Surface Functionalization and Electrical Discharge Sensitivity of Passivated Al Nanoparticles.

Passivated aluminum nanoparticles are surface functionalized to elucidate their sensitivity against an electrical discharge. Two size fractions that differ in surface morphology are investigated. Electronic interactions between the partly inert, partly energetic organic molecules used for surface functionalization and the alumina surface are analyzed in detail. The nanoparticle surfaces are modified with the well-established, inert 2-[2-(2-methoxyethoxy)ethoxy]acetic acid, whereas energetic surface modification is achieved using 1,3,5-trinitroperhydro-1,3,5-triazine or the acidic and aromatic 2,4,6-trinitrophenol. A mechanistic model for the chemical surface functionalization of Al nanoparticles is hypothesized and corroborated by comprehensive optical and Fourier transform infrared spectroscopy studies. The surface structures are adjusted by developing a tunable stabilization procedure that prevents sedimentation and hence increases the saturation concentration in the liquid phase and finally affects the sensitivity character in view of electrical discharge ignition of dry powders. Detailed material characterization is conducted using transmission electron microscopy, combined with energy-dispersive X-ray spectroscopy and various absorption spectroscopy techniques (steady state in the infrared and ultraviolet/visible regime). The adjustment of surface structures of the distinct Al nanoparticle samples offers a valuable tool for tuning and tailoring the reactivity, sensitivity, stability, and energetic performances of the nanoparticles, and thereby enables the safe use of these multipurpose nanoparticles.

Introduction

Aside from their use in diverse industrial settings and as pharmaceuticals, food additives, cosmetics, and other various household products,[1] aluminum/alumina nanoparticles are used as effective pigments for automotive coatings (as “silver”) and as coloring pigments that are generated by an interfering thin TiO2 shell.[2] In addition, Al/Al2O3 nanoparticles are also used for increasing the modulus of elasticity of polymers (e.g., poly(methyl methacrylate), about 25% higher compared to the untreated polymer) or for vinyl ester resin nanocomposites.[3−5] However, the focus of the present work is set on energetic applications for dual use in defense and civil environments, where they are essentially employed as an effective fuel component.[6] For this purpose, aluminum core nanoparticles are often combined with oxidizing agents (e.g., WO3 or KMnO4), for thermites, or by adding additional 1,3,5-trinitroperhydro-1,3,5-triazine (RDX), for energetic nanocomposites,[7] and further as additional fuel for explosive mixtures such as high-brisance explosive (HBX) formulations (e.g., “HBX3” is RDX plus nitrocellulose, calcium chloride, and calcium silicate 31.3 ± 3%, TNT 29.0 ± 3%, aluminum 34.8 ± 3%, and wax plus lecithin 4.9 ± 1%).[8] Therefore, a detailed knowledge of diverse possible interactions between aluminum/alumina nanoparticles and the interface active molecules/material for better and safer processing and deeper understanding of aluminum-based nanoparticle systems is required. Here, the surface modification of passivated aluminum nanoparticles by surface functionalization with organic molecules is shown. In this regard, scientific literature on the nature of chemical interaction between compounds and Al/Al2O3 surfaces is actually rare. The chemical interaction between Al nanopowders and RDX at higher temperatures (>70 °C) was observed by Kwok et al. and has been an exception up to now.[9] Kwok and co-workers shed light on the consequences of a chemical interaction between Al and RDX (namely, the generation of NO2 and N2O) but not on chemical interactions between Al and RDX with respect to bond formation or similar aspects.

However, chemical surface functionalization of Al-based nanoparticles by organic molecules in the literature is mostly motivated by tuning and tailoring Al/Al2O3 nanoparticle properties (e.g., higher solvent miscibility and disperse stability in water or carbon dioxide, reactivity).[10−12] In general, the molecule that is foreseen to be bound to a nanoparticle surface requires an anchor group, which forms a covalent bond (no hydrogen bonds or van der Waals interactions). The lack of an adequate anchor group would lead to a physically modified particle by simple physical adsorption of molecules. Chemical surface functionalization can be realized via an inorganic or organic pathway and is shown for pure aluminum nanoparticles, pure alumina nanoparticles, and passivated aluminum (core–shell) nanoparticles. Then, coatings can be amorphous or crystalline. The latter could be epitaxially or nonepitaxially grown.[13] Kaplowitz et al. report an inorganic coating of pure aluminum nanoparticles with iron oxide for nanoenergetic fomulations.[14] Other inorganic coatings of passivated aluminum or pure alumina nanoparticles were successfully prepared out of alkaline earth metal oxides or silanes.[15,16] In contrast to alumina-coated Al nanoparticles, pristine alumina nanoparticles cannot operate as fuel particles because they are already oxidized. Nevertheless, chemical modification procedures of alumina nanoparticle surfaces can be applied for Al/Al2O3 core/shell nanoparticles because the organic compounds only chemically interact with the Al2O3 surface, and physicochemical properties of those particular surfaces should be the same. Jouet et al. succeeded in coating pure Al nanoparticles with perfluoroalkyl carboxylic acids.[17] On the other hand, passivated Al nanoparticles were reported to be surface-modified with acrylic monomers.[10] Kappagantula et al. showed that the surface coverage of Al fuels with an organic acid shell can be used to control their reactivity with regard to the flame propagation velocity.[12] In conjunction with the surface functionalization of Al2O3, different kinds of anchor groups have been reported, most of them being organic acids and carboxylates. In general, functional acid groups are well known to bind to many kinds of oxides because oxides natively develop an OH-terminated surface that is attacked by acid groups, resulting in a condensation reaction and formation of an ether (oxygen) bridge.

Here, a higher stability against sedimentation inside the liquid phase for effective manufacturing and treatment of Al nanoparticles in acetone and ethanol is achieved. Second, a safer handling of Al nanoparticles is attained. This implies that a lower sensitivity against ignition of Al nanoparticles, especially against ignition of dry powders by electrostatic discharge (ESD), is realized. For this purpose, two different Al nanoparticle fractions with mean particle sizes of 50 and 100 nm were systematically surface functionalized with inert and also with energetic organic compounds. As an inert compound, the stabilizing 2-[2-(2-methoxyethoxy)ethoxy]acetic acid (TODA) is tested and for energetic surface modification, RDX and the acidic 2,4,6-trinitrophenol (TNP) are chosen. In addition, a detailed spectroscopy (transmission/absorption), microscopy (electron microscopy/high-resolution-TEM/energy-dispersive X-ray spectroscopy (EDS)), and sensitivity-based analysis is provided.

Results and Discussion

The sizes and morphology of the raw materials were systematically investigated using transmission electron microscopy (TEM). TEM images in Figure visualize Al nanoparticles of two fractions, which exhibit a mean size of xmean = 50 nm (Al_50) (a) and a mean size of xmean = 100 nm (Al_100) (b). Both Al_50 and Al_100 samples have spherical shapes. Whereas the surfaces of the Al_50 samples are smooth, the Al_100 fraction possesses rough and porous surfaces. Dynamic light scattering (DLS) scans of the different Al nanoparticle charges dispersed in the liquid phase (ethanol or acetone) are shown in the Supporting Information (SI) (Figure S1). The particle size distributions (PSDs) were collected for both fractions, Al_50 and Al_100. Both PSDs, in ethanol as well as in acetone, tend to be larger than xmean, as specified by the manufacturer. Al_100 is ∼60% larger for ethanol, ∼20% larger for acetone, and Al_50 is ∼380% larger for both solvents. This discrepancy might follow from agglomeration and insufficient redispersion of the primary nanoparticles. This implies that agglomeration takes place with increasing probability the smaller the nanoparticles are, the larger the surface-to-volume ratio (sa/vol) is, and the more efficient the attractive interface interactions such as van der Waals forces will become.

Figure 1

TEM images of (a) spherical Al nanoparticles with a mean size of 50 nm (“Al_50”) and (b) Al nanoparticles with a spherical shape and a mean size of 100 nm (“Al_100”). The Al nanoparticles in (a) exhibit smooth surfaces, whereas the surfaces of the Al nanoparticles in (b) are much rougher and porous.

Higher reactivity stems from the enhanced reactivity of surface molecules compared to molecules within the bulk.[18,19] Concerning higher reactivity, Al nanoparticles are favorable when compared to their homologue micromaterial.[9,18] Surface modification through chemical functionalization obviously squeezes the PSD. For instance, xmean of pristine Al nanoparticles (170 nm) dispersed in ethanol decreases through surface functionalization to 120 nm with TODA and to 140 nm with TNP. RDX has no major effect on the PSD (see SI, Figure ). A perceivable and major shift is finally evaluated as a hint for the breakage of agglomerates and an improved stabilization of primary particles through chemical surface functionalization.

Al nanoparticles are surrounded by a predominantly amorphous passivation (oxide) shell, which exists under ambient conditions. The oxidation of the Al nanoparticles results in a several nm thick amorphous (sometimes semicrystalline) alumina coating.[14] The elemental distribution of a precursor Al_100 nanoparticle fraction (without surface modification) is derived from TEM/EDS measurements and is visualized in Figure .

Figure 2

(a) Dark-field TEM image of an “Al_100” sample in scanning mode: an Al-core nanoparticle (and two likely agglomerated smaller satellites) with particle sizes between 180 and 60 nm can be seen. (b) refers to the EDS scan for aluminum, (c) for oxygen, and in (d) all scanning modes are combined. The blue corona illustrates the oxide shell around the Al nanoparticles.

Figure shows a dark-field TEM analysis in scanning mode (a), the associated EDS scans probing “aluminum” (b, 1.5 keV) and “oxygen” (c, 0.5 keV), as well as an overlay of (b) and (c), demonstrating the large oxygen content inside the Al2O3 corona of an Al_100 sample. One rather large particle (∼180 nm) center and some agglomerated satellites are depicted. The aluminum oxide shell appears as a blue corona with a significantly higher oxygen content compared to the inner particle section (Figure d). The TEM images in combination with EDS mapping allow the determination of the thickness of the Al2O3 shell, which is 3 nm for both fractions, Al_50 and Al_100. Thereupon, the oxide content of the Al_50 sample is larger than that inside the Al_100 fraction. This implies that the mass ratio m(Al):m(Al2O3) of the Al_50 fraction amounts to 1.5:1, whereas that for Al_100 is larger with 4.15:1.

In the following, real chemical bonding between diverse organic molecules, taking one energetic representative (TNP), one inert steric dispersant, and an oxygen carrier agent (TODA), is presented because interactions of these compounds are still underrepresented in the literature, or are absent. Spectroscopy studies (Fourier transform infrared spectroscopy (FTIR); ultraviolet/visible (UV/vis)) were performed to investigate the electronic interactions between passivated Al nanoparticles and RDX, TNP, and TODA at the interface. Figure shows the fingerprint region of FTIR-attenuated total reflectance (ATR) transmission spectra containing both Al_50 and Al_100 fractions in the form of dry powders. In the case of Al_50_RDX and Al_100_RDX, a loss of the majority of RDX through the washing procedure can be derived from very weak RDX vibrational modes within the FTIR transmission spectra (thus, it is not shown here, see the SI). This leads to the conclusion that RDX does not form covalent bonds with the Al2O3 surface. We assume that RDX does only weakly interact with the Al2O3 surface and is attached at the surface via physisorption.

Figure 3

FTIR transmission fingerprint region of pure surface compounds and modified Al nanoparticles of the TNP (a) and TODA (b) charges. On the right, assignment of the vibrational bands of the FTIR transmission spectra of TNP- and TODA-modified Al_50 and Al_100 nanoparticles.

The FTIR spectra of the TODA and TNP surface-modified Al nanoparticles exhibit significantly different vibrational signatures that were successfully assigned to the respective group frequencies. A clear shift of single stretching vibration modes could be detected for the TNP-functionalized (table in Figure a) and the TODA-functionalized samples (table in Figure b).

TNP vibration bands of functionalized surfaces are retained after three washing steps and are blue-shifted—except for the O–H stretching vibration peak, which is quenched because of surface binding. The spectrum of the free TNP C=C-stretching vibration exhibits a band at 1629.7 cm–1, whereas the surface-bound one is altered to 1634.53 cm–1. This difference in the wavenumber of the vibrational bands stems from the obviously higher aromatic character of the benzene ring when surface bound via the deprotonated OH group (the electronic density, also from Al2O3, is withdrawn into the electron-poor aromatic system via the NO2 groups, which offer a negative inductive and mesomeric effect). In fact, the intensity of the C=C vibrational peak decreases with the simultaneously rising intensity of the aromatic peak at Al_50_TNP (1616.21 cm–1) as well as at Al_100_TNP (1614.28 cm–1).

For Al_50_TODA, just small redshifts of all vibrational modes besides a blueshift of C–H deformation vibrations (1454.20 and 1355.84 cm–1) are detected. All vibrational modes of surface-bound TODA (Al_100) are red-shifted. This indicates the decrease in the bond strength of all intramolecular bonds of TODA due to the electronic interaction with the Al2O3 surface. The diverse shifts of the vibrational C–H bands in the spectra of Al_50 (blue) and Al_100 (red) presumably arise from the differences in the roughness, curvature, and degree of porosity of the diverse fractions, which is not completely understood.

Concerning the binding of TODA, we assume that it anchors via its carboxyl group to the Al2O3 surface. This is unambiguously shown by a shift of the C=O vibrational band from 1739.64 cm–1 (free molecule) to 1733.85 and 1724.20 cm–1 in the case of the surface-bound molecules at Al_50 and Al_100, respectively. The optical properties of RDX-, TNT-, and TODA-treated Al_50 and Al_100 nanoparticles were studied employing UV/vis absorption spectroscopy (Figure ).

Figure 4

UV/vis absorption spectra of RDX, TNP, and TODA, and pristine and RDX-, TNP-, and TODA-terminated Al_50 (top) and Al_100 nanoparticles (bottom).

The predominant characteristics of all UV/vis spectra arise from light scattering through the dispersed solid phase inside the liquid phase in the Al nanoparticle suspensions. The UV/vis absorption spectrum of aliphatic RDX shows a strong absorption onset at 242 nm (5.12 eV), which is assigned to the electronic n → π* transition. In contrast, absorption due to this electronic transition is drastically reduced for RDX molecules, physisorbed at the Al2O3 surface. This indicates that only a small concentration of RDX molecules are really attached to the surface of Al nanoparticles.

Owing to the aromatic character of TNP, the UV/vis absorption spectrum offers two bands at 209 nm (5.93 eV) and 357 nm (3.45 eV). Moreover, these absorption bands include shoulders at 236 nm (5.28 eV) and 416 nm (2.98 eV). For a detailed energy diagram of TNP and its electronic transitions please see Figure S3. The first peak and the first shoulder at 209 nm and 236 nm, respectively, are assigned to π2 → π1* and π1 → π2* transitions. The second band at 357 nm and its corresponding shoulder (416 nm) are related to the n → π1* and n → π2* transitions. In contrast, the UV/vis spectra of TNP-functionalized Al_50 and Al_100 give two absorption bands at 236 nm (5.28 eV) and 288 nm (4.31 eV). The absorbance values of the bands of Al_100_TNP are less intensive compared to those of Al_50_TNP because pristine Al_100 UV/vis absorption spectrum offers an absorption minimum in the relevant range. The absorption bands within the spectra of the TNP-treated Al nanoparticles correspond to the blue-shifted n → π1* and n → π2* transitions of TNP: The chemical bond between TNP and the Al2O3 surface increases the electron density in the aromatic system of TNP caused by the electron-withdrawing effect of the NO2 groups (see also FTIR results). This leads to an increase in the intramolecular bond strength of TNP. This increase is recognizable by a higher energy level of π* orbitals and therefore by blueshifts within the absorption spectra.

The UV/vis absorption spectrum of TODA exhibits bands at 214 nm (5.79 eV) and 266 nm (4.66 eV) according to the π → π* and n → π* transitions. Also, the spectrum of Al_100_TODA gives two absorption bands at 243 nm (5.10 eV) and 289 nm (4.29 eV). In contrast to pristine TODA, these bands are red-shifted. These redshifts occur due to a decrease in the intramolecular bond strength in TODA, which is a result of the bond formation between TODA and the amorphous Al2O3 surface. However, only one absorption band is ascertainable within the UV/vis absorption spectrum of Al_50_TODA because the second one is superimposed by light scattering. Electronic transitions and their exact values are presented in detail within energetic charts in Scheme .

Scheme 1

Energy Level Diagrams of TNP-Terminated Al_50 (a) and Al_100 (b) Nanoparticles and TODA-Terminated Al_50 (c) and Al_100 (d) Nanoparticles

Finally, one can conclude that results from UV/vis spectroscopy fit well with the results obtained from FTIR. That means, real chemical and covalent bonding, and, moreover, a specific electronic interaction between TNP, TODA, and the Al2O3 surface is substantive, whereas RDX does not manifest chemical affinity to the Al2O3 particle surface. Supported by presented spectroscopy results, applied compounds are finally found to act with the Al2O3 nanoparticulate surface via the proposed reaction mechanisms sketched in Scheme . TNP and TODA are identified to form covalent bonds to the Al2O3 surface, depicted in Scheme b,c. For RDX, the interaction with the Al2O3 surface takes place via physisorption. In the case of TNP, we propose that, first, protonation of the OH-terminated surface occurs due to the highly acidic character of TNP (pKa(TNP) = 0.29 (24));[20] second, the condensation and electronic attack of the nucleophile is executed, and finally, bonds are formed. Surface binding of TODA (c) follows the same mechanism as (b) and occurs via the carboxyl group (pKa(TODA = 4)).[21] A possible collapse of the TODA (black arrow in (c)) conformation may result from O–H interactions.

Scheme 2

Reaction Mechanisms of (a) RDX, (b) TNP, and (c) TODA at the Amorphous Al2O3-Surface

Reaction mechanisms of (a) RDX, (b) TNP, and (c) TODA at the amorphous Al2O3-surface (marked in gray), as being OH-terminated (see FTIR); the dotted lines in (a) represent H bonds, (b) suggests a possible mechanism for the chemical bonding of TNP to the surface; clockwise: (i) deprotonation of TNP, (ii) protonation of −OH and condensation, (iii) attack of the nucleophile, and finally (iv) bond formation. (c) offers the result of a comparable mechanism; however, oxo groups from TODA can likely interact with −H at the surface to support H-bonding and a flat orientation of the TODA molecule on top of the Al/Al2O3 nanoparticle (black arrow).

Furthermore, the prerequisite condition for good processability of suspensions is a sufficiently high stability against sedimentation (stabilization) and high solid percentage within the liquid phase. Therefore, we examined the dispersibility of pristine and surface-modified Al/Al2O3 nanoparticles in acetone and determined the respective saturation concentrations, which are listed in Table . Stabilization against sedimentation is apparent with an increase in the saturation concentration after a time interval of 3 days. In this context, an increased saturation concentration through surface functionalization also mirrors an efficient interaction of the used surface-active compounds with the used particle fractions. As expected, suitably surface-modified Al nanoparticles demonstrate a higher dispersion stability (higher amount of saturation concentration). Differences between pristine Al_50 and Al_100 are likely based on the mean overall density, which is, according to the individual core–shell volume ratio of Al_50 and Al_100 (Al/Al2O3 = 2:1 for Al_50 and 6:1 for Al_100), 3.11 g cm–3 for Al_50 and 2.88 g cm–3 for Al_100. As the higher density of solids when dispersed in a liquid phase leads to faster sedimentation, the increased saturation concentration of Al_100 versus Al_50 is coherent. From a spectroscopy point of view (FTIR), there is no hint for differing chemical surface terminations with regard to Al_50 and Al_100. Minor enhancement of the saturation concentration of functionalized particles could be obtained with RDX. In both cases, the concentration of the solid phase could be enhanced, 10% for Al_100_RDX and 16% for the Al_50_RDX. This result is explained by minor physical interaction between the two compounds, as can be derived from spectroscopy. Chemical surface functionalization of Al_100 and Al_50 nanoparticles with TNP facilitates stable dispersions, with nanoparticle concentrations increased by 127 and 28%, respectively. The highest concentration increase in Al nanoparticle dispersions with 104% (Al_50) and 133% (Al_100) was achieved with TODA. This again mirrors the efficient surface functionalization ranked as follows: RDX exhibits rather poor surface modification and therefore poor stabilization potential for Al2O3; TNP exhibits much better characteristics; and TODA reveals the highest saturation concentrations and stabilization characteristics in the case of acetone (see Table ). The latter is obviously less linked to good affinity of the molecules to the surface of the particles because TNP binds more effectively through a strengthening of the aromatic character (see the spectroscopy part), but rather linked to an obviously better attractive interaction of the TODA-modified surface to acetone and ethanol.

Table 1

Saturation Concentration Values of Pristine and RDX-, TNP-, and TODA-Stabilized Al_50 and Al_100 Nanoparticles in Acetone

sample	cm (g/L)	wt %
Al_50_pure	0.20	0.025
Al_50_RDX	0.23	0.029
Al_50_TNP	0.25	0.032
Al_50_TODA	0.40	0.051
Al_100_pure	0.24	0.030
Al_100_RDX	0.26	0.033
Al_100_TNP	0.54	0.068
Al_100_TODA	0.55	0.070

Figure displays ESD energies by spark ignition of RDX-, TNP-, and TODA-functionalized and pristine Al nanoparticles. A star represents at least one reaction out of six tests, a square represents no reaction within six tests. Congruency of two symbols is not possible but rather a narrow pair of values. For example, number 5 is Al_100 pure and offers 2.09 mJ (reacting) and 1.85 mJ (not reacting), see Table . The interval between “reaction” and “6× no reaction” increases with higher millijoules because external capacitors are used, each with a fixed farad (F) value (please see the Experimental section). Surprisingly, at first sight, energetic surface modification does not automatically lead to higher sensitivity and thus higher reactivity. Furthermore, differences between Al_50 and Al_100 with identical surface molecule modification could be derived. In this context, Table represents all sensitivity values, also including mechanical sensitivity against impact and friction of the tested material in the form of dry and loose powders. In a nutshell, smaller Al particles (Al_50) show increased sensitivity (0.31 mJ, reacting) compared to the larger fraction (Al_100, 2.09 mJ, reacting). This tendency is in line with literature results[9,18] and understandable because smaller particles show higher sa/vol and surface molecules reveal enhanced reactivity (see the first section on PSD results from DLS). Then, the surface-modified Al_50 and Al_100 nanoparticles tend to require higher spark energies for ignition. In opposition to pristine particles, Al_50_TODA is less sensitive (56.56 mJ, reacting) compared to the larger particle fraction surface modified with the same molecule (Al_100_TODA, 13.7 mJ, reacting). This fact is proposed to be a result of the absolute amount of surface molecules (Al_50 offers a higher sa/vol and more TODA was applied, see the Experimental part) and varying surface morphology: Al_50 constitutes a smooth surface where TODA molecules can easily attach for binding, whereas Al_100 bares a much more rough and porous surface (see Figure ) where TODA molecules could be hindered. According to the latter, also TODA, which is already bound to the surface, could sterically hinder additional molecules when it is not upright bound due to surface inhomogeneity. Because RDX is not binding in an appropriate manner, as derived from spectroscopy studies, both ESD sensitivity values of RDX-modified particle fractions are near those of the pristine samples (see Table ).

Figure 5

ESD map of all combinations of Al nanoparticles and surface functionalization; also including blank surfaces (pristine particles). Exact values can be taken from Table .

Table 2

Sensitivity Testing Results of the Reference Material and Combinations

system	ESD (reaction) [mJ]	ESD (6× no reaction) [mJ]	friction (6× no reaction) [N]	impact (6× no reaction) [J]
Al_100 pure	2.09	1.85	>360	>50
Al_100 RDX	2.34	2.09	>360	>50
Al_100 TNP	24.12	20.36	>360	>50
Al_100 TODA	13.7	12.22	>360	>50
Al_50_pure	0.31		>360	>50
Al_50 RDX	0.31		>360	>50
Al_50 TNP	0.31		>360	>50
Al_50 TODA	56.56	45.81	>360	>50

In the case of TNP-surface functionalization, Al_50 reacts at 0.31 mJ and Al_100 at 24.12 mJ. Here, it is suggested that TNP is better bound to Al_100 because the molecule is smaller and more compact than TODA. Consequently, surface roughness does not constitute a major hindrance, and a better desensitization could be detected compared to TODA (in line with a strengthening of the aromatic character, see the spectroscopy part) because pristine Al_50 and TNP-modified Al_50 both show ESD values of 0.31 mJ; a final evaluation is not possible, as 0.31 mJ states a limit of the device (see the Experimental section).

It is further proposed that the decrease in reactivity and increase in stability by chemical surface functionalization stem from the saturation of high reactivity sites by chemical functions at the Al2O3 particle surface. In fact, the binding of active surface molecules via condensation, as described here, reduces the overall energy by releasing the stable leaving group water (see Scheme ). Moreover, highly reactive sites can also be generated by defects in general, and, for instance, dangling bonds in particular, which are effectively saturated by surface-active molecules to reduce the overall surface energy, and therefore to reduce sensitivity to ignition. In this respect, the importance of ESD measurements in general is to make sure that the sensitivity of the material to a static discharge is below the static potential that can be developed by the human body (5–20 mJ).[6] Hence, process safety, for example, during physical mixing of Al nanoparticles, even when they are passivated, is of uttermost interest. The following materials were tested to be under 20 mJ and are therefore dangerous to handle (safety hazard) in an insecure environment: “Al_100 pure”, “Al_100_RDX”, “Al_100_TODA”, “Al_50_pure”, “Al_50_TNP”, and “Al_50_RDX”.

Regarding impact and friction, all of the tested materials are completely insensitive, >50 J and >360 N, respectively. All values for friction and impact are out of the tested range, which is defined through the individual testing mechanism and testing device (see the Experimental part).

Experimental Section

Synthesis

RDX is a well-known, highly explosive substance. For surface modification with RDX, “Hexogene M5” from Eurenco (former SNPE laboratories, Société Nationale des Poudres et des Explosifs, Paris, France) was purchased and used as received without further purification. TNP (moistened with ≥40% water, ≥98%) was purchased from Sigma-Aldrich and dried for 12 h/60 °C. TODA (techn. gr.) was purchased from Sigma-Aldrich. Chemical formula can be derived from Scheme . Acetone (HPLC grade) was purchased from Sigma-Aldrich. Ethanol (abs.) was purchased from Carl Roth GmbH & Co. KG.

Aluminum nanoparticles (Al_50) with an average diameter of 50 nm were purchased from Nanotechnologies, Inc. Aluminum nanoparticles (Al_100) with an average diameter of 100 nm were purchased from Intrinsiq Materials, Inc. In a first step, suspensions of Al_50 and Al_100 in acetone (w = 0.10 wt %, cm = 0.79 g/L) were produced by adding Al_50 or Al_100 (150 mg) to acetone (190 mL) and by a subsequent 5 min treatment of sonication inside an ultrasonic bath.

For surface modification of Al_50 or Al_100, 1 mL of solutions of RDX, TNP, and TODA in acetone were added to the previously produced aluminum suspensions. The concentrations of the used compounds are listed in Table . A monolayer of molecules was calculated for each particle size and “precursor” in wt %. This value was multiplied by 2 to ensure that at least a monolayer in the case of efficient surface functionalization is formed. The mixtures were treated thrice by ultrasonication for 20 min with an interruption of 5 min between each period. Then, the nanoparticles were collected by centrifugation (5000 rpm, 20 min) and washed thrice with acetone (190 mL) through centrifugation and redispersion cycles. The resulting suspension was allowed to age for 3 days. Afterward, 50 mL of the supernatant was extracted. Acetone was removed via rotation evaporation and the Al nanoparticles were weighed to calculate the saturation concentration of the suspensions. Then, acetone of the remaining suspensions was removed by rotation evaporation and the particles were collected. As references, Al_50 and Al_100 were also treated by this procedure without adding any molecular surface function. “Al_X_Y” is decrypted as “aluminum_size_surface motif”.

Table 3

Concentrations of the Used Surface-Active Molecules

chemical	Al_50 (mM)	Al_100 (mM)
RDX	51.7	27.9
TNP	51.7	27.9
TODA	76.9	41.4

Analysis Methods

TEM images were recorded on a JEOL, ARM200CF (Tokyo, Japan), with a nominal point resolution of 0.8 Å at Scherzer defocus. A cold FEG Gun 80–200 kV provided the electron source. Semiquantitative elemental analysis and chemical mapping were performed using an energy-dispersive X-ray spectroscopy (EDS) system.

The UV–vis–NIR-absorption spectra of the samples were recorded on a Cary 5000 spectrophotometer system from Agilent Technologies (Santa Clara). A macroquartz glass cell from Hellma Analytics (Müllheim, Germany) with an optical light path of 10 mm was taken for both measurement techniques (DLS (see the SI), UV/vis). FTIR spectra were recorded using the tensor 27 from Bruker Optik GmbH (Ettlingen, Germany) equipped with a standard ATR-unit MIRacle from Pike (Madison). The ATR signal was converted by the Bruker Opus 6.5 software into transmission data.

A full-fledged sensitivity investigation, strictly according to the Federal Institute for Materials Research and Testing guidelines (www.bam.de; BAM; Berlin; Germany), was additionally executed using classical test devices for impact (fall/drop hammer, BAM), friction (Julius Peters K.G. Berlin, Germany), and ESD (OZM Research, ESD 2008, Hrochův Týnec, Czech Republic). In this context, loose and dry powders were investigated. The resulting values are listed when at least one of the six tested samples reacts (statistical relevance: 98.4%), following the recommendation of ref (22). “>” indicates a limit of the test device. For ESD, external ceramic capacitors were used. Each corresponds to an exact value in Faraday. Therefore, ESD tests cannot be continuously performed. Intervals between two ESD values get larger when measuring higher ESD values (quadratic function). The apparatus is described in detail elsewhere.[23] For safety reasons, an additional ESD value is given when six tested samples of one specific material does not react. Please note that ESD values less than 0.31 mJ (capacitor 45 pF, 6 kV) cannot be adequately detected because no defined sparks are generated at this limit.

Conclusions

Two aluminum nanoparticle fractions that differ in mean single particle size (50 and 100 nm) and surface roughness were surface-modified by diverse organic motifs. In this context, inert and energetic organic molecules were chosen for surface functionalization: energetic RDX, TNP, and inert TODA partly interact with the Al surface. Whereas TODA and TNP were proven to covalently bind to the Al2O3 surface via their carboxyl group (TODA) and acidic hydroxyl group (TNP), respectively, RDX only interacts physically with the Al2O3 surface. These findings were elaborated and unambiguously corroborated upon performing FTIR transmission and UV/vis absorption spectroscopy experiments. Additionally, this work provides a reasonable reaction mechanism for Al2O3 surface modification with applied compounds. The successful surface functionalization with TNP and TODA is evident from the enhanced stability of the respective dispersions against sedimentation in ethanol and acetone. The most interesting and safety-relevant result is the observation of decreased sensitivity against ignition through energy by an electric discharge (spark). The TODA modification lowers Al-powder sensitization at least 40 times, which is, in fact, also a major improvement for safety reasons regarding unwanted ignition of aluminum powders. Even energetic compounds like TNP can provide enhanced desensitization of pure Al core with Al2O3 shell nanoparticles. Finally, as a result of the varying surface roughness, differences in terms of the two Al fractions were found for the binding of identical molecules.