Shock Tube Measurement for the Dissociation Rate Constant of

Subscriber access provided by Northern Illinois University

Article

Shock Tube Measurement for the Dissociation Rate Constant of Acetaldehyde using Sensitive CO Diagnostics Shengkai Wang, David F. Davidson, and Ronald K. Hanson J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.6b03647 • Publication Date (Web): 15 Aug 2016 Downloaded from http://pubs.acs.org on August 23, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry A is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Shock Tube Measurement for the Dissociation Rate Constant of Acetaldehyde using Sensitive CO Diagnostics

Shengkai Wang, David F. Davidson*, Ronald K. Hanson

High Temperature Gasdynamics Laboratory, Mechanical Engineering, Stanford University, CA 94305, USA

*: Corresponding author. Email: [email protected]; Fax: +1(650)-723-1748

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 40

Abstract

The rate constant of acetaldehyde thermal dissociation, CH3CHO = CH3 + HCO, was measured behind reflected shock waves at temperatures of 1273 - 1618 K and pressures near 1.6 atm and 0.34 atm. The current measurement utilized sensitive CO diagnostics to track the dissociation of CH3CHO via oxygen atom balance, and inferred the title rate constant (k1) from CO timehistories obtained in pyrolysis experiments of 1000ppm and 50ppm CH3CHO / Ar mixtures. By using dilute test mixtures, the current study successfully suppressed the interferences from secondary reactions, and directly determined the title rate constant as k1(1.6 atm) = 1.1 x 1014 exp(-36700 K/ T) s-1 over 1273-1618 K, and k1(0.34 atm) = 5.5 x 1012 exp(-32900 K/ T) s-1 over 1377-1571 K, with 2σ uncertainties of approximately +/- 30% for both expressions. Example simulations of existing reaction mechanisms updated with the current values of k1 demonstrated substantial improvements with regards to the acetaldehyde pyrolysis chemistry.


Page 3 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction As modern combustion research strives to face future fossil fuel shortages and increasingly stringent regulations on engine emissions, bio-derived oxygenated fuels have received increasing attention as alternative energy sources or clean fuel additives to suppress hazardous emissions.1 However, despite its beneficial impacts on soot and carbon monoxide emissions,2 the combustion of biofuels, especially bio-alcohols, is generally linked to increased formation of aldehydes,3-6 which if not removed properly can lead to environmental pollution and potential threats to public health.7 Thus an essential topic of biofuel combustion research involves accurate monitoring of aldehydes concentrations and better understanding of aldehydes chemistry, especially their removal reactions via either radical abstraction or direct thermal dissociation.8,9

Recently, our research group at Stanford University developed a set of accurate and sensitive high temperature laser absorption diagnostics for two major aldehyde species, namely formaldehyde (CH2O) and acetaldehyde (CH3CHO), and successfully demonstrated them in a shock tube kinetics study.10 In these demonstration experiments we measured CH3CHO timehistories during the pyrolysis of 1% CH3CHO/Ar mixtures, which were later utilized by combustion kinetics modelers to validate and optimize their reaction mechanisms.11 A comparison of simulations using various reactions mechanisms12-16 and our previous measurement results (see Fig. 1) revealed large discrepancies between these simulations as well as deviations of these simulations from the experimental data. These differences suggested the need for further investigations of the acetaldehyde pyrolysis system that we report here.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 40

Figure 1. Comparison of measured and simulated CH3CHO profiles during 1% CH3CHO/Ar pyrolysis. Dashed lines: simulations with USC Mech II,12 which are almost identical to simulations with JetSurF 2.0;13 short dashes: simulations with ARAMCO;14 dash dots: simulations with the Cong and Dagaut mechanism;15 dotted lines: simulations with the Chatelain et al. mechanism.16 Note that both USC Mech II/ JetSurF 2.0 and ARAMCO underpredict the CH3CHO decay rates, while the other mechanisms overpredict the decay rates. Figure adapted from Mével et al.11

A sensitivity analysis on our previous results (Fig. 2) suggested that the following four reactions mainly governed the CH3CHO removal rate, among which the unimolecular decomposition of aldehydes (R1) was recognized to be one of the dominant reactions: CH3CHO (+M) = CH3 + HCO (+M) (R1) CH3 + CH3CHO = CH4 + CH3CO

(R2)

CH3 + CH3 (+M) = C2H6 (+M)

(R3)

H + CH3CHO = H2 + CH3CO

(R4) 4 ACS Paragon Plus Environment

Page 5 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. CH3CHO sensitivity, ∂ln 𝜒𝜒𝐶𝐶𝐶𝐶3𝐶𝐶𝐶𝐶𝐶𝐶 (t)/ ∂lnk i , of the time-histories shown in Fig. 1, highlighting the reactions with a maximum sensitivity larger than 0.05. Simulated using USC Mech II. Figure adapted from Wang et al.10

Due to the heavy cross-interferences of the above reactions (R1-R4) with each other in a highconcentration CH3CHO pyrolysis system, in our previous experiment we were not able to directly extract their rate constants. Therefore new experiments exploiting more dilute mixtures were designed and conducted to isolate these reactions and accurately determine their rate constants, with the current study focusing on the reaction rate constant of R1 (k1).

Experimental Method Experiments were conducted in two similar high-purity shock tube facilities at Stanford University. One is a 14.13-cm inner diameter shock tube with a driver section of 3.3 m and a driven section of 8.5 m, which allowed for steady test time of about 2 ms under the typical shock conditions of the current study. The other is a 15.24-cm diameter shock tube with a driver section 5 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 40

of 3.7 m and a driven section of 10 m, which allowed for steady test time of about 1.5 ms. Test gas mixtures were manometrically prepared in stainless steel mixing tanks that were preheated to 50C to prevent wall condensation. Research grade acetaldehyde (>99.5% purity, supplied by Sigma-Aldrich) was purified through a freeze-pump-thaw procedure to remove dissolved air, and then vaporized and mixed with high-purity argon (5.0 grade, supplied by Praxair). The mixtures were prepared following a double- or triple-dilution procedure, and were stir-mixed with magnetic stirrers for at least 2 hours during each dilution. To remove possible residual impurities, the shock tubes and the mixing assemblies were routinely pumped to 6 µtorr between shock experiments. Further details about the Stanford shock tube facilities have been documented in our previous studies.10,17-20

Time-resolved species time-histories were measured behind reflected shock waves using laser absorption diagnostics accessed through a pair of optical ports on each shock tube at a location of 2 cm away from the endwall. Instead of monitoring the decay of CH3CHO, the current study utilized more sensitive diagnostics to monitor the stable product formed during its pyrolysis, CO, which correlated with the removal of CH3CHO by a 1:1 ratio (concentrations of other oxygenated products, such as CH3CO and HCO, are negligibly low due to their very short lifetimes). On the 14.13-cm diameter shock tube, a single-pass (SP) CO absorption diagnostic employing a quantum cascade laser near 4.6 µm was set up, which has a 2σ detection limit of about 20ppm CO. Set up on the 15.24-cm diameter shock tube was a more advanced CO diagnostic capable of achieving sub-ppm level detection limit. This improvement in detection limit was achieved by using a cavity-enhanced absorption spectroscopy (CEAS) system directly coupled to the shock tube, which significantly increased the effective absorption path length and 6 ACS Paragon Plus Environment

Page 7 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


hence the detection sensitivity by up to about two orders of magnitude.21,22 These highly sensitive CO diagnostics allowed us to lower the CH3CHO concentration to 1000ppm and 50ppm while still maintaining excellent signal-to-noise ratios (Fig. 3), and hence significantly suppressed the influences of secondary reactions and temperature changes (to less than 8K during the pyrolysis of 1000ppm CH3CHO and less than 1K for 50ppm CH3CHO), leading to accurate determination of k1. Further details about the CO diagnostics and their example applications can be found in our previous studies.18,21,22

Figure 3. Example CO time-histories measured during the pyrolysis of CH3CHO in argon. (a): 1000ppm CH3CHO / Ar, 1447 K, 1.60 atm; (b) ~50ppm CH3CHO / Ar, 1494 K, 1.49 atm

Rate Constant Determination As shown in Fig. 3, the rate constant k1 is inferred by best-fitting the simulated CO time-history from a comprehensive reaction mechanism, USC Mech II,12 to our experiment result. Also shown in the figure are CO time-histories with k1 perturbed by +/-30%, which highlight the sensitivity of the current rate constant measurement. This is further illustrated by a formal CO 7 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 40

sensitivity analysis displayed in Fig. 4, where the CO sensitivity is defined as the ratio of the percentage change in the CO mole fraction (χCO) to the percentage change in the reaction rate constant under investigation (ki): 𝑆𝑆𝑖𝑖 = (𝜕𝜕𝜒𝜒𝐶𝐶𝐶𝐶 /𝜒𝜒𝐶𝐶𝐶𝐶 )/(𝜕𝜕𝑘𝑘𝑖𝑖 /𝑘𝑘𝑖𝑖 ). As the figure clearly shows, the CO sensitivity is dominated by the acetaldehyde thermal dissociation reaction, R1, whereas sensitivities of other reactions, especially the radical abstraction reactions, are relatively low.

Figure 4. CO sensitivity, ∂ln 𝜒𝜒𝐶𝐶𝐶𝐶 (t)/ ∂lnk i , of the time-histories shown in Fig. 3. (a): 1000ppm

CH3CHO / Ar, 1447 K, 1.60 atm, highlighting the reactions with a maximum sensitivity larger than 0.05; (b): 50ppm CH3CHO / Ar, 1494 K, 1.49 atm, highlighting the reactions with a maximum sensitivity larger than 0.02. All sensitivities are calculated using USC Mech II with

updated values of k1 from the current study. Note that the interferences from R2-R4 have been substantially suppressed as compared to Fig. 2

A detailed uncertainty analysis, which includes secondary reactions and other uncertainty sources, is shown in Fig. 5. Rate constants of the secondary reactions, e.g. k2, k3 and k4, are assigned an uncertainty factor of 2, in accordance with recommendations by Baulch et al.23 The 8 ACS Paragon Plus Environment

Page 9 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


resulting uncertainties in k1 are determined from a brute-force approach, i.e. by adjusting the individual values of k2, k3 and k4 to their uncertainty limits and calculating the subsequent changes in the best-fit inference of k1. Uncertainty in the reflected shock temperature (T5), translates to about +/-18% and +/-12% uncertainty in k1 in the single-pass and the CEAS measurements, respectively. Fitting uncertainty, determined by varying k1 to fit the upper and lower envelopes of the measured CO profile, is about +/-10% for the single-pass absorption experiments and about +/-15% for the CEAS experiments. Uncertainties in the CO absorption cross-section (+/-5%) and in the CEAS gain factor (+/-5%) are quoted from recent studies,21,24 and the uncertainty in the CH3CHO concentration of the manometric mixture is estimated to be +/-5%, a value consistent with our previous work.10,17 The overall (root-mean-squared) 2-σ uncertainty in k1 is calculated to be +28% / -32% for the single-pass measurement at 1447 K and +24% / -26% for the CEAS measurement at 1494 K.

Figure 5. Uncertainty analysis for the measurement of k1. (a): 1000ppm CH3CHO / Ar, 1447 K, 1.60 atm; (b) 50ppm CH3CHO / Ar, 1494 K, 1.49 atm



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 40

Results and Discussions A total of 22 reflected shock experiments were conducted at temperatures between 1273 K and 1618 K and pressures near 1.6 atm and 0.34 atm (see Table 1), which resulted in an Arrhenius expression for the rate constant of R1 as k1(1.6 atm) = 1.1 x 1014 exp(-36700 K/ T) s-1 over 12731618 K, and k1(0.34 atm) = 5.5 x 1012 exp(-32900 K/ T) s-1 over 1377-1571 K, with 2σ uncertainties of approximately +/- 30% for both expressions. CO time-histories of all shock experiments are available in the supporting material attached with this manuscript. Previously, experimental studies of k1 using other approaches have been reported in the literature, and we now compare our current results with these earlier measurements in Fig. 6. Gupte et al.25 examined the pyrolytic decomposition of acetaldehyde and measured k1 through shock tube laser-schlieren experiments at temperatures of 1550 - 2400 K and pressures of 0.05 – 0.66 atm. As shown in Fig. 6a, the data from Gupte et al. agree in trend with the current results. They appear to have an activation energy that is similar to the current study, with a small offset in the magnitude of k1 probably resulting from the pressure dependence (both studies are conducted in the pressure fall-off region). Bentz et al.26 investigated k1 at temperatures ranging from 1250 to 1650 K and pressures of about 1.3, 2.9 and 4.5 atm using H-atom resonance absorption spectrometry in a shock tube. Fig. 6b shows that the 1.3 atm data from Bentz et al. agree very well with the current data at a slightly higher pressure (1.6 atm), but their higher pressure data appear to be much nosier and seem to suggest stronger pressure dependence over 1300-1600 K and at near-atmospheric pressures than observed in the current study. Recently, Sivaramakrishnan et al.27 used a similar method as Bentz et al. to measure k1 over temperatures of 1200 – 1800 K and pressures of 0.26 – 1.3 atm. Yet their data, as displayed in Fig. 6c, show more scatter and agree less well with the current study. 10 ACS Paragon Plus Environment

Page 11 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1. Summary of the current measurement of k1, with comparison to previous theoretical calculations Current Measurement T (K)

P (atm)

k1 (s-1)

2-σ Uncertainty

Previous Theoretical Calculations Harding et al.

Gupte et al.

USC- Mech II

1000ppm CH3CHO / Ar 1618

1.45

1.4 x 104

+25% / -28%

9.5 x 103

1.1 x 104

1.7 x 103

1522

1.54

3.5 x 103

+27% / -30%

2.6 x 103

3.1 x 103

5.5 x 102

1447

1.60

1.0 x 103

+28% / -32%

8.2 x 102

9.8 x 102

2.0 x 102

1382

1.66

3.0 x 102

+30% / -34%

2.7 x 102

3.2 x 102

7.0 x 101

1331

1.72

9.7 x 101

+32% /- 37%

1.0 x 102

1.2 x 102

2.8 x 101

1273

1.75

3.1 x 101

+36% / -40%

3.0 x 101

3.5 x 101

8.7 x 100

1571

0.30

3.8 x 103

+24% / -26%

2.0 x 103

2.5 x 103

2.6 x 102

1520

0.33

2.1 x 103

+28% / -29%

1.0 x 103

1.3 x 103

1.5 x 102

1453

0.36

7.5 x 102

+31% / -31%

4.0 x 102

5.1 x 102

6.7 x 101

1377

0.41

2.0 x 102

+29% / -28%

1.1 x 102

1.5 x 102

2.3 x 102

50ppm CH3CHO / Ar 1566

1.40

7.1 x 103

+25% / -27%

4.6 x 103

5.6 x 103

8.8 x 102

1519

1.47

3.7 x 103

+24% / -26%

2.4 x 103

2.9 x 103

5.1 x 102

1494

1.49

2.5 x 103

+24% / -26%

1.7 x 103

2.0 x 103

3.7 x 102

1464

1.52

1.5 x 103

+24% / -27%

1.0 x 103

1.3 x 103

2.4 x 102

1423

1.55

7.3 x 102

+24% / -28%

5.3 x 102

6.4 x 102

1.3 x 102

1363

1.60

2.4 x 102

+23% / -27%

1.8 x 102

2.2 x 102

4.9 x 101

1559

0.30

4.0 x 103

+28% / -30%

1.7 x 103

2.1 x 103

2.2 x 102

1524

0.31

2.4 x 103

+25% / -27%

1.1 x 103

1.4 x 103

1.5 x 102

1487

0.31

1.3 x 103

+24% / -25%

6.2 x 102

7.9 x 102

9.4 x 101

1460

0.32

9.5 x 102

+22% / -24%

4.1 x 102

5.3 x 102

6.7 x 102

1425

0.34

5.2 x 102

+23% / -25%

2.4 x 102

3.1 x 102

4.3 x 101

1399

0.37

3.4 x 102

+25% / -28%

1.6 x 102

2.1 x 102

3.1 x 101



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 40

Figure 6. Comparison of current results with previous experimental studies of (a) Gupte et al.,25 (b) Bentz et al.,26 and (c) Sivaramakrishnan et al.28 12 ACS Paragon Plus Environment

Page 13 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The results of the current measurement are also compared with recent theoretical calculations. Fig. 7 shows the RRKM predictions of k1 from Harding et al.28 and Gupte et al.,25 evaluated at the mean pressures of the current study (1.6 atm and 0.34 atm). In the 1.6 atm case, both predictions lie within the error limits of the current data. A more detailed comparison at the individual shock conditions, as shown in Table 1, suggest that Gupte et al. agree better with the present study on the high-temperature end, whereas Harding et al. agree better on the lowtemperature end. In the 0.34 atm case, however, both theories underpredict k1 by 30-50%, possibly resulting from their uncertainties in characterizing the pressure dependence of k1. In fact, about a factor of 1.3-1.8 difference in k1 is observed between the 1.6 atm and the 0.34 atm measurements, with more pronounced pressure-dependence occurring at higher temperatures, whereas both theories predict slightly stronger pressure dependence (about a factor of 2.2 difference in k1) between these two pressures. Also included in Fig. 7 is the evaluation of the original k1 expression in USC-Mech II, which is seen to underpredict k1 by about half an order of magnitude at 1.6 atm, and about one order of magnitude at 0.34 atm.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 40

Figure 7. Comparison of the current measurement with previous theoretical calculations at pressures near (a) 1.6 atm, and (b) 0.34 atm

It is worth noting that the following roaming channel of acetaldehyde dissociation was also recognized by Harding et al.:28 CH3CHO (+M) = CH4 + CO (+M)

(R1a)

with a theoretically predicted branching ratio (BRroam) of about 10%. This was supported by the experimental results of Sivaramakrishnan et al.27 with a measured BRroam of 23+/-9%. However, the roaming channel was not included in USC-Mech II, the base mechanism used in the current study. As the current measurement of k1 is inferred from the formation rate of CO, which can be estimated via quasi-equilibrium approximation as 𝑑𝑑[𝐶𝐶𝐶𝐶]/𝑑𝑑𝑑𝑑 = 2𝑘𝑘1 [𝐶𝐶𝐶𝐶3𝐶𝐶𝐶𝐶𝐶𝐶] + 𝑘𝑘1𝑎𝑎 [𝐶𝐶𝐶𝐶3𝐶𝐶𝐶𝐶𝐶𝐶], inclusion of this roaming channel would effectively scale the current rate expression for k1 by a factor of 1-BRroam/2. Because this roaming correction (7-16%) is well below our measurement uncertainty, we chose not to include it in the current rate expression. Contributions from other


Page 15 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


reaction channels (listed below) are negligible (

Shock Tube Measurement for the Dissociation Rate Constant of

Recommend Documents