Mechanistic Analysis and Characterization of Intermediates in the Phosphane-Catalyzed Oligomerization of Isocyanates

Article abstract of DOI:10.1002/chem.201804016

The mechanism of the oligomerization of aliphatic isocyanates catalyzed by trialkylphosphanes has been studied through low temperature ³¹P and ¹⁵N NMR spectroscopy combined with computational chemistry. A revised mechanism is proposed that contains several (spiro)cyclic pentacoordinate phosphorous intermediates. Previously reported spectroscopic data of a transient intermediate has been reevaluated and assigned to a cyclic intermediate containing a P?N bond by experiments with ¹⁵N-labeled isocyanate. ¹³C, ¹⁵N, and ³¹P NMR shifts that support this assignment have been calculated using quantum chemical methods.

Full text of DOI:10.1002/chem.201804016

A Journal of
Accepted Article
Title: Mechanistic Analysis and Characterization of Intermediates in the
Phosphane-Catalyzed Oligomerization of Isocyanates
Authors: Hendrik Zipse, Julian Helberg, and Yohei OE
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201804016
Link to VoR: http://dx.doi.org/10.1002/chem.201804016
Supported by

10.1002/chem.201804016
Chemistry - A European Journal
COMMUNICATION
Mechanistic Analysis and Characterization of Intermediates in the
Phosphane-Catalyzed Oligomerization of Isocyanates
Julian Helberg,[a] Yohei OE,[b] Hendrik Zipse*[a]
Abstract: The mechanism of the oligomerization of aliphatic
isocyanates catalyzed by trialkylphosphanes has been studied
through low temperature 31P and 15N NMR spectroscopy combined
with computational chemistry. A revised mechanism is proposed that
contains several (spiro)cyclic pentacoordinate phosphorous
intermediates. Previously reported spectroscopic data of a transient
intermediate has been reevaluated and assigned to a cyclic
intermediate containing a P-N bond by experiments with 15N-labeled
isocyanate. 13C, 15N and 31P NMR shifts that support this assignment
have been calculated using quantum chemical methods.
catalyzed oligomerization of isocyanates, Goodman et al.
proposed another pathway to the trimers 3 and 4 by reaction of
monoadduct 6 with uretdione 2. They also suggested stabilization
of intermediate 7 as a five membered ring in which O coordinates
to P (labeled 7O in this study).[5]
The 31P NMR signal at −54.0 ppm assigned by Horvath et al. to
the acyclic, tetracoordinated phosphane intermediate 7L does
neither remotely fit the chemical shift (>0 ppm) usually associated
with tetracoordinated phosphorus,[2a,6] nor the one published for
monoadduct 6 of azaphosphatrane 14 and phenyl isocyanate
(+29.46 ppm) by Verkade.[7] This prompted us to reinvestigate the
mechanism proposed for the phosphane-catalyzed oligo-
merization of alkyl isocyanates 1, using a combination of
computational NMR shift predictions and experimental NMR
measurements employing 15N labeled isocyanate. Herein, we
The oligomerization of aliphatic isocyanates 1 using Lewis base
catalysts is an important industrial process for the synthesis of
building blocks for highly durable polyurethane (PU) coatings.[1]
The use of oligomers provides a higher degree of safety during
the polymerization process due to significantly lower vapor
pressures and provides the basis for tuning polymer properties.[2]
Addition of (aliphatic) isocyanurates 3 often enhances the
physical properties of polymers, for example in the production of
flame retardant materials used for the lamination of electrical
devices.[3] Depending on the Lewis base catalyst and the reaction
conditions, the oligomerization of isocyanates generates variable
amounts of uretdiones 2, isocyanurates 3, iminooxadiazinediones
4, and higher oligomers (Scheme 1A). In the case of
show
a quantitative assignment for the signals reported
previously and in consequence, propose a new mechanism
including key intermediate 7N for the phosphane-catalyzed
oligomerization of aliphatic isocyanates (Scheme 1B).
trialkylphosphane-catalyzed
oligomerizations
of
aliphatic
isocyanates, the reaction is believed to be initiated through attack
of phosphane 5 at the central carbon atom of isocyanate 1 and
formation of zwitterion 6 as a first transient intermediate.
Subsequent reaction of 6 with a second equivalent of isocyanate
then generates acyclic, tetracoordinated phosphane intermediate
7L, whose detection by low temperature 31P and 13C NMR
spectroscopy was reported by Horvath et al. for the
oligomerization of aliphatic isocyanate 1 catalyzed by tri-alkyl-
phosphane 5.[2a,4] The reaction then either proceeds through
elimination of catalyst 5 and (reversible) formation of uretdione 2,
or addition of a third isocyanate monomer and elimination of
catalyst 5 to form isocyanurate 3 or iminooxadiazinedione 4.
Based on a computational study on the azaphosphatrane 14-
Scheme 1. Simplified mechanism showing key intermediates in the phosphane-
catalyzed oligomerization of aliphatic isocyanates 1 in previous studies (A) and
in the current study (B).
[a]
[b]
J. Helberg, Prof. H. Zipse, , Germany
Department of Chemistry, LMU München
Butenandtstrasse 5-13, 81377 München (Germany)
E-mail: zipse@cup.uni-muenchen.de
http://www.cup.lmu.de/oc/zipse/
Prof. Y. OE
Graduate School of Life and Medical Sciences
Department of Biomedical Information, Faculty of Life and Medical
Sciences, Doshisha University
Initially, we reproduced the kinetic measurements and low-
temperature NMR experiments published by Horvath et al. and
were able to confirm these results while using a different analytical
method for the kinetics.[2a] Figure 1 shows the time-resolved
1–3 Tatara Miyakodani, Kyotanabe, Kyoto 610-0394 (Japan)
1
oligomerization of n-hexyl isocyanate 1a measured by H NMR
which indicates that the initial formation of dimer 2a and
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201804016
Chemistry - A European Journal
COMMUNICATION
subsequent formation of trimers 3a and 4a proceeds in a tightly
coupled fashion.
Scheme 2. Conceivable structures for intermediate 7ab and H+-trapped model
compound 6abxHCl.
This allows the differentiation of the possible intermediates shown
in Scheme 2 based on their 31P/15N couplings and 15N chemical
shifts. 1a15 was obtained in three steps starting from heptanoic
anhydride and 15NH4Cl (see SI). In order to obtain a stable
reference compound structurally related to the potential
intermediate, the H+-trapped phosphonium 6ab×HCl and its 15N-
labeled analogue 6ab15×HCl were synthesized in situ by mixing
of equimolar amounts of isocyanate 1a and tri-n-
butylphosphonium chloride 16b (see SI). This resulted in next-to-
quantitative formation of 6ab×HCl as verified by 1H, 31P, 13C NMR
spectroscopy and high resolution mass spectroscopy. The 31P
NMR spectrum of compound 6ab×HCl shows a singlet at a typical
"phosphonium" chemical shift of +29.6 ppm that turns into a
doublet (J = 28.4 Hz) in the case of the 15N-labled 6ab15×HCl. The
15N NMR spectrum of this compound shows a doublet of doublets
at −236.54 ppm, formed by 1H/15N coupling (J = 90.7 Hz) and
31P/15N coupling (J = 28.4 Hz).
Figure 1. Time-resolved oligomerization of 1a observed by 1H NMR
spectroscopy.
Figure 2. 31P{1H} NMR spectra of the reaction of 1a and 5b (50 mol-%): top: at
−20 °C; bottom: spectra at different temperatures.
Low-temperature NMR measurements were performed under
neat conditions on oligomerization reactions of isocyanate 1a
catalyzed by 50 mol-% 5b. Figure 2 (top) shows the proton
decoupled 31P NMR spectrum at −20 °C with integrals for two
species: catalyst 5b and intermediate 7ab. Bis-adduct 7ab
appears as a temperature-dependent signal at −55.7 ±0.4 ppm
with an integral of approx. 0.5% of free phosphane 5b that
broadens and shifts with increasing temperature (bottom). Since
we suspected the signal of intermediate 7ab to represent one of
the cyclic, pentacoordinate intermediates 7Nab or 7Oab rather
than acyclic, tetracoordinated 7Lab (Scheme 2), the low
temperature NMR measurements were repeated with 15N labeled
n-hexyl isocyanate 1a15.
Figure 3. NMR spectra of the reaction of 1a15 and 5b (50 mol-%) at −40 °C:
31P{1H} spectrum showing intermediate 7Nab15 to form a doublet of doublets
(top) and 15N spectrum with the corresponding doublets (bottom).
Repeating the earlier NMR measurements at −40 °C using 1a15
and 50 mol-% 5b yields a proton decoupled 31P spectrum where
the intermediate signal appears as a well-defined doublet of
doublets with coupling constants of J = 50.6 Hz and J = 27.7 Hz
(Fig. 3). 15N NMR shows the corresponding doublets at
−218.2 ppm (J = 31.5 Hz) and −243.7 ppm (J = 53.6 Hz).
This article is protected by copyright. All rights reserved.

10.1002/chem.201804016
Chemistry - A European Journal
COMMUNICATION
Figure 4. Potential energy surface of the oligomerization of ethyl isocyanate 1c catalyzed by triethylphosphane 5c [in kJmol−1].
The minor deviation in the coupling constants results most likely
from the 15N NMR being measured in a proton coupled fashion.
This is required because the doublet at −243.7 ppm does not
appear in proton decoupled 15N NMR, presumably due to Nuclear
Overhauser Effects. The doublet at -218.2 ppm gives a coupling
constant of J = 27.5 Hz in the proton decoupled measurements in
complete agreement with the 31P NMR results (see SI).
rather polar conditions in the oligomerization mixture.[9] Figure 4
shows the potential energy surface (gas phase ∆H298 and solution
phase ∆G298) for the triethylphosphane-(5c)-catalyzed oligo-
merization of ethyl isocyanate 1c. In the following, the discussion
focusses on the gas-phase enthalpy values (∆H298). It is
noteworthy, that acyclic intermediate 7Lcc is much less stable
than the corresponding cyclic isomers 7Occ and 7Ncc, which are
both located significantly lower in energy. Of these two cyclic bis-
adducts, 7Ncc is slightly preferred.
Since the 15N atoms show coupling only to the 31P atom,
assignment of both signals is possible based on the spectroscopic
data for 6ab15×HCl. Comparison of the coupling constants in 15
N
NMR spectroscopy clearly shows, that the doublet generated by
intermediate 7Nab at −218.17 ppm (J = 27.5-31.5 Hz) represents
the nitrogen atom next to the P-bound carbonyl group. Since the
other 15N doublet at −243.74 ppm (J = 53.6 Hz) shows a much
larger coupling constant and therefore stronger coupling, we see
this as evidence of the second 15N atom being closer to the
phosphorous atom in the detected intermediate. This strongly
supports assignment of cyclic, pentacoordinate phosphane
7Nab15 as the observed intermediate as highlighted in Figure 3.
The preference of phosphorous atoms to form cycles by bonding
to isocyanate nitrogen atoms rather than the oxygen atoms is not
unprecedented. Neutral cyclic monoalkyl phosphanes structurally
similar to 7Nab were generated by condensation of phospholenes
and isocyanates, and more recently Grützmacher et al. reported
several cyclic, anionic P-N bond containing species found in the
cyclo-oligomerization of isocyanates by “P−“ anions.[8]
Having obtained experimental evidence that the observed
intermediate is cyclic, pentacoordinate phosphane 7Nab, we
decided to use computational chemistry to further elucidate the
reaction mechanism. We found that consistent results were
obtained at the B2PLYP/cc-pVTZ//B3LYP/6-31+G(d,p) level of
theory and used a SMD(CHCl3) solvation model to simulate the
Scheme 3. Proposed reaction mechanism of the (alkyl) phosphane – catalyzed
oligomerization of alkyl isocyanates formulated for all-ethyl substituents in
analogy to the computational study.
This article is protected by copyright. All rights reserved.

10.1002/chem.201804016
Chemistry - A European Journal
COMMUNICATION
While initial formation of the cyclic intermediates 7Occ and 7Ncc
is possible through reaction of monoadduct 6cc with another
monomer 1c, interconversion is also possible via transition state
TSRcc located +58.8 kJmol−1 above 7Ncc. Uretdione 2c is
exclusively formed from cyclic intermediate 7Ncc through
transition state TSD0cc. Dimer 2c is then able to react with
monoadduct 6cc forming previously unreported spiro-
intermediate 9cc, which is conceptually similar to the spiro-
tetramers formed by reductive oligomerization in the presence of
silanes reported by Süss-Fink.[10] Acyclic, trimeric intermediate
8cc is either formed by interconversion of 9cc through transition
state TSD2cc or by addition of one molecule of 1c to cyclic adduct
7Occ through transition state TS3cc. Trimers 3c and 4c are finally
obtained from 8cc via the low-lying transition states TS5cc and
TS4cc. Based on both the experimental and computational
results, we propose the revised reaction mechanism shown in
Scheme 3.
In conclusion, we provide here a comprehensive study on the
phosphane-catalyzed oligomerization of aliphatic isocyanates,
which includes the validation of earlier 31P NMR and kinetic
measurements,[2a] and the combined experimental/theoretical
analysis of 15N and 31P NMR spectral data in reactions of 15N-
labeled isocyanates. Calculation of the potential energy surface
leads us to a revised reaction mechanism featuring previously
unknown spiro intermediate 9 formed from uretdione 2. Cyclic
intermediate 7N is responsible for the signals visible at low
temperature NMR measurements as shown by the prediction of
31P, 15N and 13C chemical shifts. Further evidence was obtained
from the 31P-15N couplings and the multiplicities found for signals
in low temperature NMR measurements. The new mechanism
provides the basis for a better understanding of an important
industrial process and its possible improvement through the
development of new catalyst systems. This study also shows the
potential of the employed methodology to solve further
For all intermediates found in the revised mechanism, 13C, 15N,
and 31P NMR chemical shifts were predicted at the B3LYP/IGLO-
III//B3LYP/6-31+G(d,p) level of theory (see SI). The methodology
used combines theoretically calculated shieldings for the ethyl-
substituted systems shown in Figure 4 with a system of
increments for the influence of longer alkyl chain substituents. The
predicted 31P chemical shifts perfectly fit our assignment of cyclic,
pentacoordinate intermediate 7Nab to the signals reported by
Horvath and Richter (theor. predicted -55.2 ppm, exp. found -55.6
unanswered
questions
of
isocyanate
activation
by
organocatalysts.[11]
Acknowledgements
The authors wish to thank Dr. D. Stephenson (LMU München) for
recording the low-temperature NMR spectra, Dr. F. Richter
(Covestro AG) for fruitful discussions and the Leibnitz
Supercomputer Centre (www.lrz.de) for generous allocation of
computational resources.
ppm here and -55.0 ppm in ref. 2).[2a] In addition, the predicted 31
P
chemical shift for acyclic intermediate 7Lab of +29.3 ppm is that
for a "typical" phosphonium species. In case of 15N NMR
spectroscopy, the predicted signals vary more from the measured
ones, probably because of the partially anionic character of the
nitrogen atoms in the intermediates that was not reflected in our
set of reference compounds. Still, 7Nab is the only intermediate
whose predicted NMR shifts fit for all three measured nuclei in
complete agreement with all other results.
Keywords: Computational chemistry, isotopic labelling, Lewis
base organocatalysis, NMR spectroscopy, reaction mechanisms
[1]
[2]
a) C. Six, F. Richter in Isocyanates, Organic, Ullman's Encyclopedia of
Industrial Chemistry, Wiley-VCH, 2003; b) F. Richter, Nachr. Chem. 2007,
55, 380 – 384.
a) Z. Pusztai, G. Vlad, A. Bodor, I. T. Horvath, H. J. Laas, R. Halpaap, F.
U. Richter, Angew. Chem. Int. Ed. 2006, 45, 107 – 110; Angew. Chem.
2006, 118, 113 – 116. b) H. J. Laas, R. Halpaap, J. Pedain, J. Prakt.
Chem. 1994, 336, 185 – 200.
[3]
a) S. M. Raders, J. G. Verkade, J. Org. Chem. 2010, 75, 5308 – 5311; b)
H. A. Duong, M. J. Cross, J. Louie, Org. Lett. 2004, 25, 4679 – 4681.
U. Tracht, F. U. Richter, Macromol. Symp. 2013, 324, 33 – 40.
J. N. Gibb, J. M. Goodman, Org. Biomol. Chem. 2013, 11, 90 – 97.
a) B. E. Mann, J. Chem. Soc., Perkin Trans. 2 1972, 1, 30 – 34; b) B.
Maryasin, H. Zipse, Phys. Chem. Chem. Phys. 2011, 13, 5150 – 5158,
c) M. M. Kayser, K. L. Hatt, D. L. Hooper, Can. J. Chem. 1991, 69, 1929
– 1939.
[4]
[5]
[6]
[7]
[8]
J. S. Tang, J. G. Verkade, Angew. Chem. Int. Ed. 1993, 32, 896 – 898;
Angew. Chem. 1993, 105, 934 – 936.
a) F. U. Richter, Chem. Eur. J. 2009, 15, 5200 – 5202; b) D. Heift, Z.
Benko, H. Grützmacher, A. R. Jupp, J. M. Goicoechea, Chem. Sci. 2015,
6, 4017 – 4024.
[9]
a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648 – 5652; b) S. Grimme, J.
Chem. Phys. 2006, 124, 034108-1; c) A. V. Marenich, C. J. Cramer, D.
G. Trular, J. Phys. Chem. B 2009 113, 6378 – 6396.
[10] a) G. Süss-Fink, G. Herrmann, G. F. Schmidt, Polyhedron 1988, 7, 2341
– 2344; b) G. Herrmann, G. Süss-Fink, Chem. Ber. 1985, 118, 3959 –
3965; c) G. Süss-Fink, G. Herrmann, U. Thewalt, Angew. Chem. Int. Ed.
1983, 22, 880 – 881; Angew. Chem. 1983, 95, 899 – 900.
Figure 5. Comparison of predicted and experimentally obtained 31P, 15N and
13C NMR shifts.
[11] a) H. Sardon, A. Pascual, D. Mecerreyes, D. Taton, H. Cramail, J. L.
Hedrick, Macromolecules 2015, 48, 3153 – 3165; b) J. Alsarraf, Y. A.
This article is protected by copyright. All rights reserved.

10.1002/chem.201804016
Chemistry - A European Journal
COMMUNICATION
Ammar, F. Robert, E. Cloutet, H. Cramail, Y. Landais, Macromolecules
2012, 45, 2249 − 2256.
This article is protected by copyright. All rights reserved.

10.1002/chem.201804016
Chemistry - A European Journal
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 2:
COMMUNICATION
Julian Helberg, Yohei OE, Hendrik
Zipse*
Page No. – Page No.
Mechanistic Analysis and
Characterization of Intermediates in
the Phosphane-Catalyzed
Oligomerization of Isocyanates
Text for Table of Contents
This article is protected by copyright. All rights reserved.