Highly selective and immortal magnesium calixarene complexes for the ring-opening polymerization of rac-lactide

Article abstract of DOI:10.1002/cctc.201402041

The lithiation of 1,3-dipropoxy-p-tert-butylcalix[4]arene (LH₂) followed by reaction with nBuMgBr in THF resulted in the formation of the hetero-bimetallic complex [Li(THF)Mg(nBu)L] (1). By contrast, the treatment of tripropoxy-p-tert-butylcalix[4]arene (LH) with nBu₂Mg afforded a mononuclear complex [LMg(nBu)] (2). Single-crystal XRD studies revealed that in both structures the calix[4]arene adopts a cone conformation, and a Li cation resides in the cavity for 1. Both 1 and 2 were active for the ring-opening polymerisation of rac-lactide. Compound 2 displayed not only exceptional activity (100 equivalents, 3 min, 92% conversion, BnOH, room temperature) but also high selectivity (probability to form an r dyad, P_r=0.85) and exhibited an immortal character in THF. Surprisingly, compound 2 also showed isotactic bias (P_r=0.30-0.36) and an immortal character if toluene was employed as the solvent; 2D J-resolved ¹H NMR spectroscopy was employed for the assignment of the stereoselectivity.

Full text of DOI:10.1002/cctc.201402041

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201402041
Highly Selective and Immortal Magnesium Calixarene
Complexes for the Ring-Opening Polymerization of
rac-Lactide
Mark J. Walton,[a] Simon J. Lancaster,[a] and Carl Redshaw*[b]
The lithiation of 1,3-dipropoxy-p-tert-butylcalix[4]arene (LH2)
followed by reaction with nBuMgBr in THF resulted in the for-
mation of the hetero-bimetallic complex [Li(THF)Mg(nBu)L] (1).
By contrast, the treatment of tripropoxy-p-tert-butylcalix[4]ar-
ene (L’H) with nBu2Mg afforded a mononuclear complex
[L’Mg(nBu)] (2). Single-crystal XRD studies revealed that in
both structures the calix[4]arene adopts a cone conformation,
and a Li cation resides in the cavity for 1. Both 1 and 2 were
active for the ring-opening polymerisation of rac-lactide. Com-
pound 2 displayed not only exceptional activity (100 equiva-
lents, 3 min, 92% conversion, BnOH, room temperature) but
also high selectivity (probability to form an r dyad, Pr =0.85)
and exhibited an immortal character in THF. Surprisingly, com-
pound 2 also showed isotactic bias (Pr =0.30–0.36) and an im-
mortal character if toluene was employed as the solvent; 2D
1
J-resolved H NMR spectroscopy was employed for the assign-
ment of the stereoselectivity.
Introduction
Polymers with an inherent biode-
gradability, of which poly(lactic
acid) (PLA) and poly(caprolac-
tone) (PCL) are two of the most
common, have gained significant
interest because of their use in
biomedical devices.[1–3] The pro-
duction of PLA by using metal-
based catalysts for ring-opening
polymerisation (ROP) is consid-
ered to be the most convenient
Scheme 1. Previously reported Mg and Zn compounds for the ROP of lactide.
preparative route, primarily be-
cause of the ability to control the
molecular weight with low polydispersity.[3] Since Coates and
co-workers published their seminal work on zinc and magnesi-
um b-diiminate complexes, a large number of Mg catalysts
have appeared in the literature.[4,5] The drive towards the es-
tablishment of a highly active, selective and immortal catalyst
for the ROP of rac-lactide has seen a multitude of ligand sys-
tems employed, examples of which include iminopheno-
lates,[6,7] b-diiminates,[5,8] salen[9] and heteroscorpionates.[10,11]
However, the use of calix[4]arene-based ligand sets in this area
remains scant,[12] and indeed, in the case of Mg, there are few
reported complexes.[13] We note that recently there has been
a resurgence of Mg-based catalysts, for example, Chisholm
et al. utilised b-diiminate magnesium compounds for the ROP
of rac-lactide (Scheme 1, I).[8] The catalyst exhibited exception-
ally high activity as well as a heterotactic bias if THF was
added to the polymerisation system; however, the addition of
excess alcohol resulted in solvolysis and ligand loss and as
a result the system was unsuitable for ‘immortal’ polymeri-
sation.[8] Wang et al. explored the use of pyridyl-functionalised
alkoxy zinc and magnesium complexes that exhibited immortal
polymerisation of l-lactide and e-caprolactone (Scheme 1, II).[14]
The Mg catalyst employed was able to polymerise e-caprolac-
tone even in the presence of 500 equivalents of benzyl alcohol
as a chain-transfer agent to give the expected molecular
weight. The pyridyl alkoxy magnesium catalyst also demon-
strated an immortal character for the ROP of l-lactide using
triethanolamine as a chain-transfer/activation agent.[14]
[a] M. J. Walton, Dr. S. J. Lancaster
Energy Materials Laboratory
School of Chemistry
University of East Anglia
Norwich, NR4 7TJ (UK)
[b] Prof. C. Redshaw
Department of Chemistry
University of Hull
Hull, HU6 7RX (UK)
E-mail: C.Redshaw@hull.ac.uk
Chuang et al. have used tridentate pyrazolonate magnesium
catalysts (Scheme 1, III),[15] and although they gave lower activi-
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201402041.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1892 – 1898 1892

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
ties for the ROP of rac-lactide than the catalysts reported by
Chisholm et al. and Coates et al.,[5,8] they exhibited immortal
and stereoselective behaviour (probability to form an r dyad,
Pr =0.87). Generally, ligands that are monoanionic are chosen
for reaction with Mg precursors as they will inevitably lead to
a metal that still contains a viable nucleophilic group for ROP,
which may be the reason that calix[4]arenes are used rarely.
Vigalok et al. have had success with zinc alkyl based calix[4]ar-
enes and although the dialkoxycalix[4]arene ligand is dianionic
if deprotonated, its use leads to a dimetallic complex that still
contains a nucleophilic group (Scheme 1, IV).[16] Herein, an ex-
ploration of calix[4]arene-based Mg catalysts is reported. We
1
have used 2D J-resolved H NMR spectroscopy as an alterna-
tive to homonuclear decoupled NMR spectroscopy to assign
the stereoselectivity.
Results
Synthesis and structural studies
Figure 1. ORTEP representation of 1. Hydrogen atoms, tert-butyl groups and
minor disordered components have been removed for clarity. Displacement
ellipsoids are drawn at the 50% probability level. Selected bond lengths [ꢁ]
and angles [8]: Li(1)ÀO(5) 1.867(5), Li(1)ÀO(2) 1.903(5), Li(1)ÀO(4) 1.904(5),
Mg(1)ÀO(1) 2.373(2), Mg(1)ÀO(2) 1.9261(19), Mg(1)ÀO(3) 2.302(2), Mg(1)À
O(4) 1.9356(19), Mg(1)ÀC(61) 2.146(3), C(61)ÀMg(1)ÀLi(1) 176.77(16), O(5)À
Li(1)ÀMg(1) 179.2(3), O(2)ÀMg(1)ÀC(61) 137.75(13), O(4)ÀMg(1)ÀC(61),
131.36(13).
The synthesis of [Li(THF)Mg(nBu)L] (1) and [L’Mg(nBu)] (2) is
shown in Scheme 2.
Compound 1: The reaction of lithiated 1,3-dipropoxy-p-tert-
butylcalix[4]arene (LLi2) with two equivalents of n-butyl magne-
an n-butyl group, is bound to the four oxygen atoms of the
lower rim; the MgÀO bonds to the alkoxy groups [O(1)/O(3)] at
~2.34 ꢁ are, as expected, somewhat longer than those to the
phenolic groups [O(2)/O(4)] at ~1.93 ꢁ. The Mg and Li metal
centres are 2.670(5) ꢁ apart, and the Mg centre adopts a dis-
torted trigonal bipyramidal geometry. The THF O atom, Li, Mg
and C of the n-butyl group are all essentially linear as repre-
sented by the angles between O(5)ÀLi(1)ÀMg(1) and C(61)À
Mg(1)ÀLi(1) of 176.77(16) and 179.2(3)8, respectively. There is
also a slight difference in angle around the Mg centre between
the non-propoxy O atoms O(2)/O(4) and C(61) in the equatorial
plane; the angle between O(2) and C(61) is slightly larger than
that of O(4) and C(61), 137.75(13) versus 131.36(13)8.
Similar structures based on alkali and alkali earth metals
have been reported previously by Floriani and co-workers;[13]
in particular, the treatment of 1,3-dicyclopentoxy-p-tert-butyl-
calix[4]arene with magnesium anthracene led to the formation
of [{p-tBu-calix[4](OCyp)2-(O)2}Mg(thf)] (V). The structure of V is
similar to that of 1; compound 1 contains both a Li atom and
THF molecule within the cavity, whereas V contains only a THF
molecule. The presence of the Li centre within the cavity of
1 forces the calixarene further into an elliptical conformation
as shown by the bond angles between C(24)ÀO(2)ÀMg(1)/
C(8)ÀO(4)ÀMg(1). Each of the MgÀO bonds are extended in
1 compared to that of V (Table 1). Compound 1 has been char-
Scheme 2. Synthesis of 1 and 2. i) 1) 2 nBuLi, THF, 08C, 1 h, 2) nBuMgBr, THF,
08C, 1 h. ii) nBu2Mg, THF, 3 h, 08C.
sium bromide led to the formation of 1 in 32% yield following
crystallisation from a THF/light petroleum solution at room
temperature. Single-crystal X-ray crystallography (Figure 1) re-
vealed, rather than the formation of a dimagnesium alkyl com-
plex, that only one of the lithiated oxygen atoms reacted with
the Grignard reagent, which thus formed a hetero-bimetallic
complex. The lithium cation was found to reside inside the cal-
ix[4]arene cavity with a THF molecule, similar to our previous
observations of other metallocalix[4]arene systems.[17–20] Nota-
bly, the THF molecule was not observed in either the 1H or
13C NMR spectra, which suggests that the THF molecule is
labile enough to be lost in vacuo. The Mg centre, which bears
1
acterised by H and 13C NMR spectroscopy, mass spectrometry,
elemental analysis and IR spectroscopy.
Compound 2: Tripropoxy-p-tert-butylcalix[4]arene (L’H) was
synthesised according to the method of Zhong et al.[21] and
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1892 – 1898 1893

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
contrast the calixÀORÀMg angles are 121.1(2)–121.6(2)8 for the
axial and 133.8(2)8 for the equatorial positions. Compound 2
Table 1. Selected structural data for 1 and V.[13]
1
Bond length [ꢁ]/Angle [8]
1
V
has also been characterised by H and 13C NMR spectroscopy,
mass spectrometry, elemental analysis and IR spectroscopy.
Mg(1)ÀO(1)/(3)
Mg(1)ÀO(2)/O(4)
Mg(1)ÀL
2.373(2)/2.302(2)
1.9261(19)/1.9356(19)
2.146(3)
166.04(16)
162.96(16)
2.232(4)
1.849(4)
2.033(4)
147.2(4)
147.2(4)
Discussion
C(24)ÀO(2)ÀMg(1)
C(8)ÀO(4)ÀMg(1)
Given that Mg compounds have been shown to have excep-
tional activity for the ROP of rac-lactide, higher than their Zn
counterparts,[5,8,14] we initially attempted the synthesis of a di(a-
lkyl magnesium)calix[4]arene, similar to the Zn species of Viga-
lok et al.,[16] by the reaction of two equivalents of di-n-butyl-
magnesium and 1,3-dipropoxy-p-tert-butylcalix[4]arene in THF.
However, this led to the immediate formation of a white pre-
cipitate that we were unable to characterise because of its in-
solubility in common solvents. We suspected that one equiva-
lent of the magnesium alkyl precursor reacted with two of the
phenolic groups rather than the formation of a bimetallic cal-
ix[4]arene as encountered previously in Zn chemistry.[16] To
overcome this, we first lithiated the 1,3-dipropoxy-p-tert-butyl-
calix[4]arene by reaction with n-butyllithium in THF, which re-
moved any phenolic protons. Subsequently, the lithiated cal-
ix[4]arene was reacted with two equivalents of nBuMgBr. This
resulted in the formation of 1, in which only one equivalent of
the Grignard reagent had reacted with the lithiated calix[4]ar-
ene. The reaction proceeded without the formation of a precip-
itate, although it is probable there is the formation of a so-
called ‘turbo-Grignard reagent’ with any excess Grignard re-
agent,[22] the fate of the presumably formed LiBr is currently
unknown. Reaction of the lithiated calix[4]arene with an excess
of nBuMgBr led to the same product. To synthesise a monome-
tallic Mg species, we employed a tripropoxy-p-tert-butylcalix[4]-
arene ligand, which reacted with nBu2Mg in THF to form 2 in
50% isolated yield after crystallisation from a concentrated
light petroleum solution.
was treated with one equivalent of di-n-butylmagnesium in
THF. Single crystals of the product 2·pentane, [L’Mg(nBu)] with
a disordered alkane molecule, suitable for an XRD study, were
grown from a saturated light petroleum solution. Although the
electron density from single-crystal XRD studies indicates a dis-
ordered pentane molecule, it is probable that a number of dif-
ferent alkane molecules from the petroleum fraction used
occupy the calixarene cavity. The solid structure of the com-
pound contains disordered solvent within the cavity, and
a magnesium n-butyl fragment is again bound to the lower
rim of the calix[4]arene (Figure 2). The Mg centre adopts a dis-
Polymerisation using the Mg catalysts
Initially, we attempted the polymerisation of rac-lactide by
using benzyl alcohol (BnOH) as an activator for 1 (100:1:1) in
THF (10 mL). However, if this reaction was quenched with
excess methanol, the formation of methyl-(RS)-lactate rather
than any polymerisation products was observed (Schemes 3
and S1). Clearly, polymerisation did not occur, and a species ca-
pable of ring-opening rac-lactide is generated from reaction
with MeOH on quenching. Sobota and co-workers have report-
ed a Mg catalyst for the chemoselective ring-opening of rac-
lactide similar to our failed quenching method.[25] For subse-
quent screening, we used a drop of dilute HCl (0.1m) to
Figure 2. ORTEP representation of 2·(pentane). Hydrogen atoms, tert-butyl
groups and a pentane molecule located in the calixarene cavity have been
removed for clarity. Displacement ellipsoids are drawn at the 50% probabili-
ty level. Selected bond lengths [ꢁ] and angles [8]: Mg(1)ÀO(1) 1.860(4),
Mg(1)ÀO(2) 2.247(3), Mg(1)ÀO(3) 2.145(3), Mg(1)ÀO(4) 2.255(3), Mg(1)ÀC(54)
2.152(5), O(1)ÀC(1) 1.309(5), O(1)ÀMg(1)ÀO(3) 128.07(15), O(1)ÀMg(1)ÀC(54)
127.0(2), O(3)ÀMg(1)ÀC(54) 104.89(17), O(2)ÀMg(1)ÀC(54) 101.62(18), O(4)À
Mg(1)ÀC(54) 100.88(18), C(1)ÀO(1)ÀMg(1) 175.8(3), O(2)ÀMg(1)ÀO(4)
156.69(12).
ordered trigonal bipyramidal geometry, and the axial O(2)À
MgÀO(4) bond angle is 156.69(12)8. The bond length between
the phenolic O atom and Mg, O(1)ÀMg(1), 1.860(4) ꢁ is signifi-
cantly shorter than the OÀR bond lengths, 2.145(3)–2.255(3) ꢁ,
as expected. The equatorial ROÀMg bond, O(3)ÀMg(1), is the
shortest of the three, 2.145(3) versus 2.255(3) and 2.247(3) ꢁ.
The C(1)ÀO(1)ÀMg(1) bond angle is almost linear, 175.8(3)8; in
Scheme 3. Reaction of 1 with rac-lactide in excess MeOH.
ChemCatChem 2014, 6, 1892 – 1898 1894
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Table 2. ROP of rac-lactide using 1 and 2.
[d]
[c,d]
Run
Catalyst
Solvent
M/ROH
t [min]
Conversion[a] [%]
Pr[a,b]
Mn,GPC
Mn,cald
PDI
1
2
3
4
5
6
7
8
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
toluene
THF
100:1 (MeOH)
100:1 (MeOH)
100:1 (MeOH)
100:0
100:1 (iPrOH)
100:1 (tBuOH)
100:1 (BnOH)
100:2 (MeOH)
100:4 (MeOH)
100:1 (MeOH)
100:1 (MeOH)
100:1 (MeOH)
100:0
100:1 (iPrOH)
100:1 (tBuOH)
100:1 (BnOH)
100:2 (BnOH)
100:4 (BnOH)
100:1 (BnOH)
100:2 (BnOH)
100:4 (BnOH)
60
60
60
480
60
60
60
90
120
120
120
120
30
30
30
3
9.4
35
55
trace
6.3
12
8.8
80
94
55
65
61
28
97
95
92
95
93
94
99
–
–
–
–
–
–
–
0.41
0.42
0.49
0.73
0.30
–
0.79
0.79
0.85
0.78
0.80
0.35
0.35
0.36
–
2.33
15.4
–
–
–
–
5.08
7.96
–
–
–
–
1.09
1.22
–
–
–
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
THF
toluene
THF
THF
THF
THF
–
–
–
4.46
1.79
15.5
12.4
11.7
–
13.6
15.6
14.8
8.92
3.61
10.5
6.17
3.57
5.78
3.40
7.96
9.40
8.82
–
14.0
13.8
13.4
6.90
3.38
13.5
7.20
2.82
1.15
1.19
1.12
1.44
1.98
–
1.46
1.40
1.25
1.34
1.32
1.54
1.54
1.50
9
10
11
12
13
14
15
16
17
18
19
20
21
THF
THF
toluene
toluene
toluene
5
5
5
5
5
99
Conditions: Polymerisation performed by using 60 mmol catalyst at 208C, [LA]0 =0.6m, 10 mL solvent, ROH taken from an ROH/toluene solution. [a] Deter-
mined by NMR spectroscopy. [b] Probability to form an r dyad. [c] kgmolÀ1. [d] Calculated from ([LA]0/[Mg]0)ꢂconversion (%)ꢂ144.13; with the addition of
alcohol, ([LA]0/[ROH]0)ꢂconversion (%)ꢂ144.13+ROH. Mn GPC values corrected by Mark–Houwink factors (0.58) from polystyrene standards in THF.[23,24]
quench the polymerisation. We found that the use of one
equivalent of MeOH in combination with 1 was more active
for the ROP of rac-lactide in dichloromethane (CH2Cl2) rather
than THF or toluene (Table 2, runs 1–3, 55 vs. 35 and 9.4%, 100
equivalents rac-lactide, 60 min). The molecular weights of the
polymer obtained in CH2Cl2 were almost double the expected
values. The degradation of a magnesium butyl compound in
CH2Cl2 was also observed by Chisholm et al.;[8] the magnesium
butyl group has reacted with CH2Cl2 to form a magnesium
chloride moiety incapable of ROP, which leads to a higher than
expected monomer/catalyst ratio. Compound 1 had a low ac-
tivity if isopropanol, tert-butanol or benzyl alcohol were used
instead of methanol, and was inactive without the addition of
any alcohol.
The use of toluene as the solvent with BnOH also gives
a highly active catalytic system with complete conversion of
rac-lactide over 5 min (Table 2, runs 20–22) with good chain
length control and an immortal character.
To assign the stereoselectivity of the polymer produced, we
1
used 2D J-resolved H NMR spectroscopy rather than the more
common homonuclear decoupled spectroscopy. 2D J-resolved
spectroscopy separates the 1D spectrum of PLA (Figure 3) so
that the coupling constants appear on the y axis.[26] A projec-
tion on to the x axis essentially removes all coupling from the
entire spectrum. The stereoselectivity of the polymer can be
easily assigned by reference to the literature.[27] This has the
advantage over the traditional homonuclear decoupled spec-
troscopy used to assign the stereoselectivity as no information
has to be entered manually, which allows an automated ex-
periment. The resulting spectrum from 2D J-resolved spectros-
copy for the assignment of PLA is shown in Figure 4. Com-
pounds 1 and 2 give essentially atactic PLA if CH2Cl2 is used as
the solvent (Table 2, runs 8–10, Pr =0.41–0.49). Compound 2
shows high selectivity for heterotactic PLA in THF (Table 2,
run 14–18, Pr =0.78–0.85) and, surprisingly, isotactic PLA in tol-
uene (Table 2, run 12 and 19–21, Pr =0.30–0.36). The effect of
THF on the selectivity has been discussed previously by Chis-
holm et al.[8] and there are many other examples.[6,11,15,28] We
do not as yet have an explanation for the reversed selectivity
with toluene. Although a number of Mg catalysts have been
explored for the immortal and highly active polymerisation of
l-lactide,[14,29–31] compound 2 is the only catalyst that has ex-
hibits both highly active immortal and stereoselective ROP of
rac-lactide of which we are aware.
In contrast to 1, compound 2 showed increased activities in
THF and toluene, rather than CH2Cl2 (Table 2, runs 10–12). The
molecular weight was higher than expected in all three sol-
vents, much more so in CH2Cl2 than in THF or toluene, which
indicates the degradation of the catalyst. The addition of
iPrOH, tBuOH or BnOH rather than MeOH led to increased ac-
tivities, especially in the case of BnOH, which gave 92% con-
version of 100 equivalents of rac-lactide over 3 min (Table 2,
run 16). The polymerisation screening of 2 with the addition of
MeOH in THF afforded a lower activity than if no alcohol was
added, which indicates that MeOH can deactivate the catalytic
system. The molecular weight of the polymers obtained in THF
using iPrOH, tBuOH and BnOH were close to the expected
values, and additional benzyl alcohol also acts as a chain-trans-
fer agent to control the resultant chain length and give the
catalytic system an ‘immortal’ character (Table 2, runs 16–18).
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1892 – 1898 1895

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Figure 3. 1H NMR spectrum from quenched PLA (Table 2, run 16).
a hetero-bimetallic Li/Mg calix[4]arene [Li(THF)Mg(nBu)L] (1), in
which the Li atom is situated in the calixarene cavity and thus
prevented from reaction with n-butylmagnesium bromide. By
contrast, the reaction between tripropoxy-p-tert-butylcalix[4]ar-
ene and di-n-butylmagnesium in THF resulted in the formation
of [L’Mg(nBu)] (2). Compounds 1 and 2 were characterised by
single-crystal XRD, IR and NMR spectroscopy, elemental analy-
sis and mass spectrometry. Both compounds were active for
the ring-opening polymerisation of rac-lactide. Hetero-bimetal-
lic 1 can convert (94%) 100 equivalents of rac-lactide in di-
chloromethane/methanol in 2 h. Compound 2 is a highly
active and selective ring-opening polymerisation catalyst; 100
equivalents of rac-lactide can be converted to poly(lactic acid)
(PLA; probability to form an r dyad, Pr =0.85, 92%) in 3 min,
and it also leads to the immortal polymerisation of rac-lactide
if THF or toluene is activated with benzyl alcohol (Table 2,
runs 16–21). We have also used a new method for the determi-
nation of the stereoselectivity of the PLA produced; 2D J-re-
solved spectroscopy allows a quick and easy assignment. Com-
pound 2 gives either isotactic or heterotactic bias PLA, which
depends on the solvent employed (THF: Table 2, runs 16–18,
Pr =0.79–0.85; toluene: run 12 and 19–21, Pr =0.30–0.36).
Experimental Section
1
Figure 4. 2D J-resolved H NMR spectrum of the methine region (Table 2,
run 16).
General
All manipulations were performed under an atmosphere of N2
using standard Schlenk and cannula techniques or in a convention-
al N2-filled glovebox. Solvents were heated to reflux over an appro-
priate drying agent and distilled and degassed prior to use. Ele-
mental analyses were performed by the microanalytical service at
London Metropolitan University. NMR spectra were recorded by
using Bruker Ascend 500/300 MHz spectrometers at 298 K; chemi-
cal shifts are referenced to the residual protio impurity of the deu-
Conclusions
We have synthesised two Mg-based calix[4]arene compounds
using either 1,3-dipropoxy-p-tert-butylcalix[4]arene (LH2) or tri-
propoxy-p-tert-butylcalix[4]arene (L’H). The use of nBuLi/1,3-di-
propoxy-p-tert-butylcalix[4]arene resulted in the isolation of
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1892 – 1898 1896

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
terated solvent. IR spectra (Nujol mulls) were recorded by using
PerkinElmer 577 and 457 grating spectrophotometers. LH2 and L’H
were synthesised by the reported procedures.[21,32] rac-Lactide was
purchased from Sigma–Aldrich and used without further purifica-
tion. Gel-permeation chromatography (GPC) was performed by
using a Polymer Laboratories PL-GPC 50 using THF at 0.5 mLminÀ1
flow rate and 308C, corrected by the Mark–Houwink factor (0.58).
33.4, 33.0, 32.6, 32.2, 31.6, 30.4, 29.6, 21.6, 21.5, 13.5, 13.1, 8.2, 7.9,
7.5 ppm.
Polymerisation procedure
Solutions of rac-lactide and catalyst were prepared separately
using the required solvent. The required amount of alcohol, from
a standard alcohol solution in toluene, was added to the catalyst.
The rac-lactide solution was added to the catalyst solution and
stirred for the allotted time at RT under N2. Aliquots (0.5–1.0 mL)
were taken out of the stirred solution if required and quenched
with 1 drop of 0.1m HCl. The aliquots were then dried and ana-
Synthesis of [Li(THF)Mg(nBu)L] (1)
LH2 (4.00 g, 5.46 mmol) was dissolved in THF (50 mL). The solution
was cooled to 08C and nBuLi (10.9 mmol, 1.6m in hexanes,
6.81 mL) was added dropwise. The orange solution was allowed to
warm to RT and stirred for 1 h. The Grignard reagent nBuMgBr
(10.9 mmol), prepared freshly by the reaction between nBuBr
(1.18 mL, 10.9 mmol) and Mg turnings (ꢀ0.3 g) in THF (15 mL),
was added to the lithiated solution at 08C. The solution was al-
lowed to warm to RT and stirred for 1 h. The THF solution was ex-
tracted, then concentrated to approximately 10 mL, and light pe-
troleum (40 mL) was added. After the solution was allowed to
stand at RT overnight, yellow needles of the product formed
(1.54 g, 32% yield). MS (atmospheric solids analysis probe; ASAP):
m/z: 875.6 [MÀCH3]+, 841.5 [MÀPrÀLi+H]; IR (attenuated total re-
flectance; ATR): n˜ =2956 (s), 2904 (m), 2872 (m), 1627 (w), 1600 (w),
1479 (s), 1389 (m), 1362 (m), 1308 (m), 1249 (w), 1191 (m), 1123
(m), 1094 (m), 1043 (w), 999 (w), 870 (m), 830 (w), 797 (w),
531 cmÀ1 (m); elemental analysis calcd (%) for C58H83LiMgO5
(891.52) C 78.14, H, 9.38; found: C 78.29, H 9.44; 1H NMR (C6D6,
sample dried in vacuo for 3 h leads to loss of THF): d=7.27 (s, 4H,
ArH), 7.10 (s, 4H, ArH), 4.53 (d, 4H, J=12.7 Hz, endo-CH2), 4.16 (br t,
4H, OCH2), 3.31 (d, 4H, J=12.7 Hz, exo-CH2), 2.38 (m, 2H,
MgCH2CH2CH2CH3), 1.95 (sex, 2H, J=7.22 Hz, MgCH2CH2CH2CH3),
1.80 (sex, 4H, J=7.84 Hz, OCH2CH2CH3), 1.43 (s, 18H, tBu), 1.32 (t,
3H, J=7.31 Hz, MgCH2CH2CH2CH3), 1.05 (s, 18H, tBu) 0.63 (t, 6H,
J=7.31 Hz, OCH2CH2CH3), 0.47 ppm (m, 2H, MgCH2CH2CH2CH3);
13C NMR (C6D6, sample dried in vacuo for 3 h leads to loss of THF):
d=152.2, 146.7, 137.6, 134.9, 130.5, 128.6, 125.8, 125.2, 78.8, 35.4,
34.2, 34.1, 33.8, 32.5, 32.4, 31.4, 25.2, 14.8, 9.8, 7.0 ppm.
1
lysed by H NMR spectroscopy and GPC.
Crystallography
Crystals were mounted in oil on a glass fibre and fixed in the cold
N2 stream of a Rigaku Saturn724+ diffractometer equipped with
MoKa radiation (l=0.71073 ꢁ) at 100(2) K. The structures were
solved by direct methods and refined by full-matrix least squares
on F2. All hydrogen atoms were placed in calculated positions. Re-
finement was performed using the SHELX-2013 program.[33] The
crystallographic data for
1 and 2 are presented in Table 3.
CCDC 984215 (1) and CCDC 984216 (2·pentane) contain the sup-
plementary crystallographic data for this paper. These data can be
obtained free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
Table 3. Crystallographic data for 1 and 2·(pentane).
Compound
1
2·(pentane)
Formula
Formula weight
C58H83LiMgO5
891.49
C57H82MgO4, C5H12
927.68
Crystal system
Space group
monoclinic
C2/c
monoclinic
P21/c
Unit cell dimensions
Synthesis of [L’Mg(nBu)] (2)
a [ꢁ]
b [ꢁ]
c [ꢁ]
a [8]
b [8]
g [8]
V [ꢁ3]
Z
26.5541(19)
22.8307(16)
21.4201(15)
90
16.4117(7)
13.4182(7)
25.8860(18)
90
L’H (2.0 g, 2.58 mmol) was dissolved in THF (30 mL), cooled to 08C
and nBu2Mg (1.0m in heptane, 2.58 mmol, 2.58 mL) was added
dropwise. After complete addition, the solution was allowed to
warm to RT and then stirred for 3 h. The volatiles were removed in
vacuo, and the product was extracted using light petroleum
(30 mL). The light petroleum solution was concentrated to approxi-
mately 15 mL and upon standing overnight colourless needles
formed. Yield (1.1 g, 50%). MS (EI): m/z: 774 [MÀMgnBu]+; IR (ATR):
n˜ =2956 (s), 2871 (s), 1480 (s), 1390 (w), 1361 (m), 1299 (w), 1261
(w), 1200 (m), 1121 (m), 1105 (m), 1042 (m), 1008 (m), 985 (m), 870
(m), 799 (m), 635 (w), 532 cmÀ1 (m); elemental analysis calcd (%)
for C57H82MgO4 (855.56) C 80.02, H 9.66; found C 79.72, H 9.44;
1H NMR (C6D6): d=7.40 (s, 2H, ArH), 7.39 (s, 2H, ArH), 6.99 (s, 4H,
ArH), 4.58 (d, 2H, J=12.3 Hz, endo-CH2), 4.52 (d, 2H, J=12.1 Hz,
endo-CH2), 4.33 (m, 4H, OCH2), 3.59 (br t, 2H, OCH2), 3.46 (d, 2H,
J=12.3 Hz, exo-CH2), 3.38 (d, 2H, J=12.1 Hz, exo-CH2), 2.28 (m, 2H,
MgCH2CH2CH2CH3), 2.07–1.80 (m, 8H), 1.62 (s, 9H, tBu), 1.40 (s, 9H,
tBu), 1.31 (t, 3H, J=7.3 Hz, MgCH2CH2CH2CH3), 0.79 (s, 18H, tBu)
0.56 (m, 9H, OCH2CH2CH3), 0.28 ppm (m, 2H, MgCH2CH2CH2CH3);
13C NMR (C6D6): d=160.2, 151.2, 148.8, 148.7, 148.5, 146.3, 135.2,
133.7, 133.8, 130.9, 127.5, 125.4, 123.4, 122.9, 79.6, 78.2, 40.5, 34.8,
124.770(2)
90
93.318(7)
90
10667.2(13)
8
5690.9(6)
4
T [K]
100(2)
1.110
100(2)
1.083
Dcalcd [Mg/mÀ3
]
Absorption coefficient,
0.079
0.075
m [mmÀ1
]
Crystal size [mm3]
0.01ꢂ0.07ꢂ0.17
27.5
0.02ꢂ0.10ꢂ0.11
27.5
2qmax [8]
Reflections measured
Unique reflections, Rint
Reflections with F2 >2s(F2)
Transmission factors
(max., min.)
Number of parameters
R1, wR2 [F2 >2s(F2)]
R1, wR2 (all data)
64853
26366
12206
9632
7056
5353
1.000, 0.635
1.000, 0.299
625
621
0.074
0.090
0.192
0.275
Largest difference peak and
0.728, À0.354
0.922, À0.339
hole [eꢁÀ3
]
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1892 – 1898 1897

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[14] Y. Wang, W. Zhao, D. T. Liu, S. H. Li, X. L. Liu, D. M. Cui, X. S. Chen, Orga-
nometallics 2012, 31, 4182.
Acknowledgements
[15] H.-J. Chuang, H.-L. Chen, J.-L. Ye, Z.-Y. Chen, P.-L. Huang, T.-T. Liao, T.-E.
Tsai, C.-C. Lin, J. Polym. Sci. Part A 2013, 51, 696.
[16] E. Bukhaltsev, L. Frish, Y. Cohen, A. Vigalok, Org. Lett. 2005, 7, 5123.
[17] A. Arbaoui, C. Redshaw, M. R. J. Elsegood, V. E. Wright, A. Yoshizawa, T.
Yamato, Chem. Asian J. 2010, 5, 621.
[18] C. Redshaw, D. Homden, D. L. Hughes, J. A. Wright, M. R. J. Elsegood,
Dalton Trans. 2009, 1231.
[19] C. Redshaw, X. M. Liu, S. Z. Zhan, D. L. Hughes, H. Baillie-Johnson,
M. R. J. Elsegood, S. H. Dale, Eur. J. Inorg. Chem. 2008, 2698.
[20] V. C. Gibson, C. Redshaw, W. Clegg, M. R. J. Elsegood, Chem. Commun.
1997, 1605.
1
Colin MacDonald is thanked for help with 2D J-resolved H NMR
experiments, Chrisa Pateraki for help with GPC analysis and
Joseph Wright for help with single-crystal XRD analysis. We are
grateful to the EPSRC National Crystallography Service Centre in
Southampton and the EPSRC National Mass Spectroscopy Service
Centre, Swansea for data.
Keywords: calixarenes
·
heterometallic
complexes
·
[21] Y. Shen, Q. Tang, C. Zhang, W. Zhong, Synlett 2012, 23, 741.
[22] D. R. Armstrong, P. Garcꢅa-ꢆlvarez, A. R. Kennedy, R. E. Mulvey, J. A. Par-
kinson, Angew. Chem. Int. Ed. 2010, 49, 3185; Angew. Chem. 2010, 122,
3253.
[23] I. Barakat, P. Dubois, R. Jꢄrꢇme, P. Teyssiꢄ, J. Polym. Sci. Part A 1993, 31,
505.
[24] J. Baran, A. Duda, A. Kowalski, R. Szymanski, S. Penczek, Macromol.
Rapid Commun. 1997, 18, 325.
[25] A. Grala, J. Ejfler, L. B. Jerzykiewicz, P. Sobota, Dalton Trans. 2011, 40,
4042.
[26] C. Ludwig, M. R. Viant, Phytochem. Anal. 2010, 21, 22.
[27] T. K. Sen, A. Mukherjee, A. Modak, S. K. Mandal, D. Koley, Dalton Trans.
2013, 42, 1893.
[28] H.-Y. Chen, L. Mialon, K. A. Abboud, S. A. Miller, Organometallics 2012,
31, 5252.
[29] V. Poirier, T. Roisnel, J.-F. Carpentier, Y. Sarazin, Dalton Trans. 2011, 40,
523.
[30] C. Romain, V. Rosa, C. Fliedel, F. Bier, F. Hild, R. Welter, S. Dagorne, T.
Aviles, Dalton Trans. 2012, 41, 3377.
magnesium · NMR spectroscopy · ring-opening polymerization
[1] S. Mecking, Angew. Chem. Int. Ed. 2004, 43, 1078; Angew. Chem. 2004,
116, 1096.
[2] E. S. Place, J. H. George, C. K. Williams, M. M. Stevens, Chem. Soc. Rev.
2009, 38, 1139.
[3] O. Dechy-Cabaret, B. Martin-Vaca, D. Bourissou, Chem. Rev. 2004, 104,
6147.
[4] M. Cheng, A. B. Attygalle, E. B. Lobkovsky, G. W. Coates, J. Am. Chem.
Soc. 1999, 121, 11583.
[5] B. M. Chamberlain, M. Cheng, D. R. Moore, T. M. Ovitt, E. B. Lobkovsky,
G. W. Coates, J. Am. Chem. Soc. 2001, 123, 3229.
[6] W. Yi, H. Ma, Inorg. Chem. 2013, 52, 11821.
[7] M. Bouyhayi, Y. Sarazin, O. L. Casagrande, J. F. Carpentier, Appl. Organo-
met. Chem. 2012, 26, 681.
[8] M. H. Chisholm, K. Choojun, J. C. Gallucci, P. M. Wambua, Chem. Sci.
2012, 3, 3445.
[9] S. Song, H. Ma, Y. Yang, Dalton Trans. 2013, 42, 14200.
[10] L. F. Sꢃnchez-Barba, A. Garcꢄs, J. Fernꢃndez-Baeza, A. Otero, C. Alonso-
Moreno, A. Lara-Sꢃnchez, A. M. Rodrꢅguez, Organometallics 2011, 30,
2775.
[31] Y. Wang, W. Zhao, X. Liu, D. Cui, E. Y. X. Chen, Macromolecules 2012, 45,
6957.
[32] K. Iwamoto, K. Araki, S. Shinkai, Tetrahedron 1991, 47, 4325.
[33] G. Sheldrick, Acta Crystallogr. Sect. A 2008, 64, 112.
[11] A. Garcꢄs, L. F. Sꢃnchez-Barba, J. Fernꢃndez-Baeza, A. Otero, M. Honrado,
A. Lara-Sꢃnchez, A. M. Rodrꢅguez, Inorg. Chem. 2013, 52, 12691.
[12] D. M. Homden, C. Redshaw, Chem. Rev. 2008, 108, 5086.
[13] G. Guillemot, E. Solari, C. Rizzoli, C. Floriani, Chem. Eur. J. 2002, 8, 2072.
Received: February 12, 2014
Published online on May 21, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1892 – 1898 1898