1974 1029On the Anomalously Large Electrical Conductivity of AmmoniumPerchlorate-Ammonia (1 /2)By Stanley V. Moore, Gari P. Owen, John M. Thomas,* and John 0. Williams, Edward Davies ChemicalA.c. measurements on the diammonia compound of ammonium perchlorate (AP) along with electrical data relatingto rubidium perchlorate, its ammonia compound, and liquid ammonia, suggest that the exceptional room tempera-ture conductivity of AP,2N H, arises from protonic migration.Laboratories, University College of Wales, Aberystwyth SY23 1 NEON cooling single crystals of ammonium perchlorate (AP) attributed i n transitu, to the electret properties oft o liquid nitrogen temperatures in the presence of amm0nia.l Upon their removal from the measuringcurrents were recorded. The effect was tentatively 469.ammonia gas anomalously large and 'Ornewhat erratic 1 D.C. Lain6 and R. C . Sweeting, Phys. Letters, 1971, 36A1030 J.C.S. Daltonsystem at room temperature, the (originally single-crystal) samples were found to be loosely packed, poly-crystalline, masses. The uptake of ammonia, recordedmicrogravimetrically in separate experiments, wasanomalously large and far in excess of that correspondingto either physical adsorption or chemi~orption.~J Theabove effects may be rationalized in terms of the reactionof AP with ammonia since it is now known that AP,394in common with other alkali metal (M) perch lor ate^,^forms compounds of the type MClO,,nNH, where m areintegers ranging from 1 to 6. Clearly a given ammoniacompound is stable under a definite equilibrium pressureof ammonia and addition of an excess of ammonia maylead to its dissolution in the compound.Confirmationof these changes is readily obtainable by simple opticalmicroscopy (see micrographs in refs. 2 and 3), and bymicrob alance techniques .The electrical conductivity of AP-ammonia com-pounds was considered to be particularly interestingbecause of the possibility of proton conduction in suchsystems, there being potentially facile proton transfer inthe NH4+ + NH, Bronsted conjugate pair. Measure-ments are reported here on the electrical conductivityof the system NH,C104,2NH3 in the range 257-298 Ktaken employing an ax. technique (1592 Hz).? Theammonia compound formation was achieved, knowingthe data of Fridman and Pa~tevich,~ by admission intothe measuring cell of the calculated amount of ammoniaand maintenance of the correct gas pressure during thecourse of the experiment.Inert, stainless steel, elec-trodes were employed with the ammonia compound (aliquid in the temperature range studied) contained in aTeflon cup.At room temperatures the a.c. conductivity of NH4-C104,2NH3 is ca. 10l2 times as large as that of an APdisc (Figure) and is characterized by an activationenthalpy of ca. 0.25 eV (24 kJ mol-I). Analogousmeasurements were carried out on rubidium perchlorate(RP). Even though it is not feasible, because of theslowness of the reaction, to identify the various ammoniacompounds of RP (not hitherto reported in the literature),the temperature variation of the conductivity of asample of RP after the establishment of an equilibriumpressure of ammonia was also recorded (Figure). Theconductivity is lower by a factor of some lo5 than thatof NH4C10,,2NH,.In suitable systems 6,7 large values of the absoluteconductivity together with a very low activation en-? A.c.methods were employed in preference to d.c. methods tominimize electrochemical deposition and polarization effects.1 Since this paper was submitted a report has appeared(A. Potier and D. Rousselet, J . Chim. Phys., 1973, 70, 873) onthe electrical conductivity of the closely analogous oxoniumperchlorate ([H30+] [ClO,-1). It has been established that in theorthorhombic Pnma phase (which is the same space group asthat of ammonium perchlorate), the exceptionally large intrinsicconductivity arises primarily from the facile transport of theproton.The proton transport mechanism is obviously verysimilar to that which is discussed here.thalpy of conduction strongly indicate a mechanisminvolving intermolecular (and possibly some intra-molecular) proton jumps. Recently Delpuech and co-workers 8 have computed by ab initio methods an energybarrier of ca. 0.1 eV for the proton jump from NH,+ toNH, in solution so that our roughly determined activa-tion enthalpy, which encompasses both the intra- andinter-molecular terms, suggests protonic motion. Fur-thermore, the difference between the AP and RPclloo 10[ arl(Q-lm-lK)cooling -&I I I I 2.5 3.0 3.5 4.01 0 3 ~TLog,, CJT US.lO3/T dependence * for NH,C10,-ammonia (1/2)(0) and RbC10,-ammonia under ammonia 400 Torr (a). Thebroken line refers to extrapolated a.c. measurements a t 1600Hz on an NH,C10, disc. The difference between heating andcooling cycles for RbC10,-ammonia is attributed to theinherent irreproducibility in this system.* A. B. Lidiard in ‘ Handbuch der Physik,’ vol. 20, ed. S.Flugge, Springer, Berlin, 1957; G. P. Owen, J. >I. Thomas, andJ . 0. Williams, J.C.S. Faraday I , 1972,2356.ammonia compounds together with the fact that theconductivity of NH4C10,,2NH, is ca. lo6 times as largeas that of liquid ammonia further supports the proposalthat protons from NH,’ species are the dominantcharge carriers.$We thank the Ministry of Defence for their support.[3/2037 Received, 4th October, 19731S. V. Moore, Ph.D. Thesis, University of Wales, Aberyst-wyth, 1974.3 S. V. Moore and J . M. Thomas, Proc. 7th Int. Symp. onReactivity of Solids, Bristol, 1972, eds. J . S. Anderson, M. 131.Roberts, and F. S. Stone, Chapman and Hall, London.A. I. Fridman and I. V. Pastevich, RRUSS. J . Inovg. Chew.,1968, 18, 60.C. Smeets, Nature W . Tijdschr., 1935, 17, 83,. 213.G. P. Owen, J. M. Thomas, and J. 0. Williams, J.C.S.7 T. M. Herrington and L. A. K. Staveley, J . Phys. Chew.* J. J. Delpuech, G. Serratrice, A. Strich, and A. Veillard,9 International Critical Tables, vol. 6, McGraw-Hill, NewDalton, 1972, 808.Solids, 1964, 25, 921.J.C.S. Chem. Comm., 1972, 817.York, 1929, p. 142