Chemistry of the strong electrophilic metal fragment [99Tc(N)(PXP)]2+ (PXP = diphosphine ligand). A novel tool for the selective labeling of small molecules

Article abstract of DOI:10.1021/ja0200239

Monosubstituted [M(N)Cl₂(POP)] [M = Tc, 1; Re, 2] and [M(N)Cl₂(PNP)] [M = Tc, 3; Re, 4] complexes were prepared by reaction of the precursors [M(N)Cl₄]^- and [M(N)Cl₂(PPh₃)₂] (M =

Full text of DOI:10.1021/ja0200239

Published on Web 08/30/2002
Chemistry of the Strong Electrophilic Metal Fragment
[99Tc(N)(PXP)]2+ (PXP ) Diphosphine Ligand). A Novel Tool
for the Selective Labeling of Small Molecules
Cristina Bolzati,‡ Alessandra Boschi,† Licia Uccelli,† Francesco Tisato,*,‡
Fiorenzo Refosco,‡ Aldo Cagnolini,‡ Adriano Duatti,*,† Sushumna Prakash,†
Giuliano Bandoli,§ and Andrea Vittadini|
Contribution from the ICIS - C.N.R., Corso Stati Uniti, 4, 35127 PadoVa, Italy,
Laboratory of Nuclear Medicine, Department of Clinical and Experimental Medicine, UniVersity
of Ferrara, Via L. Borsari, 46, 44100 Ferrara, Italy, Department of Pharmaceutical Sciences,
UniVersity of PadoVa, Via Marzolo, 5, 35131 PadoVa, Italy, and ISTM - C.N.R., Via Marzolo,
1, 35131 PadoVa, Italy
Received January 3, 2002
Abstract: Monosubstituted [M(N)Cl2(POP)] [M ) Tc, 1; Re, 2] and [M(N)Cl2(PNP)] [M ) Tc, 3; Re, 4]
complexes were prepared by reaction of the precursors [M(N)Cl4]- and [M(N)Cl2(PPh3)2] (M ) Tc, Re) with
the diphosphine ligands bis(2-diphenylphosphinoethyl)ether (POP) and bis(2-diphenylphosphinoethyl)-
methoxyethylamine (PNP) in refluxing dichloromethane/methanol solutions. In these compounds, the
diphosphine acted as a chelating ligand bound to the metal center through the two phosphorus atoms.
Considering also the weak interaction of the heteroatom (N or O) located in the middle of the carbon
backbone connecting the two P atoms, we found that the coordination arrangement of the diphosphine
ligand could be viewed as either meridional (m) or facial (f), and the resulting geometry as pseudooctahedral.
The heteroatom of the diphosphine ligand was invariably located trans to the nitrido linkage, as established
by X-ray diffraction analysis of the representative compounds 2m and 4f. Density functional theoretical
calculations showed that in POP-type complexes the mer form is favored by approximately 6 kcal mol-1
,
whereas mer and fac isomers are almost isoenergetic in PNP-type complexes. A possible role of noncovalent
interactions between the phosphinic phenyl substituents in stabilizing the fac-isomer was also highlighted.
The existence of fac-mer isomerism in this class of complexes was attributed to the strong tendency of
the two phosphorus atoms to occupy a reciprocal trans-position within the pseudooctahedral geometry.
The switching of P atoms between cis- and trans-configurations was confirmed by the observation that the
fac isomers, 1f and 2f, were irreversibly transformed, in solution, into the corresponding mer isomers, 1m
and 2m, thus suggesting that fac complexes are more reactive species. Theoretical calculations supported
this view by showing that the lowest unoccupied orbitals of the fac isomers are more accessible to a
nucleophilic attack with respect to those of the mer ones. Furthermore, the large participation of the Cl
orbitals to the HOMO, which is a metal-ligand π* antibonding in the complex basal plane, shows that the
Tc-Cl bonds are labile. As a consequence, facial isomers could be considered as highly electrophilic
intermediates that were selectively reactive toward substitution by electron-rich donor ligands. Experimental
evidence was in close agreement with this description. It was found that fac-[M(N)Cl2(PXP)] complexes
easily underwent ligand-exchange reactions with bidentate donor ligands such as mercaptoacetic acid
(NaHL1), S-methyl 2-methyldithiocarbazate (H2L2), diethyldithiocarbamate sodium salt (NaL3), and N-acetyl-
L-cysteine (H2L4) to afford stable asymmetrical heterocomplexes of the type fac-[M(N)(Ln)(POP)]+/0 (5-8)
and fac-[M(N)(Ln)(PNP)]+/0 (9-14) comprising two different polydentate chelating ligands bound to the same
metal center. In these reactions, the bidentate ligand replaced the two chloride atoms on the equatorial
plane of the distorted octahedron, leaving the starting fac-[M(N)(PXP)]2+ (X ) O, N) moieties untouched.
No formation of the corresponding symmetrical complexes containing two identical bidentate ligands was
detected over a broad range of experimental conditions. Solution-state NMR studies confirmed that the
structure in solution of these heterocomplexes was identical to that established in the solid state by X-ray
diffraction analysis of the prototype complexes fac-[M(N)(HL2)(POP)][BF4] [M ) Tc, 7; Re, 8] and fac-[Tc-
(N)(HL2)(PNP)][BF4], 11. In conclusion, the novel metal fragment fac-[M(N)(PXP)]2+ could be utilized as an
efficient synthon for the preparation of a large class of asymmetrical, nitrido heterocomplexes incorporating
a particular diphosphine ligand and a variety of bidentate chelating molecules.
Introduction
140 keV), and the recent advent of the â-emitting radionuclides
186Re (t1/2 ) 3.8 d, Eâmax ) 1.07 MeV) and 188Re (t1/2 ) 0.7 d,
The widespread use in diagnostic nuclear medicine of the
metastable isomer 99mTc, a pure γ-emitter with almost ideal
physical properties for nuclear imaging (t1/2 ) 6.06 h, Eγ )
Eâmax ) 2.11 MeV) as promising candidates for the application
of injectable radiopharmaceuticals to the therapy of malignant
and degenerative diseases, has made the inorganic chemistry
studies with group 7 elements technetium and rhenium very
attractive. The ultimate goal of these investigations is to
elucidate the molecular structure of 99mTc- and 188/186Re-based
agents produced in very low concentrations at the micro- and
* To whom correspondence should be addressed. (F.T.) Phone: ++049-
8295958. Fax: ++0498702911. E-mail: tisato@icis.cnr.it.
† University of Ferrara.
‡ ICIS - C.N.R.
§ University of Padova.
| ISTM - C.N.R.
9
11468
J. AM. CHEM. SOC. 2002, 124, 11468-11479
10.1021/ja0200239 CCC: $22.00 © 2002 American Chemical Society

Electrophilic Metal Fragment [99Tc(N)(PXP)]2+
A R T I C L E S
Chart 1
nanomolar scales (“noncarrier added” level, NCA). This can
be conveniently achieved through the comparison of their
chemical and physical properties with those of the corresponding
compounds, prepared at the macroscopic scale (“carrier added”
level, CA) with the long-lived â-emitting isotope 99Tc and with
the naturally occurring mixture of cold Re isotopes, and fully
characterized by standard analytical and spectroscopic methods.
Indeed, it is apparent that the possibility to clearly establish the
chemical identity of a specific tracer is of utmost importance
for the elucidation of its biological behavior.
In past years, the chemistry of 99mTc-radiopharmaceuticals1,2
has been expanded through the introduction of an efficient
method for the production of 99mTc-species containing a terminal
[
99mTc(N)]2+ group at the NCA level.3 Despite this, studies
devoted toward the elucidation of the inorganic chemistry
underlying nitrido Tc species are still rare.4 Electrochemical
investigations on nitrido Tc-compounds revealed a strong
reluctance of this moiety to undergo redox processes,5 as it is
more frequently observed with oxo-complexes. This property,
together with the stability of the nitrido group observed over a
wide range of pH values, makes nitrido-containing complexes
interesting candidates for nuclear medicine applications. In this
study, we describe the chemical reactivity of the metal fragment
[Tc(N)(PXP)]2+, composed by a diphosphine ligand (PXP)
coordinated to a [TcV(N)]2+ core. Actually, the metal fragment
[Tc(N)(PXP)]2+ was found to react with bidentate chelating
ligands, having π-donor atoms as coordinating substituents, to
form the heterocomplexes [Tc(N)(PXP)(YZ)]0/+ characterized
by an asymmetrical arrangement of polydentate ligands around
the TctN group. The most salient feature of these reactions
was that no formation of the corresponding symmetrical bis-
substituted complexes [Tc(N)(YZ)2]0/2- and [Tc(N)(PXP)2]2+
was detected under a broad range of experimental conditions.
As an illustration of this chemical behavior, we report here
a general and reliable route for the synthesis of asymmetrical
nitrido Tc(V) and Re(V) heterocomplexes. Using this procedure,
we will show that reactions of the labile precursors [M(N)Cl4]-
and [M(N)Cl2(PPh3)2] (M ) Tc, Re) with a particular class of
diphosphine ligands having a heteroatom incorporated in the
chain connecting the two phosphorus atoms, such as bis[(2-
diphenylphosphino)ethyl]ether (POP) and bis[(2-diphenylphos-
phino)ethyl]methoxyethylamine (PNP) (see Chart 1), afford
monosubstituted [M(N)Cl2(POP)] (M ) Tc, 1; Re, 2) and
[M(N)Cl2(PNP)] (M ) Tc, 3; Re 4) complexes, respectively.
In these monosubstituted complexes, PXP acts as a classical
bidentate diphosphine ligand through the two P atoms. However,
considering also the weak interaction of the diphosphine
heteroatom always positioned trans to the nitrido linkage, we
found that the resulting complexes can be viewed to possess
either a meridional (m) or a facial (f) arrangement, as established
by X-ray diffraction analyses of 2m and 4f. Facial isomers are
“activated” intermediates which readily react with nucleophilic
ligands such as mercaptoacetic acid sodium salt (NaHL1),
S-methyl 2-methyldithiocarbazate (H2L2), diethyldithiocarbamate
sodium salt (NaL3), and N-acetyl-L-cysteine (H2L4) (see Chart
1), yielding stable asymmetrical heterocomplexes of the type
fac-[M(N)(Ln)(POP)]+/0 (5-8) and fac-[M(N)(Ln)(PNP)]+/0 (9-
14), as outlined in Scheme 1.
A preliminary report on the formation of nitrido Tc(V) and
Re(V) asymmetrical heterocomplexes has been previously
communicated.6
Experimental Section
Caution! 99Tc is a weak â-emitter (Eâ ) 0.292 MeV, t1/2 ) 2.12 ×
105 years). All manipulations were carried out in laboratories approved
for low-level radioactivity using monitored hoods and gloveboxes.
When handled in milligram amounts, 99Tc does not present a serious
health hazard because common laboratory glassware provides adequate
shielding. Bremsstrahlung is not a significant problem due to the low
energy of the â-particles. However, normal radiation safety procedures
must be used at all times, especially with solid samples, to prevent
contamination and inhalation.
Reagents. Technetium as [NH4][99TcO4] was obtained from Oak
Ridge National Laboratory. Samples were dissolved in water and treated
with excess aqueous ammonia and H2O2 (30%) at 80 °C prior to use
to eliminate residual TcO2. Solid samples of purified ammonium
pertechnetate were obtained by slow evaporation of the solvent with
heating at 40 °C. General literature methods were applied to prepare
the precursor complexes [As(C6H4)4][99Tc(N)Cl4]7 and [99Tc(N)Cl2-
(PPh3)2],8 as well as [n-Bu4N][Re(N)Cl4]9 and [Re(N)Cl2(PPh3)2].10
Rhenium as a fine metal powder was obtained as a gift from H. C.
Starck GmbH, Goslar, Germany. It was first oxidized to perrhenate
and then reduced to suitable Re(V) compounds prior to use. Mercap-
toacetic acid sodium salt (NaHL1), diethyldithiocarbamate sodium salt
(NaL3), and N-acetyl-L-cysteine (H2L4) were purchased from Aldrich
(1) (a) Future of Nuclear Medicine, Part 1: Marketing Research Forecasts, J.
Nucl. Med. 1998, 39/2, 27N. (b) Future of Nuclear Medicine, Part 2:
Assessment of the Diagnostic Radiopharmaceutical Markets (2001-2020),
J. Nucl. Med. 1998, 39/3, 20N.
(2) (a) Jurisson, S.; Lydon, J. D. Chem. ReV. 1999, 99, 2205. (b) Liu, S.;
(6) Bolzati, C.; Boschi, A.; Duatti, A.; Prakash, S.; Uccelli, L.; Bandoli, G.;
Tisato, F.; Refosco, J. Am. Chem. Soc. 2000, 122, 4510.
(7) Baldas, J.; Bonnyman, J.; Williams, G. A. Inorg. Chem. 1986, 25, 150.
(8) Abram, U.; Lorenz, B.; Kaden, L.; Scheller, D. Polyhedron 1988, 7, 285.
(9) Abram, U.; Braun, M.; Abram, S.; Kirmse, R.; Voigt, A. J. Chem. Soc.,
Dalton Trans. (1972-1999) 1998, 231.
Edwards, D. S. Chem. ReV. 1999, 99, 2235.
(3) Pasqualini, R.; Comazzi, V.; Bellande, E.; Duatti, A.; Marchi, A. Int. J.
Appl. Radiat. Isot. 1992, 43, 1329.
(4) (a) Bandoli, G.; Dolmella, A.; Porchia, M.; Refosco, F.; Tisato, F. Coord.
Chem. ReV. 2001, 214, 43. (b) Tisato, F.; Refosco, F.; Bandoli, G. Coord.
Chem. ReV. 1994, 135/136, 325.
(10) Chatt, J.; Falck, C. D.; Leigh, G. J.; Paske, R. J. J. Chem. Soc. A 1969,
2288.
(5) Pasqualini, R.; Duatti, A. J. Chem. Soc., Chem. Commun. 1992, 1354.
9
J. AM. CHEM. SOC. VOL. 124, NO. 38, 2002 11469

A R T I C L E S
Bolzati et al.
Scheme 1
Chimica, and S-methyl 2-methyldithiocarbazate (H2L2) was prepared
as previously described.11 Common laboratory solvents and other
chemicals were used as received.
132.7, and 138.4 (aromatic C). 31P NMR (200 MHz, 85% H3PO4,
CDCl3): - 21.5 (s).
Syntheses of Technetium and Rhenium Complexes. fac-[M(N)-
Cl2(POP)] (M ) Tc, 1f; M ) Re, 2f). Both complexes are prepared
with a similar procedure, detailed here for the technetium complex.
To a stirred pink suspension of [M(N)Cl2(PPh3)2] (0.110 g, 0.155 mmol)
in dichloromethane (50 mL) was added a solution of POP (0.085 g,
0.194 mmol) in dichloromethane (15 mL). The mixture was refluxed
for 30 min. Within 5 min of reflux, the solution became transparent,
giving first an orange and finally a yellow color. After the solvent was
removed by a gentle stream of N2, the crude solid was washed with
dichloromethane (5 × 2 mL) and then with diethyl ether (3 × 2 mL)
to afford a bright yellow powder which was dried under a vacuum
pump overnight. 1f (yield 83%). Anal. Calcd for C28H28NOP2Cl2Tc:
C, 53.69; H, 4.50; N, 2.23. Found: C, 52.99; H, 4.38; N, 2.07. IR
(KBr, cm-1): 1435 (s), 1099 (s), 1068 (m) [ν(Tc(N))], 695 (s). 1H NMR
(CD2Cl2, ppm): 2.88 (m, 2H) and 3.08 (m, 2H) [P-CH2], 3.89 (m,
2H) and 4.21 (m, 2H) [O-CH2], 6.86-7.98 (20H, Harom). TLC: C18
(90/10, MeCN/H2O) Rf 0.81 (tailed). 2f (yield 57%). Anal. Calcd for
C28H28NOP2Cl2Re: C, 47.12; H, 3.95; N, 1.96. Found: C, 46.68; H,
4.07; N, 1.90. IR (KBr, cm-1): 1434 (s), 1095 (m) (Re-P), 1086 (m),
1063 (m) [ν(Re(N))], 696 (s). 1H NMR (CD2Cl2, ppm): 2.86 (m, 2H)
and 3.15 (m, 2H) [P-CH2], 3.90 (m, 2H) and 4.34 (m, 2H) [O-CH2],
6.85-8.00 (20H, Harom). 31P NMR (CD2Cl2, ppm): 16.9 (s). ESI MS
(m/z, % abundance): 678 [M - Cl]+, 24; 650 [M - Cl - C2H4]+,
100. 2f can be also prepared starting from [n-Bu4N][Re(N)Cl4], as
detailed here. To a stirred clear yellow solution of [n-Bu4N][Re(N)-
Cl4] (0.090 g, 0.154 mmol) in dichloromethane (10 mL) was added a
solution of POP (0.082 g, 0.185 mmol) in dichloromethane (10 mL).
The solution was refluxed for 90 min under N2. Its color turned yellow
and then orange. After the solvent was partially removed by a gentle
stream of di-nitrogen, a yellow solid appeared. It was filtered off,
washed with dichloromethane (2 mL), and dried under a vacuum pump
(yield 73%). The filtrate contains the meridional isomer (vide infra).
mer-[M(N)Cl2(POP)] (M ) Tc, 1m; M ) Re, 2m). Both complexes
are prepared with a similar procedure, detailed here for the technetium
complex. To a stirred clear solution of [As(C6H4)4][Tc(N)Cl4] (0.138
g, 0.216 mmol) in acetonitrile (15 mL) (or [n-Bu4N][Re(N)Cl4] in the
case of rhenium) was added a solution of POP (0.141 g, 0.318 mmol)
in acetonitrile (9 mL). Upon addition of the ligand, suddenly a clear
orange solution was observed which gave a yellow-orange suspension
within 30 min. The stirring was continued for 1 h at room temperature
(the reaction requires 40 min reflux in acetonitrile for the rhenium
derivative). The orange powder was separated from the pale green
supernatant by filtration, washed with acetonitrile (3 × 1 mL) and
diethyl ether (3 × 2 mL), and dried under a vacuum pump. 1m (yield
67%). Recrystallization from dichloromethane/methanol solutions gave
crystals suitable for X-ray diffraction analysis. Anal. Calcd for C28H28-
NOP2Cl2Tc: C, 53.69; H, 4.50; N, 2.23. Found: C, 53.90; H, 4.76; N,
2.28. IR (KBr, cm-1): 1437 (s), 1101 (s), 1072 (m) [ν(Tc(N))], 696
(s). 1H NMR (CDCl3, ppm): 2.90 (m, 4H; P-CH2), 3.75 (m, 4H;
O-CH2), 7.34-7.91 (20H, Harom). 31P NMR (CDCl3, ppm): 31.2 (bs).
TLC: SiO2 (1/2/1.5, EtOH/CHCl3/C6H6) Rf 0.92; C18 (90/10, MeCN/
H2O) Rf 0.86. 1m can be also prepared starting from [Tc(N)Cl2(PPh3)2]
Physical Measurements. Elemental analyses (C, H, N, S) were
performed on a Carlo Erba 1106 elemental analyzer. FT IR spectra
were recorded on a Nicolet 510P Fourier transform spectrometer in
the range 4000-200 cm-1 in KBr mixtures using a Spectra-Tech
diffuse-reflectance collector accessory for technetium compounds or
on a Mattson 3030 Fourier transform spectrometer in the range 4000-
400 cm-1 in KBr pellets for rhenium compounds. Proton and 31P NMR
spectra were collected on a Bruker AC-200 instrument, using SiMe4
as internal reference (1H) and 85% aqueous H3PO4 as external reference
(31P). Mass spectra of selected rhenium compounds (ca. 10-6
M
solutions) were recorded on a LCQ instrument (Finnigan, Palo Alto,
CA). The complexes have been directly injected by a syringe pump at
a flow of 5 µL/min. Thin-layer chromatography (TLC) was performed
using sheets from Riedel-de Haen (Silica gel 60 F 254).
Syntheses of Diphosphine Ligands. Bis[(2-diphenylphosphino)-
ethyl]ether (POP). In a 250 mL two-necked flask was added
n-butyllithium 1.6 M in hexane (7.2 mL, 11.5 mmol) to a THF solution
(40 mL) containing diphenylphosphine (2 mL, 11.5 mmol) under N2.
The mixture was cooled to -50 °C in an acetone/liquid di-nitrogen
slurry. 2-Bromoethyl ether (0.73 mL, 5.75 mmol) dissolved in THF (5
mL) was added dropwise over a period of 15 min using a pressure
equalizing funnel. The initial red colored solution becomes colorless
at the end of the dihalide addition. The reaction solution was left to
reach room temperature. After standing overnight, the resulting golden
solution was transferred via cannula into a separating funnel, treated
with degassed water (40 mL), and then with freshly distilled (under
N2) diethyl ether (3 × 15 mL). The collected organic phases were
rotoevaporated until a colorless waxy oil appeared, which was purified
by overnight suction in a vacuum pump (10-2 mmHg). Yield 42%. 1H
3
NMR (200 MHz, Me4Si, CDCl3): 2.31 (t, J ) 8 Hz, -CH2-PPh2,
3
4H), 3.49 (q, J ) 8 Hz, O-CH2-, 4H), 7.28-7.43 (-PPh2, 20H).
13C NMR (200 MHz, Me4Si, CDCl3): 28.7 (d, 1JCP ) 13 Hz, -CH2-
2
PPh2), 67.9 (d, JCP ) 23 Hz, O-CH2-). 31P NMR (200 MHz, 85%
H3PO4, CDCl3): - 24.1 (s).
POP ligand undergoes oxidation to the corresponding phosphineoxide
when in contact with air. However, when stored under di-nitrogen
atmosphere and kept in the refrigerator, POP is stable for months. In
any case, small amounts of oxide byproducts do not interfere in the
reaction procedures.
Bis[(2-diphenylphosphino)ethyl]methoxy-ethylamine (PNP). This
was synthesized according to the method reported by Sacconi and
1
Morassi.12 PNP ligand is a solid material indefinitely stable in air. H
NMR (200 MHz, Me4Si, CDCl3): 2.11 (m, -CH2-PPh2, 4H), 2.62
3
(m, -CH2-N(CH2-)-CH2-, 6H), 3.23 (s, -OCH3, 3H), 3.29 (t, J
) 6 Hz, -CH2-OCH3, 2H), 7.25-7.45 (-PPh2, 20H). 13C NMR (200
1
MHz, Me4Si, CDCl3): 25.1 (d, JCP ) 13 Hz, -CH2-PPh2), 50.1 (d,
2JCP ) 23 Hz, -CH2-CH2-PPh2), 52.8 (>N-CH2-CH2-OCH3), 58.8
(>N-CH2-CH2-OCH3), 70.8 (>N-CH2-CH2-OCH3), 128.4, 128.5,
(11) Akbar, A.; Livingstone, S. E.; Philips, D. J. Inorg. Chim. Acta 1973, 7,
179.
(12) Morassi, R.; Sacconi, L. J. Chem. Soc. A 1971, 492.
9
11470 J. AM. CHEM. SOC. VOL. 124, NO. 38, 2002

Electrophilic Metal Fragment [99Tc(N)(PXP)]2+
A R T I C L E S
according to the method detailed here. To a stirred pink suspension of
[Tc(N)Cl2(PPh3)2] (0.077 g, 0.108 mmol) in acetonitrile (32 mL) was
added a solution of POP (0.069 g, 0.155 mmol) in acetonitrile (10 mL).
The mixture was refluxed for 30 min. Within 5 min of reflux, the
solution became transparent, giving an orange color. After the solvent
was removed by slow evaporation, the crystalline orange solid was
washed with acetonitrile (3 × 1 mL) and then with diethyl ether (3 ×
2 mL) to afford bright orange crystals (yield 78%). In addition, 1m
can be prepared by dissolving 1f in hot acetonitrile. Standing one week
in a stoppered flask, the solution deposited bright orange crystals of
the meridional isomer.
1076 (m), 1063 (m) [ν(Tc(N))], 695 (s). 1H NMR (CDCl3, ppm): 2.39-
4.45 (12H, various CH2 groups), 3.41 (dd, 2H; S-CH2, AB), 6.82-
7.98 (20H, Harom). 31P NMR (CDCl3, ppm): 30.3 (bs). 6: Once the
yellow solid was produced, it was then dissolved in the minimum
amount of dichloromethane (ca. 3 mL) and treated with water (2 × 4
mL). The organic phases were desiccated with anhydrous MgSO4; after
filtration, slow evaporation of the solvent gave a pure yellow solid
(yield 68%). Anal. Calcd for C30H30NP2SO3Re: C, 49.17; H, 4.12; N,
1.91; S, 4.37. Found: C, 50.07; H, 4.02; N, 2.03; S, 4.85. IR (KBr,
cm-1): 1639 (vs) [ν(COO)], 1436 (s), 1283 (s), 1101 (s, Re-P), 1075
1
(m, [ν(Re(N))], 696 (s). H NMR (CDCl3, ppm): 2.65-3.45 (12H,
various CH2 groups), 3.39 (dd, 2H; S-CH2, AB), 6.85-8.00 (20H,
H
fac-[M(N)(HL2)(POP)][BF4] (M ) Tc, 7; M ) Re, 8). Both
2m. Yield 71%. Anal. Calcd for C28H28NOP2Cl2Re: C, 47.12; H,
3.95; N, 1.96. Found: C, 47.34; H, 4.00; N, 2.03. IR (KBr, cm-1):
1435 (s), 1098 (s) (Re-P), 1078 (m), 1063 (m) [ν(Re(N))], 695 (s).
1H NMR (CDCl3, ppm): 2.91 (m, 4H; P-CH2), 3.72 (m, 4H; O-CH2),
7.35-7.95 (20H, Harom). 31P NMR (CDCl3, ppm): 25.5 (s). ESI MS
(m/z, % abundance): 736 [M + Na]+, 44; 714 [MH]+, 52; 678 [M -
Cl]+, 82; 650 [M - Cl - C2H4]+, 100. 2m can be also prepared by
dissolving 2f in hot acetonitrile. Standing one week in a stoppered flask,
the solution deposited bright orange crystals of the meridional isomer.
fac-[Tc(N)Cl2(PNP)], 3f. To a stirred pink suspension of [Tc(N)-
Cl2(PPh3)2] (0.110 g, 0.155 mmol) in dichloromethane (50 mL) was
added a solution of PNP (0.097 g, 0.194 mmol) in dichloromethane
(10 mL). The mixture was refluxed for 40 min. After 5 min of reflux,
the mixture became a transparent yellow solution. Removal of the
solvent by slow evaporation gave a glassy yellow solid which was
vigorously stirred in diethyl ether and washed with the same solvent
(3 × 5 mL) to afford a pure yellow powder (yield 94%). Anal. Calcd
for C31H35N2OP2Cl2Tc: C, 54.47; H, 5.16; N, 4.10. Found: C, 54.65;
H, 5.22; N, 4.35. IR (KBr, cm-1): 1435 (s), 1103 (s, Tc-P), 1055 (m)
[ν(Tc(N))], 696 (s). 1H NMR (CD2Cl2, ppm): 2.53-3.52 (12H, various
CH2 groups), 3.24 (s, 3H, OCH3), 6.85-7.95 (20H, Harom). 31P NMR
(CDCl3, ppm): 35.5 (bs).
arom). 31P NMR (CDCl3, ppm): 21.0 (s).
complexes are prepared as detailed for the technetium complex. To a
stirred yellow solution of 1f (0.055 g, 0.087 mmol) in dichloromethane
(25 mL) at reflux was added a solution of S-methyl 2-methyldithio-
carbazate (H2L2) (0.119 g, 0.877 mmol) in ethanol (5 mL). After 10
min, an excess of triethylamine (87 µL, 0.653 mmol) was added, and
the reflux continued for 1 h. No color change was observed throughout
the reaction. After the solvent was removed by slow evaporation, the
resulting yellow oil was dissolved in methanol (10 mL), filtered, and
the filtrate was treated with an excess of NaBF4 (0.059 g, 0.546 mmol)
dissolved in methanol (5 mL). The solution was stirred overnight at
room temperature, and then the solvent was removed by slow
evaporation. The resulting yellow oily solid was dissolved in dichlo-
romethane (5 mL), and unreacted NaBF4 was filtered off. After the
solvent was removed, the yellow solid was dissolved in acetone; slow
evaporation of the solvent gave yellow microcrystals which were
washed with diethyl ether (2 × 3 mL) and dried overnight under a
vacuum pump (yield 73%). Recrystallization from ethanol/diethyl ether
gave suitable crystals for X-ray diffraction analysis. Anal. Calcd for
C31H35N3OS2P2TcBF4: C, 47.88; H, 4.53; N, 5.40; S, 8.25. Found: C,
48.55; H, 4.12; N, 5.40; S, 8.00. IR (KBr, cm-1): 1437 (s), 1059 (s,
BF4), 696 (s). 1H NMR (CD2Cl2, ppm): 2.48-4.15 (8H; P-CH2-
CH2-O), 2.65 (s, 3H; S-CH3), 3.48 (s, 3H; N-CH3), 5.17 (s, 1H;
N-H), 6.91-8.01 (20H, Harom). 31P NMR (CDCl3, ppm): 29.7 (bs),
23.7 (bs). 8 (yield 80%). Anal. Calcd for C31H35N3OS2P2ReBF4: C,
43.05; H, 4.08; N, 4.86; S, 7.41. Found: C, 43.54; H, 4.23; N, 5.01; S,
7.93. IR (KBr, cm-1): 1436 (s), 1100 (s, Re-P), 1062 (vs, BF4), 1044
fac-[Re(N)Cl2(PNP)], 4f. To a stirred yellow solution of [n-Bu4N]-
[Re(N)Cl4] (0.080 g, 0.138 mmol) in dichloromethane (10 mL) was
added a solution of PNP (0.060 g, 0.180 mmol) in dichloromethane (5
mL). During the addition of the ligand, the solution turned red-brown.
The solution was refluxed for 30 min, and then it was cooled to room
temperature and concentrated by rotoevaporation to ca. 3 mL. A white
solid was filtered off. Addition of methanol (15 mL) to the filtrate gave
a yellow solid which was filtered off and washed with diethyl ether (3
mL). It was then dried overnight under a vacuum pump (yield 46%).
Slow diffusion of ethanol into a dichloromethane solution afforded
yellow crystals suitable for X-ray diffraction analysis. Anal. Calcd for
C31H35N2OP2Cl2Re: C, 48.31; H, 4.58; N, 3.63. Found: C, 49.27; H,
4.45; N, 3.66. IR (KBr, cm-1): 1435 (s), 1114 (m), 1101 (s, Re-P),
1063 (m) [ν(Re(N))], 692 (s). 1H NMR (CDCl3, ppm): 2.65-3.45 (12H,
1
(s, BF4), 699 (s). H NMR (CDCl3, ppm): 2.55-4.11 (8H; P-CH2-
CH2-O), 2.69 (s, 3H; S-CH3), 3.55 (s, 3H; N-CH3), 6.85-8.05 (20H,
H
arom). 31P NMR (CDCl3, ppm): 20.6 (d), 11.7 (d). ESI MS (m/z, %
abundance): 778 [MH]+, 100.
fac-[M(N)(L1)(PNP)] (M ) Tc, 9; M ) Re, 10). Both complexes
are prepared as detailed for the technetium complex. To a bright yellow
solution of 3f (0.031 g, 0.045 mmol) dissolved in a dichloromethane/
ethanol mixture (1:1, 10 mL) were added neat triethylamine (10 µL,
0.065 mmol) and 2-mercaptoacetic acid sodium salt (NaHL1) (0.021
g, 0.182 mmol) under stirring at room temperature. The solution was
refluxed for 2 h and, after cooling, was submitted to a gentle stream of
N2 until a yellow precipitate appeared. The solid was filtered off and
washed with water (3 mL), ethanol (5 mL), and diethyl ether (2 × 5
mL), and dried under vacuum (yield 69%). Anal. Calcd for C33H37N2P2-
SO3Tc: C, 56.40; H, 5.31; N, 3.99; S, 4.56. Found: C, 55.97; H, 5.44;
N, 3.78; S, 4.34. IR (KBr, cm-1): 1628 (vs, ν(COO)), 1437 (s), 1103
(s, Tc-P), 1063 (m) [ν(Tc(N))], 698 (s). 1H NMR (CDCl3, ppm):
2.33-3.52 (12H, various CH2 groups), 3.21 (s, 3H, OCH3), 3.46 (dd,
2H; S-CH2, AB), 6.85-7.93 (20H, Harom). 31P NMR (CDCl3, ppm):
31.2 (bs), 25.5 (bs). 10: In the case of rhenium, after reflux and cooling,
removal of dichloromethane by a gentle stream of nitrogen gave a bright
yellow precipitate (unreacted starting material) which was filtered off.
The solution was taken to dryness, and the resulting powder was
dissolved in dichloromethane (10 mL) and treated with water (3 × 5
mL). The organic phase was desiccated with anhydrous MgSO4; after
filtration, addition of diethyl ether (10 mL) gave a pale yellow
various CH2 groups), 3.19 (s, 3H, OCH3), 6.85-8.00 (20H, Harom). 13
C
NMR (CDCl3, ppm): 25.2, 50.0, 52.8, 58.7, 70.7, 128.3-138.5 (Carom).
31P NMR (CDCl3, ppm): 16.4 (s). ESI MS (m/z, % abundance): 771
[MH]+, 67; 735 [M - Cl]+, 100; 703 [M - OMe]+, 36.
fac-[M(N)(L1)(POP)] (M ) Tc, 5; M ) Re, 6). Both complexes
are prepared with a similar procedure, detailed here for the technetium
complex. To a yellow solution of 1f (0.038 g, 0.060 mmol) in
acetonitrile (8 mL) was added solid 2-mercaptoacetic acid sodium salt
(NaHL1) (0.010 g, 0.090 mmol) under stirring at room temperature.
The mixture was refluxed for 30 min (90 min reflux in the case of
rhenium; an equivalent amount of neat triethylamine helps to speed up
the reaction) until a yellow precipitate appeared. The mixture was halved
in volume, and a solid was filtered off and washed with dichloromethane
(5 mL). Unreacted ligand remained on the filter, and the yellow filtrate
was left to evaporate slowly. A yellow solid deposited with time. After
filtration, the product was washed with a few drops of acetonitrile and
diethyl ether. 5 (yield 73%). Anal. Calcd for C30H30NP2SO3Tc: C,
55.81; H, 4.68; N, 2.17; S, 4.96. Found: C, 56.12; H, 4.42; N, 2.09; S,
4.67. IR (KBr, cm-1): 1622 (vs, ν(COO)), 1438 (s), 1099 (s, Tc-P),
9
J. AM. CHEM. SOC. VOL. 124, NO. 38, 2002 11471

A R T I C L E S
Bolzati et al.
Table 1. Crystallographic Data for Compounds 2m, 4f, 7, 8, and 11
2m
4f
7‚EtOH
8
11
formula
fw
C28H28Cl2NOP2Re
713.5
C31H35Cl2N2OP2Re
770.6
C33H40BF4N3O2P2S2Tc
821.5
C31H35BF4N3OP2S2Re
864.7
C34H43BF4N4OP2S2Tc
834.6
cryst syst
space group
a, Å
b, Å
c, Å
orthorhombic
Pbca (No. 61)
20.048(4)
13.737(3)
20.285(4)
monoclinic
Cc (No. 9)
13.365(3)
18.259(4)
14.294(3)
116.63(3)
3118(1)
4
1.642
0.71073
42.0
23
0.027
monoclinic
P21/c (No. 14)
16.405(3)
15.042(3)
15.228(3)
97.39(3)
3726(1)
4
1.464
0.71073
6.4
23
0.044
monoclinic
P21/c (No. 14)
15.065(3)
14.705(3)
16.318(3)
108.31(3)
3432(1)
4
1.674
0.71073
38.1
23
0.058
monoclinic
P21/c (No. 14)
11.389(2)
11.056(2)
29.871(6)
93.10(3)
3756(1)
4
1.476
0.71073
6.3
23
0.094
â, deg
V, Å3
Z
5586(2)
8
1.697
0.71073
46.8
23
0.064
0.161
D
calcd, g cm-3
λ(Mo KR), Å
µ(Mo KR), cm-1
T, °C
Ra
Rw
b
0.063
0.110
0.158
0,198
2
2
2
a ∑||Fo| - |Fc||/∑|Fo|. b {∑[w(Fo - Fc )2]/∑[w(Fo )2]}1/2
.
precipitate, which was filtered off, washed with diethyl ether, and dried
under vacuum (yield 69%). Anal. Calcd for C33H37N2P2SO3Re: C,
50.17; H, 4.72; N, 3.54; S, 4.06. Found: C, 51.17; H, 4.52; N, 3.71; S,
3.87. IR (KBr, cm-1): 1636 (vs, ν(COO)), 1436 (s), 1101 (s, Re-P),
1062 (m) [ν(Re(N))], 697 (s). 1H NMR (CDCl3, ppm): 2.31-3.42 (12H,
various CH2 groups), 3.21 (s, 3H, OCH3), 3.52 (dd, 2H; S-CH2, AB),
6.83-7.98 (20H, Harom). 31P NMR (CDCl3, ppm): 18.0 (d), 15.1 (d).
ESI MS (m/z, % abundance): 831 [M + Na]+, 58; 791 [MH]+, 100.
Anal. Calcd for C36H45N3P2S2ORePF6: C, 43.54; H, 4.57; N, 4.23; S,
6.45. Found: C, 43.72; H, 4.51; N, 4.41; S, 6.63. IR (KBr, cm-1):
1434 (s), 1100 (s, Re-P), 1061 (m) [ν(Re(N))], 840 (vs, PF6), 693 (s).
1H NMR (CDCl3, ppm): 1.19 (m, 6H; N-CH2-CH3), 2.47-3.50 (15H,
various CH2 groups + -OCH3), 3.69 (m, 4H; N-CH2-CH3), 7.01-
8.05 (20H, Harom). 31P NMR (CDCl3, ppm): 21.6 (s), -144.0 (septet,
PF6).
fac-[Re(N)(L4)(PNP)], 14. To a yellow solution containing 4f (0.150
g, 0.194 mmol) and triethylamine (54 µL, 0.388 mmol) dissolved in a
dichloromethane/methanol mixture (7/3, 30 mL) was added H2L4 (0.063
g, 0.388 mmol) dissolved in methanol (10 mL) under stirring at room
temperature. The mixture was refluxed for 2 h while the color was
gradually changing from bright to pale yellow. The solution was left
to reach room temperature, and dichloromethane was taken off by a
gentle stream of N2. A small amount of a yellow solid (unreacted
starting material) was filtered off, and the filtrate was taken to dryness
by rotoevaporation. The residue was redissolved in dichloromethane
(10 mL) and treated with water (3 × 5 mL). The organic phase was
anhydrified over MgSO4 and then treated with diethyl ether (20 mL).
A pale yellow solid was collected by filtration (yield 86%). Anal. Calcd
for C36H42N3P2O4SRe: C, 50.16; H, 5.03; N, 4.87. Found: C, 50.74;
H, 4.81; N, 4.79. TLC chromatography revealed the presence of two
products (Rf of 0.6 and 0.78, respectively), which are separated by
fac-[M(N)(HL2)(PNP)][BF4] (M ) Tc, 11; M ) Re, 12). Both
complexes are prepared as detailed for the technetium complex. A
solution of H2L2 (0.056 g, 0.412 mmol) in methanol (5 mL) was added
to a stirred yellow suspension of 3f (0.047 g, 0.068 mmol) in methanol
(15 mL). The mixture was refluxed for 30 min (2 h in the case of
rhenium) and then filtered. The filtrate was treated with an excess of
NaBF4 (0.060 g, 0.550 mmol) dissolved in methanol (5 mL) under
stirring at room temperature for an additional 30 min. The solvent was
then removed by slow evaporation, and the solid was dissolved in
dichloromethane (10 mL). After filtration, the solvent was removed
by a gentle N2 stream, and the resulting glassy yellow solid was
redissolved in methanol (5 mL). A yellow crystalline material was
deposited after slow evaporation of the solvent, washed with a 1:2
methanol/diethyl ether mixture (4 × 2 mL) and diethyl ether (3 × 2
mL), and dried under a vacuum pump (yield 77%). Anal. Calcd for
C34H42N4OP2S2TcBF4: C, 48.93; H, 5.07; N, 6.71; S, 7.68. Found: C,
48.99; H, 5.23; N, 6.31; S, 7.11. IR (KBr, cm-1): 1437 (s), 1066 (s,
BF4), 698 (s). 1H NMR (CDCl3, ppm): 1.73-4.20 (12H, various CH2
groups), 2.61 (s, 3H; SCH3), 3.20 (s, 3H; NCH3), 3.55 (s, 3H, OCH3),
6.89-7.96 (21H, Harom + N-H). 31P NMR (CDCl3, ppm): 28.3 (bs),
23.0 (bs). TLC: SiO2 (1/2/1.5, EtOH/CHCl3/C6H6) Rf 0.21; C18 (90/
10, MeCN/H2O) Rf 0.47. 12 (yield 80%). Anal. Calcd for C34H42N4P2S2-
OReBF4: C, 44.29; H, 4.59; N, 6.08; S, 6.98. Found: C, 44.01; H,
4.54; N, 6.20; S, 6.58. IR (KBr, cm-1): 1437 (s), 1102 (s, Re-P),
1
column chromatography on silica. H NMR (CDCl3, ppm): Isomer I,
2.25 (s, 3H; -C(O)-CH3), 3.24 (s, 3H; -OCH3), 2.45-3.52 (14H;
various CH2 groups), 5.01 (m, 1H; S-CH2-CH(NHCOCH3)), 6.82-
7.75 (20H, -PPh2), 8.10 (d, 1H; S-CH2-CH(NHCOCH3)). Isomer
II: 2.04 (s, 3H; -C(O)-CH3), 3.22 (s, 3H; -OCH3), 2.23-3.50 (14H;
various CH2 groups), 5.05 (m, 1H; S-CH2-CH(NHCOCH3)), 6.81
(d, 1H; S-CH2-CH(NHCOCH3)), 6.83-7.90 (20H, -PPh2). 31P NMR
(CDCl3, ppm): Isomer I, 14.6 (d), 18.3 (d). Isomer II: 15.4 (d), 17.0
(d).
X-ray Crystallography. All of the measurements for the five
complexes were made at ambient temperature on a Nicolet Siemens
R3m/V diffractometer with graphite-monochromated Mo KR radiation.
Crystallographic data are summarized in Table 1. The structures were
solved by heavy-atom methods, expanded using Fourier techniques,
and refined by full-matrix least-squares method on F 2.13 For 2m, 4,
and 8, the non-hydrogen atoms were refined anisotropically. The H
atoms were placed in idealized positions and allowed to ride with the
parent atoms to which each was bonded. In 7, the atoms of the solvent
EtOH molecule, affected by high thermal motion, were refined
isotropically. The diffracting ability of 11 fell off rapidly with increasing
Bragg angle, and much of the higher angle data collected were flagged
1
1062 (vs, BF4), 700 (s). H NMR (CDCl3, ppm): 1.60-3.58 (12H,
various CH2 groups), 2.62 (s, 3H; SCH3), 3.19 (s, 3H; NCH3), 3.60 (s,
3H, OCH3), 6.14 (d, 1H; N-H), 6.85-8.01 (20H, Harom). 31P NMR
(CDCl3, ppm): 18.7 (d), 11.2 (d). ESI MS (m/z, % abundance): 835
[MH]+, 100.
fac-[Re(N)(L3)(PNP)][PF6], 13. To a yellow solution of 4f (0.050
g, 0.065 mmol) in dichloromethane (10 mL) at reflux was added NaL3
(0.015 g, 0.065 mmol) dissolved in methanol (3 mL) under stirring.
The color of the solution immediately changed from yellow to orange.
After 2 h, the solution was left to reach room temperature, and then it
was taken to dryness by rotoevaporation. The residue was dissolved in
few milliliters of methanol, and NH4PF6 (0.045 g, 0.130 mmol)
dissolved in methanol (3 mL) was added. Addition of diethyl ether
(20 mL) yielded an orange powder which was filtered off, washed with
diethyl ether (2 × 5 mL), and dried under a vacuum pump (yield 70%).
(13) Sheldrick, G. M. SHELXTL/NT, Version 5.10; Bruker AXS Inc.: Madison,
WI, 1999. Sheldrick, G. M. SHELXL-97, Program for crystal structure
refinement; University of Gottingen: Germany, 1997.
9
11472 J. AM. CHEM. SOC. VOL. 124, NO. 38, 2002

Electrophilic Metal Fragment [99Tc(N)(PXP)]2+
A R T I C L E S
Table 2. Selected Bond Distances (Å) and Angles (deg) for 2m,
4f, 7‚EtOH, 8, and 11
2m
4f
2m
4f
Re-P(1) 2.434(4) 2.391(3) P(1)-Re-P(2)
152.7(1)
98.6(1)
84.1(1)
94.6(3)
91.4(3)
Re-P(2) 2.447(4) 2.394(2) Cl(1)-Re-Cl(2) 157.7(1)
Re-Cl(1) 2.412(4) 2.422(3) N(1)-Re-P(1)
Re-Cl(2) 2.424(4) 2.446(2) N(1)-Re-P(2)
103.2(5)
104.1(5)
Re-N(1) 1.66(1)
1.666(7) N(1)-Re-Cl(1) 100.6(5) 107.0(3)
Re-O
2.505(9)
N(1)-Re-Cl(2) 101.8(5) 106.4(3)
Re-N(2)
2.638(9) N(1)-Re-O
N(1)-Re-N(2)
180.0(7)
164.2(7)
7‚EtOH (M ) Tc) 8 (M ) Re)
7 (M ) Tc) 8 (M ) Re)
M-P(1)
M-P(2)
M-S(1)
M-N(1)
M-N(2)
M-O(1)
2.432(1)
2.440(1)
2.388(1)
1.623(4)
2.038(3)
2.758(5)
2.566(6) P(1)-M-P(2)
98.2(1) 102.2(2)
2.425(5) S(1)-M-N(2) 78.7(1) 83.8(4)
2.350(5) N(1)-M-P(1) 95.7(1) 101.0(6)
1.56(1) N(1)-M-P(2) 96.3(1) 96.0(6)
2.04(1) N(1)-M-S(1) 107.7(1) 104.9(6)
2.56(1) N(1)-M-N(2) 110.2(2) 102(1)
N(1)-M-O(1) 159.7(2) 162(1)
Figure 1. Schematic representations of the four basic arrangements
considered for theoretical calculations.
interaction, which may involve geometry and possibly electronic state
modifications with respect to the isolated species.
To establish a closer contact with experiment, we define a complex
formation energy as
7‚EtOH
11
11
Tc-P(1)
Tc-P(2)
Tc-S(1)
Tc-N(1)
Tc-N(3)
Tc-N(2)
2.432(5)
2.440(5)
2.385(5)
1.61(1)
2.06(1)
2.81(1)
P(1)-Tc-P(2)
S(1)-Tc-N(3)
N(1)-Tc-P(1)
N(1)-Tc-P(2)
N(1)-Tc-S(1)
N(1)-Tc-N(3)
N(1)-Tc-N(2)
97.9(2)
79.8(4)
97.2(5)
93.0(4)
104.7(4)
110.0(6)
161.8(8)
[Tc(N)Cl4]2- + PXP f [Tc(N)Cl2(PXP)] + 2Cl- + ∆Ef (2)
This reaction is directly related to the actual preparation of the
complexes (vide infra), and it can be decomposed in the following
processes:
Tc(N)Cl2 + 2Cl- f [Tc(N)Cl4]2- + ∆E1
Tc(N)Cl2 + PXP f [Tc(N)Cl2(PXP)] + ∆E2
(3)
(4)
as weak and bore negative intensity. In addition, during the refinement
it became apparent that the BF4 group was disordered, and only the
heavy atoms were refined anisotropically. Features of the final
difference maps showed peaks in chemically unreasonable positions
which were considered to be noise. Selected bond lengths and angles
are given in Table 2.
-
These involve energy variations ∆E1 and ∆E2, where the former is
common to all of the complexes, and the latter is specific of each of
them. Obviously, the equivalence ∆Ef ) ∆E2 - ∆E1 holds. All ETS
analyses will be referred to process 4.
Computational Details. All of the calculations were performed by
running ADF 2.3.0,14 a program package based on the density functional
theory (DFT). Calculations were run within the generalized gradient
approximation (GGA), with the inclusion of the Becke’s15 nonlocal
exchange correction and of the Lee-Yang-Parr16 corrections for
correlation. An uncontracted triple-ú-STO basis set was used for the
4s, 4p, 4d, 5s, and 5p shells of Tc. For all of the other elements, a
double-ú basis set augmented by one polarization function was used,
that is, 1s, 2p for H; 2s, 2p, 3d for C, O, and N; 3s, 3p, 3d for P and
S. Electrons in lower shells were included in the core and kept frozen
during the SCF procedure. We also carried out some test calculations
where triple-ú sets for all of the elements were adopted; differences in
formation energies (vide infra) changed by no more than 0.4 kcal/mol.
Bond energies (BE) were analyzed by using the extended transition
state (ETS) method of Ziegler.17 This method considers the molecule
to be formed of fragments, whose BE is decomposed as follows:
Convenient reference energies should be given for the Tc(N)Cl2 and
PXP moieties. For the former, the choice is irrelevant because it is just
an intermediate, so we simply take the most stable of the geometries
assumed in the examined complexes. In contrast, more caution should
be observed when choosing the reference state for PXP. Although all
of the PXP species here considered are structurally similar, their
conformational flexibility coupled to the different stereochemical
demands of the X substituents is likely to originate different equilibrium
conformers. Thus, we decided to explore the conformational space of
the free PXP ligands with Cerius 2 software.18 The structures corre-
sponding to the five most stable conformers obtained in this way were
used as input geometries for new DFT optimizations. The most stable
of the resulting structures was finally taken as a reference state for
each PXP fragment. While this procedure does not warrant that global
minima for the free ligands are actually found, it certainly allows one
to keep errors in ∆Ef within negligible (<1 kcal mol-1) levels. Our
calculations were carried out on both the fac and the mer pseudooc-
tahedral isomers of each compound. In the fac case, both asymmetric
(C1) and planar-symmetric (indicated as “fac(Cs)” in Figure 1) structures
were considered. In addition, five-coordinated structures where the
heteroatom X is detached from Tc (“cis”) were performed. Furthermore,
we considered, besides the Ph-substituted PXP ligands, also simplified
models, where phenyl rings are replaced by hydrogens, hereafter PXPH.
This seemingly rough approximation is actually justified because H
and Ph have similar electron donor/acceptor properties and are not
directly involved in the metal-ligand bonds. This choice is attractive
because it allows one to better understand the role played by phenyl
substituents on the complex stability and because it allows one to
∆E ) ∆Esteric + ∆Eoi + ∆Eprep
(1)
where a negatiVe ∆E means a positiVe BE. ∆Esteric and ∆Eoi are,
respectively, the change in steric and in orbital interaction energy
occurring when the fragments are assembled to form the molecule.
∆Eprep is the energy required to “prepare” the fragments for the
(14) (a) Amsterdam Density Functional Package, Version 2.3.0; Vrije Univer-
siteit: Amsterdam, The Netherlands, 1997. (b) Baerends, E. J.; Ellis, D.
E.; Ros, P. Chem. Phys. 1973, 2, 41. (c) Te Velde, G.; Baerends, E. J. J.
Comput. Phys. 1992, 99, 84. (d) Fonseca Guerra, C.; Visser, O.; Snijders,
J. G.; Baerends, E. J. In Methods and Techniques in Computational
Chemistry; Clementi, E., Corongiu, G., Eds.; STEF: Cagliari, 1995; Chapter
8, p 305.
(15) Becke, A. D. Phys. ReV. A 1988, 38, 3098.
(16) Lee, C.; Yang W.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
(17) Ziegler, T.; Rauk, A. Theor. Chim. Acta 1977, 46, 1.
(18) Conformer Search and Conformer Analysis, in Cerius 2, version 4.0;
Molecular Simulation Inc., 1999.
9
J. AM. CHEM. SOC. VOL. 124, NO. 38, 2002 11473

A R T I C L E S
Bolzati et al.
quickly examine the relative stability and the main bonding features
for a large number of complexes, limiting the time-consuming
calculations required by Ph-substituted ligands only to the most
interesting cases.
In all of the calculations herein reported, we also replaced the pendant
methoxyethyl group of the PNP ligand with a simpler methyl group.
We checked that this substitution does not affect significantly the
energetics on PNPH complexes.
Basis set superposition errors (BSSE) were computed in selected
cases by the counterpoise method. We found that its effect is negligible
in the case of PXPH ligands, yielding a uniform ∼3 kcal/mol increase
of the complex formation energy. On the contrary, we found that BSSE
plays a role when Ph-substituted PXP ligands are concerned (see below).
We adopted a nonrelativistic Hamiltonian in our calculations. We found
that inclusion of relativistic effects can change by ∼1 kcal/mol the
relative stability of the isomers of a given [Tc(N)Cl2(PXP)] complex.
teranion. ESI mass spectra of [Re(N)Cl2(PXP)] intermediate
complexes show the peaks corresponding to the molecular ion
and fragmentations consistent with the loss of one chloride
group. Complexes containing the PNP ligand show an additional
peak corresponding to the fragmentation of the terminal methoxy
group incorporated in the pendant amine function.
31P NMR spectra of all rhenium complexes display signals
in agreement with diphosphine coordination, the pertinent peaks
moving downfield in the 16.4-35.5 ppm positive region from
the -24.1 and -21.5 ppm values exhibited by uncoordinated
POP and PNP ligands, respectively. Despite the diamagnetism
exhibited by all of these asymmetrical nitrido-M(V) complexes,
in accordance with the low-spin d2 configuration typical of both
distorted octahedral or distorted square-pyramidal species bear-
ing a multiple oxo- or nitrido-core,4 31P NMR signals of
technetium complexes show very broad profiles at ambient
temperature, which become narrower somewhat on lowering
the temperature. Such behavior was previously attributed to the
coupling of the 31P nuclei with the quadrupolar 99Tc center.19
The singlet shown by monosubstituted complexes 1-4 indicates
magnetic equivalence of the diphosphine P donors, in agreement
with the symmetry exhibited by these molecules (vide infra).
On the contrary, asymmetrical heterocomplexes 5-12 and 14
display a two doublets pattern, according to the magnetic
unequivalence of the two P nuclei. In fact, each P faces a
different trans-substituent (S and O in the case of L1 and L4,
and S and N in the case of HL2 chelates). In the rhenium
Results
Synthesis. POP and PNP ligands were prepared by alkylation
of the diphenylphosphine lithium salt with the appropriate
dihalides 2-bromoethyl ether and bis(2-chloroethyl)(methoxy-
ethyl)amine, respectively.
Reactions of the labile precursors [M(N)Cl4]- and [M(N)-
Cl2(PPh3)2] (M ) Tc, Re) with a slight excess of POP in
refluxing dichloromethane yielded bright yellow monosubsti-
tuted fac-[M(N)Cl2(POP)] complexes (1f, 2f) in high yields.
Facial isomers were sometimes contaminated by an orange
impurity (then characterized as the meridional isomer), which
was more soluble in chlorinated solvents. Similar ligand-
exchange reactions carried out in high boiling solvents afforded
the bright orange-red mer-[M(N)Cl2(POP)] derivatives (1m,
2m). Facial isomers could be quantitatively converted into the
meridional isomers in hot acetonitrile solutions in a few hours.
The conversion was operating also at room temperature, and in
chlorinated solvents, but at a reduced rate. Facial intermediates
reacted further with an excess of mercaptoacetic acid (NaHL1)
or S-methyl 2-methyldithiocarbazate (H2L2) in refluxing dichlo-
romethane/alcohol mixtures giving the pale-yellow asymmetric
heterocomplexes fac-[M(N)(L1)(POP)] (5, 6) and fac-[M(N)-
(HL2)(POP)]+ (7, 8). Under the synthetic conditions utilized,
no asymmetrical bis-substituted species of the type [M(N)(Ln)2]
or [M(N)(POP)2] were collected.
When a slight excess of PNP reacted with [M(N)Cl4]- or
[M(N)Cl2(PPh3)2], in refluxing dichloromethane solutions, only
fac-[M(N)Cl2(PNP)] complexes (3f, 4f) were separated by
precipitation with methanol. Repeated attempts to obtain the
meridional isomers by changing the reaction conditions (high
boiling solvents, prolonged time) always failed. The PNP-facial
isomers reacted with bidentate NaHL1, H2L2, NaL3, and H2L4
ligands to give neutral fac-[M(N)(L1)(PNP)] (9, 10) and fac-
[Re(N)(L4)(PNP)] (14), and charged fac-[M(N)(HL2)(PNP)]+
(11, 12) and fac-[Re(N)(L3)(PNP)] (13) compounds.
2
heterocomplexes, the magnitude of the JPP coupling constant
(ca. 12 Hz) further supports a cis-P coordination, confirmed in
the solid state by X-ray diffraction analyses (vide infra).
The facial/meridional isomerism of the monosubstituted 2f/
2m species is evidenced by NMR spectroscopy. By dissolving
the facial isomer 2f in dichloromethane, we observed four
distinct multiplets for the POP-methylene protons according to
their diastereotopic nature, while the phenyl protons are spread
over a wide region (6.85-8.00 ppm). In fact, both phenyl
substituents at P and geminal protons of the POP chain are
positioned close or remote with respect to the terminal nitrido
group, thereby experiencing different magnetic environments.
With the time, the signal pattern becomes less complicated: (i)
the 31P singlet moves from 16.9 to 25.5 ppm, (ii) only two
multiplets are observed in the methylene region, and (iii) the
aromatic protons fall in a restricted window (7.35-7.95 ppm).
All of these data are consistent with the rearrangement of the
facial into the more symmetrical meridional isomer. A similar
behavior is exhibited by the technetium analogues 1f and 1m,
with the rate of conversion much faster than that in the case of
rhenium.
[M(N)Cl2(PNP)] complexes 3f and 4f show proton and
phosphorus NMR spectra consistent with a molecular structure
similar to that exhibited by 1f and 2f. Hence, a facial arrange-
ment of the PNP ligand is postulated and then confirmed by
X-ray diffraction analysis. However, no isomerization into the
meridional form is observed under the conditions above-
described for the POP-analogues.
Theoretical Calculations. Aiming at the understanding of
such a different behavior, we have performed DFT theoretical
studies on a series of monosubstituted [Tc(N)Cl2(PXP)] com-
plexes. We first carried out calculations on [Tc(N)Cl2(PXPH)]
Characterization. Elemental analyses, as shown in the
Experimental Section, are in good agreement with the proposed
formulation. In the IR spectra, all of the nitrido-M(V) complexes
exhibit absorptions typical of the coordinated aryl-diphosphine
ligands at ca. 1100 and 700 cm-1, as well as medium bands in
the 1075-1052 cm-1 region attributable to ν[M(N)]. Additional
vibrations in the region 1639-1622 cm-1, characteristic of the
mercaptoacetato group, appear in complexes 5, 6, 9, and 10,
whereas charged compounds 7, 8, 11, and 12 show intense bands
around 1050 cm-1 characteristic of the tetrafluoroborato coun-
(19) Abram, U.; Lorenz, B.; Kaden, L.; Scheller, D. Polyhedron 1988, 7, 285.
9
11474 J. AM. CHEM. SOC. VOL. 124, NO. 38, 2002

Electrophilic Metal Fragment [99Tc(N)(PXP)]2+
A R T I C L E S
Figure 3. Contour plot of the wave function associated to the HOMO of
fac-[Tc(N)Cl2(PXPH)] in the (approximate) complex basal plane. Contour
levels are (0.01, (0.02, (0.04, and (0.08 au.
Table 3. Theoretical Formation Energiesa and Relevant ETS
Energy Termsb (kcal mol-1) for Complexes of PNP and POP
Ligands
∆E
prep
∆E
oi
f
PXP Tc(N)Cl2 ∆E
a′
a′′
total
∆E
steric
fac (Cs)
fac
PNP 11.0
POP 12.3
2.8
1.9
46.5 -60.7 -70.9 -132.2 -24.2
45.7 -56.1 -69.5 -126.2 -18.7
PNP 11.4
POP 13.0
2.2
1.3
49.9
50.0
-136.6 -25.5
-129.2 -18.7
mer
PNP 7.8 12.8
POP 7.5 9.6
39.9 -88.2 -43.9 -133.2 -25.7
42.0 -85.4 -44.3 -130.9 -24.6
Figure 2. One-electron energy levels for [Tc(N)Cl2(PXPH)] model
complexes. The dotted boxes indicate the energy interval spanned by the
Cl-based combinations. Note the different orientation of the molecules with
respect to the axis framework.
a Refer to eq 2. b Refer to eq 4.
orbitals, whose lobes are pointing away from P or Cl atoms
and are consequently suitable for a nucleophilic attack. In
contrast, all of the low-energy empty combinations of the mer
isomers lie in planes containing P and/or Cl atoms and are in
consequence less accessible to incoming nucleophiles. Thus, a
higher reactivity is expected for fac isomers.
model compounds. Frontier one-electron energy levels for the
complexes of PNPH and POPH ligands are reported in Figure 2
for both the fac(Cs) and the mer structures. These levels
correspond to the Tc d orbitals, the six p combinations of the
chlorine ligands (as a whole indicated by a dashed box), the
nitride p orbitals (which give rise to a σ lone pair, slightly Tc-N
bonding,20 and to two π Tc-N bonds), the lone pair(s) of the
diphosphine heteroatom X (N or O) involved in a X f Tc dative
bond, and the n-/n+ lone pair P combinations which give rise
to the Tc-P bonds. We point out that the splitting of the
antibonding d levels is quite similar to that computed for
[Tc(N)Cl4]2- by Baldas and co-workers,21 and it does not change
significantly in the examined complexes, suggesting that the
origin of the different stereochemistry of PNP and POP
complexes is not straightforwardly related to the metal-ligand
interaction. Useful insight into the reactivity can be gained by
a closer look at the frontier region, where metal-based anti-
bonding combinations are present. In all of the cases, the HOMO
corresponds to a metal-ligand π* combination in the complex
basal plane. However, because of the different orientation of
the molecules with respect to the axis framework (see Figure
2), it is labeled as dx2-y2 and dxy for the fac and the mer isomers,
respectively. We also point out (see Figure 3) that the HOMO
has a large participation of the Cl p states, whereas no significant
contribution from the P orbitals is present. This means that the
Tc-Cl bonds should be more labile with respect to the Tc-P
ones. It is also interesting to examine the lowest unoccupied
MOs. For the fac isomers, they are based on the dxz and dyz Tc
Complex formation energies, as well as ETS analyses of the
BE for [Tc(N)Cl2(PXPH)] (PXP ) PNP, POP, P-P), in several
geometrical arrangements were computed and are available as
Supporting Information. These data allow one to extract an
estimate for the X-Tc bond strength by comparing the a′ orbital
energy terms of the PXPH complexes with the value pertinent
to the PPH one. Resulting differences (7 and 3 kcal mol-1 for
PNPH and POPH, respectively, for both the fac(Cs) and the mer
coordinations) are in fair agreement with total energy variations
(6 and 1 kcal mol-1) computed for the fac f cis processes,
where the X-Tc bond is disrupted. Furthermore, the increase
of the ∆Eoi observed when passing from the Cs to the C1 fac
structure, which is particularly evident for [Tc(N)Cl2(PNPH)],
suggests that the Cs f C1 distortion of this complex should be,
at least in part, driven by a strengthening of the Tc‚‚‚N
interaction. This is in tune with the increase of the Mulliken
overlap population (from 0.153 to 0.176e) and with the decrease
in the interatomic distance (from 2.993 to 2.811 Å) observed
when passing from the symmetric to the asymmetric structure.22
We turn now to examine the results of calculations performed
on complexes of phenyl-substituted PXP ligands. Formation
energies reported in Table 3 show that the mer-[Tc(N)Cl2(POP)]
complex is by far favored over the fac species. The mer form
(20) Wheeler, R. A.; Whangbo, M.-H.; Hughbanks, T.; Hoffmann, R.; Burdett,
J. K.; Albright, T. A. J. Am. Chem. Soc. 1986, 108, 2222.
(21) Baldas, J.; Heath, G. A.; Macgregor, S. A.; Moock, K. H.; Nissen, S. C.;
Raptis, R. G. J. Chem. Soc, Dalton Trans. (1972-1999) 1998, 2303.
(22) We do not report here the optimized geometries. However, we wish to
mention that we find a very good agreement between theoretical and X-ray
structural data for 2m.
9
J. AM. CHEM. SOC. VOL. 124, NO. 38, 2002 11475

A R T I C L E S
Bolzati et al.
Scheme 2
Stimulated by a referee comment, we performed additional
calculations to investigate the kinetic factors which may govern
the fac f mer isomerization. To this end, we have considered
the relevant local minima (I - IV) of the two [Tc(N)Cl2(PXP)]
structures, in agreement with the reaction pathway sketched in
Scheme 3. According to a general behavior of nitrido-containing
compounds, which display a strong tendency to form five-
coordinate species, it is reasonable to assume a dissociative
mechanism involving the loss of a chloride group from the fac-
form I to promote the formation of II.26 The following fac(II)
f mer(III) isomerization of the five-coordinate activated species
is consistent with a favorable cis f trans rearrangement of
π-acceptor phosphorus in a tbp array. In the final step, the halide
is engaged again in coordination to afford the resulting six-
coordinate mer-form IV. Furthermore, the PXP moiety can be
either in an asymmetric “twisted” conformation (as that assumed
in the fac complexes, see Figure 1) or in a symmetric
conformation. Thus, the isomerization can be decomposed into
two processes: the P-Tc-P angle opening and the conforma-
tional change of PXP. Because bond angles are usually harder
degrees of freedom than dihedrals, and have low energy barriers,
we assumed that the latter process occurs before the chloride
exit, so that it does not affect the overall barrier. In consequence,
the rate-determining step is expected to be the fac(II) f mer(III)
conversion of five-coordinate activated species. A linear transit
(LT) path connecting the two [Tc(N)Cl2(PXP)] structures was
devised, and we found that the formation energy of the highest-
energy linear transit structures is approximately the same for
the two PXP complexes, thereby generating an isomerization
barrier higher by ∼7 kcal mol-1 in the case of the PNP species.
Although the energy gap is again rather small, these calculations
confirm that the PNP ligand tends to kinetically hinder the fac
f mer conversion.
Figure 4. Structure of the benzene dimer in the parallel displaced geometry.
is preferred, albeit by only 0.2 kcal mol-1, also for [Tc(N)Cl2-
(PNP)]. Although such a small energy difference is certainly
within the error bar of our calculations, it could be interesting
to study more carefully the factors which could influence the
structure of the complex. A peculiar feature to be considered is
the presence, in fac isomers, of two facing phenyl rings, which
could interact through noncovalent “π-π” bonds.23 These
interactions are known to play an important role in determining
the structure of biomolecules and in supramolecular chemistry,
but, unfortunately, are not recovered by currently developed
GGA functionals.24 In Figure 4 is given a sketch of the so-
called “parallel displaced” (PD) benzene dimer, which is
characterized by two geometrical parameters: R1, the distance
between the ring planes, and R2, the distance between the
projections of the two ring centers. Most recent ab initio
correlated-function calculations25 found that R1 ) 3.4-3.6 Å
and R2 ) 1.5-1.8 Å. Remarkably, both theoretical parameters
obtained for [Tc(N)Cl2(PNP)] (R1 ) 3.5 Å, R2 ) 1.5 Å) and
crystal data of 4f ([Re(N)Cl2(PNP)]), in which R1 ) 3.5 Å, R2
) 1.7 Å, fall within these ranges, corroborating the hypothesis
of a possible role of π-π bonds in stabilizing the fac structure.
The strength of noncovalent interactions occurring between
aromatic rings is still an unsettled issue, because a strong
dependence is found both on the electron correlation correction
and on the size of the basis set.25 For the PD benzene dimer,
recent estimates give Eπ-π ≈ 2 kcal mol-1 25
.
However, we
cannot simply add this stabilization energy to the ∆Ef of the
fac isomers, because we must also correct it for the BSSE error
related to the phenyl orbitals, which affects the ∆Ef of the fac
isomer. To this end, we computed the interaction energy of a
PD benzene dimer by using the geometrical parameters extracted
by the facing phenyl rings of fac-[Tc(N)Cl2(PNP)]. We obtained
a repulsiVe 0.6 kcal mol-1 interaction energy, with a BSSE of
2.8 kcal mol-1. Thus, if we assumed that the phenyl-phenyl
interaction is realistically modeled by the benzene dimer model,
the ∆Ef of fac-[Tc(N)Cl2(PNP)] should be more positive by 2.8
kcal mol-1. However, inclusion of noncovalent interactions
implies a further (negative) correction given by -(0.6 + Eπ-π).
The two corrections cancel each other out almost exactly, so
that ∆Ef values of Table 3 turn out to be (accidentally) correct.
Description of the Structures. The structures mer-[Re(N)-
Cl2(POP)] 2m (Figure 5) and fac-[Re(N)Cl2(PNP)] 4f (Figure
6) consist of discrete monomeric and neutral nitrido complexes
packed with no intermolecular contacts shorter than the van der
Waals radii sum, while the structures of fac-[M(N)(HL2)(POP)]+
(M ) Tc, 7‚EtOH; M ) Re, 8) (Figure 7) and fac-[Tc(N)-
(HL2)(PNP)]+, 11 (Figure 8), are cationic with BF4 as
-
counteranions. Although we are prone to adopt the octahedral
description (fac and mer coordination of the PXP ligands instead
of cis and trans notation of diphosphine ligation) for all of the
complexes, the arrangement around the metal is actually
intermediate between the octahedral and square-pyramidal
geometries. Hence, 2m is better described as distorted octahedral
(N(1))(Re-O angle of 180.0 (7)°) (Figure 5), while the distorted
square-pyramidal geometry appears more appropriate for 7, 8,
and 11 (N(1))(M-O(N)trans of 159.7 (2), 162 (1), and 161.8
(23) (a) Burley, S. K.; Petsko, G. A. Science 1985, 229, 23. (b) Mu¨ller-Dethlefs,
K.; Hobza, P. Chem. ReV. 2000, 100, 143. (c) Hunter, C. A.; Lawson, K.
R.; Perkins, J.; Urch, C. J. J. Chem. Soc, Perkin Trans. 2 2001, 651.
(24) (a) Mejier, E. J.; Sprik, M. J. Phys. Chem. 1996, 105, 8689. (b) Tsuzuki,
S.; Lu¨thi, H. P. J. Chem. Phys. 2001, 114, 3949.
(25) (a) Tsuzuki, S.; Uchimaru, T.; Tanabe, K. J. Mol. Struct (THEOCHEM)
1994, 307, 107. (b) Hobza, P.; Selzle, H. L.; Schlag, E. W. J. Am. Chem.
Soc. 1994, 116, 3500. (c) Hobza, P.; Selzle, H. L.; Schlag, E. W. J. Phys.
Chem. 1996, 100, 18790. (d) Jaffe, R. L.; Smith, G. D. J. Chem. Phys.
1996, 105, 2780. (e) Tsuzuki, S.; Uchimaru, T.; Mikami, M.; Tanabe, K.
Chem. Phys. Lett. 1996, 252, 206. (f) Tsuzuki, S.; Uchimaru, T.; Mikami,
M.; Tanabe, K. Chem. Phys. Lett. 2000, 319, 547.
(26) ESI mass spectrometry of fac-[Re(N)Cl2(POP)] 2f and fac-[Re(N)Cl2(PNP)]
4f shows a fragmentation pattern consistent with the loss of one halide
group. In addition, such a dissociative mechanism is favored in polar
solvents (acetonitrile) as compared to less polar ones (dichloromethane).
9
11476 J. AM. CHEM. SOC. VOL. 124, NO. 38, 2002

Electrophilic Metal Fragment [99Tc(N)(PXP)]2+
A R T I C L E S
Scheme 3
Figure 5. ORTEP diagram of 2m, showing 35% probability ellipsoids.
Figure 8. ORTEP diagram of 11, showing 35% probability ellipsoids.
(2)O one, the angles between the two triangular faces being
4.1°. The four donor atoms deviate up to 0.05 Å from the mean
equatorial plane. In 4f, the Re atom is +1.09 Å away from the
P(1)P(2)N(1) plane and -1.57 Å from the Cl(1)Cl(2)N(2), the
angle between the two faces being 11.6°. The interligand angles
in the equatorial plane severely depart from the ideal 90° (from
83.8° to 98.6°), and the metal is displaced from the mean
equatorial plane by 0.52 and 0.39 Å in 2m and 4f, respectively,
toward the nitrido N(1) apical atom. As a consequence, the
N(1)-M-Cl angles are always greater than 90° (from 91.4° to
107.0°). For complexes 7, 8, and 11, the equatorial plane is
formed by P2SN donor atoms, and the metal atom is out from
the plane by +0.48, +0.44, and +0.43 Å, respectively, toward
N(1). In 7, the metal is +1.04 Å from the P(1)P(2)N(1) plane
and -1.49 Å from S(1)N(2)O(1), the angle between the two
faces being 13.9°. The corresponding three values in 8 and in
11 are +0.97 Å, -1.38 Å, 13.0° and +1.05 Å, -1.50 Å, 9.3°,
respectively. The planar bidentate HL2 ligand realizes around
Tc the slightly twist-envelope five-membered ring Tc-S(1)-
C(5)-N(3)-N(2) (torsion angles in the range from -18.2° to
17.3° for 7 and from -5.7° to 5.1° for 11) which makes a
dihedral angle of 26.9° in 7 (29.7° in 11) with the P(1)-Tc-
P(2) plane. The bite distance and angle of the diphosphine ligand
in the monosubstituted mer-complex 2m average 4.73 Å and
152.4°, while in the monosubstituted fac-complex 4f the
pertinent values are 3.63 Å and 98.6°.
Figure 6. ORTEP diagram of 4f, showing 35% probability ellipsoids.
The dominating feature of all of the structures is the strong
trans-labilizing effect exerted by the nitrido group on the
M-O(N)trans bond, which results in dramatically longer than
usually observed M-O(N) single bonds (2.505(9), 2.758(5),
2.638(9), and 2.81(1) Å in 2m, 7, 4f, and 11, respectively).7
All other metrical parameters are unexceptional, except for 8
Figure 7. ORTEP diagram of 7‚EtOH, showing 35% probability ellipsoids.
(8)°, respectively) (Figures 6 and 7). As a measure of the
octahedral distortion of 2m, the metal atom is +1.01 Å from
the Cl(1)P(1)N(1) plane, while it is -1.63 Å from the Cl(2)P-
9
J. AM. CHEM. SOC. VOL. 124, NO. 38, 2002 11477

A R T I C L E S
Bolzati et al.
(vide infra). In particular, the Tc-S distance (2.388(1) in 7 and
2.385(5) Å in 11) is intermediate within the mean value of 2.308
Å found in 179 Cambridge Structural Database27 entries for
complexes with CN five and the value of 2.408 Å for 128 data
in complexes with CN six. Despite the different diphosphine
incorporated, the technetium cations of 7 and 11 are virtually
superimposable, the rms being 0.05 Å when the fitting is
performed using the 13 atoms nearest to Tc. On the contrary,
comparison of the distances and angles in the virtually iso-
structural 7 and 8 complexes shows that the Tc-O(1) interaction
(separation of 2.758(5) Å) is much weaker than the correspond-
ing Re-O(1) one (2.56(1) Å). In the rhenium complex 8, despite
the short Re(N(1)) distance of 1.56(1) Å, the trans-labilizing
effect exerted by the nitrido group seems not to be operating,
and an exceptionally long Re-P(1) bond distance (2.566(6) Å)
is observed. The inner sphere angles display significant differ-
ences as well, the N(1)-M-N(2) angle being 110.2(2)° and
102(1)° in 7 and 8, respectively. Although differences in the
metrical parameters are precedented in isostructural nitrido-Tc-
(V) and -Re(V) compounds and explained in terms of the
lanthanide contraction,28 it is difficult to find a reasonable
justification for the abnormally long Re-P(1) distance. In any
case, this feature cannot be classified as an artifact in the X-ray
structure determination.
acceptor donor-atom set and, in particular, strongly support the
conclusion that a mixed P2S2 arrangement of atoms bound to a
TctN group should give rise to highly stable complexes.
The arguments given above were successfully employed here
to obtain a novel class of nitrido Tc(V) complexes, which were
designed to possess the unprecedented characteristic that two
different “bidentate” ligands were bound to the same metal
center.
In more detail, it would be possible to achieve an asym-
metrical arrangement of “bidentate” ligands around the metal
center by joining separately the two P atoms and the two S
atoms as illustrated in Scheme 2. Hence, the resulting hetero-
complex would be formed, for instance, by two different ligands
such as a diphosphine ligand and a bidentate dithiolate ligand.
A necessary consequence of this hypothetical view is that the
metal fragment [Tc(N)(P-P)]2+, resulting from the coordination
of a diphosphine ligand to the TctN group, should behave as
a strong acid center exhibiting selective reactivity toward the
nucleophilic bidentate ligand having a π-donor set of coordinat-
ing atoms.
According to our basic idea, a reasonable approach to this
synthetic strategy required the preliminary reaction of labile
nitrido precursors with appropriate diphosphine ligands to afford
the electrophilic metal fragment [Tc(N)(P-P)]2+. Diphosphines
having short spacers between the two P atoms were not suitable
frameworks because, in this situation, unstable octahedral
complexes such as [M(N)Cl(dmpe)2]+ and the dimeric [Tc(N)-
Cl(dippe)]22+ (dippe ) diisopropylphosphinoethane) were col-
lected.32 Oily mixtures without isolation of any pure compound
were instead recovered by using diphosphines having longer
methylene spacers such as 1,4-bis(diphenylphosphino)butane or
1,6-bis(diphenylphosphino)hexane. On the contrary, incorpora-
tion of a heteroatom in the chain interposed between the
phosphorus donors12,33 gave finally stable monosubstituted
species of the type [M(N)Cl2(POP)] and [M(N)Cl2(PNP)]. The
X-ray analyses of 2m and 4f evidenced that, upon coordination,
both diphosphines virtually act as tridentate chelate by placing
the phosphorus donors in a reciprocal trans- (2m) or cis-position
(4f), and the heteroatom invariably trans with respect to the
MtN linkages. Although the observed metal-heteroatom
distances are quite long (2.505(9) and 2.638(9) Å in 2m and
4f, respectively), this additional interaction appears to play a
crucial role in providing a further stabilization for these
intermediate monosubstituted species.
Discussion
The chemistry of nitrido-technetium(V) and nitrido-rhenium-
(V) compounds is dominated by the presence of the terminal
nitrido group, which is known to be among the strongest electron
donor agents.29 Square-pyramidal (sp) and trigonal bipyramidal
(tbp) geometries represent the two ideal arrangements of
classical five-coordinated nitrido Tc(V) and Re(V) species. As
a general behavior, sp geometry, where the MtN group
occupies an apical position, is mostly supported by two identical
bidentate ligands containing π-donor atoms positioned on the
basal plane, as in the case of [M(N)(Et2dtc)2]30 complexes (Et2-
dtc ) diethyldithiocarbamate). Conversely, an appropriate
mixing of π-donor and π-acceptor coordinating atoms allows
for the obtainment of the less common tbp array, as illustrated
by the disubstituted nitrido Tc(V) complexes [Tc(N)(PS)2]
containing phosphinothiol ligands.31 The structural characteriza-
tion of these latter compounds showed that the two phosphorus
atoms occupy the two apical sites and the two sulfur atoms and
the nitrido nitrogen atom are located on the trigonal plane of
the tbp geometry. These structural features reveal the strong
preference of π-acceptor phosphorus atoms to achieve a
reciprocal trans-position as opposed to the tendency of π-donor
atoms to assume a mutual cis-configuration. Challenge reactions
of nitrido Tc(V) bis-dithiocarbamato species toward ligand
replacement by phosphinothiols afford quantitatively the disub-
stituted [Tc(N)(PS)2] complexes with complete removal of the
dithiocarbamate substituents. These findings demonstrate the
superior coordination properties of the mixed π-donor/π-
Two monosubstituted isomers, fac-[M(N)Cl2(POP)] and mer-
[M(N)Cl2(POP)], were isolated with the POP ligand. The fac-
form was kinetically favored, being obtained at lower temper-
ature (i.e., refluxing dichloromethane), while the corresponding
mer-isomer was isolated after refluxing in acetonitrile. Quantita-
tive conversion of the fac-isomer into the mer-isomer was
achieved at room temperature, thus indicating that the meridional
complex is the thermodynamically favored product.
Conversely, only the facial isomer was isolated when the
diphosphine PNP was utilized, and no conversion into the
meridional form was observed under the same experimental
conditions adopted for POP-type isomerization. DFT theoretical
(27) Allen, F. H.; Davies, J. E.; Galloy, J. J.; Johnson, O.; Kennard, O.; Macrae,
C. F.; Mitchell, E. M.; Smith, J. M.; Watson, D. G. J. Chem. Inf. Comput.
Sci. 1991, 31, 187.
(28) Deutsch, E.; Libson, K.; Vanderheyden, J.-L. Technetium and Rhenium in
Chemistry and Nuclear Medicine 3; Raven Press: New York, 1990; p 13.
(29) Nuget, W. A.; Haymore, B. L. Coord. Chem. ReV. 1980, 31, 123.
(30) Baldas, J.; Bonnyman, J.; Pojer, P. M.; Williams, G. A. J. Chem. Soc.,
Dalton Trans. (1972-1999) 1981, 1798.
(31) Bolzati, C.; Boschi, A.; Uccelli, L.; Malago`, E.; Bandoli, G.; Tisato, F.;
Refosco, F.; Pasqualini, R.; Duatti, A. Inorg. Chem. 1999, 38, 4473-4479.
(32) Archer, C.; Dilworth, J.; Griffiths, D.; McPartlin, M.; Kelly, J. J. Chem.
Soc., Dalton Trans. (1972-1999) 1992, 183.
(33) (a) Bianchini, C.; Peruzzini, M.; Romerosa, A.; Zanobini, F. Organome-
tallics 1995, 14, 3152. (b) Bianchini, C.; Casares, J. A. C.; Peruzzini, M.;
Romerosa, A.; Zanobini, F. J. Am. Chem. Soc. 1996, 118, 4585.
9
11478 J. AM. CHEM. SOC. VOL. 124, NO. 38, 2002

Electrophilic Metal Fragment [99Tc(N)(PXP)]2+
A R T I C L E S
at the very low concentration scale (10-6-10-9 mol dm-3
utilized for the preparation of radiopharmaceuticals with the
γ-emitting metastable isomer Tc-99m. Under these conditions,
the metal fragment [99mTc(N)(PNP)]2+ behaves as an efficient
precursor for the preparation of asymmetrical complexes with
a wide class of bidentate ligands which can also incorporate
biologically active fragments.34
calculations appear to support this synthetic evidence. In fact,
the mer-arrangement is favored by a few kcal mol-1 in the case
of POP complexes, whereas fac- and mer-isomers are predicted
to be practically isoenergetic for complexes with PNP. As cited
above, because fac-isomers are the kinetic products, it appears
there are no thermodynamic factors favoring the rearrangement
fac f mer in PNP complexes. In this connection, looking for
kinetic components which may govern the fac f mer isomer-
ization, we found that the formation energy of the highest-energy
linear transit structures is essentially the same for the two PXP
complexes, thereby generating an isomerization barrier higher
by ∼7 kcal mol-1 in the case of the PNP complex. Although
this energy gap is again rather small, it is consistent with an
unfavorable fac-II f mer-III rearrangement of PNP complexes
(see Scheme 3).
)
Conclusions
The original approach of combining π-donor and π-acceptor
ligands coordinated to the [MtN]2+ group (M ) Tc, Re) has
proven successful for the synthesis of nitrido complexes with a
mixed asymmetrical coordination sphere. Reactions with diphos-
phine ligands PXP afforded complexes of the type [M(N)Cl2-
(PXP)] (PXP ) POP, PNP). In these compounds, π-acceptor
atoms exhibit a strong tendency to place each other in a
reciprocal trans-position, as shown by the representative
complexes mer-[M(N)Cl2(POP)]. This behavior sharply deter-
mines the chemical properties of these complexes, in particular,
by preventing the formation of symmetrical species containing
four π-acceptor atoms. However, insertion of a tertiary amino
group in the diphosphine backbone bridging the two phosphorus
induces the π-acceptor atoms to lock in a reciprocal cis-
arrangement, as found in fac-[M(N)Cl2(PNP)] complexes. The
N heteroatom provides also an additional contact with the metal
in the position trans with respect to the [MtN] linkage, thereby
expanding the resulting coordination sphere to a pseudoocta-
hedral geometry. In such an environment, the facial arrangement
of the diphosphine ligand allows for the easy displacement of
the halide groups to promote the formation of asymmetrical
heterocomplexes. Thus, as confirmed by DFT calculations,
monosubstituted fac-[M(N)Cl2(PXP)] derivatives act as key
intermediates to yield stable bis-substituted heterocomplexes fac-
[M(N)(PXP)(Ln)]0/+ with bidentate ligands L. This approach can
be successfully applied at the noncarrier added level with the
99mTc isomer to prepare a novel class of diagnostic imaging
agents.35 Studies on the derivatization of suitable bidentate
ligands for conjugation to biologically active groups are in
progress.
Moreover, DFT studies showed that facial coordination of
the PNP ligand grants the minimization of the steric constraints
of the spacer and increases the efficiency of the N f metal
interaction. It also allows an additional noncovalent interaction
between the two phenyl rings attached at the cis-positioned P
donors. The HOMO of the fac-PNP complexes (see Figure 2)
corresponds to a metal-ligand π* interaction in the complex
basal plane involving only the participation of the Cl ligands.
Thus, M-Cl bonds are expected to be substitution labile.
Furthermore, the lowest unoccupied orbitals of the facial
complexes (see Figure 2) are mainly composed of Tc d orbitals
whose lobes point away from the Cl and P atoms and are, in
consequence, quite accessible to attacking nucleophiles. Hence,
in monosubstituted fac-complexes, the metal fragment [M(N)-
(PXP)]2+ should be viewed as a “robust” moiety, almost
completely inert toward both substitution and redox reactions,
that selectively react with incoming nucleophiles after removal
of the labile halide groups. Considering the additional advantage
provided by coordination with chelating agents, ligand-exchange
reactions with some representative bidentate ligands summarized
in Chart 1 were carried out to afford the asymmetrical
heterocomplexes fac-[M(N)(PXP)(Ln)] and fac-[M(N)(PXP)-
(HLn)]+. In the former compounds, the combination of a thiolate
sulfur atom with a carboxylate oxygen atom forms the π-donor
coordinating set, whereas a thioketo sulfur and an amido
nitrogen or two dithiocarbamato sulfurs act as π-donor atoms
in the latter species. The possibility to use ligands with several
combinations of two π-donor atoms makes this procedure of
wide applicability. On the contrary, the monosubstituted mer-
complex with the POP ligand did not easily undergo substitution
reactions, and more drastic conditions were required. This fact
is in agreement with the low accessibility of the lowest
unoccupied MOs of the mer complexes (see Figure 2), as well
as with the remarkable steric hindrance exerted by the mer
arrangement of POP that makes it difficult for the incoming
bidentate ligand to span the two residual positions through
replacement of Cl atoms, as evidenced in the X-ray crystal
structure of 2m.
Acknowledgment. The authors are indebted to Nihon Medi-
Physics, Tokyo, Japan, for financial support of this work. Anna
Moresco is also gratefully acknowledged for elemental analyses.
Supporting Information Available: X-ray crystallographic
files, in CIF format, for the structure determinations of 2m-
11. Table containing a summary of computational details on
complex formation energies as well as ETS analyses of the BE
for [Tc(N)Cl2(PXPH)] model complexes of H-substituted PXP
ligands (PXP ) PNP, POP, P-P) (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
JA0200239
(34) Boschi, A.; Bolzati, C.; Benini, E.; Malago`, E.; Uccelli, L.; Duatti, A.;
Piffanelli, A.; Refosco, F.; Tisato, F. Bioconjugate Chem. 2001, 12, 1035.
(35) Duatti, A.; Bolzati, C.; Uccelli, L.; Refosco, F.; Tisato, F. PCT Int. Appl.
WO98 27 100 (Cl. CO7F9/50), 1998.
It is worthy to note at this point that a completely similar
behavior was observed when the same reactions were conducted
9
J. AM. CHEM. SOC. VOL. 124, NO. 38, 2002 11479