In our continuous efforts to develop different additives fulfilling our lemma “Choosing the Right Peptide Coupling Reagent for Each Reaction”, we have prepared and assayed different oxime analogs. Although OxymaPure has been shown to be unbeatable, some of the new oxime-based derivatives have been found to possess interesting properties. Thus, Oxyma-B (9)[23] has shown to be even better than OxymaPure in minimizing racemization and Amox (4)[24] to be very convenient for the protection of amines with the 9-fluorenylmethyloxycarbonyl (Fmoc) group avoiding the formation of dimers associated to the high reactivity of the active species, mainly the chloride derivative.
It is well known that the quality of an active ester is intrinsically associated with the strength of the conjugate acid. In that regard and to rationalize our previous results, and more importantly for the development of new ones, we have performed an in silico study using ACD/Percepta software[25] and the pKa GALAS algorithm available in it to calculate the acid ionization constant values of the various oximes and other additives (Table 1).[26] Like the pKa Classic method, which is a variation of a classical Hammett-Taft approach and is available as an alternative within said software, the GALAS algorithm is based on analogous fundamental considerations.[26-28] However, instead of largely relying on equations and parameters quantified by other authors, it is developed entirely in-house by ACD/Labs, parameterized “from scratch” using an internal training set of >18,000 compounds with available experimental pKa measurement data. The custom nature of the pKa GALAS model allows for greater flexibility in using various ad-hoc adjustments and modifications, going beyond the scope of the concepts considered in the classic Hammett-Taft approach where needed. One of them is the concept of so-called “fundamental micro constant”—a micro-pKa value for an ionizable group in a hypothetical state of an uncharged molecule, which is then used to calculate a corresponding micro constant for that group in any protonation state by introducing the corrections for charges. In total, the algorithm utilizes a database of 4600 ionization centers, a set of ca. 500 various interaction constants and four interaction calculation methods for different types of interactions, producing a full range of micro constants from which pKa macro constants are obtained. The latter are experimentally measurable values associated with a particular ionization stage of any given ionizable group. Very often, when ionizable groups in a particular protonation state possess pKa micro constants of comparable magnitude, several of them undergo (de)protonation simultaneously in an isolated ionization stage and make a collective influence toward the corresponding macro-pKa value. pKa GALAS provides full and detailed insights into this relationship between the macroscopic pKa values of the molecule and the microscopic pKa constants of individual groups and extent of their dissociation in each ionization stage. This was the main reason for selecting pKa GALAS versus pKa Classic for this investigation.
First, the pKa of some non-oxime additives were calculated (Table 1). However, using this method 1-hydroxybenzotriazole-based additives did not show any pKa values. The 1‑hydroxybenzotriazoles can form zwitterionic species (HB+A−) via two tautomeric equilibria (Figure 4). Those zwitterionic species possesses zero net charge and show low or negative pKa values.[29] pKa values found in the literature for HOBt and HOAt are 4.60 and 3.28, respectively.[30]
Then, pKa of some oxime coupling reagent additives reported by our group and others were calculated, then of some oximes described in the literature or commercially available, and finally, some unknown oximes. The pKa values of oximes are divided into four categories and indicated with color code (if pKa values < 4—yellow, 4 to 5—dark green, 5 to 7—light green, 7 to 9—light orange, > 9—brown).
The first conclusion that we can get from Table 1 is that overall, the results obtained agree with what was expected. Thus, OxymaPure (1) and their close ester derivatives (2, 3) are experimentally considered to be the best and this correlates with their acidity, which is also superior for the most part compared to the other derivatives. In this regard, our group has demonstrated that OxymaPure is more efficient than Amox (4), N-Oxyma (5), Dmox (6), PipOX (7), MorOx (8), Oxyma-B (9), and Oxyma-T (10),[23, 24, 35] and the calculation outlined in Table 1 confirm that all of them have a higher pKa. Of course, the acidity of the oxime depends on the electron-withdrawing groups adjacent to oxime. Among the oximes described, the presence of cyano is key for their acidity, and the pair cyano-ester (1-3) is superior to cyano-amide (4-8), and these to the cyano-aromatic group (11-14). The superiority of OxymaPure (1) over HOPO (22) can also been explained by the higher acidity of the former.
The surprising results are the acidity of Oxyma-B, because it is considered to be a substitute for OxymaPure but its acidity is not very high. However, its good performance could be explained by the presence of the carbonyl groups oriented in the same direction as the N-OH group in Oxyma-B playing an assisted basic catalytic role, thereby enhancing the nucleophilicity of the amine function during the coupling (Figure 5). A similar effect has been described for HONM (19), HOAt, and N-ethoxycarbonyl-2-ethoxy-1,3-dihydroquinoline (EEDQ).
Amox, which has an acidity lower than OxymaPure, has demonstrated that when used in combination with 9-fluorenylmethanol as a mixed carbonate (Fmoc-Amox) it is able to introduce the Fmoc group in amino acids without the formation of dipeptides as occurs to a greater extent with Fmoc-Cl and a lesser extent with Fmoc-OxymaPure (Figure 6). In this same regard, hydroxyimino-2-phenylacetonitrile (11), which forms part of Boc-ON [2-(Boc-oxyimino)-2-phenylacetonitrile] and was proposed by Ito for the safe protection of amines with tert-butoxycarbonyl (Boc) group,[36-38] shows a pKa that confirms its moderate reactivity and therefore the absence of formation of Boc-dipeptides during the introduction of the Boc group (Figure 6). Finally, the pKa of HOSu also confirms that Fmoc-OSu is a good reagent to avoid this side-reaction.
Our group has demonstrated that the oxime derivative of Meldrum’s acid (HONM) reacts with DIC rendering the corresponding adduct (Figure 7). Because this reaction is preferred, HONM is not a good additive in combination with DIC for peptide coupling, since it mostly reacts with DIC leading to peptide formation in low yield.[40]
Recently, Kolis and co-workers have observed that OxymaPure also reacts with DIC.[54] Although, in this case, the formation of the adduct is taking place to a much lesser extent than with HONM, it can cyclize with the generation of HCN (Figure 8). These results have been corroborated by Pawlas[55] and co-workers, and our own group.[56, 57]
In this context, and although this side-reaction takes place to a very low extent and in only certain cases, there is interest in finding oxime derivatives with no cyano groups. Taking into account both the availability of their synthesis and the pKa, out of four nitro derivatives (46-49) only one ethyl—2-(hydroxyimino)-2-nitroacetate (46)—fulfils those requirements. Admittedly, the high value of uncertainty, indicating a relatively lower quality of pKa predictions for these nitro derivatives, could be the source of some concern. However, absolute values aside, the error margin being essentially equal for these four compounds (46-49), and cyano and nitro groups being very similar in their electronic activity profile, allows for an interpretation of the general trends. The latter for the group of four nitro compounds (46-49) is fully in line with common chemical intuition, and the corresponding trends in the series of cyano analogs, which are predicted with a much higher certainty, i.e., that a dinitro compound, just as a dicyano one, will be more acidic compared to a mononitro/monocyano derivative, and the latter, in its own turn, will be a stronger acid than a mononitro/monocyano-phenyl analog. Specifically, pKa (47) << pKa (46) << pKa (48) ~ pKa (49) is analogous to pKa (15) << pKa (1) << pKa (11). In this context, concerns regarding prediction accuracy do not interfere with the conclusion, that ethyl 2-(hydroxyimino)-2-nitroacetate (46) should be the most promising cyano-free alternative candidate of all nitro compounds considered here.