July 31, 2005

Reduced Phosphorus Compounds in the Environment

Reduced phosphorus compounds (phosphorus with oxidation number less than +5) are and often ignored component of the global phosphorus cycle. This paper highlights environments in which reduced phosphorus compounds may be significant, reviews their importance in industrial applications and products, and examines their possible production by anaerobic bacteria or the steel industry. The role of reduced phosphorus in corrosion of iron and steel is worthy of additional research, since reduced phosphorus compounds can catalyze iron corrosion, and corrosion can liberate significant quantities of reduced phosphorus to the environment. The recent discovery that volcanic emissions can contain phosphorus is also deserving of additional study, given its potential impact on the global phosphorus cycle and the likelihood that reduced phosphorus compounds may be present.

KEYWORDS: iron corrosion, nutrients, phosphine, phosphite, phosphorus

Fundamental understanding of phosphorus behavior is a key to solving environmental problems of plant, animal and microbe nutrition, eutrophication, corrosion, and geochemistry. Researchers of such problems nearly always assume that phosphorus in natural systems occurs exclusively in the +5 oxidation state as orthophosphate, polyphosphates, organophosphates, and paniculate phosphates. This assumption has never been explicitly proven.

Phosphorus is known to occur in at least 7 oxidation states, including the phosphides (-3), diphosphide (-2), tetraphosphide (- 0.5), elemental phosphorus (0), hypophosphite (+1), phosphite (+3), and phosphate (+5). We term any phosphorus species with an oxidation state lower than (+5) "reduced phosphorus." Extraterrestrial phosphorus in iron meteorites is commonly in the form of schreibersite (Fe, Ni)3P or perryite (Ni, Fe)5(Si1P)2, both of which are phosphides (WHO, 1988; Friel and Goldstein, 1976; Pasek et al., 2004). It has long been understood that Jupiter's upper atmosphere contains hydrogen phosphide (Larson et al, 1977), that reduced phosphorus compounds might color the Great Red Spot (Prinn and Lewis, 1975), and that reduced phosphorus species are likely dominant phosphorus species in the crust of Jupiter and similar giant planets (Lewis, 1969). In addition, the earth's volcanic gases can be rich in phosphorus (Obenholzner et al., 2003) and contain traces of methyl phosphine (CH5P) (Wahrenberger, 1997), suggesting that reduced phosphorus compounds may be present in the earth's crust. This is not surprising given the thermodynamic arguments of Lewis (1969).

Gulick (1955) proposed that phosphite or hypophosphite serves as a major source of phosphorus for the initiation of life on earth, and Schwarz (1972) speculated that the high temperature in the earth's core and lightning discharges could convert phosphate to reduced forms by reaction with carbon. Later, he successfully produced phosphite salts from phosphate with simulated lightning in the lab (Glindemann et al., 1999). Furthermore, screibersite produces P^sub 2^O^sub 7^ when it reacts with water, and this molecule is building block of ATP (Stiles, 2004).

Even if these relatively exotic reactions are ignored, a growing body of research has indicated that the modern terrestrial phosphorus cycle on earth is not limited to phosphates. As detailed in the next section, reduced phosphorus compounds are seeing increasing industrial use, which in turn will directly impact the speciation of phosphorus released to the environment. The goal of this review is to summarize the chemistry and biology of reduced phosphorus species and to discuss practical implications of phosphorus chemistry from the perspective of steel and corrosion.



Phosphorus is the 11th most common element on earth. It exists mostly in phosphate rocks (primarily as calcium phosphate), and the earth's crust contains an average of about 0.1% phosphorus (Wazer, 1961). Over time and because of weathering of rock surfaces, phosphates became available to organisms (Duley, 2001), with a very slow turnover rate of more than 1 Gy (109 years) as restricted by diagenesis (Pierrou, 1979). As a result, phosphorus is frequently a limiting nutrient in freshwater, coastal, and estuarine systems. On the other hand, eutrophication is often caused by excessive inputs of phosphate via domestic, industrial, and diffuse agricultural sources, and it is well known that control of tropic status requires limitation of phosphate inputs.

Phosphorus is unique among the major nutrients (carbon, oxygen, nitrogen, phosphorus) in that it is often assumed to lack a gaseous species for atmospheric transport. However, the recent work of Glindeman et al. (1996a, 1996b, 1996c, 1999) and Han et al. (2000, 2003) unambiguously confirmed that volatile phosphine gas (i.e., hydrogen phosphide or PH3) can be detected in the earth's atmosphere at trace levels. Several sources of phosphine have been also identified by these authors. Unfortunately, perhaps because of the erroneous assumption that phosphorus is nonvolatile, study of total transport through this mechanism has received minimal attention, and all atmospheric phosphorus transport is assumed to be via phosphate dust.

Atmospheric transport of phosphorus is significant. Pierrou (1979) estimated that atmospheric fallout of phosphorus is in the range of 3.6-9.2 Tg (1 Tg = 10^sup 12^ g) P/yr for terrestrial ecosystems (6.3-12.8 Tg P/yr for the earth), and Graham and Duce (1979) attempted to quantify P flux from land to the atmosphere and estimated to be 4.3 Tg P/yr. If this phosphorus was uniformly dissolved in the world average annual rainfall of about 400 χ 103 km^sup 3^, it would suggest an average phosphate concentration of about 10 ppb in rainwater. The potential importance of atmospheric phosphorus loading to oligotrophic lakes was seemingly confirmed by Lewis et al. (1985), who quantified soluble phosphate in rainwater to remote mountain lakes and determined it accounted for 25% of the total annual phosphate flux to the watershed. Interestingly, this phosphate was not associated with dust or pollen. Atmosphere input of soluble P to the coastal ocean was estimated to be 12 10^sup 10^ mol P yr^sup -1^ (Benitez-Nelson, 2000). This is about 10% of dissolved inorganic phosphate from rivers (Duce, 1986; Delaney, 1998). The issue of atmospheric phosphate is discussed in later sections in light of the results of Glindeman et al. (2003).

The phosphate industry directly and indirectly produces many reduced phosphorus compounds. The United States is the largest producer and consumer of phosphate rock in the world. In 1997, the marketable production of phosphate rock in the United States was 32% of the world total production (United States, 45.9 million metric tons; world total, 143 million metric tons). In 2001, the marketable production and sale of phosphate rock decreased worldwide due to decreased demand for fertilizer (USGS, 2001), and the U.S. share of production dropped to 25%.

A report complied by Centre Europeen d'Etudes des Polyphosphate (CEFIC, 1997) stated that fertilizers account for 80% of overall phosphorus chemical production, with the balance used in detergents (12%), animal feeds (5%), and special applications (3%) (CEFIC, 1997). Whether this statistical data refers to Europe or worldwide could not be determined. Since the first commercial modern phosphate industry began to produce fertilizer in the mid-19th century, demand increased continuously, especially after the 1950s, and reached a range of 25-35 million metric tons after the 1970s (Figure 1) (IFIA, 2002).

FIGURE 1. Phosphorus fertilizer consumption in the world and the United States.

The end product of the thermal route is elemental phosphorus (reduced phosphorus), whereas phosphates are produced via the wet route. Wet acid processes were first used in Europe more than 20 years ago, but the products were of low purity and were limited to agriculture usage (fertilizer and animal feed supplements) until the early 1990s, when the United States developed a technology called the purified wet acid process (Reisner, 1991). The cost of electricity is a competitive disadvantage to the thermal route, and as a result it has gradually declined to the point where only 10% of total U.S. phosphorus rock is used in this process (Reisner, 1991; Johnson, 1997; USGS, 2002). In fact, one of the last two thermal elemental phosphorus plants in the United States was closed at the end of 2001. Production of elemental phosphorus has also dropped worldwide over the past 10 years (USGS, 2002). However, at least some thermal plants are needed, since elemental phosphorus meets the unique requirements of many higher value phosphorus derivatives.

Elemental Phosphorus and Its Derivatives

In 1669, German physician and alchemist Hennig Brand first isolated elemental phosphorus from urine. There are two main allotropes of phosphorus: white and red. White phosphorus is poisonous and highly flammable. It is used to produce many different phosphorus compounds, including red phosphorus, by heating to 250C (Jefferson Lab, 2002). There is also a black and brown form of phosphorus although they are not as common. Black phosphorus is made \by heating white phosphorus in the presence of mercury catalyst and a seed crystal of black phosphorus (Jefferson Lab, 2002), whereas brown phosphorus is made by condensing phosphorus vapor at liquid nitrogen temperatures. Brown phosphorus reverts back to a red and white phosphorus mixture at above 0C (Corbridge, 1985). About 264,000 metric tons of elemental ?4 capacity was available in Northern America as of the year 2000, 85% of which is burned to P^sub 2^O^sub 5^, hydrated to phosphoric acid, and then converted to other phosphate compounds (Brummer et al., 2000) (Figure 2).

The remaining 15% of elemental P^sub 4^ production is used as direct reactant for specialty products (Brummer et al., 2000), which can include P^sub 2^S^sub 5^, PCl^sub 3^, POCl^sub 3^, P^sub 2^O^sub 5^, and hypophosphite, with much smaller amounts converted to PH^sub 3^, red P, phosphonate, and other phosphorus derivatives (Brummer et al., 2000). They are then released to consumers through various products including fertilizers (phosphite fertilizers), fungicide, insecticides, herbicides, rodenticides, fumigants, flame retardants, chemical intermediates, and other industrial products (Table 1).

FIGURE 2. Phosphate compounds derived from P^sub 4^ at thermal processing plant; 85% of the total elemental P is converted to phosphates.

It is recognized that contamination from white phosphorus munitions is an environmental concern. Military bases, including Long Harbour, Newfoundland, Canada, MuscleShoals, AL, Pine Bluff Arsenal, AR, Ft. McCoy, WI, Eagle River Flats, AL, and Aberdeen Proving Ground (Chesapeake Bay), MD, have documented white phosphorus contamination problems (Idler, 1969; Jangaard, 1972; Dacre and Rosenblatt, 1974; Blumbergs et al, 1973; Simmers et al., 1995; Walsh et al., 1999; Buchanan et al., 1989), and cases of white phosphorus contamination can be expected worldwide. White phosphorus is persistent in saturated sediments (Walsh and Collins, 1996) and has caused death of waterfowl, including ducks, geese, and swans (Sparling and Federoff, 1997; Sparling et al, 1999; Roebuck et al, 1998). A dose of 1-12 mg/kg to duck gizzards caused kidney and blood problems (Coburn et al., 1950), whereas trout, salmon, cod, and herring are killed by waterborne white phosphorus concentrations of 0.5-2.5 μg/L (Zitko et al., 1970; Fletcher et al., 1970, 1972). Many measures such as barriers have been installed to minimize the danger from white phosphorus contamination in the environment (Pochop et al., 2000; Walsh et al., 1999, 2000).

TABLE 1. Some Major Phosphorus Compounds and Their Applications

Phosphine and Phosphides

Phosphine (i.e., hydrogen phosphide or PH^sub 3^) has been known to science since the birth of modern chemistry, and its discovery was credited to Gengembre in 1783 (von Meyer, 1891). Phosphine gas has been identified in the environment in a variety of locations (Devai et al., 1988; Devai and Delaune, 1995; Glindemann et al, 1996a, 1996b, 2003; Burford and Bremner, 1972; Eismann et al, 1997a, 1997b, 1997c; Cao et al., 2000; Han et al., 2000) (Table 2). Hydrogen phosphide or phosphine (PH^sub 3^) is approximately 2 times more toxic to humans than hydrogen cyanide (Latimer, 1952) and has a shortterm exposure limit of 1 mg/m^sup 3^ (WHO, 1988). A concentration of 9 mg/m^sup 3^ can be tolerated for several hours without symptoms (Meaklim, 1998), and the National Health and Medical Research Council (NHMRC, 1992) in Australian recommended an environmental "action" level of 4 mg/m^sup 3^ (NHMRC, 1992).

Several accidents in the United States have been attributed to phosphine in fumigation or phosphine released from cargo shipping (WHO, 1988). One source cites several thousand cases of phosphide poisoning in India (Gard, 1999). It has been speculated that phosphine may be involved in sudden infant death syndrome, and spontaneous ignition of phosphine has been suspected to be responsible for ghostly green lights over marshes termed the "will- o-the-wisp" phenomena (Wilson et al, 1980; WHO, 1988; Richardson, 1990; Atlas et al., 1993). The phosphides are currently known to be introduced to the environment from degradation of corroding metals such as iron, from the anaerobic biosphere, from the combustion of coal (e.g. brown coal) in power plants, from burning of landfill gas and biogas, and through their use as grain fumigants and rodenticides (Mosher, 1988; WHO, 1988; Glindemann et al., 1996a, 1996b, 1996c, 1996d, 1998, 2003).

In Australia, as many as 20 chemicals were once used as fumigants for grain storage, but Banks (1994) noted that only methyl bromide and phosphine were commonly used. Phosphine is actually on its 'way to becoming the only allowable grain fumigant, since methyl bromide is an ozone-depleting compound and may be banned. Fumigation with phosphine is also of concern due to its toxicity to humans (Banks, 1994). A new approach advocating low-concentration phosphine fumigation with a mixture of 50-500 ppm phosphine and 4-10% carbon dioxide was invented that slightly lessens the adverse potential impacts of phosphine (Mueller, 1995). Also, to ensure the safety and effectiveness of fumigation, phosphine fumigation detector cards (phoscards) have been distributed to farmers (Emery et al., 2000). Phosphine used in fumigation can be obtained in several forms, including aluminum phosphide-based pellets or tablets, phosphine gas cylinders, or using apparatus to generate phosphine on-site (Waterford, 1998). Aluminum phosphide solids are the most common source of phosphine for fumigation. According to Degesch America, which is the only metal phosphide fumigant producer in North America, the annual worldwide demand for metal phosphide-based fumigants was about 9800 metric tons in 2002 (Degesch America, Inc., personal communication, 2002).

TABLE 2. Phosphine and Other Phosphide in the Environment

Phosphine is also a common dopant in the electronic industry. In 1979, 6 million L phosphine gas in various concentrations (equivalent to about 300,000 L pure phosphine) was used by a total 42 electronics companies (LaDou, 1983). A 30% annual growth of this usage by the electronic industry was once projected, although data on current use could not be found (SRI, 1982).

In the atmosphere, phosphine reacts very rapidly with hydroxyl radical. Frank and Rippen (1987) calculated a half-life for atmospheric phosphine of 28 h, and the final product of the reaction is phosphate, which returns to earth in rainfall. Thus, phosphine emissions would contribute to atmospheric transport of phosphorus, an idea not discussed in Pierrou (1979) or other works that consider global phosphate cycling, but which is unambiguously proven by Glindeman et al. (1996a, 1996b, 1999, 2000, 2003).

The most common phosphine analysis is by gas chromatography (GC) either with a flame photometric detector (FPD) (Hilton and Robinson, 1972; Hashimoto et al., 1985) or nitrogen-phosphorus detector (NPD) (Glindemann et al., 1995, 1996a, 1998; Morton et al., 2003). Gas chromatography-mass spectrometry (GC-MS) is another method used for phosphine analysis (Glindemann et al., 2003; Morton et al., 2004). GC-FPD and GC-MS are both 1000 times less sensitive than GC-NPD. The most sensitive phosphine analysis method is GC-NPD with cryofocussion (Glindemann et al., 1996a), which is capable of detecting 0.1 ppt (v/v) for a 50 ml sample and 0.01 ppt for 500 ml.

It is worthwhile to consider potential sources of phosphine relative to total estimated atmospheric phosphorus transport. For instance, recently phosphine has been measured with gas chromatography-mass spectroscopy (GCMS) in a few locations in the troposphere at about 1 ng/m^sup 3^ (Glindeman et al., 2003). Considering 2 10^sup 18^ m^sup 3^ as the rough volume of the troposphere, atmospheric phosphine would be 2 10^sup 9^ g. Compared to 4.3 Tg (4.3 10^sup 12^ g) P/yr of atmospheric phosphorus transport (Graham and Duce, 1979), phosphine in the troposphere at any one time is about 0.05% of the total global atmospheric P flux. If about half of the atmospheric phosphine pool was converted to phosphate daily (Frank and Rippen, 1987) and the lost phosphine was then replaced from new sources to maintain steady state, about 10% of the global atmospheric phosphorus flux would be linked to phosphine.

The origins of relatively high concentrations of phosphine measured in the troposphere ([approximate]1 ng/m^sup 3^) have not yet been clearly denned (Glindeman et al., 2003). We note that iron meteorites contain an average of about 1 wt% of phosphorus, and it is estimated that up to 7 10^sup 7^ mol of reduced phosphorus/pyr could have come to the primitive earth as schreibersite. Current loading from this source is estimated at 1 10^sup 7^ mole P/pyr (Schwartz, 1972). If all of this was somehow released as phosphine, it would account for

Mueller (1993) did a calculation on worldwide phosphine production from fumigation based on annual metal phosphide production and usage of 4000-4600 tons/yr, which is about one-half of the current estimate for phosphide demand that was obtained for this work (Degesch America, Inc., personal communication, 2002). Assuming complete release of this phosphine to the atmosphere, this would account for only about 0.2% (based on 4.3 Tg P/yr) (Graham et al., 1979) of annual atmospheric phosphorus loading. However, given that phosphine fumigant use is highly concentrated geographically and temporally, the use of fumigant is likely to be a significant local contributor to atmospheric phosphorus in certain situations (Frank and Rippen, 1987; Pratt, 1999).

It was also determined that phosphine averaged 9 ng/m^sup 3^ in emissions of a coal burning power plant (Glindemann et al., 2003). More plants would have to be assessed before worldwide significance of this source could be determined, si\nce it is commonly known that phosphorus content of coal varies dramatically with the source (0.001-0.25% P by weight). However, none of the sources just described seems capable of explaining the relatively high concentrations observed in the troposphere by Glindemann et al. (2003). The recent observation that volcanic emissions can be rich in phosphorus is therefore worthy of additional study (e.g., Obenholzner et al., 2003), since phosphine and phosphorus content of gases is not routinely monitored nor is oxidation state of phosphorus routinely determined in ash or lava.


TABLE 3. Key Thermodynamic Constants of Some Relevant Redox Processes in Natural Waters and for Reduced Phosphorus Species (Pourbaix et al., 1962; Woods et al., 1987; Stumm and Morgan, 1996)

Because of their illegal use in the manufacture of methamphetamine (an illegal drug), the manufacture and sale of hypophosphite salts (e.g. sodium hypophosphite) and elementary phosphorus (red P and white P) have been under control since 2000 (Drug Enforcement Administration [DEA], 2000). Phosphites are widely marketed either as a fungicide or as a potentially superior substitute source of plant phosphorus nutrient (Guest and Grant, 1991; McDonald et al., 2001). Although it is commonly accepted that phosphites are excellent fungicides, and their use recently received a boost as the only approved method of treating sudden oak death (Lee, 2003), the claim regarding their potential as a fertilizer is more controversial.

Studies of phosphite use as a fertilizer started after World War II (1950s). Maclntire et al. (1950) reported a definite nutritional response to phosphite compared to a control without phosphorus. Although phosphorous acid (H^sub 3^PO^sub 3^) and calcium phosphite were toxic to plants in the first year, in later years they were beneficial. This is likely because the phosphite was oxidized to phosphate by microorganisms in soil and plant tissues (Adams et al., 1953; Casida, 1960; Malacinski and Konetzka, 1966; Bezuidenhout et al., 1987).

For years thereafter, most of the focus was on phosphites use as a fungicide. This situation changed in the 1990s, when Lovatt (1990) promoted the foliar application of phosphite as more efficient than phosphate foliar application. Phosphite formulations were presented as cost-effective replacements for traditional phosphate soil applications. Phosphite fertilizers were promoted as beneficial to the environment for the lower amounts of phosphorus used, which would result in reduced eutrophication of freshwater ponds, lakes and streams. However, the side benefits of antiviral, antibacterial, and antifungal, activity are also mentioned (Lovatt, 1990).

The idea that foliar phosphate might be more efficient seems reasonable. Based on Holford's (1997) result, Rickard (2000) pointed out that a significant fraction (>80%) of phosphorus applied as phosphate is adsorbed, precipitated, or converted to other organic forms by soil. Most of this cannot be utilized by the plant and is lost. Absorption of phosphite by soil is less than phosphate (Ruthbaum et al., 1964), and therefore losses by this mechanism would be expected to be less as well. In field trials, universities, research organizations, and individuals have proven that plants treated by phosphite fertilizer grew statistically better than plants treated without fertilizer. In quite a few experiments, crops also gave a better response to foliar applied phosphite fertilizer than to foliar applied phosphate fertilizer (Rickard, 2000). There are currently many phosphite fertilizer products in the market, and most are applied using a foliar formula (J. L. Peterson, Biago Western, personal communication, 2003). However, the price (about $22/kg P^sub 2^O^sub 5^) is much higher than traditional soil phosphate fertilizer (44-77 US cents/kg of P^sub 2^O^sub 5^). (J. L. Peterson, Biagro Western, personal communication, 2003; Manitoba Agriculture and Food, 1999).

On the other hand, the potential downside of phosphite fertilizer has been noted. For instance, it has been argued that the oxidation process of phosphite to phosphate by microorganisms is very slow and may take months or even a year, giving concern about residual phosphite. Phosphites are also known to disrupt the acclimation of plants to phosphorus deficiency since they induce a characteristic plant phosphorus starvation response (McDonald et al., 2001). Finally, it is also noted that the benefits of phosphites versus phosphates might be due to its fungicidal action. The labeling of phosphites as fertilizers circumvents the requirement that they be registered as fungicides through the U.S. EPA Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) registration, a lengthy and expensive process (Callahan, 2001). Thus, there is an incentive to promote their use as fertilizer, even if its action is fungicidal.

In the MacIntire et al. (1950) review paper, it was concluded that a small percentage (

Phosphite-containing industrial waste is becoming a problem in certain high-tech industries. Hypophosphite (PO^sup 3-^^sub 2^) has been used in electrodeless plating processes such as those used in compact-disk manufacture (Ohtake, 1995). The final product of metal plating is wastewater containing a high concentration of phosphite (HPO^sup 2-^^sub 3^) and organic acids. It is commonplace to oxidize phosphite to phosphate before further disposal with, for example, well-developed physiochemical or biological nutrient removal processes. However, because of the high concentration of organic acid, oxidation is difficult (Ohtake et al., 1996).

Organic Reduced Phosphorus

Phosphonates are organophosphorus compounds with a very stable C- P bond. Since the discovery of 2-aminoethylphosphonic acid (AEP) in 1959 (Horiguchi and Kandatsu, 1959), over the next 20 yr phosphonates were identified in over 80 animal species including humans. Phosphonates are naturally synthesized by various organisms (Hilderbrand, 1983). About 3% of total phosphorus in natural plankton is present as phosphonates (Kittredge et al., 1969). It has been proven that mammals do not synthesize phosphonates by themselves, but these compounds are obtained through ingestion in the case of humans or absorption in the case of goats and cows (Hilderbrand, 1983). However, no systematic investigation has been conducted on the occurrence of phosphonates in living organisms.

Only procaryotic microorganisms (both gram-positive and gramnegative bacteria), some yeasts and fungi are able to degrade phosphonates by cleaving the C-P bond (Kononova and Nesmeyanova, 2001). A range of bacteria capable of degrading phosphonate is widely present in the environment (Schowanek and Verstraete, 1989). AEP is used by some bacteria as the sole source of carbon, nitrogen, and phosphorus (Cook et al., 1978). 2-Phosphonoacetaldehyde hydrolase or phosphonatase, phosphonoacctate hydrolase, and C-P lyase are major enzymes that can catalyze degradation of phosphonates (Wackett et al., 1987; Kononova and Nesmeyanova, 2001). Metcalf and Wanner (1991) noted that bacteria that can metabolize phosphonate could also metabolize phosphite, suggesting that a similar pathway was involved for both in biology.

There is some ambiguity on the oxidation state of phosphorus in phosphonates. Some researchers have reported that phosphonates are natural analogues of phosphates (Engel, 1977; Hilderbrand, 1983). Boenig et al. (1982) stated that hydrogen bound to phosphorus has very little acidic or hydric character, indicating that the average oxidation state of P in phosphonate is close to +4. Freedman and Doak (1956) regarded phosphonate as a derivative of phosphorous acid (+3). Corbridge (1985) specified the oxidation state of phosphonic acid or phosphite esters as +3. Schwartz (1997) stated that phosphonates such as phosphonic acids have an unsaturated bond, implying that the oxidation state of P is less than 5. Of course, ascribing formal oxidation states is not always possible, but it is worth noting that some parties define these organic phosphorus compounds as "reduced P."

As has been speculated for phosphite, it is believed by some that phosphonates might have been present on an evolving earth. Cooper et al. (1992) identified five of the eight possible alkyl phosphonic acids (I) in Murchison meteorite. Pasek et al. (2004) detected methylphosphonic acid (MPA) and ethylphosphonic acid (EPA) in Murchison meteorite. The methylphosphonic acid (MPA) and ethylphosphonic acid (EPA) concentration detected in the Murchison extract was significant (9 nmol/mg meteorite and 6 nmol/mg meteorite, respectively), de Graaf et al. (1995, 1997) proposed the possible synthesis of phosphonic acid (R-O^sup 2-^^sub 3^) by ultraviolet irradiation in meteorites or comets, which later provided organic phosphorus to the prebiotic earth.

The industrial synthesis of phosphonates started after the discovery of the Arbuzov reaction in 1905. Phosphonates are widely used as herbicides (e.g. glyphosate), insecticides, antibiotics, flame extinguishers, corrosion inhibitors, enzymes, chemical additives, and drugs (Kononova and Nesmeyanova, 2001). Some phosphonic acids, such as methylphosphonic acid and ethylmethylphosphonic acid, are degradation products of some chemical warfare agents (Munro et al., 1999).

Glyphosate \is perhaps the most prominent phosphonate. Its empirical formula is C^sub 3^H^sub 8^NO^sub 5^P (WHO, 1994). It has been used as herbicide since 1971. There are dozens of glyphosate- formulated products in the world herbicide market (Pesticide news, 1996). Annually, glyphosate products were sold for a total of US$ 1.2 billion, which accounts for 60% of global nonselective herbicide sales (PJB Publication Ltd., 1995). In the United States during 1991, 18.7 million pounds of glyphosate was applied on 13 to 20 million acres annually (U.S. EPA, 1993).


There is circumstantial evidence (Sawyer, 1973), but nothing yet considered conclusive or that has been reproduced, that certain anaerobic bacteria can reduce phosphate directly (Table 4). While the phosphine that has been detected in the environment is most commonly present at trace levels, under anaerobic conditions reduced phosphorus is currently believed to be a substantial component of the P cycle. Most prominently, although we do not agree with all of their interpretations, Devai et al. (1988) estimated that about 25- 50% of the total phosphorus removal (based on their measurements of phosphate in sewage influent and effluent) might be accounted for by phosphine gas emission. Stronger evidence for microbial phosphate reduction was gathered by Tsubota (1959), who determined that 100 mg/ L of hypophosphite-P and significant phosphite were produced in an anaerobic soil culture initially containing 2 g/L orthophosphate-P. No study to date has confirmed these important findings in relation to the environmental P cycle. Interestingly, environmental engineers who are responsible for phosphorus removal in sewage plants assume phosphorus is present only as phosphates, and the potential for reduced phosphorus compounds has never been noted in that extensive literature, nor is emission of phosphine noted as a potential removal mechanism in sewage treatment plants.

TABLE 4. Bacteria and Reduced Phosphorus in the Environment


Steel Making

Steel making has large impacts on phosphate and reduced phosphorus compounds at various stages of production (Figure 3). The process starts with iron ore, which can contain from 0.02 to 1% phosphorus by weight. The ore is heated up to 1300-1600C in a blast furnace to produce pig iron, and massive quantities of iron slag (12.3 million metric tons in the United States in 2003; USGS, 2003a) are produced to rid the ore of excess phosphate. World production of pig iron was 580 million metric tons in 2001 (USGS, 2003b). Pig iron end products can contain between

The pig ion is then used as a feedstock for steel making. In some cases, it is desirable to add ferrophosphorus to improve the mechanical properties of the final iron/steel (Crowson, 2001), reduce energy costs by lowering the casting temperature, enhance abrasion resistance, and improve corrosion resistance (Corbridge, 1978; Epstein, 1936; Stoughton, 1923). However, a high phosphorus content (e.g. >0.5 wt%) often leads to the detrimental formation of brittle iron phosphide (steadite) network (Makar and Rajani, 2000; Schipper et al., 2001). As a result, the phosphorus content of iron in drinking-water pipes may vary from 0.005 to 0.20% depending on intended use (Epstein, 1936; Achte, 1993).

The ferrophosphorus additive contains 50-60 wt% Fe and 18-28 wt% P (Fe^sub 2^P plus a small amount of Fe^sub 3^P and FeP) (Brummer et al., 2000). This additive is produced during thermal processing of phosphate rock. It was reported that in 2001 about 5070 metric tons of ferrophosphorus was added to steel in the United States as an additive (Fenton, 2001). Another waste stream termed "steel slag" was produced during manufacture of the final steel product, and the volume was 6.6 million metric tons in the United States during 2001 (USGS, 2003a). In 2002, >900 millions metric tons of steel was produced worldwide, about 13% of which (117 millions metric tons) was produced in North America (Steelnews, 2003).

FIGURE 3. Phosphorus in steel making.

Assuming that the representative average phosphorus content for all steel types is 0.04% (Shieldalloy Metallurgical Corporation, 2002), about 360,000 metric tons of phosphorus is produced annually within steel. Thus, phosphorus ending up in finished steel products may be about 5% of that mobilized for use in fertilizer, and it is therefore quantitatively significant. This percentage does not include phosphorus present in waste slag produced during steel making. No information could be found on the phosphorus content of iron or steel slag, though it has been reported to vary between 0.1 and 8% phosphorus by weight (e.g., http://soilsearth.massey.ac.nz/ cybsoil/article/slag.htm).

Friel conducted research proving that at high temperature (875C), phosphate reduction to metal phosphide can occurs within metal grains (Friel et al., 1976), thereby explaining the origins of metal phosphides such as schreibersite in meteorites. In addition, phosphorus ore is reduced to elemental phosphorus at temperatures above 1100C. Thus, the temperatures present in steel blast furnaces (1300-1600C) or other furnaces used to make steel (>l600C) are more than adequate to reduce phosphate, and as a result, the phosphorus in steel is often described as an Fe/Fe^sub 3^P (iron phosphide) eutectic.


Iron corrosion is one of the most important economic and aesthetic problems facing society. The AWWA and U.S. EPA estimate that public expenditures to replace degraded drinking water infrastructure will exceed $325 billion over the next 20 yr, which is more than 8 times the cost of all existing and proposed water treatment expenditures (Davies et al., 1998; AWWA, 1999), and studies in Australia, Great Britain, Japan, and the United States estimate that all metallic corrosion consumes about 3 to 4% of the gross national product for each country (Bennet et al., 1978; Uhlig and Revie, 1985; Iverson, 1983, 1987).

For more than 30 years Iverson has adamantly argued that production of a "volatile phosphorus compound" was a key event in catastrophic iron corrosion (Iverson, 1968, 1975, 1981, 1983, 1987, 1998). In the presence of this hypothesized compound, Iverson measured corrosion rates as high as 1250 μA/cm^sup 2^ in sea water even without oxygen present, which is a level capable of penetrating 0.6 inches deep into steel after just 1 yr. Though Iverson also identified iron phosphides (Fe^sub 3^P) in the scale (rust) layer that formed, his theory has never been accepted because it is thought that the micrograms per liter levels of phosphate typically present in natural aquatic systems are too low to support a significant effect from phosphorus, he never forwarded a mechanism by which catalysis might occur (Iverson, 1998), and the biologic pathway for volatile phosphorus compound formation is contrary to much of the conventional wisdom regarding difficult reduction of phosphates.

The redox potential (ε) for corrosion of iron, Fe(s) = Fe^sup 2+^ + 2e^sup -^, is not lower than that for reduction reactions of phosphorus species, which means iron itself cannot reduce phosphorus species under standard conditions. However, atomic hydrogen, which is produced on the surfaces of iron under anaerobic conditions, has extremely low ε, and is therefore capable of reducing phosphorus (Tables 2 and 3). In fact, in the presence of nascent hydrogen, phosphine could be manufactured by electrolysis of phosphorus (Boenig et al., 1982). It is therefore thermodynamically possible that reduced phosphorus is produced by reaction with atomic hydrogen on the surface of metals during corrosion at room temperature, although it is not possible to reduce phosphate in aqueous solution.

To better illustrate these concepts, a new Eh-pH diagram of the P- Fe-H^sub 2^O system was created (Figure 4) for this work using available constants for iron phosphides and reduced phosphorus species (Appendix A; Pourbaix et al., 1966; Kubaschewski et al., 1993; Woods and Garrels, 1987; Wagman et al., 1982; Beilstein Commander, 2000). Constants for ferrous, ferric, and phosphate species are those used in Geochemist's Workbench version 3.0.3 (Bethke, 1996; Delany et al., 1990). Compared to Pourbaix's classic thermodynamic diagram of the P system (1966), it is clear that in the presence of atomic hydrogen on the surface of iron metal, iron phosphide (Fe^sub 2^P) and phosphine CPHj) could be stable over a relatively broad Eh-pH range.

FIGURE 4. Potential-pH diagram for phosphorus-iron-water system, at 25C. [Fe^sup 2+^] = 10^sup -4^, [dissolved phosphorus species] P = 10^sup -5^; P^sub H2^ = 0.01 atm; P^sub o2^ = 0.001 atm. Area between two dashed lines is stability domain of H2O under these specific conditions.

Returning to the Iverson theory, a recent investigation by Glindeman et al. (1998) provides insights that overcome some limitations. First, many critics of the Iverson theory (and Iverson himself) never considered that phosphorus is present as a contaminant in all iron pipes as a result of iron manufacturing. While exploring rusting metal as a potential source of gaseous reduced phosphorus, Glindemann et al. noted that addition of 3.4 mg/ L hydrogen sulfide to water released up to 20% of the total phosphorus within iron chips as phosphine (Glindemann et al., 1998). Among the materials tested by Glindemann et al., sulfide seemed unique in its ability to induce phosphine release from iron. Thus, a volatile phosphorus compound is naturally produced from iron during corrosion, and biological reduction of phosphates is not necessary.

As to the issue of corrosion catalysis, a review reveals that transition metals such as iron form strong reactive complexes with phosphine and the related phosphite ligands (Collman and Hegedus, 1980). These metal complexes are a key starting poi\nt for a wide range of catalyzed redox reactions of industrial importance, including hydrognation of alkenes using Wilkensons catalyst and electroless plating of iron and nickel (Corbridge, 1985; Collman and Hegedus, 1980; Pignolet, 1983; Alyea et al., 1982; Durney, 1984). More specifically, research into the hydrogen degradation of ferrous alloys has established that certain dissolved gases including H^sub 2^S, AsH^sub 3^, and PH^sub 3^ can strongly sorb to metal surfaces and catalyze atomic hydrogen formation and subsequent detachment of gaseous H^sub 2^ from iron surfaces (Oriani et al., 1985). This process is termed "promotion" and the phosphine is a "promoter." While the main focus of the previous work has been on subsequent diffusion of this atomic hydrogen into the metal and resultant degradation of its structure, we note that removal of H^sub 2^ gas from iron surfaces is often 'the rate-limiting step in iron corrosion (Oriani et al., 1985; Iverson, 1968, 1975, 1981, 1983, 1987, 1998; Flis, 1991). Thus, phosphine and other reduced phosphorus compounds may be expected to catalyze anaerobic iron corrosion, as has been proven by Bala (1986).

Other observations in the hydrogen degradation literature are also important (Oriani et al., 1985). First, promoters are most effective at about room temperature and in a particular concentration range. Catalysis with arsine gas, a phosphine analog, is maximum when the soluble arsenic concentration is between 1 and 1000 ppb. It is also interesting that promoters are only effective as uncharged molecules, thereby explaining why hydrogen sulfide catalysis occurs only in acid solution where H^sub 2^S is dominant (Iyer and Pickering, 1989; Iyer et al, 1990), while phosphine and arsine are expected to be effective over the entire natural water pH range (Oriani et al., 1985). As a final and practical matter, very severe atmospheric corrosion of copper, iron, and even gold has been attributed to phosphine gas use during fumigation (Bond et al., 1984).

Thus, there is ample evidence supporting the idea that reduced phosphorus could catalyze iron corrosion under conditions present in some situations, such as water distribution system pipelines. There are also other interesting pathways that could lead to formation of the toxic PH^sub 3^ in drinking water systems. We note that in 2001, 56% of 380 water utilities responding to a survey added phosphoric acid corrosion inhibitors to their finished drinking water at a concentration between 1 and 3 mg/L as PO^sub 4^ (McNeill and Edwards, 2002). First, it remains possible that biofilm bacteria might reduce these phosphates directly to PH^sub 3^, or to other reduced P species, under some circumstances as noted earlier (Tables 2-4). second, Sugishma et al. (1994) believed that even reagent- grade phosphoric acid (H^sub 3^PO^sub 4^) contains significant concentrations of phosphite contaminants unless treated by H^sub 2^O^sub 2^ addition and 18 h of heating. We consider it likely that the food-grade phosphoric acid added to drinking water would contain these contaminants as well. In addition to possible direct addition of phosphite to drinking water and disproportionation, PH^sub 3^ could be directly formed from reaction with atomic hydrogen at metal surfaces (Tables 2 and 3, Figure 4), even though the redox potential of other reducing agents is not sufficient to cause direct reduction.

If the hypothesis of Sugishma et al. were proven true, it is possible that phosphorus inhibitors added to iron metal or to drinking water might actually worsen iron corrosion, with catastrophic consequences under at least some circumstances. Interestingly, there is some support for this in the literature. For instance, although it was not the focus of his study of atmospheric phosphine release, Glindemann et al. (1998) measured corrosion rates for iron known to contain very low P and iron containing very high P. Under identical conditions and in the presence of added H^sub 2^S, the low-P iron did not corrode significantly (0% weight loss) whereas the high-P iron corroded rapidly (6% weight loss). Thus, sulfides alone did not cause significant corrosion unless a significant source of reduced phosphorus species was present in the metal.

With respect to other documented adverse effects from orthophosphate "inhibitors," Weimer et al. (1988) studied anaerobic iron corrosion in the presence of sulfate-reducing bacteria (SRB) and differing orthophosphate levels in the water. In general, carbon steel corrosion rates increased at higher phosphate levels, with average corrosion rates of 25 rnpy at 1.6 mg/L PO^sup -3^^sub 4^ in the water versus

Similarly, in high-pressure boilers the term "phosphate corrosion" has been used to describe severe pitting of steel that accompanies hydrogen evolution under some circumstances (De Romero et al., 1999). The idea that reduced phosphorus, and not ortbo-f, is involved seems quite attractive. In these cases concentrations of aqueous ortho-P are observed to decrease markedly through a phenomenon popularly termed "phosphate hideout," only to inexplicably reappear later (Herro et al., 1995). It is currently believed that the missing phosphate deposits in scales and then redissolves. However, it is also possible the missing phosphate has simply been converted into a reduced phosphorus species, which is not detected by the standard analytical approaches, but which is still soluble and possibly catalyzes corrosion in an adverse way. Consistent with some of this hypothesis, recent research has demonstrated that phosphate inhibitors had a very significant adverse impact on iron corrosion under a wide range of circumstances if aerated water was held stagnant for 2-3 days before water changes in the pipe (McNeill and Edwards, 2000).

It is also worth considering the potential impact of phosphine gas release from corroding steel on the atmospheric component of the global P cycle. To our knowledge, no compilations exist on the mean weighted phosphorus content of iron ore, slag, or even steel, so only rough estimates can be attempted. Considering that 1000 million metric tons of iron ore is mined annually and with a weight of 0.1% phosphorus as a representative mean value, iron and steel making would mobilize 1 million metric tons of phosphorus worldwide or 0.06 million tons in the United States each year (USGS, 2002). If all of this phosphorus was reduced to phosphides and then eventually released as phosphine, this would constitute about 23% of the 4.3 million metric tons roughly estimated by Graham and Duce (1979) as P flux from land sources to the atmosphere worldwide. If it is assumed that an average content of finished steel is 0.04% worldwide, this would represent about 9% of the P flux from land to atmosphere. Of course, this phosphine would not be released all at once, but the release would occur slowly as the steel degrades, and it is not certain what percentage of the phosphorus in steel would be released as phosphine. However, it is very clear that steel represents a potentially large reservoir for release of true volatile phosphine to the atmosphere, especially given that older steel manufacturing processes led to products with much higher phosphorus contents than modern steel. In summary, we strongly support the link established by Glindeman et al. (1998) between rusting iron and atmospheric phosphine, and the calculations presented herein suggest that this source could be quantitatively significant in the global phosphorus budget.


There is a wide variety of circumstances in which reduced phosphorus compounds might be produced or encountered in environmental systems. Use of phosphides, phosphites, and phosphine is increasing in some industries. Steel making would appear to be the most significant potential source of reduced phosphorus compounds, either in waste products or in the finished product, and it is interesting that no detailed studies of phosphorus speciation in steel wastes, steel plant emissions, and slag waste products have been conducted. It is possible that phosphine emission from various products could contribute significantly to the global atmospheric P cycle, and that this loading could be concentrated locally. More study is needed to determine the relative contributions of dust, pollen and phosphine-derived phosphate in atmospheric loading. Likewise, studies should be conducted on the potential leaching of reduced phosphorus to drinking water from pipes, and the role of phosphorus as a potential catalyst of iron corrosion should be closely examined.

Overall, the review provides compelling evidence that reduced phosphorus species may be present in the environment at significant concentrations, and environmental engineers should explicitly test the common assumption that all phosphorus occurs exclusively as phosphate. To our knowledge, within the environmental engineering field, no one has ever even considered the possible existence of reduced phosphorus compounds, much less quantified them directly. In addition, a synthesis of disparate corrosion research findings leads to new hypotheses regarding the significance of reduced phosphorus in anaerobic iron corrosion. If such species were considered, it is possible that many inconsistencies and limitations of existing theories would be resolved. If these hypotheses are verified, it is quite possible that existing approaches to iron corrosion mitigation actually make corrosion worse under at least some circumstances, with profound economic, aesthetic, and (possibly) human health implications given the toxicity of reduced phosphorus \species.


Achte, vllig neu bearbeitete Auflage, Gmelin, Handbook of inorganic and orgamometallic Chemistry, 8th ed. eds. Koschel, D., and Schfer, H., SpringerVerlag, New York, 1993.

Adams, F., and Conrad, J.P. Transition of phosphite to phosphate in soils. Soil Sd. 75, 361, 1953.

Alyea, E.G., and Meek, D.W. Catalytic Aspects of Metal Phosphine Complexes. Advances in Chemistry Series 196. ACS, Washington, DC, 1982.

APTC (Australian Postharvest Technical Conference). Adelaide, August 2000.

Atlas, R.M., and Bartha, R. Microbial Ecology: Fundamentals and Applications, 3rd ed., Benjamin Cummings. Redwood City, CA, 1993.

AWWA. $325 Billion for pipes, AWWA Mainstream, 3, 1999.

Bala, H. Effect of phosphine on acid corrosion of iron, Werkstoffe Korros, 37, 644, 1986.

Banks, HJ. Fumigation-an endangered technology? In Proc. 6th Int. Working Conf. on Stored-Product Protection, Canberra, Australia, April 1994.

Beilstein Commander (CrossFire Beilstein, or Beilstein Handbook of Organic Chemistry). MDL Information Systems, Inc., 2000.

Bethke, C.M. Geochemical Reaction Modeling: Concepts and Applications. Oxford University Press, Oxford, UK, 1996.

Benitez-Nelson, C. Phosphorus in the upper ocean. Ciencia al Dia InternationalArticulo 1 Biologia, Geociencias, 2000.

Bennet, L.H., Kruger, J., Parker, R.L, Passalia, E., Reimann, C., Ruff, A.W., and Yakowitz, H. Economic Effects of Metallic Corrosion in the United States, Part 1. A report to Congress by the National Bureau of Standards, Washington, DC, 1978.

Bezuidenhout, JJ., Darvas, J.M., and Kotze, J.M. The dynamics and distribution of phosphite in avocado trees treated with phosetyl- Al, S. African Avocado Growers' Assoc. Yearbook 10, 101, 1987. 1

Blumbergs, P., Lillmann, R.C., Gault, R., Hatto, D.L, Ash, A., and Stevens, C.L. Chemical process studies for commercially unavailable materials. Edgewood Arsenal contract report EACR 1510- 2, Edgewood Arsenal, MD, 1973, 139.

Boenig, I.A., Crutchfield, M.M., and Heitsch, C.W. Phosphorus compounds, In KirkOthmer, Encyclopedia of Chemical Technology, 3rd d., Vol. 17. Wiley Interscience, New York, 1982, 490.

Bond, EJ., Dumas, T., and Hobbs, S. Corrosion of metals by the fumigant phosphine, J. Stored Prod. Res., 20(2), 57, 1984.

Brummer, J.R., Keely, J.A., and Munday, T.F. Phosphorus, Kirk- OthmerEncyclopedia of Chemical Technology (online article), December 2000.

Buchanan, G., Compton, H., and Wrobel, J. Search of a white phosphorus munitions disposal site in Chesapeake Bay. Proc. U. S. EPA Forum on Remediation of Superfund Sites Where Explosives Are Present, San Antonio, TX, December 1989.

Burford, J.R., and Bremner, J.M. Is phosphate reduced to phosphine in waterlogged soils? Soil. Biol. Biochem. 4, 489, 1972.

Callahan, M. Promotion of fertilizer in oak disease war called "illegal." April 2001. Accessed May 27, 2005 at http:// www.greenbrae.org/news/042901pressdemocrat.html

Cao, H.R, Liu, J., Zhuang, Y.H., and Glindemann, D. Emission sources of atmospheric phosphine and simulation of phosphine formation Sci. China Ser. B Chem. 43(2), 162, 2000.

Casida, L.E. Microbial oxidation and utilization of orthophosphite during growth. J. Bacterial. 80, 237, I960.

CEFIC. The phosphate file, Centre Europen d'Etudes des Polyphosphate, Belgium, 1997.

Chemical Market Reporter Α. 251(4), 41(1), January 27, 1997.

Chemical Market Reporterb. 251(5), 49(1). February 3, 1997. :

Clariant. Phosphorus Specialities. May 2000. :

Coburn, D.R., Dewitt, J.B., Derby, J.V. Jr., and Ediger, E. Phosphorus poisoning in water fowl, J. Am. Pharm. Assoc., 39, 151, 1950.

Collman, J.P., and Hegedus, L.S. Principles and Applications of Organotransition Metal Chemistry, University Science Books, Mill Valley, CA, 1980.

Cook, A.M., Daughton, C.G., and Alexander, M. Phosphonate utilization by bacteria, J Bacteria. 133(1), 85, 1978.

Cooper, G.W., Onwo, W.M., and Oronin, J.R. Alkyl phosphonic acids and sulfonic acids in the Murchison meteorite, Geochm. Cosmochim. Acta 56, 4109, 1992.

Corbridge, D.E.C. Phosphorus: An Outline of Its Chemistry, Biochemistry and Tech- " nology, Elsevier Scientific, Amsterdam, 1978.

Corbridge, D.E.C., Phosphorus: An Outline of Its Chemistry, Biochemistry and Technology, 3rd d., Elsevier Science, Amsterdam, 1985.

Crowson, R.A. Use of a proprietary geophysical technique to locate valuable ferrophosphorus materials at abandoned elemental phosphorus plants. 16th Annual Regional Phosphate Conference, Lakeland, FL, October 18-19, 2001.

Dacre, J.C., and Rosenblatt, D.H. Mammalian toxicology and toxicity to aquatic organisms of four types of waterborne munitions pollutants-An extensive literature evaluation. U.S. Army Medical Bioengineering Research and Development Laboratory report TR-7403, 1974, 166.

Davies, C., Fraser, D.L., Hertzler, P.C., and Jones, R.T. USEPA's infrastructure needs survey. J. AWWA, 89(12), 30, 1998.

Delaney, M.L. Phosphorus accumulation in marine sediments and the oceanic phosphorus cycle, Global Biogeochemical cycles, 12(4)m, 563, 1998.

Delany, J.M., and Lundeen, S.R. The LLNL thermochemical database. Lawrence Livermore National Laboratory Report UCRL-21658, 1990, 150.

De Romero, M.F., Ferez, O., and Navarro, A. Phosphate attack as caustic corrosion in high pressure boiler tubes, Mater. Perform. 3, 72, 1999.

de Graaf, R.M., Visscher, J., and Schwartz, A. W. A plausibly prebiotic synthesis of phosphonic acids, Nature 378(30), 474, 1995.

de Graaf, R.M., Visscher, J., and Schwartz, A.W. Reactive phosphonic acids as prebiotic carriers of phosphorus. J. Mol. Evol. 44, 237, 1997.

Devai, I., Felfoldy, L., Wieener, H., and Plosz, S. Detection of phosphine: New aspects of the phosphorus cycle in the hydrosphere, Nature 333, 343, 1988.

Devai, L, and Delaune, R.D. Evidence for phosphine production and emission from Lousiana and Florida marsh soils. Org. Geochem., 23, 277, 1995.

Drug Enforcement Administration, Justice (DEA). Rule number 198P, Control of red phosphorus, white phosphorus and phosphorous acid (and its salts) as list I chemicals, 2000. Accessed May 27, 2005 at http://www.deadiversion. usdoj.gov/fed_regs/rules/2000/fr0925.htm

Duce, R.A. The impact of atmospheric nitrogen, phosphorus, and iron species on marine biological productivity. In The Role of Air- Sea Exchange in Geochemical Cycling, d., P. Buat-Menard, D. Reidel, Dordrechthe, The Netherlands, 1986.

Duley, B. Recycling phosphorus by recovery from sewage. Rhodia Consumer Specialties UK Ltd., Grampian, UK, 2001.

Durney, LJ. d., Electroplating Engineering Handbook, 4th d., Van Nostrand Reinhld, New York, 1984.

Eismann, F., Glindemann, D., Bergmann, A., and Kuschk, P. Soils as source and sink of phosphine, Chemosphere 35(3), 523, 1997a.

Eismann, F., Glindemann, D., Bergmann, A., and Kuschk, P. Effect of free phosphine on anaerobic digestion, Wat. Res. 31(11), 2771, 1997b.

Eismann, F., Glindemann, D., Bergmann, A., and Kuschk, P. Balancing phosphorus in manure fermentation, J. Environ. Sci. Health B32(6), 955, 1997c.

Emery, R.N., and Kostas, E. Phosphine fumigation detector cards for farmers-the phoscard, Australian Postharuest Technical Conference, Adelaide, Australia, 2000.

Engel, R. Phosphonates as analogues of natural phosphates, Chem. Rev. 77(3), 345, 1977.

Epstein, S. The Alloys of Iron and Carbon, vol. 1. McGraw-Hill, New York, 1936.

Fenton, M.D. Ferroalloys. USGS minerals publications, 2001.

Fletcher, G.L., Hoyle, RJ., and Home, D.A. Yellow phosphorus pollution: Its toxicity to seawater-maintained brook trout (Salvalinus fontinalis) and smelt (Osmerus mordaz), J. Fish Res. Board Can. 27, 1379, 1970.

Fletcher, G.L., and Hoyle, RJ. Acute toxicity of yellow phosphorus to Atlantic cod (Gadus morhud) and Atlantic salmon (Salmo salar) smolts, J. Fish Res. Board Can., 29, 1295, 1972.

Flis, J. d., Trans. Skup, A. Corrosion of Metals and Hydrogen- Related Phenomena, Elsevier, New York, 1991.

Foster, T.L., and Winans, L., Jr. Anaerobic utilization of phosphite/phosphine as a sole source of phosphorus: Implication to growth in the Jovian environment, Life Sd. Space Res. 15, 81, 1977.

Frank, R., and Rippen, G. Verhalten von phosphin in der atmosphere [Fate of phosphine in the atmosphere]. Lebenstmitteltechnik July/August, 409, 1987.

Freedman, L.D., and Doak, G.O. The preparation and properties of phosphonic acids, Chem. Rev. 57, 479, 1956.

Friel, JJ., and Goldstein, J.I. An experimental study of phosphate reduction and phosphorus-bearing lunar metal particles, Proc. Lunar Sd. Conf. 7th, p. 791, 1976.

Gard, V.K. How to prevent Celphos poisoning, Agric. Tribune, November 1999

Gassmann, G., and Glindemann, D. Phosphine (PH3) in the biosphere, Angew. Chem. Int. Ed. Engl. 32(5), 761, 1993.

Gassmann, G., and Schorn, F. Phosphine from harbor surface sediments, Naturwissenscbafien 80, 78, 1993.

Gassmann, G. Phosphine in the fluvial and marine hydrosphere, Mar. Chem. 45, 197, 1994.

Gassmann, G., and van Beusekom, J.E.E. Offshore atmosphere phosphine, Naturwissenschaften 83, 129, 1996.

Gatti, N. New red phosphorus masterbatches find new application areas in thermoplastics, Plastics Addit. Compound. 4, 34, 2002.

Glindemann, D., Stottmeister, U., and Bergmann, A. Free phosphine from the anaerobic biosphere, Environ. Sd. Pollut. Res. 3(1), 17, 1996a.

Glindemann, D., and Bergmann, A. Phosphine in the lower terrestrial troposphere, Naturwissenschaften 83, 131, 1996b.

Glindemann, D., and Bergmann, A. Spontaneous emission of phosphane from animal slurry treatment processing, Zentralblatt fur Hygiene und Umweltmedizin 198, 49, 1995.

Glindemann, D., Morgenstern, P., Wennrich, R., Stottmeister, U., and Bergmann, A. Toxic oxide deposits from the combustion of landfill gas and biogas, Environ, Sd. Pollut. Res. 3(2), 75, 1996c.

Glindemann, D., Bergmann, A., Stottmeister, U., Eismann, F., and Gassmann, G. Tropospheric phosphine and its possible sources, Proc. EUROTRACSymposium 96, 147, 1996d.

Glindemann, D., Eismann, R, Bergmann, A., Kuschk, P., and St\ottmeister, U. Phosphine by bio-corrosion of phosphide-rich iron, Environ. Sd. Pollut. Res. 5(2), 71, 1998.

Glindemann, D., de Graaf, R.M., and Schwartz, A.W. Chemical reduction of phosphonate on the primitive earth, Origin of Life and Evolution of Biosphere 29, 555, 1999.

Glindemann, D., Edwards, M., and Kuschk, P. Phosphine gas in the upper troposphere, Atmos. Environ. 37, 2429, 2003.

Graham, W.F., and Duce, R.A. Atmospheric pathways of the phosphorus cycle, Geochim. Cosmochim. Acta 43, 1195, 1979.

Guest, D., and Grant, B.R. The complex action of phosphonates as antifungal agents, Biol. Rev. 66, 159, 1991.

Gulick, A. Phosphorus as a factor in the origin of life, Am. Scientist 43, 479, 1955.

Guiyang RC. Chemical Co. Ltd. Product catalog, 2002.

Han, S.H., Zhuang, Y.H., Liu, J.A., and Glindemann, D. Phosphorus cycling through phosphine in paddy fields, Sd. Total Environ. 258(3), 195, 2000.

Hashimoto, S., Fujiwara, K., and Fuwa, K. Determination of phosphate ion by gas chromatography with the phosphine generation technique, Anal. Chem. 57(7), 1305, 1985.

Herro, H.M., and Benweg, A. Phosphate corrosion in high-pressure boiler, MP 49, December 1995.

Hilderbrand, R.L., ed., The Role of Phosphonates in Living Systems. CRC Press, Boca Raton, FL, 1983.

Holford, I.C.R. Soil phosphorus:Its measurement, and its uptake by plants, Aust.J. Soil Res. 35, 227, 1997.

Hilton, H.W., and Robison, W.H. Fate of phosphide and phosphine in soil-water environment. J. Agric. Food Chem. 20(6), 1209, 1972.

Horiguchi, M., and Kandatsu, M. Isolation of 2-aminoethane phosphonic acid from rumen protozoa. Nature 184, 901, 1959.

Idler, D.R. Coexistence of a fishery and a major industry in Placentia Bay. Chem. Can. 21, 16, 1969.

International Fertilizer Industry Association. Global Agriculture Situation and Fertilizer Demand; IFA, Paris, France, 2002.

Iverson, W.P. Corrosion of iron and formation of iron phosphide by Desulfovibrio desulfuricans, Nature 217, 1265, 1968.

Iverson, W.P. Anaerobic corrosion: Metals and microbes in two works, Dev. Ind. Microbiol. 16, 1, 1975.

Iverson, W.P. An overview of the anaerobic corrosion of underground metallic structures, evidence for a new mechanism, underground corrosion. ASTM STP 741. ASTM, Philadelphia, 1981.

Iverson, W.P. Microbial Corrosion. The Metals Society, London, 46, 1983.

Iverson, W.P. Microbial Corrosion of Metals. Advances in Microbial Ecology, vol. 32, ed. A. I. Laskin, Academic Press, San Diego, CA, 1987.

Iverson, W.P. Possible source of a phosphorus compound produced by sulfatereducing bacteria that cause anaerobic corrosion of iron, Mater. Perform. 31(5), 46, 1998.

Iyer, R.N., and Pickering, H.W. Analysis of hydrogen evolution and entry into metals for the discharge recombination process, J. Electrochem. Soc. 136(9), 2463, 1989.

Iyer, R.N., Takeucki, L, Zamanzadeh, M., and Pickering, H.W. Hydrogen sulfide effect on hydrogen entry into iron-? mechanistic study, Corros. Sd. 46(6), 460, 1990.

Jangaard, P.M. Effects of elemental phosphorus on marine life. Circular No. 2. Atlantic Region Office, Research and Development, Fisheries Research Board of Canada, 313, 1972.

Jefferson Lab Science Education. Southeastern Universities Research Association for the U.S. Department of Energy, Newport News, VA, 2002.

Jenkins, R.O., Morris, T.A., Craig, P.J., Ritchie, A.W., and Ostah, N. Phosphine gemeration by mixed and monoseptic cultures of anaerobic bacteria, Sd. Total Environ. 250, 73, 2000.

Johnson, D. Phosphorus derivatives showing muscle. Chemical Market Reporter 252(24), 4(2), 1997.

Kerr, J. 2001. Rodenticide usage on arable farms. Scottish Agriculture Science Agency, 1998, Edinburgh, UK, copyright reserved 2001.

Kittredge, J.S., Horiguchi, M., and Williams, P.M. Aminophosphonic acids: Biosynthesis by marine phytoplankton, Comp. Biochem. Physiol. 29, 859, 1969.

Kononova, S. V., and Nesmeyanova, M.A. Phosphonates and their degradation by microorganisms. BioProt Network, Moscow, 2001.

Kubaschewski, O., Alcock, C.B., and Spencer, PJ. Material Thermochemistry, 6th ed., Pergamon Press, Oxford, 1993.

LaDou, J. Occupational health hazards in the microelectronics industry, Scand. J. Work Environ. Health 9, 42, 1983.

Larson, H.P., Treffers, R.R., and Fink, U. Phosphine in the Jupiter's atmosphere: The evidence from high-altitude observations at 5 micrometers, Astrophys. J. 211, 972, 1977.

Latimer, W.M. The Oxidation States of the Elements and Their Potentials in Aqueous Solutions, 2nd ed., Prentice Hall, New York, 1952.

Lee, A.K. New treatment approved for sudden oak death. California Aggie October 8, 2003.

Lewis, J. S. Observability of spectroscopically active compounds in the atmosphere of Jupiter, Icarus 10, 393, 1969.

Lewis, W.M., Jr., Grant, M.C., and Hamilton, S.K. Evidence that filterable phosphorus is a significant atmospheric link in the phosphorus cycle, Oikos 45, 428, 1985.

Liu, J.A., Yahui, C.H.Z., Kuschk, P., Eismann, F. and Glindemann, D. Phosphine in the urban air of Beijing and its possible sources. Water Air Soil Pollut. 116(3-4), 597, 1999.

Lovatt, C. J. Formulation of phosphorus fertilizer for plants. U.S. Patent 5,514,200, 1990.

Lucas, N.B., Warncke, D.D., and Thorpe, V.A. Phosphite injury to corn. AgronJ. 71, 1063, 1979.

MacIntire, W.H., Winterberg, S.H., Hardin, LJ., Sterges, A.J., and Clements, L.B. Fertilizer evaluation of certain phosphorus, phosphorous, and phosphoric materials by means of pot cultures, Agron.J. 42, 543, 1950.

Makar, J.M., and Rajani, B.B. Grey cast iron water pipe metallurgy, J. Mater. Civil Eng. 12(3), 245, 2000.

Malacinski, G., and Konetzka, W.A. Bacterial oxidation of orthophosphate, J. Bacteriol. 91, 578, 1966.

Manitoba Agriculture and F