Environ. Sci. Technol. 2005, 39, 4369-4376
Analysis of Reduced Phosphorus in Samples of Environmental Interest SIYUAN C. MORTON,† DIETMAR GLINDEMANN,† XIAORONG WANG,O XIAOJUN NIU,O AND M A R C E D W A R D S * ,† Department of Civil and Environmental Engineering, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061, and State Key Laboratory of Pollution Control and Resource Reuse, Nanjing University, Nanjing, P. R. China
The combination of ion chromatography (IC) and inductively coupled plasma emission spectroscopy (ICP-ES) was used for the sensitive and specific detection of hypophosphite (PO2), phosphite (PO3), methylphosphonic acid (MPA), and phosphate (PO4). Application of this technique to a wide range of environmental samples proved that reduced phosphorus was present in some situations including process water from thermal phosphorus plants, drinking water contacting cast iron, and phosphorus corrosion inhibitor used in water treatment and in sewage wastewater. Preliminary testing did not detect high concentrations of reduced phosphorus and phosphine in situations where it was previously reported to be very important, including anaerobic digesters in wastewater treatment plants. The new IC-ICP-ES technique is a promising tool for use in corrosion and soil research where phosphites are likely to be present.
spread use. Reduced phosphorus products are routinely used in environmental applications including fertilizer, fungicide, insecticide, herbicide, rodenticides, fumigants, flame retardant, chemical intermediates, or additives to many industrial products (1). Specifically, phosphites were promoted for use as a superior phosphorus fertilizer after World War II (6, 7), with purported nutritional benefit supplementing its wellknown fungicide properties. Use of phosphites recently received another boost, since it is the only approved treatment for Sudden Oak Death (8) and it is an approved fertilizer for “organic” farming. At present, there are both positive and negative opinions about using phosphites as fertilizers (7, 9-11), but in the meantime its use is increasing. Another reduced phosphorus compound is phosphine (hydrogen phosphide, PH3), a compound approximately 2 times more toxic to humans than hydrogen cyanide (12). Some researchers have postulated that PH3 volatilization is the major cause of phosphorus removal in sewage plants. For instance, in a widely cited Nature article, Devai attributed up to 50% of phosphorus removal in one sewage plant to PH3 emissions (13, 14). Oddly, environmental engineers responsible for phosphorus removal in sewage plants have never mentioned PH3 volatilization as a possible removal mechanism. In addition to inorganic reduced phosphorus, some researchers have claimed that organic phosphorus species such as phosphonates also represent a manifestation of reduced phosphorus (1, 15, 16). Some of these compounds are quite common. For example, glyphosate is one of the most significant herbicides in the world. It accounts for more than 60% of the global nonselective herbicide sales of U.S. $1.2 billion annually (17). Methylphosphonic acid (MPA, although not a common phosphonate) is the simplest phosphonate species and is a possible degradation product of chemical warfare agents and other phosphorus products (18). This study will focus on examining the presence of reduced phosphorus in the environmental samples of interest.
Introduction Phosphorus chemistry is critical to understanding behavior of natural and engineered systems. In many research fields including environmental engineering, soil testing, and aqueous corrosion, it is nearly always assumed that phosphorus in natural systems occurs exclusively in the +5 oxidation state as orthophosphate, polyphosphates, organophosphates, and particulate phosphates. However, a thorough review (1) has illustrated that the environmental phosphorus cycle is known to occasionally include many reduced phosphorus species such as phosphides (-3), diphosphide (-2), tetraphosphide (-0.5), elemental phosphorus (0), hypophosphite (phosphinates, +1), phosphite (phosphonates, +3), and so forth. In this paper, we term any phosphorus species with an oxidation state lower than (+5) to be “reduced phosphorus.” Reduced phosphorus may be introduced to the environment from many different sources, for example, from corroding metals such as iron (1-5). Steel and steel slags are also major sources of reduced phosphorus (1). In addition, reduced phosphorus is produced during thermal processing of phosphate ore, and industrial products including elemental phosphorus, hypophosphite, and phosphine are in wide* Corresponding author phone: (540)231-7236; fax: (540)231-7916; e-mail:
[email protected]. † Virginia Polytechnic Institute and State University. O State Key Laboratory of Pollution Control and Resource Reuse. 10.1021/es0401038 CCC: $30.25 Published on Web 05/06/2005
2005 American Chemical Society
Phosphorus Analysis. In traditional environmental engineering (water and wastewater treatment), the standard phosphorus speciation method operationally divides phosphorus into reactive P, acid hydrolyzable P, and organic P. These fractions can be further subdivided into soluble and total fractions. There is no consideration that reduced phosphorus species might be present (19, 20). The most common methods of phosphorus analysis and speciation are based on colorimetric methods (20). However, they all involve conversion of phosphorus species to orthophosphate through digestion and can only be used to detect reduced phosphorus species by difference in concentration before and after digestion. In P analysis conducted in soil and steel making fields, the operational definition is different, but P is still detected as phosphate. For example, on the basis of the specific chemical terminology, P is divided into water extractable, CaCl2 extractable, 0.1 M NaOH extractable, HCl extractable, Mehlich extractable P. None of these standard approaches could identify reduced phosphorus species if they were present. Many analytical techniques have been developed to measure glyphosate and its metabolite in the environment. Some of them are very sensitive. For example, Wigfield reported a simplified liquid chromatographic technique with a detection limit of 1 ug/L (21). Ion chromatography (IC) with conductivity detection has been used to separate hypophosphite, phosphite, and VOL. 39, NO. 12, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4369
phosphate. However, its low sensitivity and interferences prevent its use to relatively complicated environmental matrixes (22). Researchers have developed other approaches to measuring reduced phosphorus species. These include paper chromatography (23), high-performance liquid chromatography with a flow injection system as postcolumn reaction detector (HPLC) (24), gas chromatograph (GC) (25), combined gas chromatograph-mass spectroscopy (GC-MS) (26), and high-performance ion chromatography (27). Even though each had a different analytical focus and some drawbacks, they offer the possibility to quantify different reduced phosphorus in various environmental samples (Table 1). Inductively coupled plasma emission spectroscopy (ICPES) offers reasonable sensitivity for P containing compounds and detects all common phosphorus oxidation states equally (20). In this work, a new method combining ion chromatography (IC) and inductive coupled plasma-emission spectrum (ICP-ES) was developed. It was aimed at identification and quantification of some inorganic and organic reduced phosphorus in water samples including hypophosphite, phosphite, and methylphosphonic acid (MPA). It was also tested on glyphosate analysis. This method was applied to a wide range of environmental samples to shed light on the occurrence of reduced phosphorus compounds in natural and engineered systems.
Materials and Methods IC-ICP System. A Dionex IC (DX 300) was connected to a JOBIN-YVON 2000 ICP-ES to separate and measure reduced P species such as hypophosphite (PO2), phosphite (PO3), and methylphosphonic acid (MPA). The ICP was operated in a manual mode. Na2CO3 (3 mM) pre-degassed with helium (U.S. EPA method 300.1) was used as IC eluent and was pumped through IC and ICP at a flow rate of 1.5 mL/min. Three columns including a metal-free trap column (MFC), an anion exchange AG10 guard column, and an AS9 analytical column were connected in sequence for the IC. The metalfree trap column was specially designed for online cleanup of ionic transition metals to protect the other two IC columns from metal impurities in the samples. Several ICP wavelengths for phosphorus analysis were tested. The wavelength near 177.4 nm was selected because it had the lowest and most stable detection limit with the least interference from cations, such as calcium, that coelute with hypophosphite. The typical detection limit was 10 ug/L in injected samples. Samples with phosphate concentration higher than 1 mg/L were diluted to about 100 ug/L. Standard solutions containing sodium hypophosphite monohydrate (99%, Acros), methylphosphonic acid (98%, Sigma-Aldrich), phosphorous acid (98%, Acros), and sodium phosphate dibasic anhydrous (100.1%, Fisher Scientific) were used for calibration and spikes (Figure 1). These four phosphorus species eluted in less than 6 min. The oxidation of hypophosphite to phosphite and phosphate by permanganate ion has been studied (28-31). A possibility of applying this mechanism to provide further verification of the existence of hypophosphite in samples was investigated. To illustrate its use, at room temperature about 22 °C, 2.5 mg/L KMnO4-Mn was spiked into 100 ug/L phosphorus standards. After 20 min, 40% of the hypophosphite was oxidized to phosphite (∼30%), and 10% was oxidized to phosphate (Figure 2). Spiking KMnO4 can therefore confirm the presence of hypophosphite even if high concentrations of salts such as calcium (e.g., 1800 mg/L Ca) and chloride (e.g., 1000 mg/L Cl) were to coelute and provide a positive interference. In initial tests, samples were stored in a 4 °C refrigerator for several weeks before analysis with the IC-ICP. Samples were tested with and without spikes of phosphite and hypophosphite. No losses of phosphite or hypophosphite 4370
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 12, 2005
were observed in filtered (with 0.45-µm pore size filter) samples maintained at cold temperatures, indicating that these compounds are relatively stable under the conditions studied. Phosphine Analysis. GC/MS with a cryofocussion was used to analyze phosphine gas. The protocols were as follows: 100 uL gas was directly injected into a flexible fused silica capillary column (5% phenyl silicon phase, 0.25-mm inner diameter). The first half-meter of the column was inserted in a stainless steel loop and carried through a hole on top of the GC into a removable Dewar jar filled with liquid nitrogen (-196 °C or -321 °F). After 2 min of focusing, the loop was lifted out of the liquid nitrogen and heated (to 150 °C or 302 °F) with a resistance heater. Phosphine was identified by its mass spectrum (masses 31-34) and quantified by comparison with a certified gas standard (14 mg P/m3). The detection limit of this method is 0.1 mg/m3 PH3-P. Production and Collection of Samples to be Tested. A wide range of environmental samples were examined on the basis of a literature review that establishes where reduced phosphorus might be present (1, 32). Liquids were filtered with 0.45-µm pore size filter before IC-ICP analysis. Three general types of samples were collected. The first were liquid samples from various sources including a phosphorus plant, soils spiked with reduced phosphorus, sewage treatment plants, and iron corrosion experiments. Preparation of each liquid sample is described in the following paragraphs. (1) Samples from an inoperable elemental phosphorus plant were tested to evaluate the application of the IC/ICP method on samples with high concentrations of reduced P. Two sources of process water and three kinds of treated process water (with different methods, Table 2) were tested. Because the concentration of reduced P compounds, such as hypophosphite and phosphate, in the samples was high, it was necessary to dilute the samples. A Standard Method protocol (4500-P) persulfate digestion and ascorbic acid colorimetric measurement (19) was used to measure the total phosphorus for comparison to the total measured by ICICP. (2) Five 500-mL solutions of different phosphorus species (500 µg/L P) were exposed to soil to analyze their behavior and use of the method in soil research. These solutions included (a) distilled-deionized water as a control solution, (b) sodium phosphate dibasic solution, (c) “nitri-phite” magnum phosphite fertilizer 2-40-16 (Biagro Western Sales, Visalia, CA), (d) methylphosphonic acid, and (e) glyphosate. Two kinds of soils were tested, one of which (soil 1) was not supporting plant growth because of obstruction from light and another (soil 2) was sediment from near a pond with a high density of ducks. The phosphorus solutions were mixed with 100 g of each soil separately, except that glyphosate was only mixed with soil 1. Samples were kept at room temperature and mixed once every day. On Day 1 and Day 7, samples were taken for the IC/ICP analysis. (3) To test for possible bioproduction of reduced P during wastewater treatment (13, 14), influent and effluent samples were collected from major processes at two sewage treatment plants including screening, primary clarifier, anaerobic reactor, aerobic activated reactor, nitrification reactor, and anaerobic digesters. These samples were immediately filtered, stored at 4 °C, and then analyzed for reduced phosphorus within 24 h. (4) The analytical method was also applied to several preliminary experiments involving iron corrosion. First, a sample was set up in 10-3 M NaCl solution to test the production of reduced phosphorus from cast iron filings by adding 10 g cast iron into 50 mL 10-3 M NaCl solution. Two to three milliliters of the solution was collected and filtered at Day 7, 21, 71, 134, and 139 and then was tested for phosphite
TABLE 1. Comparison of Different P Analysis Methods reference
method
P species analyzed
method principle
detection limit
VOL. 39, NO. 12, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
Ryder, D. S. (22)
single-column ion chromatography (IC)-conductivity detector
hypophosphorous acid, phosphorous acid, orthophosphoric acid (or salts)
single column IC, direct analysis
0.2 mg/L
Mehra, M. C. and Pelletier, C. (40)
IC using indirect ultraviolet detection mode
hypophosphite, phosphite, and orthophosphate
0.5-1.5 mg/L
Roos, G. H. P et al. (27)
high-performance ion chromatography (HPIC)
phosphite (in plant samples)
single-column IC with UV detector, indirect ultraviolet mode single-column HPIC
Hirai, Y. et al. (24)
high-performance liquid chromatography (HPLC) with a flow injection system as postcolumn reaction detector (spectrophotometer) gas-liquid chromatography (GC)
sodium hypophosphite, sodium phosphite, and potassium phosphate
reduced phosphorus species were separated by HPLC, oxidized to and detected as orthophosphate
sub-mg/L
elementary phosphorus
P was partially subtracted by suitable organic solvent, then detected by GC
2 × 10-6 mg/L
capillary column GC-flame ionization detection capillary column GCnitrogen/phosphorus detection or ion trap
phosphonic acids
detected as trimethylsilyl (TMS) derivatives detected as tert-butyldimethylsilyl (tBDMS) deravatives
1 Ng as TMS derivative (1- uL injection volume)* 0.3 Ng as tBDMS derivative (1- uL injection volume) or 0.1 Ng (3-uL injection volume)*
Addison, R. F. and Ackman, R. G. (25) Graaf, R. M. et al. (41, 42) Smillie, R. H. et al. (26)
* Simple, possible interference from some cations.
phosphite
3-5 mg/L
note* a. simple b. possible interference from some cations, such as Cl-, F-, NO3-, SO42-, etc. a. less interference compared to direct IC detetion b. not very sensitive a. cost effective, relatively high sample throughput b. not very sensitive not simple
a. sensitive b. subtraction efficiency could vary; specific for elemental phosphorus TMS derivative not stable a. sensitive, tBDMS derivatives pretty stable b. rigorous and complicated
9
4371
FIGURE 1. Chromatograph of 100 ug/L phosphorus standard mixture (IC-ICP method).
FIGURE 2. Oxidation of hypophosphite by KMnO4 (on 100 ug/L standard mixture) before spiking in KMnO4; (b) 20 min after spiking in 2.5 mg/L KMnO4-Mn.
TABLE 2. Detailed Information of Wastewater Treatment Plants and Their Anaerobic Digesters wastewater treatment plant number 1 2 3 4 5 6 7 8 9 10 11
state/province
digester temperature (°C)
digester SRT (day)
type of industrial/nondomestic wastewater streams
% of industrial/ nondomestic wastewater
Virginia Pennsylvania California Iowa Virginia Delaware California California California California Ontario
37 36 36 30 36 34 37 52 37 36 36
38 20 27 40 27 28 14 16 21 22 22
restaurants, hotels, offices not available hospitals, metal plating, food processing packing houses, food processing food processing metal processing, volatile organic compounds metal plating, electronic industries food processing, metal processing laundry refineries, metal plating food processing
not available not available 6 34 10 9 5 15 9 20 7
and hypophosphite. In another test, a control group included four different types of 20-mL samples: pure distilled and deionized H2O, 10-3 M NaCl, 10-3 M HCl, and 10-3 M NaOH. Four grams of cast iron filing were added to each of these background solutions. Three parallel groups of samples were 4372
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 12, 2005
run under identical conditions but with an initial 100 ug/L P spike of hypophosphite, phosphite, and phosphate. This experiment was conducted at 67-85 °C. Samples were collected and filtered for the analysis. The third experiment was a repeat of the last one, but exposure was in an autoclave
TABLE 3. Result for Samples from Phosphorus Production Plantc,d IC/ICP (mg/L P) sample treatmentb
multistage treated with H2O2 raw water another raw water
hypophosphite 10.2 ( 0.64 1710 ( 80 4900 ( 480
0.65a
phosphite 2.30 ( 0.26 N/A 445 ( 130a 1550 ( 100a
colorimetric (mg/L P) phosphate
1.53 ( 6890 ( 140 2500 ( 360 1110 ( 130
0.33a
total P
total P
reactive P
14.06 6896.52 4661.12 7607.7
14.75 ( 0.14 6250 ( 130 6170 ( 72 10400 ( 140
0.7 ( 0.03 5870 ( 58 3000 1450
a not correcting matrix affect because spike-recovery was outside range 100(20%. b H O +lime treated process water blended with untreated 2 2 process at 100:1 ratio, then treated with FeCl3. c Standard deviations were based on triplicate measurements. Treated water (with H2O2) only did d one measurement for hypophosphite because of the high PO4-P content. If not specified, standard deviations of Hach measurements were zero.
at 121 °C and 32 psi for 1 h, the (duplicated) sample volume was 10 mL instead of 20 mL, the initial phosphate spike was 500 ug/L, and no phosphite or hypophosphite spike was tested. (5) Some other phosphorus chemical products, including normal phosphate fertilizer and representative phosphorus corrosion inhibitor used in water treatment plant (zinc orthophosphate), were also screened with the IC-ICP method. A second general type of sample was generated by contacting solids (slag, digestor sludge, fly ash) from different environmental sources with DI water or mild acid/base, prior to filtration and analysis by IC-ICP. A variety of steel slag samples including two basic slag samples, one ladle slag, one desulfurization slag, and one basic oxygen process (BOP) slag were used. One gram of each slag was placed in 50 mL DI H2O in a separate glass bottle. A parallel sample was set up, but its pH was held at 3 using HCl. Samples were manually stirred once everyday. After about 1 week, liquid was withdrawn and filtered for IC-ICP analysis. Wastewater treatment plant anaerobic digester liquid sludge samples and anaerobic digester sludge cake from 11 wastewater treatment plants were sampled (Table 2). The extracted solutions were filtered and then frozen before analysis. Three kinds of solid wastes (bottom ash, fly ash, and lime waste) were collected from a coal power plant waste in Blacksburg, VA. Depending on solid density, 5 or 10 g solid was placed in 100 mL DI H2O. Samples were kept in the dark for 2 days, stirred, and then filtered. The final general type of sample was that of gas produced during wastewater or sludge incubation. Samples of wastewater influent, predigestion (wasted activated sludge, primary sludge, thickened waste activated sludge, thickened primary sludge), and postdigestion (digested liquid sludge, stored digested sludge, dewatered digested cake, piled dewatered cake) were incubated in closed gastight 500 mL PET beverage bottles at 22 °C for up to 49 days. The bottle headspace gas was sampled and measured several times during the incubation period. On the day of measurement, 100 uL headspace gas was injected into GC/MS after cryofocussion.
Results and Discussion Process Water from a Phosphorus Plant. A high concentration (g/L level) of hypophosphite and phosphite were confirmed in process water samples from the thermal phosphorus plant (Table 3). The results tested using colorimetric standard methods were consistent with previously defined behavior of phosphorus species (20) in that after digestion total phosphorus was recovered but only orthophosphate was measured without digestion. In general, there was reasonable agreement between the colorimetric determination of orthophosphate and that of the IC-ICP-ES method. After peroxide treatment of the process water, virtually all (>99.99%) of the phosphorus was present as phosphate. Some process water treated with peroxide and lime was blended with a small volume of untreated water
(ratio ) 100:1) and then was treated with FeCl3. FeCl3 was not as effective as peroxide in removing hypophosphite and phosphate. Because of the very high salt content in these samples, a matrix correction based on standard spiking and recovery was applied to most of the results. The IC-ICP method quantified much more phosphorus than did the colorimetric analysis for one sample, and the recovery of spiked phosphorus to this particular sample was 90%, but glyphosphate (70%), phosphite (35%), and MPA (45%) removals were much lower. This result was consistent with Holford (33) and Rickard’s finding (10) that phosphate fertilizers were rapidly removed by soils through adsorption and precipitation and rapidly become unavailable to plants. The phosphite fertilizer tested in this work was removed by soils at a much slower rate, consistent with Ruthbaum’s result (34). We noted that this tendency would make it possible for phosphite fertilizers to reach deeper plant roots in soil, but it also makes phosphite more susceptible to runoff problems. It was not clear whether the reduced phosphorus was removed by direct sorption to the soil or whether it was first oxidized to phosphate. Glyphosate’s retention time (about 24 min) was much longer than the other phosphorus species in the standard, but even 10 ug/L glyphosate was still readily detectable by the method. Wastewater Plant Samples. No detectable reduced phosphorus was observed in water samples from one of the two wastewater plants, but 1-2% of the total P in the activated sludge reactor effluent was detected as hypophosphite (and confirmed by the permanganate oxidation test) at the other plant. This plant used biological activated sludge with addition of ferric chloride (FeCl3) to assist phosphorus removal. Since ferric chloride is often produced from pickling of steels and other metals in concentrated hydrochloric acid, it is possible that the hypophosphite was introduced by the FeCl3. In our effort to evaluate the presence of reduced phosphorus in wastewater digesters, more than 20 filtered sludge or sludge cake extract samples were tested and no detectable reduced phosphorus such as hypophosphite or phosphite was detected. Also, no phosphine above the detection limit of 0.1 mg/m3 was detected in the headspace of incubated samples from 11 wastewater treatment plants in North America. Coupled with the finding of Glindemann et al. (35) that digesters contained only up to 0.1 mg P/m3, it seems VOL. 39, NO. 12, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4373
FIGURE 3. Fate of different phosphorus compounds spiked to soil samples.
FIGURE 4. Reduced phosphorus production from cast iron filing (10-3 M NaCl solution). clear that phosphine in digester gas is typically below the 1 mg/m3 level. In summary, the Devai et al. (13) finding of 382 mg/m3 phosphine may be atypical. However, it may be that the Devai (13) method was subject to interference from other gases which are eliminated using the Glindemann et al. (35) protocol. Regardless, no evidence could be obtained that massive anaerobic biological phosphate reduction occurs in sewage treatment plants, since all soluble phosphorus detected in digestor samples was greater than 99.99% as phosphates. It is well-known that bacteria accumulate phosphate during normal growth in their biomass and that this is exploited as the phosphate removal mechanism in sewage treatment (36-38). Therefore, the hypothesis published in Nature that most phosphate removal in sewage plants occurs via reduction to phosphine seems to be without basis. Cast Iron Corrosion. All tests with cast iron were conducted without stirring. A significant amount (almost 300 ug/L) of phosphite (PO3-P) was produced from cast iron filings in 139 days in a 10-3 M NaCl solution at room temperature (Figure 4). This confirmed the expectation that cast iron could be a source of reduced phosphorus (1, 20). The release rate of phosphite was increasing as the test progressed. Interestingly, when the experiment was repeated at 6785 °C, the spiked hypophosphite disappeared much faster. 4374
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 12, 2005
In fact, after 24 h the removal was 80-100% (data not shown) except in NaOH solution where its removal was 20% after 24 h. In most heated samples, much more phosphite was produced early in the experiment when compared to the same samples without heating (Figure 5), and in the sample with 10-3 M NaOH, PO3- production peaked early in the experiment and then declined (Figure 6). In contrast to the room-temperature result, after about 1 week, there was a significant amount of PO4-P (100-400 ug/L) in all heated samples. Loss of the PO4 spike to the iron sample was 80% in the first day, but concentrations gradually increased to the same level as other samples after one week. The final test of cast iron corrosion in 10-3 M NaCl, 10-3 M HCl, and 10-3 M NaOH was conducted in an autoclave. After 1 h, almost all spiked phosphate disappeared from solution. Again, only a small amount of PO2-P was detected in all samples, except NaOH with spiked PO4 which had the highest PO2-P of all the experiments (Figure 7). It is interesting that PO2 was so much higher in the NaOH sample spiked with PO4. Additional experiments are necessary, but it is possible that a portion of the phosphate was reduced to hypophosphite. Steel Slag Samples. No phosphite or hypophosphite were detected in the two commercially available basic slag samples, but about 500-700 ug/L PO4-P was present in the basic slag samples acidified to pH 3. There was no phosphate in the slags exposed to pure water.
FIGURE 5. Phosphite production in 10-3 M HCl solution under different temperature.
FIGURE 6. Phosphite productions in 10-3 M NaOH solution under different temperature.
A significant amount of PO2-P (about 75 ug/L) was detected in desulfurization slag, as confirmed by disappearance of the peak after spiking in KMnO4, with production of 65 ug/L PO4-P. This is a very strong evidence that leachable hypophosphite was present in this sample. Only about 6% of the slag dissolved as determined by gravimetric testing indicating about 0.01% of the dissolved material mass was hypophosphite. A PO2 peak also present in three other slag samples (BOP, ladle, and desulfurization slag) acidified to pH 3 at an apparent concentration range of 60-400 ug/L P. Dosing KMnO4 caused disappearance of this peak, but an increase of either PO3 or PO4 peak was not observed. It is possible that in acidic solution, other cations such as Fe2+ or Ca2+ precipitated the PO3 or PO4 produced from hypophosphite oxidation. Within a week, the PO2 peak disappeared. Coal Power Plant Samples and Phosphate Fertilizer. In samples collected from the coal power plant, a small peak was observed as PO2-P in only one sample. However, further
FIGURE 7. Phosphate and reduced phosphorus from cast iron corrosion (in autoclave). VOL. 39, NO. 12, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4375
testing indicated that it did not disappear during addition of permanganate. This is possibly a false signal due to very high levels of Ca2+. No detectable reduced phosphorus was found in samples of phosphate fertilizer. Zinc Orthophosphate Corrosion Inhibitor. About 1% (50 mg/L) hypophosphite was detected in the zinc orthophosphate corrosion inhibitor. This seems to confirm the hypothesis of Sugishima that traces of phosphite are present in reagent-grade phosphoric acid (39). This is important because the trace contaminant reduced the operational effectiveness of phosphoric acid fuel cells (39).
(22)
(23)
(24)
Supporting Information Available This work was supported by the National Science Foundation (NSF) under grant BES-0201849. The opinions, findings, conclusions, or recommendations are those of the authors and do not necessarily reflect the views of NSF.
(25)
(26)
Literature Cited (1) Morton, S. C.; Edwards, M. A. Reduced phosphorus in the environment. Crit. Rev. Environ. Sci. Technol., in press. (2) Mosher, E. The chemical control of phosphine gas generation during the machining of nodular cast iron. J. Soc. Tribologists Lubrication Eng. 1988, 45 (7), 445-450. (3) World Health Organization. Phosphine and Selected Metal Phosphides; Environmental Health Criteria 73; Geneva, Switzerland, 1988. (4) Gassmann, G.; Glindemann, D. Phosphine (PH3) in the biosphere. Angew. Chem., Int. Ed. Engl. 1993, 32, 2 (5), 761-763. (5) Glindemann, D.; Eismann, F.; Bergmann, A.; Kuschk, P.; Stottmeister, U. Phosphine by bio-corrosion of phosphide-rich iron. Environ. Sci., Pollut. Res. 1998, 5 (2), 71-74. (6) Guest, D.; Grant, B. R. The complex action of phosphonates as antifungal agents. Biol. Rev. 1991, 66, 159-187. (7) McDonald, A. E.; Grant, B. R.; Plaxton, W. C. Phosphite (phosphorous acid): its relevance in the environment and agriculture and influence on plant phosphate starvation response. J. Plant Nutr. 2001, 24 (10), 1505-1519. (8) Lee, A. K. New treatment approved for sudden oak death. The California Aggie October 8, 2003. http://www. californiaaggie.com/article/?id=305. (9) Lovatt, C. J. Formulation of phosphorus fertilizer for plants. U.S. Patent 5,514,200, 1990. (10) Rickard, D. A. Review of phosphorus acid and its salts as fertilizer materials. J. Plant Nutr. 2000, 23 (2), 161-180. (11) Callahan, M. Promotion of fertilizer in oak disease war called “illegal”. Press Democrat, April 29, 2001. http://www. greenbrae.org/news/042901-pressdemocrat.html. (12) Latimer, W. M. The Oxidation States of the Elements and Their Potentials in Aqueous Solutions, 2nd ed.; Prentice Hall: New York, 1952. (13) Devai, I.; Felfoldy, L.; Wieener, H.; Plosz, S. Detection of phosphine: newaspects of the phosphorus cycle in the hydrosphere. Nature 1988, 333, 343-345. (14) Devai, I.; Delaune, R. D. Evidence for phosphine production and emission from Louisiana and Florida marsh soils. Org. Geochem. 1995, 23, 277-279. (15) Freedman, L. D.; Doak, G. O. The preparation and properties of phosphonic acids. Chem. Rev. 1956, 57, 479-523. (16) Corbridge, D. E. C. Phosphorus: An Outline of its Chemistry, Biochemistry and Technology, 3rd ed.; Elsevier: Amsterdam, 1985. (17) PJB Publication Ltd. Agrow World Crop Protection News, No. 230, April 14, 1995, p 1. (18) Munro, N. B.; Talmage, S. S.; Griffin, G. D.; Waters, L. C.; Watson, A. P.; King, J. F.; Hauschild, V. The sources, fate, and toxicity of chemical warfare agent degradation products. Environ. Health Perspect. 1999, 107 (12), 933-974. (19) APHA; AWWA; WPCF. Standard Methods for the Examination of Water and Wastewater, 20th ed.; 1998. (20) Morton, S. C.; Glindemann, D.; Edwards M. A. Phosphates, phosphites, and phosphides in environmental samples. Environ. Sci. Technol. 2003, 37 (6), 1169-1174. (21) Wigfield, Y. Y.; Lanouette, M. Simplified liquid chromatographic determination of glyphosate and metabolite residues in envi-
4376
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 12, 2005
(27)
(28)
(29)
(30)
(31)
(32)
(33) (34)
(35)
(36) (37)
(38)
(39)
(40)
(41) (42)
ronmental water using post-column fluorogenic labelling. Anal. Chim. Acta 1990, 233, 311-314. Ryder, D. S. Separation of orthophosphoric, phosphorous and hypophosphorous acids using single-column ion chromatography with conductivity detection. J. Chromatogr. 1986, 354, 438-441. Ben-Yoseph O.; Sparkes, M. J.; Dixon, H. B. F. A rapid spectrophotometric estimation of individual phosphates and phosphonates: its application to chromatography of sugar phosphates and their phosphonate analogues. Anal. Biochem. 1993, 32, 195-198. Hirai, Y.; Yoza, N.; Ohashi, S. Flow injection system as a postcolumn reaction detector for high-performance liquid chromatography of phosphinate, phosphonate and orthophosphate. J. Chromatogr. 1981, 206, 501-509. Addison, R. F.; Ackman, R. G. Direct determination of elemental phosphorus by gas-liquid chromatography. J. Chromatogr. 1970, 47, 421-426. Smillie, R. H.; Grant, B. Determination of phosphate and phosphite in plant material by gas chromatography-mass spectrometer and ion chromatography. J. Chromatogr. 1988, 455, 253-261. Roos, G. H. P.; Loane, C.; Dell, B. Facile high performance ion chromatographic analysis of phosphite and phosphate in plant samples. Commun. Soil. Sci. Plant Anal. 1999, 301 (17, 18), 23232329. Wojciechowski, K. T.; Malecki, A.; Prochowska-Klisch, B. The influence of pH on the kinetics of oxidation of hypophosphite by permanganate ion. Bull. Acad. Pol. Sci. Ser. Chem. 1995, 43 (2), 125-133. Wojciechowski, K. T.; Malecki, A. Mechanism and kinetics of hypophosphite oxidation by permanganate ion. Int. J. Chem. Kinet. 1999, 31 (10), 737-743. Za´hon×c6i-BudO Ä , EÄ .; Sima´ndi, L. I. Oxidation of hypophosphorous acid by permanganate ion. Inorg. Chim. Acta 1991, 181, 149-151. Idriss, K. A.; Hammam, A. M.; Seleim, M. M.; El-Komi, Z. R. Oxidation with permanganate in presence of fluoride: determination of hypophosphite. Talanta 1980, 27, 561-566. Morton, S. C. Phosphorus in the environmental and its role in anaerobic iron corrosion. Ph.D. dissertation, Virginia Polytechnic Institute and State University, Blacksburg, VA, 2003. Holford, I. C. R. Soil phosphorus: its measurement, and its uptake by plants. Aust. J. Soil Res. 1997, 35, 227-239. Ruthbaum, H. P.; Baille, W. J. H. The use of red phosphorus as a fertilizer. part 4. phosphite and phosphate retention in soil. New Zealand J. Sci. 1964, 7, 446-451. Glindemann, D.; Stottmeister, U.; Bergmann A. Free phosphine from the anaerobic biosphere. Environ. Sci., Pollut. Res. 1996, 3 (1), 17-19. Barnard, J. L. A review of biological phosphorus removal in the activated sludge process. Water SA 1976, 2, 136-144. Levin, G. V.; Topol, G. J.; Tarnay, A. G. Operation of full scale biological phosphorus removal plant. J. Water Pollut. Control Fed. 1975, 47, 577-590. Comeau, Y.; Hall, K. J.; Hancock, R. E.; Oldham, W. K. Biochemical model for enhanced biological phosphorus removal. Water Res. 1986, 20 (12), 1511-1521. Sugishima, N.; Hinatsu, J. T.; Foulkes, F. R. Phosphorus acid impurities in phophoric acid fuel cell electrolytes. J. Electrochem. Soc. 1994, 141 (12), 3325-3331. Mehra, M. C.; Pelletier C. Ion chromatographic determination of hypophosphite, phosphite and orthophosphate ions along with chloride, nitrate and sulfate ions using indirect ultraviolet detection. Anal. Sci. 1990, 6, 431-434. Graaf, R. M. de; Visscher, J.; Schwartz, A. W. A plausibly prebiotic synthesis of phosphonic acids. Nature 1995, 378 (30), 474-477. Graaf, R. M. de; Visscher, J.; Schwartz, A. W. Reactive phosphonic acids as prebiotic carriers of phosphorus. J. Mol. Evol. 1997, 44, 237-241.
Received for review November 10, 2004. Revised manuscript received March 14, 2005. Accepted March 23, 2005. ES0401038
