Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2021 Mar 25;6(13):9204–9212. doi: 10.1021/acsomega.1c00556

Locally Sustainable Biodiesel Production from Waste Cooking Oil and Grease Using a Deep Eutectic Solvent: Characterization, Thermal Properties, and Blend Performance

Neelam Khan †,*, Sang H Park †, Lorraine Kadima †, Carlove Bourdeau †, Evelyn Calina †, Charles Warren Edmunds ‡, David P Pursell †
PMCID: PMC8028156  PMID: 33842789

Abstract

graphic file with name ao1c00556_0010.jpg

As part of local sustainability efforts, biodiesel was synthesized via transesterification using a deep eutectic solvent (DES) without further washing from on-campus, dining facility waste cooking oil and grease. Before moving forward with repurposing used DES as a solvent in chemistry teaching labs, we determined the suitability of the biodiesel as an alternative fuel blended with diesel to power campus utility vehicles. Biodiesel components were characterized by gas chromatography–mass spectrometry (GC–MS), Fourier transform infrared spectroscopy (FTIR), nuclear magnetic resonance spectroscopy (1H NMR), viscometer, differential scanning calorimetry (DSC), and evolved gas analysis during pyrolysis with a thermogravimetric analyzer coupled with FTIR (TGA-FTIR). The four major components of fatty acid methyl esters (FAMEs) in the biodiesel were methyl oleate, methyl linoleate, methyl palmitate, and methyl stearate. Kinematic viscosity over typical temperature ranges was within optimal values recommended by the American Biodiesel Standard (ASTM D6751), with a 30:70 biodiesel/diesel blend experimental viscosity of 3.43 cSt at 40 °C and a calculated viscosity of 10.13 cSt at 0 °C. The pure biodiesel’s cold-temperature onset of crystal formation is −10.1 °C versus −16.4 °C for a 30:70 biodiesel/diesel blend. Pyrolysis indicates good thermal stability, however, with an increased CO2 evolution in the blended fuel at higher temperatures as compared to that in the pure biodiesel and the pure diesel. Combustion gas analysis indicates virtually complete combustion of the blended fuel to CO2 and H2O with only trace amounts of CO. Overall results indicate that the biodiesel synthesized using DES is a suitable fuel for campus utility vehicles in the local moderate temperature climate and affords increased local sustainability by using used DES repurposed in our chemistry teaching labs.

1. Introduction

Biofuels have achieved a measure of acceptance and are used across many industries, including the military. In fact, the U.S. Navy operated an entire fleet with an aircraft carrier, its aircraft, and support ships using biofuels; the fleet performed at the same level as when using petroleum-based fuels, although at a higher cost.1 Biofuels are produced from a variety of sources to include vegetable oils, ethanol, algae, wood stock, and waste products from human activities. Biodiesel is a sustainable alternative to petroleum diesel fuel, especially when using feedstock from nonedible oils that do not compete with food production, which is particularly important in developing economies.2 To increase production capacity from nonedible feedstocks, there has been much work recently on enzymatic conversion of sea-borne algae into biodiesel.3 Nevertheless, investigations continue in more traditional approaches using transesterification of triglycerides, from either pure oils or waste cooking oil and grease, into fatty acid methyl/ethyl esters.4−7 The generation of glycerol-based byproducts of transesterification synthesis is problematic, as these byproducts often require additional resources and further processing to become useful materials rather than waste needing disposal, all of which increases costs and reduces sustainability aspects of biodiesel.8

One approach to minimize nonuseful byproducts of transesterification is by incorporating deep eutectic solvents (DESs) in the production process, such as choline chloride/glycerol in the presence of a base. The DES efficiently extracts excess glycerol from the biodiesel product, reducing production costs associated with isolating, washing, and drying the biodiesel product. In addition, the DES may be recycled for use multiple times in the production process.9−13 Many local communities, including college campuses such as ours, emphasize local sustainability practices. Biodiesel synthesis from waste cooking oil and grease using DES supports these practices as the biodiesel may be used as a locally sourced alternative to petroleum diesel to power campus operations, maintenance, and security utility vehicles.14 As an added benefit, after multiple uses in the biodiesel production process, the used DESs may be repurposed as green solvents in our undergraduate chemistry teaching laboratories for Diels–Alder, olefin cross-metathesis, ring-closing metathesis, and many other reactions, further reducing operating costs of our chemistry program while emphasizing local sustainability practices.9,15,16

Our objective with this current investigation is to ensure that biodiesel produced from our local dining hall waste cooking oil and grease via base-catalyzed transesterification using a DES is suitable for use as a fuel and is of a character similar to our biodiesel produced without DES. To accomplish this, we will report on the composition of the biodiesel and biodiesel/diesel blends by gas chromatography–mass spectrometry (GC–MS) and nuclear magnetic resonance spectroscopy (1H NMR) and their key properties of viscosity (viscometer), cold-temperature crystallization (differential scanning calorimetry (DSC)), pyrolysis, and evolved gasses (thermogravimetric analysis-Fourier transform infrared spectroscopy (TGA-FTIR)) and finally a brief analysis of combustion exhaust gas (FTIR) while using the biodiesel/diesel blended fuel to power a diesel generator.

2. Results and Discussion

2.1. GC–MS Analysis

The fatty acid methyl esters (FAMEs) profile of the biodiesel generated from waste cooking oil and grease was determined by the GC–MS analysis method described in Section 4. The retention times of individual peaks of the gas chromatogram were verified against a FAME standard mixture, and individual FAMEs were identified using the MS database (NIST library data). The combined waste cooking oil and grease was generated from pure canola oil and pure peanut oil used in the cooking process of the campus dining facility. C18 fatty acids are the major fatty acid constituents of the synthesized biodiesel with the identified FAMEs being methyl oleate, methyl linoleate, methyl palmitate, and methyl stearate, and their analysis and identification are consistent with previous studies.14,17,18 The biodiesel synthesized using DES FAME composition is shown in Table 1, and GC chromatogram is shown in Figure 1.

Table 1. FAME Composition of Biodiesel Synthesized from Waste Cooking Oil and Grease Using DES.

retention time (min) FAMEs structure molecular weight weight (%)
12.60 methyl palmitate C16:0 270 13.7
14.08 methyl linoleate C18:2 (9Z, 12Z) 294 24.3
14.13 methyl oleate C18:1 (Cis-9) 296 53.7
14.36 methyl stearate C18:0 298 8.3
Open in a new tab

Figure 1.

Figure 1

Open in a new tab

GC chromatogram of fatty acid methyl esters of biodiesel synthesized from waste cooking oil and grease (DB-5 ms column).

2.2. NMR Analysis

The 1H NMR results [δ/ppm] are as follows: δ = 0.88, triplet, 3H, −CH3; δ = 1.28, singlet, 22H, aliphatic CH2; δ = 2.02, multiplet, 2H, −CH2–Csp2H; δ = 2.22, triplet, 2H, O=C–CH2–; δ = 2.80, triplet, 1H, −HCsp2–CH2–Csp2H–; δ = 3.55, singlet, 3H, CH3–O–; and δ = 5.31, multiplet, 2H, −CH=CH–. These results are consistent with our 1H NMR (Bruker, 600 MHz) results for a mixture of FAME molecules resulting from the transesterification of waste cooking oil and grease without using DES, as well as other biodiesel 1H NMR results.7,14,19,20

2.3. Temperature-Dependent Kinematic Viscosity

Figure 2 shows the variation of experimental kinematic viscosities for B100, B30, and ultralow sulfur diesel (ULSD) with temperature. Kinematic viscosities for B100 and ULSD were 5.31 and 2.59 cSt at 40 °C, meeting the B100 viscosity range (1.9–6 cSt) at 40 °C from the American Biodiesel Standard (ASTM D6751).21 The kinematic viscosity of B30 at 40 °C was 3.43 cSt, which is comparable to our previous result of 3.1 cSt for biodiesel synthesis without DES.14 It increases with a decrease in temperature, which is important in cold weather environments as high viscosity impacts flow through the fuel system and particularly fuel injection and nebulization.

Figure 2.

Figure 2

Open in a new tab

Variation of experimental kinematic viscosity with temperature for ULSD, B100, and B30.

To further investigate the low-temperature viscosity of our biodiesel using the DES, we used the experimental data to develop a fitted equation that enables the calculation of viscosity in the temperature regime where DSC data indicates the onset of crystallization. The experimental values of viscosity were fitted according to eq 1, which is the modified Andrade-type regression equation to characterize the temperature dependence of kinematic viscosity22−24

2.3. 1

where η is the kinematic viscosity in cSt, A, B, and C are constants for each specific fluid, and T is the temperature in K. Equation 1 can be considered as a second-order polynomial equation in 1/T. Figure 3 shows the graph of loge(η) and inverse temperature (T).

Figure 3.

Figure 3

Open in a new tab

Variation of loge of experimental kinematic viscosity with inverse temperature for ULSD, B100, and B30.

The values of constants A, B, and C are derived using the polynomial curve fitting of the graph. Table 2 reports the values of constants for ULSD, B100, and B30. R2 values indicate that our experimental data is a good fit with the regression equation (eq 1).

Table 2. Viscosity Correlation Constants for ULSD, B100, and B30.

fuel type A B C R2
ULSD –2.8799 587.19 191 412 0.9999
B30 0.2209 –1446.9 551 532 0.9982
B100 1.2555 –1872.7 626 512 0.9998
Open in a new tab

Using eq 1 and correlation constants, we calculated low-temperature kinematic viscosity values in a temperature range −20 to 10 °C, as shown in Table 3. Previous researchers have experimentally determined low-temperature viscosity for low sulfur petroleum diesel, pure FAMEs (to include the individual components of our biodiesel), and biodiesel/diesel blends.25

Table 3. Comparison of Experimental and Calculated Kinematic Viscosities.

    experimental kinematic viscosity (cSt)
 calculated kinematic viscosity (cSt) Andrade’s equation
 absolute-difference (exp – calc)
temp (K) temp (°C) ULSD B100 B30 ULSD B100 B30 ULSD B100 B30
253.15 –20       11.3 37.9 22.5      
263.15 –10       8.29 24.20 14.7      
273.15 0       6.27 16.4 10.13      
283.15 10       4.86 11.7 7.32      
296.15 23 3.60 7.9 5.03 3.62 7.96 5.07 0.02 0.06 0.04
303.15 30 3.14 6.71 4.27 3.13 6.65 4.26 0.01 0.06 0.01
313.15 40 2.59 5.31 3.43 2.58 5.28 3.40 0.01 0.03 0.03
333.15 60 1.82 3.59 2.38 1.84 3.59 2.33 0.02 0.00 0.05
353.15 80 1.38 2.63 1.66 1.37 2.65 1.72 0.01 0.02 0.06
373.15 100 1.07 2.1 1.38 1.08 2.09 1.36 0.01 0.01 0.02
Open in a new tab

The two major components of our biodiesel are methyl oleate and methyl linoleate. For a 50:50 blend of methyl oleate/methyl linoleate, these researchers obtained a kinematic viscosity of 19.50 cSt at −10 °C and 13.05 cSt at 0 °C, which are comparable to our calculated values at these temperatures. For a 30:70 blend of methyl linoleate/commercial low sulfur diesel, they obtained a kinematic viscosity of 9.51 cSt at 0 °C, comparable to our calculated value of 10.13 cSt at 0 °C for B30. Our calculated low-temperature kinematic viscosity is therefore reasonable, especially for 0 °C and higher and, in this case, close to the accepted range of kinematic viscosity for biodiesel at 40 °C. From this, we anticipate that our biodiesel production using DES produces a B30 blended fuel with viscosity properties that enable the operation of campus vehicles virtually year-round in the moderate Georgia climate.

2.4. DSC Analysis

The B100 DSC cooling curve in Figure 4 shows two events. The first event begins with a peak onset at −10.1 °C and includes a second peak with an onset at −18.7 °C. These two peaks are part of the broad initial first event in the B100 cooling curve, may be viewed as the onset of liquid–liquid crystallization in sequence of the samples’ saturated fatty acids methyl stearate (C18:0; 8.2%) with a ΔH of −6.62 J/g (−ΔH exothermic) followed by methyl palmitate (C16:0; 13.7%) with a ΔH of −19.9 J/g. The second event in the B100 curve is an intense and sharp peak beginning at −57.6 °C and represents freezing of the unsaturated components methyl linoleate (C18:2-9Z,12Z; 24.3%) and methyl oleate (C18:1-9Z; 53.7%) with a ΔH of −54.8 J/g. The ULSD cooling curve shows one event with an onset of −19.2 °C for a sharp initial peak followed by a broad tail and represents the onset of crystal formation with a ΔH of −6.66 J/g. The B30 cooling curve shows one event in a broad feature with an onset of −16.4 °C and represents the onset of crystal formation in the blended fuel with a ΔH of −13.2 J/g. The DSC results for B100 produced from waste cooking oil and grease are consistent with other works. B100 has a total degree of unsaturation (DU) of 102.5, calculated following the procedure of other researchers.6,26−30 The B100 sample included two distinct features, including a broad peak for the first event with an onset of −10.1 °C and the second event exhibited a sharp feature with an onset at −57.6 °C. These are consistent with other findings, indicating a higher-onset temperature for saturated components and a lower-onset temperature for unsaturated components.27,31 The B100 onset of −10.1 °C is in general agreement with the DU trends found for methyl tallowate (DU 73.6, onset 7.0°C), soybean oil methyl ester (DU 84.5, onset 2.20 °C), hemp methyl ester (DU 128.5, onset −5.27 °C), methyl soyate (DU 143.8, onset −6.5 °C), and Aleurties moluccanus methyl ester (DU 180.4, onset −9.96 °C).30,31,34

Figure 4.

Figure 4

Open in a new tab

DSC cooling curves with peak onsets 1, 2, and 3 indicated.

The addition of our biodiesel to ULSD to produce the B30 blend decreases the crystallization onset temperature by 6.3 °C compared to that of pure biodiesel, from −10.1 °C for B100 to −16.4 °C for B30, and is consistent with similar trends for biodiesel/diesel blends versus diesel found by other researchers.6,30,33 These observations support the idea that crystallization of these fuels depends on the spatial arrangement, length, and degree of saturation of the fatty acid chains, which inhibit the molecular packing leading to lower crystallization onset temperatures, and that adding diesel to biodiesel results in a crystallization onset temperature lower than pure biodiesel but higher than pure diesel.5,26,30,27,32,35 The DSC cooling curve results shown in Figure 4 are summarized in Table 4.

Table 4. DSC Cooling Curve Data Summary.

  event 1
 event 2
  onset T (°C) peak T (°C) ΔH (J/g) onset T (°C) peak T (°C) ΔH (J/g) onset T (°C) peak T (°C) ΔH (J/g)
ULSD –19.2 –20.7 –6.66            
B30 –16.4 –29.1 –13.2            
B100 –10.1 –12.5 –6.62 –18.7 −22.0 –19.9 –57.6 –58.5 –54.8
Open in a new tab

2.5. TGA-FTIR Analysis

The TGA weight loss and derivative weight loss curves shown in Figure 5 indicate a continuous weight loss event for ULSD, B100, and B30 with the samples containing biodiesel showing stages in the single weight loss event. The ULSD curve shows one continuous weight loss event with an onset of 119 °C, maximum weight loss at 211 °C, and encompasses 98.6% of the weight. The B100 curve shows one continuous weight loss event with three stages, a major stage followed by two smaller stages occurring at higher temperatures. The B100 major weight loss stage has an onset of 230 °C, maximum weight loss at 264 °C, and encompasses 86.6% of the weight. The onset of the B100 second weight loss stage is 296 °C, maximum weight loss at 341°C, and encompasses 10.4% of the weight. The third stage for B100 has an onset of 391 °C, maximum weight loss at 437 °C, and encompasses 2.9% of the weight. The continuous B100 weight loss event with three stages encompasses 99.9% of the total weight.

Figure 5.

Figure 5

Open in a new tab

TGA weight loss and derivative weight loss curves.

The B30 curve has a single weight loss event with two stages. The B30 first stage onset is 154 °C with a maximum weight loss at 248 °C and encompasses 93.5% of the weight, while the second stage onset is 337 °C with a max weight loss at 387 °C and encompasses 5.9% of the weight. The continuous B30 weight loss event with two stages encompasses 99.4% of the weight. This second stage of the B30 weight loss is noteworthy as it corresponds to increased CO2 evolution and will be discussed in more detail in the TGA-FTIR evolved gas analysis. The TGA weight loss results are summarized in Table 5.

Table 5. TGA Weight Loss Curve Data Summary.

  stage 1
 stage 2
 stage 3
  onset T (°C) max wt loss T (°C) wt loss (%) onset T (°C) max wt loss T (°C) wt loss (%) onset T (°C) max wt loss T (°C) wt loss (%)
ULSD 119 211 98.6            
B30 154 248 93.5 337 387 5.9      
B100 230 264 86.6 296 341 10.4 391 437 2.9
Open in a new tab

Researchers have found similar weight loss curves for diesel, biodiesel/diesel blends, and biodiesel. Diesel typically has the lowest onset temperature (50–120 °C depending on the source and composition of the diesel), followed by biodiesel/diesel blends, which have a higher-onset temperature than diesel. Biodiesel typically has a higher-onset temperature than the blends, and depending on the biodiesel source and DUs, the onset ranges from 150 to 230 °C.29,38−44 These researchers noted only one continuous weight loss event for diesel, biodiesel/diesel blends, and biodiesel. However, researchers examining waste cooking oil and grease biodiesel with a DU of 110.58 found a single continuous weight loss event, while their derivative weight loss curve exhibited four stages, similar to the three stages we describe.26 For their biodiesel, they found the first stage onset of 191.9 °C with a max weight loss at 270 °C and although not reported but displayed in their derivative weight loss curves, their second stage has a max weight loss at 380 °C, the third stage has a max weight loss at 430 °C, and the fourth stage has a max weight loss at 550 °C. These observations are consistent with the findings for our B100 from waste cooking oil and grease with a DU of 102.5, the first stage onset at 230 °C, the second stage max weight loss at 341 °C, and the third stage max weight loss at 437 °C. In total, these results are consistent with the notion that the larger and more complex the molecules in the fuel—including chain length, degree of unsaturation, E/Z confirmation, and extent of intermolecular attractive forces—the higher the onset temperature and, in general, reflect the colligative property of the boiling point elevation of solute/solvent mixtures.

The FTIR and TGA-FTIR spectra in Figure 6 provide information about the liquid samples of ULSD, B30, and B100 and the evolved gasses during weight loss events upon heating. The ULSD liquid shows the typical C–H alkane stretching modes at 2950, 2922, and 2850 cm–1; C–H alkane bending, scissor, and wagging modes at 1458 and 1378 cm–1; it lacks the C–H alkene stretching modes at just greater than 3000 cm–1, indicating that the ULSD is composed of fully saturated hydrocarbons; it lacks the strong and broad feature of O–H stretching modes at 3400 cm–1, indicating that the ULSD is free of water; and important for contrast with the B30 and B100 samples, it lacks the C=O ester stretching mode at 1745 cm–1. The B100 liquid shows the C–H alkene mode at 3017 cm–1, indicating the unsaturation of components in the mixture in agreement with the GC–MS and 1H NMR results; the C–H alkane stretching modes at 2950, 2922, and 2850 cm–1; C–H alkane bending, scissor, and wagging modes at 1458 and 1378 cm–1; it has an intense C=O methyl ester mode at 1745 cm–1; the O–CH3 asymmetric deformation mode of the methyl ester at 1435 cm–1; and quite significantly, it lacks the strong and broad feature of the O–H stretching mode at 3400 cm–1, indicating that the B100 is free of water and glycerol, even though our production process with DES included only the separatory funnel segregation of components and no washing and drying of the biodiesel product. As expected, the B30 liquid spectra possess modes of both the ULSD and the B100 and it is free of water and glycerol as well. The FTIR spectra of Figure 6 for the first stage max weight loss events of ULSD (211 °C), B30 (248 °C), and B100 (264 °C) are essentially identical to those for the corresponding liquid ULSD, B30, and B100, indicating volatilization of the sample components during the pyrolysis. For all three samples, the majority of the weight is lost in this first stage as discussed in the TGA results and is summarized in Table 5.

Figure 6.

Figure 6

Open in a new tab

FTIR spectra of liquid samples compared to those of gas phase at maximum TGA weight loss events.

In addition, the first stage max weight loss of the ULSD (211 °C) exhibits a small quantity of CO2, indicated by the CO2 asymmetric stretching mode at 2343 cm–1 and the CO2 bending mode at 668 cm–1. In contrast to the ULSD, the first stage max weight loss of the B30 (248 °C) and B100 (264 °C) displays only the slightest hint of these CO2 vibrational modes. Figure 7 shows 3-D (intensity–wavenumber–temperature) plots and demonstrates that as the temperature is increased after the first stage weight loss event for the ULSD, the intensity of C–H alkane bending, scissor, and wagging modes at 1458 and 1378 cm–1 is significantly reduced and in proportion to the reduction of the C–H alkane stretching modes at 2950, 2922, and 2850 cm–1.

Figure 7.

Figure 7

Open in a new tab

TGA-FTIR spectra of liquid samples and sample gas evolved over the heating range of 30–650 °C.

The intensity of the ULSD CO2 modes increases slightly to a relatively constant but small amount up to a maximum temperature of 650 °C. As the temperature is increased for the B100, stage 2 and stage 3 of the continuous weight loss events occur with the continued presence of the ester mode at 1745 cm–1, indicating that carbonyl compounds, and perhaps the full FAMEs, continue to volatilize even up to a maximum temperature of 650 °C. For B100, the intensity of the CO2 vibrational modes remains constant and very low throughout the entire heating range, indicating very little production of CO2 during pyrolysis. From these differences in the ULSD and B100 curves, one may conclude that the ULSD produces more CO2 than the B100 as the pyrolysis continues to higher temperatures. These findings are similar to those of other researchers.26−30,31−38,36−45

The spectra in Figure 7 for B30 throughout the temperature range present some interesting contrasts to those of the ULSD and B100. One would expect the B30 to demonstrate characteristics of both the ULSD and B100 pyrolysis spectra. Like the B100 spectrum, the B30 spectrum does indicate that most of the ester is lost in the initial stage of weight loss, but a very small amount of the ester is lost in the higher-temperature second stage as well, as indicated by the very low intensity ester C=O stretching mode at 1745 cm–1 at higher temperatures. Like the ULSD spectrum, the B30 spectrum also indicates that most of the diesel components are lost in the initial stage of weight loss, indicated by the sharp reduction in the C–H stretching modes and C–H bending, scissor, and wagging modes with only a slight intensity of these modes through the maximum heating temperature of 650 °C. However, unlike the ULSD and B100 spectra, the B30 spectrum has a significant increase in CO2 evolution beginning at the stage 2 onset temperature of 337 °C, as indicated by the large increase in intensity of the 2343 and 668 cm–1 vibrational modes of CO2. This large increase in CO2 evolution for B30, in comparison to that for ULSD and B100, continues from the stage 2 onset at 337 °C through heating all the way to 650 °C.

To further investigate this difference in CO2 evolution for the B30 compared to that for B100, we examined the integrated area of the intensity of the ester C=O stretching mode at 1745 cm–1 and the CO2 asymmetric stretching mode at 2342 cm–1 for B30 and B100 over the entire pyrolysis heating range from 30 to 650 °C. While the integrated area of the intensity of these modes is not a direct indication of the amount or concentration in the samples, comparison of these intensities does provide an indication of the relative differences in amounts. Based on the 30% biodiesel composition of the B30, one would expect the integrated area of the ester mode throughout the temperature range from 30 to 650 °C in the B100 to be approximately 3.3 times greater than the integrated area of the ester mode throughout the temperature range in the B30. Our integrated area ratio result was 3.7, slightly larger than expected and indicating that perhaps some of the C=O in the ester moiety in B30 was evolving in some manner other than the direct volatilization of the FAME of the biodiesel. Similar analysis for the integrated intensity of the CO2 mode over the entire heating range from 30 to 650 °C indicates that the B30 produces 8.2 times more CO2 than the B100, as shown qualitatively in Figure 7. The ester C=O stretching mode and the CO2 antisymmetric stretching mode we are comparing and contrasting are both intense features in the IR spectrum because of their large transition dipole moments, but one must be cautious to directly compare the two modes to determine absolute quantities or amounts of components. What is apparent is that the B30 ester C=O stretching mode is perhaps slightly less than expected, while the B30 CO2 asymmetric stretching mode is substantially larger than expected in comparison to the B100. While these results require further investigation, one might perhaps conclude that in the B30 there is a chemical rearrangement or reaction between the diesel and FAME components that results in some of the ester moiety of the FAME evolving as CO2 rather than as the carbonyl compound. As noted, this observation requires further investigation before drawing fully substantiated conclusions.

2.6. FTIR of Combustion Exhaust Gas Analysis

The FTIR combustion exhaust gas spectrum in Figure 8 shows signatures of three distinct gasses, CO2, H2O, and CO.46−50 At the center of the spectrum is the sharp, intense feature of the CO2 asymmetric stretch at 2348 cm–1 from combustion of the B30 in the generator. Just to the right of this CO2 stretch is the low-intensity asymmetric stretch feature for 13CO2 centered at 2284 cm–1.

Figure 8.

Figure 8

Open in a new tab

FTIR of diesel generator combustion exhaust gas fueled with B30. The inset shows the rotation–vibration band of CO.

The other major features in the spectrum correspond to water, the combustion gas partner of CO2, with the H2O asymmetric stretch at 3754 cm–1, the symmetric stretch at 3655 cm–1, and the broad bending–rotation band at 1594 cm–1. In addition to CO2 and H2O, the spectrum also shows a low-intensity CO rotation–vibration band at 2143 cm–1 with the inset showing the distinct rotation–vibration signature of heteronuclear diatomic CO.3 There is no ester C=O stretching mode at 1743 cm–1 and only very slightest intensity of C–H alkane and alkene stretching modes from 2800 to 3200 cm–1 and CH4 modes at 3030 cm–1, indicating virtually complete combustion of the B30 in the generator to CO2 and H2O with insignificant amounts of other combustion products. These results are consistent with our combustion exhaust gas analysis using fuel synthesized from waste cooking oil and grease without using DES.14

3. Conclusions

Synthesis of biodiesel from waste cooking oil and grease using DES efficiently produces a locally sourced, sustainable fuel. The biodiesel requires no washing during production, offers multiple use recycling of the DES in repeated synthesis, and provides subsequent repurposing of the DES as a solvent for use in teaching laboratories. The GC–MS and 1H NMR results indicate that the biodiesel synthesized using DES is a blend of saturated and unsaturated FAMEs and in both, the pure biodiesel form and as a biodiesel/diesel blended fuel meet viscosity standards under normal climate temperatures. As cold-temperature operations are a practical concern, crystallization onset temperature determined via DSC is −10.1 °C for B100 and −16.4 °C for B30 as compared to that of −19.2 °C for ULSD, sufficient for year-round operations in Georgia and other similar climates. Likewise, the calculated cold-temperature viscosity for B30 at 0 °C is 10.13 cSt, sufficient for cold-temperature flow through the vehicle fuel lines and injection system. TGA-FTIR pyrolysis experiments up to 650 °C demonstrate good thermal stability for B100. The evolved gas analysis presents the interesting phenomenon for B30, during stage 2 of the weight loss, of perhaps chemical rearrangement or reaction between the diesel and the biodiesel FAME components to produce a larger than expected amount of CO2 compared to pure B100 and pure diesel. The gas-phase FTIR of combustion exhaust while using the B30 fuel to power a diesel generator indicates nearly complete combustion to CO2 and H2O with only a very small quantity of CO and virtually no other combustion products. As a result, we conclude that the biodiesel synthesized from dining operations, waste cooking oil and grease using DES, is suitable for use as a blended fuel to power campus utility vehicles year-round in the local climate and contributes to our goal of local sustainability.

4. Experimental Section

4.1. Materials

Waste cooking oil and grease from the local on-campus fast food restaurants were the feedstock for the synthesized biodiesel using a DES. Methanol (certified ACS, Fisher Chemical), sodium hydroxide (certified ACS, Fisher Chemical), glycerol (certified ACS, Fisher Chemical), choline chloride (>98%, Alfa Aesar), a standard mixture of 37 FAMEs (Restek Corporation), and dichloromethane (high-performance liquid chromatography (HPLC) grade, Acros) were purchased from Fisher Scientific (Somerville, NJ). The ultralow sulfur diesel (ULSD) fuel was obtained from a local commercial gasoline station.

4.2. Biodiesel Synthesis

Synthesis followed the general procedures of similar work.4,10,51,53 The DES was prepared with a 2:1 mole ratio of choline chloride/glycerol (34.2:25.8 g), stirred until well mixed. The DES was activated with NaOH (6 g of NaOH, which is 1% of the mass of waste oil and grease used, and the NaOH is crushed, ground, and dried in a 100 °C oven for 12 h and cooled to room temperature prior to use), stirred until well mixed. The waste cooking oil and grease were vacuum-filtered (Fisherbrand filter paper, P4 grade) to remove nonfluid materials. After filtering, 600 g of waste cooking oil and grease was mixed with 120 g of methanol in a 1 L round-bottom flask, stirred under a condensing column, and heated to 65 °C. The activated DES (66 g) was added to the preheated waste cooking oil and grease/methanol mixture, stirred, and heated to 65 °C under a condensing column for 3 h. The reaction mixture was transferred to a separatory funnel and allowed to gradually cool to an ambient temperature (22 °C) over 12 h while the reaction components were partitioned in the separatory funnel. Biodiesel (clear and slightly yellow color liquid, 466 g) was separated for analysis without any additional washing. Samples analyzed were a 100% ultralow sulfur diesel (sample name ULSD), a 100% biodiesel synthesized from waste cooking oil and grease using DES (sample name B100), and a 30:70 volume ratio mixture of the biodiesel synthesized from waste cooking oil and grease using DES and ULSD (sample name B30).

4.3. Gas Chromatography–Mass Spectrometry (GC–MS)

The biodiesel samples were analyzed using a Shimadzu QP2010S Gas Chromatography–Mass Spectrometer (GC–MS) with a DB-5 ms column (30 m, 0.25 mm i.d., 0.25 μm film thickness) in scan mode to determine the fatty acid methyl ester (FAME) composition. The carrier gas was helium with a constant pressure of 52.0 kPa. The sample of 1.0 μL in methylene chloride was injected using split mode with a split ratio of 1:5 at an injection temperature of 250 °C. The GC oven temperatures were held at 50 °C for 1 min and increased at 20 °C/min to 200 °C and 10 °C/min to 250 °C with finally holding for 10 min. The GC–MS interface was maintained at 250 °C, and the MS ion source temperature was 230 °C.

4.4. Nuclear Magnetic Resonance Spectroscopy (NMR)

1H NMR scans were conducted using an Anasazi instruments Eft-90 (90 MHz) spectrometer interfaced to Anasazi aii 5.170.213 and NMR Utility Transform Software. Chemical shifts were referenced to an internal standard, tetramethylsilane (TMS), using CDCl3 solvent.

4.5. Temperature-Dependent Kinematic Viscosity

Kinematic viscosity was measured according to the standard ASTM-D445 method. A Cannon-Ubbelohde viscometer along with a Poly-Science viscosity bath was used to collect the efflux time (time for biodiesel to flow a certain distance in the viscometer under gravity) for B100, B30, and ULSD at temperatures varying from room temperature to 100 °C. Kinematic viscosity was calculated by multiplying the efflux time with the viscometer calibration constant.

4.6. Differential Scanning Calorimeter (DSC)

B100, B30, and ULSD were examined using DSC. Samples of 16.0 μg were sealed in aluminum pans, placed in the nitrogen-purged furnace with an adjacent reference sealed aluminum pan, cooled to −70 °C, heated to 50 °C at a rate of 10 °C/min, and then cooled to −70 °C at a rate of 10 °C/min using a PerkinElmer DSC6000. PerkinElmer Pyris software controlled the DSC.

4.7. Thermogravimetric Analysis Coupled with Infrared Spectroscopy (TGA-FTIR)

B100, B30, and ULSD were examined using TGA-FTIR for evolved gas analysis. Samples of 25.0 μL in alumina crucibles were used with a PerkinElmer TGA 8000. A TL8000 transfer line maintained at 270 °C coupled the TGA to a PerkinElmer Frontier FTIR with a 10 cm gas cell. Nitrogen was the purge gas through the TGA with a balance purge flow rate of 70 mL/min and a sample purge rate of 50 mL/min. The TGA temperature was increased at 10 °C/min from 30 to 650 °C. Infrared spectra were collected at 55 scans/min over the range of 4000–600 cm–1 at a 4 cm–1 resolution. The TGA was controlled using PerkinElmer Pyris software, and PerkinElmer TimeBase controlled the IR.

4.8. Combustion Gas Fourier Transform Infrared Spectroscopy (FTIR)

ULSD was used to power a Generac XD 5 kW diesel generator to operating temperatures, the fuel was then changed from ULSD to B30, and the combustion exhaust gas was captured. A Schlenk line connected to a 10 cm gas cell with CaF2 windows at 23 °C was evacuated to 0.5 Torr, and the combustion exhaust gas was then loaded into the line and gas cell to a pressure of 760 Torr at 23 °C. The exhaust gas was analyzed with a PerkinElmer Spectrum One spectrophotometer from 4000 to 1000 cm–1 at a 0.5 cm–1 resolution using signal averaging over 64 scans.

The authors declare no competing financial interest.

References

  1. Chambers A. M.; Yetiv S. A.. The Great Green Fleet. In Naval War College Review, 2011, Vol. 64, p 5.
  2. Pinzi S.; García I.; López-Giménez F. J.; Castro M. D.; Dorado G.; Dorado M. The Ideal Vegetable Oil-based Biodiesel Composition: A Review of Social, Economical and Technical Implications. Energy Fuels 2009, 23, 2325–2341. 10.1021/ef801098a. [DOI] [Google Scholar]
  3. Hossain S.; Razzak S. A.; Alshater A.; Moniruzzaman M.; Hossain M. M. Recent Advances in Enzymatic Conversion of Microalgal Lipids into Biodiesel. Energy Fuels 2020, 34, 6735–6750. 10.1021/acs.energyfuels.0c01064. [DOI] [Google Scholar]
  4. Cong W.; Wang Y.; Li H.; Fang Z.; Sun J.; Liu H.; Liu J.; Tang S.; Xu L. Direct production of biodiesel from waste oils with a strong solid base from alkalized industrial clay ash. Appl. Energy 2020, 264, 114735 10.1016/j.apenergy.2020.114735. [DOI] [Google Scholar]
  5. Uğuz G.; Atabani A. E.; Mohammed M. N.; Shobana S.; Uğuz S.; Kumar G.; Al-Muhtaseb A. Fuel stability of biodiesel from waste cooking oil: A comparative evaluation with various antioxidants using FT-IR and DSC techniques. Biocatal. Agric. Biotechnol. 2019, 21, 101283 10.1016/j.bcab.2019.101283. [DOI] [Google Scholar]
  6. Atabani A. E.; Shobana S.; Mohammed M. N.; Uğuz G.; Kumar G.; Arvindnarayan S.; Aslam M.; Al-Muhtaseb A. Integrated valorization of waste cooking oil and spent coffee grounds for biodiesel production: Blending with higher alcohols, FT–IR, TGA, DSC and NMR characterizations. Fuel 2019, 244, 419–430. 10.1016/j.fuel.2019.01.169. [DOI] [Google Scholar]
  7. Carlson J. S.; Monroe E.; Dhaoui R.; Zhu J.; McEnally C.; Shinde S.; Pfefferle L.; George A.; Davis R. Biodiesel Ethers: Fatty Acid Derived Alkyl Ether Fuels as Improved Bioblendstocks for Mixing-Controlled Compression Ignition Engines. Energy Fuels 2020, 34, 12646–12653. 10.1021/acs.energyfuels.0c01898. [DOI] [Google Scholar]
  8. Paiva A.; Craveiro R.; Aroso I.; Martins M.; Reis R.; Duarte A. Natural Deep Eutectic Solvents – Solvents for the 21st Century. ACS Sustainable Chem. Eng. 2014, 2, 1063–1071. 10.1021/sc500096j. [DOI] [Google Scholar]
  9. Smith E. L.; Abbott A. P.; Ryder K. S. Deep eutectic solvents (DESs) and their applications. Chem. Rev. 2014, 114, 11060–11082. 10.1021/cr300162p. [DOI] [PubMed] [Google Scholar]
  10. Merza F.; Fawzy A.; AlNashef I.; Al-Zuhair S.; Taher H. Effectiveness of using deep eutectic solvents as an alternative to conventional solvents in enzymatic biodiesel production from waste oils. Energy Rep. 2018, 4, 77–83. 10.1016/j.egyr.2018.01.005. [DOI] [Google Scholar]
  11. Šalić A.; Tusek A.; Gojun M.; Zelić B. Biodiesel purification in microextractors: Choline chloride based deep eutectic solvents vs water. Sep.Purif. Technol. 2020, 242, 116783 10.1016/j.seppur.2020.116783. [DOI] [Google Scholar]
  12. Gu L.; Huang W.; Tang S.; Tian S.; Zhang X. A novel deep eutectic solvent for biodiesel preparation using a homogeneous base catalyst. Chem. Eng. J. 2015, 259, 647–652. 10.1016/j.cej.2014.08.026. [DOI] [Google Scholar]
  13. Liu Y.; Hanzhao Y.; Liu J.; Dong W.; Cao Z.; Hu X.; Zhou Z. Acidic deep eutectic solvents with long carbon chains as catalysts and reaction media for biodiesel production. Renewable Energy 2020, 162, 1842–1853. 10.1016/j.renene.2020.09.121. [DOI] [Google Scholar]
  14. Park S. H.; Khan N.; Lee S.; Zimmermann K.; Derosa M.; Hamilton L.; Hudson W.; Hyder S. A.; Serratos M.; Sheffield E. C.; Veludhandi A.; Pursell D. Biodiesel Production from Locally Sourced Restaurant Waste Cooking Oil and Grease: Synthesis, Characterization, and Performance Evaluation. ACS Omega 2019, 4, 7775–7784. 10.1021/acsomega.9b00268. [DOI] [Google Scholar]
  15. Bumbaugh R. E.; Ott L. Preparing and Testing Novel Deep Eutectic Solvents from Biodiesel Co-Product Glycerol for Use as Green Solvents in Organic Chemistry Teaching Laboratories. ACS Symp. Ser. 2020, 1351, 113–130. 10.1021/bk-2020-1351.ch007. [DOI] [Google Scholar]
  16. Florindo C.; Oliveira F. S.; Rebelo L. P.; Fernandes A.; Marrucho I. Insights into the Synthesis and Properties of Deep Eutectic Solvents Based on Cholinium Chloride and Carboxylic Acids. ACS Sustainable Chem. Eng. 2014, 2, 2416–2425. 10.1021/sc500439w. [DOI] [Google Scholar]
  17. Kostik V.; Memeti S.; Bauer B. Fatty acid composition of edible oils and fats. J. Hyg. Eng. Des. 2013, 4, 112–116. [Google Scholar]
  18. Knothe G.; Steidley K. R. A comparison of used cooking oils: a very heterogeneous feedstock for biodiesel. Bioresour. Technol. 2009, 100, 5796–5801. 10.1016/j.biortech.2008.11.064. [DOI] [PubMed] [Google Scholar]
  19. Zhao H.; Zhang C.; Crittle T. D. Choline-based deep eutectic solvents for enzymatic preparation of biodiesel from soybean oil. J. Mol. Catal. B 2013, 85–86, 243–247. 10.1016/j.molcatb.2012.09.003. [DOI] [Google Scholar]
  20. Neiles K. Y.; Bowers G.; Chase D. T.; VerMeulen A.; Hovland D. E.; Bresslour-Rashap E.; Eller L. R.; Koch A. S. Teaching Collaborations and Scientific Practices through a Vertically Scaffolded Biodiesel Laboratory Experience. J. Chem. Educ. 2019, 96, 1988–1997. 10.1021/acs.jchemed.9b00008. [DOI] [Google Scholar]
  21. Knothe G.; Steidley K. R. Kinematic viscosity of biodiesel fuel components and related compounds. Influence of compound structure and comparison to petrodiesel fuel components. Fuel 2005, 84, 1059–1065. 10.1016/j.fuel.2005.01.016. [DOI] [Google Scholar]
  22. Tat M. E.; Gerpen J. V. The kinematic viscosity of biodiesel and its blends with diesel fuel. J. Am. Oil Chem. Soc. 1999, 76, 1511–1513. 10.1007/s11746-999-0194-0. [DOI] [Google Scholar]
  23. Azian M. N.; Kamal A. A.; Panau F.; Ten W. K. Viscosity estimation of triacylglycerols and of some vegetable oils, based on their triacylglycerol composition. J. Am. Oil Chem. Soc. 2001, 78, 1001–1005. 10.1007/s11746-001-0378-9. [DOI] [Google Scholar]
  24. Esteban B.; Riba J.; Baquero G.; Rius A.; Puig R. Temperature dependence of density and viscosity of vegetable oils. Biomass Bioenergy 2012, 42, 164–171. 10.1016/j.biombioe.2012.03.007. [DOI] [Google Scholar]
  25. Knothe G.; Steidley K. R. Kinematic viscosity of biodiesel components (fatty acid alkyl esters) and related compounds at low temperatures. Fuel 2007, 86, 2560–2567. 10.1016/j.fuel.2007.02.006. [DOI] [Google Scholar]
  26. Atabani A. E.; Kulthoom S. A. Spectral, thermoanalytical characterizations, properties, engine and emission performance of complementary biodiesel-diesel-pentanol/octanol blends. Fuel 2020, 282, 118849 10.1016/j.fuel.2020.118849. [DOI] [Google Scholar]
  27. Atabani A. E.; Al-Rubaye O. K. Valorization of spent coffee grounds for biodiesel production: blending with higher alcohols, FT-IR, TGA, DSC, and NMR characterizations. Biomass Convers. Biorefin. 2020, 10.1007/s13399-020-00866-z. [DOI] [Google Scholar]
  28. Bencheikh K.; Atabani A. E.; Shobana S.; Mohammed M. N.; Uğuz G.; Arpa O.; Kumar G.; Ayanoğlu A.; Bokhari A. Fuels properties, characterizations and engine and emission performance analyses of ternary waste cooking oil biodiesel–diesel–propanol blends. Sustainable Energy Technol. Assess. 2019, 35, 321–334. 10.1016/j.seta.2019.08.007. [DOI] [Google Scholar]
  29. Misutsu M. Y.; Cavalheiro L.; Ricci T. G.; Viana L. H.; Oliveira S. C.; Júnior A. M.; Oliveira L. C.. Thermoanalytical Methods in Verifying the Quality of Biodiesel. In Biofuels—Status and Perspective; IntechOpen, 2015; Chapter 12; pp 1164–1223. [Google Scholar]
  30. Chen W.; Wang Y.; Ding M.; Shi S.; Yang Z. Crystallization behaviors and rheological properties of biodiesel derived from methanol and ethanol. Fuel 2017, 207, 503–509. 10.1016/j.fuel.2017.06.121. [DOI] [Google Scholar]
  31. Dunn R. Thermal analysis of alternative diesel fuels from vegetable oils. J. Am. Oil Chem. Soc. 1999, 76, 109–115. 10.1007/s11746-999-0056-9. [DOI] [Google Scholar]
  32. Dunn R. Fuel Properties of Biodiesel/Ultra-Low Sulfur Diesel (ULSD) Blends. J. Am. Oil Chem. Soc. 2011, 88, 1977–1987. 10.1007/s11746-011-1871-3. [DOI] [Google Scholar]
  33. Vijayan S. K.; Victor M. N.; Sudharsanam A.; Chinnaraj V. K.; Nagarajan V. Winterization studies of different vegetable oil biodiesel. Bioresour. Technol. Rep. 2018, 1, 50–55. 10.1016/j.biteb.2018.02.005. [DOI] [Google Scholar]
  34. García-Pérez M.; Adams T. T.; Goodrum J.; Das K. C.; Geller D. P. DSC studies to evaluate the impact of bio-oil on cold flow properties and oxidation stability of bio-diesel. Bioresour. Technol. 2010, 101, 6219–6224. 10.1016/j.biortech.2010.03.002. [DOI] [PubMed] [Google Scholar]
  35. Lazdoviča K.; Liepiņa L.; Kampars V. Comparative wheat straw catalytic pyrolysis in the presence of zeolites, Pt/C, and Pd/C by using TGA-FTIR method. Fuel Process. Technol. 2015, 138, 645–653. 10.1016/j.fuproc.2015.07.005. [DOI] [Google Scholar]
  36. Li H.; Niu S.; Lu C.; Wang Y. Comprehensive Investigation of the Thermal Degradation Characteristics of Biodiesel and Its Feedstock Oil through TGA–FTIR. Energy Fuels 2015, 29, 5145–5153. 10.1021/acs.energyfuels.5b01054. [DOI] [Google Scholar]
  37. Dwivedi G.; Sharma M. Experimental investigation on thermal stability of Pongamia Biodiesel by thermogravimetric analysis. Egypt. J. Pet. 2016, 25, 33–38. 10.1016/j.ejpe.2015.06.008. [DOI] [Google Scholar]
  38. Chand P.; Reddy C. V.; Verkade J.; Wang T.; Grewell D. Thermogravimetric Quantification of Biodiesel Produced via Alkali Catalyzed Transesterification of Soybean oil. Energy Fuels 2009, 23, 989–992. 10.1021/ef800668u. [DOI] [Google Scholar]
  39. Díaz-Ballote L.; Gómez-Hernández K.; Vega-Lizama E.; Ruiz-Gómez M. A.; Maldonado L.; Hernández E. Thermogravimetric approach for assessing the oxidation level of a biodiesel sample. Quím. Nova 2018, 41, 492–496. 10.21577/0100-4042.20170199. [DOI] [Google Scholar]
  40. Pragas M. G.; Boey P. L.; Shafida A. H. Biodiesel from Adsorbed Waste Oil on Spent Bleaching Clay using CaO as a Heterogeneous Catalyst. Eur. J. Sci. Res. 2009, 33, 347–357. [Google Scholar]
  41. Ong H. C.; Chen W.; Singh Y.; Gan Y.; Chen C.; Show P. L. A state-of-the-art review on thermochemical conversion of biomass for biofuel production: A TG-FTIR approach. Energy Convers. Manage. 2020, 209, 112634 10.1016/j.enconman.2020.112634. [DOI] [Google Scholar]
  42. Jameel A. G. A.; Han Y.; Brignoli O.; Telalović S.; El-Baz A.; Im H.; Roberts W. L.; Sarathy S. M. Heavy fuel oil pyrolysis and combustion: Kinetics and evolved gases investigated by TGA-FTIR. J. Anal. Appl. Pyrolysis 2017, 127, 183–195. 10.1016/j.jaap.2017.08.008. [DOI] [Google Scholar]
  43. Yang H.; Yan R.; Liang D.; Chen H.; Zheng C. Pyrolysis of Palm Oil Wastes for Biofuel Production. Asian J. Energy Environ. 2006, 7, 315–323. [Google Scholar]
  44. Li H.; Feng-sheng L.; Ma X.; Cui P.; Gao Y.; Yu M.; Guo M. Effects of biodiesel blends on the kinetic and thermodynamic parameters of fossil diesel during thermal degradation. Energy Convers. Manage. 2019, 198, 111930 10.1016/j.enconman.2019.111930. [DOI] [Google Scholar]
  45. Mahamuni N. N.; Adewuyi Y. G. Fourier Transform Infrared Spectroscopy (FTIR) Method To Monitor Soy Biodiesel and Soybean Oil in Transesterification Reactions, Petrodiesel–Biodiesel Blends, and Blend Adulteration with Soy Oil. Energy Fuels 2009, 23, 3773–3782. 10.1021/ef900130m. [DOI] [Google Scholar]
  46. Oliveira J. S.; Montalvão R.; Daher L.; Suárez P.; Rubim J. C. Determination of methyl ester contents in biodiesel blends by FTIR-ATR and FTNIR spectroscopies. Talanta 2006, 69, 1278–1284. 10.1016/j.talanta.2006.01.002. [DOI] [PubMed] [Google Scholar]
  47. Rank D.; Pierre A.; Wiggins T. Rotational and vibration constants of CO. J. Mol. Spectrosc. 1965, 18, 418–427. 10.1016/0022-2852(65)90048-2. [DOI] [Google Scholar]
  48. Mina-Camilde N.; Manzanares I.; Caballero J. Molecular Constants of Carbon Monoxide at v = 0, 1, 2, and 3: A Vibrational Spectroscopy Experiment in Physical Chemistry. J. Chem. Educ. 1996, 73, 804–807. 10.1021/ed073p804. [DOI] [Google Scholar]
  49. Boyd D.; Thompson H. W.; Williams R. Vibration-rotation bands of methane. Proc. R. Soc. London, Ser. A 1952, 213, 42–54. 10.1098/rspa.1952.0109. [DOI] [Google Scholar]
  50. Tremblay A.; Cao P.; Dubé M. Biodiesel Production Using Ultralow Catalyst Concentrations. Energy Fuels 2008, 22, 2748–2755. 10.1021/ef700769v. [DOI] [Google Scholar]
  51. Huang W.; Tang S.; Zhao H.; Tian S. Activation of Commercial CaO for Biodiesel Production from Rapeseed Oil Using a Novel Deep Eutectic Solvent. Ind. Eng. Chem. Res. 2013, 52, 11943–11947. 10.1021/ie401292w. [DOI] [Google Scholar]

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES