nernst diffusion layer treatment of a typical electrochemical system gives an equation for mass transport of: i lim = nFD (c bulk - c surface ) Equation 20 Where ilim is the limiting current n is the number of transferred electrons F is the Faraday Constant D is the diffusion co-efficient C is the concentration is the diffusion layer thickness The limiting current may be increased by the use of ultrasound because this thins the diffusion layer thickness by increasing the agitation within the fluid. Hence there is an enhancement of mass transport due to the influence of ultrasound. This increase in mass transport has been observed in a number of experiments carried out by both Compton et al 66 and Walton and co-workers.67

34

Chapter 2

Experimental Set-up and Protocol

2.1 Materials
2.1.1 Chemicals All the chemicals utilised in this work, listed below, were of ANLAR purity (unless otherwise stated) and all solutions made up using doubly distilled water. Supplier Fisher Chemicals UK Chemicals Used Sodium Hydroxide (NaOH)

Sodium sulphate (Na2SO4) Sodium chloride (NaCl)

Hydrochloric acid (HCl)

Sulphuric acid (H2SO4) Di-sodium Hydrogen orthophosphate (Na2HPO3) Aldrich Hydrogen peroxide (H2O2) 3-Aminophtalhydrazine (Luminol) Diaminoethanetetra-acetic acid EDTA (CH2N[CH2COOH]2)2 BOC, UK Argon (Ar)

Albright and Wilson, UK Ltd

Phosphoric acid (H3PO4)

36

2.2 Methods
2.2.1 Ultrasound Probe The ultrasound source in all experiments was a Branson Sonifier 250 variable power 20 kHz ultrasound generator with piezo electric transducer horn. The transducer horn was used in combination with a variety of 10mm diameter cylindrical titanium tips including a standard flat tip and a tip ended in a 450 wedge section, both supplied by Branson Ultrasonics Ltd, UK.

2.2.2 Photomultiplier Tube All non-spatially-resolved light intensity measurements were performed using an Electron Tubes Ltd, model QL30F photomultiplier working into a type A1 transconductance amplifier. Light was focussed onto the pmt photocathode using a 3cm-diameter compound glass lens with a 10cm focal length.

2.2.3 Low Light Camera Luminescence imaging and spatially resolved light intensity measurements were carried out using a Merlin Low Light camera with fibre optic interface (LTC 216F40E) and type (CCU 2025) control unit supplied by Custom Cameras. The camera was fitted with a Ziess 55-mm f 2 lens. All images were captured with integration over either 16 or 64 frames (at a sample rate 25 frames per second) to improve the signal to noise ratio and produce clearer images. Greyscale images were digitised in the form of a 320 x 240 resolution matrix of 8-bit pixels using a Xciplite software.

2.2.4 Galvanostat The galvanostat apparatus comprised of a Wenking Instrument LB75 Potentiostat operated in galvanostatic mode by passing current through a standard 1 resistor as shown Figure 2.6. This apparatus was used for all experiments, which involved electrolytic scale loosening.

Figure 2.1: A illustration showing the configurations of the ultrasound transducer horn tips a) the standard flat and b) the 45 angled wedge.

38

2.2.5 Function Generator A Thandar TH501 model, 5MHz Function Generator, RS Components Limited was used in conjunction with the galvanostat (described above) to produce current pulse waveforms.

2.2.6 Potentiostat The potentiostat used for all potentiostatic and potentiodynamic experiments was a Solartron Instruments electrochemical measurement unit (Model SI 1280) under computer control, utilising Omega Pro DC software.

2.3 Calibration of the Ultrasound Probe

2.3.1 Calorimetry Before any of the experimental work could be carried out a calibration of the ultrasonic probe was undertaken to determine the ultrasound output power associated with each of the defined generator intensity settings. This was achieved by the use of calorimetry68. A rise in temperature produced by the addition of any given amount of energy (in this case ultrasound) to a body is determined by its heat capacity. Thus when the body in question is a volume of water the input of energy from the ultrasound probe can be calculated from: m c T t Equation 21

Power (W) =

Where

m c T t

is the mass of water (in grams) is the specific heat capacity (water = 4.2 JK-1) is the change in temperature (C) is the time taken for the observed temperature rise (in seconds)

39

2.3.2 Equipment Utilised in the Calibration of the Ultrasound Probe Figure 2.2 illustrates schematically the experimental set-up used for the calibration of the ultrasound probes power output.

2.3.3 Method A 100ml of distilled water was measured out and placed in the polystyrene cup within the insulated beaker. The solution was then allowed to equilibrate for 5 to 10 minutes until a constant temperature reading was achieved. After equilibration, the water was then sonicated at generator intensity one, with readings every 10 seconds until an appreciable temperature rise was observed (usually a few degrees Celsius). This was then repeated for the individual generator intensities, the results of, which can be observed in Table 2-1 and graphs Figure 2.3

2.3.4 Results Intensity Energy Input - m x

c x T

Time - S (seconds) 404

Power Output (Watts) 5.2

(Joules)

5 x (4.2 x 100) = 2100

8 x (4.2 x 100) = 3360

228.78

14.68

10 x (4.2 x 100) = 4200

148.7

28.24

8 x (4.2 x 100) = 3360

78.78

42.65

20 x (4.2 x 100) = 8400

148.95

56.39

10 x (4.2 x 100) = 4200

59.35

70.8

Table 2-1: Power Outputs for the respective ultrasound generator settings

A D E

C
Figure 2.2: Diagram of the apparatus used for calibrating the ultrasound probe power output. A = 20 kHz ultrasound generator, B = ultrasound probe, C = insulated beaker, D = polystyrene reaction vessel, E = electronic thermometer

50

45

40

Temperature

35

30

25

20 0 50 100 150 200 250 300 350 400 450 500

Time (seconds)

Figure 2.3: Calibration curves for ultrasound generator power output levels: x 1, s2, v3,

4, 5, q6.

2.4 Luminol Sonochemilumunescence Experiments

2.4.1 Experimental Set-up The apparatus was set-up as illustrated below. The ultrasound was introduced into the buffered solution of EDTA, luminol and peroxide (e) via the ultrasonic transducer horn (a) at varying power levels. The temperature of the solution was maintained at 50C by the constant circulation of thermostated water (f) around the cell (e). For the capture of non-spatially resolved SCL data the apparatus was used in conjunction with (d) the photomultiplier tube, focussed on the area directly beneath the tip of the ultrasound transducer horn with a 10cm focal length, compound glass lens. For the capture spatially resolved data a low light (CCD) camera was employed.

b d c

f e f

Figure 2.4: Schematic diagram of the experimental set-up (a) ultrasound transducer; (b) thermostated cell; (c) optically flat window; (d) photon multiplier tube/low light camera; (e) buffered solution of EDTA, luminol and peroxide; (f) thermostatic fluid; (g) lightproof box.

43

2.5 Wire Cleaning Experiments

2.5.1 Wire Samples Wire supplied by Garphyttan Wire Limited was used throughout all the experiments. This wire consisted of steel with small amounts of carbon (0.50%), silicon (1.40%), chromium (0.70%) and manganese (0.70%). As a result of the manufacturing process there was an oxide layer of approximately 15 m (determined by microscopic analysis) covering the surface of the metal. 2.5.2 Sample Preparation Test samples for electrolytic/ultrasonic cleaning consisted of 15cm lengths of wire complete with a watertight tip of epoxy resin. A coating of insulating tape was utilised to provide a predetermined surface area for electrolysis (See Figure 2.5).

Steel wire

Exposed area

Waterproof epoxy resin tip

Insulating tape Figure 2.5: An illustration of the wire sample preparation.

44

B
To working electrode

From potentiostat output

A
To potentiostat reference inputs

Figure 2.6: Set-up of the 1 resistor to provide galvanic output from the potentiostat. A = 1V square waveform, B = 1 resistor, C = 1A square waveform.

45

Figure 2.7 illustrates the equipment used for the ultrasonic used for the ultrasonic probe distance/cleaning experiments. In light of earlier research
69,70

all

cleaning experiments employed 2 minutes of square wave current with a 95% anodic duty cycle, frequency 1 Hz and 1A of current provided by the galvanostat/potentiostat (C) via the function generator (B) to soften the surface scale to ensure that all observed descaling was entirely due to ultrasound intensity alone. After the electrolytic pre-treatment the wire was then subjected to 10 second bursts of ultrasound (the minimum time required by the generator to produce a stable ultrasound field) for a maximum of 120 seconds in total. After application of ultrasound the surface was examined and the percentage of the remaining scale visually assessed. The experiments were repeated for a number of probe to wire distances and at a number of ultrasound output powers.

B A

D C F E

Figure 2.7: Diagrammatic illustration of the experimental set-up for the probe distance/cleaning tests. (A = 20 kHz generator, B = function generator, C = potentiostat/galvanostat, D = ultrasound probe, E = wire sample, F = graphite counter electrode, G = inert plastic support

47

2.5.3 Electrolyte Preparation All experiments utilised a 10% w/w solution of sodium sulphate at pH 7. A standard volume of solution (1400ml) contained in a beaker - thermostated at 50C by a stainless steel coil connected to a Grants Model L14 waterbath - was allowed to reach a thermal equilibrium before any pH adjustments were made. Adjustments were carried out by the addition of small quantities of sulphuric acid or sodium hydroxide. Measurement of the pH was made using a Jenway (Model 3071) pH Meter that was calibrated prior to each use with buffers of pH 4 and pH 10 at 20C. The cleaning solution was then subjected to 45 minutes sonication in an ultrasound bath to ensure that total degassing of the liquid and optimum ultrasound efficiency for the descaling process.

2.5.4 Electrolytic Current Pulsed anodic current is generated by connection of the galvanostat to a function generator (see Figure 2.6) which allowed the anodic duty cycle, current waveform and frequency to be set. Anodic duty cycle is defined as the percentage time during which the sample is anodically polarised during the total current time period:

Anodic Duty Cycle =

anodic period 100 total period Equation 22

2.5.5 Current Density Current density on the wire samples was kept constant at 1Acm-2 by keeping the current passing between the working and the counter electrode at 1A and the exposed area on the prepared wire sample constant at 1cm2, this can be calculated from: Current density = current exposed area Equation 23

48

2.6 Sonotrode Experiments

Experiments were carried out to determine the current voltage characteristics (Polarisation Curves) of hydrogen evolution and oxygen reduction at the tip of the titanium sonoprobe, with and without ultrasonic stimulation. The experimental set-up is illustrated in Figure 2.8 and measurements were recorded with a Solartron Instruments electrochemical measurement unit (Model SI 1280) under computer control, utilising Omega Pro DC software. The reference electrode used for these experiments was a saturated calomel electrode (SCE) which has a stable potential of 0.268 Volts when compared to the standard hydrogen electrode (SHE). For experiments that involved deaerated conditions the electrolyte solution was purged with pre-purified argon (supplied by BOC, UK) for a period of thirty minutes to ensure total removal of the air present. Cathodic polarisation curves were obtained by using a Linear Sweep Voltametry (LSV) technique
71

. LSV involves the potential of the electrode of

interest (working electrode) being varied relative to a fixed reference (reference electrode potential) at a constant rate. The flow of current between the working electrode and the electrode solution is recorded as a function of the working electrodes potential. For all experiments to determine the cathodic polarisation characteristics of the tip of the titanium sonoprobe, the LSV potential sweep commenced at 2.5 V and was swept at a rate of 0.014 Vs-1 to 0 V verses SCE. To ensure that all voltammograms would be reproducible, the potential of the sonoprobe tip was cycled from 2.5 V to 0 V at a rate of 0.014 Vs-1 without ultrasonic stimulation, and before the collection of data, until a stable pattern was observed (typically < three cycles.)

To sonic horn control unit Transducer

Electrochemical measuring unit

Computer

Perspex Ar inlet for degassing INSULATED SONIC HORN Graphite counter electrode Ti tip Magnetic stirrer
Figure 2.8: Schematic Diagram of sonotrode and electrochemical cell arrangement (Reproduced from Ref. 72

Electrolyte Reference electrode

Chapter 3 The Visualisation of Ultrasonically Induced Cavitation with Luminol

51

3.1.1 Introduction Alkaline solutions of luminol have been known for quite sometime to emit light when exposed to a source of power ultrasound. The observed light intensity is significantly more (orders of magnitude) intense than the visible sonoluminescence caused by the direct sonication of aerated water 63,73,74,75,76 and is thought to be due to an oxidative chemiluminescent process involving OH generated by ultrasound. This method of sonochemically inducing the chemiluminescence of luminol has been used in a number of studies to investigate the cavitation mechanics 63,76. The spatial distribution of cavitation in solution has also been determined by using this set-up in conjunction with either a scanning fibre optic probe photographically capturing images
74,75 75

or by

. As for the kinetics and mechanisms

77,78

involved in the sonochemiluminescence of luminol not a great deal is known. Previous studies of both luminol chemiluminescence chemiluminescence
79,80

and electrogenerated

in aqueous solution have demonstrated that pH and

hydrogen peroxide concentration have a dynamic effect on the emitted light intensity. Investigation of sonoluminescence involving luminol with additional H2O2 is affected by divalent transition metal cations
81,82,83,84

that catalyse a non-

sonochemical/background chemiluminescence. Solution contamination by transition metals is introduced through reagent impurities or via leachate from equipment, which is nigh on impossible to avoid totally. The solution to the problem is to add a chelating agent, like EDTA, which is able to complex and thus prevent the catalytic actions of the divalent transition metal cations 81,84.

3.2 Materials
3.2.1 Chemicals All chemicals used in this work (detailed below) were of ANLAR grade and all solutions created with doubly distilled water.

3.2.2 Experimental Equipment The experimental set-up used is illustrated in Figure 2.4. Image analysis and iso-luminescence contour plot generation was performed using Surfer cartography software obtained from Golden Software Ltd. In all cases where digitised images were subject to quantitative analysis care was taken that the image

52

bitmap contained no areas of saturation i.e. that all the 8 bit pixel values fell between zero and 255. Care was also taken that the cameras depth of field was adequate to keep the luminescent feature in sharp focus. The object distance was typically 40cm.

3.3 Experimental Details

3.3.1 Kinetic Measurements (i) Effect of pH and H2O2 concentration

Experiments to determine the effect of H2O2 concentration on the intensity of luminol sonoluminescence over a range of solution pH (pH 7-13) at an ultrasound power of 60W. A luminol stock solution containing 10-3M of luminol, aqueous phosphate buffer Na2HPO4 (0.1M) and 5 x 10-6M EDTA was prepared using doubly distilled water. A 50ml aliquot of this stock solution was taken, heated to 50C and the pH adjusted by addition of either 0.3M aqueous NaOH or 0.1M aqueous H3PO4. Variation in concentration of H2O2 was achieved by additions of known volumes of 0.02M aqueous H2O2 prepared by volumetric dilution of a 30% stock. All H2O2 additions came from prepared solutions that were less than 12 hours old to ensure no decay of the peroxide concentration. The aliquot was then placed in the reaction cell, allowed to thermally equilibrate for 5 minutes before sonication by the ultrasound probe. Results were plotted as relative intensities verses hydrogen peroxide concentration. (See Figure 3.1)

(ii)

Effect of Ultrasound on Sonochemical Luminescent Intensity.

50ml of the luminol stock solution (10-3M luminol, 0.1M Na2HPO4, 5 x 10-6 EDTA) was adjusted to pH 12 with 0.3M aqueous sodium hydroxide solution. This was then placed in the glass reaction cell and allowed to equilibrate to 50C before addition of 10-4M H2O2. The solution was then sonicated at each of the calibrated ultrasound generator intensities (see section 2.3) and the results plotted as a graph of power input (W) verses the relative intensity of sonoluminescence. (Figure 3.2)

53

(iii)

Effect of EDTA

A solution containing 10-3M luminol, 0.1M Na2HPO4 and 10-4M H2O2 was prepared. A 50ml aliquot of this solution was adjusted to pH 12 by using 0.3M aqueous sodium hydroxide, placed in the glass reaction vessel and allowed to equilibrate to 50C. Before sonication, a silent or background reading was taken from the photomultiplier tube with another reading recorded during sonication. The experiment was then repeated with the addition of 5 x 10-7M EDTA and the results recorded by the photomultiplier tube. EDTA was then added in a stepwise fashion up to a maximum of 5 x 10-5M.Results were recorded prior to (background/silent) and during sonication. These results were plotted as relative background (unsonicated) or sonicated intensities against concentration of EDTA and are illustrated in Figure 3.3.

3.3.2 Image Capture and Analysis The apparatus was used in conjunction with a solution comprising of luminol (10-3M), H2O2 (10-4M), Na2HPO4 (0.1M) and EDTA (0.02M) adjusted to pH 12 as previously described, with sodium hydroxide. Experiments were initially carried out on the flat, 10mm diameter probe tip. Sonication of the luminol solution was carried out at the calibrated ultrasound generator intensities, two, four and six (see 2.3) with an individual image being captured for each separate intensity by CCD video camera. These images were stored on PC via the video capture unit as greyscale images, digitised in the form of a 320 x 240 resolution matrix of 8-bit pixels using a Xciplite software prior to further analysis using Surfer cartography software. The experiments were then repeated with a wedge shaped tip (also 10 mm in diameter).

3.4 Results and Discussion

3.4.1 Kinetic Investigations The Influence of H2O2 concentration ([H2O2]) on spatially unresolved SCL intensity was studied over a range of solution pH at constant temperature and using constant ultrasound power. Figure 3.1 shows the relative intensity of SCL emission (ISCL) as a function of [H2O2] obtained from a solution containing luminol and

54

EDTA at pH values between pH 7 and pH 13. It may be seen from Figure 3.1 that any given value of [H2O2] ISCL increases monotonically with solution pH up to pH 12. Above pH 12 the trend becomes reversed and ISCL is decreased at pH 13. It may also be seen from Figure 3.1 that at pH 10 increasing [H2O2] has no significant effect on ISCL. At pH > 10 however, ISCL increases monotonically with [H2O2] for [H2O2] < 10-4M. The effect is most pronounced at pH 12 where the value of ISCL approximately doubles as [H2O2] increases from 10-6M to [H2O2] = 10-4M. At [H2O2] > 10-4M the trend is reversed and ISCL decreases with increasing H2O2 concentration. The dependence of ISCL on ultrasound power was investigated at pH 12 using a solution containing luminol, EDTA and 10-4M H2O2. Figure 3.2 shows the substantially linear relationship (linear correlation coefficient = 0.9998) relationship observed between ISCL and ultrasonic output power under these conditions. The observed dependencies of ISCL mechanisms shall be discussed in the subsequent sections. The effect of EDTA in suppressing the background (silent)

chemiluminescence of aqueous luminol / H2O2 is illustrated in Figure 3.3. Figure 3.3 shows the relative intensity of light emitted by a solution containing luminol and H2O2 at pH 12 as a function of EDTA concentration, under both sonicated and silent conditions. It may be seem from Figure 3.3 that the addition of EDTA has little effect on the ISCL, which decreases by < 5% as the EDTA concentration increases from zero to 5 x 10-5M. However it may also be seen from Figure 3.3 that the intensity of chemiluminescence under silent conditions decreases monotonically with increasing EDTA concentration and is reduced by approximately 95% at an EDTA concentration of 5 x 10-5M. Aqueous solutions of divalent transition metal cations such as Cu2+, Fe2+, Co2+ and Mn2+ are well known to catalyse the chemiluminescence of luminol in the presence of dissolved O2 and/or H2O2.81,82,83 In the presence of H2O2, significant increases in emitted light intensity may be produced by even trace concentrations of transition metal cation.83 The exact mechanism of this transition metal catalysis is not known but the production of HO through a Fenton type reaction has been suggested as an important step81 i.e.

1.4 1.2

ISCL, arbitrary units

1.0 0.8 0.6 0.4 0.2 0.0 10

-6

10

-5

10 [H2O2] , M

-4

10

-3

Figure 3.1: The effect of H2O2 concentration on spatially unresolved ISCL at different values of solution pH. x pH 7, pH 8, w pH 9, v pH 10, s pH 11, q pH 12, pH 13 (10-3M luminol, 10-4M EDTA, Temperature 50C ultrasound power 70W).

1.6 1.4

light intensity, arbitrary units

1.2 1.0 0.8 0.6 0.4 0.2 0.0 0 20 40 60 80 100

Ultrasound power, Watts

Figure 3.2: Relationship between ultrasound power output power and spatially unresolved ISCL (10-3M luminol, 10-4M H2O2, 10-4M EDTA, Temperature 50C, pH 12).

1.2 1.0

light intensity, arbitrary units

0.8 0.6 0.4 0.2 0.0 0 1 2 3

-5

[EDTA], M x 10

Figure 3.3: The effect of EDTA concentration on spatially unresolved intensity of luminol chemiluminescence under: v sonicated and q silent conditions (10-3M luminol, 10-4M H2O2, pH 12, Temperature 50C).

58

M 2 + + H 2 O 2 M 3+ + HO - + HO Reaction 1 It is also known that the complexation of divalent transitional metal cations by chelating agents such as EDTA greatly diminishes their ability to catalyse luminol chemiluminescence.81,83 On the basis of the above it is proposed that the levels of background light emission observed in the absence of added EDTA result from the luminol chemiluminescence catalysed by traces of transition metal cation present in the reagents and/or water used in the experiments. It is further proposed that the observed suppression of background chemiluminescence by EDTA result from the complexation and deactivation of the trace metal cations. The finding that ISCL is not significantly reduced at EDTA concentrations up to 5 x 10-5 M implies that EDTA acts specifically to inhibit the background chemiluminescence. This suggests that the EDTA cannot be acting as a reductive quencher for HO or O2- radicals (vidi infra) and so tends to confirm the hypothesis that is the properties of EDTA as a chelating agent that are important here. 3.4.2 Mechanism: sonochemical generation of OH and O2The propagation of ultrasound waves in aqueous solution leads to cyclic pressure variations, which cause the nucleation, growth and collapse of microscopic cavitation bubbles filled with gas and/or vapour
85,86

(see section 1.4.2.)

Furthermore, it has been shown that the extremely high local temperatures and pressures may be generated during the collapse or implosion of such bubbles.85,86 Consequently, it is generally accepted that it is within the cavitation bubble, or the layer of solution immediately contacting the cavitation bubble, that the sonochemical effects of molecular activation and dissociation take place.85,86,87, 88 In air saturated water the principal sonochemical dissociation processes involve the homolytic cleavage of H2O and dissolved O2.89,90,91

H 2 O H + HO Reaction 2 O2 O + O Reaction 3

59

Such cleavage products may then recombine or participate in further reactions:

HO + HO H 2 O 2 Reaction 4 O + H 2 O HO + HO Reaction 5 H + O 2 HO 2

Reaction 6 HO 2 + HO 2 H 2 O 2 + O 2 Reaction 7 Both hydroxyl radicals (HO) and hydrogen atoms (H) have been detected in ESR spin trapping experiments when water containing permanent gases in solution is subjected to ultrasound.61,89,92 The production of hydroperoxyl radical (HO2) has been presumed from the involvement of this species in specific sonochemical reactions.93,
94, 95

Furthermore, the sonochemical generation of H2O2 is well

documented, and reactions 4-7 are reported to be significant routes of H2O2 formation91,92,96 It should be noted that HO2 is itself a weak acid with a pKa = 4.8 97, which causes it to immediately dissociate at pHs that are neutral or alkali in nature to form the superoxide radical (O2-).

HO 2 H + + O 2

Reaction 8 H2O2 is also a weak acid with a pKa = 10-11.65, and dissociates to give the hydroperoxyl anion (HO2-). The HO radical has a redox potential of 2.8V 98 and is readily capable of oxidising both H2O2 and HO2-. Hence when significant

60

concentrations of H2O2 are present in the sonicated solution the following reactions become of critical importance OH + H 2 O 2 k8 O 2 + H 3 O + Reaction 9 OH + HO 2 k9 O 2 + H 2 O
-

Reaction 10 k8 and k9 are second order rate constants and have been determined to have the values of k8 = 3.7 x 107 M-1s-1 and k9 = 6.7 x 109 M-1s-1 respectively.99 Thus leads the rate of O2- production through reactions 9 and 10 is predicted to increase with pH as a result of H2O2 dissociation. There have been a number of mechanistic pathways proposed that contribute to the oxidative chemiluminescence of luminol and the inter-relationship of such pathways is far from straight forward.
77,99,100

The principal mechanism that is

proposed for sonoluminescence for the experimental conditions utilised is illustrated by Scheme 1 and is similar to that proposed for gamma-ray radiolysis induced luminol chemiluminescence.77 Luminol is known to be a weak dibasic acid with first and second pKa values of 6.3 101 and approximately 13.80 It may therefore be understood that over the experimental pH range the predominant luminol species will be the luminol monoanion (I). Step (i) in Scheme 1 shows the oxidation of the luminol monoanion (I) to produce the diazaquinone radical anion (II). Step (ii) is the reaction of the diazaquinone radical anion (II) with the superoxide radical, O2to form the hydroperoxide addition product (III). (III) is a weak acid, pKa = 10.4,99 and it is only the monoanion form of (III) which decomposes through step (iii) to give the excited state of the aminophthalate monoanion (IV).77,99 The neutral form of (III) decomposes via a dark reaction (iv) to give the starting material (I) and O2. Step (iii) is thought to proceed via a concerted mechanism involving an unstable endoperoxide intermediate99 and the aminophthalate product (IV) relaxes to the ground state with emission of light at 430nm.

-O2 + H+ (iv)

(I)

NH2

(II)
N N

NH2

(III)
N N

NH2

HO OOH N N O

OH (i)
O

O2 (ii)

pKa 10.4

-H

(IV)

NH2

OH NH2 O O O

HO OO N N

(iii)

+ N2

Scheme 1: Reaction pathways of luminol sonogenerated chemiluminescence (SCL)

62

If we consider SCL occurring in the absence of added H2O2, conditions approximated by the lowest H2O2 concentration data in Figure 3.1, O2- will be produced directly through the dissociation of HO2 produced by reaction 6. Some H2O2 will be generated internally through reactions 4 and 7 and this may be oxidised to O2- through reactions 9 and 10. In addition O2- is produced through the slow reaction of (II) with O2 through reaction 11 below. Thus sufficient O2- will be present for light to be emitted through Scheme 1. The monotonic increase in ISCL with pH seen in Figure 3.1 at pH 12 may be explained by the progressive dissociation of the hydroperoxide (III).99 The decrease in ISCL at pH > 12 is consistent with the known decrease in quantum yield of aminophthalate at high pH.99 The finding that increasing H2O2 concentration only increases ISCL at pH 10 may be explained if we assume the rate of reaction 10 is fast enough to significantly increase the steady state concentration of O2- whereas the rate of reaction 9 is not.77 This assumption would seem reasonable given the relative values of k9 and k10. Regarding the ISCL maxima observed in the pH 12 and pH 13 data shown in Figure 3.1, it may be understood that reaction 10 and step (i) of Scheme 1 are in direct competition for sonochemically generated HO. Furthermore, it has been shown that this competition leads to maximum light emission occurring when the steady state concentrations of (II) and O2- are equal.77 Thus the observed reduction in ISCL values at H2O2 concentrations > 10-4M may be ascribed to the depletion of HO through reaction 10 suppressing reaction step (i) and leading to a condition where steady state concentration of (II) < concentration of O2-. The second order rate constant for step (i) of Scheme 1 has been reported as 8.7 x 109 M-1s-1,102 i.e. similar to the value of k10 and close to the diffusion limit. This would suggest that at pH 12 maximum light emission should occur when the concentration of luminol and hydrogen peroxide are approximately equal, provided (II) and O2- are consumed at a similar rate. The finding that ISCL is maximal at luminol:H2O2 concentration ratio of ~10:1 suggests that (II) is actually consumed more rapidly than O2-. This would seem probable given the additional reactions in which (II) is known to be involved (vidi infra).

63

In addition to involvement in step (ii) of Scheme 1 (kii 10-7M-1 s-1)77 (II) has been shown to react slowly (forward rate constant kf (O2 + II) ~ 550 M-1s-1) with molecular oxygen through the formal equilibrium II + O 2 O 2 + I + diazaquinone Reaction 11 to give a neutral diazaquinone product.100 (II) is also known to undergo rapid selfrecombination and quantitative dismutation (k12 = 5 x 108 M-1s-1)99,103 to give (I) and the same neutral diazaquinone. II + II + H 2 O I + diazaquinone + HO Reaction 12 The neutral diazaquinone product my either be destroyed through hydrolysis (k (OH- + diazaquinone) 108 M-1s-1)99 or react with HO2- to produce the hydroperoxide adduct (III) (k (HO2- + diazaquinone) 108 M-1s-1)99. The small value of kf for reaction 11 implies that the formal equilibrium is never actually attained. In comparison with (II) O2- is relatively stable with respect to dismutation to O2 and H2O2 over the range of experimental pH used here.104 Furthermore, the steady state concentration of O2- may be augmented through reaction 11 and the reaction of neutral diazaquinone with HO2-. One consequence of the competition between reaction 10 and step (i) of Scheme 1 is that ISCL is primarily determined by the ratio of luminol and H2O2 concentrations. The absolute concentrations of luminol and H2O2 are immaterial provided they are greater than the steady state concentration of HO.77 Under conditions of alkaline pH and constant H2O2: luminol concentration ratio it has been shown that the integrated intensity of light emission is linearly dependent on -ray pulse radiolytic dose, and hence HO yield.77 Thus the linear dependence of ISCL on ultrasound output power shown in Figure 3.2 is consistent with rate of sonochemical HO generation being directly proportional to the ultrasound power entering the solution. It should be noted that the non-linear, and even inverse relationships, have been reported between ISCL and ultrasound at high transducer
-

64

power levels.73,74 However, such relationships have not been observed under the stated experimental conditions.

3.4.3 SCL Image Analysis Figure 3.4 and Figure 3.5 show images of SCL in the region of a plane and wedge ended 1cm-diameter cylindrical titanium sonoprobe tips, respectively. These images were each obtained by the integration of 64 video frames, captured at 25frame sec-1. For both figures (3.4 and 3.5) a primary lobe or plume of luminescence can be seen extending into solution such that the luminescent plume is co-axial ultrasound transducer horn. In addition, the wedge shaped sonoprobe tip displays secondary lobes of luminescence that are centred on the base angles of the wedge section joins the main body of the tip. The localisation of SCL activity in Figure 3.4 is similar to that reported previously in the case of a plane ended 20kHztransducer horn.76 Furthermore, the SCL images obtained using the planar sonoprobe tip were uninfluenced by the volume of the sonicated solution, or the dimensions of the sonication cell, provided that these were such that resonance and a standing wave pattern did not arise. It is difficult to extract quantitative information from the images shown in Figure 3.4 and 3.5 by eye. However the digital nature of the images permit a full analysis. Figure 3.6 and Figure 3.7 show the iso-luminance contour plots of spatially resolved ISCL data calculated from the pixel values associated with Figure 3.4 and 3.5 respectively. Iso-luminance contour lines were constructed using the inverse distance squared method of weighted interpolation and are spaced at intervals corresponding to 10% of maximum light intensity. It may be seen from Figure 3.6 that in the case of the plane tip, maximum ISCL values are located at, or very near, the transducer-solution interface and decay rapidly with distance from that interface. However, Figure 3.6 shows that, in the case of the wedge tip, an area of maximum ISCL values occur approximately 0.25mm from the tip vertex and that this area is elongated coaxially with the principal axis of the transducer horn. Images similar to, of SCL emission at the plane ended sonoprobe tip were obtained over a range of transducer power levels. These images were analysed with the intention of determining the extent to which spatial distribution of SCL activity was influenced by ultrasound intensity. In all cases, ISCL was found to decay exponentially with perpendicular distance (d) from the transducer surface.

65

Figure 3.4: Image of luminol SCL activity, proximal to the standard, flat; 10mm diameter titanium sonoprobe tip. (10-3M luminol, 104M H2O2, 0.02M EDTA, Temperature 50C power output - 30W).

66

Figure 3.5: Image of luminol SCL activity, proximal to the wedge shaped, 10mm diameter, titanium sonoprobe tip. (10-3M luminol, 104M H2O2, 0.02M EDTA, Temperature 50C power output - 30W).

67

B-B

A-A

Figure 3.6: Iso-luminance contour plot of SCL activity proximal to the plane ended, 10mm diameter, titanium sonoprobe tip. (Contours spaced at 10 percent light intensity intervals).

68

Figure 3.7: Iso-luminance contour plot of sono-chemiluminescent activity for the wedge-ended 10mm diameter, titanium sonoprobe tip. (Contours spaced at 10 percent light intensity intervals).

69

However the half-length of this exponential decay was found to decrease significantly with increasing transducer output power. Figure 3.8 shows a semilogarithmic plot of normalised ISCL values a line coaxial with the principal axis of the ultrasound horn, as indicated by the line A-A in Figure 3.6. The data in Figure 3.8 correspond to at least three half-lengths of ISCL-d decay. The solid lines shown in Figure 3.8 were constructed by least squares linear regression and exhibit gradients of 4.3 0.1 cm-1 and 10.3 0.4cm-1 for transducer output powers of 5W and 70 W respectively. Both data sets exhibit a linear correlation coefficient > 0.98. By contrast the radial distribution of ISCL in the luminescent plume generated by the plane ended sonoprobe tip was substantially independent of ultrasound output power. Figure 3.8 shows the distribution of normalised ISCL (image pixel values) along a line normal to principal axis ox the ultrasound transducer horn, as indicated by the line B-B in Figure 3.6 for ultrasound powers 5W and 70W. It may be seen from Figure 3.9 that neither the shape nor the absolute widths of the radial ISCL distribution are changed significantly by the change in ultrasound output power. Furthermore, the diameter of the sonoluminescent plume did not change rapidly with perpendicular distance, d, from the transducer surface, i.e. the plume was not strongly divergent. In no case did the width-at-half-maximum value of the radially resolved ISCL distribution increase by more than 10% in going from d = 2.5mm to d = 5mm. In the previous section it has been proposed that ISCL is proportional to the rate of sonochemical HO generation. However, the extent to which Figure 3.4 to Figure 3.7 can be interpreted, quantitatively, as maps of cavitational intensity is unclear. One unknown quantity is the extent to which light scattering from cavitation bubbles
105

contributes to the SCL images. It has been previously stated,

in the case of luminol SCL stimulated by a titanium tipped 20kHz-ultrasound horn, that Luminescence is located on the surface of the of the titanium horn.76 For this statement to be literally true the appearance of a the luminescent plumes observed in Figure 3.4 and Figure 3.6 could be explained by the surface of the titanium tip acting like a plane mirror and reflecting a beam of light out into solution. Scattering of light would then cause the beam to be visible. That this is not the case is made evident by Figure 3.5 and Figure 3.7, where the direction of the main plume of luminescence remains co-axial with the long axis of the transducer horn (possibly

Ln (normalised ISCL)

-1

-2

-3

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

distance from surface, cm

Figure 3.8: Normalised ISCL as a function of perpendicular distance d from the plane ended sonoprobe tip i.e. along the line A-A shown in Figure 3.6 at various ultrasound power values. (10-3M luminol, 104M H2O2, Temperature 50C, and ultrasound power: q 5W, 70W.)

1.4 1.2

normalised ISCL

1.0 0.8 0.6 0.4 0.2 0.0 -0.2 -0.5 0.0 0.5

distance from centre axis, cm

Figure 3.9: Normalised radial ISCL distribution proximal to the plane ended sonoprobe tip i.e. along the line B-B shown in Figure 3.6 at various ultrasound power values. (10-3M luminol, 104M H2O2, Temperature 50C, and ultrasound power: q 5W, 70W.)

73

due to the constructive interference of waves propagating from the tip surfaces) even though the plane surfaces of the wedge ended sonoprobe tip lie at approximately 45 to that axis. Thus, we can conclude that, whilst there is probably a contribution from scattering, the light intensity distribution shown in Figure 3.4 to Figure 3.7 derive principally from the spatial distribution of SCL activity. The next difficulty in interpreting images shown in Figure 3.4 (and its ilk) is the fact that they representations, in two dimensions, of a three dimensional phenomenon. The light contributing to a single point on the image will therefore derive from a volume of luminescent solution. If the object distance is sufficiently great that the light rays entering the camera lens are effectively parallel, and the absorption and scattering of light in solution is negligible, the observed light intensity will be linearly dependant on the optical path length (L) through the luminescent volume along the axis of observation. Given that the spatially resolved ISCL data in Figure 3.4 and Figure 3.5 derive from an approximately cylindrical plume of SCL activity the radially resolved ISCL profiles shown in Figure 3.9 are expected to derive substantially from radial variation of L. Conversely, As the diameter of the luminescent plume changes very little over the first few millimetres from the transducer surface, i.e. L is approximately independent of d, the axially resolved ISCL data shown in Figure 3.8 are expected to correspond closely to the axial distribution of sonochemical activity. Thus, the exponential decay in ISCL shown in Figure 3.8 reflects an exponential fall off in the rate of sonochemical HO generation with perpendicular distance, d, from the transducer surface.

3.4.4 Acoustic Attenuation in Cavitating Water. Given the preceding arguments, the axially resolved ISCL d data shown in Figure 3.8 may be used to characterise the propagation of ultrasound travelling waves in the volume of cavitating solution proximal to the transducer surface. In order to facilitate the necessary analysis it is assumed that the axially resolved ISCL values are proportional to the local acoustic intensity (I, Wcm-2). The assumption that, microscopically, ISCL I is not unreasonable given the linear macroscopic relationship between spatially unresolved ISCL values and transducer output power shown in Figure 3.2. Obviously, this relationship will only apply, macroscopically or microscopically, above the cavitation threshold.

74

When a plane acoustic wave propagates through a homogenous medium the intensity of the wave decreases with the distance from the radiation source due to the absorption of acoustic energy and its conversion into heat. Absorption results from: viscous effects, thermal conduction and chemical relaxation processes occurring within the medium.106,107 The acoustic intensity, I, at some distance, d, from a source of intensity I0 is given by: I = I 0 exp (-2d) Equation 24 Where is the acoustic absorption coefficient.106,107 For acoustic intensities below the cavitation threshold below the cavitation threshold the value of depends predictably on the mechanical and thermodynamic properties of the medium and increases with the square of the acoustic frequency, f,106,107 For water the quantity /f2 has a measured value of 21 x 10-17 cm-1 s2 over a wide range of frequencies,108 implying = 8.6 x 10-8 cm-1 at 20 kHz. However, bubbles, such as those produced through cavitation, are known to be effective absorbers and scatterers of acoustic energy109,110,111,112,113. Sound absorption occurs through the damping bubble oscillations by: viscous, re-radiative and thermal conduction mechanisms,109 and the absorption cross-section of a bubble near its resonant frequency may be 1000 times its geometrical cross section
110

. For these reasons the value of is predicted to

increase significantly in the presence of cavitation104,113,114 and will depend on the number concentration and size distribution of cavitation bubbles. However, under these conditions the ultrasound wave will be subject to multiple scattering from cavitation bubbles and any experimental value of would be more properly regarded as an attenuation coefficient containing both absorption and scattering contributions. Equation 24 is immediately consistent with the exponential form of the axially resolved ISCL d data shown in Figure 3.8, given the assumption that I ISCL as argued above. This being the case, the gradients of the lines shown in Figure 3.8 correspond to the values of 4.3 0.1 cm-1 and 10.3 0.4 cm-1 at I0 values of 6.4 Wcm-2 and 89 Wcm-2 respectively. The finding that, in the presence of acoustic cavitation, values in water may increase by > 8 orders of magnitude at the

75

cavitation producing frequency implies that, when present, cavitation is the predominant mechanism of acoustic energy absorption. It also helps to explain the shielding effect,113 significantly reducing acoustic intensities elsewhere in a sonicated aqueous solution. The observed increase in with I0 suggests that either the number concentration of cavitation bubbles increases with acoustic intensity, or there is an increase in their individual adsorption cross-sections, or both. This finding also tends to support the notion that an enhancement of acoustic absorption through cavitation may contribute to the phenomenon of decoupling34 whereby the efficiency of energy transfer from the ultrasound transducer a liquid medium decreases progressively with increasing I0 at high I0 values.

3.4.5 Conclusions The sonogenerated chemiluminescence (SCL) of aqueous luminol is strongly influenced by pH and by the concentration of H2O2. In the presence of 10-4 M H2O2 the intensity of SCL is linearly proportional to ultrasound transducer output power. EDTA (10-4M) reduces the background (silent) chemiluminescence of luminol/ H2O2 solutions by >95% whilst minimally affecting the intensity of SCL. These findings are consistent with SCL light emissions following the decomposition of a hydroperoxide adduct formed through the reaction of luminol mono-anion with sonogenerated HO and O2-. Spatially resolved light intensity information derived from digitally captured SCL video images may be analysed to provide quantitative data on the spatial distribution of sonochemical activity in solution, provided variations in optical pathlength are taken into account. SCL intensity (ISCL) decays exponentially with perpendicular distance (d) from a planar ultrasound transducersolution interface and that the decay half-length decreases with increasing transducer output power. Acoustic attenuation coefficients () in cavitating solution may be estimated non-invasively using ISCL-d data by assuming a linear macroscopic relationship between ISCL and transducer input power. The values thus obtained increase with transducer power and may be >108 times greater than values for homogenous water

76

Chapter 4 Determination of the Effect of Ultrasound Intensity and Proximity on Wire Cleaning Kinetics

77

4.1 Introduction
Previous studies115,116 have shown that oxide heat scale may be removed from the surface of steel wires by a combination of electrolysis and ultrasonication in neutral (or near neutral) electrolyte. It is thought that electrolysis serves to disrupt the bonds, which hold the scale to the metal surface. Ultrasound then acts to break up and remove the loosened scale.115,116 No synergistic interaction has been found between the electrolytic current and ultrasound. Best results are obtained when these two methodologies are applied separately, with electrolysis preceding ultrasonication. The work to be described was aimed at investigating the influence of ultrasound transducer power and transducer-wire distance in determining the rate of removal of an electrolytically loosened heat scale. A further aim was to relate the rate of scale removed to the intensity of cavitation proximal to the scaled surface. It is generally assumed that it is cavitation and the impingement of microjets generated by the collapsing cavitation bubbles, which are responsible for the surface cleaning effects of ultrasound. (See section 1.2.8). However, the difficulties in quantifying cavitational activity in a spacial resolved manner make this hypothesis hard to test. Here the relationships demonstrated in the preceding chapter between ultrasound transducer power, distance and cavitational driven sonochemical activity are correlated with scale removal rates.

4.2 Experimental Details

4.2.1 Samples All the work was carried out on a wire used in piston ring manufacture (described in section 2.5.1). The wire, consisting of a high carbon, silicon manganese steel, was covered with a heat scale layer of approximately 15 m. Preparation of the steel wire samples followed the protocol outlined in section 2.5.2 to obtain a current density of 1 Acm-2 during all electrolytic treatments.

4.2.2 Method The experimental set-up for the kinetics of wire descaling experiments is shown in Figure 2.7. All experiments featured a constant, 95% anodic duty cycle of 1 Hz frequency with square wave characteristics. All electrolytic baths consisted of

78

a 10% w/w aqueous solution of sodium chloride adjusted to pH 7 and thermostated at 50 C by a stainless steel coil connected to a waterbath. Ultrasound was applied at a variety of intensities and wire to probe distances as outlined below.

4.2.3 Ultrasonic Configuration Variation of Intensity and Probe Distance Tip to Wire Distance The distance of the wire samples from the tip of the ultrasound probe was varied between 5-25mm in 5mm steps by the use of an adjustable inert plastic support stand. After the initial electrolytic pre-treatment the wire was then exposed to a 10-second burst of ultrasound from the probe (this was the shortest time that produced a stable ultrasound field). After visually inspecting the surface of the wire to determine the progress of the cleaning, the cycle of ultrasound/inspection was repeated until either the sample was 100% scale free or a consistent percentage of scale remained. The experiments were repeated for the various probe to wire distances at calibrated ultrasound intensities 1, 2, 4 and 6. (See section 2.3.4)

4.2.4 Measurement of Surface Cleaning. The prepared wire sample and graphite counter electrode were immersed in the 10% sodium sulphate solution and connected to the galvanostat in preparation for electrolysis (See Figure 2.7.) For all the wire samples used the wire was prepared in such a way that the surface current density was equivalent to 1Acm-2 (see Figure 2.4.) The subsequent cleaning was achieved by an initial two minutes electrolytic treatment of the wire followed by 10-second bursts of ultrasound as described in previous section. Experiments were repeated three times in total for each of the different conditions employed from which an average cleaning rate was determined. Cleaning progress was monitored by withdrawal of the wire sample from the sodium sulphate solution after each 10-second exposure to the ultrasound. An assessment of the remaining heat scale soiling was determined by the viewing of the upper surface (hemi-cylinder see Figure 4.4) of the wire through a millimetre grid. Estimated results are presented as percentage of remaining soil verses total sonication times.

79

(a) 5.2W

100

Percentage scale remaining

80

60

40

20

0 0 20 40 60 80 100 120 140

Cleaning Time (seconds)

(b) 14.68W

100

Percentage scale remaining

80

60

40

20

0 0 20 40 60 80 100 120 140

Cleaning time (seconds)

80

(c) 42.65W

100

Percentage scale remaining

80

60

40

20

0 0 20 40 60 80 100 120 140

Cleaning time (seconds)

(d) 70.8W

100

Percentage scale remaining

80

60

40

20

0 0 20 40 60 80 100 120 140

Cleaning Time (seconds)

Figure 4.1: The variation in the percentage of scale removed with distance from the ultrasound probe tip at various power outputs of (a) 5.2W, (b) 14.68W, (c) 42.65W, (d) 70.8W (s-5mm, v-10mm, -15mm,x-20mm,q-25mm.)

81

NB These DO NOT reflect total experimentation time, ONLY the accumulative of the times that the wire was exposed to ultrasound

4.3 Results and Discussion

4.3.1 Influence of Ultrasound Power and Transducer Surface Distance. Figure 4.1 shows the fraction of the exposed metal surface which remains covered with scale as a function of time for various wire-probe distances at probe ultrasound powers of a) 5.2W b) 14.6W c) 45W and d) 70.8W. It maybe seen from Figure 4.1 that for each probe output power the rate of scale removal increases markedly with decreasing probe-wire distances. It may also be seen from Figure 4.1 that at any given probe-wire distance the rate of scale removal increases with increasing probe output power. It should be noted that the minimum experimental periods of sonication was 10 seconds and at higher probe output powers and lower probe-wire distances the surface is completely de-scaled after 10 seconds of sonication. This means descaling times indicated in Figure 4.1 under these conditions must be regarded as minimum estimates of the true descaling time Figure 4.2 shows the fraction of the exposed wire surface, which has been de-scaled after 120 seconds of ultrasonication as a function of transducer output power at various transducer-wire distances. The data in Figure 4.2 is replotted from Figure 4.1 and represents a crude estimation of scale removal rates over the 120 seconds experimentation period. Figure 4.2 serves to illustrate in a more condensed fashion how the rate of scale removal increases with increasing transducer output power and increases with decreasing transducer-wire distance. Again with the caveat that the fastest rates of scale removal are underestimated due to the experimental procedure used. A more refined estimate of scale removal rate was made by constructing tangents to the scale area verses time plots shown in Figure 4.1 at time zero. It may be appreciated that the gradient of such a tangent represents an estimate of the initial rate of scale removal (Ri). Figure 4.3 shows a semi-logarithmic plot of Ri as a function of transducerwire distance for various values of transducer output power. It may be seen from Figure 4.3 that for Ri values < 10 the plots of lnRi verses transducer wire distance are approximately linear. The slopes of the linear portions of the lnRi-distance plots were all approximately equivalent at 3.5cm-1 0.1 cm-1, which correlates quite readily with the calculated value of 4.3 0.1 cm-1 at lower ultrasound intensities.

1.0

Fractional Area Coverage (Ac/A)

0.8

0.6

0.4

0.2

0.0 0 10 20 30 40 50 60 70

Power (Watts)

Figure 4.2: Graph illustrating the changes in the area of the wire surface covered with scale as a fraction of the total surface area (Ac/A) with applied ultrasound density for distances of q5mm; v10mm; w15mm; x 20mm; + 25mm; from the surface of the sonoprobe.

lnRi

-2

-4

-6 0.0 0.5 1.0 1.5 2.0 2.5

Distance from probe tip(cm)

Figure 4.3: Graph to show the changes in initial cleaning rate (ln Ri) with distance from the sonoprobe tip for various ultrasound power densities: s= 5.2W. q= 14.68W; v= 42.65W;w= 70.8W.

84

Figure 4.3 shows clearly that the maximum measured value of Ri is 10 and all the experimental plots converge to this limit as Ri increases. Once again the limiting value of Ri almost certainly arises from the minimum experimental period of 10 seconds and the error that this introduces when de-scaling is rapid. The finding that (non-limiting) Ri values vary with exponentially with transducer-wire distance is directly consistent with the exponential relationship between cavitationally driven sonochemical activity and distance from the transducer surface shown in the previous chapter. Furthermore, the slope of the ln Ri-distance plots are of a similar magnitude to the cavitation field acoustic coefficients () values of 4.3 0.1 cm-1 and 10.3 0.4 cm-1 determined by the analysis of luminol SCL image data. These findings tend to support the hypothesis that it is cavitation, and its related properties, which are responsible for the ultrasonic removal of electrolytically loosened oxide scale from the wire surface. It further implies an approximately linear dependence of de-scaling rate on ultrasound intensity.

4.3.2 Ultrasound Shadowing. It remains to explain the observation that only the half of the wire surface proximal to the ultrasound transducer becomes significantly de-scaled over the 120 seconds experimental period (see Figure 4.4). It may be appreciated from the preceding section that the observed exponential decrease in scale removal rate with transducer-surface distance might be considered as a contributing factor. That is to say the distal side of the wire is more distant than the ultrasound transducer and is therefore more slowly de-scaled. However, the experimental ln RI-distance scores are insufficient to explain the marked dissimilarity in observed descaling rates An alternative explanation is that the wire-solution interface scatters ultrasonic energy such that ultrasound intensity is significantly attenuated in solution lying behind the wire. That is to say that the wire casts an ultrasound shadow thus reducing the intensity of the cavitational activity near the distal portion of the wire surface. To test this hypothesis a luminol SCL imaging experiment was carried out, as described in the preceding chapter. In this experiment a prepared wire sample (See section 2.5.2) was placed 5mm from the transducer tip and sonicated. The image was captured and analysed as before.

85

A)
Ultrasound Probe

Wire

Oxide

Ultrasound Probe

Wire

Hemicylinder of cleaned wire

B) Figure 4.4: The hemi-cylindrical cleaning of the oxide scale of wire: A) diagrammatic illustration and B) the appearance of the wire surface after combined electrolytic-ultrasonic de-scaling.

86

Figure 4.5: Image of luminol SCL activity proximal to the plane ended, 10mm diameter, titanium sonoprobe tip during wire de-scaling. (10-3M luminol, 104M H2O2, 0.02M EDTA, Temperature 50C power output 30W.)

87

Figure 4.6: False colour iso-luminance contour plot of SCL activity proximal to the plane ended, 10mm diameter, titanium sonoprobe tip during wire cleaning.

88

Both Figure 4.5 and Figure 4.6 clearly show a reduction in luminol SCL in the region of solution immediately behind the wire. The SCL intensity at points A and B in Figure 4.6 was 26 and 12 pixel values respectively. Assuming the linear relationship between ultrasound intensity and SCL light intensity argued in the preceding chapter these values suggest that the ultrasound intensity immediately behind the wire is less than half the ultrasound intensity found immediately in front of the wire. This difference offers a possible explanation for the observed difference in the cleaning rate between the two sides of the wires surface.

89

Chapter 5

Hydrogen Evolution at the Titanium Sonotrode.

90

5.1 Introduction
The cleaning of metal surfaces electrolytically involves, by definition, some type of electrolytic cell. This usually consists of two electrodes, (I) the working electrode, which for metal cleaning purposes is usually the item to be cleaned and (II) the counter electrode, which is traditionally an inert substance that allows the passage of current (e.g. graphite). Both of these electrodes are immersed in an ionic conductor like an aqueous salt solution, which completes the electronic circuitry and allows the passage of electric current between the electrodes. Effects caused by Power Ultrasound (15-40 kHz) cavitation on electrode processes is known to fall into five distinct modes:

(i)

Mass transport enhancement (see section 1.6.2) caused by increased turbulence and microstreaming 67,85,117,118 , Continuous electrode surface activation 119, Radical, ion or other high energy intermediate formation 117, Product desorption 120 and The increase in heterogeneous electron transfer within the chemical processes 121.

(ii) (iii) (iv) (v)

Traditionally ultrasonic baths and probes have been utilised to indirectly stimulate electrodes 122. More recent research carried out by Compton, Ecklund et al
123

as well as Reisse, Francois and co-workers 124 has, however, indicated that a far

greater rate enhancement can be achieved by using the tip of an ultrasound transducer horn as an electrode. Direct stimulation of these so-called Sonotrodes causes the observed increase in observed reaction rate. For this study experiments were carried out to investigate the beneficial effects that the direct application of ultrasound would have, if any on hydrogen evolution (Reaction One) and oxygen reduction (Reaction Two) in aqueous electrolyte. This was achieved by substituting the working electrode for a sonotrode in this case the tip of a titanium ultrasound horn.

91

2H + (aq) + 2e - H 2 (g) (1)

O 2 (aq) + 2H 2 O + 4e - 4HO - (aq) (2) Both of these reactions play an important part in both cathodic acid pickling, AC based electrolytic de-scaling and metallic corrosion in aqueous systems. (The presence of cavitation phenomena causing increased corrosion rates being of notable importance
126,127 125

.) Each of the reactions outlined above involve complex,

multistep mechanisms that incorporate the presence of chemisorbed intermediates and whose reaction could be affected by any over the previously stated modes

of action (i) (v). The present chapter aims to determine to what extent intense ultrasound cavitation influences such processes.

5.2 Experimental Details

5.2.1 Methodology (i) Aerated

The experimental set-up employed is illustrated in Figure 2.8. The electrolytic cell bath consisted of a 10% w/w solution of sodium sulphate (Na2SO4) adjusted to pH 7 with 0.1M NaOH. The power level used throughout all experiments was 26Wcm-2, which was determined by calorimetry
68

(see section
123

2.3.1). Temperature control was provided by a Grants Y14 waterbath, which circulated thermostated water through a stainless steel cooling coil and

maintained electrolyte temperature constant (to within +/- 1C). A Solatron 1280 potentiostat controlled by computer was used to carry out all voltametric measurements at a temperature of 30C. The collection of data for the linear sweep voltammograms involved a sweeping rate of 0.014Vs-1 of the sonotrode potential between 2.5V to 0V verses SCE (Standard Calomel Electrode). To ensure that the resultant voltammogram was reproducible the potential of the sonotrode was cycled from 2.5 to 0V at a rate of 0.014V-1 without sonication, prior to data collection, until a stable pattern was observed (typically < three cycles).

92

(ii)

Deaerated

For experiments that involved deaerated conditions, the electrolyte was purged for 30 minutes with pre-purified argon (supplied by BOC) prior to measurements being carried out.

5.3 Results and Discussion

The results of the investigations are displayed as Tafel plots. Figure 5.1 shows the voltametric responses under deaerated conditions and Figure 5.2 the effect of aerated electrolyte. Evolution of hydrogen was clearly observable at potentials <-1.6V. Both Figure 5.1 and Figure 5.2 show that the distinctive appearance of the Tafel plot is more or less unchanged for sonicated and silent conditions, the only difference being that the values of current density increase in the presence of ultrasound, for all potentials. The reproducibility, reversibility and magnitude of the observed sonoelectrochemical effect are emphasised by Figure 5.3 and Figure 5.4. For deaerated and aerated conditions respectively these figures (5.3 and 5.4) demonstrate how the sonotrodes current responses are time dependent/transient when short pulses of sonication are imposed on it at constant potential. The result of such bursts of ultrasound in deaerated electrolyte at a potential of 1.3V can be found in Figure 5.3 whereas Figure 5.4 displays the response of the sonotrode in aerated electrolyte at a constant potential of 1.25V. For both conditions, the changes in current density caused by application of ultrasound are seen to be instantaneous and reversible. Such results mean that bulk heating effects that can be caused by ultrasound are not responsible for producing the rapid increases in current density. Titanium sonotrodes rapidly form a TiO2/TiO3 layer (which acts like an n-Type semiconductor) when exposed to air. This layer has been found to limit electrochemical reactions especially anodic- at titanium sonotrodes
123

but the

results in Figure 5.1 to Figure 5.4 show little evidence of this. The figures demonstrate that the titanium sonotrode is more than acceptable cathode for reaction (1) and (2) at the stated potentials and that sonication leads to a substantial enhancement of reaction rate.

Potential (V)

-1

-2

-3

-6

-5

-4

-3

-2

-1

Log I (A cm-2)
Figure 5.1 Tafel plot of titanium sonotrode voltametric response under silent (---) and sonicated (___) conditions in deaerated neutral 10% wt Na2SO4 solution

Potential (V)

-1

-2

-3

-6

-5

-4

-3

-2

-1

Log I (A cm-2)
Figure 5.2: Tafel plot of titanium sonotrode voltametric response under silent (---) and sonicated (___) conditions in aerated neutral 10% wt Na2SO4 solution

0.0

-0.5
Current / mA cm-2)

-1.0

-1.5

-2.0

-2.5 150 200 250

Time (s)
Figure 5.3: Transient (time dependent) current response of the titanium sonotrode in neutral 0.7M aqueous sodium sulphate at 30C (Deaerated, 5 second ultrasound pulses.

0.0

-1.0 Current / mA cm-2

-2.0

-3.0

-4.0 50 100 150

Time (s)
Figure 5.4: Transient (time dependent) current response of the titanium sonotrode in neutral 0.7M aqueous sodium sulphate at 30C (Aerated, 1 second ultrasound pulses

97

For the reaction under deaerated conditions (Figure 5.1) it can be assumed that reaction (1) is entirely responsible for the observed cathodic currents. The Tafel plot under silent conditions is more or less linear between 1.25 and 1.6V with a Tafel slope of 190mV decade-1. This corresponds to reaction (1) being activation controlled. When the Tafel plot enters the 1.7V region it curves away suggesting that reaction becomes current limited. On the application of ultrasound the Tafel plot becomes anodically shifted by a value in the region of 500mV and there is also an observed change in the linear regions Tafel slope, which increases by about 80mV decade-1 to 270mV decade-1. A change in Tafel slope indicates that reaction (1) has undergone a change in mechanism or rate determining step caused by the ultrasound. The enhancement of current density at constant potential by sonication has the greatest magnitude within the linear, activation controlled region of the Tafel plot and show an increase approximate to a factor of five. A less dramatic effect is seen in the current limited portion of the plot where there is a factor of two increase observed. As the greatest effect of ultrasound is seen in the activationcontrolled region of reaction (1) it would be sensible to conclude that ultrasound interacts through modes (ii) - (v) with the electrochemical process. In the case of aerated conditions, the silent Tafel plot (Figure 5.2) possesses an unmistakable current plateau that spans the 0.9 to 1.4V region. This is caused by the O2 in reaction (2) on the sonotrode surface becoming mass transport limited
128

. Therefore it can be assumed that at potentials cathodic of 1.4V, reaction (2)

predominates whereas at potentials anodic of 1.4V a combination of reaction (1) and (2) leads to the cathodic currents seen. The increases in current density at constant potential caused by sonication are most profound (greater that one order of magnitude) in the region anodic of 1.4V. The presence of the diffusion limited current plateau also becomes obscured in the sonicated Tafel plot. Between 0.9 and -1.4V the current density must be due to mode (i) and suggests that the Nernst diffusion layer (see section 1.6.3) has been reduced tenfold by the sonication. At potentials anodic of 0.9V where reaction (2) is not limited by mass transport, modes (ii) (v) may play a role but mode (i) is still likely to make significant additions to the current increase caused by the sonication.

98

Appendices

99

Paper: Hydrogen evolution and oxygen reduction at a titanium sonotrode Chem. Commun. (1998)

References

101

R. Muncaster, A-Level Physics Second Edition, Stanley Thorn Publishers Ltd.

(1986).
2 3 4 5

J. P. Joule, Philos.Mag, Vol. (III) 1847, 30, pp76. J. & P. Curie, Compt. Red. Acad. Sci. Paris (1880), 91, pp294. J. & P. Curie, Compt. Red. Acad. Sci. Paris (1881), 91, pp1137. T. J. Mason, J.P. Lorimer, Sonochemistry: Theory, Applications and Uses of

Ultrasound in Chemistry, Ellis Horwood Publishers Ltd. Chapter One.

6 7 8 9

Microsoft Encarta 98 Encyclopaedia, British Edition. J. Thornycroft, S.W.Barnaby, Instit. Civil Eng., 1895, 122, pp51. Branson Ultrasonic Technical Literature, Ultrasonic Plastic Welding. T.G. Leighton, M.J.W. Pickworth, A.J. Walton, British Journal of Radiology,

Vol. 60, (714), 1987, pp614.

10

D. Esminger, Ultrasonics: Fundamentals, Technology, Applications, (Second

Edition) Marcel Dekker Inc, (1988) pp438-450

11

K. S. Suslick, Ultrasound (Its Chemical, Physical and Biological Effects), UCH

Publishers, New York (1995/1996), pp213-219.

12 13 14 15

A. Kuhn, The Process of Steel Pickling, Steel Tech. (1988) Chap. 1. R. A. Geckle, Metal Finishing Guidebook, pp144 (1980). N. S. Holt, R. Walker, Product Finishing, 33, 7, pp12 (1980). K.R. Trethewey, J. Chamberlain, Corrosion for Students of Science and

Engineering, Langman Scientific and Technical, New York, (1988), Chapter 17.
16

K. Sachs, C.W. Tuck, Surface Oxidation of Steel in Industrial Furnaces,

Reheating for Hot Working, Iron and Steel Institute, London, (1968), pp2.
17 18

H. Engell, Arch. Eisenht., 28 (1957), pp109-115. F.A. Cotton, G. Wilkinson, Advanced Inorganic Chemistry, 5thEd., Wiley &

Sons, New York, (1988), pp709-724.

19 20 21

F. Koch, J.B. Cohen, Acta Crysatllography, B25, (1969) pp275. A.K. Cheetum, B.E.G. Fender, R.I. Taylor, J. Phys C., V, (1971) pp2160. Per Kofstad, High Temperature Corrosion, Elsevier Applied Science Publishers

Ltd., London, (1988) Chapter 2.

22

P.D. Edmonds (Ed.), Methods of Experimental Physics Ultrasonics, (Vol.

19), pp20-27

102

23

C.D. Hogman (Ed.), Handbook of Chemistry and Physics, 28th Edition, Chemical

Rubber Publishing Co. (1944).

24

E.J. Pugh, Electrolytic and Ultrasonic Cleaning of Steel Wire in Neutral Salt

Solutions, M.Phil. University of Wales, Swansea (1996), Chapter One.

25

T. J. Mason, J.P. Lorimer, Sonochemistry: Theory, Applications and Uses of

Ultrasound in Chemistry, Ellis Horwood Publishers Ltd, pp61.

26

S.F. Lancelot, Electrolytic and Ultrasonic Cleaning of Steel Wire in Neutral

Sodium Tripolyphosphate Solution, M.Phil. University of Wales, Swansea (1997), Chapter One
27 28

E. Klein, J. Acoust .Soc. Amer, 20 (1948), pp601. F.G. Blake, The Tensile Strength of Liquids: A Review of Literature, Tech.

Memo. No. 9, Harvard Research Laboratory, (1949).

29 30

M. Kornfeld, L. Suvorov, J. Appl. Phys., 15 (1944), pp495. T.J. Mason, Chemistry with Ultrasound, Elsevier Science Publisher Ltd., Essex

UK, (1990).
31

D. Esminger, Ultrasonics: Fundamentals, Technology, Applications, (Second

Edition) Marcel Dekker Inc, (1988) pp66.

32

P.D. Edmonds (Ed.), Methods of Experimental Physics Ultrasonics, (Vol.

19), pp385.
33 34

G. Kurtze, Nachr. Akad. Wiss Goettingen, AII (1958) T.J. Mason, Practical Sonochemistry (Users Guide to Application in Chemistry

and Chemical Engineering), Ellis Horwood Limited, London, (1991) Chapter 1.

35

T.J. Mason, Sonochemistry (The Uses of Ultrasound in Chemistry), The Royal

Society of Chemistry, (1990) Chapter 1.

36

K.R. Trethewey, J. Chamberlain, Corrosion for Students of Science and

Engineering, Langman Scientific and Technical, New York, (1988), Chapter Four.
37

C.M.A. Brett, M. Oliveira-Brett, Electrochemistry Principles, Methods and Corrosion, 3rd Ed., Butterworth and Heinemann, Oxford, 1 (1994), Chapter One J Jones, Honours II Electrochemistry and Corrosion Courses University of

Applications, Oxford University Press, (1994), pp14-16.

38 39

Wales, Chemistry Department 1997.

40

K.R. Trethewey, J. Chamberlain, Corrosion for Students of Science and

Engineering, Langman Scientific and Technical, New York, (1988), Chapter Seven.

103

41

A.J. Bard, L.R. Faulkner, Electrochemical Methods, Fundamentals and

Applications, John Wiley and Sons Inc. Chichester, (1980), Chapter One
42 43 44

P.D. Lickiss, V.E. McGrath, Chem. Britain, 1996, 32, pp47. R.G. Compton, R.A.W. Dryfe, Prog. Reaction Kinetics, 1995, 20, pp245. G.J. Price (Ed.), Current Trends in Sonochemistry, The Royal Society of

Chemistry, London 1992.

45 46 47

T.J.Mason, J.P.Lorimer, D.J.Walton, Ultrasonics, 1990, 28, pp333. C. Degrand, J. Chem.Soc., Chem. Commun., 1986, pp1113. O.J. Murphy, S. Srinivasan, B.E. Conway (Eds.), Electrochemistry in

Transition: From the 20th to the 21st Century, Plenum, New York (1992), pp397.
48

A. Durant, J-L. Delaplancke, R. Winand, J. Reisse, Tetrahedron Lett., 1995,

36, pp4257.
49 50 51 52 53

A.J. Bard, Anal. Chem., 1963, 35, pp1125. H.D. Dewald, B.A. Peterson, Anal. Chem., 1990, 62, pp779. http:// www.phys.ox.ac.uk/:8000 R.G. Compton, F-M. Matysik, Electroanalysis, 1996, 8, pp218. F. Marken, T.O. Rebbit, J. Booth, R.G. Compton, Electroanalysis via the

internet (@ http:// www.phys.ox.ac.uk/:8000)

54

A.M.O. Brett, Fifth Meeting of the European Society of Sonochemistry, 7-11

July 1996, Cambridge, UK.

55 56

T.G. Leighton, The Acoustic Bubble, Academic Press, London, 1994. K.S. Suslick, J.W. Goodale, P.F. Schubert, H.H. Wang, J. Am. Chem. Soc.,

1983, 105, pp5781.

57

K.S. Suslick, D.A. Hammerton, R.E. Cline Jr., J. Am. Chem. Soc., 1986, 108,

pp5641.
58 59 60 61 62 63 64

A. Kotronarou, G. Mills, M.R. Hoffmann, J. Phys. Chem., 1991, 95, pp3630. T.P. Caulier, M. Maeck, J. Reisse, J. Org. Chem., 1995, 60, pp272. V. Misik, L.J. Kirschenbaum, P. Riesz, J. Phys. Chem., 1995, 99, pp5970. K. Makino, M.M. Mossaba, P. Riesz, J. Phys. Chem., 1983, 87, pp1369. W.J Tomlinson, Adv. Sonochem., 1990, 1, pp173. A.Henglein, D. Herburger, M. Gutierrez, J.Phys. Chem., 1992, 96, pp1126. T. Ando, P. Bauchat, A. Foucaud, M. Fujita, T. Kimura, H. Sohmiya,

Tetrahedron Lett., 1991, 32, pp6379.

104

65

G.

Prentice,

Electrochemical

Engineering Principles, Prentice-Hall

International (1991) pp151

66

R.G. Compton, J.C. Ecklund, S.D. Page, T.J. Mason, D.J. Walton, J. Appl.

Electrochem., 1996, 26, pp775.

67

D.J. Walton, S.S. Phull, A. Chyla, J.P.Lorimer, T.J. Mason, L.D. Burke, M.

Murphy, R.G. Compton, J.C. Ecklund, S.D. Page, J. Appl. Electrochem., 1995, 25, pp1083.
68

T.J. Mason, Practical Sonochemistry (Users Guide to Application in Chemistry

and Chemical Engineering), Ellis Horwood Limited, London, (1991) pp45.

69

E.J. Pugh, Electrolytic and Ultrasonic Cleaning of Steel Wire in Neutral Salt

Solutions, M.Phil. University of Wales, Swansea (1996), Chapter Two.

70

S.F. Lancelot, Electrolytic and Ultrasonic Cleaning of Steel Wire in Neutral

Sodium Tripolyphosphate Solution, M.Phil. University of Wales, Swansea (1997), Chapter Two
71

C.M.A. Brett, M. Oliveira-Brett, Electrochemistry Principles, Methods and

Applications, Oxford University Press, (1994), Chapter Nine.

72 73 74

H.N. McMurray, D.A. Worsley, B.P. Wilson, Chem.Commun., (1998), pp887. A. Henglein, R. Ulrich, J. Lilie, J. Am. Chem. Soc., (1989), 111, pp1974 E. Gonze, Y. Gonthier, P. Boldo, A. Bernis, Chem. Eng. Sci., (1998), 53,

pp523.
75

V. Renaudin, N. Gondrexon, P. Boldo, C. Ptrier, A. Bernis, Y. Gonthier,

Ultrasonics Sonochemistry, (1994), 1, S81.

76

C. Ptrier, M-F. Lamy, A. Francony, A. Benahcene, B. David, V. Renaudin,

N. Gondrexon, J.Phys. Chem., (1994), 98, pp10514

77 78

G. Mernyi and J.S. Lind, J. Am. Chem. Soc., (1980), 102, pp5830 J.S. Lind, G. Mernyi and T.E. Eriksen, J. Am. Chem. Soc., (1983), 105,

pp7655
79 80 81 82

E.J. Vitt, D.C. Johnson, J. Electrochem. Soc., (1991), 138, pp1637 S. Sakura, Analytica Chim. Acta., (1992), 262, pp49 G. Bottu, J. Bioluminescence Chemiluminescence, 1991, 6, pp147. L.L. Klopf, T.A. Niemann, Anal. Chem. 1983, 55, pp1080.

105

83

W.R. Seitz, D.M. Hercules, in Chemiluminescence and Bioluminescence,

M.J. Cormier, D.M. Hercules, L. Lee, Eds. Plenum Press: New York, 1973, pp427429.
84

L.J. Krika, J. Xiaoying, G.H.G. Thorpe, B.E. Edwards, J. Voyta, I.

Bronstein, J. Immunoassay, 1996, 17, pp61

85 86 87 88

K.S. Suslick, Science, 1990, 247, pp1439. K.S. Suslick, Scientific American, 1989, 260, pp80. A. Henglein, Ultrasonics, (1987), 25, 6. R.E. Verall and C.M. Sehgal, Ultrasound Its Chemical, Physical and

Biological Effects, K.S. Suslick (Ed.), VCH Publishers: New York, 1988: Chapter 6, pp227
89 90 91 92

K. Makino, M.M. Mossaba, P. Riesz, J. Am. Chem. Soc., 1982, 104, pp3537. C.-H Fischer, E.J. Hart, A. Henglein, J. Phys. Chem., 1986, 90, pp1954 E.J. Hart, A. Henglein, J. Phys. Chem., 1987, 91, pp3654 C. Ptrier, M-F. Lamy, A. Francony, A. Benahcene, B. David, V. Renaudin,

N. Gondrexon, J.Phys. Chem., (1994), 98, pp10514

93 94 95 96 97 98 99

K.S. Suslick and M.W. Grinstaff, J. Am. Chem. Soc., (1990), 112, pp7807 E.J. Hart and A. Henglien, J. Phys. Chem., (1985),89, pp4342 B. Lippet, J.M. McCord and I. Fridovich, J. Biol. Chem., (1972), 247, pp3736 E.J. Hart and A. Henglien, J. Phys. Chem., (1986),90, pp1954 J. Rabani and S.O. Neilsen, J. Phys. Chem., (1969), 73, pp3736 J.H. Baxendale, Radiat. Res. Suppl., (1964), 4, pp114 J.S. Lind, G. Mernyi and T.E. Eriksen, J. Am. Chem. Soc., (1983), 105,

pp7655
100

G. Mernyi and J.S. Lind and T.E. Eriksen, J. Phys. Chem., (1984), 88,

pp2320
101 102 103

L. Erdey, I. Buzs and K.Vigh, Talanta, (1966), 13, pp463 J.H. Baxendale, J.Chem. Soc.,Faraday Trans. 1, (1973), 69, pp1665 T.E. Eriksen, J.S. Lind, G. Mernyi, Chem. Soc., Faraday Trans. 1, 1983, 79,

pp1493
104

F. Ross, A.B. Ross (Eds.), Natl. Stand. Ref. Data Ser., Natl. Bur. Stand., 1977,

59
105

B.P. Barber, S.J. Putterman, Physical Review Letters, 1992, 69, pp3839

106

106

T.J. Mason, Sonochemistry (The Uses of Ultrasound in Chemistry), The Royal

Society of Chemistry, (1990) Chapter 2, p12

107

T. J. Mason, J.P. Lorimer, Sonochemistry: Theory, Applications and Uses of

Ultrasound in Chemistry, Ellis Horwood Publishers Ltd, pp25.

108 109

F.E. Fox, G.D. Rock, J. Acoust. Soc. Am., 1941, 12, pp505. R.E. Apfel, in Methods of Experimental Physics Vol 19, Ultrasonics, P.D.

Edmunds, Ed., Academic Press. New York, 1981,Chapter 7, pp355.

110

G.N. Pace, A. Cowley, G.A.M. Campbell, J. Acoust. Soc. Am., 1997, 102,

pp1474.
111 112 113 114

H.R. Suiter, J. Acoust. Soc. Am., 1992, 91, pp1383. S. Hilgenfeldt, D. Lohse, M. Zomack, Europ. Phys. J. B., 1998, 4, pp247. T.G. Leighton, Ultrasonics Sonochem., 1995, 4, pp247. N.A. Watkin, G.R. terHaar, I. Rivens, Ultrasound Med. Biol., 1996, 22,

pp483.
115

E.J. Pugh, Electrolytic and Ultrasonic Cleaning of Steel Wire in Neutral Salt

Solutions, M.Phil. University of Wales, Swansea (1996)

116

S.F. Lancelot, Electrolytic and Ultrasonic Cleaning of Steel Wire in Neutral

Sodium Tripolyphosphate Solutions, M.Phil. University of Wales, Swansea (1997)

117 118 119

R.G. Compton, J.C. Ecklund, S.D. Page, J. Phys. Chem., 1995, 99, pp4211. J. Klima,C. Bernard, C.Degrand, J. Electroanal. Chem., 1994, 367, pp297. R.G. Compton, J.C. Ecklund, S.D. Page, G.W.H. Sanders, J. Booth, J. Phys.

Chem., 1995, 98, pp12,410.

120 121

D.J. Walton, L.D. Burke, M.M. Murphy, Electrochem Acta, 1996, 41, pp2747. R.G. Compton, J.C. Ecklund, S.D. Page, T.O. Rebbit, J. Chem.Soc, Dalton

Trans., 1995, pp389.

122 123

C.G. Jung, F. Chapelle, A. Fontana, Ultrasonics Sonochem., 1997, 4, pp117. R.G. Compton, J.C. Ecklund, F. Marken, D.N. Waller, Electrochem. Acta,

1996, 41, pp315.

124

J.Riesse, H. Francois, J. Vandercammen, O. Fabre, A. Kirche-de-

Mesmaeker, C. Maerschalk, J.L. Delplanke, Electrochem. Acta, 1994, 39, pp37.

125

D.J. Godfrey, in Corrosion, ed. L.L. Shreir, Newnes-Butterworth (1976), Vol.

1, pp124-132.

107

126

A. Damjanovic, in Modern Aspects of Electrochemistry, ed. J.OM.Bockris and

B.E. Conway, Butterworths (1969), no.5, pp369-485.

127

J.OM.Bockris, Sum Shah, Surface Electrochemistry, Plenum, New York,

(1993).
128

H.S.Wroblowa, S.B. Qaderi, J. Electroanal. Chem., 1990, 279, pp231.