**2.1 Fabrication of self-assembled templates**

Amphiphilic di-block copolymers can self-assemble into reverse micelles when dissolved in a solvent that selectively dissolves only the apolar block of the copolymer. These reverse micelles can be obtained or be induced into attaining a spherical morphology, with a feature size determined by their aggregation number. The micelle aggregation number is not only a function of the molecular weight and composition of the copolymer but also the quality of the solvent used and the presence of additives. As a consequence, the size can be varied independent of the molecular weight, using solvent quality and concentration of additives as useful handles to fine-tune template geometries and eventually those of resulting plasmonic arrays. The spherical micelles are solvent-laden colloidal structures which can be readily organized on a variety of surfaces to yield a 2D hexagonally ordered dot-array templates. These templates are subjected to physical or chemical means of pattern transfer to produce nanopatterns of desired materials. Pattern formation on surfaces using copolymer reverse micelle approach is fundamentally different from that of microphase separation in block copolymer thin films [37–39]. The size of reverse micelles in solution, which eventually determines the feature sizes on the surface, is independently variable by engineering the solvent quality or the use of additives. The standard deviation of the templates on the surface is determined by that of the micelles in solution phase before deposition, which in turn is governed by the intermicellar exchange process. The exchange process is slow due to the slower diffusion for larger molecular weight polymers or when solvent with high selectivity to the corona-forming blocks is employed [40]. During the spin-coating process, the solvent-laden micelles in solution deform on the surface to assume an ellipsoidal shape, with partial fusion of corona from adjacent reverse micelles resulting in the globally continuous organic film presenting periodic contrast in topography with an ultrathin film (<5 nm) in the background. The pitch of the ensuing pattern on the surface can be varied in steps <5% of its mean value, through control over evaporation rate or the concentration of the micelles in solution. The evaporation rate can be controlled using spin or dip coating speeds. The ability to vary the lattice periodicity within a certain window is attributed to a range of distances for which the PS blocks from the corona of adjacent micelles can still meet upon deformation and film formation. When this condition is not met, namely, at low solution concentrations or high spin speeds, the reverse micelles are spaced too far apart, resulting in patchy coverage. On the other hand, at high solution concentrations or at low spin speeds, the excess concentration beyond what is necessary toward a monolayer appears as multilayers [36]. The topography of the reverse micelle film is a variable that can be determined by the relative humidity in the ambient environment during the coating process. This is attributed to the change in moisture that is likely to concentrate at the polar core-forming PVP and consequently increasing the interfacial tension resulting in resistance to collapse and as a result a higher topography. Under optimal conditions of coating, micelle arrays can be produced with a standard deviation of <10% in geometric attributes across a complete wafer.

In a specific example shown in **Figure 2**, reverse micelles of polystyreneblock-poly (2-vinylpyridine) (PS-b-P2VP) are obtained from dilute solutions of m-xylene. Here, a copolymer with a molecular weight of 81.5 kDa and a PDI of 1.10, at a concentration of 0.5% w/w in m-xylene, is spin-coated on a clean silicon surface at 5500 rpm resulting in a hexagonally ordered array with feature heights of 20 nm and pitch of 66 nm with typical standard deviations <15% in all geometric

**117**

*Nanoplasmonic Arrays with High Spatial Resolutions, Quality, and Throughput…*

variables. During the spin-coating process, the solvent-laden micelles in solution deform on the surface to assume an ellipsoidal shape, with the corona from adjacent reverse micelles coming together resulting in the globally continuous thin film presenting periodic topography and ultrathin film (<5 nm) in the background. The center-to-center distance or the pitch of the template arrays could be systematically decreased in steps of <5 nm between 45 and 60 nm by increasing the solution concentrations from 0.6 to 1% at a fixed spin-speed of 5000 rpm or decreasing spin-speeds from 9000 to 2000 rpm at a fixed concentration of 0.7%. The ranges were found to constitute an optimal window of conditions where a continuous

*Schematic representation of the self-assembly of amphiphilic di-block copolymer into reverse micelles and their* 

Among key limitations encountered by self-assembly-based approaches, in general, is the scaling up to practically large areas while ensuring high consistency and reproducibility. The issues of reproducibility arise mainly due to the sensitivity of the process outcome to environmental parameters. Such sensitivity also limits process scalability, due to inconsistencies encountered when coating large areas like full wafers and to limited batch-to-batch reproducibility. These issues are true also for the case of self-assembly of amphiphilic copolymers, and this can be addressed only by adequate investigations directed at mapping the impact of different environmental variables on the process outcomes. Several sources of variability were identified and addressed for the self-assembly of amphiphilic copolymers, including the presence of moisture and contaminants in solution; history of preparation (agitation and incubation), temperature, and humidity; differences in surface roughness or surface energy (e.g., due to organic or particulate contaminants on the surface); changes to solution concentrations due to solvent evaporation during use; and inadequate mixing of polymer. Under optimal conditions, the assemblies of reverse micelle feature exhibit standard deviations lower

A specific outcome of optimization of the templates represented in **Figure 2c** on a 100 mm silicon wafer can be seen in **Figure 3**, which shows the distribution for height, diameter, pitch, and nearest neighbors, at different regions of the wafer. Typical characterization involves AFM topography for heights and diameters (within errors of AFM tip-convolution) and SEM top view and image analysis using ImageJ or MATLAB. A representative AFM image recorded in the tapping mode is shown in **Figure 3b**. The characterization is critical for each batch of samples, and it is possible to scale the process to several batches of wafers [41]. Points A, B, C, and D correspond to four positions representing systematically increasing radial distances from the center to the edge of the wafer. AFM measurements at these points show standard deviation <15% for geometric variables and <10% variation of their mean values across the full wafer. Voronoi analysis of the AFM images shows a predominance of

*DOI: http://dx.doi.org/10.5772/intechopen.89064*

*subsequent assembly to form two-dimensional arrays on the surface.*

uniform film was obtained.

**Figure 2.**

than 15% across full wafers.

**2.2 Reproducibility and scalability**

*Nanoplasmonic Arrays with High Spatial Resolutions, Quality, and Throughput… DOI: http://dx.doi.org/10.5772/intechopen.89064*

**Figure 2.**

*Nanoplasmonics*

**2. Spatially controlled fabrication of nanoscale templates**

Amphiphilic di-block copolymers can self-assemble into reverse micelles when dissolved in a solvent that selectively dissolves only the apolar block of the copolymer. These reverse micelles can be obtained or be induced into attaining a spherical morphology, with a feature size determined by their aggregation number. The micelle aggregation number is not only a function of the molecular weight and composition of the copolymer but also the quality of the solvent used and the presence of additives. As a consequence, the size can be varied independent of the molecular weight, using solvent quality and concentration of additives as useful handles to fine-tune template geometries and eventually those of resulting plasmonic arrays. The spherical micelles are solvent-laden colloidal structures which can be readily organized on a variety of surfaces to yield a 2D hexagonally ordered dot-array templates. These templates are subjected to physical or chemical means of pattern transfer to produce nanopatterns of desired materials. Pattern formation on surfaces using copolymer reverse micelle approach is fundamentally different from that of microphase separation in block copolymer thin films [37–39]. The size of reverse micelles in solution, which eventually determines the feature sizes on the surface, is independently variable by engineering the solvent quality or the use of additives. The standard deviation of the templates on the surface is determined by that of the micelles in solution phase before deposition, which in turn is governed by the intermicellar exchange process. The exchange process is slow due to the slower diffusion for larger molecular weight polymers or when solvent with high selectivity to the corona-forming blocks is employed [40]. During the spin-coating process, the solvent-laden micelles in solution deform on the surface to assume an ellipsoidal shape, with partial fusion of corona from adjacent reverse micelles resulting in the globally continuous organic film presenting periodic contrast in topography with an ultrathin film (<5 nm) in the background. The pitch of the ensuing pattern on the surface can be varied in steps <5% of its mean value, through control over evaporation rate or the concentration of the micelles in solution. The evaporation rate can be controlled using spin or dip coating speeds. The ability to vary the lattice periodicity within a certain window is attributed to a range of distances for which the PS blocks from the corona of adjacent micelles can still meet upon deformation and film formation. When this condition is not met, namely, at low solution concentrations or high spin speeds, the reverse micelles are spaced too far apart, resulting in patchy coverage. On the other hand, at high solution concentrations or at low spin speeds, the excess concentration beyond what is necessary toward a monolayer appears as multilayers [36]. The topography of the reverse micelle film is a variable that can be determined by the relative humidity in the ambient environment during the coating process. This is attributed to the change in moisture that is likely to concentrate at the polar core-forming PVP and consequently increasing the interfacial tension resulting in resistance to collapse and as a result a higher topography. Under optimal conditions of coating, micelle arrays can be produced with a standard deviation of <10% in geometric attributes across a

In a specific example shown in **Figure 2**, reverse micelles of polystyreneblock-poly (2-vinylpyridine) (PS-b-P2VP) are obtained from dilute solutions of m-xylene. Here, a copolymer with a molecular weight of 81.5 kDa and a PDI of 1.10, at a concentration of 0.5% w/w in m-xylene, is spin-coated on a clean silicon surface at 5500 rpm resulting in a hexagonally ordered array with feature heights of 20 nm and pitch of 66 nm with typical standard deviations <15% in all geometric

**2.1 Fabrication of self-assembled templates**

**116**

complete wafer.

*Schematic representation of the self-assembly of amphiphilic di-block copolymer into reverse micelles and their subsequent assembly to form two-dimensional arrays on the surface.*

variables. During the spin-coating process, the solvent-laden micelles in solution deform on the surface to assume an ellipsoidal shape, with the corona from adjacent reverse micelles coming together resulting in the globally continuous thin film presenting periodic topography and ultrathin film (<5 nm) in the background. The center-to-center distance or the pitch of the template arrays could be systematically decreased in steps of <5 nm between 45 and 60 nm by increasing the solution concentrations from 0.6 to 1% at a fixed spin-speed of 5000 rpm or decreasing spin-speeds from 9000 to 2000 rpm at a fixed concentration of 0.7%. The ranges were found to constitute an optimal window of conditions where a continuous uniform film was obtained.

### **2.2 Reproducibility and scalability**

Among key limitations encountered by self-assembly-based approaches, in general, is the scaling up to practically large areas while ensuring high consistency and reproducibility. The issues of reproducibility arise mainly due to the sensitivity of the process outcome to environmental parameters. Such sensitivity also limits process scalability, due to inconsistencies encountered when coating large areas like full wafers and to limited batch-to-batch reproducibility. These issues are true also for the case of self-assembly of amphiphilic copolymers, and this can be addressed only by adequate investigations directed at mapping the impact of different environmental variables on the process outcomes. Several sources of variability were identified and addressed for the self-assembly of amphiphilic copolymers, including the presence of moisture and contaminants in solution; history of preparation (agitation and incubation), temperature, and humidity; differences in surface roughness or surface energy (e.g., due to organic or particulate contaminants on the surface); changes to solution concentrations due to solvent evaporation during use; and inadequate mixing of polymer. Under optimal conditions, the assemblies of reverse micelle feature exhibit standard deviations lower than 15% across full wafers.

A specific outcome of optimization of the templates represented in **Figure 2c** on a 100 mm silicon wafer can be seen in **Figure 3**, which shows the distribution for height, diameter, pitch, and nearest neighbors, at different regions of the wafer. Typical characterization involves AFM topography for heights and diameters (within errors of AFM tip-convolution) and SEM top view and image analysis using ImageJ or MATLAB. A representative AFM image recorded in the tapping mode is shown in **Figure 3b**. The characterization is critical for each batch of samples, and it is possible to scale the process to several batches of wafers [41]. Points A, B, C, and D correspond to four positions representing systematically increasing radial distances from the center to the edge of the wafer. AFM measurements at these points show standard deviation <15% for geometric variables and <10% variation of their mean values across the full wafer. Voronoi analysis of the AFM images shows a predominance of

#### **Figure 3.**

*Demonstration of scalability combined with uniformity on full wafers of optimized coatings: (a) Schematic of radially separated points from center to edge of 100 mm wafer, (b) where tapping mode AFM measurements are performed (image indicated for point a), (c) the distribution of height and pitch across full wafer, (d) with mean values plotted as function of radial distance from center, and error bars showing standard deviation in corresponding feature dimensions at a single point, (e) histogram of nearest neighbors showing predominantly six nearest neighbors as expected for hexagonal assembly, maintained across the wafer, and (f) representation of nearest neighbors using Voronoi analysis, with the features colored corresponding to the number of nearest neighbors as indicated in (e).*

six nearest neighbors as expected for hexagonal packing, which is uniformly maintained across the wafer (**Figure 3e, f**). The outcome clearly demonstrates the feasibility for reliable scaling up of the technique to cater to nanostructures over large areas.

## **3. Nanoplasmonic arrays by pattern transfer**

Fabrication of plasmonic nanoarrays starting from organic templates relies on pattern transfer approaches, for example, template-guided growth, deposition, or etching, to define noble metal nanoarrays with the conservation of pitch from

**119**

**Figure 4.**

*Nanoplasmonic Arrays with High Spatial Resolutions, Quality, and Throughput…*

the original template. The pattern transfer approaches offer large flexibility in the geometry of the individual features, thus making it possible to fabricate plasmonic nanoarrays of different types, for example, nanoparticle cluster arrays, nanopillars arrays, nanorods, or nanodiscs. The pattern transfer parameters provide independent control over the size, shape, and aspect ratio of the features and should be optimized to ensure that they do not affect the spatial arrangement, uniformity, and reproducibility from the original template. Although the pattern transfer approaches are common in semiconductor fabrication, extending them to work at the scale of few nanometers requires rigorous optimization and quality assurance to ensure low standard deviations and reproducibility in geometries. Further in this section different pattern transfer methods to reach such three different plasmonic

*Fabrication of plasmonic nanoarrays with different profiles, size, and distributions of metal nanostructures can be attained by control over pattern transfer processes. In all cases, the advantages of the original templates* 

Clusters (used interchangeably with "aggregates" in this report) of metal nanoparticles behave differently from their isolated counterparts due to the collective optical behavior arising out of plasmonic coupling between the constituent nanoparticles [45–50]. Clusters of nanoparticles are known to behave as hot particles, with significantly enhanced electromagnetic fields at the inter-particle junctions [51–54]. Consequently, clusters exhibit higher extinction cross sections, with hot spots that can be excited at lower energies than the isolated particles. The aggregation-induced color change of gold nanoparticle suspensions caused by the analyte of interest has been the basis of several biological assays [55]. Such random aggregation typically results in a large distribution in the number of particles per cluster, with a distinct lack of control over those numbers. To achieve clusters with desired optical properties, it is essential to be able to produce them with a narrow distribution in the size, shape, and spatial arrangement between nanoparticles within the cluster and between clusters in an array. Such clusters were demonstrated using templates

nanoarrays have been discussed in detail (**Figure 4**).

*including the uniformity and scalability are preserved [42–44].*

**3.1 Nanoparticle cluster arrays**

*DOI: http://dx.doi.org/10.5772/intechopen.89064*

*Nanoplasmonic Arrays with High Spatial Resolutions, Quality, and Throughput… DOI: http://dx.doi.org/10.5772/intechopen.89064*

**Figure 4.**

*Nanoplasmonics*

**118**

**Figure 3.**

*neighbors as indicated in (e).*

six nearest neighbors as expected for hexagonal packing, which is uniformly maintained across the wafer (**Figure 3e, f**). The outcome clearly demonstrates the feasibility for reliable scaling up of the technique to cater to nanostructures over large areas.

*Demonstration of scalability combined with uniformity on full wafers of optimized coatings: (a) Schematic of radially separated points from center to edge of 100 mm wafer, (b) where tapping mode AFM measurements are performed (image indicated for point a), (c) the distribution of height and pitch across full wafer, (d) with mean values plotted as function of radial distance from center, and error bars showing standard deviation in corresponding feature dimensions at a single point, (e) histogram of nearest neighbors showing predominantly six nearest neighbors as expected for hexagonal assembly, maintained across the wafer, and (f) representation of nearest neighbors using Voronoi analysis, with the features colored corresponding to the number of nearest* 

Fabrication of plasmonic nanoarrays starting from organic templates relies on pattern transfer approaches, for example, template-guided growth, deposition, or etching, to define noble metal nanoarrays with the conservation of pitch from

**3. Nanoplasmonic arrays by pattern transfer**

*Fabrication of plasmonic nanoarrays with different profiles, size, and distributions of metal nanostructures can be attained by control over pattern transfer processes. In all cases, the advantages of the original templates including the uniformity and scalability are preserved [42–44].*

the original template. The pattern transfer approaches offer large flexibility in the geometry of the individual features, thus making it possible to fabricate plasmonic nanoarrays of different types, for example, nanoparticle cluster arrays, nanopillars arrays, nanorods, or nanodiscs. The pattern transfer parameters provide independent control over the size, shape, and aspect ratio of the features and should be optimized to ensure that they do not affect the spatial arrangement, uniformity, and reproducibility from the original template. Although the pattern transfer approaches are common in semiconductor fabrication, extending them to work at the scale of few nanometers requires rigorous optimization and quality assurance to ensure low standard deviations and reproducibility in geometries. Further in this section different pattern transfer methods to reach such three different plasmonic nanoarrays have been discussed in detail (**Figure 4**).
