Partitioning of Organofluorine Compounds in the Environment

CHAPTER 2

Partitioning of Organofluorine Compounds in the Environment David A. Ellis 1, Thomas M. Cahill 3, Scott A. Mabury 2, Ian T. Cousins 4, Donald Mackay 5 1 2 3 4 5

Chemistry Department, University of Toronto, Toronto, Ontario M5S 3H6, Canada. E-mail: [email protected] Chemistry Department, University of Toronto, Toronto, Ontario M5S 3H6, Canada. E-mail: [email protected] Canadian Environmental Modelling Centre, Trent University, Peterborough, Ontario K9J 7B8, Canada. E-mail: [email protected] Canadian Environmental Modelling Centre, Trent University, Peterborough, Ontario K9J 7B8, Canada. E-mail: [email protected] Canadian Environmental Modelling Centre, Trent University, Peterborough, Ontario K9J 7B8, Canada. E-mail: [email protected]

This chapter describes the partitioning properties of organofluorine compounds in the environment. Partitioning in the air-water-octanol system is discussed using these pure substances as surrogates for real environmental systems consisting of air, water, natural organic matter and lipids. It is shown that the substitution of fluorine for hydrogen in alkanes, aromatics and carboxylic acids causes significant changes in properties such as vapor pressure, aqueous solubility, octanol-water partition coefficient and acid dissociation constant. These changes are quite different in magnitude from the corresponding changes caused by other better-studied halogens, namely chlorine and bromine. Perfluorinated substances have unique properties imparted by their minimal intermolecular interactions. The environmental implications of these properties are illustrated using simple multimedia partitioning models showing that organofluorine compounds behave quite differently than organochlorine analogs and more closely resemble the corresponding non-halogenated compounds. Keywords. Fluorine, Organofluorine, Haloacetic acids, Fluorocarbon, Halogen, Hydrofluoro-

carbon, Perfluorinated compounds, Halocarbon

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

1

Introduction

2

Elemental Fluorine . . . . . . . . . . . . . . . . . . . . . . . . . . 65

3

Physico-Chemical Properties

3.1 3.2 3.3

Short-Chain Fluorocarbons . . . . . . . . . . . . . . . . . . . . . 66 Fluorinated Benzenes . . . . . . . . . . . . . . . . . . . . . . . . 69 Fluorinated Acetic Acids . . . . . . . . . . . . . . . . . . . . . . . 71

4

Implication for Environmental Fate – Evaluative Modeling of the Fluoro- and Chlorobenzenes . . . . . . . . . . . . . . . . . . . . 75

5

Conclusions

6

References

. . . . . . . . . . . . . . . . . . . . 65

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 The Handbook of Environmental Chemistry Vol. 3, Part N Organofluorines (ed. by A.H. Neilson) © Springer-Verlag Berlin Heidelberg 2002

64

D.A. Ellis et al.

List of Abbreviations BCF CFC DDT DFA EPIWIN EQC HCFC HFC Ka KOW MFA PCB TFA VP

bioconcentration factor chlorofluorocarbon para-dichlorodiphenyltrichloroethane difluoroacetic acid Estimations Programs Interface for Windows Equilibrium Criterion Model hydrochlorofluorocarbon hydrofluorocarbons equilibrium acid dissociation constant octanol-water partition coefficient monofluoroacetic acid polychlorinated biphenyl trifluoroacetic acid vapor pressure

1 Introduction Since the publication of Rachel Carson’s “Silent Spring” in 1962, there has been a continuing focus on the environmental fate and effects of organochlorine compounds such as PCBs, “dioxins” and pesticides such as DDT and Mirex. The physico-chemical effects of substituting chlorine for hydrogen in organic molecules are now well known and highly predictable. In general, this substitution causes a marked increase in molar mass, which results in a reduced vapor pressure. The larger molecule has a larger surface area which usually induces a stronger hydrophobic effect. This is manifested as a decrease in water solubility and a corresponding increase in the octanol-water partition coefficient, KOW . The substitution of a hydrogen atom with chlorine also imparts stability to degradation by both abiotic reactions, as is the case with OH radical attack, and biotic reactions such as microbial degradation. The combination of increased hydrophobicity and stability results in molecular characteristics that heighten environmental concerns. Much less attention has been paid to the other halogens, although there is increasing concern about organobromine substances such as brominated fire retardants. To a first approximation, bromine appears to behave similarly to chlorine, but there are quantitative differences attributable to the size and mass of the bromine atom, i.e. bromine is approximately twice the mass of chlorine and has a slightly larger atomic radius. The carbon-chlorine bond is also considerably stronger than the carbon-bromine bond, namely 397±29 compared with 280±21 kJ/mol respectively. Organofluorine compounds have received relatively little attention when compared with organochlorine compounds.An obvious exception was the study of the volatile chlorofluorocarbons (CFCs) and related compounds that were used as refrigerants and were implicated as agents responsible for stratospheric ozone depletion. Although there have been several reviews of the commercial [2] and biological aspects of organofluorine com-

Partitioning of Organofluorine Compounds in the Environment

65

pounds, the first comprehensive critical review of their environmental sources, fate and effects was by Key et al. [26]. The goals of this review are to: 1. compile selected available data on the environmentally relevant physico-chemical properties of organofluorine compounds; 2. discuss the nature of the physico-chemical properties induced in the molecule through fluorination and, highlight the differences between fluorinated molecules and their corresponding chlorine and bromine analogs; 3. apply an evaluative environmental model based upon the physico-chemical properties of a selected class of fluorinated organics in order to investigate their likely environmental distribution relative to their non-halogenated and chlorinated counterparts. It transpires that, as is often the case, the first member of a “homologous series”, in this case fluorine, displays a number of unique characteristics that impart unusual partitioning properties in both model systems and the environment.

2 Elemental Fluorine Although this review focuses on the organofluorine compounds, a brief discussion of the properties of elemental fluorine is useful to assist in understanding some of the properties of organofluorine compounds. Table 1 shows some of the physical properties of fluorine relative to chlorine and bromine. Fluorine gas, F2 , is the strongest known oxidizing agent and it can directly react with most elements. The extreme reactivity of fluorine is exemplified by the fact that even water “burns” in the presence of fluorine gas to form hydrofluoric acid and oxygen. Hydrocarbons react directly with F2 to form fluorinated hydrocarbons. For example: CH4 +4 F2 Æ 4 HF+CF4 Fluorine can even react directly with the heavier noble gases, such as xenon, to form fluorinated compounds (XeF2 , XeF4 , XeF6) [11]. Fluorine has an electronegativity of 4.0 on the Pauling scale, which is the highest of all the elements. This extreme electronegativity results in fluorine drawing electrons from adjacent atoms, thus creating polar bonds. For example, the electronegativity difference between fluorine and carbon is 1.5 units, which indicates that the electrons are shared unequally between the two atoms and that there is a partial negative charge on the fluorine and a partial positive charge on the carbon.

3 Physico-Chemical Properties In this section we review the physico-chemical properties of organofluorine compounds focusing on the highly fluorinated compounds where fluorine is the dominant functional group. The properties of simple fluorocarbons, fluorobenzenes and fluorinated acetic acids are described and discussed, with particular

66

D.A. Ellis et al.

attention paid to the influence imparted to the molecule by fluorine. Some compounds, such as the widely used herbicide trifluralin, contain a trifluoro group, but the physico-chemical properties of the molecule are dominated by other functional groups, so they are not considered here. 3.1 Short-Chain Fluorocarbons

Fluorine is commonly used in the synthesis of volatile short-chain halocarbons that are used as refrigerants, aerosol propellants and foam blowing agents. Initially, the chlorofluorocarbons (CFCs) were used for this purpose, but they were implicated in stratospheric ozone depletion [39] and have subsequently been replaced by the hydrofluorocarbons (HFCs) [51]. The CFCs were typically fully halogenated methanes while the HFCs are generally partially fluorinated ethanes such as HFC 134a (1,1,1,2-tetrafluoroethane), which is the most commonly used HFC. The effect of fluorination on the vapor pressure of short chain hydrocarbons illustrates its unusual behavior. For alkanes, the substitution of one fluorine for hydrogen results in a decrease of the vapor pressure of the compound (Fig. 1). In this respect, fluorine behaves in a consistent fashion with respect to the other halogens. In each case, the vapor pressure of the monofluorocarbon is less than the vapor pressure of the parent compound by a factor of 2.5 to 5.0. The addition of multiple chlorine and bromine atoms to the hydrocarbon causes a similar trend in vapor pressure reduction; while the addition of multiple fluorines to a hydrocarbon results in anomalous behavior. Figure 2 shows the effect of adding multiple fluorines to the volatility (as expressed by boiling

Fig. 1. Influence of monohalogenation on the vapor pressure of the short chain hydrocarbons.

Hydrocarbon data from Lide et al. [29], fluorocarbon data from [6, 20, 35, 41], and the other halocarbon data from Mackay et al. [31]


67

Fig. 2. The boiling points of the fluorinated, chlorinated and brominated methanes. (Data from

[29])

point) of methane. The addition of one or two fluorines to methane causes the expected increase in the boiling point of the compound, which indicates that the partially fluorinated methanes are less volatile than methane. However, the addition of the third and fourth fluorines cause a decrease in boiling point, thus indicating that they are more volatile than the partially fluorinated methanes. The increase in volatility occurs despite the increase in molar mass. In contrast to fluorine, the addition of each chlorine or bromine to methane causes a consistent increase in the boiling point or a decrease in vapor pressure, which is expected since the molecule has a higher molar mass. A similar trend is observed with the fluorinated ethanes where the partially fluorinated ethanes have higher boiling points (–47.3 to 30.7 °C) than either ethane (–88.6 °C) or perfluoroethane (–79 °C) [29]. Fluorine is therefore unique in that perfluorination results in an increase in vapor pressure over the partially fluorinated hydrocarbons.

Fig. 3. Anticipated dipole-dipole interaction between two partially-fluorinated methanes

68

D.A. Ellis et al.

Table 1. Selected properties of fluorine, chlorine and bromine atoms where “X” is halogen [29]

Atomic mass Ionic radius (nm) Bond length in CX4 (nm) C–X Bond energy (kJ/mol) Electronegativity

Fluorine

Chlorine

Bromine

18.998 0.133 0.1323 552 4.0

35.453 0.181 0.1767 397±29 3.0

79.904 0.196 0.1935 280±21 2.8

A suggested mechanism for this anomalous behavior is that the high electronegativity of the fluorine creates a polar bond with the carbon. The unequal sharing of electrons creates a dipole moment in the partially fluorinated methanes. The dipole-dipole interaction between molecules (Fig. 3) results in a reduction of the vapor pressure and increases the boiling point. Tetrafluoromethane, however, has no overall net dipole moment since polarization of the C–F bonds occurs equally in all directions. The observed change in the boiling point trend observed for the tri- and tetrafluoromethanes, when compared with their other halogen counterparts, can be explained through changes in the atomic size and bond length (Table 1). As previously described, each halogen causes a polarization of the bond between it and the central carbon atom. The shorter bond length between carbon and fluorine results in the exterior negative shell being held more tightly to the carbon atom. These two inherent properties renderthe carbon atom less susceptible to intermolecular interactions causing these molecules to show enhanced volatility.

Fig. 4. Comparison of boiling points of the perfluoroalkanes to the normal alkanes. Fluorocarbon data from [1, 4, 8, 22, 29, 36–38, 43, 44]. Hydrocarbon data from [7, 12, 18, 21, 24, 29, 40, 42, 47, 48, 50]

69


Table 2. Comparison of the boiling points of noble gases, perfluorocarbons and hydrocarbons

of similar molecular weights. (Data from [29]) Molecular weight analogs (molecular wt. in parenthesis)

Boiling point (°C)

Difference (°C) between noble gas and organic b.p.

Kr (83.8) CF4 (88.0) Hexane (86.2)

–152 –129 69

– 23 221

Xe (131.3) C2F6 (138.0) Nonane (128.3)

–107 –79 151

– 28 258

–62 4 287

– 66 349

Rn (222) C4F10 (238.0) Hexadecane (226.5)

A related and unusual property of the perfluorinated hydrocarbons is that they become more volatile than the parent hydrocarbon when the chain length exceeds four carbons. Figure 4 shows the boiling point for hydrocarbons and perfluorocarbons of a variety of carbon numbers. Perfluorinated linear chain aliphatic compounds exhibit unique properties when compared to those of nonfluorinated analogs. The origins of these unusual properties lie in the interaction between neighboring –CF2– units. Strong electronic repulsion of adjacent fluorine atoms results in the backbone of the chain being held rigid. It is this rigidity, coupled with the large partial negative charge associated with each fluorine atom, results in the vapor pressure of these molecules being substantially higher than would be predicted based purely upon the mass of the molecules. It is interesting to compare the noble gases, perfluorocarbons and normal alkanes of similar molecular weights as in Table 2. In each case, the perfluorocarbon behaves more like a noble gas than a hydrocarbon of corresponding molecular weight. This suggests that there are minimal intermolecular interactions in perfluorocarbons and that their volatility may be related more to their molar mass than the hydrocarbons. In conclusion, the addition of fluorine to a hydrocarbon generally results in a drop in vapor pressure, as would be expected in a molecule of higher molar mass and increased dipole moment. The perfluorocarbons behave differently and have higher vapor pressures indicating that dipole induced intermolecular interactions in the liquid phase are reduced. 3.2 Fluorinated Benzenes

The effects of adding a single fluorine to benzene is similar to that of the alkanes. Figure 5 shows a decrease in the vapor pressure resulting from monofluorination, which is similar to that observed for the alkanes. The decrease in vapor pressure is consistent with the vapor pressure decrease caused by the addi-

70

D.A. Ellis et al.

Fig. 5. The octanol-water partition coefficient and vapor pressure of the monohalogenated ben-

zenes at 25 °C. (Data from [31])

tion of other halogens to benzene. In addition, the octanol-water partition coefficient (KOW) also increases by the addition of a single fluorine. The increase in log KOW is probably related to the increase in molar volume or surface area. Whereas multiple chlorination of benzene causes a consistent change in the physico-chemical properties of the molecule, the influence of multiple fluorination is less predictable. Figure 6 shows the boiling points of the fluorinated and chlorinated benzenes. An increase in fluorination of benzene has little effect on

Fig. 6. The boiling points for the fluorinated, chlorinated and brominated benzenes. The “boiling point” of hexachlorobenzene actually represents a sublimation point. Data for pentabromobenzene and hexabromobenzene where not available. All data from [29]


71

Fig. 7. The octanol-water partition coefficient (KOW) for the fluorinated and chlorinated ben-

zenes. (Chlorobenzene data from [31]; fluorobenzene data from [3, 17, 19, 23, 52])

the boiling point of the molecule. The partially fluorinated benzenes are slightly less volatile than either benzene or hexafluorobenzene. In contrast, the addition of chlorine or bromine to benzene increases the boiling point of the molecule by consistent increments as a function of the number of halogens. The effect of multiple fluorination on the octanol-water partition coefficient also contrasts with the behavior of chlorine. Figure 7 shows the log KOW of the fluorobenzenes and the chlorobenzenes. While the addition of each chlorine causes a consistent increase in log KOW , the addition of fluorine initially causes an increase in log KOW , but then the addition of the last two fluorines causes a decrease. This behavior results in difficulties in establishing quantitative structural property relationships based on constant atom or group contributions. 3.3 Fluorinated Acetic Acids

Fluorinated acetic acids are frequent degradation products of some organic fluorine compounds. Trifluoroacetic acid has received considerable attention since it is a known breakdown product of certain pesticides [13] and the HFC refrigerants and aerosol propellants [51]. Chlorodifluoroacetic acid may be formed via the degradation of CFC-113 in the stratosphere or through the direct tropospheric degradation of hydrochlorofluorocarbons (HCFCs) such as HCFC-142b [33]. Fluorinated acetic acids are now ubiquitous in the environment [5, 9, 14, 25, 45]. Table 3 gives the melting and boiling points for acetic acid and the fluorinated acetic acids, the latter indicating their relative volatilities. The replacement of one hydrogen with a fluorine increases the boiling point of the acid, which indicates

72

D.A. Ellis et al.

Table 3. The melting point and boiling point of the fluorinated acetic acids

Chemical

Melting point (°C)

Boiling point (°C)

CH3COOH CH2FCOOH CHF2COOH CF3COOH

16.6 35.2 –0.3 –15.2

118.9 165 134.2 72.4

Data from Lide [29].

that monofluoroacetic acid (MFA) is less volatile than acetic acid. The decreased volatility of the mono-fluorinated species follows the same general trend as the fluorinated alkanes where the first fluorine added to the molecule decreases the vapor pressure. The boiling point of difluoroacetic acid (DFA) is lower than MFA, which indicates that DFA is more volatile than MFA. Lastly, trifluoroacetic acid (TFA) has the lowest boiling point, and thus the greatest volatility. This trend is similar in nature to the vapor pressure of the alkanes where the species with perfluorinated carbons have higher vapor pressures than partially fluorinated species. In contrast to the alkanes, in which methane was more volatile than tetrafluoromethane, TFA is more volatile than the parent compound of acetic acid. The addition of halogens to acetic acid increases the acid-dissociation constant (Ka) of the acids. Ka is the equilibrium constant between the unionized and the ionized forms of the acid and is numerically defined as [H3O+][A–]/[HA]. Larger values of Ka imply stronger acids that are more likely to lose protons and be found in the ionized form. Figure 8 shows the effect of halogen substitution on the Ka of the monohalogenated acetic acid. Fluorine causes the greatest increase in Ka of the monohalogenated acetic acids, which indicates that MFA is more easily ionized than the other monohaloacetic acids.

Fig. 8. The influence of halogen electronegativity on the acid-dissociation constants of the

haloacetic acids. (Data from [46])


73

Fig. 9. The influence of multiple fluorination of the acid-dissociation constant of the fluorinated acetic acids. An increase in the number of fluorines causes an increase in Ka that indicates that the highly fluorinated acetic acids are more prone to ionization. (Data from [46])

The Ka is observed to increase with increasing fluorination (Fig. 9). The Ka values indicate that the perfluorinated acids will be completely dissociated under normal aqueous and physiological conditions. Upon deprotonation of the acid, the resultant negative charge is stabilized through the bond inductive effect of the fluorine atoms (Fig. 10). Thus, as the number and magnitude of electronegative elements that are contained within the acetic acid increases, so too does the Ka . Perfluorination of carboxylic and sulfonic acids also imparts a high degree of chemical stability or resistance to degradation. The stability of perfluorinated compounds is demonstrated by the lack of microbial degradation of perfluorinated carboxylates and sulfonates compared to their partially fluorinated counterparts. Recent studies by Ellis et al. [15], Emptage et al. [16] and Cahill et al. [10] have shown that TFA is stable under anaerobic and field microcosm conditions. Earlier research indicated that reductive dehalogenation of TFA could occur [49], although these results have not been replicated by other re-

Fig. 10. Inductive stabilization of negative charge on trifluoroacetate

74

D.A. Ellis et al.

Fig. 11. Examples of fluorinated compounds that are microbially degradable and compounds

that not been shown to be degraded by microbial mechanisms. The primary difference between the compounds is the presence of one or more hydrogens that represent a reactive site. (sulfonate data from [27], fluorinated acetate data from [16], trichloroacetate data from [15])

searchers [16]. A recent study by Kim et al. [28] demonstrated that TFA could be degraded in an anaerobic reactor at elevated temperatures in the presence of ethanol, so TFA degradation may occur in the field, but probably at rates too slow to be readily observed. Partially fluorinated acetates, especially fluoroacetate that is highly toxic to biota, have been shown to be degradable by microbes (references in Chap. 7 and [16]). Key et al. [27] demonstrated that only fluorinated sulfonates that contained hydrogen could be degraded by bacteria. In all cases, the compounds that lacked hydrogen were recalcitrant to degradation by aerobic bacteria [27]. Figure 11 shows the substances that were degraded compared to the non-metabolized compounds. In contrast to trifluoroacetate, trichloroacetate is degradable in field microcosm conditions [15].


75

4 Implication for Environmental Fate – Evaluative Modeling of the Fluoro- and Chlorobenzenes A full evaluation of the probable environmental fate of the fluorinated organic compounds discussed earlier is beyond the scope of this chapter. It is, however, important to appreciate the cardinal significance of these physico-chemical properties in determining their environmental fate. This is illustrated by a discussion of fluorobenzenes and a comparison with the analogous chlorobenzenes. As noted earlier, there may be a perception that a compound with fluorine substituents will behave somewhat similarly to its chlorinated analog. This proves not to be the case for partitioning, since fluorinated compounds clearly display anomalous behavior in the bromine-chlorine-fluorine series. The calculated behavior of a chemical in a model environment provides a basis for evaluating its environmental fate.An environmental impact assessment of a compound relies on a number of factors: its chemical properties, which largely determine the phase into which the chemical will partition, its tendency to bioaccumulate, the possibility of long-range transport, and the primary mechanisms for its loss that determine its environmental persistence. In addition, numerical results from modeling provide “benchmark” environmental data for a chemical that can be compared with those for chemicals with established properties in a comparable environment. The Equilibrium Criterion or EQC model [30] is a widely used evaluative model that treats an area of 105 km2 with about 10% of the area being covered by water. The temperature in the EQC environment is set at 25 °C, which is a common temperature at which physico-chemical properties are measured. Evaluative modeling assessments usually progress through three stages of complexity; Levels I, II and III. Each subsequent level requires more detailed information or includes additional processes providing a step-wise increase in understanding of the chemical behavior in the environment. Mackay [32] and Mackay et al. [30] provide a complete description of these calculations with examples at each level of complexity. The calculations are conducted using the concept of fugacity which simplifies the equations and facilitates interpretation of the results. Here we undertake evaluative fate modeling of benzene and selected fluoroand chlorobenzenes, specifically; monofluorobenzene, 1,2,4-trifluorobenzene, hexafluorobenzene (or perfluorobenzene), monochlorobenzene, 1,2,4-trichlorobenzene and hexachlorobenzene (or perchlorobenzene). The aim is to determine how substitution of fluorine and chlorine on the benzene ring affects environmental partitioning and fate. The physico-chemical data required for the EQC model are vapor pressures, water solubilities, octanol-water partition coefficients, melting points and reaction rate constants in air, water, soils and sediments. Physico-chemical property data for benzene and the chlorobenzenes were taken from a handbook of physico-chemical properties [31]. Melting points for the fluorobenzenes were taken from the CRC Handbook of Chemistry and Physics [29] and vapor pressures, water solubilities and octanol-water partition coefficients were obtained from Ellis and Mabury of the University of Toronto (personal communication).

76

D.A. Ellis et al.

Table 4. Physico-chemical property data for fluoro- and chlorobenzenes used as inputs to the EQC model

Compound

Benzene Monofluorobenzene 1,2,4-Trifluorobenzene Hexafluorobenzene Monochlorobenzene 1,2,4-Trichlorobenzene Hexachlorobenzene a b c

d

Molar mass

Water solubility (g/mol) (g/m3) a

Vapor pressure (Pa) a

Log Melting Assumed 1st order reaction KOW a point half lives (hours) c, d (°C) b

78.1

1780

12700

2.13

96.1

1610

12300

2.82

132.1

961

7670

186.1

120

112.6 181.5 284.8

Water Soil Sediment

209 1480

2960 13300

–42.2

372 1480

2960 13300

3.09

– 5.5

249 4320

8640 38900

2010

1.33

5.3

1500 4320

8640 38900

484

1580

2.8

–45.6

40

61

4.1

16

467 2520

5040 22700

0.0023 5.5

231

951 4320

8640 38900

0.005

5.53

Air

333

444

888

4000

Benzene and chlorobenzene data taken from Mackay et al. [31], fluorobenzene data obtained from Ellis and Mabury, University of Toronto by personal communication. Obtained from CRC handbook of chemistry and physics [29]. Estimated using the Syracuse Research Corporation EPIWIN software [34]. The following EPIWIN setting were used: BIOWIN (Ultimate)/AOP programs were used for estimating air, water, soil and sediment half-lives, the more conservative Alternate BIOWIN half-life values were used, and BIOWIN half-life factors water :soil:sediment were set to 1:2:9. The EPIWIN settings for calculation of half-life values are described in Meylan [34] and are not repeated here. It is noted that the reaction half lives estimated for hexafluorobenzene are conservative as it is expected that the compound will have oxidative and hydrolytic reaction rates markedly greater than hexachlorobenzene in all compartments of the environment.

First-order reaction rate constants were estimated using the Estimations Programs Interface for Windows (EPIWIN) software [34]. EPIWIN is a Windowsbased estimation program commercially available from the Syracuse Research Corporation that can estimate physico-chemical properties and reaction rates from structure alone. Although reaction rate constants for benzene and the chlorobenzenes have previously been estimated by Mackay et al. [31], it was decided, for consistency, to use EPIWIN for estimating the reaction rate constants of half-lives for all the compounds modeled. It must be appreciated that the halflives are subject to considerable error and variability, especially when they exceed a month in duration. For example, the assigned half-lives of 8640 h in soil for HCB and HFB (nearly one year) are highly speculative, although it is likely that they are correct in magnitude. Physico-chemical property and reaction rate data used as input to the EQC model are summarized in Table 4. A summary of results from the EQC Level I and Level II modeling is presented in Table 5. Level I EQC modeling indicates that under equilibrium, and steady state conditions, the monochloro- and trichlorobenzene partition mostly to air,

77

Partitioning of Organofluorine Compounds in the Environment Table 5. Summary of results of Level I and II simulations using the EQC model

Compound

Benzene Monofluorobenzene 1,2,4-Trifluorobenzene Hexafluorobenzene Monochlorobenzene 1,2,4-Trichlorobenzene Hexachlorobenzene a

Level I distribution

Level II

% in Air

% in Water

% in Soil

% in Sediment

99.0

0.88

0.11

0.002

98.9

0.67

0.39

0.009

99.0

0.47

0.51

0.011

99.8

0.16

0.003

0.00007

98.0

1.24

0.70

0.015

81.8

1.47

16.3

8.42

0.32

89.2

BCF a

6.7

Overall % Loss by residence reaction time (days) 3.2

24.9

33

3.5

15.7

62

3.3

21.8

4.0

4.4

32

3.5

17.4

0.36

630

4.4

13.1

1.98

16000

1.1

43

13.7

BCF is the bioconcentration factor deduced as 0.05 KOW , i.e. equilibrium partitioning into fish of 5% lipid.

whereas the hexachlorobenzene partitions mostly to the organic carbon component of soils. Thus each addition of chlorine tends to decrease the volatility and increase the hydrophobicity of the chlorobenzenes. The tendency for the more highly chlorinated, hydrophobic chlorobenzenes to be transported into biota is demonstrated by the bioconcentration factor (BCF) from water to fish for hexachlorobenzene being 2000 times that of benzene. In contrast, increasing the level of fluorination does not markedly increase hydrophobicity; indeed the hexafluorinated compound actually is less hydrophobic than benzene. The fluorobenzenes partition almost exclusively to air, i.e. their partitioning behavior is similar to benzene and monochlorobenzene. Hexafluorobenzene has unusual partitioning properties due to its low hydrophobicity and relatively high vapor pressure, so hexafluorobenzene, unlike its chlorinated analog, does not appreciably partition to soils or bioaccumulate. EQC Level II modeling illustrates the increasing persistence of the chlorobenzenes with increasing levels of chlorination. Increasing half-lives and increasing tendency to partition to soil combine to increase the overall residence time markedly as the degree of chlorination increases. Although the half-lives of the fluorobenzenes similarly increase with increasing degree of fluorination, the overall residence time (or persistence) does not markedly increase because increasing fluorination does not increase the amount partitioning to the soil and sediment compartments in which there is slower chemical degradation. Losses of the fluorobenzenes are mainly controlled by advection in air flowing out of the model world. Although hexafluorobenzene has a longer reaction half-life in air, its overall residence time in the evaluative environment is not much longer than benzene.

78

D.A. Ellis et al.

Table 6. EQC Level III results: chemical amounts in each medium based on single and multiple emissions Compound

Benzene

Monofluorobenzene

Emission medium

Amount at steady state (kg) (percent in brackets) Air

Water

Soil

Sediment

Overall residence time (days)

Air Water Soil “Total”

75,100 (99) 46,500 (15) 73,500 (49) 195,000 (37)

419 (0.55) 259,000 (84) 3,216 (2.2) 263,000 (49)

99.9 (0.13) 61.9 (0.02) 72,600 (49) 72,700 (14)

1.96 (0.003) 1212 (0.40) 15.0 (0.01) 1230 (0.23)

3.2 13 6.2 7.4

Air Water Soil “Total”

84,200 (99) 52,400 (17) 79,700 (27) 216,000 (31)

358 (0.42) 258,000 (82) 2,390 (0.80) 260,000 (37)

336 (0.40) 209 (0.07) 219,000 (73) 219,000 (31)

5.95 (0.007) 4,280 (1.4) 39.8 (0.01) 4,330 (0.62)

3.5 13 13 9.7

78,200 (99) 53,000 (16) 76,300 (21) 207,000 (26)

252 (0.32) 278,000 (82) 1,840 (0.51) 280,000 (36)

406 (0.52) 275 (0.08) 286,000 (79) 287,000 (37)

8.96 (0.01) 9,870 (2.9) 65.5 (0.02) 9,940 (1.3)

3.3 14 15 11

95,600 (100) 65,108 (19) 95,500 (93) 256,000 (47.6)

105 (0.11) 275,000 (81) 651 (0.64) 275,000 (51)

10.4 (0.01) 7.11 (0.002) 6,080 (5.9) 6,090 (1.1)

0.25 (0.0003) 661 (0.19) 1.57 (0.002) 662 (0.12)

4.0 14 4.3 7.5

82,600 (99) 39,500 (16) 62,100 (17) 184,000 (26)

506 (0.61) 204,000 (83) 2,800 (0.75) 207,000 (29)

475 (0.57) 227 (0.09) 310,000 (83) 311,000 (44)

6.3 (0.008) 2,540 (1.0) 34.8 (0.009) 2,580 (0.37)

3.5 10 16 9.8

86,900 (91) 55,300 (12) 34,600 (0.79) 177,000 (3.6)

1,010 (1.0) 280,000 (59) 2,980 (0.07) 284,000 (5.7)

7,140 (7.5) 4,550 (0.96) 4,360,000 (99) 4,370,000 (88)

482 (0.51) 134,000 (28) 1,426 (0.03) 136,000 (2.7)

4.0 20 180 69

92,600 (57) 47,000 (1.1) 2,030 (0.02) 142,000 (0.85)

2,200 (1.4) 294,000 (6.7) 691 (0.006) 297,000 (1.8)

37,200 (23) 18,900 (0.43) 12,200,000(100) 12,200,000 (73)

30,300 (19) 4,050,000 (92) 9,520 (0.08) 4,090,000 (24)

6.8 180 510 233

1,2,4-Trifluorobenzene Air Water Soil “Total” Hexafluorobenzene Air Water Soil “Total” Monochlorobenzene Air Water Soil “Total” 1,2,4-Trichlorobenzene Air Water Soil “Total” Hexachlorobenzene Air Water Soil “Total”

EQC Level III calculations allow non-equilibrium conditions to exist between connected media at steady state. They are useful in determining how media of release affects environmental fate and can identify important transformation and interphase transport processes. Table 6 shows the amount of chemical present in each medium of the EQC model environment, and the chemical persistence at steady state for individual emissions to air, water and soil, as well as a “Total” for simultaneous emissions to each compartment. The “Total” mass balance is the sum of the three individual, single compartment emission mass balances because the system of equations is linear.


79

The Level III EQC results for monochlorobenzene and 1,2,4-trichlorobenzene indicate that emissions tend to remain in the media of release, and are removed from the system by advection or chemical degradation before substantial partitioning to other media takes place. Hexachlorobenzene, which has a lower vapor pressure and higher hydrophobicity, is shown to partition rapidly out of air and water. Emissions to air partition significantly to soil, whereas water emissions partition to sediment. If emitted to soil, the chlorobenzenes are predicted to remain in that medium with only the relatively volatile benzene and monochlorobenzene partitioning significantly to air. The medium of release is also shown to affect greatly the environmental persistence of the chlorobenzenes. Direct emissions to soil lead to the longest environmental residence times because of lack of advection and slow degradation within this medium. Air emissions on the other hand are relatively rapidly removed from the evaluative environment by a combination of advection and more rapid degradation. It should be noted that rapid removal by advection is advantageous from the viewpoint of the receiving environment, but it merely transfers the contaminant to another “downwind” region by long range transport. The substance is thus not persistent locally, but can be persistent globally. This is the case with hexachlorobenzene and probably with hexafluorobenzene. The Level III EQC results for the fluorobenzenes are similar to those for benzene and monochlorobenzene in that a large fraction of emissions tends to remain in the medium of release, although emissions to soil tend to repartition rapidly to air. For example, hexafluorobenzene emissions to soil do not remain

Fig. 12. Diagram showing EQC Level III output for hexachlorobenzene

80

D.A. Ellis et al.

Fig. 13. Diagram showing EQC Level III output for hexafluorobenzene

in the soil because of the compound’s relatively high vapor pressure and very low hydrophobicity. Thus, unlike its chlorinated analog, hexafluorobenzene has a short local environmental residence time irrespective of the medium of discharge. Figures 12 and 13 compare Level III EQC output for hexachloro- and hexafluorobenzene. In summary, the Level I, II and III EQC results indicate that the environmental partitioning and fate of the fluorobenzenes are similar to those of benzene. Substituting fluorines does not cause an increase in hydrophobicity and a reduction in vapor pressure to the same extent as does chlorine substitution. Thus, in terms of environmental partitioning, fluorine substitution on the benzene ring is approximately equivalent to hydrogen substitution. Stability to degrading reactions does, however, increase with addition of fluorines because of the relative strength of the carbon-fluorine bond. The perfluorinated benzene appears to be a special case in that its hydrophobicity is very low, which increases its tendency to partition to air. A similar analysis could be undertaken for the halogenated alkanes and alkenes, resulting in generally similar conclusions that fluorine substitution does not cause a marked increase in hydrophobicity or a marked decrease in volatility. For organic acids such as the carboxylic acids, the substitution of fluorine tends to reduce pKa and increase the degree of ionization, and thus these substances have a greater affinity for the aqueous phase. They are also more stable than their chlorinated analogs as is exemplified by the extreme persistence of trifluoroacetic acid [15].


81

5 Conclusions This brief review of the partitioning properties of organofluorine compounds between the media of air, water and octanol as surrogates for environmental media has shown that fluorine as a substituent for hydrogen confers quite different properties from that of the more widely studied chlorine and bromine. The low atomic mass and a lower volume of fluorine do not cause the consistent increase in hydrophobicity and decrease in volatility induced by chlorine. The substituent effects of fluorine are smaller and generally less predictable, at least using existing structure-properties approaches. In particular, polyfluorinated compounds exhibit unique behavior probably because of significant intramolecular interactions (Sect. 3.1) and minimal intermolecular interactions. The strong electronegativity of fluorine causes appreciable dissociation of carboxylic acids and the strong C–F bond imparts considerable stability to degrading reactions. Evaluative mass balance modeling shows that the partitioning behavior of fluorinated benzenes is remarkably similar to that of benzene by exhibiting appreciable volatility and minimal hydrophobicity. Persistence is, however, greatly increased. Clearly there is a need to improve our knowledge about the partitioning and reactivity properties of this important class of organic compounds in order that their environmental fate and effects can be more fully and accurately evaluated.

6 References 1. Aten AHW Jr (1976) The boiling points of perfluoroparaffins. J Fluorine Chem 8:93–94 2. Banks RE, Smart BE, Tatlow JC (eds) (1994) Organofluorine chemistry principles and commercial applications. Plenum Press, New York 3. Bechalany A, Roethlisberger T, El Tayar N, Testa B (1989) Comparison of various non-polar stationary phases used for assessing lipophilicity. J Chromatogr 473:115–124 4. Benning AF, Park JD (1946) Fluorinated organic compounds. US Pat 2,490,764 5. Berg M, Müller SR, Mühlemann J, Wiedmer A, Schwarzenbach RP (2000) Concentrations and mass fluxes of chloroacetic acids and trifluoroacetic acid in rain and natural waters in Switzerland. Environ Sci Technol 34:2675–2683 6. Booth HS, Swinehart CF (1935) The critical constants and vapor pressures at high pressure of some gaseous fluorides of Group IV. J Am Chem Soc 57:1337–1342 7. Brekke T, Aksnes DW, Sletten E, Stocker M (1988) Solvent shifts and component identification of hydrocarbon mixtures by carbon-13 nuclear magnetic resonance spectrometry. Anal Chem 60:591–596 8. Brown JA (1963) Physical properties of perfluoropropane. J Chem Eng Data 8:106–108 9. Cahill TM, Seiber JN (2000) Regional distribution of trifluoroacetate in surface waters downwind of urban areas in northern California, U.S.A. Environ Sci Technol 34:2909–2912 10. Cahill TM, Thomas CM, Schwarzbach SE, Seiber JN (2001) Accumulation of trifluoroacetate in seasonal wetlands in California. Environ Sci Technol 35:820–825 11. Cotton AF, Wilkinson G (1998) Advanced inorganic chemistry, 5th edn. Wiley, New York, USA 12. Daniewski AR, Dabrowska H, Piasek Z, Urbanski T (1962) Infrared absorption spectra of some urea inclusion compounds. J Chem Soc 2340–2343

82

D.A. Ellis et al.

13. Ellis DA, Mabury SA (2000) The aqueous photolysis of TFM and related trifluoromethylphenols. An alternate source of trifluoroacetic acid in the environment. Environ Sci Technol 34:623–637 14. Ellis DA, Martin JW, Muir DCG, Mabury SA (2000) Development of an F-19 NMR method for the analysis of fluorinated acids in environmental water samples. Anal Chem 72:726– 731 15. Ellis DA, Hanson ML, Sibley PK, Shahid T, Fineberg NA, Solomon KR, Muir DCG, Mabury SA (2001) The fate and persistence of trifluoroacetic and chloroacetic acids in pond waters. Chemosphere 42:309–318 16. Emptage M, Tabinowski J, Odom JM (1997) Effect of fluoroacetates on methanogenesis in samples from selected methanogenic environments. Environ Sci Technol 31:732–734 17. Gago FA,Alvarez-Builla J, Elguero J (1987) Hydrophobicity measurements by HPLC: a new approach to p constants. J Liq Chromatogr 10:1031–47 18. Galska-Krajewska A (1961) Quaternary positive-negative system of acetic acid-pyridinenonane-p-xylene. Bull Acad Pol Sci Ser Sci Chim 9:455–459 19. Garst JE (1984) Accurate, wide-range, automated, high-performance liquid chromatographic method for the estimation of octanol/water partition coefficients. II. Equilibrium in partition coefficient measurements, additivity of substituent constants, and correlation of biological data. J Pharm Sci 73:1623–1629 20. Grosse AV, Wackher RC, Linn CB (1940) Physical properties of the alkyl fluorides and a comparison of the alkyl fluorides with the other alkyl halides and with the alkyls of the elements of period II. J Phys Chem 44:275–296 21. Haberditzl W, Koeppel H (1967) Faraday spectroscopy. I. Magnetooptic rotatory dispersion of organic liquids. Inorg Chem Org Chem Biochem Biophys Biol 22:691–698 22. Haszeldine RN, Smith F (1951) Organic fluorides. VI. The chemical and physical properties of certain fluorocarbons. J Chem Soc 603–608 23. Inel Y, Iseri R (1997) The octanol-water partition coefficient of benzene derivatives based on three dimensional structure directed molecular properties. Chemosphere 35:993–1002 24. Jones LB, Foster JP (1967) The nucleophilicity of chloride ion toward carbonyl carbon. J Org Chem 32:2900–2901 25. Jordan A, Frank H (1999) Trifluoroacetate in the environment. evidence for sources other than HFC/HCFCs. Environ Sci Technol 33:522–527 26. Key BD, Howell RD, Criddle CS (1997) Fluorinated organics in the biosphere. Environ Sci Technol 31:2445–2454 27. Key BD, Howell RD, Criddle CS (1998) Defluorination of organofluorine sulfur compounds by Pseudomonas sp. strain D2. Environ Sci Technol 32:2283–2287 28. Kim BR, Suidan MT, Wallington TJ, Du X (2000) Biodegradability of trifluoroacetic acid. Environ Engin Sci 17:337–342 29. Lide DR (ed) (2000) CRC handbook of chemistry and physics, 81st edn. CRC Press, Boca Raton, FL, US 30. Mackay D, Di Guardo A, Paterson S, Cowan C (1996) Evaluating the environmental fate of a variety of types of chemicals using the EQC model. Environ Toxicol Chem 15:1627–1637 31. Mackay D, Shiu WY, Ma KC (2000) Physical-chemical properties and environmental fate and degradation handbook. CRCnetBASE 2000, Chapman and Hall CRCnetBASE, CRC Press LLC., Boca Raton, FL. (CD-ROM) 32. Mackay D (2001) Multimedia environmental models: the fugacity approach, 2nd edn. Lewis Publishers, Boca Raton, Florida 33. Martin JW, Franklin J, Hanson ML, Solomon KR, Mabury SA, Ellis DA, Scott BF, Muir DCG (2000) Detection of chlorodifluoroacetic acid in precipitation: a possible product of fluorocarbon degradation. Environ Sci Technol 34:274–281 34. Meylan W. (1999) EPIWIN v.3.04 [computer program: USA EPA Version for Windows], Syracuse Research Corporation, Syracuse, NY, US 35. Michels A, Wassenaar T (1948) Vapor pressure of methyl fluoride. Physica, 14:104–110 36. Miller WT Jr, Bergman E, Fainberg AH (1957) Perfluoroalkylzinc compounds. I. The preparation and properties of perfluoroalkylzinc halides. J Am Chem Soc 79:4159–4164


83

37. Mohler FL, Bloom EG, Lengel JH,Wise CE (1949) Mass spectra of some cyclic and straightchain fluorocarbons. J Am Chem Soc 71:337–339 38. Moldavsky DD, Furin GG (1998) The purification of perfluorinated compounds for commercial use. J Fluorine Chem 87:111–121 39. Molina MJ, Rowland RS (1974) Stratospheric sink for chlorofluoromethanes: chlorine atom catalyzed destruction of ozone. Nature 249:810–812 40. Nose A, Kudo T (1990) Studies of reduction of various organic compounds with the nickel(II) chloride-zinc system. Chem Pharm Bull 38:2097–2101 41. Parish WR, Sitton DM (1982) Vapor-liquid equilibrium data for the propane, n-butane, isobutane, and propylene/isopropyl fluoride systems at 30 and 50 °C. J Chem Eng Data 27:303–306 42. Paul S, Chaudhury RT, Bhattacharyya B, Majumdar DK (1972) Suitability of Henderson’s partition function in the computation of heat capacity and compressibility of liquids. Indian J Chem 10:505–508 43. Pedler AE, Smith RC, Tatlow JC (1972) Synthesis and dehydrofluorination of some polyfluoroalkanes. J Fluorine Chem 1:337–345 44. Scott RL (1948) The solubility of fluorocarbons. J Am Chem Soc 70:4090–4093 45. Scott BF, Mactavish D, Spencer C, Strachan WM, Muir DCG (2000) Haloacetic acids in Canadian lake waters and precipitation. Environ Sci Technol 34:4266–4272 46. Serjeant EP, Dempsey B (1979) Ionisation constants of organic acids in aqueous solution. IUPAC Chemical Data Series, No 23. Pergamon Press, Oxford, England 47. Streng AG (1971) Miscibility and compatibility of some liquefied and solidified gases at low temperatures. J Chem Eng Data 16:357–359 48. Takahashi J, Kawabata Y, Yamada K (1965) The utilization of hydrocarbons by microorganisms.V. Screening of yeast for cell production from hydrocarbons and their RNA contents. Agr Biol Chem (Tokyo) 29:292–299 49. Visscher PT, Culbertson CW, Oremland RS (1994) Degradation of trifluoroacetate in oxic and anoxic sediments. Nature 369:729–731 50. Wadsworth WS Jr, Emmons WD (1962) Bicyclic phosphites. J Am Chem Soc 84:610–617 51. Wallington TJ, Schneider WF,Worsnop DR, Nielsen OJ, Sehested J, Debruyn WJ, Shorter JA (1994) The environmental impacts of CFC replacements – HFCs and HCFCs. Environ Sci Technol 28:320A-325A 52. Yalkowsky SH,Valvani SC (1980) Solubility and partitioning. I. Solubility of nonelectrolytes in water. J Pharm Sci 69:912–922