Environ. Sci. Technol. 2004, 38, 6662-6668
Sorption-Controlled Degradation Kinetics of MCPA in Soil P I A H . J E N S E N , * ,†,‡ HANS CHR. B. HANSEN,† JIM RASMUSSEN,‡ AND OLE S. JACOBSEN‡ Department of Natural Sciences, The Royal Veterinary and Agricultural University (KVL), Copenhagen, Denmark, and Department of Geochemistry, Geological Survey of Denmark and Greenland (GEUS), Copenhagen, Denmark
The relationship between sorption strength and degradation kinetics has been studied for the pesticide MCPA in a sandy top- and subsoil. After adding two types of sorbents (crushed peat and activated carbon) in various amounts to the sandy soils, sorption, desorption, and mineralization of 14C-MCPA were measured. The obtained Freundlich constants (KF) varied between 0.7 and 27.2 mg(1-nF)‚LnF/kg, and the first-order mineralization rate constants varied between 0.001 and 0.128 d-1. The results showed an inverse relationship between sorption strength and mineralization. A higher KF value corresponded to a smaller mineralization rate and less mineralization. A correlation coefficient of r2 ) 0.934 between the log-transformed Freundlich desorption coefficient (KF,des) and the log-transformed mineralization rate constant (k) was obtained. After 7, 14, 22, and 35 days of incubation, soil samples were consecutively extracted by water, methanol, and 5 M NaOH to separate the remaining 14C into 3 different pools. The extractions showed that the mineralization only proceeded from the water extractable pool of MCPA. Thin-layer chromatography revealed a formation of small amounts of metabolites; 95%. MCPA has a pKa-value of 3.07 (18). Soil. The soil used in this study was an arable sandy soil from Fladerne Bæk, Jutland, Denmark (Typic Fragiorthod (22)). MCPA has been used on the crops for 15 years and has been detected in the groundwater below. Soil samples were collected from the topsoil (A-horizon) (0-25 cm) and the subsoil (BC-horizon) (60-160 cm) with a total organic carbon content (org. Ctot) of 1.8% and 0.1%, respectively (Table 1). The top- and subsoil are hereafter referred to as A and C, respectively. The moist soil was kept in the dark at 5 °C until use. Sorbents. Peat was collected from Lille Vildmose (a former raised bog) and was air-dried and crushed ( 0.938 for all but one soil-sorbent mixture (CPM, desorption); the calculated sorption and desorption coefficients are listed in Table 2. The obtained KF,des is generally higher than the corresponding KF. This hysteresis is an often observed phenomenon (25, 26), which is due to irreversibility in the sorption process caused by e.g. slow kinetics of desorption, diffusion-limited transport within the particles and/or irreversible binding of MCPA or metabolites to the soil particles. 6664
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 24, 2004
Peat and activated carbon alone showed a very high sorption of MCPA with KF values of 12 and 104 times that of the nonamended topsoil. The value of nF obtained for peat is close to unity (0.92), showing sorption strength to be almost independent of the MCPA concentration. However, nF obtained for activated carbon is below unity (0.56), corresponding to a curved isotherm and bonding strengths decreasing with increasing MCPA content. The nF values for the different soil-sorbent mixtures are all close to unity (0.891.04), except for mixtures with a high content of activated carbon (ACM and CCM) (0.60 and 0.44). The nF values of these two mixtures are in the same range as for activated carbon alone (0.56) which is in agreement with activated carbon being the dominating sorbent in these mixtures. Calculations demonstrate that the total sorption in activated carbon amended soils is equal to the sum of the individual sorption of MCPA to activated carbon and the soil, while peat amended soils showed slightly higher sorption than the sum of sorption by the individual components. The sorption strength as determined by KF, increased in the following order for both top- and subsoils: N < PL < PM < CL < PH < CM < CH. Sorption was only slightly increased in the PL samples, and this mixture was therefore excluded from further experiments. Mineralization. The MCPA mineralization kinetics were similar for all soils. The curves of cumulated mineralization showed a sigmoid curvature with three segments: an initial lag-phase where no or little mineralization took place (I), a steep part with rapid mineralization (II), and a plateau where mineralization ceased (III) (Figure 1). The distinction between the segments was made by inspection of the curves. Segment II can be described by first-order kinetics (equation 2)
Ct ) C0‚exp(-k‚t)
(2)
where C0 and Ct are the amounts of 14C in the soil at times 0 and t, respectively, t is the incubation period after the lagphase (days), and k is the first-order rate constant (day-1) (3). Ct is equal to the amount of 14C initially added minus the amount detected as 14CO2. The rate constants are listed in Table 3 together with the total amounts mineralized after 35 days of incubations relative to the initial added 14C (M35) and the duration of the lag-phase for all soil-sorbent mixtures. In the topsoil samples, the different lag-phases were mostly of the same duration (8-9 days), though usually a little longer for the soils amended with activated carbon (812 days). For the subsoil material, a longer lag-phase was
FIGURE 1. Example of a mineralization curve for MCPA in a nonamended topsoil (AN) sample divided into lag-phase (I), steep part (II), and plateau (III).
TABLE 3. Kinetic Parameters Characterizing the Mineralization Curves Obtained for the Soil-Sorbent Mixturesd soil/sorbent
Na
lag-phase (days)
kb (day-1)
M35c (%)
ACL ACM APM APH AN CCL CCM CPM CPH CN
3 3 3 3 3 3 3 3 3 3
8-12 12 8-9 8-9 9 7-14
0.050 ( 0.004 0.004 ( 0.003 0.088 ( 0.004 0.032 ( 0.004 0.128 ( 0.005 0.017 ( 0.024 0.001 ( 0.001 0.090 ( 0.019 0.003 ( 0.001 0.117 ( 0.008
42 ( 2 5.3 ( 3.8 47 ( 2 40 ( 1 56 ( 3 17 ( 23 1.3 ( 1.4 54 ( 13 5.2 ( 1.1 63 ( 3
14-27 14-17 13-20
a Number of replicates. b First-order mineralization rate coefficient for the steep part of the mineralization curve (Figure 1, segment II). c Relative accumulated amount of 14CO up to day 35. d Errors refer to 2 standard deviations.
generally observed compared with topsoils. In addition, the duration of the lag-phase varied more in subsoil samples (7-27 days) than in the topsoil samples. The highest rate constant of the steep part of the curve (k) among the topsoils was seen in the nonamended soil (AN), and k decreased according to amendment in the following order: N > PM > CL > PH > CM. For the subsoil material, CCL was impossible to rank due to large variations of k among the triplicates. The
other soil-sorbent mixtures had the same ranking order as the topsoil samples, though no significant differences were seen between the k values of CPM and CN. Ranking according to M35 instead of k gave similar results. Sequential Extraction. From the mineralization experiments and the extraction steps, 14C-mass balances over time were calculated (Figure 2). The control samples containing no soil and sorbents, which were extracted at day 35, showed a total recovery of > 98%. The total recovery of 14C for the soil-sorbent samples varied between 99.6 ( 2.2% (CN, day 7) and 66.4 ( 3.4% (APH, day 22). The recoveries were generally better in mixtures where little or no mineralization had taken place. Loss of 14C as 14CO2 during sampling is negligible (24), and the 14C not recovered is assigned to a pool with a binding stronger than the pool extractable with 5 M NaOH. For samples where mineralization was observed, a decrease in the amount of 14C in the EWater pool proportional to the 14CO2 developed was seen; e.g. for ACL a decrease in the EWater pool of 19% is seen from day 7 to day 14 corresponds to a development of 14CO2 of 17% of the added 14C-MCPA. Mineralization took place in all mixtures containing 14C in EWater, except for CPH and some replicates of CCL, where only insignificant mineralization appeared. In soils amended with 0.05% activated carbon (CM) no 14C was extracted by MilliQwater, and almost no mineralization was observed either. While EMeOH slightly declined with time, ENaOH showed a slight increase in all of the soil-sorbent mixtures, except CCM. TLC Analysis for Tracing Metabolites. The methanol extracts, which represent the sum of EWater and EMeOH, contained low amounts of metabolites. Most distinct TLC bands were observed for the extracts from the nonamended topsoil (An), samples for which the highest mineralization was observed; only results for the nonamended topsoil are shown. The TLC analysis showed a decrease in the amount of MCPA in the extract over time that corresponded with the production of 14CO2 (Figure 3a). However, the magnitude of the produced 14CO2 was substantially smaller than the decrease of MCPA in the methanol extract. In addition to the band originating from MCPA, four low-intensity bands appeared (Figure 3b-e). The MCPA standard used for spiking the soil contained small amounts of impurities, which are shown by the levels of the bands 2-3 at day 0. Band 1 represents a fraction unable to migrate under the given circumstances, and the intensity of the band increased during the incubation period. The intensity of band 2 also increased during the incubation period. The intensity of band 3 declined
FIGURE 2. Mass balances for MCPA obtained for the different soil-sorbent mixtures sampled at four different times (7, 14, 22, 35 days after start) during incubation: (a) topsoil-sorbent mixtures and (b) subsoil-sorbent mixtures. VOL. 38, NO. 24, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6665
between the log-transformed desorption coefficient (KF,des) and the log-transformed mineralization rate constant (k) (Figure 4a). This shows MCPA sorption and the mineralization rate to be strongly correlated. Correlating the log-transformed mineralization rate constant with the log-transformed KF instead of KF,des results in correlation of r2 ) 0.772 showing the sorption coefficient to correlate as well (Figure 4b). These correlations demonstrate that mineralization is strongly retarded due to sorption and strongly indicates that the mineralization of MCPA proceeds from the solution phase. The correlations were made excluding a few outliers (CPH and CCL). A variable duration of the lag-phases in the subsoil samples (7-27 days) was observed, which may be due to an initial uneven distribution of MCPA degrading microorganisms between the subsoil samples caused by the relatively small amounts of soil used in the incubation experiment (2 g). Hence, the large variations in mineralization between triplicates in CCL may be attributed to different microbial population densities. The unexpected low mineralization observed for CPH may be due to preferential degradation of the peat compared with MCPA (27). The high C/N ratio (58) in the added peat can further support this explanation, since the low N content may result in N limited conditions for the microorganisms. The low pH in these samples is not expected to have influenced the mineralization, since it did not effect the mineralization of 14C-Na-acetate.
FIGURE 3. Fractionation of 14C in methanol extracts from the nonamended topsoil (AN) analyzed by TLC. Extraction performed after 1 to 14 days of incubation. (a) Relative amount of 14C-MCPA in the extracts and the extent of mineralization measured as accumulated 14CO2. (b-e) Relative amount of 14C in the different chromatographic bands other than the band originating from MCPA. from day 0, which indicates that the band represents an impurity originating from the added standard. Band 4 occurred in the extracts from day 9 and later on.
Discussion Correlation between Mineralization and Sorption. Comparing the mineralization rate constant (k) with the sorption strength for the soil-sorbent mixtures, an inverse relationship is observed. The soil with the largest fraction of MCPA sorbed (subscript CM) has the lowest rate of mineralization and vice versa. A linear correlation (r2 ) 0.934) was obtained
Formation of Metabolites. The results obtained by TLC showed 3 bands (band 1, 2, and 4), of which the intensity increased during the incubation period (Figure 3). The intensity of band 1 increased from day 9, which corresponds to the period, where mineralization increased. Band 1 did not migrate during TLC and is tentatively attributed to 14C associated to immobile humic material or built into microbial biomass during the degradation process. The intensity of the two other bands (2 and 4) increased from day 6 and day 9, respectively, which corresponds with the period of mineralization, and these bands are likely to be metabolites formed during degradation of MCPA. One of these bands may represent 4-chloro-2-methylphenol, which is reported as the MCPA metabolite most often found in soil (18, 28). Crespin et al. (28) found the largest amount of this metabolite 8 days after application, where it accounted for 8% of the initially applied MCPA. In our experiment the total activity of the different bands, excluding the band originating from MCPA application, reached a maximum at day 13, where they accounted for 7% of the initially added 14C. If only bands assigned to metabolites (band 2 and 4) are included, the sum accounted for only 4%, demonstrating that only little accumulation of metabolites takes place during the degradation period.
FIGURE 4. Log-log relationships of mineralization of MCPA versus MCPA sorption. (a) The first-order mineralization rate coefficient (k) for segment II (Figure 1) versus the Freundlich desorption coefficient (KF,des). (b) k versus the Freundlich sorption coefficient (KF). See text for explanation of outliers (CPH, CCL). Regression line is drawn excluding outliers. 6666
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 24, 2004
The production of 14CO2 was substantially smaller than the corresponding decrease of 14C in the methanol extract. This is likely to be due to incorporation of some 14C into humic material or microbial biomass during the degradation process (17, 29, 30). This is also demonstrated to some extent by the increase in band 1, but the larger fraction of the 14C incorporated is not expected to be extracted by methanol and will not show up in this TLC analysis. The TLC results demonstrated that when the first degradation step from MCPA to a metabolite proceeded, the next steps leading to CO2 and some incorporation quickly followed, leaving only small pools of metabolites present. Hence, most of the 14C measured in EWater and EMeOH can be attributed to unmodified MCPA. Relationship between Pools and Mineralization. EWater consists of MCPA in solution as well as the readily desorbable MCPA. An inverse correlation between the amount of 14CMCPA mineralized and the amount of 14C in this pool was observed (Figure 2); a depletion of EWater over time resulted in almost an equal increase in 14CO2. Hence, in soil-sorbent mixtures with practically no 14C in EWater only very little mineralization was observed (ACM, CCM). This indicates that only MCPA in EWater was available for microbial degradation. An exception was seen for CPH, where a large EWater was present, but only a little mineralization took place, which could be due to degradation of peat instead of MCPA as discussed above. EMeOH consists of MCPA being sorbed more strongly than MCPA in EWater. EMeOH in general slightly decreased over time, and two processes can be responsible for the depletion. As EWater depletes, some of the sorbed MCPA may desorb to the solution and then become available for mineralization. Over time the binding of MCPA can also change from reversible to irreversible, which will involve a change from EMeOH to ENaOH. ENaOH consists of bound residues and strongly sorbed MCPA/metabolites. In the peat- and the nonamended subsoil no ENaOH was present until mineralization occurred. This coincidence between ENaOH formation and mineralization strongly indicates that ENaOH represents residues formed during the degradation process due to incorporation of degraded 14C-MCPA into humic material or microbial biomass. For activated carbon amended subsoils and for the topsoil-sorbent mixtures the picture is less clear, since ENaOH was seen even in the cases where mineralization had not taken place, indicating some kind of irreversible sorption. Though, during mineralization ENaOH increased, but at the same time EMeOH declined as described above, indicating that ENaOH consists of irreversible sorbed MCPA as well as 14C incorporated into humic material or microbial biomass. Environmental Implications. The study showed that mineralization of MCPA almost entirely takes place from the water extractable pool, corresponding to MCPA in solution or easily desorbed fractions. A highly significant inverse relationship between sorption and mineralization was observed; the best correlation was obtained correlating the degradation rates with the Freundlich desorption coefficients. Sorption coefficients and mineralization are often measured parameters when estimating the fate of a pesticide in soil. The better relationship between the Freundlich desorption coefficient and mineralization found in this study shows the importance of including desorption in such estimations.
Acknowledgments The authors thank for the work done by the technical staff at the Geological Survey of Denmark and Greenland.
Supporting Information Available Data used for estimation of Freundlich sorption/desorption parameters (Table 2) and data used for estimation of kinetic
parameters (Table 3) in Figures S1-4. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Topp, E.; Vallaeys, T.; Soulas, G. In Modern soil microbiology; van Elsas, J. D., Trevors, J. T., Wellington, E. M. H., Eds.; Marcel Dekker: New York, 1997; pp 547-575. (2) Rao, P. S. C.; Bellin, C. A.; Brusseau, M. L. In Sorption and degradation of pesticides and organic chemicals in soil. SSSA Special Publication Number 32; Wisconsin, 1993; pp 1-26. (3) Ogram, A. V.; Jessup, R. E.; Ou, L. T.; Rao, P. S. C. Effects of sorption on biological degradation rates of (2,4-dichlorophenoxy)acetic acid in soils. Appl. Environ. Microbiol. 1985, 49, 582-587. (4) Scow, K. M.; Johnson, C. R. Effect of sorption on biodegradation of soil pollutants. Adv. Agron. 1997, 58, 1-56. (5) Dyson, J. S.; Beulke, S.; Brown, C. D.; Lane, M. C. G. Adsorption and degradation of the weak acid mesotrione in soil and environmental fate implications. J. Environ. Qual. 2002, 31, 613618. (6) Bolan, N. S.; Baskaran, S. Biodegradation of 2,4-D herbicide as affected by its adsorption-desorption behaviour and microbial activity of soils. Austr. J. Soil Res. 1996, 34, 1041-1053. (7) Helweg, A. Degradation of 14C-maleic hydrazide in soil as influenced by adsorption on activated carbon. Weed Res. 1975, 15, 129-133. (8) Moyer, J. R.; Hance, R. J.; McKone, C. E. The effect of adsorbents on the rate of degradation of herbicides incubated with soil. Soil Biol. Biochem. 1972, 4, 307-311. (9) Barriuso, E.; Houot, S.; SerraWittling, C. Influence of compost addition to soil on the behaviour of herbicides. Pestic. Sci. 1997, 49, 65-75. (10) Guo, L.; Jury, W. A.; Wagenet, R. J.; Flury, M. Dependence of pesticide degradation on sorption: nonequilibrium model and application to soil reactors. J. Contam. Hydrol. 2000, 43, 45-62. (11) Guo, L.; Wagenet, R. J.; Jury, W. A. Adsorption effects on kinetics of aldicarb degradation: equilibrium model and application to incubation and transport experiments. Soil Sci. Soc. Am. J. 1999, 63, 1637-1644. (12) Spadotto, C. A.; Hornsby, A. G. Soil sorption of acidic pesticides: Modeling pH effects. J. Environ. Qual. 2003, 32, 949956. (13) Socı´as-Viciana, M. M.; Ferna´ndez-Pe´rez, M.; VillafrancaSa´nchez, M.; Gonzale´z-Pradas, E.; Flores-Cespedes, F. Sorption and leaching of atrazine and MCPA in natural and peat-amended calcareous soils from Spain. J. Agric. Food Chem. 1999, 47, 12361241. (14) Baskaran, S.; Bolan, N. S.; Rahman, A.; Tillman, R. W. Effect of exogenous carbon on the sorption and movement of atrazine and 2,4-D by soils. Austr. J. Soil Res. 1996, 34, 609-622. (15) Oi, M. Time-dependent sorption of imidacloprid in two different soils. J. Agric. Food Chem. 1999, 47, 327-332. (16) Cupples, A. M.; Sims, G. K.; Hultgren, R. P.; Hart, S. E. Effect of soil conditions on the degradation of cloransulam-methyl. J. Environ. Qual. 2000, 29, 786-794. (17) Krieger, M. S.; Merritt, D. A.; Wolt, J. D.; Patterson, V. L. Concurrent patterns of sorption-degradation for oryzalin and degradates. J. Agric. Food Chem. 1998, 46, 3292-3299. (18) Roberts, T. R.; Hutson, D. H.; Lee, P. W.; Nichols, P. H.; Plimmer, J. R.; Roberts, M. C. Metabolic Pathways of Agrochemicals. Part 1; The Royal Socity of Chemistry: Cambridge, U.K., 1998. (19) Helweg, A. Degradation and adsorption of 14C-MCPA in soil influence of concentration, temperature and moisture content on degradation. Weed Res. 1987, 27, 287-296. (20) Mu ¨ ller, M. D.; Buser, H. R. Conversion reactions of various phenoxyalkanoic acid herbicides in soil. 1. Enantiomerization and enantioselective degradation of the chiral 2-phenoxypropionic acid herbicides. Environ. Sci. Technol. 1997, 31, 19531959. (21) Thorstensen, C. W.; Lode, O. Laboratory degradation studies of bentazone, dichlorprop, MCPA and propiconazole in Norwegian soils. J. Environ. Qual. 2001, 30, 947-953. (22) Soil Survey Staff Keys to Soil Taxonomy, 8th ed.; Pocahontas Press: Blacksburg, Virginia, U.S.A., 1999. (23) OECD. OECD guidelines for testing of chemicals. Proposal for updating guideline 106. Adsorption/desorption using a batch equilibrium method; 1997; pp 1-42. (24) Rasmussen, J.; Jensen, P. H.; Holm, P. E.; Jacobsen, O. S. Method for rapid screening of pesticide mineralization in soil. J. Microbiol. Met. 2004, 57, 151-156. VOL. 38, NO. 24, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6667
(25) Jenks, B. M.; Roeth, F. W.; Martin, A. R.; McCallister, D. L. Influence of surface and subsurface soil properties on atrazine sorption and degradation. Weed Sci. 1998, 46, 132-138. (26) Haberhauer, G.; Pfeiffer, L.; Gerzabek, M. H.; Kirchmann, H.; Aquino, A. J. A.; Tunega, D.; Lischka, H. Response of sorption processes of MCPA to the amount and origin of organic matter in a long-term field experiment. Eur. J. Soil Sci. 2001, 52, 279286. (27) Albarra´n, A.; Celis, R.; Hermosı´n, M. C.; Lo´pez-Pin ˜eiro, A.; OrtegaCalvo, J. J.; Cornejo, J. Effects of solid olive-mill waste addition to soil on sorption, degradation and leaching of the herbicide simazine. Soil Use Man. 2003, 19, 150-156. (28) Crespı´n, M. A.; Gallego, M.; Valca´rcel, M.; Gonza´lez, J. L. Study of the degradation of the herbicides 2,4-D and MCPA at different depths in contaminated agricultural soil. Environ. Sci. Technol. 2001, 35, 4265-4270.
6668
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 24, 2004
(29) Houot, S.; Benoit, P.; Charnay, M. P.; Barriuso, E. Experimental techniques to study the fate of organic pollutants in soils in relation to their interactions with soil organic constituents. Analusis 1997, 25, M41-M45. (30) Robertson, B. K.; Alexander, M. Growth-linked and cometabolic biodegradation: possible reason for occurrence or absence of accelerated pesticide biodegradation. Pestic. Sci. 1994, 41, 311318. (31) Allen, S. E. Chemical analysis of ecological materials; Blackwell Scientific Publications: U.K., 1989.
Received for review April 20, 2004. Revised manuscript received September 23, 2004. Accepted September 27, 2004. ES0494095
