Statistical analysis of DNT detection using chemically functionalized microcantilever arrays

Filippo Bosco, M. Bache, E.-T. Hwu, C.H. Chen, S.S. Andersen, K.A. Nielsen, Stephan Sylvest Keller, Jakob Jeppesen, I.-S. Hwang, Anja Boisen

    Research output: Contribution to journalJournal articleResearchpeer-review


    The need for miniaturized and sensitive sensors for explosives detection is increasing in areas such as security and demining. Micrometer sized cantilevers are often used for label-free detection, and have previously been reported to be able to detect explosives. However, only a few measurements from 1 to 2 cantilevers have been reported, without any information on repeatability and reliability of the presented data. In explosive detection high reliability is needed and thus a statistical measurement approach needs to be developed and implemented. We have developed a DVD-based read-out system capable of generating large sets of cantilever data for vapor and liquid phase detection of 2,4-dinitrotoluene (DNT). Gold coated cantilevers are initially functionalized with tetraTTF-calix[4]pyrrole molecules, specifically designed to bind nitro-aromatic compounds. The selective binding of DNT molecules on the chemically treated surfaces results in significant bending of the cantilevers and in a decrease of their resonant frequencies. We present averaged measurements obtained from up to 72 cantilevers being simultaneously exposed to the same sample. Compared to integrated reference cantilevers with non-selective coatings the tetraTTF-calix[4]pyrrole functionalized cantilevers reveal a uniform and reproducible behavior.
    Original languageEnglish
    JournalSensors and Actuators B: Chemical
    Pages (from-to)1054-1059
    Publication statusPublished - 2012


    • Explosives detection
    • Cantilever sensor
    • DVD-ROM
    • Optical read-out
    • Calix[4]pyrrole
    • Tetrathiafulvalene


    Dive into the research topics of 'Statistical analysis of DNT detection using chemically functionalized microcantilever arrays'. Together they form a unique fingerprint.

    Cite this