Green Hydrothermal Synthesis of 5‑Phenyl-1H-tetrazole through Cycloaddition [3 + 2] Using Co(II) as Catalyst: Statistical Optimization by Response Surface Methodology (RSM) and DFT Calculations

Abstract

5-Phenyl-1H-tetrazole (5-PHTA) and its derivatives exhibit relevant properties in pharmacology, corrosion, and coordination chemistry. In this work, a new methodology was proposed to obtain 5-PHTA under hydrothermal conditions using the [3 + 2] cycloaddition reaction between C₆H₅CN and NaN₃, with Co­(CH₃COO)₂·4H₂O as the catalyst and water as the solvent. This new method is an ecological and economical alternative that utilizes water instead of traditional organic solvents, such as DMSO, DMF, ethyl acetate, and n-hexane, for the synthesis and purification of 5-PHTA, reducing the E-factor and carbon footprint. The yield of the reaction was optimized by applying the statistical technique RSM on variables A (molar ratio C₆H₅CN/NaN₃), B (molar ratio Co²⁺/NaN₃), and C (reaction time), where the coordinates A = 2.09, B = 0.48, and C = 6.29 h produced a maximum yield of 81%. 5-PHTA structure was confirmed by single-crystal X-ray diffraction, ¹H-NMR, elemental analysis, FT-IR, and UV–vis. DFT calculations (B3LYP/def2-TZVP) were performed at 433.150 K and 1.45 atm, using water as the reaction solvent. The results demonstrate the crucial role of Co­(II) as a catalyst, via the formation of intermediates that are both kinetically and thermodynamically favored relative to the tested free-catalyst systems.

Introduction

5-Aryltetrazoles have been widely studied due to their relevant properties in various fields of application, particularly in pharmacology, where compounds such as Losartan, Valsartan, and Irbesartan are notable for their hypotensive properties, or Tazanoplast, used as an antihistaminic agent (Figure ). Moreover, the presence of the tetrazole ring gives 5-aryltetrazoles the capability of acting as anticorrosive agents on metal surfaces in both saline and acidic conditions , because they generate a passivating layer resulting from the interaction between the metal and tetrazole nitrogen that protects the metallic surface from corrosion (Figure ). On the other hand, these heterocycles also act as polydentate ligands in the synthesis of coordination complexes in their neutral (depending if they are in their 1H or 2H tautomeric conformation) and anionic form, having in the first case binuclear (μ-2,3; μ-2,4; μ-3,4; μ-1,3; μ-1,4) and trinuclear (μ-2,3,4; μ-1,3,4) coordination modes, while for tetrazolates, coordination could be binuclear (μ-1,2; μ-2,3; μ-1,3), trinuclear (μ-1,2,3; μ-1,2,4), and even tetranuclear (μ-1,2,3,4) due to the symmetry generated by the elimination of the acidic hydrogen , (Figure ).

1.

Examples of 5-aryltetrazoles with pharmacological properties.

2.

Passivation of metallic surfaces by the action of 5-phenyl-1H-tetrazole and 5-phenyl-2H-tetrazole.

3.

Possible polydentate coordination modes of 5-phenyl-1H-tetrazole, 5-phenyl-2H-tetrazole, and 5-phenyltetrazolate.

The set of applications described above encourages the search for new alternatives for the synthesis of this type of compound that are not found in nature. 5-Phenyl-1H-tetrazole (5-PHTA) is the basic unit structure of more complex 5-aryltetrazoles, and it is studied by its own properties as well as to predict the behavior of its substituted derivatives. The most used reaction to obtain it is the cycloaddition [3 + 2] or 1,3-dipolar cycloaddition between aromatic nitriles and azides, which act as dipolarophiles and dipoles, respectively. In general terms, the cycloaddition process begins with the activation of C₆H₅CN through a catalyst, whether metallic or nonmetallic, which polarizes the −CN bond. This polarization creates a reactive site that facilitates interaction with NaN₃, leading to the formation of the heterocycle.

In the case of nonmetallic catalysts, they reported SiO₂–H₂SO₄, cuttlebone, and SO₃H-carbon, which allowed the formation of the tetrazole ring of 5-PHTA. They activated the nitrogen atom from the nitrile through a hydrogen bond produced on the catalyst’s surface.

On the other hand, metallic catalysts activate the −CN bond by the subtraction of electronic density through the coordination bond M–NC. These catalysts are the most utilized in this synthetic route. They can be found in different forms, including γ-Fe₂O₃NPs, PtNPs-oxidized graphene, AgNPs, AuNPs, CuONPs/aluminosilicate, PtNPs anchored to carbon or ZnONPs over reduced graphene oxide, modified metallic zeolites, coordination complexes, , and inorganic salts like AgNO₃, Yb­(OTf)₃.·H₂O, ZrOCl₂·8H₂O, CuSO₄·5H₂O, ZnBr₂, PbCl₂, and CuCl.

Specifically, cobalt-based catalysts have recently attracted considerable attention in the current research. Co­(II) dispersed on Zeolite Y (CoY), Cobalt­(II) complex with a tetradentate ligand N,N-bis­(pyridin-2-ylmethyl)­quinolin-8-amine, a complex of cobalt stabilized on the surface of modified boehmite nanoparticles (Co-(PYT)­2@BNPs), Cobalt­(II) complex on nanodiamond-grafted polyethylenimine@Folic Acid, cobalt–nickel on magnetic mesoporous hollow spheres, and Ce₃O₄ nanoparticles are the recently reported cobalt-containing catalysts that have achieved 5-PHTA high reaction yields.

In addition to the catalyst, other critical factors such as temperature and reaction time were reported with a broad range of values, ranging from 75 °C to 153 °C for the temperature of reaction, while for time, reactions worked well from minutes to days, depending on each system.

Finally, the solvent choice and purification strategy substantially affect the environmental profile of the synthesis. Dimethylformamide (DMF) and dimethyl sulfoxide (DMSO) are commonly used to prepare 5-phenyl-1H-tetrazole because they dissolve both benzonitrile (C₆H₅CN) and sodium azide (NaN₃) completely and tolerate the high temperatures required for this transformation. Because the product is also soluble in these high-boiling solvents, isolation typically requires additional steps, for example, liquid–liquid extraction with ethyl acetate and n-hexane followed by recrystallization or chromatographic purification, which increases solvent use and waste. These downstream requirements should be considered when assessing the overall greenness of the procedure.

Some catalyst works in water as the solvent for this synthesis: ZnBr₂ (50% catalyst, 24 h, reflux, 76% yield), FeTSPP (5% catalyst, 20 min, 60 °C, 97% yield), TBAHS (25% catalyst, 6 h, 85 °C, 87% yield), and steelmaking slag (mixture of many oxides) (50 mg, 1 h, RT, 89% yield). A benefit of water as a solvent in this system is the mitigation of explosion hazards from azides, due to its high heat capacity. It is important to note that an aqueous sodium azide solution is very stable at reflux.

Theoretically, the DFT theory was utilized to determine the role of several catalysts in the cycloaddition of nitriles and azides. The reduction of energy barriers to produce the tetrazole ring demonstrated that the presence of catalysts such as ZnBr₂, AlCl₃, dialkyltin oxide-trimethylsilyl azide, and ammonium is pivotal for obtaining those tetrazole products.

To optimize the yield of reaction, it was considered to use the response surface methodology (RSM), which is a statistical technique recently applied to the organic synthesis field with relevant examples of achievements as it was demonstrated in the case of the alkyl oxidation of valencene to forming selectively benzyloxy valencene using the Kharasch–Sosnovsky reaction or with the Friedel–Crafts acylation of phloroglucinol where it was utilized a natural zeolite as catalyst. In comparison to the traditional method of optimizing reaction efficiencies, well known as OVAT (one variable at a time), the RSM provides quantitative information about the influence of each variable on the response, and it brings a reasonable description of the path that researchers must follow to achieve optimization.

Knowing the pros and drawbacks of the reported methodologies, this work proposes a greener method for 5-phenyl-1H-tetrazole synthesis, using Co­(CH₃COO)₂·4H₂O as a catalyst and water as a solvent under hydrothermal conditions. Response surface methodology (RSM) was used to optimize the yield of the reaction. Atom economy, yield, and E-factor are calculated as green chemistry metrics. Moreover, density functional theory (DFT) was applied to obtain information about the reaction mechanism.

Materials and Methods

Materials

All of the purchased reagents (C₆H₅CN, NaN₃, Co­(CH₃COO)₂·4H₂O, and HCl (cc)) were used without further purification. Deionized water was used as the solvent for the reaction and purification.

General Experimental Procedure

Defined amounts of C₆H₅CN, NaN₃, and Co­(CH₃COO)₂·4H₂O (from T1 to T37) with 2 mL of deionized water were poured into a Pyrex tube (ø 1.3 cm, 10 cm high) reactor and mixed for homogenization (pink mixture). Then, the tube reactor was closed and heated at 160 °C utilizing the Spectroquant TR 420 thermoreactor. When reactions ended, the color turned violet, and the tube was allowed to cool at room temperature; 20 mL of cold HCl 4N was added to it. The acidified mixture was transferred into a beaker and stirred for 30 min before being filtered. White needles of 5-phenyl-1H-tetrazole were obtained after recrystallization in water. Scheme describes the hydrothermal reaction performed. R _f: 0.85 (silica gel; absolute ethanol). UV (Heλios ϒ) (EtOH) = 207.5, 241 nm. FT-IR (Nicolet iS10) v in cm^–1 (KBr) = Benzene CH (3129, 3077, 3055), NH stretching (2981, 2915, 2837, 2701, 2655, 2610, 2548, 2483, 2458), (CN) tetrazole stretching 1486. ¹H NMR (NMR Bruker Advance 500 MHz, DMSO-d ₆): 16.90 (br, 1H, N–H), 8.05–8.07 (m, 2H), 7.62–7.63 (m, 3H) ppm. Elemental analysis (Leco CHNS-932 analyzer): Experimental: C 57.77%, H 4.10%, N 38.12%. Calculated: C 57.53%, H 4.14%, N 38.33%.

1. Hydrothermal Synthesis of 5-Phenyl-1H-tetrazole.

Full characterization of 5-phenyl-1H-tetrazole includes crystallographic studies (Rigaku Smart Lab X-ray diffractometer). Cu Kα radiation was used at 40 kV and 25 mA in a scanning range of 10–80° (2θ) at a rate of 8° min^–1.

Statistical Analysis and Optimization

Data collection was fitted to mathematical models to correctly obtain regression equations and optimize coordinates using Minitab 18 software. Indicators used to verify the fit of models were p _value, R ², and R _adjust ².

DFT Studies of the Intrinsic Reaction Coordinate (IRC)

Theoretical calculations were computed under the same conditions of experimental assays using density functional theory (DFT) to complete the experimental evidence of the reaction during the cyclization process in the presence of Co­(II) ion as a catalyst. The geometries of the three representative fragments selected as a template for the reagents, the transition state, and the product were fully optimized at the B3LYP/def2-TZVP level. − Grimme’s D3 approximation and the SMD solvation model , were included with the functional to account for long-range interactions and to achieve solvent effects, respectively (see Cartesian Coordinates in Supporting Information Part 3). The intrinsic reaction coordinate (IRC) was employed at the same level of theory to ascertain the energy profile for the formation of the cobalt­(II) complexes of 5-phenyl-1H-tetrazole, ensuring that the transition state accurately corresponds to the respective reactants and products. All calculations were performed at 433.150 K and 1.45 atm using the Gaussian 16 package.

Results and Discussion

Synthesis

In all of the experiments performed, varying the proportions of the reagents, a rapid color change was always observed, from pink to violet, with the formation of a precipitate. This was probably a cobalt complex that could not be identified. This indicates that the reaction proceeded in a heterogeneous medium despite the high temperature (160 °C).

After the reaction system was cooled and the medium acidified, 5-PHTA is released but precipitated in the aqueous medium. Purification was easily achieved after filtration and recrystallization from water. The purity of 5-PHTA was high enough. No signal of any impurity was detected by ¹H-NMR. Efforts are underway to extend this methodology to the preparation of 5-phenyl-1H-tetrazole derivatives incorporating substituents with varied electronic properties.

Optimization of Reaction Efficiency by RSM

Preliminary Tests

After performing preliminary tests, it was found that although this cycloaddition [3 + 2] must follow the stoichiometric rate 1:1 between the reagents, NaN₃ must act as the limiting reagent. These results agree with a previous report in which 5-phenyl-1H-tetrazole is obtained, together with its respective amide, by the collateral reaction of hydration (Supporting Information Part 1).

Afterward, it was considered a first-order model, with 4 central points to obtain a factorial design 2⁴ with 4 replicas. The ANOVA results of this model reveal that terms A–C and their second- and third-order interactions exceed the significance threshold because of their p _value ≤ 0.05, which means that they significantly influence the yield of reaction (Y%). At the same time, no curvature was observed (p _value = 0.624). Regarding the most significant factors that affect the efficiency of reaction, it was demonstrated that the amount of catalyst, time, and their interaction of second order were the main factors responsible for the yield of this cycloaddition in this experimental zone. To find a curved zone that represents the maximum possible yield achievable for this system, we followed the steepest ascent methodology by a new first-order model adjusted to a new experimental zone. There, curvature was significant (p _value = 0.000), and the linear model did not fit, which means that new experiments were required to describe the region appropriately (see Supporting Information Part 1).

Second-Order Model

A new rotatable central composite model is defined in Table (real values), where α = ±1.68179 (coded values). The experimental results are presented in Table , which was constructed after performing 20 new experiments.

1. Experimental Region of the Second-Order Model (Real Values) to Optimize the 5-PHTA Yield.

independent factors	description	level −α	lower level	centre level	higher level	level +α
A	$\mathrm{m}\mathrm{olar}\mathrm{rate}\left(\frac{\mathrm{mol}{\mathrm{C}}_6{\mathrm{H}}_5\mathrm{CN}}{\mathrm{mol\; Na}{\mathrm{N}}_3}\right)$	0.69	1.0	1.5	2.0	2.34
B	$\mathrm{m}\mathrm{olar\; rate}\left(\frac{\mathrm{m}\mathrm{o}\mathrm{l}{\mathrm{C}\mathrm{o}}^{2+}}{\mathrm{m}\mathrm{o}\mathrm{l}\mathrm{N}\mathrm{a}{\mathrm{N}}_3}\right)$	0.2342	0.3554	0.5331	0.7108	0.8320
C	time (h)	4.32	5	6	7	7.68

2. Results of the Second 2³ Factorial Design with Replicates at the Central and Axial Points (Real Values) to Optimize the 5-PHTA Yield.

test/factor	$\mathrm{A}\left(\frac{\mathrm{mol}{\mathrm{C}}_6{\mathrm{H}}_5\mathrm{CN}}{\mathrm{mol\; Na}{\mathrm{N}}_3}\right)$	$\mathrm{B}\left(\frac{\mathrm{mol}{\mathrm{Co}}^{2+}}{\mathrm{mol\; Na}{\mathrm{N}}_3}\right)$	C (h)	Y (%)
T17	1.0	0.3554	5	14
T18	1.0	0.3554	7	20
T19	1.0	0.7108	5	46
T20	1.0	0.7108	7	44
T21	2.0	0.3554	5	73
T22	2.0	0.3554	7	75
T23	2.0	0.7108	5	71
T24	2.0	0.7108	7	69
T25	1.5	0.5331	6	74
T26	1.5	0.5331	6	71
T27	1.5	0.5331	6	73
T28	1.5	0.5331	6	73
T29	1.5	0.5331	6	74
T30	1.5	0.5331	6	71
T31	0.6591	0.5331	6	13
T32	2.3409	0.5331	6	86
T33	1.5	0.2342	6	52
T34	1.5	0.8320	6	70
T35	1.5	0.5331	4.3182	72
T36	1.5	0.5331	7.68	74

The statistical significance of the factors was studied by performing the ANOVA analysis of the second-order model, where the following hypotheses were postulated at 95% confidence (Table ):

$$H_0(\mathrm{null\; hypothesis})=\mathrm{f}\mathrm{actor\; is\; not\; significant\; for\; the\; regression}(p_{}\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}\mathrm{e}>0.05)$$

$$H_{\mathrm{a}}(\mathrm{alternative\; hypothesis})=\mathrm{f}\mathrm{actor\; is\; significant\; for\; the\; regression}(p_{}\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}\mathrm{e}\le 0.05)$$

3. ANOVA of Second Factorial Design 2³ with Central and Axial Points to Optimize the 5-PHTA Yield.

source	degree of freedom	sum of squares	mean square	P values	significance
model	8	8837.28	1104.66	0.000	yes
linear	3	6474.28	2158.09	0.000	yes
A	1	6021.70	6021.70	0.000	yes
B	1	448.61	448.61	0.001	yes
C	1	3.7	3.7	0.699	no
square	3	1833.00	611.41	0.000	yes
A2	1	1465.25	1465.25	0.000	yes
B2	1	521.83	521.83	0.001	yes
C2	1	45.40	45.40	0.206	no
2-term interactions	2	530.00	265.00	0.003	yes
AB	1	512.00	521.00	0.001	yes
BC	1	18.00	18.00	0.415	no
error	11	276.47	25.13
lack of fit	6	297.14	44.52	0.002	yes
pure error	5	9.33	1.87
total	19	9113.75

Table shows that the data fitted to the model significantly (p = 0.000), with square (p = 0.000) and linear components (p = 0.000). Moreover, it was demonstrated that factors associated with time have the least impact on the response in this experimental region, and they did not influence the efficiency of the reaction. On the other hand, factors related to the rate of C₆H₅CN/NaN₃ in the forms A and A² influenced dramatically the yield, followed by factors that depend on the concentration of the catalyst (with the terms B², AB, and B). This information can also be observed in the Pareto diagram illustrated in Figure , where the preponderance of each term on Y(%) is evidenced in proportion to the length of bars.

4.

Pareto chart of standardized effects for the second-order model (α = 0.05) to optimize the 5-PHTA yield. Only variables A and B, as well as their quadratic components, exceed the significance threshold (2.20).

At the same time, the equation fitted to the regression model (real values)

$$Y(\%)=-318.3+211.0\mathrm{A}+421.1\mathrm{B}+26.3\mathrm{C}-40.33{\mathrm{A}}^2-190.6{\mathrm{B}}^2-1.77{\mathrm{C}}^2-90.0\mathrm{AB}-8.44\mathrm{BC}$$ 1

presented correlation coefficients of R ² = 96.97% and R _adjust ² = 94.76%, which indicate a high fitting of the second-order model and the experimental data.

The obtained equation is constructed by the influence of 3 variables; however, to analyze it graphically, time (C) was fixed at 6 h, whereas molar rate C₆H₅CN/NaN₃ (A) and molar rate Co²⁺/NaN₃ (B) were plotted to predict their influence on the efficiency of reaction (Y%) in the contour plot and surface response graphs described in Figures and , respectively.

5.

Contour plot of 5-PHTA yield versus molar rate C₆H₅CN/NaN₃ and Co²⁺/NaN₃ when time is constant (6 h).

6.

Response surface of 5-PHTA yield versus molar rate C₆H₅CN/NaN₃ and Co²⁺/NaN₃ when time is constant (6 h).

Optimization

To find the coordinate that maximizes the efficiency of reaction, it is needed to solve eq by applying the following derivatives:

$$\frac{\partial Y\%}{\partial A}=\frac{\partial Y\%}{\partial B}=\frac{\partial Y\%}{\partial C}=0$$ 2

Utilizing Minitab 18 software, it was determined that the coordinates which maximize this function were: A _opt = 2.09, B _opt = 0.48, and C _opt = 6.29 h, where it was predicted to have a theoretical maximum efficiency of 84%. In contrast, experimentally, under this condition (T37), 81% was achieved (see Supporting Information Part 1).

Despite the theoretical stoichiometry of the cycloaddition being 1:1, the obtained value of A _opt (mol C₆H₅CN/mol NaN₃), in the optimized coordinates, was 2.09, which is approximately 2:1. A collateral reaction can explain this variation: the hydration of the nitriles. On the other hand, B _opt (mol Co²⁺/mol NaN₃) was practically 0.5, which can be associated with the neutralization of charges that must exist between cobalt cation (2+) and 2 azide anions (−1) during the formation of the intermediate. It means that we have a molar ratio of C₆H₅CN/NaN₃/Co­(II) of 1:0.5:0.25. This ratio contrasts with other systems. For example, a ratio of 1:1.5:0.1 was observed with (NH₄)₂Ce­(NO₃)₆ as a catalyst. The presence of the Co­(II) catalyst is crucial to the activation of the −CN group (yields without catalyst varied between 0% and 5%).

Evaluation of Green Chemistry Metrics

A deep comparison between the present work and two references that used water as a solvent of reaction to produce 5-phenyl-1H-tetrazole on a laboratory scale is summarized in Table . Further details are also available in Supporting Information Part 1.

4. Green Chemistry Metrics Comparison between the Current and Previous Synthesis of 5-PHTA.

green metrics	reference X	reference Y	this work	this work (recovering water)
atom economy (AE)	71%	71%	71%	71%
percentage yield (%Y)	87%	89%	81%	81%
E-factor (EF)	489	2720	369	85

Regarding atom economy (AE), all procedures followed the synthetic route [3 + 2] cycloaddition between NaN₃ and C₆H₅CN followed by acidification with HCl to obtain 5-phenyl-1H-tetrazole, thus, all of them got AE = 71% after the union of benzonitrile, the azide anion, and the hydrogen of the acid, excluding NaCl that is formed as a byproduct (Supporting Information Part 1). Yields (Y%) in all cases are roughly the same, ranging from 81% to 89%, indicating a maximum variation of 8% that is commonly observed in laboratory-scale reactions.

The primary differences between the procedures lie in the solvents used for purification and their respective amounts. While acetyl acetate and n-hexane were used to purify 5-PHTA, , our method employs water for recrystallization, thereby reducing the carbon footprint associated with those solvents.

It has been demonstrated that other methods require further purification with significant amounts of solvents, which contribute to increasing the E-Factor of the total process (see Supporting Information Part 1); nevertheless, our initial procedure itself is a greener methodology because of its relatively low E-factor (EF = 369) in comparison to other methods that are 1.33 and 7.37 times less ecological. Finally, an additional procedure was performed to recover the acid water after filtration; the pH of the recovered solution varied slightly from −0.3 to 0.15, conserving their acidic properties to be used in another acidification process. Additionally, water collected from recrystallization was also employed in a new recrystallization of 5-PHTA. Those improvements generated a much more environmentally friendly methodology (EF = 85) when compared with other procedures, 5.75 and 32.00 times less sustainable. A summary of the two proposed procedures from the optimized conditions found is illustrated in Scheme .

2. Optimized Conditions for the Synthesis of 5-PHTA and Summary of Green Chemistry Metrics.

Characterization

Crystal X-ray Diffraction

X-ray diffraction (XRD) analysis of 5-phenyl-1H-tetrazole was conducted on a colorless plate-like crystal obtained through the slow cooling technique from an aqueous solution. The compound is crystallized in an orthorhombic system, specifically within the Ama2 space group, CCDC 237392. The unit cell revealed Z = 4. Theoretical density of 1.428 g/cm³, parameters of cell (a, b, c, α, β, and γ), bond length, and bond angles were found according to the literature (Supporting Information Part 2). Based on crystallographic information, the tautomer 5-phenyl-1H-tetrazole predominates in the solid state.

Computational Predictions for the Formation of 5-Phenyl-1H-tetrazole

Three energy profiles were evaluated to identify the most feasible cycloaddition mechanism in solution. Based on previous suggestions, we considered only Co²⁺ and N₃ ^– as directly involved in the cyclization process with benzonitrile, while the counterions Na⁺ and acetate were assumed to be nonparticipatory. Coordinates of optimized structures involved in this study are found in Supporting Information Part 3.

The first proposed mechanism (neutral complex) of exergonic reaction at 433.15 K and 1.45 atm starts with the formation of a quartet metallic complex (R-CoN₃) from azide anions, cobalt­(II), and benzonitrile molecule (Ph–CN) with a change in the free energy of −28.7 kcal/mol (Figure ). Cyclization results from a rearrangement featuring significant angle distortion and a relative barrier of 31.1 kcal/mol starting from the formatted R-CoN₃ complex. The associated angle strains are 104.7° (C_nitrile–N_azide–N_azide) and 142.3° (N–N–N in TS-CoN₃) (Figure ). The ring-closing step is completed with the formation of complex P-CoN₃ and the reduction of the bond N–N–N azide angle from 142.3° to 110.3°.

7.

Energy profile for [3 + 2] cycloaddition reaction of neutral cobalt­(II) benzonitrile complex with N₃ ^–, computed at SMD-uB3LY/def2-TZVP level.

8.

Geometries of the stationary points involved in the cobalt­(II) 5-phenyltetrazolate-azide complex formation optimized at the B3LYP/def2-TZVP level.

On the contrary, both a higher energy barrier (ΔG ^± = 34.1 kcal/mol) and free reaction energy (Δ_rxn G = – 21.8 kcal/mol) were observed in the profile of cationic cobalt­(II) 5-phenyltetrazolate complex formation (P–Co⁺) (Figure ). These findings suggest a more feasible neutral P-CoN₃ complex formation at a stoichiometric rate of 1:2 of benzonitrile and ion azide. However, even though the theoretical result shows a better preference for this mole ratio, experimental evidence revealed that the highest yield was achieved in the system with a higher quantity of benzonitrile, possibly because of the formation of aromatic amides. Angular changes that led to cycloaddition are evidenced in Figure .

9.

Energy profile for [3 + 2] cycloaddition reaction of cationic cobalt­(II) benzonitrile complex with N₃ ^–, computed at the SMD-uB3LY/def2-TZVP level.

10.

Geometries of the stationary points involved cationic cobalt­(II) 5-phenyltetrazolate complex formation optimized at the B3LYP/def2-TZVP level.

As in the previous results, DFT calculation at the same environmental conditions and theory level in the formation of tetrazole anion by [3 + 2] Huisgen addition of azide anion with benzonitrile showed a higher energy barrier (ΔG ^± = 33.3 kcal/mol) and the highest free reaction energy (Δ_rxn G = – 12.0 kcal/mol) (Figure ). These results confirm that the process in aqueous solvent is less feasible in the absence of Co­(II), in agreement with our experimental approaches, even though a similar N–N–N azidic angle strain was observed in all cases (Figure ).

11.

Energy profile for anionic [3 + 2] cycloaddition reaction of benzonitrile with N₃ ^–, computed at the SMD-B3LY/def2-TZVP level.

12.

Geometries of the stationary points involved in the 5-phenyl-1H-tetrazole formation optimized at the B3LYP/def2-TZVP level.

Surprisingly, the theoretical calculations revealed lower energy barriers when DMF was used as the solvent, likely due to the stronger affinity of the organometallic intermediate for the organic medium. This phenomenon was demonstrated experimentally (Supporting Information Part 2), as significant yields were obtained under the optimized conditions in shorter reaction times simply by replacing water with DMF. Additionally, this interpretation is further supported by the observation that, in the absence of a catalyst, the computed reaction energies remain unchanged regardless of the solvent employed, as summarized in Table . The optimized coordinates of the species involved in the cycloaddition process in DMF under the studied conditions are provided in Supporting Information Part 3.

5. Theoretical Calculation of ΔG ^± and Δ_rxn G for Possible [3 + 2] Cycloaddition Mechanisms between Ion Azide and Benzonitrile in Water and DMF (B3LYP/def2-TZVP Level, Temperature = 433.150 K, Pressure = 1.45 atm).

	ΔG ± (kcal/mol)		Δrxn G (kcal/mol)
proposed mechanisms	aqueous medium	DMF medium	aqueous medium	DMF medium
neutral Co(II) complex	31.1	9.0	–28.7	–23.6
cationic Co(II) complex	34.1	10.5	–21.8	–43.4
uncatalyzed system	33.3	33.8	–12.0	–11.2

Conclusions

It was possible to achieve the desired hydrothermal synthesis of 5-phenyl-1H-tetrazole using water as a solvent and Co­(CH₃COO)₂·4H₂O as a new catalyst, which joins the list of metallic salts that promote the cycloaddition [3 + 2] between sodium azide and benzonitrile. Furthermore, by understanding the solubility properties of the resulting product, we implemented a greener purification method using only water (E-factor between 85 and 369), with high yields (81–84%). Statistical methodology RSM was supported mathematically to determine the influence of each one of the studied factors in specific experimental zones, finding that the presence of the catalyst is pivotal to the activation of the −CN group (obtained yields without catalyst varied between 0% and 5%). The optimized coordinate suggested that despite presenting a 1:1 stoichiometry between dipole and dipolarophile, an excess of nitrile (approximately 2:1) was needed to conduct the reaction for the studied system. On the other hand, each cobalt­(II) cation required two azide anions to form a neutral species, while the optimum achieved was around 6 h, which is a reasonable time for this kind of reaction.

Theoretical calculations indicate that the cycloaddition proceeds favorably when an intermediate neutral complex comprising Co­(II), benzonitrile, and two azide anions is formed. Furthermore, the proposed cationic and free-catalyst systems underscore the pivotal role of cobalt­(II) as a catalyst and emphasize charge neutrality in the aqueous medium employed for our synthesis. Finally, despite not presenting shorter energy barriers in aqueous media than DMF, our approach is a preferable alternative due to its lower environmental emissions.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c07521.

• Statistical optimization, crystal X-ray diffraction data of 5-phenyl-1H-tetrazole, and computational details (PDF)

Conceptualization, J.A.-G.d.V., J.S.-C.; Methodology, J.A.-G.d.V.; Software, V.G.-V.; Validation, J.A.-G.d.V., V.G.-V.; Investigation, J.A.-G.d.V.; Resources, J.A.-G.d.V., J.S.-C.; WritingReview and Editing, J.A.-G., J.S.-C., V.G.-V.; Supervision, J.S.-C.

The authors declare no competing financial interest.