Almost all methods for the manufacture of electrodes are based on electrode position of the lead dioxide from a bath containing its salts. Diirkes, in German Patent 2,259,821, describes a thermal method for making a lead-coated electrode, the lead being subsequently oxidized to lead dioxide. The patent describes the application of a paste of ferrous or zinc chloride and powdered quartz and lead, followed by a drying and then a firing at 6OWC. But apart from this example, all other methods described here are electrochemical. We shall, for reasons which will become apparent, discuss graphite-based anodes separately. Unless otherwise mentioned, all the coatings described here, derived as they are from acid baths, are preponderantly of the b -PbO2 variety. much more limiteed amount of work has been done on the preparation and behaviour of a-PbO2 coatings.
The important earlier patents are D.A.S. 1,182,211 and U.S. 2,945,791 (1960) and D.A.S. 1,496,962 (1966), all from the Pacific Engineering Corp. The second of these specifies addition of a surface-active agent (such as alkyl phenoxy polyoxethylene ethanol type) to reduce gassing, increase throwing power and give a more compact deposit. Nitric acid is also stated to give better throwing power. The stepwise reduction in current density is mentioned in both the first and the second of these patents. All the Pacific patents are characterized by details of "bath management" which are trivial on the laboratory scale, but become crucial when plating on an industrial scale. Typical bath composition from the earlier patents might then be: lead nitrate, 50-400 g L-1; copper nitrate trihydrate, 0-20 g l-1 nickel nitrate trihydrate, I 0 g l-1; sodium fluoride, 0- 5 g L-1; surface active agent, 0-7S g L-1; nitric acid, 0-4 g L-1; and pH = 1.5. This bath is operated at 45-70oC at currents which may be constant (e.g. 25 mA cm-2) or increase or decrease during the plating. The fluoride ions may serve to etch the oxides present on graphite. The patent D.A.S. 1,496,962 somewhat amplifies the question. The nickel is stated to act as a grain-size refiner the copper to prevent deposition (and thus depletion) of lead on the cathode. This it does presumably by reducing the overvoltage for hydrogen evolution at the cathode. (Wabner41 has tested PbO2 deposits for the presence of metals such as Cu or Ni and finds no detectable traces.) This patent also claims advantages in the maintenance of the level of dissolved iron below 0- 02 g L-1
Further reading shows that in D.A.S. 1,496,962 (which is equivalent more or less to U.S. Patent 3,463,707), the emphasis is on removal of iron(to less than 0.02 g L-1) and chlorides, and obviating the need to use the relatively costly Ni and Cu salts previously employed. These baths thus contained only sodium fluoride and nitric acid, apart from the lead nitrate at a concentration of 200 g l-1 . Thangappan and Nachippan6 describe the benefits to be obtained by deposition of the PbO2 in a bath filled with fluidized inert particles, i.e. one in which the mass transport of species has been enhanced compared with normal conditions. A much smoother deposit is claimed. Huss and Wabner7 use paraffin-filled graphite and evacuate it before deposition, to ensure absence of gases, while Narasimham8 describes electrolytic pretreatment of graphite in alkali with a 24-h "soak" period. Little has been published in the PbO2-graphite area, although an interesting glass-fibre/PbO2 composite "sleeve" to fit graphite rods has been described." Graphite is anisotropic in its electrical conductance, and Wabner" suggests this may lead to poor current distribution and thus increased corrosion. He also states that the graphite is capable of being "burnt" by the PbO2, i.e. oxidized, and reports that this can lead with time to formation of a void between the graphite and the inner surface of the PbO2 (which would presumably in any case be reduced). This constitutes another reason (according to him) why such anodes are comparatively short-lived, although the "Pepcon" commercial anodes are stated to last for 2 years approximately when fitted in hypochlorite cells. Narasimham and Udupa,48 in their recent and excellent paper reviewing preparation and uses of these anodes, do not support the views of Wabner. In their work, they describe how anodes in rod (75 cm long, 20 cm diam.) or plate (90 cm X 18 cm x 2 cm) form behave, lasting for as long as 2 years. Although these authors conclude by stating that Ti might be a better substrate, it is quite clear that graphite-based anodes are a wholly workable proposition.
Although graphite-based anodes are still in use, notably by the Pacific Engineering Corp. for their chlorate and hypochlorite production, the growth is now seen to be in titanium-based electrodes which allow thinner and mechanically superior cell designs. The first mention of these (as distinct from graphite) was in U.S. Patent 3,463,707 or British Patents 1,189,183 (1970) and 1,159,241 or French Patent 1,534,453. In the entire treatment, the question of the formation and deleterious effects of the passive oxide layer must always be borne in mind. As Wabner points out, theconcept of an electrode based on a Ti (a strongly reducing metal) substrate coated with 4-valent, highly oxidizing PbO2 goes against all expectations of success. The thermal coefficient of expansion of Ti (in contrast to graphite) is close to that reported for PbO2, thus minimizing thermal shock problems.50 While pinholes in the outer PbO2 layer will not, with these substrates lead to catastrophic corrosion, it will nevertheless allow a passive film high resistance to spread on the (Ti) substrate surface beneath the PbO2 layer.
The above-mentioned problems of deposition, whether anodiccathodic, on Ti have given rise to several patents aimed at solving the problem. Thus German Patent 1,170,747 discusses a Cr3+ and F- etch bath as pretreatment in electrodeposition, and U.S. Patent 2,734,8 specifies a pickling in 90oC HClfor 5-15 min with or without a previous HF/HNO3etch or a molten NaOH descale bath. U.S. Patent 3,207,6 79calls for anodization of the Ti to a "Yellow colour" prior to Pb deposition (but with a Pt "flash").
There are dozens of patents describing the best way in which to prepare PbO2 anodes. There is no way of knowing, in most cases, the true worth of the ideas proposed, for there is almost no example of performance or appearance of an anode so prepared compared with one in which the step was omitted. The best procedure is to detail all the steps, it being understood that several of these may well be omitted.
Preliminary cleaning and treatment. We here consider mechanic abrasion and degreasing. De Nora'suggests sandblasting with fine silica sand followed by degreasing with benzene-a solvent virtually pro-scribed in many cases today. Trichloroethylene is now usually specified. They note that a pre-etch in boiling HCl leads to irregular PbO2 deposition, with a tendency to cracking. Barrett" suggests that the sandblasting is vital to ensure a good "keying-in" of PbO2 to the Ti substrate, and that choice of correct sandblasting conditions is vital.
Preplating stages. Prior to electrodeposition, almost all procedure specify some means of removing the titanium oxide (although as seen above, its formation is also specified). U.K. Patent 1,373,611 specifies a removal by cathodic reduction of the oxide. In theory this is convenient in that it requires no separate stage but simply a reversal of electrode polarity, although subsequent work shows H2SO4 is the best cathodization solution. Most other sources specify an etch in HCl, HF, NaF or oxalic acid. D.O.S. 2,012,506 specifies 10 min at 25oC of a 0.5 m HF/4 m NaF solution, another etch being specified in Belgian Patent 702,806. A completely different idea is proposed by Wabner et al.11 They suggest that the deeper scratches which can be seen under the microscope after mechanical treatment act as latent fault sites, even though they appear to be evenly.covered over by the lead dioxide. After many hours operation, these spring open again. Carborundum inclusions, and hollows with covered patches of oxide, are other sources of failure. These workers etch for 1h in boiling oxalic acid (15%). After a short time, when the oxide has been etched away, a violent hydrogen evolution is seen, the metal dissolving to form a reddish-brown titanium oxalato(III) complex. Figures 7 and 8 show S.E.M. photographs illustrating the "stormy-sea" effect of this process. In the next stage (to which the authors state they attach the greatest importance), the electrode, after washing, is boiled in TI(IV) oxalate and oxalic acid (0.2 m and 1.25 m, respectively). The authors state that neither with the naked eye nor with the S.E.M. does this surface show any difference from the one seen after the previous etch. The effect is only seen after plating. The whole idea is taken from D.O.S. 2,306,957 and it is worth quoting what the authors themselves stress," namely, that if the PbO2 deposit is chemically stripped from a Ti substrate so treated, the "effect" remains and a fresh PbO2 deposit on that surface constitutes an anode as good as the original one. Whether this argues in favour of some compound or species formed on the surface, or simply reflects the physical state of subdivision of the surface, is not known. We would opt for the latter suggestion. In a somewhat similar vein, D.O.S. 2,023,292 suggests an intermediate firing process of the Ti following the application of an aqueous solution of CrCl, or similar compounds which are listed there.
Wabner" reports many experiments aimed at illustrating the effect that the TI(IV) pretreatment has on the substrate. For example, even at 60oC he states that TiO, formation is inhibited in air, in contrast to the untreated metal. By means of S.E.M. photographs, he shows how the PbO2 deposits more evenly on treated Ti instead of forming isolated clumps, and it is suggested that PbO2 growth into fissures is likewise facilitated, such fissures in their uncovered state being postulated as future electrode-failure sites. Time-potential plots of Ti in the etch bath and in PBNO, plating baths all reveal differences in behaviour, and Wabner quotes the work of Thomas and Nobe@' as well as that of Goldberg and Parry- in showing that TI(IV) ions inhibit corrosion of the metal apparently by facilitating passivation. This work is possibly the source of inspiration underlying the TI(IV) pretreatment concept.
It is impossible to condense here the extensive work of Wabner, much of which has been published. We do not believe that any organic. compound could survive prolonged contact with PbO2 without itself undergoing oxidation and that if the "magic" of the TI(IV) treatment is to be explained, it lies either in the subsequent decomposition of the oxalato or perchlorate to form a stable inorganic species such as the titanoplumbate, or in some effect on the morpbology of the surface. In so far as the performance of such treated anodes can be matched in every way, as far as can be ascertained, by well-prepared Ti alone, we opt for the latter suggestion. D.O.S. 2,344,645, which again raises problems of electrode failure after operation at high anodic potentials or long terms, seeks to avoid this failure (due to oxide passivation on the Ti) by creating an interlayer. Plasma-arc spraying of 0- 15-cm thick layers of carbides or borides of Ta or Ti gives a conducting layer which inhibits oxide formation. A grain size of 40-90 lim is suggested. Such a suggestion recalls earlier ideas of Beer based on Ti anodes with a nitride coating (D.A.S. 1,421,047), although irr this case the coating was intended to be the outer one in contact with solution.
The electrodeposition process. Lead dioxide may be anodically deposited from a variety of baths. One of the best critical evaluations of these comes from the work of Hampson and Bushrod.12 Since commercially viable depositions must operate at reasonable current densities, Hampson et al. excluded baths based on fluoborates, plumbites and silicofluorides, since at current densities above 5 mA cm-2, these gave highly stressed, flaky and poorly adherent deposits. Nitrate and perchlorate salts of lead alone gave satisfactory deposits. With the latter, solutions of 2 M lead perchlorate and 1 M.acid gave good deposits at up to 50 mA cm-2, provided the pH did not fall below 0. Acidity was controlled by circulation of the solution through a bed of PbO. Hampson also reports that lead growths on the cathode were a problem and that this was minimized by using the lowest possible currentdensity on the cathode, i.e. having a larger cathode:anode area ratio. Current efficiencies were approx. 100%. In almost all patents, the nitrate bath is preferred, possibly because of the explosive hazards of perchloric acid. Other workers specify pH 1.5-2.5 as the best condition.
Reporting on the nitrate bath, Hampson showed that the major phenomenon" in the nitrate bath was the build-up of nitrite, which, because it can be re-oxidized to nitrate at the anode, leads to loss in current efficiency. A graph shows that deposition efficiency ranges from 100% (nitrate absent) to less than 20% when 6% of nitrite is present. At higher current densities, the nitrite is further reduced to ammonium ions. The nitrite levels were held down by circulation of the electrolyte through Pb3O4 Alternatively, H202 may be added to remove NO2. These authors report no effect of temperature in the range 20-40oC, although commercially, baths are operated at 6OoC or higher. The U.S. Bureau of Mines" reports cracking when deposition takes place above 70oC.
Narasimham and Udupa18 have studied the throwing power of the nitrate bath which they state passes through a maximum, but against what is unclear. The effect of nitric acid and surface active agents on throwing power was referred to previously. Oddly very few of the patents cited here refer to the stress which occurs in electrodeposited PbO2. It is again Hampson and Bushrod12 who study this, and their results are not easy to summarize. Simple nitrate baths gave compressively stressed deposits.
An equally thorough study of the effect of current density, temperature and organic additives on stress of deposited on nickel and graphite is due to Narasimham and Udupa.13 In the later paper they explored the use of lead acetate as a buffer to relieve stress. Smith" also discusses the effect of deposition conditions on stress, higher stresses occurring in more dilute electrolytes. Current density below 70 mA cm-2 had little effect while well-stirred solutions gave stress-force deposits. Organics, notably sodium acetate, acted most effectively as stress relievers. Indian workers with much experience in this field appear to lay special emphasis on CTAB (cetyl trimethylammonium bromide) for stress relief and a very perfect surface. This has been described," as has its rate of loss" by oxidation etc.
Commercial baths operate at 60-70oC. Wabner 41 states that lower bath temperatures result in anodes giving higher overpotentials in subsequent use (for example, for oxygen evolution) and this probably results from the higher overpotentials obtained during the deposition process at lower temperatures and so there is thicker TiO2 film formation. All contain approx. 200 g l-1 lead nitrate. Some have 10 ml conc. HNO, l-1, others contain copper and/or nickel salts as described earlier. De Nora (U.K. Patent 1,192,344) operates at 300g l-1 lead nitrate and 3 g l-1 copper salts with 0.92 g l-1 "'Tergitol" (a sodium alkyl sulphate), and uses 250 A dm-2 current density. Bath compositions designed to produce correctly stressed anodes (for primary batteries) are discussed in U.K. Patent 1,340,914. It is pointed out that stress can be relieved by heating the finished electrode in air and allowing it to cool slowly. The 360 g l-1 nitrate solution is unbuffered at pH 3-4 and no other bath constituents are used except in one example where a two-stage deposit was built up, the inner one as above and the outer one with a lead acetate/sodium nitrate bath. (It should be pointed out that both this patent and Hampson's work cited above relate to depositions on nickel. However, the findings are probably of general applicability.) U.K. Patent 1,373,611 uses a bath with additions of Cu, Ni salts, free nitric acid, sodium fluoride and heptafluobutanol operating at 70oC, and uses a two-step deposition rate, the first 20 mA cm-2 the second 60 mA cm-2.
Among other ideas and suggestions we can include (frequently unspecified benefits) are D.O.S. 2,012,506 (effects of 010gg l-1
urea addition), D.O.S. 2,306,957 (the laying down of two discrete PbOz layers, the outer one harder than the inner one), the effect of current interruption on PbO@ structure, and the use of glass beads in the plating bath to avoid pitting due to oxygen bubble entrapment." The importance of stirring or agitating the bath is universally stressed. One paper uses a fluidized (silica) bed to improve quality and uniformity of deposits.' In spite of the widespread use of nitrate baths, advocates exist for other baths, for example, Lazarev34 finds H2SiF6+ CH3COOH as good as the nitrate bath -, a similar bath has been advocated, while sulphamate baths have also been patented.36
Miscellaneous work on deposition. Ghosh 16 has studied the effect of various levels of superimposed a.c. on the d.c. used for electro-deposition of the lead dioxide. Certain differences are noted in the case of deposition of a-forms but no significant changes are seen during deposition of the, B-form. A series of Soviet papers describe lead dioxide deposition from trilonate (EDTA) baths.17 However, no indication is given as to the benefits so obtained, although the work is thorough in respect of stress measurement etc.
Narasimham et al.29 have studied the effect of cathode geometry on the mode of deposition (mainly thickness) of the PbO2 at the anode.
Wabner@' maintains that deposition using ultrasonic probes results in a2OO-300 mV lower overvoltage for oxygen evolution when the electrode is in service.
Glassy PbO2. Wabner" has reported the formation of a "glassy" variation of PbO2. It is formed by deposition from a conventional nitrate type bath to which has been added some Ti(IV). The precise formula is given as follows: 2.5 ml of TiCl, are hydrolysed in 300 ml distilled water, neutralized with NH3 and the hydroxide/oxihydrate so formed is well washed with water. It is then redissolved in HNO3(10 ml) and allowed to stand at 20oC overnight. Finally 66 g of lead nitrate are added, the solution is made up to 200 ml (corresponding to I N lead nitrate) and clarified with active charcoal. The deposition takes place at 40 mA cm-2 and at 20oC. It was shown that the glassy deposit was a-PbO2 of the "Plattnerite" form (in fact without any of the traces of a -PbO2 normally found) and that the grain size was approx. 600 A in contrast to the 6000 A of normally prepared deposits. Thus methods such as S.E.M. failed to resolve any structure in such deposits at X 10,000 magnification. These are highly stressed and, after standing in water, or presumably in service, the morphology changes and visible crystalline structure begin to develop. The alkaline "tempering" process described elsewhere also has the same effect after 12 h. This interesting form of the dioxide may well find certain uses in analytical applications-for example, where its low specific surface area gives the benefits normally associated with polished and smooth metals, and where its apparent lack of long-term stability is less important.
Post-treatment of anodes. Wabner and Fritz" describe an electrochemical post-treatment consisting of cyclic voltammetry between + 700 mVand +2280 mV. The effects of this are apparent to the naked eye (colourchange from light grey to brownish-black) while an S.E.M. examination (as shown in this work) depicts a change from smooth rounded contours of the immediately formed anode to the mass of needles after the formation. A further cycling, this time between the more anodic limits of 1650 mV to 2350 mV is stated to give a more reproducible electrode for kinetic measurements.
We have already described the effects of heating electrodes in air to remove strain in the deposit. A similar proposal is found i D.O.S. 2,306,957 where the coated anode is suspended in weak hot alkali for a period of time.
Regeneration of anodes. As they become more widely used, methods of regenerating older anodes, removing the coating and corrosion products and redepositing fresh PbO2 will become available. To date, the only reference appears to be a Japanese procedure" for boiling the anodes in nitric and oxalic acid mixtures.
The form of the anode. The traditional graphite-based anodes were and still are rods of 2-10 cm diameter. Adhesion of the coating to these was usually good. The advent of metallic substrates posed problems. Attempts to deposit the dioxide on continuous expanses of flat section lead to progressive failure, even when the initial defect is only localized. Expanded mesh has proved to be an ideal form on which to coat PbO2 while another idea described in a U.S. Bureau of Mines report" and in D.O.S. 2,012,506 is to drill holes in flat sheet at regular intervals. The lead dioxide, with its good throwing power, deposits into and behind these holes and anchorage is thereby provided. Another comment made by several workers is the importance of avoiding sharp edges and corners on the form to be coated. If these are present, nodular growths of lead dioxide will form there.
The foregoing description of anode preparation relate to coatings which consist predominantly of B-phase PbO2. Little has been written concerning the deposition or behaviour of the a-phase although it is known to exhibit a lower oxygen overvoltage. Ruetschi25 describes the deposition of a-PbO2 on Pt. A detailed description of an a-PbO2 deposition bath is also given by Grigger19 using alkaline lead tartrate. However, the work reported by Grigger19 in fact relates to nitrate (e.g. ,B-phase) deposits. Lartey26 describes the formation of a very ductile but thin coating of a -PbO2 which is of good adherence and is black, shiny and "slippery", apparently having been deposited in laminae which flake off when a given thickness is exceeded. The Faradaic efficiency of the deposition, even at 5 mA cm-2was c. 30%. Carr and Hampson27 have also reviewed baths for alpha phase deposition while Issa et al.28 have prepared deposits of mixed a- and B-PbO2 (92%, 80% and 60% a-phase) using formaldehyde as a reductant in th edeposition bath. The study of Gancy30 covers both Faradaic efficiency and also stoichiometry, as well as other details. Clarke 31 states that during or after PbO2 deposition from alkaline baths, some sort of luminescence is observed and that deposits exhibit a high degree of preferred orientation.
Recently, Soviet workers have patented a bath based on acetone (6-8 g L-1), urea (3-5 g L-1) and lead acetate (7-13 gL-1) in a 30-5O g L-1 NaOH solution.37
In spite of the many recipes for their preparation, few studies of the completed article exist. Largely for this reason, it is very difficult to evaluate the comparative merits of the various means of preparation.
Metallographic (S.E.M.) studies of surface textures are reported by Wabner7,11,41 and in the U.S. Bureau of Mines study;15 X-ray analysesof composition (85% B-phase, 15% a-phase) are cited in Belgian Patent 702,806 and also by Wabner and Fritz.11 Rusin51 shows how the a : B ratios vary with PbNO3 concentration and current density. A very fine study of the kinetics of deposition15 of a-PbO2 is reported by Fleischmann and Liler20 and this is only one of a series of fundamental studies devoted to the electrocrystallization process by Fleischmann.22,23 A more recent kinetic study of the electrodeposition and cathodic reduction of PbO2 on conductive tin oxide is due to Laitinen and Watkins.24 The best physical study of (electrical) properties of the PbO2 is probably that of Mindt.21 Among his more important findings are the increase in resistivity with time (over 12 days) and the effect of loss of oxygen on this. Also the fact that resistivity of the a -phase is greater than that of the B-phase, by an order of magnitude (pB = 10-2 fl cm, pa = 10-3 fl cm). In terms of the composite PbO2-TiO2, anode, Wabner points out that B-PbO2 and TiO2 are isomorphous (rutile) differing only by 7% in lattice size and that formation of a mixed crystal might be possible. This can be considered both in the context of the Pb-Ti interface and the binding processes there and also in the context of a "modified" PbO2 + TiO2 anode along the lines mentioned above. However, such an electrode could not, one feels; be made by purely electrodeposition processes.
Wabner also speculates as to the possible existence of other compounds in the system. He mentions lead titanates or titanoplumbates, and the importance of such species, which have been referred in the literature 42 remains to be resolved.* He quotes the work of Haufe et al.47 who show that the electrical conductance of TiO2 can be readily increased by several orders of magnitude when certain other oxides are added and this fact again may or may not be relevant to making better composite anodes. A range of TiO2 compositions can also exist with differing electrical properties.52 Scientifically as opposed to technologically, Wabner has published more than any other worker on the PbO2/Ti composite and his writings therefore deserve careful analysis.
His use of a.c. techniques to investigate processes occurring PbO2/Ti anodes has been criticized (see Chapter 9) and Canagarat and Hampson49 illustrate some of the pitfalls of this technique.
The basis of his "oxalate" pretreatment is hard to understand Possibly it does create on the surface of the Ti a monolayer or so of species which impedes the oxide passivation which might otherwise take place when the anodic deposition occurs. Randle,43 measure the potential of the Ti anodes during conventional deposition from nitrate bath onto simply etched (but otherwise untreated) Ti, notes the when the potential of deposition is greater than 2.0 V vs N.H.E., the resulting electrode is useless or has at best a life of a few hours before failure due to high electrical resistance occurs. Wabner points out that electrodes pretreated with the oxalato method retain their special advantages even when the PbO2 is stripped off after some considerable time and redeposited. In view of the extremely strong oxidizing proparties of PbO2, it seems inconceivable that any oxalate remains unoxidized after being so coated. Wabner believes that attempts obtain current-potential data with PbO2/Ti anodes are useless because of the large (and possibly indeterminate) iR drop across the PbO2 interface which he finds impossible to separate from the overall drop. This may well be so.
The main problem with Wabner's findings is the weight of evidence to contradict them or at least to restrict their validity. In fact, anodes manufactured in the first commercially successful place,44 based on PbO2/Ti with no interlayer or special pretreatment such as describes, have been operating in hundreds of installations in the U.K. and elsewhere (mainly in electroflotation plants) with a proven lifetiime of 2 years or longer. In further apparent contradiction to the assertions of Wabner,44 the anodes made and sold by Messrs Morgett Electrochemicals Ltd. are frequently stored in air for extended periods with no known deleterious effects on their lifetime, instead of leading to early passivation failure he suggests would take place. It is not possible here to state whether Wabner's assertions stem from experiences with poorly prepared anodes or whether the passivation failures he describes occur sooner in his solutions (2 N -- H2SO4) than they do in the chloride- containing media in which commercial anodes frequently operate Certainly passivation would occur more easily in the former than in the latter solutions. Wabner's other suggestion that 100 mA cm-2is the maximum permissible current rating on account of temperature rise from ohmic eating in the interlayer region also appears to be dubious. Electrodes have been operated by us and others for 20-30 days in chloride media at 400 mA cm-2 without showing the voltage increase he describes. In conclusion, the TI(IV) treatment does appear to reduce the rate of Ti passivation, but using properly prepared electrodes, such a procedure seems unnecessary. A discussion of corrosion of composite anodes is given on p. 299.
There seems little doubt that the PbO2/metal (especially Ti) anode has an important future in the electrochemical process industry Experience in our own laboratories and elsewhere has shown that a considerable degree of "know-how" is required before an anode capable of operating successfully over a long period can be made. Even now, it not clear what limits the lifetime of anodes, whether build-up of highly resistive films or simple loss of PbO2 are the main problems. From the large and diverse "recipes" for their manufacture, it might well be deduced that none of the elaborate "inter-layers" or pre/post-treatments are really necessary. Future work will probably therefore proceed in two directions,. These are:
(i) Study of the importance (or otherwise) of interlayers either
metallic or of TiN or similar compounds.
(ii) Study of mixed oxide coatings PbO2-MOx.
In this respect, the history of the chlorine anode which has evolved from a RuO2 through a RuO2TiO2 to a RuO2TiO2SnO2 electrocatalyst will undoubtedly be borne in mind. However, Mindt21 argues that high concentrations of "dopants" will be needed to cause any significant effects. In this connection, for example, USSR Patent 49S,714 40 describes the preparation of "resistive materials" based on PbO2, Sb2O, and other oxides. But knowing that, if well made, simple PbO2/Ti anodes have already a lifetime of 2 or more years, any improvements will have to show marked benefits especially if they are difficult to carry out.
As should now be obvious, the morphology of the lead dioxide anode can assume a variety of forms when examined under the optical or scanning electron microscope. Dr D. Wabner has kindly provided a selection of micrographs which are shown on the following pages and which represent a typical range of PbO2 surfaces.
We have seen elsewhere that the mode of corrosion of PbO2/Pb anodes takes place basically not at the surface but rather at the Pb(I 1) interlayer between the metal and the fully oxidizing PbO2. In the composite anode which forms the subject of this chapter, no such intermediate valence state exists (at least not as far as we know) and one would therefore expect such anodes to corrode more slowly, if at all. They do indeed show much higher corrosion resistance than their PbO2/Pb analogues. A recent paper by Lartey" summarizes published data and shows new corrosion information on these composites. The findings of Lartey are that corrosion is broadly related in Tafel fashion to the potential of the anode, and that effects of Cl- concentration are important mainly in as much as whether or not they "depolarize" the anode. Work has continued on this aspect of the problem and there are indications that the rate of corrosion (as shown by Lartey to be more or less constant over a 2-week period) subsequently declines to reach a lower steadystate value. Though these are only preliminary findings and subject to confirmation, it is worth noting that similar effects have been found in long-term sea-water corrosion tests of platinized titanium or similar precious metal-coated Ti anodes. An explanation was offered there in terms of a high initial loss rate due to poorly adherent metal grains on the outer surface of the anode, which exposed a more strongly adherent layer beneath.
The corrosion products of PbO2 anodes in laboratory tests appear to be mainly a powdery suspension of PbO2 which may subsequently settle out at the bottom of the reaction vessel. This poses a dilemma in that it is difficult to explain what is in effect a corrosion not involving any net chemical change in electrochemical (or indeed, chemical) terms. One alternative explanation can be attributed to purely mechanical attrition of the PbO@ due, for example, to the scouring effect of bubbles of gas formed at the electrode, If this is so, it should be possible to vary the mechanical properties of the deposited dioxide so as to achieve a mechanically more resistant structure. The means by which this might be done are obvious from the foregoing pages. There is, however, a less orthodox explanation which is as follows. It is not widely appreciated that the evolution of a gas at an electrode surface causes substantial potential fluctuations at the electrode-electrolyte interface. With a large electrode, such fluctuations will appear to be averaged out, but there is no doubt that they do occur, and can be of the order of 0- S V. The effect of these fluctuations is to reduce the overvoltage of the system momentarily. It can therefore be postulated that the corrosion protection afforded by the anodically formed PbO2 is periodically lost when a portion of the electrode is covered by a bubble. In such a case, the potential might fall back to the "sulphate" region. More rapid dissolution would then take place. The dissolved sulphate would, however, be reoxidized by contact with the anode, to PbO2, although under these conditions much of it would not deposit onto the anode in a mechanically satisfactory fashion. The result would then correspond with what is seen-a build-up of suspended PbO2 in the solution. Work is presently under way on several fronts, aimed at the elucidation of this question.
1. .J. Cotton and I. Dugdale, 3rd International Symposium on Batteries, Bournemouth, 1962,
pp. 170-183. Pergamon (1963). Also British Patent 869,618.
2. P. Faber, in "Power Sources 4" (D. H. Collins, ed.), p. 52S and discussion. Oriel Press,
Newcastle on Tyne (1973).
3. F. Beck, Ber. Bunsenges. 79, 233 (197S).
4. B. N. Kabanov, Electrochim. Acta 9, 1197 (1964).
5. A. T. Kuhn (ed.), "Industrial Electrochemical Processes". Elsevier, Amsterdam (1971).
6. R. Thangappan and S. Nachippan, Indian Patent 105731 (1967).
7. R. Huss and D. Wabner, Metalloberfldche 8, 305 (1974).
8, K@ C. Narasimham and H. V. Udupa, 1. Electrochem. Soc. jap. 29, 137 (1961).
9. 0. De Nora, British Patent 1,192,344.
10. F. Barrett, private communication.
11. D. Wabner and H, P. Fritz, Z. Naturforsch. 31B, 39 and 45 (1976).
12. N. Hampson and C. Bushrod, Brit. Corrosionj. 6,129 (1971); Trans. Inst.
Metal Finish 48, 131 (1970).
13. K. C. Narasimham and H. V. K. Udupa, Electrochim. Acta 15,1619 (1970);
16, 1301 (1971).
14. J. F. Smith, Trans. Inst. Metal Finish 53, 83 (197S). Also British Patent
1,340,914.
15. U.S. Bureau of Mines R.I. 8111 "@ad Dio.ide-Plated Ti Anodes..."
L. W. Higley, W. M. Dressel (1976).
16. X. Y. Ghosh, Electrochim. Acia 14, 161 (1969).
17. V. A. Volgina and E. A. Nechaev, Sov. Electrochem. 9, 984 (1973); 9,1717 (1973).
18. K. C. Narasimham and H. K. Udupa, "Current Sciences" 30, 139 (1961).
19. J. C. Grigger, H. C. Mille@ and F. D. Loomis, J. Electrochem. Soc. 105, 100
(19S8). See also U.S. Patents 2,872,405 and 2,945,790.
20. M. Fleischmann and M. Liler, Trans. Far. Soc. 54,1370 (1958).
21. W. Mindt, J. Electrochem. Soc. 116, 1076 (1969).
22. M. Fleischrnan and H. R. Thi,sk, Trans. Fai. Soc@ 51, 71 (1955).
23. M. Fleischman and H. R. Thirsk, Electrochim. Acta 2, 22 (1960).
24. H. Laitinen and N. H. Watkins, J. Electrochem. Soc. IZ3, 804 (1976).
25. P. Ruetschi, R. T. Angstadt and B. Cahan, J. Electrochem. Soc. 106, 547 (19sq).
26. R. Lartey, Ph.D. thesis, Salford University (1976).
27 J. P. Carr and N. Hampson, Chem Rev. 72, 679 (1972).
28. I. M. Issa, M. S. Abdelal and A. A. El Miligy, J. Appi. Efectrochem. 5, 271 (1975).
29. K. C. Narasiniham, S. Sundarajan and H. K. Udupa, Bull. Acad. Polon.,
Ser. Sci. 13,619 (1965).
30. A. B. Gancy, J. Electroahem. Soc. 116, 1496 (1969).
31. R. L. Clarke, private communication.
32. K. C. Narasimham and H. Udupa, J. Appl. Electrochem. 6,189 (1976).
33. K. C. Narasimham and H. Udupa, Can. J. Chem. 53, 3327 (1975).
34 .V. Lazarev, Izv. Vyszh. Ucheb. Zaved, Khim. 18, 1336 (1975).
35. Japanese Patent 69,13,370
36. German Off Patent 2,200,901
37. USSR Patent 456,054
38. Japanese Patent 72,38742
39. Japanese Patent 72,18,844
40. USSR Patent 495,714
41. D. Wabner, Habil Thesis T. U. Munich (1976).
42. V. S. Krikorov Izv. Akad. Nank. SSSR Neorg. Mater. 11, 461 (1975).
43. T. Randle, private communication.
44. F. Barrett (Morgett Electrochemicals Ltd.), private communication.
45. N. T. Thomas and K. Nobe, J. Efectrochem. Soc. 119,14SO (1972).
46. I. B. Goldberg and E. P. Parry, J. Electroanal. Chem. 54 (1974) 427.
47. K. Hauffe, H. Gruenwaid, Z. Elektrochem. 56,937 (1952).
48. K. C. Narasimham and H. V. Udupa, J. Electrochem. Soc. 123,1294 (1976).
49. S. G. Canagaratna and N. A. Hampson, Surface Technol. 5, 163 (1977).
50. W. Paimaer, Z. Elektrochem. 29, 415 (1923).
51. A. 1. Rusin, J. Appl. Chem. USSR 43, 2656 (1970).
52. D. Bergner, Chem. Ing. Technol. 47,137 (1975).
53. R. Lartey and A. T. Kuhn, Corrosion-NACE 32,73 (1977).
P> 47. K. Hauffe, H. Gruenwaid, Z. Elektrochem. 56,937 (1952).
48. K. C. Narasimham and H. V. Udupa, J. Electrochem. Soc. 123,1294 (1976).
49. S. G. Canagaratna and N. A. Hampson, Surface Technol. 5, 163 (1977).