Regioselective protonation of the η4-anthracene ligand in [Mn(η4-C14H10)(CO)3] to give an η5-anthracenyl complex [Mn(η5-C14H11)(CO)3] that undergoes a low-energy 1,4-hydride exchange reaction

Article abstract of DOI:10.1021/om990735g

Protonation of the terminal (η⁴-anthracene) manganese tricarbonyl anion [Mn(η⁴-C₁₄H₁₀)-(CO)₃] ^- gives the terminal (η⁵-anthracenyl) manganese tricarbonyl complex [Mn(η⁵-C₁₄H₁₁)-(CO)₃]. Single-crystal X-ray diffraction analysis confirms that the proton adds to the hydrocarbon ring, and labeling studies indicate that protonation is regiospecific and occurs at the endo site. Magnetization transfer experiments establish rapid spin population transfer from the exo site to the para ring position. This is proposed to involve an intermediate metal hydride, and dynamic NMR studies have been used to determine that the free energy barrier to this exchange is ca. 14.5 kcal/mol at and below ambient temperatures. This is interpreted in terms of a modification of the mechanism of Lamanna and Brookhart for the isomerization within 6-exo-¹H-[Mn(η⁵-C₆HD₅)(CO) ₃].

Full text of DOI:10.1021/om990735g

Organometallics 2000, 19, 661-671
661
Regioselective P r oton a tion of th e η4-An th r a cen e Liga n d
in [Mn (η4-C14H10)(CO)3]- To Give a n η5-An th r a cen yl
Com p lex [Mn (η5-C14H11)(CO)3] Th a t Un d er goes a
Low -En er gy 1,4-Hyd r id e Exch a n ge Rea ction
J acqueline M. Veauthier,† Albert Chow,‡ Gideon Fraenkel,*,‡
Steven J . Geib,*,† and N. J ohn Cooper*,†
Departments of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260, and
Ohio State University, Columbus, Ohio 43210
Received September 20, 1999
Protonation of the terminal (η4-anthracene) manganese tricarbonyl anion [Mn(η4-C14H10)-
(CO)3]- gives the terminal (η5-anthracenyl) manganese tricarbonyl complex [Mn(η5-C14H11)-
(CO)3]. Single-crystal X-ray diffraction analysis confirms that the proton adds to the
hydrocarbon ring, and labeling studies indicate that protonation is regiospecific and occurs
at the endo site. Magnetization transfer experiments establish rapid spin population transfer
from the exo site to the para ring position. This is proposed to involve an intermediate metal
hydride, and dynamic NMR studies have been used to determine that the free energy barrier
to this exchange is ca. 14.5 kcal/mol at and below ambient temperatures. This is interpreted
in terms of a modification of the mechanism of Lamanna and Brookhart for the isomerization
within 6-exo-1H-[Mn(η5-C6HD5)(CO)3].
In tr od u ction
anthracene.5a Our recent observation that the η4-an-
thracene complex [Mn(η4-C14H10)(CO)3]- (1-) can be
formed by substitution of anthracene for benzene in
[Mn(η4-C6H6)(CO)3]- (Scheme 1)4 potentially offers a
broad approach to reductively activated polyaromatic
hydrocarbons and has led us to investigate the reactivity
of 1- with electrophiles.
Recent work from our group1-4 and others5-7 has
established that arenes can be activated by coordination
to highly reduced metal centers. The arenes are most
typically mononuclear arenes like benzene, and in the
activated form the arenes are bound to the metal center
in a mode in which the arene does not adopt the
familiar, planar, π-coordination geometry in which all
the arene atoms involved in the π-bonding network bond
to the metal center; the arenes are forced instead into
a reduced hapticity bonding mode (η4 or η2), and the
locus of reactivity is then on the arene because of the
loss of aromaticity in the lower hapticity coordination
state.
The thermodynamic reduction product of anthracene,
9,10-dihydroanthracene, forms as a result of dihydrogen
addition to the middle ring of anthracene, but with the
anthracene ligand bound terminally to the metal in 1-
protonation should occur at the end ring. We now wish
to report that this is indeed the case and that protona-
tion gives the η5-anthracenyl complex [Mn(η5-C14H11)-
(CO)3] (2), but that formation of the new C-H bond is
readily reversible and that 2 is fluxional through a rapid
metal-mediated 1,4-hydride migration. There is no
evidence in the structural data for any unusual ground-
state interaction between the endo C-H and the metal
center. While this work was in progress, an analogous
1,4-hydride shift within [Mn(η5-C10H9)(CO)3] was re-
ported by Georg and Kreiter.8
A few examples are known in which polyaromatic
hydrocarbons are also activated through reduced hap-
ticity coordination, but examples to date are limited to
naphthalene ligands2,5,6b,c and an Os(II) compound of
† University of Pittsburgh.
‡ Ohio State University.
(1) (a) Leong, V. S.; Cooper, N. J . J . Am. Chem. Soc. 1988, 110, 2644,
and references therein. (b) Corella, J . A.; Cooper, N. J . J . Am. Chem.
Soc. 1990, 112, 2832.
(2) (a) Leong, V. S.; Cooper, N. J . Organometallics 1988, 7, 2058.
(b) Thompson, R. L.; Lee, S.; Rheingold, A. L.; Cooper, N. J . Organo-
metallics 1991, 10, 1657.
(3) (a) Thompson, R. L.; Geib, S. J .; Cooper, N. J . J . Am. Chem. Soc.
1991, 113, 8961. (b) Lee, S. J .; Cooper, N. J . J . Am. Chem. Soc. 1991,
113, 716.
(4) (a) Lee, S.; Geib, S. J .; Cooper, N. J . J . Am. Chem. Soc. 1995,
117, 9572. (b) Lee, S. Ph.D. Thesis, University of Pittsburgh, 1992.
(5) (a) Harman, W. D. Chem. Rev. 1997, 97, 1953. (b) Winemiller,
M. D.; Harman, W. D. J . Am. Chem. Soc. 1998, 120, 7835.
(6) (a) Rieke, R. D.; Daruwala, K. P.; Schulte, L. D.; Pankas, S. M.
Organometallics, 1992, 11, 284. (b) Henry, W. P.; Rieke, R. D. J . Am.
Chem. Soc. 1983, 105, 6314. (c) Rieke, R. D.; Henry, W. P.; Arney, J .
S. Inorg. Chem. 1987, 26, 420.
Exp er im en ta l Section
Gen er a l P r oced u r es. All reactions and manipulations
were carried out under an atmosphere of dry, oxygen-free
nitrogen by means of standard Schlenk techniques or a
Vacuum Atmospheres drybox unless otherwise stated. Glass-
ware was soaked in KOH-saturated 2-propanol for ca. 12 h
and rinsed with water and acetone before it was dried. Glass
reaction vessels were oven-dried and flamed under vacuum
before use. Krytox fluorinated grease was used for all glass
joints unless otherwise stated.
(7) Son, S. U.; Lee, S. S.; Chung, Y. K. J . Am. Chem. Soc. 1997,
119, 7711.
(8) Georg, A.; Kreiter, C. G. Eur J . Inorg. Chem. 1999, 651.
10.1021/om990735g CCC: $19.00 © 2000 American Chemical Society
Publication on Web 01/26/2000

662 Organometallics, Vol. 19, No. 4, 2000
Veauthier et al.
Sch em e 1
Air sensitive filtrations were performed using positive
nitrogen pressure to force solutions through specially
atures were recorded at room temperature. Polystyrene was
used as the external standard (1601 cm-1 peak). 1H NMR
spectra were recorded on a Bruker AC 300 spectrometer at
300 MHz. 13C NMR spectra were recorded on a Bruker AC
a
constructed filtration cannula. This was made by fitting the
port end of a stainless steel cannula with a piece of hardened
filter paper (Whatman’s No. 5 and glass microfiber were used
separately or layered) and then securing it with Teflon tape.
The port of the cannula was made by welding a stainless steel
filter support to one end of the cannula (Popper).
2
300 at 75 MHz. H NMR spectra were recorded on a Bruker
DPX-300 NMR spectrometer at 46 MHz. 1H-1H COSY and
1H-1H NOESY NMR spectra were recorded on a Bruker DPX-
300 NMR spectrometer at 300 MHz. Mass spectra were
recorded on a J EOL-SX102A mass spectrometer. Peaks are
reported in units of mass per charge (m/e) as percentages of
the base peak.
Solutions were stirred magnetically. A bath temperature of
-78 °C was maintained with dry ice in acetone.
All solvents were freshly distilled under nitrogen before use.
Tetrahydrofuran (THF) and diethyl ether (Et2O) were predried
over sodium ribbon. THF was distilled from potassium, and
Et2O was distilled from sodium/benzophenone ketyl. Methyl-
ene chloride (CH2Cl2) was freshly distilled from CaH2. Benzene
(C6H6) was distilled from sodium. n-Pentane (C5H12) was
stirred over concentrated H2SO4 for more than 24 h, neutral-
ized with K2CO3, and distilled from CaH2. All deuterated NMR
solvents were purchased from Cambridge Isotope Labs and
used as received.
Unless otherwise stated, all reagents were used as received.
Manganese carbonyl [Mn2(CO)10] was purchased from Strem
Chemicals Inc. Aqueous hexafluorophosphoric acid (HPF6),
trifluoroacetic acid (CF3CO2H(D)), anthracene (C14H10), and
bis(triphenylphosphoranylidene)ammonium chloride ([PPN]Cl)
were purchased from Aldrich. Aluminum chloride (AlCl3) was
purchased from EM Science.
Dyn a m ic NMR. Variable temperature was achieved by
replacing the normal air stream to the Bruker Aspect 3000
with a stream of nitrogen gas. The gas was passed from a 20
L Dewar of liquid nitrogen through a Dewared Pyrex glass
tube. The gas was heated as needed as determined by a
thermocouple on the insert wall below the receiver coil.
Temperatures within the NMR probe were calibrated against
methanol and found to be accurate to within 0.2 °C. Samples
for variable-temperature NMR were prepared in C6D6CD3 in
5 mm NMR tubes, degassed by several freeze/pump/thaw
cycles, and then sealed under vacuum.
Population transfer studies were carried out by selectively
saturating one resonance peak at a time and then applying a
nonselective π/2 observation pulse to record the spectrum.
Pulse widths of π and π/2 were measured for each temperature
and were determined by obtaining a well-phased Fourier-
transformed FID at low power and then adjusting the pulse
width setting several times to observe the Fourier-transformed
FID (using the same phase angle as that used at low power)
which is exactly π out of phase as the π pulse width. The π/2
pulse width is then one-half the width of the π pulse.
X-r a y Cr ysta llogr a p h y. Crystals for X-ray diffraction
studies were coated with fluorolube, then mounted on a glass
fiber and coated with epoxy. X-ray data were collected on a
Siemens P3 diffractometer using graphite-monochromatized
Mo KR radiation (λ ) 0.71073 Å). Data processing and graphics
were done with the Siemens SHELXTL package program. All
non-hydrogen atoms were located and refined anisotropically.
All hydrogen atoms in [Mn(η5-C14H11)(CO)3] were located and
refined isotropically, and all hydrogen atoms in [PPN][Mn(η4-
Potassium anthracenide [K(C14H10)] was prepared by dis-
solving freshly cut potassium metal (Aldrich) in a solution of
anthracene (1.5 equiv) in THF and stirring 8 h under a
nitrogen atmosphere at room temperature. Potassium anthra-
cenide solutions were stored under nitrogen in Suba-sealed
Schlenk vessels at -70 °C.
P r ep a r a tion of [Mn (η6-C6H6)(CO)3]P F 6. Samples of [Mn-
(η6-C6H6)(CO)3]PF6 were prepared by a variation of the method
described by Wilkinson and co-workers.9 In a typical reaction,
a mixture of 5.63 g (24.4 mmol) of [Mn(CO)5Cl] and 4.50 g (33.7
mmol) of AlCl3 dissolved in 52.0 mL of anhydrous C6H6 was
refluxed for 9 h. The reaction flask was quenched with 100
mL of ice water to give an upper organic layer and lower
aqueous phase. A separatory funnel was used to collect the
lower aqueous phase that was washed with 3 × 50 mL of
pentane (or until the pentane layer became clear), while the
upper organic layer was discarded. The clean yellow aqueous
solution was treated with excess cold aqueous HPF6 to
precipitate a pale yellow powder. This was collected on a
medium Bu¨chner filter frit, washed with H2O, methanol, and
Et2O, and then dried under vacuum to yield 6.39 g (17.7 mmol,
72%) of [Mn(η6-C6H6)(CO)3]PF6: IR (νCtO only, Nujol) 2090 (s),
C
14H10)(CO)3] were calculated and placed in idealized positions
(dC-H ) 0.96 Å).
Micr oa n a lysis. Microanalyses were performed by Atlantic
Microlab, Norcross, GA.
Rea ction of [Mn (η6-C6H6)(CO)3]P F 6 w ith K(C14H10) a n d
Isola tion of [P P N][Mn (η4-C14H10)(CO)3]. A fast addition (<5
s) of 2.2 equiv of K(C14H10) to a slurry of 0.601 g (1.66 mmol)
of [Mn(η6-C6H6)(CO)3]PF6 in THF at -78 °C resulted in an
immediate change from a yellow slurry to an orange-brown
solution and finally to a orange-brown slurry. IR spectra of a
sample removed from the orange-brown slurry showed con-
sumption of the starting material and the appearance of new
1
2020 (s, br) cm-1; H NMR (300 MHz, CD3CN) δ 6.48 (s, 6H,
Ph).
Infrared spectra were recorded on a Perkin-Elmer model
783 spectrophotometer. Solution spectra were recorded with
samples in Teflon-sealed solution cells fit with Suba Seal septa
and thoroughly purged with nitrogen before use. All infrared
spectra of samples taken from solutions at specified temper-
νCtO bands (IR, THF: 1940 (s), 1845 (s), and 1820 (s) cm-1
)
assigned to [Mn(η4-C6H6)(CO)3]K. The slurry dissolved near
-42 °C. The solution was allowed to warm to room tempera-
ture and was stirred overnight. Solid [PPN]Cl (0.810 g, 1.41
(9) Winkhaus, G.; Pratt, L.; Wilkinson, G. J . Chem. Soc. 1961, 3807.

Protonation of the η4-Anthracene Ligand
Organometallics, Vol. 19, No. 4, 2000 663
mmol) was added and stirred at room temperature for 3 h.
The solvent was removed, and the orange-brown powder was
redissolved in CH2Cl2 and filtered via cannula to give a clear
dark orange solution, which was concentrated and layered with
Et2O (1:3, CH2Cl2/Et2O) to give large orange crystals of [PPN]-
[Mn(η4-C14H10)(CO)3] (1.17 mmol, 70.5%): IR (νCtO only, THF)
1940 (s), 1845 (s), 1820 (s) cm-1; IR (νCtO only, CH2Cl2) 1945
(s), 1850 (s), 1825 (s) cm-1; 1H NMR (300 MHz, CD2Cl2) δ 7.07
(dd, J 5-4 ) 6.05 Hz, J 5-4′ ) 3.31 Hz, 2H, H5,5′), 6.86 (dd, J 4-5
) 6.05 Hz, J 4-5′ ) 3.31 Hz, 2H, H4,4′), 6.34 (s, 2H, H3,3′), 5.62
(dd, J 2-1 ) 4.91 Hz, J 2-1′ ) 2.80 Hz, 2H, H2,2′), 2.79 (dd, J 1-2
dark red pentane solution formed, while much of the yellow
[PPN][Mn(η4-C14H10)(CO)3] remained behind. The red pentane
solution was filtered. IR spectra of the red solution showed
the presence of a neutral complex (IR: νCtO, pentane, 2020
(vs), 1942 (vs), 1925 (vs) cm-1). There was also a band at 1700
(m) cm-1 assigned to [PPN][CF3CO2]). Slow cooling of the
pentane solution to -20 °C gave small wine-red crystals of
[Mn(η5-C14H10D)(CO)3] (0.236 mmol, 42%): 1H NMR (300 MHz,
C6D5CD3, 257 K) δ 7.35 (d, J 8-9 ) 8.25 Hz, 1H, H8), 7.30 (s,
1H, H6), 7.17 (d, J 11-10 ) 8.25, 1H, H11), 6.98 (m, 2H, H9,10),
6.83 (s, 1H, H13), 6.06 (d, J 4-3 ) 5.3 Hz, 1H, H4), 4.95 (t, J 3-4
∼ J 3-2 ) 6.3 Hz, 1H, H3), 3.30 (d, J 2-3 ) 7.4 Hz, 1H, H2), 2.65
(s, 1H, Hexo); 2H NMR (46 MHz, C6H5CH3, 257K) δ 2.84 (s, 1D,
Dendo); 13C NMR (75 MHz, C6D5CD3, 257 K) δ 225 (s(br), 3C,
CtO), 125-135 (8C, C6-13), 108 (s, 1C, C5), 107 (s, 1C, C14),
97.0 (d, J C4-H4 ) 168 Hz, 1C, C4), 75.3 (d, J C3-H3 ) 179 Hz,
) 4.91 Hz, J 1-2′ ) 2.80 Hz, 2H, H1,1′). Anal. Calcd for C53H40
-
MnNO3P2: C, 74.39; H, 4.71; N, 1.64. Found: C, 74.47; H, 4.77;
N, 1.63.
Rea ction of K[Mn (η4-C14H10)(CO)3] w ith CF 3CO2H. A
fast addition of 2.0 equiv of K(C14H10) to a slurry of 0.200 g
(0.553 mmol) of [Mn(η6-C6H6)(CO)3]PF6 in THF at -78 °C
resulted in a change from a yellow slurry to an orange-brown
solution and finally to an orange-brown slurry. IR spectra of
a sample removed from the orange-brown slurry showed
consumption of the starting material and the appearance of
1C, C3), 50.7 (d, J C2-H2 ) 159 Hz, 1C, C2), 29.9 (dt, J C1-Hexo
)
159 Hz, 1C, C1); mass spec (EI) m/z, 319 (14.9), 291 (66.0),
263 (50.0), 235 (48.5), 180 (100), 152 (9.0), 89 (6.5), 76 (7.5),
55 (16.2).
new νCtO bands (IR, THF: 1940 (s), 1845 (s), 1820 (s) cm-1
)
Resu lts
assigned to [Mn(η4-C6H6)(CO)3]K. The slurry dissolved near
-42 °C. The solution was allowed to warm to room tempera-
ture and was stirred overnight to give [Mn(η4-C14H10)(CO)3]K.
Addition of 43 mL (0.558 mmol) of CF3CO2H resulted in an
immediate color change from orange to dark red. IR spectra
of the red solution showed the presence of a neutral complex
(IR: νCtO, THF, 2010 (s), 1928 (s), 1915 (s) cm-1). There was
also a band at 1695 (m) cm-1 assigned to [PPN][CF3CO2]. The
solvent was removed, and the red-brown powder was redis-
solved in n-pentane and filtered via cannula to give a dark
red solution. Wine-red microcrystals of [Mn(η5-C14H11)(CO)3]
(65% spectroscopic yield) were obtained by slowly cooling the
concentrated pentane solution to -20 °C.
Exploration of the reactivity of [Mn(η4-C14H10)(CO)3]-
required access to the anion in synthetically convenient
quantities. The existing preparation of the compound,
however, gives access to the anion as a [K(Kryptofix
222)]+ salt, which is too expensive for routine work.4
To develop a more convenient preparation of the anion,
we investigated metathesis of the initial anthracene
substitution product K[Mn(η4-C14H10)(CO)3] with other
cations.
Reaction of [Mn (η6-C6H6)(CO)3]P F6 with K(C14H10
)
a n d Isola tion of [P P N][Mn (η4-C14H10)(CO)3]. The
synthesis of [PPN][Mn(η4-C14H10)(CO)3] is similar to
that of[K(Kryptofix222)][Mn(η4-C14H10)(CO)3]([K(Krypto-
fix 222)]1).4 In both procedures, the reduction of [Mn-
(η6-C6H6)(CO)3]PF6 with 2.2 equiv of K[C14H10] at -78
°C in THF resulted in the consumption of the cationic
starting material (this was indicated by a change from
a yellow slurry to an orange-brown solution and by the
appearance of CtO stretching bands in the infrared at
1940 (s), 1845 (s), and 1820 (s) cm-1). The solution was
allowed to warm to ambient temperature and was
stirred overnight. Eighty-five percent of 1.0 equiv of
[PPN]Cl was then added to the THF solution, and the
mixture was stirred for 3 h. The solvent was removed
to leave an orange powder, which was recrystallized
from CH2Cl2 and Et2O (1:3) to give large orange crystals
of [PPN][Mn(η4-C14H10)(CO)3] ([PPN]1, 70%). The na-
ture of the Mn(-I) anion in 1- was established by its
1H NMR spectrum, which is similar to that of [K(Krypto-
fix 222)]1.
The yield of [PPN]1 is similar to that of [K(Kryptofix
222)]1, but metathesis must be carried out with less
than a stoichiometric amount of [PPN]Cl because of the
less than quantitative yield of [PPN]1. Use of a full
equivalent of [PPN]Cl results in the presence of excess
[PPN]Cl, and coprecipitation of [PPN]1 and [PPN]Cl
then inhibits the purification of [PPN]1.
The use of [K(18-crown-6)]+ as a counterion for 1- was
also examined. The formation of [K(18-crown-6)]1 oc-
curred in a yield similar to that for the formation of
[PPN]1, but crystals of [K(18-crown-6)]1 grown from
THF/Et2O powdered under vacuum, presumably as a
consequence of desolvation of Et2O of crystallization.
Rea ction of [P P N][Mn (η4-C14H10)(CO)3] w ith CF 3CO2H
a n d Isola tion of [Mn (η5-C14H11)(CO)3]. To a slurry of 0.198
g (0.232 mmol) of [PPN][Mn(η4-C14H10)(CO)3] in 20 mL of
n-pentane was added 17.9 mL (0.232 mmol) of CF3CO2H. A
dark red pentane solution formed, while much of the yellow
[PPN][Mn(η4-C14H10)(CO)3] remained behind. The red pentane
solution was filtered, and subsequent treatments of the yellow
solid in pentane with CF3CO2H continued to give red solutions
that were collected by filtration. IR spectra of the red solution
confirmed the presence of a neutral complex (IR: νCtO
,
pentane, 2020 (vs), 1942 (vs), 1925 (vs) cm-1). The infrared
spectrum also contained a band at 1700 (m) cm-1 assigned to
[PPN][CF3CO2], which is sparingly soluble in pentane and
Et2O. Slow cooling of the pentane solution to -20 °C gave 0.047
g of small X-ray quality wine-red crystals of [Mn(η5-C14H11)-
(CO)3] (0.148 mmol, 38%): 1H NMR (300 MHz, C6D5CD3, 257
K) δ 7.35 (d, J 8-9 ) 8.23 Hz, 1H, H8), 7.31 (s, 1H, H6), 7.17 (d,
J 11-10 ) 8.23 Hz, 1H, H11), 6.96 (dt, 2H, H9,10), 6.84 (s, 1H,
H
13), 6.06 (d, J 4-3 ) 5.6 Hz, 1H, H4), 4.96 (t, J 3-4/J 3-2 ) 6.4
Hz, 1H, H3), 3.32 (td, J 2-3/J 2-1endo ) 5.9 Hz, 1H, H2), 2.84 (dd,
J 1endo-1exo ) 15.4 Hz, J 1endo-2 ) 4.9 Hz, 1H, Hendo), 2.66 (d,
J 1exo-1endo ) 15.4 Hz, 1H, Hexo); 13C NMR (75 MHz, C6D5CD3,
257 K) δ 224 (s(br), 3C, CtO), 125-135 (8C, C6-13), 108 (s,
1C, C5), 106 (s, 1C, C14), 96.6 (d, J C4-H4 ) 168 Hz, 1C, C4),
74.9 (d, J C3-H3 ) 176 Hz, 1C, C3), 50.4 (d, J C2-H2 ) 166 Hz,
1C, C2), 30.0 (t, J C1-Hexo ∼ J C1-Hendo ) 130 Hz, 1C, C1); 1H-1H
NOESY NMR (300 MHz, C6D5CD3, 257 K) large correlation
for δ 7.31 and 6.06 (H6 and H4), small correlation for δ 6.84
and 6.06 (H13 and H4); mass spectrum (EI) m/z 318 (5.9), 290
(25.0), 262 (17.9), 234 (53.0), 179 (100), 152 (17.0), 97 (6.0), 83
(7.9), 69 (13.0), 55 (26.8). Anal. Calcd for C17H11MnO3: C,
64.17; H, 3.48. Found: C, 63.24; H, 3.58.
Rea ction of [P P N][Mn (η4-C14H10)(CO)3] w ith CF 3CO2D
a n d Isola tion of [Mn (η5-C14H10D)(CO)3]. To a slurry of 0.479
g (0.560 mmol) of [PPN][Mn(η4-C14H10)(CO)3] in 25 mL of
n-pentane was added 43.0 mL (0.558 mmol) of CF3CO2D. A

664 Organometallics, Vol. 19, No. 4, 2000
Veauthier et al.
Ta ble 1. Cr ysta l Da ta , Da ta Collection
P a r a m eter s, a n d Agr eem en t F a ctor s for
[P P N][Mn (η4-C14H10)(CO)3]
formula
C53H40MnNO3P2
fw
855.74
temp, K
223(2) K
wavelength, Å
cryst syst
space group
a, Å
0.71073
triclinic
P1h
12.384(7)
b, Å
14.147(9)
c, Å
15.085(9)
R, deg
117.07(5)
109.05(5)
95.22(5)
2132(2), 2
1.333
0.431
888
â, deg
γ, deg
V, Z, Å3
F(calcd), Mg/m3
µ, mm-1
F(000)
crystal size, mm
θ range, deg
limiting indices
0.21 × 0.27 × 0.32
1.81-24.00
-7 e h e 8, -16 e k e 16,
-17 e l e 16
5334
no. of reflns collected
no. of indep reflns
abs corr
F igu r e 1. ORTEP drawing of the anion in [PPN][Mn(η4-
4984 (Rint ) 0.0549)
C
14H10)(CO)3] (35% probability ellipsoids).
ψ scan
refinement method
no. of data/restraints/params
goodness of fit on F2
final R indices [I > 2σ(I)]
R indices (all data)
full-matrix least squares on F2
4984/0/541
This makes the isolation of [K(18-crown-6)]1 difficult,
and the [PPN]+ salt of 1- is preferred.
1.034
R1 ) 0.0425, wR2 ) 0.0903
R1 ) 0.0694, wR2 ) 0.1042
The availability of 1- as the PPN+ salt as well as the
crystallographically characterized [K(Kryptofix 222)]+
salt4 offered an opportunity to test the assumption that
the solid-state structures of η4-arene complexes of
[Mn(CO)3]- are primarily controlled by electronic factors
at the metal center and not by interactions with the
counterions or packing forces.
largst diff peak and hole, e Å-3 0.292 and -0.231
Ta ble 2. Cr ysta l Da ta , Da ta Collection
P a r a m eter s, a n d Agr eem en t F a ctor s for
[Mn (η5-C14H11)(CO)3]
formula
C17H11MnO3
fw
318.2
The molecular structure of [PPN]1 (Figure 1) was
established by an X-ray diffraction analysis as described
in the Experimental Section. The structure of the [Mn-
(η4-C14H10)(CO)3]- anion in [PPN]1 is shown in Figure
1. Selected bond lengths and angles are listed in Table
3. The most notable feature of [PPN]1 is the dihedral
angle of 37.8(2)° between the coordinated diene and the
exo naphthalene. This angle is much larger than the
previously reported dihedral angle of 17.3° for the η4-
anthracene ligand in [Ni(η4-C14H10)(η2-depe)]10 but is
similar to the dihedral angle of 31.5° for the η4-
temp, K
wavelength, Å
cryst syst
space group
a, Å
208(2) K
0.71073
monoclinic
P21/c
8.482(5)
12.915(11)
12.105(7)
90
90.30(4)
90
1326.1(15), 4
1.594
1.001
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Z, Å3
F(calcd), Mg/m3
µ, mm-1
F(000)
anthracene ligand in [Ti(η4-C14H10)(η2-C14H10)(η5-C5-
648
cryst size, mm
θ range, deg
limiting indices
0.28 × 0.24 × 0.18
11
Me5)]-
and compares well with the η4-naphthalene
2.31-24.05
ligand in [PPN][Mn(η4-C10H8)(CO)3]2b (dihedral angle of
37.1°) and the η4-anthracene ligands in [K(Kryptofix
222)]1 (34.9(2)°)4b and [Ti(η6-C14H10)(η4-C14H10)(η2-
dmpe)] (35.5°).11
0 e h e 9, 0 e k e 14,
-13 e l e 13
2230
no. of reflns collected
no. of indep reflns
abs corr
2083 (Rint ) 0.0344)
ψ scan
Str u ctu r a l Com p a r ison of [P P N][Mn (η4-C14H10)-
(CO)3] with [K(Kr yptofix 222)][Mn (η4-C14H10)(CO)3].
Figures 2 and 3 depict least-squares fits of the anion in
[K(Kryptofix 222)]14b to that of the anion in [PPN]1. The
solid line is the anion in [PPN]1, and the dashed line is
the fitted structure. In Figure 2, the fit is made between
all the non-hydrogen atoms of the anions in [K(Kryptofix
222)]1 and [PPN]1. The average deviation between
atomic positions within the structures is 0.214 Å. If the
carbonyl ligands are not included in the fit, then the
deviation between the two structures is only 0.102 Å
(Figure 3). We conclude that the structures of the anions
refinement method
no. of data/restraints/params
goodness of fit on F2
final R indices [I > 2σ(I)]
R indices (all data)
full-matrix least squares on F2
2083/0/235
1.008
R1 ) 0.0493, wR2 ) 0.1073
R1 ) 0.0845, wR2 ) 0.1264
largst diff peak and hole, e Å-3 0.562 and -0.354
are almost identical and that the nature of the counte-
rions has remarkably little effect on key structural
parameters, such as the folding of the anthracene
ligand.
Rea ction s of [Mn (η4-C14H10)(CO)3]- w ith Elec-
t r op h iles a n d Syn t h esis of [Mn (η5-C14H11)(CO)3].
Protonation of 1- was carried out at ambient temper-
atures by addition of 1.0 equiv of CF3CO2H to an orange-
yellow slurry of [PPN][Mn(η4-C14H10)(CO)3] in n-pentane
(Scheme 2). This resulted in a change to a wine-red
(10) Boese, R.; Stanger, A.; Stellberg, P.; Shazar, A. Angew. Chem.,
Int. Ed. Engl. 1993, 32, 1475.
(11) Seaburg, J . K.; Fischer, P. J .; Young, V. G.; Ellis, J . E. Angew.
Chem., Int. Ed. 1998, 37, 155.

Protonation of the η4-Anthracene Ligand
Organometallics, Vol. 19, No. 4, 2000 665
water stable crystals of [Mn(η5-C14H11)(CO)3] (2, 38%).
Fast recrystallization at -78 °C gave wine-red micro-
crystals of 2.
Ta ble 3. Selected Bon d Len gth s (Å) a n d An gles
(d eg) w ith in [P P N][Mn (η4-C14H10)(CO)3]
Mn-C(4)
Mn-C(5)
Mn-C(16)
Mn-C(17)
Mn-C(1)
Mn-C(2)
Mn-C(3)
O(3)-C(3)
2.073(4)
2.178(5)
2.183(4)
2.075(4)
1.780(5)
1.782(5)
1.767(7)
1.173(5)
C(4)-C(5)
C(16)-C(17)
C(4)-C(17)
C(5)-C(6)
C(15)-C(16)
C(6)-C(15)
O(1)-C(1)
O(2)-C(2)
1.435(6)
1.418(6)
1.401(6)
1.453(6)
1.464(6)
1.449(6)
1.168(5)
1.159(5)
The low yield in the protonation of [PPN]1 partially
reflects poor surface contact between [PPN]1 and CF3-
CO2H in pentane, and we have examined the protona-
tion reaction in THF, CH3CN, and CH2Cl2, all of which
are good solvents for [PPN]1. The protonation product
of [PPN]1 is unstable in THF and CH3CN, and attempts
to form 2 from solutions of [PPN][Mn(η4-C14H10)(CO)3]
in THF or CH3CN resulted in orange, pentane insoluble
complexes with IR spectra that indicated the presence
of unidentified neutral manganese carbonyl complexes.
The product of protonation of [K(18-crown-6)]1 was also
unstable in THF and CH3CN. The protonation of [PPN]-
1 to form 2 did occur in CH2Cl2 (as is indicated by the
observation of CtO stretching bands at 2010 (s), 1930
(s), and 1912 (s) cm-1 and a color change from orange
to wine red), but removal of CH2Cl2 resulted in the
decomposition of 2 into anthracene and unidentified
manganese complexes.
C(1)-Mn-C(2)
C(2)-Mn-C(3)
C(3)-Mn-C(1)
C(4)-C(5)-C(6)
98.2(2)
98.2(2)
91.4(2)
118.4(4)
Mn-C(1)-O(1)
Mn-C(2)-O(2)
Mn-C(3)-O(3)
C(17)-C(16)-C(15)
178.5(4)
177.4(4)
177.9(5)
119.0(4)
We next examined the protonation of K[Mn(η4-C14H10)-
(CO)3] in THF to determine if the presence of a
coordinating counterion affected the course of the reac-
tion. The addition of 1.0 equiv of CF3CO2H to a solution
of K[Mn(η4-C14H10)(CO)3] formed in situ in THF did give
2 (65% spectroscopic yield). The solvent was immedi-
ately removed and the powder redissolved in pentane.
Only microcrystalline 2 was obtained by this method,
however, and the presence of excess anthracene frus-
trated attempts to isolate pure 2 by this procedure. It
should be noted that complex 2 decomposes in THF after
30 min at ambient temperature into anthracene and
unidentified manganese complexes.
F igu r e 2. Least-squares fit for anions in [PPN][Mn(η4-
C14H10)(CO)3] and [K(Kryptofix 222)][Mn(η4-C14H10)(CO)3],
excluding the hydrogen atoms from the fitting program.
Protonation could also be achieved by addition of 1.0
equiv of HPF6 to a pentane slurry of [PPN]1 to give 2,
but this offered no advantages over the CF3CO2H
approach.
Attempts to add other electrophiles to 1- by reaction
of [PPN]1 or K1 with PhCH2Cl and CH3I were unsuc-
cessful, and no reaction was observed. Preliminary
reactions of [PPN]1 with CF3SO3CH3 indicate the
formation of a neutral manganese carbonyl complex, but
we have been unable to isolate a pure compound.
F igu r e 3. Least-squares fit for the anions in [PPN][Mn-
(η4-C14H10)(CO)3] and [K(Kryptofix 222)][Mn(η4-C14H10)-
(CO)3], excluding the hydrogen atoms and the CO ligands
from the fitting program.
Str u ctu r a l Ch a r a cter iza tion of [Mn (η5-C14H11)-
1
(CO)3]. The H NMR spectrum of 2 has some unusual
features (see below), and its molecular structure was
therefore determined by an X-ray diffraction analysis
(Figure 4). Selected bond lengths and angles for 2 are
listed in Table 4.
Sch em e 2
Complex 2 has a nearly planar dienyl fragment, with
the B and C aromatic rings attached at the first two
atoms of the dienyl ligand and bent slightly below the
plane (9.98(2)°). The cyclohexadienyl ring is folded about
C2-C14, with a dihedral angle of 33.4(3)°. This dihedral
angle is slightly smaller than the range of values
determined for other cyclohexadienyl manganese com-
plexes, ranging from 42° for [Mn(η5-C6H7)(CO)3],12 41°
for [Mn{(η5-C2H5O2C)CH-C6H6}(CO)3],13 40° for [Mn-
{η5-C6H3(OCH3)3CH2CN}(CO)3],14 39.6° for [Mn(η5-
PhC6H6)(CO)2NO][BF4],15 39.6° for [Mn{(C2H5O)2P(O)-
pentane solution containing a neutral manganese car-
bonyl complex (as indicated by the observation of CtO
stretching bands at 2020 (s), 1942 (s), and 1925 (s)
cm-1). The wine-red powder was recrystallized slowly
from n-pentane at -15 °C to give small, red, air and
(12) Churchill, M. R.; Scholer, F. R. Inorg. Chem. 1969, 8, 1950.
(13) Mawby, A. J . Organomet. Chem. 1973, 55, C39

666 Organometallics, Vol. 19, No. 4, 2000
Veauthier et al.
Ta ble 5. Com p a r ison of Ar om a tic C-C Bon d
Len gth s (Å) in [Mn (η5-C14H11)(CO)3] to Sim ila r
Or ga n om eta llic Com p lexes a n d An th r a cen e
[Mn(η5-C14H11)(CO)3]
1.389 ( 0.028
1.401 ( 0.014
1.399 ( 0.016
1.409 ( 0.032
1.409 ( 0.029
[Mn(η5-C6H7)(CO)3]12
[Mn{η5-C6H6-exo-(Me3SiCN2)}(CO)3]18
[Mn{η5-C8H6NH-exo-(CMe2CN)}(CO)3]20
32
C
14H10
the Mn-C(2,3,4) is ∼2.115(5) Å (∼2.137 Å for [Mn{η5-
C8H6NH-exo-(CMe2CN)}(CO)3]20), Mn-C5 is 2.296(5) Å,
and Mn-C14 is 2.485(5) Å (2.266 and 2.434 Å for [Mn-
{η5-C8H6NH-exo-(CMe2CN)}(CO)3]20).
The average C-C bond length for the aromatic ring
system in 2 is compared with similar figures for related
complexes in Table 5. The geometry of the “Mn(CO)3”
fragment in 2 is unexceptional.
F igu r e 4. ORTEP drawing of [Mn(η5-C14H11)(CO)3] (35%
probability ellipsoids).
Ta ble 4. Selected Bon d Len gth s (Å) a n d An gles
(d eg) w ith in [Mn (η5-C14H11)(CO)3]
1
Va r ia ble-Tem p er a tu r e H NMR Sp ectr a of [Mn -
1
(η5-C14H11)(CO3)]. The H NMR spectra of complex 2
C(1)-C(14)
C(1)-C(2)
C(2)-C(3)
C(3)-C(4)
C(4)-C(5)
C(5)-C(14)
C(5)-C(6)
C(14)-C(13)
O(1)-C(15)
O(2)-C(16)
1.503(7)
1.487(7)
1.378(8)
1.392(7)
1.433(6)
1.412(7)
1.385(6)
1.380(7)
1.133(6)
1.145(6)
Mn-C(2)
Mn-C(3)
Mn-C(4)
Mn-C(5)
Mn-C(14)
Mn-C(15)
Mn-C(16)
Mn-C(17)
O(3)-C(17)
Mn-cya
2.155(5)
2.095(5)
2.095(5)
2.296(5)
2.485(5)
1.790(5)
1.779(6)
1.728(6)
1.167(6)
1.757(5)
at various temperatures are shown in Figure 5. The
temperature-dependent changes in the 1H NMR spectra
of 2 are fully reversed when the sample is warmed. It
was not feasible to examine spectra at temperatures
much above ambient, because the compound is ther-
mally unstable. At room temperature (297 K), lines in
the spectrum are broad and provide minimal structural
information.
Mn-C(15)-O(1)
Mn-C(16)-O(2)
Mn-C(17)-O(3)
Mn-cy-C(15)a
Mn-cy-C(16)a
178.3(5)
178.1(5)
178.2(4)
120.8(2)
120.2(2)
C(17)-Mn-C(16)
C(17)-Mn-C(15)
C(16)-Mn-C(15)
Mn-cy-C(17)a
93.2(2)
88.1(2)
98.9(2)
127.7(2)
108.0(4)
The peaks in the 297 K spectrum at 6.10, 4.99, and
3.35 δ each integrate as one proton, while the group of
peaks centered at 2.76 δ integrates as two protons. The
remaining six protons resonate in the aromatic region
of the spectrum. As the sample is cooled to 287 K, the
group of peaks centered at 2.76 δ begins to sharpen and
a shoulder forms at 6.89 δ. At 277 K, two distinct
resonances can be observed at 2.68 and 2.83 δ, the peaks
between 6.50 and 3.00 δ all begin to show H-H
coupling, and the peaks in the aromatic region of the
spectrum begin to sharpen. Finally, at 257 K definite
splitting patterns that correspond to the assignments
listed in Table 6 are observed for each proton resonance
of [Mn(η5-C14H11)(CO)3]. The endo and exo protons were
assigned according to the Karplus relation:21 the torsion
angle between the Hexo and Hortho protons is 85.1°,
between the Hendo and Hortho protons the torsion angle
is 30.9°, and the angle that is closer to 90° will
correspond to a smaller coupling constant. In complex
2, the vicinal coupling constant observed for each
doublet in the peak centered at 2.84 δ is 4.9 Hz, but
vicinal coupling is not observable for the doublet at 2.66
δ. The resonance at 2.66 δ is therefore assigned to the
exo proton and the resonance at 2.84 δ is assigned to
the endo proton of [Mn(η5-C14H11)(CO)3], with a vicinal
coupling constant to the ortho hydrogen smaller than
the typical 6 Hz value1a as a consequence of the small
dihedral angle subtended by the methylene group.
The temperature-dependent broadening of the 1H
NMR spectrum was unexpected, in part because the
related cyclohexadienyl complexes [Mn(η5-C6H7)(CO)3]22
and [Mn(η5-C10H9)(CO)3]2b do not exhibit unusual 1H
NMR spectra at ambient temperatures. The origin of
the broadening was difficult to determine at first
because so many resonances were affected in such a
C(1)-C(2)-C(14)
a
cy ) centroid of the dienyl portion of the anthracenyl ligand.
(η5-C6H6)}(CO)3],16 38° for [Mn{η5-C6H6CH2C(NCH3Ph)d
W(CO)5}(CO)3],17 36.5° for [Mn(η5-PhC6H6)(CO)3],15 36.8°
for [Mn{η5-C6H6-exo-(Me3SiCN2)}(CO)3],18 36° for [{Mn-
(CO)3}2-{µ-(η5-C6H6CHCN-η5-C6H6)}],14 to 35° for [Mn-
{η5-C6H6CCH3ClCO2CH3)(CO)3].19 The small dihedral
angle in complex 2 most likely reflects the influence of
the fused π system attached at C14 and C5, and a
similar decrease in dihedral angle to 29.6° is observed
for [Mn{η5-C8H6NH-exo-(CMe2CN)}(CO)3],20 which also
contains a fused aromatic ligand. The torsion angles
between the endo and exo C-H bonds on C1 and the
vicinal C-H bond on the dienyl ligand in 2 are 85.1°
for H2-C2-C1-Hexo and 30.9° for H2-C2-C1-Hendo
;
the latter is unusually large for a cyclohexadienyl
complex of “Mn(CO)3”, as required by the small folding
of the cyclohexadienyl ring.
An additional consequence of the presence of the fused
π system is the lengthening of the manganese to carbon
bonds at the fused carbons. For example, in complex 2
(14) Rose, E.; Le Corre-Susanne, C.; Rose-Munch, F.; Renard, C.;
Gagliardini, V.; Teldji, F.; Vaisserman, J . Eur. J . Inorg. Chem. 1999,
421.
(15) Ittel, S. D.; Whitney, J . F.; Chung, Y. K.; Williard, P. G.;
Sweigart, D. A. Organometallics 1988, 7, 1323.
(16) Lee, T.; Yu, H. B.; Chung, Y. K.; Hallows, W. A.; Sweigart, D.
A. Inorg. Chim. Acta 1994, 224, 147.
(17) Rose-Munch, F.; Le Corre-Susanne, C.; Balssa, F.; Rose, E.;
Vaisserman, J .; Licandro, E.; Papagni, A.; Maiorana, S.; Meng, W.;
Stephenson, G. R. J . Organomet. Chem. 1997, 545-546, 9.
(18) Reau, R.; Reed, R. W.; Dahan, F.; Bertrand, G. Organometallics
1993, 12, 1501.
(19) Balssa, F.; Gagliardini, V.; Le Corre-Susanne, C.; Rose-Munch,
F.; Rose, E.; Vaisserman, J . Bull. Soc. Chim. Fr. 1997, 134, 537.
(20) Ryan, W. J .; Peterson, P. E.; Cao, Y.; Williard, P. G.; Sweigart,
D. A.; Baer, C. D.; Thompson, C. F.; Chung, Y. K.; Chung, T. M. Inorg.
Chim. Acta 1993, 211, 1.
(21) Karplus, M. J . Chem. Phys. 1959, 30, 11.
(22) Lamanna, W.; Brookhart, M. J . Am. Chem. Soc. 1980, 102,
3490.

Protonation of the η4-Anthracene Ligand
Organometallics, Vol. 19, No. 4, 2000 667
1
F igu r e 5. Variable-temperature H NMR spectra of [Mn(η5-C14H11)(CO)3] (300 MHz, C6D5CD3).
Ta ble 6. 1H NMR Sp ectr u m Assign m en ts (p p m ) for
narrow band irradiation of specified nuclei (H*). The
convention for subtraction was [(H* at x ppm spectrum)
- (H* at 0.00 ppm spectrum)].
[Mn (η5-C14H11)(CO)3]
When the peak at 6.06 δ (or 2.66 δ) is irradiated, the
difference spectrum contains two negative peaks, one
very large at 6.06 δ (or 2.66 δ) and a small negative
peak at 2.66 δ (or 6.06 δ), which implies that magneti-
zation is transferred between the exo proton and the
proton para to the methylene group in the cyclohexa-
dienyl ligand. Irradiation applied at 2.84 δ causes only
the peak at 2.84 δ to invert after subtraction of the
spectra, establishing that there is no exchange between
the endo proton and other ring protons. Irradiation at
4.96 δ (or 3.32 δ) gives an additional negative peak at
3.32 δ (or 4.96 δ), and irradiation at 7.17 δ (or 7.31 δ)
yields a negative peak at 7.31 δ (or 7.17 δ).
These exchange spectra suggest that the exo proton
is exchanging with the proton in the para orientation.
This would account for all of the observed magnetization
transfers, and a low barrier to such an exchange would
be consistent with a metal-mediated intermediate
(Scheme 3) in which the proton of the trifluoroacetic acid
adds first to the metal to give the metal hydride
intermediate A and then moves up to the ring into one
of two symmetrically equivalent positions. What the
NMR data suggest is that this migration is reversible.
The magnetization transfer data incidentally confirm
the previous assignment of the endo 1H resonance based
on the Karplus relationship, and the exchange sequence
agrees with all of the assignments made for 1H NMR
spectra of complex 2.
Hexo
Hendo
H2
H3
H4
H13
H9,10
H11
H6
2.66
2.84
3.32
4.96
6.06
6.84
6.96
7.17
7.31
7.35
(d, J exo-endo ) 15.4 Hz)
(dd, J endo-exo ) 15.4 Hz, J endo-2 ) 4.9 Hz)
(t, J 2-endo ∼ J 2-3 ) 5.9 Hz)
(t, J 3-2 ∼ J 3-4 ) 6.4 Hz, 1H)
(d, J 4-3 ) 5.6 Hz, 1H)
(s, 1H)
(dt, 2H)
(d, J 11-10 ) 8.23 Hz, 1H)
(s. 1H)
H8
(d, J 8-9 ) 8.23 Hz, 1H)
uniform manner. To explore the origin of the line
broadening, magnetization transfer experiments were
conducted.
Ma gn etiza tion Tr a n sfer a n d NOE Differ en ce
Spectr a of [Mn (η5-C10H9)(CO)3].23 NOE (nuclear Over-
hauser enhancement) difference spectra of 2 were
recorded at 257 K. The pulse sequence for NOE experi-
ments involves irradiating one resonance of the spec-
trum and then applying a π/2 observation pulse to
record the FID. If exchange occurs with the irradiated
proton, magnetization will be transferred from the
irradiated proton to one or more other protons in the
system and will be observed through a reduction in
signal strength wherever exchange occurs.
Qu a n t it a t ive An a lysis of F lu xion a l P r ocess in
[Mn (η5-C14H11)(CO)3]. Simulations of the dynamic 1H
NMR spectra,24 in terms of the proposed mechanism,
were carried out in order to determine the rate of
exchange between the two proton environments and
provide insight into the underlying dynamic processes.
Figure 6 shows the difference spectra obtained by
subtracting NMR spectra recorded with and without
(23) (a) Sanders, J . K.; Hunter, B. K. In Modern NMR Spectros-
copy: A Guide for Chemists; Oxford University Press: New York, 1990,
Chapters 1-3, 6, and 17. (b) Noggle, J . H.; Schirmer, R. E. In The
Nuclear Overhauser Effect; Academic Press: New York, 1971, Chapters
1 and 7. (c) Forsen, S.; Hoffman, R. A. J . Chem. Phys. 1963, 39, 2892.
(24) The DNMR3 program (http://www.arl.hpc.mil/PET/cta/ccm/
software/qcpe/QCPE2/index.html) was used to simulate the 1H NMR
spectra: (a) Kleier, D. A.; Binsch, G. QCPE 1970, 11, 165. (b) Binsch,
G.; Kessler, H. Angew. Chem., Int. Ed. Engl. 1980, 19, 411.

668 Organometallics, Vol. 19, No. 4, 2000
Veauthier et al.
F igu r e 6. NOE difference spectra of [Mn(η5-C14H11)(CO)3] at 257 K (300 MHz, C6D5CD3). The difference spectra are labeled
with letters that correspond to the peak being irradiated. In spectrum g*, the peak at 6.84 is inverted because of the
breadth of the irradiation at 7.35 δ.
Sch em e 3
Ta ble 7. Ra tes of Exch a n ge for 1,4-Hyd r id e Sh ift
Spectra were acquired between 257 and 297 K at 5 K
in [Mn (η5-C14H11)(CO)3] Ca lcu la ted fr om Lin e
intervals, and line shapes for the region between 6.5 δ
and 2.5 δ were simulated for each spectrum. Attempts
Sh a p e An a lysis
k (s-1
“outside”
)
k (s-1
“inside”
)
k (s-1
“mean”
)
∆G
(kcal/mol)
to carry out a least-squares fit of simulated to experi-
mental line widths for all peaks in the spectrum did not
converge, but convergent fits were obtained when the
resonances were treated in two groups, one comprised
of the “outside” (exo, endo, and para) proton resonances,
and the other comprised of the “inside” (ortho and meta)
proton resonances. Figures 7 and 8 show stacked plots
of the simulated and calculated spectra for the two types
of fits. Because each proton in 2 is uniformly affected
by the hydride migration reaction (Scheme 3), the line
shape data obtained for each group can be considered
collectively. The simulated exchange rates at each
temperature, the averaged (“mean”) exchange rates at
each temperature, and the corresponding activation
energies for the 1,4-hydride migration between the two
equivalent sites on the anthracene ligand are sum-
marized in Table 7. The Eyring plot25 of the data shown
in Figure 9 gives rise to an enthalpy of activation for
T (K)
257
262
267
272
277
282
287
292
297
4.0
6.5
4.0
5.5
4.0
6.0
14.3
14.3
14.4
14.3
14.4
14.5
14.5
14.5
14.6
10.5
13.5
20.0
29.0
50.0
82.5
108.0
8.50
20.5
28.0
40.0
50.0
80.0
125.0
9.50
17.0
24.0
34.5
50.0
81.25
116.25
the migration of 12.3 ( 0.2 kcal/mol and a small
negative entropy of activation of -7.7 ( 0.8 eu.
Deu ter iu m La belin g a n d Kin etic Isotop e Effect
on F lu xion a lity w ith in [Mn (η5-C14H10D)(CO)3]. The
reaction of [PPN][Mn(η4-C14H10)(CO)3] with 1.0 equiv of
(26) Failure to always incorporate a full equivalent of deuterium
into 3 (1H NMR confirms 90% incorporation) prevented us from
performing line shape analysis on 3. Early 1H NMR samples of 3 show
full incorporation of deuterium into the endo position and allowed us
to estimate rates based upon the experimental line widths. In no case
have we seen incorporation of deuterium into the exo position. Similar
observations have been reported for [Mn(η5-C6H7(CO)3] and [Mn(η5-
(25) (a) Sandstrom, J . In Dynamic NMR Spectroscopy; Academic
Press: London, 1982; Chapter 7. (b) Glasstone, S.; Laidler, K.; Eyring,
H. In The Theory of Rate Processes; McGraw-Hill: New York, 1941; p
195. (c) Eyring, H. Chem. Rev. 1935, 17, 65.
C
10H9(CO)3]; see ref 2b.

Protonation of the η4-Anthracene Ligand
Organometallics, Vol. 19, No. 4, 2000 669
1
F igu r e 7. Variable-temperature H NMR spectra of [Mn-
(η5-C14H11)(CO)3] (300 MHz, C6D5CD3) with calculated line
shapes of the ortho and meta (inside) proton resonances,
observed vs calculated (*).
F igu r e 9. Eyring plot for the 1,4-hydride shift in [Mn(η5-
C
14H11)(CO)3] in C6D5CD3. The errors reported in the
activation parameters are based on the method of least
squares and are calculated from the standard deviations
in the slope and y-intercept.
the protonated complex implies that there is consider-
able bond breaking involving the C-H(D) bond in the
transition state for the exchange reaction and not much
compensating bond making, and this inference is con-
sistent with the transition state shown in the proposed
mechanism (see below, Scheme 5).
Discu ssion
We have shown that protonation of the terminal (η4-
anthracene) manganese tricarbonyl anion [Mn(η4-C14H10)-
(CO)3]- (1-) occurs at the end ring to give the (η5-
anthracenyl) manganese tricarbonyl complex [Mn(η5-
C14H11)(CO)3] (2), and this metal-mediated addition of
the corresponding anthracene therefore has a regiose-
lectivity complementary to that observed in conven-
tional organic reductions of anthracene. Single-crystal
X-ray diffraction analysis confirms the addition of the
proton to the hydrocarbon ring, and labeling studies
establish that protonation is regiospecific and occurs at
the endo site.
Dynamic 1H NMR studies and magnetization transfer
experiments establish that 2 is fluxional and that an
overall 1,4-hydride shift reaction equilibrates the two
sides of the anthracenyl ligand on the NMR time scale.
The data that we have been able to acquire delineate
the process to the level of detail shown in Scheme 3,
but this does not fully describe the mechanism of the
reaction: the classic study by Lamanna and Brookhart22
on the thermal isomerization of 6-exo-1H-[Mn(η5-C6-
1
F igu r e 8. Variable-temperature H NMR spectra of [Mn-
(η5-C14H11)(CO)3] with calculated line shapes of the endo
and exo (outside) proton resonances, observed vs calculated
(*).
CF3CO2D proceeds quantitatively to give [Mn(η5-
C14H10D)(CO)3] (3, 42%) under the same reaction condi-
tions reported for the formation of 2 (see above).
Complex 3 was recrystallized slowly from n-pentane at
-15 °C to give small red air stable crystals. Deuteration
occurs primarily in the endo position of the η5-bound
ring.26
The 1H NMR spectrum of 3 exhibits temperature-
dependent broadening similar to that observed in the
spectra of 2, but when spectra of 3 are compared with
spectra of 2 as the temperature is decreased, the 1H
NMR resonances of 3 sharpen more quickly than the
resonances of 2. It is clear that complex 3 participates
in an exchange sequence similar to that established for
complex 2 (Scheme 3), but at a slower rate. We could
not obtain convergent simulations of spectra of complex
3,26 but a kinetic isotope effect (kH/kD) of 1.88 could be
estimated on the basis of the line widths of the experi-
mental spectra.27 The observation that the exchange
sequence is slower for the deuterated complex than for
(27) (a) Thompson, R. L.; Lee, S.; Geib, S. J .; Cooper, N. J . Inorg.
Chem. 1993, 32, 6067. (b) Binsch, G. Top. Stereochem. 1968, 3, 97. (c)
Binsch, G. In Dynamic Nuclear Magnetic Resonance Spectroscopy;
J ackman, L. M., Cotton, F. A., Eds.; Academic Press: New York, 1975;
Chapter 3.

670 Organometallics, Vol. 19, No. 4, 2000
Veauthier et al.
Sch em e 4
migration within 6-exo-1H-[Mn(η5-C6HD5)(CO)3] (4). This
adds two additional steps to the mechanism in Scheme
3: an η5 to η3 hapticity shift in the dienyl ligand that
precedes the carbon to manganese hydride shift, and
1,2 shifts of the metal around the polyene ligand in the
intermediate η4-diene complex 6a . These steps are
invoked to explain why the intramolecular carbon to
manganese hydride shift can lead to isomerization, since
the observed isomerization requires that the hydride
return to a different carbon atom when it moves back
to the metal center. This proposal is supported by its
ability to provide a comprehensive framework for the
interpretation not only of isomerization within 4 but also
of hydride migration within phenyl- and methyl-
substituted analogues of 4 as first established and
analyzed by Pauson.28
Sch em e 5
Fluxionality within 2 poses problems similar to those
addressed by Lamanna and Brookhart, since there must
be a process by which the hydride in intermediate A
returns from the metal to a different carbon atom in
the hydrocarbon ligand without violating microscopic
reversibility. We propose, however, an interpretation
(Scheme 5) that differs in two important respects from
Scheme 4:
(1) The η5 to η3 shift in Scheme 4 is less attractive
today in light of our understanding of how agostic and
other σ donation interactions maintain 18-electron
configurations in low-valent environments;29 we there-
fore propose a migration step in which increased inter-
action of the hydride with the metal center is synchro-
(28) (a) Munro, G. A. M.; Pauson, P. L. J . Chem. Soc., Chem.
Commun. 1976, 134. (b) Pauson, P. L.; Segal, J . A. J . Chem. Soc.,
Dalton Trans. 1975, 1677. (c) Pauson, P. L.; Segal, J . A. J . Chem. Soc.,
Dalton Trans. 1975, 1683.
(29) (a) Crabtree, R. H. Angew. Chem., Int. Ed. Engl. 1993, 32, 789.
(b) Brookhart, M.; Green, M. L. H.; Wong, L.-L. Prog. Inorg. Chem.
1988, 36, 1. (c) Brookhart, M.; Green, M. L. H. J . Organomet. Chem.
1983, 250, 395.
HD5)(CO)3] both facilitates and requires a more detailed
discussion of the reaction pathway.
The Lamanna and Brookhart study proposes the
mechanism shown in Scheme 422 for hydride (deuteride)

Protonation of the η4-Anthracene Ligand
Organometallics, Vol. 19, No. 4, 2000 671
nous with deligation of the carbons carrying the aromatic
substituent.
of 42° for [Mn(η5-C6H7)(CO)3]12 and would place the endo
hydrogen closer to the metal in the anthracenyl complex
and hence facilitate the migration.
(2) We do not believe it is necessary to invoke a 1,2-
manganese shift step. Return of the hydride to carbon
“y” instead of degenerate return to carbon “x” can be
explained in terms of fluxionality within the four-legged
piano stool and averaging of the carbonyl ligands and a
hydride in A; with the sterically undemanding hydride
ligand, such an isomerization could have a low energy
barrier. Alternatively, the hydride in the intermediate
A could be positioned symmetrically between carbon “x”
and carbon “y” as shown in Scheme 5, and return to
carbon “x” or carbon “y” would then occur with equal
probability. Such a symmetric positioning of the hydride
in the intermediate could contribute to the low height
of the energy barrier observed for isomerization of 2.
The data to hand (or readily accessible) do not provide
a definitive basis for choosing between Scheme 4 and
Scheme 5, and a mechanism analogous to that in
Scheme 4 is a feasible interpretation on the assumption
that k1 . k2 and that there is a thermodynamic
preference for the anthracenyl isomer in which the
aromatic substitution is in the 2,3 (or the degenerate
5,6) position.
This interpretation is, however, difficult to reconcile
with our understanding of the free energy reaction
coordinates in these systems, and the small 14.5 kcal/
mol isomerization barrier in 2 argues that Scheme 5
offers a more satisfying interpretative framework. It is
reasonable, for example, that the aromatic substituent
on the dienyl ligand in 2 could stabilize the transition
state for the synchronous hydride shift shown in Scheme
5 and that it could lower the free energy difference
between 2 and A below the 14-19 kcal/mol energy
difference suggested for 4 and 6a by Lamanna and
Brookhart;22 additionally the 33.3° dihedral angle for
the folding of the methylene group in the dienyl frag-
ment in [Mn(η5-C14H11)(CO)3] is smaller than the value
It is much more difficult to interpret the marked
difference between the barriers to isomerization ob-
served for 4 and 2 within the context of Scheme 4. The
kinetic isotope effect that we have observed in the
isomerization of 2 establishes that the rate-determining
step must involve C-H bond cleavage, and the 1,2-
manganese shift would then have to have a lower
activation barrier than this step. The 1,2-manganese
shift barrier is, however, a combination of two free
energy differences: the base ∆G between 2 and inter-
mediate A, argued by Lamanna and Brookhart to be
about 14-19 kcal/mol in the parent cyclohexadienyl
system,22 and the activation barrier to the 1,2-manga-
nese polyene shift itself, established by Whitesides30 and
Maitlis,31 as reviewed by Lamanna and Brookhart,22 to
be ca. 14 kcal/mol. It is difficult to see how the aromatic
substituent in 2 could so dramatically lower both of
these components of the free energy difference that their
sum could become significantly less than the observed
14.5 kcal/mol barrier to the rate-determining C-H bond
cleavage step.
We conclude that we prefer the analysis of the
fluxional process proposed in Scheme 5; this offers a
pathway close to the least motion pathway for the
process since it makes movement of the hydrogen atom,
by far the lightest atom in the system, the central
feature of each step in the isomerization. The discovery
that the barrier to isomerization in 2 is dramatically
less than the barrier to isomerization established for 4
may also require minor modification of the interpreta-
tion previously proposed for the isomerization of 4.
Ack n ow led gm en t. This work was financially sup-
ported by the NSF through grant numbers CHE 9632202
(N.J .C.) and CHE 9615116 (G.F.).
Su p p or tin g In for m a tion Ava ila ble: Complete X-ray
crystallographic data for [PPN]1 and 2 and complete dynamic
1H NMR spectra for the line shape analysis of complex 2 are
available at http://www.pubs.acs.org.
(30) Whitesides, T. H.; Budnik, R. A. Inorg. Chem. 1976, 15, 874.
(31) Kang, J . W.; Childs, R. F.; Maitlis, P. M. J . Am. Chem. Soc.
1970, 92, 720.
(32) (a) Mathieson, A.; Robertson, J . M.; Sinclair, V. C. Acta
Crystallogr. 1950, 3, 245. (b) Mason, R. Acta Crystallogr. 1964, 17,
547.
OM990735G