ADVANCES IN INFRARED TECHNOLOGY FOR BREATH ALCOHOL TESTING WITH PARTICULAR REFERENCE TO THE INTOXIMETER MODEL 3000 M, R. FORRESTER INTOXIMETERS INC "f ST. LOUIS, MQ, U.S.A. Between 1964 and 1966, Aerojet General Corporation, of Azusa, California, investigated Gas Chromatography and Infrared techniques for application in the breath alcohol testing field. They did not succeed at that time in perfecting either instrument, but it was only a matter @f time before their concepts became reality. Shortcomings in breath alcohol testing programs then were more often due to human error than to instrument failure. An ideal instrument would be less operator and subject dependent, yet safeguarded against hidden malfunctions. Richard A. Harte in 1970, introduced the fir st commercial infrared breath tester. It was the Omicron Intoxilyzer. The infrared system required no chemicals or gases and was self-contained in the sense that it only needed a HOv, 60 cycle power supply, Since it was electronic it could readily be automated. Potentially, its main weakness Wc(S its lack of sp cificity. 1. Harte, R.A., A New Alcohol Detection Technique for Control in the Drinking Driver Problem, Thirty^ Third Annual A.A.F.S. Conference, 1977. 677
The Intoxilyzer consisted of an infrared photometer which measured the infrared absorbance at 3.39 microns in a 600 cc vented chamber. The folded path length was approximately 9 feet long. Interference filte r s at the end of the path isolated monochromatic light of 3.39 microns and was monitored by an infrared photo conductor which processed it for quantitative display on the basis of a 2100:1 breath-blood ratio. The instrument found early use in parts of I llin o is, California and Delaware. From a functional point of view the Intoxilyzer was relatively automatic and simple to operate. Cost per test was reasonable since a test only required a mouthpiece and a printout form. Results were displayed numerically and printed on a multiple copy form, leaving l i t t l e room for subjective interpretation. Also, calibration was reported to be factory set and permanent. Early problems related to the d ifficu lty of giving a deep lung sample because of the excessive back pressure required and the 600 cc chamber size. Simulators could easily be hooked up backwards and solution blown into the chamber. The manufacturer claimed the instrument was specific for ethanol, but in fact it was not. It not only read out acetone as alcohol, but also magnified 2 the reading. ' 2. Forrestor, M.R., A New Generation of Breath Testing Instruments and Their Response to Interfering Substances Found in Breath, Seventh Inter- National Conference on Alcohol, Drugs and Traffic Safety, Melbourne - January 23-28, 1977. 678
The mechanical deficiencies have largely been eliminated in the Intoxilyzer as the sampling system has been opened up and the 600 cc volume has been reduced by nearly 50% in some models. The instrument's sen sitivity to acetone has been reduced from three times that of ethanol to a l i t t l e less than a 1:1 relationship. It is reported that the Model 4011-AS, as developed by the CMI Corporation, combines a chopper and f ilt e r making it possible to detect acetone in a breath sample. If the acetone is over a.01% level the unit deactivates the display and printout since it cannot apparently quantitate and subtract the interference. In the United States constant probing by defense attorneys has forced an ongoing evolution in breath testing instrumentation. The human element was too prevalent in the Drunkometer; the consistency of the CO2 ratio was questionable in the Portable Intoximeter; the chromic acid ampul became suspect in the photo-electric instruments since results were not later verifiable. Preserved samples were called for through Supreme Court Decisions in states like Colorado and Arizona. Testing systems free of interfering substances gained favor as they became available. Electronic advances made possible more innovations. Automation of instrument zeroing, sure collection of deep lung breath samples and standardization could all now be perfected. Computerization made tamper-proof data possible to record, and/or communicate to other locations. Obviously, legal probing has resulted in better instruments and programs. Instrument operations have been sim plified, but instruments themselves have become more sophisticated. The IntoximeteR Model 3000 represents the latest state of the art. It was developed in 1979 by Caldetect, Inc. of 679
Richmond, California. Caldetect has employed the latest infrared technology. A single-pass beam of 3.39 micron wavelengths travels 11" through a sp lit chamber and analyzes a breath sample in one side of the chamber for its alcohol content. The volume of the chamber is just 70 cc, thus enhancing the probability of a deep lung breath sample. (See Figure No. 1) Optionally, this chamber can be used to capture and deliver a breath sample to a tube containing anhydrous magnesium perchlorate. This preserved sample may then be analyzed at a later date by one of several published methods. ^ Carefully designed flow lines and monitoring make it both easy to give a sample and also assure that deep lung breath is sampled. The sp lit chamber fa c ilita te s automatic zeroing of the instrument and assures sta b ility. Combining the infrared system with a special semi-conductor configuration, acetone can be sensed, quantitated and eliminated from any sample reading. (See Figure No. 2) The acetone is detected by comparing the output of the solid state semi-conductor to the IR 3. B iasotti, A.A., and Bradford, L.W., Quantitative Determination of Ethanol by Gas Liquid Chromatography after Collection from Vapor Phase of Anhydrous Magnesium Perchlorate, Journal of Forensic Science Society, Vol. 9, Nos. 1-2, 1969, pp. 65-74. Forrester, M.R., The DPC Intoximeter-Analytical Procedure, Information and Suggestions for the Chemist - Intoxim eters, Inc., 1901 Locust Street, S t. Lqyi s, MO 63103. 680
CHOPPER FILTER I R DETECTOR COMPONENTS FIGURE 1 FIGURE 2 681
output. In itia lly the semi-conductor sensor is automatically calibrated by the computer, so that its output equals the IR output when an ethanol sample is analyzed. If the breath sample has acetone present, the output from the semi-conductor cell increases much more rapidly than the IR channel and this difference is used to compute the amount of acetone present and also to correct the acetone contribution to the IR reading. If acetone was present in a sample, both the display and printout will note the fact, but the alcohol reading is not affected. The input board and computer side of the instrument allows several important innovations in breath testing technology. Mainly, the instrument can be preprogrammed to collect and provide specific data. For example, the instrument name and serial number along with the local department's name can be made a permanent part of each printout. Time and date of a test are automatically recorded. The operating procedure can be programmed with or without options which are selected or rejected by the operator with a "yes" or "no" on the input board, and programs can be changed only by supervisory personnel. A test is initiated by pressing the START BUTTON. The information needed is requested by the instrument on the display panel. Typically this might include Operator's Name, License Number, Subject's Name. The instrument then automatically runs a Purge, a Blank, an Internal or External Standard, depending on how it is programmed, and fin ally the display reads Blow Until STAR--SUBJECT. If the subject blows long enough to get deep lung breath into the chamber a star appears in the display and the sample is automatically taken. If a preserved sample is to be collected the display requests the sample tube be mounted. The sample chamber is then purged through this tube and fin ally the display requests the tube be disconnected. Any number of test records may be 682
obtained by pushing the PRINT key at the completion of a test. If a preserved sample was collected the fact will be noted on the printout. It is important to note that each step must be completed satisfa cto rily before the computer program will advance to the next step of the test. For instance, if the license number does not have the correct number of d igits, if the unit is not clean and therefore does not give a blank reading, or if the standard does not read within specification s, the instrument will inform the operator of the problem and either ask for a correction or recycle and attempt the step one more time before closing down. Samples of the display, tellin g the operator that the external standard is low and needs changing or that a sample has been aborted because the subject did not give a deep lung sample are seen below. The plug-in programmed chip and reprogrammable chip make the Model 3 0 0 0 program very flexib le. Features such as external standards, preserved samples, and additional m em ory capability can be added to the basic instrument at a later time if it should be desirable..std,070 LO =ABORT ABORT... PURGE. Each piece of data in a successful test is stored in an internal memory. Presently, the Model 3000 can store data from the last 100 tests. This information can be recovered in a number of different ways. For instance, 1) all tests a given operator has run can be recovered, or, 2) a distribution of test results versus time can be plotted, or 3) a summary of all the tests in m em ory can be recovered. 683
Of greater significance is the potential of tying the Model 3 0 0 0 into an existing computer net so that data can be routinely or instantly transmitted to a central location without being altered, lost or delayed. The IntoximeteR 3 0 0 0 has been designed to be compatible with RS 232-C interfaces which are the standard interface for telephone lin es, teletpye and computer terminals. Figure No. 3 shows a block diagram of the simplest type of data transmission. The Model 3 0 0 0 is programmed to send, on command from the keyboard, either individual test data on a single subject or any of the programmed summaries previously mentioned. This mode of transmission requires the operator at the instrument to in itia te the transmission and an operator at the receiving end to make sure that the computer is on line to receive this transmission. Instrument Serial No. Location Ident. No. Operator's Name Subject's Name Date - Time of Test Test Results (Figure No. 3) 684
The rate at which data can be transmitted depends upon the receiving equipment. Some computers can handle 19,000 characters per second, telephone systems average 300 characters per second, and teletypes range from 100 to 300 characters per second. Transmission speed is particularly important when considering a second mode of data transmission. It requires the instrument to be programmed to report certain data on demand. The Model 3 0 0 0 then waits until it has been interrogated by a central computer before sending back its data. This handshaking mode of transmission allows the central computer to be time shared with other data terminals. The interrogation code for the central computer may be programmed into the IntoximeteR 3 0 0 0 from the keyboard which means if the instrument is moved to a different location it can be reprogrammed for that location. Further, the a b ility to send all the test information back to a central location allows a regulatory agency to monitor the instruments' operation. Known standard solutions may be mailed to the agencies and their results sent back to the regulatory agency by the Model 3 0 0 0 over phone lines or computer net. Potential savings are sign ificant in terms of man hours and costs of periodic maintenance and quality control programs. Monitoring current data is also possible so that planning of selective enforcement programs and their evaluation may be more effective. By combining the sim plicity and speed of an infrared absorption technique that provides sp ecificity for ethanol with the automation and data treatment of modern computer technology, the IntoximeteR Model 3 0 0 0 is the most accurate and complete breath alcohol testing instrument available today. 685