Secondary Electron Energy Matters
A Key to Probing Polymers Nano-Morphology?
- Fig. 1: Spectra from spider silk force reeled at different speeds: (a) Raman, (b) SE, (c) correlation between Raman and SE for Raman peak at 830 cm-1
- Fig. 2: Cryo- fractured silk reeled at 5 mm/s: (a) SEM image Ep = 1keV, (b) intensity profile across lipid layer (appears in orange in false color inset)
- Fig. 3: SE images of PA 12, Ep = 1keV, using different SE energy ranges: (a) 4.9-5.9 eV crystalline lamella brightest, (b) 5.9-7.4 eV disordered matrix brightest
Secondary Electron (SE) Imaging in Scanning Electron Microscopes is fundamental to Materials Science including polymers - yet it lacks nanoscale chemical information, unless samples are stained. Change comes when sets of images using SEs with defined energy ranges (Hyperspectral SE Imaging) can be collected to build SE spectra that allow identification of particular energy ranges in terms of chemistry/polymer morphology either by modelling or comparison with other spectroscopy techniques.
Polymer nano-morphology determines the functionality of polymeric materials systems. In particular, many spider silks deliver outstanding mechanical performance but even after decades of research there is still insufficient knowledge of the silk nano-structure, because polymer morphology is complex and hierarchical. Silks and synthetic polymers (e.g. Polyamide (PA)) alike comprise structures that can arrange in different ways, including beta-sheet layered structures and disordered domains. The local arrangement of these strongly influences the properties because it will determine how such structural elements can interact. Yet, the ability to see such nanoscale arrangements in materials (not thin electron transparent sections) is limited because characterization tools tend to excel either in chemical identification (e.g. Raman or Infrared Spectroscopy) or fast nano-scale imaging (e.g. Scanning Electron Microscopes (SEM)). Is it possible to combine the best of both? After all SE have spectra, but they have been little exploited because they are not routinely collected on polymers, and there is no well developed theoretical framework that explains the spectra. How can this be changed?
How Can SE Spectra of Beam Sensitive Materials such as Spider Silk Be Collected?
SE spectra collection did not focus on polymers where a high detection efficiency is crucial due to the potential of beam damage. Fortunately, chemical signatures appear when the electron incident beam energy, Ep, is ~1keV, at which some SEMs can obtain sub-nanometer resolution  and charging can be minimized such that metal coatings can be avoided .
This is crucial as such coating would mask the spectral information of the polymer. Without the standard practice of applying a metal coating, beam damage must be managed by choosing appropriate Ep and scan strategies [3,4] available in modern field emission SEMs. These often feature detectors with through the lens or in-lens designs . allowing for energy selective detection [4, 5]. The SE spectra for spider silk (fig. 1) were collected in a Helios Nanolab 660 3G UC using automated collection with the i-FAST software . Cryo-SEM samples were prepared using a Quorum PP3010T attached to Helios NanoLab: A Nephila edulis silk bundle of a defined reeling speed kept under tension by the laser-cut frame was applied along a thin silicon support with carbon adhesive tape on both ends. The silk bundle (approx. 50 fibres) on the silicon support was covered by a cryo-glue/graphite suspension before another silicon support was applied. Fibre ends were cut and fibres mounted upright on a stub before submerging in liquid nitrogen slush and transfer to the PP3010T preparation chamber held at -180°C where the fibres were cleaved.
Nephila dragline silk was supplied by the Oxford Silk Group (Oxford University Department of Zoology) obtained from a single specimen, in a single force reeling session.
Raman spectroscopy was performed on single Nephila silk fibres using a Renishaw inVia Raman microscope (785 nm, 50 mW power).
Making the Most of Complex SE Spectral Data – Example Spider Silk
Figure 1a shows the Raman spectra and figure 1b the SE spectra of Nephila edulis spider silk, when reeled at different speeds. Clear differences exist in both types of spectra, yet Raman spectra exhibit many more distinct peaks, which respond differently to change in reeling speed. The Raman spectra of proteins contain vibrational bands due to the polypeptide backbone and amino acid side-chains which can be affected by molecular orientation . Molecular orientation and crystallinity of polymers is also expressed in SE spectra . Thus identifying spectral regions in Raman and SE spectra that show similar behavior with reeling speed could shine some light on the changes occurring in the silk. For example, the changes in Raman intensity at 830 cm-1 (amino acid Tyr, present in core ) correlates with the SE intensity changes at 5.1eV (fig. 1c). This is also consistent with silk from a different species (Nephila inaurata), where the energy region >3 eV was dominated by SE emission from the silk core, as established by plasma etching and SE spectroscopy, whereas the outer lipid layer shows predominately SE emission <3eV . The increase in dominat peak energy from <3 eV to > 3eV with reeling speed (fig. 1b) can be understood when considering that SE spectroscopy is a surface sensitive technique- similar to the better known Auger Electron Spectroscopy but requiring much less beam current. Our results imply that the protein signal becomes more prominent at higher reeling speeds, potentially due to a thinner coating layer allowing the protein SE signal to ‘shine through’. Although the thickness of the lipid coating with reeling speed has not been investigated, Atomic force microscopy results do show a clear change in the surface structures with reeling speed . A rough estimate of the thickness of the outermost silk layer is obtained from the cryo-SEM image in figure 2a. Here the outer layer, protrudes from the fracture plane of the rest of the fibre and it’s thickness can be estimated to be <30nm (fig. 2b). As the SE emission at 5.1eV is very weak in the SE spectra taken from spider silk with the same reeling speed in plan view geometry (fig. 1b), 30 nm is an upper limit for the information depth at this energy. A systematic study of combined cryo-cross sections and SE spectra collection in plan view on a larger set of samples and reeling speeds could enable layer thickness measurements from simple plan view SEHI image sets such as published previously .
Using SEHI for Imaging - Example PA
SEHI reveals different phases in synthetic polymers as has been shown on the examples Poly(3-hexylthiophene) , Polypropylene  and PA12 (fig. 3). The SE spectrum (fig. 3a) has several peaks. The morphological origin for the contrast that is linked to these peaks is investigated by extracting images from the same stack from which SE spectra were derived. Thus, images are noisy (to keep the dose during stack collection low) but show that the first peak relates to the distribution of crystalline lamella (fig. 3b) and the second peak relates to the disorder matrix (fig. 3b). Once suitable energies regions are determined, higher quality images can be collected using these regions only.
Conclusion and Outlook
SEHI can deliver information for polymer/protein assemblies and can provide insights to nanoscale morphology in these materials that do affect mechanical properties , or electronic properties . Yet, only a narrow range of instrumentations has been shown to be able to deliver the required spectral and image resolution. We expect this instrument base to broaden which will lead to wider accessibility. Another challenge is the full interpretation of spectra – which should be achievable by more extensive datasets and in combination with machine learning. Given the vast progress in the latter in the field of electron microscopy and spectroscopy, it is likely that SEHI can become a key to probing nano-morpholgy in the not too distant future.
The authors thank Fritz Vollrath and Alex Greenhalgh (Oxford University, Department of Zoology) for providing the spider silk. CR, CH, and NS thank EPSRC for funding (EP/N008065/1, EP/K005693/1, studentship 1816190), K.A. thanks EPSRC for funding as part of the EPSRC MAPP Hub (EP/P006566/1).
Cornelia Rodeburg, Nicola A. Stehling, Kerry J. Abrams, Chris Holland
Department Materials Science & Engineering, University of Sheffield, UK
Prof. Dr. Cornelia Rodenburg
Department Materials Science & Engineering
University of Sheffield
 Laurent Roussel, Debbie J. Stokes, Ingo Gestmann, Mark Darus, Richard J. Young: Extreme high resolution scanning electron microscopy (XHR SEM) and beyond, Proc. SPIE Scanning Microscopy 2009, 73780W: 1-9 (2009) doi: 10.1117/12.821826
 J.H. Butler, D.C. Joy, G.F. Bradley, and S.J. Krause, 1995: Low-voltage scanning electron microscopy of polymers, Polymer, 36(9): 1781-90 (1995)
 Deborah L. Vezie, Edwin L. Thomas, and W. Wade Adams: Low-voltage, high-resolution scanning electron microscopy: a new characterization technique for polymer morphology, Polymer 36(9): 1761-1779 (1995)
 Robert C. Masters, Nicola Stehling, Kerry J. Abrams, Vikas Kumar, Martina Azzolini, Nicola M. Pugno, Maurizio Dapor, Andreas Huber, Philip Schäfer, David G. Lidzey and cornelia Rodenburg: Mapping Polymer Molecular Order in the SEM with Secondary Electron Hyperspectral Imaging. Advanced Science 6(5):1801752 (2019) doi:10.1002/advs.201801752
 Quan Wan, Kerry J. Abrams, Roberts C. Masters, Abdullah C.S. Talari, Ihtesham U. Rehman, Fredrik Claeyssens, Chris Holland, Cornelia Rodenburg: Mapping Nanostructural Variations in Silk by Secondary Electron Hyperspectral Imaging, Advanced Materials 29(47): 1703510 (2017) doi: 10.1002/adma.201703510
 Thierry Lefèvre, François Paquet-Mercier, Jean-François Rioux-Dubé, Michel Pézol: Review Structure of Silk by Raman Spectromicroscopy: From the Spinning Glands to the Fibers, Biopolymers 97(6): 322-36 (2011) doi: 10.1002/bip.21712
 Nicola A. Stehling, Kerry J. Abrams, Chris Holland, Cornelia Rodenburg: Revealing Spider Silk's 3D Nanostructure Through Low Temperature Plasma Etching and Advanced Low-Voltage SEM, Frontiers in Materials 5: 84 (2019) doi: 10.3389/fmats.2018.00084
 Ning Du, Xiang Yang Liu, Janaky Narayanan, Lian Li, Matthew Lek Min Lim, Daiqin Li: Design of Superior Spider Silk: From Nanostructure to Mechanical Properties, Biophysical Journal 91: 4528–4535 (2006) doi: 10.1529/biophysj.106.089144
 Kerry J Abrams, Quan Wan, Nicola A. Stehling, C. Jiao, Abdullah C. S. Talari, Ihtesham Rehman, and Cornelia Rodenburg: Nanoscale mapping of semi‐crystalline polypropylene, physica status solidi c 14(12): 1700153 (2017) doi: 10.1002/pssc.201700153