Influence of Ligand Shell Composition upon Interparticle Interactions

Oct 25, 2015 - ... and solution small-angle X-ray scattering provided additional insight ... Aurora L. GinzburgLisa TruongRobert L. TanguayJames E. Hu...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/Langmuir

Influence of Ligand Shell Composition upon Interparticle Interactions in Multifunctional Nanoparticles Zachary C. Kennedy, Carmen E. Lisowski, Dumitru S. Mitaru-Berceanu, and James E. Hutchison* Department of Chemistry and Biochemistry, 1253 University of Oregon, Eugene, Oregon 97403, United States S Supporting Information *

ABSTRACT: The interactions of nanoparticles with biomolecules, surfaces, or other nanostructures are dictated by the nanoparticle’s surface chemistry. Thus, far, shortcomings of syntheses of nanoparticles with defined ligand shell architectures have limited our ability to understand how changes in their surface composition influence reactivity and assembly. We report new synthetic approaches to systematically control the number (polyvalency), length, and steric interactions of omega-functionalized (targeting) ligands within an otherwise passivating (diluent) ligand shell. A mesofluidic reactor was used to prepare nanoparticles with the same core diameter for each of the designed ligand architectures. When the targeting ligands are malonamide groups, the nanoparticles assemble via cross-linking in the presence of trivalent lanthanides. We examined the influence of ligand composition on assembly by monitoring the differences in optical properties of the crosslinked and free nanoparticles. Infrared spectroscopy, electron microscopy, and solution small-angle X-ray scattering provided additional insight into the assembly behavior. Lower (less than 33%) malonamide ligand densities (where the binding group extends beyond the periphery of diluent ethylene glycol ligands) produce the strongest optical responses and largest assemblies. Surprisingly, nanoparticles containing a higher surface number of targeting ligand did not produce an optical response or assemble, underscoring the importance of an informed mixed ligand strategy for highest nanoparticle performance.

■

INTRODUCTION Hybrid nanoparticles (NPs) that present specific functional groups on their exterior are being widely investigated as therapeutics,1 drug delivery platforms,2 chemical and biological sensors, catalysts, and active agents in separations.3 Such nanomaterials often incorporate multiple surface chemistries that each impart specific functions to the material. For example, drugs have been appended to NPs containing additional functionalities that target specific tissues or sites,4 protect the particle from metabolic degradation,5 and enhance imaging of the particle in situ.6 Through careful modification of the surface, it has been possible to alter pharmacokinetic profiles, increase therapeutic efficiency, and ultimately reduce cytotoxicity.7 The use of NPs with multiple functionalities has been investigated for a wide range of nanomaterials including magnetic,8 quantum dots,9 noble metal and metal oxides nanoparticles,10,11 polymers,12 dendrimers,13 and silica.14 The benefits and current challenges associated with this approach with respect to drug development were outlined recently by Cheng et al.15 A barrier to the design and synthesis of NPs with multifunctional ligand shells is the integration of multiple functional groups in a convenient, scalable, and cost-effective manner. The key challenges include (i) gaining adequate control of the core material size and morphology while simultaneously introducing the desired types and numbers of ligands on the surface, (ii) reliably and predictably controlling © XXXX American Chemical Society

the ratio of ligands within the shell and (iii) developing characterization strategies to assess shell compositions and guide nanoparticle probe development. Surface functional groups are typically introduced by grafting to the outside of the NP, performing ligand exchange, or directly introducing the groups during NP synthesis.16,17 It is often difficult to control and measure the number of groups incorporated. These challenges are exacerbated by batch-to-batch variation and the reality that typical syntheses yield polydisperse distributions of core sizes which require laborious purification strategies. In addition, many syntheses of multifunctional NPs yield ligand shells that have heterogeneous distributions of the ligands.18 The heterogeneity resulting from multifunctional NP synthesis has been explored by separating subpopulations of multifunctional NPs by ligand number using HPLC,19 but assessing the individual populations was limited due to the small quantity of material isolated. The consequences of different numbers of targeting ligands (which defines the polyvalency) have received little attention. Surface plasmon resonance (SPR) has been used to study binding avidities of ssDNA to complementary sequences on a model dendrimer system.20 In a second example, a specialized bionconjugation strategy was used to vary the number of active Received: August 20, 2015 Revised: October 22, 2015

A

DOI: 10.1021/acs.langmuir.5b03096 Langmuir XXXX, XXX, XXX−XXX

Langmuir

■

Article

METHODS Materials. All materials were used as received unless indicated below: HAuCl4·H2O (99.9%) (Strem); 2-[2-(2chloroethoxy)-ethoxy]ethanol (99%), 3,4-dihydro-2H-pyran (97%), p-toluenesulfonic acid monohydrate (98.5%), phosphorus tribromide (99%), sodium thiosulfate pentahydrate (99.5%), sodium borohydride (98%, caplets), sodium hydride (95%, dry), 4-toluenesulfonyl chloride: purified and recrystallized36 (Aldrich); sodium hydroxide, sodium iodide (Mallinckrodt); and N 1 ,N 1 ,N 3 ,N 3 -tetramethylmalonamide (97%), N1,N1,N3,N3-tetraethylmalonamide (97%) (TCI America). The Bunte salt analogue of 2-(7-mercaptoheptyl)N1,N1,N3,N3-tetramethylmalonamide (3) was synthesized as previously reported.37 The Bunte salt analogue of 2-[2-(2mercaptoethoxy)-ethoxy)ethanol (MEEE) was synthesized as previously reported.38 Column chromatography was performed using 40−63 μM silia-P flash silica gel (Silicycle). Deionized water (18.2 MΩ·cm) was obtained using a Barnstead Nanopure Diamond system. Flow nanoparticle syntheses were driven using Kloehn syringe pumps (P/N 54022) and Kloehn 10 and 25 mL syringes. The flow system was created using IDEX Teflon tubing (0.75 mm ID, WO# 0554152) and Teflon Tmixers. Lengths of tubing were used in assembling the reactor, to keep residence and mixing constant for all flow rates. Synthesis of Mixed Malonamide/Ethylene Glycol Gold Nanoparticles. Mixed monolayer gold nanoparticles were synthesized directly by reduction of HAuCl4 (aq.) in the presence of mixed Bunte salt ligand precursors in a mesofluidic reactor. The mesofluidic reactor setup used was outlined previously, and reactions were performed with minor modifications.39 Aqueous solutions were prepared to enable three successive mesofluidic syntheses at each reaction condition. Thus, 30 mL of 5 mM HAuCl4, 30 mL of 1 mM Bunte salt mixed ligands, and 60 mL of 1 mM NaBH4 were prepared. The NaBH4 solution was prepared just prior to synthesis, as this reagent also undergoes undesirable hydrolysis in water. In addition, 1.2 mL of 1 M NaOH was added to the NaBH4 solution to further slow the rate of NaBH4 hydrolysis. For subsequent characterization and cross-linking experiments, free ligands and unreacted starting materials were removed from nanoparticles produced using diafiltration.40 Mixed ligand AuNP products with functional ligands 1, 3, and 4 were diafiltered with 60 volume equivalents of nanopure H2O (typically 1.8 L) using a 70 kDa membrane (Pall Corporation), whereas AuNPs with malonamide 2 were diafiltered similarly using a 10 kDa membrane. NMR Spectroscopy of Purified Nanoparticle Samples for Analysis of Purity and of Decomposed Nanoparticles for Quantifying Mixed Ligand Compositions. Approximately 50% of the purified nanoparticle solution was lyophilized and redispersed in 0.5 mL deuterated solvent. An initial spectrum was acquired at 500 MHz with 64 scans and a relaxation delay of 1 s to confirm that all free ligands and synthetic byproducts were removed. The absence of sharp peaks (due to free ligands) and the presence of characteristic broad peaks indicated that all of the ligands were bound to the AuNP surface. Quantification of bound mixed ligands was initiated by adding approximately 2 mg of I2 to the NMR tube directly. The mixture was shaken vigorously and allowed to react in ambient conditions for ∼10 min. All spectra showed that the ligands had been oxidized to form the corresponding disulfides. Integration of peaks attributed to malonamide and of

ligands on superparamagnetic iron oxide NPs to influence cellular targeting.21 Combined, these studies suggest that binding avidity may be enhanced at intermediate ligand polyvalency. Still, important fundamental questions remain regarding how nanoparticle binding and assembly is impacted by the number of active ligands and the molecular architecture surrounding those ligands. To address these questions we developed a new synthetic strategy to control the ligand shell composition in multifunctional gold nanoparticles (AuNPs) and used the resulting NPs to understand how ligand shell composition influences nanoparticle assembly. We chose AuNPs for two reasons. First, AuNPs (diameter >2 nm) have a plasmon absorption in the visible region22 that allows us to investigate nanoparticle assembly in solution by measuring their optical spectra.23 By using malonamide-containing functional groups as molecular recognition sites in the shell, we can observe an optical response corresponding to the assembly reaction induced by the addition of trivalent lanthanide ions (Ln3+). Second, AuNPs are aptly suited as a multifunctional ligand-stabilized NPs because they can easily be functionalized with nearly all types of electron-donating molecules, resulting in a relatively inert NP with high stability and biocompatibility.24−26 The presence of multiple ligand functionalities on AuNPs has been shown to affect material properties such as fate,27 uptake,28 and binding strength toward analytes. Although there are methods available to produce specific sizes of AuNPs with functionalized ligands that control the NP physical properties, controlling the compositions of mixed ligand shells on the surface of AuNPs is an ongoing challenge. Ligand exchange methods are often performed on preformed AuNPs to introduce mixed ligand shells; however, these reactions can be slow (in the case of thiol-for-thiol exchange),29 and it is difficult to control the extent of exchange necessary to control mixed ligand ratios.30 Direct synthesis of multifunction particles is possible using a modified Brust-Schiffrin synthesis reported by Choo et al.31 In this case, it is difficult to predictably control the monolayer composition and core size because the ligands employed influence both shell composition and core size in ways that are not currently understood. Finally, few studies have been conducted on systems that yield watersoluble NPs with systematically controlled ligand shell compositions. Generally, a large (>10 nm) AuNP with a loosely bound citrate ligand shell is first produced and then one or more exchange steps with excess functional ligands are used to yield more multifunctional materials.32−34 In this report, we describe a new, flow-based synthetic approach that allows us to reproducibly and systematically control the polyvalency (number of targeting ligands bound) of the AuNPs. In a single step, AuNPs with ligand shells containing defined ratios of targeting ligands (malonamides) and diluent ligands (ethylene glycol (EG)) are prepared. A key to successfully and precisely tuning ligand shell composition was the design of an ethylene glycol-based linker for use in both the targeting and the diluent ligands. Nanoparticles with defined ligand shell compositions were used to investigate how the polyvalency influences NP assembly reactions in solution and how polyvalency can be tuned to maximize binding interactions.35 We show that the polyvalency plays a significant role in NP assembly in the presence of trivalent lanthanides, with lower numbers of targeting ligands (∼15%) eliciting the maximum extent of assembly. B

DOI: 10.1021/acs.langmuir.5b03096 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

respectively. All samples containing EG-malonamides that yielded an original colorimetric response were found to be reversible. The detection limit of the nanoparticle materials was determined through a spectral titration of Eu(NO3)3. Increased sensitivity occurred with higher AuNP concentrations, and thus starting absorbance values were fixed at ∼1 A.U. (at 500 nm) for the samples, for higher concentrations while staying well within the saturation limits of the instrument. The absorbance spectrum of purified mixed malonamide/MEEE AuNPs was used as the blank. Successive 1 μL aliquots of appropriate Eu3+ concentration were added until a measurable absorbance change at 550 nm was observed. The detection limit was defined as an increase >3× the noise of the instrument (∼0.002 A.U.). For upper sensing limit measurements, AuNPs were diluted to yield a starting absorbance of ∼0.5 at 500 nm. The AuNPs were again used as a blank, and successive 2 μL aliquots of 2 μM Eu(NO3)3 were added. Absorbance changes at 550 nm were then monitored, and the upper limit was determined after an additional aliquot produced no further change in the absorbance value. Nanoparticle Cross-Linking Investigation with Transmission Electron Microscopy (TEM), SAXS, and Fourier Transform Infrared (FT-IR) Spectroscopy. TEM analysis of purified nanoparticle samples was performed on a FEI Tecnai G2 Spirit TEM operating at 120 kV. AuNP samples were prepared for analysis by floating amine-functionalized SMART grids (Dune Sciences) on top of a drop of diluted solution (either AuNPs only, AuNPs + 20 μM Eu3+, or AuNPs + 20 μM Eu3+ + 20 μM EDTA, pH 7.5 with Tris buffer) for ∼60 s. The grids were allowed to dry in ambient conditions before imaging occurred. SAXS agglomeration analyses of purified AuNPs were performed by first measuring the scattering pattern of malonamide/MEEE-AuNPs, followed by a second measurement of AuNPs with Eu3+ present in solution. Specifically, mixed malonamide/MEEE-AuNPs (200 μL, 0.5 A.U. @ 500 nm) were exposed to line-collimated X-rays for 40 s exposure and averaged over 64 scans. To assess the solution phase agglomeration caused by malonamide binding interactions with Eu3+ present, scattering patterns were obtained on malonamide/MEEE-AuNP solutions at the same concentration as dispersed particles but also containing Eu(NO3)3 (20 μM final). The raw data was background and dark subtracted, desmeared, and imported into IGOR pro as indicated in the core size determination discussion. AuNPs only (dispersed) were fit using the Modeling II macro in IRENA assuming spherical particles, yielding a Gaussian distribution, and were successfully modeled using the LSQF method as a dilute system (with no contribution from a structure factor). Upon addition of Eu3+, AuNP samples with >50% malonamide content in the ligand shell displayed little to no change to the scattering pattern and were also fit as dilute systems (Figure S7). Scattering patterns of Eu3+ + AuNP samples with ≤50% malonamide content revealed a first diffraction peak indicative of regular ordering in solution and were fit using the Modeling II macro and LSQF method, again assuming spherical particles and Gaussian distribution, however, with a hard sticky spheres structure factor.42 FT-IR analyses of AuNPs were obtained using a Thermo Scientific Nicolet 6700 spectrometer. Lyophilized, purified AuNPs (∼0.5 mg) were added to KBr and pressed into pellets. Measurements were recorded in transmittance mode under a

diluent MEEE was performed to quantify mixed ligand ratios; full details and example NMR spectra are contained within the Supporting Information. Nanoparticle Core Size Determination Using Small Angle X-ray Scattering. Nanoparticle sizes resulting from the direct syntheses performed were determined in solution at synthesis concentrations using small-angle X-ray scattering (SAXS). The use of SAXS (i) permits rapid characterization of the core diameters necessary to measure the many samples produced during the course of the syntheses described, (ii) measures large, statistically representative samples and (iii) avoids sample drying effects that often influence size distributions by TEM. These methods were validated by TEM previously41 and during this work. Briefly, NP samples were exposed to monochromated X-rays from a Long Fine Focal spot (LFF) sealed X-ray tube (Cu 1.54 Å) powered by a generator at 2 kW focused by multilayer optics, measured with a Roper CCD in a Kratky camera. The Anton Paar SAXSess, in line collimation mode, was set to average a minimum of 50 scans for 20−40 s exposures. The corresponding dark current and background scans were subtracted from the data before desmearing using the beam profile in Anton Paar SAXSQuant software. Upon import to IGOR Pro the desmeared data were reduced to 200 points (with a 5 data point boxcar average) matching the number of bins to be fit. The size distribution of the sample was then determined by using the size distribution macro in the IRENA package.42 The SAXS patterns were fit using the maximum entropy method, assuming spherical particles (confirmed with TEM), to yield a histogram of volume distribution binned by diameter. For each sample, polydispersity and average core size were determined by fitting a Gaussian function to the histogram distribution. Characterization of Sensing Behavior by UV−vis Spectroscopy. All measurements were performed using a Mikropack DH-2000 UV−vis−NIR light source equipped with an Ocean Optics USB2000 spectrophotometer. Absorbance of purified AuNPs was measured in a quartz cuvette cleaned with aqua regia and rinsed with nanopure water in between all measurements. Mixed malonamide/MEEE-AuNPs were diluted (to 0.5 A.U. at 500 nm) to avoid exceeding the saturation limit of the instrument while the solution volume was maintained at 2 mL total and an initial absorption spectrum was obtained. For determination of colorimetric sensing capabilities, 10 μL of 4 mM Eu(NO3)3 (aq.) was added to the cuvette while stirring. Sensing occurred rapidly (seconds), and the resulting absorbance spectrum was recorded after 2 min. The change in absorbance at 550 nm from the original spectrum to the final Eu3+ added spectrum was used for the optimization of the sensing response as a function of % malonamide. Reversibility of the prepared nanoparticle sensors that demonstrated a colorimetric sensing response for Eu3+ was determined by recording the initial absorbance spectrum of a rapidly stirring solution containing 1.98 mL purified mixed ligand NPs and 20 μL of 1 M Tris buffer (pH = 7.5). Tris buffer was utilized to improve the chelating ability of EDTA and increase the overall binding affinity. An addition of Eu(NO3)3 (aq., 20 μM) as indicated above was added followed by a subsequent addition of EDTA (aq., 20 μM). Spectra were obtained after each addition. The recorded absorbance value at 550 nm versus the addition of Eu3+ and then EDTA were reviewed. Reversibility was defined as the increase and subsequent decrease in absorbance at 550 nm (back to the original value) following the addition of Eu3+ and EDTA, C

DOI: 10.1021/acs.langmuir.5b03096 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir stream of dry air with resolution of 8 cm−1 averaged over 128 scans. For assessment of Eu3+ binding, lyopholized AuNPs produced from the same batch were dissolved in 0.49 mL H2O and aqueous Eu3+ (10 μL, 1 mM Eu(NO3)3) for Eu3+ = 20 μM. The mixture was incubated for 30 min, lyophilized to a powder, pressed into a KBr pellet, and spectra were obtained.

nonspecific binding.43 We tried to incorporate 3 by ligand exchange reactions from preformed mercaptoethoxyethoxyethanol-stabilized gold nanoparticles (MEEE-AuNPs),44,45 but found that it was difficult to control the ratio of ligands within the shell. AuNPs containing a ligand shell of 3 and MEEE were next prepared by direct synthesis (Scheme 2). A total of 34 trials using various ratios of MEEE and 3 were performed (Supporting Information). Although the NPs were stable and water-soluble, the extent of assembly upon Eu3+ addition varied widely from batch to batch. We hypothesized that the variations in response might be due to differences in compositions of the NPs resulting from inadequate mixing of the reagents during synthesis. In these cases, NP formation occurs faster than the reagents can be thoroughly mixed, leading to variations in reagent ratios during the NP synthesis. We evaluated flow synthesis using a mesofluidic reactor as an approach to improve mixing and the reproducibility of these reactions. We have used these reactors previously and found that at higher flow rates (>60 mL/min) they can be used to produce specific nanoparticle sizes with less than 2% batch-tobatch variation.39 In that work, the pH-dependent speciation of the gold precursor permits precise control over the nanoparticle core diameter over a range of 2−10 nm. The use of this approach, coupled with rapid size determination with smallangle X-ray scattering (SAXS), produces working curves that guide the synthesis of NPs with a desired core size containing a specific, single-component ligand shell.46 As a result it is possible to precisely control the core size while installing a desired ligand shell during synthesis. Here, we extend that approach and show that it is possible to achieve this same degree of reproducibility, while carrying out these reactions rapidly, minimizing compositional variability and controlling the polyvalency of targeting ligands within mixed shell NPs. 3/MEEE-AuNPs were synthesized using a mesofluidic reactor with varying ratios of 3 and the diluent (MEEE). Average NP core sizes determined using SAXS are shown in Figure 1A. The core diameter increased (from 2.6−5.9 nm) with increasing feed % of 3 (10−100%). As the feed ratio of 3 increased, the solubility of the NPs in water decreased and, above 75% 3, the particles tend to precipitate during rigorous purification and lyophilization. Although the malonamide functionality is water-soluble, the hydrophobic contribution of the linker appears to dictate solubility and/or stability at higher ligand shell densities. We also evaluated the composition of the ligand shell and compared it to the feed ratio. Ideally, a mixed ligand shell of desired composition could be prepared by simply premixing the ligands at the desired feed ratio. To evaluate the composition, synthesized NPs were extensively purified using diafiltration, concentrated by lyopholization, and redispersed in deuterated

■

RESULTS AND DISCUSSION To probe how ligand shell composition influences NP assembly in solution, the surface architecture on AuNPs must be controlled. Specifically, we aimed to understand how this architecture, including the polyvalency of the targeting (or recognition) functionality and the steric interactions around the targeting groups, influences the assembly process. We used malonamide ligands as recognition elements on the monolayer surface. Previous work in our laboratory demonstrated that the use of these ligands leads to rapid AuNP assembly in the presence of trivalent lanthanides, producing a colorimetric response that can sense these ions.37 Those functionalized AuNPs were readily synthesized by reduction of HAuCl4 in the presence of a malonamide ligand linked by a heptyl chain to a terminal Bunte salt functionality 3 (Scheme 1). As sensors, Scheme 1. Structural Components of Malonamide Ligand 3 Used for Colorimetric Recognition of Ln3+

these NPs possessed a few shortcomings: their core size was difficult to control, their assembly was irreversible, they possessed limited dynamic range, and became unstable toward coalescence during sensing. Here we aimed to use the optical changes to monitor the assembly chemistry. We hypothesized that the assembly reactions leading to the optical response could be studied more effectively if the extent of cross-linking could be decreased, perhaps by reducing the density of malonamide groups in the ligand shell. Synthesis of Mixed Malonamide/Ethylene Glycol AuNPs with Specific Mixed Monolayer Compositions. Short-chain EG functionalities were selected as diluent ligands to decrease the number of malonamide-containing ligands in the shell. EG-based ligands have been shown to produce watersoluble NPs that are stable, biocompatible, and resistant to

Scheme 2. Direct Synthesis of AuNPs Containing a Mixture of Malonamide and Ethylene Glycol Ligands

D

DOI: 10.1021/acs.langmuir.5b03096 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 1. (A) Average core sizes of AuNPs functionalized through direct synthesis with 3/MEEE as determined by SAXS. The core size increases as more 3 is introduced in the mixed ligand feed stock. (B) The ligand shell compositions for 3/MEEE that result from varying the input feed ratio of ligand 3 to MEEE. The ligand shell composition does not track with the feed ratio. (C) Average core sizes of AuNPs functionalized with 1/MEEE determined by SAXS. The data show only a minor core size increase with increased amounts of 1. (D) The ligand shell compositions for 1/MEEE that result from varying the input feed ratio of ligand 1 to MEEE. These data indicate a direct relationship between the input feed ratio of ligand 1 and the actual composition on AuNPs. The error bars for SAXS measurements are one SD of the mean diameter. Ligand shell compositions were determined by iodine decomposition and subsequent 1H NMR analysis. The estimated error of 1H NMR analysis is ±2.5%.

solvent (10% DMSO-d6/90% D2O). Iodine was added to oxidize surface bound ligands and release them as the corresponding disulfides.47 Quantification by 1H NMR (see Supporting Information for full details) of characteristic peaks of the disulfides showed that the ligand composition did not match the feed ratio except at very low ratios. The ligand 3 was underrepresented in the shell at higher feed ratios (e.g., only 24% of 3 was introduced at feed ratio of 50%) (Figure 1B). Although the nanoparticle core size and ligand composition at lower percentages of 3 were reproducible, we aimed to gain greater control over the ligand shell composition across the entire composition range. Given the relatively poor water solubility of 3, it is plausible that the ligands associate during AuNP formation or that the rates of incorporation of the ligands with different tethers vary as it does in 2-D self-assembled monolayers on Au.48 In any case, AuNPs with high feed ratios of 3 (>75%) irreversibly aggregated after lyophilization and could not be redispersed for NMR analyses. The combination of the varying sizes across a range of mixed ligand ratios, the difficulty of incorporating large amounts of malonamide into the ligand shell, and the poor NP stability during manipulation prompted us to redesign the malonamide ligand. Our strategy was to replace the alkyl linker with an EG linker that would increase water solubility and enhance the compatibility of the malonamide with the diluent EG ligand. We hypothesized that installation of an EG linker on the malonamide would not only aid in construction of the NPs during direct synthesis with regards to core size and final composition but also increase the steric stabilization and solubility in water. The length of the linker was chosen so as to extend the malonamide headgroup just beyond the terminus of the diluent ligand. The synthesis of 1 (a malonamide linked to the Bunte salt through a triethylene glycol chain) was performed as described in the Supporting Information.

1/MEEE-AuNPs were synthesized as described for 3/ MEEE-AuNPs. SAXS analyses of AuNPs with feed ratios of 1 ranging from 10 to 100% showed that the core size diameter only varied from 2.6 to 3.3 nm over the range of feed ratios with excellent reproducibility (Figure 1C) and polydispersity values ranging from 19 to 26%. All particles that have 1 within the ligand shell were easily concentrated, lyophilized to dryness, and readily redispersed. 1H NMR spectra show no signals from free ligand, suggesting that the NPs remain stable throughout the drying and redispersion steps. 1 H NMR analysis revealed a direct correlation between the feed ratios of the input ligands to the actual composition on the nanoparticles (Figure 1D). These data suggest that the kinetics of the Bunte salt passivation with the Au surface, given an identical linker, are similar regardless of the terminal headgroup. In this case, matching the linker of the functional malonamide to the diluent MEEE ligand in a mesofluidic reactor proved powerful for synthesis of mixed ligand AuNPs with greatly increased stability and readily tunable ligand ratios. Although the close relationship of the composition to the feed ratio is surprising, the general concept that uniform mixing of ligands on a surface is favored when the linker types and length are similar has precedence in the self-assembled monolayer literature.49−51 Assembly of Mixed Malonamide/EG AuNPs Induced by Trivalent Lanthanide Ions. To test our hypothesis that NP assembly processes might be more reproducible and reversible for lower numbers of malonamide ligands in the shell, we conducted UV−vis, SAXS, and TEM studies. In our original work, the reported AuNPs were highly selective for Ln3+ and sensitive down to concentrations as low as 50 nM; however, the NPs underwent irreversible assembly. AuNPs containing 25% malonamide were investigated to determine whether assembly is more reversible with fewer cross-linking ligands on the surface. E

DOI: 10.1021/acs.langmuir.5b03096 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 2. Reversible assembly and disassembly of malonamide-functionalized AuNPs. (A) UV−vis spectra showing the SPR absorption for 1/MEEE (10:90)-AuNPs (black trace), the same NPs after addition of 20 μM Eu3+ (red trace), and after addition of EDTA (blue dotted trace). (B) SAXS patterns of 1/MEEE (10:90)-AuNPs (black trace) and after the addition of 20 μM Eu3+ (red trace). In both cases, the open circles are the scattering data and the solid lines are fits to those data. The first diffraction peak (q1 = 0.17 Å−1) suggests the formation of assemblies in solution. (C) Representative TEM images of dispersed 1/MEEE (10:90)-AuNPs (left), after introduction of Eu3+ (middle), and after EDTA addition (right). Note that some association of the nanoparticles on the grids is apparent in the left and right panels of panel C. SAXS measurements confirm that these NPs are fully dispersed in solution and have an average core diameter of 2.6 nm.

1/MEEE (10:90)-AuNPs (where reversible sensing was observed by UV−vis). The original purified nanoparticle sample displayed a monomodal distribution with no apparent ordering in solution and was modeled as a dilute system (Figure 2B, fit represented by black line). Upon addition of 20 μM Eu3+, the same AuNP sample displayed regular ordering of nanoparticles as evidenced by a first diffraction peak (q1) at 0.17 Å−1. The mean sphere volume fraction (η) was found to be 0.37 by modeling the data as polydisperse spherical particles with a hard spheres structure factor utilizing the Percus−Yevick approximation,54 indicating the probability of finding a nanoparticle near one or more other nanoparticles is high. A correlation distance of 3.6 nm was determined experimentally for the cross-linked AuNPs,55 a value that is comparable to the maximum predicted NP−NP distance (4.2 nm). A full discussion of the modeling parameters, calculations of the NP−NP spacing (Figure S6), and determination of the extent of agglomeration induced by Eu3+ is found in the Supporting Information. Although the scatter in the data found at higher q is more significant than for measurements carried out with synchrotron X-ray sources at these concentrations, this scatter does not hinder the data interpretation as previously reported by others using lab-scale instruments.56 TEM was used as an additional method to corroborate and visualize the assembly process. Figure 2C (left) shows individual 1/MEEE (10:90)-AuNPs prior to Eu3+ addition. When 20 μM Eu3+ was added to the NP solution, agglomerates are observed (Figure 2C, middle). After EDTA addition to the

Initially, the assembly of 1/MEEE (10:90)-AuNPs was probed by monitoring UV−vis absorption changes upon introduction of Eu(NO3)3 to the NPs.52 The presence of Eu3+ (20 μM, aq.) caused an immediate red-shift and increase in absorption of the SPR band strongly suggesting that the NPs were cross-linking and assembling (Figure 2A, red trace). No precipitation of particles was observed over the course of several hours, and the solution maintained similar absorption characteristics indicating no further NP transformations were occurring (such as core fusion). An assessment of NP disassembly was performed by introducing EDTA, a strong chelator. EDTA is known to have a strong affinity for Eu3+ (log K = 17.35)53 and was selected to scavenge the ions from the assemblies and break the cross-linked structure on the nanoparticles. After EDTA introduction (aq., 20 μM, pH = 7.5 buffered with Tris), the SPR absorption band blue-shifted and decreased in intensity, reverting to the original optical spectrum of individual nonagglomerated AuNPs (Figure 2A, blue dotted trace). Addition of excess Eu3+ to this solution resulted in AuNP assembly, suggesting that the AuNPs remain intact during the disassembly process. Taken together, these optical studies suggest that the assembly process is reversible. The assembly behavior was further evaluated by SAXS to determine whether interparticle interactions were present in solution. While traditionally a technique reserved for synchrotron X-ray sources, we were able to perform SAXS agglomeration measurements on a lab-scale instrument. Scattering patterns were obtained for an aqueous solution of F

DOI: 10.1021/acs.langmuir.5b03096 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir assemblies, images similar to the starting NPs are obtained (Figure 2C, right), suggesting that the cross-links between AuNPs were broken when the Eu3+ ions were chelated by EDTA. A visible spectral titration was performed to determine the upper and lower limits for the Eu3+-induced assembly process. Despite the smaller number of malonamide ligands in the shell, these materials had a similar lower limit for detection as our previous system (∼250 nM). On the other hand, the structural changes led to a 60-fold increase in the upper limit (30 μM) (Figure 3). This increase is likely the result of the increased NP

Figure 4. Sensing response, measured as a change in absorbance at 550 nm when exposed to 20 μM Eu3+, of mixed malonamide 1/ MEEE-AuNPs as a function of % malonamide on the AuNP surface. Solid points represent actual composition values obtained from 1H NMR, and hollow points are shown as the input feed ratio. Error in the % of malonamide was estimated as ±2.5% (from 1H NMR). Absolute error in the sensing response is estimated as 0.01−0.02 for AuNPs based upon noise (∼0.002 A.U.) in UV−vis measurements at AuNP concentrations as indicated in the Methods section.

lowest loadings of 1 (0.5−1%) little assembly was observed. The greatest extent of assembly occurs at intermediate malonamide loadings of 10−33%. At the higher loadings, the assembly upon Eu3+ addition decreases significantly around 50% and was completely inhibited at >74%. SAXS scattering patterns of 1/MEEE (74:26)-AuNPs show minimal change to the form factor when Eu3+ is added with no apparent NP-NP correlation. These results suggest that no agglomerates are formed, corroborating the lack of assembly and subsequent colorimetric response by UV−vis (Figure S7). At the lowest loadings, an average of only 0.5−1 malonamide/AuNP is expected.57 Smaller changes in absorbance make sense in this case because, although the AuNPs can assemble, there are too few malonamide ligands available to produce larger agglomerates that lead to stronger absorbances. As the amount of 1 is increased, the formation of larger assemblies becomes more likely as the number of potential cross-linkers increases. It was unexpected that at the highest ratios, where larger quantities of malonamide were available for cross-linking, no sensing of Eu3+ was found. This lack of sensing at high cross-linking ligand concentration was further investigated and is discussed in the next section. Our results illustrate the role that polyvalency of the NP plays in maximizing the recognition and assembly process of nanomaterials. It has often been suggested that the highest avidity for NP binding occurs at high ligand loadings. For example, Celiz et al. found that NPs containing hydrogen bonding groups on their periphery agglomerated only when loaded at 100% perhaps due to a decrease in intraparticle bonding and increase in interparticle binding at the highest loading.58 In contrast to those findings, our results suggest maximum assembly at intermediate malonamide loadings, consistent with recent findings for AuNP biomolecule conjugates where low to intermediate densities have been shown to produce the largest binding avidity for DNA to complementary sequences on polymeric dendrimers,20 on 2-D self-assembled Au monolayers of mixed thiols59 and for

Figure 3. Visible spectral titration of 1/MEEE (10:90)-AuNPs upon addition of Eu3+ used to determine its sensitivity and sensing limits.

stability relative to our originally reported system with 3 (100%)-AuNPs37 and/or the production of smaller agglomerates that remain dispersed during the assembly process. The redesigned NP with a diluted malonamide (1) ligand shell senses trivalent lanthanide ions, provides high stability to the nanoparticle, and the resulting cross-linking process initiated by Eu3+ addition was found to be reversible. In addition, the assembly of mixed malonamide/EG AuNPs with a functional EG linker and a diluted ligand shell allowed for significant improvement (∼60-fold) in the dynamic range for sensing Eu3+. Influence of Malonamide Polyvalency on Eu3+Induced Assembly. Multifunctional AuNPs containing varying ligand shell compositions from 0.5 to 100% malonamide 1 were prepared in order to determine the effects of malonamide concentration (polyvalency) on the assembly characteristics of the AuNPs in the presence of Eu3+. We hypothesized that an increase in malonamide content within the ligand shell would result in increased cross-linking and a linear response in sensing up to a point where steric interactions might inhibit binding at the highest malonamide concentrations. The colorimetric response measured by UV−vis was used to probe Eu3+-induced assembly behavior as a function of malonamide polyvalency. AuNPs with ligand shells ranging from 0.5−100% 1 (verified by 1H NMR for compositions above 5% 1) diluted with MEEE were mixed with 10 μL of 4 mM Eu3+ (final conc. = 20 μM), and the change in absorbance at 550 nm was plotted as a function of % malonamide 1 (Figure 4). In all cases where NP assembly occurred, the response could be reversed with the addition of EDTA. Three assembly regimes were found in these experiments. For NPs with the G

DOI: 10.1021/acs.langmuir.5b03096 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 5. (A) Schematic of mixed malonamide-AuNPs with structures of ligands (1, 2, and 4) and the sensing interaction with Eu3+ in a 2:1 ligand to metal fashion from neighboring particles. (B) The sensing response of shorter EG tethered 2/MEEE-AuNPs as a function of surface density of malonamide with longer EG tethered 1/MEEE-AuNPs for comparison.

Figure 6. (A) The diamide region of FT-IR spectra obtained with a mixed ligand shell of 2/MEEE (27:73)-AuNPs (black trace) and after addition of Eu3+(red trace). (B) 2 (100%)-AuNPs before (black trace) and after Eu3+ addition (red trace). Both pairs show a similar red-shift in the diamide stretch upon Eu3+ addition indicating that binding to Eu3+ is occurring, even though no assembly takes place in the case of 2 (100%)-AuNPs.

nanoparticle surface, it was expected that increasing steric interactions of neighboring diamides could interfere with Ln3+ cross-linking between neighboring AuNPs. We hypothesized that the intensity of the colorimetric sensing response of these nanomaterials could be improved by increasing plasmon coupling by reducing the length of the functional malonamide. On the other hand, this modification would significantly increase steric crowding. In order to address these hypotheses, a Bunte salt precursor malonamide featuring a single ethylene glycol unit tether, 2, was synthesized in an analogous fashion to 1 (Scheme S2). Mixed 2/MEEE-AuNPs were synthesized over a range of feed ratios. Core sizes varied only slightly from 2.6−2.8 nm from feed ratios of 10−100% 2

recognition of RCA120 lectin using AuNPs functionalized with lactose functionalities (20−65% monolayer density).60 Influence of Sterics at the Periphery of a Malonamide-Functionalized AuNP Ligand Shell on Ln 3+ Triggered Assembly. Cross-linking of our NP sensor requires accommodation of a malonamide functionality from adjacent NPs. Acyclic malonamides, such as tetramethylmalonamide (TMMA) are known to have a preferred trans orientation of the two carbonyls in a free state, but require a cis geometry (overcoming moderate steric strain) for binding Ln3+.61 When the malonamide group becomes crowded by neighboring PEG chains (in the case where the headgroup is buried) or as more malonamides were loaded on the H

DOI: 10.1021/acs.langmuir.5b03096 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

is minimized, and the binding site is extended past the other nonbinding functional ligands.

(Figure S4), and all particles were readily soluble and stable during further manipulation. I2 decomposition and 1H NMR were used to determine ligand compositions and both ligands were incorporated readily similar to the feed ratio. The sensing behavior of these nanomaterials toward Eu3+ was probed as a function of % malonamide 2 and is shown in Figure 5B (blue trace). At ratios of 1−14% of 2, the sensing response was strong and transitioned from too few malonamides (1− 5%) to a peak assembly at intermediate binding density (14%). The shorter EG-tether is thought to bring the AuNPs closer together and enhance the colorimetric response relative to 1 in these cases. However, a decline in the assembly was found from 27 to 49%. Again high ratios of 2 (72−100%)-AuNPs, did not respond to Eu3+. These results suggest that steric crowding dominates as more malonamides were introduced into the mixed ligand shell with a shorter ligand. Partially burying the malonamide headgroup within the ligand shell contributes to the poorer sensing of Eu3+ relative to 1 in the region from 27 to 49% and complete attenuation at the highest surface concentrations. To further assess our hypothesis that increased steric crowding was preventing cross-linking in AuNPs with high polyvalency of malonamides, we used FT-IR spectroscopy to study the shifts in the CO stretching mode of the malonamide groups before and after introduction of lanthanide ions. Specifically, the carbonyl peaks were investigated for NPs that display strong, reversible colorimetric sensing of Eu3+ (2/ MEEE (27:73)-AuNPs) and a homofunctionalized AuNP material (2 (100%)-AuNPs) where no assembly occurs. It has been demonstrated through experiments and density functional theory (DFT) calculations that the TMMA carbonyl stretch shifts from ∼1650 cm−1 to 1600 cm−1 upon Ln3+ binding.62,63 IR spectra of lyophilized particles (Figure 6A) show the expected shift to lower frequency in the carbonyl stretch indicative of Eu3+ binding that is consistent with the observed sensing behavior on the AuNPs with low to intermediate numbers of malonamides. The diamide region on AuNPs containing all malonamides (Figure 6B), shows a similar shift in the carbonyl stretch upon addition of Eu3+, indicating that binding of the ligand to the lanthanide ion is still occurring even though assembly does not occur. Apparently the steric crowding of the increased number of malonamides prohibits interparticle cross-linking and the close proximity of neighboring malonamides within a single NP’s ligand shell favors intramolecular binding. In order to place further steric demands near the binding site of the diamide and investigate the steric influence on NP-NP cross-linking, malonamide ligand 4 containing ethyl groups in place of the methyl groups on tetramethylmalonamide ligand 1, was synthesized (Scheme S1). When introduced along with MEEE in a direct synthesis, the resulting AuNPs ranged in size from 2.5−3.6 nm as the feed ratio of 4 (Figure S5) varied from 10−100%. The increased steric bulk near the malonamide site on 4-AuNPs further inhibited recognition of Eu3+ relative to 1 (Figure S8). On the basis of corroborative visual observations, analysis by microscopy and SAXS, and the IR shifts of the carbonyl groups, it is evident that the surrounding architecture plays an important role in analyte-induced assembly in addition to the polyvalency. IR experiments indicate the interaction between a Eu3+ ion and AuNP surface bound malonamide ligands occurs at any surface ligand concentration. However, NP−NP cross-linking is only favored at low to intermediate binding ligand valencies and is optimized when headgroup size

■

CONCLUSIONS We have demonstrated a new approach to produce multifunctional AuNPs with readily tunable polyvalencies of recognition ligands. The resulting NPs are water-soluble, can be dried and redispersed, and are stable in the presence of added lanthanide ions. A key to success was the design of ligands with common linker structures that mix readily, and thus are equally represented, in the ligand shell. A one-step, mesofluidic synthesis makes it possible to produce AuNPs with reproducible and precisely defined core sizes and ligand shell compositions. The approach provides a reliable alternative to ligand exchange where controlling the ligand shell density reproducibly has been difficult. This new strategy should be readily extended to produce AuNPs containing controlled ligand ratios in mixed-monolayer protected gold nanoparticles for a wide range of desirable functionality. Although the use of sodium borohydride might limit the utility of this synthesis in the presence of sensitive ligands (e.g., those containing biomolecules), the approach can still be employed by incorporating a defined number of linking ligands that can be subsequently coupled to such species. We found that the polyvalency of recognition sites has a strong influence on the assembly of AuNPs upon addition of a cross-linking ion. For malonamide-functionalized AuNPs in the presence of Eu 3+, NPs with intermediate numbers of malonamide ligands produce the largest assemblies, whereas those with higher numbers of ligands yield little to no NP−NP cross-linking. The reduced cross-linking at higher polyvalencies appears to result from steric crowding of the binding sites that favors intraparticle binding of the malonamides at the expense of interparticle binding. Furthermore, increased polyvalency of reactive groups within the passivating monolayer does not necessarily lead to enhanced binding or assembly and can even inhibit the desired interactions. To optimize binding and targeting of functional nanoparticles, careful consideration of both the ligand polyvalency and the steric environment are important. Binding can be enhanced by employing recognition sites at low to intermediate numbers and ensuring those groups extend beyond the periphery of any stabilizing ligand shell.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.5b03096. Supporting Figures S1−S9, ligand syntheses and characterization data, additional experimental details, and calculations (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by the NSF Partnerships for Innovation: Building Innovation Capacity program (IIP1237890). The CAMCOR TEM and XPS facilities are I

DOI: 10.1021/acs.langmuir.5b03096 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

(20) Li, M.-H.; Choi, S. K.; Leroueil, P. R.; Baker, J. R., Jr. Evaluating binding avidities of populations of heterogeneous multivalent ligandfunctionalized nanoparticles. ACS Nano 2014, 8, 5600−5609. (21) Elias, D. R.; Poloukhtine, A.; Popik, V.; Tsourkas, A. Effect of ligand density, receptor density, and nanoparticle size on cell targeting. Nanomedicine 2013, 9, 194−201. (22) Liu, X.; Atwater, M.; Wang, J.; Huo, Q. Extinction coefficient of gold nanoparticles with different sizes and different capping ligands. Colloids Surf., B 2007, 58, 3−7. (23) Pramod, P.; Thomas, K. G. Plasmon coupling in dimers of Au nanorods. Adv. Mater. 2008, 20, 4300−4305. (24) Mirkin, C. A.; Letsinger, R. L.; Mucic, R. C.; Storhoff, J. J. A DNA-based method for rationally assembling nanoparticles into macroscopic materials. Nature 1996, 382, 607−609. (25) Turkevich, J.; Stevenson, P. C.; Hillier, J. Nucleation and growth processes in the synthesis of colloidal gold. Discuss. Faraday Soc. 1951, 11, 55−75. (26) Giersig, M.; Mulvaney, P. Preparation of ordered colloid monolayers by electrophoretic deposition. Langmuir 1993, 9, 3408− 3413. (27) Nativo, P.; Prior, I. A.; Brust, M. Uptake and intracellular fate of surface-modified gold nanoparticles. ACS Nano 2008, 2, 1639−1644. (28) Poon, Z.; Chen, S.; Engler, A. C.; Lee, H-i.; Atas, E.; von Maltzahn, G.; Bhatia, S. N.; Hammond, P. T. Ligand-clustered “patchy” nanoparticles for modulated cellular uptake and in vivo tumor targeting. Angew. Chem., Int. Ed. 2010, 49, 7266−7270. (29) Templeton, A. C.; Wuelfing, W. P.; Murray, R. W. Monolayer − protected cluster molecules. Acc. Chem. Res. 2000, 33, 27−36. (30) Caragheorgheopol, A.; Chechik, V. Mechanistic aspects of ligand exchange in Au nanoparticles. Phys. Chem. Chem. Phys. 2008, 10, 5029−5041. (31) Choo, H.; Cutler, E.; Shon, Y.-S. Synthesis of mixed monolayerprotected gold clusters from thiol mixtures: Variation in the tail group, chain length, and solvent. Langmuir 2003, 19, 8555−8559. (32) Avvakumova, S.; Galbiati, E.; Pandolfi, L.; Mazzucchelli, S.; Cassani, M.; Gori, A.; Longhi, R.; Prosperi, D. Development of U11functionalized gold nanoparticles for selective targeting of urokinase plasminogen activator receptor-positive breast cancer cells. Bioconjugate Chem. 2014, 25, 1381−1386. (33) Nowinski, A. K.; White, A. D.; Keefe, A. J.; Jiang, S. Biologically inspired stealth peptide-capped gold nanoparticles. Langmuir 2014, 30, 1864−1870. (34) Afifi, M. M.; Austin, L. A.; Mackey, M. A.; El-Sayed, M. A. XAV939: From a small inhibitor to a potent drug bioconjugate when delivered by gold nanoparticles. Bioconjugate Chem. 2014, 25, 207− 215. (35) Albanese, A.; Tang, P. S.; Chan, W. C. W. The effect of nanoparticle size, shape, and surface chemistry on biological systems. Annu. Rev. Biomed. Eng. 2012, 14, 1−16. (36) Purification of Laboratory Chemicals, 5th ed.; Armarego, W. L. F., Chai, C. L. L., Eds.; Butterworth Heinemann: Cornwall, U.K., 2003; Chapter 4, p 370. (37) Lisowski, C. E.; Hutchison, J. E. Malonamide-functionalized gold nanoparticles for selective, colorimetric sensing of trivalent lanthanide ions. Anal. Chem. 2009, 81, 10246−10253. (38) Lohse, S. E.; Dahl, J. A.; Hutchison, J. E. Direct synthesis of large water-soluble functionalized gold nanoparticles using bunte salts as ligand precursors. Langmuir 2010, 26, 7504−7511. (39) Elliott, E. W.; Haben, P. M.; Hutchison, J. E. Subnanometer control of core size during mesofluidic synthesis of small (Dcore < 10 nm) water-soluble, ligand-stabilized gold nanoparticles. Langmuir 2015, DOI: 10.1021/acs.langmuir.5b02419. (40) Sweeney, S. F.; Woehrle, G. H.; Hutchison, J. E. Rapid purification and size separation of gold nanoparticles via diafiltration. J. Am. Chem. Soc. 2006, 128, 3190−3197. (41) McKenzie, L. C.; Haben, P. M.; Kevan, S. D.; Hutchison, J. E. Determining nanoparticle size in real time by small-angle x-ray scattering in a microscale flow system. J. Phys. Chem. C 2010, 114, 22055−22063.

supported by grants from the W.M. Keck Foundation, the M.J. Murdock Charitable Trust, the Oregon Nanoscience and Microtechnologies Institute, and the University of Oregon. The University of Oregon NMR facilities are supported by NSF CHE-0923589.

■

REFERENCES

(1) Park, K.; Lee, S.; Kang, E.; Kim, K.; Choi, K.; Kwon, I. C. New generation of multifunctional nanoparticles for cancer imaging and therapy. Adv. Funct. Mater. 2009, 19, 1553−1566. (2) Sanvicens, N.; Marco, M. P. Multifunctional nanoparticles − Properties and prospects for their use in human medicine. Trends Biotechnol. 2008, 26, 425−433. (3) de Dios, S. A.; Díaz-Garcia, M. E. Multifunctional nanoparticles: Analytical prospects. Anal. Chim. Acta 2010, 666, 1−22. (4) Yu, M. K.; Park, J.; Jon, S. Targeting strategies for multifunctional nanoparticles in cancer imaging and therapy. Theranostics 2012, 2, 3− 44. (5) Amoozgar, Z.; Yeo, Y. Recent advances in stealth coating of nanoparticle drug delivery systems. Wiley Interdiscip. Rev. Nanomed. Nanobiotechnol. 2012, 4, 219−233. (6) Bao, G.; Mitragotri, S.; Tong, S. Multifunctional nanoparticles for drug delivery and molecular imaging. Annu. Rev. Biomed. Eng. 2013, 15, 253−282. (7) Yong, K.-T.; Roy, I.; Swihart, M. T.; Prasad, P. N. Multifunctional nanoparticles as biocompatible targeted probes for human cancer diagnosis and therapy. J. Mater. Chem. 2009, 19, 4655−4672. (8) Gao, J.; Gu, H.; Xu, B. Multifunctional magnetic nanoparticles: Design, synthesis, and biomedical applications. Acc. Chem. Res. 2009, 42, 1097−1107. (9) Susumu, K.; Uyeda, H. T.; Medintz, I. L.; Pons, T.; Delehanty, J. B.; Mattoussi, H. Enhancing the stability and biological functionalities of quantum dots via multifunctional ligands. J. Am. Chem. Soc. 2007, 129, 13987−13996. (10) Murphy, C. J.; Gole, A. M.; Stone, J. W.; Sisco, P. N.; Alkilany, A. M.; Goldsmith, E. C.; Baxter, S. C. Gold nanoparticles in biology: Beyond toxicity to cellular imaging. Acc. Chem. Res. 2008, 41, 1721− 1730. (11) Liong, M.; Lu, J.; Kovochich, M.; Xia, T.; Ruehm, S. G.; Nel, A. E.; Tamanoi, F.; Zink, J. I. Multifunctional inorganic nanoparticles for imaging, targeting, and drug delivery. ACS Nano 2008, 2, 889−896. (12) Bertin, P. A.; Gibbs, J. M.; Shen, C. F.; Thaxton, C. S.; Russin, W. A.; Mirkin, C. A.; Nguyen, S. T. Multifunctional polymeric nanoparticles from diverse bioactive agents. J. Am. Chem. Soc. 2006, 128, 4168−4169. (13) Majoros, I. J.; Myc, A.; Thomas, T.; Mehta, C. B.; Baker, J. R., Jr. PAMAM Dendrimer-based multifunctional conjugate for cancer therapy: Synthesis, characterization, and functionality. Biomacromolecules 2006, 7, 572−579. (14) Lin, Q.-S.; Haynes, C. L. Synthesis and characterization of biocompatible and size-tunable multifunctional porous silica nanoparticles. Chem. Mater. 2009, 21, 3979−3986. (15) Cheng, Z.; Al Zaki, A.; Hui, J. Z.; Muzykantov, V. R.; Tsourkas, A. Multifunctional nanoparticles cost versus benefit of adding targeting and imaging capabilities. Science 2012, 338, 903−910. (16) Neouze, M.-A.; Schubert, U. Surface modification and functionalization of metal and metal oxide nanoparticles by organic ligands. Monatsh. Chem. 2008, 139, 183−195. (17) Sperling, R. A.; Parak, W. J. Surface modification, functionalization, and bioconjugation of colloidal inorganic nanoparticles. Philos. Trans. R. Soc., A 2010, 368, 1333−1383. (18) Mullen, D.; Banaszak Holl, M. M. Heterogeneous ligandnanoparticle distributions: A major obstacle to scientific understanding and commercial translation. Acc. Chem. Res. 2011, 44, 1135−1145. (19) Mullen, D. G.; Fang, M.; Desai, A.; Baker, J. R., Jr.; Orr, B. G.; Banaszak Holl, M. M. A quantitative assessment of nanoparticle − ligand distributions: Implications for targeted drug and imaging delivery in dendrimer conjugates. ACS Nano 2010, 4, 657−670. J

DOI: 10.1021/acs.langmuir.5b03096 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir (42) Ilavsky, J.; Jemian, P. R. Irena: Tool suite for modeling and analysis of small-angle scattering. J. Appl. Crystallogr. 2009, 42, 347− 353. (43) Zheng, M.; Huang, X. Nanoparticles comprising a mixed monolayer for specific bindings with biomolecules. J. Am. Chem. Soc. 2004, 126, 12047−12054. (44) Shon, Y. S.; Cutler, E. Aqueous synthesis of alkanethiolateprotected Ag nanoparticles using Bunte Salts. Langmuir 2004, 20, 6626−6630. (45) Woehrle, G. H.; Warner, M. G.; Hutchison, J. E. Ligand exchange reactions yield subnanometer, thiol-stabilized gold particles with defined optical transitions. J. Phys. Chem. B 2002, 106, 9979− 9981. (46) In the present study, mixed ligand shell AuNPs in the 2−4 nm range were targeted. By adjusting the pH of the gold precursor solution, one can adjust the speciation of the gold precursor and target larger core sizes that have similar mixed ligand shells. (47) Templeton, A. C.; Hostetler, M. J.; Kraft, C. T.; Murray, R. W. Reactivity of monolayer-protected gold cluster molecules: steric effects. J. Am. Chem. Soc. 1998, 120, 1906−1911. (48) Folkers, J. P.; Laibinis, P. E.; Whitesides, G. M. Self-assembled monolayers of alkanethiols on gold: the adsorption and wetting properties of monolayers derived from two components with alkane chains of different lengths. J. Adhes. Sci. Technol. 1992, 6, 1397−1410. (49) Smith, R. K.; Reed, S. M.; Lewis, P. A.; Monnell, J. D.; Clegg, R. S.; Kelly, K. F.; Bumm, L. A.; Hutchison, J. E.; Weiss, P. S. Phase separation within a binary self-assembled monolayer on Au{111} driven by an amide-containing alkanethiol. J. Phys. Chem. B 2001, 105, 1119−1122. (50) Folkers, J. P.; Laibinis, P. E.; Whitesides, G. M.; Deutch, J. Phase behavior of two-component self assembled monolayers of alkanethiolates on gold. J. Phys. Chem. 1994, 98, 563−571. (51) Bain, C. D.; Whitesides, G. M. Formation of two-component surfaces by the spontaneous assembly of monolayers on gold from solutions containing mixtures of organic thiols. J. Am. Chem. Soc. 1988, 110, 6560−6561. (52) At the outset of this study, a range of trivalent lanthanide salts were assessed and produced similar binding characteristics to Eu3+. This lack of selectivity for particular lanthanides was also seen in our original report.37 (53) Bulman, R. A. Mobilization of lanthanides through the terrestrial biosphere. In The Lanthanides and Their Interrelations with Biosystems; Sigel, A., Sigel, H., Eds.; Marcel Dekker, Inc: The Netherlands, 2003; Vol. 40, Chapter 2, pp 40−68. (54) Percus, J. K.; Yevick, G. J. Analysis of Classical Statistical Mechanics by Means of Collective Coordinates. Phys. Rev. 1958, 110, 1−13. (55) Pabisch, S.; Feichtenschlager, B.; Kickelbick, G.; Peterlik, H. Effect of Interparticle Interactions on Size Determination of Zirconia and Silica Based Systems − A Comparison of SAXS, DLS, BET, XRD, and TEM. Chem. Phys. Lett. 2012, 521, 91−97. (56) Polte, J.; Erler, R.; Thünemann, A. F.; Sokolov, S.; Ahner, T. T.; Rademann, K.; Emmerling, F.; Kraehnert, R. Nucleation and growth of gold nanoparticles studied via in situ small angle x-ray scattering at millisecond time resolution. ACS Nano 2010, 4, 1076−1082. (57) Based upon the calculated footprint of MEEE on the AuNP surface for a diameter of 2.6 nm. The surface area (SA) was calculated to be 21.2 nm2 from SA = 4π*(D/2)2. The number of ligands on an MEEE-AuNP = (SA)*(packing density of sphere)/(footprint of MEEE). The packing density of a sphere was assumed to be 0.826, and the calculated footprint of MEEE is 0.1775 nm2 from J. Am. Chem. Soc. 2005, 127, 2172−2183.10.1021/ja0457718 (58) Celiz, A. D.; Lee, T.-C.; Scherman, O. A. Polymer-Mediated Dispersion of Gold Nanoparticles: Using Supramolecular Moieties on the Periphery. Adv. Mater. 2009, 21, 3937−3940. (59) Frederix, F.; Bonroy, K.; Laureyn, W.; Reekmans, G.; Campitelli, A.; Dehaen, W.; Maes, G. Enhanced performance of an affinity biosensor interface based on mixed self-assembled monolayers of thiols on gold. Langmuir 2003, 19, 4351−4357.

(60) Takae, S.; Akiyama, Y.; Otsuka, H.; Nakamura, T.; Nagasaki, Y.; Kataoka, K. Ligand density effect on biorecognition by PEGylated gold nanoparticles: Regulated interaction of RCA120 lectin with lactose installed to the distal end of tethered PEG strands on gold surface. Biomacromolecules 2005, 6, 818−824. (61) Lumetta, G. L.; Rapko, B. M.; Garza, P. A.; Hay, B. P.; Gilbertson, R. D.; Weakley, T. J. R.; Hutchison, J. E. Deliberate design of ligand architecture yields dramatic enhancement of metal ion affinity. J. Am. Chem. Soc. 2002, 124, 5644−5645. (62) Den Auwer, C. D.; Charbonnel, M. C.; Drew, M. G. B.; Grigoriev, M.; Hudson, M. J.; Iveson, P. B.; Madic, C.; Nierlich, M.; Presson, M. T.; Revel, R.; et al. Crystallographic, x-ray absorption, and IR studies of solid- and solution-state structures of tris(nitrato) N, N, N′, N′,-tetraethylmalonamide complexes of lanthanides. Comparison with the americium complex. Inorg. Chem. 2000, 39, 1487−1495. (63) Sandrone, G.; Dixon, D. A.; Hay, B. P. Conformational analysis of malonamide, N, N′,-dimethylmalonamide, and N, N, N′, N′tetramethylmalonamide. J. Phys. Chem. A 1999, 103, 3554−3561.

K

DOI: 10.1021/acs.langmuir.5b03096 Langmuir XXXX, XXX, XXX−XXX