Covid-19: Biosensors, BioFETs and Labs-on-a-ChipFollow article
The ESA-backed B-LiFE mobile field laboratory set up at Piedmont is enabling the Italian authorities to test thousands of key workers for COVID-19.
Crucial to monitoring and forecasting the spread of a virus during a pandemic is the acquisition of accurate infection data. Traditional manual laboratory techniques have proved to be inadequate during the Covid-19 pandemic which has left forecast models starved of real data. As a result, the preferred model used by the UK government provided an estimate of deaths due to the virus in the range from 20,000 to 500,000. With no other information, the ‘worst-case’ scenario was accepted and a full ‘lockdown’ ordered to avoid risk of the NHS being totally overwhelmed. So, two months on from my previous post on Covid-19 testing, have things improved? At the time of writing, traditional manual ‘test and trace’ methods involving thousands of trained staff are having to be used, following the withdrawal of the NHS smartphone app due to technical problems. It won’t work properly with iPhones. Fortunately, technology has a lot more to offer in the fight to halt the spread of disease. First and foremost, what’s needed is a portable instrument which can carry out a non-invasive test on someone in the field and return a result in real-time. Progress has been slow, but scientists and engineers are getting there.
Automated test analysis
Tests that produce reliable results (very few false positives/negatives) when attempting to detect incredibly small quantities of a disease ‘marker’ in a given sample of say, saliva, can involve many processes and take hours to perform. The most popular method for Coronavirus RNA is called – deep breath - Reverse Transcription Polymerase Chain Reaction or RT-PCR. This involves a process to convert any RNA present to DNA followed by ‘amplification’ where the trace DNA is copied many times until sufficient exists to be detected. Amplification requires thermal cycling and can take hours. There are many methods of detection available, each with their own advantages: a popular one consists of adding a special coloured dye which fluoresces green in the presence of the DNA. RT-PCR is just one of many possible ‘manual’ tests that need a well-equipped laboratory with skilled technicians. Recent advances in automation have led to the production of machines such as the SAMBA II that do most of the work and can be operated on the hospital ward by personnel with limited amounts of training. These are amazing instruments, but still a long way from providing an instant, reliable result in the field.
Biosensors and Biometrics
What we really need is a portable diagnostic unit the size of a smartphone that can provide an analysis of a biological sample without the need for elaborate laboratory equipment and exotic chemicals. Something like the fictional Tricorder from the Star Trek TV series. A Biosensor is required which can measure some biological parameter, turning it into a form which can be processed by electronics. The function of the green boxes in Fig.1 will be readily understood by any electronics engineer: the Biorecognition Element or Bioreceptor will mystify all but experienced biochemists. The bioreceptor is the key component and must be tailored precisely so that when a particular biological or chemical entity is presented to it, usually in a liquid sample, it produces a unique recognition output which the transducer converts to an electrical signal for the electronics.
Because we are talking about virus detection, the biorecognition element in the diagram consists of a structure coated in antibody receptors for the relevant antigen DNA. In this case, the binding of the antigen to the receptor causes a variation in the electrical charge on the surface of the transducer, so providing a ‘virus detected’ signal. One snag with this type of detection surface is that it cannot be reused (not yet anyway). Some systems are suitable for continuous sensing: elements based on an enzyme-triggered catalytic chemical reaction work because the enzyme is not consumed. The design of recognition elements for specific applications like Covid-19 is the most difficult part of biosensor design and takes the most development effort [1, 2].
Not long after the Field-Effect Transistor (FET) was invented in the 1960’s - in many ways a superior form of the Bipolar Junction Transistor (BJT) - its special properties were harnessed to make a device that combined both chemical/biological detector and transducer functions – the BioFET. There are many experimental variations of hardware, but the basic arrangement for a biological receptor type is shown in Fig.2. This diagram illustrates the principle of operation and not the precise design of a particular chip. You can see that the structure is very similar to that of a MOSFET with the Gate electrode replaced by a reference voltage probe in an insulated ‘tank’ containing an aqueous solution (sample dissolved in pure water). The biorecognition element lies at the bottom between the sample solution and the gate insulator.
The operation of an enhancement-mode MOSFET is quite simple: with zero Gate voltage, the semiconductor channel between the Drain and Source does not conduct: ID = 0. When a large enough voltage is applied to the Gate pin, an electric field is set up across the Gate insulator which causes the channel to conduct and a current ID to flow with a value dependant on the voltage VDS applied between the Drain and Source pins. Hence the name, Field-Effect Transistor. A Bipolar transistor is current-controlled, meaning that a measurable current has to flow in the Base in order for a much larger current to flow between its Collector and Emitter terminals. Being voltage-controlled, the FET sees only a very tiny current flowing in the Gate – it is after all, attached to a block of insulating material!
It's very high Gate impedance makes the FET very sensitive to tiny changes of input and a much better choice than the BJT for a biosensor. Looking at Fig.2 again, notice that we now have a fixed Gate voltage or Reference VG which sets up a ‘bias’ electric field across the dissolved sample and bioreceptor. Before using the test sample, the device is calibrated with pure water to get a base value for ID. When the solution with the DNA sample is used instead, any target elements present will bind to the receptors, changing the electric charge on the bioreceptor base. This will modify the electric field set up by the reference and bring about a corresponding change in ID. It’s the measurement of the change in current that makes this system so sensitive. Any transducer trying to detect an absolute value, no doubt in the presence of electrical noise is going to have a tough time, leading to a high proportion of false positive/negative results. It’s not easy creating a working biosensor: this student thesis explains some of the trials and tribulations in some detail .
Wearable diagnostics and tracking
Early biosensor devices were only Biometric in function, measuring basic things like temperature and pulse rate, even electrical signals in the brain. Nowadays, most people assume the term ‘Biometric Data’ refers only to fingerprints, facial dimensions and other parameters associated with security. Actually the ‘fitness band’ or smartwatch on your wrist contains biometric sensors to monitor the body’s response to physical exercise.
A much more ‘serious’ biometric medical instrument has come into common use recently, partly because of the Covid-19 pandemic – the Pulse Oximeter. It measures the level of oxygen in your blood simply by beaming light from two different-coloured LEDs into a finger and measuring how much of each is absorbed. The next generation of Apple watch will be able to provide this information using the same method. Pulse Oximeters are widely available and thanks to the simple electronics, inexpensive. Any Covid patient in hospital will have one fitted and they’re even recommended for home use. The popularity of the device is due to a mysterious but potentially deadly effect of a Covid infection called Happy Hypoxia. A sufferer’s blood oxygen level can drop dangerously with little outward sign, leading to organ failure and death unless it’s treated quickly.
A smartwatch system that can actually diagnose a Covid infection is currently being evaluated. It combines biometric sensing with AI software to provide an early warning to the wearer. This could potentially have a big impact on stopping the spread of a virus, because the patient would begin self-isolation during the no-symptom but very infectious phase of the disease. If everyone wore one of these devices, with the information sent automatically to a central health authority, then we would at last obtain sensible data for the forecast models…. If the privacy issue can be handled.
The next logical development of automated (‘wet’) sample testing is to combine multiple processes on to a single ‘chip’. Of course, that means finding a mechanism for moving tiny amounts of liquid around inside the device between processes and sensing elements. The need for fluid transportation has led to a whole new engineering discipline – Microfluidics. If this technology becomes practical and reliable, then maybe a simple Tricorder is not so far away.
Despite the huge amount of research effort being applied, we are a long way from developing the universal sensor required to make a ‘real’ Tricorder. The field of Artificial Intelligence has the same challenge: engineers are trying to create practical solutions before scientists have gained sufficient understanding of the problem. In the meantime, mobile testing units like this one, called B-LiFE supported by the European Space Agency, with its satellite communications and Earth observation facilities for outbreak tracking may help limit the spread of the next pandemic.
 https://par.nsf.gov/servlets/purl/10124972 Guide to Selecting a Biorecognition Element for Biosensors
 https://www.mdpi.com/2079-6374/8/2/35/htm Sensors Based on Bio and Biomimetic Receptors in Medical Diagnostic, Environment, and Food Analysis
 http://uu.diva-portal.org/smash/get/diva2:1372483/FULLTEXT01.pdf Field-effect transistor based biosensing of glucose using carbon nanotubes and monolayer MoS2
If you're stuck for something to do, follow my posts on Twitter. I link to interesting articles on new electronics and related technologies, retweeting posts I spot about robots, space exploration and other issues.