Introduction

Air pollution has become an inescapable reality in the lives of both city dwellers and villagers around the world. Therefore, air quality monitoring in both urban areas and country sides have been getting necessary in virtue of the daily massive production of harmful gases, ashes, and hydrocarbons. Such pollutants have caused negative effects on the environment, for instance, acid rain, photochemical smog, greenhouse effect, and so forth. Accordingly, there is a great demand for different types of approaches to detect low concentration of harmful vapors [1,2,3,4], such as NO2, methanol, toluene, chlorobenzene. Traditional gas detection methods consist of mass spectrometry, gas chromatography, and sensors [5]. The former two suffer from disadvantages of high price, bulky volume. New gas testing approaches currently include field-effect transistors (FETs) [4], chemical nanodevices based on dioxide graphene [6].

Nitrogen dioxide is a kind of brown red, highly active gaseous substance. The production of nitrogen dioxide mainly comes from the release of high-temperature combustion, for instance exhaust emissions from motor vehicles and power plants [7]. Nitrogen dioxide is the main damage to the respiratory system. At the initial stage of inhalation, only slight eye and upper respiratory irritation symptoms may occur in human body, such as pharynx discomfort, dry cough. [8,9,10]. As a result, it is very necessary to develop quantitative and qualitative detection methods. Esmaeilzadeh et al. [11] fabricated gas sensor to test NO2 gas with different concentrations in the temperature range of 450–550 °C. Zhang et al. [12] developed field-effect transistor chemical sensors based on carbon nanotube exhibited slow response time. Plastic optical fiber (POF) sensors based on conducting graft polymer were also made to detect below 1 ppm of NO2 in a dry atmosphere [13]. Shortcomings including elevated operating temperatures, slow response time, and sophisticated structures in these ways are obstructions to be got over. Since the 1990s, researchers have followed with interest the study and usage of optical waveguide (OWG) sensors in the domain of optical communication [14]. OWG gas sensors have been developed to measure trace amounts of detrimental gases at home and abroad [15, 16]. At present, there are few reports on the use of OWG gas sensors [17] to detect nitrogen dioxide gas even though OWG sensors possess the edges of fast response, the potentials for high sensitivity, system compactness, and intrinsically safe detection; what’s more, they can be fabricated at a very low cost [18].

The type of sensitive material has great influence on the performance of sensors [19, 20], so it is important to choose the right sensitive material. As a result, there is an increasing demand for more sensitive gas-sensing compounds, which are capable of detecting toxic gases [21,22,23,24]. Organic materials are drawing tremendous attention as gas-sensing element materials owing to their low cost, easily tunable chemical structures, and ease of processing in solution to develop solid-state membranes for sensor device applications [25]. Porphyrin and related compounds have a large π-aromatic system [26], which leads to interesting optical properties [27, 28], including high optical absorption coefficient in the UV–Visible region [29]. As is known to all, when porphyrins are exposed to certain gas molecules, there occur some prominent optical and electrical changes to porphyrins [30,31,32]. Porphyrins have been considered as one of the most prospective sensing substances for optical gas transducers [33]. There are not many reports on the optical waveguide gas sensors based on porphyrin [17].

In this work, we report a simple realization of OWG gas sensor based on the free-base porphyrin (Fig. 1), employing spin coating process to fabricate sensing element, and investigate its sensing performance with NO2. Film morphology was studied by SEM before and after exposure to NO2. Film stability was observed for 20 days with UV–Vis spectroscopy, and reproducible stability of the films upon exposure to gas analyte (NO2) was detected (2, 5, 12, 20 days after the preparation of thin films). In addition, film thickness was detected before and after exposure to certain concentration of NO2 gas.

Figure 1
figure 1

Chemical structure of meso-5,10,15,20-tetrakis-(4-tertbutyl phenyl) porphyrin

Full size image

Experimental details

Instruments and reagents

Instruments

UV-1780 ultraviolet spectrophotometer spectroscopy (SHIMADZU, Japan); N-1100-type rotary evaporator (Beijing Oriental Science & Technology Development Co. Ltd); SHZ-D(III) circulating water vacuum pump (Henan Gongyi yingyu Yuhua Instrument Factory); KW-4A-type spin coater (Shanghai Chemat Technology); Tin-diffused glass OWG detection system (homemade apparatus); FE-SEM (SU8019 field emission scanning electron microscope from Hitachi, Japan); thin film thickness gauge (SGC-10-type thin film thickness gauge from Tianjin Port East Technology Development Limited by Share Ltd); gas detection tube (working range: 2–200 ppm, Gastec, Beijing Municipal Institute, Beijing, China), D8-type X-ray diffraction apparatus (Bruker, Germany).

Reagents

Chloroacetic acid, pyrrole, anisole, chloroform, etc. All reagents are analytically pure and purchased from Beijing Bailingwei Technology Co. Ltd.

Preparation of porphyrin thin-film/Tin-diffused glass OWG device

Meso-5,10,15,20-tetrakis-(4-tertbutyl phenyl) porphyrin (abbreviated as porphyrin) was synthesized by Adler–longo method [34]. The electronic absorption spectra of the porphyrin chloroform solution and its thin film were studied, as given in Fig. 2. It could be observed from Fig. 2 that the UV–Vis absorption spectra of porphyrin solution in chloroform and thin film were similar, but there was a small difference in the shape of Soret band. In chloroform, porphyrin had narrow Soret band; Soret band became substantially broader when porphyrin molecules were coated on glass substrate as a thin film. Plus, the Soret band of solution and film appeared at 420, 432 nm, respectively. That is to say, the characteristic peak of porphyrin was red-shifted 12 nm after porphyrin is being paved into thin film, indicating that the head-to-tail (J-type) aggregation might be involved upon formation of porphyrin aggregates in films [35, 36]. These changes have been previously observed in porphyrin films [37,38,39]. In general, when the porphyrin molecules were evenly paved on glass slide, the aggregation effect was minimized and the absorption spectrum of a porphyrin film was identical with that of solution [35].

Figure 2
figure 2

UV–Vis absorption spectra of porphyrin in chloroform and porphyrin thin film

Full size image

Then, certain amount of porphyrin was dissolved in 10 mL chloroform to obtain different mass percentages of porphyrin solutions. The solution was fixed on the surface of Tin-diffused glass waveguide by spin coating method (about 1 cm wide porphyrin film was retained in the middle of the slide). In spin coating method, the first rotating speed of the spin coater was 450 rpm, with a rotating time of 5 s, and the second rotating speeds were 1500, 1800, and 2100 rpm, respectively, with a rotating time of 25 s. The as-fabricated OWG components were dried in vacuum circumstance for 24 h at room temperature to take sensing measurements.

Optical waveguide gas testing system

The Tin-diffused glass OWG detection system (homemade apparatus) was composed of laser, reflector, chamber, Tin-diffused glass OWG gas-sensing element, prism, photomultiplier, PC, and carrier (Fig. 3) [40]. The semiconductor laser beam (λ = 650 nm) was introduced into the waveguide layer using prism coupler [41] (glass prism, n = 1.75, matching liquid of diiodomethane as matching liquid, n = 1.74). When certain wavelength laser passed through the first prism into the Tin-diffused guided layer, light in the guided wave layer appeared in the form of the total reflection of the evanescent wave getting into the sensitive layer [42]. Intensity of the output light was monitored employing a photomultiplier detector and was kept an account of by a computer. In every measurement, a new syringe was used to inject 20 cm3 of gas analytes onto surface of sensing element, and then out from the vent. All measurements were carried out at ambient temperature.

Figure 3
figure 3

Schematic view of optical waveguide gas testing system

Full size image

Testing gas

Specific concentration of toluene, chlorobenzene, and other organic gases (carbon tetrachloride, methanol, ethanol, ammonia, methylamine, dimethylamine and, trimethylamine) was gained through the natural volatilization method injecting corresponding solution into a 600-mL standard vessel; according to the ideal gas equation [43]:

$$ PV = nRT, $$
(1)

where n is the amount of substance of gas; \( T = \left( {273 + 20} \right)K \); R is a constant of about 8.314 J/(mol K); V is volume of the standard vessel, which is equal to Vgas; and P represents saturated vapor pressure of volatile organic compounds (at 20 °C), it is inferred that the amount of volatile organic compound measured in standard vessel remains unchanged after its natural volatilization. In other words, n is certain. And since \( n = \frac{m}{M} \), \( m = \rho V_{\text{liquid}} \), ρ is density of the liquid at 20 °C; M is molecular weight of the volatile organic compound; m stands for quality of volatile organic compound. So:

$$ PV_{\text{gas}} = nRT = \frac{{RT\rho V_{\text{liquid}} }}{M} $$
(2)
$$ V_{\text{liquid}} = \frac{{PV_{\text{gas}} M}}{RT\rho } $$
(3)
$$ C_{\text{gas}} = \frac{n}{{V_{\text{gas}} }} = \frac{m}{{MV_{\text{gas}} }} $$
(4)

And there exists a relation between Cgas and ppm value [44]:

$$ ppm = \frac{{22.4*C_{\text{gas}} }}{M} $$
(5)

Certain amount of copper powder was weighed into the same specification vessel; then, certain amount of concentrated nitric acid was added to prepare specific concentration of NO2 gas; SO2 was obtained through the reaction between a certain amount of sodium sulfite and concentrated hydrochloric acid; H2S was obtained via the reaction between a certain amount of ferrous sulfide and concentrated hydrochloric acid. Different amounts of standard target gases were deliquated with pure air to get desired concentrations in the second 600 mL standard vessel: A certain amount of standard target gas from the first standard vessel is injected into the second standard vessel to obtain lower concentration of gas analytes by syringe and then diluted with pure air. The amount of gas (n) in this dilution process is certain. By this way, very low concentrations of analyte gases were achieved. In order to verify the accuracy of the gas concentration, gas concentration in the standard vessel is detected by gas tube after complete volatilization of the liquid (VOCs) and after complete reaction of reagents (inorganic gases).

Results and discussion

Detecting principle

In an optical waveguide sensor, interactivity between the sensitive film and measured gases would cause slight changes in optical property (transmittance or absorbance) of the sensitive film, which will give rise to great changes of output light intensity that is related to absorbance and thickness of the film [45]:

$$ I = I_{0} \left( {1 - \alpha Nd_{e} } \right) $$
(6)

In this calculation, I is output light intensity, I0 stands for input light intensity, α denotes absorbance coefficient of the sensitive film, N is refractive numbers of the guided light on the surface of OWG with length L, and d e represents the real distance of light propagation, respectively. This relation indicates that when absorbance coefficient of the sensitive layer increases (α), output light intensity will decrease; the intensity of output light will increase when the absorption coefficient of sensitive layer decreases. In this experiment, dry air played roles of carrier and dilution gas. Before testing starts, dry air is introduced into the system and baseline is still stable. When the measured gas enters the chamber, interaction between the measured gas and sensitive elements may occur, which may be chemical reaction or physical absorption or neither. When the interaction occurs between the measured gas and sensitive element, it either absorbs a part of the incident light or reflects the incident light or neither absorbs nor reflects the input light. When the light is absorbed (α increases), the output light intensity decreases, while the reflected light leads to the enhancement of output light intensity (α decreases); when sensing element does not interact with the measured object, the output light intensity remains unchanged (α unchanged), that is, the baseline is stable [46, 47].

UV-visible spectroscopy

Porphyrin film coated on glass slide was exposed to toluene, chlorobenzene and other volatile organic compounds (VOCs) and NO2, H2S, and SO2 gases, and absorbance of the film was tested before and after exposure to gas analytes (as depicted in Fig. 4). These results are given in Table 1. Detection of absorbance was performed by an ultraviolet spectrophotometer spectroscopy. When porphyrin films were exposed to VOCs, there were no palpable changes in absorbance of the film; while porphyrin films were exposed to those three kinds of inorganic gases, the Soret band was red-shifted and four characteristic Q peaks generated to one stronger peak; absorbance of the film was augmented in the range of 610–750 nm. At 650 nm, the changes of film absorbance induced by NO2, H2S, and SO2 gases were 0.0257, 0.0167, 0.0165 nm, respectively. At the same wavelength, maximum absorbance change value was caused by NO2. Besides, when porphyrin membranes were contacted with those three kinds of inorganic gases, the color of the membrane changed from purple to green (as shown in Fig. 3).

Figure 4
figure 4

UV–Vis absorption spectra of film and its change after exposure to certain concentration of several inorganic gases

Full size image
Table 1 Absorbance change of sensitive film after exposure to certain concentration of several inorganic gases
Full size table

The XRD patterns of the film before and after exposure to analytes (NO2, SO2, and H2S) at 2θ values from 5 to 80 are given in Fig. 5. The diffraction peak position of the thin film was basically unchanged after contacting with the measured gases, while the diffraction peak intensity of the film varied after contacting with the measured gases. When exposed to NO2, intensity of the strong diffraction peak of the film suffered from decrease (peak intensity variation value was about 90); exposed to SO2, there was no palpable changes in film XRD patterns; exposed to H2S, there emerged an increase in intensity of the strong diffraction peak of the film and shift to higher 2θ values compared with film. Based on the above changes, we can speculate that the crystallinity of the films changed after contacting with the measured gases. Plus, it was worth mentioning that after the contact with those three kinds of analytes, the bending degree of film’s XRD diffraction peak was more obvious at 40°θ. Combining the UV absorption spectrums and XRD patterns of the thin film before and after exposure to those three kinds of analytes, it could be deduced that change of thin film caused by NO2 was larger than that of the other two.

Figure 5
figure 5

XRD patterns of porphyrin film before and after exposure to analytes

Full size image

Gas exposure

Porphyrin thin-film/Tin-diffused glass OWG elements were immobilized in OWG gas testing system for exposure to VOCs and the three kinds of inorganic gases. Responses of the porphyrin-based glass OWG sensor to 100 ppm VOCs and three kinds of inorganic gases were measured, which are directly related to thickness of the sensitive film, and film thickness has something to do with concentration of solutions and rotating speeds of spin coater [48, 49]. Thus, solution concentrations and rotating speeds were optimized (Fig. 6). Signal change was calculated through the following formula [42]:

$$ \Delta I \left( {{\text{signal}}\,{\text{intensity}}\, {\text{change}}} \right) = I_{\text{gas}} - I_{\text{air}} , $$
(7)

where Igas denotes output light intensity in the gas, while Iair is output light intensity in the surrounding air. As displayed in Fig. 6, when exposed to VOCs, sensing element had exhibited relatively small responses to all VOCs compared with inorganic gases. Plus, among those three kinds of inorganic gases, sensing component had showed much larger response to NO2 than the other two kinds of vapors. The porphyrin sensing elements under different conditions (diverse mass concentrations with diverse rotating speeds) displayed responses to the selected analytes, among which sensing element with mass concentration of 0.07% and rotating speed of 1800 rpm exhibited the best response to the NO2 vapor (Fig. 6b). Signal intensity change value (ΔI) for NO2 gas ran up to 650 (maximum variation value) at room temperature upon that condition (under mass concentration of 0.07% and rotating speed of 1800 rpm). Hence, it could be determined that this condition was the best condition for preparation of sensitive films to achieve better sensitive layer. Then, further measurements all were carried out under the best condition. From Fig. 6, it was not difficult to find out that sensing component showed a little response to VOCs. So in our later work, we had selected five VOCs as analytes to conduct further detection (Fig. 7). In addition, upon exposed to NO2 or H2S or SO2 gas, as-described device showed fast responses to these three inorganic gases with slow recovery times; no matter which gas was in contact with the element, color of the film changed from purple to green, making it difficult to carry out continuous online test for these three gases upon the same one element. The as-described device showed higher response to NO2 compared with H2S and SO2. Furthermore, based on the UV–Vis absorption spectra and XRD patterns of porphyrin film before and after exposure to those three kinds of gases (changes caused by NO2 were more obvious than the other two), we selected NO2 as analyte together with the five VOCs mentioned above to take dynamic measurement in our homemade testing system.

Figure 6
figure 6

Line chart of preparation conditions of the porphyrin-based Tin-diffused film optical waveguide device prepared with various rotating speeds 1500–2100 rpm: a–c where films were developed from porphyrin solution with mass concentration of 0.05, 0.07, 0.09%, respectively

Full size image
Figure 7
figure 7

Dynamic response curve of the porphyrin thin-film Tin-diffused glass OWG sensor to 100 ppm gas analytes

Full size image

In this work, selectivity response of the as-presented OWG element to selected gases was measured by applying a semiconductor laser beam illuminant. As displayed in Fig. 7, in the process of gas measurement, as described above, dry air was functioned as carrier and dilution gas; it was made to transfer the measured gas to the chamber. When air was introduced into the system, baseline was still stable. The response time is perceived as the time required to reach 90% of the equilibrium value following exposure to the measured gas, and recovery time is seen as the time required for sensing element to return to within 10% of the initial response in air when release of the gas [50]. Described sensitive device exhibited fast, strong response to selected objects, especially to the NO2. Upon exposed to the former five kinds of 100 ppm VOCs, device showed fast response and recovery times within a few seconds. While exposed to the same concentration of NO2, device also revealed fast response time (3 s), but slow recovery time (10 min or so, Fig. 7). It can be estimated from Fig. 7, in which signal change of NO2 (ΔI = 650) was five times more than that of other gases. We also could infer that the sensing device revealed good selectivity and high sensitivity to NO2. Intensity of the output light reduced with the existence of the object analyte (NO2) as displayed in Fig. 7. As given in Fig. 4, porphyrin sensitive film was exposed to measured gas (NO2), which resulted in an increase of absorbance (at 650 nm) of the sensitive film. It indicated that interaction between sensitive material and measured gas triggered augment of light absorption (α increased). Therefore, when the light passes through surface of the sensitive element, interaction between sensitive element and NO2 caused absorption of some incident light, which brought decrease of the output light intensity. This result showed in accordance with Eq. (6). Another point worth mentioning is that the optical waveguide gas sensor has the characteristic of fast response. Hence, there may be sudden change in light intensity.

It is necessary to go a step further in verification of porphyrin sensing element’s selectivity to NO2. Hence, two kinds of gas mixtures were prepared: In the first group, there were toluene, chlorobenzene, carbon tetrachloride, methanol, ethanol, H2S, SO2, and NO2; in the second group, there contained all the gases in the first group except for NO2; and concentration of each gas was 100 ppm. As shown in Fig. 8a, b, when in contact with mixed gas containing NO2, as-prepared testing device revealed much bigger output light intensity than that of contact with pure NO2, indicating that the sensitive element also had a certain response to other gases, which caused increase in signal intensity. What’s more, the response of sensitive element to gas mixture containing NO2 was larger than that of gas mixture that did not contain NO2. It could be deduced that as-mentioned sensitive element did have a selective response to NO2.

Figure 8
figure 8

Dynamic response curve of the porphyrin thin-film Tin-diffused glass OWG sensor to gas mixture with NO2 and without NO2

Full size image

In the interest of study correlation between NO2 concentration and response of porphyrin OWG element to NO2, elements were exposed to 100, 10, 1 ppm, 100, and 10 ppb NO2 under the selected best condition, respectively. With the increase of NO2 concentration, output light intensity was reduced, implying concentration dependency performance of sensing device (Fig. 9).

Figure 9
figure 9

Linear relationship diagram of light intensity variation value of OWG sensing element upon exposure to different concentration of NO2

Full size image

The response strength was calculated by Eq. (7). Corresponding response of sensing element was achieved even upon exposure to low concentration of NO2 (10 ppb) (Fig. 9). It is noticeable that there was a good linear relationship between gas concentration and light intensity variation value.

Characterization of the sensitive film

It is well known that efficiency of gas sensor is greatly dependent on the structure of sensing materials [51]. Due to its large specific surface area, high porosity, and excellent accessibility for gaseous analytes, porous structure is highly attractive to sensing materials [52]. Prepared porphyrin film was characterized by FE-SEM, and results are given in Fig. 10. It can be seen from Fig. 10a that surface of the film has presented irregular multi-pore structures, and pore sizes were between 37 and 68 nm, which were in the mesoporous and large pore range. And these disordered hole structures are beneficial to gas adsorption [53, 54], which increases the reaction between gas molecules and sensing films [55]. As displayed in Fig. 10b, upon exposure to NO2 gas, film morphology was different from the former, and multi-hole structures were not very obvious compared with Fig. 10a, emerging many cracks on surface of the film, which was maybe that NO2 gas was absorbed by described disorganized porous structures, altering surface of the film. Then, that might be the reason why the sensing element presented long recovery time.

Figure 10
figure 10

SEM images of film before (a) and after (b) exposure to NO2 gas

Full size image

In order to better describe properties of sensitive membrane (under the best sensing condition mentioned above), thickness of thin film was measured by thin film thickness gauge. Five points were selected on surface of the film before and after exposure to NO2 gas, and average thicknesses of the film before and after exposure were 71.3 and 76.8 nm, respectively (Table 2). After exposure to NO2 gas, there was an increment in film thickness and average thickness variation of membrane was 5.5 nm, which could imply that NO2 gas was taken up by the membrane. This result was well matched with SEM images of the film.

Table 2 Film thickness and its variation after exposure to NO2
Full size table

Contact of porphyrin film with NO2, SO2, H2S resulted in disappearance of characteristic Soret band of porphyrin and emerging of a novel band about 458 nm (Fig. 4), which was in accordance with reports from other researchers applying distinct free-base porphyrins [45, 56,57,58]. These spectral changes were well put down in many documents, and in view of the fact that porphyrins are electron-rich substances and NO2 is a strong oxidizing material, a charge transfer process in the form of oxidation is the most probable mechanism for such behavior [59,60,61], which can be described in Fig. 11. SO2 is also a oxidizing agent, of which oxidizing capacity is lower than that of NO2 [62, 63], so there may be the same mechanism between porphyrin and SO2 [64]. The reaction mechanism can be divided into two procedures. In the first procedure, porphyrin was oxidized by NO2. This brings about changes in the infrared, but not in the visible range; in the second procedure, another NO2 molecule reacts covalently with the reaction product of the first procedure and the second step stimulates the color variation from purple to green. This makes them pleasant campaigners for one-time use of NO2 transducer. And for H2S, there was the same phenomenon when in contact with porphyrin film. Unlike the former two, H2S is a reducing substance [65], and according to some studies, metalloporphyrins can oxidize H2S [66, 67]. However, there are no relevant papers to explain the reaction mechanism between free-base porphyrin and H2S, employing redox reaction. According to interaction mechanism between metalloporphyrins and H2S, we suppose that porphyrin may act as oxidant when in contact with H2S. Further efforts are still needed to explain the mechanisms of the interactions between porphyrin and these gases.

Figure 11
figure 11

Reaction mechanism diagram of porphyrin to NO2

Full size image

Film stability was probed by recording membrane UV–Vis spectra till 20 days from preparation of the membrane (Fig. 12, left). During this period, specimen was kept in the dark to prevent from contacting with pollutant gases. Influence of time on porphyrin film had produced one notable result. Film presented an increase in absorbance intensity, which may be attributed to volatilization of solvent in porphyrin film. Shape of the spectra upon 20 days had showed no changes. As drawn in Fig. 12 (right), the response values of sensing element to NO2 gas were detected after several days of film preparation (2, 5, 12, 20 days after the preparation of thin films), and the RSD value was 1.66%. According to Fig. 12, it could be inferred that the film was fairly stable and had good reproducibility.

Figure 12
figure 12

UV–Vis absorption spectra of porphyrin film (2, 5, 12, 20 days after the preparation of thin films) and response values of sensing element to NO2 (2, 5, 12, 20 days after the preparation of thin films)

Full size image

Conclusion

A free-base porphyrin-based planar OWG sensor has been developed by spin coating method. UV–Vis spectra of the porphyrin solution and its film had been compared. The sensor element exhibited good sensitivity to selected analytes and selectivity to NO2 gas with fast response time and slow recovery time. When in contact with NO2, H2S, and SO2, the color of the porphyrin film changed from purple to light green. SEM image results of the film before and after exposure to NO2 had exhibited an accordance with the film thickness variation. And as-mentioned sensing element could detect very low concentration of NO2 with detection limit of 10 ppb. In order to highlight the selectivity of the sensing element to NO2, the mixed gases with NO2 and without NO2 were tested; the results showed that the response of the sensing element to the mixed gases with NO2 was much larger than that of without NO2. The reaction mechanism was explained in detail by oxidation–reduction reaction. Compared with other sensors, the prepared porphyrin thin-film/Tin-diffused glass optical waveguide element has merits of low cost, simple structure, easy preparation, high sensitivity, and fast response and will pave a path for the exploration of optical sensors.