ARTICLE pubs.acs.org/JAFC
Sorption Desorption of Indaziflam in Selected Agricultural Soils Diego G. Alonso,*,† William C. Koskinen,§ Rubem S. Oliveira, Jr.,† Jamil Constantin,† and Suresh Mislankar# †
Center for Advanced Studies in Weed Research (NAPD), Agronomy Department (NAPD/UEM), Universidade Estadual de Maringa, Av. Colombo 5790, Maringa, PR 87020-900, Brazil § Agricultural Research Service, U.S. Department of Agriculture, 439 Borlaug Hall, 1991 Upper Buford Circle, St. Paul, Minnesota 55108, United States # Environmental Fate and Exposure Assessment, Bayer Crop Science LP, Stilwell, Kansas 66085, United States ABSTRACT: Indaziflam, a new alkylazine herbicide that inhibits cellulose biosynthesis, is under current development for soil applications in perennial crops and nonagricultural areas. Sorption and desorption of indaziflam in six soils from Brazil and three soils from the United States, with different physical chemical properties, were investigated using the batch equilibration method. Sorption kinetics demonstrated that soil solution equilibrium was attained in 0.99). Kf values of the Brazilian oxisols ranged from 4.66 to 29.3, and 1/n values were g0.95. Sorption was positively correlated to %OC and clay contents. U.S. mollisol Kf values ranged from 6.62 to 14.3; 1/n values for sorption were g0.92. Kf values from mollisols were also positively correlated with %OC. These results suggest that indaziflam potential mobility, based solely on its sorption coefficients, would range from moderate to low in soil. Desorption was hysteretic on all soils, further decreasing its potential mobility for offsite transport. KEYWORDS: adsorption, hysteresis, alkylazine, soil properties
’ INTRODUCTION Indaziflam (N-[(1R,2S)-2,3-dihydro-2,6-dimethyl-1H-inden-1-yl]6-[(1R)-1-fluoroethyl]-1,3,5-triazine-2,4-diamine) (Figure 1) is a new herbicide active ingredient, classified as a member of the new chemical class “alkylazine”. The herbicide acts on meristematic growth and inhibits cellulose biosynthesis (CBI). Similarly to other CBI herbicides, it is used for preemergence control and results in the inability of weed seedlings to grow. It is used at rates between 25 and 100 g of active ingredient (ai) per hectare and controls annual grasses such as crabgrass, goosegrass, and annual bluegrass as well as 65 other grasses and broadleaf weeds.1 It is also currently in development for use in residential and commercial areas (lawns, ornamentals, and hardscapes including patios, walkways, etc.), turf (parks, cemeteries, golf courses, sod farms, sports fields, and commercial lawns), field-grown ornamentals and Christmas trees, commercial nursery and landscape plantings, and forestry sites.2 Despite the low application rates, it appears to have long residual activity for most proposed uses as a result of its long persistence in soil (t1/2 = 150).2 Like other preemergence herbicides applied to soil, it is necessary to understand the fate and behavior of this herbicide in soil systems and, consequently, the potential risk of contamination of water resources. Sorption desorption processes are important in determining the fate and distribution of agrochemicals in the soil/ water environment, because they determine the amount of pesticide that can reach the target organism and the amounts that can be volatilized, degraded, and leached.3 Sorption desorption is dependent on the properties of the chemical and physicochemical composition of the soil.4 Among the main factors affecting the sorption potential of soil are organic matter content, soil pH, and clay contents. Besides clay contents, the type of clay is crucial in the process of sorption.5,6 r 2011 American Chemical Society
As a result, research on indaziflam sorption and desorption in a wide range of soils from different regions of the world is needed to determine the effects of physicochemical properties on these processes. Very limited information on indaziflam is available so far. Indaziflam applied at 35, 52.5, and 70 g ai ha 1 for early preemergence (early-PRE), PRE, and early-postemergence (POST) timings on the basis of soil temperature, respectively, controlled smooth crabgrass at 89 100%.7 With regard to sorption to soil, the only available information is that provided by the label of the registered product,1 which characterized it as low to medium mobility in soil. This study was aimed at characterizing the sorption and desorption coefficients of indaziflam in soils from tropical and glacial regions and is the first published information on this subject.
’ MATERIALS AND METHODS Soils. Six Brazilian oxisols and three U.S. mollisols, previously untreated with indaziflam, with various physicochemical characteristics were selected for this study. Samples were collected from the 0 10 cm depth, air-dried, and passed through a 2 mm sieve. The physicochemical properties of soils are given in Table 1. Sand and clay contents were determined by the hydrometer method. Soil pH was measured in a 1:2 soil/deionized water mixture. The organic carbon (OC) content was determined by loss on ignition.8 Chemicals. 14C-Labeled (triazine-2,4,-14C) indaziflam and pure analytical standards were graciously provided by Bayer Crop Science (Wuppertal, Germany). Solutions ranging from 0.99 to 0.033 μmol L 1 Received: July 26, 2011 Revised: November 6, 2011 Accepted: November 9, 2011 Published: November 09, 2011 13096
dx.doi.org/10.1021/jf203014g | J. Agric. Food Chem. 2011, 59, 13096–13101
Journal of Agricultural and Food Chemistry
ARTICLE
Figure 1. Main physicochemical properties of indaziflam.1
Table 1. Physicochemical Properties of Soils clay mineralogya soil
soil classificationb
origin (city, state)
pH H2O OC (%) sand (%) clay (%) texturec
major
minor
BRA1
Marialva, PR
oxisol Rhodic Hapludox (Latossolo Vermelho distroferrico)
6.0
1.60
27
65
C
K, H, 2:1
BRA3
Pres. Castelo
oxisol
5.4
0.50
88
10
LS
K, H, 2:1, Gib Go, MgH, An, Qz
5.9
2.05
38
57
C
K, H, Gib
Go, Qz
6.2
2.17
39
59
C
K, H, Gib
VHE
Branco, PR BRA7
Vilhena, RO
Typic Hapludox
Gib, MgH, An, Go, Qz, VHE
(Latossolo Vermelho distrofico) oxisol
Rhodic hapludox
(Latossolo Vermelho distrofico) BRA8
Rio Verde, GO
oxisol
Rhodic hapludox
(Latossolo Vermelho distrofico) BRA11
Santa Maria, RS
Arenic albaqualfs (Planossolo Haplico eutrofico)
6.0
1.06
62
17
SL
K, 2:1
VHE, Qz
BRA12
Barra do
Typic quartzipsaments (Neossolo
6.5
0.61
92
7
S
K, Go
H, Qz
8.1
2.18
56
19
L
nad
na
8.3
1.10
54
23
L
na
na
6.0
2.52
33
15
SL
na
na
Bugres, MT USA1
Morris, MN (footslope)
USA2
Morris, MN (shoulder)
USA3
Rosemount, MN
Quartzar^enico ortico) mollisol
Typic calciaquolls
(Maryland sandy loam) mollisol
Typic calciudolls
(Buse sandy clay loam) mollisol Typic hapludolls (Waukegan silt loam)
a K, kaolinite; H, hematite; Gib, gibbsite; Go, goethite; VHE, vermiculite with Al-hydroxy interlayer; An, anatasio; Qz, quartz; MgH, maghemite; 2:1, smectites and/or vermiculites. b According to Soil Taxonomy and Brazilian Soil Science Society. c C, clay; S, sand; SL, sandy loam; LS, loamy-sand. d na, information not available.
were prepared in 0.01 N CaCl2, and 14C-indaziflam was added to give a solution radioactivity of ∼72 Bq mL 1. Final solutions contained 94% of the total sorption found after a 48 h period occurring within the first 2 h after soil solution contact. Sorption kinetics did not differ in the two soils studied, and equilibrium was reached simultaneously in soil USA2, which had OC content 1.9 times greater than soil USA1. Other weak acid herbicides such as pyrithiobac sodium, 2,4-D, aminocyclopyrachlor, picloram, and aminopyralid had similarly fast sorption.5,10 12 The equilibrium was attained in 0.99) (Figure 3; Table 2). The values of 1/n (0.92 1.03) indicate an approach to linearity of the sorption isotherm, which means that sorption was minimally concentration-dependent. However, a small decline of 0.1 in the 1/n values may result in ∼3% of the difference between Koc values from the highest concentration solution and lowest concentration solution used in the experiment. For the soil with the lowest 1/n value of 0.92, USA3, the difference between Koc values from the highest initial concentration solution and lowest initial concentration solution was ∼26%. Because the majority of the 1/n values were not significantly different from each other, Kf values were used for comparisons.
Table 2. Freundlich Sorption Parameters for Indaziflam for Selected Soils soil
Kf (μmol (1 1/n) L1/n kg 1)
Kf,oc (μmol (1 1/n) L1/n kg 1)
1/n sorption
R2
BRA1
19.10 (18.71 19.50)a
1194 (1169 1219)a
0.95 ( 0.01
0.9999
BRA3
BRA7
BRA8
BRA11
BRA12
USA1
a
4.66 (4.55 4.78)
29.27 (27.43 31.23)
9.00 (8.65 9.37)
11.87 (11.09 12.71)
5.23(5.05 5.43)
10.86 (9.05 13.03)
933 (911 956)
1428 (1338 1524)
415 (399 432)
1120 (1046 1199)
858 (827 889)
498 (415 598)
0.95 ( 0.01
1.03 ( 0.01
0.98 ( 0.01
0.98 ( 0.02
0.99 ( 0.01
0.95 ( 0.05
0.9999
0.9996
0.9997
0.9993
0.9997
0.9949
USA2
6.62 (6.40 6.85)
602 (582 623)
0.94 ( 0.01
0.9998
USA3
14.26 (13.88 14.66)
566 (551 582)
0.92 ( 0.01
0.9999
initial concn desorption (μmol L
1
)
1/n desorption
R2
H
0.99
0.04
0.6531
0.04
0.03
0.03
0.7107
0.03
0.99
0.29
0.8256
0.3
0.03
0.23
0.8758
0.24
0.99
0.09
0.8607
0.09
0.03
0.08
0.9624
0.08
0.99
0.11
0.8025
0.12
0.03
0.13
0.8905
0.13
0.99
0.25
0.9543
0.26
0.03
0.24
0.9222
0.25
0.99
0.24
0.9428
0.25
0.03
0.19
0.9082
0.19
0.99
0.13
0.8629
0.14
0.03
0.09
0.8264
0.10
0.99 0.03
0.28 0.18
0.8274 0.9736
0.29 0.19
0.99
0.16
0.9585
0.18
0.03
0.12
0.9211
0.13
Numbers in parentheses are confidence intervals (Kf, Kf,oc) or standard deviation of the mean (1/n). 13098
dx.doi.org/10.1021/jf203014g |J. Agric. Food Chem. 2011, 59, 13096–13101
Journal of Agricultural and Food Chemistry
ARTICLE
The Kf values of Brazilian soils ranged from 4.66 (BRA3) to 29.3 μmol(1 1/n) L1/n kg 1 (BRA7) (Table 2). For these soils, Kf was positively correlated (P < 0.05) with soil properties %OC (r = 0.67**) and %clay (r = 0.69**). However, when soil BRA8 was excluded, correlations improved for %OC (r = 0.99**) and clay
Figure 3. Freundlich isotherm fit to describe indaziflam sorption (solid line) and desorption (dashed line) at concentrations of 0.99 and 0.033 μmol L 1 for soils BRA3, BRA7, USA2 and, USA3. Bars associated with each symbol represent the standard deviation of each mean value.
contents (r = 0.88**). For this case, linear equations that describe Kf as a function of carbon content (Kf = 0.2773 + 0.0633%OC, R2 = 0.99**) and as a function of clay content (Kf = 1.8054 + 2.3532%clay, R2 = 0.78**) were calculated. Normalizing Kf for % OC reduced the variability of the sorption coefficients. The range of variation in Koc and Kf,oc was 7 variation for Kf. Some variability in Koc and Kf,oc is expected due to the variability in organic matter and the functional groups of indaziflam that preclude strictly hydrophobic behavior.13 Whereas use of Koc and Kf,oc reduced variability for sorption among soils and should be used for modeling purposes, contributions from clays cannot be totally be excluded. For soils with a clay mineral fraction (fcm) to organic C fraction (foc) > 25 (which is the case for three of the six soils, see Table 3), then clay mineral fractions can be significant in sorption.13 The cause of BRA8 sorption not fitting the correlation with OC and clay contents is not known. It is interesting to note that although the physicochemical characteristics of BRA8 and BRA7 are very similar, BRA8 contains small amounts of vermiculite with Al-hydroxy interlayer, which can contribute to increased sorption in this soil. Further studies should be conducted to elucidate the role of mineral fraction on indaziflam sorption. Although indaziflam is a weak acid (pKa = 3.5), there was no correlation between sorption Kf and pH. The lack of correlation may be due to the narrow range in pH among soils in this study, which ranged from 5.4 to 6.5, pH levels at which indaziflam is substantially ionized (>99% anionic). Oliveira et al.14 also
Table 3. Distribution Coefficients (Kd), Sorption Coefficients (Koc), Percentage of Applied Indaziflam Sorbed to Soil, and Desorption Kd for Selected Soils desorption3Kdb (L kg 1)
initial concn (μmol L 1)
sorption Kd (L kg 1)
BRA1
0.99
21.14 ( 0.41c
1321
90
39.19 ( 2.96
0.03
24.71 ( 0.6
1544
91
46.48 ( 0.59
BRA3
BRA7
BRA8
BRA11
BRA12
USA1
USA2
USA3 a
Koc (L kg 1) a
soil
sorbed (%)
0.99
4.86 ( 0.31
972
66
7.04 ( 0.37
0.03
5.66 ( 0.19
1132
70
8.55 ( 0.26
0.99
27.44 ( 2.57
1339
91
38.95 ( 2.53
0.03
25.6 ( 1.33
1249
91
39.14 ( 2.44
0.99 0.03
9.42 ( 0.53 10.29 ( 0.54
434 474
80 79
15.8 ( 0.41 16.78 ( 0.99
0.99
12.44 ( 0.91
1173
82
16.18 ( 0.76
0.03
12.82 ( 0.57
1209
83
16.75 ( 0.57
0.99
5.22 ( 0.38
855
66
8.18 ( 0.21
0.03
5.5 ( 0.19
902
68
9.3 ( 0.25
0.99
12.96 ( 0.16
595
84
19.6 ( 0.25
0.03
15.74 ( 0.53
722
86
25.82 ( 2.18
0.99
7.17 ( 0.39
652
75
9.98 ( 047
0.03
8.60 ( 0.48
782
78
13.41 ( 0.21
0.99
16.72 ( 0.18
664
87
23.52 ( 2.72
0.03
22.70 ( 1.1
901
90
32.02 ( 0.18
Koc = Kd/(%OC) 100. b Desorption Kd calculated after the third desorption step. c Mean Kd value ( standard deviation of the mean. 13099
dx.doi.org/10.1021/jf203014g |J. Agric. Food Chem. 2011, 59, 13096–13101
Journal of Agricultural and Food Chemistry reported no correlation between pH and sorption of other weak acid herbicides to selected Brazilian soils. It has been suggested that organic matter of variably charge soils with high clay contents becomes more accessible for herbicide sorption at high pH levels due to conformational changes,15 which in turn may explain some of the differences observed in sorption coefficient values. Despite being primarily anionic at the pH of the soils, indaziflam was substantially sorbed, possibly due to the triazinediamine group. Similar effects have been observed for sulfonylaminocarbonyltriazolinone herbicides, which have a pKa ∼ 2. The greater than expected sorption was presumably due to the soil particle surfaces and the heterocyclic rings.16 For the mollisols, Kf values ranged from 6.62 (USA2) to 14.3 μmol(1 1/n) L1/n kg 1 (USA3) (Table 2; Figure 3). Significant linear correlations (P < 0.05) indicated that soil %OC (r = 0.97**) was positively correlated with Freundlich sorption coefficients, whereas pH (r = 0.87**) and clay contents (r = 0.998**) were negatively correlated to indaziflam Kf. Linear equations that described Kf as a function of soil pH (Kf = 10.5429 0.2908pH, R2 = 0.763**), clay contents (Kf = 30.0341 1.0429%clay, R2 = 0.996**), and %OC (Kf = 0.0626 + 0.1887%OC, R2 = 0.60**) were significant. Normalizing Kf for %OC reduced the variability of sorption, and Kf,oc values for mollisols ranged from 498 (USA1) to 602 (USA2). The negative correlation between soil pH and Kf indicated that the extent of sorption increases as the soil solution pH decreases. Sorption of another weak acid (metsulfuron-methyl, pKa = 3.3), such as indaziflam, was also significantly decreased when the equilibrium pH was increased to pH 6.0 6.1.17 However, any conclusions should be tempered by the fact that there were only three soils, two of which had pH levels ∼8.1. Lower Kf,oc values for mollisols and soil BRA8 compared with Kf,oc for other oxisols also suggest the pH effect on the sorption, because the OC contents for USA1, USA3, and BRA8 are higher than other oxisol OC contents. According to EPA criteria, the range of Kf,oc found in the present work would rank indaziflam into low (500 2000) to medium (150 500) mobility classes.18 Another reference to rank indaziflam sorption to soil is given by IBAMA (Brazilian Environmental Agency);19 under this classification, a Kf range from 0 to 24 would rank the pesticide as low sorption, and 25 < Kf < 49 would rank it as moderately sorbed. Kf values found for indaziflam are within this range of values (Table 2). Desorption. Desorption was hysteretic from oxisols, 1/ndesorption < 1/nsorption (Table 2). Figure 3 shows indaziflam desorption isotherms in soils BRA3 and BRA7 at highest and lowest concentrations in this study. Hysteresis coefficients, H, ranged from 0.04 to 0.26 (at 0.99 μmol L 1) and from 0.03 to 0.25 (at 0.03 μmol L 1). H was negatively correlated (P < 0.05) with Kf values (r = 0.71**) and soil properties %OC (r = 0.78**) and %clay (r = 0.92**). Linear equations that described H as a function of soil Kf (H = 24.4446 68.2196Kf, R2 = 0.50**), clay contents (H = 77.6479 253.4219%clay, R2 = 0.85**), and %OC (H = 2.2657 5.6609%OC, R2 = 0.61**) were significant. Hysteresis suggests that a portion of applied indaziflam is very strongly/irreversibly bound to soil and does not readily desorb. Desorption hysteresis has been reported to be caused by a variety of mechanisms for a large number of herbicide soil systems.20 Even sorbed weak acid herbicides have exhibited no desorption from soil.16 As with the soils from Brazil, mollisols were also hysteretic. For these soils H ranged from 0.14 to 0.29 (at 0.99 μmol L 1) and from 0.10 to 0.19 (at 0.03 μmol L 1) (Table 2). Indaziflam
ARTICLE
desorption isotherms in soils USA2 and USA3 at highest and lowest concentrations in this study are shown in Figure 3. No correlations were found between H and Kf or pH for these soils. A negative correlation was found between H and %OC (r = 0.72**), described by the linear equation (H = 3.1577 7.1324%OC, R2 = 0.51**). Potential Transport. The leachability index model Groundwater Ubiquity Score (GUS)21 was used to determine indaziflam’s potential offsite mobility, that is, to leach to groundwater. According to this model, a chemical would be classified as a “leacher” if GUS > 2.8, as a “nonleacher” if GUS < 1.8, and as “transitional” if 1.8 < GUS < 2.8. In oxisols, assuming a soil degradation half-life (t1/2) = 150 days,2 indaziflam would be considered to be a leacher (GUS = 3.0) for the lowest Kf,oc (415) soil and transitional (GUS = 1.84) for the highest Kf,oc (1428) soil. For the Brazilian soils, indaziflam would be considered to be a leacher only for the lowest Kf,oc when t1/2 > 106 days, whereas for the highest Kf,oc, t1/2 would have to be >2053 days. In mollisols, assuming a t1/2 = 150 days,2 GUS ranged from 2.83 to 2.66 for the lowest (498) and highest Kf,oc (602), so indaziflam would range from leacher to transitional. However, for these soils, considering Kf,oc ranged from 498 to 602, this herbicide would be considered to be a leacher when t1/2 > 141 days for the lowest Kf,oc soil or when t1/2 > 196 days for the highest Kf,oc soil. Pesticides with t1/2 > 60 days are classified as highly persistent.22 Although in tropical regions such as Brazil, the indaziflam half-life is probably
