PLATINUM METALS REVIEW

A quurterly survey of reseurch on the platinum metuls urrd of dwelopments in their applications in industry

VOL. 13 APRIL 1969 NO. 2

Contents

The Electrosynthesis of Organic Compounds 46

Measurement of Electrodeposit Thickness 52

Electron Concentration as a Guide to Alloying Behaviour 53

A Rhodium-Platinum Probe for Flame Velocity Measurement 56

Ruthenium-Platinum Oxide Catalysts for Hydrogenation Reactions 57

Complexes of Substituted Olefins with Salts of the Platinum Metals 60

Making Corrosion-Resistant Brazed Joints in Stainless Steel 61

Thermocouples under Neutron Bombardment 64

The Platinum Metals in the Measurement of Temperature 65

Palladium-Titanium Alloy in Chemical Plant 67

Sir William Crookes 68

Abstracts 73

New Patents 79

Communications should be addressed to The Editor, Platinum Metals Review Johnson, Mutthey & Co., Limited, Hatton Garden, London, E.C.1 The Electrosynthesis of Organic Compounds

By Professor Martin Fleischmann, Ph.D., and Derek Pletcher, Ph.D. Department of Chemistry, University of Southampton

reactions such as substitutions and cyclisa- The decelopment of electrochemical tions. routes in the industrial production of In order to explain the interest in electro- organic compounds is beginning to chemistry it is convenient to contrast electro- attract considerable interest. Already chemical reactions with homogeneous or two large-scale processes are in opera- heterogeneous reductions and oxidations tion, one producing adiponitrile from using hydrogen and air or oxygen. The free acrylonitrile, an intermediate step in the energy change, AGO, of these processes is manufacture of Nylon 66, and the other equivalent to a cell potential, E”, given by producing lead tetra-ethyl anti-lcnock AG= --nFE’ compounds. This article reviews the By referring to a scale of free energies or basic electrochemistry involved in this electrode potentials, Figure I, it is evident type OJ; synthesis and outlines the great that such spontaneous reactions are only pos- possibilities being opened up by ad- sible within the potential range limited by the vances in technique, in reactor design reduction of oxygen or the oxidation of and in the development of new types of hydrogen. This driving force only amounts electrode structures in which platinum to approximately 0.5 eV or 10 kcals/mole. By and its associated metuls will play an contrast, it is possible to carry out electro- important part. chemical reactions at potentials between +3.5 V and -2.5 V even in aqueous solution, if suitable electrolytes and electrodes are In recent years research in the field of chosen. Thus, the driving force for an elec- organic electrosynthesis has received consider- trode process is of the order of 3eV or 60 able stimulation from the forecast that, with kcalslmole. Electrochemical reactions there- the advent of nuclear power, electricity will fore enable one to introduce considerable become cheaper compared to chemical oxi- energy into molecules at low temperature and dants and reductants. Furthermore, the the order of magnitude of this energy is, in reasons for the failure of earlier workers to fact, that required to break chemical bonds. achieve selective reactions have become ap- Figure 2 compares the energy ranges in which parent and by controlling the electrode different methods of activating molecules will potential, the solution conditions (solvent, be effective. pH, concentration of species, etc.) and by Clearly, in view of the energy which may choice of suitable electrode materials, some be introduced, it is not surprising that many measure of selectivity is now possible. It “high energy” chemicals that are used as therefore seems likely that electrochemical oxidmts and reductants in synthesis are pre- techniques will have an increasing role in pared electrochemically (e.g. sodium, man- preparative organic chemistry, both in indus- ganese dioxide, chlorine) and there are try and in the laboratory as a standard method obvious advantages in avoiding such inter- for oxidations and reductions as well as other mediates.

Platinum MetalsTev., 1969, 13, (2), 46-52 46 + 4.0

0,+ ~H,O+ 6e --+60H- + 3.0 C10, + e _jC10,- carbornurn

cO3- -+ e +Co2- i2.0 C10- + H,O + 2e +C1- + 20H- T :z - H- radical OL- 2Hz0 + 4e __j_ 40H radical cation T 1.0 C1, - 2e __j 2C1 Te

2H ~2e+HA 0 substrate 1 I -e - L radical Zn2' + 2e 4 Zn 1.0 radical anion +HT

i - 2.0 Na- + e dNa +e

-3.0 carbanion1 Solvated electrons

Fig. 1 The scale of free energies or electrode potentials. This shows the approxi- mate potentials (in volts) at which various electron transfer reactions take place and indicates the intermediates that may be formed in electrode processes

The intermediates formed in electrode pro- of nitrobenzene where phenylhydroxylamine cesses are indicated in Figure I and are may be produced at low negative potentials entirely consistent with the large driving and aniline at more negative potentials (I). force of such reactions. While the figure is not drawn to scale, it nevertheless shows the approximate points at which radical ions, - H,O radicals, carbanions and carbonium ions would be expected to form. Thus it can be seen that by controlling the electrode potential - H,O a measure of selectivity may be achieved. The -aNHz classical example of this idea is the reduction A further example is the oxidation of 9, IO-

0 1.0 3.5 6.0 8.0 12.0 ev I I I I I photochemistry radiation chemistry f

electrochemistry corona discharge <- <------Fig. 2 The diagram illustrates the range in energy (I electron volt = 23 k.cal male-') which may be introduced into molecules by various 'high energy' techniques

Platinum MetalsTev., 1969, 13, (2) 47 diphenylanthracene at platinum electrodes in trode activity rapidly falls with time. In view aprotic solvents. At low potentials this leads of thc wide range in potential over which the to the radical cation and at high potentials to inert metals may be used the polarisation can the dication (2). be reversed for short controlled periods of time and the electrode reactivated cathodic- ally. In this way controlled successful oxida- tion of hydrocarbons may be achieved. Intermediates of the kind shown in Figure I will be formed in many electrode reactions. For example, carbonium ions will be gener- Q ated at platinum electrodes during the oxida- tions of alkyl halides (4), carboxylic acids (5), hydrocarbons (6), and amines (7). ? RI- >R BIL ,e RCOO--+R-I CO, t 2e

RH --R- I H-' i2e

Carbanions and dianions are generated in the It is also possible to exert a measure of con- reduction of alkyl halidcs, aromatic hydro- trol over electrode reactions by maintaining carbons (g), quinones ( IO), nitrocompounds the electrode at a series of potentials for con- (I I), and activated olefms (12), e.g. trolled durations. In this way it is possible, RI , 2e- .R-i I- for example, to switch successively from 0H-H+ e--0H,- oxidation to reduction and thus to change the + products of the reaction (3). CH, :CH - CN 2e-- - Cg2-cH - CN

Radicals arc also frequently produced as in the oxidation of carboxylic acids (13) and of carbanions (14, IS), and the re- duction of some alkyl halides (8), ketones (13)~quinones (17) and tetra-alkylammonium salts (IS), e.g.

RCOO.-+R' +CO,+e 4Hi+4c @N-XG - xm CH(COOEt),--->CH(COOEt), +e H,K SII

Al(CH,),----Al(CH,), CCHs'+e

A further important application of pulse electrolysis is likely to be the control of ele- 0. trode activity by control of the electrode 0 history. For example, in the use of platinum 0,CO -H+'++O,C.OH metal electrodes for the oxidation of hydro- carbons in non-aqueous solutions the elec- RI -ie--R' +I-

Platinum MetalsTev., 1969, 13, (2) 48 In order to study these highly reactive The intermediates which are formed react intermediates, it is clearly necessary to use in similar ways to those observed when they unreactive solvents and electrode materials. are generated in homogeneous processes, Indeed, it is the advent of aprotic polar although their reactions will often be modi- solvents in which the intermediates have an fied by the adsorption on the electrode. In appreciable half life that has allowcd the typical reactions, radicals will dimerise (Z I), characterisation of many intermediates by electrochemical and spectroscopic methods z EtOOC (CH,), COO- -*e > (in particular ESR). In many other cases the zCO,+EtOOC (CH,), COOEt nature of the intermediates has been inferred attack double bonds (zz), from the products of the reaction and by analogy to known reaction mechanisms. EtOOCCOO e->C02t-COOEt butaAe%-t EtOOC -CH, -CH =CH-CH;-+dimer Reactions at Platinum Electrodes or react with certain electrode metals (23), The generation of the reactive species naturally also demands inert electrode sub- R Mg I --e+MgI : R. 'b+ PbR, strates and platinum metals have been widely Carbanions or dianions formed from aro- used in aprotic solvents over the range +3.0 matic hydrocarbons will protonate to form di- to -3.oV. On the other hand, in water or hydroaromatic compounds (9) or undergo other protic solvents the use of platinum other typical coupling reactions such as that metals has been conked to anodic reactions with carbon dioxide (17). because the decomposition of solvent to form hydrogen is strongly favoured. However, at H COO- the present time a considerable research ze + zCOz effort is being devoted to investigations of the oxidation and substitution of organic com- H 'coo- pounds and for these processes the use of inert platinum electrodes is of key import- Similarly, the dianions formed from activated ance. Indeed, most of the oxidative reactions olefines will react with unsaturated com- listed in this article have been carried out on pounds as in platinum metal electrodes. Cg, - CH -CN + CH, = CHCN +2H I--+ The formation of the intermediates has NC -(CHJ-CN been written as if the electrode surface is not directly involved in the reaction step. In and intramolecular reactions leading to many cases, however, the reactive species will cyclisation are also possible (24), be adsorbed on the surface and the catalytic role of the electrode will be of key importance. For example, the reductive hydrogenation of olefines is dependent on the use of platinum blacks (19) as is the dissociative adsorption of Carbonium ions are naturally highly reactive saturated hydrocarbons or methanol as an essential step in the overall reactions on fuel and usually attack the solvent or deprotonate to form olefines. For example, in acetonitrile cell electrodes (20). Again the coupling of radicals produced from carboxylic acids in amides are formed (4) the Kolbe or Brown-Walker reaction is R++CH,CN--+CH,-C-+=N-R % dependent on the electrode surface, being CHa--CO -NHR favoured by platinum in aqueous solutions (13). and in methanol ethers are the products (25)

Platinum MetalsTev., 1969, 13, (2) 49 as shown by the reaction

MeOH -Me0 (o> OMC I OhIc In other cases carbonium ions show their generation of CO". In this way reaction typical rearrangements as in the oxidation of occurs within a zone adjacent to the electrode neopentyl iodide (4), ring formation (26), and the actual electrode process e.g. CO"-+CO~~' te CH, -CH, -CH,-COO- LZe% serves to "drive" the homogeneous catalytic CO, I CH, -CH,-CH, - H++~~2-~~2reaction (30). Other examples of such \/ CHZ reactions are the oxidation of propylene by Hgrl ring expansion (27), or ring contraction (28), CH,-CH-CH,i 4HgZ t HzO---t e.g. CH, = CH -CHO T 4H- TaHgq ' CH ,COO- CH where the mercuric ion is regenerated electro- chemically (31). These indirect reactions (yH -25 @ - co, which involve highly oxidising conditions can again only be achieved by the use of inert platinum metal electrodes. The oxidation of propylene to propylene oxide CH,-CCH-CH, +ClO-- The substitution, for example cyanation or CH2-CH CH, 1 C1 acetoxylation, of aromatic compounds also \/ 0 takes place via cationic intermediates since by electrochemically generated hypochlorite the reaction will only take place at potentials (32) and the reaction of olefines with carbon at which the hydrocarbons are oxidised even monoxide and methanol in the presence of if the substituting anion is oxidised at lower platinum carbonyls to give methyl esters of potentials (29). p unsaturated acids (33) Industrial Processes in Operation C,H,CR-CH,ICO I -0Me- z It will be seen that many of these reactions C,H,CR CHCOOMe lead to industrially useful products, for are further interesting examples of indirect example, diterminal substituted compounds. processes. Indeed, the formation of lead tetra-alkyl and Some of the most extreme oxidations of adiponitrile have been made the basis of carried out at present are the perfluorinations new commercial synthetic processes by Nalco of aliphatic hydrocarbons in anhydrous and Monsanto respectively. hydrogen fluoride. It is not clear, however, Figure I also indicates the potential regions whether these proceed via the generation of in which a number of inorganic intermediates oxidised aliphatic species or of fluorine (34, are generated. For example, Col'I may be 35). generated in situ at an electrode and allowed Figure I also shows that highly reducing to react with an organic substrate with re- species may be generated and the ultimate

Platinum MetalsTev., 1969, 13, (2) SO limit is, in fact, the formation of solvated simplifying the work-up by avoiding the intro- electrons in suitable solvents such as ammonia, duction of reagents which must subsequently amines, or hexamethyl phosphoramide (36, be removed. 37). The electrons will react with aromatic hydrocarbons and give di- or tetrahydro Development of Reactors derivatives. and Electrodes HH A further major development is likely to be the construction of new electrochemical reactors which will permit an easier scale-up. In recent years the bulk of the effort has been H H devoted to the development of highly active It may be noted that benzene cannot be catalytic electrodes for fuel cells, the aim being directly reduced at electrodes. the complete oxidation of hydrocarbons to carbon dioxide. It is likely that further re- Importance for search will lead to structurcs capable of giving, Large-scale Synthesis for example, controlled partial oxidations but It will be apparent that many of the inter- utilising many of the advantages inherent in mediates which have been described are of the design of the porous fuel cell electrodes. considerable importance for large-scale syn- The development of these and other new thesis. A major aim of these synthetic types of electrode structures such as fluidised procedures in the coming years will un- beds and packed beds and the development of doubtedly be the build-up of complex mole- membranes suitable for separating electrode cules from relatively simple starting materials compartments will undoubtedly lead to the as might be achieved by the reaction of construction of low cost reactors having a radicals with olefines. high throughput. The operation of many of Linked to this aim will be the activation of these electrodes will be dependent on the use unreactive substrates such as carbon dioxide of finely divided platinum metals imbedded in and saturated hydrocarbons by utilising the the structure to give a high surface area for high driving force of electrochemical pro- the surface reactions at low cost. There is cesses. Of equal importance to the develop- every prospect, therefore, that the chemical ment of new synthetic routes will be the use engineering technology to make use of the of electrochemical procedures to increase the chemical advantages of electrosynthetic pro- selectivity of existing reactions and to reduce cedures on the industrial scale will be avail- the number of steps in a synthesis as well as able.

References F. Haber, Z. Electrochem., 1898, 90, 3263 9 M. E. Peover, “Electroanal. Chemistry”, M. E. Peover and B. S. White, J. Electroanal. Edward Arnold, London, 1967, vol. 11, p. I Chem., 1967,13,93 10 M. E. Peover, J. Chem. SOL.,1962, 4540 M. Fleischmann, I. N. Petrov and W. F. K. w~~~-J~~~~,“Electrochemistry-proceed- I I A. H. Maki and D- H. Geske, 7. Am. Chem. ings of 1st Australian Conference”, Pergamon 1961J 837 Press, London, 1964, p. 500 12 M. M. Baizer,J. Electrochem. Soc., 1964, 111, L. L. Miller and A. K. Hoffmann, J. Am. 215 Chem. Soc., 1967,89,593 13 B. C. L. Weedon, Advances in Organic Chem., L. Ebersonand K. Nyberg, ActaChim. Scand., 1960, I, I 1964, 18, I567 14 T. Okubo and S. Tsutsumi, Bull. Chem. SOC. M. Fleischmann and D. Pletcher, Tetrahedron Letters, 1968, 6255 japan,1964, 3,, 1794 K. K. Barnes and C. K. Mmn, J, Org. Chem., I5 H. Lehmkuhl, R. Schaefer and K. Ziegler, Chem.-Ing.-Tech., 1964, 36, 612 1967, 32, 1474 F. Lantelme and M. Chemla, Electrochim. 16 P. Zuman, D. Barnes and A. RyvolovA- Acta, 1965, 10, 657 Kejharova, Disc. Farada.y Soc., 1968, 45, 202

Platinum MetalsTev., 1969, 13, (2) 51 I7 R. Berkey, S. Wawzonek, E. W. Blaha and 27 E. J. Corey, N. L. Bauld, R. T. La Londe, M. E. Runner, J. Electrochem. SOC.,1956,103, J. Casanova and E. T. Kaiser, J. Am. Chem. 456 Soc., 1960, 82, 2645 I8 M. Finkelstein, R. C. Peterson and S. D. 28 A. J. Baggerley and R. Brettle,?. Chem. SOC., Ross, Electrochim. Acta, 1965, 10, 465 c, org., 1968, 2055 19 S. H. Langer and H. P. Landi, J. Am. Chem. 29 N. L. Weinberg and H. R. Weinberg, Chem. SOC.9 1963, 85, 3043 Rev., 1968, 68, 449 20 E. Gileadi and B. Piersma. “Modern Asuects -70 M. Fleischmann et al,.- unuublished work of Electrochemistry”, Butierworths, LoLdon, 31 F. Goodridge, R. E. Plirnley, I. D. Robb and 1966, vol. IV, p. 47 J. M. Coulson, Brit. Pat. Appln. 16364j67 21 M. Ya. Fioshin, A. J. Kamneva, Sh. M. 32 J. A. M. LeDuc, U.S. Pat. 3,288,692 (1966); Itenburg, L. I. Kazakova and Yr. A. Ershov, Chem. Abs., 1967,24,43238 p. Khim. Prom., 1963, 263 T. Inoue and S. Tsutsumi, Am. Chem. SOC., 22 M. Ya. Fioshin, L. A. Mirkin, L. A. Salrnin 33 J. and A. Kornienko, Zh. Vxes. Obshchestva im 1965,87,3525 D. I. Mendeleeva, 1965,10,238 34 J. Burdon and J. C. Tatlow, “Advances in 23 E. M. Martlett, Ann. N.Y. Acad. Sci., 1965, Fluorine Chemistry”, Butterworths, London, 125, 12 1960, I, 129 24 J. D. Anderson, M. M. Baizer and J. P. 35 S. Nagase, Fluorine Chem. Rev., 1967, I, 77 PetrovichJ. Org. Chem., 1966, 31, 3890 36 H. W. Sternberg, R. E. Markby and I. 25 B. Wladislaw and H. Viertler, Chem. Ind. Wender,J. Electrochem. SOC.,1966, 113, 1060 (London), 1965, 39 37 R. A. Benkeser and E. M. Kaiser,J. Am. Chem. 26 W. J. Koeh1,J. Am. Chem. SOC.,1964,86,4687 SOC.,1963,852858

Measurement of Electrodeposit Thickness The use of the beta back-scatter technique thickness by reference to a calibration graph. for the non-destructive measurement of the Variations of this equipment provide the thickness of electrodeposited coatings - and facility to pre-set a minimum thickness or a the basic design of a commercial instrument, range of thicknesses, pass or fail lights being the Beta 750 - were described some two years activated by the count figure obtained. ago (Platinum Metals Rev., 1967, 11, 13). A direct reading instrument, the Beta 752, The success of this equipment has now led incorporates a meter reading of deposit to the introduction of an extended range of thickness in micro-inches, based upon cali- instruments to meet specialised needs and to brated scales for specific combinations of satisfy the increasingly stringent demands for deposit and basis metal. This instrument is consistency and accuracy, particularly where particularly suitable where large numbers of electrodeposits such as rhodium, palladium tests must be carried out and maximum and gold are involved. accuracy is not a vital consideration. The Beta 751 provides a four-figure digital The units are produced by Panax Equip- read out of the beta count over a pre-selected ment Limited and marketed by Johnson time, this figure being related to deposit Matthey.

The Beta 752 unit is fitted with a calibrated scale for the direct reading of electro- deposit thickness. Zero and full-scale deflection points are set by adjusting the two potentiometers

Platinum MetalsTev., 1969, 13, (2) 52 Electron Concentration as a Guide to Alloying Behaviour ENGEL-BREWER AND THE PLATINUM METALS

By A. s. Darling, Ph.D.,M.I.Mech.E. Research Laboratories, Johnson Matthey & Co Limited

Some unexpected reactions inplatinum metal alloys have been successjidlly predicted by the Engel- Brewer relationships. Although the basic principles involved are still uncertain, interesting possibilities emerge and several possible applicationsfor the reactions suggested by this theory are proposed.

The practical metallurgist is rarely able to relating to I, 2 and 3 (s+p) electrons per make use of theoretical predictions and much atom. of the interest aroused by the Engel-Brewer Over the three transition series the bonding correlation stemmed from its claims of effectiveness of a d electron falls to a minimum direct utility in the design of transition metal when the number of unpaired d electrons alloys. Although Hume-Rothery (I) has arrives at a maximum of 5 per atom. The demonstrated the unsound foundations of magnitude of the d bond strength increases some parts of the theory most of the failures with progress from the first to the third he discussed were associated with the long period. metallurgy of copper, silver and gold. In The stability of the crystal structures as the transition element field considerable predicted from Brewer’s calculations is successes have been achieved and one is determined by the difference between the encouraged to consider the practical im- total bonding energy and the original pro- plications of these ideas and how these might motional energy assumed. Tungsten offers lead to the development of new commercially a good illustration of the methods of cal- viable materials. culation adopted. The ground state of tungsten is d4 s2, and the two s electrons Pure Metals would appear to indicate a close-packed In his early attempts (2) to correlate the hexagonal crystal structure. The pro- stable crystal structure of metals with the motional energy associated with the change electron configuration of free atoms Engel from the d5 s configuration is only 8 predicted that bond strength was determined kcal/mole, however. Since this slight ex- almost completely by the number of d penditure of energy releases six electrons electrons participating. This conclusion has for bonding the most stable structure, since been qualified by Brewer (3, 4) who associated with the single s electron is of now suggests that over the main range of the course b.c.c. transition elements the bonding energy of Such arguments are used to explain the the s and p electrons is greater than that of change from b.c.c. to c.p.h. structure on the d electrons. The type of crystal structure passing from tungsten and molybdenum to adopted is still, however, regarded as being elements of higher atomic number. If the determined by the s and p electrons alone, additional electrons entered the d orbital, the b.c.c., c.p.h., and f.c.c. structures cor- pairing would occur, the number of electrons

Platinum MetalsTev., 1969, 13, (2), 53-56 53 available for bonding would decrease, and simple electronic concentrations are trans- the stability of the structure would fall lated into configurations which satisfy the catastrophically. Such a fall is avoided by a Engel correlation. Conclusions based on change in electron configuration. Thus, if these translated values could in most in- the configuration d6 s is changed to d5 sp, all stances have been reached on the basis of five d-orbitals are made available for bonding simple electronic concentrations alone. Such and the close-packed hexagonal structure is correlations were considered by Robins (5) thereby stabilised. In the neighbourhood of who in 1959 suggested that a tendency tungsten the promotional energies associated exists for the number of bonding electrons with such changes in Configuration are per atom to equal half the co-ordination claimed to be negligible compared to the number. This tendency explains why improved bonding energies obtained. Move- tungsten and molybdenum, each presumed ments to the right of osmium and ruthenium to have six valency electrons and an effective involve a change from the d sp con- co-ordination number of 12 have the highest figuration to that of d" s p2, which by placing melting-points in their respective series. three electrons outside the d shell makes more The change in crystal structure on passing electrons available for bonding and stabilises from tungsten to rhenium helps to maintain the f.c.c. structure. this two-to-one ratio and thus stabilises the Increasing refinement in Brewer's treat- hexagonal structure. ment has shown that such integral con- figurations can only be approximate, and the Alloys optimum configurations required for the In the prediction of alloy-phase equilibrium most stable structures are non-integral. Thus Brewer (4) made considerable deviations the optimum configuration for ruthenium is from the integral electron concentration d 6.3 s p u.7 rather than the d6 s p con- values used for pure metals. In the copper, figuration originally proposed. silver and gold alloys the treatment was The promotional energies involved in such largely unsuccessful and it has not been configuration changes increase as the right- explained, for example, why the composition hand side of the three transition series are ranges of rs and p brass overlap. approached and this increase is used to The treatment of transition metal alloys explain the decrease in melting-point and has been rather more successful. The (stp) increase in vapour pressure with increasing electron concentration ranges associated with atomic weight. The assumption that rnelting- the b.c.c., c.p.h., f.c.c. phases are respectively points and vapour pressures provide a true 1.0-1.5, 1.7-2.12 and 2.5-3. Taking the index of crystalline stability is, of course, very basic electron configuration of molybdenum dubious, and the errors involved in an and tungsten as (4dS) (5s)' Hume-Rothery attempt to predict the melting-point of (I) has used these concentration ranges PO gallium from its vapour pressure would be calculate the solubility limits shown in the considerable. table in these two metals. Osmium and iridium have the highest Reasonable agreement has been obtained clastic moduli of any of the metals. They also particularly when the solvent and solute have the lowest compressibilities and it does are close together in the Periodic Table. The not seem likely therefore that bonding effect of temperature upon solubility does not processes deteriorate rapidly as soon as clearly emerge from the various publications tungsten is left behind. involved. This is obviously of great import- The main weakness of the configurational ance, and Selman (6)has recently shown that adjustments referred to above has been well the solubility of platinum in molybdenum exposed by Hume-Rothery (I). Thus the increases from approximately 2 atomic per

Platinum MetalsTev., 1969, 13, (2) 54 Molybdenum

cent at 1400°C to more than 10 atomic per necessary. If an excess of platinum is cent at 1700T. Although Brewer states employed the end product is a dilute solution that “the Engel correlations make thermo- of the transition element in platinum, and the dynamic predictions and must therefore be main function of the platinum is to keep the valid for the entire temperature range” (7), activity of the carbide-forming element at a it would be interesting to see a practical low level. Molybdenum and tungsten application of this philosophy to the cal- carbide are readily dissociated in this way culation of a temperature solubility curve. although there should, according to the Brewer-Engel correlation be no driving Stable Platinum Metal Compounds force for this reaction. Molybdenum carbide An important feature of the Engel-Brewer is by no means as stable as zirconium carbide. theory is the prediction that metals from Both these compounds are, however, far less opposite ends of the second and third transi- stable than alumina, which has been shown tion series should react strongly with each (9) to be readily dissociated under un- other to form compounds of high stability. oxidising conditions in the presence of This subject has been briefly referred to in a platinum, which takes the aluminium into previous article (8). solid solution at temperatures above 14oo3C. By reacting platinum with stable carbides Accurate thermodynamic measurements on Brewer (8) formed an estimate of the free zirconium-platinum alloys are obviously re- energies of formation of these compounds. quired. The electron negativities of the two Thus, platinum reacted with zirconium metals are very different and a fairly high heat carbide to produce the hexagonal compound of reaction would be suggested by this fact ZrPt,. Since zirconium carbide has a free alone (I). The transference of charge in such energy of formation of -- 47 kcal/mole, the a reaction would, however, be in the opposite free energy of formation of ZrPt, was pre- direction to that proposed by Brewer. sumed to be higher. It is difficult, however, to pursue such arguments to their logical Some Practical Applications conclusion. It is known that all the carbides Whatever the mechanism of the reactions can be dissociated by heating in contact with involved it seems that practical attention the platinum metals and in no instance does might well be devoted to the varied lines of the formation of a compound appear to be thought suggested by the theory. Exothermic

Platinum MetalsTev., 1969, 13, (2) 55 fuses are one possibility. Here the hafnium described by Brewer. By suitably propor- wire would be sheathed with an appropriate tioning the amount of palladium added to the thickness of palladium. Such a fuse wire carbide mixture improved oxidation resist- would withstand corrosive conditions for ance might be achieved with stronger chem- long periods without failure, but would, if ical bonding between the cement and the overheated disrupt explosively and provide carbide. Appreciable improvements might instantaneous protection for the circuit. result even if the palladium replaced only Applications might also exist in the field a small proportion of the cobalt binder (11). of age-hardening and dispersion-strengthened alloys. Thus small quantities of yttrium and References I W. Hume-Rothery, Progress in Materials palladium, dissolved stoichiometrically in Science, 1967, 13, (5) an inert metal matrix should react upon 2 N. Engel, Kem. Maanesblad, 1949, Nos 5, 6, suitable heat treatment to produce a finely 8J 9, I0 divided compound of considerable strengthen- 3 L. Brewer, Prediction of High Tcmperature Metallic Phase Diagrams, U.C.R.L. - 10701, ing power (10). UC-25 Metals, Ceramics and Materials, 1963 Interesting possibilities can be discerned 4 L. Brewer, High Strength Materials, Ed. V. F. Zackay, New York, John Wiley, 1965 in the sintered carbide field. Cobalt, the 5 D. A. Robins,J. Less-Common Metals, 1959, normal cement used for the liquid-phase 1,396 bonding of carbides is unfortunately not very 6 G. L. Selman, Platinum Metals Rev., 1967, 132 resistant to oxidation at high temperatures 11, 7 L. Brewer, Acta Metall., 1967, 15, 553 and this restricts the high temperature 8 A. S. Darling, Platinum Metals Rev., 1967, applications of such materials. Palladium or XI, 138 a palladium-based alloy would be more 9 A. S. Darling, Metals Muter. 1968, 2, (I), 28 resistant to oxidation and would also react 10 Johnson Matthey, Provisional Patent Appl. 29958/67,3520/33 chemically with the carbide in the way 11 Ikid., 12720/69

A Rhodium-Platinum Probe for Flame Velocity Measurement A probe for velocity vector measurements a stainless steel in flames consists of a spherical tip with one water - cooled probe pressure point on the axis and four others body. The stainless spaced equally around it at 40" to the central steel hypodermic ex- point. Pressure differences between opposite tension tubes and pairs of these holes, when referred to the water jacket then pass pressure at the central point, give gas flow through a cylindrical direction and velocity. Stainless steel probes brass block, carrying are suitable for cooled airflows or, when a 360" scale to indi- water-cooled, for large industrial furnaces, cate angle of rota- but in the relatively much smaller dimensions tion, to a cluster of of gas turbine combustion systems the appli- unions. This assem- cation of this technique depends on devising bly can be used in a method permitting the construction of flame temperatures compact probes in a manner suitable for up to 1800°C. use at kerosene-air flame temperatures. A technique of construction has recently been developed at the National Gas Turbine A complete water-cooled Establishment in which both the spherical probe unit in which the probe head and the five lengths of hypodermic rhodium-platinum head tubing are fabricated in 20 per cent rhodium- is nickel-brazed to the platinum, the whole thcn being brazed to stainless steel probe body

Platinum MetalsTev., 1969, 13, (2) 56 Ruthenium- Platinum Oxide Catalysts for Hydrogenation Reactions

A CRITICAL COMPARISON OF PUBLISHED RESULTS

By G. C. Bond, DSC., F.R.I.c., and Dennis E. Webster, Pm. Research Laboratories, Johnson Matthey & Co Limited

Since the publication of earlier papers from the Johnson Matthey Research Laboratories on the catalytic properties of mixed platinum group metal catalysts, two other groups have published their jindings. These are in certain respects in disagreement with each other and with the original work and results relating to the ruthenium-platinum system are critically discussed in this paper.

A technique for the preparation of intimate groups have used other solvents) we decided homogeneous binary mixtures of oxides of to check their observations. The purpose of the platinum group metals by a modification this paper is to give a brief summary of our of the Adams method has been described in a further work on the ruthenium-platinum previous publication (I). These mixed oxides system and to compare it critically with the are in the main easily reduced by hydrogen other published results. under ambient conditions to yield what we believe to be finely divided alloys of the two Surface Areas metals, and many of the catalysts produced It is unlikely that the reported differences in this way are much more active in a variety in catalytic properties are significantly due to of hydrogenations than either component differences in the surface areas of the starting oxide alone. Our previous publications oxides. Values of surface areas given by the describe results obtained with ruthenium- Engelhard workers (4) (their method is not platinum catalysts (I) and a number of other described) are in approximate agreement with mixed metals (2). ours obtained using low temperature nitrogen Two other groups of workers have since adsorption with a Perkin-Elmer-Shell Sorpto- reported their own work on these systems. meter (e.g. about 100 dg-1 for PtO,): The results published (3) by D. V. Sokol'skii Sokol'skii's values (3), obtained by low tem- and his associates at Alma Ata for the perature adsorption of air, are lower by a ruthenium-platinum system are in broad factor of about 2.5, but all groups agree there agreement with ours in the composition range is a tendency for the area of the oxide to fall which we originally covered, but they with increasing ruthenium content. In our reported a further substantial activity maxi- hands the surface area for a given composition mum (at about 90 per cent ruthenium) which is not closely reproducibIe from one prepara- we had not observed. The work of P. N. tion to another, the variation being much Rylander and his colleagues (4) of Engelhard greater than is the case with catalytic activity. Industries on this system is in complete and We therefore consider it inappropriate to quote utter disagreement with both ours and specific activities using areas of the oxides. Sokol'skii's, but because the conditions have Incidentally neithcr Rylander nor Sokol'skii not been identical (and in particular the other appear to have assayed their mixed oxides.

Platinum MetalsTev., 1969, 13, (2), 57-60 57 Now the Russian authors also report that NominalNominal andand ActualActual RutheniumRuthenium ContentsContents the second maximum in the reduction of (wt%)(wt%) ofof MixedMixed Ruthenium-PlatinumRuthenium-Platinum OxidesOxides o-nitrophenol is much more pronounced NominalNominal ActiinlActiinl when the solvent is aqueous decinormal KOH, 2.5 I .63 i.e., when the reactant is the o-nitrophenoxide 5 2.75 ion. In order to check this and to obtain i0 6.32 further information on the effect of pre- 20 15.2-16.9 reduction of the catalyst, we have studied the 50 43.8 reduction of o-nitrophenol in this solvent in the normal manner and also using catalyst pre- We find a substantial loss of ruthenium reduced by shaking in hydrogen under during the preparation, presumably as the ambient conditions before adding the re- volatile tetroxide. The figures given in the actant. The results, together with Sokol’skii’s, table are representative of our findings. are shown in Figure 2. With catalysts not pre- Reduction of Aromatic Nitro-compounds In our earlier publications (I, z), we demonstrated the existence of a very sharp activity maximum for the reduction of nitro- benzene in methanol at 25 atomic per cent ruthenium: a similar but less pronounced maximum was shown with o-nitroaniline. We have confirmed the generality of the maximum in this position by obtaining results for 2,4-dinitrotoluene in acetone/isopropanol (see Figure I): the rate at the maximum is about three times less than for nitrobenzene. Sokol’skii and his associates (3) report results for the reduction of o-nitrophenol in 96 per cent ethanol. These are also shown in Figure I, as are our own rates for the reduction of this compound in methanolic solution. It is clear that the reaction is enormously faster in methanol than in 96 per cent ethanol (by a factor of about 30): it is however known that water can be a powerful catalyst poison in the reduction of aromatic nitro-compounds. However, Sokol’skii observes a maximum at about 35 atomic per cent ruthenium, and, significantly, a second smaller maximum at about 80 atomic per cent ruthenium. As our catalysts are inactive when containing more than 85 atomic per cent ruthenium because they are not reduced under our experimental conditions, we infer that Sokol’skii’s catalysts are pre-reduced in some manner, although his publications give no details.

Platinum MetalsTev., 1969, 13, (2) 58 reduced we find a broad maximum between 10 and 50 atomic per cent ruthenium with a slight dip which may or may not be genuine. Pre-reduction lowers the activity slightly in this region but certainly causes ruthenium- rich catalysts to be active, although our second maximum is far less marked than is Sokol’skii’s. As the solvents are identical in this case, differences can only be caused by differences in catalyst preparation or experi- mental procedure. We turn now to examine the results of the Engelhard workers. Their results for the reduction of nitrobenzene in glacial acetic acid are so much at variance with ours obtained using methanol as solvent, that we

decided to repeat our work using their solvent. These results are shown in Figure 3. Unfor- tunately it is not possible to make a direct comparison of rates even for PtO, as the Engelhard papers (4), do not mention the weight of catalyst used: however it would appear from our results that theirs refer to 10 mg of catalyst and the line through their points is shown in Figure 3 based on this assumption. The disagreement, which is profound and inexplicable, is not commented on by Rylander. Glacial acetic acid is anyhow an unsatisfactory solvent, the maximum rate of nitrobenzene hydrogenation in it being some three times less than when methanol is the solvent (see also Figure 3). It is abundantly clear that rates of hydro- genation with this catalyst are very much

Platinum MetalsTev., 1969, 13, (2) 59 dependent on the kind of solvent used, but group metal catalysts. Solvent effects are other factors must be operating: their nature partly responsible, but the precise method of will only be revealed by further work. preparing the mixed oxides and differences in experimental procedure are other possible fac- Other Reactions tors which may account for the discrepancies, Sokol'skii and his associates report (3) for Ruthenium-platinum oxide catalysts, their the hydrogenation of olefins (cyclohexene and preparation and use, are covered by British Patent Nos 1,099,406 and 1,016,058 and U.S. I-heptene) using ruthenium-platinum cata- Patent No. 3,305,402, and are the subject of lysts that maximum rates occur at 45 and 88 pending patent applications in a number of other atomic per cent ruthenium. We originally countries. References showed the existence of a maximum rate for References the hydrogenation of maleic acid in methanol I G. C. Bond and D. E. Webster, Platinum Metals Rev., r965, 9, 12 at about 30 atomic per cent ruthenium and 2 G. C. Bond and D. E. Webster, Platinum have since found a maximum at about the Metals Rev., 1966,10,10 Same concentration in glacial acetic acid. 3 D. V. Sokol'skii, K. K. Dzhardamalieva, A. G. Sarmurzina and T. Tonmanov, Dukl. Akad. One Of the interesting features Of our Nauk S.S.S.R., 1967, 176, 1093; D. V. Sokol'skii, A. Sarmurzina and K. earlier work was that, with acetylenic sub- Sokol'skii, A. G. Sannurzlna and K. K. Dzhardamalieva, Tr. Inst. Khim. Nauk, Akad. stances such as methYlbutYnol, maximum Nauk Kazakh. S.S.S.R., 1967, 17, IOO rates occurred with very low Concentrations 4 P. N. Rylander, L. Hasbrouck, S. G. Hindin, I.1- Karpenko,KaQefiO~ G. Pond and S.S- Starrick, Engel- of ruthenium: the Engelhard group report hard Ind. Tech. Bull., 1967, 8, 25; P. N. a similar observation (almost the only Rylander, L. Hasbrouck, S. G. Hindin, R. (4) Iverson, I. Karpenko and G. Pond, Engelhard point of agreement between us) but their Ind.r$;hf-Btl;;fp:g, Tech, Bull., 1967, z:p8, 93 Pond, maximum rate is only about 20 per cent greater than for PtO, whereas we found a maximum rate about 2.2 times greater. Complexes of Substituted Other Catalyst Systems olefins with Salts ofthe Platinum Metals Comparison of our results with those of the Engelhard group for rhodium-platinum cata- One Of the small but parts Of demonstrates organometallic chemistry concerns the com- bts again the existence Of a plexes formed between salts of the transition large solvent effect. With this system glacial metals-more particularly those of the plati- acetic acid is a much better solvent than num metals-and substituted olefins. The methanol for nitrobenzene hvdrogenation:-- discovery of this class of compound has not with methanol we find rates which vary Only advanced OUT knowledge Of chemistry and of chemical bonding, but has stimulated somewhat erratically with catalyst composi- our understanding of catalytic processes both tion, the maximum rate (at 39 atomic Per cent in solution and at solid surfaces. A recent rhodium) being only about 50 per cent greater review by Richard Jones of the ICI Petro- than for PtO,. However, with glacial acetic chemical and Polymer Laboratory (Chem. acid as solvent the maximum rate occurs at Rev., 1968, 68, 785) provides a thorough survey of this field. Among the type of the same composition but is four times substituents discussed are halogens, where greater than for PtO,. Rylander reports the perfluoro-oleiins behave in some respects factor to be about 2.5. Again we cannot com- quite differently from simple olefins, and pare absolute rates for the reasons given. amine and cyanide groups, both of which may also chelate to the central metal atom. Com- Conclusions plexes of this type provide an increased range of stabilities by comparison with their un- Substantial and inexplicable differences substituted analogues, and may well find in- are shown between the results of the various dustrial uses in the tailor-making of catalysts groups which have worked on mixed platinum of specific activity.

Platinum MetalsTev., 1969, 13, (2) 60 demonstrates Making Corrosion-Resistant Brazed Joints in Stainless Steel ADVANTAGES OF PALLADIUM-BEARING BRAZING ALLOYS

By M. H. Sloboda, Dipl. Ing. Research Laboratories, Johnson Matthey & Co Limited

Stainless steel parts brazed with standard low-melting brazing alloys are susceptible to a specijic kind of corrosion by water or humid atmosphere, generally known as interface or crevice corrosion. One of the advantages of brazing alloys of the silver-copper-palladium type is that they will produce joints in stainless steel resistant to this form of failure.

Joints made in stainless steel with certain fast. Cases are known of stainless steel low-melting brazing alloys (of the silver- parts brazed with unsuitable materials falling copper-zinc-cadmium type) and exposed to apart after a week’s exposure to the action the action of water or humid atmosphere are of ordinary tap water. susceptible to corrosion of a rather unusual Another distinguishing feature of the corro- kind. sive attack of this kind, which is usually The attack in this case consists in the disso- referred to as interface or crevice corrosion, lution of a very thin steel layer along the joint is that neithcr the brazing alloy nor the stain- interface, as a result of which the bond be- less steel show any obvious symptoms of tween the steel and the brazing alloy is com- corrosion damagc. Although iron dissolved pletely destroyed. A failure of this kind is at the joint interface is precipitated as a loose illustrated in Fig. I, showing a laboratory red rust deposit outside the joint area, once stainless steel joint specimen in an advanced this corrosion product has been removed (by stage of corrosion. design or by accidcnt), there is nothing to Since the quantity of metal that has to be indicate that the separation of the brazing dissolved to produce this effect is extremely alloy from the steel was caused by corrosion. small, the rate of bond destruction is very In fact, the usual reaction of an observer

Fig. 1 A ferritic stainless steel (EN 60) specimen brazed with a 50Ag-lSCu- 16Zn-19Cd alloy and held for eight days in running tap water. The specimen was bent to reveal the extent of interface corrosion result- ing in the destruction of the bond between the brazing alloy and the steel

Platinum MetalsTev., 1969, 13, (2), 61-64 61 +-Brazing alloy

+Stainless steel

Fig. 2 Showing a blister formed in an austenitic steel (EN 58e) specimen brazed with a 54Ag-25Pd-21Cu alloy in dissociated ammonia under a flux rover and held fur 106 days in running tap ioater. ( x 50)

+-Sleel/solder interface +-Partially uxidised steel layer

Fig. 3 A magniJied view of a region of the specinten uf Fig. 2 shoat ing the presence uf a partially oxidised steel layer at a certain distance from the joint interface along a plane coinciding with the path of the crach whose formation led to blistering. ( % 550) faced with a failure of this kind is to con- brazing alloys with which joints resistant to clude that the joint fell apart because no bond interface corrosion can be made, each of these formation had actually taken place. materials has certain limitations. And so, The mechanism of interface corrosion, most alloys that are convenient to use from which now appears to be a much more com- the purely brazing point of view will produce plex phenomenon than first thought, has not corrosion-resistant joints but only when used yet been elucidated. However, extensive on stainless steel of the nickel-bearing austen- researches in this field have made it possible itic type. On the other hand, low temperature to classify various standard and specially brazing alloys that are suitable (from the developed brazing alloys in the order of their standpoint of interface corrosion resistance) usefulness for joining stainless steel when for joining both austenitic and ferritic stain- corrosion resistznce of the joints produced is less steels are often distinguished by some- an important consideration. what inferior flow characteristics. Finally, the It has been established for instance that, corrosion resistance of stainless steel parts although there are several low-melting silver brazed with low-melting alloys of this kind

Platinum MetalsTev., 1969, 13, (2) 62 may often be substantially affected by certain stainless steel parts is too low to make the changes in the composition of fluxes used steel surface wettable by the brazing alloy. during brazing. The usual remedy in such cases is to use The only known materials that combine the both a reducing atmosphere and a small freedom from the above limitations with quantity of a suitable brazing flux. However, many of the desirable properties of various it has been recently found that this practice silver solders are certain gold-base and may have unexpected and rather puzzling palladium-bearing alloys. The latter are consequences. exemplified by the 54Ag-z5Pd-z1Cu com- If an alloy is intrinsically capable of pro- position which is one of the Johnson Matthey ducing corrosion-resistant joints, it makes no noble metal brazing alloys known as Palla- difference whether the brazing is done in a braze 950; its special advantages as a filler reducing atmosphere or under a flux cover. material for joining stainless steel can be When, however, both a reducing atmosphere summarised as follows : and a flux are used in brazing with such an (i>Pallabraze 950 will produce interface- alloy (e.g. Pallabraze 950)~the resultant joints corrcsion resistant joints in both ferritic -while resistant to interface corrosion-may and austenitic stainless steels. be subject to failure of a different kind. (ii) The corrosion resistance of joints made When observed under laboratory condi- with this alloy is not affected by changes tions on specimens tested in running tap in the composition of the brazing flux, water, the failure consisted in the formation of especially by the presence of free boron which is sometimes added to brazing fluxes to improve their flow-promoting properties and stability at elevated temperatures. (iii) Joints made with Pallabraze 950 com- bine the resistance to interface corrosion with a relatively high strength at tem- peratures of up to 400°C. (iv) Stainless steel parts brazed with this alloy are not damaged by erosion due to excessive interalloying during brazing. (v) Due to its palladium content, Pallabraze 950 may be safely used when inter- granular penetration of steel by a molten brazing alloy might otherwise lead to brazing failures. (vi) Pallabraze 950 has excellent wetting and free-flowing characteristics and can be used both for brazing in air under a flux cover and for fluxless brazing in a reducing atmosphere. In this last connection a rather unusual Fig. 4 Showing the absence of any obvious differences in the structure of effect is worth reporting. joints brazed in an austenitic stainless It is sometimes found in industrial practice steel (ELN58e) with a 54Ag-25Pd-21cu that the oxide-reducing potential of a brazing ulloy: (top) in dissociated ammonia; (middle) under u flux cover in air, and furnace atmosphere which is sufficiently pure (bottom)in thepresence of both a reducing (dry) to preserve the initial bright finish of atmosphere and a brazing flux. ( Y 200)

Platinum MetalsTev., 1969, 13, (2) 63 a crack inside the joint. The crack was propa- the as-brazed condition revealed no differ- gated in the steel itself along a plane parallel ences between joints made with the same to, and at a distance of several microns from, alloy under a flux cover, in a reducing atmo- the joint interFace; it was often associated with sphere, and in the presence of both these the presence of a partially oxidised steel layer. fluxing media (Fig. 4). The crack plane approximately marked the Further studies will be necessary to estab- boundary of the zone into which palladium lish the precise conditions leading to joint and copper had diffused into the steel during failure of the kind described above and to brazing. The only external evidence of the ascertain that the effects observed were not development of this failure was the formation due to some extraneous, as yet unidentified of blisters in regions where the brazing alloy factors. Nevertheless, the evidence available coating was sufficiently thin. Photomicro- so far is sufficient tu advise against using a graphs of a specimen that had failed in this brazing flux in combination with a reducing way are rcproduced in Figs. 2 and 3. atmosphere in the joining of stainless steel It should be added in conclusion that parts that may be exposed to the influence of metallographic examination of specimens in water or humid atmosphere in service.

Thermocouples Under Neutron Bombardment CHANGES IN VOLTJME AND COMPOSITION Under severe neutron radiation thermo- mutation effects being confined to the couples are known to be unstable although platinum. The palladium content of palla- little data on the changes in composition which dium-platinum alloys is not greatly affected occur has hitherto been published. A recent although small quantities of mercury, gold, report by C. B. T. Braunton, D. N. Hall and cadmium, silver, iridium and rhodium are C. M. Ryall, of the Atomic Energy Research formed by transmutation. Molybdenum- Establishment, Harwell, (U.K.A.E.A., platinum alloys are considerably more stable A.E.R.E. - R5837, 1968)~now provides in- than rhodium-platinum alloys when irradiated formation which will greatly simplify the being comparable to ruthenium-platinum selection of thermocouple materials for ex- alloys in this respect. periments carried out under radiation at high Tungsten, rhenium and tantalum all suffer temperatures. Under such conditions lattice severely in a neutron flux. Some 40 per cent damage anneals out and changes of thermo- of the rhenium is lost after one year at a flux electric behaviour can be predicted from the of z x ~o'~n/cm~/sec;tungsten is rapidly compositional changes. These have been replaced by rhenium, rhenium by osmium, computed from differential equations which and tantalum by tungsten. The compositional were developed to describe the exponential changes in molybdenum are small. transmutation of isotopes, and curves defining Platinum alloys increase in volume under the composition of the various alloys at suc- neutron irradiation by amounts ranging up cessive stages of time and radiation are pro- to z per cent after one year at Io1%/cm2/sec. vided in this report. Tungsten, tantalum and rhenium decrease in Pure platinum when irradiated produces volume to a considerably greater extent. only gold and mercury, each reaching a A significant conclusion of this report is maximum concentration of rather less than that tungsten-rhenium thermocouples will I per cent after a year's exposure to a flux of suffer more damage, and are likely to be less ~o~~n/cm~/sec.Under similar conditions, stable than platinum-based thermocouples however, the rhodium content of rhodium- under similar conditions of neutron bombard- platinum alloys is almost- completely con- ment. It will be interesting to see whether sumed, to form palladium, mercury, gold and the results of the experimental programme iridium. now being carried out in mixed thermal and Ruthenium alloys are far more stable. The fast neutron fluxes, referred to in this report, ruthenium remains unchanged, all trans- confirm this prediction. A. S. D.

Platinum MetalsTev., 1969, 13, (2) 64 The Platinum Metals in the Measurement of Temperature THE NEW INTERNATIONAL TEMPERATURE SCALE OF 1968 By C. R. Barber, BSC., F.1nst.P. Quantum Metrology Division, National Physical Laboratory, Teddington

A revised definition of the temperature scale the extension is made in terms of the platinum entitled the International Practical Tempera- resistance thermometer which is now the ture Scale of 1968 (IPTS-68) was introduced interpolation instrument of the Scale over the by the Comite International des Poids et wide range from -259.34”C(13.81K) to Mesures (CIPM) in October 1968 to replace 630.74T. The platinum to be used in the the IPTS-48. The new Scale was put into thermometer is required to be very pure; the operation at the National Physical Laboratory completed thermometer must have a value of on January Ist, 1969, and temperature values R(Ioo’C) the resistance ratio not less than quoted in documents issued by the Labora- R(o”C) tory from that date will be changed by the 1.39250. This value corresponds to 1.3920 on appropriate amounts. It is necessary to make the old scale. these changes throughout science and in- To conform with common usage the Scale dustry, but as will be seen below the changes is expressed in Kelvin temperatures below for many purposes will not be very significant. 273.15K and in Celsius temperatures above. Platinum plays an important role in the new From 13.81K to z73.15K the resistance of Scale for the construction of the platinum : the thermometer is measured at a number of 10per cent rhodium-platinum thermocouple fixed points and then the calibration is and the platinum resistance thermometer obtained by difference from the reference which are the chosen standard instruments function for the realisation of the Scale below the Te8=[A0+ ZiEI Ai(lnW)i]K freezing point of gold. R(T) The difference in numerical values of tem- where W = The coefficients perature introduced by the IPTS-68 are R(273.15)’ shown in Table I. They result from the A0 A,, are specified in the text of the adjustments made to the IPTS-48 to give Scale. The function has been evaluated and agreement with thermodynamic values of is available in the form of a detailed table. temperature obtained from gas-thermometer The differences from the function, AW, are measurements. These adjustments consist of expressed by a polynomial for each of the assigning the best available values of thermo- ranges 13.81K to zo.z8K, zo.zSK to 54.361K, dynamic temperature to the defining fixed 54.361K to 90.188K and 90.18SK to z73.15K. points and modifications to the interpolation The last mentioned range was formerly procedures in the parts of the Scale where the defined by the Callendar-Van Dusen equation platinum resistance thermometer is the speci- but we now use the differcnce from the refer- fied interpolation instrument. ence function given by the equation. The IPTS-68 extends down to 13.81K, Aw=A 1 B (*-I) the triple point of equilibrium hydrogen, I00 c IOO‘C whereas the IPTS-48 stopped short at the Where t,, T,,-z73.15K and the constants boiling point of oxygen (TG8(02)=go.188K); A and B are determined by measurements of

Platinum MetalsTev., 1969, 13, (2), 65-67 65 Table I Approximate Differences between 1PTS-68 and IPTS-48 in- kelvins I t.,'C 0 -I0 -20 - 30 - 40 - 50 -80 I I I -ioo 0022 0013 o 003 -0.006 --0.013 -0.013 -0.005 0.007 0.012 (O.O( at -k 0% P It: - o o ooo 0.006 o 012 -0.018 0.024 -0.029 -'2.033 0.029 -0.022 t,"C 0 I0 20 -30 40 -50 60 70 80 90 i00

0 0000 -0004 0007 -0.009 -0.OIO 0.010 -0.010 -0.008 -0.006 - 0.003 0.000 IOO 0000 0004 0007 0.012 0.016 0.020 0.025 0.029 0.034 0.038 0.043 200 0041 0047 0054 0.054 0.058 0.061 0.064 0.067 0.069 0.071 0,073 300 0073 0074 OOj5 0.076 0.077 0.077 0.077 0.077 0.077 0.076 0.076 400 0076 0075 0075 0.075 0.074 0.074 0.074 0.075 0.076 0.077 0.079 500 o 079 0082 0085 o.oS9 0.094 0.IOC 0.108 0.116 0.126 3.137 0.1~0 600 o 150 o 165 o 182 c.200 0.23 0.25 0.28 0.31 0.34 0.36 0.39 700 0 39 0 42 045 0.47 0.50 0.53 0.56 0.58 0.61 0.64 0.67 800 067 0 70 072 0.75 0.78 0.81 0.84 0.87 0.89 0.92 0.95 900 095 09s 101 I .04 1.07 1.10 1.12 1.15 1.18 1.21 1.24 xooo I 24 127 I 30 1-33 1.36 1.39 1.42 1.44 - - - -

tcrT 0 100 200 600 700 800 900 1000

I000 1.5 17 1.8 1.0 2.2 2.4 2.6 2.8 3.0 3.2 2000 32 35 3 7 4.0 4.2 4.5 4.8 5.0 5.3 5.6 5.9 3000 5 9 6 2 6 5 -6.9 7.2 -7.5 7.9 a.1 -8.6 9.0 Y .3

W at 10o-C and -182.962"C (90.188K). Thus the experimental procedure for calibra- tion in this range of temperature is simpler than formerly, requiring calibration at three It is clear that the coefficient cc for a particular points only instead of four, since it is not specimen of platinum will be unchanged on dependent on measurements at the zinc the new Scale but the 6 coefficient will be point. different because of the change in the value The Callendar equation is still used for the assigned to zinc point. For example a 6 co- range from o'C to 630.74 C but a correction efficient of 1.492 on the old scale becomes term is added, so that the calibration pro- about 1.497 on the new. cedure is to measure the resistance of the Over the range from 630.74T to the gold thermometer at the ocC point (obtained by point (1064.43"C) the platinum : 10per cent way of the triple point of water), the boiling rhodium-platinum thermocouple is used to point of water and the freezing point (100°C) define the Scale as in the IPTS-48, but now of zinc (419.58"C). It will be observed that the values of the fixed points are modified. only the temperature assigned to the freezing The emf temperature relation is point of zinc has been changed, being 419.50S"c on the 1948 scale. The Callendar E a ~ btss ' ct,,' equation is then used to determine an inter- where the constants a, b and c are deter- mediate value of tl mined from the measurements made at

t' t' 630.74'C 50.2'C, 961.93"C (F.P. of silver) t'='/, (W--I) -; 6 (=p)IO0"c and 106443°C (F.P. of gold). Now tl is corrected by an amount which Above the gold point the IPTS-68 is varies with temperature but is the same for defined as before by the Planck law of radia- all thermometers which meet the specification tion using the freezing point of gold as the for the Scale reference temperature and a new value of c2,

Platinum MetalsTev., 1969, 13, (2) 66 the second radiation constant, of 0.0143SS Table II metrc kelvin. The numerical values of tem- New Freezing Points of Metals perature in this range show a net increase over the 1948 values, the higher value of the gold IPTS-68 IPTS-48 point causing an increasc and the higher Silver 961.9332 960.8”C I value of c2 a decrease. I The values of the freezing points of the Gold 1064.43 1063 1 metals relevant to this discussion and given Palladium 1554 1552 1 either as defining fixed points of the tempera- Platinum 1772 I769 ture scale or in the list of secondary fixed Rhodium 1963 1960 I points are set out in Table I1 with the corre- Iridium 2447 2443 I sponding IPTS-48 values for comparison, The freezing points of silver and gold are defining fixed points of the Scale and there- perature which is defined as “the fraction fore the temperatures assigned are the best 1/273.16 of the temperature of the triple available values of thermodynamic tempera- point of water”. It is appropriate to refer to a ture. The other values in Table I1 are ob- temperature difference in kelvins even when tained from the IPTS-48 values corrected the temperatures are expressed in degrees according to the differences in Table I and Celsius, and this practice has been followed then rounded to the nearest kelvin. in the text of the Scale. The IPTS-68 embodies the decision of the The full text of the English version of the IPTS CIPM to discontinue the use of “degree is obtainable from Her Majesty’s Stationery Ofice. The official French text is obtainable from kelvin, symbol “K” in favour of the “kelvin” the International Bureau of Weights & Measures, to describe the unit of thermodynamic tem- F92, Skvres, France.

Palladium-Titanium Alloy in Chemical Plant The natural resistancc of titanium to concentrated chloride solutions. His results corrosion in oxidising conditions has been show that 0.I per cent palladium is inadequate extended to reducing conditions by the addi- to give titanium protection in reducing con- tion of small amounts of palladium, and the ditions but that 0.5 and 1.0 per cent alloys wider use of titanium in chemical plant which both give sufficient protection. this permits has been the subject of comment He warns that pilot tests to determine the in this journal, most recently on the work of most suitable palladium content are necessary Takamura in Japan (I). in extreme conditions, where 0.2 per cent is Whereas Takamura used the 0.13 per cent insufficient. However, short-term tests may palladium-titanium alloy in hot concentrated be inadequate as time is needed for the forma- chloride solutions, W. R. Fischer, of Friedrich tion of the palladium-rich protective film. Krupp, Essen, has now shown that the Krupp The weldability of the alloys is comparable standard alloy containing 0.2 per cent to that of ordinary titanium but the usual pre- palladium is most generally useful in extended cautions are necessary. Pure dry argon should tests with a number of chloride solutions (2). be used to prevent access of oxygen and Adequate protection against crevice corrosion nitrogen to the weld and iron and othcr and pitting was obtained. However, in ex- metallic impurities must be excluded. ceptional cases an alloy with more than 0.2 Fischer adds that a special palladium- per cent palladium may be necessary where tantalum-titanium alloy has yet higher cor- particularly awkward angles occur in fabri- rosion resistance, especially in non-oxidising cated chemical plant. acids, and is available for special purposes. Fischer measured weight losses and poten- F. J. S. of 0.1, tials palladium-titanium alloys with 0, I Platirium Metals Rev., 1968, 12, (2), 53 0.5, and 1.0 per cent palladium in hot sul- 2 W. R. Fischer, Tech. Mitt. Krupp Werksher., phuric acid and hydrochloric acids and in 1969, 27, (11, 19

Platinum MetalsTev., 1969, 13, (2) 67 Sir William Crookes INVESTIGATIONS ON IRIDIUM CRUCIBLES AND THE VOLATILITY OF THE PLATINUM METALS

By J. c. Chaston, PLD., A.R.S.M. Johnson Matthey & Co Limited

William Crookes was born on June 17th, important technical and scientific publica- 1832, and after attending courses in chemistry tions throughout the world and, through its at the College of Science, remained for three correspondence pages, a forum for many years as an assistant under Professor Hoffman. notable - and often amusing - discussions. He then taught for nearly a year as lecturer Only fifteen months later Crookes had a in chemistry at Chester Training College, piece of good fortune when he was able to and was known as “Professor” ever after- announce on March 3oth, 1861, the dis- wards. He had some private means, and in covery of a new element which he later 1858 came to London and set up as a con- christened Thallium. The discovery is sultant with a small private laboratory at his often described as the result of a spectroscopic home at 20 Mornington Road, near Regent’s examination. In fact, Crookes very simply - Park. though no less to his credit - observed an In December 1859, he founded the weekly unexpected, and hitherto unreported, bright journal The Chemical News and remained as green flash in the course of a flame test when editor for the next sixty years. T?ze Chemical a Small sample of the seleniferous residues News prospered from the first and was widely from sulphuric acid manufacture at Tilkerode, read. It provided excellent summaries of in the Hartig mountains, was supported on a

Sir William Crookes, O.M., F.R.S. 1832-1919 President ofthe Royal Society from 1913 to 1916, Sir William Crookes did fi-fty years ago this month at the age qf 87. Most of his working lifr spanned an era in which a man could take very nearly the whole of science as his province. It was thus almost inevitable that the platinum metals should have been included among his many interests, and his obserrations on the loss of weight of the platinum group metals when heated in air have indeed been quoted as authori- tative for nearly haEfa century

Platinum MetalsTev., 1969, 13, (2), 68-72 68 In this detail from a painting by Il. Jumyn Brooks at the Royal Institution the front row shozos, lefi to right, Mr A. J. BaEfour (then Prime Minister), Sir William Crookcs. Crichton Brozme, Lord Ruyleigh, George Matthey, Ludwig Mond and Oliver Lodge. The occasion uws the demonstration by Dewar in 1904 of the properties of liquid hydrogen platinum wire in a gas flame. The material developed Crookes glass to protect the eye was left over from earlier work on sulpho- against the ultra-violet. In his later years, selenides in which he had engaged at the like Sir Oliver Lodge, he explored the College of Science. Crookes was a good mysteries of spiritualism. Lord Kelvin is experimenter, and he followed up his initial said to have remarked of him that “Crookes observation by isolating the new element and started more absolutely new hares than any investigating its properties. A sample was other man among his scientific contempor- inevitably shown at the Great Exhibition of aries”. 1862 and recognition very quickly followed All his work was done in his private (too quickly, some declared) by his election laboratory in his house, first in Mornington as a fellow of the Royal Society in Novem- Road and after 1880 at 7 Kensington Park ber 1863. Gardens (which he claimed to be the first in His determination of the atomic weight England to be lighted by electricity) and of thallium, reported in 1873, was a first-class much of it was financed by grants from the piece of work and the figure he obtained, Royal Society. In his later years he was 203.642, equivalent to 204.03 when corrected involved in a variety of commercial enter- to modern values, compares very favourably prises. with the 1961 International value of 204.37. He was prominent in every area of Scientific His flair for careful observation continued activity in London for nearly sixty years and to serve him well. He devised the radio- was widely honoured. He was elected meter - though his explanation of its action President of the Institution of Electrical was soon discredited - he described the Engineers in 1890, of the British Association Crookes dark space in the cathode discharge in 1898, and of the Royal Society in 1913. tube - but was less than profound in his He was awarded the Order of Merit in 1910. theoretical views on electrical discharges No scientific meeting of note failed to wel- through gases - he separated Uranium X, come him, and in scores of paintings and he devised the spintharoscope, and he photographs of scientific gatherings of the

Platinum MetalsTev., 1969, 13, (2) 69 period his distinguished bearded figure is Johnson and Matthcy” and one of thcse has prominent in the foregound. been preserved to this day. Crookes noted The early issues of The Chemical News that the iridium crucibles resisted fusion of devoted a remarkable amount of space to the many fluxes, including caustic soda, and were platinum metals. The first three issues not attacked by molten lead, zinc, nickel, serialised a translation of a classic paper by gold and iron. Temperatures are not recorded. Deville and Debray, “On platinum and its One paragraph introduces without comment a associated metals”, which includes a des- quite unexpected characteristic. “Copper cription of a long-forgotten use of the lime melted in the crucible for some time”, furnace for the fire-refining of platinum. The Crookes wrote, makes it “hot rotten, i.e. it is base metals are slagged off, gold and brittle when hot. But if the copper is well palladium vaporised and collected in an burnt off at a high heat the iridium reverts earthenware tube, and osmic acid finally to its original condition”. No explanation is trapped in a vessel full of ammonia. Part of the given as to how copper could be “burnt off’. osmium is deposited on the tube walls in the One other passage reveals a common mis- metallic state. Later issues included a reprint conception. “Heated for some hours over of a paper by Woolcott Gibbs, describing a Bunsen burner insufficiently supplied with the treatment of osmiridium residues from air the iridium crucible is unaffected and Californian gold and from Siberia, and a the deposit of carbon easily burns away, translation of Claus’ classic paper “On leaving the surface of the metal uninjured. certain properties of the platinum metals, All chemists know how seriously a platinum particularly ruthenium”. crucible is attacked in these circumstances”. It was not until 1908 that Crookes pub- In fact, a clean platinum crucible is quite lished any original observations on the unattacked, though it may be disastrously platinum metals, but in a communicacion embrittled if it contains certain compounds read in May of that year to the Royal Society (particularly arsenates or phosphates) which he was the first to “draw the attention of are reduced in the yellow flame and alloy chemists to the great advantages of using with the platinum to form brittle phases. crucibles of pure iridium instead of platinum Many chemists even today attribute the in laboratory work”. trouble to attack on platinum by carbon. The crucibles were supplied by “Messrs These details apart, in advocating the use

One of the iridium crucibles made for Crookes in 1908 by .fohnson Matthey

Platinum MetalsTev., 1969, 13, (2) 70 The jirst page of Crookes’ paper on “The Volatility of Metds of the Platinum Group” read to the Royal Society in March 1912, showing the platinum-wound furnace in which his inuestiga- lion iias carried out

of iridium (and rhodium) crucibles for certain formal hexagonal plates with a brilliant chemical operations, Sir William Crookes metallic lustre and perfect crystalline form”. was far in advance of his times. It is only within the last decade that any serious application has been made of iridium cruci- bles, although they are now employed widely for growing laser crystals and also for melting special glasses. The measurements reported by Crookes to the Royal Society in March 1912 of the loss of weight of the platinum metals when heated in air are better known. They stemmed from his observation of the “fine glistening dust” deposits of platinum on the porcelain tube of a furnace wound with platinum ribbon. The furnace had failed through localised thinning of the winding after operating at 1500°C, and under the micro- scope the deposits appeared as “beautifully

Crookes’ results for the losses in weight of the platinum metals when heated in air at 1300°C. It will be seen that the loss is expressed as a percentage of the weight, not of the unit area of surface, in spite qf the diffprpnres in size and form of the specimens

Platinum MetalsTev., 1969, 13, (2) 71 The weight losses of palladium and iridium in air at 900°C. No loss was recorded for platinum or rhodium at this ternperdure

Crookes accordingly set out to determine the loss of weight of the various platinum metals, using as his specimens a platinum crucible weighing 150 grains, an iridium crucible “as hard as steel” of about the same weight, a polished plate of palladium weighing 58 grains, a rhodium specimen (pre- sumably as sheet) weighing 32 grains, and a “highly polished plate” of ruthenium. The platinum at the higher temperature almost weight and dimensions of the ruthenium are certainly surprised Crookes, but he rejected not specified; and it is of interest to reflect the possibility that these could be due to a that there must be few, even today, who have volatile oxide on the evidence that the seen highly polished ruthenium sheet. The crystals deposited on the walls of a platinum- choice of grains instead of grammes as units wound furnace were metallic platinum. The of weight is explained in a footnote: “I have concept of oxygen acting as a carrier to used my standard platinum grain weights transport platinum from a hotter source for nearly fifty years, and they are too valuable (where platinum oxide concentration is high) to discard in favour of gramme weights, which to a cooler surface (yet above the decompo- would demand many months’ work on them sition temperature of the oxide) was of course to bring them to the state of accuracy into not available to him. which I have now got the grain weights”. It was a little unfortunate, perhaps, that In devising the arrangements for heating, the upper temperature chosen was 13ooOC. Crookes was concerned to avoid the possibility At this temperature it is true that the vapour as he said of “the action known as ‘air washing’ pressure of PtO, in air is rather higher than (particles from a white-hot semi-molten that of Rho, so that rhodium surfaces lose surface being mechanically carried away weight less than platinum ones. This led to by a current of impinging air)”. Today it is a belief that at high temperatures in general not clear why the surfaces should have been rhodium is the less volatile in air; actually thought to be “semi-molten” at 13oo0C, the above 1300°C the vapour pressure of Rho, highest temperature used. Accordingly, the increases very rapidly and soon greatly ex- heating was carried out in a vertical resistance- ceeds that of PtO,. heated tube furnace, closed at the top and With iridium, rhodium and ruthenium it “nearly closed at the place where the crucible was obvious visually that oxidation occurs and stood and almost completely obstructed at Crookes endcavoured to ascertain whether the lower end”. In this way, “the platinum iridium would volatilise in a vacuum at was in almost still air”. 1300°C. Experimental difficulties, however, The experiments were made at 130oCC prevented him from reaching a definite and 900”C, the samples being removed for conclusion. weighing every z hours, the total time of Crookes was certainly among the first to heating being 30 hours. The results are study the behaviour of the platinum metals shown in Figs. 9 and 10 of his paper, re- when heated in air to high temperatures, produced here, and have been widely quoted though in this as in other directions he was ever since. The relatively high losses of an experimenter rather than a philosopher.

Platinum MetalsTev., 1969, 13, (2) 72 ABSTRACTS of current literature on the platinum metals and their alloys

PROPERTIES iine field at 6oConuclei in Co-Pt as a function of applied field suggest the formation of a spin- Fusion Curves for Graphite, Tungsten and compensated state at low temperatures. Platinum up to 60 kbar L. F. VERESHCHAGIN and N. S. FATEEVA, Zh. eksper. Effect of Chemical Composition and Heat teor. Fiz., 1968, 55, (4), 1145-1152 Treatment Conditions on the Magnetic Optical measurements on Pt agree well with those Properties of High Coercivity Cobalt-Platinum obtained by thermal analysis. The temperature Alloys of fusion of Pt increases linearly with pressure. YU. A. GRATSIANOV and A. G. RABIN'KIN, Sb. Tr. Tsent. Nauch.-Issled. Inst. Chern. Met., 1968, (54), Rheological Properties of Platinum at High 90-97 Temperatures Tests on cast and homogenised samples of 24,25, v. 0. SHESTOPAL, Fiz. Metal. Metalloved., 1968, 26.16, and 29.5 wt?;, Co-Pt alloys showed that 26, (6), 1127-1130 the highest magnetic properties are attained with Measurements of the shear modulus, internal cast samples, e.g. 24-25?] Co-Pt has maximum friction and creep of annealed and unannealed magnetic energy 40 kJ;'m3, B,=79oo gauss, H, decreases for >25:,; Co-Pt, samples of Pt were carried out between 1300 and H,=3950 Oe. 1980°K and results are depicted graphically. i.e. highest magnetic properties occur for 23-25 uo Co-Pt, where concentration disordering is less important. Controlled cooling rates lead to more Coherent and Incoherent Precipitation in homogeneous and more dispersed structures due the Gold-Platinum System to prolonged keeping of samples at 770-830"C, J. WEISE and V. GEROLD, z.Metallkunde, 1968, 59, the temperature favourable to nucleation. (121, 904-909 Resistivity studies of IS, 20, 60 and 65 wt", Au- Platinum-based Permanent-magnet Alloys Pt wires during isothermal ageing and during re- H. C. ANGUS, Proc. Znstiz Elect. Eng., 1968, 115, heating to the homogenisation temperature (IZ), 1849-1852 showed two rates which can be attributed to Studies of changes of magnetic properties of coherent and to incoherent precipitation. The PtCo and PtFe when part or all the Pt is replaced metastable coherent miscibility gap in the phase by Pd, Rh, Ir, and Ru with various preparations diagram was -100-200 deg C below the stable and heat treatments show that the pseudobinary incoherent gap. The temperature dependence alloys possess no better magnetic properties. of the rate of resistivity change at the start of Careful heat treatment of sintered and worked or annealing 65 wtyb Au-Pt agreed with spinodal of vacuum-melted PtCo gives energy products of decomposition theory or a nucleation mechanism. 9.5 and 10.0 x IO~Gs-Oe, and coercive forces of 5.0 and 4.5 x io30e respectively. Effect of Deformation on the Physico- mechanical Properties of an Alloy of Study of the Diffusion of Normal Hydrogen Platinum with Silver and of Hydrogen Enriched with Para- s. N. PAVLOVA, Fix.-khim. Mekhan. Mazer., 1968, hydrogen through a Palladium-Silver Mem- 4, (41, 490-492 brane. Determination of the Activation Studies of the deformation of 0.1-3 mm Ag-Pt Energies alloy wires after quenching from IOOOTshowed L. PITOUSSI, c. r., St. c., 1968, 267, (26), 1742- that wires of different diameter had different I745 tendencies towards work-hardening. 99'>t0de- Activation energies measured during diffusion formation was required to obtain high strength through 23% Ag-Pd were 4.9 kcal/g.atom in a 0.1 mm wire but for wires 0.2 mm diameter (*Io%) for normal H, and 3.54 kcal/g.atom even 99.8qi deformation only produced a UTS of (AIO;,:) for H, enriched with parahydrogen. I 70 kg,i mm2. Electrical Resistivity Measurements in Anomalous Low Temperature Susceptibility Palladium-Hydrogen Alloys of Dilute Pt-Co Alloys c. T. HAYWOOD and L. VERDINI, Can. J. Phys., J. C. GALLOP and I. A. CAMPBELL, Solid state 1968,46, (r8), 2065-2071 Commun., 1968,6, (II), 831-833 Studies of Pd and Pd-H at 2-300°K show that at Nuclear orientation mcasurements of the hyper- IO"K

Platinum MetalsTev., 1969, 13, (2), 73-79 13 ETn, with n=3.1 for pure Pd but n decreasing to smallest grain size of all, possesses the highest H,. 2.3 when H/Pd=0.25. p1 depends on time and Thermal ordering increases the grain size and rate of cooling at high H concentrations and low thereby decreases H,. temperatures, Residual resistivity is lower for faster cooling. The increase in p, due to I at.",, The Surface Tension and Density of the H, in Pd is of the same order as that of inter- Liquid Alloys Pd-Fe, Pd-Cr, Pd-Si stitials in other f.c.c. metals but is less than for v. F. UKHOV, E. L. DUBININ, 0. A. ESIN and N. A. H, in b.c.c. Ta and Nb at room temperature. VATOLIN,Z~.$Z.Khim., 1968~42,(IO), 2631-2634. Liquid Fe-Pd alloys are nearly ideal. Isotherms The Diffusion of Hydrogen in Cu,Pd Alloy of surface tension for molten Cr-Pd are smooth. v. B. VYKHODETS, v. A. GOL'TSOV and r. v. GEL'D, In molten Si-Pd, molecular groupings in the ratio Fiz. Metal Metalloved., 1968,26, (S), 933-935 of the PdSi compound are maintained during Results are plotted and tabulated for studies of isotherms of surface tension. the activation energy Q for Hz diffusion in ordered and disordered Cu,Pd by measuring the Alloys with Extra Specific Electrical Resist- rate of penetration p, the diffusion coefficient D ance and Low Temperature Coefficient of and the solubility S. Resistance in a Wide Range of Temperature M. P. RAVDEL' and O. I. EVDOKIMOVA, Sb. Tr. Tsent. Investigation of Alloys of the Pd,Al-Ag Nauch.-Zssled. Inst. Chern. Met., 1968, (54), 26-34 System Alloying Ni with Mn-Pd does not alter the resist- L. A. PANTELEIMONOV, D. N. GUBIEVA and L. v. ivity p or the temperature coefficient ct of resist- AZIKONDA, Vest. moskov. Univ., Ser. II, Khim., ance of the alloy but annealing Mn-Pd-Ni at 1968, 23, (6), 87-90 300°C alters both p and 5. 0.3 mm diameter rods The phase diagram and mlcrohardness of these of 6556 Mn-I59/;, Pd-IS?,;, Ni-S%, Cu have: p~- alloys are depicted. zyohm.m,ar=3.1o-~perdegCat -60to +120"C, ~(=(7-8).10-~ per deg C at 120-300"~ e.m.f. On the Relation of Spontaneous Magnetisa- versus Cu E -0.7 mV/deg C, o =62-68 kg/mm2, 8-20-28 Repeated heating and cooling tion to Composition of the Ordered Alloys y,. between -60 and +3oo'C does not change their Fe(PtarPdl-r)3 electrical properties. Protracted annealing at s. K. SIDOROV, v. v. KELAREV and A. I. KOZLOV, 300°C does not alter their structure or properties Fiz. Metal. Metalloved, 1968, 26, (5), 776-780 but at 500-600°C reduces p and leads to brittle- A theory is developed for magnetisation of the ness and decomposition. ordered Fe(Pt,Pd,-,), alloys in which ferro- magnetic-antiferromagnetic exchange reactions Properties of Alloys of Manganese-Paila- occur between the atomic components. diurn, Manganese-Germanium, and Man- ganese-Gallium Systems On the Nature of the Coercive Force and Ibid., 35-42 Structure of Equiatornic Iron-Palladium The search for a high-resistivity alloy showed _4lloys that when quenching high Mn content alloys L. M. MAGAT, A. S. ERMOLENKO, 6. V. IVANOVA, from the y region a f.c. tetragonal lattice appears G. M. MAKAROVA and YA. S. SHUR, Ibid., (3),5I 1-516 which upon alloying becomes f.c.c. Resistivity of The maximum value of the coercive force of equi- Mn-Pd quenched from the y region is higher than atomic Fe-Pd occurs in the single phase regu- that of Mn-Cu or of Mn-Ni. Resistivity p of lated state. The coercive force of hardened and 20-32.7"A Pd-Mn varies slightly with tempera- annealed samples is explained by delay in ordering ture at 0-35o'C. p decreases abruptly above ferromagnetic domain boundaries at antiphase 350°C for all Pd-Mn alloys and dilatometric boundaries. At first the deformed samples after curves indicate a phase transformation at 35o'C. annealing are much more ordered than the har- dened and annealed samples but crystallites of A Palladium-Magnesium Alloy Phase of the tetragonal phase are so small that they appear Co,A15 Type to be the main factor determining the size of the L. WESTIN, Acta. Chem. Scad., 1968, 22, (8), coercive force. 2 574-2 580 X-ray studies of hexagonal PdMg,,.5 give Thin Films of Iron-Palladium Alloys with dimensions a-8.646-8.660 A and ~-8.175- High Coercivity 8.169 A. The phase has a range of homogeneity. YA. S. SHUR, L. M. MAGAT, A. A. GLAZER, E. V. SHCHERBAKOVA and N. N. SHCHEGOLEVA, Ibid., (S), The Structure of the PuPd Compound 938-939 A. V. BEZNOSIKOVA, E. 8. SMOTRITSKAYA and N. T. Studies of FePd showed that the coercivity H, CHEBOTAREV, Atomnaya Enerz., 1968, 25, (5), is greatest when the crystallites of the alloy are 430-43 I smallest. It follows that a film, consisting of the Pu,Pd, has lattice parameters a =7.028 &o.oo j,

Platinum MetalsTev., 1969, 13, (2) 74 b-4.57IiO.001, C-5.658~0.002 A. Inter- reactions involving Ir and Rh complexes. Data atomic distances have been determined and are discussed are from determinations of the role of tabulated and a model is depicted. the reacting gas by studies of a series of reactions of one metal complex MLi with a variety of Magnetic Susceptibility of Alloys of Plu- addenda, and by Comparison of kinetic, equilib- tonium with Palladium rium and thermochemical data from these N. T. CHEBOTAREV, W. N. SOKURSKII, M. A. reactions and the electronic and geometric pro- ANDRIANOV and A. A. IVANOv, Ibid., 431-433 perties of the corresponding adducts ((XY) MLJ. Magnetic susceptibilities of Pu-Pd alloys are tabulated and shown in graphical form. Thermal Decomposition of Rhodium, Iridium, and Ruthenium Chlorides Thermomagnetic Measurements on Alloys A. E. NEWKIRK and D. w. MCKEE,J. Catalysis, 1968, of the Platinum Metals with Chromium IIY(41,370-377 A. KUSSMAN, K. MijLLER and E. RAUB, z.Metall- Decomposition studies of RhCl,, IrCI, and RuC1, kunde, 1968, ~9~ (II), 859-863 hydrates in air and in H, were studied by thermo- Alloys of Pt metals with Cr which have similar gravimetric analysis and reduction studies were structures have similar magnetic properties, made for dispersions on low- and high-area sub- F.c.c. LI, phases of Cr with 25-30 at.n/, Pt, Pd strates. Reduction to the respective metals was or Ir have a distinct maximum of magnetisation complete at 105, 190, and -350°C. In air, with typical ferromagnetic behaviour. Saturation dechlorination and oxidation was complete at 890, magnetisation at - 193°C is -0.39 Vs/m2 for Pt, 680 and -40°C respectively. B,C, which has -0.15 Vs/m' for Pd, 4.06 Vs/mZ for Ir and low area, has little effect on the reduction tem- the respective Curie points are 900, 400 and perature of the Rh salt but A1,0,, which has high 170°C. Tetragonal PdCr with L1, structure is area, is associated with desorption of decomposi- weakly ferromagnetic. Rh-Cr, Ru-Cr and 0s-Cr tion products. possess no Lr, structure and are very weakly ferromagnetic. F.c.c. solid solutions rich in Cr The Metallic Nature of Osmium Dioxide are anti-ferromagnetic with high Nee1 points. J. E. GREEDAN, D. B. WILLSON and T. E. HAAS, Inorg. Chem., 1968, 7, (I I), 2461-2463 Single crystals of OsO, were prepared and elec- CHEMICAL COMPOUNDS trical conductivity and Hall effect measurements Mass Spectrometric Determination of the were made. The data can be rationalised on the basis of a wholly collective electron approach in- Dissociation Energies of Gaseous RuC, IrC volving the concepts of narrow bands and high and PtB density of states which are often encountered in N. S. MCINTYRE, A. VANDER AUWERA-MAHIEU and the transition metals and their compounds. J. DROWART, Trans. Faraday soc., 1968, 64, (II), 3006-3010 RuC, IrC and PtB molecules have been detected ELECTROCHEMISTRY with dissociation energies Do7 (RuC)=151.0= 3.0, D,"(IrC)= 148.453.0, Do"(PtB)-rr3.3 IL Hydrogen Adsorption Equilibrium on Pla- 4.0 kcal 'mole. tinum Electrodes I. TELCS and M. J~Y,Acta Chim. Acad. Sci. Hung., Preparation and Properties of Some Ternary 1968958, (3)3 275-285 Selenides and Tellurides of Rhodium A review of experimental evidence and theory of R. H. PLOVNICK and A. WOLD, Inmg. Chem., 1968, HLadsorption equilibrium on Pt electrodes shows 7, (12), 2596-2598 that most probably the observed effects are due MRh,X,, where M is Cr, Co, Ni and X is Se, Te, to the several states of H, adsorption on Pt, all of and Rh,Te, have the monoclinic Cr,S,-type which obey the Langmuir adsorption isotherm to structure with space group I2/m, except for a good approximation. There are probably 44 CoRh,Te, and NiRh,Te,, which are trigonal such states. with space group PT mr. The symmetry of these compounds with defect NiAs structure depends Diffusion Coefficient of Hydrogen in Pal- on whether metal ion vacancies are ordered or ladium Alloy Electrodes randomly arranged in alternate metal layers. J. BERSIER and A. K~~SSNER,U.S. Rept N68-339x4, I968>9PP Reversible Activation of Covalent Molecules Diffusion coefficient of H, in 2300 Ag-Pd alloy by Transition Metal Complexes. The Role of and in this alloy with 0.5, 0.9,2.0 and 3.0 at.';; B the Covalent Molecule additions was determined as a function of H, L. VASKA, Accounts chem. Rex., 1968, I, (11), concentration and temperature at 30-30oT and 335-344 varies with H, concentration as D(C)-12B A review of recent work on the mechanism of [RT AC(r C)~,'~C(A~LH(C))].

Platinum MetalsTev., 1969, 13, (2) 75 On the Behaviour of a Ruthenium Electrode- severity conditions without loss of stability for u.0.P.'~new Platforming catalyst. Im- catalyst in Solutions of Sulphuric and Oxalic R-16 Acid at Various Temperatures proved product distribution selectivity at lower pressures permits dehydrocyclisation of paraffins R. B. TARAN and G. P. KHOMCHENKO, Vest. moskov. and dehydrogenation of naphthenes to create Univ., Ser, Khim., 1968, 23, (6), 83-86 II, aromatics with a minimum of hydrocracking to The rate of catalytic hydrogenation of CH,NO? light hydrocarbons, i.e. yield of motor fuel on Ru was - twice that of electrochemical reformate of a given octane number is higher at reduction but the difference decreased as the lower operating pressures. The aromatics- temperature rose. creating reactions are favoured by lower pressures. Plants which have switched to this catalyst have Catalytic and Electrochemical Reduction of not yet taken full advantage of R-16because of Nitromethane on Ruthenium Electrode- their different operating pressures. See also catalyst Chern. W., 1969, 104, (z),60-61. L. P. MASHKOVA, A. I. PLETYUSHKINA and G. P. KHOIMCHENKO, Ibid., 104-106 Results of Testing of Platinum on Alumina Further studies of these reactions in 0.1 N H,SO,. Catalysts by Hydrogenating Rcnzene to Cy clohexane LABORATORY APPARATUS YU. I. KOZOREZOV and N. M. PIKALO, Khim. Prornyshlennost', 1968, (I I), 821-822 AND TECHNIQUE C,H, hydrogenation tests using a modified Pt,A1,0, catalyst showed that the latter is ex- Correction for Temperature Loading and tremely active, selective and stable, and is suitable High Gas Pressure Effects for the Constant- for industrial use. temperature Hot-wire Anemometer H. HASAN and J. c. DENT, Br. J. appl. Phys., J. Investigation of the Conversion of Cyclo- Phys. D., 1969,2, (I), 85-92 hexene on Pt-Zeolite Catalyst under Pressure Tests on Pt and 3or;,,Ir-Pt wires enabled a of Hydrogen calibration curve to be plotted for a constant- v. I. GARANIN, u. M. KURKCHI and KH. M. MINACHEV, temperature hot-wire anemometer using the alloy Kinet. Kataliz, 1968, 9, (s), 1080-1085 at high temperature in a high-temperature and Studies of cyclohexene cyclohexane isomer- high-density gas flow. The temperature co- and isations on decationised zeolite, on CaY and on efficient of resistance for Ir-Pt was deter- 30:; 0.5":, Pt/CaY show that on PtlCaY the isomerisa- mined and also its thermal conductivity. tion of cyclohexane proceeds via formation of cyclohexene on Pt with subsequent isomerisation Electrolytic Cell with a Rotating Double on the acid centres of the zeolite. The isomerisa- Ring Electrode for High Frequencies in the tion of cyclohexene depends on temperature. Measurement of Fast Reactions G. TRI~ORN,A. HEINDRICHS and w. VIELSTICH, The Effect of Heat Treatment on the Cata- Messfechnik, 1968,76, (9), 224-229 lytic Activity of Platinum Catalysts The construction of a double ring Pt electrode S. A. KHASSAN, S. G. FEDORKINA, G. I. EMELYANOVA system is described together with requirements and V. P. LEBEDEV, Zh. fiz. Khim., 1968, 42, (IO), for its use at high speeds of roration. 2507-25 12 Sintering studies on Pt black in vacuum at 300, BRAZING AND WELDING 400, 500, 600 and 700' C indicate a second order relation between changes in the activity and sur- Bits for Soldering Irons face area. Sintering produces a fixed activity and H. C. ANGUS, R. D. BERRY and B. JONES, Engng surface area for each temperature. The relation Muter. Design, 1968, 11, (IZ), 1965-1968 between activity and roasting temperature is ex- ponential. Given the activity at two temperatures, The suitability of Ru for the bits of soldering irons is limited by its cost and so methods have an equation relating activity to temperature can been developed to localise it on bit surfaces and be derived. Kinetics of decomposition of HaOz also to protect the non-functional shank surfaces. and of hydrogenation of cyclohexene on sintered The most promising tipped bit has an A1 shank black are similar. Specific activity is unrelated to to which is welded Ru facing. the duration and temperature of sintering but change of surface area and size of crystals are,

HETEROGENEOUS CATALYSIS Hydroisomerisation of Normal Pentane over a Zeolite Catalyst UOP Catalyst Improves Yield of Aromatics A. VOORHIES and P. A. BRYAKT, A.I.Ch.E.J., 1968, Eur. chem. News, 1969, 15, (361, Jan. 3), 34 14, (6), 852-5356 Reduced operating pressures are claimed at high A series of screening tests on zeolite hydro-

Platinum MetalsTeo., 1969, 13, (2) 76 isomerisation catalysts showed that the most accumulated on it. CH, was the sole product of active is Pdihydrogen mordenite. Typical test hydrogenolysis at 245°C on Ir of CPHB-x, conditions were 550cF, 450 lb;inz, H,:C,H,,= C6H1,,-1, etc. Comparing wires with A1,0,- 3.4, 8 g n- C5Hl,!h.g. catalyst. Neither gas to supported catalysts, reactions seem characteristic particle mass transfer, nor macropore diffusion of the metal rather than of the type of catalyst. nor chemical reactions at Pd sites are rate- limiting. HOMOGENEOUS CATALYSIS Hydrogenation and Hydrogenolysis. X. The Hydrogenation of Methoxy- and Ethoxy- Homogeneous Catalysis in the Reactions of anilines with Rhodium Catalyst Olefinic Substances. XI. Homogeneous Cata- S. NISHIMURA, H. UCHINO and H. YOSHINO, Bull. lytic Hydrogenation of Short-chain Olefins Chem. SOC.Japan, 1968, 41, (9), 2194-2195 with Dichlorobis-(tripheny1phosphine)plat- Hydrogenation of alkoxyanilines with alkali- inum (11) -Tin(II) Chloride Catalyst promoted Rh catalysts showed that Rh oxide pre- R. W. ADAMS, G. E. BATLEY and J. c. BAILAR,J. Am. pared by fusion of RhC1, with LiN03gives high Chem. SOC.,1968,90, (22), 6051-6056 yields of mcthoxy- and ethoxycyclohexylamines Studies of homogeneous hydrogenation of short- from methoxy- and ethoxyanilines respectively. chain olefins with PtCl,(PPh,), and SnC1,. <6 hydrogenolysis occurred and secondary zH,O catalyst mixtures showed that mono- amine formation was negligible. Studies with olefins hydrogenate rapidly only when the double LiNO, replaced by NaNO,, LiOH and NaOH bond is terminal. Reduction of hexadiene and are reported and less satisfactory yields were pentadiene isomers does not stop necessarily at obtained. Best conditions appear to be 90"c, the monoene stage. Dienes react most when both high pHI, and use of LiNO, with RhCl, when double bonds are terminal. Conjugation is not a -2% catalyst: substrate is required. necessary step before hydrogenation; short-chain conjugated diene isomers inhibited further reac- Hydrogenation of Olefins. Part 4. Reaction tion. Long-chain conjugated dienes are reduced of n-Butenes with Hydrogen Catalysed by since their larger size prevents formation of stable Alumina-supportedRuthenium and Osmium catalyst-diene complexes. Unsaturated nitriles are hydrogenated when the double bond is terminal; G. C. BOND, G. WEBB and P. B. WELLS, Trans. substituent groups on the P-C atom do not Furaday SOC.,1968, 64, (11), 3077-3085 hinder this and the CN- group remains intact. Studies in a static system of product distributions, orders of reaction and activation energies for hydrogenation and isomerisation of but-I-ene, A Palladium-catalysed Synthesis of Benzyl cis- and trans-but-a-ene at 0-70°C using Esters from Methylhenzenes I rnol.?< Ru/a-Al,Os and at 6o-14o0C using D. R. BRYANT, J. E. MCKEON and B. c. REAM,^. org. 1 mol.% Os/a-Al,O, show that Ru gives more Chem., 1968, 33, (II), 4I23-4127 isomerisation than 0s. The suggested mechanism Liquid phase production of benzyl esters from is isomerisation by loss of one H atom from methylbenzenes takes place at moderate tempera- adsorbed C$Hgto yield adsorbed C,H, and H, tures using Pd(OAc),-Sn(OAc), catalyst and and hydrogenation where the rate is determined I atm air, e.g. toluene in CH,COOH gives benzyl by H addition to adsorbed C,H, for each C,H,. acetate. The secondary oxidation product ben- Rate-determining steps for cis-trans isomerisation zylidine diacetate is formed at high conversions. and double-bond migration are discussed. Other methylbenzenes behave similarly. Xylenes undergo selective diacctoxylation in CH,COOH The Hydrogenation of Alkadienes. Part 11. to give a,a'-diacetates rather than u,a-di- The Hydrogenation of Buta-1,S-diene Cata- acetates, lysed by Rhodium, Palladium, Iridium, and Platinum Wires Palladium-catalysed Reactions of Unsatur- P. B. WELLS and A. J. BATES, J. Chem. SOC.,A, ated Compounds in Non-aqueous Solvents. inorg. phys. theor., 1968, (12), 3064-3069 Synthesis of Vinyl Acetate via Palladium Studies of buta-1,3-diene hydrogenation over Rh, Salt/Ethylene Complexes Pd, Ir, and Pt wires showed that C,H, composi- R. VAN HELDEN, C. F. KOHLL, D. MEDEMA, G. tions depended on temperature but not on activity, VERBERG and T. JONICHOFF, Rec. Trav. chim. initial pH2 or conversion up to 50:/u. C,H,, Pays-bas, 1968, 87, (ro), 961-991 yield was zero on Pd and 1-15O/: on the other The rate of the reaction of C,H, with Pd(OAc), wires. But-r-ene was the main product; trans-cis in CH,COOH to produce vinyl acetate increases ratios for but-a-ene were determined. Chemi- considerably on addition of small amounts of sorption of buta-I,?-diene by interaction of only NaOAc, LiCl or LiC1, Fe(OAc), to the system. one double bond became more important as the The reaction is first order with respect to both temperature rose. Hydrogenation activity of dean Pd(OAc), and the additions. When using PdCI, Ir was replaced by hydrogenolysis as C residues and NaOAc the rate is proportional to them.

Platinum MetalsTev., 1969, 13, (2) 77 Catalytic Asymmetric Hydrogenation Em- addition to RhH(CO)(PPh,) a, which is also ploying a Soluble, Optically Active, Rhodium formed by dissociation when RhH(CO)(PPh,), is Complex dissolved in C,H, or other solvents. Rapid hydro- W. S. KNOWLES and M. J. SABACKY, Chem. Commun., formylation of alkenes occurs at 250°C, I atm with RhH(CO)(PPh,), catalyst, giving a ratio of -20 19683 (22)s 145-1446 A catalyst precursor which contains optically for the formation of straight- and branched-chain active tertiary phosphine ligands is prepared by aldehydes from alk-1-enes. H exchange and isomerisation of alkenes with RhH(CO)(PPh,), dissolving RhL,Cl, complexes with L =tertiary are described. phosphine in I : I v/v C,H, : C,H,OH containing triethylamine and pressurising with H,. Examples of its use are described. The Preparation and Reactions of Hydrido- chlorotris(triphenylphosphine)ruthenium(II) Catalysis of Hydrogen Transfer by a Method Including Homogeneous Catalytic Hydro- Hypothetically Similar to Fermentation. genation of Alk-1-enes VI. High-activity Hydrogenation Catalysts P. s. HALLMAN, B. R. MCGARVEY and G. WILKINSON, on the Basis of x-Complexes of Rhodium(1) Zbid., 3143-3150 with Amino Acids RuClH(PPh,),.C,H, was prepared from RuCl, (PPh,), and H, in ambient conditions in the 0. N. EFIMOV, M. L. KHIDEKEL', v. A. AVILOV, P. s. presence of a base such as triethylamine. RuBrH CHEKRII, 0. N. EREMENKO and A. G. OVCHARENKO, (PPh,),.C,H,, RuClH(C,HJ(PPh,),, [RuClH Zh. obshch.Khim., 38, 1969, (12), 2668-2677 (bipyr)(PPh,) ,] ,, and RuH,(CO)(PPh,), were also Studies of Rh complex catalysts with N-phenyl- prepared. RuClH(PPh,), is the most active anthranilic acid and L-tyrosine showed that low- catalyst yet discovered for homogeneous hydro- valent compounds are stabilised at the expense of genation of alk-1-enes in C,H, or C,H,.CH,. reactions with aromatic rings and carboxyl groups. It is highly specific but kinetic study is difficult Both in solid and solution states the complex and slow poisoning occurs. is supposedly the dimer H[Rh,(phen),Cl] but H decomposes it to the monomer. The Rh complexes catalyse D exchange with H,O. A Carbonylation of Acetylene with [Ru(CO)JS hydrogenation mechanism for trans-stilbene is as Catalyst suggested. H, is derived during hydrogenations P. FINO, G. BRACA, G. SBRANA and A. CUCCURU, of fumaric acid and acetylene derivatives by the Chem. & Znd., 1968,(49, Dec. 7), 1732-1733 cis-scheme. Hydroquinone was prepared by carbonylation of C,H, under anhydrous conditions, in either tetra- hydrofuran or dioxan as solvent, in the presence Selective Homogeneous Hydrogenation of of [Ru(CO)J,. Highest yields occurred at A&-1-enes Using Hydridocarbonyltris-(tri- pHa=5 - 10atm. Yields up to 65% hydroquinone phenylphosphine)rhodium(I) as Catalyst were obtained by reacting C,H, with CO and C. O'CONNOR and G. WILKINSON,J. Chem. SOC.,A, H,O or alcohols in the presence of [Ru(CO)~]~at inurg. phys. theor., 1968, (11), 2665-2671 I50-25O0C and relatively low pco. RhH(CO)(PPh,), efficiently catalyses homogene- ous hydrogenation of unsaturated RCH =CH2 compounds. For hex-I-ene and dec-I-ene in FUEL CELLS CBHasolution the rate law is rate=k, K, c [S][A]/ (I + K,[S]), where [S] and [A] are respective con- Electrocatalysts for the Direct Electro- centrations of &-I-ene and catalyst, C is H, chemical Oxidation of n-Octane in Fuel Cells concentration in solution, K, is the formation con- E. J. CAIRNS and J. PAYNTER, 3. Electrochem. Soc., stant for an alkene intermediate complex, and k, 1968,115,(IZ), 1218-1224 is the rate constant for the rate-determining H, Activity of supported and unsupported Pt pre- activation step. High selectivity for reduction of pared in various ways was compared with that of alk-I-ene is due to steric factors caused by the commercial Pt black. An increase in the specific bulky triphenylphosphine groups. area of a C substrate gave a corresponding in- crease in electrocatalytic activity up to at least Hydroformylation of Alkenes by Use of 200 m2/g. Some types of Pt/C had 6-18 times Rhodium Complex Catalysts the activity of Pt black. No current or voltage cycling occured with mol.o/o HF electrolyte at D. EVANS, A. OSBORN and G. WILKINSON, Zbid., 36 J. 105°C in contrast to other acid electrolytes. (12),3133-3142 The inhibition period of trans-RhX(CO)(PR,) a catalysts, where X=halogen and R=aryl, is Structure and Performance of Hydrophobic removed by addition of hydrogen halide acceptors Gas Electrodes and the halide complex forms a hydrido-species A. D. D. TANTRAM and A. C. C. TSEUNG, Nature, by hydrogenolysis. The principal catalytic species 1969, 221, (5176, Jan. XI), 167-168 seems to be RhH(CO),(PPh,), formed by CO Electron micrography showed that a catalyst pre-

Platinum MetalsTev., 1969, 13, (2) 78 pared by mixing Pt black with a dispersion of tively. The thermal em.€. hysteresis loop of PTFE and applying it to a metal screen before some metal-sheathed mineral-insulated thermo- drying and curing at 300°C consists of porous couples is explained; the maximum discrepancy aggregates of Pt intermingled with PTFE. Per- from this loop is 23pV at 550°C by calculation, formance is very dependent on microstructure. 14pV by experiment on a single thermocouple. Whereas hydrophilic electrodes become flooded Mechanical hysteresis of the loop may cause with electrolyte hydrophobic electrodes become fracture. A reliable thermocouple of this type full of gas and the whole thickness of the electrode needs matching expansion coefficients of sheath is effective. The effect of catalyst surface area was and thermocouple. studied using graphite powder as the catalyst. Comparison of Platinum Resistance Thermo- meters between 63K and 373.15K. Part I. CHEMICAL TECHNOLOGY Part I1 The Separation of Hydrogen from Ammonia M. R. M. MOUSSA, H. VAN DIJK and M. DURIEUX, Synthesis Exhaust Gases by the Method of Physics, 1968, 40, (I), 33-48, 49-60 Diffusion through Palladium Twelve Pt resistance thermometers were com- pared with a reference one at 63-273.15"K and at YU. K. BAICHTOK, M. B. AIZENBUD, V. A. KURKOVSKII, 373.15% using a constant temperature bath with A. S. FURMANOV, I. YA. AZBEL' and 2. V. KORBUTOVA, the thermometers inserted in a Cu block im- Khim. Promyshlennust', 1968, (IO), 744-746 mersed in liquid iso-C,H,,, C,H,, 02>or N, and NH, synthesis is more economic when H, in the the differences of the reduced resistances W= exhaust gas is purified and when Ar is recovered. R(t),'R(O"C) of them were determined. The Tests on such gas and on pure N,-H, mixtures number of calibration points to determine the compared pure Pd and 5yb Ni-Ion, Ag-Pd foils W-T relation was studied and discussed. for H, purification by diffusion at 400 and 500°C. Both gas mixtures gave similar results. Ag dusted A Direct-reading Bridge for a Platinum on Pd foil activated the latter for H, diffusion. Resistance Thermometer C. w. VAx DER WAL and L. c. E. STRUIK, 3. sci. TEMPERATURE Znstrum., J. Phys. E, 1969, 2, (2), 143-145 A Wheatstone bridge for a Pt resistance thermo- MEASUREMENT meter was modified so that the angle of rotation of a linear potentiometer is proportional to the The Effect of Stress on the Thermal EMF of temperature with bridge sensitivity fairly inde- Platinum-Platinum/lS Rhodium Thermo- pendent of temperature. Although not intended couples for high precision thermometry, irrespective of E. s. MORGAN, Br. J. appl. Phys., 7. Phys. D, 1968, the inaccuracy of the thermometer itself the 1, (XI), 142r-1429 reading accuracy is to better than z0.1 deg C Variations in thermal e.m.f. with stress of Pt and at 0-500 C and the bridge can be used for both 13 :{> Rh-Pt wires with junctions at 20 and 550°C temperature control and temperature measure- are -47 and -12 pV deg C-lkg-lcm-z respec- ment.

NEW PATENTS METALS AND ALLOYS Pd with a minor amount of Cr capable of forming a stable refractory compound or alloying Kh or Methods of Improving the Mechanical Pt (or Rh, Pt with another Pt-metal) with a minor Properties of Metals and their Alloys amount of an element capable of forming a stable JOHNSON MATTHEY & CO. LTD refractory compound (the element may be Be, British Patent 1,134,492 Mg, Al, Si, Th, U or a transition metal) and Pt metal alloy articles are given greater strength heating (~ooo"C)the alloy in a gas (e.g. air or 0,) by cold working and annealing to give oriented to form the refractory compound (e.g. oxide) in recrys tallisation. the alloy. Improvements in and Relating to the Treat- CHEMICAL COMPOUNDS ment of Platinum Group Metals and Alloys JOHNSON MATTHEY 82 CO. LTD New Chemical Compounds British Patent 1,139,897 IMPERIAL CHEMICAL INDUSTRIES LTD An alloy having properties of oxidation resistance British Patent 1,138,867 and high mechanical strength is made by alloying These new chemical compounds are Pt group

Platinum MetalsTev., 1969, 13, (2), 79-84 79 metalalkyl carbonyl halides offormulaR. M. COX, Vacuum-metallised Electrode where R is alkyl, X is halogen and M is Group NATIONAL RESEARCH CORP. VIII noble metal, e.g. C,H,,Pt.COCI. They are U.S. Patent 3,410,785 made by reacting MX with HCOOH and an A Ti electrode for electrochemical systems is olefine. No uses are given for the products. vacuum coated with Pt

Platinum MetalsTev., 1969, 13, (2) 80 Simultaneous Preparation of Saturated Ali- using a noble metal catalyst, such as Pd, deposited phatic Hydrocarbons Containing 14C Atoms on a mixture of gibbsite and 1-20'i(, kaolin which and Benzene with its Homologues from has a surface area of more than 150 m',/g. Light Virgin Naphtha STAMICARBON N.V. British Patent 1,133,263 Fluorop yridines This process is catalysed by a dehydrogenation NATIONAL RESEARCH DEVELOPMENT CORP. British Patent 1,134,651 catalyst comprising 0.1-21: Pt on AI,O, and/or SO9. Trifluoro- or tetrafluoropyridines can be obtained by hydrogenating chlorofluoropyridines, e.g. Preparation of Dinitriles from Acrylonitrile using PdC as catalyst. E. I. DU PONT DE NEMOURS & CO. British Patent 1,133,900 Paraffinic Jet Fuel by Hydrucracking Wax ESSO RESEARCH & ENGINEERING CO. The acrylonitrile is heated at 100-2oo'C and 3.5- British Patent 1,137,639 635 Kg,/cme in the presence of metallic Ru, and gaseous H,. Jet fuel is produced by cracking a paraffinic wax feedstock over a catalyst, such as I'd on a zeolite, Hydrocatalytic Conversion of Hydrocarbons at 300-9oo',F, 500-3000 psig and 0.1-10 viv:h with added H,. BRITISH PETROLEUM CO. LTD British Patent 1,134,014 Preparation of Dicarboxylic Acids The selective hydrocracking of n-paraflins is E. I. DU PONT DE NEMOURS & CO. catalysed by a Group VI or VIII metal or oxide British Patent 1,i38,131 deposited on a decationised mordenite having a SiO, : A1,0, ratio of 14: i. Pt is a suitable metal. Cyclic olefines are oxidised to dicarboxylic acids by HNO, in the presence of an 0s-V catalyst, Catalysts e.g. in Os0,-V20j or in metal form. JOHNSON MATTHEY & CO. LTD Preparation of Chlorine-substituted Aro- British Patent 1,134,111 matic Amines Catalysts for oxidation, reduction and other GENERAL ANILINE & FILM COW. chemical reactions are intimate homogeneous British Patent 1,138,567 mixtures (not pure mechanical mixtures) of one or more oxides of a Pt metal (excluding 0s) and These amines are produced by the catalytic one or more oxides of Fe, Co, Ni and Cu. hydrogenation of corresponding chloroaromatic nitro compounds using a new catalyst system. This comprises (a)o.or-o.08'1,;l of Pt black A Process for Starting up a Reactor for the 1-5(>;, or Pt or Pd on C or A1,0,, and (b) 0.05-0.5')0 tri- Catalytic Conversion of Hydrocarbons or a phenyl or tritolyl phosphite. Hydrocarbon Mixture SHELL INTERNATIONAL RESEARCH MIJ. N.V. Preparation of Unsaturated Organic Esters British Pateiit 1,134,144 NATIONAL DISTILLERS & CHEMIC4L CORP. A process for starting up a reactor containing a British Patent i,I39,210 catalyst comprising Pt and halogen for the cata- The contact catalysts in a process for the prepara- lytic conversion of hydrocarbons or a hydro- tion of an unsaturated organic ester from an carbon mixture, in which a fresh catalyst or a alkene and an 0,-containing gas comprise (a) a catalyst regenerated with the aid of 0,, steam and Pd- or Pt-group metal or metal compound sup- halogen (the Pt being present on the catalyst as a ported on C, and (b) a Pd- or Pt-group metal or Pt-halogen compound or complex) is first dried metal supported on A1,0,. by means of a preferably previously dried oxidative gas or gas mixture at a temperature of Oxidation of Ammonia at least 480°C and at most 570°C till the H,O INVENTA A.G. content of the reactor off-gas is less than 150 British Patent i,i39,849 ppmv. The catalyst, whether in the presence of the hydrocarbon feed to be converted or not, is NH, is oxidised by mixing with 0, in volume ratio NH, : 0, of I : 1.10-1.27 and passing the subsequently reduced by means of dry at a €I2 mixture with steam over a Pt or Pt-Rh catalyst at temperature of 500°C. a specified flow velocity at 800-950°C. The catalyst may be a Pt-Rh grid containing 90% Pt, Process for Reducing the Polymerisation ro:: Rh. and/or Condensation Tendency of Un- saturated Benzines Rich in A4romaticSub- Preparation of Aryl Thiols stances UNIROYAL INC. U.S. Patent 3,399,238 VEB LEUNA-WERKE WALTER ULBRICHT C,H, and p-toluene thiols are produced by the British Patent 1,<34,599 hydrogenation of the corresponding aryl sul- Petrol fractions are stabilised by hydrogenation phuric acid over a PtS catalyst.

Platinum MetalsTev., 1969, 13, (2) 81 Production of 1-Raphthol NHa andior a tertiary amine in the presence of a UNION CARBIDE CORP. U.S. Patent 3,402,tfO catalyst containing Ru, Ni or Co. The dehydrogenation of 3,4-dihy&o-1-(zH)- naphthalenone is catalysed by a supported mix- Treatment of Combustihle Wastes ture of 0.25-5 wt0&Pt metal and 0.25-r.5 wtoh UNIVERSAL OIL PRODUCTS CO. of Nd in the form of an inorganic compound, U.S. Patent 3,409,390 e.g. NdCl,. Noxious exhaust gases are oxidised by contacting under oxidising conditions with a catalytic com- Hydrocracking Catalyst posite essentially of A1,0,, a Pt-metal(o.01 %-I :,; STANDARD OIL CO. U.S. Patent 3,403,094 by wt) and an alkaline earth component (>I "L). The catalyst comprises a known active Group VIB or VIII metal such as Pt deposited on a Catalytic Isomerisation of Saturated Hydro- fluorided Si02-Al,0, cracking catalyst and pro- carbons moted with an alkali metal or alkaline earth metal. UNIVERSAL OIL PRODUCTS CO. US.Patent 3,409,682 Group VIII Aluminosilicate Catalyst - The catalyst for isomerisation of saturated hydro- PULLMAN INC. U.S. Patent 3,403,108 carbons comprises at least one active catalytic An improved catalyst for unsaturated ester pro- component, preferably Pt, on an A1,03 support duction consists of an alkaline metal alumino- in which not more than ZOO:, of a finely divided silicate with about 0.1-15 wt", of a Group VIII crystalline aluminosilicate is dispersed. metal having an atomic number of at least 4and a Fe(III), Cu, Cr or W halide, e.g. PdC1, and CuCI, on a zeolite. Catalytic Isomerisation of Xylene UNIVERSAL OIL PRODUCTS CO. Selective Hydrogenation of Ole fines US.Patent 3,409,685 SINCLAIR RESEARCH INC. U.S. Patent 3,404,192 Alkyl aromatic hydrocarbons are isomerised in The hydrogenation of straight chain olefines in contact with a Group VIII metal catalyst (e.g. the presence of branched olefines is catalysed by a Pt) on an Al,O,!alumino-silicate matrix and also Pt catalyst produced by treating a deactivated in contact with about O.OOI-Z.O wty& of an S- zeolite with H,PtCl,. containing isomerisation promoter. See also U.S. Patent 3,409,686. Hydrocracking Process UNION OIL CO. OF CALIPORNIA Dehydrogenation of Cycloparafiic Hydro- U.S. Patent 3,405,056 carbons When the H2S concentration is low, the catalyst UNIVERSAL OIL PRODUCTS CO. for hydrocracking may be a zeolite promoted with U.S. Patent 3,409,690 a partially (coke) - deactivated Group VIII noble The catalyst for dehydrogenating cyclo-paraffinic metal (especially Pd). hydrocarbons to aromatic hydrocarbons com- prises a refractory inorganic oxide (e.g. A1,0,) Hydroxylammonium Salts containing at least one metal (e.g. Pt) from Group FARBENFABRIKEN BAYER A.G. VIII of the periodic table, chemically combined U.S. Patent 3,406,011 with a metal subfluoride (e.g. AlF) vapour. See These salts are produced by the reduction of NO also U.S. Patent 3,409,699. with H, in the presence of noble metal catalysts, e.g. Pt/C, which are periodically or continuously Catalytic Olefines Isomerisation treated with O2to maintain the catalyst selectivity ETHYL CORP. and activity. US.Patent 3,409,702 a-Olefines (4-24C) are effectively isomerised to Oxidation of Alcohols to Carboxylic Acids p-olefines in a process using a catalyst of mixtures SIXELL OIL co. U.S. Patent 3,407,220 of Group VIII metals on an inert support, e.g. Saturated aliphatic straight chain primary alco- Rh, Pd, Ru and Pt on charcoal. hols are selectively oxidised to the corresponding carboxylic acids in the presence of branched chain Catalytic Reforming acids by using finely dispersed Pt as the catalyst. UNIVERSAL OIL PRODUCTS CO. U.S. Patent 3,410,789 Production of Methyl Pentamethylenedi- Gasoline or naphtha fractions are reformed in amine in Two Stages contact with a refractory inorganic oxide (A1,0,) NATIONAL DISTILLERS & CHEMICAL CORP. containing a Group VIII metal (Pt) which has U.S. Patent 3,408,397 been contacted with vapours of AlF or a Si sub- An appropriated dinitrile is reduced over a Pd or fluoride and a S-containing component (tert- Pt catalyst and then the product is reacted with butyl mercaptan). See also U.S. Patent 3,410,790.

Platinum MetalsTev., 1969, 13, (2) 82 Catalytic Transalkylation Improvements in or Relating to the Produc- UNIVERSAL OIL CO. U.S. Patent 3,410,921 tion of Silicon Compounds A polyalkylated aromatic compound is trans- IMPERIAL CHEMICAL INDUSTRIES LTD alkylated by reacting the compound in admixture British Patent 1,140,424 with H,, with alkylatable aromatic compound in Organic silicone compounds (e.g. C1,SiCH,CH1 contact with a catalyst. The catalyst is preferably SiCl,) are prepared by contacting a compound a Group VIII metal (e.g. Ni, Pt, Pd; especially containing at least one Si-H bond with an Pt) on an alumino-silicate support under trans- organic compound containing at least one alkylating conditions. acetylenic or olefinic bond in the presence of a catalyst of general formula RhX(R’R”R”’Y), Butyraldehyde Production where X is an anionic element or group, R’, R, JOHNSON MATTHBY & CO. LTD R”’ are organic groups and Y is As, Sb or Bi, German Patent 1,278,420 e. g. RhCI(Ph,P),. The hydrogenation of crotonaldehyde in the liquid phase using a trickle column reactor is Oxidation of Aromatic Compounds catalysed by Pd,/Al,O,. LUMMUS CO. British Patent 1,141,238 Aromatic compounds, having at least one un- HOMOGENEOUS CATALYSIS substituted position on the ring and free from any Production of Unsaturated Compounds unsaturated substituent groups, are oxidised to phenols in the presence of an alcohol and a IMPERIAL CHEMICAL INDUSTRIES LTD Group noble metal compound, e.g. CcHo British Patent 1,138,366 VIII in the presence of PdCl,, CuC12 and ethylene Unsaturated diesters and diethers are produced glycol gives phenol. by contacting a conjugated diene with a carboxylic acid or an organic hydroxy compound in the Production of Lactones and Ketones presence of a salt (not a nitrate or sulphate) or co- IMPERIAL CHEMICAL INDUSTRIES LTD ordination compound of Pd, Pt, Ni, Fe or Co, British Patent 1,141,353 e.g. palladous acetate, chloride or acetyl- acetonate. Dienes are converted to cyclic ketones and lac- tones by reaction with CO in the presence of Pd Improvement in Catalytic Reactions and acid ions. The Pd is preferably in complexed G. WILKINSON British Patent 1,138,601 form as in PdI,(PBu,),. Hydrogenation, hydroformylation or carbonyla- tion reactions are catalysed by a complex of a Pt Production of Polynuclear Aromatic Com- metal halide or pseudohalide and either (a) an pounds organic nitrile or (b) a N, P, As, Sb, Bi, S or Se GULF RESEARCH & DEVELOPMENT CO. compound. For example (Ph,As),RhCI, catalyses U.S. Patent 3,401,207 the hydrogenation of hexyne-1 to C,H,,. Polynuclear compounds are produced from aromatic compounds having at least one labile Decomposition of Formic Acid H atom by contacting them with a noble metal IMPERIAL CHEMICAL INDUSTRIES LTD salt of an organic acid in the presence of a strong British Patent 1,138,946 acid; halide and nitrate ions must be absent. A The selective destruction of HCOOH in the typical catalyst consists of Pd acetate and HClO,. presence of CH,COOH is achieved by oxidation with Oz using a catalyst system comprising a Saturated Aliphatic Nitrile Production soluble Pt compound and a redox system, e.g. ASAHI KASEI K.K.K. PdCl,, LiC1, LiOAc and Cu(OAc),. US.Patent 3,407,223 These nitriles are produced by the reaction of Production of Unsaturated Aliphatic Nitriles C,H, or C,H, and HCN at 200-600°C in the ASAHI KASEI K.K.K. presence of I’d, Rh or their salts, H,PO, and British Patent 1,139,398 ammonium thiocyanate, A1 thiocyanate or AlC1,. An unsaturated aliphatic nitrile is produced by contacting a gaseous mixture of an olefinic hydro- Catalytic Ester Production carbon, HCN and 0 with a catalyst selected from SHELL OIL CO. Pd metal, Pd compounds and mixtures consisting U.S. Patent 3,407,224 predominantly of at least one of PdCl,, CsBr and The reaction of conjugated dienes and carboxylic SO,. Alternatively high temperature (-300°C) acids to produce 2,7-alkadienyl esters is catalysed may be used while feeding in simultaneously a by a Pt, Pd or Ru compound activated by a gaseous mixture of halogenated aliphatic nitrile phenoxide anion. In an example PdClz activated and hydrogen halide or halogenated aliphatic by sodium phenate (NaOPh) is used to catalyse nitrile, unsaturated aliphatic nitrile and hydrogen the production of z,7-octadienyl acetate from halide. butadiene and CH,COOH.

Platinum MetalsTev., 1969, 13, (2) 83 Solutions of Noble Metal Compounds Pt group metal wound round helically with a LUAlMUS CO. U.S. Patent 3,410,807 plastic insulator. Stable homogeneous solutions suitable for the catalytic oxidation of olefines, aromatic hydro- carbons, CO and SO, are compounds of Group CHEMICAL TECHNOLOGY VIII metals with a concentration of metal ion Improvements in or Relating to Gas Per- between 0.01 and 0.0001 mol/l, mono- or poly- meable Diaphragm hydric alcohols containing about 0.1-120; by vol. of H,O. The Group VIII metal compound is ENERGY CONVERSION LTD British Patent 1,139,501 preferably PdCl,. A gas-permeable diaphragm comprises a non- porous gas-permeable metallic member (Pd) Catalytic Addition of Si-H to --C and C G C C adhering face to face with a non-porous sheet of UNION CARBIDE CORP. U.S. Patent 3,410,886 polymeric material permeable to at least the same Si-C compounds are produced by contacting a gases as the metal. compound containing at least one Si-H with an aliphatic organic compound having C =I C or Hydrogen-Permeable Membrane and Hydro- C=C in the presence of a Pt (11) complex for sufficient time to form the Si-C bond. gen Permeating Assembly JAPAN GAS-CHEMICAL CO. British Patent 1,140,952 FUEL CELLS A H,-permeable membrane of Pd or I'd-alloy (0.3-0.005 mm thick) is characterised by a Fuel Cells and other Galvanic Cells plurality of convexities protruding in the direction ALLMANNA SVENSKA ELEKTRISKA A.B. of the thickness of the membrane and concavities British Patent 1,133,921 surrounding these convexities so that it has a Porous electrodes, e.g. made from Ni powder, wavy configuration longitudinally and laterally. are now modified by adding 0.005-0.1:, Pt The I'd-alloy may be Pd-Ag-Au. metal before sintering so that the added metal lies in the active surface. ELECTRICAL AND Fuel Cell Electrode ELECTRONIC ENGINEERING STANDARD OIL CO. OF INDIANA U.S.Patent 3,405,007 Semiconductor Devices Catalytic metals are deposited on porous con- C.K.D. PRAHA, OBOROW PODNIK ducting supports by impregnating the support British Patent 1,135,068 with a compound of the metal (e.g. H,PtCl,) Junction problems between semiconductor wafers and reducing carboxylic acid and then heating it. in pressure contact with electrode contact plates are overcome by applying a layer of Rh to both Spinel-Ruthenium Catalysed Electrodes contact surfaces. UNION CARBIDE CORP. G'.s.Patent 3,405,010 Electrodes especially suitable for fuel cells consist Electrical Resistance Element of a support treated with a heavy metal salt, an INTERNATIONAL BUSINESS MACHINES CORP. A1 salt and a Ru salt and then heated at 700- U.S. Patent 3,411,122 goo"C to form the required spinel. The heavy A laminated electrical resistance structure con- metal may be Au and Ag, among other metals. sists of an insulating base supporting a conductive layer of a finely divided metal and metal oxide Hydrogen Control in Fuel Cells (e.g. Pd-Ag, Pd oxide or In oxide) and a layer of PROTECH INC. U.S. Patent 3,407,094-5 a polyimide condensation product. H, is generated from carbonaceous fuel reforma- tion for passage to a fuel cell where a permeable Pd membrane is used to control current genera- TEMPERATURE tion. In 3,407,095 a Pd-containing film is used. MEASUREMENT Thermistor Conipositions CATHODIC PROTECTION E. I. DU PONT DE NEMOURS & CO. U.S. Patent 3,408,31 I Anode Assembly for the Internal Cathodic Thin film thermistors with a negative coefficient Protection of Pipes against Corrosion of resistance are produced by applying and firing W. C. HERAEUS G.m.b.H. (on a substrate) a mixture of finely divided Pd British Patent 1,133,285 mixed with up to 60"; Ag and an inorganic The anode is made with a surface of Ti, Ta, Nb binder as before but now also a finely divided or an alloy of one of these metals coated with a mixture of Co,Ol and MnO,.

Platinum MetalsTev., 1969, 13, (2) 84