Application of a Conductive Polymer Electronic-nose Device to Identify Aged Woody Samples Alphus Dan Wilson Forest Insect and Disease Research USDA Forest Service, Southern Hardwoods Laboratory Stoneville, MS, USA e-mail: dwilson02@fs.fed.us Abstract— The identification of aged woody samples is often a difficult task as a result of weathering and physical deterioration over time which removes or obscures distinguishing anatomical features and characteristics required for visual taxonomic determinations. Fortunately, the chemical characteristics of aged woods usually are preserved better than physical characteristics if the wood remains dry in storage. All wood types, determined by the particular plant species from which woody samples are derived, produce and release a unique complex of volatile organic compounds that distinguish individual wood types when headspace volatiles (containing these unique chemical mixtures) are collectively analyzed using an electronic gas-sensing device such as an electronic nose. The advantage of electronic-nose devices over conventional analytical-chemistry instruments, typically used in laboratory chemical analyses, is that the woody source (plant species) from which headspace volatiles are derived may be identified without having to identify individual chemical compounds present in the headspace analyte mixture. Methods were developed for a conductive polymer type electronic nose gas- sensing device, the Aromascan model A32S, to accurately identify aged woody samples derived from wood pieces held in dry storage for long periods of time. An aroma library was developed using diagnostic aroma profile databases (electronic aroma signature patterns) from known woods of numerous tree species. The A32S electronic nose was capable of distinguishing between 44 wood types, providing correct identification determinations at frequencies ranging from 92- 99%. The distribution of aroma class components, defined by wood type for each sample analyzed, also could be determined to indicate the relatedness of volatile aroma components that each sample analyte had in common with individual wood aroma classes. This information was useful for determining the taxonomic relatedness of wood types (plant species) based on the headspace volatiles that were produced. Furthermore, principal component analysis provided precise statistical numerical values (quality factors of significance) that indicated the chemical relatedness between wood volatiles based on pairwise comparisons of organic chemical mixtures from individual wood types. Keywords- electronic aroma detection; e-nose technologies; volatile organic compounds; wood identification I. INTRODUCTION There are numerous situations where the identity of wood types must be known for many commercial or industrial applications, scientific research, or forensic analyses. Conventional methods used for the determination of wood type identities involve examinations of macroscopic and microscopic anatomical characteristics of wood tissues. The identification of wood types becomes more difficult if the wood is exposed to adverse environmental conditions (such as weathering) that result in physical deterioration of the wood, masking or diminishing diagnostic anatomical characteristics required for visual taxonomic determinations. By contrast, most chemical characteristics of wood are not lost during aging or weathering as long as the wood is stored in a dry state, even over extended periods of time. Traditional chemical and microscopic methods used for wood identification hitherto are cumbersome and less reliable because they often require extensive and expensive sample preparation and time-consuming analyses. Thus, there is a need for an analytical device that quickly identifies organic samples such as wood types without the high cost of conventional chemical analyses. Electronic-nose devices are designed to produce digital electronic signatures of volatile organic compounds (VOCs) released from any source [1-3]. Unlike other analytical instruments, these devices have the capability of identifying organic samples from the VOCs they release without having to identify individual chemical compounds present in volatile mixtures [4-6]. A variety of different sensor types have been developed for various applications including optical sensors [7], metal oxides [8, 9], semiconductive polymers [10-13], and conductive polymers [14-15]. The agricultural and food industries have utilized electronic aroma detection (EAD) technologies to evaluate food quality and product aromas [16-18], food storage life and freshness [19-20], detect industrial wastes [21-22], diagnose plant diseases [23], and for many other applications requiring gas-detection technologies [24-27]. The purposes of this study were to 1) determine if an electronic-nose (e-nose) device, the conductive polymer (CP)-type Aromascan A32S e-nose, has the capability of identifying and distinguishing between dried aged specimens of various temperate North American wood types based on headspace volatiles (given the reduced amount of volatiles released from aged wood), to 2) evaluate the effectiveness (accuracy) of wood-type determinations, and to 3) assess whether e-nose aroma data outputs provide indications of 77 Copyright (c) The Government of USA, 2012. Used by permission to IARIA. ISBN: 978-1-61208-208-0 SENSORDEVICES 2012 : The Third International Conference on Sensor Device Technologies and Applications