Effects of N-Decane Substitution on Structure and Extinction Limits of Formic Acid Diffusion Flames

Effects of N-Decane Substitution on Structure and Extinction Limits of Formic Acid Diffusion Flames

Item Type Preprint

Authors Alfazazi, Adamu; Li, Jiajun; Xu, Chaochen; Es-sebbar, Et-touhami;

Zhang, S. Xiaoyuan; Abdullah, Marwan; Younes, Mourad; Sarathy, Mani; Dally, Bassam

Citation Alfazazi, A., Li, J., Xu, C., Es-sebbar, E., Zhang, S. X., Abdullah, M., Younes, M., Sarathy, M., & Dally, B. (2023). Effects of N-Decane Substitution on Structure and Extinction Limits of Formic Acid Diffusion Flames. https://doi.org/10.2139/ssrn.4479861

Eprint version Pre-print

DOI 10.2139/ssrn.4479861

Publisher Elsevier BV

Rights This is a preprint version of a paper and has not been peer reviewed. Archived with thanks to Elsevier BV.

Download date 21/06/2023 06:13:51

Link to Item http://hdl.handle.net/10754/692670

1

Effects of n-decane substitution on structure and extinction limits

2

of formic acid diffusion flames

3

4 Adamu Alfazazi^a*, Jiajun Li^a,Chaochen Xu^a, Et-touhami Es-sebbar^a, S. Xiaoyuan Zhang^a, 5 Marwan Abdullah^b, Mourad Younes^b, S. Mani Sarathy^a, Bassam Dally^a

6

7 ^aKing Abdullah University of Science and Technology (KAUST), Clean Combustion 8 Research Center (CCRC), Thuwal 23955-6900, Saudi Arabia

9 ^b Saudi Aramco, Dhahran, 31311, Saudi Arabia 10

11

12 Corresponding Authors:

13 Adamu Alfazazi

14 adamu.fazazi@kaust.edu.sa 15

16 Professor Bassam Dally 17 Bassam.dally@kaust.edu.sa 18

19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

34

Preprint not peer reviewed

36 Abstract

37 Formic acid (FA) is a low-carbon fuel that can be produced from renewable hydrogen (H2) and 38 carbon dioxide (CO2). However, transitioning completely to low carbon fuels might be lengthy 39 and challenging, so it makes sense to gradually introduce low carbon fuels into the energy sector 40 in conjunction with existing fuels. One practical approach could be blending FA with existing 41 fossil fuels, such as kerosene in aviation. This study investigates the behavior of flames 42 comprising FA and its blends with n-decane. The extinction limits of pure FA and FA/n-decane 43 mixtures were measured in a counterflow non-premixed laminar flame setup at various blending 44 ratios. Additionally, the flame structure was studied by measuring temperature and species 45 distribution using probe-based sampling. The results indicate that replacing FA with n-decane 46 in the FA-N2 flame greatly enhances the flame's reactivity. Blending the fuels also led to a 47 decrease in CO2 production in the flame, while increasing the flame temperature and the 48 concentration of H2 and CO species. Experimental results were modeled using an updated FA 49 and n-decane model. The kinetic model agrees with the experimental trends, but slightly 50 overpredicts H2, CH4 species, and flame temperature. The kinetic model was also used to 51 investigate the kinetic coupling between FA and n-decane in non-premixed laminar flames.

52 Kinetic analyses indicate that HOCO and OCHO intermediates play a crucial role in pure FA 53 flames by reacting with active radicals like H and OH to produce CO2 and CO, while in n- 54 decane blended flames, fuel decomposition proceeds through two other pathways leading to H2

55 and H2O2.

56 57 58 59 60

61 Key Words: Formic Acid; Flame Extinction; non-premixed flames; Counterflow flames

Preprint not peer reviewed

62 1.0 Introduction

63 The rise in global energy consumption has led to a significant increase in greenhouse gases and 64 pollutants such as unburned hydrocarbons, carbon monoxide, and nitrous oxides. Emissions 65 resulting from burning fossil fuels in the transportation and power industries constitute 50% of 66 the overall greenhouse gas emissions in the United States [1]. To meet the climate ambitions 67 stated by many countries, there is a need to shift to lower carbon and sustainable fuels.

68 As the world races to adopting clean energy source, hydrogen is often viewed as a promising 69 energy carrier and fuel. Hydrogen burns without carbon-related emissions. However, hydrogen 70 is a gas at room temperature and highly flammable, which render it difficult to transport and 71 store. One solution is to use a hydrogen carrier such as ammonia (NH3) or formic acid (FA) to 72 store and transport hydrogen. Compared to pressurized hydrogen, formic acid (FA) is liquid at 73 room temperature and can store almost twice as much energy at equal volume [2, 3]. This allows 74 for a more compact design of storage tanks in vehicles. Formic acid can be produced from 75 renewable H2 and captured CO2, biomass decomposition, or as an industrial by-product via 76 photoelectric catalytic reduction of CO2 [4, 5]. Formic acid decomposes at mild temperature 77 conditions to release H2, which is an important attribute for on-demand release and utilization 78 in vehicles that are using hydrogen fuel cells [2].

79 The development of models that could be used to understand the combustion behavior of any 80 candidate fuel in realistic energy systems, relies on fundamental experimental data for 81 validation. These data include flame speed, ignition delay time, flame extinction and speciation.

82 However, as mentioned recently by Sarathy et al. [4], most previous studies on FA focus on 83 understanding its decomposition behavior in shock tube and flow reactors, and only a few 84 reported its flame behavior. Marshall et al. [6] developed a detailed kinetic model for FA 85 oxidation. They tested their model using old laminar burning velocity (LBV) data by Wilde et

Preprint not peer reviewed

87 measurement, especially under rich conditions. Sarathy et al. [4] developed a FA kinetics model 88 by merging the FA sub mechanism with Aramco 2.0 base chemistry and measured the LBV of 89 mixtures containing HOCOOH, H2, and CO2 in spherical flames. They compared their 90 measurements using both Marshall et al. [6] mechanism and their own mechanism. Their 91 developed mechanism better reproduces the experimental trend, especially under lean and 92 stoichiometric conditions. Osipova et al. [8] used the molecular beam mass spectrometer to 93 measure the laminar premixed flame structure of HCOOH/H₂/O₂/Ar in a flat flame burner. Both 94 models, developed by [6, 9], captured the mole fractions of the reactants, but over-predicted the 95 concentration of H2, H2O2 and HO2 species.

96 Lavadera et al. [10] measured the LBV of FA+CH4+Air flames using the heat flux method and 97 noted disparities in their measured flame speed with the data reported by Yin et. al. [11], 98 especially under rich conditions. They updated the Glarborg et al. [6] mechanism by modifying 99 the rate constant for HOCO(+M)=H+CO2(+M), which significantly improved the performance 100 of the model even at rich conditions. In another study, Yin et al. [12] used a high-level quantum 101 calculation to obtain the rate coefficients of missing reactions in Glarborg et al. [6] kinetic 102 mechanism. They developed a detailed chemical kinetic mechanism that includes low 103 temperature oxidation and pyrolysis reactions of FA. Their model reproduced speciation data 104 from FA pyrolysis and oxidation in JSR and flame speed data by [4, 11] under rich and lean 105 conditions.

106 All previous studies focused on flame speed data and were performed in homogeneous systems 107 where the effects of transport are negligible. In practical combustion systems such as turbines 108 and compression ignition engines, combustion takes place in the presence of concentration and 109 temperature gradients, such that reactants and intermediate products diffusion can affect the

Preprint not peer reviewed

110 reaction kinetics, and therefore, can contribute to flame propagation, emission and stability 111 [13-17].

112 The low LBV of FA is a drawback as a potential fuel for practical applications. As a result, 113 many studies have explored blending FA fuel with reactivity promoters such as CH4, H2, and 114 H2/CO2. A similar procedure is used for low reactivity hydrogen-carrying fuels such as 115 ammonia [18-21]. The aim of this study is to investigate the potential of formic acid as a fuel 116 source by analyzing its behavior in non-premixed diffusion flames. The counterflow 117 configuration is used to understand the flame stability of FA and its blends with n-decane 118 (nC10H22), a relevant surrogate fuel for kerosene in combustion research [22, 23]. The extinction 119 limits of both pure FA flame and flames of FA/n-decane blends under various conditions will 120 be measured. The flame structure will also be analyzed by measuring temperature and species 121 distribution using gas chromatography. Additionally, the study tests the predictive capability of 122 a recently developed chemical kinetics model for these fuel blends [12] in non-homogeneous 123 environments, such as those found in turbines.

124 2 Description of the Experimental and Numerical Procedure 125 2.1 Experimental Procedure

126 Figure 1 shows a schematic illustration of the counterflow configuration used in this study. The 127 burner consists of two opposing ducts, each having an internal diameter of 20 mm. The upper 128 burner carries an oxidizer stream, while the lower burner contains the fuel stream. These jets 129 flow into the mixing layer between the two ducts. The fuel stream in this case comprises the 130 fuel (pure FA, pure n-decane or FA + n-decane mixture) and nitrogen. The liquid fuel is injected 131 using a high precision syringe pump. The liquid fuel is first vaporized and then introduced into 132 the counterflow system. A thermocouple is used to monitor the temperature of the fuel stream 133 inside the vaporizer and at the exit of the fuel outlet. The flow rates of air and nitrogen are

Preprint not peer reviewed

134 adjusted by computer-regulated mass flow controllers. The two ducts are separated by a distance 135 L. The fuel mass fraction, temperature, density of the fuel stream, and the component of the 136 fuel flow velocity normal to the stagnation plane at the exit of the fuel outlet were Yf,1, T1, ρ1, 137 and V1, respectively. The oxygen mass fraction, oxidizer temperature, density, and the oxidizer 138 flow velocity normal to the stagnation plane at the exit of the oxidizer outlet were YO2,2, T2, ρ2, 139 and V2, accordingly. All experiments were conducted assuming plug flow conditions and 140 keeping the momenta of the counterflowing streams equal (𝜌𝑉𝑖2,=1,2). The strain rate, a2, Eq.

141 (1), is defined as the gradient of the normal component of the flow velocity [24].

142 𝑎₂= ^2|𝑉2|_𝐿 (1 +^|𝑉1|_|𝑉2| ^𝜌_𝜌¹

2) (1)

143 Extinction experiments were carried out at T₂ = 298 K, and T₁ kept close to the boiling point of 144 the fuel mixture. To measure the extinction limits, the flame is stabilized at a given fuel and 145 oxidizer mass fraction Yf,1, and YO2. Subsequently, the velocities V1 and V2 were gradually 146 increased by increasing the flow rates while maintaining a balanced momentum between the 147 two counterflowing streams until the flame was extinguished. The corresponding strain rate at 148 extinction, a_2,E, is then calculated using Eq. (1) above. The strain rate at stable flaming condition 149 is lower than the extinction strain rate. The process is repeated to verify each data point. The 150 accuracies of the strain rate was estimated at 7% of the recorded values. The experimental 151 repeatability of the reported strain rate at extinction was 3% of the recorded value. The flame 152 temperatures were measured using an Omega (Pt/10% Rh-Pt) thermocouple with a wire diameter of 153 0.075 mm. The measured temperatures are corrected according to [25, 26] by considering the heat 154 transfer due to radiation and convection between the thermocouple bead and the surrounding.

155 The gas products were quantified using a gas chromatograph (GC, Thermo Scientific TRACE 156 1610). The GC was fitted with two pre-columns (Porapak N), two analytical columns (5Å 157 MoleSieve), one mixed pre-column (TG-WAXMS+TR-1), one HP AL/S column, and two 158 TCDs and one FID. Ultra-high purity He and N

Preprint not peer reviewed

2 were used as carrier gases. The gas product

159 was injected directly into the sampling loop of the GC by a vacuum-assisted gas sampling 160 system. The TCD and FID detectors showed good linearity over a wide concentration range 161 toward H2, O2, CO, CO2, CH4, C2H2, C2H4, and C2H6. The error for H2 speciation is less than 162 5%. Similarly, the errors for CO, CO2, and hydrocarbons are less than 4.0%.

163 2.2 Numerical Procedure

164 The Yin et al. [12] mechanism for FA, was used for the simulation. The mechanism was 165 extended to include the recently developed reaction mechanisms for n-decane. Brunialti et al.

166 [27] developed a methodology for automatic generation of predictive lumped and detailed 167 mechanisms for normal and branched alkanes including n-decane. A comprehensive detailed 168 reaction mechanism for combustion of alkanes was generated by employing an updated version 169 of MAMOX++ software, where recent progress in the low temperature reaction classes and rate 170 rules are incorporated. The developed kinetic mechanism agreed well with experimental data 171 in the literature, including the speciation data measured in the JSR and global ignition delay 172 times [28-30].

173 Extinction simulations were carried out using the 1-D extinction solver on CHEMKIN-PRO.

174 The solver uses an arc length continuation method to generate the S-curve [16, 31]. The 175 procedure involves establishing a stable flame using the OPPDIF code at conditions close to 176 extinction, and then the solution is restarted in the extinction solver. A 2-point extinction 177 method with 1000 steps was used. Convergence factors GRAD and CURV = 0.1 were used to 178 control the maximum gradients and curvatures allowed between grid points. The 179 multicomponent mixture formulation was used to determine species diffusive coefficients and 180 fluxes.

Preprint not peer reviewed

181

182 Figure 1. Schematic illustration of the counterflow facility

183 3 Results and Discussion 184 3.1 Extinction measurement

185 For validation purposes, the extinction strain rates of n-decane-N2 were measured and compared 186 to previous experimental data by Humer et al. [32]. The results are shown in Fig. 2. Overall, the 187 present experimental results show a good agreement with the previous data. As Formic acid 188 exhibits low reactivity, it cannot sustain a flame in air. Consequently, an oxidizer enriched with 189 oxygen (XO2=0.5) is employed to determine its extinction limit (see Fig. 2). As illustrated in 190 Fig. 2, the extinction limit of both flames increases as the fuel mass fraction is increased. FA- 191 N2 flame is more prone to extinction as compared to n-decane-N2. Extinction limit measures 192 flames' resistance to local extinction; therefore, flames with higher extinction limits are more 193 resistant and correspondingly more stable. In other words, the area under the lines in Fig. 2 194 represents a stabilized and burning flame region. Above the lines, the flame cannot withstand 195 the leakages of heat and radicals, and therefore the flame is extinguished. This trend in reactivity 196 of FA results also agree with results reported by Sarathy et al. [4] where they reported a very

N2 N2

Fuel stream

FA supply L

N2 Supply Decane MFCs, syringes

Stagnation Plane

N2 Oxidizer Stream N2

O2 supply

N2 supply

N2 supply MFC

MFC MFC

MFC

Preprint not peer reviewed

197 low LBV for formic acid. Worth noting is that n-decane has more active sites for H2 abstraction, 198 and when it reacts in flames it populates active radicals which sustain the flame.

199

200 Figure 2. The extinction strain rate a2,E as a function of the fuel mass fraction for FA-N2 and n- 201 decane-N2 fuels. Purple and Red stars are present measurements, while red circles are 202 measurements from Humer et al., 2007 [32]. Lines are numerical simulations.

203 Figure 3 shows the dependence of the extinction strain rate on the fuel mass fraction for 204 nitrogen-diluted n-decane/air and n-decane blended FA/air flames. The blending ratio is defined 205 as the liquid volume ratio between n-decane and FA, and this is systematically varied from 10%

206 n-decane (i.e. 90% FA) to 100% n-decane. The results show an increase in the reactivity of FA- 207 N2 flame as more FA is substituted by n-decane in the mixture. This shows the ability of n- 208 decane to improve the flammability limit of FA in opposed flow diffusive systems. Also as 209 illustrated in Fig. 3, the model agreed with the observed experimental trend for both fuel 210 mixtures. The flame structure and kinetic analysis below rationalize why the extinction strain 211 rate of FA is significantly affected by n-decane substitution.

212

Preprint not peer reviewed

213

214 Figure 3. Mass fraction of fuel as a function of strain rate at extinction, a2,E in the counterflow 215 diffusion flame. Symbols represent experimental measurement, dash-dot lines represent 216 modeling predictions.

217 3.2 Measured and simulated flame structure

218 Figure 4 illustrates the temperature measurements along with the measured concentration 219 profiles of CO, CO2, and H2 for two types of flames: FA-N2 and a 30% n-decane substituted 220 flame. These measurements were conducted at strain rate = 41 s^-1 with an oxidizer containing 221 50% O2. The H2 concentration and temperature of the nitrogen-diluted FA flame showed a 222 significant increase upon the substitution of 30% n-decane into the fuel stream. Additionally, 223 the n-decane substituted flame exhibited a higher concentration of CO and H2 compared to the 224 FA-N2 flame. The impact of CO on the observed higher reactivity of the substituted flame is 225 analyzed in detail in the following section. It is worth noting that the maximum concentration 226 of CO2 in the nitrogen-diluted FA flame was found to be approximately 21000ppm, whereas it 227 reduced to 18000ppm with n-decane substitution.

Preprint not peer reviewed

228

229

230 Figure 4. Measured temperature and species concentration of laminar non-premixed 231 counterflow flames with the fuel blends. Blue lines and symbols represent the profiles for the 232 n-decane substituted flame, and black lines and symbols represent the profiles for the FA-N2 233 flame. The vertical error bars represent the errors related to the concentration results; the 234 horizontal error bars represent the errors related to the probe sampling position, which are ± 235 0.075 mm for each experimental data.

236 Figure 5a presents both simulated and measured temperature and species concentration profiles 237 for FA-N2 flame. The model quantitatively reproduces CO2 concentration profiles, but it over- 238 predicts the measured temperature and species of H2 and CO in the FA-N2 flame. Similarly, 239 Fig. 5b illustrates the measured and simulated flame structure of n-decane substituted flame.

240 The model reproduces CO2, CO, C2H2, C2H4 species but slightly overpredicted flame 241 temperature and CH4 concentration. Moreover, a better agreement is observed for H2 in 242 comparison to the FA-N2 flame in Fig. 5a.

Preprint not peer reviewed

243

244

245 Figure 5a. Measured and simulated flame temperature and species concentration of FA-N2 246 flame. Lines are numerical simulation while symbols plus lines are experimental measurements.

247 The vertical error bars represent the errors related to the concentration results; the horizontal 248 error bars represent the errors related to the probe sampling position, which are ± 0.075 mm for 249 each experimental data.

Preprint not peer reviewed

250

251

Preprint not peer reviewed

252

253

254 Figure 5b. Measured and simulated flame temperature and species concentration of n-decane 255 substituted flame. Lines are numerical simulation while symbols plus lines are experimental 256 measurements. The vertical error bars represent the errors related to the concentration results;

257 the horizontal error bars represent the errors related to the probe sampling position, which are 258 ± 0.075 mm for each experimental data.

259 3.2.1 Effects of n-decane substitution on fuel consumption

260 Reaction path flux analyses were performed at 65% fuel consumption to compare fuel 261 consumption pathways between the FA-N2 and the n-decane substituted FA flames as illustrated 262 in Fig. 6. The 30% n-decane substituted FA flame is used. These analyses were conducted under 263 the same conditions depicted in Fig. 5. For the FA-N2 flame, approximately 76% of the fuel 264 decomposition proceed via HOCO intermediate route, where FA undergoes H abstraction 265 reaction by OH and H to produce H2O and H2 as the secondary products. Around 24% of FA 266 consumes hydroxyl radical (OH) to produce OCHO and H2O. About 94% of HOCO 267 intermediates in the FA-N2 flame dissociated into OH and CO, while the remaining react with

Preprint not peer reviewed

268 hydrogen radicals (H) to produce H2O and CO. Almost 100% of OCHO intermediate dissociates 269 into CO2 and H radicals.

270 In the 30% n-decane substituted FA flame, two additional pathways were observed: about 4.3%

271 of the fuel dissociated directly into H2 and CO2, and 5.6% of the fuel reacted with hydroperoxy 272 radicals to produce HOCO and hydrogen peroxide. The remaining fuel proceeded through the 273 pathways of HOCO and OCHO intermediates to generate CO, CO2, and other secondary 274 radicals.

275 The primary difference between these flames is related to the consumption of FA and of course, 276 the presence of n-Decane fragments. In FA-N2 flames, fuel decomposition primarily involves 277 HOCO and OCHO intermediates that consume H and OH active radicals to produce CO and 278 CO2. In contrast, two other pathways leading to H2 and H2O2 production were observed in the 279 30% n-decane substituted FA flame. The remaining fuel in this case proceeded through the 280 chain propagation route to produce HOCO, which further reacted to yield CO and CO2. 281 Increasing the CO, OH, and H concentration increases the rate of heat release in the flame 282 through the CO+OH=CO2+H and H+O2=OH+O reactions, as demonstrated in the sensitivity 283 analyses shown in the later section. Additionally, in the FA-N2 flame, 6.8% of CO was 284 consumed via the HCO formation route (CO+H+MHCO+M), with most of the HCO reacting 285 with O2 to produce HO2 and CO again. Moreover, while 100% of HOCO produced CO and OH 286 in the 30% substituted n-decane flame, around 3.7% and 2.5% of HOCO reacted with H radicals 287 to respectively produce CO, CO2 and H2O in the FA-N2 flame. Overall, the radicals-scavenging 288 reactions observed in FA-N2 flames, along with the pathways leading to the production of H2O2

289 and H2 observed in n-decane blended flames, could provide an explanation for the lower 290 reactivity in FA-N2 flames and the increase in reactivity resulting from n-decane substitution.

Preprint not peer reviewed

291

292 Figure 6. Reaction flux analyses showing major pathways for fuel and intermediates 293 consumption. Red are percentages for FA flame, gray represents percentages for the 30% n- 294 decane substituted FA flame.

295 To understand which reactions affect the extinction limits of the FA-N2 and 30% n-decane 296 substituted flames, brute force sensitivity analyses were conducted and are presented in Fig. 7.

297 The analyses used percentage sensitivity [33, 34], which is defined as:

298 % S =

(

^а𝐸𝑖_а^{― а}_{𝐸𝑛𝑜𝑟𝑚}^{𝐸𝑛𝑜𝑟𝑚}

) ∗ 100

(2)

299 where а_𝐸𝑖 is the extinction strain rate when the rate constant is doubled (k = 2*k) and а_{𝐸𝑛𝑜𝑟𝑚} 300 is the extinction strain rate for an unchanged rate in the mechanism.

301 Prior to the brute force sensitivity analyses, a temperature sensitivity analysis was performed 302 under the same conditions to identify the most sensitive reactions [35]. Increasing the rate of 303 reactions with positive sensitivity enhances reactivity, resulting in higher extinction limits, 304 while decreasing their rates decreases reactivity and vice versa. Both flames are positively 305 impacted by branching reactions that produce O, OH, and H radicals. They are also sensitive to 306 heat-producing reactions such as CO+OHCO2+H. Although not shown in Fig. 7, this reaction

HO O

(Formic Acid)

O O CO₂

CO

+OH/-H2O(24%)(16.5%)

-H(100%)(100%)

+H/-H₂O(3.7%) -OH(93.5%)(100%) +OH/-H

2O(44%) (34.7%) +H/-H

2(32%) (37%) +HO2/-H2O2

(5.6%)

C O OH

+H/-H2 O(2.5%) +OH/-H

(93%) (94%) +HO

2/-OH (5%) -H

2(4 .3%)

OCHO

HOCO

Preprint not peer reviewed

307 accounts for most of the heat released in the FA-N2 flame. The FA-N2 flame's reactivity is 308 negatively impacted by FA fuel fragments, especially HOCHO+OH HOCO+H₂O and 309 HOCO+HCO+H2O. In this case, OH is consumed to produce HOCO and H2O, which 310 scavenges H radical to form CO and H2O. As previously discussed in the reaction flux analyses, 311 subsequent CO reactions in the FA-N2 flame result in the formation of HCO, which reacts with 312 H radicals to produce HO2 and CO. In contrast, HOCO producing reactions are positively 313 sensitive in the n-decane substituted flame, where nearly 100% of HOCO proceeds through the 314 CO and OH formation route. As shown in Fig. 7, the n-decane substituted flames are negatively 315 affected by beta-scission products of n-decane fragments, which produce intermediates that 316 undergo subsequent reactions to produce highly branched and resonantly stable products.

317

318 Figure 7. Sensitivity analysis of FA-N2, and n-decane substituted flames to extinction 319 3.2.1 Effects of n-decane substitution on FA flame structure

320 The simulated species profiles in Fig. 5 are combined and shown as illustrated in Fig. 8, in order 321 to trace the fuel consumption, radicals formation and production of CO/CO2 species. The 322 maximum flame temperature for FA-N2 flame is ~1600K, however with n-decane substitution, 323 the flame temperature increase by more than 400K. The lower flame temperature in FA-N2

Preprint not peer reviewed

325 decreases the overall reactivity of the flame. It is worthy to note that H, OH, and O radicals are 326 higher in the n-decane substituted FA flame, and can diffuse across the stagnation plane to 327 sustain the flame, by initiating more fuel abstraction. This improves the flames’ stability and 328 increases its resistance to extinction. It can be seen in Fig. 8, that the fuel and the oxidizer are 329 consumed earlier in the n-decane substituted flame, which is an indication of fast oxidation as 330 a result of the enhanced activity of the radicals produced by the flame.

331

332 Figure 8. Effects of n-decane substitution on the structure of the FA flame. Dashed line 333 represents the FA-N2 flame; while straight lines are 30%, n-decane substituted flame.

334 Simulation are carried out at strain rate (41 s^-1), YF = 0.5 and XO2 in oxidizer is 0.5.

335 4.0 Conclusions

336 The present study investigated the flame structure of formic acid (FA) and n-decane blends in 337 a counterflow laminar non-premixed flame configuration. Both experimental and simulations 338 were conducted in this study. The results indicate that replacing FA with n-decane in the FA- 339 N2 flame greatly enhances the flame's reactivity of FA-based flames and reduces CO2

340 production while increasing the concentrations of H2 and CO species in the flame. The 341 experimental results were simulated using Yin et al. chemical kinetics mechanism [12], which 342 was extended to include Brunialti and Zhang et al.’s [27] n-decane sub-chemistry. The model 343 was in good agreement with the measured extinction limits and flame structure, and helped to

Preprint not peer reviewed

344 elucidate the kinetic coupling between FA and n-decane in non-premixed laminar flames.

345 Results from kinetic analyses revealed that HOCO and OCHO are important intermediates for 346 the FA-N2 flames, while n-decane blended flames decompose through two additional pathways 347 leading to H2 and H2O2 formation. Sensitivity analyses showed that both types of flames are 348 sensitive to heat-producing and branching reactions, as well as H abstraction reactions of FA 349 by OH radical. The FA-N2 flame is negatively sensitive to HOCO radicals producing reactions, 350 while n-decane substituted flames show positive sensitivity as nearly 100% of HOCO reactions, 351 the flame proceeds through the CO and OH formation route. The 30% n-decane substituted 352 flame is also sensitive to reactions involving n-decane fuel fragments. Finally, the n-decane 353 substituted FA flame exhibited the highest extinction limit due to its higher flame temperature 354 and concentration of active radicals. These findings provide valuable insights into the viability 355 of FA and its blends, with higher hydrocarbons as potential alternative fuel sources for practical 356 applications.

357 CRediT authorship contribution statement

358 Adamu Alfazazi: Conceptualization, Methodology, Investigation, Experiment/numerical 359 simulation, Formal analysis, Writing – original draft. Jiajun Li: Methodology, Experiment, 360 numerical simulations and plots. Chaochen Li: Analytical measurements. Et-touhami Es-sebbar:

361 Experimental measurements. S. Xiaoyuan Zhang: Chemical kinetics model development.

362 Marwan Abdullah, Mourad Younes: funding acquisition, review. S. Mani Sarathy: funding 363 acquisition, writing, review & editing, conceptualization. Bassam Dally: Supervision, Formal 364 analysis, Writing – review & editing.

365 Acknowledgements

366 The authors wish to thank the Saudi Aramco Research and Development Center for funding 367 this project under research agreement number RGC/3/3837-01-01.

Preprint not peer reviewed

368 Reference:

369 [1] U.S.E.P. Agency, The 2022 EPA Automotive Trends Report, Report No. EPA-420-S-22-001, 2022.

370 [2] X. Wang, Q. Meng, L. Gao, Z. Jin, J. Ge, C. Liu, W. Xing, Recent progress in hydrogen production 371 from formic acid decomposition, Int. J. Hydrog. Energy 43 (2018) 7055-7071.

372 [3] Z. Xin, J. Zhang, K. Sordakis, M. Beller, C.X. Du, G. Laurenczy, Y. Li, Towards Hydrogen Storage 373 through an Efficient Ruthenium-Catalyzed Dehydrogenation of Formic Acid, ChemSusChem 11 (2018) 374 2077-2082.

375 [4] S.M. Sarathy, P. Brequigny, A. Katoch, A.M. Elbaz, W.L. Roberts, R.W. Dibble, F. Foucher, Laminar 376 Burning Velocities and Kinetic Modeling of a Renewable E-Fuel: Formic Acid and Its Mixtures with H2 377 and CO2, Energ. Fuel 34 (2020) 7564-7572.

378 [5] X. Wang, Q. Meng, L. Gao, Z. Jin, J. Ge, C. Liu, W. Xing, Recent progress in hydrogen production 379 from formic acid decomposition, International Journal of Hydrogen Energy 43 (2018) 7055-7071.

380 [6] P. Marshall, P. Glarborg, Ab initio and kinetic modeling studies of formic acid oxidation, Proc.

381 Combust. Inst. 35 (2015) 153-160.

382 [7] E. de Wilde, A. van Tiggelen, Burning velocities in mixtures of methyl alcohol, formaldehyde or 383 formic acid with oxygen, Bulletin des Sociétés Chimiques Belges 77 (1968) 67-75.

384 [8] K.N. Osipova, S.M. Sarathy, O.P. Korobeinichev, A.G. Shmakov, Chemical structure of atmospheric 385 pressure premixed laminar formic acid/hydrogen flames, Proc. Combust. Inst. 38 (2021) 2379-2386.

386 [9] M. Christensen, A.A. Konnov, Laminar burning velocity of acetic acid + air flames, Combust. Flame 387 170 (2016) 12-29.

388 [10] M. Lubrano Lavadera, A.A. Konnov, Laminar burning velocities of methane + formic acid + air 389 flames: Experimental and modeling study, Combust. Flame 225 (2021) 65-73.

390 [11] G. Yin, Q. Gao, E. Hu, J. Xu, M. Zhou, Z. Huang, Experimental and kinetic study on laminar flame 391 speeds of formic acid, Combust. Flame 220 (2020) 73-81.

392 [12] G. Yin, J. Xu, E. Hu, Q. Gao, H. Zhan, Z. Huang, Experimental and kinetic study on the low 393 temperature oxidation and pyrolysis of formic acid in a jet-stirred reactor, Combust. Flame 223 394 (2021) 77-87.

395 [13] S. Deng, P. Zhao, D. Zhu, C.K. Law, NTC-affected ignition and low-temperature flames in 396 nonpremixed DME/air counterflow, Combust. Flame 161 (2014) 1993-1997.

397 [14] A.T. Holley, Y. Dong, M.G. Andac, F.N. Egolfopoulos, Extinction of premixed flames of practical 398 liquid fuels: Experiments and simulations, Combust. Flame 144 (2006) 448-460.

399 [15] A. Alfazazi, A. Al-Omier, A. Secco, H. Selim, Y. Ju, S.M. Sarathy, Cool diffusion flames of butane 400 isomers activated by ozone in the counterflow, Combust. Flame 191 (2018) 175-186.

401 [16] A. Alfazazi, U. Niemann, H. Selim, R.J. Cattolica, S.M. Sarathy, Effects of Substitution on 402 Counterflow Ignition and Extinction of C3 and C4 Alcohols, Energ. Fuel 30 (2016) 6091-6097.

403 [17] A. Alfazazi, G. Mairinger, H. Selim, K. Seshadri, S. Mani Sarathy, Counterflow ignition and 404 extinction of FACE gasoline fuels, Proc. Combust. Inst. 38 (2021) 2323-2331.

405 [18] R.C. da Rocha, M. Costa, X.-S. Bai, Chemical kinetic modelling of ammonia/hydrogen/air ignition, 406 premixed flame propagation and NO emission, Fuel 246 (2019) 24-33.

407 [19] J.W. Ku, S. Choi, H.K. Kim, S. Lee, O.C. Kwon, Extinction limits and structure of counterflow 408 nonpremixed methane-ammonia/air flames, Energy 165 (2018) 314-325.

409 [20] A.M. Elbaz, S. Wang, T.F. Guiberti, W.L. Roberts, Review on the recent advances on ammonia 410 combustion from the fundamentals to the applications, JFUECO. 10 (2022).

411 [21] A. Alfazazi, E.-t. Es-sebbar, X. Zhang, B. Dally, M. Abdullah, M. Younes, S.M. Sarathy, Counterflow 412 flame extinction of ammonia and its blends with hydrogen and C1-C3 hydrocarbons, Applications in 413 Energy and Combustion Science 12 (2022) 100099.

414 [22] S.H. Won, W. Sun, Y. Ju, Kinetic effects of toluene blending on the extinction limit of n-decane 415 diffusion flames, Combust. Flame 157 (2010) 411-420.

416 [23] S. Jahangirian, S. Dooley, F.M. Haas, F.L. Dryer, A detailed experimental and kinetic modeling 417 study of n-decane oxidation at elevated pressures, Combust. Flame 159 (2012) 30-43.

Preprint not peer reviewed

418 [24] K. Seshadri, F.A. Williams, Laminar flow between parallel plates with injection of a reactant at 419 high reynolds number, Int. J. Heat Mass Transf. 21 (1978) 251-253.

420 [25]

421 S.C. R. Correcting thermocouple measurements for radiation loss a critical review. In: editor^editors.

422 Proceedings of the 33'd National Heat Transfer Conference; 1999; Albuquerque, New Mexico. p.

423 [26] K.B. Brady, X. Hui, C.-J. Sung, K.E. Niemeyer, Counterflow ignition of n-butanol at atmospheric 424 and elevated pressures, Combust. Flame 162 (2015) 3596-3611.

425 [27] S. Brunialti, X. Zhang, T. Faravelli, A. Frassoldati, S.M. Sarathy, Automatically generated detailed 426 and lumped reaction mechanisms for low- and high-temperature oxidation of alkanes, Proc.

427 Combust. Inst., doi:https://doi.org/10.1016/j.proci.2022.08.084(2022).

428 [28] U. Pfahl, K. Fieweger, G. Adomeit, Self-ignition of diesel-relevant hydrocarbon-air mixtures under 429 engine conditions, Symposium (International) on Combustion 26 (1996) 781-789.

430 [29] H.K. Ciezki, G. Adomeit, Shock-tube investigation of self-ignition of n-heptane-air mixtures under 431 engine relevant conditions, Combust. Flame 93 (1993) 421-433.

432 [30] O. Herbinet, B. Husson, Z. Serinyel, M. Cord, V. Warth, R. Fournet, P.-A. Glaude, B. Sirjean, F.

433 Battin-Leclerc, Z. Wang, M. Xie, Z. Cheng, F. Qi, Experimental and modeling investigation of the low- 434 temperature oxidation of n-heptane, Combust. Flame 159 (2012) 3455-3471.

435 [31] N. Liu, S. Mani Sarathy, C.K. Westbrook, F.N. Egolfopoulos, Ignition of non-premixed counterflow 436 flames of octane and decane isomers, Proc. Combust. Inst. 34 (2013) 903-910.

437 [32] S. Humer, A. Frassoldati, S. Granata, T. Faravelli, E. Ranzi, R. Seiser, K. Seshadri, Experimental and 438 kinetic modeling study of combustion of JP-8, its surrogates and reference components in laminar 439 nonpremixed flows, Proc. Combust. Inst. 31 (2007) 393-400.

440 [33] J.C.G. Andrae, P. Björnbom, R.F. Cracknell, G.T. Kalghatgi, Autoignition of toluene reference fuels 441 at high pressures modeled with detailed chemical kinetics, Combust. Flame 149 (2007) 2-24.

442 [34] A. Alfazazi, O.A. Kuti, N. Naser, S.H. Chung, S.M. Sarathy, Two-stage Lagrangian modeling of 443 ignition processes in ignition quality tester and constant volume combustion chambers, Fuel 185 444 (2016) 589-598.

445 [35] S.Y. Mohamed, L. Cai, F. Khaled, C. Banyon, Z. Wang, M.J. Al Rashidi, H. Pitsch, H.J. Curran, A.

446 Farooq, S.M. Sarathy, Modeling Ignition of a Heptane Isomer: Improved Thermodynamics, Reaction 447 Pathways, Kinetics, and Rate Rule Optimizations for 2-Methylhexane, J. Phys. Chem. 120 (2016) 2201- 448 2217.

449

Preprint not peer reviewed