Will FIB replace TEM as king of the hill?
In my previous article, I discussed the ways in which the SEM and specifically FIB-SEM has improved in the past 20 years. In this time there has been massive uptake of the field-emission sources and more recently, stages that deliver a bias to decelerate primary electrons to sub-kilovoltages. The effect of these two things is to remove the need for conductive coating and offer theoretical sub-nanometer resolution of the most delicate features, meaning that they can compare with uncorrected TEMs.
A further area of notable improvement in SEM technology is the detector systems. In days gone by, you had a simple choice, secondary or backscattered electrons. Over the past few years it has become common to have in-lens detectors (SE and BSE) and a ‘normal’ SE detector in the chamber, frequently referred to as the SE2 detector. These detectors can and often are run simultaneously or mixed to give composite images of complementary information. If you include these detectors on a FIB-SEM and couple with a FE-SEM column and stage decelerator and monochromator (another trick to improve the energy spread of the electrons) you’re talking one serious piece of kit!! The cost of such an instrument? Well, don’t quote me, but I’ll guess at a little under 1 million GBP. Now that might sound like a lot, but look at what you’re getting - near nanometre resolution imaging, TEM lamella prep, serial sectioning for 3D data and the less fussy sample prep associated with FIB and SEM.
Okay, I hear you say, ‘so it can take pretty pictures and prepare TEM lamellae, but I still need a TEM to analyse the lamella’. Well not necessarily, because SEM and FIB-SEM instruments can be fitted with a STEM (scanning transmission EM) detector, also known as TSEM or transmission SEM detector. With this in mind, you may seriously start to wonder, ‘do I need a TEM anymore??’ Perhaps this was the thought in the minds of the management at Carl Zeiss Microscopy, when they recently announced their intention to stop producing TEMs altogether.
A schematic of an in-lens detector system, courtesy of Hitatchi
On paper, why would you spend £1.5m on just one TEM, when you can buy a FIB-SEM and HR-SEM for the same money and have two instruments rather than one? Let’s not forget the added advantage of the ease of use, ease of sample preparation from bulk or larger samples, and relatively shorter time to train new users, all of which are appealing to multi-user and interdisciplinary labs, where an instrument needs to be configured to be multi-purpose for a wide appeal.
Okay, maybe I’m getting a little carried away by my love of FIB. Am I saying that we don’t need TEMs anymore? Of course not! There will always be a need for the ‘king’ of electron microscopy, but I feel it will be in the category of high resolution and performance microscopy. Look at the trend, both JEOL and FEI offer corrected TEMs and from recent marketing pitches seem to be tailoring specialised instruments for both materials and life-science applications, which are specially adapted to deal with the demands of these areas with correction and automation to increase through-put. We must not forget Hitachi who offer TEMs adapted for biosciences and holography and a Cs (spherical aberration) corrected STEM instrument. All the top end instruments are encased in a protective enclosure, thereby isolating the column from sources of drift or interference and often can be operated remotely from a separate room if desired.
FEI’s Titan Krios comes in a range of specialised versions (Courtesy of FEI)
All in all, there seems to be a shift in thinking, as far as I see it. TEM is still the king, and always will be, offering the highest resolution using customised configurations in specialist fields to deliver the results that matter. However, the FIB-SEM (and FE-SEM) will dominate in all but the most demanding applications, where there is a need for fast answers, high throughput and multidisciplinary users. As we all know microscopes are tools to aid with the delivery of solutions and provide answers to questions, and we have to know which tool to use for which job without over-complicating things.
What do you think? Let me know and check back soon for another post.