SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Prochnow, L. I.
Right arrow Articles by Taylor, R. W.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Prochnow, L. I.
Right arrow Articles by Taylor, R. W.
Agricola
Right arrow Articles by Prochnow, L. I.
Right arrow Articles by Taylor, R. W.
Related Collections
Right arrow Soil Chemistry
Right arrow Soil Fertility and Productivity
Right arrow Wetland Soils
Published in Soil Sci. Soc. Am. J. 67:1551-1563 (2003).
© 2003 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA

DIVISION S-8—NUTRIENT MANAGEMENT & SOIL & PLANT ANALYSIS

Synthesis, Characterization, and Agronomic Evaluation of Iron Phosphate Impurities in Superphosphates

L. I. Prochnowa, S. H. Chien*,b, E. F. Dillardb, E. R. Austinb, G. Carmonab, J. Henaob, U. Singhb and R. W. Taylorc

a Dep. of Soil and Plant Nutrition, Univ. of São Paulo/ESALQ, C.P. 9, 13418-900, Piracicaba, Brazil
b IFDC, P.O. Box 2040, Muscle Shoals, AL 35662
c Dep. of Plant and Soil Sciences, Alabama A&M Univ., Normal, AL 35762

* Corresponding author (nchien{at}ifdc.org).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Two of the most common impurities found in superphosphates (single superphosphate [SSP] and triple superphosphate [TSP]) in the forms of Fe3KH8(PO4)6·6H2O and Fe3KH14(PO4)8·4H2O were synthesized (H8-syn and H14-syn, respectively), characterized, and agronomically evaluated to investigate cost-effective means to optimize the utilization of phosphate rocks (PRs) containing Fe impurities. A solubility study showed that more P was released from both compounds as pH increased in the 0.01 M KCl solutions (pH 3.0–7.5) and more P was released from H14-syn than H8-syn. The two Fe-K-P compounds were mixed and compacted with monocalcium phosphate (MCP) at 0, 25, 50, 75, and 100% of total P as MCP. In a greenhouse study, rates of P were applied at 0, 10, 20, 40 and 80 mg P kg-1 from H8-syn, H14-syn, and MCP, while the compacted mixtures were applied only at 40 mg P kg-1 to an Ultisol (thermic Rhodic Kanhapludults, pH 5.3) cropped with upland and flooded rice (Oryza sativa L.) for 65 d. The results showed that P uptake and dry-matter yield were greater with H14-syn than H8-syn for both crops and both compounds were more effective for flooded rice than upland rice. The calculated values of relative agronomic effectiveness (RAE) of H8-syn and H14-syn with respect to MCP were 32 and 72% in dry-matter yield for upland rice and 55 and 102% for flooded rice, respectively. To reach 90% of maximum dry-matter yield obtained with MCP, it required approximately 43 and 35% of total P as water-soluble P (WSP) in the mixtures of H8-syn and H14-syn with MCP for upland rice and only 17 and 11% for flooded rice, respectively.

Abbreviations: AOAC, Association of Official Analytical Chemists • EDS, energy-dispersive spectroscopy • H8, Fe3KH8(PO4)6·6H2O • H8-syn, synthesized H8 • H14, Fe3KH14(PO4)8·4H2O • H14-syn, synthesized H14 • ICDD, International Center for Diffraction Data • MCP, monocalcium phosphate • NAC, neutral ammonium citrate • PDF, powder diffraction file • PR, phosphate rock • RMSE, root mean square of error • SEM, scanning electron microscope • SSP, single superphosphate • TSP, triple superphosphate • WSP, water-soluble P


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
WHEN MARGINAL-GRADE PR containing a high content of cationic impurities is used to produce WSP fertilizers, some water-insoluble P compounds are often present in the final products. These are primarily Fe and Al phosphates (Gilkes and Lim-Nunez, 1980; Lehr, 1980; White, 1976). As the fertilizer industry becomes more dependent on the use of marginal-grade PR, higher levels of impurity compounds in acidulated P fertilizers can be expected. A better understanding of how these compounds form in fertilizer processing, how they react in the soil system, and availability to crops is essential for P fertilizer producers and users to obtain and manage these heterogenous fertilizers in a cost-effective manner.

Agronomic evaluation of P fertilizers obtained from processing marginal-grade PR was reported by Bartos et al. (1992), Mullins et al. (1995), Mullins and Sikora (1995), and Prochnow et al. (1998). The results showed that water solubility of these acidulated P fertilizers required for maximum crop yield was lower than that accepted by some legislation (Official Journal of the European Communities, 1975; Brasil, 1982). However, little attempt has been made to adequately characterize the impurity compounds present in acidulated P fertilizers, especially in SSP and TSP or to produce large quantities under laboratory conditions for agronomic evaluation.

In SSP and TSP, two of the most common water-insoluble P impurities are the generic compounds H8 and H14 (Lehr et al., 1967; Frazier et al., 1991). These compounds preferentially form the K-containing compounds, but in the absence of K in solution, Na-containing compounds will form [(Fe,Al)3NaH8(PO4)6·6H2O or (Fe, Al)3NaH14(PO4)8·4H2O] or even as the analog H9 or H15 members [(Fe, Al)3H9(PO4)6·6H2O or (Fe,Al)3H15(PO4)8·4H2O] (Frazier et al., 1991; Sullivan et al., 1991). According to Lehr et al. (1967) and Frazier et al. (1991), the Na members of these compounds are metastable and can rearrange and further precipitate as FeNaH5(PO4)3·H2O. Lehr et al. (1967) refer to these K-containing compounds as forming continuous isomorphous series with Na+, H+, and even NH+4 replacing K and Al replacing Fe. A study of the phase system Fe2O3–K2O-P2O5–H2O at 25°C conducted by Frazier et al. (1989) showed that H8 preferentially forms at P2O5 concentrations lower than 39% and H14 forms at P2O5 concentrations in the range of 39 to 70%. This information suggests that H8 and H14 would be primarily formed in SSP and TSP, respectively. For example, Prochnow et al. (2003) found that the amounts of H8 in three SSP fertilizers produced in Brazil were 0.9, 3.8, and 7.0%.The particular process and conditions of P fertilizer production, including the post-precipitation of H8 and H14 in phosphoric acid can modify this scenario with the presence of H8 and H14 in SSP and TSP.

Based on agronomic evaluations, researchers concluded that H8 was a poor source while H14 was a readily available source of P for plants (Taylor et al., 1960; Lindsay and Taylor, 1961; Gilkes and Lim-Nunez, 1980; Frazier et al., 1991). However, all the agronomic experiments used the impurity compounds alone, instead of mixing them with MCP as they are in commercial SSP and TSP. It has been shown that P availability to plants of water-insoluble P sources, mainly apatite, can be increased by the addition of WSP (Chien et al., 1987; Menon et al., 1991; Chien et al., 1996). Little information is available on the possible effect of WSP on these water-insoluble Fe-K-P compounds in SSP and TSP.

Lehr et al. (1967) proposed a preparation procedure to obtain various impurity compounds present in P fertilizers, but initial tests with H8 and the H14 showed that the methodology yielded very low quantities of material. This limited the scale for agronomic testing of these compounds in greenhouse facilities or under field conditions. There is a need to develop a proper laboratory procedure that can produce large quantities of specific cationic impurities present in P fertilizers for agronomic evaluation. The compounds also need to be characterized, including their solubility. The information is essential to understand agronomic data of fertilizers produced from marginal-grade PR.

The objectives of this study were to (i) develop a procedure that would yield sufficient amounts of H8 and H14 for agronomic evaluation, (ii) characterize these compounds by chemical analysis and appropriate instrumental techniques, (iii) study the effect of pH on the solubility of both compounds, (iv) evaluate the agronomic effectiveness of these compounds under aerobic and reduced soil condition, and (v) determine the effect of water solubility on the agronomic effectiveness of these compounds.


   MATERIAL AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Synthesis of H8 and H14 Compounds
The H8 and H14 compounds were prepared based on the information provided by Lehr et al. (1967) and a Fe2O3–K2O-P2O5–H2O phase system study conducted by Frazier et al. (1989). After several trials the following procedure was used to make adequate materials for this study. The H8 compound was synthetized by combining 204.4 g of Fe2(SO4)3·nH2O [77% Fe2(SO4)3], 22.0 g of K2SO4, 406.0 g of H3PO4 (85%), and 468.0 g of distilled water in a 1-L beaker covered with a watch glass. This system was stirred and heated for 48 h once the solution reached 70°C. The solids were collected by vacuum filtration with a medium glass fritted funnel, washed three times with distilled water, once with acetone, and air-dried. This process yielded 4.4 g of material. To re-establish the initial concentrations of Fe, K, and P, fresh solution containing the same amounts of Fe2(SO4)3·nH2O, K2SO4, and H3PO4 was added to the original filtrate and the process was repeated two additional times with a total of 45 g of H8 being obtained. An increase in H8 production during the second and third batches was attributed to the influence of nucleation. The H14 compound was obtained following the same procedure except that 102.0 g of Fe2(SO4)3·nH2O, 11.0 g of K2SO4, 810.0 g of H3PO4, and 77 g of H2O were used. The initial yield was 57 g, and reprocessing one more time produced an additional 52 g. As the compounds were prepared, the types of crystals formed were observed by optical microscopy. This helped guide the process and obtain the correct compound based on crystal morphology. The compounds synthesized under laboratory conditions were identified as H8-syn and H14-syn.

Characterization of the Synthesized Compounds
Chemical analysis, optical microscopy, x-ray, infrared, scanning electron microscopy (SEM) and energy dispersive x-ray spectrometer (EDS) techniques were used to characterize H8-syn and H14-syn.

Samples from the two compounds were analyzed for total contents of P, Fe, K, and SO4, soluble P in neutral ammonium citrate (NAC) and water, free water, and water of hydration. The analytical procedure for total P, Fe, and K involved the digestion of 1.00 g of sample using a mixture of concentrated HClO4/HNO3 at 2:1 volume ratio followed by filtration through Whatman No. 42 filter paper (Whatman International Ltd., Maidstone, UK). The filtrate was collected in a 500-mL volumetric flask and diluted to volume with distilled water. The total P concentration was determined by the molybdovanadophosphate procedure (Association of Official Analytical Chemists [AOAC], 1999). The concentration of Fe was determined using a Perkin-Elmer Model 400 inductively coupled plasma atomic emission spectrometer (Perkin-Elmer Corporation, Norwalk, CT) and K was analyzed using a Perkin-Elmer Model 460 atomic absorption spectrophotometer (Perkin-Elmer Corporation, Norwalk, CT). The AOAC procedure (1999) was used to analyze SO4 and soluble P in NAC and water. Determination of free water was conducted using the vacuum desiccation method (AOAC, 1999). After removing the free water, the water of hydration was determined by distillation according to the n-amyl alcohol method (Duncan and Brabson, 1969). Total water was obtained by the summation of free water and water of hydration.

Based on the elemental chemical analysis of P, Fe, K, SO4, and water of hydration, an empirical formula was calculated for the compounds. In the calculation it was assumed that SO4 was present as residual Fe2(SO4)3. This assumption was based on the fact that K2SO4 has a much higher solubility than Fe2(SO4)3 (Weast, 1989) and, consequently, should dissolve during the process of synthesis.

The two Fe-K-P compounds in the mother liquor after being washed with water and dried with acetone were examined using the Olympus Model PM-10A polarizing light microscope (Olympus Corporation of America, New Hyde Park, NY) to obtain the morphological and optical data including crystal system, habit, and refractive indices. Refractive indices were measured by the Becke line method (Wahlstrom, 1979) with certified refractive index oils.

Twenty-gram samples of H8-syn and H14-syn were prepared for x-ray analysis by grinding with 10 mL of Freon for 7 min in an impact grinder. The samples were scanned with a Siemens D-500 x-ray powder diffractometer (Bruker AXS, Inc., Madison, WI) from 5.0 to 65.0° 2{theta} using a step size of 0.02° 2{theta} and a dwell-time of 3 s, with Cu radiation produced at a power setting of 40 kV and 30 mA. The data were analyzed using Jade 5.0 software (Materials Data Inc., Livermore, CA) and the powder diffraction file (PDF) database (International Center for Diffraction Data [ICDD], 1999) to which XRD patterns for Fe-Al-P compounds as described by Frazier et al. (1991) were added.

Both compounds were examined by Perkin-Elmer Model 283 infrared spectrophotometer (Perkin-Elmer Corporation, Norwalk, CT) in compacted KBr pellets with sample concentration of 0.33% by weight. The KBr pellet was prepared by thoroughly mixing and grinding 1 mg of each compound with 100 mg of spectrographic-grade KBr and then mixing with an additional 200 mg of KBr to obtain the desired dilution. The KBr mixture was placed in an evacuable die, degassed, and compressed at 140 MPa to prepare a transparent pellet. The spectra were background corrected and normalized to 100%.

A portion of H8-syn and H14-syn was placed on individual sample stubs using double-stick tape and coated with C. The samples were scanned in an AMR Model 1000A (Advanced Metals Research Corporation, Bedford, MA) SEM with accelerating voltage of 15 to 30 kV using a Thermo NORAN EDS (NORAN Instruments, Middleton, WI) operating in the range of 0 to 10 kV. The scanning procedure consisted of looking for structures in the sample with defined micromorphology.

Solubility Study
A 0.01 M KCl solution with pH adjusted to 3.0, 4.0, 5.0, 5.5, 6.0, 6.5, 7.0, or 7.5 with HCl or KOH was used to study the solubility of H8, H14, and an apatitic Patos PR (110 g P kg-1) from Brazil. Thirty milliliters of each of these solutions was added to a plastic snap cap vial in which a 100-mg sample of each P source had been added. The samples were shaken continuously until the solution pH reached equilibrium near the target pH values. The pH was adjusted to the target pH every 24 h by adding drops of 0.01, 0.1, or 1.0 M solution of KOH or HCl. Amounts of OH- or H+, in mmol, added to the target pH solution and the final pH values at equilibrium are presented in Table 1.


View this table:
[in this window]
[in a new window]
  		Table 1. Amount of OH- or H+ added as KOH or HCl, respectively, to the target pH solution and final pH at equilibrium in the 0.01 M KCl solution.

 
After reaching equilibrium, the samples were filtered with a 0.2-µm syringe filter and P concentration was measured by the method of Murphy and Riley (1962). Phosphorus released from each P source was expressed as a percentage of the amounts of P added.

Agronomic Evaluation
Phosphorous Sources
To simulate fertilizers having a wide range of WSP, mixtures of MCP and H8-syn or H14-syn were prepared at 0, 25, 50, 75, and 100% of total P as MCP. All P source were thoroughly homogenized and compacted into 10-g pellets at 140 MPa. The resulting pellets were crushed and screened to particle sizes between 3.35 and 2.00 mm.

Greenhouse Experiment
A greenhouse study was conducted with a Hiwassee soil (thermic Rhodic Kanhapludults) containing 390 g kg-1 sand, 220 g kg-1 silt, and 390 g kg-1 clay (texture class: clay loam). The soil had 3 mg P kg-1 of resin-extractable P (van Raij and Quaggio, 1983), 1.7 mg P kg-1 of extractable P by Pi test (Menon et al., 1989), 10.4 cmolc kg-1 of cation-exchange capacity, pH 5.3 in water (1:1 ratio), and 18 g kg-1 of organic matter (Nelson and Sommers, 1996).

Granulated H8-syn, H14-syn, and MCP were mixed thoroughly with soil for upland rice and flooded rice at rates to supply 0, 10, 20, 40, and 80 mg P kg-1 as total P. The mixtures containing H8-syn or H14-syn and MCP were applied only at 40 mg P kg-1. In addition, a comparison of powdered and granulated H8-syn was made at 40 mg P kg-1 for flooded rice. Other nutrients were added at rates of 200 mg N kg-1 as urea and 200 mg K kg-1 as KCl. A nutrient solution was added to supply 96 mg Mg, 135 mg S, 5 mg Cu, 11 mg Zn, 8 mg Mn, 2 mg B, and 0.2 mg Mo per kilogram of soil. The experiment was conducted with a randomized complete block design with three replications for each treatment.

Eight seeds of upland rice (cultivar IR-47686) were planted per pot of 3 kg of soil at a depth of {cong}1 cm and subsequently thinned to two per pot 10 d after germination. The pots with upland rice were watered using de-ionized water to maintain 75% field moisture capacity during the entire experiment. Flooded rice (cultivar IR-36) seedlings were grown for 4 wk and then transplanted with one hill of four plants per pot of 7 kg soil. The soil was flooded for 2 wk before incorporating P sources followed by transplanting of rice seedlings. After transplanting, the flooded water was kept at 2 cm of water above the soil surface until the plants were harvested.

The plants were harvested by cutting the stalk 65 d after seed planting or rice transplanting, followed by drying at 60°C for 2 wk, weighed, and ground. The plant tissue samples were digested with H2SO4–H2O2 (Linder and Harley, 1942) and P concentration was determined by the method of Murphy and Riley (1962).

Data Analysis
The relationship between dry-matter yield or P uptake and rate of P applied was evaluated using regression procedures (SAS Institute, 1985). A dummy variable multiple regression analysis as described by Chien et al. (1988) was performed. The dummy variable takes the value of 1 for the P source being considered and 0 for other P sources. This resulted in a common intercept and a single value of root mean square of error (RMSE) and R2 for the three regression equations (one for each P source). Three response functions (linear, semi-log, and square root) were tested to describe the relationship between the parameters studied, and the one with the lowest RMSE value was chosen. The models tested were as follows:

[1]

[2]

[3]
where Yi is the dry-matter yield or P uptake obtained with source i, X is the rate of P applied, ßi is the slope of the response function for source i, ßo is the common intercept, and {epsilon}i is the error term of the fitted model.

The relative agronomic effectiveness (RAE) was calculated for H8-syn and H14-syn as the ratio of the two slopes for a given model:

[4]
where ßi is the slope of the response function of H8-syn or H14-syn and ßMCP is the slope of the response function of the standard source of P (MCP). This expression ranks H8-syn and H14-syn with respect to MCP according to their agronomic potential to produce a yield response (Chien et al., 1990). Use of RAE made possible the comparison of the effectiveness of P sources between upland rice and flooded rice, even when utilizing different pot sizes, rice cultivars, and planting methods.

To determine if there was a statistically significant difference between two P sources in the range of P rates applied, an F value (= t2) was calculated according to the formula:

[5]
where ßia is the slope of the response function for the first P source tested, ßib is the slope for the second P source tested, SE(ßia) is the standard error for ßia, and SE(ßib) is the standard error for ßib.

The relationship between dry-matter yield and WSP for the mixtures of H8-syn or H14-syn with MCP was evaluated using a quadratic-plateau response model (SAS Institute, 1985). The model implies that if a plateau can be found in the range of WSP rates applied, the regression model can be used to explain dry-matter yield response to WSP. It assumes Xo as WSP where the plateau (Pl) is reached when Y = A + BX + CX2 if X <= Xo and Y = Pl if X > Xo. This is possible when the two sections meet at Xo and the curve is continuous and smooth (first derivative with respect to X is the same at Xo), that is, Xo = -B/2C and Pl = A - B2/4C. This procedure determines if no response will occur after a certain percentage of WSP (Xo) in the fertilizer mixtures.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Characterization
Values of the concentrations of total P, Fe and K, water of hydration and the calculated formula (Table 2) are very close to those expected for H8 and H14 when considering the chemical composition of its empirical formula. The amounts of SO4 in H8-syn (1.11%) and H14-syn (0.11%) were accounted as Fe2(SO4)3·7H2O remaining in the materials when calculating the formula of each compound.


View this table:
[in this window]
[in a new window]
  		Table 2. Chemical composition of H8 and H14 synthesized at laboratory conditions.

 
The AOAC (1999) defines available P as the sum of WSP and citrate-soluble P. Based on this definition, 95 and 97% of the total P were available for H8-syn and H14-syn, respectively (Table 2). The amounts of WSP of both compounds were very low and therefore, their high values of available P were mainly due to citrate-soluble P. The high citrate-soluble P of H8-syn seems to contradict the results presented by Lehr et al. (1967) and Frazier et al. (1991). They described this compound as citrate insoluble and a very poor source of P for plants. These authors referred to the agronomic data of Lindsay and Taylor (1961) and Taylor et al. (1960), but none of the cited reports showed the actual data for citrate-soluble P. Only agronomic data revealed the poor performance of this compound as a source of P for maize (Zea mays). Gilkes and Lim-Nunez (1980) reported a low citrate-soluble P (3.7%) of H8-syn that was prepared by the method proposed by Lehr et al. (1967). It is possible that our higher citrate-soluble P in H8-syn than that reported by Gilkes and Lim-Nunez (1980) might be due to different crystallite sizes of H8-syn prepared by different methods. Unlike H8-syn, the high citrate solubility of H14-syn obtained in our study is in agreement with Lehr et al. (1967) and Frazier et al. (1991).

Lehr et al. (1967) and Frazier et al. (1991) referred to the H8 compound as citrate insoluble, which was probably based on the observed poor agronomic effectiveness obtained by Taylor et al. (1960) and the discussion by Lindsay and Taylor (1961). The conclusion that a P source should be plant available solely based on its high citrate solubility, or vice-versa, can be misleading. It is important to consider that citrate is a chelating ligand that can complex Fe with a high formation constant (pK = 12.3), higher than the formation constants of citrate with Ca (pK = 4.3) or Al (pK = 9.1) (Norvell, 1991). Also Fe-Al-P minerals are more soluble in solutions having a high pH than in those having a low pH (Lindsay, 1979). The solubility of these Fe-P compounds is expected to be higher than that of Ca-P compounds in NAC solution. Therefore, it is likely that H8-syn, which contained Fe-P compounds, could be readily soluble in NAC because of complex formation of Fe-citrate. However, P released from Fe-P compounds by the NAC procedure is not necessarily plant available in the soil system. Chien (1993) suggested that solubility results for apatitic PR containing Ca-P should be interpreted with caution when compared with non-apatitic PR containing Fe-Al-P. Prochnow et al. (1998) also found a very poor relationship between citrate-soluble P of different sources of SSP containing varying amounts of Fe and Al and their agronomic effectiveness for maize. They concluded that the AOAC method was not an adequate index for evaluating the P availability of fertilizers with high amounts of Fe-Al-P compounds.

Both synthesized compounds were found to be essentially homogeneous with the optical data identical to H8 (hexagonal rods; uniaxial (+); {omega} = 1.595, {epsilon} = 1.601) and H14 (orthorhombic-hexagonal shaped plates; biaxial (-); {alpha} = 1.556, ß = 1.600, {gamma} = 1.606).

X-ray powder diffraction patterns obtained for the H8-syn and H14-syn samples and x-ray patterns for these compounds from PDF file (ICDD, 1999) or from the publication of Frazier et al. (1991) are similar in terms of peak location and intensity (Fig. 1) . There is an apparent difference when comparing the x-ray spectra for the H8 compound after 46° 2{theta}. This is due to the fact that the original paper published by Haseman et al. (1950) on the x-ray patterns for this compound, which is used by the PDF, contained information only up to 46° 2{theta}. New x-ray diffractometry studies may be necessary if more precise peak location and intensity are needed in the range of 46 to 65° 2{theta} for the H8 compound.



View larger version (31K):
[in this window]
[in a new window]
  		Fig. 1. X-ray diffractogram and powder diffraction file (PDF) data for compounds H8-syn (A) and H14-syn (B).

 
The infrared spectra for the samples of H8-syn and H14-syn (Fig. 2) have similar peak locations and intensities as reported by Lehr et al. (1967). The SEM photomicrographs and the EDS analysis of specific areas of crystals in the samples of H8-syn and H14-syn are shown in Fig. 3 and 4 , respectively. The morphology of the compounds, rod crystals for H8-syn and pseudo-hexagonal plates for H14-syn are in accordance with the morphology suggested by Frazier et al. (1989). They showed the illustrated photomicrographs of Fe3H9(PO4)6·6H2O (H9) and Fe3H15(PO4)8·4H2O (H15) and suggested that H8 and H14 would have similar morphology to H9 and H15, respectively. Our EDS analysis of specific areas of both crystals suggests that the compounds consisted basically of Fe, K, and P. No other morphological structures were found in the samples of H8-syn and H14-syn shown in Fig. 3 and 4.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 2. Infrared spectra for compounds H8-syn (A) and H14-syn (B).

 


View larger version (44K):
[in this window]
[in a new window]
  		Fig. 3. SEM photomicrograph of subsample of the H8-syn showing (A) rod crystals of Fe2.8K1.1H8(PO4)6·4.1H2O (scale = 15 µm) and (B) EDS analysis of the specific square area.

 


View larger version (46K):
[in this window]
[in a new window]
  		Fig. 4. Scanning electron microsope (SEM) photomicrograph of subsample of the H14-syn showing (A) pseudohexagonal plates of Fe3.0K0.9H14(PO4)8.4.3H2O (scale = 15 µm) and (B) EDS analysis of the specific square area on top of the main crystal.

 
Results of chemical analysis, optical microscopy, x-ray difractrometry, infrared, morphology of the compounds and EDS analysis of areas of the crystals confirm that the method of preparation was successful and that H8-syn and H14-syn can be considered as Fe3KH8(PO4)6·6H2O and Fe3KH14(PO4)8·4H2O, respectively. The proposed procedure yielded 45 g of H8-syn in three cycles and 109 g of H14-syn in two cycles, an adequate amount for agronomic evaluation under greenhouse conditions.

Solubility Study
The amounts of P released in 0.01 M KCl from Patos PR, H8-syn, and H14-syn as influenced by pH at equilibrium are shown in Fig. 5 . It shows that P released from Patos PR decreased with increasing pH. At pH 5.5, almost no dissolution of Patos PR was observed. This is consistent with solubility studies on other PR sources that show a decrease in solubility as pH increases (Chien et al., 1975). Also, the results are in agreement with agronomic evaluation of apatitic PR that generally shows increasing PR effectiveness with decreasing pH (Chien and Menon, 1995; Engelstad et al., 1974; Khasawneh and Doll, 1978).



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 5. Phosphorus (% of total P) released at equilibrium from Patos phosphate rock, H8-syn, and H14-syn as related to pH of a 0.01 M KCl solution.

 
Unlike Patos PR, a positive relationship between P released and solution pH was observed with H8-syn and H14-syn (Fig. 5). The results also show that P released from H8-syn was significantly lower than H14-syn within the pH values tested. For example, 49% of the total P in H14-syn was released at pH 6.0 whereas only 11% was released from H8-syn. The results of the solubility study suggest that agronomic effectiveness of P fertilizers containing high amounts of Fe-K-P compounds, similar to H8-syn and H14-syn, would increase as soil pH increases. It is also expected that P fertilizers containing H14 would perform better than fertilizers containing H8. The present solubility study thus supports the conclusion by Taylor et al. (1960) that H8 was a poor source of P for maize. It also supports the suggestion by Lehr et al. (1967) and Frazier et al. (1991) that H14 would be a better source of available P to plants than H8. The solubility data in Fig. 5 further suggest that non-apatitic PR containing Fe-Al-P can be more agronomically effective than apatitic PR containing Ca-P in alkaline soils. In fact, this has been reported by Chien (1999), who found that a calcined non-apatitic Christmas Island C-grade PR containing Fe-Al-P minerals (crandallite and millisite) was a good source of P for upland rice grown on an alkaline soil (pH 7.9) whereas a reactive apatitic Gafsa PR was totally ineffective.

Both H8-syn and H14-syn were highly citrate soluble and had about the same citrate solubility (about 95% of total P) (Table 2). However, data in Fig. 5 show that H8-syn had a lower solubility in 0.01 M KCl than H14-syn within the pH range of 3.0 to 7.5. This suggests that some Fe-P compounds can exhibit high solubility in NAC as previously discussed, but low solubility in soil solution. As the fertilizer industry more intensively uses marginal-grade PR to manufacture P fertilizers with lower water solubility and higher amounts of Fe-P impurities, the AOAC method may not be suitable to estimate the P availability of such fertilizers to crops.

As shown in Table 1, increasing equilibrium solution pH for both Fe-K-P compounds required base to maintain pH and resulted in more P released (Fig. 5). This indicates that the reactions involved in the incongruent dissolution of either compound release H+ and P. Furthermore, no Fe was found in the solutions at equilibrium with either Fe-K-P compounds. Based on the present results and the information provided by Taylor et al. (1960) and Dillard and Frazier (1983), the following reactions are proposed to explain the hydrolysis of H8-syn and H14-syn:

[6]

[7]

These reactions indicate that the hydrolysis of H14-syn should result in a lower solution pH with more P released than H8-syn (Fig. 5 and Table 1) because more H3PO4 is produced from H14-syn (4 mol) than that from H8-syn (2 mol). The absence of Fe in solution at equilibrium can be explained by the precipitation of FePO4·nH2O, which is almost water-insoluble.

Agronomic Evaluation
The square root model gave the best description of the relationship between dry-matter yield and P rate for both crops, whereas the linear model was better for the relationship between P uptake and P rate. Regression estimates for the models and the efficiency of H8-syn and H14-syn in providing P to the plants as compared with MCP are presented in Table 3 and Fig. 6 and 7 . Phosphorus uptake efficiency followed the order: MCP > H14-syn > H8-syn for upland rice and MCP = H14-syn > H8-syn for flooded rice. Both H8-syn and H14-syn were more effective in increasing dry-matter yield and P uptake when applied to flooded rice than to upland rice. For example, RAE of H8-syn increased from 33% for upland rice to 75% for flooded rice in P uptake. The corresponding increase in RAE of H14-syn was from 73% for upland rice to 104% for flooded rice.


View this table:
[in this window]
[in a new window]
  		Table 3. Regression estimates for the models describing the relation between dry matter-yield (DMY) or P uptake and P rate, and relative agronomic effectiveness (RAE) of H8-syn and H14-syn with respect to MCP for upland and flooded rice.

 


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 6. Dry-matter yield of upland rice (A) and flooded rice (B) as affected by P source and P rate applied. Models followed by the same letter indicate their slopes are not statistically different (p <= 0.05).

 


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 7. Uptake of P by upland rice (A) and flooded rice (B) as affected by P source and P rate applied. Models followed by the same letter indicate their slopes are not statistically different (p <= 0.05).

 
The better performance of H14-syn than H8-syn for both rice crops can be explained by the solubility study of H8-syn and H14-syn. More P was released from H14-syn than H8-syn in 0.01 M KCl with a pH range from 2.9 to 7.5 at equilibrium (Fig. 5). Our results are in agreement with Taylor et al. (1960), Lindsay and Taylor (1961), and Frazier et al. (1991). Taylor et al. (1960) and Lindsay and Taylor (1961) referred to the H8 compound as having low agronomic performance as a source of P. More recently, Frazier et al. (1991) described H8 and H14 compounds as a very poor source and readily available source of P for plants, respectively.

The higher agronomic performance of H8-syn and H14-syn when applied to flooded rice may have resulted from increased dissolution of Fe-P by a reduction of Fe+3 to Fe+2 on flooding, the same mechanism for the dissolution of soil native Fe-P minerals (Willet, 1991). Additionally, the solubility of H8-syn and H14-syn increased with an increase in solution pH (Fig. 5). Since flooding soil generally increases soil pH, H8-syn and H14-syn would have higher P dissolution and agronomic performance in flooded soil than upland soil.

According to the AOAC definition, citrate-soluble P is considered plant available. Both H8-syn and MCP had very high citrate solubility (95 and 100% of total P, respectively). However, H8-syn was much less effective than MCP in increasing dry-matter yield or P uptake (Table 3). The present results thus question the suitability of the AOAC definition of plant available P that has been adopted by the fertilizer industry and government legislation for many years. As explained by Chien (1993), the high citrate solubility of Fe-P compounds is mainly due to the formation of a stable Fe-citrate complex that often overestimates its actual plant availability in soil. The low solubility of H8-syn in 0.01 M KCl correlated better than NAC solubility with the low agronomic effectiveness of H8-syn as observed in the present study.

Table 4 shows that dry-matter yield and P uptake obtained with powdered H8-syn were significantly higher compared with the granular form for flooded rice. This suggests that acidulated P fertilizers containing high amounts of H8 compound would perform better as a P source to plant when applied in powder form. Future research will be needed to compare the agronomic effectiveness of powdered H8-syn and powdered SSP or TSP containing the H8 with respect to MCP for upland and flooded rice.


View this table:
[in this window]
[in a new window]
  		Table 4. Agronomic effectiveness of H8-syn applied at 40 mg total P kg-1 in powdered and granular form for flooded rice.

 
Segmented regression models, WSP required to attain the plateau of dry-matter yield, and 90% of the plateau are presented in Table 5. The plateau was attained at 66.6% of WSP in the mixtures of H8-syn with MCP and 70.9% of WSP in the mixtures of H14-syn with MCP for upland rice. For flooded rice, 29.3 and 39.1% of WSP were required in the mixtures of H8-syn or H14-syn with MCP, respectively, to reach the plateau. To reach 90% of the plateau, it required even lower WSP content, 42.7% in the mixtures of H8-syn and MCP and 34.6% in the mixtures of H14-syn with MCP for upland rice. For flooded rice, the requirements were only 16.7% in the mixture of H8-syn with MCP and 10.6% in the mixture of H14-syn with MCP.


View this table:
[in this window]
[in a new window]
  		Table 5. Quadratic-plateau model for mixtures of H8-syn or H14-syn with MCP for the maximum dry-matter yield (DMY) of upland and flooded rice and the relationship between DMY and percentage of water-soluble P (WSP) in the mixtures required to attain plateau or 90% of plateau.

 
The present results do not support the suggestions by Gilkes and Lim-Nunez (1980) and Frazier et al. (1991) that water-insoluble but citrate-soluble Fe-K-P compounds in acidulated P fertilizers should be deemed as undesirable for plant growth. These impurity P compounds not only can provide some plant available P by themselves but also their P availability may be further enhanced by the presence of MCP. Future research is needed to confirm this enhancement effect of WSP on P availability of Fe-K-P compounds.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Chemical and instrumental analyses confirm that the method of preparation was successful in obtaining Fe3KH8(PO4)6·6H2O (H8-syn) and Fe3KH14(PO4)8·4H2O (H14-syn). The procedure is suitable to prepare sufficient amounts of these Fe-P impurities present in commercial SSP and TSP for greenhouse evaluation. Dissolution of these compounds increased with increasing equilibrium pH in 0.01 M KCl. The amounts of P released from H14-syn were higher than H8-syn within the pH range tested. These results suggest that the agronomic effectiveness of fertilizers containing these compounds will increase as soil pH increases. It is also expected that the P fertilizers containing H14-syn will perform better than those containing H8-syn for crop production.

The agronomic results show that the Fe-P impurity compound in the form of H14-syn was more effective than the form of H8-syn as a source of P in increasing dry-matter yield and P uptake by upland and flooded rice. Both H8-syn and H14-syn compounds were more effective when applied to the flooded soil than to the upland soil. Flooding the soil can create a reduced soil condition and an increase of soil pH that favor dissolution of Fe-P. The present results also show that acidulated fertilizers with only about 11 to 17% of WSP were able to produce 90% of maximum dry-matter yield of flooded rice attained by MCP with 100% WSP. For upland rice, fertilizers required higher P water solubility ranging from about 35 to 43%. All these requirements are lower than that required by current legislation (90–93% WSP) adopted by some countries. The water-insoluble but citrate-soluble Fe-P compounds present as impurities in acidulated P fertilizers should not be deemed as totally undesirable for crop production.


   ACKNOWLEDGMENTS
 
The senior author expresses his gratitude to FAPESP (Foundation for Supporting Research in the State of São Paulo, Brazil) for the research scholarship that made this study possible and to the staff of the International Fertilizer Development Center (IFDC) for their help during the study. He also thanks Dr. T. Ranatunga of Alabama A&M University for infrared spectra of H8-syn and H14-syn.

Received for publication January 29, 2002.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Agron. J.Home page
E. A. B. Francisco, S. H. Chien, L. I. Prochnow, E. R. Austin, M. C. M. Toledo, and R. W. Taylor
Characterization and Greenhouse Evaluation of Brazilian Calcined Nonapatite Phosphate Rocks for Rice
Agron. J., May 7, 2008; 100(3): 819 - 829.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
L. I. Prochnow, S. H. Chien, G. Carmona, E. F. Dillard, J. Henao, and E. R. Austin
Plant Availability of Phosphorus in Four Superphosphate Fertilizers Varying in Water-Insoluble Phosphate Compounds
Soil Sci. Soc. Am. J., January 25, 2008; 72(2): 462 - 470.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Prochnow, L. I.
Right arrow Articles by Taylor, R. W.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Prochnow, L. I.
Right arrow Articles by Taylor, R. W.
Agricola
Right arrow Articles by Prochnow, L. I.
Right arrow Articles by Taylor, R. W.
Related Collections
Right arrow Soil Chemistry
Right arrow Soil Fertility and Productivity
Right arrow Wetland Soils

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome