An experimental and modeling study of 2-methyl-1-butanol oxidation in a jet-stirred reactor

Combustion and Flame 161 (2014) 3003–3013

Contents lists available at ScienceDirect

Combustion and Flame j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / c o m b u s t fl a m e

An experimental and modeling study of 2-methyl-1-butanol oxidation in a jet-stirred reactor Zeynep Serinyel a,b,⇑, Casimir Togbé a, Guillaume Dayma a,b, Philippe Dagaut a a b

CNRS–ICARE, 1C Avenue de la Recherche Scientifique, 45071 Orléans cedex 2, France Université d’Orléans, Collegium Sciences et Techniques, 1 rue de Chartres, 45067 Orléans cedex 2, France

a r t i c l e

i n f o

Article history: Received 19 March 2014 Received in revised form 28 May 2014 Accepted 5 June 2014 Available online 3 July 2014 Keywords: Jet-stirred reactor 2-Methyl-1-butanol Oxidation Chemical kinetics

a b s t r a c t In an effort to understand the oxidation chemistry of new generation biofuels, oxidation of a pentanol isomer (2-methyl-1-butanol) was investigated experimentally in a jet-stirred reactor (JSR) at a pressure of 10 atm, equivalence ratios of 0.5, 1, 2 and 4 and in a temperature range of 700–1200 K. Concentration profiles of the stable species were measured using GC and FTIR. A detailed chemical kinetic mechanism including oxidation of various hydrocarbon and oxygenated fuels was extended to include the oxidation chemistry of 2-methyl-1-butanol, the resulting mechanism was used to simulate the present experiments. In addition to the present data, recent experimental data such as ignition delay times measured in a shock tube and laminar flame speeds were also simulated with this mechanism and satisfactory results were obtained. Reaction path and sensitivity analyses were performed in order to interpret the results. Ó 2014 The Combustion Institute. Published by Elsevier Inc. All rights reserved.

1. Introduction Given the ongoing worldwide energy demand and non-sustainable character of fossil fuels, biofuels have recently been given a lot of interest. Among these, bio-ethanol dominates over 90% of the total biofuel production and is currently being used as a first generation biofuel despite its low energy density (21 kJ/mL) compared to gasoline, and its hygroscopic character. Recently larger alcohols have been the focus of many studies due to their higher energy density and better solubility in gasoline. Although many experimental studies now exist on alcohols up to C4, fewer studies are available for larger (C P 5) alcohols. As far as the C6 alcohols are concerned, Togbé et al. studied n-hexanol oxidation in a jet stirred reactor and a spherical bomb [1], Heufer and co-workers measured ignition delay times of n-hexanol at high pressures in a shock tube and a rapid compression machine [2]. Among the pentanol isomers, iso-pentanol is the most widely studied one; its oxidation has been investigated in a jet stirred reactor at an operating pressure of 5–10 atm and various equivalence ratios [3,4], in laminar flames [4,5] as well as in shock tubes and a rapid compression machine for ignition delay times [4,6,7]. Similarly, experimental ⇑ Corresponding author at: Institut de Combustion Aérothermique Réactivité et Environnement, Université d’Orléans, 1C avenue de la Recherche Scientifique, 45071 Orléans cedex 2, France. E-mail address: [email protected] (Z. Serinyel).

kinetic studies exist on the n- isomer; Togbé et al. [8] studied n-pentanol oxidation in a jet stirred reactor and in laminar flames, Tang et al. [6] and Heufer et al. [2] studied auto-ignition of this biofuel behind reflected shocks. The present study focuses on a mono-methylated isomer of pentanol; 2-methyl-1-butanol (2-MB). Atsumi et al. [9] studied non-fermentative pathways to synthesize larger alcohols (than ethanol) including 2-MB, from glucose, which is a renewable source. To the best of our knowledge, the only kinetic studies available on 2-MB combustion focus on global reactivity. Li and coworkers [10] recently reported laminar flame speed measurements of this alcohol in a spherically propagating flame at unburned mixture temperatures of 393, 433 and 473 K and pressures of 1, 2.5, 5 and 7.5 bar, they found that flame speeds of 2-MB were higher than iso-octane and lower than ethanol. Tang and co-workers measured ignition delay times of 2-MB behind reflected shock waves at 1 and 2.6 atm and over a temperature range of 1100–1500 K, along with n- and iso-pentanol [6], they also developed a kinetic mechanism in order to represent their results and concluded that further improvements would be needed. They observed that 2-MB reactivity was between the latter two. As far as 2-MB is concerned, no speciation data is reported so far. This study aims to provide new kinetic data through a detailed product analysis of 2-MB oxidation in a jet stirred reactor. Four dilute mixtures with equivalence ratios from 0.5 to 4.0 were investigated at 10 atm at a residence time of s = 0.7 s between 700 and

http://dx.doi.org/10.1016/j.combustflame.2014.06.004 0010-2180/Ó 2014 The Combustion Institute. Published by Elsevier Inc. All rights reserved.

3004

Z. Serinyel et al. / Combustion and Flame 161 (2014) 3003–3013

1200 K. A chemical kinetic mechanism is used to represent the present data as well as the available literature data discussed above.

2. Experimental The jet stirred reactor experimental setup used here has been described earlier [8,11]. The reactor consists of a 4 cm diameter fused silica sphere (42 cm3) equipped with four nozzles of 1 mm i.d. Prior to the injectors, the reactants were diluted with nitrogen (<100 ppm H2O, <50 ppm O2, <1000 ppm Ar, <5 ppm H2 from Air Liquide) and mixed. A high degree of dilution (0.1–0.15% mol. of fuel) was used, reducing temperature gradients and heat release in the JSR. The reactants were high-purity oxygen (99.995% pure form Air Liquide) and high-purity 2-methyl-1-butanol (>99% pure from Aldrich, CAS 137-32-6). The reactants were preheated before injection to minimize temperature gradients inside the reactor. A Shimadzu HPLC pump (LC10 AD VP) with an on-line degasser (Shimadzu DGU-20 A3) was used to deliver the fuel to an in-house atomizer-vaporizer assembly maintained at 200 °C. Good thermal homogeneity along the vertical axis of the reactor (gradients of ca. 1 K/cm) was observed during the experiments by thermocouple measurements (0.1 mm Pt–Pt/Rh-10%, located inside a thin-wall fused silica tube). The reacting mixtures were sampled using a Table 1 Structures and enthalpies of formation of some selected species (kcal mol

OH

OH

OH

. Butoh2m DfH° =

Butoh2m-1 DfH° = 28.98

71.88

Butoh2m-2 DfH° = 28.38

O

O

55.45

.

.

Butal2m-a DfH° = 20.85

Butal2m-t DfH° = 18.35

1

, at 298 K) in 2-MB oxidation.

.

Buto2m DfH° = 20.58

But2m1d DfH° = 8.68

OH

OH

Butoh2d DfH° = 37.36

Butoh1d DfH° = 42.23

O

.

.

Butal2m-4 DfH° = 6.65

Butal2m-5 DfH° = 6.65

methane

2-MB

8.0E-04

Butoh2m-5 DfH° = 23.78

O

Butal2m-s DfH° = 8.45

O.

.

Butoh2m-4 DfH° = 23.08

O

1.6E-03

1.0E-03

OH

OH

OH

Butoh2m-3 DfH° = 25.28

O

. Butal2m DfH° =

.

.

movable fused silica low-pressure sonic probe. The samples were sent to analyzers via a Teflon heated line maintained at 200 °C. Analyses were performed online using a FTIR spectrometer (10 m path length, resolution of 0.5 cm 1, 200 mBar in the cell) and offline after collection and storage in 1 L Pyrex bulbs [12]. Off-line analyses were performed using gas chromatographs (GC) equipped with capillary columns (0.32 mm i.d.: DB-624 and CP-Al2O3–KCl, and 0.53 mm i.d.: Carboplot-P7), a TCD (thermal conductivity detector), and an FID (flame ionization detector). Two GC–MS (Varian quadrupole V1200) operating with electron ionization (70 eV) were used for product identification. The experiments were performed at steady state, at a constant mean residence time, s, of 0.7 s and a constant pressure of 10 atm, with the reactants continually flowing in the reactor while the temperature of the gas inside the reactor was increased stepwise. A good repeatability of the measurements and a reasonably good carbon balance (typically 100 ± 10%) were obtained in these experiments. Hydrogen and oxygen balances were also checked and found to be within ±10% for most of the experimental points except for a few ones where they reach 12–15%. Experimental uncertainties can be summarized as follows: uncertainty related to temperature measurements is expected to be less than 10 K, and that related to residence time it is less than 5%, inlet uncertainty is less than 5% for the reactants and that of the measured species is determined as <10% for concentrations higher

8.0E-04

1.2E-03

ethylene

6.0E-04

6.0E-04 8.0E-04

4.0E-04

4.0E-04

2.0E-04

4.0E-04 2.0E-04 0.0E+00

0.0E+00 700

800

2.0E-04

900

1000

1100

1200

0.0E+00 700

800

900

1000

1100

1200

8.0E-05

propene

6.0E-05

1.0E-04

4.0E-05

5.0E-05

2.0E-05

800

900

1000

6.0E-05

2-methylbutanal

1.5E-04

700

1100

1200

propanal

4.0E-05

2.0E-05

0.0E+00

0.0E+00 700

800

900

1000

1100

1200

0.0E+00 700

800

900

1000

1100

1200

700

800

900

1000

1100

1200

Fig. 1. Experimental mole fraction profiles of some main oxidation products of 2-MB as a function of temperature (u = 4 profiles are normalized to an initial fuel concentration of 1000 ppm).

3005

Z. Serinyel et al. / Combustion and Flame 161 (2014) 3003–3013

than 10 ppm. There is also a contribution of the sampling technique to global uncertainty however this cannot be quantified. These factors contribute to the uncertainty of the measured mole fractions profiles. 3. Chemical kinetic mechanism In an attempt to represent the present and literature experimental data, a chemical kinetic sub-mechanism of 2-MB was

developed. An existing detailed mechanism, which includes oxidation mechanisms of various alcohols such as 1-pentanol, 1-hexanol, n-butanol and iso-pentanol [1,3,8,13] was used for simulations. This mechanism is basically extended to cover reactions involving 2-MB oxidation and those of related products. Thermochemical properties of the new species belonging to the 2-MB sub-mechanism were calculated using the Thergas software [14], which uses group additivity methods proposed by Benson [15]. A nomenclature consistent with the iso-pentanol mechanism is 8.0x10 -3

2-MB CH4

1.0x10 -3

CO CO2

6.0x10

H2

Mole fraction

Mole fraction

C2H4

8.0x10 -4

H2O

6.0x10 -4 4.0x10 -4

-3

O2/3

4.0x10 -3

2.0x10 -3 2.0x10 -4 0.0

0.0 700

800

900

1000

1100

700

1200

800

900

T (K) 2.5x10 -4

1.0x10 -4

2.0x10

C2H2

Mole fraction

Mole fraction

-5

6.0x10 -5 4.0x10 -5

-4

1.5x10 -4 1.0x10 -4

0.0

0.0 700

800

900

1000

1100

700

1200

800

900

6.0x10 -5

1000

1100

1200

T (K)

T (K) 9.0x10 -5

1-C4H8

2-methylbutanal propanal acetone 2-butanone

trans 2-C4H8 cis 2-C4H8 -5

Mole fraction

Mole fraction

1200

5.0x10 -5

2.0x10 -5

4.0x10

1100

formaldehyde acetaldehyde 2-propenal

C3H6 C2H6

8.0x10

1000

T (K)

2.0x10 -5

0.0

6.0x10 -5

3.0x10 -5

0.0 700

800

900

1000

1100

1200

700

800

900

T (K)

1000

1100

1200

T (K)

4.0x10 -5

1.2x10 -5

2-methyl-1-butene isobutene 2,3-dimethyloxirane

methacrolein 1,3-butadiene isoprene

Mole fraction

Mole fraction

3.0x10 -5

2.0x10 -5

8.0x10 -6

4.0x10 -6

1.0x10 -5

0.0

0.0 700

800

900

1000 T (K)

1100

1200

700

800

900

1000

1100

1200

T (K)

Fig. 2. Experimental and simulated mole fraction profiles of stable species in 2-MB oxidation in the JSR (u = 0.5, p = 10 atm, s = 0.7 s).

3006

Z. Serinyel et al. / Combustion and Flame 161 (2014) 3003–3013

provided for some selected species, along with their enthalpies of formation at 298 K, in Table 1. All kinetic simulations were performed using Chemkin-II [16]. The kinetic mechanism used in this study is provided as supplementary material. It is mainly based on 1-pentanol and iso-pentanol oxidation schemes [3,8] and only some main features

are discussed here. As far as fuel reactions are concerned, water elimination, unimolecular dissociations, bimolecular initiation reactions with O2 and hydrogen abstraction reactions by O, OH, H, HO2, CH3, HCO, C2H5 and C2H3 were considered. In our experimental conditions as well as the literature conditions, H-abstraction reactions by OH, H and HO2 are the most important

9.0x10 -3

2-MB CH4

1.0x10 -3

H2O CO CO2

8.0x10

H2

6.0x10 -4 4.0x10 -4

O2

6.0x10 -3

Mole fraction

Mole fraction

C2H4 -4

3.0x10 -3

2.0x10 -4 0.0 700

0.0 800

900

1000

1100

1200

700

800

900

T (K) 1.6x10 -4

3.0x10 -4

C3H6 C2H6

1100

1200

formaldehyde acetaldehyde 2-propenal

2.5x10 -4

C2H2

Mole fraction

1.2x10 -4

Mole fraction

1000

T (K)

8.0x10 -5

2.0x10 -4 1.5x10 -4 1.0x10 -4

4.0x10 -5 5.0x10 -5 0.0 700

0.0 800

900

1000

1100

1200

700

800

T (K)

4.0x10

-5

3.0x10

-5

1000

8.0x10 -5

1-C4H8

1200

2-methylbutanal propanal acetone 2-butanone

trans 2-C4H8

Mole fraction

6.0x10 -5

2.0x10 -5

4.0x10 -5

2.0x10 -5

1.0x10 -5

0.0

0.0 700

800

900

1000

1100

700

1200

800

900

4.0x10 -5

2-methyl-1-butene isobutene 2,3-dimethyloxirane

1100

1200

methacrolein 1,3-butadiene isoprene propyne

1.0x10 -5

3.0x10 -5

Mole fraction

8.0x10 -6

2.0x10 -5

1.0x10 -5

0.0 700

1000

T (K)

T (K)

Mole fraction

1100

T (K)

cis 2-C4H8

Mole fraction

900

6.0x10 -6 4.0x10 -6 2.0x10 -6 0.0

800

900

1000

T (K)

1100

1200

700

800

900

1000

1100

1200

T (K)

Fig. 3. Experimental and simulated mole fraction profiles of stable species in 2-MB oxidation in the JSR (u = 1, p = 10 atm, s = 0.7 s).

3007

Z. Serinyel et al. / Combustion and Flame 161 (2014) 3003–3013

ones. These reactions yield the fuel radicals shown in Table 1, which then decompose into smaller radicals and molecules. For hydrogen abstraction reactions by OH radical, rate constants proposed by Zhou et al. for n-butanol [17] were adopted, but the rate constant of abstraction involving the tertiary C–H bond was considered analogous to that of an alkane and was estimated based on the reference [18]. Hydrogen abstraction reactions by all other radicals were estimated based on the iso-pentanol mechanism and the references therein [3]. Fuel radical isomerization reactions involving 5– and 6–membered transition states were also

1.6x10 -3

considered as proposed by Zheng and Truhlar for butoxyl radicals [19]. These include isomerization between fuel radicals butoh2m4 and butoh2m-1 as well as between 2-methylbutoxy radical (buto2m) with radicals butoh2m-3, butoh2m-4 and butoh2m-5. Dissociation rate constants by C–H bond fission were estimated in the radical–radical recombination direction (fuel radical + H) as proposed by Harding et al. [20]. Similarly for the C–C bond fission reactions, recommendations by Tsang [21,22] were used for C2H5 + CH3
CHCH2OH and CH3 + CH3CH(CH2)(CH2OH) reactions in analogy with the recombination reactions iC3H7 + C2H5 = iC5H12

2-MB CH4 C2H4 H2

Mole fraction

Mole fraction

1.2x10

-3

8.0x10 -4

4.0x10 -4

5.0x10

-3

4.0x10

-3

3.0x10

-3

2.0x10

-3

1.0x10

-3

800

900

1000

1100

CO CO2 O2

0.0 700

0.0 700

H2O

1200

800

900

2.0x10 -4

C2H6 C2H2

Mole fraction

Mole fraction

2.5x10

-4

1.2x10 -4 8.0x10 -5 4.0x10 -5 0.0 700

800

900

2.0x10

-4

1.5x10

-4

1.0x10

-4

5.0x10

-5

1000

1100

0.0 700

1200

800

900

T (K)

1-C4H8 trans 2-C4H8

4.0x10 -5

1000

1100

1200

6.0x10

-5

4.0x10

-5

2.0x10

-5

2-methylbutanal propanal acetone

cis 2-C4H8

Mole fraction

Mole fraction

1200

T (K)

5.0x10 -5

3.0x10 -5 2.0x10 -5 -5

0.0 700

0.0 700

800

900

1000

1100

1200

800

900

T (K) 5.0x10 -5 4.0x10

1100

formaldehyde acetaldehyde 2-propenal

C3H6

1.6x10 -4

1.0x10

1000

T (K)

T (K)

2-methyl-1-butene isobutene propane

-5

1000

1100

1200

T (K)

isoprene methacrolein 1,3-butadiene benzene

1.5x10-5

2-butanone allene propyne 2,3-dimethyloxirane

8.0x10 -6 6.0x10 -6

3.0x10 -5

1.0x10-5 4.0x10 -6

2.0x10 -5 -6

5.0x10

2.0x10 -6

1.0x10 -5 0.0

0.0 700

800

900

1000

1100

1200

0.0 700

800

900

1000

1100

1200

700

800

900

1000

1100

Fig. 4. Experimental and simulated mole fraction profiles of stable species in 2-MB oxidation in the JSR (u = 2, p = 10 atm, s = 0.7 s).

1200

3008

Z. Serinyel et al. / Combustion and Flame 161 (2014) 3003–3013

(k = 1.15  1014 T 0.35 cm3 mol 1 s 1) and iC4H9 + CH3 = iC5H12 (k = 1.94  1011 T 0.32 cm3 mol 1 s 1) respectively. For the recombination of CH2OH and CH3CH2
CHCH3 the rate constant was estimated based on Tsang’s recommendation of k = 1.45  1013 cm3 mol 1 s 1. Forward (dissociation) rate constants were then evaluated via microscopic reversibility. The rate constant for the C–O bond fission forming OH + CH3CH2 CH(CH3)(CH2) is expected to be less likely to occur compared to C–C bond fission reactions based on its higher bond dissociation energy. Its rate constant was taken as analogous to the reaction iso-pentanol ? OH + CH3CH(CH3)CH2
CH2 [3]. The water

Mole fraction

2.0x10 -3 1.5x10

2-MB CH4

4.0x10 -3

H2O

CO

CO2

O2

C2H4 H2

Mole fraction

2.5x10 -3

elimination reaction (2-MB ? 2-methyl-1-butene + H2O) rate constant was estimated based on the study by Carstensen and Dean [23] where heats of reaction and barriers for water elimination were calculated on a CBS-QB3 level of theory for various alcohols. Fuel radicals decomposition rate constants were estimated, either in addition or dissociation direction, mainly based on the work by Curran [24] and the rate rules described in [25,26]. Hydrogen elimination from the fuel radicals or oxidation of those can lead to several C5 unsaturated alcohols. Formation and consumption reactions of these species were also considered although their formation is not competitive compared to beta

-3

1.0x10 -3

3.0x10 -3

2.0x10 -3

1.0x10 -3

5.0x10 -4 0.0 700

800

900

1000

1100

0.0 700

1200

800

900

T (K) 3.0x10 -4

C3H6 C2H2

2.0x10 -4 1.5x10 -4 1.0x10 -4 5.0x10

2.0x10 -4

1.0x10 -4

0.0 700

800

900

1000

1100

1200

700

800

900

T (K)

1000

1100

1200

T (K) 1.0x10 -4

6.0x10 -5

2-methylbutanal propanal acetone

1-C4H8 trans 2-C4H8

8.0x10

cis 2-C4H8 -5

Mole fraction

Mole fraction

1200

-5

0.0

4.0x10

1100

acetaldehyde formaldehyde 2-propenal

3.0x10 -4

C2H6

Mole fraction

Mole fraction

2.5x10

-4

1000

T (K)

2.0x10 -5

-5

6.0x10 -5 4.0x10 -5 2.0x10 -5

0.0 700

800

900

1000

1100

0.0 700

1200

800

900

T (K) 8.0x10 -5

2-methyl-1-butene isobutene propane

6.0x10 -5

5.0x10 -5 4.0x10 -5

isoprene methacrolein 1,3-butadiene benzene

1.2x10 -5

3.0x10 -5

4.0x10 -5

1000

1100

1200

T (K)

2-butanone allene propyne 2,3-dimethyloxirane

8.0x10 -6

2.0x10 -5 4.0x10 -6

2.0x10 -5

1.0x10 -5 0.0

0.0

0.0 700

800

900

1000

1100

1200

700

800

900

1000

1100

1200

700

800

900

1000

1100

Fig. 5. Experimental and simulated mole fraction profiles of stable species in 2-MB oxidation in the JSR (u = 4, p = 10 atm, s = 0.7 s).

1200

3009

Z. Serinyel et al. / Combustion and Flame 161 (2014) 3003–3013

scission reactions in our conditions. Moreover these species were not observed experimentally. C4 enols can be formed by beta scission of the fuel radicals, among which but-2-en-1-ol formation is slightly more important, oxidation mechanism of this unsaturated alcohol is defined in the iso-pentanol mechanism [3]. In addition to these, propenol and vinyl alcohol reactions are also included [27– 29]. Another important molecule in 2-MB oxidation is 2-methylbutanal, which can be formed by dissociation reactions of the fuel radicals: butoh2m-1 and buto2m. A sub-scheme involving the oxidation of this molecule was therefore added into the mechanism. Unimolecular dissociation reactions of 2-methylbutanal were adapted from a recent experimental study by Rosado-Reyes and Tsang [30]. Hydrogen abstraction from 2-methylbutanal by small radicals and dissociation of the 2-methylbutanal radicals were also considered. 4. Results and discussion 4.1. Jet-stirred reactor results Four mixtures with equivalence ratios of u = 0.5, 1, 2, and 4 were investigated in the JSR at an operating pressure of 10 atm and a constant mean residence time of 0.7 s. Inlet fuel concentration was 0.10% for the first three mixtures and 0.15% for the mixture with u = 4. Carbon balances were checked and found to be no more than ±10%. Oxidation intermediates and products were quantified via GC and FTIR and are presented in Figs. 1–5. Methane, ethane, ethylene, acetylene, propene, 2-methyl-1-butene, and butene isomers (1-, 2cis and trans, iso-) were the abundant hydrocarbon species observed for all mixtures. Other hydrocarbon species such as isoprene and 1,3-butadiene were also observed with smaller concentrations than the former ones. Propyne and allene were also quantified with concentrations less than 10 ppm for stoichiometric and rich mixtures. Oxygenated intermediates quantified in all mixtures include formaldehyde, acetaldehyde, 2-methylbutanal, propanal, acrolein, acetone, and methacrolein. Very small amounts of iso-butanal was also detected in mixtures with u = 1, 2 and 4 at a maximum of 5 ppm for the u = 4 mixture. Butanal was detected with trace amounts of 2–3 ppm for all mixtures around 820–830 K, as well as traces of cyclopentene (2 ppm) for the rich mixtures. Formation of benzene was observed for the rich mixtures, at temperatures greater than 1000 K with a maximum value of 15 ppm and 47 ppm at 1200 K, for u = 2 and u = 4, respectively. Mole fraction profiles of the fuel and some important intermediate species are presented in Fig. 1 on a logarithmic scale in y-axis, in order to compare relative reactivity at different equivalence

OH

+ C2H5

.

.

butal2m

+ CH2OH

O

butoh2m-3

OH

+ OH

O + HCO

+ C2H5

butoh2m-4

O

O

.

OH

butoh2m-2

O + CH3

.

OH

.

O

buto2m-5

OH

+R

.

1

5

.

+R

2

4

butoh2m-1

O

O

OH

OH

3

.

O

O

ratios. In the present conditions fuel consumption starts occurring around 770 K for all the mixtures investigated. For the lean and stoichiometric mixtures almost all fuel is converted around 950 K, whereas this happens in the vicinity of 1100 K for the rich mixtures. This reflects in the profiles of main hydrocarbon intermediates such as methane, ethylene, and propene where almost the same trend is observed up to 800 K, however these species’ mole fractions peak at higher temperatures as the equivalence ratio increases. On the other hand, a main oxygenated product, 2-methylbutanal was observed to have similar mole fraction profiles at all conditions, mostly due to the fact that it is a product of the fuel radical butoh2m-1 decomposition and this radical is produced at similar percentages at all conditions by H-abstraction from fuel. Comparisons of the experimental data and kinetic model predictions are presented in Figs. 2–5. General trends and mole fraction profiles of the main oxidation products are well captured by the model. As far as fuel decay is concerned, the kinetic model predicts the differences in reactivity, and agrees with the experimental data at all conditions. Among the oxidation products, the kinetic model under-predicts water at all conditions, this could be due to a combination of analytical and model uncertainties. In fact, some points where disagreement is more significant, the hydrogen balance is around + 10% or slightly more, given that H2 predictions are good, the discrepancy can be attributed to water measurements. Similarly for O2 predictions, for the lean mixture O2 balance reaches 67% less than the theoretical quantity which can partly explain the over-prediction, for the rich mixtures at the highest temperatures of interest, oxygen seems to remain in the system while the model consumes it, these are few points where the balance is about +8–12%. The model has difficulties reproducing small and highly unsaturated species such as allene, propyne and 1,3-butadiene, the former ones are quantified at quantities less than 10 ppm while the production of the latter one, although less than 10 ppm for u = 0.5, 1 and 2 mixtures, reaches 30 ppm around 1100 K for the richest (u = 4) mixture. Formation of 1,3-butadiene is mostly due to C4H7 radicals therefore depends on the oxidation of butenes. Also it is partly formed by the beta scission of the CH2 CHCH(CH3)(CH2) radical. It is mainly consumed by addition reactions on the double bond. Uncertainties related to these reactions may cause an over-consumption. Another interesting point to note is on the benzene profiles for the rich mixtures, the model almost reproduces well benzene mole fractions at u = 4, however fails to produce it at u = 2. Reaction flux and sensitivity analyses were performed in order to delineate oxidation paths and important reactions.

+ CH2OH

+ CH3

. + C2H4

O

Fig. 6. Major reaction pathways in the oxidation of 2-MB (900 K, u = 1).

OH

.

3010

Z. Serinyel et al. / Combustion and Flame 161 (2014) 3003–3013

ch3+ho2 = ch4+o2 2ho2 = h2o2+o2 ch2o+oh = hco+h2o 2ho2 = h2o2+o2 c2h3+o2 = ch2o+hco ho2+oh = h2o+o2 butoh2m-1+o2 = butal2m+ho2 butoh2m+ho2 = butoh2m-1+h2o2 c2h5cho+ho2 = c2h5co+h2o2 c3h5-a+ho2 = c3h5o+oh c2h4+oh = c2h3+h2o ch3cho+ho2 = ch3co+h2o2 butoh2m+oh = butoh2m-5+h2o butoh2m-1 = c2h5+c2h5cho ch2o+ho2 = hco+h2o2 c2h3+o2 = ch2cho+o h2o2(+M) = 2oh(+M) ch3+ho2 = ch3o+oh butoh2m+oh = butoh2m-1+h2o -0.6 -0.5 -0.4 -0.3 -0.2 -0.1

0

2ho2 = h2o2+o2 2ho2 = h2o2+o2 ch3+ho2 = ch4+o2 sc4h9 = c3h6+ch3 ch2o+oh = hco+h2o butoh2m-3 = ch3+butoh2d butoh2m+oh = butoh2m-3+h2o but3m1d = c4h713+ch3 butoh2m+ho2 = butoh2m-4+h2o2 ch2o+ho2 = hco+h2o2 ch3+ho2 = ch3o+oh butoh2m+ho2 = butoh2m-3+h2o2 butal2m+ho2 = butal2m-a+h2o2 butoh2m+oh = butoh2m-1+h2o butoh2m+ho2 = butoh2m-2+h2o2 butoh2m+ho2 = butoh2m-1+h2o2 h2o2(+M) = 2oh(+M) x 0.5

0.1 0.2 0.3 0.4

-0.2 -0.15 -0.1 -0.05

Sensitivity coefficient

0

0.05 0.1 0.15 0.2

Sensitivity coefficient

(b)

(a)

Fig. 7. First order sensitivity analysis of the oxidation of 2-MB in JSR conditions for the stoichiometric mixture (a) 900 K and (b) 800 K.

6.0E-05

1.6E-03

3-MB

acrolein (3-MB) acrolein (2-MB)

2-MB 5.0E-05

propanal (3-MB) propanal (2-MB)

1.0E-04 8.0E-05

1.2E-03 4.0E-05

6.0E-05 3.0E-05

8.0E-04

4.0E-05 2.0E-05 4.0E-04

2.0E-05

1.0E-05 0.0E+00

0.0E+00

0.0E+00 700

800

900

1000

1100

700

1200

800

(a)

900

1000

1100

700

1200

800

900

(b)

1.5E-04

1.0E-04

i-butene (3-MB) i-butene (2-MB)

1.2E-04

8.0E-05

9.0E-05

6.0E-05

6.0E-05

4.0E-05

1000

1100

1200

(c) 2-butene (3-MB) 2-butene (2-MB)

6.0E-05

1-butene (3-MB) 1-butene (2-MB)

5.0E-05 4.0E-05 3.0E-05 2.0E-05

3.0E-05

2.0E-05

0.0E+00

1.0E-05

0.0E+00 700

800

900

1000

1100

1200

0.0E+00 700

800

(d)

900

1000

1100

1200

700

800

900

(e) 6.0E-04

propene (3-MB) propene (2-MB)

5.0E-04

1000

1100

1200

(f) acetaldehyde (3-MB) acetaldehyde (2-MB)

2.0E-04

1.5E-04

4.0E-04 3.0E-04

1.0E-04

2.0E-04 5.0E-05 1.0E-04 0.0E+00

0.0E+00 700

800

900

1000

(g)

1100

1200

700

800

900

1000

1100

1200

(h)

Fig. 8. Fuel decomposition and selected product mole fractions in 2-MB (this study) and 3-MB oxidation [3], u = 4, s = 0.7 and p = 10 atm. (cis- and trans- forms of 2-butene are plotted together).

3011

Z. Serinyel et al. / Combustion and Flame 161 (2014) 3003–3013

4.2. Reaction flux and sensitivity analyses in JSR conditions

ch3+ho2 = ch4+o2 ho2+oh = h2o+o2 butoh2m+h = butoh2m-3+h2 hco+o2 = co+ho2 ch3+ch3(+m) = c2h6(+m) butoh2m+h = butoh2m-4+h2 c2h3+o2 = ch2o+hco c2h3+o2 = ch2cho+o hco+m = h+co+m c2h4+oh = c2h3+h2o c3h5-a+ho2 = c3h5o+oh butoh2m+oh = butoh2m-1+h2o butoh2m = c2h5+c3h61oh butoh2m = ch2oh+sc4h9 h+o2 = o+oh (x 0.5) ch3+ho2 = ch3o+oh

Major reaction pathways in the oxidation of 2-MB are presented in Fig. 6, for the stoichiometric mixture at 890 K which corresponds to the temperature of maximum rate of fuel consumption. In these conditions, 85% of the fuel is consumed by OH attack abstracting hydrogen; the remaining part is consumed by attacks of H and HO2. About 35% of the fuel forms butoh2m-1, the a-radical, while 24% forms the secondary radical butoh2m-3, 15% forms butoh2m2, the tertiary radical, and a total of 26% gives the primary radicals, butoh2m-4 and butoh2m-5. Fuel radicals then decompose into smaller radicals and molecules. In our mechanism we have assumed that the butoh2m-1 radical forms propanal (butoh2m1 = C2H5CHO + C2H5) instead of prop-1-en-1-ol (CH3CHCHOH), which was not observed experimentally. On the other hand, a substantial production of propanal is quantified, almost 50% of the radical butoh2m-1 forms propanal while 45% of it gives 2-methylbutanal. Hydrogen abstraction from this aldehyde mostly yields the aldehydic (46%) radical, 17% of it yields the secondary radical, 16% yields the tertiary radical and a total of 16% of 2-methylbutanal gives the primary radicals. Further decomposition of the aldehydic radicals leads to the formation of propene. Secondary radicals mostly (90%) form 2-butenes and the rest (10%) can form 2-butenal, which was not observed experimentally. The tertiary radicals yield methacrolein by beta scission, about 78% of the methacrolein is formed via this pathway. The radical butoh2m-2 mostly yields (90%) 2-methyl-1-butene, which is observed at all experimental conditions. This molecule in turn, can form 2-butanone by OH addition on the double bond (trace amounts observed experimentally) or give iso-butene by H addition on the double bond or isoprene by first hydrogen abstraction then hydrogen elimination from the formed radical, both of which were observed with the former one being more abundant (10 ppm of iso-butene at u = 1 and 45 ppm at u = 4). The secondary radical butoh2m-3 is responsible for the majority of the 2butene formation. On the other hand the primary radical butoh2m-4 mostly (70%) isomerizes to the 2-methylbutoxy radical buto2m, which yields the sec-butyl radical (CH3CH2
CHCH3) via C–C bond fission. 53% of the propene is formed by beta scission of the sec-butyl radicals in these conditions. The other primary fuel radical butoh2m-5 mainly yields 1-butene (70%), and prop-2-en1-ol to a lesser extent. This specie was not observed experimentally either, and its production/consumption has no significant effect on the system. A first order sensitivity analysis was performed on the fuel, Fig. 7, for the stoichiometric mixture at 800 K corresponding to about 45% fuel consumption and at 900 K when the rate of fuel decomposition is at its highest. Although most of the fuel sensitive reactions are similar at both temperatures, some differences are observed. Hydrogen abstraction reactions by OH and HO2 radicals forming the fuel radicals butoh2m-1 and butoh2m-2 are sensitive at both temperatures with fuel + HO2 being more important at

-0.6

In an attempt to understand the effect of the methyl group position on reactivity and product distribution of these two pentanol isomers, the present experimental results were compared to the results obtained for iso-pentanol (3-MB) oxidation by Dayma et al. [3] at the exact same conditions (10 bar, residence time of 0.7 s, u = 4, and 1500 ppm inlet fuel mole fraction). Fuel consumption and mole fractions of the most distinctive compounds observed in the oxidation of these isomers are presented in Fig. 8. According to this figure, we observe higher reactivity for 2-MB compared to 3-MB, at 800 K where 13% and 46% of the conversion occurred for 3-MB and 2-MB, respectively. This also implies that species produced during oxidation of both alcohols and that are common to both, such as CO, CO2, H2O, methane, ethylene, ethane, acetylene, propene, formaldehyde, acetaldehyde and acetone also show this difference in reactivity in their mole fraction profiles. In Fig. 8b–h, only the species that show a notable difference during oxidation of 2-MB and 3-MB are presented. It can be observed from Fig. 8d–f that mole fraction profiles of butene isomers do not evolve similarly. In 3-MB oxidation, iso-butene mole fractions are five times higher than those in 2-MB oxidation due to the fact that iso-butene is readily formed by beta scission of 3-MB fuel radical, where the radical site is on the tertiary carbon, CH3
C(CH3)(CH2CH2 OH). On the other hand, 2-butene and 1-butene formations are more important in 2-MB oxidation given that they are formed most importantly by fuel radical decomposition, i.e. by the reactions: butoh2m-3 = 2-C4H8 + CH2OH and butoh2m-5 = C4H8-1 + CH2OH.

Ignition delay time (μs)

Ignition delay time (μ μs)

0.4

4.3. Comparison of the jet-stirred reactor experimental results between 2-methylbutanol (2-MB) and iso-pentanol (3-MB)

1000

100 phi 0.25 phi 0.50 phi 1

0.25 % fuel (ϕ = 0.5)

1000

100 2.6 atm 1 atm

10

10 8

0.2

800 K and fuel + OH, at 900 K. Formation and consumption of propanal (by metathesis) and the reactions between vinyl radical and O2 are also sensitive at 900 K, while H-abstraction from 2-methylbutanal appears as a reactivity promoting reaction at 800 K.

10000

10000 K / T

0

Fig. 10. Sensitivity analysis of 2-MB ignition delay time (0.5% fuel, 1 atm, 1170 K).

0.5 % fuel (1 atm)

7

-0.2

Sensitivity coeffcient

10000

6

-0.4

9

6

7

8

10000 K / T

Fig. 9. Ignition delay times of 2-MB measured in a shock tube [6] (a) 0.5% fuel, 1 atm (b) 0.25% fuel, u = 0.5.

3012

Z. Serinyel et al. / Combustion and Flame 161 (2014) 3003–3013

393 K 433 K 473 K

Laminar flame speed (cm/s)

Laminar flame speed (cm/s)

80 80

60

40

60

1.0 bar 2.5 bar 7.5 bar

40

20

0

20 0.6

0.8

1.0

1.2

1.4

0.6

0.8

1.0

1.2

1.4

Equivalence ratio

Equivalence ratio

Fig. 11. Experimental [10] and calculated laminar flame speed of 2-MB/air mixtures (a) p = 1 bar, Tu = 393, 433 and 473 K, (b) Tu = 433 K, p = 1, 2.5 and 7.5 bar.

We should also note that some of the species formed are specific to the oxidation of each fuel given their structural difference. As an example, 2-MB produces 2-methylbutanal and 2-methyl-1-butene while 3-MB produces 3-methylbutanal and 3-methyl-1-butene. Moreover, production of species such as propyne, allene, isoprene, and methacrolein was observed during oxidation of both alcohols, all below 15 ppm. 4.4. Shock tube ignition delay time simulations The present mechanism was used to simulate the ignition delay times of dilute 2-MB/O2/Ar mixtures behind reflected shock waves presented in a recent study by Tang and co-workers [6], in which they compare relative reactivity of n-pentanol, 3-MB and 2-MB. The results are presented in Fig. 9. The temperature range for the 2-MB experiments is 1170–1640 K with reflected shock pressure of 1 atm and 2.6 atm. The kinetic model compares quite well with these data at T > 1250 K, except at the lowest temperature of interest, where the model remains slower by a factor of two at the most, around 1200 K. A logarithmic sensitivity analysis on ignition delay time was performed for the mixture with u = 0.5 at 1170 K and 1 atm (Fig. 10) by perturbing (increasing and decreasing) the A-factors of the temperature sensitive rate constants by a factor of two. The main chain branching reaction (H + O2 = O + OH) is the most sensitive one as is always the case under such conditions, followed by the reactions between CH3 and HO2 radicals with the chain terminating pathway (CH4 + O2) being the most inhibiting reaction for this system. Among the fuel reactions, hydrogen abstraction reaction by H has a small positive (inhibiting) sensitivity as it competes for H atoms with the main chain branching reaction. Otherwise fuel decomposition reactions appear to have negative (promoting) sensitivity coefficients. We can observe that ignition delay times are quite sensitive to the fuel unimolecular decomposition reactions, 2-MB = C2H5 + CH3
CHCH2OH and 2-MB = CH2OH + CH3CH2
CHCH3 unlike JSR experimental conditions. However, the reaction system C2H3 + O2 shows up again, implying its importance over a wide range of experimental conditions. 4.5. Laminar flame speed simulations In a recent study by Li and co-workers [10], laminar flame speeds of spherically propagating 2-MB/air flames were measured at several pressures (1, 2.5, 5 and 7.5 bar) and initial temperatures (393, 433 and 473 K). The authors compared 2-MB flame speeds to those of ethanol and n-butanol and found that the reactivity of 2-MB was close to that of n-butanol and lower than that of ethanol. We tested our kinetic scheme against these results and the comparisons are presented in Fig. 11. The kinetic model captures the experimental tendency of increasing flame speed with increasing unburned gas temperature and decreasing with pressure. It

also shows very good predictions as a function of the mixture equivalence ratio. 5. Conclusions Jet-stirred reactor experiments were carried out for dilute 2-MB mixtures at four different equivalence ratios (u = 0.5, 1, 2 and 4) at a residence time of s = 0.7 s at 10 atm of operating pressure and between 700 and 1200 K. Concentration profiles of stable species were obtained using GC and FTIR analyses. A kinetic sub-scheme was developed and integrated in a larger kinetic mechanism including oxidation chemistry of various alcohols. This mechanism was used to simulate the present experimental data as well as shock tube ignition delay times and laminar flame speeds taken from the literature. The model comparison was found to be satisfactory representing the global reactivity but needs further improvement especially in reproducing small unsaturated species. Oxidation intermediates and products from 2-MB and isopentanol (3-MB) oxidation were compared at same operating conditions and initial fuel concentrations. It was found that 2-MB conversion was faster than that of 3-MB for temperatures higher than 800 K, which is consistent with Tang et al. [6] study on the ignition delay times of C5 alcohols, where 2-MB has higher overall reactivity compared to 3-MB. The position of the methyl group has an effect on the product distribution of these two pentanol isomers, especially in terms of aldehydes and unsaturated hydrocarbons. Acknowledgments The research leading to these results has received funding from the European Research Council under the European Community’s Seventh Framework Programme (FP7/2007-2013)/ERC grant agreement n° 291049 2G-CSafe. References [1] C. Togbé, P. Dagaut, A. Mzé-Ahmed, P. Diévart, F. Halter, F. Foucher, Energy Fuels 24 (11) (2010) 5859–5875. [2] K.A. Heufer, J. Bugler, H.J. Curran, Proc. Comb. Inst. 34 (2013) 511–518. [3] G. Dayma, C. Togbé, P. Dagaut, Energy Fuels 25 (2011) 4986–4998. [4] S.M. Sarathy, S. Park, B.W. Weber, W.J. Wang, P.S. Veloo, A.C. Davis, C. Togbé, C.K. Westbrook, O. Park, G. Dayma, Z.Y. Luo, M.A. Oehlschlaeger, F.N. Egolfopoulos, T.F. Lu, W.J. Pitz, C.J. Sung, P. Dagaut, Combust. Flame 160 (12) (2013) 2712–2728. [5] W. Liu, A.P. Kelley, C.K. Law, Proc. Comb. Inst. 33 (2010) 995–1002. [6] C. Tang, L. Wei, X. Man, J. Zhang, Z. Huang, C.K. Law, Combust. Flame 160 (2013) 520–529. [7] T. Tsujimura, W.J. Pitz, F. Gillespie, H.J. Curran, B.W. Weber, Y. Zhang, C.-J. Sung, Energy Fuels 26 (2012) 4871–4886. [8] C. Togbé, F. Halter, F. Foucher, C. Mounaim-Rousselle, P. Dagaut, Proc. Comb. Inst. 33 (1) (2011) 367–374. [9] S. Atsumi, T. Hanai, J.C. Liao, Nature 451 (2008) 86–89. [10] Q. Li, E. Hu, Y. Cheng, Z. Huang, Fuel 112 (2013) 263–271.

Z. Serinyel et al. / Combustion and Flame 161 (2014) 3003–3013 [11] P. Dagaut, M. Cathonnet, J.P. Rouan, R. Foutatier, A. Quilgars, J.C. Boettnet, F. Gaillard, H. James, J. Phys. E – Sci. Instrum. 19 (3) (1986) 207–209. [12] P. Dagaut, F. Lecomte, S. Chevailler, M. Cathonnet, Combust. Sci. Tech. 139 (1998) 329–363. [13] P. Dagaut, S.M. Sarathy, M.J. Thomson, Proc. Comb. Inst. 32 (2009) 229–237. [14] C. Muller, V. Michel, G. Scacchi, G.M. Côme, J. Chim. Phys. 92 (1995) 1154– 1178. [15] S.W. Benson, Thermochemical Kinetics, Wiley, New York, 1976. p.^pp. [16] R.J. Kee, F.M. Rupley, J.A. Miller, in: SAND89-8009, Sandia National Laboratories, Livermore (CA), 1989. [17] C.-W. Zhou, J.M. Simmie, H.J. Curran, Combust. Flame 158 (2011) 726–731. [18] R. Atkinson, IJCK 18 (1986) 555–568. [19] J. Zheng, D.G. Truhlar, Phys. Chem. Chem. Phys. 12 (2010) 7782–7793. [20] L.B. Harding, Y. Georgievskii, S.J. Klippenstein, J. Phys. Chem. A 109 (2005) 4646–4656.

[21] [22] [23] [24] [25] [26] [27] [28] [29] [30]

3013

W. Tsang, J. Phys. Chem. Ref. Data 17 (2) (1988) 887–951. W. Tsang, J. Phys. Chem. Ref. Data 19 (1) (1990) 1–68. H.-H. Carstensen, A.M. Dean, ACS Symp. Ser. 1052 (2010) 201–243. H.J. Curran, IJCK 38 (2006) 250–275. F. Buda, R. Bounaceur, V. Warth, P.A. Glaude, R. Fournet, F. Battin-Leclerc, Combust. Flame 142 (2005) 170–186. B. Heyberger, F. Battin-Leclerc, V. Warth, R. Fournet, G.M. Côme, G. Scacchi, Combust. Flame 126 (2001) 1780–1802. M.V. Johnson, S.S. Goldsborough, Z. Serinyel, P. O’Toole, E. Larkin, G. O’Malley, H.J. Curran, Energy Fuels 23 (2009) 5886–5898. G. da Silva, J.W. Bozzelli, Chem. Phys. Lett. 483 (2009) 25–29. J.-X. Shao, C.-M. Gong, X.-Y. Li, J. Li, Theor. Chem. Acc. 128 (2001) 341–348. C.M. Rosado-Reyes, W. Tsang, IJCK (2013), http://dx.doi.org/10.1002/ kin.20828.