**1. Introduction**

This paper is the extended version of the proceedings paper presented at the 1st International Electronic Conference on Biosensors, 2–17 November 2020 [1].

A biosensor allows for the fast detection and quantification of a biological analyte, without pre-enrichment steps. It is characterized by two components: a biological element, which is often key to the specificity, and a transducer [2]. The biological element frequently takes the form of a bioreceptor, which is bound to the surface of the transducer. This binding requires several steps of surface modification or functionalization, which can be complex, time-consuming and/or expensive. On top of that, functionalization can shorten the lifespan of the biosensor and often puts heavy requirements on the storage conditions. These bottlenecks can be avoided by adding the biological element into the sample volume instead of binding it to the transducer. Among these biological elements are endolysins: produced by bacteriophages or bacterial cells, they have the capability to specifically digest the cell wall of specific bacterial strains. They have proven to be powerful specificity means for the detection of bacteria [3–5].

The transducer can rely on electrical, optical, thermal, or magnetic signals. Optical transducers however outperform most physical transducers as there is no influence of the

**Citation:** Vercauteren, R.; Leprince, A.; Mahillon, J.; Francis, L.A. Porous Silicon Biosensor for the Detection ofBacteria through Their Lysate. *Biosensors* **2021**, *11*, 27. https://doi.org/doi:10.3390/ bios11020027

Received: 24 December 2020 Accepted: 16 January 2021 Published: 20 January 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

nature of the sample or of external disturbances on the signal [6]. Optical biosensors can take many forms: waveguides [7], ring resonators [8], refractometers [9,10], surface plasmon resonators [11], optical fibers [12], or even photonic crystals [13]. Porous silicon (PSi) is one widely used optical transducer for biosensors [14]. Its benefits include a large surface area, unique optical properties and a low production cost. While PSi photonic crystal [15], ring resonators [16,17], and photoluminescent sensors [18] have been demonstrated, most PSi-based optical transducers act as interferometers and rely on changes in the effective optical thickness (EOT) of the porous layer as means of detection. The EOT depends on both the effective refractive index and the thickness of the porous layer and can be quantified using reflective interferometric Fourier transform spectroscopy (RIFTS) [19,20]. The RIFTS method consists in shining a halogen light perpendicularly to the sensor surface and measuring its reflection. The reflection takes the form of Fabry-Pérot fringes due to the interferences between the light reflections from the top and the bottom interfaces of the porous film. By applying a Fourier transform to this fringe pattern, it is possible to extract their frequency, which takes the shape of a peak. The position of this peak translates the EOT. When an analyte penetrates the porous matrix, it affects the refractive index of the layer medium and induces a wavelength shift in the interference pattern; this is translated by a shift of the EOT value.

Recent works using the RIFTS as transducing mechanism of PSi-based transducers target the detection of, for instance, bacterial surface proteins [21], heat shock protein 70 [22], or bovine mastitis biomarkers [23,24], but this method has also been applied for the detection of bacteria [25–27]. These bacterial detectors however lacked sensitivity, with detection limits unable to go below 10<sup>3</sup> colony forming units per mL (CFU/mL). Current bacteria detection techniques, such as polymerase chain reaction (PCR) or enzyme-linked immunosorbent assay ELISA, while time-consuming, can easily reach detection limits of 10 CFU/mL [28]. This insufficient sensitivity is attributed to the hindered diffusion of bacteria into the porous matrix. Solutions, such as the electrokinetic transport for the preconcentration of analyte, have been proposed but remain to be tested on bacteria [29]. Recently, porous silicon interferometers have also been combined with gold nanoparticles for localized surface plasmon spectroscopy (LSPS), enhancing the fringe pattern contrast and increasing the sensitivity of the porous layer [30,31].

Another approach to increase the sensitivity of PSi biosensor is the fabrication of PSi membranes (PSiMs). Instead of a close-ended PSi layer over which the analyte must flow, an open-ended PSi membrane allows the analyte to flow through the porous matrix [32]. PSi membranes have been fabricated as early as the 1990s [33], but the interest in PSiMs for sensing applications only arose recently [34–38]. PSiMs can be self-supported, meaning still partly attached to the silicon substrate, or freestanding. Recently, lateral PSiMs have also been fabricated [39–41]. For biosensing applications, PSiMs have been found to increase both the response time and amplitude of the sensors [34–38].

In this work, we combine the benefits of selective endolysins and PSiM-based transducers for the fabrication of an innovative biosensor which enables the fast and label-free detection of bacteria through their lysate. The biosensing platform operates in two steps: a first step is the selective lysis of the targeted bacteria by an endolysin in a vial; and secondly, the optical monitoring of the bacterial lysate filtering through a PSiM using the RIFTS method. We demonstrate this concept with a selective optical detection of *Bacillus cereus* lysate in PBS using the recently characterized PlyB221 endolysin, encoded by the Deep-Blue phage targeting *B. cereus* [42]. The specificity is confirmed by replacing the targeted bacteria with *Staphylococcus epidermidis*, for which no optical detection was observed. To illustrate the versatility of our detection platform, the targeted bacteria strain was then switched to *S. epidermidis*, using lysostaphin as selective lytic agent.

#### **2. Materials and Methods**
