Renewable plant-based soybean oil methyl esters as alternatives to organic solvents

Abstract

The physico-chemical properties of soybean oil methyl ester (SBME), better known as biodiesel, of importance to its use as a solvent in liquid–liquid separations have been examined. Partition coefficients of several organic species between SBME–water have been determined and compared to log P (1-octanol–water). The free energy of transfer of a methylene group has been obtained and the solvent properties of the SBME–water system determined from distribution data of a small solute set using Abraham's generalized solvation equation. Solute distribution behavior is similar to that found for conventional organic solvent–water systems, but is most similar to other vegetable oils such as olive oil. When ionizable solutes are partitioned in the SBME–water system at differing pH, the neutral species show the highest distribution. Partitioning is dependent on the solute's ability to form hydrogen bonds between water and its charged state. Metal ions (e.g., Fe³⁺, Co²⁺, and Ni²⁺) exhibit moderate partitioning to the SBME phase from water only in the presence of extractants. Actinides (UO₂²⁺, Am³⁺) exhibit significant partitioning to the SBME from aqueous solutions with the use of octyl(phenyl)-N,N-diisobutylcarbamoylmethyl phosphine oxide (CMPO). Soybean oil methyl ester may be a suitable “green” alternative for the replacement of volatile organic solvents in liquid–liquid extractions in selected applications.

---

Introduction

An important component of green chemistry is the search for new solvents with which to replace volatile organic compounds (VOCs). Ideally, such solvents should be of low toxicity, low volatility, and derived from renewable resources.¹ We have a long standing interest in the design and development of environmentally benign solvents for liquid–liquid extraction.^2–4 One possible route to renewable solvents having a ‘green’ character with some potential to address pollution prevention/remediation concerns, may be to base the starting materials upon agrochemicals rather than petroleum based raw materials.^5,6 Here, we examine the solvent properties of transesterified soybean oil, better known as biodiesel, as related to liquid–liquid partition from water.

World production of vegetable oils is recorded at 88.7 million metric tons, with over 10 million metric tons of vegetable oils produced in the United States.⁷ U. S. oilseed crops include soybean (by far the predominant seed oil crop), corn, cotton, sunflower, flax, and rapeseed (canola). The major use for vegetable oils is, obviously, in food products, however, some vegetable oils and their ester derivatives are finding increasing industrial usage. Two important uses of transesterified soybean oil are in soybean oil inks and in diesel fuel.⁸ Also, vegetable oils have recently found use as a cheap, non-toxic solvent for the synthesis of nanoparticles⁹ and quantum dots.¹⁰ Regulatory pressure to reduce VOC emissions is leading to an increasing use of such solvents in the coatings industry.¹¹

Soybean oil methyl esters (SBMEs, e.g., SoyGold^® 1000) are composed of a mixture of fatty acids (Table 1). Palmitic and stearic acid are saturated fatty acids and comprise approximately 15% of the final mixture. The remaining three major fatty acids, oleic acid, linoleic acid, and linolenic acid have varying degrees of unsaturation and comprise approximately 83% of the final mixture. By far the major fatty acid present in soybean oil is linoleic acid. Linoleic acid is an 18-carbon long-chain fatty acid with double bonds at the 9- and 12-carbon atoms.

Table 1 Major fatty acids (%) and ratio of # of carbons to carbon–carbon double bonds in soybean oil^ab

Ratio (name)	16 : 0 (Palmitic)	18 : 0 (Stearic)	18 : 1 (Oleic)	18 : 2 (Linoleic)	18 : 3 (Linolenic)	Other
a From data compiled in ref. 12. b Cis-unsaturation.
%	11	4	22	53	7.5	2.5
Structure

---

SBME is produced by the transesterification of soybean oil and methanol. It is non-toxic, non-hazardous, and biodegradable. It is neither a hazardous air pollutant (HAP), nor an ozone-depleting chemical (ODC).

Transesterification chemistry (used in the production of SoyGold^® 1000) is a more modern innovation, and involves the combination of organically derived oils or fats with a small chain alcohol such as methanol or ethanol in the presence of a catalyst (most often the catalyst is a base) to form fatty acid esters. This process is akin to saponification chemistry, the alkaline hydrolysis of fatty acid esters, in which triglycerides are reacted with sodium or potassium hydroxide to produce glycerol and fatty acid salts (soap). SBME has a high flash point (218 °C) and low vapor pressure (<2 mm Hg).¹³

Materials such as SBME may have an important role to play in replacing the previous generation of solvents with materials of a more benign nature derived from renewable resources. However, there is a need to understand the physico-chemical properties of these materials before they can find widespread application. The solvent property data we report here is intended to inform that understanding, and begin to place SBME in context with other common solvents.

Experimental

Reagents and tracers

SoyGold^® 1000, was used as received from Ag Environmental Products L.L.C. (Lenexa, KS, USA). Extractants 1-(2-pyridylazo)-2-naphthol, PAN, and (1-thiazolylazo)-2-naphthol, TAN, were obtained from Lancaster Synthesis, Inc. (Windham, NH, USA) and used as received. Octyl(phenyl)-N,N-diisobutylcarbamoylmethyl phosphine oxide, CMPO, was obtained from Strem Chemicals (Newburyport, MA, USA) and used without further purification. All water was purified using a Barnstead (Dubuque, IA, USA) commercial deionization and polishing system. All other chemicals were of reagent grade.

The ¹⁴C-labeled organic solutes were purchased from Sigma (St. Louis, MO, USA). Upon receipt, the tracers were diluted to an activity of ca. 0.06–0.08 µCi µL⁻¹ for use as the ‘spike’ in the partitioning experiments. The hydrophilic tracers were diluted in deionized water and the hydrophobic tracers were diluted in their unlabelled form.

The cobalt tracer (as ⁶⁰CoCl₂ in 0.5 M HCl) was obtained from New England Nuclear (Boston, MA, USA). The iron tracer (as ⁵⁹FeCl₃ in 0.1 M HCl) and the nickel tracer (as ⁶³NiCl₂ in 0.1 M HCl) were both obtained from Amersham Pharmacia Biotech (Piscataway, NJ, USA). The actinide tracers (²³³UO₂Cl₂ and ²⁴¹AmCl₃) were obtained from Isotope Products Laboratories (Valencia, CA, USA). All metal ion radiotracers, except for ²³³U, were used as received for partitioning studies, or diluted with deionized water to an activity of ca. 0.03 µCi µL⁻¹.

Stock solutions of the uranyl nitrate radiotracer solution were cleaned of impurities (²²⁸Th and its daughters from decay of ²³²U impurities) using TEVA resin (EIChrom Industries, Darien, IL, USA).¹⁴ 20 µL of ²³³UO₂²⁺ stock solution was diluted in 10 mL of 5 M HCl and the solution was added to a column comprised of 1 g TEVA resin preconditioned with 5 M HCl. After loading, 20 mL of 0.5 M HCl was added to the column to elute ²³³U, which was collected, evaporated to dryness and dissolved in 0.1 M HNO₃.

Water content

The equilibrium water content of SBME was determined using a volumetric Aquastar Karl Fischer titrator (EM Science, Gibbstown, NJ, USA) with Composite 5 solution as the titrant and anhydrous methanol as the solvent. Each sample was at least 0.1 g and triplicate measurements were performed on each sample.

Radioanalytical measurements

Cobalt-60, iron-59, nickel-63, and americium-241 were analyzed by their characteristic gamma ray emission with a Packard Cobra II Auto-Gamma counting system (Packard Instrument Co., Inc., Meriden, CT, USA). Carbon-14 and uranium-233 activities were followed by beta decay analysis using Ultima Gold scintillation cocktail and a Packard Tri-Carb 1900 TR Liquid Scintillation Analyzer (Packard Instrument Co., Inc., Meriden, CT, USA).

Partitioning studies

Equal volumes of water and SoyGold^® were added to form the biphasic system studied. In liquid–liquid partition studies, the organic and metal ion species distribution ratios were determined radiometrically using previously described methods.¹⁵ For each determination, equal aliquots of SoyGold^® and water were mixed. Each system was equilibrated by vortexing for 2 min, followed by 2 min of centrifugation (2000 g). A tracer quantity of organic or metal ion species was then added to the two-phase system. The system was then twice vortexed for 2 min followed by 2 min of centrifugation (2000 g). Equal aliquots of each phase were then removed for radioanalysis. The distribution ratio (D) was calculated by the ratio of the activity in the top SoyGold^® phase divided by the activity in the lower aqueous phase.

The distribution ratios reported here are the average of at least two measurements and are typically accurate to ±5%. Where necessary, the aqueous phase pH was measured using a Fisher Accumet pH meter and adjusted to acidic conditions using sulfuric acid or to basic conditions using sodium hydroxide. Multiple linear regression analyses were performed using Stat Box v2.5 (Grimmer Logiciels, 34 Rue de Dunkerque, Paris, France).

Solvent characterization

The SBME used in this study (SoyGold^® 1000) is identified as methyl soyate with the CAS registry number 67784-80-9.¹³ It can be described as a pale yellow liquid with a mild odor. SBME is insoluble in water and is incompatible with strong oxidizing agents. While no specific fire hazard problems during shipment have been reported, the possibility of spontaneous combustion (“oily rag” problem) is inherent. With a boiling point over 315 °C at 760 mm Hg and a specific gravity of 0.882 g mL⁻¹ at 25 °C, SBME has most commonly been used as ‘biodiesel’ in the United States. In the following discussion, we address the effectiveness of SBMEs as environmentally-friendly, alternative organic diluents.

Results and discussion

Organic solute partitioning

Partitioning data for 20 solutes partitioned as the neutral species in an SBME–water biphasic system are shown in Fig. 1, where they are compared to published values of these solutes' octanol–water partition coefficient (log P). As is to be expected, there is a significant relationship between the octanol–water partition coefficient and the partition coefficient in the SBME–water system. However, there are also, again as is to be expected, significant differences in detail.

Fig. 1 Correlation of partitioning data between SBME–water and 1-octanol–water biphasic systems.

The relationship between the octanol–water partition coefficient and the distribution in the SBME–water system can be summarized in the form of eqn (1):

Log D (SBME) = b + m(log P) (1)

where log D (SBME) is the distribution in the SBME–water system, log P is the distribution in the 1-octanol–water system, and b and m are constants. Here, b = −0.69 and m = 1.04 (R² = 0.88). Thus, the distribution coefficient in the SBME–water system is in general somewhat less than one order of magnitude lower (ca. 0.7 log) than that in the 1-octanol–water system, and a 1-octanol–water partition coefficient of about 0.7 marks the transition between water and oil phase preference in the SBME–water system.

Determination of ΔG_CH2

ΔG_CH2 is generally recognized as a measure of hydrophobicity related to the relative free energy of cavity formation to accommodate a solute in each of the two phases in a l–l biphasic system.^17,18 As a result of the limited molecular interactions possible for solutes, it is characterized here by an incremental increase in methylene groups. However, solutes differ in their relative hydrophobicity, and this arises from differences in the total of all solute–solvent interactions possible in each of the phases. Such complexity is not always captured by relationships to 1-octanol–water partitioning or by quantities such as the free energy of transfer of a methylene group, and thus it is usually necessary to measure several such properties for any given system. Here we discuss our determinations of ΔG_CH2 for the SBME–water system.

Fig. 2 shows the distribution of several n-alcohols in relation to their alkyl chain length in the SBME–water system, as well as literature data for their distribution in an olive oil–water system¹⁶ and the conventional 1-octanol–water system. These relationships can be described by eqn (2):

ln K = C + En_c (2)

where K is the partition coefficient, C is a constant related to the hydration properties of the phases, n_c refers to the number of carbons in the alkyl chain of the solutes partitioned, and E is a constant defined as the average ln K increment per CH₂ group.^17–19 From this data the free energy of transfer of a methylene group (ΔG_CH2) may be obtained from eqn (3) in kcal mol⁻¹:

ΔG_CH2 = −RTE (3)

where R is the universal gas constant, T is the absolute temperature, and E is the slope of the line in Fig. 2 (plotted from eqn (2)).^17–19

Fig. 2 Distribution of short chain alcohols in the SBME–water (▲) system compared to distribution in the olive oil–water (▼) and 1-octanol–water (●) systems.

Table 2 presents ΔG_CH2 values for various solvent systems to allow comparison with the present SBME system and with the olive oil–water system. Also included in Table 2 are ΔG_CH2 values for selected ionic liquid (IL) salt–salt²⁰ and polyethylene glycol (PEG)–salt aqueous biphasic systems (ABS)⁴ that have been studied in our laboratory. From Table 2, one notes that based on the ΔG_CH2 for the various solvents listed, SBME is very near to that observed for xylene and would suggest that SBME is a likely alternative to xylene as a solvent. Indeed, Hendrickson et al. have discovered that SBME is a suitable replacement for xylene as a developing solvent for photopolymerizable printing plates.²¹

Table 2 Free energy of transfer of a methylene group for various solvent–water systems

Solvent–water system	−ΔGCH2/kcal mol^−1	Ref.	Solvent–water system	−ΔGCH2/kcal mol^−1	Ref.
a Present work. b Calculated in present work with data from Ref. 16.
Hexane	1.10	18	Diisopropyl ether	0.68	18
Chloroform	0.85	18	SBME	0.67	—^a
Benzene	0.84	18	Xylene	0.64	18
Octanol	0.79	18	Olive oil	0.61	—^b
Octane	0.77	18	n-Butanol	0.54	18
Dodecane	0.77	18	Methyl ethyl ketone (MEK)	0.43	18
Methyl isobutyl ketone (MIBK)	0.72	18	PEG–salt (general)	0.142–0.696	4
IL salt–salt : 1-butyl-3-methylimidazolium chloride/K3PO4 or K2HPO4 or K2CO3	0.304–0.728	20	PEG-2000 (40% w/w)/(NH4)2SO4 (1.68 M)	0.214	4

---

Table 2 and Fig. 2 illustrate the similarity between the solvent properties of the currently investigated SBME–water system and literature data for the olive oil–water system. The free energy of transfer between the phases of the SBME system is much less than that typical of solvents dominated by van der Waals forces and much more similar to solvents typified by molecular interactions or by a significant water content at equilibrium. However, the equilibrium water content of the SBME phase is relatively low (0.35% wt/wt), and since it might be expected that the predominant intermolecular forces in an SBME phase would be dispersion forces, it is possible that molecular packing features contribute to the relative weakness of cavity forces in the olive oil and SBME phases implied by the relatively low free energy of CH₂ transfer value obtained in these systems.

Linear free energy relationships

The complexity of solute–solvent interactions is generally assumed to be encapsulated in some form of a generalized solvation equation which can be represented by eqn (4):⁴

Some Property (SP) = cavity terms + polarity terms + hydrogen bonding terms + constant (4)

Linear free energy relationships (LFER) based upon the generalized solvation equation have been widely used to model many processes, such as partitioning in aqueous–organic systems, solubility, and transport across biological membranes.^22–24 Abraham's generalized solvation equation, eqn (5):

log SP = c + rR₂ + sπ₂^H + a∑α₂^H + b∑β₂^H + vV_x (5)

has been used to characterize solvation in a wide variety of solvent systems including the description of partition in various aqueous–organic systems^25–27 and aqueous–micellar systems.^28,29 We have previously used Abraham's approach to characterize solute distribution in a PEG-2000/(NH₄)₂SO₄ ABS⁴ and for ionic liquids (ILs).³⁰

In essence, the log of SP, in this case the partition coefficient of a series of solutes in a given system, can be related to the solute property descriptors of each solute through eqn (5). The solute distribution ratios for a very limited set of solutes determined in the SBME–water system and their corresponding Gibbs's energy related solute property descriptors are shown in Table 3.^24,26,31

Table 3 Abraham's property descriptors and distribution coefficients in SBME–water of various solutes

Solute	Abraham parameters						log D
	π 2 ^H	∑α2^H	∑β2^H	Vx	R 2	Ref.
1,2,4-Trichlorobenzene	0.081	0	0	1.0836	0.98	23	2.99
1,4-Dichlorobenzene	0.75	0	0.02	0.9612	0.825	23	2.75
Chlorobenzene	0.65	0	0.07	0.8388	0.718	23	2.41
Toluene	0.52	0	0.14	0.8573	0.601	23	2.33
p-Toluic acid	0.90	0.60	0.38	1.073	0.73	25	1.28
Salicyclic acid	0.84	0.71	0.38	0.9904	0.89	26	1.39
Benzene	0.52	0	0.14	0.7164	0.61	23	2.23
Acetophenone	1.01	0	0.49	1.0139	0.818	23	1.54
Methyl iodide	0.43	0	0.13	0.5077	0.676	23	1.27
Pentanol	0.42	0.37	0.48	0.8718	0.219	23	0.52
Butanol	0.42	0.37	0.48	0.7309	0.224	23	−0.096
Phthalic acid	1.60	0.82	0.75	1.147	0.85	26	−1.24
Propylamine	0.35	0.16	0.61	0.631	0.225	23	0.41
Propanol	0.42	0.37	0.48	0.59	0.236	23	−0.39
Isopropanol	0.36	0.33	0.56	0.59	0.212	23	−0.87
Acetic acid	0.65	0.61	0.45	0.4648	0.265	23	−1.26
Ethanol	0.42	0.37	0.48	0.4491	0.246	23	−1.07
Acetonitrile	0.90	0.04	0.33	0.4042	0.237	23	−0.81
Methanol	0.44	0.43	0.47	0.3082	0.278	23	−1.44

---

The descriptors are the McGowan volume (V_x), the excess molar refraction (R₂), the solute dipolarity–polarizability (π₂^H), and the solutes' overall hydrogen-bond acceptor basicity and hydrogen-bond donor acidity, ∑β₂^H and ∑α₂^H, respectively. The sign and magnitude of the regression coefficients obtained from eqn (5) (r, s, a, b, and v), by multiple linear regression of the solute descriptors on the log of the partition coefficient, may be considered proportional to the solvent properties of the phases corresponding to the appropriate solute descriptor. Thus, r corresponds to the relative strengths of the solute–solvent interactions determined by the excess molar refractivity of the solute, and a corresponds to the relative solvent hydrogen-bond basicity of the phases, and so on. Finally, c is a constant of proportionality.

The use of as few as twenty solutes in a multiple linear regression is unlikely to give very satisfactory results, and thus the results we report can only be considered a preliminary characterization of the solvent properties of the SBME–water partitioning system. Nevertheless, Table 4 shows the relative contributions of the regression coefficients to the partitioning of the solutes in the SBME–water system. Only solute acidity, solute basicity, and solute volume are significant at the 95% level. The most significant factor is solute basicity, for which the regression coefficient is large and negative. The SBME phase is significantly less acidic than the aqueous phase and basic (in the sense of Lewis) solutes have a strong preference for the aqueous phase.

Table 4 Relative contributions of the coefficients of the generalized solvation equation^a

Variable	Coefficient	Correlation/Y	Student's t	Probability
a Adjusted R^2 = 0.90; multiple R^2 = 0.96; F statistic = 55.34
a	−1.67	−0.61	2.87	0.0117
b	−3.61	−0.80	4.80	0.0002
v	3.20	0.59	6.98	4.41× 10^−6
Constant	−0.012

---

The next most significant parameter is volume, which is comparatively small and (typical of solvent–water partitioning) positive. The v parameter may be considered to be a measure of the relative hydrophobicity of the system and is similar to ΔG_CH2 in that it reflects the difference in the free energy required for cavity formation between the water and SBME phases. Typically, this parameter is large in solvent–water systems (Table 4) reflecting the fact that for most solvents, there are few intermolecular forces hindering cavity formation other than dispersion forces, whereas the free energy of cavity formation in aqueous phases is dominated by its extensively hydrogen bonded nature. There is a significant penalty to cavity formation in the aqueous phase due to the structuredness of water. However, the magnitude of this parameter is not as great as for most pure van der Waals solvents, which usually indicates either there is significant structure to the SBME phase, or that (and of course it amounts to the same thing) there is significant water present in the organic phase at equilibrium; for example, in the 1-octanol–water system. However, in the case of SBME, we have measured an equilibrium water content of only 0.35% wt/wt. It may be thought in the case of olive oil and SBME that molecular shape features play a role in determining the magnitude of this parameter.

Finally, the a coefficient is relatively small and negative, suggesting that the oil phase is less basic than the aqueous phase, it is also the least significant factor of those which are significant at the 95% confidence level. Again, it should be remembered that the solute set is small and the above interpretation tentative.

The full solvation equation for the 20 solute SBME–water partition is given in eqn (6):

Log D_(SBME) = −0.11 − 0.78π₂^H − 1.58∑α₂^H − 2.90∑β₂^H + 3.19V_x − 0.59R₂ F = 35.5 (6)

The coefficients are slightly different from those shown in Table 4 because in Table 4 only three terms are used (a, b, and v), and in eqn (6) all 5 terms have been used (a, b, v, s, and r). Both equations give the same result for Log D within statistical limits; however, the effect of the extra terms is to change the values of the terms in the first equation. The effect ought to be small because the extra terms are of small statistical significance, nevertheless, they do have an effect.

These results may be directly compared to a similar LSER derived for the olive oil–water system²⁷ shown in eqn (7).

Log D_(OO) = −0.011 − 0.8π₂^H − 1.47∑α₂^H − 4.92∑β₂^H + 4.17V_x + 0.58R₂ F = 5841 (7)

The similarity in the roles of solute acidity, basicity, and volume in the SBME–water systems and the olive oil–water system is readily apparent. Surprisingly, even terms which were least significant in the regression of eqn (6), follow the olive oil–water regression quite closely.

Partitioning vs. pH

The partition of several ionizable organic species was examined in relation to changes in the aqueous phase pH in the SBME–water system. Quite frequently in chemical processing by liquid–liquid extraction, recourse is to perform forward (partition into the extracting organic phase) and back (partition into a fresh aqueous phase) partitioning to effect purification of particular species (e.g., in the liquid–liquid extraction of penicillin).³²Fig. 3 shows partitioning data for several ionizable organic compounds in SBME–water systems. The data confirm that distribution is higher for the uncharged form than for the charged form. For acidic species this means that there is a relative preference for the organic (SBME) phase at low pH and a relative preference for the aqueous phase at high pH. This phase preference is reversed where the ionizable group is basic. For the relatively weak acids and the base shown in Fig. 3, a forward and back extraction of these species is possible in principle, although there may still be mass transfer concerns to be addressed.

Fig. 3 pH dependent partitioning of ionizable solutes: 1 acetic acid (pK_a = 4.75) ; 2 phthalic acid (pK_a1 = 2.89, pK_a2 = 5.51); 3 propylamine (pK_a = 10.71); 4 salicylic acid (pK_a1 = 2.97, pK_a2 = 13.40); 5 p-toluic acid (pK_a = 4.36).

Metal ion partitioning

In general, the distribution and extraction of metal ion species into organic phases is limited by the requirement to solvate (hydrate) the metal ion in the organic phase.³³ This usually can only be achieved by the use of an extractant which can replace the metal ion hydration layer and provide a molecular surface compatible with solvation in the organic phase. Fig. 4 illustrates the partition of several metal ion species in the SBME–water system and demonstrates that this system is no exception. In all cases the metal ion species partitions strongly to the aqueous phase unless an extractant is included in the system.

Fig. 4 The distribution of metal ions (Co²⁺, Fe³⁺, Ni²⁺) in the absence and presence of PAN and TAN (each at 0.1 mM) vs. pH.

For the transition metal ions (Co²⁺, Ni²⁺, Fe³⁺) examined, no useful extraction is achieved in the absence of an extractant. In the presence of the extractants PAN and TAN, the distribution coefficient of all the metal ions increases, except at the lowest pH. This increase is pH dependent, being highest under the most basic conditions. This is commonly found in the extraction of metal ion species by solvent extraction and is due to the increase in the complexation constant with pH and forms a useful technical means of metal ion purification. Extractants and organic phases may be loaded at high pH and unloaded at low pH in a similar forward and back extraction process as mentioned above for the extraction of organic acids and bases. Additionally, some selectivity in metal ion separation may be secured by this means. In the SBME–water system, there appears to be little difference between the performance of PAN and TAN which are most effective in the extraction of Co²⁺, but appear not practically useful for the extraction of Ni²⁺ and Fe³⁺. Undoubtedly, suitable extractants could be found for the extraction of these and other metal ion species using the SBME–water system.

In the TRUEX process for the extraction of uranium and transuranic ions, 0.2 M CMPO is used as the extractant, 1.2–1.4 M tri-n-butylphosphate (TBP) as a phase modifier, and the diluent is paraffinic hydrocarbons.³⁴ Schultz and Horwitz,³⁴ using the TRUEX formulation, reported the highest distribution values for UO₂²⁺ and Am³⁺ to be 10³ and 10, respectively, at 6 M HNO₃. Mathur et al., using dodecane as the diluent, 0.2 M CMPO and 1.2 M TBP, reported distribution ratios of 10² for UO₂²⁺ and 10 for Am³⁺ at 1–6 M HNO₃.³⁵ We have examined the partitioning of UO₂²⁺ (Fig. 5) and Am³⁺ (Fig. 6) in a similar system but with SBME as a replacement for the paraffinic hydrocarbon diluent. Without CMPO, the distribution ratios fall well below 1, indicating the preference for the aqueous phase. For the uranyl cation, addition of 0.2 M CMPO significantly improves the distribution ratios up to 10² at all nitric acid and NaNO₃ concentrations studied.

Fig. 5 The distribution of UO₂²⁺ in the SBME–water system without extractant (●), with 0.2 M CMPO (■), and with 0.2 M CMPO–1.2 M TBP (▲) vs. HNO₃ (filled symbols, solid lines) or NaNO₃ (open symbols, dashed lines). Distribution ratio from SBME–water = 0.0044.

Fig. 6 The distribution of Am³⁺ in the SBME–water system without extractant (●), with 0.2 M CMPO (■), and with 0.2 M CMPO–1.2 M TBP (▲) vs. HNO₃ (filled symbols, solid lines) or NaNO₃ (open symbols, dashed lines). Distribution ratio from SBME–water = 0.0082.

As with the uranyl cation, the distribution ratios for Am³⁺ are well below 1 and although there is an increase in preference for the SBME phase with the addition of CMPO, the distribution ratios are still below 1 and favor the aqueous phase up to 0.1 M NO₃⁻, where they then begin to increase above 1 and favor the SBME phase approaching 10² at 5 M NO₃⁻. This behavior would allow the separation of UO₂²⁺ from Am³⁺ in this system.

The inclusion of 1.2 M TBP for either UO₂²⁺ or Am³⁺ only made small improvements in the distribution ratios and does not appear to be needed as a phase modifier in the SBME–water systems as observed for traditional dodecane–water systems;^34,35 the distribution values remain comparable without its use. The elimination of TBP could simplify and reduce the costs of these systems.

Conclusions

We have examined the partitioning of neutral, ionizable, and metal ion solutes in biphasic systems formed with soybean oil methyl ester and water. For neutral solutes, distribution behavior is similar to comparative solvent–water systems and is most similar to the olive oil–water system as judged by application of the Abraham generalized solvation equation. For the ionizable solutes, distribution behavior in this system is similar to the behavior of similar solutes in the presence of organic diluents in traditional solvent extraction. Extraction of metal ions using SBME is dependent on the presence of coordinating extractants to carry metal ions into the SBME phase. Such behavior is comparable to traditional organic solvent extraction. Soybean oil methyl esters thus represent a class of solvents that can be considered as a medium for the development of liquid–liquid extraction processes. Solute behavior during extraction is comparable to that found for traditional solvents, thus current models and process practices may, in many cases, be easily adapted to the use of this solvent–water system.

Acknowledgements

This research was supported by the Division of Chemical Sciences, Geosciences, and Biosciences, Office of Basic Energy Research, U. S. Department of Energy (Grant DE-FG02-96ER14673). The authors wish to thank Ag Environmental Products of Lenexa, Kansas for the gracious gift of the soybean oil methyl ester, SoyGold^® 1000, used in this study.

References

1. P. T. Anastas, Green Chemistry as Applied to Solvents, in Clean Solvents: Alternative Media for Chemical Reactions and Processing, ed. M. A. Abraham, L. Moens, ACS Symposium Series 819, American Chemical Society, Washington, DC, 2002, pp. 1–9.
2. J. G. Huddleston, H. D. Willauer, R. P. Swatloski, A. E. Visser and R. D. Rogers, Room Temperature Ionic Liquids as Novel Media for ‘Clean’ Liquid–Liquid Extraction, Chem. Commun., 1998, 16, 1765–1766.
3. S. K. Spear, A. E. Visser, H. D. Willauer, R. P. Swatloski, S. T. Griffin, J. G. Huddleston and R. D. Rogers, Green Separation Science & Technology: Replacement of Volatile Organic Compounds in Industrial Scale Liquid/Liquid or Chromatographic Separations, in Green Chemistry and Engineering, ed. P. T. Anastas, L. G. Heine, T. C. Williamson, ACS Symposia Series 766, American Chemical Society, Washington, DC, 2001, pp. 206–221.
4. H. D. Willauer, J. G. Huddleston and R. D. Rogers, Solvent Properties of Aqueous Biphasic Systems Composed of Polyethylene Glycol and Salt Characterized by the Free Energy of Transfer of a Methylene Group between the Phases and by a Linear Solvation Energy Relationship, Ind. Eng. Chem. Res., 2002, 41, 2591–2601.
5. N. E. Kob, Dibasic Ester: A Low Risk, Green Organic Solvent Alternative, in Clean Solvents: Alternative Media for Chemical Reactions and Processing, ed. M. A. Abraham, L. Moens, ACS Symposium Series 819, American Chemical Society, Washington, DC, 2002, pp. 238–253.
6. L. R. Lynd, C. E. Wyman and T. U. Gerngross, Biocommodity Engineering, Biotechnol. Prog., 1999, 15, 777–793.
7. Oil Crops Situation and Yearbook Outlook, Market and Trade Economics Division, Economic Research Service, U. S. Department of Agriculture, October 2001, OCS–2001.
8. M. O. Bagby, Products from Vegetable Oils: Two Examples, in Agricultural Materials As Renewable Resources: Nonfood and Industrial Applications, ed. G. Fuller, T. A. McKeon, D. D. Bills, ACS Symposium Series 647, American Chemical Society, Washington, DC, 1996, pp. 248–257.
9. Y. Koltypin, N. Perkas and A. Gedanken, Commercial Edible Oils as New Solvents for Ultrasonic Synthesis of Nanoparticles: The Preparation of Air Stable Nanocrystalline Iron Particles, J. Mater. Chem., 2004, 14, 2975–2977.
10. S. Sapra, A. L. Rogach and J. Feldman, Phosphine-Free Synthesis of Monodisperse CdSe Nanocrystals in Olive Oil, J. Mater. Chem., 2006, 16, 3391–3395.
11. B. Lavers, A Fresh Look at Solvents – Vegetable vs. Mineral Oils, Asia Pac. Coat. J., 2000, 13, 14–18.
12. The Lipid Handbook, ed. F. D. Gunstone, J. L. Harwood and F. B. Padley, Chapman & Hall, London, 2nd edn, 1994, pp. 53–146.
13. Ag Environmental Products, L. L. C. 2001, Material Safety Data Sheet, http://www.soygold.com/products/solvents/SG1000/SG1000_MSDS.pdf, last accessed 02/05/2007.
14. E. P. Horwitz, M. L. Dietz, R. Chiarizia, H. Diamond, S. L. Maxwell, III and M. R. Nelson, Separation and Preconcentration of Actinides by Extraction Chromatography using a Supported Liquid Anion Exchanger: Application to the Characterization of High-Level Nuclear Waste Solutions, Anal. Chim. Acta, 1995, 310, 63–78.
15. J. G. Huddleston, S. T. Griffin, J. Zhang, H. D. Willauer and R. D. Rogers, in Methods in Biotechnology, Vol. 2: Bioremediation Protocols, ed. D. Sheehan, Humana Press, Totowa, NJ, 1998, pp. 1–17.
16. T. Kaneko, P.-Y. Wang and A. Sato, Partition Coefficients of Some Acetate Esters and Alcohols in Water, Blood, Olive Oil and Rat Tissues, Occup. Environ. Med., 1994, 51, 68–72.
17. B. Y. Zaslavsky, L. M. Miheeva, N. N. Mestechkina and S. V. Rogozhin, Physico-Chemical Factors Governing Partition Behaviour of Solutes and Particles in Aqueous Polymeric Biphasic Systems: I. Effect of Ionic Composition on the Relative Hydrophobicity of the Phases, J. Chromatogr., 1982, 253, 139–148.
18. B. Y. Zaslavsky, L. M. Miheeva and S. V. Rogozhin, Parameterization of Hydrophobic Properties of Aqueous Polymeric Biphasic Systems and Water-Organic Solvent Systems, J. Chromatogr., 1981, 212, 13–22.
19. B. Y. Zaslavsky, N. D. Gulaeva, E. A. Djafarov, E. A. Masimov and L. M. Miheeva, Phase Separation in Aqueous Poly(ethylene glycol)-(NH₄)₂SO₄ Systems and Some Physicochemical Properties of the Phases, J. Colloid Interface Sci., 1990, 137, 147–156.
20. N. J. Bridges, K. E. Gutowski and R. D. Rogers, Investigation of Aqueous Biphasic Systems Formed from Solutions of Chaotropic Salts with Kosmotropic Salts (Salt-Salt ABS), Green Chem., 2007, 9, 177–183.
21. C. M. Hendrickson and D. C. Bradford, Developing Solvent for Photopolymerizable Printing Plates. US Pat. 6 582 886, 2003.
22. M. H. Abraham, H. S. Chadha, F. Martins, R. C. Mitchell, M. W. Bradbury and J. A. Gratton, Hydrogen Bonding. Part 46. A Review of the Correlation and Prediction of Transport Properties by an LFER Method: Physicochemical Properties, Brain Penetration and Skin Permeation, Pestic. Sci., 1999, 55, 78–88.
23. J. A. Platts, M. H. Abraham, D. Butina and A. Hersey, Estimation of Molecular Linear Free Energy Relationship Descriptors by a Group Contribution Approach. 2. Prediction of Partition Coefficients, J. Chem. Inf. Comput. Sci., 2000, 40, 71–80.
24. M. H. Abraham, J. Andonian-Haftvan, G. S. Whiting, A. Leo and R. S. Taft, Hydrogen Bonding. Part 34. The Factors that Influence the Solubility of Gases and Vapours in Water at 298 K, and a New Method for its Determination, J. Chem. Soc., Perkin Trans. 2, 1994, 1777–1791.
25. A. Pagliara, G. Caron, G. Lisa, F. Weizheng, P. Gaillard, P.-A. Carrupt, B. Testa and M. H. Abraham, Solvatochromic Analysis of Di-n-butyl ether/Water Partition Coefficients as Compared to Other Solvent Systems, J. Chem. Soc., Perkin Trans. 2, 1997, 2639–2643.
26. M. H. Abraham, H. S. Chadha, G. S. Whitingand and R. C. Mitchell, Hydrogen Bonding. 32. An Analysis of Water-Octanol and Water-Alkane Partitioning and the ΔLog P Parameter of Seiler, J. Pharm. Sci., 1994, 83, 1085–1100.
27. M. H. Abraham, C. E. Green and J. A. Platts, University College London, London, Private communication, 2000.
28. F. H. Quina, E. O. Alonso and J. P. S. Farah, Incorporation of Nonionic Solutes into Aqueous Micelles: A Linear Solvation Free Energy Relationship Analysis, J. Phys. Chem., 1995, 99, 11708–11714.
29. M. F. Vitha and P. W. Carr, An LSER Study of the Effects of Surfactant Chain Length on the Chemical Interactions Governing Retention and Selectivity in Micellar Electrokinetic Capillary Chromatography using Sodium Alkyl Sulfate Elution Buffers, Sep. Sci. Technol., 1998, 33, 2075–2100.
30. J. G. Huddleston, G. A. Broker, H. D. Willauer and R. D. Rogers, Free-Energy Relationships and Solvatochromatic Properties of 1-Alkyl-3-Methylimidazolium Ionic Liquids, in Ionic Liquids: Industrial Applications for Green Chemistry, ed. R. D. Rogers, K. R. Seddon, ACS Symposium Series 818, American Chemical Society, Washington D. C., 2002, pp. 270–288.
31. M. H. Abraham, Scales of Solute Hydrogen-Bonding: Their Construction and Application to Physicochemical and Biochemical Processes, Chem. Soc. Rev., 1993, 73–83.
32. T. A. Hatton, Liquid–Liquid Extraction of Antibiotics, in Comprehensive Bio/technology: The Principles, Applications and Regulations of Bio/technology in Industry, Agriculture and Medicine, ed. M. Moo-Young, Pergamon, Oxford, 1985, pp. 439–449.
33. Principles and Practices of Solvent Extraction, ed. J. Rydberg, C. Musikas and G. R. Choppin, Marcel Dekker, Inc., New York, 1992.
34. W. W. Schultz and E. P. Horwitz, The Truex Process and the Management of Liquid TRU Waste, Sep. Sci. Technol., 1988, 23, 1191–1210.
35. J. N. Mathur, M. S. Murali, P. R. Natarajan, L. P. Badheka and A. Banerji, Extraction of Actinides and Fission Products by Octyl(phenyl)-N,N-diisobutylcarbamoylmethyl-phosphine Oxide from Nitric Acid Media, Talanta, 1992, 39, 493–496.

---