Characterization of nanometric inclusions via nanoprojectile impacts

It has been shown that sputter and secondary ion yield from nanoparticles, NPs, in sizes below a few tens of nanometers are size dependent.¹ We refer here to dimensions which are not sufficient for complete projectile energy deposition in a direct collision between the incident ion and the nano-object. Grazing and interfacial collisions further complicate the nature and abundance of the ejecta. The nonequivalency of impacts can clearly affect the accuracy of NP analysis. However, when NPs are ultrasmall (<10 nm), they can be completely fragmented/atomized in one collision; thus, all impacts should be equivalent. We examine here this case and its consequences for characterizing ultrasmall NPs and their chemical environment. The samples were aggregations of gold atoms, specifically 55, 147, or 225 atoms encapsulated within a poly(amido amine) (PAMAM) dendrimer. Encapsulation within a dendrimer structure prevents agglomeration of NPs, prevents passivation of catalytic surfaces, and can be used to control catalytic rates.^2–4 While most analysis techniques for the characterization of nanoparticles produce an averaged measurement due to the blending of subensembles within a population, it is possible to obtain accurate analysis of discrete ensembles by measuring each nano-object individually. This approach allows for a nuanced method that returns both averaged data and separates subensembles based on the treatment of the data after analysis. The emission of characteristic gold adducts from the nanoparticle is detailed below.

The experiments were run in the event by event bombardment-detection regime, where a single projectile hits a NP coupled with the individual recording of the secondary ions (SIs) from each impact. To be practical, this approach requires a projectile generating a large ion multiplicity.⁵ The projectile of choice was Au₄₀₀⁴⁺, itself a nanoprojectile of ∼2 nm in diameter.

Au NPs were reduced from a solution of HAuCl₄ by excess NaBH₄ in the presence of generation 7 PAMAM dendrimers. The reduction to zero valent gold causes the aggregation and growth of gold into NPs, the size of which is controlled by the original concentration of HAuCl₄ in solution.⁶ Au₅₅, Au₁₄₇, and Au₂₂₅ were produced, corresponding to nanoparticle diameters of 1.3, 1.5, and 1.7 nm, respectively (see supplementary material).⁷ Solutions of dendrimer encapsulated nanoparticles (DENs) were then drop cast onto Si (1 0 0) wafers after cleaning by sonication in absolute ethanol. Upon evaporation of excess solvent, a layer of DENs was produced where the gold NPs are isolated from one another due to the enveloping dendrimer structure.

The instrument used in this study is a custom built secondary ion mass spectrometer utilizing an Au liquid metal ion source (LMIS). This LMIS generates a continuum of gold clusters from which Au₄₀₀⁴⁺ is mass selected, through the use of a Wein filter, for use as the analysis projectile. The ion source is installed on a platform biased at 120 kV, which provides the projectile acceleration toward a −10 kV target, giving a total projectile energy of 520 keV.⁸ Deflectors and a pulsing system are used to reduce the Au₄₀₀⁴⁺ beam down to individually impacting projectiles at a rate of ∼1000 Hz. Ejected electrons are deviated using a magnetic prism to a microchannel plate (MCP) detector that starts a time of flight measurement. Ejected SIs are mass analyzed using a reflectron ToF mass spectrometer and then are detected using an eight anode MCP detector.⁵ 

Due to the pulsing of the bombardment, each projectile impact can be recorded separately as individual events with each event separated in both time and space.⁹ Several million impacts are collected and then processed using a program developed in-house (surface analysis and mapping of projectile impacts). A conventional mass spectrum can be obtained by summing all events together. A subset of mass spectra containing an ion of interest can be selected, termed a coincidence mass spectra. This coincidence mass spectrum contains all ions coemitted with the ion of interest. Ions present in the coincidence mass spectra represent ions coemitted from a single impact and therefore must be colocated within the sample surface. Examples of this method have been demonstrated on cell surfaces¹⁰ and the environment of catalytically active molecules.¹¹ 

Dendrimer encapsulation allows for the individual bombardment of a single nanoparticle at a time. Chemical information about the surrounding area can be obtained through coemitted ions, but first characteristic ions must be classified. Table I shows ions that are representative of different chemical moieties within the sample. Ions arise from an impact on the Au NP or the surrounding dendrimer scaffolding. The dendrimer has two distinct chemical configurations, both of which are formed from the fragmentation of a dendrimer chain upon bombardment. The difference arises from whether or not the chain has undergone reductive damage from the NaBH₄ used during the synthesis of the Au NP. In the total mass spectrum (Fig. 1), there are peaks that arise from impacts on the native and reduced dendrimer structure, as well as on the Au NPs.

With the coincidental methodology, it is possible to select SIs that arise from a specific impact. Since Au₂CN⁻ and Au(CN)₂⁻ are present only in spectra with the bombardment of a Au DEN and not in a blank dendrimer sample (data not shown), keying in on these two adducts allows for the selection of impacts of the Au₄₀₀⁴⁺ projectile with the Au NP without worrying about gold contribution from the projectile. With Au(CN)₂^– set as the coincidental condition while evaluating the yield of Au₂CN⁻, the emission of the gold adduct is enhanced as the size of the Au NP increases (Fig. 2). This agrees with previous studies that have shown an increase in gold clusters from nanoparticles as the size of the NP increases.¹ 

Utilizing characteristic ions identified as belonging to different portions of the sample make it possible to use the coincidental methodology to probe the chemical environment around the NP. Using Eq. (1) where Y_A,B is the coincidental yield of ion A with ion B and Y_A is total yield of ion A, the coemission of two ions can be evaluated. If ion (A) is coincidental with another ion (B) that signifies the two ions are coemitted from an individual impact on the sample. The ratio of the coincidental yield to the total yield, where ratios above 1 show coemission, and therefore colocation, while ratios under 1 show chemical segregation of the two ions

When ions belonging to the native dendrimer structure are emitted, there is an increase in coemission of characteristic gold ions. For example, when impacts containing C₅H₉NO⁻ are selected, Au(CN)₂⁻ the coincidental yield is three times larger than the total yield. Thus, there must be chemical colocation of less than 10 nm of the Au NP with dendrimer branches that have not undergone reductive damage. Conversely, it is shown that the coemission of gold ions is lowered when reduced dendrimer fragment ions are coemitted. The coincidental yield of Au(CN)₂⁻ with C₇H₁₁N⁻ is 1% of the total yield; therefore, the NP does not colocate with damaged dendrimer regions during encapsulation (Fig. 3). These results show that the Au NP preferentially grows in native dendrimer structure environments during synthesis, which could be used to direct growth of NPs in future systems.

The Au NP size also affects the chemical environment of the dendrimer, as shown in Fig. 4. As the nanoparticle size increases from 55 to 225 atoms, there is a clear increase in the segregation of native and reduced dendrimer structures. This is evidenced by the decrease in the ratio of coincidental yield to total yield that signifies a reduction in the coemission of two ions. Interestingly, characteristic gold ions have greater coemission with native dendrimer structures and greater segregation from reduced dendrimer structures for larger NPs. This correlates well with the idea that Au NPs prefer native dendrimer structures as we expect greater reductive damage to the dendrimer during synthesis of larger NPs due to an increase in NaBH₄ used.

This study describes the effect of the size of ultrasmall nanoparticles within a nanodomain upon the emission of characteristic ions by bombardment with massive projectiles. Along with the successful detection and analysis of ultrasmall gold nanoparticles with as few as 55 atoms, we show that for NPs containing between 55 and 225 atoms, there is a linear relationship between their size and SI emission. This suggests a constant ionization probability for ultrasmall nanoparticles of the same composition, with an overall ion yield dependent upon the number of atoms available within the emission volume of Au₄₀₀⁴⁺. This volume encompasses the entire NP in these ultrasmall systems, so all atoms of the NP are sputtered.

Due to complete sputtering, differences in impact parameters are not observed. This is a distinction from larger NP systems, where the increased size provides for different types of impacts on the NP and hence nonequivalency of impacts.¹ 

Our data show that the gold nanoparticles preferentially locate within specific chemical environments, namely, in undamaged dendrimer branches. Dendrimer reduction is linked to the size of the nanoparticle, with greater damage seen in larger NP samples. Simultaneously, preferential colocalization of the gold nanoparticle with the native dendrimer structure increases with the size of the nanoparticle, likely due to the greater segregation of native and reduced structures. This example illustrates the feasibility of chemical characterization within 10 nm of individual ultrasmall nanoparticles, providing information not obtainable by other techniques. Enhanced chemical understanding could lead to increased insight into particle synthesis, in turn facilitating directed growth of nanoparticles within a system by tuning the amount of reductant introduced during synthesis.

This work was supported by the National Science Foundation Grant No. CHE-1308312.