Photosubstitution reactions of M(CO)4(1,10-phenanthroline) (M = Mo, W). Influence of entering ligand, irradiation wavelength, and pressure

Article abstract of DOI:10.1021/om960531p

The influence of the entering nucleophile, irradiation wavelength, and pressure on the quantum yield for the photosubstitution of M(CO)₄phen (M = Mo, W) to produce M(CO)₃-(L)phen (L = PMe₃, PPh₃) was investigated in toluene at 298 K. From the pressure dependence of the quantum yield apparent volumes of activation could be determined as a function of irradiation wavelength. These could be analyzed in terms of contributions arising from dissociative ligand field excitation and associative metal-to-ligand charge transfer excitation. The influence of steric hindrance on the entering ligand (PPh₃ > PEt₃ > PMe₃) controls the contribution of the associative charge-transfer photosubstitution reaction. An overall mechanistic picture is presented and discussed in reference to available literature data.

Full text of DOI:10.1021/om960531p

572
Organometallics 1997, 16, 572-578
P h otosu bstitu tion Rea ction s of
M(CO)4(1,10-p h en a n th r olin e) (M ) Mo, W). In flu en ce of
En ter in g Liga n d , Ir r a d ia tion Wa velen gth , a n d P r essu r e
Wen-Fu Fu† and Rudi van Eldik*
Institute for Inorganic Chemistry, University of Erlangen-Nu¨rnberg, Egerlandstrasse 1,
91058 Erlangen, Germany
Received J uly 2, 1996X
The influence of the entering nucleophile, irradiation wavelength, and pressure on the
quantum yield for the photosubstitution of M(CO)4phen (M ) Mo, W) to produce M(CO)3-
(L)phen (L ) PMe3, PPh3) was investigated in toluene at 298 K. From the pressure
dependence of the quantum yield apparent volumes of activation could be determined as a
function of irradiation wavelength. These could be analyzed in terms of contributions arising
from dissociative ligand field excitation and associative metal-to-ligand charge transfer
excitation. The influence of steric hindrance on the entering ligand (PPh3 > PEt3 > PMe3)
controls the contribution of the associative charge-transfer photosubstitution reaction. An
overall mechanistic picture is presented and discussed in reference to available literature
data.
In tr od u ction
and PPh3.11 It could be shown that LF excitation
resulted in a dissociative (D) substitution mechanism,
whereas MLCT excitation resulted in a dissociative
interchange (Id) substitution mechanism. Steric hin-
drance on the entering nucleophile (PR3) had no influ-
ence on the nature of the substitution mechanism. The
latter was only controlled by the excitation wavelength.
We have now undertaken a detailed study of the
photosubstitution behavior of Mo(CO)4phen and W-
(CO)4phen as a function of irradiation wavelength for
PMe3 and PPh3 as entering ligands. The nucleophile
concentration and pressure dependences of these reac-
tions clearly demonstrate the sensitivity of the underly-
ing substitution mechanism to the size of the central
metal atom and the entering ligand, as controlled by
the population of LF and MLCT excited states.
q
The apparent volume of activation, ∆Vφ , obtained
from the pressure dependence of the quantum yield for
a photochemical reaction has become an interesting
parameter for the elucidation of the intimate mecha-
nism of photoinduced reactions of inorganic and orga-
nometallic complexes.1-7 In an earlier study on the
photosubstitution reactions of M(CO)4phen (M ) Mo,
W; phen ) 1,10-phenanthroline) with PEt3, the values
q
of ∆Vφ pointed to a dissociative mechanism for ligand
field (LF) excitation as compared to an associative
mechanism for metal-to-ligand charge-transfer (MLCT)
excitation.2,8 These measurements were performed
under limiting wavelength conditions, during which
either LF or MLCT photochemistry was observed. Little
is known about the fine tuning of such processes as
controlled by the irradiation wavelength, the nature and
concentration of the entering nucleophile (PR3), the
central metal atom (Cr, Mo, W), and the applied
pressure. A systematic variation of these variables
should enable a control over the nature of the photo-
substitution mechanism. In this respect the pressure
variable can provide crucial information on the intimate
nature of the photochemical mechanism,1-8 as has been
Exp er im en ta l Section
M(CO)4phen (M ) Mo, W) were synthesized photochemically
under an argon atmosphere as previously described12-17 and
recrystallized from isooctane/dichloromethane.18 Their UV-
vis absorption spectra were found to be in good agreement with
those cited in the literature.8,19 M(CO)6 (M ) Mo, W) and 1.0
M PMe3 solution in toluene were purchased from Aldrich and
used without further purification. Triphenylphosphine (PPh3,
Merck) was recrystallized twice from ethanol. Elemental
analyses were performed by Beller Analytical Laboratory
(Go¨ttingen, Germany). Anal. Found for PPh3: C, 82.17; H,
6.19; P, 11,74. Found for Mo(CO)4phen: C, 49.51; H, 2.23; N,
6.59. Found for W(CO)4phen: C, 40.01; H, 1.83; N, 5.87.
the case for many related thermal processes.9,10
similar detailed study was recently performed on the
photosubstitution reaction of Cr(CO)4phen with PMe3
A
† On leave from the Department of Chemistry, Yunnan Normal
University, Kunming, 650092, People’s Republic of China.
X Abstract published in Advance ACS Abstracts, J anuary 1, 1997.
(1) Ford, P. C.; Crane, D. R. Coord. Chem. Rev. 1991, 111, 153.
(2) Wieland, S.; van Eldik, R. Coord. Chem. Rev. 1990, 97, 155.
(3) Stochel, G.; van Eldik, R. Coord. Chem. Rev., in press.
(4) van Eldik, R. Pure Appl. Chem. 1993, 65, 2603.
(5) Ryba, D. W.; van Eldik, R.; Ford, P. C. Organometallics 1993,
12, 104.
(6) Rossenaar, B. D.; van der Graaf, T.; van Eldik, R.; Langford, C.
H.; Stufkens, D. J .; Vlcek, A., J r. Inorg. Chem. 1994, 33, 2865.
(7) Friesen, D. A.; Lee, S, H.; Nashiem, R. E.; Mezyk, S. P.; Waltz,
W. L. Inorg. Chem. 1995, 34, 4026.
(10) Hubbard, C. D.; van Eldik, R. In Chemistry under Extreme or
Non-Classical Conditions; van Eldik, R., Hubbard, C. D., Eds.; Wiley/
Spektrum: New York/Heidelberg, 1997; Chapter 2.
(11) Fu, W. F.; van Eldik, R. Inorg. Chim. Acta, in press.
(12) Bock, H.; tom Dieck, H. Angew. Chem. 1966, 78, 549.
(13) Bock, H.; tom Dieck, H. Chem. Ber. 1967, 100, 228.
(14) Brunner, H.; Herrmann, W. A. Chem. Ber. 1972, 105, 770.
(15) Schadt, M. J .; Lees, A. J . Inorg. Chem. 1986, 25, 672.
(16) Marx, D. E.; Lees, A. J . Inorg. Chem. 1987, 26, 620.
(17) Manuta, D. M.; Lees, A. J . Inorg. Chem. 1983, 22, 3825.
(18) Vichova, J ; Hartl, F.; Vlcek, A., J r. J . Am. Chem. Soc. 1992,
114, 10903.
(8) Wieland, S.; Bal Reddy, K.; van Eldik, R. Organometallics 1990,
9, 1802.
(9) van Eldik, R.; Merbach, A. E. Comments Inorg. Chem. 1992, 12,
341.
(19) Manuta, D. M.; Lees, A. J . Inorg. Chem. 1986, 25, 1354.
S0276-7333(96)00531-6 CCC: $14.00 © 1997 American Chemical Society

Photosubstitution of M(CO)4(1,10-phenanthroline)
Organometallics, Vol. 16, No. 4, 1997 573
F igu r e 2. UV-vis absorption spectra recorded during the
irradiation of 1.9 × 10-4 M W(CO)4phen and 0.3 M PMe3
in toluene at 546 nm and 298 K. The dot-dash line
represents the absorption spectrum of W(CO)3(PMe3)phen.
F igu r e 1. UV-vis absorption spectra recorded during the
irradiation of 2.6 × 10-4 M Mo(CO)4phen and 0.3 M PMe3
in toluene at 546 nm and 298 K. The dot-dash line
represents the absorption spectrum of Mo(CO)3(PMe3)phen.
Toluene (Aldrich spectroscopic grade) was distilled over sodium
under an Ar atmosphere and saturated with high-purity Ar
prior to use.
Photochemical measurements were performed in a thermo-
stated ((0.1 °C) high-pressure cell20 that was equipped with
two windows and a calibrated photodiode mounted on the rear
window, situated inside a black metal compartment and placed
on top of a magnetic stirrer on an optical rail. Test solutions
were transferred under an Ar atmosphere into a “pillbox”
optical cell21 and stirred with a Teflon-coated magnetic bar,
using a combination of Schlenk and double-tip-needle tech-
niques and a specially designed filling system.22 The pillbox
was then placed into the high-pressure cell, which was filled
with n-heptane as the pressure-transmitting medium.
Monochromatic light at 313, 336, 366, 403, 436, 486, and
546 nm was selected from a high-pressure mercury lamp
(Osram HBO 100/2) using Oriel interference filters. It was
necessary to stir the solution effectively during the irradiation.
The incident light intensity was determined by ferrioxalate
actinometry23-25 for irradiation at 313-436 nm, whereas
actinochrome N 475/610 (Amko)26,27 was used for irradiation
at 486 and 546 nm. In all the photolysis experiments the
concentration of reactants and products was monitored by
UV-visible spectroscopy using a Cary 1 spectrophotometer.
The photosubstitution quantum yields were corrected for inner
filter effects.
F igu r e 3. UV-vis absorption spectra recorded during the
irradiation of 1.8 × 10-4 M W(CO)4phen and 0.3 M PMe3
in toluene at 313 nm and 298 K.
indicates that the reaction does not seem to be compli-
cated by subsequent substitution processes. This is in
agreement with the tendency of M(CO)4phen complexes
to undergo photosubstitution according to reaction 1.8
M(CO)4phen + PR3
9
hν8
fac-M(CO)3(PR3)phen + CO (1)
M ) Mo, W; R ) Me, Ph
Similar spectral changes were observed during the
photolysis of W(CO)4phen in the presence of PMe3 (see
Figure 2), and the clean isosbestic points once again
demonstrate that the formation of W(CO)3(PMe3)phen
is not affected by subsequent reactions or the apparent
buildup of any intermediate species.
In the case of photolysis in the LF band at λ < 350
nm (see ref 8 and literature cited therein), the primary
formation of M(CO)3(PR3)phen is followed by the sub-
sequent formation of M(CO)2(PR3)2phen according to
reaction 2 (see Figures 3 and 4), for which a second
Resu lts a n d Discu ssion
Sp ectr a l Ch a r a cter iza tion . Absorption spectra
were recorded during the photoinduced substitution
reaction of M(CO)4phen (M ) Mo, W) with PMe3 and
PPh3. Figure 1 illustrates a typical observed spectral
sequence following irradiation (546 nm) of Mo(CO)4phen
in a toluene solution containing the PMe3 nucleophile.
The characteristic MLCT bands (see ref 8 and literature
cited therein) of both Mo(CO)4phen and Mo(CO)3(PMe3)-
phen appear at 495 and 616 nm, respectively, and the
presence of clear isosbestic points at 383 and 532 nm
M(CO)3(PR3)phen + PR3
cis-M(CO)2(PR3)2phen + CO (2)
9
hν8
(20) Fleischmann, F. K.; Conze, E. G.; Kelm, H.; Stranks, D. R. Rev.
Sci. Instrum. 1974, 45, 1427.
M ) Mo, W; R ) Me, Ph
(21) le Noble, W. J .; Schlott, R. Rev. Sci. Instrum. 1976, 47, 770.
(22) Wieland, S.; van Eldik, R. Rev. Sci. Instrum. 1989, 60, 955.
(23) Lee, J .; Seliger, H. H. J . Chem. Phys. 1964, 40, 519.
(24) Bowman, W. D.; Demas, J . N. J . Phys. Chem. 1976, 80, 2434.
(25) Hatchard, C. G.; Parker, C. A. Proc. R. Soc. London, Ser, A 1956,
235, 518.
(26) Brauer, H. D.; Schmidt, R. J . Photochem. Photobiol. 1983, 37,
587.
(27) Schmidt, R.; Drews, W.; Brauer, H. D. J . Am Chem. Soc. 1980,
102, 2791.
series of clean isosbestic points are observed. Figure 4
illustrates the steric effect of the entering ligand on the
subsequent photoreaction at 366 nm photolysis. For-
mation of Mo(CO)2(PPh3)2phen seems to occur easily in
a subsequent reaction, due to the labilization of CO by
PPh3 on Mo(CO)3(PPh3)phen during LF excitation,

574 Organometallics, Vol. 16, No. 4, 1997
Fu and van Eldik
F igu r e 4. UV-vis absorption spectra recorded during the
irradiation of 2.8 × 10-4 M Mo(CO)4phen and 0.2 M PPh3
in toluene at 366 nm and 298 K.
Ta ble 1. Qu a n tu m Yield for Rea ction 1 a s a
F u n ction of [P Me3] a n d Ir r a d ia tion Wa velen gth a t
Am bien t P r essu r e
λirr ) 366 nm
λirr ) 436 nm
[PMe3],
M
φd × 102
φp × 102
φd × 103
φp × 103
Mo(CO)4phena
0.0010 3.83 ( 0.05 3.09 ( 0.12
0.0015 3.79 ( 0.06 3.29 ( 0.05
0.0020 3.73 ( 0.01 3.41 ( 0.17
0.0030 3.63 ( 0.01 3.62 ( 0.02
0.0050 4.03 ( 0.03 3.89 ( 0.04
0.0060 4.16 ( 0.08 3.93 ( 0.05
1.92 ( 0.01
2.57 ( 0.15
1.34 ( 0.11
2.06 ( 0.03
3.67 ( 0.15
5.10 ( 0.00
3.23 ( 0.02
4.90 ( 0.19
9.30 ( 0.15
0.010
0.020
0.030
4.19 ( 0.24 4.08 ( 0.02
4.60 ( 0.13 4.74 ( 0.13 10.15 ( 0.12
5.41 ( 0.36 5.48 ( 0.34 14.39 ( 1.00 14.37 ( 0.48
W(CO)4phenb
0.0015 1.57 ( 0.10 0.77 ( 0.02
0.0030 1.51 ( 0.06 0.91 ( 0.06
0.0060 1.42 ( 0.10 1.05 ( 0.05
0.54 ( 0.03
0.73 ( 0.03
1.09 ( 0.08
1.64 ( 0.07
2.84 ( 0.06
4.09 ( 0.03
0.21 ( 0.05
0.43 ( 0.03
0.79 ( 0.04
1.37 ( 0.03
2.95 ( 0.04
4.42 ( 0.09
F igu r e 5. φtot as a function of [PMe3] and irradiation
wavelength for the photosubstitution of Mo(CO)4phen at
298 K.
0.010
0.020
0.030
1.39 ( 0.05 1.20 ( 0.03
1.42 ( 0.01 1.47 ( 0.01
1.55 ( 0.02 1.52 ( 0.02
Ta ble 2. Otot for Rea ction 1 a s a F u n ction of [P Me3]
a n d Ir r a d ia tion Wa velen gth for M ) Wa a t 298 K
a
[Mo(CO)4phen] ) 2.48 × 10-4 M; φd, disappearance of
Mo(CO)4phen; φp, formation of Mo(CO)3(PMe3)phen. b [W(CO)4phen]
) 2.42 × 10-4 M; φd, disappearance of W(CO)4phen; φp, formation
of W(CO)3(PMe3)phen.
φ × 102
[PMe3],
M
λirr
)
λirr
)
λirr
)
λirr
)
366 nm
436 nm
486 nm
546 nm
0.010
0.020
0.050
0.10
0.15
0.20
1.20 ( 0.03 0.14 ( 0.03 0.23 ( 0.02 0.19 ( 0.01
1.47 ( 0.01 0.30 ( 0.04
which is very similar to that found for the photosubsti-
tution reaction of Cr(CO)4phen.11 During this photoly-
sis, the color of the solution clearly changes from red to
violet and then to green.28,29 This reaction can be
prevented by using shorter irradiation times during LF
excitation. Due to the significantly lower quantum yield
and the different nature of the photosubstitution mech-
anism in the case of MLCT excitation (see below),
significant interference of the subsequent reaction (2)
is not observed under such conditions.
2.23 ( 0.12 0.79 ( 0.03 0.94 ( 0.05 0.94 ( 0.04
2.98 ( 0.19 1.42 ( 0.01 1.86 ( 0.03 1.89 ( 0.07
3.59 ( 0.05 2.18 ( 0.01
4.04 ( 0.08 2.77 ( 0.05 3.40 ( 0.15 3.14 ( 0.04
4.34 ( 0.38 4.39 ( 0.04
0.30
φLF × 102 1.27 ( 0.15 0.04 ( 0.03 0.26 ( 0.18 0.23 ( 0.13
b × 10
1.49 ( 0.13 1.39 ( 0.03 1.43 ( 0.10 1.42 ( 0.08
a
Conditions: [W(CO)4phen] ) 2.42 × 10-4 M, λirr ) 366 nm,
λirr ) 436 nm; [W(CO)4phen] ) 2.20 × 10-4 M, λirr ) 486 nm, λirr
) 546 nm.
Con cen tr a tion Dep en d en ce of Qu a n tu m Yield .
The quantum yields for the LF and MLCT photosub-
stitution reactions, on the basis of available literature
data,8,19,30 are expected to exhibit characteristic concen-
tration dependences. The overall quantum yield for
reaction 1, φtot ) φLF + φMLCT, was therefore measured
as a function of [PMe3] and irradiation wavelength.
Preliminary measurements, summarized in Table 1,
indicated that at low [PMe3] the quantum yield for the
disappearence of M(CO)4phen was higher than that for
the formation of M(CO)3(PMe3)phen. This was ascribed
to a small contribution of a decomposition process in
the absence of excess PMe3. Plots of φtot (based on the
formation of M(CO)3(PMe3)phen) as a function of [PMe3]
and irradiation wavelength are summarized for M ) Mo
in Figure 5, whereas the corresponding data for M ) W
are summarized in Table 2. The plots clearly show the
deviation from linear behavior at low [PMe3] due to the
participation of side reactions under such conditions as
discussed above. The data show that φtot increases with
increasing [PMe3], and the contribution of the intercept
increases with increasing excitation energy, i.e. decreas-
ing irradiation wavelength. These trends are typical
(28) Houk, L. W.; Dobson, G. R. Inorg. Chem. 1966, 5, 2119.
(29) Houk, L. W.; Dobson, G. R. J . Chem. Soc. A 1966, 317.
(30) van Dijk, H. K.; Servaas, P. C.; Stufkens, D. J .; Oskam, A. Inorg.
Chim. Acta 1985, 104, 179.

Photosubstitution of M(CO)4(1,10-phenanthroline)
Organometallics, Vol. 16, No. 4, 1997 575
for parallel reaction pathways, viz. a concentration-
independent LF and concentration-dependent MLCT
reaction, and can be considered as an important piece
of information regarding the nature of the photosubsti-
tution mechanism. This observation is of fundamental
importance in the interpretation of the wavelength
dependence of photosubstitution reactions usually mea-
sured at a fixed ligand concentration. The difference
in the concentration dependences of LF and MLCT
photosubstitution reactions can lead to an unexpected
increase in quantum yield on increasing the excitation
wavelength due to a concentration-independent process
going to a concentration-dependent process and can lead
to a wrong interpretation in terms of excited-state
effects only.31 The competition between a concentration-
independent and a concentration-dependent pathway
requires that wavelength dependence should be studied
as a function of entering ligand concentration.
The dependence of φtot on [PR3] can be expressed by
eq (3), where φLF represents the intercept and b the
slope of the plots. The values of φLF and b strongly
φtot ) φLF + φMLCT
) φLF + b[PR3]
(3)
depend on the irradiation wavelength, as shown in
Table 2. Similar experiments were performed for the
bulkier PPh3 ligand. Typical plots of φtot versus [PPh3]
as a function of irradiation wavelength are shown in
Figure 6, which clearly demonstrate that φ decreases
with increasing irradiation wavelength, in good agree-
ment with results reported before.2,8,11 The ligand
concentration independence of φ is typical for dissocia-
tive LF photochemistry. Even for the excitation at 436
nm for M ) Mo no meaningful dependence of φ on
[PPh3] was observed.19 This indicates that MLCT
excitation presumably also results in a dissociative
substitution process, in contrast to that reported above
for the significantly smaller entering nucleophile PMe3.
In the case of M ) W, the plot of φtot versus [PPh3]
shown in Figure 6b clearly indicates the contribution
of an associative reaction path during MLCT excitation
at 546 nm.31 In an effort to obtain further mechanistic
insight, the pressure dependence of these reactions was
studied at different excitation wavelengths.
F igu r e 6. φtot as a function of [PPh3] and irradiation
wavelength for the photosubstitution of M(CO)4phen at 298
K: (a, top) M ) Mo; (b, bottom) M ) W.
P r essu r e Dep en d en ce of Qu a n tu m Yield . In
order to investigate the effect of pressure on this
photosubstitution reaction, it is essential to separate the
pressure dependence of φLF and φMLCT. φLF makes a
significant contribution at short irradiation wavelengths
such that the pressure dependence of φtot must be
studied as a function of [PR3] in order to separate the
F igu r e 7. φtot as a function of [PMe3] and pressure for the
photosubstitution of Mo(CO)4phen at 336 nm and 298 K.
contribution of φLF and φMLCT
. A typical example is
) W. These data clearly demonstrate that, for the
photosubstitution of CO by PMe3, LF excitation strongly
favors a dissociative reaction path whereas MLCT
excitation favors an associative reaction path.
shown in Figure 7, from which it follows that the
intercept decreases and the slope increases with in-
creasing pressure. The values of φLF and b (according
to eq 3) are summarized as a function of pressure in
Table 3. The results clearly show that φLF is character-
ized by a strongly positive apparent volume of activa-
tion, whereas b is characterized by significantly negative
volumes of activation. In the case of M ) Mo, the latter
value is even more negative than those observed for M
This mechanistic changeover can also be seen in an
overall manner when the pressure dependence of these
substitution reactions is studied at various irradiation
wavelengths and [PMe3], for which some typical data
are reported in Table 4. The typical plots of ln φ versus
pressure in Figure 8 show how the pressure dependence
of φ changes as a function of irradiation wavelength.
According to Table 4 the pressure dependence of φtot
(31) Lindsay, E.; Vlcek, A., J r.; Langford, C. H. Inorg. Chem. 1993,
32, 2269.

576 Organometallics, Vol. 16, No. 4, 1997
Fu and van Eldik
Ta ble 3. Su m m a r y of OLF a n d b a s a F u n ction of Ir r a d ia tion Wa velen gth , Meta l, a n d P r essu r e for th e
Rea ction
M(CO)4phen + PMe3 9
hν8 M(CO)3(PMe3)phen + CO
λirr ) 336 nm
λirr ) 366 nm
pressure, MPa
0.1
50
100
150
φLF × 102
b × 10
φLF × 102
b × 10
Mo(CO)4phena
5.12 ( 0.65
8.82 ( 1.07
11.64 ( 0.35
14.02 ( 0.81
5.26 ( 0.42
3.33 ( 0.70
2.48 ( 0.23
2.06 ( 0.53
4.08 ( 0.92
2.99 ( 0.83
3.16 ( 1.28
1.93 ( 1.47
5.34 ( 1.40
8.38 ( 1.26
9.47 ( 1.94
13.53 ( 2.24
∆Vq, (cm3 mol-1
)
+15.4 ( 2.1
-16.4 ( 2.9
+10.9 ( 3.5
-14.4 ( 2.1
W(CO)4phenb
1.85 ( 0.27
2.07 ( 0.21
2.46 ( 0.14
2.85 ( 0.02
0.1
50
100
150
1.18 ( 0.18
1.00 ( 0.14
0.80 ( 0.09
0.61 ( 0.01
4.17 ( 7.07
3.06 ( 12.5
2.62 ( 7.99
1.95 ( 6.61
1.40 ( 0.11
1.92 ( 0.19
2.29 ( 0.12
2.65 ( 0.10
∆Vq, (cm3 mol-1
)
+10.9 ( 0.8
-7.3 ( 0.4
+12.1 ( 1.0
-10.4 ( 1.4
a
b
Conditions: [M] ) 2.48 × 10-4 M; solvent toluene; T ) 298 K. Conditions: [M] ) 2.42 × 10-4 M; solvent toluene; T ) 298 K.
Ta ble 4. Otot (×102) a s a F u n ction of [P Me3], Ir r a d ia tion Wa velen gth , a n d P r essu r e for P h otosu bstitu tion of
M(CO)4p h en Com p lexes in Tolu en e a t 298 K
pressure, MPa
[M] × 104,
[PMe3],
M
∆Vq
,
φ
λirr, nm
M
0.1
50
100
150
cm3 mol-1
Mo(CO)4phen
336
336
336
366
366
366
2.48
2.48
2.48
2.48
2.48
2.48
0.02
0.05
0.10
0.02
0.05
0.10
6.47 ( 0.19
7.52 ( 0.27
10.49 ( 0.27
4.74 ( 0.13
7.39 ( 0.32
9.17 ( 0.28
5.40 ( 0.16
4.91 ( 0.04
8.14 ( 0.15
14.18 ( 0.49
4.49 ( 0.10
8.79 ( 0.29
12.29 ( 0.78
5.10 ( 0.16
8.70 ( 0.05
16.22 ( 0.27
3.99 ( 0.17
9.73 ( 0.24
15.08 ( 1.66
+4.0 ( 1.7
-2.8 ( 1.1
-7.2 ( 0.2
+2.4 ( 1.1
-4.7 ( 0.5
-7.9 ( 0.7
7.24 ( 0.02
12.33 ( 0.91
4.30 ( 0.04
7.76 ( 0.38
11.15 ( 0.11
W(CO)4phen
366
366
366
436
436
436
2.42
2.42
2.42
2.42
2.42
2.42
0.02
0.05
0.10
0.02
0.05
0.10
1.47 ( 0.01
2.23 ( 0.12
2.98 ( 0.19
0.29 ( 0.02
0.79 ( 0.03
1.42 ( 0.00
1.35 ( 0.02
2.13 ( 0.06
3.03 ( 0.04
0.36 ( 0.01
1.08 ( 0.05
1.92 ( 0.02
1.25 ( 0.08
2.09 ( 0.10
3.23 ( 0.01
0.45 ( 0.04
1.23 ( 0.05
2.30 ( 0.10
1.19 ( 0.11
2.03 ( 0.12
3.47 ( 0.01
0.52 ( 0.02
1.39 ( 0.08
2.65 ( 0.10
+3.5 ( 0.3
+1.5 ( 0.2
-2.6 ( 0.5
-9.8 ( 0.6
-9.1 ( 1.6
-10.2 ( 1.3
Ta ble 5. Volu m es of Activa tion for
P h otosu bstitu tion of CO in M(CO)4(p h en )
Com p lexes by th e Liga n d P R3 a t Differ en t
Excita tion Wa velen gth sa
∆Vqφ, cm3 mol-1
Mo(CO)4phen
W(CO)4phen
PMe3 PPh3
+3.2 ( 0.6
λirr
,
nm
PMe3
PPh3
313 +14.4 ( 3.1
336
366
-2.8 ( 1.1
-4.7 ( 0.5
+7.6 ( 1.3
+6.7 ( 1.5
+1.5 ( 0.2 +10.9 ( 0.6
-5.8 ( 0.8 +13.9 ( 1.5
403 -12.3 ( 0.1
436 -13.3 ( 0.9 +11.0 ( 1.6
486 -13.3 ( 3.2
546 -10.7 ( 2.5
-9.1 ( 1.6
-10.3 ( 2.5
-9.6 ( 1.6
+8.1 ( 1.5
+3.4 ( 1.6b
-8.2 ( 0.8c
Experimental conditions: [M] ) (1.90-2.48) × 10-4 M; [PMe3]
) 0.0506 M, [PPh3] ) 0.100 M; solvent toluene; T ) 298 K. b [PPh3]
) 0.320 M. c [PPh3] ) 0.305 M.
a
F igu r e 8. Plots of ln φtot versus pressure as a function of
irradiation wavelength for the reaction
φMLCT) in Table 3 characterize the dissociative and
associative nature of the photosubstitution process,
respectively.
W(CO)4phen + PMe3 9
hν8 W(CO)3(PMe3)phen + CO
Experimental conditions: [W(CO)4phen] ) 2.1 × 10-4
M; [PMe3] ) 0.05 M; T ) 298 K; toluene solvent.
In the case of PPh3 as entering nucleophile, typical
plots of ln φ versus pressure are shown in Figure 9 and
the corresponding apparent volumes of activation are
summarized in Table 5. It follows from these data that
substitution of CO by PPh3 on Mo(CO)4phen follows a
dissociatively activated process at all irradiation wave-
lengths. The trend in ∆Vφq to smaller values at 546 nm
irradiation for M ) Mo suggests a possible changover
exhibits an overall trend from slightly positive to
significantly negative volumes of activation on increas-
ing both the irradiation wavelength and the concentra-
tion of PMe3. This is a logical consequence of the [PMe3]
dependence of φtot as expressed by eq 3. It follows that
the reported ∆Vq values for φLF and b (representing

Photosubstitution of M(CO)4(1,10-phenanthroline)
Organometallics, Vol. 16, No. 4, 1997 577
Sch em e 1
determined from the effect of pressure on the overall
photosubstitution quantum yield. Since quantum yield
is a composite quantity, its pressure dependence may
also be influenced by the effect of pressure on various
photophysical processes that contribute to its value.1-3
For that reason it is necessary to analyze the above
interpretation in more detail. Photoexcitation of M-
(CO)4phen leads to the population of LF and MLCT
excited states, depending on the irradiation wave-
length.8,30,31 Detailed studies have resulted in different
energy diagrams that represent the relative location of
singlet and triplet states, as well as possible intersystem
crossings (isc).18,31,32 An overall summary of possible
activation and deactivation routes is presented in
Scheme 1. It is generally known that radiative deac-
tivation (kr) is slow, since these complexes exhibit almost
no emission.19,33 Decay of the excited state can occur
via nonradiative deactivation (kn) and photochemical
F igu r e 9. Plots of ln φtot versus pressure as a function of
irradiation wavelength for the reaction
M(CO)4phen + PPh3 9
hν8 M(CO)3(PPh3)phen + CO
LF
conversion (kp and kpMLCT), and photochemistry is
suggested to occur from the spin-singlet states. We
could clearly demonstrate in our earlier work that the
LF and MLCT excited states exhibit their own charac-
teristic photosubstitution mechanism.2,8 In general LF
excitation results in the activation of the M-CO bonds
and leads to dissociative ligand substitution reactions.
In the case of MLCT excitation, the nature of the excited
state will determine the nature of the ligand substitu-
tion mechanism. Partial delocalization of the electron
density from the π*(diimine) orbital, occupied in the
MLCT state, to the σ*(M-Cax) orbital may occur via
vibronic mixing.12 This will in turn result in the
dissociation of the axial CO ligand. Alternatively,
delocalization on the phen ligand will cause a decrease
in the electron density on the metal center (formally
then a 17-electron species) and an increase in electro-
philicity. The latter will induce an associative attack
of a nucleophile on the metal center and result in a
nucleophile concentration dependence of φ. Thus, both
dissociative and associative reaction modes are feasible
in the case of pure MLCT excitation. According to
Scheme 1, a back-population from MLCT to LF is not
ruled out and will be included in the observed concen-
(a, top) M ) Mo; (b, bottom) M ) W. Experimental
conditions: [Mo(CO)4phen] ) 1.80 × 10-4 M; [W(CO)4-
phen] ) 1.98 × 10-4 M; λirr ) 366, 403 and 436 nm,
[PPh3] ) 0.100 M; λirr ) 546 nm, M ) Mo, [PPh3] )
0.320 M; λirr ) 546 nm, M ) W, [PPh3] ) 0.305 M; T )
298 K; toluene solvent.
to an interchange (Id) photosubstitution mechanism
during MLCT excitation. Although the MLCT state
obviously favors an associative reaction path for the
smaller ligands (PMe3 and PEt38), the bulkier PPh3
nucleophile forces the systems to follow an Id reaction
route. In a similar way, MLCT excitation also induces
an Id mechanism for the reaction of Cr(CO)4phen with
the smaller PMe3 nucleophile, which is due to the
smaller metal center in the chromium complex. For the
photosubstitution of CO by PPh3 on W(CO)4phen, the
results in Table 5 and Figure 9b clearly show a
changeover from a dissociative process at low-wave-
length irradiation to an associative process at high-
wavelength irradiation. A comparison of the data for
M ) Mo and W indicates a significant difference in these
systems during MLCT excitation at 546 nm. The W
complex does exhibit an associative substitution mode
under these conditions, which is in line with the [PPh3]
dependence reported in Figure 6b. Thus, steric conges-
tion as determined by the size of the central metal and
the entering nucleophile controls the intimate nature
of the photosubstitution mechanism.
tration-independent φLF
.
In general, the observed quantum yield for reaction
1 will be the sum of the contributions resulting from
(32) Vlcek, A., J r.; Vichova, J .; Hartl, F. Coord. Chem. Rev. 1994,
132, 167.
(33) Wrighton, M. S.; Morse, D. L. J . Organomet. Chem. 1975, 97,
405.
Over a ll Com p a r ison . The interpretation outlined
above is based on the apparent volumes of activation

578 Organometallics, Vol. 16, No. 4, 1997
Fu and van Eldik
Sch em e 2
LF and MLCT excitation. Expressions for the wave-
length-dependent φLF and φMLCT in terms of [L], kn, and
kp have been dealt with in the past.8 Crucial for the
MLCT
present study is the dependence of φp
on the
concentration and nucleophilicity of the entering ligand
L. In the case of relatively low quantum yields, kn
.
kp and eq 4 applies. This is in good agreement with the
φtot ) φLF + φMLCT
MLCT
) (kpLF/knLF) + (kpMLCT[L]/kn
) (4)
empirically found relationship in (3). The dependence
of φMLCT on [L] will depend on the role of the associative
reaction pathway. On the assumption that nonradiative
deactivation is not very pressure sensitive, which has
been shown to be the case in systems where it was
possible to measure the effect of pressure on the lifetime
of the excited state,1-3,34,35 the pressure dependence of
φLF and φMLCT will represent that of the photochemical
caused by the PPh3 nucleophile can prevent an associa-
tive ligand substitution process. In fact a cone angle of
145° seems to be critical in controlling the nature of
addition of nucleophiles to a metal center, as recently
demonstrated for addition to the H2Os3(CO)10 cluster.37
The steric crowding can also account for the operation
of a dissociative interchange mechanism during the
displacement of CO by PPh3 on MLCT excitation, where
bond cleavage (Mo-CO) will be the controlling factor
(Scheme 2). On the basis of these results we cannot
distinguish whether the dissociative nature arises from
the MLCT state itself or from the thermal promotion of
MLCT to LF states. In the case of MLCT excitation of
W(CO)4phen, the quantum yield does increase with
increasing [PPh3] and the corresponding volume of
activation also supports the operation of an associative
mechanism. Thus, there is a significant difference
between Mo(CO)4phen and W(CO)4phen in their ability
to bind PPh3 during MLCT excitation. This trend can
only be related to steric effects associated with the metal
complex in the excited state and the entering nucleo-
phile.
The results of this study have clearly demonstrated
that the intimate nature of the photosubstitution mech-
anism can be controlled by the nature of the metal
center, the nature and concentration of the entering
nucleophile, the irradiation wavelength, and the applied
pressure. It is in particular the entering nucleophile
concentration dependence and the associated pressure
dependence that reveal the intimate nature of the
photosubstitution mechanism. The variables mentioned
enable a systematic “tuning” of the selected photosub-
stitution route.
LF
rate constants kp and kpMLCT, respectively. In the case
of a dissociative mechanism operating for both LF and
MLCT excitation, this will result in a concentration-
independent φtot
. Alternatively, if MLCT excitation
results in an associative ligand substitution mechanism,
φtot will exhibit a linear dependence on [L] as found in
the present study for L ) PMe3. It is important to note
that in considering the contributions of φLF and φMLCT
in eq 4, the MLCT state can be considered as a bound
state but the LF state cannot, which leads to a wave-
length-dependent contribution from φLF . On the other
hand, as indicated above, it is the contribution from
φMLCT that will depend on the ligand concentration, such
that φtot can exhibit rather complicated or unexpected
wavelength dependences at a fixed ligand concentra-
tion.31 It follows that the wavelength dependence of φtot
should always be studied as a function of ligand
concentration in such systems.
In the case of the W complex, irradiation at 366 nm
(Figure 5) also exhibits a meaningful ligand concentra-
tion dependence and factor b (Table 2) remains practi-
cally constant on increasing the irradiation wavelength
to 546 nm. This can be considered as clear evidence
for the nonradiative population of the 3MLCT state
(Scheme 1), from which the associative substitution
reaction occurs. Thus, the associative photosubstitution
reactions occur from the lowest, presumably triplet
MLCT state as suggested in the literature.31 The data
in Table 2 furthermore demonstrate that φLF reduces
to practically zero within the experimental error limits
at higher irradiation wavelength where only the MLCT
state is populated.
Ack n ow led gm en t. We gratefully acknowledge fi-
nancial support from the Deutsche Forschungsgemein-
schaft and Fonds der Chemischen Industrie. We ap-
preciate many fruitful discussions with Prof. A. Vlcek,
J r.
No dependence of φtot on [L] was found for excitation
in the region from LF to MLCT in the case of L ) PPh3
and M ) Mo. The cone angles of PMe3 and PPh3 are
118 and 145°, respectively,36 such that steric crowding
OM960531P
(34) Skibsted, L. H.; Weber, W.; van Eldik, R.; Kelm, H.; Ford, P.
C. Inorg. Chem. 1983, 22, 541.
(35) Kim, H. B.; Hiraga, T.; Uchida, T.; Kitamara, N.; Tazuke, S.
Cood. Chem. Rev. 1990, 97, 81.
(36) Tolman, C. A. Chem. Rev. 1977, 77, 313.
(37) Neubrand, A.; Poe¨, A. J .; van Eldik, R. Organometallics 1995,
14, 3249.