- Email: [email protected]
Acetaldehyde adsorption on TiO2: Influence of NO2 preliminary adsorption
Accepted Manuscript Acetaldehyde adsorption on TiO2: influence of NO2 preliminary adsorption F. Thevenet, L. Olivier, F. Batault, L. Sivachandiran, N. Locoge PII: DOI: Reference:
S1385-8947(15)00924-9 http://dx.doi.org/10.1016/j.cej.2015.06.084 CEJ 13856
To appear in:
Chemical Engineering Journal
Received Date: Revised Date: Accepted Date:
12 May 2015 17 June 2015 18 June 2015
Please cite this article as: F. Thevenet, L. Olivier, F. Batault, L. Sivachandiran, N. Locoge, Acetaldehyde adsorption on TiO2: influence of NO2 preliminary adsorption, Chemical Engineering Journal (2015), doi: http://dx.doi.org/ 10.1016/j.cej.2015.06.084
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Acetaldehyde adsorption on TiO2: influence of NO2 preliminary adsorption F.Thevenet, L. Olivier, F. Batault, L. Sivachandiran, N. Locoge Mines Douai, SAGE, F-59508 Douai, France Abstract Titanium dioxide (TiO2) is a widespread metal oxide used in catalysis and photocatalysis, or coupled to non-thermal plasma for volatile organic compound (VOC) oxidation. Adsorption is a key step in the advanced heterogeneous oxidation processes; adsorption is partially addressed in that it is mainly investigated on fresh TiO2 surfaces. However, the treatment of real effluents combines various pollutants with VOCs; among them, NOx are characterized by reactive adsorption properties on TiO2 leading to irreversibly adsorbed species on the surface. The aim of this work is to determine the impact of NO2 preliminary adsorption on the subsequent adsorption of a model VOC: acetaldehyde. Breakthrough methods are used to (i) characterize acetaldehyde adsorption on fresh and NO2 covered TiO2 surface, (ii) control the coverage of TiO2 surface by irreversibly adsorbed NOx(ads) species. In a first step, acetaldehyde adsorption on TiO2 has been characterized. In a second step, acetaldehyde adsorption has been achieved on TiO2 surface after NO2 exposure corresponding to different surface coverages. Then, acetaldehyde adsorption parameters have been determined. Subsequently, it has been possible to plot the evolution of acetaldehyde reversibly and irreversibly adsorbed quantities, and reversible fraction adsorption constant, as a function of TiO2 surface coverage by NOx-(ads) species, so called TiO2 surface NO2 ageing. Interestingly, both adsorbed acetaldehyde fractions are highly impacted by the presence of NOx-(ads) species on TiO2 surface. The irreversibly adsorbed fraction is considerably decreased since the lowest values of TiO2 surface coverage by NOx-(ads) species. Hypotheses related to competitive adsorption are not sufficient to explain the observed impact, suggesting that NOx-(ads) species modify TiO2 surface chemistry and acetaldehyde reactive adsorption. The reversibly adsorbed fraction of acetaldehyde is highly impacted as well; reversibly adsorbed amounts are decreased since the lowest surface coverage by NOx-(ads) species. The corresponding adsorption constants are abruptly increased as soon as TiO2 surface has been exposed to NO2, suggesting the formation of a new adsorption mode for acetaldehyde on NO2 exposed TiO2 surface. This paper evidences the considerable impact of NO2 on TiO2 adsorption properties regarding VOCs. Considering that NOx may accumulate on TiO2 the long term behaviour of such processes should be investigated further taking into account the NOx vs. VOC interaction for adsorption.
Keywords: adsorption; TiO2; NO2; acetaldehyde; surface coverage
1
1. Introduction In the field of advanced heterogeneous oxidation processes dedicated to pollution remediation, adsorption plays a key role regarding the system performances. Among various reactive solids, metal oxides are widely used as sorbent, catalyst, or plasma coupling materials. The purpose of these processes is generally to induce oxidation, and even mineralization, of organic species. Thermal-catalysis, photocatalysis and plasma-catalysis are the most widespread techniques. Adsorption and desorption of pollutants from solid surfaces are key steps in these heterogeneous techniques. Indeed, adsorption makes possible the catalytic treatment of pollutants; similarly, it expands the residence time of compounds to be treated in plasma-catalyst combination. In addition, desorption of obtained products makes possible the renewal of pollutants on the surface and prevents from poisoning. Nevertheless, in both catalytic and plasma-catalytic methods, the catalyst deactivation has been reported by several authors [1][2][3]. Three kinds of species are typically pointed out to explain the deactivation of oxidation heterogeneous processes: NOx, COx and reaction intermediates [3][4]. NOx and COx may be present in the effluent to be treated or produced by the process itself. Among various metal oxides, TiO2 is a material of interest combining catalytic and photocatalytic properties. Moreover, due to its high relative permittivity, it is classically coupled to plasma devices for environmental applications. Interestingly, it is characterized by distinctive surface properties regarding adsorption and complexation [5][6][7]. However, irrespectively of the oxidation technique, the long term performances of this material is poorly addressed whereas knowledge improvements on pollutant adsorption and desorption may be fruitful to understand the system performances and ageing. Acetaldehyde is a typical VOC of indoor air, it is reported in almost all kinds of buildings with concentrations around 18 µg.m-3 (93 ppb) [8]. Exceptionally high levels may reach 140 µg.m-3 (720 ppb) [9]. The main sources of acetaldehyde are glues, paints, deodorants and fuel additives[8][10]. Moreover, it is frequently reported that the oxidation of hydrocarbons (HC) leads to the formation of acetaldehyde as a side product [11][12][13][14].Qualitative studies have been dedicated to the adsorption of acetaldehyde on TiO2 surface; two adsorption modes are generally reported at ambient temperature: (i) a weak adsorption mode on hydroxyl groups present on TiO2 surface and (ii) a stronger adsorption mode on Ti4+ sites (Lewis acidic sites). Batault et al. [15] investigated quantitatively acetaldehyde adsorption on clean TiO2 surface. They related the observed reversible and irreversible adsorbed fraction with the reported adsorption modes. Interestingly, acetaldehyde adsorption on TiO2 is a reactive process since two acetaldehyde adsorbed molecules may undergo an aldolic condensation into 2-butenal via 3-hydroxy-butanal as an intermediate [13][14][16]. NOx (NO and NO2) are formed in thermal power plants or combustion engines. Besides, when non-thermal plasma is ignited in air, it produces higher oxides of nitrogen such as NO, N2O, NO2 and HNO3, in concentrations ranging from some hundreds of ppb to some hundreds of ppm, depending on the plasma deposited energy [17][18][19]. In non-thermal plasma performed under pure and dry air, most of NOx are contributed by NO2 because of the fast oxidation of NO by plasma produced O3 [20]. In indoor air, levels of NO and NO2 respectively range from tens of ppb to few hundred ppb [21]. National Institute for Occupational Safety and Health (NIOSH) suggested that the Immediately Dangerous to Life and Health (IDLH) concentration of NO2 is 20 ppm. Studies focused on risk assessment have shown that high outdoor NO2 concentration observed in residential areas contributes to increased respiratory and cardiovascular diseases and mortality [22]. Air treatment systems based on advanced heterogeneous oxidation aim at converting VOCs; however, the material surface is exposed to the wide diversity of species present in the air stream or generated by the system itself. Among them, NOx may dramatically influence VOC adsorption processes and decomposition mechanisms [23][24], which may also impact the system efficiency 2
and ageing. Former works did not extensively study the interactions between VOCs and such copollutants into heterogeneous advanced oxidation processes, especially the crucial adsorption step. At room temperature, under dark condition and under dry air flow, NO does not adsorb on TiO2 surface. In contrast, without any UV radiation, under dry air and at 296 K, NO2 is known to undergo reactive adsorption on TiO2 surface. This leads to the formation of adsorbed species. Gaseous NO may also be evolved along NO2 reactive adsorption process on TiO2 [25][26][27][28]. Interestingly, Sivachandiran et al. [27] recently proposed the following mechanism to describe NO2 adsorption on TiO2 surface under dark conditions ( Equation 1 to Equation 3): Equation 1
2 NO2 (gas)
2 NO2 (ads)
Equation 2
2 NO2 (ads)
NO3- (ads) + NO+(ads)
Equation 3
NO
+ (ads)
2-
+O
-
NO2
(TiO2 S.L.)
(ads)
Authors evidenced that, as soon as the TiO2 surface coverage “θ" by NO2-(ads) and NO3-(ads) exceeds 0.2, then the reaction corresponding to Equation 4 occurs: Equation 4
-
NO2
(ads)
-
+ NO2 (ads)
NO3
(ads)
+ NO (gas)
The global mechanism can be summarized as follows (Equation 5): Equation 5
3 NO2 (ads) + O2- (TiO2 S.L.)
2 NO3- (ads) + NO (gas)
From adsorbed NO2, various species are produced: some are reaction intermediates and subsequently do not accumulate on the TiO2 surface, like the highly reactive nitrosonium ion NO+ and nitrite ion NO2-; some are evolved in the gas phase, like NO; and some are irreversibly adsorbed on TiO2 like the nitrate species produced from Equation 2 and Equation 4. Both nitrate species produced are reported as irreversibly adsorbed on Ti4+ sites [25][26][27][28]. As a consequence, NO2 adsorption onto TiO2 leads to the formation of irreversibly adsorbed nitrates species which could be considered as poisons for the catalyst. Therefore, NO2 adsorption on the surface may be responsible for TiO2 surface deactivation, or ageing. The main impact of such adsorbed species on the photocatalyst surface could be the decrease of VOC sorption which may considerably hinder the very first step of any catalytic process. The objective of the present work is to assess the impact of NO2 on the adsorption of acetaldehyde on TiO2 surface under dark conditions. In a first step, an experimental procedure is proposed to control TiO2 surface coverage, i.e. ageing, by NO2. Then, single acetaldehyde adsorption is investigated on fresh TiO2 surface in order to quantify (i) the total adsorbed amount, (ii) the reversibly and irreversibly adsorbed fractions of acetaldehyde, and (iii) acetaldehyde adsorption constant. Finally, acetaldehyde adsorption parameters are determined and discussed as a function of TiO2 surface coverage by NO2 ad-species.
3
2. Experimental 2.1 Materials and gas stream preparation The TiO2 P25-Degussa material (70% anatase, 30% rutile) used in the present work was provided and used as a powder. The specific area of the material has been measured using the BET (Brunauer, Emmett and Teller) N2 adsorption method at 77 K and the determined value is 57 ± 3 m² g-1. Typical photocatalyst mass used for each experiment is 15.0 ± 0.05 mg. Gas cylinders, 99.9% purity, are provided by Praxair (Belgium) with nominal concentrations ranging from 35 to 500 ppm balanced in N2. Experimental air stream is provided via two coupled devices fed by a classical air compressor. The compressed air is first sent into Claind (Lenno, Italy) AZ-2020 catalytic zero air generators, which is coupled to a Pressure Swing Adsorption (PSA) device. The remaining impurity levels in the air stream are lower than the analytical system detection limits: VOCs < 0.1 ppb, CO2 < 10 ppb and CO < 80 ppb. Moisture level is about 2 ppm. This air stream is used for infrared spectrometers purge and as carrier gas during each experiment. 2.2 Adsorption setup The experimental bench used for this study is presented in Figure 1. It consists in three parts: (i) gas generation setup, (ii) reactor and (iii) analytical device. The gas generation part includes the gas cylinders and zero air generation system, coupled to mass flow controllers (MFCs). The reactor is made of a U-shaped vertical quartz tube with 4 and 6 mm inner and outer diameter respectively. The catalyst bed is positioned in a widened section of the U-shaped tube of 15 mm inner diameter and maintained with a quartz wool plug. The reactor is inserted in a tubular oven and the temperature of the catalyst bed is monitored using a K-type thermocouple set in contact with the catalytic bed.
2.3 Analytical device The Fourier Transform Infrared spectrometer (FTIR) is equipped with a 10 m long path cell and liquid N2 cooled Mercury Cadmium Telluride (MCT) detector with 0.5 cm-1 spectral resolution. Spectra are acquired using Result-3 software with 16 scans per spectrum. Quantifications and data processing were performed using TQ-Analyst software. The detection limits for the species of interest are reported in Table 1. As reported in Figure 2, the quantification of CH3CHO, NO, NO2 and N2O have been performed using the framed absorbance bands. More precisely, acetaldehyde has been quantified with the Q-branch associated to ν(C=O), between 1745 cm-1 and 1743 cm-1 and with the Q-branch associated to δ(CH3), between 1396 cm-1 and 1394 cm-1. N2O has been quantified with the R-branch of the band associated to ν(N-N) between 2223 cm-1 and 2260 cm-1. NO has been quantified with the R-branch of the band associated to ν(N=O), between 1936 cm-1 and 1878.6 cm-1. NO2 has been quantified with the P-branch of the band associated to νas (NO2) between 1606 cm-1 and 1580 cm-1. 2.4 Experimental Procedure Figure 3 represents the cyclic experimental procedure used to assess the influence of NO2 on TiO2 surface regarding acetaldehyde adsorption. For all the experiments the total flow rate has been kept constant as 600 mL.min-1 unless otherwise mentioned: (i) cleaning of the catalyst surface by maintaining the reactor at 703 K under air for 1h, (ii) surface ageing of TiO2 by adsorbing 30 ppm of NO2 in dry air, (iii) acetaldehyde adsorption until breakthrough, (iv) flushing the reactor under dry air. In particular, the step (ii) is skipped when acetaldehyde adsorption has to be performed on 4
fresh TiO2 surface. Moreover, to estimate the acetaldehyde adsorption isotherm, the step (iii) has been performed for various acetaldehyde concentrations like 5, 10, 15, 20, 50, 100, 200, and 300 ppm. All the experiments have been performed under dark condition and repeated three times in order to evaluate their reproducibility. It has been formerly evidenced that N2O does not show any adsorption on TiO2 surface under our experimental conditions [27]. Thus, N2O is used in the experiments to determine the system mixing curve in order to achieve accurate quantifications of adsorbed species using the classical breakthrough method.
5
3. Results and discussion
The influence of adsorbed NO2 on acetaldehyde adsorption over TiO2 is investigated. Firstly, NO2 adsorption is performed on clean TiO2 surface and the surface coverage is determined for different NO2 adsorption times. Secondly, acetaldehyde is adsorbed on fresh and/or NO2 aged TiO2 surface. Moreover, for all experiments the adsorbed amounts of acetaldehyde are quantified and the adsorption constants are determined and discussed.
3.1. Control of TiO2 surface coverage: NO2 adsorption on TiO2 at 296 K Figure 4 shows the temporal profiles of NO2 and N2O breakthrough curves and NO transient formation curve on fresh TiO2 surface. NO2 adsorption on TiO2 is performed under dark condition and in dry air with a constant flow rate of 600 mL min-1. At the reactor inlet, 30 ppm of NO2 and 5 ppm of tracer N2O are concurrently introduced. As can be seen in Figure 4, during adsorption, NO2 concentration at the reactor downstream increases until the surface is saturated. As demonstrated by Sivachandiran et al. [27] NO is evolved in the gas phase after 6 minutes of adsorption. And the NO concentration reaches the maximum value between 15 and 20 minutes of adsorption. When the surface tends to reach saturation, NO concentration gradually decreases. In particular, it reaches zero when the surface is completely saturated. Both NO2 and NO temporal profiles are overlaid with the system mixing curve obtained from the temporal profile of N2O. Five different adsorption experiment sets have been carried out, they differ by their adsorption times ranging from 150 s to 5000 s. Each set of experiment has been repeated four times in order to assess the reproducibility. Considering the mechanism proposed by Sivachandiran et al. [27] and recalled in the introduction, the stable ad-species present on saturated surface are : (i) irreversibly adsorbed NO2(ads) and (ii) NO3 (ads) species. Regarding NO3 (ads) species, Hadjiivanov et al. [29] reported that NO2 adsorption on TiO2 surface leads to the formation of monodentate as well as bidentate NO3-(ads) species on Ti4+ sites. Although NO2-(ads) is a reaction intermediate, it coexists with NO3-(ads) on partially covered surface. Therefore, in the following discussion the notation of NOx-(ads) is used to represent the adsorbed NO2-(ads) and NO3-(ads) species. Subsequently, the coverage rate θ is used to characterize the surface coverage due to NO2 adsorption. The parameter θ is defined as the ratio between NOx-(ads) adsorbed amount along partial adsorption and the total amount of NOx-(ads) adsorbed on saturated TiO2 surface. As reported in Figure 4, from NO2 adsorption breakthrough curve, it is possible to determine the total amount of NO2 consumed on TiO2 surface and the total amount of evolved NO for various adsorption times. The total amount of NO2 consumed is obtained by integrating the area between the NO2 breakthrough curve and the N2O system mixing curve and the total amount is expressed in µmol m-2. Similarly, the amount of NO produced is quantified directly from NO concentration temporal profile. Consumed NO2 values are reported in Table 2. In the following, q(NOx-(ads) ) represents the sum of NO2-(ads) and NO3-(ads) adsorbed amounts per surface unit. As reported by Sivachandiran et al. [27] the proposed mechanism allows assessing the following points: (i)
There is a θ threshold regarding the production of NO in the gas phase.
(ii)
Below this θ threshold, the stoechiometric ratio between consumed NO2 and NOx(ads) is 2:2 (Equations 1 to 3).
(iii)
Beyond the θ threshold, the stoechiometric ratio is 3:2 (Equation 5). 6
Thus, q(NOx-(ads)) is quantified by multiplying the consumed NO2 amount by 1 or 2/3, depending on the adsorption process advancement. The values related to saturated surface are considered as references to calculate θ when θ < 1. The uncertainty values are determined from the dispersion of the measurements obtained during the repeated experiments. Table 2 reports the amounts of NO2 consumed per surface unit for the different adsorption times and the corresponding q(NOx-(ads)) amounts and θ values. Based on Table 2, four different surface coverages (0.05, 0.1, 0.55 and 0.92) can be considered to evaluate the impact of NO2 on acetaldehyde adsorption. The first two are lower than NO formation threshold and the last two are above NO formation threshold. Acetaldehyde adsorption parameters, i.e. irreversibly adsorbed fraction (qirr), reversibly adsorbed fraction (qrev) and adsorption constant (K) have been determined for each selected values of θ .
3.2 Acetaldehyde adsorption on clean TiO2 surface Figure 5 shows acetaldehyde breakthrough curve overlaid with the N2O system mixing curve during breakthrough and flushing steps on clean TiO2 surface. During adsorption, 20 ppm of acetaldehyde balanced with dry air is sent to the reactor inlet. The acetaldehyde concentration at the reactor outlet increases till the adsorption equilibrium is reached. Then the reactor is flushed under dry air. During flushing, the reversibly adsorbed acetaldehyde fraction desorbs. Flushing is stopped as the acetaldehyde concentration returns to zero. The total amount of acetaldehyde adsorbed, i.e. irreversibly and reversibly adsorbed fractions, at equilibrium is quantified by integrating the area between the acetaldehyde breakthrough curve and the system mixing curve. Similarly, the reversibly adsorbed amount is quantified by calculating the area between both curves during flushing. The irreversibly adsorbed amount is then quantified by subtracting the reversibly adsorbed amount from the total adsorbed amount. Obtained values on fresh TiO2 surface are 8.61 µmol m-² ± 10% for the irreversibly adsorbed fraction and 0.42 µmol m-² ± 10% for the reversibly adsorbed fraction, corresponding to a total adsorbed amount of 9.03 µmol m-² ± 15%. These experimental data obtained with 20 ppm inlet concentration of acetaldehyde are consistent with the recent work of Batault et al. [15] and the corresponding adsorption isotherms. This result confirms that acetaldehyde adsorption is mostly irreversible (95%), on clean TiO2 surface, and highlights the predominance of strong chemical interaction between acetaldehyde oxygen and Ti4+ surface Lewis acid sites [13]. Furthermore, adsorption of acetaldehyde on clean TiO2 surface has been investigated with acetaldehyde inlet concentrations ranging from 5 to 300 ppm. For each inlet concentration, reversible and irreversible fractions have been quantified, they exhibit contrasted behaviours. Indeed, the irreversibly adsorbed fractions is poorly influenced by acetaldehyde inlet concentration, which suggests that even with 5 ppm inlet concentration, all irreversible adsorption sites can be covered. In contrast, the reversibly adsorbed fraction of acetaldehyde shows a typical Langmuir isotherm behaviour with respect to acetaldehyde inlet concentration. This isotherm, determined on clean TiO2 surface, is reported on Figure 7 (a). It enabled the calculation of the maximum reversibly adsorbed amount on clean TiO2 surface as : qrevm = 1.6 ± 0.2 µmol m-² and the adsorption constant of reversibly adsorbed acetaldehyde on clean TiO2 surface as 0.010 ± 0.002 ppm-1.
7
3.3 Acetaldehyde adsorption on preliminary NO2 exposed TiO2 surface Acetaldehyde adsorption and desorption have been performed for the various selected values of θ, i.e. various surface coverages by NOx-(ads) species. Acetaldehyde, NO, NO2 and N2O concentrations are monitored at the reactor downstream. For each θ values reported in Table 2, acetaldehyde adsorption has been repeated three times. It has to be noted that neither NO nor NO2 have been monitored at the reactor outlet during acetaldehyde adsorption on NOx-(ads) covered TiO2 surface. This finding emphasizes the fact that acetaldehyde adsorption does not affect the chemisorbed NO2- and/or NO3- species on TiO2 surface under our experimental conditions. For each experiment both irreversibly and reversibly adsorbed fractions of acetaldehyde have been quantified and correlated to θ values.
3.3.1 Impact of NO2 preliminary adsorption on acetaldehyde irreversibly adsorbed fraction Figure 6 shows the amount of irreversibly adsorbed acetaldehyde per surface unit as a function of θ, i.e. NOx-(ads) surface coverage. The vertical error bars are determined by the minimum and maximum qirr values quantified from the repeated experiments. The horizontal error bars correspond to the uncertainty related to θ as reported in Table 2. It can be noticed in Figure 6 that the irreversibly adsorbed fraction decreases with the increase of TiO2 surface coverage by NOx-(ads) species. At high surface coverage, irreversibly adsorbed acetaldehyde (qirr) reaches a minimum value of 1.3 ± 0.5 µmol m-², which corresponds only to 12% of the irreversibly adsorbed amount on fresh TiO2 surface. This result emphasizes the negative impact of NO2 preliminary adsorption on acetaldehyde irreversible adsorption. The first hypothesis that can be proposed to explain this significant impact is the competitive adsorption between one NO2-(ads) and/or NO3-(ads) ad-species and one acetaldehyde molecule to access Ti4+ sites. If this hypothesis were valid, it would mean that the adsorption of one NOx-(ads) species would prevent the adsorption of one acetaldehyde molecule. However, a significant discrepancy can be pointed out regarding the corresponding ratio. Indeed, Table 3 reports the quantified amounts per surface unit of irreversibly adsorbed acetaldehyde for the various θ values. From these values –∆qirr has been calculated. It represents the absolute value of the difference between irreversibly adsorbed acetaldehyde (qirr) on fresh TiO2 surface and irreversibly adsorbed acetaldehyde (qirr) on corresponding NO2 exposed TiO2 surface (θ). In addition to that, the –∆qirr / q(NOx-(ads)) ratio is also calculated for each θ value. It represents the average number of acetaldehyde molecule prevented from adsorbing per one NOx-(ads) species present on TiO2 surface. Remarkably, at low θ, this ratio reaches values as high as 8 to 10, meaning that the presence of one NOx-(ads) species on the surface prevents the adsorption of 8 to 10 acetaldehyde molecules. Subsequently, a simple competitive adsorption is not satisfactory to explain the negative impact of NO2 ageing on acetaldehyde adsorption. Remarkably, the –∆ qirr / q(NOx-(ads)) ratio is not constant with the different investigated θ values. Indeed, it drops down to 2.2 for θ = 0.92 which is the lowest values. The impact of NOx-(ads) on acetaldehyde irreversible adsorption is remarkable; the observed phenomena may result from the combination of two main effects related to the pre-adsorption of NOx-(ads) species: (i)
According to the literature [13][14][16][30], acetaldehyde adsorption scarcely leads to the adsorption of single acetaldehyde molecules on TiO2 surface, but mostly produces 2butenal through β-aldolization and dehydration reactions. Such reactive adsorption indirectly promotes acetaldehyde adsorption on TiO2 surface by consuming adsorbed acetaldehyde. 2-butenal, as well as acetaldehyde, adsorbs on Ti4+ sites through carbonyl 8
function groups. Thus, it can be suggested that the competitive adsorption of NOx-(ads) on Ti4+ sites may also impact 2-butenal adsorption. Therefore, it can be suggested that the covering of one Ti4+ site by a NO2-(ads) or NO3-(ads) species may prevent the irreversible adsorption of two acetaldehyde molecules. Even if this consideration could not be sufficient to explain the massive negative impact of NOx-(ads) pre-adsorption on acetaldehyde adsorption, since it would decrease the –∆ qirr / q(NOx-(ads)) ratios only by a factor 2, it has to be considered as one of the possible impacts of pre-adsorbed NOx-(ads) species. (ii)
NOx-(ads) species may not only impact 2-butenal adsorption but also 2-butenal formation. As reported by Raskó and Kiss [13] 2-butenal is produced through aldol reaction by consuming one lattice oxygen O2-(lattice) from TiO2 surface. Furthermore, as reported in Equation 3, on TiO2 surface, the intermediate species NO+ necessarily consumes lattice oxygen O2-(lattice) to produce NO3-. Therefore, it can be proposed that the pre-adsorption of NO2 on TiO2 surface could significantly decrease the amount of available O2-(lattice) required for 2-butenal formation. As a consequence, 2-butenal formation could be hindered, which would considerably decrease irreversibly adsorbed acetaldehyde fraction.
Apart from irreversibly adsorbed acetaldehyde, the reversibly adsorbed fractions on aged TiO2 have also been quantified and discussed.
3.3.2 Impact of NO2 preliminary adsorption on acetaldehyde reversibly adsorbed fraction Figure 7 reports the adsorption isotherms of the reversibly adsorbed acetaldehyde fractions for different TiO2 surface ageing by NO2 at 296 K. Isotherms are obtained from acetaldehyde adsorption and flushing experiments as described in section 3.3.1. For each θ value, experiments have been performed for different acetaldehyde inlet concentrations ranging from 5 to 300 ppm. It can be assumed in Figure 7 that, irrespectively of the surface ageing θ, the reversibly adsorbed fraction (qrev) depends on the gas phase acetaldehyde concentration accordingly to Langmuir model [31], as described in Equation 6. Equation 6
࢘ࢋ࢜ = ࢘ࢋ࢜ ×
ࡷ. ାࡷ.
In Equation 6, qrev (µmol m-²) represents the amount of acetaldehyde adsorbed on TiO2 surface at equilibrium, qrevm (µmol m-²) is the maximum adsorbed amount corresponding to surface saturation, K represents the adsorption constant, and C is acetaldehyde gas-phase concentration at equilibrium. The fitting curves reported on Figure 7 were obtained from regression using Langmuir model on experimental data. Regressions are used to determine the values of K and qrevm. Figure 8 reports the evolution of qrevm and K as a function of θ, from fresh (θ = 0) to aged (θ = 0.92) TiO2 surface by NO2. It can be noticed on Figure 8 (a) and (b) that, compared to clean TiO2 surface, the various surface ageing strongly impact qrevm as well as K. Indeed, the presence of NOx-(ads) on the surface decreases the amount of reversibly adsorbed acetaldehyde even for the lowest values of θ. For instance, as soon as θ reaches 0.05, qrevm is decreased by 65%. Interestingly, qrevm slightly varies with increasing θ, but uncertainties on qrevm do not make possible the determination of any clear tendency. Considering that the reversibly adsorbed fraction of acetaldehyde mainly occurs on OH surface group, the first hypotheses to explain this behaviour would be the decrease of OH surface groups by NOx-(ads) species. Indeed, Hadjiivanov et al. [29] demonstrated that, on TiO2 surface, the adsorbed NO2 and/or NOX- species increase the surface acidity by consuming , the basic surface OH groups. Therefore, it can be proposed that the lowering of surface OH group density due to NOXspecies may decrease the fraction of reversibly adsorbed acetaldehyde. 9
Beside, a second hypothesis could be raised; indeed, the presence of NOx-(ads) species may considerably change the adsorption mode of reversibly adsorbed acetaldehyde. This hypothesis is supported by the observtion of K variations with θ. Surprisingly, the adsorption constant of acetaldehyde is multiplied by a factor ranging from 7 to 11 as soon as NOx-(ads) species are present on TiO2 surface. This phenomenon suggests that the interaction between acetaldehyde and aged TiO2 surface is probably changed in nature, compared to clean TiO2 surface. The ionic structure of TiO2 adsorbed NOx-(ads) species may induce stronger electrostatic interactions between acetaldehyde and the aged photocatalyst surface. This hypothesis is supported by the work of Rodriguez et al. [28] who reported that adsorbed NOX- species modify the thermochemical stability of O vacancies on TiO2 surface and facilitates their migration from bulk to the surface which could provide new adsorption sites. The investigation of TiO2 surface ageing and subsequent acetaldehyde adsorption using diffuse reflectance infrared spectroscopy could help identifying the nature of reversible adsorption sites.
10
Conclusions
The adsorption of acetaldehyde on TiO2 has been qualified and quantified on fresh and NO2 aged TiO2 surface. First, quantitative data have been obtained regarding acetaldehyde adsorption on fresh TiO2. Second, based on the reactive adsorption mechanism of Sivachandiran et al. [27], it has been evidenced that it is possible to control the coverage rate of TiO2 surface by NOx-(ads) species, namely NO2-(ads) or NO3-(ads). Finally, acetaldehyde adsorption parameters have been determined as a function of the surface coverage by NOx-(ads) species. Interestingly, the quantities of irreversibly and reversibly adsorbed fractions, as well as the adsorption constant K, are impacted by the preliminary adsorption of NO2. NO2 preliminary adsorption tends to decrease the amount of irreversibly adsorbed acetaldehyde. This effect is characterized by the –∆qirr / q(NOx-(ads) parameter and varies with NOx-(ads) species. The effect of direct competitive adsorption combined with the hindering of acetaldehyde condensation on the surface are proposed to explain the observed phenomenon. Similarly, the reversibly adsorbed fraction of acetaldehyde is negatively impacted by NO2 preliminary adsorption. The first hypothesis proposed to explain this behaviour if the consumption of OH groups by NOx-(ads) species. However, considering the sudden variation of the adsorption constant K for the lowest surface coverages by NOx-(ads) species, it is suggested that a new adsorption mode of acetaldehyde is induced by the presence of NOx-(ads) species on TiO2 surface, this could be confirmed using in-situ spectroscopy of adsorbed species such as DRIFTS (Diffuse Reflectance Infrared Fourier Transform Spectometry). This study quantitatively evidences for the first time the remarkable impact of NO2 preliminary adsorption on the consecutive adsorption of a VOC on TiO2. It is clearly shown that the presence of NOx-(ads) species on TiO2 surface considerably modifies the surface chemistry and the subsequent VOC sorption properties of TiO2. As a consequence, this result questions the long term behaviour of TiO2 when used as a coupling material in-side or downstream a plasma discharge. Considering that, in the case of plasma-material coupling, the role of TiO2 is highly related to its sorption properties, long term performances of such processes should be addressed. Similarly, in catalytic or photocatalytic processes, the decrease of VOC adsorption on TiO2 may considerably hinder the global conversion rate of VOCs. These aspects should be investigated in order to give new insights on the long term performances of heterogeneous oxidation processes under real operating conditions.
Acknowledgements Authors want to thank the Institut Carnot M.I.N.E.S. for its financial support in the framework of PhotoCair project, as well as Vincent Gaudion for his valuable technical assistance.
11
References [1] R.M. Alberici, M.C. Canela, M.N. Eberlin, W.F. Jardim, Catalyst deactivation in the gas phase destruction of nitrogen-containing organic compounds using TiO2/UV–VIS, Applied Catalysis B: Environmental 3-4 (2001) 389-397. [2] T.H. Lim, S.M. Jeong, S.D. Kim, J. Gyenis, Photocatalytic decomposition of NO by TiO2 particles , Journal of Photochemistry and Photobiology A 134 (2000) 209-217. [3] H. Wang, Z. Wu, W. Zhao, B. Guan, Photocatalytic oxidation of nitrogen oxides using TiO2 loading on woven glass fabric, Chemosphere 66 (2007) 185-190. [4] O. Debono, F. Thevenet, P. Gravejat, V. Hequet, C. Raillard, L. Lecoq, N. Locoge, Toluene photocatalytic oxidation at ppbv levels: Kinetic investigation and carbon balance determination , Applied Catalysis B: Environmental 106 (2011) 600-608. [5] H. Park, W. Choi, Effects of TiO2 Surface Fluorination on Photocatalytic Reactions and Photoelectrochemical Behaviors, Journal of Physical Chemistry B 108 (2004) 4086-4093. [6] S. Bahri, C.M. Jonsson, C.L. Jonsson, D. Azzolini, D.A. Sverjensky, R.M. Hazen, Adsorption and surface complexation study of LDOPA on rutile (α-TiO2) in NaCl solutions, Environmental Science & Technology 45 (2011) 3959-3966. [7] F. Thevenet, J. Couble, M. Brandhorst, J.L. Dubois, E. Puzenat, C. Guillard, D. Bianchi, Synthesis of Hydrogen Peroxide Using Dielectric Barrier Discharge Associated with Fibrous Materials, Plasma Chemistry and Plasma Processing 30 (2010) 489-502. [8] C. Marchand, B. Bulliot, S. L. Calvé, P. Mirabel, Aldehyde Measurements in indoor environmentin Strasbourg (France), Atmosperic Environment 40 (2006) 1336-1345. [9] Q. Liu, Y. Liu, M. Zhang, Personal exposure and source characteristics of carbonyl compounds and BTEXs within homes in Beijing, China , Building & Environment 61 (2013) 210-216. [10] R. Atkinson, Atmospheric chemistry of VOCs and NOx , Atmospheric Environment 34 (2000) 2063-2101 [11] C. Marchand, S. L. Calvé, P. Mirabel, N. Glasser, A. Casset, N. Schneider, F. de Blay, Indoor aldehydes concentrations and determinants in 162 homes in Strasbourg (France), Atmospheric Environment 42 (2008) 505-516 [12] J. M. Coronado, S. Kataoka, Isabel Tejedor-Tejedor, M. A. Anderson, Dynamic phenomena during the photocatalytic oxidation of ethanol and acetone over nanocrystalline TiO2: simultaneous FTIR analysis of gas and surface species, Journal of Catalysis 219 (2003) 219-230 [13] J. Raskó, J. Kiss, Adsorption and surface reactions of acetaldehyde on TiO2 , CeO 2 and Al2O3, Applied Catalysis A: General 287 (2005) 252-260 [14] M. Singh, N. Zhou, D. K. Paul, K. J. Klabunde, IR spectral evidence of aldol condensation: Acetaldehyde adsorption over TiO2 surface, Journal of Catalysis 260 (2008) 371-379 [15] F. Batault, F. Thevenet, V. Hequet, C. Rillard, L. Le Coq, N. Locoge, Acetaldehyde and acetic acid adsorption on TiO2 under dry and humid conditions, Chemical Engineering Journal 264 (2015) 197-210 [16] Z. Topalian, B. I. Stefanov, C. G. Granqvist, L. Österlund, Adsorption and photo-oxidation of acetaldehyde on TiO2 and sulfatemodified TiO2: Studies by in situ FTIR spectroscopy and micro-kinetic modeling, Journal of Catalysis 307 (2013) 265-274 [17] H.H. Kim, S.M. Oh, A. Ogata, S. Futamura, Decomposition of gas-phase benzene using plasma-driven catalyst (PDC) reactor packed with Ag/TiO2 catalyst, Applied Catalysis B: Environmental 56 (2005) 213–220. [18] J.V. Durme, J. Dewulf, W. Sysmans, C. Leys, H.V. Langenhove, Combining non-thermal plasma with heterogeneous catalysis in waste gas treatment: A review, Applied Catalysis B: Environmental 74 (2007) 161-1690. [19] X. Fan, T.L. Zhu, M.Y. Wang, X.M. Li, Removal of low-concentration BTX in air using a combined plasma catalysis system, Chemosphere 75 (2009) 1301-1306. [20] H.H. Kim, Nonthermal Plasma Processing for Air-Pollution Control: A Historical Review, Current Issues, and Future Prospects, Plasma Process and Polymers 1 (2004) 91-110.
12
Figure 1 - General scheme of the adsorption experimental setup.
Figure 2 - FTIR gas phase spectra of CH3CHO, N2O, NO2 and NO. The framed regions are used for quantifications.
Figure 3 - Experimental procedure used to assess the influence of NO2 preliminary adsorption on TiO2 surface regarding acetaldehyde adsorption. Acetaldehyde adsorption is carried out on fresh and NO 2 aged TiO2 surface under dark condition, dry air and at 296 K.
Figure 4 - Temporal profiles of NO2 and NO monitored at the reactor downstream under dark conditions and at 296 K. For adsorption, 30 ppm of NO2 and 5 ppm of N2O balanced with zero air are concurrently sent at the -1 reactor inlet with a total flow rate of 600 mL min .
Figure 5 - Acetaldehyde breakthrough and flushing curves overlaid with N2O system mixing curves on clean TiO2 surface. During adsorption, 20 ppm of acetaldehyde and 5 ppm of N2O are concurrently sent to the -1 reactor inlet. Adsorption is performed under dry air with 600 mL min total flow rate, at 296 K and under dark conditions.
Figure 6 - Irreversibly adsorbed acetaldehyde fractions on TiO 2 as a function of i.e. NOx (ads) surface coverage. Acetaldehyde adsorption on fresh and aged TiO2 is performed under dry air and at 296 K. -
Figure 7 - Acetaldehyde adsorption on fresh and aged TiO 2 at 296 K. Adsorption isotherms of the reversibly adsorbed fraction (qrev) are determined for different values: (a) 0, (b) 0.05, (c) 0.1, (d) 0.55 and (e) 0.92. Acetaldehyde inlet concentration was varied from 5 to 300 ppm.
Figure 8 - Evolution of a) maximum reversibly adsorbed amounts of acetaldehyde (qrevm), and b) reversibly adsorbed acetaldehyde adsorption constant (K), reported as a function of NO2 ageing on TiO2 surface, i.e. . These data are determined from Langmuir model regression using experimental isotherms.
[21] C.J. Weschler, H.C. Shields, D.V. Naik, Indoor Chemistry Involving O3, NO, and NO 2 as Evidenced by 14 Months of Measurements at a Site in Southern California, Environmental Science & Technology 28 (1994) 2120-2132. [22] A. Chaloulakou, I. Mavroidis, I. Gavriil, Compliance with the annual NO2 air quality standard in Athens. Required NOx levels and expected health implications, Atmospheric Environment 42 (2008) 454-465. [23] S. Matsuda, H. Hatano, A. Tsutsumi, Ultrafine particle fluidization and its application to photocatalytic NOx treatment, Chemical Engineering Journal 82 (2001) 183-188. [24] F.L. Toma, G. Bertrand, D. Klein, C. Coddet, Photocatalytic removal of nitrogen oxides via titanium dioxide, Environmental Chemistry Letters 2 (2004) 117-121. [25] K. I. Hadjiivanov, Identification of neutral and charged NxOy surface species by IR spectroscopy, Catalysis Review 42 (2000) 71144 [26] V. H. Grassian, Heterogeneous uptake and reaction of nitrogen oxides and volatile organic compounds on the surface of atmospheric particles including oxides, carbonates, soot and mineral dust: Implications for the chemical balance of the troposphere, International Review in Physical Chemistry 20 (2001) 467-548 [27] L. Sivachandiran, F. Thevenet, P. Gravejat, A. Rousseau, Investigation of NO and NO2 Adsorption Mechanism on TiO2 at Room Temperature, Applied Catalysis B: Environmental 142-143 (2013) 196-204, 2013. [28] J. A. Rodriguez, T. Jirsak, G. Liu, J. Hrbek, J. Dvorak, A. Maiti, Chemistry of NO2 on oxide surfaces: formation of NO 3 on TiO2(110) and NO2 O vacancy interactions, Journal of American Chemical Society 123 (2001) 9597-9605 [29] K.I. Hadjiivanov, V. Bushev, M. Kantcheva, D. Klissurski, Infrared spectroscopy study of the species arising during nitrogen dioxide adsorption on titania (anatase), Langmuir 10 (1994) 464-471 [30] J. E. Rekoske, M. A. Barteau, Kinetics, Selectivity, and deactivation in the aldol condensation of acetaldehyde on anatase titanium dioxide, Industrial and Engineering Chemistry Research 50 (2010) 41-51 [31] I. Langmuir, The adsorption of gases on plane surfaces of glass, mica and platinum, Journal of American Chemical Society 40 (1918) 1361-1403
13
bypass air
AZ
PSA
MFC 2
N2O
FTIR
MFC 1 vent
MFC 3
NO2
vent
thermocouple
MFC 4 oven
CH3CHO
TiO2 quartz wool reactor part
Figure 1 - General scheme of the adsorption experimental setup.
Absorbance (a. u.)
CH3CHO
Absorbance (a. u.)
1900
1800
20 ppmv
1700
1600
1500
1400
NO2
1700
N2O
1300
2 ppmv
1650
1600
1550
5 ppmv
2300
2260
2220
2180
NO
1500
2000
2140
2 ppmv
1950
1900
1850
1800
1750
1700
Wavenumbers (cm-1)
Figure 2 - FTIR gas phase spectra of CH3CHO, N2O, NO2 and NO. The framed regions are used for quantifications.
Thermal treatment Dry Air, 703 K
Surface coverages by NO2 30 ppm dry air, 296K OR
CH3CHO breakthrough Dry Air, 296 K
Flushing Dry Air, 296 K
Figure 3 - Experimental procedure used to assess the influence of NO2 preliminary adsorption on TiO2 surface regarding acetaldehyde adsorption. Acetaldehyde adsorption is carried out on fresh and NO2 aged TiO2 surface under dark condition, dry air and at 296 K.
30 system mixing curve
CC (ppm) (ppm)
NO2
20
10 NO
0
0
30
60
90
120
150
Time (min) Time (min) Figure 4 - Temporal profiles of NO2 and NO monitored at the reactor downstream under dark conditions and at 296 K. For adsorption, 30 ppm of NO2 and 5 ppm of N2O balanced with zero air are concurrently sent at the reactor inlet with a total flow rate of 600 mL min-1.
[acetaldehyde] (ppm)
system mixing curve
20 15
10 5 0 0
50
100
150
Time (min) Figure 5 - Acetaldehyde breakthrough and flushing curves overlaid with N2O system mixing curves on clean TiO2 surface. During adsorption, 20 ppm of acetaldehyde and 5 ppm of N2O are concurrently sent to the -1 reactor inlet. Adsorption is performed under dry air with 600 mL min total flow rate, at 296 K and under dark conditions.
Figure 6 - Irreversibly adsorbed acetaldehyde fractions on TiO 2 as a function of i.e. NOx (ads) surface coverage. Acetaldehyde adsorption on fresh and aged TiO2 is performed under dry air and at 296 K. -
a) = 0
b) = 0.05
c) = 0.1
d) = 0.55
e) = 0.92
q rev (µmol/m²)
Acetaldehyde (ppm)
Figure 7 - Acetaldehyde adsorption on fresh and aged TiO2 at 296 K. Adsorption isotherms of the reversibly adsorbed fraction (qrev) are determined for different values: (a) 0, (b) 0.05, (c) 0.1, (d) 0.55 and (e) 0.92. Acetaldehyde inlet concentration was varied from 5 to 300 ppm.
a)
b)
Figure 8 - Evolution of a) maximum reversibly adsorbed amounts of acetaldehyde (qrevm), and b) reversibly adsorbed acetaldehyde adsorption constant (K), reported as a function of NO2 ageing on TiO2 surface, i.e. . These data are determined from Langmuir model regression using experimental isotherms.
Table 1 - FTIR detection limits (DL) for the species of interest.
-
Table 2 - Amounts of NO2 consumed per surface unit for various adsorption times. Total q(NOx (ads)) is calculated as the total amount of NO2 (ads) and NO3 (ads) per surface unit. Covering of TiO2 surface by NO2 is represented by the parameter varying from 0 to 1.
Table 3 - Acetaldehyde adsorbed amounts on fresh and NO2 exposed TiO2 surface. Decreases in adsorbed acetaldehyde amounts –qirr and –qirr / q(NOx (ads)) are determined from fresh and NO2 exposed acetaldehyde adsorption data.
Compound NO NO2 N2O CH3CHO
DL (ppb) 700 50 50 500
Table 1 - FTIR detection limits (DL) for the species of interest.
Adsorption time (s)
Total consumed NO2 (µmol m ²)
Total q(NOx (ads)) (µmol m ²)
150
0.19 ± 0.1
0.19 ± 0.05
0.05 ± 0.02
300
0.38 ± 0.05
0.38 ± 0.05
0.10 ± 0.02
1450
3.03 ± 0.3
2.02 ± 0.15
0.55 ± 0.05
2400
5.01 ± 0.4
3.34 ± 0.27
0.92 ± 0.05
5000
5.47 ± 0.4
3.65 ± 0.27
1
-
Table 2 - Amounts of NO2 consumed per surface unit for various adsorption times. Total q(NOx (ads)) is calculated as the total amount of NO2 (ads) and NO3 (ads) per surface unit. Covering of TiO2 surface by NO2 is represented by the parameter varying from 0 to 1.
Acetaldehyde qirr
-Δqirr
µmol m ²
µmol m ²
0.00
8.61 ± 0.28
0
Ø
0.05
7.07 ± 0.34
1.52
8.0
0.10
4.66 ± 0.68
3.93
10.3
0.55
1.65 ± 0.09
6.94
3.4
0.92
1.29 ± 0.41
7.3
2.2
θ
-
-
-Δqirr / q(NOx-(ads))
Table 3 - Acetaldehyde adsorbed amounts on fresh and NO2 exposed TiO2 surface. Decreases in adsorbed acetaldehyde amounts –qirr and –qirr / q(NOx (ads)) are determined from fresh and NO2 exposed acetaldehyde adsorption data.
Evolution of acetaldehyde irreversibly adsorbed fraction (qirr) on TiO2 at 296 K as a function of TiO2 surface coverage by NOx- adsorbed species () obtained by preliminary NO2 adsorption on TiO2.
Acetaldehyde adsorption parameters on TiO2 are quantitatively determines TiO2 surface coverage by NO2 derived adsorbed species has been controlled NO2 preliminary adsorption considerably impacts acetaldehyde adsorption parameters
S1385-8947(15)00924-9 http://dx.doi.org/10.1016/j.cej.2015.06.084 CEJ 13856
To appear in:
Chemical Engineering Journal
Received Date: Revised Date: Accepted Date:
12 May 2015 17 June 2015 18 June 2015
Please cite this article as: F. Thevenet, L. Olivier, F. Batault, L. Sivachandiran, N. Locoge, Acetaldehyde adsorption on TiO2: influence of NO2 preliminary adsorption, Chemical Engineering Journal (2015), doi: http://dx.doi.org/ 10.1016/j.cej.2015.06.084
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Acetaldehyde adsorption on TiO2: influence of NO2 preliminary adsorption F.Thevenet, L. Olivier, F. Batault, L. Sivachandiran, N. Locoge Mines Douai, SAGE, F-59508 Douai, France Abstract Titanium dioxide (TiO2) is a widespread metal oxide used in catalysis and photocatalysis, or coupled to non-thermal plasma for volatile organic compound (VOC) oxidation. Adsorption is a key step in the advanced heterogeneous oxidation processes; adsorption is partially addressed in that it is mainly investigated on fresh TiO2 surfaces. However, the treatment of real effluents combines various pollutants with VOCs; among them, NOx are characterized by reactive adsorption properties on TiO2 leading to irreversibly adsorbed species on the surface. The aim of this work is to determine the impact of NO2 preliminary adsorption on the subsequent adsorption of a model VOC: acetaldehyde. Breakthrough methods are used to (i) characterize acetaldehyde adsorption on fresh and NO2 covered TiO2 surface, (ii) control the coverage of TiO2 surface by irreversibly adsorbed NOx(ads) species. In a first step, acetaldehyde adsorption on TiO2 has been characterized. In a second step, acetaldehyde adsorption has been achieved on TiO2 surface after NO2 exposure corresponding to different surface coverages. Then, acetaldehyde adsorption parameters have been determined. Subsequently, it has been possible to plot the evolution of acetaldehyde reversibly and irreversibly adsorbed quantities, and reversible fraction adsorption constant, as a function of TiO2 surface coverage by NOx-(ads) species, so called TiO2 surface NO2 ageing. Interestingly, both adsorbed acetaldehyde fractions are highly impacted by the presence of NOx-(ads) species on TiO2 surface. The irreversibly adsorbed fraction is considerably decreased since the lowest values of TiO2 surface coverage by NOx-(ads) species. Hypotheses related to competitive adsorption are not sufficient to explain the observed impact, suggesting that NOx-(ads) species modify TiO2 surface chemistry and acetaldehyde reactive adsorption. The reversibly adsorbed fraction of acetaldehyde is highly impacted as well; reversibly adsorbed amounts are decreased since the lowest surface coverage by NOx-(ads) species. The corresponding adsorption constants are abruptly increased as soon as TiO2 surface has been exposed to NO2, suggesting the formation of a new adsorption mode for acetaldehyde on NO2 exposed TiO2 surface. This paper evidences the considerable impact of NO2 on TiO2 adsorption properties regarding VOCs. Considering that NOx may accumulate on TiO2 the long term behaviour of such processes should be investigated further taking into account the NOx vs. VOC interaction for adsorption.
Keywords: adsorption; TiO2; NO2; acetaldehyde; surface coverage
1
1. Introduction In the field of advanced heterogeneous oxidation processes dedicated to pollution remediation, adsorption plays a key role regarding the system performances. Among various reactive solids, metal oxides are widely used as sorbent, catalyst, or plasma coupling materials. The purpose of these processes is generally to induce oxidation, and even mineralization, of organic species. Thermal-catalysis, photocatalysis and plasma-catalysis are the most widespread techniques. Adsorption and desorption of pollutants from solid surfaces are key steps in these heterogeneous techniques. Indeed, adsorption makes possible the catalytic treatment of pollutants; similarly, it expands the residence time of compounds to be treated in plasma-catalyst combination. In addition, desorption of obtained products makes possible the renewal of pollutants on the surface and prevents from poisoning. Nevertheless, in both catalytic and plasma-catalytic methods, the catalyst deactivation has been reported by several authors [1][2][3]. Three kinds of species are typically pointed out to explain the deactivation of oxidation heterogeneous processes: NOx, COx and reaction intermediates [3][4]. NOx and COx may be present in the effluent to be treated or produced by the process itself. Among various metal oxides, TiO2 is a material of interest combining catalytic and photocatalytic properties. Moreover, due to its high relative permittivity, it is classically coupled to plasma devices for environmental applications. Interestingly, it is characterized by distinctive surface properties regarding adsorption and complexation [5][6][7]. However, irrespectively of the oxidation technique, the long term performances of this material is poorly addressed whereas knowledge improvements on pollutant adsorption and desorption may be fruitful to understand the system performances and ageing. Acetaldehyde is a typical VOC of indoor air, it is reported in almost all kinds of buildings with concentrations around 18 µg.m-3 (93 ppb) [8]. Exceptionally high levels may reach 140 µg.m-3 (720 ppb) [9]. The main sources of acetaldehyde are glues, paints, deodorants and fuel additives[8][10]. Moreover, it is frequently reported that the oxidation of hydrocarbons (HC) leads to the formation of acetaldehyde as a side product [11][12][13][14].Qualitative studies have been dedicated to the adsorption of acetaldehyde on TiO2 surface; two adsorption modes are generally reported at ambient temperature: (i) a weak adsorption mode on hydroxyl groups present on TiO2 surface and (ii) a stronger adsorption mode on Ti4+ sites (Lewis acidic sites). Batault et al. [15] investigated quantitatively acetaldehyde adsorption on clean TiO2 surface. They related the observed reversible and irreversible adsorbed fraction with the reported adsorption modes. Interestingly, acetaldehyde adsorption on TiO2 is a reactive process since two acetaldehyde adsorbed molecules may undergo an aldolic condensation into 2-butenal via 3-hydroxy-butanal as an intermediate [13][14][16]. NOx (NO and NO2) are formed in thermal power plants or combustion engines. Besides, when non-thermal plasma is ignited in air, it produces higher oxides of nitrogen such as NO, N2O, NO2 and HNO3, in concentrations ranging from some hundreds of ppb to some hundreds of ppm, depending on the plasma deposited energy [17][18][19]. In non-thermal plasma performed under pure and dry air, most of NOx are contributed by NO2 because of the fast oxidation of NO by plasma produced O3 [20]. In indoor air, levels of NO and NO2 respectively range from tens of ppb to few hundred ppb [21]. National Institute for Occupational Safety and Health (NIOSH) suggested that the Immediately Dangerous to Life and Health (IDLH) concentration of NO2 is 20 ppm. Studies focused on risk assessment have shown that high outdoor NO2 concentration observed in residential areas contributes to increased respiratory and cardiovascular diseases and mortality [22]. Air treatment systems based on advanced heterogeneous oxidation aim at converting VOCs; however, the material surface is exposed to the wide diversity of species present in the air stream or generated by the system itself. Among them, NOx may dramatically influence VOC adsorption processes and decomposition mechanisms [23][24], which may also impact the system efficiency 2
and ageing. Former works did not extensively study the interactions between VOCs and such copollutants into heterogeneous advanced oxidation processes, especially the crucial adsorption step. At room temperature, under dark condition and under dry air flow, NO does not adsorb on TiO2 surface. In contrast, without any UV radiation, under dry air and at 296 K, NO2 is known to undergo reactive adsorption on TiO2 surface. This leads to the formation of adsorbed species. Gaseous NO may also be evolved along NO2 reactive adsorption process on TiO2 [25][26][27][28]. Interestingly, Sivachandiran et al. [27] recently proposed the following mechanism to describe NO2 adsorption on TiO2 surface under dark conditions ( Equation 1 to Equation 3): Equation 1
2 NO2 (gas)
2 NO2 (ads)
Equation 2
2 NO2 (ads)
NO3- (ads) + NO+(ads)
Equation 3
NO
+ (ads)
2-
+O
-
NO2
(TiO2 S.L.)
(ads)
Authors evidenced that, as soon as the TiO2 surface coverage “θ" by NO2-(ads) and NO3-(ads) exceeds 0.2, then the reaction corresponding to Equation 4 occurs: Equation 4
-
NO2
(ads)
-
+ NO2 (ads)
NO3
(ads)
+ NO (gas)
The global mechanism can be summarized as follows (Equation 5): Equation 5
3 NO2 (ads) + O2- (TiO2 S.L.)
2 NO3- (ads) + NO (gas)
From adsorbed NO2, various species are produced: some are reaction intermediates and subsequently do not accumulate on the TiO2 surface, like the highly reactive nitrosonium ion NO+ and nitrite ion NO2-; some are evolved in the gas phase, like NO; and some are irreversibly adsorbed on TiO2 like the nitrate species produced from Equation 2 and Equation 4. Both nitrate species produced are reported as irreversibly adsorbed on Ti4+ sites [25][26][27][28]. As a consequence, NO2 adsorption onto TiO2 leads to the formation of irreversibly adsorbed nitrates species which could be considered as poisons for the catalyst. Therefore, NO2 adsorption on the surface may be responsible for TiO2 surface deactivation, or ageing. The main impact of such adsorbed species on the photocatalyst surface could be the decrease of VOC sorption which may considerably hinder the very first step of any catalytic process. The objective of the present work is to assess the impact of NO2 on the adsorption of acetaldehyde on TiO2 surface under dark conditions. In a first step, an experimental procedure is proposed to control TiO2 surface coverage, i.e. ageing, by NO2. Then, single acetaldehyde adsorption is investigated on fresh TiO2 surface in order to quantify (i) the total adsorbed amount, (ii) the reversibly and irreversibly adsorbed fractions of acetaldehyde, and (iii) acetaldehyde adsorption constant. Finally, acetaldehyde adsorption parameters are determined and discussed as a function of TiO2 surface coverage by NO2 ad-species.
3
2. Experimental 2.1 Materials and gas stream preparation The TiO2 P25-Degussa material (70% anatase, 30% rutile) used in the present work was provided and used as a powder. The specific area of the material has been measured using the BET (Brunauer, Emmett and Teller) N2 adsorption method at 77 K and the determined value is 57 ± 3 m² g-1. Typical photocatalyst mass used for each experiment is 15.0 ± 0.05 mg. Gas cylinders, 99.9% purity, are provided by Praxair (Belgium) with nominal concentrations ranging from 35 to 500 ppm balanced in N2. Experimental air stream is provided via two coupled devices fed by a classical air compressor. The compressed air is first sent into Claind (Lenno, Italy) AZ-2020 catalytic zero air generators, which is coupled to a Pressure Swing Adsorption (PSA) device. The remaining impurity levels in the air stream are lower than the analytical system detection limits: VOCs < 0.1 ppb, CO2 < 10 ppb and CO < 80 ppb. Moisture level is about 2 ppm. This air stream is used for infrared spectrometers purge and as carrier gas during each experiment. 2.2 Adsorption setup The experimental bench used for this study is presented in Figure 1. It consists in three parts: (i) gas generation setup, (ii) reactor and (iii) analytical device. The gas generation part includes the gas cylinders and zero air generation system, coupled to mass flow controllers (MFCs). The reactor is made of a U-shaped vertical quartz tube with 4 and 6 mm inner and outer diameter respectively. The catalyst bed is positioned in a widened section of the U-shaped tube of 15 mm inner diameter and maintained with a quartz wool plug. The reactor is inserted in a tubular oven and the temperature of the catalyst bed is monitored using a K-type thermocouple set in contact with the catalytic bed.
2.3 Analytical device The Fourier Transform Infrared spectrometer (FTIR) is equipped with a 10 m long path cell and liquid N2 cooled Mercury Cadmium Telluride (MCT) detector with 0.5 cm-1 spectral resolution. Spectra are acquired using Result-3 software with 16 scans per spectrum. Quantifications and data processing were performed using TQ-Analyst software. The detection limits for the species of interest are reported in Table 1. As reported in Figure 2, the quantification of CH3CHO, NO, NO2 and N2O have been performed using the framed absorbance bands. More precisely, acetaldehyde has been quantified with the Q-branch associated to ν(C=O), between 1745 cm-1 and 1743 cm-1 and with the Q-branch associated to δ(CH3), between 1396 cm-1 and 1394 cm-1. N2O has been quantified with the R-branch of the band associated to ν(N-N) between 2223 cm-1 and 2260 cm-1. NO has been quantified with the R-branch of the band associated to ν(N=O), between 1936 cm-1 and 1878.6 cm-1. NO2 has been quantified with the P-branch of the band associated to νas (NO2) between 1606 cm-1 and 1580 cm-1. 2.4 Experimental Procedure Figure 3 represents the cyclic experimental procedure used to assess the influence of NO2 on TiO2 surface regarding acetaldehyde adsorption. For all the experiments the total flow rate has been kept constant as 600 mL.min-1 unless otherwise mentioned: (i) cleaning of the catalyst surface by maintaining the reactor at 703 K under air for 1h, (ii) surface ageing of TiO2 by adsorbing 30 ppm of NO2 in dry air, (iii) acetaldehyde adsorption until breakthrough, (iv) flushing the reactor under dry air. In particular, the step (ii) is skipped when acetaldehyde adsorption has to be performed on 4
fresh TiO2 surface. Moreover, to estimate the acetaldehyde adsorption isotherm, the step (iii) has been performed for various acetaldehyde concentrations like 5, 10, 15, 20, 50, 100, 200, and 300 ppm. All the experiments have been performed under dark condition and repeated three times in order to evaluate their reproducibility. It has been formerly evidenced that N2O does not show any adsorption on TiO2 surface under our experimental conditions [27]. Thus, N2O is used in the experiments to determine the system mixing curve in order to achieve accurate quantifications of adsorbed species using the classical breakthrough method.
5
3. Results and discussion
The influence of adsorbed NO2 on acetaldehyde adsorption over TiO2 is investigated. Firstly, NO2 adsorption is performed on clean TiO2 surface and the surface coverage is determined for different NO2 adsorption times. Secondly, acetaldehyde is adsorbed on fresh and/or NO2 aged TiO2 surface. Moreover, for all experiments the adsorbed amounts of acetaldehyde are quantified and the adsorption constants are determined and discussed.
3.1. Control of TiO2 surface coverage: NO2 adsorption on TiO2 at 296 K Figure 4 shows the temporal profiles of NO2 and N2O breakthrough curves and NO transient formation curve on fresh TiO2 surface. NO2 adsorption on TiO2 is performed under dark condition and in dry air with a constant flow rate of 600 mL min-1. At the reactor inlet, 30 ppm of NO2 and 5 ppm of tracer N2O are concurrently introduced. As can be seen in Figure 4, during adsorption, NO2 concentration at the reactor downstream increases until the surface is saturated. As demonstrated by Sivachandiran et al. [27] NO is evolved in the gas phase after 6 minutes of adsorption. And the NO concentration reaches the maximum value between 15 and 20 minutes of adsorption. When the surface tends to reach saturation, NO concentration gradually decreases. In particular, it reaches zero when the surface is completely saturated. Both NO2 and NO temporal profiles are overlaid with the system mixing curve obtained from the temporal profile of N2O. Five different adsorption experiment sets have been carried out, they differ by their adsorption times ranging from 150 s to 5000 s. Each set of experiment has been repeated four times in order to assess the reproducibility. Considering the mechanism proposed by Sivachandiran et al. [27] and recalled in the introduction, the stable ad-species present on saturated surface are : (i) irreversibly adsorbed NO2(ads) and (ii) NO3 (ads) species. Regarding NO3 (ads) species, Hadjiivanov et al. [29] reported that NO2 adsorption on TiO2 surface leads to the formation of monodentate as well as bidentate NO3-(ads) species on Ti4+ sites. Although NO2-(ads) is a reaction intermediate, it coexists with NO3-(ads) on partially covered surface. Therefore, in the following discussion the notation of NOx-(ads) is used to represent the adsorbed NO2-(ads) and NO3-(ads) species. Subsequently, the coverage rate θ is used to characterize the surface coverage due to NO2 adsorption. The parameter θ is defined as the ratio between NOx-(ads) adsorbed amount along partial adsorption and the total amount of NOx-(ads) adsorbed on saturated TiO2 surface. As reported in Figure 4, from NO2 adsorption breakthrough curve, it is possible to determine the total amount of NO2 consumed on TiO2 surface and the total amount of evolved NO for various adsorption times. The total amount of NO2 consumed is obtained by integrating the area between the NO2 breakthrough curve and the N2O system mixing curve and the total amount is expressed in µmol m-2. Similarly, the amount of NO produced is quantified directly from NO concentration temporal profile. Consumed NO2 values are reported in Table 2. In the following, q(NOx-(ads) ) represents the sum of NO2-(ads) and NO3-(ads) adsorbed amounts per surface unit. As reported by Sivachandiran et al. [27] the proposed mechanism allows assessing the following points: (i)
There is a θ threshold regarding the production of NO in the gas phase.
(ii)
Below this θ threshold, the stoechiometric ratio between consumed NO2 and NOx(ads) is 2:2 (Equations 1 to 3).
(iii)
Beyond the θ threshold, the stoechiometric ratio is 3:2 (Equation 5). 6
Thus, q(NOx-(ads)) is quantified by multiplying the consumed NO2 amount by 1 or 2/3, depending on the adsorption process advancement. The values related to saturated surface are considered as references to calculate θ when θ < 1. The uncertainty values are determined from the dispersion of the measurements obtained during the repeated experiments. Table 2 reports the amounts of NO2 consumed per surface unit for the different adsorption times and the corresponding q(NOx-(ads)) amounts and θ values. Based on Table 2, four different surface coverages (0.05, 0.1, 0.55 and 0.92) can be considered to evaluate the impact of NO2 on acetaldehyde adsorption. The first two are lower than NO formation threshold and the last two are above NO formation threshold. Acetaldehyde adsorption parameters, i.e. irreversibly adsorbed fraction (qirr), reversibly adsorbed fraction (qrev) and adsorption constant (K) have been determined for each selected values of θ .
3.2 Acetaldehyde adsorption on clean TiO2 surface Figure 5 shows acetaldehyde breakthrough curve overlaid with the N2O system mixing curve during breakthrough and flushing steps on clean TiO2 surface. During adsorption, 20 ppm of acetaldehyde balanced with dry air is sent to the reactor inlet. The acetaldehyde concentration at the reactor outlet increases till the adsorption equilibrium is reached. Then the reactor is flushed under dry air. During flushing, the reversibly adsorbed acetaldehyde fraction desorbs. Flushing is stopped as the acetaldehyde concentration returns to zero. The total amount of acetaldehyde adsorbed, i.e. irreversibly and reversibly adsorbed fractions, at equilibrium is quantified by integrating the area between the acetaldehyde breakthrough curve and the system mixing curve. Similarly, the reversibly adsorbed amount is quantified by calculating the area between both curves during flushing. The irreversibly adsorbed amount is then quantified by subtracting the reversibly adsorbed amount from the total adsorbed amount. Obtained values on fresh TiO2 surface are 8.61 µmol m-² ± 10% for the irreversibly adsorbed fraction and 0.42 µmol m-² ± 10% for the reversibly adsorbed fraction, corresponding to a total adsorbed amount of 9.03 µmol m-² ± 15%. These experimental data obtained with 20 ppm inlet concentration of acetaldehyde are consistent with the recent work of Batault et al. [15] and the corresponding adsorption isotherms. This result confirms that acetaldehyde adsorption is mostly irreversible (95%), on clean TiO2 surface, and highlights the predominance of strong chemical interaction between acetaldehyde oxygen and Ti4+ surface Lewis acid sites [13]. Furthermore, adsorption of acetaldehyde on clean TiO2 surface has been investigated with acetaldehyde inlet concentrations ranging from 5 to 300 ppm. For each inlet concentration, reversible and irreversible fractions have been quantified, they exhibit contrasted behaviours. Indeed, the irreversibly adsorbed fractions is poorly influenced by acetaldehyde inlet concentration, which suggests that even with 5 ppm inlet concentration, all irreversible adsorption sites can be covered. In contrast, the reversibly adsorbed fraction of acetaldehyde shows a typical Langmuir isotherm behaviour with respect to acetaldehyde inlet concentration. This isotherm, determined on clean TiO2 surface, is reported on Figure 7 (a). It enabled the calculation of the maximum reversibly adsorbed amount on clean TiO2 surface as : qrevm = 1.6 ± 0.2 µmol m-² and the adsorption constant of reversibly adsorbed acetaldehyde on clean TiO2 surface as 0.010 ± 0.002 ppm-1.
7
3.3 Acetaldehyde adsorption on preliminary NO2 exposed TiO2 surface Acetaldehyde adsorption and desorption have been performed for the various selected values of θ, i.e. various surface coverages by NOx-(ads) species. Acetaldehyde, NO, NO2 and N2O concentrations are monitored at the reactor downstream. For each θ values reported in Table 2, acetaldehyde adsorption has been repeated three times. It has to be noted that neither NO nor NO2 have been monitored at the reactor outlet during acetaldehyde adsorption on NOx-(ads) covered TiO2 surface. This finding emphasizes the fact that acetaldehyde adsorption does not affect the chemisorbed NO2- and/or NO3- species on TiO2 surface under our experimental conditions. For each experiment both irreversibly and reversibly adsorbed fractions of acetaldehyde have been quantified and correlated to θ values.
3.3.1 Impact of NO2 preliminary adsorption on acetaldehyde irreversibly adsorbed fraction Figure 6 shows the amount of irreversibly adsorbed acetaldehyde per surface unit as a function of θ, i.e. NOx-(ads) surface coverage. The vertical error bars are determined by the minimum and maximum qirr values quantified from the repeated experiments. The horizontal error bars correspond to the uncertainty related to θ as reported in Table 2. It can be noticed in Figure 6 that the irreversibly adsorbed fraction decreases with the increase of TiO2 surface coverage by NOx-(ads) species. At high surface coverage, irreversibly adsorbed acetaldehyde (qirr) reaches a minimum value of 1.3 ± 0.5 µmol m-², which corresponds only to 12% of the irreversibly adsorbed amount on fresh TiO2 surface. This result emphasizes the negative impact of NO2 preliminary adsorption on acetaldehyde irreversible adsorption. The first hypothesis that can be proposed to explain this significant impact is the competitive adsorption between one NO2-(ads) and/or NO3-(ads) ad-species and one acetaldehyde molecule to access Ti4+ sites. If this hypothesis were valid, it would mean that the adsorption of one NOx-(ads) species would prevent the adsorption of one acetaldehyde molecule. However, a significant discrepancy can be pointed out regarding the corresponding ratio. Indeed, Table 3 reports the quantified amounts per surface unit of irreversibly adsorbed acetaldehyde for the various θ values. From these values –∆qirr has been calculated. It represents the absolute value of the difference between irreversibly adsorbed acetaldehyde (qirr) on fresh TiO2 surface and irreversibly adsorbed acetaldehyde (qirr) on corresponding NO2 exposed TiO2 surface (θ). In addition to that, the –∆qirr / q(NOx-(ads)) ratio is also calculated for each θ value. It represents the average number of acetaldehyde molecule prevented from adsorbing per one NOx-(ads) species present on TiO2 surface. Remarkably, at low θ, this ratio reaches values as high as 8 to 10, meaning that the presence of one NOx-(ads) species on the surface prevents the adsorption of 8 to 10 acetaldehyde molecules. Subsequently, a simple competitive adsorption is not satisfactory to explain the negative impact of NO2 ageing on acetaldehyde adsorption. Remarkably, the –∆ qirr / q(NOx-(ads)) ratio is not constant with the different investigated θ values. Indeed, it drops down to 2.2 for θ = 0.92 which is the lowest values. The impact of NOx-(ads) on acetaldehyde irreversible adsorption is remarkable; the observed phenomena may result from the combination of two main effects related to the pre-adsorption of NOx-(ads) species: (i)
According to the literature [13][14][16][30], acetaldehyde adsorption scarcely leads to the adsorption of single acetaldehyde molecules on TiO2 surface, but mostly produces 2butenal through β-aldolization and dehydration reactions. Such reactive adsorption indirectly promotes acetaldehyde adsorption on TiO2 surface by consuming adsorbed acetaldehyde. 2-butenal, as well as acetaldehyde, adsorbs on Ti4+ sites through carbonyl 8
function groups. Thus, it can be suggested that the competitive adsorption of NOx-(ads) on Ti4+ sites may also impact 2-butenal adsorption. Therefore, it can be suggested that the covering of one Ti4+ site by a NO2-(ads) or NO3-(ads) species may prevent the irreversible adsorption of two acetaldehyde molecules. Even if this consideration could not be sufficient to explain the massive negative impact of NOx-(ads) pre-adsorption on acetaldehyde adsorption, since it would decrease the –∆ qirr / q(NOx-(ads)) ratios only by a factor 2, it has to be considered as one of the possible impacts of pre-adsorbed NOx-(ads) species. (ii)
NOx-(ads) species may not only impact 2-butenal adsorption but also 2-butenal formation. As reported by Raskó and Kiss [13] 2-butenal is produced through aldol reaction by consuming one lattice oxygen O2-(lattice) from TiO2 surface. Furthermore, as reported in Equation 3, on TiO2 surface, the intermediate species NO+ necessarily consumes lattice oxygen O2-(lattice) to produce NO3-. Therefore, it can be proposed that the pre-adsorption of NO2 on TiO2 surface could significantly decrease the amount of available O2-(lattice) required for 2-butenal formation. As a consequence, 2-butenal formation could be hindered, which would considerably decrease irreversibly adsorbed acetaldehyde fraction.
Apart from irreversibly adsorbed acetaldehyde, the reversibly adsorbed fractions on aged TiO2 have also been quantified and discussed.
3.3.2 Impact of NO2 preliminary adsorption on acetaldehyde reversibly adsorbed fraction Figure 7 reports the adsorption isotherms of the reversibly adsorbed acetaldehyde fractions for different TiO2 surface ageing by NO2 at 296 K. Isotherms are obtained from acetaldehyde adsorption and flushing experiments as described in section 3.3.1. For each θ value, experiments have been performed for different acetaldehyde inlet concentrations ranging from 5 to 300 ppm. It can be assumed in Figure 7 that, irrespectively of the surface ageing θ, the reversibly adsorbed fraction (qrev) depends on the gas phase acetaldehyde concentration accordingly to Langmuir model [31], as described in Equation 6. Equation 6
࢘ࢋ࢜ = ࢘ࢋ࢜ ×
ࡷ. ାࡷ.
In Equation 6, qrev (µmol m-²) represents the amount of acetaldehyde adsorbed on TiO2 surface at equilibrium, qrevm (µmol m-²) is the maximum adsorbed amount corresponding to surface saturation, K represents the adsorption constant, and C is acetaldehyde gas-phase concentration at equilibrium. The fitting curves reported on Figure 7 were obtained from regression using Langmuir model on experimental data. Regressions are used to determine the values of K and qrevm. Figure 8 reports the evolution of qrevm and K as a function of θ, from fresh (θ = 0) to aged (θ = 0.92) TiO2 surface by NO2. It can be noticed on Figure 8 (a) and (b) that, compared to clean TiO2 surface, the various surface ageing strongly impact qrevm as well as K. Indeed, the presence of NOx-(ads) on the surface decreases the amount of reversibly adsorbed acetaldehyde even for the lowest values of θ. For instance, as soon as θ reaches 0.05, qrevm is decreased by 65%. Interestingly, qrevm slightly varies with increasing θ, but uncertainties on qrevm do not make possible the determination of any clear tendency. Considering that the reversibly adsorbed fraction of acetaldehyde mainly occurs on OH surface group, the first hypotheses to explain this behaviour would be the decrease of OH surface groups by NOx-(ads) species. Indeed, Hadjiivanov et al. [29] demonstrated that, on TiO2 surface, the adsorbed NO2 and/or NOX- species increase the surface acidity by consuming , the basic surface OH groups. Therefore, it can be proposed that the lowering of surface OH group density due to NOXspecies may decrease the fraction of reversibly adsorbed acetaldehyde. 9
Beside, a second hypothesis could be raised; indeed, the presence of NOx-(ads) species may considerably change the adsorption mode of reversibly adsorbed acetaldehyde. This hypothesis is supported by the observtion of K variations with θ. Surprisingly, the adsorption constant of acetaldehyde is multiplied by a factor ranging from 7 to 11 as soon as NOx-(ads) species are present on TiO2 surface. This phenomenon suggests that the interaction between acetaldehyde and aged TiO2 surface is probably changed in nature, compared to clean TiO2 surface. The ionic structure of TiO2 adsorbed NOx-(ads) species may induce stronger electrostatic interactions between acetaldehyde and the aged photocatalyst surface. This hypothesis is supported by the work of Rodriguez et al. [28] who reported that adsorbed NOX- species modify the thermochemical stability of O vacancies on TiO2 surface and facilitates their migration from bulk to the surface which could provide new adsorption sites. The investigation of TiO2 surface ageing and subsequent acetaldehyde adsorption using diffuse reflectance infrared spectroscopy could help identifying the nature of reversible adsorption sites.
10
Conclusions
The adsorption of acetaldehyde on TiO2 has been qualified and quantified on fresh and NO2 aged TiO2 surface. First, quantitative data have been obtained regarding acetaldehyde adsorption on fresh TiO2. Second, based on the reactive adsorption mechanism of Sivachandiran et al. [27], it has been evidenced that it is possible to control the coverage rate of TiO2 surface by NOx-(ads) species, namely NO2-(ads) or NO3-(ads). Finally, acetaldehyde adsorption parameters have been determined as a function of the surface coverage by NOx-(ads) species. Interestingly, the quantities of irreversibly and reversibly adsorbed fractions, as well as the adsorption constant K, are impacted by the preliminary adsorption of NO2. NO2 preliminary adsorption tends to decrease the amount of irreversibly adsorbed acetaldehyde. This effect is characterized by the –∆qirr / q(NOx-(ads) parameter and varies with NOx-(ads) species. The effect of direct competitive adsorption combined with the hindering of acetaldehyde condensation on the surface are proposed to explain the observed phenomenon. Similarly, the reversibly adsorbed fraction of acetaldehyde is negatively impacted by NO2 preliminary adsorption. The first hypothesis proposed to explain this behaviour if the consumption of OH groups by NOx-(ads) species. However, considering the sudden variation of the adsorption constant K for the lowest surface coverages by NOx-(ads) species, it is suggested that a new adsorption mode of acetaldehyde is induced by the presence of NOx-(ads) species on TiO2 surface, this could be confirmed using in-situ spectroscopy of adsorbed species such as DRIFTS (Diffuse Reflectance Infrared Fourier Transform Spectometry). This study quantitatively evidences for the first time the remarkable impact of NO2 preliminary adsorption on the consecutive adsorption of a VOC on TiO2. It is clearly shown that the presence of NOx-(ads) species on TiO2 surface considerably modifies the surface chemistry and the subsequent VOC sorption properties of TiO2. As a consequence, this result questions the long term behaviour of TiO2 when used as a coupling material in-side or downstream a plasma discharge. Considering that, in the case of plasma-material coupling, the role of TiO2 is highly related to its sorption properties, long term performances of such processes should be addressed. Similarly, in catalytic or photocatalytic processes, the decrease of VOC adsorption on TiO2 may considerably hinder the global conversion rate of VOCs. These aspects should be investigated in order to give new insights on the long term performances of heterogeneous oxidation processes under real operating conditions.
Acknowledgements Authors want to thank the Institut Carnot M.I.N.E.S. for its financial support in the framework of PhotoCair project, as well as Vincent Gaudion for his valuable technical assistance.
11
References [1] R.M. Alberici, M.C. Canela, M.N. Eberlin, W.F. Jardim, Catalyst deactivation in the gas phase destruction of nitrogen-containing organic compounds using TiO2/UV–VIS, Applied Catalysis B: Environmental 3-4 (2001) 389-397. [2] T.H. Lim, S.M. Jeong, S.D. Kim, J. Gyenis, Photocatalytic decomposition of NO by TiO2 particles , Journal of Photochemistry and Photobiology A 134 (2000) 209-217. [3] H. Wang, Z. Wu, W. Zhao, B. Guan, Photocatalytic oxidation of nitrogen oxides using TiO2 loading on woven glass fabric, Chemosphere 66 (2007) 185-190. [4] O. Debono, F. Thevenet, P. Gravejat, V. Hequet, C. Raillard, L. Lecoq, N. Locoge, Toluene photocatalytic oxidation at ppbv levels: Kinetic investigation and carbon balance determination , Applied Catalysis B: Environmental 106 (2011) 600-608. [5] H. Park, W. Choi, Effects of TiO2 Surface Fluorination on Photocatalytic Reactions and Photoelectrochemical Behaviors, Journal of Physical Chemistry B 108 (2004) 4086-4093. [6] S. Bahri, C.M. Jonsson, C.L. Jonsson, D. Azzolini, D.A. Sverjensky, R.M. Hazen, Adsorption and surface complexation study of LDOPA on rutile (α-TiO2) in NaCl solutions, Environmental Science & Technology 45 (2011) 3959-3966. [7] F. Thevenet, J. Couble, M. Brandhorst, J.L. Dubois, E. Puzenat, C. Guillard, D. Bianchi, Synthesis of Hydrogen Peroxide Using Dielectric Barrier Discharge Associated with Fibrous Materials, Plasma Chemistry and Plasma Processing 30 (2010) 489-502. [8] C. Marchand, B. Bulliot, S. L. Calvé, P. Mirabel, Aldehyde Measurements in indoor environmentin Strasbourg (France), Atmosperic Environment 40 (2006) 1336-1345. [9] Q. Liu, Y. Liu, M. Zhang, Personal exposure and source characteristics of carbonyl compounds and BTEXs within homes in Beijing, China , Building & Environment 61 (2013) 210-216. [10] R. Atkinson, Atmospheric chemistry of VOCs and NOx , Atmospheric Environment 34 (2000) 2063-2101 [11] C. Marchand, S. L. Calvé, P. Mirabel, N. Glasser, A. Casset, N. Schneider, F. de Blay, Indoor aldehydes concentrations and determinants in 162 homes in Strasbourg (France), Atmospheric Environment 42 (2008) 505-516 [12] J. M. Coronado, S. Kataoka, Isabel Tejedor-Tejedor, M. A. Anderson, Dynamic phenomena during the photocatalytic oxidation of ethanol and acetone over nanocrystalline TiO2: simultaneous FTIR analysis of gas and surface species, Journal of Catalysis 219 (2003) 219-230 [13] J. Raskó, J. Kiss, Adsorption and surface reactions of acetaldehyde on TiO2 , CeO 2 and Al2O3, Applied Catalysis A: General 287 (2005) 252-260 [14] M. Singh, N. Zhou, D. K. Paul, K. J. Klabunde, IR spectral evidence of aldol condensation: Acetaldehyde adsorption over TiO2 surface, Journal of Catalysis 260 (2008) 371-379 [15] F. Batault, F. Thevenet, V. Hequet, C. Rillard, L. Le Coq, N. Locoge, Acetaldehyde and acetic acid adsorption on TiO2 under dry and humid conditions, Chemical Engineering Journal 264 (2015) 197-210 [16] Z. Topalian, B. I. Stefanov, C. G. Granqvist, L. Österlund, Adsorption and photo-oxidation of acetaldehyde on TiO2 and sulfatemodified TiO2: Studies by in situ FTIR spectroscopy and micro-kinetic modeling, Journal of Catalysis 307 (2013) 265-274 [17] H.H. Kim, S.M. Oh, A. Ogata, S. Futamura, Decomposition of gas-phase benzene using plasma-driven catalyst (PDC) reactor packed with Ag/TiO2 catalyst, Applied Catalysis B: Environmental 56 (2005) 213–220. [18] J.V. Durme, J. Dewulf, W. Sysmans, C. Leys, H.V. Langenhove, Combining non-thermal plasma with heterogeneous catalysis in waste gas treatment: A review, Applied Catalysis B: Environmental 74 (2007) 161-1690. [19] X. Fan, T.L. Zhu, M.Y. Wang, X.M. Li, Removal of low-concentration BTX in air using a combined plasma catalysis system, Chemosphere 75 (2009) 1301-1306. [20] H.H. Kim, Nonthermal Plasma Processing for Air-Pollution Control: A Historical Review, Current Issues, and Future Prospects, Plasma Process and Polymers 1 (2004) 91-110.
12
Figure 1 - General scheme of the adsorption experimental setup.
Figure 2 - FTIR gas phase spectra of CH3CHO, N2O, NO2 and NO. The framed regions are used for quantifications.
Figure 3 - Experimental procedure used to assess the influence of NO2 preliminary adsorption on TiO2 surface regarding acetaldehyde adsorption. Acetaldehyde adsorption is carried out on fresh and NO 2 aged TiO2 surface under dark condition, dry air and at 296 K.
Figure 4 - Temporal profiles of NO2 and NO monitored at the reactor downstream under dark conditions and at 296 K. For adsorption, 30 ppm of NO2 and 5 ppm of N2O balanced with zero air are concurrently sent at the -1 reactor inlet with a total flow rate of 600 mL min .
Figure 5 - Acetaldehyde breakthrough and flushing curves overlaid with N2O system mixing curves on clean TiO2 surface. During adsorption, 20 ppm of acetaldehyde and 5 ppm of N2O are concurrently sent to the -1 reactor inlet. Adsorption is performed under dry air with 600 mL min total flow rate, at 296 K and under dark conditions.
Figure 6 - Irreversibly adsorbed acetaldehyde fractions on TiO 2 as a function of i.e. NOx (ads) surface coverage. Acetaldehyde adsorption on fresh and aged TiO2 is performed under dry air and at 296 K. -
Figure 7 - Acetaldehyde adsorption on fresh and aged TiO 2 at 296 K. Adsorption isotherms of the reversibly adsorbed fraction (qrev) are determined for different values: (a) 0, (b) 0.05, (c) 0.1, (d) 0.55 and (e) 0.92. Acetaldehyde inlet concentration was varied from 5 to 300 ppm.
Figure 8 - Evolution of a) maximum reversibly adsorbed amounts of acetaldehyde (qrevm), and b) reversibly adsorbed acetaldehyde adsorption constant (K), reported as a function of NO2 ageing on TiO2 surface, i.e. . These data are determined from Langmuir model regression using experimental isotherms.
[21] C.J. Weschler, H.C. Shields, D.V. Naik, Indoor Chemistry Involving O3, NO, and NO 2 as Evidenced by 14 Months of Measurements at a Site in Southern California, Environmental Science & Technology 28 (1994) 2120-2132. [22] A. Chaloulakou, I. Mavroidis, I. Gavriil, Compliance with the annual NO2 air quality standard in Athens. Required NOx levels and expected health implications, Atmospheric Environment 42 (2008) 454-465. [23] S. Matsuda, H. Hatano, A. Tsutsumi, Ultrafine particle fluidization and its application to photocatalytic NOx treatment, Chemical Engineering Journal 82 (2001) 183-188. [24] F.L. Toma, G. Bertrand, D. Klein, C. Coddet, Photocatalytic removal of nitrogen oxides via titanium dioxide, Environmental Chemistry Letters 2 (2004) 117-121. [25] K. I. Hadjiivanov, Identification of neutral and charged NxOy surface species by IR spectroscopy, Catalysis Review 42 (2000) 71144 [26] V. H. Grassian, Heterogeneous uptake and reaction of nitrogen oxides and volatile organic compounds on the surface of atmospheric particles including oxides, carbonates, soot and mineral dust: Implications for the chemical balance of the troposphere, International Review in Physical Chemistry 20 (2001) 467-548 [27] L. Sivachandiran, F. Thevenet, P. Gravejat, A. Rousseau, Investigation of NO and NO2 Adsorption Mechanism on TiO2 at Room Temperature, Applied Catalysis B: Environmental 142-143 (2013) 196-204, 2013. [28] J. A. Rodriguez, T. Jirsak, G. Liu, J. Hrbek, J. Dvorak, A. Maiti, Chemistry of NO2 on oxide surfaces: formation of NO 3 on TiO2(110) and NO2 O vacancy interactions, Journal of American Chemical Society 123 (2001) 9597-9605 [29] K.I. Hadjiivanov, V. Bushev, M. Kantcheva, D. Klissurski, Infrared spectroscopy study of the species arising during nitrogen dioxide adsorption on titania (anatase), Langmuir 10 (1994) 464-471 [30] J. E. Rekoske, M. A. Barteau, Kinetics, Selectivity, and deactivation in the aldol condensation of acetaldehyde on anatase titanium dioxide, Industrial and Engineering Chemistry Research 50 (2010) 41-51 [31] I. Langmuir, The adsorption of gases on plane surfaces of glass, mica and platinum, Journal of American Chemical Society 40 (1918) 1361-1403
13
bypass air
AZ
PSA
MFC 2
N2O
FTIR
MFC 1 vent
MFC 3
NO2
vent
thermocouple
MFC 4 oven
CH3CHO
TiO2 quartz wool reactor part
Figure 1 - General scheme of the adsorption experimental setup.
Absorbance (a. u.)
CH3CHO
Absorbance (a. u.)
1900
1800
20 ppmv
1700
1600
1500
1400
NO2
1700
N2O
1300
2 ppmv
1650
1600
1550
5 ppmv
2300
2260
2220
2180
NO
1500
2000
2140
2 ppmv
1950
1900
1850
1800
1750
1700
Wavenumbers (cm-1)
Figure 2 - FTIR gas phase spectra of CH3CHO, N2O, NO2 and NO. The framed regions are used for quantifications.
Thermal treatment Dry Air, 703 K
Surface coverages by NO2 30 ppm dry air, 296K OR
CH3CHO breakthrough Dry Air, 296 K
Flushing Dry Air, 296 K
Figure 3 - Experimental procedure used to assess the influence of NO2 preliminary adsorption on TiO2 surface regarding acetaldehyde adsorption. Acetaldehyde adsorption is carried out on fresh and NO2 aged TiO2 surface under dark condition, dry air and at 296 K.
30 system mixing curve
CC (ppm) (ppm)
NO2
20
10 NO
0
0
30
60
90
120
150
Time (min) Time (min) Figure 4 - Temporal profiles of NO2 and NO monitored at the reactor downstream under dark conditions and at 296 K. For adsorption, 30 ppm of NO2 and 5 ppm of N2O balanced with zero air are concurrently sent at the reactor inlet with a total flow rate of 600 mL min-1.
[acetaldehyde] (ppm)
system mixing curve
20 15
10 5 0 0
50
100
150
Time (min) Figure 5 - Acetaldehyde breakthrough and flushing curves overlaid with N2O system mixing curves on clean TiO2 surface. During adsorption, 20 ppm of acetaldehyde and 5 ppm of N2O are concurrently sent to the -1 reactor inlet. Adsorption is performed under dry air with 600 mL min total flow rate, at 296 K and under dark conditions.
Figure 6 - Irreversibly adsorbed acetaldehyde fractions on TiO 2 as a function of i.e. NOx (ads) surface coverage. Acetaldehyde adsorption on fresh and aged TiO2 is performed under dry air and at 296 K. -
a) = 0
b) = 0.05
c) = 0.1
d) = 0.55
e) = 0.92
q rev (µmol/m²)
Acetaldehyde (ppm)
Figure 7 - Acetaldehyde adsorption on fresh and aged TiO2 at 296 K. Adsorption isotherms of the reversibly adsorbed fraction (qrev) are determined for different values: (a) 0, (b) 0.05, (c) 0.1, (d) 0.55 and (e) 0.92. Acetaldehyde inlet concentration was varied from 5 to 300 ppm.
a)
b)
Figure 8 - Evolution of a) maximum reversibly adsorbed amounts of acetaldehyde (qrevm), and b) reversibly adsorbed acetaldehyde adsorption constant (K), reported as a function of NO2 ageing on TiO2 surface, i.e. . These data are determined from Langmuir model regression using experimental isotherms.
Table 1 - FTIR detection limits (DL) for the species of interest.
-
Table 2 - Amounts of NO2 consumed per surface unit for various adsorption times. Total q(NOx (ads)) is calculated as the total amount of NO2 (ads) and NO3 (ads) per surface unit. Covering of TiO2 surface by NO2 is represented by the parameter varying from 0 to 1.
Table 3 - Acetaldehyde adsorbed amounts on fresh and NO2 exposed TiO2 surface. Decreases in adsorbed acetaldehyde amounts –qirr and –qirr / q(NOx (ads)) are determined from fresh and NO2 exposed acetaldehyde adsorption data.
Compound NO NO2 N2O CH3CHO
DL (ppb) 700 50 50 500
Table 1 - FTIR detection limits (DL) for the species of interest.
Adsorption time (s)
Total consumed NO2 (µmol m ²)
Total q(NOx (ads)) (µmol m ²)
150
0.19 ± 0.1
0.19 ± 0.05
0.05 ± 0.02
300
0.38 ± 0.05
0.38 ± 0.05
0.10 ± 0.02
1450
3.03 ± 0.3
2.02 ± 0.15
0.55 ± 0.05
2400
5.01 ± 0.4
3.34 ± 0.27
0.92 ± 0.05
5000
5.47 ± 0.4
3.65 ± 0.27
1
-
Table 2 - Amounts of NO2 consumed per surface unit for various adsorption times. Total q(NOx (ads)) is calculated as the total amount of NO2 (ads) and NO3 (ads) per surface unit. Covering of TiO2 surface by NO2 is represented by the parameter varying from 0 to 1.
Acetaldehyde qirr
-Δqirr
µmol m ²
µmol m ²
0.00
8.61 ± 0.28
0
Ø
0.05
7.07 ± 0.34
1.52
8.0
0.10
4.66 ± 0.68
3.93
10.3
0.55
1.65 ± 0.09
6.94
3.4
0.92
1.29 ± 0.41
7.3
2.2
θ
-
-
-Δqirr / q(NOx-(ads))
Table 3 - Acetaldehyde adsorbed amounts on fresh and NO2 exposed TiO2 surface. Decreases in adsorbed acetaldehyde amounts –qirr and –qirr / q(NOx (ads)) are determined from fresh and NO2 exposed acetaldehyde adsorption data.
Evolution of acetaldehyde irreversibly adsorbed fraction (qirr) on TiO2 at 296 K as a function of TiO2 surface coverage by NOx- adsorbed species () obtained by preliminary NO2 adsorption on TiO2.
Acetaldehyde adsorption parameters on TiO2 are quantitatively determines TiO2 surface coverage by NO2 derived adsorbed species has been controlled NO2 preliminary adsorption considerably impacts acetaldehyde adsorption parameters