Imprimir Republish

Chemical engineering

Artificial nose detects gases

Petroleum sector invests in portable sensor for areas of risk

Two years ago, researchers from Pernambuco succeeded in analyzing electronically and identifying the aromas and vintages of different types of wine with the help of the computer. Entirely Brazilian, the technology has now been perfected and mastered in all its stages, and is being adapted to detect toxic and explosive gases in oil rigs and refineries. The fruit of a partnership between the Catholic University of Pernambuco (Unicap) and the Federal University of Pernambuco (UFPE) and funds of R$ 1 million passed down in 2001, the electronic nose is among the main projects approved by the oil industry sectorial fund, CTPetro.

The merger between electronics and “smelling” takes place by means of electricity conducting polymers, which are materials that alter their electrical resistance in the presence of gases. The team from Unicap’s Polymers Laboratory makes use of polypyrrole, doped with reagents which give it electrical properties at room temperature. The material is deposited on slides of conductive glass – the sensors of the electronic nose. “The innovation lies in designing the best formula for these reagents”, explains chemical engineer Edson Gomes de Souza, from Unicap. Each slide is connected by wires to a device that measures the electric current and is linked to a computer that stores the measurements.

The system for collecting the data was developed in the University from Pernambuco, through a graphic language software called LabView. The alterations in electrical conductivity generated by the gases are transformed into patterns stored in the database. The information is interpreted by neural network systems, “trained” by UFPE’s artificial intelligence team to recognize the odors corresponding to the various graphs of conductivity.

Imperceptible difference
The project started out with the detection of odors of the simpler products in molecular terms, like the organic solvents, ethanol and methanol. “They are substances with differences that are imperceptible to the human sense of smell, but one of them, methanol, can kill”, points out physician Francisco Luiz dos Santos, from Unicap. With this successful experience, the group went on to distinguish electronically different types of wine – white, red, and rosé. The artificial nose also managed to differentiate wine from whisky, which was when the scientists decided to test the identification of Brazilian red wines of the vintages of 1995, 1996 and 1997.

The more recent ones have a higher content of sulfur dioxide, detected by the sensor. “The technological challenge was to form distinct patterns for the different vintages”, says Santos. The system was 90% accurate. The results, which have already yielded one thesis for a master’s degree and two for a doctorate, were published in the scientific magazine, Synthetic Metals.

The technology for electronic noses is one of those that is growing most in the world, on the steps of the scientific effervescence in the polymers sector. The electrical properties of these materials gave the Nobel Prize for Chemistry to Japanese and American scientists in 2000. From food industries to help in diagnosing diseases associated with the odors exhaled by the organism, the applications of this technology are countless. The focus in Brazil is now the petroleum sector. A gas center with a control panel was installed in Unicamp in November last year, at a cost of R$ 30,000, to check on the sensibility of the sensors to methane, ethane, butane, propane and carbon monoxide.

These are gases that are explosive and toxic, some of them odorless to the human being. Chemical engineers are perfecting the architecture of the test tubes and using new reagents. The objective is to increase from eight to 30 the number of sensors contained in the artificial nose, which in the future will be used to detect quickly and accurately.

New model
At the same time, the researchers from Pernambuco developed a third generation of artificial noses. There is a portable prototype that can be carried into risky industrial areas. The program for miniaturizing the “nose” received R$ 500,000 from CTPetro. All the artificial intelligence programs are being adapted to the new model. The previous prototype worked inside an executive style briefcase, which contained a laptop computer and a gas collecting apparatus with the sensors. In the new version, the screen of the laptop is replaced by a mini-display of liquid crystal, and the computer’s processor by a microcontroller chip. The objective is to process the data online and identify the gases immediately. “We already have the technology. The delay is on the commercial side”, says Santos.

From start to finish, the process was developed in Brazil. It all began seven years ago, when the physicist’s wife, systems analyst with UFPE Marizete Silva Santos, was concluding her thesis for a master’s degree on multimedia centers for serving the public. “These devices were beginning to reproduce human senses, like sight, hearing and touch, but smell was missing”, Santos says. “As a physicist, I committed myself to finding a solution.” The subject was being studied in the United States and in Europe, but literature was sparse. The discovery happened by chance. As he was working on electricity conducting polymers, the physician decided to sprinkle alcohol on some of them to see what would happen.

He verified the change in electrical resistance. And he started to put together the first, rudimentary prototype, fixing the material in needles for injection syringes. As it used to happen in Japan in the age of the transistors, patient and careful trainees would bring the tips of the needles together under the microscope. But none of this was a novelty. “We knew that prototypes of this kind already existed in England”, Santos points out. It was better that way. “In the isolation of the northeast, we started the work from scratch, and today we are in the vanguard.”

Source of funds for universities

Pernambuco does not produce petroleum, but it is in the front line of research into this fuel and its derivatives. UFPE has 20 projects approved by the Financier of Studies and Projects (Finep) and by the National Council of Scientific and Technological Development (CNPq), with R$ 5 million in funds from CTPetro. Another six are cooperative projects, carried out in several states of the north and northeast, under the coordination of the university from Pernambuco.

As it is happening with the electronic nose, several groups are focusing their research on the oil and natural gas sector, which is the owner of one of the largest sources of funds for universities. “We have never received so much money”, celebrates chemical engineer César Abreu, the coordinator of UFPE’s Catalytic Process Laboratory. Just the Department of Chemical Engineering, responsible for the Laboratory, is running 13 projects, worth R$ 3.7 million.

One of the most important is the one for the catalytic reform of natural gas, a technology based on chemical reactions to transform the product into substances of higher added value. At the moment, natural gas is only used in processes of burning, automobile combustion, or in the generation of heat and electricity. But it has nobler uses. Associated with carbon dioxide – a pollutant carried away by the smokestacks of alcohol distilleries -, it generates carbon monoxide and hydrogen. These are raw materials in the production of liquid fuels like synthetic gasoline, methanol, heavy hydrocarbons and hydrogen. “In liquid form, these fuels generate less risks in their transport and storage”, says Abreu.