Hydroxyapatite-nanosphere supported ruthenium(0) nanoparticle catalyst for hydrogen generation from ammonia-borane solution: kinetic studies for nanoparticle formation and hydrogen evolution

Abstract

The development of readily prepared effective heterogeneous catalysts for hydrogen generation from ammonia-borane (AB; NH₃BH₃) solution under mild conditions still remains a challenge in the field of “hydrogen economy”. In this study, we report our finding of an in situ generated, highly active ruthenium nanocatalyst for the dehydrogenation of ammonia-borane in water at room temperature. The new catalyst system consists of ruthenium(0) nanoparticles supported on nanohydroxyapatite (RuNPs@nano-HAp), and can be reproducibly prepared under in situ conditions from the ammonia-borane reduction of Ru³⁺ ions exchanged into nanohydroxyapatite (Ru³⁺@nano-HAp) during the hydrolytic dehydrogenation of ammonia-borane at 25 ± 0.1 °C. Nanohydroxyapatite-supported ruthenium(0) nanoparticles were characterized by a combination of advanced analytical techniques. The sum of their results shows the formation of well-dispersed ruthenium(0) nanoparticles with a mean diameter of 2.6 ± 0.6 nm on the surface of the nanospheres of hydroxyapatite by keeping the host matrix intact. The resulting RuNPs@nano-HAp are highly active catalyst in the hydrolytic dehydrogenation of ammonia-borane with an initial TOF value of 205 min⁻¹ by generating 3.0 equiv. of H₂ per mole of ammonia-borane at 25 ± 0.1 °C. Moreover, they are sufficiently stable to be isolated and bottled as solid materials, which can be reused as active catalyst under the identical conditions of first run. The work reported here also includes the following results: (i) monitoring the formation kinetics of the in situ generated RuNPs@nano-HAp by hydrogen generation from the hydrolytic dehydrogenation of ammonia-borane as the reporter reaction. The sigmoidal kinetics of catalyst formation and concomitant dehydrogenation fits well to the two-step, slow nucleation, followed by autocatalytic surface growth mechanism, P → Q (rate constant k₁) and P + Q → 2Q (rate constant k₂), in which P is Ru³⁺@nano-HAp and Q is the growing, catalytically active RuNPs@nano-HAp; (ii) the compilation of kinetic data for the RuNPs@nano-HAp catalyzed hydrolytic dehydrogenation of ammonia-borane depending on the temperature and catalyst concentration to determine the dependency of reaction rate on catalyst concentration and activation parameters (E_a, ΔH^#, and ΔS^#) of the reaction.

---

Introduction

Nowadays, the development of safe and efficient hydrogen storage technologies is one of the most important and challenging concerns in hydrogen-based energy policies,^1,2 which would enable the changeover from fossil fuels to renewable energy sources.^3,4 For this reason, numerous studies have been performed for the development of materials with high volumetric and gravimetric storage capacity because the low density of hydrogen makes it difficult to store in compressed or liquefied form.³ In this context, various porous materials,^5–8 boron-based chemical hydrides^9–11 and boron–nitrogen compounds^12–16 have been tested for chemical hydrogen storage. Among these materials, ammonia-borane (AB; NH₃BH₃) has been found to be significantly better suited for this purpose because of the following advantages: (i) AB has high gravimetric hydrogen storage capacity (19.6 wt%); (ii) it has low molecular weight (30.7 g mol⁻¹); and (iii) it is non-flammable and non-explosive under standard conditions.¹⁷ These significant properties make AB unique compared to metal hydrides/B–N compounds or porous materials, where hydrogen release and uptake can be controlled by temperature and pressure. Although hydrogen can be released from AB by its thermal decomposition,¹⁸ dehydrocoupling,¹⁹ and alcoholysis,²⁰ there is significant interest in transition metal-catalyzed hydrolytic dehydrogenation¹² because of favorable kinetics and mild reaction conditions. The hydrolytic dehydrogenation of AB generates 3 equiv. of H₂ (per mole of AB) at room temperature (RT) under air only in the presence of a suitable catalyst (1).^12,17,21,22

Despite the difficulty in recycling the hydrolysis product metaborate anion, hydrogen generation from the hydrolytic dehydrogenation of ammonia-borane has several features, which make it promising for potential applications: (i) AB has high solubility in water (33.6 g per 100 g H₂O);²³ (ii) the reaction proceeds at an appreciable rate only in the presence of a suitable catalyst at room temperature; (iii) the hydrolytic dehydrogenation of AB is exothermic (ΔH = −155.97 kJ mol⁻¹).²⁴ Several transition metals or their compounds have been tested as catalysts in hydrogen generation from the hydrolysis of AB.²⁵ Of particular importance is the ruthenium-based heterogeneous catalyst showing high activity in the hydrolytic dehydrogenation of AB under mild conditions.^26–33 Unfortunately, most catalysts used in these schemes suffer from difficult isolation,^28,31,32 low activity^29,30,33 low stability,^31,32 and time-consuming synthesis procedures.^27,28,32 Therefore, the development of a readily accessible, highly active and reusable catalyst that operates under mild conditions remains a challenge in this field.

Recently, hydroxyapatite ([Ca₁₀(OH)₂(PO₄)₆], HAp)³⁴ has generated significant interest in view of its potential application as a catalyst support with attractive properties.^35–37 Particularly, HAp has the following advantages as a catalyst support: (i) reduced mass transfer limitations because of the absence of structural porosity, (ii) high ion-exchange and adsorption capacity, and (iii) low surface acidity that prevent side reactions from the support itself.⁶ These physico-chemical properties of HAp prompted us to focus on the use of HAp matrix for the stabilization of metal nanoparticles. We recently reported that ruthenium(0) nanoparticles supported on the surface of HAp is highly active, long-lived and reusable catalyst providing 87 000 turnovers with an initial turnover frequency of 137 min⁻¹ for hydrogen generation from the hydrolysis of AB at 25.0 ± 0.1 °C.³⁸ In addition, we have also shown that the reduction of the particle size of support materials from the micro-size to nano-size regime (from >1 μm to <100 nm) results in a tremendous increase in external surface area, numerous exchange sites and lower mass transfer limitations for catalytic applications.^37,39

Herein, we report the use of colloidal hydroxyapatite nanospheres with particle size smaller than 50 nm as support for the stabilization of ruthenium(0) nanoparticles. Ruthenium(0) nanoparticles supported on the surface of nano-HAp, hereafter referred to as RuNPs@nano-HAp, show remarkable catalytic performance in terms of activity and reusability in the hydrolytic dehydrogenation of AB at room temperature under air. RuNPs@nano-HAp can be reproducibly prepared by ion exchange of Ru³⁺ ions with Ca²⁺ ions of the nano-HAp matrix, followed by the in situ AB reduction of Ru³⁺ ions on the surface of colloidal HAp during the hydrolytic dehydrogenation of AB. After a short induction time, the resulting RuNPs@nano-HAp catalyzes the dehydrogenation of aqueous AB solution at room temperature with an initial turnover frequency (TOF) of 205 min⁻¹. More importantly, nano-HAp dispersed ruthenium(0) nanoparticles feature notable resistance against agglomeration and leaching throughout catalytic runs. When the isolated RuNPs@nano-HAp are reused, they retain 92% of their initial catalytic activity even at fifth reuse in the hydrolytic dehydrogenation of AB.

Experimental

Materials

Synthetic hydroxyapatite nanocrystalline powder ([Ca₅(OH)(PO₄)₃]_x, ≥ 99.995% trace metals basis), ruthenium(III) chloride trihydrate (RuCl₃·3H₂O), ammonia-borane (NH₃BH₃, 97%), D₂O and BF₃·(C₂H₅)₂O were purchased from Sigma-Aldrich. Deionized water was distilled with a water purification system (Thermo Scientific Barnsted Nanopure System). All the glassware and Teflon-coated magnetic stir bars were cleaned with acetone, followed by copious rinsing with doubly deionized water under ultrasonication, and finally dried in an oven at 423 K.

Characterization

XRD data for nano-HAp, Ru³⁺@nano-HAp and RuNPs@nano-HAp were collected using Rigaku X-ray Diffractometer (Model, Miniflex) with Cu-Kα (30 kV, 15 mA, = 1.54051 Å) radiation at room temperature. TEM samples were prepared by dropping one drop of dilute suspension on copper-coated carbon TEM grid, and the solvent was then evaporated at room temperature under reduced pressure (10⁻³ Torr). Conventional TEM was carried out on a JEOL JEM-200CX transmission electron microscope operating at 120 kV. Ruthenium content of RuNPs@nano-HAp was determined by inductively coupled plasma atomic emission spectroscopy (ICP-OES; ULTIMA 2-HORIBA Jobin–Yvon) after the powdered sample was completely dissolved in the mixture of HNO₃–HCl in a 1 : 3 ratio. Nitrogen adsorption–desorption experiments were carried out at 77 K using a NOVA 3000 series (Quantachrome Inst.) instrument. The sample was outgassed under vacuum at 393 K for 3 h before the adsorption of nitrogen. XPS analysis was performed on a Kratos AXIS ultra imaging X-ray photoelectron spectrometer using monochromatic Al Kα radiation (1486.6 eV, the X-ray tube working at 15 kV, 350 W and pass energy of 23.5 keV). At the end of the hydrolysis reaction, the resulting solutions were filtered and the filtrates were collected for ¹¹B NMR analysis. ¹¹B NMR spectra were recorded on a Bruker Avance DPX 400 with an operating frequency of 128.15 MHz. D₂O and BF₃·(C₂H₅)₂O were used as a lock and an external reference, respectively.

Preparation of ruthenium(III)-exchanged nanohydroxyapatite (Ru³⁺@nano-HAp) precatalyst and the general procedure for the in situ generation of ruthenium(0) nanoparticles supported on nanohydroxyapatite (RuNPs@nano-HAp)

Ruthenium(III) cations were introduced into Ca²⁺@nano-HAp by ion exchange of 200 mg of Ca²⁺@nano-HAp in 10 mL of an aqueous solution of 5.23 mg of RuCl₃·3H₂O (0.02 mmol) for 12 h at room temperature. The sample was then filtered by suction filtration using a Whatman-1 filter (Ø = 9 cm), washed three times with 100 mL of deionized water, and dried at 353 K in the oven. The in situ formation of RuNPs@nano-HAp and the concomitant hydrolytic dehydrogenation of AB were performed in a Fischer–Porter (F–P) pressure bottle connected to a line by Swagelock tetrafluoroethylene (TFE)-sealed quick connects and to an Omega-PX209-100GI pressure transducer interfaced through an Omega-UWPC-2-NEMA wireless transmitter to a computer using an Omega-UWTC-REC2-DV2 wireless receiver. The progress of individual dehydrogenation reactions was followed by monitoring the pressure of H₂ gas on Omega-data logging and using the recording software program. In a typical experiment, 100 mg of Ru³⁺@nano-HAp (4.16 μmol Ru) was weighed and placed in a new 22 × 175 mm Pyrex culture tube containing a new 5/16 in. × 5/8 in. stir bar. The culture tube was then placed inside the F–P bottle, and the entire set-up was placed inside a constant-temperature circulating water bath thermostated at 25 ± 0.1 °C. Next, 10.0 mL of 100 mM AB solution was rapidly added to the F–P bottle by a syringe with a long needle and the reaction timer was started (t = 0 min). When no more hydrogen generation was observed, the experiment was stopped, the F–P bottle was closed and disconnected from the line, and hydrogen pressure was released. Finally, a small aliquot from the reaction solution in the culture tube was withdrawn for ¹¹B NMR analysis.

Data handling and curve fit for hydrogen generation data

Raw pressure versus time data collected with the computer-interfaced transducer were exported from Omega software program and imported into OriginPro 8, which was then converted into the volume of hydrogen (mL). Curve fitting of equivalent H₂ generated per NH₃BH₃ versus time data to the Finke–Watzky two-step mechanism^40,41 was performed, as described elsewhere^40,41,42 using the software package OriginPro 8, which is a nonlinear regression subroutine and uses a modified Levenberg–Marquardt algorithm.⁴³

Isolability and reusability of RuNPs@nano-HAp in hydrolytic dehydrogenation of ammonia-borane

After the first run of hydrolytic dehydrogenation of 100 mM AB in 10 mL H₂O catalyzed by 100 mg RuNPs@nano-HAp (4.16 μmol Ru) at 25 ± 0.1 °C, the catalyst was isolated by suction filtration, washed three times with 20 mL of deionized water, dried under N₂ gas purging at room temperature, and then transferred into the glove box. The dried samples of RuNPs@nano-HAp were weighed and used again in the hydrolytic dehydrogenation of 100 mM AB in 10 mL H₂O, and the same procedure was repeated up to the fifth catalytic cycle. The results were expressed as the percentage of the retained initial catalytic activity of RuNPs@nano-HAp in the hydrolytic dehydrogenation of AB versus the number of catalytic runs.

Results and discussion

In situ formation of RuNPs@nano-HAp during the hydrolytic dehydrogenation of AB starting with Ru³⁺@nano-HAp precatalyst and the characterization of the resulting RuNPs@nano-HAp

Ruthenium(0) nanoparticles supported on nano-hydroxyapatite, referred to as RuNPs@nano-HAp, can be reproducibly prepared by the ion-exchange⁴⁴ of Ru³⁺ ions with Ca²⁺ ions of the HAp matrix followed by the AB reduction of the resulting Ru³⁺@nano-HAp precatalyst during the hydrolytic dehydrogenation of AB. The progress of nanoparticle formation and associated hydrolytic dehydrogenation of AB was followed by monitoring the changes in hydrogen pressure, which were then converted into the equivalent H₂ generated per mole of AB using the known 3 : 1 H₂/AB stoichiometry (1). Fig. 1 shows the plot of equivalent hydrogen generated per mole of AB versus time for the hydrolytic dehydrogenation of AB starting with Ru³⁺@nano-HAp precatalyst at 25 ± 0.1 °C. The formation kinetics of RuNPs@nano-HAp catalyst can be obtained by the hydrolysis of AB as reporter reaction,^40,45,46 (Scheme 1), in which A is the added precatalyst Ru³⁺@nano-HAp and B is the growing RuNPs@nano-HAp catalyst. The hydrolytic dehydrogenation of AB will report and intensify the amount of RuNPs@nano-HAp catalyst, Q, present if the hydrolysis rate is fast in comparison to the rate of nanoparticle formation.

Fig. 1 Plot of the generation of equivalent H₂ per NH₃BH₃ versus time (min) for in situ generated RuNPs@nano-HAp catalyzed hydrolytic dehydrogenation of AB starting with Ru³⁺@nano-HAp precatalyst ([AB] = 100 mM; [Ru] = 0.25 mM in 10.0 mL H₂O) at 25 ± 0.1 °C and its curve fit (red) to F–W two-step nucleation and autocatalytic growth mechanism⁴⁰ (k₁ = (3.27 ± 0.19) × 10⁻² min⁻¹, k₂ = (14.65 ± 0.47) M⁻¹ min⁻¹ and R² = 0.997) for ruthenium(0) nanoparticles formation.

Scheme 1 Minimalistic, two-step nanoparticle nucleation and autocatalytic surface growth mechanism⁴⁰ for in situ generated RuNPs@nano-HAp during the hydrolytic dehydrogenation of AB.

The sigmoidal kinetic can be seen in Fig. 1, which is an example of all the data collected under different conditions (vide infra), and fits well by the Finke–Watzky two-step nucleation and the autocatalytic growth mechanism of nanoparticle formation.⁴⁰ The observation of a sigmoidal dehydrogenation curve and its curve fit to a slow, continuous nucleation A →B (rate constant k₁) followed by autocatalytic surface growth A + B →2B (rate constant k₂) kinetics is indicative of the formation of a metal(0) nanoparticle catalyst from a precatalyst in the presence of a reducing agent.^40–42 Rate constants determined from the nonlinear least-squares curve fit from Fig. 1 are k₁ = 3.27 × 10⁻² min⁻¹ and k₂ = 14.65 M⁻¹ min⁻¹ (k₂ has been corrected to the stoichiometry factor of 400, as described elsewhere).⁴⁵

After generating 3 equiv. of H₂ along with the result of AB (δ = −23 ppm, q) conversion to metaborate (δ = 9 ppm, s) by ¹¹B NMR spectroscopy,⁴⁷ resulting RuNPs@nano-HAp were isolated by filtration and characterized by several analytical techniques. Fig. 2 depicts the XRD patterns of nano-HAp, Ru³⁺@nano-HAp and RuNPs@nano-HAp (with 0.51 wt% Ru loading as determined by ICP-OES) altogether, and the comparison of them clearly shows that the incorporation of ruthenium(III) ions into nano-HAp and the reduction of ruthenium(III) ions forming the ruthenium(0) nanoparticles on the surface of nano-HAp cause no detectable alteration in the framework lattice and no loss in the crystallinity of nano-HAp. The morphology and size of RuNPs@nano-HAp were investigated by transmission electron microscopy (TEM). TEM images of RuNPs@nano-HAp in different magnifications shown in Fig. 3 reveal the existence of ruthenium(0) nanoparticles with an average diameter of 2.6 ± 0.6 nm (Fig. 3d), which corresponds to Ru(0)_∼650 nanoclusters⁴⁸ on the surface of HAp-nanospheres of 60–80 nm size. N₂ adsorption–desorption isotherms of nano-HAp and RuNPs@nano-HAp (0.51 wt% Ru loading) are given in Fig. 4.

Fig. 2 Powder X-ray diffraction (P-XRD) patterns of (a) nano-HAp, (b) Ru³⁺@nano-HAp precatalyst, (c) RuNPs@nano-HAp.

Fig. 3 (a–c) TEM images at different magnifications, (d) corresponding size histogram of RuNPs@nano-HAp.

Fig. 4 Nitrogen-adsorption–desorption isotherms of nano-HAp (up) and RuNPs@nano-HAp (down).

Both the isotherms show Type III shape, reflecting the absence of micropores (<2 nm).⁴⁹ On passing from nano-HAp to RuNPs@nano-HAp, the surface area reduced from 40 to 28 m² g⁻¹. The decrease in surface area is associated with the presence of ruthenium nanoparticles on the nano-HAp surface. Furthermore, no hysteresis loop was observed in the N₂ adsorption–desorption isotherm of RuNPs@nano-HAp, indicating that the procedure followed for the preparation of RuNPs@nano-HAp does not create any mesopores within the framework of nano-HAp. The oxidation state of ruthenium in the RuNPs@nano-HAp sample was investigated by X-ray photoelectron spectroscopy (XPS). Fig. 5 shows the high resolution Ru 3d and 3p XPS spectra of RuNPs@nano-HAp, which shows three prominent peaks at 287.2, 284.7, and 465.9 eV, readily assigned to Ru(0) 3d_3/2, Ru(0) 3d_5/2, and Ru(0) 3p_3/2, respectively,⁵⁰ indicating the complete reduction of Ru³⁺ species in situ during the hydrolytic dehydrogenation of AB. Compared to the value of metallic ruthenium 3d and 3p peaks, the slight shift (ΔE_b = 0.4 eV–0.6 eV) observed in RuNPs@nano-HAp toward a higher energy value can be attributed to the peculiar electronic properties of the HAp matrix.³⁵

Fig. 5 High resolution (a) Ru 3d and (b) Ru 3p XPS spectra of RuNPs@nano-HAp.

Control experiments: (i) catalytic reactivity of host material (nano-HAp) in the hydrolytic dehydrogenation of AB and (ii) effect of ruthenium loading on the catalytic activity of RuNPs@nano-HAp in the hydrolytic dehydrogenation of AB

Before testing the catalytic activity of in situ generated RuNPs@nano-HAp in hydrogen generation from the hydrolysis of ammonia-borane, one needs to investigate whether sole nano-HAp can catalyze the same reaction under identical conditions. In the control experiments starting with sole nano-HAp, no hydrogen generation was observed during the hydrolysis of AB at 20, 25, 30, 35 and 40 °C at the same time intervals as the ones used for the RuNPs@nano-HAp catalyzed hydrolytic dehydrogenation of AB (vide infra). Next, the hydrolytic dehydrogenation of AB was performed starting with Ru³⁺@nano-HAp precatalyst with different ruthenium loadings in the range of 0.32–4.20 wt% to determine the effect of ruthenium loading on the catalytic activity of RuNPs@nano-HAp. Fig. 6a shows the plots of equivalent hydrogen generated per mole of AB versus time for the hydrolysis of AB starting with Ru³⁺@nano-HAp precatalyst at 25 ± 0.1 °C with different metal loading.

Fig. 6 (a) Plot of the generation of equivalent H₂ per NH₃BH₃ versus time (min) for in situ generated RuNPs@nano-HAp catalyzed hydrolytic dehydrogenation of AB starting with Ru³⁺@nano-HAp precatalysts with Ru loadings of 0.32, 0.51, 1.17, 2.62 and 4.20 wt% (in all [AB] = 100 mM and 100 mg Ru³⁺@nano-HAp precatalyst in 10.0 mL H₂O) at 25 ± 0.1 °C. (b) Plot of initial TOF values versus Ru loadings (wt%).

Variation in catalytic activity with ruthenium loading given in Fig. 6 reflects the relative number of accessible ruthenium(0) atoms on the surface of nanoparticles supported on the nano-HAp spheres. The fraction of accessible ruthenium(0) atoms depends on the interaction between the host surface and nanoparticles along with the size of nanoparticles. The highest catalytic activity is obtained by Ru³⁺@nano-HAp precatalyst containing 0.51 wt% Ru. An informative TEM image obtained from RuNPs@nano-HAp sample with 4.20 wt% Ru loading indicates the existence of larger ruthenium(0) nanoparticles (3.6 ± 0.9 nm) than that of the sample with 0.51 wt% Ru loading.

Determination of activation parameters (E_a, ΔH^#, ΔS^#) and rate dependency on [Ru] for the hydrolysis of AB catalyzed by the in situ generated RuNPs@nano-HAp

We determined activation energy and activation parameters (E_a, ΔH^# and ΔS^#) for the hydrolytic dehydrogenation of AB catalyzed by the in situ generated RuNPs@nano-HAp from the temperature-dependent kinetic data as shown in Fig. 7. Observed rate constants were calculated from the nearly linear portions of the curves at different temperatures (Table 1) and used for the construction of Arrhenius and Eyring–Polonyi plots given in Fig. 8a and b, respectively; activation energy, E_a = 55 ± 2 kJ mol⁻¹; activation enthalpy ΔH^# = 51 ± 2 kJ mol⁻¹ and activation entropy ΔS^# = −51 ± 5 J mol⁻¹ K⁻¹ for the hydrolysis of AB catalyzed by the in situ generated RuNPs@nano-HAp. The activation energy value is comparable with the majority of previously employed Ru-based catalysts (vide infra). The small value of activation enthalpy and the large negative value of activation entropy imply an associative mechanism in the transition state for the hydrolytic dehydrogenation of AB catalyzed by the in situ formed RuNPs@nano-HAp. The rate constants k₁ and k₂ obtained from the curve fit of the data to the two-step mechanism and the k₂/k₁ ratios are given in Table 1 together with the induction time and hydrogen generation rates obtained from the nearly linear portions of the curves for the hydrolysis of AB at different temperatures. While induction period decreases, the k₁ and k₂ rate constants increase with increasing temperature, as expected. The inverse relationship between induction time and rate constant k₁, which has been known for a long time,⁴⁰ is also observed in our case. Another important point that can be concluded from the inspection of data listed in Table 1 is the large values of the k₂/k₁ ratio, which implies the high level of kinetic control on the formation of ruthenium(0) nanoparticles supported on nano-HAp. The kinetics of the hydrolytic dehydrogenation of AB catalyzed by RuNPs@nano-HAp was also studied depending on the catalyst concentration to establish the rate law for catalytic transformation. Fig. 9a shows the plots of equivalent H₂ generated per mole of AB versus time during the hydrolytic dehydrogenation of AB starting with Ru³⁺@nano-HAp precatalyst (0.51 wt% Ru) in different ruthenium concentrations at 25.0 ± 0.1 °C. A fast dehydrogenation starts after an induction time of 4.0–0.3 min.

Fig. 7 Plot of the generation of equivalent H₂ per NH₃BH₃ versus time (min) and its curve fit (red) to the F–W two-step nucleation and autocatalytic growth mechanism⁴⁰ for the in situ generated RuNPs@nano-HAp catalyzed hydrolytic dehydrogenation of AB starting with Ru³⁺@nano-Hap (0.51 wt% Ru) precatalyst at different temperatures (in all [AB] = 100 mM; [Ru] = 0.25 mM in 10.0 mL H₂O).

Table 1 k₁, k₂, t (induction time), k₂/k₁ and k_obs (dehydrogenation rate) of in situ generated RuNPs@nano-HAp catalyzed hydrolytic dehydrogenation of AB ([AB] = 100 mM; [Ru] = 0.25 mM in 10.0 mL H₂O) depending on temperature

Entry	T (°C)	k1 (min^−1) × 10^2	k2 (M min)^−1	k2/k1 (M^−1) × 10^2	t (min)	kobs (mM min^−1)
1	20	1.75 ± 0.09	44.77 ± 1.09	25.58 ± 1.95	1.50	11.06
2	25	3.27 ± 0.19	59.21 ± 1.89	18.10 ± 1.62	0.83	15.88
3	30	5.35 ± 0.35	94.61 ± 3.47	17.68 ± 1.81	0.58	25.77
4	35	9.29 ± 0.66	131.59 ± 5.73	14.17 ± 1.62	0.33	35.04
5	40	17.41 ± 1.31	157.73 ± 9.01	9.06 ± 1.19	0.17	44.87

---

Fig. 8 (a) Arrhenius plot (y = 24.948 − 6599x and R² = 0.988) and (b) Eyring–Polanyi plot (y = 17.665 − 6122x and R² = 0.981) for the in situ generated RuNPs@nano-HAp catalyzed hydrolytic dehydrogenation of AB starting with Ru³⁺@nano-HAp precatalyst.

Fig. 9 (a) Plot of generation of equivalent H₂ per NH₃BH₃ versus time (min) and its curve fit (red) to F–W two-step nucleation and autocatalytic growth mechanism⁴⁰ for hydrolytic dehydrogenation of AB starting with Ru³⁺@nano-HAp precatalyst (in all [AB] = 100 mM in 10.0 mL H₂O) at different ruthenium concentrations as given on graph at 25 ± 0.1 °C, (b) ln(k_obs) versus ln([Ru]) graph (y = 3.748 − 0.919x and R² = 0.998).

Dehydrogenation rate, determined from the nearly linear portion of the plots, increases with catalyst concentration (Table 2). Plotting hydrogen generation rate versus ruthenium concentration, both on logarithmic scales, gives a straight line with a slope of 0.92 (Fig. 9b). Thus, an apparent first-order dependence on catalyst concentration is observed. As seen from Fig. 9a, experimental data for all ruthenium concentrations fit well to the Finke–Watzky two-step nanoparticle formation mechanism,⁴⁰ which provides the rate constants k₁ of the slow, continuous nucleation, P → Q, and k₂ of the autocatalytic surface growth, P + Q → 2Q (Table 2). Expectedly, nucleation rate constant k₁ for nucleation increases with increasing [AB]/[Ru] ratio, and rate constant k₂ for surface growth decreases with increasing ruthenium concentration. More importantly, k₂/k₁ ratio decreases with increasing ruthenium concentration, indicating that nanoparticle formation becomes less kinetically controlled in more concentrated solution.

Table 2 k₁, k₂, k₂/k₁ and k_obs (dehydrogenation rate) of the in situ generated RuNPs@nano-HAp catalyzed hydrolytic dehydrogenation of AB at 25 ± 0.1 °C depending on [Ru] concentrations

Entry	[AB] (mM)	[Ru] (mM)	k1 (min^−1) × 10^2	k2 (M min)^−1	k2/k1 (M^−1) × 10^−2	kobs (mM min^−1)
1	100	0.063	2.02 ± 0.06	65.28 ± 1.53	32.30 ± 1.60	4.60
2	100	0.125	2.88 ± 0.12	71.12 ± 1.90	24.70 ± 1.70	9.72
3	100	0.250	3.27 ± 0.19	59.21 ± 1.89	18.10 ± 1.62	15.88
4	100	0.500	3.48 ± 0.32	55.76 ± 2.21	16.02 ± 2.11	25.22
5	100	1.000	3.59 ± 0.46	46.70 ± 2.16	13.01 ± 0.23	39.72

---

In addition to activity, the reusability of RuNPs@nano-HAp, as another crucial measure in heterogeneous catalysis, was also examined in the hydrolysis of AB. When an isolated and dried sample of RuNPs@nano-HAp catalyst is reused in AB hydrolysis, RuNPs@nano-HAp still acts as an active catalyst; they almost retain their inherent catalytic activity (>94%) even at complete conversion (≥99%).

Effect of the size of host HAp matrix on the catalytic activity of guest ruthenium(0) nanoparticles “activity comparison of RuNPs@nano-HAp with RuNPs@micro-HAp”

Recently, we reported³⁸ the activity of ruthenium(0) nanoparticles supported on micro-sized HAp (≥1 μm) in hydrogen generation from the hydrolysis of AB at 25.0 ± 0.1 °C. Because both the catalysts were prepared in the same manner by using the same support but in different particle sizes, we can compare the size effect of support materials on the catalytic activity of guest ruthenium(0) nanoparticles. Ruthenium(0) nanoparticles supported on nano-HAp show an activity of TOF = 205 min⁻¹ in hydrogen generation from the hydrolysis of AB at 25.0 ± 0.1 °C, while ruthenium(0) nanoparticles supported on micro-HAp show an activity of TOF = 137 min⁻¹ in the same reaction. This notable difference in the activity of ruthenium(0) nanoparticles supported on nano-HAp and microo-HAp can be attributed to the larger surface area of nano-HAp spheres than the micro-HAp particles and to the smaller size of ruthenium(0) nanoparticles supported on nano-HAp (2.56 ± 0.61 nm) than that of the nanoparticles supported on the micro-HAp particles (4.70 ± 0.70 nm). Both these phenomena are the consequences of the reduction of particle size of HAp matrix from micro-size to nano-size regime (from >1 μm to <100 nm).

Conclusions

In summary, our study on the preparation and characterization of ruthenium(0) nanoparticles supports on the surface of hydroxyapatite nanospheres as well as their catalytic application in hydrogen generation from the hydrolysis of ammonia-borane along with the detailed kinetics of both nanoparticle formation and hydrogen evolution has led to the following conclusions and insights:

(i) For the first time, nanohydroxyapatite supported ruthenium(0) nanoparticles, RuNPs@nano-HAp, were reproducibly prepared from the in situ reduction Ru³⁺@nano-HAp during the hydrolytic dehydrogenation of ammonia-borane at room temperature;

(ii) The characterization of the resulting novel catalytic material using ICP-OES, XRD, XPS, TEM and N₂-adsorption–desorption techniques reveals the formation of well-dispersed Ru(0)_∼650 nanoparticles (2.56 ± 0.61 nm) on the surface of hydroxyapatite nanospheres by keeping the host matrix intact;

(iii) Ruthenium(0) nanoparticles supported on nano-hydroxyapatite show notable catalytic activity (lower-bound TOF_inital = 205 min⁻¹) among all the heterogeneous ruthenium catalysts tested in the hydrolytic dehydrogenation of AB at room temperature;

(iv) Ruthenium(0) nanoparticles show high stability against the sintering and leaching, which makes them highly reusable catalyst; when redispersed they almost retain their initial activity even at the fifth catalytic run in the hydrolytic dehydrogenation of ammonia-borane, releasing 3 equiv. of H₂ per mole of ammonia-borane;

(v) Quantitative kinetic studies depending on catalyst concentration and temperature reveal that the hydrolytic dehydrogenation of ammonia-borane catalyzed by ruthenium(0) nanoparticles is first-order for catalyst concentration. The activation parameters were also determined for the hydrolytic dehydrogenation of ammonia-borane catalyzed by ruthenium(0) nanoparticles supported on nano-hydroxyapatite. Small values of activation energy and enthalpy (E_a = 55 kJ mol⁻¹, ΔH^# = 51 kJ mol⁻¹) and negative value of activation entropy (ΔS^# = −51 J mol⁻¹ K⁻¹) are indicative of an associative mechanism in the transition state for the catalytic hydrolysis of ammonia-borane;

(vi) All the kinetic data, collected for the formation of ruthenium(0) nanoparticles and concomitant hydrolytic dehydrogenation of ammonia-borane under various experimental conditions, fit well to the two-step mechanism for the nanoparticle formation:⁴⁰ continuous nucleation A → B (rate constant k₁) followed by autocatalytic surface growth A + B →2B (rate constant k₂). Large values of the k₂/k₁ ratio obtained at different conditions are indicative of the high level of kinetic control in the formation of ruthenium(0) nanoparticles from the reduction of ruthenium(III) ions exchanged to the nano-hydroxyapatite;

(vii) The reporter reaction method developed by Finke et al.⁵¹ for the catalytic hydrogenation of olefins and aromatics was shown to work in the case of ruthenium(0) nanoparticle formation from the reduction of ruthenium(III) ions exchanged to nano-hydroxyapatite during the hydrolytic dehydrogenation of ammonia-borane. Monitoring hydrogen generation from the catalytic hydrolysis of ammonia-borane provides an indirect route to follow the nucleation and autocatalytic surface growth of metal(0) nanoparticles from the reduction of a precatalyst in an aqueous medium.

Acknowledgements

Partial support by the Turkish Academy of Sciences is gratefully acknowledged. MZ thanks FABED for their partial support to his research.

Notes and references

1. US Department of Energy (DOE), Basic Research Needs For the Hydrogen Economy, Report of the Basic Energy Sciences Workshop on Hydrogen Production, Storage and Use, http://science.energy.gov/%7E/media/bes/pdf/reports/files/nhe_rpt.pdf.
2. Lighting the Way Towards a Sustainable Energy Futures, Inter Academy Council (I.A.C.) Report Amsterdam, 2007.
3. J. Turner, G. Sverdrup, K. Mann, P. G. Maness, B. Kroposki, M. Ghirardi, R. J. Evans and D. Blake, Int. J. Hydrogen Energy, 2008, 32, 379.
4. Annual Energy Outlook 2005 With Projections To 2025, Energy Information Administration, February, 2005, http://www.eia.doe.gov/oiaf/aeo/pdf/0383(2005).pdf.
5. S. H. Lim, J. Luo, Z. Zhong, W. Ji and J. Lin, Inorg. Chem., 2005, 44, 4124.
6. J. Weitkamp, M. Fritz and S. Ernst, Int. J. Hydrogen Energy, 1995, 20, 967.
7. N. B. McKeown and P. M. Budd, Chem. Soc. Rev., 2006, 35, 675.
8. N. L. Rossi, J. Eckert, M. Eddaoudi, D. T. Vodak, J. Kim, M. O'Keefe and O. M. Yaghi, Science, 2003, 300, 1127.
9. M. Au and A. Jurgensen, J. Phys. Chem. B, 2006, 110, 7062.
10. M. Zahmakiran and S. Özkar, Langmuir, 2009, 25, 2667.
11. E. Rönnebro and E. H. Majzoub, J. Phys. Chem. B, 2007, 111, 12045.
12. H. L. Jiang, S. K. Singh, J. M. Yan, X. B. Zhang and Q. Xu, ChemSusChem, 2010, 3, 541.
13. A. Staubitz, A. P. M. Robertson and I. Manners, Chem. Rev., 2010, 110, 4079.
14. C. W. Yoon and L. G. Sneddon, J. Am. Chem. Soc., 2006, 128, 13992.
15. X. Huang, X. Chen, T. Yisgedu, E. A. Meyers, S. G. Shore and J. C. Zhao, Inorg. Chem., 2011, 50, 3738.
16. S. Karahan, M. Zahmakiran and S. Özkar, Int. J. Hydrogen Energy, 2011, 36, 4958.
17. C. W. Hamilton, R. T. Baker, A. Staubitz and I. Manners, Chem. Soc. Rev., 2009, 38, 279.
18. G. Wolf, J. Baumann, F. Baitalow and F. P. Hoffmann, Thermochim. Acta, 2000, 343, 19.
19. C. A. Jaska, K. Temple, A. J. Lough and I. Manners, J. Am. Chem. Soc., 2003, 125, 9424.
20. S. Çalışkan, M. Zahmakiran and S. Özkar, Appl. Catal., B, 2010, 93, 387.
21. H. L. Jiang and Q. Xu, Catal. Today, 2011, 170, 56.
22. M. Zahmakiran, T. Ayvali, S. Akbayrak, S. Çalışkan, D. Çelik and S. Özkar, Catal. Today, 2011, 170, 76.
23. F. H. Stephens, V. Pons and R. T. Baker, Dalton Trans., 2007, 2613.
24. M. E. Bluhm, M. G. Bradley, R. Butterick, U. Kusari and L. G. Sneddon, J. Am. Chem. Soc., 2006, 128, 7748.
25. A Web of Science literature search showed that in total >120 studies exist according to the terms “catalytic hydrolysis of ammonia-borane”. Please see recent excellent reviews given in ref. 12, 17 and 21, which briefly survey these catalysts systems.
26. H. Y. Liang, G. Z. Chen, Z. Desinan, R. Rosei, F. Roseiand and D. L. Ma, Int. J. Hydrogen Energy, 2012, 37, 1792.
27. S. Akbayrak and S. Özkar, ACS Appl. Mater. Interfaces, 2012, 4, 6302.
28. H. Can and Ö. Metin, Appl. Catal., B, 2012, 125, 304.
29. M. Zahmakiran, Mater. Sci. Eng., B, 2012, 177, 606.
30. H. B. Dai, X. D. Kang and W. Ping, Int. J. Hydrogen Energy, 2010, 35, 10317.
31. F. Durap, M. Zahmakiran and S. Özkar, Int. J. Hydrogen Energy, 2009, 34, 7223.
32. Ö. Metin, Ş. Şahin and S. Özkar, Int. J. Hydrogen Energy, 2009, 34, 6304.
33. S. Basu, A. Brockman, P. Gagare, Y. Zheng, P. V. Ramachandran, W. N. Delgass and J. P. Gore, J. Power Sources, 2009, 188, 238.
34. J. C. Elliot, Structure and Chemistry of Apatites and Other Calcium Orthophosphates, Elsevier, Amsterdam, 1994.
35. K. Kaneda and T. Mizugaki, Energy Environ. Sci., 2009, 2, 655.
36. M. Zahmakiran, Y. Tonbul and S. Özkar, Chem. Commun., 2010, 46, 4788.
37. M. Zahmakiran, Y. Román-Leshkov and Y. Zhang, Langmuir, 2012, 28, 60.
38. S. Akbayrak, P. Erdek and S. Özkar, Appl. Catal., B, 2013, 142–143, 187.
39. Y. Tonbul, M. Zahmakiran and S. Özkar, J. Am. Chem. Soc., 2010, 132, 6541.
40. M. A. Watzky and R. G. Finke, J. Am. Chem. Soc., 1997, 119, 10382.
41. M. Zahmakiran and S. Özkar, Inorg. Chem., 2009, 48, 8955.
42. S. Çalışkan, M. Zahmakiran, F. Durap and S. Özkar, Dalton Trans., 2012, 41, 4976.
43. W. H. Press, B. P. Flannery, S. A. Teukolsky and W. T. Vetterling, Numerical Recipies, Cambridge University Press, Cambridge, UK, 1989.
44. D. W. Breck, Zeolite Molecular Sieves, Krieger, Malabal, FL, 1984.
45. J. A. Widegren, J. D. Aiken, S. Özkar and R. G. Finke, Chem. Mater., 2001, 13, 312.
46. J. A. Widegren, M. A. Bennett and R. G. Finke, J. Am. Chem. Soc., 2003, 125, 10301.
47. P. V. Ramachandran and P. D. Gagare, Inorg. Chem., 2007, 46, 7810.
48. Using the equation N = N_opV/101.07, where N_o = 6.022 × 1023, p = 12.45 g cm⁻³ and V = (4/3)πr³, the number of metal atoms in spherical 2.56 nm ruthenium(0) nanoclusters is estimated to be 147.
49. S. Storck, H. Bretinger and W. F. Maier, Appl. Catal., A, 1998, 174, 137.
50. C. Wagner, W. M. Riggs, L. E. Davis, J. F. Moulder and G. E. Muilenberg, Handbook of X-ray Photoelectron Spectroscopy, Perkin-Elmer, 1979.
51. Y. Lin and R. G. Finke, J. Am. Chem. Soc., 1994, 116, 8335.

---