research papers
acid (3.60 g) was added to the reaction mixture. Heating was
stopped once the reagents dissolved, and the mixture was
stirred at room temperature for a week. Ethanol (20 ml) was
removed via rotary evaporation and hexane (15 ml) was
added. After chilling, an orange precipitate formed. The
product (yield 0.81 g, 15.7%) was collected by vacuum filtra-
tion. 1H NMR (400 MHz, DMSO, ppm): ꢁ 7.91 (d, 4H), 7.58 (t,
2H), 7.47 (t, 4H), 6.53 (m, 2H), 6.40 (m, 2H). Single crystals
were obtained via vapor diffusion of hexane into a chloroform
solution of the product at room temperature.
normalized values determined by neutron diffraction results
(Allen et al., 2004) for (1) and (2). An additional pyridine-2-
carboxylate was included in interaction energy calculations for
(1) and (2) in order to maintain charge neutrality. For (3),
H-atom positions were optimized to their energy minima using
the M06-2X/6-31G(d,p) functional/basis set combination.
3. Results and discussion
Compounds (1) and (2) are isostructural salts. Views of (1) and
(2) displaying the atom-labeling schemes are shown in Figs. 1
and 2. The cation resides across a crystallographically imposed
twofold rotation axis and the asymmetric unit contains the
pyridine-2-carboxylate (paꢁ) anion and one-half of the
benzene-1,2-diaminium (bdaH22+) dication for (1) and the 4,5-
dimethylbenzene-1,2-diaminium (Me2bdaH22+) dication for
(2). A search of the Cambridge Structural Database (CSD;
Groom et al., 2016) yielded only one organic structure con-
taining a benzene-1,2-diaminium moiety (CSD refcode
ZEXBEM; Amirthakumar et al., 2018).
The hydrogen-bonding network for (1) is shown in Fig. 3.
An extensive ring system involving protonated amine donors
with a pyridine-ring N atom and carboxylate O atoms as
acceptors is observed. The carboxylate and diaminium
synthons result in R22(9) rings with each aminium group as
donor to a different oxygen acceptor. An R21(5) ring is the
product of a bifurcated aminium donor with a pyridine N atom
and one of the carboxylate O atoms as acceptors. This
hydrogen-bonding motif has been observed for other pyri-
2.2. Refinement
Crystal data, data collection and structure refinement
details are summarized in Table 1. For all compounds, H atoms
bonded to C atoms were refined using a riding model, with
˚
C—H = 0.95 A for aromatic C atoms and, for (2), C—H =
˚
0.98 A for methyl H atoms. For all structures, Uiso(H) =
kUeq(C), where k = 1.2 for H atoms bonded to aromatic C
atoms and 1.5 for H atoms bonded to methyl C atoms. For (1)
˚
and (2), the N—H bond lengths were restrained to 0.91 (2) A;
for (3), the N—H bond lengths were refined freely. For all
compounds, the isotropic diplacement parameters of the H
atoms bonded to N and O atoms were refined freely.
Large K values were noted in the analysis of variance for (1)
and (2). However, the K value is large only for weak reflec-
tions [Fc/Fc(max) is lower than 0.005 for (1) and 0.020 for (2)].
2.3. Hirshfeld surface, fingerprint plots, and interaction
energy calculations
Hirshfeld surfaces, fingerprint plots, interaction energies,
and energy frameworks (Turner et al., 2015) were calculated
using CrystalExplorer17 (Turner et al., 2017). Interaction
energies for (3) were calculated employing the CE-B3LYP/6-
31G(d,p) functional/basis set combination and are corrected
for basis set superposition energy (BSSE) using the counter-
poise (CP) method (Boys & Bernardi, 1970). The interaction
energy is broken down as
˙
dine-2-carboxylate salts (Zesławska et al., 2017). A third ring,
R24(8), involves a single O atom from two different carboxylate
acceptors and a single aminium group from two different
cations, each behaving as a two-H-atom donor. The joined
rings propagate along [001].
A weak ꢀ–ꢀ interaction is observed in (1) with
ꢁ
ꢁ
ꢁ
˚
Cg(pa )ꢀ ꢀ ꢀCg(pa ) = 3.9360 (16) A, where Cg(pa ) is the
pyridine ring centroid. The corresponding distance in (2) is
˚
4.6018 (15) A. As seen in Fig. 3, rings involved in the ꢀ-
Etot ¼ keleE0ele þ kpolE0pol þ kdisEd0 is þ krepE0rep
;
stacking are related by the twofold screw axes. The large
difference in stacking distance can be accounted for by the
disparity in the b-axis lengths (Table 1 and Fig. 4). Both (1)
where the k values are scale factors, E0ele represen0ts the
electrostatic component, E0pol the polarization energy, E dis the
dispersion energy, and E0rep the exchange–repulsion energy
(Turner et al., 2014; Mackenzie et al., 2017). The C—H bond
lengths were converted to normalized values based on neutron
diffraction results (Allen et al., 2004).
Interaction energy calculations were also performed on
molecules in the gas phase using SPARTAN’16 (Wavefunc-
tion, 2016). Density functional theory (DFT) calculations
using the M06-2X (Zhao & Truhlar, 2008) functional with a
6-31G(d,p) basis set were employed for the determination of
interaction energies, which were corrected for BSSE
employing the CP method (Boys & Bernardi, 1970). Atomic
coordinates obtained from the crystallographic analysis were
used for all non-H atoms. Because bond lengths obtained for
H atoms from X-ray crystallographic analyses are inaccurate,
the positions of the H atoms were adjusted based on
Figure 1
View of the molecular structure of (1), showing the atom-labeling scheme.
Displacement ellipsoids for non-H atoms are drawn at the 50%
3
probability level. [Symmetry code: (a) ꢁx + 1, y, ꢁz + .]
2
ꢂ
Acta Cryst. (2019). C75
Powers and Geiger
Hydrogen bonding in benzene-1,2-diamine salts and cocrystal 3 of 7