Superamidines 2. Synthesis of the bulky ligand N,N'-bis- (2,6-diisopropylphenyl)-trifluoroacetamidine and its molybdenum carbonyl complex

Article abstract of DOI:10.1139/v00-063

N,N'-bis-(2,6-diisopropylphenyl)trifluoroacetamidine has been prepared for the first time from 2,6-diisopropylaniline and the trifluoroacylation reagent TFAP via the imidoylchloride. The crystal structure of the amidine was determined, indicating that it

Full text of DOI:10.1139/v00-063

583
Superamidines 2. Synthesis of the bulky ligand
N,N′-bis-(2,6-diisopropylphenyl)-trifluoroacetamidine
and its molybdenum carbonyl complex1
René T. Boeré, Vicki Klassen, and Gotthelf Wolmershäuser
Abstract: N,N′-bis-(2,6-diisopropylphenyl)trifluoroacetamidine has been prepared for the first time from 2,6-
diisopropylaniline and the trifluoroacylation reagent TFAP via the imidoylchloride. The crystal structure of the amidine
was determined, indicating that it crystallizes in the Z-anti tautomer, in contrast to the nonfluorinated analogue, which
is E-anti in the solid state. In solution, as indicated by NMR spectroscopy, it exists in two isomeric forms. The
amidine reacts with Mo(CO)6 to produce a coordination complex with Mo(CO)3 in which the ligand is also in the
Z-anti geometry, the metal is η6-coordinated to the imino-2,6-diisopropylphenyl ring, and the amino N-H unit is di-
rected towards the metal, as determined by a single-crystal X-ray structure. Unlike the analogous nonfluorinated
acetamidine, there is no indication of an intermediate in which the neutral amidine is coordinated in a monodentate
fashion to an Mo(CO)5 unit, which we now attribute to the predominant geometry of the ligand, both in the solid state
and in solution, being Z-anti. The high steric bulk of this superamidine ligand apparently prevents the formation of a
metal–metal bonded Mo2(amidinate)4 as observed previously in a redox reaction between N,N′-diphenylbenzamidine
and Mo(CO)6 under similar thermal reaction conditions.
Key words: trifluoromethyl, superamidine, amidine, molybdenum, carbonyl, coordination.
Résumé : On a réalisé la première synthèse de la N,N′-bis-(2,6-diisopropylphényl)-trifluoroacétamidine à partir de la
2,6-diisopropylaniline et du réactif de trifluoroacétylation, TFAP, par le biais du chlorure d’imidoyle. On a déterminé la
structure cristalline de l’amidine qui indique qu’elle cristallise sous la forme du tautomère Z-anti, qui diffère de la situ-
ation avec l’analogue non fluoré qui, à l’état solide, existe sous la forme E-anti. Sur la base des données de R.M.N.,
en solution, il existe dans les deux formes isomères. L’amidine réagit avec le Mo(CO)6 pour donner un complexe de
coordination avec le Mo(CO)3 dans lequel, selon la structure déterminée par diffraction des rayons X sur un cristal
unique, le ligand est aussi dans une géométrie Z-anti, le métal forme une η6-coordination avec le noyau imino-2,6-
diisopropylphényle et l’unité N-H de l’amino est orientée vers le métal. Contrairement à ce qui a été observé avec
l’acétamidine analogue non fluorée, il ne semble pas exister d’intermédiaire dans lequel l’amidine neutre serait
coordinée de façon monodentate avec une unité de Mo(CO)5; on attribue maintenant cette situation à la géométrie
prédominante du ligand qui est Z-anti tant à l’état solide qu’en solution. Le grand encombrement stérique qui l’on
retrouve dans ce ligand superamidine empêche la formation d’un Mo2(amidinate)4 avec liaison métal-métal comme on
l’a observé antérieurement dans une réaction redox entre la N,N′-diphénylbenzamidine et le MO(CO)6 dans des condi-
tions de réactions semblables.
Mots clés : trifluorométhyle, superamidine, amidine, molybdène, carbonyle, coordination. Boeré et al. 589
Introduction
taining the 2,6-diisopropylphenyl (Dip) substituent on nitro-
gen (6). Such “super-bulky” substituents were originally
developed to stabilize low-coordinate main-group elements
of the third and subsequent periods. It is our contention that
the use of super-bulky substituents with conventional nitro-
gen-based ligands will lead to unique patterns of reactivity.
So far, we have demonstrated that it is possible to synthesize
Amidines and amidinate anions are useful ligands in the
general development of coordination chemistry (1, 2), and
their complexes with the Group 13 elements catalyze the
polymerization of ethylene (3–5). We have recently reported
the preparation of the first N,N′-disubstituted amidines con-
Received June 24, 1999. Published on the NRC Research Press website on May 10, 2000.
This paper is dedicated to Professor Christopher J. Willis on the occasion of his 65th birthday.
R.T. Boeré2 and V. Klassen. Department of Chemistry, University of Lethbridge, Lethbridge, AB T1K 3M4, Canada.
G. Wolmershäuser. Fachbereich Chemie der Universität Kaiserslautern, Erwin-Schrödinger-Straße, D-67663 Kaiserslautern,
Germany.
1Part of this work was presented at the 15th International Symposium on Fluorine Chemistry, UBC, Vancouver, B.C., August 2–7,
1997.
2Author to whom correspondence may be addressed. Telephone: (403) 329-2045. Fax: (403) 329-2057. e-mail: boere@uleth.ca
Can. J. Chem. 78: 583–589 (2000)
© 2000 NRC Canada

584
Can. J. Chem. Vol. 78, 2000
Scheme 1.
amidines 1a-c with Dip groups on nitrogen. The ligands dis-
solve in typical NMR solvents (CDCl3) as at least two iso-
meric forms, and so far, three different isomers have been
detected in the solid state by X-ray crystallography (6).
However, whereas analogous nonbulky amidines such as
N,N′-diphenylbenzamidine produce quadruply bonded dimoly-
bdenum(II) compounds (7) in thermal reactions with
Mo(CO)6 in hydrocarbon solvents, under these conditions
the superamidines have only provided two different kinds of
neutral coordination compounds. Thus superamidine 1a re-
acts with molybdenum carbonyl in hot n-heptane to produce
an N-coordinated LMo(CO)5 complex 2a, but on further re-
action this complex loses two further CO molecules to form
the half-sandwich LMo(CO)3 complex 3a. By contrast,
amidines 1b, c form LMo(CO)3 complexes 3b, c with no de-
tectable intermediates in thermal reactions with Mo(CO)6
(6).
The imidoylchloride does not react directly with a second
equiv of 4 as the nonfluorinated analogues do (6). Instead,
the lithium anilide was produced, and this reacted with the
imidoyl chloride to produce the desired amidine directly.
However we have found that to purify this raw product con-
version first to the hydrochloride, followed by reforming the
free amidine, 1d, using 23% sodium hydroxide solution was
required. All of these compounds are new, and have been
fully characterized by elemental analysis, mass spectroscopy
In this work we report results on a further superamidine
with Dip groups on N and a trifluoromethyl substituent on
the central carbon atom, 1d. The structure and reactivity of
this compound is distinct from that of its unfluorinated ana-
logue 1a, and more closely resembles that of the aryl deriva-
tives 1b, c, apparently for structural rather than electronic
reasons. In thermal reactions of 1d with Mo(CO)6 the only
observed product is the LMo(CO)3 complex 3d.
1
as well as H and 13C NMR spectroscopy. The details are
given in the Experimental section.
Scheme 2.
Results and discussion
Ligand synthesis
Our synthesis follows the basic methodology outlined by
Maringgele and Meller (8) for the synthesis of several N,N′-
diaryl-N′-trimethylsilyltrifluoroacetamidines (i.e., with the re-
maining hydrogen on N being replaced with an Si(CH3)3
group). Each step of the procedure required extensive opti-
mization, and in particular, differed significantly from that of
the nonfluorinated analogues 1a-c. The synthetic route is
outlined in Scheme 1, and the particulars of each step are
detailed in the Experimental section. Commercially available
2,6-diisopropylaniline, 4, was converted to the correspond-
ing amide of trifluoroacetic acid, 5, using the trifluoroacyl-
ation reagent 2-(trifluoroacetoxy)pyridine (TFAP) (9).
Dehydration and chlorination using 1.2 equiv of PCl5 in
refluxing SOCl2 produced the imidoylchloride, 6, which was
isolated and purified by vacuum distillation in 63% yield.
Maringgele and Meller (10) obtained a variety of N-aryl-
trifluoromethylimidoyl chlorides as byproducts in only
10–20% yield from reactions of amides with BCl3. The 19F
NMR chemical shifts of these imidoyl chlorides are found at
δ = –71.0 to –72.0 ppm, in excellent agreement with the
–71.48 ppm we find in the NMR spectrum of 6.
Isomers and tautomers
In CDCl3 solution, 1d exists in two isomeric forms as
1
shown by H, 13C, and 19F NMR spectroscopy. Previously
reported trifluoroacetamidines with a variety of nonbulky
aryl groups also show isomerism in solution (8). For exam-
ple, an amidine with a 2,6-dimethylphenyl group on one ni-
trogen and a phenyl group on the other displays two singlets
at δ = –61.9 and –60.4 ppm in the 19F NMR spectrum (8).
Similarly in 1d there are two closely spaced singlets at δ =
–67.24 and –66.84 ppm. The difference in absolute shift is
attributable to protonation vs. silylation at the amino nitrogen.
We have previously shown that superamidines can exist in
a variety of isomeric and tautomeric forms (6). The four
possible isomers of the trifluoromethyl derivative are shown
in Scheme 2, using the nomenclature devised by Perrin (11)
© 2000 NRC Canada

Boeré et al.
585
Fig. 1. PLUTO diagram of 1d, showing the atom numbering scheme. The ligand is in the Z-anti conformation. The rotationally disor-
dered CF3 group has been modeled by two trigonal F3 units with occupancy of 0.85, F(1)-F(3), and 0.15, F(4)-F(6).
Table 1. Crystal data and structure refinement.a
Compound
1d
3d
Formula
FW
C26H35F3N2
432.56
C29H35F3MoN2O3
612.53
T (K)
293(2)
293(2)
Crystal system
Space group
a (Å)
b (Å)
c (Å)
Monoclinic
C2/c
Monoclinic
P21/n
17.4663(17)
14.6527(12)
20.7594(15)
106.654(10)
5090.1(7)
12.0559(11)
16.5327(10)
15.3228(14)
95.270(11)
3041.2(4)
β (°)
V (Å3)
Z
8
4
Dc (mg/m3)
Crystal
1.129
1.338
colourless block
0.40 × 0.34 × 0.30
21505 (0.0403)
3871 / 30 / 301
R1 = 0.0421, wR2 = 0.0943
0.992
yellow plate
0.60 × 0.26 × 0.08
26067 (0.0338)
4787 / 0 / 364
R1 = 0.0287, wR2 = 0.0653
0.946
Size (mm)
Rfl. Collected (Rint)
Data / restraints / parameters
Final R [I > 2σ(I)]
Gof (all data)
aWeighting schemes used: 1d, w = 1/[σ2(Fo2) + (0.0500P)2]; 3d, w = 1/[σ2(Fo2) + (0.0420P)2] where P = (Fo2 + 2Fc2)/3.
based on the Cahn–Ingold–Prelog sequence rules. The low-
est-energy isomer of 1d, as minimized in an AM1 calcula-
tion, is the Z-anti form. Relative to this, the tautomeric
E-syn isomer lies only 1.2 kJ mol–1 higher in energy, while
the Z-syn isomer lies 11.0, and the E-anti 26.7 kJ mol–1
higher in energy. Calculations on 1a, however, indicate
that the E-anti isomer for the methyl ligand 1a lies only
15.5 kJ mol–1 higher than Z-anti. These calculations do not
factor in the energy of hydrogen bonding, which apparently
can stabilize the E-anti form as seen in the solid state for the
methyl analogue 1a. The typical energy of a single H-bond
is ~20 kJ mol–1 (12). We have determined the single-crystal
X-ray structure of 1d (see Fig. 1), which indicates that it
crystallizes preferentially as the Z-anti isomer. There are no
short intermolecular contacts in the crystal. Apparently the
steric bulk of a trifluoromethyl group is sufficiently large to
cancel out the energetic advantage for the formation of an
H-bonded dimer, which is possible only in the E-anti con-
formation. Important bond distances and angles are listed in
Table 1.
Häfelinger and Kuske (13) have defined the parameter ∆CN
=
d(C—N) – d(C=N) for the central N—C—N linkage found
in all amidines, using crystallographic structural data. This
parameter ranges from 0 in the structure of 1a, to 0.178 Å
in 2,6-cis-dimethylpiperidyl-N-phenyl-2,2-dimethylpropion-
amidine, an amidine where conjugation is minimized due to
bulky substituents on nitrogen and carbon (d(C—N) =
1.441(5); d(C=N) = 1.263(5) Å), and which crystallizes in
© 2000 NRC Canada

586
Can. J. Chem. Vol. 78, 2000
Table 2. Selected interatomic distances and angles.
Table 3. Solution IR data.a
ν(N-H)
Distances (Å)
Angles (°)
ν(N-C-N)
ν(CϵO)
Compound
(cm–1)
(cm–1)
(cm–1)
1d
1d
3d
3453, 3356
1657
1665
—
N(1)—C(1)
1.270(2)
1.364(2)
0.094
1.426(2)
1.444(2)
1.509(3)
0.92(2)
C(1)-N(1)-C(9)
119.49(15)
128.55(15)
126.85(16)
115.45(17)
117.68(18)
1971, 1906, 1883
N(2)—C(1)
C(1)-N(2)-C(21)
N(1)-C(1)-N(2)
N(1)-C(1)-C(2)
N(2)-C(1)-C(2)
an-Heptane solution in 0.2 mm NaCl solution cell.
a
∆
CN
N(1)—C(9)
N(2)—C(21)
C(1)—C(2)
N(2)—H(2)
meta hydrogen atoms, in the expected 1:2 intensity ratio.
The magnitude of these shifts are typical of those observed
for η6-arenes coordinated to an Mo(CO)3 unit. The solution
IR spectrum (Table 3) showed three bands in the carbonyl-
stretching region, a very strong singlet at about 1966 cm–1,
and a slightly less intense doublet centred at about 1885 cm–1.
This is the pattern expected for a fac-octahedral tricarbonyl
metal complex with highly asymmetric coligands (16).
The single-crystal X-ray structure of 3d is presented in
Fig. 2. The same “baseball catcher’s mitt,” structure is ob-
served as previously obtained for the methyl 3a and tolyl 3b
structures, in which the amidine ligand, itself in the Z-anti
tautomer, has “caught” the Mo(CO)3 fragment with the in-
side face of the imino-bearing ring. With respect to the
C(1)=N(1) double bond, the metal complex also has Z sym-
metry. The N-H of the amino group corresponds to the
“catcher’s thumb,” and there is a close contact of 3.14(3) Å
between H(1) and Mo, less than a reasonable guess at the
sum of their van der Waals’ radii (the van der Waals radius
for molybdenum is not contained in standard compilations;
we have used values of neighbouring elements to provide
this estimate). The similarity in structures of 3d, 3a, and 3b
is striking. The amidine unit in 3d is bent further away from
the molybdenum atom than in 3a. This is evidenced by a 5°
larger N(1)-C(1)-N(2) angle, and a 3° smaller N(1)-C(1)-
C(2) angle. Other metric parameters between the two struc-
tures are extremely similar.
3d
N(1)—C(1)
1.270(3)
1.343(3)
0.073
1.527(3)
1.402(3)
1.454(3)
0.88(3)
1.940(3)
1.912(4)
1.947(4)
3.14(3)
N(1)-C(1)-N(2)
N(1)-C(1)-C(2)
N(2)-C(1)-C(2)
C(1)-N(1)-C(9)
C(1)-N(2)-C(21)
C(33)-Mo(1)-C(34)
C(33)-Mo(1)-C(35)
C(34)-Mo(1)-C(35)
Mo(1)-H(1)-N(1)
128.1(2)
113.5(2)
118.4(2)
126.6(2)
128.3(2)
83.63(13)
82.96(14)
85.77(16)
N(2)—C(1)
a
∆
CN
C(1)—C(2)
N(1)—C(9)
N(2)—C(21)
N(2)—H(2)
Mo(1)—C(33)
Mo(1)—C(34)
Mo(1)—C(35)
Mo(1)—H(2)
a
∆
= d(C-N) – d(C=N).
CN
the rarely observed Z-syn form (14). In the crystal structure
of 1d, ∆CN = 0.094, the largest we have yet observed for a
superamidine structure. The metric parameters of 1d other-
wise closely parallel those of 1c, the only other
superamidine so far which crystallizes exclusively in the
Z-anti form, for which ∆CN = 0.057 (6). The C—N single
bonds in 1c and 1d are identical within experimental error at
1.364 Å. Thus the origin of the larger ∆CN value in 1d is the
shorter C=N double bond, which at 1.270(2) is almost 0.04
Å shorter in the fluorinated amidine. However, these values
are not unprecedented. In the structure of N′-p-bromophenyl-
N-(p-tolyl)acetamidine, which crystallizes in the E-anti iso-
mer, d(C–N) = 1.366(8) and d(C=N) = 1.273(8) Å, leading
to a ∆CN = 0.093 (15).
Progress of reaction with Mo(CO)6
The thermal reactions with Mo(CO)6 of all the super-
amidines that we have studied do not go on to the formation
of bridged multiply bonded dimolybdenum species (6, 7).
This is undoubtedly due to the high steric bulk of the Dip
groups. The question that remains, however, is why 1d re-
acts more like the aryl amidines 1b and 1c, in going directly
to an η3-Mo(CO)3 complex, than the methyl amidine 1a,
which first forms an N-bonded Mo(CO)5 complex, and only
on further heating produces the η3-Mo(CO)3 complex?
We believe that this difference in reactivity can be traced
to the structural forms of the free ligand in the solid state
and in solution. Only 1a crystallizes in the E-anti form as an
H-bonded dimer, unlike 1b, 1c, and 1d. Furthermore, 1a is
much less soluble in n-heptane than the others, undoubtedly
due to the dimer structure. Thus visual inspection of the
thermal reaction of 1a with Mo(CO)6 clearly shows that un-
dissolved ligand is present during much of the course of the
reaction. This is different from what is observed for the
other three superamidines, which dissolve readily as the
temperature of the solvent is increased. We hypothesize that
if the three more soluble amidines dissolve in the Z-anti ge-
ometry they have in the solid, reaction of ligand in this con-
figuration promotes conversion to the η6-Mo(CO)3 complex.
On the other hand, if the hexacarbonyl reacts with a
Reaction with Mo(CO)6
Amidine 1d was heated with Mo(CO)6 in refluxing n-
heptane under an atmosphere of N2. The course of the reac-
tion was easily monitored by solution infrared spectroscopy
and we observed that 1d reacted slowly, with no observable
evidence by solution IR spectroscopy of any intermediate, to
the final product, a bright yellow crystalline solid, which
was produced in 25% yield. The structure of the compound
was indicated by H NMR and IR spectroscopy to be a half-
sandwich “piano-stool” complex of Mo(CO)3, in which the
metal is coordinated to one of the electron-rich aromatic
1
1
rings of the ligand rather than through nitrogen. In the H
NMR spectrum, the signals due to the aromatic hydrogen at-
oms of one of the diisopropylphenyl groups are shifted
1.45–1.49 ppm upfield, while the other signals remain at the
same chemical shift. In the free ligand, the meta and para
hydrogen resonances are not distinguishable, but in the com-
plex, the upfield-shifted signals show a resolved pseudo-
triplet for the para and a resolved pseudo-doublet for the
© 2000 NRC Canada

Boeré et al.
587
Fig. 2. PLUTO diagram of 3d, showing the atom numbering schemes. The ligand is also in the Z-anti conformation, with the Mo atom
coordinated to the inside of the aryl ring attached to the imino nitrogen atom. The rotationally disordered CF3 group has been modeled
by two trigonal F3 units with occupancy of 0.87, F(1)-F(3), and 0.13, F(4)-F(6).
hydrogen-bonded dimer in the E-anti geometry before this
can isomerize to the Z-anti form, it forms a metal complex
in which this geometry is preserved. We have previously
shown that conversion of the η1-Mo(CO)5 complex to the η6-
Mo(CO)3 complex is accompanied by E to Z ligand
isomerism (6).
unit blocking the aromatic ring that is responsible for the
shielding effect.
Experimental section
Is there any evidence that the amidines exist in solution in
the Z-anti form? Consider the combined NMR spectroscopic
parameters presented in Table 4 for all the species we have
so far prepared. The proton (and 19F, where applicable)
NMR parameters of the dominant species in CDCl3 solution
for the free amidines in each case are strikingly similar to
those of the metal tricarbonyl derivatives. If, as seems likely,
the latter dissolve in the Z-anti geometry found in the solid
state, the correspondence of the NMR spectoscopic parame-
ters supports this isomer as the predominant form of the free
ligand in CDCl3 as well. Furthermore, this is also the lowest
energy form in the gas phase as indicated by MO calcula-
tions.
General
2,6-diisopropylaniline, 2-hydroxypyridine, trifluoroacetic
anhydride, thionyl chloride, phosphorus pentachloride, n-
butyl lithium (Aldrich), and molybdenum carbonyl (Strem)
were commercial products, and used as received, except that
the aniline was vacuum distilled and stored under nitrogen
before use. Solvents were reagent grade, and distilled from
sodium wire (toluene, THF), P2O5 (CH2Cl2), or LiAlH4 (n-
heptane). 2-(Trifluoroacetoxy)pyridine (TFAP) was prepared
from 2-hydroxypyridine and trifluoroacetic anhydride by the
literature method (9). Reactions involving metal carbonyl
and organolithium reagents were performed under an atmo-
sphere of purified N2 using a dry-box, Schlenk ware, and
vacuum-line techniques; all other procedures were per-
formed in vessels open to the atmosphere but protected by
CaCl2 drying tubes. Infrared spectra were recorded on a
BOMEM MB102 Fourier transform spectrometer, and are
KBr pellets unless otherwise specified. NMR spectra were
acquired at 250.13 (1H) and 62.90 (13C) MHz on a Bruker
AC250-F spectrometer referenced to TMS, and on a Bruker
AMX300 at 282.41 (19F) MHz referenced to external C6F6
in C6D6 at –163.00 ppm on the CFCl3 scale, and are in
CDCl3 unless otherwise specified. Mass spectra were re-
corded by the Mass Spectrometry Center, University of Al-
berta, Canada and in the Fachbereich Chemie, Universität
Kaiserslautern. Elemental analyses were performed by MHW
The differences that do exist between the NMR spectro-
scopic parameters of free ligand and complex are readily at-
1
tributed to variations in ring shielding effects. Thus the H
signal of the amino NH atom is further upfield in the ligand,
where in the Z-anti form it experiences ring current shield-
ing from the Dip group on the imino nitrogen. Similarly, one
CH3 group is significantly further upfield than the rest. As-
suming that the Dip group on the amino nitrogen rocks back
and forth on the C—N bond, one set of methyl groups would
also experience ring current shielding from the ring attached
to the imino nitrogen. In the Mo(CO)3 complexes, both types
of anomalously shielded protons revert to more normal
chemical shifts. This is consistent with the metal carbonyl
© 2000 NRC Canada

588
Can. J. Chem. Vol. 78, 2000
Table 4.1H and 19F NMR chemical shifts.
b
b
b
b
Compound
1a majorc,d 5.40
1a minorc,d 5.51
NH
Aromatic
X
iPr CHa iPrCHa
iPr CH3
1.32
iPr CH3
1.26
iPr CH3
1.19
iPr CH3
1.04
7.29–7.04m
buried
7.29–6.97 m
buried
7.35–6.70 m
buried
7.34–7.06 m
buried
1.88f
3.23
3.21
1.84f
3.31 br
2.95 br
3.16
buried between 1.35 and 1.15
1.23 0.98 0.88
buried 1.4–1.2
buried 1.0–0.8j
1b majorc
1b minorc
1c majorc
1c minorc
1d majore
1d minore
2ac
5.67
6.88
5.66
6.89
5.50
6.05
7.37
7.98
8.20
8.18
8.07
2.27g 3.24
1.34
1.34
buried
3.74h 3.23
3.37 br
3.45 br
3.00 br
3.16
1.23
0.99
buried
1.20
0.90
0.8
1.00
buried
2.93 br
3.03
buried 1.1–1.3
–67.24i
–66.84i
1.42f
3.18
1.34
1.27
3.25 br
3.16
2.82 br
3.15
buried between 1.4–1.2
7.4–7.16 m
1.42
1.29
1.28
1.28
1.28
1.30
1.25
1.27
1.27
1.27
1.27
1.24
1.20
1.20
1.26
1.11
1.22
0.86
0.88
1.22
3ac
7.35–7.17 m, 5.82 tr, 5.60 d
7.29–6.96 m, 5.79–5.66 m
7.29–7.06 m, 6.69 d, 5.79–5.66 m
7.37–7.17 m, 5.89 t, 5.57 d
1.79f
3.19
2.81
3bc
2.27g 3.17
2.85
2.84
2.72
3cc
3.74h 3.17
–67.54i
3.14
3de
aHeptet with J 7 Hz.
bDoublet with J 7 Hz.
cRef. (6).
i
dThere is evidence for a third species as well, with a single sharp Pr CH signal, which could be the H-bonded dimer in solution.
eThis work.
f δCH3 group on amidine backbone.
g δCH3 group on 4-position of phenyl ring.
h δCH3O group on 4-position of phenyl ring.
i δCF3 group, 19F NMR.
jUpfield doublet is evident centred on 0.80 ppm.
Laboratories, Phoenix, AZ, U.S.A. and by the analytical ser-
vices of the Fachbereich Chemie, Universität Kaiserslautern.
7.18 (m, 3 H), 2.62 (hept, 6.9 Hz, 2 H), 1.18 (d, 6.9 Hz, 12
H). 13C NMR: 140.47, 135.93, 133.94 (quart, 49 Hz),
126.35, 123.47, 116.65 (quart, 277 Hz), 28.64, 22.97. 19F
NMR: –71.48 (s, CF3). Mass spec: 290.7 (M+ on 35Cl, 29%),
Preparation of CF3C{O}NH(2,6-iPrC6H3), 5
i
+
256 (M–Cl+, 75%), 240 (M–CH4Cl), 177 (C6H3 Pr2NH2 ,
2,6-Diisopropylaniline (90.9 g, 513 mmol) in 250 mL of
dry diethyl ether was added drop wise to TFAP (98.0 g,
513 mmol) in 300 mL of the same solvent, with cooling to
maintain 23°C. After stirring 20 h, the solvent was removed,
and the residues extracted with 1.6 L of dH2O by vigorous
stirring for 2 h, filtering and drying to give 128.6 g of pure 5
(471 mmol, 92% yield), mp 158–159°C. An analytical sam-
ple was obtained by recrystallizing from ethanol. IR: ν(N-H)
i
+
48%), 162 (C6H3 Pr2 , 100%). Anal. calcd. for C14H17ClF3N:
C 57.64, H 5.87, N 4.80%; found: C 57.57, H 5.85, N
5.00%.
Preparation of CF3C{N-2,6-iPr2C2H3}NH(2,6-iPr2C6H3),
1d
210 mL of 1.5 M nBuLi (hexane solution) was added via
cannula to 44 mL (233 mmol) of 2,6-diisopropylaniline in
500 mL of THF in a 1 L flask equipped with a reflux con-
denser. After stirring a further 30 min, 68 g (233 mmol) of 6
in 200 mL THF was added dropwise with cooling. After
stirring 30 min, the mixture was heated to reflux for 10 h,
filtered, and the filtrate evaporated. The resulting oil was
treated with 1 L of 95% ethanol followed by 600 mL cHCl.
After evaporation to dryness, the residues were divided into
five parts, each dissolved in 200 mL ethanol, reacted with
200 mL of 23% NaOH, and extracted three times with
100 mL of toluene. The toluene phase was washed with
NH4Cl and then water, and evaporated to an oil. Extended
pumping on this oil produced crystals, which could be
recrystallized from n-heptane, to give 16 g of colourless
crystals of 1d (38 mmol, 16% yield). An analytical sample
was produced by recrystallizing a second time from n-
heptane, mp 94–98°C. IR: ν(N-H) 3451, 3358, ν(N-C-N)
1
3254, ν(C=O) 1709 cm–1. H NMR: 7.54 (s, 1 H, conc. de-
pend.), 7.40–7.20 (m, 3 H), 2.95 (hept, 6.9 Hz, 2 H), 1.19 (d,
6.9 Hz, 12 H. 13C NMR: 157.11 (quart, 37 Hz), 146.17,
129.63, 128.10, 123.98, 116.34 (quart, 289 Hz), 28.83,
23.58. 19F NMR: –75.47 (s, CF3). Mass spec: 272.8 (M+,
+
+
34%), 258 (M–CH3 , 12%), 204 (M–CF3 , 100%). Anal.
calcd. for C14H18F3NO: C 61.5, H 6.6, N 5.1%; found: C
61.5, H 6.7, N 5.2%.
Preparation of 4-CF3C{Cl}N(2,6-iPr2C6H3), 6
5 (20.0 g, 73.2 mmol) was refluxed in 15 mL of SOCl2
(excess) for 0.5 h (oil bath at 80–90°C), after which one
equiv of PCl5 (15.2 g, 73.2 mmol) was added in small incre-
ments. After a further 1.5 h reflux and NMR monitoring of
the reaction, a further 2.5 g (12 mmol) PCl5 was added,
which after 3 h led to complete consumption of 5. The bath
temperature was raised to 140°C and the remaining SOCl2
was distilled off. The residues were fractionally distilled un-
der vacuum, producing 6 as a colourless liquid, bp 48°C / 4
× 10–2 mbar (13.5 g, 46.3 mmol, 63% yield). IR (neat):
1
1655 cm–1. H NMR: 7.37–7.06 (m, 6H), 5.50 (s, 1 H), 3.18
(hept, 6.9 Hz, 2H), 3.03 (hept, 6.7 Hz, 2 H), 1.34 (d, 6.8 Hz,
6 H), 1.27 (d, 6.8 Hz, 6 H), 1.20 (d, 6.7, 6 H), 1.00 (d,
6.7 Hz, 6 H). 13C NMR: 147.51, 143.28 (quart, 33 Hz),
1
ν(Χ=Ν) 1703, ν(Χ−Φ) 1287, 1217, 1165. H NMR: 7.20–
© 2000 NRC Canada

Boeré et al.
589
140.62, 138.17, 131.18, 129.03, 123.46, 118.10 (quart,
279 Hz), 28.48, 25.32, 24.16, 22.41, 22.10. 19F NMR:
–67.24 (s, CF3, major species), –66.84 (s, CF3, minor spe-
cies). Mass spec: 432.27486 (M+, 57%), 389 (M–iPr+, 79%),
Acknowledgments
The Natural Sciences and Engineering Research Council
of Canada, the University of Lethbridge research fund, and
the Alexander von Humboldt-Foundation supported this
work. R.T.B wishes to thank Prof. Dr. O.J. Scherer and Dr.
G. Wolmershäuser for their hospitality during a sabbatical
leave at Kaiserslautern and acknowledges the access given to
the facilities of the Fachbereich Chemie of the University.
We thank staff at the Chemistry Department of the Univer-
sity of Calgary for acquiring the 19F NMR spectra, and one
of the referees for many helpful suggestions.
i
+
i
256 (M-C6H3 Pr2NH , 100%), 214 (CF3NC6H3 Pr+, 13%),
i
+
177 (C6H3 Pr2NH2 , 73%). Anal. calcd. for C26H35F3N2: C
72.19, H 8.16, N 6.48%; found: C 71.95, H 7.99, N 6.59%.
Preparation of η6-[CF3C{N-2,6-iPr2C6H3}NH(2,6-
iPr2C6H3)]Mo(CO)3, 3d
Three grams (6.9 mmol) of 7 and 1.83 g (6.9 mmol) of
Mo(CO)6 were loaded under N2 into a side-arm flask under
N2. One hundred fifty millilitres of dry n-heptane was then
added, and the mixture heated to reflux. Progress of the re-
action was monitored by solution IR, and heating discontin-
ued after 42 h. Removal of the solvent followed by
recrystallization from n-heptane provided 1.02 g (1.7 mmol,
25% yield) of 3d as bright yellow crystals, which were ana-
lytically pure, mp 145°C dec. IR: ν(N-H) 3462, 3208,
ν(C=O) 1969, 1865, ν(N-C-N) 1665 cm–1. 1H NMR: 8.07 (s,
1 H), 7.37–7.17 (m, 3 H), 5.89 (t, 6.5 Hz, 1 H), 5.57 (d,
6.5 Hz, 2 H), 3.14 (hept, 6.8 Hz, 2 H), 2.72 (hept, 6.9 Hz, 2
H), 1.28 (d, 6.9 Hz, 6 H), 1.27 (d, 6.8 Hz, 6 H), 1.26 (d, 6.9,
6 H), 1.22 (d, 6.8 Hz, 6 H). 13C NMR spectrum not obtained
due to sample instability. 19F NMR: –67.54 (s, CF3). Mass
spec: 614.16607 (M+ based on 98Mo, 31%), 558 (M–2CO+,
4%), 530 (MoL+, 100%), 486 (MoL–iPr+, 2%), 432 (L+,
References
1. G.A. Dushenko, I.E. Mikhailov, A. Zschunke, G. Reck, B.
Schulz, C. Mugge, and V.I. Minkin. Mendeleev Comm. 67
(1999).
2. A. Kasani, R.P.K. Babu, K. Feghali, S. Gambarotta, G.P.A.
Yap, L.K. Thompson, and R. Herbst-Irmer. Chem. Eur. J. 5,
577 (1999).
3. A. Littke, N. Sleiman, C. Bensimon, D.S. Richeson, G.P.A.
Yap, and S.J. Brown. Organometallics, 17, 446 (1998).
4. M.P. Coles and R.F. Jordan. J. Am. Chem. Soc. 119, 8125
(1997).
5. S. Dagorne, L.A. Guzei, M.P. Coles, and R.F. Jordan. J. Am.
Chem Soc. 122, 274 (2000).
6. R.T. Boeré, V. Klassen, and G. Wolmershäuser. J. Chem. Soc.
Dalton Trans. 4147 (1998).
7. F.A. Cotton, T. Inglis, M. Kilner, and T.R. Webb. Inorg. Chem.
14, 2023 (1975).
8. W. Maringgele and A. Meller. Z. Anorg. Allg. Chem. 445, 107
(1978).
i
36%), 389 (L–iPr+, 62%), 256 (M-C6H3 Pr2NH +, 72%), 177
i
+
(C6H3 Pr2NH2 , 56%). Anal. calcd. for C26H35F3MoN2O3: C
56.86, H 5.76, N 4.57%; found: C 57.00, H 5.60, N 4.65%.
9. T. Keumi, M. Shimada, T. Morita, and H. Kitajima. Bull.
Chem. Soc. Jpn. 63, 2252 (1990).
10. W. Maringgele and A. Meller. Monatsh. Chem. 110, 63
(1979).
11. C.L. Perrin. In The chemistry of the amidines and imidates.
Vol. 2. Edited by S. Patai and Z. Rappoport. Wiley, Chichester.
1991. Ch. 3, p. 152.
12. P. Atkins. Physical chemistry, 6th ed. Freeman, N.Y. 1998.
p. 666.
13. G. Häfelinger and K.H. Kuske. In The chemistry of the
amidines and imidates. Vol. 2. Edited by S. Patai and Z.
Rappoport. Wiley, Chichester. 1991. Ch. 1, pp. 1–100.
14. G. Gilli and V. Bertolasi. J. Am. Chem. Soc. 101, 7704 (1979).
15. E. Cisazk, M. Gdaniec, and Z. Kosturkiewicz. Acta
Crystallogr. B43, 1362 (1978).
Crystallography
Data collected on a Stoe IPDS diffractometer using
monochromated Mo Kα radiation (0.71073 Å) by the φ-oscil-
lation method. An analytical absorption correction was car-
ried for 3d, but no correction was necessary for 1d. Both
structures were solved by direct methods using SHELXS-97
(17) and refined by full-matrix least-squares on F2 using
SHELXL-97 (18) using all reflections. For 1d all the hydro-
gen atoms except the one attached to N were localized geo-
metrically and a riding model was used for refinement. The
CF3 group is rotationally disordered and was modeled by
two trigonal F3 units with occupancies of 0.85 and 0.15. For
3d, all the hydrogen atoms were localized, and a similar dis-
order model was applied to the F3 unit, with refined occu-
pancies of 0.87 and 0.13.3
16. C.M.
Lukehart.
In
Fundamental
transition
metal
organometallic chemistry. Brooks/Cole, Monterey, CA. 1985.
p. 81.
Electronic structure calculations
17. G. Sheldrick. SHELXS-97, University of Göttingen. 1997.
18. G. Sheldrick. SHELXL-97, University of Göttingen. 1997.
19. HyperChem Pro Release 5.1. HyperCube, Waterloo, Ontario.
All calculations were performed using the AM1 method
as implemented in HyperChem 5.1 running on a Pentium II
computer under Windows 95 (19).
3 Detailed structure reports including the H-atom positions and the anisotropic temperature factors have been deposited. Copies of materials
on deposit may be purchased from the Depository of Unpublished Data, Document Delivery, CISTI, National Research Council Canada, Ot-
tawa, Canada K1A 0S2.
© 2000 NRC Canada