A highly sensitive oscillating in-situ spectroscopy experimental method and its data processing algorithm

Abstract

The present invention belongs to the technical field of spectroscopy, and particularly relates to a highly sensitive oscillating in-situ spectroscopy experimental method and its data processing algorithm. By means of an external periodic perturbation oscillation source, a stable periodic oscillation response signal is formed within the in-situ experimental system, and further by using Fourier transform means, tiny experimental signals can be captured, and signal response time series information can be obtained. This experimental method and data processing algorithm can eliminate the need for collecting spectroscopic background signals, can more truly reflect the spectroscopic characteristics of the in-situ experimental process, are suitable for capturing tiny spectroscopic signal information, and are suitable for various in-situ spectroscopy experimental processes.

Description

Technical Field

[0001] The invention belongs to the technical field of spectroscopy, and in particular relates to a sensitively resolving oscillating in-situ spectroscopy experimental method and a data processing algorithm thereof. Background Art

[0002] In-situ spectroscopy is an important means of research in chemistry and other related fields recently, especially in the field of catalysis. In-situ spectroscopy experiments are an almost indispensable experimental means to reveal the nature of reactions. Spectroscopy can explain important information of the target research system. For example, infrared spectroscopy can be used to reveal the structure and semi-quantitative information of adsorbed species, ultraviolet-visible spectroscopy can reveal the characteristic structural information of solid materials, and nuclear magnetic resonance spectroscopy can reveal the local neighboring structural information of solid materials, etc. In-situ experiments are essentially based on spectroscopy experiments, and further introduce temperature and gas to get closer to the real research environment, so as to more accurately reveal the spectral characteristics of the target research system under the real environment.

[0003] Specifically, "in situ" corresponds to two different concepts in English. One is insitu, which refers to the real-time collection of spectral information at the reaction position, emphasizing the spatiotemporal synchronization of in situ experiments. The other is operando, which refers to the simulated environment in in situ experiments being close to the real environment. In fact, in many in situ experimental design processes, the above two principles cannot be taken into account. For example, in a common in situ experiment design, the target system is pre-contacted with one of the reactants A until the spectral signal is stable, and then the reactant B is introduced to observe the spectral change characteristics. In this experiment, since A and B are not introduced at the same time, the experimental environment is always rich in component A or rich in component B, and operando cannot be achieved. In addition, in some in situ experimental processes, in order to highlight the operando characteristics, A and B are introduced into the in situ reaction device at the same time. At this time, the spectrum obtained is the spectrum under real reaction conditions, but at this time, the important reaction intermediates are often too active, the concentration is too low, and the signal intensity is low in spectroscopy, making it difficult to capture. Therefore, common in situ experimental designs have certain limitations to a greater or lesser extent.

[0004] On the other hand, in-situ experiments usually require the pre-collection of background spectra for subsequent data processing. However, during in-situ experiments, due to the long duration of the experiment, the analysis target may change slowly over time, such as sample aging and rearrangement, which may distort the background. In particular, when the sample rearranges, the subsequent background is distorted and it is difficult to re-collect the background spectrum.

[0005] In in-situ experiments, there is another type of problem, which is the problem of time-series analysis. Usually, this depends on the need for the highest possible spectral acquisition frequency, but this will lead to a problem of reduced spectral quality. Generally speaking, the signal-to-noise ratio of the spectrum is proportional to the square root of the acquisition time. In order to pursue the time-series analysis of reactive species, it will inevitably lead to a decrease in spectral quality, thus making spectral analysis difficult. Summary of the Invention

[0006] The object of the present invention is to provide a highly sensitive oscillating in-situ spectroscopy experimental method and its data processing algorithm.

[0007] The technical solution of the present invention is as follows:

[0008] A highly sensitive oscillating in-situ spectroscopy experimental method, by introducing a stable excitation oscillation signal source in an in-situ experiment where the spectrum reaches stability, the stable excitation oscillation signal source perturbs the in-situ spectrum with periodic excitation oscillation, tracks and collects the in-situ spectroscopy characteristic information affected by the periodic excitation oscillation perturbation, and records each piece of in-situ spectroscopy information until the collected in-situ spectroscopy information presents stable periodic oscillation.

[0009] Further, the in-situ spectroscopy sampling frequency is higher than the signal generation frequency of the introduced oscillation signal source.

[0010] Further, the stable excitation oscillation signal source is the periodic oscillation of the in-situ cell temperature, gas flow rate or gas composition.

[0011] Further, in the highly sensitive oscillating in-situ spectroscopy experimental method, by changing the excitation oscillation source frequency, signals with different response frequencies are obtained. Further increasing the excitation oscillation source frequency, the signals that cannot be captured in the amplitude spectrum are slow response signals.

[0012] Further, the highly sensitive oscillating in-situ spectroscopy experimental method is applied to in-situ infrared spectroscopy, in-situ Raman spectroscopy, in-situ ultraviolet-visible-near-infrared spectroscopy, in-situ synchrotron radiation spectroscopy and in-situ nuclear magnetic resonance spectroscopy.

[0013] Further, an oscillating in-situ spectroscopy data processing method, by intercepting multiple pieces of stable periodic oscillation spectroscopy data recorded, and performing complex variable function Fourier transform or real variable function Fourier transform on the intercepted multiple data. When performing complex variable function Fourier transform, the zero-th term is taken to represent the in-situ spectrum experiment background, the M-th term is taken to represent the intensity of the spectral change caused by the oscillation signal transformation, and the phase of the M-th term represents the characteristic signal time series caused by the oscillation signal transformation, where M is the number of oscillation periods.

[0014] Further, it specifically includes the following steps:

[0015] S1: Starting from the acquisition of a spectroscopically presented periodic oscillation, continuously record several spectra. Denote each spectrum as f n (x). There are N spectra in total. Each spectrum has several data points, denoted as x 1,n , x 2,n , …, x i,n , …, where x is a spectral characteristic quantity, which can be wavenumber, wavelength, etc., n is the serial number of the recorded spectrum, and x 1,n is the recorded signal value corresponding to the first wavenumber, wavelength, etc. of the nth spectrum, and f n (x) represents the nth spectrum;

[0016] S2: Perform a discrete Fourier transform on all spectral data. Then, for each data point, there is:

[0017]

[0018] Each combination forms a sequence F k (X). For each corresponding term number k, there are several data points, denoted as X 1,k , X 2,k , …, X i,k , …, where j is the imaginary unit, and X 1,k is the kth term after performing the Fourier transform on the ith signal point in the spectrum for N discrete time series points;

[0019] S3: Extract the term with k = 0 from the data in S2 to form a spectrum, denoted as F0(X). The F0(X) has multiple data points, and the multiple data points are X 1,0 , X 2,0 , …, X i,0 , …, and each term is respectively

[0020]

[0021] Then F0(X) / N is a spectrum obtained after a data transformation, and this spectrum is the background curve in the in-situ spectroscopy experiment;

[0022] S4: Extract the term with k = M from S2 to form a spectrum, denoted as F M (X). The F M (X) has multiple data points, which are X 1,M , X 2,M , …, X i,M , …, and each term is respectively

[0023]

[0024] The amplitudes of S2 form a spectrum, denoted as A M (X). The A M(X) has multiple data points, which are A 1,M , A 2,M , …, A i,M , …, each item is respectively

[0025]

[0026] The phases of the said S2 form a spectrogram, denoted as argm(X), and the argm(X) has multiple data points, which are arg 1,M , arg 2,M , …, arg i,M , …, each item is respectively

[0027]

[0028] Among them, each point A in the amplitude spectrogram i,M represents the response magnitude degree of the i-th signal point in the spectrogram to the excitation oscillation signal, and each point arg in the phase spectrogram i,M represents the response delay degree of the i-th signal point in the spectrogram to the excitation oscillation signal.

[0029] Furthermore, in the said oscillating in-situ spectroscopy data processing method, by changing the frequency of the excitation oscillation source, signals with different response frequencies are obtained. Further increasing the frequency of the excitation oscillation source, the signals that cannot be captured in the amplitude spectrum are slow response signals.

[0030] Compared with the prior art, the beneficial effects of the present invention are mainly reflected in:

[0031] 1. It truly realizes the in-situ experimental means of in_situ (real-time acquisition) and operando (close to the real environment). The spectroscopic properties shown by the experimental method provided by the present invention are close to the real application scenario, and there will be no large deviation of a certain experimental variable.

[0032] 2. The experiment does not need to pre-collect the background spectrogram, and the real background spectrogram can be naturally obtained by using the algorithm provided by the present invention.

[0033] 3. It can effectively reflect the small changes in the spectrogram superimposed on a strong background and capture the signal characteristics of the real changes. Description of the Drawings

[0034] Figure 1 is the flow chart of the present invention;

[0035] Figure 2 are the background spectrogram and amplitude spectrogram obtained from the data in the embodiment of the present invention;

[0036] Figure 3 is the phase spectrogram obtained from the data in the embodiment of the present invention. Detailed Embodiments

[0037] The following will describe in more detail a highly sensitive oscillating in-situ spectroscopy experimental method and its data processing algorithm of the present invention in conjunction with schematic diagrams, in which the preferred embodiments of the present invention are shown. It should be understood that those skilled in the art can modify the present invention described herein while still achieving the advantageous effects of the present invention. Therefore, the following description should be understood as a broad guidance for those skilled in the art and not as a limitation on the present invention.

[0038] In order to solve the problem that existing in-situ spectroscopy methods cannot simultaneously take into account insitu (i.e., real-time acquisition) and operando (experimental conditions close to the real environment), and at the same time solve the problem of difficult background spectrum acquisition in the real environment and the overly weak response of tiny transient processes being hidden in the background signal, a new in-situ spectroscopy experimental method and its data processing algorithm are developed.

[0039] The principle is outlined here. In an in-situ experiment, under an external condition, it will gradually tend to be stable, that is, reach an equilibrium state, and thus obtain the spectral signal in this equilibrium state. If a small periodic excitation oscillation perturbation is applied to this external condition, the collected spectrum will also generate a dynamic following perturbation in the original equilibrium state and gradually reach a dynamic periodic stable perturbation. At this time, if the spectrum is analyzed using Fourier transform, the zero-order term, that is, the signal that does not generate a perturbation with the change of the excitation source, should be the original equilibrium state signal, which is equivalent to the background in this in-situ experiment; and in Fourier transform, the signal with the same frequency as the excitation source perturbation frequency, that is, the information sensitive to the change of the excitation source.

[0040] For the in-phase excited signal, it can be split into two main information: amplitude and phase. If the target signal is greatly affected by the excitation signal, it is easy to form a strong amplitude signal; if the target signal is not related to the excitation signal, the amplitude is small. Therefore, the degree of excitation of the excitation signal can be obtained by analyzing the amplitude spectrum of the excited signal. For example, consider an in-situ infrared spectroscopy experiment, and consider simulating a real gas-solid catalytic reaction process. The in-situ cell is filled with a solid catalyst and continuously supplied with reaction gases A and B, so as to in-situ simulate the adsorption and reaction process of A and B on the catalyst. At this time, taking the concentration of gas B as the excitation signal, that is, modulating and fluctuating the B concentration periodically above and below the target real concentration, then during the in-situ reaction process, the monitored infrared spectrum will also show a periodic response. According to the algorithm provided by the present invention, the amplitude spectrum can be further obtained. Obviously, the adsorbed species corresponding to gas B has a strong signal in the amplitude spectrum, indicating that the fluctuation of the B gas concentration will significantly affect the concentration of its corresponding adsorbed species. In addition, an intermediate species C may have a strong signal in the amplitude spectrum, while an intermediate species D may have a weak signal in the amplitude spectrum, indicating that the concentration of C strongly depends on the concentration of B in the gas phase, while the concentration of D is hardly affected by the concentration of B.

[0041] The in - frequency stimulated signal can also obtain phase - spectrum information. Due to the influence of diffusion, adsorption - desorption, and reaction in in - situ experiments, different target substances actually appear in an obvious chronological order, which is manifested as an obvious time difference in the phase spectrum. Still taking the in - situ infrared experiment mentioned in the previous paragraph as an example, assuming that the reaction between A and B is a multi - step reaction, first an intermediate compound C is formed, and finally the product E is formed. Then, when the concentration of B is used as the excitation signal to form fluctuations, it can be observed that the phase of the infrared band corresponding to B is ahead, C follows closely, and the phase of E is the last. Thus, using the algorithm of the present invention, the time sequence of the sequential reaction can be clearly revealed.

[0042] Finally, a brief description of the background spectrum information is given. The zero - order term obtained by this algorithm is essentially the average value of several spectra. Since the stimulated signal oscillates periodically, after taking the average value, the oscillations cancel each other out, and what is obtained is the background spectrum. For the above example, what the background spectrum actually shows is the spectroscopic information of the equilibrium adsorption state on the catalyst in this in - situ experimental environment. If the background spectrum is compared with the phase spectrum, it can be known which adsorbed species actually react under the change of the concentration of B, that is, the species active with respect to B, and it can also be known which adsorbed species do not actually react under the change of the concentration of B, that is, the species that are only an inert adsorbed species with respect to B. From the perspective of a simple catalytic reaction, the inert adsorbed species only occupy the adsorption sites and may not contribute to the activity, thus more valuable information may be revealed.

[0043] The present invention will be specifically described below through specific embodiments.

[0044] Consider the in - situ infrared experimental process of a certain actual catalyst used for selective catalytic reduction (SCR) denitration. Here, in - situ diffuse reflectance spectroscopy is used to collect the in - situ infrared spectra of the catalyst at a certain temperature, a certain pressure, a certain NO concentration, a certain NH3 concentration, and a certain O2 concentration. And the experiment is carried out and data collection and calculation analysis are carried out according to the flow chart as Figure 1 shown.

[0045] Step 1. First, continuously pass 500 ppm NH3, 500 ppm NO, and 2 vol.% O2 into the catalyst, and maintain it at 120 °C, and continuously collect infrared spectra until the spectra are stable. Subsequently, the gas is switched back and forth between two atmospheres of 500 ppm NH3 + N2 and 500 ppm NO + 2 vol.% O2 + N2, with a switch every 5 min and a switching period of 10 min. At this time, continuous spectral collection is carried out, the spectral collection interval is 10 s, and the real - time spectral line at 1450 cm -1The signal intensity at that location is measured until the infrared absorbance signal intensity at that location reaches a stable periodic oscillation. At this time, spectra are collected continuously for 5 cycles, that is, 50 minutes, and all spectra f n (x) within the 5 cycles are completely recorded and saved. Since the spectral acquisition frequency is 0.1 Hz, within the five cycles, n = 0, 1, …, 300. Each spectrum f n (x) contains 1738 data points, corresponding to wavenumbers from 4000 cm -1 to 650 cm -1 for a total of 1738 evenly distributed wavenumber data points. That is, for the nth spectrum f n (x), the index i of the data point x i,n has 1738 possible values.

[0046] Step 2. At this time, Fourier transform is performed on the data. 300 spectra are taken out, and for each wavenumber, the corresponding data are taken out in sequence to form a sequence of 300 elements. For example, all the values corresponding to the wavenumber of 4000 cm -1 are taken out to obtain the sequence {x4000 cm -1 ,n}. Fourier transform is performed on this sequence to obtain

[0047]

[0048] Similarly, Fourier transform is performed on the data points corresponding to all wavenumbers in sequence to obtain the following transformation results.

[0049]

[0050] The obtained X i,k is arranged according to the wavenumber to obtain a spectroscopic sequence F k (X). Here, for each wavenumber, for example, for the wavenumber of 4000 cm -1 , the index k in X i,k should range from 0 to 300. Each wavenumber has an X i,k sequence, and a total of 1738 X i,k sequences form the spectroscopic sequence F k (X).

[0051] Step 3. For each sequence point corresponding to each wavenumber in F k (X), the term with k = 0 is taken out, and thus a background spectrum F0(X) can be obtained from 1738 points:

[0052]

[0053] F0(X) is a one-dimensional array composed of 1738 data points, and each data point corresponds to the corresponding wavenumber. Essentially, it is an infrared spectrum.

[0054] Step 4. According to the experimental operation description, each excitation period is 10 min, and the sampling frequency is 0.1 Hz, which is equivalent to the number of samples per period M = 60. Therefore, from each sequence point corresponding to each wavenumber in F k (X), k = M terms are taken, so as to obtain a complex sequence F M (X) consisting of 1738 points:

[0055]

[0056] For each complex number in this complex sequence F M (X), its amplitude and phase are separately taken out, and an amplitude spectrum A M (X) and a phase spectrum argm(X) are formed:

[0057]

[0058] At this time, both the amplitude spectrum A M (X) and the phase spectrum argm(X) are one-dimensional arrays composed of 1738 real numbers. Essentially, they are equivalent to two spectral diagrams corresponding to the corresponding wavenumbers. The former has the same data dimension as the original infrared spectral data, and the phase spectrum data is dimensionless and should take values between -π and +π.

[0059] For the convenience of further elaborating on the data processing results, the background spectrum and the amplitude spectrum obtained from the above data are plotted in Figure 2 and the phase spectrum is plotted in Figure 3 for a brief schematic description.

[0060] In Figure 2 , it can be seen that in an atmosphere where the catalyst coexists with NH3, NO, and O2, two adsorption forms of NH3 can be observed. One is the form adsorbed on the Lewis acid sites, denoted as the L signal, and the other is the form adsorbed on the acid sites, denoted as the B signal. In the background spectrum, in contrast, the B signal is stronger than the L signal. However, in the phase spectrum, it can be observed that the L signal is stronger than the B signal, which is opposite to the background spectrum. This means that, in fact, when the adsorbed NH3 participates in the reaction, the reaction rates of the two forms are not exactly corresponding to their adsorption amounts. Although the adsorption amount of NH3 in the L form is less than that in the B form, its reaction activity is higher. This may provide important information for the interpretation of the active sites of this catalytic reaction, and this information is difficult to obtain in conventional in-situ infrared experiments.

[0061] In Figure 3It can be seen that the phase of the B signal is larger than that of the L signal. Considering the duration of one period, this phase difference approximately corresponds to a time interval of 20 s. Therefore, this means that in the oscillation experiment, the change response of NH3 in the B form is about 20 s earlier than that in the L form. For example, in this catalytic field, it is generally believed that NH3 can evolve between the L form and the B form on the catalyst surface. One view is that the L form is the truly active reaction intermediate species, while the B form is the reservoir of the L form. Therefore, the consumption of NH3 in the B form first may mean that once the L form participates in the reaction and is consumed, the B form replenishes it. As a result, the concentration of the L form reaches dynamic stability, and only the consumption of the B form can be observed spectroscopically. Therefore, the evidence of this phase spectrum may support the previous hypothesis. This information is difficult to capture in common in-situ experiments, especially the information on the signal transformation order with a 20 s time difference. Affected by noise and the capabilities of instrument hardware in experiments, it is extremely difficult to capture.

[0062] The above are only the preferred embodiments of the present invention and do not impose any limitation on the present invention. Any person skilled in the art, without departing from the scope of the technical solution of the present invention, makes any form of equivalent replacement or modification and other changes to the technical solution and technical content disclosed by the present invention, all of which fall within the content of the technical solution of the present invention and still belong to the protection scope of the present invention.

Claims

1. A highly sensitive oscillating in-situ spectroscopy experimental method, characterized in that, In the in-situ experiment where the spectrum reaches stability, a stable excitation oscillation signal source is introduced. The stable excitation oscillation signal source perturbs the in-situ spectrum with periodic excitation oscillation, tracks and collects the in-situ spectroscopic characteristic information subjected to the periodic excitation oscillation perturbation, and records each piece of in-situ spectroscopic information until the collected in-situ spectroscopic information presents stable periodic oscillation; The stable excitation oscillation signal source is specifically as follows: First, continuously introduce 500 ppm NH3, 500 ppm NO, and 2 vol.% O2 into the catalyst, and maintain it at 120 °C, continuously collect infrared spectra until the spectra are stable; Subsequently, switch the gas back and forth between two atmospheres of 500 ppm NH3 + N2 and 500 ppm NO + 2 vol.% O2 + N2, switch every 5 min, and the switching period is 10 min.

2. The highly sensitive oscillating in-situ spectroscopy experimental method according to claim 1, characterized in that The sampling frequency of the in-situ spectroscopy is higher than the signal generation frequency of the introduced oscillation signal source.

3. The method for performing a resolution-sensitive oscillating in-situ spectroscopy experiment according to claim 1, wherein The stable excitation oscillation signal source is the periodic oscillation of the in-situ cell temperature, gas flow rate, or gas components.

4. The highly sensitive oscillating in-situ spectroscopy experimental method according to claim 1, characterized in that By changing the frequency of the excitation oscillation source, signals with different response frequencies are obtained. Further increasing the frequency of the excitation oscillation source, the signals that cannot be captured in the amplitude spectrum are slow response signals.

5. The method for performing a highly sensitive oscillating in-situ spectroscopy experiment according to claim 1, wherein The oscillation in-situ spectroscopy experimental method with high resolution and sensitivity is applied to in-situ infrared spectroscopy, in-situ Raman spectroscopy, in-situ ultraviolet-visible-near-infrared spectroscopy, in-situ synchrotron radiation spectroscopy, and in-situ nuclear magnetic resonance spectroscopy.

6. A method for processing oscillating in-situ spectroscopy data, based on the oscillating in-situ spectroscopy experiment method with high resolution and sensitivity described in any one of claims 1-5, characterized in that, Intercept multiple stable periodic oscillation spectroscopic data records, and perform complex variable function Fourier transform or real variable function Fourier transform on the intercepted multiple data. When performing complex variable function Fourier transform, take its zero term to represent the in-situ spectrum experiment background, take its Mth term to represent the intensity of the spectroscopic change caused by the oscillation signal transformation, and take the phase of its Mth term to represent the characteristic signal time series caused by the oscillation signal transformation, where M is the number of oscillation periods.

7. The oscillating in-situ spectroscopy data processing method according to claim 6, wherein Specifically, it includes the following steps: S1: Starting from the acquisition of a spectrogram showing periodic oscillations, continuously record several spectrograms. Denote each spectrogram as f n (x). There are N spectrograms in total. Each spectrogram contains a number of data points, denoted as x 1,n , x 2,n , …, x i,n , …, where x is a spectral characteristic quantity, which can be wavenumber, wavelength, etc., n is the serial number of the recorded spectrogram, and x 1,n is the recorded signal value corresponding to the first wavenumber, wavelength, etc. of the nth spectrogram, and f n (x) represents the nth spectrogram; S2: Perform discrete Fourier transform on all spectral data, then for each data point, there is: Each combination forms a sequence F k (X), for each corresponding number of terms k, there are several data points, denoted as X 1,k , X 2,k , …, X i,k , …, where j is the imaginary unit, X 1,k is the k-th term after performing the Fourier transform on the i-th signal point in the spectrogram for N discrete time series points; S3: Take out k = 0 items from the S2 data to form a graph, denoted as F0(X). The F0(X) has multiple data points, and the multiple data points are X 1,0 , X 2,0 , …, X i,0 , …, and each item is respectively Then F0(X) / N is a spectrum obtained after data transformation, and this spectrum is the background curve in the in-situ spectroscopy experiment; S4: Take out k = M items from S2 to form a spectrogram, denoted as F M (X), where the F M (X) has multiple data points, which are X 1,M , X 2,M , …, X i,M , …, and each item is respectively The amplitudes of the said S2 form a spectrum, denoted as A M (X), and the said A M (X) has multiple data points, being A 1,M , A 2,M , …, A i,M , …, and each term is respectively The phases of the S2 form a spectrum, denoted as argm(X), and the argm(X) has multiple data points, namely arg 1,M , arg 2,M , …, arg i,M , …, and each term is respectively Among them, each point A in the amplitude spectrum diagram i,M represents the response magnitude of the i-th signal point in the spectrum diagram to the excitation oscillation signal, and each point arg in the phase spectrum diagram i,M represents the response delay degree of the i-th signal point in the spectrum diagram to the excitation oscillation signal.

8. The oscillating in-situ spectroscopy data processing method according to claim 7, wherein By changing the frequency of the excitation oscillation source, signals with different response frequencies are obtained. Further increasing the frequency of the excitation oscillation source, the signals that cannot be captured in the amplitude spectrum are slow response signals.

Images