The effect of water on the acidity of TiO2 and sulfated titania

https://doi.org/10.1016/j.apcata.2006.01.033Get rights and content

Abstract

Temperature-programmed desorptions (TPD) of isopropylamine (IPA), NH3, and pyridine were compared with diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) of pyridine to determine the effect of H2O on the acidities of three titania catalysts. Brønsted acidity increased in the following order: Degussa P-25 TiO2 < TiO2 synthesized in our laboratory (p-TiO2) < sulfated titania (SO42−/TiO2). Although the traditional interpretation of pyridine infrared spectra showed an apparent increase in Brønsted acidity upon treating SO42−/TiO2 with H2O, IPA TPD spectra showed that H2O displaced IPA from approximately one-third to one-half of the Brønsted sites. Similarly, H2O treatment prior to TPD displaced significant amounts of both NH3 and pyridine. The primary effect of H2O is displacement of strongly adsorbed basic probe molecules from Brønsted sites, rather than the conversion of Lewis sites to Brønsted sites.

Introduction

Solid acid catalysts, especially strong acids such as sulfated titania (SO42−/TiO2), are being studied extensively in an effort to replace liquid-phase acids. Although previous studies [1], [2], [3], generally focusing on sulfated zirconia, concluded that acidity is essential for several important catalytic reactions, the correlation between the type of acidity (Brønsted or Lewis) and catalyst activity is being debated. Some reports [4], [5], which used n-butane isomerization as a probe reaction, suggested that Lewis sites are active whereas others [3], [6], [7] indicated that Brønsted sites are important. To complicate the debate on the importance of Brønsted and Lewis acidity, many studies have proposed that water, present in the feed or produced as a reaction product, converts Lewis sites to Brønsted sites. Although previous studies [3], [6], [8], [9], [10], [11] measured the effect of surface hydration on catalyst activity during acid-catalyzed reactions, surface processes cannot be understood completely until and unless the effect of hydration on the catalytic acidity is investigated. That is, the activity of catalysts can be understood more fully once the changes in the nature of acidity due to hydration are determined. The effect of H2O on acidity is important because water is present in the product stream of almost all hydrocarbon degradation reactions and also in Fischer–Tropsch synthesis [12]. Although the majority of studies focused on sulfated zirconia, no studies were identified that contradicted these findings for TiO2 or SO42−/TiO2.

Previously, infrared spectroscopy (IR) of NH3 illustrated that dry catalysts showed evidence of both Brønsted and Lewis acids sites, and H2O increased IR bands corresponding to NH4+ at the expense of those from NH3 [13], [14], [15], [16], [17]. This observation has lead previous studies to conclude that H2O converts Brønsted sites to Lewis sites. Some studies [4], [5], [18] also suggested that water converts Lewis sites into Brønsted sites that are inactive for n-butane isomerization. In contrast, Tierney and co-workers [19] suggested that water converted Lewis sites to Brønsted sites, which increased catalytic activity. Still other studies [20], [21] indicated that water alone may not produce Brønsted sites that are strong enough to protonate probe molecules but that a combination of sulfation and water treatment is required.

Previous studies reported both the advantages and the disadvantages of hydration on activity of solid acid catalysts and several [5], [19], [22], [23], [24], [25], [26], [27], [28] suggested that an “intermediate” water content maximizes catalyst activity [8], [9], [10], [29], [30]. In the majority of these studies, strong acids such as sulfated zirconia or SO42−/TiO2 were studied by calcining the catalyst at different temperatures, resulting in the loss of water to different extents, and then observing the activity of catalysts during butane isomerization. In fact, most investigations that studied the effect of H2O on acidity deduced their conclusions on acidity enhancements or reductions by measuring the butane isomerization activity of catalysts [4], [5], [8], [9], [11], [28], [31], [32], [33]. Although this method shows how H2O affects activity, it does not elucidate the reason for it. Another limitation of this method is that, typically, the catalytic activity enhancement or reduction is assumed to be due to a change in the number Brønsted sites even though the role played by Lewis acidity is still being debated.

Typically, NH3 or pyridine TPD has determined the strength and number of acid sites on solid acid catalysts [34], [35], but the inability of these experiments to distinguish between Lewis and Brønsted sites [36] allows only identifying sites as ‘weak’ or ‘strong’ based on desorption temperature. Thus, comparing different catalysts by NH3 or pyridine TPD is difficult because the distribution of Lewis and Brønsted sites is not determined. Because catalyst acidity is characterized incompletely, changes in catalyst activity due to differences in the type of acidity cannot be determined. Although one study suggested that H2O can be used to increase the sensitivity of NH3 TPD in distinguishing between Brønsted and Lewis sites, as H2O may displace NH3 from Lewis acid sites [37], the accuracy would depend on how H2O was dosed, its amount, and whether or not it displaced NH3 from Brønsted sites.

Gorte and co-workers [26], [36], [38], [39], [40], [41] suggested that the best method to quantify Brønsted site densities is TPD of amines. Alkyl ammonium ions form by protonation of amines at Brønsted sites and decompose into alkene and NH3 in a narrow temperature range through a reaction similar to Hoffman elimination. This method measures accurately the Brønsted site densities with a sensitivity as low as 2 μmol/g catalyst [42]. Because all of the alkene and NH3 desorb from the catalyst surface in a narrow temperature range [26], [36], [38], [39], [40], [41], the strength of the Brønsted sites cannot be determined by amine TPD. One study [43] suggested that NH3 TPD can be used to measure the strength of acid sites [43] while isopropylamine (IPA) quantifies the Brønsted sites.

The objective of this work is to study the effect of H2O on the acidity of solid acid catalysts through a comparative study of SO42−/TiO2, precipitated-TiO2 (p-TiO2) and Degussa P-25 TiO2. Degussa P-25 was studied because it is the standard catalyst used in photocatalytic oxidation. Because p-TiO2 and SO42−/TiO2 were synthesized using the same method, except for the addition of H2SO4, the effect of sulfation [20], [21] on the conversion of sites from Lewis to Brønsted could be determined. Temperature-programmed desorption studies were carried out using IPA, NH3, and pyridine as probe molecules to compare methods for determining the number and strength of acid sites. Coadsorbing these probe molecules with H2O determined the extent to which H2O converted Lewis sites to Brønsted sites. Although the correlation between activity and acidity is well established, to our knowledge no effort has been made to quantify directly the acidity changes due to hydration. Previous attempts have been made to quantify Brønsted acid sites using IR spectroscopy [11], but the accuracy of this method is questionable [10], especially when compared to TPD of IPA. Thus, in this work TPDs of IPA, NH3, and pyridine were used in conjunction with diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) to quantify acid sites, investigate the strength of acid sites, and to determine the effect of H2O on the acidity of solid acid catalysts.

Section snippets

Catalyst preparation

In a typical precipitated-TiO2 (p-TiO2) synthesis, 0.1 mol titanium butoxide was dissolved in 1.5 mol anhydrous ethanol and this solution was added drop wise to a water–ethanol solution (11 M) until the water-to-butoxide ratio reached 50:1. A white precipitate was produced and the mixture was stirred at room temperature for 3 h. The prepared Ti(OH)4 was powdered below 100 mesh and calcined at 723 K for 4 h before final use. Sulfated TiO2 was prepared by pouring approximately 30 cm3 of aqueous H2SO4

Results

During TPD, IPA formed propene above 600 K on Brønsted sites via a Hoffman elimination type reaction in agreement with studies by Gorte and co-workers [26], [36], [38], [39], [40], [41]. Fig. 1 shows that SO42−/TiO2 had the most Brønsted acid sites, followed by p-TiO2 and P-25. Unreacted IPA that desorbed during TPD quantified the weaker adsorption sites as shown in Fig. 2; approximately 170, 150, and 130 μmol/g catalyst desorbed from SO42−/TiO2, p-TiO2, and P-25, respectively.

For comparison, TPD

Catalyst acidity comparison

Table 1 shows clearly that SO42−/TiO2 has the greatest number of Brønsted acid sites, followed by p-TiO2, with Degussa P-25 having the fewest. The results agree with previous studies [52], [53] that concluded SO42−/TiO2 was a strong solid acid, often described as a superacid. Although IPA TPD quantifies directly the Brønsted sites, it only estimates indirectly the number of Lewis sites. Gorte and co-workers [26], [36], [38], [39], [40], [41] proposed that the unreacted IPA that desorbed below

Conclusion

Water does not convert a significant number of sites from Lewis to Brønsted. Instead, the primary effect of H2O is that it competes with basic probe molecules for both types of acid sites. The results show that H2O displaces readily all probe molecules used in this study (IPA, NH3 and pyridine) from both Lewis and Brønsted sites. Note that H2O displaces strong bases from strong Bronsted sites, which are expected to be the active sites in many acid-catalyzed reactions. Sulfated titania appears

Acknowledgments

This material is based upon work supported by the National Science Foundation under Grant CTS 0223008. Acknowledgment is made to the donors of the Petroleum Research Fund, administered by the American Chemical Society for support of this research.

References (57)

  • A. Corma et al.

    Appl. Catal. A: Gen.

    (1994)
  • D.A. Ward et al.

    J. Catal.

    (1994)
  • G. Yaluris et al.

    J. Catal.

    (1996)
  • R.A. Comelli et al.

    J. Catal.

    (1995)
  • C. Morterra et al.

    J. Catal.

    (1994)
  • B.H. Davis et al.

    Catal. Today

    (1994)
  • M.R. Gonzalez et al.

    Catal. Today

    (1997)
  • J.M. Kobe et al.

    J. Catal.

    (1996)
  • M.R. Gonzalez et al.

    J. Catal.

    (1996)
  • E.P. Parry

    J. Catal.

    (1963)
  • J. Engelhardt et al.

    J. Catal.

    (1990)
  • A. Clearfield et al.

    Catal. Today

    (1994)
  • F. Babou et al.

    J. Catal.

    (1995)
  • L.K. Noda et al.

    Catal. Today

    (2003)
  • K. Arata

    Adv. Catal.

    (1990)
  • M. Bensitel et al.

    Mater. Chem. Phys.

    (1987)
  • J.G. Tittensor et al.

    J. Catal.

    (1992)
  • R.A. Keogh et al.

    J. Catal.

    (1995)
  • F. Babou et al.

    Stud. Surf. Sci. Catal.

    (1994)
  • A. Corma et al.

    J. Catal.

    (2004)
  • M. Nagai et al.

    Catal. Today

    (1997)
  • A.I. Biaglow et al.

    J. Catal.

    (1991)
  • A.I. Biaglow et al.

    J. Catal.

    (1993)
  • A.I. Biaglow et al.

    J. Catal.

    (1994)
  • R.J. Gorte

    Catal. Lett.

    (1996)
  • F. Arena et al.

    Appl. Catal. A: Gen.

    (1998)
  • D.S. Muggli et al.

    J. Catal.

    (1998)
  • D.S. Muggli et al.

    J. Catal.

    (1999)
  • Cited by (52)

    View all citing articles on Scopus
    1

    Present address: Chemical Engineering and Material Science Department, University of California at Davis, Davis, CA 95616, United States.

    View full text