Piezo nanopositioning – a key enabler of super-resolution microscopy

Thomas Bocher
Physik Instrumente (PI) GmbH & Co. KG, Auf der Roemerstrasse 1, 76228 Karlsruhe, Germany
Thomas obtained a Master’s degree and a PhD in physics from the Freie Universität Berlin before joining Laser-und Medizin-Technologie GmbH as a project manager. He was the founder and CEO of several start-up and sales and service companies, mainly in the medical devices industry, including Vivoscopic GmbH and Liko GmbH. Thomas has 10 years’ experience as a product manager and business unit manager in routine microscopy at Carl Zeiss Microscopy, plus seven years as business unit manager in compact magnetic resonance spectrometers at Bruker BioSpin. Since 2018, he has been the segment marketing manager for microscopy and life sciences at Physik Instrumente (PI).
The days of simply looking down the lens of a microscope are fast being eclipsed by superresolution microscopy that allows researchers to investigate and understand intracellular processes and interactions at a molecular level. During the past two decades, advances in the field have led to significantly higher resolved images, faster processing times and further insight into the functionality of molecular machines within cells. We discuss how nanometre precision positioning has been a major enabler for rapid advances in this field, and the importance of nanopositioning for superresolution microscopy applications.
Corresponding author
Dr. Thomas Bocher, Head of Segment Marketing for Microscopy and Life Sciences Physik Instrumente (PI) GmbH & Co. KG, Auf der Roemerstrasse 1, 76228 Karlsruhe, Germany
The development of super-resolution microscopy (SRM) has led to renewed interest in the use of optical microscopy in a number of areas of life sciences research, especially neuroscience and molecular cell biology. First proposed in 1978, this technique overcomes the so-called Abbe diffraction limit – described by German physicist Ernst Abbe over 100 years earlier in 1873 – allowing structures of less than 200 nanometres to be distinguished with fluorescence-based light microscopy. 
Although there was some development in the field over the following quarter of a century, it remained a niche technique until around 15 to 20 years ago, when researchers began combining a number of separate super-resolution technologies – including light sheet, 4Pi, STED (stimulated emission depletion), PALM (photoactivation localisation microscopy) and STORM (stochastic optical reconstruction microscopy) – to create systems capable of achieving spatial resolutions down to around 20 nanometres. This development led to Eric Betzig, Stefan W. Hell and William E. Moerner being awarded the 2014 Nobel Prize in Chemistry, and has greatly enhanced our understanding of many intracellular processes and molecular interactions.  
SRM allows cell biologists to non-destructively investigate cellular structures in previously unattainable detail, and its use has grown rapidly, particularly since the arrival of commercially-available super-resolution systems.
One of the main benefits of using these optical microscopy systems is that, unlike other methods offering comparable resolution – such as scanning electron microscopy (SEM) – they can be used to investigate metabolic interactions in living cells. The rise of SRM has therefore prompted many laboratories to revisit imaging-based research in areas such as cell division and intracellular transport – which had previously been hindered by diffraction-limited image resolution – by combining selective labelling with the ability to distinguish co-localised molecules of interest using spectrally-separated fluorescent dyes. 
As with many new research tools, the development of SRM technologies has been driven primarily by academia, with a majority of the systems in use today originating either directly from research labs or via university spin-off companies. The first generation of these instruments were, by necessity, built in house by physicists using standard parts from other microscope platforms.
Given the 10-fold improvement in resolution achieved by SRM techniques, the ability to position samples and objectives accurately was obviously a key requirement to be able to create such systems. Fortunately, piezo-based positioning systems – with nanometre resolution – were already commercially available when SRM development began, and were widely used by other high resolution imaging techniques such as atomic force microscopy (AFM) and SEM. As a result, piezo-based nanopositioners became a key enabler for the development of SRM techniques. 
Piezo – derived from the Greek word for pressure – technology was discovered in 1880 by brothers Jacques and Pierre Curie, after they identified that applying pressure to crystals such as quartz and tourmaline generated an electrical charge.
This effect constitutes conversion of mechanical energy into electrical energy – coined the ‘direct piezoelectric effect’ – in crystals and ceramics with non-symmetric elementary cells below the Curie temperature (Figure 1). The pair went on to discover that conversely, applying an electric field to certain crystals can deform the material, resulting in a change of its length – the inverse piezoelectric effect. 

Figure 1
. Perovskite crystal structure of PZT (1) Structure above Curie temperature, with Ti or Zr in non-piezoelectric, symmetrical cubic structure. (2) Structure below Curie temperature, showing the non-symmetrical tetragonal structure which allows piezoelectric conversion Image: PI Ceramic
The piezoelectric effects in natural monocrystalline materials, such as the afore-mentioned quartz and tourmaline, are relatively small. However, engineered polycrystalline ferroelectric ceramics – including barium titanate (BaTiO3) and lead zirconate titanate (PZT) – display much larger displacements or induce larger electric voltages. These piezoceramic materials can also be doped with ions – for example Ni, Bi, La, Nd or Nb – to optimise the piezoelectric effects and parameters, allowing each piezo component to be individually adapted according to the intended application, such as sample stage actuators or Z-focussing for microscope objectives. 
Two types of piezo actuators have become established – low voltage (up to 120 V) multilayer actuators made from ceramic tapes of 20 to 100 μm in thickness, and high voltage (up to 1000 V) actuators produced with 0.5 to 1 mm thick ceramic layers. Both have low energy consumption levels and, with static operation, they consume little power and do not generate heat when at rest, much like an electrical capacitator.
As the piezoelectric effect is related to electric fields, piezo actuators do not produce magnetic fields, nor are they affected by them. In simple terms, piezo actuators can be considered to be ‘spring and mass’ systems. The stiffness of the actuator depends on the Young’s modulus of the ceramic – around 40 GPa, which is 75 per cent less than steel – the cross section and length of the active material, and a number of non-linear parameters. 
Typical actuators have stiffnesses between 12 and 2,000 N/μm, and compressive limits of between 100 and 100,000 N, allowing them to have a high load capacity and making them insensitive to shock or vibrations. However, to deal with tensile stresses, an external preload spring, or casing with integrated preload, is required. Adequate measures must also be taken to protect the piezoceramic – which is both vacuum and clean room compatible – from torque, shear and bending forces. 
Piezoelectric actuators have long been considered the gold standard for the fine positioning of both sample stages and objectives in a broad range of microscopy techniques. The main advantage of this technology is that movement is controlled by converting electrical energy into mechanical motion. In theory, the piezoceramic material itself has unlimited resolution, allowing precise, stable and reproducible fine positioning down to the sub-nanometre range, and typical travel ranges between ten and a few hundred micrometres for linear actuators.
As these piezo materials have no moving parts, the resulting motion is completely friction free and not subject to wear, making them virtually maintenance free and capable of several billion cycles without measurable changes in their behaviour. 
Despite traditional light microscopes not requiring this level of precision, many of these products have been in widespread use for over 20 years. This is thanks to their combination of positional accuracy, high speed movement and improved stability compared to motorised drives. 
Piezo actuators offer extremely rapid response times – down to a few microseconds – and can achieve acceleration rates of more than 10,000 g. Combined with their direct-drive design and robust nature, this allows piezo ceramics to offer very fast start/stop behaviour, and velocities to several hundred millimetres per second. Many of the latest generation piezo flexure actuators combine 0.5 nanometre precision with very long travel – up to nearly 500 micrometres in some cases – and rapid settling times of less than 20 milliseconds, making them ideal for fast scanning and high throughput microscopy applications.
Conversely, piezoceramics can also be used to provide exceptionally high resolution motion at a low, constant speed – down to single encoder counts per second – in combination with suitable control and regulation processes. This proven track record across multiple microscopy disciplines has made piezo actuators the natural choice for laboratories developing the first generation of SRM platforms; they were both familiar to researchers and already offered the performance required.
It may seem serendipity that the piezo actuators already in widespread use offered resolution almost two orders of magnitude higher than was required for existing light microscopy applications, but the other characteristics of this technology – exceptional stiffness, excellent reproducibility and extremely rapid movement – made it the logical choice. The advent of SRM has now allowed optical microscopy to fully exploit the capabilities of these existing electromechanical elements. 
High-resolution microscopy techniques must be capable of accurate and reproducible positioning of imaging elements or samples. 
Applications range from precise Z-focus positioning of light microscope objectives in the depths of field, to exact positioning of tiny biological samples for 4Pi Resolft 2-Photon super-resolution fluorescence microscopy. Also in AFM, SEM and TEM, and correlative microscopy, versatile nanopositioning solutions greatly support even the most demanding positioning tasks.
Piezo actuators are currently used for three main applications in SRM: Z-positioning of objectives for autofocussing; manipulation of illumination systems (such as in light sheet and 4Pi techniques); and positioning of sample stages. Each of these applications obviously has unique requirements in terms of range of travel, reproducibility and settling times but, in general terms, system integrators want piezo-controlled movements to be as rapid, precise and stable as possible. 
The exceptional resolutions which can be achieved using SRM – and resulting decrease in the field of view – has naturally led to a significant increase in the number of images required to complete a full scan of even a single plane of an individual cell. For other techniques offering similar image resolution, for example SEM, this increase in the number of images required, and subsequently the time taken per scan, is not a significant issue, as they can only be used with dead cells or other inanimate materials, which will not change over the course of the scan.
In contrast, SRM allows metabolic processes to continue throughout image acquisition. The attributes, and resulting speed, of the piezo actuators therefore directly influence the quality of experimental results as, the faster a scan can be completed, the better it represents a ‘snapshot’ of a single time point. Add to this the ability to scan through the entire Z stack to create a 3D render of a cell or tissue, and any improvements that can speed up imaging – longer travel for sample stage movement, reduced drift, shorter settling times, etc. – become increasingly important. 
When performing quantitative image analysis, it is essential that users can bookmark the position of cells or structures of interest, then quickly and consistently relocate those same structures at a later time or on a different instrument. This is particularly important for longitudinal (time-lapse) studies to understand metabolic, molecular and information transport on a cellular level, as these processes often have very short timescales.
Bidirectional reproducibility of the positioning – the accuracy with which the system can return to a specific X and Y position – is therefore becoming increasingly important, and piezo-nanopositioners can reliably move the sample or objective to the correct place consistently and precisely. By allowing the motion of the stage to be measured directly with a capacitive sensor – without any influence from the drive or guide elements – this set-up ensures optimum repeatability, outstanding stability and stiffness, and fast-responding control.
Total internal reflection fluorescence microscopy (TIRFM) is a comparatively simple and inexpensive super-resolution fluorescence microscopy method that has been used for many years, and TIRFM systems – such as the GATTAscope (Figure 2) – benefit greatly from piezo-based solutions1. For this technique, a laser beam hits the glass surface of a microscope slide, microplate or Petri dish at a total internal reflection angle to stimulate fluorescence.

Figure 2
Example of a TIRF microscope – the GATTAscope from GATTAquant. Image: GATTAquant.
Because the electromagnetic field of this beam only penetrates the sample on the other side of the glass plate to a depth of 100 to 200 nm, the stimulated area in the Z plane is reduced, offering better Z resolution than standard wide field or confocal fluorescence microscopy. Piezo solutions enable the precise positioning of lasers and Z-drive that are key to successful super-resolution imaging in the Z-direction.
Similarly, light sheet fluorescence microscopy – as used by the Janelia research group2 – uses a unique planar illumination approach, orthogonal to the observation direction (Figure 3). This minimises photobleaching and phototoxicity for living specimens, allowing 3D and live cell imaging of large specimens over an extended period of time.

Figure 3
. Schematic representation of the IsoView light sheet fluorescence microscope.  
(a) Left: four orthogonal arms provide simultaneous light sheet illumination and fluorescence detection. The specimen is located at the centre of this arrangement, and volumetric imaging is performed by sweeping light sheets across the sample with objective piezo positioners. 
Right: close-up view of sample chamber showing objectives (O1-O4) and piezo positioners.
(b) In mode 1, four views are acquired by alternating light sheet illumination (I) and fluorescence detection (D) along orthogonal axes, with temporal separation used to avoid cross-talk between views. 
(c) In mode 2, illumination and detection are performed at the same time in all arms, and cross-talk is avoided via spatially matched beam scanning and confocal detection using a phase offset in orthogonal arms (spatial separation). 
(d) In mode 3, different illumination wavelengths are used in orthogonal arms simultaneously, and four views are acquired for both channels by alternating wavelengths (spectral separation).
Examining small organisms in this way requires samples to be positioned to within a range of a few millimetres in XY with a bidirectional repeatability of 100 nm, which can be achieved efficiently with piezo-driven stages and objectives.
The high throughput nature of piezo-based stages allows imaging of large-surface objects – for example high magnification image recording of objective slides and microplate wells – within seconds. During image recording, piezo-controlled stages precisely and continuously move the object, either in ‘stop-and-go’ mode, or at a constant velocity in ‘fly-over’ mode. It is also possible to combine this scanning process with real-time data handling and image-processing steps to ensure that even CPU-intensive tasks, such as the superimposed tiling processes, run with limited delay. This allows individual images to be merged seamlessly into the overall image even while acquisition is in progress, while continuously adjusting the focus during scanning.
Super-resolution microscopy is driving rapid advances in our understanding of many cellular processes and molecular interactions. The latest generation SRM systems are achieving spatial resolutions of around 20 nanometres, and the sub-nanometre positioning precision offered by current piezo-based drives will help to increase this resolution even further.
Despite this, microscope integrators have been demanding longer travel ranges, greater stiffness and higher speeds since the construction of the very first SRM systems, and the development of customised solutions to meet this growing demand will be vital as optical microscopy pushes further into the nanodimension.
1. Schmied, J. et al. DNA Origami Nanopillars as Standards for Three-Dimensional Superresolution Microscopy. Nano Lett, 2013, 13 (2) 781-785.
2. Chhetri, R. et al. Whole-animal functional and developmental imaging with isotropic spatial resolutions. Nat Meth, 2015, 12, 1171-1178.
Website developed by S8080 Digital Media