Open Access
Add to BibSonomy
Abstract
Large amount of wastewater is produced by washing machines and dishwashers, which are used in a daily basis. This domestic wastewater generated in households or office buildings (also called greywater) is drained directly to the drainpipes without differentiation from that with fecal contamination from toilets. Detergents are arguably the pollutants most frequently found in greywater from home appliances. Their concentrations vary in the successive stages in a wash cycle, which could be taken into account in a rational design of home appliances wastewater management. Analytical chemistry procedures are commonly used to determine the pollutant content in wastewater. They require collecting samples and their transport to properly equipped laboratories, which hampers real time wastewater management. In this paper, optofluidic devices based on planar Fabry-Perot microresonators operating in transmission mode in the visible and near infrared spectral ranges have been studied to determine the concentration of five brands of soap dissolved in water. It is found that the spectral positions of the optical resonances redshift when the soap concentration increases in the corresponding solutions. Experimental calibration curves of the optofluidic device were used to determine the soap concentration of wastewater from the successive stages of a washing machine wash cycle either loaded with garments or unloaded. Interestingly, the analysis of the optical sensor indicated that the greywater from the last water discharge of the wash cycle could be reused for gardening or agriculture. The integration of this kind of microfluidic devices into the home appliances design could lead to reduce our hydric environmental impact.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
The optimal management of water resources around the world, especially in highly populated cities or during drought periods, is a major concern of current societies. This challenging issue is getting increasingly important linked with global warming effects. Consequently, policy makers face sustainable environmental policies and new technical solutions related to the rational use of water are demanded [1–3].
Average daily use per person of domestic fresh water varies in different countries [4]. For example, in North America is about 200 l, in Spain 130 l, in several countries like Ethiopia, Niger or Mozambique well below 30 l. About 10-15% is used in home appliances such as washing machines or dishwashers [5].
Water saving can be achieved reusing domestic wastewater from households (apart from sewage), also known as greywater. Normally, greywater is produced in sinks, showers, and home appliances such as washing machines or dishwashers. After adequate treatment, greywater can be safely reused with a positive impact on the reduction of the demand on public water supply. One of the ways to decrease freshwater demands in developed or urban environments is through the use of greywater systems for toilet flushing and irrigation. In both cases, greywater may require automatic analysis and eventually mild treatment. The actual treatment of greywater depends on the contamination level of the wastewater. Currently there exist precise chemical analytical methods to determine the concentration of pollutants in water. Typical analytical measurements aim to determine the pH, conductivity, and surfactant, phosphates and nitrogen content (which often come from kitchen and laundry activities). Nevertheless, they require sampling and often costly analysis in properly equipped analytical laboratories [6–8].
A typical detergent is composed of a mixture of different types of surfactants to achieve the desired cleaning and foaming properties. In practice, detergents are the most abundant chemicals in household greywater. They are regularly delivered to the drainpipes in showers, sinks, washing machines, dishwashers, and similar appliances.
In this paper, we determine the content of detergent in water using optical monitoring by means of optofluidic microsystems instead of other more traditional methods based on analytical chemistry. Indeed, a simple Fabry-Perot microcavity (FPM) is used to determine detergent concentration in soapy water. We show that the optical response of these optofluidic devices depends on the specific soap and its concentration. Thus, the optical responses to several commercial detergents are determined. Additionally, the water discharged at different stages of a wash cycle from a commercial automatic washing machine is analyzed. Interestingly, the optical response to each of the water discharges produced during a laundry cycle is different. In light of these results, the implementation of these devices in household water quality control systems is discussed.
2. Materials and methods
2.1 Nominal ethanol-water mixtures and soapy water samples
Nominal ethanol-water mixtures used as references as well as dilutions of several detergents in tap water were studied in this investigation. The ethanol (EtOH) and water mixtures’ EtOH concentration ranged from 0 to 111 mg/ml, with steps of around 10 mg/ml.
The detergents considered were Sodium Lauryl Sulphate (SLS), an anionic surfactant used in many cleaning and hygiene products, and four commercial soap brands (easily available in most supermarkets in Spain), namely Fairy (CDi), Bosque Verde (CDii), Marsella (CDiii), and WIPP (CDiv) diluted in water, with concentrations ranging from 0 to about 58 mg/ml, with increments of around 7.2 mg/ml. All dilutions were prepared by building a stock of the highest studied concentration and then diluting this stock in distilled water.
Soapy laundry water came from the discharges of a Candy CS 13102D3/1-S automatic washing machine (program "Daily 59’-20$^{\circ }$C"). In this program, the wash cycle is performed at room temperature, that is, there is no heating of the inlet water. The outlet of the appliance was redirected to a container where the amount of water discharged was collected. A sample of each discharge produced during the laundry cycle was recovered. Two laundry cycles were recorded: one with the washing machine unloaded, and another with it fully loaded (8 kg) with clothing.
2.2 Optofluidic FPM fabrication
Two optofluidic FPM were used in this work. They were designed to act as multimode interference filters that modulate the transmitted signal at certain spectral ranges. The first one showed the interference pattern in the visible spectral range from about 500 to 580 nm (FPM@VIS), and the second one in the NIR range from about 800 to 880 nm (FPM@NIR).
The optofluidic microcavities consisted of a sealed volume delimited by two parallel fused silica plates coated with dichroic Bragg mirrors at the inner side of the plates and separated by approximately 170 $\mathrm {\mu }$m. The Bragg mirrors were fabricated by stacking a series of high refractive index H (Nb$_{2}$O$_{5}$, refractive index at 550nm of 2.29), and low refractive index L (SiO$_{2}$, refractive index at 550nm of 1.46) thin films with the layered structure given in Table 1, where the number preceding the L or H labels indicate the physical thickness (in nm) of the corresponding layer. The reflectance of these dielectric mirrors is shown in Fig. 1. Further details on the manufacturing of the multilayer system can be found in [9,10]. Similar FPM configurations have been reported for other optical sensing applications [11].
Fig. 1. Reflectance of the Bragg mirrors considered in the optofluidic microcavities. The grey vertical lines show the spectral ranges used for analysis.
Table 1. Layered structure of Bragg mirrors in the optofluidic microcavities.
The greywater samples were circulated in these cavities through tubing fittings located in one of the plates. The sealing material was PDMS.
2.3 Experimental setup
The optical response of the optofluidic microsystem is interrogated with a small 3 W halogen lamp powered by a 5 V voltage source. A pinhole is placed in front of the lamp and the light is collected by a convergence lens located at the focal distance so that a collimated beam is obtained. A mirror is used to redirect the beam at the normal direction of the optofluidic microcavity. An additional pinhole is used to provide an incident beam with a diameter of about 0.2 mm. An optical bunch fiber with a core diameter around 200 $\mathrm {\mu }$m is used to collect the transmitted light at the normal direction. The optical fiber is launched to a spectrometer (Andor SR-500i-B2) coupled to a CCD (Andor iDus 416) detector, which are controlled by a personal computer, see Fig. 2.
Fig. 2. Scheme of the experimental setup. L, lense; P, pinhole; M, mirror; I, inlet tube; O, outlet tube, OF, optical fiber; SP, spectrometer; CCD, charged coupled device; PC, personal computer.
2.4 Standarized analysis of soapy samples
The following physical and chemical parameters were analyzed on the greywater samples collected from the washing machine water discharges at several stages of a wash cycle: conductivity, determined by measurements based on UNE-EN 27888 norm; total molecular detergent concentration, determined by UV-VIS spectrometry based on the UNE-EN ISO 16265:2012 and SM 5540 D norms by an ENAC accredited laboratory; and total phosphorous and total oxidized nitrogen concentrations, determined by spectrophotometry assays based on UNE-EN ISO 6878 and SM 4500–NO-3 E norms.
3. Results and discussion
3.1 Calibration of the optofluidic FPM
The emission spectrum of the halogen lamp, used as the illumination source, shows the characteristic broad emission band ranging from the visible to the NIR spectral regions. FPMs have been designed to act as multimode interference filters that modulate the transmitted signal at certain spectral ranges. In fact, two different FPM have been tested. One showed the interference pattern in the visible spectral range (from about 500 to 580 nm), and the second one in the near infrared (NIR) range (from about 800 to 880 nm). The transmitted signal of the microcavities shows maxima at their optical resonances. The spectral wavelengths at which they occur are given by [12]:
where $m$ is the mode number, $\lambda _m$ is the resonance wavelength, and $n$ and $d$ the refractive index and thickness of the cavity.The transmission spectra of the FPM@VIS (spectral range from 520 to 523 nm) and FPM@NIR (830 to 838 nm) filled with water are given in Fig. 3(a) and 3(b). The optical resonances correspond to the maxima in the transmitted signal. Simulated FPM modes have been calculated using Eq. (1), when $n$ refers to the refractive index of water at that wavelength (1.3341 and 1.3290 for the visible and NIR ranges, respectively), and a Gaussian spectral shape of the modes. The best fitting to the experimental data was obtained for a $d$ parameter of 170 $\mathrm {\mu }$m for both cavities, and $m$ varied from 872 to 868 and from 540 to 536 in the FPM@VIS and FPM@NIR, respectively.
Fig. 3. Detail of the transmission spectra (solid black line) and simulated spectra (solid red line) of (a) FPM@VIS and (b) FPM@NIR filled with water.
If the refractive index of the cavity medium changes, the positions of the FPM modes are expected to change too [9]. In order to test the ability of our optical system to detect small changes in refractive index of the liquid circulating through the microcavities, a set of water-ethanol mixtures was prepared. The refractive index of the mixtures depend on the ethanol concentration [13]. According to Eq. (1), the spectral shift of a FPM mode, $\Delta \lambda _m$ due to a refractive index change, $\Delta n$, is given by
The dependence of the spectral shift on the wavelength suggests that higher spectral shifts could be expected for FPM modes at longer wavelengths, for a given $\Delta n$.
The refractive index sensitivity of the FPMs gives the change in the wavelength of the resonant modes over refractive index unit (RIU) change. Refractive index sensitivities of about 390 nm/RIU and 630 nm/RIU were obtained from Eq. (2) for FPM@VIS and FPM@NIR, respectively. These values are relatively close to that reported for optofluidic ring resonators, around 570 nm/RIU [14]. Nevertheless, a large range of refractive index sensitivities have been reported in the literature depending on the actual resonator, as, for example, 3.26 nm/RIU for a high-Q crystalline MgF$_{2}$ whispering gallery mode small resonator [15]. Values of 130 and 102 nm/RIU have been observed in self assembled Rhodamine B doped PMMA hemispherical microlasers and in polymer FP microresonators, respectively [16,17]. A refractive index sensitivity of 1368 nm/RIU was reported at the wavelength of 1600 nm for an open cavity FP fiber resonator [18].
Both FPM@VIS and FPM@NIR have been tested using the same water-ethanol mixtures. The spectral position of the FPM modes at 520.3 nm ($m$ = 872) and at 830.9 nm ($m$ = 546) of the visible and NIR microcavities, respectively, have been measured as a function of the ethanol concentration in the mixture at room temperature, see Fig. 4. The peak positions were determined using a simple peak-finding algorithm, based on a Gaussian fitting procedure to the transmission modes around the selected spectral range. It is worth mentioning that the determination of the peak positions by this procedure is rather independent on small changes in the intensity of the transmission modes. Taking into account the difference in refractive index between water-ethanol mixtures and pure water $\Delta n$ ($\Delta n$ = refractive index of the ethanol-water mixture – refractive index of water) reported in the literature [13], and Eq. (2), the corresponding $\Delta \lambda _m$ values have also been calculated and they are given in Fig. 4 (dotted black line), where a good agreement with the experimental spectral shifts can be found.
Fig. 4. Experimental (symbols) and simulated (dashed lines) spectral shifts $\Delta \lambda _m$ of FPM@Vis (circles) and FPM@NIR (squares) modes as a function of the ethanol concentration. Linear least squares fittings of the experimental data are included as red lines.
The absolute sensitivity of the microcavities for the detection of ethanol concentration in water, S, is given by the slope of the linear regression of the experimental data in Fig. 4, described by
C denoting concentration of the solution. As expected, the absolute sensitivity increases at longer wavelengths. The sensitivity of the visible FPM@VIS was 0.025 nm/(mg/ml), while it increases to 0.042 nm/(mg/ml) for the FPM@NIR. That is an increase of 68%, in agreement with Eq. (2). Consequently, the rest of the studies, devoted to the monitorization of detergents in greywater, were conducted with the FPM@NIR sensor due to its better performance.
Sodium lauryl sulphate (SLS) is a surfactant, frequently included in many detergents and household cleaning products. The concentration of SLS in cleaning products depends on the product and manufacturer, but typically ranges from 1% to 30% [19]. Therefore, we have measured the spectral shift of the FPM@NIR modes as a function of the SLS concentration in water, see Fig. 5(a) and 5(b). A linear dependence is observed, yielding an absolute sensitivity of 0.07 nm/(mg/ml).
Fig. 5. (a) Detail of transmission spectra of SLS at different concentrations. They have been vertically shifted for clarity. Tilted arrows were added to show the mode shifts. Spectral shifts of the resonance modes as a function of the soap concentration for (b) SLS; and for (c) CDi, CDii, CDiii and CDiv brands. Dashed lines are linear fits to the experimental data.
Additionally, four commercially available detergents were investigated. Two of them are cleaning products for dishwashing (CDi and CDii) and two for laundry (CDiii and CDiv). The results are given in Fig. 5(c). In all the cases, a red shift of the FPM modes was detected as the soap concentration in water increases. However, different slopes have been found for the different brands. Thus, the absolute sensitivities S obtained are 0.022, 0.023, 0.014, and 0.025 nm/(mg/ml) for CDi, CDii, CDiii, and CDiv, respectively (see Table 2).
Table 2. Absolute sensitivities and LODs of the FPM@NIR for SLS and the commercial detergents.
Another important parameter to characterize the performance of a sensor is the limit of detection (LOD), defined as the smallest concentration of soap, $C_{min}$, which can be detected by the sensor. It is inversely proportional to the absolute sensitivity of the device:
where $\Delta \lambda _{min}$ corresponds to the smallest change of the spectral position of the FPM mode that can be experimentally detected. In order to determine the accuracy in the determination of the spectral position of a FPM mode, 100 repeatability measurements of five consecutive FP modes were conducted. The standard deviation of the spectral position distributions provided an accuracy of 0.025 nm. Taking into account this value, the LOD for SLS, CDi, CDii, CDiii, and CDiv were 0.36, 1.14, 1.09, 1.79, and 1.09 mg/mL, respectively (see Table 2).3.2 Application to laundry greywater
It has been proposed that greywater coming from laundry might be reused with little to no treatment for gardening, lawn keeping and agriculture [20–22]. However, greywater with a high pollutant concentration such as surfactants damage the soil and hinder the growth of some crops [23,24]. For example, it has been found that watering lettuce and okra with a 5.0 mg/ml detergent-water dilution results in the eventual death of the plants. On the other hand, when watered with lower detergent concentration (0.1-1.0 mg/ml) the plants grew without significant issues [25]. The usability of laundry greywater from a washing machine will depend on the soap concentration, which varies among the water discharges during a wash cycle. Thus, it is interesting to investigate the response of the device to soapy samples corresponding to the successive rinse discharges from a laundry cycle.
A typical laundry cycle may have one or more rinse and discharge steps. One would expect a reduction of detergent and dirt with each discharge, since the greywater would be becoming progressively cleaner with each rinse.
Soapy water samples from an automatic washing machine appliance during a laundry cycle were collected. CDiv soap brand was used (100 g). The wash cycle had three water discharges: a first one from the initial wash (1st discharge), a second one from a first rinse (2nd discharge), and a third one from the last rinse and spin cycle (3rd discharge). Soapy samples from an empty (unloaded) and a fully filled with 8 kg of garments (loaded) washing machine were analyzed. They were tested in the FPM@NIR optical transducer, and their optical response compared to the same tap water used by the washing machine.
The spectral shifts of the FPM modes with respect to tap water corresponding to the unloaded and loaded situations for the 3 discharges during the wash cycle described above are displayed in Fig. 6. Error bars represent the standard deviation of the FPM spectral position determination obtained through the peak fitting procedure. As a general behavior, a redshift of the modes’ positions in soapy water with respect to clean tap water is found. Besides, note that the redshift decreases with the successive discharges. A shift of 0.34 nm was detected in the first discharge of the fully loaded washing machine, and then it dropped to 0.12 and 0.04 nm in the second and third discharges, respectively. In the case of the unloaded washing machine, the spectral shift of the first discharge was 0.27 nm and it drastically fell to 0.05 and 0.00 for the following discharges.
Fig. 6. Spectral shifts of the FPM-NIR modes corresponding to the three water discharges from a loaded and unloaded wash cycle in an automatic washing machine.
In the case of the unloaded wash cycle, the soap concentration in the water discharges can be directly calculated from the calibration curve, see Fig. 5(c) and Eq. (3), considering the specific sensitivity parameter of the CDiv soap (Table 2). The results are given in Table 3. In the unloaded wash cycle, there are no clothes to retain the soap. Therefore, the first discharge drags most of it. In the second discharge the soap concentration drops one order of magnitude, and during the spin cycle the greywater piped out of the washing machine is indistinguishable from tap water. We wish to remark that fluctuations in parameters such as lamp intensity, turbidity of the water sample, scattering due to dirt or non-dissolved particles may affect the intensity of the transmission modes. However, in our case, the intensity change was less than a factor of two in the discharge water compared to the reference tap water samples. Moreover, the standard deviation of the experimental determination of the spectral distribution of the FPMs was the same for both and, consequently, they did not affect the uncertainty of the spectral shift measurements.
Table 3. Detergent concentration of the water discharged at the different stages for the unloaded and loaded wash cycle.
The loaded wash cycle is different and requires a more detailed analysis. The spectral shifts of the greywater discharges are larger than in the unloaded case. This is due to the presence of impurities from the clothes being washed. Diluted salts, oils, or other organic compounds may affect the refractive index of the discharged streams and contribute to the spectral shifts of the FPM modes. Therefore, it seems reasonable to assume that the spectral shifts measured in the greywater discharges from the loaded wash cycle had two contributions, one from the detergent supplied to the wash cycle and another from substances from the clothes.
To gain more insight into the detergent content in the greywater discharged from the successive stages of the wash cycle, we have performed independent chemical analysis of several parameters that might be interesting to be followed in a real case of wastewater analysis. Table 4 shows standardized analysis of total surfactant concentration (anionic, cationic and non-ionic species), conductivity, total phosphorous, and total oxidized nitrogen of samples from the first, second and third water discharges for the loaded and unloaded washing machine cases described above, as well as for tap water, used as a reference.
Table 4. Physico-chemical parameters obtained from standardized analytical methods.
The determination of the surfactant content in the soapy samples (see Table 4) is a very interesting parameter for our analysis. The actual formulations of commercial detergents include surfactants, among other compounds. The surfactant concentration will be proportional to the detergent’s concentration in the greywater. Indeed, the rate of decrease of the total surfactant concentrations in the unloaded wash cycle agrees with the calculated detergent concentrations, within the accuracy provided by the measurements.
Our rationale to quantify the actual detergent concentration in the discharged greywater of the loaded wash cycle was as follows. The first water discharge carries the largest amount of detergent (together with other impurities from the clothes), the next discharge occurs after the second wash and rinse, and the final one subsequent to the last rinse and spin step. Therefore, we assume that the spectral red shift of the third greywater discharge (0.03 nm) is basically due to detergent, which was still absorbed by the clothes and, consequently its detergent content can be directly estimated from Eq. (3) and the absolute sensitivity of FPM@NIR for CDiv (Table 2), providing a value of 1.1 ${\pm }$ 0.4 mg/ml. As the detergent concentration in the greywater must be proportional to the analysed surfactant content, the detergent concentration of the other two water discharges were calculated (see Table 3). The greywater from the first discharge has similar soap concentration to the one in the unloaded cycle, but additional impurities, which explains the higher spectral shift of the loaded laundry versus the unloaded. The second and third discharges of the loaded cycle carry a similar amount of detergent, but in the second discharge there is still a significant amount of impurities from the clothes to account for the observed FPM mode spectral shift.
Concerning the rest of the parameters of Table 4, conductivity measurements can be easily correlated to salts that are part of the formulation of the soaps dissolved in water. Furthermore, total oxidized nitrogen and phosphorous species, as well as total detergent, are signature of contaminants in water. These parameters have been chosen to illustrate the correlation between the automatic analysis performed with the Fabry-Perot microcavity and standard parameters considered in wastewater analysis.
The conductivity of surfactants depends on the nature of ions formed after ionization, the nature of the solvent used, the temperature, and the presence of additives like salts. Generally, a decrease in specific conductivity with dilution is due to the replacement of ions by colloidal particles, which, although conducting, have a lower equivalent conductivity than the ions from which they are formed [26]. Indeed, the conductivity decreases with the number of water discharges for both loaded and unloaded laundry samples. Moreover, the values are higher in the loaded laundry sample as expected because of the contribution of additional organic and inorganic compounds which may be present in the garments. Indeed, the conductivity observed for the third water discharge of the unloaded laundry is similar to that of tap water. These results show the same tendency observed for the spectral shifts of the FPM modes (see Fig. 6).
Regarding the chemical analytical assays, the total phosphorous concentration is below the detection limit in tap water, and it is only detected in the first water discharge, being higher in the loaded laundry because of additional contributions of the garments, as expected. The total oxidized nitrogen concentration decreases with the number of water discharges and is similar to that of tap water for the third discharge of the unloaded laundry, which is in agreement with the trend of Fig. 6. In summary, the chemical analytical measurements show the same trends and their interpretation agree with the one applied to the optical measurements in the FPM.
According to previous reports on the negative impact of soaps in grey water for irrigation [23,25], in our case, the 1st discharge of the loaded cycle should not be used for gardening or agriculture without further treatment. The 2nd discharge has notably lower soap concentration. Nevertheless, it should be noted that additional impurities are present in the discharged water from the loaded wash cycle. Only the 3rd water discharge could be reasonably reused for those purposes, taking into account that the obtained detergent concentration is close to the 1 mg/ml value reported for safe plant irrigation and well below the established harmful level of 5 mg/ml [25]. However, in practice, the actual soap concentration values depend on the quantity of soap and clothes introduced by the user, and on the amount of water used by the washing machine. This, in turn, varies among washing machine models and with the mass of the laundry. Therefore, a FPM integrated in the washing machine would be able to assess the actual degree of soap concentration of discharged water, as shown in this study. This information could automatically be used to switch the discharge system to allow recycling the greywater discharge for irrigation. However, the detergent concentrations obtained with our device are close to the LOD. Therefore, to improve the robustness and reliability of such a discharge system, it would be desirable to lower the LOD and to improve the accuracy in the determination of the safe detergent concentration threshold for plant irrigation. Further research is planned to reduce the experimental uncertainty and to shift the FPM resonances to longer wavelengths, both of which would have an impact in the reduction of the LOD.
We are aware that the proposed method is based on changes in the refractive index of the streams flowing through the microcavity. The method lacks selectivity against any particular compound. However, for those cases in which detergents are expected to be the main pollutant of greywater, FPM@NIR transducers may be used to control the water quality. In practice, this kind of devices could be coupled to dishwashers, washing machines or even certain sinks. Given the absolute sensitivity of the device to a specified cleaning product, the spectral shift of the FP modes could be used to determine the pollutant soap concentration in water and, consequently, determine the adequate treatment or the feasibility of reuse.
Optical methods based on fluorescence have been previously reported mainly related to detection of optical brighteners or fluorescent whitening agents in waste waters [27]; or for optical detection of human fecal contamination of water [28,29]. However, the results presented in this paper demonstrated a new approach to detect soap contamination in household wastewater based on an interferometric microdevice that could, hopefully, lead to reuse of part of the greywater produced at home and, therefore, contribute to more rational water consumption habits. From the application point of view, we have to take into account the temperature dependence of refractive index of water. For the case of pure water it is reported that there is about a 0.1% (0.3 %) decrease in refractive index when rising the temp between 20$^{\circ }$C and 40$^{\circ }$C (60$^{\circ }$C) [30]. Thus, calibration procedures are required to perform the analysis, because peak shifts of the same order of magnitude are expected for changes of few mg/ml of soap than for few degrees in the liquid under analysis. Additionally, the actual temperature of the discharge greywater in the FPM should be measured with accuracy. Alternatively, a thermal bath could be used to thermalize both the discharge greywater and the reference tap water samples. In this way, the measurements of the transmission spectral shifts between both samples would be conducted at the same temperature avoiding uncertainties due to changes in temperature.
For actual integration in wash machines or in dishwashers, a compact power deconvolution method can be used to interrogate the FPM. With this interrogation technique, a monochromatic laser light is transmitted through the FPM and detected with a photodiode or photoresistor. The power of the transmitted light depends on the composition of the liquid inside the FPM [31]. In this way, the quality of the discharged water can be estimated from a simple and cheap electrical measurement, without the need of large and expensive optical analysers.
4. Conclusions
Automatic control of the quality of greywater may be of great interest to identify its adequate treatment prior to reuse. Arguably, the most frequently found contaminants of greywater are detergents. In this paper, remote optical detection of detergent pollutants in water was demonstrated using FPM resonators. The resonant optical modes of the microcavities were observed in the transmission spectra of the devices. Two spectral ranges were investigated, visible and NIR. The spectral positions of the FPM modes are sensitive to the refractive index of the fluidic medium. The absolute sensitivity of the NIR FPM microcavity was 68% higher than in the visible FPM resonator.
A spectral red shift of the FPM modes was observed as the concentration of soap was increased. Calibration curves of the FPM@NIR were conducted for SLS and four commercially available soap brands. Furthermore, soapy grey water samples from the discharges of an actual automatic washing machine (for both loaded and unloaded washing cycles) were investigated with the FPM@NIR. Comparison of the measured red shifts with the calibration curves allowed an estimation of the detergent concentration of the greywater discharges. Moreover, the trends and interpretation of the chemical analysis of the greywater samples agree with those remotely achieved through the FPM analysis.
Despite the lack of selectivity of the FPM, which is an intrinsic limitation of refractometric sensors, they can be useful for those cases in which the main pollutant of greywater is a specific commercial detergent. For example, if integrated within the outlet pipes of dishwasher or washing machines, for a given specific calibration of the sensor to a certain detergent brand, the spectral shifts of the FPM modes can be used to determine its concentration. The FPM can be interrogated with a laser and a photodiode or photoresistor, providing a compact, simple and cheap device. Since the remote optical analysis proposed in this research provides real time detection of detergent content, it can be used to automatically control the water discharge set-up to eventually reuse some greywater streams for gardening or plant irrigation, for instance. This would hopefully contribute to a more rational use of domestic water.
Funding
Agencia Canaria de Investigación, Innovación y Sociedad de la Información (ProID2020010067); Agencia Estatal de Investigación (PID2019-107335RA-I00); European Regional Development Fund (PID2019-107335RA-I00, PID2019-110430GB-C21).
Acknowledgments
We wish to thank Servicio de Medio Ambiente from Universidad de La Laguna, for technical assistance with the chemical analytical assays of greywater samples.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are available in [32].
References
1. J. F. Velasco-Muñoz, J. A. Aznar-Sánchez, L. J. Belmonte-Ureña, and I. M. Román-Sánchez, “Sustainable water use in agriculture: A review of worldwide research,” Sustainability 10(4), 1084 (2018). [CrossRef]
2. E. Z. Berglund, J. E. Pesantez, A. Rasekh, M. E. Shafiee, L. Sela, and T. Haxton, “Review of modeling methodologies for managing water distribution security,” J. Water Resour. Plan. Manag. 146(8), 03120001 (2020). [CrossRef]
3. H. H. Ngo, W. Guo, R. Y. Surampalli, and T. C. Zhang, Green Technologies for Sustainable Water Management (American Society of Civil Engineers, 2016).
4. The International Water Association, “International statistics for water services 2018,” (2018). Available at https://iwa-network.org/wp-content/uploads/2018/10/IWA-2018-Statistics-Flyer-Web-ready-revised.pdf.
5. A.-F. Morote, J. Olcina, A.-M. Rico, and M. Hernandez, “Water management in urban sprawl typologies in the city of alicante (southern spain): New trends and perception after the economic crisis?” Urban Sci. 3(1), 7 (2019). [CrossRef]
6. D. M. Ghaitidak and K. D. Yadav, “Characteristics and treatment of greywater–a review,” Environ. Sci. Pollut. Res. 20(5), 2795–2809 (2013). [CrossRef]
7. A. Maimon, E. Friedler, and A. Gross, “Parameters affecting greywater quality and its safety for reuse,” Sci. Total Environ. 487, 20–25 (2014). [CrossRef]
8. Y. Boyjoo, V. K. Pareek, and M. Ang, “A review of greywater characteristics and treatment processes,” Water Sci. Technol. 67(7), 1403–1424 (2013). [CrossRef]
9. F. Lahoz, I. R. Martín, J. Gil-Rostra, M. Oliva-Ramirez, F. Yubero, and A. R. Gonzalez-Elipe, “Portable ir dye laser optofluidic microresonator as a temperature and chemical sensor,” Opt. Express 24(13), 14383–14392 (2016). [CrossRef]
10. F. Lahoz, I. R. Martín, D. Walo, J. Gil-Rostra, F. Yubero, and A. R. Gonzalez-Elipe, “A compact and portable optofluidic device for detection of liquid properties and label-free sensing,” J. Phys. D: Appl. Phys. 50(21), 215103 (2017). [CrossRef]
11. O. Sayginer, A. Chiasera, L. Zur, S. Varas, L. Thi Ngoc Tran, C. Armellini, M. Ferrari, and O. S. Bursi, “Fabrication, modelling and assessment of hybrid 1-d elastic fabry perot microcavity for mechanical sensing applications,” Ceram. Int. 45(6), 7785–7788 (2019). [CrossRef]
12. B. E. A. Saleh and M. C. Teich, Fundamentals of Photonics (Wiley Series in Pure and Applied Optics), vol. 5 (John Wiley & Sons, 1991).
13. R. J. Jiménez Riobóo, M. Philipp, M. A. Ramos, and J. K. Krúger, “Concentration and temperature dependence of the refractive index of ethanol-water mixtures: Influence of intermolecular interactions,” Eur. Phys. J. E 30(1), 19–26 (2009). [CrossRef]
14. X. Fan and I. M. White, “Optofluidic microsystems for chemical and biological analysis,” Nat. Photonics 5(10), 591–597 (2011). [CrossRef]
15. F. Sedlmeir, R. Zeltner, G. Leuchs, and H. G. Schwefel, “High-q mgf2 whispering gallery mode resonators for refractometric sensing in aqueous environment,” Opt. Express 22(25), 30934 (2014). [CrossRef]
16. V. D. Ta, R. Chen, D. M. Nguyen, and H. D. Sun, “Application of self-assembled hemispherical microlasers as gas sensors,” Appl. Phys. Lett. 102(3), 031107 (2013). [CrossRef]
17. J. Gil Rostra, K. Soler-Carracedo, L. L. Martín, F. Lahoz, and F. Yubero, “Thermo-optic response of meh-ppv films incorporated to monolithic fabry-perot microresonators,” Dyes Pigm. 182, 108625 (2020). [CrossRef]
18. C. Wu, Z. Liu, A. Ping Zhang, B.-O. Guan, H.-Y. Tam, X. Wang, J. Xu, Y. Zhu, K. L. Cooper, and A. Wang, “In-line open-cavity fabry-pérot interferometer formed by c-shaped fiber fortemperature-insensitive refractive index sensing,” Opt. Express 22(18), 21757–21766 (2014). [CrossRef]
19. C. A. Bondi, J. L. Marks, L. B. Wroblewski, H. S. Raatikainen, S. R. Lenox, and K. E. Gebhardt, “Human and environmental toxicity of sodium lauryl sulfate (sls): Evidence for safe use in household cleaning products,” Environ. Health Insights 9, EHI.S31765 (2015). [CrossRef]
20. O. R. Al-Jayyousi, “Greywater reuse: towards sustainable water management,” Desalination 156(1-3), 181–192 (2003). [CrossRef]
21. B. Jeppesen, “Domestic greywater re-use: australia’s challenge for the future,” Desalination 106(1-3), 311–315 (1996). [CrossRef]
22. S. Finley, S. Barrington, and D. Lyew, “Reuse of domestic greywater for the irrigation of food crops,” Water, Air, Soil Pollut. 199(1-4), 235–245 (2009). [CrossRef]
23. R. K. Misra, J. H. Patel, and V. R. Baxi, “Reuse potential of laundry greywater for irrigation based on growth, water and nutrient use of tomato,” J. Hydrol. 386(1-4), 95–102 (2010). [CrossRef]
24. M. Lado and M. Ben-Hur, “Treated domestic sewage irrigation effects on soil hydraulic properties in arid and semiarid zones: A review,” Soil Tillage Res. 106(1), 152–163 (2009). [CrossRef]
25. B. Sawadogo, M. Sou, and N. Hijikata, “Effect of detergents from greywater on irrigated plants: Case of okra (abelmoschus esculentus) and lettuce (lactuca sativa) (desert technology 11 international conference),” J. Arid Land Studies 24, 117–120 (2014).
26. N. K. Chaudhary, A. Bhattarai, B. Guragain, and A. Bhattarai, “Conductivity, surface tension, and comparative antibacterial efficacy study of different brands of soaps of nepal,” J. Chem. 2020, 1–13 (2020). [CrossRef]
27. C. Hagedorn, M. Saluta, A. Hassall, and J. Dickerson, “Fluorometric detection of optical brighteners as an indicator of human sources of water pollution. part i. description and detection of optical brighteners” (2005).
28. P. G. Hartel, C. Hagedorn, J. L. McDonald, J. A. Fisher, M. A. Saluta, J. W. Dickerson, L. C. Gentit, S. L. Smith, N. S. Mantripragada, K. J. Ritter, and C. N. Belcher, “Exposing water samples to ultraviolet light improves fluorometry for detecting human fecal contamination,” Water Res. 41(16), 3629–3642 (2007). [CrossRef]
29. D. Dubber and L. W. Gill, “Suitability of fluorescent whitening compounds (fwcs) as indicators of human faecal contamination from septic tanks in rural catchments,” Water Res. 127, 104–117 (2017). [CrossRef]
30. A. N. Bashkatov and E. A. Genina, “Water refractive index in dependence on temperature and wavelength: a simple approximation,” Proc. SPIE 5068, 393–395 (2003). [CrossRef]
31. M. F. Domingues, C. Tavares, N. Alberto, A. Radwan, P. André, and P. Antunes, “High rate dynamic monitoring with fabry-perot interferometric sensors: An alternative interrogation technique targeting biomedical applications,” Sensors 19(21), 4744 (2019). [CrossRef]
32. S. de Armas-Rillo, “Original data for the journal article “Optical monitoring of detergent pollutants in greywater”,” Universidad de La Laguna Institutional Repository (2022), https://riull.ull.es/xmlui/handle/915/27421.
