This document is the accepted manuscript version of a published work that
appeared in final form in Chem. Commun., 2019, 55, 2380-2383. DOI:
10.1039/c8cc08884e.
Halogen Bonding Effects on the Outcome of Reactions at Metal
Centres‡†
Lucas Carreras,a Jordi Benet-Buchholz,a Antonio Franconetti,b Antonio Frontera,b Piet W. N. M. van
Leeuwenc and Anton Vidal-Ferran*a,d

Key findings regarding the effects of ligand preorganisation via
halogen bonding on the outcome of reactions at rhodium are
reported. An unprecedented halogen bonding-mediated oxidative
addition of CArI bonds to rhodium with efficient formation of
cyclometallated species deserves special mention.
Non-covalent interactions are at the core of supramolecular
chemistry.1,2 Whilst interactions such as hydrogen bonding,2b
ionic,2c-g metal-ligand,2h-j anion-π3 or cation-π4 interactions
have been widely exploited, non-conventional interactions
such as halogen,5 chalcogen,6 pnictogen7 or tetrel bonding 8
have been less extensively studied. 9 Of the latter interactions,
halogen bonding has recently gained importance due to its
inherent supramolecular directionality and strength that is
independent of solvent polarity.5 In contrast to the numerous
reports of halogen bonding applied to crystal engineering or
functional materials,10 its application to solution chemistry has
progressed at a much slower pace and mainly involves
molecular recognition events,11 organocatalysis11 and, to a
lesser extent, transition metal based catalysis.12,13 Our group
has pioneered the use of halogen bonding as the driving force
to construct the backbone of rhodium chelates.12 As indicated
in Scheme 1, our approach relies on the halogen-bondmediated assembly of a metal centre and two suitably
engineered building blocks (each of which incorporates an
appropriate halogen-bond donor and acceptor group,
respectively, together with a P-based ligating group for the
metal centre). We successfully used the resulting rhodium
complex XBphos-Rh as an efficient catalyst for the
hydroboration of terminal alkynes.12 This breakthrough in the

use of halogen bonding for easily constructing metal catalysts
prompted us to further explore the use of this interaction for
influencing the coordination sphere at the metal centre. By
expanding the structural diversity of our approach, we aimed
to gain new insights into the reactivity of these complexes and
induce transformations at the metal that would not take place
in the absence of halogen bonding. Herein, we report our work
on the formation of five- and six- coordinate rhodium
complexes, in which halogen bonding drives ligand
preorganisation and determines the structure of the final
products.
The underlying strategy for the preparation of XBphos-Rh,
relied on the in situ preparation of a putative [Rh(CO)2]X
intermediate from [Rh(Cl)(CO)2]2 and a halide scavenger (i.e.
NaBArF). Although this strategy proved to be useful for the
preparation of four-coordinate rhodium complexes such as
XBphos-Rh, herein we broaden our approach to include a
wider array of rhodium precursors.
We began by studying [Rh(nbd)2]BF4 as the rhodium precursor
in the complexation reaction, since phosphorus-containing
complexes derived from this cationic metal precursor have
found wide application in pivotal organic transformations 14
such as (enantioselective) hydrogenations. 15 When an
equimolar mixture of ligands 1 and 2 were added to
stoichiometric amounts of [Rh(nbd) 2]BF 4, heterocomplex
[Rh(nbd)(1)(2)]BF4 3 was isolated in 90% yield (Scheme 2, top).
X-ray analysis confirmed the structure of the rhodium complex
derived from 1 and 2 (Scheme 2, bottom) as a 5-coordinate
square-pyramidal complex. In addition to a coordinated
n o r b o r n a d i e n e u n i t , X - r a y a n a l ys i s a l s o r e ve a l e d

aInstitute

of Chemical Research of Catalonia (ICIQ) & The Barcelona Institute of
Science and Technology, Av. Països Catalans 16, 43007 Tarragona, Spain.
bDepartament de Química, Universitat de les Illes Balears (UIB), Ctra. de
Valldemossa km 7.5, 07122 Palma de Mallorca, Spain.
cLPCNO, INSA-Toulouse, 135 Avenue de Rangueil, F-31077 Toulouse, France.
dICREA, Pg. Lluís Companys 23, 08010 Barcelona, Spain.
E-mail: avidal@iciq.cat
‡Dedicated to Prof. Pablo Espinet on the occasion of his 70 th birthday.
† Electronic Supplementary Information (ESI) available: Experimental and
theoretical details, spectroscopic and crystallographic data (CCDC 18745071874513). For ESI and crystallographic data in CIF or other electronic format, see
DOI: 10.1039/c8cc08884e.

Scheme 1. Supramolecular halogen-bonded XBphos-Rh catalyst.

|1

Scheme 3. Top: reaction of ligands 1 and 2 with [Rh(acac)(CO)2]. Bottom: reaction of
ligand 1 with complex 5.

Scheme 2. Top: Preparation of [Rh(nbd)(1)(2)]BF4 complex 3. Bottom: X-Ray structure
of 3 with hydrogen atoms and the BF4- unit having been omitted for the sake of clarity.
Colour scheme: C: black, P: orange, Rh: green, F: green, N: blue, I: purple. Atomic
displacement ellipsoids are drawn at a 50% probability.

phosphines 1 and 2 coordinated in a relative cis-arrangement
as well as a I–Rh interaction in the apical position (2.7726(5)
Å). This arrangement of the P-ligands containing the halogenbond donor and acceptor motifs in the slightly distorted
square-pyramidal environment (P–Rh–P angle = 95.31(5)°)
does not allow the required linear N···I–C alignment for
effective halogen bonding (N···I–C angle = 59.1(2)°). However,
preorganisation of ligands 1 and 2 via halogen bonding
interactions (i.e. in situ formation of supramolecular
bisphosphine 1·2) before coordination to the rhodium centre
accounts for the high selectivity obtained in the formation of
heterocomplex [Rh(nbd)(1)(2)]BF4 3 (90% yield), with the
corresponding homocomplexes not being detected.16 These
observations are in agreement with the results of DOSY
experiments of ligands 1, 2 and a 1:1 mixture of 1 and 2, in
which the value of the diffusion coefficient of the ligand
mixture is lower than the same values for free ligand 1 or 2. A
lower diffusion coefficient is associated to a larger molecular
size, thus pointing to a favourable halogen-bond-mediated
assembly of the two ligands in solution (see ESI†).
We then turned our attention to the use of [Rh(acac)(CO)2] as
the rhodium precursor. The combination of this derivative and
P-ligands has been widely employed to generate active
hydroformylation catalysts.17 In this case, when an equimolar
mixture of ligands 1 and 2 was added to stoichiometric
amounts of [Rh(acac)(CO)2], 31P{1H} NMR analysis showed that
complex [Rh(acac)(CO)(1)] 4 was formed, with iodosubstituted phosphine 2 remaining uncoordinated (Scheme 3,
top). Therefore, we envisaged that a stepwise coordination
protocol using first ligand 2 followed by phosphine 1 could
lead to the target complex [Rh(acac)(1)(2)]. The reaction
between [Rh(acac)(CO)2 ] and ligand 2 efficiently led to
complex [Rh(acac)(CO)(2)] 5, which was successfully isolated.
Unfortunately, the reaction of 5 with equimolar amounts of
the ligand 1 led to complex 4 (Scheme 3, bottom) by ligand
displacement. The halogen bonding interaction between
ligands 1 and 2 does not appear to be strong enough to
displace the acetylacetonate ligand. Furthermore, the lower

2|

donicity of the P-electron pair in 2 (see ESI† for the
determination of the Tolman electronic parameter 18 and the
percent buried volume19) compared to 1 ensured that 2
remained uncoordinated.
We then moved to the [Rh(acac)(C2H4)2] precursor,20 with the
reasoning that the ethylene ligands would be more labile and
thus more readily displaced by the supramolecular phosphine
assembly compared to carbonyl ligands. Interestingly, the
reaction of [Rh(acac)(C2H4)2] with phosphines 1 and 2 led to
complex 6 that incorporated both phosphines and an
acetylacetonate unit. A closer inspection of the NMR spectral
data pointed to the formation of a Rh(III) complex in which a
C–I oxidative addition process had taken place (Scheme 4, left).
Single crystals of complex 6 suitable for X-ray analysis were
obtained and the structure of the Rh(III) complex was
confirmed (Fig. 1a). Indeed, it is reported in the literature that
ortho-haloarylphosphines such as 2 undergo oxidative addition
(OA) processes with rhodium (and other transition metals) at
high temperatures or under UV/Vis irradiation.21 According to
these literature reports, the mild conditions employed in the
complexation reaction of [Rh(acac)(C2H4)2] with 1 and 2 should
not favour the CAr–I oxidative addition process.22 However, as
we observed OA even at 25 °C, we wondered whether ligand
preorganisation around the rhodium centre due to halogen
bonding between 1 and 2 could be facilitating the OA process.
It has been demonstrated for ligands 1 and 2 that the
geometry of the resulting complexes depends on the nature of
the precursor. Moreover, the iodine atom is coordinated to
the metal centre in both XBphos-Rh and 3. In the case of sixcoordinate complex 6, we reasoned that a transient N···I–C
alignment via halogen b onding with simultaneo us
coordination of the I-atom to the rhodium centre could be
taking place, ultimately favouring the oxidative addition of the
C–I bond to the electron-rich rhodium centre. To support this
statement, we decided to study the interplay between
[Rh(acac)(C2H4)2], ligand 2 and 3-pyridyldiphenyl-phosphine 7,
where alignment of the N···I–C atoms with simultaneous
coordination of the P-ligating groups is not possible. Therefore,
oxidative addition using ligand 7 should be disfavoured with
respect to the same process employing halogen-bond
complementary ligand 1. Unlike in the case of the reaction of 1
with 2, where chemoselective formation of the oxidative
addition Rh(III) complex 6 was observed, the reaction of 7 and

Scheme 5. Preparation of complex 9.

Scheme 4. Reactivity of 2-iodo-3,4,5,6-tetrafluorophenyldiphenylphosphine 2 and
[Rh(acac)(C2H4)2] with 2-pyridyldiphenylphosphine 1 and 3-pyridyldiphenylphosphine 7.

2 led to homocomplex 8 (Scheme 4, right) with no traces of the
OA product being detected by NMR analysis. To further
demonstrate the influence of halogen bonding in this
transformation, we monitored both complexation reactions by
31P NMR spectroscopy. It is interesting to note that, whilst no
heterocomplex for 7 and 2 was observed by 31P NMR analysis,
this technique clearly demonstrated the formation of rhodium
species containing the halogen-bonded 1·2 assembly (see ESI†
for details).
Theoretical calculations were also carried out in order to
understand the selective formation of complex 6 (see ESI† for
details). The examination of the reaction path reveals the
formation an intermediate displaying a short and directional
halogen bonding contact N···I‒C (see Fig. 2a). In fact, the
molecular electronic potential (MEP) surface of ligand 2
reveals the existence of a strong σ-hole (ca. 30 kcal·mol1, see
Fig. SI 50) on the extension of the C–I bond. This halogen
bonding interaction enlarges the C–I distance from 2.11 Å in
the ligand to 2.14 Å in the intermediate thus facilitating the
subsequent oxidative addition. Interestingly, the calculation of
the equivalent intermediate using phosphine 7 (no halogenbond is possible, see Fig. SI 51) is 6.9 kcal·mol1 less stable,
thus explaining the absence of oxidative addition reaction (see
Scheme 4). The TS structure is characterised by a lengthening
of the CAr‒I bond (from 2.14 to 2.20 Å) as well as a shortening
of the C‒Rh distance (1.06 Å). The N···IC halogen- bond is less
directional and longer in the TS, but the interaction still exists
as evidenced by the non-covalent interaction plot (NCIplot, see
Fig. SI 52). Overall, these results illustrate that the combined
use of an electron-rich metal precursor such as
[Rh(acac)(C2H4)2] together with complementary ligands 1 and 2

Fig. 1. Crystal structures of 6 (a) and 9 (b). Hydrogen atoms have been omitted
for the sake of clarity. For the colour scheme, see caption to Scheme 4.

efficiently leads to a new class of cyclometallated rhodium(III)
complexes.23
Analogous behaviour was observed for the reaction between 1
and 2 and [Rh(Cl)(CO)2]2 in the absence of a halide scavenger.12
In this case, the CO ligands were preferentially displaced by
phosphines 1 and 2. The complementarity of the halogenbonded complex 1·2 for simultaneous complexation of the two
phosphine groups to the rhodium centre and the absence of
strong pi-backbonding carbonyl ligands facilitated the
formation of complex 9 arising from coordination of the Patoms and a CAr–I oxidative addition process. The structure of
the final complex was determined by standard spectroscopic
techniques and confirmed by X-ray analysis (Scheme 5 and Fig.
1b). Interestingly, the pyridine group is coordinated to the
rhodium centre with the two halogen ligands placed axially
and the two phosphino groups coordinated in a transarrangement.24
In conclusion, these results illustrate how the careful selection
of the rhodium precursor and the geometry of the halogenbond donor and acceptor ligands determines the outcome of
the Rh–P complexation chemistry, either to P–I–P pincer-like
Rh(I) complexes with a permanent halogen bonding
interaction (XBphos-Rh),12 or to the P–P Rh(III) complexes
described in this work that arise from an unprecedented
halogen-bond-driven CAr–I oxidative addition process. It has
been both experimentally and computationally demonstrated
that the N···I-halogen-bond interaction and the electron
density at the metal centre are key factors that control the
reactivity towards oxidative addition processes as observed in

Fig. 2. M06-2X/LANL2DZ optimised structures of the intermediate and transition
state to yield 6. Distances in Å. Halogen bonds are represented by green dashes.

|3

6 and 9. Selectivity in the primary sphere coordination
chemistry at rhodium has also been controlled by employing
building blocks 1 and 2. Ligand preorganisation by in situ
formation of supramolecular halogen-bonded complex 1·2
appears to translate to a selective formation of complex 3.
Further work is currently underway to study the application23
and catalytic activity of these new complexes, to ultimately set
the basis for the use of halogen bonding in transition metal
chemistry and catalysis.
The authors would like to thank MINECO (CTQ2014-60256-P,
CTQ2017-89814-P, CTQ2017-85821-R, and Severo Ochoa
Excellence Accreditation 2014–2018 SEV-2013-0319) and the
ICIQ Foundation for the financial support. L. C. thanks MINECO
for a FPI-SO predoctoral fellowship (BES-2015-071872). The
Université Fédérale Toulouse et Midi-Pyrénées is thanked for
an IDEX Chaire d'attractivité. A.F. thanks the MINECO/AEI from
Spain for a “Juan de la Cierva” contract. Drs. E. Martin and S. J.
Prathapa are acknowledged for X-ray crystallographic data
acquisition. We also thank Ms R. Somerville for proof reading
the manuscript.

8
9
10
11
12
13

14

15
16
17

Conflicts of interest

18

There are no conflicts to declare.

19

Notes and references
1
2

3
4
5
6
7

4|

J. M. Lehn, Science, 1993, 260, 1762-1763.
For example, for a review on the use of these interactions in
supramolecular catalysis, see: (a) M. Raynal, P. Ballester, A.
Vidal-Ferran and P. W. N. M. van Leeuwen, Chem. Soc. Rev.,
2014, 43, 1660-1733. For selected examples from our group
exploiting non-covalent interactions, see: Hydrogen bonding:
(b) A. Desmarchelier, X. Caumes, M. Raynal, A. Vidal-Ferran,
P. W. N. M. van Leeuwen and L. Bouteiller, J. Am. Chem. Soc.,
2016, 138, 4908-4916. Ionic interactions: (c) I. Mon, D. A.
Jose and A. Vidal-Ferran, Chem. - Eur. J., 2013, 19, 27202725. (d) H. Fernández-Pérez, I. Mon, A. Frontera and A.
Vidal-Ferran, Tetrahedron, 2015, 71, 4490-4494. (e) A. VidalFerran, I. Mon, A. Bauzá, A. Frontera and L. Rovira, Chem. Eur. J., 2015, 21, 11417-11426. (f) L. Rovira, M. Vaquero and
A. Vidal-Ferran, J. Org. Chem., 2015, 80, 10397-10403. (g) L.
Rovira, H. Fernández-Pérez and A. Vidal-Ferran,
Organometallics,
2016, 35, 528-533.
Metal-ligand
interactions: (h) J. Etxebarria, A. Vidal-Ferran and P.
Ballester, Chem. Commun., 2008, 5939-5941. (i) H.
Degenbeck, A.-S. Felten, J. Etxebarria, E. C. Escudero-Adán, J.
Benet-Buchholz and A. Vidal-Ferran, Cryst. Growth Des.,
2012, 12, 2719-2723. (j) H. Degenbeck, A.-S. Felten, E. C.
Escudero-Adán, J. Benet-Buchholz, L. Di Bari, G. Pescitelli and
A. Vidal-Ferran, Inorg. Chem., 2012, 51, 8643-8645.
D. Quiñonero, C. Garau, C. Rotger, A. Frontera, P. Ballester,
A. Costa and P. M. Deyà, Angew. Chem., Int. Ed., 2002, 41,
3389-3392.
D. A. Dougherty, Acc. Chem. Res., 2013, 46, 885-893.
G. Cavallo, P. Metrangolo, R. Milani, T. Pilati, A. Priimagi, G.
Resnati and G. Terraneo, Chem. Rev., 2016, 116, 2478-2601.
W. Wang, B. Ji and Y. Zhang, J. Phys. Chem. A, 2009, 113,
8132-8135.
G. Sánchez-Sanz, C. Trujillo, M. Solimannejad, I. Alkorta and
J. Elguero, Phys. Chem. Chem. Phys., 2013, 15, 14310-14318.

20
21

22
23

24

A. Bauzá, T. J. Mooibroek and A. Frontera, Angew. Chem.,
Int. Ed., 2013, 52, 12317-12321.
Y. Zhao, Y. Cotelle, N. Sakai and S. Matile, J. Am. Chem. Soc.,
2016, 138, 4270-4277.
A. Priimagi, G. Cavallo, P. Metrangolo and G. Resnati, Acc.
Chem. Res., 2013, 46, 2686-2695.
R. Tepper and U. S. Schubert, Angew. Chem., Int. Ed., 2018,
57, 6004-6016.
L. Carreras, M. Serrano-Torné, P. W. N. M. van Leeuwen and
A. Vidal-Ferran, Chem. Sci., 2018, 9, 3644-3648.
For examples in the use of halogen bonding applied to
transition metal catalysis, see ref. 12 and: (a) V. N. G.
Lindsay, W. Lin and A. B. Charette, J. Am. Chem. Soc., 2009,
131, 16383-16385. (b) N. Q. Shixaliyev, A. V. Gurbanov, A. M.
Maharramov, K. T. Mahmudov, M. N. Kopylovich, L. M. D. R.
S. Martins, V. M. Muzalevskiy, V. G. Nenajdenko and A. J. L.
Pombeiro, New J. Chem., 2014, 38, 4807-4815.
For selected examples, see: (a) A. Togni, C. Breutel, A.
Schnyder, F. Spindler, H. Landert and A. Tijani, J. Am. Chem.
Soc., 1994, 116, 4062-4066. (b) Y. Yamanoi and T. Imamoto,
J. Org. Chem., 1999, 64, 2988-2989. (c) T. Imamoto, K. Sugita
and K. Yoshida, J. Am. Chem. Soc., 2005, 127, 11934-11935.
J. M. Brown, Organometallics, 2014, 33, 5912-5923.
Homocomplexes [Rh(nbd)(1)2]BF4 or [Rh(nbd)(2)2]BF4 were
not observed during the complexation reaction. See ESI†.
P. W. N. M. Van Leeuwen and C. Claver, Eds., Rhodium
Catalyzed Hydroformylation, Kluwer, 2000.
A. Roodt, S. Otto and G. Steyl, Coord. Chem. Rev., 2003, 245,
121-137.
Percent buried volume (%V Bur) was used as steric descriptor,
see: H. Clavier and S. P. Nolan, Chem. Commun., 2010, 46,
841-861. SambVca 2.0 web application was used to calculate
%VBur, see: L. Falivene, R. Credendino, A. Poater, A. Petta, L.
Serra, R. Oliva, V. Scarano and L. Cavallo, Organometallics,
2016, 35, 2286-2293, and ESI† for details.
The reactivity observed for precursor [Rh(acac)(C 2H4)2] was
identical to those observed for related precursor
[Rh(acac)(cod)]. For details see ESI†.
For a selected example see: (a) J. C. Besteiro, P. Lahuerta, M.
Sanaú, I. Solana, F. A. Cotton, R. Llusar and W. Schwotzer,
Polyhedron, 1988, 7, 87-96, and references cited therein. For
a review on ortho-metallated transition metal complexes
derived from tertiary phosphines see: (b) F. Mohr, S. H.
Privér, S. K. Bhargava and M. A. Bennett, Coord. Chem. Rev.,
2006, 250, 1851-1888.
The preparation of complex 6 has been reported under
oxidative addition conditions. See ref. 12.
Cyclometallated Rh(III) complexes have been reported as
therapeutic agents in medicinal chemistry, see: (a) Y.
Geldmacher, M. Oleszak and W. S. Sheldrick, Inorg. Chim.
Acta, 2012, 393, 84-102. (b) C.-H. Leung, S. Lin, H.-J. Zhong
and D.-L. Ma, Chem. Sci., 2015, 6, 871-884.
The complexation of phosphines 7 and 2 using [Rh(Cl)(CO)2]2
as the Rh(I) source led to a mixture of hetero- and
homocomplexes, but no OA products were observed.