Nanomaterials often possess physicochemical properties that are distinct from their dissolved and bulk analogues. This quality makes them very attractive for use in a wide range of products, but also presents challenges for assessing their impact on human health and the environment. One of the most fundamental measurements to make of nanomaterials is to measure their size and concentration. Single-particle inductively coupled plasma–mass spectrometry (spICP-MS) is one of the few analytical methods capable of detecting, counting, and sizing metal-containing nanoparticles at ultratrace levels in real-world samples. In this interview, Martin Hassellöv, a professor of chemistry in the Environmental Nanochemistry Group at the University of Gothenburg, in Sweden, talks about the current state of spICP-MS for nanoparticle measurements and his work to advance this method.
Could you describe briefly how spICP-MS works?
Basically, compared to normal ICP-MS, where one strives to obtain as steady a signal as possible, in spICP-MS one utilizes the fact that bursts of ions are generated as metal nanoparticles pass through the plasma. By collecting data at much faster acquisition rates than in normal ICP-MS it is possible to probe such ion bursts as individual particle events for dilute samples, and analyze the frequency of events (directly proportional to the particle number concentration) as well as the signal intensity of particle events, which is proportional to the mass of element in each particle (related to size).
What can spICP-MS do in terms of the detection and measurement of nanoparticles that standard ICP-MS cannot?
In essence, spICP-MS is an element-specific particle counter that also provides size information, so it is very different from the bulk measurements that we are familiar with from standard ICP-MS.
What is the importance of dwell time in spICP-MS, and how do you optimize it?
In conventional spICP-MS the dwell time needs to be optimized; as you lower it, you improve the signal-to-noise ratio and lower the probability of multiple particle coincidences, but you also increase the probability of incomplete particle events. We found that 5 ms was a good compromise.
How do you assess and optimize nebulization efficiency in spICP-MS?
There is an indirect method that is commonly used (using a reference nanoparticle to back-calculate), but for direct determinations we have found that compared to only measuring the waste flow (liquid mass balance), accuracy could be improved by analyzing the metal content of the spray chamber waste, because some analyte-droplet partitioning may occur in the spray chamber.
What approaches can be used to improve the ability to discriminate nanoparticles from the background?
The ability to discriminate nanoparticles from the background or dissolved ions in the sample or diluting water is a key issue when developing the method for small nanoparticles. Remember that the mass concentration scales with size cubed so small nanoparticles contain very little mass compared to their larger counterparts. In a study we published in 2012 (1), we first investigated the previously suggested 3σ of the background signal as a detection threshold, but found that gave many false positives, and that 5σ would be more appropriate. Still, this approach of “conventional" spICP-MS has a lot of difficulty discriminating whether a dwell time of approximately 5 ms contains a small nanoparticle or if it is only ionic noise contributing to the signal.
What are the limits of detection of spICP-MS — in terms of nanoparticle size and concentration levels — and what factors affect those limits?
An interesting feature of spICP-MS is that in principle it is not limited by the mass concentration; if one can detect the particles then it is just a matter of how long an acquisition time (how much sample volume) one can afford to obtain statistically valid results. The theoretical lowest detectable size depends on the element (monoisotopic is best), stoichiometry, and ion transmission, but for something like gold nanoparticles the theoretical size is around 5 nm or slightly below.
What is the new “fast spICP-MS” method and how is it different from conventional spICP-MS?
In “conventional” spICP-MS one applies a dwell time that is a bit longer (typically 5 ms) than the extent of a nanoparticle event to avoid having many incomplete particle events (fronts or tails of events). In the new “fast” method that we recently demonstrated (2), we applied a much faster acquisition rate (0.1 ms) so that a true real-time capture of the events with 4–6 data points across the peak is obtained. That improves the smallest detectable size, and increases the dynamic range in terms of particle concentration because the probability of multiple particle events is drastically reduced.
How does your noise deconvolution approach improve the ability to detect and quantify smaller nanoparticles? How much improvement have you seen with this approach?
The noise deconvolution is based on the conventional data acquisition for a 5-ms dwell time, but a range of blanks and low-concentration dissolved standards is acquired and the signal is fitted to a Pólya–Gaussian probability mass function, which fits better than a Poisson-Gaussian probability mass function (3). Then it is possible to fit, deconvolute, and remove the dissolved background signal from the nanoparticle and background signal. This improves the lowest detectable size and makes it possible to discriminate and analyze 10-nm gold particles.
How accurate is spICP-MS or fast spICP-MS when measuring nanoparticles in real-world samples? Can you give an example of the results you have seen so far with either method?
The accuracy in real world samples still need to be investigated because there are no reference materials or suitable reference methods for environmental samples. But we have started to investigate the accuracy of the conventional spICP-MS in simple solutions (2). In terms of size determinations, if nebulization efficiency has been thoroughly determined then spICP-MS is comparable to reference methods such as electron microscopy. We have started to compare the accuracy of concentration measurements, but improved reference methods are needed to make final conclusions.