Nuclear magnetic resonance (NMR) was first adopted by the medical world in the 1980s for the diagnosis of brain, muscle, and bone diseases. The widespread medical use of the technique is now expanding into new fields. NMR will tell supermarket shoppers whether the fruit is sweet, the mayonnaise is low-fat as advertised, or the olive oil has been adulterated. A number of studies conducted by researcher Luiz Alberto Colnago and his team at Embrapa Instrumentation, a division of the Brazilian Agricultural Research Corporation (Embrapa) based in the city of São Carlos, in inland São Paulo State, have demonstrated these potential uses that might be commercially exploitable in the future. The Embrapa technology will be developed into a commercial product by Fine Instrument Technology (FIT), a Brazilian company also based in São Carlos.
Nuclear magnetic resonance tomography has been used for controlling the internal quality of in natura fruit since the 1980s, by analyzing the effects of impacts, low temperatures, physiological defects, and the consequences of infestation by pests and diseases. “NMR has also been used to study the variability of compounds found in grapes, and to sort tomatoes according to firmness and maturity,” Colnago adds.
The problem is that the devices used in these studies are little different from their medical counterparts, which can cost millions of dollars. Either that or the analyses can only be performed with the fruit cut into pieces, or with the juice poured into little tubes. In other words, the products cannot be analyzed in their own packaging. So Colnago went ahead and developed low-cost methods and equipment that allow a non-invasive analysis of in natura or packaged foods using nuclear magnetic resonance. “When it comes to analyzing fruit, fresh meats, and packaged commercial products, we have no direct competition. There are no low-cost devices out there,” he says. Colnago says that Professor Michael McCarthy at the University of California, Davis has been using costly tomographic equipment to study vegetables. “He has analyzed mainly avocados, plums, and olives,” the Embrapa researcher reports. “However, the analytical principle he developed is different from ours. It takes longer and the analyses cannot be performed using low-cost equipment.”
Like similar devices, the NMR developed in São Carlos consists of a magnet, a radio transceiver station, and a computer. The radio transceiver, manufactured by Tecmag, was imported from the United States. Colnago developed the antenna. The products to be analyzed can be packaged as usual, as long as it is not a metal container or a Tetra Pak carton, as these materials are impervious to radio waves. The magnets in the device magnetize the products. Pulse-modulated radio signals (electromagnetic energy) at a frequency of nine megahertz (MHz) are generated by an AM radio transmitter and directed into the sample by the antenna. “After the product has been irradiated, it induces a signal – known as a magnetic resonance signal – in the antenna. The signal is then amplified, converted to a digital format, and stored in a computer.”
Resonance is possible because the atomic nuclei of each chemical element in the material analyzed absorbs energy at specific frequencies, making it possible to distinguish each type of atom in a molecule. This allows obtaining the chemical and biochemical composition of the products analyzed in an NMR machine. A major advantage is that Colnago’s NMR reading is ready in less than a second, whereas a medical imaging procedure can take up to 40 minutes. “The output from the machines that are used for food is not an image, but only a measurement of the time it takes for the radio signal to disappear, which is in proportion to the product’s viscosity,” Colnago explains. “The more viscous, the faster the signal will disappear.”
A rock is an extremely viscous material. As such, its signal will disappear immediately. A signal from water, which has a low viscosity, will last about three seconds. And this property of NMR is also why the signal will disappear more quickly when a piece of fruit is sweeter – and therefore more viscous – than when its sugar content is low. The same notion applies to measuring the fat content in a jar of mayonnaise or for checking whether a bottle of olive oil has been adulterated with soybean oil, for instance. When it comes to meat, the technique allows users to check three different properties: tenderness, juiciness, and intramuscular fat content.
Despite costing a lot less than their medical counterparts, the initial models designed by the researchers at Embrapa Instrumentation and funded by FAPESP are still too expensive – not to mention too big – for use in supermarkets. So Colnago has been developing smaller, more economical versions. One of them cost around R$40,000 to make. Another model, which is smaller and flat, and allows the product to be placed on top, costs R$5,000.
This smaller, circular device is currently under development to be sold by FIT, a company established in 2006 in the city of São Paulo and later installed in São Carlos. “The company was initially founded to sell medical magnetic resonance equipment made in China, and develop components for these machines until they became Chinese-Brazilian products. Shortly after, a multinational acquired the Chinese company and FIT was left with no product to sell, but it continued to develop its projects,” says FIT project manager Daniel Martelozo Consalter. “In 2009, we took the first step towards building a Brazilian-made magnetic resonance machine. We were awarded one grant from the Brazilian Innovation Agency (Finep) and another from the Qualification Program for Human Resources in Strategic Areas (Rhae) at the National Council for Scientific and Technological Development (CNPq),” says Consalter.
FIT used the grant money to develop five receiver coils (NMR equipment components) that can be used in knee, wrist, elbow, and temporomandibular joint imaging, as well as a control module, also known as a digital spectrometer, which is already available for sale under the commercial name SpecFIT. “This control module is the basis for any magnetic resonance machine,” says Consalter. “During this project, we entered into a partnership with the São Carlos Institute of Physics at USP to develop the SpecFIT.”
1. Development and validation of time domain NMR spectrometers and methods for non-destructive analysis of food (nº 2012/20247-8); Grant mechanism Regular Research Awards; Principal investigator Luiz Alberto Colnago (Embrapa); Investment R$222,188.58 (FAPESP).
2. Analysis of agrifood products by flow NMR (nº 2009/09526-0); Grant mechanism Regular Research Awards; Principal investigator Luiz Alberto Colnago (Embrapa); Investment R$405,449.82 (FAPESP).
COLNAGO, L.A. et al. Why is inline NMR rarely used as industrial sensor? Challenges and opportunities. Chemical and Engineering Technology. v. 37, p. 191-203. fev 2014.