Weakly coordinating anions: Crystallographic and NQR studies of halogen-metal bonding in silver, thallium, sodium, and potassium halomethanesulfonates

Article abstract of DOI:10.1021/ic010715i

³⁵Cl, ^79,81Br, and ¹²⁷l NQR (nuclear quadrupole resonance) spectroscopy in conjunction with X-ray crystallography is potentially one of the best ways of characterizing secondary bonding of metal cations such as Ag⁺ to halogen donor atoms on the surfaces of very weakly coordinating anions. We have determined the X-ray crystal structure of Ag(O₃SCH₂Cl) (a = 13.241(3) A; b = 7.544(2) A; c = 4.925(2) A; orthorhombic; space group Pnma; Z = 4) and compared it with the known structure of Ag(O₃SCH₂Br) (Charbonnier, F.; Faure, R.; Loiseleur, H. Acta Crystallogr., Sect. B 1978, 34, 3598-3601). The halogen atom in each is apical (three-coordinate), being weakly coordinated to two silver ions. ¹²⁷I NQR studies on Ag(O₃SCH₂I) show the expected NQR consequences of three-coordination of iodine: substantially reduced NQR frequencies v₁ and v₂ and a fairly small NQR asymmetry parameter ν. The reduction of the halogen NQR frequency of the coordinating halogen atom in Ag(O₃SCH₂X) becomes more substantial in the series X = Cl < Br < I, indicating that the coordination to Ag⁺ strengthens in this series, as expected from hard-soft acid-base principles. The numbers of electrons donated by the organic iodine atom to Ag⁺ have been estimated; these indicate that the bonding to the cation is weak but not insignificant. We have not found any evidence for the bonding of these organohalogen atoms to another soft-acid metal ion, thallium. A scheme for recycling of thallium halide wastes is included.

Full text of DOI:10.1021/ic010715i

Inorg. Chem. 2002, 41, 2032−2040
Weakly Coordinating Anions: Crystallographic and NQR Studies of
Halogen−Metal Bonding in Silver, Thallium, Sodium, and Potassium
Halomethanesulfonates
Gary Wulfsberg,* Katherine D. Parks, Richard Rutherford, Debra Jones Jackson, Frank E. Jones,
Dana Derrick, and William Ilsley
National Center for Applications of NQR Spectroscopy in Inorganic Chemistry, Department of
Chemistry, Middle Tennessee State UniVersity, Murfreesboro, Tennessee 37132
Steven H. Strauss, Susie M. Miller, and Oren P. Anderson
Department of Chemistry, Colorado State UniVersity, Ft. Collins, Colorado 80523
T. A. Babushkina† and S. I. Gushchin‡
Institute of Biophysics, Ministry of Health, Moscow 123182, Russia
E. A. Kravchenko and V. G. Morgunov
Institute of General and Inorganic Chemistry, Russian Academy of Sciences,
117907 Moscow, GSP-1, Russia
Received July 5, 2001
35Cl, 79,81Br, and 127I NQR (nuclear quadrupole resonance) spectroscopy in conjunction with X-ray crystallography
is potentially one of the best ways of characterizing secondary bonding of metal cations such as Ag+ to halogen
donor atoms on the surfaces of very weakly coordinating anions. We have determined the X-ray crystal structure
of Ag(O SCH Cl) (a ) 13.241(3) Å; b ) 7.544(2) Å; c ) 4.925(2) Å; orthorhombic; space group Pnma; Z ) 4)
3
2
and compared it with the known structure of Ag(O SCH Br) (Charbonnier, F.; Faure, R.; Loiseleur, H. Acta Crystallogr.,
3
2
Sect. B 1978, 34, 3598−3601). The halogen atom in each is apical (three-coordinate), being weakly coordinated
to two silver ions. 127I NQR studies on Ag(O SCH I) show the expected NQR consequences of three-coordination
3
2
of iodine: substantially reduced NQR frequencies ν1 and ν2 and a fairly small NQR asymmetry parameter η. The
reduction of the halogen NQR frequency of the coordinating halogen atom in Ag(O SCH X) becomes more substantial
3
2
in the series X ) Cl < Br < I, indicating that the coordination to Ag+ strengthens in this series, as expected from
hard−soft acid−base principles. The numbers of electrons donated by the organic iodine atom to Ag+ have been
estimated; these indicate that the bonding to the cation is weak but not insignificant. We have not found any
evidence for the bonding of these organohalogen atoms to another soft-acid metal ion, thallium. A scheme for
recycling of thallium halide wastes is included.
Introduction
natometal cations, (porphyrinato)Fe+, and various transition
metal organometallic cations which are important industrially
as catalysts, and which require “vacant” coordination sites.
It is now acknowledged that there is no such thing as a
noncoordinating anion; the focus is now on seeking very
weakly basic, easily dissociated neutral ligands and weakly
coordinating anions1 that act as good leaving groups from
latent coordination sites.2 A symposium on weakly coordi-
nating anions was held at the spring 1998 American
Chemical Society meeting in Dallas.3
Chemists have long sought “noncoordinating” anions to
serve as inert counterions in salts of very elusive, very
reactive cations such as silylium ions, R3Si+, porphyri-
* To whom correspondence should be addressed. E-mail: wulfsberg@
mtsu.edu.
† Current address: A. N. Nesmeyanov Institute of Organoelement
Compounds, Russian Academy of Sciences, 117813 Moscow, Russia.
‡ Current address: Department of Chemistry, Perm State University,
Perm, Russia.
2032 Inorganic Chemistry, Vol. 41, No. 8, 2002
10.1021/ic010715i CCC: $22.00 © 2002 American Chemical Society
Published on Web 03/28/2002

Weakly Coordinating Anions
Weakly coordinating anions should be as nonbasic4 as
possible: they should have low charge and a large size over
which the charge can be dispersed. Examples of this type of
Table 1. Average Metal-Organohalogen Distances for Halocarbon and
Halocarborane Complexes of Silver, Silicon, Thallium, and Cesium
complex
d(M‚‚‚X)
excess da CN(Ag)b
ref
-
anion include methanesulfonate (CH3SO3 ), tetraphenyl-
Ag(CH3CB11Cl11)
Ag(CB11H6Cl6)
Ag(CB11H6Br6)
Ag(CB11H6I6)
2.882 Å
2.780
2.824
0.37 Å
0.27 Å
0.14 Å
0.10 Å
0.15 Å
0.16 Å
0.15 Å
0.86 Å
0.24 Å
0.40 Å
0.39 Å
0.32 Å
0.20 Å
0.00 Å
0.43 Å
0.31 Å
6
6
6
5
c
d
d
d
e
e
e
f
c
g
g
h
h
h
borate ([B(C6H5)4]-), and 1-carba-closo-dodecaborate
(CB11H12)-.5 Coordinating ability of the anion is further
reduced by substituting the outside surface of the anion with
weakly coordinating functional groups containing very
electronegative atoms such as halogen, to give anions such
2.946
iPr3Si(CB11H6Cl6)
iPr3Si(CB11H6Br6)
iPr3Si(CB11H6I6)
2.323 Å
2.479 Å
2.661 Å
Tl(CB11H6Br6)‚2C7H8 3.469 Å
Cs(HCB11Br11) 3.733 Å
-6
as CF3SO3 and [B(C6F5)4]-.7 For the most important
Cs(HCB11I6Br5)‚THF 3.89 Å (Br)
Cs(HCB11I6Br5)‚THF 4.07 Å (I)
industrial application requiring weakly coordinating anions,
the metallocene process of producing stereoregular polym-
erization of alkenes,8 the most active catalyst of all is
obtained when (CB11H6X6)- (X ) Cl, Br, I) anions are
present.9 One might postulate that this is connected with the
fact that most weakly coordinating anions now being
investigated have fluorinated surfaces, and fluorine is a hard
base, more likely to coordinate to the (presumably) hard
zirconium(IV) active site in the metallocene catalysts, while
the carborane anions have softer surfaces. If this postulate
is valid, then the softest carborane anion, the iodinated one,
should show the weakest coordination to zirconium and,
conversely, the strongest coordination to the (more easily
isolated) silver salt. The crystal structures of all six silver
salts Ag(CB11H6X6) and Ag(CH3CB11H5X6) have been
determined,10 but the varying numbers, types, and bond
distances of donors to silver did not present an easily
analyzed pattern.
[Ag(Cl2CH2)n]+
[Ag(Br2CH2)n]+
[Ag(I2CH2)n]+
Ag(ClCH2SO3)
Ag(BrCH2SO3)
2.832 Å
2.865 Å
2.851 Å
2.945 Å
2.971 Å
6
6
4
6
6
this work
i
a Average distance in excess of M-X single bond distance ) sum of
covalent radii except in the case of Cs+, for which the single-bonded metallic
radius is used. b Coordination number of Ag+. c Ref 13a. d Ref 10a. e Ref
9b. f Ref 18. g Ref 13b. h From sources cited in Table 1 of ref 31. i Ref 16.
secondary bonding correspond to such a shallow potential
well that the differences merely represent variations in what
is needed to achieve good packing at the lowest energy in
the solid state?
The normal criterion for weak secondary bonding is that
the metal-donor distance should be greater than the sums
of single covalent or ionic radii but less than the sums of
van der Waals radii. But van der Waals radii are notoriously
hard to determine, because one must first be confident that
there is indeed no bonding in the direction in which the
contacts are measured. As an illustration of the difficulties,
solid halocarbons commonly pack with halogen-halogen
distances that are less than the sums of their van der Waals
radii. Consequently, one must ask whether the van der Waals
radii are incorrect, or differ in different directions around
an organohalogen atom, or whether there are hitherto-
unsuspected secondary bonding interactions present.12
In Table 1, we summarize metal-halogen distances found
in silver, silicon, thallium, and cesium salts of weakly
coordinating anions and related ligands.13 We can attempt
to analyze these distances by computing the excess bond
distance: the amount that the observed Ag-X distance is
in excess of the distance expected for a normal single bond
between Ag and X. For a neutral halogen atom in contact
with a cation, the types of radii that should be summed to
give the expected distance are not entirely clear, but the most
sensible results are obtained from the sums of covalent radii,
which are 2.51 Å for Ag-Cl, 2.66 Å for Ag-Br, and 2.85
Å for Ag-I. How much longer than the sums of covalent
radii must the metal-halogen contact be before we can
decide that it no longer represents coordination? Overall, it
may be seen that excess bond distances are broadly similar
for Ag, Cs, and Si, although cations of these elements would
be expected to bond in quite different ways to halogens.
Although X-ray crystallography is the most readily avail-
able and most commonly used method of detecting weak
secondary bonding11 of ligands to metal ions, it leaves several
important questions unanswered. Secondary bond distances
commonly show wide ranges, even in chemically equivalent
bonds in the same complex (e.g., from 2.640 to 2.926 Å in
Ag(CB11H6Cl6)). Is there any bonding significance to the
differences in contact distances commonly found, or does
(1) Strauss, S. H. Chem. ReV. 1993, 93, 927-42. Seppelt, K. Angew.
Chem., Intl. Ed. Engl. 1993, 32, 1025-1027.
(2) Strauss, S. H. Chemtracts: Inorg. Chem. 1994, 6, 1-13.
(3) Dagani, R. Chemical Superweaklings. Chem. Eng. News 1998, 76 (May
4), 49-54. This work was first presented in that symposium (Wulfs-
berg, G.; Rutherford, R.; Jackson, D.; Jones, F.; Jones, M.; Derrick,
D.; Strauss, S.; Terao, H.; Babushkina, T. A.; Gushchin, S. I. Presented
at the National Meeting of the American Chemical Society, Dallas,
TX, March 1998; Paper INOR 102).
(4) Wulfsberg, G. Inorganic Chemistry; University Science Books:
Sausalito, CA, 2000; pp 68-73, 116-120.
(5) Reed, C. A. Acc. Chem. Res. 1998, 31, 133-139.
(6) Lawrance, G. A. Chem. ReV. 1986, 86, 17-33.
(7) Lambert, J. B.; Zhang, S.; Stern, C. L.; Huffman, J. C. Science 1993,
260, 1917-1918.
(8) Jordan, R. F. AdV. Organomet. Chem. 1991, 32, 325-382. Brintzinger,
H. H. Fischer, D.; Mu¨lhaupt, R.; Rieger, B.; Waymouth, R. M. Angew.
Chem., Intl. Ed. Engl. 1995, 34, 1143-1170. Deutsch, C. H. Finding
Flexibility in Plastics: High Technology Could Add New Life to an
Old Product. New York Times, Sept 9, 1997, p C1, C6.
(9) (a) Xie, Z.; Bau, R.; Reed, C. A. Angew. Chem., Intl. Ed. Engl. 1994,
33, 2433-2434. (b) Xie, Z.; Manning, J.; Reed, R. W.; Mathur, R.;
Boyd, P. D. W.; Benesi, A.; Reed, C. A. J. Am. Chem. Soc. 1996,
118, 2922-2928.
(12) Desiraju, G. R.; Parthasarathy, R. J. Am. Chem. Soc. 1989, 111, 8725-
8726.
(13) (a) Xie, Z.; Tsang, C.-W.; Sze, E.; Yang, Q.; Chan, D.; Mak, T. C.
W. Inorg. Chem. 1998, 37, 6444-6451. (b) Tsang, C.-W.; Yang, Q.;
Sze, E.; Mak, T. C. W.; Chan, D.; Xie, Z. Inorg. Chem. 2000, 39,
5851-5858.
(10) (a) Xie, Z.; Wu, B.-M.; Mak, T. C. W.; Manning, J.; Reed, C. A. J.
Chem. Soc., Dalton Trans. 1997, 1213-1217. (b) Xie, Z.; Tsang, C.-
W.; Xue, F.; Mak, T. C. W. J. Organomet. Chem. 1999, 577, 197-
204.
(11) Alcock, N. W. AdV. Inorg. Chem. Radiochem. 1972, 1, 1-58.
Inorganic Chemistry, Vol. 41, No. 8, 2002 2033

Wulfsberg et al.
Table 2. Elemental Analyses of Halocarbon Derivativesa
%C %H
calcd found calcd found calcd found
Hence, determining whether a long metal-donor atom
contact actually signifies coordinate covalent (secondary)
bonding requires confirmation by another method. Most
spectroscopic methods cannot reliably detect the subtle
bonding changes that follow the weak coordination of a
halogenated anion or other ligand to a metal ion. We propose
that, for ligands or anions in which surface donor atoms are
chlorine, bromine, or iodine, the method of choice to
supplement X-ray crystallography for characterizing weak
bonding may be halogen (35Cl, 79,81Br, 127I) nuclear quadru-
pole resonance (NQR) spectroscopy.
%hal
compound
AgBrCH2SO3
4.26 4.24 0.72
5.64 5.63 0.95
3.17 2.40 0.53 <0.5
0.67 28.35 25.91
0.93 37.50 37.12
KBrCH2SO3
TlBrCH2SO3
KClCH2SO3
AgClCH2SO3
TlClCH2SO3
AgICH2SO3
21.12 20.91
7.12 7.05 1.20
5.06 5.26 0.85
3.60 3.65 0.60
3.63 2.73 0.61
1.08 21.02 21.12
0.59 14.93 12.73
0.59 10.62 10.31
0.62 38.59 34.01
1.42 28.30 30.34
0.82 48.79 47.66
TlICH2SO3‚1/12C12H24O6 5.36 5.23 1.12
KICH2SO3
4.62 4.66 0.78
a NaClCH2SO3‚1/3H2O weight loss upon dehydration (between 30 and
131 °C): Calcd, 3.79%. Found by TGA, 3.94%.
A limitation of NQR spectroscopy is the requirement for
gram-scale samples of polycrystalline solid materials; such
quantities are not affordable or even available for Ag-
(CB11H6X6).14 Hence, to begin the study of the metal-ion
coordinating characteristics of weakly coordinating haloge-
nated anions, we have substituted a much more affordable
in our experience the percent halogen obtained in silver compounds
of halocarbons is always somewhat low. TGA measurements were
obtained on a Perkin-Elmer TGA-7 thermogravimetric analyzer
between ambient temperature and 200 °C. Proton nuclear magnetic
resonance (1H) data were obtained in D2O solution using sodium
3-trimethylsilylpropionate-2,2,3,3-d4 (TMSP) as internal reference
on a Bruker AC 200 MHz spectrometer equipped with a FT Aspect
3000 signal processing system. 35Cl NQR spectra were measured
at 77 K on a Decca NQR spectrometer.19 79Br, 81Br, and 127I NQR
spectra were measured on a Wilks NQR-IA instrument20 at ambient
temperature and then were more accurately recorded at 77 K and
room temperature on a pulse IS-3 NQR spectrometer. 127I NQR
measurements of ν1 (∼300 MHz) and their temperature dependence
were made in Japan on a superregenerative NQR spectrometer with
Zeeman modulation, the oscillator of which had a tank circuit with
a small Lecher-line in addition to the usual LC parts, or on an ISSH-
1-12 pulse NQR spectrometer (SKB IRE, Russia);21 those obtained
above 360 MHz were measured on a homemade pulsed NQR
spectrometer.22 Since doing this work, we have acquired a state-
of-the-art RITEC pulse Fourier transform 35Cl NQR spectrometer,23
which has been built to implement two-dimensional nutation NQR
spectra (useful in obtaining asymmetry parameters).24
Crystallographic Study. The diffraction pattern from a colorless
crystal (0.12 × 0.16 × 0.20 mm3) was studied on a Bruker R3
diffractometer using the Bruker-Nonius XSCANS software. The
unit cell constants reported were determined by refinement against
the setting angles of 25 widely spread reflections, and the intensity
data were measured by θ/2θ scans to a maximum in 2θ of 55°. In
addition to Lorentz and polarization corrections, a semiempirical
absorption correction based on ψ scans of selected reflections over
a range of setting angles was performed.
-
set of anions, the halomethanesulfonate anions, ICH2SO3
-
-
(1-), BrCH2SO3 (2-), and ClCH2SO3 (3-). Fluorinated
anions of this type are suggested for use as noncoordinating
anions in the “Notice to Authors” of this journal;15 results
on related weakly coordinating anions will be reported
elsewhere. Although the oxygen ends of these anions are
the main sites of coordination, the crystal structure of Ag-
(2)16 shows that it has Ag-Br contacts very similar to those
found in Ag(CB11H6Br6). In this study, we have determined
the X-ray structure of Ag(3) and find not only that it is
isomorphous and isostructural with Ag(2) but also that it
shows very similar Ag-Cl contacts to those found in Ag-
(CB11H6Cl6). Although twinning problems prevented us from
determining the structure of Ag(1), we shall show that the
NQR evidence suggests that it also shows similar charac-
teristics to silver chloro- and bromomethanesulfonate. Study
of the NQR spectra of salts of the halomethanesulfonate
anions should show what happens when an anion generally
expected to be weakly coordinating to typical cations such
as Na+ and K+ does coordinate to a soft cation such as Ag+
via halogen atom(s).
The thallium(I) ion is another classical +1-charged soft-
acid metal ion which has been found by X-ray crystal-
lography17 to coordinate to organohalogen atoms of a weakly
coordinating ligand and to halogen atoms in the weakly
coordinating anion (CB11H6Br6)-.18 This study includes an
NQR investigation of the coordinating ability of thallium(I)
ion in its halomethanesulfonate salts.
The structure was solved by direct methods (Bruker-Nonius
SHELXTL) and refined by full matrix least squares techniques on
F. The final model employed anisotropic displacement coefficients
to model motion of all non-hydrogen atoms. The single crystallo-
graphically unique hydrogen atom was placed in an idealized
position and allowed to ride on the carbon atom to which it was
attached. The final electron density map was reasonably flat, with
the only significant peak (1.89 e Å-3) being due to residual electron
density from the Ag(I) cation.
Experimental Section
All reactions involving the photosensitive silver and thallium
halomethanesulfonates were carried out with red-light illumination.
Elemental analyses were obtained from Galbraith Laboratories,
Knoxville, TN, and are shown in Table 2. It should be noted that
(19) Smith, J. A. S. J. Sci. Instrum. 1968, 1, 8-14.
(20) Truett, W. L. Prepr. Symp.sAm. Chem. Soc., DiV. Fuel. Chem. 1967,
11, 303-307.
(21) Pavlov, B. N.; Safin, I. A.; Semin, G. K.; Fedin, E. I.; Shtern, D. Ya.
Vestn. Ross. Akad. Nauk 1964, 40.
(14) Reed, C. A. Personal communication.
(15) Notice to Authors. Inorg. Chem. 2001, 39 (1), 14A.
(16) Charbonnier, F.; Faure, R.; Loiseleur, H. Acta Crystallogr., Sect. B
1978, 34, 3598-3601.
(22) Gushchin, S. I.; Kolchanov, V. N. Radiospektroskopiya (Perm) 1978,
110-111.
(17) Hurlburt, P. K.; Anderson, O. P.; Strauss, S. H. Can. J. Chem. 1992,
70, 726-731.
(23) Petersen, G. L.; Bray, P. J.; Marino, R. A. Z. Naturforsch., A: Phys.
Sci. 1994, 65-70.
(18) Mathur, R. S.; Drovetskaya, T.; Reed, C. A. Acta Crystallogr., Sect.
C 1997, 53, 881-883.
(24) Harbison, G. S.; Slokenbergs, A.; Barbara, T. H. J. Chem. Phys. 1989,
90, 5292-5298.
2034 Inorganic Chemistry, Vol. 41, No. 8, 2002

Weakly Coordinating Anions
Materials. Insoluble silver(I) sulfite was precipitated upon
reaction of solutions of AgNO3 and Na2SO3. Slightly soluble
thallium(I) sulfite was synthesized from waste thallium halides,
which were first boiled with ethanol and concentrated HCl to
convert any soluble thallium compounds to insoluble TlCl, while
leaving organic contaminants in the ethanol. The mixed halides
were filtered off, dried, and then heated with concentrated H2SO4
to drive off HCl(g), HBr(g), and I2(g), while producing soluble Tl2-
SO4 and perhaps some TlHSO4. Yellow Tl2SO3 in good purity was
precipitated using an excess of 1.0 M freshly opened Na2SO3.25
Calcd % SO32- in Tl2SO3: 16.4%. Found: 15.4% iodimetrically.26
Because thallium sulfite and sulfate have similar solubilities, the
sulfate contaminates product made from older, air-oxidized Na2-
SO3.
filtered, and the precipitate was washed with five 20-mL portions
of dry methanol, leaving behind the expected amount of AgCl.
Evaporation of the combined filtrate in a vacuum desiccator gave
a total of 3.01 g (12.7 mmol, 74% yield) of crystalline product.
Results
Preparation of Compounds. Halomethanesulfonates are
usually prepared by refluxing solutions of sulfite ion with
dihalomethanes:
SO32- + CH2X2 f XCH2SO3- + X-
(1)
We synthesized Na(1)‚H2O and K(1) in this manner using
the procedure of Ossenbeck.27 We also prepared the thallium
salt from Tl2SO3 and CH2I2 using 5 mol % of 18-crown-6
as a phase transfer catalyst, but purer product was obtained
almost as quickly by omitting the crown ether. The corre-
sponding reaction using Ag2SO3 and 18-crown-6 gave very
low yields, presumably because of the low solubility of Ag2-
SO3. Although Ag(1) has previously been prepared by the
neutralization of Ag2CO3 with ICH2SO3H (obtained by ion-
exchange from Na(1)),28 we found it more convenient to react
Na(1) with AgNO3 in hot ethanol, filtering off the precipi-
tated NaNO3 and allowing crystallization of the product as
the filtrate evaporated. All iodomethanesulfonates were
Tl(1). Freshly prepared Tl2SO3 (4.53 g, 9.28 mmol) was dissolved
in 50 mL of deionized water in a Schlenk flask under Ar in an
∼85 °C hot water bath. This solution was stirred with CH2I2 (4.00
g, 14.9 mmol); a yellow precipitate of TlI gradually formed during
the first hour. After 2 h, 3.01 g (9.09 mmol) of TlI was filtered off
and washed with 30 mL of hot water. The volume of the combined
filtrate was reduced in a vacuum desiccator, producing at first 0.693
g mixed yellow and white crystals (possibly including some
unreacted Tl2SO3). After filtering these off, the volume was reduced
nearly to dryness to give 3.06 g (7.19 mmol, 77.5% yield) of white
Tl+(1)-, which was washed with 1.5 mL of cold deionized water.
1
The product was characterized successfully by H NMR in D2O
and NQR.
Tl(1)‚1/nC12H24O6 (n ≈ 12). Tl2SO3 (4.89 g, 10.0 mmol) in 20
mL of deionized water in a Schlenk flask under Ar was mixed
with CH2I2 (4.02 g, 15.0 mmol) and 18-crown-6 (0.13 g, 0.49
mmol); the mixture was stirred in the absence of light for 2 days.
However, no reaction was observed until the solution was heated
to about 85 °C in a hot water bath for about 1 h, whereupon TlI
precipitated, was filtered off, and was washed with 20 mL of hot
deionized water. The filtrate was allowed to evaporate; the 2.90 g
of air-dried crystals were then rinsed with 10 mL of ethanol and
then with 50 mL of hot methanol to give 2.00 g of the final product.
1
characterized by H NMR peaks at δ 4.39-4.41 in D2O.
Na(2) was synthesized by refluxing aqueous Na2SO3 and
CH2Br2 for 4 days,29 but we could find no satisfactory way
to separate it from the byproduct NaBr by fractional
crystallization. Refluxing aqueous K2CO3, CH2Br2, and 3 mol
% 18-crown-6 for 2 days produced K(2); fractional crystal-
lization from water produced first KBr crystals and then the
product, which was recrystallized from water. Tl(2) was
similarly produced by refluxing CH2Br2 and aqueous Tl2-
SO3 and was easily separated from insoluble TlBr and
unreacted, slightly soluble Tl2SO3; more complete reaction
but less pure product resulted from the addition of catalytic
amounts of 18-crown-6. Our synthesis of Ag(2) was a
modification of the procedures of Charbonnier16 and Senning:28
we passed our unseparated NaBr-Na(2) mixture through a
strongly acidic cation-exchange resin to give a mixture of
HBr and BrCH2SO3H. This was reacted with Ag2O, AgBr
was filtered off, and the filtrate was slowly evaporated to
give crystalline Ag(2). The reaction of this salt with KBr
was used to prepare analytically pure K(2). All bromo-
methanesulfonates were characterized by 1H NMR peaks at
δ 4.43-4.44 in D2O.
1
The washed product gave two H NMR peaks of roughly equal
intensity at δ 4.41 (methylene) and at at δ 3.70 (crown ether) in
D2O solution. Elemental analysis (Table 2) also confirmed the
1
presence of a small amount (about /12 mol) of 18-crown-6. This
product gave very strong 127I NQR signals, which were different
from those of crown ether-free Tl(1), and which were too strong
to be due to the small possible yield of the crown ether adduct for
which n ) 1.
Ag(1). Solutions of 2.68 g (16.8 mmol) of AgNO3 in 20 mL of
hot ethanol and of 4.11 g (15.7 mmol) of Na(1)‚H2O in 165 mL of
hot ethanol were mixed and then allowed to cool to room
temperature. The solution was filtered to remove precipitated
NaNO3, and the filtrate was placed in a vacuum desiccator. After
12 h, crystals had appeared; the mother liquor was decanted from
1
1.62 g (4.9 mmol, 31% yield) of product which gave a single H
The 7 day reaction of aqueous Tl2SO3 and CH2Cl2 in the
presence of 18-crown-6 did not prove a satisfactory method
of preparing Tl(3) although reaction occurred and NMR
spectra suggested the formation of the crown ether adduct
[Tl(18-crown-6)](3). The thallium salt is produced instead
by the reaction of ClCH2SO2Cl (Alfa) with Tl2CO3 in
NMR peak in D2O at δ 4.44. Evaporation of the mother liquor
gave an additional 3.34 g of product, the 1H NMR of which
indicated some contamination.
Ag(3). Chloromethanesulfonyl chloride (Alfa, 2.55 g, 17.1 mmol)
was placed in 40 mL of dry methanol in a Schlenk flask under Ar.
Ag2O (3.97 g, 17.1 mmol) was added. Stirring overnight caused
the dark color of the Ag2O to disappear. The solution was then
(27) Ossenbeck, A.; Hecht, G. Iodomethane Sulphonic Acid and Homo-
logues Thereof. U.S. Patent 1,842,626, Jan 26, 1932.
(28) Senning, A.; Buchholt, H. C.; Bierling, R. Arzneim.-Forsch. 1976,
26, 1800-1809.
(29) Truce, W. E.; Abraham, D. J.; Son, P. J. Org. Chem. 1967, 32, 990-
995.
(25) Seubert, K.; Elten, M. Z. Anorg. Allg. Chem. 1892, 2, 434-436. Reuter,
B.; Levi, H. W. Z. Anorg. Allg. Chem. 1952, 270, 100-113.
(26) Standard Methods for the Examination of Water and Wastewater, 14th
ed.; APHA-AWWA-WPCF: Washington, DC, 1975; pp 508-9.
Inorganic Chemistry, Vol. 41, No. 8, 2002 2035

Wulfsberg et al.
methanol overnight.
ClCH2SO2Cl + Tl2CO3 f TlCl(s) + Tl(3) + CO2(g) (2)
Silver salt Ag(3) was similarly prepared using Ag2O in place
of Tl2CO3; this synthesis was simpler than that earlier used
by Senning.28 The sodium salt was similarly prepared using
Na2CO3 and Ag2O and was extracted from insoluble AgCl
using methanol.
2ClCH2SO2Cl + Ag2O + Na2CO3 f
2AgCl(s) + 2Na(3) + CO2(g) (3)
Potassium salt K(3) was prepared from the silver salt by
reaction with aqueous KCl, or in the same manner as the
sodium salt. Chloromethanesulfonates were characterized by
1H NMR peaks at δ 4.50-4.52 in D2O.
Figure 1. Thermal ellipsoid plot (50% probability) of the structure of
compound Ag(3), with atom labeling indicated. Hydrogen atoms are drawn
as spheres of arbitrary radius. Bond lengths and angles are given in Table
4.
The use of 3 mol % of the phase transfer catalyst 18-
crown-6 seemed to speed up the reaction of aqueous sulfite
with dihalomethanes. The catalyst gave trouble in the case
of thallium salts, which seem to form substantial amounts
of crown-ether adducts of stoichiometry Tl(XCH2SO3)‚1/n-
(C12H24O6), which give distinctive NQR spectra and which
are difficult to separate from uncomplexed Tl(XCH2SO3).
Both NMR spectra and elemental analysis suggested that n
could be as high as 12, although the results were not
definitive because this product was not obtained pure. Using
1 mol of 18-crown-6, the pure, more normal crown ether
adduct Tl(2)‚C12H24O6 (i.e., n ) 1) was produced; this did
not give the NQR spectrum of the product for which n was
about 12 (81Br ν ) 236.9 MHz at rt).
Crystal Structure of Ag(3). The crystal structure of Ag-
(3) was determined at 173 K so that it could be compared
with the previously published structure of Ag(2);16 the two
compounds are isomorphous and isostructural. Unfortunately,
the crystals of Ag(1) were twinned, so their crystal structure
could not be determined; later in the paper, we present
evidence from 127I NQR spectroscopy which suggests that
it at least has a very similar structure.
Table 3. Crystallographic Data for Ag(O3SCH2Cl)
chemical formula
a
b
c
V
Z
Ag(O3SCH2Cl)
13.241(3) Å
7.544(2) Å
4.925(2) Å
492.0(3) Å3
4
fw
237.4
space group
T
Pnma (No. 62)
173 K
λ
0.7107 Å
Dcalcd
µ
3.20 g cm-3
0.49 cm-1
R1
wR2
0.058 (observed reflections)a
0.084 (all data)b
2 2
a R1 ) ∑||Fo| - |Fc||/∑|Fo|. b wR2 ) {∑[w(Fo2 - Fc2)2]/∑[w(Fo ) ]}1/2
Table 4. Bond Lengths (Å) and Angles (deg) with esd’s
.
Ag(1)-O(1)
Ag(1)-O(2A)
Ag(1)-Cl(1)
S-O(1)
2.436(4)
2.367(4)
2.9454 (15)
1.465(6)
1.449(4)
2.367(4)
Ag(1)-O(1A)
Ag(1)-O(2B)
Cl(1)-C
2.436(4)
2.367(4)
1.799(8)
1.499(4)
1.764(9)
2.436(4)
S-O(2)
S-O(2C)
S-C
O(1)-Ag(1B)
Table 3 presents a summary of the crystallographic data
for Ag(3); Table 4 summarizes the bond lengths and bond
angles. Figure 1 shows the labeling of the atoms and the
thermal ellipsoids. Figure 2 illustrates the coordination
environment of Ag+ in the salt. The silver(I) ion is ap-
proximately octahedrally coordinated by four oxygen donor
atoms from sulfonate groups at distances of 2.367 and 2.436
Å and two chlorine donor atoms from chloromethane groups
at a distance of 2.945 Å. All bond angles at Ag+ are within
5° of the ideal angles. The primary bond lengths are
unremarkable. As is commonly the case, there is no
significant lengthening of the primary C-Cl bond upon
formation of the secondary bonds to Ag+. The structure is
in striking contrast to the layer type of structure commonly
found for halogenated and nonhalogenated organosulfonates
of the hard-acid group 1 and 2 cations, which feature a layer
of metal ions surrounded exclusively by hard-base oxygen
donor atoms from two layers of sulfonate groups (and
sometimes water), leaving the organic group (and terminal
halogens if present) in a layer far from the metal ion, but
O(2)-Ag(1A)
O(1)-Ag(1)-O(1A)
O(1A)-Ag(1)-O(2A)
O(1A)-Ag(1)-O(2B)
O(1)-S-O(2)
180.0(1) O(1)-Ag(1)-O(2A)
94.3(2) O(1)-Ag(1)-O(2B)
85.7(2)
94.3(2)
85.7(2) O(2A)-Ag(1)-O(2B) 180.0(1)
112.6(2) O(1)-S-C
105.9(2) O(1)-S-O(2C)
115.2(3) C-S-O(2C)
103.3(4)
112.6(2)
105.9(2)
O(2)-S-C
O(2)-S-O(2C)
Ag(1)-O(1)-S
129.3(1) Ag(1)-O(1)-Ag(1B) 101.5(2)
S-O(1)-Ag(1B)
Cl(1)-C-S
129.3(1) S-O(2)-Ag(1A)
109.4(4) C-Cl(1)-Ag(1)
79.6(2)
131.9(3)
105.62(19)
Ag(1)-Cl(1)-Ag(1B)
nearer to a layer of organic groups (and terminal halogen
atoms) from the next set of (anion)(cation)(anion) layers.30
Figure 3 shows the coordination environment of the
chlorine atom (crosshatched) in the chloromethanesulfonate
ion. It is three-coordinated by its carbon atom at 1.799 Å
and two silver ions at 2.945 Å. The bond angles C-Cl-Ag
and Ag-Cl-Ag are 105.6° and 79.6°, respectively. In our
(30) Shubnell, A. J.; Kosnic, E. J.; Squattrino, P. J. Inorg. Chim. Acta 1994,
216, 101-112. Jansen, M.; Korus, G. Z. Anorg. Allg. Chem. 1997,
623, 1625-1632.
2036 Inorganic Chemistry, Vol. 41, No. 8, 2002

Weakly Coordinating Anions
Br(12) in Ag(CB11H6Br6) and Br(12) in Ag(CH3CB11H5Br6),
are apical. Only 2 (or perhaps 3) apical organoiodine atoms
were found among the over 50 examples of coordinated
organoiodine atoms in Powell’s34 extensive studies of io-
docarbon ligands coordinated to Ag+.
In Table 1, we note the following average excess bond
distances for two-coordinate halogen atoms: (a) in diha-
lomethanes CH2X2 coordinated to silver ions, 0.32 Å for X
) Cl, 0.20 Å for X ) Br, and 0.00 Å for X ) I; (b) in
halogenated carborane anions (CB11H6X6)- coordinated to
silver ions, 0.27 Å for X ) Cl, 0.14 Å for X ) Br, and 0.05
Å for X ) I; and (c) in (CB11H6X6)- coordinated to (i-Pr)3Si+
ions, 0.15-0.16 Å regardless of the halogen involved. For
halocarbons and halocarboranes bonded to the soft acid Ag+,
the excess bond distance decreases as the halogen atom
becomes softer, while for halocarbons bonded to the hard
siliconium cation, there is no such trend.
Figure 2. Thermal ellipsoid drawing showing the coordination environment
of Ag(3).
We find (Table 5) that the Ag‚‚‚Cl secondary bond
distances in Ag(XCH2SO3) are 2.945 Å for X ) Cl (this
study) and 2.971 Å for X ) Br.16 These give excess bond
distances of 0.43 Å for X ) Cl and 0.31 Å for X ) Br.
These excess bond distances are quite comparable to those
previously found for the apical halogen atoms in [Ag(1,2-
C2H4Cl2)OTeF5]2 and Ag(CB11H6Br6), respectively. They are,
however, on the order of 0.1 Å longer than the excess bond
distances found for the same halogen atoms acting as
bridging ligands to silver. The same is true of the apical
iodine atoms of the first two iodotoluene and iodobenzene
adducts listed in Table 5, but the third one is anomalous: It
has two Ag-I contacts, one with zero excess bond distance
(as is found with bridging iodine atoms) and one with a very
large 0.46 Å excess bond distance. This suggests that possibly
this iodine atom is more nearly bridging than apical, so it
will not be included in this analysis of apical halogen atoms.
Some theoretical treatments of NQR parameters predict a
great sensitivity to bond angles at the halogen atom. Apical
halogen atoms, like bridging ones, generally show C-X-
Ag angles between 90° and 109° (Table 5). In addition, there
is a Ag‚‚‚X‚‚‚Ag angle, which in contrast shows wide
variability (between 61.4° and 144.5°); this angle is 79.6°
in Ag(3) and 79.1° in Ag(2).
Figure 3. Packing diagram of the structure of Ag(3) showing the
pseudooctahedral AgO4Cl2 coordination spheres. An Ag atom is shown at
the center of the diagram; Cl atoms are crosshatched (X). The Ag-Cl bonds,
at 2.9454(15) Å, are shown as dashed lines. The C-Ag-Cl and Ag-Cl-
Ag(1B) angles are 105.62 (19)° and 79.6 (2)°.
earlier NQR study of complexes of neutral halocarbons with
silver ions,31 the complexes featured two-coordinate halogen
atoms bridging between carbon and silver. In this study, we
have three-coordinate apical halogen atoms, each halogen
atom being bonded to carbon through the primary covalent
bond and to two different silver ions through two secondary
bonds.
Apically coordinated organohalogen atoms are rather
unusual; crystallographic data for these atoms are sum-
marized in Table 5. Only 1 of the 12 coordinated orga-
nochlorine atoms reported32 in the dichloromethane and
dichloroethane adducts of silver(I), Cl(3) in [Ag(1,2-C2H4-
Cl2)OTeF5]2, was an apical chlorine atom. Only 2 of the 31
crystallographically different coordinated halogen atoms
found in 7 silver(I) hexahalocarborane salts by Xie et al.,9,10,33
NQR Spectra: Theory.35-37 The two fundamental prop-
erties of a quadrupolar nucleus in a compound or ion are its
quadrupole coupling constant, e2Qqzz/h (in MHz), which
measures the deviation of the electronic environment of the
nucleus from spherical symmetry, and its asymmetry pa-
(33) Xie, Z.; Jel´ınek, T.; Bau, R.; Reed, C. J. Am. Chem. Soc. 1994, 116,
1907-1913.
(34) Powell, J.; Horvath, M.; Lough, A. J. Organomet. Chem. 1993, 456,
C27-C28. Powell, J.; Horvath, M.; Lough, A. J. Chem. Soc., Chem.
Commun. 1993, 733-5. Powell, J.; Horvath, M.; Lough, A. J. Chem.
Soc., Dalton Trans. 1996, 1669-1677. Powell, J.; Lough, A.; Saeed,
T. J. Chem. Soc., Dalton Trans. 1997, 14137-14138. Powell, J.;
Horvath, M.; Lough, A.; Phillips, A.; Brunet, J. J. Chem. Soc., Dalton
Trans. 1998, 637-645.
(31) Wulfsberg, G.; Robertson, J.; Babushkina, T. A.; Gushchin, S. I.; Terao,
H.; Powell, J. Z. Naturforsch., A: Phys. Sci. 2000, 145-150.
(32) Newbound, T. D.; Colsman, M. R.; Miller, M. M.; Wulfsberg, G. P.;
Anderson, O. P.; Strauss, S. H. J. Am. Chem. Soc. 1989, 111, 3762.
Colsman, M. R.; Newbound, T. D.; Marshall, L. J.; Noirot, M. D.;
Miller, M. M.; Wulfsberg, G. P.; Frye, J. S.; Anderson, O. P.; Strauss,
S. H. J. Am. Chem. Soc. 1990, 112, 2349.
(35) Buslaev, Yu. A.; Kravchenko, E. A.; Kolditz, L. Coord. Chem. ReV.
1987, 82, 1-237; (a) p 52-58.
(36) Lucken, E. A. C. Nuclear Quadrupole Coupling Constants; Academic
Press: London, 1969; (a) p 39; (b) Chapter 7; (c) p. 139.
(37) Semin, G. K.; Babushkina, T. A.; Yakobson, G. G. Nuclear Quad-
rupole Resonance In Chemistry; Wiley: New York, 1975; (a) pp 202-
204; (b) Chapter 5.
Inorganic Chemistry, Vol. 41, No. 8, 2002 2037

Wulfsberg et al.
Table 5. Average Crystallographic Results for Apically-Coordinated Halocarbons and Halo Anions
compound
ref
r(C-X)
r(Ag-X)
excess
C-X-Ag
Ag-X-Ag
Ag(ClCH2SO3)
[Ag(Cl2C2H4)OTeF5]2,
apical Cl only
Ag(BrCH2SO3)
Ag(CB11H6Br6)
Ag4(O2CCF3)4(H2O)2-
(p-IC6H4CH3)2
Ag2(O2CCCl3)2-
(HO2CCCl3)(IC6H5)
Ag2(O2CCF3)2(IC6H5)
1.799
1.75
2.945
2.915,
3.000
2.971
2.940
2.920,
2.933
2.927,
2.970
2.853,
3.309
0.43
0.40
0.49
0.31
0.28
0.07,
0.08
0.08,
0.12
0.00,
0.46
105.6
98.9,
106.3
103.6
79.6
119.2
a
b
c
c
c
1.938
2.119
2.093
2.082
79.1
144.53
61.4
d
102.2,
101.2
102.3,
94.4
99.0, d
a Ref 34. b Ref 17. c Ref 36. d Not reported.
rameter η (a dimensionless fraction that ranges from 0.00
to 1.00), which measures the deviation of the electronic
environment from axial symmetry. Neglecting the asymmetry
of the solid-state crystalline environment or the effects of
distant substituents, a singly bonded organohalogen atom has
axial symmetry but does not have spherical symmetry, so
has a nonzero e2Qqzz/h but an (approximately) zero η. On
forming a secondary bond at a bond angle other than 180°,
the halogen atom also loses its axial symmetry, so it also
has a nonzero η. In our earlier study of secondary-bonded
iodocarbons coordinated to one Ag+ ion, we found values
of η ranging from 0.20 to 0.42.31
The potential value of the fundamental NQR parameters
is that they are readily computable in principle. Rigorous
computations on the ionic compounds in this study would
in practice be quite formidable, because of the heavy atoms
involved and because the compounds are ionic, and the
quadrupole parameters are sensitive to the long-range ionic
forces present in the lattice. Fortunately, the NQR parameters
can be fairly directly related to the populations of the valence
orbitals of the halogen atom through simple Townes-Dailey
computations35,36b that often give reasonable estimates of
valence orbital populations, particularly in molecular species.
This theory assumes that core electron shells, which are filled
and have spherical symmetry, contribute nothing to e2Qqzz/h
or ν. It also assumes that external charges can be neglected
as well. In practice, this is a rough approximation, particularly
in ionic compounds, where the external ionic charges can
change the polarization of the halogen core electrons. Thus,
the NQR frequencies of group 1 salts of halo anions (in
which cations are very close to halogen atoms) can increase
by as much as 10% as the cation is increased in radius from
K+ through Cs+ to (CH3)4N+,35a while the NQR frequencies
of layered group 1 metal chloroacetates (in which cations
are close only to oxygen atoms)40 decrease by about 3% from
Na+ to Rb+.41 Hence, we have taken care in this study to
emphasize data for ions of comparable six-coordinate Shan-
non-Prewitt radii: Na+ (1.16 Å), Ag+ (1.29 Å); K+ (1.52
Å); Tl+ (1.64 Å). We do not know, however, whether these
are isomorphous and isostructural; if not, changes in structure
type could produce additional shifts of NQR frequencies.
Under these circumstances, the Townes-Dailey model only
gives us an estimate of what more sophisticated computations
might be able to tell us.
An NQR experiment measures not these fundamental
5
parameters, but the NQR frequencies ν. For a spin I ) /2
nucleus such as 127I, there are two NQR frequencies for each
crystallographically inequivalent atom. If it is known or can
be assumed that these two frequencies come from the same
iodine atom, then it is possible to determine e2Qqzz/h and η
based on the relationships38
ν1 ) 3/20(e2Qqzz/h)(1 + 5/54η2)
ν2 ) 6/20(e2Qqzz/h)(1-11/54η2)
(4)
(5)
3
For I ) /2 nuclei such as 35Cl, 79Br, and 81Br, there is
only one ν, which is related to the two fundamental
parameters e2Qqzz/h and η by the equation
ν ) 1/2(e2Qqzz/h)(1 + η2/3)1/2
(6)
If η is zero or even close to zero it has a negligible effect on
ν (an η of 0.17 raises the value of ν by only 0.5%). Thus, as
long as η does not exceed 0.17, it is often assumed that the
fundamental property, e2Qqzz/h, is 2 times ν. However, it is
possible to obtain η and therefore e2Qqzz/h by the Fourier
analysis of the slow beats in the spin-echo envelope of the
pulse NQR signal of polycrystalline samples;39 this was done
in external magnetic fields of 32 Πat 300 K for polycrys-
talline Ag(2) and 48 Πat 77 K for polycrystalline K(2).
Within this simple model, e2Qqzz/h is a function of the
imbalance in population of the valence orbitals of a halogen
atom:
e2Qqzz/h ) -e2Qqat/h(Nz - 1/2(Nx + Ny))
(7)
where e2Qqat/h is the atomic quadrupole coupling constant
(+109.746 MHz for 35Cl, -643.032 MHz for 81Br, -769.756
MHz for 79Br, and +2292.712 MHz for 127I)36c and Nz, Nx,
(38) Creel, R. B.; Brooker, H. R.; Barnes, R. G. J. Magn. Reson. 1980,
41, 146-149.
(39) Sapozhnikov, Yu. E.; Yasman, Ya. A. Bull. Acad. Sci. USSR, Phys.
Ser. (Engl. Transl.) 1978, 42, 120-123. IzV. Akad. Nauk SSSR, Ser.
Fiz. 1978, 42, 2148-2151. Current modifications described in:
Morgunov, V. G.; Kravchenko, E. A. Prib. Tekh. Eksp. 1988, 6, 199-
200. Kravchenko, E. A. Russ. Chem. ReV. 1999, 68, 787-805.
Additional modifications to be described elsewhere.
(40) Elizabe´, L.; Kariuki, B. M.; Harris, K. D. M.; Tremayne, M.; Epple,
M.; Thomas, J. M. J. Phys. Chem. B 1997, 101, 8827-8831.
Ehrenburg, H.; Hasse, B.; Schwartz, K.; Epple, M. Acta Crystallogr.,
Sect. B 1999, 55, 517-524.
(41) David, S.; Guibe´, L.; Gourdji, M. New J. Chem. 1995, 19, 37-46.
2038 Inorganic Chemistry, Vol. 41, No. 8, 2002

Weakly Coordinating Anions
Table 6. 127I NQR Frequencies and Parameters at 77 Ka
and Ny are the populations of the valence p orbitals indicated.
For a single-bonded halogen atom, Nx and Ny are 2.00 each,
so Nz (which is less than 2) is easily computed from
compound
ν1, rt
ν1
ν2
-e2Qqb
η
c
CH2I2
286.88
286.95
288.19
288.90d
300.32
301.81d
300.28
568.36
1897.37 0.086
K(ICH2SO3)
573.19(2) 1911.0 0.03
576.24(4) 1920.1 0.01
577.53(4) 1925.2 0.02
599.80(4) 1999.9 0.03
602.29(8) 2008.4 0.04
599.76(25) 1999.7 0.03
e2Qqzz/h ) -e2Qqat/h(Nz - 2)
(8)
Na(ICH2SO3)‚H2O
The Townes-Dailey expression for η shows that it arises
from an imbalance in the populations of the valence px and
py orbitals:
Tl(ICH2SO3)‚
1/n(18-crown-6)
Tl(ICH2SO3)
281.44(8) 287.51(20)
279.85(30) 286.52(40)
277.85(8) 284.81(20)
η ) (e2Qqxx - e2Qqyy)/e2Qqzz = 3/2(Nx - Ny)(e2Qqat/e2Qqzz)
Ag(ICH2SO3)
262.28(30) 264.85(42) 527.54
1759.7 0.05
(9)
a Signal-to-noise (S/N) ratios are given in parentheses. b By theory, the
quadrupole coupling constants should have negative values, although this
is not determined in the experiment. c Sources: Biryukov, I. P.; Voronkov,
M. G.; Safin, I. A. Tables of Nuclear Quadrupole Resonance Frequencies;
Israel Program for Scientific Translations: Jerusalem, 1969; p 105. Semin,
G. K.; Babushkina, T. A.; Yakobson, G. G. Nuclear Quadrupole Resonance
In Chemistry; Wiley: New York, 1975; p 423. d Signal of doubled intensity,
therefore likely due to two iodine nuclei.
For three-coordinate (apical) halogen atoms, the population
of all valence p orbitals of the halogen atom are different
from 2.00; potentially, there are more unknowns than
measurables.42 Fortunately, the two secondary bonds in most
of the compounds listed in Table 5, including the silver
chloro- and bromomethanesulfonates, are identical by sym-
metry. We will assume that they are in the iodomethane-
sulfonate as well, so we take Ny ) Nx.
If the secondary-bonding orbitals are pure unhybridized
valence p orbitals bonding at 90° angles to the main bonding
orbital and at 90° angles to each other, then by eq 9, because
Nx ) Ny, η ) 0. Equation 7 then becomes
crystal lattices, this makes relatively little difference in their
spectra: they are all similar to the corresponding halocarbon
in their electronic effects as measured by NQR. We could
then use either the K+ salt or the neutral dihalomethane as
the reference with very similar results.
NQR Spectra: Results and Discussion. We begin with
the results and analysis of the 127I NQR spectra (Table 6);
these are the most valuable because e2Qqzz/h and η can be
obtained directly from them. As expected, η for the free
ligand CH2I2 is close to zero (0.086). In the K+, Na+, Ag+,
and crown-complexed Tl+ salts of this anion, η is found to
range from 0.01 to 0.04, which suggests the absence of two-
coordinated (bridging) halogen atoms31 and either the absence
of any metal coordination to the iodine or, if it is present,
that it is apical with bond angles close to 90°. The two NQR
frequencies and the quadrupole coupling constant of each
iodine atom have similar values in CH2I2 and in all of the
metal iodomethanesulfonates except that of silver, suggesting
the absence of metal-iodine coordination. Applying eq 8,
we can estimate Nz, the population of the iodine 5pz orbital
involved in σ bonding to carbon, to be 1.16-1.17 electrons
in CH2I2 and the potassium (and likely the crown-free
thallium) salts and 1.13 electrons in the sodium and crown-
complexed thallium salts.
Turning to the data for the Ag+ salt, we observe that its
quadrupole coupling constant is substantially (7.3%) lower
than that of the reference CH2I2 and is similarly lower than
the coupling constants of the other metal salts. We also note
that η for the Ag+ salt is close to zero. These two findings
taken together suggest that the iodine atom in this salt is
apical and is coordinated to two Ag+ ions with bond angles
close to 90°. Hence, eq 10 ought to apply to this salt, giving
us the result that Nz - Ny ) -0.77 electrons. This does not
allow us to solve for Ny unless we make an additional
assumption, that Nz remains about 1.16 electrons after
complexation (which is supported in general by observations
that carbon-iodine bond lengths are unaltered upon com-
plexation).31 Then, we estimate that Nx ) Ny ) 1.92 electrons,
that is, that the iodine atom has donated 0.08 electrons to
each of the two coordinated silver ions, for a total of 0.16
e2Qqzz/h ) -e2Qqat/h(Nz - Ny)
(10)
Because Ny is between 2.00 and Nz, it can be seen by
comparing eqs 10 and 8 that e2Qqzz/h moves closer to zero
for apical halogen atoms, and (applying eqs 4-6) the halogen
NQR frequencies of apical halogens are reduced as compared
to those of the original terminal halogen atoms.
The crystallographic data show that the angles at the apical
halogen atoms in Ag(3) and Ag(2) are not equal to 90°.
Under these conditions, η becomes greater than zero, because
the valence px and py orbitals can no longer be equally
involved in the bonding. It should also be noted that
asymmetry parameters are also very sensitive to the polar-
izing effects of external charges (ions).
It is important, in assessing the effects of coordination of
a halocarbon or a halogenated anion to a metal ion such as
Ag+, to have a reference set of NQR parameters of that
species when it is not coordinated. Because K+ is a hard
acid likely to coordinate only to the sulfonate oxygen atoms,
and because K+ has the same charge as Ag+ and has about
as similar a size as is possible, we take it as the best reference
salt. The suitability of K+ salts as reference compounds will
be supported by the observation that, for these and several
other halogenated anions that we will describe elsewhere, it
is observed that solid halogenated compound X-CnHm-R
has about the same halogen NQR frequency when R ) X
-
as when R ) SO3 K+. Thus, although the various potassium
salts are undoubtedly not isostructural or isomorphous either
with the Ag+ salt or with each other, and thus have diverse
(42) Because of the lower symmetry of a pyramidal apical halogen atom,
the z direction of the quadrupole coupling tensor may not coincide
exactly with the direction of the C-X bond, in which case the right
side of eq 9 does not apply exactly.
Inorganic Chemistry, Vol. 41, No. 8, 2002 2039

Wulfsberg et al.
Table 7. 81Br and 79Br NQR Frequencies (MHz)a
Table 8. 35Cl NQR Frequencies (MHz)a
81Br,
298 K
81Br,
77 K
79Br, e2Qq,
compound
298 K
77 K
compound
77 K
77 K
η
b
CH2Cl2
35.991
Na(ClCH2SO3)‚1/3H2O
37.827(3)
36.707(3)
36.645(3)
34.545(4)
36.500(4)
36.261(2)
36.171(3)
35.915(17)
b
CH2Br2
236.18
282.72
K(BrCH2SO3) 230.36(13) 236.60(25)
472.07 0.12 ( 0.03c
Tl(BrCH2SO3) 229.64
233.82
279.86
Ag(BrCH2SO3) 218.21(30) 220.80(100) 264.36 438.08 0.22 ( 0.05d
Ag(ClCH2SO3)
K(ClCH2SO3)
34.151(3)
36.026(2)c
35.699(2)c
a Signal-to-noise (S/N) ratios are given in parentheses. b Source: Semin,
G. K.; Babushkina, T. A.; Yakobson, G. G. Nuclear Quadrupole Resonance
In Chemistry; Wiley: New York, 1975; for the R phase, p 402. The value
for 81Br was computed from the listed value for 79Br. c At 77 K. d At 300
K.
Tl(ClCH2SO3)
34.690(7)
a Signal-to-noise (S/N) ratios are given in parentheses. b Source: Semin,
G. K.; Babushkina, T. A.; Yakobson, G. G. Nuclear Quadrupole Resonance
In Chemistry; Wiley: New York, 1975; p 314. c Measured at 195 K.
electrons donated. This is virtually the same total as is
donated by the two-coordinated iodine atoms in silver
complexes of iodocarbons.31 But each of the two secondary
bonds in Ag(1) has only half the bond order that is found
with a two-coordinated iodine atom, which goes along with
the observation that secondary bond distances are longer with
three-coordinate iodine atoms.
Table 6 also lists some NQR frequencies recorded at room
temperature. NQR frequencies are almost always lower at
room temperature than at 77 K, but we have previously noted
that this is less so when intermolecular (or in this case,
interionic) coordination is involved.43 Thus, ν1 drops by about
6 MHz (2%) in the thallium salt (no coordination) but only
by about 2.5 MHz (1%) in the silver salt (interionic
coordination present).
Bromine NQR frequencies are shown in Table 7. The
frequencies of the two bromine isotopes were detected as
shown and found to be in the ratio of their atomic quadrupole
coupling constants, as expected. We found the NQR fre-
quencies of K(2) and Tl(2) to be very similar to that of the
parent bromocarbon CH2Br2, while the frequencies of
Ag(2) are substantially (6.5%) lower than that of the
reference CH2Br2 or of the other metal salts.
The measured η of K(2) is relatively small (0.12 ( 0.03)
but is not so small (0.22 ( 0.05) in Ag(2), which may be
connected with the bond angles C-Br-Ag and Ag-Br-
Ag being substantially different than 90°. Hence, eq 10
probably cannot be applied to this salt. The temperature
dependence of the 81Br NQR frequencies is, as expected,
less in the silver salt (2.6 MHz or 1.2%) than in the potassium
and thallium salts (6.3 MHz or 2.7% and 4.18 MHz or 1.8%,
respectively).
suggests that it has, in common with the other salts, an
unusual feature: a three-coordinate organohalogen donor
atom which is coordinated to two Ag+ ions by weak
secondary covalent bonds. The bond lengths and bond angles
indicate that these salts are good models for very weakly
coordinating carborane anions in Ag(CB11H6X6) and for
weakly coordinating halocarbons in salts of [Ag(CH2X2)n]+
(X ) Cl, Br, I). The NQR frequencies of the three silver
salts are lower than those of the other salts and of the parent
dihalomethanes; the degree of frequency lowering grows as
the halogen becomes softer (Cl, 4.0% < Br, 6.5% < I, 7.3%).
Even though thallium(I) is also a soft acid and has been found
capable of coordinating chlorocarbons, our evidence indicates
that it does not coordinate the halogen atoms in the thallium
halomethanesulfonates. For the first time, the fundamental
NQR parameters, the quadrupole coupling constant e2Qqzz/h
and the asymmetry parameter η, have been obtained for
halogenated weakly coordinated organic anions, in the silver
halomethanesulfonates. Asymmetry parameters so far tend
to be smaller for three-coordinated (apical) organohalogen
atoms (0.05 and 0.22) than two-coordinate (bridging) orga-
nohalogen atoms (0.20-0.42), but they are not necessarily
close to zero. We have used the simple Townes-Dailey
analysis of NQR frequencies to estimate that there is indeed
a significant degree of covalent bonding between the Ag and
organohalogen atoms in Ag(1): a total of about 0.16
electrons are donated by the iodine atom to its neighbor silver
ions. This suggests that although the coordination of such
an anion to silver ion may be weak, it is by no means
negligible.
Chlorine NQR frequencies recorded at 77 K and either
195 K or room temperature are shown in Table 8. We found
the NQR frequencies of K(3) and Tl(3) are close to that of
CH2Cl2, while the frequency of Ag(3) is substantially (4.0%)
lower than that of the reference CH2Cl2 or of the other metal
salts. Again, the temperature dependence for the silver salt
is much less (0.4 MHz or 1.2%) than that in the thallium
salt (1.2 MHz or 3.3%).
Conclusions. We have determined the crystal structure
of a silver salt of a weakly coordinating anion, Ag(3), and
found it to be isostructural and isomorphous with Ag(2).
Crystals of Ag(1) were twinned, so the crystal structure could
not be determined, but analysis of the NQR spectrum
Acknowledgment. G.P.W. would like gratefully to
acknowledge the financial support of the Middle Tennessee
State University Committee for Faculty Research and the
MTSU Foundation Special Projects Committee, especially
for their funding enabling international cooperation with the
Russian Academy of Sciences and the Russian Ministry of
Health, the technical assistance of Jack Ross, Andrienne
Friedli, and Celeste Mathews, and the assistance of Ron
Caple in facilitating cooperation with our Russian co-
workers. We also thank Robert West for the loan of the Wilks
NQR spectrometer.
Supporting Information Available: Four X-ray crystallo-
graphic files, in CIF format. This material is available free of charge
via the Internet at http://pubs.acs.org.
(43) Wulfsberg, G.; Brown, R. J. C.; Graves, J.; Essig, D.; Bonner, T.;
Lorber, M. Inorg. Chem. 1978, 17, 3426-3432.
IC010715I
2040 Inorganic Chemistry, Vol. 41, No. 8, 2002