sensors Article
Towards the Development of a Low-Cost Device for the Detection of Explosives Vapors by Fluorescence Quenching of Conjugated Polymers in Solid Matrices Liliana M. Martelo 1,2, * ID , Tiago F. Pimentel das Neves 3 , João Figueiredo 3 , Lino Marques 3 , Alexander Fedorov 3 ID , Ana Charas 4 , Mário N. Berberan-Santos 2 and Hugh D. Burrows 1 1 2 3
4
*
Department of Chemistry, University of Coimbra, 3004-535 Coimbra, Portugal;
[email protected] Centro de Química-Física Molecular (CQFM) and the Institute of Nanoscience and Nanotechnology (IN), Instituto Superior Técnico, University of Lisbon, 1049-001 Lisbon, Portugal;
[email protected] Institute of Systems and Robotics (ISR), University of Coimbra, 3030-290 Coimbra, Portugal;
[email protected] (T.F.P.d.N.);
[email protected] (J.F.);
[email protected] (L.M.);
[email protected] (A.F.) Instituto de Telecomunicações, Instituto Superior Técnico, Av. Rovisco Pais, 1049-001 Lisbon, Portugal;
[email protected] Correspondence:
[email protected]; Tel.: +351-218-419-259
Received: 28 September 2017; Accepted: 30 October 2017; Published: 3 November 2017
Abstract: Conjugated polymers (CPs) have proved to be promising chemosensory materials for detecting nitroaromatic explosives vapors, as they quickly convert a chemical interaction into an easily-measured high-sensitivity optical output. The nitroaromatic analytes are strongly electron-deficient, whereas the conjugated polymer sensing materials are electron-rich. As a result, the photoexcitation of the CP is followed by electron transfer to the nitroaromatic analyte, resulting in a quenching of the light-emission from the conjugated polymer. The best CP in our studies was found to be poly[(9,9-dioctylfluorenyl-2,7-diyl)-co-bithiophene] (F8T2). It is photostable, has a good absorption between 400 and 450 nm, and a strong and structured fluorescence around 550 nm. Our studies indicate up to 96% quenching of light-emission, accompanied by a marked decrease in the fluorescence lifetime, upon exposure of the films of F8T2 in ethyl cellulose to nitrobenzene (NB) and 1,3-dinitrobenzene (DNB) vapors at room temperature. The effects of the polymeric matrix, plasticizer, and temperature have been studied, and the morphology of films determined by scanning electron microscopy (SEM) and confocal fluorescence microscopy. We have used ink jet printing to produce sensor films containing both sensor element and a fluorescence reference. In addition, a high dynamic range, intensity-based fluorometer, using a laser diode and a filtered photodiode was developed for use with this system. Keywords: conjugated polymers; explosives detection; trace analysis; optical sensor; luminescence sensor
1. Introduction Part of the extensive research in conjugated polymers (CPs) and conjugated polyelectrolytes (CPEs) is motivated by their capacity as sensitive fluorescent materials for chemo- and biosensing. They offer a broad range of possibilities for transforming analyte receptor interactions, as well as nonspecific interactions, into observable (transducible) responses [1,2]. Amplified quenching in fluorescent CP was introduced by Swager and Zhou [3] and opened the way for novel sensory materials using this important class of conjugated polymers. In 1998, Yang and co-workers [4] used a fluorescence quenching transduction mechanism together with the amplifying nature of conjugated polymers to develop a material highly sensitive to 2,4,6-trinitrotoluene (TNT) vapors, the major explosive Sensors 2017, 17, 2532; doi:10.3390/s17112532
www.mdpi.com/journal/sensors
Sensors 2017, 17, 2532
2 of 12
Sensors 2017, 17, 2532
2 of 13 explosive component of landmines. One peculiarity of nitroaromatics which may be used in detection based on fluorescence techniques is their electron-accepting capability. CPs are promising for redox sensing because they are normally electron donors. donor is further component of landmines. One peculiarity of nitroaromatics which This may be usedbehavior in detection based enhanced in their delocalized pi excited states. This excited state delocalization is crucial because the on fluorescence techniques is their electron-accepting capability. CPs are promising for redox sensing resultingthey exciton migration alongdonors. the polymer chainbehavior increases the frequency of in interaction with a because are normally electron This donor is further enhanced their delocalized bound quencher, in this case the nitroaromatic analytes, which contributes to improve detection pi excited states. This excited state delocalization is crucial because the resulting exciton migration sensitivity. For these reasons, nitroaromatic analytes can efficiently emission by along the polymer chain increases the frequency of interaction with aquench boundthe quencher, in of thisCP case photoinduced electron transfer process. As a practical result, photoexcitation of the conjugated the nitroaromatic analytes, which contributes to improve detection sensitivity. For these reasons, polymer is followed bycan electron transfer to thethe nitrated organic compounds, resulting in a quenching nitroaromatic analytes efficiently quench emission of CP by photoinduced electron transfer of the CP fluorescence. Fluorescence quenching sensing methods are promising for rapid and process. As a practical result, photoexcitation of the conjugated polymer is followed by electron transfer sensitive detection of explosives vapors, and possess major advantages, including high sensitivity to the nitrated organic compounds, resulting in a quenching of the CP fluorescence. Fluorescence signal output and operational simplicity [5]. for rapid and sensitive detection of explosives vapors, quenching sensing methods are promising The detection of explosives is a high majorsensitivity quest forsignal security in and many civilian simplicity and military and possess major advantages, including output operational [5]. environments, and is usually carried out through the sensing of the vapor emitted by the explosives, The detection of explosives is a major quest for security in many civilian and military or of markers and present with them. sensorsthe must satisfy several criteria, as explosives, sensitivity, environments, is usually carriedThese out through sensing of the vapor emittedsuch by the reversibility and the capability for real-time signal processing. For nitroaromatic explosives, sensing or of markers present with them. These sensors must satisfy several criteria, such as sensitivity, of a few parts per billion or less of the analyte vapor is mandatory, and should be accompanied with reversibility and the capability for real-time signal processing. For nitroaromatic explosives, sensing of rapid and, ideally, reversible changes in the sensor output. Some assessments of explosives a few parts per billion or less of the analyte vapor is mandatory, and should be accompanied with rapid containing have been performed, and output. it has been thatofthe concentration of TNT is and, ideally,soils reversible changes in the sensor Someindicated assessments explosives containing soils around 10–100 ng/kg. The vapor concentration is even lower, around the 100 pg/kg to have been performed, and it has been indicated that the concentration of TNT is around 10–100 ng/kg. 100 vapor fg/kg level [6]. For in-field of such system bein-field highly The concentration is even detection lower, around the materials, 100 pg/kga toportable 100 fg/kg levelwould [6]. For beneficial. detection of such materials, a portable system would be highly beneficial. In order issues, we we havehave developed a newa conjugated polymer-based optical In order to toaddress addressthese these issues, developed new conjugated polymer-based sensor of trace explosives vapors. For the chemosensory material, we have used hairy-rod optical sensor of trace explosives vapors. For the chemosensory material, we have used hairy-rod polymers [7], [7], an an important important class class of of π-conjugated π-conjugated polymers, polymers, such such as as poly(fluorene-2,7-diyl)s poly(fluorene-2,7-diyl)s (PFs). (PFs). polymers These have haveexcellent excellent photoluminescence quantum yields, good thermal good These photoluminescence quantum yields, good thermal stability,stability, and goodand solubility solubility in several solvents [8]. The linear side chain poly[9,9-dioctylfluorene-2,7-diyl] (PFO) in several solvents [8]. The linear side chain poly[9,9-dioctylfluorene-2,7-diyl] (PFO) and and its its homologue poly[9,9-dioctylfluorenyl-2,7-diyl)-co-bithiophene]) (F8T2)were wereused usedinin this this study study homologue poly[9,9-dioctylfluorenyl-2,7-diyl)-co-bithiophene]) (F8T2) (Figure1).1).Detailed Detailed spectroscopic photophysical properties these polymers been (Figure spectroscopic and and photophysical properties of theseofpolymers have beenhave presented presented elsewhere [9]. elsewhere [9].
Figure 1. Structures of the CPs used: (A) poly[9,9-dioctylfluorene-2,7-diyl] (PFO); (B) poly[(9,9Figure 1. Structures of the CPs used: (A) poly[9,9-dioctylfluorene-2,7-diyl] (PFO); (B) dioctylfluorenyl-2,7-diyl)-co-bithiophene] (F8T2); and (C) poly[(9,9-dioctylfluorenyl-2,7-diyl)-alt-copoly[(9,9-dioctylfluorenyl-2,7-diyl)-co-bithiophene] (F8T2); and (C) (1,4-benzo-2,10,3- thiadiazole)] (F8BT). poly[(9,9-dioctylfluorenyl-2,7-diyl)-alt-co-(1,4-benzo-2,10,3- thiadiazole)] (F8BT).
For For many many practical practical applications, applications, itit is is desirable desirable to to incorporate incorporate CPs CPs in in an appropriate appropriate porous porous inert inert matrix. In this work, we used ethyl cellulose (EC) to incorporate the CPs. Both CPs exhibit high matrix. In this work, we used ethyl cellulose (EC) to incorporate the CPs. Both CPs exhibit high sensitivity sensitivity in in ethyl ethyl cellulose films when exposed to nitrobenzene (NB) and 1,3-dinitrobenzene (DNB) vapors. These vapors. These are are chosen chosen as as models models or or markers markers of of more more common common nitroaromatic nitroaromatic explosives, explosives, such such as as TNT TNT or or RDX. RDX. EC EC is is the the most most common common insoluble insoluble cellulose cellulose derivative derivative used used and and is is available available in in a variety of of viscosity viscosity grades, grades, according according to to the the molecular molecular weight weight range range of of the the products. products. The The molecular molecular weight weight affects the mechanical properties, which have fundamental importance for producing intact films, affects the mechanical properties, which have fundamental depending on the application [10]. Plasticizers are generally used to improve the mechanical properties
Sensors 2017, 17, 2532
3 of 13
of a polymer matrix. This occurs because the plasticizer can decrease the intramolecular forces between the polymer chains, reducing the glass transition temperature and increasing the permeability of the polymer matrix to gases or other analytes [11]. In this work, we use polyethylene glycol and polypropylene glycol with different molecular weights as plasticizers to improve the mechanical properties and permeability of our polymer matrices. This contribution is divided in two parts. First, we report absorption, emission spectra, and fluorescence lifetimes of the PFO and F8T2 in ethyl cellulose films, the structural characterization of the thin films, and then discuss the ability of these materials to sense TNT model compounds. In the second part, we study the improved polymer matrices produced by the introduction of plasticizers which increase the sensitivity to TNT-like compounds when compared with the non-plasticized ones. We also develop other methods for CP device preparation in the solid matrix, such as ink jet printing technology: in this case we added an internal reference, a CP whose fluorescence is not quenched by the TNT-like molecules, to provide potential for ratiometric sensing. In this condition, we print different zones with the two CPs, and use as “paper” the non-plasticized ethyl cellulose matrix. 2. Materials and Methods 2.1. Materials Ethyl cellulose of viscosity grade 100 cP, was acquired from Sigma-Aldrich (St. Louis, MO, USA) and used without any treatment. The conjugated polymers, poly[9,9-dioctylfluorene-2,7-diyl] (PFO, Mw ≥ 20,000) poly[9,9-dioctylfluorenyl-2,7-diyl)-co-bithiophene]) (F8T2, Mn > 20,000) and poly(9,9-dioctylfluorene-alt-benzothiadiazole) (F8BT, Mn ± 17,000-23,000) were from Sigma-Aldrich. Solutions for film preparation were made by dissolving ethyl cellulose and the CPs (200–500 ppm) in toluene (GPS grade, Carlo Erba Reagents) at room temperature. Nitrobenzene (ACS reagent, 99%) and 1,3-dinitrobenzene (99%) were from Sigma-Aldrich. 2.2. Film Preparation and Ethyl Cellulose Plasticization Films were prepared by solution casting of a mixture of ethyl cellulose and the CPs (200–500 ppm) from toluene at room temperature. To ensure good optical quality of the films, the solvent was evaporated slowly at room temperature (72 h) and the last traces removed in an oven at 60 ◦ C for 10 min. No differences in fluorescence behavior were observed when samples were left at this temperature for longer times. Plasticized films were obtained by the addition of 1–10 wt % of polyethylene glycol (600 and 3400) and polypropylene glycol (average Mw = 1000) to EC, followed by its dissolution in the solvent. The thicknesses of the films were 400–600 µm as measured with a micrometer (Etalon Rolle, Switzerland). 2.3. Ink Jet Printing A FUJIFILM Dimatix Materials printer DMP-2800 Series was used for printing films. This is suitable for printing the CPs on an appropriate matrix. Toluene solutions of the CPs were used to fill disposable cartridges that have 16 individually-tunable, piezo-actuated nozzles. Cartridges are available for dispensing 10 pL or 1 pL drops. Drops were printed by voltage-driven deformations of a membrane wall of a chamber behind each nozzle. The segments of this action make up a waveform that is optimized for each ink, as well as the intended print job. 2.4. Luminescence Characterization The UV spectroscopic measurements were performed at room temperature with a Shimadzu UV-3101PC UV-VIS-NIR spectrometer, using cells of 1.0 cm optical path length for solution measurements. The emission and excitation spectra were recorded with a Horiba Jobin-Ivon SPEX Fluorolog 3–22 fluorescence spectrometer. The Fluorolog consists of a modular spectrofluorimeter with double-grating monochromators for excitation (200–950 nm range, optimized in the UV with a blaze angle at 330 nm) and emission (200–950 nm range, optimized in the visible and with a blaze angle at
Sensors 2017, 17, 2532
4 of 12
with double-grating Sensors 2017, 17, 2532
monochromators for excitation (200–950 nm range, optimized in the UV with a 4 of 13 blaze angle at 330 nm) and emission (200–950 nm range, optimized in the visible and with a blaze angle at 500 nm). The bandpass for excitation and emission was 5 nm with a wavelength accuracy of 500 excitation and emission was 5 nm with a wavelength accuracy of ±detector 0.5 nm. ± 0.5nm). nm.The Thebandpass excitationfor source consisted of an ozone-free 450 W xenon lamp. The emission The excitation source consisted of an ozone-free 450 W xenon lamp. The emission detector employed employed was a Hamamatsu R928 photomultiplier, with a photodiode as the reference detector. The was a Hamamatsu R928 photomultiplier, with a photodiode as cuvette the reference detector. fluorescence fluorescence quenching of the film was measured in a sealed containing theThe nitrobenzene or quenching of the film was measured in a sealed cuvette containing the nitrobenzene or 1,3-dinitrobenzene 1,3-dinitrobenzene vapors at room temperature (293 K). vapors at room temperature (293 K). Time-resolved picosecond fluorescence intensity decays were obtained by the single-photon Time-resolved fluorescence decays were obtained the single-photon timing method withpicosecond laser excitation, with theintensity set-up described elsewhere [12].byDecay data analysis timing method with laser excitation, with the set-up described elsewhere [12]. Decay data analysis with a sum of exponentials was achieved by means of a Microsoft Excel spreadsheet specially with a sumfor of lifetime exponentials was that achieved by means of a Microsoft Excel designed analysis considers deconvolution with the spreadsheet instrument specially response designed function for lifetime analysis that considers deconvolution with the instrument response function (IRF) [13]. (IRF) [13]. 2.5. Structural Characterization 2.5. Structural Characterization A (Leica TCS-SP5) TCS-SP5) equipped equipped with with aa CW CW Ar Ar ion ion laser laser (458, (458, A confocal confocal laser laser scanning scanning microscope microscope (Leica 465, 488, 496, and 514 nm) and a pulsed Ti:sapphire (Spectra-Physics Mai Tai BB, 710–990 nm, 100 fs, 465, 488, 496, and 514 nm) and a pulsed Ti:sapphire (Spectra-Physics Mai Tai BB, 710–990 nm, 100 fs, 80 MHz) was used to obtain images of the films. 80 MHz) was used to obtain images of the films. SEM a Hitachi S2400 microscope and the images were recorded by software SEM was wasperformed performedwith with a Hitachi S2400 microscope and the images were recorded by Quantax (Bruker; Billerica, MA, USA). Samples were coated with gold and registered at 50×atand software Quantax (Bruker; Billerica, MA, USA). Samples were coated with gold and registered 50× 500 magnification. and×500× magnification. 2.6. Sensor Prototype 2.6. Sensor Prototype A portable device (Figure 2) was developed and tested for the study of the fluorescence quenching A portable device (Figure 2) was developed and tested for the study of the fluorescence of the CP films by the nitroaromatic vapors. A Sony SLD3135 laser diode (405 nm; 50 mW) with quenching of the CP films by the nitroaromatic vapors. A Sony SLD3135 laser diode (405 nm; 50 integrated monitoring photodiode was used as excitation source and a Vishay BPW34 Silicon PIN mW) with integrated monitoring photodiode was used as excitation source and a Vishay BPW34 photodiode covered with a green filter (Edmund Optics #43-934) to avoid excitation light from Silicon PIN photodiode covered with a green filter (Edmund Optics #43-934) to avoid excitation light reaching the measuring photodiode was used for the detection. The general architecture is described from reaching the measuring photodiode was used for the detection. The general architecture is elsewhere [14]. described elsewhere [14].
Figure Figure 2. 2. Prototype Prototype sensor sensor device. device.
3. Results and Discussion The ground-state absorption spectrum of F8T2 was obtained in toluene solution and in ethyl cellulose films (Figure 3). In the case of the thin film a new new band band is is observed observed at at longer longer wavelengths wavelengths (492 nm). nm). A number of possible explanations exist for this this new new band. band. However, detailed detailed analysis analysis suggests that it is associated with the so-called β phase of the F8T2 [15]. The β phase is a metastable the so-called β F8T2 [15]. β state with part part of of the theCP CPin inaarigid rigidextended extendedstructure, structure,and and can formed through evaporation of can bebe formed through evaporation of an an appropriate solvent, treatment film oneofofthe theother otherphases phasesby by solvent solvent vapor, vapor, or by appropriate solvent, by by treatment of of thethe film ororone keeping the polymer in a restricted environment. The β-phase peak was observed originally by Bradley and co-workers [16] in PFO polystyrene films and PFO solutions in poor solvents (such as
Sensors 2017, 17, 2532 Sensors 2017, 17, 2532 Sensors 2017, 17, 2532
5 of 12 5 of 12 5 of 13
keeping the polymer in a restricted environment. The β-phase peak was observed originally by keeping the polymer in a restricted environment. The β-phase peak was observed originally by Bradley and co-workers [16] in PFO polystyrene films and PFO solutions in poor solvents (such as Bradley and co-workers in PFO polystyrene films PFO in poor solventsof(such as methylcyclohexane) [17].[16] We observed this new band at and 492 nm insolutions the absorption spectrum the PFO methylcyclohexane) [17]. We observed this new band at 492 nm in the absorption spectrum of the methylcyclohexane) We4), observed this new band 492 nm polymer in the absorption of the ethyl cellulose films [17]. (Figure even though F8T2 is a at less rigid than PFO,spectrum strong support PFO ethyl cellulose films (Figure 4), even though F8T2 is a less rigid polymer than PFO, strong PFO ethyl cellulose films (Figure 4), even though F8T2 is a less rigid polymer than PFO, strong has been presented from steady-state and time-resolved fluorescence and fluorescence anisotropy support has been presented from steady-state and time-resolved fluorescence and fluorescence support has been presented from steady-state andthis time-resolved fluorescence and fluorescence measurements for the formation of the β-phase with polymer [18]. anisotropy measurements for the formation of the -phase with this polymer [18]. anisotropy measurements for the formation of the -phase with this polymer [18].
Figure 3. 3. Absorption Absorptionspectrum spectrumof of the the F8T2 F8T2 ethyl ethyl cellulose cellulose film film (solid (solid line) line) and and in in toluene toluene solution solution Figure Figure 3. Absorption spectrum of the F8T2 ethyl cellulose film (solid line) and in toluene solution (dashed line). (dashed line). (dashed line).
The presence of NB vapor vapor doesnot not changesignificantly significantly the absorption spectrum (Figure 4) of The presence presence of absorption spectrum (Figure 4) of The of NB NB vapordoes does notchange change significantlythe the absorption spectrum (Figure 4)the of the PFT2 or PFO ethyl cellulose films, suggesting the absence of ground-state complexation. PFT2 or PFO ethyl cellulose films, suggesting the absence of ground-state complexation. the PFT2 or PFO ethyl cellulose films, suggesting the absence of ground-state complexation.
Figure 4. (A) Absorption spectraofofthe the F8T2film film in the presence (dashed line) and absence (solid Figure 4. 4. (A) (A) Absorption inin thethe presence (dashed line) andand absence (solid line) Figure Absorption spectra spectra of theF8T2 F8T2 film presence (dashed line) absence (solid line) of nitrobenzene vapor. (B) Absorption spectra of the PFO film in the presence (dashed line) and of nitrobenzene vapor; (B) Absorption spectra of the of PFO infilm the presence (dashed(dashed line) and absence line) of nitrobenzene vapor. (B) Absorption spectra thefilm PFO in the presence line) and absence (solid line) of nitrobenzene vapor. Both CPs are incorporated in an ethyl cellulose matrix. (solid line) of nitrobenzene vapor. Both CPs are incorporated in an ethyl cellulose matrix. absence (solid line) of nitrobenzene vapor. Both CPs are incorporated in an ethyl cellulose matrix.
The F8T2 emission spectrum has a maximum at 545 nm (Figure 5). In the presence of NB vapor The The F8T2 F8T2 emission emission spectrum spectrum has has aa maximum maximum at at 545 545 nm nm (Figure (Figure 5). 5). In Inthe thepresence presence of of NB NB vapor vapor we observed a decrease of the emission intensity of about 42%. In the case of the PFO, the emission we observed a decrease of the emission intensity of about 42%. In the case of the PFO, the emission we observed a decrease of the emission intensity of about 42%. In the case of the PFO, the emission showed a structured fluorescence spectrum between 400 and 600 nm, attributed to at least three showed showed aa structured structured fluorescence fluorescence spectrum spectrum between between 400 400 and and 600 600 nm, nm, attributed attributed to to at at least least three three vibronic components. We observed only a 34% drop in the fluorescence emission intensity in the vibronic vibronic components. components. We Weobserved observedonly onlyaa 34% 34% drop drop in in the the fluorescence fluorescence emission emission intensity intensity in in the the presence of NB vapor at 445 nm. As mentioned, the CPs are good electron donors and their presence ofNB NBvapor vapor at 445 As mentioned, gooddonors electron and their presence of at 445 nm. nm. As mentioned, the CPsthe areCPs goodare electron anddonors their fluorescence fluorescence is quenched by NB through photoinduced electron transfer. The amplifying nature of fluorescence by NB through photoinduced electron transfer. The amplifying nature of is quenched is byquenched NB through photoinduced electron transfer. The amplifying nature of the exciton
Sensors 2017, 17, 2532
6 of 13
Sensors 2017, 17, 2532
6 of 12
delocalization in the conjugated polymers makes them highly sensitive materials to quenching by the exciton delocalization in the conjugated polymers makes them highly sensitive materials to nitroaromatic vapors. quenching by nitroaromatic vapors.
Figure 5. 5. Fluorescence Fluorescence spectra line) and presence of saturated NB vapors at room Figure spectra in inthe theabsence absence(solid (solid line) and presence of saturated NB vapors at temperature (dashed line) of: ethyl cellulose film, film, λexc =λ410 in ethyl exc = nm, 410 and nm, (B) andF8T2 (B) F8T2 in room temperature (dashed line)(A) of:PFO (A) in PFO in ethyl cellulose cellulose film, film, λexc =λexc 490= nm. ethyl cellulose 490 nm.
Time-resolved fluorescencedecays decaysof of in thin ofcellulose ethyl cellulose at the Time-resolved fluorescence CPsCPs in thin films films of ethyl recordedrecorded at the maximum maximum emission wavelength arewith well sums fittedof with sums of three exponentials Thisprevious agrees emission wavelength are well fitted three exponentials (Table 1). This(Table agrees1).with with previous studies [18] showing triple exponential decays of F8T2 in methylcyclohexane (MCH), studies [18] showing triple exponential decays of F8T2 in methylcyclohexane (MCH), with lifetimes with lifetimes ps,20440 20 ps. The time longest (650 ps) is assigned to the of 650 ps, 440of ps,650 and ps. ps, Theand longest decay (650decay ps) istime assigned to the β-conformation β-conformation and thelifetime intermediate lifetime ps) to the α-conformation [18]. The and the intermediate (440 ps) to the(440 α-conformation [18]. The shortest timeshortest (20 ps)time may (20 ps) may result from solvent/conformational relaxation or intramolecular energy transfer from result from solvent/conformational relaxation or intramolecular energy transfer from non-ordered to non-ordered to segments. ordered chain segments. of PFO in toluene solution also show a complex ordered chain Studies of PFOStudies in toluene solution [19] also show[19] a complex decay, a sum decay, twoexponentials or even three aretobeing good fits.ofAabout fast of two aorsum evenofthree areexponentials being required obtainrequired good fits.toAobtain fast component component of about 20 ps is found, and is more important (with greater amplitude) at the onset of 20 ps is found, and is more important (with greater amplitude) at the onset of the emission band. the band. An intermediate component is also around 90decay ps and a predominant An emission intermediate component is also observed around 90 psobserved and a predominant time around 360 ps decay time around 360 ps is observed independent of the emission wavelength and attributed to the is observed independent of the emission wavelength and attributed to the PFO intrinsic fluorescence PFO intrinsic fluorescence lifetime. In thin films, the decay is again described by a of sum of lifetime. In thin films, the decay is again described by a of sum of three exponentials, however, atthree long exponentials, however, at long wavelengths, it is dominated by a long component of 3 ns, attributed wavelengths, it is dominated by a long component of 3 ns, attributed to the presence of photooxidized to the presence photooxidized species, such as keto defects other populated emissive defects, which are species, such asof keto defects and other emissive defects, which and are easily by efficient energy easily populated by efficient energy migration [20]. migration [20]. Table Table 1. 1. Decay Decaytimes timesand andamplitudes amplitudesof ofthe thePFO PFOand andF8T2 F8T2in inethyl ethylcellulose cellulosethin thinfilm filmwithout withoutNB NB vapors and in the presence of NB vapors. vapors and in the presence of NB vapors. /ns change (%) ** NB /ns ( *) /ns ( *) /ns ( *) τfluor change (%) ** NB τ1 /ns (f1 *) τ2 /ns (f2 *) τ3 /ns (f3 *) τaverage /ns Without 0.03 (0.10) 0.43 (0.55) 0.62 (0.35) 0.40 F8T2 65 0.03 (0.10) 0.04 (0.51) 0.43 (0.55) 0.29 0.62 0.40 WithWithout 0.005 (0.07) (0.42)(0.35) 0.14 F8T2 65 With 0.005 (0.07) 0.04 (0.51) 0.29 (0.42) 0.14 Without 0.16 (0.26) 0.31 (0.63) 1.70 (0.11) 0.43 PFO 72 Without0.05 (0.37) 0.16 (0.26) 0.08 (0.52) 0.31 (0.63) 0.55 1.70 0.43 (0.11)(0.11) 0.12 PFOWith 72 With 0.05 (0.37) 0.08 (0.52) 0.55 (0.11) 0.12 * computed from the individual lifetimes and pre-exponential factors: , and , = * computed from the individual lifetimes and pre-exponential factors: α1 , α2 and α3 , f 1 = α1 τ1 /(α1 τ1 + /( + + ), = /( + + ) and = 1 − ( + ). ** computed α2 τ2 +
α3 τ3 ),
f2
=
α2 τ2 /(α1 τ1 +
α2 τ2 +
τ ( without NB vapors −τaverage with NB vapors as: ( average ) × 100.
α3 τ3 ) and f 3
) × 100.
=
1 − ( f 1 + f 2 ).
** computed as:
τaverage without NB
The Thedrop drop in in the the average average lifetimes lifetimes (τ⁄τ (τ⁄τ0)0 )ofofF8T2 F8T2and andPFO, PFO,resulting resultingfrom fromthe thepresence presenceof ofNB NB vapors, is 35% and 28%, respectively. Comparing these values with those measured in steady-state vapors, is 35% and 28%, respectively. Comparing these values with those measured in steady-state conditions, conditions,42% 42%and and34% 34%for forF8T2 F8T2and andPFO, PFO,respectively, respectively,ititisisconcluded concludedthat thatthe thequenching quenchinginduced induced by the nitroaromatics is predominantly a dynamic process. This is important as it favors reversibility of the sensing system.
Sensors 2017, 17, 2532
7 of 13
by the nitroaromatics is predominantly a dynamic process. This is important as it favors reversibility Sensors 2017, 17, 2532 7 of 12 of the sensing system. To Toincrease increasethe thesensitivity sensitivitytowards towardsdetection detectionof ofnitroaromatic nitroaromaticvapors vaporsof ofthe theP8T2 P8T2ethyl ethylcellulose cellulose films, tested the effect of adding different plasticizers to the ethyl Plasticized films,we wehave have tested the effect of adding different plasticizers to cellulose the ethylmatrix. cellulose matrix. films show a very high sensitivity towards nitroaromatics when compared with the non-plasticized Plasticized films show a very high sensitivity towards nitroaromatics when compared with the ones, probably due higher permeability (Figure 6). For (Figure example, only 1% using (w/w)only of the non-plasticized ones,toprobably due to higher permeability 6).using For example, 1% plasticizer PEG 3400, we obtained in a short time (three minute) of 2,3-dinotrobenzene, DNB, vapor (w/w) of the plasticizer PEG 3400, we obtained in a short time (three minute) of 2,3-dinotrobenzene, exposure a fluorescence of quenching 95% (Figureof6B). DNB, vapor exposure aquenching fluorescence 95% (Figure 6B).
Figure Figure6.6. (A) (A) Emission Emission spectra spectra of ofthe theF8T2 F8T2ethyl ethylcellulose cellulosefilm filmwith with1% 1%(w/w) (w/w) of of PEG PEG 3400 3400 in inthe the −6 M) vapor absence (solid line) and after three minutes of exposure to nitrobenzene (8.74 × 10 −6 absence (solid line) and after three minutes of exposure to nitrobenzene (8.74 × 10 M) vapor (dashed (dashed lines); (B) Emission of the F8T2 ethyl cellulose film 1%of(w/w) of PEG 3400absence in the lines). (B) Emission spectra spectra of the F8T2 ethyl cellulose film with 1%with (w/w) PEG 3400 in the −5 M) vapor absence (solid line) and after three minutes of exposure to 2,3-dinitrobenzene (1.83 × 10 (solid line) and after three minutes of exposure to 2,3-dinitrobenzene (1.83 × 10−5 M) vapor (dashed (dashed lines). Excitation wavelength lines). Excitation wavelength was 450was nm.450 nm.
We from Figure Figure66a astronger strongerfluorescence fluorescence quenching caused by the vapors. We can can see from quenching caused by the DNBDNB vapors. This This higher quenching efficiency may arise from thehigher higherelectron electronaffinity affinity of of DNB, outweighing higher quenching efficiency may arise from the outweighing its its lower vapor pressure [21]. In all the studied plasticized films, the sensitivity towards nitroaromatic lower vapor pressure [21]. In all the studied plasticized films, the sensitivity towards nitroaromatic vapors vaporsincreases increaseswhen whencompared comparedwith withthe theneat neatethyl ethylcellulose cellulosefilms, films,as ascan canbe beseen seenfrom fromthe thedecrease decrease in influorescence fluorescencelifetimes lifetimesininTable Table2.2. Table 2. Decay times of F8T2 in ethyl cellulose thin films with different plasticizers in the presence and Table 2. Decay times of F8T2 in ethyl cellulose thin films with different plasticizers in the presence in the absence of nitroaromatic vapors. and in the absence of nitroaromatic vapors. NB NB τ τ τfluor. average average average without NB withoutaverage with NBwith . Decrease (%) * NB vapors /ns NB vapors /ns Decrease (%)* vapors/ns vapors/ns Neat Ethyl Cellulose 400 340 15 Neat Ethyl Cellulose 1513 400 380 340 330 1% PEG 3400 PEG 600 1% PEG 1% 3400 1321 380 380 330 300
1% PEG 600
τ
or DNB vapors 21 300 without NB *380 computed as: ( average τ
−τaverage
DNB DNB τ τ τfluor. average without average with average without NB average with . Decrease Decrease (%) * NB vapors /ns DNB vapors (%)* vapors/ns DNB vapors 400 300 25 400 300 375 200 47 25 380 180 53 47 375 200 with NB or DNB vapors
average without NB or DNB vapors
* computed as: (
380
) × 100. 180
53
) × 100.
By comparing the values in the lifetime attenuation for the plasticized and non-plasticized films (TableBy 2),comparing we observethe that the addition of onlyattenuation 1% by weight of plasticizer thenon-plasticized neat ethyl cellulose values in the lifetime for the plasticizedtoand films matrix the sensitivity of F8T2 towards these vapors by ca. 5% and (Table increases 2), we observe that the addition of only 1% bynitroaromatic weight of plasticizer to the neat(NB), ethyl29%, cellulose 35% (DNB). The molecular weight of the plasticizer not appearvapors to have significant matrix increases the sensitivity of F8T2 towards thesedoes nitroaromatic byaca. 5% (NB),influence 29%, and on the(DNB). increase inmolecular the sensitivity of the CP,plasticizer but it is noted a lowerto molecular weight one seems 35% The weight of the does that not appear have a significant influence to a slightly stronger quenching NB that andaDNB. may be a result better onfacilitate the increase in the sensitivity of the CP, with but it both is noted lowerThis molecular weight oneof seems to compatibility with the EC matrix, and, hence, production of a more amorphous structure. facilitate a slightly stronger quenching with both NB and DNB. This may be a result of better The thermal behavior of theseand, plasticizers films was of studied. ethyl cellulose has a glass compatibility with the EC matrix, hence, production a more Neat amorphous structure. ◦ transition (Tg) atof130–133 C [22]. As expected, addition of plasticizer decreases Tg The temperature thermal behavior these plasticizers films wasthe studied. Neat ethyl cellulose has athe glass
transition temperature (Tg) at 130–133 C [22]. As expected, the addition of plasticizer decreases the Tg of the samples. For example, using 25% (w/w) PEG 400 the Tg drops to 70 C [21]. This strong decrease means that there is a good compatibility between the polymer matrix and the plasticizer. In
Sensors 2017, 17, 2532
8 of 13
Sensors 2017, 17, 2532 8 of 12 of the samples. For example, using 25% (w/w) PEG 400 the Tg drops to 70 ◦ C [21]. This strong decrease means that there is a good compatibility between the polymer matrix and the plasticizer. In general, general, plasticizers reduce polymer interchain interactions by distributing themselves plasticizers reduce polymer interchain interactions by distributing themselves homogeneously within homogeneously within the polymer, hence increasing the free volume. However, a reduction the the polymer, hence increasing the free volume. However, a reduction in the Tg value down toinnear Tg value down to near room temperature will result in an increase in chain mobility and, room temperature will result in an increase in chain mobility and, consequently, could enhance the consequently, of could the crystallization of films by reducing the phenomenon energy required for lead this crystallization filmsenhance by reducing the energy required for this process. This would process. This phenomenon would lead to structural changes resulting in loss of transparency. The to structural changes resulting in loss of transparency. The thermal stability of the films was studied thermal stability of fluorescence the films wasintensity studied with by monitoring theoffluorescence intensity withexperiments an increase by monitoring the an increase the temperature. These of the temperature. These experiments (Figure were performed in the absence and in the presence (Figure 7) were performed in the absence and in7)the presence of DNB vapors. of DNB vapors.
Figure Figure 7. 7. Fluorescence Fluorescence attenuation attenuation of of the the F8T2 F8T2 ethyl ethyl cellulose cellulose film film with with1% 1%(w/w) (w/w) of of PEG PEG 3400 3400 in in the the absence (solid line) and presence (dashed line) of DNB vapor with as a function of the temperature. absence (solid line) and presence (dashed line) of DNB vapor with as a function of the temperature. Excitation Excitation wavelength wavelength was was 450 450 nm. nm.
It can be seen seen that that the the increase increase of of DNB DNB vapor vapor pressure pressure with with the the temperature temperature is not the major factor involved. Instead, the temperature dependence appears to result from from the decrease decrease in Tg upon PEG addition. The morphology morphologyofofthe the films studied by confocal fluorescence microscopy. The CPCP films waswas studied by confocal fluorescence microscopy. Typical Typical images images obtained this technique areinshown 8, in green spots represent the obtained from thisfrom technique are shown Figurein8,Figure in which thewhich greenthe spots represent the emission emission ofThese the CPs. These films do not exhibit bulk phase separation at the magnifications studied, of the CPs. films do not exhibit bulk phase separation at the magnifications studied, but some but someaggregation polymer aggregation can beespecially observed, in the ethyl cellulose film containing polymer can be observed, in especially the ethyl cellulose film containing F8T2. F8T2.The surface morphology of the films was studied using scanning electron microscopy (SEM) (Figure 9). The surface of neat ethyl cellulose film (not shown) is rather smooth, compact, and featureless. However, in the case of ethyl cellulose, F8T2 blends, phase separated zones are observed for concentrations above the incorporation capacity of CPEs into ethyl cellulose films, Figure 9A. The addition of a plasticizer to the neat ethyl cellulose can be seen in Figure 9B to introduce some porosity. This film has pores with diameters between 1.5 and 3 µm, randomly distributed. The formation of these pores in the plasticized films may explain, in part, the increased sensitivity to the nitroaromatic vapors observed by fluorescence quenching of the CPs.
factor involved. Instead, the temperature dependence appears to result from the decrease in Tg upon PEG addition. The morphology of the CP films was studied by confocal fluorescence microscopy. Typical images obtained from this technique are shown in Figure 8, in which the green spots represent the emission of the CPs. These films do not exhibit bulk phase separation at the magnifications studied, Sensors 2017, 17, 2532 9 of 13 but some polymer aggregation can be observed, especially in the ethyl cellulose film containing F8T2.
Sensors 2017, 17, 2532
9 of 12
Figure 8. Typical images of (A) PFO and (B) F8T2 in neat ethyl cellulose film, observed by confocal microscopy. Excitation wavelength was 458 nm and emission wavelength in the range of 510–700 nm. These pictures have 512 × 512 pixels, using a pinhole of 1 AU, zoom 4×, and 400 Hz. The scale bars are 25 µm.
The surface morphology of the films was studied using scanning electron microscopy (SEM) (Figure 9). The surface of neat ethyl cellulose film (not shown) is rather smooth, compact, and featureless. However, in the case of ethyl cellulose, F8T2 blends, phase separated zones are observed for concentrations above the incorporation capacity of CPEs into ethyl cellulose films, Figure 9A. The Figure 8. of Typical images of (B) cellulose F8T2 in neat cellulose film, observed by confocal addition a plasticizer to (A) thePFO neatand ethyl can ethyl be seen in Figure 9B to introduce some microscopy. Excitation wavelength was 458 nm and emission wavelength in the range of 510–700 porosity. This film has pores with diameters between 1.5 and 3 µm, randomly distributed.nm. The These pictures havepores 512 × usingfilms a pinhole of 1 AU, in zoom , and 400 Hz.sensitivity The scale to bars formation of these in512 thepixels, plasticized may explain, part,4× the increased the are 25 µm. nitroaromatic vapors observed by fluorescence quenching of the CPs.
Figure 9. SEM images of ethyl cellulose films containing (A) F8T2 and (B) F8T2 and 1% PEG 600. Figure 9. SEM images of ethyl cellulose films containing (A) F8T2 and (B) F8T2 and 1% PEG 600. The The scale barsare are50 50µm µmand and9 µm. 9 µm. scale bars
Fluorescence quenching based sensors normally require a reference material, the Fluorescence quenching based sensors normally require a reference material, suchsuch thatthat the degree degree of measured quenching by measured byofthe ratio of signals from sensor and reference materials. Weproduced have of quenching the ratio signals from sensor and reference materials. We have such asystem ratiometric system using ink-jet microprinting (IJMP) (Figure 10). IJMP The IJMP allows suchproduced a ratiometric using ink-jet microprinting (IJMP) (Figure 10). The allows direct direct deposition of minuscule quantities of the CPs onto the ethyl cellulose film substrate (thickness deposition of minuscule quantities of the CPs onto the ethyl cellulose film substrate (thickness 240 µm). 240 µm). The diameter and uniformity of the microdot can be controlled by modifying substrate The diameter and uniformity of the microdot can be controlled by modifying substrate surface surface chemistry and ink preparation [21]. chemistryWe and ink preparation [21]. have used F8T2 as sensor material, and have incorporated another CP, F8BT (Figure 1), We have used F8T2oxidizable as sensorand material, haveany incorporated F8BT (Figure which is not readily does notand exhibit fluorescence another changes CP, in the presence of 1), which is not readily oxidizable andserving does notasexhibit any fluorescence changes in thethe presence DNB DNB and NB vapors, thereby an internal reference. Table 3 shows analysisof of and fluorescence NB vapors, decays therebyofserving as an internal reference. Table 3 shows the analysis ofabsence fluorescence the imprinted ethyl cellulose sensor by IJMP. In the presence and of decays of the imprinted ethylwe cellulose sensor by IJMP. In the presence and absence of concentrated concentrated DNB vapors, observed that F8BT did not exhibit any significant attenuation of its fluorescence lifetime, in contrast to the quenching observed with F8T2. In this case, we can use this system as a sensor with an internal reference for nitroaromatic vapors.
Sensors 2017, 17, 2532
10 of 13
DNB vapors, we observed that F8BT did not exhibit any significant attenuation of its fluorescence lifetime, in contrast to the quenching observed with F8T2. In this case, we can use this system as 2017, 17, 2532 10 of 12 aSensors sensor with an internal reference for nitroaromatic vapors.
Figure 10. Photograph of the P8T2 IJMP imprinted film on ethylcellulose under UV light. Table 3. 3. Decay Decay times times of of the the F8T2 F8T2 and and F8BT F8BT imprinted imprinted by by IJMP IJMP in in ethyl ethyl cellulose cellulose film in presence presence or or film in Table absence of nitroaromatics vapors. absence of nitroaromatics vapors.
F8T2
F8BT
F8T2 average without DNB
average with DNB
τ average without vapors/ns DNB vapors /ns
τ average with vapors/ns DNB vapors /ns
1. 8
1. 8
0.86
.
øfluor. Decrease (%)* Decrease (%) * 52
0.86 * computed as: (
* computed as:
52
F8BT average without NB
τ average without vapors/ns NB vapors /ns
0.79
0.79
τ NB vapors −τaverage with NB vapors ) ( average without τaverage without NB vapors
.
average with
τ average with DNB vapors
DNB vapors
0.78
0.78 ) × 100.
τfluor. Decrease (%)* Decrease (%) * 1 1
× 100.
If we compare the values in the lifetime attenuation for the F8T2 and F8BT (Table 3), we observe If we compare thedecreases values inby the52%, lifetime attenuation forinsignificant the F8T2 and F8BT (Table we lifetime. observe that the F8T2 lifetime whereas there are changes in the 3), F8BT that the F8T2 lifetime decreases by 52%, whereas there are insignificant changes in the F8BT lifetime. The measured sensitivity is equivalent to that obtained in plasticized ethyl cellulose films. The measured equivalent tosetup that obtained in of plasticized films. The sensorsensitivity was testedisin a dynamic composed two massethyl flow cellulose controllers (MFCs Dwyer The sensor was tested in a dynamic setup composed of two mass flow controllers Dwyer GFC-2102), one controlling the flow of clean air and the other controlling the flow(MFCs of saturated GFC-2102), one controlling the flow of clean air and the other controlling the flow of saturated nitrobenzene vapor, obtained from a bubbler at constant temperature and constant pressure. Both nitrobenzene vapor, obtained from a bubbler at constant temperature(PIC24FV16KM202) and constant pressure. Both MFCs MFCs are controlled from MATLAB through a microcontroller that sets the are controlled from MATLAB through a microcontroller (PIC24FV16KM202) that sets the references to references to the MFCs, defining the composition of the nitrobenzene-clean air mixture. As expected, the MFCs, defining the composition of the nitrobenzene-clean air mixture. As expected, the proposed the proposed differential approach guaranteed very good sensitivity and fast response from the differential approach guaranteed very good sensitivity fast11response from the response electronics sidea electronics side while providing high-level input signals.and Figure shows the sensor after while providing high-level input signals. Figure 11 shows the sensor response after a 20-second 20-second exposure to minute NB vapors. exposure to minuteapplications, NB vapors. the fast response to the presence of a small concentration of For in-field For in-field fast response to the the presence oftime a small concentration of nitroaromatics is aapplications, very positivethe characteristic. Although recovery is rather long (several nitroaromatics is a very positive characteristic. Although the recovery time is rather long (several minutes), this can be acceptable for scenarios where the detection of frequent changes in the analyte minutes), can be acceptable for scenarios where the detection of frequent changes in the analyte level is notthis required. level is not required.
Figure 11. Attenuation of the fluorescence emission of the F8T2 in ethyl cellulose when exposed to NB (1.7 × 10−5 M) vapors for 20 s.
the proposed differential approach guaranteed very good sensitivity and fast response from the electronics side while providing high-level input signals. Figure 11 shows the sensor response after a 20-second exposure to minute NB vapors. For in-field applications, the fast response to the presence of a small concentration of nitroaromatics is a very positive characteristic. Although the recovery time is rather long (several Sensors 2017, 17, 2532 11 of 13 minutes), this can be acceptable for scenarios where the detection of frequent changes in the analyte level is not required.
Figure 11. Attenuation of the fluorescence emission of the F8T2 in ethyl cellulose when exposed to NB Figure 11. Attenuation of the fluorescence emission of the F8T2 in ethyl cellulose when exposed to (1.7 × 10−5 M) vapors for 20 s. NB (1.7 × 10−5 M) vapors for 20 s.
4. Conclusions We have studied fluorene-based conjugated polymers as fluorescence sensing materials for nitroaromatic vapors with the overall goal of detecting explosives using such polymers. The best conjugated polymer in our studies was found to be poly(9,9-dioctylfluorenyl-2,7-diyl]-co-bithiophene] (F8T2). It is stable, has a good absorption between 400 and 450 nm, a strong and structured fluorescence around 550 nm. A 96% quenching of fluorescence, accompanied by a corresponding decrease in the fluorescence lifetime, is seen on exposure of the plasticizer film of F8T2 ethylcellulose to the model compounds nitrobenzene (NB) and 1,3-dinitrobenzene (DNB) vapors, both from the family of common explosives vapors. Furthermore, it was demonstrated that ink-jet microprinting can be used as a convenient approach to easily and rapidly fabricate films containing these sensors and inert reference materials, with the same sensitivity of plasticized ethyl cellulose films towards nitroaromatic vapors. A sensor prototype based on the F8T2 conjugated polymer was developed and tested. The ability of the sensor to detect small quantities of nitrobenzene was confirmed. The sensor prototype showed very fast response (a few seconds) to the presence of small concentrations of the target analyte, but also showed a large recovery time, which limits its potential applications. This slow recovery may result from the designed sampling chamber and also from the time required for desorption of the nitrobenzene molecules from the polymer surface. Both of the above aspects will be addressed in a future designs, optimizing polymer thickness and porosity, and optimizing the shape and arrangement of the sampling chamber. Acknowledgments: This work was partially carried out in the framework of TIRAMISU project. This project is funded by the European Community’s Seventh Framework Program (FP7/2007-2013) under grant 284747. L Martelo was supported by Fundação para a Ciência e Tecnologia (FCT, Portugal) with a postdoctoral fellowship (SFRH/BPD/121728/2016). HDB is grateful for funding from the Coimbra Chemistry Centre (CQC), which is supported by FCT through the programs UID/QUI/UI0313/2013 and COMPETE. AC also thanks FCT for funding through the program UID/EEA/50008/2013. AF also thanks to the FCT for funding through (SFRH/BPD/111301/2015). Author Contributions: L. Martelo conceived and performed the experiments, under the supervision of H. Burrows and M. N. Berberan-Santos; A. Fedorov performed the time-resolved fluorescence experiments; A. Charas helps in the conception of the ink jet printing films; J. Figueiredo designed and implemented the sensor prototype shown in Figure 2; T. Neves characterized the sensor’s response and Lino Marques supervised the works relative to the prototype design, implementation and characterization. Conflicts of Interest: The authors declare no conflict of interest.
Sensors 2017, 17, 2532
12 of 13
References 1. 2.
3. 4. 5. 6.
7.
8. 9.
10.
11.
12.
13. 14. 15.
16. 17.
18. 19.
McQuade, T.D.; Pullen, E.A.; Swager, T.M. Conjugated polymer-based chemical sensors. Chem. Rev. 2000, 100, 2537–2574. [CrossRef] [PubMed] Gu, C.; Huang, N.; Gao, J.; Xu, Y.; Jiang, D. Controlled synthesis of conjugated microporous polymer films: versatile platforms for highly sensitive and label-free chemo- and biosensing. Angewandte 2014, 53, 4850–4855. [CrossRef] [PubMed] Thomas, W.S., III; Joly, D.G.; Swager, M.T. Chemical sensors based on amplifying fluorescent conjugated polymers. Chem. Rev. 2007, 107, 1339–1386. [CrossRef] [PubMed] Gu, C.; Huang, N.; Wu, Y.; Xu, Y.; Jiang, D. Design of highly photofunctional porous polymer films with controlled thickness and prominent microporosity. Angewandte 2015, 54, 11540–11544. [CrossRef] [PubMed] Yang, J.; Swager, T.M. Fluorescent porous polymer films as TNT chemosensors: Electronic and structural effects. J. Am. Chem. Soc. 1998, 120, 11864–11873. [CrossRef] Hancock, F.L.; Deans, R.; Moon, J.; Swager, M.T. Amplifying fluorescent polymer detection of bioanalytes. In Proceedings of the Chemical and Biological Early Warning Monitoring for Water, Food, and Ground, Boston, MA, USA, 2001. Available online: https://www.spiedigitallibrary.org/conference-proceedings-ofspie/4575/1/Amplifying-fluorescent-polymer-detection-of-bioanalytes/10.1117/12.456910.short?SSO=1 (accessed on 29 October 2017). Knaapila, M.; Garamus, V.M.; Dias, F.B.; Almásy, L.; Galbrecht, F.; Charas, A.; Morgado, J.; Burrows, H.D.; Monkman, A.P. Influence of solvent quality on the self-organization of archetypical hairy rods−branched and linear side chain polyfluorenes: Rodlike chains versus “Beta-Sheets” in solution. Macromolecules 2006, 39, 6505–6512. [CrossRef] Wegner, G. Nanocomposites of hairy-rod macromolecules: Concepts, constructs, and materials. Macromol. Chem. Phys. 2003, 204, 347–357. [CrossRef] Fonseca, S.M.; Pina, A.L.G.; Seixas de Melo, J.; Burrows, H.D.; Chattopadhyay, N.; Alcácer, L.; Charas, A.; Morgado, J.; Monkman, A.P.; Asawapirom, U.; et al. Triplet-state and singlet oxygen formation in fluorene-based alternating copolymers. J. Phys. Chem. B 2006, 110, 8278–8283. [CrossRef] [PubMed] Andersson, H.; Hjärtstam, J.; Stading, M.; Corswant, V.C.; Larsson, A. Effects of molecular weight on permeability and microstructure of mixed ethyl-hydroxypropyl-cellulose films. Eur. J. Pharm. Sci. 2013, 48, 240–248. [CrossRef] [PubMed] Martelo, L.; Jiménez, A.; Valente, A.J.M.; Burrows, H.D.; Marques, A.T.; Forster, M.; Scherf, U.; Peltzer, M.; Fonseca, M.S. Incorporation of polyfluorenes into poly(lactic acid) films for sensor and optoelectronics applications. Polym. Int. 2012, 61, 1023–1030. [CrossRef] Menezes, F.; Fedorov, A.; Baleizão, C.; Valeur, B.; Berberan-Santos, M.N. Methods for the analysis of complex fluorescence decays: Sum of Becquerel functions versus sum of exponentials. Meth. Appl. Fluoresc. 2013, 1, 015002. [CrossRef] Berberan-Santos, M.N. Unpublished work, 2009. Neves, T.; Marques, L.; Martelo, L.; Burrows, H.D. Conjugated polymer-based explosives sensor: Progresses in the design of a handheld device. IEEE Sens. 2014, 11, 1415–1418. Knaapila, M.; Dias, F.B.; Garamus, V.M.; Almásy, L.; Torkkeli, M.; Leppänen, K.; Galbrecht, F.; Preis, E.; Burrows, H.D.; Scherf, U.; et al. Influence of side chain length on the self-assembly of hairy-rod poly(9,9-dialkylfluorene)s in the poor solvent methylcyclohexane. Macromolecules 2007, 40, 9398–9405. [CrossRef] Bradley, D.D.C.; Grell, M.; Long, X.; Mellor, H.; Grice, W.A.; Inbasekaran, M.; EWoo, P.E. Influence of aggregation on the optical properties of a polyfluorene. Proc. SPIE 1997, 3145, 254–259. Dias, F.B.; Morgado, J.; Maçanita, A.L.; Da Costa, F.P.; Burrows, H.D.; Monkman, A.P. Kinetics and thermodynamics of poly(9,9-dioctylfluorene) β-phase formation in dilute solution. Macromolecules 2006, 39, 5854–5864. [CrossRef] Rodrigues, R.F.; Charas, A.; Morgado, J.; Maçanita, A. Self-organization and excited-state dynamics of a fluorene−bithiophene copolymer (F8T2) in solution. Macromolecules 2010, 43, 765–771. [CrossRef] Monkman, A.; Rothe, C.; King, S.F.; Dias, F. Polyfluorene photophysics. In Polyfluorenes; Springer: Berlin/Heidelberg, Germany, 2008.
Sensors 2017, 17, 2532
20.
21. 22.
13 of 13
Dias, F.B.; Knaapila, M.; Monkman, A.P.; Burrows, H.D. Fast and slow time regimes of fluorescence quenching in conjugated polyfluorene−fluorenone random copolymers: The role of exciton hopping and dexter transfer along the polymer backbone. Macromolecules 2006, 39, 1598–1606. [CrossRef] Salinas, Y.; Martinéz-Máñez, R.; Marcos, D.M.; Sancenón, F.; Costero, M.A.; Parra, M.; Gil, S. Optical chemosensors and reagents to detect explosives. Chem. Soc. Rev. 2012, 41, 1261–1296. [CrossRef] [PubMed] Pabst, O.; Perelaer, J.; Beckert, E.; Schubert, U.S.; Eberhardt, R.; Tünnermann, A. All inkjet-printed piezoelectric polymer actuators: Characterization and applications for micropumps in lab-on-a-chip systems. Org. Electron. 2013, 14, 3423–3429. [CrossRef] © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
