1. Introduction

The unique configuration of guided wave propagation in the close vicinity of a microfludic capillary structure, which exists in the geometry of a microstructured optical fibre (MOF), constitutes a versatile platform for the development of novel and functional photonic devices [1,2]. In MOFs, the guiding mode that is confined in the microstructured solid or hollow core partially interacts with the medium infiltrated capillary, while it is subjected to refractive index, scattering or optical absorption changes, affecting its phase and power carrying properties. There are several examples where the micrometric diameter capillaries of MOFs are infiltrated using gas [3] or liquid [4] matrices for developing specific sensing [5] or actuating/switching [6] fibre devices. The inscription of Bragg or long-period gratings into microstructured optical fibres has been proven to be a quite efficient interrogation scheme [7,8]. Infiltrated Bragg and long period microstructured optical fibre gratings have been used as actuating devices by employing external heating elements [7] [9] or voltage trimming [10]; while in other cases have been used as ultra-sensitive refractometers [11,12].

Herein, we spectrally tune a Bragg grating inscribed in a microstructured optical fibre by utilizing an infiltrated ferrofluidic medium. A ferrofluid is a stable colloidal suspension of sub-domain magnetic micro-/nano-particles dispersed inside a liquid carrier [13]. In the presence of a magnetic field the magnetic moments of the particles orient along the field lines and magnetization of the fluid occurs. Such magnetisation results in volume and viscosity changes, which in turn and under specific conditions can be exploited in controllable liquid spatial translation. Ferrofluids have attracted significant attention due to their rheological and optical and thermal properties in applications such as microfluidic micropumps [14], in-line optical power modulators [15], in adaptable lithographic techniques [16] and in-fiber gain control [17]. Recently, we have used ferrofluids as long-period gratings outcladdings for tuning their transmission spectrum in an actuation mode, wherein the ferrofluidic medium is translated across the long-period grating length [18]. In this manuscript, we experimentally demonstrate that by using a static magnetic field and suitable surface functionalisation processing of the microstructured fibres capillaries, we can efficiently translate a dense ferrofluidic medium along the fibre length; and accordingly tune the spectrum of a Bragg grating reflector inscribed in the fibre core. The spectral data obtained are studied in transmission and reflection mode, for indentifying spectral shift and extinction ratio changes in the spectrum of the Bragg grating interrogator. This MOF device is the first demonstration step towards the development of magnetically tuneable opto-fluidic devices [19] utilising dispersed magnetic nanoparticles solutions.

2. Experimental

The microstructured fibre that was used, had 5 holes of 20.8μm diameter, forming an outer core of 16.1μm, which includes a 3.5%wt Ge doped socket of diameter 8.5μm (see Fig. 1a ). In that fibre a 4mm long Bragg grating was inscribed using a 1067.73nm phase mask and a 193nm excimer laser [20]. The grating reflected two major modes, located at 1545.57nm (0^th order) and 1541.26nm (1^st order), with strengths 12dB and 2.6dB, respectively. Modal simulations depicted that the 0^th order mode is confined into the Ge-doped socket, while the 1^st order mode is defined by the surrounding capillary structure, extending at a greater area.

 

Fig. 1 (a) Scanning electron microscope photo of the MOF used. The fibre cladding diameter is 125μm. (b) Transmission spectrum of the MOF grating after PVP functionalisation.

Download Full Size | PDF

The commercial hydrocarbon based ferrofluid EMG905, manufactured by Ferrotec and having 7.1% of volume concentration of magnetite (Fe₃O₄) was used. The EMG905 ferrofluid exhibits a 40mT saturable magnetization, 9mPas viscosity and the size of the Fe₃O₄ dispersed particles is 10nm, while its carrier liquid is a synthetic isoparaffinic solvent. The EMG905 ferrofluid is practically opaque since its absorption loss at 1550nm was measured to be 6.84μm⁻¹, while its refractive index was measured to be 1.58±0.01.

Initially, the fibre capillaries were functionalized using Polyvinylpyrrolidone (PVP) diluted in water [21]. The PVP was infiltrated using a vacuum pump and then expelled out from the capillaries by employing compressed nitrogen, for leaving a thin overlayer onto the capillary walls. The functionalized fibre was baked for 20h at 70°C in a vacuum oven to let the PVP layer dry into the capillaries. The grating spectrum after this step functionlisation is presented in Fig. 1b. The PVP infiltration was used for enhancing the hydrophilic properties of the surface of the fibre capillaries, and finally achieving a double effect: improving the mobility of the ferrofluid inside them and leaving a reduced residue onto the capillaries. Then, the EMG905 ferrofluid was infiltrated into the fibre by exploiting capillarity, while it was transferred along the length of the fibre and towards the grating area using a 200mT static magnetic field. The length of the ferrofluidic medium infiltrated was 6mm±0.5mm approximately, longer than that of the grating for ensuring full overlap between them. The uncertainty related to the ferrofluidic length is due to the differential infiltration of the five capillary channels constituting the fibre microstructure. The infiltration progresses at a different rate for each channel, resulting in a central “dark” ferrofluidic length where all channels are full; and a short (<0.5mm) “grey” length where individual channels are empty.

A stepper motor was employed to precisely move at a constant speed the magnetic fluid along the fibre capillaries. The infiltrated grating was spectrally characterized simultaneously in transmission and reflection mode, using a broadband source and two optical spectrum analyzers. The optical signal was injected into the microstructured optical fibre for reaching the grating first and then the ferrofluidic infiltrated section. The above state was considered as State A (see Fig. 2a ), from which the ferrofluid was translated towards the grating vicinity. Accordingly, the grating is half-length covered with ferrofluid when the distance Ox is equal to a nominal value of 2mm (see Fig. 2a). Finally, when the ferrofluid is placed between the grating and the light injection side, we consider it as State B.

 

Fig. 2 (a) Schematic of the ferrofluid infiltrated MOF-Bragg grating. For the specific position of the ferrofluid the MOF Bragg grating actuator is at State A. (b) Overview picture of a ferrofluid infiltrated MOF with the actuating magnet on the bottom. Minimum scale=1/32 inch. (c) Close view of the MOF, for depicting the “grey” length of the in-fibre infiltrated ferrofluid.

Download Full Size | PDF

3. Results and discussion

Transmission spectral measurements were obtained from the infiltrated MOF Bragg reflector for both of the guiding modes scattered (see Fig. 3 and 4 ). In Fig. 3 a contour plot of the transmission spectral data is presented; while in Fig. 4 the summary of the spectral shifts and grating notch strength changes are appended for forward and backward translation operation.

 

Fig. 3 Transmission spectra contour graph of the ferrofluid infiltrated MOF Bragg reflector, for different values of the coordination Ox defining the overlap of the ferrofluid with the grating.

Download Full Size | PDF

 

Fig. 4 Transmission data for Bragg wavelength shift and grating strength (extinction ratio) changes, for the 0th and the 1^st guiding modes scattered by the infiltrated, MOF- reflector. The shadowed area resembles the 4mm grating length, assisting visualization. Black and red color traces refer to forward and backward translation of the ferrofluid for investigating repeatability.

Download Full Size | PDF

The data of Fig. 3 show that once the ferrofluid starts to overlap with the grating length the spectral response of the periodic structure undergoes two fundamental changes: first, it shifts to slightly higher wavelengths and second, the extinction ratio of the grating reduces. The last effect is associated with two optical properties of the ferrofluid: its extreme optical absorption and its refractive index that is higher than silica, inducing leaky propagation when infiltrated into the fibre capillaries. In such interaction scheme, the grating length overlapping with the magnetic fluid exhibits reduced diffraction efficiency, depending upon the overlap of each guiding mode with the infiltrated capillary region.

This effect is more explicitly depicted from the data presented in Fig. 4, wherein the Bragg scattering corresponding to the 1^st guiding mode, exhibits greater spectral shifts –compared to the 0^th mode-, while almost vanishing in extinction ratio when is fully overlapping with the ferrofluid. The maximum wavelength shift for the 0^th order mode is ≈0.1nm (Fig. 4a) and the corresponding extinction ratio in transmission is greater than 2.5dB (Fig. 4b); while the 1^st mode is shifted more than 0.18nm (Fig. 4c) and its 2.5dB strength is completely erased (see Fig. 4d). Moreover, in the data of Fig. 4 the ferrofluidic actuator is tested for backward and forward operation, for investigating the effect of ferrofluid residue onto the capillaries walls and its effect on the spectra obtained. For the experimental conditions applied (namely, surface functionalisation conditions and translation speed of the ferrofluid, namely, ≈0.5mm/sec) the spectral hysterisys effect appears to be minimal. Spectral hysterisys effects for backward or forward operation of the MOF may be associated with imperfect surface functionalisation or with spatially confined clotting of the ferrofluid. The above experiments were repeated using the same MOF actuator several times, while leading to reproducible results. Nonetheless, we found that the lifetime of such device is dependent upon the stability of the infiltrated ferrofluid, which is usually of the order of few days.

The effect of the translation of the ferrofluid along the MOF Bragg grating length L for different guiding modes is directly dependent upon the scattering efficiency of each mode, and more specifically upon the effective grating length L_eff of each scattered mode, where $L_{eff}=\frac{\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.L\sqrt{R}}{a\tan \text{h}\left(\sqrt{R}\right)}$ and R the mode reflectivity. By assuming uniform inscription profile and no-chirping effects, the effective length of the Bragg grating for the 0^th mode was estimated to be L_eff ^0th=1.39mm and for the 1^st guiding mode notch L_eff ^1st=2.47mm.

By considering these two L_eff values, the behaviour of the results presented in Fig. 3 and 4 becomes clearer. As the ferrofluid enters the grating region (positive points on Ox axis in Fig. 3) the 1^st order mode is affected first, while the fundamental grating mode undergoes significant spectral changes at a later phase and to a lesser extent. The above grating design parameter, also defines the spatial sensitivity of the specific in fibre actuator with respect to the spectral changes obtained.

An interesting spectral effect was observed for specific operation conditions of the MOF- device. When the ferrofluid is not overlapping with the grating, while being positioned either to the left or right side of the grating (see Fig. 5a and d ), the spectral results obtained in transmission mode (Fig. 5b, e) from the actuator are identical irrespectively of light excitation direction, denoting bidirectional operation. Bidirectional operation is not observed in reflection mode (Fig. 5 c, f), where the reflection measured when the ferrofluid is placed on the side of light excitation and before the grating length (see Fig. 5 d at State B). In such a case, the reflection measured is practically vanished for the higher order mode and substantially reduced for the fundamental mode peak (see Fig. 5f). This is due to the fact that the injected light is subjected twice in the absorption of the ferrofluidic length, once when reaches the grating; and then when the light is reflected back by it. Moreover, by exploiting this propagation effect and by using Beer’s absorption law, for a known ferrofluid length, we have estimated the transmission loss for the fundamental mode due to the ferrofluid to be −4.0dB/cm, approximately.

 

Fig. 5 Spectral measurements in transmission and reflection for different positions of the ferrofluid with respect to the Bragg grating. (a) State A (d) State B.

Download Full Size | PDF

The actuating operation of the ferrofluid infiltrated Bragg grating MOF was also tested using a stepper motor translation stage and light excitation using a tuneable laser emitting at the 1.5μm band. The laser was tuned at the center of each Bragg spectral notch, and its reflection and transmission was measured using a power-meter head. Power modulation results in transmission and reflection mode, are presented in Fig. 6 .

 

Fig. 6 Power modulation results for two cycles, in transmission and reflection obtained for the 1st guiding mode. Stepper motor speed 0.5mm/sec.

Download Full Size | PDF

The power modulation measured for the 1^st order mode in transmission is of the order of 2dB, mainly related to the bandwidth and spectral shape of the grating, while exhibiting good repeatability. Modulation figures of 7dB approximately, are obtained in reflection from the 1^st guiding mode; nonetheless, the repeatability of the actuation is decreased after repetitive cycles, possibly due to ferrofluidic residue deposited onto the capillary walls. Since reflection mode is not bidirectional while being quite sensitive to absorption loss changes, any residue left on the fibre walls will degrade the actuating behavior. Power modulation results of ≈6dB strength have been also obtained from the 0^th guiding mode in reflection.

4. Conclusion

We have presented a microstructured optical fibre Bragg grating actuator utilizing an infiltrated ferrofluidic matrix into the fibre capillaries. Details were provided for the functionalisation of the capillaries of the MOF and the ferrofluid infiltration process. The results presented refer to the spectral changes induced for different overlaps of the ferrofluid with the grating length, using a static magnetic field. Wavelength shifts of the Bragg peaks of the order of 0.2nm and significant modulation of the grating strength were measured. The MOF device was examined with respect to its actuating operation for different light launching directions. We are currently investigating the possibility of exploiting this ferrofluidic infiltrated MOF in the development of magnetic field miniature probes; as well as, of shear-stress and magnetic valves sensing units of sub-millimetre spatial accuracy. We are working on the optimization of the fibre capillaries wettability using suitable silanisation methods and with respect to the ferrofluid carrier, for increasing the actuation speed of the device.