<?xml-model href='http://www.tei-c.org/release/xml/tei/custom/schema/relaxng/tei_all.rng' schematypens='http://relaxng.org/ns/structure/1.0'?><TEI xmlns="http://www.tei-c.org/ns/1.0">
	<teiHeader>
		<fileDesc>
			<titleStmt><title level='a'>Linker-Templated Structure Tuning of Optical Response in Plasmonic Nanoparticle Gels</title></titleStmt>
			<publicationStmt>
				<publisher></publisher>
				<date>10/06/2022</date>
			</publicationStmt>
			<sourceDesc>
				<bibl> 
					<idno type="par_id">10434688</idno>
					<idno type="doi">10.1021/acs.jpcc.2c05353</idno>
					<title level='j'>The Journal of Physical Chemistry C</title>
<idno>1932-7447</idno>
<biblScope unit="volume">126</biblScope>
<biblScope unit="issue">39</biblScope>					

					<author>Murari Singh</author><author>Zachary M. Sherman</author><author>Delia J. Milliron</author><author>Thomas M. Truskett</author>
				</bibl>
			</sourceDesc>
		</fileDesc>
		<profileDesc>
			<abstract><ab><![CDATA[Gel assemblies of functional nanoparticles, reversibly associated into percolating networks using bifunctional linking molecules, offer promise as versatile materials platforms. Molecular linkers can be customized to template interparticle spacing and modify colloidal network attributes, enabling design for structuredependent properties. Mechanical properties of gels are commonly studied by molecular simulation, but simulating the optical response of large-scale, disordered assemblies has been computationally intractable, limiting our understanding of light-matter interactions in structurally complex plasmonic networks. Here, we use a recently developed mutual polarization method, capable of predicting optical properties for large disordered configurations of spherical particles, together with an experimentally informed coarse-grained model to study the behavior of plasmonic linker gels. The simulation results demonstrate how blends of short and long linkers with the same average molecular weight can be chosen to deliberately modulate structure-dependent near-and far-field spectral features of the colloidal gel while preserving gel mechanical properties. Linker selection can also be used to prepare gel networks with qualitatively different mechano-optical responses. The structural changes occurring under strain shed light on possible origins of experimentally observed red-and blue-shifting of the optical extinction of plasmonic nanocomposites under a uniaxial extension.]]></ab></abstract>
		</profileDesc>
	</teiHeader>
	<text><body xmlns="http://www.tei-c.org/ns/1.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink">
<div xmlns="http://www.tei-c.org/ns/1.0"><head>&#9632; INTRODUCTION</head><p>Assemblies of metal nanoparticles, whose conduction electrons interact collectively with light by localized surface plasmon resonance (LSPR), display rich optical properties. <ref type="bibr">[1]</ref><ref type="bibr">[2]</ref><ref type="bibr">[3]</ref><ref type="bibr">[4]</ref> Structure-dependent LSPR coupling in these materials produces frequency-tunable absorption and strong electromagnetic fields in gaps between closely spaced nanoparticles ("hot spots"), enabling applications from molecular detection and characterization to energy harvesting and data storage. <ref type="bibr">3,</ref><ref type="bibr">[5]</ref><ref type="bibr">[6]</ref><ref type="bibr">[7]</ref><ref type="bibr">[8]</ref><ref type="bibr">[9]</ref><ref type="bibr">[10]</ref><ref type="bibr">[11]</ref><ref type="bibr">[12]</ref><ref type="bibr">[13]</ref><ref type="bibr">[14]</ref> This sensitivity to nanoparticle spatial organization can render plasmonic assemblies responsive, capable of dynamically reconfiguring due to physicochemical, <ref type="bibr">[15]</ref><ref type="bibr">[16]</ref><ref type="bibr">[17]</ref><ref type="bibr">[18]</ref><ref type="bibr">[19]</ref><ref type="bibr">[20]</ref><ref type="bibr">[21]</ref><ref type="bibr">[22]</ref><ref type="bibr">[23]</ref><ref type="bibr">[24]</ref> electromagnetic, <ref type="bibr">21,</ref><ref type="bibr">[24]</ref><ref type="bibr">[25]</ref><ref type="bibr">[26]</ref><ref type="bibr">[27]</ref><ref type="bibr">[28]</ref><ref type="bibr">[29]</ref> or mechanical <ref type="bibr">[30]</ref><ref type="bibr">[31]</ref><ref type="bibr">[32]</ref><ref type="bibr">[33]</ref><ref type="bibr">[34]</ref><ref type="bibr">[35]</ref><ref type="bibr">[36]</ref><ref type="bibr">[37]</ref><ref type="bibr">[38]</ref><ref type="bibr">[39]</ref><ref type="bibr">[40]</ref><ref type="bibr">[41]</ref><ref type="bibr">[42]</ref> stimuli to form structures with distinct optical near-and far-field spectral features.</p><p>Bifunctional molecules that link sites (e.g., ligands) on neighboring nanoparticles offer a powerful handle for design of functional colloidal assemblies. Linkers can be selected to deliberately template colloid spacing and valence in threedimensional networks, as demonstrated for plasmonic clusters, <ref type="bibr">27,</ref><ref type="bibr">[43]</ref><ref type="bibr">[44]</ref><ref type="bibr">[45]</ref><ref type="bibr">[46]</ref><ref type="bibr">[47]</ref><ref type="bibr">[48]</ref><ref type="bibr">[49]</ref> superlattices, <ref type="bibr">[50]</ref><ref type="bibr">[51]</ref><ref type="bibr">[52]</ref><ref type="bibr">[53]</ref><ref type="bibr">[54]</ref><ref type="bibr">[55]</ref> and gels. <ref type="bibr">22,</ref><ref type="bibr">23,</ref><ref type="bibr">[56]</ref><ref type="bibr">[57]</ref><ref type="bibr">[58]</ref> In recent studies, doped metal oxide nanocrystal dispersions underwent pronounced structural changes upon linker-mediated gelation by dynamic covalent bonding, <ref type="bibr">22,</ref><ref type="bibr">23,</ref><ref type="bibr">59</ref> and the resulting LSPR coupling led to substantial modulation of their infrared extinction spectra. Since mechanical properties of colloidal gels are also sensitive to network architecture, <ref type="bibr">[60]</ref><ref type="bibr">[61]</ref><ref type="bibr">[62]</ref><ref type="bibr">[63]</ref> linked colloidal gel assembly presents an attractive platform for preparing hybrid materials with targeted optical and mechanical properties and distinctive mechano-optical response.</p><p>Though mechanical properties of amorphous materials are commonly studied by molecular simulation, modeling the optical response of nanoparticle assemblies with complex structural features involving large numbers of particles has been computationally intractable. <ref type="bibr">59</ref> Accordingly, how local structuring affects resonant light-matter interactions in disordered plasmonic materials such as colloidal gels with strong LSPR coupling remains poorly understood. <ref type="bibr">22,</ref><ref type="bibr">23,</ref><ref type="bibr">59</ref> We recently introduced a highly efficient mutual polarization method (MPM) <ref type="bibr">59</ref> capable of rapidly computing optical properties for arbitrary configurations of thousands of spherical nanoparticles with periodic boundary conditions, including mixtures of colloids with different dielectric functions. Here, we use this method together with configurations generated from an experimentally informed coarse-grained model <ref type="bibr">64,</ref><ref type="bibr">65</ref> to explore via simulation how linking molecules can be selected to deliberately modify the structure in colloidal gels for the desired optical and mechanical properties (Figure <ref type="figure">1</ref>).</p><p>Our simulation results establish linker molecular weight distribution as a powerful tool for designing plasmonic linker gels. As previously hypothesized, <ref type="bibr">64</ref> we found that optical extinction spectra can be tuned largely independent of gel elastic properties by employing different blends of short and long linkers with fixed average molecular weight. Short linkers strongly constrain the distance between nanoparticle surfaces they connect, creating small interparticle gaps. Increasing the proportion of these gaps strengthens LSPR coupling in the network, leading to high near-field enhancement and broadening the optical extinction spectra. <ref type="bibr">22,</ref><ref type="bibr">23,</ref><ref type="bibr">59</ref> On the other hand, as in polymer networks, <ref type="bibr">[66]</ref><ref type="bibr">[67]</ref><ref type="bibr">[68]</ref><ref type="bibr">[69]</ref> the elastic moduli and elongation behavior of the resulting nanoparticle networks depend on average linker length, but are otherwise insensitive to the details of the linker-size distribution. Together, this modeling predicts that linker composition provides a convenient experimental handle for modulating the optical response of gels without sacrificing key aspects of their mechanical properties.</p><p>The simulations also allowed us to test how linker mixtures can be chosen to program the optical response of plasmonic gels to changes in their macroscopic strain state. Experiments on polymer-metal nanocomposites have produced qualitatively distinct polarization-dependent spectral changes due to stretching, <ref type="bibr">31,</ref><ref type="bibr">[33]</ref><ref type="bibr">[34]</ref><ref type="bibr">[35]</ref><ref type="bibr">38</ref> but detailed information about how the mechano-optical properties reflect modifications to local arrangement of nanoparticles in real-space is lacking. Unlike for nanocomposites, the plasmonic nanoparticles of linked colloidal gels are themselves elements of the percolating network, which tightly couples their spatial organization to the macroscopic strain state of the material. Using our model linker gel, we study how this coupling leads to optical responses that reproduce the diverse trends observed in experiments and relate them to linker-templated structural modifications of the networks under uniaxial strain.</p></div>
<div xmlns="http://www.tei-c.org/ns/1.0"><head>&#9632; METHODS</head><p>Coarse-Grained Model. Plasmonic nanoparticle-linker networks were studied via molecular dynamics simulations in the canonical ensemble using LAMMPS (Large-scale Atomic/ Molecular Massively Parallel Simulator), <ref type="bibr">70</ref> fixing temperature and volume V of a triply periodic, square cuboid-shaped cell. Interactions were described by the coarse-grained PolyPatch (Polymer linked, Patchy colloid) model, <ref type="bibr">64,</ref><ref type="bibr">65</ref> which treats bifunctional linking molecules as linear bead-spring chains and nanoparticles as spherical colloids with patches that can each form at most one attractive bond of strength &#949; with a chain end (Figure <ref type="figure">1b</ref>). Gelation occurs by slowly "cooling" a single-phase fluid to conditions of high &#946;&#949;, where &#946; = 1/k B T and k B is the Boltzmann constant. The mechanical response of the networks to uniaxial deformation can then be readily probed via molecular dynamics simulations.</p><p>Linkers were modeled as fully flexible chains with repulsive chain-segment interactions, mimicking good-solvent conditions, <ref type="bibr">71</ref> and segments were treated as spherical beads of diameter d and mass m, joined to their neighbors by finitely extensible nonlinear elastic bonds. Linker stiffness is also a tunable parameter of the PolyPatch model; a discussion about its effects on colloidal network structure and properties, as well as synthetic strategies for its systematic variation in experiments, is presented elsewhere. <ref type="bibr">65</ref>  . We studied three systems with single-component linkers, small chains with M = 2, 5, and 8 segments, respectively, as well as binary linker mixtures with &#10216;M&#10217; &#8801; (N 1 M 1 + N 2 M 2 )/N l = 5, where N i represents the number of linkers (i.e., M i , segment chains) of component i (Table <ref type="table">1</ref>). Configurations were rendered using OVITO 2.9.0. 72 Mechanical Properties. As described elsewhere, <ref type="bibr">64</ref> slow annealing was used to prepare gels. Molecular dynamics simulations were initially equilibrated for 0.5 &#215; 10 4 &#964; without any linker-patch bonding (&#946;&#949; = 0), where = md 2 . The integration time step was 0.001 &#964;, and T was controlled by a Langevin thermostat with friction coefficient 0.1 m/&#964; applied to each particle. Linker-patch bonding energy was increased to &#946;&#949; = 10 over 10 4 &#964;, followed by a 10 4 &#964; equilibration period. Linker-patch energy was then further ramped by k B T over a 0.5 The Journal of Physical Chemistry C &#215; 10 4 &#964; period, followed by reequilibration for 1.5 &#215; 10 4 &#964;. This process was repeated nine times to reach &#946;&#949; = 20. Final configurations saved from this annealing protocol were thermalized for 2 &#215; 10 3 &#964; at &#946;&#949; = 100 and subsequently used as the initial states for the uniaxial elongation simulations described below. In the gel states, over 85% of linkers formed bridges between patches on neighboring nanoparticles, while the rest either bonded to a single patch (with a dangling end) or bonded to two patches on the same colloid, forming a selfloop. No self-loops form for M = 2 and 5 linkers because their end-to-end distance is shorter than the patch-patch separation on a colloid.</p><p>To quantify the mechanical response of gels to uniaxial extension, we performed a series of deformation simulations in which one dimension of the simulation cell L z was elongated at a steady rate of 10 -4 &#964; -1 . The other two dimensions, L x and L y , were simultaneously and isotropically contracted at a rate to conserve volume, consistent with a Poisson's ratio of 1/2 (commonly observed for soft materials like polymer melts and solutions, where the bulk modulus is higher than the shear modulus). <ref type="bibr">71</ref> Strains in the range &#955; = L z /L z,0 &#8712; [1, 4] were simulated, where L z,0 was the initial box dimension in the z direction. Bonds between linkers and patches break and reform during deformation. The engineering (or nominal) stress was computed from the diagonal components of stress tensor, &#963; n &#8801; &#963; true /&#955; = [&#963; zz -0.5(&#963; xx + &#963; yy )]/&#955;, where &#963; true is the true stress. For small uniaxial deformations, the simulation data obeys the classical form of the stress-elongation relationship, &#963; eng = G(&#955; -1/&#955; 2 ), observed experimentally for polymer networks and predicted by network elasticity theory. <ref type="bibr">71</ref> Using this form, we extracted G from simulated uniaxial elongation data, averaging over 10 independent initial configurations and deformations in three independent deformation directions (x, y, and z). From the averaged &#963; n across samples, we extract the maximum stress n UTS and corresponding strain &#955; UTE . Optical Properties. We calculated the extinction spectra and distributions of near-field intensity for our simulated nanoparticle networks using a recently developed manybodied, mutual polarization method (MPM) . <ref type="bibr">59</ref> The dielectric function of the plasmonic colloids are modeled with a Drude dielectric function,</p><p>, where &#969; p is the plasma frequency, &#947; is the damping coefficient, &#949; &#8734; is the high-frequency permittivity, and &#949; 0 is the vacuum permittivity. The permittivity of the background medium &#949; m is assumed real and frequency-independent over the range of interest. The patches and linkers are ignored in the optical calculations. We set , where &#956; m is permeability of the background medium. In terms of these units, the optical calculations apply to a range of materials, with different colloid compositions and background media encoded through specific choices of * p and &#949; m . For example, typical values for colloidal tin-doped indium oxide (ITO) nanoparticles are &#949; m = 2&#949; 0 , * = 10000 cm p 1 , &#947; = 500 cm -1 , and &#949; &#8734; = 4&#949; 0 .</p><p>The N induced particle dipoles p i in an electric field E 0 of magnitude E 0 = |E 0 | are computed by solving the linear system of equations for each frequency &#969;</p><p>where </p><p>is the dipole polarizability, d c is the particle diameter, &#949; p (&#969;) is the particle permittivity at &#969;, &#949; m is the permittivity of the background medium (assumed pure real), r = x ix j , x i is particle i's position, r = |r|, = r r r/ , and I is the identity tensor. The sum in eq 1 runs over all particles and their periodic images and is evaluated using an efficient Ewald summation method. <ref type="bibr">73</ref> This linear system is solved iteratively using the generalized minimal residual method (GMRES) to a relative error tolerance of 10 -3 . The ensemble-averaged extinction cross-section per nanoparticle volume is </p><p>where the sum runs over all particles and periodic images and is evaluated using the same Ewald method as eq 1. The locations of local maxima in the field intensity, that is, "hotspots", are approximated as The Journal of Physical Chemistry C</p><p>For quiescent gels, extinction spectra and the near-field distributions are averaged over x, y, and z polarizations and 10 gel configurations from independent molecular dynamics simulations. The near-field distributions are sampled at the frequency corresponding to maximum | | p for each gel. For strained gels, perpendicular-polarization extinction spectra are averaged over two polarizations (E e x 0 and E e y 0 ) perpendicular to the elongation axis e ( ) z . Both the parallel-and perpendicular-polarization extinction spectra are averaged over 10 independent trajectories at each strain value, and &#969; LSPR is the frequency corresponding to the maximum value of the extinction spectra.</p><p>Structural Properties. The directional radial distribution functions g &#8741; and g &#8869; were computed with the help of projected distances</p><p>The quantity g &#8741; (r) is the radial distribution function computed only between pairs of particles with r &#8869; &#8804; d c /2, and g &#8869; (r) is computed only between pairs of particles with r &#8741; &#8804; d c /2. We calculate these radial distribution functions through conditional binning</p><p>where h = 0.004d c is the bin spacing and &#961; is the number density.</p><p>The three-dimensional structure factor S(q) was computed on a grid of wavevectors q = [2&#960;&#954; x /L x , 2&#960;&#954; y /L y , 2&#960;&#954; z /L z ], where &#954; i = -N i /2, ..., (N i -1)/2 and L i are the box dimensions, by dividing the simulation box into N x N y N z approximately cubic cells with centers r = [n x L x , n y L y , n z L z ], where n i = -N i /2, ..., (N i -1)/2. Selecting N x = N y = 256 in the smallest box dimension and N z = N x L z /L x (rounded to nearest integer) to maintain uniform grid density, we counted the number of particles in each cell &#961;(r), used a fast Fourier transform to get q ( ), then computed</p><p>. Using projected wavenumbers</p><p>q I e e , ( )</p><p>we computed the 2D S(q &#8741; , q &#8869; ) by constructing a grid [2&#960;&#954; &#8741; / L &#8741; , 2&#960;&#954; &#8869; /L &#8869; ], where &#954; i = 0, 1, 2, ..., and averaging all S(q) values in the same bin. We averaged g(r) over 10 independent gel trajectories and 1000 configurations per gel. S, g &#8741; , and g &#8869; are computed over a single gel trajectory at each strain.</p></div>
<div xmlns="http://www.tei-c.org/ns/1.0"><head>&#9632; RESULTS AND DISCUSSION</head><p>Simulated uniaxial deformation of the model nanoparticle networks highlights how linker molecular weight influences mechanical properties (Figure <ref type="figure">2</ref>). Adopting short (M = 2) linking molecules produced rigid assemblies with higher shear modulus G, but lower ultimate tensile strength n UTS and elongation &#955; UTE , owing to the limited extensibility of the short linkers. In contrast, employing longer, flexible linking chains (M = 8) resulted in a softer (lower G) but stronger (higher n UTS ) and more stretchable (higher &#955; UTE ) network. The intermediate, 5-mer colloidal gel also exhibited considerable strength and stretchability, with a modest enhancement of rigidity compared to the 8-mer gel. Deformation simulations of three different colloidal networks with a fixed average linker length (&#10216;M&#10217; = 5), including blends of short (M 1 = 2) and long (M 2 = 8, 10, or 12) linkers address the sensitivity of the mechanical properties to the linker size distribution. Strikingly, the computed shear modulus and ultimate elongation of these networks were equivalent to those of the 5-mer gel. The ultimate tensile strengths of the blended-linker gels were 20-35% lower than that of the 5-mer assembly and in between those of the dimer and 8-mer gels, as expected. The slightly reduced tensile strengths of blendedlinker gels follows from the ability of linkers longer than M = 5 to form self-loops instead of bridges. These results indicate that the colloidal gels with binary linker blends show behavior consistent with experiments on polymer networks and predictions of network elasticity theory. <ref type="bibr">[66]</ref><ref type="bibr">[67]</ref><ref type="bibr">[68]</ref><ref type="bibr">[69]</ref> If the average chain length is fixed, other details of the length distribution&#65533; whether it be for polymers in a network or linkers in a colloidal gel&#65533;do not significantly impact shear elasticity or elongation behavior.</p><p>The model further suggests a design opportunity for plasmonic linker gels, where optical response is known to be sensitive to the presence of close nanoparticle separations, which give rise to pronounced LSPR coupling. <ref type="bibr">22,</ref><ref type="bibr">23</ref> By choosing blends with different proportions of small linkers (but the same molar average linker length), it should be possible to deliberately modulate the number of close nanoparticle contacts and, hence, the structure-dependent near-and far-field optical response of the network, while leaving gel mechanical properties largely unchanged.</p><p>To elucidate how the choice of linker affects network microstructure, we computed the colloid-colloid radial distribution function g cc (r) (Figure <ref type="figure">3a</ref>). The quantity The Journal of Physical Chemistry C N c g cc (r)/V characterizes the average colloid number density a distance r away from a colloid center. Focusing on assemblies with a single linker type, there is stark, qualitative difference between the spacing of colloids when short (M = 2) or longer (M = 5 and 8) linkers are employed. The short-linker network's radial distribution function features a sharp colloidcolloid "contact" peak and a nearby second peak corresponding to colloid surfaces separated by the dimer bond. <ref type="bibr">64</ref> Gels with only longer linkers lacked the contact peak altogether and instead displayed a broader first peak at larger separations, allowing the M = 5 and M = 8 chains to access more configurational degrees of freedom in extended states. <ref type="bibr">71</ref> Significantly, networks with binary linker mixtures all exhibited the contact peak, the magnitude of which was correlated to the proportion of short linkers present.</p><p>The linker-templated structural differences produced a diverse range of far-field optical responses as reflected by the computed extinction cross section C(&#969;) (Figure <ref type="figure">3b,</ref><ref type="figure">c</ref>). The 8-  Colloid-colloid static structure factors S(q) for wavevectors parallel q &#8741; and perpendicular q &#8869; to strain direction: (a, f) prior to strain and (b, g) in strain. Colloidcolloid one-dimensional radial distribution function parallel g &#8741; (r) and perpendicular g &#8869; (r) to strain direction: (c, h) prior to strain and (d, i) in strain. Configuration snapshots in strain (e, j).</p></div>
<div xmlns="http://www.tei-c.org/ns/1.0"><head>The Journal of Physical Chemistry C</head><p>mer gel exhibited a symmetric extinction spectrum resembling the isolated particle spectrum in Figure <ref type="figure">3b</ref>, consistent with weak LSPR coupling and expected based on the lack of a contact peak in g cc (r). By comparison, the extinction cross section for the dimer gel was significantly broadened, reflecting strong LSPR coupling. This broadening embodies far-field signatures of both red-shifting and blue-shifting, consistent with a structure that contains contact pairs with an isotropic distribution of orientations relative to the incident field (Figure <ref type="figure">3b</ref>). As expected, the 5-mer and binary linker gels displayed cross sections consistent with LSPR coupling-induced broadening intermediate to that of the 8-mer and dimer networks.</p><p>Despite similarities in the far-field optical properties of the 5mer and the binary linker networks, differences become apparent upon examination of their near-field responses (Figure <ref type="figure">3d,</ref><ref type="figure">e</ref>). Assemblies with linker mixtures have an increased probability of finding nanoparticles with strong local electric fields near their surfaces. Near-field enhancements were systematically more pronounced for networks with greater proportions of short linkers and hence a higher prevalence of nanoparticles in contact, that is, higher values of g cc (r = d c ). That hot-spot enhancements can be realized without sacrificing network stiffness or stretchability (Figure <ref type="figure">2</ref>) emphasizes an interesting advantage of adopting blends of short and long linkers for designing plasmonic nanoparticle linker gels.</p><p>Differences in the linker-templated structure for the gels also set the stage for distinct responses to mechanical deformation (Figure <ref type="figure">4</ref>). For the 8-mer gel network, the static structure factor and the one-dimensional colloid radial distribution functions established that strain distorts the first-neighbor coordination shell, significantly increasing nearest-neighbor distances parallel to strain, while slightly decreasing separations in the perpendicular direction. Inspection of a strained configuration confirmed that these structural changes approximately correspond to affine, uniaxial deformation. For the mixed M = (2, 12) linker gel, qualitatively different deformation-induced structural changes were evident. Despite little change in the perpendicular direction, deformation enhanced structural correlations in the direction parallel to stretching, as evidenced by an increase in height of the static structure factor and the development of short-range ordering reminiscent of cluster formation in the radial distribution function. <ref type="bibr">74</ref> The configuration snapshot of the mixed linker gel visually confirms that its structural changes reflect a spatially heterogeneous nanoparticle distribution (Figure <ref type="figure">4j</ref>). Short linkers promote clustering of plasmonic nanoparticles under uniaxial deformation, while longer linkers extend to accommodate the macroscopic strain in the network.</p><p>Accordingly, the computed optical extinction cross sections of our model gels display distinct polarization-dependent responses to uniaxial strain, evidenced by qualitatively different shifts in their LSPR peak frequencies (Figure <ref type="figure">5</ref>). For the 8-mer gel, the extinction maximum showed blue-shifting for the polarization parallel to the extension direction, and moderate red-shifting for the perpendicular polarization. This is behavior anticipated with affine deformation <ref type="bibr">75</ref> such as that seen in Figure <ref type="figure">4</ref> for the 8-mer gel, with particles moving farther apart in the direction parallel to extension (weakening LSPR coupling), but closer together in the perpendicular direction (strengthening LSPR coupling). Though experimental systems have been prepared that show this intuitive behavior, <ref type="bibr">35,</ref><ref type="bibr">38</ref> the opposite trend, pronounced red-shifting of the LSPR peak frequency for the parallel polarization and moderate blueshifting in the perpendicular polarization, is more commonly observed for nanocomposites, <ref type="bibr">31,</ref><ref type="bibr">[40]</ref><ref type="bibr">[41]</ref><ref type="bibr">[42]</ref> where the plasmonic nanoparticles have a tendency to cluster. The simulated colloidal gel with a binary mixture of M 1 = 2 and M 2 = 12 linkers, which also formed nanoparticle clusters under uniaxial strain (Figure <ref type="figure">4</ref>), displayed this more typical, though less intuitive, mechano-optical response (Figure <ref type="figure">5b</ref>). Unlike in the plasmonic nanocomposites, where the degree of nanoparticle clustering and structure-dependent optical response is extremely challenging to control, the present results suggest that such changes can be deliberately engineered in linked plasmonic nanoparticle networks via a macroscopic control parameter: the proportion of short linkers (i.e., the composition of the binary linker blend).</p></div>
<div xmlns="http://www.tei-c.org/ns/1.0"><head>&#9632; CONCLUSIONS</head><p>The simulation results reported here, leveraging the recently introduced MPM for computing optical properties of disordered assemblies <ref type="bibr">59</ref> and the experimentally motivated coarse-grained PolyPatch model, <ref type="bibr">64,</ref><ref type="bibr">65</ref> establish how binary mixtures of short and long linkers can be chosen to deliberately template structure and modulate near-and far-field optical and mechano-optical responses of plasmonic nanoparticle linker gels. By changing the composition in a blend of short and long linkers, one can assemble plasmonic networks with qualitatively different mechano-optical responses, capturing the diversity of behaviors observed in experiments and providing insights into their structural origins. This framework can be extended in future studies, and even coupled with inverse methods, <ref type="bibr">76</ref> to design assembly of complex plasmonic gels with novel structure-dependent properties, including hierarchical networks with planet-satellite motifs <ref type="bibr">77</ref> or interlaced double networks. <ref type="bibr">78</ref> It should also prove useful for studying linkertemplated compositional ordering and optical response of binary plasmonic nanoparticle gels, complementing what is known about plasmon coupling in binary nanocrystal superlattices. <ref type="bibr">59,</ref><ref type="bibr">[79]</ref><ref type="bibr">[80]</ref><ref type="bibr">[81]</ref><ref type="bibr">[82]</ref> Finally, it may be interesting to consider networks with a much wider range of linker molecular weights and linker-to-colloid ratios, for example, characterizing both Poisson's ratio and optical response in plasmonic composites mimicking the behavior of hyper-auxetic networks with ultralow cross-linking density. <ref type="bibr">83</ref>   The Journal of Physical Chemistry C</p></div>
<div xmlns="http://www.tei-c.org/ns/1.0"><head>&#9632; AUTHOR INFORMATION</head></div><note xmlns="http://www.tei-c.org/ns/1.0" place="foot" xml:id="foot_0"><p>https://doi.org/10.1021/acs.jpcc.2c05353 J. Phys. Chem. C 2022, 126, 16885-16893 Downloaded via UNIV OF TEXAS AT AUSTIN on May 9, 2023 at 20:16:08 (UTC).See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.</p></note>
			<note xmlns="http://www.tei-c.org/ns/1.0" place="foot" xml:id="foot_1"><p>https://doi.org/10.1021/acs.jpcc.2c05353 J. Phys. Chem. C 2022, 126, 16885-16893</p></note>
		</body>
		</text>
</TEI>
