Background/Objectives:Polydimethylsiloxane (PDMS) is a non-invasive and versatile material often used for non-invasive collection of skin-emitted volatile organic compounds (VOCs), with potential applicability in acute and pre-critical care settings. However, most existing PDMS-based methodologies rely on extensive sample preparation and environmental control, limiting their feasibility in time-sensitive clinical contexts. Methods:We conducted a proof-of-concept pilot study in four healthy volunteers to evaluate whether a simplified skin-contact PDMS sampling procedure can capture detectable VOCs and preserve individual-level variation. PDMS strips were applied directly to the skin with minimal preparation, and collected VOCs were analyzed using gas chromatography-mass spectrometry. Donor-associated variability was assessed using Bray-Curtis dissimilarity, and variability in VOC detection was evaluated across body sites. Results:Skin-contact PDMS sampling detected 160 VOCs across four participants. The mean within-donor Bray-Curtis dissimilarity was 0.308, compared with a mean between-donor dissimilarity of 0.347. Preliminary permutation testing showed distinguishable donor profiles (p-value = 0.004). VOC detection variability differed across body sites, with lower coefficients of variation at the forehead, neck, and wrist than at the ankle. Conclusions:Under simplified sampling conditions, skin-contact PDMS captured individual-associated VOC profiles with lower within-donor variability than between-donor variability. These findings support the feasibility of PDMS-based skin VOC sampling in minimally controlled settings. Further validation in larger and clinically relevant cohorts is warranted to assess the utility of PDMS-sampled skin VOCs as potential biomarkers for early disease detection.

Water vapor extraction from human exhaled breath is critical for breath analysis, enabling lower humidity interference in volatile compound sensing and separate analysis of exhaled breath condensate. Thus, any condenser surface must support efficient vapor extraction and condensate collection from high-humidity breath samples. This study investigates the effect of aluminum surface modification on water vapor extraction and condensate collection from humid air simulating exhaled breath in a 110 & times; 40 & times; 5 mm rectangular flow channel. Aluminum surfaces were modified via alkali etching and anodization to achieve superhydrophilicity (contact angles: 15.6 deg +/- 2.5 deg and 12.9 deg +/- 1.13 deg) and further treated with perfluorooctyltriethoxysilane to obtain superhydrophobicity (contact angles: 154.3 deg +/- 1.9 deg and 165.5 deg +/- 1.1 deg). Surface modification did not significantly affect bulk heat and mass transfer within 28-36 degrees C and 95% relative humidity, and the data agreed with the comsol model within 15% for temperature and 20% for water vapor concentration. Droplet growth behavior differed between treatments, where alkali-treated surfaces exhibited pinned droplets with larger departure diameters (2.061 +/- 0.319 mm), while anodized surfaces had smaller, unpinned droplets (0.351 +/- 0.151 mm). Under gravity-driven collection, superhydrophilic surfaces outperformed superhydrophobic ones, with anodized superhydrophobic surfaces showing the lowest collection efficiency.

The use of lithium chloride as an internal standard for capillary electrophoresis analysis of CAFO waste is examined. Lithium is chosen due to its ostensible exclusion from CAFO waste and its long elution time due to complexation with crown ethers. An alkaline mixture of lithium, sodium, and potassium was tested via capillary electrophoresis in a buffer containing 18-crown-6 ether and the elution times when normalized by the elution time of lithium were significantly more consistent than the absolute elution times. Poultry samples were then analyzed to identify anion concentration levels. Introduction Capillary electrophoresis is a separation technique where different charged ions are separated by their electrophoretic mobility, as a function of the size and charge of the molecule. Separation of the analyte occurs in a capillary filled with a buffer solution with a large potential difference across the capillary. Besides the charged ions, the buffer moves through the capillary due to electroosmotic flow, in which an electric double layer forms at the capillary surface due to ionized hydroxyl groups on the glass surface. The electric double layer pulls the bulk solution as it moves towards the cathode. Poor repeatability in capillary electrophoresis is known [1][2] and is often attributed to fluctuations in electroosmotic flow or other interactions with the capillary wall. There are quite a few methods analysts use to overcome these issues such as capillary coatings (surfactant-modified capillaries), frequent and thorough washing of the capillary (sometimes with weak hydroxide solutions), or relative migration time. Extensive effort has been put forth to characterize wastewater from concentrated animal feeding operations (CAFOs) or other agricultural sources [3] and lithium, generally being absent in such samples, is thus a potential internal standard. Methods A capillary electrophoresis instrument was constructed according to figure 1 using pumps and valves from LabSmith, a custom-built capacitively-couple contactless conductivity detector (C4D), and an EMCO CB101N DC-DC converter. All capillaries are 100 um inner diameter except for the sampling arm which is 75 um to have higher fluidic resistance during rinsing. The data from the C4D is recorded via an Arduino. The buffer used was 12 mM histidine and 2 mM 18-crown-6 ether in deionized water balanced to a pH of 4 with acetic acid. A voltage difference of -10 kV was applied between the high voltage and ground electrodes. Before any experiments were run, the capillary was cleaned with a mixture of 0.1 M NaOH solution and then deionized water for 30 and 60 minutes respectively. Lastly the capillary was flushed with buffer for 60 minutes. Mixtures of varying concentrations between 0.5 and 10 mM of the three alkaline metals, lithium, sodium, and potassium, were injected into the capillary electrophoresis setup via the sampling reservoir. Between each run, the capillary was flushed with buffer. Results and Conclusions Several of the electropherograms with constant lithium concentration, after lithium normalization, are shown in figure 2. Before normalization, the percent relative standard deviation of sodium and potassium were 32.6% and 28.6% respectively. After normalization, the percent relative standard deviations decreased to 5.4% and 7.6% respectively. Even with a crude electrophoresis instrument, lithium normalization has greatly increased the reproducibility of analysis. Acknowledgements This material is based upon work supported by the National Science Foundation under Grant No. 2133576. References [1] Nowak, P. M., Woźniakiewicz, M., Gładysz, M., Janus, M., & Kościelniak, P. (2017). Improving repeatability of capillary electrophoresis—a critical comparison of ten different capillary inner surfaces and three criteria of peak identification. Analytical and Bioanalytical Chemistry , 409 , 4383-4393. [2] Shihabi, Z. K., & Hinsdale, M. E. (1995). Some variables affecting reproducibility in capillary electrophoresis. Electrophoresis , 16 (1), 2159-2163. [3] Bradford, S. A., Segal, E., Zheng, W., Wang, Q., & Hutchins, S. R. (2008). Reuse of Concentrated Animal Feeding Operation Wastewater on Agricultural Lands. Journal of Environmental Quality , 37 (S5), S-97-S-115. Figure 1

In this paper, the directed message-passing neural network architecture is used to predict several quantities of interest in gas chromatography: retention times, Clarke-Glew 3-point thermodynamic parameters for simulation, and retention indices. The retention index model was trained with 48,803 training samples and reached 1.9–2.6% accuracy, whereas the thermodynamic parameters and retention time were trained by using 230 training data samples yielding 17% accuracy. Furthermore, the accuracy as a function of the number of training samples is investigated, showing the necessity of large, accurate datasets for training deep learning-based models. Lastly, several uses of such a model for the identification of compounds and the optimization of GC parameters are discussed.

One-dimensional gas chromatography (1D-GC) stationary phases are generally classified according to their relative polarity into non-polar, semi-polar, and polar columns. In comprehensive two-dimensional gas chromatography (GC×GC), it is the polarity difference between the two tandem-assembled stationary phases that determines the selectivity of the column ensemble. This polarity difference is called orthogonality, and GC×GC column sets can be broadly categorized into four groups based on the direction of the serial coupling between the primary and secondary columns. A significant portion of the GC×GC column sets in use today are operated in forward-orthogonality mode, which means that the secondary column is more polar than the primary column. A growing number of reported GC×GC applications operate in reversed-orthogonality configurations, where the secondary column is less polar than the primary column. Very few examples exist of non-orthogonal column sets because of the fact that there is limited additional selectivity that the secondary column can offer over the separation that has already been achieved in the primary column when the two phases are either identical or close in polarity. The fourth group of GC×GC column sets, called hybrid orthogonality, involves the coupling of stationary phases with peculiar selectivity differences that manifest themselves in the bi-dimensional separation plots. In this work, we are presenting a normalized approach to GC×GC column set characterization that is based on the use of a reference mixture of standards called the Century Mix. The Century Mix contains 100 chemical probes of different functionalities that span a reasonable range of volatilities and polarities to capture the selectivity profile of any GC×GC column set. The Century Mix also contains important chemical probes that are used for 1D-GC column characterization (such as the Grob mix and the Rohrschneider/McReynolds compounds) to make some connections between the 1D-GC building block columns and the GC×GC column sets. We finally also outline some other important metrics of comparison that should be taken into consideration to assess the overall performance of a GC×GC system.

This study presents an analytical approach for calculating retention indices of unknown compounds in the 2nd dimension of GC×GC systems by using an n-alkane-aromatic hydrocarbon (AH) reference mixture. We demonstrated that, by running a reference mixture containing a series of n-alkanes and three AH compounds in a GC×GC system, with the same run parameters as the target sample, we can mathematically obtain the retention index of unknown compounds. An analytical calculation procedure was developed and verified by using 59 test compounds. The results show that the calculated second dimension retention indices are in good agreement with those obtained from 1-D GC–MS runs, with a high coefficient of determination (R² = 0.996) and an average 2nd dimension retention index error of 14 points with standard deviation of 11.7.

A pipeline for the analysis of GC×GC-MS data is presented. The pipeline consists of retention index matching that improves compound identification as well as several methods for removing background compounds of no interest and column bleed. The efficacy of retention index matching is demonstrated via sample mixtures of 26, 26, and 23 compounds. The pipeline in practice is able to reduce the number of matched compounds by half even with lenient screening parameters.

CAFO (Concentrated animal feed operations) wastewater is a promising source for fertilizer to apply to soils in agriculture. However the overapplication of nutrients is just as much of a problem as underapplication [1,2]. The waste material requires analysis of the ionic composition prior to enrichment as a fertilizer. Current methods for ion concentration analysis including Ion Chromatography, Flow Injection Analysis, and High Pressure Liquid Chromatography, which are laboratory instruments. Although some development of portable instruments has been undertaken by NASA/JPL and other groups [3], these techniques are not yet suitable for low cost portable analysis systems. Capillary Electrophoresis methods have been applied to analysis of lake water [4], biological samples [5], food [6], agricultural products [7] providing a rapid analysis at lower cost than laboratory tests. In this study simulated capillary electrophoresis for the separation of several essential plant nutrients in their anionic forms: sulfate, nitrate, chloride, and phosphate was undertaken. The relative concentrations for these anions were chosen to be similar to those found in [8.9] with a 1:100 dilution. The conditions for the separation are listed in Table 1. Several buffer solutions were simulated including borate, sodium phosphate, and the combination of lactic acid and histidine. Separations using borate and sodium phosphate caused the four anions to elute in less than one minute in most cases, which could lead to their misidentification as they are eluting too rapidly. Separations using lactic acid and histidine were slower, but consequently had more time between each peak allowing for an easier detection. The total separation time can be tailored through the concentration of the two buffer components while still allowing greater time between elutions. References [1] Goyal, Sham S., and Ray C. Huffaker. "Nitrogen toxicity in plants." Nitrogen in crop production (1984): 97-118. [2] Rashid, Ghazunfar, et al. "Influence of Nitrogen Fertilizer on Nitrate Contents of Plants: A Prospective Aspect of Nitrate Poisoning in Dairy Animals." Pakistan Journal of Zoology 51.1 (2019). [3] R. D. Kidd,et al., “Ion Chromatography-on-a-Chip for Water Quality Analysis,” 45th International Conference on Environmental Systems ICES, 12-16 July 2015, Bellevue, Washington, paper # 2015-141. [4] C. B. Freitas, R. C. Moreira, M. G. de Oliveira Travares, W. K. T. Coltro, “ Monitoring of nitrite, nitrate, chloride and sulfate in environmental samples using electrophoresis microchips coupled with contactless conductivity detection,” Talanta, Vol. 147, pp. 335-341 (2016). [5] E. Morcos, N. P. Wiklund, “Nitrite and nitrate measurement in human urine by capillary electrophoresis,” Electrophoresis, Vol. 22, pp. 2763-2768 (2001). [6] N. Oztekin, M. S. Nutku, F. B. Erim, “Simultaneous determination of nitrite and nitrate in meat products and vegetables by capillary electrophoresis,” Food Chemistry, Vol. 76, pp. 103-106 (2002). [7] M. Timm, B. M. Jorgensen, “Simultaneous determination of ammonia, dimethylamine, trimethylamine, and trimethylamine-N-oxide in fish extracts by capillary electrophoresis with indirect UV-detection,” Food Chemistry, Vol. 76, pp. 509-518 (2002). [8] Chastain, John P., et al. "Swine manure production and nutrient content." South Carolina confined animal manure managers certification program. Clemson University, SC (1999): 1-17. [9] Bradford, Scott A., et al. "Reuse of concentrated animal feeding operation wastewater on agricultural lands." Journal of Environmental Quality 37.S5 (2008): S-97. Figure 1

Hydrogen sulfide, Nitrogen Dioxide, and sulfur dioxide are highly toxic gases with health effects at ppb and ppm levels in air. These compounds occur in industrial operations as well as in environmental atmospheric polluted air. Traditionally, microfabricated electrochemical gas sensors use photolithography processes for patterning of the electrodes and defining the electrode geometry for high surface area to increase the sensor gas sensitivity. We present herein a series of gas sensor designs fabricated via laser ablation of sputtered Au films. For these sensors, the laser ablation step itself takes less than two hours to pattern up to 60 sensors. The sensors are fabricated on a porous hydrophobic substrate made of layers of a Teflon woven mesh, and the 400 nm gold film was sputtered onto a 100 nm tungsten adhesion layer. A total of 60 sensors was patterned in a 6 by 10 grid per substrate with each sensor having a footprint of 15 mm x 15 mm. An image of some of the sensors and their corresponding designs are illustrated in figure 1. Laser ablation was performed by an Optec WS-Flex USP femtosecond 1030 nm laser. The pulse frequency was set to 400 kHz and both the speed and jump speed were set to 200 mm/s. The laser power is nominally 4 W and the power percentage setting was determined for each substrate by patterning a test array onto the substrate in increments of 2%. From this test array, the lowest power that completely ablated the gold could be determined and that was used for the remaining sensor ablation. Images of the etched surface area shown in the figure 1. Surface analysis with EDS using SEM microscope indicated the composition of the layers and effective removal of the gold from the Teflon sheet without damage to the Teflon porosity in the woven mesh, figure 2. The sensor electrodes were assembled by lamination between several plastic layers and then an electrolyte was added. The assembled sensors were diced from the assembled 10x6 sensor wafer and individual sensors were connected to a potentiostat and tested via exposure to the various pollutant gases. The current response of these amperometric sensors was measured and found to be linear with respect to concentration in the low ppm range. This work illustrates an alternative to photolithography for the preparation of thin film gas porous electrodes for use in amperometric gas sensors. Figure 1

Gas chromatography is a common technique for the separation and analysis of mixtures of volatile organic compounds (VOCs). In many applications, such as breath analysis and room air sampling, the concentration of analytes is so low that preconcentration is necessary to selectively trap VOCs over a longer period of time, before analysis. The goal of this project is to develop an energy efficient, compact vapor preconcentrator for air sampling. We present thin-film preconcentrators (PCF) with integrated heaters and temperature sensors that were coated with polydimethylsiloxane (OV-1), poly(2,6-diphenylphenylene oxide) (Tenax-TA) and a mixture of OV-1 and Tenax-TA. The polymer thin-films were fabricated using a dynamic coating method with 20mg/ml OV-1 in a 1:1 mixture of DCM and pentane, 10mg/ml of OV-1 and 5mg/ml of Tenax-TA in chloroform, or 20 mg/ml of Tenax-TA in chloroform. The OV-1 and OV-1/ Tenax-TA PCFs were initially tested by connecting them to the inlet and FID of a gas chromatograph with 1 m of 50um guard column. A volume of 0.2 µl of an alkane was injected via the heated inlet 175 oC at 1 psi, after any breakthrough had passed, the GC was pressurized to 8 psi and the PCFs were heated using an integrated platinum film heater. The GC was kept at a constant temperature of 40°C during these tests while the PCF was independently heated outside of the GC oven. The adsorption ratio of these experiments is summarized in table 1. Alkanes above decane such as dodecane proved difficult to desorb, even with the integrated heater reaching temperatures of approximately 260°C. The air sampling experiments for each PCF were performed for decane and benzene using a dynamic headspace sampling method with a constant flow of nitrogen induced by a mass flow controller at a rate of 0.3 ml/min. at a pressure of 5psi. A six-way valve connected the PCF to the GC for desorption at a pressure of 15psi, after a set period of time. Analysis with Agilent 5890GC using a 20 m Rxi-1ms column, temperature ramp from 40°C to 180°C at a rate of 20°C/min, and FID detector. The results for the headspace sampling are summarized in figure 1. For benzene the thin-films with Tenax-TA saturate much more quickly than the OV-1 film and the pure Tenax-TA film which has the lowest capacity. For decane, the thin films with Tenax-TA still saturate faster but the largest capacity was observed for the mix of OV-1 and Tenax-TA over this limited range of sampling times. The authors would like to thank the Georgia Tech Research Institute for funding this research out of Independent Research and Development (IRAD) funds. Figure 1

Low-cost, fast and accurate antibody detection is important for viral infection testing, especially for in cases of epidemics or pandemics. Currently, most rapid antibody detection methods are based on lateral flow immunoassay tests. [1] Beads offer a convenient way to capture and concentrate antigens, antibodies, and other biological molecules but determining the concentration or number can be a challenge, in particular at lower concentration levels. What we would like to achieve is the ability to detect much lower concentration levels, and in principle count individual molecules or a bound complex labelled with the magnetic bead using a micro-solenoid. In this paper, a three-dimensional, hollow channel, micro-solenoid is investigated to improve sensitivity, with antibody testing using magnetic microbeads bound to the target antigen. Previous work has shown the use of magnetic microbeads for microfluidic devices [2]. The micro-solenoid is simulated using COMSOL multiphysics to test various geometric parameters such as bead size, coil size, and bead velocity for the goal of detecting and counting antibodies bound to the beads. A time dependent model was constructed in COMSOL using the magnetic fields and moving mesh modules. In order to simplify the simulation, a 2D axisymmetric model was used and the model was first solved using a stationary step. The bead is represented by a short cylinder, and the solenoid as a uniform diameter annular cylinder, 390 um in length. The coil was simulated through the coil interface under the magnetic fields module allowing it to be represented as single uniform object with variable number of turns, conductivity, and cross-sectional area. For several sets of constant bead size, coil size, and bead magnetization, the maximum voltage reached was found to be a linear function with respect to velocity. In terms of coil radius, the simulated voltage response decreased quadratically with increasing radius however this was only significant when the bead size was appreciable to the coil size for simulations featuring coil radii of 15, 25, 35, and 50 um with 2, 5, 10, and 12.5 um radii beads. The data is presented in figure 1A and 1B. The results for 10 and 12.5 um radius bead are the only ones with appreciable curvature. The simulated voltage response increased quadratically with increasing bead size and the curvature of the underlying quadratic fit increased with decreased coil radius. Future avenues of study include 3 dimensional designs and simulations and fabrication using microfabricated coils. [1] D. A. Mistry, J. Y. Wang, M.-E. Moeser, T. Starkey, and L. Y. W. Lee, “A systematic review of the sensitivity and specificity of lateral flow devices in the detection of SARS-CoV-2.,” BMC Infect. Dis. , vol. 21, no. 1, p. 828, Aug. 2021, doi: 10.1186/s12879-021-06528-3. [2] Kyu Sung Kim and Je-Kyun Park, “Magnetic force-based immunoassay using superparamagnetic nanoparticles in microfluidic channel,” in The 13th International Conference on Solid-State Sensors, Actuators and Microsystems, 2005. Digest of Technical Papers. TRANSDUCERS ’05., Jun. 2005, vol. 1, pp. 81-84 Vol. 1. doi: 10.1109/SENSOR.2005.1496364. Figure 1

Introduction Micro GC systems based upon MEMS fabrication technology have been developed for portable analysis systems to enable detection and quantification of volatile organic compounds over the past twenty years (1). In particular, work at Sandia National Laboratories, has pioneered development of these miniature systems (2). Novel designs for pre-concentrators have also been developed to assist with sample injection into these low flow rate systems (3). We have studied micro-GC columns made by silicon to silicon direct bonding (4), and integrated a micro-GC column 6m length, with an ion trap mass spectrometer for this purpose (5). To inject the sample into the column, our initial method was to use a syringe injection into a heated septum with Agilent 6890. However, for automated operation and method of sampling, pre-concentration of the volatile organic compounds is required, followed by rapid heating for sample injection for analysis. A microfabricated pre-concentrator can fulfill this function by providing a high surface area and a compact platform with reduced thermal mass, compared to commercial desorption tubes. Method We designed a pre-concentrator (MPC) with an integrated platinum heater. The MPC is fabricated using standard photolithography, deep reactive ion etching and wafer bonding. The layout of the geometry in the preconcentrator is shown in Figure 1. Pillars 50 um in diameter are located in the channel to provide a high surface area. The silicon surface is liquid coated with Tenax TA (Tenax 60-80 mesh from Supelco) and PDMS (OV-1 from Ohio Valley Specialty) in an organic solvent, dichloromethane, typically a dozen times to build up a layer of approximate thickness less than a micrometer. Results Sample collection from air flow was modelled with COMSOL and indicates the pressure drop for an air flow of 600sccm a pressure drop of less than 3 psi at room temperature. Figure 3 shows results of the model with a channel depth of 350 um. To evaluate the performance, the pre-concentrator it is mounted inside a commercial Agilent 6890 GC system. A 2m length of Guard Column, 50 um ID was selected to minimize the dead-volume in the fluidic connections between the sample injection port, MPC and the flame ionization detector. Liquid injection of decane at volume of 0.02uL, using a syringe into Agilent 6890 heated injector at 275°C, with a 100:1 split, at a helium flow rate of 0.005 ml/min. Typical results, shown in Figure 4A for heating profile, indicate complete absorption, and no break-through, followed by rapid desorption with heating Figure 4B. The flow rate could be increased at the outlet during the desorption cycle using electrical current of 130 mA to the heater approx.180 ohms, provided rapid heating to 180°C, based upon resistance temperature calibration of the heater and temperature sensor, located on the lid of the MPC. Conclusions This miniature pre-concentrator provided the opportunity to build a small thermal mass designs using silicon microfabrication. With an integrated platinum heater, rapid heating at low power consumption of 5W was achieved, capable of desorption of the captured analyte to detection with FID in GC system. References [1] M. Akbar, M. Restaino, M. Agah, “Chip-scale gas chromatography: From injection through detection,” Microsystems Nanoengineering, Vol. 1, pg. 15938 (2015). [2] in J. J. Whiting, E. B. Myers, R. P. Manginell, M. W. Moorman, K. Pfeifer, J. M. Anderson, C. S. Fix, C. Washburn, A. Staton, D. Porter, D. Graf, D. R. Wheeler, J. Richards, K. E. Achuythan, M. Roukes, R. J. Simonson, "μChem Lab: twenty years of developing CBRNE detection systems with low false “μChemLab: twenty years of developing CBRNE detection systems with low false alarm rates”Proceedings of SPIE, Vo.l. 11010, Chemical, Biological, Radiological, Nuclear, and Explosives (CBRNE) Sensing XX, 1101012 (17 May 2019); doi: 10.1117/12.2518778. [3] Tian, Wei-Chang, H. K. L. Chan, Chia-Jung Lu, S. W. Pang, E. T. Zellers, “ Multiple-stage microfabricated preconcentrator-focuser for micro gas chromatography system,” J. Microelectromachanical Systems, Vol. 14, No. 3, pp 498-507 (2005). [4] M. Navaei, A. Mahdavifar, J.-M. Dimandja, G. McMurray, P. J. Hesketh “All silicon micro GC column temperature programming using axial heating,” Micromachines, Vol 6, pp. 865-878 (2015). [5] Tzu-Hsuan Chang, D. Struk, M. Navaei, V. M. Doroshenko; V. Laiko; E. Moskovets; K. Novoselov, J. D. Dimandja, Peter J. Hesketh, “Separation of Volatile Organic Compounds using of MEMS-GC Integrated Heater for Ion Trap Mass Spectrometer,” revised, Sensors and Actuators B, 2020. Figure 1

In this paper we report the first integration of a silicon microfabricated gas chromatography column with an ion trap mass spectrometer. The MEMS-column is fabricated in all silicon materials, with an integrated platinum resistive heater and temperature sensors. This design enables low power operation and rapid temperature heating, which are both useful for temperature programmed separations. The ion trap mass spectrometer has miniaturized electronics, including a compact RF generator, miniature detector system and control electronics which form the walls of the vacuum chamber. The integrated system has been demonstrated to separate a model mixture containing chemicals from different functional groups in under 22 min, while only consuming less than 15 W of power (30 W with vacuum pump).

Abstract MEMS fabrication technology has been applied to build micro-GC systems for portable analysis and detection of volatile organic compounds for environmental monitoring, medical diagnostics and food safety and security (1). Pioneering work at Sandia National Laboratories developed - these miniature systems (2). Novel designs for pre-concentrators have been developed to provide efficient sample injection into these systems (3). We have integrated a 6m length micro-GC column with an ion trap mass spectrometer (4). To inject the sample into the column, our initial method was to use a syringe injection into a heated septum with Agilent 6890. However, for automated operation and method of sampling, pre-concentration of the volatile organic compounds is required, followed by rapid heating, for sample injection for analysis. A microfabricated pre-concentrator can fulfill this function by providing a high surface area in a compact platform with reduced thermal mass, compared to commercial desorption tubes. The design has an integrated platinum heater and was fabricated using standard photolithography, deep reactive ion etching and wafer bonding. To evaluate the performance, the pre-concentrator is mounted inside a commercial Agilent 6890 GC system. Liquid injections 0.02uL, using a syringe, with a 100:1 split, at a helium flow rate of 0.005 ml/min. The flow rate was increased during the desorption cycle, and the preconcentrator was heated by electrical current of providing approx. 5W to enable rapid heating to 180°C. This miniature pre-concentrator provides a small thermal mass design and was capable of fast desorption of several analytes with detection by FID. References [1] M. Akbar, M. Restaino, M. Agah, “Chip-scale gas chromatography: From injection through detection,” Microsystems Nanoengineering, Vol. 1, pg. 15938 (2015). [2] in J. J. Whiting, E. B. Myers, R. P. Manginell, M. W. Moorman, K. Pfeifer, J. M. Anderson, C. S. Fix, C. Washburn, A. Staton, D. Porter, D. Graf, D. R. Wheeler, J. Richards, K. E. Achuythan, M. Roukes, R. J. Simonson, "μChem Lab: twenty years of developing CBRNE detection systems with low false “μChemLab: twenty years of developing CBRNE detection systems with low false alarm rates”Proceedings of SPIE, Vo.l. 11010, Chemical, Biological, Radiological, Nuclear, and Explosives (CBRNE) Sensing XX, 1101012 (17 May 2019); doi: 10.1117/12.2518778. [3] Tian, Wei-Chang, H. K. L. Chan, Chia-Jung Lu, S. W. Pang, E. T. Zellers, “ Multiple-stage microfabricated preconcentrator-focuser for micro gas chromatography system,” J. Microelectromachanical Systems, Vol. 14, No. 3, pp 498-507 (2005). [4] Tzu-Hsuan Chang, D. Struk, M. Navaei, V. M. Doroshenko; V. Laiko; E. Moskovets; K. Novoselov, J. D. Dimandja, Peter J. Hesketh, “Separation of Volatile Organic Compounds using of MEMS-GC Integrated Heater for Ion Trap Mass Spectrometer,” Sensors and Actuators B, Vol. 307, pg. 127588 (2020).

Introduction Microfabricated gas chromatography systems based upon MEMS fabrication technology are under development for portable chemical vapor analysis and chemical detection. In particular, Sandia National Laboratories has pioneered development of these miniature systems (1). Current methods for coating microfabricated gas chromatography columns are based upon deposition from a saturated mixture of siloxanes in a solvent, which are adapted from techniques developed for coating of open-tubular capillary columns (2). This static method is more successful with silica open tube columns which have a uniform cross-section area, however in a microfabricated column, a rectangular cross-section is produced in the channel. The square corners tend to have a thicker layer of material deposited using this process and hence these thicker layers provide a different retention time for the analytes which results in peak broadening and reduction in resolution and peak capacity of the column. Alternative stationary phases have been investigated (3), and using chemical vapor deposition (CVD) they can be grown with uniform thickness, for example carbon nanotubes (CNTs) (4) The CNTs have high surface area and are stable at high temperatures. Method A silicon substrate is used for the fabrication of a microfabricated spiral column, 6m in length using standard photolithography and deep reactive ion etching fabrication processes (5). An ultra-thin layer of iron catalyst is deposited by sputtering into the etched silicon channel. Using a Black Magic plasma enhanced CVD growth a layer of CNTs are formed thickness 0.5 to 1 micrometer. Figure 1 shows an SEM micrograph of the deposited CNTs layer which coats the sides of channel and base with uniform thickness. Next a layer of gold is deposited for eutectic bonding of the silicon lid in order to enclose the channel making the column. Finally high temperature annealing is carried out at 450 oC to increase bond strength. Results The column is mounted inside a commercial Agilent 6890 GC system oven with a short length of guard column, approximately 0.5m length, connect between the split injector and the flame ionization detector. A liquid injection of three compounds, hexane, octane and decane, volume of 0.02uL, was carried out with a 100:1 split and injector temperature of 275°C. The helium carrier gas flow rate was 0.1 ml/min. A typical chromatogram is shown in Figure 2. The McReynolds probes for polarity were also injected, and the Kovats indices were evaluated for comparison with OV-1 (Ohio Valley, OH) polymer stationary phase. Conclusions The use of CNT’s for a stationary phase may offer some advantages with micro-GC column fabrication because the quality of the layer can be inspected prior to bonding. We can observe the low polarity of the CNT stationary phase compared to the poly-siloxane However, it may be possible to chemically functionalize the surface for the CNTs for more polar compound separation and analysis. References [1] in J. J. Whiting, E. B. Myers, R. P. Manginell, M. W. Moorman, K. Pfeifer, J. M. Anderson, C. S. Fix, C. Washburn, A. Staton, D. Porter, D. Graf, D. R. Wheeler, J. Richards, K. E. Achuythan, M. Roukes, R. J. Simonson, "μChem Lab: twenty years of developing CBRNE detection systems with low false “μChemLab: twenty years of developing CBRNE detection systems with low false alarm rates” Proceedings of SPIE, Vo.l. 11010, Chemical, Biological, Radiological, Nuclear, and Explosives (CBRNE) Sensing XX, 1101012 (17 May 2019); doi: 10.1117/12.2518778. [2] C. F. Poole, The Essence of Chromatography, Elsevier, 2003. [3] I. Azzouz, J. Vial, D. Thiébaut, R. Haudebourg, K. Danaie, P. Sassiat, J. Breviere, “Review of stationary phases for microelectromechanical systems in gas chromatography: feasibility and separations,” Analytical and Bioanalytical Chemistry, Vol. 406, pp. 981–994 (2014). [4] W. Intrchom, S. Mitra, “Analytical sample preparation, preconcentration and chromatographic separation on carbon nanotubes,” Current Opinion in Chemical Engineering, Vol. 16, pp. 102-114 (2017). [5] M. Navaei, A. Mahdavifar, J.-M. Dimandja, G. McMurray, P. J. Hesketh “All silicon micro GC column temperature programming using axial heating,” Micromachines, Vol 6, pp. 865-878 (2015). Figure 1

Miniature gas chromatography (GC) columns of length 6m were fabricated in silicon using eutectic bonding to achieve a column with improved temperature control. The micro GC column includes an integrated platinum thin film heater and temperature sensors for closed loop temperature programing. In order to achieve low power operation the GC column was mounted inside the vacuum chamber of an ion-trap mass spectrometer. This approach provides much lower power consumption than a laboratory instrument. The ion trap electronics were miniaturized and operate at lower power, to demonstrate a portable instrument able to operate at with less than 25W, including vacuum pumps. This instrument will have a range of applications from environmental monitoring, to food safety and medical diagnostics, including breath analysis. Static coating of the stationary phase in the microfabricated column was carried out by injection using a solvent and extraction at reduced pressure to improve resolution. This paper reports the results of different stationary phases investigated to separate mixtures of volatile organic compounds, specifically OV-1 and OV-215. The properties of the stationary phases where characterized using a novel probe mixture for microcolumn evaluation and six McReynolds probes to identify the Kovats indices and the polarity of the column. The dependence of polarity on the coating technique and temperature of the column during analysis was also investigated. Results are presented with both an FID and ion-trap mass spectrometer to achieve rapid temperature programmed analysis of mixtures of up to 18 organic compounds.

This paper is on the development of physical-based gas sensor for detection of highly toxic and corrosive ammonia. COMSOL simulation was used to study the effects of geometry on sensor sensitivity. The uniform cross-section micro-wire is shown to have the highest sensitivity to the surrounding gas. In order to further improve the sensor sensitivity, 3-Omega technique is combined with a differential electrical measurement. The proposed method enhances the sensor's performance for detection of low concentration gas ammonia in nitrogen, and reaches five hundred ppm for ammonia gas.

This paper describes advances in glass substrates for automotive cameras. Typical camera sensors are housed in ceramic packaging and are displaced from related processing components whereas glass substrates allow integration of image sensing and processing components yielding higher data rates while not sacrificing signal strength. Proper design of the glass substrate using thermal vias and coefficient of thermal expansion matching can negate distortion caused by thermal expansion.

This paper reports on characterization and optimization of a low-noise Thermal Conductivity Detector (TCD) for detection of ammonia. Ammonia is a highly toxic agent and is widely produced by of chickens in a chicken farm. There are studies that show the presence of ammonia decreases the quality and quantity of egg and meat production, and may inhibit growth in cell lines. Research has shown that high level of ammonia will create about 5-10% runts in a flock [1]. Therefore, accurate detection ammonia is important issue for food quality control and also labor safety. The commercially available ammonia sensors on the market often have short battery life, baseline drift, selectivity problem, false alarms and need frequent recalibration. Currently most ammonia sensors rely on the chemical interaction of the gas and a specific material which chemical properties changed by the gas. The drawbacks of these sensors are the reliability of the sensors under different environmental conditions and limited life-time. On the other hand, gas sensors which primarily use physical properties of the gas to sense, such as Thermal Conductivity Detector (TCD) is potentially more stable [2]. The working principle of thermal conductivity detectors is based on relating bridge resistivity to its temperature, and bridge temperature depends on the heat loss to the surrounding gas. Here we present a comprehensive COMSOL model to study the effect of the micro-bridge geometry and the heat transfer phenomenal through the gas and substrate. The purpose of the geometry design is to reduce the heat loss to the substrate at the micro-bridge anchors, to improve the efficiency of the sensor, and increase the heat loss through the gas phase to improve the detection limit. For the purpose of this study, we investigated the effect of three differed designed depicted in figure 1. Based on the simulation data, the micro-bridge with the uniform cross-section shows the highest ratio of heat loss through the gas relative to total power of the sensor. This has improved the sensitivity of the sensor by 18% for gases with low thermal conductance such as ammonia. To improve the sensitivity of our TCD sensor by changing the working temperature, the temperature of the sensor has a tremendous effect on the noise level and, consequently, limit of detection (LOD). Ammonia and nitrogen have comparable thermal conductivity at room temperature, but the slope of change of their thermal conductivity is not equal. We have exploited of change in thermal conductivity of gases for different sensor operating temperatures by increasing the sensor power. As a result, the increase of working temperature of sensor has improved sensitivity and reduced the LOD. In conclusion, we have analyzed operation of micro-bridge for robust performance to detect ammonia, which has comparably similar thermal conductivity to the nitrogen. Robust sensitivity of gas detection is also being investigated by 3-Omega techniques, for read-out circuit, increase in the working temperature of the sensor and having better Signal-to-Noise Ratio (SNR). There is also more room in improvement of material properties including having higher thermal coefficient resistivity. Figure 1