Ultratrace Measurement of Acetone from Skin Using Zeolite: Toward

Jul 15, 2015 - Analysis of gases emitted from human skin and contained in human breath has received increasing attention in recent years for noninvasi...
2 downloads 0 Views 4MB Size
Article pubs.acs.org/ac

Ultratrace Measurement of Acetone from Skin Using Zeolite: Toward Development of a Wearable Monitor of Fat Metabolism Yuki Yamada,*,†,‡ Satoshi Hiyama,‡ Tsuguyoshi Toyooka,‡ Shoji Takeuchi,§ Keiji Itabashi,∥ Tatsuya Okubo,∥ and Hitoshi Tabata† †

Department of Bioengineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan Research Laboratories, NTT DOCOMO, Inc., 3-6 Hikarinooka, Yokosuka, Kanagawa 239-8536, Japan § Institute of Industrial Science, The University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8505, Japan ∥ Department of Chemical System Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan ‡

S Supporting Information *

ABSTRACT: Analysis of gases emitted from human skin and contained in human breath has received increasing attention in recent years for noninvasive clinical diagnoses and health checkups. Acetone emitted from human skin (skin acetone) should be a good indicator of fat metabolism, which is associated with diet and exercise. However, skin acetone is an analytically challenging target because it is emitted in very low concentrations. In the present study, zeolite was investigated for concentrating skin acetone for subsequent semiconductor-based analysis. The adsorption and desorption characteristics of five zeolites with different structures and those hydrophobicities were compared. A hydrophobic zeolite with relatively large pores (approximately 1.6 times larger than the acetone molecule diameter) was the best concentrator of skin acetone among the zeolites tested. The concentrator developed using zeolite was applied in a semiconductor-based gas sensor in a simulated mobile environment where the closed space was frequently collapsed to reflect the twisting and elastic movement of skin that would be encountered in a wearable device. These results could be used to develop a wearable analyzer for skin acetone, which would be a powerful tool for preventing and alleviating lifestyle-related diseases.

M

for skin acetone would provide a powerful tool for preventing and alleviating lifestyle-related diseases. However, development of such a device is challenging. This is because skin acetone is typically emitted at concentrations of only several tens of parts per billion (ppb), which is too low for detection by small commercially available sensors. For example, semiconductorbased gas sensors have limits of detection around 200 ppb.4 By contrast, existing high-sensitivity methods for skin acetone analysis use large apparatus, such as gas chromatographs (GCs) or liquid chromatographs combined with skin gas collection bags,13 cooling concentrators,14 trapping filters,15 or solid-phase microextraction cartridges,16−18 which are not suitable for development into a wearable device. To develop a wearable device, either the size of the high-sensitivity devices or the sensitivity requirements of the small devices needs to be reduced. Here, to reduce the detector sensitivity requirements, concentration of the acetone emitted from skin was investigated (Figure 1). Zeolite was selected for concentration

ore than 200 compounds have been identified in human breath, and some of these are associated with various diseases, physiological changes, and physical conditions. Consequently, breath analysis has received increasing attention in recent years for noninvasive clinical diagnoses and health checkups.1,2 Among the many compounds in human breath, acetone is expected to be a good indicator of fat metabolism, which is associated with diet,3,4 aerobic exercise,5,6 and diabetes.7,8 Obesity increases the risk of lifestyle-related diseases, and enabling patient monitoring of breath acetone concentrations could play a pivotal role in day-to-day management of dieting and diabetes control.3,9 However, with breath analysis, it is difficult to motivate patients to blow into a measurement device on a daily basis. Therefore, a simpler method that requires less user input is required. Many compounds, including acetone, are emitted from human skin,10−12 and breath and skin acetone concentrations are correlated.13 Compared with breath acetone, skin acetone is advantageous for analysis because it is continuously emitted and analysis requires no active action on the part of the user, unlike the deep exhalation required for breath analysis. Skin analysis is also more precise than breath analysis as it excludes factors such as the flow rate of exhalation. A wearable analyzer © XXXX American Chemical Society

Received: January 23, 2015 Accepted: July 3, 2015

A

DOI: 10.1021/acs.analchem.5b00296 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 1. Schematic diagram of our proposed method using zeolite to concentrate the very low concentrations of acetone emitted from human skin. (a) Acetone is continuously emitted from human skin. (b) Emitted acetone is adsorbed into zeolite pores. (c) Adsorbed acetone is desorbed by heating the zeolite, and then the acetone is detected by a semiconductor-based gas sensor.

Showa K.K. (Tokyo, Japan). This zeolite has a MFI structure, and its pores are slightly larger than those of FER-156. The zeolites 390HUA, 385HUA, and 350HUA were purchased from Tosoh Corp. (Tokyo, Japan). These zeolites have FAU structures and their pores are larger than those of HISIV3000. The SiO2/Al2O3 ratio of the zeolite was used as an indicator of the zeolite hydrophobicity, with a high ratio corresponding to high hydrophobicity. On the basis of this grading, 390HUA has higher hydrophobicity than 385HUA and 350HUA (see Table 1). Adsorption/Desorption of Pure Acetone Gas into/ from the Zeolites. An aliquot (5 mg) of zeolite powder was placed in a 16.9 mL glass vial (CV-140, Osaka Chemical Co., Ltd., Osaka, Japan) and the open vial was heated at 200 °C for 20 min on a hot plate to remove any volatile impurities from the zeolite. The vial was then sealed and pure acetone gas diluted with nitrogen gas was injected into the vial using a syringe so that the final concentration of acetone in the vial was 800 ppb. This initial acetone concentration was used as a reference to evaluate the acetone adsorption/desorption characteristics of each zeolite, because the concentration in the vial without zeolite samples will stay the same during the experiments. After 15 min, the acetone concentration in the vial was measured by GC (SGEA-P2, FIS Inc., Hyogo, Japan) to determine the acetone adsorption capacity of each zeolite. Then, the vial was opened and the gas it contained was exchanged with clean air. After resealing the vial, the vial was heated at 200 °C for 5 min on the hot plate. The acetone concentration in the sealed vial was then measured using the GC, and this was used to determine how much acetone desorbed from each zeolite. Development of the Concentrator Using Zeolite. Zeolite powder (4.42 g) was mixed with 4.42 g of silica sol (SNOWTEX N30G, Nissan Chemical Industries, Ltd., Tokyo, Japan), 0.35 g of carboxymethyl cellulose ammonium salt (Wako Pure Chemical Industries, Ltd., Osaka, Japan), and 20 mL of distilled water. A platinum coil heater (300 × 300 × 500 μm) was then dipped into the mixture. Any of the mixture that adhered to the coil was air-dried on the coil surface at room temperature. This dip and dry process was repeated several times, and then the coil heater was used to calcine the air-dried mixture at 500 °C for 1 h. The obtained calcined mixture weighed 0.6 mg, and was used to concentrate either pure acetone gas or acetone emitted from the skin of human volunteers. Adsorption/Desorption of Pure Acetone Gas into/ from the Concentrator in Repeated Measurements. The

of skin acetone, which was then detected using a semiconductor-based gas sensor. Zeolites are microporous materials that adsorb gaseous molecules, and are widely used as adsorbents, ion exchangers, and catalysts.19 It is commonly used in automotive exhaust treatment 20 and chemical industries,21 and has been used to coat a quartz crystal microbalance for gas sensing.22 More than 200 structure types of zeolites have been recognized, and it is expected that a specific zeolite could be targeted to efficient acetone adsorption/desorption agent by the selection of appropriate structure type and its hydrophobicity. The aims of this work were: (1) to investigate and compare the adsorption/desorption properties of acetone gas with various zeolites with different structures and those hydrophobicities, and to identify the best zeolite for concentration of skin acetone, (2) to verify whether the concentrated skin acetones can be measured using a semiconductor-based gas sensor, and (3) to confirm the feasibility of the developed method for analysis of acetone emitted from the skin of human volunteers. To the best of our knowledge, this is the first study to demonstrate the concentration of skin acetones using zeolite and their subsequent detection using a semiconductor-based gas sensor. These results could help us in developing a wearable analyzer for skin acetone that could be used for self-monitoring of fat metabolism in daily life.



EXPERIMENTAL SECTION

Zeolites Used in This Study. Five representative zeolites with different structures and hydrophobicities were selected (Table 1). The zeolite FER-156 was synthesized by a research group of one of the authors (T. Okubo) according to a published method.23 This zeolite has a FER structure, and its pore size is comparable with the acetone molecule diameter (4.6 Å). The zeolite HISIV3000 was purchased from Union Table 1. Characteristics of Zeolites Used in This Studya FER-156 HISIV3000 390HUA 385HUA 350HUA

Structure type

Pore size [Å]

SiO2/Al2O3 [mol/mol]

FER MFI FAU FAU FAU

3.5−5.4 5.1−5.6 7.4 7.4 7.4

156 800 500 100 10

a

Codes consisting of three capital letters, such as FER, MFI, and FAU, indicate a topology of zeolite frameworks and are assigned by the structure commission of the International Zeolite Association. B

DOI: 10.1021/acs.analchem.5b00296 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry concentrator was heated at 330 °C for 30 s with the coil heater to remove any volatile impurities from the concentrator and then moved into a 6.4 mL glass vial. The vial was sealed and pure acetone gas diluted with nitrogen gas was injected into the vial using a syringe so that the final concentration of acetone in the vial was 800 ppb. After 15 min, the acetone concentration in the vial was measured using the GC to determine the acetone adsorption capacity of the concentrator. The concentrator was then moved into another 6.4 mL glass vial and the vial was sealed. The concentrator was heated at 330 °C for 30 s and the acetone concentration in the sealed vial was measured using the GC to determine how much acetone desorbed from the concentrator. The above experiments were repeated nine times. Characterization of the Concentrator Using Acetone Emitted from Skin. Skin acetone samples were collected from six healthy male volunteers aged in their 20s and 30s. A capped glass vial with the bottom removed was attached on the skin of the left forearm of each volunteer to create a closed space. This vial contained the concentrator. Acetone samples emitted from the skin were collected and adsorbed into the concentrator over 3, 6, 9, 12, or 15 min. After sample adsorption, the concentrator was transferred into a 16.9 mL glass vial, which was then sealed and heated at 330 °C for 30 s with the coil heater. The acetone concentration in the vial was measured using the GC. Accuracy of Our Proposed Method in Skin Acetone Measurements. Skin acetone samples were collected twice from six healthy volunteers. A capped 16.9 mL glass vial with the bottom removed was attached on the skin of the left forearm of each volunteer to create a closed space (Figure 2).

portion of the concentrator were placed far enough away from the skin surface. In addition, the heated portions were limited to the micrometer scale and thus their radiant heat did not affect the skin surface. The acetone concentration in the vial was measured using the acetone sensor. For comparison, acetone samples emitted from the skin were collected over 10 min using the same glass vial without the concentrator and acetone sensor. The acetone concentration in the vial was measured using the GC. Skin Acetone Sensing under Real Conditions. A 6.4 mL bottomless glass vial containing the concentrator and the acetone sensor was attached to the skin of the left forearm of a human volunteer to create a closed space. Skin acetone was collected over 20 min. The closed space was intentionally collapsed in a cyclic manner during the collection. This was performed by manually detaching the vial to open the closed space, and then attaching it again to close the space. The open/ close process was repeated every 30 s during the 20 min collection period. This process replicates the conditions that could occur in a wearable device during daily activities, where twisting and elastic movement of the skin surface could temporarily collapse the closed space of the sensor. After the collection time was complete, the concentrator was heated at 330 °C for 30 s using the coil heater to desorb the adsorbed skin acetone. These processes were repeated three times in succession, and changes in the sensitivity of the acetone sensor were continuously monitored during the experiments. For comparison, the same experiments were conducted without equipping the vial with the concentrator.



RESULTS AND DISCUSSION Five representative zeolites were studied to determine how pore size and hydrophobicity affected the concentration of skin acetone. The adsorption and desorption characteristics of FER156, HISIV3000, and 390HUA zeolites for pure acetone gas were compared (Figure 3). With an acetone concentration of 880 ppb in a sealed glass vial, more than 800 ppb acetone was adsorbed by all three zeolites after a 15 min incubation period (Figure 3a). These results indicate that approximately 92% of the acetone was adsorbed into the zeolites, and their adsorption capacities were similar. In contrast, desorption of acetone by heating was different for the three zeolites. The zeolite 390HUA showed much better desorption performance than the other zeolites, and approximately 85% of the adsorbed acetone was desorbed from this zeolite under the present conditions (Figure 3b). The differences in desorption are presumably caused by the structural differences of the zeolites. Acetone would desorb from zeolite 390HUA more easily than the other zeolites because its pores are approximately 1.6 times larger than the acetone molecule diameter. If the zeolite pores are too large, acetone and the zeolite will not strongly interact. This indicates that there is a trade-off between adsorption capacity and desorption kinetics, and this could be used to select a zeolite to act as the best concentrator for a specific molecule. To investigate the effect of the differences in zeolite hydrophobicity, the adsorption and desorption characteristics of pure acetone gas with the FAU type zeolites (390HUA, 385HUA, and 350HUA) were compared (Figure 4). When an acetone concentration of 780 ppb was generated in the sealed glass vial, more than 740 ppb acetone was adsorbed by all three of the zeolites after a 15 min incubation period (Figure 4a).

Figure 2. Photograph of skin acetone collection and measurement from human skin on the left forearm. A closed space is created using a glass vial containing the concentrator and acetone sensor. The inset shows a photograph of the concentrator, which is composed of 390HUA zeolite. The scale bar corresponds to 500 μm.

This vial contained the concentrator and a semiconductorbased acetone sensor4 made of platinum-doped tungsten oxide (FIS Inc., Hyogo, Japan). We used a sensor that has high sensitivity for acetone and low sensitivity for the other gases emitted from human skin, such as acetaldehyde and ammonia (Figure S1, Supporting Information). Acetone samples emitted from the skin were collected and adsorbed into the concentrator over 10 min. After sample adsorption, the concentrator was heated at 330 °C for 30 s with the coil heater equipped inside the concentrator. The volunteers did not feel heat because the heater and the corresponding heated C

DOI: 10.1021/acs.analchem.5b00296 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 3. (a) Acetone (pure gas) adsorption characteristics of three zeolites with different structures. The error bars represent the standard error from three measurements. (b) Desorption characteristics of the three zeolites. The error bars represent the standard error from three measurements.

Figure 4. (a) Acetone (pure gas) adsorption characteristics of three zeolites with different hydrophobicities. The error bars represent the standard error from three measurements. (b) Desorption characteristics of the three zeolites. The error bars represent the standard error from three measurements.

These results indicate that approximately 95% of the acetone was adsorbed into the zeolites, and their adsorption capacities were similar. Desorption of acetone from the three zeolites by heating was quite different. Zeolite 390HUA showed much better desorption performance than the other zeolites (Figure 4b). Approximately 81% of the adsorbed acetone desorbed from zeolite 390HUA, whereas only around 3% of the adsorbed acetone desorbed from zeolite 350HUA. These results indicate that hydrophilic zeolites, such as 350HUA, strongly adsorb polar molecules such as acetone. These results show that a hydrophobic zeolite with relatively large pores will be most suitable for use as a concentrator of skin acetone. Among the five zeolites examined in this study, 390HUA is the best choice as a concentrator of acetone. Next, the acetone adsorption capacity and the reproducibility of acetone adsorption/desorption of the calcined 390HUA concentrator were investigated (Figure 5). The concentrator used in this study (0.6 mg) was able to adsorb up to 688 ppb pure acetone gas in an average incubation time of 15 min for 843 ppb of generated pure acetone gas. This adsorption capacity is feasible because the skin acetone emission rate per unit time and area will be 2 to 3 ppb, and it will thus take 3.8 to 5.7 h to reach 688 ppb collected skin acetone gas. Assuming that our proposed method will be implemented in wearable devices to monitor fat metabolism, a 15 min to 1 h collection time will be sufficient for this application and saturation of the zeolite pores will not occur in such a collection time. We also found that 93% ± 3% of the adsorbed pure acetone desorbed

Figure 5. Adsorption and desorption characteristics of the concentrator for pure acetone in repeated measurements.

from the concentrator for the nine distinct measurements. These results indicate that our proposed method has high reproducibility. We then verified adsorption/desorption of acetone with the concentrator made of 390HUA zeolite in human volunteers. Gas samples were collected and adsorbed into 390HUA over 3, D

DOI: 10.1021/acs.analchem.5b00296 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry 6, 9, 12, or 15 min. Figure 6 shows that three representative patterns of the emission flux of skin acetone observed from

Figure 7. Scatter plots of skin acetone concentrations in 10 min collection time detected by our proposed system (concentrator and acetone sensor) and by GC. Figure 6. Relationship between acetone collection time and concentration of skin acetone desorbed from the concentrator. The experiments were conducted in the order of decreasing collection time, and thus each of the lapse times from the start of the experiment corresponds to the sum of the collection times.

proposed method tended to be slightly lower than those obtained from the GC. This is presumably because some of the adsorbed skin acetone molecules remain in the concentrator, as shown in Figure 5. To investigate the feasibility of our proposed method, skin acetone was measured using the semiconductor-based acetone sensor with and without the concentrator (390HUA zeolite) under simulated real life conditions. To achieve this, the closed space was intentionally collapsed in a cyclic manner during the 20 min gas collection (Figure 8). The sensitivity of the acetone

three volunteers among the six. Results obtained from the rest of three volunteers were similar to either Subject A or Subject C, and thus these results were not shown in this figure for clarity. Assuming that the emission rate of skin acetone per unit time and area is constant, the relationship between desorbed acetone and collection time should theoretically be linear. The emission rate, however, fluctuates over time according to the lifestyle behavior, such as the diet and exercise of each subject. Considering that the experiments were conducted in the order of decreasing collection time, the emission rates for the longer collection times (e.g., 12 and 15 min, which correspond to 30 and 45 min in lapse time from the start of the experiment) may have changed from the rate at the start of the experiment. Thus, we assumed that the emission rate was constant during the first 9 min in lapse time from the start of the experiment, and found that the skin acetone concentration desorbed from the 390HUA zeolite linearly increased with increasing collection time. These results show that 390HUA zeolite can adsorb/ desorb skin acetone as well as pure acetone gas. The obtained results that the skin acetone concentration desorbed from the 390HUA zeolite did not linearly increase after the first 9 min in lapse time indicate that the emission rates significantly increased in Subject B and decreased in Subject C. The slopes of the curves in Figure 6 show the differences in skin acetone emission among the volunteers. Among the six volunteers, Subject B emitted acetone at the highest rate per unit time and area during the experiments. To investigate the accuracy of our proposed method, the concentrations of skin acetone in 10 min collection time detected by our proposed system (with the concentrator and acetone sensor) and those determined by GC were compared (Figure 7). The plots show that there is a strong correlation between the concentrations of skin acetone obtained from our proposed method and those obtained by the GC (R2 = 0.90, P < 0.001), confirming that our proposed method is practical with a reasonable range of measurement deviation. It should be noted that the skin acetone concentrations obtained from our

Figure 8. Continuous monitoring of the sensitivities of the semiconductor-based acetone sensor with and without the concentrator under simulated real conditions. The closed space was intentionally collapsed in a cyclic manner during each 20 min collection time. Acetone detection is possible when the sensitivity is less than 0.8.

sensor without the concentrator drastically decreased from 1.00 to 0.83 within the first 5 min from the start of the measurement because of the increased concentration of skin acetone in the closed space. It then plateaued at around 0.83 because of the intentional and cyclic collapsing of the closed space. Such collapsing is very likely in daily life, where some activities are accompanied by twisting and elastic movement of the skin surface. According to the results of the six volunteers in Figure E

DOI: 10.1021/acs.analchem.5b00296 Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry



6, the skin acetone emission rate per unit time and area was 2 to 3 ppb in average, and it will thus take 7 to 10 min to reach 20 ppb of collected gas. This indicates that it will be difficult to measure skin acetone using only a sensor even if state-of-the-art acetone sensors that can measure 20 ppb of breath acetone24 are used, because such collapsing will take place in real life during the gas collection time of 7 to 10 min. After the 20 min collection time, the sensitivity increased back to 1.00 because the closed space was opened and the enclosed gas was completely exchanged with clean air. The sensitivity is indicated by the ratio of the electrical resistance of the gas sensor in air (Rair) to the electrical resistance of the gas sensor in the target gas (R). Thus, it will be difficult to clearly distinguish the signal from the noise as R/Rair → 1. We set the limit of detection to R/Rair = 0.8 to avoid false detection and to conduct reliable gas sensing. Considering that the sensitivities of the acetone sensor without the concentrator were always R/ Rair > 0.8, we concluded that the acetone sensor alone would not be able to detect skin acetone under the simulated conditions. In comparison, the sensitivities of the acetone sensor with the concentrator were all much less than 0.8 when the adsorbed skin acetone was desorbed every 20 min. Therefore, skin acetone could be easily detected under the simulated conditions with the concentrator made from zeolite. We also found that almost the same peak sensitivities of the acetone sensor with the concentrator (0.53 ± 0.01) were obtained from the three measurements. These results indicate that our proposed method has high reproducibility even if the closed space for skin acetone collection is in high humidity because of the water vapor emitted from the skin surface. It should be noted that the sensitivities of the acetone sensor with the concentrator also plateaued after the first 5 min, and were higher than those of the acetone sensor without the concentrator. This is because skin-emitted acetone adsorbed into the concentrator and the adsorbed acetone was retained until the concentrator was heated for desorption. The ratio of skin acetone measured by the acetone sensor with the concentrator to skin acetone measured by the acetone sensor without the concentrator was 7.5. This magnitude of concentrated skin acetone could help to reduce the detector sensitivity requirement. These results indicate that our proposed method is feasible and shows potential for implementation in wearable devices.

Article

ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in the text. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b00296.



AUTHOR INFORMATION

Corresponding Author

*Y. Yamada. E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors thank Mariko Hanada, Kazuo Onaga, Junko Yanagitani, Aki Uesaka, Hitoshi Katayama, Dr. Katsuyuki Tanaka, and Takeo Tsunemi of FIS Inc. for their help in developing the concentrators.



REFERENCES

(1) Manolis, A. Clin. Chem. 1983, 29, 5−15. (2) Cao, W.; Duan, Y. Clin. Chem. 2006, 52, 800−811. (3) Kundu, S. K.; Bruzek, J. A.; Nair, R.; Judilla, A. M. Clin. Chem. 1993, 39, 87−92. (4) Toyooka, T.; Hiyama, S.; Yamada, Y. J. Breath Res. 2013, 7, 036005. (5) King, J.; Kupferthaler, A.; Koc, H.; Teschi, S.; Teschi, G.; Mekisch, W.; Schubert, J.; Hinterhuber, H.; Amann, A. J. Breath Res. 2009, 3, 027006. (6) Sakai, H.; Ishikawa, S.; Ueda, H.; Kimura, Y. Adv. Exerc. Sports Physiol. 2011, 16, 97−100. (7) Wang, C.; Mbi, A.; Sphepherd, M. IEEE Sensors J. 2010, 10, 54− 63. (8) Storer, M.; Dummer, J.; Lunt, H.; Scotter, J.; McCartin, F.; Cook, J.; Swanney, M.; Kendall, D.; Logan, F.; Epton, M. J. Breath Res. 2011, 5, 046011. (9) Righettoni, M.; Tricoli, A.; Pratsinis, S. E. Anal. Chem. 2010, 82, 3581−3587. (10) Mochalski, P.; King, J.; Unterkofler, K.; Hinterhuber, H.; Amann, A. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2014, 959, 62−70. (11) Bernier, R. U.; Kline, L. D.; Bamard, R. D.; Schreck, E. C.; Yost, A. R. Anal. Chem. 2000, 72, 747−756. (12) Dormont, L.; Bessière, J.-M.; Cohuet, A. J. Chem. Ecol. 2013, 39, 569−578. (13) Turner, C.; Parekh, B.; Walton, C.; Spanel, P.; Smith, D.; Evans, M. Rapid Commun. Mass Spectrom. 2008, 22, 526−532. (14) Naitoh, K.; Inai, Y.; Hirabayashi, T.; Tsuda, T. Anal. Chem. 2000, 72, 2797−2801. (15) Sekine, Y.; Toyooka, S.; Watts, F. S. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2007, 859, 201−207. (16) Musteata, M. F.; Pawliszyn, J. J. Biochem. Biophys. Methods 2007, 70, 181−193. (17) Duan, C.; Shen, Z.; Wu, D.; Guan, Y. TrAC, Trends Anal. Chem. 2011, 30, 1568−1574. (18) Dormont, L.; Bessière, J.-M.; McKey, D.; Cohuet, A. J. Exp. Biol. 2013, 216, 2783−2788. (19) Roth, J. W.; Nachtigall, P.; Morris, E. R.; Č ejka. Chem. Rev. 2014, 114, 4807−4837. (20) Deka, U.; Lezcano-Gonzalez, I.; Weckhuysen, M. B.; Beal, M. A. ACS Catal. 2013, 3, 413−427. (21) Perego, C.; Ingallina, P. Green Chem. 2004, 6, 274−279. (22) Huang, H.; Zhou, J.; Chen, S.; Zeng, L.; Huang, Y. Sens. Actuators, B 2004, 101, 316−321. (23) Kamimura, Y.; Kowenje, C.; Yamanaka, K.; Itabashi, K.; Endo, A.; Okubo, T. Microporous Mesoporous Mater. 2013, 181, 154−159.



CONCLUSIONS A skin acetone measurement method was developed using zeolite to concentrate the acetone gas before analysis with a semiconductor-based gas sensor. Among the five zeolites tested, 390HUA zeolite, a hydrophobic zeolite with relatively large pores, was the best for concentration of skin acetone. Concentration of the acetone by the zeolite and subsequent semiconductor-based gas sensor analysis was achieved in a simulated mobile environment, where the closed space was frequently collapsed to simulate the twisting and elastic movement of the skin surface that could occur with a wearable device. These results could be used to develop a wearable analyzer for skin acetone that could be used for self-monitoring of fat metabolism in daily life. In the future, we will further investigate the characteristics of our proposed system under different environmental conditions, such as temperature, humidity, and gas composition. The next major step will be miniaturization and implementation in a wearable device, such as a wristwatch. F

DOI: 10.1021/acs.analchem.5b00296 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry (24) Righettoni, M.; Tricoli, A.; Gass, S.; Schmid, A.; Amann, A.; Pratsinis, S. E. Anal. Chim. Acta 2012, 738, 69−75.

G

DOI: 10.1021/acs.analchem.5b00296 Anal. Chem. XXXX, XXX, XXX−XXX