At present, infrared spectroscopy, nuclear magnetic resonance and other commonly used technologies for fiber fingerprint research at home and abroad have the problems of complex and time-consuming detection processes; Traditional chemical testing and physical testing are not only inefficient, but also result in chemical reagent contamination, and belong to destructive testing, which cannot meet the needs of rapid and non-destructive identification.
The article proposes the use of electronic nose technology to extract trace odor information from fiber material volatilization, establish an odor fingerprint spectrum, and achieve rapid, convenient, and non-destructive identification of textile fibers.

01
AIRSENSE electronic nose
AIRSENSE electronic nose is a compact, fast, and efficient detection system consisting of a sensor array, signal preprocessing circuit, and computer pattern recognition system. Its sensor array includes multiple metal oxide gas sensors, which can accurately detect various volatile substances such as aromatic compounds, hydroxyl compounds, ammonia, etc. When the electronic nose comes into contact with the fiber sample, the sensor will undergo oxidation-reduction reactions with the trace gases evaporated from the fiber, causing changes in the conductivity of the sensor material and generating unique resistance change signals, which constitute the "odor fingerprint" of each fiber. This odor based recognition method does not require pre-processing such as burning or cutting of the sample, fundamentally achieving the goal of non-destructive testing.

02
Details of the experimental plan
1
Experimental materials
Take 5g each of 9 types of undyed bulk fiber materials, including acrylic fiber, silk, modal fiber, acetate fiber, nylon, spandex, cotton, polyester, and cotton/polyester 70/30 blended fabric to be identified. During testing, each sample was not burned
2
instrument and equipment
AIRSENSE's electronic nose from Germany contains 10 metal oxide sensors with varying sensitivity to different volatile compounds.
3
Experimental steps
Set or record test parameters: temperature 28 ℃, relative humidity 50%, sample bottle capacity 250mL, headspace generation time 30min, continuous sampling 80 times, obtain. nos type data file
Firstly, connect the AIRSENSE electronic nose to the computer and run its supporting software, then set the various detection parameters of the electronic nose The sample intake flow rate is set to 400mL, the sensor rinsing time is 180s, and the sampling time is 120s
03
Experimental results and analysis
Identification results of 6 types of single fibers
After sampling for 60 seconds, the response values of the electronic nose sensor tend to stabilize, indicating that the signal is stable and can be used for analysis.
There are differences in the main response sensors of different fibers:
Acrylic and acetate fibers: mainly respond to S1 and S3
Silk and spandex: mainly responding to S6 and S8
Modal: mainly responds to S1, S3, S6
Nylon: mainly responds to S2 and S3
Even if some fibers have the same primary response sensor (such as acrylic and acetate fibers), the response curves formed due to the G/Go deviation values of the main components and differences in the secondary components are different, which can distinguish fiber categories.

Data analysis (LDA vs PCA)
Identification results of cotton polyester blended fabric

Analysis method selection: Due to the more concentrated distribution of LDA data points and easier identification, LDA is used to analyze the data of cotton, polyester, and cotton polyester 70/30 blended fabrics.
Result performance:
The sample points of cotton and polyester fibers are concentrated and clearly defined.
The sample points of cotton polyester blended fabric are concentrated in the sample point areas of cotton and polyester, and most of the data points are close to cotton fibers (consistent with the actual composition ratio of cotton/polyester 70/30).
Conclusion: AIRSENSE electronic nose can accurately distinguish between cotton fibers and polyester fibers in cotton polyester blended products.

04
research findings
AIRSENSE electronic nose combined with linear discriminant analysis (LDA) and principal component analysis (PCA) can effectively identify six single fibers: acrylic fiber, silk, modal, acetate fiber, nylon, and spandex. Among them, LDA has more concentrated data points and better identification effect.

The AIRSENSE electronic nose can accurately distinguish between cotton and polyester fibers in cotton polyester blended fabrics, providing a fast, convenient, and non-destructive new method for identifying textile fibers.