Introduction

If one accepts “picric acid” as a shorter name for 2,4,6-trinitrophenol, then 2,4,6-trinitrophenyl is picryl, and 2,4,6-trinitroaniline can have the convenient shorter name of picramide, which will be used in this paper. One of the most stable free radicals is 2,2-diphenyl-1-picrylhydrazyl (DPPH, 1), due to electronic and steric factors [1, 2]. The electronic delocalization of the unshared and odd electrons around the two hydrazinic nitrogen atoms is equivalent to consider DPPH as a push–pull aminyl [3, 4], with an electron-donor diphenylamino group and a bulky electron-acceptor picryl (2,4,6-trinitrophenyl) group. We were able to obtain various other persistent push–pull aminyls, one of which 4 was formed by oxidizing N-methoxypicramide (N-methoxy-2,4,6-trinitroaniline, 2) [5,6,7]. An interesting deep blue betainic compound 5 was obtained from the reaction of 1 and 2 [8]. Its structure was confirmed by X-ray crystallography [9], and we report now some of its physical properties.

Results and discussion

Synthesis and structure

The reaction of two moles of the stable free radical DPPH 1 with one mole of the N-methoxypicramide 2 leads instantaneously to a blue compound that can be detected by TLC. This reaction occurs in two steps (i and ii, Scheme 1): the first one (i) is a redox process, followed by the second one (ii), as a result reaction involving of radicals coupling 1 and 4, resulting in the blue compound 1-picryl-2-phenyl-2-(4-picrylamidephenyl) diazenium betaine 5.

scheme 1

The reaction presented in Scheme 1 is unusual. The betaine 5 results in high yield [8] (~80%) because it is favored, first of all by the stability of the two free radicals (DPPH, 1 and N-methoxypicrylaminyl, 4), then by the redox environment which ensures the regeneration of the free radical DPPH (because the betaine 5 is an oxidant for the hydrazine 3 formed from DPPH). The stability of hydrazyl 1 and the aminyl 4 free radicals is due to the electronic push–pull effect for nitrogen-centered free radicals [2, 3] (an effect also named capto-dative stabilization [10] or merostabilization [11], mainly for carbon-centered free radicals).

We describe in the following cyclic voltammetry (CV) measurements that demonstrated the stability of the compound 2 in the repetitive cyclic redox process (Fig. 1a), when the free radical 4 is the intermediate compound in the oxidation step. The qualitative analysis by CV was completed by vertex CV (Fig. 1b).

Fig. 1
figure 1

a Cyclic voltammetry of compound 2 in acetonitrile in the presence of tetra-n-butylammonium perchlorate (TBAP 10−1 M as supporting electrolyte), scan rate: 100 mV s−1, 50 mV−1, and 30 mV s−1. Peak potentials (V vs. Ag/AgCl) for oxidation (E pa = 0.67 V) and reduction (E pc = 0.60 V); b the “reactivity” test of the compound 2 by cyclic voltammetry (anodic ↔ cathodic) measurements (in acetonitrile in the presence of TBAP) by multiple cyclic voltammograms at constant speed (100 mV s−1) and by vertex cyclic voltammogram with the following potentials steps 0.2 V → −0.8 V → 1.7 V → 0.2 V

Full size image

Congeners 6a–6f of the betaine 5 (with analogs of picryl groups indicated in Scheme 2) were synthesized [12] in good yields according to the mechanism presented in the same Scheme 1.

scheme 2

The properties of the seven betaines 5 and 6a–6f are similar [12]: blue colored and redox properties. Recent research [9] by single-crystal X-ray diffractometry analysis confirmed the structure of the betaine 5 which has two resonance formulas 5A (quinonoid) and 5B (benzenoid) presented in Scheme 3 (abbreviations: Pic = picryl and Ph = phenyl). The geometry of betaine 5 in Scheme 3 reflects the X-ray findings, with the phenyl group practically orthogonal to the remainder of the molecule. The quinonoid structure 5A is dominant (as shown by the bond lengths [9]) and accounts for the deep blue color and metallic luster of compound 5 [9]. The aromaticity of the central para-phenylene group is insufficient for compensating the energetic drawbacks of resonance structure 5B.

scheme 3

Redox properties and cyclic voltammetry for betaine 5

The betaine 5 (blue color) leads to compound 7 (brick-red color) by a chemical reducing process (Scheme 4). This process is reversible [8, 12].

scheme 4

The redox properties of the betaine 5 were studied by CV (Fig. 2). The CV of betaine 5 indicates a reversible one-electron reduction in acetonitrile at E pc = 0.042 V and a remarkable oxidation potential at E pa = 0.106 V for 100 mV s−1 scan rate. In this process, the betaine 5 is stable after multiple CV cycles; characteristic cathodic and anodic peaks for a reversible redox process are shown (Fig. 2). The peak splitting (ΔE p = E pa − E pc) was 64 mV. It was observed that compound 5 is very stable because the height of the current peak does not decrease when the potential follows vertex mode, but it depends on whether the starting potential is very close to cathodic range. Thus in the vertex mode (starts from 0.2 V) both anodic and cathodic peaks maintain the same intensity and potential showing a well-defined cathodic peak current for reduction at E = 0.042 V and anodic peak current for oxidation at E = 0.106 V (for scan rate 100 mV s−1). However, with starting potential 0.5 V, both the anodic and cathodic peaks decrease. The first pair peak is attributed to the intermediate radical anion 8 formed by one-electron reduction of betaine 5 (Scheme 5).

Fig. 2
figure 2

Cyclic voltammograms in acetonitrile of compound 5 (1 mM) in presence of 0.1 M TBAP (on platinum vs Ag/Ag+ electrodes); scan rate: 10, 30, 50, 100, 200, 500, and 1000 mV s−1 with vertex potential 0.2 V → −0.8 V → 1.7 V → 0.2 V

Full size image
scheme 5

EPR measurements

Betaine 5 dissolved in methylene chloride is reduced in the presence of crystalline sodium ascorbate. Some 15-crown-5 ether is added to this system in order to ensure the solubility of ascorbate into the organic medium. This process is associated with a change of color from blue to brick-red. The EPR spectrum recorded for this mixture evidenced a free radical with five main lines due to the two hydrazinic 14N nuclei with about the same hyperfine coupling constants (Fig. 3, a N = 8.9 G, g = 2.0038 relatively similar to DPPH [1]), which disappears over time.

Fig. 3
figure 3

The EPR spectrum in methylene chloride of the radical-anion 8 (the reaction product of compound 5 in presence of solid sodium ascorbate and 15-crown-5 ether)

Full size image

The EPR signal is in fact attributed to the intermediary radical-anion 8 formed by one-electron reducing of betaine 5 (Scheme 5). The disappearance of EPR signal in time evidenced a further reduction process, resulting in the formation of dianion 9 (Scheme 5). A more detailed discussion of the radical-anion 8 presented in Scheme 5 and of its EPR spectra will be presented in the “Conclusion” section.

By methylation of the reduced form 7, the monomethylated 10 and dimethylated derivatives 11 were obtained (Scheme 6) [12]. Oxidation of the monomethylated derivative 10 led to the neutral radical 12, whose EPR spectrum also showed five main lines (with a N = 8.6 G) [12].

scheme 6

Magnetism

Magnetic susceptibility was measured at room temperature for crystals of the betaine 5 yielding the value χ molar = −535.5 × 10−6 cm3/mol; for the crystals of the reduced form 7, we obtained χ molar = −894.9 × 10−6 cm3/mol. These measurements showed that both compounds 5 and 7 are diamagnetic. The higher diamagnetism of 7 that of 5 is probably due to the aromaticity of 7.

Spectral properties

The absorption spectra of the betaine 5 and its reduced form 7 in solid state are presented in Fig. 4.

Fig. 4
figure 4

Absorption spectra in solid state of the blue dye 5 and its reduced form 7

Full size image

The color of the crystals prompted us to record absorption spectra from the microcrystalline powder using an integrating Ulbricht sphere [9]; in this case a bathochromic shift was observed at λ = 742 nm (Δλ = 32 nm comparative with normal spectra presented in Fig. 4).

According to the solvent polarity, betaine 5 in solution shows a weak hypsochromic effect (shifts of λ max to shorter wavelength) which increases with the solvent polarity (negative solvatochromy, Fig. 5).

Fig. 5
figure 5

Absorption spectra of the betaine 5 in various solvents with different polarities (E T(30) parameters)

Full size image

Conclusion

1-Picryl-2-phenyl-2-(4-picrylamidophenyl) diazenium betaine 5 has an interesting structure formed by homolytic reactions in a two-step process. The two bulky electron-withdrawing groups (EWGs) provide the electronic effect along the hydrazinic nitrogens (Scheme 7). Such structures explain the special properties of this type of betaines 5 and 6a–6f, allowing future variations on this theme. It will be interesting to test other bulky EWGs such as aromatic rings with NO2, SO2R, COR, as well as smaller-volume substituents such as cyano, triflate, etc. The fact that 1-cyano-2,2-bis(3,5-di-tert-butylphenyl)hydrazyl and 1-cyano-2,2-diphenyl hydrazyl [13,14,15] have a stability comparable to that of DPPH indicates that electronic effects predominate over steric effects in explaining the stability of hydrazyls.

scheme 7

For compound 5 several physicochemical properties were enhanced with CV, EPR, magnetic, and spectral measurements. The investigation by CV for the betaine 5 results in the formation of a radical-anion by one-electron reduction; this process is confirmed by EPR. In solid state, compound 5 is diamagnetic. In solution, compound 5 shows a weak negative hypsochromic effect with increasing solvent polarity.

According to Linnett’s theory [16, 17] taking into account electronic spin (denoted by α and β), the stability of DPPH is due to increasing the bond order of the N–N bond from 1 to 1.5 (with three electrons connecting the two hydrazinic nitrogens in DPPH 1 as seen in Scheme 8). This was confirmed by measuring the intramolecular rotation barrier [18] and by X-ray crystallography [19,20,21]. All non-hydrogen atoms in DPPH and in nitroxides (both the picryl group and, respectively, an oxygen atom are EWGs) have electron octets. A generalized Linnett formula for the radical-anion 13 is presented in Scheme 8 where picryl groups are the EWGs.

scheme 8

One can understand why the radical-anion 8 has an EPR spectrum so similar to that of DPPH, essentially because in 8 (or in general 13), instead of the two phenyl groups in DPPH 1, one the two phenyl groups is replaced by a para-substituted phenyl. The structure of Scheme 8 also explains how the odd-electron is almost equally distributed between the two hydrazinic nitrogen atoms. Hyperfine coupling constants with protons, amino and nitro 14N nuclei are much smaller, so that only the prevailing electron spin distribution due to the two 14N hydrazinic nuclei is reflected in structures 1 and 13.

Experimental

Compound 5 was obtained and purified according to the published procedure [8, 12]. For TLC we used Silicagel plates 60F254 from Merck. Commercially available compounds: (+)-sodium l-ascorbate, 15-crown-5 ether from Sigma-Aldrich; tetra-n-butylammonium perchlorate (TBAP) from Fluka; the spectroscopy solvents were spectral purity from Sigma-Aldrich.

The reversible redox process presented in Scheme 4 was monitored spectrally in visible range or (silicagel plates 60F254, toluene, twice).

Cyclic voltammetry (CV) was performed under the conditions mentioned in Figs. 1 and 2. The CV measurements were performed with a conventional three-electrode glass cell by a PAR-273-A potentiostat, and all solutions were prepared by using acetonitrile. As reference electrode, an Ag/AgCl electrode immersed in a 0.1 M supporting electrolyte solution was used, linked to the main compartment of the cell by a Vycor plug. A platinum disk (surface area, 0.07 cm2) and a platinum wire were used as the working and counter electrode, respectively. As supporting electrolyte, tetra-n-butylammonium perchlorate (TBAP) was used at a concentration of 0.1 M.

The magnetic measurements were performed at room temperature on a Lake Shore’s fully integrated Vibrating Sample Magnetometer (VSM) system 7404. This equipment allows to take low-cost M(T,H) magnetic data from 100 to 950 K using nitrogen, and argon gas, in variable magnetic field [0–21,700] Oe.

The electron paramagnetic resonance (EPR) measurements were performed under the conditions mentioned in Fig. 3. EPR spectra were recorded on a JEOL FA100 EPR spectrometer. The following parameters were used: center field 3222 G, sweep width 100 G, microwave power 1 mW, modulation amplitude 1 G, microwave frequency 9.04 GHz, sweep time 240 s and time constant 0.1 s.

In solution, the UV–Vis spectra (Fig. 5) were recorded using a Perking Elmer, Lambda 35, UV–Vis spectrometer at a scan rate of 480 nm/min and slit, 1 nm. In solid, the absorption spectra (Fig. 4) were recorded using a spectrophotometer Perkin Elmer Lambda 35, equipped with an integrating sphere. As reference, a certified reflectance standard, spectralon, was used and the measurements were carried out in the range 900–200 nm. The sample holder is 8° wedge and the used parameters were: data interval, 1 nm; scan speed, 120 nm/min and slits, 4 nm.