The development of sensing systems that can detect trace amount of hydrogen peroxide (H2O2) remains a major challenge in the field of bio-imaging,1 healthcare2 and anti-terrorism.3 Continuing efforts have been focused on detecting H2O2 by different strategies such as infrared/Raman spectroscopy,4 mass spectrometry,5 electrochemical method,6 colorimetric7 and fluorimetric detection.8 Generally these methods either require expensive bulky equipment and tedious sample preparation, or have poor selectivity and limits of detection. Therefore, the development of contrast agents that allow the expedient detection of H2O2 at low concentration is in demand. Peroxalate ester and its derivatives are typical chemiluminogenic compounds highly sensitive to H2O2.9 The reaction between peroxalate ester groups and H2O2 can generate a high-energy intermediate that may transfer energy to the fluorescent dye, resulting in efficient luminescence. This has been proved to be a promising strategy for the expedient and sensitive detecting of H2O2.1, 10 However, they often suffer from poor portability and H2O2 sensing in solution only. Due to the widespread use and toxicity (OSHA PEL = 1 ppm) of H2O2, vapor phase monitoring of H2O2 has become an important industrial health issue.11
To date, there has been a move towards making miniaturized sensors with nanowires due to the high surface-to-volume ratios that facilitate the diffusion of gas molecules into the structures.12 Most recently, the waveguiding single-nanowire has been demonstrated for gas sensing with fast response and high sensitivity with the evanescent coupling method.13 The sensitive detection of a variety of gaseous specimens (e.g. H2, NH3, H2O) on a single-nanowire scale has been realized.14 Following our interest in the design of novel systems for the rapid and sensitive detection of chemical vapors,12e here we propose the design of a core/sheath organic nanowire-based H2O2 sensor. In such a structure, the single-crystalline organic luminescent nanowire was chosen as the core and the H2O2-sensitive peroxalate ester as the shell (Figure 1a). Compared with single-component or uniformly doped nanowires,12e, 14a the core/sheath structures possess unique advantages as optical waveguide sensors: 1) the chemo-selective shell is influenced exclusively by the special analyte, providing identity information of the analytes; 2) the nanowire core is insensitive to the analyte, so the full length of the wire can be put in the analyte atmospheres. There is no need to isolate the light in-/out-coupling positions from the sensing region; 3) the large aspect ratio of the nanocables allows for fast diffusion of the gaseous analyte molecules throughout the shell, leading to an instant capture of the vapor species and a waveguiding amplification of the evanescence coupling.15
In this work, the nanoprobe was fabricated by decorating the single-crystalline 9,10-bis(phenylethynyl) anthracene (BPEA) nanowire with chemo-reactive bis(2,4,5-trichloro-6-carbopentoxy-phenyl) oxalate (CPPO) (see Figure S1 for molecular structures) cladding for rapid and sensitive optical detection of gaseous H2O2. The chemiluminescence (CL) measurement indicates that the CPPO layer is highly sensitive and selective to H2O2 vapors; and the evanescent coupling between the core and shell enables optical changes in the nanowire waveguide, which could be used for spectral analysis in the visible region. The optical waveguide response of the nanowires exposed to H2O2 vapors is remarkably fast (35 ms). The cable-like optical waveguide sensing platform complements the present nanowire field-effect sensors with the ability to rapidly and selectively monitor the amplified optical variation across the wire waveguiding, showing that a single nanocable is very promising for developing optical probes to detect individual response on chip.
Due to the poor mechanical property of CPPO, it is essential to decorate it on single crystal nanowires to fabricate complex structures with better mechanical and optical properties. As shown in Figure 1b, spheric beads with weak blue fluorescence were obtained by drop-casting the ethanol solutions of CPPO (1 mM). When small amount (∼2 mol%) of BPEA (in acetone) were added into the CPPO solution, the beads became uniformly green emissive as shown in Figure 1c (called dye-doped CPPO, or ddCPPO), suggesting that the BPEA molecules were well compatible with the CPPO at low doping ratio. With the increase of the BPEA amount, the excess BPEA monomers began to aggregate and form wire-like structures inside the ddCPPO. Figure 1d shows that BPEA nanowires with bright yellow emission were embedded in the ddCPPO due to the aggregation-induced red-shift of the luminescence of BPEA (Figure S2).16 With the further increase of BPEA ratio, cable-like structures could be achieved, in which the BPEA wires are totally or partially wrapped by the ddCPPO shells as displayed in Figure 1e,f. For comparison, only yellow emissive nanowires were obtained from pure BPEA (Figure 1g).
As is mentioned above, the blend solvents of acetone and ethanol were applied in the synthesis of the wires (see Supporting Information for experiment details), which is essential for the formation of core/sheath structures. Acetone is a good solvent for both BPEA and CPPO, while the solubility of BPEA is low in ethanol. Therefore, with the evaporation of acetone, the decrease of solubility together with the strong intermolecular π–π interactions provide driving forces for the BPEA molecules to aggregate and self-assemble quickly into one-dimensional nano- to submicro-structures.16 After that, the CPPO emulsion is adsorbed onto the surfaces of the BPEA wires, which then induces the gradual crystallization to form the core/sheath structures. The mechanism for the formation of the core/sheath wire structures is illustrated in Figure 2a. Figure 2b–d show the typical fluorescence microscopy, SEM, and TEM images of a single wire, the average core diameter is about 600 nm, and the shell thickness is about 300 nm. The length of the wire could be tuned from tens to hundreds of micrometers, which gives a large aspect ratio.
The large aspect ratio could facilitate the absorption and diffusion of gaseous analyte throughout the shells, leading to an instant capture of the vapor species and a waveguiding amplification of the evanescence coupling. Here the mechanism for H2O2 triggered reaction in the ddCPPO shell is shown in Figure S3.9 First, H2O2 react with the peroxalate ester groups, generating a high-energy dioxetanedione. This cyclic intermediate may transfer energy to the fluorescent dye, via the chemically initiated electro-exchange luminescence (CIEEL) mechanism, decomposing to form carbon dioxide and leading to the luminescence. Figure 3a depicts the H2O2 initiated luminescent spectra of the doped nanocables after exposed to H2O2 (ca. 56.5 ppm) for different periods of time, which were recorded by a fluorescence spectrometer without photoexcitation. The generated CL spectra (λ = 450–600 nm) are almost identical with the PL of the BPEA monomers shown in Figure S2, which indicates that the first singlet excited states of BPEA monomers are the emitting species in the CL process.16 The CL spectra were recorded at five different time points, and the intensity experienced an increasing stage first and then decreased slowly with further increase of the exposure time.
The reaction kinetics was investigated with a modified inert gas flow system and an ultra-weak CL analyzer (see Figure S4 for the setup and methods). Upon exposure to the H2O2 vapor, the luminescence of the core/sheath wires was chemically generated with high sensitivity according to the mechanism shown in Figure S3. The CL intensities were monitored as a function of the H2O2 concentration (CH2O2) ranging from 62 ppb to 56.5 ppm. The CL was measured as the integral of the CL intensity in relative light units (RLU) over the total reaction period. The typical calibration curve (peak area versus ppm H2O2) is given in Figure 3b, in which each point represents three repeated measurements (n = 3) with standard deviations ranging from 6% to 10%.
The plot shows a broad dynamic range over four orders of magnitude with a near-linear detection range from 62 ppb to 1.24 ppm. The average intensities rise with the increase of H2O2 concentration, but tend to be saturated when the H2O2 partial pressure goes up, which can be ascribed to the surface coverage of the adsorbed molecules.12d As shown in Figure 3c, at low H2O2 concentrations (≤1.24 ppm), the core/sheath wires exhibit a near-linear response to the concentration, and the curve can be well fitted with the following equation:
with n = 5 and R = 0.97. The detection limit for H2O2 is 40 ppb based on 11 blank determinations (k = 3),17 showing a much higher sensitivity than most of the existing commercial vapor phase H2O2 detection methods with detection limits ranging from 0.1 to 1 ppm. As demonstrated in Figure 3d, the cable wires show very high sensing selectivity for H2O2, with minimal CL signals on exposure to other volatile reactive oxygen species (ROS), because peroxalate chemiluminescence requires the generation of a transient 1,2-dioxetanedione intermediate, which can be formed only by H2O2.18
The results indicate that the BPEA/CPPO cable-like structures are both sensitive and selective to H2O2 vapors, which could be used for the monitoring of trace amount H2O2 vapors. However, the problem lies in that it needs long time for the CL apparatus to collect enough photon signals to read out in the detection, which is not applicable in some cases when rapid sensing is required.3a It has been reported that the optical waveguide of single composite nanowire was sensitive to the surface changes of the wires, and the output signals could be amplified by the waveguiding, which can be used for the rapid and sensitive detections.13b,c, 14b Furthermore, with recent advances in photonic applications of organic single nanowire, such as lasers19 and detectors,15c, 20 the highly compact and flexible single-nanowire optical sensing scheme may open up vast opportunities for very fast detection in physical, chemical, and biological applications with high sensitivity and small footprint.21
Herein, we employ the as-prepared core/sheath wires for H2O2 sensing based on a H2O2-dependent evanescent power leakage of a waveguiding nanowire. In practical sensing applications, the sensitive element is usually immersed in or exposed to a liquid such as aqueous solution, and single-mode operation is generally required in waveguide-based optical sensing when intensity detection is used to achieve high sensitivity. Also, only when the nanowire works under single-mode condition can it leave a large amount of guided field outside the wire.22 Therefore, here we consider the single-mode waveguiding properties of organic nanowires for optical sensing in the air. Figure 4a provides an overview of the waveguide device prepared for the evanescent sensors. The sensing experiments were carried out by placing the sensing element in a sealed glass chamber with a gas flowing system. The applied core/sheath wire was 25 μm in length and 1 μm in diameter. The waveguide was put on the bottom of the chamber, which provided full access of analyte to the surface of the optical cavity. To perform the H2O2 detection, we launched 351 nm excitation light on one tip of the wire, and measured the light output from the other tip, while changing the surrounding by circulating diluted H2O2 gas inside the chamber (see Figure S5 for the experimental setup).
As depicted in Figure 4b, once the gas mixture containing 6 ppm H2O2 was injected into the sensing region, the output of the waveguide was drastically attenuated (ca. 50%) with the input luminous flux remaining unchanged. The corresponding spectra of the input and output light as a function of exposing time are shown in Figure 4c and d, respectively. The monotonic decrease of the PL output with H2O2 exposure can be used for H2O2 sensing. Here, the sensor response is defined as the difference in optical transmission between the H2O2 free and H2O2 exposed signals over the initial transmittance level. The relative change of optical transmission T, (where T = I/I0, and I0 is the output intensity of the waveguide before gas exposure) of a nanocable waveguide at room temperature is plotted versus H2O2 concentration in Figure 5a. The data displays a saturation point of ca. T/T0 = 0.6 occurring at approximately 10 ppm H2O2 within the response regime between 0 and 20 ppm H2O2. The lowest H2O2 concentration detected thus far is 1 ppm; however, our true limit of detection is currently masked by the large noise level associated with coupling a free space laser into a free-standing cable.13b,c, 14b
Similar runs carried out on BPEA waveguides exposed to H2O2 gas and other nanocable waveguides exposed to only air or argon did not show any optical response. The dependence can be explained as following: when exposed to H2O2 atmosphere, the refractive index of the nanocable decreases due to the generation of surface traps of ddCPPO layer by H2O2 molecules, which leads to a monotonous decrease in index contrast between the nanocable and the substrate, and subsequently in optical confinement of the guided light. As a result, the waveguiding light tends to suffer high evanescent leakage due to its high fraction of evanescent waves.14a
The instant response of the sensor was investigated by introducing sudden changes of the H2O2 vapors in the chamber,14c with typical time-dependent transmittance shown in Figure 5b. The estimated response time of the sensor is about 35 ms when H2O2 vapor jumps from 6 ppm to 12 ppm, which is 1 to 2 orders faster than those of existing H2O2 sensors.11b The remarkably fast response of the sensor can be attributed to the small diameter and large surface-to-volume ratio of the nanocable that enable rapid diffusion and capture of the H2O2 molecules, as well as fast signal retrieval using the optical approach.15b Additional improvement in the waveguide design will have a significant impact on the speed of these devices, pushing detection times well into the microsecond regime. The path length required for this waveguide sensing platform to achieve reasonable sensitivity and fast response is extremely short compared to other fiber-based gas detectors. Here we demonstrate a 25 μm interaction length, which is substantially shorter than the previously reported 2 mm to 1 cm interaction lengths.23 We achieve this by utilizing the high surface area, highly reactive CPPO claddings, and exposing them to the evanescent field of the BPEA waveguides.
It should be noted that the core/sheath nanoprobe can distinguish H2O2 from ozone and other organic volatile peroxides (e.g., benzoyl peroxide, peroxybenzoic acid, teralin hydroperoxide and tert-butyl hydroperoxide), as is clearly shown in Figure 5c, because the CPPO cladding is insensitive to those organic peroxides.24 Water molecules are ubiquitous in the environment and humidity is a parameter that must be considered in a nanowire-based device for reliable H2O2 sensing.15 We studied the influence of relative humidity (RH) on the response of gaseous H2O2 by measuring the optical transmission of the BPEA/CPPO nanowire in different humidity. Figure 5d shows that the influence of humidity on the sensor is quite limited. Such a result may be the comprehensive result of the following two factors: on one hand, the response of the nanocable to H2O2 is counteracted by RH due to the competition of H2O and H2O2 in the vapor diffusion; on the other hand, the reaction of CPPO and H2O2 can be accelerated in the presence of H2O molecules, enhancing the sensitivity of nanoprobe to H2O2. The fast response and high selectivity of this waveguide sensor can offer several advantages such as the possibility of remote sensing and feasibility of multiplexing the information from different sensors in single nanowire, which may open interesting perspectives for highly selective detection in complex chemical and biological environments on the optical chip.
In summary, we have developed a general approach to preparing organic core/sheath nanowires with waveguiding core and chemiluminogenic cladding. The cladding was sensitive and selective to the H2O2 molecules proved by the measurement of chemiluminescence responses. By utilizing the evanescence wave coupling of the core and shell, we demonstrated an organic single-wire optical sensor with both fast response and high selectivity for H2O2 vapor probing. The use of a single nanowire for optical waveguiding not only bestows the sensor with a small footprint, fast response, and high sensitivity, but also enables its integration into the chip. Since the chemical properties of organic materials could be effectively tailored through rational molecular design, organic single-nanowire optical sensors for a variety of specimens can be realized on the basis of the above-mentioned optical detecting scheme.