Filming the Ultrafast & Ultrasmall

by. Dr. Khalid Siddiqui

Moviemaking is an art. It can be an arduous task requiring a lot of planning and work at various levels as well being very time consuming and expensive. Many things must be just right to pull off a masterpiece: the lighting, audio, and the settings all have to create that special atmosphere into which we can immerse ourselves. However, it takes more than just production value and effects to make things perfect: the actors need to deliver also, not to mention that the script must be engaging to capture the imagination of the audience. These are the movies that we are all familiar with and love, and yet, there is a whole different world of moviemaking out there producing a unique class of “movies” which do not make it to the public theatres, and where art turns into science. These are molecular movies or in other words, movies that follow the motions of atoms and molecules as they occur [1]. Molecular movies are made in laboratories around the world equipped with the fastest cameras in existence: the molecular camera. The actors in these movies are atoms and molecules, the scripts are written by nature itself, and scientists merely act as camera crew. The field that encompasses these endeavours is known as the field of ultrafast structural dynamics and is concerned with imaging matter to understand it better and control it.

Structure-Function Relationship

The idea that structure defines function is a central one in many science fields including biology and material sciences [5]. It is one of the major motivations in ultrafast structure dynamics. An easy way to visualise and appreciate this dependence is by thinking about a paper clip. It is essentially a looped steel wire, and its distinct shape (structure) gives it its functionality: fastening paper sheets together. Moulding it into different shape, e.g. by straightening it out, changes this functionality.

Things are obviously a lot more complex at the atomic level where, depending on the system, a small local change could have large consequences which could be reversible or permanent. Ultrafast methods allow us to pinpoint what this relationship is for different systems.

Matter, or any substance that has mass and occupies volume, is made up of atoms (from Greek atomos for indivisible) — the fundamental building block. It exists in five different forms: solids, liquids, gases, plasma (a gas of ions), and Bose-Einstein condensates with the latter being made exclusively in scientific labs. Therefore, nearly everything in our physical world embodies matter and as such, the study of it is very important to understand and improve the universe in which we live in. For example, understanding how plants convert sunlight into chemical energy at remarkable efficiencies can help us to design highly efficient energy devices and potentially solve issues relating to energy economy. We can learn from mother nature and devise new ways to make new drugs that will create effective treatments for diseases such as Alzheimer’s or cancer. All of these are very worthwhile goals. But, where to begin?

The key starting point in the quest to understand and achieve these goals is to follow the dynamics, i.e. the evolution over time of the fundamental microscopic processes which are responsible for physical, chemical, and biological phenomena. However, as one can probably guess, it is easier said than done. There are two major fundamental challenges that need to be overcome.

First, we need to see the atoms! Atoms are extremely tiny: typical size of an atom is about one angstrom, 1 Å = 10-10 metre. They easily evade the human eye which cannot see or distinguish objects smaller than 1 micron or 10-6 metre [2]. However, microscopes (instruments which magnify very small objects to examine them in detail) can. Light microscopes have become so powerful that they can view objects as small as 20 nm (1 nm = 10-9 metre) [3]. However, that is still not enough, as the size of our target (the atom) is still at least 100 times smaller than what light microscopes are able to currently resolve. What prevents them from viewing smaller objects like atoms? It turns out that this is due to diffraction of light (diffraction is a phenomenon involving wave interference, a topic of discussion for another day) which blurs out features smaller than half the wavelength of light. This is known as Abbe diffraction limit [4], named after the German physicist Ernst Abbe. Simply put, the wavelength of the probing source must be a lot smaller than the size of object under investigation to provide detailed information about them, and what this means for our quest to observe atoms is that we need to overcome the Abbe diffraction limit to achieve atomic resolution. An easy way to remember this is to imagine putting a flea and tyrannosaurus rex (T-Rex) in a room and asking them to draw up a picture of each other in as much detail as possible. Obviously, the flea — due to its much smaller size compared to the T-Rex, will be the only one to succeed in the task whereas the T-Rex will not see the flea at all!

A T-rex and a flea attempt to draw a portrait of each other.

Similarly, visible light (400 nm to 800 nm) used in light microscopes with its wavelength on the order of hundreds of nanometres — which is at least three orders of magnitude (10n, where n is order of magnitude so when n = 3, 103 = 1000 times) larger than atomic size, is unable to “see” atoms. We therefore need a different probe which should have wavelength smaller than 1 Angstrom. It turns out that electrons and X-rays are ideal in this regard. Both have wavelengths on the order of tens of picometres (10-12 metre), with electrons featuring shorter wavelengths than X-rays. Electrons are charged particles that are commonly employed in transmission electron microscopes (TEM) to image materials and elucidate their physical properties. X-rays are photons (packets of electromagnetic radiation and have properties of light) which find uses in X-ray machines for medical scans, and in scientific research to study material properties.

Figure 1: The shutter speed must higher than the phenomenon being captured. The idea is illustrated here with an example: photography of a water fountain. (Left) Picture taken with shutter faster than the rate at which water sprinkles out of the fountain allowing each drop to be photographed. (Right) Picture taken with much slower shutter which smears out the individual water drops. Note that faster shutter means less light reaches the detector and so the picture appears darker. Image credit: David Strever Photography [6].

So, this solves one hurdle: we now have a tool that is sensitive to atoms! But what about the second challenge? The world of atoms and molecules is a dynamic one, so we must consider their motion. Even atoms in solid objects, like the pencil in your hand, are moving tiny distances. But, how long does it take for an atom to move approximately 1 Å in a solid? The speed of atomic motion can be approximated as 1 km s-1 (typical speed of sound in solids) and a quick, back of the envelope calculation gives us about 10-13 sec or 100 femtoseconds (see definition) for the time it takes for an atom to move 1 Å. This place an upper limit on how fast a detector we use to follow atoms needs to be to capture their displacement in real time (see Figure 1. Note that in TEM and other static microscopies, only average positions of atoms are obtained which still carry important information, but the dynamic information is lost). As we have just established, we need femtosecond-level time resolution to track atomic motions or to put in a different way: the camera to make our molecular movies must have an effective shutter speed on the order of 100 fs or shorter! Needless to say, no camera is fast enough to approach this target, and it was a big challenge (considered practically impossible) to be able to watch atomic motions until about a couple of decades ago [7]. Thanks to advances in lasers (highly coherent and directional light, more on them next time!) and leaps in ultrafast optics (field that works with application of ultrashort lasers) and particle accelerator technologies, it is now possible to produces extremely short bursts of X-rays and electron pulses lasting a duration of only a few femtoseconds that can be used to make movies of molecules. Either can be produced in university labs or in large facilities. However, femtosecond X-rays with high flux (in technical language brightness but the definition is more complex and goes beyond the scope of this article) are created in facilities know as free-electron-lasers (FELs or XFELs) [8]. FELs/XFELs can be very big (Figure 2) and expensive costing billions of dollars.

Figure 2: (Left) Linac Coherent Light Source (LCLS) free-electron laser facility located at Stanford, USA is one of the fastest molecular cameras around today. It utilises X-rays and the linear accelerator highlighted in blue is approximately 2 miles (3.2 kilometres) long. (Right) The High Repetition-rate Electron Scattering (HiRES) based in Berkeley Lab is an alternative molecular camera which takes molecular snapshots using electrons as “flash light” and is one of the most advanced apparatus of its kind around [12].  Image credit: SLAC & Berkeley Lab.

Now that we have a source that can see atoms and can keep up to speed with their movements as well, we can make a molecular movie! (just like the gif below!) With this capability at hand, it can be very interesting to also test how we can control the processes we are filming. For example, we might be interested in knowing how sudden changes in the temperature affect a protein structure or how applying pressure changes the way atoms behave in a solid. Developing these motion pictures of atoms and molecules are providing scientists with extremely valuable information about important phenomena in biology, chemistry, and physics. Things we have learned from ultrafast movies include the primary steps of vision (which involves a twist of carbon-carbon double bond in Rhodopsin — a pigment of photoreceptor cells in retina and occurs in a mere ~30 fs [9]), critical structural changes that could help explain the high-efficiency of the next-generation solar cells [10] and clues which may assist scientists in their chase for one of the holy grails in physics: room temperature superconductors [11].

A molecular movie showing film shots of ultrafast heating in Bismuth using ultrafast electron scattering/diffraction. A diffraction pattern is captured at different moments in time after the call for “Action” and contains atomic information. As with any photograph taken with a Kodak camera, the film containing the negatives needs to be developed. Likewise, the diffraction pattens are inverted via a mathematical process to produce familiar pictures of atoms.

Given the evidence at hand, there is no doubt that the next big feature to come out of the “molecular Hollywood” will further advance our understanding of fundamental processes that will help drive modern breakthroughs in technologies. While they may never win an Oscar (could land a Nobel or two!), these movies nevertheless are true masterpieces which shall remain timeless.  

Definition: What is a femtosecond?

Femtosecond is a unit of time. It is equivalent to a millionth of a billionth of a second or 10-15 seconds. To appreciate how short this time is, consider a simple exercise. Take a single sheet of A4 paper and let’s say that it corresponds to 1 second, so that if you cut it into two equal halves then each half will be exactly half a second or if you were to quarter it then each piece would be 0.25 seconds. Now imagine being handed a pair of scissors and being asked to cut one whole A4 sheet into one thousand trillion equally sized pieces. Impossible! There is no way you will be able to do that, but hypothetically speaking, if you do manage this impossible feat then each piece will be exactly a femtosecond, the timescale of the atomic world!

Discussion questions


[1]. J.R. Dwyer et al., Phil. Trans. R. Soc. A., 364, 741–778 (2006).

[2]. W. T. Welford, Advanced Light Microscopy. Vol. 1. Principles and Basic Properties  (Informa UK Limited, 1989).

[3]. S. W. Hell & J. Wichmann, Opt. Lett., 19, 780 (1994).

[4]. N. Gray, Nat Cell Biol., 11, S8 (2009).

[5]. R. Reis & I. Mores, Biochem Soc Trans., 47, 47–61 (2019).


[7]. L. R. Khundkar & A. H. Zewail, Annu. Rev. Phys. Chem., 41, 15-60 (1990).

[8]. P. Emma et al., Nature Photon., 4, 641–647 (2010).

[9]. P. Johnson et al., Nature Chem., 7, 980–986 (2015).

[10]. Wu et al., Sci. Adv., 3, e1602388 (2017).

[11]. M. Först et al., Phys. Rev. Lett., 112, 157002 (2014).

[12]. D. Filippetto & H. Qian, J. Phys. B: At. Mol. Opt. Phys. 49 104003 (2016).