Ethanol is a colourless organic chemical, which is often referred to as alcohol or ethyl alcohol. Its attractive solvent properties of being easily soluble in water and other organic compounds means that it is one of the key chemicals used in modern industrial processes and consumer products. Its uses throughout society are manifold, including as a preservative, an anti-bacterial agent, an astringent in personal care products, as an antidote, an anti-infective and rubbing alcohol in medicines, as a solvent in paints, lacquers and varnishes, as well as an ingredient in intoxicating alcohol beverages and as an additive for flavouring and preserving food [1]. Since the 1970s, interest in the use of ethanol as a renewable fuel or partial substitute for gasoline has grown significantly. It is considered widely as a renewable alternative for fossil-based chemicals such as bioplastics and an additive for ethanol fuel blends [2,3]. In addition, the traditional drinks industry produces ethanol from yeast fermentation in brewing and wine manufacture and the distillation of spirits leads to a huge range of ethanol fortified alcoholic beverages [4].

The proliferation of ethanol uses as described above also increases the demand for its measurement in other sectors such as environmental health and safety, emission control, new biofuel production processes, in pharmaceutical and industrial product development, in breath analysis and in food quality assessment [5]. Analysis of ethanol in many processes is necessarily mandated by the relevant regulating agencies, e.g., the food and drugs administration (FDA) in the USA. A variety of methods can be used to measure the concentration of ethanol in aqueous solution. Commonly used detection methods include enzymatic measurement [6], Raman spectroscopy [7], UV/NIR spectroscopy [8], dichromatic oxidation spectrophotometry [9], refractive index (RI) analysis [8], gas chromatography (GC) [10], high-performance liquid chromatography (HPLC) [11], pycnometry [12], densimetry [12], hydrometry [12], capillary electrophoresis [13] and colorimetric methods [14]. However, the use of these methods has many disadvantages including low reproducibility, potential sample loss, long analysis time, complex offline sample preparation procedures and bulky and expensive instrumentation. Dichromatic oxidation spectrophotometry, pycnometry, hydrometry and densimetry suffer from large sample loss and require a moderate to long time for analysis [9,12,15,16]. The accuracy of the measurement is also highly dependent on the operator’s knowledge and the sample’s temperature [17]. Enzymatic methods are characteristically known for their low accuracy, reproducibility and enzyme stability. Modular Raman spectrometry necessitates precautionary measures for laser use and yield a detection limit of only 1% (v/v) ethanol [18,19]. Complicated calibration procedures are needed for near-infrared spectroscopy and, hence, it can be expensive and time consuming [20]. On the other hand, RI analysis is a relatively simple method, but the accuracy is highly dependent on temperature and can only be used for simple solvents (as opposed to complex mixtures) [21]. When compared to standard chromatography techniques such as GC and HPLC, capillary electrophoresis has lower accuracy. GC is currently considered the most reliable method for ethanol concentration measurement in clinical samples and for alcoholic drinks and is, therefore, the most widely used. Despite the benefits of chromatography techniques, they can be relatively slow (often requiring pre-concentration), complex and costly due to the large quantities of expensive organics needed [22]. Ethanol detection using stimuli-responsive hydrogels and piezoresistive pressure sensors has also been recently reported, where the ethanol concentration of a vodka product (“Wodka Gorbatschow”) with a specified value of 37.5 vol% ethanol was measured [23]. Many industries and research sectors, e.g., alcoholic beverage production and clinical/medical applications, require a simpler, more convenient and higher throughput determination of ethanol concentration [9]. The advantages and disadvantages of the commonly used ethanol detection methods are summarised in Table 1.

Recent advances in industrial practice including the emergence of Industry 4.0 for process automation and manufacture has meant that real-time monitoring is a crucial part of those processes. Big data is a major element of the Industrial Internet of Things (IIoT), which results from the generation and collection of massive data sets from sensors used in diagnosis, process monitoring, product manufacturing, health, safety, and quality control. Bulk optical sensing technology has the unique advantage of being immune to external electromagnetic interference and is potentially highly accurate, specifically in resolving very small changes in RI compared to existing commercially available technologies, but typically requires delicate alignment and coupling mechanisms, which increase sensor size, complexity and reduce stability, e.g., through susceptibility to mechanical vibration. On the other hand, optical sensors based on optical fibres are becoming a highly versatile, rugged and potentially cost-effective alternative due to their capability for being miniaturised, readily integrated with electro-optical components or electronic systems, feasible for real-time and remote sensing, and light weight, as well as having a minimised need for precise alignment and coupling [30,31].

Optical fibres have been investigated for their potential use in sensing applications since the early 1980s, and several advances have been made in the fields of optical fibre chemical sensing and biosensing [32]. At that time, researchers also started exploring the concept of optical fibre-based ethanol sensors [33,34]. Wolfbies et al. (1988) used enzymatic oxidation of ethanol to create an optical fibre ethanol biosensor [34]. The sensor layer included an oxygen-sensitive fluorescing indicator that detected a drop in local oxygen partial pressure due to enzymatic oxidation. The sensor detected ethanol concentrations in the range of 50 to 500 mmol L⁻¹, with an accuracy of ±4 mmol L⁻¹ at 100 mmol L⁻¹. Since then, a diverse range of mechanisms have been investigated for measuring ethanol concentration using optical fibre sensors and, in some cases, real applications have been further explored using these sensing schemes. In general, these sensors have delivered encouraging results and demonstrated great potential for ethanol sensing in a wide range of applications. Figure 1 is a graphical summary of a selection of applications for ethanol sensing and optical fibre sensing schemes utilised for ethanol sensing extracted from recent literature [35,36,37,38,39,40,41,42,43] and includes sensor design parameters and typical output responses. Several articles have recently outlined and reflected on the advancement of optical fibre chemical and biosensors from various perspectives [31,32,44,45,46,47,48]. There are abundant possibilities and potential for fabricating highly effective optical fibre ethanol sensors in view of the rapid development of this technology and related functional materials [31]. This article focuses on a review of recent optical fibre sensing developments for ethanol measurement in aqueous solutions, including a perspective on their design and response characteristics. A comparison of their respective performance is provided, which points to their applicability for use in current and future full-scale industrial measurement systems. The continuously developing needs of real time measurement and the emergence of the internet of things (IoT) sets the background to this article, which is intended to inspire and focus further research and wider utilization of these sensors in the commercially significant industrial and medical sectors.