Introducing MARC: A new concept for multiplexed sensing on a photonic sensor

We are developing an optical sensing device for multiplexed measurements of biomarkers in blood and have applied for a patent. The new sensor design named MARC, which is short for Mach-Zehnder interferometer assisted ring resonator configuration, makes it easier to multiplex and has a larger dynamic range.

The Lab-on-a-Chip (LOC)-project is developing an optical sensing device for multiplexed measurements of biomarkers in blood. We are using what’s called an optical ring resonator as the sensing element. The reason for this choice is that the ring resonator has high sensitivity and is fast, small, and fabricated with techniques already used by the semiconductor industry. All of these things put together means that with careful design we can place tens or hundreds of sensors on a chip that is just a couple of square centimeters (or smaller) allowing for fast multiplexed sensors that can be easily mass produced.

The new sensor design makes it easier to multiplex and has a larger dynamic range than a standard ring resonator. We have named this sensor design Mach-Zehnder interferometer assisted ring resonator configuration, or MARC for short, and have applied for a patent. PhD student Mukesh Yadav is the main inventor and prof. Astrid Aksnes, prof. Dag Roar Hjelme and PhD student Jens Høvik are co-inventors.

To explain how MARC works, we first have to look at how a standard optical ring resonator works. The device consists of a straight waveguide and a circular waveguide placed very close to each other (Figure 1).

Figure 1: Schematic of a ring resonator.

The waveguides have a higher refractive index than their surroundings, so that when light is coupled into one end of the straight waveguide, the light travels through the waveguide due to total internal reflection and can be collected and measured at the other end. The ring is placed so close to the waveguide that light passing through the waveguide can, given the right conditions, “jump” from the waveguide and start traveling around and around in the ring instead. This happens because light propagating through a waveguide due to total internal reflection will penetrate the surroundings of the waveguide as an evanescent wave with a penetration depth of a few hundred nanometers. As long as the ring is well within this distance from the straight waveguide the light can “feel” the ring and potentially make the jump. The right condition for this jump is given by the resonance condition of the ring:

2πr* neff=m* λ

meaning that it depends on the ring radius (r), the wavelength of the light (λ) and the effective refractive index (neff) of the ring. The effective refractive index will depend on the waveguide itself and the local environment of the ring, due to the penetration depth of the evanescent wave.

This “jumping light” can be observed if you at one end of the waveguide send in light that sweeps over a set of wavelengths, and then measure the intensity of the light that comes out at the other end. You can easily observe the wavelengths that are able to make the jump as distinct and clear dips in the measured intensity at the output. This is what we use for sensing, as any change in the effective refractive index (i.e. the local environment of the ring) will give a change in which wavelengths that can make the jump (Figure 2).

Figure 2: (a) Schematic representation of a ring resonator. According to the resonance condition, only selected wavelengths can propagate in the ring and distinct resonance peaks appear in the output spectrum. (b) Molecular binding occurs if a sample of the analyte is in close proximity to the adsorbed layer on top of the waveguide. This leads to a resonance wavelength shift Δλ.

By immobilizing antibodies for a specific protein on the surface of the ring resonator, the protein we want to detect will be captured by the antibodies and attach to the sensor surface. This attachment will displace some water, giving a small refractive index change that we can detect as a shift in the resonance wavelength, i.e., the dip in intensity at the output. This shift will also be concentration dependent, as the more protein we capture the more water is displaced and the larger the shift will be. In this way, we can utilize this sensor setup to measure the concentrations of proteins in solution (like biomarkers in blood).

Figure 3: A schematic representation of the biosensing principle on a ring resonator. Immobilized antibodies (pink) capture the biomolecule of interest (green) as it flows over the sensor. The captured biomolecule causes a local refractive index change, resulting in a measurable shift in the resonating wavelength (from blue to red). This shift is concentration dependent and can be used to determine the amount of the biomolecule that is present. Figure made using

Given that these ring resonator sensors are fast, selective, and sensitive and so small that we can place tens to hundreds of them on a 1x1 cm chip, you might now wonder why we need to improve on them with a new design and patent? Well, as we see from the resonance condition, there will be more than one wavelength that can resonate in the ring. For larger changes in the refractive index, the shift in resonance can be so large that the resonance shift is equal to or larger than the distance between two subsequent resonating wavelengths. It will then be hard to know if a small or a large shift in the resonating wavelength has occurred. One way to fix this is to use a laser light source that sweeps over the wavelength window so fast that you can track the shift as it happens. These laser lights sources are, however, very expensive and therefore not an option when the goal is to make an affordable sensor setup.

To solve this challenge, we have designed the MARC sensor that combines a Mach-Zehnder interferometer with a ring resonator. This design gives a more complex signature of the output signal than a single dip in the intensity for a certain wavelength (Figure 4).

Figure 4: MARC and different “fingerprint” transmission spectra.

Since it is more complex, it will also repeat itself over a larger range of wavelengths giving a larger dynamic range of the sensor before it overlaps with itself and makes analysis difficult. Another selling point for this design is that this complex signal can then be used as a fingerprint of a specific ring resonator. By changing the angle between the input and output ports, we can make MARC-sensors with different fingerprints that can be detangled after data acquisition using an algorithm. This increases the sensors capability for multiplexed detection. In the paper describing this design (ref. we could show that we could measure changes in temperature. We are currently working on experiments to show the sensors potential for biosensing.

By Nina Bjørk Arnfinnsdottir
Published May 26, 2021 12:12 PM - Last modified May 26, 2021 12:37 PM