Frustrations of a Physical Chemist in the AFM World
Usually the SPM blogs I write are more of the “hooray for AFM” variety where I focus on all the wonderful capabilities and new information that AFM does provide. And indeed, there are many. But lately, as I was working with a fellow SEM’er side-by-side, I become frustrated – even jealous - as I was reminded of a well-known gap in AFM capability. This frustration comes about because although in my “day” job I am primarily a user and researcher of atomic force microscopes, at heart I am still a physical chemist.
After all, I entered this field by obtaining chemical information with a scanning tunneling microscope (STM). Specifically, my graduate research was on unambiguous identification of chiral molecules using STM contrast. Identity and separation of chiral molecules has been a hot area of chemical research for decades as these molecules are isomers that are non-superimposable mirror images of each other (e.g. like your hands), but can have very different functions, which is especially critical in areas like the pharmaceutical industry. The famous example from the 1960’s is thalidomide, which was widely prescribed to treat morning sickness in pregnant women. The problem was that only the right handed (R) isomer was effective, while the left-handed (S) isomer induced severe birth defects. Pharmaceutical companies to this day spent tremendous resources on identifying and separating chiral isomers.
We identified “chemical marker groups” or chemical functionalities that appeared with particular STM tunneling contrast so that we could unambiguously identify functional groups on a molecule, and thus identify its chirality directly from the STM image. In our case, the functional groups that we worked with in the molecule 2-bromohexadecanoic acid were bromine (bright contrast), carboxylic acid (dark contrast), hydrogen (small yellow round feature), and the hydrocarbon backbone. But the problem is chemical marker groups are applicable to only a handful of functionalities – not exactly widely applicable.
STM image of 2-Br-hexadecanoic acid mixed with hexadecanoic acid
Back to my SEM study….We were studying a metal surface with some nanometer -sized flakes (both in diameter and thickness) and had no idea what those flakes were. We could pretty quickly get the size and even morphology of these flakes, but where did they come from and how did they land on this surface? Well, first my colleague just collected a back scattered electron (BSE) image with not much more than a click of a button to confirm that indeed the flakes were a different chemical composition than the metallic substrate – very useful information. But then he was able to switch to his energy dispersive x-ray spectrometer (EDS) which simply detects the emitted x-rays that result from the impinging electron beam. Because different elements emit x-rays at very specific energies, this information is elementally very specific. And voilà, we were able to quickly ascertain that the impurity flakes were in fact Nickel. AFM would have never been able to provide such information. Feeding this information back to the customer, they were able to figure out where the flakes came from.
That chemical information remains an obvious, and rather gaping hole, in AFM technology. Perhaps it’s unfair since AFM is based on an inherently mechanical contact and so there is no obvious way to extract any chemical information. We are very good at extracting loads of other kinds of information – mechanical, electrical, magnetic, and even thermal – but we come up short in anything that is chemically identifiable. And as a chemist, that’s pretty darn frustrating. How awesome would it be just to get a tiny amount of chemical information – differentiating elements would be great! Identifying elements unambiguously would be really great! Bonding information of molecular species fantastic! I know that x-ray based techniques or techniques that generate x-rays have an inherent advantage in this world because elements detect x-rays at specific energies, but can’t a chemist still hope?
Sure, there are “hybrid” technologies that try to marry AFM with other techniques to fill in that gap. Probably the most common one at this point is AFM-IR where infrared spectra (organic molecular bonding information) can be obtained at localized points on the surface using the AFM cantilever as a platform either through a photothermal method or a near-field scattering method. Raman is another technique that has been successfully married to AFM where you can collected localized AFM and confocal Raman images on the same area (for higher resolution you can try tip-enhanced Raman spectroscopy, which I would only recommend for the most expert AFM users). And there are instruments now that are hybrid AFM-SEM instruments, where one of the goals is specifically to add on EDS capability to AFM imaging.
But these hybrid techniques are not trivial to operate. The sum of a difficult technique plus an easy technique still results in a more difficult technique. The sum of 2 difficult techniques is, well, a nonlinear operation resulting in a really difficult technique.
So yes, the physical chemist in me is envious of my SEM colleague and the ease of which he can point and shoot and get an unambiguous spectrum of x-ray intensity of a spot on the surface. Or even get a map based on chemical composition. We’ll just have to add this to the list of where AFM needs improvement….
On the other hand, which I save for a future blog, AFM has a lot to offer the synthetic chemist. Impressive work from labs like that of Leo Gross of IBM has delighted us for the past decade with spectacular sub-molecular resolution images of molecules and even perform in situ reactions. However for better or worse, the synthetic chemist in me remains a very small component…..