Acidic properties of fiberglass materials - Springer Link

Jointly published by Akadémiai Kiadó, Budapest and Springer, Dordrecht

React.Kinet.Catal.Lett. Vol. 92, No. 2, 303−309 (2007) 10.1007/s11144-007-5204-3

RKCL5204 ACIDIC PROPERTIES OF FIBERGLASS MATERIALS Tatiana S. Glazneva*, Vera P. Shmachkova, Ludmila G. Simonova and Evgeny A. Paukshtis Boreskov Institute of Catalysis, Novosibirsk 630090, Russia

Received August 30, 2007, accepted September 14, 2007

Abstract Acidic properties of fiberglass materials were investigated using the adsorption of NH3 and the rate of isopropanol dehydration. It is shown that the specific catalytic activity of such materials and amount of the Brønsted acid sites per their surface unit (100 Å2) exceed considerably those in zeolite HZSM-5. Keywords: Fiberglass materials, isopropanol dehydration, DRIFT spectroscopy

INTRODUCTION Catalysts based on leached fiberglass (FG) materials exhibit high activity in many gas- and liquid-phase catalytic processes [1-11]. According to [5-10], the catalytic properties of FG catalysts depend on the chemical composition of the FG material, conditions of its treatment and methods of the deposition of the active component. _________________________ *

Corresponding author. E-mail: [email protected] 0133-1736/2007/US$ 20.00. © Akadémiai Kiadó, Budapest. All rights reserved.

304

GLAZNEVA et al.: FIBERGLASS MATERIALS

The most typical impurities in industrial silicate FG materials are Al introduced into the glass with an initial raw material, and Na, which cannot be removed completely during leaching. It is known that the presence of even small amounts of Al in silicate materials results in their high acidity and influences distinctly their catalytic activity. For traditional silica alumina catalysts there are numerous techniques for measuring their acidic properties, including the chemical adsorption of ammonia and determining the rate of model reactions. However, the acidic properties of FG materials have not been studied so far. As a model reaction in studying the acidic properties of catalysts, isopropanol dehydration is widely used. It is known that this reaction proceeds on both the Lewis (LAS) and the Brønsted acid sites (BAS) with the formation of propylene and diisopropyl ether, and that acetone forms in a side reaction of isopropanol dehydrogenation [12, 13]. In this work, the acidic and catalytic properties of silicate FG materials were studied using the chemical adsorption of NH3 and isopropanol dehydration. Table 1 Some properties of the FG materials used Sample

Al (wt.%)

Na (wt.%)

SBET (m2/g)

Wpr (mol/g⋅h)

Wa (mol/g⋅h)

Concentration of BAS (µmol/g)

Concentration of LAS (µmol/g)

FG-1 FG-2

0.15 0.53

0.03 0.05

0.7 0.6

0.069 0.049

0.010 0.018

13 36

21 7.2

FG-3 HZSM-5

1.25 2.30

0.08 0.10

1.0 600

0.210 4.2

0.015 1.2

48 165

3.2 91

Wpr – propylene formation rate, Wa – acetone formation rate

EXPERIMENTAL Samples For investigation, we used a commercial sodium silicate FG textile (75–80% SiO2, 16–20% Na2O, 0.1–4% Al2O3) (Steklovolokno, Belarus). The textile was made of threads with a diameter of 1.5 mm spun from elementary fibers of 7– 10 μm in diameter. The initial FG textile was leached in a solution of 5% HNO3 providing a high degree (> 98%) of Na extraction [5], whose residual content in


305

our samples did not exceed 0.1 wt.%. In the experiments we studied three samples of leached FG materials (Table 1, samples FG-1, FG-2 and FG-3) with different amounts of Al impurities. When studying the behavior of FG in isopropanol dehydration, HZSM-5 zeolite was used as a model sample (Si/Al = 17, degree of Na exchange to hydrogen > 95%). The specific surface area (SBET, m2/g) was determined by the BET method using argon thermal desorption. The content of Al and Na in the samples was measured with an ICP-AES (Baird). Sample characterization The catalytic activity of FG materials in isopropanol transformation was studied by the flow-circulating method at 150–400оC. The concentration of the alcohol in the mixture was adjusted by regulating the flow rates of the alcohol vapor and a gas-diluent N2. The mixture of the effluent vapors was analyzed by the GC method with a FID detector. Time on stream was 35–70 min. It was shown that at concentrations of the alcohol vapors in the reaction mixture equal to or exceeding 40%, the reaction rate did not depend on the alcohol concentration, i.e., the reaction had a zero order with respect to alcohol. The acidity of the catalysts was characterized by the alcohol dehydration rate in this range of alcohol concentrations. Initial catalytic activities were obtained by linear extrapolation of the reaction rates to zero time which were also measured in this concentration range. The concentrations of BAS and LAS in the FG materials were determined by the DRIFT spectroscopy of adsorbed NH3 using the intensities of the NH4+ bands and the bands of LAS complexes with NH3 (1430 cm–1 and 1590 cm–1, respectively). A sample in a ceramic holder was placed into a vacuum chamber. After evacuation at 200°C, the DRIFT spectra were recorded using a Shimadzu FTIR-8300 spectrometer with a diffuse reflectance attachment DRS-8000. Then, NH3 was adsorbed on the sample at 200°C and a pressure of 50 torr. The excess NH3 was removed by evacuation at 200°C. The DRIFT spectra were recorded at room temperature with a 4 cm–1 resolution and accumulation of 50 scans. For quantitative measurements, the intensities of the bands (A1430 and A1590) were normalized to the intensity I of a 1900 cm–1 band, which belongs to SiO2 overtones, and which is proportional to the amount of the FG material analyzed. For the calculation, we used the ratios of the integral absorption A1430/I and A1590/I measured for zeolites Hβ and HZSM-5. Before the H2O adsorption/desorption experiments, the FG sample was evacuated at 450°C. After that, the DRIFT spectra were taken. Then, the sample was kept at room temperature and at the pressure of saturated water vapors for a few days. The water content was determined using the intensity of a band at 1630 cm–1.

306


RESULTS AND DISCUSSION Table 1 shows the specific surface areas and contents of Al and Na in the FG materials studied. One can see that the FG materials have specific surface areas around 0.7–1.0 m2/g. Acidic properties according to DRIFT of NH3 adsorption The DRIFT spectra of adsorbed NH3 show (Table 1) that NH4+ ions (ν = 1430 cm–1) which characterizing BAS and the complexes of NH3 with LAS (ν = 1590–1610 cm–1) were formed in all the samples.

5

+

NH4 4

F(R)

3 2 1

NH3

0 -1

1400

1600

1800

-1

2000

Wavenumber (cm )

Fig. 1. DRIFT spectrum of NH3 adsorbed on the sample FG-3

Figure 1 shows the DRIFT spectra of NH3 adsorbed on the FG-3 sample. The concentration of BAS and LAS in the FG materials measured by DRIFT spectroscopy of adsorbed NH3 is shown in Table 1. For the FG materials with an increased Al content, the concentration of BAS increases, whereas the concentration of LAS changes without any visible correlation. Unfortunately, the position of the NH4+ band does not allow the determination of the strength of BAS, because its position (ν = 1430 cm–1) is affected not only by the strength of BAS interacted with NH3, but also by the nearest NH4+ surroundings.


307

However, we suppose that the BAS strength in FG materials is only slightly different from the strength of BAS in silica-alumina catalysts with similar composition [14]. The concentration of BAS is approximately proportional to the Al content. Catalytic activity in isopropanol transformation A study of the catalytic behavior of the FG materials in the isopropanol dehydration showed that the reaction started at 300°C. At 300–350°C, together with the formation of propylene and acetone, diisopropyl ether is also formed. At 400°C, the formation of only propylene and acetone is observed for all the samples. FG materials, as typical acid catalysts, deactivate with time. In our case, the activity decreased in 1 h less than by 50%. Therefore, to compare the activities of the FG materials and their acidic properties by NH3 adsorption, the initial activity of the catalysts was used. Table 1 shows the rate of propylene and acetone formation for the studied FG materials at 400°C. The activity of the samples in alcohol dehydration increases with the increase in the Al amount. The reaction rate referred to the BAS concentration as measured in NH3 adsorption experiments remains constant within the accuracy of one order of magnitude. For FG materials, the specific catalytic activity (0.08–0.21 mol/m2 h), as well as the amount of BAS per the surface unit (10.8–36 sites/100 Å2) exceed considerably those for the zeolite (0.007 mol/m2 h and 0.17 sites/100 Å2, respectively). The amount of BAS per surface unit in the FG materials also exceeds considerably that of a monolayer (ca. 5–6 sites /100 Å2). Therefore, it is reasonable to suppose that BAS are located, at least partly, in the bulk of the glass fibers. On the other hand, the dehydration rate of isopropanol referred to the BAS concentration for the FG materials is less than that for the zeolite, i.e. the catalytic reaction takes place in thin subsurface layers of the fibers, and not all the sites measured by NH3 adsorption are involved in the reaction. This is in a good agreement with the data published elsewhere [7] concerning Pt particles located at 100 Å depth of a glass fiber. The capability of molecules to be removed from or to penetrate into the bulk of FG is demonstrated by the data on water adsorption/desorption obtained in DRIFT experiments. The overall amount of molecular water in the FG-3 is 5 wt.% (2700 µmol/g). This amount exceeds considerably the monolayer parameter (10 µmol/g), which proves the presence of water in the bulk of the FG materials. Figure 2 shows the water content in the FG-3 sample in dependence of the heating temperature and rehydration time. At low temperatures (100–150°C), the water content is much higher for the sample calcined in air than that calcined in vacuum (curves 1 and 2). Therefore, an

308


3

-1

Intensity of band at 1630 cm (arb. un.)

equilibrium exists between water in the gas phase and that in the FG material. At 450°C, almost all molecular water is removed. As we can see (curve 3), the sample FG-3 calcined at 450°C adsorbs water at room temperature for 4 days almost to its initial level. Therefore, the water adsorption/desorption processes in the FG materials at temperatures below 450°C are reversible. Similar processes are assumed to take place in the case of isopropanol and NH3. One can expect that the reversibility of water penetration and removal from FG is determined by the FG structure, since the FG materials are actually viscous liquids. In [15], it was found that the diffusion coefficients of water in the bulk of a glass matrix can be comparable with the diffusion coefficients in liquids.

1

2

3

2 1

0 0

200

400 o

Temperature ( C)

0

1

2

3

4

5

6

Time (days)

Fig. 2. The intensity of an IR band at 1630 cm–1 which is considered to be proportional to the water content in the sample FG-3 versus the heating temperature and rehydration time: 1 – heated in air, 2 – heated in vacuum at 10–3 torr, 3 – rehydration at room temperature and the corresponding pressure of the saturated water vapor (2 vol. %)

As stated above, the activity of the FG materials referred to the BAS number is less than that for the zeolite. The possible reason for that could be the penetration of NH3 throughout the FG bulk, whereas the catalytic reaction takes place only in thin subsurface layers. As it will be shown in the next publication, the method of FG preparation provides a uniform distribution of Al throughout the thickness of the glass fiber. The ratio of BAS to Al achieves 0.8, and since


309

the BAS number could not exceed the number of Al atoms, we conclude that NH3 penetrates throughout the full thickness of the glass fiber. Thus, the DRIFT spectroscopy gives a possibility to measure the overall amount of protonated NH3 molecules in FG materials. However, the method of NH3 adsorption overestimates considerably the number of BAS that are accessible for the reagent molecules in the FG materials. Therefore, it is more preferable to characterize the acidic properties of FG materials via their activity in the reaction of isopropanol dehydration, due to a sufficient peculiarity of the penetration of the NH3 molecules into the FG bulk. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

L. Kiwi-Minsker, I. Yuranov, E. Slavinskaia, V. Zaikovskii, A. Renken: Catal. Today, 59, 61 (2000). V. Holler, K. Radevik, I. Yuranov, L. Kiwi-Minsker, A. Renken: Appl. Catal. B, 32, 143 (2001). B. Louis, C. Tezel, L. Kiwi-Minsker, A. Renken: Catal. Today, 69, 365 (2001). Yu. Matatov-Meytal, V. Barelko, I. Yuranov, M. Sheintuch: Appl. Catal. B, 27, 127 (2000). L.G. Simonova, V.V. Barelko, O.B. Lapina, E.A. Paukshtis, V.V. Terskikh, V.I. Zaikovskii, B.S. Bal’zhinimaev: Kinet. Catal., 42, 693 (2001). L.G. Simonova, V.V. Barelko, E.A. Paukshtis, O.B. Lapina, V.V. Terskikh, V.I. Zaikovskii, B.S. Bal’zhinimaev: Kinet. Catal., 42, 828 (2001). L.G. Simonova, V.V. Barelko, A.V. Toktarev, V.I. Zaikovskii, V.I. Bukhtiyarov, V.V. Kaichev, B.S. Bal’zhinimaev: Kinet. Catal., 42, 837 (2001). L.G. Simonova, V.V. Barelko, A.V. Toktarev, A.F. Cherashev, V.A. Chumachenko, B.S. Bal’zhinimaev: Kinet. Catal., 43, 61 (2002). B.S. Bal’zhinimaev, V.V. Barelko, A.P. Suknev, E.A. Paukshtis, L.G. Simonova, V.B. Goncharov, V.L. Kirillov, A.V. Toktarev: Kinet. Catal., 43, 542 (2002). B.S. Bal’zhinimaev, L.G. Simonova, V.V. Barelko, A.V. Toktarev, V.I. Zaikovskii, V.A. Chumachenko: Chem. Eng. J., 91, 175 (2003). Yu. Matatov-Meytal, M. Sheintuch: Appl. Catal. A, 231, 1 (2002). O.V. Krylov: In: Heterogeneous Catalysis in Chemical Industry, p. 437. Goskhimizdat, Moscow 1955. M.A. Aramendia, V. Borau, C. Jimenez, J.M. Marinas, A. Porras, F.J. Urbano: J. Catal., 161, 829 (1996). E.A. Paukshtis: IR Spectroscopy for Heterogeneous Acid-Base Catalysis, Nauka, Novosibirsk 1992. E.M. Sadovskaya, A.P. Suknev, L.G. Simonova, E.A. Paukshtis, B.S. Bal’zhinimaev: Kinet. Catal., 47, 131 (2006).