Kinetics and mechanism for the metalation reaction of 5,10,15,20-tetrakis(pentafluorophenyl)porphyrin with diaquabis(1,1,1,5,5,5-hexafluoropentane-2,4-dionato)nickel(II) in supercritical carbon dioxide

Article abstract of DOI:10.1246/bcsj.76.999

The nickel(II) ion incorporation reaction into 5,10,15,20-tetrakis(pentafluorophenyl)porphyrin (H₂tpfpp) to form the [Ni(tpfpp)] complex has been kinetically investigated using diaquabis(1,1,1,5,5,5-hexafluoropentane-2,4dionato)nickel(II), [Ni(hfac)₂(H₂O)₂], in supercritical carbon dioxide (scCO₂) medium in a spectrophotometric cell at various temperatures and pressures. The metalation reaction is first order with respect to porphyrin under the conditions where [Ni(hfac)₂(H₂O)₂] is in a large excess relative to H₂tpfpp. The saturation dependence was observed for the conditional first-order rate constants as a function of the mole fraction (X_Ni) of the excess [Ni(hfac)₂(H₂O)₂]. Such dependence suggests that the metalation reaction consists of a fast pre-equilibrium of the formation of the outer-sphere association between [Ni(hfac)₂(H₂O)₂] and H₂tpfpp and a first-order rate-determining nickel(II) ion incorporation into the porphyrin core. The conditional first-order rate constants k_obs are well expressed by k_obs = Kkx_Ni/(1 + KX_Ni), where K is the equilibrium constant for the formation of the outer-sphere association and k is the first-order rate constant for the nickel(II) ion incorporation. The thermodynamic and kinetic parameters are obtained as follows: ΔH° = 14.4 ± 1.3 kJ mol^-1 at 20.5 MPa, ΔAS° = 133 ± 4 J mol^-1 K^-1 at 20.5 MPa and ΔV° = 15 ± 1 cm³ mol^-1 at 333.3 K for the pre-equilibrium; ΔH^≠ = 110 ± 5 kJ mol^-1 at 20.5 MPa, ΔS^≠ = 22 ± 16 J mol^-1 K^-1 at 20.5 MPa, and ΔV≠=81 ± 8 cm³ mol^-1 at 333.3 K for the rate-determining step. The positive change in enthalpy and volume and significant positive change in entropy strongly indicate that the outer-sphere association is derived from the desolvation of reactants in scCO₂.

Full text of DOI:10.1246/bcsj.76.999

ꢀ 2003 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 76, 999–1005 (2003)
999
Kinetics and Mechanism for the Metalation Reaction of 5,10,15,20-Tet-
rakis(pentafluorophenyl)porphyrin with Diaquabis(1,1,1,5,5,5-hexa-
fluoropentane-2,4-dionato)nickel(b) in Supercritical Carbon Dioxide
#
Shijun Liu, Yasuhiro Inada, and Shigenobu Funahashi
ꢀ
Laboratory of Analytical Chemistry, Graduate School of Science, Nagoya University, Chikusa, Nagoya 464-8602
(Received November 15, 2002)
The nickel(b) ion incorporation reaction into 5,10,15,20-tetrakis(pentafluorophenyl)porphyrin (H2tpfpp) to form
the [Ni(tpfpp)] complex has been kinetically investigated using diaquabis(1,1,1,5,5,5-hexafluoropentane-2,4-
dionato)nickel(b), [Ni(hfac)2(H2O)2], in supercritical carbon dioxide (scCO2) medium in a spectrophotometric cell at
various temperatures and pressures. The metalation reaction is first order with respect to porphyrin under the conditions
where [Ni(hfac)2(H2O)2] is in a large excess relative to H2tpfpp. The saturation dependence was observed for the con-
ditional first-order rate constants as a function of the mole fraction (xNi) of the excess [Ni(hfac)2(H2O)2]. Such depen-
dence suggests that the metalation reaction consists of a fast pre-equilibrium of the formation of the outer-sphere asso-
ciation between [Ni(hfac)2(H2O)2] and H2tpfpp and a first-order rate-determining nickel(b) ion incorporation into the
porphyrin core. The conditional first-order rate constants kobs are well expressed by kobs ¼ KkxNi=ð1 þ KxNiÞ, where
K is the equilibrium constant for the formation of the outer-sphere association and k is the first-order rate constant
for the nickel(b) ion incorporation. The thermodynamic and kinetic parameters are obtained as follows:
ꢀHꢁ ¼ 14:4 ꢂ 1:3 kJ molꢃ1 at 20.5 MPa, ꢀSꢁ ¼ 133 ꢂ 4 J molꢃ1 Kꢃ1 at 20.5 MPa and ꢀVꢁ ¼ 15 ꢂ 1 cm3 molꢃ1 at
¼j
¼j
333.3 K for the pre-equilibrium; ꢀH ¼ 110 ꢂ 5 kJ molꢃ1 at 20.5 MPa, ꢀS ¼ 22 ꢂ 16 J molꢃ1 Kꢃ1 at 20.5 MPa,
and ꢀV ¼ 81 ꢂ 8 cm3 molꢃ1 at 333.3 K for the rate-determining step. The positive change in enthalpy and volume
¼j
and significant positive change in entropy strongly indicate that the outer-sphere association is derived from the desol-
vation of reactants in scCO2.
Supercritical fluids (SCFs) have been used as a solvent for
liquid chromatography and extraction in the field of analytical
chemistry, as one of the new reaction media for the preparation
of new chemical compounds in the chemical industry, and as a
unique solvent with a wide-ranging variation in density for the
kinetic study in the field of physical chemistry.1{10 The pro-
mising applications of SCFs have been successfully developed
in various fields, such as biotechnology11 and material
processing.12 Among the supercritical fluids, attention has
particularly focused on supercritical carbon dioxide (scCO2)
since it is inert, nontoxic, environmentally acceptable, inex-
pensive, nonflammable, noncorrosive and has a low critical
has been extensively studied for many kinds of metal ions in
a variety of solvents in order to clarify the metalation mechan-
ism of porphyrins.19{29 However, there is no report about the
metalation mechanism of porphyrins in supercritical fluids ex-
cept for the preliminary results of our study.30
On the other hand, in 1960, Fleischer and Wang31 first pro-
posed the so-called sitting-atop (SAT) complex of the proto-
porphyrin dimethyl ester in chloroform on the basis of visible
spectra, infrared spectra, and composition of the SAT
complex. After this report, the reports on the detection of
the SAT complex have been very limited, occurring only in pe-
culiar systems. We think that a crucial point to form the SAT
complex is to guarantee the thermodynamic and kinetic retar-
dation of the dissociation of protons on the pyrrole rings in the
SAT complex. Therefore, we must select solvents with a very
low Brꢁnsted basicity. Acetonitrile has a very low Brꢁnsted
basicity and is a coordinating and aprotic solvent with the sig-
nificantly high dielectric constant of 36, which is favorable for
coordination ability and solubility. In fact, we succeeded in
the detection of the sitting-atop (SAT) copper(b) complex of
TPP (5,10,15,20-tetraphenylporphyrin) in acetonitrile, where
two pyrrolenine nitrogen atoms in the Cu(b)-SAT complex co-
ordinate to the metal ion and two protons on the pyrrole nitro-
gen still remain.29 The structure parameters around the
copper(b) ion in the Cu(b)-SAT complex, determined by a
fluorescent XAFS method, have shown an axially elongated
and equatorially distorted six-coordinate geometry.32 We
ꢁ
temperature (31.1 C) and a moderate critical pressure (7.28
MPa).
Supercritical fluid extraction (SFE) using an extractant-
scCO2 mixture has been recognized to be promising as an ad-
vanced method for the separation of metals from a liquid for
the purpose of analytical pretreatment or hydrometal-
lurgy.13;14 Therefore, quantitative investigation concerning
the complexation and ligand-substitution reactions of metal
ions in scCO2 must be important.
The metalloporphyrin formation reaction is one of the im-
portant processes from both analytical and bioinorganic points
of view.15{18 The kinetics of the metalloporphyrin formation
# On leave from the Department of Chemistry, Xiangtan Univer-
sity, China

1000 Bull. Chem. Soc. Jpn., 76, No. 5 (2003)
Metalation of Porphyrin in Supercritical CO2
cell was controlled to within ꢂ0:2 MPa for at least 24 h, and
the temperature of the cell was constant within ꢂ0:2 K.
measured the rates of the formation reaction of the SAT com-
plexes for a series of transition metal(b) ions in acetonitrile,
using the stopped-flow technique, and we proposed the me-
chanism where there is a rapid deformation equilibrium of
the porphyrin ring prior to the rate-determining step of the
bond rupture of a coordinated solvent molecule on the
metal(b) ion.32 Furthermore, we measured the rates of the de-
protonation reaction of the Cu(b)-SAT complex by some
Brꢁnsted bases and indicated that the rate-determining step
is the attack of the base to the proton on the pyrrole nitrogen
in the SAT complex.33 Finally the unified mechanism relevant
to the porphyrin metalation mechanism has been proposed.34;35
The solubility of metal complexes and porphyrins in scCO2
is usually very low. However, the fluorinated metal ꢀ-diketo-
nates and fluorinated porphyrins possessed excellent solubility
in scCO2.36{39 In this context, in order to clarify the metala-
tion mechanism of porphyrins in scCO2 as a specific solvent,
we have investigated the kinetics for the Ni(b) ion incorpora-
tion reaction into 5,10,15,20-tetrakis(pentafluorophenyl)por-
phyrin (H2tpfpp) with diaquabis(1,1,1,5,5,5-hexafluoropen-
tane-2,4-dionato)nickel(b), [Ni(hfac)2(H2O)2], in scCO2. In
addition, the effect of water added as a cosolvent has been ex-
amined.
Sampling was carried out as follows. A certain amount of
H2tpfpp solution in hexane was first introduced into the cell. After
complete vaporization of the hexane by heating the cell at 338 K,
the desired amount of [Ni(hfac)] nH2O powder was then intro-
ꢄ
duced, and the cell was heated at 343 K for several tens of
minutes. In order to replace the air inside the system, 0.1 MPa
CO2 gas was introduced into the cell after closing the outlet stop
valve, and then was slowly released by opening the outlet valve.
This procedure was done three times. After a given temperature
was established, the cell was filled with CO2 to the desired pres-
sures, the contents of the cell were stirred by a Teflon-coated steel
bar, and then the absorption spectra of the reaction system were
recorded as a function of reaction time.
Results and Discussion
Characterization of [Ni(hfac) ] nH O. The water con-
tent was unknown for the commercially available reagent
2 Á
[Ni(hfac)2] nH2O. A TGA of the commercially available re-
2
ꢄ
agent and recrystallized reagent showed three weight loss steps
at about 353 K, 423 K, and 473 K, respectively. The first one
started at about 343 K and ended at about 363 K with a 3.5%
weight loss for the commercial reagent and 5.6% weight loss
for the recrystallized reagent. When the commercial and re-
crystallized reagents were dried at 343 K overnight under at-
mospheric pressure, only two weight loss steps at about 423
K and 473 K were observed in the TGA of the dried
reagent. The results of the Karl Fischer method showed that
the water content is (10:2 ꢂ 0:2)% in the commercially avail-
able reagent and (7:1 ꢂ 0:2)% in the dried reagent. The IR
spectra at 3465 cmꢃ1 clearly indicated the existence of water
in both the commercially available and the dried reagent. It
is obvious that there are two kinds of water in
Experimental
Reagents. H2tpfpp and [Ni(hfac)2] nH2O were obtained from
ꢄ
the Aldrich Co. H2tpfpp (Anal. Calcd: C, 54.22%; H, 1.04%; N,
5.75%; Found: C, 54.20%; H, 1.04%; N, 5.78%) was used without
further purification. The scCO2 solvent was prepared using liquid
carbon dioxide (99.99%, Showa Tansan). The commercial re-
agent of [Ni(hfac)2] nH2O was recrystallized from an acetone/
ꢁ
ꢄ
water (50% volume) solution and dried overnight at 70 C before
use. [Ni(tpfpp)] was prepared as follows: nickel(b) acetate tetra-
hydrate (Wako, 58.5 mg, 0.235 mmol) was added to a DMF
(Wako, 30 cm3) solution of H2tpfpp (Aldrich, 31.4 mg, 0.032
mmol). The mixture was refluxed for 12 h and then added to
100 cm3 of water containing NaCl (5 g). The precipitate was col-
lected by filtration and washed well with water. Recrystallization
from ethanol/water gave a red powder product. Anal. Calcd for
[Ni(tpfpp)]: C, 51.24; H, 0.78; N, 5.43%. Found: C, 50.98; H,
1.30; N, 5.28%. The absorbance peaks of the UV-vis spectra of
[Ni(tpfpp)] appeared at 396, 521, and 555 nm in scCO2, and at
404, 524 and 558 nm in hexane.
[Ni(hfac)2] nH2O; one is the coordinated water which directly
ꢄ
binds to the nickel(b) ion and remains after heating at 343 K,
and the other is the uncoordinated water which releases at 343
K. These results, accompanied by the results of the elemental
analysis, support the compositions of [Ni(hfac)2(H2O)2] H2O
ꢄ
for the commercially available reagent (Calcd: H2O, 10.25%;
C, 22.77%; H, 1.52%; Found: H2O, 10.2%; C, 22.73%; H,
1.31%) and [Ni(hfac)2(H2O)2] for the dried reagent (Calcd:
H2O, 7.07%; C, 23.58%; H, 1.18%; Found: H2O, 7.1%; C,
22.64%; H, 1.13%). The latter was used for the kinetic mea-
surements.
Measurements. Thermal gravity analysis (TGA) was carried
out with a Mettler-Toledo TGA/SDTA 851 instrument. Infrared
(IR) spectra (KBr pellets) were recorded on a Nicollet AVATAR
36-FT-IR spectrophotometer. Magnetic susceptibilities were mea-
sured at room temperature using an MSB-AUTO magnetic sus-
ceptibilitiy balance. A CA-07 moisture meter (Mitsubishi) was
used to determine the water content of the solid sample by dissol-
ving the sample in acetonitrile.
Kinetic Measurements. The details of the measurements of
the reaction rates in scCO2 using the spectrophotometric cell,
the scCO2-stopped-flow instrument, and the scCO2 measurement
system were previously described.30;39 In this study, a newly con-
structed spectrophotometric supercritical fluid cell with a volume
of 3.43 cm3 and an optical path length of 0.991 cm was used in the
UV-1600 spectrophotometer (Shimadzu) for the kinetic
measurements. The cell has two sapphire windows, which are
transparent over the range of 250–800 nm. The pressure in the
The magnetic measurement indicated that [Ni(hfac)2-
(H2O)2] is paramagnetic in the solid state with the moment
of 2.86 B.M. at 300 K.
The UV-visible spectra of
[Ni(hfac)2(H2O)2] in scCO2 (315 and 645 nm at 343.8 K
and 20.5 MPa, similar to that in chloroform at 315 and 641
nm) suggested that this species is a monomer with an octahe-
dral structure in scCO2.40{42
Nickel(b) complexes with a number of pentane-2,4-dione
derivatives having alkyl substituents in the 3-position have
been prepared, and the substituents in the 3-position of the
2,4-diketonates have a significant influence on the tendency
of the nickel(b) complex to adopt either a monomeric or a tri-
meric structure as the result of the electronic effects associated
with the substituent.43 Both the mono- and trimeric deriva-

S. Liu et al.
Bull. Chem. Soc. Jpn., 76, No. 5 (2003) 1001
Table 1. The Experimental Values of kobs for the Reaction
System of {H2tpfpp + [Ni(hfac)2(H2O)2]g in scCO2 at
20.5 MPa and Various Temperatures. The Concentration
of H2tpfpp is 4:8 ꢅ 10ꢃ6 mol dmꢃ3
T/K
CNi=10ꢃ4 mol dmꢃ3
xNi=10ꢃ5
kobs=10ꢃ4 sꢃ1
328.3
1.45
2.19
3.76
4.81
7.02
9.58
13.0
15.8
0.839
1.27
2.17
2.78
4.05
5.53
7.47
9.13
0:889 ꢂ 0:033
1:15 ꢂ 0:03
1:64 ꢂ 0:04
2:05 ꢂ 0:03
2:15 ꢂ 0:04
2:53 ꢂ 0:06
2:54 ꢂ 0:05
2:63 ꢂ 0:14
333.3
1.14
1.27
1.60
2.17
3.80
4.86
6.99
9.23
11.2
13.3
16.2
0.687
0.765
0.961
1.31
2.28
2.92
4.20
5.55
6.75
7.99
9.74
1:64 ꢂ 0:06
1:63 ꢂ 0:06
1:92 ꢂ 0:06
2:58 ꢂ 0:05
3:59 ꢂ 0:08
3:87 ꢂ 0:08
4:55 ꢂ 0:09
5:25 ꢂ 0:07
4:90 ꢂ 0:09
5:13 ꢂ 0:07
5:23 ꢂ 0:12
Fig. 1. Spectral change for the reaction system of {H2tpfpp
+ [Ni(hfac)2(H2O)2]g in scCO2 at 333.3 K and 20.5 MPa,
2:17 ꢅ 10ꢃ4 mol dmꢃ3 of [Ni(hfac)2(H2O)2], and
4:75 ꢅ 10ꢃ6 mol dmꢃ3 of H2tpfpp. The spectra at
t ¼ 10, 20, 30, 40, 60, 120, 240, and 480 min were
depicted. In the inset, the absorbances at 500 nm (A500
referred A470 versus the reaction time (t) were plotted.
)
338.6
1.07
2.14
3.24
4.41
9.51
13.3
16.1
0.673
1.35
2.04
2.78
5.98
8.36
10.1
2:79 ꢂ 0:09
4:95 ꢂ 0:09
7:27 ꢂ 0:15
8:28 ꢂ 0:23
10:2 ꢂ 0:1
10:4 ꢂ 0:3
10:6 ꢂ 0:1
tives of nickel(b) readily form monomeric adducts with etha-
nol, where the nickel atom has an octahedral environment,
with the four oxygen atoms of the pentadionato fragment lying
in the equatorial plane and two ethanol oxygen atoms in the
axial positions.44 In fact, [Ni(hfac)2] readily forms octahedral
complexes in the presence of coordinating ligands including
the donating solvents such as water.45{48 As a result, the
two waters in [Ni(hfac)2(H2O)2] are expected to be in the trans
position.
Kinetics of Metalloporphyrin Formation between
[Ni(hfac)2(H2O)2] and H2tpfpp. An example of the ob-
served change in the UV-vis absorption spectrum is shown
in Fig. 1. The spectral change in Fig. 1 was characteristic
for the formation of a metalloporphyrin, and the final absorp-
tion spectrum of the reaction product was very consistent with
that of [Ni(tpfpp)], which was independently prepared. All the
kinetic measurements in scCO2 were carried out under the
pseudo-first order conditions where [Ni(hfac)2(H2O)2] is in
large excess relative to H2tpfpp. The absorbance at 500 nm
(A500) corresponding to the decrease in H2tpfpp was found to
change as an exponential function, whereas there is almost
no absorbance at 500 nm for [Ni(hfac)2(H2O)2] and
[Ni(tpfpp)]. The change in A500 (referred to A470) as a function
of the reaction time (t) was then analyzed by a non-linear least-
squares calculation according to the following equation:
343.8
0.623
0.685
0.897
1.09
1.17
2.14
3.77
4.76
4.81
6.94
9.53
9.68
12.5
14.4
0.411
0.452
0.592
0.718
0.772
1.41
2.49
3.14
3.17
4.58
3:08 ꢂ 0:14
3:87 ꢂ 0:13
5:03 ꢂ 0:20
6:15 ꢂ 0:15
6:49 ꢂ 0:17
10:1 ꢂ 0:3
12:8 ꢂ 0:2
14:0 ꢂ 0:2
15:1 ꢂ 0:6
15:2 ꢂ 0:5
15:4 ꢂ 0:6
17:1 ꢂ 0:2
16:8 ꢂ 0:2
18:1 ꢂ 0:6
6.29
6.39
8.24
9.52
to 14:30 ꢅ 10ꢃ6 mol dmꢃ3. The rate for the formation of the
metalloporphyrin, [Ni(tpfpp)], was confirmed to be first order
in porphyrin.
The kobs values were determined at a constant concentration
of H2tpfpp (4:8 ꢅ 10ꢃ6 mol dmꢃ3) by varying the total concen-
tration of CNi at various temperatures (Table 1) and various
pressures (Table S1: Supplementary data). The kobs values
are plotted in Fig. 2 versus the mole fraction (xNi) of
[Ni(hfac)2(H2O)2], which is independent of temperature and
pressure. The plots at various temperatures (Fig. 2A) and var-
ious pressures (Fig. 2B) indicated the curvature and eventual
A500 ¼ A1 ꢃ ðA1 ꢃ A0Þ exp ðꢃkobstÞ
ð1Þ
where A and A0 are the absorbance at t ¼ 1 and 0, respec-
1
tively, and kobs is the conditional first-order rate constant.
At 333.3 K and 30.0 MPa, the average value of kobs was
ð3:94 ꢂ 0:20Þ ꢅ 10ꢃ4 sꢃ1 for six runs at a constant [Ni-
(hfac)2(H2O)2] concentration (CNi ¼ 1:6 ꢅ 10ꢃ3 mol dmꢃ3
)
and the varied concentrations of H2tpfpp from 2:90 ꢅ 10ꢃ6

1002 Bull. Chem. Soc. Jpn., 76, No. 5 (2003)
Metalation of Porphyrin in Supercritical CO2
Table 2. The Obtained Values of K and k for the Reaction
System of {H2tpfpp + [Ni(hfac)2(H2O)2]g in scCO2 at
Various Temperatures and Pressures
T/K
P/MPa
KaÞ=104
kbÞ=10ꢃ4 sꢃ1
328.3
333.3
333.3
333.3
333.3
338.6
343.8
20.5
20.5
25.0
30.0
35.0
20.5
20.5
4:61 ꢂ 0:61
4:89 ꢂ 0:50
4:79 ꢂ 0:58
4:65 ꢂ 0:51
4:53 ꢂ 0:57
5:34 ꢂ 0:89
5:87 ꢂ 0:57
3:34 ꢂ 0:15
6:58 ꢂ 0:23
5:72 ꢂ 0:23
4:77 ꢂ 0:18
4:36 ꢂ 0:19
13:0 ꢂ 0:7
21:1 ꢂ 0:7
a) Evaluated thermodynamic parameters: ꢀHꢁ ¼ 14:4 ꢂ 1:3
kJ molꢃ1 at 20.5 MPa, ꢀSꢁ ¼ 133 ꢂ 4 J molꢃ1 Kꢃ1 at 20.5
MPa, and ꢀVꢁ ¼ 15 ꢂ 1 cm3 molꢃ1 at 333.3 K. b) Evaluated
kinetic parameters: ꢀH ¼ 110 ꢂ 5 kJ molꢃ1 at 20.5 MPa,
¼j
¼j
¼j
ꢀS ¼ 22 ꢂ 16 J molꢃ1 Kꢃ1 at 20.5 MPa, and ꢀV
¼
81 ꢂ 8 cm3 molꢃ1 at 333.3 K.
kobs ¼ KkxNi=ð1 þ KxNiÞ
ð2Þ
The dissociation of H2O in the preequilibrium step can be
ruled out, because the absorbance changes were reproduced
by an exponential function within the error level of the absor-
bance measurements even at the lowest CNi (6:23 ꢅ 10ꢃ5
mol dmꢃ3). If H2O is dissociated at the pre-equilibrium step,
the observed absorbance changes should not obey the expo-
nential function, because the concentration of H2O increases
with the progress of the reaction. Moreover, if the inner-
sphere complex, such as the so-called sitting-atop [(H2-
tpfpp)Ni(hfac)2], is formed, and two H2O molecules are disso-
ciated at the preequilibrium step, the interaction in [(H2-
tpfpp)Ni(hfac)2] must be strong, then [(H2tpfpp)Ni-
(hfac)2] should give a spectrum which is different from the
free H2tpfpp and metalloporphyrin, [Ni(tpfpp)]. However,
we did not observe such a spectrum during the reaction.
The experimental values of kobs at a constant temperature
and pressure listed in Table 1 and Table S1 were analyzed
by a least-squares calculation according to Eq. 2, and the ob-
tained K and k values at various temperatures and pressures
are summarized in Table 2.
Fig. 2. Dependence of the conditional rate constants (kobs
)
on the mole fraction (xNi) of [Ni(hfac)2(H2O)2] for the re-
action system of {H2tpfpp + [Ni(hfac)2(H2O)2]g in
scCO2. In A were given the kobs values at 20.5 MPa
and various temperatures, (a): 328.3 K, (b): 333.3 K, (c):
338.6 K, and (d): 343.8 K. In B were given the kobs values
at 333.3 K and various pressures, (e): 20.5 MPa, (f): 25.0
MPa, (g): 30.0 MPa, and (h): 35.0 MPa. The solid curves
were calculated using the obtained K and k values which
are given in Table 2.
Under the isobaric conditions, the K value is related to the
changes in enthalpy (ꢀHꢁ) and entropy (ꢀSꢁ) by Eq. 3 for
the chemical equilibrium, and k is related to the activation en-
¼j
¼j
thalpy (ꢀH ) and activation entropy (ꢀS ) by Eq. 4 for the
chemical reaction.
ln K ¼ ꢃꢀHꢁ=RT þ ꢀSꢁ=R
ln ðkh=kBTÞ ¼ ꢃꢀH =RT þ ꢀS =R
ð3Þ
ð4Þ
¼j
¼j
Scheme 1.
where h is Planck’s constant and kB is Boltzmann’s constant.
The obtained values of K and k at 20.5 MPa listed in
Table 2 were analyzed by a least-squares calculation using
Eqs. 3 and 4, respectively. The thermodynamic and kinetic
saturation. Such a saturation dependence invokes Scheme 1
for the reaction between H2tpfpp and [Ni(hfac)2(H2O)2]. In
Scheme 1, K is the equilibrium constant for the fast preequili-
brium to form the outer-sphere association intermediate, [Ni(h-
parameters were evaluated as ꢀHꢁ ¼ 14:4 ꢂ 1:3 kJ molꢃ1
,
¼j
ꢀSꢁ ¼ 133 ꢂ 4 J molꢃ1 Kꢃ1, ꢀH ¼ 110 ꢂ 5 kJ molꢃ1, and
fac)2(H2O)2] H2tpfpp, and k is the first-order rate constant for
ꢄ
ꢀS ¼ 22 ꢂ 16 J molꢃ1 Kꢃ1 at 20.5 MPa.
¼j
the rate-determining formation of the metalloporphyrin pro-
duct, [Ni(tpfpp)]. The kobs value is then expressed in terms
of K, k, and xNi by the following equation:49
Furthermore, the pressure dependence of K and k at a con-
stant temperature, which are given in pressure-independent

S. Liu et al.
Bull. Chem. Soc. Jpn., 76, No. 5 (2003) 1003
units, i.e., mole fraction units, is expressed by Eqs. 5 and 6.
@ ln K=@P ¼ ꢃꢀVꢁ=RT
@ ln k=@P ¼ ꢃꢀV =RT
ð5Þ
ð6Þ
¼j
where ꢀVꢁ is the change in volume for the preequilibrium (K),
¼j
and ꢀV is the activation volume for the rate-determining step
(k). The obtained values of K and k at 333.3 K listed in
Table 2 were analyzed by a least-squares calculation using
Eqs. 5, and 6, respectively. The determined values are ꢀVꢁ ¼
¼j
15 ꢂ 1 cm3 molꢃ1 and ꢀV ¼ 81 ꢂ 8 cm3 molꢃ1 at 333.3 K.
As is apparent from the relation shown in Fig. 2, just after
starting the reaction, the outer-sphere association intermediate,
[Ni(hfac)2(H2O)2] H2tpfpp, is almost quantitatively (ca. 82%)
ꢄ
formed at the high concentration of [Ni(hfac)2(H2O)2] as much
as CNi ¼ 1:7 ꢅ 10ꢃ3 mol dmꢃ3 (xNi ¼ 9:82 ꢅ 10ꢃ4) at 328.3 K
and 20.5 MPa. Under such conditions, the measured spectrum
for the reaction system of {H2tpfpp + [Ni(hfac)2(H2O)2]g
within 10 min just after starting the reaction is shown to be al-
most the same as the sum of the two spectra of H2tpfpp and
[Ni(hfac)2(H2O)2]. This finding indicates that the interaction
in the outer-sphere association is not strong enough to change
Fig. 3. Dependence of the conditional rate constants (kobs
)
on the concentrations of H2O (CW) for the reaction system
of {H2tpfpp + [Ni(hfac)2(H2O)2] + H2Og in scCO2 at
20.5 MPa and 343.8 K.
The concentrations of
the absorption spectrum of the free porphyrin.
The
[Ni(hfac)2(H2O)2] and H2tpfpp are 4:9 ꢅ 10ꢃ4 mol dmꢃ3
and 4:8 ꢅ 10ꢃ6 mol dmꢃ3, respectively. The point at
CW ¼ 0 (filled circle) was obtained from the known values
of K and k using Eq. 2. The solid curve is calculated using
the obtained values.
[Ni(hfac)2(H2O)2] molecule is almost planar because two
waters in [Ni(hfac)2(H2O)2] are in the trans position. Since
the porphyrin core of H2tpfpp also has a planar structure,
H2tpfpp and [Ni(hfac)2(H2O)2] molecules may interact be-
tween their molecular planes in the outer-sphere association
complex.
Kinetics of Metalloporphyrin Formation between
[Ni(hfac)2(H2O)2] and H2tpfpp in the Presence of a Large
Amount of Added Water. The addition of a large amount
of water to the reaction system made the reaction rate of the
metalloporphyrin formation much slower, and the absorbance
change during the reaction was observed as an exponential
function. The conditional first-order rate constants (kobs) at
343.8 K and 20.5 MPa were determined by the nonlinear
least-squares calculation according to Eq. 1 (given in Table
S2). Figure 3 shows the relation between the value of kobs
and the concentration of added water (Cw). The relation of
the retardation of the metalloporphyrin formation rate indi-
cates the production of unreactive water adduct species that
are formed by the interaction between water molecules and
[Ni(hfac)2(H2O)2].
Scheme 2.
With the knowledge of the values of K ¼ ð3:87 ꢂ 0:38Þ ꢅ 103
molꢃ1 dm3 and k ¼ ð2:11 ꢂ 0:07Þ ꢅ 10ꢃ3 sꢃ1 at 343.8 K and
20.5 MPa determined according to Eq. 2, we can estimate
the values of ꢀ1 ¼ ð4:0 ꢂ 0:7Þ ꢅ 102 molꢃ1 dm3 and of
ꢀ2 ¼ ð2:6 ꢂ 0:6Þ ꢅ 104 molꢃ2 dm6 at 343.8 K and 20.5 MPa
by the least-squares fitting.
Mechanism of Metalloporphyrin Formation in scCO2.
CO2 has a modest polarizability, no dipole moment, and a
large quadrupole moment. Usually fluorinated compounds
are more soluble relative to hydrogenated molecules. Re-
cently, Dardin et al.51 have presented NMR evidence for spe-
cific solute-solvent interactions between CO2 and fluorinated
compounds in scCO2 which come from the van der Waals in-
teraction between the fluorinated sites in the solute and carbon
dioxide. Interestingly, the strong face-to-face stacking interac-
tion of benzene and hexafluorobenzene is explained by a rever-
sal of the quadrupole moment of hexafluorobenzene because of
the electron-withdrawing fluorine substituents, allowing a net
electrostatic attraction between the p-faces to occur.52 In ad-
It has been reported that [Ni(hfac)2(H2O)2] has the ability to
form a bis-adduct by hydrogen bonding between the aqua li-
gand and the nitrogen in the adduct ligand.50 The composition
of [Ni(hfac)2] 5H2O has also been reported.48 As a result, the
ꢄ
hydrogen bonding between the aqua ligand in
[Ni(hfac)2(H2O)2] and the oxygen of H2O in the outer sphere
is expected to form the water adduct, [Ni(hfac)2-
(H2O)2] 2H2O, in scCO2. Therefore, in the presence of a
ꢄ
large excess of added water, we can propose the reactions in
Scheme 2, where ꢀ1 and ꢀ2 are the equilibrium constants
for the water adduct in the outer sphere. According to
Scheme 2, if CNi is the sum of the concentrations of
[Ni(hfac)2(H2O)2], [Ni(hfac)2(H2O)2] H2O, and [Ni(hfac)2-
(H2O)2] 2H2O, we have the following relation:
ꢄ
ꢄ
2
kobs ¼ kKCNi=ð1 þ ꢀ1CW þ ꢀ2CW þ KCNiÞ
ð7Þ

1004 Bull. Chem. Soc. Jpn., 76, No. 5 (2003)
Metalation of Porphyrin in Supercritical CO2
thalpy (110 kJ molꢃ1).
Tables S1 and S2 as supplementary data are available by a
request to the authors.
dition, the solubility of H2tpfpp in scCO2 has been shown to
increase with the increasing density of scCO2.39
In the present case, the H2tpfpp and [Ni(hfac)2(H2O)2] mo-
lecules are solvated by the solvent CO2 molecules in scCO2.
The solvation of two separate molecules of H2tpfpp and
[Ni(hfac)2(H2O)2] may be more oriented than that in one asso-
ciated molecule due to the larger molecular surface. The out-
er-sphere association by the interaction between H2tpfpp and
[Ni(hfac)2(H2O)2] leads to the desolvation of the solvated
molecules. Therefore, the positive change in enthalpy (14.4
kJ molꢃ1) and the large positive value for the entropy change
(133 J molꢃ1 Kꢃ1) of the pre-association equilibrium will be
expected. As described above, the outer-sphere association
complex for the pre-equilibrium is not a SAT complex.
Furthermore, it is reasonable that the formation of the water
adduct, [Ni(hfac)2(H2O)2] H2O and [Ni(hfac)2(H2O)2] 2H2O,
This work was supported by Grants-in-Aid for Scientific
Research (Nos. 11354009, 11554030, and 12874094) from
the Ministry of Education, Science, Sports and Culture. S.
L. is a postdoctoral fellow of the Japan Society for the Promo-
tion of Science for foreign researchers in Japan.
References
1
K. P. Johnston and J. M. L. Penninger, ‘‘Supercritical Fluid
Science and Technology,’’ American Chemical Society,
Washington (1989).
2
F. V. Bright and M. E. P. McNally, ‘‘Supercritical Fluid
ꢄ
ꢄ
Technology: Theoretical and Applied Approaches in Analytical
Chemistry,’’ American Chemical Society, Washington (1992).
between [Ni(hfac)2(H2O)2] and water molecules by hydrogen
bonding strongly inhibits the interaction between H2tpfpp
and [Ni(hfac)2(H2O)2]. The decrease in the equilibrium con-
stant (K) with the increase in the density of scCO2 is expressed
by the positive value of 15 cm3 molꢃ1 for the volume change.
The formation of the outer-sphere association will become
more difficult when more CO2 molecules are involved in the
solvation shell and surround the solute molecules.
3
E. Kiran and J. F. Brennecke, ‘‘Supercritical Fluid Engi-
neering Science: Fundamentals and Applications,’’ American
Chemical Society, Washington (1993).
4
(1995).
5
P. G. Jessop, T. Ikariya, and R. Noyori, Science, 269, 1065
E. Kiran, P. G. Debenedetti, and C. J. Peters, ‘‘Supercriti-
cal Fluids: Fundamentals and Applications,’’ Kluwer Academic
Publishers, Dordrecht (1998).
According to the recent investigations on the metalation re-
action of porphyrins,29;32{34 it has been concluded that the nu-
cleophilic attack of the first pyrrolenine nitrogen of the por-
phyrin is the rate-determining step for the overall metal ion
incorporation. Therefore, during the activation process of
the present Ni(b) ion incorporation into H2tpfpp, it is expected
that the nucleophilic attack of the pyrrolenine nitrogen of
H2tpfpp to the central metal ion occurs together with the dis-
sociation of a water at the axial position. At the same moment,
the dissociation of the donating oxygen of the ꢀ-diketonate li-
gand with a negative charge should accompany the abstraction
of a pyrrole amine proton of H2tpfpp, because the charged spe-
cies is quite unstable in scCO2 that has a low dielectric con-
stant.
6
T. Clifford, ‘‘Fundamentals of Supercritical Fluids,’’
Oxford University Press, Oxford (1999).
7
8
S. B. Hawthorne, Anal. Chem., 62, 633A (1990).
R. M. Smith, ‘‘Supercritical Fluid Chromatography,’’ The
Royal Society of Chemistry, Cambridge (1993).
M. D. L. de Castro, M. Valcarcel, and M. T. Tena, ‘‘Ana-
9
lytical Supercritical Fluid Extraction,’’ Springer-Verlag, New
York (1994).
10 L. T. Taylor, ‘‘Supercritical Fluid Extraction,’’ John Wiley
& Sons, New York (1996).
11 E. J. Beckman, Science, 271, 613 (1996).
12 D. J. Dixon, G. Luna-Barcenas, and K. P. Johnston, Poly-
mer, 35, 3998 (1994).
13 F. Dehghani, T. Wells, N. J. Cotton, and N. R. Foster, J.
Supercrit. Fluids, 9, 263 (1996).
14 H. Wu, Y. Lin, N. G. Smart, and C. M. Wai, Anal. Chem.,
68, 4072 (1996).
15 D. K. Lavallee, Mol. Struct. Energ., 9, 279 (1988).
16 D. K. Lavallee, ‘‘The Chemistry and Biochemistry of N-
Substituted Porphyrins,’’ VCH, New York (1987).
17 M. E. Blackwood, Jr., T. S. Rush, c, A. Medlock, H. A.
Dailey, and T. G. Spiro, J. Am. Chem. Soc., 119, 12170 (1997).
18 M. M. Conn, J. R. Prudent, and P. G. Schultz, J. Am. Chem.
Soc., 118, 7012 (1996).
19 P. Hambright and P. B. Chock, J. Am. Chem. Soc., 96,
3123 (1974).
20 M. J. Bain-Ackerman and D. K. Lavallee, Inorg. Chem.,
18, 3358 (1979).
21 J. Turay and P. Hambright, Inorg. Chem., 19, 562 (1980).
22 S. Funahashi, K. Saito, and M. Tanaka, Bull. Chem. Soc.
Jpn., 54, 2695 (1981).
23 R. F. Pasternack, G. C. Vogel, C. A. Skowronek, R. K.
Harris, and J. G. Miller, Inorg. Chem., 20, 3763 (1981).
24 S. Funahashi, Y. Yamaguchi, and M. Tanaka, Bull. Chem.
Soc. Jpn., 57, 204 (1984).
For the rate-determining step, there are two components for
the activation volume: a repulsive or intrinsic part due to the
change in occupied volume between the reactants and the tran-
sition state, and a part due to attractive solute-solvent
forces.53{55 For the unimolecular decomposition reaction,
the repulsive part of the activation volume is usually positive
due to an increase in the bond length as bonds are broken.53
The attractive part due to solute-solvent forces is, however, de-
pendent on the isothermal compressibility of supercritical
fluids and the polarities of transition state and reactants. This
part may be a large negative value54 or a negligible value.55
In the present case, since CO2 is a non-coordinating solvent
and the experimental conditions are far from the critical point,
we can assume that the activation volume for the rate-deter-
mining step is dominated by the repulsive contribution in the
outer-sphere association complex, [Ni(hfac)2(H2O)2] H2-
ꢄ
tpfpp. This assumption suggests that the release of H2O and
the dissociation of the donating oxygen of the ꢀ-diketonate li-
gand in the transition state cause the large repulsive contribu-
¼j
tion, which accounts for the positive ꢀV value (81
cm3 molꢃ1) and the large positive value of the activation en-

S. Liu et al.
Bull. Chem. Soc. Jpn., 76, No. 5 (2003) 1005
25 S. Funahashi, Y. Yamaguchi, and M. Tanaka, Inorg.
Chem., 23, 2249 (1984).
43 A. Dohring, R. Goddard, P. W. Jolly, C. Kruger, and V. R.
Polyakov, Inorg. Chem., 36, 177 (1997).
26 L. R. Robinson and P. Hambright, Inorg. Chim. Acta, 185,
17 (1991).
44 V. R. Polyakov and M. Czerwinski, Inorg. Chem., 40, 4798
(2001).
27 J. Takeda, T. Ohya, and M. Sato, Inorg. Chem., 31, 2877
(1992).
45 A. Caneschi, D. Gatteschi, J. P. Renard, P. Rey, and R.
Sessoli, Inorg. Chem., 28, 2940 (1989).
28 M. Inamo, A. Tomita, Y. Inagaki, N. Asano, K. Suenaga,
M. Tabata, and S. Funahashi, Inorg. Chim. Acta, 256, 77 (1997).
29 Y. Inada, Y. Sugimoto, Y. Nakano, Y. Itoh, and S.
Funahashi, Inorg. Chem., 37, 5519 (1998).
30 Y. Inada, H. Sato, T. Horita, T. Nagamura, and S.
Funahashi, J. Supercrit. Fluids, 21, 205 (2001).
31 E. B. Fleischer and J. H. Wang, J. Am. Chem. Soc., 82,
3498 (1960).
32 Y. Inada, Y. Nakano, M. Inamo, M. Nomura, and S.
Funahashi, Inorg. Chem., 39, 4793 (2000).
33 Y. Inada, T. Yamaguchi, H. Sato, and S. Funahashi, Inorg.
React. Mech., 2, 277 (2000).
46 M. H. Dickman, Acta Crystallogr., Sect. C, 56, 58 (2000).
47 L. C. Porter, M. H. Dickman, and R. J. Doedens, Inorg.
Chem., 27, 1548 (1988).
48 K. Arakawa and K. Tanikawa, Bunseki Kagaku, 16, 812
(1967).
49 R. G. Wilkins, ‘‘The Study of Kinetics and Mechanism of
Reaction of Transition Metal Complexes,’’ Allyn and Bacon, Bos-
ton (1974). As pointed out in page 28 in Ref. 49, the same kinetic
behavior as in Eq. 2 is observed if the complex was viewed as a
‘‘dead-end’’ intermediate in rapid equilibrium and if the reaction
proceeded bimolecularly between [Ni(hfac)2(H2O)2] and H2tpfpp.
Kinetics alone will not distinguish between these two mechan-
isms. Arguments over which of the two is preferred must then
be based on the chemical considerations. As discussed later, the
mechanism expressed by Eq. 2 must be plausible. A related sub-
ject has been reported: Ref. 27.
34 M. Inamo, N. Kamiya, Y. Inada, M. Nomura, and S.
Funahashi, Inorg. Chem., 40, 5636 (2001).
35 S. Funahashi, Y. Inada, and M. Inamo, Anal. Sci., 17, 917
(2001).
36 J. A. Darr and M. Poliakoff, Chem. Rev., 99, 495 (1999).
37 A. F. Lagalante, B. N. Hansen, and T. J. Bruno, Inorg.
Chem., 34, 5781 (1995).
38 M. Ashraf-Khorassani, M. T. Combs, and L. T. Taylor,
Talanta, 44, 755 (1997).
39 H. Sato, Y. Inada, T. Nagamura, and S. Funahashi, J.
Supercrit. Fluids, 21, 71 (2001).
40 F. A. Cotton and J. P. Fackler, Jr., J. Am. Chem. Soc., 83,
2818 (1961).
50 D. Luneau, P. Rey, J. Laugier, E. Belorizky, and A. Cogne,
Inorg. Chem., 31, 3578 (1992).
51 A. Dardin, J. M. DeSimone, and E. T. Samulski, J. Phy.
Chem. B, 102, 1775 (1998).
52 A. P. Jr. West, S. Mecozzi, and D. A. Dougherty, J. Phys.
Org. Chem., 10, 347 (1997).
53 S. D. Hamann and R. S. Bradley, ‘‘High-Pressure Physics
and Chemistry,’’ Vol. 2, Academic Press, London and New York
(1963).
41 J. P. Fackler, Jr. and F. A. Cotton, J. Am. Chem. Soc., 83,
3775 (1961).
42 J. P. Fackler, Jr., M. L. Mittleman, H. Weigold, and G. M.
Barrow, J. Phys. Chem., 72, 4631 (1968).
54 K. P. Johnston and C. Haynes, AIChE J., 33, 2017 (1987).
55 Q. Ji, C. R. Lloyd, E. M. Eyring, and R. van Eldik, J. Phys.
Chem. A, 101, 243 (1997).