"Diagnosis" of domestic near infrared spectroscopy instrument and its application development

Professor Chen Bin's team of Jiangsu University has been engaged in the application of near infrared spectroscopy in the rapid detection of food and agricultural products since 1996. Recently, the editor of instrument information network interviewed Professor Chen Bin to discuss the development status of domestic near infrared spectroscopy instruments.

Near infrared spectroscopy instrument has gone through nearly 50 years of development. In this 50 years, great changes have taken place in the instrument design, performance and measurement methods. In recent years, the digitization of near infrared spectroscopy instruments has been continuously improved. Equipped with powerful computer and chemometric analysis software, it has been more widely used. Especially in developed countries such as Europe and America, near infrared spectroscopy instrument has been regarded as an essential instrument in quality management experiment.

However, in terms of instrument hardware, although China has made some achievements, there is still a considerable gap compared with the international advanced level. For example, the main core components of the instrument still rely on imports, the accessories for measuring different types of samples are not perfect, and the overall performance index and intelligent level of the spectrometer need to be improved.

Professor Chen Bin's team of Jiangsu University has been engaged in the application of near infrared spectroscopy in the rapid detection of food and agricultural products since 1996. Recently, the editor of the instrument information network interviewed Professor Chen Bin to discuss the current situation of the development of domestic near infrared spectroscopy instruments. Professor Chen Bin spoke frankly about the gap between domestic and imported instruments, the "bottleneck" in the development of domestic instruments, and even the future development direction.

Chen Bin, Professor of Jiangsu University

There is still a gap between domestic and imported NIRS

Since the 1990s, some near infrared spectroscopy instruments have been produced in China. After nearly 30 years of development, there are more and more domestic manufacturers of near-infrared spectrometers, such as concentrating technology, Beifen Rayleigh, Huaxia scientific innovation, Shanghai Lingguang, Nanjing Zhongdi, Kaiyuan Shengshi, Changguang Sibo, etc. However, Chen Bin believes that "the near infrared spectroscopy instruments produced by these enterprises have not reached the level of stable mass production. The main reason is not that they do not have the capacity of mass production, but that all production enterprises do not have the sales volume of mass production."

"The stability, especially the long-term stability, of domestic NIRS instruments is still far behind that of those of foreign well-known companies; The signal-to-noise ratio is also unsatisfactory; In addition, there is an urgent need to strengthen the research and development of near infrared spectroscopy instruments in standardization, model transfer, Internet of things, technical standards, application demonstration and other key theories or technologies. These basic requirements in NIR analysis restrict the application of domestic instruments in practical production Chen Bin said.

As for the gap between domestic and imported NIRS instruments, Chen Bin thinks that there are mainly the following aspects:

(1) Domestic NIRS instruments have poor long-term stability, poor SNR accuracy, few types of measurement accessories, single function, and unsatisfactory humanization of software, so research institutes and other application units are unwilling to accept domestic NIRS instruments;

(2) The domestic near-infrared spectroscopy instrument basically adopts the traditional manufacturing process of general spectroscopy instrument, and lacks a certain selection and matching detection platform of optical and electronic components. These factors accumulate to the whole machine, resulting in poor reliability and signal-to-noise ratio, which can not meet some basic requirements of near-infrared spectroscopy analysis;

(3) The application industry does not have enough understanding of NIR analysis technology, the instrument enterprises do not have the ability to establish calibration model according to the user's requirements, the instrument users lack basic knowledge, and do not have the ability of secondary development of the model. In addition, the salesmen of import instrument companies are often application engineers, who have the ability to make plans before sales, guide the establishment of calibration models during sales, and be responsible for model maintenance and instrument maintenance after sales; The sales personnel of domestic instrument companies are out of touch with the application, which restrict the expansion of domestic NIR instrument market share.

The enthusiasm of production is not high, and the formulation of standards is seriously lagging behind

In 2012, Xiangshan Science conference entitled "key technical problems, application and development strategy of near infrared spectroscopy analysis in China" was held in China. It can be seen that the state attaches great importance to the development of near infrared spectroscopy instruments. Therefore, it is necessary to find out the bottlenecks in the development of domestic NIRS instruments; For these "bottlenecks", how can we quickly improve the level and scale of domestic NIRS instruments and applications?

As for the "bottleneck" in the development of domestic NIRS instruments and its solutions, Chen Bin thinks that:

(1) Most researchers and users don't understand it well

The research work of near infrared spectroscopy in China has a history of nearly 30 years. However, because most of them are in the stage of laboratory research, there are still many researchers who do not fully understand their application ability and fail to give full play to their characteristics; Emphasis on improving the prediction accuracy of algorithm research, light detection method to reduce the error of research. The general users don't understand the practical application of NIRS thoroughly, and they don't have the ability of secondary development and application. These two reasons make the domestic near infrared spectroscopy analysis technology can not be widely applied to the actual production.

(2) No model, no application

A stable and reliable calibration model is the key to the application of NIR analysis technology to production practice. The basis of model establishment is for specific materials and specific detection indicators. Instrument manufacturers need to do the basic work well in advance to ensure that the user can use the instrument immediately after purchasing. However, the general instrument manufacturers can not have the ability of chemical analysis and chemometrics at the same time. The key to solve this problem is to build a large number of practical calibration models for instrument manufacturers and users, and provide model maintenance.

(3) No quantity, no quality

The improvement of instrument quality can not be completely solved overnight. In addition to the quality assurance system of design and every link, as a production enterprise, it needs to have a certain production batch guarantee, so as to accumulate a lot of production experience, improve the manufacturing process, establish the necessary process detection platform, and finally achieve the purpose of improving the quality of the instrument. Therefore, expanding the application range of near infrared spectroscopy and increasing the market share of domestic instruments can improve the enthusiasm of manufacturers and improve the quality of instruments. This is a relationship of mutual restriction. On the contrary, without market share, it is meaningless to talk about the quality of domestic instruments.

(4) No standard, no promotion

Near infrared spectroscopy (NIRS) is a kind of practical technology. Only when it is well applied in practical production is the real way out. To be promoted in the industry, we must get the help of government departments. At present, the national standard of near infrared spectroscopy is very few, the technology still stays in the internal quality control system of enterprises, and the detection data obtained in many cases are not recognized by the industry. Therefore, the introduction of relevant national standards as soon as possible and the recognition of the industry can expand the application range of near infrared spectroscopy analysis technology. In this regard, industry associations and professional associations have great prospects, and their work plays a great role in speeding up the localization process of near infrared spectroscopy instruments.

(5) Strengthen the basic research of analytical technology

Although many years of efforts have been made by scientific researchers, some problems of NIR analysis technology have not been fundamentally solved, and most of the time, it is still based on experience or matter of fact. For example: in the aspect of information extraction, how to evaluate the effect of a certain spectral pretreatment? Up to now, we have not seen a generally accepted method to evaluate theoretically, or just draw a conclusion of which pretreatment method is better by comparing the subsequent modeling effect (under a certain modeling method). Obviously, this is a method of "blind people feel the elephant"; What is the classification method, proportion and data range of calibration set and prediction set? And so on, the solution to the problem basically remains in the accumulation of experience

There are many similar problems, which need the majority of scientific research workers to do careful work. If we still stay in the form of "using some advanced instrument + some spectral preprocessing method + some modeling algorithm = solving the detection of a certain index", it is not very helpful to improve the application level of near-infrared spectroscopy technology.

Both high-end and dedicated, networking is the direction

Chen Bin believes that near infrared spectrometers are now in the ascendant stage of development. Compared with foreign advanced instruments, the following aspects may be the development trend of domestic near infrared spectrometers in the near future:

(1) Fourier transform is still the mainstream product

At present, the research-based near infrared spectroscopy instruments are basically Fourier type, and the requirements of Fourier transform hardware technology are high. So far, China can not independently develop and produce such instruments. More than 95% of the domestic FT-NIR market share is occupied by foreign brands.

Due to the use of laser positioning, the Fourier transform instrument has good wavelength stability, high spectral resolution and large luminous flux, which can minimize the difference between instruments. It has unique advantages in spectral standardization, and is convenient for the transfer of calibration model between multiple instruments, even between instruments of different companies. However, the modeling technology of Fourier transform instrument is demanding and the price is relatively high; The users are mainly concentrated in universities, research institutes and powerful industries (drug administration, tobacco, entry-exit, petroleum, military industry, etc.). The research and development of this kind of instrument needs a large amount of investment in the early stage, and needs a long time of technology accumulation. The localization of this kind of instrument needs the support of national policies, and it takes a long time to develop with the strength of the enterprise itself.

(2) Miniaturization, specialization and portability are the development direction

In order to meet the various needs in the production process, we need to develop a variety of practical instruments. We must not adopt the thinking mode of "one machine with multiple uses". This not only wastes resources, but also makes the instrument unable to take into account all aspects, often making the instrument unable to work at the best working point.

Miniaturization - in order to save space and facilitate online installation;

Specialization - can maximize the performance of the instrument to meet the detection of specific content. There are many links in the production process, only a few parameters need to be detected, sometimes only one parameter needs to be detected. The cost, accuracy, stability and reliability of the R & D special instruments will be greatly improved;

Portability - to meet the detection of field and emergency, this kind of instrument should not only consider the basic indicators of the instrument, but also consider the energy consumption, especially the energy consumption of the light source, as well as the cumulative working time, anti vibration, working environment, etc.

(3) Intellectualization, networking and data sharing are the contents of the instrument

In order to reduce the requirements of the instrument for users, and to make the instrument always work at the best working point in all kinds of environments, from the aspect of the instrument, it is to make the instrument itself smart and actively adapt to the changes of users and environment. The intellectualization of the instrument is the most basic requirement for the development of all analytical instruments.

The emergence of network instruments makes the analytical instruments from tangible to intangible. For laboratory instruments, a new concept of "analytical instruments used in fixed places + computer + network = Network Instruments" has been formed; For mobile and on-site portable instruments, a new concept of "mobile analytical instrument + mobile control platform + mobile network = mobile cloud instrument" should also be formed. We hope to have a convenient and simple new instrument system such as "spectrum sensor + mobile phone control platform + mobile network = mobile cloud spectrum detection system". The network instrument not only shortens the development cycle of the instrument, greatly saves the cost of the instrument, but also achieves the real sense of data sharing, resource sharing and talent sharing. In a sense, analytical instruments need more network technology and more cloud service system; Near infrared spectroscopy technology needs network instruments and mobile cloud spectrum detection system.

Due to the particularity of NIR Analysis (not only learning how to use instruments), it is unrealistic to require users to learn chemometrics. Data sharing is the best solution to give full play to the performance of NIR instruments. It can enable a master proficient in data processing and spectral analysis to provide timely and accurate services for users of one or more instruments in different places.

(4) Unity, cooperation, mutual benefit and win win cooperation promote the development of near infrared spectroscopy technology

Near infrared spectroscopy analysis technology involves chemistry, mathematics, computer application, signal processing, physical optics, spectral instruments and professional knowledge in the industry. Therefore, the development of this technology requires the participation of talents from various aspects. We hope that we can form a real joint research team based on the consensus of unity, cooperation, mutual benefit and win-win situation, so that we can achieve the goal in a short time, Only by jointly promoting the rapid development of near infrared spectroscopy analysis technology in China, can we develop a near infrared spectroscopy instrument suitable for China's national conditions, which can be used in the industry, and promote the process of practical research of the technology.

Interview editor: Liu Fengqiu