I run an experiment to test the dispersing property of oxalic acid on TiO2 (anatase) colloids, only by mixting the carboxylic aqueous solution with the titania colloid. My results showed that a small quantity of oxalic acid is enough to promote particles dispersion. In a previous similar experiment reported in literature (paper 1) after mixting the TiO2 (anatase) Nps suspensions with oxalic acid (in different amounts) and waiting for 24h, the samples were dried at 80°C and analyzed by ATR FT-IR spectroscopy in order to understand if the carboxylic acid was bound to the TiO2 surface. From the results obtained they confirmed a chemisorption of oxalic acid on the TiO2 Nps. However, In another paper (paper 2) TiO2 nanoparticles were treated with 2 different carboxylic acids (para-Bromobenzoic , BBA and Dodecanoic, DDA) in two different ways : 1) only by mixting ; 2) in autoclave at 100°C (in both case the time of the treatment was 24h). After the treatment was completed, the samples were dried at 50°C under vacuum overnight and then thermal treated in nitrogen atmosphere at 240°C to remove all physisorbed carboxylic acids. Then analyzing the samples by FT-IR , they found that the mixing method gives only weak signals belonging to a possible TiO2 covalently bound to the carboxylic acids (TiO2-OOCR), whereas for the autoclaved samples the signals corrisponding to TiO2-OOCR was clearly visible. These results led to the conclusion that when TiO2 anatase are only mixed with a carboxylic acid there is mainly a physisorption, whereas to have a chemisorption a differnt approach is needed.
Having said that, assuming that the surface of TiO2 has several -OH groups, I know from organic chemistry that they can react with carboxylic acid through an esterification like reaction. This reaction is acid catalyzed (since it's slow) and is an equlibrium that lies on the products (ester-like species) but it's not quantitative.
The reason why in paper 2 the autoclave method gives TiO2-OOCR , is the temperature that decrease the activation enery of the reaction? and If so, is it possible to obtain covalently bound TiO2-OOCR by only mixting a carboxylic acid with TiO2 by adding a concentrated acid?