Titration

Precise knowledge of acid-base behaviour of mineral surfaces can be of great importance to realistically describe sorption processes since retardation processes (sorption, precipitation, etc.) of the majority of relevant radio nuclides (RN) are pH dependent. Potentiometric acid-base titration experiments of mineral surfaces have been widely applied (e.g. Arnold et al. (2001), Lützenkirchen et al. (2012 a, b), Schwarz et al. (1983)) to characterize mineral electrostatic behaviour in terms of points of zero charge, surface charge densities, and net surface proton excess. In this study a combination of three state-of-the-art titration techniques was used to characterize mineral surface charge properties.

Discontinuous electrolyte and mass titration experiments in combination with continuous potentiometric acid-base titrations were used to get insight in mineral surface protonation and deprotonation states as a function of e.g. pH, ionic strength (IS), and solid-liquid ratio (SLR). For many pure mineral oxides all three techniques result in identical pH values (pHppzc, Tab. 1) where the surface charge density is zero. Natural or highly complex mineral surfaces do not exhibit a pHppzc due to intrinsic mineral characteristics (e.g. Lützenkirchen et al. (2012 a), Preocanin and Kallay (1998), Noh and Schwarz (1988)); hence, the definition of surface charge properties of these minerals proves to be rather challenging.

Besides gathering experimental data, their interpretation, evaluation, and simulation are key aspects. To realistically describe complex geochemical processes in hard rock or sedimentary environments deterministic thermodynamic sorption models (TSMs) have been developed over the past approximately 50 years. Deterministic TSMs describe mineral-water interfaces during sorption processes and simulate thermodynamic processes such as protonation and deprotonation reactions, surface complexation, cation exchange processes, surface precipitation, and dissolution processes, as well as solution speciation. The notion "deterministic" refers to models that are governed by parameters and predefined initial model boundary conditions. No randomness or stochastic correlations are included. Uncertainties are external to the deterministic model. Therefore, utilized parameter values, material properties, system boundary conditions, etc. should be well known.

Surface complexation parameters (SCPs) are utilized to describe thermodynamic processes (inverse modeling) and to physicochemically characterize system properties with TSMs. These SCPs are determined from, e.g., titration experiments. Here, SCPs comprise protonation and deprotonation constants (pK values), conditional cation exchange equilibrium coefficients (selectivity coefficients logKc), counter and coion associations coefficients (here for Na+ and ClO4, the surface site densities (SSDs), capacitances (C1, C2), and the specific surface area (SSA).

Exemplarily, results of quartz titration experiments are illustated. Detailed information on SCP determination via inverse modeling and surface charge properties of muscovite and orthoclase is provided in Britz (2018).

Quartz

Fig. 1 depicts mass (A) and electrolyte (B) titration data of quartz. Titration results must be assessed with caution since surface impurities of the natural quartz mineral potentially bias experimental raw data. Interestingly, depending on the initial pH condition (pHini), mass titration results approached two different end points (pHmass, Fig. 1 A):

  • Experiments initiated at pHini 3.0 did not show any change with increasing solid content. This implied an ideal, pristine (hydr)oxide silica surface (free of relevant basic or acidic impurities); pHmass = 2.9 and agreed well with cited silica pHppzc (e.g. Polly et al. (2010)).
  • In contrast, measurements that were started at pHini = 4.5 converged towards pHmass = 6.5 which contradicted the former observation.

 

Results supported the assumption that two "types" of silica surface functional groups might exist which behave differently depending on surrounding pH conditions. One "type" of sites behaved pH independent whereas as a second "type" of sites evolved pH depend properties. This observation could be related to findings of Lützenkirchen (1996) who observed Cd sorption processes in unexpectedly low pH ranges at silica solid-solution interfaces. Lützenkirchen (1996) had to assume that besides Si-OH groups the examined surface was partially composed of cation exchange sites due to small amounts of Na-impurities. Interestingly, similar findings were made in the present study: Quartz also showed minor surface impurities (data provided by Britz (2018)) and Eu sorption processes were also observed at pH < 4 in this study (data provided by Britz (2018)). These unexpected results highlight that assumingly well-studied surfaces like silica surfaces still bear unknowns that have yet to be uncovered.