NOTES
Jan., 1962 Cotton, et aL,7 have .re-examined the infrared spectrum of Fe(C0)b and failed to observe two bands crucial Lo O’Dwyer’s argument. Other infrared i n v e ~ t i g a t i o n sand ~ ~ ~ several normal coordinate a,nalyses10-12support the D ~ I structure. , The Raman shifts observed recently by Stammreich, et al.,13 also are consistent only with the non-polar structure. In the present investigation the dielectric constant, E’, and dielectric loss, I:”, of pure liquid Fe(C0)h were measured a t wave lengths of 1.25 and 3.22 cm. with an apparatus described else~ h e r e . 1 The ~ dielectric constant also was measured a t radiofrequencies. The results are shown in Table I. TABLE I Wave length, om.
e’
E”
1.25 3.22 577 m.
2.6209 2.6178 2.6257
0.00986 0.00465
T,
see.
3.35 x 28.5 X
PI
D
0.15 0.10
.......... .. All measurements were taken at 20’ on a triplydistilled s8ample of iron pentacarbonyl and all operations were carried out .in subdued light. Using the data of Table I, estimates of the magnitude of the dipole moment were made from the equation15 p% =
. .. . .
27kT( 1 4- d+)E” (e’ 2) 2 4 7 r ~ A ~
+
( 1)
187
some typical non-polar substances. The loss values for these are shown in Table 11. Whiffen18 has suggested that the losses in nonpolar substances arise from the relaxation of small instantaneous dipole moments induced in the molecules by inter-molecular collisions in the liquid state. It is probable that a similar mechanism will account for the loss found in iron pentacarbonyl. This molecule possesses three infrared active fundamental vibrations between 90 and 110 cm.-le7 It is these low-lying fundamentals which contribute heavily to the atom polarization and make the molecule easily susceptible to collisioninduced dipole moments. Nickel tetracarbonyl has only two infrared active low-lying fundamentals and might be expected to have a smaller atom polarization than the iron compound. The limited data available4 seem to confirm this. In Fe(CO)s the measured loss a t 1.25 em. is higher than the loss at 3.22 cm., which is consistent with the variation of loss with frequency observed by Whiffenl* in several non-polar compounds. The findings of the present work are consistent with the D 3 h structure of the undeformed Fe(C0)6 molecule, the discrepancy between the total polarization and the electron polarization being due to a large atom polarization contribution and the presence of collision-induced instantaneous dipoles. Acknowledgment.-We wish to thank Professor C. P. Smyth and Dr. W. E. Vaughan for their encouragement and many helpful discussions. One of us (E. N. DiC.) is indebted to the Esso Foundation for financial assistance in the form of a fellowship.
The values of the relaxation time, 7, were determined in a soiriewhat approximate manner from the linear plot of E’ us. E ’ ’ u ,assuming ~~ a Debye (18) D. H. Whiffen, Trans. Faraday Soo., 46, 124 (1950). line shape. Dipole moments of 0.15 and 0.10 D were calculated from the 1.25 and 3.22 cm. data, respectively. Although there is a large discrepancy in the calculated relaxation times, 7,for the THE FORMATION CONSTANTS OF T H E two sets of data, it was found that using values of TANTALUM FLUORIDE SYSTEM. 11. 7 between. the two calculated values tended to TANTALUM ELECTRODE POTENTIAL alter the calculated dipole moment, p, only t o a STUDIES1 small extent (0.01 to 0.03 0 ) in both cases. EquaBY LOUISP. VARGAAND HARRYFREUND tion I yields a value of p = 0.04 D for benzene.15 The values! found for the dielectric loss of Fe(CO)& Contributzon f r o m the Department of Chemzstry, Oregon State University Corvallzs, Oregon and the U. 8.Bureau of Mznes, Albang, Ofegon are considerably larger than those noted for some Received August 0 , 1961 other non,-polar molecules. Bleaney17 measured In a previous report2 it was demonstrated that TABLE I1 potentiometric hydrogen ion measurements and lor^ VALUESOF NON-POLAR SUBSTANCES anion exchange distribution studies in perchloric Wave length, om. 1.35 3.2 acid solutions of Ta(V) and fluoride ion yielded Benzene: O.OOl2 0.00050 the formation function between fluoride ligand Carbon tetrachloride 0.001’75 0,00069 numbers 4 and 9. This paper describes tantalum electrode-hydrogen electrode potential measure(7) F. A. Cotton, A. Danti, J. S. Waugli and R. W. Fesaenden, J . Chem. Phys., 29, 1427 (1958). ments made in 1 molal perchloric acid over the same (8) R. K. Sheline and K. S. Pitzer, J . Am. Chem. Soc., 72, 1107 fluoride ion concentration range covered in the (1950). previous studies. Apparent reversible behavior (9) W.F. Edge11 (private communication quoted in ref. 7). was found at fluoride ion concentrations above 4 X (10) W. G. Fateley and E. R. Lippincott, Spectrochim. Acta, 10, 8 (1957). molar in agreement with Haissinsky and co(11) H. Murs’a and K. Kawai, J . Chem. Phys., 28, 616 (1958). workers. (12) C. W. F. T. Pistorius and P. C. Haarhoff, zbid., 31, 1439 Since several tantalum fluoride species exist at (1959). (13) H. Stammreich, 0. Sala and Y . Tavares, ibid., 30, 856 (1958). (14) W. M. Ileston, Jr., A. D. Franklin, El. J. Hennelly and C. P. Smyth, J . Am. Chem. Sot:., 72, 3443 (1950). (15) W. M. IXeston, Jr., and C. P. Smyth, ibid., 71, 99 (1950). (18) R. H.Cole, J . Chem. Phgs., 23, 493 (1955). (17) B. Bleaney. J. €3 N. Loubaer and R . P. Penroae, Proc. Phyr. SUc. (Landon$, 69, 185 (1847).
(1) Abstracted in part from the Ph.D. thesis of Louis P. Varga, Oregon State University, June, 1960. Presented a t the 139th National Meeting, Am. Chem. Soc., St. Louis, Mo., March 30, 1961. (2) L. P. Varga and H. Freund, J . Phys. Chem., 66,21 (1962). (3) M. Haissinsky, Comate ztern. thermodynam. et cinet. electrochim. Compt. rend. reunaon, 222 (1951); M. Hsissinsky, A. Coohe and M. Cottin, Jd chzm. phus., 44, 234 (1947)d
Nma
188
I I
.
I
I
-r
_. - I 3 -0.1
-3.4
I
-4.8
I
I
I
- 4.0
I
I
-3.0
I.
I
I
-2.8
I
I
-3.0
I
I
- 2.2
L O G (F-1, M.
Fig. l.--.*Reversible or normal; X irrcversible. Tantalum electrode cell potentials and the average tantalum fluoride ligand number aa a function of log (F-). The least squares quadratic equation for the otential curve drawn is Eeell = 0.769 0.182 lo (F-P O.O13[log (F-)]* Interpretation of the potenti3 data beyond the limits od the insert is riot justified using a quadratic function.
+
+
equilibrium a t the surface of the electrode) the cell reaction with the hydrogen electrode may be written TaO
+ 5 H + + fiF- = TaFi6-i + 5/2 H1
(1)
The corresponding Nernst equation at 25' is 0.0591
5log
(TaFi6-i) (Ha)'/:
(H+)b
(2)
where Eo'contains the activity coefficient quotient term. Partial differentiation of Eaellwith respect to log (F-) gives (3)
where it is seen that the slope of the Eceil vs. log (I?-) plot a t constant metal and acid concentration is a direct measure of a.4 It should be noted that the derivation of equation 3 requires no knowledge of the tantalum species reversible to the metal electrode and analysis of potential data by equation 3 will give no information as t o the species measured. Experimental Procedure.-Potential measurements were made between tantalum metal electrodes and a hydrogen electrode without liquid junction in a series of solutions containing 1 molal perchlorate ion and 0.02 to 12 molar hydrofluoric acid. Repeat potential measurements were made at 15 to 30 minute intervals until two or three measurements indicated the abscnce of drift. The average tantalum electrode potentials have been tabulated.' (4) I. Leden, Z. phyeik. Chem., A188,160 (1941).
Vol. 66
All solutions were analyzed for total fluoride content by titration with thorium nitrate. The tantalum Concentration increased slowly due to solution of the tantalum electrode in the acid fluoride medium. The hydrogen ion concentration of each solution was determined from potcntial measurements of the hydrogen electrode against a normal calomel elcctrode with liquid junction as detailed in reference l. The fluoride ion concentrations were calculated from the total H F and hydrogen ion concentrations as described previously.2 Hydrofluoric and perchloric acids used for solution make-up were analytical reagent grade. The tantalum metal electrodes were made from Fansteel thin dental sheeting and from rolled tantalum sheet produced at the U. S. Bureau of Mines, Albany) Oregon. Spectrographic analyses of both metal samples have been reported.' The potentiometer and associated equipment were the same as used previously.2 The cell container was partially immersed in a water-bath a t 25 f 0.1') as was the calomel electrode. Plastic equipment was used throughout for all fluoride-containing solutions. Polarization Studies.-A potential of 3.0 to 4.5 volts applied between the tantalum electrode as cathode and an adjacent bright platinum eIectrode as anode was intended as an electropolishing device. Since we were not dealing with a uniform surface such as a mercury pool, it was anticipated that vigorous evolution of hydrogen gas or some similar mechanical or chemical action by electrolysis would condition the natural crystalline nature of the tantalum electrode so that its potential would be reproducible. With solutions of log (F-) greater than -3.4, it was observed that after such treatment the tantalum electrode potential changed rapidly at first, finally reaching a steady value in the range considered reversible. In log (F-) conccntrations of -3.4 and below, erratic behavior was encountered suggesting that an oxide layer was dissolved at least partially a t the higher fluoride concentration but not a t the lower concentration. The data are shown plotted in Fig. 1. After anodic polarization of the tantalum electrode for one minute periods up to 0.040 microampere, the electrode potential returned to the original steady-state value after a few seconds in solutions of the higher fluoride concentration. Again the tantalum electrode behaved as if an oxide coating mere dissolved slowly by the aqueous acid fluoride to givc a surface which behaved in a normal manner. Whether the tantalum electrode was ever behaving reversibly toward some tantalum species in solution generally could not be proved by polarization tests, The tests did show that immersion in acidic HF solutions of sufficient concentration caused the tantalum elcctrode to become active and approach a potential in a predictable rangc. It is doubtful whethcr the exchange current for deposition of tantalum metal from solution approached the magnitude of the reverse current but this mechanism for reversible behavior is clouded by the continued forward corrosion reaction of the tantalum metal in the acid fluoride solution. Assuming that the experimental potentials approached the reversible potentials, an analysis may be made.
189
NOTES
Jan., 1962
GASEOUS O S l D E S OF RHENIUM1 Results and Conclusions -The experimental potentials of the cell Ta/Ta(V), EF(1 m C104-)/ BYMARTIN H. STUDIER H2Pt as a function of log (F-) are plotted in Fig. 1. ArOonna National Labwotory, Argonna, IIItnnis The vertical scatter of t,hc potential is inherent in Received AugW 14. 1081 the behavior of solid metal electrodes, even a t the higher fluoride concentrations. A least squares fit Several gaseous rhenium oxides have been deof the 78 data points considered to be in the re- tected with a Bendix Time of Flight Mass Spectromversible or normal range was made t o a quadratic eter.2 During a study of surface ionization13 function. The resultant equation neutral species were volatilized (at temperatures below 500’) from rhenium surfaces on which samples in nitric acid had been evaporated. Ions EDoii= 0.760 + 0.182 log (F-) + O.O13[10g (F-)]* (4) produced by electron bombardment of the gaseous is shown as the curve plotted in Fig. 1. The first species were identified by their masses and corresponded to the empirical formulas derivative of equation 4 d E,,ir/d log (F-)= 0.182
+ 0.026 log (F-) (5)
yields an analytical expression for the slope which may be substituted directly into equation 3 to give values of f i as a function of log (F-). Such a formation curve is shown as an insert in Fig. 1 over the fluoride ion concentration range for which the data would appear to be most applicable. This formation curve for the tantalum fluoride system is seen to be about one-half of an fi unit higher in this fluoride region than the formation curve found previously.2 This agreement is satisfactory considering the difficulties of the system, and the assumption that the cell approaches reversible behavior appears justified. No new calculation of a set of formation constants was made, but the results did strengthen the evidence for the existence of the species TaFg-4, although in a statistical sense only. The solubility of Ta+& may be inferred from the studies by assuming from Fig. 1 that Eofor the reaction of equation 1 was near 0.35 v. a t the higher fluoride concentrations. Upon subtracting the tantalum fluoride half-reaction of equation 1 from Lati~ner’s~ half-reaction for the oxidation of tantalum to the pentoxide one obtains 2TaF,,+“
+ 51120 =
T d ) 6
+ 2nF- + lOII+, EO
0.4 v. (6)
The free energy change favors the reaction from left to right as written. This agrees with ohservation siuce the salt IC2TaT’7when dissolved in pure watcr is hydrolyzed to give an acid solution. Very thin oxide films arid freshly precipitated hydrated tantalum oxide dissolve in sufficiently strong hydrofluoric acid, but the massive crystalline oxide such as is formed after sintering the hydrate a t 1000°for several hours is hardly affected by the cold acid. This work was supported by a cooperative agreement between Oregon State University and the U. S. Bureau of Mines, Albany, Oregon. (5) W. M. Latimer, “Oxidation Potentials.” 2nd ed., Prontice HaU, Englewood Cliffs, N. J.. 1962.
Monomers
Dimers
Re +
Re+
Re0 + Reo*+ Reo: + Reo, +
Re20 Re202 %*O: Reno, R e z O a +,&%Ob + RerOs+ Re20T+,Re90~++ +
+
+
+
+
Gaseous oxides still were observed after they had been evaporated from the filament and the source had cooled to room temperature. As the source was warmed gradually the intensity of the oxide ion beams increased and the oxides were observed to fractionate with respect to each other. Since RezO, has the highest mass of any oxide observed, it must be a primary gaseous product. All the lower dimeric oxides are primarily fragmentation products of the ionizing electron beam. Relatively high electron energies are required to produce the lower mass dimeric ions. I n addition, a t a given electron energy the dimeric forms were found to remain in constant ratio to each other with large variations in time, temperature and vapor pressure. The highest mass monomeric oxide, ReOd, is also a primary gaseous product for it was frequently observed in the absence of oxides of greater mass. Although both Reo3+ and Reo2+ are formed in abundance by fragmentation of higher oxides, marked fractionation of Reo2+, Reo3+and Reo4+ with respect to each other and to the dimeric forms suggests that both Reoz and Reo3 have an independent gaseous existence. Reo+ and Re+ wcrc observed as fragmentation products only. It is of interest to note that the oxide ion Reoa+ appears a t masses 233 and 235. It is possible that, it may interfere with uranium isotopic aiialyscs when rhenium filaments are used in surface ionization sources. (1) Based on work performed under tho auspiees of the U. S Atomic Energy Commission. Presrnted in part at the Ninth Annual Meeting (June, 19611 of the A.S.T.M. Committee E14 on Mass Spectrometry. (2) D. B. Harrington. “Encyclopedia of Spectroscopy,” Reinhold Publ. Corp., New York, N. Y..1V60, pp. 628-1347, (3) M. 11. Studier, E. N. Sloth and L. P. Moore, J . Phys. Chem, 66, 133 (1962).