Probing PAH Formation from Heptane Pyrolysis in a Single-Pulse Shock Tube

ABSTRACT

To improve our capabilities to model surrogate fuels, particularly in respect to polycyclic aromatic hydrocarbons (PAHs) and soot formation, the pyrolysis of n-heptane is studied in a single-pulse shock across a wide temperature range (900–1700 K) at 20 bar nominal pressure and 4 ms residence time. Three different initial fuel mole fractions are considered, 103, 502, and 2000 ppm of n-heptane in argon. Fuel and intermediate species, including aromatics up to three-ring structures, are measured using gas chromatography and mass spectrometry diagnostics. An ongoing detailed chemical kinetic model for PAH chemistry has been updated to successfully capture the fuel decomposition, the formation of small hydrocarbons, and the concentration of the main PAH products. Major reaction pathways to PAHs are highlighted as well as the role of important intermediate species.

KEYWORDS:

• Heptane
• Pyrolysis
• Single-pulse shock tube
• Kinetic modeling
• Polycyclic aromatic hydrocarbons (PAHs)

Introduction

n-Heptane is widely used as a single-component surrogate for diesel fuel and as a key reference fuel for gasoline (Curran et al. 1998). Thus, its pyrolytic and oxidation chemistry has been studied for decades. At high temperatures, it dissociates into small hydrocarbon products, mainly C₂, C₃, and C₄, which, in absence of oxygen or at fuel-rich conditions, may subsequently react to form larger aromatic species such as the polycyclic aromatic hydrocarbons (PAHs). In general, understanding PAH formation during combustion processes paves the way for the researchers to develop soot formation kinetic models, which can help them improve combustion efficiency and reduce emissions. Hence, a detailed investigation of n-heptane pyrolysis can reveal the fuel breakdown and subsequent molecular weight growth processes.

n-Heptane pyrolysis has long been pursued using numerous experimental configurations and methodologies that cover a wide range of pressure, temperature, and initial fuel concentration conditions. The list of literature investigations concerning n-heptane pyrolysis is reported in Table 1, and it includes only the studies performed in absence of oxygen as a detailed review on the oxidation chemistry of n-heptane is beyond the scope of the present work.

Table 1. Literature review on experimental investigations concerning n-heptane pyrolysis.

Literature	Facility	Conditions	Measured products
(Appleby, Avery, and Meerbott 1947)	flow reactor/multi-isothermal distillation for products determination	NC7H14, T =823–903 K, p = 1 and 8.7 atm, τ*= not mentioned	C1-C4 speciation
(Murata and Saito 1974, 1975)	Flow-tube reactor/gas chromatography	NC7H14 in N2, T =973 K, p = 1 atm, τ= 0.09 to 46.1 sec.	C1-C6 speciation
(Bajus et al. 1979)	Tubular flow reactor/gas chromatography	Steam to n-heptane 3:1, T =953–1033 K, p = 1 atm, τ= 0.14 sec	C1-C6 speciation
(Aribike and Susu 1988, 1988)	Annular reactor/gas chromatography	n-heptane in N2, T =933–1053 K, p = 1 atm, τ= 0.4–1.02 sec	C1-C6 speciation
(Pant and Kunzru 1996)	Annular tubular reactor/gas chromatography	n-heptane in N2, T =953–1023 K, p = 1 atm, τ= 0.08–0.58 sec	C1-C6 speciation
(Held, Marchese, and Dryer 1997)	Variable pressure flow reactor/gas chromatography	1430 ppm n-heptane in N2, T =1085 K, p = 3 atm, τ= 0.6 sec	C1-C6 speciation
(Davidson, Oehlschlaeger, and Hanson 2007)	Shock tube/laser absorption	100–500 ppm n-heptane in Ar, T =1100–1560 K, p = 1.6–2 atm	CH3 concentration time-histories
(Chakraborty and Kunzru 2009)	Tubular reactor/gas chromatography	n-heptane in Ar, T =793–953 K, p = 0.1–2.93 MPa τ= 0.2–14.8 sec	C1-C6 speciation
(Garner, Sivaramakrishnan, and Brzezinski 2009)	Shock tube/gas chromatography	100 ppm n-heptane in Ar, T =1000–1350 K, p = 25 and 50 atm, τ= 1–3 ms	C1-C5 speciation
(Zamostny et al. 2010)	micro-pyrolysis unit (Shimadzu Pyr-4A)/gas chromatography	2µl n-heptane in N2, T =1083 K, p = 400 KPa, τ= 0.2–0.4 sec	C1-C6 speciation
(Pilla, Davidson, and Hanson 2011)	Shock tube/CO2 laser for C2H4 measurement	300 ppm n-heptane in Ar, T =1350–1950 K, p = 1.3–3.3 atm	C2H4 concentration time-histories
(Yuan et al. 2011)	Flow reactor/tunable synchrotron vacuum ultraviolet photoionization mass spectrometry	2% n-heptane in Ar, T =780–1780 K, p = 400 Pa	C1-C6 speciation
(Pyun et al. 2013)	Shock tube/DFG laser for CH4 measurement and CO2 laser for C2H4 measurement	1% n-heptane in Ar, T =1200–1600 K, p = 1.5 atm, τ= 1.5 ms	CH4 and C2H4 concentration time-histories
(Yasunaga et al. 2017, 2018)	Shock tube/gas chromatography	1% n-heptane in Ar, T =1000–1500 K, p = 1–2.5 atm, τ= 1.5–2.2 ms	C1-C4 speciation
(Li et al. 2021)	Shock tube/gas chromatography	3% n-heptane in Ar, T =1200–2100 K, p = 2.2–2.8 atm τ= not mentioned	C1-C4 speciation

*τ is the reaction time in shock tube experiments and residence time in flow reactor experiments.

Among the reported investigations, only few articles provide detailed speciation measurements. The pioneering work by Appleby, Avery, and Meerbott (1947) employed a flow technique to explore the thermal breakdown of n-heptane between 823 and 903 K at pressures of 1 and 8.7 atm. Murata and Saito (1974, 1975) studied the decomposition of paraffins and 1-olefins, including n-heptane, at 973 K and 1 atm, and developed a molecular model to explain the product distribution. Both Bajus et al. (1979), Pant and Kunzru (1996), and Chakraborty and Kunzru (2009) examined the steam cracking of n-heptane and determined the overall breakdown rate, by monitoring the formation of stable small hydrocarbon products using gas chromatography. Aribike and Susu (1988, 1988) proposed a mechanistic model for n-heptane pyrolysis based on product distribution data acquired in a pulse reactor at atmospheric pressure and reaction durations of 0.40–1.02 s at 933–1053 K. Held and his coworkers (1997) developed a simplified high-temperature pyrolysis and oxidation model for n-heptane that is validated against a variety of datasets, including their flow reactor experiments. More recent investigations have employed shock tube techniques for kinetic studies. Various time-history measurements were performed at Stanford University focusing on CH₃ (Davidson, Oehlschlaeger, and Hanson 2007), C₂H₄ (Pilla, Davidson, and Hanson (2011) and Pyun et al. (2013)), and CH₄ (Pyun et al. 2013) products. The single-pulse technique coupled to gas chromatography was also employed by Garner, Sivaramakrishnan, and Brzezinski (2009) to investigate the pyrolysis of saturated and unsaturated C₇ hydrocarbons, including n-heptane, in the temperature range 1000–1350 K at pressures of 25 to 50 atm with reaction times of 1–3 ms using a high-pressure shock tube. Yasunaga et al. (2017, 2018) carried out a series of works to address the decomposition of 1% n-heptane in argon using a single-pulse shock tube at pressures of 1.0–2.5 atm and temperatures of 1000–1500 K for reaction durations ranging from 1.5 to 2.2 ms. Li et al. (2021) conducted n-heptane and iso-octane pyrolysis experiments in a shock tube across a temperature range of 1200–2100 K and pressure range of 2.2–2.8 atm. Finally, among other techniques, Yuan et al. (2011) reported speciation measurements in n-heptane pyrolysis at low pressure (400 Pa) and temperatures ranging from 780 to 1780 K using a flow reactor and synchrotron vacuum ultraviolet photoionization mass spectrometry.

According to the above review, past research has focused on the fuel decomposition reactivity and the formation of small hydrocarbons up to C₆ intermediates. No studies have been conducted to investigate the formation of PAHs from n-heptane pyrolysis, though this latter focus is highly required for the development of clean combustion technologies. For this reason, the present work aims to provide shock tube speciation datasets spanning from small hydrocarbons to three-ring aromatics from n-heptane pyrolysis experiments at conditions relevant to modern combustion devices (20 bar, 900–1770 K). Our most recent kinetic model (Hamadi et al. 2022) has been updated to simulate and interpret the newly obtained experimental data. Kinetic insights into the formation of aromatic molecules from n-heptane pyrolysis will be addressed based on the experimental findings and kinetic modeling analyses.

Shock tube pyrolysis experiments

Pyrolysis experiments with three different n-heptane concentrations are performed using the high-purity single-pulse shock tube facility with dilute mole fractions of 103 ppm, 502 ppm, and 2000 ppm in bulk argon. A detailed description of the shock tube apparatus is well documented in our previous publications (Sun et al. 2022), so only a brief description is given here. The single-pulse shock tube apparatus consists of the driven (length: 6 m; inner diameter: 78 mm) and the driver (length: 3.7 m; inner diameter: 120 mm) sections, separated by a double diaphragm. The shock tube is equipped with a dump tank (volume 150 L) located near the diaphragm section, making it operate in a single pulse fashion. The low-pressure section is heated to 90°C to avoid condensation (or absorption) of fuel and reaction products. It is also pumped down to below 10⁻⁵ mbar with a turbo-molecular pump before each experiment. To avoid carbon deposits, the inner surface of the shock tube is cleaned regularly.

Four pressure sensors (CHIMIE METAL A25L05B) are mounted on the side wall of the driven section. The adjacent sensors are 150 mm apart, and the last sensor is 82 mm from the end wall. The distance and time difference between two successive pressure sensors are used to calculate the average incident shock velocity, which is then used to determine T₅ and p₅ behind the reflected shock wave by solving the conservation equations (Hugoniot 1887, 1889; Rankine 1870). The calculated T₅ is subject to an uncertainty of ± 30 K due to the attenuation of the shockwave and the physical dimension of the pressure sensors (Hamadi et al. 2022). A PCB Piezotronics pressure sensor, shielded by a layer of room-temperature vulcanizing (RTV) silicone, is located on the end wall of the driven section to record the pressure history of each shock, from which the corresponding reaction time is determined (Tang and Brezinsky 2006). The reaction time for each experiment is defined as the time interval between the arrival of the incident shock wave and the time point at which the pressure drops to 80% p₅ as a result of the quenching rarefaction waves. For the current experimental configuration, the nominal reaction time is 4 ms.

Stable species are withdrawn from the shock tube via an air-actuated valve and analyzed by the analytical system, which consists of three parts: an Agilent GC series 7890B, a Thermo Trace GC Ultra, and a Thermo DSQ mass spectrometer. The Agilent GC, designed to measure PAHs up to four rings, is equipped with a flame ionization detector (FID) coupled to a DB-17 ms column for heavy species separation, and a thermal conductivity detector coupled to a Molsieve 5A column for monitoring the absence of air and the dilution level by helium. An external valve box, coupled to the Agilent GC, diminishes the heavy compounds loss during sample storage and injection by maintaining the temperature at 320°C. The Thermo Trace GC is also equipped with an FID detector, connected to a HP Plot Q column for measuring light species up to mono-aromatics. The gas samples were also injected into the DSQ mass spectrometer, but in this case the MS analyses did not provide any additional information as all the heavy PAH compounds detected had been previously identified in our sequential studies on aromatic fuels.

The quantification of the measured species is based on the calibrations of the FID responses performed before the experiments. Standard gas mixtures are used for the calibration of the light species representing the C₁-C₅ hydrocarbons except for diacetylene (C₄H₂) and triacetylene (C₆H₂), whose calibration factors are obtained from high-temperature acetylene (C₂H₂) pyrolysis experiments through carbon atom conservation. Similarly, liquid fuels are used to calibrate benzene, heptane, toluene, styrene, and phenylacetylene. For all of the above calibrations, mixtures are prepared in stainless steel vessel using the partial pressure method. For small PAHs up to three rings, the calibration procedure differs. Here, the calibrations use gas-phase mixtures prepared in a heated glass vessel (200°C) (Sun et al. 2021b) to minimize surface adsorption (Comandini, Malewicki, and Brezinsky 2012). Uncertainties in species mole fractions are 5% for directly calibrated small species and 10% to 15% for PAH species calibrated in the gas phase. These uncertainties derive from the reproducibility of the single measure when repeated several times and the uncertainties in the composition of the initial reference calibration mixtures which reflect in the accuracy of the calibration curves.

The experiments are conducted with helium (Air Liquide, purity>99.995%) as the driver gas and dilute mixtures of n-heptane (Sigma – Aldrich, purity≥99.5%) in argon bath gas (Air Liquide, purity>99.9999%). The experimental gas mixtures are prepared in a 136 L electropolished stainless-steel cylinder that is evacuated in advance below 10⁻⁵ mbar with a turbomolecular pump. Before performing the experiments, the prepared gas mixtures are left overnight to homogenize, and the actual pre-shock compositions are analyzed by GC. The Supplementary Material provides all experimental results, consisting of the post-shock conditions (T₅, p₅), the measured pressure profiles, the defined reaction time, and the mole fraction measurements for individual species.

Kinetic modeling

The current work is a continuation of our previous serial work to build a kinetic model that can accurately predict the pyrolytic PAH formation under high-pressure and temperature conditions (the last version being published in (Hamadi et al. 2022)). The n-heptane sub-mechanism of Zhang et al. (2016) is chosen as a substitute for the lumped heptane chemical mechanism. The n-heptane sub-mechanism involves unimolecular decomposition and H-abstraction reactions from the fuel and the corresponding fuel radicals. The reactions of cyclohexadiene with molecular hydrogen (CYC₆H₈+H) and subsequent fragmentation and the corresponding rate coefficients reported in (Wang, Villano, and Dean 2015) are incorporated into the current kinetic model. The rate coefficients of the unimolecular decomposition reaction of 1-pentene (C₅H₁₀-1) and 1-hexene (C₆H₁₂-1) are an update from the work of (Shao et al. 2020). Additionally, the rate coefficients of the H-abstraction reaction of ethylene (C₂H₄) are taken from the same LLNL PAH model (Shao et al. 2020).

Thermochemical data for most species are from the CRECK model (Pejpichestakul et al. 2019), as it is the basis of the current model (Sun et al. 2021c). For species introduced by reactions added or updated in this work, the thermochemical data are from the same publications. The mechanism file and thermochemical data for the current kinetic model are provided in the Supplementary Material.

All simulations in the present work are performed using the homogeneous reactor model of the COSILAB software. The nominal reaction time of 4.0 ms and constant pressure of 20 bar are used for the simulations. The assumption of constant pressure, typically used in simulations of speciation results from single-pulse shock tube experiments, has been well justified in previous publications (Han, Mehta, and Brezinsky 2019; Tang and Brezinsky 2006). However, it is suggested that certain reactions involving resonantly stabilized radicals may occur during the post-shock quenching period, resulting in an increase in the mole fractions of some products (Manion, Sheen, and Awan 2015; Mertens et al. 2018). In addition, depending on the characteristics of the experimental facility (small-bore shock tubes) or conditions (higher fuel concentrations) (Ferris, Davidson, and Hanson 2018; Mehta, Wang, and Brezinsky 2022), the constant-pressure assumption may be erroneous. Simulations with measured pressure profiles are performed over a longer time period (10.0 ms) for all n-heptane sets. The time-dependent concentrations of the major products and important radicals are shown in Figure S1 in the Supplementary Material. The plateaus in the concentration profiles of the main products are reached before the quenching process which indicates that their formation is already completed at the specified reaction time. The reaction systems currently studied do not produce large amounts of highly stabilized radicals such as propargyl and benzyl, so their concentrations are too low to significantly affect the amount of final products as shown in Figures S2-S4.

Similar to the vast literature on n-heptane experimental chemistry, several detailed kinetic models have been developed and optimized in the past decades based on the available data. Some of the most recent models, such as NUIG (Zhang et al. 2016), JetsurFv2 (Wang et al. 2010), CRECK (Pejpichestakul et al. 2019), and LLNL PAH (Shao et al. 2020), were also used to simulate the current experiments for comparison purpose (Figures S5–S7). The current model is compared to NUIG and JetsurFv2 for small hydrocarbon predictions, as well as the CRECK and LLNL PAH models for both small hydrocarbon and aromatic species predictions. In comparison to the other tested models, the LLNL PAH model is in good agreement with our experiments except for over-predictions for benzene, phenylacetylene, and naphthalene.

The current kinetic model for heptane pyrolysis is tested against pertinent speciation data reported in literature to assess its capability to predict results under different conditions (Garner, Sivaramakrishnan, and Brzezinski 2009; Held, Marchese, and Dryer 1997; Yasunaga et al. 2017, 2018; Yuan et al. 2011). There are a variety of experimental setups and, as a result, simulation methodologies involved. For the shock tube experiments (Garner, Sivaramakrishnan, and Brzezinski 2009; Yasunaga et al. 2017, 2018), each data point is simulated with the “homogeneous reactor” by entering the measured T₅, p₅ and reaction time. The “plug-flow” model is employed to simulate heptane pyrolysis speciation in a flow tube (Yuan et al. 2011). As for the experiments performed with the variable-pressure flow reactor (Held, Marchese, and Dryer 1997), an isobaric and adiabatic zero-dimensional reactor is used to simulate the species mole fraction time histories by providing the initial temperature, pressure, and chemical compositions. Figures S8–S12 in the Supplementary Material show the comparisons between the literature data and the simulations with the current model.

Results and discussions

In this section, experimental and simulated species concentrations as a function of post-shock temperature T₅ will be compared to validate the prediction capabilities of the present kinetic model. Only the data from 502 and 2000 ppm of n-heptane are presented in the discussion, while the ones for 103 ppm are given in Figure S2 in the Supplementary Material. In general, the model can mimic the mole fraction profiles of observed pyrolysis species and capture their decomposition and formation temperature windows within the experimental errors. To gain insight into the chemistry of heptane degradation and aromatics growth, modeling analytical approaches such as rate of production (ROP) analyses and sensitivity analyses are applied.

Fuel decomposition and the formation of small hydrocarbons

Figure 1 presents the experimental and simulated mole fraction profiles of n-heptane and C₁–C₆ acyclic species in the three separate sets. The kinetic model satisfactorily predicts the measured fuel conversion profiles as well as the formation of small intermediates in the individual studied cases. Olefins begin to form during the early stage of the fuel consumption. To determine the relationship between various alkene production and fuel consumption, Figure 2 displays initial fuel consumption pathways based on integrated rate-of-production (ROP) analyses over 4 ms in the three reaction systems. The temperatures chosen to perform these ROP calculations are 1160 K for 103 and 502 ppm heptane pyrolysis, and 1130 K for 2000 ppm heptane pyrolysis, where roughly half of the fuel is consumed in each case. The reaction scheme shows that n-heptane is largely consumed via the H-abstraction reactions, which produce 1-, 2-, 3-, and 4-heptyl radicals, and via the C-C bond scission reaction, which produces n-propyl (NC₃H₇) and 1-butyl (PC₄H₉) radicals. Other C-C bond fission reactions that produce methyl and 1-hexyl radicals, as well as ethyl and 1-pentyl radicals, contribute only slightly to the thermal decomposition of n-heptane. Following the decomposition of the major heptane radicals, the various alkene products are produced. Propylene (C₃H₆, Figure 1f) and 1-pentene (C₅H₁₀-1, Figure 1q) are directly formed through the unimolecular decomposition of 2- and 4-heptyl radicals, respectively (see Figure 2). The 3-heptyl radical allows a modest but significant fraction of methyl radical (CH₃) and l-hexene (C₆H₁₂-1, Figure 1t) to be formed in favor of NC₃H₇ and l-butene (C₄H₈-1, Figure 1j). The allyl (C₃H₅-A)+CH₃ recombination reaction is the second important C₄H₈-1 formation channel at 1160 K (around 30% according to the ROP analyses). C₃H₅-A is produced mostly from the unimolecular decomposition of C₅H₁₀-1 (C₅H₁₀-1= ethyl (C₂H₅) + C₃H₅-A) and C₆H₁₂-1 (C₆H₁₂-1= NC₃H₇+ C₃H₅-A), and partly from the C₃H₆ through H-abstraction reactions with methyl radical and hydrogen atom. The other butene isomer, 2-butene (C₄H₈-2, Figure 1k), is produced mainly through isomerization and H-assisted isomerization of C₄H₈-1, with increasing contribution from C₅H₁₀-1 + CH₃ reaction along with the initial fuel concentration (from 10% with 103 ppm to 30% with 2000 ppm initial n-heptane mole fraction at 1170 K). Ethylene (C₂H₄, Figure 1c) is formed via numerous reaction pathways as shown in Figure 2, including 1-heptyl decomposition, thermal decomposition of the 1-pentyl (C₅H₁₁-1), 1-butyl, l-propyl and ethyl radicals formed in 1-,2-,3-and 4-heptyl decomposition, respectively, and further H-abstraction reactions of the other olefinic intermediates, namely C₃H₆ and C₄H₈-1. The dominant C₂H₄ consumption route is the unimolecular fragmentation into vinyl radicals (C₂H₃) and H-atoms. Aside from alkenes, the recombination of methyl and ethyl drives propane formation (C₃H₈, Figure 1i), which starts at the early phases of heptane decay. Pentadiene (LC₅H₈, Figure 1s) is the result of the unimolecular decomposition of pentenyl (NC₅H₉-3) and heptenyl radicals (NC₇H₁₃), which are formed through dehydrogenation of C₅H₁₀-1 (Figure 1q) and 1-heptene (NC₇H₁₄), respectively. NC₇H₁₄ (not measured in the current study) is produced entirely by the unimolecular breakdown of C₇H₁₅-1 (Figure 2).

Figure 1. Measured (symbols) and simulated (lines) mole fraction profiles of fuel and small intermediates as a function of post-shock temperature T₅ in heptane pyrolysis.

Figure 2. Fuel consumption pathways at 1160 K in 103 and 502 ppm n-heptane pyrolysis and at 1130 K in 2000 ppm n-heptane pyrolysis, 20 bar. The percentage numbers (103 ppm heptane: black normal; 502 ppm heptane: blue italic; 2000 ppm heptane: green underlined) are the contributions by the corresponding reactions to the consumption of the species on the source side. The reaction path analyses are based on the integrated ROP analyses over 4 ms.

The remaining small hydrocarbons are formed through subsequent reactions. The remarkable mole fractions of methane (CH₄) and ethane (C₂H₆) in n-heptane pyrolysis (Figure 1 b, e, respectively) arise from the abundant CH₃ production via NC₃H₇ = CH₃ + C₂H₄ reaction at low temperatures (>1200 K), H + C₃H₆ = CH₃ + C₂H₄ reaction at moderate temperatures (1200–1350 K), and unimolecular decomposition of C₂H₆ at high temperatures (>1400 K). Acetylene (C₂H₂, Figure 1d) is predominantly formed by the decomposition of C₂H₃ and marginally by the reaction of H atoms with allene (C₃H₄-A, Figure 1g) and propyne (C₃H₄-P, Figure 1h). The C₃H₄ isomers, C₃H₄-P and C₃H₄-A, are formed at temperatures where the fuel is nearly depleted. C₃H₆ consumption leads to C₃H₄-A, which isomerizes to C₃H₄-P. Four C₄H₆ isomers are observed: 1-butyne (1-C₄H₆, Figure 1n), 2-butyne (2-C₄H₆, Figure 1o), 1,2-butadiene (1,2-C₄H₆, Figure 1m) and 1,3-butadiene (1,3-C₄H₆, Figure 1l). Quantitative mole fraction profiles for C₄H₆ isomers are rarely reported in previous investigations on n-heptane pyrolysis. The current model captures 1,3-C₄H₆, 1,2-C₄H₆, and 1-C₄H₆ profiles well; nevertheless, discrepancies exist between the experimental and simulated 2-C₄H₆ speciation profiles with an overprediction of the experimental profiles. The peak mole fraction of 1,3-C₄H₆ is fifty times greater than that of the other three isomers. ROP analyses are done across the temperature formation windows to better explain the higher peak mole fractions of 1,3-butadiene. The formation of the dominant isomer 1,3-C₄H₆ largely relies on the 1-C₄H₈ consumption through the stepwise dehydrogenation via 1-buten-3-yl (C₄H₇1–3) radical as an intermediate. As we observed above, 1-butene is one of the main alkenes from the direct decomposition of the heptyl radicals, thus its large concentrations (Figure 1j) drive the formation of 1,3-C₄H₆. The other C₄H₆ isomers are mainly produced by bimolecular recombination reactions. 1,2-C₄H₆ is mainly formed through CH₃ + propargyl (C₃H₃) reaction, where C₃H₃ is produced from the decomposition of C₃H₄-P and C₃H₄-A. Reactions between CH₃ and C₃ molecules/radicals (C₃H₄-A and C₃H₃) govern the formation of 1-C₄H₆. The production of 2-C₄H₆ mostly depends on the isomerization reactions of 1,2-C₄H₆ and just moderately by 1,3-C₄H₆ isomerization reactions. Regarding C₅ species, cyclopentadiene (C₅H₆, Figure 1r) is mainly produced through the dehydrogenation of cyclopentene (CYC₅H₈) and slightly through the reaction C₃H₅-A+C₄H₆=C₂H₅+C₅H₆. The 1-penten-4-yl radical (C₅H₉14) formed by the C₃H₅-A+C₂H₄ recombination reaction can undergo cyclization reaction to produce cyclopentenyl radical (cyC₅H₉), which is the main source of CYC₅H₈ through its β-C – H scission. Though not shown in Figure 1, CYC₅H₈ mole fractions are measured in our investigation and are available in the Supplementary Material.

Aromatic growth

Figure 3 depicts the experimental and simulated mole fraction profiles of several common mono-aromatic hydrocarbons (MAHs) and bicyclic and tricyclic PAHs. For all the aromatics, the current model accurately captures the experimental trends regarding the shapes and the peak concentrations. The major aromatic formation pathways are identified through ROP analyses and will be addressed in this section. The reactions that account for the formation of these aromatics are similar in all the reaction systems; however, the product concentrations increase with the fuel mole fraction due to the increased amount of carbon in the reaction system.

Figure 3. Measured (symbols) and simulated (lines) mole fraction profiles of MAHs and PAHs as a function of post-shock temperature T₅ in n-heptane pyrolysis.

One of the key subjects covered in kinetic studies of acyclic fuel consumption under fuel-rich or pyrolytic conditions is the production of the first aromatic ring. The MAH species detected in n-heptane pyrolysis include benzene (C₆H₆), toluene (C₇H₈), phenylacetylene (C₆H₅C₂H), styrene (C₆H₅C₂H₃), and ethylbenzene (C₆H₅C₂H₅). C₆H₆ (Figure 3a) is the most abundant MAH species formed during n-heptane pyrolysis. Integrated ROP analyses at 1250 K and 1450 K are used to identify the reaction pathways leading to C₆H₆ production as shown in Figure 4. Three sources are responsible for C₆H₆ production in heptane pyrolysis: the dehydrogenation of 1,4-cyclohexadiene (CYC₆H₈) that is mainly produced from C₂H₃+C₄H₆ recombination followed by cyclization, H-abstraction and β-scission (Wang, Villano, and Dean 2015), the reaction sequence 2C₃H₃ → fulvene→ C₅H₅CH₂-1/C₅H₅CH₂-2 →C₆H₆ (Miller and Klippenstein 2003), and the C₃H₃ self-recombination. The relative importance of the above-mentioned channels varies with temperature. The first two pathways control C₆H₆ formation in the low-temperature range of our study, whereas the latter two at higher temperatures. It is worth mentioning how the pathway through 1,4-cyclohexadiene is also strongly dependent on the initial fuel concentration. Besides, the C₃H₅-A+C₃H₃ recombination and the fulvene isomerization have minor contributions to C₆H₆ throughout the temperature range (see Figure 4).

Figure 4. The reaction pathways leading to benzene formation at (a) T₅ = 1250 K, (b) T₅ = 1450 K in the pyrolysis of n-heptane. The percentage numbers (502 ppm heptane: blueitalic; 2000 ppm heptane: green underlined) represent the contributions to benzene formation by the corresponding reactions.

The formation of C₇H₈ (Figure 3b) largely depends on the reactions of C₃H₃ with C₄H₆ and but-2-yn-1-yl radical (C₄H₅-2), respectively. The formation of C₆H₅C₂H (Figure 3c) and C₆H₅C₂H₃ (Figure 3d) mostly relies on the reactions of phenyl with C₂H₂/C₂H and C₂H₄, respectively. C₅H₅+C₃H₃ reaction and C₄H₆ self-recombination also slightly contribute to C₆H₅C₂H₃ at low temperatures. C₆H₅C₂H₅ is mainly formed through benzyl (C₇H₇)+CH₃. C₄H₆ self-recombination also has a non-negligible contribution to C₆H₅C₂H₅.

The major goal of this work is to unravel the mechanisms responsible for PAH formation during n-heptane pyrolysis. The smallest PAH detected in this study is indene (C₉H₈, Figure 3e). C₉H₈ mainly originates from the bimolecular reaction C₉H₇+H=C₉H₈ at low temperatures (T₅<1400 K), where C₉H₇ is exclusively formed from C₄H₂+C₅H₅ reaction. As the temperature increases, the isomerization of 1-phenyl-propyne (C₆H₅C₃H₃P_1) and phenyl-allene (C₆H₅C₃H₃A) and C₇H₇+C₂H₂ reaction become the dominant C₉H₈ formation channels. The corresponding rate parameters have been validated in our previous investigations on the pyrolysis of propylene and propyne (Sun et al. 2021a), and C₂ and C₃ addition to toluene and benzene (Hamadi et al. 2022; Sun et al. 2021a, 2022). C₆H₅C₃H₃P_1 is largely formed through the molecule+radical reaction C₆H₅C₂H+CH₃, whereas the radical+radical reaction C₆H₅+C₃H₃ is the major source of C₆H₅C₃H₃A.

C₉H₈ consumption mostly yields indenyl radical (C₉H₇) through H-abstraction and unimolecular decomposition. C₉H₇ participates in the formation of the other PAHs, including naphthalene (C₁₀H₈) and acenaphthylene (C₁₂H₈), which were found to be the most prevalent in the current study (Figure 3f,g). The formation mechanism of C₁₀H₈ includes several possible reaction pathways that are displayed in Figure 5 based on ROP analysis done at 1525 K. As can be seen in Figure 5, benzofulvene (C₉H₆CH₂) is a vital precursor of naphthalene. According to the ROP-analyzed results, the formation of C₉H₆CH₂ relies on three major channels: i) the unimolecular decomposition of methyl-indene radical (C₉H₆CH₃-1) following the production of 1-methyl-indene (C₉H₇CH₃-1) through C₉H₇+CH₃ recombination; ii) the recombination between fulvenallyl (C₇H₅) and C₃H₃. (C₇H₅ mainly comes from C₇H₇ decomposition and the C₃H₃+C₄H₂ recombination); iii) the bimolecular reaction between o-benzyne (o-C₆H₄) and vinylacetylene (C₄H₄). (o-C₆H₄ is mostly derived from C₇H₇ breakdown). Besides the isomerization of benzofulvene, other potential sources of naphthalene include the consumption of methyl-indene radicals (C₉H₇CH₂ and C₉H₆CH₃-1), C₇H₅+C₃H₃ reaction, and the hydrogen abstraction acetylene addition (HACA) route through phenylacetylene radical (C₆H₄C₂H) leading to naphthyl (C₁₀H₇) radicals which will recombine with H atom (not shown in Figure 5).

Figure 5. The reaction pathways leading to naphthalene formation at T₅ = 1525 K in the pyrolysis of heptane. The percentage numbers (502 ppm heptane: blue italic; 2000 ppm heptane: green underlined) represent the contributions to naphthalene formation by the corresponding reactions.

1-methyl-indene (C₉H₇CH₃-1), 1-methyl naphthalene (C₁₀H₇CH₃-1), acenaphthylene (C₁₂H₈), 1-ethynyl-naphthalene (C₁₀H₇C₂H_1) and fluorene (C₁₃H₁₀) are among the major C₁₀–C₁₃ PAHs identified in this study. As previously stated, C₉H₇CH₃-1 (Figure 3k) is a key precursor for the naphthalene-benzofulvene isomer pair. The C₉H₇+CH₃ recombination produces the majority of C₉H₇CH₃-1. C₁₂H₈ (Figure 3g) is one of the most abundant PAHs in n-heptane pyrolysis. The C₉H₇+C₃H₃ recombination and the subsequent dehydrogenation and ring-rearrangement processes (Jin et al. 2019) prevails C₁₂H₈ formation throughout the temperature range. The formation of the remaining PAHs involves the participation of naphthyl radical (C₁₀H₇-1): C₁₀H₇-1+CH₃ recombination is the primary source of C₁₀H₇CH₃-1 (Figure 3l); the HACA route C₁₀H₇-1+C₂H₂ = C₁₀H₇C₂H_1+H controls the formation of C₁₀H₇C₂H_1 (Figure 3m) and it also partly accounts for C₁₂H₈ formation (27% at 1525 K); C₁₀H₇-1+C₃H₄-P is a minor route leading to C₁₃H₁₀ (Figure 3n). In this case, the C₉H₇+vinylacetylene (C₄H₄) reaction governs the C₁₃H₁₀ formation throughout the temperature range.

Four different C₁₄H₁₀ species are identified and quantified in 2000 ppm heptane pyrolysis, including the dominant phenanthrene (PC₁₄H₁₀), and its isomers, diphenylacetylene (C₆H₅CCC₆H₅), 9-methylene-fluorene (C₁₃H₈CH₂), and anthracene (AC₁₄H₁₀). Reaction pathways leading to C₁₄H₁₀ isomers formation at 1500 K are presented in Figure 6. C₆H₅CCC₆H₅ (Figure 3o) and C₁₃H₈CH₂ (not shown in Figure 3) are the products of the C₆H₅C₂H+C₆H₅ addition-elimination reaction (Sun et al. 2020). AC₁₄H₁₀ (Figure 3p) is totally formed from C₇H₅ self-recombination. PC₁₄H₁₀ (Figure 3h) mainly comes from the C₇H₇ self-recombination and AC₁₄H₁₀ isomerization. Other reaction channels which contribute to PC₁₄H₁₀ formation include: the C₇H₅ self-recombination, H-assisted isomerization of C₆H₅CCC₆H₅ and C₁₃H₈CH₂, and the reaction of C₉H₇ with cyclopentadienyl radical (C₅H₅), which results from the consumption of C₅H₆ and the addition of C₃H₃ to C₂H₂.

Figure 6. The reaction pathways leading to the formation of C₁₄H₁₀ isomers at T₅ = 1500 K in the pyrolysis of n-heptane. The percentage numbers (502 ppm n-heptane: blue italic; 2000 ppm n-heptane: green underlined) represent the contributions to C₁₄H₁₀ isomers formation by the corresponding reactions. The dashed arrows represent multi-step reaction.

Conclusion

This research highlights PAH products from n-heptane pyrolysis as a significant new contribution to the literature on the chemistry of this important surrogate component. The experiments were carried out in a single-pulse shock tube coupled to gas chromatographic techniques at a nominal pressure of 20 bar throughout a temperature range of 900–1700 K. Updates to our ongoing PAH kinetic model demonstrate satisfactory predictive performances for the speciation measurements obtained in this work as well as those reported in literature studies on n-heptane pyrolysis. Heptane mainly decomposes through C-C fission and H-abstraction reactions leading to the formation of heptyl, n-butyl and n-propyl radicals. These Alkyl radicals further undergo beta-fission reactions leading to the formation of alkenes, namely, ethylene, propylene, 1-butene, 1-pentene and 1-hexene. The consumption of alkenes results in key intermediates such as methyl, acetylene, ethyl, propargyl, and 1,3-butadiene, which play a significant role in the formation and growth of aromatics. In particular, the formation of benzene relies on C₂H₃+C₄H₆ reaction through the intermediate cyclohexene, the H-addition and isomerization of fulvene, and C₃+C₃ reactions, more specifically, propargyl self-recombination and propargyl+allene reaction. Toluene is the product of the addition reaction of C₃H₃ to 1,3-butadiene. Indene formation mainly depends on C₄H₂+C₅H₅ reaction at low temperatures and on the consumption of C₆H₅C₃H₃P_1 and C₆H₅C₃H₃A, and the C₇H₇+C₂H₂ channel at high temperatures (>1400 K). Indenyl’s subsequent interactions with methyl, propargyl and vinylacetylene lead to the production of naphthalene, acenaphthylene and fluorene, respectively. Naphthyl radicals further participate in the formation of other bigger PAHs including 1-methyl naphthalene and 1-ethynyl naphthalene. Benzyl self-recombination and anthracene isomerization were analyzed to be the greatest contributors to phenanthrene formation. A small amount of phenanthrene is produced through C₆H₅C₂H+C₆H₅ and C₉H₇+C₅H₅ reaction channels. Finally, the findings will help researchers better understand how PAHs form during the combustion of surrogate fuels.

Acknowledgements

This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement n° 756785).

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/00102202.2023.2182199

Additional information

Funding

The work was supported by the H2020 European Research Council [grant agreement n° 756785].