A Novel Method for Generating H₂ by Activation of the μAl-Water System Using Aluminum Nanoparticles

Featured Application

Aluminum nanoparticles are integrated with micron-scale particles to enhance hydrolytic sensitivity to water and accelerate hydrogen gas production.

Abstract

A method is described for activation of the reaction of room temperature water with micron-scale aluminum particles (μAl) by the addition of poly(epoxyhexane)-capped aluminum nanoparticles (Al NPs). By themselves, Al NPs react vigorously and completely with water at ambient temperatures to produce H₂. While pure μAl particles are unreactive toward water, mixtures of the μAl particles comprising 10 to 90% (by mass) of Al NPs, demonstrated appreciable hydrolytic activation. This activation is attributed to the reaction of the Al NPs present with water to produce a basic solution. Speciation modelling, pH studies, and powder X-ray diffraction analysis of the hydrolysis product confirm that the pH change is the key driver for the activation of μAl rather than residual heat from the exothermicity of Al NP hydrolysis. A mechanism is proposed by which the nonreactive aluminum oxide layer of the μAl is eroded under basic conditions. Mixtures 10% by mass of Al NPs can be used to produce the optimal quantity of H₂.

1. Introduction

Aluminum, Al, is the most abundant metal in the earth’s crust. When Al reacts with water, it produces H₂ gas and aluminum hydroxide, Al(OH)₃ [1]. The H₂ produced can be harnessed as a clean, ecologically friendly energy source, making it a desirable fuel for many applications [2,3,4,5,6]. The development of methods for H₂ production is motivated by a vision to move away from fossil fuels due to their finite availability and the environmental damage caused by their combustion [7,8,9,10]. There are many ways to generate H₂ including water electrolysis [11], chemical reactions [12], steam reforming of hydrocarbons [13], and from biological sources [14]. These methods are not amenable to portable-use systems, as they are either too expensive or initiated under extreme conditions. Transporting pure H₂ has challenges with its low volumetric energy density and high flammability. Thus, it is necessary to develop safe and economically feasible H₂ delivery systems that can be utilized on-demand. Aluminum is deemed as a possible source of energy storage for direct H₂ generation as long as its surface is passivated efficiently, effectively and safely [15].

Pure aluminum spontaneously reacts with water due to its high reduction potential (E° = −1.66 V) but can easily form a thin oxide layer when in contact with air, which leads to remarkably low reactivity [16]. This thin oxide layer develops into a 75 to 100 nm thick layer in a few months. It is also insoluble in, and impermeable to, water and prevents contact between the metal and water, hindering access to the energy-rich aluminum metal core [17]. Thus, μAl powders remain almost unhydrolyzed due to the thick oxide layer formation [18,19]. Because of this, it is important to find a method to activate the surface of μAl to allow access to the energetically rich and reactive metallic core. This problem can be solved by mixing organic capped Al nanoparticles (Al NPs) with μAl. Strategies for the activation of aluminum and aluminum alloys also have importance in additive manufacturing processes [20].

When the problems related to surface oxidation and agglomeration are overcome, Al NPs demonstrate fast reaction rates, exceptional combustion efficiency, and higher volumetric energy densities in comparison to conventional materials [21,22]. Due to their high surface-to-volume ratio, nanoparticles can have exceptionally strong interactions with other components in a nanocomposite. Thus, nanoaluminum surfaces may be stabilized by passivation and surface coated with organic polymers or polymerizable substrates to control their size and safety in energetics applications [23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40].

Our lab has previously reported that epoxyalkane-capped Al NPs demonstrated moderate air stability, as well as high active aluminum content [41,42,43]. Of these, epoxyhexane-capped Al NPs have the least protective cap owing to the shorter alkyl side-chain and react rapidly and completely with water to generate H₂ at room temperature, and produce basic conditions that can be exploited for the benefit of μAl hydrolytic activation. Thus the Al NPs could work as an indirect promoter by generating Al(OH)₃ to facilitate H₂ production from μAl at room temperature. Wang et al. have previously reported that Al(OH)₃ could catalytically promote H₂ generation from the μAl-water system and was more effective compared with other promoters [44,45,46,47].

Several composite samples of μAl and Al NPs were prepared with a variety of ratios by mass and analyzed for their hydrogen generation efficiency upon hydrolysis. pH measurements and pH-dependent speciation modelling were used to determine whether to attribute the μAl activation to the exothermicity of the Al NPs’ hydrolysis or the resulting pH change. The optimal composition was determined based on the notion that the most useful, cost-effective energetic material will generate the maximal quantity of hydrogen from the minimal quantity of Al NPs in the mixture.

2. Materials and Methods

2.1. Reagents and Materials

Titanium (IV) isopropoxide (99.999% trace metals basis), N,N-dimethylethylamine alane solution (0.5 M in toluene), 1,2-epoxyhexane (97%), disodium ethylenediaminetetraacetic acid dihydrate (Na₂EDTA·2H₂O, 99%), zinc sulfate heptahydrate (ZnSO₄·7H₂O, 99.999% trace metals basis), and xylenol orange were purchased from Sigma-Aldrich (Milwaukee, WI, USA). Sodium chloride was purchased from Fisher Scientific (Waltham, MA, USA) and used to make saturated aqueous solutions. Micron-scale aluminum (μAl, 30 μm average diameter particles) was purchased from Sigma-Aldrich and used without further purification. Toluene was distilled under inert argon over Na or 4A molecular sieves to remove trace water and oxygen. The nitric acid and sodium hydroxide pellets used in the gas titration apparatus were purchased from Sigma-Aldrich.

2.2. Synthesis of Al NPs

The following is a slight modification of the previously described synthesis of epoxyhexane-capped aluminum nanoparticles [43]. All reactions were carried out on a Schlenk line under high purity nitrogen or argon. For each reaction, 6.5 mL (3.25 mmol) of 0.5 M N,N-dimethylethylamine alane solution in toluene was syringed into a Schlenk flask containing a further 10 mL of toluene. The flask was fitted with a reflux condenser and the solution was heated to 85 °C. Another solution was prepared by mixing 99 μL of titanium (IV) isopropoxide with 10 mL of toluene and then 0.8 mL of this solution was syringed into the stirred, heated alane solution, leading to a change from colorless to black. The 1,2-epoxyhexane capping agent (40 μL, 10:1 Al:epoxide molar ratio) was immediately added directly to the reaction mixture. After vigorous stirring of the reaction mixture at 85 °C for 30 min, the solvent was removed under reduced pressure and the solid residue dried in vacuo.

2.3. Characterization of Al NPs

Powder X-ray diffraction (PXRD) measurements were performed using a Rigaku Miniflex 600 diffractometer. The Rigaku had a Cu K(α) source and a scintillation counter detector. The ICDD Crystallographic Database was used to compare PXRD patterns. The instrument software used Scherrer analysis to determine nanocrystallite dimensions.

Images of nanoparticle morphology and dimensions were obtained from a JEOL 1200 EX TEM instrument, which operated at 80 kV and had 200,000× magnification.

The DSC/TGA instrument used to obtain thermal profiles was a TA Instruments SDT Q600 simultaneous DSC/TGA instrument. The instrument had a constant flow rate of 50 mL/min of N₂, temperature range (25–800) °C, as well as a 10 °C/min temperature ramp. The samples were held in alumina cups.

The active and total aluminum contents of the Al NPs were determined by previously used methods of quantitative hydrolytic H₂ evaluation and complexometric Zn²⁺−EDTA back-titration, respectively [48,49,50].

2.4. Hydrolysis Activation Experiments

Five samples composed of various mass ratios of μAl with Al NPs were prepared. Each sample weighed a total of 50 mg precisely, with Al NPs mass % values of 10, 25, 50, 75, and 90%, the remaining mass being the μAl particles. A sixth sample of pure μAl without added Al NPs was also studied. The solid fuels were placed in an inverted-Y-shaped reaction vessel connected by flexible rubber tubing to a burette filled with saturated NaCl (aq) (in order to minimize H₂ dissolution), which was in turn connected via flexible plastic tubing to a glass funnel. The set-up is depicted in Figure 1. In the Y-tube side-arm was placed exactly 25.0 mL of deionized water. A timer was started upon tipping the water onto the Al solid sample and the H₂ (g) volume collected measured in the burette. By appropriate leveling of the water in the funnel, the H₂ pressure was recorded as the measured ambient atmospheric pressure at that time. Gas volume measurement was carried out until there appeared to be little visually discernible change in volume from one minute to the next (15 min for all samples). Samples were then left to sit in water for 3 days to allow any residual Al to hydrolyze. White solid residues from these samples were isolated by evaporating the water in an oven at 120 °C and then analyzed by PXRD.

3. Results and Discussion

3.1. Synthesis and Characterization of Al NPs

The Al NPs used in this study were synthesized by a slight modification of a previously described method [43]. The products were characterized by PXRD, which confirmed the presence of the crystalline Al cores (fcc Al, ICDD # 00-004-0787) with an average nanocrystallite diameter of 12 nm. Some polycrystallinity was observed in TEM analysis, with an average size of 30 nm. DSC/TGA also confirmed the core-shell nature of the nanoparticles, where an organic cap is removed at temperatures up to 460 °C, while aluminum oxidation onset takes place at 565 °C. All-in-all, these diagnostic characterizations (Supplementary Materials) closely matched those previously reported. Hydrolytic H₂ formation and complexometric Zn²⁺−EDTA back-titration also confirmed an active aluminum content (Al⁰ as a fraction of all aluminum present) of 83 ± 5% in the samples used for the μAl hydrolysis activation study.

3.2. Hydrolysis Activation Study

In the absence of any catalytic promoter, μAl particles hydrolyze so slowly as to be unobservable. Al NPs, on the other hand, react rapidly and completely with water to form Al(OH)₃ (s) (Equation (1)). We hypothesize that the Al(OH)₃ produced in this more rapid reaction then promotes the observable acceleration of the μAl reaction with water (Equation (2)). Alternatively, the exothermicity of Al NP hydrolysis in Equation (1) could be the driver for the pursuant reaction in Equation (2); the following will discern which of these is the more likely.

$$2\mathrm{Al}\mathrm{NPs}\left(s\right)+6{\mathrm{H}}_2\mathrm{O}\left(l\right)\to 2\mathrm{Al}{\left(\mathrm{OH}\right)}_3\left(s\right)+3{\mathrm{H}}_2\left(g\right)$$

(1)

$$2\mathsf{\mu }\mathrm{Al}\left(s\right)+6{\mathrm{H}}_2\mathrm{O}\left(l\right)\stackrel{\mathrm{Catalytic}\mathrm{Al}{\left(\mathrm{OH}\right)}_3}{\to }2\mathrm{Al}{\left(\mathrm{OH}\right)}_3 \left(s\right)+3{\mathrm{H}}_2\left(g\right)$$

(2)

Experiments were thus designed where μAl and Al NPs were combined with various mass ratios, from 10 to 90% Al NPs, and then their hydrolysis reactions were studied using a gas burette assembly shown in Figure 1. Any measured volume of hydrogen produced, V_mix, at atmospheric pressure, P, was deemed to have originated only from active Al⁰, either from the Al NPs (primary source) or from the μAl (secondary source) and was therefore the sum of V₁ and V₂, respectively (Equation (3)). To calculate V₂, the reaction to produce H₂ (with volume V₁) by the hydrolysis of the same quantity of Al NPs without the μAl present was also carried out. Volume V₂ was then used to determine the number of moles, n₂, of Al⁰ reacting from the μAl only, according to the ideal gas law (Equation (4)). Finally, the ‘activated μAl’ was quantified as the stochiometric equivalent mole amount of μAl⁰ that reacted to produce the n₂ moles of H₂ relative to the total quantity of μAl used in the experiment, n_μAl (Equation (5)) and expressed as a percentage. Allowing for oxide passivation, the latter quantity was adjusted by coefficient f, the mass fraction of active Al⁰ ($ℳ$_Al g/mol) in the mass of μAl used, m_μAl, which was calculated to be 0.99 based on a spherical core volume fraction for 30 μm particles with a 100 nm thick oxide shell. The ‘activated μAl’ was plotted versus the mass % of Al NPs used in the mixed μAl/Al NPs samples and is shown in Figure 2.

$$V_{mix}=V_1+V_2$$

(3)

$$n_2=\frac{P{V}_2}{RT}=\frac{P\left(V_{mix}-V_1\right)}{RT}$$

(4)

$$\mathrm{Activated} \mathsf{\mu }\mathrm{Al}=\frac{\frac{3}{2}{n}_2}{n_{\mathsf{\mu }\mathrm{Al}}}\times 100\%=\frac{\frac{3}{2}P\left(V_{mix}-V_1\right)\times {ℳ}_{Al}}{RT\times f{m}_{\mathsf{\mu }\mathrm{Al}}}\times 100\%$$

(5)

The inset bar chart of Figure 2 shows that in the absence of Al NPs, there was no activation toward hydrolysis with less than 1% of the μAl reacting. Integration with just 10% Al NPs by mass incurred substantial activation with ca. 52% of the μAl consumed by the time H₂ generation appeared to have visually ceased. As the proportion of Al NPs was raised, a steady increase in μAl activation was observed, reaching ca. 59% corresponding to 90% Al NPs by mass. As the main plot in Figure 2 shows, there is a simple linear relationship between these two quantities between 10 and 90% Al NPs by mass. The activation range thus spans only 7% and given the need to maximize H₂ production with the minimum incorporation of Al NPs, our results suggest that a composite 9:1 μAl/Al NPs material (by mass) is the best possible formulation. Indeed, when the total moles of active μAl⁰ per unit mass of Al NPs is considered (Figure 3), it can be seen that this was highest at 0.204 mol/g for the 9:1 μAl/Al NPs sample and decreased rapidly before becoming steady at ca. 0.035 mol/g in the limit for samples with the highest proportions of Al NPs.

The hydrolysis of the μAl/Al NPs composite material is a complex process that is kinetically very difficult to model. However, tracking the hydrogen gas production from 50 mg samples as a function of reaction time (Figure 4a) clearly revealed a dual phase bimodal process. Irrespective of mass ratio, a burst of hydrogen production occurred within ca. 1 min. The rate at which hydrogen was produced during this initial phase is linearly dependent upon the mass % of Al NPs in the sample, confirming first order kinetics (Figure 4b). This is entirely consistent with our previously determined kinetic profile for the hydrolysis of poly(N-isopropylacrylamide)-capped Al nanoparticles [40]. This was followed by a steadier gas formation phase, which settled at around 0.4 mL/min during the final 3 min of measurement, irrespective of the mass % of Al NPs in the sample. This is to be expected, given the now-dominant surface-area limited hydrolysis of the much larger μAl particles. It is also important that the reaction environment created by the Al NPs hydrolysis (vide infra) affords a stable and consistent supply of hydrogen from the μAl, which can be helpful for maintaining steady fuel generation for a device such as a hydrogen fuel cell.

For the 50% Al NPs/50% μAl sample after 3 days sitting in water, analysis by PXRD of the recovered, postreaction residue (Figure 5) determined a crystalline composition that was 94% Al(OH)₃ (s) and 6% elemental Al(s), presumably unreacted μAl. No AlO(OH) was detected in this measurement. The Al(OH)₃ formation is typically expected in most aluminum hydrolysis reactions below 280 °C [4,51]. Additionally, Al(OH)₃ is an affirmed catalytic promoter of this reaction [44,45,46,47]. In the case of our μAl/Al NPs composite samples, there was rapid consumption of the Al NPs to produce Al(OH)₃ in the initial activation phase of the reaction. This is possible because the oligo(epoxyhexane) cap on the nanoparticles is easily compromised by pure water. The Al(OH)₃ produced by Al NPs hydrolysis then catalytically drives the μAl hydrolysis phase by raising the reaction pH and compromising the micron-scale particle oxide shell. Higher proportions of Al NPs in the composite mixture will clearly generate more Al(OH)₃ during the activation, further elevating the pH, as was observed, and further activating the μAl toward the hydrolysis and ultimately increasing the hydrogen production.

The hydrolysis of aluminum nanoparticles to bulk Al(OH)₃ is a highly exothermic process (ΔH° = −1094 kJ/mol) [52]. It is conceivable, therefore, that energy released by this swift reaction could drive pursuant μAl particle hydrolysis. However, monitoring the temperature of the reaction solution using a thermocouple probe recorded a temperature increase of just 5–7 °C. Thus, it would seem the heat release rate is not reasonably sufficient to initiate degradation of the Al₂O₃ shell protecting the μAl particles and accelerate the μAl reaction. Most micron-scale and bulk forms of Al do not react appreciably with neutral water without some form of ball milling, chemical activation and/or elevated temperatures (≥35 °C) [1,10,53,54,55,56]. On the other hand, the Al₂O₃ protective shell is vulnerable to alkaline pH [1]. A sufficient increase in reaction pH might compromise the oxide layer to the extent that μAl hydrolysis is expedited. Since the Al(OH)₃ produced by Al NP hydrolysis is capable of promoting such pH-driven μAl hydrolysis to produce more Al(OH)₃, it could be viewed as a catalytic enhancement. Increasing the proportion of Al NPs in the composite material naturally increases the rate of initial Al(OH)₃ formation in the reaction mixture (supported, of course, by the observed increase of the initial rate of formation of the other product, H₂). We also measured a corresponding increase in the pH of the reaction solution at the end of the 15 min reaction monitoring period (Figure 6). Indeed, with the value of the slope close to 3, the dependence of the (basic) pH on the log(Al NP amount reacting) closely matches the 3:1 hydroxide-to-Al stoichiometry of the Al NP hydrolysis equation, which supports the rapid and complete consumption of the nanoparticles in the initial phase. The measured increase in resulting pH with the increasing Al NP proportion is offset by the reduced fraction of μAl in the sample. The result is the same steady rate of μAl hydrolysis observed in the final reaction phase of Figure 4a for all Al NP mass % samples.

The solution pH measured at the culmination of the hydrogen formation experiments ranged from 8 to 11 for the lowest and highest Al NPs mass % samples, respectively. This basic pH regime is consistent with Al³⁺ speciation analysis for the aqueous solutions shown in Figure 7a (and determined by equilibria equations shown in Appendix A), bearing in mind that Al(OH)₃ is a sparingly soluble salt. Thus, while most of the Al(OH)₃ precipitates out of solution, observed as a white solid, the thermodynamically equilibrated composition in solution ranges from ca. 50% Al(OH)₃ (aq)/50% (Al(OH)₄⁻ (aq) at pH 8 to almost entirely Al(OH)₄⁻ (aq) at pH 11. These equilibria systems also confirm that the Al₂O₃ shell protecting the μAl particles is most likely to be compromised at a pH greater than 8, yielding primarily Al(OH)₄⁻ ions. As a result, the total saturated concentration of aqueous Al species increased by three orders of magnitude as expected in the pH range 8–11 (Figure 7b, Appendix B). This, therefore, is the most likely mode of activation of the μAl toward hydrolysis for this composite material.

4. Conclusions

The sensitivity and reactivity of aluminum nanoparticles (Al NPs) were exploited to activate the hydrolytic reactivity of micron-scale aluminum particles (μAl) through the catalytic promotion by Al(OH)₃ and the associated increase in solution pH rather than by thermal activation. This method provides for the fabrication of a simple binary admixture of μAl with Al NPs that could be packaged as a portable solid and used to produce hydrogen gas on demand, such as that which might be needed for a fuel cell. Given that Al NPs are typically much more expensive and hazardous than μAl, this work demonstrates that it is beneficial to formulate a small (≤10%) proportion of the composite material as Al NPs relative the μAl to achieve the optimal μAl activation. Proportions of Al NPs >10% would likely not provide a substantially cost-effective performance benefit improvement in terms of overall H₂ production.