Susanne
Wintzheimer
a,
Tim
Granath
a,
Antonia
Eppinger
a,
Manuel Rodrigues
Goncalves
b and
Karl
Mandel
*ac
aChair of Chemical Technology of Materials Synthesis, Julius-Maximilians-University Würzburg, Röntgenring 11, D97070 Würzburg, Germany
bInstitute of Experimental Physics, Ulm University, Albert-Einstein-Allee 11, D-89069 Ulm, Germany
cFraunhofer Institute for Silicate Research, ISC, Neunerplatz 2, D97082 Würzburg, Germany. E-mail: karl-sebastian.mandel@isc.fraunhofer.de
First published on 31st January 2019
Superparamagnetic iron oxide nanoparticles can be assembled to form anisotropic microrod supraparticles with the assistance of a magnetic field during synthesis. Optionally, these iron oxide microrods can furthermore be coated with a thin silica shell. Due to their anisotropic structure, both microrod types can be aligned in a magnetic field while being dispersed in a matrix material which can be cured during the alignment of the microrods. In this way, an anisotropic magnetic composite is obtained. Interestingly, it was observed that the optical extinction properties for visible light in such a composite are direction dependent, which can be explained by using appropriate models based on Maxwell equations. Based on the understanding of this principle, a clever approach for a hidden code could be proposed which is obtained from mixing pure iron oxide and silica coated microrod supraparticles in such an anisotropic composite. The hidden code, which comes down to obtaining a single value eventually, can only be revealed when knowing that the system needs to be measured with a certain “twist”.
Herein, we present a system which is made in a very simple way and which breaks with the classical distinction between the two approaches “graphical” and “optical” codes and mixes both principles. For this purpose, a mixture of two nanostructured objects is spatially distributed in a matrix (=“graphical code” principle). Then, the integral signal needs to be read out (=“optical code” principle), however, very importantly, from two directions (=“digital graphical code” principle). This ultimately results in one single value obtained with the detector (=neither graphical nor optical but simply a single number result). This single value eventually depends on the ratio of the initially spatially distributed two objects and therefore can be adjusted in a way that a code can be maintained by selecting a combination of the two objects.
To obtain silica-coated superparamagnetic microrods (type B), 40 mL ethanol and 1 mL ammonia were added to 10 mL of the as-prepared type A microrod dispersion. Then, 100 μL TEOS was added dropwise into the reaction mixture. This recipe is a modified version of the well-known Stöber synthesis to create colloidal silica. However, due to the selected very low concentrations of TEOS, the condensation of the hydrolysed ethxoy-silane is of the heterogenous type, i.e., no silica nanoparticles are formed homogenously in solution but rather a heterogenous Si-Ox nucleation takes place on the microrod surface, leading to the growth of a thin silica shell on the surface of the iron oxide. After 8 h agitation, the product was magnetically separated, decanted and washed with H2O and finally redispersed in 10 mL H2O.
In order to incorporate the synthesised microrods into a matrix, the obtained supraparticle dispersion (type A or B) was diluted to a volume of 200 μL with a pH-adjusted HNO3-solution (pH = 3.6) and 3.5 g of a hot agar solution (25 mg agar dissolved in 5 mL H2O) was added and intermixed within a cuvette. The cuvette was closed and cured at room temperature standing next to a magnet at a distance of 13 cm. To investigate the anisotropic optical properties, UV/Vis measurements were performed, having the microrods orientated once parallel and once perpendicular to the beam path, simply by rotating the cuvette by 90°.
When studying the extinction spectra of the composites, containing microrods of either only type A or only type B, it was found that they differ, depending on whether the microrods are parallel or perpendicularly aligned in the composite with respect to the light propagation (see Fig. 3a for the measurement setup). As the microrods are anisotropic, an anisotropic behaviour of the composite was expected. However, surprisingly, the extinction was not shifted equally over all wavelengths but rather non-linearly. Thus, from superimposing the recorded extinction spectra measured in parallel and perpendicular directions relative to the microrod alignment, an intersection in the spectral curves is obtained. Interestingly, the position of this intersection is at considerably lower wavelengths for type A microrods (Fig. 3b), namely at 413 nm, compared to type B microrods (Fig. 3c) which is at 550 nm. It is worth noting that the extinction measurements do not necessarily need to be exactly parallel or perpendicular with respect to the microrod direction; in fact, the intersection is invariant (i.e., always is at the same wavelength) for any two superimposed extinction spectral curves obtained from having measured the composite from any two angles. Only the width of splitting of the intensity of the curves below and above the wavelength of intersection varies as a function of the angle between the two directions of measurement (Fig. S2 in the ESI†). It should also be noted that the signal can be obtained in a well reproducible way (at least 5 measurements were carried out from which a variance below 10 nm in wavelength was obtained): it is insensitive to slight changes during preparation and easy to measure, but with the special feature of an unexpected signal behaviour.
In order to understand this complex behaviour of light interaction with the microrods, theoretical modelling was performed based on full electromagnetic calculations (Maxwell equations) using the computer program COMSOL Multiphysics. For theoretical modelling, the following parameters were chosen: The microrod is represented by a homogeneous spheroid with a diameter of 100 nm and a length of 2 μm. For simplification the surrounding medium is assumed to be water (refractive index n = 1.33) instead of agar-agar. The calculations used a spherical computational domain of diameter D = 4 μm. A perfectly matching layer (PML) of 300 nm thickness was used to attenuate the scattered light from the spheroids at the external boundary. The outer boundary was defined as a scattering boundary condition, adequate for solving the scattering problems. The excitation used is a plane wave of electric field amplitude 1 V m−1. Graphical representations of the computational domain used in all calculations are shown in Fig. 4.
The spheroid has its longest axis parallel to the x-axis, or along the z-axis. The direction of the excitation plane wave (k-vector) is along the negative z-axis. The polarization direction was chosen, either along the x-axis (parallel to the spheroid), or along the y-axis (perpendicular to the spheroid), or at 45° to the x-axis. The materials used in the simulations were Fe2O3 or Fe3O4 to take the different oxidation states of Fe(III) and Fe(II) of iron oxide into account, potentially present in the synthesised microrods. The optical constants (n – real value and k – imaginary value) of these materials, required for solving the electromagnetic problem, were taken from the publication of M. R. Query,17 which were made available online at https://refractiveindex.info/. As Fe2O3 is a birefringent material, simulations were carried out with n and k both for ordinary and extraordinary rays.
COMSOL permits to obtain the near-fields inside, at the surface and surrounding the spheroids, and the far-fields at the internal boundary of the PML layer in order to calculate the total cross sections (scattering, absorption and extinction). The distributions of the electric near-field at the surface of the spheroids (exemplarily for the case of Fe3O4) for an excitation wavelength of 600 nm are presented in Fig. 5a and b. In Fig. 5c and d the power outflow (time averaged Poynting vector) at the inner surface of the PML spherical surface is shown (exemplarily for the case of Fe3O4). For this calculation, COMSOL takes a spherical imaginary surface at a very long distance compared to the wavelength and size of the scattering spheroid. The blue colour represents far-field light leaving the boundary, whereas the yellow colour corresponds to light entering the spherical boundary (Fig. 5c), in agreement with the definition of the plane wave propagation direction. The surface patterns obtained for incident light polarized parallel or perpendicular to the spheroid show considerable differences in the distribution of fringes. This is due to the scattering properties of the elongated spheroid interacting with light at two different polarizations. These patterns are presented for qualitative comparison only. The quantitative experimental characterization of the radiation patterns for a well-defined polarization direction of the excitation would require an angle resolved spectroscopy technique.
As the simulations were done with polarized light for the illumination of the spheroid, the extinction cross sections are also dependent on the polarization direction. In order to achieve a better comparison with experimental results, where non-polarized light was used, the cross sections were averaged over 4 different polarization angles (−45°, 0°, 45°, and 90°, see Fig. S3 in the ESI†). The obtained simulated extinction cross sections for Fe2O3 (ordinary and extraordinary rays) and Fe3O4 in the case of parallel and perpendicular light propagation with respect to the orientation of the spheroids are depicted in Fig. 6. All simulated cross sections show an extinction of light by the microrods in the complete studied wavelength range (300 to 800 nm) while the extinction intensity depends on the direction and wavelength of the propagating light as well as on the spheroid material. The simulated extinction accounts for the absorption and scattering of the studied materials (which is exemplarily shown for a Fe3O4 spheroid in Fig. S4 in the ESI†). The acquired graphs underline the anisotropic extinction behaviour of the studied spheroid with a higher extinction at shorter wavelengths (∼300 to 450 nm) of light propagating perpendicular in comparison to light parallel. This behaviour switches at longer wavelengths (∼450 to 600 nm), resulting in an intersection of both curves (at ∼350 or 450 nm depending on the material), which correlates with the obtained experimental outcomes (Fig. 3). However, a second intersection at longer wavelength ranges (at ∼550 or 600 nm depending on the material) of the simulated extinction cross sections does not reflect the experimental findings.
This discrepancy may on the one hand be explained by the shift of the curves one notices when comparing the theoretical with the experimental extinction curves: the experimentally assessed intersection is found at longer wavelengths indicating that a second intersection may also exist but has not been measured as it is out of the range of the measurement system at over 800 nm. On the other hand, in the simulated examples only one geometry was calculated. Changes in the extinction cross-sectional spectra for smaller or larger spheroids or for another refractive index of the surrounding medium are possible. The simulations do not consider the polydispersity of the particle size either. Nevertheless, the theoretical modelling is supporting the experimental results as it clearly shows an anisotropic extinction behaviour of the microrods due to their shape, resulting in at least one intersection of the extinction as a function of the wavelength of light propagating parallel and perpendicular to the microrods. Additionally, it underlines the material dependence of the extinction spectra and therefore of the position of the intersection.
Summing up so far, it has been found and – by means of theoretical modelling – verified (in terms of validity) that magnetically oriented microrods yield direction dependent extinction behaviour. This anisotropic behaviour can be exploited to determine a very well-defined point of intersection when measuring the spectral transmittance from two arbitrary but different directions. Moreover, the position of the intersection point is determined by the optical material properties of the microrods, which can be used to create unique codes. For this purpose, quantitative mixtures of microrods of type A and type B were prepared (namely: A:
B (v%/v%) = 100
:
0, 50
:
50, 40
:
60, 30
:
70, 20
:
80, 10
:
90 and 0
:
100), aligned in a magnetic field and fixed in a matrix to obtain a solid composite. From this “graphical” code (=spacial arrangement of objects) an integral signal (transmitted light) was obtained and spectrally resolved (=“optical code”). However, only by performing this procedure at two different angles (=“binary graphically”; security feature that demands “asking the right question”), one unique intersection at one specific wavelength (=“numerical code”) could be obtained. The wavelength at which this intersection could be observed was found to correlate very well with the quantitative ratio of the mixtures of type A and type B microrods (Fig. 7). Thus, this ultimately allows predefining a unique code number by adjusting a certain mixture of the two microrod types, which can only be detected if the “reader” is aware of “the little twist”.
Footnote |
† Electronic supplementary information (ESI) available: Magnetisation measurements, extinction curves depending on the microrod orientation and simulated extinction cross sections depending on the polarization. See DOI: 10.1039/c8na00334c |
This journal is © The Royal Society of Chemistry 2019 |