Fe-ZSM-5 catalysts were prepared by six different methods (a-f). Samples provided by Prof. Dr. W. Grünert were prepared by methods a-c and e using two different H-ZSM-5 supports which were obtained from commercial Na-ZSM-5 with different Si/Al ratio (≈ 14 and ≈ 40, the latter with high density of structure defects, e.g. silanol nests, labeled A and B, respectively) by treating with 0.1 M HCl. The first one (A) was used for all six preparations, the second one (B) was used only for the CVD method for comparative purposes.

Samples provided by Prof. Dr. Pérez-Ramírez were prepared by methods a (chemical vapor deposition), d (liquid ion exchange) and f (hydrothermal synthesis followed by steam activation). Two different commercial H-ZSM-5 (Si/Al ≈14) and NH₄-ZSM-5 (Si/Al ≈37) supports (Degussa and Zeolyst) were used for CVD and liquid ion exchange methods respectively.

This preparation method was proposed by H.-Y. Chen and W. M. H. Sachtler [31]. Samples provided by Prof. Dr. W. Grünert were prepared according to the following procedure. Fe ions were introduced into the dried parent H-ZSM-5 matrices (A and B) by evaporating anhydrous FeCl₃ in an inert atmosphere into the pores of the zeolite followed by washing with 1 l or 10 l water per 5 g catalyst (W1 and W10, respectively) and calcination in air. The standard calcination procedure [36,55] was as follows: samples were heated in air from 293 K to 423 K at 2 K/min and kept isothermally at this temperature for 15 min, further heated to 873 K with a heating rate of 5 K/min and kept isothermally at this temperature for 2 h. Besides, a heating rate of 0.5 K/min was also used to elucidate the influence of different heating rates on the properties of the catalysts. Details on the preparation procedure and pretreatment conditions of these zeolites can be found elsewhere [36,55].

Throughout this thesis sample labels reflecting the preparation history are used. The label is composed of symbols for the type of ZSM-5, Fe introduction route, washing intensity and heating rate during calcination. Thus, A(CVD,W1,C0.5) means a material in which Fe was introduced via CVD into ZSM-5 of type A, washed with 1 l water per 5 g catalyst and calcined with a heating rate of 0.5 K/min. Missing symbols mean that the respective step has been omitted (i.e., A(CVD,W1) is A(CVD,W1,Cx) before calcination with a ramp of x K/min). The Fe content of the samples has been determined in the group of Prof. Dr. W. Grünert by ICP analysis. It amounted to 5.4 wt.-% for A(CVD, W1), 5.0 wt.-% for A(CVD,W10) and 2.6 wt.-% for B(CVD,W1) (Table 3.1).

A sample provided by Prof. Dr. Pérez-Ramírez was also prepared by the same CVD procedure as proposed by H.-Y. Chen and W. M. H. Sachtler [31]. Fe ions were introduced into a dried parent H-ZSM-5(A) (Si/Al ≈14) by using anhydrous FeCl₃ followed by washing and calcination in static air at 823 K for 5 h [58]. The resulting catalyst is labeled as A'(CVD,W1,C2). The chemical composition of the catalysts was determined by ICP and is presented in Table 3.1.

Solid state ion exchange was performed by mixing 7 g of H-ZSM-5(A) and 3.5 g of FeCl₃ ^.6 H₂O. The mixture was placed into a porcelain boat and heated in nitrogen flow (50 ml/min) at 573 K for 1 h followed by washing with 1 l of double deionized water and calcined according to the standard calcination procedure [36,55]. The same washing and calcination procedures are also used for samples in subsections c, d and e. Henceforth, these two steps are omitted from the sample labeling for the sake of clarity. The Fe content of the sample determined by ICP analysis amounted to 5.2 wt.-% for A(SSIE). The resulting sample is labeled as A(SSIE)5.2 (Table 3.1). For some experiments uncalcined sample was used and is mentioned so.

This method comprises intense grinding of the parent H-ZSM-5(A) with FeCl₃ ^.6 H₂O, followed by 2-3 short-time washing steps (0.5 l water per 2 g catalyst). Calcination was performed according to the standard procedure in air at 873 K for 2 h [33,55]. The Fe content of the sample determined by ICP analyses amounted to 0.5 wt.-% for A(MR). The sample is labeled as A(MR)0.5 (Table 3.1). For some experiments uncalcined sample was used and is mentioned so.

A sample provided by Prof. Dr. Pérez-Ramírez was prepared by conventional aqueous ion exchange. Diluted solutions (0.30 mM) of Fe(NO₃)₃. 9H₂O and NH₄-ZSM-5 (Zeolyst with Si/Al ≈37) were vigorously stirred at 293 K for 15 h without control of the pH. The ion-exchanged zeolite was then filtered, washed thoroughly, dried and finally calcined in static air at 823 K for 5 h. The Fe-content of the catalyst determined by ICP is 1.4 wt% (Table 3.1). This catalyst is labeled as Fe-ZSM-5(LIE)1.4 [58].

R. Q. Long and R. T. Yang proposed this preparation method [106]. Ion exchange was performed in Ar flow in order to prevent oxidation of Fe⁺² to Fe⁺³ during exchange, by using different amounts of iron powder in 200 ml of 0.1 M HCl and 2 g of H-ZSM-5(A) support stirred at 293 K for 5 days. The ion exchanged zeolites were then filtered, washed with deionized water, dried and finally calcined according to the standard procedure in air at 823 K with a heating rate of 5 K/min. The iron percentage was determined by ICP. The resulting five catalysts are labeled as: A(ILIE)0.2 (that means a sample prepared by improved liquid ion exchange and containing 0.2 wt.-% iron (Table 3.1)), A(ILIE)0.3, A(ILIE)0.6, A(ILIE)0.7 and A(ILIE)1.2 [115].

As synthesized, calcined and steamed Fe-silicalite, Fe-ZSM-5 and Fe-beta samples provided by Prof. Dr. Pérez-Ramírez were prepared by isomorphous substitution of iron in the zeolite framework. The preparation procedure is described in detail elsewhere [41,58,154,155]. Briefly, for Fe-silicalite, a solution of Tetraethylorthosilicate, tetrapropyl ammonium hydroxide and NaOH was dropwise added to an iron nitrate solution while for Fe-ZSM-5, aluminum nitrate was additionally added as Al source. The gel was then treated in a static air at 448 K for 5 days in a stainless steel autoclave lined with Teflon. The solid was filtered and washed with deionized water. The as-synthesized sample was calcined in air at 823 K and converted into the H-form by exchange with ammonium nitrate solution followed by calcination at 823 K for 5 h.

For Fe-beta the preparation procedure was modified using TEAOH as templating agent and HF was added as a mineraliser agent. The gel was heated in a Teflon-lined steel autoclave at 423 K for 10 days under static conditions. The solid was washed, dried at 353 K, and calcined at 853 K for 3 h.

Finally, the calcined Fe-zeolites were activated in flowing steam at ambient pressure (water partial pressure of 300 mbar and 30 ml STP min^-1 of N₂ flow) at 873 K during 5 h. A ramp of 2 K min^-1 was used during heating up and cooling down the sample in N₂. The resulting catalysts were labeled as: c-Fe-silicalite, c-Fe-ZSM-5, c-Fe- beta, ex-Fe-silicalite ex-Fe-ZSM-5 and ex-Fe- beta that mean samples calcined (c) and steam activated (ex) respectively. The chemical composition of the catalysts was determined by ICP and is presented in Table 3.1.

The parent siliceous SBA-15 material was provided by Dr. H. Kosslick (ACA Berlin) and was prepared by an optimized synthesis procedure, details of which are described in [156]. The BET surface area and the average pore diameter of the Si-SBA-15 support were determined to be 558 m²/g and 7-8 nm respectively.

Iron was introduced into the parent SBA-15 by impregnation. 0.1092 g of iron acetylacetonate, Fe(acac)₃ (purity > 97 %, Fluka Chemicals), was dissolved in 10 ml of acetone. The resulting solution was added dropwise to 1.6517 g of the SBA-15 support, which had been pretreated in air at 673 K for 30 min to remove adsorbed moisture. The resulting sample was dried at 293 K and subsequently calcined in airflow at 823 K for 2 h with a heating rate of 5 K/min. The washing step was omitted for this sample. The Fe% of the catalyst was determined by ICP to be 0.95 Fe Wt.% (Table 3.1). The resulting catalyst was labeled as: (Fe-SBA-I)0.95, that means iron deposited in SBA-15 support via impregnation method and contains 0.95 wt.% iron. The structure of SBA-15 consists of a well-ordered hexagonal array of linear cylindrical channels (d = 7.5 nm) [157,158].

Catalyst

Fe( wt%)

Si/Al

Catalyst

Si/Al

Fe( wt%)

A(CVD,W1)

5.4

14

A(ILIE)0.2

14

0.2

A(CVD,W1,C0.5)

5.4

14

A(ILIE)0.3

14

0.3

A(CVD,W1,C5)

5.4

14

A(ILIE)0.6

14

0.6

A(CVD,W10)

5.0

14

A(ILIE)0.7

14

0.7

A'(CVD,W1,C2)

5.0

14

A(ILIE)1.2

14

1.2

B(CVD,W1)

2.6

44

Fe-ZSM-5(LIE)1.4

37

1.4

B(CVD,W1,C5)

2.6

44

c-Fe-silicalite

∞

0.68

A(SSIE)5.2 uncalcined

5.2

14

ex-Fe-silicalite

∞

0.68

A(SSIE)5.2

5.2

14

c-Fe-ZSM-5

31

0.67

A(MR)0.5 uncalcined

0.5

14

ex-Fe-ZSM-5

31

0.67

A(MR)0.5

0.5

14

c-Fe-beta

31.5

0.61

(Fe-SBA-I)0.95

∞

0.95

ex-Fe-beta

31.8

0.63

Table 3.1. Chemical composition of the catalysts as determined by ICP-OES [55,58,64,159]

EPR spectra in the X-band (ν≈ 9.5 GHz) were recorded with the cw-spectrometer ELEXSYS 500-10/12 (Bruker) using a microwave power of 6.3 mW, a modulation frequency of 100 kHz and a modulation amplitude of 0.5 mT. The  magnetic  field  was  measured  with  respect to the standard 2,2-diphenyl-1-picrylhydrazyl hydrate (DPPH). For temperature-dependent measurements in the range from 90 to 293 K, a commercial variable temperature control unit (Bruker) and a conventional EPR sample tube was used, while in situ EPR measurements during calcination in the range 293 < T/K < 773 and in situ measurements in the presence of reactants were performed in a home-made EPR flow reactor (Fig. 3.1) which was placed into the rectangular X-band cavity of the spectrometer [147]. Temperature programmer (Eurotherm) and a gas dosing system with mass flow controllers (Bronkhorst) were attached to the system for controlling temperature and gas flows. In situ EPR experiments with NH₃, isobutane, CO, NO and N₂O were performed at 623 K since at this temperature all catalysts have shown substantial activity in steady-state catalytic experiments. For all experiments 50 mg of granular catalyst with a particle size of 125-200 μm was used.

Low temperature NH₃ interaction experiments with A(MR)0.5 and A(CVD,W1,C5) were performed at 293 K according to the following sequence: (a) pretreatment in air (10 ml/min) at 773 K for 1 h and cooling to 293 K, (b) changing the flow from air to 1% NH₃/He (10 ml/min) for 1 h followed by flushing with N₂ at the same temperature for 15 min. Spectra were recorded after each treatment.

Fig. 3.1. Home-made EPR flow reactor for in situ measurements [147].

For NH₃ or isobutane-SCR, the catalysts were treated according to the following sequence: (a) pretreatment in air (20 ml min^-1) at 773 K for 1 h and cooling to 623 K (denoted as the reaction temperature), (b) changing the flow from air subsequently to total feed, i.e., 0.1% NO, 0.1% reducing agent and 2% O₂/He (reducing agent: NH₃ or isobutane, respectively) (GHSV for NH₃ and isobutane-SCR was 30,000 h^-1) for 1 h, (c) switching to air (20 ml min^-1) for 1 h and for 15 min at 773 K to ensure complete reoxidation of iron species having been reduced in step (b), followed by cooling to 623 K again, (d) changing flow from air to 0.1% NH₃/He (34.7 ml/min) or 0.1% isobutane/He (34.7 ml/min) for 1 h, and (f) switching to 0.1% NO/He (34.7 ml/min) for 1 h.

To study the influence of temperature on the redox behaviour of different isolated Fe sites, subsequent interactions of 0.1% NH₃/He and 0.1% NO/He with Fe-ZSM-5 were also performed at 773 K in the following sequence: (a) pretreatment in air (20ml min^-1) at 773 K for 1 h, (b) changing the flow from air to 0.1% NH₃/He (34.7 ml/min) for 1 h, and (f) switching to 0.1% NO/He (34.7 ml/min) for 1 h.

In situ experiments with N₂O and CO were performed according to the following sequence: (a) heating in air (15 ml/min) at 773 K for 1 h and cooling to 623 K (denoted as the reaction temperature), (b) changing the flow from air to 2 vol.% N₂O/He (12.5 ml/min) for 1 h, (c) switching to 2 vol.% CO/He (12.5 ml/min) for 1 h, and finally (d) switching back to 2 vol.% N₂O/He (12.5 ml/min) at reaction temperature.

Additional in situ EPR experiments over A'(CVD,W1,C2) were performed at atmospheric pressure at 623 K with mixtures of 1 vol.% N₂O + 1 vol.% CO in He (25 ml/min) and 0.66 vol.% N₂O + 1.34 vol.% CO in He (18.7 ml/min). For on-line analysis of N₂ formed during N₂O-CO reaction, the outlet of the EPR flow reactor was connected to a GC 17AAF capillary gas chromatograph (Shimadzu) equipped with a 30 m x 0.32 mm molecular sieve 5A column (CP-7534, Chrompack) and a thermal conductivity detector.

UV/VIS-DRS measurements were performed with a Cary 400 spectrometer (Varian) equipped with a diffuse reflectance accessory (praying mantis, Harrick, Fig. 3.2). To reduce light absorption, catalysts were diluted with α-Al₂O₃ (calcined at 1473 K for 4 hours), also used as white reference sample, in a ratio of 1:3, 1:5 or 1:10. The UV/VIS-DR spectra were collected in reflectance mode and evaluated after converting them into the Kubelka-Munk-Function (F(R)).

Fig. 3.2. In situ DRS cell: left praying mantis (Harrick) diffuse reflection attachment and right stainless steel reaction chamber.

UV/VIS-DR spectra were deconvoluted into subbands with Gaussian line shape by means of GRAMS/32 software (Galactic), which uses a least-square fitting algorithm. In practice, the peak fitting was run until satisfactory statistical results were obtained, i.e. until the solution was converged. This occurs when reduced χ² (Chi squared) has reached a minimum and five successive iterations have not significantly improved the fit. A fitting is considered to be converged if χ² value is lower than 3. The spectra were interpreted by comparing positions of the peaks, obtained from deconvolution, with the literature data.

In situ measurements at elevated temperatures and in flowing gas mixtures were carried out in a heatable reaction chamber (Harrick) (Fig. 3.2) equipped with a temperature programmer (Eurotherm) and a gas dosing system with mass flow controllers (Bronkhorst).

For elucidating the influence of calcination conditions on the nature of the Fe sites, the effect of calcination on selected catalysts was studied by performing in situ calcination in air using the standard calcination procedure (section 3.1.1.a).

The reduction/reoxidation behavior of selected catalysts was analyzed by treatment in a flow of 20 vol.% H₂ in N₂ (10 ml/min) and air (10 ml/min) respectively at 773 K at ambient pressure for 1 h. After the respective treatment, the catalysts were cooled to 293 K in the same mixture and the spectrum was recorded.

Kinetic studies of the reduction/reoxidation were performed with A(ILIE)0.3 and A(ILIE)1.2 at 673 K. The time dependence of the absorbance (at 238 and 290 nm for isolated Fe⁺³ ions and 350 nm for iron oxide clusters) was measured in reducing and oxidizing atmospheres. The obtained redox kinetic curves were evaluated according to a pseudo-first order rate law using the following equation:

Abs _t = Abs _{t = ∞} + (Abs _{t = 0} ― Abs _{t = ∞}) e^{― kt} (3.1)

Here, Abs _{t = 0}, Abs _t and Abs _{t = ∞} are the absorbance values at a given wavelength at the start of the experiment, at time t and after reaching steady state respectively.

The reduction and reoxidation treatments were performed according to the following sequence: (a) pretreatment in air (10 ml/min) at 773 K for 1 h and cooling to 673 K, (b) changing the flow from air to 1% NH₃/He (10 ml/min) for 2 h, and (c) switching to air (10 ml/min) for 2 h. After treatments (a) and (b) the pipeline was flushed with N₂ for 5 min. Spectra were recorded after each treatment.

In situ measurements with NH₃, isobutane, CO, NO and N₂O were performed according to the same sequence and with the same gas composition and flow rate as described for in situ EPR experiments.

The surface acidity of the Fe-zeolites was studied by FT-IR spectroscopic analysis of adsorbed pyridine. Spectra were recorded using a Bruker IFS 66 spectrometer equipped with a heatable and evacuable reaction cell with CaF₂ windows, connected to a gas dosing and evacuation system. The zeolite powder was pressed into self-supporting wafers with a diameter of 20 mm and a weight of 50 mg. Prior to pyridine adsorption, the samples were pretreated in flowing air (30 ml/min) at 673 K for 1 h followed by cooling to 373 K. Then, pyridine was adsorbed at 373 K for 1 h by bubbling an argon flow (60 ml/min) through a pyridine-containing saturator (at 293 K). Physisorbed pyridine was removed by evacuation for 30 min at 373 K and infrared spectra were recorded at the same temperature with 2 cm^-1 resolution and 100 scans.

NO adsorption experiments on pre-oxidized and pre-reduced surfaces of H-ZSM-5(A), A(CVD,W1,C5), A(MR)0.5, A(ILIE)0.3 and A(ILIE)1.2 were performed according to the following sequence: samples were oxidatively pretreated in air flow (30 ml/min) at 673 K for 1 h followed by cooling to 293 K and initial spectra were recorded for oxidized surfaces. NO adsorption was performed at 293 K with 1% NO/He (40 ml/min) for 1 h followed by evacuation at the same temperature for 30 min. Spectra were recorded before and after evacuation.

Subsequently, the samples were heated in air at 673 K for 1 h. At this temperature, the flow was changed from air to 2 or 10% NH₃/He (40 ml/min) for 1 h followed by evacuation at the same temperature for 30 min. Afterwards samples were cooled to 293 K and evacuated for 30 min and initial spectra were recorded for reduced surfaces. NO interaction with prereduced zeolites was studied by switching to 1% NO/He (40 ml/min) at 293 K for 1 h followed by evacuation at the same temperature for 30 min and spectra were recorded before and after evacuation with 2 cm^-1 resolution and 100 scans.

In situ FT-IR experiments over H-ZSM-5(A), A(ILIE)0.2 and A(ILIE)1.2 with isobutane and NO were performed at 623 K. The catalysts were treated according to the following sequence: (a) pretreatment in air (30 ml/min) at 673 K for 1 h and cooling to 623 K, (b) changing the flow from air subsequently to total feed, i.e., 1% NO/He (30 ml/min), 1% isobutane/He (30 ml/min) and air (30 ml/min) for 1 h. Spectra were recorded after different time intervals after which samples were evacuated and spectra were recorded again. Subsequently, samples were heated again in air (30 ml/min) at 623 K for 1 h and spectra were recorded at the same temperature.

The selective reduction of NO with isobutane or NH₃ was studied in a catalytic micro-flow reactor with a product analysis scheme that combined calibrated mass spectrometry, gas chromatography and non-dispersive IR photometry (NH₃). Feed gases containing 1000 ppm NO, 1000 ppm reductant (isobutane or NH₃), 2 % O₂ in He were charged onto the catalyst at GHSV of 30,000 and 42,000 h^-1 for isobutane-SCR and 750,000 h^-1 for NH₃-SCR. Generally, the catalytic runs were started with a thermal activation and stabilization treatment of the catalysts in flowing He at 823 K (isobutane) or at 873 K (NH₃). The activities were measured from the higher to the lower reaction temperatures. Under experimental conditions, the only reaction product of NO observed in the limits of experimental accuracy was N₂, i.e. the NO conversions given are equal to N₂ yields [32,36,55,64].

Activity measurements were carried out in a parallel-flow reactor system, similar to that described in [160], using 50 mg of catalyst (125-200 μm) and space velocities W/F(N₂O)_o of 3⋅10⁵ and 9⋅10⁵ g s mol^-1 at atmospheric pressure, where W/F(N₂O)_o is the ratio between the catalyst mass and the molar flow of N₂O at the reactor inlet. Feed mixtures containing N₂O (1.5 mbar) and CO (0-1.5 mbar) in He were applied and conversions of both components were measured in the range between 475-1100 K. The product gases were analyzed by online gas chromatography. During the experiments, the mass balances of N, C, and O closed at > 98%. N₂O and CO conversions were calculated from the amount of N₂ and CO₂ formed, respectively [58,161,162].