Environ. Sci. Technol. 2005, 39, 8295-8299
Phosphine from Rocks: Mechanically Driven Phosphate Reduction? ,†
DIETMAR GLINDEMANN,* MARC EDWARDS,† AND PETER MORGENSTERN‡ Department of Civil and Environmental Engineering, Virginia Tech, 418 Durham Hall, Blacksburg, Virginia 24061, and Department of Analytical Chemistry, UFZsCentre for Environmental Research, Leipzig-Halle, Permoserstrasse 15, D-04318 Leipzig, Germany
Natural rock and mineral samples released trace amounts of phosphine during dissolution in mineral acid. An order of magnitude more phosphine (average 1982 ng PH3/ kg rock and maximum 6673 ng PH3/kg rock) is released from pulverized rock samples (basalt, gneiss, granite, clay, quartzitic pebbles, or marble). Phosphine was correlated to hardness and mechanical pulverization energy of the rocks. The yield of PH3 ranged from 0 to 0.01% of the total P content of the dissolved rock. Strong circumstantial evidence was gathered for reduction of phosphate in the rock via mechanochemical or “tribochemical” weathering at quartz and calcite/marble inclusions. Artificial reproduction of this mechanism by rubbing quartz rods coated with apatitephosphate to the point of visible triboluminescence, led to detection of more than 70 000 ng/kg PH3 in the apatite. This reaction pathway may be considered a mechanochemical analogue of phosphate reduction from lightning or electrical discharges and may contribute to phosphine production via tectonic forces and processing of rocks.
Introduction Phosphine (PH3) is a reactive, toxic gas with a carbide/garliclike odor, which is widely applied to fumigate stored grain (1, 2). “Natural” phosphine gas has also been found in biogas from digesters and sediments (3), in gases from landfills, composts, and animal manure (4), as a trace gas in the atmospheric phosphorus cycle (5, 6), and in off-gases from heated clay (7). “Matrix-bound phosphine” has also been detected in food (8) and in other environmental samples including water and aquatic sediments (9, 10) and soils (11) by acid or alkaline digestion. One reference cited health concerns arising from PH3 emissions to air during flotation of crushed phosphorite mineral (12). The potential sources of phosphine in the environment are of interest. Elemental phosphorus is industrially produced from apatite and carbon at 1500 K (13). Phosphine can then be produced by hydrolysis of the elemental phosphorus. Other natural mechanisms of phosphine production have been investigated. Potential microbial reduction of phosphate to phosphine remains controversial (14-16). Significant release of phosphine and other reduced phosphorus compounds occurs during corrosion of iron that contains phosphorus as part of the metal alloy (17-19). Electrical spark * Corresponding author phone: ++49-345-6879948; fax: ++49345-6871333; e-mail:
[email protected] or
[email protected]. † Virginia Tech. ‡ UFZsCentre for Environmental Research. 10.1021/es050682w CCC: $30.25 Published on Web 09/21/2005
2005 American Chemical Society
plasma discharges can reduce phosphate in apatite to phosphite at a yield as high as 25% (20, 21), while also forming phosphine and odor (22). In the geological environment, sparks and odor are readily produced within pressured sandstone structures in mines (23) or by striking together triboluminescent and piezoelectric minerals such as quartz (24). It has been suggested that hydrogen gas evolution before earthquakes is enhanced by mechanochemical reaction of fractured rock surfaces with water (25). Different theories of mechanochemistry explain these effects. It is well-known that mechanical fracture of solids can increase chemical activity due to the production of fresh surfaces that can produce thermodynamically metastable products at a very small yield (26, 27). The mechanochemical effects have been attributed to production of photons, volatile ions, and recombination products after localized overheating (plasma), athermal bond-rupturing, excitation of vibration, and atomic dislocations. (28, 29). These observations led us to conduct experiments monitoring natural phosphine in rocks and examining phosphate reduction to phosphine via mechanical action.
Experimental Section Sample Description. The natural rock samples (Table 1) were obtained from a geological collection at the University of Leipzig. The monocrystal mineral samples (apatite, calcite) and apatite-mica rock were obtained from the mineralogical collection of the Geosciences Department at Virginia Tech. The NaCl monocrystal was obtained from Korth-Kristalle, Germany. Fresh samples of synthesized vivianite (Fe3(PO4)2) and FePO4 were prepared in our laboratory by precipitation. After the samples were dried, the presence of crystalline vivianite and FePO4 were confirmed by X-ray diffraction. Quartz glass rods (0.006 m in diameter) were obtained from a local glass shop. Total P in Rocks. The analysis of total P as reported in Table 1 was conducted by wavelength dispersive X-ray fluorescence (WDXRF), Siemens SRS 3000. Approximately 5-g pieces of rock were pulverized in an agate mill. Glass disks were produced for XRF by melting 1 g of sample powder with 7 g of lithium tetraborate at 1200 °C. The analyses were calibrated by the use of reference standards NIST 2689, NIST 2691, GBW 07402, GBW 07406, GBW 07407, NIST 2704, GBW 07310, and GBW 07311. The P content of the apatite monocrystals was calculated from the empirical apatite formula Ca5(PO4)3(OH)0.33F0.33Cl0.33. Quartz Mineral in Rocks. The quartz mineral content was provided along with the sample and was significant in the rock samples phosphorite (2%), marl (10%), sandstone (50%). gneiss (30%), clay (15%), quartz pebble (85%), granite (50%), and mica rock (10%), in this study. The quartz content of the samples basalt, glassy lava, and green slate was not detectable (97% 2.5 97% 3 97% 5 182 500 >97% 5 182 500 >97% 5 182 500
NaCl (optical window crystal) calcite apatite (HB 707) apatite (HB 890) apatite (HB 474) NaCl (optical window crystal) calcite apatite (HB 707) apatite (HB 890) apatite (HB 474)
97% 3 97% 3 97% 5 182 500 1425 >97% 5 182 500 2693 >97% 5 182 500 3217
n.d. n.d. 5 × 10-9 5 × 10-9 3 × 10-8
n.d. 39% 61% 89% 29%
n.d. n.d. 8 × 10-9 1 × 10-8 2 × 10-8
79% 25% 16% 5% 27%
7 × 10-8 1 × 10-6 2 × 10-6
90% 20% 32%
5 × 10-6 3 × 10-6 3 × 10-4
31% 55% 83%
rocks, HCl-soluble Splinters, Mechanical Energy Input Low >97% 2.5 73 >97% 2.5 88 >97% 4 11
limestone, freshwater limestone, saltwater marble
Powder, by Rubbing Two Equal Splinters, Mechanical Energy Input High >97% 2.5 73 346 >97% 2.5 88 252 >97% 3 11 2951
limestone, freshwater limestone, saltwater marble quartz, glass powder (rubbing) phosphorite (oligocene) marl green slate with actinolite sandstone gneiss (metamorphic granite) clay basalte quartz rich pebble granite volcanic rock (glassy lava) powderized by mortar in air crumbling pieces (1 mm) powderized by mortar in CH4
Vivianite (Fe3(PO4)2), unground Vivianite (Fe3(PO4)2), mortar-ground FePO4, unground FePO4, mortar-ground no POx addition +PO4 (phosphate) +PO3 (phosphite) +PO2 (hypophosphite) a
5 92 20
5%
7
unknown
9941
37%
rocks, poorly HCl-soluble, powderized by rubbing 22% 2 85000 64% 3 290 18% 4 303 54% 3 520 28% 6 527 24% 2.5 529 46% 6 3332 13% 7 39 7% 6.5 697 5% 7.5 1956
64 157 391 693 964 1248 3271 4966 5834 7562
8 × 10-10 5 × 10-7 1 × 10-6 1 × 10-6 2 × 10-6 2 × 10-6 1 × 10-6 1 × 10-4 8 × 10-6 4 × 10-6
38% 18% 105% 18% 6% 53% 97% 65% 49% 71%
brittle apatite- and mica-rich rock 2 15 000 2 12 000 2 15 500
2246 2724 4081
1 × 10-7 2 × 10-7 3 × 10-7
16% 13% 36%
n.d. n.d. n.d n.d.
n.d. n.d. n.d. n.d.
1 × 10-4 2 × 10-4 1 × 10-4 1 × 10-4
9% 36% 85% 38%
34% 24% 33%
phosphate chemicals, small precipitated/dried grains >97% 2 173 184 >97% 2 173 184 >97% 4 205 298 >97% 4 205 298
97% 3 11 1538
Matrix-bound phosphine and yield are normalized for the dissolved proportion of the rock.
rubbing together two rock pieces. Material for analysis was only collected from fresh internal surfaces, thereby eliminating the potential for contamination from iron tools used to collect the samples. Splinter samples were obtained with very little mechanical energy input by manually crushing two pieces and breaking a splinter down to 0.1 g of mass for analysis. Only the HClsoluble rocks (marble or lime) and mineral monocrystals (apatite, calcite, or NaCl) could be analyzed with this method. Powder samples were obtained with relatively high mechanical energy input by rubbing two rock or crystal pieces at hand pressure. A mass of 0.1 g of powder was collected for analysis. One powder sample of an apatite-rich mica rock was produced by using an agate mortar, because the rock was too brittle to use the manual pulverization approach. 8296
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 21, 2005
While no mechanical device was available to measure the necessary energy input to produce these splinter and powder samples, it is estimated that producing a powder via mechanical rubbing uses at least an order of magnitude more energy than producing a splinter by a striking mechanism. The Mohs hardness (Table 1, Figure 1) is also expected to be related to the energy required to produce power via rubbing. The length of time of mechanical rubbing to produce the powder increased as much as 20 times as the reported Mohs hardness (Table 1) increased. Preparing “Synthesized Rock Samples” Combining the Essential Minerals Quartz and/or Apatite. These samples were devised to show that the reduction of phosphate to phosphine could be mechanically activated through triboluminescence of rubbed (cracked) quartz coated with apatite.
FIGURE 1. Matrix-bound phosphine of pulverized rocks versus rock hardness from compiled data from Table 1, comprising marble, limestones, and poorly HCl-soluble rocks.
FIGURE 2. Matrix-bound phosphine of powders created by rubbing quartz glass rods and apatite monocrystals. The yield of phosphine for the largest signal is still as low as 4 × 10-07 mg PH3/kg apatite P. The embedded image shows the optical triboluminescence of the two crossed and rubbed quartz glass rods used. The quartz rods and apatite monocrystals were approximately 5-mm-thick. First, two cleaned quartz glass rods with unknown P content were rubbed manually to the point of visible triboluminescence (see image in Figure 2) in phosphinefree air. This produced a quartz powder suitable for analysis of matrix phosphine for the quartz (phosphine per dissolved fraction of estimated 5% of the quartz powder). Second, two apatite (HB 474) monocrystals were rubbed manually to analyze the resulting apatite powder for analysis of phosphine. No triboluminescence was expected or observed from rubbing apatite. Third, one apatite crystal (HB 474) was rubbed against one quartz glass rod, resulting in apatite powder for analysis of phosphine. No visible triboluminescence was expected or observed when rubbing apatite against quartz. Fourth, the quartz glass rod used for the previous experiment, with an attached apatite film, was rubbed against another quartz glass rod to the point of visible triboluminescence. The apatite-impregnated segment was then cut from the quartz rod and analyzed for matrix phosphine. A control quartz rod without apatite attached was also analyzed by this same method. The apatite dissolved was determined by the mass difference of the rod after HCl treatment, since bulk quartz did not significantly dissolve in acid. The reported results are averages from three independent experiments (standard deviation ranging from 27% to 37%). Analysis of Matrix-Bound Phosphine in Rock Samples. The analysis of matrix-bound phosphine in rock samples is similar to that for the analysis of phosphine pesticide residues in phosphine-fumigated food, where the food sample is digested in mineral acid and the produced phosphine gas is quantified (8). Matrix-bound phosphine in a rock sample in this work is defined as the mass of liberated phosphine gas
per mass of HCl digested rock sample. The cause of the liberated phosphine gas during HCl dissolution of the rock could be desorption of adsorbed phosphine, hydrolysis of phosphide, or disproportionation if reduced species such as elemental phosphorus were present. For the measurement, a 0.1-g rock splinter or powder was heated for 5 min in 3 mL of 1-N HCl in a nitrogen atmosphere. The total mass of formed phosphine was measured by purging the solution with nitrogen and analyzing the gas by cryo-trap gas chromatography (GC) as described elsewhere (5). The accuracy/precision of the GC method using PH3 standards was 12%. The mass of the rock sample that dissolved during the test was determined as the weight difference of rock solids before and after acidic digestion, filtration, and drying the solid filtrate at 105 °C. Matrix-bound phosphine as reported in Table 1 was calculated by dividing the total mass of produced phosphine by the mass of the sample that dissolved on acidic digestion. The percent solubility of the rock sample as reported in Table 1 was calculated by dividing the mass of the rock sample that dissolved by the initial mass of the sample. The detection limit of matrix-bound phosphine in rock was 3 ng/kg. Triplicate HCl digestions and measurements were conducted on all samples. The average relative standard deviation of matrix phosphine was (45% and ranged from 5% to 105% (Table 1). Statistical tests for significance in this work were conducted using a t-test at 95% confidence. The normalized yield of phosphine (mg PH3/mg P, Table 1), was determined by dividing the experimentally determined matrix-bound phosphine (ng/kg, reported in Table 1) by the total P content (mg/kg, reported in Table 1).
Results and Discussion Matrix-bound phosphine was detected in all mineral and rock samples at low levels (Table 1). The yield of PH3 as fraction of total P in the dissolved part of the samples ranged from 10-4 to 10-9 mg PH3/mg P (Table 1). The HCl-soluble mineral apatite, splinters, and powders contained phosphine at a similar order of magnitude (9223217 ng/kg). It can be concluded that the bulk interior of these apatites contains natural phosphine, and statistical tests indicated that pulverization did not significantly increase the phosphine content of these samples. Splinters and powders of the minerals calcite, vivianite, and NaCl yielded nearly undetectable phosphine. The laboratory samples of iron phosphates produced from reagent grade chemicals did not form detectable phosphine with or without grinding. Consequently, mechanical action alone is not sufficient to reduce phosphate to phosphine. Rocks, HCl-Soluble. Splinters of limestone and marble produced with low mechanical energy input contained a low level of phosphine below 100 ng/kg. However, powders produced with high mechanical energy input contained significantly more phosphine in the case of limestone powder (5-70 times increase) and in marble powder (about 100 times increase). Rocks, Poorly HCl-Soluble, Pulverized by Rubbing. All of the rocks that were sparingly soluble in HCl and that were pulverized by high mechanical energy input liberated matrixbound phosphine (Table 1). Phosphine in rock powders is positively correlated with the Mohs hardness (Figure 1, R2 ) 0.713), which in turn was related to the higher estimated manual energy requirements to pulverize the samples. The correlation of phosphine with the phosphorus content of the rocks is positive as expected but is very weak (R2 ) 0.1867) even if the outlier phosphorite is not included. The rock with the highest P content (phosphorite) yielded the lowest phosphine (64 ng/kg), but quartz pebbles with the lowest P content had nearly the highest phosphine (4966 ng phosphine VOL. 39, NO. 21, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
8297
per kg of dissolved quartz) and the highest yield (10-4 mg phosphine per mg phosphorus in the dissolved quartz). Powder Produced with a Mortar from a Brittle Apatiteand Mica-Rich Rock. Such a powder resulted in 2246 ng/kg phosphine and a low yield of about 10-7, with no significant difference to loose grains obtained from the same rock. Because the authors had initially speculated that this phosphate rich rock would yield much more phosphine, an air atmosphere was exchanged for a methane atmosphere. The concentration of phosphine approximately doubled (4081 ng/kg), but the yield was still very low. Rock Powder of Marble, HCl-Amended with 0.1 M OxyAcids of Phosphorus (POx). If acids of P(I), P(III), and P(V) species were added onto the marble at a concentration of 0.1 M, then phosphine yields were not significantly increased. It is therefore unlikely that phosphine production occurs via a classic chemical reduction pathway starting with any of these species. “Artificial Rock Samples” (Figure 2). When the apatitecoated quartz glass rod was rubbed against another quartz glass rod, very high levels of matrix-bound phosphine were created (>70 000 ng phosphine/kg dissolved apatite, Figure 2). The creation of phosphine in this artificial sample may be analogous to that occurring in prior observations for natural quartzite or granite rock, where quartz is coaggregated with apatite crystallites. Control samples with apatite rubbing against apatite (4800) or apatite versus quartz (14 100) produced much less phosphine and no visible triboluminescence. Rubbing two clean quartz rods resulted in about 10 000 ng phosphine per kg dissolved quartz, which would translate into a very high yield of mg phosphine per mg phosphorus in the dissolved quartz (with unknown but likely orders of magnitude lower P content compared to apatite). Evaluation of Gases from Striking Rocks for Free Phosphine and Odor. Striking or rubbing of a pair of the triboluminescent rocks (quartzite and marble) or of quartz glass rods resulted in yellow sparks and in a pungent odor similar to an electrical discharge. The occasional analysis of the odorous gas phases produced during the pulverization of the rocks in gas bags did not result in enough “free phosphine” from the dry powder to cause any odor. Therefore we speculate that the odor from striking triboluminescent minerals is caused by the gases nitrogen monoxide (NO) and ozone (O3) or other compounds, but it is not created by phosphine. Other odorous volatile emissions of rocks could be elemental chlorine or clusters of atoms (28). Health problems caused by free PH3 in air during carbonate flotation of crushed phosphorite ore during mining were reported (12). We speculate that acidic flotation chemicals (phosphoric acid and fatty acids) combined with the mechanical action caused liberation of matrix-bound phosphine at levels deemed to be a human health concern. Discussion of a Phosphine-Forming Mechanism. Reduction of phosphate in apatite by dissociation in thermal plasma and trapping of the formed phosphorus species by quenching (rapid cooling) with air has been demonstrated (37). Formation of elemental phosphorus (P4, P2, and atomic P, precursors of phosphine) is expected above temperatures of 2500 K, and phosphorus atomization is almost complete above 5000 K. It has been established that powerful reduction reactions are possible via mechanochemistry, and the temperature in the activation zone can be 3500-5000 K (35). These temperatures exceed the 1500 K required to industrially produce phosphorus and phosphine from phosphate in the presence of chemical reductants (13, 36). In addition, hydrogen gas can be formed at room temperature during the grinding of quartz or granite rock wetted with pure water (38, 39), which demonstrates the existence of strong reductive mechanochemical forces. “Mechanocatalytic” reduction of water to 8298
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 21, 2005
FIGURE 3. Proposed conditions of phosphine production in a propagating crack of quartz, allowing phosphate to thermally dissociate partially into reduced forms of phosphorus, which are preserved when the plasma is quenched and generate matrix-bound phosphine. H2 also occurs after rubbing of water-suspended oxide powders such as Fe3O4 or NiO onto quartz glass, with reported conversion of mechanical energy to H2 production as high as 4.3% (40). Marginal phosphate reduction to phosphine in the absence of chemical reductants has also been observed during experiments with millimeter-sized sparks (20-22). In combination, it is likely that the above mechanochemical factors are responsible for de novo phosphine production in this work. This finding, along with the identification of trace phosphine in nearly all rock samples, has implications for future studies of environmental phosphorus chemistry. Trace phosphine may have been incorporated into the rock structure when it was first formed or created from reduction of phosphate during crystallization and pressurization over geologic time scales. But after input of intensive mechanical energy, up to 100 times more new or de novo matrix-bound phosphine is present in limestone and marble after mechanochemical effects of pulverization. We cannot completely rule out the possibility that traces of natural iron phosphides were in these samples, initially recalcitrant to HCl dissolution but dissolving more effectively after pulverization, yielding higher phosphine. However, the fact that de novo phosphine was clearly produced using quartz rods coated with apatite after rubbing to the point of visible triboluminescence is strong circumstantial support for the mechanochemical or “tribochemical” pathway to phosphine production at quartz and calcite/marble inclusions. It seems likely that natural geo-mechanical crack stress could produce natural phosphine via a similar mechanochemical mechanism. Specifically, mechanical weathering causes stress-cracking of quartz crystallites. The flash of triboplasma at the crack tip (Figure 3) causes the mechanochemical reduction of phosphate impurities in the quartz to matrix-bound phosphine. The triboplasma atomizes some phosphate material and quenching of the plasma leads to a metastable recombination of phosphorus atoms resulting in matrix-bound phosphine. Phosphine is then formed in the natural plasma without any chemical reductants present, although at a low yield (yield of 4 × 10-7 mg PH3/mg apatite P as per Figure 2). If chemical reductants (e.g., hydrogen, methane, CO, graphite, and diamond) were included in quartz (30-32), then we speculate that the yield could be increased. Reductants such as H2 and CH4 could also be formed in situ in the crack zone of quartz by water splitting. Apatite crystals that are separated from the quartz crystals cannot be reduced to phosphine because they are not triboluminescent. Therefore, phosphorite, apatite monocrystals, and iron phosphates yielded low phosphine per total P content in Table 1. The result that natural and pulverized rocks and minerals contain phosphine is of environmental significance because of the relative rarity of reduced phosphorus in the environment and the importance of the atmospheric pathway in the global phosphorus cycle (41). Another review (16) noted that
the atmospheric loading of phosphorus is quite high and assumed that the phosphorus is exclusively due to phosphate present in dust. Established release of gaseous phosphine to the atmosphere from steel-making and phosphine fumigants could account for a small but significant fraction of atmospheric phosphine loading. To the extent that natural mechanochemistry yields phosphine that was loaded to the atmosphere, it should be considered in the global atmospheric phosphorus cycle.
Acknowledgments This work was supported by the National Science Foundation (NSF; Grant No. BES-0201849). The opinions expressed are those of the authors and not necessarily those of the NSF. We thank R. J. Bodnar, J. Gilman, A. W. Schwartz, and K. P. Thiessen for consultation, the mineral collection of the Geosciences Department at Virginia Tech, and the mineral collection of the faculties for Mineralogy and Geo-Sciences, University of Leipzig.
Literature Cited (1) World Health Organization (WHO). Phosphine and Selected Metal Phosphides; Environmental Health Criteria 73; World Health Organization: Geneva, Switzerland, 1988. (2) Gmelin Handbook of Inorganic and Organometallic Chemistry: Phosphorus; Springer: Berlin, 1993; Suppl. Vol. C1. (3) Devai, I.; Felfo¨ldy, L.; Wittner, I.; Plo´sz, S. Detection of phosphine: New aspects of the phosphorus cycle in the hydrosphere. Nature 1988, 333, 343-345. (4) Glindemann, D.; Stottmeister, U.; Bergmann, A. Free phosphine from the anaerobic biosphere. Environ. Sci. Pollut. Res. Int. 1996, 3, 17-19. (5) Glindemann, D.; Bergmann, A.; Stottmeister, U.; Gassmann, G. Phosphine in the lower terrestrial troposphere. Naturwissenschaften 1996, 83, 131-133. (6) Glindemann, D.; Edwards, M.; Kuschk, P. Phosphine (PH3) gas in the upper troposphere. Atmos. Environ. 2003, 37, 2429-2433. (7) Heller-Kallai, L.; Miloslavski, I.; Aizenshtat, Z.; Halicz, L. Chemical and mass spectrometric analysis of volatiles derived from clays. Am. Mineral. 1988, 73, 376-382. (8) Nowicki, T. W. J. Gas-liquid-chromatographic and flame photometric detection of phosphine in wheat. J. Assoc. Off. Anal. Chem. 1978, 61, 829-836. (9) Gassmann, G.; Glindemann, D. Phosphane in the Biosphere. Angew. Chem., Int. Ed. Engl. 1993, 32, 761-763. (10) Niu, X. J.; Geng, J. J.; Wang, X. R.; Wang, C. H.; Gu, X. H.; Edwards, M.; Glindemann, D. Temporal and spatial distributions of phosphine in Taihu Lake, China. Sci. Total Environ. 2004, 323, 169-178. (11) Eismann, F.; Glindemann D.; Bergmann A.; Kuschk P. Soils as source and sink of phosphine. Chemosphere 1997, 35, 523533. (12) Proshkina, T. A. Industrial hygiene during the flotation concentration of phosphorite ores. Gig. Tr. Prof. Zabol. 1971, 15, 8-11 (in Russian). (13) Jacob, K. D.; Reynolds, D. S.; Hill, W. L. Reduction of tricalciumphosphate by carbon. J. Ind. Eng. Chem. 1929, 21, 1126-32. (14) Schink B.; Friedrich, M. Phosphite oxidation by sulphate reduction. Nature 2000, 406, 37. (15) Roels, J.; Verstraete, W. Biological formation of volatile phosphorus compounds. Bioresour. Technol. 2001, 79, 243-250. (16) Morton, S. C.; Edwards, M. Reduced phosphorus compounds in the environment Crit. Rev. Environ. Sci. Technol. 2005, 35, 333-364. (17) Glindemann, D.; Eismann, F.; Bergmann, A.; Kuschk, P.; Stottmeister, U. Phosphine by bio-corrosion of phosphide-rich iron. Environ. Sci. Pollut. Res. Int. 1998, 5, 71-74. (18) Morton, S. C.; Glindemann, D.; Edwards, M. A. Phosphates, phosphites, and phosphides in environmental samples. Environ. Sci. Technol. 2003, 37 1169-1174.
(19) Roels, J.; Verstraete, W. Occurrence and origin of phosphine in landfill gas. Sci. Total Environ. 2004, 327, 185-196. (20) Glindemann, D.; De Graaf, R. M.; Schwartz, A. W. Chemical reduction of phosphate on the primitive earth. Origins Life Evol. Biosphere 1999, 29, 555-561. (21) DeGraaf, R. M.; Schwartz, A. W. Reduction and activation of phosphate on the primitive earth. Origins Life Evol. Biosphere 2000, 30, 405-410. (22) Glindemann, D.; Edwards, M.; Schrems, O. Phosphine and methylphosphine production by simulated lightningsA study for the volatile phosphorus cycle and cloud formation in the earth atmosphere. Atmos. Environ. 2004, 38, 6867-6874. (23) Canaval, R. Luminous phenomena in the construction of mine headings. Z. Prakt. Geol. 1910, 17, 440-446. (24) Johnsen, A. Sparks and odors produced by the striking together of minerals. Centr. Min. Geol. 1919, 227-232. (25) Ito, T.; Nagamine, K.; Yamamoto, K.; Adachi, M.; Kawabe, I. Preseismic hydrogen gas anomalies caused by stress-corrosion process preceding earthquakes. Geophys. Res. Lett. 1999, 26, 2009-2012. (26) Thiessen, K. P. J. On the origin of increased activity in mechanochemistry of solids. Chim. Phys. 1986, 83, 717-724. (27) Boldyrev, V. V. Mechanochemistry and mechanical activation. Metabstable, mechanically alloyed and nanocristalline materials, Parts 1 and 2. Mater. Sci. Forum 1996, 225, 511-519. (28) Urakaev, F. K.; Boldyrev, V. V. Correlation between the yield of volatile products and the characteristics of the propagation of cracks in crystals. Russ. J. Phys. Chem. 2000, 74, 1339-1344. (29) Gilman, J. Mechanochemistry. Science 1996, 274, 65-70. (30) Hall, D. L.; Bodnar, R. J. Methane fluid inclusions from granulitessA product of hydrogen diffusion. Geochim. Cosmochim. Acta 1990, 54, 641-651. (31) Barker, C.; Underwood, W. D. Analysis of gases in individual fluid inclusions in natural minerals using a dual mass spectrometer system. Analyst 1992, 117, 1407-1411. (32) Cesare, B. Graphite precipitation in C-O-H fluid inclusions: Closed system compositional and density changes, and thermobarometric implications. Contrib. Mineral. Petrol. 1995, 122, 25-33. (33) Mysen, B. O.; Ryerson, F. J.; Virgo, D. The structural role of phosphorus in silicate metls. Am. Mineral. 1981, 66, 106-117. (34) Rhodes, H. E.; Apai, G. Impurities in silicon dioxide: Atomic states of phosphorus and arsenic. Mater. Res. Soc. Symp. Proc. 1987, 82, 103-108. (35) Chapman, G. N.; Walton, A. J. J. Triboluminescence of glasses and quartz. Appl. Phys. 1983, 54, 5961-5965. (36) Klimovich, A. I. Theoretical bases of the reduction of calcium orthophosphate by gaseous reducing agents in the solid phase. Zh. Prikl. Khim. (Leningrad) 1975, 48, 732-737. (37) Krasovskaya, L. I.; Mosse, A. L. Thermodynamic analysis of the reacting gaseous system N + O + P. High Energy Chem. 1976, 10, 210-213. (38) Schrader, R.; Wissing, R.; Kubsch, H. Z. Surface chemistry of mechanically activated quartz. Z. Anorg. Allg. Chem. 1969, 365, 191-198. (39) Kita, I.; Matsuo, S.; Wakita, H. J. Molecular hydrogen generation by reaction between water and crushed rock: An experimental study on molecular hydrogen degassing from the active fault zone. J. Geophys. Res., B 1982, 87 (13), 10789-10795. (40) Ikeda, S.; Takata, T.; Komoda, M.; Hara, M.; Kondo, J. N.; Domen, K.; Tanaka, A.; Hosono, H.; Kawazoe, H. Mechano-catalysissA novel method for overall water splitting. Phys. Chem. Chem. Phys. 1999, 1, 4485-4491. (41) Glindemann, D.; Edwards, M.; Liu, J. A.; Kuschk, P. Phosphine in soils, sludges, biogases and atmospheric implicationssA review. Ecol. Eng. 2005, 24, 457-463.
Received for review April 9, 2005. Revised manuscript received July 11, 2005. Accepted August 8, 2005. ES050682W
VOL. 39, NO. 21, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
8299
