Envlron. Sci. Technol. 1993, 27, 500-505
(11) Russel, J. D.; Cruz, M.; White, J. L.; Bailey, G. W.; Payne, W. R.; Pope, J. D.; Teasley, J. I. Science 1968, 160, 1340-1342. (12) Terce, M.; Calvet, R. Z. Pflanzenkrankh. Pflanzenschutz, Sonderheft 1977,8, 237-243. (13) Talbert, R. E.; Fletchall, 0. H. Weeds 1964, 13, 46-52. (14) Borggard, 0. K.; Streibig, J. C. Acta Agric. Scand. 1988, 38, 293-301. (15) Abraham, M. H.; Grellier, P. L.; Prior, D. V.; Duce, P. P.; Morris, J. J.; Taylor, P. J. J. Chem. Soc., Perkin Trans. 2 1989,699-711. (16) Abraham, M. H.; Grellier, P. L.; Prior, D. V.; Morris, J. J.; Taylor, P. J. J. Chem. Soc., Perkin Trans.2 1990,521-529. (17) Nadler, E. B.; Rappoport, Z. J. Am. Chem. SOC.1989, I l l , 213-223. (18) Nagano, M.; Nagano, N. 1.;Higuchi, T. J. Phys. Chem. 1967, 71, 3954-3959. (19) Gurka, D.; Taft, R. W. J. Am. Chem. SOC.1969, 91, 4794-4801. (20) Slasinsky,F. M.; Tustin, J. M.; Sweeney, F. J.; Armstrong, A. M.; Ahmed, Q. A.; Lorand, J. P. J. Org. Chem. 1976,41, 2693-2699. (21) Lorand, J. P.; Nelson, J. P.; Gilman, R. D.; Staley, K. L.;
Chambers, J. R.; Kirk, H. D.; Moeggenborg, K. J.; Farlow, D. L. J. Phys. Org. Chem. 1990,3,659-669. (22) Granot, J. J. Magn. Reson. 1983, 55, 216-224. (23) Taft, R. W.; Klingensmith,G. B.; Ehrenson, S. J.Am. Chem. SOC.1965,87, 3620-3626. (24) Slejko, F. L.; Drago, R. S. Inorg. Chem. 1973,12,176-183. (25) Jackel, H.; Stamm, H. Arch. Pharm. (Weinheim,Ger.) 1988, 321, 213-219. (26) Chudek, J. A.; Foster, R.; Mackay, R. L.; Page, F. M.;
Twiselton, D. R. J. Chem. SOC.,Faraday Trans. 1 1988,84, 1145-1152. (27) Huyskens, P. L. J. Am. Chem. Soc. 1977,99,2578-2582. (28) Frange, B.; Abboud, J.-L. M.; Benamou, C.; Bellon, L. J. Org. Chem. 1982,47,4553-4557. (29) Deranleau, D. A. J . Am. Chem. SOC.1969,91,4044-4049. (30) Bailey, W. F.; Monahan, A. S. J. Chem. Educ. 1978, 55, 489-493. (31) Abraham, M. H.; Grellier, P. L.; Prior, D. V.; Taft, R. W.;
Morris, J. J.; Taylor, P. J., Laurence, C.; Berthelot, M.; Doherty, R. M.; Kamlet, M. J.; Abboud, J.-L. M.; Sraidi, K.; Guiheneuf, G. J. Am. Chem. Soc. 1988,110,8534-8536. (32) Levy, G. C.; Lichter, R. L. Nitrogen-15 Nuclear Magnetic Resonance Spectroscopy;Wiley: New York, 1981;Chapter 4. (33) Wilson, M. A.; Vassallo, A. M.; Perdue, E. M.; Reuter, J. H. Anal. Chem. 1987.59. 551-558. (34) Schnitzer, M.; Preston, C. M. Soil Sei. SOC.Am. J. 1987, 51. 639-646. (35) Moelwyn-Hughs,E. A. The Chemical Statics and Kinetics of Solutions; Academic Press; New York, 1971; p 35. (36) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. Rev. 1988, 88, 899-926.
Received for review June 1,1992. Revised manuscript received November 4,1992. Accepted November 10,1992. We thank the Ceiba-Giegy Corp. for their financial support and for samples of atrazine. We gratefully acknowledge financial support from the Federal Hatch Project 3289, the Wisconsin Graduate School Project 920284, and the NRI Competitive Grants ProgramlUSDA Grant 91-37012-6795.
Cooperative Hydrogen Bonding of Atrazine Gereon J. Welhouse and Willlam F. Bleam"
Department of Soil Science, University of Wlsconsln-Madison, Madison, Wisconsin 53706 Formation constants are determined for hydrogen-bond complexes between atrazine and compounds commonly found in soil organic matter. Weak to moderately strong complexes (Kf = 2-30 L/mol) are formed with amine, hydroxyl, and carbonyl functional groups. Strong complexation is observed with the carboxylic acid (Kf = 212 L/mol) and amide (Kf = 276 L/mol) functional groups, which interact cooperatively with atrazine by simultaneously donating and accepting a hydrogen bond. These results confirm that hydrogen bonding provides a mechanism for atrazine adsorption to soil surfaces and identifies functional groups that have high affinity for atrazine.
C1
H
I
R3
Atrazine
1
Introduction We recently reported that atrazine, 1, forms complexes with hydrogen bond donors and acceptors in CCll solvent (I). The formation constants for hydrogen-bond complexes range from weak to moderate (Kf = 2-30 L/mol) for these compounds, which interact with atrazine by forming a single donor-acceptor hydrogen bond. The compounds used in our previous study are reference compounds that establish universal hydrogen-bond activity (2, 3). In this paper we report formation constants for complexation between atrazine and functional groups found in soil organic matter. Two of these compounds, pyrrolidinone (NHP) and acetic acid (HOAC), have the potential to form strong cooperative hydrogen bonds with atrazine (4). Atrazine hydrogen-bond activity is calibrated with the a and 0 parameters than rank atrazine reactivity on 500 Envlron. Scl. Technol., Voi. 27, No. 3, 1993
R3
Isomer - R1 2
3 4
H H Et Et
Et Et H H
114
H
iPr
iPr
H iPr H
H
iPr
Et = ethyl; iPr = isopropyl
universal hydrogen-bonding scales ( 2 , 3 ) . Atrazine's parameters (a = 0.42; p = 0.50) demonstrate that it can donate and accept hydrogen bonds. We now test the predictive power of the a and @ parameters by estimating complexation constants from the a and p parameters of atrazine and the compounds used in this study and comparing these with the experimental results. As in our previous study, we measure the change in NH chemical shifts of atrazine as it is titrated with a complexing agent. The resulting titration curve data are fit to a 1:l complexation model using linear regression to obtain the formation constants.
0013-936X/93/0927-0500$04.00/0
0 1993 American Chemlcal Society
Experimental Methods The titration procedure and linear regression procedure used to obtain the formation constants were described previously ( I ) . Atrazine chemical shifts at different concentrations of complexing agents are observed by NMR spectroscopy and fit to a 1:l complexation model using an iterative linear regression procedure (5). The NH-i proton on the isopropylamino side chain of isomer 1 is used as the observed signal, and formation constants calculated from this site are assumed to represent overall atrazine reactivity. This proton signal is used because its chemical shift changes in response to complexation and is the signal least affected by atrazine dimerization (6). Dimerization interferes with the ethylamino NH proton signals of isomers 1 and 2 so much that their NMR chemical shifts could not be used to find formation constants by linear regression. Two of the complexing agents used in this study (NHP and HOAC) are known to dimerize to a significant extent in organic solvents. Dimerization reduces the solution concentration of these two compounds and affects the formation constants for complexation with atrazine. The regression equation was therefore corrected by including dimerization constants in the mass balance relations used for the original derivation. The dimerization constant for NHP was found using an iterative linear regression procedure (7). Our estimate, 380 L/mol, is larger than the 180 L/mol reported in the literature (8). We were unable to measure the formation constant acetic acid from the NMR chemical shifts (perhaps because of interference from proton exchange), so the literature value of 80 L/mol for butyric acid was used (9). We attempted to correct for dimerization of both atrazine and the complexing agent. This “corrected”regression equation either did not converge or gave unrealistic chemical shifts and formation constants when it did converge. Therefore, formation constants reported here were obtained under the assumption that dimerization does not interfere with complexation and may be lower than the true formation constants.
.0056
.0025
h_r
.0017
AI, 3-e 2;i
I
I
0.0
I 7.5
5.5
65
45
PPM Flgure 1. Chemical shifts of abazlne NH protons (In ppm from Internal TMS) at dlfferent concentrations of NHP complexing agent (in mollL). All spectra were recorded at 298 K in CCL4 solvent and at a constant atrazine concentration of 0.007 mollL. Spikes at 6.5 and 7.3 ppm are from solvent impurities. Peaks are labeled accordlng to the isomer (1, 2, or 3) and side chain (I = isopropylamlno,e = ethylamino) of the NH proton responsible for the slgnal (6). 3200 3000
Results Figure 1shows the change in the atrazine NH chemical shifts in response to added complexing agent NHP. There are eight peaks in this partial spectrum, one for each atrazine NH proton in each of the four isomers. The signal labeled 1-i is the isopropylamino NH proton of isomer 1 and is the signal used in linear regression to find the formation constants. The change in chemical shift is more or less uniform for all signals except those assigned to NH-i of isomer 2 ( 2 4 , NH-e of isomer 3 (3-e), and the signals for isomer 4. The lower response of these signals to the addition of complexing agent suggests that they are less reactive than the NH-i signal used for regression. However, these signals together make up only -25% of the total atrazine (based on the relative peak areas) and do not significantly affect the overall activity calculated with NH-i of isomer 1. The chemical shifts of NH-i are plotted against the concentration of NHP in Figure 2. The solid curve represents chemical shifts estimated by linear regression and shows strong complexation of atrazine, with 76% existing as the complexed species at an NHP concentration of 0.022 mol/L (dashed line). The data points sample over 90% of the curve, well above the 76% needed to “prove” the complexation model (IO). The formation constants and chemical shifts for atrazine hydrogen-bond complexes are presented in Table I.
2800
chem. shift
2600 2400
2200 2000
I
I
I
I
-3
-2
-1
0
log PWl“ Flgure 2. Plot of atrazlne NH-i chemical shifts (in Hz relative to TMS) vs the concentration of NHP (In moilL). The concentration of this complexing agent ranges from 0.0004 to 0.22 moi/L. The curve is calculated from the formation constant and NKI chemlcai shift of the complex found by llnear regression of the data points. The dashed line lndlcates the point at which 76% of atrazine Is compiexed.
Moderate complexation (Kf between 2 and 20 L/mol) is observed with all the compounds except NHP and HOAC, which display strong complexation (Kf > 200 L/mol). In general, the trends in the atrazine NH chemical shifts calculated for the complexes parallel the formation conEnviron. Sci. Technol., Vol. 27, No. 3, 1993 501
Table I. Formation Constants and Proton Chemical Shifts of Atrazine Hydrogen-Bond Complexes compd pyrrolidine cyclopentanone pyrrole N-methylpyrrolidone pyrrolidinone acetone ethanol phenol methyl acetate acetic acid
abbr
Kf"sd
dcbsd
CNH 11 (2.6) 2560 (108) CP 3.5 (0.8) 2331 (46) PYRR 2.6 (2.5) 2321 (158) NMP 14 (4.7) 2535 (158) NHP 276 (27) 3134 (45) ACE 2.3 (1.4) 2390 (190) EtOH 3.3 (2.4) 2757 (140) PHEN 16 (4.4) 2252 (49) MEAC 4.8 (0.8) 2248 (66) HOAC 212 (36) 3198 (60)
CIA"' 0.51 0.52 0.40 0.77 0.91 0.67 0.50 0.80 0.70 0.90
Formation constants (in L mol) obtained by linear regression of NMR chemical shift data. 6Chemical shift of the observed NH proton in the hydrogen-bond complex (in Hz relative to TMS). 'Portion of the titration curve sampled by the data. Sampling less than 0.76 introduces additional uncertainty to the regression estimates of Kf and dC. dNumbers in parentheses are the 95% confidence intervals.
stants; the largest chemical shifts are observed with the strongest complexing agents. In a previous paper (I), we determined the universal hydrogen-bond a and p parameters of atrazine. The predictive power of these parameters is tested on the compounds in Table I by calculating expected formation constants with eq 1 (11)and comparing these with those log K = 7.354aHPH- 1.094 (1) observed for each compound. Equation 1 is used to calculate formation constants for complexes of atrazine with the monofunctional compounds pyrrolidine (CNH), cyclopentanone (CP), pyrrole (PYRR), N-methylpyrrolidinone (NMP), and acetone (ACE). The results are presented in Table 11. The compounds CNH, CP, NMP, PYRR, and ACE have only one a or p parameter to be matched up against the complimentary P or a parameter for atrazine. Ethanol (EtOH) and phenol (PHEN), however, can act as hydrogen-bond donors and acceptors and have both a and @ parameters. They are able to form two distinct hydrogen-bond complexes. For example, EtOH ( a = 0.33) can donate a hydrogen bond to atrazine and form a complex. Conversely, EtOH can accept (@= 0.44) a hydrogen bond donated by atrazine to form a different complex. For compounds that have this dual nature, formation constants are calculated by applying eq 1 separately to two sets of complimentary parameters and the overall formation constant is the sum of the two formation constants (log K , + log KBa)for two separate complexes. +he formation constant predicted this way for methyl acetate (MEAC) is only 1/3 the observed value (see Table 11). This may be due to the fact that only one hydrogenbonding parameter is reported for MEAC, yet it has two distinct functional groups (carbonyl and ether) that can accept hydrogen bonds. [Surprisingly, the one parameter assigned to MEAC (0= 0.40) is less than that for the carbonyl of acetone (0.52).] Including parameters for both the carbonyl (P = 0.40) and ether (p = 0.50) functional groups and applying eq 1 for these two sets of complimentary parameters gives a predicted value of 4.2 L/mol, much closer to the observed value of 4.8 L/mol. It is possible that two B values are needed to adequately describe the hydrogen-bonding potential of MEAC. Equation 1 applied to NHP and HOAC gives predicted formation constants of 13.5 and 9.3 L/mol, respectively, much lower than the observed formation constants, 276 and 212 L/mol. It appears the donating and accepting 502 Envlron. Scl. Technol., Vol. 27, No. 3, 1993
Table 11. Comparison between Predicted and Observed Formation Constants for Atrazine Hydrogen Bond Complexes K, compd" CNH CP PYRRe NMP' NHP ACE" EtOH PHEN' MEAC HOAC atrazine water
ab
Pb
predC
obsd
0 0 0.41 0 0.38
0.71 0.50 0 0.77 0.72
13 2.8 2.4 19 306
11 3.5 2.3 14 276
0 0.33 0.60 0 0.55
0.50 0.44 0.22 0.40 0.42
2.8 3.1 12 1.4 153
2.3 3.3 16 4.8 212
0.42 0.35
0.50 0.38
98r 2.8
671
"Abbreviations are taken from Table I. b~ and @ hydrogenbonding parameters taken from the literature (2, 3). Predicted formation constants (in L/mol) calculated with eq 1 or 2 and the complimentary OL and 0 parameters of atrazine and each complexing agent. dObserved formation constants from Table I. e Indicates a compound that was used to obtain the atrazine a and P parameters. f Predicted and observed (6)dimerization constants for atrazine in CC14
functional groups of NHP and HOAC do not act independently to form two distinct complexes (as did EtOH and PHEN) but instead form a cooperative arrangement in which the donor and acceptor interactions reinforce each other in the sample complex. Equation 1 is modified to allow cooperativity by including both donor and acceptor interactions in one complex, as described in eq 2. Formation constants calculated log K = 7.354(a& + @AaB) - 1.094 (2) with eq 2 are 306 for NHP and 153 L/mol for HOAC, much closer to the observed quantities.
Discussion These results indicate that cooperative interactions greatly enhance the hydrogen-bonding potential of atrazine. A good illustration of this cooperative enhancement is provided by comparing the results for NMP with NHP. Replacing the NCH, group with NH adds hydrogen-bond donating potential to the existing accepting potential and increases formation constants from 14 to 276 L/mol. Similarly, replacing OCH, of MEAC with OH in HOAC increases activity from 5 to 212 L/mol. The cooperative interaction of NHP and HOAC with atrazine is pictured in Figures 3 and 4. These interactions are the same as those used to describe hydrogen-bond complexation of adenine with the imide functional group, which has Kf = 220 L/mol(I2), and with the carboxylic acid functional group, which has Kf = 160 L/mol (9). A similar cooperative arrangement has been used to describe complexation of 2-aminopyridineby acetic acid (13). The arrangement of ring nitrogen and side-chain NH functional groups in these molecules resembles that present in atrazine. These results also demonstrate that some compounds engage in cooperative interactions while others do not, even though they have both hydrogen-bond donating and accepting parameters. PHEN and EtOH have a and parameters about as large as HOAC and NHP (Table 11),yet their formation constants for complexation with atrazine are much smaller. This is surprising considering the reports of cooperative hydrogen bonding in the self-association of PHEN (14) and EtOH (15, 16).
H
H
H
I
0
H I
wo
+ O w o
r -
I
I
CH3
CNH
CP
PYRR
NMP
11
3.5
2.6
14
o0 B
HAC
CH3
C1
NHP
bo
280
Atrazine
Flgure 3. Formatlon constants (In L/moi) for complexatlon between atrazlne and the NHP serles of compounds. The strong complexation observed between atrazlne and NHP Is attributed to the cooperative lnteractlon shown. Both molecules donate and accept hydrogen bonds because of favorable orbital overlap and resonance stabillration. H I
/o CH2
I
CH3
ETOH 3.3
PHEN 16
H I
ACE
MEAC
2.3
4.8
CI
I
O CH3 Y O
HAC 210 CH3
Flgure 4. Formation constants (In Llmol) for complexation between atrazine and the M A C series of compoonds. The strong complexation observed between atrazine and HOAC Is attributed to the cooperative Interaction shown. Both molecules donate and accept a hydrogen bonds because of favorable orbital overlap and resonance stablllration.
One explanation of this behavior is provided by the nonbonded orbital (NBO) representation of hydrogen bonding, which emphasizes the role of molecular geometry and orbital overlap (4). Atrazine interacts strongly with NHP and HOAC because there is favorable geometry for orbital overlap. PHEN and EtOH do not have this favorable geometry for cooperative interaction with atrazine. There are at least two factors that determine the strength of the cooperative interaction: (1) the geometry
Figure 5. Resonance structures for atrazlne, NHP, and HOAC. Atomic orbitals which contain "lone pair" electrons adjacent to the aromatic triazine ring or the double bond of a carbonyl group rehybrldlze into sp2 orbitals. Electron density Is then easily redlstrlbuted through the resuitlng *-electron system.
of overlap between HBD and HI3A orbitals in the complex and (2) the ability to redistribute electronic charge. Favorable overlap geometry can be estimated visually from molecular models constructed to scale, and the distribution of electronic charge can be estimated by resonance structure descriptions of the molecules (6). The resonance structures for NHP, HOAC, and atrazine are shown in Figure 5. The separation of charge implied by these structures reinforces the dipolar contribution to hydrogen bonding. The NBO description of hydrogen bonding can be combined with the (Y and b parameters to examine atrazine interactions in the aqueous soil environment. For example, the hydrogen-bonding interaction with water can be estimated by assuming that the water hydroxyl has the same geometry as the hydroxyl in EtOH or PHEN. The result is poor orbital overlap and a small formation constant for an atrazine-water complex of 2.8 L/mol calculated with eq 1. It is reasonable to expect that water does not compete well against atrazine for cooperative functional groups in soil organic matter, especially if these interactions take place in hydrophobic domains which limit water access. Many functional groups capable of cooperative interaction with atrazine has been identified in soil organic matter, including carboxylic acids, amides, and quinones (17,18).Atrazine can interact with these functional groups and be adsorbed as the neutral species; proton transfer from carboxylic acids is not required. This is consistent with adsorption of atrazine attributed to interactions with carboxylic acids (1S22). The hydrogen-bonding parameters reported in this procedure apply to CCl, solutions at 298 K. They are valid for hydrogen-bonding interactions that occur in the interlayer of organoclay adjuvants (23)and in the hydrophobic domains in soil organic matter to the extent that these domains resemble an organic solvent with low (