Easily accessible, hydrocarbon-soluble, crystalline, anhydrous lanthanide (Nd, La, and Y) phosphates

Article abstract of DOI:10.1039/c2dt31505j

Nd, La and Y triphosphates were prepared via the reaction of potassium ionol ethyl phosphate with the corresponding lanthanide nitrates or chlorides in water. According to the X-ray diffraction data, the recrystallised reaction products were dimers. The p

Full text of DOI:10.1039/c2dt31505j

Dalton
Transactions
View Article Online
View Journal
PAPER
Easily accessible, hydrocarbon-soluble, crystalline,
anhydrous lanthanide (Nd, La, and Y) phosphates†
Cite this: DOI: 10.1039/c2dt31505j
Ilya E. Nifant’ev,*a,b Alexander N. Tavtorkin,a Andrei V. Shlyahtin,b
Sof’ya A. Korchagina,a,c Inna F. Gavrilenko,a Natalya N. Glebovaa and
Andrei V. Churakovd
Nd, La and Y triphosphates were prepared via the reaction of potassium ionol ethyl phosphate with the
corresponding lanthanide nitrates or chlorides in water. According to the X-ray diffraction data, the
recrystallised reaction products were dimers. The products did not contain water, were readily soluble in
hydrocarbon solvents and demonstrated promising catalytic properties in the polymerisation of buta-
diene and DL-dilactide.
Received 9th July 2012,
Accepted 5th October 2012
DOI: 10.1039/c2dt31505j
www.rsc.org/dalton
Introduction
Lanthanide derivatives have found wide use in organic syn-
thesis and homogeneous catalysis.1 The use of lanthanide
compounds as catalysts of butadiene 1,4-cis-polymerisation is
their most important application in homogeneous catalysis.2
Usually, these processes utilise neodymium carboxylate and
phosphate salts with lipophilic anions, such as neodymium
versatate, NdV, 13 (a mixture of isomers of different α,α-disub-
stituted decanoic acids that is also referred to as neodymium
neodecanoate), neodymium tris[bis(2-ethylhexyl)phosphate],
NdP, 24,5 and neodymium tris[bis(para-nonylphenyl)phos-
phate] 36 (see Scheme 1).
The use of lipophilic anions makes neodymium derivatives
soluble in hydrocarbons in which the process of polymeris-
ation is most efficient.2 However, due to the high electronic
and steric unsaturation of the neodymium ions in compounds
such as NdV and NdP, these neodymium derivatives form oli-
gomers of uncertain composition due to neodymium coordi-
nation to the oxygen atoms of the neighbouring molecules.7,8
Therefore, the “solutions” of NdV and NdP represent non-
homogeneous and viscous gels and the reactivity of neody-
mium in such derivatives is substantially reduced and critically
depends on the method of salt preparation. Obviously, the
Scheme 1 The structures of neodymium carboxylate and phosphate salts with
lipophilic anions.
presence of different Nd species in the gels of NdV and NdP
not only complicates the application of these compounds as
catalysts and reduces the potential process parameters (broad-
ens the polydispersity of the polymer and reduces the activity
of the catalyst)7 but also complicates the study of Nd-catalysed
polymerisation, as it is impossible to reliably identify the indi-
vidual intermediates.
There are several approaches that can be used to suppress
the oligomerisation of neodymium salts in solution. Kwag suc-
ceeded in preparing the adduct of neodymium tris-neodecano-
ate with an additional equivalent of the carboxylic acid and
showed that the activity of this monomeric complex is almost
one order of magnitude higher than the activity of standard
neodymium tris-neodecanoate.7 Yet another possible method
for suppressing the oligomerisation of neodymium carboxy-
lates and phosphates is the use of bulky and more structurally
rigid anions. Previously, this strategy was utilised by Anwander
in the preparation of neodymium derivatives based on ortho-
substituted benzoic acids.9 Anwander was able to not only
isolate individual neodymium carboxylates but also to unam-
biguously characterise several products obtained by alkylation
with trimethylaluminium. These works as well as the investi-
gation by Evans,10 who isolated intermediate chlorine-contain-
ing lanthanide compounds and studied the role of chlorine in
aA. V. Topchiev Institute of Petrochemical Synthesis RAS, 29 Leninsky prosp., Moscow
119991, , Russia
bChemistry Department, M. V. Lomonosov Moscow State University, Moscow 119991,
Russia. E-mail: inif@org.chem.msu.ru
cDepartment of Chemistry, Moscow State Pedagogical University, 1, M. Pirogovskaya,
Moscow 119991Russia
dN. S. Kurnakov Institute of General and Inorganic Chemistry RAS, 31 Leninsky
prosp., Moscow 119991, Russia
†CCDC reference numbers 856748, 856749 and 866340. For crystallographic data
in CIF or other electronic format see DOI: 10.1039/c2dt31505j
This journal is © The Royal Society of Chemistry 2012
Dalton Trans.

View Article Online
Paper
Dalton Transactions
neodymium catalysis, provided considerable progress in the geometrical parameters for 5 and 6 are shown in Fig. 1. The
understanding of neodymium catalysis of diene polymeris- central atoms have CN = 7 and possess an irregular coordi-
ation. Unfortunately, due to the difficulty in obtaining the nation polyhedron. The lengths of the non-bridging M–O
benzoic acids used and the moderate performance of the bonds vary within 2.3259(17)–2.4305(16) Å for 5 and 2.386(3)–
resulting neodymium complex catalysts, no further studies of 2.488(2) Å for 6, while the M–O(14) bridging bonds are signifi-
the Anwander-type catalysts derived from the substituted cantly longer (2.7425(17) Å for 5 and 2.793(2) Å for 6). In
benzoic acids were performed.
general, the La–O distances are approximately 2.5% longer
During the current work, we demonstrated that the oligo- than the corresponding Nd–O bonds. In both complexes, the
merisation of lanthanide salts in solutions can be completely metal atoms are linked by four phosphate ligands. Two of the
prevented by the use of bulky and structurally rigid ionol- metal atoms are tridentate chelating bridges (with central P(1)
containing phosphate anions and that this approach is atoms) and two are bidentate bridges (with central P(2)
beneficial for the catalytic properties of the compounds. The atoms). Additionally, the metal centres bear terminal chelating
corresponding phosphoric acid can be easily prepared from ligands with central P(3) atoms. In the ligands, all the phos-
cheap and readily available reagents such as ionol, ethanol phorous atoms are tetrahedral with O–P–O angles lying within
and POCl3.
101.8(1)–115.9(2)°. The P–O distances clearly split into three
non-overlapping ranges: P–OMet (1.487(3)–1.5106(16) Å); P–OEt
(1.557(3)–1.576(3) Å) and P–OAr (1.580(3)–1.603(3) Å). The sig-
nificant shortening of the P–OMet bonds (in comparison with
the single P–OEt and P–OAr bonds) is a result of their partial
Results and discussion
We undertook the synthesis of lanthanide (Nd, La, and Y) double-bond character.
derivatives of previously unknown but readily accessible ionol
The structure of the bimetallic complex 7 is presented in
ethyl phosphoric acid 4. This acid, as we demonstrated, can be Fig. 2. Both yttrium atoms possess a strongly distorted octa-
easily obtained in acceptable yields by the hydrolysis of the hedral coordination environment. The metal atoms are linked
corresponding acid chloride, which is in turn prepared by by four bidentate bridging ligands (with central P(1), P(2), P(3)
either the reaction of sodium ionolate with EtOP(O)Cl2 or by and P(4) atoms). Additionally, the metal centres bear terminal
successive reactions of POCl3 with lithium ionolate and chelating ligands containing central P(5) and P(6) atoms. The
ethanol (Scheme 2).
Y–O bonds formed by the terminal ligands (2.306(3)–2.317(3)
The reaction of the potassium salt of acid 4 with neody- Å) are noticeably longer than those formed by the bridging
mium nitrate in water at a reactant ratio of 3 : 1 gives a high phosphate esters (2.223(3)–2.267(3) Å). All four bridging
yield of crude neodymium tris-phosphate 5 formed as a pre- ligands exhibit significant asymmetry. Thus, the P(n)–O(n3)–Y
cipitate, which can be purified by recrystallisation from angles are approximately 35° less than the corresponding P(n)–
hexane. Lanthanum and yttrium tris-phosphates 6 and 7, O(n4)–Y angles (n = 1–4). As expected, the P–OAr and P–OEt dis-
respectively, can be prepared by a similar method to the neody- tances are ∼0.06 Å longer than the P–OMet bond lengths. The
mium derivative. According to the X-ray diffraction data, com- basal planes of the oppositely lying terminal ligands (specifi-
plexes 5–7 are dimers.
cally, P(5), O(53), O(54) and P(6), O(63), O(64)) are almost per-
The crystallographic data, data collection and refinement pendicular. The angle between these ligands is 89.1(2)°. In
parameters for 5, 6 and 7 are given in the Experimental section contrast, the same planes in the structures of 5 and 6 (namely
(Table 3). The bimetallic phosphate complexes 5 and 6 are iso- P(3A), O(33A), O(34A) and P(3), O(33), O(34)) are parallel due to
structural (Fig. 1). The metal core of 5 and the main the symmetry of these molecules.
Scheme 2 Synthesis of ionol ethyl phosphoric acid 4 and phosphates 5–7 of this acid.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2012

View Article Online
Dalton Transactions
Paper
Fig. 1 Molecular structures of 5 and 6. (a) Full structure, displacement ellipsoids are drawn at a 50% probability level. Hydrogen atoms are omitted for clarity. (b)
Metal core in the structures of 5 and 6. For clarity, the ethyl and 2,6-di-t-butyl-4-methylphenoxy groups are drawn as circles and denoted as Et and Io, respectively.
Selected bond distances (Å) and angles (°) in 5: Nd(1)⋯Nd(1A) 4.0172(6), Nd(1)–O(13) 2.3796(16), Nd(1)–O(14)bridge 2.7425(17), Nd(1)–O(14A) 2.3860(16), Nd(1)–
O(23) 2.3698(16), Nd(1)–O(24A) 2.3259(17), Nd(1)–O(33) 2.4215(16), Nd(1)–O(34) 2.4305(16), P(1)–O(11)Et 1.5601(18), P(1)–O(12)Ar 1.5862(17), P(1)–O(13)Met
1.4985(18), P(1)–O(14)Met 1.5106(16), P(2)–O(21)Et 1.5641(19), P(2)–O(22)Ar 1.5961(18), P(2)–O(23)Met 1.4966(18), P(2)–O(24)Met 1.4910(18), P(3)–O(31)Et 1.5679(18),
P(3)–O(32)Ar 1.5765(16), P(3)–O(33)Met 1.5028(17), P(3)–O(34)Met 1.5050(17), P(1)–O(13)–Nd(1) 104.96(9), P(1)–O(14)–Nd(1) 89.62(8), P(1)–O(14)–Nd(1A)
167.44(10), Nd(1)–O(14)–Nd(1A) 102.91(5), P(2)–O(23)–Nd(1) 138.34(10), P(2)–O(24)–Nd(1A) 142.59(10), P(3)–O(33)–Nd(1) 96.66(8), P(3)–O(34)–Nd(1) 96.23(8).
Selected bond distances (Å) and angles (°) in 6: La(1)⋯La(1A) 4.1139(8), La(1)–O(13) 2.442(3), La(1)–O(14)bridge 2.793(2), La(1)–O(14A) 2.448(3), La(1)–O(23)
2.424(2), La(1)–O(24A) 2.386(3), La(1)–O(33) 2.479(3), La(1)–O(34) 2.488(2), P(1)–O(11)Et 1.557(3), P(1)–O(12)Ar 1.584(3), P(1)–O(13)Met 1.496(3), P(1)–O(14)Met
1.508(3), P(2)–O(21)Et 1.570(3), P(2)–O(22)Ar 1.603(3), P(2)–O(23)Met 1.501(3), P(2)–O(24)Met 1.487(3), P(3)–O(31)Et 1.576(3), P(3)–O(32)Ar 1.580(3), P(3)–O(33)Met
1.496(3), P(3)–O(34)Met 1.504(3), P(1)–O(13)–La(1) 105.04(13), P(1)–O(14)–La(1) 90.14(12), P(1)–O(14)–La(1A) 166.63(16), La(1)–O(14)–La(1A) 103.23(8), P(2)–
O(23)–La(1) 138.79(15), P(2)–O(24)–La(1A) 143.11(16), P(3)–O(33)–La(1) 97.22(12), P(3)–O(34)–La(1) 96.62(12).
Owing to their good solubility in non-polar solvents, for the structures of 5–7 presented herein are dimeric due to the
example in chloroform-d1 and benzene-d6, tris-phosphates can steric requirements of the bulky 2,6-di-t-butyl-4-methylphenoxy
be studied by NMR. In the 31P NMR (CDCl3) spectrum of groups. No short intermolecular contacts were found in the
yttrium derivative 7, the signals of the terminal (−1.55 ppm) structures of 5–7. Thus, by using bulky and structurally rigid
and bridging phosphates (−14.42 ppm) are clearly distin- phosphate ligands, we succeeded in completely preventing the
guished. In lanthanum derivative 6, bridging phosphates oligomerisation of phosphates 5–7. As a result, the compounds
differ in their X-ray diffraction data but still produce only one we obtained were not only readily soluble in hydrocarbon sol-
signal in the 31P NMR spectrum (−14.50 ppm) due to the vents but also formed non-viscous solutions.
similar nature and fast chemical exchange on the NMR time
To elucidate the effect of oligomerisation on the catalytic
scale. Terminal phosphates are responsible for a separate properties of neodymium phosphate 5, we studied its catalytic
signal (−3.95 ppm). Unfortunately, the paramagnetism of the properties in the polymerisation of butadiene and compared
Nd(III) ion precludes recording informative NMR spectra of the results with the properties of the commercial neodymium
complex 5. Nevertheless, the identity of structures 5 and 6 in 2-ethylhexyl phosphate 2 and neodymium tris[bis(para-nonyl-
the crystalline state and the similarity of the chemical proper- phenyl) phosphate] 3, which represents the derivative of the
ties of lanthanum and neodymium suggest that in solution, lipophilic but not a very bulky bis(para-nonylphenyl) phospho-
the structures of these compounds are also identical. Thus, ric acid. The catalyst is usually prepared by the reaction of the
the dimeric phosphate complexes in unpolar and non-coordi- neodymium derivative with an organoaluminium reagent
nating solvents have the same dimeric structures as they have (AlR3, AlHR2) and chlorinating reagent (Et2AlCl, Et3Al2Cl3).2
in the crystalline state.
According to the mechanism of the catalysis proposed by
Only a few structures of homoleptic lanthanide complexes Kwag,7 the catalytic site represents a compound of the type
with phosphate diesters are currently known, specifically, com- XNdRCl (X = phosphate or carboxylate, R = alkyl). We prepared
plexes containing La, Ce, Pr, Nd, Sm and Eu.11 Only the sim- the catalyst via alkylation of phosphate 5 with diisobutylalumi-
plest symmetric dimethyl and diethyl phosphate ligands nium hydride in the presence of butadiene followed by the
(RO)2PO2 were used in these investigations and all of the addition of ethylaluminium sesquichloride (Et3Al2Cl3). Table 1
reported structures were polymeric in the crystal. In contrast, summarises the experimental data obtained. Surprisingly, it
This journal is © The Royal Society of Chemistry 2012
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Fig. 2 Molecular structure of 7. (a) Full structure, displacement ellipsoids are drawn at a 30% probability level. Hydrogen atoms are omitted for clarity. (b) Metal
core in the structure of 7. For clarity, the ethyl and 2,6-di-t-butyl-4-methylphenoxy groups are drawn as circles and denoted as Et and Io, respectively. Selected bond
distances (Å) and angles (°) in 7: Y(1)–O(13) 2.267(3), Y(2)–O(14) 2.237(3), Y(1)–O(24) 2.236(3), Y(2)–O(23) 2.252(3), Y(1)–O(33) 2.236(3), Y(2)–O(34) 2.223(3), Y(1)–
O(44) 2.226(3), Y(2)–O(43) 2.237(3), Y(1)–O(53) 2.317(3), Y(2)–O(63) 2.306(3), Y(1)–O(54) 2.315(3), Y(2)–O(64) 2.308(3), P(1)–O(11)Et 1.566(4), P(1)–O(13)Met
1.489(4), P(1)–O(12)Ar 1.595(4), P(1)–O(14)Met 1.489(4), P(2)–O(21)Et 1.551(4), P(2)–O(23)Met 1.499(3), P(2)–O(22)Ar 1.582(4), P(2)–O(24)Met 1.488(4), P(3)–O(31)Et
1.557(4), P(3)–O(33)Met 1.498(4), P(3)–O(32)Ar 1.581(4), P(3)–O(34)Met 1.486(4), P(4)–O(41)Et 1.570(4), P(4)–O(43)Met 1.491(4), P(4)–O(42)Ar 1.578(4), P(4)–O(44)Met
1.492(4), P(5)–O(51)Et 1.551(4), P(5)–O(53)Met 1.493(4), P(5)–O(52)Ar 1.565(4), P(5)–O(54)Met 1.502(4), P(6)–O(61)Et 1.561(4), P(6)–O(63)Met 1.501(4), P(6)–O(62)Ar
1.557(4), P(6)–O(64)Met 1.506(4), Y(1)⋯Y(2) 4.0770(7), P(1)–O(13)–Y(1) 122.9(2), P(1)–O(14)–Y(2) 164.8(2), P(2)–O(23)–Y(2) 123.66(19), P(2)–O(24)–Y(1) 164.1(2),
P(3)–O(33)–Y(1) 127.3(2), P(3)–O(34)–Y(2) 159.8(2), P(4)–O(43)–Y(2) 127.4(2), P(4)–O(44)–Y(1) 161.0(2), P(5)–O(53)–Y(1) 95.96(17), P(5)–O(54)–Y(1) 95.79(17), P(6)–
O(63)–Y(2) 95.74(16), P(6)–O(64)–Y(2) 95.49(17).
Table 1 Butadiene polymerisation by complexes 5, 2 and 3 at 50 °C
Run
Complex
Nd : AlHiBu2 : Cl
Yield
1,4-cis, %
1,4-trans, %
1,2-, %
0.1
Mn × 10−3
Mw × 10−3
Mw/Mn
Ref.
1
2
3
4
5
1 : 15 : 2.5
—
—
>99
90
100
>99
97.7
96.0
95.7
95.1
2.2
2.7
a
237.9
479.6
2.01
This work
3
6
5
2
a
0.8
185
120.2
325
160.9
1.76
1.34
2*
1 : 15 : 2.5
4.1
This work
Conditions: hexane, Vtotal = 20 ml, [butadiene] = 1.5 M, [Nd] = 1.5 × 10−4 M, [i-Bu2AlH] = 4.5 × 10−3 M, [Cl] = 3.75 × 10−4 M, reaction
time = 2 h.
was found that the presence of bulky ortho-substituents in
The application of organic neodymium phosphates, usually
phosphate 5 does not reduce the catalytic activity of this com- neodymium 2-ethylhexyl phosphate 2, is not limited to the
pound but rather increases it compared with 3 (cf. runs 1 and stereoregular polymerisation of conjugated dienes. It has been
2). Moreover, the prevention of the oligomerisation in 5 results shown that catalytic systems based on organic neodymium
in
a
markedly different microstructure of the polymers phosphates can be used in the polymerisation of acetylene,12
obtained using 5 rather than oligomeric 2 and 3. For com- acrylonitrile,13 methyl methacrylate,14 and styrene,15 in the
plexes 2 and 3, only a 95–96% content of the 1,4-cis-com- copolymerisation of carbon dioxide and epichlorohydrin16 and
ponent can be achieved but the use of 5 under the same in the copolymerisation of isoprene and epichlorohydrin.17
conditions allows one to increase this value to nearly 98%. Because of our interest in the polymerisation of dilactide, we
Note that a similar increase in the content of the 1,4-cis-com- investigated the effect of the non-oligomeric nature of phos-
ponent was observed by Kwag during his study of monomeric phates 5–7 on their ability to initiate dilactide polymerisation.
neodymium tris-neodecanoate.7 Thus, we demonstrate that The ability to use neodymium 2-ethylhexyl phosphate 2 in this
the suppression of oligomerisation in a solution of neody- process has previously been demonstrated18 and the appli-
mium phosphate 5 has a pronounced influence on its catalytic cation of some carboxylates also reported.19 In both cases, the
performance in 1,4-cis-butadiene polymerisation by both results were moderate and a reasonable conversion was
enhancing the catalyst activity and increasing the degree of achieved only at temperatures above 100 °C. In our opinion,
process stereoregularity.
this rather low activity of the studied phosphates and
Dalton Trans.
This journal is © The Royal Society of Chemistry 2012

View Article Online
Dalton Transactions
Paper
Table 2 Results of the polymerisation of D,L-dilactide catalysed by complexes 5, 6, 7, and 2
Run
Complex
T, °C
Conversion (NBR), %a
Mn, NMR × 10−3 b
Mn, calc × 10−3 c
Mn × 10−3 (GPC)
Mw/Mn (GPC)
1
2
3
4
5
6
6
5
2
7
6
5
80
80
80
80
60
60
>99
>99
85
6.9
8.7
5.4
6.1
4.6
4.1
7.2
7.2
6.1
5.4
7.2
5.4
8.2
8.3
6.6
6.5
7.0
4.7
1.3
1.2
1.1
1.1
1.1
1.2
75
>99
75
Conditions: nD,L : ncomplex : nEt Al : niPrOH = 100 : 1 : 2 : 1, where nD,L, ncomplex, nEt Al, niPrOH are the initial amounts of the dilactide, the
3
3
complex, Et3Al and iPrOH, respectively.a The conversion was determined by integrating the signals of the methine protons of the
b
dilactide (5.04 ppm in CDCl3) and polylactide (5.12–5.26 ppm in CDCl3). Mn, NMR was determined by integrating the signals of the
c
methine protons of the polylactide and the methine protons of the polylactide terminal units (4.19 ppm in CDCl3). Mn, calc = 144 ×
dilactide amount/Et3Al amount × dilactide conversion.
carboxylates is due exclusively to oligomerisation of the neody- polymerisation of butadiene and DL-dilactide. It is likely that
mium salts. Indeed, other more sophisticated neodymium the lanthanide triphosphates based on ionol ethyl phosphoric
derivatives described recently can polymerise dilactide under acid and their related compounds could be successfully uti-
mild conditions at high rates.20
lised for many applications that require soluble lanthanide
We studied the polymerisation of dilactide catalysed by the derivatives of a definite composition.
“phosphate”/[Et3Al]/[iPrOH] = 1/2/1 system, where the “phos-
phate” was represented by derivatives 5–7, which were syn-
thesised in this work, or by the commercial neodymium
2-ethylhexyl phosphate 2. The catalyst was prepared by alkyl-
ation of phosphates 2 and 5–7 with two equivalents of Et3Al,
Experimental
Synthesis
and the addition of 1 equivalent of isopropanol initiated the In the synthesis of 2,6-di-tert-butyl-4-methylphenyl ethyl phos-
polymerization of the dilactide.19 Complexes 5 and 6 proved to phoric acid, commercial ionol (Merk) and EtOPOCl2, prepared
be quite active catalysts of polylactide synthesis, whereas the by a known procedure,21 were used. The lanthanide triphos-
activities of 7 and 2 were markedly lower, see Table 2. Indeed, phates were prepared from Nd(NO3)3 × 6H2O, LaCl3 × 7H2O
at 80 °C, 5 and 6 provide an almost quantitative conversion of and YCl3 × 6H2O (Reakhim). The hexane used in the catalyst
the dilactide, while in the presence of 7 and 2, the reaction synthesis and polymerisation was dried by reflux over sodium
goes only to approximately two-thirds completion. Interest- wire. D,L-Dilactide (Aldrich) was recrystallised from toluene and
ingly, a decrease in the reaction temperature to 60 °C results in dried in vacuo. Isopropanol (Lab-Scan, 99.8%) was dried and
considerable differentiation of the activities of the lanthanum distilled from CaH2. Toluene (Sigma-Aldrich, 99.8%) was dried
and neodymium phosphates. While the conversion on the and distilled from sodium. NMR spectra were recorded on a
lanthanum complex 6 is almost quantitative, in the presence Bruker Avance 400 instrument in CDCl3 at 30 °C. The micro-
of the neodymium complex 5, it is only 65%. Thus, we demon- structures of polybutadiene were measured by IR spectroscopy
strated that for lactide polymerisation, the non-associated indi- (IFS-66v/s Bruker) according to published data.22 Gel per-
vidual catalyst 5 has a higher catalytic activity than the meation chromatography was performed on a high-pressure
oligomeric neodymium 2-ethylhexyl phosphate 2. An interest- chromatograph equipped with a Styrogel HR 5E, HR 4E system
ing finding is that the lanthanum phosphate 6 is a markedly of columns using a Rheodyne injector with a 200 μl loop and a
more active catalyst than the neodymium analogue 5, which Waters 2410 refractometer. THF was used as the eluent.
substantially surpasses the yttrium complex 7.
Synthesis of 2,6-di-tert-butyl-4-methylphenyl-ethyl phosphoric
acid (4)
Method A: A 1-litre flask with a reflux condenser was charged
Conclusion
with ionol (110 g, 0.5 mol), sodium (15 g, 0.65 mol) and
Thus, in this work we synthesised Nd, La and Y triphosphates toluene (200 ml). The mixture was stirred at reflux until the
based on ionol ethyl phosphoric acid. These compounds are hydrogen evolution ceased (∼5 h). Then, EtOPOCl2 (89.62 g,
remarkable for two reasons: first, they can be easily prepared 0.55 mol) was added with stirring and ice water cooling. After
in water but do not contain water and second, they are individ- 1 h, the mixture was filtered to remove excess sodium, the fil-
ual non-oligomerised compounds of a definite composition. trate was concentrated in vacuo, dissolved in pyridine (200 ml)
These features not only provide good solubility of the triphos- and water (20 ml) was then added. The mixture was stirred for
phates in hydrocarbon solvents but the compounds also quali- 3 h and concentrated in vacuo, HCl was added to a highly
tatively surpass the conventional oligomeric organic acidic pH (∼50 ml) and the mixture was extracted with hexane
phosphates in their catalytic properties, such as in the (500 ml). The organic phase was separated, dried with sodium
This journal is © The Royal Society of Chemistry 2012
Dalton Trans.

View Article Online
Paper
Dalton Transactions
sulphate and concentrated in vacuo to ∼200 ml. The precipitate 7H2O (3.71 g, 10 mmol) in water (100 ml) was added to the
was filtered off and washed with cold hexane (2 × 50 ml). Yield obtained solution with stirring. After 4 h of stirring, the
77 g (47%).
mixture was filtered and washed with water (3 × 100 ml). The
Method B: A two-necked flask was charged with ionol (11 g, precipitate was dried twice in a vacuum desiccator over CaCl2
50 mmol) and anhydrous diethyl ether (30 ml), and butyl- and then in vacuo to a constant weight. The resulting sub-
lithium (2 ml, 50 mmol) was added dropwise. After 30 min, stance was dissolved in benzene (20 ml). The solution was con-
POCl3 (7 ml, 76 mmol) was added and the mixture was left centrated to half its volume and added to pentane (10 ml). The
overnight. Then, the mixture was filtered and concentrated precipitated crystals were filtered off, washed with pentane
and the product was used without further purification. The and dried in vacuo. Yield 5.1 g (51%). 1H NMR (400 MHz,
yield of 2,6-di-tert-butyl-4-methylphenyl phosphorodichloridate CDCl3): δ = 0.48 (12H, br.s, OCH2CH3), 1.16 (6H, br.s,
1
was quantitative (16.9 g). H NMR (400 MHz, CDCl3): δ = 1.54 OCH2CH3), 1.46 (36H, s, tert-Bu), 1.52 (18H, s, tert-Bu), 2.23
(18H, s, tert-Bu), 2.36 (3H, d, J = 2.38, Me), 7.21 (2H, s, Ar). 31P (12H, s, Me (Ar)), 2.32 (6H, s, Me (Ar)), 3.68 (8H, br.s,
NMR (161 MHz, CDCl3): δ = 1.61.
OCH2CH3), 3.99 (4H, br.s, OCH2CH3), 7.02 (8H, s, m-H (Ar)),
Pyridine (0.95 ml, 12 mmol) was added to a solution of 2,6- 7.11 (4H, s, m-H (Ar)). 31P NMR (161 MHz, CDCl3): δ = −3.95
di-tert-butyl-4-methylphenyl phosphorodichloridate (3.37 g, (2P, L2), −14.50 (4P, L4). Anal. C51H84LaO12P3 calcd C 54.64, H
10 mmol) and ethanol (11 mmol) in methylene dichloride 7.55; found C 54.49, H 7.89.
(6 ml). After 2 days, the mixture was concentrated, the residue
was dissolved in pyridine (10 ml) and water (2 ml) was added.
After 10 min, the pyridine was distilled off, the residue was dis-
solved in hexane (50 ml) and the solution was washed with
10% hydrochloric acid (50 ml). The organic phase was separ-
ated and the solvent was distilled off. The residue was recrys-
tallised from hexane. The yield of 2,6-di-tert-butyl-4-
methylphenyl ethyl phosphoric acid was 1.97 g (60%). 1H NMR
(400 MHz, CDCl3): δ = 1.23 (3H, t, J = 7.02, OCH2CH3), 1.42
(18H, s, tert-Bu), 2.27 (3H, s, Me (Ar)), 3.95 (2H, m, OCH2CH3),
7.05 (2H, s, Ar). 13C NMR (100 MHz, CDCl3): δ = 15.80 (d, J =
7.31, OCH2CH3), 31.86 (s, tert-Bu), 35.83 (s, Me (Ar)), 63.79 (d,
J = 5.12, OCH2CH3), 127.56 (d, J = 1.85, 3-C), 132.87 (d, J = 1.46,
4-C), 142.25 (d, J = 3.66, 2-C), 146.00 (d, J = 8.78, 1-C). 31P NMR
(161 MHz, CDCl3): δ = −5.17. C17H29O4P calcd C 62.18, H 8.90;
found C 62.07, H 8.81.
Synthesis of yttrium tris[bis(2,6-di-tert-butyl-4-methylphenyl
ethyl) phosphate] (7)
A solution of KOH (0.35 g, 6.3 mmol) in water (5 ml) was
added dropwise to a suspension of finely ground acid (2.07 g,
6.3 mmol) in water (20 ml) until the acid dissolved and the
mixture was neutralised. YCl3 × 6H2O (0.61 g, 2 mmol) in
water (10 ml) was added to the obtained solution with stirring.
After 4 h of stirring, the mixture was filtered and washed with
water (3 × 20 ml). The precipitate was dried twice in a vacuum
desiccator over CaCl2 and then in vacuo to a constant weight.
The resulting substance was dissolved in hexane (20 ml). The
solution was concentrated by 50% and cooled to −20 °C. The
precipitated crystals were removed by filtration, washed with
cooled pentane and dried in vacuo. Yield 1.21 g (60%). 1H
NMR (400 MHz, CDCl3): δ = 0.45 (6H, br.s, OCH2CH3), 0.88
(12H, br.s, OCH2CH3), 1.46 (36H, s, tert-Bu), 1.52 (18H, s, tert-
Bu), 2.22 (12H, s, Me (Ar)), 2.30 (6H, s, Me (Ar)), 3.69 (8H, br.s,
OCH2CH3), 3.93 (4H, br.s, OCH2CH3), 7.03 (8H, s, m-H (Ar)),
7.09 (4H, s, m-H (Ar)). 31P NMR (161 MHz, CDCl3): δ = −1.55
(2P, L2), −14.42 (4P, t, JP–Y = 9.91, L4). C51H84O12P3Y calcd C
57.19, H 7.91; found C 57.05, H 8.30.
Synthesis of neodymium tris[(2,6-di-tert-butyl-4-methylphenyl
ethyl) phosphate] (5)
A solution of KOH (6.74 g, 120 mmol) in water (100 ml) was
added with stirring to a suspension of finely ground acid 4
(39.41 g, 120 mmol) in water (400 ml). After complete neutral-
isation, water (500 ml) was added and the resulting suspen-
sion was filtered and placed into a 2-litre flask equipped with a
mechanical stirrer. Neodymium nitrate (16.65 g, 38 mmol) in
X-ray structure determination of 5–7
water (200 ml) was added with stirring. After 4 h of stirring, Single crystals of complexes 5–7 were obtained by the slow
the mixture was filtered and washed with water (4 × 200 ml). evaporation of saturated solutions in hexane. Experimental
The precipitate collected on the filter was resuspended in intensities were measured on a Bruker SMART APEX II
benzene (500 ml) and filtered. The organic phase was dried diffractometer using graphite monochromatized Mo–Kα radi-
over sodium sulphate. Benzene was evaporated in vacuo and ation (λ = 0.71073 Å) using a ω-scan mode. Absorption correc-
the residue was recrystallised from 50 ml of hexane. The crys- tions based on measurements of equivalent reflections were
talline precipitate was filtered, washed with cold hexane applied.23,24 The structures were solved by direct methods and
(30 ml) and dried in vacuo. Yield 31.74
g
(74%). refined by full matrix least-squares on F2 with anisotropic
C51H84NdO12P3 calcd C 54.38, H 7.52; found C 54.77, H 7.52.
thermal parameters for all non-hydrogen atoms (except for the
t-Bu ligands in Y). As for Y, 36 methyl groups from the t-Bu
ligands were refined isotropically due to the high thermal
motion caused by free rotation along the CAr–CAlk bonds. In
Synthesis of lanthanum tris[bis(2,6-di-tert-butyl-4-
methylphenyl ethyl) phosphate] (6)
A solution of KOH (1.77 g, 31.5 mmol) in water (10 ml) was all the structures, hydrogen atoms were placed in calculated
added to a suspension of finely ground acid (10.34 g, positions and refined using a riding model. The structure of Y
31.5 mmol) in water (100 ml) until the acid dissolved. LaCl3 × contains two sites partially occupied by solvent n-hexane
Dalton Trans.
This journal is © The Royal Society of Chemistry 2012

View Article Online
Dalton Transactions
Paper
Table 3 Crystal data, data collection and refinement parameters for 5, 6 and 7
Complex
5
6
7
Formula
Fw
C102H168Nd2O24P6
2252.66
C102H168La2O24P6
2242.00
C108H182O24P6Y2
2228.18
Colour, habit
Cryst size, mm
Cryst syst
Space group
T (K)
Colourless, block
0.20 × 0.15 × 0.10
Triclinic
Colourless, plate
0.10 × 0.05 × 0.01
Triclinic
Colourless, block
0.25 × 0.20 × 0.20
Triclinic
ˉ
ˉ
ˉ
P1
P1
P1
173
173
296
a (Å)
14.273(3)
14.921(3)
15.089(3)
65.781(2)
75.178(3)
82.048(3)
2831.1(9)
1
14.344(3)
15.019(3)
15.110(3)
65.955(3)
74.968(3)
81.995(3)
2868.9(9)
1
15.049(2)
20.351(3)
22.399(3)
81.292(2)
87.985(2)
76.987(2)
6606.5(16)
2
b (Å)
c (Å)
α (°)
β (°)
γ (°)
V (Å3)
Z
ρcalc (g cm−3
)
1.321
1.298
1.120
μ (mm−1
F(000)
)
1.057
0.882
1.005
1182
1176
2380
θ range (°)
1.48 to 28.00
29 369
2.34 to 25.50
24 590
2.28 to 25.50
56 736
Total no. of reflns
Unique reflns
Rint
13 617
10 682
24 561
0.0259
11 903
0.0587
7945
0.0495
No. with I > 2σ(I)
No. of variables
R1 (I > 2σ(I))
14 593
628
628
1052
0.0310
0.0827
1.075
0.0444
0.0857
0.952
0.0709
wR2 (all data)
0.2282
GOOF on F2
1.034
Largest diff peak/hole, e Å−3
0.972/−0.456
0.968/−0.801
0.928/−0.603
molecules. The crystallographic data, data collection and covered with a septum, a solution was prepared by mixing a
refinement parameters for 5, 6 and 7 are given in Table 3.
1.9 M solution of Et3Al (1.4 ml) in toluene with toluene
(8.6 ml) and isopropanol (0.10 ml). After stirring for 3 min,
0.5 ml of this solution was withdrawn into the dilactide
ampoule. The ampoule was sealed and placed into a bath pre-
heated to the desired temperature and kept for 6 h. Then, the
ampoule was opened and the solvent was evaporated. The
small amount of the resulting substance was dissolved in
CDCl3. Conversion of the dilactive was determined by compari-
son of the integration of the signals of the polymer with the
corresponding signals of the residual dilactide.
Preparation of the polymerisation catalyst from 5
A glass reactor was charged under argon with neodymium
derivative 5 (0.113 g, 0.1 mmol) and hexane (7.5 ml). Buta-
diene (0.15 ml, 2 mmol) was recondensed (−78 °C). The
mixture was cooled to −78 °C and a 1 M solution of DIBAL-H
(1.5 ml) in hexane was added with stirring. After 1 h, a 0.83 M
solution of Me3Al2Cl3 (1 ml) in hexane was added to the reac-
tion mixture. The mixture was stirred for 20 h at 20 °C and was
then considered ready for use.
Butadiene polymerisation
Notes and references
The catalyst prepared from 5 (0.3 ml, 0.003 mmol) was placed
into an ampoule in an argon flow. Hexane (17 ml) was added
and butadiene (2.2 ml, 30 mmol) was recondensed (−78 °C).
The ampoule was sealed and heated to 50 °C. After 2 h, the
ampoule was opened and the reaction mixture was poured into
methanol (50 ml). The precipitated polymer was collected,
washed twice with methanol and dried in vacuo. Yield 1.61 g
(99.4%).
1 (a) S. Kobayashi, R. Anwander, Lanthanides: Chemistry and
Use in Organic Synthesis. Topics in Organometallic Chemistry,
vol. 2, Springer-Verlag, Berlin, Heidelberg, 2001;
(b) S. Kobayashi, M. Sugiura, H. Kitagawa and
W. W.-L. Lam, Chem. Rev., 2002, 102, 2227;
(c) A. L. Reznichenko, K. C. Hultzsch, Zh. Zhang, D. Cui,
B. Wang, F.T. Edelmann, T. Li, J. Jenter, P. Roesky, Molecu-
lar Catalysis of Rare-Earth Elements. Structure and Bonding,
ed. P. W. Roesky, Springer-Verlag, Berlin, Heidelberg, 2010,
vol. 137.
Dilactide polymerisation
A glass ampoule was charged with a stirrer bar, D,L-dilactide
(1.000 g, 6.9 × 10−3 mol) and the corresponding metal triphos-
phate (6.9 × 10−5 mol) was evacuated in vacuo (2 × 10−1
mmHg), filled with dry argon, and covered with a septum.
Toluene (4 ml) was then added. In another dry ampoule
2 L. Friebe, O. Nuyken and W. Obrecht, Adv., Polym., Sci.,
2006, 204, 1.
3 G. Sylvester, J. Witte and G. Marwede, Appl. EP 0007027,
1980.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans.

View Article Online
Paper
Dalton Transactions
4 (a) A. A. Berg, Y. B. Monakov, V. P. Budtov and 14 L. Jiang, Z. Shen and Y. Zhang, Eur. Polym. J., 2000, 36,
S. R. Rafikov, Vysokomol. Soedin. Ser. B, 1978, 20, 295; 2513.
(b) Z. Shen, J. Ouyang, F. Wang, Z. Hu, Y. Fu and B. Qian, 15 L. B. Wu, B. G. Li, K. Cao and B. F. Li, Eur. Polym. J., 2001,
J. Polym. Sci., Part A: Polym. Chem. Ed., 1980, 18, 3345.
5 P. Laubry, Appl. WO 02/38636 (to Michelin), 2002.
6 Y. Hatsutori, Y. Inoki and T. Ikematsu, Appl. JP 60023406A
(to Asahi Chemical Ind.), 1985.
37, 2105.
16 X. Chen, Y. Zhang and Z. Shen, Chin. J. Polym. Sci., 1994,
12, 28.
17 Y. Zhang, R. Zheng and Z. Shen, Chin. Sci. Bull., 1995, 40,
7 G. Kwag, Macromolecules, 2002, 35, 4875.
1182.
8 (a) C. Scharf, A. Ditze, K. Schwerdtfeger, S. Fuermeier, 18 J. Sun, L. Wu, L. Feng and Z. Shen, Zhejiang Daxue Xuebao
T. Bruhn, D. E. Kaufmann and J. C. Namyslo, Metall. Mater.
Trans. B, 2005, 36B, 429; (b) J. Thuilliez, P. Kiener and
E. Guerin, US 20120130055 (to Michelin).
Ziran Kexueban, 1988, 22, 203 (Chem. Abstr. 1989,
110:173825).
19 (a) J. Zhang, Z. Gan, Z. Zhong and X. Jing, Polym. Int., 1998,
45, 60; (b) Z. Zhong, D. Yu, F. Meng, Z. Gan and X. Jing,
Polym. J., 1999, 31, 633.
9 (a) A. Fischbach, F. Perdih, P. Sirsch, W. Scherer and
R. Anwander, Organometallics, 2002, 21,
4569;
(b) A. Fischbach, F. Perdih, E. Herdtweck and R. Anwander, 20 A. Buchard, R. H. Platel, A. Auffrant, X. F. Le Goff, P. Le
Organometallics, 2006, 25, 1626. Floch and C. K. Williams, Organometallics, 2010, 29, 2892.
10 (a) W. J. Evans, D. G. Giarikos and J. W. Ziller, Organometal- 21 A. M. Modro and T. A. Modro, Org. Prep. Proced. Int., 1992,
lics, 2001, 20, 5751; (b) W. J. Evans and D. G. Giarikos,
Macromolecules, 2004, 37, 5130.
11 F. H. Allen, Acta Crystallogr., Sect. B: Struct. Sci., 2002, B58,
380.
24, 57.
22 F. Ciampelli, D. Morero and M. Cambini, Makromol. Chem.,
1963, 61, 250.
23 Bruker, APEX2, SADABS and SAINT, Bruker AXS Inc.,
Madison, Wisconsin, USA, 2008.
12 L. Yu, M. Yang and Z. Shen, Cuihua Xuebao, 1985, 6, 260.
13 L. Jiang, Y. Zhang and Z. Shen, Eur. Polym. J., 1997, 33, 24 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystal-
577.
logr., 2008, A64, 112.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2012