The role of charge in 1,2,3-triazol(ium)-based halogen bonding activators

Article abstract of DOI:10.1039/c8cc00527c

The influence of charge on the performance of monocationic and dicationic triazol(ium)-based halogen bond donors was investigated. Next to the activity in a halide abstraction benchmark reaction, halogen bonding was also evaluated via X-ray structural analyses and isothermal titration calorimetry.

Full text of DOI:10.1039/c8cc00527c

ChemComm
View Article Online
View Journal
COMMUNICATION
The role of charge in 1,2,3-triazol(ium)-based
halogen bonding activators†
Cite this:DOI: 10.1039/c8cc00527c
a
a
a
b
Alexander Dreger, Elric Engelage, Bert Mallick, Paul D. Beer and
a
Received 21st January 2018,
Accepted 16th February 2018
Stefan M. Huber
*
DOI: 10.1039/c8cc00527c
rsc.li/chemcomm
The influence of charge on the performance of monocationic and
dicationic triazol(ium)-based halogen bond donors was investigated.
Next to the activity in a halide abstraction benchmark reaction,
halogen bonding was also evaluated via X-ray structural analyses
and isothermal titration calorimetry.
Fig. 1 Previously used dicationic halogen bonding based activator (left),
Halogen bonds (XBs) are highly directional non-covalent inter-
actions between the electrophilic region of a covalently bound
halogen atom and a Lewis base. For strong halogen bonding,
monocationic (middle) and neutral variant of this study (right). Z = triflate
F
or BAr
4
.
1
the respective XB donor (i.e., the halogen-based Lewis acid)
requires an electron-withdrawing core structure, and often
In 2015, Beer et al. introduced strong bidentate (but mono-
perfluorinated or cationic (hetero)arenes are used. Until now, cationic) halogen bond donors based on a bis(iodotriazolyl)-
0
R,R
halogen bonding has been successfully applied in many areas, pyridinium motif (3,5-Py(X) /Z, Fig. 1 middle) for anion
1
3
including crystal engineering, supramolecular chemistry and recognition. Since these compounds represent an interesting
2
–4
organocatalysis.
compromise between overall charge assistance but neutral
Halogen bond donors based on triazolium moieties were first halogen-bond-donating moieties, we became interested to test their
5
introduced in the context of an anion-binding rotaxane. Bidentate performance as Lewis acidic activators, especially in comparison to
Oct,Me
6
triazolium-based activators were then used in a benchmark reaction our previous dicationic systems (1,3-Ph(I)
/OTf, Fig. 1 left).
for halide abstraction (the solvolysis of benzhydryl bromide 1 in In doing so, our main goal was to elucidate the relevance of
6
14
acetonitrile), and their activating power was found to be similar to charge on the actual halogen bonding groups.
7
8
that of bidentate imidazolium or pyridinium-based halogen bond
The syntheses of all XB donors employed in this study are
donors. Subsequently, neutral bidentate triazole derivatives were shown in Scheme 1. First, a modified Cu(I)-catalysed azide–alkyne
9,10
investigated as anion receptors,
but their binding strength to 1,3-dipolar cycloaddition reaction of bis-alkyne 4 with benzyl azide
chloride was found to be markedly lower than that of their dicationic or octyl azide and sodium iodide led to the corresponding
R
analogues. This is in line with previous findings that charge-assisted bis(iodotriazolyl)pyridine derivatives 3,5-Py(I) in good yields
11
15
XB donors are stronger Lewis acids and are thus preferentially (Scheme 1; for nomenclature see Fig. 1). Subsequent regio-
used in organic synthesis. On the other hand, halogen bond donors selective alkylation at the pyridine core with methyl triflate
based on cationic core structures also feature some intrinsic dis- (MeOTf), octyl triflate (OctOTf) or N-methyl bis(trifluoromethyl-
0
R,R
advantages, e.g. low solubility in apolar solvents and competition by sulfonyl)amide (MeNTf
2
) provided XB donors 3,5-Py(I) /Z
12
16
Lewis basic counterions.
(Z = OTf, NTf
2
) in good yields (71–98%). In order to include
a variant with noncoordinating tetrakis(3,5-bis(trifluoromethyl)-
F
Oct,Me
F
a
b
phenyl)borate (BAr
), the corresponding salt 3,5-Py(I)
/BAr
4
Department of Chemistry and Biochemistry, Ruhr-Universit a¨ t Bochum,
Universit a¨ tsstraße 150, 44801 Bochum, Germany. E-mail: stefan.m.huber@rub.de
Department of Chemistry, University of Oxford, Mansfield Road, Oxford,
OX1 3TA, UK
4
was obtained from salt metathesis of the triflate salt with NMe
BAr in chloroform in quantitative yield. Non-iodinated reference
compounds were obtained via similar procedures (see the ESI†).
The activity of these XB donors was then tested in the bench-
mark reaction shown in Scheme 2. Similarly to our previous
studies, the XB donor activates the C–Br bond of benzhydryl
4
F
4
16
†
Electronic supplementary information (ESI) available: Syntheses as well as
kinetic, titration and crystallographic data. CCDC 1817655 and 1817656. For
ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/
c8cc00527c
17
This journal is ©The Royal Society of Chemistry 2018
Chem. Commun.

View Article Online
Communication
ChemComm
Scheme 1 Synthesis of XB donors: for X = I (ii) octyl azide or benzyl azide (2.2 eq.), CuOTf
2
(4 eq.) or Cu(ClO
)
4 2
(H
2
O)
6
(4 eq.), NaI (8 eq.), TBTA (2 mol%),
DBU (2 eq.), MeCN : THF (1 : 1), r.t., 1–3 d. (iii) MeNTf
2
(1.5 eq.), toluene : EtOAc (1 : 1), microwave, 130 1C, 1 h. (iv) MeOTf or OctOTf (1.2–1.5 eq.), CH
2
Cl
2
,
F
ꢀ
78 1C, 1–3 d. (v) 1.3 eq. TMA–BAr
4
, CHCl
3
, 12 h. For X = H: (ii) octyl azide or benzyl azide (2 eq.), CuSO
Cl
, ꢀ78 1C, 1 d.
4
(H O) (0.05 eq.), Na ascorbate (0.15 eq.),
2 5
CH Cl : H O (1 : 1), 5 d. (iv) MeOTf or OctOTf (1.2–1.5 eq.), CH
2
2
2
2
2
Scheme 2 Friedel–Crafts alkylation as the benchmark reaction for the
activation of the carbon–bromine bond by charge-assisted halogen bond
donors. (i) 1 eq. activator (XB donor or reference compound), 1 eq. Cs
CDCl (1.6 mM), rt.
2 3
CO ,
3
bromide 1 to generate the benzhydryl cation. Deviating from
our earlier benchmark reaction, which featured the solvolysis of
the intermediate, in this study the cation was quenched with
one equivalent of 1,3,5-trimethoxybenzene 2 as an external
nucleophile. A stoichiometric amount of the respective XB
donor was used and Cs
2
CO
3
was added to each reaction to rule
out hidden acid catalysis.
1
The reaction was monitored by H-NMR spectroscopy (see
Fig. 2 for kinetic profiles and Table 1 for yields of product 3).
No significant background reaction occurs in the absence of
activators (o5% yield after 30 h). Activation by Brønsted acids
can also be ruled out, since 5 mol% of triflic acid (HOTf)
showed no activation. As a non-iodinated reference compound
Oct,Me
3,5-Py(H)
/OTf induced no product formation, any activity
0
R,R
of iodinated 3,5-Py(I) /Z compounds cannot be explained by
hydrogen bonding or by simple electrostatic interactions and is
thus likely based on halogen bonding.
Oct
The non-cationic bis(triazole) derivative 1,3-Ph(I) (cf. Fig. 1)
showed no activity, indicating that some form of charge assis- Fig. 2 Yield vs. time diagram for the formation of compound 3 in the
1
benchmark reaction. The yield of compound 3 according to H-NMR
tance is essential for this reaction. The highest rate acceleration
Oct,Me
F
analysis. The error is approximately 5%.
was obtained in the presence of 3,5-Py(I)
/BAr
2% of compound 3 after 2.5 h (Table 1, entry 5). However, after
h decomposition of the XB donor was observed (reaching (42% for 3,5-Py(I)
4
, yielding
7
3
Oct,Me
Oct,Me
/OTf after 5 h and 53% for 3,5-Py(I)
/
B13% of the decomposed material after 11 h), and thus the NTf after the same time; see Table 1, entries 6 and 7). Variation of
2
interpretation of the observed performance needs some caution. the alkylation pattern did not decisively alter the overall outcome:
In contrast, no indications of decomposition were obtained for changing the methyl group on the pyridinium to octyl led to a
ꢀ
the analogous triflate and NTf
2
salts, but – as could be expected slight increase in yield, which is likely still within the error of
Oct,Oct
from the more coordinating nature of the counterions – these experiment (48% yield for 3,5-Py(I)
/OTf after 5 h, entry 8).
activators also induced somewhat less product formation Substituting the octyl groups on the triazolium moieties with
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2018

View Article Online
ChemComm
Communication
Table 1 Yields of product 3 in the reaction of Scheme 2 after approx. 5 h
and 12 h in the presence of different activation reagents
Table 2 Results of isothermal titration calorimetry experiments of various
mono- and di-cationic XB donors with halides in chloroform at 30 1C
a
b
a
b
a
ꢀ1
DH [kJ molꢀ1] ꢀTDS [kJ molꢀ1
#
Activating reagent
—
Yield [%] (5 h)
Yield [%] (12 h)
#
Host
X
K [M
]
]
Oct,Me
Oct,Me
Oct,Me
Oct,Me
Oct,Me
Oct,Me
Oct,Me
Oct,Me
5
1
2
3
4
5
6
7
8
9
1
1
1
1
o5
o5
o5
o5
o5
o5
o5
o5
1
2
3
4
5
6
7
8
9
3,5-Py(I)
3,5-Py(I)
3,5-Py(I)
3,5-Py(I)
3,5-Py(I)
3,5-Py(I)
3,5-Py(I)
3,5-Py(I)
1,3-Ph(I)
/OTf Cl 1.35 ꢁ 10 ꢀ19.34
ꢀ10.42
ꢀ8.67
ꢀ7.05
ꢀ5.94
ꢀ6.74
ꢀ4.42
ꢀ7.67
ꢀ3.06
ꢀ15.85
ꢀ15.81
ꢀ12.81
—
5
HOTf
3,5-Py(H)
1,3-Ph(I)
/OTf Br 1.59 ꢁ 10 ꢀ21.50
Oct,Me
4
/OTf
/OTf
/NTf
/NTf
/NTf
/BAr
/BAr
I
9.26 ꢁ 10 ꢀ21.77
Oct
5
2
2
2
Cl 2.53 ꢁ 10 ꢀ25.41
Oct,Me
F
b
b
5
3,5-Py(I)
3,5-Py(I)
3,5-Py(I)
3,5-Py(I)
3,5-Py(I)
3,5-Py(I)
3,5-Py(I)
1,3-Ph(I)
1,3-Ph(I)
/BAr
4
81
87
Br 2.59 ꢁ 10 ꢀ24.66
Oct,Me
Oct,Me
5
/OTf
/NTf
/OTf
42
53
48
44
37
56
69
63
61
52
I
1.53 ꢁ 10 ꢀ25.66
F
F
5
2
4
Cl 3.16 ꢁ 10 ꢀ24.25
Oct,Oct
Bn,Oct
5
4
Br 3.57 ꢁ 10 ꢀ29.15
Oct,Me
Oct,Me
Oct,Me
Oct,Me
Oct,Me
4
/OTf
/OTf Cl 9.09 ꢁ 10 ꢀ12.45
Bn,Me
Bn,Me
5
0
1
2
3
/NTf
/OTf
2
10 1,3-Ph(I)
11 1,3-Ph(I)
12 1,3-Ph(I)
13 1,3-Ph(I)
/OTf Br 1.60 ꢁ 10 ꢀ14.36
c
c
4
o5
10
/OTf
/BAr
/BAr
I
1.82 ꢁ 10 ꢀ17.68
Oct,Me
Oct,Me
F
b
/OTf
/BAr
54
35
68
38
4
Cl —
Br —
—
—
F
c
c
F
b
4
4
—
a
1
b
a
Yields according to H-NMR analysis. Partial decomposition of the
Tetraoctylammonium salts were used as guests; c(guest) = 10 mM;
c(host) = 1 mM. The guest solution was titrated into the host solution.
No accurate fitting data could be obtained.
c
3
XB donor. Low solubility in CDCl .
b
benzyl groups again led to virtually identical results for the
F ꢀ
ꢀ
ꢀ
Bn,Oct
counterions (BAr 4 4 NTf2 4 OTf , entries 1–8). This underlines
the relevance of ion pairing and may also explain the similar Lewis
acidity (and activation potential) of mono- and dicationic XB donors:
ion pairing in the latter is likely stronger, which might compensate
for their potentially stronger XB donor properties.
N –Oct compound 3,5-Py(I)
Py
/OTf (44% yield, entry 9). The
corresponding benzylated N –Me compounds showed slightly
Py
Bn,Me
(
37% yield for 3,5-Py(I)
2
/NTf ) to drastically lower perfor-
Bn,Me
mance (no yield for 3,5-Py(I)
/OTf). At least in the latter
case, this can be attributed to the low solubility of the XB donor
in chloroform.
Finally, X-ray structural analyses helped to shed light on
the binding mode of the monocationic XB donors to anions.
The result of the single-crystal X-ray analysis of the XB donor
6
Most importantly, the previously studied dicationic XB donor
Oct,Me
1
5
,3-Ph(I)
/OTf was only slightly more active (54% yield after
Bn,Me
Oct,Me
3,5-Py(I)
/OTf is shown in Fig. 3. Two strong halogen bonds
h) than its monocationic analogue 3,5-Py(I)
/OTf (42%,
1
8
between iodine atoms of the XB donor and oxygen atoms of
triflate can be found in the solid state. The iodine–oxygen bond
distances (2.78 Å and 2.99 Å) are shorter than the sum of the
van der Waals radii (3.5 Å) and the binding angle C–I–O is
nearly linear (1731 and 1661), as expected for halogen bonding.
see above). This clearly demonstrates that bidentate charge-
assisted halogen bonding does not automatically also necessitate
dicationic XB donors. A comparable performance as an activator
is also feasible for monocationic species. In the current example,
this is particularly noteworthy since the Lewis structure formalism
2
0
The previously reported crystal structure of the dicationic XB
(!) does not indicate the delocalization of the cationic charge
Bn,Me
donor 1,3-Ph(I)
/OTf featured halogen bonds to triflate
towards the triazole moieties.
The results of the activation studies are supported by isother-
mal titration calorimetry (ITC) measurements (see Table 2). Both
8
anions with I–O distances of 2.88 Å and 2.90 Å.
From a solution of 3,5-Py(I)
(1) in acetonitrile, we obtained a single-crystal of 3,5-Py(I)
Bn,Oct
/OTf and benzhydryl bromide
Bn,Oct
Oct,Me
/Br
the monocationic XB donor 3,5-Py(I)
/OTf and its dicationic
Oct,Me
(Fig. 4). The bromide is bound in a bidentate fashion by the XB
variant 1,3-Ph(I)
/OTf feature nearly the same binding strength
5
ꢀ1
to bromide in chloroform (1.6 ꢁ 10 M , entries 2 and 10). The
binding to chloride is also very similar (entries 1 and 9), whereas a
notable deviation is found for iodide complexation. Somewhat
Oct,Me
surprisingly, the monocationic system 3,5-Py(I)
/OTf turned
out to be the stronger Lewis acid. A clear difference between the
two types of XB donors is the relative contribution of the enthalpic
and entropic parts. For the monocationic XB donor, the enthalpic
Oct,Me
part is decidedly more favourable compared to 1,3-Ph(I)
/OTf.
On the other hand, the entropic contribution is negligible for the
former, but constitutes an important favourable contribution in
the latter (similar to the binding of bis(iodoimidazolium)-derived
19
XB donors).
1
3
As had already been observed, the association strength
decreases in the order chloride 4 bromide 4 iodide for all XB
donors tested. The influence of the counterion could only be
investigated for the monocationic species, as the dicationic
ones suffered from solubility issues. As expected, the binding
2
1
Fig. 3 ORTEP plot of the X-ray structural analysis of compound
Bn,Me
3,5-Py(I)
/OTf (ellipsoids at 50% probability). Selected bond distances [Å]
and angles [1]: I1–O3 = 2.780, I2–O2 = 2.987, I1–C8 = 2.056, I2–C17 = 2,072,
strength of the XB donors follows the lesser Lewis basicity of the C8–I1–O3 = 173.62, and C17–I2–O2 = 166.57.
This journal is ©The Royal Society of Chemistry 2018
Chem. Commun.

View Article Online
Communication
ChemComm
respective dicationic analogues: complete anion exchange by
metathesis is realized more easily, and the solubility in less
polar organic solvents is generally superior. This already
Oct,Me
F
became apparent in this study, as 3,5-Py(I)
/BAr exhibits
4
Oct,Me
F
much better solubility in chloroform than 1,3-Ph(I)
/BAr 4
Oct,Oct
(
see footnote 18). As a further example, 3,5-Py(I)
/OTf exhibits
Oct,Me
much better solubility in diethylether than 1,3-Ph(I)
/OTf. As
a consequence, the utilization of monocationic motifs like the one
studied here might allow to combine the ‘‘best of both worlds’’
in halogen-bonding organocatalysis: the performance of charge-
assisted XB donors and the solubility of (generally weaker)
fluorinated XB donors in less polar media.
This project has received funding from the European
Research Council (ERC) under the European Union’s Horizon
2020 research and innovation programme (Grant Agreement
No. 638337) and from the Fonds der Chemischen Industrie.
21
Fig. 4 ORTEP plot of the X-ray structural analysis of compound
Bn,Oct
3,5-Py(I)
/Br (ellipsoids at 50% probability). Selected bond distances [Å] Conflicts of interest
and angles [1]: I1–Br = 3.185, I2–Br = 3,227, I1–C22 = 2.075, I2–C32 = 2,091,
C22–I1–Br = 171.70, and C32–I2–Br = 166.51.
There are no conflicts to declare.
Notes and references
donor with I–Br distances of 3.19 Å and 3.23 Å. These distances are
again significantly shorter than the sum of the van der Waals radii
1
G. R. Desiraju, P. S. Ho, L. Kloo, A. C. Legon, R. Marquardt,
P. Metrangolo, P. Politzer, G. Resnati and K. Rissanen, Pure Appl.
Chem., 2013, 85, 1711.
20
(
1
3.83 Å) and the binding angle C–I–Br is close to linear (1711 and
Oct,Me
661). A similar crystal structure of 3,5-Py(I)
/Cl was previously
2 D. Bulfield and S. M. Huber, Chem. – Eur. J., 2016, 22, 14434.
3 L. C. Gilday, S. W. Robinson, T. A. Barendt, M. J. Langton, B. R. Mullaney
and P. D. Beer, Chem. Rev., 2015, 115, 7118.
13
reported by Beer et al. In all reported crystal structures of
0
R,R
monocationic XB donors with the 3,5-Py(I) /Z core motif, both
4
G. Cavallo, P. Metrangolo, R. Milani, T. Pilati, A. Priimagi, G. Resnati
triazolyl moieties are in plane with the pyridinium ring. In the
and G. Terraneo, Chem. Rev., 2016, 116, 2478.
5 N. L. Kilah, M. D. Wise, C. J. Serpell, A. L. Thompson, N. G. White,
K. E. Christensen and P. D. Beer, J. Am. Chem. Soc., 2010, 132, 11893.
F. Kniep, L. Rout, S. M. Walter, H. K. V. Bensch, S. H. Jungbauer,
E. Herdtweck and S. M. Huber, Chem. Commun., 2012, 48, 9299.
7 S. M. Walter, F. Kniep, E. Herdtweck and S. M. Huber, Angew. Chem.,
Int. Ed., 2011, 50, 7187.
F. Kniep, S. M. Walter, E. Herdtweck and S. M. Huber, Chem. – Eur. J.,
2012, 18, 1306.
R. Tepper, B. Schulze, P. Bellstedt, J. Heidler, H. Gorls, M. Jager and
U. S. Schubert, Chem. Commun., 2017, 53, 2260.
0 R. Tepper, B. Schulze, M. J ¨a ger, C. Friebe, D. H. Scharf, H. G o¨ rls and
U. S. Schubert, J. Org. Chem., 2015, 80, 3139.
1 S. M. Walter, S. H. Jungbauer, F. Kniep, S. Schindler, E. Herdtweck
and S. M. Huber, J. Fluorine Chem., 2013, 150, 14.
2 S. H. Jungbauer, S. M. Walter, S. Schindler, L. Rout, F. Kniep and
S. M. Huber, Chem. Commun., 2014, 50, 6281.
3 S. W. Robinson, C. L. Mustoe, N. G. White, A. Brown, A. L. Thompson,
P. Kennepohl and P. D. Beer, J. Am. Chem. Soc., 2015, 137, 499.
4 Triply-charged systems featuring two iodotriazolium groups next to
the pyridinium core were also prepared, but suffered from very bad
solubility and decompositon (dealkylation).
Bn,Me
crystal structure of 1,3-Ph(I)
/OTf, on the other hand, the
triazolium groups are out of plane with the central benzene
6
6
ring. As a consequence of the planar geometry of the core
0
R,R
structure of 3,5-Py(I) , p conjugation will allow some delocaliza-
tion of the cationic charge.
8
In summary, a comparison of halogen bond donors based
on either a bis(iodotriazolium)benzene or a bis(iodotriazolyl)-
pyridinium motif allowed elucidation of the influence of charge
9
1
1
1
1
1
(dicationic vs. monocationic) on the Lewis acidity of these
compounds. In a halide abstraction benchmark reaction, the
performance of both types of XB donors was very similar.
Calorimetric measurements also confirmed that overall the Lewis
acidity of the monocationic XB donors is roughly in the same range
as the one of the previously studied dicationic bidentate variants.
The titrations also revealed that XB donors with less Lewis basic
F ꢀ
ꢀ
ꢀ
anions bind stronger to halides (BAr 4 4 NTf2 4 OTf ), in
1
1
5 W. S. Brotherton, R. J. Clark and L. Zhu, J. Org. Chem., 2012, 77, 6443.
6 S. H. Jungbauer and S. M. Huber, J. Am. Chem. Soc., 2015, 137, 12110.
agreement with the performance in the test reaction. X-ray struc-
tural analyses indicate that the cationic charge is likely delocalized 17 N. Watanabe, A. Matsugi, K. Nakano, Y. Ichikawa and H. Kotsuki,
Synlett, 2014, 438.
8 For the corresponding BAr
activity is observed for the dicationic XB donor 1,3-Ph(I)
(35% yield after 5 h), again likely due to low solubility.
9 S. M. Walter, F. Kniep, L. Rout, F. P. Schmidtchen, E. Herdtweck and
S. M. Huber, J. Am. Chem. Soc., 2012, 134, 8507.
0 A. Bondi, J. Phys. Chem., 1964, 68, 441.
over the planar bis(iodotriazolyl)pyridinium motif.
F
4
1
compounds, even a slight decrease in
These findings may have important implications for the design
of strong halogen-bonding-based activators or organocatalysts:
apparently, a monocationic backbone is sufficient to ensure strong
charge-assisted bidentate halogen bonding. The monocationic
Oct,Me
F
/BAr
4
1
2
species, however, feature some distinct advantages over their 21 L. J. Farrugia, J. Appl. Crystallogr., 2012, 45, 849.
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2018