Objective  In in-situ analysis applications, where the promise of laser-induced breakdown spectroscopy (LIBS) is significant, the use of fixed analysis distances often proves impractical. Zoom-LIBS, with its greater flexibility, emerges as a more viable solution. The typical structure of a zoom-LIBS system employs a coaxial light path for both excitation and collection. This design enables precise laser focusing on the object's surface at various distances by adjusting the focal length of the telescope. However, the adjustment in focal length induces changes in the plasma state and system efficiency. Consequently, the LIBS equipment collects plasma spectra with varying intensities and characteristics. Integrating these spectra directly into a quantitave model for accurate inversion results becomes challenging. Moreover, standardizing plasma spectra after accurately calibrating equipment parameters at all analysis distances is a complex task, especially considering the diverse application scenarios of zoom-LIBS equiptment. To address these challenges, this paper proposes a reference pulse diagnostic method. This innovative approach allows the diagnosis of plasma temperature without the need for calibration equipment. Subsequently, the plasma spectra collected at different analysis distances can be corrected, facilitating accurate zoom-LIBS quantitative analysis.

  Methods  The reference pulse diagnostic method consists of three key steps. Firstly, two lasers with power densities P₁ and P₂ are used for plasma excitations, where one serves as an analysis pulse and the other as a reference pulse. Following the principles of plasma radiation theory, the ratio of the intensity of atomic lines at the same wavelength in the two acquired spectra is computed. This ratio, denoted as V₁, is a function of the plasma temperature. In the seconde step, the ratio of the intensity ratio of the ion line to the atomic line in the spectra generated by both the excitation pulse and the reference pulse is calculated. This ratio, denoted as V₂, is also a function of the plasma temperature. In the third step, the plasma temperature is determined by solving for it using the values of V₁ and V₂. Once the diagnostic method is completed, and through spectral standardization or alternative techniques, the plasma spectra collected at different analysis distances are restored to their respective distances. This facilitates the establishment of a quantitative model, enabling the realization of zoom-LIBS quantitative analysis.

  Results and Discussions   A set of aluminum-based standard samples was used to verify the zoom-LIBS quantitative analysis based on the proposed reference pulse diagnostic method. Initially, the results obtained from this method were compared with the Saha-Boltzmann method, revealing a maximum relative error of not more than 8%. To highlight the correction effect of this method, a basic multivariate linear model (principal element) and univariate linear models (for other elements) were used to establish a quantitative model at 2 m for nine standard samples. Additionally, four standard samples were analyzed at distances of 1.5 m, 2 m and 3 m, respectively. For elements with content exceeds 1%, only the relative error of the abundance of Si in sample No. 11 was around 16%, while the rest were all less than 8%. Most elements with a content of less than 1% exhibited a relative error ranging between 10% and 30%. This demonstrates that the proposed method can achieve quantitative analysis of zoom laser-induced breakdown spectroscopy in an in-situ detection environment, broadening the application space of laser-induced breakdown spectroscopy technology. Four standard samples were analyzed at distances of 1.5 m, 2 m, 2.4 m, 2.7 m and 3 m. In the inversion results, except for the relative error of Si element in sample 11, which is 16% for elements with a content of more than 1%, most of the rest are less than 8%. Additionally, most of the elements with a content of less than 1% exhibited a relative error between 10% and 30%. Theses results confirm that the proposed method can achieve quantitative analysis of zoom-LIBS in an in-situ analysis environment.

  Conclusions  While some error amplification may occur due to the division form used in the final step of the reference pulse diagnostic method, it facilitates a relatively accurate diagnosis of plasma temperature. Further precision in diagnosis results could be achieved by carefully selecting more appropriate spectral lines. An alternative and potentially superior approach involves directly performing plasma spectral correction based on the values of V₁ and V₂. This quantitative analysis method for zoom-LIBS, rooted in the physical model of plasma radiation, holds the potential to broaden the application scenarios of zoom-LIBS. Additionally, it serves as a valuable reference for the design of zoom-LIBS equipment.