Langmuir 2006, 22, 749-755
749
Structural and Morphological Consequences of High-Temperature Treatments of Hydroxyapatite in the Absence or Presence of HCl Vapor M. I. Zaki,*,† H. Kno¨zinger,‡ and B. Tesche§ Chemistry Department, Faculty of Science, Minia UniVersity, El-Minia 61519, Egypt, Department Chemie und Biochemie, UniVersita¨t Mu¨nchen, Butenandtstrasse 5-13, Haus E, D-81377 Mu¨nchen, Germany, and Max Planck-Institut fu¨r Kohlenforschung, Kaiser-Wilhelm-Platz 1, 45466 Mu¨lheim/Ruhr, Germany ReceiVed June 18, 2005. In Final Form: NoVember 8, 2005 Hydroxyapatite [HAP; Ca5(PO4)3(OH)], a biocompatible, osteoconductive material, was perceived, in the present investigation, to mimic a healthy bone mineral. Structural and morphological properties of its bulk and surface were examined versus high-temperature (up to 900 °C) thermal treatments in air or wet HCl gas atmosphere, using thermogravimetry, X-ray powder diffractometry, N2 sorptiometry, scanning electron microscopy, X-ray energy dispersive spectroscopy, and ex- and in situ infrared spectroscopy. CO, CDCl3, and methylbutynol were used as infrared probe molecules. Results obtained revealed that, in the absence of HCl, the bulk crystalline structure and the chemical composition of HAP were stable during high-temperature treatments. The surface exposed isolated Lewis acid sites (Ca2+) and reactive base sites (Ox- and/or OH-) that chemisorbed atmospheric CO2 molecules with the formation of surface carbonate species (CaCO3). It is assumed that surface OH groups may interact with atmospheric oxygen molecules, leading to the formation and incorporation of peroxide (O22-) species. In the atmosphere of wet HCl, HAP was shown to suffer loss of chemical integrity, facilitated by its carbonated domains, as well as disintegration (or erosion) of particle aggregates and creation of what appeared to be deep groves.
1. Introduction Osteoporosis, meaning abnormalities in the amount and structure of bone that result in reduced bone strength and an increased risk of fractures,1 is a disease affecting 200 million people worldwide, for which there is currently no cure.2 Bone is composed chiefly of the mineral hydroxyapatite [Ca5(PO4)3(OH); denoted HAP], the fibrous protein collagen, blood vessels, and two specific classes of cells, namely, osteoblasts and osteoclasts. Throughout life, bone undergoes two phases of change: (i) a modeling phase, which runs from birth until the age of ∼30 years and leads to bone growth, restructuring, and calcium enrichment; and (ii) a remodeling phase, in which old bone is dissolved (resorbed) by the osteoclast cells (acid-secreting cells) and new bone becomes deposited (mineralized) by the osteoblast cells.1-3 In a healthy organism, a balance is achieved between remineralization and resorption. However, remodeling is controlled by a complex series of regulatory elements including hormones. Thus, accessory hormones and growth factors may disturb this biochemical balance, leading to skeletal disorders such as osteoporosis.4 It is worth noting that the enzyme vacuolar ATPase (V-ATPase) has the job of pumping protons from within the osteoclast cells. This reduces the pH in the cell’s environment to well below 5, that is, acidic enough to dissolve the HAP in bone.1,2,4 Remedial attempts to treat osteoporosis have been leaning on either hormonal or nonhormonal treatments. The latter * Corresponding author. † Minia University. ‡ Universita ¨ t Mu¨nchen. § Max Planck-Institut fu ¨ r Kohlenforschung. (1) Stevenson, J. C.; Lindsay, R. Osteoporosis; Chapman & Hall: Cambridge, 1998. (2) Kee, T.; Dixon, N. Chem. Br. 2001, 37, 38. (3) Skodit, H.; Russell, R. G. G. Cytokine and Bone Metabolism; Gowen, M., Ed.; CRC Press: Baton Rouge, LA, 1992; pp 1-70. (4) Nelson, N.; Harvey, W. R. Physiol. ReV. 1999, 51, 361.
type has been gaining more ground, since hormonal treatments are frequently accompanied by side effects, such as increased risk of certain cancers.5 Increasingly popular nonhormonal treatments are (i) treatment with Vitamin D, and (ii) treatment with a methylene bisphosphonate (MBP) class of drugs. The first treatment plays a role in increasing calcium absorption, whereas, in the second treatment, MBP molecules bind strongly to the calcium at the surface of HAP crystals in growing or regenerating bone.6,7 MBP molecules are specifically concentrated at sites of new bone formation and where bone has been exposed by osteoclast resorptive activity. The fact that MBPs are pyrophosphate analogues in which the [P-O-P] bond has been replaced by a [P-C-P] function, makes MBPs far more stable toward enzyme-catalyzed hydrolysis, but allows them to maintain a good deal of pyrophosphate character, especially in their ability to bind to HAP.6 The above facts and considerations should render the search for surface properties related to the development of osteoporosis worth attempting. We have therefore launched a comprehensive investigation, employing a wide range of materials, research methods, and techniques. The present communication only presents and discusses results of examinations focusing on HAP bulk and surface thermochemical and morphological stabilities to heating in the absence and presence of HCl vapor at elevated temperatures. These conditions, though at variance from the physiological conditions in a living organisms,1 were intended to enhance impacts of the thermal and acid treatments. In these examinations, HAP, which has been proven to be biocompatible8 and osteoconductive9,10 and can be easily processed to matrices with interconnecting pores to allow for bone intergrowth, has (5) Cosman, F.; Lindsay, R. Osteoporosis Primer; Henderson, J. E., Goltzman, D., Eds.; Cambridge University Press: Cambridge, U.K., 2000; p 291. (6) Parfitt, A. M. Am. J. Med. 1991, 91, 425. (7) Flanagan, A. M.; Chambers, T. J. Bone Miner. 1989, 6, 33. (8) Ghorbani, M.; Afshar, A.; Ehsani, N.; Saeri, M. R.; Sorrell, C. C. Int. J. Eng., Trans. B 2002, 15, 173. (9) Heise, U.; Osborn, J. F.; Duwe, F. Int. Orthop. 1990, 14, 329.
10.1021/la051644t CCC: $33.50 © 2006 American Chemical Society Published on Web 12/15/2005
750 Langmuir, Vol. 22, No. 2, 2006
Zaki et al.
been perceived to mimic a healthy bone. The results were obtained using ex- and in situ studies employing Fourier transform infrared (IR) spectroscopy, thermogravimetry (TG), X-ray powder diffractometry (XRD), scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDXS), and nitrogen sorptiometry. 2. Experimental Section 2.1. Materials. HAP (Ca5(PO4)3(OH) ) pentacalciumhydroxide triphosphate) was a white powder, AR-grade Aldrich product; as per the manufacturer, it was dried at 60-65 °C for several days in the final step of its synthesis. In the present work, it was calcined in air for 5 h at 700 or 900 °C. The calcination products are denoted below as HAP(700) and HAP(900), respectively. They were stored over self-indicating silica gel until further use. The HCl solution (36 wt %; d ) 1.18 g/mL), the vapor of which was used for the acid treatments, was a Prolabo product. It was used as supplied. CO, CDCl3, and methylbutynol [MBOH; (CH3)2C(OH)CtCH] were used as IR surface probe molecules, and N2 was the adsorptive gas employed to determine the specific surface area. CO (99.997%) and N2 (99.99%) were products of Linde. CDCl3 (99.9 atm% D; 1% v/v TMS) and MBOH (AR-grade) were supplied by Aldrich. Both liquids were thoroughly deaerated before application, using freezepump-thaw cycles. 2.2. Methods and Techniques. TG was carried out by heating (at 20 °C/min) small amounts (10 ( 2 mg) of HAP up to 900 °C, in flowing air (50 cm3/min), using an automatically recording Du Pont 591 thermogravimetric analyzer. The on-line TA Instruments Thermal Analyzer System (Du Pont Corp.) facilitated data acquisition and handling. XRD was performed at ambient temperature on a Scintag XDS 2000 powder diffractometer using monochromatized Ni-filtered CuKR radiation (λ ) 1.543 Å). The goniometer was operated at 30 kV and 30 mA, at a 0.1-mm divergence slit-width and a 0.5-mm receiving slit-width. The diffractograms were recorded over the 2θ range 10-80° at a step size of 1°, step speed of 1.2°/min, and a count time of 50 s. To warrant a trustworthy comparative assessment of the results, the test sample mass was maintained constant to within (2 mg. N2 sorptiometry was carried out with 500 ( 10-mg test samples, following a 2-h outgassing at 100 °C and ∼10-3 Pa, and cooling to liquid nitrogen temperature (-195 °C), using a purpose-made sorptiometer. The specific surface area (m2/g) was obtained by Brunauer-Emmett-Teller (BET) analysis11 of the N2 adsorption data. SEM spectra were recorded on a Hitachi S3500N scanning electron microscope equipped with an Oxford energy analytical attachment using Oxford-Inka spectrum processing software. With only a few exceptions, the samples were investigated between 10 and 25 kV at a working distance ranging from 5 to 7 mm. The samples were mounted on an adhesive carbon support, which was carried by a metal disk. To ensure good contact with the adhesive surface and satisfactory electrical conductivity of the samples, these were coated with a gold film (approximately 8 nm thick) in a Balzers vacuum coating unit. Ex situ IR spectroscopy was carried out using a Bruker IFS 66 Fourier transform spectrometer, equipped with a liquid-nitrogencooled mercury cadmium telluride (MCT) detector and an on-line spectra acquisition and handling system powered by OPUS 3.1v software. Small amounts (500 °C. It must, however, be noted that a possible alternative type of interaction, namely, the adsorption-absorption of atmospheric CO2, cannot be excluded with certainty. XRD powder diffractograms obtained for HAP and its calcination products at 700° and 900 °C are compared in Figure 2A. The diffraction peaks displayed for the three test samples occur similarly at the following d spacings: 3.43, 3.18, 3.09, 2.8, 2.72, 2.63, 2.26, 1.94, 1.88, and 1.84 Å. These d spacings are very close to those filed in the JCPDS card no. 09-0432 (reviewed partially in the inset in Figure 2A) for HAP crystallized in the hexagonal P63/m space group, containing hexagon-helicoidal channels.18,19 The standard lattice parameters therein reported are a ) 9.418 Å and c ) 6.884 Å, and the OH ions have been found19 to occur specifically in the channels lying along the 63 axes with orientation strictly parallel to the principle c-axis. Perhaps the sole detectable change observed in Figure 2A lies in the obvious intensification of the peaks upon calcination at 900 °C. Hence, the XRD results are in line with the TG results in confirming the bulk structural stability of HAP when heated in air to 900 °C. To quantify the enhanced peak intensities resulting from heating at 900 °C, which are indicative of improved crystallinity, the diffraction peaks at 2θ ) 22.5-37.5° for HAP and HAP(900) were deconvoluted. The results shown in Figure 2B indicate that the crystallization is accompanied by a general intensification of all peaks resolved, although most prominently those located at 2θ values of 27.923 (d ) 3.18 Å; [102] plane), 28.88 (3.09 Å; [210]), 33.20 (2.72 Å; [300]), and 35.80° (2.50 Å; [301]). The areas of these peaks are shown (Figure 2B) to be enhanced by a factor of approximately 2 upon increasing the calcination temperature to 900 °C. Strong IR absorption peaks observed for HAP and its calcination products are displayed in the ex situ spectra compared in Figure 3A. Not shown are two weak absorptions at 1459 and 1416 cm-1. According to earlier IR studies,16,19-23 the absorptions at 3572 and 637 cm-1 are respectively assigned to the stretching (17) Gmelin Institut fu¨r Anorganische Chemie und Grenzgebiete in der MaxPlank-Gesellschaft zur Fo¨rderung der Wissenschaft. Gmelins Handbuch der Anorganischen Chemie, Calcium Teil B, Lieferung 3, 8; Meyer, R. J., Pietsch, E. H. E., Eds.; Verlag Chemie: Weinheim, Germany, 1961; pp 857-860. (18) Elliot, J. C.; Mackie, P. E.; Young, R. A. Science 1973, 180, 1055. (19) Kay, M. I.; Yong, R. A.; Posner, A. S. Nature 1964, 204, 1050. (20) Gadaleta, S. J.; Paschalis, E. P.; Betts, F.; Mendelsohn, R.; Boskey, A. L. Calcif. Tissue Int. 1996, 58, 9. (21) Gadaleta, S. J.; Gericke, A.; Boskey, A. L.; Mendelsohn, R. Biospectroscopy 1996, 2, 353. (22) Sivakumar, M.; Kumar, T. S. S.; Shantha, K. L.; Rao, K. P. Biomaterials 1996, 17, 1709. (23) Joschek, S.; Nies, B.; Krotz, R.; Go¨pferich, A. Biomaterials 2000, 21, 1645.
Figure 2. (A) XRD peaks observed, at the corresponding d spacings (Å) and reflecting planes, for HAP and its calcination products (the inset d spacings and corresponding I/Io values (%) are those filed for standard Ca5(PO4)3(OH) in the JCPDS file indicated). (B) Deconvolution results of diffraction peaks at 2θ ) 22.5 -37.5°, with corresponding relative change in the integrated peak area [HAP/ HAP(900)].
and lattice liberational modes of OH groups. The absorptions at 1101-1025 and 964 cm-1 can be attributed to ν1 and ν3 vibrations of PO43- groups, whereas those at 603, 574, and 566 cm-1 are due to its ν4 vibrations. The two weak bands at 1459 and 1416 cm-1 are respectively assignable to νas and νsOCO vibrations of the carbonate impurity species.21 The ν1 and ν3 vibrations are strongly overlapping and form a broad absorption contour (Figure 3A); these bands and the absorptions due to the ν4 vibrations were therefore deconvoluted. The results, exhibited for HAP and HAP(900) in Figure 3B, resolve the ν1 and ν3 contour into 8 peaks occurring at 1150, 1131, 1102, 1072, 1048, 1022, 993,
752 Langmuir, Vol. 22, No. 2, 2006
Zaki et al.
Figure 3. (A) Ex situ IR spectra taken of KBr-supported samples of HAP and its calcination products. (B) Deconvolution results of the peaks observed for HAP and HAP(900) at the frequency ranges indicated, with corresponding relative change in the integrated peak area [HAP/HAP(900)]
and 963 cm-1, as well as 8 peaks (at 655, 633, 613, 603, 586, 577, 566, and 548 cm-1) for the ν4 vibrations. The resolved peak frequencies are very close to those resolved earlier, using either second-derivative23,24 or two-dimensional (2D)21 IR spectroscopy. Accordingly,21 the peaks at 1102, 993, and 963 cm-1 are due to ν(PO43-) vibrations in an apatitic/stoichiometric environment, whereas those at 1131, 1048, and 1022 cm-1 arise from ν(PO43-) vibrations in a nonstoichiometric/acid phosphate environment. Moreover, the band at 1150 cm-1 is attributable to ν(PO43-) vibrations in an apatitic environment of poorly crystalline HAP.21 When correlated with the XRD peak intensity modifications accompanying the heat-enhanced crystallization (Figure 2B), the IR deconvolution results (Figure 3B) may help, considering the intensification of the following ν(PO43-) absorptions as being most indicative of the crystallization progress: (i) the ν1 and ν3 vibrations at 1131 and 993 cm-1, and (ii) the ν4 vibrations at 603 and 566 cm-1. The absorption due to the OH lattice liberational mode of vibration at 633 cm-1 is also shown to intensify with crystallization, together with the ν(OH) absorption at 3572 cm-1 (Figure 3A). Except for the absorption at 1131 cm-1, all of the crystallization-reminiscent ν(PO43-) absorptions are consistently those arising from phosphates in an apatitic/stoichiometric environment.21 The enhanced OH absorptions may suggest that HAP crystallites grow preferentially in the c-direction, along which lie the structural channels accommodating the OH ions.15 In the meantime, this enhanced OH absorption may appear contradictory to the expected consumption of OH radicals in the formation of peroxide species at high temperatures, as depicted in the reaction equation given above, unless a compensation effect involving OH-site ordering/OH-consumption is considered. 3.1.2. The Surface. BET analysis of N2 adsorption data, determined at liquid nitrogen temperature, resulted in specific surface areas (m2/g) of 59.2 for HAP, 40.7 for HAP(700), and 11.7 for HAP(900), compared to 87-100 m2/g for natural bone minerals,25,26 17-82 m2/g for synthetic HAP,26,27 and e1 m2/g for sintered HAP.14 Thus, the present results reveal that the test (24) Walters, M. A.; Leung, Y. C.; Blumenthal, N. C.; LeGeros, R. Z.; Konsker, K. A. J. Inorg. Biochem. 1990, 39, 193. (25) Misra, D. N.; Bowen, R. L.; Mattamal, G. J. Calcif. Tissue Res. 1978, 26, 139. (26) Wood, N. V. Science 1947, 105, 531. (27) Jarcho, M.; Dombrowski, L. J.; Salsbury, R. L.; Bondley, B. A. J. Dent. Res. 1978, 57, 917.
Figure 4. In situ IR spectra taken of adsorbed CO molecules (at 85 K) on HAP(700) as a function of the gas pressure [1 Torr ) 133.3 Pa].
HAP maintains a high surface accessibility (59.7-40.7 m2/g), even after calcination at temperatures up to 700 °C, which is an essential behavior for HAP biocompatibility.14 Accordingly, HAP and HAP(700) were preferred as test materials for the IR surface spectroscopic studies performed in the present investigation. To identify adsorption sites exposed on HAP, in situ IR spectra were taken from CO/HAP(700) near the liquid nitrogen temperature (85 K). Figure 4 compares the spectra obtained at 22002100 cm-1 as a function of CO pressure (i.e., surface coverage). The spectrum obtained following exposure to 2.7 kPa of CO displays a composite, strong absorption centered around 2167 cm-1 and a weak one at 2140 cm-1. By deconvolution of the composite absorption, two additional maxima at 2173 and 2160 cm-1 were resolved. At lower pressures, the peak at 2140 cm-1 is weakened considerably until it completely disappears at 399 Pa, whereas the shoulder at 2160 cm-1 is shown to persist, even at pressures as low as 133 Pa, and to disappear completely at still lower pressures. In contrast, the peaks at 2173 and 2167 cm-1
High-Temperature Treatments of Hydroxyapatite
remain just as strong down to a 133 Pa. However, they weaken considerably at lower pressures, particularly the latter peak. Concomitantly, a peak emerges at 2180 cm-1, becoming the strongest feature at 0.13 kPa. The band resisted a prolonged evacuation (for 20 min) at the 85 K, keeping its position on the frequency scale. Its location is thus CO-coverage independent. An analogous set of IR absorptions (at 2185, 2167, and 2145 cm-1) was observed previously28 for high coverages of CO adsorbed on CaNaY zeolite at 85 K. Upon evacuation at the same temperature, the 2145 cm-1 band was found to disappear completely, the band at 2167 cm-1 was shifted to higher frequency (to 2175 cm-1) and then disappeared completely, and the 2185 cm-1 band was shifted to higher frequency (to 2191 cm-1) and weakened significantly, leading eventually to the emergence of a weak, but stable band (at 2201 cm-1) at 10-3 Pa. On the basis of literature data,29-33 Hadjiivanov and Kno¨zinger28 assigned the 2145 cm-1 band to physically adsorbed CO, the 2167 cm-1 band to Na+(CO)2 species, and the 2185 cm-1 band to Ca2+(CO)3 species. The high-frequency shift of the 2167 cm-1 band to 2175 cm-1 at reduced coverage was attributed28 to the conversion of Na+(CO)2 into Na+-CO species, whereas the shift of the 2185 cm-1 band to 2191 cm-1 was considered to indicate a decarbonylation of the Ca2+(CO)3 species to yield Ca2+(CO)2 species. A further decarbonylation of the latter species to give Ca2+-CO has been suggested28 to give rise to the lowcoverage stable band at 2201 cm-1, in agreement with other reports.34,35 It is worth mentioning that these authors28 considered (i) the appearance of only one absorption band for the tri- and geminal carbonyl species to account for CO molecules behaving as independent oscillators (i.e., weakly adsorbed), and (ii) the independence of the location of the 2201 cm-1 band on the coverage to reveal that the adsorption sites (Ca2+) are isolated. In view of the close analogy between the reported IR behavior of CO/CaNaY and that observed for the present CO/HAP(700) system, the present 2167 and 2173 cm-1 bands (Figure 4) may be assigned to Ca2+(CO)3 and Ca2+(CO)2 species, respectively, and the emergence of the 2180 cm-1 band (at 0.13 kPa), most probably at the expense of the 2173 cm-1 band, is due to a decarbonylation of the geminal species into the terminal Ca2+CO species. Thus, the coverage-independent frequency of the latter species can be considered suggestive of isolated Ca2+ sites on HAP(700). On the other hand, according to Spielbauer et al.36 the 2160 cm-1 band may be ascribed to CO molecules hydrogenbonded to phosphate species. In support of this assignment, the ν(OH) absorption peak of HAP(700) was red-shifted from 3572 to 3552 cm-1 following CO adsorption. To probe the presence of basic sites on HAP, in situ IR spectra were taken of adsorbed CDCl3 on HAP(700). The spectrum taken of CDCl3/HAP(700) at RT (Figure S1, Supporting Information) displays a single absorption band at 2247 cm-1. According to Paukshtis et al.37 and Zaki et al.,38 this band is due to ν(CD) (28) Hadjiivanov, K.; Knozinger, H. J. Phys. Chem. B 2001, 105, 4531. (29) Paukshtis, E.; Soltanov, R.; Yurchenko, E. React. Kinet. Catal. Lett. 1983, 22, 147. (30) Bordiga, S.; Escalona Platero, E.; Otero Arean, O.; Lamberti, C.; Zecchina, A. J. Catal. 1992, 137, 179. (31) Spoto, G.; Zecchina, A.; Bordiga, S.; Ricchiardi, S.; Martra, G.; Leofanti, G.; Petrini, G. Appl. Catal., B. 1994, 3, 151. (32) Pieplu, T.; Poignant, F.; Vallet, A.; Saussey, J.; Lavalley, J.-C. Stud. Surf. Sci. Catal. 1995, 96, 619. (33) Hadjiivanov, K.; Kno¨zinger, H. J. Catal. 2000, 191, 480. (34) Tsyganenko, A.; Otero Arean, C.; Escalona, Platero, E. Stud. Surf. Sci. Catal. 2000, 130, 3143. (35) Li, P.; Xiang, Y.; Grassian, V. H.; Karsen, S. C. J. Phys. Chem. B 1999, 103, 5058. (36) Spielbauer, D.; Mekhemer, G. A. H.; Riemer, T.; Zaki, M. I.; Kno¨zinger, H. J. Phys. Chem. B. 1997, 101, 4681. (37) Paukshtis, E. A.; Yurchenko, E. N. Russ. Chem. ReV. 1983, 52, 242.
Langmuir, Vol. 22, No. 2, 2006 753
vibrations of either physisorbed or chemisorbed CDCl3 molecules on weak basic sites. The fact that the adsorption of CDCl3 was carried out at RT, and that the observed band persisted against a brief evacuation at the same temperature, may account for chemisorption. The absence of absorptions at higher (up to 2260 cm-1) or lower (down to 2210 cm-1) frequencies than the observed one (2247 cm-1) may exclude the presence of sites on HAP(700) capable of forming bonds with the molecule Cl atom (Cl2(D)CCl...-HO) or with the deuterium atom (Cl3C-D...-O), respectively.37,38 Figure S1 also shows in situ IR spectra taken of MBOH/ HAP(700) at RT and 200 °C. MBOH [(CH3)2C(OH)CtCH] has been successfully used to probe reactive surface basic sites.39-41 The absorptions displayed in the RT spectrum occur mostly at frequencies lower than those reported41 or observed (Figure S2, Supporting Information) for the free molecule: ν(OH), 3645; ν(HCt), 3329; ν(CH), 2996, 2947, and 2892; ν(CtC), 2129; (CH), 1463, 1370, and 1327; δ(OH), 1212; ν(CO), 1180; ν(C-C), 1125 cm-1. The largest red-shifts are those observed for the ν(OH) (3575(ad) vs 3645(free) cm-1) and ν(HCt) (3291(ad) vs 3329(free) cm-1) absorptions of adsorbed MBOH at RT. This observation suggests that the adsorption of MBOH occurs essentially via hydrogen-bonded species40 and the interaction of the acetylenic group with strong Lewis base sites (tC-H...O2-) and/or Lewis acid sites involving the π-electrons.40,41 Upon increasing the adsorption temperature to 200 °C, remarkable changes occur in the spectra taken from the adsorbed (Figure S1) and gas-phase species (Figure S2) of MBOH molecules. The intensities of the molecular vibrations of MBOH decreased significantly. The remaining ν(HCt) are located at two different frequencies (3325 (sharp) and 3258 (broad) cm-1), and strong absorptions emerged at 1683, 1558, and 1416 cm-1. The changes observed in the gas-phase spectrum (Figure S2) indicate an almost 80% conversion of MBOH into acetone and acetylene molecules.41,42 The splitting of the ν(HCt) absorption into two absorptions is indicative of MBOH adsorption via two types of acetylenic interactions with the surface: with Lewis acid (Ca2+) and basic (O2- or OH-) sites.40-42 The latter type of interaction is responsible for the greater red-shift of the ν(HCt) frequency (to 3258 cm-1) relative to that of the former type (to 3325 cm-1).40-42 Consistently, the catalytic conversion of MBOH into acetone and acetylene has been reported to occur on surfaces exposing strong basic sites as well as Lewis acid sites.43 According to Panov and Fripiat44 and Zaki et al.,42 the adsorption and surface condensation of acetone molecules thus produced are responsible for the emerging absorptions at 16831416 cm-1 in the 200 °C spectrum of adsorbed MBOH. Acetone condensation (aldol-condensation-like) products, such as diacetone alcohol (DAA) and mesityl oxide (MSO), have been found to give rise to similar absorptions when adsorbed.42,44 Hence, the present IR results for adsorption and surface reactions of MBOH reveal the presence on HAP(700) of reactive basic sites (Oxand/or OH- sites), together with Lewis acid sites associated with coordinatively unsaturated Ca2+ ions. 3.2. Heating in HCl Vapor. 3.2.1. Chemical Consequences. IR spectra (4000-1000 cm-1) of HAP(700) recorded before and (38) Zaki, M. I.; Hussein, G. A. M.; Mansour, S. A. A.; Ismail, H. M.; Mekhemer, G. A. H. Colloids Surf. A. 1997, 127, 47. (39) Lahousse, C.; Bachelier, J.; Lavalley, J.-C.; Lauron-Pernot, H.; Le Govic, A. M. J. Mol. Catal. 1994, 87, 329. (40) Fouad, N. E.; Thomasson, P.; Knozinger, H. Appl. Catal., A 2000, 194, 213. (41) Hasan, M. A.; Zaki, M. I.; Pasupulety, L. J. Mol. Catal. A 2002, 178, 125. (42) Zaki, M. I.; Hasan, M. A.; Pasupulety, L. Langmuir 2001, 17, 768. (43) Lauron-Pernot, H.; Luck, F.; Popa, J. M. Appl. Catal. 1991, 78, 213. (44) Panov, A.; Fripiat, J. J. Langmuir 1998, 14, 3788.
754 Langmuir, Vol. 22, No. 2, 2006
Zaki et al.
Figure 5. In situ IR difference spectra obtained for HCl/HAP(700) by absorption subtraction of the spectra taken of HAP(700) before exposure to the acid atmosphere at the temperatures indicated.
after exposure to the atmosphere of wet HCl vapor at RT to 400 °C are shown in Figure S3 (Supporting Information). The observed peaks in the spectrum taken before exposure to the wet acid atmosphere characterize the pure (the ν(OH) at 3578 cm-1 and the set of satellite peaks at 2148-2002 cm-1)23 and carbonated (the peaks at 1459 and 1416 cm-1)45 HAP. Relative to the strong phosphate absorptions (at
