Sign up for email alerts of new Fluid Journal issues!
Fluid Journal : Winter 2017
15 The Fluid Journal Winter 2017 “Blending approach possesses tremendous potential to reduce total P inputs.” Figure 4. Oxalate extractable phosphorus is the 0-8mm dish section. Percentages expressed as portion of total phosphorus by aqua regia digestion. Table 2. Bulk-XANES results expressing primary solid phosphorus species as a percent of the total. Treatment Dish Section Ca-P Clay adsorbed P Fe adsorbed P Al Adsorbed P Red. Χ2 Control - Water 0-8 68 24 8 - 0.025 8-15.5 72 7 21 - 0.005 MAP 0-8 68 - 22 10 0.005 8-15.5 68 12 20 - 0.005 APP 0-8 46 25 29 - 0.015 8-15.5 67 - 14 19 0.005 80/20 MAP/APP 0-8 44 15 41 - 0.009 8-15.5 68 - 9 23 0.006 content of soil, primarily ferrihydrite, but in recent years some researchers are using the procedure to estimate the fraction of P associated with Fe and aluminum (Al) (Yan et al. 2016). Because calcium oxalates are highly insoluble, an ammonium acetate (pH = 5.5) pretreatment is employed to eliminate possible interference of the Fe estimation in calcareous soils. Phosphorus, naturally, is associated with some of the Ca removed, so quantification of the P in the pretreatment was necessary as well as in the ammonium oxalate extraction solution. Figure 4 suggests that treatment influences P partitioning. Soils receiving MAP appeared to favor the fraction dissolved by the oxalate treatment, while the APP and 80/20 blend, to a lesser extend, favored the pretreatment dissolved fraction. Baifan and Yichu (1989) found that similar ammonium acetate extractions selectively removed calcium associated P in addition to the water-soluble fraction. These results corroborate well with the resin extractable data as the pretreatment is a less chemically “aggressive” solution that removes a fraction that could be interpreted as more labile. Synchrotron investigation. Bulk- XANES (X-ray Absorption Near Edge Structure) analysis was completed to compliment and validate the information gathered through traditional wet chemical methods. This technique is useful in identifying the major P species present in each sample. As expected, the control soils were dominated by apatite, a highly insoluble Ca-P mineral, explaining the low resin extractability of the native P. Soils tested with MAP exhibited similar partitioning with the exception of a relatively large portion sorbed to ferrihydrite. This again supports the resin extractable P data in that approximately one third of the total P may be available to be used by crops. Phosphorus relegated to the apatite fraction was lessened when APP was involved. Both the APP and 80/20 blend treatments were found to contain greater proportions of sorbed P that is less tightly bound and thus more labile. In the 80/20 blend, a higher concentration seems to be associated with ferrihydrite as compared to the APP treated soil. Polyphosphates have been shown to possess a high affinity for iron, and the wet chemical investigation did insinuate that a portion of the P remaining was PP. Unfortunately, at this time, we do not possess a large library of PP mineral and sorbed species XANES spectra that would allow for more precise assessment of this fraction. Digestion and resin extractable P results indicate that very little P diffused outside of the first section, therefore the XANES results (Table 2) for the 8-15mm sections more closely resemble the