*2.3. Methods*

Kinetic data were obtained both chromatographically and spectrophotometrically. Observed rate constants, *k*obs, were obtained by fitting the percentage of yield-time or absorbance-time data to the integrated first order Equation (1), where *X* stands for the experimental measured UV–VIS absorbance or product yields. Runs were done at *T* = 35 ± 0.1 ◦C (2MBD and 3MBD) and at *T* = 60 ± 0.1 ◦C (4MBD) with aryldiazonium ions, ArN<sup>+</sup> 2 , as the limiting reagents. The reported *k*obs values are the average of duplicate or triplicate experiments, with deviations lower than 7%.

$$\ln\left(\frac{X\_t - X\_{\infty}}{X\_0 - X\_{\infty}}\right) = -k\_{\text{obs}}t\tag{1}$$

Spectrophotometric kinetic data were obtained by monitoring ArN<sup>+</sup> 2 loss at a suitable wavelength to minimize interferences with other components of the solution (dediazoniation products) as much as possible. Linear variations (not shown) in the absorbance of ArN<sup>+</sup> 2in aqueous and MeOH solutions up to [ArN<sup>+</sup> 2] = 2.0 × 10−<sup>4</sup> M (cc. ≥ 0.999) were

found, keeping with the predictions of Beer's law. ArN<sup>+</sup> 2 solutions were prepared by dissolving the corresponding aryldiazonium salt in the appropriate acidic (HCl) MeOH/H2O mixtures to diminish diazotate formation [2]. Final concentrations were approximately 1.0 × 10−<sup>4</sup> M and [HCl] = 0.01 M. The stock ArN<sup>+</sup> 2 solutions were kept in the dark at low temperatures ( *T* < 5 ◦C) to minimize their photochemical and/or spontaneous decomposition (Zollinger, 1994) and were used immediately or within a time period of less than 60 min.

Preliminary HPLC experiments showed that, in the absence of ClO− 4 , up to five decomposition products were detected to different extents. The most frequent were cresols, ArOH, and the methyl phenyl ethers (ArOMe), and, on occasion, depending on the particular experimental conditions, chlorotoluenes (when investigating Cl− dependence), as well as the reduction product, toluene (ArH), that was detected in low yields when employing high percentages of MeOH in the solvent mixture. The calibration curves for all these products were obtained by employing authentic commercial samples. Linear (cc. > 0.999) calibration absorbance–concentration plots were obtained and employed for converting HPLC peak areas into concentrations. The following equation Yield = 100 [dediazoniation product]/[ArN<sup>+</sup> 2 T] was employed to calculate the yields of the various dediazoniation products.

When studying the effects of the perchlorate salts, a new chromatographic peak was found that was not observed in its absence, with a retention time lower than that of ArOH but higher than that of the front peak (see Figure S1, Supplementary Materials). The area of this new peak increased linearly upon the increasing [LiClO4]. In dediazoniations, where sulfate ions were employed, the aryldiazonium sulfate anion, Ar-OSO− 4 was formed and could be isolated [46,47]. As the possibility of the formation of the aryldiazonium perchlorate derivative ArOClO3 exists, which is explosive in the solid state [3,38], we made no attempts in identifying nor isolating this new product. In spite of this, the failure in identifying this new derivative does not invalidate the main conclusions of the work because: (i) the peak area is proportional to the added [LiCLO4] and (ii) dediazoniation products are formed competitively and their rate of formation is the same as that of other dediazoniation products (as determined from the variations in its peak areas with time), and equal to that of ArN<sup>+</sup> 2 loss.

The rates of the formation of dediazoniation products were obtained by employing a derivatization method, as described elsewhere [48]. To minimize side reactions that may occur upon the injection of ArN<sup>+</sup> 2 in the HPLC system (metal parts, solvent, etc.), TRIS buffer ([TRIS] = 0.05 M) solutions of the coupling agen<sup>t</sup> 2N6S, that allows for the rapid formation of a stable azo dye, were employed [48]. The dediazoniations were quenched at convenient times, as described elsewhere [42,48]. The derivatization reaction was carried out under pseudo-first order conditions ([2N6S] > 20 [ArN<sup>+</sup> 2 ]). The final pH was adjusted to a pH of approximately 8, because naphthoxide ions are much more reactive than their protonated forms. It is not advisable to use lower acidities, because the competing reactions of aryldiazonium ions with OH<sup>−</sup>, to form diazotates, becomes significant [2].

The experimental conditions were chosen so that the coupling reaction was essentially over by the time the reagents were mixed, i.e., azo dye formation is much faster than dediazoniations (at least 100 times faster) and ArN<sup>+</sup> 2 is effectively quenched at any solvent composition. Figure 1 illustrates the determination of the rates of product formation (ArOMe, Figure 1A) and of ArN<sup>+</sup> 2 loss by monitoring the decrease in the absorbance of the azo dye formed (Figure 1B).

**Figure 1.** Illustrative examples of determination of the rate constant for product formation (ArOMe) by employing the chromatographic (**A**) and ArN+2 loss by monitoring the decrease in the absorbance of the azo dye formed in the derivatization method (**B**) under the same experimental conditions. The *k*obs values obtained were similar (~12 × 10−<sup>4</sup> s<sup>−</sup>1) with differences lower than 10%. Experimental conditions: 99.5% MeOH/H2O, [2MBD] = 1 × 10−<sup>4</sup> M, [HCl] = 0.01 M, *T* = 35 ◦C.
