1041-1135 (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/LPT.2014.2344857, IEEE Photonics Technology Letters 1 Optimizing the signal to noise ratio of microcavity sensors M. Imran Cheema, Ce Shi, Andrea M. Armani, Senior Member, IEEE and Andrew G. Kirk, Member, IEEE Abstract—Optical microcavities provide a sensitive platform for biological and chemical sensing applications. Previous work has focused on optimizing the device performance by increasing the quality factor of the cavity. However, this approach overlooks the impact of the quality factor on the signal to noise ratio of the sensor. Here, our purpose is to show the existence of optimum parameters, both theoretically and experimentally, for achieving maximum signal to noise ratio in the microcavity sensors. Although toroidal cavities are used here, the present work is easily generalizable to any cavity geometry, enabling the performance optimization of a wide range of microcavity sensors. Index Terms—Microcavity sensors, signal to noise ratio, quality factor I. I NTRODUCTION O VER the last few years, optical microcavities have been utilized in a wide range of biological and chemical sensing applications, ranging from the detection of cytokines in serum to heavy water detection. Typically, the sensing experiments involve measuring a change, in either the resonant wavelength, Δλ, or the quality factor, ΔQ, of the cavity, in response to a refractive index or optical absorption change [1], [2]. One motivation for pursuing microcavity based sensors over alternative devices is their high quality factors (Q) or long photon lifetimes at well-defined resonant wavelengths. Additionally, because it is currently believed that when using the wavelength shift approach, the narrow linewidth inherent in higher Q devices enables smaller sensing events to be resolved; research has focused on maximizing the Q of the cavity [3]. However, without considering the various noise mechanisms present in a microcavity sensor, it has been suggested, via simulations, that the noise in the wavelength measurements (Δλ) depends upon the Q of a microcavity [4], [5]. Because the signal to noise ratio (SNR) governs the overall performance of a sensor system, the balance between the signal and the noise should be the focus of any optimization process. Here, by signal we mean changes in the resonant wavelength, Δλ, or the quality factor, ΔQ, of a microcavity in response to a sensing event. One way to improve the SNR is to increase the magnitude of the signal. For the wavelength shift measurement approach, enhancement of the signal has been achieved by replacing Imran Cheema is with the Electrical Engineering Department, National University of Computer & Emerging Sciences, Lahore, Pakistan. email: imran.cheema@nu.edu.pk Ce Shi, and Andrea M. Armani are with the Mork Family Department of Chemical Engineering and Materials Science, University of Southern California, Los Angeles, California, USA. email: armani@usc.edu Andrew Kirk is with the Department of Electrical and Computer Engineer- ing, McGill University, Montreal, Canada. e-mail: andrew.kirk@mcgill.ca the passive microcavity with a microlaser [6] or by attaching plasmonic nanoparticles to microcavities [7]. On the other hand, methods to increase the magnitude of the signal in the Q measurement technique have not yet been demonstrated. To increase the SNR by decreasing the noise, researchers have utilized interferometric techniques [8] for the wavelength mea- surements, and phase shift-cavity ring down spectroscopy [9] for the Q measurements. Here, without implementing the alternative schemes for either the signal enhancement or the noise reduction, we show that a microcavity sensor can achieve an optimum SNR by simply optimizing its dimensions and Q. Interestingly, the optimum Q is not the highest possible Q of the cavity. It is important to note that, for a sensing application, there are various methods of measuring a change in Q or in wavelength. For example, the Q of a microcavity can be determined by either a linewidth measurement [10] or by using phase shift-cavity ring down spectroscopy [9]. Similarly, the wavelength can be tracked by either continuously scanning the laser across the resonance and tracking the minimum [11], or fixing a laser at a single frequency and monitoring the change in transmission [12]. In the present work we focus on the linewidth change and the peak tracking approaches for the Q and the wavelength measurements, respectively, for a microtoroidal cavity. However, the work presented here is applicable to any testing methodology and microcavity geometry. In the present letter, we optimize the SNR of a microtoroidal cavity [13], that is utilized as a refractometric sensor at 765nm, by measuring change in the wavelength, Δλ, and change in the quality factor, ΔQ, of the cavity. The behavior of the signals (Δλ, ΔQ), and the noise in the wavelength measurements (σ λ ), as a function of the Q of the cavity, is determined by both simulations and experiments. Due to experimental challenges, the behavior of the noise in the quality factor measurements (σ Q ), as a function of the Q of the cavity, is only modeled. It should be noted that in this letter when we refer to the term noise it means the standard deviation of a normal distribution. After establishing the dependance of the signals and the noise on the Q of the cavity, we determine the optimum parameters (quality factor and cavity dimensions) to maximize the SNR of the sensor. II. SIGNAL ANALYSIS In this section, we provide the modeling and the experi- mental results for the sensing signals, change in the wave- length (Δλ) and change in the quality factor (ΔQ) of a refractometric sensor based upon a microtoroidal cavity.