Aluminium-catalysed isocyanate trimerization, enhanced by exploiting a dynamic coordination sphere

Article abstract of DOI:10.1039/c9cc03339d

Main-group metals are inherently labile, hindering their use in catalysis. We exploit this lability in the synthesis of isocyanurates. For the first time we report a highly active catalyst that trimerizes alkyl, allyl and aryl isocyanates, and di-isocyanates, with low catalyst loadings under mild conditions, using a hemi-labile aluminium-pyridyl-bis(iminophenolate) complex.

Full text of DOI:10.1039/c9cc03339d

ChemComm
View Article Online
View Journal
COMMUNICATION
Aluminium-catalysed isocyanate trimerization,
enhanced by exploiting a dynamic coordination
Cite this:DOI: 10.1039/c9cc03339d
sphere†‡
Received 30th April 2019,
Accepted 14th May 2019
Mohammed A. Bahili,ab Emily C. Stokes,
Robert C. Amesbury,a
a
Darren M. C. Ould,a Bashar Christo,a Rhian J. Horne,a Benson M. Kariuki,
a
Jack A. Stewart,a Rebekah L. Taylor,a P. Andrew Williams,a Matthew D. Jones,
c
DOI: 10.1039/c9cc03339d
a
a
Kenneth D. M. Harris
and Benjamin D. Ward
*
rsc.li/chemcomm
Main-group metals are inherently labile, hindering their use in
catalysis. We exploit this lability in the synthesis of isocyanurates.
For the first time we report a highly active catalyst that trimerizes
alkyl, allyl and aryl isocyanates, and di-isocyanates, with low catalyst
loadings under mild conditions, using a hemi-labile aluminium-
pyridyl-bis(iminophenolate) complex.
Scheme 1 Trimerization of isocyanates.
Catalytic applications of Earth-abundant main-group metals, boards and are responsible for their mechanical and thermo-
such as aluminium, have compelling environmental and eco- setting properties.10
nomic benefits. In contrast to the extensively-used platinum
Isocyanates can be trimerized by amines,11 phosphines,12–15
group metals, they are non-toxic, inexpensive, and readily avail- N-heterocyclic carbenes,16 inorganic salts,17 phosphides,18,19 and
able;1 furthermore, the requirement to reprocess the metals via main group and transition metal complexes,20–28 but existing
complex, expensive, and polluting recycling processes is there- catalysts often require high catalyst loadings, high temperatures
fore less critical. However, the widespread use of main group (e.g. Z100 1C for amines and halides),11,17 can give a mixture of
metals in catalysis is hindered by the fact that they often exhibit oligomers, and/or fail to trimerize both alkyl and aryl isocyanates.
labile and unpredictable coordination chemistry, leading to In this communication we report the first catalyst that excels in
ligand redistribution and poorly-defined catalyst species. We all of these aspects, specifically an aluminium complex that selec-
demonstrate here that such lability can actually be exploited as a tively trimerizes alkyl, allyl and aryl isocyanates, and di-isocyanates,
part of the catalyst design to give a dynamic coordination at mild temperatures, with low catalyst loadings.
environment that enhances catalytic performance.
Preparation of aluminium complexes. The 6-coordinate aluminium
Trimerization of isocyanates (RNCO, Scheme 1) is the most complex [Al(Salpy)(OBn)] (3) incorporating the pyridyl-bisimino-
rapid, economical and atom-efficient route to isocyanurates, which phenolate (Salpy) ligand (1)29 was found to trimerize isocyanates
are used in a diverse range of applications, such as medicines,2–5 rapidly at ambient or modest temperatures; this is remarkable since
selective anion binding,6 microporous materials7,8 and coating the 5-coordinate pyridyl-free analogue [Al(Salpn)(OBn)] [Salpn =
materials.9 However, isocyanurates’ principal application is in bis(salicylidene)propylenediamine]30 was found to be unable to
the construction industry; they are the key component in rigid trimerize isocyanates, even under forcing conditions.§ The high
polyurethane foams which are used ubiquitously as insulation activity of 3 is surprising, since a coordinatively-saturated complex is
expected to be less active than a coordinatively-unsaturated one.
Reaction of the Salpy pro-ligand with AlMe3 afforded [Al(Salpy)-
Me] (2) in good yield (Scheme 2). Further reaction of 2 with
alcohol/phenol afforded [Al(Salpy)(OR)] [R = Bn, CH2Ph, (3) or
Tol, 4-C6H4Me, (4)].
Isocyanate trimerization. The reaction of 3 with phenyl iso-
cyanate (PhNCO; 500 equiv.) completely solidified within 1 h at
50 1C or 18 h at 25 1C (Table 1), corresponding to essentially
quantitative conversion of the isocyanate to N,N0,N00-triphenyl
isocyanurate. NMR and mass spectra indicate that the crude
product mixture is pure (only trimer is detected), and the bulk
a School of Chemistry, Cardiff University, Main Building, Park Place,
Cardiff CF10 3AT, UK. E-mail: WardBD@Cardiff.ac.uk
b Department of Science, College of Science, University of Basrah, Basrah, Iraq
c Department of Chemistry, University of Bath, Claverton Down, Bath BA2 7AY, UK
† The data supporting the results presented in this article are freely available via
the Cardiff University Data Catalogue at http://dx.doi.org/10.17035/d.2019.
0069570379.
‡ Electronic supplementary information (ESI) available: Full experimental details
and characterizing data, Cartesian coordinates of all calculated species. CCDC
1896206–1896216. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/c9cc03339d
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun.

View Article Online
Communication
ChemComm
(TDI), phenylene di-isocyanate (PDI) and methylenedi(phenyl-
isocyanate) (MDI) were all trimerized using complex 3 (both
PDI and MDI were reacted in the melt as they are solids at room
temperature). For TDI and MDI, there was clear evidence (from
mass spectra) for oligomeric species arising from the reaction
of both isocyanate moieties, and simple trimers were also
evident. The lower intensity nNCO infrared band for the product
obtained from MDI compared to those for TDI indicates a
greater propensity for MDI to form higher-order oligomers.
Mechanistic aspects. The structures of 2 and 4 were con-
firmed by single-crystal XRD (S2.3 and S2.4, ESI‡)35 and by
elemental analysis and NMR spectroscopy. A second set of
resonances was observed for 2 and 3 (but not for 4, suggesting
that the second species is present in very low concentration). In each
case, several signals due to the minor component are obscured by
those of the major component, but the data are consistent with the
two components being isomers; the pyridyl H6 signal for the minor
isomer has a comparable chemical shift to the uncoordinated
ligand (8.53 ppm for 2; 8.58 ppm for 3), while the pyridyl H6 signal
for the major component is further downfield (8.69 ppm for 2
and 9.10 ppm for 3). Since the chemical shift of H6 is indicative of
the coordinative state of the pyridyl,36 our observations suggest
that the minor isomer has a de-coordinated pyridyl.
Scheme 2 Aluminium complexes 2–4.
product was characterized by powder X-ray diffraction (XRD).
Unit cell determination and profile fitting (using the Le Bail
method) led to an excellent fit to the experimental powder XRD
pattern (S1.1, ESI‡). The unit cell was in excellent agreement
with that of the monoclinic polymorph of the cyclotrimer
reported previously.31 Rietveld refinement of the powder XRD
data confirmed that the crystalline product was the monoclinic
polymorph of the cyclotrimer, from which the molecular struc-
ture of the product was confirmed (see ESI‡ for more details).32
From our combined analyses, we conclude that the alumi-
nium catalyst is completely selective (within detection limits) in
forming the cyclotrimer, in contrast to other catalytic systems
that give various side products in addition to the isocyanurate
and/or require higher catalyst loadings at higher temperatures.17,33
Complex 3 is active for the trimerization of alkyl, aryl and
allyl isocyanates, as well as di-isocyanates (Table 1), including
those of commercial relevance (especially di-isocyanates), and
those that can be incorporated into polymers as cross-linking
agents (e.g. allyl isocyanate).34 From our results, it is clear that
complex 3 is an exceptionally versatile catalyst, especially when
considering that only low catalyst loadings are required.
Di-isocyanates find commercial use in rigid polyurethane
foams and are routinely used in cavity wall, roof, and under-
floor insulation. In the present work, toluene di-isocyanate
From 1H NMR spectra of 3 measured between 298 K and
323 K, the equilibrium constant for the major " minor component
interchange was determined at each temperature and a van’t Hoff
plot (S1.6, ESI‡) gave DH~ = 13.6 ꢀ 4.2 kJ molꢁ1 and DS~ = 29.5 ꢀ
13.5 J Kꢁ1 molꢁ1. Hence, DG~ at 298 K is estimated to be
4.8 kJ molꢁ1. The small value of DS~ is consistent with a
unimolecular process rather than a monomer–dimer equilibrium,
and the increase in entropy on going from the major to the minor
component suggests that the major component has the coordinated
pyridyl. The two components give almost identical diffusion
coefficients in DOSY NMR spectra (S1.6 and Fig. S5, ESI‡).
The mechanism for the trimerization of MeNCO by [Al(Salpy)-
(OMe)] (5calc) was probed using DFT calculations. It is common
for isocyanate trimerization to proceed via nucleophilic attack on
the isocyanate carbon,10 and there are several examples of
isolated species arising from the insertion of isocyanates into
metal–ligand bonds.18,23,37–39 This leads us to propose a catalytic
cycle (Scheme 3) based on a repeated coordination–insertion
mechanism.¶ A crucial component of this mechanism is that
pyridyl dissociation must occur before the isocyanate can associate
to allow the 1,2-migratory insertion of the methoxide onto the
isocyanate carbon; importantly, de-coordination of the pyridyl is
facile (S to INT1: DG = +16 kJ molꢁ1). Coordination of the isocyanate
to the aluminium (INT2) is endergonic (Grel = +52 kJ molꢁ1); INT2
lies only slightly lower in energy than the transition state TS1 and is
therefore likely to be synchronous with the insertion step.
Table 1 Catalytic data for the trimerization of isocyanates and di-isocyanates
by [Al(Salpy)(OBn)] (3)
Entry
Substrate
T (1C)
ta (h)
Yieldb (%)
1
2
PhNCO
PhNCO
25
50
25
50
25
25
50
50
50
25
25
25
25
50
100c
25
50c
18
1
48
18
6
48
18
0.5
3.5
3
98
95
25
98
93
95
95
97
76
77
59
98
5
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
4-Me-C6H4NCO
4-Me-C6H4NCO
4-F-C6H4NCO
2-F-C6H4NCO
4-CF3-C6H4NCO
4-Cl-C6H4NCO
4-MeO-C6H4NCO
C6H5CH2NCO
H2CQCHCH2NCO
EtNCO
3
Bader’s quantum theory of atoms in molecules analysis of
INT2 indicates a bond-critical point (BCP) between the associated
isocyanate carbon and the methoxide oxygen, (electron density
r = 0.09 e Åꢁ3) which indicates that the isocyanate is pre-
organized for the migratory insertion of methoxide, i.e. the BCP
lies along the reaction coordinate. All migratory insertion steps
(TS1 to TS3) can be described (natural bonding orbital analyses)
0.5
18
18
1.5
1.5
48
tBuNCO
tBuNCO
1,4-(NCO)2C6H4 (PDI)
1,3-(NCO)2-4-Me-C6H3 (TDI)
CH2(4-NCO-C6H4)2 (MDI)
96
97
89
90
a
Time taken for reaction mixture to solidify or until no further conversion
b
c
attained. Isolated yields. Reaction carried out in the melt.
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2019

View Article Online
ChemComm
Communication
Scheme 3 Proposed catalytic cycle for the trimerization of MeNCO using [Al(Salpy)(OMe)] (5calc). [Al] = [Al(k4-Salpy)]; [Al]* = [Al(k5-Salpy)]. Gibbs
energies (Grel) in kJ molꢁ1 calculated using M06-2X: cc-pVTZ/cc-pV(T+d)Z are shown alongside each label. The free energy profile is provided in the
ESI.‡ Boxed sections highlight the pyridyl hemi-lability and its role in enhancing the catalytic activity.
as a methoxide/imine sp2 hybrid donating to an empty p orbital
of the isocyanate carbons (S3.2, ESI‡).
As demonstrated in this work, the presence of a hemi-labile
pyridyl donor makes the aluminium alkoxide complex [Al(Salpy)-
The product from the migratory insertion initially affords an (OBn)] (3) a highly active and selective catalyst for the trimerization
O,O0-coordinated isocarbamate ligand (INT3), but as this is of isocyanates under mild conditions. Calculations suggest that the
endergonic relative to INT2, the insertion is expected to be pyridyl donor facilitates the partial de-coordination of the growing
readily reversible without intervention of the pyridyl. However, trimer, thereby disfavouring the de-insertion reaction. Whilst this is
pyridyl coordination affords a monodentate isocarbamate that an important development for isocyanurate synthesis, the concept
has significantly lower energy (INT4, Grel = +11 kJ molꢁ1). of an additional donor group enhancing catalytic performance
Rather than re-coordination of the OMe, imine coordination is may have profound implications for the use of labile main group
more favourable, giving OC1 which must be an off-cycle inter- metals in a wider range of catalytic reactions, and we anticipate
mediate as the imine lone pair is unavailable for the subsequent significant research opportunities with the aim of further
migratory insertion to isocyanate. The energetics of INT4, OC1 exploiting this concept.
and INT3 support a resting state pyridyl-imine coordination
This research was supported by the Iraqi Ministry of Higher
equilibrium between INT4 and OC1; although re-formation of Education (studentship to MAB), the Cardiff Partnership Fund
INT3 is energetically viable, an INT4/OC1 equilibrium locks the (postdoctoral fellowship to ECS), Kingspan Insulation Ltd, and
isocarbamate into a configuration that conformationally disfa- the EPSRC (access to the National crystallography Service and
vours OMe re-coordination by placing it away from the Al centre, the National Mass Spectrometry Facility, studentship to RCA,
reducing the likelihood of de-insertion and promoting the EP/L016443/1). Access to the Advanced Research Computing
forward reaction, thus explaining the importance of the pyridyl facility (ARCCA) at Cardiff University is gratefully acknowledged.
donor in enhancing catalytic performance, consistent with experi-
mental observations.
Insertion of subsequent isocyanates proceeds in a similar
manner to the first. As the stepwise formation of an isocyanurate
Conflicts of interest
There are no conflicts to declare.
affords a linear series of amide linkages with rigid 1201 bond
angles, it follows that, after the third isocyanate insertion has
taken place, the OMe group at the terminus is necessarily close to
the imino nitrogen from the third isocyanate insertion, leading to
Notes and references
§ Reactions with [Al(Salpn)(OBn)] were performed at 20–100 1C for
24 hours. Isocyanurate was obtained only for 4-fluorophenylisocyanate
with o15% conversion, i.e. with the substrate expected to have the most
cyclization to afford the isocyanurate and the regeneration of 5calc
.
For the overall reaction, DG is calculated to be ꢁ139 kJ molꢁ1
.
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun.

View Article Online
Communication
ChemComm
+
electrophilic isocyanate carbon. No reaction was observed for other 19 D. Heift, Z. Benko, H. Gru¨tzmacher, A. R. Jupp and J. M. Goicoechea,
substrates. Chem. Sci., 2015, 6, 4017–4024.
¶ Stoichiometric reactions on an NMR tube scale gave rise to products 20 A. Hernan-Gomez, T. D. Bradley, A. R. Kennedy, Z. Livingstone,
consistent with a sequential insertion of isocyanates, as predicted by S. D. Robertson and E. Hevia, Chem. Commun., 2013, 49, 8659–8661.
our hypothesized mechanism. Since, the modus operandi of 21 X. Zhu, J. Fan, Y. Wu, S. Wang, L. Zhang, G. Yang, Y. Wei, C. Yin,
nucleophile-based catalysts is well established, and several isolated H. Zhu, S. Wu and H. Zhang, Organometallics, 2009, 28, 3882–3888.
species reported in the literature relate to isocyanate insertion into M–L 22 Y. Wu, S. Wang, X. Zhu, G. Yang, Y. Wei, L. Zhang and H. Song,
´
´
bonds, there is tangible experimental verification for the validity of the
mechanism shown in Scheme 3. See ESI‡ (S1.8) for full details.
Inorg. Chem., 2008, 47, 5503–5511.
23 X. Zhou, L. Zhang, M. Zhu, R. Cai, L. Weng, Z. Huang and Q. Wu,
Organometallics, 2001, 20, 5700–5706.
24 H. R. Sharpe, A. M. Geer, H. E. L. Williams, T. J. Blundell, W. Lewis,
A. J. Blake and D. L. Kays, Chem. Commun., 2017, 53, 937–940.
25 A. J. Bloodworth and A. G. Davies, J. Chem. Soc., 1965, 6858–6863.
26 F. Moghaddam, M. Dekamin and G. Koozehgari, Lett. Org. Chem.,
2005, 2, 734–738.
27 A. Flamini, A. M. Giuliani and N. Poli, Tetrahedron Lett., 1987, 28,
2169–2170.
´
1 K. V. Ragnarsdottir, Nat. Geosci., 2008, 1, 720–721.
2 P. Gibbons, D. Love, T. Craig and C. Budke, Vet. Parasitol., 2016, 218,
1–4.
3 M. Ghosh and M. J. Miller, J. Org. Chem., 1994, 59, 1020–1026.
4 A. Bosco, L. Rinaldi, G. Cappelli, A. Saratsis, L. Nisoli and
G. Cringoli, Vet. Parasitol., 2015, 212, 408–410.
5 A. P. Murray and M. J. Miller, J. Org. Chem., 2003, 68, 191–194.
6 M. Mascal, I. Yakovlev, E. B. Nikitin and J. C. Fettinger, Angew.
Chem., Int. Ed., 2007, 46, 8782–8784.
28 S. G. Lee, K.-Y. Choi, Y.-J. Kim, S. Park and S. W. Lee, Dalton Trans.,
2015, 44, 6537–6545.
7 Y. Zhang, S. N. Riduan and J. Y. Ying, Chem. – Eur. J., 2009, 15, 1077–1081.
8 E. Preis, N. Schindler, S. Adrian and U. Scherf, ACS Macro Lett., 2015,
4, 1268–1272.
9 H. Ni, A. D. Skaja, R. A. Sailer and M. D. Soucek, Macromol. Chem.
Phys., 2000, 201, 722–732.
10 M. F. Sonnenschein, Polyurethanes: science, technology, markets, and
trends, Wiley, Hoboken, New Jersey, 2015.
11 Y. Taguchi, I. Shibuya, M. Yasumoto, T. Tsuchiya and K. Yonemoto,
Bull. Chem. Soc. Jpn., 1990, 63, 3486–3489.
29 R. Shakya, A. Jozwiuk, D. R. Powell and R. P. Houser, Inorg. Chem.,
2009, 48, 4083–4088.
30 H. Du, X. Pang, H. Yu, X. Zhuang, X. Chen, D. Cui, X. Wang and
X. Jing, Macromolecules, 2007, 40, 1904–1913.
31 A. Usanmaz, Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst.
Chem., 1979, 35, 1117–1119.
32 K. D. M. Harris, Advanced X-Ray Crystallography, Springer, Berlin,
Heidelberg, 2012, vol. 315, pp. 133–177.
33 F. Paul, S. Moulin, O. Piechaczyk, P. Le Floch and J. A. Osborn, J. Am.
Chem. Soc., 2007, 129, 7294–7304.
´
´
12 Z. Pusztai, G. Vlad, A. Bodor, I. T. Horvath, H. J. Laas, R. Halpaap
and F. U. Richter, Angew. Chem., Int. Ed., 2006, 45, 107–110.
34 W. Sun, X. Yan and X. Zhu, J. Appl. Polym. Sci., 2011, 122, 2359–2367.
13 S. M. Raders and J. G. Verkade, J. Org. Chem., 2010, 75, 5308–5311. 35 S. J. Coles and P. A. Gale, Chem. Sci., 2012, 3, 683–689.
14 J. Tang, T. Mohan and J. G. Verkade, J. Org. Chem., 1994, 59, 4931–4938. 36 S. Friedrich, M. Schubart, L. H. Gade, I. J. Scowen, A. J. Edwards and
15 X. Liu, Y. Bai and J. G. Verkade, J. Organomet. Chem., 1999, 582, 16–24.
16 H. A. Duong, M. J. Cross and J. Louie, Org. Lett., 2004, 6, 4679–4681. 37 L. Orzechowski and S. Harder, Organometallics, 2007, 26, 2144–2148.
17 Y. Nambu and T. Endo, J. Org. Chem., 1993, 58, 1932–1934.
38 H. Wang, H.-W. Li and Z. Xie, Organometallics, 2003, 22, 4522–4531.
18 W. Yi, J. Zhang, L. Hong, Z. Chen and X. Zhou, Organometallics, 39 W. Uhl, J. S. Bruchhage, M. Willeke, A. Hepp and J. Ko¨sters, Eur.
M. McPartlin, Chem. Ber./Recl., 1997, 130, 1751–1759.
2011, 30, 5809–5814.
J. Inorg. Chem., 2016, 2721–2730.
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2019