1. Introduction

Diatoms, a form of photosynthetic algae, exist in almost every terrestrial freshwater or saltwater (or moist) environment. Indeed, their efficient photosynthetic machinery and prevalence are such that diatoms are responsible for a significant portion of the oxygen and biomass generation on earth [1]. Diatoms also play a key role in water quality management and environmental protection [2]. More recently, diatoms have attracted appreciable attention as a source of renewable energy, since they utilize sunlight to generate organic materials that can be used to produce biofuel [3]. Furthermore, the cell walls of diatoms, known as frustules, are comprised of silica that is formed in complex three-dimensional (3-D) hierarchical (micro-to-nanoscale) morphologies [1]. Each of the tens of thousands of diatom species forms a frustule with a unique 3-D morphology decorated with fine (down to nanoscale) features (e.g., pores, protuberances, channels) that are preserved with high fidelity upon reproduction [1]. The combination of 3-D complexity, variety, reproducibility, and scalability (through sustained biological reproduction) of frustules formed by diatoms under ambient conditions exceeds state-of-the-art micro/nanoscale fabrication technologies [4, 5].

A new diatom research direction, diatom nanotechnology, has begun to emerge in recent years, for which major aims include understanding how diatoms build their frustules (to allow for the development of diatom-inspired 3-D self-assembly strategies) and learning how to utilize such frustules, or chemically-modified versions of such frustules, for applications such as biophotonics, gas sensing, catalysis, energy harvesting, microfluidics, etc [4–15]. Despite such growing technological interest, and the historical fascination of researchers with the intricacy and beauty of diatom frustules (Fig. 1(a)), the varied functions of diatom frustules are only partially understood. The optical behavior of these natural objects has started to attract some attention. For example, the valves of Coscinodiscus wailesii frustules have been shown to focus light like an optical lens [16] and to exhibit wavelength-dependent transmission [17, 18]. Melosira variance frustules have been reported to absorb mainly blue light [19]. The pore structure of the girdle bands of Coscinodiscus granii frustules have been hypothesized to yield photonic crystal modes for the guiding of blue-green light [20].

 

Fig. 1 Optical images of: a) an arrangement of different diatom silica frustules (created by K. D. Kemp and available at http://www.diatoms.co.uk/pg.htm), and b) living C. wailesii diatom cells. The scale bar represents 100 µm.

Download Full Size | PDF

Considering the attention given to the light-guiding properties of synthetic silica-based photonic crystal fibers for telecommunications [21–25], we have been inspired to directly interrogate, for the first time, the position-dependent optical transmission of silica diatom frustules possessing quasi-periodic pore arrays, given their photonic crystal-like structures. In this communication, we report the results of our recent experimental investigation of the diffraction and transmission of coherent supercontinuum light from diatom frustules. We show that the valve of the diatom Coscinodiscus wailesii exhibits well-defined resonant bands that are consistent with a photonic crystal, albeit without a complete band gap. Furthermore, the associated optical filtering behavior was sensitive to changes in the periodic pore nanostructure at various locations on the C. wailesii valve.

2. Sample preparation

2.1 Diatom culturing, harvesting, and frustule cleaning

The Coscinodiscus wailesii diatom strain was acquired from CCMP (https://ncma.bigelow.org) and cultured in our laboratory using the L1 culture medium (also purchased from CCMP) in 100 ml glass flasks under a 20 W fluorescence lamp with 12/12 h of light/dark cycles. Culturing was conducted at 18°C with the L1 culture medium renewed every 6 weeks. Figure 1(b) reveals an optical image of living C. wailesii diatoms, with the girdle view shown for one cell (appearing as a cylinder) and the valve view (appearing as a circular disk) shown for another cell. In the valve view orientation, the chloroplasts within the diatom cells are seen as bright circular features. The C. wailesii valves possess a relatively large diameter (≥ 130 µm, which is among the largest of valves for diatom frustules [1]).

Diatoms were harvested from the culturing medium using a fine net (10 µm openings), and the residual medium solution was removed by rinsing in distilled water. The collected diatom sample was then immersed in an aqueous 50% hydrogen peroxide solution at 85°C for 4-6 h to remove the organic material. After rinsing with distilled water, the sample was rinsed in ethanol and then stored in ethanol.

2.2 Optical characterization

To allow for manipulation and optical interrogation, individual C. wailesii valves were attached to the end of a supporting optical fiber. Cleaned diatom frustules in ethanol were spread on a microscope coverslip and the ethanol was allowed to evaporate. Individual frustule valves were then glued to the end of an optical fiber (Corning SMF28) using precision translation stages under an optical microscope. A UV curable epoxy was first applied to the fiber end, which was then brought into contact with the valve of interest. After contact was achieved, a UV lamp was used to cure the epoxy. Secondary electron images of such a C. wailesii valve attached to the end of the supporting optical fiber are shown in Fig. 2.

 

Fig. 2 a) Secondary electron (SE) image of a single C. wailesii valve attached to a supporting optical fiber (scale bar represents 100 µm). b) A higher magnification SE image of the valve in a), revealing a nanostructured, quasi-periodic pore pattern (scale bar represents 1 µm).

Download Full Size | PDF

To evaluate the optical properties of individual diatom valves, the experimental setup shown in Fig. 3 was utilized. Light was provided with a fiber-based supercontinuum source from NKT (SuperK COMPACT). This source consisted of a white light broadband spectrum extending from 400 nm to beyond 1700 nm. The light output from the fiber tip was passed through the individual diatom valve (which was attached to a different, supporting fiber as described above). The separation distance between the diatom valve and the output fiber of the supercontinuum was kept to 50-100 µm, so as to achieve a light spot size of only ~20 µm on the valve. Such a fine imposed light spot allowed us to probe different regions of the individual valve to evaluate the influence of the variation of the pore structure, which occurred on length scales larger than the spot size. The light transmitted through the single valve was collected by another optical fiber and then sent to an optical spectrum analyzer (Ando AQ-6315A) for analyses. All three optical fibers used in the experiment (the support fiber, the source fiber, and the collection fiber) were attached to 3-D translation stages for fine adjustment.

 

Fig. 3 a) Experimental setup for measurement of the tunable filter behavior of individual frustule valves. b) Higher magnification optical image of the central region where the diatom valve is located. Light from a coherent supercontinuum source emerged from the optical fiber (SMF28) on the left to illuminate the diatom valve which was attached to a supporting optical fiber (vertical fiber in the image on the right). The valve-supporting fiber was also attached to a xyz stage for precision position adjustment relative to the light source. The distance from the diatom valve to the end of the output fiber was ~50 µm. The transmitted light through the diatom valve was collected by the optical fiber on the right (right horizontal fiber in the image on the right) and sent to an optical spectrum analyzer for analysis. The diameter of all the fibers used in the experiment was ~125 µm.

Download Full Size | PDF

3. Results and discussion

The analyses obtained from the OSA are shown in Fig. 4. The black curve is the spectrum associated with the supercontinuum source itself (i.e., without passage through the diatom frustule valve). Reductions (dips) in the transmitted power were observed for light passed through the C. wailesii diatom valve. Such reductions in transmitted light intensity were consistent with diffraction of those wavelengths away from the center spot. Interestingly, the dips in light intensity measured at different locations along the valve were found to occur at different wavelengths in a manner consistent with strong diffraction at the corresponding wavelengths; that is, dips at shorter wavelengths were observed at the outer rim of the valve whereas longer wavelength resonances were found near the center of the valve. Such wavelength variations in transmitted light intensity followed changes in the spacing of the pores on the valve; that is, the pores located near the center of the diatom valve tended to be spaced closer together than pores located near the outer rim of the valve, as revealed in the secondary electron images in Fig. 5.

 

Fig. 4 Optical spectra of the supercontinuum light transmitted through different areas of the silica C. wailesii valve. The laser spot diameter at the valve surface was ~20 µm. Dips observed in the power of light transmitted through the valve at different valve locations were consistent with position dependent diffraction associated with different pore spacings. The spectra corresponds to regions near the center (red) approximately between the center and the outer valve edge (green) and near the outer valve edge (blue). The inset reveals an optical photograph of the C. wailesii valve attached to the end of the supporting optical fiber.

Download Full Size | PDF

 

Fig. 5 SE images of the valve of a C. wailesii frustule (scale bars represent 10 μm). The average pore spacing in the top right image (obtained from a location near the center of the diatom) was 4.26 ± 0.06 μm, while that in the bottom right image (obtained from a location closer to the edge of the valve) was 4.66 ± 0.04 μm.

Download Full Size | PDF

To observe the diffraction patterns obtained from the C. wailesii valve, we replaced the collection fiber with a flat piece of white paper. Figure 6 reveals a few photographs of typical patterns that were generated as the result of light interaction with the periodic pore structure of the same valve. Figure 6(a) reveals the case where the valve was not in the light path, which yielded just a white circular spot (as expected for the supercontinuum source). When the C. wailesii valve was placed in the light path, we observed a colorful pattern associated with diffraction of different applied wavelengths at different angles (note: the central portion of the pattern from the supercontinuum source remained after the interaction). Upon moving the valve relative to the laser beam, the color of the central laser spot changed, which was consistent with a change in the local diffraction maximum. The diffracted light also exhibited a hexagonal pattern, which was consistent with the quasi-hexagonal pore structure of the valve (as shown in the SE images in Figs. 2 and 5).

 

Fig. 6 Optical images of the diffraction patterns obtained from the different regions of a single siliceous C. wailesii diatom frustule valve. a) No diatom was present in the beam path, so the laser beam is shown as just a circular white spot. b-d) The frustule valve was in the beam path. The color of the central spot changed according to the location of the partial photonic band gap. The hexagonal scattering patterns resulted from the hexagonal periodic pore structure of the diatom valve.

Download Full Size | PDF

The diatom diffraction was modeled using standard diffraction theory [26]. Figure 7 shows the simulation results revealing the six-fold symmetry consistent with the experimental observations in Fig. 6. For this simulation, the refractive index of the diatom silica was taken to be approximately 1.45, the surrounding medium was air, and the light was assumed to be normally incident on a flat periodic structure with the hole spacings determined above and indicated in Fig. 5.

 

Fig. 7 Simulated diffraction patterns for normal light incidence and for two different center-to-center hole spacings (4.25 μm on the left and 4.65 μm on the right), in correspondence with Fig. 5. These simulations were consistent with experimental observations, shown in Fig. 6, of changes in optical diffraction patterns upon translation of the frustule valve with respect to the incident laser spot.

Download Full Size | PDF

With the index contrast of 1.45 between silica and air, it was not expected that the diatom frustules would exhibit complete 2-D photonic band gaps in the region of measurement, as an index contrast of about 2 is generally needed for clear 2-D photonic band gaps [25]. Note that such 2-D band gaps could be accessed with a tightly focused optical beam that includes a range of incidence angles, or by increasing the angle of incidence between the light beam and the diatom frustule by rotating the frustule. Figure 8 shows the 2-D TE and TM dispersion relations obtained for a planar, hexagonal array of pores in silica using dimensions similar to those for the C. wailesii valve inner plate (i.e., a hole diameter of 2.6 μm, a center-to-center hole spacing of 4.5 μm, and a silica thickness of 0.2 μm). Note the high density of bands as the actual measurement range is approached, indicating the lack of band gaps in this system at this low index contrast.

 

Fig. 8 Simulated dispersion relation for TE (red) and TM (blue) modes for a planar, hexagonal array of pores in silica, with a spacing of 4.5 μm and a radius of 1.3 μm similar to those for the C. wailesii valve inner plate.

Download Full Size | PDF

We have demonstrated that the silica-based valve of a representative centric diatom species (Coscinodiscus wailesii) exhibits distinct tunable filtering effects produced by position-dependent diffraction. Furthermore, we have found that the wavelengths at which such effects occur are correlated to the diatom valve pore structure; that is, the wavelength of the band-gap is directly related to the periodicity of the pore pattern on the valve. Ultimately, genetic modifications may also be used to tailor the periodic pore structure of diatom frustules to fit the needs of various applications (i.e., genetically engineered micro/nanodevices) [7]. Given the wide variety of quasi-periodic pore patterns available among existing centric diatom species, along with the reproducibility (due to genetic control), scalability (due to massively-parallel biological reproduction), and potentially structural tailorability of diatom frustules [1, 4, 5], we believe that diatoms provide attractive biological platforms for studying light/matter interactions and potentially for creating a variety of photonic devices.