Early diagnosis of infectious disease-causing agents such as viruses is essential for clinical and point of care (POC) applications [1,2,3]. Since the viruses have extremely small size and can infect all living beings, i.e., humans, animals, and plants. Therefore, their precise and accurate detection is of significant interest. Viruses usually live in host’s living cells, where they replicate. Thus, their detection is a complicated and challenging task due to the complex nature of the medium in which they exist. Since the discovery of first virus, i.e., tobacco mosaic virus (TMV) in 1892, a large number of tests and toolkits have been developed for different types of viruses [4,5]. Especially, modern tools like enzyme-linked immunosorbent assay (ELISA) and polymerized chain reaction (PCR) amplification are highly sensitive and selective protocols for virus recognition [6,7]. Disposable kits having immobilized enzymes or DNA offer simple and rapid results in minimal cost [8,9]. However, such tests are more suited for qualitative detection of viruses in remote areas.

Regardless of various established techniques for the detection and investigation of infectious viruses, the development of smart biosensors and diagnostic devices for quantitative determination of viruses is mandatory [10,11,12,13]. The major impetus in the research and development of such virus sensors is their practical features such as the ease of operation, simple and straightforward device fabrication, possibility to integrate synthetic or natural antibodies, rapid response, high selectivity and cross-sensitivity, portability, and miniaturization capability. It is due to these advantages that biosensors have not only been widely employed for virus recognition [14], but have also been applied for the detection of microorganism such as pathogenic bacteria [15,16] and yeast [17,18,19,20], living blood cells [21,22,23], and different diseases biomarkers [24,25].

A typical chemical sensor is comprised of natural and-or synthetic receptors coated on a suitable transducer such as optical, electrochemical, magnetic, gravimetric or mass-sensitive etc. [26]. Among many other factors, the selection of a transduction device mainly depends on the nature and physicochemical properties of the sensitive layer material that undergoes changes when exposed to the analyte. For instance, in case of an optical sensor, the optical properties of the selective material (or receptor layer) change due to its interactions with the analyte of interest. However, if the receptor layer is optically inactive, a labeling species or indicator is introduced that translates the receptor-analyte interactions into a recognizable optical signal [27]. Thus, the inclusion of a labeling molecule to the receptor layer introduces the desired sensing feature. Since all receptor surfaces do not necessarily possess the optical, electrochemical, or magnetic characteristics, the addition of a labeling agent causes more complexity in the receptor’s binding mechanism and may also affect its sensitivity, which are obvious disadvantages of the labeling techniques.

In this bargain, acoustic or mass-sensitive devices offer label-free detection of analytes, because mass is a fundamental property of any analyte. Thus, the acoustic wave devices are considered as universal mass-sensitive or gravimetric transducers. These transducers include: (a) the bulk acoustic wave devices such as quartz crystal microbalance (QCM) [28,29]; (b) the surface acoustic wave (SAW) resonators [30,31]; and (c) the shear transverse wave (STW) resonators [32,33], which are frequently studied in a combination with different receptor materials for molecular recognition and biomedical diagnosis. The mass-sensitive biosensors produce direct shift in frequency in the event of receptor-analyte interaction and analyte recognition. Among the aforementioned acoustic devices, QCM based chemical and biosensors have been extensively studied for the detection and quantification of a wide range of analytes from small molecules and ions to biological macromolecules and pathogenic species, e.g., viruses.

Figure 1 shows the principle of a QCM based gravimetric sensor for virus recognition. A QCM transducer, if coated with a suitably selective receptor, is capable of binding virus particles as well as virus proteins. In principle, QCM is highly sensitive to the changes in mass loading and depending upon the nature of analyte (e.g., virus particles) these little changes in mass due to selective virus binding can be detected easily. The selective receptors therefore play a major role in virus recognition and quantification due to their ability to interact and bind with viruses. Hereby, we present a comprehensive review of the gravimetric or mass-sensitive viral diagnostic devices based on QCM and a broad spectrum of synthetic and natural receptors.

The key objective of this work is to deliver an up-to-date literature review and critical analysis of different approaches employed in the design and application of QCM based biosensors and diagnostic devices for the exclusive recognition of different types of viruses. The substance is divided into two sections in which the basic QCM device design and assembly, and different types of receptor materials with special emphasis on the nature, fabrication method, and sensor performance are discussed. In the end a comparison of different receptor layers is provided along with the benefits and limitations of each receptor. This work also highlights the major achievements and the future research viewpoint.

This section presents a brief description of the principle, design, and fabrication of QCM-based gravimetric transducer. During the last few decades, QCM measuring the frequency shift and the electrochemical QCM, i.e., EQCM [34,35] simultaneously computing the electrochemical and frequency changes have been broadly designed for sensing different analytes in the gaseous and liquid phases. The pioneering work of Sauerbrey [36] on calculating the frequency changes as a function of mass adsorbed or deposited on the surface of gravimetric devices has laid the foundation for their applications in chemical and biosensors.

In a typical QCM sensor construction, a thin AT-cut quartz wafer having metal electrodes on opposite sides is connected with oscillator circuit which drives QCM to resonate at characteristic frequency. A bulk transverse wave is generated that propagates in perpendicular direction of quartz surface. The crystal surface is usually covered with a certain receptor material which binds with target analyte and thus leads to increase in the rigid mass of crystal. This results a change in fundamental resonance frequency of crystal and may be used to precisely determine the adsorbed mass on crystal surface therefore, referring it as sensor signal.

QCM, due to its extremely high sensitivity towards minor changes in mass, is considered as one of the best platforms for such applications. If combined with suitable selective material or receptor layer, QCM transforms into an exceptionally smart tiny balance capable of detecting miniscule alterations in surface mass, i.e., usually in the range of ng/cm² [37].

The sensor response of a typical QCM device, i.e., shift in frequency (Δf), is directly proportional to the square of its fundamental resonating frequency and the mass deposited [36], as given in the Equation (1) below: (1) Δ f = − f o 2 · Δ m A ρ q μ q where f_o is the fundamental resonant frequency of QCM, Δm is the change in mass, A is the piezoelectrically active area of QCM in cm², ρ_q is the density of quartz (i.e., 2.648 g/cm³), and µ_q is the shear modulus of an AT-cut quartz crystal (i.e., 2.947 × 10¹¹ g/cm·s²).

The Equation (1) is modified for measurements in liquid medium [38,39], i.e., when one face of QCM is in contact with a liquid, the liquid loading effect on frequency change can be summarized as given in Equation (2) below: (2) Δ f = − f o 3 / 2 ρ l η l π ρ q μ q where ρ_l is the density of the liquid, and η_l is the viscosity of the liquid. It is important to mention here whenever QCM devices are studied in liquid phase, the acoustic properties of liquid medium have to be taken into account for sensor measurements.

Nonetheless, these equations show that the sensitivity of the QCM device can be improved by increasing its primary frequency, i.e., if the fundamental frequency of a QCM is doubled, its response would be increased by a factor of four. However, this can only be achieved by reducing the thickness of QCM wafer that will result in mechanically fragile devices. Thus, this parameter practically limits the fabrication of devices beyond certain frequency (f_o). The other way to increase the sensitivity of QCM devices is to work with overtones, which require special oscillator circuit having band pass filter for attenuation of harmonic resonances. However, this would also lead to an increase in noise level and ultimately making small improvement in signal to noise ratio. Therefore, an increase in the fundamental resonance frequency of crystal is the more appropriate method of tuning sensitivity.

Typically, QCM having frequency in the range of 5–20 MHz are used in liquid phase measurements. Although, a few reports suggest the use of QCM with frequency as high as 110 MHz for liquid phase operation [40]. This is remarkably high frequency and the authors report that substantially high detection limit is achieved in the detection of M13-phages in liquid medium, i.e., the sensitivity improved by a factor of 200 while using 56 MHz QCM as compared to 19 MHz QCM device [40].

Another important aspect in determining the sensitivity of a QCM device is the nature and design of electrodes on QCM wafers [41,42]. If the diameter of the electrode is bigger, the available surface for selective receptor layer integration would be large, which ultimately leads to higher sensor response. For electrode fabrication on QCM wafer, gold or any other noble metal paste can be applied via screen printing. Many of the commercially available QCM have fused metal electrodes that afford better support on QCM surface. Furthermore, the integration of receptor layer with QCM electrode should be firm and stable, i.e., the receptor material should not be deteriorated during the fabrication and testing processes.

Figure 2 shows the design of a single, dual, and tetra-electrode QCM wafer for detecting viruses. A multichannel QCM having more than one electrodes for sensing different analytes is an innovative design [43]. Albeit this unique tetra-electrode, QCM design has not been implemented for simultaneous recognition of different viruses or virus sub-types, but it could be adopted as a single transducer platform for different viruses. However, it is important to mention that there should be a minimal distance between the multiple electrodes to prevent the cross-talk between electrical signals and thereby reducing signal-to-noise ratio for improved sensitivity. Furthermore, in liquid phase measurements, one side of the QCM remains in air to avoid excessive damping problems.

These are some general but imperative design and fabrication parameters that regulate the ultimate QCM device sensitivity. The following study emphasizes a variety of synthetic receptors based on imprinted polymers, natural antibodies, DNA and aptamers for the detection of different viruses in combination with QCM devices.

A number of academic research articles and review papers have been published previously that emphasize the QCM’s working principle and study their performance in a wide range of complex biochemical processes taking place at the receptor-analyte interface [44,45,46,47]. These investigations highlight the surface chemistry and mechanistic events during the attachment or binding of various target analytes. This work is primarily focused on different types of synthetic and natural receptors that directly influence QCM device sensitivity and selectivity for improved recognition of viruses. In case of virus sensing, the recognition characteristics are governed by nature of receptor/selective layer. In the following sections, we shall describe the virus recognition principle, mechanism, and performance including sensitivity and selectivity of various receptors. The potential receptors for selective sensing of different viruses are divided into four classes based on their nature and virus-binding mechanism: (a) Synthetic antibodies based on molecularly imprinted polymers; (b) natural antibodies; (c) DNA; (d) aptamers; and (e) other biomacromolecular receptors such as proteins.

This article presents a review of the gravimetric viral diagnostic systems consisting of a selective layer or receptor that captures viruses and a QCM transducer that translates viral binding events into legible sensor signals. This work provides a comparative study of the assembly and performance of different types of receptors such as synthetic antibodies, natural antibodies, DNA probes, and aptamers. In the past 15–20 years, a number of reports have been published on combining these receptors with QCM to develop rapid, low-cost, reliable, sensitive, and specific biosensors for label-free recognition of viruses, or viral DNA and surface proteins. In summary, we report the competing advantages and drawbacks of various receptors based on their nature, immobilization and fabrication strategies, cost-effectiveness, stability, and sensing performance.

Firstly, gravimetric or mass-sensitive biosensors based on QCM are one of the very few devices that can virtually detect and report anything that has mass ranging from small molecules to microorganisms. A QCM is particularly useful for sensing viruses, which are neither fluorescent nor electrically conducting and thus are difficult to detect by optical or electrochemical methods. QCM is also highly sensitive to very small changes in surface mass, i.e., in the range of ng/cm² [37]. Moreover, QCM can be combined with synthetic or naturally occurring receptors easily for selective and specific recognition of any target, e.g., viruses in this case. Table 4 provides a qualitative comparison of the most commonly used virus recognition technologies with QCM based gravimetric viral diagnostics.

Apart from many advantages of QCM devices as a general gravimetric transducer, it suffers from certain inherent problems. For example, in order to enhance sensitivity, fundamental frequency of QCM has to be increased and correspondingly the thickness of quartz sheet has to be reduced. Decreasing thickness of quartz sheet would make QCM device more fragile and mechanically unstable thus, limiting its practical use when extremely high sensitivity is desired in chemical sensing. The viscosity of the in-contact medium also has strong influence on QCM damping as highly viscous analyte solution would lead to increase frequency fluctuations that make measurements difficult at trace levels. Furthermore, the integration of QCM with micro/nano electromechanical systems is still in its infancy due to material constraints and other issues. In future, some of these problems need to be addressed to achieve the true potential of QCM-based gravimetric virus sensors.

Secondly, positive and negative aspects of different receptors are briefly discussed. Synthetic antibodies based on MIPs can be fabricated easily using different approaches such as soft-lithography and epitope imprinting. The primary advantage of soft-lithographic surface imprinting procedure is to transmit complete geometrical and surface chemical characteristics into MIPs that leads to high specificity. On the other hand, typical challenge in this technique is to produce a good number of recognition centers for viruses that can lead to readable sensor signal. Another limitation of soft-lithography is the difficulty to imprint smaller biomacromolecules such as viral DNA or surface proteins, which can undergo conformational changes.

Conversely, the epitope approach can be used to imprint epitopes of viral surface proteins to detect different viruses. The difficulty in epitope imprinting however originates from the choice of specific fragments that can be imprinted and are capable of selectively capturing the target [150]. In addition, certain reports suggest that whole-protein imprinting is more efficient as compared to epitope imprinting [151]. Synthetic antibody replica designed via dual imprinting process present an exceptional case of selectivity and sensitivity for viruses and small molecules, but they have not gained popular backing for their application to other viruses or bioanalytes. Thus, it is difficult to conclude about its future based on few examples.

Natural antibodies (NAb) are proteins with inherent sensitivity and specific binding capacity towards antigens, i.e., viruses. They are extremely selective and possess high binding affinity for target viruses. However, they have limited life span and cannot be used in harsh environments fearing their decomposition. For instance, the fragile structure of NAb limits their applications in real-life samples such as serum due to the presence of enzymes that may damage their structure [152]. Furthermore, it is believed that their protein character may lead to their deactivation during repeated binding-and-regeneration cycles [153,154].

When compared to fully synthetic antibodies (MIPs), DNA and aptamers are also less stable. DNA and aptamers are biomolecules, which consist of nucleic acids. They can be degraded by microorganisms. Also, the degradation by RNases and DNases can occur. RNases and DNases are problematic since they are an important part of the immune systems’ defense against viruses in numerous organisms and thus are almost omnipresent. As a result, special care has to be taken to make them RNase/DNase free. Currently there are two strategies which have been developed to circumvent this problem. The first approach is to synthesize a “mirror” analog of the aptamer that retains the original properties, but is not cleaved by nucleases [155]. The second approach is to include local modifications of the ribose 2’ sites in the aptamer chain which prevents cleavage by the enzymes [156,157].

3D structure of aptamers for instance in comparison with other NAb depends more on the surrounding medium. As a consequence, the sensing parameters and the quality of the sensors varies more when different solutions are used for sensing. Additionally, the high affinity of the NAb, DNA, and aptamers with their target can be challenging. Although high binding affinities are usually desired, it may lead to irreversible sensor signal preventing the receptor layer to be regenerated. Generally, longer regeneration times or harsher conditions are needed in that case than e.g., for MIP, which usually have a lower affinity for their target. Specifically, the aptamers are composed only of 4 different nucleic acids, while proteins have 20 amino acid building blocks and thousands of monomers are available commercially [158]. As a result there is less chemical flexibility, when using aptamers [159].

Finally, we believe that MIPs, NAb, DNA as well as aptamers are of great value for viral recognition systems due to their excellent sensitivity and selectivity towards a variety of disease-causing viruses. However, the NAb, DNA, and aptamers are worthy of further investigations to enhance their long-term stability, to develop new immobilization and fabrication technologies with the objective of lowering cost and effort, to optimize their concentration, orientation, binding efficiency for reversible binding, and to advance their easy regeneration preventing any functionality damages and-or degradation.

Synthetic antibodies based on MIPs, on the other hand, fairly deserve a more concentrated research effort to make them a viable alternative of natural receptors, and more so to commercialize MIP-QCM based gravimetric viral diagnostic systems because the MIPs are relatively more stable, selective and show comparable sensitivity, and are easy to fabricate and regenerate thus lowering their overall cost and analysis time. In future, these approaches may lead to commercial availability of cheaper and more reliable viral diagnostics.