Syntheses, crystal structures, and density functional theory calculations of the closo-[1-M(CO)3(η4-E9)]4- (E = Sn, Pb; M = Mo, W) cluster anions and solution NMR spectroscopic characterization of [1-M(CO)3(η4-Sn9)]4- (M = Cr, Mo, W)

Inorg. Chem. 2002, 41, 86−107

Syntheses, Crystal Structures, and Density Functional Theory

4

4-

Calculations of the closo-[1-M(CO) (η -E )] (E ) Sn, Pb; M ) Mo, W)

3

9

Cluster Anions and Solution NMR Spectroscopic Characterization of

4

4-

†

[

1-M(CO) (η -Sn )] (M ) Cr, Mo, W)

3

9

‡

,‡

‡

§

||

Janette Campbell, H e´ l e` ne P. A. Mercier,* Holger Franke, David P. Santry, David A. Dixon, and

Gary J. Schrobilgen*

,‡

Department of Chemistry, McMaster UniVersity, Hamilton, Ontario L8S 4M1, Canada,

Snowbird Software, Inc., 71A Leland Street, Hamilton, Ontario, L8S 3A1, Canada, and

William R. Wiley EnVironmental Molecular Sciences Laboratory, Pacific Northwest National

Laboratory, 906 Battelle BouleVard, P.O. Box 999, KI-83, Richland, Washington 99352

Received July 2, 2001

4

)]⁴^-(E ) Sn, Pb; M ) Mo, W) anions have been obtained by extracting the binary alloys KSn2.05

followed by reaction with M(CO) ‚mes (M

The closo-[1-M(CO)

3

(η -E

9

and KPb2.26 in ethylenediamine (en) in the presence of 2,2,2-crypt or in liquid NH

3

)

Mo, W) or Cr(CO)

3

·tol in en or liquid NH

3

solution. Crystallization of the molybdenum and tungsten salts was induced by

4

vapor diffusion of tetrahydrofuran into the en solutions. The salts [2,2,2-crypt-K]

4

[1-M(CO)

3

9

(η -Sn )]‚en (M ) Mo, W) crystallize

in the triclinic system, space group P 1h , Z ) 4, a ) 16.187(3) Å, b ) 25.832(4) Å, c ) 29.855(5) Å, r ) 111.46(1)°, â )

02.84(2)°, γ ) 92.87(2)° at −95 °C (M) Mo) and a ) 17.018(3) Å, b ) 27.057(5) Å, c ) 28.298(6) Å, R ) 66.42(3)°, â

1

)

4

3 2 4 3 9

76.72(3)°, γ ) 87.27(3)° at 20 °C (M) W). The salts (CO) M(en) [2,2,2-crypt-K] [1-M(CO) (η -Pb )]·2.5en (M ) Mo, W)

crystallize in the triclinic system, space group P 1h , Z ) 2, a ) 16.319(3) Å, b ) 17.078(3) Å, c ) 24.827(5) Å, R ) 71.82(3)°,

â ) 83.01(3)°, γ ) 81.73(3)° at −133 °C (M) Mo) and a ) 16.283(4) Å, b ) 17.094(3) Å, c ) 24.872(6) Å, R ) 71.62(2)°,

4

4-

â ) 82.91(2)°, γ ) 81.35(2)° at −153 °C (M) W). The [1-M(CO)

3

(η -Sn

9

)] anions were also characterized in liquid NH

⁴^-, NMR studies show that the

3

solution by ¹Sn, Sn, and Mo NMR spectroscopy. Unlike their fluxional precursor, nido-Sn

19

117

95

9

4

)]⁴^-anions are rigid on the NMR time scale. All possible inter- and intraenvironmental couplings, J(¹¹⁹^,¹¹⁷Sn−

[

1-M(CO)

3 9

(η -Sn

¹¹⁹^,¹¹⁷Sn), J(

¹¹⁹^,¹¹⁷Sn−¹⁸³W), and one J(¹¹⁹^,¹¹⁷Sn− Mo) coupling, have been observed and assigned. Complete spin−spin coupling

95

constant assignments were achieved by detailed analyses and simulations of all spin multiplets that comprise the ¹¹⁹Sn and

117

Sn NMR spectra and that arise from natural abundance tin isotopomer distributions and from natural abundance ¹⁸³W, in the

3 9 3 9

case of [1-W(CO) (η -Sn (η -Sn

)]⁴^-. Both the solid state and solution structures of the [1-M(CO) )]⁴^-anions are based on a

4

closo-bicapped square antiprismatic structure in which the transition metal occupies a cap position. The cluster structures are

consistent with Wade’s rules for 22 (2n + 2) skeletal electron systems. Electron structure calculations at the density functional

theory (DFT) level provide fully optimized geometries that are in agreement with the experimental structures. Complete assignment

of the NMR spectra was also aided by GIAO calculations. The calculated vibrational frequencies of the E ⁴

(

-

and [1-M(CO)

9

3

-

η⁴-E )]⁴^-anions are also reported and are used to assign the solid-state vibrational spectra of the [1-M(CO) )]⁴^-anions.

(η -E

4

9

3

9

2

05

Tl, and ²⁰⁷Pb NMR spectroscopy, studied the homonuclear

Introduction

_R_u_d_o_l_p_h_a_n_d_c_o_-_w_o_r_k_e_r_s_,₁-4 who first identified naked

cluster anions of the group 13-15 elements by solution Sn,

E₉⁴(E ) Sn, Pb) and the mixed series of isoelectronic

^aⁿⁱ^oⁿ^s^Sⁿ9-x^G^ex , Sn9-x^P^b^,^aⁿ^d^Sⁿ8-yPb Tl (x ) 0-9,

y ) 0-4). The E

-

4

-

4-

5-

119

x

y

4

-

9

anions (E ) Si, Ge, Sn, Pb) and their

†

Dedicated to Professor M. F. Hawthorne in appreciation for his

contributions to the quality of American Chemical Society journals during

(1) Rudolph, R. W.; Wilson, W. L.; Parker, F.; Taylor, R. C.; Young, D.

C. J. Am. Chem. Soc. 1978, 100, 4629.

(2) Rudolph, R. W.; Taylor, R. C.; Young, D. C. In Fundamental Research

in Homogeneous Catalysis; Tsutsui, M., Ed.; Plenum: New York,

1979; pp 997-1005.

(3) Rudolph, R. W.; Wilson, W. L.; Taylor, R. C. J. Am. Chem. Soc.

1981, 103, 2480.

his term as Editor of Inorganic Chemistry (1969-2000); presented at the

8

2

4th Canadian Society for Chemistry Conference, Montr e´ al, Qu e´ bec, May

6-30, 2001, Abstract No. 904-IN3.

*

To whom correspondence should be sent. E-mail: SCHROBIL@

MCMASTER.CA.

‡

McMaster University.

Snowbird Software, Inc.

Pacific Northwest National Laboratory.

§

(4) Wilson, W. L.; Rudolph, R. W.; Lohr, L. L.; Taylor, R. C.; Pyykko,

P. Inorg. Chem. 1986, 25, 1535.

||

86 Inorganic Chemistry, Vol. 41, No. 1, 2002

Published on Web 12/13/2001

4

(η -E

9

)]⁴^-Cluster Anions

closo-[1-M(Co)

3

mixed analogues possess (2n + 4) skeletal electrons and are

electron deficient. The structures are predicted to be, by the

polyhedral skeletal electron pair theory (PSEPT) or Wade’s

in solution. Local density functional theory (LDFT) calcula-

4

-

tions for Pb

9

show that the C4V structure is favored in the

gas phase. The low barrier to C4V-D3h interconversion is

5

-7

rules,

nido-clusters derived from the parent ten-vertex

also consistent with the low vibrational frequencies of the

3-

4-

13

deltahedron, a bicapped square antiprism. The X-ray crystal

Pb

9

and Pb

9

anions, intimating that the clusters are very

4

- 8

4- 9,10

4- 11,12

4- 13,14

structures of the Si

9

, Ge

9

,

Sn

9

,

and Pb

9

deformable, and with the observation that fluxional behavior

4

-

5-

anions have been obtained and shown to possess nido-

structures of ∼C4V symmetry, in accord with the PSEPT

predictions.

in Sn

9

and TlSn

8

3

is not slowed in liquid NH at

3

temperatures as low as -74 °C.

1

7,18

Rudolph and co-workers

the preparation of transition metal derivatives of the main-

3 4

group metal clusters obtained by the reaction of M′(PPh )

were also the first to report

The fluxionality of these nine-atom clusters has prevented

their detailed structural characterization in solution by NMR

1

19

207

205

4

-

5-

4-

spectroscopy. The observation of single Sn, Pb, and Tl

with Sn

9

and Sn

8

Tl (M′ ) Pt, Pd) and Pb

9

(M′ ) Pt)

resonances and single J( Sn- Sn), J(²⁰⁷Pb-¹¹⁷^,¹¹⁹Sn),

119

117

4

-

in en (ethylenediamine). The resulting Sn

9

M′(PPh

3

)

2

,

2

03,205

117,119

Sn), and J(²⁰³^,²⁰⁵Tl-²⁰⁷Pb) spin-spin cou-

5

-

4-

J(

Tl-

Sn

8

TlM′(PPh

3

)

2

, and Pb

9

PtL

2

(L ) PPh

3

and/or en)

31

119

195

plings indicates that the three chemical environments of the

anions were shown to be fluxional in solution by P, Sn,

Tl, and

resonances were reported. Rudolph and co-workers

2

05

207

C

4V structure are exchange-averaged by means of a facile

Pb NMR spectroscopy; however, no

Pt

were

1-4

17,18

intramolecular rearrangement process.

Sn

In the case of

intramolecular exchange averages

4- 1-4

4- 1,2,4,13

9

and Pb

9

,

unable to establish whether the clusters were nido- or closo-

structures by use of multi-NMR spectroscopy.

the unique apical E(1), four coplanar E(2,3,4,5) atoms

proximate to the apical atom, and the four coplanar E(6,7,8,9)

atoms comprising the open nido-face of the cluster to a single

chemical environment on the NMR time scale. Rapid

To date, the only closo-transition metal derivatives of

electron-deficient naked metal main-group clusters that have

been reported and structurally characterized by X-ray crystal-

4

-

rearrangement of the Pb

9

cluster prevents the observation

4

4-

19,20

21

lography are the [1-Cr(CO)

3

(η -E

6

9

)] (E ) Sn,

Pb ),

of J(²Pb- Pb) spin-spin couplings, because of chemical

and magnetic equivalence. This contrasts with Sn

07

207

5

4- 20

2- 22

[

(

2-W(CO)

3

(η -Sn

9

)] , Sn

[M(CO)

5

]

6

,

and Ge [M-

6

4

-

9

, which

2- 22

CO) ]

5 6

(M ) Cr, Mo, W) cluster anions. In accord with

possesses two natural abundance magnetically active iso-

the (2n +2) rule for closo-clusters having 22 and 14 skeletal

electrons, both the endo-skeletal Cr(CO) , W(CO) and exo-

skeletal M(CO) groups donate zero electrons to the skeletal

bonding of these anions. The [1-Cr(CO)

1

19

117

119

topes, Sn and Sn. A single exchange-averaged J( Sn-

3

117

4-

Sn) coupling constant is observed for Sn

9

, which results

5

in an overlapping series of satellites corresponding to the

summation of the binomial multiplet subspectra arising from

4

4-

3

(η -E

9

)] and

5

4-

[2-W(CO)

3

(η -Sn

9

)] anions are representative of the reac-

1

19

117

0

4-

the isotopomers Sn

x

Sn

y

Sn9-(x+y) (x + y ) 0-9),

tion products expected between nido-clusters and a zero-

0

where Sn denotes tin isotopes having zero spin. The

fluxional behavior of the E

electron donor. Two related closo-metalloborane isomers

4

-

9

cluster anions is attributed to

-

derived from B

9

H

12 and having 22 skeletal electrons have

rapid intramolecular interconversion between nido-mono-

capped square antiprismatic (C4V) and closo-tricapped trigonal

prismatic (D3h) structures by means of a C2V transition

5

4

- 23,24

also been reported, i.e., closo-[(η -C

5 5 9 9

H )-1-Ni(η -B H )]

5

- 23,24

and closo-[(η -C

5

H

5

)-2-Ni(η -B

9 9

H )] .

In view of the facile intramolecular exchange behavior

1

3,15,16

state.

The mechanism is presumed to be operative for

4-

exhibited by the 22-electron nido-Sn

9

and nido-Pb

9

the related heteroanions. The barrier to C4V-D3h intercon-

anions and by related isovalent anions, it was of interest to

determine what type, if any, intramolecular exchange pro-

cesses might arise when the zero-electron donor transition

version in nine-atom clusters has been calculated using a

2

-

9 9

points-on-a-sphere model for the pseudo-isoelectronic B H

15

4-

16

anion and for Sn

9

by SCF-MO-CNDO calculations and

metal fragments M(CO)

3

4-

(M ) Cr, Mo, W) are coordinated

anions.

indicate a preference for D3h symmetry in the gas phase. The

barrier is less than 0.1% of the total orbital energy of either

symmetry, suggesting that interconversion is likely to occur

to the nido-faces of E

9

At the point of submission of this paper, we learned of an

4

-

independent study on the [M(CO)

3

Sn

9

]

(M ) Cr, Mo, W)

(

5) Wade, K. AdV. Inorg. Radiochem. 1976, 18, 1.

anions by Eichhorn et al., who have obtained similar spectra,

6) Mingos, D. M. P. Acc. Chem. Res. 1984, 17, 311.

7) Mingos, D. M. P. J. Chem. Soc., Chem. Commun. 1983, 706.

8) von Schnering, H. G.; Somer, M.; Kaupp, M.; Carrillo-Cabrera, W.;

Baitinger, M.; Schmeding, A.; Grin, Y. Angew. Chem., Int. Ed. Engl.

(17) Teixidor, F.; Luetkens, M. L., Jr.; Rudolph, R. W. J. Am. Chem. Soc.

1983, 105, 149.

1

998, 37, 2359.

9) Diehl, L.; Khodadadeh, K.; Kummer, D.; Str a¨ hle, J. Chem. Ber. 1976,

09, 3404.

10) Belin, C. H. E.; Corbett, J. D.; Cisar, A. J. Am. Chem. Soc. 1977, 99,

163.

(18) Luetkens, M. L., Jr.; Teixidor, F.; Rudolph, R. W. Inorg. Chim. Acta

1984, 83, L13.

(19) Eichhorn, B. W.; Haushalter, R. C.; Pennington, W. T. J. Am. Chem.

Soc. 1988, 110, 8704.

(

1

(

7

(20) Kesanli, B.; Fettinger, J.; Eichhorn, B. W. Chem.-Eur. J. 2001, in

(

11) Corbett, J. D.; Edwards, P. A. J. Am. Chem. Soc. 1977, 99, 3313.

12) Burns, R. C.; Corbett, J. D. Inorg. Chem. 1985, 24, 1489.

13) Campbell, J.; Dixon, D. A.; Mercier, H. P. A.; Schrobilgen, G. J. Inorg.

Chem. 1995, 34, 5798.

14) Korber, N.; Fleischmann, A. J. Chem. Soc., Dalton Trans. 2001, 383.

15) Guggenberger, L. J.; Muetterties E. L. J. Am. Chem. Soc. 1976, 98,

press.

(21) Eichhorn, B. W.; Haushalter, R. C. J. Am. Chem. Soc., Chem. Commun.

1990, 937.

(22) Renner, G.; Kircher, P.; Huttner, G.; Rutsch, P.; Heinze, K. Eur. J.

Inorg. Chem. 2001, 973.

(23) Leyden, R. N.; Hawthorne, M. F. J. Chem. Soc., Chem. Commun.

1975, 310.

(

7221.

(

16) Burns, R. C.; Gillespie, R. J.; Barnes, J. A.; McGlinchey, M. J. Inorg.

(24) Leyden, R. N.; Sullivan, B. P.; Baker, R. T.; Hawthorne, M. F. J.

Am. Chem. Soc. 1978, 100, 3758.

Chem. 1982, 21, 799.

Inorganic Chemistry, Vol. 41, No. 1, 2002 87

Campbell et al.

4

Table 1. Summary of Crystal Data and Refinement Results for [2,2,2-crypt-K]4[1-M(CO)3(η -Sn9)]·en and

4

M(CO)3(en)2[2,2,2-crypt-K]4[1-M(CO)3(η -Pb9)]·2.5en (M ) Mo, W)

4

[

2,2,2-crypt-K]4[1-M(CO)3(η -Sn9)]‚en

M(CO)3(en)2[2,2,2-crypt-K]4[1-M(CO)3(η -Pb9)]‚2.5en

formula

C77H152K4MoN10O27Sn9

C77H152K4N10O27Sn9W

C87H180K4Mo2N17O30Pb9

C87H180K4N17O30Pb9W2

fw

2970.64

P 1h (2)

3046.54

P 1h (2)

4157.47

P 1h (2)

4333.29

P 1h (2)

space group (no.)

a (Å)

b (Å)

16.187(3)

25.832(4)

29.855(5)

111.46(1)

102.84(2)

92.87(2)

11208.3(33)

4

17.018(3)

27.057(5)

28.298(6)

66.42(3)

76.72(3)

87.27(3)

11608.13(44)

4

16.319(3)

17.078(3)

24.827(5)

71.82(3)

83.01(3)

81.73(3)

6483.7(21)

2

16.283(4)

17.094(3)

24.872(6)

71.62(2)

82.91(2)

81.35(2)

6473.9(24)

2

c (Å)

R (deg)

â (deg)

γ (deg)

V (Å )

Z

3

T (°C)

-95

20

0.71073

1.743

3.095

0.1215

0.3794

-133

-153

λ (Å)

0.71073

1.760

0.71073

2.130

0.71073

2.223

Fcalcd (g cm^-³)

-

1

µ (mm

)

2.293

0.0982

0.2759

12.022

0.0700

0.1781

13.621

0.0544

0.1227

a

R1

b

wR2

a

R1 ) ∑|Fo| - |Fc|/∑|Fo| for I > 2σ(I). wR2 ) [∑[w(Fo²- Fc ) ]/∑w(Fo ) ] for I > 2σ(I).

b

2 2

2 2 1/2

but of lower resolution, and have interpreted the spectra in

4

5

4-

terms of a dynamic equilibrium between the η - and η

2

0

isomers.

The present paper²⁵reports the crystal structures of the

4

4-

heteroatomic closo-[1-M(CO)

3

(η -E

9

)] anions (E ) Sn, Pb;

M ) Mo, W) and the comprehensive solution characteriza-

clusters (Figure 1). All four anions exhibit the bicapped

square antiprismatic geometries predicted by PSEPT for (2n

4

4-

tions of the closo-[1-M(CO)

3

(η -Sn

9

)] (M ) Cr, Mo, W)

anions by ¹Sn, Sn, and Mo NMR spectroscopy. A

19

117

95

+ 2) skeletal electron systems and contain M(CO)

which occupy the otherwise open square faces of the parent

3

fragments

4

4-

preliminary study of the closo-[1-Mo(CO)

3

9

(η -Pb )] anion

207

4-

4

4-

by Pb NMR spectroscopy is also described. The solution

NMR studies provide a means to study the effect on the

cluster dynamics upon coordination of a zero-electron donor

nido-E

9

anions. The closo-[1-M(CO)

3

9

(η -E )] anions are

4

4-

isostructural with the known [1-Cr(CO)

3

5

(η -E

9

)] ana-

)-1-Ni(η -

1

9-21

4

logues

as well as with the closo-[(η -C

H

5 5

4-

)]^-metalloborane

23,24

group to an otherwise fluxional nido-E

9

cluster. In contrast

B

9

H

9

and the heteroatom-substituted

2

- 26

- 27

28

29

with the study of Eichhorn et al., the present solution study

XB

9

H

9

(X ) BH , CH , NH, S ) clusters. The

4

4-

4

4-

definitively establishes that the closo-[1-M(CO)

3

(η -Sn

9

)]

[1-M(CO)

group symmetry, whereas the [1-M(CO)

possess C point group symmetry. The symmetry differences

arise from the orientation of the M(CO) group relative to

9

cage. The E-E contacts (Table 2) within the E -

3

(η -Pb

9

)] anions possess essentially C

s

point

4

4-

(M ) Cr, Mo, W) anions are rigid on the NMR time scale.

3

(η -Sn

9

)] anions

Density functional theory calculations have been done at the

LDFT level to support and aid in the assignments of the

experimental NMR and vibrational spectra and to provide a

better understanding of the nature of the bonding between

naked main-group metal clusters and transition metals.

1

3

0

the E

framework of [1-M(CO)

9

4

4-

3

9

(η -E )] can be classified into

three categories: the shortest E-E contacts are those from

the apical E(1) atoms to the E(2,3,4,5) planes and those

within the E(6,7,8,9) planes; longer contacts are observed

Results and Discussion

X-ray Crystal Structures of the closo-[1-M(CO)

3

(η⁴-

)] Anions (E ) Sn, Pb; M ) Mo, Cr). Details of

4

between the two E -planes of the square antiprisms; and the

longest contacts exist between the adjacent E atoms of the

4

-

E

9

synthetic and crystal growing procedures are given in the

Experimental Section and details of data collection param-

eters and other crystallographic information are given in

Table 1. The most significant E-E and E-M contacts, E‚‚‚E

long contact distances, and M-C and C-O bond lengths in

E(2,3,4,5) planes. A comparison of the E-E contacts

4

-

between the nido-E

9

precursors and the closo-[1-M(CO)

3

-

4

4-

(η -E

9

)] anions reveals that coordination of the M(CO)

3

fragments to the E(6,7,8,9) planes results in elongation of

the E-E distances within these planes. The E-E distances

4

4-

the [1-M(CO)

3

9

(η -E )] anions are given in Table 2. Bond

(26) Hawthorne, M. F.; Pilling, R. L.; Stokely, P. F. J. Am. Chem. Soc.

angles are included in the Supporting Information. The

bicapped square antiprismatic geometries observed for the

1965, 87, 1893.

(

27) Knoth, W. H. Inorg. Chem. 1971, 10, 598.

28) Arafat, A.; Baer, J.; Huffman, J. C.; Todd, L. J. Inorg. Chem. 1986,

4

4-

[1-M(CO)

3

9

(η -E )] clusters are shown in Figure 1 along

25, 3797.

(

29) Pretzer, W. R.; Rudolph, R. W. J. Am. Chem. Soc. 1973, 95, 931.

30) For structural conformers possessing Cs symmetry, one CO group

eclipses an E atom of the E(2,3,4,5) plane while it is staggered with

respect to the E(6,7,8,9) plane. All three CO groups of the C1

conformers are gauche with respect to the E atoms of the E(2,3,4,5)

and E(6,7,8,9) rings, so no CO groups lie in a mirror plane of the

ME9 bicapped square antiprism.

with the atom numbering scheme used throughout this paper.

+

The structures of the (2,2,2-crypt-K ‚en) cations and the

(

25) The X-ray crystal sructures, experimental NMR spectra with their full

assignments, and infrared data were originally reported in the Ph.D.

Thesis of J. Campbell (McMaster University, 1997).

88 Inorganic Chemistry, Vol. 41, No. 1, 2002

4

(η -E

9

)]⁴^-Cluster Anions

closo-[1-M(Co)

3

4

4-

Table 2. The E-E, E-M, M-C, and C-O Distances (Å) in the [1-M(CO)3(η -E9)] (E ) Sn, Pb; M ) Mo, W) Anions

Sn9Mo(CO)3⁴

- a

Sn9W(CO)3⁴^-^a

Pb9Mo(CO)3⁴^-

Pb9W(CO)3⁴^-

E-E Contacts from the Apical E(1) Atom to the E(2,3,4,5) Plane

E(1)-E(2)

E(1)-E(3)

E(1)-E(4)

E(1)-E(5)

2.959(2), 2.967(3)

2.955(5), 2.954(4)

2.946(3), 2.966(3)

2.978(3), 2.977(3)

2.935(3), 2.937(3)

3.0661(10)

3.0330(11)

3.0557(10)

3.0737(9)

3.0417(9)

2.919(2), 2.950(2)

2.958(2), 2.961(3)

2.954(2), 2.947(2)

3.0533(10)

3.0326(14)

3.0407(9)

E-E Contacts and E‚‚‚E Long Contact Distances within the E(2,3,4,5) Plane

E(2)-E(3)

E(2)-E(5)

E(3)-E(4)

E(4)-E(5)

E(2)‚‚‚E(4)

E(3)‚‚‚E(5)

3.187(3), 3.216(3)

3.170(2), 3.155(3)

3.283(2), 3.146(3)

3.137(2), 3.189(2)

4.527(3), 4.502(3)

4.507(2), 4.482(2)

3.209(4), 3.201(3)

3.167(4), 3.161(3)

3.192(3), 3.189(3)

3.225(3), 3.213(3)

4.550(4), 4.529(3)

4.495(3), 4.495(3)

3.3094(14)

3.3165(12)

3.3200(12)

3.3289(14)

4.696(2)

3.3076(10)

3.3282(11)

3.3050(11)

3.3086(10)

4.691(1)

4.690(2)

4.678(1)

E-E Contacts between the E(2,3,4,5) and E(6,7,8,9) Planes

E(2)-E(6)

E(2)-E(9)

E(3)-E(6)

E(3)-E(7)

E(4)-E(7)

E(4)-E(8)

E(5)-E(9)

E(5)-E(8)

2.976(2), 2.946(2)

2.984(3), 3.000(3)

2.979(3), 2.985(3)

2.968(4), 2.970(3)

2.977(3), 2.978(3)

2.990(3), 3.001(3)

2.963(3), 2.972(3)

2.981(4), 2.962(3)

2.969(3), 2.977(3)

3.1399(9)

3.1184(9)

3.1136(8)

3.0854(10)

3.1703(8)

3.1471(10)

3.1289(9)

3.0813(10)

3.1364(8)

2.978(2), 2.926(2)

2.950(2), 2.939(2)

2.961(2), 2.959(2)

2.946(2), 2.984(2)

2.939(2), 2.968(2)

2.998(2), 2.984(2)

2.995(2), 2.980(2)

3.1248(11)

3.1527(10)

3.0844(13)

3.1082(9)

3.1059(9)

3.1200(9)

3.0765(11)

E-E Contacts and E‚‚‚E Long Contact Distances within the E(6,7,8,9) Plane

E(6)-E(9)

E(6)-E(7)

E(7)-E(8)

E(8)-E(9)

E(7)‚‚‚E(9)

E(6)‚‚‚E(8)

3.064(2), 3.082(2)

3.096(2), 3.102(2)

3.053(2), 3.082(2)

3.029(2), 3.073(2)

4.329(2), 4.356(2)

4.327(2), 4.368(2)

3.095(3), 3.094(3)

3.049(3), 3.035(3)

3.094(3), 3.099(3)

3.115(3), 3.080(3)

4.412(3), 4.374(3)

4.323(3), 4.329(3)

3.1489(11)

3.2026(12)

3.2911(11)

3.1842(12)

4.556(2)

3.2962(10)

3.2171(10)

3.1466(9)

3.1948(10)

4.517(1)

4.513(1)

4.571(1)

E-M Contacts

E(6)-M(10)

E(7)-M(10)

E(8)-M(10)

E(9)-M(10)

2.976(2), 2.993(2)

2.984(2), 2.999(2)

3.004(2), 3.004(2)

3.028(2), 2.946(2)

2.947(2), 2.955(2)

3.029(2), 3.020(2)

2.997(2), 2.999(2)

2.952(2), 2.951(2)

3.052(2)

3.0081(8)

3.0347(10)

3.0768(10)

2.9902(9)

3.0055(12)

2.9852(11)

3.0840(13)

M-C Bond Lengths

1.88(3), 1.98(3)

1.87(2), 1.94(2)

1.95(2), 1.87(2)

M(10)-C(11)

M(10)-C(12)

M(10)-C(13)

1.89(2), 1.77(2)

1.87(2), 1.86(3)

1.88(2), 1.84(2)

1.950(13)

1.926(10)

1.926(12)

1.939(9)

1.928(13)

1.935(12)

C-O Bond Lengths

1.22(3), 1.09(3)

1.26(3), 1.17(3)

1.12(3), 1.24(3)

C(11)-O(11)

C(12)-O(12)

C(13)-O(13)

1.16(2), 1.27(2)

1.21(2), 1.20(3)

1.21(2), 1.27(3)

1.19(2)

1.175(13)

1.189(14)

1.189(11)

1.186(14)

1.171(14)

a

Values are for the two noncrystallographically equivalent anions defined in the asymmetric unit.

2

- 33

within the E(6,7,8,9) planes increase in the order Cr < Mo

W, with these increases being larger for the lead series

OH)(µ

3

-O)

3

(OEt)

3

{(CO)

The Pb-Mo distances in [1-Mo(CO)

are comparable to the Pb-Mo bond length in (C

5 7 7 5

W} Sn ] , and (CO) WSn[W-

3

4

4-

<

(CO)

5

]

2

.

3

(η -Pb

)(CO)

but they are longer than the Pb-Mo bonds in

9

)]

than for the tin series. The distances within the E(2,3,4,5)

plane show the opposite, less pronounced trend over the

chromium triad, but are larger than those of the E(6,7,8,9)

plane coordinated to the metal. The E-E contacts from the

capping E(1) atoms to the adjacent E(2,3,4,5) plane and

H

5 5

3

-

3

5

MoPbPh

[(C HMo]

lengths in the [1-M(CO)

3

,

3

6

H

5 5

)

2

Pb(O

2

CCH

3

)

2

4

.

The M-C and C-O bond

4-

3

(η -E

9

)] anions (Table 2) are in

close agreement with other values reported in the litera-

4

4-

19-22,32-34

between the two square planes in the [1-M(CO)

3

(η -E

9

)]

ture

and are equal, within experimental error, to

(en) reported in this study.

The structure of M(CO) (en) and the coordination of en

4-

anions are similar to those in the E

9

anions.

those of M(CO)

3

2

The Sn-Mo [2.977(2)-3.028(2) Å] and Sn-W [2.949(2)-

3

2

3.030(2) Å] distances (Table 2) are significantly longer than

solvent molecules are discussed in the Supporting Informa-

the sums of their respective single-bond metallic radii [2.695

tion.

3

1

4

)]⁴^-(M ) Cr,

and 2.703 Å]. Similarly, the Pb-Mo [2.9852(11)-

.0839(13) Å] and Pb-W [2.9902(2)-3.0768(10) Å] dis-

tances are also longer than the sums of their respective single-

Characterization of the [1-M(CO)

Mo, W) Anions by NMR Spectroscopy. The [1-M(CO) -

3

(η -Sn

9

3

1

(32) Veith, M.; Kunze, K. Angew. Chem., Int. Ed. Engl. 1991, 30, 95.

bond metallic radii [2.834 and 2.842 Å, respectively]. The

(

33) Kircher, P.; Huttner, G.; Zsolnai, L.; Driess, A. Angew. Chem., Int.

4

4-

Sn-M distances in the [1-M(CO)

anions are longer than those in Sn

3

(η -Sn

6

9

)] (M ) Mo, W)

Ed. Engl. 1998, 37, 1666.

34) Huttner, G.; Weber, U.; Sigwarth, B.; Scheidsteger, O.; Lang, H.;

Zsolnai, L. J. Organomet. Chem. 1985, 282, 331.

35) Struchkov, Y. T.; Anisimov, K. N.; Osipova, O. P.; Kolobova, N. E.;

Nesmeyanov, A. N. Dokl. Akad. Nauk SSSR 1967, 172, 107.

2

- 22

t

[M(CO)

Mo-Sn(O Bu)

]

5 6

, Tl(O Bu)

3

2

t

32

Sn-Mo(CO)

5

,

(CO)

5

3

In-Mo(CO)

5

,

3

[(µ -

(

31) Pauling, L. In Nature of the Chemical Bond, 3rd ed.; Cornell University

(36) Kubicki, M. M.; Kergoat, R,; Guerchais, J.-E. J. Chem. Soc., Dalton

Trans. 1984, 1791.

Press: Ithaca, NY, pp 116, 256, 261.

Inorganic Chemistry, Vol. 41, No. 1, 2002 89

Campbell et al.

spacings corresponding to J(¹¹⁹Sn-¹¹⁹Sn), J( Sn-¹¹⁷Sn),

119

1

17

117

117,119

Sn- Mo), and J(¹¹⁷^,¹¹⁹Sn-¹⁸³W), and

95

J( Sn- Sn), J(

the satellite-to-central peak intensity ratios are consistent with

bicapped square antiprismatic MSn structures having ap-

parent C4V point symmetries (unhindered M(CO) rotation)

on the NMR time scale and with the solid-state geometries

9

3

4

4-

of the [1-M(CO)

3

9

(η -Sn )] anions.

The NMR study not only shows that the structures of these

anions are retained in solution, but in contrast with the nido-

4

-

Sn

9

precursor, the 10-vertex closo-MSn cages are stereo-

chemically rigid on the NMR time scale. Assignments of

1

17

119

183

the three tin resonances and the major Sn, Sn, and

W

4

4-

satellite spectra of the [1-M(CO)

3

9

(η -Sn )] anions that led

to the deduction of their solution structures are discussed

below.

(a) Chemical Shifts. The most deshielded resonance is

readily assigned to the apical Sn(1) based on its relative

intensity. The assignments of the resonances corresponding

to the two planes, Sn(2,3,4,5) and Sn(6,7,8,9), are based on

their relative atomic charges (see below and Computational

Results) and are confirmed by the relative magnitudes of

1

19

117

119

117

the Sn- Sn, Sn- Sn, and Sn- Sn couplings

between these planes and the apical Sn(1) atom (see Spin-

4

4-

Spin Coupling in [1-M(CO)

3

9

(η -Sn )] ). As a consequence,

the most shielded of the three resonances is assigned to the

Sn(2,3,4,5) plane adjacent to the apical Sn(1) atom and that

having intermediate shielding is assigned to the Sn(6,7,8,9)

plane for each anion of the series.

4

4-

Figure 1. Structures of the (a) [1-Mo(CO)3(η -Pb9)] , (b) [1-W(CO)3-

4

4-

4

4-

4

4-

(

η -Pb9)] , (c) [1-Mo(CO)3(η -Sn9)] , and (d) [1-W(CO)3(η -Sn9)]

anions. Thermal ellipsoids are shown at the 50% probability level.

The ¹¹⁹Sn NMR resonance of the apical Sn(1) atom in

_η4_-_S_n₎_]4- anions were structurally characterized in liquid

9

solutions by observation of the spin- /

Sn, and Mo at their natural abundance levels. All Sn

(

4

4-

1

119

[1-M(CO)

3

9

(η -Sn )] is very significantly deshielded [2493.2

NH

3

2

nuclides Sn,

1

17

95

119

(Cr), 2125.1 (Mo), and 2448.2 (W) ppm] relative to the Sn-

(2,3,4,5) [-521.9 (Cr), -682.1 (Mo), and -735.7 (W) ppm]

and Sn(6,7,8,9) environments [-213.9 (Cr), -402.4 (Mo),

and -496.9 (W) ppm]. A pronounced deshielding of the

apical atoms in bicapped antiprismatic clusters has also been

117

and Sn NMR spectra were recorded at 0 °C, while the

1₁₉_S_n_,117

95

4

4-

Sn, and Mo spectra of [1-Mo(CO)

3

9

(η -Sn )] were

also recorded at 30 °C in order to minimize the effects of

solvent viscosity on the rate of quadrupolar relaxation and

line broadening.

2

6-29,38-40

observed in heteroatom-substituted boranes.

For

11

5

119

117

95

example, in the B NMR spectrum of the nonfluxional [(η -

The experimental Sn, Sn, and Mo NMR spectra for

4

-

23,24

4

4-

C H

5 5

)-1-Ni(η -B

ing to the B(1) atom appears at 73.6 ppm, whereas those

corresponding to the two B -planes appear at 29.0 and -1.3

9

H

9

)] cluster,

the resonance correspond-

the [1-M(CO)

in liquid NH

tin environments in a 4:4:1 intensity ratio along with

3

(η -Sn

9

)] (M ) Cr, Mo, W) anions at 0 °C

3

are depicted in Figures 2-6 and display three

4

_a_c_c_o_m_p_a_n_y_i_n_g1 Sn, Sn, and W satellites. The obser-

vation of three tin resonances indicates that the M(CO)

groups are unhindered and rotate sufficiently rapidly about

the Sn(1)-M pseudo 4-fold axis of the MSn cage, so the

M(CO) tripods do not lower the 4-fold symmetries of MSn

cages on the NMR time scale. The possibility that the Sn-

2,3,4,5) and Sn(6,7,8,9) rings are rotating relative to each

19

117

183

37

ppm; however, the resonances could not be unambiguously

assigned to the B(2,3,4,5) and B(6,7,8,9) planes. The shift

to higher frequency for the apical B atom located opposite

the apical heteroatom in a closo-cluster has been termed the

3

9

3

8-40

“

antipodal effect”.

Addition of the M(CO)

in appreciable changes in the Sn chemical shifts of the

3

9

4

-

3

group to the Sn

9

cluster results

1

19

(

4

4-

[1-M(CO)

3

(η -Sn

9

)] anions relative to that of Sn

behavior contrasts with the relative insensitivity of the Sn

chemical shifts for the Sn

9

.

This

other has been considered. This intramolecular exchange

process is definitively shown in the subsequent discussion

to be inconsistent with the observed number of interplanar

coupling paths. The chemical shifts and spin-spin coupling

constants extracted from the satellite spectra are summarized

in Tables 3 and 4. The number of environments, the satellite

1

19

4-

5-

M′(PPh )

9 3 2

8 3 2

and Sn TlM′(PPh )

anions upon replacement of a Pt(PPh )

3 2

group with a Pd-

17,18

119

(PPh )

3 2

group.

The weighted averages of the Sn NMR

4

4-

resonances of the [1-M(CO)

3

9

(η -Sn )] anions all occur to

4

4-

(

37) In addition to the three intense resonances of the [1-W(CO)3(η -Sn9)]

(38) He rˇ m a´ nek, S. Chem. ReV. 1992, 92, 325.

(39) He rˇ m a´ nek, S.; Hnyk, D.; Havlas, Z. J. Chem. Soc., Chem. Commun.

1989, 1859.

(40) B u¨ hl, M.; Schleyer, P. v. R.; Havlas, Z.; Hnyk, D.; He rˇ m a´ nek, S. Inorg.

Chem. 1991, 30, 3107.

4-

119

117

anion and the nido-Sn9 anion (-1291.9 ppm, J( Sn- Sn) ) 280

Hz)), two weak singlets were observed in the Sn9 /W(CO)3‚mes

system, but they could not be assigned due to their low intensity [δ-

4

-

1

19

119

(

Sn) ) -639 ppm; δ( Sn) ) -712 ppm].

90 Inorganic Chemistry, Vol. 41, No. 1, 2002

4

(η -E

9

)]⁴^-Cluster Anions

closo-[1-M(Co)

3

Figure 2. The ¹¹⁹Sn (111.922 MHz) NMR spectra of the Sn(1), Sn(2,3,4,5), and Sn(6,7,8,9) environments in the [1-W(CO)3(η -Sn9)]⁴^-anion. Simulated

4

spectra are the right-hand traces.

significantly higher frequency of Sn ⁴^-

solvent, -10 °C) and shift progressively to lower frequency

(-1287 ppm, NH

The redistribution of charge is also reflected in the ⁹⁵Mo

9

3

chemical shifts of (CO) Mo‚mes (-1918.4 ppm) and [1-Mo-

3

4

-

4

4-

95

relative to Sn

9

in the series Cr (-50 ppm), Mo (-246

(CO) (η -Sn )] (-2002.6 ppm), where Mo of the anion

3

9

ppm), and W (-276 ppm) (Table 3). The global deshielding

is shielded by 84.2 ppm relative to M(CO) ·mes (present as

3

4

-

accompanying coordination of the open vertex of the Sn

cluster to a zero-electron donor transition metal center is the

result of electron transfer from the electron rich Sn cage to

the π-acid transition metal center. The aforementioned Sn

9

excess reagent in the solution) and is more shielded than

95

41

the Mo resonance of Mo(CO) (-1865 ppm). Interest-

6

9

ingly, the chemical shift of the apical Sn(1) atom in the Mo

derivative is not intermediate with respect to those of the Cr

117

1

19

and Sn chemical shift trends correlate with the calculated

charge delocalizations from the E cages and the charge

buildups on the transition metals (see Computational Results).

9

(

41) Vosegaard, T.; Skibsted, J.; Jakobsen, H. J. J. Phys. Chem. A 1999,

103, 9144.

Inorganic Chemistry, Vol. 41, No. 1, 2002 91

Campbell et al.

Figure 3. The ¹¹⁷Sn (106.945 MHz) NMR spectra of the Sn(1), Sn(2,3,4,5), and Sn(6,7,8,9) environments in the [1-W(CO)3(η -Sn9)]⁴^-anion. Simulated

4

spectra are the right-hand traces.

and W analogues, as was observed for the other two tin

resonances, but is significantly more shielded.

¹¹⁹Sn satellite patterns in addition to intense central lines.

The latter arise from the most abundant isotopomer families

b) Spin-Spin Couplings. The natural abundance ¹¹⁷Sn

119

Sn Sn

tense satellite patterns are doublet subspectra arising from

0

M(CO)

4-

and Sn Sn

117

0

4-

. The most in-

(

8

3

8

M(CO)

3

and ¹Sn NMR spectra of the three tin environments of the

19

4

4-

117

119

117

[

1-M(CO)

3

(η -Sn

9

)] anions each display complex Sn and

the J( Sn- Sn), J( Sn- Sn), and J( Sn- Sn) cou-

92 Inorganic Chemistry, Vol. 41, No. 1, 2002

4

(η -E

9

)]⁴^-Cluster Anions

closo-[1-M(Co)

3

Figure 4. The ¹¹⁹Sn (111.922 MHz) NMR spectra of the Sn(1), Sn(2,3,4,5), and Sn(6,7,8,9) environments in the [1-Cr(CO)3(η -Sn9)]⁴^-anion. Simulated

4

spectra are the right-hand traces.

plings of the most abundant isotopomer families, ¹¹⁹Sn

⁰Sn

-

.

in the ¹¹⁹Sn and ¹¹⁷Sn NMR spectra of [1-W(CO)

(η -Sn

4

)] ,

4-

2

7

4-

3

183

9

4-

119

117

0

4-

117

0

119

0

4-

M(CO)

3

,

Sn Sn Sn

7

M(CO)

3

, and Sn

2

7

Sn M(CO)

3

which arise from the isotopomer families Sn Sn

8

W(CO)

3

1

17

0

183

4-

The next most significant contributors to the satellite spectra

and Sn Sn

8

W(CO)

3

.

1

19

0

4- 119

117

0

are the isotopomer families Sn

3

Sn

6

M(CO)

3

,

Sn

2

Sn -

Two categories of tin-tin spin-spin couplings are dis-

tinguishable among the various absolute isotopomers that

comprise an isotopomer family, namely, those that arise from

4

plane and those

that arise from interenvironmental coupling between the Sn

4

-

119

117

⁰Sn

4-

117

0

Sn

M(CO)

subspectra. Tungsten satellite doublets ( W, I ) /

natural abundance) are observed for all three tin environments

6

M(CO)

3

,

Sn Sn

2

6

M(CO)

3

, and

Sn

3

Sn

6

-

4

-

3

, which give rise to weak and more complex

183

1

2

, 14.28%

intraenvironmental coupling within each Sn

4

Inorganic Chemistry, Vol. 41, No. 1, 2002 93

Campbell et al.

Figure 5. The ¹¹⁹Sn (111.922 MHz) NMR spectra of the Sn(1), Sn(2,3,4,5), and Sn(6,7,8,9) environments in the [1-Mo(CO)3(η -Sn9)]⁴^-anion. Simulated

4

spectra are the right-hand traces.

4

2

are observed as single ¹¹⁷Sn satellite doublets in the ¹¹⁹Sn

planes or between an Sn

4

plane and the apical tin atom.

The ratios J( Sn- Sn)/J(¹¹⁹Sn- Sn) and J( Sn- Sn)/

) are used to verify

the assignment of Sn/ Sn doublet satellite pairs to the

interenvironmental couplings J( Sn- Sn), J( Sn- Sn),

J

ratios are equal to the ratio of

the gyromagnetic ratios of the Sn and Sn nuclides, γ-

119

117

119

117

NMR spectrum and as single Sn satellites in the Sn

NMR spectrum. Edge- and diagonal-type intraenvironmental

couplings are expected for each plane, e.g., J(Sn(2)-Sn(4))

and J(Sn(2)-Sn(3)) for Sn(2,3,4,5) and J(Sn(6)-Sn(8)) and

J(Sn(6)-Sn(7)) for Sn(6,7,8,9). Doublet satellites associated

with couplings along diagonal paths are readily assigned,

because they are half the intensity of the doublet satellites

arising from the edge couplings. The edge couplings typified

119

117

J( Sn- Sn) (hereafter referred to as R

J

117

119

117

1

17

117

and J( Sn- Sn). The R

119

117

1

19

117

43

(

γ

Sn)/γ( Sn) ) 1.046 (hereafter referred to as R ). In

contrast, the intraenvironmental couplings, J( Sn- Sn),

119

117

94 Inorganic Chemistry, Vol. 41, No. 1, 2002

4

(η -E

9

)]⁴^-Cluster Anions

closo-[1-M(Co)

3

Sn (8.58%) and ¹¹⁷Sn (7.61%) and that of ¹⁸³W (14.28%)⁴³

4

4-

in the case of [1-W(CO)

3

(η -Sn

9

)] . In the case of [1-Cr-

4

4-

4

4-

(

CO)

3

(η -Sn

9

)] and [1-Mo(CO)

3

9

(η -Sn )] , complete qua-

drupolar collapse of the transition metal-tin couplings was

assumed. Full spectral simulations were achieved by assum-

ing C4V symmetry for the MSn cage and required several

9

thousand absolute isotopomers (see Experimental; Simulation

of NMR Spectra). The most abundant isotopomers and their

1

19

117

subspectra that comprise the Sn and Sn NMR spectra

4

4-

of the [1-M(CO)

3

9

(η -Sn )] anions are given in Tables 5

and 6. The simulated spectra are in excellent agreement with

the experimental spectra (Figures 2-5), accounting for all

of the observed spectral features including distortions arising

from second-order effects (vide infra) and for all coupling

paths consistent with the rigid bicapped square antiprismatic

Figure 6. The ⁹⁵Mo (32.593 MHz) NMR spectrum of the [1-Mo(CO)3-

geometries of the MSn

tin spin-spin couplings and their assignments is given in

Table 4. Because the spectral analyses for all three [1-M(CO)

9

clusters. The complete set of tin-

4

4-

(η -Sn9)] anion.

3

-

42

4

4-

by J(Sn(2)-Sn(3)) and J(Sn(6)-Sn(7)) were differentiated

(η -Sn )] anions are very similar, only the major isotopomer

Sn subspectra for

[1-W(CO) (η -Sn )] are described in detail in the ensuing

9

1

19

117

by their sensitivity to varying M, with the most variation

contributions and their

Sn and

4

4-

expected for the plane η -coordinated to the transition metal.

3

9

This is also reflected in both the calculated and observed

Sn-Sn distances, and in the calculated charges (see Com-

putational Results). The unequivocal assignments of the

intraenvironmental couplings allowed for definitive chemical

shift assignments of the Sn(2,3,4,5) and Sn(6,7,8,9) planes

and of the interenvironmental couplings J(Sn(1)-Sn(2)) and

J(Sn(1)-Sn(6)). The assignments of the remaining two

interenvironmental couplings, J(Sn(2)-Sn(6)) and J(Sn(2)-

Sn(8)), cannot be differentiated, because their associated

doublet satellites have the same intensities and the spectral

simulation (vide infra) is unable to distinguish between

coupling paths having the same symmetry relationships.

discussion.

The experimental and simulated ¹¹⁹Sn and ¹¹⁷Sn NMR

4

4-

spectra of the [1-W(CO) (η -Sn )] anion are depicted in

Figures 2 and 3. The resonance assigned to the apical Sn(1)

3

9

1

83

(2448.2 ppm) consists of a singlet flanked by W satellites

1

19

183

119

[J( Sn- W) ) 125 Hz] as well as by several sets of

-

1

17

Sn and

Sn satellites. The two most intense sets of

117

119

symmetrically disposed Sn and Sn satellites are assigned

to the two interenvironmental Sn-Sn couplings. The smaller

set of couplings, J( Sn- Sn) ) 357 Hz, J( Sn- Sn)

) 342 Hz, and J( Sn- Sn) ) 335 Hz (R ) 1.046), is

1

19

119

117

1

17

117

J

assigned to the coupling between Sn(1) and the nonadjacent

Within the Sn

couplings are smaller [e.g., J(Sn(2)-Sn(3)) and J(Sn(6)-

Sn(7))] than couplings whose paths pass through the cluster

8

square antiprism, we note that adjacent

basal Sn(6,7,8,9) plane, which is coordinated to the W(CO)

3

1

19

119

group, while the larger set of couplings, J( Sn- Sn) )

1

19

117

1091 Hz, J( Sn- Sn) ) 1043 Hz, and J( Sn- Sn) )

[e.g., J(Sn(2)-Sn(4)) and J(Sn(6)-Sn(8))]. Accordingly,

995 Hz (R ) 1.046), is assigned to couplings between Sn-

J

42

J(Sn(2)-Sn(6)) and its equivalents have been assigned to

(1) and the adjacent basal Sn(2,3,4,5) plane. The next most

the smaller value (Table 4; also see Computational Results,

intense satellites result from the doublet-of-doublets sub-

119

117,119

(

g) Coupling Constants).

spectra corresponding to the J( Sn(1)-

Sn(2)) doublet

The Sn and ¹¹⁹Sn NMR spectra of the [1-M(CO)

1

17

(η⁴-

119

3

whose transitions are further split into a doublet by J( Sn(1)-

4-

117,119

Sn

9

)] anions were fully assigned based on the above

Sn(6)).

The ¹¹⁷Sn and ¹¹⁹Sn resonances assigned to the basal Sn-

(2,3,4,5) plane (-735.7 ppm) and to the middle Sn(6,7,8,9)

plane (-496.9 ppm) each consist of a singlet flanked by a

considerations and confirmed using the multinuclear NMR

simulation program ISOTOPOMER.⁴⁴Spectra were simu-

1

119

lated using the natural abundances of the spin- /

2

nuclei

-

1

83

119

183

set of W doublet satellites, J( Sn- W) ) 47 and 125

(

42) With the exception of J(M-Sn(1)), all couplings listed and discussed

117

183

Hz and J( Sn- W) ) 45 and 119 Hz, respectively, and

in the paper represent one of several equivalent coupling paths. These

J coupling equivalencies are as follows: J(Sn(1)-Sn(2)) ) J(Sn(1)-

Sn(3)), J(Sn(1)-Sn(4)), J(Sn(1)-Sn(5)); J(Sn(1)-Sn(6)) ) J(Sn(1)-

Sn(7)), J(Sn(1)-Sn(8)), J(Sn(1)-Sn(9)); J(Sn(2)-Sn(3)) ) J(Sn(3)-

Sn(4)), J(Sn(4)-Sn(5)), J(Sn(2)-Sn(5)); J(Sn(2)-Sn(4)) ) J(Sn(3)-

Sn(5)); J(Sn(6)-Sn(7)) ) J(Sn(7)-Sn(8)), J(Sn(8)-Sn(9)), J(Sn(6)-

Sn(9)); J(Sn(6)-Sn(8)) ) J(Sn(7)-Sn(9)); J(Sn(2)-Sn(6)) ) J(Sn(2)-

Sn(9)), J(Sn(3)-Sn(7)), J(Sn(3)-Sn(6)), J(Sn(4)-Sn(8)), J(Sn(4)-

Sn(7)), J(Sn(5)-Sn(8)), J(Sn(5)-Sn(9)); J(Sn(2)-Sn(8)) ) J(Sn(2)-

Sn(7)), J(Sn(3)-Sn(8)), J(Sn(3)-Sn(9)), J(Sn(4)-Sn(6)), J(Sn(4)-

Sn(9)), J(Sn(5)-Sn(7)), J(Sn(5)-Sn(6)); J(M-Sn(2)) ) J(M-Sn(3)),

J(M-Sn(4)), J(M-Sn(5)); J(M-Sn(6)) ) J(M-Sn(7)), J(M-Sn(8)),

J(M-Sn(9)). These equivalencies were also used for averaging the

bond lengths and bond orders given in Table 7.

1

19

117

119

117

several sets of Sn and Sn satellites. The Sn ( Sn)

satellites corresponding to the interenvironmental coupling

between the Sn(2,3,4,5) and Sn(6,7,8,9) planes and the apical

(44) Santry, D. P.; Mercier, H. P. A.; Schrobilgen, G. J. ISOTOPOMER,

A Multi-NMR Simulation Program, version 3.02NTF.; Snowbird

Software, Inc.: Hamilton, ON, Canada, 2000.

(45) Emsley, J. W.; Feeney, J.; Sutcliffe, L. H. High-Resolution Nuclear

Magnetic Resonance Spectroscopy, Pergamon Press: London, 1965;

Vol. 1, Chapter 8, pp 310-320.

(46) Gay, I. D.; Jones, C. H. W.; Sharma, R. D. J. Magn. Reson. 1989, 84,

501.

(

43) Mason, J. In Multinuclear NMR; Mason, J., Ed.; Plenum Press: New

(47) Kennedy, J. D.; McFarlane, W. J. Chem. Soc., Dalton Trans. 1976,

1219.

York, 1987; Appendix, pp 626-628.

Inorganic Chemistry, Vol. 41, No. 1, 2002 95

Campbell et al.

a

b

4-

4

4-

Table 3. Calculated and Experimental Tin NMR Chemical Shifts (ppm) for Sn9 and [1-M(CO)3(η -Sn9)] (M ) Cr, Mo, W) Anions

Sn9⁴

-

Sn9Cr(CO)3⁴^-

Sn9Mo(CO)3⁴^-

Sn9W(CO)3⁴^-

atom(s)^c

Sn(1)

GIAO

expt

GIAO

expt

GIAO

expt

GIAO

expt

-530

-272

-608

-450

3069

655

592

895

2493.2

-521.9

-213.9

-50.0

2666

543

539

777

2125.1

-682.1

-402.4

-245.9

2727

440

487

715

2448.2

-735.7

-496.9

-275.8

Sn(2,3,4,5)

Sn(6,7,8,9)

d

Sn(1-9)

-1291.9

a

Calculated chemical shifts are given relative to Sn(CH3)4 (2934 ppm). Values obtained from the ¹¹⁷Sn and ¹¹⁹Sn NMR spectra recorded at 0 C in

b

o

liquid NH3. The cluster numbering schemes are given in Figure 1. ^dAverage chemical shifts.

c

Consequently, the ¹¹⁹Sn (¹¹⁷Sn) satellites in the ¹¹⁹Sn (¹¹⁷Sn)

spectrum having J/ν δ ) 0.0368 (0.0353) exhibit a more

severe second-order perturbation than those having J/ν δ )

Table 4. Spin-Spin Coupling Constants (Hz) for the

4

4-

a,b

[

1-M(CO)3(η -Sn9)] (M ) Cr, Mo, W) Anions

o

intraenvironmental J(¹¹⁷^,¹¹⁹Sn-¹¹⁷^,¹¹⁹Sn), Hz

o

anion

Sn(2)-Sn(3) Sn(2)-Sn(4) Sn(6)-Sn(7) Sn(6)-Sn(8)

0.0141 (0.0136). Similar second-order effects are noted for

_S_n₉_C_r₍_C_O₎₃4

-

the Cr and Mo analogues, and more severe second-order

effects have been observed in the Sn NMR spectra of the

129

165

154

621

673

700

544

324

238

656

739

745

4

-

119

Sn9Mo(CO)3

4

-

Sn9W(CO)3

-

_δ₎₂₀₉₅₀_H_z_,_J₍119Sn-119Sn) )

Sn[Sn(CH

3

)]

3

anion [ν

o

4

6

_i_n_t_e_r_e_n_v_i_r_o_n_m_e_n_t_a_l_J₍₁17,119_S_n_-117,119Sn), Hz

5185 Hz, J/ν

o

δ ) 0.25] and ((CH

3

)

2

SnS)

3

47

o

[ν δ ) 388 Hz,

1

19

119

J( Sn- Sn) ) 198 Hz, J/ν δ ) 0.51].

o

anion

Sn(1)-Sn(2) Sn(1)-Sn(6) Sn(2)-Sn(6) Sn(2)-Sn(8)

The experimental and simulated Sn and ¹¹⁷Sn NMR

119

Sn9Cr(CO)3⁴

Sn9Mo(CO)3⁴

-

1048

1098]

1042

1091]

1043

328

[342]

351

[367]

342

(335)

[357]

393

783

[821]

975

[1021]

943

(901)

[985]

spectra of the chromium and molybdenum clusters (Figures

[

-

4

375

[391]

363

(348)

[378]

4 and 5) are very similar to those of the [1-W(CO)

3

(η -

[

4-

9

Sn )] anion but are somewhat broader, with the line widths

Sn9W(CO)3⁴

-

being broadest for the Cr anion (see Experimental; Simula-

tion of NMR Spectra). The second-order effects observed

(

[

995)

1091]

1

17

119

for the Sn and Sn resonances of the Sn

4

planes of

)]⁴^-are also evident for the tin resonances

of the chromium and molybdenum analogues; however, the

interenvironmental J(¹¹⁹Sn-M), Hz

4

[

1-W(CO)

3

(η -Sn

9

anion

Sn9Mo(CO)3

Sn(1)-M

75^c

Sn(2)-M

Sn(6)-M

75^c

125

(119)

4

-

(M ) ⁹⁵Mo)

119

117

broad line widths of the Sn and Sn resonances of [1-Cr-

4

-

183

Sn9W(CO)3

125

47

(45)

(M ) W)

4

4-

(CO)

3

9

(η -Sn )] preclude observation of these effects for

(116)

the smaller Sn(2)-Sn(6) type interenvironmental couplings.

a

The numbering scheme is given in Figure 1. ^bValues denote J( Sn-

119

95

1

17

83

117 117 117

Although Mo and Cr possess quadrupolar isotopes ( Mo, I

Sn), whereas values in parentheses denote J( Sn- Sn) or J( Sn-

4-

97

W) in the case of Sn9W(CO)3 and those in square brackets denote

) 5/2, 15.72% natural abundance; Mo, I ) 5/2, 9.46%

J(¹¹⁹Sn- Sn). Values are the unresolved Sn and Sn satellites on

119

c

117

119

53

43

natural abundance; Cr, I ) 3/2, 9.55% natural abundance),

9

5

95

the Mo resonance at δ( Mo) ) -2002.6 ppm (30 °C).

the broader line widths are not attributable to quadrupolar

relaxation, because the natural abundances of these nuclides

are too low. The line widths of the Cr and Mo anions showed

no significant variation when recorded at 30, 0, and -70 °C

or at different field strengths (7.046 and 11.744 T), thus

apparently ruling out exchange and relaxation by quadru-

Sn(1) atom are also symmetrically disposed about the central

line in the Sn ( Sn) spectrum. However, the Sn ( Sn)

satellites corresponding to the two interenvironmental cou-

pling paths between the two Sn

and Sn(2)-Sn(8), do not symmetrically flank the central

line in the Sn ( Sn) spectrum. The R

interenvironmental couplings [1.041 for Sn(2)-Sn(6) and

.045 for Sn(2)-Sn(8)] are in excellent agreement with R

117

119

117

4

-planes, e.g., Sn(2)-Sn(6)

4

2

48

49

polar and CSA (chemical shift anisotropy) mechanisms.

1

19

117

J

ratios for these

_T_h_e95Mo NMR spectrum of [1-Mo(CO)

obtained at 30 °C in liquid NH (Figure 6) and consists of a

singlet (-2002.6 ppm) flanked by

4

9

(η -Sn )] was

4-

3

1

γ

.

117,119

95

Sn satellites [J( Mo-

119

The centroids of satellites arising from J( Sn(2)- Sn(6))

117,119

117

119

Sn) ) 75 Hz] for which the individual Sn and Sn

117

119

)

378 Hz, (J( Sn(2)- Sn(6)) ) 348 Hz) and J( Sn(2)-

Sn(8)) ) 985 Hz (J( Sn(2)- Sn(8)) ) 901 Hz) are

95

components were not resolved. Inability to observe the Mo

satellites in the Sn and Sn spectra at 30 °C and the

observation of only a single set of

assignment of the J( Mo-

coupling path. However, the J( W- Sn) and J( W-

Sn) couplings corresponding to coupling between tungsten

1

19

117

119

shifted relative to the central line of each environment, and

to high frequency and to low frequency by 1.2 and 8.8 (0.8

and 7.7) Hz for the Sn(6,7,8,9) and Sn(2,3,4,5) environments,

respectively, in the ¹Sn ( Sn) spectrum. These asym-

117,119

Sn satellites prevented

Sn) coupling to a specific

95

117,119

183

119

183

19

117

1

19

metries arise because the magnitudes of J( Sn(2)-

and the apical Sn(1) atom and between tungsten and the Sn-

1

19

Sn(6)) (J( Sn(2)- Sn(6))) and J( Sn(2)-¹¹⁹Sn(8))

117

119

117

(

6,7,8,9) ring have been assigned in their respective Sn

17

117

(

J(¹Sn(2)- Sn(8))) are a significant percentage of the

119

and Sn spectra and are essentially equal in magnitude. It

frequency difference between the Sn(2,3,4,5) and Sn(6,7,8,9)

95

117,119

is therefore reasonable to assign J( Mo-

Sn) ) 75 Hz

environments, ν

spectrum, yielding J/ν

0.0353), respectively, and are attributed to second-order

o

δ ) 26 736 (25 557) Hz in the ¹Sn ( Sn)

19

117

o

δ ratios of 0.0141 (0.0136) and 0.0368

(48) Howarth, O. In Multinuclear NMR; Mason, J., Ed.; Plenum Press:

New York, 1987; Chapter 5, pp 150-152.

(

49) Howarth, O. In Multinuclear NMR; Mason, J., Ed.; Plenum Press:

4

5

effects that give rise to weakly coupled AB spin systems.

New York, 1987; Chapter 5, p 149.

96 Inorganic Chemistry, Vol. 41, No. 1, 2002

4

(η -E

9

)]⁴^-Cluster Anions

closo-[1-M(Co)

3

Table 5. The Most Abundant Isotopomers and Subspectra Comprising the ¹¹⁹Sn and ¹¹⁷Sn NMR Spectra of the [1-W(CO)3(η -Sn9)] Anion

4

4-

1

19

117

0

¹¹⁹Snx Sny Sn9-x-y-

117

0

¹¹⁹Snx Sny Sn9-x-y-

117

0

Snx Sny Sn9-x-y-

1

83

a

_m_u_l_t_i_p_l_i_c_i_t_yb

apical Sn(1)

183_W_z_W₁_-_za

_m_u_l_t_i_p_l_i_c_i_t_yb

Sn(2-5) belt

183

_W_z_W₁_-_za

_m_u_l_t_i_p_l_i_c_i_t_yb

Sn(6-9) belt

WzW1-z

fractional

isotopomer

fractional

¹¹⁹^,¹¹⁷Sn

¹¹⁹^,¹¹⁷Sn isotopomer

environ abundance ¹¹⁹Sn ¹¹⁷Sn

a

x

y

z

environ abundance^a¹¹⁹Sn ¹¹⁷Sn

x

y

z

environ abundance^a¹¹⁹Sn ¹¹⁷Sn

x

y

z

1

0

1

0

1

0

(1)

(1, 2-5)

(1, 6-9)

0.0179

0.0030

0.0065

S

-

1

0

1

0

1

0

(2-5)

0.0716

0.0119

0.0065

S

d10

d1

-

1

0

1

(6-9)

0.0716

0.0119

S

d9

-

d11

d1

d2

-

(2-5)

-

d1′′

d2′′

(1,2-5)

d1′′

1

0

(1,6-9)

(2,6)

(2,8)

(6,7)

(6,8)

0.0065

0.0260

0.0130

0.0065

d2

d3

d4

d7

d8

d2′′

d3′′

d4′′

d7′′

d8′′

1

2

1

0

(2,6)

(2,8)

(2,3)

(2,4)

0.0260

0.0130

0.0065

0.0073

d3

d4

d5

d6

d1′

d3′′

d4′′

d5′′

d6′′

-

2

0

(1, 2-5)

(1, 6-9)

0.0073

d1′

d2′

-

(1,2-5)

2

1

0

1

(1,6-9)

(2,6)

(2,8)

(6,7)

(6,8)

(6-9)

0.0073

0.0147

0.0073

0.0037

0.0635

0.0106

d2′

d3′

d4′

d7′

d8′

-

S

2

0

1

2

0

1

0

(2,6)

(2,8)

(2,3)

(2,4)

(2,5)

(1,2-5)

0.0147

0.0073

0.0037

0.0635

0.0106

0.0058

d3′

d4′

d5′

d6′

-

0

1

2

0

1

0

(1)

(1, 2-5)

(1, 6-9)

0.0159

0.0026

0.0058

-

S

d11′′

d1′′

d2′′

d10′′

d1′′

-

d9′′

0

2

0

(1,6-9)

(2,6)

(2,8)

(6,7)

(6,8)

0.0058

0.0115

0.0058

0.0029

-

d2′′

d3′′

d4′′

d7′′

d8′′

0

2

0

(2,6)

(2,8)

(2,3)

(2,4)

0.0115

0.0058

0.0029

-

d3′′

d4′′

d5′′

d6′′

a

Natural abundances of the spin- /2 nuclides used to calculate isotopomer abundances were taken from ref 43: ¹¹⁷Sn, 7.61%; ¹¹⁹Sn, 8.58%; ¹⁸³W, 14.28%.

1

15

b

The natural abundance of Sn (0.35%) is too low to contribute detectable isotopomer subspectra and is combined with the spinless tin nuclides. S denotes

1

19

117

119

183

119

a singlet; d denotes a doublet arising from J( Sn- Sn) or J( Sn- W); d′ denotes a doublet arising from J( Sn- Sn); d′′ denotes a doublet arising

1

17

117

183

from J( Sn- Sn) or J( Sn- W). The right-hand subscript on d, d′, and d′′ denotes the coupling path, i.e., 1 ) J(Sn(1)-Sn(2)), 2 ) J(Sn(1)-Sn(6)),

3

) J(Sn(2)-Sn(6)), 4 ) J(Sn(2)-Sn(8)), 5 ) J(Sn(2)-Sn(3)), 6 ) J(Sn(2)-Sn(4)), 7 ) J(Sn(6)-Sn(7)), 8 ) J(Sn(6)-Sn(8)), 9 ) J(Sn(6)-W), 10 )

J(Sn(2)-W), 11 ) J(Sn(1)-W). With the exception of J(Sn(1)-W), each of the designated couplings represents two or more equivalent coupling paths (see

ref 42).

Table 6. The Most Abundant Isotopomers and Subspectra Comprising the ¹¹⁹Sn and ¹¹⁷Sn NMR Spectra of the [1-M(CO)3(η -Sn9)] Anions Where

4

4-

(M ) Cr, Mo) (see Table 5 for footnotes)

multiplicity^b

apical Sn(1) ¹¹⁹Snx¹¹⁷Sny Sn9-x-yM^a

multiplicity^b

Sn(2-5) belt ¹¹⁹Snx¹¹⁷Sny Sn9-x-yM^a

multiplicity^b

Sn(6-9) belt

fractional

isotopomer

fractional

isotopomer

fractional

isotopomer

1

¹⁹Snx¹¹⁷

0

a

0

Sny Sn9-x-yM

¹¹⁹^,¹¹⁷Sn environ abundance ¹¹⁹Sn ¹¹⁷Sn

a

x

y

¹¹⁹^,¹¹⁷Sn environ abundance^a¹¹⁹Sn ¹¹⁷Sn

x

y

¹¹⁹^,¹¹⁷Sn environ abundance^a¹¹⁹Sn ¹¹⁷Sn

x

y

1

0

1

(1)

(1,2-5)

(1,6-9)

0.0209

0.0076

S

d1

d2

-

1

0

1

(2-5)

0.0835

0.0076

S

d1

-

1

0

(6-9)

0.0835

S

-

d1′′

d2′′

(1,2-5)

d1′′

1

(1,6-9)

(2,6)

(2,8)

(6,7)

(6,8)

0.0076

0.0152

0.0076

d2

d3

d4

d7

d8

d2′′

d3′′

d4′′

d7′′

d8′′

1

2

1

0

(2,6)

(2,8)

(2,3)

(2,4)

0.0152

0.0076

0.0086

d3

d4

d5

d6

d1′

d3′′

d4′′

d5′′

d6′′

-

2

0

(1,2-5)

(1,6-9)

0.0086

d1′

d2′

-

(1,2-5)

2

1

0

(1,6-9)

(2,6)

(2,8)

(6,7)

(6,8)

0.0086

0.0171

0.0086

0.0043

0.0741

d2′

d3′

d4′

d7′

d8′

-

S

2

0

1

2

(2,6)

(2,8)

(2,3)

(2,4)

(2-5)

(1,2-5)

0.0171

0.0086

0.0043

0.0741

0.0067

d3′

d4′

d5′

d6′

-

S

0

1

2

(1)

(1,2-5)

(1,6-9 )

0.0185

0.0067

-

S

d1′′

d2′′

(6-9)

-

d1′′

0

2

(1,6-9)

(2,6)

(2,8)

(6,7)

(6,8)

0.0067

0.0135

0.0067

0.0034

-

d2′′

d3′′

d4′′

d7′′

d8′′

0

2

(2,6)

(2,8)

(2,3)

(2,4)

0.0135

0.0067

0.0034

-

d3′′

d4′′

d5′′

d6′′

4

-

to one or both of the analogous coupling paths in [1-Mo-

reaction of Pb

9

3

with Mo(CO) ‚mes in en initially gave a

4

4-

(

CO)

Characterization of the [1-Mo(CO)

by NMR Spectroscopy. Solutions obtained from the reac-

3

(η -Sn

9

)] .

red-brown solution and led to the observation of two broad,

equally intense Pb resonances at -1869 ppm (∆ν1/2 )

1200 Hz) and -3100 ppm (∆ν1/2 ) 990 Hz). Each resonance

was accompanied by a broad satellite doublet corresponding

4

)]⁴^-Anion

207

3

(η -Pb

9

4-

tions of Mo(CO)

periods of time required for acquisition of the Pb NMR

spectra in en and liquid NH , decomposing to give emerald

green NH solutions characteristic of the paramagnetic Pb

anion, lead mirrors, and pale yellow en solutions. The

3

‚mes with Pb

9

are unstable over the

207

to J( Pb- Pb) ≈ 4500 Hz and is respectively assigned

3

to the Pb(6,7,8,9) and Pb(2,3,4,5) planes of the [1-Mo(CO) -

3

3-

4

4-

3

9

(η -Pb

9

)] anion by analogy with the tin chemical shift

assignments for the [1-M(CO)

1

3

4

4-

3

9

(η -Sn )] anions (see Char-

Inorganic Chemistry, Vol. 41, No. 1, 2002 97

Campbell et al.

the Mo/Sn and Cr/Pb anions optimized as C symmetry, the

remaining anions optimized as C symmetry.

4

)]⁴^-(M ) Cr, Mo, W)

acterization of the [1-M(CO)

3

(η -Sn

9

1

Anions by NMR Spectroscopy, (a)). Unfortunately, the rapid

decomposition of these solutions prevented the observation

of the apical Pb(1) atom resonance. Samples prepared in

s

The calculated distances are in good agreement with the

experimental ones, although they are all longer by 0.10-

4

-

liquid NH

3

and recorded at -70 °C, however, yielded three

0.15 Å than in the experimental structures of the E

and

9

4

4-

resonances having considerably narrower line widths (Figure

S3). The two low-frequency resonances also displayed well-

resolved ²Pb satellites, with two satellite doublets appearing

on the higher frequency resonance, i.e., -1934 ppm (∆ν1/2

[1-M(CO) (η -E )] anions (Table 7 and also see Supporting

3

9

Information). The fact that the calculated bond distances are

longer than the experimental ones is consistent with our

07

4

- 13

previous results for Pb₉

. The calculated M-E distances

207

)

230; J( Pb- Pb) ) 660 and 4630 Hz) and -3450 ppm

are in better agreement with experiment than the E-E

distances. Both families of clusters have comparable bond

distances, with those of the Sn clusters being slightly shorter

than those of the Pb clusters, as expected. The trends

observed experimentally upon coordination of the M(CO)₃

group are reproduced, i.e., the distances within the basal ring

07

207

(

∆ν1/2 ) 200 Hz; J(²Pb- Pb) ) 4580 Hz). Although the

chemical shifts of the latter resonances are significantly

shifted to lower frequency relative to those in en at room

temperature, they are readily assigned to the pair of

resonances observed at room temperature in en solution based

on their satellite couplings. The sensitivity of the chemical

shifts to temperature and solvent is not unexpected, as these

effects have been shown to be large for a number of heavy

NMR nuclides. The third resonance at 27 ppm (∆ν1/2

50 Hz), assigned to Pb(1), does not display resolved

satellites but does show pronounced broadening near its base,

which is consistent with an unresolved pair of satellites

corresponding to that observed on the resonance assigned

to the Pb(6,7,8,9) plane (J(²Pb- Pb) ) 660 Hz). The

4

η -coordinated to the transition metal expand while those in

the middle ring contract (see Structural Characterization by

X-ray Crystallography). This is paralleled by increases and

decreases in the bond orders, respectively (Table 7). The

bond order increases are surprisingly large considering the

change of less than 0.1 Å for these distances (<3% of the

bond distance). As shown by the bond orders, the interactions

within the E(2,3,4,5) ring are somewhat weaker than the

interactions between this ring and the apical E(1) atom as

well as between this ring and the E(6,7,8,9) ring.

5

0

)

2

07

207

2

07

observation of two of the four interenvironmental Pb-

2

07

4

4-

Pb couplings expected for a rigid [1-Mo(CO)

anion establish that the fluxionality of the Pb

been modified. The most likely assignments for the observed

3

(η -Pb

9

)]

Coordination of the transition metal leads to overall

reductions in the E-E bond orders relative to those of the

9

moiety has

4-

E₉anions, and as expected, those of the basal ring are most

affected. The E(6,7,8,9) ring bond orders decrease so that

they are lower than the E(1)-E(2,3,4,5) bond orders by ∼0.1,

but they are still larger than the bond orders in the E(2,3,4,5)

ring. The bond orders between diagonal pairs of E atoms

are essentially zero in the basal E(6,7,8,9) ring, in contrast

with the diagonal bond orders in the E(2,3,4,5) ring. There

is a weak interaction between the apical E(1) atom and the

transition metal atom of 0.07-0.08 (Sn) and 0.06-0.15 (Pb)

and another weak interaction between the middle ring and

the transition metal, which increases from 0.06 (Sn) and 0.04

(Pb) for Cr, to 0.16 (Sn) and 0.15 (Pb) for Mo, and to 0.14

(Sn) and 0.16 (Pb) for W. The bond orders between the basal

ring and the transition metal show a significant increase from

Cr (0.50, Sn; 0.44, Pb) to Mo (0.74, Sn; 0.58, Pb) and W

(0.78 Sn; 0.65, Pb). The interactions of the metal with the

Pb cluster are smaller for Cr and W than with the Sn cluster

207

couplings are to apical-adjacent plane coupling, J( Pb(1)-

207

Pb(2)) ) 660 Hz, and to the interplanar coupling, J( Pb-

07

(

2)-²Pb(8)) ) 4580/4630 Hz. By analogy with the tin

anions, the two remaining interplanar couplings are expected

to be less than 660 Hz and are presumably not resolved

because of broad line widths. Alternatively, the Pb planes

4

may be rotating relative to each other and rapidly enough

on the NMR time scale to average the two interplanar

couplings to a single coupling. The broad line widths in liquid

NH

relaxation arising from the CSA (chemical shift anisotropy)

mechanism.⁴⁸It seems likely that the resonances of the Pb

3

at -70 °C may, in part, also arise from spin-lattice

4

planes are further broadened in en solution at 0 °C as a result

of additional chemical exchange processes that may be

dissociative in nature followed by rapid rearrangement of

9

the Pb cluster and/or involve intraanion migration of the

9

4

5

Mo center from an η -apical position to an η -position in a

belt with concomitant migration of a Pb atom to an apical

position. The latter rearrangement has been proposed by

Rudolph⁵¹and Eichhorn.

Computational Results. (a) Geometries and Bond

Orders. The fully optimized geometries show that the E

anions have C4V symmetry and are based on monocapped

square antiprisms consistent with PSEPT. The M(CO)

groups η -bonded to the nido-faces of the E

optimized gas-phase structures. In all cases for the M(CO)

complexes, no symmetry was used in the calculation. While

but are comparable for the Mo cluster.

4-

(b) Charges and Valencies. For the E₉anions, the apical

atoms charges are the most negative followed by the

E(6,7,8,9) rings of their nido-faces, with slightly larger

charges on the apical atoms (Table 7). The Mayer valencies

20

4

-

4-

9

range from 2.88 to 2.90 for Sn and from 2.92 to 3.00 for

9

4

-

Pb . Bonding of the M(CO) group to the nido-face of an

9

4

3

-

3

E₉anion leads to a significant change in the electron

populations, with significant delocalization of charge from

4

4-

9

anions in the

4

-

3

the E₉cluster onto the M(CO) group. The largest change

3

is calculated for the basal ring that binds to the transition

metal. For Cr, this ring has charges of -0.13 (Sn) and -0.25

(50) Brevard, C.; Granger, P. In Handbook of High-Resolution Multinuclear

NMR, John Wiley & Sons: 1981; p 40.

(Pb) on each of the E(6,7,8,9) atoms. In the case of tin, the

(51) Pretzer, W. R.; Rudolph, R. W. J. Am. Chem. Soc. 1976, 98, 1441.

basal ring atom charges for Mo (+0.03) and W (+0.04) are

98 Inorganic Chemistry, Vol. 41, No. 1, 2002

4

(η -E

9

)]⁴^-Cluster Anions

closo-[1-M(Co)

3

Table 7. Calculated Bond Distances, Mayer Bond Orders, Mulliken Charges and Mayer Valencies in the E9⁴and [1-M(CO)3(η -E9)] Anions (E )

-

4

4-

a

Sn, Pb; M ) Cr, Mo, W)

Sn9⁴

-

Sn9Cr(CO)3⁴^-Sn9Mo(CO)3⁴^-Sn9W(CO)3⁴^-

Pb9⁴^-

Pb9Cr(CO)3⁴^-Pb9Mo(CO)3⁴^-Pb9W(CO)3⁴^-

Bond Distances and Bond Orders

3.112 [0.56] 3.21 [0.64]

E(1)-E(2)

E(1)-E(6)

E(1)-M

3.113 [0.60]

[0.13]

3.117 [0.55]

[0.11]

3.114 [0.56]

[0.10]

3.196 [0.61]

[0.10]

3.20 [0.56]

[0.10]

3.199 [0.61]

[0.09]

[0.10]

[0.08]

[0.11]

[0.07]

[0.08]

[0.15]

[0.08]

[0.06]

E(2)-E(3)

E(2)-E(4)

E(2)-E(6)

E(2)-E(8)

E(2-5)-M

E(6)-E(7)

3.420 [0.35]

4.84 [0.11]

3.133 [0.60]

[0.14]

3.352 [0.44]

4.74 [0.11]

3.146 [0.52]

[0.09]

[0.06]

3.196 [0.50]

4.52

2.935 [0.50]

1.42

0.14

2.00

0.07

3.382 [0.41]

4.78 [0.11]

3.133 [0.51]

[0.09]

[0.16]

3.229 [0.45]

4.56

3.068 [0.74]

1.33

0.27

1.86

0.11

3.378 [0.42] 3.50 [0.36]

4.78 [0.11] 4.94 [0.07]

3.132 [0.50] 3.23 [0.63]

3.419 [0.47]

4.84 [0.07]

3.233 [0.54]

[0.06]

[0.04]

3.266 [0.57]

4.62

3.025 [0.44]

1.49

0.16

1.97

0.08

3.437 [0.46]

4.86 [0.06]

3.236 [0.51]

[0.05]

[0.15]

3.297 [0.52]

4.66

3.155 [0.58]

1.34

0.29

1.81

0.10

3.437 [0.46]

4.86 [0.06]

3.232 [0.48]

[0.05]

[0.16]

3.303 [0.48]

4.67

3.155 [0.65]

1.38

0.27

1.80

0.08

[0.09]

[0.14]

[0.11]

3.134 [0.66]

4.43

3.239 [0.43] 3.24 [0.69]

b

E(6)-E(8)

4.58

E(6)-M

M-C

M-O

C-O

3.079 [0.78]

1.35

0.26

1.85

0.09

C-C

Charges

E(1)

E(2-5)

E(6-9)

M

C

O

-0.529

-0.397

-0.474

-0.423

-0.370

-0.130

-0.950

+0.106

-0.323

-0.399

-0.350

+0.026

-1.267

-0.018

-0.327

-0.398

-0.358

+0.042

-1.185

-0.054

-0.330

-0.591

-0.365

-0.488

-0.485

-0.336

-0.246

-0.503

+0.110

-0.339

-0.459

-0.350

-0.060

-0.668

-0.069

-0.341

-0.457

-0.360

-0.018

-0.717

-0.093

-0.345

c

+

1.376

1.597

+2.000

+2.305

+2.064

+2.338

+0.968

+1.187

+1.712

+1.901

+1.880

+2.031

∆(6-9)

∆(1-9)

+

Valencies

E(1)

E(2-5)

E(6-9)

M

C

O

2.894

2.882

2.903

2.768

2.861

2.932

6.989

3.768

2.251

2.755

2.855

3.041

8.470

3.620

2.226

2.755

2.853

3.030

8.584

3.577

2.218

3.001

2.924

2.960

2.926

2.892

2.939

6.884

3.763

2.226

2.912

2.914

2.987

7.864

3.646

2.198

2.912

2.915

2.970

8.261

3.581

2.191

a

For simplicity, and because the E9 unit is very close to C4V point symmetry in all the anions, only average values are given. A full listening can be found

in Tables S1 and S2 of the Supporting Information. The atom numbering schemes used for the bond distances and bond orders are given in Figure 1. Charges

and valencies are per atom in the designated rings. The bond order is essentially zero. ^c∆(m-n) denotes [total charge on atoms m-n of [1-M(CO)3(η -

b

4

E9)] ] - [total charge on atoms m-n of E9⁴^-].

4-

actually predicted to be positive, with increased electron

transfer to tungsten. The transition metal charges in the Sn

clusters are negative: -0.95 for Cr, increasing to -1.27 for

Mo, and then decreasing to -1.18 for W. The negative

Comparisons of the atomic charge differences between the

4

-

4

4-

E

9

and [1-M(CO)

reveal that 86-93% of the charge that is delocalized onto

the M(CO) originates from the basal E(6,7,8,9) ring that is

3

(η -E

9

)] anions, ∆(m-n) (Table 7),

3

4

charges on the transition metals in the Pb

9

clusters are about

η -bonded to the transition metal. The apical E(1) atoms

donate 6-9% of the delocalized charge, while the middle

E(2,3,4,5) rings donate comparable amounts of charge, i.e.,

1-10%. The large charge depletions from the E(6,7,8,9)

rings are consistent with the assigned E-E spin-spin

half that of the Sn analogue and show an increase in the

negative charge as the atomic number of the transition metal

increases. There is a decrease of 0.11-0.13 (Sn, Pb) e on

the apical E(1) atom, whereas the middle E(2,3,4,5) ring

shows only a small change on the order of 0.03-0.05 (Sn)

4

couplings and chemical shifts of the E rings (see Charac-

4

4-

and 0.005-0.03 (Pb) e per atom. The valencies in the E

9

terization of the [1-M(CO)

3

(η -Sn

9

)] (M ) Cr, Mo, W)

)] Anions by NMR Spectroscopy).

c) NBO Populations. The NBO populations (Table 8)

show that the valence s orbital populations do not change

on complexation to M(CO) . The p orbital populations show

a significant change, and it is from the p orbitals that the

bulk of the charge is transferred to the M(CO) fragment.

For the E fragment, the charge from the apical atom is

transferred from the in-plane p orbital perpendicular to the

-axis of the E cluster. There is very little change in the

orbitals along the C -axis. The total p orbital populations in

4

4-

clusters show only small changes on complexation with the

metal, suggesting a strong electrostatic interaction (Table 7).

The valency for the Cr is about 7.0 (Sn) and 6.9 (Pb), as

would be expected for interactions with four E atoms and

three CO groups. The valencies for the Mo and W anions

increase to 8.5-8.6 (Sn) and 7.9-8.3 (Pb).

and [1-Mo(CO)

3

(η -Pb

9

(

3

There is less charge transferred from the Pb

9

cluster to

9

the M(CO) groups. The differences are most apparent for

3

the Pb atom charges of the basal Pb(6,7,8,9) ring, which,

unlike the Sn analogues, are always negative, and the charges

on the transition metals, which are always less negative for

C

4

9

4

the middle ring change significantly for the [1-M(CO) (η -

Sn )] series and show some charge transfer for the

9

3

4

4-

4

4-

[

1-M(CO)

is also seen in the transition metal valencies and in the Pb-M

bond orders, which are lower for the Pb clusters.

3

(η -Pb

9

)] than for [1-M(CO)

3

(η -Sn

9

)] . This

4

4-

[1-M(CO)

3

9

(η -Pb )] series. However, the charge-transfer

9

per atom for the middle E(2,3,4,5) ring is smaller than the

Inorganic Chemistry, Vol. 41, No. 1, 2002 99

Campbell et al.

Table 8. Natural Atomic Orbital Populations for the Valence Atomic Orbitals of the E9⁴^-and [1-M(CO)3(η -E9)] (E ) Sn, Pb; M ) Cr, Mo, W)

4

4-

Anions

Sn9⁴

-

Sn9Cr(CO)3⁴^-

Sn9Mo(CO)3⁴^-

Sn9W(CO)3⁴^-

(1)

(2,3,4,5) (6,7,8,9)

(1)

(2,3,4,5) (6,7,8,9)

Cr

(1)

(2,3,4,5) (6,7,8,9)

Mo

(1)

(2,3,4,5) (6,7,8,9)

W

s

1.67

1.17

0.86

0.85

1.69

0.83

0.88

0.94

1.67

0.93

1.09

0.80

1.70

1.19

0.74

1.69

0.86

1.01

0.76

1.66

0.77

0.97

0.43 1.70

1.20

1.69

0.87

0.73

1.02

1.66

0.85

0.87

0.90

0.43 1.70

1.19

1.69

0.86

0.98

0.77

1.67

0.81

0.98

0.76

0.50

px

py

pz

dxy

dxz

0.75

0.74

0.75

1.22

1.21

1.35

1.37

1.09

1.25

1.29

1.19

1.18

1.28

1.32

1.29

dyz

dx2-y2

dz2

1.35

1.24

Pb9⁴

-

Pb9Cr(CO)3⁴^-

Pb9Mo(CO)3⁴^-

Pb9W(CO)3⁴^-

(2,3,4,5) (6,7,8,9)

(1)

(2,3,4,5) (6,7,8,9)

(1)

(2,3,4,5) (6,7,8,9)

Cr

(1)

(2,3,4,5) (6,7,8,9)

Mo

(1)

W

s

1.75

1.08

0.86

1.77

0.82

0.85

0.91

1.76

0.90

1.04

0.80

1.76

1.09

0.77

1.77

0.85

0.86

1.74

0.74

0.66

1.02

0.41 1.77

1.10

1.77

0.86

0.77

0.92

1.75

0.81

0.72

1.00

0.42 1.77

1.09

1.77

0.84

0.81

0.88

1.76

0.80

1.01

0.70

0.51

px

py

pz

dxy

dxz

0.76

0.77

0.78

1.18

1.31

1.36

1.08

1.09

1.22

1.28

1.15

1.16

1.23

1.32

1.29

dyz

dx2-y2

dz2

1.34

1.23

charge transfer for the apical E(1) atom. There is a significant

charge transfer from the basal E(6,7,8,9) ring to the M(CO)

orbitals on the apical Sn(1) atom that are perpendicular to

the C -axis.

Several of the highest occupied molecular orbitals as well

3

4

fragment. The largest amounts of charge transferred per basal

E atom are for Cr (0.31 e for E ) Sn and 0.32 e for E )

Pb). The least amounts of charge transferred are for Mo,

where the calculated amounts are 0.20 and 0.21 e for Sn

and Pb, respectively. These values can be compared with

transfers from the apical atom of 0.17-0.21 e for Sn and

4

4-

as the LUMO are depicted for [1-Cr(CO)

Figure 8. The focus of the discussion is on the orbitals

involving the Sn fragment and/or its interaction with the

Cr(CO) fragment that also provides a qualitative picture of

the bonding of the M(CO) and E fragments for the

₎_]4- anions of the series consid-

ered in this study. The HOMO and NHOMO of [1-Cr(CO)

3

(η -Sn

9

)] in

9

3

9

0

.16-0.17 e for Pb. The metal atoms have ∼0.4 e in the

4

remaining [1-M(CO)

3

(η -E

9

valence s orbitals of the lighter metals Cr and Mo and ∼0.5e

in the s orbital of the W atom. The d orbital populations for

all of the transition metals range from ∼1.1 to ∼1.4 e.

3

-

4

4-

(

η -Sn

Sn fragment, and there is a small delocalization of charge

into the d and d orbitals of the transition metal. The next

9

)] are formed from the degenerate HOMO of the

9

(

d) Molecular Orbitals. The highest occupied molecular

xz

yz

4-

orbitals for Sn

9

are shown in Figure 7. The HOMO for

orbital, HOMO-2, involves a mixing of the d_zorbital on

2

the cluster is a degenerate orbital of E symmetry and is

predominantly composed of radial p orbitals involving the

Sn(6,7,8,9) atoms of the nido-face. The NHOMO involves

a set of radial p orbitals on the basal Sn(6,7,8,9) atoms that

constructively interact to place a significant density in the

center of the basal ring, perfectly suited for interacting with

the Cr with the HOMO-3 orbital of the Sn cluster so that

the radial p orbital on the apical Sn(1) atom along the C₄-

9

52

axis has some interaction with the transition metal atom.

There are also some metal-carbonyl interactions in the

HOMO-3. The HOMO-5 orbital involves the mixing of the

2

NHOMO on the Sn fragment with the d_zorbital on Cr,

9

vacant orbitals on an incoming M(CO)

also some p character on the apical Sn(1) along the C

3

fragment. There is

-axis

leading to a bonding interaction. The HOMO-6 orbital

involves the interaction of the d and d orbitals with the

4

xz

yz

polarized toward the center of the cluster that is antibonding

relative to the density in the center of the basal Sn(6,7,8,9)

ring. The next orbital, HOMO-2, is degenerate with radial

density in the basal and middle ring planes. The HOMO-3

HOMO-2 orbital of the Sn cluster, again leading to a

9

bonding interaction between the Sn and Cr(CO) fragments.

9

3

The HOMO-12 orbital involves Cr-CO interactions as well

4-

as Cr-Sn interactions. The Sn₉orbital involved in HOMO-

12 of the transition metal cluster anion is HOMO-6. The

final orbital that is shown is a very strongly polarized

4

orbital is essentially pure p in character along the C -axis

with some small interactions with the basal Sn(6,7,8,9) atoms.

The orbital HOMO-4 involves the bonding between the

adjacent middle ring atoms, and there is a corresponding

orbital, HOMO-6, for the basal ring atoms. Between is a

doubly degenerate orbital, HOMO-5, involving the radial p

2

9

chromium d_zorbital interacting with the Sn fragment. The

LUMO of the anion involves an antibonding Cr-CO

interaction coupled with a p orbital on the apical Sn(1) atom

that is perpendicular to the C -axis. Several of the orbitals

4

that are depicted place electron density within the cluster

framework and possibly help account for the larger E-E

nuclear spin-spin couplings observed between nonadjacent

(

52) NHOMO is defined as the next highest occupied molecular orbital,

i.e., the second highest occupied molecular orbital. The remaining

occupied orbitals are defined by their difference from the HOMO;

for example, HOMO-2 is the third highest occupied molecular orbital.

3

tin and lead atoms (see Characterization of the [1-M(CO) -

100 Inorganic Chemistry, Vol. 41, No. 1, 2002

4

(η -E

9

)]⁴^-Cluster Anions

closo-[1-M(Co)

3

Figure 7. Wave function plots of the valence molecular orbitals for the Sn9⁴^-anion at the LDFT level contoured at 0.04.

4

4-

Figure 8. Wave function plots of the LUMO and valence molecular orbitals for the [1-Cr(CO)3(η -Sn9)] anion at the LDFT level contoured at 0.04.

η⁴-Sn

Anions by NMR Spectroscopy).

)]⁴^-(M ) Cr, Mo, W) and [1-Mo(CO)

(η -Pb

4

)]⁴^-

(e) Infrared and Raman Spectra of the E

9

4-

and

(

9

3

9

4

4-

[1-M(CO)

3

(η -Sn

9

)] Anions (E ) Sn, Pb; M ) Cr, Mo,

Inorganic Chemistry, Vol. 41, No. 1, 2002 101

Campbell et al.

‚mes⁵⁶as well as for Mo(CO)

(Table S4, footnotes f and g) can be

attributed to the low effective masses of the organic groups

coordinated to the M(CO) moieties.

The carbonyl stretching frequencies in [1-M(CO)

W). The experimental Raman and infrared frequencies and

Mo(CO)

3

‚mes, and W(CO)

(en)

3

-

4-

4

4-

their assignments for the E

9

and [1-M(CO)

3

(η -Sn

9

)]

(en) and W(CO)

2

3

2

anions are listed in Tables S3 and S4 of the Supporting

Information, respectively, along with the calculated frequen-

cies. The experimental Raman and infrared frequencies for

3

₍_η4_-

3

4

4-

[

1-M(CO)

3

(η -Sn

9

)] are listed in the Experimental Section

4-

-1

9

E )] (1681-1830 cm ) (see Experimental Section, Infra-

(

Infrared Spectroscopy and Raman Spectroscopy).

red Spectroscopy) are shifted to lower frequencies relative

(

i) E ⁴

9

-

. The only experimental vibrational data reported

3

to those of their M(CO) precursors and are among the

56

4-

thus far for the E

9

anions is for Sn

Sn (91, 102, 133, and 149 cm ) and

[Sn

work). Calculated vibrational frequencies have been previ-

9

4

in K Sn

9

(99 and

lowest C-O stretching frequencies reported for terminally

-

1

-1 53

1

46 cm ) and Cs

4

9

bonded metal carbonyls. The ν(CO) values of the [1-M(CO) -

3

-

1

4

4-

4

for [2,2,2-crypt-K]

4

9

] (108, 128, 148, 168 cm ; present

(η -E )] anions are similar to those of the [1-M(CO) (η -

9

3

-

55

Pn )] (Pn ) P, As, Sb; M ) Cr, Mo, W) anions. The

7

4

- 13

4

4-

ously reported for Pb

9

.

low ν(CO) values for [1-M(CO) (η -E )] are consistent

3

9

4

-

with enhanced metal-CO d f π* back-bonding resulting

from the buildup of negative charge on the transition metal

The calculated vibrational frequencies for the Sn

9

and

Pb ⁴

-

anions were derived from the fully optimized C4V

9

geometries of the anions at the LDFT level, and are all below

atoms that results from electron donation from the E

9

cage

-

1

50 cm , showing that the clusters are quite deformable.

into the frontier d orbitals of the metal (see (b) Charges and

4-

4

4-

As expected, the Pb

than those of Sn

9

modes appear at lower frequencies

Valencies). The C-O stretches of the [1-M(CO)

3

(η -Pb

9

)]

4-

. The experimental Raman spectrum of

anions are lower in frequency than those of the corresponding

9

4-

+

4

4-

Sn

9

in the 2,2,2-crypt-K salt comprises four broad peaks

[1-M(CO)

3

9

(η -Sn )] anions, and this trend is reproduced

having frequencies that are in good agreement with the

calculated values, although it is clear from their broadness

that each of these peaks spans a range of frequencies. The

calculated values are slightly underestimated. The low-

frequency values indicate that there is little separability in

terms of stretches, bends, torsions, etc., so we provide only

general assignments.

in the calculated values. The experimental and calculated

trends appear to be counterintuitive, however, suggesting that

4

-

4-

Pb

to the conclusion drawn from a consideration of the changes

in the calculated charge distributions in the E cage and on

the transition metal upon coordination of the M(CO) group.

is the better π donor (see

9

is a better π donor than Sn

9

. The trend is opposite

9

3

4

-

The latter clearly shows that Sn

9

4

4-

(b) Charges and Valencies). It appears that the relative CO

stretching frequencies are not a reliable gauge of the relative

π donor strengths of the coordinated E cages.

9

(

ii) [1-M(CO)

calculated at the LDFT level for their fully optimized C

point group symmetries, which is also the symmetry observed

3

9

(η -E )] . The vibrational frequencies were

1

4

4-

The features in the range 417-698 cm^-¹are mostly

associated with M-C stretching motions and M-C-O

bends. These modes were too weak to be observed in the

for the [1-M(CO)

tures.

3

9

(η -Sn )] anions in their crystal struc-

The three highest frequency modes are readily assigned

to the CO stretches, with the highest frequency being

assigned to the in-phase C-O stretch. The two low-frequency

bands are approximately equal and are assigned to the out-

of-phase C-O stretching modes, which are degenerate in

4

4-

Raman spectra of [1-M(CO)

3

9

(η -Sn )] . There are small

contributions from the M-E stretches in several of these

-

1

bands. The frequencies below 150 cm are associated with

the motions of the E fragment: the E-M stretches; the

9

C-M-C, E-M-C, and E-M-E bends; and the M(CO)

3

an isolated M(CO) fragment. There is good agreement

torsion. The torsional frequencies for rotation of the M(CO)

3

between the calculated and observed values, with the

calculated values being 30-80 cm^-higher than the experi-

mental values, as expected for calculated frequencies that

do not include anharmonic corrections. As observed experi-

mentally, the lowest C-O stretches are calculated for the

Cr complexes, consistent with the largest dfπ* interaction

between the CO groups and the transition metal of lowest

atomic number. With the exception of [1-Cr(CO)

the three vibrational bands expected for the ν(C-O) modes

are observed for each [1-M(CO)

1

moieties relative to the basal planes of the E cages are very

9

low. This is in accord with the tin NMR spectra of the

4

4-

207

[

1-M(CO)

of the [1-Mo(CO)

with free rotation of the M(CO)

3

9

(η -Sn )] anions and the Pb NMR spectrum

4

4-

3

(η -Pb

9

)] anion, which are consistent

tripods (see Characterization

)] (M ) Cr, Mo, W) and [1-Mo-

(CO) (η -Pb )]^Aⁿⁱ^oⁿ^s^b^y^N^M^R^S^p^e^c^t^r^o^s^c^o^p^y⁾^.^T^h^e^M^E9

3

4

4-

of the [1-M(CO)

3

(η -Sn

9

4

4-

4

4-

3

(η -Sn

9

)] ,

3

9

cage modes and various bends associated with the cage atoms

are strongly coupled, preventing a simplified description of

4

4-

3

9

(η -E )] anion in the

the modes, and further demonstrates that the M(CO)

3

and

infrared spectra, which are in good agreement with the

4

-

_c_a_l_c_u_l_a_t_e_d_v_a_l_u_e_s_._A_s_p_r_e_v_i_o_u_s_l_y_n_o_t_e_d_,5

4,55

E₉are rather strongly bonded. The experimental Raman

the observation

‚tol,

4

4-

spectra of the [1-M(CO)

3

(η -Sn

9

)] anions exhibit several

of only two ν(C-O) bands for the precursors Cr(CO)

3

broad lines in this region (see Experimental Section, Raman

4

-

Spectroscopy) and, as in the case of the Sn

9

anion, each

(

53) von Schnering, H. G.; Baitinger, M.; Carrillo-Cabrera, W.; Bolle, U.;

Curda, J.; Grin, Y.; Heinemann, F.; Llanos, J.; Peters, K.; Schmeding,

A.; Somer, M. Z. Anorg. Allg. Chem. 1997, 623, 1037.

54) Deganello, G. In Transition Metal Complexes of Cyclic Polyolefins;

Academic Press: New York, 1979; pp 197-199.

55) Charles, S.; Eichhorn, B. W.; Rheingold, A. L.; Bott, S. J. Am. Chem.

Soc. 1994, 116, 8077.

band is comprised of several overlapping bands. Again, the

(

(56) The CO stretching frequencies for the transition metal carbonyl

-1

precursors are Cr(CO)3‚C6H6 (1970, 1894 cm ), Mo(CO)3‚mes (1957,

-

1

-

1

1883 cm ), and W(CO)3‚mes (1946, 1879 cm ).

102 Inorganic Chemistry, Vol. 41, No. 1, 2002

4

(η -E

9

)]⁴^-Cluster Anions

closo-[1-M(Co)

3

experimental frequencies are in good agreement with the

calculated ranges of values.

As a consequence, there is a smaller contribution from the

apical Sn(1) atom to the density inside the Sn(6,7,8,9) ring

for the higher lying orbitals. These results all show a

significant change in the electronic structure at the apical

Sn(1) atom and are consistent with the large deshielding of

the chemical shift that is observed upon coordination of the

(

f) Chemical Shifts. The standard used for the relative

chemical shift calculations was Sn(CH (2934 ppm).

Although the calculated chemical shifts are too high (Table

), several experimental trends are correctly predicted. The

3 4

)

3

apical Sn(1) is the most deshielded environment and is

M(CO)

3

moiety.

(g) Coupling Constants. The magnitudes of J(¹¹⁷^,¹¹⁹Sn-

calculated to be deshielded by ∼2000 ppm with respect to

117,119

the Sn

4

planes, which is comparable with experiment. Most

Sn) can be correlated with the s character in the skeletal

bonding and are consistent with an enhanced Fermi contact

importantly, the weighted averages of the experimental and

4

-

57,58

calculated (in parentheses) tin chemical shifts of Sn

9

and

contribution

to the tin-tin spin-spin couplings within

4

4-

of the [1-M(CO)

3

(η -Sn

9

)] anion series exhibit the same

these clusters. The Fermi contact contribution to the spin-

spin coupling is proportional to the s electron densities at

the coupled nuclei and is expected to be an important

contributor to the spin-spin coupling mechanism when

orbitals of high s character are involved in the coupling

4-

trends, i.e., -1291.9 (-450) ppm, Sn

9

< -275.8 (715)

ppm, W < -245.9 (777) ppm, Mo < -50 (895) ppm, Cr.

As observed experimentally, the smallest Sn(1) chemical shift

is that predicted for the Mo complex. The averages for the

other eight Sn atoms are in error by 992, 1033, and 1080

ppm for Cr, Mo, and W, respectively, which are similar to

1

19

117

mechanism. The weighted averages for the J( Sn- Sn)

4

4-

couplings of the nonfluxional [1-M(CO) (η -Sn )] anions,

3

9

4

-

the error in the average chemical shift for Sn

9

. The

589 Hz (Cr), 617 Hz (Mo), and 596 Hz (W), are significantly

1

19

117

calculated differences in the chemical shifts of the two Sn

4

larger than the exchange averaged J( Sn- Sn) coupling

4

4-

planes are small for the [1-M(CO)

(

3

(η -Sn

4-63 ppm) when compared with that calculated for Sn

336 ppm) and therefore do not allow definitive assignments

9

)] anion series

in Sn₉(256 Hz, -10 °C in liquid NH solvent) and in

Sn Tl (372 Hz, 30 °C in liquid NH solvent). The NBO

3

4

-

5-

9

8

3

populations (vide supra) show that the s orbital populations

4

4-

of the Sn

4

planes to be made for the [1-M(CO)

3

(η -Sn

9

)]

remain unaltered upon coordination of the M(CO) group to

3

4-

anions whose experimental differences are much greater

the Sn₉and Pb₉clusters with the donated electron density

(

239-308 ppm). It is noteworthy that only in the case of

coming from the p orbitals. In the case of the [1-M(CO) -

3

4

4-

tungsten is the experimental trend reproduced. The differ-

ences between calculated and experimental values are

attributed to the need for a better basis set and/or for the

need to include relativistic effects. In addition, the calculated

chemical shifts represent gas-phase values, and heavy atom

chemical shifts display large solvent-induced effects on their

chemical shifts that may be on the order of several hundred

ppm (see Characterization of the [1-Mo(CO)

Anion by NMR Spectroscopy and ref 50).

The fact that the apical chemical shift is significantly

deshielded on complexation can be seen from many of the

(η -Sn )]

anions, the total p orbital populations are

9

diminished over the triad by 1.53 (Cr), 1.10 (Mo), and 1.44

4-

(W), with respect to those of Sn₉, resulting in enhancement

of the s character of the cluster bonding in these anions and

1

19

117

corresponding increases in their average J( Sn- Sn)

couplings.

As noted above ((f) Chemical Shifts), there are significant

4

4-

3

9

(η -Pb )]

bond orders between the basal and apical atoms in the Sn

9

4

4-

moieties of the [1-M(CO)

3

(η -Sn

9

)] anions and between

the apical atom and the transition metal. These interactions

117,119

183

likely account for the fact that the J(

Sn- W) coupling

electronic structure indicators and is consistent with the

“

basal and apical atoms in the Sn

between the two apical atoms, Sn(1) and W, is similar in

magnitude to the coupling between W and the adjacent Sn-

antipodal effect”.³

9,40

The large bond orders between the

fragment suggest that

1

19

117

9

(

6,7,8,9) plane. The observation that the J( Sn- Sn)

significant changes in the electronic environment at the apical

Sn(1) atom occur when charge is transferred from the basal

Sn(6,7,8,9) ring to the metal on complexation. In addition,

the significant bond order between the apical Sn(1) atom

and M shows that the M further perturbs the electronic

environment about the apical atom. In addition, the molecular

couplings between nonadjacent tin atoms are larger than

between adjacent ones may be accounted for in similar terms.

Conclusion

4

)]⁴^-and closo-[1-W(CO)

(η⁴-

The closo-[1-Mo(CO)

3

(η -E

9

3

4

-

E )] cluster anions (E ) Sn, Pb) have been structurally

9

orbitals of the Sn

binding to the Cr(CO)

anion shows a radial p orbital along the C

9

fragment show a significant change on

4

-

characterized in the solid state by vibrational spectroscopy

3

moiety. The NHOMO of the Sn

-axis mixing with

9

and X-ray crystallography. The 22 (2n + 4) electron nido-

4

-

4

E

9

anions are found to η -coordinate the zero-electron

a set of radial p orbitals in the basal plane, and most of the

density on the apical atom is inside the Sn(6,7,8,9) ring

donor groups M(CO) (M ) Cr, Mo, W) to their vacant

3

apical positions, resulting in 10-atom (2n + 2) heteronuclear

(Figure 7). In contrast, this orbital in the complex (HOMO-

clusters that are based on bicapped square antiprisms. In

5

) has essentially no density on the apical atom (Figure 8).

4

-

4-

contrast with the fluxional nido-Sn

MSn cages of the closo-[1-M(CO)

9

and nido-Pb

9

, the

Rather, the HOMO-3 of Sn

Cr complex and mixes strongly with the d ²

9

is now the HOMO-2 of the

orbital on Cr.

4

4-

9

3

(η -Sn

9

)] (M ) Cr,

z

This orbital, which had more s character and was polarized

essentially completely external to the ring, now has more p

character and is only partially polarized external to the ring.

(

57) Jameson, C. J. In Multinuclear NMR; Mason, J., Ed.; Plenum Press:

New York, 1987; Chapter 4, p 89.

(58) Pople, J. A.; Santry, D. P. Mol. Phys. 1964, 8, 1.

Inorganic Chemistry, Vol. 41, No. 1, 2002 103

Campbell et al.

Mo, W) anions have been shown by ¹¹⁹Sn, ¹¹⁷Sn, and ⁹⁵Mo

NMR spectroscopy to be rigid and nonfluxional on the NMR

99%) and tetrahydrofuran (Fisher, 99.9%) were initially dried over

CaH

British Drug Houses, 99.8%), respectively, for several weeks and

then vacuum distilled onto and stored over fresh CaH powder and

2

powder (British Drug Houses, 99.5%) and sodium pieces

(

3 3

time scale in liquid NH solutions, whereas the M(CO)

2

groups are freely rotating. The simulation and summation

_o_f1 Sn and

19

117

sodium, respectively, for at least an additional week prior to use.

Anhydrous ammonia (Matheson, 99.99%) was condensed under

vacuum at -78 °C from the commercial cylinder into a previously

vacuum-dried tube containing freshly cut sodium metal and stored

at -78 °C for at least 1 week prior to use.

Sn subspectra arising from all possible

isotopomer families and their absolute isotopomers have

provided definitive assignments and interpretations of the

complex satellite spectra that result from spin-spin coupling

1

17

119

183

among the natural abundance Sn, Sn, and W nuclides

KPb2.26 and KSn2.05. The alloys were prepared as previously

19

117

of the MSn

Sn atoms were significantly deshielded relative to those of

both Sn planes, and this finding is reproduced by GIAO

calculations of the tin NMR shieldings. Lead-207 NMR

9

. The ¹Sn and Sn chemical shifts of the apical

12,13

described:

KPb2.26 (K, 0.4978 g, 12.73 mmol; Pb, 5.9621 g,

28.775 mmol), KSn₂_.₀₅(K, 0.6356 g, 16.26 mmol; Sn, 3.9553 g,

33.325 mmol).

4

)]⁴^-(E ) Sn, Pb; M ) Cr, Mo, W) Solutions

3

[1-M(CO) (η -E

9

4

4-

for NMR Spectroscopy. Apparatus and techniques for NMR

studies of the [1-Mo(CO)

NH show that the fluxionality of the Pb

modified. Density functional theory calculations at the local

3

(η -Pb

9

)] anion in en and liquid

63

sample preparation have been previously described. Solutions of

3

9

moiety has been

4

4-

the [1-M(CO)

red solutions of Sn

binary KSn2.05 alloy in liquid NH

Cr(CO) ‚tol or M(CO)

3 9

(η -Sn )] anions were prepared by reaction of deep

4-

9

, which were obtained by extraction of the

(

LDFT) level have been used to derive the energy-minimized

3

, with a ca. 40% molar excess of

4-

4

4-

geometries of the E

9

and [1-M(CO)

3 9

(η -E )] anions and

‚mes with respect to Sn ⁴

-

. The alloys were

3

9

to study the nature of the bonding in these anions, corroborate

specific trends among their NMR chemical shifts, and assign

extracted for at least 1 week and the tricarbonyl complexes were

added 1 day prior to sample isolation and NMR spectroscopic

characterization. The following reagent amounts were used: [1-Cr-

4

-

the experimental infrared and Raman spectra of the Sn

9

4

4-

4

4-

and [1-M(CO)

calculated charges and orbital populations show that the E

cages are π donors to the transition metal centers and that

the E plane bonded to the transition metal is the most charge-

3

(η -Sn

9

)] anions (M ) Mo, W). The

(CO) (η -Sn )] (KSn₂_.₀₅, 0.1392 g, 0.4929 mmol; Cr(CO) ·tol,

3

9

3

4

4-

9

0.0456 g, 0.200 mmol), [1-Mo(CO)

3 9

(η -Sn )] (KSn2.05, 0.1556

g, 0.5510 mmol; Mo(CO) ‚mes, 0.0458 g, 0.153 mmol), [1-W(CO) -

3

4

4-

(

η -Sn

9

)] (KSn2.05, 0.1546 g, 0.5474 mmol; W(CO)

3

‚mes, 0.0701

)] solution,

4

4-

g, 0.181 mmol). In the case of the [1-Mo(CO)

3

(η -Pb

9

depleted, donating the bulk of the electron density to the

transition metal.

the KPb2.26 alloy was extracted in both en (KPb2.26, 0.1088 g, 0.214

mmol; Mo(CO) ‚mes, 0.0146 g, 0.049 mmol) and liquid NH

KPb2.26, 0.0972 g, 0.192 mmol; Mo(CO) ‚mes, 0.0164 g, 0.055

mmol).

Crystal Growth. Single crystals of [2,2,2-crypt-K] salts of the

3

(

3

Experimental Section

Apparatus and Materials. The majority of compounds used

and prepared are highly air sensitive. Manipulations were performed

under rigorously anhydrous conditions and in the absence of oxygen

on a glass vacuum line or in a two-station nitrogen atmosphere

drybox as previously described.⁵⁹

Potassium metal (British Drug Houses, >99%) was cleaned as

previously described,⁶⁰and freshly cut samples were handled in a

drybox. Lead shot (British Drug Houses, 99.9%), tin granules

+

4

4-

[

3 9

1-M(CO) (η -E )] anions that were suitable for X-ray structure

determinations were grown from en solutions by slow vapor phase

diffusion of THF into these solutions. Crystals were isolated and

64

mounted as previously described; specific details follow.

4-

Solutions of Sn

9

were prepared by extraction of KSn2.05 into

en in the presence of 2,2,2-crypt. Addition of Mo(CO)

g, 0.119 mmol; ca. 40 mol % excess with respect to Sn

3

‚mes (0.0359

4-

9

) to the

(Baker Analyzed Reagent, 99.9%), and 2,2,2-crypt (1,10-diaza-4,7,-

red en extract containing 0.0948 g (0.252 mmol) of 2,2,2-crypt (ca

0 mol % deficit with respect to KSn2.05; 0.1069 g, 0.3785 mmol)

resulted in a red-brown solution. An analogous procedure was

13,16,21,24-hexaoxabicyclo[8.8.8]hexacosane; Merck, 99%) were

4

vacuum dried prior to handling in the drybox. Tricarbonylmesity-

lenemolybdenum(0) and tricarbonyltolylchromium(0) were prepared

4

4-

employed for the preparation of [1-W(CO)

3

(η -Sn

)] (M ) Mo, W) solutions for crystal growth.

The following amounts of reagents were used: KSn2.05 (0.1093 g,

.3870 mmol), W(CO) ‚mes (0.0480 g, 0.124 mmol), and 2,2,2-

crypt (0.0852 g, 0.226 mmol); KPb2.26 (0.0935 g, 0.184 mmol),

Mo(CO) ‚mes (0.0312 g, 0.104 mmol), and 2,2,2-crypt (0.1287 g,

.3418 mmol); KPb2.26 (0.0848 g, 0.167 mmol), W(CO) ‚mes

0.0407 g, 0.105 mmol), and 2,2,2-crypt (0.1222 g, 0.3246 mmol).

9

)]

and

by heating M(CO)

excess of mesitylene (British Drug Houses, 99%) and toluene

British Drug Houses, 98%), respectively, under reflux in a nitrogen

atmosphere, as described in refs 61 and 62. A small amount of dry

tetrahydrofuran (THF) was added to wash any M(CO) back into

the reaction flask. The resulting precipitate was suction filtered and

recrystallized from a CH Cl (Caledon, 99.5%)/hexane (Caledon)

6

[Alfa Inorganics, 99% (Mo), 98% (Cr)] in an

4

4-

3 9

[1-M(CO) (η -Pb

(

0

3

6

3

0

(

3

2

mixture. Tricarbonylmesitylenetungsten(0) (Strem Chemicals, 98%)

was transferred into the drybox as received and was exposed to

the atmosphere of the drybox for at least 2 days prior to use.

All solvents were thoroughly dried, transferred by vacuum

distillation, and stored in round-bottom flasks equipped with glass/

Teflon stopcocks (J. Young). Ethylenediamine (Fisher Scientific,

The resulting solutions were also red-brown in color. After 24 h,

THF was allowed to vapor diffuse into the en solutions over a period

of several days. Large (>3 mm) parallelepiped crystals were isolated

from the en/THF mixtures and were used for the crystal structure

determinations. Note that the use of an excess or deficit (up to 20

mol %) of 2,2,2-crypt in the Mo and W tin solutions resulted in

the immediate precipitation of amorphous material upon addition

of liquid THF to the en solutions. Crystals used in this study had

(

59) Borrmann, H.; Campbell, J.; Dixon, D. A.; Mercier, H. P. A.; Pirani,

A. M.; Schrobilgen, G. J. Inorg. Chem. 1998, 37, 1929.

60) Campbell, J.; DiCiommo, D. P.; Mercier, H. P. A.; Pirani, A. M.;

Schrobilgen, G. J.; Willuhn, M. Inorg. Chem. 1995, 34, 6265.

61) Muetterties, E. L.; Bleeke, J. R.; Sievert, A. C. J. Organomet. Chem.

(63) Burns, R. C.; Devereux, L. A.; Granger, P.; Schrobilgen, G. J. Inorg.

Chem. 1985, 24, 2615.

(64) Pirani, A. M.; Mercier, H. P. A.; Dixon, D. A.; Borrmann, H.;

Schrobilgen, G. J. Inorg. Chem. 2001, 40, 4823.

1979, 178, 197.

62) Nichols, B.; Whiting, M. C. J. Chem. Soc. 1959, 551.

104 Inorganic Chemistry, Vol. 41, No. 1, 2002

4

(η -E

9

)]⁴^-Cluster Anions

closo-[1-M(Co)

3

the following dimensions: 0.51 × 0.44 × 0.33 mm (Sn/Mo), 0.54

KBr pellets. The KBr pellets were prepared in the drybox by mixing

KBr, which had been previously dried under dynamic vacuum, with

3

×

0.41 × 0.34 mm (Sn/W), 0.57 × 0.39 × 0.25 mm (Pb/Mo),

3

and 0.52 × 0.43 × 0.31 mm (Pb/W).

the M(CO)

3

‚mes precursors, crystals of the [2,2,2-crypt-K]

4

[1-

Multinuclear Magnetic Resonance Spectroscopy. The ¹¹⁹Sn

M(CO) (η -Sn )]‚en, or M(CO) (en) [2,2,2-crypt-K] [1-M(CO) -

3

4

9

3

2

4

3

_a_n_d1 Sn (0 °C, NH

17

(η -Pb )]‚2.5en (M ) Mo, W). The resulting mixtures were pressed

4

3

) NMR spectra were recorded on a Bruker

9

95

119

into pellets and sealed inside a chamber equipped with AgCl

windows. The chamber was then placed inside the sample compart-

ment of the spectrometer, which was closed and purged with dry

AC-300 (7.046 T) pulse spectrometer. The Mo (0 °C, NH

3

), Sn

₎_a_n_d207_P_b₍₀_°_C_,_e_n_a_n_d_N_H

3

) NMR spectra

(

-70 °C, 0 °C, NH

3

were recorded on a Bruker AM-500 (11.744 T) pulse spectrometer.

Spectra were routinely obtained without locking (field drift < 0.1

Hz h ) using 10-mm probes broad banded over the frequency

2

N gas for at least 15 min prior to data acquisition. The spectra

consisted of 16 scans acquired with a resolution of (4 cm-1 and

-

1

ranges 14-121 MHz (7.046 T) and 23-202 MHz (11.744 T). The

a 5 kHz scan speed and were simultaneously subtracted from the

95

background which was recorded prior to spectral acquisition: Sn/

spectrometer frequencies were 32.593 MHz ( Mo), 104.631 MHz

207

117

119

Mo, 1703, 1718, 1830 cm-1; Sn/W, 1701, 1711, 1822 cm-1; Pb/

(

Pb), 106.945 MHz ( Sn) and 111.922 MHz ( Sn, 7.046 T).

Mo, 1702, 1736, 1813 cm ; Pb/W, 1686, 1697, 1804 cm^-¹.

-1

Free-induction decays were typically accumulated in 16K or 32K

memories. Spectral width settings of 25 and 100 kHz were

employed, yielding data point resolutions of 3.05 and 6.10 Hz/

data point and acquisition times of 0.328 and 0.164 s, respectively.

Relaxation delays were not applied. Typically, 10 000-100 000

transients were accumulated, depending on the concentrations and

sensitivities of the nuclides under study. Pulse-width settings

corresponding to a bulk magnetization tip angle, θ, of ∼90° were

3

Raman Spectroscopy. The FT Raman spectra of the [1-M(CO) -

4

4-

(η -Sn )] (M ) Mo, W) salts were recorded on a Bruker Equinox

9

65

55/FRA 106 spectrometer as described previously. The spectra

were recorded at -100 °C using a defocused laser spot of 400 mW

(15% less at the sample) power. Eight (Mo) and four (W) blocks

of 500 scans were averaged for a combined acquisition time of

120 (Mo) and 60(W) min. The randomly orientated single crystals

were sealed in Pyrex Lindeman capillaries inside a drybox; the

19

117

207

95

1

2.5 (¹Sn), 13.0 ( Sn), 25.0 ( Pb), and 27.0 ( Mo) µs. Line

broadening parameters used in the exponential multiplication of

capillaries had been previously dried at 250 °C under vacuum for

117

119

4-

the free induction decays were 3-10 Hz for the Sn, Sn, and

at least 1 day. The Raman spectrum of the Sn

9

anion salt was

95

Mo NMR spectra and 50-100 Hz for the ²⁰⁷Pb NMR spectra.

recorded on a Bruker RFS 100 FT Raman spectrometer as described

previously.⁶⁶The spectrum was recorded at -130 °C using a laser

power of 178 mW and averaged over 400 scans. The sample was

a microcrystalline powder sealed in a Pyrex melting point capillary

inside a drybox; the capillary had been previously dried at 250 °C

under vacuum for at least 1 day. The spectral resolutions were 4

Zero-filling to 128K of memory and 0-3 Hz line broadening factors

were employed for resolution enhancement of the ¹Sn and Sn

Lorentzian lines and for observation of all J couplings. Variable-

temperature spectra were recorded by using the variable-temperature

controllers of the spectrometers, and temperatures (accurate to (

19

117

-

1

-1

1

°C and stable to within ( 0.1 °C) were checked by placing a

cm : Sn/Mo, 93, 100, 108, 117, 124, 140 cm ; Sn/W, 83, 92,

-

1

copper constantan thermocouple into the sample region of the

probes. Samples were allowed to equilibrate for at least 5 min while

spinning before spectral accumulations were begun. The respective

108, 124, 139, 164 cm .

X-ray Crystallography. (a) Collection and Reduction of X-ray

Data. The data sets were collected on a P4 Siemens diffractometer

equipped with a Siemens SMART 1K CCD area detector and a

rotating anode with graphite-monochromated Mo KR radiation (λ

) 0.71073 Å) using the program SMART.⁶⁷The diffraction data

collection consisted of a full ψ rotation at ø ) 0° using (1200 +

50) 0.3° frames, followed by a series of short (100 frames) $ scans

at various ø and ψ settings to fill the gaps. The detector was located

at 3.991 cm from the crystal, and a complete data set was acquired

at a 2θ setting of 330°. A complete sphere of data was collected,

to better than 0.8 Å resolution. The data were reduced with the

program SAINT,⁶⁷which applied Lorentz and polarization correc-

tions to the three-dimensionally integrated diffraction spots. The

program SADABS⁶⁸was used for the scaling of the diffraction data

and the application of an empirical absorption correction based on

redundant reflections.

nuclei were referenced externally to samples of neat (CH

3 4

) Pb, neat

(

CH Sn, and 1 M aqueous Na MoO at 30 °C. The chemical shift

3

)

4

2

4

convention used was a positive (negative) shift signifies a chemical

shift to high (low) frequency of the reference sample.

Simulation of NMR Spectra. The ¹¹⁷Sn and ¹¹⁹Sn NMR spectra

4

4-

3 9

of the [1-M(CO) (η -Sn )] (M ) Cr, Mo, W) anions were

simulated using the computer program ISOTOPOMER, as described

previously.⁴⁴The process begins with the calculation of the weight

of each isotopomer family belonging to a cluster using binomial

coefficients and the isotopic abundances. The program has a

threshold option allowing the rejection of those isotopomer families

whose percent abundances are less than or equal to a preselected

value; the threshold applied for the simulations described in this

-6

work was 1 × 10 % relative abundance. The selected isotopomer

families are further expanded into absolute isotopomers by taking

all permutations of their isotopes among the sites within the

molecule. The NMR Hamiltonian is then calculated for each of

the absolute isotopomers and solved, and a subspectrum is

calculated. The resultant subspectra are weighted according to their

parent absolute isotopomer’s weight and added together to produce

the total simulated NMR spectrum for the cluster. Spectra in the

present study were not iterated. The number of absolute isotopomers

used for each environment is as follows: Sn(1), 4273 (Cr, Mo)

and 4600 (W); Sn(2,3,4,5) or Sn(6,7,8,9), 10 989 (Cr, Mo) and

(b) Solution and Refinement of the Structures. The Siemens

69

SHELXTL PLUS software package was used for the solution and

refinement of the crystal structures. The XPREP⁶⁹program was

used to confirm the unit cells and the crystal lattices. Direct methods

were used to solve the structures in the appropriate space groups

which generally located the positions of the polyanions and most

(

65) Borrmann, H.; Campbell, J.; Dixon, D. A.; Mercier, H. P. A.; Pirani,

A. M.; Schrobilgen, G. J. Inorg. Chem. 1998, 37, 6656.

66) Fir, B. A.; Mercier, H. P. A.; Sanders, J. C. P.; Dixon, D. A.;

Schrobilgen, G. J. J. Fluorine Chem. 2001, 110, 89.

67) SMART and SAINT, Release 4.05, Siemens Energy and Automation

Inc., Madison, WI, 1994.

12 917 (W). The line widths were adjusted to fit the experimental

spectrum. The chemical shifts and coupling constants used are those

reported in Tables 3 and 4, respectively.

68) Sheldrick, G. M., Siemens Area Detector Absorption Corrections

(SADABS), Personal communication, 1996.

Infrared Spectroscopy. The FT infrared spectra were recorded

on a Bio-Rad spectrometer (FTS-40) at ambient temperatures on

69) Sheldrick, G. M., SHELXTL-Plus, Release 5.03, Siemens Analytical

X-ray Instruments Inc., Madison, WI, 1994.

Inorganic Chemistry, Vol. 41, No. 1, 2002 105

Campbell et al.

+

were done with the program Gaussian 98.⁹³These calculations were

done at the local level with the basis sets given above with the

following changes: for Sn, a set of d polarization functions was

added,⁹⁴and the DZVP basis set was used for C, O, Cr, and Mo.

The NMR calculations for Sn were done with the program Gaussian

of the non-hydrogen atoms of the (2,2,2-crypt-K ) cations. Full-

matrix least-squares refinements of the positions and isotropic

thermal parameters of the assigned atoms and successive difference

Fourier syntheses revealed the positions of the remaining non-

+

hydrogen atoms of (2,2,2-crypt-K ) and any solvent molecules. In

4

93

95-98

the crystal structure of [2,2,2-crypt-K]

(

4

[1-Mo(CO)

3

(η -Sn

9

)]‚en, two

98 using the GIAO approach for treating the origin problem.

+

2,2,2-crypt-K ) cations were found to be disordered. The disorder

was treated by using a rigid body refinement (FRAG in the program

The calculations were done with the DZVP basis set of Godbout

et al.,⁸¹except for W, where the effective core potential (ECP) of

69

99

SHELXTL PLUS ). Disordered solvent molecules were generally

not located until the atoms of the anions and/or cations were refined

with anisotropic thermal parameters. The final structure solutions

involved the refinement of the non-hydrogen atoms of nondisor-

dered anions, cations, and solvent molecules with anisotropic

thermal parameters (Supporting Information), assigning the calcu-

lated positions of the hydrogen atoms [d(C-H) ) 0.96 Å, Uiso(H)

fixed to 1.2Ueq(C)], and by setting the weight factor to the

recommended values.

Computational Results. The calculations were done at the

The DFT geometry

optimizations and frequency calculations were done with the

on SGI computers. The calculations were done

at the local level with the potential fit of Vosko et al. for

correlation and Slater exchange.⁷⁹The calculations were done with

the Stuttgart group with their basis set was used. Sixty electrons

were included in the W ECP, the so-called small core ECP.

Acknowledgment. We thank the Natural Sciences and

Engineering Research Council (NSERC) of Canada for

support in the form of a research grant. We also thank

NSERC and the Ontario Ministry of Education and Training

for the award of graduate scholarships to J.C. and the German

Academic Exchange Service (DAAD) for the award of a

travel grant to H.F. We acknowledge Dr. Charles F. Campana

density functional theory (DFT) level.⁷

0-74

of Bruker AXS, Inc., Madison, WI, for acquiring a data set

4

for (CO)

3

Mo(en)

2

[2,2,2-crypt-K]

4

[1-Mo(CO)

3

(η -Pb

9

)]·2.5en

program Dgauss⁷

5-77

and Nancy Kawai of Bruker Optics, Billerica, MA, for

78

4

4-

recording the Raman spectra of the [1-M(CO)

3

(η -Sn

9

)]

(M ) Mo, W) salts. The density functional theory calcula-

the following basis sets: for Sn, a double-ú basis set was used

_w_i_t_h_t_h_e_E_C_P_o_f_K_u_e_c_h_l_e_e_t_a_l_._;₈0 for Pb, a polarized double-ú basis

tions were performed, in part, in the Molecular Sciences

Computing Facility (MSCF), William R. Wiley Environ-

mental Molecular Sciences Laboratory, Pacific Northwest

National Laboratory. The MSCF is funded by the Office of

Biological and Environmental Research, U.S. Department

of Energy. The Pacific Northwest National Laboratory is

operated by Battelle for the U.S. Department of Energy under

Contract DE-AC06-76RLO 1830.

_s_e_t_w_a_s_u_s_e_d_w_i_t_h_t_h_e_E_C_P_o_f_K_u_e_c_h_l_e_e_t_a_l_._;₈0 for Cr, C, and O,

81

the DZVP2 polarized double-ú basis set was used; and for Mo

and W, the 6s5p3d basis set with the ECP of Kuechle was used.⁸⁰

The A1 fitting set or the fitting sets derived for pseudopotentials

within DGauss were used.⁸²^,⁸³The geometries were optimized by

using analytic gradient methods, and second derivatives were also

calculated analytically. Atomic valencies following Mayer⁸⁴^-⁸⁷were

calculated. In addition, natural bond orbital (NBO) analyses⁸

8-92

+

Supporting Information Available: The [2,2,2-crypt-K -en]

(

70) Parr, R. G.; Yang, W. Density Functional Theory of Atoms and

Molecules; Oxford University Press: New York, 1989.

71) Labanowski, J.; Andzelm, J., Eds. Density Functional Methods in

Chemistry; Springer-Verlag: New York, 1991.

4

cation in [2,2,2-crypt-K]

Mo(CO) (en) complex in (CO)

)]‚2.5en (Figure S2); the Pb (104.613 MHz) NMR spectra

of the Sn(1), Sn(2,3,4,5), and Sn(6,7,8,9) environments in the

4

[1-M(CO)

3

(η -Sn

9

)]‚en (Figure S1); the

3

2

3

Mo(en)

2

[2,2,2-crypt-K]

4

[1-M(CO) -

3

4

207

(

η -Pb

9

(

72) Ziegler, T. Chem. ReV. 1991, 91, 651.

73) Salahub, D. R. In Ab Initio Methods in Quantum Chemistry-II; K. P.

Lawley, Ed.; J. Wiley & Sons: New York, 1987; p 447.

74) Jones, R. O.; Gunnarsson, O. ReV. Mod. Phys. 1989, 61, 689.

75) Andzelm, J. W.; Wimmer, E.; Salahub, D. R. In The Challenge of d

and f Electrons: Theory and Computation; Salahub, D. R., Zerner,

M. C., Eds.; ACS Symposium Series 394, American Chemical

Society: Washington DC, 1989; p 228.

4

4-

[

3 9

1-Mo(CO) (η -Pb )] anion (Figure S3); complete set of nonav-

eraged calculated bond lengths and bond angles (Table S1) and

(

Mayer valencies, Mayer bond orders, and Mulliken Charges (Table

4-

4

4-

S2) for the E

9

3 9

and [1-M(CO) (η -E )] anions (E ) Sn, Pb; M

)

Cr, Mo, W]; experimental and calculated frequencies and

(

76) Andzelm, J. In Density Functional Theory in Chemistry; Labanowski,

J. K., Andzelm, J. W., Eds.; Springer-Verlag: New York, 1991; pp

(93) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,

M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.;

Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels,

A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone,

V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.;

Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;

Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;

Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.;

Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R.

L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara,

A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,

W.; Wong, M. W.; Andres, J. L.; Gonzalez, A. C.; Head-Gordon, M.;

Replogle, E. S.; Pople, J. A. Gaussian 98, A.6; Gaussian, Inc.:

Pittsburgh, PA, 1998.

(94) Huzinaga, S.; Andzelm, J.; Klobukowski, M.; Radzio-Andzelm, E.;

Sakai, Y.; tatewaki, H. Gaussian Basis Sets for Molecular Calcula-

tions; Physical Sciences Data Series 16; Elsevier: Amsterdam, 1984.

(95) Cheeseman, J. R.; Trucks, G. W.; Keith, T. A.; Frisch, M. J. J. Chem.

Phys. 1996, 104, 5497.

(96) London, F. J. Phys. Radium (Paris) 1937, 8, 397.

(97) Ditchfield, R. Mol. Phys. 1974, 27, 789.

155-174.

(

77) Andzelm, J. W.; Wimmer, E. J. Chem. Phys. 1992, 96, 1280. DGauss

is a density functional program which is part of Unichem and is

available from Oxford Molecular.

(

78) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.

79) Slater, J. C. Phys. ReV. 1951, 81, 385.

80) Kuechle, W.; Dolg, M.; Stoll, H.; Preuss, H. Mol. Phys. 1991, 74,

1245.

(

81) Godbout, N.; Salahub, D. R.; Andzelm, J. W.; Wimmer, E. Can. J.

Chem. 1992, 70, 560.

82) Chen, H.; M. Krasowski, M.; Fitzgerald, G. J. Chem. Phys. 1993, 98,

8710.

83) Lee, C.; Chen, H. Unpublished results. See the UniChem manual,

Versions 3.0-5.0.

(

84) Mayer, I. Chem. Phys. Lett. 1983, 97, 270.

85) Mayer, I. Theor. Chim. Acta 1985, 67, 315.

86) Mayer, I. Int. J. Quantum Chem. 1986, 29, 73.

87) Mayer, I. Int. J. Quantum Chem. 1986, 29, 477.

88) Reed, A. E.; Curtiss, L. A.; Weinhold, F.Chem. ReV. 1988, 88, 899.

89) Foster, J. P.; Weinhold, F. J. Am. Chem. Soc. 1980, 102, 7211.

90) Reed, A. E.; F. Weinhold, F. J. Chem. Phys. 1983, 78, 4066.

91) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985,

(98) Wolinski, K.; Hinton, J. F.; Pulay, P. J. Am. Chem. Soc. 1990, 112,

8251.

8

3, 735.

(99) Kuechle, W.; Dolg, M.; Stoll, H.; Preuss, H. J. Chem. Phys. 1994,

100, 7535 and references therein.

(

92) Reed, A. E.; Weinhold, F. J. Chem. Phys. 1985, 83, 1736.

106 Inorganic Chemistry, Vol. 41, No. 1, 2002

4

)]⁴^-Cluster Anions

9

closo-[1-M(Co)

3

(η -E

assignments for the E ⁴^-

(Table S3) and [1-M(CO)

(η -E

4

)]⁴^-(Table

4

9

3

9

4 3 9

crypt-K] [1-M(CO) (η -Sn )]‚en, where M ) Mo, W. This material

S4) anions (E ) Sn, Pb; M ) Cr, Mo, W]; and an X-ray

is available free of charge via the Internet at http://pubs.acs.org.

crystallographic file in CIF format for the structure determinations

4

of (CO)

3

M(en)

2

[2,2,2-crypt-K]

4

[1-M(CO)

3

(η -Pb

9

)]‚2.5en and [2,2,2-

IC010695K

Inorganic Chemistry, Vol. 41, No. 1, 2002 107

Article Doi

Syntheses, crystal structures, and density functional theory calculations of the closo-[1-M(CO)3(η4-E9)]4- (E = Sn, Pb; M = Mo, W) cluster anions and solution NMR spectroscopic characterization of [1-M(CO)3(η4-Sn9)]4- (M = Cr, Mo, W)

DOI: 10.1021/ic010695k

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ic010695k

Full text of DOI:10.1021/ic010695k

Products guided by the article

R&D Labs maybe for 12083-24-8

Relevant to this article

Hot Product