Largeamount of Work Has Been Done Inrecent Years Onrare-Earth/3-Diketonate
Largeamount of Work Has Been Done Inrecent Years Onrare-Earth/3-Diketonate
Largeamount of Work Has Been Done Inrecent Years Onrare-Earth/3-Diketonate
RARE-EARTH
TRISHEXAFLUOROACETYLACETONATES AND
RELATED COMPOUNDS*
M A R Y F R A N C E S R1CHARDSON'~, W I L L I A M F. W A G N E R and D O N A L D E. S A N D S
Department of Chemistry, University of Kentucky Lexington, Kentucky 40506
INTRODUCTION
A LARGE a m o u n t of w o r k has been done in recent years on rare-earth/3-diketonate
chelates, as many of these compounds have been found to exhibit laser activity
[1,2]. Among the compounds of interest are the hexafluoroacetylacetonates,
particularly the tetrakis salt and adducts of the tris chelate of europium [3,4]. The
tris chelates themselves, particularly the state of hydration, have been studied
very little. There is some indication that the extent of hydration changes as the
rare-earth series is traversed, as the lanthanum[3] and neodymium[5] chelates
are reported to form monohydrates, and the europium [3,6] and gadolinium [6, 7]
chelates form dihydrates. Several states of hydration might be expected to exist
for a given chelate, in analogy to the rare-earth acetylacetonates for which mono-
and trihydrated chelates are found for all rare-earths [8] and the dihydrated acetyl-
acetonate is found for the rare-earths from lantlmnum through samarium[9]. It
*Abstracted in part from a thesis submitted by M. F. Richardson to the Graduate School at the
University of Kentucky in partial fulfillment of the requirements for the degree of Doctor of Philos-
ophy.
tNational Science Cooperative Graduate Fellow, 1962-66; Present address: Aerospace Research
Laboratories ARC, Wright Patterson Air Force Base, Ohio 45433.
may be mentioned that Steinberg, Mashall, and Glassner[7] have found that
gadolinium hexafluoroacetylacetonate dihydrate is very stable with respect to loss
of water, and failed to lose water even when stored over P4Ot0 in v a c u o for several
months. This behavior contrasts to that of the rare-earth acetylacetonate hydrates,
which form the anhydrous chelate in the course of a few days in v a c u o over
M g ( C I 0 4 ) 2 [9].
Many tetrakis salts of the type QLn(hfac)4 have been produced by Melby,
Rose, Abramson, and Cads [4]. (Q is a substituted ammonium or alkali metal
ion, Ln is a rare-earth ion, and Hhfac is hexafluoroacetylacetone). Few attempts
have been made to traverse the series of rare-earths for a given Q in the tetrakis
salt. This is of interest because of the gradual decrease in the size of the rare-
earth ion as the atomic number increases, which may make possible a change in
co-ordination number and/or geometry. Lippard, Cotton, and Legzdins [ 10] have
evinced interest in problems of this type, and have determined the crystal structure
of the volatile salt, CsY(hfac)4. The yttrium ion is co-ordinated dodecahedrally by
the eight oxygen atoms from the four ligands. However, a square antiprismatic
arrangement is also a reasonable arrangement for eight ligand atoms [ 11], and has,
in fact, been found for tris(acetylacetonato)diaquoyttrium(III)monohydrate [ 12].
The present study was made in order to provide information regarding the
preparation and properties of rare-earth hexafluoroacetylacetonates. Infra-red
spectra and X-ray powder patterns have been obtained for the following series of
chelates: Ln(hfac)s.nH20, NHaLn(hfac)a'nH~O, pyH(hfac)4.nH20, and Ln(hfac)2.
CFsCO2-2H20, where pyH is the pyridinium ion.
EXPERIMENTAL
The rare-earth oxides were obtained in 99.9 per cent purity from the Michigan Chemical Company
and the Lindsay Chemical Company. Rare-earth oxides were converted to the chlorides by digestion
with 12 M HC1 and evaporation of the excess acid. The resulting chlorides were dissolved in water to
give solutions containing 1.0-1.2 M LnCIs. Hexafluoroacetylacetone, obtained from the Peninsular
Chemical Company, was shaken with water until precipitation of the dihydrate ceased; the solid was
filtered, washed with cold water, dried in air, and stored in a dark bottle until it was used in this form
for most preparations. Trifluoroacetic acid from Matheson Coleman and Bell was distilled before use.
Rare-earth trishexafluoroacetylacetonate hydrates, Ln(hfac)3.nH20, were prepared by a slightly
modified method of Halverson, Brinen, and Leto [3]. As an example, 10 ml of 1.2 M NdCI3 were placed
in a separatory funnel. An ethereal solution of NH4hfac was prepared by dissolving 8-7 g (0-036 moles)
of Hhfac.2H20 in 100 ml of ethyl ether and adding 2.5 ml NH4OH (-0.036 moles). The aqueous
NdCI3 solution was then extracted twice, each time with 50 ml of the ethereal NH4hfac. (Extraction
was not complete.) The organic phase was washed with 10 ml H20 and dried over Na2SO4. The ether
was evaporated in a stream of dry air. The resulting mixture of oil and crystals was extracted several
times with 100 ml portions of boiling hexane solvent. The extracts were combined and allowed to cool.
The resulting crystals were separated by filtration through a Buchner funnel, air dried for a few hours,
and stored in plastic-capped glass vials. Subsequent analyses showed that Nd(hfac)3.3H20 was the
product. Analogous trihydrated chelates were obtained for La and Pr, while this method gave dihy-
drated chelates, Ln(hfac)3.2HzO, for Sm through Lu. The yields of the tris chelate hydrates were
40-50 per cent. Analytical data for the tris chelates are shown in Table 1.
After the hexane extractions, a residue remained which was only slightly soluble in hot benzene,
indicating that it was not the tris chelate. The residue was found to consist of two compounds: the
ammonium tetrakis salt, NHiLn(hfac)4, (yield ~<1 per cent) and a mixed ligand chelate, Ln(hfac)oCF3
10. S.J. Lippard, F. A. Cotton and P. Legzdins, J . A m . chem. Soc. 88, 5930 (1966).
11. R. V. Parish, Coordn. chem. Rev. 1,439 (1966).
12. J. A. Cunningham, D. E. Sands and W. F. Wagner, lnorg. Chem. 6, 499 (1967).
Rare-earth trishexafluoroacetylacetonates and related compounds 1277
CO2"2H20 (yield 5-10 per cent). A single extraction of the residue with hot benzene served to remove
the ammonium tetrakis salt, thus leaving pure mixed ligand chelate as the final residue. Table 2 gives
the analytical data for these mixed ligand compounds.
Three methods were used to prepare NH4Ln(hfac)4. In the first, 1 g (~0.00125 mole) of the tris
chelate hydrate was dissolved in the minimum amount of benzene ( - 1 0 m l ) and 0.17ml Hhfac
(0.00125 mole) was added. Ammonia was then bubbled into the solution until precipitation ceased. The
second method involved dissolving 1 g of the tris chelate hydrate in 1 ml acetone and adding 1 ml of an
acetone solution which contained the stoichiometric amount of NH4hfac (0.17 ml Hhfac, 0-08 ml
conc. N H4OH). The acetone was evaporated and the residue recrystallized from benzene. Anhydrous
chelates were obtained in essentially quantitative yields by these two methods. The third method
% Ln.,O3 % H20 %C %H %F
Ln n Calc. Found Calc. Found* Calc. Found Calc. Found Calc. Found
*Percentage of water was determined by Karl Fischer Titrations; the average error is approxi-
mately --_0-1%.
CAs Ln~OT.
% Ln~O:~ % H~O% C %C %H %F
Ln Calc. Found Calc. Found* Calc. Found Calc. Found Calc. Found
*Percentage of water was determined by Karl Fischer titrations; the average error is approxi-
mately_+0-2%.
t A s Pr4OT.
1278 M . F . R I C H A R D S O N , W. F. W A G N E R and D. E. S A N D S
involved the reaction of the aqueous rare-earth chloride with aqueous NI-I4hfac, taken in a 1:4 mole
ratio of LnCI3 to NH4hfac. For instance, 5.8 g Hhfae.2H~O and 1-5 rnl NH4OH were mixed in water
and diluted to a volume of - 15 ml. This solution was added dropwise, with stirring, to 5 ml 1.2 M
SmCI3. The precipitate of the tetrakis salt was collected on a Buchner funnel. The yield of NH4Sm
(hfac)4 was 25 per cent (the high solubility of the tetrakis salt in water limits the yields). The anhydrous
tetrakis salts were obtained for Sm through Lu, while the monohydrated chelates, NH4Ln(hfac)4.H~O,
were obtained for La through Nd.
Two methods were used to prepare pyridinium rare-earth tetrakishexafluoroacetylacetonates,
pyHLn(hfac)4. One method was essentially the same as the second method for the ammonium salt; the
resulting compound was recrystallized from benzene or from a benzene-hexane mixture. The second
method involved mixing hot solutions of the tris chelate hydrate (1 g in 5 - 6 m l benzene) and the
stoichiometric amount of pyHhfac (0.17 ml Hhfac and 0.10 ml pyridine in 5-6 ml benzene). Enough
benzene was added to dissolve any precipitate which appeared. The solution was filtered and cooled,
whereupon the salt crystallized. Large, beautiful needles were obtained for the rare-earth chelates
from neodymium through lutetium, and plates were obtained for the lanthanum and praseodymium
compounds. The lanthanum and praseodymium salts were subsequently found to be monohydrates,
while the other salts were anhydrous. (The Nd chelate was converted to the monohydrate, however,
by storing pyHNd(hfac)4 in a humidity cabinet for a few hours.) Analytical data for the ammonium and
pyridinium tetrakis salts are given in Table 3.
It was necessary not to exceed the 1 : 1 ratio of pyridiaium hexafluoroacetylacetonate to the rare-
earth tris chelate for the lanthanum, praseodymium, and neodymium complexes if the tetrakis salt
was the desired product. Use of a 2 : 1 ratio resulted in the precipitation of (pyH)~Ln(hfac)~ (Ln is La,
Pr, or Nd). The powder pattern and infra-red spectrum of the lanthanum compound are given in
Tables 4 and 5, respectively. Anal. Calc. for (pyH)2La(hfac)~: La,zO.~, 12.22; H.,O, 0.0; C, 31.45: H,
1'27; N, 2.10; F, 42.7%. Found: La~zO3, 12-32; H~O, 0.0; C, 30.86; H, 1.22; N, 2.22; F, 42.8%.
The rare-earth oxide content was determined by converting the chelate to the oxalate, and then
igniting the oxalate to the oxide at I000° for an hour. Conversion to the oxalate was necessary because
of the volatility of the hexafluoroacetylacetonates. The water content in the samples was determined
by Karl Fischer titrations, using methanol as a solvent. The end point was determined visually. The
reagent was standardized before each use with sodium tartrate dihydrate. Carbon, hydrogen, nitrogen,
and fluorine analyses were performed by Crobaugh Laboratories in Charleston, West Virginia, by
Weiler and Strauss in England, and by Mr. Daryl Sharp and Miss Betty Snider at the University of
Kentucky, using a F & M Model 185 C H N analyzer.
It should be noted that titration with "old" Karl Fischer reagent gave a precipitate, which caused a
premature, rapidly fading end point. This end point continued to fade even after a considerable
amount of reagent, more than enough to titrate the water present, had been added. The precipitate was
found to contain the rare earth ion, but was not studied further. These difficulties with precipitation
were not met when freshly prepared reagent was used, so new reagent was made up periodically.
Infra-red spectra from 600 to 4000 cm -1 were obtained on a Beckman IR-8 spectrophotometer.
The polystyrene bands at 906, I ! 54, 1583, 2850, and 3027 cm -~ were used to calibrate the spectrum of
each compound. Spectra were obtained under conditions of fairly high resolution (3-4 cm -~ in the region
600-2000 cm-1; 4-6 cm -1 in the region 2000-4000 cm-~). All spectra were taken as mulls in Nujol
mineral oil, Fluorlube oil (Fisher LG-160), or hexachlorobutadiene. The last two mulling agents were
used in the regions in which Nujol absorbed strongly.
The powder patterns were determined on a Norelco X-ray diffraction unit. The powdered samples
were packed in 0.2 mm glass or 0.3 mm quartz capillary tubes. Photographs were taken on Kodak film
in a 114.6 mm Debye camera. The radiation was CuK~ filtered through a nickel foil. Line positions on
the film were determined to the nearest 0.05 mm. Intensities were estimated visually. The general
diffuseness of the back reflections made corrections for film shrinkage impossible in most cases;
however, in the few cases where good back reflections were available, correction for film shrinkage
was almost negligible.
Glove bags obtained from Instruments for Research and Industry were used to maintain a dry
atmosphere over compounds suspected of moisture sensitivity. The purging gas was air which had
been passed through columns of silica gel and magnesium perchlorate. All analytical operations,
including preparation of mulls for i.r. spectra and capillary tubes for X-ray powder patterns, were
carried out inside dry bags for these hygroscopic compounds.
Rare-earth trishexafluoroacetylacetonates and related compounds 1279
~ m. r--
~ .2.,
tr-~
+l
._E
,¢
t~
i _
Y.
"0
[-
~Z
.=_
¢o
'O
ZZZZZZ ~ m ~ m ~ ~
1280 M.F. RICHARDSON, W. F. WAGNER and D. E. SANDS
The attraction which serves to hold this third water molecule in the lattice of
the trihydrated chelate is an intriguing problem. A clue to the mode of interaction
may be given by the crystal structure of NH4La(hfacL.H20, whose determination
is now being undertaken by one of us (D.E.S.). The i.r. spectra show that the
hydrogens of the water molecule in this tetrakis salt are not hydrogen-bonded.
The hydrated rare-earth hexafluoroacetylacetonates could be sublimed
unchanged in vacuo at 100°. This was somewhat surprising (even though sublima-
tion of the neodymium chelate has been reported[l 3], as hydration normally has
an adverse effect on volatility and thermal stability[14]. It seems likely that the
hydrated chelate first loses water so that the anhydrous chelate actually is the
subliming species. The anhydrous chelate on the end of the cold finger may then
recombine with the water produced by the initial decomposition of the chelate.
This decomposition would be similar to that observed for the monohydrated rare-
earth heptafluorodimethyloctanedione chelates, Ln(fod)3"H20; Margrave[15]
has subjected these chelates to a mass spectrometric study and found a large peak
at m/e = 18 but no peak corresponding to the species Ln(fod)3.H20 +.
Ln(hfac)2CF:~CO2.2H20. These mixed ligand chelates were obtained as the
major by-product of the preparation of Ln(hfac)3.nH.,O. The yield of the mixed
chelates was increased by the addition of ammonium trifluoroacetate to the
ethereal ammonium hexafluoroacetylacetonate. After extracting the rare-earth
chloride solution and evaporating the ether in the usual manner, the residue was
extracted several times with small portions of hot benzene in order to remove
Ln(hfac)~.nH20 and NH4Ln(hfac)4 . The residue was then recrystallized from
large amounts of benzene.
The mixed chelates are extremely soluble in oxygen-containing organic
solvents, slightly soluble in benzene, and insoluble in hexane and in water.
The mixed chelates were obtained even when freshly recrystallized from water
hexafluoroacetlyacetone dihydrate was used for the preparation of the tris chelate.
The reaction of hexafluoroacetylacetone with water to give trifluoroacetic acid
and trifluoroacetone is essentially the reverse of the condensation reaction used to
prepare hexafluoroacetylacetone. Also, Fry and Pirie[16] have found that heat
and light decompose ethanolic europium benzoylacetonate; the ultimate reaction
products are ethyl acetate and acetophenone. The trifluoroacetate ester could not
form in the present system because of the absence of alcohol, but the trifluoro-
acetic acid produced could react with the tris chelate to give the observed
Ln(hfac)2CF3CO2.2H20. These mixed chelates are analogous to the acetyl-
acetonate-acetate mixed complexes, Ln(acac)2CH3CO2-H20, which are some-
times found during recrystallization of rare-earth acetlyacetonates[9]. Also,
Charles [17] has found that mixed ligand chelates of the type Eu(DBM)2X form
in preference to Eu(DBM)3 even when excess DBM is present (DBM is the
dibenzoylmethanate anion; X is acetate, propionate, or benzoate.)
13. R. W. Moshier and R. E. Sievers, Gas Chromatography of Metal Chelates, p. 35. Pergamon
Press, Oxford (1965).
14. C. S. Springer, Jr., D. W. Meek and R. E. Sievers, lnorg. Chem. 6, 1105 (1967).
15. J. L. Margrave, Rice University, Houston, Texas. Private communication, 16 June, 1967.
16. F. H. Fry and W. R. Pirie, J. chem. Phys. 43, 3761 (1965).
17. R. G. Charles, J. inorg, nucl. Chem. 26, 2195 (1964).
1282 M . F . RICHARDSON, W. F. WAGNER and D. E. S A N D S
°~
© ~ EEEE E E E~E~E~ ~E~E ~
£
'~" "g~ ~t~ t"q r'~ t'~ ~ 'O "~" ~ ~ ~ ~ eq ~ ~ a~° t"~
° ° ' ~~" " ~eq
" ~ ¢:~
r'" I"-.-
~ t---~ ~o ~.~
Rare-earth trishexafluoroacetylacetonates and related compounds 1283
Ca
z ~ '~-~ °
z Ca~8~
o o . ~
~'~ ~ ~'~ ~
Ca 0 Ca Ca
0 ~ ~ ¢-, ¢,,
"2,
ka
£
o~
14
Ca e-
Z
Ca
g_~ z
~aCa
ca.l=
Ca Ca
"~ooo
1284 M. F. RICHARDSON, W. F. WAGNER a n d D . E. S A N D S
T a b l e 5. I n f r a - r e d s p e c t r a l d a t a f o r s o m e r a r e - e a r t h
hexafluoroacetylacetonate d e r i v a t i v e s a' b
789 mst
771 mst
610w
1286 M . F . R I C H A R D S O N , W. F. W A G N E R and D. E. S A N D S
18. H. Bauer, J. Blanc and D. C. Ross, J. A m . chem. Soc. 86, 5125 (1964).
Rare-earth trishexafluoroacetylacetonates and related compounds 1287
different. It is possible that the same situation occurs in the ammonium tetrakis
salts, particularly since the i.r. spectra of the a- and B-forms of the lower rare-earth
tetrakis salts are different both from each other and from the spectra of the salts of
gadolinium through lutetium. Significant, albeit small, differences are found in
the 1600-1615, 1150-1260, 745-775, and 660-670 cm-' regions (Table 5). On
the other hand, the yttrium, erbium, and lutetium salts, which have a different
crystal structure from the gadolinium and holmium salts, have i.r. spectra identical
to those of the last two compounds. The lack of change in the i.r. spectrum
(especially with regard to band splittings) is probably indicative of a lack of change
in the molecular structure. GiUard, Silver, and Wood[19] have shown that the
degree of splitting of a particular group frequency is a function of the molecular
symmetry of the chelate.
The peaks enclosed in parentheses in Table 5 are those which are found in
hydrated NH4Ln(hfac)4, but which are absent after the sample has been dried
until the crystals turn opaque. It is interesting that the a- and B-forms of NH4Nd-
(hfac)4 seem to have different modes of interaction with the water of hydration, as
is evidenced by the differences in the O - H stretching region. The hydrated/3-
form has a sharp weak peak at 3680 cm -1 and a broad shoulder at 3480 cm-',
which shows that one of the water hydrogens is hydrogen-bonded but not the
other. The hydrated a-form (and also NH4Ln(hfac)4.HzO) has two sharp peaks, at
3700 and 3604 c m -1, which is fairly strong evidence that both of the water hydro-
gens are only very weakly hydrogen-bonded (the free O - H stretching frequencies
of gaseous water are at 3756 and 3657 cm -1 [20].
PyHLn(hfac)4.nHeO. Both monohydrated and anhydrous pyridinium salts
were found for lanthanum, praseodymium, and neodymium, but for the rare-earths
above neodymium only the anhydrous tetrakis salt was found. The stability of the
hydrated neodymium compound depends on the prevailing relative humidity, but
the lanthanum and praseodymium tetrakis salt hydrates are stable in air and only
lose water when stored in a desiccator over magnesium perchlorate.
The europium tetrakis salt fluoresces bright pink under u.v. light, while
Eu(hfac)3"2H20 exhibits only a weak pink fluorescence. On the other hand, the
terbium tetrakis salt fluoresces greenish-yellow, but not as strongly as does
Tb(hfac)a-2H20.
The X-ray powder patterns show that there are three isomorphous series of
pyridinium salts, the first for the hydrated chelates of lanthanum, praseodymium,
and neodymium, the second for the anhydrous salts of the rare-earths from samar-
ium through terbium, and the third for salts of the rare-earths from holmium
through ytterbium. The i.r. spectra of all of the anhydrous pyridinium tetrakis
salts are exactly alike, in spite of the structural change which occurs between
terbium and holmium. Again, we accept this as evidence that no significant
structural changes have occurred on the molecular level.
The i.r. spectra of the hydrated pyridinium salts have a relatively strong,
broad peak at 3445 cm-', due to hydrogen-bonded water, and a series of peaks at
3100-3300 cm-' due to the pyridinium ion. Dehydration of these salts results in
19. R. D. Gillard, H. G. Silver and J. L. Wood, Spectrochim. A cta 20, 63 (1964).
20. K. Nakamoto, Infrared Spectra of Inorganic and Coordination Compounds, p. 83. Wiley, New
York (1963).
1288 M . F . R I C H A R D S O N , W. F. W A G N E R and D. E. S A N D S
the disappearance of the peak at 3445 cm -1 as well as in slight shifts and increased
intensities of the bands at 3100-3300 cm -1. These 3100-3300 cm -1 bands have
been found to be strong when hydrogen-bonding by the pyridinium ion takes
place[21]. Thus, the decrease in intensity of these bands which occurs during
hydration may signify that the water molecules are being hydrogen-bonded in the
sites previously occupied by pyridinium ions.
A certain degree of regularity is evident in the behavior of all of the tris and
tetrakis chelates of the lower rare-earths, in that two states of hydration exist. It
is tempting to relate the addition of an "extra" molecule of water to the size of the
rare-earth ion by postulating co-ordination of the water molecule to the rare-earth
to give a nine-co-ordinate central ion. Co-ordination should become less likely as
the ionic size of the rare-earth decreases, which might explain why the upper
rare-eath tris and tetrakis salts do not add an extra water molecule. On the other
hand, there is some evidence against co-ordination, namely, the lack of hydrogen
bonding in at least some of the hydrated chelates. The crystal structure of
Y(acac)3.3H20[12], part of which is reproduced in Fig. 1 shows that two of the
water molecules (O1 and 0 9 ) are co-ordinated and the third (02) is hydrogen-
bonded to O1. However, the i.r. spectra [9] show that all of the water hydrogens
are hydrogen bonded in Y(acac)3.3H20. Examination of the structure suggests
that the oxygen atoms of the chelate rings are close enough to the co-ordinated
water molecules for hydrogen bonding to occur, and hydrogen-bonding between
water molecules and the acetylacetonate ring oxygens has been postulated by
Davis and Fackler for several transition metal acetylacetonates [22]. Thus, if this
argument is correct and hydrogen bonding does occur between the co-ordinated
water molecules and the oxygen atoms of the acetylacetonate ring, it is difficult to
see how the water molecule in NH4Ln(hfac)4.H20 could be co-ordinated but not
hydrogen bonded [23].
Tetrakis salts other than the hexafluoroacetylacetonates have been observed
to form adducts. Triethylammonium tetrakis(benzoyltrifluoroacetato)europate(I II)
has been reported as the monohydrate [2] and piperidinium tetrakis(dibenzoyl-
methanato)europate(III) forms an adduct with dimethylformamide[1]. In the
latter case, fluorescence spectra have given evidence that the europium ion is
nine-co-ordinate~l.
One other interesting series of compounds was prepared in the course of the
present study: the pentakis salts, (pyH)2Ln(hfac)5. These salts were found only
for lanthanum through neodymium; attempts to make the corresponding samarium
chelate met with failure. The powder pattern and i.r. spectrum of the lanthanum
pentakis salt are given in Tables 4 and 5, respectively. The crystal and mole-
cular structure of this compound should be quite interesting, since there is the
possibility that the rare-earth ion is ten-coordinated. To our knowledge, there is
only one other example of a ten-coordinated rare-earth complex, and that is
09 OI--2"-'?."." 02.
.%°
".,¢.
Fig. 1. Positions of the water molecules in Y(acac)3"3H20. The O1, 0 2 , 0 9 , O1', O2',
0 9 ' are oxygen from the water molecules; and the dotted lines represent hydrogen bonds.
The water molecules evidently stabilize a particular crystal structure (and thus
co-ordination number and/or geometry) because of hydrogen-bonding. It is also
interesting that the anhydrous pyHLn(hfac)4 series has only two crystal forms,
which may reflect the steric requirements of the pyridinium ion.
A c k n o w l e d g e m e n t s - W e appreciate partial support of this work by the U.S. Atomic Energy Com-
mission (Contract No. At(40-1)2124). We also thank the National Science Foundation for the
Cooperative Graduate Fellowship, 1962-1966, for Mary Frances Richardson.
24. M. D. Lind, B. Lee andJ. L. Hoard, J . A m . chem. Soc. 87, 1611 (1965).