HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (1) Citing Articles via Google Scholar Google Scholar Articles by Mansell, R. S. Articles by Downs, W. C. Search for Related Content PubMed PubMed Citation Articles by Mansell, R. S. Articles by Downs, W. C. GeoRef GeoRef Citation Agricola Articles by Mansell, R. S. Articles by Downs, W. C. Related Collections Water Quality Water Pollution Ground Water Quality

TECHNICAL REPORT
Ground Water Quality

A Transport Model with Coupled Ternary Exchange and Chemisorption Retention for Hydrazinium Cations

^a Soil & Water Science, P.O. Box 110290, 2169 McCarty Hall, Univ. of Florida, Gainesville, FL 32611-0290
^b Civil and Environmental Engineering, Brigham Young Univ., 368 Clyde Building, Provo, UT 84602

* Corresponding author (rsm{at}gnv.ifas.ufl.edu)

Received for publication October 27, 2000.

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS MATHEMATICAL MODEL RESULTS AND DISCUSSION CONCLUSION REFERENCES

 
A numerical model was developed to describe the fate and transport of hydrazinium and competing Ca²⁺ and H⁺ cations applied in acidic solutions to columns of Ca²⁺/H⁺–saturated sandy soil during steady saturated flow conditions. Instantaneous ternary H⁺–Ca²⁺–N₂H⁺₅ cation exchange using the Gaines–Thomas approach was combined with second-order, irreversible, kinetic chemisorption of exchange-phase N₂H⁺₅ ions as major retention mechanisms for N₂H⁺₅. Exchange-mediated chemisorption is assumed to occur as chemical binding of N₂H⁺₅ ions located on carboxyl-group exchange sites to nearby carbonyl groups, consequently decreasing the effective soil cation exchange capacity (CEC). Comparison of simulated and observed breakthrough curves (BTCs) for concentrations of N₂H⁺₅ and Ca²⁺ ions in column effluent was used in model evaluation. The cation transport model with cation exchange coupled with exchange-mediated chemisorption provided a valid first approximation for N₂H⁺₅ transport.

Abbreviations: BTC, breakthrough curve • CEC, cation exchange capacity

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS MATHEMATICAL MODEL RESULTS AND DISCUSSION CONCLUSION REFERENCES

 
THE basic, difunctional, polar hydrazine molecule (N₂H₄) is a powerful reducing agent (Schmidt, 1984). Hydrazine and its derivatives are used widely in the synthesis of various pesticides, as rocket fuel, as an antioxidant, and as an oxygen scavenger (Schmidt, 1984). Extensive commercial usage of this toxic compound combined with high solubility in water implicate hydrazine as a potential surface and ground water contaminant. For example, in 1991 a train accident in California resulted in the spillage of about 1500 L (400 gallons) of liquid hydrazine onto the landscape (Reed, 1991). Hydrazine is a toxic environmental carcinogen suspected of being a human carcinogen (Braun and Zirroli, 1983).

For acid aqueous conditions (pH < pK_a = 7.96), hydrazine exists primarily as the protonated hydrazinium cation N₂H⁺₅. At pH 4 approximately 99.99% of hydrazine molecules exist as nonmetal, inorganic N₂H⁺₅ cations. Acidic N₂H⁺₅ solutions have been shown to be stable (Cuy and Bray, 1924), in contrast to alkaline conditions, where N₂H₄ undergoes abiotic autooxidation (Ellis et al., 1960).

Transport of both N₂H⁺₅ cations and N₂H₄ molecules during water flow in soils is greatly dependent upon removal from the aqueous phase during interaction with organic and mineral components. Sorption of both N₂H⁺₅ and N₂H₄ is recognized to be strongly influenced by humic substances and to a lesser extent by clay minerals (Hayes et al., 1988). Because of high polarity, N₂H₄ molecules are miscible with polar solvents such as water and immiscible in nonpolar solvents such as hydrocarbons. Hydrazinium cations have been proposed to undergo multiprocess sorption (retention) due to: (i) reversible ion exchange (physical sorption) with native ions present on the soil exchange phase (Isaacson and Hayes, 1984); (ii) partially reversible complexation (nonspecific sorption) of N₂H⁺₅ ions with cations on exchange sites associated with reactive soil constituents (Isaacson and Hayes, 1984; Griffith et al., 1980); (iii) irreversible condensation (i.e., chemisorption) of N₂H⁺₅ ions with carbonyl groups of humic components (Schnitzer and Skinner, 1965; Isaacson and Hayes, 1984); and (iv) microbiological degradation of N₂H⁺₅ ions by soil microorganisms (Ou, 1987; Moliner and Street, 1989a). Although these mechanisms may be difficult to distinguish experimentally, each limits hydrazinium mobility by lowering the concentration in the mobile soil–solution phase. Chemisorption and biodegradation result in irreversible removal of N₂H⁺₅ ions from the mobile soil solution and thus provide potential means for minimizing transport during water flow through the vadose zone. Complexation may be partially irreversible and thus also may be an important retention mechanism. Griffith et al. (1980) showed that hydrazine can be retained through complexation with cations on exchange sites.

Cation exchange was demonstrated by Isaacson and Hayes (1984) to be a major mechanism for retention of monovalent N₂H⁺₅ ions by pH-4 aqueous suspensions of H⁺– and Ca²⁺–saturated humic acids during continuous flow in reaction cells. At pH 4 hydrazinium ions were reported to exchange more readily with H than Ca ions on humic acids (water insoluble at this acidity level). In addition to simple cation exchange, N₂H⁺₅ adsorption occurred by condensation reaction (chemisorption) with humate carbonyl groups, as well as nonspecific sorption involving both weak and strong bonding. Maximum condensation reaction rates for hydrazine and, to a lesser extent, N₂H⁺₅, occur for solution pH values close to the pK_a. As pH decreases these rates decrease but the relative contribution of ion exchange tends to increase. On average, 58% of the N₂H⁺₅ ions sorbed by humic acids was determined to be reversibly and 42% irreversibly sorbed, respectively, as determined by repeated washings with concentrated salt and acid solutions (Isaacson and Hayes, 1984).

Moliner and Street (1989b) reported hydrazine adsorption in pH-4 aqueous suspensions of Na-saturated kaolinite clay and Arredondo fine sand (loamy, siliceous, semiactive, hyperthermic Grossarenic Paleudult; Ap and E2 horizons) to depend upon surface functional groups as well as suspension pH. Although exchange was a primary retention mechanism for N₂H⁺₅ cations in kaolinite suspensions, N₂H⁺₅ retention in the soil suspensions involved the combined effect of ion exchange and specific sorption. Both mechanisms in the soil were predominantly associated with organic matter components. Sorbed N₂H⁺₅ in the Arredondo soils was totally irreversible as determined by KCl salt flushing, in contrast to 10% in the clay. Only 5 and 8% of sorbed hydrazinium were released by further extraction of Ap and E2 soils with HCl acid, respectively. Total retention was greater in the surface Ap soil, which had a higher organic matter content than E2. Hydrazinium adsorption was more strongly influenced by humic substances than by clay minerals, although the clay mineral (kaolinitic) contents exceeded the organic matter contents in both Arredondo soil horizons. A charge fraction of exchange-phase N₂H⁺₅ ions in the Arredondo soils was hypothesized to undergo specific sorption (exchange-mediated chemisorption) by forming hydrogen bonds with neighboring carbonyl organic-surface functional groups.

Multiprocess retention behavior for cations in soil is not unique to inorganic, nonmetallic N₂H⁺₅ cations but is a common occurrence for inorganic heavy metals such as Pb (Selim and Amacher, 1997), as well as for organic cations such as paraquat herbicide (Burns et al., 1973), although specific mechanisms may vary greatly among ion species. Rapid reversible retention of heavy metals in soils is generally attributed to nonspecific ion exchange (instantaneous) and slower retention (kinetics) to high-affinity, specific sorption (inner-sphere complexation, surface precipitation, and penetration into clay mineral lattices) at soil surfaces.

A multicomponent, multiprocess mathematical model was developed here to describe the fate and transport of monovalent hydrazinium (N₂H⁺₅) cations applied in anoxic, acid solution to columns of Ca-saturated sandy soil during steady saturated flow conditions. The model used instantaneous ternary H⁺–Ca²⁺–N₂H⁺₅ cation exchange on organic matter as a major retention mechanism for N₂H⁺₅ combined with second-order, irreversible, kinetic chemisorption of a fraction of N₂H⁺₅ ions in the exchange phase. The exchange-mediated chemisorption process was envisioned to decrease the effective soil CEC by deactivation of exchange sites. Cation exchange was described using the Gaines–Thomas approach, and H⁺ ions were taken to be strongly competitive relative to N₂H⁺₅ and Ca²⁺ ions for exchange sites on soil organic matter. Retention of N₂H⁺₅ ions within the transport model was simply described by coupling chemisorption as a secondary process with a primary process of cation exchange.

	MATERIALS AND METHODS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS MATHEMATICAL MODEL RESULTS AND DISCUSSION CONCLUSION REFERENCES

 
Bulk samples of three horizons—Ap, E1, and E2—from a profile of Arredondo fine sand were collected near Gainesville, FL. Organic carbon and clay contents were greatest for the surface Ap horizon, less in the underlying E1 horizon, and least in the E2 horizon (Table 1). Particle-size analysis for mineral components was performed by the pipette method of Gee and Bauder (1986). The percentage of organic carbon was determined for each soil horizon by dry combustion in an induction furnace (LECO [St. Joseph, MI] Model 523-300) following the procedure of Nelson and Sommers (1982). Organic C content in the Ap horizon was 13.14 times that in the E2 horizon and 2.43 times that in the E1 horizon. Soil pH determined using 2 (mL):1 (g) suspensions in 0.01 M CaCl₂ was 4.46, 5.03, and 5.13 for the Ap, E1, and E2 horizons. Soil samples were air-dried, sieved through a 2-mm screen, and exchange sites were saturated with Ca using CaCl₂.

Column effluent was collected in glass test tubes using an automatic fraction collector. Effluent in odd-numbered tubes were analyzed for pH and Ca²⁺ concentrations. A glass-calomel electrode with a pH meter was used to determine pH. Effluent in even-numbered tubes was acidified with 1 M HCl and later analyzed for hydrazinium concentration.

A modified colorimetric method reported by Watt and Crisp (1952) was used for hydrazine analyses. Small aliquots of column effluent were placed into 25-mL volumetric flasks along with 15 mL of 4-dimethylaminobenzaldehyde (PDBA) solution. An intense orange color forms as hydrazine reacts with PDBA. The solution was diluted and stabilized by the addition of 1 M HCl to bring the column to 25 mL. Color intensity was read with a spectrophotometer as percentage transmission and then converted to absorbance. Stock solutions of hydrazine were used to prepare a standard solution of concentration versus absorbance. A linear standard curve was used to convert absorbances to hydrazine concentrations.

Breakthrough curves (BTCs) for Ca²⁺, N₂H⁺₅, and H⁺ providing concentrations (mol_c m^-3) in column effluent versus pore volumes of effluent were prepared.

	MATHEMATICAL MODEL

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS MATHEMATICAL MODEL RESULTS AND DISCUSSION CONCLUSION REFERENCES

 
A numerical model was developed to describe the transport of three cations—Ca²⁺, N₂H⁺₅, and H⁺—that undergo competitive instantaneous exchange under conditions of exchange-mediated chemisorption of N₂H⁺₅ cations, which resulted in a variable effective CEC for the soil. Exchange sites were assumed to be dominated by carboxyl groups associated with soil organic matter. Chemisorption was assumed in the model to occur as essentially irreversible condensation of exchange-phase N₂H⁺₅ ions with carbonyl groups of humic components of the soil (Schnitzer and Skinner, 1965; Isaacson and Hayes, 1984; Moliner and Street, 1989b).

An established approach for modeling convective–dispersive transport of multiple-species cations (Rubin and James, 1973; Valocchi et al., 1981; Mansell et al., 1993) that uses instantaneous exchange of metal cation species in soil with constant CEC was modified to describe instantaneous ternary H⁺–Ca²⁺–N₂H⁺₅ cation exchange with second-order, irreversible, kinetic chemisorption of N₂H⁺₅ ions as a major retention mechanisms for N₂H⁺₅. Chemisorbed hydrazinium cations in the exchange phase were assumed to be so tightly bound as to be essentially nonexchangeable and thus deactivate associated exchange sites. Chemisorption was thus assumed to decrease the effective CEC. The soil contained both organic matter and kaolinitic clay, but selectivity properties of exchange sites were assumed to depend predominantly upon soil organic matter. Although the general order of ion preference for exchange sites was assumed to be H⁺ > Ca²⁺ > N₂H⁺₅, the presence of chemisorption greatly enhances the relative preference of soil sorption sites for N₂H⁺₅.

A critical new element in the proposed transport model for ternary cations is the inclusion of exchange-mediated chemisorption of N₂H⁺₅ ions, which provides a quasicatalytic behavior. Cation exchange sites in the model developed here were assumed to be predominantly negatively charged COO^- on organic matter surfaces since the acid nature of humic substances is usually attributed to ionization of carboxyl (COOH) and phenoxy OH groups (Stevenson, 1982). Chemisorption of exchange-phase N₂H⁺₅ ions is assumed to be kinetic, irreversible, and limited to a fraction of exchange sites in close proximity to carbonyl groups. The irreversibility is further assumed to kinetically decrease the magnitude of the effective soil CEC (i.e., the number of active exchange sites). Resulting temporal and spatial decreases in S_Tef consequently alters competitive exchange, as well as transport of N₂H⁺₅, Ca²⁺, and H⁺ ions.

Total concentrations for hydrazinium, calcium, and protons in both aqueous and solid phases of the soil (mol_c m^-3 of bulk soil) at any depth and time were assumed to be:

[1]

where the subscript n can be k (hydrazinium), l (calcium), or m (hydrogen); C_n is the concentration (mol_c m^-3 of solution) of the indicated cation in solution phase, S_n is concentration (mol_c Mg^-1) of the indicated cation in the exchange phase; _n is an arbitrary constant that is unity for hydrazinium and zero for all other ions; S^'_k is the concentration (mol_c Mg^-1) of chemisorbed hydrazinium; t is time; z is soil depth; is volumetric water content (m³ m^-3); and is soil bulk density (Mg m^-3).

For modeling purposes, effective soil CEC_eff is defined here as a variable parameter with magnitude specified to be less than or equal to soil CEC (S_T):

[2]

Thus, S_Tef (mol_c Mg^-1) is subject to decrease with time at a rate equal to minus the rate of chemisorption of exchange phase N₂H⁺₅ according to:

[3]

Hydrazinium cations located on organic matter carboxyl-group exchange sites were assumed to undergo simplified Langmuir-type kinetic chemisorption (Selim and Amacher, 1997) for those sites located in close proximity to organic matter carbonyl groups according to:

[4]

where k_f is a forward kinetic rate coefficient, k_b (s^-1) is a backward rate coefficient, and S^'_k-max is the maximum potential concentration (mol_c Mg^-1) of chemisorbed hydrazinium on exchange sites. S^'_k-max = S_T can be expressed as the number of exchange sites capable of actively supporting chemisorption, where is a designated fraction of S_T. The forward reaction rate in Eq. [4] is proportional to the number of unoccupied chemisorption sites and to the concentration of hydrazinium cations in the exchange phase. The overall chemisorption reaction may be assumed to be either very slowly reversible if k_f >> k_b or to be simply irreversible if k_b = 0. The latter choice is assumed in all simulations reported here. In chromatography, chemisorption is often considered as a two-step process whereby a rapid initial step involving physical adsorption (a weak bond with <15 kcal mol^-1, such as cation exchange) is followed by a slow final step involving formation of a chemical bond with 20 to 100 kcal mol^-1 (Giddings, 1965). Only a fraction of physically adsorbed molecules may actually undergo chemisorption, and once a molecule is chemisorbed it is expected to either desorb slowly or not at all.

Competitive ternary cation exchange, chemisorption of exchange phase hydrazinium, and convective dispersive transport of three ternary cation species—N₂H⁺₅, Ca²⁺, and H⁺—where S_Tef decreased with chemisorption of exchange phase N₂H⁺₅, were described for conditions of steady saturated water flow using a system of three coupled, nonlinear, partial differential equations:

[5]

where n = k, l, or m depending on the cation in question; _n is described previously; D is the hydrodynamic dispersion coefficient (m² s^-1); and is the pore velocity (m s^-1). The first and second terms on the right-hand side (rhs) of Eq. [5] represent dispersion and convection transport processes, respectively. The fourth term on the rhs of Eq. [5] represents a sink term for chemisorption of hydrazinium as defined in Eq. [4].

The third terms on the rhs of transport Eq. [5] was generated using an approach developed by Rubin and James (1973) and Valocchi et al. (1981) based upon an assumption of constant S_T(z,t) and instantaneous exchange:

[6]

[7]

[8]

where r_k, r_l, r_m represent the valences of the hydrazinium, calcium, and hydrogen ion species, respectively. The last term on the rhs of Eq. [5] was developed here from first principles of mass balance to correct Eq. [6], [7], and [8], respectively, for temporal changes in S_Tef(z, t) resulting from exchange-mediated chemisorption of hydrazinium. Although chemisorption of hydrazinium decreased S_Tef, solution normality was unaltered by the model.

Equations [6], [7], and [8] assume a generalized Gaines–Thomas binary exchange coefficient K_ij for species i and j (Rubin and James, 1973; Valocchi et al., 1981) defined as

[9]

where parameters C_i and C_j represent concentrations of two arbitrary ion species i and j in the solution phase, S^*_i = S_i/S_T and S^*_j = S_j/S_T are charge fractions of ions i and j in the exchange phase, and S_T is assumed constant. A convenient dimensionless form for the Gaines–Thomas binary exchange selectivity coefficient during conditions of constant solution normality can be expressed as:

[10]

where C^*_i and C^*_j represent charge fractions of ion species i and j in the solution phase. For homovalent exchange K^*_ij = K_ij, but for heterovalent exchange the numerical value for K^*_ij must be adjusted for solution normality since K^*_ij K_ij. Values of K^*_ij greater than 1.0 imply preference of cation i over j, values of unity indicate no preference, and values of less than 1.0 imply preference of cation j over i.

Once a numerical solution for C^*_i (x,t) for a cation species i has been obtained, corresponding values for S^*_i(x,t) can be calculated using a general expression for an ion exchange isotherm where N cation species are involved (Valocchi et al., 1981):

[11]

The charge fraction of hydrogen can be calculated by difference:

[12]

Exchange coefficients for all possible pair-wise combination of cations are required for calculations involving Eq. [11]. In this paper, K^*_ml (and the reciprocal K^*_lm) were known, K^*_km (and the reciprocal K^*_mk) were estimated, and K^*_kl (and the reciprocal K^*_lk) were calculated using the triangle rule:

[13]

For a valence combination of r_k = 1, r_m = 1, and r_l = 2, respectively, for hydrazinium, hydrogen, and calcium, an analytical solution for the resulting quadratic equations for S_k and S_l is reported as Eq. [11] through [13] in Mansell et al. (1993). Soil CEC was replaced with effective CEC, S_Tef, rather than S_T, which is utilized within those equations.

Mass transport was simulated during the application of a pulse of aqueous solution containing a concentration C_I-k of N₂H⁺₅ ions during steady flow (_o = pore water flux) in columns of saturated (_o = volumetric water content at saturation) soil containing a calcium and proton concentrations C_0-l and C_0-m. Influent CaCl₂ (C_I-l) solutions were acidified to the values cited above for Ap, E1, and E2 to provide corresponding influent proton concentrations C_I-m-ap, C_I-m-e1, and C_I-m-e2, respectively. Initial influent normalities C_I-T were maintained approximately constant at C_I-T-ap = C_I-l + C_I-m-ap, C_I-T-e1 = C_I-l + C_I-m-e1, and C_I-T-e2 = C_I-l + C_I-m-e2, respectively, for Ap, E1, and E2 soils throughout the experiments. Initial soil solution normalities C_0-T and influent solution normalities C_I-T were approximately 10 mol_c m^-3. Initially, the soil exchange capacity S_Tef = S_T was assumed to be occupied by both divalent calcium and monovalent hydrogen ions with no chemisorbed N₂H⁺₅ ions present. The influent pulse of width (number of pore volumes) and hydrazinium concentration C_I-k was applied during 0 < t < t₁ where t is time. Initial and boundary conditions used in simulations included:

[14]

[15]

[16]

[17]

[18]

[19]

where n = l or m and L (m) is the column length.

A Crank–Nicholson finite-differencing technique (Remson et al., 1971) with iteration was used to simultaneously solve Eq. [5] through [8], using Eq. [1] through [4], in accordance with the initial and boundary conditions (Eq. [14]–[19]) in order to obtain C_k(z,t), C_l(z,t), and C_m(z,t); and Eq. [11] and [12] for S_k(z,t), S_l(z,t), and S_m(z,t). Values for z and maximum t were 2.7 mm and 5 s, respectively. Initial concentrations for each cation species in the soil were specified for the solution phase in order to solve for C_k(z,t), and the exchange isotherm equations were used to provide corresponding S_k values.

	RESULTS AND DISCUSSION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS MATHEMATICAL MODEL RESULTS AND DISCUSSION CONCLUSION REFERENCES

 
Observed breakthrough curves (BTCs; Fig. 1–6) for N₂H⁺₅, Ca²⁺, and H⁺ cations in effluent from calcium-saturated soil columns during miscible displacement of acidic solutions of N₂H⁺₅ and Ca⁺ ions provide three key observations related to transport and retention. (i) Comparative relationships between the three cation species suggests competitive ion exchange to be an important solute retention process during transport. (ii) Mass losses of applied N₂H⁺₅ ions during miscible displacement indicates the occurrence of an irreversible retention reaction such as chemisorption during transport. (iii) Asymmetry of accompanying calcium BTCs suggests that chemisorption occurring for exchange phase N₂H⁺₅ ions results in an effective decrease in soil CEC.

View larger version (22K): [in this window] [in a new window]   Fig. 1. Observed (discrete points) N2H+5 and Ca2+ breakthrough curves (BTCs) in effluent from an Arredondo Ap soil column injected with a pulse of high hydrazinium concentration (22.15 molc m-3). Simulated BTCs with (smooth lines) and without chemisorption (broken lines) of exchange-phase N2H+5 are given.

View larger version (18K): [in this window] [in a new window]   Fig. 6. Observed (discrete points) N2H+5 and Ca2+ breakthrough curves (BTCs) in effluent from an Arredondo E2 soil column injected with a pulse of medium hydrazinium concentration (5.45 molc m-3). Simulated BTCs with (smooth lines) and without (broken lines) chemisorption of exchange-phase N2H+5 are given.

View larger version (20K): [in this window] [in a new window]   Fig. 2. Observed (discrete points) N2H+5 and Ca2+ breakthrough curves (BTCs) in effluent from an Arredondo Ap soil column injected with a pulse of medium hydrazinium concentration (5.64 molc m-3). Simulated BTCs with (smooth lines) and without (broken lines) chemisorption of exchange-phase N2H+5 are given.

View larger version (19K): [in this window] [in a new window]   Fig. 3. Observed (discrete points) N2H+5 and Ca2+ breakthrough curves (BTCs) in effluent from an Arredondo E1 soil column injected with a pulse of high hydrazinium concentration (15.00 molc m-3). Simulated BTCs with (smooth lines) and without (broken lines) chemisorption of exchange-phase N2H+5 are given.

View larger version (19K): [in this window] [in a new window]   Fig. 4. Observed (discrete points) N2H+5 and Ca2+ breakthrough curves (BTCs) in effluent from an Arredondo E1 soil column injected with a pulse of medium hydrazinium concentration (5.70 molc m-3). Simulated BTCs with (smooth lines) and without (broken lines) chemisorption of exchange-phase N2H+5 are given.

View larger version (19K): [in this window] [in a new window]   Fig. 5. Observed (discrete points) N2H+5 and Ca2+ breakthrough curves (BTCs) in effluent from an Arredondo E2 soil column injected with a pulse of high hydrazinium concentration (16.66 molc m-3). Simulated BTCs with (smooth lines) and without (broken lines) chemisorption of exchange-phase N2H+5 are given.

Model parameters used in simulations were either taken from the constants of a given soil (Table 1), a given experiment (Table 2), or were estimated by sensitivity analyses on solute BTCs. The estimated parameters included exchange coefficients, the dispersion coefficient, and the specific soil characteristics of S_T and S^'_k-max. Determination of all pair-wise exchange coefficients is critical to model simulations of ternary ion transport and/or exchange. A numerical value for the dimensionless Gaines–Thomas binary selectivity coefficient K^*_ml of 47.08 (Table 2), indicating a strong preference of carboxyl exchange sites for H⁺ relative to Ca²⁺ ions, was obtained from exchange isotherms reported (van Beinum et al., 2001) for idealized soil organic matter (synthetic alginate beads having carboxyl exchange sites). The CEC or carboxyl-site density for alginate (5000 mol_c Mg^-1) is similar to that for humic acids in soils. Preference of exchange sites for H ions in humic materials generally exceeds that for divalent Ca, which exceeds Na (Stevenson, 1982); although for pure clays and resins H selectivity is generally similar to monovalent Na (Rhue and Mansell, 1988). Hydrogen ions were reported by Rhue and Mansell (1988) to have a large preference over Ca and Na metal cations for exchange sites in Cecil sandy loam (fine, kaolinitic, thermic Typic Kanhapludult). Numerical values for K^*_kl = 0.750 and K^*_km = 0.126 were obtained by calibration of the transport model (considering exchange without chemisorption) using solute BTCs for Column 1 and using the triangle rule, respectively (Table 1). The selectivity values were taken to apply to all three soils. The general order of preference of soil exchange sites for cation species was thus taken to be H⁺ > Ca²⁺ > N₂H⁺₅, indicating hydrazinium to be a weak competitor for exchange sites relative to calcium and hydrogen ions.

Critical model input parameters D, S_T, S^'_k-max, and k_f were obtained for Ap, E1, and E2 soils by performing sensitivity analyses calibration of S_Tef BTCs for Columns 1 (Fig. 1), 3 (Fig. 3), and 5 (Fig. 5), respectively. An attempt was made to hold as many parameters as possible constant across all three soils. Also, an assumption was made that hydrazinium chemistry was essentially similar across soils. Sensitivity analysis revealed a calibrated dispersion coefficient of 3.0 x 10^-7 m² s^-1 to be adequate for all three soils. Cation exchange capacity values were obtained first by performing simple ternary cation exchange. Values of S^'_k-max were then obtained by performing ion exchange coupled with chemisorption. Calibrated values for CEC, S^'_k-max, and k_f parameters for Ap, E1, and E2 soils are presented in Table 2.

Irreversible losses (mass not recovered in column effluent) of hydrazinium from pulses of solution applied during miscible displacement experiments were assumed primarily attributable to irreversible specific sorption or chemisorption (Table 3). Applied amounts of N₂H⁺₅ ranged from 1.62 to 1.83 mol_c for pulses with medium concentrations and 3.37 to 7.20 mol_c for pulses with high concentrations. Highest percentage losses of hydrazinium were associated with the lowest values of C_0-k in the influent pulses. For a given value of C_0-k, percentage losses, organic contents, and CECs were in the order Ap > E1 > E2 for the three soil horizons. Simulated minus observed losses of applied N₂H⁺₅ ranged from 1.4 to -7.9% indicating adequate approximation by the model. Observed and simulated masses of nonrecovered N₂H⁺₅ ranged from 0.07 to 1.63 mol_c and from 0.12 to 1.43 mol_c, respectively.

Hydrogen concentrations in the column effluent increased during exchange of hydrazinium with calcium cations for Ap, E1, and E2 soils (Fig. 7; BTCs not shown for E1 and E2 soils). The elevated concentrations of hydrazinium corresponded with those of protons. Although exchangeable hydrogen is well known to compete effectively with metal and organic cations in exchange reactions involving soil organic matter (Stevenson, 1982), the proton concentration in the solution phase was very small relative to the calcium and hydrazinium concentrations. Humic acids and other components of soil organic matter generally behave as weak-acid polyelectrolytes. The chemical behavior of soil organic matter in the pH range from 3 to 9 has been described as a polycarboxylic acid, which adsorbs metal cations with the release of H⁺ ions, and under very acid conditions, behaves like an uncharged polymer (Bloom and McBride, 1979).

View larger version (17K): [in this window] [in a new window]   Fig. 7. Observed (discrete points) H+ breakthrough curve (BTC) in effluent from an Arredondo Ap soil column injected with a pulse of high hyrdazinium concentration (22.15 molc m-3). Simulated BTCs with (smooth lines) and without (broken lines) chemisorption of exchange-phase N2H+5 are given.

Hypothetical exchange isotherms for N₂H⁺₅ and Ca²⁺ for Ap soil were prepared using an algorithm developed by Bloom and Mansell (2001) from binary exchange selectivity coefficients used in the transport simulations (Fig. 8). The difference between and initial (C_o) and final (C_e) N₂H⁺₅ concentrations represents the quantity of charge removed from the solution phase that is assumed here to be attributable to retention by soil components due to exchange and/or chemisorption. The quasi multiprocess sorption isotherm where the quasi charge fraction S^*_kl as S_kl/S_T rather than S_kl/S_Tef demonstrates a major effect of chemisorption within the relatively low N₂H⁺₅ concentration range 0 < C^*_kl < 0.20 that occurred within the column transport experiments. Use of this algorithm provides a convenient means to numerically separate exchange and sorption isotherms for cations that undergo multiprocess retention by soil components. For C^*_kl 0 and S^*_kl 0 the value of S^'_k = S^'_k-max (3.5 mol_c Mg^-1). Although experimental determination of sorption isotherms for cations subject to multiprocess retention is relatively easy, experimental determination of the corresponding isotherm isolating pure ion exchange could be a difficult task.

View larger version (20K): [in this window] [in a new window]   Fig. 8. Simulated exchange isotherms (solid circles; upper line) for N2H+5 for Arredondo Ap soil calculated from binary exchange selectivity coefficients used in transport simulations. A sorption isotherm (open circles; lower line) that includes both ion exchange and N2H+5 chemisorption was obtained by dividing the sorbed phase concentration by ST rather than STef. The terms Co and Ce represent initial and final N2H+5 concentrations in the solution phase.

The multiprocess model reported here for convective–dispersive transport of N₂H⁺₅ cations in Arredondo fine sand includes predominant retention by ion exchange coupled with secondary chemisorption. Using a minimal set of retention processes and input parameters, the model accurately described observed complex transport of N₂H⁺₅, Ca²⁺, and H⁺ ions during steady flow through columns of water-saturated Arredondo fine sand. The model is not comprehensive with regard to retention reactions because processes such as biodegradation were omitted. Because the model was developed long after transport experiments were performed, measurements for a number of critical input parameters for the model and definitive experiments were not available. Independent experiments should include application of multiple pulses of isotopically labeled hydrazinium to soil columns over a range of water flux and exchange and/or transport experiments using soil initially treated with highly competitive cations, such as barium, to effectively block exchange of N₂H⁺₅ ions so as to verify the role of exchange to chemisorption. Independent measurement of the sorption isotherms is also needed.

	CONCLUSION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS MATHEMATICAL MODEL RESULTS AND DISCUSSION CONCLUSION REFERENCES

 
A multiprocess model was developed to describe transport of three cations—Ca²⁺, N₂H⁺₅, and H⁺—that undergo competitive instantaneous exchange under conditions of exchange-mediated chemisorption of N₂H⁺₅ cations, which resulted in a variable effective CEC for the soil. Exchange sites were assumed dominated by carboxyl groups associated with soil organic matter. Chemisorption was assumed in the model to occur as irreversible condensation of exchange-phase N₂H⁺₅ ions with carbonyl groups of humic components of the soil.

Simulated breakthrough curves (BTCs) for concentrations of N₂H⁺₅ and Ca²⁺ ions in column effluent approximated observations. Corresponding N₂H⁺₅ and Ca²⁺ BTCs were characteristically asymmetrical, exhibiting strong tailing that was dependent upon influent N₂H⁺₅ concentration, and provided incomplete recovery of applied hydrazinium. Unrecovered N₂H⁺₅ was attributed to exchange-mediated chemisorption. Although the concentrations of H⁺ ions in solution were several orders of magnitude less than the other two cation species, simulations described a general trend of depressions in effluent pH associated with N₂H⁺₅ breakthrough. Increased proton concentration in the solution phase occurred primarily by cation exchange with no evidence of generation by molecular degradation of N₂H⁺₅.

	ACKNOWLEDGMENTS

 
This research was partially supported by earlier funding from the United States Air Force Environics Division, Tyndall Air Force Base, FL (No. F08635-83-C-0136).

	NOTES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS MATHEMATICAL MODEL RESULTS AND DISCUSSION CONCLUSION REFERENCES

 
Florida Agric. Exp. Stn. Journal Ser. no. R-07853.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS MATHEMATICAL MODEL RESULTS AND DISCUSSION CONCLUSION REFERENCES

 

• Bloom, P.R., and M.B. McBride. 1979. Metal ion binding and exchange with hydrogen ions in acid-washed peat. Soil Sci. Soc. Am. J. 43: 687–692.[Abstract/Free Full Text]
• Bloom, S.A., and R.S. Mansell. 2001. An algorithm for generating cation exchange isotherms from binary selectivity coefficients. Soil Sci. Soc. Am. J. (in press).
• Braun, B.A., and J.A. Zirrolli. 1983. Environmental fate of hydrazine fuels in aqueous and soil environments. Final Rep. ESL-TR-82-45. Eng. and Services Lab., Tyndall Air Force Base, FL.
• Brusseau, M.L. 1998. Multiprocess nonequilibrium and nonideal transport of solutes in porous media. p. 63–82. In H.M. Selim and L. Ma (ed.) Physical nonequilibrium in soils. Ann Arbor Press, Chelsea, MI.
• Burns, I.G., M.H.B. Hayes, and M. Stacey. 1973. Some physico–chemical interactions of paraquat with soil organic materials and model compounds. II. Adsorption and desorption equilibria in aqueous suspensions. Weed Res. 13:79–90.
• Cuy, E.J., and W.C. Bray. 1924. The oxidation of hydrazine. II. The effect of oxygen on the decomposition of hydrazine: The reactions with ferricyanide in alkaline solutions and dichromate in acid solution. J. Am. Chem. Soc. 46:1786–1795.
• Ellis, S.R.M., G.V. Jeffreys, and P. Hill. 1960. Oxidation of hydrazine in aqueous solution. J. Appl. Chem. 10:347–352.
• Gee, G.W., and J.W. Bauder. 1986. Particle-size analysis. p. 383–409. In A. Klute (ed.) Methods of soil analysis. Part 1. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.
• Giddings, J.C. 1965. Dynamics of chromatography. Part I. Principles and theory. Marcel Dekker, New York.
• Griffith, S.M., J. Silver, and M. Schnitzer. 1980. Hydrazine derivatives at Fe⁺³ sites in humic materials. Geoderma 23:299–302.
• Hayes, M.H.B., K.Y. Chia, and T.B.R. Yormah. 1988. Interactions of hydrazines with colloidal constituents of soils. p. 94–107. In D.A. Stone and F.L. Wiseman (ed.) The 3rd Conf. on the Environ. Chem. of Hydrazine Fuels. Hazardous Materials Tech. Center, Rockville, MD.
• Isaacson, P.J., and M.H.B. Hayes. 1984. The interaction of hydrazine hydrate with humic acid preparations at pH 4. J. Soil Sci. 35:79–92.
• Mansell, R.S., W.J. Bond, and S.A. Bloom. 1993. Simulating cation transport during water flow in soil: Two approaches. Soil Sci. Soc. Am. J. 57:3–9.
• Moliner, A.M., and J.J. Street. 1989a. Decomposition of hydrazine in aqueous solutions. J. Environ. Qual. 18:483–487.[Abstract/Free Full Text]
• Moliner, A.M., and J.J. Street. 1989b. Interactions of hydrazine with clays and soils. J. Environ. Qual. 18:487–491.[Abstract/Free Full Text]
• Nelson, D.W., and L.E. Sommers. 1982. Total carbon, organic matter. p. 539–577. In A.L. Page et al. (ed.) Methods of soil analysis. Part 2. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.
• Ou, L.T. 1987. Microbial degradation of hydrazine. Bull. Environ. Contam. Toxicol. 39:78–85.[Medline]
• Reed, M. 1991. Anatomy of a rail disaster. Los Angeles Times, 4 August, p. 1, Metro Section.
• Remson, I., G.M. Hornberger, and F.J. Molz. 1971. Numerical methods in subsurface hydrology. Wiley-Interscience, New York.
• Rhue, R.D., and R.S. Mansell. 1988. The Effect of pH on sodium-calcium and potassium-calcium exchange selectivity for Cecil soil. Soil Sci. Soc. Am. J. 52:641–647.[Abstract/Free Full Text]
• Rubin, J., and R.V. James. 1973. Dispersion-affected transport of reacting solutes in saturated porous media: Galerkin method applied to equilibrium-controlled exchange in unidirectional steady water flow. Water Resour. Res. 9:1332–1356.
• Schmidt, E.W. 1984. Hydrazine and its derivatives: Preparation, properties, applications. John Wiley & Sons, New York.
• Schnitzer, M., and S.I.M. Skinner. 1965. The carbonyl group in a soil organic matter preparation. Soil Sci. Soc. Am. Proc. 29:400–405.
• Selim, H.M., and M.C. Amacher. 1997. Reactivity and transport of heavy metals in soils. Lewis Publ., Boca Raton, FL.
• Stevenson, F.J. 1982. Humus chemistry. John Wiley & Sons, New York.
• Valocchi, A.J., R.L. Street, and P.V. Roberts. 1981. Transport of ion-exchanging solutes in groundwater: Chromatographic theory and field simulation. Water Resour. Res. 17:1518–1527.
• Van Beinum, W., J.C.L. Meeussen, and W.H. van Riemsdijk. 2001. Modelling transport of protons and calcium ions in an alginate gel bead system: The effects of physical nonequilibrium and nonlinear competitive sorption. Environ. Sci. Technol. (in press).
• Watt, G.W., and J.D. Crisp. 1952. A spectrophotometric method for the determination of hydrazine. Anal. Chem. 24:2001–2008.

This article has been cited by other articles:

				S. A. Bloom and R. S. Mansell An algorithm for generating cation exchange isotherms from binary selectivity coefficients Soil Sci. Soc. Am. J., September 1, 2001; 65(5): 1426 - 1429. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (1) Citing Articles via Google Scholar Google Scholar Articles by Mansell, R. S. Articles by Downs, W. C. Search for Related Content PubMed PubMed Citation Articles by Mansell, R. S. Articles by Downs, W. C. GeoRef GeoRef Citation Agricola Articles by Mansell, R. S. Articles by Downs, W. C. Related Collections Water Quality Water Pollution Ground Water Quality