Article pubs.acs.org/ac
Visualizing Nanoparticle Dissolution by Imaging Mass Spectrometry Christopher Szakal,* Melissa S. Ugelow, Justin M. Gorham, Andrew R. Konicek, and R. David Holbrook Materials Measurement Science Division, National Institute of Standards and Technology, 100 Bureau Drive, Gaithersburg, Maryland 20899-8371, United States S Supporting Information *
ABSTRACT: We demonstrate the ability to visualize nanoparticle dissolution while simultaneously providing chemical signatures that differentiate between citrate-capped silver nanoparticles (AgNPs), AgNPs forced into dissolution via exposure to UV radiation, silver nitrate (AgNO3), and AgNO3 /citrate deposited from aqueous solutions and suspensions. We utilize recently developed inkjet printing (IJP) protocols to deposit the different solutions/suspensions as NP aggregates and soluble species, which separate onto surfaces in situ, and collect mass spectral imaging data via timeof-flight secondary ion mass spectrometry (TOF-SIMS). Resulting 2D Ag+ chemical images provide the ability to distinguish between the different Ag-containing starting materials and, when coupled with mass spectral peak ratios, provide information-rich data sets for quick and reproducible visualization of NP-based aqueous constituents. When compared to other measurements aimed at studying NP dissolution, the IJP-TOF-SIMS approach offers valuable information that can potentially help in understanding the complex equilibria in NP-containing solutions and suspensions, including NP dissolution kinetics and extent of overall dissolution.
D
distinct physical particle characteristics. Therefore, subtle chemical changes that influence particle morphology or chemical artifacts introduced by sample preparation/analysis methods may not be apparent by these characterization techniques. Despite the demand for NP size and surface morphology characterization via convincing image analysis,19,20 the need for better characterization of NP dissolution has not benefited from chemical imaging, which is likely caused by a lack of available techniques and/or rigorous, quantifiable information. The lack of available methods for NP chemical characterization is especially pronounced when attempting to either quantitatively or qualitatively identify the chemical form of the soluble species. (For example, does the AgNP dissociate completely to Ag+ or is the Ag part of a soluble complex species?) Additionally, visible alterations in physical particle characteristics may be a secondary consequence if the primary interest is, for example, the production of soluble ionic species (i.e., NP dissolution). If chemical images can be used to differentiate between dissolved solution constituents and NPs, the results could provide convincing chemical characterization to enhance current NP analysis methods. The purpose of this work is to study the Ag-based dissolution from citrate-capped AgNPs and acquire insights into the aqueous behavior of dissolved Ag. As such, we utilized newly combined inkjet-printing (IJP)- and time-of-flight secondary ion mass spectrometry (TOF-SIMS)-based protocols21 for
issolution is an important phenomenon for many engineered nanomaterials (ENMs) and is of specific interest for silver nanoparticles (AgNPs).1,2 The inherent biocidal properties of silver ions, the ability to synthesize AgNPs with unique morphologies and surface coatings, and the somewhat tunable silver ion release rates have led to AgNPs being incorporated into textiles, plastics, paints, and sprays.3−7 AgNP dissolution in aqueous suspensions is typically assessed through either chemical or physical measurements. The most commonly employed chemical measurement technique is to quantify the Ag concentration in a “nanoparticulate-free” solution,1,8,9 whereas physical measurement methods may involve thorough dimensional analysis of NP surfaces.10 Techniques such as single particle inductively coupled plasma mass spectrometry can measure both soluble and nanoparticulate phases,11 thereby providing both chemical and physical information, but can require substantial dilution of the test media and therefore may suffer from artifact introduction.12 The need to simultaneously obtain both chemical and physical information of AgNP solutions without significant sampling manipulation procedures is substantial. From thin films to synthesizing hybrid nanoparticles, valuable insight into the structure, function, and behavior of NPs has been critical for their proper design and implementation.13,14 Correspondingly, advances in nanotechnology have flourished because of the development and implementation of advanced imaging techniques.15−18 Imaging-based NP characterization is readily obtained by, for example, scanning and transmission electron microscopy and atomic force microscopy. Common to these imaging techniques is the reliance on directly measuring This article not subject to U.S. Copyright. Published 2014 by the American Chemical Society
Received: December 20, 2013 Accepted: March 10, 2014 Published: March 10, 2014 3517
dx.doi.org/10.1021/ac404153q | Anal. Chem. 2014, 86, 3517−3524
Analytical Chemistry
Article
radiation as previously described.22 The diluted AgNP suspension was split into two aliquots: 27 mL used for 5 d of UV radiation (“forced dissolution”) and 34 mL used for 5 d of dark storage. A third sample was diluted from the stock to the same concentration as above (6.56 mgAg/L) immediately before inkjet printing (“as-synthesized”). UV−vis spectroscopy was performed on a Shimadzu (Columbia, MD) UV-1800 spectrophotometer which had been calibrated using NIST SRM 203425 and NIST SRM 2031.26 Spectra were acquired at 1 nm/s over a range of 600 nm with one sweep per sample for both the stock suspension and the forced dissolution suspension. DLS analysis was performed on a Zetasizer Nano (Malvern Instruments, Westborough, MA) in 173° backscatter mode at room temperature. The intensity-based distributions for the assynthesized and UV-exposed (forced dissolution) suspensions are representative of the z-average ± 1 standard deviation for 6 and 3 separate measurements, respectively. UV−vis and DLS data are provided in Figure S-1 in the Supporting Information. Inkjet Printing (IJP). Detailed IJP parameters are described elsewhere, including information on the use of the substrate system.21 Briefly, 5 mm × 5 mm × 0.5 mm silicon (111) wafers (Virginia Semiconductor, Fredericksburg, VA) were ethanolcleaned and dried under nitrogen gas. Precision drop-ondemand IJP was performed with a modified MicroFab JetLab 4 printer (MicroFab Technologies, Inc., Plano, TX) for the four types of suspensions/solutions: as-synthesized, forced dissolution, AgNO3, and AgNO3/citrate. Samples were printed into (3 × 3) arrays at 1 mm pitch, where each array element consisted of 10 repeat depositions of 100 drops (with 30 s pauses between each deposition to allow for full drying) to produce a total of 1000 drops (approximately 50 nL total deposition) per array element. For a given starting material, the (3 × 3) array is printed in less than 1 h (including programmed wait times), with each array element drying on the order of seconds or less. Other session-optimized IJP parameters were within the value ranges reported elsewhere for piezo excitation voltages, ejection frequencies, drop dwell times, rise times, fall times, and back pressure.21 TOF-SIMS. Mass spectral data acquisition parameters are described elsewhere.21 Briefly, raw data files (both spectra and images) were acquired with an IonTOF IV TOF-SIMS (IonTof USA, Inc., Chestnut Ridge, NY) instrument, with operating pressures in the low 10−6 Pa (10−8 Torr) range. Bi3+ primary ions at 25 keV were directed at the inkjet-printed drop arrays in 7 ns pulse widths at 10 kHz repetition rates (approximately 0.04 pA of pulsed current). Digital rasters produced (256 × 256) pixel images at 20 pulses/pixel of 500 μm × 500 μm areas to interrogate entire inkjet drops and 50 μm × 50 μm areas for mass spectral peak ratio data, with each image pixel made up of one full spectral scan. The total analysis time from starting material solution/suspension, through inkjet printing, and culminating with chemical image data (including replicates) is on the order of a few hours. Raw and Simulated Line Scan Data. Raw data line scans were completed with the TOF-SIMS instrument software for 107 Ag+ mass-specific images. Individual images were 16-pixel binned, and lines were drawn across the inkjet depositions, as indicated in Supporting Information Figures S-2 and S-3. Simulated line scans were generated in MATLAB (version R2011a, Mathworks, Natick, MA) and Origin 9 (OriginLab, Northampton, MA). The Gaussian relationship [y = exp(−1/2 × (α × n/(N/2))2)] was used within MATLAB with varying
obtaining image data aimed at evaluating NP dissolution and related chemical phenomena. In that work, it was found that IJP could readily produce samples from NP suspensions by depositing the NPs into aggregate rings. Subsequent TOFSIMS spectral data was used to probe differences/similarities in NP surface chemistry. For the Ag-based work presented here, we capitalized on previous work that forced dissolution of the AgNP into soluble Ag species via exposure to UV light22 and combined the aforementioned IJP-TOF-SIMS approach to reproducibly differentiate the chemical changes occurring in AgNP-based suspensions and related solutions. For the inkjetprinted suspensions studied herein, we examine the resulting images and mass spectral peak ratios in the context of previously reported Ag and citrate solution equilibria as well as their potential effects on particulate formation within such aqueous-based environments.
■
EXPERIMENTAL SECTION Terminology for Ag-Based Solution and Suspension Constituents. We refer to Ag0 as-synthesized in the nanoparticle form as “AgNP”. The AgNP suspension is exposed to UV radiation to convert the AgNP into a dissolved form and is referred to throughout the text as “forced dissolution AgNP”. The form of silver in solution is complex, and measuring such differences within solutions and suspensions is one of the main thrusts of this work. Because silver atoms in a dissolved state can result in different chemistries, we use “dissolved Ag” to refer to soluble Ag species that can be either fully dissociated as Ag+ and/or as an Ag complex. We refer to fully dissociated silver as “Ag+” when discussing solution chemistry but also note that mass spectrometric techniques will detect “Ag+” regardless of whether it is fully dissociated, part of a soluble complex, or in the Ag0 form. Nanoparticle (NP) Synthesis and Solution-Based Characterization. Nominal 20 nm diameter, citrate-capped, AgNP suspensions were synthesized using a chemical reduction method similar to previously published work.23 Briefly, 0.0460 g of silver nitrate and 0.0786 g of trisodium citrate (Sigma Aldrich, St. Louis, MO) were added in series to a beaker filled with 0.350 L of 92 °C milli-Q purified water (90% conversion of Ag+ to the nanoparticulate form and was subsequently diluted to a final volume of 0.14 L. The resulting stock AgNP suspension, which was previously found to be stable for the time associated with this study,24 was filtered through a 0.2 μm syringe filter to remove any dust or aggregates and diluted to 6.56 mgAg/L with milli-Q water immediately prior to forced dissolution by UV-exposure. TraceSelect Ultra water (Fluka, St. Louis, MO) was used as a diluent for each stock batch of synthesized NPs as reported previously.21 UV-Based Forced Dissolution and Initial Characterization of Ag Nanoparticle (AgNP) Suspensions. The forced dissolution of AgNPs to dissolved Ag was accomplished by exposing the AgNPs to UV radiation. A Rayonet reactor (Southern New England Ultraviolet Company, Branford, Connecticut) with a 300 nm light source was used for UV 3518
dx.doi.org/10.1021/ac404153q | Anal. Chem. 2014, 86, 3517−3524
Analytical Chemistry
Article
Figure 1. Schematic of inkjet printing (IJP) and TOF-SIMS approach to measuring solution species for NP vs. dissolved states. (a) Sample solutions and/or suspensions of each Ag-containing starting material; (b) each solution/suspension is individually inkjet-printed into droplet arrays as indicated in the text; (c) each droplet is individually analyzed for simultaneous image and mass spectral information by TOF-SIMS.
Figure 2. Replicate SIMS images of both the as-synthesized AgNP suspension (panels a,e,i and b,f,j) and the forced dissolution AgNP suspension (panels c,g,k and d,h,l) for individual inkjet-printed droplets: (a−d) total ion images (sum of all intensities of all detected secondary ions) for each inkjet-printed droplet, (e−h) 107Ag+ chemical map for each inkjet-printed droplet, and (i−l) RGB overlay of 107Ag+ (red), Na2OH+ (green), and substrate 28Si+ (blue). All images are 500 μm × 500 μm. Images e−l uniformly adjusted by +40 contrast and +40 brightness, with overlays k and l having the red channel thresholded from approximately 2% to 50% of maximum signal using the instrument software.
values of α to provide different width Gaussian distributions describing the general Ag SIMS image line scan distributions that can be obtained for different sample types (as opposed to Gaussian fits of actual line scan data). Variations on the Gaussians were performed in Origin 9 spreadsheets for flat signal regions and height differences. Safety Considerations. All pertinent safety considerations related to the inkjet printing of NPs from the suspensions were outlined in a prior publication and were followed for this work.21
dissolution, and in particular dissolved Ag, are plagued by rapid NP transformation,24,27 low concentrations of soluble species in natural environments,28 and complexation by surface capping agents.29,30 By exploiting the innate differences of how NPs and dissolved material interact with surfaces following inkjet printing, we devised a set of experiments to probe whether imaging mass spectrometry (TOF-SIMS) can help elucidate surface chemistry and NP vs. dissolved fraction differences for commonly used citrate-capped AgNPs. In Figure 1, the process from solution or suspension starting material through imaging mass spectrometry is presented. Each starting solution/ suspension, AgNPs (citrate-capped), forced dissolution AgNPs, Ag+ (from AgNO3), and Ag+/citrate (from AgNO3/ Na3C6H5O7), is deposited via IJP to produce arrays of accumulated material as described elsewhere.21 The IJP approach is advantageous because no solution-based sample preparation steps such as acid digestion or centrifugation are
■
RESULTS AND DISCUSSION Studying AgNP Dissolution with Imaging Mass Spectrometry. The incorporation of IJP and TOF-SIMS into the realm of studying NP dissolution was based on previous work for investigating differences in NP surface chemistry.21 The measurement difficulties of studying NP 3519
dx.doi.org/10.1021/ac404153q | Anal. Chem. 2014, 86, 3517−3524
Analytical Chemistry
Article
Figure 3. Replicate SIMS images of both an AgNO3 solution (panels a,e,i and b,f,j) and a AgNO3/citrate solution (panels c,g,k and d,h,l) for individual inkjet-printed droplets: (a−d) total ion images (sum of all intensities of all detected secondary ions) for each inkjet-printed droplet, (e−h) 107 Ag+ chemical map for each inkjet-printed droplet, and (i−l) RGB overlay of 107Ag+ (red), Na2OH+ (green), and substrate 28Si+ (blue). All images are 500 μm × 500 μm. Images e−l uniformly adjusted by +40 contrast and +40 brightness, with overlays k and l having the red channel thresholded from approximately 2% to 50% of maximum signal using the instrument software.
each droplet deposition appears to be in a ring within a ring, where the larger diameter represents the original liquid droplet region, and the smaller diameter represents the final drying area of the NP solution deposition. Although the total ion images are not chemically specific, the IJP-TOF-SIMS approach allows for the collection of each ion channel at a specific m/z to be mapped on a pixel-by-pixel basis. These chemical images can be used to examine the extent to which consistent IJP deposition leads to surface chemistry-specific distributions dictated by the original Ag species in the starting solution/suspension. The ionspecific 107Ag+ signals (used here as a chemical indicator of Ag location within the image fields, regardless of chemical speciation) in Figure 2e,f form a ring of deposited NP aggregates on the outside portion of the final droplet drying diameter (inner ring in total ion images). In Figure 2i,j, an RGB overlay of the Ag+ (red), Na2OH+ (green, and an intense signal indicator of the ionic solution constituents), and substrate silicon (blue) shows the separation of the NPs from the other mostly ionic and water-soluble suspension species. The AgNPs, which did not readily dissociate into the printed aqueous suspensions, remain on the outside of the drying water droplet while more ionic and water-soluble species travel to the final drying location of the inkjet drops. Since the fresh citratecapped AgNP suspensions depicted in the Figure 2 panel data a,e,i and b,f,j were not exposed to any stimuli that would produce dissolved Ag from the AgNPs, the ring-based deposition patterns are immediately recognizable as stemming from a NP-containing starting material (an assertion made from this data as well as that from previous work).21 SIMS total ion images of the forced dissolution AgNP suspension deposition in Figure 2c,d display one immediate difference from those in Figure 2a,b: the smaller diameter ring feature is not present, suggesting that the majority of silver is no longer in AgNP form. Further inspection of the 107Ag+ ionspecific images in Figure 2g,h reveals that the Ag+ signals are
required. It is known that such solution-based sample preparation can alter NP surface chemistry and hence soluble species formation,9,31 with the alternative here being a change in physical state of the medium from liquid phase to solid phase rather than inducing chemical changes. Because of the fast drying of nL volumes of each deposited solution or suspension, the impact of solution-based chemical transformations on drying patterns is minimized while interactions between the solutions/suspensions and substrate are maximized. For example, our previous work using these procedures showed that citrate-capped TiO2 NPs remained on the outside of the printed droplets and deposit in a ring, whereas ionic species within the aqueous suspensions deposited in the center of the droplets. Ionic species have a higher association with water compared to the NPs and are therefore transported with the water to the final drying spot, giving rise to a form of in situ chromatography directly on the substrate. The postulate for the current experiments is to exploit this ionic behavior to discern if Ag-based starting solutions/suspensions contain AgNPs or dissolved Ag. Once deposited, substrates are imaged by TOFSIMS with a focus on mapping Ag distributions across the depositions in order to quickly assign the class of Ag-containing starting material, including attribution of Ag0-based species vs. dissolved forms. The combination of IJP and TOF-SIMS allows for this study of solution-based chemistry differences in a way that would not be possible with TOF-SIMS alone. Visualizing AgNP vs. Dissolved Ag Differences. Figure 2 displays chemical images of inkjet-printed droplets for the assynthesized citrate-capped AgNP suspension (panels a,e,i and b,f,j) and the forced dissolution AgNP suspension (panels c,g,k and d,h,l). In the top row, Figure 2a−d, the SIMS total ion images (which represent the sum of all detected ions in the mass spectrum at each pixel) are displayed for two of the droplet depositions from both the as-synthesized and forced dissolution arrays. For the as-synthesized data (Figure 2a,b), 3520
dx.doi.org/10.1021/ac404153q | Anal. Chem. 2014, 86, 3517−3524
Analytical Chemistry
Article
Figure 4. Simulated representations of SIMS Ag intensity line scan data across inkjet-printed depositions for AgNO3, AgNO3 and sodium citrate, AgNP, and forced dissolution AgNP. y-axis scales are presented as arbitrary units from 0 to 1, and x-axis scales are presented as arbitrary units from −80 to 80 (for line scan distances from center of inkjet-printed deposition). Sample raw data line scans from Figures 2 and 3 are presented in the insets for comparison as follows: (a) AgNO3 with inset from Figure 3f, (b) AgNO3/citrate with inset from Figure 3g, (c) AgNP with inset from Figure 2e, and (d) forced dissolution AgNP with inset from Figure 2h. Additional raw data line scans are included in the Supporting Information Figures S-2 and S-3.
are they similar to the distributions of forced dissolution AgNP. Instead, all silver signals are localized with the Na2OH+, suggesting fully water-soluble ionic character, sharing the same final drying location as the water droplets. Also visible are satellite drops that can occasionally form during inkjet printing, in addition to the main droplet pathway. The forced dissolution AgNP data (from Figure 2g,h and Figure 2k,l) show an appreciable amount of 107Ag+ codeposited with the Na2OH+ but also a distinct silver deposition pattern throughout the entirety of the larger diameter droplet deposition. While AgNP dissolution is indeed occurring after UV exposure as expected, the Figure 2 image data reveal that the system is not purely dissociated Ag+, as seen with AgNO3 in Figure 3. The addition of 25 mg/L of sodium citrate to the 10 mg/L AgNO3 solution was meant to more realistically mimic the solution environment of the citrate-capped forced dissolution AgNPs. Total-ion SIMS images in Figure 3c,d of the AgNO3/ citrate inkjet-printed solution reveal deposition patterns that appear like a combination of those for the as-synthesized and UV-exposed (forced dissolution) images in Figure 2, particularly the similar noncircular and amorphous deposition shapes of the final droplet drying locations. Some spreading of the 107Ag+ signal similar to the citrate-capped forced dissolution AgNPs is exhibited in Figure 3g,h and, although not as pronounced as in the Figure 2g,h images, is still suggestive of dissolved but not purely dissociated Ag-containing species. We envision the ability to monitor images of a solution/ suspension of potential NP-containing starting materials and use the IJP-TOF-SIMS approach to differentiate classes of chemistry occurring in solution as shown in Figures 2 and 3. Rather than relying on images alone, line scans of the Ag+ signals can be used to generate data shapes that serve as an alternative means of differentiating between the varying solution/suspension chemistries. A simulated set of summary
deposited across the entire droplet diameter with slightly higher signal intensities in the center of the final droplet drying location. Due to the preparation method, this deposition pattern without the NP ring can only occur if the Ag-based starting material has dissolved or become ionic in nature, as similarly reported for TiO2 NPs.21 Fully dissociated Ag+, however, would be expected to localize with other watersoluble ionic constituents to the final drying diameter as shown in the green channel of Figure 2i,j. Since the images display 107 Ag+ signals throughout the wider deposition diameter, the dissolved Ag species can be either of partial ionic character (such as slightly soluble Ag2O) or of a soluble mixed inorganic/ organic species. Additionally, the RGB overlays in Figure 2k,l are in sharp contrast with those in Figure 2i,j, in that the forced dissolution AgNP suspension material does not form a distinct ring with respect to the final droplet drying location marked by the Na2OH+ but rather is loosely distributed around the entire droplet diameter. To our knowledge, this is the first time that NP dissolution has been chemically identified and imaged, where differences between NP-based and soluble signatures of the same nominal material by chemical image inspection can be differentiated. However, we wanted to verify the IJP and TOF-SIMS ability to differentiate the solution/suspension behavior of dissolved species and compare to the forced dissolution AgNP data in Figure 2. As such, SIMS images were acquired for fully dissociated Ag+ from an AgNO3 solution (Figure 3 panels a,e,i and b,f,j) as well as the same AgNO3 solution with added citrate (Figure 3 panels c,g,k and d,h,l). In Figure 3a,b, the total ion images are not immediately informative for determining chemical similarities/differences for either AgNP or dissolved Ag from NP dissolution data. However, the chemically specific images in Figure 3e,f and Figure 3i,j show 107Ag+ signals that do not form deposition rings like would be expected of AgNPs nor 3521
dx.doi.org/10.1021/ac404153q | Anal. Chem. 2014, 86, 3517−3524
Analytical Chemistry
Article
synthesized AgNPs. The solution-based equilibria that create the IJP-based surface distributions either can lead to the Ag+ chemical maps by the presence of dissolved Ag−ligand complexes (with citrate, breakdown products of citrate, or other organic ligands) and/or could be from destabilized NPs that agglomerate and no longer have the size range and surface chemistry to deposit in a ring formation. Similarly, the AgNO3/ citrate images in Figure 3g,h show the solution complexity by likely being representative of solution chemistry at the time of inkjet printing that was either starting to allow nanoparticle formation (visible by a slight ring in Figure 3h) or producing Ag−ligand dissolvable complexes, or both. The likelihood is a combination of these effects is possible in systems studied herein and more careful experimentation would be needed along with data from other techniques to fully understand mechanistic details, including inkjet-printing and TOF-SIMS imaging at specific time points in solution/suspension to provide potential kinetic data. When trying to determine the dissolution behavior of NPs, more information is always helpful, particularly if chemical signatures are acquired. Since a full mass spectrum is recorded at each image pixel, we can also study the differences in the peak ratios of specific spectral signatures to enhance the ability to measure NP dissolution. As described in previous work,21 the data sampling across (500 × 500) μm images is not high enough to monitor subtle chemical changes in the mass spectra, so (50 × 50) μm images with the same number of pixels were acquired near each droplet deposition edge, providing a 100× increase in sampling density. Figure 5 shows Ag+/X+ peak area
line scan data shapes for SIMS Ag intensity signals across inkjet-printed depositions is presented in Figure 4a−d based on actual raw data line scans (Figure 4 insets and Supporting Information Figures S-2 and S-3). As indicated in Figure 4, the line scan shapes for each Ag-based starting material are as follows: Figure 4a, maximum Ag+ signal in the center of the droplet deposition and approaching zero signal intensity toward the outside of the deposition area (AgNO3); Figure 4b, lower maximum Ag+ signal in the center of the deposition with a small signal level across the entire droplet deposition (AgNO3/ citrate); Figure 4c, droplet centers mostly void of Ag signal with two local signal maxima (AgNP); and Figure 4d, a small maximum Ag+ signal in the center of the droplet deposition with a pronounced elevated background signal level across the entire droplet deposition (forced dissolution AgNP). The Figure 4d forced dissolution line scans are combinations of those in Figure 4a (highest signal in center area) and Figure 4b (nonzero Ag intensity across deposition area) while being entirely different than Figure 4c. Such differentiation allows for rapid identification of whether AgNPs stay as NPs in solution/ suspension or dissolve into their surrounding matrix. With the ability to differentiate between different Agcontaining solution/suspension starting materials being presented, one may naturally inquire about the chemistry that can lead to such inkjet-printed depositions, as well as Ag+ chemical maps and line scan data. A more detailed look at the published literature associated with AgNPs, dissolved Ag, and citrate can help provide insight into the complex equilibria that are occurring in these systems and that the IJP-TOF-SIMS approach can be informative for helping to understand the chemistry occurring in such solutions/suspensions. AgNP dissolution often implies that Ag0 has oxidized to Ag+ and that the free silver ion exists in solution. However, additional solution constituents, namely, the often-used stabilizing citrate molecule, can alter the aqueous chemical identity of the dissolved silver. For example, the stabilizing efficacy of the citrate molecule depends on precise concentrations of the molecule in the solution/suspension. If the citrate concentration is too low, NP size due to agglomeration can increase whereas too high of a citrate concentration can lead to high levels of destabilized particles that can actually grow via different solution mechanisms.32 Several theories across many decades have been developed to explain the particle formation behavior in solution. Such nucleation theory comparisons between those such as Lamer (self-nucleation) and Turkevich (the organizer model where molecules like citrate can help induce nanoparticle formation of metal atoms merely by being present) reveal that exact solution chemistries leave largely unpredictable behavior where a wide range of equilibria can occur.33 Among those are the idea that Ag−molecule complexes can aid in particle growth (particularly at low citrate concentration) via Agx(citrate) + Agx(citrate) yielding Ag2x(citrate),32 that metal ion nucleation can occur via association with ligand molecules to the form MxLyn+,33 and that high concentrations of excess citrate can produce dissolved Ag-citrate ligands of the form of [Ag3(C6H5O7)n+1]3n−.34 The reality is that any or all of these possibilities, among others, may be observed in Figures 2 and 3. However, the plurality of evidence suggests that forced dissolution of Ag following UV exposure does occur and is likely to be at least partly due to breakdown of the stabilizing citrate molecule.22 Even if a low amount of citrate molecule is left in the suspension after forced dissolution of AgNPs, the Figure 2g,h images represent this new system state relative to the as-
Figure 5. Plot of Ag+/X+ mass spectral peak area ratios as a function of the as-synthesized, forced dissolution, AgNO3, and AgNO3/citrate solutions/suspensions. Each column grouping is for a different peak ratio as indicated in the axis labels. Error bars represent n = 11 measurements for the AgNP-based data and n = 3 measurements for the AgNO3-based data. If no error bars are present, only one analysis contained enough denominator counts for a ratio to be calculated.
ratios for each Ag-containing starting material as calculated using methods described elsewhere,21 for small hydrocarbons as well as the Ag3 species. For the Ag+/CH3+ and Ag+/C2H3+ peak ratios, one general observation is that the values are highest by a significant margin for the AgNO3 solution and next highest for the deposited as-synthesized AgNP solutions. This is expected, as the amount of organic material in solution that codeposits with the Ag from AgNO3 solutions is expected to be low while the AgNPs should still be associated with some level of citrate (lowering the ratio some). Additionally, by far the lowest Ag+ 3522
dx.doi.org/10.1021/ac404153q | Anal. Chem. 2014, 86, 3517−3524
Analytical Chemistry
Article
chemical changes from dissolution kinetics. Additional SIMSbased studies are also aimed at determining the chemical signature differences between the surfaces and bulk of NP aggregates, as well as solution-based time and storage changes to NP chemistry.
hydrocarbon ratios are measured in the forced dissolution case, where small fragment ions like the ones measured here are expected to be high in the mass spectra because of UV-induced citrate molecule degradation along with the widespread Ag+ images showing low concentration of silver in any one area. The AgNO3/citrate solution, as described earlier, is largely a chemical hybrid of the other deposited solutions/suspensions, and because those solution equilibria have multiple chemical pathways, it is no surprise that the peak ratios exhibit mixed behavior with respect to the other starting materials, especially for the Ag+/C3H5+ data. The other interesting peak ratio data in Figure 5 involves the Ag+/Ag3+ peak ratios. Ag cluster ions have previously been useful target ions for studying secondary ion emissions in SIMS.35 The Ag3+ can only be removed, ionized, and detected if there are enough nearest neighbor atoms to support the ion formation from either direct sample emission or gas phase recombination above the sample plane following SIMS ion beam bombardment. The Ag3+ formation is both related to the bond energies of silver atoms and ions to other species as well as the local concentrations of Ag atoms with respect to other surrounding Ag atoms. On the other hand, detected Ag+ is indicative of both dissolution (either preformed or ionically bonded silver atoms) and SIMS removal and subsequent ionization of prevalent Ag material (regardless of number of nearest Ag atom neighbors). Therefore, an increase in the Ag+/ Ag3+ peak area ratio could indicate more dissolved Ag than NPbased Ag0 since fewer nearest neighbors of core Ag0 would make it dramatically less likely for the Ag3+ to be either removed or ionized and subsequently detected. The plot in Figure 5 shows that the inkjet-printed forced dissolution AgNP suspensions have substantially higher Ag+/Ag3+ ratios (by approximately a factor of 2) than the corresponding assynthesized AgNP, AgNO3, and AgNO3/citrate solutions. It would seem that, despite the myriad of different chemical equilibria involving Ag, AgNP formation, and stability, as well as citrate concentrations mentioned earlier, something unique is occurring after forced dissolution of the AgNPs following UV radiation. The ability to combine substrate-based separation chemistry from inkjet printing and differentiate the metal ion distributions via SIMS imaging and peak ratio data for subtly different solutions and suspensions suggests the promise that the approach described herein can be useful in NP dissolution studies for other systems.
■
ASSOCIATED CONTENT
S Supporting Information *
Figures and text related to nanoparticle solution characterization and raw line scan data related to Figures 2−4. This information is available free of charge via the Internet at http:// pubs.acs.org/.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Ph: (301) 975-3816. Fax: (301) 417-1321. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We thank Dr. M. Verkouteren for his assistance and advice in optimizing the inkjet printing component of this work, Dr. J. Gigault for ICPMS Ag concentration data, and Dr. A. Herzing for assistance in producing the simulated line scan data. Certain commercial equipment, instruments, or materials are identified to adequately specify the experimental procedure. Such identification does not imply recommendation or endorsement of the National Institute of Standards and Technology, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose.
■
REFERENCES
(1) Liu, J. Y.; Hurt, R. H. Environ. Sci. Technol. 2010, 44 (6), 2169− 2175. (2) Marambio-Jones, C.; Hoek, E. M. V. J. Nanopart. Res. 2010, 12 (5), 1531−1551. (3) Elzey, S.; Grassian, V. H. J. Nanopart. Res. 2010, 12 (5), 1945− 1958. (4) Fabrega, J.; Fawcett, S. R.; Renshaw, J. C.; Lead, J. R. Environ. Sci. Technol. 2009, 43 (19), 7285−7290. (5) Gottschalk, F.; Nowack, B. J. Environ. Monit. 2011, 13 (5), 1145− 1155. (6) Liu, J. Y.; Sonshine, D. A.; Shervani, S.; Hurt, R. H. ACS Nano 2010, 4 (11), 6903−6913. (7) Zhang, F.; Wu, X. L.; Chen, Y. Y.; Lin, H. Fibers Polym. 2009, 10 (4), 496−501. (8) Benn, T. M.; Westerhoff, P. Environ. Sci. Technol. 2008, 42 (11), 4133−4139. (9) Kittler, S.; Greulich, C.; Diendorf, J.; Koller, M.; Epple, M. Chem. Mater. 2010, 22 (16), 4548−4554. (10) Kent, R. D.; Vikesland, P. J. Environ. Sci. Technol. 2012, 46 (13), 6977−6984. (11) Mitrano, D. M.; Lesher, E. K.; Bednar, A.; Monserud, J.; Higgins, C. P.; Ranville, J. F. Environ. Toxicol. Chem. 2012, 31 (1), 115−121. (12) Ulrich, A.; Losert, S.; Bendixen, N.; Al-Kattan, A.; Hagendorfer, H.; Nowack, B.; Adlhart, C.; Ebert, J.; Lattuada, M.; Hungerbuhler, K. J. Anal. At. Spectrom. 2012, 27 (7), 1120−1130. (13) Champion, J. A.; Katare, Y. K.; Mitragotri, S. Proc. Natl. Acad. Sci. 2007, 104 (29), 11901−11904. (14) Shang, L.; Dong, S. J.; Nienhaus, G. U. Nano Today 2011, 6 (4), 401−418.
■
CONCLUSIONS We presented the ability to visualize the effects of silver nanoparticle dissolution from Ag0 to a forced dissolution state and to differentiate the subtle changes in solution/suspension deposition chemistry via imaging mass spectrometry coupled with inkjet printing technology. Despite the complexities of chemical equilibria associated with solutions and suspensions containing Ag, AgNP, and/or stabilizing citrate molecules, mass spectral images and peak ratio signatures can provide insight into the different chemistries occurring in such liquid samples that are rapidly converted to the solid phase at specific time points. The work outlined in this study can easily be extended to study the onset of NP dissolution for a variety of metal and metal oxide-based NP systems versus variables such as time and storage conditions. The combined chemical imaging of mass spectrometry with the robust sample preparation of inkjet printing can be used as powerful tools to study more complicated NP solution chemistry questions, such as subtle 3523
dx.doi.org/10.1021/ac404153q | Anal. Chem. 2014, 86, 3517−3524
Analytical Chemistry
Article
(15) Helveg, S.; Lopez-Cartes, C.; Sehested, J.; Hansen, P. L.; Clausen, B. S.; Rostrup-Nielsen, J. R.; Abild-Pedersen, F.; Norskov, J. K. Nature 2004, 427 (6973), 426−429. (16) Midgley, P. A.; Weyland, M. Ultramicroscopy 2003, 96 (3−4), 413−431. (17) Thomas, J. M.; Gal, P. L. Electron microscopy and the materials chemistry of solid catalysts; Elsevier Academic Press Inc: San Diego, 2004; Vol. 48; pp 171−227. (18) Utsunomiya, S.; Ewing, R. C. Environ. Sci. Technol. 2003, 37 (4), 786−791. (19) Pettitt, M. E.; Lead, J. R. Environ. Int. 2013, 52, 41−50. (20) Powers, K. W.; Brown, S. C.; Krishna, V. B.; Wasdo, S. C.; Moudgil, B. M.; Roberts, S. M. Toxicol. Sci. 2006, 90 (2), 296−303. (21) Szakal, C.; McCarthy, J. A.; Ugelow, M. S.; Konicek, A. R.; Louis, K.; Yezer, B.; Herzing, A. A.; Hamers, R. J.; Holbrook, R. D. J. Environ. Monit. 2012, 14 (7), 1914−1925. (22) Gorham, J. M.; MacCuspie, R. I.; Klein, K. L.; Fairbrother, D. H.; Holbrook, R. D. J. Nanopart. Res. 2012, 14, 10. (23) MacCuspie, R. I. J. Nanopart. Res. 2011, 13 (7), 2893−2908. (24) Gorham, J. M.; Rohlfing, A. B.; Lippa, K. A.; MacCuspie, R. I.; Hemmatti, A.; Holbrook, R. D. J. Nanopart. Res. 2014, DOI: 10.1007/ s11051-014-2339-9. (25) Weidner, V. R.; et al. Holmium oxide solution: wavelength standard from 240 to 640 nm--SRM 2034; U.S. Dept. of Commerce, U.S. Govt. Print. Off: Gaithersburg, MD, Washington DC, 1986. (26) Mavrodineanu, R.; Baldwin, J. R. Metal-on-quartz filters as a standard reference material for spectrophotometry--SRM 2031; U.S. Govt Print. Off.: Gaithersburg, MD, Washington DC, 1980. (27) Glover, R. D.; Miller, J. M.; Hutchison, J. E. ACS Nano 2011, 5 (11), 8950−8957. (28) Peters, A.; Simpson, P.; Merrington, G.; Rothenbacher, K.; Sturdy, L. Bull. Environ. Contam. Toxicol. 2011, 86 (6), 637−641. (29) Liu, J. Y.; Sonshine, D. A.; Shervani, S.; Hurt, R. H. ACS Nano 2010, 4 (11), 6903−6913. (30) Pillai, Z. S.; Kamat, P. V. J. Phys. Chem. B 2004, 108 (3), 945− 951. (31) He, D.; Bligh, M. W.; Waite, T. D. Environ. Sci. Technol. 2013, 47 (16), 9148−9156. (32) Henglein, A.; Giersig, M. J. Phys. Chem. B 1999, 103 (44), 9533−9539. (33) Finney, E. E.; Finke, R. G. J. Colloid Interface Sci. 2008, 317 (2), 351−374. (34) Djokic, S. Bioinorg. Chem. Appl. 2008, DOI: 10.1155/2008/ 436458. (35) Sun, S. X.; Szakal, C.; Winograd, N.; Wucher, A. J. Am. Soc. Mass Spectrom. 2005, 16 (10), 1677−1686.
3524
dx.doi.org/10.1021/ac404153q | Anal. Chem. 2014, 86, 3517−3524
