Comparing Micro- & Nano-spectroscopy Tools: Raman, PiFM, ToF-SIMS, & SEM

Background

Many analytical techniques can provide researchers with crucial insights into how our world works at the micro and nano scales. Since each technique has its own strengths, using many analytical techniques in concert can be an extremely effective approach. PiFM (photo-induced force microscopy) is one of the only imaging techniques that can provide direct chemical data with sub-5 nm resolution. Similarly, PiF-IR (photo-induced force infrared) spectroscopy is one of the only options for getting IR absorption spectra on monolayer materials. These specifications sound impressive on their own, but the real utility lies in an instrument’s ability to help people solve important scientific questions.

A recent Nature Communications paper used a variety of techniques, named the micro-spectroscopy toolbox, to analyze polyolefin catalysts [1]. The goal was to explore the formation of polyethylene on a catalyst model that is more relevant to the industrial production of polyethylene. These data provide an excellent showcase of how different advanced nanoscale analytical tools can complement one another to efficiently study complex chemical systems.

The model

The authors focused on polyolefin catalysts which, due to their hierarchically complex nature, are usually studied using a simplified planar model system whereas highly spherical catalyst particles are used industrially. To bridge the gap between the spherical catalysts and planar catalyst model, the authors introduce a spherical cap model that can be analyzed by all the techniques in their micro-spectroscopy toolbox.

The material discussed is a Ziegler-type catalyst spherical cap model which is based on a moisture-stable LaOCl framework designed to support a TiCl₄ pre-active site. This is created to be as consistent as possible with the industrially relevant MgCl₂ framework that is technically and experimentally limited due to high moisture sensitivity. Therefore, the authors were inspired to design the LaOCl support matrix because it would provide strong SEM imaging contrast due to the high atomic weight of the lanthanide, and because of its stability in ambient conditions. The primary goal behind this new spherical cap model is to provide a system in which to study ethylene polymerization process ex-situ, but in a system more like the highly spherical industrial framework.

The toolbox

The micro-spectroscopy toolbox discussed consists of IR PiFM, PiF-IR, Raman microscopy, FIB-SEM-EDX, XPS, and ToF-SIMS. These techniques were all used to examine the interplay between the catalyst and the ethylene polymer phases formed after polymerization times ranging from 1 minute to 60 minutes. The authors were able to make several insightful observations using each of the tools in their toolbox.

Results comparison

Results from Raman

Besides PiFM and PiF-IR, Raman is the only other vibrational (and therefore chemical) analysis technique used. In this paper they used Raman microscopy to provide an efficient method of mapping the -CH stretching vibrations in the 2700–3100 cm⁻¹ range. This will visualize the local thickness differences of the polyethylene formed in the spherical cap, but with a limited resolution of only 360 nm in the best-case scenario.

Using Raman, the authors were able to determine that starting at the 2 min ethylene polymerization time, the formation of polyethylene can clearly be seen at both the center and edges of the spherical cap. However, the signals were stronger near the edges. At the longer polymerization times (5, 20 and 60 min) the trends seen with Raman microscopy were that the polyethylene layers grew more consistent across the spherical cap, and the layers grew to be thicker. Therefore, they conclude that polyethylene yields increased as a function of polymerization time.

Results from PiFM

As the authors note, PiFM has a much higher spatial resolution than any of the other chemical analysis techniques used in this study. Therefore, PiFM allows them to map the crystalline polyethylene via the -CH₂– bending vibrations at 1461 cm⁻¹ (B_1u) and 1471 cm⁻¹ (B_2u). These PiFM maps are correlated to the topographic information obtained simultaneously by the AFM part of the Vista One instrument (referred to as a Vistascope).

The PiFM images corroborate the conclusions based off the Raman data – that the thickness, and therefore yield, of the polyethylene increases as a function of polymerization time. However, because PiFM has such high spatial resolution, they were able to see more details than Raman provided.

Here, they can see that even at the early polymerization times, there are well-defined and intertwined polyethylene fibers growing outwards of the LaOCl spherical cap. These fibers extend towards the Si(100) substrate, and at the 60 min polymerization time the polyethylene layer reaches a thickness of 5 µm. Interestingly, the lower left corner of the 60 min image shows a LaOCl fragment in cyan that lies on top and within the magenta polyethylene fibers near the center of the spherical cap.

While the overall conclusions drawn from these PiFM images are similar to those drawn from the data provided by Raman microscopy, the significantly higher resolution does provide some additional insights into the morphology of these samples by quantifying the thickness of the polyethylene fibers much more precisely.

Results from PiF-IR

PiF-IR refers to IR spectra taken on a PiFM instrument. In this study, the authors found that for the 1- and 2-min ethylene polymerization times the doublet of peaks (1461 cm⁻¹ and 1471 cm⁻¹ absorption peaks associated with -CH₂– bending ) associated with crystalline polyethylene were not present. Instead, they just saw a broad amorphous band at 1463 cm⁻¹.

By performing multivariate curve resolution (MCR) analysis on the PiF-IR spectra they determined the fraction of the crystalline components that contribute to the spectra taken for each of the polymerization times (see figure below).

Based on these results, the authors conclude that there is a steep increase in crystallinity up until 10 min. Then, there is a saturation as the polymer forms an HDPE-like PE layer that is thicker.

Therefore, these PiF-IR spectra reveal a significant amount of information about the polymerization process that Raman imaging or even PiFM imaging alone could not provide. This impressive result is even hypothesized to be related to a conclusion based off the study’s ToF-SIMs data that is discussed later.

Results from FIB-SEM-EDX

Besides the chemical analysis of the polyethylene fibers performed using the vibrational techniques, the authors also used FIB-SEM to study the morphologies of the samples as a function of polymerization time (see below). In the FIB-SEM images, there is a clear atomic contrast between the low atomic weight polyethylene, the intermediate atomic weight Si(100) substrate, and the high atomic weight LaOCl framework, via the detection of backscattered electrons. Therefore, it is quite easy to differentiate these materials even without any chemical data, which was one of the original reasons LaOCI was chosen for this experiment.

Based on these images, the authors are able to observe how the spherical cap matrix fragments over time. In the initial stages, the polyethylene fibers are extruded from cracks on the surface where they begin to locally peel off some of the LaOCI. This happens mainly in the center of the spherical cap. At the edges of the matrix, the polyethylene rapidly fills internal macroporous cavities and cleaves the LaOCI from within. After about 10 to 20 min, both of these fragmentation models are observed throughout the spherical cap. Finally, at 60 min, they observe that internal fragmentation model becomes the dominate fragmentation pathway.

Results from XPS

In this study XPS was used to study the LaOCl surface, both in its pristine condition and after grafting on TiCl₄ but before the ethylene polymerization. By comparing these measurements to some bulk reference materials, the authors concluded that XPS gave evidence for the coordination of Ti⁴⁺ on the LaOCl surface. They also saw a minor presence of Ti³⁺ species as well. Therefore, XPS was critical for double checking their synthesis processes to make the spherical cap models, but it was not useful for studying the catalyst itself.

Results from ToF-SIMS

Time-of-flight secondary ion mass spectrometry (ToF-SIMS) was used by the authors to verify the formation of LaOCl during their synthesis of these spherical cap models. Additionally, this technique was used to observe the surface chemistry and confirm the formation of polyethylene at all polymerization times studied. This confirmation was done by measuring the distribution of negatively charged fragments of LaOCl⁻, TiOCl⁻, and polyethylene-characteristic C₂₁H₃₁⁻ via secondary electron images.

Additionally, AFM and ToF-SIMS analysis in Figs. S8 and S11 of the paper show, “a decay in the increase of the estimated polyethylene on the external surface after roughly 2–5 min.” They surmise that this decrease in the rate of polymerization could lead to a higher ratio between the rate of crystallization and the rate of polymerization, which is consistent with the observations made via PiF-IR spectroscopy and MCR analysis.

Summary

The authors of this paper were able to use their micro-spectroscopy toolbox to make some impressive observations about the polymerization of ethylene in this LaOCl catalyst matrix. Some of the techniques in their toolbox, like XPS and ToF-SIMS were most helpful during the synthesis of the LaOCl spherical caps. ToF-SIMS did help them understand the polymerization rate of polyethylene, but the most impressive results about the rate of crystallization were made using PiF-IR spectroscopy. Raman played a key role in the authors’ understanding of where the strongest formation of polyethylene was early in the polymerization times. However, the high resolution of the PiFM images were able to not only corroborate that finding but also show the structure of the polyethylene fibers themselves. They even quantified the thickness of the polyethylene layer via PiFM and the AFM. Besides the insights from these vibrational techniques, the FIB-SEM images played a key role in understanding the fragmentation behavior of the samples as a function of polymerization time. This was made easy due to the stark contrast of the atomic weights of their materials. However, without that advantage they would have likely needed to use chemical techniques like Raman or PiFM to observe the morphologies of different materials with similar atomic weights. Overall, these data show an impressive array of results from each technique, and the authors do an excellent job of highlighting the strengths of the tools in their micro-spectroscopy toolbox.

References

1. Bossers, K.W., Mandemaker, L.D.B., Nikolopoulos, N. et al. A Ziegler-type spherical cap model reveals early stage ethylene polymerization growth versus catalyst fragmentation relationships. Nat Commun 13, 4954 (2022).