Application of helium ion microscopy to nanostructured polymer materials

2.1 Parameters of HIM as an imaging system

Helium is more massive than an electron and therefore, possesses greater momentum, with DeBroglie wavelength ∼100× smaller than a SEM beam, which allows for a higher theoretical ultimate resolution [18]. All scanning particle beam microscopes, whether they are electron or ion based, exploit the detection of secondary electrons that are generated through the interaction of a charged particle with a surface. The added momentum of the helium ion also produces a higher yield of secondary electrons (SE) than an electron beam. The typical SE yield of an SEM is around one SE per electron that hits the sample, whereas the SE yield from the HIM has been shown to be between two and eight secondary electrons per ion [21]. However, beyond the increase in secondary electron production by helium ion/surface interactions, there are other differences in the interaction of a helium ion with surface materials that generate different effects and create alternative potential for HIM.

2.2 Contrast and resolution in HIM

As with all scanning particle beam microscopy, three factors within a sample contribute to the particle/surface interaction that will affect image/contrast formation: 1) sample composition and chemistry; 2) topological variation of the sample surface; and 3) structural heterogeneity within the sample. The chemical composition of a sample determines the nature of the particle/surface interaction and, therefore, plays a major factor regarding image formation in particle beam microscopy. Owing in part to its greater mass, the interaction volume of a helium ion with a given material is smaller than that of an electron of the same energy (Figure 2) [26]. For example, a secondary electron image produced in the HIM with a 20-kV accelerating voltage resembles more a SEM image obtained using 0.5 kV accelerating voltage as can be seen in TRIM and EFS calculations [18]. Heterogeneity in the chemical composition of a sample may also generate differences in contrast. In some cases, this will permit limited chemical characterization of such materials using an HIM. This type of chemical analysis has been performed with both low voltage SEM and HIM to measure dopant concentration in semiconductor material, but the greater resolving power of HIM allows for more precise application of this technique into nanoscale features [27, 28]. While many polymer materials are composed of molecules and are nonconductive, in a similar fashion, HIM may be useful in characterizing nanoscale polymer composites doped with conductive fillers.

Figure 2

Comparison of the interaction volume of an electron and a helium ion interaction with a carbon surface. (A) Simulation of the interaction volume generated by a 10-kV electron beam bombarding a carbon surface generated by the Win-X-ray program (http://montecarlomodeling.mcgill.ca/software/winxray/features.html); the scattering image includes backscattering electrons. (B) Simulation of the interaction volume generated by a 10-kV helium ion beam bombarding a carbon surface generated by SRIM program.

Topological variation of a sample’s surface, such as the alteration to the surface roughness or the presences of high aspect ratios of surface features will alter image contrast. Another topological consideration important for image/contrast formation is shadowing, in which a surface structure blocks a portion of the sample from the ion beam. This is particularly true when imaging nanoscale topologies that are composed of nonconductive materials (as most polymeric materials where the accumulated surface charge cannot escape). Moreover, the larger interaction volume of an electron beam may completely encompass a nanoscale structure being observed creating additional resolution problems for SEM imaging of nanostructures polymeric material. In traditional SEM, the issues in imaging nonconductive biological and polymeric samples have been mediated somewhat by lowering the beam energy [29, 30]. Helium ions having more localized interactions with surface materials, have an inherent reduction in some of the potentially confounding effects observed in SEM when imaging nanostructured features with high aspect ratios, even those composed of nonconductive material [18, 20]. Furthermore, when imaging thin materials with HIM, interactions within the sample material are able to generate ion-induced SE from adjacent objects, which can be useful in fields ranging from materials science to biological sciences [18, 31].

Structural variance within a chemically homogenous sample, itself, also affects contrast formation with HIM. It may be possible to characterize the crystalline properties of polymers using ion channeling in HIM. Ion channeling is the phenomena in which an ion beam differentially interacts with a crystalline lattice of a material in an angle-dependent manner and is used as the basis of analytical techniques such as Rutherford backscattering [32–34]. Although the operating energies of the helium ion beam in HIM is significantly lower than in techniques such as Rutherford backscattering, HIM has been used to characterize a polycrystalline gold film and describe orientation-dependent contrast formation [33].

3.1 Polymer-based composites

Composite materials dominate the modern material world and are extremely important from both scientific and technological viewpoints. Polymer-based composites are widely used in construction, transportation, and aerospace industries. Many of such materials have been designed and tested to have a unique set of properties required for some niche applications. Very often, a synergistic effect can be achieved when the properties of the composite are significantly exceeding the additive sum of those of the constituting components thus offering superior mechanical, thermal, or degradation-resistance properties. In polymer-based nanocomposites, mechanical flexibility and processibility of a polymer (which is used as a continuous phase or matrix) can be combined with unique properties of a nanostructured filler (such as high electrical conductivity, magnetic properties, dielectric constant, etc.). The filler may be of inorganic or polymer nature and can be of an isotropic (nanoparticles), anisotropic (nanowires, nanofibers, nanotubes), or hierarchical (dendrimers) shape. The nanofiller can be synthesized in situ in functional polymer matrix or can be dispersed in it during fabrication protocols (solution or melt mixing) [35].

Carbon nanotubes (CNTs) have been introduced more than two decades ago and possess unique mechanical, electrical, and thermal conductivity properties, which makes them very attractive as a filler material for developing conducting polymer composites for modern engineering applications. The high aspect ratio of the CNTs significantly lowers their electrical conductivity percolation threshold compared to the conventional polymer/carbon black composites. Therefore, polymer/CNT composites have been suggested for many high tech applications including RF interference shielding coatings, memory devices, field effect transistors, and ultrasensitive physical, chemical, and biological sensors [36].

Aggregation of nanoparticles or nanotubes in the polymer matrix represents one of the major problems in the creation of polymer nanocomposites. CNTs have a natural trend for the formation of bundles – thicker fibers constituted with seven or more individual nanotubes, thus, significantly constraining the final characteristics of the material. Therefore, both new technological solutions for CNTs dispersion and nondestructive methods of their structural control are required. Figure 3A,B demonstrates CNT bundles as viewed by HIM and by AFM techniques. Visible lateral cross-sectional dimensions of thread-like features in AFM topography image (Figure 1B) are exaggerated due to the effect of AFM tip shape convolution with cylindrical features attached to a flat substrate (silicon wafer surface). The real diameter of the features can be extracted from the height distribution analysis (inset) and gives an evidence that they are aggregates (bundles of 10–30 nm diameter) constituted with seven or more individual single-walled carbon nanotubes of 1.3 nm diameter [37]. HIM image of the same nanotubes (Figure 3A) explicitly shows these aggregated structures and reveals the details of the bundle formation. Moreover, self-suspended nanotubes without substrate can be imaged. The bottom portion of Figure 3 shows comparison between imaging capabilities of the ultrasonic mode of scanning probe microscopy (so-called ultrasonic force microscopy (UFM)) and HIM in visualization of nanotubes hidden underneath a polymer composite surface. As demonstrated [38], polyurethane (PU) can be used as a flexible matrix promoting fine dispersion of the nanotubes. Figure 3D has an enhanced contrast associated with the difference of interaction of ultrasonic waves with the matrix and the filler. The PU matrix is significantly damping sound wave propagation and appears as a dark background in the image. Owing to a much higher modulus of elasticity (∼1 TPa for CNTs vs. ∼100 MPa for PU), the nanotubes appear as bright lines in UFM allowing their easy identification and mapping. HIM imaging (Figure 3C) also shows high contrast for PU-CNT systems but because of a different reason. CNTs are electrically conductive, while the matrix is a dielectric material. Charge density distribution imposed during ion beam exposure causes high contrast imaging in this case. Generally, HIM technique has an advantage of a more simple (no special sample preparation) nanometer-scale imaging of polymer-CNT composites revealing details of the nanotube dispersion and aggregation within the polymer matrix.

Figure 3

Comparison of SPM and HIM imaging of polymer-carbon nanotube composites. (A) HIM image of a CNT bundle, (B) AFM topography image of single wall carbon nanotube bundles (constituted of seven individual nanotubes on average) on silicon wafer surface (inset shows cross-sectional scan with peak heights corresponding to nanotube bundle diameters), (C) HIM image of polymer-impregnated carbon nanotubes, (D) UFM image of CNT-polyurethane nanocomposite surface [Image 3c reproduced with permission from Carl Zeiss Microscopy, Peabody, MA, USA].

3.2 Polymer nanoparticles and nanofibers

As has been mentioned already, polymers are “natural” nanomaterials due to the fact that common macromolecular dimensions fall in the range of a few to several tens of nanometers. Some methods of polymer synthesis (emulsion and suspension polymerizations) or physical techniques (self-assembly, Langmuir-Blodgett, layer-by-layer deposition) can be used for shaping of polymer chains in nanoscale objects like nanoparticles, nanofibers, or ultrathin films. A typical example of such processes is polymer latexes. Polymer colloids or latexes are common for application ranging from cosmetics and pharmaceuticals to adhesives and coating [39]. Latexes have typical diameters in the range of a few hundreds of nanometers, and therefore, the entire surface created by a few grams of such polymer may be of the order of 100 m². As a result, physical properties of latex suspensions such as their stability are almost fully determined by their surface properties. Latex suspensions are therefore good model systems for the study of surface-related problems (so-called “soft condensed matter”) [40–42]. Polymer colloidal particles can be also used as “building blocks” for construction of larger mesoscale structures. After the formation of the final film or coating, the individual particle identity is vanished to produce a homogeneous material [43]. A key advantage of such approach is that it offers fine control of the final structure at different levels from the nanoscale (within particles) to meso- and even macroscale through the creation of ordered assemblies [8].

Even more sophisticated structures can be created with core-shell nanoparticles enabling inherent molecular level compatibility of the components and otherwise mutually exclusive properties (e.g., electrically conductive shell on insulating core) [44–46]. This approach has found widespread attention especially recently in connection with development of printed electronics and 3D printing technologies [47–50]. Application of core-shell polymer nanoparticles for creation of highly porous functional materials suitable for chemical sensor applications and their characterization with HIM was demonstrated earlier [51]. Functional sensing shell has been prepared from intrinsically conductive polymer polyaniline (PANI) whose conductivity depends on the environment. Moreover, the shell could be structured at the nanoscale through application of special polymerization approaches [52–54]. Simultaneously, relatively bulky core possessed mechanical stability and was used as a solid frame for the final sensing material.

The core can introduce special functions in the system. For instance, a core-shell nanoparticle can introduce desirable physical properties like macroscopic polarization, optical transparency, mechanical toughness, and thermal processability. The nanoscale functional shell can be formed on top of relatively large (micrometer-sized) cores. As pointed out in [55] the core-shell synthetic approach enables reduction of human health and environmental risks associated with formation of aero-sols in existing nanofabrication protocols. The authors introduced the way of creation of nanostructured electrically conductive composites based on submicron core-shell particles with poly(vinylidene fluoride) core and PANI shell. The PANI component could be additionally doped with various dopants, thus, broadening the spectrum of possible functionalities. PVDF offers good processability, chemical inertness against many organic solvents, acids, and bases, and thermal stability [56]. The morphology and electrical properties for the PVDF-PANI core-shell system were studied with HIM depending on the preparation conditions (Figure 2). The possibility of preparing stable and mechanically strong PVDF/PANI (DBSA) films with excellent sensing properties has been demonstrated. HIM studies have revealed that application of short-time annealing, above the melting point of the PVDF matrix, made possible additional adjustment of the structure and conductivity properties of the material. Image A in Figure 4 shows as-prepared core-shell particles. The particles are characterized with a smooth surface and narrow size distribution. The electrical contact between conductive PANI shells is not perfect, and therefore, the deposited film still possesses insulating properties. When annealed at 180°C for a short time (Figure 4B), simultaneous softening of the PVDF core and welding of the shells leads to formation of a porous monolithic film and an electrically percolation system achieved with a relatively small (∼20%) volume fraction of the conductive component. The resulting nanoporous films consisting of tightly bound PVDF/PANI(DBSA) core-shell submicron particles with PANI shells of up to 10 nm thickness exhibited high sensitivity to gaseous ammonia with a ppb range detection limit. Interestingly, due to an incomplete doping of the polymer chains, the nanocomposite material was also sensitive to gaseous HCl, however, in a narrower concentration range.

Figure 4

HIM imaging of PVDF-PANI core-shell nanoparticles. (A) nanoparticle aggregates formed after casting from chloroform dispersion; (B) porous monolithic film formed after short term annealing of the solution cast film at 180°C (above the melting transition temperature of the PVDF core).

Polymer nanofibers are another example of nanostructured polymer materials. They can be produced via electrospinning protocols [57, 58] or, in some cases, just due to spontaneous self-assembly process [16]. Figure 5 demonstrates high applicability of HIM in direct imaging of polymer nanofibers. The left image (A) shows an electro-span poly(lactic acid) nanofiber. This polymer is an example of a biodegradable, thermoplastic, aliphatic polyester derived from renewable resources [59–63]. Several distinct forms of polylactide exist due to the chiral nature of the lactic acid. Poly-l-lactide (PLLA) is characterized with a moderate degree of crystallinity (∼37%), has T_g between 50°C and 80°C, and a T_m between 173°C and 178°C. PLLA is currently used in a number of biomedical applications and is also evaluated as a material for tissue engineering (scaffolds). Image A shows the onset of formation of mineral particles of the surface of the fibers (mineralization process). Nucleation of inorganic phase and development of mineralization can be studied with nanoscale resolution with HIM without special sample preparation requirements. A characteristic waviness of the fibers, which is a result of flow instabilities during electrospinning process, can be also observed and characterized in detail. The observed nanoscale roughness of the PLLA fibers helps in creation of tight attachment of the inorganic phase to polymer nanofiber frame. The right image (B) shows polyaniline (PANI) nanofiber spontaneously formed due to self-organization process during polymer synthesis [51]. Nanostructuring of PANI as a conductive polymer with environmentally dependable conductivity allowed the fabrication of a supersensitivity to ammonia vapor chemical sensors through simple conductometric transducer scheme [51]. Electron microscopy and diffraction have confirmed a high degree of orientation of PANI molecules within the fibers, which is in tune with the observed molecular level smoothness of their surface. The image shows that the PANI nanofiber and supporting polished silicon wafer surface have similar roughness (i.e., on a level of 0.4 nm). In both examples of Figure 5, the details of the surface roughness and texturing on the fibers could be revealed only due to the application of HIM technique. These details are lost in AFM imaging because of high local curvature of the surface (small radius of fibers) and difficulties of their tight attachment to a substrate required for high-resolution scanning. SEM imaging does not offer high resolution due to significant surface charging of insulating polymer materials. TEM does offer atomic resolution imaging and chain orientation, however, only with a complex sample preparation procedure. The surface topography details are typically lost in TEM images due to relatively low contribution from the surface to the image contrast in comparison to the “bulk fiber” electron scattering.

Figure 5

HIM imaging of polymer nanofibers: (A) electrospan PLLA nanofibers used as a scaffold (mineralization process can be seen on the surface); (B) self-organized fibers of polyaniline polymer. Insets show higher magnification images [Reproduced with permission from Carl Zeiss Microscopy, Peabody, MA, USA].

3.3 Nanoporous polymer systems

Nanoporous polymer beads represent another challenging system for microscopy studies. Porous polymer resins are prospective materials for hydrogen storage, as a sorption media as well as for sensor applications. In the latter case, a porous polymer membrane or porous beads of desirable size and high specific surface area should be functionalized to enable selective binding of targeting molecules or atoms on their surface. The adsorption of the analyte species on the surface can be then monitored either directly by detection of emitted light intensity (if the analyte molecules have fluorescence centers) or indirectly via incorporation of specially designed fluor molecules to convert radiation (X-rays, γ-rays, electrons, α-particles, or neutrons) into energy of light. Recent reviews demonstrated several successful approaches for synthesis of porous polymer particles [64, 65]. The porosity of the fabricated beads can be controlled by the type and concentration of a porogen and the degree of polymer matrix crosslinking. In our recent study, we have applied three different porogens in order to achieve a significant variation of the nanopore distribution within the same polystyrene (PS) matrix during suspension polymerization fabrication protocol.

Pore structure characterization of polymer beads is not a trivial task. The pore size distribution analysis of polymer materials can be accomplished by surface porosimetry techniques and modeling sorption-desorption curves via application of some of theoretically developed approaches [66]. This indirect analytical technique is generally accepted and gives reliable data. Alternatively, the same data can be also be extracted from small-angle X-ray or neutron scattering techniques [67, 68]. However, the direct visualization of the pore structure with pore size ∼2 nm or smaller has not been possible until recently due to technical limitations of traditional microscopy techniques (TEM, SEM, AFM). In our study, we compare the morphological features and porosity of polystyrene-based polymer resins prepared using suspension polymerization technique with the application of three different porogens resulting in three different types of porous structures. The direct nanoscale resolution imaging of PS beads surface was possible only with the application of HIM (Figure 6). Figure 6 shows comparative study of PS beads prepared with three different porogens. The left images are low-magnification pictures showing the overall quality of the beads and their surface roughness. One can see similar smooth surfaces for the samples prepared with toluene and n-heptane porogens, while the addition of a nonionic span-80 surfactant (water in oil in water, w/o/w, or double emulsion approach) introduces highly developed surface to the system. Some additional contrast associated with surface charging (the beads are studied in their native state) during imaging is clearly observed but does not affect the quality of the images. Similar pictures obtained with SEM are reported in many related publications but require thermally evaporated metallic layer on the bead surface. The evaporated metallic layer masks the intimate details on the surface and limits the resolution of SEM technique to 20–50 nm (the typical size of metallic film domains). On the contrary, high-resolution imaging of pores is still possible with HIM technique. On the right images of Figure 6, the nanometer scale details of the bead surface are clearly visible. The contrast arises from the different secondary electron emissions from the top surface and the surface of the caves (pores). Comparison of the size and distribution of dark spots representing pores show that toluene porogen promotes formation of the smallest pores (average size 2–3 nm), followed by n-heptane porogen (∼5–10 nm pore size), and by span-80 (w/o/w) system with a very broad pore size distribution (30–50 nm).

Figure 6

HIM images of nanoporous polystyrene beads prepared via suspension polymerization with three different porogens: toluene (A, B), n-heptane (C, D), and Span-80 (E, F) (see explanation in the text). Low-resolution images (left) give insight in the bead’s size distribution and surface roughness, while high-resolution images (right) allow visualization and direct measurements of the pore size and pore size distribution not available in other microscopy techniques.

3.4 Surface crystallization in ultrathin polymer films

The ability of long polymer chains to form ordered structures during the crystallization process is one of the most striking phenomena in condensed matter physics [69]. The phenomenon has been studied since the origin of polymer science both by the real space imaging methods (TEM, SEM, AFM, etc.) and diffraction techniques (electron diffraction, wide angle X-ray diffraction, small-angle X-ray scattering, reflection high-energy electron diffraction, etc.) The real space imaging methods are usually applied to probe morphology, crystallization kinetics, and melting behavior, while the diffraction techniques can detect the crystalline lattice, lamellar orientation, crystallinity, and transition temperatures [70].

The most common morphological superstructures spontaneously formed during polymer crystallization from melt are spherulites, which are polycrystalline and semicrystalline in nature. Formation of polymer single crystals can be observed when crystallization happens from dilute solutions. This leads to formation of platelet shape single crystals (lamellar) first reported with TEM technique in the later 1950s [71–73]. The lamellar single crystals have one dimension (thickness) in the range of several or tens of nanometers and lateral dimensions of a few micrometers and represent the basic unit of crystalline polymers, of which spherulites and other polycrystalline aggregate are composed.

Especially interesting is the investigation of the crystallization phenomenon in thin polymer films with the thickness comparable with the radius of gyration of polymer chains. Owing to geometrical confinement, such polymer materials possess structural organization and physical properties different from the bulk state [74–76]. Thin polymer films and coatings constitute an essential portion of modern organic materials and find applications in many branches of industry including packaging, printing technology, and plastic electronics. Not surprisingly, structural organization and crystallization phenomenon of polymer materials in ultrathin films has been a focus of several recent research efforts [74–76]. The variety of crystallization protocols reported in literature for thin polymer films depending on the film thickness can be reduced to three characteristic types: i) spherulitic (or bulk-like crystallization observed in relatively thick films with the thickness, d, much higher than the radius of gyration of the polymer chains R_g); ii) dendritic crystallization for d∼R_g; and iii) quasi-two-dimensional crystallization in the case of highly confined films (d<R_g) [77–82]. In the latter case, the crystallization process can be significantly modified by a process of polymer layer dewetting from the solid surface (in case of weak polymer-substrate interaction) or can be suppressed (in case of strong polymer-substrate interaction) [83–85].

Generally, crystallization in very thin films comparable to the R_g can be described as diffusion-limited crystallization (DLC) [77–82, 86]. This term should be understood both in the sense of limitations in chain mobility, which is increasing with temperature, and in terms of limited amount of polymer material (just a few polymer chains across the film thickness) capable for rearrangement necessary for the crystalline phase formation. The first limitation can be referred to as a kinetic constrain, while the second one is a geometrical constrain (i.e., can be treated as temperature-independent parameter). As demonstrated by several publications, both glass-transition and crystallization transition points (T_g and T_c) will be affected by the film thickness and vicinity of the solid surface [87–91]. This is due to enhanced mobility of the surface layer in comparison to the bulk material (relative contribution from the surface property will increase as the film thickness is reduced).

Morphological features of crystallization in thin polymer films have been intensively studied during the last decade [70]. Despite a number of publications on surface crystallization of various polymer systems, the real understanding of the physical process is still to be achieved. Therefore, a systematic study of the crystallization process and the final crystalline morphology for well-characterized set of polymers with application of high-resolution microscopy techniques like AFM or HIM allowing in situ characterization is highly desirable.

Polyesters constitute the main segment of polymer materials prospective for printed electronics and packaging applications. They can be used as a flexible substrate to support electronic devices, as dielectric interlayers as well as in the top encapsulating coating. Poly(ethylene terephthalate) (PET) is a one of the most common thermoplastic polymers from the polyester family. It is widely used in synthetic fibers; packaging (beverage, food, and other liquid containers) engineering and thermoforming applications [92]. Poly(ethylene 2,6-naphthalate) (PEN) is similar to PET but has higher glass and melting transition temperatures, which makes it attractive for medical (hot steam sterilization, etc.), mechanical, and chemical applications (due to its mechanical stability and chemical inertness). In our study, ultrathin PET and PEN films were prepared from polymer solutions via spin- or dip-coating procedures, which suppressed the crystallization process during polymer film fabrication. Subsequent in situ annealing of the samples at various temperatures allowed monitoring of the diffusion-limited cold crystallization process by AFM [93].

The general scheme of formation of surface crystals in thin polymer films is presented in Figure 7A. Formation of surface crystals with a special dendritic morphology is observed for ultrathin films with the thickness comparable with the radius of gyration of polymer chains. Based on our AFM studies, a phase diagram of diffusion-limited crystallization behavior (in temperature-film thickness coordinates) was constructed [93]. Existence of three characteristic layers (as shown in Figure 7A) determines the features of crystallization behavior in thin films of crystallizable polymers [13, 93]. Figure 7 shows typical surface crystal morphology as observed by AFM in comparison with HIM imaging. AFM as height distribution-sensitive technique allows better contrast in comparison to HIM imaging. Figure 7B demonstrates a dendritic PET crystal grown on top of an amorphous PET “background” film. The total PET film thickness is quite low (40 nm). The main features of the crystalline structure can be recognized. Also, application of a stop-and-go approach (i.e., repetition of annealing and AFM imaging events) of direct in situ annealing will permit characterization of the kinetics of the crystallization process. While the vertical resolution of such imaging is in a subnanometer level, the lateral resolution is limited by AFM tip-surface feature shape convolution and depends on the tip sharpness. In the best case scenario, the details of the crystalline structure smaller than 10 nm cannot be resolved. HIM technique does not suffer this limitation making it perhaps better suited for this type of characterization. Figure 7C demonstrates that crystalline beams of the dendritic structures are actually constituted with smaller crystalline domains (5–20 nm lateral dimensions).

Figure 7

(A) Bulk vs. surface crystallization phenomena in a crystallizable polymer film under condition of geometrical confinement (film thickness is below 200 nm and comparable with the characteristic size of polymer chains –the radius of gyration); (B) AFM image of PET dendritic crystals grown on the surface of otherwise amorphous film (40 nm thickness); (C) the same structure visualized by HIM (200 nm film thickness). Arrows point several dendritic PET crystals. Scale bar 1 μm.

Application of micro-imprinting protocols allowed the fabrication of polymer films with 2D (lines) and 3D (dots) type of geometrical confinement. Figure 8 shows examples of AFM and HIM imaging of such structures. The HIM technique has a higher applicability to study of the nano-imprinted structures as it does not have the same problems of high roughness or high surface curvature that limits the application of the AFM technique. In Figure 8A,B, the details of the surface roughness are lost, and information about surface topography between imprinted features is unavailable. On the contrary, HIM images (Figure 8C–H) show the real structure with vertical side walls of the imprinted features (dots or parallel lines). One can see that annealing of the structures causes partial dewetting of the PET film in the case of the dotted pattern, but no surface crystallization process is observed. Therefore, AFM and HIM post investigation of the imprinted film morphology depending on the annealing temperature revealed a remarkable stability of the amorphous phase. Compared to 1D-confined films (Figure 7), where characteristic surface crystallization was observed at T<Tg, no sign of surface crystallization process was observed even at temperatures exceeding the bulk melting transition. Further studies of the surface crystallization in PET and other crystallizable polymers are in progress.

Figure 8

AFM (A, B) and HIM (C–H) images of PET imprinted films with 2D (lines) or 3D (dots) type of geometrical confinement. (C) and (D) portions are different magnification HIM images of an as-prepared film PET film; (E) and (F) show the morphology of the same film after its annealing at temperature 5° above the T_m of PET; (G) and (H) show different magnification HIM images of a PET film with imprinted parallel lines (no sign of surface crystallization can be seen before or after annealing). AFM images do not allow high-resolution characterization of the film surface (especially) between the nano-imprinted features due to high local curvature and surface roughness of the surface.

3.5 Polymer-based devices

3.5.1 Bulk-heterojunction solar cells

Organic photovoltaic (PV) devices are considered the third generation of PV technologies following first-generation, single-junction crystalline silicon cells and second-generation, thin-film solar cells based on CdSe, copper indium gallium diselenide (CIGS), and a-Si or nano-Si [94–96]. Recent advances in solution-processable organic and inorganic semiconductors, as well as their composites, have led to the possibility of using significantly lower cost manufacturing methods, such as printing and coating, and simplify the fabrication of PV devices. One of the best candidate for low-cost manufacturing PV devices is polymer-based bulk heterojunction (BHJ) solar cells. The setup of such devices is based on a mixture of electron donor and electron acceptor materials that are sandwiched between two selective electrodes. The two intermixed photoactive materials create an active layer of the cell. Its nanoscale, heterophase character enables a very efficient charge generation throughout the bulk [97]. The morphology of BHJ cells has to be controlled at the nanoscale to achieve efficient light energy harvesting via exciton diffusion and dissociation phenomena. Polymer BHJ solar cells showed a remarkable progress during the last 3 years with the efficiency exceeding 10% for the best devices [98–101].

Probably, the most studied active materials for this type of solar cells are conjugated polymers, such as P3HT (poly-3-hexylthiophene) or other polythiophene derivatives (especially donor-acceptor type) in combination with the fullerene-based acceptor material [91, 98–101]. Such polymer solar cell can be deposited via solution-based protocols (wet deposition) or vacuum evaporation techniques on top of a transparent substrate (glass or PET) coated with a thin layer of an optically transparent electrically conductive materials (indium tin oxide – ITO) [102, 103].

Nanostructured organization of the BHJ layer is critical and can promote exciton drift to the donor-acceptor interface in cases where the components of the active layer form interdigitated structures (with dramatically enhanced specific interfacial area). These structures can be created deliberately via control of the crystalline morphology of polymer photoconductor and the degree of phase separation in the system (a proper choice of the components, and special thermal annealing protocols) [104–106]. The same structures also can be created via addition of small quantities of special molecular compatibilizers (amphiphilic or bi-anchor thiols) [106]. Ordinary (random) BHJ structure is then created spontaneously within the active layer due to phase separation of the components under thermal annealing (120–150°C). The length of the exciton diffusion puts strict requirements on the size of the polymer phase domains, not to exceed 20 nm. Other morphological restrictions include desirable alignment of the polymer chains (to promote high mobility of charge carriers) and their orientation normal to the electrode surface (this allows higher active layer thickness – i.e., higher light energy absorption with simultaneously reasonable electrical resistance). Further enhancement of the PV efficiency could be achieved if so-called organized heterojunction structure is created [107]. Such a special morphology offers multiple exciton dissociation sites and separate charge pathways, thus limiting exciton recombination. As a result, thicker and therefore more absorbing polymer films can be used as an active layer.

AFM and electron microscopy techniques were applied successfully to study the morophology of the active layer components depending on their ratio and annealing procedures [108]. Particularly, the size and orientation of polythiophene (PT) nanocrystals is responsible for BHJ cell performance due to anisotropy of charge carrier mobility along and across the polymer chains [109]. Nanometer resolution imaging of organic BHJ is a challenging goal; however, such information may assist in the development of PV devices with improved efficiency. TEM as the highest-resolution electron microscopy technique was applied to determine the nanoscale structures present in thin BHJ films (<100 nm) [110, 111]; however, such images represent a projection of structures detected through the sample thickness [112]. Unambiguous interpretation of the bright-field TEM data was difficult due to small contrast variation within the polymer layer and possible artifacts arising from thickness variation and electron beam defocusing [111, 113, 114]. Moreover, most of the PV cell organic materials showed very low stability under the electron beam exposure. To reduce the beam damage, energy filtered transmission electron microscopy (EFTEM) has been used to image P3HT/PCBM thin films [113]. The authors applied computer-aided reconstruction based on the TEM images, which has proved an existence of a three-dimensional network of P3HT nanofibers within the active layer; however, the size and distribution of the polymer domains obtained from bright-field and energy-filtered images reconstruction techniques were not consistent.

Application of HIM technique in this case has an advantage of finely focused and less destructive ion beam enabling visualization of the nanoscale compositional variation in a PV device. The components of the BHJ system show different brightness in their secondary electron (SE) emission, which depends on the ratio of π-bonds to the sum of π- and σ-bonds in an organic molecule [115]. As a result, the P3HT component appears brighter than PCBM in HIM images (Figure 9). Monitoring of a gray level distribution of the HIM images taken across the sample allowed high-resolution mapping of the BHJ layer components. Particularly, it was determined that PCBM domains are slightly elongated in shape and have their dimensions below 20 nm (∼12 nm in length), which is consistent with the observed efficient exciton dissociation in the system. Therefore, HIM was proved as a powerful method enabling ex situ (or even in situ) characterization of organic semiconductor thin films and devices [116].

Figure 9

HIM imaging of BHJ solar cell active layer constituted with P3HT electron donor and PCBM electron acceptor. High-resolution imaging and high contrast are observed because of different secondary electron yield properties of the components [Reprinted with permission from (A.J. Pearson, S.A. Boden, D.M. Bagnall, D.G. Lidzey, and C. Rodenburg, Imaging the Bulk Nanoscale Morphology of Organic Solar Cell Blends Using Helium Ion Microscopy, Nano Lett, 2011, 11(10): p. 4275–4281). Copyright (2011) American Chemical Society].

3.5.2 Polymer-based sensors

Intrinsically conducting polymers (ICP) such as polyaniline, polythiophene, and some other polymers have a proven record of their application as chemical sensor materials. These polymers have a quite high sensitivity, are relatively simple to synthesize, lightweight, and can detect gases at ambient temperature [35, 43–45, 47, 48, 117–121]. PANI differs from other ICP due to its higher stability and much lower cost. Moreover, the electrical conductivity of this polymer can be reversibly controlled in two different ways: either by electrochemical or chemical oxidation/reduction processes or by protonation/deprotonation in a simple acid-base chemical interaction. Additionally, PANI can be combined with some thermoplastic polymers to form a composite with a very low percolation threshold of conductivity. The resulting composites may effectively combine mechanical properties of the thermoplastic polymer matrix and the electronic properties of PANI to efficiently serve as an active component of an organic gas sensor device [122, 123]. Nanostructured PANI-based materials fabricated as nanofibers, core-shell particles, or nanotubes are attractive due to their ability to form thin conductive films with a porous structure [124, 125]. The latter morphology enables easy penetration of the analyte molecules inside the sensor film dramatically increasing the detection limit and sensitivity while simultaneously reducing the response time of corresponding sensor devices [126, 127].

Several recent publications show potential of nanostructured PANI-based composites as sensing materials for ammonia gas with the detection limit in a ppm range [128, 129] Application of HIM microscopy technique allowed monitoring and optimization of the morphology and porous structure of PANI-PVDF core-shell-based films during their fabrication [55, 130]. Moreover, sensitivity of the HIM contrast formation mechanism to a charge state of the material under study allowed better understanding of the physical sensing mechanism during sensor material operation. Figure 10 shows the optimized nanoporous polymer film with high sensitivity to ammonia and an example of irreversible transformation of the PANI layer due to chemical interaction of a dopant with HCl analyte. The latter process leads to formation of nanoscale dots of dopant salt clearly visible in HIM because of their higher electrical conductivity in comparison to a semiconductor PANI film.

Figure 10

HIM imaging of polymer-based sensor devices; (A, B) active layer of PVDF-PANI ppb range ammonia sensor layer exposed to ammonia for a long time (phase separation and formation of dopant-ammonia salt is observed), (C) nanoporous sensor film based on PVDF-polythiophene core-shell nanoparticles after elevated temperature sintering of the particles. Polythiophene shows high sensitivity to volatile organic solvents such as acetone, chloroform, toluene, and some others.

Figure 10C shows another example of nanostructured semiconducting polymer polythiophene (PT) prepared through the same approach – i.e., first, core-shell nanoparticles with a PVDF core and PT shell were synthesized, and then, particulate films were cast from a chloroform dispersion and sintered into a solid nanoporous film under short-term thermal treatment [130]. Nanostructured PT materials possess a ppm detection limit and sensor discrimination capability in detection of volatile organic compounds (VOCs) such as acetone, chloroform, toluene, and others) [131–133].

3.6 Imaging of biopolymers

Biological polymers or biopolymers are the molecules of life; they compose the bulk of the dry mass of all living organisms, they assume the bulk of metabolic and structural duties in the cell, and in the case of the nucleic acid biopolymers deoxyribonucleic acid and ribonucleic acid, these molecules function as the blueprint for life itself. The interest of the structural aspects of biopolymers has risen over the past 10 years as properties such as self-assembly and self-organization has been demonstrated for many of the biological molecule and biopolymer systems [134]. Self-assembling collagen arrays, DNA origami, and now protein origami clearly demonstrate how elaborate and complex nanostructured surfaces can be generated de novo from a relatively simple collection of biopolymers and little or no scaffolding using a bottom up strategy [134–139]. The hope is that larger-scale self-assembly of even more elaborate nanopatterned surface will be possible using similar principles and approaches.

There are three major classes of biopolymer: polysaccharides, polypeptides, and nucleic acids. Polysaccharides are found throughout the biosphere, alone and also conjugated to other biopolymers such as proteins, in the form of glycoproteins, or to lipids in the form of glycolipids [140, 141]. In comparison to nucleic acids and proteins, carbohydrates have larger capacity of complexity through the presence of numerous chiral centers within the sugar/monosacchride subunits, the greater possible combinations of glycosidic linkages between different carbons within the sugar subunits resulting in branching, and to modifications to individual sugar subunits such as acetylation, methylation, oxidation, and sulfation [139, 142]. Given this extraordinary potential diversity, the extent of structural diversity of polysaccharides is just being explored. Nucleic acids and polypeptides have been the most extensively studied, probably for the very reason that they are the most experimentally tractable using current molecular biological techniques [139]. Polypetides are polar polymers composed of amino acid subunits. The folded, tertiary structure of a given polypeptide is determined by the primary sequence of its amino acid chain; this folding reaction is a thermodynamically driven self-assembly process. In some cases, the folded polypeptide is the functional state, and it is this state that physically interacts with other polypeptides and other components of the cell, both structural and metabolic. However, most proteins have some level of posttranslational modification prior to generating a fully functional protein complex. Natural occurring polypeptides contain 20 amino acids, although recent genetic engineering of bacteria has allowed the incorporation of synthetic amino acids into a polypeptide chain and, thus, generating potentially new and novel proteins [143]. There is a great deal of diversity in protein tertiary and quaternary structure, ranging from globular to filamentous, and many of the elaborate structures arise from the assembly of large protein-based complexes. Nucleic acids are simpler in their primary structure when compared to polysaccharides or polypeptides, them being composed of a smaller number of basic subunits called nucleotides, which contain a sugar group, a phosphate group, and a nitrogenous base group. Nucleic acids have the potential like polypeptides to fold into elaborate three-dimension structures that play catalytic and structural roles in the cell.

Regarding the structural aspects of biopolymers, there are two general levels of inquiry where the bulk of progress has been made one at the molecular/atomic level and the other at the cellular level. Techniques such as X-ray crystallography and NMR have been used extensively in the field of structural biochemistry to determine the orientation, organization, and three-dimensional structures of many biological molecules to atomic resolution, although lower resolution can be also be informative [144]. In most cases, solving of a crystal structure allows the placement of each atom of a single biological polymer, and in some cases, the organization of individual clusters of interacting biological polymers with respect to one another. However, challenges exist. X-ray crystallography requires that the protein or other biological molecule be sufficiently pure to grow the crystal to be used in the structural solution. This has been notoriously difficult for extremely large proteins or transmembrane proteins [144, 145]. Current NMR techniques require the protein or other biological molecule to be small and soluble [146]. The other level of inquiry regarding biological structure is at the cellular level. At this level, optical microscopy, especially the recent advances in confocal microscopy and super resolution microscopy have allowed unprecedented resolution of the overall organization and structure of the cell [147, 148]. While each of these approaches is important and will be important for the nanoscientist, the focus of nanobiology and nanobiotechnology falls somewhere between the cell biological view and the atomic/molecular structural view and provides a conceptual as well as technical link between these two realms.

The study of biopolymers, their structure and assembly and organization into higher-ordered structures, is a major thrust of nanoscience, and the imaging demands fall squarely between the ultrahigh-resolution techniques (such as X-ray crystallography and NMR) and the lower-resolution optical microscopy techniques; this is a zone, which the HIM is well suited. Regardless of the nature of the biopolymer, all are composed primarily of nonconductive carbon, which limits the resolution of any particle beam microscope and pose similar issues as with the imaging of other organic polymers described in the sections above; to complicate matters, many, if not most, of the biological molecules even when purified are far more complex in both structure and content. Because of these complications, imagining biological samples, i.e., molecules in situ, in the context of a cell or a tissue requires special preparations such as labeling samples with heavy metal stains to generate contrast in the material and also coating samples with nanoscale layers of gold or other heavy metals. Recent advances in the imaging of biological samples by environmental SEM or ESEM, which operates through the use of varying the pressure and/or reducing of the operating voltage of the electron beam have permitted the examination of native biological structures [149, 150]. While such techniques have enabled the imaging of uncoated biological samples, the application of ESEM typically results is a slight loss of resolution. Helium ion microscopy by virtue of the parameters described in the section of this review above, such as the lower penetration of the surface by the helium ion allows for higher resolution with a lower energy beam and the proportionally larger numbers of secondary electrons. This has led to spectacular images of biological samples such as butterfly scales, the surface cuticle of Drosophila, and even bone [31, 151, 152]. HIM microscopy demonstrates clear advantages by generating imaging with increased resolution, clarity, and increased depth of field; however, the idea that HIM imaging is simply a high-powered SEM is wrong. The role that HIM imaging in the biological sciences, especially for the visualization of the structure and organization of biological molecules is just being tapped. For instance, chitin is a crystalline polysaccharide polymer of N-acetyl glucosamine and is the primary component of the natural composite material of insect cuticles. The insect cuticle contains an elaborate and diverse set of structures (Figure 11A) and as a material displays an enormous range of physical properties, which is dependent on the organization of chitin by specific proteins in the cuticle [153]; HIM imaging of purified chitin reveals a complex network of nanoscale pores (Figure 11B).

Figure 11

(A) HIM images of the native cuticle of Drosophila melanogaster demonstrating the elaboration of chitin polymer-based structures and sensory organs. (B) Ultrahigh-resolution image of purified chitin. Note the presences of nanoscale holes and pores.

4.1 Milling/patterning

HIM combines the advantages gained from a small beam size and high brightness, along with the increased particle momentum and SE yield, to fabricate features with a high level of control. Ion milling is a standard fabrication method that involves the removal of the substrate with a focused ion beam. Samples milled with standard FIB using gallium (Ga) ion sources often have residual contamination introduced by Ga implantation during the milling process, which can lead to alteration of the surface chemistry. Moreover, the more massive Ga ions are harder to manipulate and control than He ions. They possess a greater momentum when compared to helium ions, which often results in unwanted sample damage especially when milling at the sub-30 nm scale [154, 155]. This side effect is not present in the HIM where sub-10 nm features can be milled with high fidelity. Recently, intricate and detailed devices have been milled with HIM lithography that match the theoretical resonance quality factor limit modeled in a coaxial plasmonic nanoresonator [156]. These structures would have been impossible with Ga FIB milling. Rudneva et al. have developed a heating stage allowing milling of nanoscale patterns in Si and Ti and demonstrated that unlike other methods, HIM milling did not affect the crystalline structure of the samples when heated to a high enough temperature, even at the edges of the mill [157].

The HIM can be used in a variety of biological applications as well. Not only is it capable of imaging uncoated biological samples at a very high resolution [31, 151, 152], but it has also been used to mill nanopores with diameters below 4 nm that are comparable to similar pores formed using TEM [158]. Further studies were performed, which related the milling depth to the ion dose for both direct and transmission milling and shed light into differences in milling parameters [159]. The milling process allows for multiple diameter sizes on one sample and only takes seconds to mill one pattern, making it an ideal method for wafer-scale production.

4.2 Lithography

All of the advantages arising from the gas field ion source with the helium beam have also made HIM a promising technique for ion lithography or the patterning of a resist material by exposure to the ion beam (Figures 12 and 13). Until recently, these types of patterns have been written using electron beam lithography (EBL), or Ga FIB; however, these techniques are limited in the resolution of features by the spot size and the interaction volume of the beam itself. Micrometer range scattering due to the small particle size of the electron beam gives a large interaction volume at the surface, while the gallium beam is good up to a certain point, but loses resolution below 30 nm because of its relatively large beam size [156].

Figure 12

Nanolithography with HIM. 100 nm×100 nm square hole milled on top of 200 nm diameter PVDF particle.

Figure 13

Ion lithography using the HIM. (A) Generation of nanoscale feature using He ion implantation. The surface has a 100-nm layer of gold on a Si wafer; He ion deposition occurred with 1 μs dwell time and 10 repeats of the pattern. Inset is the pattern of deposition. (B) Ion milling using HIM, a 10-nm layer of AuPd sputter coated on a Si wafer. Pattern same as in (A) with different dwell times, starting from top left and going in a z pattern: 5 μs, 10 μs, 15 μs, and 20 μs dwell times with one repeat and milling occurred.

Initially, helium ion lithography (HIL) in hydrogen silsesquioxane was shown to be capable of achieving 20 nm pitch dense structures with little resist residue, but had considerable resist residue with 10-nm pitch structures [160]. Although there were difficulties in obtaining sub-10 nm structures, the required ion dosage was shown to be approximately an order of magnitude less than that needed with EBL. The point spread function was also measured and had a reduction in the amount of micrometer range scattering similar to what is observed with electrons. Given these advantages, in its current state, HIL has feature resolution comparable to EBL, and recently, 4-nm half pitch dense patterns have subsequently been achieved using HIL [161]. These structures were subsequently used for imprint lithography on a UV curable resist, demonstrating that high-throughput single-digit nanometer fabrication of dense patterns is possible using HIL.

HIL will continue to develop and will be a major player in single-digit nanometer fabrication and lithography. The development of other ion beam milling stations using other noble gases such as neon is now being explored as a lithography tool. The neon beam allows for an increased sensitivity to resist exposure. Neon ions have been shown to be ∼10× more sensitive than He and ∼1000× more sensitive than electrons and have resolutions equal to state-of-the-art EBL [162]. It can be sure that this technology will be refined as it continues to mature and will surely have a great effect on many research areas relying on scanning particle beam lithography. While, much of HIM lithography has been concentrated on milling and processing nonpolymeric materials such as SiO₂, as with other lithography methods, it is simply a matter of time until the precision of HIM lithography will be applied to polymers.