Open Access
Add to BibSonomy
Abstract
In this manuscript, we explored the performance of a hollow thin film array (HTFA) for the detection of HCl vapor based on fluorescence quenching. The HTFA structure was fabricated by manually stacking layers of an active thin film and a supporting film, alternately, with a hollow structure in each supporting film. The total penetration depth of vapor molecules in the HTFA sample is 2n times increased, where n is the layer number of the active thin film. We tested the sensing performance of the HTFA sample using fluorescence emission and laser emission in a Fabry-Pérot (FP) microcavity. In the fluorescence sensing, the sensing efficiency increases with the vapor concentration, and can be as high as 80% with a vapor concentration of 400 ppm. While in the laser sensing, the efficiency can achieve 100% with an external pump intensity three times of the lasing threshold at a vapor concentration of 85 ppm. The HTFA sample is not only suitable for vapor detection but also suitable for molecule detection in liquid.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Forced by the increasing number of security accidents and potential survival crises, vapor detection is currently one of the highlight research topics. Vapor with low concentration may also have serious consequences, which pushes the limits for most vapor detection methods. Fluorescence quenching has received great attention in this field due to its high sensitivity and rapid detection [1–11]. The quenching response depends on the energetics of electron transfer from the photo-excited sensing material to the analytes, the vapor pressure of the analyte, and the diffusion process of the analytes in the sensing films. Once the sensing material and the analyte are determined, the sensing sensitivity mainly depends on the diffusion process: when the analyte binds to the surface of the film, it will penetrate into the film and the fluorescence quenching occurs due to the non-radiative deactivation by an electron transfer mechanism or other interaction effects; the penetration depth usually depends on the concentration of the analyte. Thus, the sensing efficiency is defined as the ratio of the fluorescence intensity change in the sensing film during the exposure to the initial signal intensity before vapor exposure. An effective way to enhance the sensitivity is to expand the deactivation ratio by increasing the penetration depth or decreasing the thickness of the sensing films. For example, thin films with mesoscale porosity usually have large interaction surface area (i.e. an increase in the penetration depth) and fast response time [12–14]. For detection of trace vapors, such as explosive analytes in a scale of ppb, the thickness of the sensing film can be as thin as a few nanometers [1,2,6–8]; while a thinner film may suffer from less control of the morphologies and lead to less intense fluorescence signal than the thicker films, yielding sensors with a decreased signal-to-noise ratio. Another way to improve the sensitivity is to introduce laser emission as the sensing signal with the advantages of laser sensing in narrow bandwidth, fast response time, high conversion efficiency, and high signal-to-noise ratio. A wide researches [9–12,15–18] have mentioned that a distributed feedback laser (DFB) in highly directional emission, signal mode emission, and low laser threshold could offer advantages for reliable and highly sensitive vapor sensing. However it is worth noting that the sensing films for a DFB laser is usually thicker than 100 nm, for a thinner film (less than 100 nm) may be hard to support the laser mode [16]. Thus for low-concentration vapor detection, DFB lasers may suffer from a small distinction in laser thresholds or laser conversion efficiencies before and after the vapor exposure, for a low vapor concentration usually makes a small penetration depth in the sensing film.
Here, in this work, we fabricate a structure named as hollow thin film arrays (HTFA). It was assembled by manually stacking layers of an active thin film and a supporting film, alternately, with a hollow rectangular structure in each supporting film. Thus, there is a hollow channel between two adjacent active films in the structure; vapor molecules can enter into the hollow channels and interact with the two surfaces (upper and down) of each active film. The penetration depth in each layer of active film is doubled; thus the total penetration depth in the HTFA sample is 2n times increased, where n is the number of the active thin film layers (as shown in Fig. 2(a)).
Fluorescein sodium (FS), a dye which is sensitive to acid vapor, was doped in the thin active film as the fluorescence probe and hydrochloric acid (HCl) vapor was used as the the analyte, considering its hazardous nature to human health and the environment. Many researches have been reported in the detection of HCl vapor, such as quartz-enhanced photoacoustic spectroscopy [19], and quartz-enhanced photothermal spectroscopy [20] with a vapor detection limit in ppb-level.
Here, in experiment we tested the performance of the HTFA sample to HCl vapor sensing based on the fluorescence emission and lasing emissions in a Fabry-Pérot (FP) microcavity. In the fluorescence sensing, the sensing efficiency increases with the increase of the vapor concentration, and can be achieved as high as 80% at a vapor concentration of 400 ppm; while in the lasing sensing, the efficiency can reach up to 100% at an external pump intensity three times of the lasing threshold in a faster response time at a vapor concentration of 85 ppm. In the current work, the thickness of the thin active film is 270 nm. We believe that by choosing proper (chemically and physically stable) polymers, the active film can be fabricated with a much thinner thickness, for example 100 nm or less, and HCl vapor with a much lower concentration may be detected. The HTFA sample is not only suitable for vapor sensing but also suitable for molecule sensing in fluid.
2. Fabrication and assembly
The HTFA sample was assembled by manually stacking layers of active thin films and layers of supporting films alternately on the surface of a SiO2 substrate. The active layer and the supporting layer was made of Ethyl cellulose (EC) film doped with and without a FS dye, respectively. The FS dye is the probe for HCl vapor sensing.
The specific process is shown in Figs. 1(a)–1(d). First, we would like to make a layer of substrate-less thin active film, as shown in Figs. 1(a1)–1(c): a film was fabricated by spin-coating polystyrene (PS) and then EC doped with FS on a SiO2 substrate, as shown in Fig. 1(a1). Then, the fabricated film was peeled off from the SiO2 substrate, and immersed in the cyclohexane solution (Fig. 1(b)). When immersed for a while, the EC film was released from the substrate by dissolving the PS film (Fig. 1(c)), then a substrate-less thin active film was obtained. Same method can be used to fabricate a layer of substrate-less supporting film except that an hollow channel (height equals to the thickness of the supporting film) had to be made in the middle of the EC film (doped without FS) before peeling off, as depict in Fig. 1(a2); The hollow channel (width: 200 µm, length: 10 mm) was made by drawing out a straight line in the EC film using a thin blade. After complete dissolution of PS, the substrate-less active films and supporting films were picked up from cyclohexane solution, then were stacked alternately on the surface of a SiO2 substrate (or on the surface of mirror with high reflectivity) under a microscope to fabricate the HTFA structure by self-adhesion when films were dried, as shown in Fig. 1(d). All the channels in the supporting films were aligned up and down. The fabricated HTFA containing ten active thin films and ten supporting films are drawn in Fig. 1(e); Both sides of the film array along the channel direction were cut off to get channels with inlet and outlet. The inset in Fig. 1(e) schematically shows the amplification of the edge of the HTFA. The channels act as the hollow structures to allow the detected vapor to flow into the structures and interact with each thin active film from the two surfaces of it. The thickness of the active thin film and the supporting film is 270 nm and 2.8 µm, respectively; thus, the height of the channel is 2.8 µm. The width of the channel was set to be 200 µm to ensure that any two adjacent active films will not stick together. For comparison, a thick active film with thickness of 2700 nm, the total thickness of active thin films as in the HTFA sample, was spin-coated directly on the surface of a substrate, as shown in Fig. 1(f); also, a thin film with thickness of 270 nm was spin-coated directly on the surface of a substrate.
Fig. 1. Fabrication and assembly of the HTFA. (a1) PS film and then the active thin film (EC film doped with FS) spin-coated on a SiO2 substrate; (a2) PS film and then the supporting film (EC film doped without FS) spin-coated on a SiO2 substrate; (b) Films immersed in cyclohexane; (c) Active thin film and supporting film released by dissolving the PS layer; (d) Stacking layers of active thin films and supporting films alternately on the substrate of a mirror; (e) HTFA structure on a mirror. Inset schematically shows the amplification of the edge of the HTFA; (f) Thick active film spin-coated directly on the substrate of a mirror; (g) Assembled FP microcavity with the HTFA sample.
Fig. 2. Models of the three films: a HTFA sample (a), a thick active film (b), and a thin active film (c).
3. Results and discussion
3.1 Theoretical analysis
The schematic diagram of the HTFA sample and its interaction with HCl vapor is illustrated in Fig. 2(a). Due to the hollow structures of the HTFA, HCl vapor will interact with the upper and down surfaces of each active film layer. z is the penetration depth of the vapor molecules in each layer, d is the thickness of each active film, n is the number of the active film layers, and h is the thickness of each supporting film. The vapor diffusion was assumed to be uniform and up to a maximum penetration depth from the film surface when the concentration of the vapor is the same. The dependance of the fluorescence emission intensity I on the propagation distance x in the sample can be written as:
where A is a constant, P is the pump energy, α is the absorption coefficient at the pump wavelength, and Q is a quenching factor of the emission caused by the additional non-radiative emission in the region with vapor exposure; for example, in the region without vapor exposure, Q = 0. The intensity of the emitted fluorescence signal at the end of a HTFA sample before and after the vapor exposure can be obtained, respectively, by integrating Eq. (1).Similarly, the fluorescence intensity at the end of a solid active film before and after the vapor exposure can be expressed as
3.2 Fluorescence emission
First, we have to prepare the HCl vapor in a ppm scale. Liquid hydrochloric acid was dropped into a sealed container for a few hours for complete volatilization of HCl vapor. For a concentration of 1 ppm, the ratio of the volume of the HCl vapor to the volume of all the containers (including the sealed container mentioned above, a pipe and a transparent sealed box mentioned below) was set to be 1 µL/L. By evaporating 1.63 µg of pure liquid HCl, 1 µL of HCl vapor could be obtained, which was estimated from the molar mass to molar volume ratio at room temperature. The mass of pure liquid HCl can be calculated from the concentrated HCl liquid with a mass percentage of 37%.
We explored the detecting of HCl vapor in a HTFA sample based on the fluorescence emission. There were ten layers of active thin films in the HTFA sample. The experimental setup is depict in Fig. 3(a). Then the vapor diffused into a transparent sealed box with the HTFA sample in it through a pipe. A laser beam from a diode laser (λ@405 nm) was focused by a lens, then was incidence into the HTFA through the sealed box. The focal spot was about 100 µm in diameter, and was located in the hollow channels of the HTFA sample. The emitted signal was collected through an optical fiber and then was sent to a spectrometer (Ocean USB 4000) for analysis.
Fig. 3. (a) Experimental setup for HCl vapor sensing with the fluorescence emission; (b) Experimental setup for HCl vapor sensing with the lasing emission.
Figure 4(a) shows the fluorescence emission spectra from the pre-exposed and 60-min post-exposed films at a pump power P = 0.05 mW at a vapor concentration of 400 ppm. At the emission wavelength of 518 nm, the sensing efficiency can be obtained as high as η1 = 80%. While in one thick active film with the same thickness of active materials as that in the HTFA sample, the fluorescence sensing efficiency is only η2 = 3.9% at 518 nm, as shown in Fig. 4(b). The ratio of the sensing efficiencies in the two structures is η1/η2 = 20.5, which is in accordance with the theoretical calculation. The main difference in sensing efficiency is due to the hollow structures and the two-surface penetration of the vapor into each layer of the film in the HTFA sample. In an active thin film with thickness equal to 270 nm, the fluorescence emissions were obtained under the same experimental condition, as depicted in Fig. 4(c). From the results, we can see that the fluorescence emission in a thin film is very weak and full of noise. By smoothing the spectrum, a sensing efficiency η3 = 38% at the emitted wavelength of 518 nm is obtained, which is about two times smaller than that in the HTFA sample.
Fig. 4. Fluorescence emission spectra at the pump power 0.05 mW from the pre-exposed and post-exposed films in the HTFA sample (a), the thick active film (b), and the thin active film (c) at a vapor concentration of 400 ppm. (d) Dependance of the sensing efficiency on the vapor concentration in the HTFA sample at the pump power 0.05 mW.
Fluorescence emissions from the HTFA sample after 60-min vapor exposure at different vapor concentration were also experimentally measured; accordingly the sensing efficiency is calculated and plotted then in Fig. 4(d). The sensing efficiency decreases almost linearly with the decrease of the vapor concentration. For example, with a vapor concentration of 85 ppm, the sensing efficiency is calculated as 25% at the emission wavelength of 518 nm; with a vapor concentration of 10 ppm, the sensing efficiency is only 3%, which is the lowest vapor concentration we can detected from the fluorescence emission.
Lower vapor concentration/pressure will cause a smaller penetration depth z, thus a lower sensing efficiency. In the thick active film, the intensity difference of the emission signal from the pre-exposed and post-exposed films are too small to be distinguished at a low vapor concentration; while in a thin active film, the emission signal or signal difference is also very weak and noisy at the low vapor concentration. Thus our structure is more sensitive in vapor sensing.
3.3 Laser emission
To generate laser signal for sensing based on the structure of the HTFA, a FP microcavity was fabricated, as shown in Fig. 1(g). The mirrors of the cavity are dielectric-coated and have high reflectivity. The length of the FP microcavity is 30 µm. The experimental setup is depicted in Fig. 3(b). A pulsed laser (λ@450 nm, 5 ns) was normally incident into the cavity through one mirror with a focal spot size of 100 µm, and the focal spot was located in the hollow channels of the HTFA sample. The pump power was controlled with a neutral density attenuator. The emission signal was collected with an optical fiber and then sent to a spectrometer (Horiba 320). The HCl vapor after the volatilization of hydrochloric acid passed into the sealed box at a concentration of 85 ppm.
Figure 5(a) shows the spectrally integrated emission as a function of the pump power density from the pre-exposed and post-exposed HTFA sample. The lasing threshold is 51 µJ/mm2 and the laser efficiency (curve slope) is 99 /(µJ/mm) in the pre-exposed sample; after 10-min exposure, the lasing threshold increases to 317 µJ/mm2 (about 6 times of that in the pre-exposed sample) and laser efficiency decreases to 43/(µJ/mm). The large difference in the lasing threshold and laser efficiency between the pre-exposed and post-exposed samples is due to a large fraction of vapor-penetration area in the hollow structures. If we pump the sample at a pumping density Ip between the two lasing thresholds, i.e. 51 µJ/mm2 < Ip < 317 µJ/mm2, we can obtain a sensing efficiency η of about 100%. In experiment, the lasing performance is not stable near the lasing threshold, so we can choose the pumping density Ip, for example, about three times of the lower lasing threshold; as shown in Fig. 5(b), at the pumping power density Ip = 175 µJ/mm2, we can obtain a stable lasing signal at the pre-exposed HTFA sample. There are two lasing peaks representing a corresponding laser emission at λ1 = 562.8 nm and λ2 = 581.9 nm. According to the lasing mode condition of oscillation in a FP microcavity, the two lasing peaks represent a laser oscillation with a corresponding mode number m1 = 111, and m2 = 106, taking into account that the refractive index of the thin film and air is 1.47 and 1, respectively. After 10-min exposure to the HCl vapor, no signal can be detected.
Fig. 5. Spectrally integrated emission as a function of the pump power density from the pre-exposed and post-exposed HTFA sample (a) and from one layer of pre-exposed and post-exposed thick film (c). Dots: experimental data; lines: linear fit. (b) Lasing emission spectra from the pre-exposed and post-exposed HTFA sample at the pump power density of 27 µJ/mm2. (d) Lasing emissions spectra from one layer of pre-exposed and post-exposed thick film at the pump power density 175 µJ/mm2.
For comparison, one thick film with the same thickness of active materials as in the HTFA sample was tested for the lasing sensing, as shown in Figs. 5(c) and 5(d). In the pre-exposed sample, the lasing threshold is 21 µJ/mm2, and the laser efficiency is 182 /(µJ/mm); after 10-min exposure to the HCl vapor, the lasing threshold increases to 29 µJ/mm2 (1.4 times that of the pre-exposed sample), and the lasing efficiency decreases to 146 /(µJ/mm). The changes in lasing threshold and lasing efficiency between the pre-exposed and post-exposed thick film are much smaller than that between the pre-exposed and post-exposed HTFA sample, and are probably not big enough for a stable lasing sensing in experiment. Lasing emission was not detected in a thin film of 270 nm thick, for the gain medium in the film is not enough for laser oscillation in the FP microcavity.
Vapor sensing based on laser emission in our HTFA structure presents many advantages, as we discussed above, such as high sensing efficiency, fast response time than the fluorescence emission, large differences in laser behavior before and after the vapor exposure, thus high stability. While, till now, our work still have some limitations in, for example, high laser threshold and high pumping power which will cause a short lifetime of the structure in practice. Further work has to be focused on the reduction of the laser threshold; some methods may be feasible, for example, increasing the number of the layers of the active thin films, adopting fluorescence materials with high luminescence or using a FP microcavity with a much higher Q value.
4. Conclusion
In summary, we have demonstrated a HTFA sample for HCl vapor detection. The HTFA sample was constructed by manually stacking layers of active thin films and layers of supporting films (with a hollow structure in it) alternately on a substrate. The fluorescence sensing of HCl vapor with different vapor concentration was measured with a detection limit of 10 ppm. The fluorescence sensing efficiency decreases with the decrease of the vapor concentration. For example, the sensing efficiency is 80% at a vapor concentration of 400 ppm, and reduces to 25% at a vapor concentration of 85 ppm. The laser sensing of HCl vapor was also tested in the HTFA sample at a vapor concentration of 85 ppm; experimental results indicated that the lasing threshold in the post-exposed sample increased to 6 times of that in the pre-exposed sample; with a pumping energy far from the lasing thresholds, a stable sensing can be achieved with a sensing efficiency as high as 100%. The vapor can penetrate into the HTFA structure and interact with each surface of the active films, thus has more interacting area compared with the one layer of film. The device is also suitable for the detection of molecules in fluids. Our work may provide another way in highly sensitive analysis and detection of the molecules.
Funding
National Natural Science Foundation of China (11574186, 11574228, 11874276).
Disclosures
The authors declare that there are no conflicts of interest related to this article.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. Y. Liu, R. C. Mills, J. M. Boncella, and K. S. Schanze, “Fluorescent Polyacetylene Thin Film Sensor for Nitroaromatics,” Langmuir 17(24), 7452–7455 (2001). [CrossRef]
2. S. J. Toal and W. C. Trogler, “Polymer sensors for nitroaromatic explosives detection,” J. Mater. Chem. 16(28), 2871 (2006). [CrossRef]
3. M. S. Meaney and V. L. McGuffin, “Investigation of common fluorophores for the detection of nitrated explosives by fluorescence quenching,” Anal. Chim. Acta 610(1), 57–67 (2008). [CrossRef]
4. N. Venkatramaiah, A. D. G. Firmino, F. A. Almeida Paz, and J. P. C. Tomé, “Fast detection of nitroaromatics using phosphonate pyrene motifs as dual chemosensors,” Chem. Commun. 50(68), 9683–9686 (2014). [CrossRef]
5. M. Xu, J.-M. Han, C. Wang, X. Yang, J. Pei, and L. Zang, “Fluorescence Ratiometric Sensor for Trace Vapor Detection of Hydrogen Peroxide,” ACS Appl. Mater. Interfaces 6(11), 8708–8714 (2014). [CrossRef]
6. H. Q. Zhang and W. B. Euler, “Detection of gas-phase explosive analytes using fluorescent spectroscopy of thin films of xanthene dyes,” Sens. Actuators, B 225, 553–562 (2016). [CrossRef]
7. S. Shoaee, S. S. Y. Chen, H. Cavaye, A. R. G. Smith, P. L. Burn, I. R. Gentle, P. Meredith, and P. E. Shaw, “Assessing the sensing limits of fluorescent dendrimer thin films for the detection of explosive vapors,” Sens. Actuators, B 239, 727–733 (2017). [CrossRef]
8. Z. Wang, Z. Y. Wang, J. Ma, W. J. Bock, and D. Ma, “Effect of film thickness, blending and undercoating on optical detection of nitroaromatics using fluorescent polymer films,” Polymer 51(4), 842–847 (2010). [CrossRef]
9. A. Rose, Z. Zhu, C. F. Madigan, T. M. Swager, and V. Bulović, “Sensitivity gains in chemosensing by lasing action in organic polymers,” Nature 434(7035), 876–879 (2005). [CrossRef]
10. Y. Yang, G. A. Turnbull, and I. D. W. Samuel, “Sensitive Explosive Vapor Detection with Polyfluorene Lasers,” Adv. Funct. Mater. 20(13), 2093–2097 (2010). [CrossRef]
11. Y. Wang, P. O. Morawska, A. L. Kanibolotsky, P. J. Skabara, G. A. Turnbull, and I. D. W. Samuel, “LED pumped polymer laser sensor for explosives,” Laser Photonics Rev. 7(6), L71–L76 (2013). [CrossRef]
12. Y. Wang, N. B. McKeown, K. J. Msayib, G. A. Turnbull, and I. D. W. Samuel, “Laser Chemosensor with Rapid Responsivity and Inherent Memory Based on a Polymer of Intrinsic Microporosity,” Sensors 11(3), 2478–2487 (2011). [CrossRef]
13. I. A. Levitsky, W. B. Euler, N. Tokranova, and A. Rose, “Fluorescent polymer-porous silicon microcavity devices for explosive detection,” Appl. Phys. Lett. 90(4), 041904 (2007). [CrossRef]
14. D. Kim, D. Hahm, S. Kwon, J. Lim, and K. Char, “Controlled Phase Separation in Poly(p-phenyleneethynylene) Thin Films and Its Relationship to Vapor-Sensing Properties,” Langmuir 35(11), 4011–4019 (2019). [CrossRef]
15. S. Richardson, H. S. Barcena, G. A. Turnbull, P. L. Burn, and I. D. W. Samuel, “Chemosensing of 1,4-dinitrobenzene using bisfluorene dendrimer distributed feedback lasers,” Appl. Phys. Lett. 95(6), 063305 (2009). [CrossRef]
16. M. Umar, K. Min, and S. Kim, “A physically transient and eco-friendly distributed feedback laser chemosensor for detecting acid vapor,” Sens. Actuators, B 255, 3207–3215 (2018). [CrossRef]
17. H. Jung, K. Min, H. Jeon, and S. Kim, “Physically Transient Distributed Feedback Laser Using Optically Activated Silk Bio-Ink,” Adv. Opt. Mater. 4(11), 1738–1743 (2016). [CrossRef]
18. A. Camposeo, P. Del Carro, L. Persano, K. Cyprych, A. Szukalski, L. Sznitko, J. Mysliwiec, and D. Pisignano, “Physically Transient Photonics: Random versus Distributed Feedback Lasing Based on Nanoimprinted DNA,” ACS Nano 8(10), 10893–10898 (2014). [CrossRef]
19. Y. Ma, Y. He, X. Yu, C. Chen, R. Sun, and F. K. Tittel, “HCl ppb-level detection based on QEPAS sensor using a low resonance frequency quartz tuning fork,” Sens. Actuators, B 233, 388–393 (2016). [CrossRef]
20. Y. Ma, Z. Lang, Y. He, S. Qiao, and Y. Li, “Ultra-Highly Sensitive Hydrogen Chloride Detection Based on Quartz-Enhanced Photothermal Spectroscopy,” Sensors 21(10), 3563 (2021). [CrossRef]
