- Email: [email protected]
The influence of H2SO4 on soot oxidation with NO2
CARBON
5 0 ( 2 0 1 2 ) 2 1 0 0 –2 1 0 9
Available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/carbon
The influence of H2SO4 on soot oxidation with NO2 M. Mehring, M. Elsener, L. Ba¨chli, O. Kro¨cher
*
Paul Scherrer Institute, 5232 Villigen PSI, Switzerland
A R T I C L E I N F O
A B S T R A C T
Article history:
The influence of H2SO4 traces on soot oxidation by NO2 and O2 in the presence of water was
Received 7 September 2011
investigated between 100 and 700 C. When there was NO2 in the feed, the addition of
Accepted 28 December 2011
H2SO4 traces enhanced the oxidation of soot over the whole temperature range. Without
Available online 11 January 2012
NO2, there was only a slight increase in soot oxidation observed above 450 C. The HNO3 concentration in the product gas was found to increase with increasing concentrations of H2SO4. Based on these results, a mechanism was developed in which H2SO4 enhances the formation of nitronium ions (NOþ 2 ) – the active species during nitration of the aromatic carbon structures. Because nitration of aromatic soot structures is one of the major steps of soot oxidation with NO2, the increased oxidation rates can be explained by the enhanced concentration of NOþ 2 . In addition, H2SO4 is also assumed to accelerate the formation of HNO3 by the disproportionation of two NO2 molecules into HNO2 and HNO3 via protonation of the intermediate N2O4. This results in an increased HNO3 concentration, which further enhances the NOþ 2 concentration and, as a consequence, the C-oxidation rate. 2012 Elsevier Ltd. All rights reserved.
1.
Introduction
Soot in the exhaust gas of diesel engines is known to have many adverse health effects [1,2]. To mitigate these effects, limits have been placed on the release of soot emissions from diesel engines [3]. These limits can be achieved by improvements of the combustion process and the additional introduction of a diesel particulate filter (DPF) [4,5]. In the DPF, soot is accumulated on and in the walls of combustion systems leading to an increase of the pressure drop over the DPFs with time, and as a consequence also to increased fuel consumption. Therefore, the DPF has to be regenerated by oxidation of the accumulated carbon by the oxygen already present in the exhaust gas. However, because the temperatures are often too low to ensure total oxidation of the soot with oxygen, additional heat is required [4,5]. To avoid this additional heating, alternative oxidants were investigated. NO2 was found to oxidize soot at significantly lower temperatures than oxygen [6–10] and allowed for continuous regeneration of the DPFs
[11,12]. Setibudi et al. and Jaquot et al. explained this promoting effect on soot oxidation by the formation of surface oxygen functional groups formed from the reaction of NO2 with active carbon sites, which are more reactive than those formed from the reaction with oxygen [7,9]. The authors assumed that these functional groups react with O2 or NO2 in a subsequent step, evolving CO2, CO and, in the case of NO2, also NO, leading to faster consumption of solid carbon. In addition, Jequirim et al. also suggested an oxidation mechanism, where O2 adsorbs on the soot surface and forms carbon–oxygen complexes, which are attacked faster by NO2 than O2, leading to the evolution of NO, CO and CO2 [13]. It was also reported that, in the presence of NO2, the formation of acidic surface functional groups increases [14], and these groups have lower temperatures than the more basic groups formed in the presence of only O2. Because the exhaust gas of diesel vehicles usually does not contain enough NO2, Pt-based diesel oxidation catalysts (DOCs) are applied upstream of the DPFs to oxidize a part of the NO coming from the engine.
* Corresponding author: Fax: +41 56 310 2624. E-mail address: [email protected] (O. Kro¨cher). 0008-6223/$ - see front matter 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.carbon.2011.12.061
CARBON
5 0 ( 20 1 2 ) 2 1 0 0–21 0 9
In addition to NO2, there is always a small amount of SO3 and H2SO4 formed in the DOCs from the conversion of SO2 due to the presence of sulfur in fossil diesel fuel. Although the sulfur content of modern diesel fuel is very low, an influence of these compounds on soot oxidation in the DPFs cannot be excluded due to their reactive nature. In fact, Oi-Uchisawa et al. and Obuchi et al. showed in laboratory experiments with model soot that traces of SO2 further enhanced the oxidation of soot in the presence of a Pt-based oxidation catalyst, NO, O2 and water [15–18]. They concluded that over the Pt catalyst NO and SO2 were oxidized to NO2 and SO3, respectively, which was in equilibrium with H2SO4 in the presence of water. The NO2 formed oxygen surface functional groups, whose decomposition was assumed to be catalyzed by SO3 and H2SO4. To exclude the influence of the catalyst soot interaction, we performed a study on the influence of H2SO4 on the soot oxidation without catalyst. Instead of dosing NO and SO2, NO2 was dosed in the gas feed and H2SO4 was either dosed in the feed gas or impregnated on the soot prior to the oxidation experiment. Based on our results we developed a new mechanism, which explains the promoting effect of H2SO4 on the oxidation of diesel soot by NO2.
2.
Experimental
We used commercially available Printex U from Evonik as a model substance for diesel soot because it is available in sufficient amounts, has a reproducible composition, is well characterized in many studies and shows similarities in surface chemistry [10,19,20]. Cordierite monoliths (400 cpsi, V = 15.5 mL, dimensions 26.5 mm · 26.5 mm · 22 mm) were used as soot carriers. They were loaded with 60 mg soot by dip coating into a dispersion of soot and isopropyl alcohol (IPA). After dipping, the solvent was removed first with an air blower for 2 min and then dried at 85 C for 4 h in a cabinet dryer. Part of the loaded monoliths was impregnated with either 1 (±10%), 2 (±10%), 5 (±10%) or 10 (±10%) mg of a H2SO4 solution. The impregnation was performed by dipping the monoliths once for 1 s into H2SO4 solutions of different concentrations and subsequently drying at room temperature (RT). The H2SO4 concentrations of these solutions were calculated from the water absorption capacity of a monolith during one-time dipping. For the monoliths used in this work, the maximum water absorption capacity was 1.7 mL, leading to the H2SO4 concentrations shown in Table 1. Experiments testing the reactivity of Printex U in the presence of NO2 and SO3/H2SO4 were performed in a heated tubu-
2101
lar quartz reactor (length: 650 mm; inner diameter: 49 mm) as shown in Fig. 1. After the feed gases entered the reactor, they were temperature-controlled in the pre-heating zone (heating zone 1), which was filled with ceramic beads for effective heat transfer. Then the gases reached the region around the inlet of the SO3 generator (heating zone 2), which was also filled with ceramic beads. After passing the frit, the gases entered the reaction zone (heating zone 3), where the sample holder was placed. The last heating zone (heating zone 4) controlled the reactor outlet downstream of the sample, where approximately 10–15% of the product gas flow was extracted continuously through quartz pipes for gas analysis. For taking measurements, the loaded monoliths were placed in a sample holder that allowed for reproducible positioning of the samples in the reactor. The gap between the sample holder and the reactor walls was sealed with a ceramic fiber mat to avoid any bypass. The gases were provided by a gas mixing unit consisting of six mass flow controllers (Brooks 5850S). Water was generated by controlled hydrogen oxidation over a Pt catalyst (Fig. 1). NO2 was added to the basic feed gas in varying concentrations. SO3 was generated by oxidation of SO2 over a Pt catalyst (2% Pt on SiO2) in a bypass stream of 15% of the total flow. Since the SO2 conversion rates on the catalyst did not exceed 65–70%, the SO2 always had to be dosed in excess. After entering the main gas stream, which contained water, H2SO4 was formed in equilibrium with SO3 in dependence of water partial pressure and temperature. The total flow was 550 L/h at STP, resulting in a gas hourly space velocity (GHSV: quotient of gas volume flow and the catalyst volume) of 35,000 h1, which is a realistic space velocity for mobile application. The base feed gas, which was used for most of the experiments, consisted of 10% O2 + 5% H2O, with nitrogen as the balance. For the first part of the soot oxidation measurements, 0, 200 and 1000 ppm NO2 and 0, 1, 4 and 10 ppm H2SO4 were added to the base feed gas. For the second part, the H2SO4-impregnated monoliths were used, and 0, 200 and 1000 ppm NO2 was added to the base feed gas, respectively. Additionally, the soot reactivity was also investigated with either 10% O2 or 10% O2 + 5% H2O
Table 1 – H2SO4 concentrations of the H2SO4 solutions used for impregnation of the soot-loaded monoliths. H2SO4 concentration in precursor solution (mol/L) 0.060 0.030 0.010 0.006
H2SO4 on monolith (mg) 10 5 2 1
Fig. 1 – Schematic of the experimental setup for investigating the influence of H2SO4 on soot oxidation with NO2. The dotted lines symbolize communication connections of the FTIR spectrometer and the MFCs with the PC.
2102
CARBON
5 0 ( 2 0 1 2 ) 2 1 0 0 –2 1 0 9
as a comparison. All measurements were performed with a heating ramp of 10 K/min between 50 and 700 C. The gas composition was analyzed quantitatively downstream of the quartz reactor using a Nicolet Nexus FTIR spectrometer with a gas measuring cell of 2 m path length and ZnSe windows. All tubes downstream of the gas mixing unit and the FTIR cell were heated to 180 C to avoid condensation. The adsorption of H2SO4 on the ZnSe windows of the FTIR gas measuring cell did not influence the measurements because the H2SO4 concentrations were kept below 15 ppm. The spectrometer was calibrated for CO2, CO, NO2, NO, HNO3, H2O, SO2 and H2SO4. Calibration of these components, except H2SO4, was done using the standard procedure with calibration gases from gas bottles. The calibration of H2SO4 was more difficult because first a qualitative reference spectrum of H2SO4 had to be collected, and in a second step, the quantification method of the spectrometer had to be calibrated with a nitrogen-based model gas containing 10% O2 + 5% H2O + 2, 6 and 10 ppm H2SO4. For recording a reference spectrum of H2SO4, a dry nitrogen flow was saturated in a temperature-controlled wash bottle containing a concentrated H2SO4 solution. The spectrum showed absorption bands of both SO3 and H2SO4; therefore, the region between 1200 and 1260 cm1 showing only specific bands for H2SO4 [21] was used as a reference for further calibration. Quantification of the reference spectrum was performed by dosing different SO2 concentrations over the SO3 generator in the presence of 10% O2 with nitrogen as the balance. When the gas flow from the SO3 generator was mixed with the main nitrogen-based gas flow containing 5% H2O, H2SO4 was formed, which was analyzed via a wet chemical analysis method as described in [22–24]. The obtained concentrations were used for quantification with FTIR. Because the SO3/ H2SO4 equilibrium in the presence of 5% H2O at 180 C in the gas measuring cell is biased toward H2SO4, a calibration of SO3 was unnecessary. Therefore, in our study, only H2SO4 will be discussed, although under certain reaction conditions, H2SO4 and SO3 might be present in equilibrium.
3.
Results
3.1. Introductory example for the impact of H2SO4 on soot oxidation with NO2 Fig. 2 shows the trends of the gas concentrations observed during a typical soot oxidation experiment performed in this study. The inlet gas concentrations for the chosen example were 200 ppm NO2 + 10% O2 + 5% H2O + 18 ppm SO2, with nitrogen as the balance. The figure shows the outlet concentrations of CO, CO2, NO, NO2, SO2, H2SO4 and HNO3. At 170 C, the soot surface started to be oxidized by NO2 leading to the formation of oxygen surface functional groups and gaseous NO, as shown in reaction (1) where C(O) represents a surface functional group and Cbulk the carbon backbone: Cbulk C þ NO2 ! Cbulk CðOÞ þ NO
ð1Þ
Fig. 2 – Outlet concentrations of CO, CO2, NO, NO2, SO2, H2SO4 and HNO3 in a soot oxidation experiment with 200 ppm NO2 + 10% O2 + 5% H2O + SO2 + 10 ppm H2SO4 in the feed gas. The slight increase of the CO2 concentration at 180 C indicates that a part of the functional groups decomposed after a second attack by NO2: Cbulk CðOÞ þ NO2 ! Cbulk þ CO2 þ NO
ð2Þ
Since reaction (2) is a consecutive reaction of (1) it can be concluded that always more NO than CO2 is produced. In fact, Fig. 2 clearly shows that the slope of the NO concentration is more than two times steeper than the slope of the CO2 concentration between 170 and 210 C. At 10 C below the breakthrough of H2SO4, the NO2 and NO concentrations reached a plateau due to limitation by the available amount of reactive carbon structures. Above 230 C, the NO and CO2 concentrations increased again, while the NO2 concentration decreased. At 280 C, the NO2 conversion exceeded 90%, and in parallel to reaction (2), the carbon was also oxidized to CO, as indicated by the increase in the CO concentration: Cbulk C þ NO2 ! Cbulk CðOÞ þ NO
ð3Þ
At 315 C, the H2SO4 desorption reached a maximum, and the NO2 conversion reached 100% meaning that the CO and CO2 concentration should also reach a plateau if only NO2 accounts for soot oxidation. However, a rough balance showed that the amount of CO and CO2 was too large to be produced only by the dosed NO2. Therefore, it was concluded that all additional CO and CO2 formed above this temperature must be due to the decomposition of previously formed surface functional groups according to reaction (1) or due to reactions (4)–(6) with oxygen 2 Cbulk CðOÞ þ O2 ! 2 Cbulk CðOÞðOÞ ! 2 Cbulk þ 2 CO2
ð4Þ
Cbulk C þ O2 ! Cbulk CðOÞðOÞ ! Cbulk þ CO2
ð5Þ
2 Cbulk C þ O2 ! 2 Cbulk CðOÞ ! 2 Cbulk þ 2 CO
ð6Þ
With increasing temperature, the concentration of SO2, CO and CO2 increase, while the sum of the NO and NO2 concentration remained constant during the entire experiment. It is important to note that an influence of excess SO2 on the reaction was not observed. SO2 was added below 100 C, but H2SO4 did not break through until 210 C due to adsorption on the system walls and on the soot sample. NO2 was added at
CARBON
5 0 ( 20 1 2 ) 2 1 0 0–21 0 9
2103
120 C. Later we will show that the amount of added H2SO4 was too small to influence the observed carbon oxidation significantly. For the other NO2 and H2SO4 concentrations specified in the experimental section the absolute concentrations of the products changed, but the general trends were similar.
3.2.
NO2 conversion and C-oxidation rates
The influence of H2SO4 on soot oxidation with NO2 was investigated systematically in the presence of different NO2 and H2SO4 inlet concentrations by calculating the NO2 conversion and the transient C-oxidation rates from the data of the FTIR analysis. The transient C-oxidation rate was defined as the ratio of the carbon amount in the gas phase at time t (cgas,t) in [lg] (calculated from the CO, CO2, HNCO and HCN concentration in the gas phase detected in the FTIR spectrum at time t) to the carbon amount at time t (csolid,t) in [mg] remaining in the solid phase, referred to the period of time needed for the collection of a FTIR spectrum tcollect in [s] cgas;t lg 1 ¼ ð7Þ C-oxidation rate mg s csolid;t tcollect csolid,t was calculated from the total carbon amount at the beginning of the measurements, csolid,0, minus the sum of carbon desorbed up to time t. This way of calculating the C-oxidation rate is similar to a normalization of the data, which allows for a comparison of the results from different reactive gases and samples of different sizes. In comparison to C-oxidation rates referred to the initial carbon amount, transient C-oxidation rates have several advantages: They are more realistic, as the carbon conversion to CO and CO2 is referred to the actual amount of remaining carbon at the point of measurement and not to a quantity, which is only measured once at the beginning of the experiment. This results in an increasing deviation of the transient rate during carbon consumption from the traditional way of referring the oxidation rates on the initial carbon amount. Since the actual carbon amounts are considered in the denominator, the transient C-oxidation rates become significantly larger than the rates referred to the initial carbon amount especially above 400 C and are, therefore, more sensitive in this temperature range. Moreover, they directly represent the kinetics, because the rates increase with temperature and do not show a global maximum implying decreasing rates at high burn-off levels. Figs. 3 and 4 show the NO2 conversion during soot oxidation with 200 and 1000 ppm NO2 in the feed gas, respectively. Fig. 3 depicts the results from the addition of H2SO4 via the gas phase between 150 and 700 C, and Fig. 4 shows the results from adding H2SO4 by impregnation between 150 and 700 C. As a basis for comparison, the NO2 conversion in the presence of 200 and 1000 ppm NO2 without H2SO4 in the feed gas was included as reference in the figures. The conversion rates were plotted up to 99.5% carbon conversion. When 200 ppm NO2 without H2SO4 were added to the feed gas, the NO2 conversion increased with temperature from 4% at 200 C up to 100% at 450 C. This conversion level was kept until the main part of the soot sample was oxidized around
Fig. 3 – Increase in the NO2 conversion rates for the soot oxidation experiments in the presence of NO2 and H2SO4 between 150 and 700 C. Feed gas: 10% O2 + 5% H2O + 200 ppm NO2 + 0, 1, 4, 10 ppm H2SO4 and 10% O2 + 5% H2O + 1000 ppm NO2 + 0, 10 ppm H2SO4. The NO2 conversion is plotted until 99.5% carbon conversion.
Fig. 4 – Increase in the NO2 conversion rates for the soot oxidation experiments in the presence of NO2 and H2SO4 between 150 and 700 C. Feed gas: 10% O2 + 5% H2O + 200 ppm NO2 + 0, 1, 5, 10 mg H2SO4 and 10% O2 + 5% H2O + 1000 ppm NO2 + 0, 10 mg H2SO4. The NO2 conversion is plotted until 99.5% carbon conversion. 560 C, which led to a decrease of the NO2 conversion. By the addition of 1 ppm H2SO4 to the feed gas the base NO2 conversion increased from 4% to 6%. Above 200 C, NO2 conversion increased faster than without H2SO4 up to 320 C. Above 320 C the slope became smaller, most probably due to limitation of the available amount of carbon structures, which are reactive enough to be oxidized in this temperature range. However, a further increase of temperature resulted in a steeper slope again and full conversion was reached at 450 C, where even less reactive carbon structures could be rapidly oxidized by NO2. When the H2SO4 concentration was further increased to 4 and 10 ppm the base conversion increased only little. However, above 180 C NO2 conversion increased significantly faster than in the presence of 1 ppm H2SO4. The plateaus at 230 C were again due to limitation of the available amount
2104
CARBON
5 0 ( 2 0 1 2 ) 2 1 0 0 –2 1 0 9
of reactive carbon structures. For 10 ppm H2SO4 full NO2 conversion was already reached at 320 C, while in the presence of 4 ppm H2SO4 a slight flattening of the curve was observed before full conversion was achieved at 400 C. Similar to the experiment without H2SO4, the NO2 conversion decreased when the soot was nearly totally gasified. For 1000 ppm NO2 full conversion was not reached because under the applied conditions, the reaction was limited by the available amount of reactive carbon structures over the whole temperature range. As expected, total carbon oxidation was achieved at significantly lower temperatures in the presence of 1000 ppm NO2 than in the presence of 200 ppm NO2 due to the higher amount of oxidant. The addition of 10 ppm H2SO4 to the feed gas containing already 1000 ppm NO2 resulted in an increase of the NO2 conversion at lower temperatures. The first maximum was also due to limitation of the available amount of reactive carbon structures. The trends of the NO2 conversions rates for the experiments with the H2SO4-impregnated samples were similar to those obtained for the addition of H2SO4 via the gas phase (Fig. 3). However, full NO2 conversion was not achieved in the experiments with the impregnated samples. This was most probably due to the larger amounts of H2SO4, which were present at the beginning of the experiments with the H2SO4-impregnated samples compared to the experiments with H2SO4 in the gas phase, leading to a limitation of the available amount of reactive carbon structures also in this case. In Figs. 5–8 the carbon oxidation (C-oxidation) rates of Printex U are shown for different reaction conditions. The C-oxidation rates for the experiments with 10% O2 + 5% H2O + NO2 were significantly higher than those seen in experiments without NO2. For example, in Fig. 5, the C-oxidation rate observed at 300 C in the absence of NO2 had already been reached at 200 C in the presence of 200 ppm NO2. By increasing the NO2 concentration to 1000 ppm, these differences were enhanced (Fig. 7). As already expected from the trends of the NO2 conversion rates, it was possible to further increase the level of carbon
Fig. 5 – C-oxidation rates for soot oxidation experiments done in the presence of NO2 and H2SO4. Reactive gas compositions: 10% O2 + 5% H2O + 0 and 200 ppm NO2 + 0, 1, 4 and 10 ppm H2SO4. The C-oxidation rates are plotted until 99% carbon conversion.
oxidation in the presence of NO2 by adding small amounts of H2SO4. Adding 1 ppm H2SO4 enhanced the C-oxidation at 300 C by 75%, and the addition of 4 and 10 ppm H2SO4 enhanced the rate by 120% (Fig. 5). For 1000 ppm NO2, the addition of 1 ppm H2SO4 at 300 C did not enhance the C-oxidation, but at levels of 4 and 10 ppm, the rates were increased by a factor of 3 and 4, respectively (Fig. 7). For 200 ppm NO2 (Fig. 5) the C-oxidation rates approached each other above 410 C regardless of the reactive gas. For 1000 ppm NO2 (Fig. 7), the C-oxidation rates did not converge, and with 4 and 10 ppm H2SO4 in the reactive gas, the soot sample was totally gasified at a temperature 190 C lower compared to the experiments performed without NO2 and H2SO4. Similar results were obtained for oxidation of the soot samples impregnated with H2SO4 before the measurement (Figs. 6 and 8). The carbon oxidation rate was enhanced when NO2 was passed over the H2SO4-impregnated soot samples. In these figures, the C-oxidation rates seen with 0 ppm NO2 + 10% O2 + 5% H2O for a Printex U sample impregnated with 10 mg H2SO4 were added. These results do not show a significant difference between the other experiments without NO2 up to 450 C. Above 450 C, the C-oxidation rates increased slightly.
3.3.
HNO3 evolution
The trend of the HNO3 concentration was also investigated in our study because HNO3 is assumed to be an intermediate of soot oxidation in the presence of H2O and NO2 [25] and can be formed by a heterogeneous disproportionation reaction of NO2 (Eq. (4)), which was confirmed by several studies at RT concerning atmospheric chemistry [26–28]: 2 NO2 þ H2 O ! HNO3 þ HNO2
ð8Þ
HNO2 was not calibrated for these experiments, but its formation on Printex U was detected during experiments that were performed subsequent to this study under similar con-
Fig. 6 – C-oxidation rates for soot oxidation experiments done in the presence of NO2 and H2SO4. Reactive gas compositions: 10% O2 + 5% H2O + 0 and 200 ppm NO2 + 0, 1, 5 and 10 mg H2SO4. The C-oxidation rates are plotted until 99% carbon conversion.
CARBON
5 0 ( 20 1 2 ) 2 1 0 0–21 0 9
Fig. 7 – C-oxidation rates for soot oxidation experiments done in the presence of NO2 and H2SO4. Reactive gas compositions: 10% O2 + 5% H2O + 0 and 1000 ppm NO2 + 0, 1, 4 and 10 ppm H2SO4. The C-oxidation rates are plotted until 99% carbon conversion.
Fig. 8 – C-oxidation rates for soot oxidation experiments done in the presence of NO2 and H2SO4. Reactive gas compositions: 10% O2 + 5% H2O + 0 and 1000 ppm NO2 + 0, 1, 2, 10 mg H2SO4. The C-oxidation rates are plotted until 99% carbon conversion.
ditions. In the example shown in Fig. 2, only a maximum of 1 ppm HNO3 was detected between 180 and 220 C, but higher amounts of HNO3 were detected for higher NO2 concentrations, as shown in Figs. 9 and 10. For the experiment without H2SO4 in the feed gas, the formation of HNO3 was below the detection limit when only 200 ppm NO2 were dosed, but HNO3 formation increased significantly when the NO2 concentration was increased to 1000 ppm (Fig. 9). The addition of 10 ppm H2SO4 to the reactive gas led to the formation of approximately 1 ppm HNO3 at 210 C in the presence of 200 ppm NO2. In the presence of 1000 ppm NO2, the HNO3 concentration between 120 and 180 C was similar to the amounts seen for the experiment without H2SO4 in the reactive gas, but between 180 and 220 C, a peak of 8 ppm HNO3 could be observed. The addition of H2SO4 by impregnation had an even more pronounced effect (Fig. 10). For both NO2 inlet concentrations, more HNO3 was formed when the H2SO4 was impregnated as
2105
Fig. 9 – HNO3 produced during soot oxidation in the presence of NO2 and H2SO4. Feed gas composition: 200 and 1000 ppm NO2 + 0 and 10 ppm H2SO4.
Fig. 10 – HNO3 produced during soot oxidation in the presence of NO2 and H2SO4. Feed gas composition: 200 and 1000 ppm NO2 + 0 and 10 mg H2SO4.
compared to the addition of H2SO4 via the gas phase. Between 120 and 280 C, maximum HNO3 concentrations of 2 and 22 ppm were obtained for 200 and 1000 ppm NO2, respectively. For the final carbon gasification phase above 480 C, up to 2.5 ppm HNO3 was measured for the experiments with 1000 ppm NO2 in the reactive gas. For 200 ppm NO2, this increase was below 1 ppm, which was within the error of measurement. Before the discussion of the results of this study, it should be mentioned that the results obtained with the H2SO4-impregnated samples could not be quantitatively compared to the results of the experiments with H2SO4 in the feed gas. In the former case, a significant amount of H2SO4 was present at the start of the experiment, while in the latter case, the H2SO4 amount on the sample increased due to the SO3 dosing through the gas phase. However, the general qualitative trends observed were similar.
3.4.
Discussion
The results from this study again confirm that the addition of NO2 to the reactive gases containing O2 and water results in
2106
CARBON
5 0 ( 2 0 1 2 ) 2 1 0 0 –2 1 0 9
an increase in the carbon oxidation rate over the whole temperature range; this has already been described by several authors [7,8,10,11,25,29–33]. In addition, it was observed that the C-oxidation rates could be further increased when traces of H2SO4 were added to the reactive gas containing O2, H2O and NO2 (Figs. 5–8). This is in accordance with the results of Oi-Uchisawa et al. and Obuchi et al., who also observed a significant increase in the C-oxidation rate in the presence of NO2, which could be further enhanced by the addition of small amounts of H2SO4 [15–18]. However, in contrast to our experiments, Oi-Uchisawa et al. and Obuchi et al. used a mixture of a commercially available carbon black and a Pt catalyst for their investigations [15–18]. In their case, the Pt catalyst was needed for the in situ generation of NO2 and H2SO4 from NO and SO2. In addition to NO and SO2, their reactive gases also contained O2 and H2O, with nitrogen as the balance. Based on their findings, Oi-Uchisawa et al. proposed a mechanism that explains this enhancement [18]. The authors suggested that first phenols and carboxylates are formed during the partial oxidation of the carbon surface with NO2. Further oxidation of these groups with NO2 on the carbon surface would lead to their decomposition, but their electronwithdrawing character decreased the reactivity of neighboring carbon sites. The addition of H2SO4 could increase the C-oxidation rate because its strong acidity is assumed to catalyze the decomposition of the surface functional groups formed by the first attack of NO2. Furthermore, they suggested that H2SO4 could hydrolyze esters and anhydrides, leading to the decarboxylation of the hydrolysis products. After the decomposition of the oxygenated surface species, the carbon surface was ‘‘free’’ for a further oxidation cycle with NO2. The results of our work confirmed the assumptions of OiUchisawa et al. that a co-operative mechanism between H2SO4 and NO2 exists, where H2SO4 behaves more like a catalyst than a stoichiometric reactant (i.e., even traces of H2SO4 have a promoting effect). The catalytic effect of H2SO4 on the oxidation capability of NO2 was supported by our finding that, below 480 C, H2SO4 could not enhance the C-oxidation rates in the absence of NO2 (as seen in Figs. 6 and 8). Although the explanation in [18] on the influence of H2SO4 on the carbon oxidation seems to be plausible, it is questionable if the acceleration of the decarboxylation and of hydrolysis
of anhydrides and esters were the only processes responsible for the increase in the C-oxidation rates. We suggest instead (or in addition) that H2SO4 also influences carbon oxidation at the nitration of the graphitic soot structure, which is most likely one of the major steps of the soot oxidation process. It is well known that the active species during the nitration of aromatics is not NO2 or HNO3 but the more electrophilic NOþ 2 (nitronium-ion) [34,35], which is formed in traces in concentrated HNO3 solutions. HNO3, which was most likely formed according to Eq. (4), was also found during experiments in our study. Although the HNO2 concentration was not followed in this study, its formation on Printex U was confirmed by a subsequent study under similar conditions. While HNO2 decomposed quickly into NO, NO2 and H2O [36], HNO3 probably formed small amounts of NOþ 2 according to route A shown in Fig. 11. Although we found HNO3 in larger amounts in the gas phase at temperatures only up to 250 C, it was most likely formed on the soot surface over the entire temperature range and was not detected at higher temperatures because the rate of desorption was smaller than the rate of reaction. This assumption was confirmed by the HNO3 traces measured in the gas phase above 500 C (Figs. 9 and 10), which were formed during experiments with H2SO4 when up to 95% of the carbon was already burned. In this case, the rate of desorption was again higher than the rate of reaction. Furthermore, it is known that deactivated aromatics can also be nitrified when nitrosulfuric acid – a mixture of concentrated HNO3 and H2SO4 – is used, because the NOþ 2 concentration increases in the presence of a strong acid [34,35]. We suggest that the addition of H2SO4 to the reactive gas containing NO2 and H2O also promoted these favorable conditions for the formation of NOþ 2 ions on the soot surface. The strong acidity of H2SO4 increased the NOþ 2 concentration by protonation of HNO3, as shown in route B of Fig. 11, compared to the self-protonation pathway shown in route A. After nitration, the C–NO2 bond was most probably either homolytically cleaved, resulting in an ‘‘aryl radical’’ and NO2, or the nitro group was rearranged to a nitrite [37–41], which decomposed homolytically into a ‘‘phenyl radical’’ and NO. The ‘‘aryl radical’’ can react with HOOÆ radicals, formed by hydrogen abstraction from the carbon surface with O2, or directly with NO2 [42] resulting also in ‘‘phenyl
Fig. 11 – Mechanism for the formation of NOþ 2 [34]. (A) Reaction route in conc. HNO3 and (B) reaction route in nitrosulfuric acid.
CARBON
5 0 ( 20 1 2 ) 2 1 0 0–21 0 9
radicals’’. Electron rearrangement of the ‘‘phenyl radical’’ leads to the formation of a ‘‘cycloalkyl radical’’ and a carbonyl group and by oxidation of the radical with NO2 or HOOÆ to ringopening as shown in Fig. 12. The carbonyl group formed in the last step of Fig. 12 might decompose either directly under decarbonylation or can be further oxidized to a carboxylic group decomposing under decarboxylation. The ‘‘carbonyl radical’’ formed in the last step of Fig. 12 might also be oxidized to a carboxylic group or decomposes under decarbonylation leaving an ‘‘alkyl radical’’ on the soot surface. We assume this ‘‘alkyl radical’’ to be a potential active centre for the continuation of a radical chain reaction consisting of oxidation by HOÆ, HOOÆ, O2 or NO2, which is either followed by direct decarbonylation or a second oxidation step followed by decarboxylation. Based on the analytical results and considerations, it is reasonable to conclude that an increase in the nitration rate also enhanced the carbon oxidation. As already speculated, the H2SO4 acted catalytically as a proton donator during the formation of NOþ 2 , which could be regenerated after the nitration step. In addition to enhancing the formation of NOþ 2 , Figs. 9 and 10 show that the H2SO4 also enhanced the formation of HNO3. This can be explained by a closer look at the mechanism behind reaction (8). First, NO2 is dimerized to N2O4 (Eq. (9)), which disproportionates into HNO2 and HNO3 (Eq. (10)) in a second step in the presence of water. With increasing temper-
2107
ature, the equilibrium of the dimerization step is shifted to the NO2 side. In the presence of H2SO4, it is assumed that the HNO3 concentration increases because H2SO4 catalyzed the decomposition of N2O4, as illustrated in Fig. 13 by protonation of N2O4, facilitating the attack of water. Therefore, H2SO4 catalyzes not only the formation of NOþ 2 but also the formation of HNO3 – the NOþ 2 precursor 2 NO2 ! N2 O4 N2 O4 þ H2 O ! HNO3 þ HNO2
ð9Þ ð10Þ
Beside the nitration of aromatics, there might be additional non-radicalic reactions of surface groups in the presence of NO2, which can only be explained if NOþ 2 is the active species. One example for such a reaction is the oxidation of aldehydes with NO2, which proceeds via an acylnitrite [19]. Based on an analysis of the oxidation states, it is clear that a non-radicalic electrophilic attack via oxidation of the carbonyl carbon is only possible if the electrophile is NOþ 2, as shown in Fig. 14. In a subsequent step the acylnitrite most probably decomposes into NO and a COOÆ radical as already indicated by results of Muckenhuber et al. [19]. Supported by the theoretical investigations of Barco et al. [43] we assumed the COOÆ to decompose under decarboxylation leaving an ‘‘alkyl radical’’ on the surface, which reacts similar to the ‘‘alkyl radical’’ formed from the ‘‘carbonyl radical’’ in Fig. 12. A direct oxidation of soot by H2SO4 also has to be considered because an increase in the SO2 concentration in the product gas was observed with increasing temperature, even
Fig. 12 – Decomposition of the ‘‘phenyl radical’’. The benzene ring symbolizes one aromatic ring of the soot backbone.
Fig. 13 – Mechanistic details of the HNO3 and HNO2 formation by N2O4 decomposition over soot.
2108
CARBON
5 0 ( 2 0 1 2 ) 2 1 0 0 –2 1 0 9
Fig. 14 – Example reactions of surface groups with NOþ 2 : the oxidation of aldehydes to nitrites. R stands for the carbon backbone.
though the amount of SO2 added to the SO3 generator remained constant. However, the amounts of H2SO4 were too small to significantly contribute to the C-oxidation rates observed at temperatures up to 450 C (Fig. 6) Above 450 C, a slight increase in the C-oxidation rate was observed (Figs. 6 and 8), which corresponds to the formation of SO2 in Fig. 2. The results from the impregnated soot samples and those oxidized with the addition of H2SO4 in the gas phase showed that the differences were mainly attributed to differences in the actual amounts of H2SO4. The impregnated samples were more active at 120 C because the amount of H2SO4 was larger in this temperature range. During gas phase addition, adequate amounts of H2SO4 had to be accumulated before an increase in the C-oxidation rates was observed.
4.
Conclusions
The results from this study clearly show that traces of H2SO4 resulting from oxidized fuel sulfur have a significant promoting effect on soot oxidation via NO2. Based on a detailed infrared gas phase analysis, a new mechanism has been suggested to explain the activation effect of H2SO4; it most probably acts in parallel to the already published activation mechanisms. In our new mechanism, H2SO4 acts as a catalyst at two stages. First, it enhances the formation of HNO3 by catalyzing the disproportionation of N2O4, and secondly, H2SO4 most likely increases the amount of NOþ 2 , which is the known active species for the nitration of aromatic structures. Once the aromatic system is nitrified, the C–N bond is either cleaved homolytically or the nitro group is rearranged to a nitrite, which also decomposes homolytically to NO and a ‘‘phenyl radical’’. These radical species are further oxidized, leading to ring-opening and decomposition of the aromatic system by decarboxylation and decarbonylation. For applications with high-sulfur fuels, a significant influence of the soot oxidation by the sulfur is expected. However, we have clearly shown that even very small sulfur concentrations (below 10 ppm), as found in modern low-sulfur diesel fuels, have a strong catalytic effect on soot oxidation. Moreover, the situation is aggravated by the fact that SO3/H2SO4 are accumulated in the DOCs under low load conditions and can be released quickly in high concentrations, when the
exhaust temperature increases [44]. After SO3/H2SO4 has been released from the DOC they enter the DPF and react with the accumulated soot as shown in this study.
R E F E R E N C E S
[1] Bu¨nger J, Krahl J, Baum K, Schro¨der O, Mu¨ller M, Westphal G. Cytotoxic and mutagenic effects, particle size and concentration analysis of diesel engine emissions using biodiesel and petrol diesel as fuel. Arch Toxicol 2000;74(8):490–8. [2] Yamaguchi T, Yamazaki H. Cytotoxicity of airborne particulates sampled roadside in rodent and human lung fibroblasts. J. Health Sci. 2001;47(3):272–7. [3] Gazette of the European Union, EU, Regulation No. 715/2007. [4] Tschoeke H, Graf A, Stein J, Krueger M, Schaller J, Breuer N et al. Diesel engine exhaust emissions. In: Mollenhauer K, Tschoeke H, editors. Handbook of diesel engines. Berlin, Heidelberg; Springer; 2010. p. 417–85. [5] Twigg MV. Progress and future challenges in controlling automotive exhaust gas emissions. Appl Catal B-Environ 2007;70(1–4):2–15. [6] Neeft JPA, Makkee M, Moulijn JA. Diesel particulate emission control. Fuel Process Technol 1996;47(1):1–69. [7] Ehrburger P, Brilhac JF, Drouillot Y, Logie V, Gilot P. Reactivity of soot with nitrogen oxides in exhaust stream. SAE 2002–011683. [8] Jacquot F, Logie V, Brilhac JF, Gilot P. Kinetics of the oxidation of carbon black by NO2: influence of the presence of water and oxygen. Carbon 2002;40(3):335–43. [9] Muckenhuber H, Grothe H. The reaction between soot and NO2 – investigation on functional groups using TPD-MS. Top Catal 2004;30–31(1):287–91. [10] Setiabudi A, Makkee M, Moulijn JA. The role of NO2 and O2 in the accelerated combustion of soot in diesel exhaust gases. Appl Catal B-Environ 2004;50(3):185–94. [11] Stanmore BR, Tschamber V, Brilhac JF. Oxidation of carbon by NOx, with particular reference to NO2 and N2O. Fuel 2008;87(2):131–46. [12] Stanmore BR, Brilhac JF, Gilot P. The oxidation of soot: a review of experiments, mechanisms and models. Carbon 2001;39:2247–68. [13] Jeguirim M, Tschamber V, Brilhac JF, Ehrburger P. Interaction mechanism of NO2 with carbon black: effect of surface oxygen complexes. J Anal Appl Pyrol 2004;72(1):171–81.
CARBON
5 0 ( 20 1 2 ) 2 1 0 0–21 0 9
[14] Muckenhuber H, Grothe H. The heterogeneous reaction between soot and NO2 at elevated temperature. Carbon 2006;44(3):546–59. [15] Obuchi A, Ohi A, Aoyama H, Ohuchi H. Evaluation of gaseous and particulate emission characteristics of a single cylinder diesel engine. Combust Flame 1987;70(2):215–24. [16] Oi Uchisawa J, Obuchi A, Zhao Z, Kushiyama S. Carbon oxidation with platinum supported catalysts. Appl Catal BEnviron 1998;18(3–4):L183–7. [17] Oi-Uchisawa J, Obuchi A, Enomoto R, Ogata A, Kushiyama S. Outstanding effect of SO2 addition on the rate of carbon oxidation with a Pt catalyst. Chem Commun 1998;20:2255–6. [18] Oi-Uchisawa J, Obuchi A, Ogata A, Enomoto R, Kushiyama S. Effect of feed gas composition on the rate of carbon oxidation with Pt/SiO2 and the oxidation mechanism. Appl Catal BEnviron 1999;21(1):9–17. [19] Muckenhuber H, Grothe H. A DRIFTS study of the heterogeneous reaction of NO2 with carbonaceous materials at elevated temperature. Carbon 2007;45(2):321–9. [20] Neeft JPA, Nijhuis TX, Smakman E, Makkee M, Moulijn JA. Kinetics of the oxidation of diesel soot. Fuel 1997;76(12):1129–36. [21] Dene C, Himes R. Continuous measurement technologies for SO3 and H2SO4 in coal-fired power plants, EPRI, 1009812. Palo Alto, CA; 2004. [22] Koebel M, Elsener M. Schwefeltrioxidbestimmung in Abgasen nach der isoporopanolmethode – Eine kritische Betrachtung. Gefahrstoffe – Reinhaltung der Luft 1997;57:193–9. [23] Messen der Schwefeltrioxid-Konzentration. Wasserstoffperoxid-Verfahren, Titrimetrische Bestimmmungen, VDI 2462 Bl. 2. Berlin: Beuth-Verlag; 1974. [24] Messen der Schwefeltrioxid-Konzentration. 2-PropanolVerfahren, VDI 2462 Bl. 7. Berlin: Beuth-Verlag; 1985. [25] Jeguirim M, Tschamber V, Brilhac JF, Ehrburger P. Oxidation mechanism of carbon black by NO2: effect of water vapour. Fuel 2005;84(14–15):1949–56. [26] Akimoto H, Bandow H, Sakamaki F, Inoue G, Hoshino M, Okuda M. Photooxidation of the propylene-NOx-air system studied by long-path Fourier transform infrared spectrometry. Environ Sci Technol 1980;14(2):172–9. [27] Kleffmann J, Becker KH, Wiesen P. Heterogeneous NO2 conversion processes on acid surfaces: possible atmospheric implications. Atmos Environ 1998;32(16):2721–9. [28] Sakamaki F, Shiro H, Akimoto H. Formation of nitrous acid and nitric oxide in the heterogeneous dark reaction of nitrogen dioxide and water vapor in a smog chamber. Int J Chem Kinet 1980;15:1013–29. [29] Chughtai AR, Gordon SA, Smith DM. Kinetics of the hexane soot reaction with NO2/N2O4 at low concentration. Carbon 1994;32(3):405–16.
2109
[30] Lur’e BA, Mikhno AV. Interaction of NO2 with soot. Kinet Catal 1995;38(4):535–40. [31] Messerer A, Niessner R, Po¨schl U. Comprehensive kinetic characterization of the oxidation and gasification of model and real diesel soot by nitrogen oxides and oxygen under engine exhaust conditions: measurement, Langmuir– Hinshelwood, and Arrhenius parameters. Carbon 2006;44(2):307–24. [32] Shirahama N, Moon SH, Choi KH, Enjoji T, Kawano S, Korai Y, et al. Mechanistic study on adsorption and reduction of NO2 over activated carbon fibers. Carbon 2002;40(14):2605–11. [33] Stadler D, Rossi MJ. The reactivity of NO2 and HONO on flame soot at ambient temperature: the influence of combustion conditions. Phys Chem Chem Phys 2000;2(23):5420–9. [34] Bru¨ckner R. Reaktionsmechanismen. 2nd ed. Heidelberg: Springer; 2003. [35] Smith MB, March J. March’s advanced organic chemistry. 5th ed. New York: Wiley; 2001. [36] Hollemann AF, Wiberg E. Hollemann–Wiberg Lehrbuch der anorganischen Chemie. Berlin, New York: Walter de Gruyter; 1995. [37] He YZ, Cui JP, Mallard WG, Tsang W. Homogeneous gas-phase formation and destruction of anthranil from o-nitrotoluene decomposition. J Am Chem Soc 1988;110(12):3754–9. [38] Nikolaeva EV, Shamov AG, Chackov DV, Gordeev DA, Khraphovsky GM. The role of nitro-nitrite rearrangement in the mechanism of monomolecular decomposition of aliphatic nitro compounds. Butlerov Commun 2001;3:15–23. [39] Pitz WJ, Westbrook CK. A detailed chemical kinetic model for gas phase combustion of TNT. Proc Combust Inst 2007;31(2):2343–51. [40] Poon C, Mayer PM. Competing rearrangement reactions in small gas-phase ionic complexes: the internal SN2 and nitronitrite rearrangements in nitroalkane proton-bound pairs. Int J Mass Spectrom 2006;255–256:93–101. [41] Tsang W, Robaugh D, Mallard WG. Single-pulse shock-tube studies on C–NO2 bond cleavage during the decomposition of some nitro aromatic compounds. J Phys Chem 1986;90(22):5968–73. [42] Otsuka K, Takahashi R, Amakawa K, Yamanaka I. Partial oxidation of light alkanes by NOx in the gas phase. Catal Today 1998;45(1–4):23–8. [43] Barco G. The oxidized soot surface: theoretical study of desorption mechanisms involving oxygenated functionalities and comparison with temperature programed desorption experiments. J Chem Phys 2006;125(19): 194706 (1–12). [44] Kroecher O, Widmer M, Elsener M, Rothe D. Adsorption and desorption of SOx on diesel oxidation catalysts. Ind Eng Chem Res 2009;48(22):9847–57.
5 0 ( 2 0 1 2 ) 2 1 0 0 –2 1 0 9
Available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/carbon
The influence of H2SO4 on soot oxidation with NO2 M. Mehring, M. Elsener, L. Ba¨chli, O. Kro¨cher
*
Paul Scherrer Institute, 5232 Villigen PSI, Switzerland
A R T I C L E I N F O
A B S T R A C T
Article history:
The influence of H2SO4 traces on soot oxidation by NO2 and O2 in the presence of water was
Received 7 September 2011
investigated between 100 and 700 C. When there was NO2 in the feed, the addition of
Accepted 28 December 2011
H2SO4 traces enhanced the oxidation of soot over the whole temperature range. Without
Available online 11 January 2012
NO2, there was only a slight increase in soot oxidation observed above 450 C. The HNO3 concentration in the product gas was found to increase with increasing concentrations of H2SO4. Based on these results, a mechanism was developed in which H2SO4 enhances the formation of nitronium ions (NOþ 2 ) – the active species during nitration of the aromatic carbon structures. Because nitration of aromatic soot structures is one of the major steps of soot oxidation with NO2, the increased oxidation rates can be explained by the enhanced concentration of NOþ 2 . In addition, H2SO4 is also assumed to accelerate the formation of HNO3 by the disproportionation of two NO2 molecules into HNO2 and HNO3 via protonation of the intermediate N2O4. This results in an increased HNO3 concentration, which further enhances the NOþ 2 concentration and, as a consequence, the C-oxidation rate. 2012 Elsevier Ltd. All rights reserved.
1.
Introduction
Soot in the exhaust gas of diesel engines is known to have many adverse health effects [1,2]. To mitigate these effects, limits have been placed on the release of soot emissions from diesel engines [3]. These limits can be achieved by improvements of the combustion process and the additional introduction of a diesel particulate filter (DPF) [4,5]. In the DPF, soot is accumulated on and in the walls of combustion systems leading to an increase of the pressure drop over the DPFs with time, and as a consequence also to increased fuel consumption. Therefore, the DPF has to be regenerated by oxidation of the accumulated carbon by the oxygen already present in the exhaust gas. However, because the temperatures are often too low to ensure total oxidation of the soot with oxygen, additional heat is required [4,5]. To avoid this additional heating, alternative oxidants were investigated. NO2 was found to oxidize soot at significantly lower temperatures than oxygen [6–10] and allowed for continuous regeneration of the DPFs
[11,12]. Setibudi et al. and Jaquot et al. explained this promoting effect on soot oxidation by the formation of surface oxygen functional groups formed from the reaction of NO2 with active carbon sites, which are more reactive than those formed from the reaction with oxygen [7,9]. The authors assumed that these functional groups react with O2 or NO2 in a subsequent step, evolving CO2, CO and, in the case of NO2, also NO, leading to faster consumption of solid carbon. In addition, Jequirim et al. also suggested an oxidation mechanism, where O2 adsorbs on the soot surface and forms carbon–oxygen complexes, which are attacked faster by NO2 than O2, leading to the evolution of NO, CO and CO2 [13]. It was also reported that, in the presence of NO2, the formation of acidic surface functional groups increases [14], and these groups have lower temperatures than the more basic groups formed in the presence of only O2. Because the exhaust gas of diesel vehicles usually does not contain enough NO2, Pt-based diesel oxidation catalysts (DOCs) are applied upstream of the DPFs to oxidize a part of the NO coming from the engine.
* Corresponding author: Fax: +41 56 310 2624. E-mail address: [email protected] (O. Kro¨cher). 0008-6223/$ - see front matter 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.carbon.2011.12.061
CARBON
5 0 ( 20 1 2 ) 2 1 0 0–21 0 9
In addition to NO2, there is always a small amount of SO3 and H2SO4 formed in the DOCs from the conversion of SO2 due to the presence of sulfur in fossil diesel fuel. Although the sulfur content of modern diesel fuel is very low, an influence of these compounds on soot oxidation in the DPFs cannot be excluded due to their reactive nature. In fact, Oi-Uchisawa et al. and Obuchi et al. showed in laboratory experiments with model soot that traces of SO2 further enhanced the oxidation of soot in the presence of a Pt-based oxidation catalyst, NO, O2 and water [15–18]. They concluded that over the Pt catalyst NO and SO2 were oxidized to NO2 and SO3, respectively, which was in equilibrium with H2SO4 in the presence of water. The NO2 formed oxygen surface functional groups, whose decomposition was assumed to be catalyzed by SO3 and H2SO4. To exclude the influence of the catalyst soot interaction, we performed a study on the influence of H2SO4 on the soot oxidation without catalyst. Instead of dosing NO and SO2, NO2 was dosed in the gas feed and H2SO4 was either dosed in the feed gas or impregnated on the soot prior to the oxidation experiment. Based on our results we developed a new mechanism, which explains the promoting effect of H2SO4 on the oxidation of diesel soot by NO2.
2.
Experimental
We used commercially available Printex U from Evonik as a model substance for diesel soot because it is available in sufficient amounts, has a reproducible composition, is well characterized in many studies and shows similarities in surface chemistry [10,19,20]. Cordierite monoliths (400 cpsi, V = 15.5 mL, dimensions 26.5 mm · 26.5 mm · 22 mm) were used as soot carriers. They were loaded with 60 mg soot by dip coating into a dispersion of soot and isopropyl alcohol (IPA). After dipping, the solvent was removed first with an air blower for 2 min and then dried at 85 C for 4 h in a cabinet dryer. Part of the loaded monoliths was impregnated with either 1 (±10%), 2 (±10%), 5 (±10%) or 10 (±10%) mg of a H2SO4 solution. The impregnation was performed by dipping the monoliths once for 1 s into H2SO4 solutions of different concentrations and subsequently drying at room temperature (RT). The H2SO4 concentrations of these solutions were calculated from the water absorption capacity of a monolith during one-time dipping. For the monoliths used in this work, the maximum water absorption capacity was 1.7 mL, leading to the H2SO4 concentrations shown in Table 1. Experiments testing the reactivity of Printex U in the presence of NO2 and SO3/H2SO4 were performed in a heated tubu-
2101
lar quartz reactor (length: 650 mm; inner diameter: 49 mm) as shown in Fig. 1. After the feed gases entered the reactor, they were temperature-controlled in the pre-heating zone (heating zone 1), which was filled with ceramic beads for effective heat transfer. Then the gases reached the region around the inlet of the SO3 generator (heating zone 2), which was also filled with ceramic beads. After passing the frit, the gases entered the reaction zone (heating zone 3), where the sample holder was placed. The last heating zone (heating zone 4) controlled the reactor outlet downstream of the sample, where approximately 10–15% of the product gas flow was extracted continuously through quartz pipes for gas analysis. For taking measurements, the loaded monoliths were placed in a sample holder that allowed for reproducible positioning of the samples in the reactor. The gap between the sample holder and the reactor walls was sealed with a ceramic fiber mat to avoid any bypass. The gases were provided by a gas mixing unit consisting of six mass flow controllers (Brooks 5850S). Water was generated by controlled hydrogen oxidation over a Pt catalyst (Fig. 1). NO2 was added to the basic feed gas in varying concentrations. SO3 was generated by oxidation of SO2 over a Pt catalyst (2% Pt on SiO2) in a bypass stream of 15% of the total flow. Since the SO2 conversion rates on the catalyst did not exceed 65–70%, the SO2 always had to be dosed in excess. After entering the main gas stream, which contained water, H2SO4 was formed in equilibrium with SO3 in dependence of water partial pressure and temperature. The total flow was 550 L/h at STP, resulting in a gas hourly space velocity (GHSV: quotient of gas volume flow and the catalyst volume) of 35,000 h1, which is a realistic space velocity for mobile application. The base feed gas, which was used for most of the experiments, consisted of 10% O2 + 5% H2O, with nitrogen as the balance. For the first part of the soot oxidation measurements, 0, 200 and 1000 ppm NO2 and 0, 1, 4 and 10 ppm H2SO4 were added to the base feed gas. For the second part, the H2SO4-impregnated monoliths were used, and 0, 200 and 1000 ppm NO2 was added to the base feed gas, respectively. Additionally, the soot reactivity was also investigated with either 10% O2 or 10% O2 + 5% H2O
Table 1 – H2SO4 concentrations of the H2SO4 solutions used for impregnation of the soot-loaded monoliths. H2SO4 concentration in precursor solution (mol/L) 0.060 0.030 0.010 0.006
H2SO4 on monolith (mg) 10 5 2 1
Fig. 1 – Schematic of the experimental setup for investigating the influence of H2SO4 on soot oxidation with NO2. The dotted lines symbolize communication connections of the FTIR spectrometer and the MFCs with the PC.
2102
CARBON
5 0 ( 2 0 1 2 ) 2 1 0 0 –2 1 0 9
as a comparison. All measurements were performed with a heating ramp of 10 K/min between 50 and 700 C. The gas composition was analyzed quantitatively downstream of the quartz reactor using a Nicolet Nexus FTIR spectrometer with a gas measuring cell of 2 m path length and ZnSe windows. All tubes downstream of the gas mixing unit and the FTIR cell were heated to 180 C to avoid condensation. The adsorption of H2SO4 on the ZnSe windows of the FTIR gas measuring cell did not influence the measurements because the H2SO4 concentrations were kept below 15 ppm. The spectrometer was calibrated for CO2, CO, NO2, NO, HNO3, H2O, SO2 and H2SO4. Calibration of these components, except H2SO4, was done using the standard procedure with calibration gases from gas bottles. The calibration of H2SO4 was more difficult because first a qualitative reference spectrum of H2SO4 had to be collected, and in a second step, the quantification method of the spectrometer had to be calibrated with a nitrogen-based model gas containing 10% O2 + 5% H2O + 2, 6 and 10 ppm H2SO4. For recording a reference spectrum of H2SO4, a dry nitrogen flow was saturated in a temperature-controlled wash bottle containing a concentrated H2SO4 solution. The spectrum showed absorption bands of both SO3 and H2SO4; therefore, the region between 1200 and 1260 cm1 showing only specific bands for H2SO4 [21] was used as a reference for further calibration. Quantification of the reference spectrum was performed by dosing different SO2 concentrations over the SO3 generator in the presence of 10% O2 with nitrogen as the balance. When the gas flow from the SO3 generator was mixed with the main nitrogen-based gas flow containing 5% H2O, H2SO4 was formed, which was analyzed via a wet chemical analysis method as described in [22–24]. The obtained concentrations were used for quantification with FTIR. Because the SO3/ H2SO4 equilibrium in the presence of 5% H2O at 180 C in the gas measuring cell is biased toward H2SO4, a calibration of SO3 was unnecessary. Therefore, in our study, only H2SO4 will be discussed, although under certain reaction conditions, H2SO4 and SO3 might be present in equilibrium.
3.
Results
3.1. Introductory example for the impact of H2SO4 on soot oxidation with NO2 Fig. 2 shows the trends of the gas concentrations observed during a typical soot oxidation experiment performed in this study. The inlet gas concentrations for the chosen example were 200 ppm NO2 + 10% O2 + 5% H2O + 18 ppm SO2, with nitrogen as the balance. The figure shows the outlet concentrations of CO, CO2, NO, NO2, SO2, H2SO4 and HNO3. At 170 C, the soot surface started to be oxidized by NO2 leading to the formation of oxygen surface functional groups and gaseous NO, as shown in reaction (1) where C(O) represents a surface functional group and Cbulk the carbon backbone: Cbulk C þ NO2 ! Cbulk CðOÞ þ NO
ð1Þ
Fig. 2 – Outlet concentrations of CO, CO2, NO, NO2, SO2, H2SO4 and HNO3 in a soot oxidation experiment with 200 ppm NO2 + 10% O2 + 5% H2O + SO2 + 10 ppm H2SO4 in the feed gas. The slight increase of the CO2 concentration at 180 C indicates that a part of the functional groups decomposed after a second attack by NO2: Cbulk CðOÞ þ NO2 ! Cbulk þ CO2 þ NO
ð2Þ
Since reaction (2) is a consecutive reaction of (1) it can be concluded that always more NO than CO2 is produced. In fact, Fig. 2 clearly shows that the slope of the NO concentration is more than two times steeper than the slope of the CO2 concentration between 170 and 210 C. At 10 C below the breakthrough of H2SO4, the NO2 and NO concentrations reached a plateau due to limitation by the available amount of reactive carbon structures. Above 230 C, the NO and CO2 concentrations increased again, while the NO2 concentration decreased. At 280 C, the NO2 conversion exceeded 90%, and in parallel to reaction (2), the carbon was also oxidized to CO, as indicated by the increase in the CO concentration: Cbulk C þ NO2 ! Cbulk CðOÞ þ NO
ð3Þ
At 315 C, the H2SO4 desorption reached a maximum, and the NO2 conversion reached 100% meaning that the CO and CO2 concentration should also reach a plateau if only NO2 accounts for soot oxidation. However, a rough balance showed that the amount of CO and CO2 was too large to be produced only by the dosed NO2. Therefore, it was concluded that all additional CO and CO2 formed above this temperature must be due to the decomposition of previously formed surface functional groups according to reaction (1) or due to reactions (4)–(6) with oxygen 2 Cbulk CðOÞ þ O2 ! 2 Cbulk CðOÞðOÞ ! 2 Cbulk þ 2 CO2
ð4Þ
Cbulk C þ O2 ! Cbulk CðOÞðOÞ ! Cbulk þ CO2
ð5Þ
2 Cbulk C þ O2 ! 2 Cbulk CðOÞ ! 2 Cbulk þ 2 CO
ð6Þ
With increasing temperature, the concentration of SO2, CO and CO2 increase, while the sum of the NO and NO2 concentration remained constant during the entire experiment. It is important to note that an influence of excess SO2 on the reaction was not observed. SO2 was added below 100 C, but H2SO4 did not break through until 210 C due to adsorption on the system walls and on the soot sample. NO2 was added at
CARBON
5 0 ( 20 1 2 ) 2 1 0 0–21 0 9
2103
120 C. Later we will show that the amount of added H2SO4 was too small to influence the observed carbon oxidation significantly. For the other NO2 and H2SO4 concentrations specified in the experimental section the absolute concentrations of the products changed, but the general trends were similar.
3.2.
NO2 conversion and C-oxidation rates
The influence of H2SO4 on soot oxidation with NO2 was investigated systematically in the presence of different NO2 and H2SO4 inlet concentrations by calculating the NO2 conversion and the transient C-oxidation rates from the data of the FTIR analysis. The transient C-oxidation rate was defined as the ratio of the carbon amount in the gas phase at time t (cgas,t) in [lg] (calculated from the CO, CO2, HNCO and HCN concentration in the gas phase detected in the FTIR spectrum at time t) to the carbon amount at time t (csolid,t) in [mg] remaining in the solid phase, referred to the period of time needed for the collection of a FTIR spectrum tcollect in [s] cgas;t lg 1 ¼ ð7Þ C-oxidation rate mg s csolid;t tcollect csolid,t was calculated from the total carbon amount at the beginning of the measurements, csolid,0, minus the sum of carbon desorbed up to time t. This way of calculating the C-oxidation rate is similar to a normalization of the data, which allows for a comparison of the results from different reactive gases and samples of different sizes. In comparison to C-oxidation rates referred to the initial carbon amount, transient C-oxidation rates have several advantages: They are more realistic, as the carbon conversion to CO and CO2 is referred to the actual amount of remaining carbon at the point of measurement and not to a quantity, which is only measured once at the beginning of the experiment. This results in an increasing deviation of the transient rate during carbon consumption from the traditional way of referring the oxidation rates on the initial carbon amount. Since the actual carbon amounts are considered in the denominator, the transient C-oxidation rates become significantly larger than the rates referred to the initial carbon amount especially above 400 C and are, therefore, more sensitive in this temperature range. Moreover, they directly represent the kinetics, because the rates increase with temperature and do not show a global maximum implying decreasing rates at high burn-off levels. Figs. 3 and 4 show the NO2 conversion during soot oxidation with 200 and 1000 ppm NO2 in the feed gas, respectively. Fig. 3 depicts the results from the addition of H2SO4 via the gas phase between 150 and 700 C, and Fig. 4 shows the results from adding H2SO4 by impregnation between 150 and 700 C. As a basis for comparison, the NO2 conversion in the presence of 200 and 1000 ppm NO2 without H2SO4 in the feed gas was included as reference in the figures. The conversion rates were plotted up to 99.5% carbon conversion. When 200 ppm NO2 without H2SO4 were added to the feed gas, the NO2 conversion increased with temperature from 4% at 200 C up to 100% at 450 C. This conversion level was kept until the main part of the soot sample was oxidized around
Fig. 3 – Increase in the NO2 conversion rates for the soot oxidation experiments in the presence of NO2 and H2SO4 between 150 and 700 C. Feed gas: 10% O2 + 5% H2O + 200 ppm NO2 + 0, 1, 4, 10 ppm H2SO4 and 10% O2 + 5% H2O + 1000 ppm NO2 + 0, 10 ppm H2SO4. The NO2 conversion is plotted until 99.5% carbon conversion.
Fig. 4 – Increase in the NO2 conversion rates for the soot oxidation experiments in the presence of NO2 and H2SO4 between 150 and 700 C. Feed gas: 10% O2 + 5% H2O + 200 ppm NO2 + 0, 1, 5, 10 mg H2SO4 and 10% O2 + 5% H2O + 1000 ppm NO2 + 0, 10 mg H2SO4. The NO2 conversion is plotted until 99.5% carbon conversion. 560 C, which led to a decrease of the NO2 conversion. By the addition of 1 ppm H2SO4 to the feed gas the base NO2 conversion increased from 4% to 6%. Above 200 C, NO2 conversion increased faster than without H2SO4 up to 320 C. Above 320 C the slope became smaller, most probably due to limitation of the available amount of carbon structures, which are reactive enough to be oxidized in this temperature range. However, a further increase of temperature resulted in a steeper slope again and full conversion was reached at 450 C, where even less reactive carbon structures could be rapidly oxidized by NO2. When the H2SO4 concentration was further increased to 4 and 10 ppm the base conversion increased only little. However, above 180 C NO2 conversion increased significantly faster than in the presence of 1 ppm H2SO4. The plateaus at 230 C were again due to limitation of the available amount
2104
CARBON
5 0 ( 2 0 1 2 ) 2 1 0 0 –2 1 0 9
of reactive carbon structures. For 10 ppm H2SO4 full NO2 conversion was already reached at 320 C, while in the presence of 4 ppm H2SO4 a slight flattening of the curve was observed before full conversion was achieved at 400 C. Similar to the experiment without H2SO4, the NO2 conversion decreased when the soot was nearly totally gasified. For 1000 ppm NO2 full conversion was not reached because under the applied conditions, the reaction was limited by the available amount of reactive carbon structures over the whole temperature range. As expected, total carbon oxidation was achieved at significantly lower temperatures in the presence of 1000 ppm NO2 than in the presence of 200 ppm NO2 due to the higher amount of oxidant. The addition of 10 ppm H2SO4 to the feed gas containing already 1000 ppm NO2 resulted in an increase of the NO2 conversion at lower temperatures. The first maximum was also due to limitation of the available amount of reactive carbon structures. The trends of the NO2 conversions rates for the experiments with the H2SO4-impregnated samples were similar to those obtained for the addition of H2SO4 via the gas phase (Fig. 3). However, full NO2 conversion was not achieved in the experiments with the impregnated samples. This was most probably due to the larger amounts of H2SO4, which were present at the beginning of the experiments with the H2SO4-impregnated samples compared to the experiments with H2SO4 in the gas phase, leading to a limitation of the available amount of reactive carbon structures also in this case. In Figs. 5–8 the carbon oxidation (C-oxidation) rates of Printex U are shown for different reaction conditions. The C-oxidation rates for the experiments with 10% O2 + 5% H2O + NO2 were significantly higher than those seen in experiments without NO2. For example, in Fig. 5, the C-oxidation rate observed at 300 C in the absence of NO2 had already been reached at 200 C in the presence of 200 ppm NO2. By increasing the NO2 concentration to 1000 ppm, these differences were enhanced (Fig. 7). As already expected from the trends of the NO2 conversion rates, it was possible to further increase the level of carbon
Fig. 5 – C-oxidation rates for soot oxidation experiments done in the presence of NO2 and H2SO4. Reactive gas compositions: 10% O2 + 5% H2O + 0 and 200 ppm NO2 + 0, 1, 4 and 10 ppm H2SO4. The C-oxidation rates are plotted until 99% carbon conversion.
oxidation in the presence of NO2 by adding small amounts of H2SO4. Adding 1 ppm H2SO4 enhanced the C-oxidation at 300 C by 75%, and the addition of 4 and 10 ppm H2SO4 enhanced the rate by 120% (Fig. 5). For 1000 ppm NO2, the addition of 1 ppm H2SO4 at 300 C did not enhance the C-oxidation, but at levels of 4 and 10 ppm, the rates were increased by a factor of 3 and 4, respectively (Fig. 7). For 200 ppm NO2 (Fig. 5) the C-oxidation rates approached each other above 410 C regardless of the reactive gas. For 1000 ppm NO2 (Fig. 7), the C-oxidation rates did not converge, and with 4 and 10 ppm H2SO4 in the reactive gas, the soot sample was totally gasified at a temperature 190 C lower compared to the experiments performed without NO2 and H2SO4. Similar results were obtained for oxidation of the soot samples impregnated with H2SO4 before the measurement (Figs. 6 and 8). The carbon oxidation rate was enhanced when NO2 was passed over the H2SO4-impregnated soot samples. In these figures, the C-oxidation rates seen with 0 ppm NO2 + 10% O2 + 5% H2O for a Printex U sample impregnated with 10 mg H2SO4 were added. These results do not show a significant difference between the other experiments without NO2 up to 450 C. Above 450 C, the C-oxidation rates increased slightly.
3.3.
HNO3 evolution
The trend of the HNO3 concentration was also investigated in our study because HNO3 is assumed to be an intermediate of soot oxidation in the presence of H2O and NO2 [25] and can be formed by a heterogeneous disproportionation reaction of NO2 (Eq. (4)), which was confirmed by several studies at RT concerning atmospheric chemistry [26–28]: 2 NO2 þ H2 O ! HNO3 þ HNO2
ð8Þ
HNO2 was not calibrated for these experiments, but its formation on Printex U was detected during experiments that were performed subsequent to this study under similar con-
Fig. 6 – C-oxidation rates for soot oxidation experiments done in the presence of NO2 and H2SO4. Reactive gas compositions: 10% O2 + 5% H2O + 0 and 200 ppm NO2 + 0, 1, 5 and 10 mg H2SO4. The C-oxidation rates are plotted until 99% carbon conversion.
CARBON
5 0 ( 20 1 2 ) 2 1 0 0–21 0 9
Fig. 7 – C-oxidation rates for soot oxidation experiments done in the presence of NO2 and H2SO4. Reactive gas compositions: 10% O2 + 5% H2O + 0 and 1000 ppm NO2 + 0, 1, 4 and 10 ppm H2SO4. The C-oxidation rates are plotted until 99% carbon conversion.
Fig. 8 – C-oxidation rates for soot oxidation experiments done in the presence of NO2 and H2SO4. Reactive gas compositions: 10% O2 + 5% H2O + 0 and 1000 ppm NO2 + 0, 1, 2, 10 mg H2SO4. The C-oxidation rates are plotted until 99% carbon conversion.
ditions. In the example shown in Fig. 2, only a maximum of 1 ppm HNO3 was detected between 180 and 220 C, but higher amounts of HNO3 were detected for higher NO2 concentrations, as shown in Figs. 9 and 10. For the experiment without H2SO4 in the feed gas, the formation of HNO3 was below the detection limit when only 200 ppm NO2 were dosed, but HNO3 formation increased significantly when the NO2 concentration was increased to 1000 ppm (Fig. 9). The addition of 10 ppm H2SO4 to the reactive gas led to the formation of approximately 1 ppm HNO3 at 210 C in the presence of 200 ppm NO2. In the presence of 1000 ppm NO2, the HNO3 concentration between 120 and 180 C was similar to the amounts seen for the experiment without H2SO4 in the reactive gas, but between 180 and 220 C, a peak of 8 ppm HNO3 could be observed. The addition of H2SO4 by impregnation had an even more pronounced effect (Fig. 10). For both NO2 inlet concentrations, more HNO3 was formed when the H2SO4 was impregnated as
2105
Fig. 9 – HNO3 produced during soot oxidation in the presence of NO2 and H2SO4. Feed gas composition: 200 and 1000 ppm NO2 + 0 and 10 ppm H2SO4.
Fig. 10 – HNO3 produced during soot oxidation in the presence of NO2 and H2SO4. Feed gas composition: 200 and 1000 ppm NO2 + 0 and 10 mg H2SO4.
compared to the addition of H2SO4 via the gas phase. Between 120 and 280 C, maximum HNO3 concentrations of 2 and 22 ppm were obtained for 200 and 1000 ppm NO2, respectively. For the final carbon gasification phase above 480 C, up to 2.5 ppm HNO3 was measured for the experiments with 1000 ppm NO2 in the reactive gas. For 200 ppm NO2, this increase was below 1 ppm, which was within the error of measurement. Before the discussion of the results of this study, it should be mentioned that the results obtained with the H2SO4-impregnated samples could not be quantitatively compared to the results of the experiments with H2SO4 in the feed gas. In the former case, a significant amount of H2SO4 was present at the start of the experiment, while in the latter case, the H2SO4 amount on the sample increased due to the SO3 dosing through the gas phase. However, the general qualitative trends observed were similar.
3.4.
Discussion
The results from this study again confirm that the addition of NO2 to the reactive gases containing O2 and water results in
2106
CARBON
5 0 ( 2 0 1 2 ) 2 1 0 0 –2 1 0 9
an increase in the carbon oxidation rate over the whole temperature range; this has already been described by several authors [7,8,10,11,25,29–33]. In addition, it was observed that the C-oxidation rates could be further increased when traces of H2SO4 were added to the reactive gas containing O2, H2O and NO2 (Figs. 5–8). This is in accordance with the results of Oi-Uchisawa et al. and Obuchi et al., who also observed a significant increase in the C-oxidation rate in the presence of NO2, which could be further enhanced by the addition of small amounts of H2SO4 [15–18]. However, in contrast to our experiments, Oi-Uchisawa et al. and Obuchi et al. used a mixture of a commercially available carbon black and a Pt catalyst for their investigations [15–18]. In their case, the Pt catalyst was needed for the in situ generation of NO2 and H2SO4 from NO and SO2. In addition to NO and SO2, their reactive gases also contained O2 and H2O, with nitrogen as the balance. Based on their findings, Oi-Uchisawa et al. proposed a mechanism that explains this enhancement [18]. The authors suggested that first phenols and carboxylates are formed during the partial oxidation of the carbon surface with NO2. Further oxidation of these groups with NO2 on the carbon surface would lead to their decomposition, but their electronwithdrawing character decreased the reactivity of neighboring carbon sites. The addition of H2SO4 could increase the C-oxidation rate because its strong acidity is assumed to catalyze the decomposition of the surface functional groups formed by the first attack of NO2. Furthermore, they suggested that H2SO4 could hydrolyze esters and anhydrides, leading to the decarboxylation of the hydrolysis products. After the decomposition of the oxygenated surface species, the carbon surface was ‘‘free’’ for a further oxidation cycle with NO2. The results of our work confirmed the assumptions of OiUchisawa et al. that a co-operative mechanism between H2SO4 and NO2 exists, where H2SO4 behaves more like a catalyst than a stoichiometric reactant (i.e., even traces of H2SO4 have a promoting effect). The catalytic effect of H2SO4 on the oxidation capability of NO2 was supported by our finding that, below 480 C, H2SO4 could not enhance the C-oxidation rates in the absence of NO2 (as seen in Figs. 6 and 8). Although the explanation in [18] on the influence of H2SO4 on the carbon oxidation seems to be plausible, it is questionable if the acceleration of the decarboxylation and of hydrolysis
of anhydrides and esters were the only processes responsible for the increase in the C-oxidation rates. We suggest instead (or in addition) that H2SO4 also influences carbon oxidation at the nitration of the graphitic soot structure, which is most likely one of the major steps of the soot oxidation process. It is well known that the active species during the nitration of aromatics is not NO2 or HNO3 but the more electrophilic NOþ 2 (nitronium-ion) [34,35], which is formed in traces in concentrated HNO3 solutions. HNO3, which was most likely formed according to Eq. (4), was also found during experiments in our study. Although the HNO2 concentration was not followed in this study, its formation on Printex U was confirmed by a subsequent study under similar conditions. While HNO2 decomposed quickly into NO, NO2 and H2O [36], HNO3 probably formed small amounts of NOþ 2 according to route A shown in Fig. 11. Although we found HNO3 in larger amounts in the gas phase at temperatures only up to 250 C, it was most likely formed on the soot surface over the entire temperature range and was not detected at higher temperatures because the rate of desorption was smaller than the rate of reaction. This assumption was confirmed by the HNO3 traces measured in the gas phase above 500 C (Figs. 9 and 10), which were formed during experiments with H2SO4 when up to 95% of the carbon was already burned. In this case, the rate of desorption was again higher than the rate of reaction. Furthermore, it is known that deactivated aromatics can also be nitrified when nitrosulfuric acid – a mixture of concentrated HNO3 and H2SO4 – is used, because the NOþ 2 concentration increases in the presence of a strong acid [34,35]. We suggest that the addition of H2SO4 to the reactive gas containing NO2 and H2O also promoted these favorable conditions for the formation of NOþ 2 ions on the soot surface. The strong acidity of H2SO4 increased the NOþ 2 concentration by protonation of HNO3, as shown in route B of Fig. 11, compared to the self-protonation pathway shown in route A. After nitration, the C–NO2 bond was most probably either homolytically cleaved, resulting in an ‘‘aryl radical’’ and NO2, or the nitro group was rearranged to a nitrite [37–41], which decomposed homolytically into a ‘‘phenyl radical’’ and NO. The ‘‘aryl radical’’ can react with HOOÆ radicals, formed by hydrogen abstraction from the carbon surface with O2, or directly with NO2 [42] resulting also in ‘‘phenyl
Fig. 11 – Mechanism for the formation of NOþ 2 [34]. (A) Reaction route in conc. HNO3 and (B) reaction route in nitrosulfuric acid.
CARBON
5 0 ( 20 1 2 ) 2 1 0 0–21 0 9
radicals’’. Electron rearrangement of the ‘‘phenyl radical’’ leads to the formation of a ‘‘cycloalkyl radical’’ and a carbonyl group and by oxidation of the radical with NO2 or HOOÆ to ringopening as shown in Fig. 12. The carbonyl group formed in the last step of Fig. 12 might decompose either directly under decarbonylation or can be further oxidized to a carboxylic group decomposing under decarboxylation. The ‘‘carbonyl radical’’ formed in the last step of Fig. 12 might also be oxidized to a carboxylic group or decomposes under decarbonylation leaving an ‘‘alkyl radical’’ on the soot surface. We assume this ‘‘alkyl radical’’ to be a potential active centre for the continuation of a radical chain reaction consisting of oxidation by HOÆ, HOOÆ, O2 or NO2, which is either followed by direct decarbonylation or a second oxidation step followed by decarboxylation. Based on the analytical results and considerations, it is reasonable to conclude that an increase in the nitration rate also enhanced the carbon oxidation. As already speculated, the H2SO4 acted catalytically as a proton donator during the formation of NOþ 2 , which could be regenerated after the nitration step. In addition to enhancing the formation of NOþ 2 , Figs. 9 and 10 show that the H2SO4 also enhanced the formation of HNO3. This can be explained by a closer look at the mechanism behind reaction (8). First, NO2 is dimerized to N2O4 (Eq. (9)), which disproportionates into HNO2 and HNO3 (Eq. (10)) in a second step in the presence of water. With increasing temper-
2107
ature, the equilibrium of the dimerization step is shifted to the NO2 side. In the presence of H2SO4, it is assumed that the HNO3 concentration increases because H2SO4 catalyzed the decomposition of N2O4, as illustrated in Fig. 13 by protonation of N2O4, facilitating the attack of water. Therefore, H2SO4 catalyzes not only the formation of NOþ 2 but also the formation of HNO3 – the NOþ 2 precursor 2 NO2 ! N2 O4 N2 O4 þ H2 O ! HNO3 þ HNO2
ð9Þ ð10Þ
Beside the nitration of aromatics, there might be additional non-radicalic reactions of surface groups in the presence of NO2, which can only be explained if NOþ 2 is the active species. One example for such a reaction is the oxidation of aldehydes with NO2, which proceeds via an acylnitrite [19]. Based on an analysis of the oxidation states, it is clear that a non-radicalic electrophilic attack via oxidation of the carbonyl carbon is only possible if the electrophile is NOþ 2, as shown in Fig. 14. In a subsequent step the acylnitrite most probably decomposes into NO and a COOÆ radical as already indicated by results of Muckenhuber et al. [19]. Supported by the theoretical investigations of Barco et al. [43] we assumed the COOÆ to decompose under decarboxylation leaving an ‘‘alkyl radical’’ on the surface, which reacts similar to the ‘‘alkyl radical’’ formed from the ‘‘carbonyl radical’’ in Fig. 12. A direct oxidation of soot by H2SO4 also has to be considered because an increase in the SO2 concentration in the product gas was observed with increasing temperature, even
Fig. 12 – Decomposition of the ‘‘phenyl radical’’. The benzene ring symbolizes one aromatic ring of the soot backbone.
Fig. 13 – Mechanistic details of the HNO3 and HNO2 formation by N2O4 decomposition over soot.
2108
CARBON
5 0 ( 2 0 1 2 ) 2 1 0 0 –2 1 0 9
Fig. 14 – Example reactions of surface groups with NOþ 2 : the oxidation of aldehydes to nitrites. R stands for the carbon backbone.
though the amount of SO2 added to the SO3 generator remained constant. However, the amounts of H2SO4 were too small to significantly contribute to the C-oxidation rates observed at temperatures up to 450 C (Fig. 6) Above 450 C, a slight increase in the C-oxidation rate was observed (Figs. 6 and 8), which corresponds to the formation of SO2 in Fig. 2. The results from the impregnated soot samples and those oxidized with the addition of H2SO4 in the gas phase showed that the differences were mainly attributed to differences in the actual amounts of H2SO4. The impregnated samples were more active at 120 C because the amount of H2SO4 was larger in this temperature range. During gas phase addition, adequate amounts of H2SO4 had to be accumulated before an increase in the C-oxidation rates was observed.
4.
Conclusions
The results from this study clearly show that traces of H2SO4 resulting from oxidized fuel sulfur have a significant promoting effect on soot oxidation via NO2. Based on a detailed infrared gas phase analysis, a new mechanism has been suggested to explain the activation effect of H2SO4; it most probably acts in parallel to the already published activation mechanisms. In our new mechanism, H2SO4 acts as a catalyst at two stages. First, it enhances the formation of HNO3 by catalyzing the disproportionation of N2O4, and secondly, H2SO4 most likely increases the amount of NOþ 2 , which is the known active species for the nitration of aromatic structures. Once the aromatic system is nitrified, the C–N bond is either cleaved homolytically or the nitro group is rearranged to a nitrite, which also decomposes homolytically to NO and a ‘‘phenyl radical’’. These radical species are further oxidized, leading to ring-opening and decomposition of the aromatic system by decarboxylation and decarbonylation. For applications with high-sulfur fuels, a significant influence of the soot oxidation by the sulfur is expected. However, we have clearly shown that even very small sulfur concentrations (below 10 ppm), as found in modern low-sulfur diesel fuels, have a strong catalytic effect on soot oxidation. Moreover, the situation is aggravated by the fact that SO3/H2SO4 are accumulated in the DOCs under low load conditions and can be released quickly in high concentrations, when the
exhaust temperature increases [44]. After SO3/H2SO4 has been released from the DOC they enter the DPF and react with the accumulated soot as shown in this study.
R E F E R E N C E S
[1] Bu¨nger J, Krahl J, Baum K, Schro¨der O, Mu¨ller M, Westphal G. Cytotoxic and mutagenic effects, particle size and concentration analysis of diesel engine emissions using biodiesel and petrol diesel as fuel. Arch Toxicol 2000;74(8):490–8. [2] Yamaguchi T, Yamazaki H. Cytotoxicity of airborne particulates sampled roadside in rodent and human lung fibroblasts. J. Health Sci. 2001;47(3):272–7. [3] Gazette of the European Union, EU, Regulation No. 715/2007. [4] Tschoeke H, Graf A, Stein J, Krueger M, Schaller J, Breuer N et al. Diesel engine exhaust emissions. In: Mollenhauer K, Tschoeke H, editors. Handbook of diesel engines. Berlin, Heidelberg; Springer; 2010. p. 417–85. [5] Twigg MV. Progress and future challenges in controlling automotive exhaust gas emissions. Appl Catal B-Environ 2007;70(1–4):2–15. [6] Neeft JPA, Makkee M, Moulijn JA. Diesel particulate emission control. Fuel Process Technol 1996;47(1):1–69. [7] Ehrburger P, Brilhac JF, Drouillot Y, Logie V, Gilot P. Reactivity of soot with nitrogen oxides in exhaust stream. SAE 2002–011683. [8] Jacquot F, Logie V, Brilhac JF, Gilot P. Kinetics of the oxidation of carbon black by NO2: influence of the presence of water and oxygen. Carbon 2002;40(3):335–43. [9] Muckenhuber H, Grothe H. The reaction between soot and NO2 – investigation on functional groups using TPD-MS. Top Catal 2004;30–31(1):287–91. [10] Setiabudi A, Makkee M, Moulijn JA. The role of NO2 and O2 in the accelerated combustion of soot in diesel exhaust gases. Appl Catal B-Environ 2004;50(3):185–94. [11] Stanmore BR, Tschamber V, Brilhac JF. Oxidation of carbon by NOx, with particular reference to NO2 and N2O. Fuel 2008;87(2):131–46. [12] Stanmore BR, Brilhac JF, Gilot P. The oxidation of soot: a review of experiments, mechanisms and models. Carbon 2001;39:2247–68. [13] Jeguirim M, Tschamber V, Brilhac JF, Ehrburger P. Interaction mechanism of NO2 with carbon black: effect of surface oxygen complexes. J Anal Appl Pyrol 2004;72(1):171–81.
CARBON
5 0 ( 20 1 2 ) 2 1 0 0–21 0 9
[14] Muckenhuber H, Grothe H. The heterogeneous reaction between soot and NO2 at elevated temperature. Carbon 2006;44(3):546–59. [15] Obuchi A, Ohi A, Aoyama H, Ohuchi H. Evaluation of gaseous and particulate emission characteristics of a single cylinder diesel engine. Combust Flame 1987;70(2):215–24. [16] Oi Uchisawa J, Obuchi A, Zhao Z, Kushiyama S. Carbon oxidation with platinum supported catalysts. Appl Catal BEnviron 1998;18(3–4):L183–7. [17] Oi-Uchisawa J, Obuchi A, Enomoto R, Ogata A, Kushiyama S. Outstanding effect of SO2 addition on the rate of carbon oxidation with a Pt catalyst. Chem Commun 1998;20:2255–6. [18] Oi-Uchisawa J, Obuchi A, Ogata A, Enomoto R, Kushiyama S. Effect of feed gas composition on the rate of carbon oxidation with Pt/SiO2 and the oxidation mechanism. Appl Catal BEnviron 1999;21(1):9–17. [19] Muckenhuber H, Grothe H. A DRIFTS study of the heterogeneous reaction of NO2 with carbonaceous materials at elevated temperature. Carbon 2007;45(2):321–9. [20] Neeft JPA, Nijhuis TX, Smakman E, Makkee M, Moulijn JA. Kinetics of the oxidation of diesel soot. Fuel 1997;76(12):1129–36. [21] Dene C, Himes R. Continuous measurement technologies for SO3 and H2SO4 in coal-fired power plants, EPRI, 1009812. Palo Alto, CA; 2004. [22] Koebel M, Elsener M. Schwefeltrioxidbestimmung in Abgasen nach der isoporopanolmethode – Eine kritische Betrachtung. Gefahrstoffe – Reinhaltung der Luft 1997;57:193–9. [23] Messen der Schwefeltrioxid-Konzentration. Wasserstoffperoxid-Verfahren, Titrimetrische Bestimmmungen, VDI 2462 Bl. 2. Berlin: Beuth-Verlag; 1974. [24] Messen der Schwefeltrioxid-Konzentration. 2-PropanolVerfahren, VDI 2462 Bl. 7. Berlin: Beuth-Verlag; 1985. [25] Jeguirim M, Tschamber V, Brilhac JF, Ehrburger P. Oxidation mechanism of carbon black by NO2: effect of water vapour. Fuel 2005;84(14–15):1949–56. [26] Akimoto H, Bandow H, Sakamaki F, Inoue G, Hoshino M, Okuda M. Photooxidation of the propylene-NOx-air system studied by long-path Fourier transform infrared spectrometry. Environ Sci Technol 1980;14(2):172–9. [27] Kleffmann J, Becker KH, Wiesen P. Heterogeneous NO2 conversion processes on acid surfaces: possible atmospheric implications. Atmos Environ 1998;32(16):2721–9. [28] Sakamaki F, Shiro H, Akimoto H. Formation of nitrous acid and nitric oxide in the heterogeneous dark reaction of nitrogen dioxide and water vapor in a smog chamber. Int J Chem Kinet 1980;15:1013–29. [29] Chughtai AR, Gordon SA, Smith DM. Kinetics of the hexane soot reaction with NO2/N2O4 at low concentration. Carbon 1994;32(3):405–16.
2109
[30] Lur’e BA, Mikhno AV. Interaction of NO2 with soot. Kinet Catal 1995;38(4):535–40. [31] Messerer A, Niessner R, Po¨schl U. Comprehensive kinetic characterization of the oxidation and gasification of model and real diesel soot by nitrogen oxides and oxygen under engine exhaust conditions: measurement, Langmuir– Hinshelwood, and Arrhenius parameters. Carbon 2006;44(2):307–24. [32] Shirahama N, Moon SH, Choi KH, Enjoji T, Kawano S, Korai Y, et al. Mechanistic study on adsorption and reduction of NO2 over activated carbon fibers. Carbon 2002;40(14):2605–11. [33] Stadler D, Rossi MJ. The reactivity of NO2 and HONO on flame soot at ambient temperature: the influence of combustion conditions. Phys Chem Chem Phys 2000;2(23):5420–9. [34] Bru¨ckner R. Reaktionsmechanismen. 2nd ed. Heidelberg: Springer; 2003. [35] Smith MB, March J. March’s advanced organic chemistry. 5th ed. New York: Wiley; 2001. [36] Hollemann AF, Wiberg E. Hollemann–Wiberg Lehrbuch der anorganischen Chemie. Berlin, New York: Walter de Gruyter; 1995. [37] He YZ, Cui JP, Mallard WG, Tsang W. Homogeneous gas-phase formation and destruction of anthranil from o-nitrotoluene decomposition. J Am Chem Soc 1988;110(12):3754–9. [38] Nikolaeva EV, Shamov AG, Chackov DV, Gordeev DA, Khraphovsky GM. The role of nitro-nitrite rearrangement in the mechanism of monomolecular decomposition of aliphatic nitro compounds. Butlerov Commun 2001;3:15–23. [39] Pitz WJ, Westbrook CK. A detailed chemical kinetic model for gas phase combustion of TNT. Proc Combust Inst 2007;31(2):2343–51. [40] Poon C, Mayer PM. Competing rearrangement reactions in small gas-phase ionic complexes: the internal SN2 and nitronitrite rearrangements in nitroalkane proton-bound pairs. Int J Mass Spectrom 2006;255–256:93–101. [41] Tsang W, Robaugh D, Mallard WG. Single-pulse shock-tube studies on C–NO2 bond cleavage during the decomposition of some nitro aromatic compounds. J Phys Chem 1986;90(22):5968–73. [42] Otsuka K, Takahashi R, Amakawa K, Yamanaka I. Partial oxidation of light alkanes by NOx in the gas phase. Catal Today 1998;45(1–4):23–8. [43] Barco G. The oxidized soot surface: theoretical study of desorption mechanisms involving oxygenated functionalities and comparison with temperature programed desorption experiments. J Chem Phys 2006;125(19): 194706 (1–12). [44] Kroecher O, Widmer M, Elsener M, Rothe D. Adsorption and desorption of SOx on diesel oxidation catalysts. Ind Eng Chem Res 2009;48(22):9847–57.