J O U R N A L O F C H E M I S T R Y Materials N,N-Dialkylcarbamato complexes as precursors for the chemical implantation of metal cations on a silica support. Part 3† Palladium Luigi Abis,a Daniela Belli Dell’ Amico,b Carlo Busetto,a Fausto Calderazzo,*b Ruggero Caminiti,c Fabio Garbassia and Alessandra Tomeib,d aEnichem S.p.A., Centro Ricerche Novara ‘Guido Donegani’, Via G. Fauser 4, I-28100 Novara, Italy bUniversita` di Pisa, Dipartimento di Chimica e Chimica Industriale, Sezione di Chimica Inorganica, Via Risorgimento 35, I-56126 Pisa, Italy cUniversita` di Roma ‘La Sapienza’, Dipartimento di Chimica and Istituto Nazionale per la Fisica della Materia (I.N.F.M.), P.le Aldo Moro 5, I-00185 Roma, Italy dScuola Normale Superiore, Piazza dei Cavalieri 7, I-56100 Pisa, Italy Received 12th June 1998; Accepted 8th September 1998 Chemical implantation of palladium(II) has been carried out under mild conditions by reacting trans- Pd(O2CNEt2)2(NHEt2)2 with the silanol groups of amorphous silica, carbon dioxide and secondary amine being released in the process.The palladium-containing silica has been characterized and the coordination environment of the implanted cation has been defined by 13C CP MAS NMR, DRIFT and XPS spectra, and by WAXS measurements.Silica-bonded palladium(II) was reduced thermally in vacuo or with dihydrogen at room temperature. Catalytic activity in the hydrogenation of cyclohexene was found for all samples containing the silicasupported reduced palladium; the best results, with rates independent of olefin concentration, were found for the samples treated thermally (200 °C) under reduced pressure. Earlier papers from these laboratories have pointed out that dehydrated silica has been reported in the literature, with release of propylene upon reaction with the silanol groups.6h cationic implantation on silica can be carried out with tin(IV)1a and platinum(II),1b by using the corresponding N,N-dialkylcarbamato complexes, of general formula M(O2CNR2)n, as pre- Experimental cursors. As the silanol groups are the reactive sites, it was anticipated that cations would be homogeneously distributed Materials and reagents on the surface.This method therefore appears to be an useful All operations were carried out in conventional Schlenk tubes alternative to more traditional ones,2 and to the methodology under a dry dinitrogen or argon atmosphere, unless otherwise based on organometallics.3 The use of N,N-dialkylcarbamates specified.Solvents were dried according to conventional presents several advantages, which have been pointed out methods. Carbon dioxide was dried over calcium chloride. earlier;1 implanted species and their reduction products can The secondary amine was distilled from sodium prior to use.be investigated by conventional surface methods. The compound trans-Pd(O2CNEt2)2(NHEt2)2 was prepared The availability of the N,N-diethylcarbamato derivative of according to a method reported earlier.4 Commercial silica palladium(II)4 has urged us to use this compound as starting (Grace SD 3217/50; surface area, 318 m2 g-1; pore volume, material for the implantation of palladium(II) on silica; its 2.22 cm3 g-1) was heated at 160 °C at ca. 10-2 mmHg for 16 h further reduction was predicted to occur easily (E0=0.951 V5), to eliminate most of the physisorbed and chemisorbed water, both thermally and chemically. The use of reduced palladium cooled down to room temperature, and flame-sealed in vials in the hydrogenation of cyclohexene is also reported.under carbon dioxide ( label: SiO2-160). The silanol content Supported palladium represents an important technical probwas assumed to correspond to the weight loss after calcination lem;6a this metal has been implanted on inorganic supports by at 850 °C. The silanol content (expressed as mmol of OH per ion exchange,6b by metal evaporation,6c by solvent extraction gram of silica) was thus estimated to be 2.8.followed by reduction,6d by organometallic chemical vapor deposition (OMCVD) using allyl derivatives,6e and by cluster generation in the presence of an inorganic support.6f In spite Gas-volumetric and elemental analyses of this intense activity in the field, we are not aware of any The carbon dioxide content of the silica-supported palladium prior use of a non-organometallic compound of palladium for species and the carbon dioxide evolved in the course of the a chemoselective reaction with an oxide support under mild implantation reaction were determined in a thermostatted gas conditions. On the other hand, palladium on copper has been burette,7 using liquid media previously saturated with carbon obtained by chemical vapour deposition of volatile paldioxide at the temperature of the experiment.ladium(II) coordination compounds.6g Also, the chemical Elemental (C, H, N) analyses were carried out with a C. interaction of p-allyl derivatives of palladium(II) with partially Erba mod. 1106 elemental analyzer at the Microanalytical Laboratory of the University of Pisa (Faculty of Pharmacy) or in the house.Other elemental analyses were carried out by inductively coupled plasma-atomic emission spectrometry †In partial fulfilment of the requirements for the PhD Thesis of A.T., Scuola Normale Superiore of Pisa. Part 2: ref. 1(b). (ICP-AES) with a Perkin-Elmer Plasma II instrument (Pd). J. Mater. Chem., 1998, 8, 2855–2861 2855Instrumental analysis absorption, polarization and inelastic scattering of the incident X-ray radiation; thus, the total intensity, observed by an 1H and 13C NMR solution spectra were measured with a energy dispersive detector, was corrected accordingly and for Varian Gemini 200 BB instrument, and chemical shifts are the escape peak suppression as well.Experimental conditions expressed in ppm with respect to SiMe4.IR spectra were were: voltage, 45 kV; current, 35 mA; total power, 1.575 kW; measured with a FTIR Perkin-Elmer mod. 1725X instrument energy interval, 16.0–38.0 keV; h values, 26.0, 21.0, 15.5, 10.5, equipped with a KBr beam splitter and a TGS detector, in the 8.0, 5.0, 3.5, 3.0, 2.0, 1.5, 1.0, 0.5 and 0.4°; scattering parameter 4000–400 cm-1 range, using CaF2 windows for liquid samples range (q), 0.16–15.64 A° -1.Normalization to a stoichiometric and KBr or CaF2 plates for Nujol mulls. DiVusion reflectance unit volume containing one palladium atom was performed. infrared Fourier transform (DRIFT) spectra were measured The static structure function i(q) was obtained from the with the same spectrophotometer by mixing the sample with observed intensity I (E, h), and expressed as qi(q)M(q), where dry KBr under an inert atmosphere and by rapid transfer to M(q) is a sharpening factor for a given atom (silicon in this the cell (Spectra Tech).case), defined (c=0.01) as for eqn. (1). The cross polarization magic angle spinning (CP MAS) 13C M(q)=fSi2(0)/fSi2(q) exp(-cq2) (1) NMR spectra were measured at room temperature with a MSL 200 Brucker instrument operating at 50.321 MHz, The Fourier transformation of the experimental static structure chemical shifts being referred to external TMS.functions gives the radial distribution function D(r): XPS spectra were measured with a Perkin-Elmer PHI 5500 ESCA spectrometer equipped with a monochromatic X-ray D(r)=4pr2r0+ 2r p Pqmax qmin qi(q)M(q)sin(qr) dq (2) source and an aluminium anode (Al-Ka radiation, hn= 1483.6 eV), the source being maintained at 14 kV, with a In this equation, r0 is the average electronic density of the power of 200 W.Powdered samples pressed on clean indium sample [r0=.i nifi(0))2V-1], V is the stoichiometric unit foils were used and the diameter of the analyzed sample area volume, ni is the number of atoms i per unit volume and fi is was ca. 400 mm while the background pressure in the analysis the scattering factor for atom i. chamber was 10-8 Pa. For each sample, a preliminary general The EDXD measurements were carried out on both the analysis was performed, in order to detect the presence of parent silica and the palladium-containing silica. The static possible contaminants; the relevant photoemission peaks structure functions are similar, see Fig. 1, thus showing that (Pd 3d , O 1s, C 1s, Si 2p) were recorded under high resolution the presence of palladium(II) does not appreciably modify the conditions. From the photoemission peak intensity, the surface structure of the matrix. atomic concentrations were estimated, using the elemental This experimental observation made the application of the sensitivity factors method.8 Electrostatic charge was attenuated diVerence or isomorphous substitution method possible.12 In by using a low-energy flow electron gun: generally, peaks free the diVerence curve of the radial distribution functions from the typical deformations due to this phenomenon were (SiO2/PdII-SiO2), which contains the contributions due to obtained.the implanted atoms only, the area of each peak is proportional Transmission electron microscopy (TEM) images were to the number of scattering atoms and to their scattering obtained with a JEOL TEM 2010 instrument operating at factors. 200 kV. The material was ground in a mortar until a very fine The coordination environment of palladium was established powder was obtained, which was deposited on a lacy carbon by a curve fitting procedure of the experimental diVerence film supported on a standard copper grid.To avoid deteriorradial distribution function. The experimental static structure ation or contamination, the time required for sample prepfunction i(q) can be interpreted as a weighted sum of partial aration, carried out under dinitrogen, was reduced to a structure functions due to pairs of interacting atoms, by using minimum (ca. 10 min). Bright field images were used in order the Debye function i(q)13 [eqn. (3)] and by adjusting the sij to obtain the size distribution of the palladium particles. and the rij parameters, sij being the rms variation of the Wide-angle X-ray spectroscopy (WAXS) and small angle interatomic distance rij [starting parameters are those obtained X-ray spectroscopy (SAXS) data were collected with a nonin the parent compound4 trans-Pd(O2CNEt2)2(Et2NH)2, commercial energy scanning diVractometer9a equipped with an namely Pd–O 2.022(3), Pd–N 2.058(3) A° .For the non-bond- X-ray generator (water-cooled, tungsten target, 3.0 kW maxiing Pd,Si distance, reference has been made to several mum power), a germanium solid-state detector (SSD) connecmolecular complexes of transition metal cations with alkyl- ted to a multichannel analyzer by means of an electronic and aryl-silanolato ligands which have recently appeared in chain, a collimator system, step motors and sample holder.the literature,14 in addition to the structural data of an iron(III ) The X-ray tube and the detector can rotate in the vertical plane around a common centre in order to reach the appropriate 2h scattering angle.A schematic drawing of the diVactometer has been published earlier.9b WAXS, as applied to liquid10 and amorphous11 systems, allows the static structure function i(q) to be derived, the scattering parameter being q=(4p/l) sinh; 2h is the scattering angle, and l is the radiation wavelength.Since q depends on both E and h, an angular scanning with a monochromatic Xray radiation (ADXD technique) or an energetic scanning with a white X-ray beam at a fixed value of h (EDXD technique) can be performed. In the present case, the latter procedure was used. The EDXD technique presents several advantages, namely: (a) the time of measurement is strongly reduced at approximately constant statistical accuracy; (b) measurements are independent of the intensity fluctuation of the primary beam; (c) the instrument is static during the measurement, which simplifies the instrumental geometry and reduces the errors due to misalignment.On the other hand, Fig. 1 Structure function qi(q)M(q) (e.u. A° -1) vs.q (A° -1) of SiO2-160 (,) and palladium-containing silica (———). energy-dependent phenomena have to be considered, such as 2856 J. Mater. Chem., 1998, 8, 2855–2861silicate14a]. i(q)=.fi(q)fj(q) sin(qrij ) qrij exp(-1/2sij2q2) (3) The palladium containing silica samples were reduced both thermally and chemically (with dihydrogen), vide infra, and were subjected to both WAXS and SAXS measurements. In the latter case, the intensity of the scattered X-ray radiation I(E, h), with h<1°, is related to the size and shape of the scattering centres, as for the Guinier law expressed by eqn.(4),15 where Rg is the gyration radius of the scattering particle, which depends on both its shape and size. The gyration radius can be determined from the slope of the plot of lnI(q) vs.q2. Fig. 2 13C CP MAS NMR spectrum of palladium(II ) supported on SiO2 resulting from the reaction of trans-Pd(O2CNEt2)2(NHEt2)2 I(q)=I(0) exp (-1/3Rg2q2) (4) with the silanol groups (sample AT-323). Irradiating field, 50 kHz; spinning rate, 5 kHz; contact time, 5 ms; sequence recycle time, 4 s; For a given shape of the particle, the gyration radius is related number of transients, 16 640; spectral width, 20 kHz; time domain to the particle size by simple equations [for a cube, Rg2=l2/4; points, 2048; chemical shifts are referred to SiMe4.for a sphere, Rg2=(3/5) R2]. For the small-angle X-ray scattering (SAXS) measurements, the scattering angle h was 0.3, 0.4, and 0.5°, with a slit width sponded to a CO2/Pd molar ratio of 0.8. Addition of excess of about 60 mm in order to reduce the X-ray angular diver- acetic acid caused the evolution of 1.2 mol of carbon dioxide gence.A blank measurement showed that no intensity was per palladium. In another experiment (AT-203), the suspension monitored by the detector. after the reaction with the silica was filtered under carbon dioxide and the filtrate was treated with excess acetic acid: no Chemical implantation carbon dioxide was evolved.Palladium implantation on silica was carried out by the Reduction of palladium following procedure. Silica SiO2-160 (DB-14–214, 5.6 g, corresponding to 15.7 mmol of silanol groups) was added to a With dihydrogen. In a preparative experiment, the palladiumtoluene (100 cm3) solution of trans-Pd(O2CNEt2)2(NHEt2)2 containing silica was reduced at room temperature under (1.01 g, 2.08 mmol; OH/Pd molar ratio, 7.5) in a 500 cm3 flask dihydrogen at atmospheric pressure in benzene (AT-166, and the mixture was stirred at room temperature for 2 h, Found: C, 4.2; H, 1.0; N, 0.6; Pd, 2.2%), see Table 2.WAXS occasionally reducing the partial pressure of carbon dioxide and SAXS measurements were carried out on this sample.released in the process. After being recovered by filtration, the Reduction was also achieved by treating silica-supported palyellow palladium-containing silica was dried in vacuo for 20 h ladium(II) with dihydrogen at room temperature for 6 d, in (5.75 g) and the following analytical results were obtained: the absence of any solvent. AT-19, C, 5.5; H, 1.3; N, 1.5; Pd, 4.0; CO, 1.5%, corresponding In a gas volumetric experiment, the palladium-containing to a substantially quantitative yield (see Table 1) of the silica (AT-323; 0.92 g; Pd, 3.4%; 0.29 mmol of palladium) was implantation reaction and to the following molar ratios: added to cyclohexane (25 cm3) presaturated with dihydrogen CO2/Pd, 0.9; N/Pd, 2.8.The 13C CP MAS NMR spectrum and the volume of dihydrogen was measured at constant showed resonances (d, ppm from SiMe4) at 12.0 (CH3), 41.1 temperature (24.0±0.1 °C) and pressure (1 atm).The orig- (CH2), and 164.0 (O2C ), see Fig. 2. The 13C NMR spectrum inally yellow silica became brown and finally black after the of the parent compound trans-Pd(O2CNEt2)2(NHEt2)2 completion of the gas absorption (0.43 mmol, corresponding has bands ([2H8]toluene, d, ppm from SiMe4) at 14.1 to a H2/Pd molar ratio of 1.5).In a blank experiment (AT- [NH(CH2CH3)2], 14.5 [O2CN(CH2CH3)2], 41.6 [NH(CH2CH3)2], 395) carried out with SiO2-160, noH2 was found to be absorbed 46.0 [O2CN(CH2CH3)2], 164.9 (O2C). The DRIFT spectrum under the same experimental conditions. showed bands at 3259, 2978, 2939, 2882, 1598, 1549, 1483, 1463, 1426, 1382 and 1303 cm-1.The parent compound has Thermal reduction. A sample of the palladium-containing IR bands (PCTFE mull ) at 3060, 2985, 2935, 2870, 1590, silica (AT-323; 2.1 g; Pd, 3.4%; 0.67 mmol of palladium) was 1555, 1475, 1455, 1440, 1410, 1375 and 1325 cm-1. For the heated at 100 °C in vacuo (3 mmHg) for ca. 7 h; the brown XPS data and other details on the implantation reactions, product was stored under dinitrogen (AT-381, Found: C, 3.6; see Table 1.H, 0.9; N, 0.7%). The implantation reaction was monitored gas volu- Another sample (AT-323; 1.7 g; Pd, 3.4%, corresponding to metrically, using a large excess of silica. A toluene (25 cm3) 0.54 mmol of palladium) was heated at 200 °C in vacuo suspension of SiO2-160 (5.4 g, 15.1 mmol of silanol groups), (5×10-2 mmHg) for 6 h; at the end of the treatment the presaturated with carbon dioxide, was treated with the palladium( II) complex (0.20 g, 0.41 mmol, for a OH/Pd molar Table 2 Reduction of palladium ratio of 37) at 20±0.1 °C: the evolved carbon dioxide corre- Reduction XPSc TEMd Sample Precursor methoda,b Eb/eV d/nm Table 1 Implantation of palladium(II) on silica AT-166e AT-90 a¾ 335.3 n.d.DB-16–9 AT-323 a n.d. 6–8 Molar ratio Sample OH/Pd Yield (%) Pd (%) XPSaEb/eV AT-381e AT-323 b¾ 335.7 3–6 AT-408e AT-323 b 335.5 2–5 AT-19 7.5 quant. 4.0 336.3 aReduction with H2 at room temperature: a¾, in benzene; a, without AT-323b 6.9 83 3.4 336.3 solvent. bThermal reduction: b¾, T=100 °C; b, T=200 °C. cPd 3d5/2 AT-355 20.5 88 1.8 336.1 binding energy (eV) (±0.2 eV); for reference Eb values see Table 1, footnote a.dPalladium particle, mean diameter. eSamples used for aPd 3d5/2 binding energy. Reference data (eV ) (±0.2 eV): Pd(s), 335.1; PdO, 336.1.8 bSample used for WAXS experiments. WAXS experiments. J. Mater. Chem., 1998, 8, 2855–2861 2857Fig. 4 WAXS data: radial distribution function D(r) (×10-3 e2 A° -1) Fig. 3 Catalytic hydrogenation of cyclohexene. Moles of dihydrogen absorbed as a function of time ($). Solvent, cyclohexane; temperature, vs. r for SiO2-160 (,) and palladium-containing silica (———), AT-323. 24.1±0.1 °C; cyclohexene, 2.3 mmol. sample became brown (AT-408, Found: C, 0.6; H, 0.0; N, The gas volumetric data, see Experimental section, show that 0.0%). For further data concerning the reduced samples, at least 80% of palladium is chemically bonded to the silica see Table 2.surface. The XPS binding energies (336.1–336.3 eV) are to be compared with the value of 336.1 for PdO.8a,b Moreover, the Catalysis 13C CP MAS NMR data, see Fig. 2, show the presence of Palladium-containing silica (AT-408, 0.23 g; Pd, 3.4%; residual carbamato groups (resonances around 160 ppm), thus 0.073 mmol of palladium), thermally pre-treated at 200 °C was confirming that the silica-bonded palladium still maintains suspended in cyclohexane (25 cm3); the suspension was satu- part of the original coordination environment.As far as other rated with dihydrogen at 24.1±0.1 °C for 3 h and then added resonances are concerned (amine or carbamato ethyl groups), of cyclohexene (0.25 cm3, 2.4 mmol) under vigorous magnetic the intrinsic low resolution of the solid-state NMR spectrum stirring. The olefin hydrogenation was independent of the does not allow any specific assignment to be made.The IR cyclohexene concentration and substantially complete in ca. reflectance spectra show bands in the 1600–1300 cm-1 region, 5 min, corresponding to a H2/Pd molar ratio of 33.The plot which closely resemble those of the parent palladium(II) of dihydrogen absorption as a function of time is shown in compound. Fig. 3. A catalytic activity of 70 mmol of H2 min-1 gPd-1 was WAXS measurements allowed the radial distribution estimated from the plot of Fig. 3; this corresponds to an function, see Fig. 4, to be calculated, in comparison with the apparent turnover frequency of 0.12, expressed as mols of data for the parent silica.dihydrogen absorbed per mol of palladium per second. The diVerence radial distribution function was calculated A commercially available sample of palladium supported (see Experimental section) and a curve fitting procedure of on a silico-aluminate matrix (AT-416, 0.11 g; Pd, 2%; this function was carried out by Fourier transformation of the 0.021 mmol of palladium) suspended in cyclohexane (25 cm3) intensities obtained by the Debye formula of eqn.(3), using and cyclohexene added (0.3 cm3, 2.9 mmol) absorbed the the same M(q) and qmax as for the experimental data. Starting expected amount of dihydrogen for complete hydrogenation bonding and non-bonding parameters are from the literain about 6 min at 24.2±0.1 °C (for a H2/Pd molar ratio of ture:4,14 particularly relevant in this connection are the results 138).The rate of cyclohexene hydrogenation was substantially of a crystallographic study14f on a silanolato complex of independent of olefin concentration: a catalytic activity of palladium(I) showing Pd,Si non bonding distances of 2.981 250 mmol of H2 min-1 gPd-1 was estimated, corresponding to and 3.520 A° .The Debye–Waller factors (sij, A° ) are considered an apparent turnover frequency (as defined above) of 0.44 s-1. as parameters in the curve fitting procedure (sij=0.04 A° for Samples of silica-supported palladium, prereduced with rij1.50 A° ; sij=0.08 A° for 1.50<rij2.20 A° ; sij=0.13 A° for dihydrogen at room temperature or thermally at 100 °C were 2.20<rij3.50 A° ; sij=0.20 A° for rij>3.50 A° ).The experfound to catalyze cyclohexene hydrogenation, the gas absorp- imental and calculated diVerence radial distribution functions tion being dependent on olefin concentration (initial rates, are shown in Fig. 5; in the model, see Fig. 6, which gave the respectively: 2.9 and 22 mmol of H2 min-1 gPd-1).best fit of the experimental curve, each palladium atom is coordinated to one silanolato and two diethylamine ligands, Results and discussion and to a residual carbamato group as well. Implantation Reduction The N,N-diethylcarbamato complex of palladium(II), trans- The reduction of the silica-supported palladium was carried Pd(O2CNEt2)2(NHEt2)2, reacts with the silanol groups of out both with dihydrogen at room temperature or thermally partially dehydroxylated silica in toluene at room temperature.(at 100 or 200 °C) under reduced pressure. On the basis of the experimentally determined volume of carbon dioxide evolved in the reaction, see eqn. (5), the Reduction with dihydrogen. The reduction process was implantation was found to involve the evolution of about one followed gas volumetrically and found to require 1.4 mol of mol of carbon dioxide per palladium.H2 per palladium. Eqn. (6) represents the idealized reduction process for a palladium(II ) centre containing both carbamato and silanolato groups. The most eYcient reduction method, trans-Pd(O2CNEt2)2(NHEt2)2+n :Si–OHA A(:Si–O)nPd(O2CNEt2)2-n(NHEt2)2+n CO2+n NHEt2 (n1) as judged from the XPS binding energies approaching that of palladium bulk, is the treatment with dihydrogen, see Table 2.(5) 2858 J. Mater. Chem., 1998, 8, 2855–2861Fig. 7 WAXS data. Radial distribution functions D(r) (×10-3) for Fig. 5 WAXS data. Experimental (——) and calculated (,) diVerence palladium containing silica after reduction with dihydrogen at room radial distributions of silica-coordinated palladium (sample AT-323).temperature (———), sample AT-166 and for SiO2-160 (,). D(r) values (×10-3) were calculated by using the Debye formula and the interatomic parameters specified in text. 166 and the silica support were obtained, see Fig. 8. These data show that reduced palladium gives metallic crystallites, with some long-range ordered structure.Fig. 8 shows peaks at ca. 2.7, 3.9 and 4.8 A° , corresponding, respectively, to the palladium–palladium distances of the twelve nearest- (2.751 A° ), the six second-nearest- (3.891 A° ), and the twentyfour third-nearest (4.765 A° ) neighbours of the fcc crystal lattice of bulk palladium17 [a=3.8900(7) A° ], the twelve fourthnearest neighbours being at 5.502 A° .An approximate evaluation of the mean dimension of the metal particles was carried out by SAXS measurements: a value of 19±1 A° was obtained for the gyration radius, corresponding to a radius of 24±2 A° for a spherical shape and to an edge of 38±2 A° for cubic shape. The theoretical structure function qi(q)M(q) was calculated for metal particles containing 2457 palladium atoms (d= 2.750 A° ), corresponding to cubic particles with l=31 A° , the Debye–Waller factors being s=0.15 A° for 0d5.60 A° and s=0.20 A° for d>5.60 A° .The curve of the structure function superimposed on the experimental one is shown in Fig. 9. Fig. 6 Suggested model of the coordination shell of silica-bonded For the palladium-containing samples subjected to thermal palladium(II), see text; bond distances (A° ): Pd–N(1) 2.06(8), Pd–N(2) treatment under reduced pressure (at 100 or 200 °C), reduction 2.06(8), Pd–O(1) 2.02(8), Pd–O(2) 2.02(8), the Pd,Si non-bonding to palladium(0) was anticipated by the darkening of the distance is 3.08(0.13) A° .substance and by the slight decrease of the XPS binding energy values with respect to PdO (336.1 eV).On the other hand, the WAXS structure functions for both samples, see Fig. 10, are :Si–O–Pd(O2CNEt2) (NHEt2)2+H2A CO2+3 NHEt2+:Si–OH+Pd (6) similar to that of the unreduced substance, thus showing that only partial reduction had occurred (the analytical nitrogen The excess of dihydrogen absorbed with respect to the content decreasing with increasing temperature is suggestive stoichiometry of reaction (6) is presumably used to form Pd–H bonds (palladium is reported 16 to react with dihydrogen forming a so-called b phase characterized by a H/Pd molar ratio of 0.6 at room temperature, with a dihydrogen equilibrium pressure of ca. 10 mmHg; other palladium hydride phases are known16c–g) or to reduce part of carbon dioxide formed in the process. The palladium-containing silica was also treated thermally under reduced pressure; the sample showed only partial reduction to palladium(0), as suggested by the XPS data, with an Eb (335.7 eV for the sample thermally treated at 100 °C and 335.5 eV for that treated at 200 °C), slightly but signifi- cantly higher than those pertaining to bulk palladium (335.1 eV) and to the sample reduced with dihydrogen at room temperature (335.3 eV).The WAXS and SAXS data of the samples reduced with dihydrogen will be discussed first and then compared with those obtained by thermal reduction. The radial distribution functions for the reduced palladiumcontaining silica and for the parent silica are shown in Fig. 7. Fig. 8 WAXS data. Experimental DiV(r) (×10-3) of AT-166 after After correction for the contribution by the average bulk reduction with dihydrogen at room temperature.The silica contribution was subtracted. electron density (4pr2r0), the DiV(r) functions of sample ATJ. Mater. Chem., 1998, 8, 2855–2861 2859Fig. 9 Calculated structure function qi(q)M(q) vs. q for palladium metal particles (clusters of 2457 atoms, l=31 A° ) with face-centred cubic unit cell (———) compared with the experimental curve (,), AT-166.Fig. 11 TEM image (1.7 nm cm-1; ×6 000 000) of one of the palladium particles; sample DB-16.1, obtained by thermal reduction at 200 °C under reduced pressure. the TEM image of the thermally (200 °C) treated sample, see Fig. 11; the picture shows two diVerent orientations of the atomic planes within the same particle.Catalytic hydrogenations The highest catalytic activity was observed with the palladium catalyst thermally pretreated under reduced pressure at 200 °C, see Experimental section and Fig. 3. In this case, the plot of the dihydrogen absorbed vs. time shows a zero-order dependence with respect to olefin concentration (under the conditions of the experiment, the dihydrogen concentration is constant).Heterogeneous catalytic hydrogenations of unsaturated substrates have been extensively studied.18 Rates have frequently been found to be zero-order with respect to olefin concen- Fig. 10 Structure functions qi(q)M(q) (e.u. A° -1) vs. q (A° -1) of tration, and this has been attributed to the dissociative adsorp- SiO2-160 (,) and palladium-containing silica (———): (A) after tion of dihydrogen being rate determining.19 On the other thermal treatment at 100 °C (AT-381); (B) after thermal treatment at hand, poisoning may lead to a decreased adsorption rate of 200 °C (AT-408).These curves should be compared with those shown the olefin by the catalyst, thus possibly leading the overall rate in Fig. 1. to become dependent on substrate concentration. Some of the catalysts prepared in the course of this study retain secondary of palladium(II) persisting in the thermally treated samples at amine, after thermal (100 °C under reduced pressure) or chemilower temperature).cal (by dihydrogen at room temperature) activation. The Reduction to palladium(0) in all three cases (room thermal treatment at 200 °C under reduced pressure shows no temperature with dihydrogen, 100 or 200 °C) has been con- residual amine and the rate of hydrogenation is independent firmed by transmission electron microscopy (TEM) measure- of substrate concentration, in agreement with similar findings ments which have evidenced the formation of metallic particles. for other supported palladium catalysts in hydrogenation The particle size appears to increase as the temperature of the reactions.19b,20 Although the time required for completing the reduction decreases, the larger distribution being comprised reaction is ca. 5 min, a diVusion-controlled process is unlikely between 6 and 8 nm for the sample reduced with dihydrogen because of the very eYcient stirring. Similar observations were at room temperature. The apparent disagreement between the made for the silica-supported catalysts based on platinum.1b SAXS and TEM measurements concerning the particle size may be reconciled by considering that TEM measurements Conclusions may overestimate the particle size due to the fact that the contribution by the smaller sizes may become almost negligible.This paper has shown that a palladium silicate can be prepared on a partially hydroxylated amorphous silica by protonation An enlargement of one of the palladium particles is shown in 2860 J.Mater. Chem., 1998, 8, 2855–286146, 197; (c) Y. Takasu, R. Unwin, B. Tesche, A. M. Bradshaw and (the silanol groups are the reactive sites on the surface) of a M. Grunze, Surf. Sci., 1978, 77, 219; (d) G. L. Haller and hydrocarbon-soluble N,N-dialkylcarbamato derivative of pal- D.E. Resasco, Adv. Catal., 1989, 36, 173. ladium(II). The implantation reaction depends on the OH/Pd 9 (a) R. Caminiti, C. Sadun, V. Rossi, F. Cilloco and R. Felici, 25th molar ratio used; in the present case, when such a ratio is &7, Italian Congress of Physical Chemistry, Cagliari, Italy, June 17–21, the implantation reaction is substantially quantitative.This 1991; Ital.Pat., 01261484, June 23rd 1993; (b) M. Carbone, R. Caminiti and C. Sadun, J. Mater. Chem., 1996, 6, 1709. third paper of the series shows that a readily available molecu- 10 (a) R. Caminiti, R. Cucca and T. Radnai, J. Phys. Chem., 1984, lar compound of palladium(II) can be used to carry out the 88, 2382; (b) G. Paschina, G. Piccaluga and M.Magini, J. Chem. chemical implantation on the silica surface, with a presumably Phys., 1984, 81, 6201; (c) R. Caminiti, D. Atzei, P. Cucca, uniform distribution of the palladium(II ) centres. As the A. Anedda and G. Bongiovanni, J. Phys. Chem., 1986, 90, 238; palladium(II) precursor is easily prepared from the cationic (d) R. Caminiti, C. Sadun, M. Basanisi and M. Carbone, J. Mol.acetonitrile complex of palladium(II) [Pd(MeCN)4]2+ through Liq., 1996, 20, 55. 11 (a) R. Caminiti, C. Munoz Roca, D. Beltran-Porter and the oxidation of palladium metal by NO+, this paper discloses A. Z. Rossi, Z. Naturforsch., Teil A, 1988, 43, 591; (b) A. Musinu, a facile method of ultimately obtaining palladium particles on G. Piccaluga and G. Pinna, J. Non-Cryst. Solids, 1990, 122, 52; silica based on a simple and selective chemical methodology (c) A.Capobianchi, A. M. Paoletti, G. Pennesi, G. Rossi, in the preliminary step. It is easy to predict that this method- R. Caminiti and C. Ercolani, Inorg. Chem., 1994, 33, 4635; (d) ology can be extended to implant palladium(II) and thus to D. Atzei, R. Caminiti, C. Sadun, R. Bucci and A. Corrias, produce palladium particles on other inorganic, partially Phosphorus Sulfur Silicon, 1993, 79, 13. 12 (a) W. Bol, G. J. H. Gerrit and C. van Panthaleon, J. Appl. hydroxylated, matrices, such as alumina and titania and work Crystallogr., 1970, 3, 486; (b) G. Johansson and R. Caminiti, Z. is now in progress aimed at verifying this prediction. Naturforsch., Teil A, 1986, 41, 1325; (c) R. Caminiti, F.Cilloco This work was partly supported by the Consiglio Nazionale and R. Felici, Mol. Phys., 1992, 76, 681. 13 R. W. James, The Optical Principles of the DiVraction of X-rays, delle Ricerche (C.N.R., Roma), Progetto Strategico ‘Tecnologie G. Bell & Sons Ltd, London, 1950. Chimiche Innovative’. We are indebted to Chimet SpA, Badia 14 (a) M. J. Auburn, R. D. Holmes-Smith, S.R. Stobart, al Pino, Arezzo, Italy for a sample of their commercially M. J. Zaworotko, T. S. Cameron and A. Kumari, J. Chem. Soc., available oxide-supported hydrogenation catalyst (Chimet D Chem. Commun., 1983, 1523; (b) P. Braunstein, M. Knorr, 9050, lot 81; Pd 2 wt.%) and for a loan of palladium. A. Tiripicchio and M. Tiripicchio Camellini, Angew. Chem., Int. Ed. Engl., 1989, 28, 1361; (c) T.A. Budzichowski, S. T. Chacon, M. H. Chisholm, F. J. Feher and W. Streib, J. Am. Chem. Soc., References 1991, 113, 689; (d) M. Huang and C. W. DeKock, Inorg. Chem., 1993, 32, 2287; (e) A. Vizi-Orosz, R. Ugo, R. Psaro, A. Sironi, 1 (a) L. Abis, D. Belli Dell’ Amico, F. Calderazzo, R. Caminiti, F. Garbassi, S. Ianelli, G. Pelizzi, P. Robino and A. Tomei, J. Mol. M.Moret, C. Zucchi, F. Ghelfi and G. Pa�lyi, Inorg. Chem., 1994, 33, 4600; ( f) M. Knorr, P. Braunstein, A. Tiripicchio and Catal., 1996, 108, L113; (b) L. Abis, D. Belli Dell’ Amico, C. Busetto, F. Calderazzo, R. Caminiti, C. Ciofi, F. Garbassi and F. Ugozzoli, Organometallics, 1995, 14, 4910; (g) K. Su, T. Don Tilley and M. J. Sailor, J. Am. Chem. Soc., 1996, 118, 3459; G. Masciarelli, J.Mater. Chem., 1998, 8, 751. 2 (a) J. M. Tour, S. L. Pendalwar and J. P. Cooper, Chem. Mater., (h) K. W. Terry, C. G. Lugmair, P. K. Gantzel and T. Don Tilley, Chem. Mater., 1996, 8, 274. 1990, 2, 647; (b) B. Breitscheidel, J. Zieder and U. Schubert, Chem. Mater., 1991, 3, 559; (c) J. A. Schwarz, C. Contescu and 15 O. Glatter and O. Kratky, Small Angle X-ray Scattering, Academic Press, London, 1982.A. Contescu, Chem. Rev., 1995, 95, 477; (d) K. Okitsu, H. Bandow, Y.Maeda and Y. Nagata, Chem.Mater., 1996, 8, 315. 16 (a) W. Palczewska, Adv. Catal., 1975, 24, 245; (b) J. J. S. Scholten and J. A. Konvalinka, J. Catal., 1966, 5, 1; (c) H. W. King and 3 D. G. H. Ballard, Adv. Catal., 1973, 23, 267; Yu. I. Yermakov and V. Zakharov, Adv. Catal., 1975, 24, 173; H.C. Foley, F. D. Manchester, Phys. Status Solidi, 1971, 45, 527; (d) Y. P. Khodyrev, R. V. Baranova, R. M. Imamov and S. J. DeCanio, K. D. Tau, K. J. Chao, J. H. Onuferko, C. Dybowski and B. C. Gates, J. Am. Chem. Soc., 1983, 105, 3074; S. A. Semiletov, Kristallofiya, 1978, 23, 1046; (e) Y. P. Khodyrev, R. V. Baranova, R. M. Imamov and S. A. Semiletov, Izv. Akad. J. Schwartz, Acc.Chem. Res., 1985, 18, 302; Y. Iwasawa, Inorganic Oxide-attached Metal Catalysts, in, Tailored Metal Catalysts, ed. Nauk SSSR, Neorg. Mater., 1978, 14, 1645; ( f) S. A. Semiletov, R. V. Baranova, Yu. P. Khodyrev and R. M. Imamov, Y. Iwasawa and D. Reidel Publ. Co., Dordrecht, 1986; S. A. Vierko� tter, C. E. Barnes, T. L. Hatmaker, J. E. Penner- Kristallofiya, 1980, 25, 1162; ( g) R.V. Baranova, Yu. P. Khodyrev, R. M. Imamov and S. A. Semiletov, Hahn, C. M. Stinson, B. A. Huggins, A. Benesi and P. D. Ellis, Kristallofiya, 1980, 25, 1290; (h) P. Chou and M. A. Vannice, Organometallics, 1991, 10, 3803; T. J. Marks, Acc. Chem. Res., J. Catal., 1987, 104, 1; (i) C. H. Foo, C. T. Hong and C.-T. Yeh, 1992, 25, 57; S. L. Scott, P. Dufour, C. C. Santini and J.-M.Basset, J. Chem. Soc., Faraday Trans., 1989, 85, 65; ( j) S.-C. Chou, S.-H. Inorg. Chem., 1996, 35, 869; V. Vidal, A. The�olier, J. Thivolle- Lin and C.-T. Yeh, J. Chem. Soc., Faraday Trans., 1995, 91, 949, Cazat and J.- M. Basset, Science, 1997, 276, 99. and references therein; (k) Y. Sakamoto, M. Imoto, K. Takai, T. 4 A. Anillo, D. Belli Dell’ Amico, F. Calderazzo, M. Nardelli, Yanaru and K.Ohshima, J. Phys.: Condens. Matter, 1996, 8, 3229. G. Pelizzi and L. Rocchi, J. Chem. Soc., Dalton Trans., 1991, 2845. 17 (a) JCPDS, Joint Committee on Powder DiVraction Standards, 5 Handbook of Chemistry and Physics, ed. D. R. Lide, CRC Press, 1601 Park Lane, Swarthmore, PA, Card No. 5–0681; Boca Raton–Ann Arbor, London–Tokyo, 75th edn., 1994–1995, (b) H. E. Swanson and E.Tatge, NBS Circular No. 539, 1953, p. 8–21. vol. 1, p. 21; (c) J. Donohue, The Structure of the Elements, 6 (a) Z. Karpinski, Adv. Catal., 1990, 37, 45; (b) Z. Zhang, J. Wiley, New York, 1974; (d) K. L. Shelton, P. A. Merewether F. A. P. Cavalcanti and W. M. H. Sachtler, Catal. Lett., 1992, 12, and B. J. Skinner, Can. Mineral., 1981, 19, 599. 157; (c) A. Renou andM. Gillet, Surf. Sci., 1981, 106, 27; (d) F. C. 18 (a) J. Horiuti and M. Polanyi, Trans. Faraday Soc., 1934, 30, 1164; Meldrum, N. A. Kotov and J. H. Fendler, Chem. Mater., 1995, 7, (b) E. E. Gonzo and M. Boudart, J. Catal., 1978, 52, 462; 1112; (e) J.-C. Hierso, C. Satto, R. Feurer and P. Kalck, Chem. (c) G. Leclercq and M. Boudart, J. Catal., 1981, 71, 127; Mater., 1996, 8, 2481; (f ) G. Schmid, M. Harms, J.-O. Malm, J.-O. (d) N. Koga, C. Daniel, J. Han and K. Morokuma, J. Am. Chem. Bovin, J. van Ruitenbeck, H.W. Zandbergen and W. T. Fu, J. Am. Soc., 1987, 109, 3455; (e) L. C� erveny and V. Ruzicka, Adv. Catal., Chem. Soc., 1993, 115, 2046; (g)W.Lin, T. H.Warren,R.G.Nuzzo 1981, 30, 335; ( f) G. C. Bond, Surf. Sci., 1985, 156, 966; andG. S. Girolami, J. Am. Chem. Soc., 1993, 115, 11644; (h) Yu. I. ( g) M. Boudart and D. J. Sajkowski, Faraday Discuss., 1991, 92, Yermakov, B. N. Kuznetsov and V. A. Zakharov, Catalysis by 57; (h) M. Boudart and K. Tamaru, Catal. Lett., 1991, 9, 15; Supported Complexes, Elsevier, Amsterdam, 1981, ch. 2; A. (i) J. Barbier, F. Lamy-Pitara, P. Marecot, J. P. Boitiaux, Zecchina and C. Otero Area� n, Catal. Rev.—Sci. Eng., 1993, 35, 261; J. Cosyns and F. Verna, Adv. Catal., 1990, 37, 279; ( j) H. Arnold, W. Lin, B. C. Wiegand, R. G. Nuzzo and G. S. Girolami, J. Am. S. Go� bel, D. Reinig and J. Gaube, Chem. Ing. Tech., 1994, 66, 727. Chem. Soc., 1996, 118, 5977; W. Lin, R. G. Nuzzo and G. S. 19 (a) M. Boudart and D. J. Sajkowski, J. Chem. Soc., Faraday Girolami, J. Am. Chem. Soc., 1996, 118, 5988. Discuss., 1991, 92, 57; (b) M. Che and C. O. Bennett, Adv. Catal., 7 F. Calderazzo and F. A. Cotton, Inorg. Chem., 1962, 1, 30. 1989, 36, 55. 8 (a) J. F. Moulder, W. F. Stickle, P. E. Sobol and K. D. Bomben, 20 M. Boudart, Chem. Rev., 1995, 95, 661. Handbook of X-ray Photoelectron Spectroscopy, ed. J. Chastain, Perkin–Elmer Corporation, Eden Prairie, MN, 1994; (b) K. S. Kim, A. F. Gossmann, N. Winograd, Anal. Chem., 1974, Paper 8/04456B J. Mater. Chem., 1998, 8, 2855&nda