Introduction

Laser-Induced Breakdown Spectroscopy (LIBS) is widely used for elemental analysis because it enables rapid testing1, requires minimal sample preparation2, allows remote operation3, and provides real-time results4. In LIBS, a high-powered laser is focused onto a sample, where multiphoton absorption and inverse bremsstrahlung photoionization lead to plasma formation5. The resulting plasma emits light that contains the characteristic spectral signatures of the constituent elements6.

Despite these advantages, LIBS faces key limitations. Plasma produced by a single laser pulse is typically unstable, with a rapidly fluctuating spatial profile and a very short duration suitable for analysis7. Several approaches have been developed to mitigate these challenges, including beam shaping and the use of noble gas environments to improve ablation control8, as well as plasma stabilization strategies such as spatial and magnetic confinement9, 10, 11, dual-pulse LIBS 12, and microwave (MW)-assisted breakdown13, 14, 15, 16, 17, 18, 19, 20.

Among these methods, MW-assisted LIBS is particularly promising. It has demonstrated substantial improvements in plasma stability8, and reliability of elemental analysis21, 22, 23, 24, 25, 26, 27. Microwave coupling drives plasma expansion, yielding signal enhancements of up to two to three orders of magnitude compared with standard LIBS28. The enhancement correlates directly with absorbed microwave power, governed by impedance matching between the MW source and the plasma29. While standard LIBS plasmas persist for only nanoseconds, MW excitation extends plasma lifetimes to the millisecond range, even at modest laser energies of ~ 2 mJ. This effect has been verified in both metallic systems (stainless steel, zirconium)30 and oxides (alumina oxide, gadolinium oxide, zirconium oxide, and uranium oxide)29, 31. The prolonged lifetimes arise from continuous re-excitation of neutrals and ions in the plasma22.

We previously demonstrated the effectiveness of microwave-enhanced LIBS (MWE-LIBS) across a wide range of materials32, 33, reporting significant signal-to-noise ratio (SNR) improvements in alumina7, lead, stainless steel, zirconium metal, and gadolinium. MWE-LIBS also enabled improved isotopic measurements of uranium by stabilizing emissions33.

Nevertheless, MWE-LIBS has drawbacks. Elevated background noise often accompanies the prolonged plasma emission33, and matrix effects and self-absorption remain largely unmitigated34. In single-point collection geometries, the background contribution is exacerbated as more emissions are accumulated. However, using multiple collection points and combining them into a multifiber arrangement offers a pathway to suppress background contributions while preserving enhanced plasma emission. Addressing this challenge defines the core research objective of this study.

Figure 1 illustrates a critical feature of microwave-enhanced LIBS captured through A high-speed camera (640 × 280, 100,000 frames/sec, Fastcam SA-Z, Photron, UK): the dramatic expansion and stabilization of aluminum plasma under microwave irradiation. More details of the methods are found in reference8.

Fig. 1
Fig. 1
Full size image

(a) Enlarged plasma emission of laser ablated Al plasma by MW and their corresponding (b) approximated volume. Methods and similar images found in8.

Even at a low laser energy of 1.0 mJ, the plasma volume increases by more than 20-fold and persists for over a millisecond, compared with the nanosecond-scale lifetime of conventional LIBS plasma. This expansion motivates the development of advanced emission-collection strategies, since larger plasma volumes generate more photons that can be harvested efficiently. To address this, our research focuses on enhancing large-area emission collection and improving the SNR.

The intensity of the ablation plasma was enhanced by longer Delay time of the monochromator, which was sustained in an air longer time because of ME injection. At the same process, we can observe lager plasma in a air. This plasma can be collected by larger receiving optics.

By coaxial multifibra arrays with microwave excitation, we aim to significantly boost detection sensitivity. The expansion of plasma by microwave can be collected by this multifiber and increased the signal qualilty. This approach is particularly relevant for applications such as recycling aluminum alloys, where precise compositional identification is increasingly critical.

Methodology

Figure 2 presents the experimental setup of the MWE-LIBS system with multifiber emission collection. The fiber–optics arrangement was designed with a dual function: laser delivery and plasma emission collection. A 1053 nm laser (Q1, Quantum Light Instruments, Lithuania) was guided through the central fiber of a multifiber bundle (RP20 and FT200UMT, Thorlabs, NJ), while the surrounding fibers collected plasma emissions that were subsequently sustained and enlarged by microwave excitation. These emissions were then combined into a single 200 µm diameter fiber using a custom-built series of collimating and focusing lenses, engineered in-house to maximize photon throughput and minimize optical losses.

Fig. 2
Fig. 2
Full size image

Multifiber configuration in MWE LIBS.

The MW source operated at 2.45 GHz with a peak power of 1 kW and a pulse duration of 1 ms. Energy was delivered to the plasma through a helical antenna, enabling efficient coupling and expansion. This configuration provided precise control over plasma stabilization, thereby improving emission intensity and overall detection performance.

Spectroscopic measurements

Plasma emissions were analyzed using an echelle spectrometer (EMU-120/65, Catalina Scientific, AZ, USA). The instrument was equipped with a diffraction grating providing a spectral resolution of λ/50,000. Emission signals were recorded with a gate width of 1000 µs and a gate delay of 0.5 µs to optimize the SNR. Each acquisition employed an exposure time of 1.0 ms. Wavelength calibration was performed using a standard spectral lamp to ensure accuracy. Light collection was facilitated by a 400 µm core fiber coupled resulting to spot size of 240 μm. The laser energy density at 1 mJ is ~ 0.2 GW/cm2.

Enhancement equations

As reported in the previous report24, the degree of enhancement by the microwaves is expressed by the intensity enhancement factor (IEF), which is the ratio of the magnitude of the intensity of the atomic emissions under MWE-LIBS (\({I}_{MW}\)) to that under the standard LIBS (\({I}_{no MW}\)). This magnitude of intensity was obtained by subtracting the average background signal from the highest peak value. IEF is a unit-less value computed using Eq. 124.

$$IEF = {{I_{MW} } \mathord{\left/ {\vphantom {{I_{MW} } {}}} \right. \kern-0pt} {}}I_{no MW}$$
(1)

The SNR (S/N)24 was measured by dividing the ratio of the emission intensity signal (H) by three times the standard deviation (σ) of the background signal, as shown in Eq. 2. H is the difference between the intensity peak and the ‘zero base’ or the average background signal.

$${S \mathord{\left/ {\vphantom {S N}} \right. \kern-0pt} N} = {H \mathord{\left/ {\vphantom {H {3\sigma }}} \right. \kern-0pt} {3\sigma }}$$
(2)

Multifiber collection concept

Figure 3 illustrates the role of multifiber systems in enhancing emission collection. Experimental results with MW-sustained air plasma revealed an enlargement factor of up to ~ 400-fold, corresponding to an increase in plasma cross-sectional area from ~ 0.26 to ~ 14.84 mm2 [43]. In this configuration, six 200 µm fibers were employed, and their combined collection area (0.080 mm2) was further optimized by in-house focusing optics. This design ensures that a greater fraction of the expanded plasma emission is captured, thereby significantly improving the detection sensitivity of MWE-LIBS.

Fig. 3
Fig. 3
Full size image

Illustration of enlarged area collection using multifiber [43].

Aluminum alloy samples

Five aluminum alloy samples with varying Fe impurity levels were selected to evaluate the performance of LIBS and MW-LIBS. The chemical compositions of the alloys are summarized in Table 1. Among these, Al-1100 (99.1% Al, 0.5% Fe) was employed in most experiments as the representative aluminum sample, owing to its high purity and widespread use in LIBS studies. The other alloys (Al-2055, Al-5181, Al-6063, and Al-7050) were used primarily for limit of detection (LOD) analysis and calibration purposes, since they span a broader range of Fe concentrations from 0.03% to 0.25%.

Table 1 Aluminum alloy composition.
Full size table

The Alloy contains traces of Fe at low concentrations. However, each calobration sample of aluminum alloy also contains different concentrations. This lead to attempt Al limit detection at higher Al concentrations even though it is not ideal.

Results

Multifiber collection performance

The limits of single-fiber collection are clear. A single fiber only samples a small portion of the plasma, which means most of the emission is not captured. By comparison, a multifiber setup collects light over a larger area, increasing the number of photons that reach the detector and improving data quality. This makes multifiber collection a practical and effective approach for getting the most out of MWE-LIBS.

Our aim is to use this larger collection area to improve the SNR, while also applying strict noise control to keep the data reliable. As shown in Fig. 4, the Al I lines measured with the multifiber system were more than twice as intense as those recorded with a single fiber. At the same time, the background noise also increased.

Fig. 4
Fig. 4
Full size image

(a) Atomic emission of Al Alloy and (b) the background emission level for single and multifiber collection in MWE-LIBS.

The stronger signals therefore come with a cost. Multifiber collection clearly boosts emission intensity, but it also makes noise control more critical. Post-processing methods such as baseline correction are recommended and can easily address this issue, ensuring that the extra signal gain can be fully utilized without compromising data quality.

Figure 5 shows the effect of multifiber collection compared to a single-fiber setup using continuous Xenon lamp. Across the 400–800 nm range, the total emission (area under the spectrum) increased by about six times with multifiber, while the Xe I line at 479.3 nm was four times stronger. This confirms that combining six 200-µm fibers into one fiber using compression optics worked effectively.

Fig. 5
Fig. 5
Full size image

Comparison of emission spectra collected with (a) single fiber and (b) multifiber configurations. The multifiber arrangement yielded a six-fold increase in integrated broadband emission (400–800 nm) and a four-fold enhancement of the Xe I line at 479.3 nm.

The area under the emission curve represents the total number of photons detected. Since a xenon lamp emits light continuously, the detector integrates this steady photon flux over the chosen gate width (integration time). A longer gate width collects more photons in direct proportion to time, but the relative gain between multifiber and single fiber remains constant because it depends only on the optical geometry.

From a statistical perspective, the total photon yield increases through multifiber collection. Because noise grows more slowly than the signal, SNR improves as photon count rises. This makes weak spectral lines that are usually hidden in noise appear clearer, since the peaks stand out more distinctly above the background. As a result, analytical sensitivity is enhanced and detection limits are lowered. The lamp’s emission physics itself does not change, but the multifiber approach captures more of the emitted photons, leading to stronger signals and more reliable spectroscopic measurements.

The plasma is enhanced both temporally and spatially by MW. As previously reported, the deposition can be several hundred times larger. Therefore, rather than transmitting the enlarged plasma through a single optical fiber, we achieved a large volume of plasma light with a large receiving capacity.

Application to aluminum alloy

To evaluate the practical applicability of MW-enhanced LIBS, experiments were conducted on an aluminum alloy target using both single- and multifiber configurations. Figure 6a presents emission spectra obtained with and without MW excitation in the single-fiber setup. The laser pulse energy was 1 mJ, and the four strongest elemental lines were identified as Al I, Fe I, Mn I, and Cu I. Figure 6b shows the corresponding results for the multifiber configuration. In both cases, microwave excitation substantially increased emission intensities; however, the enhancement was significantly greater with multifiber collection. These spectra were acquired under identical conditions (exposure time: 1 ms; gate delay: 1 µs), demonstrating the combined advantage of MW excitation and multifiber collection for maximizing emission signal strength.

Fig. 6
Fig. 6
Full size image

Comparative emission collection in (a) single-fiber LIBS and (b) multifiber LIBS under microwave (MW) and non-MW conditions. Laser specs: λ = 1053 nm, τ = 7 ns, E_l = 0.5 mJ, Power = 1.4 GW/cm^2, M^2 = 1.96, 100 shots. Fiber core: 200 µm, 0.22 NA. Outer fiber: 200 µm, 0.22 NA. Spectrometer resolution: λ/30,000. Temporal settings: Exposure = 1 ms, Gate delay = 1 µs, Intensifier gain = 3500.

A closer inspection of the emission spectra in the range of 393.5–397 nm (Fig. 7) highlights the Al I lines at 394 and 396 nm. Both lines exhibit substantial intensity gains under MW excitation compared with standard LIBS. For the single-fiber configuration (Fig. 7a), intensities increased from ~ 8.9 × 103 a.u. without MW—which essentially corresponds to baseline noise with no distinct spectral peaks, due to the fiber’s limited overlap with the small ablation core—to ~ 8.1 × 10⁶ a.u. (394 nm) and ~ 8.9 × 10⁶ a.u. (396 nm) with MW. This behavior reflects a collection-geometry constraint: a single fiber with NA = 0.22 subtends only ~ 1.2% of 4π steradians, so small plasma volumes or lateral offsets result in negligible collected emission. Under MW excitation, the plasma expands and persists longer, increasing its overlap with the fiber acceptance cone and enabling detection. In the multifiber configuration (Fig. 7b), even greater enhancements were observed, with emission intensities rising from ~ 0.3–0.4 × 10⁷ a.u. (without MW) to ~ 2.1–2.4 × 10⁷ a.u. (with MW). These results clearly demonstrate the combined benefits of MW excitation and multifiber collection, producing emission enhancements of three to four orders of magnitude relative to conventional single-fiber LIBS.

Fig. 7
Fig. 7
Full size image

Comparison between atomic emission of Al I line in (a) single-fiber LIBS and (b) multifiber LIBS under microwave (MW) and non-MW conditions.

Energy-level dependence of Al transitions

Figure 8 compares the effect of microwave (MW) excitation on Al emission lines at different excitation energies. The left panel shows that the Al I line at 206 nm (7.6 eV) reaches a maximum IEF of ~ 400 under MW excitation, whereas lower-energy transitions exhibit much weaker enhancement. The right panel confirms that lines originating from ~ 3–4 eV states are only modestly enhanced, with maximum IEF values of ~ 7.5. This behavior illustrates the fundamental limit of photon collection alone: optimizing optics or using multifiber collection increases the number of photons retrieved, but cannot change the underlying excited-state populations in the plasma. In contrast, MW excitation directly alters plasma dynamics by extending plasma lifetime and driving re-excitation and re-ionization processes, thereby replenishing higher-energy states that would otherwise decay rapidly. The strong enhancement observed at 7.6 eV reflects the role of MW in sustaining electron collisions energetic enough to repopulate these levels. In short, photon collection amplifies existing emission, but MW excitation modifies the plasma itself to generate additional photons, with the degree of enhancement ultimately constrained by the excitation potential and population distribution of the element.

Fig. 8
Fig. 8
Full size image

Comparison of spectral line intensities of aluminum (Al I) at various energy levels with and without the application of microwaves (MW (a) single-fiber LIBS and (b) multifiber LIBS).

Figure 9 compares the SNR of Al I spectral lines at different excitation energies with and without MW excitation. The x-axis represents excitation energy (eV), while the y-axis shows the SNR on a logarithmic scale (101–10⁶). Red markers denote MW-assisted LIBS, and blue markers correspond to standard LIBS.

Fig. 9
Fig. 9
Full size image

Comparison of the SNR of aluminum (Al I) spectral lines at various energy levels with and without the application of microwaves (MW) (a) single-fiber LIBS and (b) multifiber LIBS.

For the single-fiber configuration, MW excitation produced a clear SNR enhancement at the 4.0 eV transition (208/209 nm), where SNR increased to ~ 103 compared with ~ 101 without MW. At higher (7.6 eV) and lower (3.1 eV) energy levels, the SNR remained low under both conditions, with a sharper decline observed with MW.

For the multifiber configuration, the same enhancement was observed at 4.0 eV, though the improvement was less pronounced than in the single-fiber case. At other energies, SNR values remained low, following a similar downward trend. Both single- and multifiber setups highlight the selective effect of MW excitation in boosting the SNR of the 4.0 eV Al I transition, while having limited impact at other excitation energies.

Limit of detection analysis

In Fig. 10, Calibration curves for Al I (396.4 nm) and Fe I (373.5 nm) emissions obtained with and without microwave (MW) enhancement using multifiber light collection. For both analytes, MW excitation produced substantially stronger signals and more reliable calibration behavior compared to conventional LIBS. In the case of Al, the relationship between concentration and emission intensity followed a near-linear response, with MW-enhanced signals reaching more than an order of magnitude higher than those obtained without MW. The calculated limits of detection (LOD) and quantification (LOQ) for Al improved to 0.590 wt.% and 1.966 wt.%, respectively, under MW coupling.

Fig. 10
Fig. 10
Full size image

Limit of Detection (LOD) measurements of (a) Al I at 396.4 nm and (b) Fe I at 373.5 nm, comparing calibration curves obtained with and without MW enhancement using multifiber LIBS. Error bars represent standard deviation from 10 replicate measurements.

For Fe, the calibration response exhibited a nonlinear, power-law behavior, indicating that MW coupling not only increased emission intensity but also amplified higher-order concentration dependence. This is why because that the microwave enhanced plasma is sustained in an air while the MW is injected and the effect is over several hundred times longer. The MW-enhanced Fe curve yielded a lower LOD of 0.323 wt.% and an LOQ of 0.510 wt.%, compared with significantly poorer figures of merit in the absence of MW excitation (LOD = 0.378 wt.%, LOQ = 0.678 wt.%). Unfortunately, the ratio Fe/Al did not show a good calibration fit.

MW excitation improved both the sensitivity and the usable dynamic range of multifiber LIBS. In conventional operation, the signals were close to the detection threshold, making calibration difficult. With MW coupling, the emission curves became stronger and more stable, producing clear trends that could be reliably fitted for calibration. The detailed regression parameters and analytical figures of merit are listed in Table 2.

Table 2 Regression statistics for Al I (396.4 nm) and Fe I (373.5 nm) calibration curves obtained with and without MW enhancements using multifiber LIBS.
Full size table

Synergy of multifiber and microwave enhancement

Our research examined two principal strategies to improve LIBS performance: multifiber light collection and microwave (MW) excitation. As shown in Fig. 11, multifiber collection alone increased both emission intensity and SNR by nearly two orders of magnitude compared with the single-fiber baseline. This improvement arises from the higher light-collection efficiency and broader plasma sampling enabled by the multifiber geometry.

Fig. 11
Fig. 11
Full size image

(a) Enhanced intensity and (b) SNR in multifiber configurations compared to single-fiber setups, attributed to improved light collection and broader emission capture.

MW excitation provided a further leap in performance. For single-fiber LIBS, MW coupling enhanced intensity by about 130 × and SNR by about 50 × relative to conventional operation. When combined with multifiber collection, MW excitation delivered an even stronger effect, with intensity enhanced by about 1500 × and SNR by about 500 × . Surprisingly, when comparing the single and multifiber results with MW, the enhancement is around 7.5 × which is nearly proportional to the number of fiber bunldes. These gains are attributed to the combined effects of a larger radiating plasma volume, extended emission lifetime, and improved overlap with the collection fibers.

Together, these results highlight the synergistic role of MW excitation and multifiber collection. While multifiber geometry alone mitigates the geometric inefficiencies of single-fiber systems, the addition of MW excitation unlocks its full potential, delivering strong, stable signals with substantially enhanced SNR. This synergy directly translates into improved detection limits and greater analytical reach across a wide range of materials.

To summarize, the effect of MW was about 500 times greater, while the effect of bundling fibers alone was about 7.5 times, which is nearly proportional to the number of bundled fibers. By combining both, an intensity improvement of approximately 1500 times was achieved. The MW effect also plays a clear role in attenuating high-frequency components. The implication is that in applications where maintaining low laser output and minimizing cavity size are important, we were able to demonstrate a high-performance measurement method even when using a low-performance spectrometer.

Conclusions

This study demonstrated that combining multifiber collection with microwave (MW) excitation markedly enhances LIBS performance. Multifiber collection alone increased emission intensity and SNR by nearly two orders of magnitude compared to single-fiber operation, owing to higher photon throughput and a broader plasma sampling volume. With the addition of MW excitation, the plasma was spatially and temporally sustained and expanded, yielding over three orders of magnitude improvement in intensity relative to the single-fiber baseline.

Microwave enhancement primarily functions by prolonging plasma lifetime and expanding its volume, thereby increasing the number of photons available for detection. In parallel, multifiber collection enlarges the effective sampling area, further benefiting from the spatial expansion induced by MW excitation. Together, these effects substantially improve photon capture and signal quality.

For aluminum, the MW-assisted multifiber configuration produced an almost linear calibration curve, with an LOD of 0.590 wt.% and an LOQ of 1.966 wt.%. For iron, calibration followed a nonlinear power-law relationship but still yielded improved detection limits (LOD = 0.323 wt.%, LOQ = 0.510 wt.%).

In summary, MW-assisted multifiber LIBS simultaneously improves emission intensity, SNR, and calibration performance. By accommodating both linear and nonlinear emission behaviors, this combined approach achieves lower detection limits and greater analytical reliability, underscoring its potential for sensitive LIBS measurements across a wide range of materials.