Advanced Distillation Curve Analysis on Ethyl Levulinate as a Diesel

Mar 10, 2011 - Multi-SO 3 H functionalized mesoporous polymeric acid catalyst for biodiesel production and fructose-to-biodiesel additive conversion. ...
11 downloads 18 Views 2MB Size
ARTICLE pubs.acs.org/EF

Advanced Distillation Curve Analysis on Ethyl Levulinate as a Diesel Fuel Oxygenate and a Hybrid Biodiesel Fuel Bret C. Windom,† Tara M. Lovestead,† Mark Mascal,‡ Edward B. Nikitin,‡ and Thomas J. Bruno*,† † ‡

Thermophysical Properties Division, National Institute of Standards and Technology, Boulder, Colorado 80305, United States Department of Chemistry, University of California Davis, Davis, California 95616, United States ABSTRACT: Recently, a new processing technique was developed that converts the carbohydrates found in plant biomass into ethyl levulinate, which has properties making it a possible diesel fuel oxygenate additive. Additionally, the new processing technique applied to oil-containing seeds can create a biodiesel fuel at high yields, while possibly enhancing the cold flow properties that commonly plague biodiesel fuels. The first part of this two-part study focused on ethyl levulinate as a possible diesel fuel oxygenate additive, by investigating the volatility of petroleum diesel/ethyl levulinate mixtures. Volatility was measured with the advanced distillation curve (ADC) method for mixtures containing 1, 2.5, 5, 10, and 20% ethyl levulinate (v/v) and compared with unblended petroleum diesel fuel. In addition, the concentration of ethyl levulinate was tracked during the distillation for each mixture by use of the composition explicit data channel. The second part of this study investigated fatty acidlevulinate ester biodiesel fuel blends as viable petroleum diesel fuel extenders/replacements. This was done by measuring their volatilities and comparing them to a commercially available biodiesel fuel, and also to a petroleum diesel fuel. In addition, distillate fractions were withdrawn to measure the changing composition and energy content during the distillation of the fatty acidlevulinate ester biodiesel blends and commercial biodiesel fuel.

’ INTRODUCTION Diminishing petroleum reserves, the potential of supply disruptions, and price volatility, as well as environmental considerations resulting from polluting emissions, have led to development of alternative liquid fuels and fuel additives produced from renewable feedstocks. For example, bioderived diesel fuels have shown promise as extenders for diesel fuels. Compared to gasoline engines, diesel engines have higher efficiencies, requiring less fuel for equivalent work, have longer working lifetimes, and produce less carbon monoxide emissions. On the other hand, a primary concern regarding the diesel engine is the high soot and NOx emissions, and the subsequent environmental impact. Currently, there are numerous research efforts to develop solutions to these problems, including engine and system modifications as well as the development and reformulation of diesel fuels.17 For example, many studies have focused on investigating oxygenates and their effect on soot reduction as well as their effect on the overall performance of the fuel.35,811 In fact, it has been shown that completely smoke-free operation can be obtained when 38% (mass/mass) oxygen is incorporated into the diesel fuel.12 We note that a formulation of fuel to include this amount of oxygen would not be cost-effective and would lead to increased fuel consumption; however, this figure does show the importance of oxygen in reducing soot emission. Unfortunately, many of the oxygenate additives investigated are petroleum derived, while others are very expensive and energy intensive to produce. Biodiesel fuel has many favorable properties as a possible renewable extender for petroleum diesel fuel.13 In addition to its renewability, other advantages of biodiesel include biodegradability, miscibility with petroleum diesel, increased lubricity, high flash point, and reduced emissions of particulate matter, unburned hydrocarbons, and carbon monoxide during fully warmed-up r 2011 American Chemical Society

conditions.1417 Indeed, the high lubricity is particularly helpful since the trend toward ultra low sulfur diesel fuel (ULSD) has proven problematic in this regard.1719 On the other hand, some of the known problems with biodiesel fuel include thermal and oxidative instability, poor low temperature flow properties (higher cloud point), increase in NOx emissions, increased fuel consumption due to lower net heat of combustion, and increased unburned fuel and carbon monoxide emissions under cold start conditions.2026 A number of feedstocks are currently used to produce oils that can be processed into biodiesel fuel.2736 Soybean, palm, rapeseed, sunflower seed, coconut, peanut, cottonseed oils, jatropha, algae, and cuphea seeds along with animal fats and waste oils are just a few of the feedstocks sources that have been proven technologically viable, although not necessarily economical.2732 Typically, the feedstock that is selected depends on the crop economics, geography, climate, and crop availability.32 Because of the small oil yield extracted from plant sources, however, a great deal of crop feedstock is needed for commercialization of biodiesel fuel. This requires large areas of land, and can result in competition with food crops. For these reasons, many scientists are looking for crop sources and new processing techniques that result in higher yields of fuel while requiring less land. Recently, a new process has been developed to convert carbohydrates found in raw biomass (cotton, wood, straw, other plant materials, and even newsprint) into simple organic molecules that may be further processed into either fuel or fuel additives.37 This technology uses a biphasic acid/solvent biomass Received: February 15, 2011 Revised: March 10, 2011 Published: March 10, 2011 1878

dx.doi.org/10.1021/ef200239x | Energy Fuels 2011, 25, 1878–1890

Energy & Fuels Table 1. Data on the Additive Studied in This Work91 a

ethyl levulinate :

CAS No. 539-88-8 INChI = 1S/C7H12O3/c1-3-10-7(9)5-4-6(2)8/h3-5H2,1-2H3 RMM = 144.1684 Tboil = 205.8 °C Tfus = 32.8 °C density = 1.00831 g/mL (25 °C) refractive index, Nad = 1.4225 (20 °C) flash point: 90.6 °C (TCC) ΔHcomb = 3523 kJ/mol a

Synonyms: Pentanoic acid, 4-oxo-, ethyl ester; Levulinic acid, ethyl ester; Ethyl laevulinate; Ethyl levulate; Ethyl 3-acetylpropionate; Ethyl 4-ketovalerate; Ethyl 4-oxopentanoate; Ethyl 4-oxovalerate; ELA; Ethyl ketovalerate.

digester to convert the carbohydrates (mono, di, or poly saccharides) into a mixture of 5-(chloromethyl)furfural (CMF) and levulinic acid in combined yields up to 95% (generally 7090% CMF and less than 10% levulinic acid). By use of a suitable alcohol post-treatment, the CMF/levulinic acid mixture can be converted into levulinate esters. Ethyl levulinate contains ∼14% (mol/mol) oxygen and properties enabling it to be a viable bioderived petroleum diesel fuel oxygenate additive. Some properties of ethyl levulinate are provided in Table 1. We note that these properties include a high flash point (90 °C), a normal boiling temperature (206 °C) similar to the initial boiling temperature of typical diesel fuel, and a higher energy content compared to that of many other oxygenate candidates. Preliminary tests have even shown that ethyl levulinate as an oxygenate can be blended with petroleum diesel up to 10% without a change in the cetane number.38 In addition, the same tests have shown that ethyl levulinate enhances the lubricity of the diesel fuel. We note that these results are encouraging for use of this material as a diesel blending component, but considerable additional data and testing are needed. Specifically, it must be shown that ethyl levulinate diesel blends are stable in the presence of water, and that ethyl levulinate is compatible with fuel system materials. In addition, engine performance and emissions data are also required. As a follow up study, the new carbohydrate conversion technique was applied to oil-containing seeds effectively increasing the yield in biodiesel production.39 During the typical production of a biodiesel fuel from a plant source, the oils are extracted and converted from triglycerides into single chain fatty acid esters. This is done by transesterification, in which an alcohol, typically methanol or ethanol, is mixed with the triglycerides and an acid or base catalyst, resulting in glycerol and three monoalkyl esters. Applying the biphasic acid/solvent biomass digester mentioned earlier, the carbohydrates found in the seeds can be converted into CMF and levulinic acid while simultaneously extracting the oils from the seeds.39 Following an ethanol treatment at 200 °C for 6 h, the CMF/triglyceride mixture can be converted into a mixture

ARTICLE

of fatty acid ethyl esters (FAEEs, commonly found in biodiesel fuel) and ethyl levulinate. The transesterfication of the triglycerides is catalyzed as a result of the in situ production of the catalytic acid produced during the ethanolysis of the CMF. The above process, when carried out on the seeds of the safflower plant, yields a 24% increase in fuel production when compared to standard transesterfication of the available oil from equal masses of feedstock.39 It is conceivable that if the carbohydrates found in the entire plant were processed (leaves, stalks, and seeds), an even greater yield of the levulinate component of the fuel could be achieved. As an added benefit, the high concentration of the ethyl levulinate in the safflower biodiesel fuel blend is a possible approach to enhance the cold flow properties associated with biodiesel fuel and reduce the associated performance and emission problems that can occur at low temperatures.39 In contrast, the majority of feedstocks currently used for biodiesel fuel production produce long carbon chain fatty acid esters, specifically unsaturated fatty acid esters, that are responsible for the poor cold flow properties. Oil seed biodiesel fuel blends produced in the manner described above contain many of these same fatty acid esters. The blends also contain a large amount of ethyl levulinate which has a lower melting and boiling temperature than those of any of the FAEEs. We should note that the influence of the ethyl levulinate on the oxidative stability of the FAEEs has not been investigated and is unknown at this time. Reliable measurements of the thermophysical properties of both the biodiesel fuel blend and the ethyl levulinate oxygenate will aid in better understanding how the fuel will perform, and help determine the optimal conditions for peak performance. The first part of the work presented here focuses on ethyl levulinate as a potential oxygenate additive for petroleum derived diesel fuel. Specifically, we measure the volatility (as an approximation of vapor/liquid equilibrium) of mixtures of diesel fuel with ethyl levulinate. The second part of this study investigates fatty acidlevulinate ester biodiesel blends by accurately measuring their volatility and comparing it to that of a commercially available soy-based biodiesel fuel. These two families of fluids are linked by the common carbohydrate conversion processing technique discussed above. Each has a different purpose, however, and thus for the remainder of this paper they will be addressed separately. Advanced Distillation Curve Measurement. One of the most important and informative thermophysical properties that is measured for complex fluid mixtures is the distillation (or boiling) curve.4042 Simply stated, the distillation curve is a graphical depiction of the boiling temperature of a fluid or fluid mixture plotted against the volume fraction distilled. Distillation curves are typically associated with petrochemicals and petroleum refining.43 Such curves are of great value in assessing the properties of any complex fluid mixture; indeed, the distillation curve is one of the few properties that can be used to characterize a complex fluid by providing an approximation of the vapor/liquid equilibrium. Thus, distillation curves are used commonly in the design, operation, and specification of liquid fuels such as gasoline, diesel fuel, rocket propellant, and gas turbine fuel.4446 The standard test method for the measurement of the distillation curve, ASTM D-86, provides the usual approach, yielding the initial boiling point, the temperature at predetermined distillate volume fractions, and the final boiling point.47 The ASTM D-86 test suffers however from several drawbacks, including large uncertainties in temperature measurements and little theoretical significance.48 1879

dx.doi.org/10.1021/ef200239x |Energy Fuels 2011, 25, 1878–1890

Energy & Fuels In earlier work, we described an improved method and apparatus for distillation curve measurement that is especially applicable to the characterization of fuels.4855 This method, called the advanced distillation curve (ADC), is a significant improvement over current approaches such as ASTM D-86. First, we incorporate a composition-explicit data channel for each distillate fraction (for qualitative, quantitative, and trace analysis). Sampling very small distillate volumes (525 μL) yields a composition-explicit data channel with nearly instantaneous composition measurements. Chemical analysis of the distillate fractions allows for determination of how the composition of the fluid varies with volume fraction and distillation temperature, even for complex fluids. These data can be used to approximate vaporliquid equilibrium (volatility) of complex mixtures and presents a more complete picture of the fluid under study. The ADC approach provides consistency with historical data, an assessment of the energy content of each distillate fraction and, where needed, a corrosivity assessment of each distillate fraction. Suitable analytical techniques include gas chromatography with either flame ionization detection (GC-FID) or mass spectral detection (GC-MS), element specific detection (such as GC with sulfur or nitrogen chemiluminescence detection, GC-SCD or GC-NCD), and Fourier-transform infrared spectrophotometry (FTIR).56,57 Another advantage of the ADC approach is that it provides temperature, volume, and pressure measurements of low uncertainty, and the temperatures obtained are true thermodynamic state points that can be modeled with an equation of state.5862 Such thermodynamic model development is simply impossible with the classical approach to distillation curve measurement. We have applied this metrology to hydrocarbon mixtures, azeotropic mixtures, gasolines, diesel fuels (including biodiesel fuels), crude oils, aviation fuels, and rocket propellants.4852,5457,6382 The ADC method was applied to each of the two families of fluids (the petroleum diesel fuel/ethyl levulinate mixtures and the fatty acidlevulinate ester biodiesel blends). Thermodynamically consistent distillation curves were measured for mixtures of petroleum diesel fuel and ethyl levulinate at varying concentrations. The composition-explicit data channel was used to track the amount of ethyl levulinate during the distillation of the mixtures. These measurements show the changes in thermophysical properties that can be expected when varying concentrations of ethyl levulinate are added to petroleum diesel fuel, which is important when considering a new oxygenate additive. Thermodynamically consistent distillation curves were measured for the fatty acidlevulinate ester biodiesel blends while the composition-explicit data channel was used to characterize the curve in terms of composition and available energy content. The results following the ADC analysis of the fatty acidlevulinate ester biodiesel blends were compared to results from a previously measured commercially available soy-based biodiesel fuel.52 Thus, we provide a basis of comparison among these fuels in terms of the fundamental thermophysical properties. This comparison is critical in determining the applicability and suitability of the fatty acidlevulinate ester biodiesel blends for use as renewable replacements or extenders for petroleum diesel fuel.

’ EXPERIMENTAL SECTION 1. Materials. 1.1. Study of Ethyl Levulinate as an Oxygenate. The diesel fuel used in this work was obtained from a commercial source and was stored in a tightly sealed container to preserve any volatile

ARTICLE

components. The fuel was a summer grade, low wax, ULSD fuel that incorporated a red dye (specifying off-road use), and was refined locally from petroleum of the Denver-Julesburg field. This fluid was used without any purification or modification. The composition was studied with a gas chromatographic method (30-m capillary column of 5% phenyl95%-dimethyl polysiloxane having a thickness of 1 μm, temperature program from 90 to 275 °C, 9 °C per minute) using flame ionization detection and mass spectrometric detection.8385 This analysis was unremarkable in that the typical pattern of commercial petroleum derived diesel fuel was observed. The n-hexane used as the solvent in this work was obtained from a commercial supplier. Prior to use, the n-hexane was analyzed by gas chromatography (same column as above, temperature program from 50 to 170 °C, 5 °C per minute) by use of flame ionization detection and mass spectrometric detection. These analyses revealed the purity to be approximately 99.9% (mass/mass), and the fluids were used without further purification. The ethyl levulinate used in the diesel oxygenate study was obtained as a pure fluid from a commercial source. It was a reagent grade fluid with reported purities of 9899% (mass/mass). This fluid was analyzed with chromatographic methods confirming that the purity of the fluid was in fact higher than specified. Ethyl levulinate and ULSD mixtures were prepared as stock solutions of 1, 2.5, 5, 10 and 20% oxygenate by volume. This ensured that there were no variations in composition among measurements done with each mixture. It should be noted that the solubility limit, indicated by the onset of phase separation, was measured to be approximately 16% (v/v) ethyl levulinate in ULSD at room temperature (23 °C). No phase separation was noticed in the 20% mixture at 35 °C, therefore, the 20% mixture was made up in the distillation flask allowing the heating and stirring during the distillation measurements to dissolve the ethyl levulinate into the ULSD. As we will describe later, the marginal miscibility of the 20% mixture affected our ability to characterize the distillate composition of this mixture. The miscibility issues need to be further investigated, more specifically, in terms of what minor impurities (i.e., water) and conditions (i.e., temperature) may trigger phase separations at lower blend levels. Some discussion is warranted regarding our choice of mixture ratios. Mixtures that contain 20% (v/v) of a given oxygenate additive might not be practical fuels for all circumstances, and the performance of an engine using such a mixture could potentially be poor. Even when high concentrations of oxygenates are workable or even advantageous, they may not be economically viable. Our study of a particular mixture ratio does not imply that mixtures with very high oxygenate concentrations are practical formulations for fuels, nor that we advocate such as mixtures for use as fuels. Rather, our mixtures were chosen to cover as large a range of composition as practical for the purpose of modeling the results with equations of state. Likewise, our reason for the use of reagent grade oxygenating fluids rather than the technical grade fluids that might be used industrially speaks to the same issue. 1.2. Study of the Fatty AcidLevulinate Ester Biodiesel Fuel. The acetone used as a solvent in this work was obtained from a commercial supplier, and was analyzed by gas chromatography (same column described above, temperature program from 50 to 170 °C at a heating rate of 5 °C per minute) with flame ionization and mass spectrometric detection. These analyses revealed the purity to be approximately 99%, and the fluid was used without further purification. Two batches of the safflower fatty acidlevulinate ester biodiesel blend (FLBB) produced with two different compositions were analyzed in this study. The commercial biodiesel fuel (B100) chosen for comparison to these two FLBBs was produced from a soybean feedstock via a traditional transesterification processing using methanol. The biodiesel fuel samples were stored tightly sealed in plastic bottles, and care was taken to minimize exposure to the atmosphere to limit oxidation, evaporation of the more volatile components, and the uptake 1880

dx.doi.org/10.1021/ef200239x |Energy Fuels 2011, 25, 1878–1890

Energy & Fuels of moisture. No other precautions were taken, nor were the samples physically or chemically dried. The biodiesel fuel blends were subjected to chemical analysis before the measurement of the distillation curve. They were analyzed with GCMS in scanning mode with a 30-m capillary column with a 0.1-mm coating of the stationary phase, 50% cyanopropyl50% dimethyl polysiloxane. This stationary phase provides separations primarily based upon polarity and is specifically intended for the analysis of fatty acid ester compounds that make up biodiesel fuels.69 Samples were injected with a syringe into a split/splitless injector set with a 50 to 1 split ratio. The injector was operated at a temperature of 300 °C and a constant head pressure of 82.7 kPa (12 psig). A temperature program of 80 °C for 2 min followed by temperature ramping at 8 °C per minute to 225 °C was used. The temperature was then held at 225 °C for 5 min. Mass spectra were collected for each peak from 15 to 550 relative molecular mass (RMM) units. Peaks were identified with guidance from the NIST/EPA/NIH Mass Spectral Database,83 and also on the basis of retention indices. 2. Distillation Curve Measurements. The method and apparatus for ADC measurements has been reviewed in detail elsewhere,4852 and thus only a limited description of a typical measurement is provided herein. In brief, 200 mL of each sample fuel (either the FLBB or the ULSD/ethyl levulinate mixture) was placed into the boiling flask with a 200 mL volumetric pipet (equipped with an automatic pipet) for each distillation curve measurement. Thermocouples were then inserted into the proper locations to monitor Tk, the temperature in the fluid, and Th, the temperature at the bottom of the takeoff position in the distillation head. Enclosure heating was then commenced with a four-step program based upon a previously measured distillation curve and knowledge of the neat sample composition.29 Volume measurements were made in a level-stabilized receiver, and sample aliquots were collected at the receiver adapter hammock. Due to the oxidative instabilities of the fatty acid esters in the biodiesel fuels, an argon sparge was used to remove and exclude air from the apparatus during their distillations. This procedure has been described in detail elsewhere and shown to give highly reproducible distillation curve results.29,69,81 In short, a capillary inserted into the biodiesel sample contained inside the kettle bubbled ∼5 mL/ min of argon into the sample for ∼20 min prior to heating. Once heating began, the sparge tube was removed from the fluid (to avoid affecting the fluid temperature during the distillation) and positioned directly above the boiling fluid. An argon purge of the atmosphere above the fluid was maintained throughout the distillation at a flow rate of 5 mL/min. This flow rate has no measurable effect on the observed temperatures. Because the measurements of the distillation curves were performed at an elevation of ∼1650 m resulting in an ambient atmospheric pressure of ∼83 kPa (measured with an electronic barometer), temperature readings were adjusted for what should be obtained at standard atmospheric pressure (1 atm =101.325 kPa). This adjustment was done with the modified Sydney Young equation, in which the constant term was assigned a value of 0.000109.8688 This value corresponds to a carbon chain of 12. In the chemical analysis of the ULSD sample, as well as in previous work on diesel fuel, it was found that n-dodecane can indeed represent the fluid as a very rough surrogate, not accounting for the presence of the oxygenate additive. Clearly, biodiesel fuel is better represented by a hydrocarbon chain length greater than 12. We chose not to use a predicted value for the constant term (for a larger chain) in our data presentation in order to be consistent with previous measurements, and with measurements on ULSD. We note, however, that for a carbon chain of 18, the predicted constant would be 0.000095; use of this value (in place of that for a chain length of 12) would lower the presented temperatures uniformly by 1.2 °C. The shapes of the distillation curves would be unchanged. The magnitude of the Sydney Young correction depends on the extent of the deviation from standard atmospheric pressure. The pressure adjustment resulted in a typical temperature adjustment of ∼8 °C. The actual measured temperatures

ARTICLE

Table 2. Comparison of the Initial Boiling Conditions of the Ethyl Levulinate/ULSD Mixturesa vapor rise (° C)

first drop (° C)

1% (83.1)

215.4

216.0

2.5% (83.5) 5% (83.7)

212.9 210.0

213.4 210.4

10% (83.6)

202.6

203.3

20% (83.5)

199.6

197.0

ULSD (83.9)

217.6

217.9

sample (pressure) (kPa)

a

The vapor rise temperature is the initial boiling temperature (IBT) of the mixture. These temperatures have been corrected to 1 atm (101.325 kPa) with the Sydney Young equation. The pressures at which the measurements were made are provided in the first column to permit recovery of the actual measured temperature. The uncertainties are discussed in the text.

are easily recovered from the Sydney Young equation at each measured atmospheric pressure which is reported for every measurement throughout this work. To provide the composition channel information to accompany the temperature data grid on the distillation curves, sample aliquots were withdrawn for selected distillate volume fractions. To accomplish this, aliquots of ∼7 μL of emergent fluid were withdrawn from the sampling hammock in the receiver adapter with a blunt-tipped chromatographic syringe and added to a crimp-sealed vial containing a known mass (∼1 mL) of solvent (hexane for the oxygenate study and acetone for the biodiesel fuel). Distillate samples were withdrawn at the first drop of fluid from the condenser and then at each of 11 additional volume fractions of distillate, for a total of 12 sample aliquots. Each distillate volume aliquot underwent two analyses. First, each aliquot was subjected to chemical analysis and peak identification by GC-MS as described above for the analyses of the neat samples. Once the compounds in each aliquot were identified, the aliquots were analyzed with GC-FID with external standards to determine the compound concentrations.8385 The quantitative GC-FID analyses were performed by injecting 3 μL of the solvent/sample mixtures from the crimp-sealed vials into the chromatograph with an automatic sampler. High-purity nitrogen was used as the carrier and makeup gas. The split/ splitless injection inlet was maintained at 300.0 °C, and the detector was maintained at 275.0 °C for all the analyses performed with the GC-FID. The column and temperature programs used were identical to those of the GC-MS analyses previously discussed. The GC-FID analysis was applied to the oxygenate study so that we could quantitatively track the concentration of the ethyl levulinate during the distillation. For the biodiesel blends, following standardization, the GC-FID results allowed us to quantitatively determine the composition of each distillate fraction. By knowing the distillate composition, the enthalpy of combustion for each distillate fraction was determined in a process that has been described in detail in previous work.54

’ RESULTS AND DISCUSSION 1. Study of Ethyl Levulinate as an Oxygenate. The initial boiling behavior as described by the vapor rise temperature and the temperature of the first drop of condensate from the condenser for the ethyl levulinate/ULSD mixtures is provided in Table 2. All temperatures are corrected to standard atmospheric temperature using the modified Sydney Young equation, discussed above. The temperatures provided in Table 2 represent an average of three separate measurements. All the temperatures recorded in Table 2 have uncertainties 1881

dx.doi.org/10.1021/ef200239x |Energy Fuels 2011, 25, 1878–1890

Energy & Fuels

ARTICLE

Table 3. Representative Distillation Curve Data (Given As the Average of Two Distillation Curves) for Mixtures of Diesel Fuel with Ethyl Levulinate Additive in 1, 2.5, 5, 10, and 20% (v/v) Concentrationsa 1% (83.1 kPa)

2.5% (83.5 kPa)

5% (83.7 kPa)

10% (83.6 kPa)

20% (83.9 kPa)

ULSD (83.9 kPa)

distillate volume fraction (%)

Tk (° C)

Tk (° C)

Tk (° C)

Tk (° C)

Tk (° C)

Tk (° C)

5 10

223.1 228.3

220.2 225.0

216.8 221.2

210.6 214.4

205.5 207.9

225.5 230.1

15

233.7

231.0

226.6

218.4

210.7

235.4

20

239.5

236.5

232.4

222.7

213.7

240.9

25

244.7

242.1

238.0

228.1

217.3

245.9

30

250.1

247.9

244.5

234.2

221.6

251.2

35

255.7

253.7

251.0

240.8

226.7

256.5

40

261.4

259.6

257.6

248.0

234.1

261.9

45 50

268.0 273.9

266.5 272.3

264.9 271.1

256.0 264.3

245.5 257.2

268.2 273.8

55

280.1

278.9

277.9

272.4

269.1

280.2

60

287.3

286.3

285.0

281.0

279.3

287.1

65

294.4

294.1

292.3

289.0

287.7

294.3

70

302.4

302.0

300.2

298.7

296.3

301.6

75

310.6

310.7

309.0

305.9

306.8

310.0

80

320.0

319.9

317.9

316.6

317.3

319.2

85

330.1

330.0

328.7

327.5

329.4

328.9

a

Data for the neat diesel fuel are also provided. The uncertainties are discussed in the text. These temperatures have been corrected to 1 atm with the Sydney Young equation.

Figure 1. Distillation curves of diesel fuel with 1, 2.5, 5, 10, and 20% (v/v) ethyl levulinate. Also include is the distillation curve measured for the neat ULSD. Here we present Tk, the temperature measured directly in the fluid. The hashes along the y-axis represent the initial boiling temperature (IBT) of the fluid. The uncertainties are discussed in the text.

(expressed as the average standard deviation for each volume fraction) less than 0.6 °C. The vapor rise temperature represents the initial boiling temperature (IBT) of the fluid and is marked by the rapid increase in the head temperature (Th) as the vapor begins to travel out of the kettle and into the distillation head. The boiling temperature of ethyl levulinate, at 205.8 °C, is less than the IBT (indicated by the vapor rise temperature) of neat ULSD. The addition of ethyl levulinate

lowers the IBT and has greater affect as its concentration is increased. The distillation curve data for the 1, 2.5, 5, 10 and 20% (v/v) ethyl levulinate in ULSD are provided in Table 3 and presented graphically in Figure 1. The results from a neat ULSD are also presented. The data are presented in Tk (fluid temperature measured directly). The Tk values represent the true thermodynamic state point (the temperature of the mixture itself). 1882

dx.doi.org/10.1021/ef200239x |Energy Fuels 2011, 25, 1878–1890

Energy & Fuels

ARTICLE

Figure 2. Histogram plot showing the results of the quantitative analysis for ethyl levulinate as a function of distillate volume fraction for the four starting mixtures in ULSD (1, 2.5, 5, and 10% ethyl levulinate in diesel fuel). There is no detectable ethyl levulinate in any mixture after the 60% distillate cut. The analysis of the 20% ethyl levulinate mixture is not presented in the above figure for reasons explained in the text.

Figure 3. Representative chromatograms of the neat fatty acidlevulinate ester biodiesel fuel blends. The y-axes are arbitrary units of intensity, and the x-axes are retention time (RT) in min. The details of the chromatography are discussed in the text.

The data in Table 3 and Figure 1 are the average of between two and five separate measurements performed for each fluid mixture. The uncertainty in the temperature measurements was less than 0.4 °C, and the uncertainty in the volume measurements was 0.05 mL (expressed as the standard deviation among the replicate curves). The ethyl levulinate has the greatest influence on the boiling temperatures of the early distillate fractions. As represented in Figure 1, the distillation curves deviate further from the ULSD curve as the ethyl levulinate concentration is increased. The shapes of the distillation curves, especially for the higher concentration ethyl levulinate mixtures, provide a better understanding of the evolving composition of the mixtures as the vaporization progresses. For example, the 20% ethyl levulinate mixture is initially flat because it is ethyl levulinate that is primarily being vaporized. This results in the greatest deviation from the ULSD curve at the 35% volume fraction. Following the 35% volume fraction, the slope of the curve begins to increase as the heavier components making up the diesel fuel begin to volatize. The 20% ethyl levulinate mixture does not approach the temperatures of the ULSD curve until the 85% distillate volume

fraction. For the 1 and 2.5% concentrated mixtures, the distillation curves only slightly deviate from the ULSD curve initially and coincide with it following the 45% volume fraction. The concentration of the ethyl levulinate was traced during the distillation by withdrawing and analyzing samples taken during the distillation by use of the explicit data channel. The results of these sequential analyses for the 1, 2.5, 5, and 10% mixtures can be seen in Figure 2. Similar analyses were also performed on the 20% ethyl levulinate mixture, but these measurements yielded very high uncertainties and are not presented. The high uncertainty is probably due to the onset of phase separation (liquidliquid equilibrium) in the condenser during the measurement. It is simply not possible to reproducibly sample and analyze a two phase liquid. The evolution of the ethyl levulinate concentration during the distillation follows the same trend for each mixture. The maximum concentration occurs at the 10% volume fraction and continues to decrease until the ethyl levulinate has been completely vaporized, which occurs following the 60% volume fraction. Since the boiling point of ethyl levulinate is less than 1883

dx.doi.org/10.1021/ef200239x |Energy Fuels 2011, 25, 1878–1890

Energy & Fuels

ARTICLE

Table 4. List of Components and Their Relative Amounts in the Two Batches of the Fatty AcidLevulinate Ester Biodiesel Fuel Blend Identified by GC-MS and Determined by GC-FID batch 1

batch 2

components

mass fraction (%)

mol fraction (%)

components

mass fraction (%)

mol fraction (%)

ethyl levulinate

20.05%

ethyl palmitate

5.38%

34.91%

ethyl levulinate

10.00%

19.16%

4.74%

ethyl palmitate

8.78%

ethyl stearate ethyl oleate

2.59% 58.62%

2.08% 47.39%

8.53%

ethyl stearate ethyl oleate

2.32% 64.20%

2.05% 57.10%

ethyl linoleate

13.36%

10.87%

ethyl linoleate

14.69%

13.15%

Table 5. Comparison of the Initial Boiling Conditions of the Fatty AcidLevulinate Ester Biodiesel Fuel Blends, the Soy Based B100 Sample, and Neat ULSDa vapor rise (° C)

first drop (° C)

batch 1 (83.3 kPa)

249.9

251.2

batch 2 (84.5 kPa)

281.5

283.4

B100 (82.9 kPa)

345.6

341.6

ULSD (83.9 kPa)

217.6

217.9

sample (pressure) (kPa)

a

The vapor rise temperature is the initial boiling temperature (IBT) of the mixture. These temperatures have been corrected to 1 atm (101.325 kPa) with the Sydney Young equation. The pressures at which the measurements were made are provided in the first column to permit recovery of the actual measured temperature. The uncertainties are discussed in the text.

that of the vapor rise temperature of the ULSD, one would expect the first drop to contain the highest concentration of ethyl levulinate, however, the diesel fuel contains small amounts of components that have boiling points lower than that of ethyl levulinate. The presence of these fluids therefore decreases the concentration of the ethyl levulinate in the first drop. 2. Study of the Fatty AcidLevulinate Ester Biodiesel Fuel. As discussed above, two compositions of FLBB were analyzed during this study. Prior to any distillation measurements, the compositions of the two different batches were identified by use of GC-MS and quantified by use of GC-FID measurements. The chromatograms of the neat FLBB samples can be seen in Figure 3. A list of components along with their associated mass and mole percents for the two FLBB samples is provided in Table 4. The difference between the two batches was the relative amount of ethyl levulinate (and presence of appreciable residual solvent in the second sample). Batch 1 contained approximately twice as much ethyl levulinate (by mass) as Batch 2. This reflects the natural variability in composition with respect to the oil seed source of the FLBB, which ranges from about 10% for jatropha fruits to about 20% for safflower seeds, by use of the carbohydrate/fatty acid processing technique described earlier.39 Based on previous measurements performed on the B100 sample, it is clear the ethyl levulinate is what distinguishes the FLBB from an ordinary biodiesel fuel. The initial boiling behavior of the two FLBB batches, the B100, and the ULSD are provided in Table 5. The uncertainties in these measurements were less than 0.3 °C. It is interesting to note the significant difference in the IBTs of the two FLBB batches. Batch 1, which has more ethyl levulinate, has an IBT that is ∼11% less than that of Batch 2. Because ethyl levulinate has a considerably lower boiling point compared to the IBT of the FAEEs, its concentration has the greatest effect on the initial

boiling temperature and the boiling temperatures of the early volume fractions. The addition of ethyl levulinate significantly lowers the boiling point as compared to B100, making it more comparable to the initial boiling temperature of ULSD. Representative distillation curve data for the two FLBB batches, the B100 sample, and the ULSD are given in Table 6. These data are presented in both Tk (fluid temperature measured directly) and Th (head temperature). The data are the average of two separate measurements performed for each fluid. Sample limitations prevented additional replicate measurements. The uncertainties in the temperature measurements for each fluid were less than 0.4 °C. The Tk data for each mixture are presented graphically in Figure 4. The effect of the ethyl levulinate is apparent early in the distillation curves. In addition, the concentration of the ethyl levulinate has a pronounced effect on the boiling temperatures of the early distillate fractions as seen by the differences in the curves for Batch 1 and Batch 2. Following the complete vaporization of ethyl levulinate, the FLBB distillation curves flatten and resemble that of the B100. The curves flatten because the remaining components (the FAEEs) all have similar boiling points. The points at which the curves begin to flatten are different for the two batches because of the different initial concentrations of ethyl levulinate. The ethyl levulinate does not completely vaporize in Batch 1 until the 20% volume fraction, whereas removal of the ethyl levulinate in Batch 2 occurs at approximately the 10% volume fraction. This is because there is approximately twice as much ethyl levulinate in Batch 1 compared to Batch 2 (see Table 4). Following the flattening of the curves, we note that the FLBB boils at slightly higher temperatures than the B100. This is because the FAEEs have higher boiling temperatures compared to the FAMEs which make up the B100 fluid. Beginning at ∼75% volume fraction, the FLBB and the B100 curves sharply increase. This is a result of the fatty acid esters polymerizing into larger molecules resulting in the increase in the boiling temperature of the fluid.89 While the gross examination of the distillation curves is instructive and valuable for many design purposes, the composition channel of the ADC approach provides more information that can aid in developing a better understanding of the thermophysical behavior of the fluid. A series of chromatograms for the two FLBB batches is provided in Figure 5a and 5b. The analysis of the first drop of Batch 2 showed higher concentrations of hexane and ethanol, which are two reagents used during processing of the FLBBs. The chromatograms show higher concentrations of the FAEEs in the 10% volume fraction in Batch 2 compared to the 10% volume fraction of Batch 1. This is because the ethyl levulinate in Batch 2 is nearly completely removed from the boiling fluid by the 10% volume fraction. The 50% volume fractions of both batches are nearly identical and primarily 1884

dx.doi.org/10.1021/ef200239x |Energy Fuels 2011, 25, 1878–1890

Energy & Fuels

ARTICLE

Table 6. Representative Distillation Curve Data (Given As the Average of Two Distillation Curves) for the Fatty AcidLevulinate Ester Biodiesel Fuel Blends, the Soy Based B100 Sample, and Neat ULSDa batch 1 (83.3 kPa)

a

batch 2 (84.5 kPa)

B100 (82.9 kPa)

ULSD (83.9 kPa)

distillate volume fraction (%)

Tk (° C)

Th (° C)

Tk (° C)

Th (° C)

Tk (° C)

Th (° C)

Tk (° C)

Th (° C)

5

261.5

223.3

316.7

218.7

347.6

345.1

225.5

206.8

10

280.8

227.2

351.5

291.2

350.0

345.9

230.1

212.4

15 20

318.8 354.7

214.0 350.5

356.6 357.7

349.1 351.1

350.7 351.4

346.9 347.3

235.4 240.9

219.3 224.3

25

355.8

352.2

358.5

353.1

351.7

347.8

245.9

229.2

30

356.7

353.0

359.2

352.8

352.2

348.2

251.2

235.2

35

357.1

353.4

359.9

353.0

352.8

348.5

256.5

241.0

40

357.7

353.8

360.5

353.5

353.5

349.2

261.9

246.1

45

358.4

353.8

361.4

354.0

354.0

349.5

268.2

252.0

50

358.7

354.3

362.3

354.3

354.8

349.9

273.8

257.4

55 60

359.6 360.2

355.0 355.4

363.4 364.8

354.7 355.4

355.7 356.7

350.1 350.6

280.2 287.1

264.6 271.7

65

361.1

355.8

366.6

355.7

358.6

350.9

294.3

279.0

70

362.5

356.6

368.9

356.7

360.4

351.4

301.6

285.4

75

364.4

356.8

372.2

357.0

363.4

351.9

310.0

294.7

80

368.1

357.0

377.7

358.0

369.4

351.9

319.2

303.5

85

375.1

357.6

387.1

357.9

382.3

351.9

328.9

312.6

90

391.3

350.7

404.6

350.6

404.5

347.0

-

-

The uncertainties are discussed in the text. These temperatures have been corrected to 1 atm with the Sydney Young equation.

Figure 4. Representative distillation curves of the two batches of fatty acidlevulinate ester biodiesel fuel presented as Tk (the temperature measured in the fluid). The distillation curve for a representative ULSD and a previously measured commercial soy based biodiesel fuel sample52 are also presented. Although only one curve for each fluid is shown, each curve was measured two times. The uncertainties are discussed in the text and are smaller than the symbols used.

composed of the FAEEs. This explains the flat and nearly identical distillation curves in this region. The decrease in concentration of the ethyl linoleate, which is an unsaturated carbon-carbon double bonds, observed in the 80% GC plots supports the thermal compositional transitions previously mentioned.89 The composition-explicit data channel allows the addition of thermochemical data, such as the enthalpy of combustion, to the

distillation curve.50,51,54 The total enthalpy of combustion (which we represent as ΔHc) can be found by multiplying the enthalpy of combustion of each of the pure (or individual) components by the mole fraction of that component, and then adding the contributions of the individual components to obtain the total enthalpy of combustion:50,51,54  ΔHc ¼ 1885

∑xi ð  ΔHi Þ

ð1Þ

dx.doi.org/10.1021/ef200239x |Energy Fuels 2011, 25, 1878–1890

Energy & Fuels

ARTICLE

Figure 5. (a) and (b) Representative chromatograms of four distillate volume fractions of the two fatty acidlevulinate ester biodiesel fuel blends. The major peaks are labeled in the first fraction in which they were observed. The y-axes are arbitrary units of intensity, and the x-axes are retention time (RT) in min. The details of the chromatography are discussed in the text.

1886

dx.doi.org/10.1021/ef200239x |Energy Fuels 2011, 25, 1878–1890

Energy & Fuels

ARTICLE

Table 7. Total Enthalpy of Combustion, Presented in kJ/mol, for Four Distillate Volume Fractions of the Two Fatty AcidLevulinate Ester Biodiesel Fuel Blends, the Soy Based B100, and a ULSD Sample (Uncertainties Are Presented in Parentheses) distillate volume fraction (%)

batch 1

batch 2

B100

ULSD

0.025

3352 (111)

2021 (67)

10403 (343)

5090 (168)

10

3790 (125)

4597 (152)

10468 (345)

6258 (207)

50

11151 (368)

11130 (367)

10579 (349)

9847 (325)

80

11153 (368)

11221 (370)

10689 (353)

10572 (355)

Figure 6. Enthalpy of combustion for the two fatty acidlevulinate ester biodiesel fuel blend batches presented at four distillate cuts, 0.025% (the first drop), 10%, 50%, and 80%. The uncertainties are discussed in the text.

where i refers to the individual components that have been identified or selected, and the enthalpy of mixing is ignored. We have discussed the contributions to the overall uncertainty of the total enthalpy of combustion elsewhere.50,51,54 The main sources of uncertainty in the enthalpy of combustion calculation here are due to (1) uncertainty in the values tabulated for the individual enthalpy of combustion values for each component, (2) uncertainty in the measured mole fraction, and (3) the uncertainty arising from the absence of data for experimental enthalpy of combustion for some of the constituents. There is also uncertainty in neglecting the enthalpy of mixing; however, this value has been shown previously to be less than 0.01% of the enthalpy of combustion. Additionally, there may be uncertainty in the enthalpy of combustion due to the inability to resolve very closely related isomers via the analytical protocol, the complete misidentification of a component, and neglecting components present at very low concentrations. The FLBBs contain a small number of easily resolved and identifiable peaks, thus, these analytical metrology contributions to the uncertainty are also negligible. In past work, we determined that neglecting peaks with total uncalibrated area percentages of up to 4% increased the uncertainty of the calculated enthalpy by only 1.5%. Thus, neglecting minor components in the biodiesel fuel distillate fractions does not significantly affect the uncertainty of the total enthalpy of combustion. In view of these sources of uncertainty, the overall combined uncertainty in our total enthalpy of combustion calculations (with a coverage factor k = 2)90 was less

than 3%.69 The uncertainty is dominated by the analytical measurement and determination of the component mole fractions. The total enthalpy of combustion of both FLBB batches (taking into account peaks with area % in excess of 2%) was calculated at four distillate volume fractions: 0.025% (the first drop), 10%, 50%, and 80%. Table 7 and Figure 6 present the calculated enthalpies of combustion, in kJ/mol, for the distillate volume fractions of both FLBB batches. For comparison, the previously determined enthalpies of combustion for B100 and the ULSD sample are also presented in Table 6. Notice that the molar enthalpy of combustion of the first drop for Batch 2 is less than that for Batch 1. This is due to the higher concentrations of hexane and ethanol remaining in Batch 2 following the processing. The molar energy content of Batch 2 is greater at the 10% volume fraction than Batch 1 because of the smaller amount of ethyl levulinate at this volume fraction. At higher distillate volume fractions, the energies of the two batches are identical because the composition of these distillate fractions is very similar. The presence of ethyl levulinate in the FLBB results in lower enthalpies of combustion in the early volume fractions compared to B100; however, following the exhaustion of the ethyl levulinate, the FLBB energy content becomes very similar to that of B100. An alternative representation of the enthalpy of combustion that is perhaps more practical for applied engineering would be to present ΔHc in terms of mass or volume, expressed in kJ/g 1887

dx.doi.org/10.1021/ef200239x |Energy Fuels 2011, 25, 1878–1890

Energy & Fuels

ARTICLE

Table 8. Total Enthalpy of Combustion, Presented in kJ/g and kJ/L, for Four Distillate Volume Fractions of the Two Fatty AcidLevulinate Ester Biodiesel Fuel Blends and the Soy Based B100 for Comparison (Uncertainties Are Presented in Parentheses) mass-basis distillate volume fraction (%)

batch 1 (kJ/g)

batch 2 (kJ/g)

volume-basis B100 (kJ/g)

batch 1 (kJ/L)

batch 2 (kJ/L)

B100 (kJ/L)

0.025

25 (1)

30 (1)

36 (1)

24752 (817)

23087 (762)

31517 (1040)

10

25 (1)

27 (1)

36 (1)

24881 (821)

26256 (866)

31502 (1040)

50

36 (1)

36 (1)

36 (1)

31447 (1038)

31457 (1038)

31663 (1045)

80

36 (1)

36 (1)

37 (1)

31400 (1036)

31508 (1040)

31772 (1049)

or kJ/L, respectively. The conversion to a per-mass basis requires only the relative molecular mass (RMM) of the constituents of each sample; the uncertainty of this calculation remains less than 3%. Table 8 shows that Batch 2 has a 17% higher per-mass enthalpy of combustion than Batch 1 in the first drop. This is because of the higher concentrations of hexane and ethanol in the first drop of Batch 2 (which contains more energy per-mass than ethyl levulinate) which is the primary component in the first drop of Batch 1. The FAEEs contain more energy than ethyl levulinate, which accounts for the increase in the per-mass enthalpy during the distillation for both FLBB batches. Once the ethyl levulinate has been volatilized, the per-mass energy content is equal to that of the B100 sample. Presentation of the data on a per-volume basis is also valuable. Consumers are more accustomed to thinking about fuel on a pervolume basis instead of a per-mole or per-mass basis. The conversion to a per-volume basis requires both the RMM and the density of the constituents of each sample at each distillate temperature. For the fluids measured here, reliable density data of the constituents of each sample are available as a function of the distillation temperatures, and the uncertainty of this calculation remains less than 3%.91 Table 8 indicates that the first drop of Batch 2 has a slightly lower per-volume enthalpy than that of Batch 1. This is because of the lower energy content per-volume of ethanol, which makes up nearly half of the volumetric concentration of the first drop for Batch 2. Since fatty acid esters have greater per-volume energy content than ethyl levulinate, the total per-volume enthalpies of combustion of the initial volume fractions of the FLBB are less than that of B100. As the distillation progresses and the ethyl levulinate is removed, the FLBB per-volume energy content increases to within the experimental uncertainty of the B100.

’ CONCLUSION Recently, a new processing technique was developed that converts the carbohydrates found in plant biomass into ethyl levulinate, which has properties making it a possible diesel fuel oxygenate additive. The volatility properties of ethyl levulinate/ diesel fuel were investigated using the ADC method and apparatus. Mixtures containing 1, 2.5, 5, 10, and 20% ethyl levulinate (v/v) were analyzed. The distillation curves of the oxygenated mixtures showed the greatest deviation from the ULSD curve early in the distillation, with a more pronounced deviation occurring as the concentration of the ethyl levulinate was increased. In fact, the boiling temperatures of the less concentrated mixtures (i.e., 1 and 2.5 volume percent) deviated very little from those measured for the neat ULSD. Compositionexplicit data analysis performed on each mixture revealed that detectable amounts of ethyl levulinate remained in the boiling fluid until the 60% distillate volume fraction. Based on these data,

it appears that modest additions of ethyl levulinate to diesel fuel will not significantly affect the volatility. The second part of this study focused on biodiesel fuel blends produced from safflower oil seeds via a new processing technique that converts both the oils and the carbohydrates in the seeds into a hybrid biodiesel fuel resulting in an increase in fuel yield compared to traditional processing techniques. The new biodiesel blends contain a mixture of FAEEs, commonly found in most biodiesel fuels, and ethyl levulinate. Two compositions of the new biodiesel fuel blend were analyzed with the ADC and compared to a commercial soy-based biodiesel fuel and ULSD fuel. Distillate fractions were withdrawn to measure the changing composition and energy content during the distillation. Reductions in boiling temperatures for the fatty acidlevulinate ester blends in the early distillate fractions as compared to the commercial biodiesel fuel were noticed. In addition, the enthalpies of combustion in the initial distillate fractions were lower in the fatty acidlevulinate ester blends than the commercial biodiesel. The measurements presented in this paper will assist in the evaluation of ethyl levulinate as a potential additive for ULSD. Moreover, by examining the impact of the addition of ethyl levulinate on biodiesel fuel properties (specifically the volatility), this work will assist in the evaluation of this additive to enhance the cold flow properties of biodiesel fuel.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT T.M.L. and B.C.W. both acknowledge National Academy of Sciences/National Research Council postdoctoral fellowships at the National Institute of Standards and Technology. M.M. and E. B.N. acknowledge the support of National Science Foundation grant CBET 0932391. ’ REFERENCES (1) Ålander, T. J. A.; Leskinen, A. P.; Raunemaa, T. M.; Rantanen, L. Characterization of diesel particles: effects of fuel reformulation, exhaust aftertreatment, and engine operation on particle carbon composition and volatility. Environ. Sci. Technol. 2004, 38, 2707–2714. (2) Han, Z.; Uludogan, A.; Hampson, G. J.; Reitz, R. D. Mechanism of soot and NOx emission reduction using multiple-injection in a diesel engine. SAE Technical Paper 960633, 1996. (3) Mueller, C. J.; Pitz, W. J.; Pickett, L. M.; Martin, G. C.; Siebers, D. L.; Westbrook, C. K. Effects of oxygenates on soot processes in DI diesel engines: experiments and numerical simulations. SAE Technical Paper Series 2003-01-1791, 2003. 1888

dx.doi.org/10.1021/ef200239x |Energy Fuels 2011, 25, 1878–1890

Energy & Fuels (4) Huang, Z. H.; Lu, H. B.; Jiang, D. M.; Zeng, K.; Liu, B.; Zhang, J. Q.; Wang, X. B. Performance and emissions of a compression ignition engine fueled with diesel/oxygenate blends for various fuel delivery advance angles. Energy Fuels 2005, 19, 403–410. (5) Gong, Y. F.; Liu, S. H.; Guo, H. J.; Hu, T. G.; Zhou, L. B. A new diesel oxygenate additive and its effects on engine combustion and emissions. Appl. Therm. Eng. 2007, 27, 202–207. (6) Choi, C. Y.; Reitz, R. D. An experimental study on the effects of oxygenated fuel blends and multiple injection strategies on DI diesel engine emissions. Fuel 1999, 78, 1303–1317. (7) Zheng, M.; Reader, G. T.; Hawley, J. G. Diesel engine exhaust gas recirculation--a review on advanced and novel concepts. Energy Convers. Manage. 2004, 45, 883–900. (8) Ott, L. S.; Smith, B. L.; Bruno, T. J. Composition-explicit distillation curves of mixtures of diesel fuel with biomass-derived glycol ester oxygenates: a fuel design tool for decreased particulate emissions. Energy Fuels 2008, 22, 2518–2526. (9) Smith, B. L.; Ott, L. S.; Bruno, T. J. Composition-explicit distillation curves of diesel fuel with glycol ether and glycol ester oxygenates: fuel analysis metrology to enable decreased particulate emissions. Environ. Sci. Technol. 2008, 42, 7682–7689. (10) Wang, Y.; Zhou, L. B.; Wang, H. W. Diesel emission improvements by the use of oxygenated DME/diesel blend fuels. Atmos. Environ. 2006, 40, 2313–2320. (11) Guo, H. J.; Huang, X. X.; Wang, X. J.; Zhang, Z. L. Research on a new type of oxygenate of 2-methoxy ethyl acetate as a clean diesel fuel additive. Prepr. - Am. Chem. Soc. Div. Pet. Chem. 2004, 227, U239– U240. (12) Miyamoto, N., Ogawa, H, Nurun, M. M., Obata, K, Arima, T. Smokeless, low NOx, high thermal efficiency, and low noise diesel combustion with oxygenated agents as main fuel. SAE Technical Paper 980506, 1998. (13) Knothe, G.; Krahl, J.; Van Gerpen, J. The Biodiesel Handbook; AOCS Press: Champaign, IL, 2005. (14) Yanowitz, J.; McCormick, R. L.; Graboski, M. S. In-use emissions from heavy-duty diesel vehicles. Environ. Sci. Technol. 2000, 34, 729–740. (15) Lapuerta, M.; Armas, O.; Rodreguez-Fernandez, J. Effect of biodiesel fuels on diesel engine emissions. Prog. Energy Combust. Sci. 2008, 34, 198–223. (16) Knothe, G. ”Designer” biodiesel: Optimizing fatty ester composition to improve fuel properties. Energy Fuels 2008, 22, 1358–1364. (17) Knothe, G.; Steidley, K. R. Lubricity of components of biodiesel and petrodiesel. The origion of biodiesel lubricity. Energy Fuels 2005, 19, 1192–1200. (18) Wei, D.; Spikes, H. A. The lubricity of diesel fuels. Wear 1986, 111, 217–235. (19) Barbour, R. H.; Rickeard, D. J.; Elliott, N. G. Understanding diesel lubricity. SAE Technical Paper Series 2000-01-1918, 2000. (20) Sch€onborn, A.; Ladommatos, N.; Allan, R.; Williams, J.; Rogerson, J. Effect of the molecular structure of individual fatty acid alcohol esters (biodiesel) on the formation of NOx and particulate matter in the diesel combustion process. SAE Int. J. Fuels Lubr. 2009, 1, 849–872. (21) Moser, B. R.; Williams, A.; Haas, M. J.; McCormick, R. L. Exhaust emissions and fuel properties of partially hydrogenated soybean oil methyl esters blended with ultra low sulfur diesel fuel. Fuel Process. Technol. 2009, 90, 1122–1128. (22) McCormick, R. L.; Ratcliff, M.; Moens, L.; Lawrence, R. Several factors affecting the stability of biodiesel in standard accelerated tests. Fuel Process. Technol. 2007, 88, 651–657. (23) McCormick, R. L. The impact of biodiesel on pollutant emissions and public health. Inhalation Toxicol. 2007, 19, 1033–1039. (24) Graboski, M. S.; McCormick, R. L. Combustion of fat and vegetable oil derived fuels in diesel engines. Prog. Energy Combust. 1998, 24, 125–164. (25) Fisher, B.; Knothe, G.; Mueller, C. Liquid-phase penetration under unsteady in-cylinder conditions: soy- and cuphea-derived biodiesel fuels versus conventional diesel. Energy Fuels 2010, 24, 5163–5180.

ARTICLE

(26) Cheng, A. S.; Upatnieks, A.; Mueller, C. J. Investigation of the impact of biodiesel fuelling on NOx emissions using an optical direct injection diesel engine. Int. J. Engine Res. 2006, 7, 297–318. (27) Soyatech. 2008 Soya & Oilseed Bluebook; ISSN 1099-7970; 2007; p 346. (28) Tiwari, A.; Kumar, A.; Raheman, H. Biodiesel production from jatropha oil (Jatropha curcas) with high free fatty acids: an optimized process. Biomass Bioenergy 2007, 31, 569–575. (29) Lovestead, T. M.; Windom, B. C.; Bruno, T. J. Investigating the unique properties of cuphea derived biodiesel with the advanced distillation curve method. Energy Fuels 2010, 24, 3662–3675. (30) Knothe, G.; Cermak, S. C.; Evangelista, R. L. Cuphea Oil as Source of Biodiesel with Improved Fuel Properties Caused by High Content of Methyl Decanoate. Energy Fuels 2009, 23, 1743–1747. (31) Johnson, M. B.; Wen, Z. Production of Biodiesel Fuel from the Microalga Schizochytrium limacinum by Direct Transesterification of Algal Biomass. Energy Fuels 2009, 23, 5179–5183. (32) Stroup, R. L. Feedstock Considerations for Future U.S. Producers. Biodiesel Mag. 2004, January, p 10. (33) Usta, N. Use of tobacco seed oil methyl ester in a turbocharged indirect injection diesel engine. Biomass Bioenergy 2005, 28, 77–86. (34) Moser, B.; Knothe, G.; Vaughn, S. F.; Isbell, T. A. Production and evaluation of biodiesel from field pennycress (Thlaspi arvense L.) oil. Energy Fuels 2009, 23, 4149–4155. (35) Kondamudi, N.; Mohapatra, S. K.; Misra, M. Spent Coffee Grounds as a Versatile Source of Green Energy. J. Agric. Food Chem. 2008, 56, 11757–11760. (36) Jham, G.; Moser, B.; Shah, S.; Holser, R.; Dhingra, O.; Vaughn, S.; et al. Wild Brazilian mustard seed oil methyl esters as biodiesel fuel. J Am Oil Chem Soc. 2009, 86, 917–926. (37) Mascal, M.; Nikitin, E. B. Dramatic advancements in the saccharide to 5-(chloromethyl)furfural conversion reaction. ChemSusChem. 2009, 2, 859–861. (38) McCormick, B. Renewable diesel fuels: Status of technology and R&D needs. Presented at 8th Diesel Engine Emissions Reduction Conference August 2529, 2003 Coronado, CA, (http://www1.eere. energy.gov/vehiclesandfuels/pdfs/deer_2002/session4/ 2002_deer_mccormick.pdf). (39) Mascal, M.; Nikitin, E. B. Co-processing of carbohydrates and lipids in oil crops to produce a hybrid biodiesel. Energy Fuels 2010, 24, 2170–2171. (40) Leffler, W. L. Petroleum Refining in Nontechnical Language; PennWell: Tulsa, OK, 2000. (41) Kister, H. Z. Distillation Design; McGraw-Hill: New York, 1991. (42) Kister, H. Z. Distillation Operation; McGraw-Hill: New York, 1988. (43) Riazi, M. R. Characterization and Properties of Petroleum Fractions, ASTM MNL50; American Society for Testing and Materials: West Conshohoken, PA, 2005. (44) Manovyan, A. K.; Khachaturova, D. A.; Lozin, V. V. Method for determination of distillation curve of heavy petroleum products. Chem. Technol. Fuels Oils 1983, 19, 259–261. (45) Karonis, D.; Lois, E.; Zannikos, F.; Alexandridis, A.; Sarimveis, H. A neural network approach for the correlation of exhaust emissions from a diesel engine with diesel fuel properties. Energy Fuels 2003, 17, 1259–1265. (46) Karonis, D.; Lois, E.; Stournas, S.; Zannikos, F. Correlations of exhaust emissions from a diesel engine with diesel fuel properties. Energy Fuels 1998, 12, 230–238. (47) ASTM Standard D 86-04b, Standard Test Method for Distillation of Petroleum Products at Atmospheric Pressure; ASTM International: West Conshohocken, PA, 2004. (48) Bruno, T. J. Improvements in the measurement of distillation curves. Part 1: A composition-explicit approach. Ind. Eng. Chem. Res. 2006, 45, 4371–4380. (49) Bruno, T. J.; Smith, B. L. Improvements in the measurement of distillation curves. Part 2: Application to aerospace/aviation fuels RP-1 and S-8. Ind. Eng. Chem. Res. 2006, 45, 4381–4388. 1889

dx.doi.org/10.1021/ef200239x |Energy Fuels 2011, 25, 1878–1890

Energy & Fuels (50) Smith, B. L.; Bruno, T. J. Improvements in the measurement of distillation curves. Part 3: Application to gasoline and gasoline þ methanol mixtures. Ind. Eng. Chem. Res. 2007, 46, 297–309. (51) Smith, B. L.; Bruno, T. J. Improvements in the measurement of distillation curves. Part 4: Application to the aviation turbine fuel Jet-A. Ind. Eng. Chem. Res. 2007, 46, 310–320. (52) Windom, B. C.; Bruno, T. J. Improvements in the measurement of distillation curves. 5. Reduced pressure composition-explicit approach. Ind. Eng. Chem. Res. 2011, 50 (2), 1115–1126. (53) Smith, B. L.; Bruno, T. J. Advanced distillation curve measurement with a model predictive temperature controller. Int. J. Thermophys. 2006, 27, 1419–1434. (54) Bruno, T. J.; Smith, B. L. Enthalpy of combustion of fuels as a function of distillate cut: application of an advanced distillation curve method. Energy Fuels 2006, 20, 2109–2116. (55) Bruno, T. J. Method and apparatus for precision in-line sampling of distillate. Sep. Sci. Technol. 2006, 41, 309–314. (56) Bruno, T. J.; Ott, L. S.; Smith, B. L.; Lovestead, T. M. Complex fluid analysis with the advanced distillation curve approach. Anal. Chem. 2010, 82, 777–783. (57) Bruno, T. J.; Ott, L. S.; Lovestead, T. M.; Huber, M. L. The composition-explicit distillation curve technique: Relating chemical analysis and physical properties of complex fluids. J. Chromatogr., A. 2010, 1217, 2703–2715. (58) Huber, M. L.; Lemmon, E.; Bruno, T. J. Effect of RP-1 compositional variability on thermophysical properties. Energy Fuels 2009, 23, 5550–5555. (59) Huber, M. L.; Lemmon, E.; Ott, L. S.; Bruno, T. J. Preliminary surrogate mixture models for rocket propellants RP-1 and RP-2. Energy Fuels 2009, 23, 3083–3088. (60) Huber, M. L.; Lemmon, E.; Kazakov, A.; Ott, L. S.; Bruno, T. J. Model for the thermodynamic properties of a biodiesel fuel. Energy Fuels 2009, 23, 3790–3797. (61) Huber, M. L.; Lemmon, E. W.; Diky, V.; Smith, B. L.; Bruno, T. J. Chemically authentic surrogate mixture model for the thermophysical properties of a coal-derived-liquid fuel. Energy Fuels 2008, 22, 3249–3257. (62) Huber, M. L.; Smith, B. L.; Ott, L. S.; Bruno, T. J. Surrogate Mixture Model for the Thermophysical Properties of Synthetic Aviation Fuel S-8: Explicit Application of the Advanced Distillation Curve. Energy Fuels 2008, 22, 1104–1114. (63) Lovestead, T. M.; Windom, B. C.; Nickell, J. R. R. C.; Bruno, T. J. Assessment of the Compositional Variability of RP-1 and RP-2 with the Advanced Distillation Curve Approach. Energy Fuels 2010, 24, 5611–5623. (64) Smith, B. L.; Ott, L. S.; Bruno, T. J. Composition-explicit distillation curves of commercial biodiesel fuels: comparison of petroleum derived fuel with B20 and B100. In. Eng. Chem. Res. 2008, 47, 5832–5840. (65) Smith, B. L.; Bruno, T. J. Application of a Composition-Explicit Distillation Curve Metrology to Mixtures of Jet-A þ Synthetic Fischer Tropsch S-8. J. Propul. Power 2008, 24, 619–623. (66) Smith, B. L.; Bruno, T. J. Composition-explicit distillation curves of aviation fuel JP-8 and a coal based jet fuel. Energy Fuels 2007, 21, 2853–2862. (67) Smith, B. L.; Ott, L. S.; Bruno, T. J. Composition-explicit distillation curves of diesel fuel with glycol ether and glycol ester oxygenates: a design tool for decreased particulate emissions. Environ. Sci. Technol. 2008, 42, 7682–7689. (68) Ott, L. S.; Smith, B. L.; Bruno, T. J. Advanced distillation curve measurements for corrosive fluids: application to two crude oils. Fuel 2008, 87, 3055–3064. (69) Ott, L. S.; Bruno, T. J. Variability of biodiesel fuel and comparison to petroleum-derived diesel fuel: application of a composition and enthalpy explicit distillation curve method. Energy Fuels 2008, 22, 2861–2868. (70) Ott, L. S.; Hadler, A.; Bruno, T. J. Variability of the rocket propellants RP-1, RP-2, and TS-5: application of a composition- and

ARTICLE

enthalpy-explicit distillation curve method. Ind. Eng. Chem. Res. 2008, 47, 9225–9233. (71) Ott, L. S.; Smith, B. L.; Bruno, T. J. Experimental test of the Sydney Young equation for the presentation of distillation curves. J. Chem. Thermodynam. 2008, 40, 1352–1357. (72) Ott, L. S.; Smith, B. L.; Bruno, T. J. Advanced distillation curve measurement: application to a bio-derived crude oil prepared from swine manure. Fuel 2008, 87, 3379–3387. (73) Ott, L. S.; Bruno, T. J. Modifications to the copper strip corrosion test for the measurement of microscale samples. J. Sulfur Chem. 2007, 28, 493–504. (74) Ott, L. S.; Bruno, T. J. Corrosivity of fluids as a function of distillate cut: application of an advanced distillation curve method. Energy Fuels 2007, 21, 2778–2784. (75) Lovestead, T. M.; Bruno, T. J. Application of the advanced distillation curve method to aviation fuel avgas 100LL. Energy Fuels 2009, 23, 2176–2183. (76) Lovestead, T. M.; Bruno, T. J. Comparison of the hypersonic vehicle fuel JP-7 to the rocket propellants RP-1 and RP-2 with the advanced distillation curve method. Energy Fuels 2009, 23, 3637–3644. (77) Bruno, T. J.; Wolk, A.; Naydich, A.; Huber, M. L. Composition explicit distillation curves for mixtures of diesel fuel with dimethyl carbonate and diethyl carbonate. Energy Fuels 2009, 23, 3989–3997. (78) Bruno, T. J.; Wolk, A.; Naydich, A. Composition-explicit distillation curves for mixtures of gasoline with four-carbon alcohols (butanols). Energy Fuels 2009, 23, 2295–2306. (79) Bruno, T. J.; Laesecke, A.; Outcalt, S. L.; Seelig, H.-D.; Smith, B. L. Properties of a 50/50 Mixture of Jet-A þ S-8, NIST-IR-6647; National Institute of Standards and Technology: Washington, DC, 2007. (80) Bruno, T. J.; Huber, M. L.; Laesecke, A.; Lemmon, E. W.; Perkins, R. A. Thermochemical and thermophysical properties of JP-10; NIST-IR 6640; National Institute of Standards and Technology (U.S.), 2006. (81) Bruno, T. J.; Wolk, A.; Naydich, A. Stabilization of biodiesel fuel at elevated Temperature with Hydrogen Donors: evaluation with the advanced distillation curve method. Energy Fuels 2009, 23, 1015–1023. (82) Bruno, T. J. Thermodynamic, transport and chemical properties of “reference” JP-8. In Book of Abstracts, Army Research Office and Air Force Office of Scientific Research, 2007 Contractor’s meeting in Chemical Propulsion, 2007. (83) NIST/EPA/NIH Mass Spectral Database 2005; SRD Program, National Institute of Standards and Technology, Gaithersburg, MD. (84) Lide, D. R. CRC Handbook of Chemistry and Physics, 85th ed.; Taylor and Francis CRC Press: Boca Raton, FL, 20042005. (85) Bruno, T. J.; Svoronos, P. D. N. CRC Handbook of Fundamental Spectroscopic Correlation Charts; Taylor and Francis CRC Press: Boca Raton, FL, 2005. (86) Young, S. Distillation Principles and Processes; Macmillan and Co., Ltd.: London, 1922. (87) Young, S. Fractional Distillation; Macmillan and Co., Ltd.: London, 1903. (88) Young, S. Correction of boiling points of liquids from observed to normal pressures. Proc. Chem. Soc. 1902, 81, 777. (89) Waynick, J. A. Characterization of Biodiesel Oxidation and Oxidation Products: CRC Project No. AVFL-2b; NREL/TP-540-39096: National Renewable Energy Laboratory: Golden, CO, 2005. (90) Taylor, B. N.; Kuyatt, C. E. Guidelines for Evaluating and Expressing the Uncertainty of NIST Measurement Results; NIST Technical Note 1297; Washington, DC, 1994. (91) Rowley, R. L.; Wilding, W. V.; Oscarson, J. L.; Zundel, N. A.; Marshall, T. L.; Daubert, T. E.; Danner, R. P. DIPPR Data Compilation of Pure Compound Properties; Design Institute for Physical Properties AIChE: New York, 2004.

1890

dx.doi.org/10.1021/ef200239x |Energy Fuels 2011, 25, 1878–1890