Optical fiber laser refractometer based on an open microcavity Mach-Zehnder interferometer with an ultra-low detection limit

Panpan Niu; Panpan Niu; Panpan Niu; Junfeng Jiang; Junfeng Jiang; Junfeng Jiang; Shuang Wang; Shuang Wang; Shuang Wang; Kun Liu; Kun Liu; Kun Liu; Zhe Ma; Zhe Ma; Zhe Ma; Yongning Zhang; Yongning Zhang; Yongning Zhang; Wenjie Chen; Wenjie Chen; Wenjie Chen; Tiegen Liu; Tiegen Liu; Tiegen Liu

doi:10.1364/OE.401813

1. Introduction

Optical fiber refractometers have been widely researched owing to their intrinsic advantages of miniature size, high sensitivity, anti-electromagnetic interference and low cost, which can be easily embedded in microfluidic systems for chemical, biological and medical applications [1–3]. In the various optical fiber sensors for refractive index (RI) measurement, the optical fiber with microcavity structure is more suitable for the RI sensing in fluid liquid environment. The unique integrated all-fiber microcavity can be used as a sample cell to enhance the interaction between the light field and the medium to be measured [4,5]. These RI sensors are based upon different sensing elements and principles, such as whispering gallery mode (WGM) [6,7], surface plasmon resonance (SPR) [8,9], modal interferometers [10,11], optical gratings [12,13], etc., which can integrate sample collection, injection and measurement. For modal interferometers, the core-offset MZI is usually used as the sensing element since it can be easily fabricated by fiber fusion splicer [14–16]. The slight core-offset of single-mode fiber (SMF) leads to mode field mismatch, and cladding modes can be excited to sense the change of surrounding refractive index (SRI). However, the sensitivity of the traditional core-offset MZI is limited due to the weak evanescent field on fiber surface, and the overall size of sensing element is long (in centimeter scale).

In order to improve RI sensitivity, Duan et al. proposed an open microcavity Mach-Zehnder interferometer (OMZI) formed by large lateral offset splicing, the core misalignment is increased to expose the fiber core into surrounding medium [17]. The light from lead-in SMF will be separated into the surrounding medium as well as the lateral SMF, which constructs two interferometric arms of OMZI with different materials. Because of the direct interaction between light and surrounding medium, the RI sensitivity is higher than that of the traditional core-offset MZI. Xie et al. presented a microfluidic OMZI by splicing a segment of microfiber tapered from SMF [18], a U-shaped microcavity as the sample carrier is formed by the SMF-microfiber-SMF structure with an RI sensitivity of 10⁴ nm/RIU. However, owing to the thin core and restricted numerical aperture of the lead-in/out SMF, the OMZI with the direct coupling by SMF has large insertion loss, especially in liquid environment. Additionally, in order to obtain interference spectrum with high fringe visibility, the core-offset displacement requires high precision during the manufacturing process. To mitigate this problem, Baharin et al. proposed a modified OMZI based on symmetrical offset coreless silica fiber (CSF) configuration [19], the use of CSF without fiber core improves the core-offset tolerance while keeping high sensitivity of OMZI. Wang et al. demonstrated a multi-mode interference (MMI) coupling method for OMZI [20], by optimizing the length of lead-in/out multi-mode fiber (MMF), the coupling efficiency of interference light is enhanced and the insertion loss can be reduced to ∼10 dB. But the assisted fiber with large core obviously increases the size of sensing element which is also sensitive to SRI. Moreover, the RI measurements of the aforementioned researches are achieved by discriminating the wavelength drifts of resonant notches in the interference spectra with broadband light source (BBS), and thus the detection limit (DL) is limited by the full width at half maxima (FWHM) and optical signal-to-noise ratio (OSNR) of the resonant notch [21], even though the OMZI provides an ultra-high RI sensitivity.

In this paper, a fiber laser refractometer based on MMF-assisted OMZI is proposed and experimentally demonstrated. An open microcavity packaged with microfluidic device for liquid sample is constructed by sandwiching a tiny segment SMF between two coupling MMF joints with lateral-offset splicing. Assisted with the MMF joints, the proposed OMZI has a higher core-offset tolerance, and can be inserted into the erbium-doped fiber ring laser (FRL) cavity as a wavelength selector. Because of the RI-dependent spatial filtering characteristic of OMZI, the oscillating wavelength of FRL is modulated by the SRI. By combining the high RI sensitivity of OMZI with the narrow laser spectral width of FRL, the RI sensing with ultra-low DL can be realized by measuring the lasing wavelength of FRL. The proposed fiber laser refractometer possesses an ultra-low DL of 5.87×10⁻⁶ RIU and a high-quality factor Q of 2.23×10¹⁰, which are improved by two orders of magnitude comparing to the previous researches. The method will have good potential application in microfluidic RI measurement.

2. Principle

2.1. MMF-assisted OMZI and its RI-dependent spatial filtering characteristic

The schematic diagram of the proposed MMF-assisted OMZI is shown in Fig. 1, which consists of a small segment of SMF sandwiched between two lead-in/out MMF joints with lateral-offset. The fabrication procedures of the MMF-assisted OMZI are illustrated in Fig. 2, and the main operations involve splicing, cutting and offset-splicing. Firstly, two SMF (8.2/125 µm, SMF-28e, CORNING)-MMF (62.5/125 µm, 457WY, YOFC) segments are automatically splicing by a fiber fusion splicer (LDS 2.5, 3SAE). Secondly, the redundant MMFs are cut off by using ultrasonic cleaver with precise length control, and the length of the in/out MMF joints are L_mmf= 100 µm in our experiments. Finally, the processed SMF-MMF segments are non-coaxially spliced with a tiny segment of SMF, the offset is about half of the fiber diameter to expose the cores of lead-in/out MMFs. The cross-sectional and side-sectional geometries of the prepared OMZI are shown in Fig. 2(d). The offset SMF sandwiched between coaxial MMF constructs an open microcavity, and the microcavity length L_mc is determined by the length of offset SMF. As illustrated in Fig. 1, when light is launched through the lead-in MMF, the light is separated into two parts, which propagate in the cladding of offset SMF and the microcavity, and then they are coupled into the lead-out MMF. The two light beams interfere with each other, and therefore the proposed fiber structure can be considered as an in-line MZI. The transmission intensity guided by SMF can be expressed as:

(1)$$I = {I_f} + {I_{mc}} + 2\sqrt {{I_f}{I_{mc}}} \cos ({\Delta \varphi } )$$

where I_f and I_mc are the intensities of the light propagated in the offset SMF and the microcavity, respectively. Δφ is the phase difference between I_f and I_mc, and it can be expressed as:

(2)$$\Delta \varphi = \frac{{2\pi ({n_f} - {n_{mc}}){L_{mc}}}}{\lambda }$$

where n_f and n_mc are the effective RI of the SMF cladding mode and the RI of the medium in microcavity. λ is the free space wavelength in vacuum. In order to obtain maximum extinction ratio (ER) of transmission spectrum, it is important for the OMZI to balance the intensity of I_f and I_mc [22,23], which can be achieved by optimizing the offset. However, due to the thin core, high-visibility spectrum is difficult to obtain by adjusting offset for the OMZI constructed by SMF only. For the proposed OMZI assisted with lead-in/out MMFs, the offset tolerance of fabrication and the coupling efficiency of interference light can be increased owing to the large core diameter of MMF.

Fig. 1. Schematic diagram of the proposed MMF-assisted OMZI.

Download Full Size | PDF

Fig. 2. Illustration of the fabrication procedures of MMF-assisted OMZI: (a) splicing, (b) cutting and (c) offset-splicing. (d) Cross-sectional and side-sectional geometries.

Download Full Size | PDF

When the Δφ is multiple of (2m+1)π, the resonant notch of transmission spectrum will appear for an OMZI with fixed L_mc, where m is an integer. The wavelength of notch in the transmission spectrum can be expressed as:

(3)$${\lambda _{notch}} = \frac{{2({n_f} - {n_{mc}}){L_{mc}}}}{{(2m + 1)}}$$

The free spectral range (FSR) indicates the bandwidth between two adjacent notches of transmission spectrum, which can be approximated by:

(4)$$FSR = \frac{{{\lambda ^2}}}{{({n_f} - {n_{mc}}){L_{mc}}}}$$

For the traditional in-line MZIs based on fiber mode interferometers, the RI sensitivity is limited by the weak evanescent field on fiber surface. By contrast, because of the open cavity structure of OMZI, the surrounding medium can be considered as the waveguide of the light I_mc. The OMZI can respond to slight SRI change drastically as a result of the direct interaction between light and surrounding medium, and the SRI change is reflected in the drift of transmission spectrum. The RI sensitivity of transmission spectrum can be expressed as:

(5)$$\frac{{\partial {\lambda _{notch}}}}{{\partial {n_{mc}}}} = \frac{{ - 2{L_{mc}}}}{{2m + 1}}$$

It can be seen that the RI sensitivity is negative value, and the transmission spectrum will drift to the shorter wavelength with SRI increasing.

The microcavity length is an important structural parameter that influences the transmission characteristics of OMZI, including FSR and insertion loss of transmission spectrum. Hence, the OMZIs with different microcavity lengths are simulated by using the BeamPROP module of Rsoft software. The microcavity length of MMF-assisted OMZI is increased from 500 µm to 1900 µm with a step of 100 µm, the simulated transmission spectra in air are shown selectively in Fig. 3(a) for clear observation. Meanwhile, the MMF-assisted OMZIs with microcavity lengths of L_mc=500 µm, 700 µm, 900 µm, 1200 µm, 1500 µm and 1800 µm are fabricated for verification, the measured transmission spectra in air are shown in Fig. 3(b), and the microscopic images of them are shown in Fig. 3(c). The FSRs of simulated and measured transmission spectra are evaluated, as shown in Fig. 3(d). It could be found that the FSR is reduced with the L_mc elongation, which is in accordance with the curve based on Eq. (4) about the theoretical relationship between FSR and microcavity length. Moreover, longer microcavity length leads to additional transmission losses, so the insertion loss of OMZI increases with L_mc, and the spectral quality of OMZI with longer L_mc will be obviously degraded. Suppose the interference is caused by two transmission modes, first order Taylor series expansion of phase difference Δφ at wavelength λ_x can be obtained from Eq. (2):

(6)$$\Delta \varphi \approx \Delta \varphi ({\lambda _x}) - \frac{{2\pi {L_{mc}}({n_f} - {n_{mc}})}}{{\lambda _x^2}}(\lambda - {\lambda _x})$$

For a specific pair of interference modes, $\Delta \varphi ({\lambda _x})$ is a constant, and the OMZI has a cosine interference pattern:

(7)$$\cos (\Delta \varphi ) = \cos [{2\pi F(\lambda - {\lambda_x})} ]$$

where F is the spatial frequency, and it can be expressed as:

(8)$$F = \frac{{{L_{mc}}({n_f} - {n_{mc}})}}{{\lambda _x^2}}$$

Figure 4(a) shows the simulated optical field distributions of the MMF-assisted OMZIs with L_mc=500 µm, 1000 µm and 1500 µm in air, and the contour maps of the transverse fields at the model longitudinal lengths L_Sim=750 µm, 800 µm, 1000 µm and 1500 µm of the OMZI with 500 µm length are shown in Fig. 4(b). It can be seen that the light field energy is mainly concentrated in the core center of the lead-in MMF before entering the microcavity, as the light propagates to the boundary of the fiber and the microcavity with large RI difference, the light is divided into two parts, one part is coupled into the core-offset fiber, the other penetrates into the microcavity through the exposed fiber core. The exposed core causes more light energy to enter the microcavity and improves the RI sensitivity, but also causes more light energy to be lost. After traveling along the different media paths, the two light beams are recoupled into the lead-out MMF and produce the interference. In the lead-out SMF, there are still a large number of high-order modes, and these high-order modes will be consumed completely after propagating for a certain distance in the fiber cladding. The spatial frequency spectra of the MMF-assisted OMZIs with L_mc=500 µm, 1000 µm and 1500 µm in air are calculated by fast Fourier transform (FFT) as shown in Fig. 4(c). Except for the peak at zero-point F₀=0 nm⁻¹, there are peaks at F₁=0.096 nm⁻¹, F₂=0.160 nm⁻¹ and F₃=0.281 nm⁻¹ for L_mc=500 µm, 1000 µm and 1500 µm, respectively. The frequency component at F₀=0 nm⁻¹ represents the fundamental mode LP₀₁, while F₁, F₂ and F₃ represent the high-order modes LP_nm. As the elongation of microcavity length, the spatial frequency of non-zero peak increases, but the intensity decreases. Therefore, the FSR of OMZI cannot be compressed by lengthening microcavity length L_mc indefinitely, and the L_mc should be reasonably selected considering the insertion loss and spectral quality in practice. Due to the structural features of OMZI, other higher-order modes are greatly weakened after multiple coupling, which have little influence on the interference pattern. As a result, the transmission spectrum of OMZI shows a comb shape with high ER, which is conducive to the wavelength selection and mode competition inhibition in fiber ring cavity.

Fig. 3. (a) Simulated and (b) measured transmission spectra of the MMF-assisted OMZIs with different microcavity lengths in air. (c) Microscopic images of the fabricated MMF-assisted OMZIs with L_mc=500 µm, 700 µm, 900 µm, 1200 µm, 1500 µm and 1800 µm. (d) Relationship between FSR and microcavity length.

Download Full Size | PDF

Fig. 4. (a) Simulated optical field distributions of the MMF-assisted OMZIs with L_mc=500 µm, 1000 µm and 1500 µm in air. (b) Contour maps of the transverse fields at L_Sim=750 µm, 800 µm, 1000 µm and 1500 µm of the MMF-assisted OMZI with L_mc=500 µm. (c) Spatial frequency spectra of the MMF-assisted OMZIs with L_mc=500 µm, 1000 µm and 1500 µm in air.

Download Full Size | PDF

In order to investigate the improvement of offset tolerance, the transmission spectra of the MMF-assisted OMZIs with L_mc=500 µm, 1000 µm and 1500 µm at n=1.3330 are simulated, and the offset is changed from 50 µm to 75 µm with an interval of 2.5 µm. Figure 5 gives the simulated transmission spectra of the MMF-assisted OMZIs with different offsets. With the offset increasing, the ERs of transmission spectra are magnified gradually and maximized to 25 dB at 62.5 µm. However, with the further increase of offset, the ERs begin to be weakened. The ER of transmission spectrum can be expressed as:

(9)$$ER = 10 \cdot {\log _{10}}{\left( {\frac{{1 + \sqrt {{I_f}/{I_{mc}}} }}{{1 - \sqrt {{I_f}/{I_{mc}}} }}} \right)^2}$$

Due to the direct effect of offset on the splitting ratio of I_f and I_mc, the ER will be the maximum under the optimum offset. At the offset range of 55 µm∼70 µm, the ERs of the spectra are greater than 5 dB. The proposed OMZI assisted with MMF has the large overlapping area between fiber core and offset fiber. As a result, the offset tolerance can be enhanced to 15 µm, which is higher than that of the traditional OMZI using SMF only [17], and the requirement of offset accuracy for high-visibility spectrum is reduced during the manufacturing process.

Fig. 5. Simulated transmission spectra of the MMF-assisted OMZIs with different offsets at n=1.3330: (a) L_mc=500 µm, (b) L_mc=1000 µm and (c) L_mc=1500 µm.

Download Full Size | PDF

By changing the background RI of simulated OMZIs, the evolutions of the transmission spectra with SRI are shown in Fig. 6(a), and the relationships between the selected notch wavelengths and SRIs are shown in Fig. 6(b). Simulated results show that the transmission spectra have blueshifts with SRI increasing, and the RI sensitivities of the OMZIs with L_mc=500 µm, 700 µm, 900 µm, 1200 µm and 1500 µm are −2996.017 nm/RIU, −2952.363 nm/RIU, −3066.667 nm/RIU, −2892.333 nm/RIU and −2934.193 nm/RIU, respectively, during the SRI range of 1.3330∼1.3342. When the SRI changes in a small range, the transmission spectral drift remains linear, and the RI sensitivity is not affected by the microcavity length.

Fig. 6. Simulated results of the MMF-assisted OMZIs with different microcavity lengths in different SRIs: (a) transmission spectra evolutions at different SRIs, (b) relationships between the selected notch wavelengths and SRIs.

Download Full Size | PDF

2.2. MMF-assisted OMZI intra-cavity fiber ring laser refractometer

Due to the structural features of OMZI, the OMZI with spatial filtering characteristics can provide a high-contrast ratio filter to construct the wavelength selector, which has the transmission spectrum with high ER. Meanwhile, because the medium in the microcavity of OMZI can be considered as the waveguide of the sensing light, the spatial filtering characteristics of OMZI can response to SRI change with wavelength drift. Accordingly, the wavelength tunable laser for RI sensing can be achieved by inserting the OMZI into the FRL cavity, which will have the advantage of high optical power and narrow spectral width. The proposed fiber laser refractometer is thus established according to Fig. 7(a).

Fig. 7. (a) Experiment setup of the proposed fiber laser refractometer. (b) Photo of the microfluidic device. Inserts: microscopic images of the prepared OMZI and the MMF-assisted core-offset joint.

Download Full Size | PDF

The 980 nm pump light is injected into the ring cavity through a 980/1550 nm wavelength division multiplexer (WDM). A 1 m long highly doped erbium-doped fiber (EDF, Er80-4/125, LIEKKI) is used as the active medium to provide gain for the ring cavity. An MMF-assisted OMZI serves as the lasing tunable wavelength selector for RI sensing. The ring cavity is equipped with two optical isolators (ISO 1 and ISO 2), in order to ensure unidirectional propagation and avoid the end-face reflection effect due to the structural features of OMZI. A 5:95 optical coupler (OC) is used to extract the laser output from the 5% port, and the remain 95% is re-injected into the ring cavity. The laser output is monitored by an optical spectrum analyzer (OSA, AQ6370, Yokogawa). In addition, a polarization controller (PC) is inserted into the ring cavity to adjust the polarization state.

Because the transmission spectrum of OMZI has blueshifts with SRI increasing, the lasing wavelength should be located at the long wavelength region to obtain a stable and enough tuning range. Due to the large core-offset, the total cavity loss of the ring cavity is less than 23 dB, which mainly introduced by the OMZI with a typical insertion loss of 20 dB in liquid environment. However, the total cavity loss can be compensated by the active EDF, and a stable oscillation can be generated in the ring cavity. It is worth noting that the increase of microcavity length can compress the FSR of transmission spectrum and improve the ability of wavelength selection of OMZI, but the longer microcavity will lead to more loss in the ring cavity and reduce the output laser power and OSNR. According to the relationship between the transmission characteristic and the microcavity length of OMZI mentioned above, in order to balance the number of potential lasing wavelengths and lasing linewidth within the gain bandwidth, the microcavity length of 1200 µm with lower loss and regular spectrum is selected to build the OMZI for RI sensing. As the sensing element for liquid RI measurement, the MMF-assisted OMZI with L_mc=1200 µm is packaged by microfluidic device, as shown in Fig. 7(b), and the inserts of Fig. 7(b) give the microscopic images of the prepared OMZI and the MMF-assisted core-offset joint. The packaged sensing element is placed on a semiconductor thermostat for a constant temperature environment.

The original forward-pump amplified spontaneous emission (ASE) spectrum of the ring cavity with a pump power of 450 mW is given in Fig. 8 (Black line), which has typical EDF-based ASE spectral characteristics. The transmission spectrum of OMZI at n=1.33302 (Red line) and the filtered ASE spectrum (Blue line) are shown in Fig. 8. There are three peaks within the gain bandwidth. In the closed ring cavity, the single wavelength lasing oscillation of the peak near 1560 nm is realized by adjusting the polarization dependent loss (PDL) of ring cavity with PC, which has narrower FWHM of 0.0164 nm and higher OSNR of 42 dB compared with transmission spectrum of OMZI, as shown in Fig. 8 (Green line). Due to the assistance of MMF, the transmission spectrum shows a comb shape with a high ER of 15 dB and a FWHM of 8 nm. When the OMZI with quality spatial filtering characteristics is inserted into the ring cavity with saturated EDF, the corresponding filter peak obtains the cyclic-cumulative gain and be amplified in the ring cavity. Eventually, the stimulated emission is generated at the filter peak wavelength, and the ring cavity can output a single-wavelength laser with a relative narrow FHWM without additional linewidth compression method.

Fig. 8. Original and filtered forward-pump ASE spectra, transmission spectrum of OMZI, and output laser spectrum.

Download Full Size | PDF

3. Experimental results and discussions

To investigate the RI response of the proposed fiber laser refractometer, a series of RI liquid samples are prepared by diluting the NaCl standard solution (PST007A, PHYGENE) with deionized water, the concentrations of the RI samples are increased from 0 mol/L to 0.1 mol/L with a corresponding RI range of 1.33302∼1.33402 at temperature of 26 °C [24]. The RI samples are injected into the microfluidic device by syringes, and then collected by a liquid pool after flowing through the sensor element. Before the sensing element is inserted into the ring cavity, the transmission spectra of the MMF-assisted OMZI are measured at different concentrations, as shown in Fig. 9(a), and the relationship between the notch wavelength at 1566.329 nm and the SRI by linear fitting is shown in Fig. 9(b). As the NaCl concentration increasing, the transmission spectra drift to short wavelength. The RI sensitivity can reach 3062.857 nm/RIU during 1.33302∼1.33402, which is close to the simulation result of 2892.333 nm/RIU. The sensitivity difference about 5% between the experiment and the simulation may be due to the subtle difference of geometric parameters and material properties between the ideal simulation model and the actual fabricated structure.

Fig. 9. (a) Transmission spectra of the MMF-assisted OMZI at different concentrations. (b) Linear fitting result of the relationship between notch wavelength and SRI.

Download Full Size | PDF

After 30 mins warm-up for the pump source, the RI experiment of FRL is carried out. In the experiment, each RI sample is measured five times with an interval of 1 minute, and the output laser spectra are recorded by OSA. Figure 10 shows the measured output laser spectrum evolution of the fiber laser refractometer at different concentrations. The measured results of the center wavelength changing with time at different concentrations are shown in Fig. 11. With the increasing of the concentration of NaCl solution, the center wavelength of output laser drifts to short wavelength, and the peak power remains stable relatively. Because the absorption loss caused by the increase of NaCl concentration, can be compensated by the erbium fiber working in the saturated gain state, the output laser has regular wavelength drift rather than obvious power fluctuation. The insert of Fig. 11 gives the peak wavelength (black) and the peak power (blue) change of the measured output laser in response to concentrations, the NaCl concentration sensitivity is −29.578 nm/mol·L⁻¹ during the range of 0 mol·L⁻¹∼0.1 mol·L⁻¹. The linear fitting result with error bars of the relationship between laser center wavelength and SRI are shown in Fig. 12. During the SRI range of 1.33302∼1.33402, an average RI sensitivity of −2953.444 nm/RIU is achieved, which is higher than that of the traditional fiber laser refractometer without open microcavity [25–27]. The measurement range is limited to a relatively small range of 0.001, which covers most of RI measurement requirements in biochemical and food fields. At the same time, the measurement range is also helpful for stable laser output operation. Furthermore, the output lasers at the minimum and maximum of measurement range are monitored, which operate at 1559.527 nm and 1556.577 nm, respectively, during half an hour to examine the stability of single-wavelength oscillation. The repeated scanning spectra of the output laser are shown in Fig. 13. It can be observed that the center wavelength and peak power of the output laser keep unchanged without observable fluctuations. Therefore, the single-wavelength fiber laser refractometer with high sensitivity and stability are achieved. In addition, due to the small dispersion-coefficients difference between the optical fiber and the NaCl solution, the RI measurement is not affected by the dispersion, which adopts the static wavelength demodulation method by OSA.

Fig. 10. Measured output laser spectrum evolution of the fiber laser refractometer at different concentrations.

Download Full Size | PDF

Fig. 11. Measured results of the center wavelength changing with time at different concentrations. Insert: The peak wavelength (black) and the peak power (blue) change of the measured output laser in response to concentrations.

Download Full Size | PDF

Fig. 12. Linear fitting result with error bars of the relationship between laser center wavelength and SRI.

Download Full Size | PDF

Fig. 13. Repeated scanning spectra of the output laser at (a) n=1.33302 and (b) n=1.33402.

Download Full Size | PDF

In order to evaluate the comprehensive performance of sensors, the DL and the quality factor Q are usually introduced. For refractometric sensing, the DL indicates the minimum RI change that can be measured accurately, which is a standard approach for quantifying and comparing the performance of optical resonance-based RI sensors [21]. The DL can be calculated as:

(10)$$DL = \frac{R}{S} = \frac{{3\sqrt {\sigma _{_{ampl - noise}}^2 + \sigma _{_{temp - induced}}^2 + \sigma _{_{spect - res}}^2} }}{S}$$

where S is the RI sensitivity of wavelength drift, R is the quantitative resolution which is determined by the standard deviations of amplitude noise ${\sigma _{ampl - noise}}$, spectral quantization error ${\sigma _{spect - res}}$ and temperature noise ${\sigma _{temp - induced}}$. It should be noted that the standard deviation of amplitude noise ${\sigma _{ampl - noise}}$ is related to the FWHM and OSNR of the sensing spectrum, and it can be expressed as [21]:

(11)$${\sigma _{ampl - noise}} \approx \frac{{FWHM}}{{4.5 \times {{({OSNR} )}^{0.25}}}}$$

where OSNR is the proportion to the ER of interference spectrum for resonant RI sensors. In addition, the quality factor Q is a normalized value to indicate the sensing properties of optical fiber sensors with consideration of spectral quality [28], which can be expressed as:

(12)$$\textrm{Q} = \frac{{K \cdot {S^2} \cdot OSNR}}{{FWHM}}$$

where K is a unit coefficient to normalize the physical dimension. Owing to the contribution of FRL, the OSNR and the FWHM of the sensing spectrum in our work are 42 dB and 0.0164 nm, respectively. Consequently, the calculated DL and Q can be enhanced to 5.87×10⁻⁶ RIU and 2.23×10¹⁰. The comparison of sensing features between the proposed fiber laser refractometer and other fiber core-offset based schemes is listed in Table 1. The DLs and Q values of the reported fiber refractometers are calculated based on the representative references. Compared with the refractometers based on slight core-offset joints, the proposed MMF-assisted OMZI refractometer has higher sensitivity and smaller size, which is more suitable for precise localized measurement in biochemistry and biomedicine. Although the microfiber-assisted OMZIs have higher sensitivity, the DL and Q value about two orders of magnitude better than that of the listed references are achieved by utilizing the narrower FWHM and higher OSNR of FRL in our work. Furthermore, the proposed fiber laser refractometer eliminates the need for fragile microfibers and special fibers, which has the advantages of compact size, low cost, ultra-low DL and high Q.

Table 1. Comparison of sensing features between the proposed fiber laser refractometer and other fiber core-offset based schemes.

View Table

4. Conclusion

In this paper, a fiber laser refractometer based on MMF-assisted OMZI is proposed and demonstrated. An MMF-assisted OMZI is fabricated by embedding a 1.2 mm SMF segment into two 100 µm MMF joints with lateral-offset splicing, which is introduced into the fiber ring cavity as a tunable wavelength selector. Experimental results show that the output laser wavelength of the proposed refractometer has a good linear response to SRI, and the sensitivity is as high as −2953.444 nm/RIU during the range of 1.33302∼1.33402. Due to the narrow FWHM and high OSNR of FRL, the enhanced DL of 5.87×10⁻⁶ RIU can be achieved with a high spectral quality factor Q of 2.23×10¹⁰. Furthermore, the offset tolerance of microcavity structure is enhanced by the auxiliary of large core MMF. Compared with other optical fiber refractometers, the proposed fiber laser refractometer with open microcavity has the advantages of micro sensing element, ultra-low DL and easy fabrication, which is a good candidate for the microfluidic RI measurement in biochemistry.

Funding

State Key Laboratory of Information Photonics and Optical Communications (2019KFKT007); the first rank of Tianjin 131 Innovation Talent Development Program; Tianjin Talent Development Special Plan for High Level Innovation and Entrepreneurship Team; National Key Scientific Instrument and Equipment Development Projects of China (2013YQ030915); National Natural Science Foundation of China (61675152, 61735011).

Disclosures

The authors declare no conflicts of interest.

References

1. L. Sun, T. Huang, Z. Yuan, M. Yang, Y. Huang, P. Xiao, and B. Guan, “Ultrasensitive optofluidic interferometer for online monitoring of photocatalytic reactions,” J. Lightwave Technol. 37(21), 5435–5441 (2019). [CrossRef]

2. B. Song, H. Zhang, B. Liu, W. Lin, and J. Wu, “Label-free in-situ real-time DNA hybridization kinetics detection employing microfiber-assisted Mach-Zehnder interferometer,” Biosens. Bioelectron. 81, 151–158 (2016). [CrossRef]

3. L. Li, H. Xiang, Y. Xiong, H. Zhao, Y. Bai, S. Wang, F. Sun, M. Hao, L. Liu, T. Li, Z. Peng, J. Xu, and T. Zhang, “Ultrastretchable fiber sensor with high sensitivity in whole workable range for wearable electronics and implantable medicine,” Adv. Sci. 5(9), 1800558 (2018). [CrossRef]

4. Y. Zhao, H. Zhao, R. Lv, and J. Zhao, “Review of optical fiber Mach-Zehnder interferometers with micro-cavity fabricated by femtosecond laser and sensing applications,” Opt. Laser Eng. 117, 7–20 (2019). [CrossRef]

5. A. B. Socorro-Leránoz, D. Santanoa, I. Del Villar, and I. R. Matias, “Trends in the design of wavelength-based optical fibre biosensors (2008–2018),” Biosens. Bioelectron.: X 1, 100015 (2019). [CrossRef]

6. S. Yang, Y. Wang, and H. Sun, “Advances and prospects for whispering gallery mode microcavities,” Adv. Opt. Mater. 3(9), 1136–1162 (2015). [CrossRef]

7. J. Silverstone, S. McFarlane, C. Manchee, and A. Meldrum, “Ultimate resolution for refractometric sensing with whispering gallery mode microcavities,” Opt. Express 20(8), 8284–8295 (2012). [CrossRef]

8. Y. Liu, X. Liu, S. Chen, Q. Liu, and W. Peng, “Investigation of a capillary-based surface plasmon resonance sensor for biosensing,” J. Lightwave Technol. 34(17), 4036–4042 (2016). [CrossRef]

9. X. He, H. Yi, J. Long, X. Zhou, J. Yang, and T. Yang, “Plasmonic crystal cavity on single-mode optical fiber end facet for label-free biosensing,” Appl. Phys. Lett. 108(23), 231105 (2016). [CrossRef]

10. T. Wieduwilt, J. Dellith, F. Talkenberg, H. Bartelt, and M. A. Schmidt, “Reflectivity enhanced refractive index sensor based on a fiber-integrated Fabry-Perot microresonator,” Opt. Express 22(21), 25333–25346 (2014). [CrossRef]

11. M. Smietana, M. Janik, M. Koba, and W. J. Bock, “Transition between bulk and surface refractive index sensitivity of micro-cavity in-line Mach-Zehnder interferometer induced by thin film deposition,” Opt. Express 25(21), 26118–26123 (2017). [CrossRef]

12. G. Marshall, R. Williams, N. Jovanovic, M. Steel, and M. Withford, “Point-by-point written fiber-Bragg gratings and their application in complex grating designs,” Opt. Express 18(19), 19844–19859 (2010). [CrossRef]

13. S. Warren-Smith, R. Kostecki, L. Nguyen, and T. Monro, “Fabrication, splicing, Bragg grating writing, and polyelectrolyte functionalization of exposed-core microstructured optical fibers,” Opt. Express 22(24), 29493–29504 (2014). [CrossRef]

14. X. Yu, J. Ju, J. Zhang, S. Liu, and W. Jin, “In-fiber modal interferometer formed by offset-splicing in air-core photonic bandgap fiber,” Opt. Commun. 323, 143–147 (2014). [CrossRef]

15. Y. Zhao, X. Lia, and L. Cai, “A highly sensitive Mach-Zehnder interferometric refractive index sensor based on core-offset single mode fiber,” Sens. Actuators, A 223, 119–124 (2015). [CrossRef]

16. J. Liu, X. Zhang, J. Yang, J. Kang, and X. Wang, “Common difference temperature compensation based fiber refractive index sensor through asymmetrical core-offset splicing,” Opt. Commun. 427, 261–265 (2018). [CrossRef]

17. D. Duan, Y. Rao, L. Xu, T. Zhu, D. Wu, and J. Yao, “In-fiber Mach-Zehnder interferometer formed by large lateral offset fusion splicing for gases refractive index measurement with high sensitivity,” Sens. Actuators, B 160(1), 1198–1202 (2011). [CrossRef]

18. N. Xie, H. Zhang, B. Liu, H. Liu, T. Liu, and C. Wang, “In-line microfiber-assisted Mach-Zehnder interferometer for microfluidic highly sensitive measurement of salinity,” IEEE Sens. J 18(21), 8767–8772 (2018). [CrossRef]

19. N. F. Baharin, A. I. Azmi, A. S. Abdullah, and Y. M. N. Muhammad, “Refractive index sensor based on lateral-offset of coreless silica interferometer,” Opt. Laser Technol. 99, 396–401 (2018). [CrossRef]

20. L. Wang, L. Yang, C. Zhang, C. Miao, J. Zhao, and W. Xu, “High sensitivity and low loss open-cavity Mach-Zehnder interferometer based on multimode interference coupling for refractive index measurement,” Opt. Laser Technol. 109, 193–198 (2019). [CrossRef]

21. I. M. White and X. Fan, “On the performance quantification of resonant refractive index sensors,” Opt. Express 16(2), 1020–1028 (2008). [CrossRef]

22. D. Wu, T. Zhu, M. Deng, D. Duan, L. Shi, J. Yao, and Y. Rao, “Refractive index sensing based on Mach-Zehnder interferometer formed by three cascaded single-mode fiber tapers,” Appl. Opt. 50(11), 1548–1553 (2011). [CrossRef]

23. J. Fan, J. Zhang, P. Lu, M. Tian, J. Xu, and D. Liu, “A single-mode fiber sensor based on core-offset inter-modal interferometer,” Opt. Commun. 320, 33–37 (2014). [CrossRef]

24. W. M. Haynes, D. R. Lide, and T. J. Bruno, CRC Handbook of Chemistry and Physics (CRC, 2014), Chap. 5.

25. Y. Zhao, L. Cai, and X. Li, “In-fiber Mach-Zehnder interferometer based on up-taper fiber structure with Er3+ doped fiber ring laser,” J. Lightwave Technol. 34(15), 3475–3481 (2016). [CrossRef]

26. J. Liu, M. Wang, X. Liang, Y. Dong, H. Xiao, and S. Jian, “Erbium-doped fiber ring laser based on few-mode-singlemode-few-mode fiber structure for refractive index measurement,” Opt. Laser Technol. 93, 74–78 (2017). [CrossRef]

27. X. Liu, X. Zhang, J. Yang, and X. Du, “Dual-ring dual-wavelength fiber laser sensor for simultaneous measurement of refractive index and ambient temperature with improved discrimination and detection limit,” Appl. Opt. 58(27), 7582–7587 (2019). [CrossRef]

28. X. Bai, D. Fan, S. Wang, S. Pu, and X. Zeng, “Strain sensor based on fiber ring cavity laser with photonic crystal fiber in-line Mach-Zehnder interferometer,” IEEE Photonics J. 6(4), 1–8 (2014). [CrossRef]

29. L. Cai, Y. Zhao, and X. Li, “A fiber ring cavity laser sensor for refractive index and temperature measurement with core-offset modal interferometer as tunable filter,” Sens. Actuators, B 242, 673–678 (2017). [CrossRef]

30. S. Gao, W. Zhang, Z. Bai, H. Zhang, P. Geng, W. Lin, and J. Li, “Ultrasensitive refractive index sensor based on microfiber-assisted U-shape cavity,” IEEE Photonics Technol. Lett. 25(18), 1815–1818 (2013). [CrossRef]

Sensing configuration	Description	Demodulation	Sensing element Size (mm)	Sensitivity (nm/RIU)	DL (RIU)	Q	Ref.
Slight core offset + FRL	Common SMF with slight core-offset to form MZI for active fiber laser wavelength tune	Laser output wavelength	53	−38.10	4.54×10⁻⁴	2.72×10⁶	[29]
Slight core offset + FRL	Slight core-offset SMF based MZI embedded in fiber Bragg grating with dual-FRL cavity, for temperature compensation	Laser output wavelength	50	−30.57	1.16×10⁻³	2.88×10⁵	[27]
Slight core offset + MMF	Slight core-offset SMF MZI assisted by single MMF coupling with BBS	Broadband resonant notch wavelength	46	−64.89	4.62×10⁻⁴	6.17×10⁴	[23]
MMF + CSF assisted OMZI	Tiny CSF sandwiched by MMF joints with BBS for low-loss	Broadband resonant notch wavelength	11.5	1364.343	6.18×10⁻⁴	9.31×10⁶	[20]
CSF-assisted OMZI	Core-offset CSF based MZI assisted by CSF with BBS	Broadband resonant notch wavelength	2.5	750	2.31×10⁻⁴	3.09×10⁶	[19]
Microfiber assisted OMZI	SMF MZI assisted by tapered SMF microfiber with BBS, for sensitivity improvement	Broadband resonant notch wavelength	0.688	∼13000	1.34×10⁻⁴	3.45×10⁸	[30]
Microfiber assisted OMZI	SMF MZI assisted by tapered SMF microfiber with BBS, for sensitivity improvement	Broadband resonant notch wavelength	0.352	15118.75	2.85×10⁻⁴	1.99×10⁸	[18]
Our work	MMF-assisted OMZI with active microcavity for high DL sensing	Laser output wavelength	1.4	−2953.444	5.87×10⁻⁶	2.23×10¹⁰

Optical fiber laser refractometer based on an open microcavity Mach-Zehnder interferometer with an ultra-low detection limit

Abstract

1. Introduction

2. Principle

2.1. MMF-assisted OMZI and its RI-dependent spatial filtering characteristic

2.2. MMF-assisted OMZI intra-cavity fiber ring laser refractometer

3. Experimental results and discussions

4. Conclusion

Funding

Disclosures

References

Cited By

Figures (13)

Tables (1)

Equations (12)

Optics Express