Synthesis and structure of pentavalent bismuth(V) alkoxides and ligand redistribution equilibria in solution

Article abstract of DOI:10.1021/om970760r

The six new pentavalent Bi(V) alkoxide complexes Ph₃Bi(OR)₂, 1, Ph₃BiBr(OR), 2, and Ph₄Bi(OR), 3 (a, R = C₆F₅; b R = C₆Cl₅) have been prepared. Ph₄Bi(OR) was synthesized by alcoholysis of BiPh₅ with ROH. Ph₃Bi(OR)₂ and Ph₃BiBr(OR) were products of the salt elimination reaction between Ph₃BiBr₂ and NaOR. These compounds were characterized spectroscopically and by single-crystal X-ray diffraction. In the solid state, they possess distorted trigonal bipyramidal coordination geometries. Although the mixed species Ph₃BiBr(OR) may be isolated in pure form as crystalline solids, they do not exist as pure compounds in solution owing to the very rapid redistribution equilibrium Ph₃Bi(OR)₂ + Ph₃BiBr₂ ? 2Ph₃BiBr(OR). The equilibrium constants were measured using VT NMR spec-troscopy. For comparison, the equilibrium constants for the previously unreported cross-exchange reactions between the dihalides Ph₃BiX₂ (X = F, Cl, Br) were also measured at 226 K. Trends observed in the values of K_eq correlate with the difference in the electro-negativity of the mixed X species. The thermal stabilities of Ph₃Bi(OR)₂ were examined in toluene solution and in the solid state. Pyrolysis results in elimination of ROPh for both R = C₆F₅ and C₆Cl₅.

Full text of DOI:10.1021/om970760r

Organometallics 1998, 17, 1347-1354
1347
Syn th esis a n d Str u ctu r e of P en ta va len t Bism u th (V)
Alk oxid es a n d Liga n d Red istr ibu tion Equ ilibr ia in
Solu tion
Silke Hoppe and Kenton H. Whitmire*
Department of Chemistry MS 60, Rice University, 6100 Main Street,
Houston, Texas 77005-1892
Received August 25, 1997
The six new pentavalent Bi(V) alkoxide complexes Ph3Bi(OR)2, 1, Ph3BiBr(OR), 2, and
Ph4Bi(OR), 3 (a , R ) C6F5; b R ) C6Cl5) have been prepared. Ph4Bi(OR) was synthesized
by alcoholysis of BiPh5 with ROH. Ph3Bi(OR)2 and Ph3BiBr(OR) were products of the salt
elimination reaction between Ph3BiBr2 and NaOR. These compounds were characterized
spectroscopically and by single-crystal X-ray diffraction. In the solid state, they possess
distorted trigonal bipyramidal coordination geometries. Although the mixed species
Ph3BiBr(OR) may be isolated in pure form as crystalline solids, they do not exist as pure
compounds in solution owing to the very rapid redistribution equilibrium Ph3Bi(OR)2 +
Ph3BiBr2 h 2Ph3BiBr(OR). The equilibrium constants were measured using VT NMR spec-
troscopy. For comparison, the equilibrium constants for the previously unreported cross-
exchange reactions between the dihalides Ph3BiX2 (X ) F, Cl, Br) were also measured at
226 K. Trends observed in the values of Keq correlate with the difference in the electro-
negativity of the mixed X species. The thermal stabilities of Ph3Bi(OR)2 were examined in
toluene solution and in the solid state. Pyrolysis results in elimination of ROPh for both R
) C6F5 and C6Cl5.
In tr od u ction
gel precursors for production of pentavalent bismuth
oxide materials. A problematic aspect of the choice of
alkoxide ligand is that alcohols are oxidized by bismuth-
(V).6 Complexes have been previously synthesized
which have an alkoxide ligand attached to bismuth(V),
but these complexes are stabilized by chelation of the
bismuth by other donor groups on the alkoxide ligand.
For example, Dittes7 et al. recently reported tris(aryl)-
bismuth(V) tropolonato complexes which have a seven-
coordinate bismuth, and Barton, Yamamoto, and co-
workers have produced hexacoordinate bismuth species.8,9
To avoid problems with oxidation and arylation reac-
tions, the alkoxides OR (R ) C6F5, C6Cl5) were chosen
for initial study. En route to producing the species Bi-
(OR)5, we decided to establish the ability of these OR
groups to support the bismuth(V) oxidation state by pre-
paring the complexes Ar3Bi(OR)2 and Ar4Bi(OR) via salt
metathesis reactions and alcoholysis, respectively. These
results indeed show that bismuth(V) species with these
ligands are stable. During the characterization of these
compounds, the complexes gave evidence of being sub-
stitutionally labile in solution, undergoing rapid ligand
redistribution reactions. While similar reactions have
been reported for antimony(V) species,10,11 this is the
first such report for bismuth(V) complexes.
The isolation of homoleptic Bi(V) compounds is a
challenging synthetic goal. The only discrete inorganic
bismuth(V) compound is BiF5 which has been studied
little, probably due to some reports that it tends to react
explosively with organic compounds.1 In contrast, the
pentavalent chemistry of bismuth is well-established for
organometallic compounds of the type Ar3BiX2 (Ar )
aryl; X ) F, Cl, Br, pseudohalides, Ar, etc.). The
compound with X ) I is notably missing from the list
as addition of I2 to Ar3Bi results in immediate elimina-
tion of ArI to give Ar2BiI.2 Similar elimination reac-
tions occur readily at ambient temperatures for BiR5
(R ) alkyl), but in remarkable work, Seppelt and co-
workers have managed to synthesize, isolate, and
structurally characterize the very unstable BiMe5 spe-
cies at low temperature.3 Organobismuth(V) complexes
have been studied as oxidants and as aryl transfer
reagents.4,5
In approaching the challenge of creating new inor-
ganic Bi(V) compounds, we chose to examine alkoxide
ligands for two reasons: (1) the π-donation of the oxygen
atoms might stabilize the high oxidation state and (2)
the use of alkoxide ligands could lead to suitable sol-
(1) Cotton, F. A.; Wilkinson, G. Inorganic Chemistry, 5th ed.; J ohn
Wiley & Sons: New York, 1988; p 393.
(2) Gmelin Handbuch der Anorganischen Chemie; Springer-Ver-
lag: Berlin, 1977; Vol. 47.
(6) Barton, D. H. R. Pure Appl. Chem. 1987, 59, 937.
(7) Dittes, U.; Keppler, B. K.; Nuber, B. Angew. Chem. 1996, 108,
90; Angew. Chem., Int. Ed. Engl. 1996, 35, 67.
(3) Wallenhauer, S.; Seppelt, K. Angew. Chem. 1994, 106, 1044;
Angew. Chem., Int. Ed. Engl. 1994, 33, 976.
(4) Postel, M.; Dunach, E. Coord. Chem. Rev. 1996, 155, 127.
(5) Finet, J .-P. Chem. Rev. 1989, 89, 1487.
(8) Barton, D. H. R.; Charpiot, B.; Dau, E. T. H.; Motherwell, W.
B.; Pascard, C.; Pichon, C. Helv. Chim. Acta 1984, 67, 586.
(9) Yamamoto, Y.; Ohdoi, K.; Chen, X.; Kilano, M.; Akila, K.
Organometallics 1993, 12, 2, 3297.
S0276-7333(97)00760-7 CCC: $15.00 © 1998 American Chemical Society
Publication on Web 02/27/1998

1348 Organometallics, Vol. 17, No. 7, 1998
Ta ble 1. Qu a n tities of Sta r tin g Ma ter ia ls, Yield s, a n d An a lytica l Da ta
Hoppe and Whitmire
1a
2a
3a
1b
2b
3b
ROH (g)
1.84
0.62
0.77
0.88
0.45
0.11
NaH (g)
0.60
0.20
0.20
0.10
Ph3BiBr2 (g)
yield (g) (%)
anal. found
(calcd)
3.00
1.91 (47)
2.00
1.43 (59)
1.00
1.36 (88)
1.00
1.12(85)
C, 36.85 (36.7);
H, 1.94 (1.92);
Cl, 22.7 (22.57);
Br, 9.9 (10.17)
0.42 (72)
C, 51.82 (51.44);
H, 3.1 (2.88)
0.18 (54)
C, 44.1 (44.68);
C, 41.7 (40.99);
C, 37.11 (37.11);
C, 45.90 (46.03);
H, 2.1 (1.87)a
H, 2.5 (2.15)a
H, 1.74 (1.56);
H, 2.68 (2.58);
Cl, 36.43 (36.51)
Cl, 22.24 (22.65)
19F NMR (ppm) -159.2 (4),
in acetone-d6 -168.3 (4),
-175.7 (9)
1H NMR (ppm) 8.07 (2), 7.02 (m),
in benzene-d6 6.84 (m)
131
-159.6 (4),
-166.9, -173.4,
-185.35
-168.7 (4),
-177.1 (9)
8.27 (2), 7.02 (m),
7.79 (d), 7.02 (m)
8.09 (2), 6.99 (3),
6.78 (3)
155
446.2(s), 493.7(s),
689(s), 722(s),
769(s), 982(s),
1218(w), 1353(w),
1394(s), 1437(s),
1469(s), 1528(s),
1559(s), 3057(s)
8.245 (2), 7.02 (m), 8.01 (d), 7.63 (m)
6.87 (m)
6.84 (m)
mp (°C)
124
122
125
145
IR (cm-1
)
444(s), 566(w),
616(s), 644(w),
683(s), 727(s),
446(s), 619(w),
644(w), 681(s),
724(s), 799(s),
984(s), 1012(s),
1043(w), 1180(w),
1160(s), 1196(w),
1261(s), 1302(m),
1325(m), 1437(s),
1465(s), 1502(s),
1560(s), 3060(m)
3051(w), 1642(w),
1161(w), 1561(s),
1489(s), 1461(w),
1433(w), 1300(w),
1239(w), 1183(w),
1161(w), 1050(w),
990(s), 844(w),
727(s), 683(w),
644(w), 611(w),
511(w), 444(s)
444(s), 479(s),
639(s), 679(s),
726(s), 769(s),
981(s), 1041(w),
1128(w), 1172(w),
1220(s), 1398(b),
1437(s), 1465(s),
1529(s), 1555(s),
3047(s)
3049(w), 2356(w),
1563(w), 1543(w),
1527(s), 1472(w),
1455(w), 1444(s),
1367(w), 1328(w),
1206(s), 1189(sh),
1117(w), 1050(w),
1005(w), 989(s),
844(w), 761(s),
722(s), 683(s),
644(w), 473(w),
444(s)
983(s), 1011(s),
1127(w), 1155(s),
1254(w), 1305(s),
1327(w), 1433(s),
1461(s), 1505(s),
1561(s), 1627(w),
1650(w), 3055(s)
a
The elemental analysis has been repeated several times giving these averaged values. The errors are due to cocrystallization of 1a
and 2a which coexist in solution and cannot be separated completely by fractional crystallization.
P h 4BiOC6F 5 (3a ). A solution of 0.77 g (4.1 mmol) of
C6F5OH in toluene was added to 0.5 g (0.842 mmol) of BiPh5
dissolved in toluene. The yellow reaction mixture was stirred
overnight. After complete removal of the toluene, ether (15
mL) was added to the oily residue, precipitating a bright yellow
solid. The product was recrystallized from toluene/ether by
diffusion, forming yellow block-shaped crystals suitable for
single-crystal X-ray diffraction.
Exp er im en ta l Section
Gen er a l Con sid er a tion s. All reactions were carried out
under an inert atmosphere of either argon or nitrogen using
Schlenk-line, drybox, or high-vacuum techniques.12 Solvents
were freshly distilled from the appropriate drying agents prior
to use: THF (LiAlH4 followed by Na/Ph2CO), toluene (Na),
acetone-d6 (molecular sieves), benzene-d6, and toluene-d8
(CaH2). Ph3BiBr2 was synthesized according to literature
procedures.13 The phenols C6F5OH (PCR, Inc.) and C6Cl5OH
were sublimed from molecular sieves under reduced pressure.
Sodium hydride (Aldrich), received as a mineral suspension,
was washed with hexane and dried under vacuum.14 BiPh5
was synthesized according to the literature procedure.15 NMR
spectra were recorded on a Bruker AC250 (250 MHz 1H, 235.34
MHz 19F) spectrometer; chemical shifts are reported in ppm
on a scale relative to CFCl3 (19F NMR) as δ ) 0 ppm. Infrared
spectra in the 400-4000 cm-1 region were taken on a Perkin-
Elmer 1600 series spectrometer as KBr pellets. Elemental
analyses (C, H) were performed on a Carlo Erba Instruments
NA 1500 series 2 analyzer and by National Chemical Consult-
ing, Tenafly, NJ (C, H, Cl, F, Br). Mass spectra were obtained
on a Finnigan Mat 90.
P h 4BiOC6Cl5 (3b). A solution of 0.11 g (0.42 mmol) of
C6Cl5OH in 20 mL of toluene was added to 0.25 g (0.42 mmol)
of BiPh5 dissolved in 25 mL of toluene. The yellow reaction
mixture was stirred overnight. After complete removal of the
toluene, 40 mL of THF was added to the yellow residue and
stirred for 4 h. The THF solution was filtered through
diatomaceous earth. The filtrate was concentrated and placed
in the freezer (-10 °C) where yellow hexagonal crystals
suitable for single-crystal X-ray diffraction formed.
Th er m a l Decom p osition . A saturated solution of each
compound was refluxed in toluene for 14 h. The resulting
solution was injected in a GC-MS analyzer. Under the same
conditions, compound 2a was also heated with a stoichiometric
amount of tolyl bromide, followed by GC-MS.
Gen er a l Syn th esis of 1a ,b a n d 2a ,b. The NaOR (R )
C6F5, C6Cl5) was freshly prepared for each reaction by adding
a solution of the phenol in THF to NaH (excess) suspended in
the same solvent. After filtration through diatomaceous earth,
the solvent was removed under vacuum. The resulting NaOR
was dissolved in toluene and added to a toluene solution with
stoichiometric amounts of Ph3BiBr2. The reaction mixture was
stirred at 80 °C for 12 h. After the mixture was cooled to room
temperature, the solid NaBr was removed by filtration and
the filtrate was concentrated. After addition of hexane to the
concentrate and cooling to -10 °C, the pure product crystal-
lized. The quantities of starting materials, yields, and analyti-
cal data for each compound can be found in Table 1.
Equ ilibr iu m Stu d ies. Solutions of Ph3BiBr2, 1a , and 1b
in benzene-d6 were prepared in the following concentrations:
For K(2a ) [Ph3BiBr2] ) 0.025 mol/L, [1a ] ) 0.025mol/L; for
K(2b) [Ph3BiBr2] ) 0.017 mol/L, [1b] ) 0.017 mol/L. The stock
solutions were then mixed in at least seven different ratios.
The volumes of the mixed samples were kept constant (0.75
mL). The room-temperature 1H NMR spectra of the equilib-
rium mixtures showed completely resolved peaks for the ortho-
phenyl hydrogen atoms, which were carefully integrated to
give a measurement of the concentrations of all three com-
pounds coexisting in the equilibrium. For the determination
of ∆H° and ∆S°, the 1a /Ph3BiBr2 and 1b/Ph3BiBr2 mixtures
in toluene-d8 were measured at nine different temperatures
between 178 and 312 K.
(10) Moreland, C. G.; O’Brien, M. H.; Douthit, C. E.; Long, G. G.
Inorg. Chem. 1968, 7, 834.
(11) Long, G. G.; Moreland, C. G.; Doak, G. O.; Miller, M. Inorg.
Chem. 1966, 5, 1358.
(12) Shriver, D. F.; Drezdon, M. A. The Manipulation of Air-Sensitive
Compounds; J ohn Wiley and Sons: New York, 1986.
(13) Michaelis, A.; Polis, A. Ber. Deut. Chem. Ges. 1887, 20, 55.
(14) Gordon, A. J .; Ford, R. A. The Chemist’s Companion; J ohn Wiley
and Sons: New York, 1972.
For the determination of the equilibrium constants of the
triphenyl bismuth dihalide mixtures, solutions in toluene-d8
of the following concentrations were prepared: [Ph3BiF2] )
0.0134 mol/L, [Ph3BiCl2] ) 0.0143 mol/L, [Ph3BiBr2] ) 0.0139
mol/L. Mixtures were made in NMR tubes as described above.
The 1H and 19F (for fluoride-containing mixtures) spectra were
recorded at 226 K.
(15) Wittig, G.; Clauss, K. Liebigs Ann. Chem. 1952, 26, 136.

Pentavalent Bismuth(V) Alkoxides
Organometallics, Vol. 17, No. 7, 1998 1349
Ta ble 2. Cr ysta llogr a p h ic Da ta for Com p ou n d s 1-3
1a
2a
3a
1b
2b
3b
empirical formula
fw
cryst syst
space group
a, Å
b, Å
c, Å
C
806.4
triclinic
P1h
11.212(2)
11.705(2)
13.093(3)
102.06(3)
113.41(3)
106.19(3)
1411.1(5)
2
30H15BiF10O2
C24H15BiBrF5O
703.25
triclinic
P1h
9.313(2)
10.561(2)
13.122(3)
102.45(3)
90.56(3)
115.25(3)
1132.3(4)
2
C30H20BiF5O
700.44
orthorhombic
Pbca
14.112(3)
17.844(4)
19.954(4)
90
C30H15BiCl10O2
970.90
monoclinic
P21/c
16.133(3)
12.208(2)
18.416(4)
90
114.37(3)
90
C24H15BiBrCl5O C30H20BiCl5O
785.50
monoclinic
P21/c
16.319(3)
9.901(2)
17.502(4)
90
782.69
orthorhombic
Pna21
15.918(3)
11.408(2)
15.374(3)
90
R, deg
â, deg
90
90
114.64(3)
90
2570.4(9)
4
90
90
γ, deg
vol., Å3
5024.7(19)
8
3303.9(11)
4
2791.8(10)
4
Z
density (calcd),
1.898
2.063
1.852
1.952
2.030
1.862
Mg/m3
abs coeff, mm-1
F(000)
6.340
768
9.605
660
7.077
2688
6.174
1856
8.949
1480
6.817
1504
cryst size, mm
Θ range, deg
index ranges
0.3 × 0.2 × 0.01 0.05 × 0.05 × 0.05 0.3 × 0.3 × 0.1 0.4 × 0.50 × 0.05 0.5 × 0.3 × 0.3
0.3 × 0.3 × 0.3
2.20-24.99
-15 e h e 18
-10 e k e 13
-16 e l e 16
3956
2.04-22.49
-11 e h e 12
-12 e k e 0
-13 e l e 14
3902
2.20-22.50
-10 e h e 9
0 e k e 11
-14 e l e 13
3157
2.04-23.99
0 e h e 16
0 e k e 20
0 e l e 22
3948
2.06-24.98
0 e h e 17
0 e k e 13
-19 e l e 19
4574
2.35-24.99
0 e h e 17
0 e k e 11
-20 e l e 18
4001
no. of reflns colld
no. of ind reflns
3684 [R(int) )
2962 [R(int) )
0.0349]
3948
4359 [R(int) )
0.0402]
3833 [R(int) )
0.0340]
3826 [R(int) )
0.0303]
0.0187]
max, min
1.0, 0.5075
1.0, 0.7086
1.0, 0.9449
transmission
data/restraints /
parameters
3682/0/388
1.047
2962/0/289
1.034
3947/0/334
1.002
4359/0/388
1.022
3833/0/289
1.056
3826/1/335
1.012
goodness-of-fit
on F2
final R indices
[I > 2σ(I)]
0.0315
0.0756
0.0402
0.0466
0.0454
0.0397
0.0380
0.0399
final wR indices
[I > 2σ(I)]
0.0819
0.0847
0.0737
0.0893
0.0729
R indices (all data)
0.0820
0.1376
0.0945
0.0596
0.0814
wR indices (all data) 0.0827
0.0954
0.651 and
-0.876
0.1092
0.641 and
-0.722
0.0880
0.608 and
-0.636
0.0979
1.557 and
-1.942
0.0831
0.895 and
-0.619
largest diff. peak
0.957 and
-1.357
and hole, e/Å3
X-r a y Cr ysta llogr a p h y. Crystals were examined under
mineral oil, and the ones selected for the data collection were
glued to the tip of a glass fiber and transferred to a cold
nitrogen stream of a Rigaku AFC5-S four-circle diffractometer
with the Texsan 5.0 software package using Mo KR radiation
(λ ) 0.710 73 Å).16 The temperature during data collection
was -50 °C. The data for 1a , 1b, 2b, 3a , and 3b were
corrected empirically for absorption using ψ-scans. Structures
were solved, and a full-matrix least-squares refinement on F2
was performed using the SHELXTL-PLUS 5.0 package for
PC.17,18 Hydrogen atoms were taken into account at calculated
positions. All non-hydrogen atoms were refined anisotropi-
cally. Compound 3b was refined as a racemic twin; the
absolute structure parameter is 0.522(10). Data collection
parameters for all compounds are found in Table 2.
excess phenol, showing that 3 does not undergo further
alcoholysis under the given conditions. Substituted
Bi(V) alkoxide complexes Ph3BiBrx(OR)2-x (x ) 0, 1; R
) C6F5, C6Cl5) can be synthesized via salt elimination
reactions, if the alcohols used are stable in the presence
of a bismuth(V) compound. Pentachlorophenol, pen-
tafluorophenol, and their sodium salts are not oxidized
by bismuth(V) at temperatures below 80 °C. Therefore,
the reaction between Ph3BiBr2 and 2 equiv of NaOR in
toluene according to eq 2 yields 1a and 1b cleanly in
good yields. Using only 1 equiv of NaOR, the mono-
Resu lts a n d Discu ssion
Syn th esis, Str u ctu r e in Solu tion , a n d Th er m a l
Sta bility. There are two convenient methods synthe-
size pentavalent Bi(V) alkoxides, alcoholysis and salt
elimination. The reaction between BiPh5 and ROH (R
) C6F5, C6Cl5) at ambient temperature leads to the
formation of Ph4BiOR species (see eq 1). For R ) C6F5,
substituted species Ph3BiBr(OR) can be synthesized as
(16) TEXSAN 5.0, Single-Crystal Structure Analysis Software; Mo-
lecular Structure Corp.: The Woodlands, TX, 1990.
(17) Sheldrick, G. M. SHELXL-93; University of Go¨ttingen: Go¨t-
tingen, Germany, 1993.
(18) Sheldrick, G. M. SHELXTL PLUS PC 5.0; Siemens Crystal-
lographic Research Systems: Madison, WI, 1995.
a 5-fold excess of the phenol was used to achieve a high
yield. The formation of 1 was not observed, even with

1350 Organometallics, Vol. 17, No. 7, 1998
Hoppe and Whitmire
shown in eq 3, but they exist in solution in equilibrium
with Ph3BiBr2 and Ph3Bi(OR)2 (eq 4).
Ar/O2. The decomposition end product in Ar/O2 after
heating to 600 °C is Bi2O3 in all cases, suggesting that
the bismuth(V) compound undergoes a reductive elimi-
nation which may be similar to the one in solution. The
elimination of the ethers PhOR or bromobenzene could
not be confirmed with the corresponding weight loss
percentages, probably due to the extremely exothermic
processes found in the DTA curves, which cause ad-
ditional weight loss upon overheating of the sample.
Multinuclear NMR spectroscopy was used to probe
the structures of 1-3 in solution. The 1H NMR spectra
show only one environment for the phenyl groups on
the bismuth atom for compound 1 and two environ-
ments for 3. A solution of compound 2 always contains
equilibrium concentrations of 1, 2, and Ph3BiBr2 (see
eq 4). Two of the three sets of signals in the NMR
spectra correspond to 1 and Ph3BiBr2. The third set of
peaks can be assigned to 2 and shows only one environ-
ment for the phenyl groups. The 19F NMR spectra
contain only peaks due to one type of pentafluorophe-
noxy ligand for compounds 1a and 3a . The 19F NMR
spectrum of 2a contains signals due to 1a and 2a (see
eq 4), each compound having one type of pentafluo-
rophenoxy environment. The 1H and 19F NMR data
combined are in agreement with the trigonal-bipyrami-
dal coordination sphere of the bismuth atom as found
in the solid state or with high fluctionality as common
for five-coordinate complexes. The alkoxy and halogen
ligands are situated trans to each other, and the three
phenyl groups are in the equatorial plane for compounds
1 and 2. This confirms a generally accepted rule that
the most electronegative substituents always occupy the
axial positions in EX3Y2 (E ) P, As, Sb, Bi) compounds.22
For compound 3a , a concentration dependency of the
19F shift of the para-flourine atom was found in toluene
and acetone. The 19F signals for 3a show a higher order
splitting in acetone, acetonitrile, and toluene. This
complicated splitting pattern is probably due to ex-
change processes involving solvent molecules, since the
solution color changes over time to violet in the case of
acetone and to green for acetonitrile. The nature of
those solution transformations will be the subject of a
future publication.
Furthermore, we observed that the shift of the ortho-
phenyl protons is temperature and concentration de-
pendent. A similar phenomenon was observed for other
R3EX2 (E ) P, As, Sb) compounds and attributed to the
changing electronegativity of E when ligand X was the
same.23 There is no obvious explanation for the change
in the shift of the ortho-phenyl protons when E is
constant and X is varied. We believe that not only the
electronegativity but also possible inter- and intramo-
lecular interactions such as H-X bonding between the
substituent X and the ortho-hydrogen atoms could be
important, because the alkoxide compounds show in-
termolecular interactions between the phenyl hydrogen
atoms and the halogen atoms of the alkoxy group as well
as the bromine in the solid state, as shown in the
packing diagrams (Figures 7-9).
Compounds 1-3 represent the first examples of
pentavalent Bi alkoxide complexes with a simple,
nonchelating alkoxide ligand. Asthana19 mentioned the
synthesis of 1b via reaction of Ph3BiBr2 with C6Cl5OH
in the presence of triethylamine, but this synthesis could
not be reproduced in our laboratory, and the character-
istics described for that product do not match the results
reported here.
According to Finet’s suggested mechanism for aryla-
tion reactions of aromatic alcohols with bismuth(V) aryl
compounds, an intermediate structure may be formed
which contains a bismuth(V) center with an aryloxide
ligand.5 He suggests that these intermediates are very
unstable and decompose readily to O- or C-phenylated
products, depending upon the substituents on the
aromatic alcohol. For electron-withdrawing groups,
O-phenylation is predominant. Compounds 1a -2b can
be considered models of such intermediate structures
in these organic reactions, because they form PhOR (R
) C6F5, C6Cl5) in refluxing toluene as illustrated in eq
5. The ethers were detected with GC-MS spectrometry
of the toluene solution after refluxing for 1 h.
Compounds of the type Ar3BiX(OR′) could eliminate
biphenyl,21 but no detectable biphenyl was observed to
form from 1 and 2 in refluxing toluene after 14 h. Also,
Ph3BiBr2 reacts to form PhBr and Ph2BiBr in refluxing
toluene.20 A small amount of PhBr was present in the
toluene solution of Ph3BiBr(OC6F5) after refluxing for
14 h but not in solutions of Ph3BiBr(OC6Cl5). This may
occur because a solution of Ph3BiBr(OC6F5) always
contains equilibrium concentrations of Ph3BiBr2, which
would decompose to form PhBr in hot toluene. Solutions
of Ph3BiBr(OC6Cl5) contain less Ph3BiBr2 since K(2b)
is greater than K(2a ). The formation of PhOC6F5 could
also occur via reaction of PhBr with 2a . If a such a path
exists, we can expect the formation of tolylpentafluo-
rophenyl ether (TolOC6F5) when a 1:1 solution of 2a and
tolyl bromide is heated to reflux. TolOC6F5 was not
present, indicating that a bimolecular reaction between
compound 2 and Ar-Br is not the source of the ether
PhOR.
The thermal decomposition curves of all six crystalline
compounds, 1-3 (TGA) were determined under Ar and
Solid -Sta te Str u ctu r es. Similar to many pentava-
lent Bi(V) compounds, the geometry about the bismuth
(19) Asthana, A. Indian J . Chem. A 1994, 33, 687.
(20) Challenger, F. J . Chem. Soc. 1914, 105, 2210.
(21) Barton, D. H. R.; Finet, J .-P. Pure Appl. Chem. 1987, 59,
937.
(22) Muetterties, E. L.; Mahler, W.; Packer, K. J .; Schmutzler, R.
Inorg. Chem. 1964, 3, 1299.
(23) Keck, J .-M.; Klar, G. Z. Naturforsch. 1972, 27b, 591.

Pentavalent Bismuth(V) Alkoxides
Organometallics, Vol. 17, No. 7, 1998 1351
F igu r e 5. Structure of 2b with thermal ellipsoid plot at
50% probability level. Hydrogen atoms have been omitted
for clarity.
F igu r e 1. Structure of 1a with thermal ellipsoid plot at
50% probability level. Hydrogen atoms have been omitted
for clarity.
F igu r e 2. Structure of 2a with thermal ellipsoid plot at
50% probability level. Hydrogen atoms have been omitted
for clarity.
F igu r e 6. Structure of 3b with thermal ellipsoid plot at
50% probability level. Hydrogen atoms have been omitted
for clarity.
F igu r e 3. Structure of 3a with thermal ellipsoid plot at
50% probability level. Hydrogen atoms have been omitted
for clarity.
F igu r e 7. Packing diagram of 1a with intermolecular
interactions (H-F).
except for the fourth phenyl group for 3a and 3b. The
Bi-C distances are almost constant in all compounds
and fall in the range of Bi-C(aryl) distances found in
the literature (Table 3). The planes of two of the phenyl
rings in each of the compounds 1a , 1b, and 2b are
rotated with respect to the equatorial plane defined by
the Bi atom and the three attached phenyl carbon
atoms. The phenyl ring closest to the bent alkoxy
ligand(s) lies flat in the plane and has a shorter Bi-C
distance.
F igu r e 4. Structure of 1b with thermal ellipsoid plot at
50% probability level. Hydrogen atoms have been omitted
for clarity.
The C-Bi-C angles deviate substantially from the
ideal value of 120° (Table 3). One of the angles is always
greater than 120°, which is due to the spacial require-
ments of the alkoxy group. The deviation is greater for
the sterically more crowded C6Cl5. In 1b, the two
atom in the solid state is trigonal bipyramidal for all
six compounds (Figures 1-6). The more electronegative
substituents Br and -OR (R ) C6F5, C6Cl5) occupy the
axial positions and the phenyl groups are equatorial,

1352 Organometallics, Vol. 17, No. 7, 1998
Hoppe and Whitmire
In addition to those contacts, the structure is stabi-
lized by weak π-stacking forces of the C6Cl5 rings, with
a ring to ring distance of 3.86 Å compared to graphite
sheet separation of 3.35 Å. Such π-interaction also
exists in the fluorine analogue 2a ; the distance between
two stacked pentafluorophenyl rings is 2.631 Å. In
compound 1b, Cl‚‚‚Cl contacts and C-H‚‚‚Cl hydrogen
bonds were also observed (Cl(14)‚‚‚Cl(24b) 3.364 Å,
H(55a)‚‚‚Cl(22) 2.787 Å). 3b shows only weak intermo-
lecular C-H‚‚‚Cl hydrogen bonds (Cl(24)‚‚‚H(46a)
2.895 Å).
Red istr ibu tion Rea ction s/Equ ilibr iu m . 1H NMR
spectra of a solution of single crystals of 2a showed three
sets of peaks in the ortho-phenyl-H region, two of which
are characteristic for compounds 1a and Ph3BiBr2. The
third peak set was identified as that corresponding to
2a . It was concluded that all three compounds exist in
dynamic equilibrium (eq 4). This was confirmed by the
presence of the signal of 2a in the NMR spectra of
various mixtures of Ph3BiBr2 and 1a . The same obser-
vation was made for compound 2b. The integrated peak
intensity of the ortho-hydrogen atoms on the phenyl
groups was used as a measurement for the equilibrium
concentration of the coexisting compounds. The equi-
librium occurs immediately, so that kinetic measure-
ments were not possible. The alkoxide compounds are
all similarly colored, so that no visible change results
upon mixing.
F igu r e 8. Packing diagram of 2b with intermolecular
interactions (Cl-H‚‚‚Cl, Cl‚‚‚Cl).
Equilibrium constants for both reactions (eq 4) were
calculated from NMR data according to eq 6. Constant
F igu r e 9. Packing diagram of 3b with intermolecular
interactions (C-H‚‚‚Cl).
[2]2
pentachlorophenoxy ligands, which are both bent to-
ward the same equatorial phenyl group, expand the
C-Bi-C angle to 142.6°. Compound 2b has a signifi-
cantly smaller C-Bi-C angle (131.7°), because there
is only one C6Cl5 group. In all compounds, 1 and 2, the
O-Bi-O/Br angles are close to 180° (174.1(2)-179.4-
(2)°). The Bi-O distances are shorter than the ones
found by Yamamoto for hexacoordinate Bi(V) com-
pounds9 and the Bi-O distance in Dittes’ ditropolonato
complexes.7
The packing in the crystal is influenced by weak
halogen‚‚‚hydrogen and/or chlorine‚‚‚chlorine interac-
tions. Figure 7 shows a packing diagram of compound
1a with some of the hydrogen-fluorine contacts
(F(14)‚‚‚H(53a) 2.620 Å; F(24)‚‚‚H(54b) 2.690 Å). Simi-
lar contacts exist in compound 2a (F(44)-H(26b) 2.631
Å) and compound 3a (F(24)‚‚‚H(33b) 2.743 Å). For the
pentachlorophenoxides, chlorine‚‚‚chlorine contacts simi-
lar to those described by Desiraju for chloroaromatics
can be found.24 Desiraju interprets Cl‚‚‚Cl contacts in
the range from 3.27 to 4.10 Å as “short” and, therefore,
structure stabilizing. He also states that in those
structures where the shortest Cl‚‚‚Cl contacts are at
the higher end of the range, additional stabilization
seems to be achieved through C-H‚‚‚Cl “hydrogen
bonds” in the range from 2.8 to 3.3 Å. Figure 8
illustrates both types of interactions found in compound
2b . The shortest intermolecular chlorine-chlorine
distance Cl(4)‚‚‚Cl(3b) is 3.48 Å, which is stabilized by
C-H‚‚‚Cl “hydrogen bonds” of 2.881 Å.
K )
(6)
[Ph3BiBr2] [1]
K(2a ) ) 6.5 ( 0.3 was calculated for reaction 4 with R
) C6F5 in benzene at room temperature. The equilib-
rium constant for reaction 4 with R ) C6Cl5 in benzene
at room temperature is K(2b) ) 12.8 ( 0.5. These
values show that the equilibrium is shifted further
toward the mixed-ligand product Ph3BiBr(OR) when R
) C6Cl5. To understand this tendency, variable-tem-
perature NMR spectroscopy was used to determine ∆H°
and ∆S° for the equilibria. The variable-temperature
measurements were carried out in deuterated toluene,
which has a lower freezing point than benzene. The
room-temperature equilibrium constants for toluene
solutions are on the same order of magnitude as, but
not identical to, the ones in benzene (K(2a ) ) 7.4 and
K(2b) ) 16.9).
Figure 10 shows the logarithmic dependence of K on
1/T for the formation of 2a and 2b in toluene. A simple
curve fit was applied for the calculation of ∆H° and ∆S°.
ln K(2a ) ) (87.8 J mol-1)/T + 1.71 J mol-1
ln K(2b) ) (536 J mol-1)/T + 0.993 J mol-1
(R2 ) 0.998) (R2 ) 0.991)
∆H°(2a ) ) -730 J mol-1 ∆H°(2b) ) -4457 J mol-1
∆S°(2a ) ) 14 J mol-1 K-1 ∆S°(2b) ) 8 J mol-1 K-1
(24) (a) Sarma, J . A. R. P.; Desiraju, G. R. Acc. Chem. Res. 1986,
19, 222. (b) Desiraju, G. R.; Parthasarathy, R. J . Am. Chem. Soc. 1989,
111, 8725.
In both reactions, the entropy change is essentially
zero, as expected. The enthalpy changes are also small,

Pentavalent Bismuth(V) Alkoxides
Organometallics, Vol. 17, No. 7, 1998 1353
Ta ble 3. Selected Bon d Dista n ces (Å) a n d An gles (d eg)
1a
2a
3a
1b
2b
3b
Bi-C
Bi-O
2.193(7)
2.203(8)
2.196(7)
2.228(5)
2.212(5)
2.195(14)
2.220(13)
2.212(14)
2.235(9)
2.200(11)
2.233(12)
2.222(13)
2.544(7)
2.188(10)
2.188(10)
2.209(11)
2.250(7)
2.244(7)
2.189(9)
2.214(9)
2.215(9)
2.230(6)
2.181(14)
2.209(12)
2.21(2)
2.543(9)
Bi-Br/C
C-Bi-C
2.726(2)
123.0(5)
119.5(5)
116.9(5)
88.5(4)
88.7(5)
84.5(4)
91.3(3)
94.2(4)
92.5(3)
176.6(3)
2.257(11)
115.9(4)
120.1(4)
116.7(4)
78.3(4)
81.7(4)
83.0(4)
100.5(4)
100.1(4)
96.5(4)
2.7417(14)
131.7(3)
116.1(3)
112.2(3)
91.4(3)
88.2(3)
89.3(3)
87.5(2)
91.2(2)
2.22(2)
114.8(5)
107.8(6)
128.5(5)
78.1(4)
80.1(4)
82.4(5)
100.7(5)
101.5(5)
97.6(5)
175.4(4)
121.8(3)
119.3(3)
119.0(3)
87.4(2)
91.1(2)
92.7(2)
87.6(2)
89.2(2)
91.9(2)
179.4(2)
142.6(4)
113.4(4)
103.9(4)
90.4(3)
91.3(3)
88.1(3)
89.5(3)
94.4(3)
88.5(3)
174.1(2)
C-Bi-O
C-Bi-O/Br/C
O-Bi-O or Br
92.9(2)
177.8(2)
174.9(4)
Ta ble 4. Electr on ega tivity Differ en ces a n d Keq a t
226 K for P ossible X/Y Com bin a tion s
X
Y
EN(X) - EN(Y)
Keq (226 K)
Cl
Br
Br
Br
Br
F
F
1.0
1.2
38.04 ((1.2)
59.75 ((4)
2.92 ((0.18)
8.14
Cl
OC6F5
OC6Cl5
0.2
0.52a
0.95a
28.94
a
Extrapolated from the curve in Figure 11.
F igu r e 10. ln K as a function of 1/T for equilibrium 4 in
toluene.
as anticipated from a simple bond breakage and forma-
tion analysis, with ∆H° for the formation of 2b being
about 6 times greater than the one for 2a formation.
Moreland et al. observed similar redistribution reactions
for pentavalent organoantimony dihalides. They con-
cluded that the equilibrium constants are influenced by
the R group and by the electronegativity difference of
the two halides.10 Keeping the R group constant (R )
Ph), we investigated the influence of the ligand elec-
tronegativity on the redistribution equilibria. Since
pentavalent organobismuth dihalides also undergo re-
distribution reactions (see eq 7) with each other, mixed
toluene solutions of triphenyl bismuth dihalides Ph3-
BiX2 (X ) F, Cl, Br) were analyzed by NMR spectros-
copy. At room temperature, broad and unresolved
F igu r e 11. ∆G° vs the electronegativity difference of X
and Y.
function of [EN(X) - EN(Y)] vs ∆G°. Making the very
simplified assumption that this relationship still holds
for the complex OR ligands, this graph and the
values for K(2a ) and K(2b) at 226 K can be used to
estimate the effective group electronegativity of OC6F5
and OC6Cl5. By interpolation, EN(OC6F5) equals 3.32
and EN(OC6Cl5) is 3.75. To our knowledge, the group
electronegativities for those two -OR groups (OC6F5
and OC6Cl5) have not been measured. The tabulated
group electronegativity for OC6H5 is 3.5, about mid-
way between OC6F5 and OC6Cl5. Since OC6H5 is
oxidized by Bi(V) compounds, it is not possible to obtain
an experimental value for that group for comparison.
We would have expected OC6F5 to be more electrone-
gative as a group than OC6Cl5; but since OC6Cl5 is very
bulky compared to OC6F5, as shown by the Bi-O
distances for 1a and 1b (Table 3), steric factors may also
strongly influence the redistribution equilibrium con-
stant.
Keq
Ph3BiX2 + Ph3BiY2 y z 2Ph3BiXY
(7)
peaks were observed in the ortho-phenyl-hydrogen
region. At 226 K, the three signals are separated for
all three mixtures. Two signals correspond to the
parent dihalides, and the third one was attributed to
the mixed-halide species. In Table 4, the equilibrium
constants for the dihalides corresponding to eq 7 are
listed at 226 K as well as the values for the equilibrium
constants K(2a ) and K(2b) at 226 K.
It was found that the change in the free energies of
the equilibria show a linear dependence on the differ-
ence of the Pauling-electronegativity (EN) between the
ligands X and Y, which is illustrated in Figure 11 as a
Studies on R3E(X)(Y) (E ) As, Sb) compounds allowed
the determination of equilibrium constants for different

1354 Organometallics, Vol. 17, No. 7, 1998
Hoppe and Whitmire
R groups (R ) Me, Ph, Bz) and a number of halogen
combinations (X/Y ) F, Cl, Br, I) in chloroform.10 For
Me3Sb(X)(Y) equilibria, a temperature dependence of K
was reported. The general tendencies in these systems
vary from the data reported here, which may be due to
the different solvent and different R groups. A direct
comparison to the Sb system is not possible since
Me3BiX2 compounds are not stable. The formation of a
bridged dimer structure, as shown in eq 8, is a reason-
able explanation for these redistribution reactions since
they occur in benzene and toluene, which do not support
ionic structures, and no free fluoride or -ORf groups
were observed in the 19F NMR spectra. The formation
and dissociation of the dimer must be fast on the NMR
time scale.
Ack n ow led gm en t. The Robert A. Welch Founda-
tion and the National Science Foundation are gratefully
acknowledged for their financial support.
lengths and angles, and anisotropic displacement param-
eters and ORTEP diagrams for compounds 1, 2, and 3 (34
pages). Ordering information is given on any current mast-
head page.
Su p p or tin g In for m a tion Ava ila ble: Tables of thermo-
dynamic data for Figure 10 and atomic coordinates, bond
OM970760R