Sniffing Entrapped Humans with Sensor Arrays

Earthquakes are lethal natural disasters frequently burying people alive under collapsed buildings. Tracking entrapped humans from their unique volatile chemical signature with hand-held devices would accelerate urban search and rescue (USaR) efforts. Here, a pilot study is presented with compact and orthogonal sensor arrays to detect the breath- and skin-emitted metabolic tracers acetone, ammonia, isoprene, CO2, and relative humidity (RH), all together serving as sign of life. It consists of three nanostructured metal-oxide sensors (Si-doped WO3, Si-doped MoO3, and Ti-doped ZnO), each specifically tailored at the nanoscale for highly sensitive and selective tracer detection along with commercial CO2 and humidity sensors. When tested on humans enclosed in plethysmography chambers to simulate entrapment, this sensor array rapidly detected sub-ppm acetone, ammonia, and isoprene concentrations with high accuracies (19, 21, and 3 ppb, respectively) and precision, unprecedented by portable sensors but required for USaR. These results were in good agreement (Pearson’s correlation coefficients ≥0.9) with benchtop selective reagent ionization time-of-flight mass spectrometry (SRI-TOF-MS). As a result, an inexpensive sensor array is presented that can be integrated readily into hand-held or even drone-carried detectors for first responders to rapidly screen affected terrain.

following 2 h, each volunteer maintained in a sitting position inside the chamber. In the first hour, only skin-emitted volatile compounds were targeted. To achieve this, the subjects freely inhaled and exhaled (outside-chamber) room air via a two-way non-rebreathing Y-shaped valve (Hans Rudolph Inc., USA) of a silicone head mask (V2 Mask, Hans Rudolph Inc., USA) connected with two flexible polypropylene tubes (ID = 22 mm, Flextube, Intersurgical Inc., UK) to two ports located on the side wall of the chamber. In the second hour, breath-and skin-emitted compounds were measured. For this, the volunteers disconnected the outlet tube from the mask and exhaled directly inside the chamber, while still inhaling outside air.
Therefore, during the second hour of the experiment both, breath and skin-emitted volatiles were measured. Altogether, a single experiment lasted for 140 minutes.

Selective reagent ionization time-of-flight mass spectrometer (SRI-TOF-MS)
An Ionicon Analytik (Innsbruck, Austria) type 8000 SRI-TOF-MS, was used to monitor acetone, ammonia and isoprene continuously. Chamber air was extracted through a sampling port (34 x 30 mm, i.e. detector inlet) located at the center of the sidewall of the chamber. This port was maintained at 40 °C to avoid condensation as it was sufficiently higher than the temperature of the air inside the plethysmographic chamber during entrapment of volunteers ( Figure S4b). To this, a 3 m (d i = 3.188 mm) long Teflon transfer line was connected delivering a steady flow of 100 mL min -1 to the instrument. Protonated acetone (C 3 H 7 O + , m/z = 59.0492 Th) and isoprene (C 5 H 9 O + , m/z = 69.0704 Th) were generated through reaction of the corresponding neutral molecule with H 3 O + . 35 Given that ammonia cannot be detected accurately in this mode owing to a high instrumental background, O 2 + ions were used in a combined mode. 36 The ion-source current, source voltage, source out voltage and valve whereas the standards for acetone and isoprene calibration were prepared using a method described elsewhere. 37

Sensors and Array
Chemoresistive Si-doped WO 3 (10 mol%) 38 , Si-doped MoO 3 (3 wt%) 25 and Ti-doped ZnO (2.5 mol%) 26 sensors were prepared by flame aerosol technology and directly deposited onto sensor substrates 28 . These were mounted on macor holders and installed in Teflon chambers, described in detail elsewhere. 39 The sensors were heated up to 350, 400 and 325 °C for optimal sensitivity and selectivity to acetone 39  The chemoresistive sensor responses were calculated as: where R b and R c are the sensor film resistances in background (room) and chamber air, respectively. Analyte concentrations are estimated by the sensor array, as illustrated in Fig. 2 of the paper. The individual sensor responses S ୧ are processed with a multivariate linear regression model 29 , as done with a SnO 2 -based sensor array in laboratory gas mixtures 22 .
Therein, the concentration of an analyte C ୶ is described as a superposition of each sensor response S ୧ multiplied with an individual regression coefficient a ୧,୶ and an intercept b ୶ : Due to the known interference of humidity for metal oxide-based chemoresistive gas sensors 39 , the RH is included also as a sensor response S ୧ and regression coefficient a ୧,୶ . In a "training" step, all coefficients a ୧,୶ and b ୶ were calculated using SRI-TOF-MS concentrations and the sensors responses as input. The performance of the sensor array model was then assessed by applying the regression coefficient on a separate data set (not used for "training").
This was followed by stepwise-elimination to identify multicollinearity 40 between sensors, remove insignificant sensors from the prediction and determine the best composition overall.
Limited acoustic and seismic range.
Interference by other noise.

Canines
Search and rescue dogs 3 Short operational time.
Limited numbers due to expensive training. Prone to stress.