Density Functional Theory Study of Degradation of Tetraalkylammonium Hydroxides Shaji Chempath, † James M. Boncella, § Lawrence R. Pratt, | Neil Henson, † and Bryan S. Pivovar* ,‡ Theoretical DiVision, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, National Renewable Energy Laboratory, Golden, Colorado 80401, Materials Physics and Applications, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, and Department of Chemical and Biomolecular Engineering, Tulane UniVersity, New Orleans, Louisiana 70118 ReceiVed: December 28, 2009; ReVised Manuscript ReceiVed: May 12, 2010 We report density functional theory (DFT) studies of the degradation mechanism of tetraalkylammonium cations which are of interest for anion exchange membrane fuel cells. Three mechanisms of attack by hydroxide anions are explored: an S N 2 pathway leading to alcohol formation, an ylide pathway that gives rise to unstable intermediates, and Hofmann elimination. Tetramethylammonium, ethyltrimethylammonium, and benzyltri- methylammonium are the model cations studied here. S N 2 attack on tetramethylammonium was found to have a free energy barrier of 17.0 kcal/mol at 298 K. In the case of ethyltrimethylammonium, the overall barrier for the S N 2 pathway was found to be 23.0 kcal/mol while Hofmann elimination was 12.8 kcal/mol. The ylide and S N 2 attacks on benzyltrimethylammonium show similar energy changes as in the case of tetramethylammonium. In the case of benzyltrimethylammonium, additional side reactions starting from the ylide intermediate are also shown to be feasible. We also discuss the influence of the immediate solvation shell on the reaction mechanism. A refined model in which the immediate solvation shell of hydroxide is modeled explicitly is found to have better experimental agreement than a model in which solvation is modeled implicitly. Introduction Tetraalkylammonium cations (e.g., N(CH 3 ) 4 + ) are known to degrade under aqueous conditions at high pH, primarily due to attack by the hydroxide counterion. 1-5 Among the many applications of tetraalkylammonium ions, their possible use in alkaline anion exchange membranes (AAEM) has received recent attention. 6,7 These cations can be tethered to a polymer backbone and the resulting material can be used as the polymer electrolyte membrane (PEM) in an alkaline membrane fuel cell (AMFC). 8,9 However, the tendency of these cations to react with aqueous hydroxide (OH - ) under fuel cell operating conditions is one of the limitations that is preventing AMFCs from being widely used. Target AMFC operating conditions may be 60-100 °C, where membranes may contain 10-50 absorbed water molecules per cation. If AMFCs can be successfully demonstrated they will likely displace current polymeric, acid based fuel cells in many applications. Acidic membrane fuel cells use precious metals such as Pt or Ru as the catalyst in their electrodes, whereas AMFCs allow for the use of nonpre- cious metal catalysis. 10 Recently, Lu et al. 9 demonstrated an alkaline fuel cell that did not use any noble metal catalysis. The membrane in their fuel cell used quarternary ammonium polysulfone and was demonstrated at only 60 °C, presumably because the quarternary ammonium group is not stable at higher temperatures. There is clearly a need for identifying cations that are chemically stable in the temperature range of 60-100 °C in aqueous alkaline conditions. In our previous work, 6 we reported computational studies on the stability of tetramethylammonium ([N(CH 3 ) 4 + ]) under aque- ous conditions in the presence of OH - ion. Two reaction paths, both of which lead to the same final reaction products, were identified. In the S N 2 pathway, nucleophilic attack by OH - on the CH 3 group produces trimethylamine and methanol. In the other pathway denoted as the ylide pathway, OH - abstracts a proton from a CH 3 group and produces a water molecule along with an ylide intermediate (N(CH 3 ) 3 + -CH 2 - ). This intermediate in turn reacts to give the final products, trimethylamine and methanol. By changing the dielectric constant of the medium used in our DFT calculations, we also found that the rate of degradation increases with decreasing water content in the reaction medium. Experimental results on the thermogravimetry (TG) and mass spectroscopy (MS) of N(CH 3 ) 4 OH · 5H 2 O were also reported. 6,11 An exchange between the aqueous and methyl hydrogens (as observed with deuterated water) confirmed our prediction of an ylide pathway. Almost all the previous work on AAEMs suggests that anion exchange membranes are not stable for extended periods of time in the hydroxide form at temperatures above 60 °C. Hatch and Lloyd 1 and Baumann 2 studied the stability of membranes made from polymers with benzyltrimethylammonium side chains and concluded that S N 2 attack on benzyl and methyl groups is the main reason for membrane instability above 60 °C. The Hofmann elimination reaction, which involves hydroxide attack on aliphatic hydrogens at the -position relative to the quarternary nitrogen, has also been reported as another mechanism of degradation of tetraalky- lammonium ions. 12-14 Tomoi et al. 14 modified the linker between the benzene ring and the quarternary nitrogen in anion exchange resins and found improved performance with linkers such as * To whom correspondence should be addressed: 1617 Cole Blvd, Golden, CO, 80401. Fax: 303-275-2840. Phone: 303-275-3809. E-mail: bryan_pivovar@nrel.gov. † Theoretical Division, Los Alamos National Laboratory. § Materials Physics and Applications, Los Alamos National Laboratory. | Tulane University. ‡ National Renewable Energy Laboratory. J. Phys. Chem. C 2010, 114, 11977–11983 11977 10.1021/jp9122198  2010 American Chemical Society Published on Web 06/22/2010