Inthe paper mentioned above, recently publishedin Catalysis Com-munication [1], we reported about the utilization of carbon nanotubes(MWCNT) as gold support for the catalytic oxidation of glycerol underbasic conditions. According to the paper, this catalyst would lead to ahigher selectivity to dihydroxyacetone (60%) when compared withAu/AC catalysts (18%).The quantitative analysis of the mixture was carried out by highperformance liquid chromatography (HLPC) and the compoundswere identified by comparison with standard samples. Reactant andproducts were separated with an ion exclusion column (Alltech OA1000), which is commonly used in this type of work [2–4].Dihydroxyacetone(DIHA)isareactionproductmentionedinthelit-erature by different groups even under basic conditions [4–8].Underourexperimentalconditions,theretentiontimeofthiscompoundcoin-cided exactly with one of the peaks obtained from the HPLC analysisof the reaction mixture. Therefore, in the chromatogram, this peakwas identified as DIHA. Moreover, the selectivity obtained with goldsupportedonactivatedcarbonwassimilartothatreportedbyotherau-thors, which apparently confirmed our results [9].Unfortunately, we found recently that formic acid (FORMA) was er-roneously identified as dihydroxyacetone at the time the article waspublished. This was due to the fact that both compounds have thesame retention time in the column mentioned above. Some tests werecarried out using another column (BIO-RAD Aminex HPX-87H) wherepartial separation of standard sampl es of formic acid and dihydroxyace-tone was achieved. The analysis of the reaction mixtures in this columnallowedustoconcludethattheproductinitiallyidenti fiedasdihydroxy-acetonewasinfactformicacid.Thislattercompoundisnotwidelymen-tioned as a product of glycerol oxidation, especially when conventionalcarbon materials are used as support. However, an exception wasreported recently by Prati et al. [10], when testing gold supported oncarbon nanofibers, which lead to selectivities to FORMA as high as 30%.The origin of formic acid is still under discussion. This compound isprobably obtained via the base catalyzed degradation of dihydroxyace-tone[11].Therefore,webelievethatthehighformicacidselectivityob-served in this work is still due to the peculiar porous characteristic ofmulti-walled carbon nanotubes, which may favor the initial formationof DIHA, as mentioned both in our article [1] and in the work of Pratietal.[10].Accordingly,wesuggestthatglycerolcanbeinitiallyoxidizedto DIHA; this compound acts as an intermediate prone to degradationdue to the highly basic medium used, and a fraction could be readilyconverted into FORMA, whereas the remaining is further oxidized, viaglyceraldehyde, in agreement with Fig. 1 of reference [1].Consequently, some of the results and conclusions in [1] must bemodified. Table 1 summarizes the most relevant changes.In conclusion, under basic conditions, gold supported on activatedcarbon is preferable in order to obtain a high selectivity towardsglyceric acid, product of commercial interest, and no dihydroxyace-tone is formed.References