Your body is constantly emitting thousands of different molecules. If each small molecule were a musical note, then every person would be a walking symphony. That music carries rich information about our health; you just need to be able to hear it. Materials with nanoscopic pores called metal-organic frameworks (MOFs) can be the ears for that music. By harnessing the superior adsorptive power of MOFs, we can create an electronic nose that will provide enormous socioeconomic benefits, especially in transforming healthcare.
The original super smeller
Until now, the best ear for that job — the most polished conductor, if you will — came from our canine friends. Dogs have up to 300 million olfactory receptors in their noses (compared to a mere 6 million in humans), giving them a sort of superpower smell that enables them to sense things in the environment otherwise out of reach.
For this reason, dogs are used in a variety of settings for their superior olfaction — including search and rescue, law enforcement, and, most recently, healthcare, sniffing out diseases.
The volatile organic compounds emitted by our bodies, whether emanating from our breath or the sweat on our skin, contain biomarkers for a wide range of illnesses, from various cancers to degenerative diseases such as Parkinson’s. Much of our knowledge about these biomarkers comes from clinical studies that use dogs’ smell to reliably detect various health conditions, in many cases identifying illnesses much earlier than any other known diagnostic technique.
Although the olfactory capabilities of dogs are impressive, they are nevertheless impractical for widespread, reliable health diagnostics. Rather, the noses of dogs are proof that it is possible to diagnose health via smell.
For example, in 2004, a research group publishing in BMJ presented the first robust clinical evidence that dogs could detect cancer in patients. During this study, the researchers trained dogs to detect bladder cancer by smelling urine samples from patients and healthy controls. Since then, several other studies have demonstrated that dogs can identify other cancers and a wide range of diseases.
Although the olfactory capabilities of dogs are impressive, they are nevertheless impractical for widespread, reliable health diagnostics. Rather, the noses of dogs are proof that it is possible to diagnose health via smell. As such, these discoveries have since sparked research into the use of electronic noses for biomedical applications.
Early efforts struggle
The first studies on “machine olfaction” began in the early 1960s, but the term “electronic nose,” referring to an array of chemical sensors combined with a pattern recognition system to identify gases, did not come into wide use until the late 1980s. Early applications for electronic nose research came from the food and beverage industry — replacing unreliable panels of human expert coffee/beer/wine/fragrance sniffers — as well as the military, so as not to endanger dogs who were detecting explosives.
Funding from private industry and the military led to significant advances in the development of electronic noses, notably Nathan Lewis’ Caltech Electronic Nose Project, which was based on arrays of conducting polymeric sponges. While these efforts achieved partial success in the form of some commercial e-nose products that could detect very specific VOCs in narrow operating conditions, they were still no match for the abilities of dogs.
Just as your ear has thousands of little hairs that are precisely tuned to different sound frequencies, the pores of MOFs can be precisely tuned to each kind of small molecule emitted by our bodies.
Combining advanced tech for a new approach
For me, the marriage of MOFs and e-noses is the logical next step. The ability of MOFs to selectively adsorb gases makes them ideal candidates for creating good gas sensors.
While many groups to date have succeeded in making gas sensors that use one MOF to detect one specific gas, my research group is working to create a sensor with multiple MOFs to detect multiple gases. Remember, a dog’s nose uses 300 million olfactory receptors. It is thus unlikely that an electronic nose with less than 30 sensing elements can match a dog’s nose any more than a digital camera with less than 30 pixels can match a human eye.
Going back to the music analogy: Just as your ear has thousands of little hairs that are precisely tuned to different sound frequencies, the pores of MOFs can be precisely tuned to each kind of small molecule emitted by our bodies. Prior approaches for e-noses used materials with randomly shaped pores to absorb those molecules, which obscures the health information — muddles the music, if you will.
My team is taking a different approach, designing large arrays of MOFs by using computational data science and engineering methods to identify the precise and optimal combinations of MOFs. We have already created surface acoustic wave (SAW) sensors deposited with porous MOFs, showing that the entire device can be predictively modeled to selective adsorption. By finding the right number and combination of MOF-SAW sensing elements, we are working to create an array to form a broad-spectrum portable VOC sensor for health.
Such a sensor could one day be used in doctor’s offices and hospitals to diagnose diseases prior to using other techniques. Lung cancer is one of our first targets, as there is a large body of literature on the VOCs emitted by lung cancer patients. Working with clinicians at the University of Pittsburgh medical school, we are aiming to use patients’ breaths to test our arrays and start building prototypes.
Broadening the landscape with more MOFs
While we are starting with one disease state at a time, the hope is to eventually replace a trained dog with a sophisticated gas sensor that can work across a broad spectrum of diseases. We envision a future where with one breath, you can get a health diagnosis — imagine breathing into your phone to get a continuous health readout, like a FitBit for the breath.
As is the case throughout the MOF field, the biggest obstacle to making our MOF-based e-noses a reality is the availability of many different MOFs at scale. We ultimately need hundreds of different MOFs available at a large scale to enable the mass manufacturing of medical devices with arrays that bind to the many gases needed to detect various diseases. This challenge will ease as MOFs increasingly enter commercial applications with new methods of scale-up.
If we solve this biomedical challenge with MOFs, we see its potential across all the other applications dogs are known for — detecting landmines and narcotics, as well as finding people buried by an avalanche or in the midst of a disaster.
The sense of smell opens up a new dimension of technology for humans to explore and harness for our health and well-being. MOFs give us a new way to explore that dimension and bring cutting-edge materials technology into everyday life.