Search
Close this search box.
Contact Us

Recent Advances in Continuous Fast Data Acquisition: < 100 µs for Single-Particle ICP-TOFMS

Download Application Note

Lyndsey Hendriks, Fredrik Oestlund & Martin Rittner
TOFWERK, Switzerland

Single-particle inductively coupled plasma time-of-flight mass spectrometry (sp-ICP-TOFMS) is a powerful analytical technique for the detection and quantification of inorganic nanoparticles in a wide range of applications. While the technique is rapidly evolving into a well-established tool for nanoparticle analysis, constant development and research continue to thrive to achieve the fast acquisition capabilities. The core principle of sp-ICP-MS involves measuring a particle suspension at a sufficiently low concentration while acquiring data at a high frequency. Under these conditions, discrete particle events can be detected above the background, as opposed to conventional liquid analysis, where constant steady signals are recorded. Subsequently, the number of particle events correlates directly with the particle number concentration, while the intensity of each event correlates with the elemental mass of individual particles, respectively size.

Here we evaluate the advanced capabilities of the TOFWERK icpTOF S2, which features improved fast acquisition hardware designed for high time-resolved single particle analysis. More specifically, we present single-particle data acquired at acquisition speeds down to 12 µs. As fast data acquisition rate and data processing are key factors in the accuracy of the characterization of single nanoparticles, the effect of different integration times from micro- to milliseconds is investigated in detail here with regard to the distinction of signals from the background, quantification of mass and size, as well as particle number concentration (PNC).

Experimental

Chemicals – Commercial nominal spherical gold nanoparticles (30, 50 and 100 nm) as well as 60 nm AgAu coreshells were purchased from NanoComposix. The 50 and 100 nm Au nanoparticles were diluted to a final concentration of approx. 1E5 particles/mL, while the 30 nm Au nanoparticles were diluted 10x more to reach a final PNC of 1E4 particles/mL.

Sp-ICP-TOFMS analysis – In this work, the icpTOF S2 was used, with a TOF period of 12 μs. All datasets were recorded using TOFpilot 2.14, which features advanced capabilities for millisecond data acquisition.

Data processing – All files were processed using the Liquid Reprocessing module in TOFpilot 2.14.  For microsecond integration times, the raw data was first binned to 1ms and then, in the same way as for millisecond integration times, the Compound Poisson algorithm was used to differentiate nanoparticle events from the background [1]. The data was corrected for split events. Mass and size quantification followed the procedure of Pace et al. [2].

Results

Figure 1. Overview of sp-ICP-TOFMS transient signals depending on the selected integration time: a) 12 μs, b) 120 μs and c) 1200 μs, for 60 nm AgAu coreshell nanoparticles. Thanks to the simultaneous multi-element detection capabilities of ICP-TOFMS, the silver shell (107Ag and 109Ag) as well the gold core (197Au) can be measured at the same time.

Based on its TOF period, the icpTOF S2 can measure the full mass spectrum continuously at acquisition speeds down to 12 μs. This means that within 12 μs, ions of different mass-to-charge ratios (m/Q) are measured continuously, and nothing is missed. As one nanoparticle event is about 0.5 ms [3], the signals of nanoparticles will present different profiles depending on the integration time used for the acquisition. Figure 1 provides the transient signals for 60 nm AgAu coreshell nanoparticles acquired with different resolution time, namely 12 μs, 120 μs and 1200 μs. As observed, when working with integration times shorter 500 μs, a time resolved profile can be measured for each nanoparticle (several data points), while with millisecond integration times, a single pulse is detected per nanoparticle (1-2 data points) [4]. Hence, a balance between too short and too long dwell times needs to be found.

  • Short integration times: If the dwell time is too short, the signal is dominated by Poisson Counting Statistics and can become very spiky, making it challenging to distinguish it from noise. The total signal will be spread over multiple data points, and individual counts may appear more random.
  • Long integration times: If the dwell time is too long, there is a risk of measuring multiple nanoparticles in a single pulse. This can lead to the co-occurrence of signals, making it challenging to distinguish between individual nanoparticles and obtaining accurate information about their size, composition, and concentration.

Quantification and Ultra-Fast Acquisition Time for Particle Analysis

Figure 2. Quantification of 30, 50 and 100 nm Au nanoparticles, depending on the selected integration time for a) size, b) PNC and c) mass concentration (μg/ml). With longer integration times of >1 ms, the risk of multiple particle occurrence is increased. This leads to an overestimation in the sizes for the 50 and 100 nm Au nanoparticles, as well as underestimation in PNC. The lower PNC of the 30 nm Au nanoparticles make it less affected by the different integration times as fewer concurrent events are detected, resulting in constant sizing and PNC. In all cases, a constant Au mass concentration is observed which confirms that all the particulate gold being introduced was measured.

Quantified results are presented in Figure 2. No notable difference in quantification is observed with integration times of 12 µs up to 1.2 ms. This suggest that while ultra-fast acquisition times (e.g., 12 µs) is possible, it may not be necessary for accurate quantification. On the other hand, longer integration times of > 1.2 ms pose a risk of capturing multiple nanoparticles within a single pulse, necessitating adjustments to the PNC and, indirectly, the dilution factor relative to the data acquisition frequency. At higher nanoparticle concentrations, the likelihood of co-occurring signals increases, complicating the distinction between individual nanoparticles. In this study, 50 nm and 100 nm Au nanoparticles were present at higher concentrations compared to the 30 nm Au nanoparticles.

Closer examination of the particle distribution (see Figure 3) revealed a multi-modal pattern, attributed to the simultaneous detection of multiple nanoparticles. This phenomenon results in a “hockey stick” trend observed in both the size and particle number concentration plots (Figure 2), where the total number concentration is underestimated, and particle sizes are overestimated. These findings underscore the importance of optimizing integration times and concentration levels to ensure accurate nanoparticle characterization. Longer integration times require lower PNC to prevent concurrent events, while short integration times allow for the distinction of different events at higher PNCs. Additionally, to ensure that all the gold was being measured, the Au mass concentration was determined (product of the particle mass and PNC). The constant mass concentration demonstrates that concurrent events increase the measured mass per particle but reduce the total PNC and thus do not affect the overall quantification of Au. This consistency validates the reliability of our fast data acquisition process, indicating that all Au nanoparticles are being detected regardless of integration time.

Figure 3. Overview of the size distribution at different integration times for 30 nm and 50 nm Au nanoparticles. Millisecond integration time may lead to the overlapping of two or more nanoparticles, which is reflected here in the multi-modal distribution observed for the 50 nm Au nanoparticles at 12000 µs. On the other hand, microsecond integration time has the potential to reduce such errors by recording time resolved profiles for the nanoparticles. However, it should be noted that initial PNC and subsequent dilutions should also be considered as these also play a role in the quality of the measured data. Indeed, the 30 nm Au nanoparticles were 10x more diluted and no multi-modal distribution was observed at longer integration times.

Dataset Size Consideration of Fast Acquisition Data of Single and Nanoparticles

Lastly, the acquisition of full mass spectra at high frequency also generates large datasets. Figure 4 gives an overview of the file sizes produced here, where 315 analytes were recorded over 10 minutes at different integration times. At acquisition frequency of 83.333 kHz, extremely large datasets are produced which may introduce additional stress on the data processing systems.

Figure 4. Important to note, microsecond integration time also lead to much larger datasets and then need to be binned back to 1 ms. Consequently, a compromise should be found between unnecessary large datasets while maintaining the quality of the sp-ICP-TOFMS data. Here 315 analytes were measured continuously over a 10 minute long acquisition.

Conclusion

Data acquisition rate (i.e. integration time) is an important parameter that influences the output of single-particle ICP-TOFMS analysis. As shown, depending on the integration time, the ion cloud produced by a nanoparticle will be measured into multiple lower intensity data points (microsecond integration time) or one high intensity data point (millisecond integration time). Our study demonstrates that the sizing accuracy for 30 nm, 50 nm and 100 nm Au nanoparticles remains consistent across a range of time resolutions from 12 µs to 1200 µs. Significant deviations in sizing accuracy were observed only at an extended integration time of 12000 µs, where the data showed a multimodal distribution indicative of multiple events being recorded within a single time bin. Our findings indicate that while ultra-fast acquisition at 12 µs is possible, an integration time of 1.2 ms provides a balance between accurate nanoparticle sizing and manageable datafile sizes, without compromising the integrity of the results.

The advanced features for fast acquisition for single-particle analysis were highlighted here using the icpTOF S2 but are not limited to this model. These can be applied to the the icpTOF R and icpTOF 2R allowing fast data acquisition down to 33 µs and 46 µs, respectively. Furthermore, these enhanced capabilities for fast data acquisition hold great potential for other applications such as laser ablation imaging.

References

[1] A. Gundlach-Graham, L. Hendriks, K. Mehrabi and D. Günther, Anal. Chem., 2018, 90, 11847–11855. DOI: 10.1021/acs.analchem.8b01551
[2] H. E. Pace, N. J. Rogers, C. Jarolimek, V. A. Coleman, C. P. Higgins and J. F. Ranville, Anal. Chem., 2011, 83, 9361–9369. DOI: 10.1021/ac201952t
[3] J. W. Olesik and P. J. Gray, J. Anal. At. Spectrom., 2012, 27, 1143–1155. DOI: 10.1039/C2JA30073G
[4] I. Abad-Álvaro, E. Peña-Vázquez, E. Bolea, P. Bermejo-Barrera, J. R. Castillo and F. Laborda, Anal. Bioanal. Chem., 2016, 408, 5089–5097. DOI: 10.1007/s00216-016-9515-y 

Learn More

icpTOF Product Page