Synthesis and lanthanide coordination chemistry of trifluoromethyl derivatives of phosphinoylmethyl pyridine N-oxides

Article abstract of DOI:10.1039/b905947d

A synthetic route for the formation of 2-[bis(2-trifluoromethylphenyl) phosphinoylmethyl]pyridine N-oxide (1c) and 2-[bis(3,5-trifluoromethylphenyl) phosphinoylmethyl]pyridine N-oxide (1d) was developed and the new ligands characterized by spectroscopic m

Full text of DOI:10.1039/b905947d

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Synthesis and lanthanide coordination chemistry of trifluoromethyl derivatives
of phosphinoylmethyl pyridine N-oxides†
Sylvie Pailloux,a Cornel Edicome Shirima,a Alisha D. Ray,a Eileen N. Duesler,a Karen Ann Smith,a
Robert T. Paine,*a John R. Klaehn,b Michael E. McIlwainb and Benjamin P. Hayc
Received 25th March 2009, Accepted 18th June 2009
First published as an Advance Article on the web 24th July 2009
DOI: 10.1039/b905947d
A synthetic route for the formation of 2-[bis(2-trifluoromethylphenyl)phosphinoylmethyl]pyridine
N-oxide (1c) and 2-[bis(3,5-trifluoromethylphenyl)phosphinoylmethyl]pyridine N-oxide (1d) was
developed and the new ligands characterized by spectroscopic methods and single-crystal X-ray
diffraction analyses. The coordination chemistry of 1c was examined with Yb(NO3)3 and the molecular
structure of one complex, [Yb(1c)(NO3)3(DMF)]·DMF·0.5H2O, was determined by single-crystal X-ray
diffraction methods. The ligand is found to coordinate in a bidentate fashion, and this is compared
against lanthanide coordination chemistry observed for the related ligand, [Ph2P(O)CH2] C5H4NO.
Introduction
The ligation behavior of the NOPO-class ligand 2-[(diphenyl)-
phosphinoylmethyl]pyridine N-oxide (1a) toward lanthanide ions,
Ln(III), has been examined in our group and found to produce
stable 1 : 1, 2 : 1 and 4 : 1 ligand–metal (L : M) complexes.1
=
Although the ligand resting state structure has the P O and
N–O bond dipoles rotated away from each other, in all cases
examined to date, the ligand is able to rearrange and bind to
Ln(III) ions with a relatively strain-free seven-membered bidentate
chelate coordination mode. The ready formation of 4 : 1 L : M
complexes suggests that examples of 1 might serve as useful
liquid–liquid extractants for f-element ions.2 Unfortunately 1a
only shows significant solubility in chlorocarbon solvents that are
inappropriate for practical solvent extraction purposes. Therefore,
a general synthesis for derivatives with aliphatic substituents was
sought.3 One such compound with 2-ethylhexyl substituents, [(2-
EtHx)2P(O)CH2]C5H4NO (1b), was prepared and found to be
soluble in aromatic and aliphatic solvents.4 Initial extraction
studies with 1b in dodecane revealed undesirable formation of
a third phase during extractions. However, toluene solutions of
1b showed sharp phase separation and extractions of Eu(III) and
Am(III) with 0.05 M 1b in toluene gave favorable performance.4
Although distribution ratios for 1b are not as large as found
for the two-armed tridentate analog, [(EtHx)2P(O)CH2]2C5H3NO
(2b), interest in one-armed ligands 1 as practical liquid–liquid
extraction reagents has remained strong.
aqueous solutions obtained from nitric acid dissolution of solid
nuclear materials.5–8 In a current option the process employs a car-
bamoylmethyl phosphine oxide (CMPO), Ph2P(O)CH2C(O)N(n-
Bu)2 as the actinide ion extractant and phenyl trifluoromethyl
sulfone (FS-13) as the organic diluent. As formulated, good
partitioning of low concentrations of Ln and An ions is observed;
however, reduced performance is found when high metal ion
concentrations are encountered. In part this is a result of low
organic phase solubility of the CMPO/metal complexes in FS-13.
In the present contribution, we describe the synthesis of
new NOPO ligands {[2-(CF3)C6H4)]2P(O)CH2]}C5H4NO (1c)
and {[3,5-(CF3)2C6H3)]2P(O)CH2]}C5H4NO (1d) that carry CF3
substituted aryl substituents. It was anticipated that the CF3
functionalized aryl ligands would be more easily prepared and
purified relative to 1b and that they might display improved
solubility in FS-13 compared to CMPO ligands. It was also of
interest to determine if the CF3 groups would noticeably alter the
coordination ability of the NOPO fragment due to greater steric
=
demands or increased electron withdrawal from the P O donor
center.
In this regard, new multi-component extraction processes
(UNEX) are under study in a number of programs. These typically
address separations of Cs, Sr, and actinide ions from acidic
Results and discussion
The new CF3 derivatized phosphinoylmethyl pyridine
compounds {[2-(CF3)C6H4]2P(O)CH2}C5H4N, 3c, and {[3,5-
(CF3)2C6H3]2P(O)CH2}C5H4N, 3d were isolated in good yield
(86 and 60%, respectively) from the combination of the required
phosphinoyl Grignard reagents [2-(CF3)C6H4]2P(O)MgBr and
[3,5-(CF3)2C6H3]2P(O)MgBr as summarized in Scheme 1.
aDepartment of Chemistry, University of New Mexico, Albuquerque, NM
87131, USA. E-mail: rtpaine@unm.edu
bIdaho National Laboratory, P.O. Box 1625, Idaho Falls, ID 83415, USA
cOak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831,
USA
† Electronic supplementary information (ESI) available: Crystallographic
and computational data. CCDC reference numbers 715682–715686. For
ESI and crystallographic data in CIF or other electronic format see DOI:
10.1039/b905947d
Following work-up and chromatographic purification, the com-
pounds were obtained as white (3c) and brown (3d) solids. The
7486 | Dalton Trans., 2009, 7486–7493
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
calculations suggest the appearance of a strong, low frequency
N–O stretching vibration and a weaker, higher frequency N–O
stretching vibration: 1a 1275, 1320; 1c 1273, 1329; 1d 1274,
1314 cm-1. These also can be compared against the measured
(1265 cm-1)11 and computed (1322 cm-1) vNO frequency for pyridine
N-oxide. Given the larger error associated with computed vNO
frequencies ( 50 cm-1), the small differences should not be ascribed
to electronic shifts in the pyridine N–O band due to the addition
of one or two CF3 groups to each aryl group.
The 31P NMR spectra for 3c, 3d, 1c and 1d all display a single
resonance that appears at d 30.7, 27.2, 33.0 and 28.7, respectively.
These can be compared against the values for 3a, d 30.2 and
1a, d 31.7. In each case, N-oxidation leads to a small downfield
shift (1–3 ppm) and a similar effect appears with N-oxidation of
the two-armed NPOPO ligands.1,3,4 It is interesting to note that
the resonances for 3,5-(CF3)2C6H3 derivatives 3d and 1d appear
slightly upfield of the resonances for the 2-(CF3)C6H4 3c and 1c
and the unsubstituted aryl 3a and 1a derivatives. A clear rational
for this observation is not apparent at this time but a similar
observation is found with related dithiophosphinic acids.12 The
Scheme 1
isolated intermediates were then N-oxidized with m-CPBA and
the target ligands 1c and 1d were isolated as analytically pure
crystalline solids in 60 and 58% yields, respectively. The ligands are
soluble in acetone, ethyl acetate, chloroform, methylene chloride,
acetonitrile, alcohols and DMF and insoluble in benzene and
hexane. Further, and most importantly, they are soluble in FS-
13 (1c, 0.26 M; 1d, 0.07 M) while the derivatives 1a and 1b have
solubilities < 0.001 M in FS-13.
The identities of the phosphinoylmethyl pyridines and phos-
phinoylmethyl pyridine N-oxides have been confirmed by spectro-
scopic analyses. Each of 3c, 3d, 1c and 1d display intense [M + H+]
and [M + Na+] ions in high-resolution electrospray ionization mass
spectra. Based upon earlier studies of NPO and NOPO ligands
having aryl substituents,1,3 it was expected that the vPO infrared
stretching frequency would appear in the region 1150–1200 cm-1.
In particular, the parent compounds [Ph2P(O)CH2]C5H4N (3a)
and [Ph2P(O)CH2]C5H4NO (1a) show bands assigned to vPO modes
at 1190 cm-1 (3a) and 1186 cm-1 (1a), respectively. It might be
expected that placement of electron-withdrawing CF3 groups on
1
1H and 13C{ H} NMR spectra are dominated by overlapping
resonances from the substituted aryl and pyridine rings and only
partial assignments have been made. The exo-CH2P(O) group
provides a signature doublet resonance due to coupling with the
P-atom: 1H d 4.10 (J = 14.5 Hz) (3c), 4.09 (J = 15.5 Hz) (3d), 4.35
1
(J = 14.3 Hz) (1c), 4.30 (J = 15.3 Hz) (1d); 13C{ H} d 41.6 (J =
68.1 Hz) (3c), 41.6 (J = 66.6 Hz) (3d), 33.4 (J = 92.3 Hz) (1c) and
33.9 (J = 68.7 Hz) (1d). These data can be compared with data for
1
3a: 1H d 3.88 (J = 14.2 Hz), 13C{ H} d 40.7 (J = 64 Hz) and 1a:
1
1H d 4.22 (J = 14 Hz), 13C{ H} d 30.7 (J = 66 Hz). Additional
assignments for aryl and pyridine ring resonances provided in
the Experimental section were aided by assignments for related
secondary phosphine oxides.13
The recent observation by one of us12 of short intramolecular
non-bonded P(V) ◊ ◊ ◊ F interactions between two fluorine atoms
and the phosphorus atom in [2-(CF3)C6H4]2P(S)(SH) of 2.968 and
˚
the aryl rings would result in a significant up-frequency shift of vPO
.
3.165 A led here to determinations of the molecular structures
However, in the new CF3 substituted analogs, the vPO frequencies
are assigned to strong bands at 1177 cm-1 (3c), 1186 cm-1 (3d),
1190 cm-1 (1c) and 1186 cm-1 (1d). These values are shifted little
from the vPO for 3a and 1a, and they are comparable to vPO for
Ph3PO (1191 cm-1)9 and (4-CF3C6H4)3PO (1199 cm-1).9 The lack of
a clear trend in the observed data suggest that neither N-oxidation
or CF3-substitution on the aryl substituents dramatically alter the
of 3c, 3d, 1c and 1d. In general, the molecular structures are
similar and views are shown in Figs. 1–4. A summary of selected
=
electron density in the P O donor groups.
In order to support the band assignments and to test the con-
clusions drawn from the spectra, Gaussian calculations for each
molecule were performed. The computed vPO values are 1194 cm-1
(3a), 1233 cm-1 (3c), 1215 cm-1 (3d), 1191 cm-1 (1a), 1239 cm-1
(1c) and 1201 cm -1 (1d) that are little shifted from vPO for 3a and
1a. Given the expected computational error for vPO ( 30 cm-1),
these calculated frequencies can be considered to be the same, and
comparable to the computed value of 1187 cm-1 for Ph3PO.10
Strong CF3 stretching modes for 3c, 3d, 1c and 1d seriously
complicate the fingerprint region,9 1350–1120 cm-1, where the
N–O stretching vibration also would be expected to appear.11
In the measured spectra several bands appear in this region
for both pyridine and N-oxidized compounds and a confident
assignment of vNO for 1c and 1d can not be made. Gaussian
Fig.
1
Molecular structure and atom labeling scheme for
{[2-(CF3)C6H4]2P(O)CH2}C5H4N, 3c.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 7486–7493 | 7487
©

View Article Online
Fig. 4 Molecular structure and atom labeling scheme for {[3,5-(CF3)2-
C6H3]2P(O)CH2}C5H4NO, 1d.
Fig. 2 Molecular structure and atom labeling scheme for {[3,5-(CF3)2-
C6H3]2P(O)CH2}C5H4N, 3d.
pendent [2-(CF3)C6H4]2P(O)CH2- or [3,5-(CF3)2C6H3]2P(O)CH2-
=
arm attached at the 2-position of the pyridine ring. The P O
bond vector is rotated away from the plane of the pyridine or
pyridine N-oxide by 50.7◦ (3c), 61.0◦ (3d), 44.2◦ (1c) and 69.4◦
(1d).14 The twist of these planes originates from P O/N-lone pair
=
=
or P O/N– bond pair non-bonded repulsions and the twist angles
are similar to those observed for [Ph2P(O)CH2]2C6H3N (73.3 and
-65.9◦).15 It is noted that the 2-CF3 groups in 3c and 1c appear
=
to constrain the orientation of the P O group relative to the
pyridine/pyridine N-oxide ring to a greater degree as evidenced
=
by the smaller twist angle. The P O bond lengths in 3d, 1c and
˚
1d are the same (av. 1.484 A) while the value in 3c is slightly
˚
shorter, 1.4734(2) A. Although the CF3 groups in 3d and 1d are
disordered, there appear to be no close non-bonded interactions
between the F-atoms and the P-atom in these two molecules. This
is to be expected given the 3,5-positions of the CF3 substituents
on the aryl rings. However, in 3c and 1c several F atoms make
close approaches to the backside of the P-atom. The non-bonded
˚
˚
P ◊ ◊ ◊ F distances in 3c, P ◊ ◊ ◊ F4, 2.889 A, and P ◊ ◊ ◊ F2, 3.147 A, and
˚
˚
in 1c, P ◊ ◊ ◊ F1, 2.908 A and 1c P ◊ ◊ ◊ F5, 3.191 A are shorter than the
sum of covalent radii, 3.27 A.16 These values compare well against
˚
Fig. 3 Molecular structure and atom labeling scheme for {[2-(CF3)-
C6H4]2P(O)CH2}C5H4NO, 1c.
short non-bonded distances in [2-(CF3)C6H4]2P(S)(SH) (2.968 and
12
˚
3.165 A).
bond lengths is given in Table 1. The structure of each molecule
consists of a planar pyridine or pyridine N-oxide ring with a
Before undertaking detailed extraction analyses with 1c and
1d and performance comparisons against 1a, it was important
˚
Table 1 Selected bond lengths (A) for 3c, 1c, 3d, 1d and [Yb(1c)(NO3)3(DMF)]·DMF·0.5H2O
[Yb(1c)(NO3)3(DMF)]·
DMF·0.5H2O
Bond
3c
1c
3d
1d
=
P
P–C
O
P–O
1.4734(2)
1.8262(15)
1.8396(15)
1.8175(15)
P1–O2
P1–C6
P1–C7
P1–C14
N1–O1
N1–C1
N1–C5
C5–C6
1.4822(9)
P1–O1
P1–C6
P1–C7
P1–C15
1.4874(14)
1.8077(18)
1.8178(17)
1.8070(15)
P–O2
P–C6
P–C7
P–C15
N–O1
N–C1
N–C5
C5–C6
1.4825(12)
1.8325(17)
1.8165(17)
1.8052(17)
1.303(2)
1.359(3)
1.364(3)
P–O1
P–C6
P–C7
P–C14
N1–O2
N1–C1
N1–C5
C5–C6
1.5016(11)
P–C6
P–C12
P–C19
1.8180(13)
1.8177(12)
1.8180(13)
1.3064(14)
1.3591(17)
1.3682(17)
1.4885(17)
1.8123(14)
1.8236(16)
1.8127(17)
1.3318(16)
1.351(2)
1.353(2)
1.489(2)
N–O
N–C
N–C1
N–C5
C5–C6
1.339(2)
1.342(2)
1.507(2)
N1–C1
N1–C5
C5–C6
1.344(3)
1.334(3)
1.510(2)
C–C
1.488(2)
Yb–O bond lengths for [Yb(1c)(NO3)3(DMF)]·DMF·0.5H2O
Yb–O1
Yb–O2
2.2499(11)
2.2660(11)
Yb–O3
Yb–O4
2.2575(11)
2.4079(14)
Yb–O5
Yb–O7
2.3997(13)
2.4908(13)
Yb–O8
Yb–O10
2.4055(12)
2.3831(13)
Yb–O11
2.4362(13)
7488 | Dalton Trans., 2009, 7486–7493
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
to examine the coordination behavior of 1c toward Ln(III) ions. If
differences are to appear, they should come with a small lanthanide
ion and 1c where the 2-CF3 substituents might be expected to
produce steric congestion in the neighborhood of the coordinated
Ln(III) ion. Initially, 1c and Yb(NO3)3 were combined in a 1 : 1 ratio
in MeOH. The mixture produced a solid that was recrystallized
from DMF. Infrared spectra for the complex show numerous over-
lapping bands in the region 1500–1000 cm-1 and assignments of
significant bond lengthening relative to the bond length in the free
˚
ligand 1c, 1.4822(9) A and the N-oxide bond length, 1.3318(16)
˚
˚
A, is elongated relative to that in 1c, 1.3064(14) A consistent with
strong ligand–Yb(III) ion interactions. It is also noteworthy that
=
the twist angle between the coordinated P O and N–O groups
is 66.8◦ which is larger than in the free ligand. Lastly, it is found
that both CF3 groups of the coordinated ligand are disordered
over two positions.17 Nonetheless, two F-atom positions on
˚
vNO and vPO and hence coordination shifts could not be confidently
each group result in non-bonded separations, F2 ◊ ◊ ◊ P 3.109 A,
˚
˚
˚
made. However, X-ray diffraction quality crystals were obtained
and the molecular structure of the complex was determined. A
view of the molecule is shown in Fig. 5 and selected bond lengths
are listed in Table 1. The asymmetric unit consists of one molecule
of [Yb(1c)(NO3)3(DMF)] and an outer-sphere DMF molecule
hydrogen bonded to a partially occupied (50%) water molecule.
The inner sphere around the Yb(III) contains one molecule of 1c
F5 ◊ ◊ ◊ P 3.069 A, F8 ◊ ◊ ◊ P 2.929 A and F12 ◊ ◊ ◊ P 2.881 A, that are
shorter than the sum of van der Waals radii. This is consistent
with the retention of backside P(V) ◊ ◊ ◊ F interactions in the
complex.
Given that our prior work with the related bidentate ligand
1a led to isolation and structure determinations for 2 : 1 and
4 : 1 ligand–lanthanide complexes,1 efforts were also made in
the present study to prepare 2 : 1 and 4 : 1 complexes using
the new ligand 1c, several lanthanide nitrates and a wide range
of reaction solvents. In each case, complexation was indicated by
very rapid formation of precipitates that resisted dissolution and
crystal growth. Hence, the identity of these complexes remains
undefined. The apparent low solubility of the complexes under
synthesis conditions could complicate the potential use of the
ligand as a solvent extractant since the extraction complexes
must remain soluble in the organic process solvent. However, we
note here that under extraction conditions (~0.1 M 1c in FS-13
and low [Ln(III)]), the complexes are soluble in the organic
phase.
-
and three NO3 ions all coordinated in a bidentate fashion and
an O-bonded DMF. The resulting nine-coordinate coordination
polyhedron approximates a mono-capped square antiprism. The
˚
Yb–O1(P) bond length, 2.2499(11) A, and the Yb–O2(N) bond
˚
length, 2.2660(11) A, can be compared to the related bond lengths,
˚
˚
2.252(5) and 2.2354 A [Yb–O(P)] and 2.260(4) A [Yb–O(N)]
in Yb(2a)2(NO3)3 where 2a = [Ph2P(O)CH2]2C5H3NO. There,
2a acts as a tridentate ligand.1 Structural data also have been
reported for 2 : 1 and 4 : 1 inner-sphere complexes involving
the bidentate parent ligand 1a:3 Pr(1a)2(NO3) Pr–O(P) 2.439(6),
2.474(6) A and Pr–O(N) 2.467(7), 2.426(7) A; Tb(1a)4 Tb–O(P)
2.335(5), 2.379(5), 2.399(5), 2.338(5) A and Tb–O(N) 2.398(6),
3+
˚
˚
˚
˚
2.420(6), 2.411(6), 2.395(6)A. Themajorityofthecomparable(i.e.,
Initially it was thought that the CF3 groups of 1c might
reduce the effective donor strength of the ligand through electron
M–O(P) or M–O(N)) differences between the Pr–O, Tb–O and
Yb–O bond lengths result from the differences in lanthanide ionic
radii. However, in most case, in the same molecule, the M–O(P)
bond length is shorter than the M–O(N) bond length consistent
with the P O donor group acting as the stronger donor. The
phosphoryl group bond length P–O1, 1.5016(11) A also shows
=
withdrawal from the P O donor group. It was also considered that
2-CF3 substitution on the phosphoryl aryl rings might impose
steric hindrance to formation of complexes with Ln(III) ions.
However, these factors apparently do not impact the ability of
1c to form at least the 1 : 1 complex isolated and structurally
characterized here. The next question then becomes: can one or
both of these factors be responsible for the inability to isolate 2 : 1
and 4 : 1 complexes which are formed with the parent bidentate
ligand 1a? In an attempt to address, at the least, the steric issue, we
have completed molecular modeling calculations for 2 : 1 (Fig. 6)
and 4 : 1 (Fig. 7) compositions. The starting structures were those
found from the crystal structure determinations for Pr(1a)2(NO3)3
and Tb(1a)43+. These structures were optimized with an extended
MM3 force field to obtain minimum energy gas-phase structures.
At this point, CF3 groups were added at the 2-position of each
phenyl ring and the structures re-optimized. The calculations, as
summarized in Table 2, indicate that there is no steric reason for
the 2 : 1 or 4 : 1 complexes not to form. In fact, the calculations
suggest that the CF3 substituted ligands provide slightly more
sterically favorable structures.
=
˚
Table 2 Computed molecular mechanics strain energies (kcal mol-1)
Complex
Ucomplex
nUligand
DU = Ucomplex - nUligand
Ln(1a)2(NO3)3
Ln(1c)2(NO3)3
[Ln(1a)4[(NO3)3
[Ln(1c)4](NO3)3
45.63
76.53
81.89
29.26
63.66
58.52
16.4
12.9
23.4
17.6
Fig.
5 Molecular structure and atom labeling scheme for
[Yb(1c)(NO3)3(DMF)]·DMF·0.5H2O (outer-sphere DMF and 0.5 H2O
144.89
127.32
omitted for clarity).
This journal is The Royal Society of Chemistry 2009
Dalton Trans., 2009, 7486–7493 | 7489
©

View Article Online
presence of the electron-withdrawing CF3 substituents, the CF3
decorated ligands form complexes with Ln(NO3)3. One complex,
[Yb(1c)(NO3)3(DMF)](DMF)(H2O)1 , was isolated, structurally
2
characterized and the ligand confirmed to act as a bidentate
chelate. These results suggest additional modifications to the
parent structure 1c that should further enhance ligand solubility
and complexation behavior. These modified systems currently
under development will be the focus of detailed extraction analyses
in the future.
Experimental
The organic reagents and solvents were obtained from Aldrich
Chemical Co. or VWR and they were used as received. Solvents
were dried by standard procedures and stored under dry nitrogen.
NMR spectra were recorded on a Bruker FX-250 spectrometer
using H3PO4 (31P) and Me4Si (1H, 13C) as external standards.
Downfield shifts from the reference were designated as +d. Mass
spectra were obtained from the University of New Mexico Mass
Spectrometry Facility. Infrared spectra were recorded on a Bruker
Tensor 27 spectrometer. Elemental analyses were determined by
Galbraith Laboratories.
Fig. 6 Computed, energy minimized structure for the 2 : 1 complex,
Pr(1c)2(NO3)3.
Ligand syntheses
2-(Bromomethyl)pyridine was prepared by neutralization of 2-
(bromomethyl)pyridine hydrobromide with a saturated aqueous
1
solution of NaHCO3 . CAUTION: This reagent and its solutions
should be handled in a well ventilated fume hood. The compound
is an aggressive skin irritant and contact should be avoided. The
compound has a small vapor pressure at 23 ◦C making it a
potential bronchial irritant as well. The HBr neutralized ligand
should be used immediately as it has a short shelf-life.
2-[Bis(2-(trifluoromethylphenyl))phosphinoylmethyl]pyridine (3c)
A solution of 2-bromobenzotrifluoride (8.03 mL, 59.0 mmol) in
dry THF (10 mL) was added dropwise to a stirred suspension of
Mg turnings (1.43 g, 59.0 mmol) in dry THF (10 mL) and stirred
(1 h). The temperature rose during the addition and the mixture
was then cooled (23 ◦C). A solution of diethyl phosphite (2.01 mL,
15.7 mmol) in dry THF (10 mL) was added dropwise and this
mixture was first refluxed (1 h) and then cooled (23 ◦C). A solution
of 2-(bromomethyl)pyridine (1.0 g, 7.8 mmol) in dry THF (10 mL)
was added dropwise (23 ◦C) and the resulting mixture stirred and
refluxed (12 h). The mixture was quenched with saturated aqueous
NH4Cl solution (100 mL) and the aqueous phase extracted with
CH2Cl2 (3 ¥ 20 mL). The combined organic phases were dried
(Na2SO4) and vacuum evaporated. The resulting colorless oil was
purified by flash chromatography (silica gel, 70–230 mesh, elution
MeOH–CH2Cl2 1 : 99). The compound 3c was obtained as a white
Fig. 7 Computed, energy minimized structure for the 4 : 1 complex,
3+
Tb(1c)4
.
Conclusion
◦
The objectives of this study centered on decoration of the aryl
substituents on phosphorus in ligand 1a with CF3 groups in order
to enhance the solubility of this ligand type in the fluorocarbon
solventFS-13. Itwas alsoofinteresttodetermineifCF3 decoration
would adversely impact the coordination ability of the NOPO-
class ligand. The results of the study show that both 1c and 1d are
more soluble in FS-13 than is 1a with the greater solubility found in
the 2-substituted derivative 1c. In addition, both of the new ligands
are solids and they are more easily purified than 1a. Despite the
powder (2.9 g, 86%) mp 108–110 C. Crystals of 3c were grown
from EtOAc (Found: C, 55.92; H, 3.45; N, 3.31%). C20H14NOF6P
requires C, 55.96; H, 3.29; N, 3.26%. HRMS (ESI): m/z 430.0805
[M + H+], 452.0609 [M + Na+]: C20H15NOF6P requires 430.0795.
IR (KBr, cm-1): 3075 (m), 3029 (sh, w), 2924 (sh, w), 2854 (m),
2785 (w), 2612 (w), 2091 (w), 1998 (w), 1960 (vw), 1924 (vw),
1888 (vw), 1850 (w), 1780 (vw), 1738 (m), 1657 (m), 1586 (s),
1474 (s), 1439 (s), 1402 (m), 1311 (vs), 1265 (vs), 1227 (vs), 1177
(sh, s), 1128 (vs), 1034 (vs), 888 (m), 824 (s), 776 (vs), 707 (s),
7490 | Dalton Trans., 2009, 7486–7493
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
636 (m), 592 (s). 31P NMR (CDCl3): d 30.7. 1H NMR (CDCl3): d
8.30 (d, JHH = 5.0 Hz, 1 H, py), 8.16 (dm, JHP = 14.0 Hz, 2 H),
3068 (s), 3025 (s), 2973 (m), 2924 (m), 2884 (s), 2792 (w), 2456 (w),
2094 (w), 1993 (m), 1951 (w), 1880 (m), 1812 (w), 1735 (m),
1692 (w), 1652 (w), 1585 (m), 1490 (s), 1438 (s), 1313 (vs), 1262 (vs),
1256 (vs), 1235 (vs), 1190 (vs), 1116 (vs), 1035 (s), 903 (w), 863 (m),
7.75 (m, 2 H), 7.62–7.49 (m, 6 H), 7.06 (m, 1 H), 4.09 (d, 2JHP
=
1
14.5 Hz, 2 H). 13C{ H} NMR (CDCl3): d 152.4 (d, 2JCP = 7.2 Hz,
py), 149.4 (py), 136.6 (py), 134.6 (d, JCP = 8.2 Hz), 132.2 (d, JCP
=
825 (m), 817 (m), 783 (s), 727 (m), 708 (m), 644 (m), 602 (m). 31
P
1
1
10.7 Hz), 131.9 (d, JCP = 77.4 Hz), 131.5 (d, JCP = 11.6 Hz),
NMR (CDCl3): d 32.7. H NMR (CDCl3): d 8.2 (m, 3 H), 7.85
(ddd, 3JHH = 7.5 Hz, 4JHH = 2.5 Hz, 4JHH = 2.5 Hz, 1 H), 7.80 (m,
1
131.5 (m), 127.8 (m), 126.0 (py), 124.2 (d, JCF = 258 Hz, CF3),
41.6 (d, JCP = 68.1 Hz, CH2).
2 H), 7.65 (m, 4 H), 7.2 (m, 2 H), 4.35 (d, JHP = 14.3 Hz, 2 H). 13
C
4
NMR (CDCl3): d 143.1 (d, JCP = 6.5 Hz), 139.1 (py), 133.7 (d,
JCP = 9.3 Hz), 132.2 (d, JCP = 2.8 Hz), 131.7 (dq, 2JCF = 34.0 Hz,
2JCP = 6.1 Hz), 131.4 (d, JCP = 11.7 Hz), 131.1 (d, 1JCP = 95.3 Hz),
129.1 (d, JCP = 3.8 Hz, py), 127.7 (m), 126.1 (q, 1JCF = 204 Hz),
31.9 (d, 1JCP = 72.5 Hz).
2-[Bis(3,5-bis(trifluoromethylphenyl))phosphinoylmethyl]pyridine
(3d)
A solution of 3,5-bis(trifluoromethyl)bromobenzene (5.0 mL,
29.0 mmol) in dry THF (10 mL) was combined dropwise with Mg
turnings (0.7 g, 29.0 mmol) in dry THF (10 mL). The temperature
rose during the addition and the mixture was stirred (1 h). The
mixture was cooled (23 ◦C), a solution of diethyl phosphite
(1.0 mL, 7.8 mmol) in dry THF (10 mL) was added dropwise
and then refluxed (1 h). To the resulting mixture, a solution of
2-(bromomethyl)pyridine (1.0 g, 7.8 mmol) in dry THF (10 mL)
was added dropwise (23 ◦C), the mixture stirred and then refluxed
(12 h). The mixture was quenched with saturated aqueous NH4Cl
solution (100 mL), the aqueous phase separated, extracted with
CH2Cl2 (3 ¥ 20 mL) and the combined organic phases dried
(Na2SO4). The solvent was evaporated leaving a dark oil that was
purified by flash chromatography (silica gel, 70–230 mesh, elution
MeOH–CH2Cl2 1 : 99). The compound 3d was obtained as a
brown solid (2.6 g, 60%) mp; 132–134 ◦C. X-Ray quality crystals
were obtained by recrystallization from EtOAc. HRMS (ESI): m/z
566.0536 [M + H+], 588.0352 [M + Na+]: C22H13NOF12P requires
566.0543. IR (KBr, cm-1): 3074 (m), 3020 (m), 2962 (m), 2922 (m),
2854 (m), 2789 (vw), 2644 (vw), 2557 (vw), 2273 (w), 2225 (w),
2118 (vw), 1933 (w), 1843 (w), 1619 (s), 1473 (s), 1437 (s), 1369 (vs),
1284 (vs), 1186 (vs), 1134 (vs), 1000 (w), 847 (m), 787 (m), 739 (m),
1d was isolated as a solid and purified by crystallization from
◦
EtOAc (0.24 g, 60%) mp 156–158 C (Found: C, 45.14; H, 2.25;
N, 2.42). C22H12NO2F12P requires C, 45.46; H, 2.08; N, 2.41.
HRMS (ESI): m/z 582.0487 [M + H+], 604.0312 [M + Na+]:
C22H13NO2F12P requires 582.0492. IR (KBr, cm-1): 3056 (m),
3030 (m), 2961 (m), 2922 (s), 2854 (m), 1938 (w), 1824 (w),
1737 (m), 1617 (s), 1493 (m), 1446 (s), 1366 (s), 1282 (vs), 1176 (vs),
1129 (vs), 950 (w), 908 (m), 885 (m), 836 (m), 770 (s), 683 (s),
612 (m), 583 (w), 571 (m), 527 (s). 31P NMR (CDCl3): d 28.7. 1H
3
4
NMR (CDCl3): d 8.47 (d, JHP = 11.7 Hz, 4 H) 8.01 (d, JHP
=
6.3 Hz, 1 H), 8.05 (s, 2 H), 7.60 (ddd, 3JHH = 7.8 Hz, 4JHH = 2.3 Hz,
4JHH = 2.3 Hz, 1 H), 7.20 (m, 2 H), 4.30 (d, JHP = 15.3 Hz, 2 H).
13C{ H} NMR (CDCl3): d 142.5 (d, JCP = 9.0 Hz), 139.90 (py),
1
1
2
3
134.2 (d, JCP = 100.4 Hz), 132.4 (dq, JCF = 34.2 Hz, JCP
=
2
12.7 Hz), 131.1 (d, JCP = 8.0 Hz), 127.6 (d, JCP = 4.9 Hz, py),
126.3 (d, JCP = 3.1 Hz), 125.8 (d, JCP = 2.6 Hz, py), 125.1 (d, JCP
=
2.9 Hz, py), 122.5 (q, 1JCF = 273.5 Hz), 33.4 (d, 1JCP = 68.7 Hz).
Synthesis of complex
A sample of 1c (0.2 g, 0.4 mmol) in MeOH (5 mL) was combined
with Yb(NO3)3·5H2O (0.2 g, 0.4 mmol) in MeOH (5 mL) and
1
691 (m), 616 (m), 519 (m). 31P NMR (CDCl3): d 27.2. H NMR
3
(CDCl3): d 8.43 (d, JHH = 4.0 Hz, py), 8.30 (d, JHP = 11.5 Hz,
◦
the solution stirred (23 C). During 15 min a precipitate formed
3
3
4 H), 8.05 (s, 2 H), 7.61 (t, JHH = 6.0 Hz, 1 H), 7.32 (d, JHH
=
and stirring was continued (12 h). The resulting white solid was
recovered by filtration (0.3 g, 93% based upon Yb(1c)(NO3)3).
Crystals were obtained by dissolution of the solid in hot DMF
followed by slow evaporation. IR (KBr, cm-1): 3099 (s), 3033 (s),
2941 (s), 2815 (w), 2598 (w), 2521 (m), 2458 (w), 2352 (w), 2314 (m),
2226 (w), 2166 (w), 2103 (w), 2060 (m), 2019 (m), 1984 (m),
1903 (w), 1865 (w), 1666 (vs), 1498 (vs), 1389 (vs), 1312 (vs),
1239 (vs), 1173 (s), 1123 (vs), 1032 (vs), 873 (s), 813 (s), 777 (s),
688 (s), 653 (m), 630 (m), 606 (m), 587 (m), 531 (s), 512 (s).
7.7 Hz, 1 H), 7.10 (t, 3JHH = 4.5 Hz, 1 H), 4.09 (d, 1JHP = 15.5 Hz,
2 H). 13C NMR (CDCl3): d 150.6 (d, 2JCP = 7.3 Hz, py), 149.6 (py),
137.1, 134.5 (d, 1JCP = 98.7 Hz), 132.4 (dq, 2JCF = 34.2 Hz, 3JCP
12.0 Hz), 131.5 (m, py), 126.2 (d, JCP = 3.3 Hz), 125.1 (d, JCP
=
=
4.7 Hz), 124.0 (q, JCF = 272.9 Hz, CF3), 122.7 (d, JCP = 2.5 Hz,
py), 41.0 (d, 1JCP = 66.3 Hz, CH2).
Oxidative syntheses of 1c and 1d
Samples of 3c and 3d (1 equiv.) were dissolved in CH2Cl2 (10 mL)
and combined with m-chloroperbenzoic acid (77%, 1.3 equiv). The
resulting mixtures were stirred and refluxed (12 h) then quenched
with saturated aqueous NaHCO3 solution (30 mL). The phases
were separated, the aqueous phase extracted with CH2Cl2 (3 ¥
20 mL) and the combined organic phases dried (Na2SO4).
1c was isolated as a tan solid following purification by flash
chromatography using gradient elution with MeOH in CH2Cl2
(1–10%) (0.34 g, 58%), mp 142–144 ◦C. X-Ray quality crystals
were obtained by recrystallization from MeOH (Found: C, 53.60;
H, 3.41; N, 3.09). C20H14NO2F6P requires C, 53.95; H, 3.17; N,
3.15. HRMS (ESI): m/z 446.0749 [M + H+], 468.0566 [M + Na+]:
C20H15F6NO2P requires 446.0745. IR (KBr, cm-1): 3106 (m),
Single-crystal X-ray diffraction
Single-crystals of the ligands and complex were placed in glass
capillaries and sealed. Crystal data were collected on a Bruker
X8 Apex 2 CCD-based X-ray diffractometer outfitted with an
Oxford Cryostream 700 low temperature attachment and a normal
˚
focus Mo X-ray tube (l = 0.71073 A) operated at 1.5 kW. The
data frames were integrated with Bruker SAINT software and
processed with SADABS. The structures were solved and refined
with Bruker SHELXTL.18 Lattice and data collection parameters
are presented in Table 3. Some specific comments on each structure
determination follow.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 7486–7493 | 7491
©

View Article Online
Table 3 Crystallographic and X-ray data collection parameters
[Yb(1c)(NO3)3(DMF)]·
DMF·0.5H2O
3c
1c
3d
1d
Identification code
CCDC number
Empirical formula
Mr
rpsl21
715684
C20H14NOF6P
429.29
Triclinic
¯
P1
8.9554(4)
9.6902(4)
11.0339(4)
88.599(2)
74.127(2)
88.578(2)
920.57(7)
2
1.549
0.219
223
19527
rpsl32
715686
C20H14NO2F6P
445.29
Triclinic
¯
P1
7.8219(6)
8.8196(7)
14.4992(12)
84.198(4)
81.657(4)
72.516(4)
942.18(13)
2
1.570
0.221
225
25744
rpsl20a
715683
C22H12NOF12P
565.30
Monoclinic
P21/c
11.6362(12)
19.362(3)
10.6850(13)
90
104.192(6)
90
2333.9(5)
4
1.609
0.230
rpsl22
rpec17
715685
C22H12NO2F12P
581.30
Monoclinic
C2/c
18.7789(6)
24.0769(8)
11.4618(4)
90
117.727(2)
90
4587.2(3)
8
715682
C26H29N6O13.5F6PYb
959.56
Monoclinic
C2/c
29.2303(17)
13.5964(8)
21.6915(12)
90
124.507(2)
90
7104.0(7)
8
1.794
2.781
225
100265
13060
0.0336
0.0204
0.0500
Crystal system
Space group
˚
a/A
˚
b/A
˚
c/A
a/◦
b/◦
g /◦
3
˚
V/A
Z
Dc/g cm-3
1.683
0.239
223
38656
m/mm-1
T/K
223
32662
6258
0.0172
0.0537
0.2327
Reflections collected
Reflections unique
Rint
R1 [I > 2s(I)]
wR2 (all data)
6424
7013
5720
0.0162
0.0500
0.1615
0.0237
0.0426
0.1254
0.0219
0.0473
0.1354
3c. Colorless prism 0.12 ¥ 0.28 ¥ 0.48 mm. The structure
solution and refinement were well behaved. All non-hydrogen
atoms refined anisotropically and H-atoms were included and
refined in ideal positions with fixed Uiso = 1.2Ueq of C. There
was no disorder present.
occupancies C13F1F2F3 (77%), C13F4F5F6 (23%); C20F7F8F9
(75%), C20F10F11F12 (25%).
Full descriptions of the disordered groups in 3d, 1d and the
complex are presented in ESI.†
Computational analyses
1c. Colorless prism 0.138 ¥ 0.276 ¥ 0.509 mm. The structure
solution and refinement were well behaved. All non-hydrogen
atoms refined anisotropically and H-atoms were included and
refined in ideal positions with fixed Uiso = 1.5Ueq of CH2 and
1.2Uiso of CH. There was no disorder in the structure.
Geometry optimization of ground-state structures and corre-
sponding vibrational frequencies were carried out with the
Gaussian 03 suite of programs.19 Analytic calculations of the
energy Hessian confirmed that minimum energy structures had
no imaginary frequencies. Given that Pople type triple-z are
adequate for phosphorus atoms,20 all calculations were performed
in the gas-phase with the B3YLP hybrid density functional21
and 6-311G(d,p) and 6-311++G(d,p) Pople basis sets. Based on
evaluation of the resulting vibrational frequency values produced
for both basis sets for 2a and several model compounds, i.e.
pyridine N-oxide and triphenylphosphine oxide, the 6—311G(d,p)
basis set offered the best balance between computational cost and
frequency accuracy. The geometries and energies were not cor-
rected for basis set superposition error (BSSE) since the larger basis
sets combined with Density Functional Theory were expected
to produce smaller errors.22 The harmonic frequency algorithm
employed by Gaussian 03 generates a set of atom motions for
each frequency. The visualizer option in GaussView 3.023 was
employed to identify the frequency values corresponding to vPO
and vNO. The geometry optimized coordinates and frequencies
for the compounds of interest in this paper can be found in the
supplemental information.
˚
3d. Colorless prism 0.276 ¥ 0.440 ¥ 0.460 A. All non-hydrogen
atoms were refined anisotropically and H-atoms were included and
refined in ideal positions with fixed Uiso = 1.2Ueq of C. Three of the
four CF3 groups were rotationally disordered (F on C13, C14 and
C21). The F atoms are mostly in two orientations slightly rotated
from each other.
1d. Colorless prism 0.161 ¥ 0.230 ¥ 0.322 mm. All non-
hydrogen atoms were refined anisotropically and H-atoms were
included and refined in ideal positions with fixed Uiso = 1.2Ueq of
C. Residual difference peaks were seen about all CF3 groups and
those about C21 and C22 were larger. These two CF3 were refined
in two positions with 51/49% occupancies.
Yb(1c)(NO3)3(DMF). Irregular shape 0.092
¥
0.12
¥
0.414 mm. The asymmetric unit contains [Yb(1c)(NO3)3(DMF)]
and one solvent DMF H-bonded to a partially occupied water
molecule. All non-hydrogen atoms were refined anisotropically
except on the water (O14/O15). All hydrogen atoms were included
in ideal positions: terminal methyl hydrogens with Uiso = 1.5Ueq
of C and all other H-atoms on C with Uiso = 1.2Ueq. The partially
occupied water (50% occupancy disordered over three positions)
was refined isotropically. The F atoms on the C13 and C20 were
disordered and treated with a two-position model with disorder
Acknowledgements
Financial support for this study at UNM was provided by
the U.S. Department of Energy (DOE), Chemical Sciences,
Geosciences and Biosciences Office, Office of Basic Energy
Sciences (Grant DE-FG02–03ER15419).
7492 | Dalton Trans., 2009, 7486–7493
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
15 E. Szbyk, Z. Y. Zhang, G. J. Palenik, R. C. Palenik and S. O. Colgate,
Acta Crystallogr., Sect. C, 1989, 45, 1234.
References
1 B. M. Rapko, E. N. Duesler, P. H. Smith and R. R. Ryan, Inorg. Chem.,
1993, 32, 2164.
2 K. L. Nash, C. Madic, and J. N. Mathur, Actinide Separation Science
in The Chemistry of the Actinide and Transactinide Elements, ed.
L. R. Morss, N. M. Edelstein and J. Fuger, Springer, Dordrecht, 2006,
ch 24, p. 2622.
3 X. Gan, E. N. Duesler and R. T. Paine, Inorg. Chem., 2001, 40, 4420.
4 K. L. Nash, C. Lavalette, M. Borkowski, R. T. Paine and X. Gan, Inorg.
Chem., 2002, 41, 5849.
16 J. Boudi, Phys. Chem., 1964, 68, 441.
17 Disorder occupancies are: C13F1F2F3 (77%), C13F4F5F6 (23%);
C20F7F8F9 (75%), C20F10F11F12 (25%).
18 The crystallographic software used in these structure determinations
include: G. M. Sheldrick, SHELXTL, v. 6.14, Bruker Analytical X-ray
Madison, WI, 2001; G. M. Sheldrick, SADABS, v. 2.10. Program for
Empirical Absorption Correction of Area Detector Data, University
of Go¨ttingen, Go¨ttingen, Germany, 2003; SAINTPlus, v. 7.01, Bruker
Analytical X-ray, Madison, WI, 2003.
5 V. N. Romanovskiy, I. V. Smirnov, V. A. Babain, T. A. Todd, R. S.
Herbst, J. D. Law and K. N. Brewer, Solv. Extr. Ion Exch., 2001, 19, 1.
6 J. D. Law, R. S. Herbst, T. A. Todd, V. N. Romanovskiy, V. A. Babain,
V. M. Esimantovskiy, I. V. Smirnov and B. N. Zaitsev, Solv. Extr. Ion
Exch., 2001, 19, 23.
7 R. S. Herbst, J. D. Law, T. A. Todd, V. N. Romanovskiy, V. A. Babain,
V. M. Esimantovskiy, I. V. Smirnov and B. N. Zaitsev, Solv. Extr. Ion
Exch., 2002, 20, 429.
8 V. N. Romanovskiy, V. A. Babain, M. Yu. Alyapyshev, I. V. Smirnov,
R. S. Herbst, J. D. Law and T. A. Todd, Sep. Sci. Technol., 2006, 41,
2111.
9 J. M. Casper and E. E. Remsen, Spectrochim Acta, Part A, 1978, 34, 1.
10 V. K. Manchanda and M. S. Subramanian, Aust. J. Chem., 1974, 27,
1573.
11 R. M. Silverstein, G. C. Bassler and T. C. Morrill, Spectroscopic
Identification of Organic Compounds, J. Wiley, New York, 5th edn,
1991, p 130.
12 J. R. Klaehn, D. R. Peterman, M. K. Harrup, R. D. Tillotson, T. A.
Luther, J. D. Law and L. M. Daniels, Inorg. Chim. Acta, 2008, 41, 2522.
13 C. A. Busacca, J. C. Lorenz, N. Grinberg, N. Haddad, M. Hrapchak, B.
Latli, H. Lee, P. Sabila, A. Saha, M. Sorvestavi, S. Shen, R. Varsolona,
X. Wei and C. H. Senanayoke, Org. Lett., 2005, 7, 4277.
14 These angles are between the planes C6NC1C2C3C4C5 and C6PO or
C6P1O1 or C6PO2 or C6P1O2.
19 Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel,
G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr.,
T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, M. KIene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross,
V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y.
Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G.
Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas,
D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V.
Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov,
G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J.
Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M.
Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C.
Gonzalez and J. A. Pople, Gaussian, Inc., Wallingford CT, 2004.
20 A. D. McLean and G. S. Chandler, J. Chem. Phys., 1980, 72, 5639.
21 S. P. Sousa, S. A. Fernandes and M. J. Ramos, J. Phys. Chem. A, 2007,
111, 10439.
22 Y. Zhano and D. G. Truhlar, J. Chem. Theory Comput., 2005, 1, 415.
23 GaussView, Version 3.09, R. Dennington II, K. Roy, M. Todd, E. John,
H. Ken, W. Lee and R. Gilliland, Semichem, Inc., Shawnee Mission,
KS, 2003.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 7486–7493 | 7493
©