Final Report Summary - SENSEOCEAN (SenseOCEAN: Marine sensors for the 21st Century)
Executive Summary:
The SenseOCEAN project; marine sensors for the 21st century, has been an exceptionally successful project. A number of products have been brought to the market place and are already generating sales for our SME partners. We have a series of patent applications for technology developments that have occurred during the project. The project has published a number of high impact scientific papers and we have been very effective in promoting the results of the project to a wide audience of scientists, engineers, technologists, policymakers and members of the general public.
SenseOCEAN has produced a new suite of sensors that are generally small, energy efficient and can be mass manufactured using state of the art processes (chip imprinting, 3D printing etc.) to allow cost reductions. A suite of sensors based on optode, electrochemical and voltammetric technologies, do not use any wet reagents, whilst we have a whole new family of lab on chip sensors to measure a wide range of compounds. The project has developed a control hub that allows multiple sensors to be deployed simultaneously with co-registered data collection; the system is integrated with a communication system to allow the transmission of data in real-time via the Iridium telecommunications network. The project has even developed a novel, patented antifouling system to protect the sensors for long-term deployments. On the data side, we have implemented data protocols based on internationally agreed standards, employing the World Wide Web Consortium Linked Data and Open Geospatial Consortium Sensor Web Enablement.
The sensors, both individually and as part of the integrated multifunctional sensor package, have been deployed in a wide range of marine and freshwater environments. Sensors have been used in rivers, wastewater treatment plants, under glaciers, coastal areas and in open ocean areas from the tropics to the poles. Some of the sensors that we have developed have been deployed for periods of up to two years in the marine environment.
We have developed sensors designed to address some of the key global environmental challenges such as climate change, organic pollution and eutrophication. The project has acted as a catalyst to bring all of the projects together that were funded under the original Ocean of Tomorrow call, and encouraged a sense of community to develop standardised approaches to global environmental problems. The projects have worked together on the development of a briefing document outlining how the work done under SenseOCEAN and the other projects is addressing key aspects of the Marine Strategy Framework Directive and the United Nations’ Sustainable Development Goals.
Looking to the future, many of the sensors are now being taken to the marketplace, either through direct manufacturing by our SME partners, or under licensing agreements with large multinational corporations. The interest in the sensors developed in SenseOCEAN is increasing and there is a growing interest from all sectors of the market including wastewater companies, oil companies, regulators and the research community. The development of the sensors will continue under the auspices of EU and nationally funded projects into the future.
Project Context and Objectives:
This project brought together established and world leading marine sensor developers from across Europe, with previous EU project experience including 5 SMEs, to develop a new in situ marine biogeochemical system that is highly integrated, multifunctional, cost-effective and deployable on a mass scale. The world’s oceans are under threat from a myriad of sources, climate change, eutrophication and pollution to name a few. Our understanding of the way the oceans work and their ability to respond to the challenges they are under is poorly understood. A large part of this lack of understanding is due to our inability to measure critical components of the marine biogeochemical cycles that affect the way the oceans function, and determine the effect that the oceans have on the global climate. Along with this inability to measure certain key components of the biogeochemical cycles in the oceans, we are also unable to measure them over long time scales and large spatial scales; in short, the ocean is vast and poorly monitored. Clearly, we need to sample more of the oceans. However, there are physical restrictions which can affect accessibility; some areas of the oceans are ice covered for parts of the year or weather conditions make taking ships to certain areas difficult. In addition, ship operations are very costly and data is only collected for the limited time that the ship is at sea. Ideally, we need an autonomous means of collecting data over long time periods that is able to collect the type of data we need to inform our understanding of the ocean/climate system. There are autonomous platforms in the oceans that could be used to collect an enhanced data set e.g. there are over 3800 Argo floats in the worlds oceans, but currently most are limited to physical measurements (temperature, conductivity and depth) in the top 2000m of the ocean. Our ability to measure the key biogeochemical parameters is limited by the types of sensors we presently have available.
This collaborative project aimed to provide significant advances in the ability to quantify a suite of biogeochemical parameters that are hard to measure but crucial to the scientific understanding of the oceans, the management of ocean resources and the in-situ calibration / validation of satellite Earth Observation data, as well as to the supply of data for the development of advanced biogeochemical models. This was to be achieved through innovation in combination with state of the art sensor technologies (microfabrication, lab on chip, micro and calibration free electrochemical sensors, multiparameter optodes, multispectral optical sensors) in a modular system that can be deployed across multiple ocean and environmental platforms. Prototypes were planned to be optimized for scale up and commercialization including preparation of data flow and data management architectures. These would then be tested and demonstrated on profiling floats, deep-sea observatories, autonomous underwater vehicles, and fishing vessels. Specific overall objectives of the project were:
• The development of a cost-effective, small, integrated in situ marine sensor system that measures key biogeochemical parameters and is suitable for deployment in large numbers to address measurement over wide geographical and temporal scales. The novel package of sensors will be delivered by the end of the project, by which time they will have undergone thorough development and testing, including ‘real-world’ field testing. Further to this an assessment of the commercial viability of the developed sensors and legal protection for the developed systems will be in place.
• Production of high performance analytical systems giving improved limits of detection and accuracy for pH, pCO2, Dissolved Inorganic Carbon (DIC), Alkalinity, O2, NH4+, N2O, NO2-, NO3-, PO43-, SiO44-, Fe, Mn, Coloured Dissolved Organic Matter (CDOM), Chlorophylls, photopigments, primary production, organic fluorophores, photosynthetically available radiation, optical backscatter, irradiance, radiance and transmittance in integrated systems.
• Delivery of state of the art resource (e.g. chemicals, batteries) supply and management using pressure balanced batteries and chemical “printer cartridge” cassettes. This represents a move towards a common core technology for the sensors.
• Delivery of modular hardware and software interfaces for multiple platforms e.g. ocean gliders and profiling floats as part of project members Glider Observatories and bio-Argo initiatives. This represents the development of the modular systems, the integrated systems and subsequent testing on observatories, AUV’s, buoys etc.
• Delivery of operational state of the art data management architectures conforming to internationally agreed formats with outputs of real time data and a post processed data product via existing international services.
• Web enablement to allow real-time control / access to sensor settings and data.
• Innovation in analytical technologies that can be easily applied to a wider list of parameters, e.g. the microfluidic platform could be used for further inorganic assays, or as biosensors for organics and nucleic acids. An overarching goal of SenseOCEAN is to ensure that the systems developed have the potential to be exploited for different environmental applications.
• Direct stakeholder (including fisheries, marine industry, regulators) engagement via consortium partners and our advisory panel. It is important that the sensors we produce are those required by our end user stakeholders and the data produced can inform other stakeholders to best environmental practices.
Underpinning the main objectives described above, there were other core objectives within the work packages. These were:
• Develop common core technologies for use on all sensors. Core technologies are sensor package electronics, communications systems, data management, connectors and interfaces, mechanical systems, resources and their management (power and reagents if used), shared analytical systems, and bio-fouling protection.
• Ensure the seamless integration of all developed sensors in respect to mechanical setup and electronic interfaces.
• Develop a suite of novel high performance analytical technologies to measure biogeochemical parameters, which are suitable for use in small low cost sensors that can be mass-produced. The sensor technologies and specific areas for development are:
– Multiparameter optical sensors
– Miniaturized and autonomous silicate and phosphate electrochemical cells
– Guard solutions for electrochemical microsensors
– High sensitivity electrochemical sensor for nitrous oxide
– Electrochemical sensor for CO2
– pH, CO2, and ammonia optode dyes and opto-electronics
– Multi-analyte optodes and opto-electronics
– Lab on a chip (extended parameter list, analytical performance and analytical miniaturization)
• Produce preproduction prototype of multiple integrated sensor suite and individual sensors
• Ensure design of sensors takes the manufacture into consideration
• Address robustness, ease of operation and long term performance of the sensors
• Perform detailed testing of the sensor technology and integrated sensor suites entailing:
– Laboratory testing to enable evaluation of analytical performance and validation of prototype components and systems
– Performance evaluation of sensors and sensor suites in progressively more realistic scenarios. This will include laboratory testing in synthetic and real seawater at high pressure and low temperature.
– Testing in the real environment to resolve any unforeseen environmental or operational problems
• Set up systems, resources and supply routes to manufacture large numbers of each sensor and the integrated sensor suites
• Design production systems
• Complete demonstration exercises of sensors and sensor suites in environments representative of those in pre-existing markets by deploying on underwater platforms (for example, autonomous lander, benthic observatories, autonomous underwater vehicles (AUV) and water column moorings) in North Sea shallow waters, deeper waters on European continental shelf, deeper upwelling areas off South America and the Arctic deep sea.
• Enhance dissemination of knowledge from the project (commercial and academic dissemination)
• Identify market scale and opportunities
• Develop exploitation and market penetration strategies
Project Results:
SenseOCEAN has achieved its main objective of producing an integrated multifunctional biogeochemical sensor package that had been successfully demonstrated on several platforms in varying marine conditions. In the course of achieving this overall objective, many other significant advances have been made.
The instruments and sensors that have been developed as part of the SenseOCEAN project include:
– Two electrochemical microsensors, one for CO2 and one for N2O (Aarhus, Unisense)
– A datalogger for electrochemical microsensors (Unisense)
– Optodes to measure pH, CO2, O2 and NH3. In addition, the optodes have been produced in two versions, one for integration with the SenseOCEAN unit and a standalone version (TU Graz, Pyro Science)
– Reagentless electrochemical sensors for Si and PO4 (CNRS)
– Multiparameter fluorometer, marketed as V Lux (CTG)
– Lab on chip sensors for pH, nitrite, nitrate, ammonium, phosphate, silicate, dissolved iron and manganese with sensors for dissolved inorganic carbon and total alkalinity also in development (NERC, Southampton)
– MODBUS interface and protocol unit (CTG)
The results of the sensor development work including the improvement of the analytical methods and components on which they are based, are described below in the following sections.
ELECTROCHEMICAL MICROSENSORS FOR NITROUS OXIDE AND CARBON DIOXIDE
For the N2O and CO2 electrochemical microsensors, the same detection principles (reduction at a cathode poised at highly negative electrochemical potential) and general configuration are used. The choice of cathode material and electrolyte are critical; the electrolytes have to be non-aquatic to avoid excessive zero current due to electrolysis of water with H2 evolution at the cathode.
For both sensors, oxygen interference is a major issue. Oxygen is reduced at a much less negative potential than N2O or CO2 so any oxygen reaching the cathode would interfere with the signal. Therefore, it was necessary to design an oxygen trap to place in front of the N2O and CO2 reducing cathodes. Substantial resources were spent on developing non-aqueous O2 trap solutions: 1 M diphenyl-phosphine in propylene carbonate resulted in outstanding sensor characteristics in terms of zero current and analyte sensitivity (Sveegaard et al. 2017). However, damage to the membranes and CO2 consumption by O2 reduction products required the development of an alternative aqueous O2 trap solution. The O2 trap solutions are 1 M ascorbate, pH 12 for the N2O sensors and 1 M Cr2+ solution, pH 1 in the CO2 sensors. It is essential to use an acidic solution for CO2 to avoid CO2 being trapped as bicarbonate or carbonate.
The N2O and CO2 microsensors have glass casings. For the CO2 sensor, the cathode is silver or gold. A patent is pending on the metal for the N2O sensors. The diameter of the sensor tip can vary. Sensors where a small size is more important than high sensitivity may have tip diameters down to about 20 µm, whereas higher sensitivity is obtained with sensors with larger tip diameter. The maximum diameter of the sensors constructed for N2O and CO2 is about 500 µm otherwise the O2 trap solution cannot keep up with the O2 influx through the silicone membrane.
The N2O sensor exhibits excellent long-term stability and shows a high degree of linearity of the signal. The lifetime of the sensor, even when monitoring continuously in a waste water treatment plant, is several months. This sensor has a theoretical detection limit of a few nM, however, this can only be achieved in a temperature controlled laboratory with very frequent calibration. To improve the detection limit for N2O, a STOX type sensor was constructed. The STOX oxygen sensor has revolutionized our understanding of the Oxygen Minimum Zones (OMZs), as it has a 100-1000 fold better detection limit by in situ oxygen measurements than the optical and electrochemical sensors used prior to the STOX sensor introduction in 2007 (Revsbech et al. 2009, Revsbech et al. 2011, Thamdrup et al. 2011). The STOX sensor is made with an extra cathode that, when not polarized, allows the analyte (O2 or N2O) to enter, and when polarized prevents its entry to the measuring cathode. Signal interpretation is done on the amplitude in signal between front cathode depolarized/polarised rather than on the absolute signal. The signal for the polarized front cathode may change due to lower than 100% efficiency of O2 (or N2O) reduction, but the amplitude in signal is still linearly correlated to the partial pressure of the gas.
The work on the STOX type N2O sensor is still in progress. Recently, a new combined design was implemented to increase the sensitivity, and a resolution better than ~5 nM was achieved, allowing us to analyse ambient levels of N2O in the ocean. In contrast to ordinary sensors the in situ STOX detection limit is at least as high as under laboratory conditions, as electrical noise is virtually absent with in situ work.
The CO2 sensor (Revsbech et al. submitted) that has been developed has the same architecture as the N2O sensor. It is the first functional sensor for the analysis of dissolved CO2 based on CO2 reduction that has ever been developed. It has excellent sensitivity at atmospheric concentrations. Work is in progress to lower the response time which is currently too slow (about 3 min for increasing concentrations, longer for decreasing concentrations).
ELECTROCHEMICAL SENSORS FOR SILICATE AND PHOSPHATE
Electrochemical sensors allow the detection of silicate and phosphate in seawater without the addition of liquid reagents. The electrochemical part of the sensor works by simple oxidation at fixed potential of a molybdenum electrode that produces molybdate ions (MoO42-) and H+ needed to form silico- and phospho-molybdic complexes. To reach the required acidic pH, the counter electrode is isolated behind a Nafion® membrane which limits the reduction of the protons formed at its surface. In the case of phosphate, in order to avoid the interference of silicate, a ratio of H+/MoO42- = 70 is needed. To achieve this, another molybdenum electrode is placed behind a proton exchange membrane. The oxidation of the first molybdenum electrode produces H+ ions that cross the membrane and concentrate on the cell where phosphates are sampled, enabling the acidic pH to be reached. The oxidation of the second Mo electrode gives the ratio required and only the phosphomolybdic complex is formed and detected. Both complexes are detected on a gold working electrode by cyclic voltammetry for silicate and by square wave voltammetry for phosphate.
The concentration of phosphate in the open ocean is much lower than silicate, therefore the signal recorded is much smaller. Potential pulses and a derivative method such as square wave voltammetry (SWV) are used to increase the sensitivity of the analytical method. In fact, due to a special potential modulation (potential ramp combined with short-term potential pulses) and the way the current is sampled (at the end of each pulse), the contribution of the capacitive current can be considered as negligible and the faradaic current is mainly sampled corresponding to the electronic transfer of interest. An article has been published in Talanta in 2016 about the optimisation of the SWV parameters to detect phosphate in seawater, toward the in situ electrochemical sensor
The electrochemical cells were further refined to give better results by isolating the detection step in a separate cell. The silico- and phospho-molybdic complexes are first formed in a complexation cell with the molybdenum electrodes (and membranes) and then transferred into a detection cell with a gold working electrode using a Lee-Company® solenoid pump. The flow rate of the pump was found to be particularly important in order to transfer the complexes without any dilution but also to fill and rinse efficiently the whole circuit. In order to obtain a pseudo plug flow, a pressure difference between the entrance and the exit of the pump/entrance of the cell is created allowing to drastically decrease the flow rate and increase the efficiency of the pump.
The electrochemical silicate sensor was designed and built with plugged electrodes to allow easy recovery and reconditioning, making the sensor very practical. The silicate sensor (Figure 7), is an anodized aluminium cylindrical sensor of 2.2 kg in the air, 90 mm diameter and 250 mm height without the connector. The “technical part” is placed on the “top” of the sensor, the electronics in a dry compartment and the pump in dielectric oil compartment. The housing is common for the silicate and the phosphate sensors as well as the detection cell. Only the complexation cells (with the molybdenum electrodes) are different between the two versions (silicate and phosphate). The electronics of the sensors have been developed in collaboration between LEGOS-CNRS and NKE Instrumentation. The optimisation of the electronic card for the silicate version gave very satisfying results compared to the commercial potentiostat Metrohm® with a signal increase of almost 20 %.
LAB ON CHIP SENSORS
The main lab on chip sensor development has been for nitrate, nitrite, pH, phosphate, dissolved iron and manganese, ammonium, dissolved inorganic carbon and total alkalinity sensors. Some development of dissolved organic phosphorus, dissolved organic nitrogen and particle sensors has also been carried out.
The assays for the lab on chip sensors have been further optimised to make them fit for purpose for long term precise and accurate deployments. One example is the addition of PVP to the phosphate assay, resulting in a patent submission (NE/P006833/1: A novel analytical assay for the spectrophotometric determination of phosphate in natural waters). Work is ongoing to enhance the lifetime of the silicate assay lifetime including the addition of stabilising polymers and adjusting the acid: molybdate ratio.
In addition, a comprehensive study was undertaken (and is still on-going) to assess the shelf-life of the standards and reagents used in the iron, manganese, silicate, nitrate/nitrite and phosphate sensor assays. The storage variables tested were temperature, time, fixative (for the standard solutions) and container material (various polymers with differing levels of gas permeability and light exclusion). Nitrate/nitrite, phosphate and silicate LOC reagents were found to be particularly susceptible to heat induced degradation primarily resulting in an increased blank measurement (due to discolouration of the solution). All solutions still accurately and precisely measured the determinands after 6 months storage – the detection limits increased slightly with an increased blank signal. Proposed alternative storage options include freeze-drying and gelification with phosphate and nitrate/nitrite with trials ongoing. The initial data for the Griess reagent (nitrate/nitrite assay) show that there no significant change in colour formation (absorbance) when the Griess is mixed with a NO2- standard over the initial test period (18 months) across a wide temperature range (4-35°C).
The lab on chip fluidics have been further developed to improve performance and stability. To achieve greater limits of detection a new long optical cell (approx. 10 cm) was designed although this has shown that the measured noise in the optical cell tends to scale with cell length, resulting in only minor improvements compared to shorter cells. LED intensity stability can affect measurements therefore LED monitoring Photodiodes have been incorporated into the latest LOC sensor designs so that any intensity fluctuations can be calibrated out. For systems such as pH and TA that require more than one light source a unique Y shaped light channel that focuses two light sources into one optical cell has been designed. The original nitrate sensors mixed fluids by diffusion in a very long section of channel; to improve this, a custom micromixer channel was designed, exploiting secondary Dean flows, which was optimized through simulations, tested in the lab, and then implemented on the nitrate sensors. The resulting mixer decreases the channel length by over a factor of 5 which reduces reagent use, fluidic backpressure, and flushing time.
OPTODES FOR pH, OXYGEN, CARBON DIOXIDE AND AMMONIA
The optodes represent a combination of two essential elements: a read-out device and sensing material. The optoelectronic unit is a key component of the read-out device and includes the red LED for the excitation, a set of optical filters and a photodetector. The excitation light is sent via an optical feed-through to the sensing material which emits near-infrared luminescence. The luminescence is guided back through the feed-through and is detected by the photodetector. Optodes utilise the ability of indicator dyes to change their luminescent properties depending on the analyte concentration. Whereas the basic concepts of optodes were established many years ago, their applications in practice have been limited by the availability of high performance sensing materials and dedicated compact, lightweight and low cost read-out devices specially designed for their interrogation. The teams of TU Graz and Pyro Science combined their efforts in the framework of the SenseOCEAN project to design a new generation of optodes for the measurement of several important parameters in seawater.
Optode “sensing chemistry” was developed in TU Graz. Several main challenges had to be solved:
(i) all the materials should have similar spectral properties to be compatible with the read-out devices manufactured by Pyro Science
(ii) the sensors should cover the dynamic measurement range relevant for oceanographic applications.
(iii) the materials should have high chemical robustness for long-term deployment in various environments
(iv) the materials should enable reliable referenced measurements with phase modulation technique
These challenges have been successfully addressed and new high performance sensing materials have been developed. Many different materials had to be prepared and tested throughout the project in order to find the best materials, which are described here for each sensor.
The oxygen sensors use bright benzoporphyrin dyes featuring efficient absorption in the red part of the spectrum and strong phosphorescence in near-infrared. The phosphorescence of the dyes is quenched in presence of oxygen. Due to the long decay time of phosphorescence, the sensors are self-referenced and the measured phase shift provides direct information about pO2. Importantly, we designed a palette of oxygen indicators with similar spectral properties (Pt(II) and Pd(II) porphyrins) which, in combination with various polymeric matrices, allow for tuneability of the sensitivity of the oxygen optodes. Whereas the sensors with a dynamic range from 1 to 1000 µM dissolved oxygen cover most of the applications, those operating in nanomolar range have also been prepared. Importantly, they can be used in combination with the same optoelectronic unit.
Aza-BODIPY indicator dyes are key components of the pH sensing material. These dyes are also excitable with the red light and feature NIR fluorescence. The fluorescence is strong in acidic conditions and is quenched in basic media. Since fluorescence is not a self-referenced parameter, addition of referenced material was necessary. NIR emitting inorganic phosphor Egyptian blue proved to be very suitable for this purpose. A dual Lifetime Referencing read-out scheme is applied to convert the fluorescence of the pH indicator into a referenced measurement of luminescence phase shift. The main developmental efforts were focused on: (i) selection of the indicator with dynamic range optimal for oceanographic applications; (ii) finding a polymeric matrix enabling fast response at low temperature and (iii) elaboration of methods for covalent immobilization of the indicators into the matrix. The last two tasks turned out to be particularly challenging but these challenges were successfully solved within the project. The commonly used polyurethane hydrogels were found to be unsuitable for preparation of the sensors for oceanographic applications since the sensor response at low temperatures was slowed dramatically (several hours at 5 °C). We prepared a completely new matrix based on copolymer of acryloylmorpholine and hydroxyethyl acrylamide. Several strategies for covalent immobilization of the pH indicators have been developed. The immobilization with B-O coupling (post modification of the pH indicator) proved to be particularly promising due to comparably low synthetic effort. The sensors showed fast response (seconds to minutes depending on the thickness of the sensing layer) even at low temperatures.
New materials for CO2 sensing also rely on aza-BODIPY indicators that however are not fluorescent, but show changes of absorption spectrum in the NIR region. These pH indicators are chemically and photochemically robust and have high pKa values which enable their use as highly sensitive CO2 sensors. In order to convert the colorimetric changes into reference phase shift read-out, a mixture of two phosphors (fluorescent and phosphorescent) is used. The long-term stability of the CO2 sensors had been a critical issue and this problem has been addressed in SenseOCEAN project. We developed new coating layers based on perfluorinated polymers (such as Hyflon AD) which efficiently restrict diffusion of acidic gases (such as H2S) into the sensing layer and resultant poisoning of the sensor. The new pCO2 optodes are stable for several weeks of deployment. However, the trade-off is comparably slow response time (particularly at low temperature) which must be considered e.g. in case of profiling applications. Fast-responding pCO2 sensors have also been prepared but they also show significantly faster poisoning.
Ammonia optodes are based on new aza-BODIPY dyes with pKa values in the acidic region. The dyes are immobilized into hydrophilic polymers covered with a hydrophobic porous membrane as a protective layer. The fluorescence of NIR-emitting dyes is quenched upon deprotonation in presence of ammonia. The sensors show very good sensitivity with a dynamic range of 0.2-300 µg/l of dissolved ammonia which is optimal e.g. for application in fish farming (toxicity limit of 25 µg/l for aquatic organisms). Unfortunately, the sensors are limited to measurements at pH 7.2 since in acidic media most of the analyte is converted to ammonium which is not detected by the sensor. During SenseOCEAN, we also developed a new concept for ammonia sensing based on application of ammonium-selective fluoroionophores. In contrast to ammonia sensors based on pH indicators, these sensors do not show any cross-talk to secondary and tertiary amines but are currently not sensitive enough for oceanographic applications.
Several multiparameter optode sensors were also designed and characterized, such as a dual sensor for pH and oxygen and a dual sensor for oxygen and temperature. These sensors allow for simultaneous measurement of more than 1 parameter with a single material which contains several indicators and reference materials. For the first time, we prepared an optode for simultaneous sensing of oxygen and temperature (which therefore enables compensation of the oxygen sensor for temperature effect) based on a single indicator dye. We demonstrated that the multi-parameter sensors are very promising for some applications (e.g. imaging of analyte distribution) but would require significant modification of the optoelectronic unit which was not desirable due to higher costs and loss of compatibility for other analytes.
MULITPARAMETER FLUOROMETER
CTG has developed a generic design for a new multi-parameter fluorometer that incorporates additional measurements to enable fluorescence readings to be corrected for environmental interferences. The new fluorometer will be commercialized under the name V-Lux.
V-Lux detects fluorescence from aromatic or heterocyclic compounds. Fluorescence intensity is directly proportional to concentration and the technique is recognized as one of the most sensitive detection methods available. In real world deployments, however, the measured fluorescence can be attenuated by high levels of turbidity and/or absorbance (‘colour’) in the sample, leading to an underreporting of the measured concentration, which limits the use of fluorescence for some applications. To combat this, V-Lux measures both the solution’s turbidity and its absorbance and uses a correction algorithm, developed during the project, to provide robust fluorescence measurements, which has the added benefit of extending the linear dynamic range of the fluorometer by at least a factor of 10.
V-Lux is available in two basic configurations targeting either UV or visible fluorescence.
– UV VARIANTS
At UV wavelengths there can be significant spectral overlap between fluorescing compounds, which can reduce the selectivity of the measurement. To combat this, the UV variants of V-Lux provide three fluorescence channels. A common excitation wavelength is used so that interferences are common to all three channels. More specific detection can then be achieved by taking a ratio of the three outputs to eliminate these common interference effects.
All UV variants have Chlorophyll-a and CDOM fluorescence channels to help identify false positives arising from UV fluorescence from algae and/or non-specific background CDOM fluorescence.
The optical design for the UV variants is as follows. For fluorescence measurements, modulated output from a UV LED is reflected off a dichroic filter and directed into the sample through a lens and sapphire window. The duty cycle and drive current of the UV LED modulation is optimized both to maximize sensor performance and extend calibration intervals.
Chlorophyll-a, PAH and CDOM fluorescence generated in the sample passes back through the sapphire window, is collimated and transmitted through the dichroic filter. A second dichroic filter transmits the longer wavelength Chlorophyll-a fluorescence, which is then focused onto a SiPM detector. The shorter wavelengths are reflected off the second dichroic filter and a third filter splits the PAH (reflected) from CDOM (transmitted) wavelengths.
Optical transmission is measured using a UV-enhanced photodiode positioned opposite the excitation window. A reference photodiode monitors the output of the UV LED and this signal is used to track any drift in LED output and also provides a reference for the absorbance measurement.
Turbidity measurements use an infrared LED (860 nm) mounted at 90° to the main optical axis. Output from this LED passes through an aperture and a sapphire window. Scattered light generated in the sample is then detected using the same photodiode used for the absorbance measurement.
– ALGAL VARIANT
In the V-Lux (Algae) variant, four discrete excitation wavelengths are used to target common algal pigment groups across the visible spectrum. Energy absorbed by these pigments is rapidly transferred to a Chlorophyll-a molecule and the arising fluorescence is then detected. By monitoring the changes in Chlorophyll-a fluorescence, as a function of excitation wavelength, it is possible to detect changes in algal group composition, e.g. at the onset of a cyanobacterial bloom.
The output from four low cost, high intensity, LEDs is collimated through the sample to provide a range of fluorescence excitation wavelengths across the visible spectrum, targeting the absorption of the main algal pigment groups.
Excitation light absorbed by the light harvesting pigments in algae is rapidly transferred to Chlorophyll-a in Photosystem II and a portion of this energy is then re-emitted as fluorescence at 685nm. This fluorescence passes back through two dichroic filters and is focused onto a SiPM detector. The UV LED is still incorporated in the algal variant to provide anti-biofouling protection.
– CALIBRATION
A calibration method was developed during the project that reports fluorescence output intensity relative to a traceable Quinine Sulphate fluorescence standard. This approach was essential to ensure that the outputs from all fluorometer channels can be compared directly, without reference to the specific calibration compounds used. The approach also allows different fluorometers to be compared directly, which has not been possible up until now.
DATA LOGGER FOR ELECTROCHEMICAL MICROSENSORS
Microsensors generate currents in the picoampere range, therefore highly specialised electronics are required. Prior to SenseOCEAN, the amplifiers for microsensors were integrated in large dataloggers from Unisense A/S, which could operate autonomously for up to 40 hours. In SenseOCEAN, however, a new datalogger was developed which can measure continuously for up to 14 days and has the capability to interface to other devices. The datalogger can be integrated with other sensor devices through RS-232, RS-485 or analog output. This new device allows the integration of microsensors in a larger variety of applications (see Figure 5 attached).
Unisense has developed this product even further and designed a special version for industrial applications which integrates the mass-fabricated sensors developed in the SenseOCEAN-project. (Figure 6 attached)
OPTODE/OPTOELECTRONIC UNIT
Several instruments for the read-out of the optodes have been developed by Pyro Science during the SenseOCEAN project. Two types have been produced and deployed. The first type relies on an opto-electronic unit in a pressure-resistant titanium housing and is equipped with a deep-sea connector and a Modbus board allowing for shared data logging along with other units developed by SenseOCEAN partners. The optoelectronic unit is equipped with the red LED for the excitation, a set of optical filters and a photodetector.
The second type is a stand-alone version which includes a logger and a battery. The optoelectronic unit is placed into a POM housing with a wall thickness of 15 mm, which enables an operation depth down to 500 m. Due to a highly optimised power consumption of the optode and the logger the battery capacity is sufficient for approximately 1.4 million single measurement points which is equivalent to a deployment time of more than one year with a measurement interval of 30 s. The internal logger is able to trigger periodic measurements between 1 s and 1 h and features an internal flash memory with a storage capacity sufficient for approximately 20 million data points. The data readout is performed via USB port using the standard Windows software of the device.
In both devices, an optical feed-through with a fibre thickness of 2 mm and a theoretical pressure resistance to 3000 m is used to connect the optical sensing materials with the optoelectronics. The outer end of the feed-through is equipped with an M6x0.75 outer thread. This is used to mount the sensor materials which are glued onto transparent PMMA caps with the respective internal thread. Thereby the “sensing chemistry” (and hence the analyte the optode is measuring) is easily exchangeable. Another feed-through is implemented for a Pt-100 resistance thermometer. The temperature data is especially useful for internal temperature compensation of the sensing materials. A 6-pin Subconn connector is used for power-supply and communication.
INTEGRATED SYSTEMS
The SenseOCEAN project has developed a Modbus system to handle the data and communication aspects of sensor integration, to lead to a truly multi-parameter sensor package (see specific details later). This desire for an integrated package meant each of the sensors had to be developed with this end goal in mind, coupled with a desire to also implement the integrated package on a number of marine observing platforms. An example of this effort was the work done by nke Instrumentation to test the sensor on its environment simulator of a PROVOR profiling float before integration onto the float. This test bench make it possible to model the effects of adding the sensor to the platform and its interaction with the platform hardware and software prior to deployments. All efforts were made in the design of the sensor integration onto the PROVOR float to provide easy access to all the connectors and/or reagent ports on each of the sensors. Each sensor can easily be removed from the holding without disassembling the whole structure.
The data management system uses the Modbus interface unit to collect SenseOCEAN sensor data and a communication board is used to transmit data files over satellite to an Iridium Provider Server. An exchange multi-application protocol (EMAP) is used between Modbus interface unit and the communication board. Then data files are then retrieved from the Iridium Provide Server by the British Oceanographic Data Centre.
There are some specific details related to the different sensors when used in an integrated fashion on the multiparameter sensor. The optodes and lab on chip sensors are compliant with the SenseOCEAN Modbus communications specification whereas the silicate and phosphate electrochemical sensors bypass the Modbus board and are directly connected to communication board using the EMAP protocol.
The multiparameter fluorometer is compliant with the SenseOCEAN Modbus communications specification, thus simplifying integration with other sensors using the SenseOCEAN Modbus interface unit. The fluorescence correction algorithms are implemented within the firmware so no external data processing is required.
The lab on chip sensor’s electronics architecture has been changed to provide separate RS232 and RS485 serial ports exclusively for communications with external platforms while user configuration, data download and firmware upgrade functions have been moved to a new USB interface. All three interfaces can be used simultaneously. The pin-out of the sensor’s two IE55 connectors has been standardised across all chemistries, with USB in one and power (10-16V), RS232 and RS485 in the other. The sensor’s firmware architecture has been changed to have an abstract platform interface supporting the input of time, salinity, temperature, depth and ascent/descent from the platform, and the output of measurement result (e.g. nutrient concentration, pH), measurement timestamp and sensor status information to the platform. Platform-specific drivers can now be easily added as thin wrappers around this abstract interface and we have provided 10 so far, including an RS485/Modbus driver for integration with the SenseOCEAN hub which was demonstrated successfully in the WP1 test deployments at Kiel, Southampton and Villefranche-sur-Mer. This approach has also allowed us, in other projects, to successfully integrate with platforms including gliders (Kongsberg Seaglider and BRIDGES Deep Explorer), various project-specific hubs on landers and vehicles, commercial loggers (e.g. Storm 3) and a Trident Systems Iridium unit. In all cases except the Deep Explorer (which was a lab test), these were successfully deployed at sea.
DESIGNS, PROCESSES, METHODS AND MANUFACTURING ASPECTS
– MASS-FABRICATION OF MICROSENSORS
At present, all electrochemical microsensors are hand-made, including those marketed by Unisense A/S. One of the objectives of SenseOCEAN was to develop cheaper sensors that can be used routinely in various marine settings. Significant resources have therefore been allocated to the development a new sensor design for the electrochemical microsensors that enables mass-production and with a mechanical construction that offers enhanced robustness for applications where spatial resolution is of minor importance. Needle-shaped sensors with dimensions of << 1 mm still have to be produced by hand, but it is now possible to construct sensors with microscale dimension internally, but with macroscopic dimensions on the outside. The internal micro-dimensions ensure that all the positive measuring characteristics of the microsensors, such as efficient supply of reactants to the tip region and insensitivity to stirring, are retained. Likewise, the response time, sensitivity and long-term stability of the sensors are comparable or better than the hand-made sensors.
Unisense has filed for three patents on this new sensor design and expects that the sensor design will offer unique possibilities for the company in monitoring and control applications, such as open-water applications, but also more industrial applications. Unisense intend to continue the development of the new sensor platform after the SenseOCEAN project to exploit the industrial opportunities.
– ELECTROCHEMICAL Si AND PO4 SENSOR DESIGN
The first version of the in situ electrochemical silicate sensor helped to redefine the overall mechanical design to produce a more advanced version of the sensor. Testing of the first in situ sensor clearly showed that the mechanical design was not practical for marine operations. For instance, to recover the electrodes, it was necessary to completely open the sensor, including the dry compartment containing the electronics, which meant there was a high probability of incurring damage, additionally, the waterproof and pressure tests had to be repeated each time to sensor was opened. The second, updated version allows easy recovery and reconditioning of the electrodes, and changing of other parts.
– LAB ON CHIP MANUFACTURING. METHODS AND QUALITY CONTROL
For the lab on chip sensors, several improvements were made to the manufacturing. A new conformal layer has been applied to the press plate for manufacturing. This permits many devices to be pressed in parallel despite small variations in thickness that would otherwise inhibit press uniformity. A multi-layered film is used to both distribute the force and prevent sticking to the press platen. Currently, press time dominates the total bonding process, not vapour exposure so the overall throughput has been significantly improved and cost decreased correspondingly.
As a part of the standard quality control procedures, a procedure has been developed to generate stitched optical transmission micrographs with 5 µm resolution which is sufficient to assess bond quality, layer alignment, and contamination in channels. Current methods to assemble optics are manual, but yield losses are very low (<5%) for current chip designs.
Methods to use mechanical tolerances within the current design to allow for inherently aligned optics were explored at length. Ultimately, due to batch-to-batch variation of the optical components, other approaches to improving the optics alignment procedure have been implemented, such as thermally cured with hot plates to accelerate curing. Also, test equipment to verify optical alignment and fixturing is fully scalable to eliminate any concerns of a bottleneck at this step in the production process.
Quality control procedures for the lab on chip sensors have also been streamlined making them more robust and acceptable to the external oceanographic community, with the production of a document, ‘The application of ‘analytical figures of merit’ to the nutrient sensors’, detailing the QC information required for every new sensor developed. This document describes what is required, and what is considered necessary by the oceanographic and analytical chemistry community to ensure that the analytical results from the nutrient sensors are fit for purpose. It will allow direct comparison between the sensors and other benchtop methods (i.e. using the SEAL autoanalyser, or FIA or simply using a batch method). Additionally, it will allow us to produce, with a high level of confidence, figures of merit for a whole system, rather than individual sensors. This document also combines learning from a workshop on ‘uncertainties in oceanography’ (as part of the SenseOCEAN programme) with the figures of merit document, and provides a step by step guide on how to obtain these figures (limit of detection, limit of quantification, dynamic range, limit of linearity, uncertainty: accuracy, precision and the ‘Nordtest’ approach).
– MULITPARAMETER FLUOROMETER
Various options for the mechanical design and constructing the sensor were considered. One of the approaches outlined in the original project proposal was to consider using pressure-balanced, oil-filled, housings in place of traditional pressure housings. However, this was not compatible with the need for internal refractive optics, filters and UV transparency. It quickly became clear that the optimum solution was to use a titanium housing, both for its mechanical strength and its chemical and corrosion resistance. A novel threaded endcap design was developed that reduces the number of components required to seal the housing
UNDERPINNING TECHNOLOGIES AND/OR DESIGN ELEMENTS OF SENSORS
– ELECTRONICS DESIGN FOR MULTIPARAMETER FLUOROMETER
For the multi parameter fluorometer, a key driver in the electronics development was to generate a single PCB design that could be used for both the UV and Visible variants of the fluorometer. A ‘flexi-rigid’ PCB design was selected to aid miniaturization, reduce manufacturing costs and improve reliability. Direct connections between the various PCBs avoid the need for bulky connectors and custom ribbon cables. It also saves costs in manufacture, as all the electronic components can be populated on a single PCB panel.
There are two additional PCBs: the first provides the absorbance measurement, the second allows different digital interfaces to be selected. The flexi-rigid PCB design folds around the internal optical assembly.
– ELECTROCHEMICAL SI AND PO4
The electrochemical silicate and phosphate sensors are based on a measurement process that avoids the use of liquid reagents. This process is unique in comparison with other sensor techniques for these nutrients. The associated technology and embedded electronic board are contained in a housing that fits the constraints of an ARGO float in terms of energy, mechanical design and pressure resistance. The current design is compliant up to 800 m depth and will be improved to achieve the lower limit of the ARGO float of 2000m. It provides an important step forward to avoid liquid reagents for nutrient measurement opening the scope of use the sensors for long term deployment platforms where maintenance is not possible (ocean mooring buoy, Deep sea platform or ARGO float).
– LAB ON CHIP SENSORS
For the lab on chip sensors, the core control electronics have been updated and the sensor firmware has been entirely rewritten to production standards. The new core electronics are based around a modern Atmel SAM4L microprocessor which incorporates a high performance 32-bit ARM Cortex-M4 RISC processor that is specifically designed for ultra-low power applications. User configuration and control of the sensor, firmware updates and data downloads are all now provided from a single Windows-based graphical user interface application (GUI) which connects to the sensor over USB using standard USB Mass Storage and HID interface classes. This means other Windows applications can communicate with the sensor, in particular Windows Explorer, which can now be used for fast drag-and-drop file downloads from the sensor’s 8-Gb FAT32 MicroSD card. The provision of real-time graphical and textual display of raw sensor data, a new state machine editor, and non-volatile on-board diagnostics are all available via the new sensor GUI. The additional processing performance has also allowed the addition of on-board processing of the sensor raw data to generate calibrated measurement results for all supported chemistries (nutrient concentrations and pH).
On-chip (zero dead volume) latching microvalves have been developed. The valves use a Viton membrane that is mechanically bonded into the chip. The valve is actuated by a miniature stepper motor and requires no power to stay in a particular state (open or closed). The new valves bring significant enhancements in terms of flushing volume and power consumption. Valves previously consumed 14% of the fluid budget in the nitrate sensor, whereas the new valves developed through SenseOCEAN reduce this to <1%
The service life of the lab on chip sensors has been significantly increased by the development of a new type of sliding pump seal. The seals contain an integrated spring which dynamically corrects for wear. The seals also prevent oil ingress when the sensors are deployed deep and cold (the seals were demonstrated to function in the NOC pressure test facility at 600 bar and -1C). A new syringe pump for the LOC systems which has the same performance yet 1/3 of the cost of the previous syringe pump has also been developed.
A cartridge-based reagent changing system designed to allow easy non-skilled changing of reagents in the field has been developed. On a typical deployment, a system would be able to conduct three months of hourly measurements on a single reagent pack. The packs can be prepared in the lab by a technician, and the change can be performed in the field by a non-skilled operative.
For the pH lab on chip sensor, the thermistor was redesigned. A commercial glass encased thermistor with a 4mm long glass shaft was sealed with a small quad ring and mounted on the pH microfluidic chip with a modified ¼ 28 connector. The advantage of this design is that no epoxy is used. Therefore, once the thermistor is mounted it can be positioned next to the microfluidic channel before fully tightening the ¼ 28 connector.
A method of rapid prototyping polydimethylsiloxane (PDMS) chips was developed using our current micro-milling polymethylmethacrylate (PMMA) process rather than using the more labour intensive SU8 moulding technology, which requires specialised clean room facilities. A quick and simple fabrication protocol was developed using a novel PDMS/PDMS moulding technique used to fabricate a negative structured PDMS device from a negative PMMA master.
UNDERPINNING TECHNOLOGIES FOR THE INTEGRATED SENSOR PACKAGE
– ELECTRICAL INTERFACE
Connectors are a key cost associated with oceanographic sensor suites. In SenseOCEAN standard connectors are provided either directly on the sensor or via an adaptor. The use of a standard connector also ensures that sensors are directly compatible with a wide range of existing hardware.
A ring architecture layout was adopted for sensor integration to provide maximum flexibility for adding or removing sensors. Economies of scale are achieved through the use of common cabling components. However, additional connectors are required in the ‘repeating unit’ between sensors, which increases overall costs. This approach was demonstrated successfully with a sensor deployment in Kiel during the summer of 2016.
For many deployments, the number of sensors will be pre-defined, driven by the scientific need for the data output from a particular combination of sensors, as well as ease of integration. In these cases, a potted wiring loom was considered to reduce the number of interconnections and thus reduce the cost of moulded connectors by two thirds.
– MODBUS PROTOCOL AND INTERFACE UNIT
At the beginning of the project a Modbus serial communications protocol was developed, all SenseOCEAN sensors are compliant with the protocol. A central Modbus interface unit was developed to act as the master control providing power and receiving the data from each sensor, each of which has a unique MODBUS address. At first initialisation, the Modbus interface unit polls the slave addresses and creates a list of the connected sensors. This provides plug-and-play sensor discovery, allowing for flexible number and type of sensors to be attached to the integrated sensor suite, enabling a suitable combination of sensors to be installed for different deployment scenarios.
Each sensor outputs a header containing key sensor information including name, manufacturer, model number, serial number, a globally unique identifier (GUID), the number of parameters measured and poll rate. For each parameter measured, parameter name, units and data format are also provided. The Modbus sensor header provides sensor metadata, which ensures full data traceability and supplies all the information needed for open data requirements.
The Modbus interface is also able to broadcast CTD data to all the sensors, when provided by the platform. GPS data can also be integrated from the platform.
A key additional advantage of the Modbus protocol is that it is a recognised standard widely used across different industries, which means the individual sensors can be readily integrated into third-party platforms.
– DATA MANAGEMENT
Data from the Modbus interface unit is logged to an internal SD card or output via RS232, using exchange multi-application protocol (EMAP) to a communications board developed by NKE. When these are integrated e.g. on a PROVOR float, once the float surfaces, data is transmitted over satellite to an Iridium Provider Server. British Oceanographic Data Centre (BODC) can then retrieve files from the Iridium Provider Server using FTP. BODC have developed a standard for the sensor metadata enabling plug-and-play sensor integration. A XML configuration file is transmitted with the data, providing the required SensorML metadata. New parameter mappings, standard names and SensorML have been integrated by BODC into the data management system for SenseOCEAN. BODC will publish ‘five star’ linked data using Open Geo Spatial (OGC)’s Sensor Web Enablement standards.
Providing an internal logging capability also allows the sensor to be deployed for extended periods needing only an external power supply, which significantly reduces the costs for unattended operation.
– INTEGRATION OF SENSOR SUITE ONTO PLATFORMS
CTG designed a sensor suite for deployment on a mooring/observatory. Sensors are supported across three tiers of clamp plates, each mounted along a central steel rod that can be used for profiling. The clamp plates are manufactured in acetyl and have been designed with integral slots to provide flexible regions that can be clamped around each of the sensors, thus minimising the number of separate machined bracket components needed, an approach that has been used successfully on CTG’s FastOcean APD system. The assembly sits on a fourth acetyl ‘skirt’ component.
For the float/AUV suite, significantly stricter guidelines in terms of size, weight and power than the frame-based system suitable for mooring or observatory deployments had to be met. Because of this the number of sensors deployed was restricted. NKE adopted the design principles behind the moored approach, outlined above, for a successful deployment of their PROVOR float in Villlefranche-sur-Mer in the spring of 2017,
DATA
The SenseOCEAN project aimed to enable SenseOCEAN data to be interoperable with operational data flows and international aggregators of data, producing an operational standardised data flow. The state of the art in data technology has shifted significantly since the start of the project with SenseOCEAN making a significant contribution to these developments.
– Uniquely identifying sensors identifiers for sensors
The first challenge addressed within SenseOCEAN was to enable unique machine readable identification of sensors. The solution was for each sensor to include a unique 8 character identifiers with data. This concept is being developed further in a newly forming Research Data Alliance working group to broaden the unique identification to enable globally unique identifiers akin to digital object identifiers.
– OGC standards
From the outset of the project SenseOCEAN worked closely with the Marine SWE profiles community that aim to harmonise and develop common templates for marine data following open geospatial consortium standards. The specific standards used and enhanced in SenseOCEAN are SensorML to describe sensor metadata and Observations and Measurements (O&M) for data These standards are being pursued within the environmental and marine domains on a range of European projects including all four oceans of tomorrow projects, AtlantOS and EMSODev.
At the outset of the SenseOCEAN project the formats were syntactically interoperable with the marine community, within SenseOCEAN we developed semantic interoperability where terms in the documents were constrained to controlled vocabularies. For example, salinity can be described using a precise term on the NERC vocabulary server (http://vocab.nerc.ac.uk/collection/P01/current/PSALST01/) rather than a free-text term (sal, Psal, salin, etc) enabling the record to be machine readable and automatic ingestion into OGC Sensor Observation Service systems such as those developed by the EMSO European Research Infrastructure Consortium.
SensorML templates have been developed for the sensors
– W3C standards
In parallel with the OGC standards the World Wide Web Consortium (W3C) developed a metadata standard called Semantic Sensor Network (SSN). This standard serves a similar function to OGC SensorML with the W3C standards being developed for use in the context of W3C linked. This is a generic standard for standardised metadata on the internet and is used in applications covering sectors broader than the environment and marine.
SenseOCEAN developed templates for the exposure of metadata in SSN format using the same semantic enrichment previously described of OGC SensorML.
This is not the final state of W3C and OGC standards. Current research progressing this standardisation is the OGC Internet of Things standard and this begins to align OGC and W3C standards. An updated SSN standard is also being developed called Sensor, Observation, Sample, and Actuator (SOSA) which will address deficiencies identified in SSN and improve alignment with OGC standards (https://www.w3.org/TR/vocab-ssn/#Developments).
– Serving of data and metadata
BODC developed an ontology fusing together existing available ontologies to enable metadata to be stored in a single database and exposed using both W3C SSN and OGC SensorML metadata standards.
Existing software solutions and adopted where possible as this enables rapid standardisation of results and community development. Software implemented includes:
- 52North software for OGC O&M standard and OGC SOS service
-The US NOAA ERDDAP software, including collaboration with the development team on SeaDataNet formats in ERDDAP enabling trans-Atlantic interoperability. ERDDAP also enables the output of NetCDF data and was used to develop a web interface of creation of SenseOCEAN metadata records.
- W3C linked data software
APPLICATIONS
Within SenseOCEAN, newly developed sensors were used at different locations to demonstrate their performance and usability, as detailed below.
Electrochemical silicate sensor by LEGOS/CNRS and NKE:
- Chilean coast – Mooring
Optodes (O2, CO2, pH) by TU Graz and Pyro Science
- Baltic Sea during PROSID2014 cruise on RV Salme – CTD
- Denmark, Limfjorden – Stand-alone module
- Arctic Fram Strait – TRAMPER (electronic assemblies)
- Arctic Fram Strait – AUV Paul (only pH optode)
- Antarctic, Potter Cove – Stand-alone module
- USA, Californian shelf - AUV
- USA, Californian shelf – Floating buoy
- USA, California, Moss Landing –Abalone tank
LOC (Phosphate, Nitrate, pH) by NOC
- UK, Southampton, Empress Dock - stainless steel frame at pontoon
- UK, Christchurch Harbour (deployment > 18 months)
- USA, Ohio, Maumee River – Flow-through tank
- USA, Chesapeake Bay, Chesapeake Biological Laboratory – Stainless steel frame
- USA, Hawaii, Kaneohe Bay – Stainless steel frame
- USA, Hawaii, Kaneohe Bay – Fish pond in He’eia
- Seychelles – Mooring and CTD
- Arctic Fram Strait – Mooring
- Porcupine Abyssal Plain (PAP) – Mooring and CTD
- Sweden, Gullmarfjord – Stainless steel frame at pontoon
- UK, Bridlington – AUV
- Pacific near Tortel (Chile) – underway ship-based sampling
- Stand-alone device in glacial meltwater streams at Chile (Steffen Glacier) and Greenland (Narsarsuaq and Leverett Glacier)
Electrochemical CO2 by Aarhus University and Unisense
- Sweden, Aarhus, Aarhus University – Laboratory experiments
- Germany, Wilhelmshaven
Electrochemical N2O by Aarhus University and Unisense
- Part of sensor control equipment in about 200 wastewater treatment plants worldwide.
FRRf by CTG and MPI
- Germany, Helgoland – Hypersub observatory platform
- Svalbard, Ny-Ålesund – Hypersub observatory platform
- Germany, Bremen, Max-Planck-Institute – Laboratory experiments
Multi-sensor suite by all SenseOCEAN partners, include a variety of different parameter
- Germany, Kiel - Stainless steel frame at pontoon and CTD
- UK, Southampton, Empress Dock - Stainless steel frame at pontoon
- Mediterranean Sea – PROVOR float
– APPLICATION IN SCIENCE
Within SenseOCEAN newly developed single sensors and the multi-sensor suite were employed to address different scientific questions. The seasonal cycle of nitrate in the upper Arctic Ocean at two sites, one mainly ice covered and the other mainly ice free, was revealed by NOC Lab on chip (LOC) nitrate sensors. Furthermore, SenseOCEAN optode sensor technology from Pyro Science enables monthly measurements of benthic oxygen fluxes with the TRAMPER system. Both trials obtained unique data from the remote Arctic surface and deep sea waters of the Arctic Ocean. This allows us to draw a baseline of these parameters, which is crucial to track climate-change related processes and effects in the vulnerable Arctic Ocean.
Another major topic of the United Nations’ Sustainable Development Goal (SDG) 14 is the detection of the greenhouse gas nitrous oxide (N2O). The improved electrochemical N2O sensor (University of Aarhus & Unisense) provides the ability to study production and depletion processes in the atmosphere and the ocean. Currently, an expedition is scheduled for April 2018 to the oxygen-minimum-zone (OMZ) of Mexico. The main theme of the expedition will be the transformation of N2O in OMZ regions, which will be tackled by fine-scaled in situ mapping of the N2O distribution.
Within SenseOCEAN two pH sensors types were developed, a LOC type by NOC and an optical type by TU Graz/Pyro Science, both suitable for specific purposes. Shorter measurement times and lower detection limits are the major improved factors of these sensors. Thereby, questions regarding ocean acidification, also a major topic of SDG 14, can by tackled.
Eutrophication and related ocean deoxygenation are two more topics of SDG 14, which can be tracked and monitored by sensors developed in SenseOCEAN. Oxygen optodes, developed by TU Graz/Pyro Science are suitable tools to monitor oxygen concentrations in coastal and shelf areas. A two-month long-term deployment of the sensor in the Antarctic Potter Cove (King George Island/Isla de Mayo 25) proved that the device provides suitable data storage space and battery power even under cold and rough Antarctic conditions. The improved FRRf by CTG, measuring fluorescence and by this chlorophyll, proved reliable operation as well under cold and rough conditions in the Arctic Kongsfjorden (Spitsbergen, Svalbard). Data from this sensor can be used to calculated primary production within a daily resolution and by this provide reliable data to ground truth satellite-based estimations. The LOC sensors for phosphate and nitrate are suitable tools to detect eutrophication and monitor recovery. The LOC nitrate and phosphate sensor were given an honourable mention in the recent Alliance for Coastal Technologies (ACT) nutrient challenge, performing well in freshwater, estuarine and marine systems over long periods
The CO2 concentration is a parameter of great interest for studies of climate change, ocean acidification and eutrophication. An optical CO2 sensor was developed by TU Graz/Pyro Science and a STOX type by University of Aarhus/Unisense. A few years ago, CO2 measurements in the aquatic realm were impossible. Therefore, these developments are a huge step forward. For example, the STOX CO2 sensor has been used for analysing the exchange between seawater and atmosphere during a campaign in Wilhelmshaven. Miniaturized setups will help in a variety of eco-physiological studies such as calcification of organisms (reefs, mussels, etc.) under acidified conditions.
Additionally, the newly developed sensors and particularly the multi-sensor suite can be used to obtain essential ocean variables (EOV) for the General Ocean Observatory Strategy (GOOS) and for the Deep Ocean Observatory Strategy (DOOS). In particular, the multiparameter optical sensor by CTG and the multi-sensor suite are able to measure a set of important parameters in parallel, allowing easy direct comparison of different parameters.
Other research fields will also benefit from our sensor developments. For example, LOC and optode sensors were used in an aquaculture pool and an abalone tank (Haliotis sp.) for monitoring tasks proving to be useful tools for aquaculture research. Measurements of O2 and pH are already commonly used in aquaculture research, however, CO2, nitrate, nitrite, ammonia and N2O would also be of great interest in this field. Due to the extremely small tip size of the electrochemical CO2, O2- and pH microsensors, they are able to measure e.g. inside primary productive tissues and by that give a deeper insight into primary production processes and the balanced/unbalanced production of oxygen and uptake of carbon dioxide. All the sensors are useful in biogeochemical studies, dealing with regional and global element cycling and offset phenomena like eutrophication or upwelling.
– APPLICATION IN INDUSTRY AND OTHER FIELDS
In addition to scientific research, the newly developed sensors can also be of great benefit for industrial use. Similarly to the benefit that the sensors can provide in aquaculture research, industrial aquafarms also use sensors to monitor health and productivity of their products. Industrial fish farms commonly use O2 sensors but again CO2, nitrate, nitrite, and ammonia would be of greater interest. Algae farms, like the Swedish algae factory (http://swedishalgaefactory.com/) also use pH sensors. Their productivity could benefit from the continuously monitoring of nitrate, nitrate, phosphate, ammonia, silicate, CO2 and trace metals (also developed by NOC, but at lower TRL levels at the moment). Our N2O sensor is already used in over a 100 waste water treatment facilities.
The CO2 and the N2O sensors will have a widespread application in medicine, when improvements to measure reliably within air are complete. In the field of anaesthesia, measurements of these parameters in air are of particular interest.
Furthermore, our sensors were proven to work reliably for long durations, even under rough Arctic and Antarctic conditions. The fluorometric sensor from CTG can be used to detect water pollution and petroleum leakages via the measurement aromatic hydrocarbons; aromatic hydrocarbons are important monitoring parameters in the shipping industry. Hence the SenseOCEAN suite of sensors are perfect tools for environmental monitoring tasks performed by, for example, NGOs, environmental protection agencies or nature conservation authorities.
Potential Impact:
Our world is facing major environmental problems that affect us all. These include global warming and climate change, pollution, ocean acidification, ocean oxygenation, eutrophication and loss of marine biodiversity. In turn, these can lead to fish kills, food contamination (e.g. harmful algal blooms), sea level rise (resulting in loss of habitable land) and a loss of utility of the marine system through impacts on tourism (e.g. fish kills and harmful algal blooms).
The oceans provide vital ecosystems services to humans, and these services are currently under multiple stressors that are both increasing and changing, creating complex, often unpredictable feedbacks. Defining acceptable targets for ocean health and sustainability, establishing the knowledge base needed to maintain and improve the health of oceans and developing systems to predict and respond to shocks to and from oceans all represent critical research needs. A multi-scale integrated transdisciplinary approach combining models, observation systems, analytics, experiments, and societal needs is needed to create the knowledge required to map pathways and identify trade-offs in conserving ocean health.
Climate change is one of the greatest threats to the planet, and the marine system is especially sensitive. Major contributors to global warming are the greenhouse gases (GHGs), carbon dioxide (CO2) and nitrous oxide (N2O). Carbon dioxide receives most attention due to its relatively high concentrations, however N2O though currently at levels that are approximately 1000-fold lower than CO2, has a potential global warming effect of >260 times that of CO2. Marine areas can be sources or sinks for both CO2 and N2O and both gases exchange between the atmosphere and the ocean through physical and biological processes. It is therefore necessary to be able to measure both of them to understand these interactions. Within SenseOCEAN there has been development of sensors for both CO2 and N2O, using a number of technological approaches.
The oceans store a substantial amount of carbon dioxide as dissolved inorganic carbon (DIC), which buffers increases in atmospheric CO2 concentrations with the associated consequences for our climate. The large quantities of anthropogenic CO2 taken up by the oceans has resulted in a decrease in seawater pH in a process termed ocean acidification. With increasing atmospheric CO2 levels, the capacity of the oceans to absorb CO2 is decreasing whilst ocean waters reach higher DIC concentrations, and global warming reduces the capacity for seawater to hold as much CO2. The ability of the oceans to sequester CO2 over geological timescales (100s to 1000s of years) depends on the role of physical and biological (soft tissue and carbonate) pumps. These pumps remove carbon from the upper ocean and take it to greater ocean depths, both are influenced by climate change in ways that are currently poorly understood. Against the background of a gradual anthropogenically driven increase in oceanic DIC, there are pronounced natural fluctuations and poorly constrained climate change driven variations in the functioning of the physical and biological pumps. Improved observational capabilities, involving high quality carbonate chemistry measurements, are required to enhance our mechanistic and quantitative understanding of the influences of ocean warming and acidification on oceanic carbon sequestration.
Oceanic uptake of CO2 has a range of consequences for individual biological species and communities. The decrease in carbonate ion concentrations lowers the saturation states of calcite and aragonite, minerals used by a range of organisms for building protective shells (calcification). Ocean acidification can reduce calcification and thereby ballasting of sinking organic matter, and be detrimental for organisms that lack protective mechanisms. Some regions of the oceans (e.g. polar oceans, nutrient enriched coastal waters) are experiencing pronounced reductions in calcium carbonate saturation states, with potential consequences for ecosystems. It is therefore particularly important to monitor carbonate chemistry in these ocean regions which are most vulnerable to ocean acidification.
Ocean deoxygenation is affecting both coastal regions and the open ocean and has implications for oceanic productivity and marine ecosystems. Warmer oceans resulting from climate change, lead to oxygen loss due to the reduced solubility of oxygen, and increased stratification, which inhibits the photosynthetic production of oxygen. In coastal zones, increased nutrient loading (e.g. from agricultural and industrial run-off) can lead to eutrophication (see below) that also causes oxygen depletion. The impacts of deoxygenation on ecosystems and ecosystems goods and services are numerous. Living organisms are impacted at a functional level below certain oxygen concentrations; this varies on a species by species level so it is extremely difficult to define a unique hypoxia threshold. In addition, warming and acidification will act in antagonistic ways to enhance the negative impacts of low oxygen levels. Shoaling of hypoxic waters can lead to cascading effects on benthic and pelagic ecosystems, including marine habitat compression and changes to community structure. Hypoxic events also have severe impacts on goods and services, for instance they may lead to massive fish kills in aquaculture farms which can result in large economic losses.
Eutrophication is a type of pollution caused by an excess supply of nutrients to natural waters, including fresh and marine waters. The main nutrients implicated in eutrophication are the macronutrients, nitrate and phosphate, that are used by plants in high concentrations for growth. Humans are responsible for adding high concentrations of these elements to our waterways from activities such as sewage disposal (nitrate and phosphate), atmospheric inputs (mainly nitrate) due to high temperature combustion processes and run off from agricultural use of fertilizers (nitrate and phosphate). Enhanced concentrations of nutrients in marine systems promote the rapid growth of phytoplankton, and changes in the species composition of the planktonic community. The rapid growth of phytoplankton is known as a bloom and phytoplankton blooms can have significant impacts in both marine and freshwater systems. Deleterious effects include the development of harmful algal blooms (HAB’s) and anoxia. Anoxia occurs due to a massive uptake of oxygen resulting from the microbial decomposition of the phytoplankton as they die and sink. Furthermore, strong blooms result in light shading which means plants below the bloom in the water column are unable to grow due to the lack of sunlight. Harmful algal blooms can impact aquaculture and fisheries (commercial and recreational), and the toxins produced by the phytoplankton can also lead to bathing restrictions and/or closure of beaches.
The task of understanding the global ocean system in order to monitor, comprehend and mitigate these environmental problems is enormous, and can at first seem intractable. The challenge is to measure a wide range of chemicals, processes and other factors, and to measure them on spatial and temporal scales (i.e. frequently and at closely spaced intervals over the whole ocean) that until recently would have been hugely cost prohibitive. In addition, the driving factors behind these global impacts are still poorly understood and we are often limited by our inability to measure the substance and/or process required to increase our understanding.
The traditional approach to obtaining ocean measurements is inherently costly, labour intensive and provides limited data; ocean going research vessels, with complements of marine staff and research scientists, sail to a single point in the ocean and take a series of measurements over a very short timescale (generally a week to a month). Not only is this expensive and technically challenging for some measurements, but there are regions of the ocean that are inaccessible for large periods of the year such as the Arctic and Antarctic oceans.
We are on the cusp of a more integrated ocean observing system, driven in large part by technological developments in autonomy, communications, battery technology and satellite observing. A large part of this technology push is around the platforms used for observing: shallow and deep fixed point observatories (e.g. the Porcupine Abyssal Plain Observatory), moorings, autonomous profiling floats (e.g. there are more than 3800 ARGO floats in the world’s oceans), established ship observation lines (e.g. GO-SHIP), vessels of opportunity, arrays of gliders, autonomous underwater vehicles (AUVs), and benthic crawlers and landers.
A key consideration in using such platforms is they are limited in terms of space, power, data handling and servicing. Therefore, the sensor technology deployed on these platforms must be small, energy efficient, require minimal servicing and be able to store/transfer data. In addition to satisfy the need for measurement at an increased temporal and spatial resolution the sensors should be low cost and ideally, mass producible.
SenseOCEAN has significantly advanced our ability to measure specific parameters related to of some of the major environmental problems facing us, as well as enabling these measurements to be obtained at a greater frequency and on a wider scale.
SenseOCEAN has developed three methods to measure carbon dioxide, the most important greenhouse gas and a key component of the marine carbonate system. They are, electrochemical microsensors, lab on chip sensors and optodes. Having several methods available means that there is sensor ideally suited for different environments. The electrochemical sensors have good sensitivity at partial pressures close to atmospheric and in addition to measuring CO2 in seawater are capable of analysing CO2 within sediment and plant tissue, facilitated by their microscopic tip design. The CO2 optodes can be deployed for several weeks with the optoelectronics contained in very compact pressure resistant housings, these are very cheap sensors. The lab on chip sensors have in situ calibration techniques and are also deployable for extended periods of time.
Nitrous oxide, also an important greenhouse gas, can be measured using an electrochemical microsensor which has been significantly modified and improved. The sensors can be used over a wide concentration range from nanomolar to millimolar concentrations making them applicable to a wide range of environments. They can analyse continuously for several months as has been demonstrated and they have been effectively deployed as part of a monitoring system for at over 100 wastewater treatment plants.
To fully constrain the marine carbonate system, you need a combination of two of the 4 carbonate system variables; Dissolved inorganic carbon (DIC), pH, the partial pressure of CO2 and Total alkalinity. The measurement of pH is vital for assessment of ocean acidification and understanding the marine carbonate system. SenseOCEAN has two sensors for this purpose, the pH optode and the lab on chip pH sensor. Both sensors have been demonstrated successfully in several field deployments including shallow coastal moorings, on autonomous underwater vehicles, on CTD frames and benthic crawlers. The development of in situ sensors for DIC and TA is less advanced, with only a few reported in situ instruments, relying on spectrophotometric pH or infra-red CO2 measurements. SenseOCEAN has started to develop a lab on chip TA and DIC sensor that uses a gas diffusion and conductivity measurement principle. Although at a lower technology readiness level compared to the lab on chip pH sensor (4 versus 7), once increased this will provide a full suite of carbonate system measurements.
Ocean deoxygenation requires measurement of the dissolved O2 in water. The optode oxygen sensors developed within the SenseOCEAN project will make the reliable long term monitoring of oxygen concentrations possible. Oxygen optodes rely on quenching of luminescent indicator dyes by oxygen and are characterized not only by the absence of electromagnetic interferences and analyte consumption, but also by their small size and comparatively low cost. In the stand-alone module (i.e. with optoelectronics and in a housing) developed in the SenseOCEAN project, we are now able to take autonomous long-term measurements for up to 2 years at depths down to 600 m. The deep-sea version is extremely compact and allows parallel measurement with the other SenseOCEAN developed sensors (using the MODBUS communication protocol). Importantly, the read-out modules are fully compatible with other “sensing chemistries” including pH and pCO2 developed during SenseOCEAN and also trace oxygen optode sensors suitable for precise oxygen quantification in low oxygen zones.
In addition to dissolved oxygen, the assessment and monitoring of eutrophication requires the measurement of dissolved nutrients (primarily nitrate and phosphate) concentrations (high concentrations can act as an early warning system for the onset of phytoplankton blooms) and chlorophyll a (gives an indication of algal biomass). Combined with these single indicator measurements the ability to monitor the composition of the phytoplankton community is also important in order to detect any changes that may indicate a community shift towards species that produce toxins, and thus generate a harmful algal bloom (HAB). SenseOCEAN has developed technologies to measure the compounds that the cause the blooms (nutrient concentrations) and the changes in community structure that can indicate a shift to a HAB.
SenseOCEAN has taken two approaches to nutrient measurements, wet chemical lab on chip based systems and reagentless electrochemical methods. The lab on chip approach basically miniaturizes the classical benchtop wet chemistry techniques for the determination of nitrate, nitrite and phosphate, the electrochemical techniques do not use reagents and are applicable to the measurement of silicate and phosphate. Both approaches have strengths and weaknesses and these are more or less important depending on the type of environment they are deployed in and the platforms on which they are used. During SenseOCEAN we have deployed sensors of both types in a range of environments including the tropical Pacific, a eutrophic harbour and a fjord system.
The direct effects of eutrophication can be measured using our multi parameter fluorimeter (V-Lux); by using multiple wavelengths to target common algal pigment groups, changes in algal group composition can be detected e.g. at the onset of a HAB. In addition, the FRRf (fast repetition rate fluorometer) has been widely tested in the marine environment and was deployed on an observatory in the Arctic during SenseOCEAN for over 12 months. The ability to determine community shifts in the phytoplankton community was clearly demonstrated as the year progressed.
To obtain measurements on a wider scale and at a greater frequency, SenseOCEAN has addressed the cost and ease of sensor manufacture. Ensuring that the sensors are low cost and can be produced on an industrial scale reduces barriers to mass deployment of the sensors. Once deployed in the real environment there are factors that can limit the time the sensors can be deployed, these can be fouling by marine organisms that can block detectors and flow through systems, and factors related to the breakdown of reagents used in those sensors that use chemicals. SenseOCEAN has dedicated some effort on both of these factors; biological fouling and reagent lifetime.
With an aim to reducing the cost of the manufacture of sensors we explored state of the art approaches to the whole development process. In the case of the development of our multiparameter fluorometer we were able to reduce costs and ease of manufacturing by adopting one design for several variants of the instrument. Alongside this Chelsea Technology Group has reduced costs by adopting a ‘flexi-rigid’ PCB (printed circuit board) design for their V-Lux multiparameter fluorometer, which folds around the internal assembly of the fluorometer. This aids miniaturisation (no need for bulky connectors), improves reliability and reduces manufacturing costs as all the electronic components are placed on a single PCB panel. The production of electrochemical microsensors was traditionally done in a bespoke manner, with each sensor requiring a specialised technician to hand manufacture each sensor; Aarhus University have simplified their N2O electrochemical microsensor to a 1-compartment design, the new design has enabled Unisense to employ microfabrication techniques, which means the sensors are more robust, manufacturing is consistent and at a much lower cost. The optodes developed by TU Graz and Pyro Science have been engineered to enable read out with a standardised optoelectronics unit, this is done by chemically engineering the spectral properties of the sensing materials used in their optode sensors, thereby minimising manufacturing costs. In addition, a miniaturised polymer housing for the shallow water standalone optode system (including data logger and battery) has been developed, this has reduced the volume of the device by 75% compared to the prototype.
Lab on chip based sensors were traditionally manufactured I a bespoke manner, the versions developed in SenseOCEAN use a standardised platform approach, with the adoption a new manufacturing press plate technique that allows the parallel pressing of multiple devices. New optical component alignment procedures have been implemented which are faster, and scalable, this ensures that this step does not slow the production process. A new syringe pump was also developed with the same performance as those that were traditionally bought ‘out of house’, but at a third of the cost. Looking to the near future a potentially cheaper, more efficient valve has also been identified and is undergoing testing.
Many of the issues that have previously restricted the length of in situ sensor deployments have been addressed by SenseOCEAN. Some of these issues have also restricted the location of long term deployments as frequent need for servicing does not lend itself to remote deployments. For example, the electrochemical nutrient sensor is reagentless, as are the optodes and electrochemical microsensors. The lab on chip nitrate sensors have been improved to reduce reagent use by using a custom micromixer channel. Preservation techniques to prolong reagent life for the LOC sensors have been evaluated with alginate gelification in combination with lyophilisation showing promise as a long term in situ reagent storage solution.
We have addressed the perennially taxing issues around biofouling in order to increase deployment times. On the newly developed optode systems as have designed copper guards that sit around the sensing spot on the optode, this copper kills microscopic organisms and prevents fouling. On other sensors we have developed an In situ chlorination technique and have put effort into exploring the use of ultrasonic acoustic vibration for anti-biofouling. A patent is pending for the in situ chlorination method and the ultrasonic method has been published, showing promise for future application. The optical system developed by Chelsea Technologies integrates a ultra violet flash system to kill fouling organisms on their optical windows.
Of course, designing and producing sensors, carrying out deployments and making measurements is pointless if the data is not made available so that scientists and stakeholders (governments, policy makers, etc.) can analyse and use the data. Equally important is that the users can be assured of the source and quality of the data, including calibration data and any quality assurance processing that may have occurred. Although global standards for marine data do not yet exist, SenseOCEAN has invested time around issues concerned with data availability, quality assurance and standardisation. Every sensor is uniquely identifiable (and the concept is being developed to enable globally unique identifiers), data formats follow standards that are already available and existing software solutions have been adopted. SenseOCEAN has been involved in, and continues to work on, establishing globally accepted standards for marine sensor data and meta data. The Modbus interface receives standardised data files from the SenseOCEAN developed sensors which, via the developed EMAP communication board can be transmitted using the Iridium satellite network.
DISSEMINATION & EXPLOITATION OF RESULTS
The SenseOCEAN project has been promoted at events throughout the four years of the project. Some of the events have also involved other Ocean of Tomorrow (OoT) projects such as the OoT stand at Oceanology International in March 2016 (London, UK), which was organised and coordinated by the SenseOCEAN project.
The SenseOCEAN coordinator has given several presentations to audiences including policymakers (e.g. UK government, Spanish Council, European Commission). A presentation at the EuroScience Open Forum (ESOF), in Manchester, UK, in July 2016 on how our sensors address issues in the marine strategy framework directive and plans to commercialise the sensors was well attended and well received. At an IOCCP (International Ocean Carbon Coordination Project – a UNESCO) side event at the 10th International Carbon Dioxide Conference in August 2017, the coordinator gave a presentation and lead a discussion on using autonomous sensors. In July 2017 he gave a presentation at the European Commission event ‘A New Era of Blue Enlightenment’ in Lisbon, Portugal on sensors and sampling for Atlantic Ocean observations. The COLUMBUS project has also invited presentations from the SenseOCEAN coordinator at two of their events, a brokerage event and the 2nd Annual Conference.
We have also had television and radio coverage of the project. Back in 2015, Matt Mowlem was interviewed about SenseOCEAN for BBC Radio 4’s Inside Science series, resulting in a 12-minute segment during the programme. More recently Euronews filmed some of our partners in the UK and Denmark, showcasing sensors and technology from the project for a segment devoted to SenseOCEAN in one of their ‘Futuris’ series programmes. The programme was broadcast repeatedly for a week on the Euronews TV channel and made available of their website and (www.euronews.com) and Youtube channel in 12 languages.
The SenseOCEAN coordinator and other members of the consortium (e.g. Tellabs and CTG) have contributed to Twitter, Our Policy Brief was tweeted with #SaveOurOceans to coincide with the UN Ocean Conference in June 2017.
Project partners, especially our SMEs have presented their sensors and work at many trade fairs, exhibitions, workshops and meetings aimed at the industry and commercial sectors. In addition to the Ocean of Tomorrow stand at Oceanology 2016, nke also exhibited at the event. Nke also presented the SenseOCEAN project at the Seatechweek SSCO Conference in Brest, France in 2014. Tellabs and Pyro Science have exhibited at the international Analytica trade fair in both 2014 and 2016 (Munich, Germany). Chelsea Technologies Group regularly participate in the annual UK Marine Measurements Forum (MMF) which has members from the academic, industry and public sectors. They announced their new V-Lux fluorometer, developed as part of SenseOCEAN at the 59th MMF meeting before the official launch at WEFTEC 2017 in October in Chicago, USA. Unisense also exhibited their Field Datalogger Mini at WEFTEC 2017 as well as at FEMS 2017 in Valencia, Spain and the 2017 ASLO Aquatic Sciences Meeting in Hawaii.
SenseOCEAN work has also been widely disseminated at renowned international conferences as well as at more specialised technical conferences. SenseOCEAN, NeXOS, SCHeMA and COMMONSENSE convened a joint session at the 2015 ASLO Aquatic Sciences in Granada, Spain where SenseOCEAN scientists presented some of their early results. Talks were also given at the 2016 ASLO Ocean Sciences Meeting in New Orleans, USA. One of the leading international geochemistry conferences is the Goldschmidt Conference. Presentations were given at both the 2016 and 2017 conferences by SenseOCEAN partners. Presentations from SenseOCEAN scientists were also given at the EGU General Assembly in Vienna Austria, in 2016 this conference had on the order of 15,000 attendees.
Specialised technical conferences where SenseOCEAN work has been presented include the 2016 EUROPT(R)ODE XIII Conference on Optical Chemical Sensors and Biosensors (Austria), International Electromaterials Science Symposium 2017 (Australia), the International Conference on Electrochemical Sensors 2017 (Hungary) and the Electrochemical Society Meeting (USA).
To date, SenseOCEAN has published seventeen peer-reviewed papers. A joint publication on the results from the test deployments of the integrated multifunctional sensor package is in preparation and a joint publication on societal impacts and UN SDG goal 14 and how SenseOCEAN technology can address these is also planned.
The development work and results generated as a consequence of the SenseOCEAN project have resulted in 6 patent applications, three in Denmark, two in the UK and one to the World Intellectual Property Organisation. They are related to the electrochemical microsensors and the lab on chip sensors. Three products are already on the market, the Unisense Field Datalogger Mini, the Chelsea Technologies Group V-Lux Fluorometer and the Unisense N2O Microsensor. These will undoubtedly be followed by further products from Pyro Science and products from the commercialisation of the lab on chip technology.
List of Websites:
www.senseocean.eu
Coordinator: Douglas Connelly
National Oceanography Centre, Southampton, UK
douglas.connelly@noc.ac.uk
The SenseOCEAN project; marine sensors for the 21st century, has been an exceptionally successful project. A number of products have been brought to the market place and are already generating sales for our SME partners. We have a series of patent applications for technology developments that have occurred during the project. The project has published a number of high impact scientific papers and we have been very effective in promoting the results of the project to a wide audience of scientists, engineers, technologists, policymakers and members of the general public.
SenseOCEAN has produced a new suite of sensors that are generally small, energy efficient and can be mass manufactured using state of the art processes (chip imprinting, 3D printing etc.) to allow cost reductions. A suite of sensors based on optode, electrochemical and voltammetric technologies, do not use any wet reagents, whilst we have a whole new family of lab on chip sensors to measure a wide range of compounds. The project has developed a control hub that allows multiple sensors to be deployed simultaneously with co-registered data collection; the system is integrated with a communication system to allow the transmission of data in real-time via the Iridium telecommunications network. The project has even developed a novel, patented antifouling system to protect the sensors for long-term deployments. On the data side, we have implemented data protocols based on internationally agreed standards, employing the World Wide Web Consortium Linked Data and Open Geospatial Consortium Sensor Web Enablement.
The sensors, both individually and as part of the integrated multifunctional sensor package, have been deployed in a wide range of marine and freshwater environments. Sensors have been used in rivers, wastewater treatment plants, under glaciers, coastal areas and in open ocean areas from the tropics to the poles. Some of the sensors that we have developed have been deployed for periods of up to two years in the marine environment.
We have developed sensors designed to address some of the key global environmental challenges such as climate change, organic pollution and eutrophication. The project has acted as a catalyst to bring all of the projects together that were funded under the original Ocean of Tomorrow call, and encouraged a sense of community to develop standardised approaches to global environmental problems. The projects have worked together on the development of a briefing document outlining how the work done under SenseOCEAN and the other projects is addressing key aspects of the Marine Strategy Framework Directive and the United Nations’ Sustainable Development Goals.
Looking to the future, many of the sensors are now being taken to the marketplace, either through direct manufacturing by our SME partners, or under licensing agreements with large multinational corporations. The interest in the sensors developed in SenseOCEAN is increasing and there is a growing interest from all sectors of the market including wastewater companies, oil companies, regulators and the research community. The development of the sensors will continue under the auspices of EU and nationally funded projects into the future.
Project Context and Objectives:
This project brought together established and world leading marine sensor developers from across Europe, with previous EU project experience including 5 SMEs, to develop a new in situ marine biogeochemical system that is highly integrated, multifunctional, cost-effective and deployable on a mass scale. The world’s oceans are under threat from a myriad of sources, climate change, eutrophication and pollution to name a few. Our understanding of the way the oceans work and their ability to respond to the challenges they are under is poorly understood. A large part of this lack of understanding is due to our inability to measure critical components of the marine biogeochemical cycles that affect the way the oceans function, and determine the effect that the oceans have on the global climate. Along with this inability to measure certain key components of the biogeochemical cycles in the oceans, we are also unable to measure them over long time scales and large spatial scales; in short, the ocean is vast and poorly monitored. Clearly, we need to sample more of the oceans. However, there are physical restrictions which can affect accessibility; some areas of the oceans are ice covered for parts of the year or weather conditions make taking ships to certain areas difficult. In addition, ship operations are very costly and data is only collected for the limited time that the ship is at sea. Ideally, we need an autonomous means of collecting data over long time periods that is able to collect the type of data we need to inform our understanding of the ocean/climate system. There are autonomous platforms in the oceans that could be used to collect an enhanced data set e.g. there are over 3800 Argo floats in the worlds oceans, but currently most are limited to physical measurements (temperature, conductivity and depth) in the top 2000m of the ocean. Our ability to measure the key biogeochemical parameters is limited by the types of sensors we presently have available.
This collaborative project aimed to provide significant advances in the ability to quantify a suite of biogeochemical parameters that are hard to measure but crucial to the scientific understanding of the oceans, the management of ocean resources and the in-situ calibration / validation of satellite Earth Observation data, as well as to the supply of data for the development of advanced biogeochemical models. This was to be achieved through innovation in combination with state of the art sensor technologies (microfabrication, lab on chip, micro and calibration free electrochemical sensors, multiparameter optodes, multispectral optical sensors) in a modular system that can be deployed across multiple ocean and environmental platforms. Prototypes were planned to be optimized for scale up and commercialization including preparation of data flow and data management architectures. These would then be tested and demonstrated on profiling floats, deep-sea observatories, autonomous underwater vehicles, and fishing vessels. Specific overall objectives of the project were:
• The development of a cost-effective, small, integrated in situ marine sensor system that measures key biogeochemical parameters and is suitable for deployment in large numbers to address measurement over wide geographical and temporal scales. The novel package of sensors will be delivered by the end of the project, by which time they will have undergone thorough development and testing, including ‘real-world’ field testing. Further to this an assessment of the commercial viability of the developed sensors and legal protection for the developed systems will be in place.
• Production of high performance analytical systems giving improved limits of detection and accuracy for pH, pCO2, Dissolved Inorganic Carbon (DIC), Alkalinity, O2, NH4+, N2O, NO2-, NO3-, PO43-, SiO44-, Fe, Mn, Coloured Dissolved Organic Matter (CDOM), Chlorophylls, photopigments, primary production, organic fluorophores, photosynthetically available radiation, optical backscatter, irradiance, radiance and transmittance in integrated systems.
• Delivery of state of the art resource (e.g. chemicals, batteries) supply and management using pressure balanced batteries and chemical “printer cartridge” cassettes. This represents a move towards a common core technology for the sensors.
• Delivery of modular hardware and software interfaces for multiple platforms e.g. ocean gliders and profiling floats as part of project members Glider Observatories and bio-Argo initiatives. This represents the development of the modular systems, the integrated systems and subsequent testing on observatories, AUV’s, buoys etc.
• Delivery of operational state of the art data management architectures conforming to internationally agreed formats with outputs of real time data and a post processed data product via existing international services.
• Web enablement to allow real-time control / access to sensor settings and data.
• Innovation in analytical technologies that can be easily applied to a wider list of parameters, e.g. the microfluidic platform could be used for further inorganic assays, or as biosensors for organics and nucleic acids. An overarching goal of SenseOCEAN is to ensure that the systems developed have the potential to be exploited for different environmental applications.
• Direct stakeholder (including fisheries, marine industry, regulators) engagement via consortium partners and our advisory panel. It is important that the sensors we produce are those required by our end user stakeholders and the data produced can inform other stakeholders to best environmental practices.
Underpinning the main objectives described above, there were other core objectives within the work packages. These were:
• Develop common core technologies for use on all sensors. Core technologies are sensor package electronics, communications systems, data management, connectors and interfaces, mechanical systems, resources and their management (power and reagents if used), shared analytical systems, and bio-fouling protection.
• Ensure the seamless integration of all developed sensors in respect to mechanical setup and electronic interfaces.
• Develop a suite of novel high performance analytical technologies to measure biogeochemical parameters, which are suitable for use in small low cost sensors that can be mass-produced. The sensor technologies and specific areas for development are:
– Multiparameter optical sensors
– Miniaturized and autonomous silicate and phosphate electrochemical cells
– Guard solutions for electrochemical microsensors
– High sensitivity electrochemical sensor for nitrous oxide
– Electrochemical sensor for CO2
– pH, CO2, and ammonia optode dyes and opto-electronics
– Multi-analyte optodes and opto-electronics
– Lab on a chip (extended parameter list, analytical performance and analytical miniaturization)
• Produce preproduction prototype of multiple integrated sensor suite and individual sensors
• Ensure design of sensors takes the manufacture into consideration
• Address robustness, ease of operation and long term performance of the sensors
• Perform detailed testing of the sensor technology and integrated sensor suites entailing:
– Laboratory testing to enable evaluation of analytical performance and validation of prototype components and systems
– Performance evaluation of sensors and sensor suites in progressively more realistic scenarios. This will include laboratory testing in synthetic and real seawater at high pressure and low temperature.
– Testing in the real environment to resolve any unforeseen environmental or operational problems
• Set up systems, resources and supply routes to manufacture large numbers of each sensor and the integrated sensor suites
• Design production systems
• Complete demonstration exercises of sensors and sensor suites in environments representative of those in pre-existing markets by deploying on underwater platforms (for example, autonomous lander, benthic observatories, autonomous underwater vehicles (AUV) and water column moorings) in North Sea shallow waters, deeper waters on European continental shelf, deeper upwelling areas off South America and the Arctic deep sea.
• Enhance dissemination of knowledge from the project (commercial and academic dissemination)
• Identify market scale and opportunities
• Develop exploitation and market penetration strategies
Project Results:
SenseOCEAN has achieved its main objective of producing an integrated multifunctional biogeochemical sensor package that had been successfully demonstrated on several platforms in varying marine conditions. In the course of achieving this overall objective, many other significant advances have been made.
The instruments and sensors that have been developed as part of the SenseOCEAN project include:
– Two electrochemical microsensors, one for CO2 and one for N2O (Aarhus, Unisense)
– A datalogger for electrochemical microsensors (Unisense)
– Optodes to measure pH, CO2, O2 and NH3. In addition, the optodes have been produced in two versions, one for integration with the SenseOCEAN unit and a standalone version (TU Graz, Pyro Science)
– Reagentless electrochemical sensors for Si and PO4 (CNRS)
– Multiparameter fluorometer, marketed as V Lux (CTG)
– Lab on chip sensors for pH, nitrite, nitrate, ammonium, phosphate, silicate, dissolved iron and manganese with sensors for dissolved inorganic carbon and total alkalinity also in development (NERC, Southampton)
– MODBUS interface and protocol unit (CTG)
The results of the sensor development work including the improvement of the analytical methods and components on which they are based, are described below in the following sections.
ELECTROCHEMICAL MICROSENSORS FOR NITROUS OXIDE AND CARBON DIOXIDE
For the N2O and CO2 electrochemical microsensors, the same detection principles (reduction at a cathode poised at highly negative electrochemical potential) and general configuration are used. The choice of cathode material and electrolyte are critical; the electrolytes have to be non-aquatic to avoid excessive zero current due to electrolysis of water with H2 evolution at the cathode.
For both sensors, oxygen interference is a major issue. Oxygen is reduced at a much less negative potential than N2O or CO2 so any oxygen reaching the cathode would interfere with the signal. Therefore, it was necessary to design an oxygen trap to place in front of the N2O and CO2 reducing cathodes. Substantial resources were spent on developing non-aqueous O2 trap solutions: 1 M diphenyl-phosphine in propylene carbonate resulted in outstanding sensor characteristics in terms of zero current and analyte sensitivity (Sveegaard et al. 2017). However, damage to the membranes and CO2 consumption by O2 reduction products required the development of an alternative aqueous O2 trap solution. The O2 trap solutions are 1 M ascorbate, pH 12 for the N2O sensors and 1 M Cr2+ solution, pH 1 in the CO2 sensors. It is essential to use an acidic solution for CO2 to avoid CO2 being trapped as bicarbonate or carbonate.
The N2O and CO2 microsensors have glass casings. For the CO2 sensor, the cathode is silver or gold. A patent is pending on the metal for the N2O sensors. The diameter of the sensor tip can vary. Sensors where a small size is more important than high sensitivity may have tip diameters down to about 20 µm, whereas higher sensitivity is obtained with sensors with larger tip diameter. The maximum diameter of the sensors constructed for N2O and CO2 is about 500 µm otherwise the O2 trap solution cannot keep up with the O2 influx through the silicone membrane.
The N2O sensor exhibits excellent long-term stability and shows a high degree of linearity of the signal. The lifetime of the sensor, even when monitoring continuously in a waste water treatment plant, is several months. This sensor has a theoretical detection limit of a few nM, however, this can only be achieved in a temperature controlled laboratory with very frequent calibration. To improve the detection limit for N2O, a STOX type sensor was constructed. The STOX oxygen sensor has revolutionized our understanding of the Oxygen Minimum Zones (OMZs), as it has a 100-1000 fold better detection limit by in situ oxygen measurements than the optical and electrochemical sensors used prior to the STOX sensor introduction in 2007 (Revsbech et al. 2009, Revsbech et al. 2011, Thamdrup et al. 2011). The STOX sensor is made with an extra cathode that, when not polarized, allows the analyte (O2 or N2O) to enter, and when polarized prevents its entry to the measuring cathode. Signal interpretation is done on the amplitude in signal between front cathode depolarized/polarised rather than on the absolute signal. The signal for the polarized front cathode may change due to lower than 100% efficiency of O2 (or N2O) reduction, but the amplitude in signal is still linearly correlated to the partial pressure of the gas.
The work on the STOX type N2O sensor is still in progress. Recently, a new combined design was implemented to increase the sensitivity, and a resolution better than ~5 nM was achieved, allowing us to analyse ambient levels of N2O in the ocean. In contrast to ordinary sensors the in situ STOX detection limit is at least as high as under laboratory conditions, as electrical noise is virtually absent with in situ work.
The CO2 sensor (Revsbech et al. submitted) that has been developed has the same architecture as the N2O sensor. It is the first functional sensor for the analysis of dissolved CO2 based on CO2 reduction that has ever been developed. It has excellent sensitivity at atmospheric concentrations. Work is in progress to lower the response time which is currently too slow (about 3 min for increasing concentrations, longer for decreasing concentrations).
ELECTROCHEMICAL SENSORS FOR SILICATE AND PHOSPHATE
Electrochemical sensors allow the detection of silicate and phosphate in seawater without the addition of liquid reagents. The electrochemical part of the sensor works by simple oxidation at fixed potential of a molybdenum electrode that produces molybdate ions (MoO42-) and H+ needed to form silico- and phospho-molybdic complexes. To reach the required acidic pH, the counter electrode is isolated behind a Nafion® membrane which limits the reduction of the protons formed at its surface. In the case of phosphate, in order to avoid the interference of silicate, a ratio of H+/MoO42- = 70 is needed. To achieve this, another molybdenum electrode is placed behind a proton exchange membrane. The oxidation of the first molybdenum electrode produces H+ ions that cross the membrane and concentrate on the cell where phosphates are sampled, enabling the acidic pH to be reached. The oxidation of the second Mo electrode gives the ratio required and only the phosphomolybdic complex is formed and detected. Both complexes are detected on a gold working electrode by cyclic voltammetry for silicate and by square wave voltammetry for phosphate.
The concentration of phosphate in the open ocean is much lower than silicate, therefore the signal recorded is much smaller. Potential pulses and a derivative method such as square wave voltammetry (SWV) are used to increase the sensitivity of the analytical method. In fact, due to a special potential modulation (potential ramp combined with short-term potential pulses) and the way the current is sampled (at the end of each pulse), the contribution of the capacitive current can be considered as negligible and the faradaic current is mainly sampled corresponding to the electronic transfer of interest. An article has been published in Talanta in 2016 about the optimisation of the SWV parameters to detect phosphate in seawater, toward the in situ electrochemical sensor
The electrochemical cells were further refined to give better results by isolating the detection step in a separate cell. The silico- and phospho-molybdic complexes are first formed in a complexation cell with the molybdenum electrodes (and membranes) and then transferred into a detection cell with a gold working electrode using a Lee-Company® solenoid pump. The flow rate of the pump was found to be particularly important in order to transfer the complexes without any dilution but also to fill and rinse efficiently the whole circuit. In order to obtain a pseudo plug flow, a pressure difference between the entrance and the exit of the pump/entrance of the cell is created allowing to drastically decrease the flow rate and increase the efficiency of the pump.
The electrochemical silicate sensor was designed and built with plugged electrodes to allow easy recovery and reconditioning, making the sensor very practical. The silicate sensor (Figure 7), is an anodized aluminium cylindrical sensor of 2.2 kg in the air, 90 mm diameter and 250 mm height without the connector. The “technical part” is placed on the “top” of the sensor, the electronics in a dry compartment and the pump in dielectric oil compartment. The housing is common for the silicate and the phosphate sensors as well as the detection cell. Only the complexation cells (with the molybdenum electrodes) are different between the two versions (silicate and phosphate). The electronics of the sensors have been developed in collaboration between LEGOS-CNRS and NKE Instrumentation. The optimisation of the electronic card for the silicate version gave very satisfying results compared to the commercial potentiostat Metrohm® with a signal increase of almost 20 %.
LAB ON CHIP SENSORS
The main lab on chip sensor development has been for nitrate, nitrite, pH, phosphate, dissolved iron and manganese, ammonium, dissolved inorganic carbon and total alkalinity sensors. Some development of dissolved organic phosphorus, dissolved organic nitrogen and particle sensors has also been carried out.
The assays for the lab on chip sensors have been further optimised to make them fit for purpose for long term precise and accurate deployments. One example is the addition of PVP to the phosphate assay, resulting in a patent submission (NE/P006833/1: A novel analytical assay for the spectrophotometric determination of phosphate in natural waters). Work is ongoing to enhance the lifetime of the silicate assay lifetime including the addition of stabilising polymers and adjusting the acid: molybdate ratio.
In addition, a comprehensive study was undertaken (and is still on-going) to assess the shelf-life of the standards and reagents used in the iron, manganese, silicate, nitrate/nitrite and phosphate sensor assays. The storage variables tested were temperature, time, fixative (for the standard solutions) and container material (various polymers with differing levels of gas permeability and light exclusion). Nitrate/nitrite, phosphate and silicate LOC reagents were found to be particularly susceptible to heat induced degradation primarily resulting in an increased blank measurement (due to discolouration of the solution). All solutions still accurately and precisely measured the determinands after 6 months storage – the detection limits increased slightly with an increased blank signal. Proposed alternative storage options include freeze-drying and gelification with phosphate and nitrate/nitrite with trials ongoing. The initial data for the Griess reagent (nitrate/nitrite assay) show that there no significant change in colour formation (absorbance) when the Griess is mixed with a NO2- standard over the initial test period (18 months) across a wide temperature range (4-35°C).
The lab on chip fluidics have been further developed to improve performance and stability. To achieve greater limits of detection a new long optical cell (approx. 10 cm) was designed although this has shown that the measured noise in the optical cell tends to scale with cell length, resulting in only minor improvements compared to shorter cells. LED intensity stability can affect measurements therefore LED monitoring Photodiodes have been incorporated into the latest LOC sensor designs so that any intensity fluctuations can be calibrated out. For systems such as pH and TA that require more than one light source a unique Y shaped light channel that focuses two light sources into one optical cell has been designed. The original nitrate sensors mixed fluids by diffusion in a very long section of channel; to improve this, a custom micromixer channel was designed, exploiting secondary Dean flows, which was optimized through simulations, tested in the lab, and then implemented on the nitrate sensors. The resulting mixer decreases the channel length by over a factor of 5 which reduces reagent use, fluidic backpressure, and flushing time.
OPTODES FOR pH, OXYGEN, CARBON DIOXIDE AND AMMONIA
The optodes represent a combination of two essential elements: a read-out device and sensing material. The optoelectronic unit is a key component of the read-out device and includes the red LED for the excitation, a set of optical filters and a photodetector. The excitation light is sent via an optical feed-through to the sensing material which emits near-infrared luminescence. The luminescence is guided back through the feed-through and is detected by the photodetector. Optodes utilise the ability of indicator dyes to change their luminescent properties depending on the analyte concentration. Whereas the basic concepts of optodes were established many years ago, their applications in practice have been limited by the availability of high performance sensing materials and dedicated compact, lightweight and low cost read-out devices specially designed for their interrogation. The teams of TU Graz and Pyro Science combined their efforts in the framework of the SenseOCEAN project to design a new generation of optodes for the measurement of several important parameters in seawater.
Optode “sensing chemistry” was developed in TU Graz. Several main challenges had to be solved:
(i) all the materials should have similar spectral properties to be compatible with the read-out devices manufactured by Pyro Science
(ii) the sensors should cover the dynamic measurement range relevant for oceanographic applications.
(iii) the materials should have high chemical robustness for long-term deployment in various environments
(iv) the materials should enable reliable referenced measurements with phase modulation technique
These challenges have been successfully addressed and new high performance sensing materials have been developed. Many different materials had to be prepared and tested throughout the project in order to find the best materials, which are described here for each sensor.
The oxygen sensors use bright benzoporphyrin dyes featuring efficient absorption in the red part of the spectrum and strong phosphorescence in near-infrared. The phosphorescence of the dyes is quenched in presence of oxygen. Due to the long decay time of phosphorescence, the sensors are self-referenced and the measured phase shift provides direct information about pO2. Importantly, we designed a palette of oxygen indicators with similar spectral properties (Pt(II) and Pd(II) porphyrins) which, in combination with various polymeric matrices, allow for tuneability of the sensitivity of the oxygen optodes. Whereas the sensors with a dynamic range from 1 to 1000 µM dissolved oxygen cover most of the applications, those operating in nanomolar range have also been prepared. Importantly, they can be used in combination with the same optoelectronic unit.
Aza-BODIPY indicator dyes are key components of the pH sensing material. These dyes are also excitable with the red light and feature NIR fluorescence. The fluorescence is strong in acidic conditions and is quenched in basic media. Since fluorescence is not a self-referenced parameter, addition of referenced material was necessary. NIR emitting inorganic phosphor Egyptian blue proved to be very suitable for this purpose. A dual Lifetime Referencing read-out scheme is applied to convert the fluorescence of the pH indicator into a referenced measurement of luminescence phase shift. The main developmental efforts were focused on: (i) selection of the indicator with dynamic range optimal for oceanographic applications; (ii) finding a polymeric matrix enabling fast response at low temperature and (iii) elaboration of methods for covalent immobilization of the indicators into the matrix. The last two tasks turned out to be particularly challenging but these challenges were successfully solved within the project. The commonly used polyurethane hydrogels were found to be unsuitable for preparation of the sensors for oceanographic applications since the sensor response at low temperatures was slowed dramatically (several hours at 5 °C). We prepared a completely new matrix based on copolymer of acryloylmorpholine and hydroxyethyl acrylamide. Several strategies for covalent immobilization of the pH indicators have been developed. The immobilization with B-O coupling (post modification of the pH indicator) proved to be particularly promising due to comparably low synthetic effort. The sensors showed fast response (seconds to minutes depending on the thickness of the sensing layer) even at low temperatures.
New materials for CO2 sensing also rely on aza-BODIPY indicators that however are not fluorescent, but show changes of absorption spectrum in the NIR region. These pH indicators are chemically and photochemically robust and have high pKa values which enable their use as highly sensitive CO2 sensors. In order to convert the colorimetric changes into reference phase shift read-out, a mixture of two phosphors (fluorescent and phosphorescent) is used. The long-term stability of the CO2 sensors had been a critical issue and this problem has been addressed in SenseOCEAN project. We developed new coating layers based on perfluorinated polymers (such as Hyflon AD) which efficiently restrict diffusion of acidic gases (such as H2S) into the sensing layer and resultant poisoning of the sensor. The new pCO2 optodes are stable for several weeks of deployment. However, the trade-off is comparably slow response time (particularly at low temperature) which must be considered e.g. in case of profiling applications. Fast-responding pCO2 sensors have also been prepared but they also show significantly faster poisoning.
Ammonia optodes are based on new aza-BODIPY dyes with pKa values in the acidic region. The dyes are immobilized into hydrophilic polymers covered with a hydrophobic porous membrane as a protective layer. The fluorescence of NIR-emitting dyes is quenched upon deprotonation in presence of ammonia. The sensors show very good sensitivity with a dynamic range of 0.2-300 µg/l of dissolved ammonia which is optimal e.g. for application in fish farming (toxicity limit of 25 µg/l for aquatic organisms). Unfortunately, the sensors are limited to measurements at pH 7.2 since in acidic media most of the analyte is converted to ammonium which is not detected by the sensor. During SenseOCEAN, we also developed a new concept for ammonia sensing based on application of ammonium-selective fluoroionophores. In contrast to ammonia sensors based on pH indicators, these sensors do not show any cross-talk to secondary and tertiary amines but are currently not sensitive enough for oceanographic applications.
Several multiparameter optode sensors were also designed and characterized, such as a dual sensor for pH and oxygen and a dual sensor for oxygen and temperature. These sensors allow for simultaneous measurement of more than 1 parameter with a single material which contains several indicators and reference materials. For the first time, we prepared an optode for simultaneous sensing of oxygen and temperature (which therefore enables compensation of the oxygen sensor for temperature effect) based on a single indicator dye. We demonstrated that the multi-parameter sensors are very promising for some applications (e.g. imaging of analyte distribution) but would require significant modification of the optoelectronic unit which was not desirable due to higher costs and loss of compatibility for other analytes.
MULITPARAMETER FLUOROMETER
CTG has developed a generic design for a new multi-parameter fluorometer that incorporates additional measurements to enable fluorescence readings to be corrected for environmental interferences. The new fluorometer will be commercialized under the name V-Lux.
V-Lux detects fluorescence from aromatic or heterocyclic compounds. Fluorescence intensity is directly proportional to concentration and the technique is recognized as one of the most sensitive detection methods available. In real world deployments, however, the measured fluorescence can be attenuated by high levels of turbidity and/or absorbance (‘colour’) in the sample, leading to an underreporting of the measured concentration, which limits the use of fluorescence for some applications. To combat this, V-Lux measures both the solution’s turbidity and its absorbance and uses a correction algorithm, developed during the project, to provide robust fluorescence measurements, which has the added benefit of extending the linear dynamic range of the fluorometer by at least a factor of 10.
V-Lux is available in two basic configurations targeting either UV or visible fluorescence.
– UV VARIANTS
At UV wavelengths there can be significant spectral overlap between fluorescing compounds, which can reduce the selectivity of the measurement. To combat this, the UV variants of V-Lux provide three fluorescence channels. A common excitation wavelength is used so that interferences are common to all three channels. More specific detection can then be achieved by taking a ratio of the three outputs to eliminate these common interference effects.
All UV variants have Chlorophyll-a and CDOM fluorescence channels to help identify false positives arising from UV fluorescence from algae and/or non-specific background CDOM fluorescence.
The optical design for the UV variants is as follows. For fluorescence measurements, modulated output from a UV LED is reflected off a dichroic filter and directed into the sample through a lens and sapphire window. The duty cycle and drive current of the UV LED modulation is optimized both to maximize sensor performance and extend calibration intervals.
Chlorophyll-a, PAH and CDOM fluorescence generated in the sample passes back through the sapphire window, is collimated and transmitted through the dichroic filter. A second dichroic filter transmits the longer wavelength Chlorophyll-a fluorescence, which is then focused onto a SiPM detector. The shorter wavelengths are reflected off the second dichroic filter and a third filter splits the PAH (reflected) from CDOM (transmitted) wavelengths.
Optical transmission is measured using a UV-enhanced photodiode positioned opposite the excitation window. A reference photodiode monitors the output of the UV LED and this signal is used to track any drift in LED output and also provides a reference for the absorbance measurement.
Turbidity measurements use an infrared LED (860 nm) mounted at 90° to the main optical axis. Output from this LED passes through an aperture and a sapphire window. Scattered light generated in the sample is then detected using the same photodiode used for the absorbance measurement.
– ALGAL VARIANT
In the V-Lux (Algae) variant, four discrete excitation wavelengths are used to target common algal pigment groups across the visible spectrum. Energy absorbed by these pigments is rapidly transferred to a Chlorophyll-a molecule and the arising fluorescence is then detected. By monitoring the changes in Chlorophyll-a fluorescence, as a function of excitation wavelength, it is possible to detect changes in algal group composition, e.g. at the onset of a cyanobacterial bloom.
The output from four low cost, high intensity, LEDs is collimated through the sample to provide a range of fluorescence excitation wavelengths across the visible spectrum, targeting the absorption of the main algal pigment groups.
Excitation light absorbed by the light harvesting pigments in algae is rapidly transferred to Chlorophyll-a in Photosystem II and a portion of this energy is then re-emitted as fluorescence at 685nm. This fluorescence passes back through two dichroic filters and is focused onto a SiPM detector. The UV LED is still incorporated in the algal variant to provide anti-biofouling protection.
– CALIBRATION
A calibration method was developed during the project that reports fluorescence output intensity relative to a traceable Quinine Sulphate fluorescence standard. This approach was essential to ensure that the outputs from all fluorometer channels can be compared directly, without reference to the specific calibration compounds used. The approach also allows different fluorometers to be compared directly, which has not been possible up until now.
DATA LOGGER FOR ELECTROCHEMICAL MICROSENSORS
Microsensors generate currents in the picoampere range, therefore highly specialised electronics are required. Prior to SenseOCEAN, the amplifiers for microsensors were integrated in large dataloggers from Unisense A/S, which could operate autonomously for up to 40 hours. In SenseOCEAN, however, a new datalogger was developed which can measure continuously for up to 14 days and has the capability to interface to other devices. The datalogger can be integrated with other sensor devices through RS-232, RS-485 or analog output. This new device allows the integration of microsensors in a larger variety of applications (see Figure 5 attached).
Unisense has developed this product even further and designed a special version for industrial applications which integrates the mass-fabricated sensors developed in the SenseOCEAN-project. (Figure 6 attached)
OPTODE/OPTOELECTRONIC UNIT
Several instruments for the read-out of the optodes have been developed by Pyro Science during the SenseOCEAN project. Two types have been produced and deployed. The first type relies on an opto-electronic unit in a pressure-resistant titanium housing and is equipped with a deep-sea connector and a Modbus board allowing for shared data logging along with other units developed by SenseOCEAN partners. The optoelectronic unit is equipped with the red LED for the excitation, a set of optical filters and a photodetector.
The second type is a stand-alone version which includes a logger and a battery. The optoelectronic unit is placed into a POM housing with a wall thickness of 15 mm, which enables an operation depth down to 500 m. Due to a highly optimised power consumption of the optode and the logger the battery capacity is sufficient for approximately 1.4 million single measurement points which is equivalent to a deployment time of more than one year with a measurement interval of 30 s. The internal logger is able to trigger periodic measurements between 1 s and 1 h and features an internal flash memory with a storage capacity sufficient for approximately 20 million data points. The data readout is performed via USB port using the standard Windows software of the device.
In both devices, an optical feed-through with a fibre thickness of 2 mm and a theoretical pressure resistance to 3000 m is used to connect the optical sensing materials with the optoelectronics. The outer end of the feed-through is equipped with an M6x0.75 outer thread. This is used to mount the sensor materials which are glued onto transparent PMMA caps with the respective internal thread. Thereby the “sensing chemistry” (and hence the analyte the optode is measuring) is easily exchangeable. Another feed-through is implemented for a Pt-100 resistance thermometer. The temperature data is especially useful for internal temperature compensation of the sensing materials. A 6-pin Subconn connector is used for power-supply and communication.
INTEGRATED SYSTEMS
The SenseOCEAN project has developed a Modbus system to handle the data and communication aspects of sensor integration, to lead to a truly multi-parameter sensor package (see specific details later). This desire for an integrated package meant each of the sensors had to be developed with this end goal in mind, coupled with a desire to also implement the integrated package on a number of marine observing platforms. An example of this effort was the work done by nke Instrumentation to test the sensor on its environment simulator of a PROVOR profiling float before integration onto the float. This test bench make it possible to model the effects of adding the sensor to the platform and its interaction with the platform hardware and software prior to deployments. All efforts were made in the design of the sensor integration onto the PROVOR float to provide easy access to all the connectors and/or reagent ports on each of the sensors. Each sensor can easily be removed from the holding without disassembling the whole structure.
The data management system uses the Modbus interface unit to collect SenseOCEAN sensor data and a communication board is used to transmit data files over satellite to an Iridium Provider Server. An exchange multi-application protocol (EMAP) is used between Modbus interface unit and the communication board. Then data files are then retrieved from the Iridium Provide Server by the British Oceanographic Data Centre.
There are some specific details related to the different sensors when used in an integrated fashion on the multiparameter sensor. The optodes and lab on chip sensors are compliant with the SenseOCEAN Modbus communications specification whereas the silicate and phosphate electrochemical sensors bypass the Modbus board and are directly connected to communication board using the EMAP protocol.
The multiparameter fluorometer is compliant with the SenseOCEAN Modbus communications specification, thus simplifying integration with other sensors using the SenseOCEAN Modbus interface unit. The fluorescence correction algorithms are implemented within the firmware so no external data processing is required.
The lab on chip sensor’s electronics architecture has been changed to provide separate RS232 and RS485 serial ports exclusively for communications with external platforms while user configuration, data download and firmware upgrade functions have been moved to a new USB interface. All three interfaces can be used simultaneously. The pin-out of the sensor’s two IE55 connectors has been standardised across all chemistries, with USB in one and power (10-16V), RS232 and RS485 in the other. The sensor’s firmware architecture has been changed to have an abstract platform interface supporting the input of time, salinity, temperature, depth and ascent/descent from the platform, and the output of measurement result (e.g. nutrient concentration, pH), measurement timestamp and sensor status information to the platform. Platform-specific drivers can now be easily added as thin wrappers around this abstract interface and we have provided 10 so far, including an RS485/Modbus driver for integration with the SenseOCEAN hub which was demonstrated successfully in the WP1 test deployments at Kiel, Southampton and Villefranche-sur-Mer. This approach has also allowed us, in other projects, to successfully integrate with platforms including gliders (Kongsberg Seaglider and BRIDGES Deep Explorer), various project-specific hubs on landers and vehicles, commercial loggers (e.g. Storm 3) and a Trident Systems Iridium unit. In all cases except the Deep Explorer (which was a lab test), these were successfully deployed at sea.
DESIGNS, PROCESSES, METHODS AND MANUFACTURING ASPECTS
– MASS-FABRICATION OF MICROSENSORS
At present, all electrochemical microsensors are hand-made, including those marketed by Unisense A/S. One of the objectives of SenseOCEAN was to develop cheaper sensors that can be used routinely in various marine settings. Significant resources have therefore been allocated to the development a new sensor design for the electrochemical microsensors that enables mass-production and with a mechanical construction that offers enhanced robustness for applications where spatial resolution is of minor importance. Needle-shaped sensors with dimensions of << 1 mm still have to be produced by hand, but it is now possible to construct sensors with microscale dimension internally, but with macroscopic dimensions on the outside. The internal micro-dimensions ensure that all the positive measuring characteristics of the microsensors, such as efficient supply of reactants to the tip region and insensitivity to stirring, are retained. Likewise, the response time, sensitivity and long-term stability of the sensors are comparable or better than the hand-made sensors.
Unisense has filed for three patents on this new sensor design and expects that the sensor design will offer unique possibilities for the company in monitoring and control applications, such as open-water applications, but also more industrial applications. Unisense intend to continue the development of the new sensor platform after the SenseOCEAN project to exploit the industrial opportunities.
– ELECTROCHEMICAL Si AND PO4 SENSOR DESIGN
The first version of the in situ electrochemical silicate sensor helped to redefine the overall mechanical design to produce a more advanced version of the sensor. Testing of the first in situ sensor clearly showed that the mechanical design was not practical for marine operations. For instance, to recover the electrodes, it was necessary to completely open the sensor, including the dry compartment containing the electronics, which meant there was a high probability of incurring damage, additionally, the waterproof and pressure tests had to be repeated each time to sensor was opened. The second, updated version allows easy recovery and reconditioning of the electrodes, and changing of other parts.
– LAB ON CHIP MANUFACTURING. METHODS AND QUALITY CONTROL
For the lab on chip sensors, several improvements were made to the manufacturing. A new conformal layer has been applied to the press plate for manufacturing. This permits many devices to be pressed in parallel despite small variations in thickness that would otherwise inhibit press uniformity. A multi-layered film is used to both distribute the force and prevent sticking to the press platen. Currently, press time dominates the total bonding process, not vapour exposure so the overall throughput has been significantly improved and cost decreased correspondingly.
As a part of the standard quality control procedures, a procedure has been developed to generate stitched optical transmission micrographs with 5 µm resolution which is sufficient to assess bond quality, layer alignment, and contamination in channels. Current methods to assemble optics are manual, but yield losses are very low (<5%) for current chip designs.
Methods to use mechanical tolerances within the current design to allow for inherently aligned optics were explored at length. Ultimately, due to batch-to-batch variation of the optical components, other approaches to improving the optics alignment procedure have been implemented, such as thermally cured with hot plates to accelerate curing. Also, test equipment to verify optical alignment and fixturing is fully scalable to eliminate any concerns of a bottleneck at this step in the production process.
Quality control procedures for the lab on chip sensors have also been streamlined making them more robust and acceptable to the external oceanographic community, with the production of a document, ‘The application of ‘analytical figures of merit’ to the nutrient sensors’, detailing the QC information required for every new sensor developed. This document describes what is required, and what is considered necessary by the oceanographic and analytical chemistry community to ensure that the analytical results from the nutrient sensors are fit for purpose. It will allow direct comparison between the sensors and other benchtop methods (i.e. using the SEAL autoanalyser, or FIA or simply using a batch method). Additionally, it will allow us to produce, with a high level of confidence, figures of merit for a whole system, rather than individual sensors. This document also combines learning from a workshop on ‘uncertainties in oceanography’ (as part of the SenseOCEAN programme) with the figures of merit document, and provides a step by step guide on how to obtain these figures (limit of detection, limit of quantification, dynamic range, limit of linearity, uncertainty: accuracy, precision and the ‘Nordtest’ approach).
– MULITPARAMETER FLUOROMETER
Various options for the mechanical design and constructing the sensor were considered. One of the approaches outlined in the original project proposal was to consider using pressure-balanced, oil-filled, housings in place of traditional pressure housings. However, this was not compatible with the need for internal refractive optics, filters and UV transparency. It quickly became clear that the optimum solution was to use a titanium housing, both for its mechanical strength and its chemical and corrosion resistance. A novel threaded endcap design was developed that reduces the number of components required to seal the housing
UNDERPINNING TECHNOLOGIES AND/OR DESIGN ELEMENTS OF SENSORS
– ELECTRONICS DESIGN FOR MULTIPARAMETER FLUOROMETER
For the multi parameter fluorometer, a key driver in the electronics development was to generate a single PCB design that could be used for both the UV and Visible variants of the fluorometer. A ‘flexi-rigid’ PCB design was selected to aid miniaturization, reduce manufacturing costs and improve reliability. Direct connections between the various PCBs avoid the need for bulky connectors and custom ribbon cables. It also saves costs in manufacture, as all the electronic components can be populated on a single PCB panel.
There are two additional PCBs: the first provides the absorbance measurement, the second allows different digital interfaces to be selected. The flexi-rigid PCB design folds around the internal optical assembly.
– ELECTROCHEMICAL SI AND PO4
The electrochemical silicate and phosphate sensors are based on a measurement process that avoids the use of liquid reagents. This process is unique in comparison with other sensor techniques for these nutrients. The associated technology and embedded electronic board are contained in a housing that fits the constraints of an ARGO float in terms of energy, mechanical design and pressure resistance. The current design is compliant up to 800 m depth and will be improved to achieve the lower limit of the ARGO float of 2000m. It provides an important step forward to avoid liquid reagents for nutrient measurement opening the scope of use the sensors for long term deployment platforms where maintenance is not possible (ocean mooring buoy, Deep sea platform or ARGO float).
– LAB ON CHIP SENSORS
For the lab on chip sensors, the core control electronics have been updated and the sensor firmware has been entirely rewritten to production standards. The new core electronics are based around a modern Atmel SAM4L microprocessor which incorporates a high performance 32-bit ARM Cortex-M4 RISC processor that is specifically designed for ultra-low power applications. User configuration and control of the sensor, firmware updates and data downloads are all now provided from a single Windows-based graphical user interface application (GUI) which connects to the sensor over USB using standard USB Mass Storage and HID interface classes. This means other Windows applications can communicate with the sensor, in particular Windows Explorer, which can now be used for fast drag-and-drop file downloads from the sensor’s 8-Gb FAT32 MicroSD card. The provision of real-time graphical and textual display of raw sensor data, a new state machine editor, and non-volatile on-board diagnostics are all available via the new sensor GUI. The additional processing performance has also allowed the addition of on-board processing of the sensor raw data to generate calibrated measurement results for all supported chemistries (nutrient concentrations and pH).
On-chip (zero dead volume) latching microvalves have been developed. The valves use a Viton membrane that is mechanically bonded into the chip. The valve is actuated by a miniature stepper motor and requires no power to stay in a particular state (open or closed). The new valves bring significant enhancements in terms of flushing volume and power consumption. Valves previously consumed 14% of the fluid budget in the nitrate sensor, whereas the new valves developed through SenseOCEAN reduce this to <1%
The service life of the lab on chip sensors has been significantly increased by the development of a new type of sliding pump seal. The seals contain an integrated spring which dynamically corrects for wear. The seals also prevent oil ingress when the sensors are deployed deep and cold (the seals were demonstrated to function in the NOC pressure test facility at 600 bar and -1C). A new syringe pump for the LOC systems which has the same performance yet 1/3 of the cost of the previous syringe pump has also been developed.
A cartridge-based reagent changing system designed to allow easy non-skilled changing of reagents in the field has been developed. On a typical deployment, a system would be able to conduct three months of hourly measurements on a single reagent pack. The packs can be prepared in the lab by a technician, and the change can be performed in the field by a non-skilled operative.
For the pH lab on chip sensor, the thermistor was redesigned. A commercial glass encased thermistor with a 4mm long glass shaft was sealed with a small quad ring and mounted on the pH microfluidic chip with a modified ¼ 28 connector. The advantage of this design is that no epoxy is used. Therefore, once the thermistor is mounted it can be positioned next to the microfluidic channel before fully tightening the ¼ 28 connector.
A method of rapid prototyping polydimethylsiloxane (PDMS) chips was developed using our current micro-milling polymethylmethacrylate (PMMA) process rather than using the more labour intensive SU8 moulding technology, which requires specialised clean room facilities. A quick and simple fabrication protocol was developed using a novel PDMS/PDMS moulding technique used to fabricate a negative structured PDMS device from a negative PMMA master.
UNDERPINNING TECHNOLOGIES FOR THE INTEGRATED SENSOR PACKAGE
– ELECTRICAL INTERFACE
Connectors are a key cost associated with oceanographic sensor suites. In SenseOCEAN standard connectors are provided either directly on the sensor or via an adaptor. The use of a standard connector also ensures that sensors are directly compatible with a wide range of existing hardware.
A ring architecture layout was adopted for sensor integration to provide maximum flexibility for adding or removing sensors. Economies of scale are achieved through the use of common cabling components. However, additional connectors are required in the ‘repeating unit’ between sensors, which increases overall costs. This approach was demonstrated successfully with a sensor deployment in Kiel during the summer of 2016.
For many deployments, the number of sensors will be pre-defined, driven by the scientific need for the data output from a particular combination of sensors, as well as ease of integration. In these cases, a potted wiring loom was considered to reduce the number of interconnections and thus reduce the cost of moulded connectors by two thirds.
– MODBUS PROTOCOL AND INTERFACE UNIT
At the beginning of the project a Modbus serial communications protocol was developed, all SenseOCEAN sensors are compliant with the protocol. A central Modbus interface unit was developed to act as the master control providing power and receiving the data from each sensor, each of which has a unique MODBUS address. At first initialisation, the Modbus interface unit polls the slave addresses and creates a list of the connected sensors. This provides plug-and-play sensor discovery, allowing for flexible number and type of sensors to be attached to the integrated sensor suite, enabling a suitable combination of sensors to be installed for different deployment scenarios.
Each sensor outputs a header containing key sensor information including name, manufacturer, model number, serial number, a globally unique identifier (GUID), the number of parameters measured and poll rate. For each parameter measured, parameter name, units and data format are also provided. The Modbus sensor header provides sensor metadata, which ensures full data traceability and supplies all the information needed for open data requirements.
The Modbus interface is also able to broadcast CTD data to all the sensors, when provided by the platform. GPS data can also be integrated from the platform.
A key additional advantage of the Modbus protocol is that it is a recognised standard widely used across different industries, which means the individual sensors can be readily integrated into third-party platforms.
– DATA MANAGEMENT
Data from the Modbus interface unit is logged to an internal SD card or output via RS232, using exchange multi-application protocol (EMAP) to a communications board developed by NKE. When these are integrated e.g. on a PROVOR float, once the float surfaces, data is transmitted over satellite to an Iridium Provider Server. British Oceanographic Data Centre (BODC) can then retrieve files from the Iridium Provider Server using FTP. BODC have developed a standard for the sensor metadata enabling plug-and-play sensor integration. A XML configuration file is transmitted with the data, providing the required SensorML metadata. New parameter mappings, standard names and SensorML have been integrated by BODC into the data management system for SenseOCEAN. BODC will publish ‘five star’ linked data using Open Geo Spatial (OGC)’s Sensor Web Enablement standards.
Providing an internal logging capability also allows the sensor to be deployed for extended periods needing only an external power supply, which significantly reduces the costs for unattended operation.
– INTEGRATION OF SENSOR SUITE ONTO PLATFORMS
CTG designed a sensor suite for deployment on a mooring/observatory. Sensors are supported across three tiers of clamp plates, each mounted along a central steel rod that can be used for profiling. The clamp plates are manufactured in acetyl and have been designed with integral slots to provide flexible regions that can be clamped around each of the sensors, thus minimising the number of separate machined bracket components needed, an approach that has been used successfully on CTG’s FastOcean APD system. The assembly sits on a fourth acetyl ‘skirt’ component.
For the float/AUV suite, significantly stricter guidelines in terms of size, weight and power than the frame-based system suitable for mooring or observatory deployments had to be met. Because of this the number of sensors deployed was restricted. NKE adopted the design principles behind the moored approach, outlined above, for a successful deployment of their PROVOR float in Villlefranche-sur-Mer in the spring of 2017,
DATA
The SenseOCEAN project aimed to enable SenseOCEAN data to be interoperable with operational data flows and international aggregators of data, producing an operational standardised data flow. The state of the art in data technology has shifted significantly since the start of the project with SenseOCEAN making a significant contribution to these developments.
– Uniquely identifying sensors identifiers for sensors
The first challenge addressed within SenseOCEAN was to enable unique machine readable identification of sensors. The solution was for each sensor to include a unique 8 character identifiers with data. This concept is being developed further in a newly forming Research Data Alliance working group to broaden the unique identification to enable globally unique identifiers akin to digital object identifiers.
– OGC standards
From the outset of the project SenseOCEAN worked closely with the Marine SWE profiles community that aim to harmonise and develop common templates for marine data following open geospatial consortium standards. The specific standards used and enhanced in SenseOCEAN are SensorML to describe sensor metadata and Observations and Measurements (O&M) for data These standards are being pursued within the environmental and marine domains on a range of European projects including all four oceans of tomorrow projects, AtlantOS and EMSODev.
At the outset of the SenseOCEAN project the formats were syntactically interoperable with the marine community, within SenseOCEAN we developed semantic interoperability where terms in the documents were constrained to controlled vocabularies. For example, salinity can be described using a precise term on the NERC vocabulary server (http://vocab.nerc.ac.uk/collection/P01/current/PSALST01/) rather than a free-text term (sal, Psal, salin, etc) enabling the record to be machine readable and automatic ingestion into OGC Sensor Observation Service systems such as those developed by the EMSO European Research Infrastructure Consortium.
SensorML templates have been developed for the sensors
– W3C standards
In parallel with the OGC standards the World Wide Web Consortium (W3C) developed a metadata standard called Semantic Sensor Network (SSN). This standard serves a similar function to OGC SensorML with the W3C standards being developed for use in the context of W3C linked. This is a generic standard for standardised metadata on the internet and is used in applications covering sectors broader than the environment and marine.
SenseOCEAN developed templates for the exposure of metadata in SSN format using the same semantic enrichment previously described of OGC SensorML.
This is not the final state of W3C and OGC standards. Current research progressing this standardisation is the OGC Internet of Things standard and this begins to align OGC and W3C standards. An updated SSN standard is also being developed called Sensor, Observation, Sample, and Actuator (SOSA) which will address deficiencies identified in SSN and improve alignment with OGC standards (https://www.w3.org/TR/vocab-ssn/#Developments).
– Serving of data and metadata
BODC developed an ontology fusing together existing available ontologies to enable metadata to be stored in a single database and exposed using both W3C SSN and OGC SensorML metadata standards.
Existing software solutions and adopted where possible as this enables rapid standardisation of results and community development. Software implemented includes:
- 52North software for OGC O&M standard and OGC SOS service
-The US NOAA ERDDAP software, including collaboration with the development team on SeaDataNet formats in ERDDAP enabling trans-Atlantic interoperability. ERDDAP also enables the output of NetCDF data and was used to develop a web interface of creation of SenseOCEAN metadata records.
- W3C linked data software
APPLICATIONS
Within SenseOCEAN, newly developed sensors were used at different locations to demonstrate their performance and usability, as detailed below.
Electrochemical silicate sensor by LEGOS/CNRS and NKE:
- Chilean coast – Mooring
Optodes (O2, CO2, pH) by TU Graz and Pyro Science
- Baltic Sea during PROSID2014 cruise on RV Salme – CTD
- Denmark, Limfjorden – Stand-alone module
- Arctic Fram Strait – TRAMPER (electronic assemblies)
- Arctic Fram Strait – AUV Paul (only pH optode)
- Antarctic, Potter Cove – Stand-alone module
- USA, Californian shelf - AUV
- USA, Californian shelf – Floating buoy
- USA, California, Moss Landing –Abalone tank
LOC (Phosphate, Nitrate, pH) by NOC
- UK, Southampton, Empress Dock - stainless steel frame at pontoon
- UK, Christchurch Harbour (deployment > 18 months)
- USA, Ohio, Maumee River – Flow-through tank
- USA, Chesapeake Bay, Chesapeake Biological Laboratory – Stainless steel frame
- USA, Hawaii, Kaneohe Bay – Stainless steel frame
- USA, Hawaii, Kaneohe Bay – Fish pond in He’eia
- Seychelles – Mooring and CTD
- Arctic Fram Strait – Mooring
- Porcupine Abyssal Plain (PAP) – Mooring and CTD
- Sweden, Gullmarfjord – Stainless steel frame at pontoon
- UK, Bridlington – AUV
- Pacific near Tortel (Chile) – underway ship-based sampling
- Stand-alone device in glacial meltwater streams at Chile (Steffen Glacier) and Greenland (Narsarsuaq and Leverett Glacier)
Electrochemical CO2 by Aarhus University and Unisense
- Sweden, Aarhus, Aarhus University – Laboratory experiments
- Germany, Wilhelmshaven
Electrochemical N2O by Aarhus University and Unisense
- Part of sensor control equipment in about 200 wastewater treatment plants worldwide.
FRRf by CTG and MPI
- Germany, Helgoland – Hypersub observatory platform
- Svalbard, Ny-Ålesund – Hypersub observatory platform
- Germany, Bremen, Max-Planck-Institute – Laboratory experiments
Multi-sensor suite by all SenseOCEAN partners, include a variety of different parameter
- Germany, Kiel - Stainless steel frame at pontoon and CTD
- UK, Southampton, Empress Dock - Stainless steel frame at pontoon
- Mediterranean Sea – PROVOR float
– APPLICATION IN SCIENCE
Within SenseOCEAN newly developed single sensors and the multi-sensor suite were employed to address different scientific questions. The seasonal cycle of nitrate in the upper Arctic Ocean at two sites, one mainly ice covered and the other mainly ice free, was revealed by NOC Lab on chip (LOC) nitrate sensors. Furthermore, SenseOCEAN optode sensor technology from Pyro Science enables monthly measurements of benthic oxygen fluxes with the TRAMPER system. Both trials obtained unique data from the remote Arctic surface and deep sea waters of the Arctic Ocean. This allows us to draw a baseline of these parameters, which is crucial to track climate-change related processes and effects in the vulnerable Arctic Ocean.
Another major topic of the United Nations’ Sustainable Development Goal (SDG) 14 is the detection of the greenhouse gas nitrous oxide (N2O). The improved electrochemical N2O sensor (University of Aarhus & Unisense) provides the ability to study production and depletion processes in the atmosphere and the ocean. Currently, an expedition is scheduled for April 2018 to the oxygen-minimum-zone (OMZ) of Mexico. The main theme of the expedition will be the transformation of N2O in OMZ regions, which will be tackled by fine-scaled in situ mapping of the N2O distribution.
Within SenseOCEAN two pH sensors types were developed, a LOC type by NOC and an optical type by TU Graz/Pyro Science, both suitable for specific purposes. Shorter measurement times and lower detection limits are the major improved factors of these sensors. Thereby, questions regarding ocean acidification, also a major topic of SDG 14, can by tackled.
Eutrophication and related ocean deoxygenation are two more topics of SDG 14, which can be tracked and monitored by sensors developed in SenseOCEAN. Oxygen optodes, developed by TU Graz/Pyro Science are suitable tools to monitor oxygen concentrations in coastal and shelf areas. A two-month long-term deployment of the sensor in the Antarctic Potter Cove (King George Island/Isla de Mayo 25) proved that the device provides suitable data storage space and battery power even under cold and rough Antarctic conditions. The improved FRRf by CTG, measuring fluorescence and by this chlorophyll, proved reliable operation as well under cold and rough conditions in the Arctic Kongsfjorden (Spitsbergen, Svalbard). Data from this sensor can be used to calculated primary production within a daily resolution and by this provide reliable data to ground truth satellite-based estimations. The LOC sensors for phosphate and nitrate are suitable tools to detect eutrophication and monitor recovery. The LOC nitrate and phosphate sensor were given an honourable mention in the recent Alliance for Coastal Technologies (ACT) nutrient challenge, performing well in freshwater, estuarine and marine systems over long periods
The CO2 concentration is a parameter of great interest for studies of climate change, ocean acidification and eutrophication. An optical CO2 sensor was developed by TU Graz/Pyro Science and a STOX type by University of Aarhus/Unisense. A few years ago, CO2 measurements in the aquatic realm were impossible. Therefore, these developments are a huge step forward. For example, the STOX CO2 sensor has been used for analysing the exchange between seawater and atmosphere during a campaign in Wilhelmshaven. Miniaturized setups will help in a variety of eco-physiological studies such as calcification of organisms (reefs, mussels, etc.) under acidified conditions.
Additionally, the newly developed sensors and particularly the multi-sensor suite can be used to obtain essential ocean variables (EOV) for the General Ocean Observatory Strategy (GOOS) and for the Deep Ocean Observatory Strategy (DOOS). In particular, the multiparameter optical sensor by CTG and the multi-sensor suite are able to measure a set of important parameters in parallel, allowing easy direct comparison of different parameters.
Other research fields will also benefit from our sensor developments. For example, LOC and optode sensors were used in an aquaculture pool and an abalone tank (Haliotis sp.) for monitoring tasks proving to be useful tools for aquaculture research. Measurements of O2 and pH are already commonly used in aquaculture research, however, CO2, nitrate, nitrite, ammonia and N2O would also be of great interest in this field. Due to the extremely small tip size of the electrochemical CO2, O2- and pH microsensors, they are able to measure e.g. inside primary productive tissues and by that give a deeper insight into primary production processes and the balanced/unbalanced production of oxygen and uptake of carbon dioxide. All the sensors are useful in biogeochemical studies, dealing with regional and global element cycling and offset phenomena like eutrophication or upwelling.
– APPLICATION IN INDUSTRY AND OTHER FIELDS
In addition to scientific research, the newly developed sensors can also be of great benefit for industrial use. Similarly to the benefit that the sensors can provide in aquaculture research, industrial aquafarms also use sensors to monitor health and productivity of their products. Industrial fish farms commonly use O2 sensors but again CO2, nitrate, nitrite, and ammonia would be of greater interest. Algae farms, like the Swedish algae factory (http://swedishalgaefactory.com/) also use pH sensors. Their productivity could benefit from the continuously monitoring of nitrate, nitrate, phosphate, ammonia, silicate, CO2 and trace metals (also developed by NOC, but at lower TRL levels at the moment). Our N2O sensor is already used in over a 100 waste water treatment facilities.
The CO2 and the N2O sensors will have a widespread application in medicine, when improvements to measure reliably within air are complete. In the field of anaesthesia, measurements of these parameters in air are of particular interest.
Furthermore, our sensors were proven to work reliably for long durations, even under rough Arctic and Antarctic conditions. The fluorometric sensor from CTG can be used to detect water pollution and petroleum leakages via the measurement aromatic hydrocarbons; aromatic hydrocarbons are important monitoring parameters in the shipping industry. Hence the SenseOCEAN suite of sensors are perfect tools for environmental monitoring tasks performed by, for example, NGOs, environmental protection agencies or nature conservation authorities.
Potential Impact:
Our world is facing major environmental problems that affect us all. These include global warming and climate change, pollution, ocean acidification, ocean oxygenation, eutrophication and loss of marine biodiversity. In turn, these can lead to fish kills, food contamination (e.g. harmful algal blooms), sea level rise (resulting in loss of habitable land) and a loss of utility of the marine system through impacts on tourism (e.g. fish kills and harmful algal blooms).
The oceans provide vital ecosystems services to humans, and these services are currently under multiple stressors that are both increasing and changing, creating complex, often unpredictable feedbacks. Defining acceptable targets for ocean health and sustainability, establishing the knowledge base needed to maintain and improve the health of oceans and developing systems to predict and respond to shocks to and from oceans all represent critical research needs. A multi-scale integrated transdisciplinary approach combining models, observation systems, analytics, experiments, and societal needs is needed to create the knowledge required to map pathways and identify trade-offs in conserving ocean health.
Climate change is one of the greatest threats to the planet, and the marine system is especially sensitive. Major contributors to global warming are the greenhouse gases (GHGs), carbon dioxide (CO2) and nitrous oxide (N2O). Carbon dioxide receives most attention due to its relatively high concentrations, however N2O though currently at levels that are approximately 1000-fold lower than CO2, has a potential global warming effect of >260 times that of CO2. Marine areas can be sources or sinks for both CO2 and N2O and both gases exchange between the atmosphere and the ocean through physical and biological processes. It is therefore necessary to be able to measure both of them to understand these interactions. Within SenseOCEAN there has been development of sensors for both CO2 and N2O, using a number of technological approaches.
The oceans store a substantial amount of carbon dioxide as dissolved inorganic carbon (DIC), which buffers increases in atmospheric CO2 concentrations with the associated consequences for our climate. The large quantities of anthropogenic CO2 taken up by the oceans has resulted in a decrease in seawater pH in a process termed ocean acidification. With increasing atmospheric CO2 levels, the capacity of the oceans to absorb CO2 is decreasing whilst ocean waters reach higher DIC concentrations, and global warming reduces the capacity for seawater to hold as much CO2. The ability of the oceans to sequester CO2 over geological timescales (100s to 1000s of years) depends on the role of physical and biological (soft tissue and carbonate) pumps. These pumps remove carbon from the upper ocean and take it to greater ocean depths, both are influenced by climate change in ways that are currently poorly understood. Against the background of a gradual anthropogenically driven increase in oceanic DIC, there are pronounced natural fluctuations and poorly constrained climate change driven variations in the functioning of the physical and biological pumps. Improved observational capabilities, involving high quality carbonate chemistry measurements, are required to enhance our mechanistic and quantitative understanding of the influences of ocean warming and acidification on oceanic carbon sequestration.
Oceanic uptake of CO2 has a range of consequences for individual biological species and communities. The decrease in carbonate ion concentrations lowers the saturation states of calcite and aragonite, minerals used by a range of organisms for building protective shells (calcification). Ocean acidification can reduce calcification and thereby ballasting of sinking organic matter, and be detrimental for organisms that lack protective mechanisms. Some regions of the oceans (e.g. polar oceans, nutrient enriched coastal waters) are experiencing pronounced reductions in calcium carbonate saturation states, with potential consequences for ecosystems. It is therefore particularly important to monitor carbonate chemistry in these ocean regions which are most vulnerable to ocean acidification.
Ocean deoxygenation is affecting both coastal regions and the open ocean and has implications for oceanic productivity and marine ecosystems. Warmer oceans resulting from climate change, lead to oxygen loss due to the reduced solubility of oxygen, and increased stratification, which inhibits the photosynthetic production of oxygen. In coastal zones, increased nutrient loading (e.g. from agricultural and industrial run-off) can lead to eutrophication (see below) that also causes oxygen depletion. The impacts of deoxygenation on ecosystems and ecosystems goods and services are numerous. Living organisms are impacted at a functional level below certain oxygen concentrations; this varies on a species by species level so it is extremely difficult to define a unique hypoxia threshold. In addition, warming and acidification will act in antagonistic ways to enhance the negative impacts of low oxygen levels. Shoaling of hypoxic waters can lead to cascading effects on benthic and pelagic ecosystems, including marine habitat compression and changes to community structure. Hypoxic events also have severe impacts on goods and services, for instance they may lead to massive fish kills in aquaculture farms which can result in large economic losses.
Eutrophication is a type of pollution caused by an excess supply of nutrients to natural waters, including fresh and marine waters. The main nutrients implicated in eutrophication are the macronutrients, nitrate and phosphate, that are used by plants in high concentrations for growth. Humans are responsible for adding high concentrations of these elements to our waterways from activities such as sewage disposal (nitrate and phosphate), atmospheric inputs (mainly nitrate) due to high temperature combustion processes and run off from agricultural use of fertilizers (nitrate and phosphate). Enhanced concentrations of nutrients in marine systems promote the rapid growth of phytoplankton, and changes in the species composition of the planktonic community. The rapid growth of phytoplankton is known as a bloom and phytoplankton blooms can have significant impacts in both marine and freshwater systems. Deleterious effects include the development of harmful algal blooms (HAB’s) and anoxia. Anoxia occurs due to a massive uptake of oxygen resulting from the microbial decomposition of the phytoplankton as they die and sink. Furthermore, strong blooms result in light shading which means plants below the bloom in the water column are unable to grow due to the lack of sunlight. Harmful algal blooms can impact aquaculture and fisheries (commercial and recreational), and the toxins produced by the phytoplankton can also lead to bathing restrictions and/or closure of beaches.
The task of understanding the global ocean system in order to monitor, comprehend and mitigate these environmental problems is enormous, and can at first seem intractable. The challenge is to measure a wide range of chemicals, processes and other factors, and to measure them on spatial and temporal scales (i.e. frequently and at closely spaced intervals over the whole ocean) that until recently would have been hugely cost prohibitive. In addition, the driving factors behind these global impacts are still poorly understood and we are often limited by our inability to measure the substance and/or process required to increase our understanding.
The traditional approach to obtaining ocean measurements is inherently costly, labour intensive and provides limited data; ocean going research vessels, with complements of marine staff and research scientists, sail to a single point in the ocean and take a series of measurements over a very short timescale (generally a week to a month). Not only is this expensive and technically challenging for some measurements, but there are regions of the ocean that are inaccessible for large periods of the year such as the Arctic and Antarctic oceans.
We are on the cusp of a more integrated ocean observing system, driven in large part by technological developments in autonomy, communications, battery technology and satellite observing. A large part of this technology push is around the platforms used for observing: shallow and deep fixed point observatories (e.g. the Porcupine Abyssal Plain Observatory), moorings, autonomous profiling floats (e.g. there are more than 3800 ARGO floats in the world’s oceans), established ship observation lines (e.g. GO-SHIP), vessels of opportunity, arrays of gliders, autonomous underwater vehicles (AUVs), and benthic crawlers and landers.
A key consideration in using such platforms is they are limited in terms of space, power, data handling and servicing. Therefore, the sensor technology deployed on these platforms must be small, energy efficient, require minimal servicing and be able to store/transfer data. In addition to satisfy the need for measurement at an increased temporal and spatial resolution the sensors should be low cost and ideally, mass producible.
SenseOCEAN has significantly advanced our ability to measure specific parameters related to of some of the major environmental problems facing us, as well as enabling these measurements to be obtained at a greater frequency and on a wider scale.
SenseOCEAN has developed three methods to measure carbon dioxide, the most important greenhouse gas and a key component of the marine carbonate system. They are, electrochemical microsensors, lab on chip sensors and optodes. Having several methods available means that there is sensor ideally suited for different environments. The electrochemical sensors have good sensitivity at partial pressures close to atmospheric and in addition to measuring CO2 in seawater are capable of analysing CO2 within sediment and plant tissue, facilitated by their microscopic tip design. The CO2 optodes can be deployed for several weeks with the optoelectronics contained in very compact pressure resistant housings, these are very cheap sensors. The lab on chip sensors have in situ calibration techniques and are also deployable for extended periods of time.
Nitrous oxide, also an important greenhouse gas, can be measured using an electrochemical microsensor which has been significantly modified and improved. The sensors can be used over a wide concentration range from nanomolar to millimolar concentrations making them applicable to a wide range of environments. They can analyse continuously for several months as has been demonstrated and they have been effectively deployed as part of a monitoring system for at over 100 wastewater treatment plants.
To fully constrain the marine carbonate system, you need a combination of two of the 4 carbonate system variables; Dissolved inorganic carbon (DIC), pH, the partial pressure of CO2 and Total alkalinity. The measurement of pH is vital for assessment of ocean acidification and understanding the marine carbonate system. SenseOCEAN has two sensors for this purpose, the pH optode and the lab on chip pH sensor. Both sensors have been demonstrated successfully in several field deployments including shallow coastal moorings, on autonomous underwater vehicles, on CTD frames and benthic crawlers. The development of in situ sensors for DIC and TA is less advanced, with only a few reported in situ instruments, relying on spectrophotometric pH or infra-red CO2 measurements. SenseOCEAN has started to develop a lab on chip TA and DIC sensor that uses a gas diffusion and conductivity measurement principle. Although at a lower technology readiness level compared to the lab on chip pH sensor (4 versus 7), once increased this will provide a full suite of carbonate system measurements.
Ocean deoxygenation requires measurement of the dissolved O2 in water. The optode oxygen sensors developed within the SenseOCEAN project will make the reliable long term monitoring of oxygen concentrations possible. Oxygen optodes rely on quenching of luminescent indicator dyes by oxygen and are characterized not only by the absence of electromagnetic interferences and analyte consumption, but also by their small size and comparatively low cost. In the stand-alone module (i.e. with optoelectronics and in a housing) developed in the SenseOCEAN project, we are now able to take autonomous long-term measurements for up to 2 years at depths down to 600 m. The deep-sea version is extremely compact and allows parallel measurement with the other SenseOCEAN developed sensors (using the MODBUS communication protocol). Importantly, the read-out modules are fully compatible with other “sensing chemistries” including pH and pCO2 developed during SenseOCEAN and also trace oxygen optode sensors suitable for precise oxygen quantification in low oxygen zones.
In addition to dissolved oxygen, the assessment and monitoring of eutrophication requires the measurement of dissolved nutrients (primarily nitrate and phosphate) concentrations (high concentrations can act as an early warning system for the onset of phytoplankton blooms) and chlorophyll a (gives an indication of algal biomass). Combined with these single indicator measurements the ability to monitor the composition of the phytoplankton community is also important in order to detect any changes that may indicate a community shift towards species that produce toxins, and thus generate a harmful algal bloom (HAB). SenseOCEAN has developed technologies to measure the compounds that the cause the blooms (nutrient concentrations) and the changes in community structure that can indicate a shift to a HAB.
SenseOCEAN has taken two approaches to nutrient measurements, wet chemical lab on chip based systems and reagentless electrochemical methods. The lab on chip approach basically miniaturizes the classical benchtop wet chemistry techniques for the determination of nitrate, nitrite and phosphate, the electrochemical techniques do not use reagents and are applicable to the measurement of silicate and phosphate. Both approaches have strengths and weaknesses and these are more or less important depending on the type of environment they are deployed in and the platforms on which they are used. During SenseOCEAN we have deployed sensors of both types in a range of environments including the tropical Pacific, a eutrophic harbour and a fjord system.
The direct effects of eutrophication can be measured using our multi parameter fluorimeter (V-Lux); by using multiple wavelengths to target common algal pigment groups, changes in algal group composition can be detected e.g. at the onset of a HAB. In addition, the FRRf (fast repetition rate fluorometer) has been widely tested in the marine environment and was deployed on an observatory in the Arctic during SenseOCEAN for over 12 months. The ability to determine community shifts in the phytoplankton community was clearly demonstrated as the year progressed.
To obtain measurements on a wider scale and at a greater frequency, SenseOCEAN has addressed the cost and ease of sensor manufacture. Ensuring that the sensors are low cost and can be produced on an industrial scale reduces barriers to mass deployment of the sensors. Once deployed in the real environment there are factors that can limit the time the sensors can be deployed, these can be fouling by marine organisms that can block detectors and flow through systems, and factors related to the breakdown of reagents used in those sensors that use chemicals. SenseOCEAN has dedicated some effort on both of these factors; biological fouling and reagent lifetime.
With an aim to reducing the cost of the manufacture of sensors we explored state of the art approaches to the whole development process. In the case of the development of our multiparameter fluorometer we were able to reduce costs and ease of manufacturing by adopting one design for several variants of the instrument. Alongside this Chelsea Technology Group has reduced costs by adopting a ‘flexi-rigid’ PCB (printed circuit board) design for their V-Lux multiparameter fluorometer, which folds around the internal assembly of the fluorometer. This aids miniaturisation (no need for bulky connectors), improves reliability and reduces manufacturing costs as all the electronic components are placed on a single PCB panel. The production of electrochemical microsensors was traditionally done in a bespoke manner, with each sensor requiring a specialised technician to hand manufacture each sensor; Aarhus University have simplified their N2O electrochemical microsensor to a 1-compartment design, the new design has enabled Unisense to employ microfabrication techniques, which means the sensors are more robust, manufacturing is consistent and at a much lower cost. The optodes developed by TU Graz and Pyro Science have been engineered to enable read out with a standardised optoelectronics unit, this is done by chemically engineering the spectral properties of the sensing materials used in their optode sensors, thereby minimising manufacturing costs. In addition, a miniaturised polymer housing for the shallow water standalone optode system (including data logger and battery) has been developed, this has reduced the volume of the device by 75% compared to the prototype.
Lab on chip based sensors were traditionally manufactured I a bespoke manner, the versions developed in SenseOCEAN use a standardised platform approach, with the adoption a new manufacturing press plate technique that allows the parallel pressing of multiple devices. New optical component alignment procedures have been implemented which are faster, and scalable, this ensures that this step does not slow the production process. A new syringe pump was also developed with the same performance as those that were traditionally bought ‘out of house’, but at a third of the cost. Looking to the near future a potentially cheaper, more efficient valve has also been identified and is undergoing testing.
Many of the issues that have previously restricted the length of in situ sensor deployments have been addressed by SenseOCEAN. Some of these issues have also restricted the location of long term deployments as frequent need for servicing does not lend itself to remote deployments. For example, the electrochemical nutrient sensor is reagentless, as are the optodes and electrochemical microsensors. The lab on chip nitrate sensors have been improved to reduce reagent use by using a custom micromixer channel. Preservation techniques to prolong reagent life for the LOC sensors have been evaluated with alginate gelification in combination with lyophilisation showing promise as a long term in situ reagent storage solution.
We have addressed the perennially taxing issues around biofouling in order to increase deployment times. On the newly developed optode systems as have designed copper guards that sit around the sensing spot on the optode, this copper kills microscopic organisms and prevents fouling. On other sensors we have developed an In situ chlorination technique and have put effort into exploring the use of ultrasonic acoustic vibration for anti-biofouling. A patent is pending for the in situ chlorination method and the ultrasonic method has been published, showing promise for future application. The optical system developed by Chelsea Technologies integrates a ultra violet flash system to kill fouling organisms on their optical windows.
Of course, designing and producing sensors, carrying out deployments and making measurements is pointless if the data is not made available so that scientists and stakeholders (governments, policy makers, etc.) can analyse and use the data. Equally important is that the users can be assured of the source and quality of the data, including calibration data and any quality assurance processing that may have occurred. Although global standards for marine data do not yet exist, SenseOCEAN has invested time around issues concerned with data availability, quality assurance and standardisation. Every sensor is uniquely identifiable (and the concept is being developed to enable globally unique identifiers), data formats follow standards that are already available and existing software solutions have been adopted. SenseOCEAN has been involved in, and continues to work on, establishing globally accepted standards for marine sensor data and meta data. The Modbus interface receives standardised data files from the SenseOCEAN developed sensors which, via the developed EMAP communication board can be transmitted using the Iridium satellite network.
DISSEMINATION & EXPLOITATION OF RESULTS
The SenseOCEAN project has been promoted at events throughout the four years of the project. Some of the events have also involved other Ocean of Tomorrow (OoT) projects such as the OoT stand at Oceanology International in March 2016 (London, UK), which was organised and coordinated by the SenseOCEAN project.
The SenseOCEAN coordinator has given several presentations to audiences including policymakers (e.g. UK government, Spanish Council, European Commission). A presentation at the EuroScience Open Forum (ESOF), in Manchester, UK, in July 2016 on how our sensors address issues in the marine strategy framework directive and plans to commercialise the sensors was well attended and well received. At an IOCCP (International Ocean Carbon Coordination Project – a UNESCO) side event at the 10th International Carbon Dioxide Conference in August 2017, the coordinator gave a presentation and lead a discussion on using autonomous sensors. In July 2017 he gave a presentation at the European Commission event ‘A New Era of Blue Enlightenment’ in Lisbon, Portugal on sensors and sampling for Atlantic Ocean observations. The COLUMBUS project has also invited presentations from the SenseOCEAN coordinator at two of their events, a brokerage event and the 2nd Annual Conference.
We have also had television and radio coverage of the project. Back in 2015, Matt Mowlem was interviewed about SenseOCEAN for BBC Radio 4’s Inside Science series, resulting in a 12-minute segment during the programme. More recently Euronews filmed some of our partners in the UK and Denmark, showcasing sensors and technology from the project for a segment devoted to SenseOCEAN in one of their ‘Futuris’ series programmes. The programme was broadcast repeatedly for a week on the Euronews TV channel and made available of their website and (www.euronews.com) and Youtube channel in 12 languages.
The SenseOCEAN coordinator and other members of the consortium (e.g. Tellabs and CTG) have contributed to Twitter, Our Policy Brief was tweeted with #SaveOurOceans to coincide with the UN Ocean Conference in June 2017.
Project partners, especially our SMEs have presented their sensors and work at many trade fairs, exhibitions, workshops and meetings aimed at the industry and commercial sectors. In addition to the Ocean of Tomorrow stand at Oceanology 2016, nke also exhibited at the event. Nke also presented the SenseOCEAN project at the Seatechweek SSCO Conference in Brest, France in 2014. Tellabs and Pyro Science have exhibited at the international Analytica trade fair in both 2014 and 2016 (Munich, Germany). Chelsea Technologies Group regularly participate in the annual UK Marine Measurements Forum (MMF) which has members from the academic, industry and public sectors. They announced their new V-Lux fluorometer, developed as part of SenseOCEAN at the 59th MMF meeting before the official launch at WEFTEC 2017 in October in Chicago, USA. Unisense also exhibited their Field Datalogger Mini at WEFTEC 2017 as well as at FEMS 2017 in Valencia, Spain and the 2017 ASLO Aquatic Sciences Meeting in Hawaii.
SenseOCEAN work has also been widely disseminated at renowned international conferences as well as at more specialised technical conferences. SenseOCEAN, NeXOS, SCHeMA and COMMONSENSE convened a joint session at the 2015 ASLO Aquatic Sciences in Granada, Spain where SenseOCEAN scientists presented some of their early results. Talks were also given at the 2016 ASLO Ocean Sciences Meeting in New Orleans, USA. One of the leading international geochemistry conferences is the Goldschmidt Conference. Presentations were given at both the 2016 and 2017 conferences by SenseOCEAN partners. Presentations from SenseOCEAN scientists were also given at the EGU General Assembly in Vienna Austria, in 2016 this conference had on the order of 15,000 attendees.
Specialised technical conferences where SenseOCEAN work has been presented include the 2016 EUROPT(R)ODE XIII Conference on Optical Chemical Sensors and Biosensors (Austria), International Electromaterials Science Symposium 2017 (Australia), the International Conference on Electrochemical Sensors 2017 (Hungary) and the Electrochemical Society Meeting (USA).
To date, SenseOCEAN has published seventeen peer-reviewed papers. A joint publication on the results from the test deployments of the integrated multifunctional sensor package is in preparation and a joint publication on societal impacts and UN SDG goal 14 and how SenseOCEAN technology can address these is also planned.
The development work and results generated as a consequence of the SenseOCEAN project have resulted in 6 patent applications, three in Denmark, two in the UK and one to the World Intellectual Property Organisation. They are related to the electrochemical microsensors and the lab on chip sensors. Three products are already on the market, the Unisense Field Datalogger Mini, the Chelsea Technologies Group V-Lux Fluorometer and the Unisense N2O Microsensor. These will undoubtedly be followed by further products from Pyro Science and products from the commercialisation of the lab on chip technology.
List of Websites:
www.senseocean.eu
Coordinator: Douglas Connelly
National Oceanography Centre, Southampton, UK
douglas.connelly@noc.ac.uk