Understanding the crystalline formation of triazene: N -oxides and the role of halogen?π interactions

Article abstract of DOI:10.1039/c7ce02015e

The crystallization of a series of 1-(4-halophenyl)-3-phenyltriazenide N1-oxides (1-4) and 1-(phenyl)-3-phenyltriazenide N1-oxide (5) was evaluated using the supramolecular cluster approach. This method is an efficient tool to assess the crystallization mechanism of compounds and, consequently, the steps involved in crystal formation. Compounds 1 and 4 show crystallization in two main steps while compounds 2 and 3 present three main steps, in which column formation occurs in the first step. The crystallization process for 5 occurs in 3 main steps, starting from a robust dimer formation (-16.82 kcal mol^-1). Two new parameters-N_CG% (topological and energetic contribution percentage) and NG/NC (energetic parameter/topological parameter ratio)-assisted in the interpretation of crystal growth. Compounds 1-4 showed N_CG% = 50 in the first step while compound 5 reached only 50% of the contribution in the second step. The differences in N_CG% were attributed to strong hydrogen bonds in the non-halogenated compound. The dominant parameter in each step of the crystallization process was indicated by the NG/NC parameter. The crystallization mechanism in all compounds was initially driven by an energetic process followed by a topological process. The existence of X?π interactions was shown and was observed to be a consequence of a topological process and without any major contributions to crystal formation. Thermal analysis and UV-vis spectral data were also discussed regarding the properties of these compounds.

Full text of DOI:10.1039/c7ce02015e

View Article Online
View Journal
CrystEngComm
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: M. A. P. Martins, P.
R. D. S. Salbego, G. Alves de Moraes, C. R. Bender, P. J. Zambiazi, T. Orlando, A. Berti Pagliari, C. P. Frizzo
and M. Horner, CrystEngComm, 2017, DOI: 10.1039/C7CE02015E.
This is an Accepted Manuscript, which has been through the
Volume 18 Number 1 7 January 2016 Pages 1–184
Royal Society of Chemistry peer review process and has been
accepted for publication.
CrystEngComm
Accepted Manuscripts are published online shortly after
www.rsc.org/crystengcomm
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
HIGHLIGHT
Tiddo J. Mooibroek, Antonio Frontera et al.
Towards design strategies for anion–π interactions in crystal engineering
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
rsc.li/crystengcomm

Page 1 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
Understanding the Crystalline Formation of Triazene NꢀOxides and the Role of
Halogen···π Interactions
Marcos A. P. Martins,*a Paulo R. S. Salbego,a Guilherme A. de Moraes,b Priscilla J.
Zambiazi,b Caroline R. Bender,a Tainára Orlando,a Anderson B. Pagliari,a Clarissa P.
Frizzo,a Manfredo Hörner*b
aNúcleo de Química de Heterociclos (NUQUIMHE), Department of Chemistry,
Federal University of Santa Maria (UFSM), 97105ꢀ900, Santa Maria, RS, Brazil
bNúcleo de Investigação de Triazenos e Complexos (NITRICO), Department of
Chemistry, Federal University of Santa Maria (UFSM), 97105ꢀ900, Santa Maria, RS,
Brazil
*Eꢀmail of corresponding author: marcos.nuquimhe@gmail.com
ABSTRACT
The crystallization of a series of 1ꢀ(4ꢀhalophenyl)ꢀ3ꢀphenyltriazenide N1ꢀoxides (1ꢀ4)
and 1ꢀ(phenyl)ꢀ3ꢀphenyltriazenide N1ꢀoxide (5) were evaluated using the
supramolecular cluster approach. This method was an efficient tool to assess the
crystallization mechanism of compounds and, consequently, the steps involved in
crystal formation. Compounds 1 and 4 show crystallization in two main steps
while compounds 2 and 3 present three main steps, in which column formation occurs
in the first step. The crystallization process for 5 occurs in 3 main steps, starting from
a robust dimer formation (ꢀ16.82 kcal molꢀ1). Two new parameters – NCG%
(topological and energetic contribution percentage) and NG/NC (energetic
parameter/topological parameter ratio) – assisted in the interpretation of crystal
growth. Compounds 1ꢀ4 showed NCG% = 50 in the first step while compound 5 only
reached 50% of contribution in the second step. The differences in NCG% were
attributed to strong hydrogen bonds in the nonꢀhalogenated compound. The dominant
parameter in each step of the crystallization process was indicated by the NG/NC
parameter. The crystallization mechanism in all compounds was initially driven by an
energetic process followed by a topological process. It was showed the existence of
the Xꢁꢁꢁπ interactions and observed as a consequence of a topological process and
without any major contributions to crystal formation. Thermal analysis and UVꢀVis
spectral data were also discussed regarding the properties of these compounds.

CrystEngComm
Page 2 of 35
View Article Online
DOI: 10.1039/C7CE02015E
INTRODUCTION
The understanding of crystalline networks requires the use of certain parameters in
order to understand the interactions that form the system. The geometric parameter
has been a widely used tool for describing intermolecular interactions, including
halogen interactions.1,2 Recognizing interactions helps elucidate the crystallization
processes of different molecules whose final goal is to correlate characteristics of this
process with final emergent properties. However, a broad approach using topological
and energetic data parameters in addition to geometric data for the study of crystals
remains scarce in the literature.
According to Kitaigorodskii3, the assembly route of molecules to crystals is governed
by molecules approaching each other to form the first closeꢀpack, which are then
closeꢀpacked by other layers or chains of molecules. Because of this, the
crystallization process can be viewed as a stepwise progression in which molecule
association increases system complexity, leading to the final 3D crystal. This
progression occurs analogously to Darwinian evolution4, where the process of selfꢀ
organization blends competition, cooperation, and complementarity. The Darwinian
process improves individual organization, which leads to the emergence of the
crystalline arrangement.
Following this idea, our research group proposed a method in a previous study5 where
crystallization mechanisms for isoxazoles were presented in order to better understand
crystalline formation. This proposal was based on the observation of the energetic
contributions of different dimers that formed the crystal and used the supramolecular
cluster as portion of the study. Our findings show that the supramolecular cluster has
an energetic hierarchy related to the formation of interactions; in that case, the
hydrogen bond was the strongest interaction and the first to form. The πꢀinteractions,
on the other hand, were weaker and observed in other steps of the crystallization.
Triazene1ꢀOxide derivatives are an important class of molecules due to the hard
basicity character of the coordinating atoms O and N. This characteristic imposes
electronic properties that allow molecular selfꢀorganization via intermolecular
secondary interactions6. These species have different applications in chemistry and
biology, such as in sensors, electronic displays7,8
, as well as in antileukemic and
anticancer drugs9,10. Some of these applications are a result of nonꢀcovalent
intermolecular interactions and the packaging behavior of these compounds.

Page 3 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
Therefore, our aim is to assess the comprehension of the system by studying the
analysis of crystal formation while considering the topological and energetic
normalized data of the supramolecular cluster as a tool. In pursuance of this, a series
of five 1ꢀoxide triazene was used, in addition to the proposal of crystallization
mechanisms for each system in order to observe the contrasts between different
substituents. Important information regarding the contribution of each crystallization
step will be provided. This will be done by observing the dominant parameter of the
process in addition to observing the influence of halogen···π interactions to crystal
formation. Geometric and QTAIM data were used to answer the questions on the role
of halogen···π interactions in crystal formation. Additionally, thermal analyses (TGA
and DSC) and UVꢀVis spectroscopy were carried out to determine the potential
applications of these molecules.
RESULTS AND DISCUSSION
Molecular description
Crystals suitable for Xꢀray single crystal analysis of compounds 1ꢀ5 were obtained
from the respective solutions by means of crystallization through slow evaporation of
the solvent. Figure 1 shows the molecular structures of the compounds.
Crystallographic data are shown in Table 5 and additional data regarding bond lengths
(Å) and angles (°) are demonstrated in Table S1 (ESI†).

CrystEngComm
Page 4 of 35
View Article Online
DOI: 10.1039/C7CE02015E
1
3
2
4
5
Figure 1. Projection of the molecular structures of compounds 1 to 5. H atoms are
represented with arbitrary radii. Nonꢀhydrogen atoms are represented with thermal
ellipsoids at the 50% probability level.
All compounds discussed in this study belong to the class of triazene N1ꢀoxides, such
as derivatives from the classical triazenes, in which the doubleꢀbonded terminal Nꢀ
atom in the nitrogen triad is substituted by an “oxo” group in the solid state. In
solution, these compounds are in a chemical equilibrium with the respective tautomer
form in which the oxo group is protonated resulting in a hydroxyl functional group.
The diazo(oxide)amine [ꢀN=N(→O)ꢀN(H)ꢀ] functional group can be stabilized by
substitution of the terminal N atoms with alkyl, phenyl, or aryl substituents. In the
present study, 1ꢀphenylꢀ3ꢀparaꢀhalophenyl (pꢀF, pꢀCl, pꢀBr and pꢀI) substitution was
selected in view of the comparison of its crystallographic features with that of the
unsubstituted 1,3ꢀbis(phenyl)triazene N1ꢀoxide.
Compound 1 belongs to a monoclinic crystal system in a primitive threeꢀdimensional
lattice corresponding to Bravais type P and chiral space group P21 (Z=2; Z’=1). Given
that compound 1 does not present a heavy atom with anomalous scattering, we were
unable to determine its absolute configuration. Thus, the selected molecular

Page 5 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
configuration was based on the lowest Flack value observed, which was 0.1(3). An
attempt to solve the structure considering the centrosymmetric space group P21/m in
comparison with the acentric space group P21 was fruitless. Compounds 2 and 3 are
isostructural and belong to the centrosymmetric monoclinic space group P21/c (Z = 4;
Z’ = 1). Compound 4 belongs to the orthorhombic and chiral P212121 space group
(Z=4; Z’=1), and its molecular configuration was selected based on the lowest Flack
value observed, which was 0.048(18). Compound
5 crystallizes in the
centrosymmetric monoclinic space group P21/n (Z=4; Z’=1). All molecules present
the expected trans stereochemistry regarding the N=N double bond in the [ꢀ
N=N(→O)ꢀN(H)ꢀ] diazoamine N1ꢀoxide moiety and an almost planar shape.
Molecules 1ꢀ5 deviate slightly from planarity [r.m.s. deviation from the leastꢀsquares
plane through the nonꢀhydrogen atoms: 0.0857 Å (1), 0.0353 Å (2), 0.0318 Å (3),
0.0466 Å (4), 0.0468 Å (5), respectively]. The N13ꢀN12ꢀN11ꢀO dihedral angle in the
molecules {[ꢀ0.3(2)° (1)], [ꢀ0.9(4)° (2)], [ꢀ1.1(4)° (3)], [1.4(12)° (4)] and [0.5(3)° (5)],
respectively]} are similar to the angle [N13ꢀN12ꢀN11ꢀO1 = 0.7(5)°] found in 3ꢀ
phenylꢀ1ꢀmethyltriazene 1ꢀoxide, and contribute to the nearꢀplanar conformation of
the molecules11. The average value of 1.295 Å for the N11ꢀO bond distance observed
in the molecules 1 to 5 (ESI†, Table S1) is in agreement with the N11ꢀO1 bond
distance of 1.246(4) Å observed in the similar compound 3ꢀphenylꢀ1ꢀmethyltriazene
1ꢀoxide, which indicates that the molecules adopt the Nꢀoxide tautomeric form in the
solid state11. The angles between the planes of the rings presented values of 10.32,
3.01, 2.16, 5.93, and 3.53 for 1 to 5, respectively. These values are a result of the
influence of the crystal packing on the different intermolecular interactions of each
structure.
Supramolecular cluster
The supramolecular cluster approach12 ꢀ grown from the neighboring molecules
around a reference molecule (M1) ꢀ was used to obtain information on the contact
surface (topologic parameter) and energy of interaction (energetic parameter) between
the molecules in the crystalline arrangement of each structure (see ESI†, Figures S2ꢀ
S7 and Tables S2ꢀS6). The cluster was speculated since we are considering a 3D
portion of the crystal that provides all the necessary information of the involved

CrystEngComm
Page 6 of 35
View Article Online
DOI: 10.1039/C7CE02015E
interactions in the whole crystal. This method can assess the understanding of crystal
formation without neglecting molecules that are important to the system.
The neighboring molecules around M1 represent the molecular coordination number
(MCN). These dimers are represented by the interaction between the M1 and MN
molecules (e.g., M1⋯M2, M1⋯M3, …, M1⋯M15). All molecules which share
surface contact area of the van der Waals radii with the reference molecule M1 are
considered part of the cluster. The supramolecular cluster with the surface contact
area of each dimer was obtained using the VoronoiꢀDirichlet Polyhedra (VDP) with
the ToposPro 4.0 software13. This method has already been reported5,14–16 as an
efficient tool to assess crystalline understanding, especially regarding the steps
involved in crystal formation, i.e. the crystallization mechanism5.
The different supramolecular clusters can be compared by the normalization12 of raw
topological and energetic data in order to observe the distinct behavior of the dimers
in the formation of each crystalline structure. Normalization is a tool that places all
the different data in the same scale, allowing precise comparisons and interpretations
of different systems. Hence, all N molecules in the cluster contribute in some way to
the whole system, and it is necessary to assess the contribution of each one in relation
to a standard, in our case, in relation to the number of molecules present in the cluster
(around reference molecule M1).
For example, in a cluster with an MCN of 16, all dimers must have an ideal
contribution of 1 for all considered parameters (e.g. energy), which leads to a 100%
contribution of 16. However, this is not the reality of most molecules where distinct
interactions occur, leading to different dimeric contributions. Therefore, all raw data
should be reduced to the same scale using the MCN as reference for this
normalization.
From normalized values, intermolecular interactions can be divided into four types12.
Type I represents a molecular dimer with high interaction energy on a small contact
surface (e.g. hydrogen bonds). Type II corresponds to high interaction energy on a
large contact surface (e.g., πꢁꢁꢁπ interactions). Type III corresponds to both small
interaction energy and contact surfaces with a maximum difference of ± 0.5 between
the both parameters, and Type IV is a contribution based on large contact surfaces
with low interaction energy.

Page 7 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
Topological and energetic normalized data of each dimer from the supramolecular
cluster of compounds 1, 2, and 5 (for 3 and 4, see ESI†, Figure S1) with the respective
dimer classification are described in Figure 2.
1
2
5
Figure 2. Topological and energetic normalized data of each dimer from the
supramolecular cluster of compounds 1, 2, and 5.
Figure 2 demonstrates that compound 1 has two typeꢀI dimers and two typeꢀII dimers,
which corresponds to dimers involved in hydrogen interactions and stacking,
respectively. The remaining dimers exhibited a typeꢀIII behavior with low area and
energy values. Notably, data from compound 1 is similar to data from compound 4
(see ESI†, Figure S1). Compound 2 (for compound 3, see ESI†, Figure S1) had only
one dimer with type I behavior and one dimer with type II, which is different from
what was observed in compounds 1 and 4. These data provide important information
on the formation of the crystalline structure, especially concerning the first dimers
formed (the most energetic ones). Finally, compound 5 displayed a distinct
characteristic when compared to compounds 1ꢀ4. The supramolecular cluster only
presented one typeꢀI dimer of the highest energetic contribution of all systems, and

CrystEngComm
Page 8 of 35
View Article Online
DOI: 10.1039/C7CE02015E
two typeꢀII dimers with lower contribution regarding stacking interactions when
compared to 1ꢀ4. This normalized data approach provides significant data of the
characteristics of the dimers and enables the understanding of the crystal formation,
which is discussed in the following section.
Crystallization mechanism
Since crystallization is a stepwise process that takes into account energy hierarchies,
crystallization mechanisms have been proposed for a series of compounds in order to
understand crystal formation. The normalized data previously presented was used to
assess the proposed crystallization mechanism5 for each structure. The main
difference between these proposed mechanisms and the earlier study3 is that the
normalized values were used in all steps and in the interpretation of the mechanisms.
Firstly, individual dimer data regarding stabilization energy was used to define the
highest and lowest dimer hierarchies in order to assess the steps of crystal formation.
Then, the main steps were defined, where in some cases more than one hierarchy is
present in the same crystallization step. The proposed crystallization mechanism for
compounds 1, 4, and 5 are presented in Figure 3, 4, and 5.
Once the main steps were determined, we returned to the data of the normalized dimer
in order to obtain two main parameters to assess the elucidation of the process:
•
The topological and energetic contribution for each step of the crystal growth.
The percentage of contribution (NCG%) from each given step can be found by
the sum of the normalized topological and energetic data (NC+NG) of each
step. As a result, we are able establish the steps with the highest contribution
for both parameters, i.e. surface area and energy. The step contribution can
indicate where the definitive step towards the considered crystalline structure
is.
•
The ratio between NG and NC to find the parameter that rules each step. The
NG/NC provides important information on the governing parameter in each
step, in which values over 1.0 indicate higher energetic contribution and
values under 1.0 indicate dominant topologic parameter. Values near 1.0
indicate a balance between parameters. Thus, it is possible to inform where the
changes between the dominant parameters occur.

Page 9 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
Figure 3 presents the steps of the crystallization mechanism for compound 1.
Compound 4 also presents the same general steps, and an illustration can be found in
the ESI† (Figure S8). The cluster molecules involved in the steps were depicted and
its normalized energy values were highlighted.
Figure 3. Proposal of crystallization mechanism of compound 1. NG = normalized
energy. The shaded area represents the portion in the previous step.
The proposal of the crystallization mechanism for structures 1 and 4 shows two major
steps (Figure 3). The first step occurs with the formation of interconnected columns
by hydrogen interactions with stabilization energy of ꢀ9.78 kcal molꢀ1 (NG = 2.99) for
each of the two dimers involved and energy of ꢀ8.85 kcal molꢀ1 (NG = 2.71) for the
molecule stacking of each dimer. The dimers involved in this step present the highest
portion of surface contact area and stabilization energy in relation to the whole
supramolecular cluster (see Figure 2). This step contributes 53.4% to crystal
formation when both stabilization energy (NG) and surface contact area (NC) are
considered, namely the topological and energetic parameters are taken into
consideration in the final contribution (NCG%).

CrystEngComm
Page 10 of 35
View Article Online
DOI: 10.1039/C7CE02015E
Step 2 presents the final crystal formation (3D growth) with the interaction between
the columns formed in step 1. These dimer interactions occur when stabilization
energy is below ꢀ2.7 kcal molꢀ1 (NG < 0.83). However, despite the higher number of
dimers involved, the remaining 46.6% of topological and energetic contribution for
the crystallization mechanism is occurs to form the 3D network.
A value of 1.46 was found for the first step, regarding the second parameter (NG/NC),
which indicates a strong impact from the energetic parameter in this step of crystal
growth, i.e. almost 50% more influence than the topologic factor. Moreover, a value
of 0.65 was found for the second step, demonstrating a change in the dominant
parameter in this step of the crystallization where topology plays a more relevant role
than stabilization energy.
Compounds 2 and 3 present similar clusters with isostructural behavior, which leads
to identical proposals of crystallization mechanism. Figure 4 depicts the mechanism
of compound 2 (for compound 3, see ESI†, Figure S9).
Figure 4. Proposal of the crystallization mechanism of compound 2. NG =
normalized energy. The shaded area represents the portion in the previous step.
There are three main steps in the proposal of crystallization mechanism for
compounds 2 and 3. In the first step, the two main energetic dimers form the
displaced columns. The dimer involved in a hydrogen interaction of ꢀ15.17 kcal molꢀ1
(NG = 3.31) and the stacking dimer ꢀ13.65 kcal molꢀ1 (NG = 2.98). Two dimers from

Page 11 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
the supramolecular cluster with stabilization energy of ꢀ4.48 kcal molꢀ1 (NG=1.96)
appear in this step due to the formation of the displaced columns. The dimers
considered in the first step had a NCG% of 54.6%, indicating that this first step is a
point of no return for the crystal since over 50% of all stabilization energy and surface
contact area of the crystal rely on this first step.
The second step is the interaction between the columns (formed in step 1) with
interaction energy of ꢀ6.34 kcal molꢀ1, which leads to the 2D growth of the crystal. In
this step, crystallization is increased by a NCG% of 9.6%. Finally, the third step is
responsible for the formation of the 3D crystal network. In this final step, the 2D
columns interact through dimers with lower stabilization energy below ꢀ3.2 kcal molꢀ1
(NG < 0.69). This final step represents the remaining 35.8% of the topological and
energetic contribution. The NG/NG parameter presented values of 1.17, 1.06, and
0.77 for the first, second, and third steps, respectively. These data demonstrate the
evolution of the dominant parameter through the crystallization process with a slight
dominance of the stabilization energy in the first step, and the subsequent rise of
topological governance in the other two steps.
Figure 5 demonstrates the proposal for the crystallization mechanism of compound 5,
which differs from the other supramolecular structures.
Figure 5. Crystallization mechanism of compound 5. NG = normalized energy. The
shaded area represents the portion in the previous step.
The first step for compound 5 (Figure 5) differs from the previous compounds due to
the formation of a dimer with a robust stabilization energy of ꢀ16.82 kcal molꢀ1. The

CrystEngComm
Page 12 of 35
View Article Online
DOI: 10.1039/C7CE02015E
normalized data (NG = 4.49) shows the magnitude of this dimer with a value four and
a half times greater than expected, which represents a strong hydrogen interaction.
This first step corresponds to 19.2% of contribution of the whole crystal, even though
it is only one dimer. The second step is the stacking interaction of these dimers in
order to form columns with stabilization energy of ꢀ9.44 kcal molꢀ1 (NG = 2.52) and ꢀ
3.89 kcal molꢀ1 (NG = 1.04). At this moment, there is an increase of 37% in
contribution. The third step is similar to that of compounds 1 and 4, where the
columns interact with each other with energies below ꢀ2.6 kcal molꢀ1 (NG < 0.69) in
order to form the 3D crystal lattice. The last step contributes with a NCG% of 43.8.
The crystallization behavior of this structure differs from the others in the
contribution of each step, since there is a unique first step with a strong hydrogen
interaction between the two molecules. The G/C ratio showed values of 2.72, 1.05,
and 0.63 for steps 1, 2, and 3, respectively. This data clearly demonstrates the
transition between the dominant parameter in each step. The first step with high
energy in a smaller surface area had a value almost three times higher, thus,
highlighting the importance of the energetic parameter. The transition in the dominant
parameters in the second step indicated a balance between the surface and energy
parameters. The last step, due to the presence of lower stabilization energies in
column approximation, was highly influenced by the topological parameter.
For better visualization, the three proposed mechanisms of crystallization for
compounds 1ꢀ5 are summarized in Figure 6.

Page 13 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
Figure 6. Summary of crystallization steps for compounds 1ꢀ5 with the topologic and
energetic contribution (NCG%). NG/NC parameter is also demonstrated for each stage.
The data presented in Figure 6 shows the main event in each step for compounds 1ꢀ5
until 3D crystal growth. In each step, the combined contribution considering
topological and energetic parameters is described in order to have a comprehensive
view of the steps in all the studied supramolecular structures. Compounds 1ꢀ4 had
contributions of over 50% in the first step. Compound 5, which formed a strong first
dimer, only reached a contribution of 50% when the first and second stages were
considered. However, it is possible to state that the first dimer is formed irreversibly,
since the hydrogen interaction formed corresponds to almost 20% of the entire
interaction energy of the crystal.
The NG/NC parameter allows us to observe the transition between the dominant
parameter during the crystallization process. Compounds 1ꢀ5 behaved similarly
during the steps, in other words the first dominance of the stabilization energy of the
interactions involved in the considered dimers, its subsequent decrease, and the
increase of the topological relevance to the crystal growth. The final step of the
crystallization process relies almost exclusively on a topological dependent process.

CrystEngComm
Page 14 of 35
View Article Online
DOI: 10.1039/C7CE02015E
Thus, based on such findings, differences between the angles of the rings in each
structure mentioned in the molecular description section can be better understood.
Compounds 1 and 4 have the highest angles, which is unexpected due to the
differences between F and I. However, the influence on these torsion angles relies on
crystal packing, and it was possible to observe that 1 and 4 have similar mechanisms
of crystallization achieving a similar 3D crystal network. Structures 2 and 3 are
isostructural and consequently present similar torsion values. The structure without
substituent (5) has an intermediate value, which is distinct from the others and
consistent with its different crystal packing.
Intermolecular interactions
The supramolecular cluster aided in comprehending the differences in the crystal
packing of the studied systems. From this approach, it is possible look even further
into specific interactions to assess if considered dimers are relevant to crystal packing.
In view of studies that conventionally focus on intermolecular interactions in a
geometric perspective, we will first discuss this parameter and afterwards advance to
a broader view using the supramolecular cluster as the essential portion of the study.
This discussion will focus mainly on halogen interactions and its importance to
crystal packing. These intermolecular interactions were identified according to the
standardization of the Platon program17. The arrangements involving halogenꢁꢁꢁπ
interactions for compounds 1–4 are shown in ESI† (Figure S40), where a supposed
prominence of these intermolecular interactions is demonstrated. Figure 7 presents the
orientation of the halogen atoms for compounds 1ꢀ4 in relation to the plane of the
aromatic ring forming the halogenꢁꢁꢁπ interactions. The geometric parameters for all
halogenꢁꢁꢁπ interactions are presented in Table 1.

Page 15 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
Figure 7. Orientation of the halogen atoms in relation to the plane of the aromatic
ring forming the halogenꢁꢁꢁπ type interactions for compounds 1ꢀ4.
Table 1. Geometric parameters for the halogen interactions observed in compounds
1ꢀ4.
Parameter
Dist. ra (Å)
Dist. rb(Å)
Dist. rr(Å)
β (°)
1
2
3
4
3.2654(5)
3.4127(5)
4.1018(7)
91.443(5)
44.757(4)
3.5379(4)
3.4023(4)
3.7591(5)
154.064(7)
70.169(8)
3.5332(1)
3.4054(1)
3.7499(1)
154.457(2)
70.344(2)
3.6053(1)
3.9080(2)
3.9080(2)
170.601(1)
49.518(1)
α (°)
Symmetry operators
2ꢀx, ꢀ½+
y, ꢀ
z
1+
x, ꢀ½ꢀ
y, ½+
z
1+x, ½ꢀy, ½+
z
1ꢀx, ꢀ½+y, 1½ꢀz
α
= X⋅⋅⋅Centroid–plane.
β
= C–X⋅⋅⋅Centroid.
The interactions presented in 1 and 4 are classified as localized and type (L)1,2. This is
due to their parameter ra (distance from halogen to a specific carbon) as the shortest
distance when compared to the parameters rr (distance from the halogen to the
centroid) and rb (distance from the halogen to the bond between two carbons of the
aromatic ring). Table 4 shows ra values equal to 3.2654 (5) and 3.6053 (1) Å for
compounds 1 and 4. The distance of the van der Waals radii for FꢁꢁꢁC and IꢁꢁꢁC are
3.17 and 3.68 Å, indicating a strong halogenꢁꢁꢁπ interaction for compound 4, whereas
values for distance ra are lower than the sum of the van der Waals IꢁꢁꢁC radii.

CrystEngComm
Page 16 of 35
View Article Online
DOI: 10.1039/C7CE02015E
Compounds 2 and 3 are classified as semiꢀlocalized and type (SL). Their distances of
lower values registered for parameter rb are 3.4023 (4) and 3.4054 (1) Å, respectively,
where the distance of the interactions Clꢁꢁꢁπ and Brꢁꢁꢁπ are 3.4023 (4) and 3.4054 (1)
Å. In this sense, values below the sum of the van der Waals radii, which are 3.45 and
3.55 Å, respectively, for ClꢁꢁꢁC and BrꢁꢁꢁC, indicating a significant interaction. The
halogenꢁꢁꢁπ interaction nomenclature for compounds 1 ꢀ 4 can also be attributed to the
geometric criteria. For compound 1, the Fꢁꢁꢁπ interaction shows angle β (Cꢀ
FꢁꢁꢁCentroid) as 91.443 (5)°, a value very close to the orthogonal one, thus indicating
that it is a lone pairꢁꢁꢁπ type interaction according to Mooibroek et al.18. For
compounds 2, 3, and 4, the halogenꢁꢁꢁπ interactions have a β angle of 154.064 (7),
154.457 (2), and 170.601 (1), respectively, and present values close to linearity
(180°). This indicates that these are σꢀhole type interactions.
In this geometric approach, it is possible to assess some magnitude of the Xꢁꢁꢁπ
interactions using only the atomꢀatom distances, in accordance with the established
literature. However, how can the existence of the Xꢁꢁꢁπ interactions be assured only by
means of raw geometric data? And what about the strength of those interactions, how
can it be assessed without considering an energetic approach? How can it be properly
assessed that these type of interactions are relevant to crystal formation?
QTAIM approach
The Quantum Theory of Atoms in Molecules (QTAIM)19 analysis was carried to
assess the identification and strength of the intermolecular interactions. This was done
to corroborate the intermolecular interactions observed by the geometric parameter
and to attempt to answer some of the earlier questions. Bond critical points (BCPs)
were obtained to confirm the existence of the interactions. The intermolecular X···π
interactions were confirmed for structures 1ꢀ4 (Figure 8), corroborating the observed
data in the geometric discussion in addition to other interactions observed. In the case
of structure 1, two different types of interactions (X···π and CꢀH···π) were observed,
which were different from compounds 2ꢀ4 that presented only X···π interaction.
Structure 5, which had no replacement in the phenyl ring, presented, in a similar
dimer, two CꢀH···π interactions.
The first question regarding the existence of the intermolecular interactions is then
answered: each interaction involved in the considered dimer has been identified by

Page 17 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
the BCPs in the QTAIM. The second and third questions are still pendent and, thus, it
is necessary to go beyond the energy calculation of the dimer and observed
interaction. A referential is necessary to assess the relevance of this dimer/interaction,
and the supramolecular cluster can aid us in acquiring this information. Therefore, the
energy involved in each interaction was obtained through the fragmentation of the
energy from the considered dimer using a method previously reported5,14, which is
highlighted in Figure 8.
ꢀ1.10
ꢀ1.76
ꢀ1.59
1
2
ꢀ2.61
ꢀ2.56
3
4
ꢀ1.50
ꢀ0.83
5
Figure 8. Representation of X···π (1ꢀ4) and CꢀH···π (1 and 5) interactions through
the BCPs observed by QTAIM analysis. The energy involved in each interaction is
highlighted (in kcal molꢀ1).
The QTAIM data regarding the interactions involved in those dimers are presented in
Table 2. These interactions observed by the BCPs, in a separate form, can induce the

CrystEngComm
Page 18 of 35
View Article Online
DOI: 10.1039/C7CE02015E
observer to conclude that X···π interactions are relevant to crystal formation.
However, when considering the energy of the involved dimers that range from ꢀ1.76
to ꢀ2.69 kcal molꢀ1, it becomes clear that these dimers have low energetic value when
considering total crystal energy. Consequently, this information helps answer the two
remaining questions.
Table 2. QTAIM data and atom interaction energy (GAI)b for dimers with H···π and
X···π interactions for compounds 1–5.
GM1···MN
ρINT
GAI (x···Y)
d
%
Comp.
Dimer
Interaction
(kcal molꢀ1)a
(u.a.)b
(kcal molꢀ1)c
3.7
5.4
F···π
H···π
0.004047
0.005811
ꢀ1.10
ꢀ1.59
M1···M12
M1···M13
ꢀ2.69
ꢀ1.76
ꢀ2.61
ꢀ2.56
ꢀ2.33
1
2
3
4
5
M1···M9
M1···M14
5.5
7.9
7.0
Cl···π
Br···π
I···π
0.006136
0.007093
0.005375
ꢀ1.76
ꢀ2.61
ꢀ2.56
M1···M9
M1···M14
M1···M16
M1···M19
2.8
5.0
H···π
H···π
0.002655
0.004812
ꢀ0.83
ꢀ1.50
M1···M9
M1···M16
aStabilization energy obtained by DFT calculation. bObtained using QTAIM. cInteraction energy
involved in each interaction obtained by the fragmentation. dInteraction contribution regarding the total
energy cluster.
The fragmentation of the interaction energy of the dimer in the different interactions
provides important additional data. The contribution of X···π interactions are
discernible when the whole system is considered. Moreover, X···π interactions are
responsible for less than 7.9% of the entire stabilization energy of the supramolecular
cluster, contributing 3.7, 5.5, 7.9, and 7.0% for compounds 1 to 4, respectively. This
data provides an important insight on X···π interactions and allows us to answer the
previous question of the importance of this kind of interaction for crystal formation.
This broad analysis has shown us that in these series of crystal molecules, halogen···π
interactions are a topological consequence of the crystallization process because they
are present in the last steps of the crystallization process in the approximation of the
columns. The systematic approach, which is not solely dependent on geometric
parameters, enabled us to confidently assess that halogen···π interactions do not drive
the crystallization process.

Page 19 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
Using the molecular electrostatic potential (MEPs), it is possible to observe the
regions with negative and positive potential in the molecules20. In the case of
compounds 2ꢀ4, it is possible to observe the positive region in the halogen atom (σꢀ
hole) (Figure 9), which allows the interaction σꢀhole···π observed in the previous
section to occur. Compound 1 with the fluorine atom, on the other hand, does not
present the σꢀhole, which leads to a loneꢀpair···π interaction.20 Additional MEPs
regarding the involved dimers are presented in ESI† (Figure S10).
1
2
3
4
Figure 9. Molecular electrostatic potential for compounds 1ꢀ4. The Vmax is
highlighted.
Thermal Analysis
Thermal analytical techniques can provide important information of storage and
stability of different compounds. As previously reported in the literature, triazenes are
compounds with excellent biological activity against tumor cells (e.g. TLX5
lymphoma and L1210 leukemia cells)9,10,21. Thus, understanding the thermal
behavior of these compounds is an important and integral part of developing stable
drugs22.

CrystEngComm
Page 20 of 35
View Article Online
DOI: 10.1039/C7CE02015E
DSC and TGA analysis were carried out to elucidate the thermal performance of these
structures in a broad range of temperature. Firstly, TGA screening was carried out for
compounds 1ꢀ4 in a temperature range of 25 to 900 °C in order to observe the
decomposition steps and decomposition temperature of the structures as a function of
the halogen pꢀsubstituted in the aromatic ring (see ESI†; Figures S11ꢀS15). The
decomposition percentage indicates that the decomposition of the samples is almost
complete (> 90%) in the evaluated temperature range. Nevertheless, Figure 10 shows
the TGA thermograms of compounds 1ꢀ4 in a range from 100 to 400 °C, in which the
main changes in weight loss versus temperature were observed. For a better
visualization, the TGA thermogram of 5 was omitted from Figure 10 and depicted in
the ESI.
Figure 10. Overlap of TGA thermograms of 1ꢀ4 at heating rate of 10 °C minꢀ1.
From Figure 10, it is possible to note that these structures decompose in two main
steps, the first from 90 to 220 °C and the second from 150 to 450 °C, approximately.
Furthermore, by peak differences in the derivative weight as a function of temperature
plots, it can be note that the halogens influence the decomposition pattern of triazene
Nꢀoxides. Table 3 demonstrates data regarding initial decomposition temperature (Ti),
final decomposition temperature (Tf), decomposition percentage (%), and temperature
of maximum decomposition (Td). Ti shows the temperature in which the compound
starts to lose mass and indicates that no thermal degradation based on mass loss
occurs in temperatures below Ti. Tf indicates the temperature in which mass loss is
complete. Td corresponds to the temperature of the maximum peaks in the derivative
weight (%) as a function of temperature (T) in the TGA thermogram (Figure 10).

Page 21 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
Table 3. Thermogravimetric data obtained by TGA in a heating rate of 10 °C minꢀ1.
c
d
Compound
Step
Tia(°C)
Tfb(°C)
Td (°C)
%
1
2
129.36
155.16
155.16
396.20
143.97
177.54
52.50
41.47
1
1
2
137.24
173.59
173.59
400.05
159.96
241.27
67.77
23.68
2
3
4
5
1
2
94.92
207.79
449.78
160.92
254.66
63.34
27.58
207.79
1
2
121.70
217.35
217.35
455.52
162.83
282.40
66.61
20.45
1
2
128.40
154.26
154.26
312.88
138.76
166.60
60.49
35.84
aInitial decomposition temperature. bFinal decomposition temperature. cTemperature of maximum
decomposition. dDecomposition percentage.
Halogens, in general, tend to increase the stability of triazene compounds as observed
in Table 3. An order for decomposition temperatures can be estimated for the
compounds in this study, in which: 5 < 1 < 2 < 3 < 4. This order in relation to
compounds in which X = halogen corresponds to: F < Cl < Br < I.
In the literature, halogen compounds exhibit a pattern of decomposition behavior
generally related to the strengths of the carbonꢀhalogen (C−X) bonds. The greater the
C−X bond dissociation enthalpies, the more thermally stable the halogenated
compounds23. For alkyl halides, for example, the bond dissociation enthalpies are 106,
81, 69, and 54 kcal molꢀ1 for C
−F, C−Cl, C−Br and C−I, respectively, in which the
24
stability order is: I < Br < Cl < I . This order represents the opposite trend found in
this study. We suppose that the order found for triazene Nꢀoxides is a result of two
mutual factors: (i) halogen electronegativity and (ii) atomic radius. An increase in
decomposition temperatures of the compounds as the halogens atomic radius increase
and electronegativity decrease can be highlighted. Unexpectedly, this indicates that
triazene Nꢀoxides substituted in the para position by hard substituents (lower atomic
radius and greater electronegativity, i.e. F atom) decomposed in lower temperatures
than the same compounds with soft substituents in the same position (greater atomic

CrystEngComm
Page 22 of 35
View Article Online
DOI: 10.1039/C7CE02015E
radius and lower electronegativity, i.e. I atom) that possess a more polarizable
electronic cloud.
Compounds were submitted to DSC analysis in order to determine their thermal
events (e.g. melting and crystallization temperatures). Each sample was subjected to
three cycles of heating and cooling in a safe temperature range varied according to
the decomposition temperature of the compounds present in the sample. This was
done to avoid decomposition within the DSC equipment. However, no thermal events
were observed for 1ꢀ4 in all cycles within the temperature ranges evaluated (see ESI†;
Figures S16ꢀS19). This leads us to believe that the melting point of the compounds is
likely very close to the point at which they decompose. This assumption is in
accordance with the study reported by Serebryanskaya et al.25. The authors observed
thermal decompositions for 1,3ꢀBis(2ꢀalkyltetrazolꢀ5ꢀyl)triazenes and their Fe(II),
Co(II), and Ni(II) complexes using DSC, in addition to finding that the triazene
complexes decomposed without melting at temperatures above 240 °C.
UVꢀVis
The absorption of electromagnetic radiation of the ultravioletꢀvisible by molecules,
atoms, or ions is usually associated with electronic transitions. The development of
semiꢀconductor materials with optical properties can result in optical electronic
devices, such as light emitting diode (LED) displays, applications in optical
communication, and biological labeling7,8. Therefore, the optical behavior of
compounds 1ꢀ4 in ethanol were evaluated from spectroscopy of molecular absorption
in the UVꢀVisible region (in a range of 200 to 600 nm). All the compounds show
three distinct peaks of absorbance in a range of 190 to 400 nm, which corresponds to
the UV region (Figure 11(a)). These peaks of absorbance occur when an electron is
promoted from a low energy state to a highꢀenergy electronic state by the interaction
of the absorbent medium with electromagnetic radiation. Figure 11(a) demonstrates
absorbance as a function of wavelength (λ) in different concentrations of compound 1,
which presents three peaks at 200, 234, and 348 nm (black arrows). From this data, an
absorbance versus concentration plot can be constructed for each absorption peak.
Figure 11(b) demonstrates this plot for compound 1 at 200 nm.

Page 23 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
(a)
(b)
Figure 11. (a) Absorbance versus wavelength (λ) and (b) absorbance versus
concentration for 1 at 200 nm in ethanol at 25 °C.
By using the LambertꢀBeer law, the angular coefficient of the straight line observed in
Figure 11(b) corresponds to the molar absorptivity (ε) of compound 1 in ethanol at
200 nm. The absorbance versus λ for 2ꢀ4 and the absorbance versus concentration
plots for 1 at 238 and 348 nm and 2ꢀ4 in all λ of absorption are depicted in SI. The
straight lines of all experimental data were obtained with an excellent determination
coefficient. Molar absorptivity (ε) measures the ability of a mole of substance to
attenuate a light incident on a given wavelength. It is an intrinsic property of
substances and provides an estimation of how strongly a substance absorbs radiation
of a given frequency. Table 4 shows the λ1, λ2, and λ3 of absorption peaks and the
molar absorptivity (ε) in each λ for compounds 1ꢀ4.
Table 4. Wavelengths (λ1, λ2, λ3) and molar absorptivity (ε) in each λ for compounds
1ꢀ4 in ethanol.
λ1
λ2
λ3
ε1
ε2
ε3
Compound
(nm) (nm) (nm) (L mol ꢀ1 cmꢀ1) (L mol ꢀ1 cmꢀ1) (L mol ꢀ1 cmꢀ1)
348 238 200
360 240 200
352 238 200
360 241 203
23098
12562
21238
25083
15289
6424
26757
11531
23963
21899
1
2
3
4
12257
12874
a
ε1, ε2 and ε3 correspond to molar absorptivity in λ1, λ2, λ3, respectively.
π*, n → σ* and π → π* electron
From Table 4, λ1, λ2, and λ3 were assigned to n →
transitions, respectively. The triazene chemical groups probably involved in these

CrystEngComm
Page 24 of 35
View Article Online
DOI: 10.1039/C7CE02015E
transitions are N=N, −N=N−N, and C=C, respectively. The absorption transition of
these triazene1ꢀoxide are in agreement with data previously reported in the literature
for triazene structures without coordinated oxygen26,27. Furthermore, the ε of 1ꢀ4
demonstrates that all compounds possess a strong absorbance capacity, which shows
promising optical properties.
Domingues et al.26 reported an orthoꢀsubstituted triazene with a fluorine atom
possessing greater ε value than the orthoꢀsubstituted analogue compound with a
chlorine atom. In this study, compound 1 with an aromatic ring pꢀF substituted shows
significantly larger ε1, ε2, and ε3 values than the pꢀCl substituted analogue compound
(2). Nevertheless, the ε1, ε2, and ε3 values for compounds 2 and 3 were similar to
values found for compound 1. This indicates that the electron transitions associated
with the absorbance phenomena in triazene Nꢀoxides is not highly dependent on
electronic effects caused by different halogens bond in the aromatic ring.
Conclusions
Triazene Nꢀoxide were successfully used as a model to contribute to the
understanding of crystallization mechanisms and the importance of certain
interactions, such as halogen···π interactions to crystal growth. Proposals of
crystallization mechanisms were presented for the series of molecules using
topological and energetic data obtained from the supramolecular cluster. Two new
parameters were presented to interpret the crystal formation, the CG% and the
NG/NC. It was possible to observe that compounds 1ꢀ4 (X = halogen) presented
contributions of over 50% in the first step. Compound 5 (X = H), which formed a
strong first dimer in the first step, only reached 50% of contribution when considering
the first and second step. The NG/NC parameter enabled us to observe the transition
between the dominant parameters during the crystallization process. All compounds
presented similar behavior during the steps, i.e. first a dominance of the stabilization
energy of the interactions and the final step of the crystallization process relying
almost exclusively on a topological process. With a broad analysis, it was possible to
assure the existence of Xꢁꢁꢁπ interactions and assess their participation in crystal
formation, showing that the formation of this kind of interaction in these molecules is
a consequence of a topological process. UVꢀVis and thermal analysis (TGA and DSC)
were also carried out to observe the special properties of these molecules. In addition

Page 25 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
to halogens influencing the decomposition pattern, it was also possible to assess that
these structures are stable (mass loss at T > 129 °C) and decompose in two main
stages. An increase in the decomposition temperature of the compounds as the atomic
radius of the halogens increase and electronegativity decrease can be highlighted. No
thermal events (e.g. melting and/or crystallization temperatures) were observed for
the compounds in the temperature range evaluated (below decomposition
temperatures). Furthermore, the ε of the halogenated compounds showed that all
compounds possess strong absorbance capacity, indicate promising optical properties.
Experimental
Materials and Methods
The infrared absorption spectra were performed as KBr pellets using 3 to 5 mg of
sample per 100 mg of KBr in the Bruker IR spectrophotometer. UV/Visible spectra
were recorded using a Shimadzu UVꢀ2600 spectrophotometer. For this analysis, a
standard solution with concentration of 0.5 mM was made for compounds 1ꢀ4. From
these standards solutions, 5.00 ꢂL volumes were extracted and diluted in the cuvette
containing 3 mL of ethanol. Samples were analyzed at 25 °C. The spectra were
recorded within a range of 190ꢀ600 nm at a scanning rate of 600 nm minꢀ1 (0.1 s
1
13
integration per 1.0 nm). The H and C NMR spectra were obtained on Bruker DPX
(600MHz) spectrometers using DMSOꢀd6 as solvent. Raman spectra were performed
on Bruker MultiRAM devices with laser at 1064 nm wavelength and exposure time of
3 s. Thermal analysis were performed on a TGA Q5000 (TA Instruments Inc., USA).
The heating rate used was 10 °C.minꢀ1 and inert atmosphere of N2 (50 ml.minꢀ1).
Elemental analysis technique was performed using a Perkin ꢀ Elmer CHN 2400. DSC
experiments were performed using a MDSC Q2000 (Tꢀzero™ DSC Technology, TA
Instruments Inc., USA). Dry high purity (99.999%) nitrogen gas was used as the
purge gas (50 mL min−1). The instrument was initially calibrated in standard MDSC
mode using the extrapolated onset temperatures of melting indium (156.60°C) at a
heating rate of 10°C min−1 and the heat from the fusion of indium (28.71 J g−1). Heat
capacity calibration was carried out by running standard sapphire (IJαꢀAl2O3)
measurement at the experimental temperature. The heating rate was 5 °C minꢀ1. The
instrument was initially calibrated in standard DSC mode with Indium (99.99%). The

CrystEngComm
Page 26 of 35
View Article Online
DOI: 10.1039/C7CE02015E
masses of the reference pots and lids and the samples weighed 51 ± 0.02 mg. The
samples were enclosed in aluminum pots with lids. The masses of the samples were
weighed on a Sartorius scale (M500P) with an accuracy of (± 0.001 mg). Data were
treated using TA Universal Analysis 2000 Software, version 4.5 (TA Instruments Inc.,
USA).
Synthesis
Solvents and reagents were used without prior treatment. The triazene N1ꢀoxide
compounds were synthesized starting from the diazotization reaction of the respective
4ꢀhalophenylamine in acidic medium with sodium nitrite (1:1) at 0 °C to obtain the
respective diazonium salt and, subsequent coupling with βꢀphenylhydroxylamine,
previously synthetized28, according to the methodology described in29 (Scheme 1).
Scheme 1. Synthesis path for triazene N1ꢀoxide compounds 1ꢀ4.
Synthesis of 1ꢀ(4ꢀfluorphenyl)ꢀ3ꢀphenyltriazenide N1ꢀoxide (1)
For the synthesis of 1, a solution of NaNO2 (1.25 g, 17.9 mmol) in 20 mL of water
was added to a solution of 4ꢀfluoraniline (2.0 g, 17.9 mmol) in 15 mL of a mixture of
concentrated HCl and water (1:1) cooled to 0ºC, thus forming the diazonium salt.
After 20 minutes, the βꢀphenylhydroxylamine solution (2.2 g, 17.9 mmol) in 15 mL
of concentrated acetic acid was added and the reaction mixture stirred for 40 minutes
at 0 °C. The reaction mixture was carefully pHꢀadjusted to about 5 and the yellow
opaque precipitate was vacuum filtered and recrystallized from a fraction of
ethanol/pyridine solvents, which after approximately five days of slow evaporation
resulted in yellow crystals in the form of needles. Compound 1 C12H10FN3O shows a
yield of 75%. Elementar analysis % C12H10FN3O: Calculated: C 63.40; H 5.32; N
17.06; Experimental: C 62.27; H 4.36; N 18.19. IR (KBr): ν (NꢀH) 3132.26; ν (CꢀH)
3055; ν (C=C) 1604.71; δ (NꢀH) 1515.99; ν (N=N) 1429.19; ν (NꢀNꢀN) 1332.76; ν (N

Page 27 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
ꢀ
O) 1298.13; ν (CarꢀHalog) 1220.25; ν (NꢀN) 1195.40 cmꢀ1. Raman: ν (NꢀH) 3080; ν
(CꢀH) 3050; ν (C=C) 1606; δ (NꢀH) 1003; ν (N=N) 1431; ν (CꢀC) 1401; ν (N=NꢀN)ꢀ
ring 100; ν (CarꢀHalog) 381; ν (CꢀN) 1204 cmꢀ1. NMR 1H (400 MHz DMSO – d6) δ =
12.04 (s, 1H, OH); δ = 8.09 (d, J=8.8Hz, 2H, ArH); δ = 7.60ꢀ5.53 (m, 5H, ArH); δ =
7.22 ppm (t, J=8.1Hz, 2H, ArH); NMR 13C (400 MHz DMSO – d6); δ = 159.9 (d, 1JCꢀ
4
F=238.9Hz, C4); δ = 142.8 (C7); δ =137.1 (C1 JCꢀF = 2.1Hz,); δ = 129.5 (C10); δ =
129.1 (C9, C11); δ = 119.5 (C8, C12); δ = 116.4 (d, 3JCꢀF = 7.9Hz, C2 , C6); δ =115.8 ppm
(d, 2JCꢀF = 22.7Hz, C3,C5). UV/Vis: λ = 360 nm, attributed n
15008 L mol ꢀ1.cmꢀ1; λ = 276 nm, attributed n
σ*, group (ꢀN=NꢀN), ε276 = 28153 L
mol ꢀ1.cmꢀ1; λ = 238 nm, attributed π π*, group (C=C), ε238 = 32551 L mol ꢀ1 cmꢀ1.
ꢀ π*, group (NꢀN), ε360 =
ꢀ
ꢀ
Synthesis of 1ꢀ(4ꢀchlorophenyl)ꢀ3ꢀphenyltriazenide N1ꢀoxide (2)
Compound 2 was synthesized according to the procedure described for compound 1,
but using 4ꢀchloroaniline (2.0 g; 15.74 mmol) as the primary amine. Compound 2
C12H10ClN3O shows a yield of 83%. Elementar Analysis % C12H10ClN3O: Calculated:
C 58.19; H 4.07; N 16.97; Experimental: C 58.34; H 4.27; N 16.64. IR (KBr): ν (NꢀH)
3186.07; ν (CꢀH) 3100; ν (C=C) 1598.74; δ (NꢀH) 1509.44; ν (N=N) 1418.37; ν (Nꢀ
NꢀN) 1330.07; ν (N
ꢀ
O) 1296.74; ν (CarꢀHalog) 636.90; ν (NꢀN) 1168.33 cmꢀ1.
Raman: ν (NꢀH) 3078; ν (CꢀH) 3070; ν (C=C) 1598; δ (NꢀH) 1005; ν (N=N) 1420; ν
(CꢀC) 1396 cmꢀ1; ν (N=NꢀN)ꢀring 93; ν (CarꢀHalog) 410; ν (CꢀN) 1213 cmꢀ1. NMR 1H
(600 MHz DMSO – d6) δ = 12.13 (s, 1H, OH); δ = 8.09 (d, J = 7.8 Hz, 2H, ArH); δ =
7.60 – 7.51 (m, 5H, ArH); δ = 7.41 ppm (d, J = 8.6 Hz, 2H, ArH); NMR 13C (600
MHz DMSO – d6); δ = 142.86 (C1); δ = 139.50 (C7); δ = 129.82 (C4); δ = 129.17 (C3,
C5); δ= 129.08 (C9, C11); δ = 126.13 (C10); δ = 119.65 (C2, C6); δ = 116.49 ppm (C8,
C12). UV/Vis: λ= 360 nm, attributed n
λ = 240 nm, attributed n
σ*, group (ꢀN = NꢀN), ε240 = 6424 L mol ꢀ1 cmꢀ1; λ = 200
nm, attributed π
π*, group (C=C), ε200 = 11531 Lmol ꢀ1 cmꢀ1.
ꢀ
π*, group (N=N), ε360 = 12562 Lmol ꢀ1 cmꢀ1;
ꢀ
ꢀ
Synthesis of 1ꢀ(4ꢀbromophenyl)ꢀ3ꢀphenyltriazenide N1ꢀoxide (3)
By repeating the same preparation used for compound 1, but substituting 4ꢀ
fluorophenyl for 4ꢀbromophenyl (2 g; 11.76 mmol) for compound 3. Compound 3
C12H10BrN3O had a yield of 80%. Elementar Analysis % C12H10BrN3O: Calculated: C

CrystEngComm
Page 28 of 35
View Article Online
DOI: 10.1039/C7CE02015E
49.34; H 3.45; N 14.38; Experimental: C 49.58; H 3.48; N 13.98. IR (KBr): ν (NꢀH)
3188.96; ν (CꢀH) 3090; ν (C=C) 1584.85; δ (NꢀH) 1505.15; ν (N=N) 1414.72; ν (Nꢀ
NꢀN) 1329.03; ν (N
ꢀ
O)1296.70; ν (CarꢀHalog) 609.23; ν (NꢀN) 1169.23 cmꢀ1.
Raman: ν (NꢀH)3075; ν (CꢀH)3069;ν (C=C) 1596;δ (NꢀH)1006;ν (N=N)1417;ν (Cꢀ
C)1394;ν (N=NꢀN)ꢀring92; ν (CarꢀHalog)553; ν (CꢀN)1212 cmꢀ1. NMR 1H (600 MHz
DMSO – d6) δ = 12.13 (s, 1H, OH); δ = 8.09 (d, J = 7.4 Hz, 2H, ArH); δ = 7.54 ppm
(ddd, J = 41.4, 23.7, 7.8 Hz, 7H, ArH); NMR 13C (600 MHz DMSO – d6); δ = 142.85
(C1); δ = 139.88 (C7); δ = 131.94 (C3, C5); δ = 129.84 (C10); δ = 129.18 (C9, C11); δ =
119.65 (C2, C6); δ = 116.91 (C8, C12); δ = 114.05 ppm (C4). UV/Vis: λ = 352 nm, a
attributed n
ꢀ
π*, group (N=N), ε352 = 21238 Lmol ꢀ1 cmꢀ1; λ = 238 nm, attributed n
ꢀ
σ*, group (ꢀN=NꢀN), ε238 = 12257 Lmol ꢀ1 cmꢀ1; λ = 200 nm, attributed π
ꢀ
π*,
group (C=C), ε200 = 23963 Lmol ꢀ1 cmꢀ1.
Synthesis of 1ꢀ(4ꢀiodophenyl)ꢀ3ꢀphenyltriazenide N1ꢀoxide (4)
Compound 4 was obtained by beginning with 4ꢀiodophenyl (2 g; 9.13 mmol) as the
primary amine and following the same procedure used for the synthesis of compound
1. Compound 4 C12H10IN3O shows a yield of 68%. Elementary Analysis %
C12H10IN3O: Calculated: C 42.50; H 2.97; N 12.39; Experimental: C 40.71; H 3.08; N
11.51. IR (KBr): ν (NꢀH) 3068.15; ν (CꢀH) 3050; ν (C=C) 1588.52; δ (NꢀH) 1500.44;
ν (N=N) 1417.09 cmꢀ1; ν (NꢀNꢀN) 1331.55; ν (N
ꢀ O) 1298.04; ν (CarꢀHalog) 611.30;
ν (NꢀN) 1086.98 cmꢀ1. Raman: ν (NꢀH) 3072; ν (CꢀH) 3072; ν (C=C) 1592; δ (NꢀH)
1003; ν (N=N) 1420; ν (CꢀC) 1394; ν (N=NꢀN)ꢀring 100ꢀ77 ; ν (CarꢀHalog) 1060; ν
(CꢀN) 1210 cmꢀ1. NMR 1H (600 MHz DMSO – d6) δ = 12.10 (s, 1H, OH); δ = 8.09 (d,
J =7.5Hz, 2H, ArH); δ= 7.68 (d, J = 8.2 Hz, 2H, ArH); δ = 7.61 – 7.52 (m, 3H, ArH);
δ = 7.35 ppm (d, J = 8.1 Hz, 2H, ArH); NMR 13C (600 MHz DMSO – d6); δ = 142.84
(C1); δ = 140.33 (C7); δ = 137.70 (C3, C5); δ = 129.82 (C10); δ = 129.15 (C9, C11); δ =
119.64 (C2, C6); δ = 117.24 (C8,C12), δ = 85.56 ppm (C4). UV/Vis: λ = 360 nm,
attributed n
σ*, group (ꢀN=NꢀN), ε241 = 12874 Lmol ꢀ1 cmꢀ1; λ = 203 nm, attributed π
group (C=C), ε203 = 21899 Lmol ꢀ1 cmꢀ1.
ꢀ
π*, group (N=N), ε360 = 25083 L mol ꢀ1 cmꢀ1; λ = 241 nm, attributed n
π*,
ꢀ
ꢀ

Page 29 of 35
CrystEngComm
View Article Online
DOI: 10.1039/C7CE02015E
Spectroscopic analysis
Absorptions observed in the infrared spectra of compounds 1ꢀ4 at 3132 cmꢀ1, 3186,
3189, and 3068 cmꢀ1, respectively, correspond to the
as(NꢀH) stretches of the
respective diazo(oxide)amine [ꢀN=N( O)ꢀN(H)ꢀ] chain of each compound, which is
ν
→
in accordance with what was found at 3140 cmꢀ1 for compound 51. In addition, the
bands observed at 1298, 1296, 1297, and 1298 cmꢀ1, respectively, correspond to the
characteristic stretches of the νN(→O) dative bond in accordance with what was
found at 1339 cmꢀ1 in compound 51 and demonstrate that the compounds show the
tautomeric form of the triazene N1ꢀoxide in the solid state. In the Raman spectra of
compounds 1ꢀ4, the observed bands at 3080, 3078, 3075, and 3072 cmꢀ1 were
attributed to the characteristic
hand, the vibrations observed at 381, 410, 553, and 1060 cmꢀ1 for compounds 1, 2, 3,
and 4, were attributed to the aromatic carbonꢀhalogen [ s(CarꢀX); X = F, Cl, Br, and
νs(NꢀH) stretches in triazene compounds. On the other
ν
I], respectively30,31. In the 13C NMR spectrum, the “C4” carbons bond to the
respective halogens for compound 1, 2, 3, and 4 correspond to the peak with a
1
chemical shift at δ 159.9 (d, JCꢀF=238,9Hz), δ 129.31, δ 114.58, and δ 85.06,
respectively. It is possible to highlight the decreasing disbanding effect as it decreases
along group 17 in the periodic table, an expected effect due to the electronegativity (Iꢀ
inductive effect) decreasing from fluorine to iodine and also because of the effect of
the heavy atom that increases from fluorine to iodine. Electronic absorption spectra of
compounds 1, 2, 3, and 4 were recorded in ethanol, which exhibits a strong absorption
at 348 nm (1) and 352 nm (2, 3, and 4) (Figs. S28ꢀS31). When compared with usual
electronic absorption spectra of classical triazene derivatives [RN=NꢀN(H)R], these
absorptions observed at the region of 348ꢀ352 nm where assigned to the partly
forbidden (n→π*) transition and attributed to the (N=N) chromophore group.32
Additionally, medium intensity absorptions were observed at 234 nm (1), 240 nm (2),
238 nm (3), and 241 nm for (4). These absorptions at 234ꢀ241 nm were assigned to
the (n→σ*) transition, also in analogy of that attributed to the (N=N) chromophore
group when compared with the triazenes. 32 A third strong band is observed at 200 nm
for all compounds that were assigned to the expected and characteristic (π→π*)
transition of the phenyl and pꢀhalophenyl substituents.