Dual-Mode Optical Sensing of Organic Vapors and Proteins with

Article pubs.acs.org/Langmuir

Dual-Mode Optical Sensing of Organic Vapors and Proteins with Polydiacetylene (PDA)-Embedded Electrospun Nanofibers Bryce W. Davis,† Andrew J. Burris,† Nakorn Niamnont,‡ Christopher D. Hare,† Chih-Yuan Chen,† Mongkol Sukwattanasinitt,‡ and Quan Cheng*,† †

Department of Chemistry, University of California, Riverside, Riverside, California 92521, United States Organic Synthesis Research Unit, Department of Chemistry, Faculty of Science, and Center for Petroleum, Petrochemicals and Advanced Materials, Chulalongkorn University, Bangkok 10330, Thailand

‡

S Supporting Information *

ABSTRACT: Optical sensors capable of colorimetric visualization and/or fluorescence detection have shown tremendous potential for field technicians and emergency responders, owing to the portability and low cost of such devices. Polydiacetylene (PDA)-enhanced nanofibers are particularly promising due to high surface area, facile functionalization, simple construction, and the versatility to empower either colorimetric or fluorescence signaling. We demonstrate here a dual-mode optical sensing with electrospun nanofibers embedded with various PDAs. The solvent-dependent fluorescent transition of nanofibers generated a pattern that successfully distinguished four common organic solvents. The colorimetric and fluorescent sensing of biotin−avidin interactions by embedding biotinylated-PCDA monomers into silicareinforced nanofiber mats were realized for detection of biomolecules. Finally, a PDA-based nanofiber sensor array consisting of three monomers has been fabricated for the determination and identification of organic amine vapors using colorimetry and principal component analysis (PCA). The combination of PCA and the strategy of probing analytes in two different concentration ranges (ppm and ppth) led to successful analysis of all eight amines.

■

INTRODUCTION The development of new sensor technology is driven by improving sensitivity, selectivity, response time, reproducibility, shelf life, and simplicityall of which are dependent directly on the property of the sensing materials.1 Electrospun polymer nanofibers have gained much attention as sensing materials due to their inherently high surface-area-to-volume ratio, surface functionality, superior mechanical performance, low cost, and ease of construction.2−4 Electrospun nanofibers have been incorporated into a variety of detection schemes, including colorimetric, fluorescent, and electrochemical approaches.5,6 Colorimetric readout is a desirable sensing technique because it is highly useful to field technicians and emergency responders, owing to the portability and low cost of the devices. Conjugated polymers have been broadly explored as colorimetric sensing materials due to their attractive optical and electrical properties.7 These properties primarily result from an extensive delocalized π-system and intrinsic conformational restrictions within the polymer chain. Conjugated polymers that have been used in sensing matrices include polyaniline, polyphenylene, polypyrrole, polythiophene, and polydiacetylene (PDA).8 When the backbones of these conjugated polymer chains are perturbed, the delocalized π-network induces changes in electronic absorption and emission properties. Since the backbone of these conjugated polymers can easily be modified, a variety of different surface ligands and substrate functionalities © 2014 American Chemical Society

can be applied to design a chemo/biosensor for specific analyte detection. Among the conjugated polymers reported to date, PDAs are of particular interest because they can change color from blue to red upon response to heat (thermochromism), mechanical stress (mechanochromism), organic solvents (solvatochromism), and ligand−receptor interactions (affinochromism).9 Owing to the topochemical constraints of the PDA backbone, polymerization occurs only in solids or other highly ordered structures (for instance, liposomes). The polymerization reaction is typically induced by UV irradiation at 254 nm. For bioanalysis, the analyte binding to surface bound receptors triggers a distortion of the conjugation plane (ene−yne backbone) of the polymer, which leads to chromatic phase transition in the sensor, providing colorimetric detection of the target molecule.7 This strategy allows molecular recognition and optical reporting to be built within a single supramolecular assembly and can be conveniently applied to direct, one-step colorimetric detection of a variety of analytes of interest. PDA-based sensors have been developed to detect many different analytes, including viruses, bacteria, lipophilic enzymes, ions, antibodies, peptides, proteins, and pharmacologically active compounds.10 Received: May 6, 2014 Revised: July 13, 2014 Published: July 15, 2014 9616

dx.doi.org/10.1021/la5017388 | Langmuir 2014, 30, 9616−9622

Langmuir

Article

Figure 1. Structures of diacetylene monomers investigated for electrospun fibrous colorimetric sensors.

quantitative aspects were not explored. Embedding the PDA sensor array into a nanofiber platform would increase specific surface area, which could improve sensitivity, making both qualitative and quantitative analysis possible using commercially available PDAs. Furthermore, biorecognition-induced colorimetric sensing using electrospun nanofibers will be reported. The biotin−STA interaction is one of the most widely used recognition schemes for bioanalytical sensors because the association constant is one of the highest known in biochemistry (1015).19 To our best knowledge, there has been no reported work on using PDA for the detection of biotin−streptavidin (STA) interactions with the solid support. This work may offer a new avenue for developing robust PDA sensors that meet the needs in both chemical and biological detection with qualitative or/and quantitative measurement.

However, as the diacetylene monomers must be organized in a correct geometry to facilitate topochemical polymerization, PDA formation has been limited to Langmuir−Blodgett (LB) films, self-assembled monolayers (SAMs), multilayer coatings, colloids, and immobilized colloids.11−13 Each specific platform offers certain advantages. For instance, LB films and SAMs provide a well-controlled platform for studying the fundamental aspects of PDAs, while colloidal solutions (liposomes and micelles) offer a better platform for liquid handling techniques used in drug discovery applications.14 However, when ruggedness and easeof-operation of biosensing application are required, a solidsupported PDA material is desirable. Kim and co-workers recently reported an approach for constructing silica-reinforced PDA supramolecular structures in electrospun nanofibers.9,15 They demonstrated that PDAconjugated polymers embedded into electrospun nanofibers could be used as a colorimetric sensor for various compounds. Although this work established a qualitative approach for nakedeye detection, no effort was made to differentiate the change in color using a quantitative method, which would give a better analysis of the PDA system.9 In addition, there is no discussion on the fluorescence response of the PDA materials, which can further facilitate the quantitative analysis. In this study, we report the expanded research of electrospun nanofiber sensors fabricated with conjugated PDA assemblies for both fluorescence- and colorimetric-based quantitative detection of organic solvents. The structures of the diacetylene monomers investigated in this work are shown in Figure 1. Detailed quantitative parameters including digital measurement of red− green−blue (RGB) values will be presented. We are particularly interested in characterizing array-based PDA sensors, which have emerged as a powerful tool for the detection of volatile organic compounds (VOCs), disease biomarkers, and pathogens.16−18 Colorimetric arrays surpass single-analyte detection by establishing a unique “fingerprint” for each analyte. Recently, Eaidkong and co-workers developed a paper-based array composed of 8 different PDAs that successfully classified 18 VOCs.16 The authors achieved a very high classification accuracy using only 3 of the 8 PDAs. While their sensor showed excellent qualitative performance, only saturated vapors were studied; thus,

■

EXPERIMENTAL DETAILS

Materials. All chemicals were of the highest analytical grade and were used without further purification. The DA monomers 10,12pentacosadiynoic acid (PCDA), 5,7-eicosadiynoic acid (ECDA), and 10,12-docosadiyndioic acid (DCDA) were purchased from GFS Chemicals (Powell, OH). N-(2-(2-(2-Aminoethoxy)ethoxy)ethyl)pentacosa-10,12-diynamide (PCDA-EDEA) and PCDA-biotin were prepared according to the literature procedures.10 Poly(ethylene oxide) (PEO; Mw = 300 000 g mol−1), tetraethyl orthosilicate (TEOS; reagent grade, 98%), chloroform, and streptavidin (STA) from Streptomyces avidinii (affinity-purified, lyophilized powder, salt-free) were obtained from Sigma-Aldrich (St. Louis, MO). Microscope slides were purchased from Fisher (Pittsburgh, PA), and ethanol (200 proof) was purchased from Gold Shield Chemical Co. (Hayward, CA). All protein solutions were prepared in a 10 mM phosphate buffered saline (PBS; 150 mM NaCl; pH 7.4) using Milli-Q (>18 MΩ) water. Preparation of PDA-Embedded Electrospun Nanofibers. The typical procedure for fabrication of PDA-embedded electrospun polymer nanofibers is as follows. A 3:1 EtOH/CHCl3 solution containing 1.6 wt % DA monomer and 4 wt % PEO was prepared. A TEOS solution was prepared in a separate vial with a 1:4:4 molar ratio of TEOS/EtOH/H2O (pH 1.1). For experiments involving silicareinforced nanofibers, the DA and TEOS solutions were mixed with a 10:1 weight ratio and stirred for 60 min. For all other fibers, the DA solution was stirred for 60 min. The resulting sol−gel was pumped through a capillary connected to a 25 gauge blunt tip needle (Braintree 9617

dx.doi.org/10.1021/la5017388 | Langmuir 2014, 30, 9616−9622

Langmuir

Article

Scientific, Inc.) at a rate of 0.1−1.0 mL h−1 by a syringe pump (KD Scientific, Model 200 series). The application of a high voltage (8−20 kV) to the metal syringe needle led to generation of nanofibers, which were collected on the surface of a grounded aluminum plate (tip-tocollector distance: 10−12 cm). The nanofibers were then stored in the dark. Photopolymerization of DA monomer nanofibers was carried out by UV irradiation at 254 nm (1 mW cm−2) for 5 min. Fluorescence Measurement. For fluorescence measurements, the nanofibers were collected on 11 × 50 mm cut glass slides for 45 s. The polymerized ECDA, PCDA, and PCDA-EDEA nanofibers were exposed to solvent vapor by placing the glass slides on the top cover (inside part) of a glass Petri dish. Each Petri dish contained 2 mL of organic solvent, and the fluorescence change of the nanofibers was monitored after 1 h of incubation at room temperature. For the biotin-STA tests, the polymerized PCDA-biotin nanofibers were immersed into a 250 μg mL−1 STA solution and retracted from the solution before the measurement. All fluorescence spectra were measured in a cuvette (12 × 12 × 45 mm, Fisher), recorded on a HORIBA FluoroLog spectrofluorometer at room temperature (∼25 °C). The fluorescence emission profiles were monitored from 500 to 700 nm (λEx = 490 nm). Each vapor test value is an average of three independent measurements. Colorimetric Detection. For solution-based sensing, nonwoven mats of nanofibers were collected for 10 min each onto aluminum foil and stored in the dark. Photopolymerization was carried out by UV irradiation for 5 min (2.5 min each side). Colorimetric change of the PCDA-biotin fiber mat was induced by applying 100 nL of 500 or 250 μg mL−1 STA solution. Control experiments were performed using 100 nL of 250 or 500 μg mL−1 BSA solution. The colorimetric transition was monitored using a digital microscope (MicroXplore PC200) after 1 min. For gas-phase sensing, nonwoven mats of nanofibers were collected onto aluminum foil for 8 min each and stored in the dark. Photopolymerization was carried out by UV irradiation at the power of 30 mJ/cm2. To fabricate the three-component sensor array, each polymerized nanofiber mat was then cut into pie-shaped pieces of equal size, and each of three different PDA fiber mats (PCDA, ECDA, DCDA) were put together into a single sensor. RGB values for the PDA nanofiber mat were measured immediately before and after exposure to various concentrations of eight different basic amine vapors (diisopropylamine, diisopropylethylamine, hexylamine, piperidine, pyridine, tert-butylamine, triethylamine, and tripropylamine) for 30 min at room temperature (∼25 °C). Experiments were performed in sealed glass vessels equipped with septa. The volatile amines were injected in liquid form with a glass syringe; analytes transitioned completely to the gas phase within a few minutes. Vapor concentrations were calculated in parts per million by volume (ppmv) using the ideal gas law and the physical properties of each amine. All experiments were performed in triplicate. RGB values were measured using a flatbed scanner (CanoScan LiDE 90, Canon) connected to a personal computer. The scanned images (TIF format) were normalized in Adobe Photoshop (despeckle, dust and scratches, median filters) to minimize artifacts introduced by scanner hardware noise. Color change values were calculated from each RGB measurement in terms of red chromaticity level (r)20,21

r=

analysis (PCA) was performed on the data set using R software (version 3.0.1 Mac OS, The R Foundation for Statistical Computing, Vienna, Austria). SEM and TEM Characterization. Scanning electron microscopy (SEM) images of the PDA-embedded nanofibers were obtained on a Phillips XL30-FEG. For transmission electron microscopy (TEM) characterization, PDA-embedded nanofibers were electrospun directly onto a carbon-coated copper mesh and characterized before and after UV-irradiation using a Tecnai T12 transmission electron microscope.

■

RESULTS AND DISCUSSION Fluorescence Sensing of Organic Solvent Vapor with PDA. PDA nanofiber sensors have been reported for organic vapor using a colorimetric readout.9,15 We focused our work on fluorescence detection with PDA-embedded fibers. Fabrication of PDA nanofibers was successfully realized using PEO as a matrix polymer owing to its water solubility and nontoxicity. TEOS, an alkoxide precursor, was also used to enhance the stability of the nanofibers. Photopolymerization of the DA monomer nanofibers occurs under UV irradiation at 254 nm; the appearance of blue color indicates DA polymerization has occurred. Solvatochromic changes in the nanofiber mats were found to be irreversible, affording a convenient platform for colorimetric sensing. The shelf life of the fiber mats was excellent; the unpolymerized white fibers, polymerized blue fibers, and used red fibers showed no apparent change in color or morphology over a 6 month period. Three electrospun nanoassembliesPCDA, ECDA, and PCDA-EDEAwere used for fluorescence response experiments. They exhibited good structural stability against organic solvent vapor and remained uniform (bead-free) fibrous morphology during the process. To keep the sensing conditions comparable and reduce batch-to-batch differences, all fluorescence response experiments were performed with mats containing similar amounts of fiber distribution, and the fibers had diameters of 400−600 nm. No significant morphological difference was observed in the SEM images before and after UVirradiation. However, after exposure to different organic solvent vapors for an extended period of time, the highly ordered lipid nanofibers were disrupted at different rates, which promoted the blue-to-red solvatochromism. Fluorescence detection using PDA nanofibers shows excellent sensitivity against organic vapors. This is because the PDAembedded nanofibers undergo a distinct fluorescence transition that is from none (blue phase) to fluorescent (red phase), and this varies in a solvent-dependent manner. As a result, a practical one-step combinatorial optical sensor can be created. The organic solvent vapor sensing was performed by monitoring the fluorescent transition of the fiber film. The results are shown in Figure 2 for PCDA, ECDA, and PCDA-EDEA nanofibers with 1 h incubation against chloroform, tetrahydroduran (THF), hexane, and methanol. The fluorescence increase of the PCDA and ECDA nanofibers at 640 nm had similar response patterns (hexane < methanol < chloroform < THF), with THF causing the greatest transition to red. The ECDA exhibited greater color shift as compared to PCDA, presumably because the carboxylic chain is much shorter and thus closer to the backbone of the conjugated polymer. The fluorescence response for the amineterminated PCDA-EDEA nanofibers, however, had a very different response pattern, in which the order is hexane < THF < methanol < chloroform. This illustrates a solvent-dependent response pattern that is highly useful for sensing purposes. This property of the PDA-embedded polymer nanofibers allows for a straightforward “fluorescence pattern” signaling procedure to

R R+G+B

where R (red), G (green), and B (blue) are the primary color components with possible values of 0−255. In order to make quantitative comparisons of the data, we follow the convention established in the literature for red chromaticity shift (RCS)

RCS =

rsample − r0 rmax − r0

× 100%

where rsample is the average red chromaticity of the PDA fiber mat surface after exposure to solvent vapor, r0 is the average red chromaticity of the PDA fiber prior to solvent exposure, and rmax is the maximum r value obtained for each PDA type upon exposure to the amine that caused the maximal color shift. Essentially, RCS is a normalized value describing the blue-to-red color change of each PDA fiber mat. Principal component 9618

dx.doi.org/10.1021/la5017388 | Langmuir 2014, 30, 9616−9622

Langmuir

Article

Figure 3. Histogram of the fluorescence emission profiles at 640 nm (λEx = 490 nm) of the polymerized PCDA, ECDA, and PCDA-EDEA embedded electrospun nanofibers against four organic solvent vapors (THF, chloroform, methanol, and hexane) at 25 °C for 1 h. Each value is an average of three independent measurements.

fold higher intensity in response to chloroform over THF, differentiating the two vapors. Methanol and hexane are identified by PCDA-EDEA, which showed a 4-fold higher intensity in the presence of methanol. Characterization of PCDA-Biotin Nanofibers and Colorimetric Sensing of Streptavidin. Morphological characterization of electrospun nanofibers were carried out representatively using fibers encapsulated with PCDA-biotin, as they represent new materials from this study. Figure 4 shows

Figure 4. Photographs of electrospun fiber mat embedded with PCDAbiotin before (a) and after (b) UV-irradiation (1 mW cm−1) for 5 min, (c) is the SEM image of the nanofibers containing polymerized PCDAbiotin, and (d) and (e) are TEM images of PCDA-biotin nanofibers before and after UV-irradiation.

Figure 2. Fluorescence emission profiles of the polymerized PCDA (a), ECDA (b), and PCDA-EDEA (c) embedded electrospun nanofibers after exposure of the organic solvent vapor at 25 °C for 1 h (λEx = 490 nm). Organic solvents: THF (black solid), chloroform (red dash), methanol (blue dot), and hexane (green dash-dot).

optical, SEM, and TEM images of the fiber mats and individual fibers at high magnification. From optical images of the mat before and after UV irradiation, distinctive blue color is well developed after UV light irradiation, indicating effective photopolymerization within the nonwoven fibrous mats. This confirms that the delocalized π-network and the conformational restrictions within the polymer chain backbone are retained in the electrospinning process. The SEM image of a blue-colored

differentiate between several common organic solvents, as shown in the histogram in Figure 3. The histogram shows that THF and chloroform induced higher overall fluorescence intensities compared with methanol and hexane, making the two groups of analytes easily distinguishable. PCDA-EDEA exhibited a 59619

dx.doi.org/10.1021/la5017388 | Langmuir 2014, 30, 9616−9622

Langmuir

Article

trigger of the fluorescence signal from the red color change of the fibers. While fluorescence intensity was relatively low, the blueto-red color shift was clearly visible, suggesting that the colorimetric detection strategy is superior for the STA sensor. Overall, the nanofiber-supported biotin-PCDA sensor was less sensitive compared to vesicle-based systems found in previous studies. Array-Based Colorimetric Detection of Organic Amine Vapors. The sensing capability of the PDA nanofibers was further tested with an array-based measurement in combination with pattern recognition principles for chemical identification. For this purpose, we have developed a three-component PDA nanofiber sensor array, which was tested with 8 volatile amines at 6 different concentrations, yielding a total of 576 measurements for red chromaticity shift (RCS). Upon exposure to amine vapors for 30 min, the PDA sensor array exhibited varied color shifts, as shown in Figure 6 (hexylamine). Differences in the colorimetric

PCDA-biotin nonwoven mat is shown in Figure 4c, further illustrating the large surface area-to-volume ratio formed within the electrospun film. The nanofibers exhibited well-defined, bead-free fibrous morphology with good structural stability. They were continuous, uniform, and had a diameter of approximately 400−600 nm. No significant morphological differences were observed in the SEM images before and after UV-irradiation. Figures 4d,e shows TEM images of the PEO/ TEOS electrospun nanofibers mats encapsulated with PCDAbiotin, before (Figure 4a) and after (Figure 4b) UV-irradiation. The internal structure of the nanofibers can be visualized by the diffraction contrast, which is caused by the electron beam diffraction of different crystal structures within the fiber structure.22 No dark aggregates were found within the fiber before or after UV-irradiation, indicating the PDA polymer chain is aligned, which suggests that the electrospinning process yields uniform dispersion of PDA within the polymer matrix. Furthermore, the TEM images revealed a homogeneous morphology throughout the nanofibers. The diffraction contrast change before (Figure 4d) and after (Figure 4e) UV-irradiation is attributed to the transformation of the diacetylene monomers from a rigid bundle of noncrystalline monomeric rods to flexible crystalline polymer chains. The polymerized blue nanofibers changed color to red when STA was added, which led to protein binding to the biotinylated PDA. As shown in Figure 5, both colorimetric and fluorescence

Figure 6. Flatbed scanner images illustrating the colorimetric response of polydiacetylene-embedded nanofiber sensor array to various concentrations of hexylamine vapor.

response of PDAs are dependent on alkyl chain length and headgroup identity.23 ECDA was the most sensitive in all three PDAs tested in this study, which can be attributed to weaker London forces resulting from its shorter side chains. PCDA and DCDA exhibited varied and weaker sensitivity as compared to ECDA, possibly due to the presence of a longer alkyl chain on PCDA and dual carboxyl head groups on DCDA. We then take advantage of this variability to generate both quantitative and qualitative information for the sensor array. Quantitative analysis of the RCS data shows that the sensor array exhibited a linear response to 4 of the 8 amines at ppmv concentrations (∼10−100 ppmv). The error bars in Figure 7 represent the combined reproducibility of sensor fabrication and data measurement (RGB values). We found that the primary amines hexylamine (Figure 7a) and tert-butylamine (Figure S1a) were detected by the PCDA sensor with good correlation (R2 > 0.97). Two of the linear tertiary amines, tripropylamine (Figure 7b) and diisopropylethylamine (Figure S1b), were detected by the ECDA sensor with good correlation (R2 > 0.98). The secondary amines, diisopropylamine and piperidine, were not detected by the sensor array at the low ppmv range. Pyridine, a tertiary cyclic amine, did not induce a significant RCS in the sensor array. Triethylamine was the only tertiary linear amine that was not quantitatively detected by the sensor. The results highlight the flexibility of our sensor array for quantitative analysis. Differences in PDA sensitivity, coupled with differences in analyte basicity, give the sensor array a powerful flexibility, where each component of a sensor array can be tailored to suit a particular class of analyte.

Figure 5. Optical microscope color changes of polymerized PCDAbiotin embedded electrospun fiber mat caused by the addition of 500 ng (top), 250 ng (middle), and 0 ng (bottom) streptavidin after 1 min (a). Fluorescence emission profile of the polymerized PCDA-biotin embedded electrospun nanofibers after a 5 s immersion into a 250 μg mL−1 streptavidin solution (b).

biotin-STA response can be monitored using the PDA-biotin nanofibers. Figure 5a shows the colorimetric response of microspots after 1 min exposure to 500 (top), 250 (middle), and 0 ng (bottom) of STA in 100 nL. The blue nanofibers were rapidly converted into red fibers when STA was applied to the surface of the biotin-PCDA fibers. No color change was observed when only PBS buffer solution was used as shown in Figure 5a (bottom). Two additional control experiments were performed. First, a 100 nL of a 1 mg mL−1 BSA solution was spotted onto a polymerized biotin-PCDA mat. Second, STA solutions were tested on PCDA, ECDA, and PCDA-EDEA electrospun mats. No visible color change was observed in any of these tests. Fluorescence property was further investigated with biotinPCDA fibers upon addition of STA (Figure 5b). A low but clear fluorescence emission at 640 nm was observed, indicating the 9620

dx.doi.org/10.1021/la5017388 | Langmuir 2014, 30, 9616−9622

Langmuir

Article

linear transformation that reduces dimensionality of multivariate data by clustering similar observations together. Each principal component (PC) is an orthogonal eigenvector that represents the maximal amount of variance in the data set. In essence, PCA extracts the smallest number of components that describe the most variation of the original data, with minimal loss of information. PCA on the red chromaticity shift data resulted in a total of 19 PCs; PC1 and PC2 accounted for 72.3% and 14.8% of the data variance, respectively (Figure 8a). The 2D PCA plot

Figure 8. PCA plots for differentiation of 8 different amine vapors using data from 6 concentrations (a). (b) and (c) are independent plot for 3 lowest (ppm) and 3 highest (ppth) concentrations. Ellipses represent 95% confidence intervals.

Figure 7. Calibration curves for the three-component polydiacetylene sensor array after 30 min exposure to hexylamine (a) and tripropylamine (b) vapor at ppm concentrations.

The analyte-dependent colorimetric response of the senor array to different basic amines can be explained in terms of basicity and steric hindrance. Amines with higher basicity have greater affinity for the carboxylic acid head groups on PDA. Disturbing the hydrogen bonds of neighboring carboxylic acid groups induces a conformational change in the ene−yne backbone of PDA, which is accompanied by a shift in absorbance wavelength. For example, the pKa values for hexylamine and tripropylamine are 10.56 and 10.65, respectively. Based on pKa alone, tripropylamine is expected to cause a higher RCS in the sensor array. However, tripropylamine (3°) is sterically more hindered than hexylamine (1°), which explains why hexylamine caused a higher RCS in the sensor. Principal Component Analysis (PCA). Variability in chromatic response among the 3 PDAs produces a pattern that can be analyzed to distinguish different compounds. The sensor provided a large data set with many variables (8 VOCs × 6 concentrations × 3 PDAs × triplicate analysis = 576 values for RCS); pattern recognition of the data requires a multivariate analysis method. Principal component analysis is a mathematical

in Figure 8a clearly shows 8 separate clusters, indicating that the sensor array successfully identified all 8 amines. For comparison, a PCA plot of the data of data from the 3 lowest (ppm) concentrations (Figure 8b) shows that 2 of the 8 amines are distinguished (piperidine and pyridine). Using only the 3 highest (ppth) concentrations (Figure 8c), the sensor array identified 6 of the 8 amines (tert-butylamine and diisopropylamine were not distinguishable). The error ellipses (95% CI) in Figure 8 indicate good reproducibility of the sensing system. These results highlight the ability of the sensor array to successfully classify amine vapors at subsaturation concentrations.

■

CONCLUSIONS A new sensing strategy for the differentiation of volatile organic solvents based on PDA nanofibers using fluorescence and colorimetric response patterns has been demonstrated. The organic-solvent-dependent fluorescence properties allow for a straightforward fluorescent pattern procedure to differentiate several chemical solvents. Using silica-reinforced PDA nanofiber 9621

dx.doi.org/10.1021/la5017388 | Langmuir 2014, 30, 9616−9622

Langmuir

Article

(9) Yoon, J.; Jung, Y. S.; Kim, J. M. A combinatorial approach for colorimetric differentiation of organic solvents based on conjugated polymer-embedded electrospun fibers. Adv. Funct. Mater. 2009, 19, 209−214. (10) Jung, Y. K.; Park, H. G.; Kim, J. M. Polydiacelylene (PDA)-based colorimetric detection of biotin-streptavidin interactions. Biosens. Bioelectron. 2006, 21, 1536−1544. (11) Carpick, R. W.; Sasaki, D. Y.; Marcus, M. S.; Eriksson, M. A.; Burns, A. R. Polydiacetylene films: a review of recent investigations into chromogenic transitions and nanomechanical properties. J. Phys.: Condens. Matter 2004, 16, R679−R697. (12) Okada, S.; Peng, S.; Spevak, W.; Charych, D. Color and chromism of polydiacetylene vesicles. Acc. Chem. Res. 1998, 31, 229−239. (13) Tieke, B. Polymerization of butadiene and butadiyne (diacetylene) derivatives in layer structures. Adv. Polym. Sci. 1985, 71, 79−151. (14) Reppy, M. A.; Pindzola, B. A. Biosensing with polydiacetylene materials: structures, optical properties and applications. Chem. Commun. 2007, 4317−4338. (15) Yoon, J.; Chae, S. K.; Kim, J. M. Colorimetric sensors for volatile organic compounds (VOCs) based on conjugated polymer-embedded electrospun fibers. J. Am. Chem. Soc. 2007, 129, 3038−3039. (16) Eaidkong, T.; Mungkarndee, R.; Phollookin, C.; Tumcharern, G.; Sukwattanasinitt, M.; Wacharasindhu, S. Polydiacetylene paper-based colorimetric sensor array for vapor phase detection and identification of volatile organic compounds. J. Mater. Chem. 2012, 22, 5970−5977. (17) Cheol Hee, P.; Jun Pyo, K.; Sang Wook, L.; Noo, Li. J.; Yoo, P. J.; Sang Jun, S. A direct, multiplex biosensor platform for pathogen detection based on cross-linked polydiacetylene (PDA) supramolecules. Adv. Funct. Mater. 2009, 19, 3703−3710. (18) Kolusheva, S.; Yossef, R.; Kugel, A.; Katz, M.; Volinsky, R.; Welt, M.; Hadad, U.; Drory, V.; Kliger, M.; Rubin, E.; Porgador, A.; Jelinek, R. Array-based disease diagnostics using lipid/polydiacetylene vesicles encapsulated in a sol-gel matrix. Anal. Chem. 2012, 84, 5925−5931. (19) Gonzalez, M.; Bagatolli, L. A.; Echabe, I.; Arrondo, J. L. R.; Argarana, C. E.; Cantor, C. R.; Fidelio, G. D. Interaction of biotin with streptavidin - Thermostability and conformational changes upon binding. J. Biol. Chem. 1997, 272, 11288−11294. (20) Friedman, S.; Kolusheva, S.; Volinsky, R.; Zeiri, L.; Schrader, T.; Jelinek, R. Lipid/polydiacetylene films for colorimetric protein surfacecharge analysis. Anal. Chem. 2008, 80, 7804−7811. (21) Jelinek, R.; Volinsky, R.; Kliger, M.; Sheynis, T.; Kolusheva, S. Glass-supported lipid/polydiacetylene films for colour sensing of membrane-active compounds. Biosens. Bioelectron. 2007, 22, 3247− 3251. (22) Zhang, J. F.; Yang, D. Z.; Xu, F.; Zhang, Z. P.; Yin, R. X.; Nie, J. Electrospun core-shell structure nanofibers from homogeneous solution of poly(ethylene oxide)/chitosan. Macromolecules 2009, 42, 5278− 5284. (23) Charoenthai, N.; Pattanatornchai, T.; Wacharasindhu, S.; Sukwattanasinitt, M.; Traiphol, R. Roles of head group architecture and side chain length on colorimetric response of polydiacetylene vesicles to temperature, ethanol and pH. J. Colloid Interface Sci. 2011, 360, 565−573.

mats as a recognition system, detection of biotin−STA interactions was accomplished. The specificity was achieved with a synthetic biotinylated PCDA monomer, and the target protein STA could be detected both visually and fluorescently. A PDA-based nanofiber sensor array consisting of PCDA, ECDA, and 10,12-docosadiyndioic acid (DCDA) was utilized for the determination and identification of organic amine vapors using colorimetry and principal component analysis. The combination of PCA and use of a strategy based on two different concentration ranges (ppm and ppth) led to successful analysis of all 8 amines. Four of the 8 amines were determined quantitatively at ppm concentrations. To the best of our knowledge, this is the first example of a PDA-embedded electrospun nanoassembly that combines qualitative and quantitative sensing into a single platform. Further efforts will focus on (1) modifying the substrate for better stability to environmental interactions, (2) functionalizing the electrospun nanofibers for specific capture and detection of environmental pathogens or other contaminates, and (3) improving the response of the sensor array for analysis of complex mixtures.

■

ASSOCIATED CONTENT

S Supporting Information *

Figure S1. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Tel (951) 827-2702; e-mail [email protected] (Q.C.). Author Contributions

B.W.D. and A.J.B. contributed equally. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors acknowledge financial support from NSF grant CHE-1059050 (Q.C.) and Thailand’s National Nanotechnology Center grant NN-B-22-FN9-10-52-06 (M.S.).

■

REFERENCES

(1) Zhang, Y. Z.; Lim, C. T.; Ramakrishna, S.; Huang, Z. M. Recent development of polymer nanofibers for biomedical and biotechnological applications. J. Mater. Sci.: Mater. Med. 2005, 16, 933−946. (2) Ding, B.; Li, C. R.; Miyauchi, Y.; Kuwaki, O.; Shiratori, S. Formation of novel 2D polymer nanowebs via electrospinning. Nanotechnology 2006, 17, 3685−3691. (3) McKee, M. G.; Layman, J. M.; Cashion, M. P.; Long, T. E. Phospholipid nonwoven electrospun membranes. Science 2006, 311, 353−355. (4) Huang, Z. M.; Zhang, Y. Z.; Kotaki, M.; Ramakrishna, S. A review on polymer nanofibers by electrospinning and their applications in nanocomposites. Compos. Sci. Technol. 2003, 63, 2223−2253. (5) Davis, B. W.; Niamnont, N.; Hare, C. D.; Sukwattanasinitt, M.; Cheng, Q. A. Nanofibers doped with dendritic fluorophores for protein detection. ACS Appl. Mater. Interfaces 2010, 2, 1798−1803. (6) Nakamura, H.; Karube, I. Current research activity in biosensors. Anal. Bioanal. Chem. 2003, 377, 446−468. (7) Song, J.; Cheng, Q.; Zhu, S. M.; Stevens, R. C. “Smart” materials for biosensing devices: Cell-mimicking supramolecular assemblies and colorimetric detection of pathogenic agents. Biomed. Microdevices 2002, 4, 213−221. (8) Lee, K.; Povlich, L. K.; Kim, J. Recent advances in fluorescent and colorimetric conjugated polymer-based biosensors. Analyst 2010, 135, 2179−2189. 9622

dx.doi.org/10.1021/la5017388 | Langmuir 2014, 30, 9616−9622