Preparation and performance of sensing material for diethyl ether: Optimization of working conditions by response surface methodology

Pd doped ZnNi3Al2O7 composite was synthesized by sol-gel and impregnation method. The EDS spectrum and TEM image showed that Pd atoms are uniformly distributed on the surface of ZnNi3Al2O7 with a particle size of less than 50 nm. A sensitive diethyl ether gas sensor based on cataluminescence on the composite at lower temperature than 200°C was reported. There is a good linear relationship between the luminescence intensity and the concentration of diethyl ether in the range of 0.1-75 mg/m. The detection limit (3σ) is 0.06 mg/m. The working conditions optimized by response surface methodology were analytical wavelength of 548.86 nm, reaction temperature of 109.18°C and carrier gas velocity of 125.88 ml/min, respectively. The sensitivity of the method can be increased by 4.5% under the op timized working conditions. The optimization method is universal for many multi parameter processes.


Introduction
Diethyl ether is a colorless and flammable liquid, which is very volatile and can be miscible with ethanol, acetone, benzene and chloroform. The mixture of gaseous diethyl ether and 10 times volume of oxygen can explode violently in case of fire or electric spark [1,2].
Diethyl ether in the air not only has safety problems, but also affects human health. Large concentration of diethyl ether may cause the person in contact with it to be excited, then sleepy, vomit, pale, slow pulse, hypothermia and irregular breath, even life -threatening. Most of the fatal cases involving diethyl ether reported in the second half of the 19th century and the first half of the 20th century were related to anesthesia, as diethyl ether was widely used as an anesthetic in many countries during this period [3]. When people inhale low concentration diethyl ether for a long time, they may have headache, dizziness, tiredness, drowsiness, proteinuria and erythrocytosis. Therefore, it is necessary to determine the content of diethyl ether in air simply and quickly.
Diethyl ether in air is usually determined by gas chromatography with direct injection.
Because this method requires a gas chromatograph , it cannot be done on the spot. Gas sensors are especially suitable for on-site, online or remote measurement. Up to now, there are few reports about other sensing technologies of diethyl ether besides cataluminescence (CTL). CTL, a kind of chemiluminescence (CL) emitted from heterogeneous catalytic oxidation reactions on gas-solid interface, has been considered as a promising energy transduction mechanism for fabricating gas sensors [4]. In recent years, a series of CTL sensing applications have been attempted to develop for diethyl ether [5][6][7][8][9][10][11][12][13][14] and many other molecules [15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31] at different laboratories. When using the composites to make diethyl ether gas sensor, high working temperature becomes a main obstacle. Generally, sensors need to be as small or even miniaturized as possible in order to be portable, which determines that they must be a small heat capacity system. We know that it is very easy to control the temperature accurately in the laboratory. However, it is very difficult to maintain the high temperature accuracy and stability in a small heat capacity system, which is totally different from the situation in the la boratory equipment. Our previous work has proposed that CTL baseline instability will occur in a small heat capacity system when the working temperature exceeds 200 ℃ [4]. In the past, most of the working temperature of the CTL method was more than 200℃ . So their sensitive materials are not suitable for making CTL gas sensors. Pd doped composites often exhibit low temperature catalytic oxidation activities for many molecules [32][33][34][35][36]. Our team has found that Pd/ZnNi3Al2O7 has low-temperature CTL activity for diethyl ether [37]. It may be a good attempt to fabricate gas sensor by using Pd/ZnNi3Al2O7 as sensitive material.
In the past, the working conditions of CTL method were simply determined by single factor experiment [4 -31]. We know that the premise of sing le factor experiment is to assume that there is no interaction between various factors. This assumption is beneficial to the completion of the experiment, but its defects are very obvious. So far, there is no research report about the interaction of differ ent factors on CTL signal and about using statistical model to obtain the optimal experimental conditions. Response surface method (RSM) is a statistical comprehensive test technique [38][39][40][41]. It can be used to analyze the regression relationship between test indexes (dependent variables) and multiple test factors (independent variables). It takes the response of the system as a function of many factors, and synthesizes the finite deterministic "trials" into a response surface to simulate the real state surf ace. In this way, we can use intuitive observation to select the optimal experimental conditions in the experimental design.
In this work, firstly, Pt doped ZnNi3Al2O7 was synthesized and characterized . Then, the experimental conditions were optimized by RSM, and the low temperature CTL properties of the composite were studied. Finally, a feasible method was established for determining diethyl ether by utilizing CTL at lower temperature than 200℃ .

Chemical reagents and apparatus
Zinc acetate, nickel chloride, aluminum nitrate, hydrochloric acid, malic acid, ammonia, palladium chloride and hydrazine hydrate were purchased from Beijing Chemical Regent Co., LTD. (Beijing, China). Various standard gases of ether, formaldehyde, acetaldehyde, ammonia, benzene, ethanol, sulfur dioxide, hydrogen sulfide, carbon monoxide and carbon dioxide in nitrogen were purchased from Beijing Ya -nan Gas Co., LTD. (Beijing, China). Distilled water was used throughout the whole experiment. The micro area composition and particle morphology of the nano-Pd/ZnNi3Al2O7 were investigated by scanning electron microscope (SEM, JEOL-IT500) and transmission electron microscopy (TEM, JEOL -2100) respectively.
The CTL intensities were recorded by ultraweak luminescence analyzer manufactured at the Biophysics Institute of Chinese Academy of Science (Beijing, China).

Synthesis of nano-Pd/ZnNi3Al2O7
Zinc acetate, nickel chloride and aluminum nitrate were dissolved in dilute hydrochloric acid. The solution is oscillated by u ltrasonic to clear. The malic acid was added into the solution in stirring state, and sol was formed by continuously stirring the solution more than 6 h. The pH value of the solution was adjusted to 5.4 with dilute ammonia water. The solution was stirred for 6 hours and aged for 12 hours. Then the gel was prepared by rotating evaporation. After drying and grinding, the gel is placed in a box type resistance furnace. The temperature is increased to 420 ℃ at the rate of 4 ℃/min and maintained the temperature for 5 hours. The composite powder of ZnO, NiO and Al2O3 were obtained by natural cooling.
Palladium chloride was dissolved in 10% hydrochloric acid aqueous solution, and the composite powder dispersed by ultrasonic wave was added into the solution under stirring. 4 Then, 25% hydrazine hydrate aqueous solution was continuously dripped into the solution under stirring. The solution was subjected to aging, filtering, washing, drying and calcining at 280℃ in vacuum oven, successively, to finally get nano -Pd/ZnNi3Al2O7 that the content of P d can be adjusted by changing the concentration of palladium chloride.

Measurement of CTL signals
The loading of nano-Pd/ZnNi3Al2O7, the adjustment of working temperature, the introduction of measured gas sample and the recordin g of CTL signals can refer to our previous work [29][30][31].

Experimental design for RSM optimization
According to the single factor experimental results of analysis wavelength, reaction temperature and carrier gas velocity, the Box-Behnken central composite experiment was designed, and the CTL experimental conditions of diethyl ether were optimized by RSM at three factors and three levels. The EDS spectrum in Fig. 1 shows that prepared sensitive material is Pd/ZnNi3Al2O7, because the atomic ratio of O , Ni, Al and Zn is close to 7:3:2:1. It is found that the content of Pd atom in sensitive materials has a great influence on the ir CTL activity. When the atomic percentage of Pd is less than 0. 8%, the sensitive material has CTL activity to diethyl ether only 5 when the temperature is above 200 ℃ . When the atomic percentage of Pd is more than 1.3%, in addition to diethyl ether, formaldehyde, acetaldehyde, sulfur dioxide , hydrogen sulfide and ammonia also have obvious CTL signals.

Characterization of sensitive materials
The EDS element mapping in the illustration of Figure 1 sho ws that oxygen, Nickel, aluminum, zinc and palladium were uniformly distributed in the prepared sensitive material. The TEM image in Figure 2a shows that the size of Pd/ZnNi3Al2O7 is not more than 50 nm.

Optimization of experimental conditi ons by single factor test
The effects of analysis wavelength, reaction temperature and carrier gas velocity on the CTL intensity of 5 mg/m 3 diethyl ether on sensitive materials were separately studied. Fig. 2b is a CTL spectrum of diethyl ether on nano-Pd/ZnNi3Al2O7 at 110℃ with a carrier gas velocity of 120ml/min. The CTL intensities were high at 550 nm. Fig. 2c is the temperature dependence of the CTL intensity of diethyl ether at a wavelength of 550 nm under a carrier gas velocity of 120 ml/min. As you can see, the temperature of around 110℃ will be favorable to the determination of diethyl ether. Fig. 2d is the carrier gas velocity dependence of the CTL intensity of diethyl ether at 550 nm and 110℃ . The CTL intensity increased with an increase in the carrier gas velocity below 120 ml/min, and it decreased above this velocity. 6 It can be seen that the operating conditions of simple optimization through single factor experiments are analysis wavelength 550 nm, working temperature 110℃ and carrier gas velocity 120 ml/min.

Optimization of experimental conditions by RSM
The results of single factor test are summarized. The analysis wavelength, reaction temperature, and carrier gas velocity were used as research factors. The CTL intensity was used as the response value. Three factors and three levels (see table 1) were used to carry out the Box-Behnken test design. The results of 17 response surface design trials (12 edge points plus 5 center points in Box-Behnken test design) for 5 mg/m 3 diethyl ether are shown in Table 2. The 3D surfaces and contours are plotted by Design Expert software, as shown from  reaction temperature on the CTL intensity, respectively. With the increase of the analysis wavelength and reaction temperature, the CTL intensity increases. When the analysis wavelength reaches 548.86 nm and the reaction temperature reaches 1 09.18 ℃ , the CTL intensity reaches its maximum. When the analy sis wavelength and reaction temperature continue to increase, the CTL intensity begins to decrease.    impossible to obtain the optimal operating conditions through single factor experiments, because the number of experiments needed is very large. According to the model, the maximum value of CTL intensity was 744. This is 4.5% higher than the CTL intensity under single factor optimization conditions, which will correspondingly improve the sensitivity of the method. To investigate the lifetime of the sensitive materials, an experiment was carried out by 9 continually introducing 20 mg/m 3 diethyl ether in air with the velocity of 125.88 ml/min to the surface of sensitive materials at 109.18°C, and detecting the CTL intensities once every hour at 548.86 nm. Partially recording signals are shown in Fig. 6. The results showed that the relative standard deviation (RSD) of C TL intensities was less than 2.5% for continuous 500 h detection. Further experiments showed that the RSD of the CTL intensities was within 5% for daily use above 10 months. It can be said that nano-Pd/ZnNi3Al2O7 has a long service life for diethyl ether monitoring. The results were shown in Fig. 7.  Fig. 7, where y is the CTL intensity, x is the concentration of diethyl ether and R is the correlation coefficient.

Possible luminescence mechanism
We have made a preliminary study on the mechanism of this luminescence reaction. The products of catalytic oxidation were analyzed by GC-MS after molecules in air was introduced.
When the surface temperature of catalyst is about 110 ℃ , only when there is diethyl ether in the sample gas could the tail gas change, that is to say, a small amount of acetic acid molecule can be detected. There is no change in the other m olecules in the sample gas. However, when the catalyst surface temperature is raised to about 1 60 ℃ , except for diethyl ether, both acetic acid and carbon dioxide can be detected. At the same time, many molecules in the sample gas can change at this temperature. For example, a small amount of formaldehyde molecules in the sample gas can be converted into formic acid, few of sulfur dioxide molecules can be converted into sulfur trioxide, trace of acetaldehyde molecules can be converted into acetic acid, etc.
The possible luminescence mechanism is that the active sites produced by the selected catalytic sensitive materials under specific experimental conditions have a good selective adsorption to diethyl ether molecules, and the surface adsorbed oxygen also has a strong reaction activity with it, so diethyl ether is oxidized to acetic acid. The light quantum with ν1 wavelength can be emitted when the acetic acid molecule on the excited state returns to the ground state. Few acetic acid molecules without leaving the catalyst surface are further oxidized by adsorbed oxygen to carbon dioxide, and the light quantum with ν2 wavelength can be emitted when the carbon dioxide molecule on the excited state returns to the ground state.
The latter reaction is competitive one at high temperature. The reaction mechanism can be expressed by the following formula:

Conclusions
A sensitive diethyl ether gas sensor based on Pd activated ZnNi3Al2O7 at 109.18°C was demonstrated. The detection range is 0.1-75 mg/m 3 , and the detection limit is 0.06 mg/m 3 .
The atomic percentage of 0.8-1.3% Pd in nano-composites was beneficial to low operating temperature and high selectivity f or the chemiluminescence of diethyl ether. The sensitivity of the method can be increased by 4.5% after optimizing the experimental conditions by response surface methodology.