CC BY
11УДК 544.478.1 В.Я. Супрун1
МЕЗО/МАКРО-ПОРИСТЫЕ Ag/Al2O3 КАТАЛИЗАТОРЫ ДЛЯ СЕЛЕКТИВНОГО КАТАЛИТИЧЕСКОГО ВОССТАНОВЛЕНИЯ NOX В ПРИСУТСВИИ ЭТАНОЛА
Институт Технической Химии, Университет Лейпциг, Линне Штрассе 3, г. Лейпциг, 04103, Германия e-mail: wladimir.suprun@uni-leipzig.de
Лд/Л12Оз катализаторы с мезо- и мезо макропористой структурой синтезированы методом импрегрирования промышленного Y-алюминий оксида (AO-C) и иерархически структурированного золь-гель Y-алюминий оксида (ЛО-SG) и исследованы в реакции селективного каталитического восстановления NOx в присуттсвии этанола. EtOH-SCR экспрерименты свидетельствуют о том, что Лд/ЛО-SG катализатор обладает постоянной каталитической активностью и высокой селективностью в температурном интервале 250-320°C. В отличие, Ag/AO-C катализатор вызывает непродуктивное каталитическое окисление этанола и подвергается деактивации в условиях продолижительной эксплуатации при тепературах ниже 270°C. Исследованый эффект размеров частиц катализаторов в SCR реакции свидельствует о наличии внутренних диффузионных ограничений для Ag/AO-C катализатора. Природа отложений на деактивировах катализторв ислледована методами температурно-программированно окисления, FT-IR, ESI-MS спектроскопии и термо-гравиметрического анализов.
Ключевые слова: ЛдОх/Л12О3, катализатор , SCR, NOx, этанол.
Introduction
Nitrogen oxides (NO and NO2) are amongst the major air pollutants that cause great harm to the ecological environment and human health. The selective catalytic reduction (SCR) of NOx in the presence of ammonia and excess oxygen is a well-proven and widely applied technology for NOx reduction (DeNOx) from the exhaust of stationary power plants and diesel engines [1]. The SCR of NOx with hydrocarbons or lower (C2-C4) bio-alcohols has been intensively studied as alternative technology that could be used in place of the current commercially applied NH3-SCR processes [2, 3]. As bio-ethanol or bio-butanol can be mixed to diesel fuel, such technology could be applied to clean flue gases from lean-burn engines. Moreover, the use of bio-ethanol would improve the environmental impact of such approach. From the catalysts proposed in literature [2-4] for EtOH-SCR, exclusive silver containing ones proved to be promising, as they possessed both high SCR activity, as well as resistance towards water and sulfur dioxide [5].
Wladimir Suprun 1
MESO/MACRO-POROUS Ag/Al2O3 CATALYSTS FOR SELECTIVE CATALYTIC REDUCTION OF NOX BY ETHANOL
Institute of Chemical Technology, Universität Leipzig, Linnestraße 3, 04103 Leipzig, Germany e-mail: wladimir.suprun@uni-leipzig.de
The investigation of Selective Catalytic Reduction (SCR) of NOx in the presence of ethanol (EtOH) was performed with Ag/Al2Ü3 catayyst with mesoporous and meso-macroporous structure, prepared by impregnation of commercial y-alumina (AO-C) and hierarchical structured sol-gel alumina (AO-SG). The EtOH-SCR tests showed that Ag/AO-SG catalyst possesses a persistently high activity and selectivity at temperature range of 250-320°C. In contrast, Ag/AO-C catalyst causes undesirable ethanol oxidation and is subject to deactivation in the long-term run at temperature below 270°C. Investigations of the effect of partccle size on the SCR reaction indicated the presence of internal diffusion limitation for mesoporous Ag/AO-C catalyst. The nature of the catalyst deposit was investigated by temperature programmed oxidation, FT-IR, ESI-MS and thermogravimetric analysis. Possible reactions' pathways for the catalyst deactivation during EtOH-SCR were discussed.
Keywords: AgO^AbOs, catalyst, SCR, NOx, ethanol.
Nevertheless, many open questions remain unanswered regarding the details of the mechanism of this catalytic system. In particular, the reason for the low catalytic activity and catalyst deactivation at low temperature range below 300 °C has not been identified. The low-temperature activity is highly desirable for removing NOx from diesel exhaust and especially for the so-called coolstart and stop and go operation modes. However, little attention has been paid to the influence of mass transfer properties on the EtOH-SCR efficiency. In order to overcome these difficulties, alumina based mixed oxide catalysts with hierarchical pore structure which should facilitate the diffusion of reactants and products were proposed [6-7]. The aim of this paper is to clarify the catalytic behavior of hierarchical structured silver-alumina catalysts by EtOH-SCR reaction. For this purpose, the influence of the physic-chemical characteristics of hierarchically structured macro-mesoporous silver-alumina catalyst on the activity in EtOH-SCR are investigated in comparison with the mesoporous silver-alumina catalyst supported on commercial mono-modal y-alumina.
1. Wladimir Suprun, Dr. of Sci.. (habil.), Senior Research Associate; Private Lecturer, Head of Laboratory of Heterogeneous Catalysis, Institute of Chemical Technology, Universität Leipzig, e-mail: wladimir.suprun@uni-leipzig.de
Владимир Ярославович Супрун д-р хим. наук, ст. науч. сотр., доцент, зав. лаборатории гетерогенного катализа, Институт Технической Химии, Университет Лейпциг
Дата поступления 27 ноября 2018
Experimental
Catalysts preparation. Silver alumina catalysts based on two different support materials, commercial (mesoporous i.e. mono-modal) and hierarchically structured sol-gel alumina (meso-macro, i.e. bimodal), were used. The content of Ag for silver-alumina catalyst was 3wt. % which was reported to be the optimal composition of Ag/Al2O3 catalyst for EtOH-SCR [2,3]. The commercial y-Al2O3 (Alfa-Aesar, AreaBET: 257 m2 g-1, mean pore diameter 12 nm) was loaded by impregnation with aqueous solutions of AgNO3 to achieve a metal loading of 3.0 wt. %. It was subsequently dried and calcined at 520 °C for 5 h. The resulting samples were labelled AO-C and 3Ag/AO-C, where C indicates commercial Y-alumina. The hierarchically structured y -alumina were prepared based on the procedure described by Tokodume et al. [6]. The dried alumina gel was calcined for 4 h at 520 °C and loaded by impregnation with an aqueous AgNO3 solution, then dried at 100°C for 12 h and calcined at 520 °C for 5 h. The supports and silver-alumina catalysts containing 3.0 wt. % Ag were labeled AO-SG and 3Ag/AO-SG, respectively.
Catalytic activity measurements .The catalytic activity tests were carried out in a fixed bed quartz glass reactor (L: 150 mm; ID: 6 mm) at atmospheric pressure. Before the activity test, the catalyst sample was treated at 300°C in O2/He flow for 60 min. Gas feed for EtOH-SCR reaction consisted of 500 ppm NO, 950 ppm EtOH, 6.5 vol. % O2 and 9.0 vol. % water stream and He as carrier gas. The gas space velocity was fixed at 31.000 h-1. The ethanol-water solution was adjusted by an evaporator (W-101-930-P, Bronkhorst High-Tech). The output gas was monitored on-line by an FTIR analyzer (MKS Multigas 2030). The conversion of the NO and EtOH was calculated by equation (1):
x( i ) =C(V)iniet 7 c( * ) °"'tet ■ 1 o o % (1)
c( 0 in I e t
where CfiJn/et and C(i)oulet are the concentrations of NO or EtOH in the input-feed and the reaction mixture, respectively. The observed reaction rates related to NO conversion (rNOobs/mol g-1 s-1) were calculated by multiplying the NO feed rate per gram of catalyst in the reactor by the conversion of NO at steady state at the corresponding reaction temperature (eq. 2):
rN0 = ^ 1 0 0 (2)
where FNO, XNO and W are the NO molar flow rate (mol s-1), the conversion of NOx (Eq. (1) and the catalyst weight (g), respectively. The selectivity of EtOH-SCR reaction (3) was defined according to the fraction of reductant that was used to reduce NO to N2, i.e. the ratio of the consumption rate of ethanol for the NO reduction (equation 4).
S CR S e 1 e C t ivi ty = C(N0 )in'et 'XNO 1 o 0 % (3}
6 C(Et0H)tniet -XEtOH
6 NO + EtOH + 3,5O2 - 3 N2 + 2CO2+ 6H2O (4)
The value 6 in denominator of eq. 3 represents the stoichiometric coefficient for NO in the target SCR reaction (eq. 4) [2, 3].
Characterization of the spent catalysts after SCR-EtOH. Different characterization techniques were used to determine the physico-chemical properties of the spent or deactivated 3Ag/AO-C and 3Ag/AO-SG. They include UV-vis spectroscopy (speciation of silver species), thermogra-
vimetry (TGA/DTA), temperature programmed oxidation with online mass spectroscopy (MS) analysis, temperature programmed reduction with hydrogen (H2-TPR), nitrogen sorption, ICP-EOS, C/N/H analysis for the content of silver and organic deposit (ES-MS) and spectroscopy FTIR.
Results and Discussion
EtOH-SCR activity test To investigate the effect of the nature of the alumina support on the performance of silver-alumina catalysts in EtOH-SCR reaction, activity tests were performed over 3Ag/AO-C, 3Ag/AO-SG and pure alumina supports AO-C and AO-SG in the temperature range between 150 °C and 500 °C. As shown in Fig.1, the conversion of EtOH and NO differ significantly for both the catalysts and for different alumina supports as well.
3 Ag/AO-SG —■— 3 Ag/AO-C —AO-SG —*—AO-C
100 200 300 400 500 100 200 300 400 500
Temperature / °C Temperature / °C
Figure 1. Comparison of NO (a) and EtOH (b) conversion during EtOH/SCR reaction over silver-alumina catalysts and corresponding alumina supports in the temperature range between 150 and 510 °C.
The former is in accordance with literature which reported that the y -alumina support possesses the so-called "background" activity in the EtOH-SCR process [2, 8]. The loading of alumina with silver improved the conversion of both reactants. The maximum NO conversion of 94-97 % was attained over both silver-alumina catalysts in the lower temperature range between 200 and 450 °C. In contrast, the 3Ag/AO-SG catalyst is characterized by a significantly higher SCR activity at temperatures below 300 °C, as proven by the shift of the maxima of the conversions to lower temperatures by a margin of 30-50 °C for this sample. Ethanol conversion over silver-alumina catalysts proceeded already at temperature around 300350 °C which is in line with the earlier observations of Shimizu [9] and Burch [10] who argued that NOx were very active oxidants and facilitated catalytic oxidation of hydrocarbons.
It was found by the model reaction of ethanol oxidation, that EtOH was directly involved in the catalytic combustion process. This indicates that only a part of reducing agent is being used in the target SCR reaction for reduction of NO.
The efficiency of the consumption of the reducing agent is crucial for the SCR process [3, 15], while alcohol can also be consumed for parasitic oxidation without participation in the target SCR reaction. Thus, there are several approaches for the determination of selectivity in the SCR reaction in the presence of organic reductants [2-5]. The most convincing is a definition proposed in [11, 12]. The defined parameter balances conversion of NO and of
a reducing agent according to the target SCR reaction (eq. 4, 5):
aNO + CmHn + O2 = mCO2 + a/2 N2 + n/2 H2O (5)
The selectivity of EtOH-SCR reaction for the investigated samples was calculated for the temperature range between 150 °C and 500 °C and is summarized in Figure 2. The highest SCR selectivity of about 33-38 % was reached over 3Ag/AO-SG catalyst at temperatures below 300 °C.
50 -,-,-,-,-
3 Ag/AO-SG 3 Ag/AO-C I IAO-SG EZ22AO-C
150 200 250 300 350 400 450 500 Temperature / °C
Figure 2. Comparison of the SCR-selectivity in EtOH-SCR reaction over pure alumina supports and silver alumina catalysts in the
temperatures range between 150 and 500°C.
Nevertheless, the major portion of ethanol underwent undesirable combustion already at these moderate reaction temperatures, as shown in Figure 1, 2. The SCR selectivity in the whole temperature range decreased in the sequence: 3Ag/AO-SG > 3Ag/AO-C > AO-SG > AO-C. The SCR selectivity is higher for 3Ag/AO-SG than for 3Ag/AO-C and it is also higher for AO-SG than for AO-C. This indicates that the sol-gel y -alumina support and the corresponding 3Ag/AO-SG catalyst exhibit different catalytic acidity of silver species, and different textural, hydro-phobic and acidic properties; consequently higher activity and selectivity towards reduction of NOx and/or nonproductive oxidation of EtOH than that of commercial support.
Most scientists argue that during the EtOH-SCR reaction the selective reduction of NO to N2 proceeds by the following sequence (6) with intermediate formation of HNCO and its hydrolysis to ammonia [2-5, 13], while the hydrolysis of HNCO can be responsible for the selective reduction of NO and/or deactivation of a catalyst.
EtOH ^ CH3CHO ^ [Nitro-organic] ^ HNCO ^ NH3 (6)
The results in Figure 3 show that the difference in SCR activity of both silver-alumina catalysts may be explained by higher concentration of HNCO and lower concentration of ammonia detected over 3Ag/AO-SG catalyst at 300 °C.
It appears that the hydrolysis of HNCO over more hydrophilic and more acidic 3Ag/AO-SG catalyst proceeds faster [14] under the selected reaction conditions as evidenced by its lower concentration when compared to 3Ag/AO-C. In contrast to that, the 3Ag/AO-C catalyst favors the parasitic ethanol oxidation and is also less active towards hydrolysis of the intermediate HNCO. Thus, it may be concluded that there are two reasons for the lower SCR selectivity of 3Ag/AO-C catalyst: (i) the preferred undesirable (total) oxidation of ethanol promoted by the Ag species and Y-alumina support, and (ii) lower activity
of 3Ag/AO-C catalyst towards hydrolysis of HNCO at temperatures below 300°C, resulting in the observed slip of HNCO and NH3 during EtOH-SCR reaction and catalyst deactivation as suggested by [14-16]. Apparently, both of these intermediates can contribute to the formation of various cyclic N/C organic products which can also cause the internal diffusion limitations and catalyst deactivation [7,14].
Figure 3. Formation of iso-cyanic acid (a) during EtOH/SCR reaction over alumina-supports and on silver-alumina catalysts in temperature range between 150 and 520 °C and the comparisons of the concentrations of NH3 and HNCO (b) at 300
°C
The investigation of diffusion limitations. The influence of internal diffusion limitations on NO conversion in EtOH/SCR over 3Ag/AO-C and 3Ag/AO-SG catalysts was studied by varying the grain size from 50 to 600 pm. The effect of catalyst particle size on NOx conversion in the temperature range between 150 °C and 520°C is presented for both catalysts in Figure. 4.
Figure 4. The effect of the particle size on NO conversion over 3Ag/AO-C (a) and 3Ag/AO-SG (b) catalyst and the dependence of the reaction rate of NO conversion at 175, 250, 300 and 350 °C on average particle size during EtOH-SCR on 3Ag/AO-C (c) and 3Ag/AO-SG (d) catalyst. At temperatures below 350 °C, NO conversion over the mesoporous 3Ag/AO-C catalyst is affected by
variation of the catalyst grain size (Figure 4a) while at higher temperatures no visible effect was detected. In contrast, NO conversion over meso-macroporous 3Ag/AO-SG catalyst is almost unaffected in the whole temperature range (Figure 4b). To obtain a more detailed insight, the reaction rate of NO conversion was calculated for 175, 250, 300 and 350 °C (Figure 4c,d). The results show that the reaction rate is independent on the particle size for meso-macroporous 3Ag/AO-SG catalyst, and thus intra-particle diffusion does not affect the reaction rate of EtOH-SCR over it, i.e. it is not controlled by internal diffusion. On the other hand, the observed reaction rate (rNO) for mesoporous 3Ag/AO-C catalyst decreases linearly with the increasing particle size (Figure 4c). This indicates the presence of internal mass-transfer limitation in the low temperature range (< 350°C) [7, 17, 18].
The observed linear dependence of rNO on particle diameter for Ag/AO-C catalyst is in agreement with the earlier results of Tronconi et al. [7] for NH3-SCR reactions and confirms a first order reaction with internal diffusion limitation, characterized by an inverse proportionality between reaction rate and particle diameter [19]. In the case of large particles of macro-mesoporous 3Ag/AO-SG catalyst, the intraparticle diffusion limitations are not present, allowing access of the reactants to all catalytic active centers. Interestingly, the dependence of the reaction rates on the particle size for Ag/AO-C (Figure 4c,d) shows that as the reaction temperature increases, the effect of particle size becomes small, i.e. the angle of slope becomes smaller. Thus, the effect of particle size is more pronounced at low temperatures, between 180 °C and 300 °C. This could be explained only by a change in the reaction mechanism in the investigated broad temperature range [2, 3]. As proposed by [20, 21], bulky products or intermediates can be formed, especially at low reaction temperatures, which could strongly affect the intraparticle diffusion.
Long-duration activity test and catalyst deactiva-tion. To understand the impact of intraparticle diffusion on the EtOH-SCR reaction in more detail, the long-duration activity of both silver-alumina catalysts was investigated during a 700 min run at temperatures below 300°C. The long duration activity test shows that during 700 min of TOS, a continuous decrease in NH3 and CO2 concentrations with the simultaneous increase in NO and EtOH concentrations in the outgas for the mesoporous 3Ag/AO-C sample take place, even when the feed contains excess of water and oxygen. This indicates the deactivation of 3Ag/AO-C catalyst. In contrast, a constant rise in concentration of reaction products was achieved in the first 60 minutes, followed by a nearly constant activity for the rest of 700 minutes. his is another evidence of the advantageous effect of the hierarchical meso-macropore structure of the 3Ag/AO-SG on their long duration performance. There was a substantial deficit in the atom balances especially for N and C of about 8 and 10 %, respectively, in the output gas over the 3Ag/AO-C catalyst at 275 °C. This suggests the formation of products undetected during online FTIR-analysis. Moreover, it was found that at low space velocity (GHSV= 10.000 h-1) on the addition of 200 ppm of NH3 and 220 ppm of formaldehyde (as 1.5 wt.-% aqueous solution) to the input gas, further decrease of NO and EtOH conversion and a drastic coke deposition occurred at 275 °C.
A similar long duration activity test was carried out for 3Ag/AO-C catalyst at 330 °C. It was found that higher reaction temperature alone was sufficient to pro-
vide constant concentration of products at the reactor outlet even for a reaction time over 700 minutes. This suggests the absence of deactivation phenomena at higher reaction temperatures.
Characterizations of the spent catalysts. To elucidate the differences in catalytic activity of the catalysts supported on commercial and hierarchical Y-alumina, as well as their deactivation behavior, the characterization of the spent catalysts was carried out. First, it should be mentioned, that the ICP-EOS analysis showed no significant change (less than 2 rel. %) of the total Ag content after long time run by SCR testing. Therefore, no leaching of Ag species for both 3Ag/AO-C and 3Ag/AO-SG catalysts occurred during EtOH-SCR in the presence of water.
UV-vis investigations. For the spent 3Ag/AO-C and 3Ag/AO-SG samples after 12 h of EtOH-SCR reaction at 300 °C, a change of intensity of the absorption bands in the whole UV-region between 200 and 450 nm was observed as compared with the fresh samples. For the spent 3Ag/AO-SG catalyst, a marginal decrease of highly dispersed Ag+ and large Agn+ clusters in combination with a slight increase in the content of small Agn+ cluster and metallic silver of about 1-2 rel.% was observed (Figure 5). In contrast, a significant decrease of about 12 rel. % of highly dispersed Ag cations and a corresponding increase of the fraction of metallic species were detected for the spent 3Ag/AO-C. The changes for the latter catalyst may have arisen from sintering of silver particles during the EtOH/SCR reaction [8].
50
40
5 30 c o> c o ° 20
a)
>
2 10
0
Figure 5. Comparison of the the relative content of various silver species of fresh (fr) and spent (sp) catalysts: 3Ag/AO-SG (a) and 3Ag/AO-C (b).
UV-vis results indicate that the distinct agglomeration of highly dispersed Ag+ cations under formation of metallic silver and large Agn+ cluster takes place after long-time SCR-reaction at 280 °C on the 3Ag/AO-C. Taking into account the numerous results, [8, 22-24] it may be concluded that the AgOx species immobilized on the commercial
Y-alumina surface are weakly bound to the support and undergo a cyclic oxidation/reduction more often during SCR-reaction. This results in a higher tendency of agglomeration of AgOx species during long duration operation on 3Ag/AO-C [23-25]. In contrast, the spent 3Ag/AO-SG catalyst contained nearly constant distribution of AgOx species. Thus, the silver is more strongly bonded on the surface of sol-gel Y-alumina.
The reducibility of the silver-alumina catalysts. To obtain information about reduction behavior of silver-alumina catalysts and possible reason for their unique SCR activity, the H2-TPR investigation was performed for catalysts after 12 h run in SCR at 300 °C. Both spent samples contained reducible silver species. However, a clear differ-
1 I3AO/AQ-SG fr 1 1 3Ag/AO-C fr EZ223Ag/AO-SG sp EZ3 3Ag/AO-C sp
1 - - - 6
high disp. nano cluster large cluster metalic Ag* Agn°*; n:4-8 Agn'*n>8 Ag°
ence between the H2-TPR profiles of both could be observed (Figure 6).
Figure6. H2-TPR profiles of the 3Ag/AO-C and 3Ag/AO-SG catalysts. Fresh p(a, c) and spent i.e. after SCR reaction (b, d). thick line: experimental, thin deconvoluted
Firstly, the TPR profiles for the fresh and spent 3Ag/AO-C and 3Ag/AO-SG catalysts (Figure 6) show two distinct groups of H2-consumption with a sharp peak centered at about 265 °C and broad area at higher temperature above 365 °C, whereas most H2-consumption, about 55-68 rel. % takes place in the lower temperature range. Secondly, a considerable part (35-45 %) of the AgOx species of the fresh 3Ag/AO-SG undergoes reduction at temperature range between 350 °C and 560 °C. In contrast, the spent 3Ag/AO-C catalyst shows two peaks of hydrogen uptake located at a similar temperature range compared to fresh samples and an additional peak which appears at lower temperatures of about 120 °C. Whereas the total amount of H2 consumed for the fresh and spent catalyst has a nearly constant value of 104-2 pmol/g, the distribution of the H2-uptake, especially for high temperature region (365 °C), underwent a remarkable alteration with the decrease of these species from 35 to about 13 rel. %.This indicates that during SCR reaction, an increase in the particle size of silver oxide occurred. The H2 consumption at 120 °C can be attributed to the reduction of large Ag2O clusters interacting weakly with the alumina support and those are typically easier to reduce than smaller metal oxide particles [22].
The H2-TPR profile of the spent 3Ag/AO-SG catalyst, similar to fresh samples, (Figure 6c,d) shows a main reduction peak at 260 °C and a very broad H2 consumption region between 300 °C and 560 °C. Three visible reduction areas with H2 consumption of about 7, 9 and 19 rel. % were obtained by deconvolution at Tmax at 320 °C, 420 °C and
560 °C respectively. However, the 3Ag/AO-SG catalyst after SCR run is characterized by more dispersed silver species with a relative content of about two times higher than its 3Ag/AO-C counterpart, i.e. 13 and 28 % respec-
tively. This indicates that the surface silver species in 3Ag/AO-SG are more stable towards agglomeration and/or reductive sintering under SCR conditions. This may be explained by Ag species more strongly bound on the surface of sol-gel-alumina than that on the commercial y-alumina. As a consequence, such strongly bound Agn+ species should undergo reduction with hydrogen at high temperatures (> 350 °C) as confirmed by the results of the H2-TPR experiments. Obviously, the Ag species loaded onto commercial Y-alumina phase partially undergo pronounced agglomeration during the repeated redox cycles during EtOH-SCR-reaction [26]. It is in agreement with the results of our UV/-vis experiment and literature dates of Richter et al. [24] and Furusawa et al. [27] who reported that oxygen treatment caused significant silver sintering (oxygen-aging) on silver catalyst loaded on y-alumina support.
Summing up, the isolated Ag+ cations, silver
+
nano and large Agn clusters in 3 Ag/AO-SG are more resistant towards agglomeration or sintering during long duration reaction run, due to the stronger interaction of silver species with the more hydrated surface of sol-gel-alumina support [23, 25]. In contrast, the highly dispersed or isolated Ag+ cations and Agn+ clusters in 3Ag/AO-C catalyst are more mobile on the surface of the commercial y-alumina. The obtained results are in agreement with the reports of Zhang et al. [25] who studied the influence of various types of alumina precursors (AlOOH, y-AI2O3 and Al(OH)3) and of various calcination temperatures on SCR activity of Ag catalysts in the reduction of NO in the presence of propene. Calcination temperatures above 750 °C were found to markedly accelerate sintering of AgmO clusters, resulting in a severe non-selective combustion of propene competing with the target SCR reaction [24].
Textural, thermogravimetric and FTIR investigations. The texture of the spent 3Ag/AO-C catalyst after reaction run at various reaction times and temperatures was compared to that of 3Ag/AO-SG. Additionally, the TG experiments were supplemented by on-line MS analysis to elucidate the type of products desorbed from the catalyst deposits during temperature programmed heating in the presence of air. As show in Table 1, specific surface area of the spent 3Ag/AO-C catalyst decreased continuously from 232 m2 g-1 to 151 m2 g-1 and the total pore volume from 0.68 cm g-1 to 0.4 cm g-1 after 700 min of TOS. On the other hand, the long-time run over both catalysts at 350 °C, as well as on meso-macroporous 3 Ag/AO-SG catalyst at 275 °C shows no significant drop of these parameters. Considerable amounts of deposits of about 4.38 wt. % and 10.36 wt. % of C and N respectively were detected for the spent 3Ag/AO-C catalyst after 480 min TOS at 275 °C.
Tablel. Textural properties, weight loss (■ TG) and content of deposits of the spent silver-alumina catalysts determined by nitrogen
Catalyst;reaction condition Abet a)/m2 g-1 Total PV/cm3 g-1 ATG a)/wt.-% C/N content a)/ wt.-%
3Ag/AO-C; fresh 232 0.68 - -
3Ag/AO-C; 240 min/275°C 203 0.64 12.2 3.15/ 7.34
3Ag/AO-C; 480 min/275°C 187 0.62 17.4 4.38/10.36
3Ag/AO-C; 700 min/275°C 152 0.60 23.3 7.27/14.38
3Ag/AO-C; 700 min/350°C 230 0.55 ca. 1.1 0.38/0.64
3Ag/AO-SG; 700 min/275°C 250 0.54 ca. 1.2 0.32/0.52
a) determined after vacuum treatment of the spent catalyst by 3-5 mbar at 100°C for 3 h.
After 700 min of TOF, the coke loadings were highest for the mesoporous 3Ag/AO-C sample with about 7.27 wt. % and 14.38 wt.% of C and N, respectively.
The shapes of TG and DTG profiles, as shown in Figure 6a, b, of the spent 3Ag/AO-C catalyst are similar after different TOS. The main weight loss of the sample took place between 220 °C and 420 °C with a maximum at about 370 °C. A simultaneous production of water, CO2, HNCO, HCHO and NH3 in this temperature range was detected by MS-online analysis As the TG/DTG measurements were performed in air flow, it may be concluded that the creation of these products results from the thermal decomposition or from combustion of the catalyst deposits. The weight loss at 120 °C could be assigned to detachment of weakly bound or physiosorbed water. The TG/DTA profiles indicate that by increasing run time from 240 min to 700 min, the weight loss increases from 12.2 wt. % to 23.3 wt. %, indicating that longer reaction times favor the formation of the catalyst deposits. On the other hand (Figure 6c), the maxima for Figure 6c the building of all detected products (NH3, CO2, H2O and formaldehyde) are located in the narrow temperature range of 355 ± 15°C, suggesting that the thermal stability of the catalyst deposits and/or their oxidative decomposition do not change with TOS of SCR reaction .
Additionally, the spent 3Ag/AO-C catalyst after 700 min run by EtOH-SCR reaction at 275°C was investigated by FTIR and MS spectroscopy in order to clarify the possible structure of the catalyst deposits. The Ag/AO-C catalyst loaded with melamine shows several absorption bands centered at 3469, 3418, 3338 and 3137 cm-1 which are characteristic for the deformation vibrations of amino (>NH2 or -NH) groups of melamine. Furthermore, this sample is characterized by several distinct peaks at 1630, 1560, 1470 and 814 cm-1. The peaks at 810 cm-1 and between 1350-1560 cm-1 correspond to the "aromatic vibrations" of the 1,3,5-triazine ring. [28]. Similar FTIR spectra, however with lower intensity, were detected for the spent 3Ag/AO-C catalyst. The frequency band at 3420 cm-1 can also be assigned to the hydroxyl group of the methylol (CH2-OH) fragment of mono or di-methylolamine. The three peaks at 1630, 1560, and 1470 cm-1 can be clearly assigned to methylene groups and the peak at 1170 cm-1 can be assigned to the ether group (C-O-C) of methylol-melamine [29]. These FTIR results indicate the presence of cyclic melamine-like compounds on the surface of deactivated Ag/AO-C catalyst. Furthermore, the deposits of the spent 3Ag/AO-C catalyst were isolated by ultrasonic extraction using MeOH and analyzed by ESI-MS spectroscopy. The MS spectra of the methanol solution obtained after extractions show the main peak with m/z: 126, corresponding to the melamine molecule C3N6H6. Additionally, two MS peaks of lower intensity (about 5 % compared to the main peak) with m/z =156 and 190 were detected, indicating the presence of mono-methylol-melamine and di-methylol-melamine structures, respectively. Thus it could be concluded that only a small amount (less than 5 %) of methylol-melamine derivatives may be also present on the surface of the spent 3Ag/AO-C catalyst.
Mechanism of EtOH-SCR reaction. Firstly, it is necessary to briefly outline the commonly accepted consensus on the mechanism for DeNOx EtOH-SCR reaction over 3Ag/Y-Al2O3, described in numerous reports [2, 3, 22] and also discussed in our previous study [30]. Various N/C/O-containing intermediates such as ethoxy-, acetate-, formiate-, NCO-, NC and/or HCN-species were detected by
IR over silver-alumina catalysts [4, 21, 15]. The chemistry of these intermediates, especially the formation of HNCO (-CNO *) and its transformation, are well established. It involves hydrolysis to ammonia and the reaction of NH3 with nitrous acid (HONO) with the formation of ammonium nitrite. The latter is thermally unstable and rapidly undergoes decomposition to N2 and water at ca. 110°C. [2, 4]. Despite great concern in the literature, there is still an open question concerning the behavior of isocyanate intermediates (or HNCO) in EtOH-SCR reaction, especially, for the operation at lower temperature range. Since iso-cyanic acid is stable in the gas phase, solid polymerization products like cyanuric acid, ammelide, ammeline and/or melamine can be formed via a complex network on the surface or within the porous system of the catalyst [14, 31].
The results presented here reveal that the mass transfer limitations may play an important role during EtOH-SCR for silver alumina catalyst under certain operating conditions, i.e. mesoporous 3Ag/AO-C catalyst and reaction temperatures lower than 300 °C. Because of limited internal diffusion inside the catalyst particles, the overall rate of NO conversion is thus reduced. Moreover, limited hydrolysis of HNCO can be observed over less acidic and hydrophobic 3Ag/AO-C catalyst [30]. All these factors may also favor the undesirable condensation of HNCO to melamine, typically in the small pore, and result in catalyst deactivation by operation at lower temperature range.
It is reasonable to assume that isocyanate (NCO-) and/or HNCO species adsorbed on the less hydro-philic surface of 3Ag/AO-C catalyst at lower temperatures undergo condensation which leads to formation of hetero-cyclic compound such as melamine. These species can block the active sites resulting in poisoning of the catalyst [32]. The hydrolysis of HNCO and its condensation to cy-anuric acid represent two parallel reactions. Moreover, melamine, due to its high activity, can also further react with the formaldehyde (product of ethanol oxidation or decomposition of C1, C2-nitro-organics) forming mono-, di-and methylol melamine and finally hexa-methylol mela-mine. Such reactions can irreversibly poison the catalyst. However, the MS analysis of deposit isolated from deactivated 3Ag/AO-C catalyst revealed that a main component of the "SCR-coke" consists of melamine. Methylol-melamine derivative could not be isolated by extraction due to their limited solubility in methanol.
The long duration experiments clearly indicate a progressing deactivation of 3Ag/AO-C catalyst, in contrast to the hierarchically structured 3Ag/AO-SG sample characterized by more robust behavior without loss of activity during 700 min SCR run at 275 °C. It is evident from DRIFTS, DTG/TG, TPO and elementary analysis on deactivated 3Ag/AO-C sample that the catalyst deposit mainly consists of cyclic C/N heterocyclic compounds which undergo decomposition at temperatures above 330°C.
Conclusions
The obtained results indicate that not only various physico-chemical properties such as acidity, NOx adsorption ability, but also the textural properties of Y-alumina support such as mono or bimodal structure, influence the low temperature EtOH-SCR process. It was found that the catalyst containing silver supported on hierarchically structured meso-macro (sol-gel) Y-alumina possesses higher SCR activity at temperatures below 300°C than the one supported on mono-modal (mesoporous, commercial)
y-alumina. This may be explained by enhanced catalyst activity towards soft oxidation of ethanol to acetaldehyde. Silver-alumina catalyst based on the hierarchically structured alumina is characterized by an improved internal mass transfer. The enhanced hydrophobicity of sol-gel alumina combined with the meso-macroporous nature of the support evidently favors hydrolysis of HNCO and prevents catalyst deactivation by coke deposition. The results of FTIR, DTA/TG and MS measurements of the spent 3Ag/AO-C catalyst suggest that deactivation of the catalyst obviously proceeds due to the formation of undesirable melamine-like compounds. This assumption is supported by the observed long duration activity in EtOH-SCR measurements on both the silver-alumina catalysts at moderate temperature (280-320°C).
In this study, the results an approach of new highly active hierarchical silver alumina catalyst for the reduction of NOx from mobile emission source were presented. The application of ethanol or replacement of ammonia for DeNOx-SCR technology for large emission sources such as combustion of fossil fuels (electric and heating plants, petroleum refinery or FCC-units) were not discussed here. Up to this day the application of SCR in the presence of ethanol (or bio-ethanol) as a reducing agents instead of gaseous ammonia, aqueous ammonia or urea solutions is considered as an alternative for mobile exhaust emissions worldwide. The advantage of the usage of ethanol is a safer design of the DeNOx system and full absence of ammonia leakage. On the other hand, a problem in the conventional application (cars with Euro 5 and Euro 6 norms) of urea solutions, especially in the cold season at temperatures below -7 ° C, is the flocculation of urea. This aspect is especially important for the Russian Federation, not only for the European part but especially for Ural regions and Siberia. The prevention of flocculation of urea requires additionally heating of the urea-tank and all of the transfer lines for reducing agent from the urea tank to the SCR DeNOx unit. In addition, it must be stressed, that today's world production of bio-ethanol (so-called bioethanol 2nd and 3th generation) for technical applications not only in USA, Canada but also in China and the EU is based on the usage of green biomass instead of grain. It could be expected that in the next 5-10 years, the application of DeNOx-SCR technology in the presence of C2, C4 bio-alcohol for mobile emission sources (car and truck) will be also of interest for the Russian Federation with its great motor or car pools.
Acknowledgements
This project was financially supported by the National Science Foundation of Germany (DFG, Deutsche Forschungs Gemeinschaft) Project Gl: 290/8-1.
References
1. Busca G., Lietti L., Ramisa G.,[et al.]. Chemical and Mechanistic Aspects of the Selective Catalytic Reduction of NOx by Ammonia over Oxide Catalysts: A review Appl. Catal. B. Environm. 1996. V. 18. P. 1-36.
2. H. He, Yu Y. B. Selective Catalytic Reduction of NOx over Ag/AI2O3 Catalyst. From Reaction Mechnism to Diesel Engine Test // Catal. Today. 2005. V.100. P. 37-47.
3. Johnson W. L, Fisher, G. B. T., Toops J. Mechanistic Investigation of Ethanol SCR of NOx over Ag/AI2O3 // Catal. Today. 2012. V.184. P. 166-177.
4. Yu. Y, Zhao, J. Yan, Y. Han X., [et al.]. A Cyclic Reaction Pathway Triggered by Ammonia for the Selective
Catalytic Reduction of NOx by Ethanol over Ag/Al2O3 // Appl. Catal. B. Environm. 2013. V.136-137. P. 103-111.
5. Komatsu T., Tomokuni K. K, Yamada I. Outstanding Low Temperature HC-SCR of NOx over Platinum-Group Catalysts Supported on Mesoporous Materials Expecting Diesel-Auto Emission Regulation // Catal. Today. 2006. V.116. P. 244-247.
6. Tokudome Y., Fujita K., Nakanishi K., [et al.]. Synthesis of Monolithic Al2O3 with Well-Defined Macropores and Mesostructured Skeletons via the Sol-Gel Process Accompanied by Phase Separation // Chem. Mater. 2007. V.1. 3393-3398.
7. Tronconi E., Beretta A. The Role of Inter- and Intra-Phase Mass Transfer in the SCR-DeNOx Reaction over Catalysts of Different Shapes // Catal. Today, 1999. V. 52. P. 249-258.
8. Popovych N., Kirienko P., Soloviev S., [et al.]. Selective Catalytic Reduction of NOx by C2H5OH over Ag/Al2O3/Cordierite: Effect of the surface concentration of silver // Catal. Today. 2012. V.191. P. 38-41.
9. Shimizu K., Shibata J., Yoshida H., [et al.]. Silver-Alumina Catalysts for Selective Reduction of NO by Higher Hydrocarbons: Structure of Active Sites and Reaction Mechanism // App l. Catal. B. Environm. 2001. V.30. P. 151-162.
10. Burch R., Scire S. Selective Catalytic Reduction of Nitric Oxide with Ethane and Methane on some Metal Exchanged ZSM-5 Zeolites // Appl. Catal. B. Environm. 1994. V.3. P. 295-318.
11. Shelef M. Selective Catalytic Reduction of NOx with N-Free Reductants // Chem. Review. 1995.V.95(1). P. 209-225.
12. Y. Li Y., Armor J.N. Selective Catalytic Reduction of NO with Methane on Galium Cataylsts // J. Catal. 1994. V.145. P. 1-9.
13. Lee J.H., Schmieg S. J,. Oh S. H. Improved NOx Reduction over the Staged Ag/Al2O3 Catalyst System // Appl. Catal. A. 2008. V.342. P. 78-86.
14. P. Hauck, A. Jentys, Lercher J. A. Surface Chemistry and Kinetics of the Hydrolysis of Isocyanic acid on Anatase // Appl. Catal. B: Environm. 2007. V.70. P. 9199.
15. Bion N., Saussey J., Haneda M., [et al.]. Study by in situ FTIR Spectroscopy of the SCR of NOx by Ethanol on Ag/Al2O3. Evidence of the Role of Isocyanate Species // J. Catal. 2003. V.217. P. 47-58.
16. Piazzesi G. The Catalytic Hydrolysis of Isocyanic Acid (HNCO) in the Uurea-SCR Process; PhD Thesis. Diss. ETH No. 16693. Zürich, 2006.
17. Hansen N., Krishna R., van Baten J. M., [et al.]. Experimental study of Mass-Tansfer Limitations in Fe and Cu-zeolite-based NH3-SCR Monolithic Catalysts // J. Phys. Chem. C, 2009. V.113, P.235-240.
18. Metkar P. S., Harold M.-P. Experimental Study of Mass Transfer Limitations in Fe- and Cu-Zeolite-Based NH3-SCR Monolithic Catalyst // Chem. Eng. Sci. 2011. V.66(21). P. 5192-520.
19. Boudart M., Djega-Mariadassou G., 6 Parasitic Phenomena. 6.1. Mass and Heat Transfer . P. 194-200. in: Kinetics of Heterogeneous Catalytic Reactions, Princeton University Press, Princeton, NJ, 1984 ISBN-13: 9780691640488.
20. Hauck P., Jentys A., Lercher J. A. On the Quantitative Aspects of Hydrolysisis of Isocyanic acid on TiO2 // Catal. Today. 2007. V.127. P. 165-175.
21. Yeom Y. H,. Li M, Sachtler W.M.H., [et al.]. Low-Temperature NOx Reduction with Ethanol over Ag/Y:
A Cmparison with Ag/Y-Al2O3 and BaNa/Y // J. Catal. 2007. V.246. P.413-427.
22. Bethke K. A., Kung H. H. Supported Ag Catalyst for Reduction of NO with C3H6 // J. Catal. 1997, V.172. P. 93-102.
23. Wachs E. I. Catalysis Science of Supported Vanadium // Dalton Trans. 2013. V.42. P. 11762-11769.
24. Richter M., Bentrup U., Ekelt R., [et al.]. The effect of Hydrogen on the Selective Catalytic Reduction of NO in Excess of Oxygen over Ag/Al2O3 // Appl. Catal. B. Environm. 2004. V.51. P. 261-274.
25. Zhang R., Kaliaguine S. Lean reduction of NO by C3H6 over Ag/Alumina Derived from Al2O3, AlOOH and Al(OH)3 // Appl. Catal. B. Environm. 2008. V.78. P. 275287.
26. She X., Flytzani-Stephanopoulos M., [et al.]. SO2 Induced Stability of Ag Alumina Catalyst in the SCR of NO with Methane // Appl. Catal. B: Environm. 2009. V.88. P. 98-105.
27. Furusawa T.,. Seshan K, Lercher J.A., [et al.]. Selective Reduction of NO to N2 in the Presence of Oxy-
gen over Supported Silver Catalysts // Appl. Catal. B. Environm. 2002. V.37. P. 205-2016.
28. Jones F.N., G. Chu G., Samaraweera U. Recent studies of self-condensation and co-condensation of melamine-formaldehyde resins; cure at low temperature; // Progress Org. Coatings. 1994. V. 24(1-2). 189-208
29. Braun D., Legradich V. Investgation of Mela-min-Formaldehyde Reaction Products // Angew. Makrom. Chem. 1974. V.136. P. 41-45.
30. Worch D., Suprun W., Gläser R. Fe- and Cu-Oxides Supported on Y-Al2O3 as Catalysts for the Selective Catalytic Reduction of NO with Ethanol, Part I: Catalyst Preparation, Characterization and Activity, Chemical Papers. 2014. V.68(9). P. 1228-1239.
31. Bernhard A.M., Peitz D., Elsener M., [et al.]. Hydrolysis and Thermolysis of Urea and its Decomposition Byproducts, Biuret, Cyanuric acid and Melamine over Ana-tase TiO2 // Appl. Catal. B: Environm. 2012. V.115- 116. P. 129-137.
32. Bartholomew C. H. Mechanisms of Catalyst Deactivation // App. Catal. A: General, 2001. V. 212. P. 17-60.
