In situ formation of reverse polymeric micelles in liquid alkanes to lodge alcohol micro-droplets

Abstract

An amphiphilic comb-like polymer has been synthesized in a liquid alkane medium, which involves the alkylation of glycidyl methacrylate (GMA) by 1-hexadecylamine (HDA) or 1-octadecylamine (ODA) and the in situ polymerization of the resulting alkyl methacrylate monomer. The resulting macromolecules possess a hydrophilic backbone with thickly anchored –OH and [double bond splayed left]NH groups and long aliphatic side chains extending into the alkane medium, and hence undergo self-assembly in the non-polar medium. The resulting polymeric micellar solution displays an enhanced capability to dissolve methanol or ethanol over those employing low molecular weight surfactants, such as Span® 80, according to a stability study of the resulting microemulsions. The ethanol content can be raised from the contemporary level of 15% to 23% for the same loading (by weight) of emulsifier. By the ASTM D240-09 method, the in-house formulated model diesohol (diesel/ethanol/emulsifier = 75/20/5) exhibits only a minute decline in calorific value, as compared with pristine diesel fuel.

---

Introduction

The major pollutants in exhaust gas from diesel engines consist of particulate matter, smoke density, oxides of nitrogen, polycyclic aromatic hydrocarbons and other emissions, posing health hazards.^1,2 It is therefore imperative to enhance the burning efficiency of diesel towards complete combustion. This has been pursued through structural modification of diesel engines and the introduction of pertinent oxygenates into diesel fuel to curb carbon-rich particles. Ethanol is a promising oxygenate additive for diesel because it is a mass petrochemical product and can also be obtained from renewable biomass resources. Furthermore, it also possesses a high gross calorific value.^3–5 As remarked by Durgun et al.,⁶ there are three approaches developed thus far to improve the combustion of diesel using ethanol: (1) ethanol fumigation using carburetion or a manifold injection technique,⁷ (2) a dual fuel injection technique, and (3) formation of a microemulsion in which the ethanol is highly dispersed in diesel in the presence of an appropriate emulsifier or a co-solvent. Of these three approaches, the last one is most attractive because no engine modification is required.

Physical and chemical properties affecting the miscibility between ethanol and diesel have been extensively investigated by various research groups.^8–10 To improve the solubility of ethanol in diesel fuel, ethyl acetate has also been used as a co-solvent in ethanol-based microemulsified fuel as reported by Chandra¹¹ and Ashok et al.¹² The performance and emissions of the resultant blends were assessed in an unmodified compression-ignition (CI) engine, which showed reduced values of the air pollutants of concern. A light-scattering investigation on the formation of microemulsions in an ethanol–diesel blend reported by Loh et al.¹³ provided an insight into the colloidal stability of fuels containing amphiphilic molecules, such as dodecylamine and oleic acid. Commercially available additives such as sorbitan ester (Span® 80) and soybean methyl esters (AEP-102) have been utilized by Reyes et al. as ethanol–diesel miscibility promoters.¹⁴ Following that, there were numerous publications focusing on ethanol–biodiesel–diesel microemulsions, which are also termed as EB–diesel. The trans-esterified methyl esters of soybean oil, palm oil and rapeseed oil, i.e. biodiesel, were identified as non-fossil-fuel additives since they promote ethanol–diesel miscibility as well.^15–20 On the basis of these accomplishments, exploring an oil-soluble polymeric emulsifier for greater dissolution capacity is of interest from both fundamental and application perspectives.

Although the use of polymeric emulsifiers to stabilize o/w emulsions has been extensively studied and some of them are on the market, such as the Poloxamer type (e.g., Pluronic™) and acrylic type (e.g., Pemulen™), their counterparts to stabilize the reverse emulsion (w/o), in particular in a nonpolar organic medium, are still rare. An ex situ synthesized hydrophobic polymeric emulsifier is normally difficult to dissolve in a non-polar organic medium because of the weak solvation capability of the nonpolar organic solvent, such as diesel or kerosene. The in situ strategy is therefore effective to circumvent this thermodynamic barrier (ΔH > 0) of dissolution. Carrying out the polymerization of an amphiphilic monomer in a nonpolar organic solvent could avoid the dissolution step. This design is similar to the concept described in the review article by Richez et al.,²¹ which summarizes the versatility of dispersion polymerization in non-polar media. The typical structure of such an amphiphilic monomer consists of a vinyl group with adjacent hydrophilic groups and a long aliphatic chain. These types of molecules will undertake micellization in a non-polar solvent, with their vinyl groups collecting around the core of the micelles formed, in which their spatial proximity favors polymerization.

In this study we synthesized an amphiphilic comb-like polymer in n-dodecane or diesel via the free-radical polymerization of an adduct monomer derived from the ring opening alkylation of glycidyl methacrylate with a long-chain aliphatic 1-amine. As the vinyl group and hydrophilic –OH/[double bond splayed left]NH groups are closely located in the adduct monomer, a polar inner space is formed inside each aggregation micelle, which facilitates polymerization as illustrated in Scheme 1. A homogeneous solution was then obtained. It is important to note that the polymer separated from the solution cannot be re-dissolved at all, albeit it could be slightly dissolved in polar solvents, e.g. toluene and DMF, to just meet the requirements for NMR and SEC characterization. Such irreversibility justifies the in situ synthesis as the sole route for application of an oleophilic polymer emulsifier. Moreover, the present polymerization system also permits copolymerization of the adduct monomer with the PEGylated methacrylate, for instance to tune the hydrophilic trait of the core of the polymeric micelle. Subsequently, the capability of the as-generated polymeric micellar solution in n-dodecane (5% by weight) to dissolve methanol or ethanol was evaluated through inspecting changes in the turbidity of the mixture. Fundamentally, turbidity is a light scattering-based measurement, detecting the occurrence of phase separation when larger or more emulsion particles are formed. We formulated an in-house model of diesohol through utilizing the polymeric micelles to lodge ethanol in the ExxonMobil synergy diesel fuel and the gross calorific values were determined by following the corresponding ASTM standard.

Scheme 1 In situ free-radical polymerization of the adduct monomer in an alkane medium brings about a reverse polymer micellar solution.

Experimental

1. Materials

1-Hexadecylamine (HDA, technical grade, 90%, Aldrich), 1-octadecylamine (ODA, technical grade, 90%, Aldrich), n-dodecane (>99%, Sigma-Aldrich), 2,2′-azobis(isobutyronitrile) (AIBN, 0.2 M in toluene, Aldrich), toluene (analytical reagent grade, 99.99%, Fisher Chemical), ethanol (analytical reagent grade, 99.99%, Fisher Chemical), methanol (analytical reagent grade, 99.99%, Fisher Chemical), methyl ethyl ketone (MEK, 99%, Sigma-Aldrich), 1,4-dioxane (99.8%, Sigma-Aldrich), chloroform-d (99.96 atom% D, contains 0.03% v/v TMS, Aldrich), toluene-d₈ (99.6 atom% D, Aldrich), commercial diesel fuel (ExxonMobil Synergy Diesel) and Span® 80 (Fluka) were used as received. Glycidyl methacrylate (GMA, 97%, Aldrich), poly(ethylene glycol) methyl ether methacrylate (PEGME-MA300, average M_n 300, Aldrich) and poly(ethylene glycol) methacrylate (PEG-MA360, average M_n 360, Aldrich) were passed respectively through a short column of neutral alumina to remove the inhibitor before use.

2. Synthesis of the amphiphilic comb-like polymer

In a model synthesis, HDA (3.54 g, 14.66 mmol) or ODA (3.95 g, 14.66 mmol) and n-dodecane (15 mL) were added to a one-neck round bottom flask equipped with a magnetic stirrer, which was then sealed with a rubber septum. The mixture was heated at 70 °C in an oil bath for about 20 min, to form a clear solution. GMA (2 mL, 14.66 mmol) was then introduced, using a syringe, to the solution. The mixture was stirred at 70 °C for 24 h to complete the synthesis of the respective adduct monomer, GMA-HDA or GMA-ODA. The functional group conversion was examined using FT-IR spectroscopy. Subsequently, the reaction mixture was diluted with 25 mL n-dodecane and purged with argon for 20 min before the AIBN initiator (1 mol% with respect to GMA) was introduced into this monomer solution. The solution was then stirred at 70 °C for 24 h under an argon atmosphere to complete the polymerization of the adduct monomer. Through the same procedure, three homogeneous polymeric micellar solutions (P1–P3, Table 1) were achieved. Similarly, ExxonMobil Synergy Diesel was also used as the dispersion medium in place of n-dodecane to obtain two additional polymeric micellar solutions (P4–P5). The polymers generated were sampled respectively by withdrawing a small portion of solution and then adding excess ethanol to precipitate the polymer for structural characterizations.

Table 1 A list of the comb-like polymers and their corresponding micellar solutions

Polymeric micellar solution	Adduct monomers for the comb-like polymer		Dispersion medium
	1-Aliphatic amine	Vinyl monomer
a GMA/HDA = 1/0.8 (molar basis).b GMA/PEGME-MA300 = GMA/PEG-MA360 = 4/1 (molar basis).
P1	HDA	GMA	n-Dodecane
P2	ODA	GMA	n-Dodecane
P3^a	HDA	GMA	n-Dodecane
P4	HDA	GMA	Diesel
P5	ODA	GMA	Diesel
P6^b	HDA	GMA + PEGME-MA300	Diesel
P7	HDA	GMA + PEG-MA360	Diesel
P8	ODA	GMA + PEGME-MA300	Diesel
P9	ODA	GMA + PEG-MA360	Diesel

---

3. Incorporation of the PEGylated monomer unit into the polymeric micelles

In a model synthesis, the random copolymer consisting of an adduct monomer (e.g., GMA-HDA) and a PEGylated monomer was synthesized. To realize this design, the above protocol was slightly modified by introducing PEGME-MA300 (0.84 mL, 2.93 mmol) together with 25 mL of diesel into the solution of the monomer adduct after it had been synthesized in diesel. The successive steps were kept the same as described in Section 2. Correspondingly, PEG-MA360 could also be assimilated into the comb-like copolymer through the same procedure.

4. Grafting homopolymer PGMA with HDA to prepare a control sample

Free-radical polymerization of GMA in 1,4-dioxane containing AIBN (1 mol% of the monomer) was carried out at 70 °C under an argon atmosphere for 6 h. The resulting homopolymer PGMA was separated from the reaction mixture by precipitation in a large excess of methanol, followed by vacuum drying at 50 °C. The recovered amount showed a 93% yield. A given amount of PGMA (1 g) was dissolved in 50 mL of toluene or MEK, together with equimolar HDA based on the functional group of both types. The solution was refluxed under argon for 24 h to carry out ring opening alkylation.

5. Reverse emulsion preparation by dissolving alcohol in the polymeric micellar solution

The as-prepared polymeric micellar solution was mixed with a given amount of n-dodecane to dilute the polymeric micelles to 5 wt%. After that, methanol was injected portionwise (20 μL for each addition) into the diluted micellar solution (6.5 g) with magnetic stirring at 200 rpm for 10 min to homogenize the blend. The homogeneity of the resulting solution was monitored using a turbidity meter. For this examination, the initial turbidity was taken from the micellar solution, i.e. prior to the addition of methanol. The turbidity underwent continuous and minor variation with the introduction of alcohol by the above procedure before the incipient phase instability that accompanies a turbidity jump occurred. As such, the alcohol solubility limit in wt% was the alcohol added until the injection right before the injection that caused an instant turbidity jump. It is clear that the comb-like polymers possess different methanol dissolution capacities, which are summarized in Table 2. Moreover, the diesel-based polymeric micellar solutions (P4 to P9) were diluted by the diesel and their alcohol acceptances were assessed using ethanol following the same procedure as stated above (Table 3). Pristine diesel (Control 1) and Span® 80 (5 wt%)–diesel (Control 2) were adopted respectively as control samples in this assessment.

Table 2 Formation of the reverse microemulsions^a through dissolving methanol in the polymeric micellar solutions

Reverse microemulsion	Polymeric micellar solution	Solubility limit^b (wt% MeOH)	Turbidity at phase separation (NTU)
a Determined by the turbidity measurement.b Refer to Fig. 4 and 5.c Flocculation of polymers P7 and P9 in the diesel happens prior to methanol phase separation (or breakage of the microemulsion).
Sm1	P1	3.5	2.01
Sm2	P2	3.3	5.50
Sm3	P3	3.1	6.17
Sm4	P4	3.1	1194
Sm5	P5	3.1	1091
Sm6	P6	2.8	800
Sm7	P7	2.4^c	NA
Sm8	P8	3.1	1246
Sm9	P9	1.7^c	NA

---

Table 3 A summary of the ethanol dissolution capacity of the diesel with the use of different emulsifiers

Reverse emulsion	Emulsifier	Emulsifier wt%	Solubility limit (wt% EtOH)	Turbidity at phase separation (NTU)
Control 1	NIL	0	11.2	1254
Control 2	Span® 80	5	13.4	1068
Se4_1	P4	1	12.7	806
Se4_2.5	P4	2.5	16.3	1377
Se4	P4	5	22.6	2522
Se5_1	P5	1	13.4	938
Se5_2.5	P5	2.5	16.3	1368
Se5	P5	5	23.1	2394
Se6	P6	5	19.5	1957
Se8	P8	5	19.5	942

---

6. Structural and property characterization

The purified polymer samples were characterized using ¹H NMR spectroscopy on a Bruker Ultra Shield spectrometer (400 MHz), using chloroform-d (for PGMA) or toluene-d₈ (for the comb-like polymers) as solvents. The chemical shifts were referenced to the TMS peak at δ = 0.00 ppm for CDCl₃ and δ = 2.09 ppm for the solvent peak of toluene-d₈. The FTIR spectra were obtained from a Bio-Rad Excalibur FTS-3500 FTIR spectrometer. The dissolution extent of the alcohol in the polymeric micellar solution was recorded using a LaMotte LTC3000 bench-top turbidity meter equipped with five measurement ranges (0–11, 11–110, 110–300, 300–600, 600–4000 NTU), which were calibrated using five respective EPA compliance turbidity standards, for example, the standard with 1000 NTU was used to check readings from the highest range. The rheological behaviors of the diesel-based polymeric micellar solution and the reverse microemulsion were characterized using a Brookfield RV DV-II+ Pro Viscometer at room temperature. The PGMA and P(GMA-HDA) isolated from P1 were characterized using a size exclusion chromatography (SEC) system equipped with a Waters 1515 Isocratic HPLC pump, a 717plus autosampler, a 2414 refractive-index detector, and a PLgel 5 μm Mixed-D SEC column (Agilent Technologies) using DMF as eluent, operated at 1 mL min⁻¹ and 35 °C. Thermogravimetric analysis (TGA) of the pristine diesel and a selected diesel-based polymeric micellar solution was performed on a TA Instrument (Q500 Thermogravimetric Analyzer). All tests were conducted under an airflow (60 mL min⁻¹) over a temperature range of 30–600 °C and at a scan rate of 10 °C min⁻¹. The size distribution of the micelles and emulsion particles in nonpolar media was determined using light-scattering measurements (Brookhaven 90Plus Particle Size Analyzer), with a scattering angle of 90 degrees. The laser source was a semiconductor laser diode, with a wavelength of 659 nm and laser power of 35 mW. To obtain images of the polymeric micelles and the droplets of the reverse emulsion, characterization was performed on a high-resolution transmission electron microscope (Philips CM300 FEGTEM). The polymeric micellar solution (P1) and the reverse microemulsion (20al/S_e4 in Table 4) were diluted respectively in n-heptane to make 0.5 wt% colloidal dispersions. A drop of liquid was transferred from each of the dispersions to a TEM copper grid. Upon drying under ambient conditions, the samples were ready for TEM. Selected ethanol-in-diesel emulsions were sent to Intertek Testing Services (Singapore) Pte Ltd for measurement of the gross calorific values (GCV) of them using the ASTM D240-09 standard.

Table 4 Properties of the model diesohol and combustion tests

Sample	Emulsifier wt%	Ethanol loading wt%	Gross calorific value^a (kJ kg^−1)	Viscosity^b (cP)
a Based on ASTM D240-09.b Measured shear rate of 186 s^−1.
Control 1	0	0	45 825	4.13
P4	5	0	45 425	5.28
20al/Se4	5	20	43 039	3.75
20al/Se5	5	20	42 648	3.63

---

Results and discussion

1. Solvent effects on the synthesis of the comb-like polymer imparted by spectroscopic characterization

As illustrated in Scheme 1, the ring-opening alkylation of GMA with a 1-alkylamine produces the adduct monomer. The alkylation becomes solvent-selective for a long aliphatic chain of the amine. Infrared (IR) spectroscopic studies showed an obviously higher alkylation extent for the reaction carried out in n-dodecane than in toluene (Fig. 1b vs. c), under the same synthetic conditions. The amplitude of the C=O stretching band at 1718 cm⁻¹ of the GMA-HDA adduct monomer is a clear indication of the alkylation extent since unreacted GMA has been removed from the reaction product by vacuum vaporization.

Fig. 1 FTIR spectra of (a) hexadecylamine (HDA), (b) the GMA-HDA adduct monomer synthesized in n-dodecane, and (c) the same adduct monomer synthesized in toluene.

In addition, compared with HDA, GMA-HDA exhibits a far weaker N–H stretching vibration band (∼3331 cm⁻¹) on top of the broad –OH stretching band (3700–3500 cm⁻¹), the presence of the conjugated C=C stretching vibration at 1606 cm⁻¹ and the alkene C–H stretch at 3077 cm⁻¹ (Fig. 1b). All of these validate the formation of the adduct product. The 1-alkylamine could achieve maximum stretch in n-dodecane due to their similar molecular chain structures besides the amino end group. This also facilitates a reverse micellization in which the polar 1-amino groups assemble to form the inner core of the micelle. The inner core functions as the preferential location for GMA molecules added to the micellar solution because of a large contrast in polarity between the core of the micelles and the dispersion medium. This polar enrichment creates spatial proximity between the amino groups and the epoxide groups of GMA, and hence permits efficacy of the alkylation. On the contrary, the same alkylation achieves a lower extent of reaction in toluene as indicated by the negligible C=O absorption (Fig. 1c). It is rational that the long aliphatic chain of the 1-alkylamine would become contractive in toluene relative to in n-dodecane due to the differences between the two solvents in their solubility parameters and Kauri-Butanol numbers.²² The coiled aliphatic chains would shield the terminal amino group from the epoxide group of GMA and therefore sterically retard the alkylation. Subsequent to the alkylation in the nonpolar medium, the resultant adduct product is subjected to polymerization in situ to generate polymeric micelles in either n-dodecane or the diesel medium. The two polymer solids were separated from the micellar solutions P1 and P5 respectively as model samples for characterization. The IR spectra of these two samples show the strongest aliphatic C–H stretch bands in the range of 2800–3000 cm⁻¹ and the characteristic carbonyl group of an ester. In comparison, the control sample PGMA (see Experimental section) displays the asymmetrical epoxide-ring stretching peak at 906 cm⁻¹ and the characteristic peaks of an ester, but a far weaker C–H stretching peak, as displayed in Fig. S1 (ESI†). This spectral comparison supports the generation of a comb-like polymer structure in alkanes. In addition to the IR evidence, the polymer sample separated from micellar solution P1, taken as an example, exhibits a stronger peak at a chemical shift of δ = 1.11–1.93 ppm than PGMA (Fig. S2, ESI†), which also justifies the presence of a comb-like polymer structure. This assessment is similar to the NMR study reported by Leroux et al.²³

We were also curious about whether this comb-like polymer could be synthesized through attaching HDA or ODA to the PGMA backbone in a polar solvent. The alkylation was tested in both toluene and MEK respectively. The comb-like chain formed in toluene precipitated when the alkylation reached a certain extent, as a result of entanglement of the pendant aliphatic chains which are weakly solvated in toluene as aforementioned. Alternatively, only a very low extent of alkylation in MEK could be achieved because the steric shielding by the aliphatic chain on the 1-amino group is more severe in MEK than in toluene. The long alkyl tail of aliphatic amines has very low δ_p (polarity cohesion parameter) and δ_h (hydrogen bonding cohesion parameter) values. They are closer to the δ_p and δ_h of toluene (1.4 and 2 MPa^1/2) than those of MEK (9 and 5.1 MPa^1/2).^24,25 Moreover, as PGMA is insoluble in n-dodecane, the alkylation cannot be carried out in this non-polar medium.

2. Colloidal behaviors of the polymeric micellar solution

The micellization of the comb-like macromolecules in nonpolar medium was characterized using a dynamic light scattering technique. No scattering was detected from the adduct monomer, GMA-HDA, in either n-dodecane or diesel, because its size is below the instrument detection limit (<2 nm diameter). After polymerization, P(GMA-HDA) formed in n-dodecane, sample P1 (Table 1), reveals a size distribution of the micelles centered at 15.9 nm and a polydispersity of 0.375, whereas the same polymer in diesel, sample P4, presents a distribution centered at 20.0 nm with a larger polydispersity of about 0.858, as shown in Fig. 2. The diesel contains aromatic hydrocarbons that have a stronger affinity with the backbone of the polymer and cause the micelles to experience slight expansion. Both size distributions observed match the normal size range of micelles assembled by surfactants and amphiphilic polymers.²⁶ Regarding the particulate structure of the micelles, a small amount of P1 solution was substantially diluted in n-heptane and a drop of the liquid was sampled to conduct a TEM examination. Sparse microcapsules with hollow structures were found in the sample and one of them was selected to present in Fig. 3a. The microcapsule is apparently far beyond the nanoscale as identified by the light scattering measurements. This variation likely happens because the original polymeric micelles undergo merging during dispersion in heptane. It is presumed that the colloidal stability of the micellar solutions originate from the comparable shape and size between the aliphatic side chains and the long-hydrocarbon-chain medium. Hence dispersion in n-heptane largely impairs such a size-dependent stabilization mechanism. As for the molecular weight of P(GMA-HDA), it cannot be accurately determined using size exclusion chromatography because, driven by an amphiphilic trait, the polymer keeps a particulate form in the eluent. Hence, its average molecular weight of 927k must be overestimated (Fig. S3†). Correspondingly, the PGMA sample, used as the reference for SEC, has an average molecular weight of 64k. In principle, PGMA should have a greater number average degree of polymerization than P(GMA-HDA) due to less steric hindrance to polymerization.

Fig. 2 Particle size distribution as characterized by dynamic light scattering.

Fig. 3 Transmission electron micrographs of (a) a particle obtained from drying polymeric micellar solution P1, and (b) a particle obtained from a model diesohol, 20al/S_e4.

3. Incorporation of methanol into the polymeric micellar solution

The dissolution capacity for methanol is a property related to the colloidal stability of the reverse polymer micelles (5 wt%) in n-dodecane or the diesel. With the addition of alcohol into a polymeric micellar solution, a microemulsion is first produced, since the resulting emulsion is optically transparent, in which alcohol is distributed into the interior of individual micelles. Hence, the alcohol dissolution limit reflects the embryonic transition of the microemulsion to a miniemulsion. It may be noted that pure n-dodecane has negligible methanol dissolution capacity compared with the presence of the comb-like polymer. In general, the n-dodecane-based microemulsions, S_m1, S_m2 and S_m3, exhibit greater or at least comparable methanol dissolution capacities than with the diesel-based emulsions, S_m4, S_m5, S_m6 and S_m8 (Table 2). The turbidity profiles of the S_m1, S_m2 and S_m3 microemulsions present close dissolution limits, among them varying from 3.1 to 3.5 wt% (Fig. 4). It is presumed that this narrow split reflects the structural similarities between the aliphatic moiety of the polymer and n-dodecane. S_m1 has a slightly larger capacity than S_m2 because the C₁₆ aliphatic side chain is more analogous to n-dodecane relative to the C₁₈ counterpart. Regarding S_m3, it presents the lowest capacity in the n-dodecane medium because the polymer has a lower substitution density of aliphatic chains (C₁₆) than that used in S_m1. It may also be noted that the diesel-based microemulsions, S_m4 and S_m5, stabilized by the polymer with C₁₆ and C₁₈ side chains, respectively, exhibit flat turbidity profiles before a turbidity jump to greater than 1000 NTUs (Fig. 4 and Table 2). Compared with S_m1, the phase separation at the point just past the dissolution limit is far more severe in S_m4, despite exactly the same polymeric stabilizer being used. This difference also happens with S_m2 vs. S_m5. Apparently n-dodecane defers the sudden collapse of the microemulsion structure in contrast to the diesel, likely owing to its rather complex composition.²⁷ This observation reveals a high demand for structural similarity between the polymer side chain and the hydrocarbons of the dispersion phase for the stability of microemulsions accommodating highly hydrophilic methanol.

Fig. 4 The variations in turbidity with the addition of methanol into the five reverse polymeric micellar solutions where n-dodecane or diesel is the dispersion medium.

In addition, Fig. 5 presents results from an investigation into the methanol solubility of four polymeric micellar solutions (P6 to P9) where the polymers contain PEG side chains besides the aliphatic chains (Table 1). Correspondingly, the four microemulsions S_m6 to S_m9 exhibit lower methanol dissolving limits than the preceding microemulsions by ca. 12–45%. In addition, both S_m7 and S_m9 become vulnerable when approaching their limit because the polymers start precipitating out as sparse floc from the clear liquid, indicating that the microemulsions are not yet broken. The flocculation could be attributed to methanol-prompted coalescence of the hydrophilic co-side chain PEG-360 end-capped by –OH groups. On the contrary, the microemulsion state remains in S_m6 and S_m8 before reaching their methanol-dissolving limits because both emulsions are stabilized by the polymers bearing less hydrophilic PEG-300 co-side chains end-capped by an –OCH₃ group. Similar to S_m4 and S_m5, the turbidity shooting up to ca. 1000 NTUs (Table 2) marks the methanol dissolving limits of S_m6 and S_m8. There is a lever rule between the size of the hydrophilic core and the solvation extent of the aliphatic side chains in the dispersion medium. Hence, the introduction of oligomeric PEG to the core of the micelle in effect promotes instability, as we observed from the above methanol dissolution examination.

Fig. 5 The variations in turbidity with the addition of methanol into the four reverse polymeric micellar solutions (P6 to P9) where the diesel is the dispersion medium.

4. Examination of the ethanol dissolving limit in the diesel containing polymeric micelles

Ethanol is less hydrophilic than methanol and hence has a higher solubility limit (11.2%) in the diesel (Control 1, Table 3). Moreover, introducing a hydrophobic surfactant, Span® 80 (HLB = 4.5), into the diesel brings about only a marginal improvement in ethanol solubility (Control 2). Interactions between ethanol and the hydrophilic moiety of Span® 80 might perturb the ordered assembly of the surfactant molecules. On this basis, the ethanol dissolution limit sustained by the P(GMA-HDA) polymer in diesel was examined by varying its content. It turns out that the microemulsions S_e4 S_e4_2.5 and S_e4_1 have ethanol dissolution limits of 22.6, 16.3 and 12.7 wt%, respectively (Table 3). Similarly, S_e5 shows an ethanol dissolution limit of 23.1%, which is also higher than the other two microemulsions of the same kind, S_e5_2.5 and S_e5_1 (Fig. S4†). Clearly, both S_e4 and S_e5 effectively enhance dissolution of ethanol in the diesel. In addition, these two microemulsions have similar ethanol dissolution limits despite the slightly different lengths of the aliphatic side chains between the two polymeric emulsifiers. It is also noted that S_m4 and S_m5 display rather close turbidity readings to S_e4 and S_e5 before reaching their dissolution limits, albeit the latter two microemulsions contain higher alcohol contents. As turbidity is a measure of the light scattering properties of the emulsion particles, which depend on their size, shape, concentration and refractive index, the above observation implies that the dispersion phases in the S_mx and S_ex microemulsions possess similar interfacial structures and particle sizes, viz. the higher loadings of ethanol do not lead to apparently larger microemulsion particles. This likely happens because ethanol is structurally more analogous to the hydrophilic repeating unit of the polymer main chain, –C(O)OCH₂CH(OH)CH₂NH–, than methanol and a stronger association occurs.

Correspondingly, a dry particle separated from S_e4 shows a denser interior than for its external surface (Fig. 3b). The image implies a gel state formed through swelling of the polymer main chain by ethanol in the interior of microemulsion particles. Hence, a polymer matrix is left behind after the alcohol is removed. Moreover, the particle is also significantly larger than its micelle precursor, which could be attributed not only to ethanol swelling but also to recombination of the polymer molecules during the dissolution of ethanol. It is deemed that the swelling of the hydrophilic polymer main chains by ethanol inside the emulsion droplets offers a superior dissolution capacity over low molecular weight surfactant molecules, since the swelling causes a gel structure which is mechanically much more stable than liquid droplets. Furthermore, both S_e6 and S_e8 show slightly weaker ethanol-dissolution capabilities compared to S_e4 and S_e5. It is considered that too high a degree of ethanol swelling of the hydrophilic PEG-300 side chains would perturb the balancing role of the aliphatic C₁₆ and C₁₈ side chains in S_e6 and S_e8 as proposed afore. It is now clear that the structure of the comb-like polymer, the concentration of the polymeric micellar solution used to develop a microemulsion, and the composition of the diesel affect the dissolution extent of ethanol in the diesel. Hence, performing an orthogonal experimental design in the future is essential in order to determine the optimal conditions for a stable ethanol–diesel blend with maximum ethanol loading.

Pursuant to the above study on the introduction of ethanol into diesel, S_e4 and S_e5 have been proven to possess the maximum ethanol loading capacity amid the microemulsions listed in Table 3. Therefore, 20% ethanol loadings in two formulated diesohols, labelled as 20al/S_e4 and 20al/S_e5, were used to carry out the following characterization. In Fig. 6 the shear stress–shear rate (σ–[small gamma, Greek, dot above]) relationships of these two microemulsions and the two control samples, the diesel (Control 1) and micellar solution P4, were scrutinized. The two control samples manifest Newtonian fluid behavior. Micellar solution P4 has a slightly larger viscosity than the synergy diesel, i.e. 5.28 cp vs. 4.13 cp, due to flocculation of the polymeric micelles because of their aliphatic C₁₆ chains stretching out into the continuous phase. Whereas the two microemulsions, 20al/S_e4 and 20al/S_e5, display lower σ with respect to [small gamma, Greek, dot above] than even the diesel. This phenomenon presumes that the microemulsions have a different interfacial structure to the micellar solutions and a reduced density of the dispersion phase. In addition, both microemulsions exhibit almost identical two-stage σ–[small gamma, Greek, dot above] segments according to a slight decrease in slope (or viscosity), of which the first one ends at 75 s⁻¹. The capsulated ethanol might cause a minor contraction of the pendant aliphatic side chains in diesel. From the perspective of applications, this measurement manifests the colloidal stability of the microemulsions upon shearing within the range of testing.

Fig. 6 Rheological characterization of the four listed samples at room temperature.

A combustion measurement of the synergy diesel (Control 1 in Table 4) showed a gross calorific value (GCV) of 45 825 kJ kg⁻¹, which is typical for commercial diesel. The inclusion of 5 wt% polymeric micelles into it results in a slight reduction in the GCV. If this amount of heat is approximately divided on a portion basis, the GCV of the polymer would be 2291 kJ kg⁻¹. To understand the combustion extent of the polymer in the diesel, the pyrolysis profiles of the above two samples were then assessed by TGA in an airflow (ESI, Fig. S5†). The results showed that P4 displays basically the same combustion profile as diesel. With respect to these two measurement values, 20al/S_e4 exhibited a smaller GCV than that of P4. Taking the GCV of pure ethanol of 29 734 kJ kg⁻¹ into account, the calculated combustion heat of 20al/S_e4 should be 40 430 kJ kg⁻¹ based on the contributions of the three components by their corresponding portions. Hence, the measured GCV is greater than the calculated value by 6.5%. This improvement can be attributed to the oxygen brought in by ethanol. With regards to 20al/S_e5, a slight reduction in the GCV by about 0.9% relative to 20al/S_e4 was obtained. This observation suggests that too long an aliphatic side chain of the comb-like polymer would become unfavorable for combustion.

Conclusions

The main findings of the present work are listed as follows:

1. Accomplishing the in situ synthesis of a non-ionic amphiphilic polymer composed of a hydrophilic main chain and long aliphatic side chains in a nonpolar solvent medium, which constitutes a micellar solution. This comb-like polymer was synthesized through the ring-opening alkylation of a 1-alkyl(C₁₆ or C₁₈)amine with glycidyl methacrylate and subsequent free radical polymerization. The resulting polymeric micellar solution cannot be otherwise realized because the polymer if synthesized separately cannot be dissolved in the nonpolar solvent, typically saturated alkanes.

2. The resulting polymeric micellar solution possesses the alcohol dissolution capability to form a microemulsion, whose stability is impaired to different extents by the presence of hydrophilic co-side chains, e.g. polyethylene glycol, the non-aliphatic hydrocarbons in the non-polar medium, and the increase in the hydrophilic traits of alcohol. The swelling of the hydrophilic backbone by alcohol to form a gel inside the microemulsion droplets upholds the diesohol formed.

3. A micellar solution using ExxonMobil synergy diesel as the dispersion medium can effectively dissolve up to 23 wt% ethanol. The gross calorific value measured for the model diesohol (diesel/ethanol/polymer = 75/20/5) exhibits an enhanced result that is attributed to the synergy of the ethanol and diesel.

Acknowledgements

The authors express their gratitude to the National Research Foundation of Singapore for funding this research (project code: NRF2012NRF-POC001-047).

Notes and references

1. B. Brunekreef and S. T. Holgate, Lancet, 2002, 360, 1233–1242.
2. R. O. McClellan, Annu. Rev. Pharmacol. Toxicol., 1987, 27, 279–300.
3. J. Campos-Fernández, J. M. Arnal, J. Gómez and M. P. Dorado, Appl. Energy, 2012, 95, 267–275.
4. S. Kumar, J. H. Cho, J. Park and I. Moon, Renewable Sustainable Energy Rev., 2013, 22, 46–72.
5. A. C. Hansen, Q. Zhang and P. W. L. Lyne, Bioresour. Technol., 2005, 96, 277–285.
6. Z. Sahin and O. Durgun, Energy Fuels, 2009, 23, 1707–1717.
7. A. Imran, M. Varman, H. H. Masjuki and M. A. Kalam, Renewable Sustainable Energy Rev., 2013, 26, 739–751.
8. K. R. Gerdes and G. J. Suppes, Ind. Eng. Chem. Res., 2001, 40, 949–956.
9. M. Lapuerta, R. García-Contreras, J. Campos-Fernández and M. P. Dorado, Energy Fuels, 2010, 24, 4497–4502.
10. E. Torres-Jimenez, M. S. Jerman, A. Gregorc, I. Lisec, M. P. Dorado and B. Kegl, Fuel, 2011, 90, 795–802.
11. R. Chandra and R. Kumar, Energy Fuels, 2007, 21, 3140–3414.
12. M. P. Ashok, Energy Fuels, 2010, 24, 1822–1828.
13. E. J. Silva, M. E. D. Zaniquelli and W. Loh, Energy Fuels, 2007, 21, 222–226.
14. Y. Reyes, D. A. G. Aranda, L. A. M. Santander, A. Cavado and C. R. P. Belchior, Energy Fuels, 2009, 23, 2731–2735.
15. S. Fernando and M. Hanna, Energy Fuels, 2004, 18, 1695–1703.
16. M. Lapuerta and O. Armas, Energy Fuels, 2009, 23, 4343–4354.
17. O. Armas, M. Gómez, E. J. Barrientos and A. L. Boehman, Energy Fuels, 2011, 25, 3283–3288.
18. P. Kwanchareon, A. Luengnaruemitchai and S. Jai-In, Fuel, 2007, 86, 1053–1061.
19. S. Lebedevas, G. Lebedeva, V. Makareviciene, P. Janulis and E. Sendzikiene, Energy Fuels, 2009, 23, 217–223.
20. N. Yilmaz, Energy, 2012, 40, 210–213.
21. A. P. Richez, H. N. Yow, S. Biggs and O. J. Cayre, Prog. Polym. Sci., 2013, 38, 897–931.
22. G. Knothe and K. R. Steidley, Ind. Eng. Chem. Res., 2011, 50, 4177–4182.
23. H. Gao, M. Elsabahy, E. V. Giger, D. Li, R. E. Prud’homme and J.-C. Leroux, Biomacromolecules, 2010, 11, 889–895.
24. J. Jang, V. H. Pham, S. H. Hur and J. S. Chung, J. Colloid Interface Sci., 2014, 424, 62–66.
25. C. M. Hansen, Hansen Solubility Parameters: A User’s Handbook, CRC Press, Boca Raton, FL, 2nd edn, 2000.
26. T. S. Kale, A. Klaikherd, B. Popere and S. Thayumanavan, Langmuir, 2009, 25, 9660–9670.
27. S. Puri, A. Gupta, H. Dekka, M. Vanmamalai, A. Jha, R. Malhotra and N. Raje, US Pat., 20060277819 A1, Dec 14, 2006.

---

Footnote

† Electronic supplementary information (ESI) available: FTIR and ¹H-NMR spectra for the selected purified polymer samples, SEC chromatograph for homopolymer PGMA and comb-like polymer P1, turbidity profiles with respect to the addition of ethanol into polymeric micellar solutions, pyrolysis profiles of pristine diesel and selected polymer micellar solutions, tables of tabulated turbidity measurement data. See DOI: 10.1039/c4ra14210a

---