JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 22, NO. 2, APRIL 2013 443 A Microfabricated Propofol Trap for Breath-Based Anesthesia Depth Monitoring Muhammad Akbar and Masoud Agah, Senior Member, IEEE Abstract—This paper reports a golden microtrap chip for anes- thetic depth monitoring. The chip selectively captures the propofol (2,6-diisopropylphenol) compound, which is a widely used sub- stance for anesthesia, by filtering out the other species found in human breath. The fabricated silicon-glass chip is 12 mm on each side and consists of an array of high aspect ratio parabolic reflectors inside its 7 mm × 7 mm × 0.24 mm cavity. The interior surfaces of the chip are coated with an electroplated gold layer having a surface roughness of around 6.82 nm, which is an order of magnitude higher than a gold layer deposited by electron (e)-beam evaporation. Uncoated and e-beam gold-coated chips are unable to trap propofol and other compounds found in human breath. In contrast, silicon-glass chips coated with Tenax TA (2,6 dipheny- lene oxide), a gas adsorbing polymer, captures propofol among other volatiles present in human breath. Only devices coated with an electroplated gold layer demonstrate selective affinity for the target compound propofol. For the same propofol concentration, these golden microtraps show consistent capture efficiency with less than 8% variation in the trapped propofol amount while tested with different human breath samples. These microfabricated chips have the potential to accurately quantify the amount of propofol present in human breath samples without incorporating the gas chromatography column into the testing setup, resulting in faster analysis and reduced cost and complexity. [2012-0062] Index Terms—Breath analysis, gas chromatography (GC), mi- croelectromechanical systems (MEMS) propofol sensor, micro preconcentrator (μPC). I. I NTRODUCTION A NESTHETIC agents can be administered into the body of human patients by different methods. In conventional methods, the vaporized inhalants are introduced into the body through oral or nasal means. The anesthesia depth is reflected in the change of the partial pressure of the anesthetics in the brain [1]. One of the disadvantages using this method is the requirement of special apparatus (anesthetic vaporizer) to de- liver a specific amount of anesthetic dose to the patient [2]. On the other hand, the intravenous technique relies on injecting the anesthetic agents (propofol, ketamine, and etomidate) directly into the vein. The agents gain access to the blood, after being administered by one of these methods, and become part of the circulatory system through which they are transported to the Manuscript received March 12, 2012; revised September 16, 2012; accepted October 21, 2012. Date of publication December 10, 2012; date of current version March 29, 2013. This work was supported by the National Science Foundation under Award CBET-0854242. Subject Editor R. Ghodssi. The authors are with the Virginia Tech MEMS Laboratory, Bradley Depart- ment of Electrical and Computer Engineering, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061 USA (e-mail: makbar@vt.edu; agah@vt.edu). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JMEMS.2012.2227949 desired location in the central or autonomic nervous system [3]–[6]. The accurate amount of dose necessary to induce anesthesia is very critical and is a persistent challenge for anesthesiolo- gists [7]–[9]. An inadequate amount of anesthesia may result in acute pain and accidental awakening during the surgery and is extremely traumatic. Moreover, the patient may recall conversation and other unpleasant events that occurred during the process of surgery. Similarly, overmedication of anesthetic agent can lead to serious consequences such as permanent brain damage and even mortality. Propofol (2,6-diisopropylphenol) is widely used for general sedation in intensive care units. It has a phenolic chemical structure with a distinct smell and a molecular weight of 178.27 g/mol. Monitoring the quantity of propofol injected into the blood of the patient undergoing anesthesia is very critical and usually is carried out by high-performance liquid chro- matography and gas chromatography (GC) [10], [11]. These methods are time consuming, expensive, and labor intensive. Moreover, after the injection of anesthetic agent into the patient body, 97%–99% of propofol is bound with albumin and red blood cells, and the remainder exists in blood as free type. Only free type is detected by these techniques, which makes them inaccurate for the quantification of propofol [12], [13]. Fur- thermore, almost 88% of the propofol dose can be monitored in urine sample as hydroxylated and conjugated metabolites [14], [15]. Currently, there is no medical device that can directly deter- mine the level of anesthesia based on effect-site concentration in blood [16], [17]. Therefore, anesthesiologists presently rely on observation and indirect monitoring methods, including monitoring the blood pressure, patient pulse, pain response, heart rate, and rhythm [18], [19]. Analysis of the exhaled breath provides a noninvasive method to quantify/monitor the compound of interest in the breath sample. Recent studies have revealed that there is cor- relation between the amount of propofol injected into the blood and that present in the exhaled breath of patients [20]–[23]. However, analyzing breath is not a trivial task for several reasons. First, volatile organic compounds (VOCs) are present at very low concentration, usually in the parts-per-billion range in the breath sample. Analyzing analytes at such low level of concentration is extremely difficult and cumbersome. Second, the high level of breath complexity containing more than 1200 VOCs, including ketones, aldehydes, and alcohols, makes it difficult to identify the compound of interest from the rest of the breath biomarkers found in human breath. In addition, the background level of these compounds must be accounted 1057-7157/$31.00 © 2012 IEEE