Ultra-simplified design and quantitative analysis for the optical system of compact laser-induced breakdown spectroscopy

1.   Introduction

Laser-induced breakdown spectroscopy (LIBS) technology is a new detection technique that was invented after the invention of laser technology. It has the characteristics of high speed, high sensitivity, small damage, non-contact, real-time and in-situ detection [1–3]. For example, the portable and handheld LIBS systems were used for the detection of volcanic rock by Harmon et al [4], in which the ablation size was only microns of scale, and the in-situ detection time was only a few seconds. These advantages were further strengthened with the help of ultra-fast laser technology [5, 6], resonance excitation technology [7] and beam shaping technology, etc [8, 9]. Because of this, LIBS technology has been widely used in industrial production [2, 10, 11], deep-sea exploration [12–14], Mars exploration [15, 16], biological testing and metal testing, etc. [17–19] in recent years. For example, the in-situ monitoring and quantitative analysis of composition and temperature for molten metal were realized using a combined LIBS-IR thermometry system by Zeng et al, which were the most essential parameters for process control in the metallurgical industry [20]. In the astrobiology field, García-Gómez et al presented the first-ever detection of the indigenous organic matter in rocks from the interpretation of carbon molecular forms using laser-induced plasma [21]. In whole, LIBS technology has shown its feasibility and great application potential in many fields [22–24].

As LIBS technology gradually came out of the laboratory to the application frontier in various fields, the miniaturization and integration of the technology have become the key trends driving the development of this technology [25]. However, in the aspect of system design, there is often a contradiction between the weight and volume of the system and systematic performance. In general, the better the systematic performance of the system, the more complex the system structure, the greater the volume and weight of the system, and vice versa. However, for almost any system including LIBS systems, the stability of the system is contrary to this rule, and the simpler the system is more stable. Therefore, it is very important to balance all kinds of factors such as stability, weight, and volume, especially for the handheld, portable, compact, and even ultra-compact LIBS systems [24]. For LIBS systems, the optimal design of optical systems is critical and even plays a fundamental role, which determines the excitation degree, excitation efficiency, and collecting efficiency of the plasma, to affect the spectral intensity, stability, performance of the qualitative and quantitative analysis model, etc. For example, in complex environments such as deep ocean, deep space, the handheld or portable LIBS systems that are often used in in-situ detection are generally strictly limited by weight and volume [14, 16]. The United States of America [26] and China [27] launched a Mars rover almost simultaneously in 2020, due to the limitations of rocket loads and the operational load of the Mars rover, both of them have an extremely strict volume and weight requirement for each module on the equipment, so the optical units of their LIBS systems on the two rovers were also optimized and integrated. Therefore, a compact and systematic LIBS system with a small volume and lightweight is essential and significant.

This research, for the first time, conducted a comprehensive optimization design and comparative analysis of three optical paths of LIBS system: the paraxial optical path (OP), the off-axis OP, and the reflective OP. This process aims to reveal the unique performance of each OP in terms of spectral intensity and quantitative analysis, providing a solid scientific basis for the selection of OPs in LIBS systems. We conducted detailed spectral analysis on Si, Cu, and Ti elements in aluminum alloy samples as the core research objects, and rigorously quantified key parameters such as spot quality, energy density, spectral intensity, stability, and quantitative analysis performance. Through this meticulous evaluation process, we not only gained a deep understanding of the spectral response characteristics of each element but also clarified the specific impact of different OP designs on the performance of LIBS systems. An efficient, stable, and suitable OP form for compact LIBS systems is summarized and selected, opening up new paths for the convenient application and performance optimization of LIBS technology.

2.   Materials and methods

2.1   Samples

A set of 5A66 Al alloy standard samples (Southwest Aluminum Group Limited) was used in this work, which includes E2121–E2125 5 samples. The concentration information of each element in the sample is shown in table 1. To reduce the effect of the impurities such as oxide on the surface of the sample, it was polished before the experiment. To avoid the irregularity of the sample and spectral fluctuations, a multi-point collecting method was adopted. 10 different points were selected for laser ablation on each sample surface, then 10 spectra were collected from each point, each spectrum was averaged by 20 spectra, and eventually 100 spectra were obtained from each Al alloy sample.

Table  1.  Information list of 5A66 Al alloy samples.

Standard samples	Si	Cu	Ti	Mn	Zn	Mg
E2121	0.027	0.0025	0.009	0.0017	0.0072	1.2
E2122	0.0033	0.0065	0.00084	0.0017	0.013	1.88
E2123	0.0073	0.0038	0.0042	0.0042	0.0047	1.63
E2124	0.013	0.011	0.011	0.011	0.0012	2.1
E2125	0.015	0.012	0.0006	0.0013	0.0024	2.13

 | Show Table

DownLoad: CSV

2.2   Instrumental setup

The compact LIBS system and the conventional LIBS system are consistent with the overall principle and main structure, but it only retains the most simplified devices that can achieve laser-induced and collection functions for plasma [23]. Its optical system was composed of a compact laser (wavelength: 1064 nm, energy: 6.5 mJ per pulse, frequency: 10 Hz, pulse width: 3.825 ns, diverge angle: 1.6 mrad@10 Hz, original spot size: ~ 1 mm, size: 70×45×40 cm³), spectrometer (Avantes Mini 3648, 264–437 nm, Grating: MN1800-0.25, Slit: 10 μm), laser focusing OP, plasma collection OP, and fiber, etc. The laser focusing OP is achieved through a focusing lens (single side spherical lens, focal length: 30 mm, size: 12.7 mm, accuracy: ±0.1 mm, material: ultraviolet fused quartz). In the experiment, the spectrometer operates in external triggering mode, with a delay of 95 ms between the laser external triggering signal and the spectrometer acquisition signal, and an integration time of 10 ms. 20 points are collected for each sample, and 10 spectra are collected for each point. To reduce spectral fluctuations, each spectrum is averaged from 20 spectra.

The plasma collection OP system has three types: the paraxial OP, off-axis OP, and reflective OP, as shown in figure 1. The paraxial OP is a relatively common form in the current compact LIBS system. Its main feature is that the plasma collection head is placed on the side of the laser-focusing lens at a specific angle (usually above, left, or right), and the laser-induced plasma is collected and focused by the collection head and then passed into the optical fiber, which is connected with a spectrometer. Its structure is shown in figure 1(a). In this research, the angle of the collection head and the laser beam was about 52.5°. The off-axis OP is also an optional solution for the compact LIBS system. In this structure, the laser passes through the rectangular hole of the plasma focusing lens and then is focused on the surface of the sample by the laser focusing lens, as shown in figure 1(b). The collection of plasma was through the focal lens that was placed behind the laser focusing lens, which was collected and directly introduced into the optical fiber and the spectrometer. The design diagram of the plasma focusing lens is shown in figure 1(d). In the reflective OP, the plasma emission light was reflected into the collection head by the reflector. The collection head and the reflector were placed in the front and end of the laser-focusing lens, respectively. The former and the latter were placed at an angle of 12° and 0° to the laser beam, respectively, as shown in figure 1(c). In addition, it is worth noting that the parameters such as the place, angle, distance, and lens sizes, etc. of the above three OPs were optimized to a certain extent, which were working in their relatively optimal conditions.

Figure  1.  Three types of OP: (a) paraxial OP, (b) off-axis OP, (c) reflective OP. (d) The design diagram of the plasma focusing lens used in off-axis OP.

DownLoad: Full-Size Img PowerPoint

3.   Results and discussion

3.1   Ablation spot

This study delves into the differences in spot ablation morphology and energy density among three OPs. Taking the 5A66-E2122 sample as a specific research object, a notable characteristic is that despite their different names, the paraxial and reflective OPs share an identical laser transmission route. This commonality directly leads to a high degree of similarity in both spot morphology and energy distribution density, as clearly illustrated in figure 2(a). In this figure, the circle represented by r₂ was the area of the ablation spot formed on the surface of the sample by the focused laser beam. This spot was irregular, and the energy distribution was also uneven. However, most of its energy focused on the circle represented by r₁ as shown, so it was used to approximate the energy density of the laser in our manuscript. In these two OPs, the r₁ was 181.3 μm, and as mentioned in section 2, the single pulse laser energy was 6.5 mJ, so their laser energy density was about 14.8 J/cm².

Figure  2.  (a) Ablation spot of the paraxial OP and reflective OP, (b) ablation spot of the off-axis OP.

DownLoad: Full-Size Img PowerPoint

Furthermore, the configuration of off-axis OP specifically explored, selecting a laser-focusing lens with a radius of 12.7 mm. This choice was made with consideration for the size limitations and aperture drilling difficulties associated with half-inch lenses. However, since the laser beam does not travel through the center axis of the lens but deviates from it, this change is directly reflected in the morphology of the focused spot on the sample, resulting in an elliptical shape with semi-major and semi-minor axes of r₃ = 237.4 μm and r₄ = 172.1 μm, respectively. Notably, a small bright area remains unabated in the lower-right corner of the elliptical spot, likely due to uneven energy distribution caused by the beam deviation. Ignoring this unabated area, we estimate the laser energy density for this OP using the elliptical area formula in conjunction with the known single-pulse laser energy, yielding a value of approximately 11.9 J/cm², which is slightly lower than that of the previous two OP configurations, clearly demonstrating the direct and significant impact of beam deviation on laser energy distribution and density.

It is worth noting the difference in uneven energy distribution in figures 2(a) and (b). Firstly, the first scenario is when the laser is focused through the center of the lens. In this case, the laser beam passes through the center of the focusing lens in a straight line without deflection or deformation. This is because the design of lenses is usually optimized based on their center to ensure that the beam maintains its original shape and direction when passing through. Therefore, when the laser beam is focused at the center of the lens, it can form a clear, circular focused spot, and the ablation profile generated by this spot on the material will also be circular. Then, in the second scenario, the laser is focused through a point that deviates from the center of the focusing lens. In this case, the laser beam will undergo deflection and deformation when passing through the lens. This is because the design of the lens does not provide the same focusing effect at all positions, especially when the beam deviates from the center of the lens. This deflection and deformation can cause the laser beam to form an irregular spot when focused, and the ablation profile generated by this spot on the material will also be irregular.

3.2   Spectral quality

3.2.1   Analysis of the average spectral intensity

Based on a thorough analysis of table 1, it is evident that among the aluminum alloy samples, E2122 stands out with its unique elemental composition: featuring the lowest concentrations of Si and Ti, and relatively low levels of Cu and Mn compared to the rest of the series. This characteristic renders E2122 an ideal candidate for evaluating the ultimate excitation performance of OP configurations toward trace elements. Consequently, this study focuses on E2122 and conducts a comparative analysis of spectral quality across different OPs.

Figure 3 presents a clear visual representation of the average spectra obtained from three OPs: paraxial (red), off-axis (blue), and reflective (gray). The spectral intensity of the paraxial path, denoted by the red trace, is notably higher than that of the other two paths, with the off-axis path (blue) exhibiting higher intensity than the reflective path (gray). This observation indicates that the paraxial path achieves the best overall excitation of trace elements in aluminum alloys, potentially enabling lower detection limits. Moreover, taking the specific background segment (340.65–341.87 nm) highlighted in figure 3 as an example, the spectral background intensities exhibit a similar trend to the overall spectra, with the paraxial path displaying the strongest background, followed by the off-axis path, and the reflective path having the lowest. Therefore, a detailed quantitative analysis is necessary to validate this inference further.

Figure  3.  The mean spectra of the paraxial, off-axis, and reflective OP.

DownLoad: Full-Size Img PowerPoint

3.2.2   Analysis of the spectral indicators of certain spectral lines

As shown in table 1, the aluminum alloy samples selected for this study are enriched with various critical trace metallic elements. Among them, Si serves as a vital alloying element that significantly enhances the strength, hardness, ductility, corrosion resistance, and thermal stability of aluminum alloys by forming solid solutions with aluminum. However, excessive addition of silicon can lead to increased brittleness of the material. Cu also contributes positively to enhancing these properties of aluminum alloys, but its over-addition can compromise their toughness. Ti, a commonly used additive in aluminum alloys, is typically introduced in the form of Al-Ti or Al-Ti-B master alloys. By reacting with aluminum to form the TiAl₂ phase, titanium acts as a non-spontaneous nucleus during crystallization, refining the cast and weld microstructures, and thereby optimizing the material’s structure. Therefore, precise detection of these elements is of paramount importance.

Crucially, the relatively low concentrations of these elements in the E2122 sample make it an ideal candidate for evaluating the limit detection capabilities of various OPs. Based on this, we selected Si, Cu, and Ti for in-depth investigation and conducted a quantitative analysis of their spectroscopic indicators to assess the performance of different OPs. Specifically, the analytical lines chosen were Si II 288.15 nm, Cu I 324.75 nm, and Ti I 334.94 nm, with a spectral background interval of 340.65–341.87 nm, as illustrated in figure 3.

Subsequently, we extracted the spectral intensities of these three analytical lines and calculated key parameters such as relative standard deviation (RSD), signal-to-noise ratio (SNR), and signal-to-background ratio (SBR). The results are presented in figure 4. From the figure, it is evident that the paraxial OP exhibits overwhelming superiority in spectral intensity, with intensities of 1437, 1432, and 1506 for Si II, Cu I, and Ti I, respectively, far exceeding those of the off-axis OP (90, 67, 51) and the reflective OP (4, 1, 8). Notably, the spectral intensity of the paraxial OP is over 150 times that of the off-axis OP, while the off-axis OP is more than 6 times stronger than the reflective OP.

Figure  4.  The spectral (a) intensities, (b) RSDs, (c) SNRs, and (d) SBRs of the 3 types of Ops.

DownLoad: Full-Size Img PowerPoint

Further analysis of figures 4(b) and (c) reveals significant differences in RSD and SNR among the OPs, with an overall trend consistent with spectral intensity. The paraxial OP continues to demonstrate superior performance, while the reflective OP is nearly ineffective in detecting these trace metallic elements. As for figure 4(d), although there is no significant order-of-magnitude difference in SBR among the OPs, the off-axis OP exhibits a slight advantage.

In conclusion, this study unequivocally highlights that in compact LIBS systems, the spectral quality of the paraxial and off-axis OPs is significantly superior to that of the reflective OP. This finding is of great significance for optimizing the detection performance of LIBS systems and their integration.

3.2.3   Mechanism analysis

In the LIBS field, spectral quality serves as a crucial benchmark for assessing system performance, which is intimately tied to the efficiency of the plasma collection optics. The core elements efficiency lies in the design of the collection angle and the collection area, both of which are strategically configured in three-dimensional space to determine the effective capture of plasma emission light.

Firstly, the geometric definition of the collection angle is the conical angle formed by the line connecting the edge of the collection optics lens to the ablation point. The size of this angle directly correlates with the efficiency of the OP in collecting plasma radiation. Simultaneously, the collection area, which represents the circular area where this conical surface intersects with the hypothetical hemispherical model of plasma emission, is also a key parameter in evaluating OP performance. Typically, as the distance between the collection optics and the ablation point decreases, the collection angle naturally increases, facilitating an expansion of the collection area and enhancing the intensity and stability of spectral signals.

Next, the paraxial OP stands out due to its unique layout advantage: the collection optics are strategically placed above (or to the side of) the laser focusing mirror, enabling maximum proximity to the ablation point for optimal focusing. This design not only shortens the OP, reducing light attenuation during transmission but also significantly increases the collection angle and area, thereby optimizing spectral signal intensity and various quality indicators (such as RSD).

In contrast, the reflective OP faces challenges: the plasma emission light must first undergo reflection by a mirror before reaching the collection optics. This complex OP layout not only lengthens the OP, increasing the risk of light attenuation, but also restricts the expansion of the collection angle and area. Consequently, in terms of spectral quality, the reflective OP often fails to achieve the desired performance, particularly in compact LIBS systems where its limitations are more pronounced.

In summary, based on current experimental conditions and the performance evaluation of OP designs, the reflective OP has certain limitations in the application of current compact LIBS systems. Therefore, we are more inclined to delve into the quantitative analysis of the performance of paraxial and off-axis OPs, aiming to enhance the overall performance and spectral quality of LIBS systems through optimized OP designs.

3.3   Quantitative analysis for Al alloy

By implementing a meticulous internal standardization process on the spectral data, we ensured that each analytical element was matched with its optimal internal standard line, thereby significantly enhancing the accuracy and reliability of the data. Subsequently, based on the unique spectral characteristics, we tailored the optimal fitting model for each element. For instance, in the case of the off-axis OP, Si and Ti elements, due to their spectral properties, were appropriately fitted with a linear model, whereas Cu, with its more complex response relationship, necessitated a quadratic fitting model. Conversely, in pursuit of optimal quantitative analytical capability, the off-axis OP adopted a uniform linear fitting approach for all elements.

The calibration curves presented in figure 5 clearly and intuitively demonstrate the remarkable superiority of the paraxial-axis OP in terms of quantitative performance. Taking Si as an example, the paraxial-axis OP exhibits a near-perfect fitting outcome, with a determination coefficient (R²) of 1, indicating a high degree of consistency between the model and the data. The average relative error (ARE) is as low as 2.75%, signifying minimal prediction errors. The root mean square error of cross-validation (RMSECV) is merely 0.0003wt.%, further verifying the high precision of the model. Additionally, the limit of detection (LoD) reaches 66.5 μg/g, indicating the remarkable capability of this OP to detect low-concentration elements. In contrast, the off-axis OP under identical conditions falls short in these crucial metrics, attesting to the paraxial OP’s superiority in quantitative accuracy, error control, and sensitivity.

Figure  5.  Quantitative analysis of Si, Cu, and Ti elements using paraxial OP and off-axis OP.

DownLoad: Full-Size Img PowerPoint

This advantage is also evident in the quantitative analysis of Cu and Ti elements. Specifically, Ti, present in minute concentrations (0.0006wt.%) in 5A66 aluminum alloy samples, poses a significant challenge for spectral detection. In the off-axis OP, the weak spectral intensity, suboptimal SNR, and SBR, along with a relatively large RSD, limit the overall performance of the quantitative analysis model, making it difficult to accurately capture and quantify the signals of such trace elements. These further underscores the insufficiency of the off-axis OP in detecting trace elements.

4.   Conclusion

In the exploration of OP configurations for compact LIBS system, the paraxial OP stands out as the highlight of this study with its unique design concept and exceptional performance. Its merits lie not only in the simplistic and flexible structural design but also in the significant enhancement of spectral quality and quantitative analysis efficiency. The paraxial OP simplifies the construction process of the OP and reduces operational complexity through its intuitive and easily adjustable design. This simplicity not only enhances the system’s usability but also offers users more freedom to optimize the OP configuration to accommodate diverse analytical needs. Additionally, its flexible adjustment mechanism enables the system to promptly respond to environmental changes, ensuring the stability and reliability of analysis results.

Regarding spectral quality, the paraxial OP demonstrates remarkable performance advantages. For the analysis line intensities of key elements such as Si, Cu, and Ti, the paraxial OP achieves a significant enhancement of over 6 times compared to the off-axis configuration, and even more than 150 times the improvement over the reflective OP. This innovation not only broadens the application range of LIBS systems but also improves the sensitivity and accuracy of analysis. Furthermore, the paraxial-axis OP maintains an extremely low range (10.9%–13.4%) for the spectral stability indicator RSD, providing a stable and reliable data foundation for quantitative analysis. At the quantitative analysis level, the paraxial OP also exhibits high precision and sensitivity. Its quantitative model fitting is precise, with well-controlled errors, offering users high-quality quantitative analysis results. This not only enhances the competitiveness of LIBS systems in the field of quantitative analysis but also provides users with more accurate and reliable experimental data support. Notably, although the paraxial OP may seem slightly insufficient in pursuing extreme compactness, it achieves a delicate balance between size and performance by sacrificing some compactness for superior spectral and quantitative analysis capabilities. This design philosophy reflects a profound understanding and precise grasp of practical application requirements, offering new insights into the optimized design of LIBS systems.