Written by: Ben Wilbert, Product Manager, Cole-Parmer
While most scientists who study human brain evolution tend to regard something people do very well, such as language, Dr. Tyler Lorig, Ph.D. at Washington & Lee University studies one of the least understood senses and something people tend not to do so well: smell. With an interest in brain evolution and a background in EEG analysis, he studies the olfactory (a.k.a. sense of smell) system in his lab, specifically how the brain responds to an odor. Often a neglected and under-served field of science, and past research often fraught with loose control methods, Dr. Lorig decided to develop his own olfactometer. Since an olfactometer isn’t exactly the type of device you’d find on the shelf of the hardware store—or even in a specialty catalog—Dr. Lorig built his olfactometer by using many quality products from the Cole-Parmer catalog, starting with solenoid valves. The one he devised proved to be much less expensive than one valued at $100,000 on the market, yet served all the needs of his research.
One of the challenges to studying the sense of smell is being able to appropriately regulate the amount of the stimulus. It’s difficult to determine how much of an odor you’ve delivered to a test subject. While you may try and “dose” the amount of smell, there are many factors affecting the distribution of airborne molecules that provide an odor. In studying the body’s other natural senses, it’s comparatively easy to gauge precisely how much of a stimulus is applied to a subject. Light, sound, electricity, and even flavors—although taste is a field of study filled with ambiguity itself—can be regulated to keep tight controls over a scientific test. But in testing the olfactory response, even the weight of the molecule, for instance, will affect how fast it disperses, and hence sensed by someone. Lighter molecules move quicker, so if you’ve decided to expose the test subject to five seconds of a particular odor, you will get a different result than a heavier molecule that only gets a fraction of the activity in the air within the same time frame. This means lightweight molecules will invade the nose, migrate through the mucous membrane at the top of the nasal cavity, and be sensed more so than heavier molecules. To top it all off, there is even remarkable inter-subject variability, whereby one person may sense an odor sooner or later than another person.
To minimize the effects of the inherent differences in the physical nature of odors and test subjects themselves, the equipment used has to be of such design and quality that the experiment is not compromised. All parts of Dr. Lorig’s olfactometer that could contaminate smells are high purity, hence minimizing any residual odor that would affect the experiment results. Lorig admits to first coming to Cole-Parmer while looking for some high-quality solenoid valves. He sought PTFE (wetted parts) valves because of PTFEs quality to stay clean, and not absorb errant odors. In his paper “A computer-controlled olfactometer for fMRI and electrophysiological studies of olfaction”, originally published in Behavior Research Methods, Instruments, and Computers, Lorig describes the design for an inexpensive and reliable olfactometer that he pieced together and constructed from off-the-shelf chromatography parts that required little modification. Since he would be using the olfactometer near an fMRI, the olfactometer had to obviously be free of ferrous metals, which will wreak havoc near the magnet. Overall, the instrument needed essentially seven features: (1) computer control; (2) effective delivery of a variety of odors, in series or randomly; (3) production of an odor stimulus of selectable and reliable duration in a constant airstream, without any additional type of ancillary stimulation (e.g., tactile, auditory); (4) resistance to contamination; (5) durability; (6) ease of operation, refilling, and cleaning; and (7) low cost (Lorig et al. 1999).
Following the drawing above, air from a compressor is passed through a charcoal filter to remove odors and then through particulate filters to prevent charcoal dust from being administered to test subject. After passing through the particulate filters the flow is divided and metered through variable-area flowmeters. One of the lines is always open and provides a constant low-volume air stream. The other flowmeter provides the air that will be passed over the odors. This stream is also divided and passed to two solenoid valves. Valve A is a single valve that is normally open. The other valve is a multi-port valve that can have from 1 to 6 individual normally closed solenoid valves (B1-6). To send an odor to a subject, the computer turns on valve A (stopping airflow in that line) and turns on valve Bn commencing airflow in that line. The syringe filter connected to line Bn contains odor, and the air now passes over the filter and through the manifold to the subject. Turning the valve off stops airflow over the filter paper and stops the blockage cause by actuating valve A. To avoid any increases in air flow, one non-odorized line is stopped during odor stimulation making the net change in air zero. Because the switching in the valves lead to very brief airflow changes (around 20 milliseconds) the constant flow line acts as a buffer for the airflow change, thereby reducing any extraneous sensory stimulation to the test subject.
“Some of the research done shows we are exquisitely sensitive to smells, contrary to our expectation.” states Lorig.
In relatively normal test subjects, Lorig finds people have measurable brain activity induced by odors, even when the test subject reports not smelling anything! Even when more than one chemical is blindly switched—neither reported as smelled—they render different brain responses. While brain patterns related to particular smells may evoke similar and predictable brain responses, Lorig is careful not to jump too far in his conclusions, for example, they will not indicate emotions such as fear or joy.
Lorig notes that when it comes to the extreme smells, people tend to agree across cultural boundaries as to what smells bad and what smells good. On the bad end of the spectrum, odors such as feces and cadavers evoke similar negative responses from people, and on the pleasant end, vanilla ranks universally high as a positive response from people. But in the vast midsection of the odor continuum, there is a wide variance regarding what is pleasant versus not-so-pleasant odors. Other interests include why some people find certain odors pleasant or at least tolerable, while others find them absolutely repugnant.
“Since I’ve talked to so many people about smell, they will sort of confess, ‘Oh, I really like skunk smell.’ ”
While much of the current olfaction study takes place in a research setting, Lorig states he would like to see olfaction analysis become simplified . There’s now understanding of the connection between olfaction and certain health problems. Current research examines the relationship between olfaction and maladies such as Parkinson’s Disease, Huntington’s Disease, Korsakoff’s Syndrome, Schizophrenia, Depression, and Alzheimer’s Disease (AD). Recent evidence suggests that areas in the central nervous system processing olfactory information are affected at the early stages of AD, even before the onset of cognitive decline, and that olfactory dysfunction might be an early indicator of AD (Murphy, 1999). The smell threshold is much higher for those who suffer from AD.
Aside from aiding pathological diagnosis, Lorig’s current and future toils include researching how the brain is organized, the pathways the brain uses to process odors, and the many relationships between smelling and the other senses. Cole-Parmer continues to provide scientific instruments used by these professionals to support the overall advancement of science. We thank Dr. Lorig for his time and efforts in providing valuable feedback about our products and wish him well in his future endeavors.
Lorig TS, Elmes DG, Zald DH, Pardo JV (1999) A computer-controlled olfactomenter for fMRI and electrophysiological studies of olfaction. Behav Res Methods Instrum Comput 31: 370-375.
Murphy, C., 1999. Loss of olfactory function in dementing disease. Physiol. Behav. 66 (2), 177-182.
Complete selection of Flowmeters
Complete selection of Syringe Filters
Complete selection of Valves