Anion Receptor Design: Exploiting Outer-Sphere Coordination Chemistry to Obtain High Selectivity for Chloridometalates over Chloride

Article abstract of DOI:10.1021/acs.inorgchem.5b01317

High anion selectivity for PtCl₆^2- over Cl^- is shown by a series of amidoamines, R¹R²NCOCH₂CH₂NR³R⁴ (L1 with R¹ = R⁴ = benzyl and R² = R³ = phenyl and L3 with R¹ = H, R² = 2-ethylhexyl, R³ = phenyl and R⁴ = methyl), and amidoethers, R¹R²NCOCH₂CH₂OR³ (L5 with R¹ = H, R² = 2-ethylhexyl and R³ = phenyl), which provide receptor sites which extract PtCl₆^2- preferentially over Cl^- in extractions from 6 M HCl solutions. The amidoether receptor L5 was found to be a much weaker extractant for PtCl₆^2- than its amidoamine analogues. Density functional theory calculations indicate that this is due to the difficulty in protonating the amidoether to generate a cationic receptor, LH⁺, rather than the latter showing weaker binding to PtCl₆^2-. The most stable forms of the receptors, LH⁺, contain a tautomer in which the added proton forms an intramolecular hydrogen bond to the amide oxygen atom to give a six-membered proton chelate. Dispersion-corrected DFT calculations appear to suggest a switch in ligand conformation for the amidoamine ligands to an open tautomer state in the complex, such that the cationic N-H or O-H groups are also readily available to form hydrogen bonds to the PtCl₆^2- ion, in addition to the array of polarized C-H bonds. The predicted difference in energies between the proton chelate and nonchelated tautomer states for L1 is small, however, and the former is found in the X-ray crystal structure of the assembly [(L1H)₂PtCl₆]. The DFT calculations and the X-ray structure indicate that all LH⁺ receptors present an array of polarized C-H groups to the large, charge diffuse PtCl₆^2- anion resulting in high selectivity of extraction of PtCl₆^2- over the large excess of chloride. (Figure Presented).

Full text of DOI:10.1021/acs.inorgchem.5b01317

Article
pubs.acs.org/IC
Anion Receptor Design: Exploiting Outer-Sphere Coordination
Chemistry To Obtain High Selectivity for Chloridometalates over
Chloride
†
†
†
†
‡
‡
Innis Carson, Kirstian J. MacRuary, Euan D. Doidge, Ross J. Ellis, Richard A. Grant, Ross J. Gordon,
†
,†
†
,†
†
Jason B. Love, Carole A. Morrison,* Gary S. Nichol, Peter A. Tasker,* and A. Matthew Wilson
†EaStCHEM School of Chemistry, University of Edinburgh, David Brewster Road, Edinburgh, EH9 3FJ, United Kingdom
‡Johnson Matthey Technology Centre, Sonning Common, Reading, RG4 9NH, United Kingdom
*
S Supporting Information
2
−
−
ABSTRACT: High anion selectivity for PtCl over Cl is shown by a series of amidoamines,
6
1
2
3
4
1
4
2
3
1
R R NCOCH CH NR R (L1 with R = R = benzyl and R = R = phenyl and L3 with R = H,
R = 2-ethylhexyl, R = phenyl and R = methyl), and amidoethers, R R NCOCH CH OR (L5
2
2
2
3
4
1
2
3
2
2
1
2
3
with R = H, R = 2-ethylhexyl and R = phenyl), which provide receptor sites which extract
PtCl62 preferentially over Cl in extractions from 6 M HCl solutions. The amidoether receptor
−
−
2−
L5 was found to be a much weaker extractant for PtCl6 than its amidoamine analogues. Density
functional theory calculations indicate that this is due to the difficulty in protonating the
+
amidoether to generate a cationic receptor, LH , rather than the latter showing weaker binding to
PtCl62 . The most stable forms of the receptors, LH , contain a tautomer in which the added
proton forms an intramolecular hydrogen bond to the amide oxygen atom to give a six-membered
proton chelate. Dispersion-corrected DFT calculations appear to suggest a switch in ligand
conformation for the amidoamine ligands to an open tautomer state in the complex, such that the
−
+
2
−
cationic N−H or O−H groups are also readily available to form hydrogen bonds to the PtCl6
ion, in addition to the array of polarized C−H bonds. The predicted difference in energies between
the proton chelate and nonchelated tautomer states for L1 is small, however, and the former is found in the X-ray crystal
+
structure of the assembly [(L1H) PtCl ]. The DFT calculations and the X-ray structure indicate that all LH receptors present an
array of polarized C−H groups to the large, charge diffuse PtCl6 anion resulting in high selectivity of extraction of PtCl6 over
the large excess of chloride.
2
6
2
−
2−
1
1
12
13
INTRODUCTION
The design of highly selective receptors for anions is of
considerable current interest.
complexation of inorganic anions has targeted small (hard)
anions such as chloride and phosphate, the design features
which will discriminate in favor of larger, softer anions such as
metalates is less frequently studied. An example of particular
commercial importance is the provision of high selectivity for
chloridometalate anions such as PtCl over the much smaller
Cl . Such selectivity is essential in processes to recover
Amines, amides, trialkylphosphates, and phosphine
oxides
■
14,15
have been used as solvent extractants to transport
1
−5
the chloridometalate intact as in the pH-dependent process
shown in eq 1.
for the chloridometalate anion over chloride to ensure that the
chloride transfer shown in eq 3 does not predominate. One
approach to achieve this is to develop receptors LH which
have H-bond donor groups that recognize the centers of
negative charge on the chloridometalate anion. Amido-
functionalized tertiary amines (Figure 1) have been reported
which meet this criterion, and they have indeed been found to
be stronger chloridometalate extractants than the analogous
While most of the work on
6
,7,10
The extractant must show a high selectivity
+
2−
6
−
platinum from aqueous solutions obtained from the oxidative
leaching of minerals containing the platinum group metals
PGMs) with strong hydrochloric acid. The kinetic inertness
6
−9
16−22
(
unsubstituted tertiary amines.
of the PGMs means that it is not practicable to use reagents
which displace chloride ions to form inner-sphere, organic-
soluble, complexes (eq 2) on the time scales used in solvent
A feature of the amidoamine extractants (Figure 1) is that
when the tertiary amine group is protonated to form the
cationic receptor, they are all able to form a strong internal H-
bond to a neighboring amido oxygen atom. This increases the
effective basicity of the reagent and the formation of the
10
extraction processes used to recover base metals.
+
2−
2
Lorg + 2H + [PtCl ] ⇌ [(LH) PtCl ]
aq
6 aq
2
6 org
(1)
(2)
(3)
“
proton chelate” which can align an array of C−H and N−H
groups to match the distribution of charge on the target
2−
2
Lorg + [PtCl ] ⇌ [L PtCl ] + 2Cl−
6 aq
2
4 org
aq
Lorg + H+ + Cl− ⇌ [LH.Cl]org
Received: June 11, 2015
aq
aq
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.5b01317
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
system. H and 13C nuclear magnetic resonance (NMR) spectra were
1
obtained on a Bruker AVA 500 or 600 spectrometer as solutions in
CDCl . Chemical shifts (δ’s) are reported in parts per million (ppm)
3
relative to the residual solvent (δH 7.26 and δC 77.0). Mass spectra
were recorded on a MAT 900 XP spectrometer (EI) or a Thermo-
Fisher LCQ Classic (ESI). Elemental analyses were determined by Mr.
Stephen Boyer at London Metropolitan University, School of Human
Sciences, Science Centre, London Metropolitan University, 29
Hornsey Road, London, N7 7DD. Crystal structure data (CCDC-
1
406119) were collected at 150 K on a three circle Rigaku Oxford
Diffraction SuperNova CCD diffractometer equipped with an Oxford
Cryosystems low temperature device with Cu Kα radiation (λ =
1
.541 78 Å). The new reagents L1, L3, and L5 were prepared by
23
adaptation of methods described previously.
N-Benzyl-3-(benzyl(phenyl)amino)-N-phenylpropanamide
L1). Neat 3-bromopropanoyl chloride (18.7 g, 109 mmol) was added
(
dropwise to a stirred solution of N-benzylaniline (38.5 g, 210 mmol) in
CH Cl (200 mL) at 0 °C and then stirred at room temperature for 1
Figure 1. Tripodal amidoamines which have been used as
chloridometalate extractants and the five-, six-, and seven-membered
2
2
h. The resulting mixture was filtered; N-benzylaniline (37.3 g, 204
mmol) was added to the filtrate, and the mixture refluxed for 3 h. After
filtration and evaporation of solvent, the resulting oil purified on a
silica column, eluting with 20% chloroform in hexane to give the title
“proton chelates” which can be formed during metal extraction by the
protonated reagents.
1
6
compound as a pale yellow/green oil, yield 14.5 g (32%). H NMR
chloridometalate species. This is very clearly demonstrated by
(
δ , 500 MHz, CDCl ) 6.56−7.34 (m, 20H, aromatic H), 4.91 (s, 2H,
2−
H
3
the X-ray crystal structure of a ZnCl4 assembly in which the
13
CH ), 4.46 (s, 2H, CH ), 3.60 (t, 2H, CH ), 2.50 (t, 2H, CH ).
C
2
2
2
2
receptor site directs four polarized C−H and N−H groups
NMR (δ , 125 MHz, CDCl ) 171.2, 137.3, 129.7, 129.3, 128.9, 128.5,
C
3
2−
toward the edges and the center of a face of the ZnCl
+
4
128.4, 128.1, 127.4, 54.3, 53.6, 47.7, 32.0. m/z (ESI) 421.92 (M+H ).
N-(2-Ethylhexyl)prop-2-enamide. Neat 2-ethylhexylamine (6
mL, 37 mmol) was added dropwise to a stirred solution of acryloyl
1
8
tetrahedron. This work examines whether structurally
analogous amidoethers can function in a similar manner. A
potential advantage of these latter reagents is that their lower
basicity might make it easier to subsequently release the
chloridometalate species, allowing them to be stripped by
contacting the loaded organic phase with water, generating
chloroplatinic acid as in eq 4.
chloride (3 mL, 37 mmol) in dry CH Cl (20 mL) at 0 °C. After
2 2
stirring for 30 min at room temperature, the reaction was quenched
with saturated sodium hydrogen carbonate solution (3 × 30 mL) and
the organic phase washed with water (30 mL) and dried over MgSO4.
The solvent was evaporated under vacuum to give the title compound
as a colorless oil which was used without further purification in the
1
preparation of L3, yield 4.84 g (72%). H NMR (δ , 600 MHz,
[
(LH) PtCl ] ⇌ H PtCl + 2L
H
2
6 org
2
6
org
(4)
CDCl ) 6.29 (dd, 1H, HCCH ), 6.12 (dd, 1H, CCH′H), 5.64
3
2
To this end, a hydrocarbon-soluble amidoether reagent L5
(dd, 1H, CCH′H), 3.34−3.26 (m, 2H, CH
2
NH), 1.52−1.48 (m,
1
H, CH), 1.38−1.27 (m, 8H, CH and CH ), 0.95−0.88 (m, 6H, CH ).
(
Figure 2) with a 2-ethylhexyl substituent at site R , along with
2
3
2
1
3
C NMR (δ , 125 MHz, CDCl ) 165.7, 131.0, 126.1, 42.5, 39.4, 31.0,
C
3
2
8.9, 24.3, 23.0, 14.1, 10.9.
N-(2-Ethylhexyl)-3-(methylphenylamino)-propanamide (L3).
A mixture of N-(2-ethylhexyl)prop-2-enamide (4.0 g, 22 mmol), N-
methylaniline (2.6 g, 24 mmol), and SiCl (2 mol %) was stirred and
4
heated under N at 70 °C for 16 h. The mixture was dissolved in
2
CH Cl (20 mL), washed with water (3 × 20 mL), and dried over
2
2
MgSO . The solvent was removed and the crude product purified on a
4
silica column, eluting with 10% methanol in CH Cl to give the title
2
2
1
compound as a pale yellow oil, yield 4.1 g (65%). H NMR (δ , 600
H
MHz, CDCl ) 7.28−7.26 (m, 2H, aromatic H), 6.79−6.76 (3H,
3
aromatic H), 3.69 (t, 2H, CH N(CH )), 3.2 (m, 2H, CH NH), 2.95
2
3
2
(
s, 3H, NCH ), 2.44 (t, 2H, CH CH N(CH )), 1.42−1.38 (m, 1H,
3
2
2
3
13
CH), 1.31−1.20 (m, 8H, CH ), 0.93−0.86 (m, 6H, CH ). C NMR
(
2
3
δ , 151 MHz, CDCl ) 171.5, 148.9, 129.3, 117.2, 113.1, 49.4, 42.4,
C 3
3
9.3, 38.8, 34.1, 31.0, 28.9, 24.2, 23.0, 14.1, 10.8. m/z (ESI) 290.2 (M
+
+
H ).
-Phenoxypropanoic Acid. A mixture of 3-phenyoxynitrile (9.88
g, 67 mmol) in 6 M HCl (50 mL) was heated at reflux temperature for
6 h. After being cooled, the mixture was filtered and the precipitate
Figure 2. Amidoamine (L1−L4) and amidoether (L5−L6) reagents
discussed in this work.
3
1
two amidoamine analogues (L1 and L3) have been prepared in
this work to compare their extraction properties. Density
functional theory (DFT) calculations were undertaken with
shorter chain, n-butyl analogues, L2, L4, and L6, and with the
more conformationally rigid L1, which also afforded crystals of
recrystallized from 40:60 benzene/petroleum ether (60−80 °C) to
1
give the title compound as colorless crystals, yield 6.95 g (62%). H
NMR (δ , 500 MHz, CDCl ) 7.36−7.31 (m, 2H, aromatic H), 7.06−
H
3
7
2
1
.02 (m, 1H, aromatic H), 6.96−6.92 (m, 1H, aromatic H), 4.23 (t,
13
H, CH ), 2.86 (t, 2H, CH ). C NMR (δ , 125 MHz, CDCl ) 177.4,
2
2
C
3
58.4, 129.5, 121.2, 114.7, 63.0, 34.4.
-Phenoxypropanoyl Chloride. A mixture of 3-phenoxypropa-
[
(L1H) PtCl ] suitable for X-ray structure determination.
2 6
3
noic acid (5.80 g, 35 mmol) and thionyl chloride (4.5 mL, 62 mmol)
was stirred at room temperature for 3 h. Excess thionyl chloride was
removed by vacuum distillation to give the title compound as an
orange oil which was used in the preparation of L5 without further
EXPERIMENTAL SECTION
All solvents and reagents were used as received from commercial
suppliers. Deionized water was obtained from a Milli-Q purification
■
B
DOI: 10.1021/acs.inorgchem.5b01317
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
1
20
purification, yield 6.26 g (97%). H NMR (δ , 500 MHz, CDCl )
Turkington et al. In vacuo classical molecular dynamics (MD)
H
3
7
2
1
.44−7.35 (m, 2H, aromatic H), 7.09 (dd, 1H, aromatic H), 6.99 (d,
simulations were employed to investigate the possible formation of
13
2−
H, aromatic H), 4.30 (t, 2H, CH ), 3.36 (t, 2H, CH ). C NMR (δ ,
larger assemblies of L1 and PtCl . These were performed using the
2
2
C
6
3
0
31
25 MHz, CDCl ) 171.7, 158.2, 129.7, 121.7, 114.8, 62.7, 46.7.
OPLS-AA force field with the software package LAMMPS. An
initial model, comprising two L1H , two LH, and one PtCl entities
3
+
2−
N-(2-Ethylhexyl)-3-phenoxypropanamide (L5). Neat 2-ethyl-
6
hexylamine (6.12 g, 47 mmol) was added dropwise to a stirred
solution of 3-phenoxylpropanoyl chloride (6.07 g, 33 mmol) in dry
CH Cl (75 mL). After 3 h at room temperature, the solvent was
randomly distributed in a cubic simulation cell of length 40 Å, was
constructed using Packmol. The integration time step was set to 1 fs,
and time increments accrued using the standard Velocity-Verlet
3
2
2
2
evaporated and the resulting oil purified on a silica column eluting with
0% ethyl acetate in hexane to give the title compound as a pale yellow
algorithm. Aggregation of all species to create a [(L1H)
2
{(L1H)-
1
Cl} PtCl ] system occurred very quickly (within 30 ps); in total,
2
6
1
oil, yield 6.21 g (68%). H NMR (δ , 500 MHz, CDCl ) 7.32−7.25
system dynamics were accrued for 10 200 ps, which included 500 ps
equilibration time. The simulation was run under canonical (NVT)
ensemble conditions (with the temperature thermostated at room
temperature using the Nose−Hoover thermostat system); the
́
potential energy values obtained during the production run were
collected and then averaged.
H
3
(
m, 2H, aromatic H), 6.97 (t, 2H, aromatic H), 6.90 (d, 1H, aromatic
H), 6.22 (s, 1H, NH), 4.25 (t, 2H, CH ), 3.29−3.17 (m, 2H, CH ),
2
2
2
0
1
1
.67 (t, 2H, CH ), 1.49−1.40 (m, 1H, CH), 1.37−1.21 (m, 8H, CH ),
2
2
13
.95−0.83 (m, 6H, CH ). C NMR (δ , 125 MHz, CDCl ) 170.7,
3
C
3
58.2, 129.6, 121.2, 114.5, 64.3, 42.3, 39.3, 36.9, 31.0, 28.9, 24.2, 23.0,
+
4.1, 10.9. m/z (ESI) 300.19 (M + Na ). Anal. Calcd for C H NO :
17
27
2
C 73.61, H 9.81, N 5.05. Found: C 73.71, H 9.75, N 5.09.
RESULTS AND DISCUSSION
The uptake of PtCl6 from 6 M aqueous HCl by toluene
solutions containing varying concentrations of the amidoamine
reagents L1 or L3 or the amidoether reagent L5 (Figure 3)
■
General Extraction Procedure. Analytical grade toluene was
used as the water-immiscible solvent for the extractants and deionized
water for the hexachloroplatinate solutions. Inductively coupled
plasma emission spectroscopy (ICP-OES) calibration standards were
prepared by dilution of commercially available standards. ICP-OES
was carried out on a PerkinElmer Optima 5300 DV employing a radio
frequency (RF) forward power of 1400 W, with argon gas flows of 10,
2−
−1
1
.4, and 0.45 L min for plasma, auxiliary, and nebulizer flows. Using
a peristaltic pump, sample solutions were taken up into a Gem Tip
cross-flow nebulizer and a glass cyclonic spray chamber at a rate of 2.0
−1
mL min . Solutions of extractants were prepared at concentrations
ranging from 0.10 to 1.00 M by dilution of a 1.00 M solution in
toluene. Five mL portions of these solutions were contacted with 5 mL
of chloridoplatinate solution (0.01 M Pt, 6 M HCl) and stirred
vigorously in sealed vials for 16 h at room temperature, after which the
phases were separated and 0.5 mL aliquots of each diluted to 10 mL in
1
-methoxy-2-propanol for ICP-OES analysis.
Computational Modeling. Geometry optimization calculations
were carried out for L1, L2, L4 (which is a shorter chain model system
for L3), and L6 (which similarly is a model system for L5) using the
24
Gaussian 09 program. No attempts were made to model solvation of
assemblies. The justification for this stems from the fact that the
solvation/desolvation energies in the aqueous phase are identical for
+
aq
2−
all extractions, involving only H and [PtCl6] aq, leaving only the
Figure 3. Pt recovery from aqueous solutions of 0.01 M H PtCl in 6
M HCl by equal volumes of toluene solutions of L1, L3, and L5 as a
function of extractant concentration.
differences in the solvation energies of Lorg and [(LH) PtCl ] arising
2
6
2
6 org
from variations of the R and X substituents on the ligands to influence
the correlation between calculated formation energies and the
observed strength of extraction. In the organic phase (toluene which
has a low dielectric constant), any dependence of the overall formation
energy of the complex on solvation is likely to involve only the minor
differences originating from variation of the R and X groups. Atom
coordinates for all energy-minimized structures can be found in the
indicates that L1 and L3 are stronger extractants, requiring
2
−
lower L:Pt molar ratios to remove >90% of PtCl6 from the
aqueous phase. The formation of a third phase that was not
soluble in either the bulk toluene or aqueous phases, was
observed for all of the extractions undertaken, but the material
25
Supporting Information. The B3LYP exchange/correlation func-
tional was used throughout (data reported in the Supporting
2
6
Information), alongside the highly parametrized M06-2X functional
2−
balances were good with the loss of PtCl6 from the aqueous
to assess the impact of a dispersion-corrected functional. As expected,
the dispersion correction had minimal impact on ligand protonation
energies (Table 2 and Table S1, Supporting Information), but a
marked impact on their binding energies (due to the prevalence of soft
C−H···Cl) and complex formation energies (Table 3 and Table S2,
Supporting Information). The 6-31+G(d) basis set was applied to all
atoms with the exception of platinum, for which the LANL2DZ
pseudopotential/basis set was used. Structures were considered
optimized when the forces and atomic displacements fell to within
the program default convergence criteria. Assembly formation energies
and protonation energies were calculated using the difference in
internal energy values of the sum of the products and the sum of the
individual reactants (corrected for basis set superposition error using
phase being equal to that detected in the toluene and the third
phase.
In order to define the origins of the greater extractant
strength of the amidoamines L1 and L3 over their amidoether
analogue, L5, DFT calculations were carried out to compare the
ease of protonation (ΔU ) to generate the active forms (eq 5)
p
and to compare the binding energies (ΔU ’s) of these to the
b
2−
PtCl dianion (eq 6). The overall formation energies (ΔU ’s)
of the neutral assemblies [(LH) PtCl ] for the reaction shown
6
f
2
6
in eq 7 represent the gas-phase equivalents of the solvent
extraction process shown in eq 1. In order to make valid
comparisons of these energies of reaction it is essential that the
lowest energy forms of the various entities shown in eqs 5−7
are used in the calculations. To reduce the numbers of
conformers of the 2-ethylhexyl groups in L3 and L5 which will
27,28
the Counterpoise correction method).
Natural bond order (NBO)
analysis, which uses “Lewis”-type orbitals to describe donor and
acceptor orbitals and their interactions, was performed using NBO
29
6
.0 on optimized structures in a similar manner to that used by
C
DOI: 10.1021/acs.inorgchem.5b01317
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
+
+
+
Table 1. Energies of Formation of Tautomers of Each of the Protonated Amidoamines L1H , L2H , L4H , and Amidoether
+
a
L6H , Derived from Equation 5 Using the M06-2X/6-31+G(d)/LANL2DZ Level of Theory
a
Note that for L4 and L6 energies are quoted relative to the more stable dimeric forms of the unprotonated reagents.
Figure 4. Lowest energy calculated structures of [(L1H) PtCl ] (left) and [(L6H) PtCl ] (right). The contacts a−g and a′−g′ shown have Natural
2
6
2
6
−1
Bond Order (NBO) energies ≥5 kJ mol (Figures S2 and S9, respectively, Supporting Information).
have very similar energies, the n-butyl analogues L4 and L6
were used as model systems in the DFT calculations. L2 was
used as a less rigid analogue of L1 to allow comparisons of
tertiary amides with the secondary amides L4 and L6.
Interesting issues arise when identifying the most stable gas-
phase structures and when considering whether similar
structures are likely to be the most stable forms in toluene,
which was used in the solvent extraction experiments. The
unprotonated reagents that contain secondary amide groups
+
+
L + H O ⇌ LH + H O
(5)
(6)
(7)
3
2
(
i.e., R = H) associate to form dimers, such that L4 and L6
1
form (L4)2 and (L6) , which are more stable than the
+
2−
2
2
LH + PtCl
⇌ [(LH) PtCl ]
6
2
6
intramolecular H-bonded monomers of L4 and L6, with
−1
energies of dimerization of −29.4 and −30.1 kJ mol ,
+
2−
2
L + 2H O + PtCl
⇌ [(LH) PtCl ] + 2H O
3
6
2
6
2
respectively (Figure S1 in Supporting Information). In contrast,
D
DOI: 10.1021/acs.inorgchem.5b01317
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
the tertiary amide reagents L1 and L2, which have no amido
N−H, show no propensity for dimerization.
Table 2. Energies of Formation (M06-2X/6-31+G(d)/
LANL2DZ), ΔU , of the [(LH) PtCl ] Assemblies
f
2
6
Protonation of the reagents can give rise to a number of
different structures, the stabilities of which are dependent upon
which functional group is protonated and whether or not an
intramolecular H-bond is formed. For the amidoamines it was
found in all cases that protonation of the amine nitrogen gave
rise to more stable structures than those protonated at the
carbonyl oxygen (Table 1). For the amidoether (L6),
protonation at the carbonyl oxygen is preferred to that at the
ether group. Protonation energies are more favorable for
amidoamines than for the amidoether, which can be attributed
to the higher basicity of the amine. Note that the protonation
energy for L4 is less favorable than those for L1 and L2 because
the lowest energy state for L4 was taken to be the dimer (L4)2
and two H-bonds in the dimerized form have to be broken to
Containing the Proton Chelated and Nonchelated
Tautomers of L1, L3, L4, and L6; the Protonation Energies
(ΔU ’s) Required To Generate the Most Stable Form of
p
+
these Cations, LH , and Their Binding Energies (ΔU ) to
b
2
−
PtCl
6
structure of
protonated
ligand
reagent
L1
ΔU /kJ mol− ΔU /kJ mol
1
−1
ΔU /kJ mol
a
−1
p
b
f
chelated proton
−325.7
−325.7
−917.6
−925.7
−1569.0
−1577.1
nonchelated
tautomer
L2
L4
L6
chelated proton
−319.7
−319.7
−916.4
−946.0
−1555.9
−1585.4
nonchelated
tautomer
+
chelated proton
−295.9
−295.9
−940.8
−987.7
−1532.6
−1579.5
generate L4H (Figure S1, Supporting Information). Also, the
nonchelated
tautomer
“proton chelate” form, with an internal H-bond, was found to
be the global minimum conformation for all ligand systems.
Geometry optimizations to yield the lowest energy forms of
the [(LH) PtCl ] assemblies at the M06-2X/6-31+G(d)/
chelated proton
−222.9
−222.9
−1044.2
−884.8
−1490.1
−1330.7
nonchelated
tautomer
2
6
a
LANL2DZ level contained, in the cases of amidoamines L1,
L2, and L4, the tautomer of the receptor which lacks a
ΔU = 2ΔU + ΔU .
f
p
b
“
chelated proton” (first column, Table 1). As a consequence,
in contrast to the nonchelated assembly structure which was
calculated to be lower in energy, though as the calculated
difference in energy between the two structures is small (9.1 kJ
the cationic N−H or O−H groups are readily available to form
2
6
−
H-bonds to the PtCl ion, as well as an array of polarized C−
−1
H bonds (see Figure 4, right). Note that at the B3LYP/6-
mol , see Table 2) it is not surprising that the proton chelated
structure could exist. Further weak interactions are formed with
C−H bonds from two other L1 units in the solid state (lower
part of Figure 6). This suggests that the aryl groups in L1 are
not large enough to create a sufficiently hydrophobic exterior in
a 2:1 assembly with a PtCl62 dianion, and so may not be able
to ensure high solubility in toluene. This may account for the
formation of a third phase in the extraction experiments.
3
1+G(d)/LANL2DZ lower level of theory (given in the
Supporting Information) the proton chelate form was the lower
energy state in all cases, highlighting that the balance between
internal hydrogen bond versus C−H···Cl can be tipped when
an improved description of dispersion interactions is
introduced. In contrast, a different behavior appears to be
exhibited by the amidoether L6. In the [(L6H) PtCl ]
−
2
6
assembly, the L6H+ still retains the “proton chelated”
conformation it preferentially adopts in isolation (Table 1),
with the proton formally residing on the carbonyl oxygen atom
of the amide. This renders the amido N−H group in L6
cationic, allowing it to form a very strong interaction with
PtCl62 (Figure 4, right).
Formation energies, ΔU ’s, for the most stable forms of the
f
[(LH) PtCl ] assemblies were determined, along with proto-
2
6
nation energies (ΔU ’s) and binding energies (ΔU ’s) to
p
b
2−
PtCl (Table 2). In accordance with the empirical extraction
6
data (Figure 3), which show the amidoether ligand L5 to be the
weakest extractant, the formation energy for the [(LH) PtCl ]
−
2
6
In the case of [(L4H) PtCl ], the nonchelated form of the
assembly for the amidoether model system L6 is calculated to
be the least favorable. The formation energy is dependent on
both the protonation and binding energies, and it is apparent
that the lower formation energy for [(L6H) PtCl ] can be
2
6
receptor is particularly strongly preferred, with the chelated
−1
form 46.9 kJ mol higher in energy (see Table 2, and the right
and left images of Figure 5, respectively). This arises because
the nonchelated form can use both the amido NH and the
2
6
attributed to L6 having the least favorable protonation energy;
the corresponding binding energy for this ligand is actually the
most favorable of the ligand set. For the amidoamines L1, L2,
and L4 the formation energies are broadly similar and thus offer
no clear prediction as to whether secondary or tertiary amides
will be the stronger extractants. While differences in binding
energies are apparent, they are offset by corresponding
differences in protonation energies, such that while the
secondary amide L4 has the most favorable binding energy
for the amidoamines (a consequence of the presence of the
strongly H-bonding amido N−H group), this is countered by
the least favorable protonation energy.
+
ammonium NH units to interact with the anion, leading to a
slightly stronger binding energy (see below and Supporting
Information Figures S6 and S7). The lack of similar behavior in
the case of [(L6H) PtCl ] is presumably due to the instability
of the form of L6H with the proton residing formally on the
ether oxygen atom (see Table 1).
2
6
+
All the energy-minimized structures of the [(LH) PtCl ]
2
6
assemblies have several polarized C−H bonds forming bonding
2−
interactions with the PtCl anion (Figures S2−S9, Supporting
6
Information). A similar feature is found in the solid-state
structure of [(L1H) PtCl ] (Figure 6), crystals of which were
2
6
isolated by diffusion of diethyl ether into a methanol solution of
the third phase formed during the solvent extraction experi-
ment. The solid-state structure shows that the proton which has
In a prediction of whether or not a particular reagent will be
a strong platinum extractant, the selectivity of the receptor for
PtCl62 over Cl is an important factor because the extraction
−
−
+
−
been added to generate the cationic receptor L1H is sited on
processes are competitive (see eqs 1 and 3) and Cl will be
the benzylamine nitrogen atom N and is strongly hydrogen
present in large excess in the aqueous solutions which contain
high concentrations of HCl (the data presented in Figure 3
relate to 6 M HCl). As such, in order to compare metalate
bonded to the amido oxygen atom O (N−H···O = 1.86(4) or
1.90(3) Å) to form the six-membered “proton chelate”. This is
E
DOI: 10.1021/acs.inorgchem.5b01317
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
+
Figure 5. Energy-minimized structures of [(L4H) PtCl ] with L4H units input in the “chelate-proton” conformer (left) and a “nonchelated”
2
6
−1
conformer (right). The contacts shown have NBO energies ≥5 kJ mol (Figure S6 and S7, Supporting Information).
Table 3. Energies of Formation (M06-2X/6-31+G(d)/
LANL2DZ) of the Ion Pairs [(LH)Cl] from the Most Stable
Forms of L1, L2, L4, and L6 (ΔU ); the Protonation
fCl
Energies (ΔU ) Associated with Forming the Most Stable
p
+
Form of the Cations LH ; Their Binding Energies (ΔU
)
bCl
−
to Cl ; and the Energies of the Gas-Phase Ion Exchange
a
Reaction As Expressed by Equation 9
overall
formation
energy,
energy of
anion exchange
binding energy,
reaction,
ΔU /kJ mol− ΔUbCl/kJ mol
1
−1
ΔUfCl /kJ mol
−1
ΔU /kJ mol
−1
p
ex
L1
L2
L4
L6
−325.7
−319.7
−295.9
−222.9
−409.0
−412.6
−428.5
−433.1
−734.6
−732.3
−724.4
−659.1
−107.8
−120.8
−130.7
−177.9
a
ΔUex = ΔU − 2ΔU .
f
fCl
Figure 6. Solid-state structure of [(L1H) PtCl ] with Cl, O, and N
2
6
atoms in green, red, and blue, showing contacts between ligand C−H
groups and PtCl62 (a, 2.938; b, 2.933; c, 3.210; d, 2.946; e, 2.814; f,
−
2
.947; g, 2.883; h, 2.861 Å) comparable to those in the energy-
+
−
Figure 7. Energy-minimized structures of the ion pairs, LH Cl ,
formed by L1 (left) and L6 (right).
minimized structure (Figure 4) and a space-filling representation
showing the approach of two further L1H units (purple) from
+
neighboring [(L1H) PtCl ] assemblies.
2
6
favored for all ligand systems. This information helps to
account for the observation that in solvent extraction processes
significant uptake of PtCl62 is observed from aqueous
solutions even in the presence of a large excess of chloride
−
versus chloride, the energies of formation of the ion pairs
+
−
LH Cl (ΔU , see eq 8) were calculated (Table 3). In all
fCl
6
cases, the most stable form of the ion pair has a close contact
between the N−H bond of the protonated amine and the
chloride ion (Figure 7), which is consistent with the latter being
a “hard” anion and the N−H unit being a good H-bond donor.
ions.
A feature of the ligands discussed in this work is that they
present an array of polarized N−H and C−H bonds which
favor formation of assemblies with the larger, charge diffuse,
PtCl62 ion, rather than the smaller “hard” chloride ion. It is
clear from estimates of the energies of the interaction of the
−
The energies of the gas-phase ion exchange reactions, ΔU ’s,
ex
derived by eq 9, were also calculated (Table 3). The negative
values determined in each case indicate that formation of the
chloridoplatinate assembly, rather than the chloride ion pair, is
+
2−
LH cations with PtCl6 (Figures S2 and S3, Supporting
2−
Information) that N−H···PtCl6 bonding is much stronger
F
DOI: 10.1021/acs.inorgchem.5b01317
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
2
6
−
than C−H···PtCl bonding. However, when the preorganized
+
form of a LH receptor provides several polarized C−H units
they make a significant contribution to the binding of the anion.
This is substantiated by significantly higher NBO energies for
the 14 C−H···Cl contacts a−g and a′−g′ in [(L1H) PtCl ]
2
6
(
Figure 4 and Figure S2, Supporting Information) compared to
+
the conventional hydrogen bonding interactions from the NH
groups in the isomeric forms of [(L1H) PtCl ] (Figure S3,
Supporting Information) which have L1H in a nonchelated
tautomeric form.
2
6
+
+
−
LH + Cl ⇌ [(LH)Cl]
(8)
(9)
2
−
−
2
[(LH)Cl] + PtCl6 ⇌ [(LH) PtCl ] + 2Cl
2
6
While the DFT calculations above provide insight into the
origins of the strength and selectivity of the amidoamine and
amidoether extractants, it is clear that they do not completely
reproduce the observed order of extractant strength, as the
experimental data demonstrate that L1 is a significantly better
extractant than L3, which in turn is significantly better than L5;
Figure 8. Space filling representation of the energy-minimized
structure of [(L1H)
{(L1H)Cl}
PtCl ] assembly showing the two
2
2 6
+
L1H cations in red, the two (L1H)Cl ion pairs in blue, and the
2
−
PtCl6 in green.
this trend is only partly borne out by the ΔU values (Table 2),
f
amidoether L5 can be ascribed to their greater basicity, making
formation of the cationic receptors LH more favorable. As in
which show that the formation energy of L1 almost has parity
with that of L4 (the model system for L3), which in turn has a
higher formation energy than L6 (the model system for L5). A
possible reason for this disparity is that the [(LH) PtCl ]
+
other areas of anion receptor design, the benefits of
incorporating amide units as strong hydrogen bond donors
can be offset by their propensity to bond to each other, and it is
of considerable practical significance that the tertiary amide L1,
which has no amido N−H unit, is a strong extractant. In
2
6
assembly that was used as a basis for the DFT modeling work is
not an accurate representation of those that form in the water-
immiscible phases. From the X-ray crystal structure of
2
4
−
contrast to the extraction of tetrahedral ZnCl dianions where
[
(L1H) PtCl ] (Figure 6) and the analogous energy-minimized
2
6
5
relatively simple aminoamides confer good solubility of
structures (Figures 4 and 5) it is clear that there is sufficient
2−
[(LH) ZnCl ] assemblies in hydrocarbon solvents, the
space around the PtCl6 to accommodate further ligands, and
indeed, 100% Pt-loading in extraction requires extractant to
platinum ratios greater than 2:1 (Figure 3). Investigation of the
formation of such large structures using quantum mechanical
computational methods is impractical because of the large
number of atoms involved. Moreover, the many degrees of
freedom that such a large system comprises (many of which will
relate to low energy torsional motions) render the concept of a
global minimum energy structure less relevant. Instead a
collective ensemble of low energy states, such as can be
obtained from a molecular dynamics simulation, becomes a
more appropriate representation. Preliminary results of in vacuo
simulations indicate that the potential energy of the system
2
4
2
−
extraction of PtCl6 is characterized by formation of a third
phase which could possibly contain additional extractant
molecules in the form of their hydrochloride salts. Computa-
tional techniques to model the formation of the relatively large
assemblies such as [(LH) (PtCl )Cl ] are in progress.
4
6
2
ASSOCIATED CONTENT
Supporting Information
■
*
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.5b01317.
Tables of energies and figures showing energy-minimized
structures and NBO energies (PDF)
2−
+
involving association of a PtCl6 ion with two L1H cations
and two (L1H)Cl ion pairs is 209.3 ± 121.8 kJ mol− lower
than the combined potential energies of one [(L1H) PtCl ]
1
2
6
AUTHOR INFORMATION
Corresponding Authors
E-mail: c.morrison@ed.ac.uk. Phone: 0131 650 4725.
E-mail: P.A.Tasker@ed.ac.uk.
unit and two (L1H)Cl units (Figure 8). From this it is clear
■
*
*
2−
that a PtCl6 ion can readily accommodate the interactions
with four L1-type ligands. An atomistic model of this type may
also provide an explanation for the formation of a third phase in
the extraction process because aggregation of multiple units is
likely to lead to gel-like structures.
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
CONCLUSIONS
The protonated forms of L1−L6 have been shown to be good
receptors for the PtCl anion. Computational work and X-ray
■
2−
Notes
6
The authors declare no competing financial interest.
structure determination reveal that an important feature of the
anion-binding sites is the presence of an array of polarized C−
H bonds that effectively address the diffuse negative charge on
the chloridoplatinate ion, helping the extractants L1, L3, and L5
to show selectivity for the metalate over chloride which is
present in large excess. The higher strengths of the amidoamine
extractant receptors L1 and L3 as compared to that of the
ACKNOWLEDGMENTS
■
We thank the EPSRC, Johnson Matthey, and The University of
Edinburgh for funding for Ph.D. studentships for K.J.M., I.C.,
E.D.D., and A.M.W., and EaStCHEM for access to the Research
Computing Facility.
G
DOI: 10.1021/acs.inorgchem.5b01317
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
REFERENCES
■
(
1) Busschaert, N.; Caltagirone, C.; Van Rossom, W.; Gale, P. A.
Chem. Rev. 2015, 115, 8038.
(
2) Dalla Cort, A. Ion-Pair Receptors. Supermolecular Chemistry: From
Molecules to Nanomaterials; Wiley: New York, 2012 10.1002/
780470661345.smc068.
9
(
3) Gale, P. A.; Caltagirone, C. Chem. Soc. Rev. 2015, 44, 4212−4227.
4) Kim, D. S.; Sessler, J. L. Chem. Soc. Rev. 2015, 44, 532−546.
5) Jentzsch, A. V. Pure Appl. Chem. 2015, 87, 15−41.
(
(
(
6) Turkington, J. R.; Bailey, P. J.; Love, J. B.; Wilson, A. M.; Tasker,
P. A. Chem. Commun. 2013, 49, 1891−1900.
7) Grant, R. A.; Gordon, R. J.; Woolham, S. F. International Solvent
Extraction Conference; Wurzburg, Sept 7−11, 2014.
8) Reedijk, J. Platinum Met. Rev. 2008, 52, 2−11.
9) Bernardis, F. L.; Grant, R. A.; Sherrington, D. C. React. Funct.
(
̈
(
(
Polym. 2005, 65, 205−217.
10) Wilson, A. M.; Bailey, P. J.; Tasker, P. A.; Turkington, J. R.;
Grant, R. A.; Love, J. B. Chem. Soc. Rev. 2014, 43, 123−135.
11) Ivanova, S. N.; Gindin, L. M.; Mironova, L. Y. Izvestiya
Sibirskogo Otdeleniya Akademii Naul SSSR, Seriya Khimicheskikh Nauk
964, 2, 35−43.
(
(
1
(
12) Pohlandt, C.; Fritz, J. S. Talanta 1979, 26 (5), 395−399.
13) Berg, E. W.; Senn, W. L., Jr. Anal. Chim. Acta 1958, 19, 12−17.
14) Knothe, M. Z. Anorg. Allg. Chem. 1980, 470, 216−226.
15) Mhaske, A. A.; Dhadke, P. M. J. Chem. Eng. Jpn. 2001, 34 (8),
(
(
(
1
(
052−1055.
16) Warr, R. J.; Westra, A. N.; Bell, K. J.; Chartres, J.; Ellis, R.; Tong,
C.; Simmance, T. G.; Gadzhieva, G.; Blake, A. J.; Tasker, P. A.;
Schroder, M. Chem. - Eur. J. 2009, 15, 4836−4850.
17) Bell, K. J.; Westra, A. N.; Warr, R. J.; Chartres, J.; Ellis, R.; Tong,
C. C.; Blake, A. J.; Tasker, P. A.; Schroder, M. Angew. Chem., Int. Ed.
008, 47, 1745−1748.
18) Ellis, R. J.; Chartres, J.; Henderson, D. K.; Cabot, R.;
Richardson, P. R.; White, F. J.; Schroder, M.; Turkington, J. R.;
Tasker, P. A.; Sole, K. C. Chem. - Eur. J. 2012, 18, 7715−7728.
19) Ellis, R. J.; Chartres, J.; Tasker, P. A.; Sole, K. C. Solvent Extr. Ion
Exch. 2011, 29, 657−672.
20) Turkington, J. R.; Cocalia, V.; Kendall, K.; Morrison, C. A.;
̈
(
̈
2
(
̈
(
(
Richardson, P.; Sassi, T.; Tasker, P. A.; Bailey, P. J.; Sole, K. C. Inorg.
Chem. 2012, 51, 12805−12819.
(
21) Narita, H.; Morisaku, K.; Tanaka, M.; Nagao, K.; Fuchikami, T.;
Yoshida, T.; Kuroda, K. International Solvent Extraction Conference;
Sept 7−11, 2014.
(
22) Narita, H.; Morisaku, K.; Mikiya, T. Solvent Extr. Ion Exch. 2015,
3
(
3, 407−417.
23) Al-Awadi, S. A.; Abdallah, M. R.; Dib, H. H.; Ibrahim, M. R.; Al-
Awadi, N. A.; El-Dusouqui, O. M. E. Tetrahedron 2005, 61, 5769−
777.
24) Gaussian 09, Revision D.01; Gaussian, Inc.: Wallingford CT,
009. Full citation available in the Supporting Information.
5
(
2
(
25) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−52.
(26) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215−41.
(27) Boys, S. F.; Bernardi, F. Mol. Phys. 1970, 19, 553−566.
(28) Simon, S.; Duran, M.; Dannenberg, J. J. J. Chem. Phys. 1996,
1
(
05, 11024−11031.
29) NBO 6.0; Glendening, E. D.; Badenhoop, J. K.; Reed, A. E;
Carpenter, J. E.; Bohmann, J. A.; Morales, C. M,; Landis, C. R.;
Weinhold, F. Theoretical Chemistry Institute, University of
Wisconsin: Madison, WI, 2013.
(
30) Jorgenson, W. L.; Maxwell, D. S.; Tirado-Raves, J. J. Am. Chem.
Soc. 1996, 118, 11225−11236.
(
(
31) Plimpton, S. J. Comput. Phys. 1995, 117, 1−19.
32) Martínez, L.; Andrade, R.; Birgin, E. G.; Martínez, J. M. J.
Comput. Chem. 2009, 30, 2157−2164.
H
DOI: 10.1021/acs.inorgchem.5b01317
Inorg. Chem. XXXX, XXX, XXX−XXX