8 668
HERBERT
H.H Y h L I N ,
IlARTIN
KILPATRICK ANI)
JOSEPH
I.KAT2
(CONTRIBUTIOK FROM THE C H E M I S T R Y D I V I S I O N , ARGONNEN A T I O N A L LABORATORY, A N D TliE I L L I N O I S INSTITUTE O F T E C H N O L O G Y ]
\rol. i9
DEPARTMENT O F CHEMISTRY,
The Hammett Acidity Function Ho for Hydrofluoric Acid Solutions’ B Y HERBERT H.HYMAN,n f A R T I N KILPATRICK A N D JOSEPH J. KATZ RECEIVED
FEBRUARY 28, 1957
The Hammett acidity function HOvalue has been determiued for hydrogen fluoride iti the concentration range from anhydrous material down t o about 40y0 hydrogen fluoride. The dryest hydrogen fluoride prepared had an Ho value of - 10.2. In the course of this investigation, improved spectrophotometric cells were developed for handling hydrofluoric acid solutions and the spectrum of anhydrous hydrogen fluoride in the near infrared was determined The effect of sodium fluoride and water on the near infrared hydrogen fluoride spectrum was investigated and an analytical method for water in hydrogen fluoride was developed based on measurements of the water absorption a t 1 95 p .
Liquid anhydrous hydrogen fluoride has long been recognized as a very acidic substance despite the fact that in dilute aqueous solution it behaves as a rather weak a ~ i d . ~A, wide ~ variety of compounds are readily soluble in liquid hydrogen fluoride. These solutions usually exhibit substantial ionic conductivity that can be attributed to proton transfer from the acid to the solute. Hammett has discussed the use of organic indicators for the determination of relative acidity and suggested an acidity scale based on the relative concentration of protonated and non-protonated species in the medium under i n ~ e s t i g a t i o n . ~ Paul and Longi recently have reviewed the determination and use of the Ho function. I n particular, the acidity of any solution on the Ho scale is determined by the fraction of the neutral indicator which is converted from an uncharged molecule t o a protonated species. Hammett also discussed the H- and H+ scales in which the original indicator was an anion or cation rather than a neutral molecule. For practical use of the Hoacidity functions, it is defined by the equation if0
=
pK,’
+ log (C,/CIH+ )
fur the indicator, I , is determined by a series of intercomparisms with indicators of decreasing fiKaI. The activity coefficient term for a constant charge type is supposed to be independent of the medium or indicator, and 130 approaches QH in dilute aqueous solution. Paul and Long have reviewed previous intercomparisons and extrapolations and suggest a best value for each indicator which has been studied. As they point out, these values are not entirely self-consistent within the expected limit of error. Different indicators will give different values for t h e same solution. Other workersq have yuestioned the general applicability of the I30 concept, or sought more satisfying acidity f~inctions.~I n spite of this, the Ho function is very useful in cor( 1 ) Based o n work performed under the auspices of the U .S. Atomic Energy Commission (‘2) J. H. Simons, “ F l u o r i n e Chemistry,” Chap. 6, Academic Press, h‘en York, N . Y.,1950, p. 2 2 5 . ( 3 ) L. Pailling, J . Chem. ( 4 ) I,. P. H a m m e t t , “ P h c m i s t r y . ” hfcGraw - F r i l l Hcmk Cn., K e w York, N.Y , 1RLO. ( 5 ) hI. A . Paul a n d F. A. Long. Chewt. Revs., i n press. (A) R G. Bates a n d G Schwarzenbach, H r ! u . C h i m . A c l u , 38, F!XI (1035). (71 U Giitbezahl a n d K C > r u n w ; i l J , l’iirs J u r r n N 4 i . , 7 6 , 5G2 (19331
relating and predicting for different media properties such as solubility or catalytic effects which depend on acidity. I n the highly acid, high. dielectric solvents, such as those considered here, it is probably fair to assume that a medium which shows a more negative Ho value for a specific. indicator is the more acidic medium. Hammett and his students determined the I& function for several aqueous acid solutions, including O-lGG% sulfuric a ~ i d . ~Lewis . ~ and Bigeleisen’” and more recently Brand” extended these measurements into the fuming sulfuric acid range. The development of a variety of cells with polychlorotrifluoroethylene windows, l 2 now make it possible t o study optical absorption spectra in hydrogen fluoride solutions with relative ease, and a determination of the Hammett acidity function for hydrofluoric acid solutions becomes practical. R. P. Bell has previously investigated the dilute aqueous range up to about 4G% hydrogen fluorideI3 and it has now been possible to complete the study of the mater-hydrogen fluoride system using the same indicators used by Hammett for his sulfuric acid study. Experimental Pdaterials.--The iridicators used iri this ivork w:rc -41drich Chemical Company products sold for use as Hnniiriett iiidicators and no further purification was attempted. T w o separate methods of purification were used for t h e hydrogen fluoride. For most of the earlier runs comrricrcia1 hydrogen fluoride was absorbed on sodium fluoride, aiid the resulting NaHF2 heated in vacuo a t 150” t o renio\’c volatile impurities. Hydrogen fluoride was theti regencrated by heating a t 300”. The material was stored over cubaltic fluoride in a nickel vessel. Hydrogen fluoride Tvai. distilled from this vessel as needed directly into a conductivity cell or an optical absorption cell. Hydrogen fluoride purified in this way showed conductivities in the range 4 tir G X IO-‘ ohm-’ a n . - ’ a t 0” indicating a water content crf O.Olyoor less. More recently a better quality acid has been obtained. As suggested b y Runner, Balog and Kilpatrkk, the hydrogen fluoride was purified by fractional distillatioil in an eficient c0lumn.1~ The product was collected mid handled entirely in polychlorotrifluoroethylene and goltl lined fittings thus avoiding contact with base metal (uickel). Under these conditions, batches of acid have been obtaiiied with conductivities below ohm-’ C I I I . - ~ . hLeasurat~ly liigher acidities (more negative 11” values) were found for .______ ( 5 ) I, .P. H a m m e t t and A . J . 13eyriip. i b i d . . 64, 2721 (193?), ( 0 ) I,. P. I i a m m e t t and h l . A Paul. i b i d . , 66, 827 (1931) (10) (;, K.Lewis a n d J. Higrleisen, i b i d . , 66, 1144-50 (lY43) ( I I ) J . C . D Brand. J . Clirm. Soc.. 997 (1950). ( 1 2 ) J J Katz and 1I XI. I l y m a n , l i e u . S r z . I w s t u . 24, J (13) IC. I’ XFII, IC. h‘ Udsc
