2
MgCl2 experiments performed at a molar ratio of 0.5 and 0.75
show a slower SI decrease (Fig. 6). Unlike all other experiments, nucleation and crystal growth did not occur immediately. SI profiles start to decline after several minutes, which is caused by the delayed crystal formation and can be described as the induc- tion time. Induction times were found to be within the range 7– 14 min by following the pH change within the solution (Kofina and Koutsoukos, 2005). The initial pH decrease is due to the mixing of the precursor and the bulk solution. The second pH decrease is directly linked to proton release during struvite formation (V). This second decrease in pH is not visible in the experiments conducted at molar ratio 0.5 due to the slow precipitation reaction and its compensation through NaOH titration.
MgCl2 experiments carried out at Mg:P 1 show a very similar SI decrease to the MgO experiments. There is no presence of an induction time and SI decreases fast due to immediate struvite precipitation. Both previous mentioned pH decreases are super- posed and more NaOH is titrated into the system to compensate the simultaneous mixing and precipitation which leads to an extended and apparent single pH increase (Fig. 6) at the very beginning of the process.
Despite being conducted under the same conditions, runs using MgO precursor show faster struvite precipitation and higher phosphorous recovery rates (Fig. 7). FTIR and IC analysis (data not shown) indicated very high quality struvite products with no traces of MgO or Mg(OH)2. Considering the assumption, that an excess of magnesia is responsible for the final increase in pH