RARAF –
New Developments
Our current single-particle microbeam system features an extensively upgraded spatial resolution of 0.8 μm (diameter) for the beam. Further work is continuing with the aim of obtaining a beam spot diameter of 0.5 μm. Current technology development focuses on adding and improving imaging techniques, the capabilities of our existing system and on developing new functionalities for the future.
Among these new developments are enhancements in microscopy, no-stain imaging, a new particle detector to be placed below the samples, higher throughput via point and shoot technology, and development of permanent magnet and x-ray microbeams, to be used as additional irradiation facilities.
The diameter of the beam spot for the electrostatic focusing system has continued to decrease with tuning of the focusing systems and development of new lenses, as can be seen in the table below.
The most recent development has been the installation of a compound quadrupole triplet. Two identical triplets 2.20 meters apart are organized in Russian symmetry. Their combined demagnifications have produced our first sub-micron beam spot! The process of adjusting the alignment of the lenses and tuning of the lens voltages is continuing with the aim of obtaining a beam spot diameter of 0.5 µm.
Years
Type of beam
Diameter (µm)
2001
Focused - single quadrupole quadruplet
11
2002-2003
Focused - single quadrupole quadruplet
3
2004-2005
Focused - single quadrupole triplet
3.5
2006
Focused - single quadrupole triplet
2.0
2007
Focused - single quadrupole triplet
1.3
2008
Focused - compound quadrupole triplet
0.8
Multiphoton microscopy, a laser-based, 3D imaging technique is now integrated into the Microbeam II endstation at RARAF. The multiphoton microscope was custom-designed around the Nikon Eclipse E600-FN research fluorescence microscope at the endstation and is intended for imaging cell dynamics in tissue and cell-culture samples following irradiation. Now operational, this multiphoton microscope is the first of its kind to be assembled onto a microbeam cell-irradiation platform.
Multiphoton excitation is a phenomenon where multiple photons, coincident in time and space, can act like a single photon having a superposition of energies from the incident photons. With two photon excitation, for example, two incident photons at a laser focal point can act like one photon of half the wavelength. A Chameleon (Coherent Inc.) tunable Titanium:Sapphire laser with a 705-950 nm wavelength range is the light source for the multiphoton microscope at RARAF. These infrared wavelengths are less photo-damaging and have greater penetration depth than single-photon excitation. These characteristics support exciting fluorophores at the laser focal plane, and when the imaging technique is coupled to a precision xyz-stage, optical sectioning in 3D tissue samples is possible.
Figure 1. Multiphoton microscope at RARAF for observing responses in cell cultures and in tissues after particle irradiation.
Pictured in Figure 1, the setup of the multiphoton microscope at RARAF includes a scanning laser source that is built into the endstation microscope. The laser path starts at the Ti:S laser, which is bolted to the optics table. The light path proceeds through a variable attenuator and a fast shutter before being directed vertically along the rotation axis of the microscope's pivot mount, which is used to switch between online and offline positions. A beam expander increases the beam to a size that will over-fill the back aperture of the objective lens. The platform attached to the microscope mount accommodates the scanning optics. These optics components point along an axis into the side of a custom-built extension tube for the CCD camera that is used during fluorescence microscopy. A pivoting mirror inside the extension tube selects between fluorescence microscopy or multiphoton microscopy. For multiphoton microscopy, this mirror guides the laser vertically down through both the microscope tube lens and objective lens.
Fluorophores in a sample absorb the laser light through multiphoton excitation at the laser focal point. As these markers fluoresce, signal light entering the objective is deflected towards photomultiplier detectors (H5783, Hamamatsu). Shown in figure 2 (below), signal detection is accomplished with photomultiplier tubes that are mounted to the side of the filter cube changer on the microscope. Control software monitors the correlation between the laser scan position and signal at the photomultiplier tube to produce the images.
An emission filter placed before a photomultiplier tube designates the wavelength range (window) of detected signal. With multiple fluorophores, color images can be produced using appropriate dichroic mirrors and emission filters. In figure 3 (below), the image of bovine pulmonary artery endothelial cells (FluoCells prepared slide #1, Molecular Probes) is an example of a color image produced by wavelength-specific detection. The cells in this image were stained with BODIPY FL phallacidin to label the filamentous actin (F-actin) and were counterstained with DAPI to label the nucleus. The objective lens used for imaging these cells was a Nikon CFI LU Epi Plan Fluor 50X objective, with 0.80 numerical aperture and 1.0 mm working distance. To develop this image, 20 to 30 single frames were acquired per color at a rate of 2 seconds/frame. A routine written in MATLAB® subsequently summed and combined the images into a 3D matrix for color display.
Figure 2. Fluorescent signal emitted by the sample is deflected by a dichroic mirror and collected by photomultiplier tubes
Figure 3. Multiphoton microscopy image of bovine pulmonary artery endothelial cells. (FluoCells prepared slide #1, Molecular Probes). Cell nuclei diameter ~10 microns.
With the attachment housing for two photomultiplier tubes, experiments using fluorescence resonance energy transfer (FRET) are possible. FRET optics sets are on hand for the following transitions: BFP > GFP, CFP > YFP, and Alexa 488 > Alexa 594. FRET experiments concurrent with cell irradiation are examples of monitoring post cell-irradiation dynamics using the multiphoton microscope at RARAF.
Quantitative Phase Imaging
Software-based Quantitative Phase Imaging (QPI) is now available for use in the Columbia microbeam laboratory as a non-interferometric approach to no-stain imaging. Reflected-light based images are now routinely obtained in focus and with the focus set slightly above and below the sample plane. These images are then used to approximately solve the light transport equation using the Fourier transform-based software from Iatia (Melbourne, Australia), a set of relatively new techniques that generate phase images and phase-amplitude images. The results are used to create a new 2-D map of the sample which is then used by the custom microbeam irradiation software to locate cells.
In some of the tests, some cells have been missed and there also have been false positives. In general, the quality of the images can be improved by careful tuning of the parameters for the approximate solution to the transport equation. However, to optimize the system for our regular, automated use, there is still work to be done in eliminating false cells and reducing missed cells. Several variables have been isolated and eliminated as causes: plating time, cell-type, cell phase, light color, cell growth surface, amount of medium (depth), percent of medium vs. buffer, and use of a cover glass. Continued efforts are aimed at finding a combination of these variables that may affect the images and at exploring other variables that have not yet been considered.
In the picture the QPI algorithm has been applied to normal human dermal fibroblasts under conditions conducive to rapid, automatic cell location.
Immersion-based Mirau Lens
To enable interferometry, we used high precision glass for the beamsplitter and the mirror. The surfaces interacting with the medium have a protective layer on top of the coating.
The new attachment has been tested as a water immersion lens. We confirmed that the two equal-arm light pathways are indeed restored and then provide interference fringes in the environment with sufficient contrast to perform the biological experiments. The acquired images are comparable to those that can be obtained using the commercial attachment that we used for preliminary testing.
Diagram of the immersion-based Mirau lens
The new objective will support rapidly and automatically locating the cell nuclei without stain. The "banding" in some of the examples shows the presence of both a slow drift in z-position and some very slight vibrational interference. Work is currently being done to compensate the vibrations of the work environment.
While it is possible that we ultimately will keep both forms of no-stain imaging in the endstation, QPI and IMI are being evaluated in competition with each other. Under consideration are: processing time, reliability, maintenance, and ease of use for the experimenter. QPI is more sensitive to the algorithm parameters, which makes Immersion-Mirau interferometry more reliable than QPI at this stage. At the same time, IMI is more sensitive to vibrations than QPI. QPI is a software solution and does not require introduction of fluids or cleaning, whereas IMI does. QPI does not require any additional equipment, while the immersion-Mirau interferometer requires the end-user to use a custom objective. Additional work will be done, especially on QPI reliability, sensitivity of IMI to vibrations, and the processing time for both of these modes.
Custom-built interferometric attachment with an additional window unit
Left: IMI image of normal human dermal fibroblasts taken with the interim immersion-Mirau lens and a neutral density microscope cube. Right: Same sample with a UV cube to reveal the nuclear stain.
For more details see:
Phase-based cell imaging techniques for microbeam irradiations. G.J. Ross , et al. Nucl Instr Meth B 241: 387-391 (2005)
Currently the RARAF microbeam irradiator delivers a precise number of particles by irradiating the sample and counting the particles traversing it, using a gas-based ionization chamber placed immediately above the cells. This dictates the use of thin samples (e.g., single cell monolayers covered by little or no medium). With the move to thicker samples, tissue irradiations and immersion microscopy, it becomes necessary to be able to detect the irradiating particles before they enter the sample as they may be stopped in it.
To this end we have developed a completely non-invasive, non-scattering single particle detector, which is placed in the beam line below the sample to be irradiated. The "Lumped Delay Line Detector" (LD2) is a non-scattering device based on a capacitive pickup detector, typically used for detection of ion clusters or highly charged ions within ion traps. The LD2 enables single particle irradiation of thick samples (tissue for example) by sub-micron beams. It maintains the attainable beam spot size, since it contains no material within the beam path and therefore does not induce scattering. It also obviates the current need for removing the medium from cells pre-irradiation.
During irradiation, cells to be irradiated are moved to the beam position using a high-resolution three-axis piezo-electric stage. When a collimated microbeam was being used, this was a necessary but relatively time-consuming method to position cells for irradiation. Unlike a collimated microbeam, a focused microbeam is not restricted to a single location on the exit window and therefore can be deflected to any position in the field of view of the microscope used to observe the cells during irradiation. Moving the beam to the cell position magnetically or electrostatically can be performed much faster than moving the stage.
Photograph of the Point and Shoot deflector coil.
We are developing a "Point and Shoot" targeting system for microbeam irradiation based on the wide-field magnetic split-coil deflector system from Technisches Büro Fischer (Ober Ramstadt, Germany). The deflector consists of four interlaced coils in a set of ferrite rings. The coils are paired to create two axes that are orthogonal to each other. When wired in series, a pair of coils on opposite sides will create a temporary dipole magnetic field, which deflects the beam in the direction of the right hand rule. A dual deflection amplifier, optimally matched to these coils, has been purchased from the same company to drive the coil. This system has been used for the microbeam facility at Gesellschat für Schwerionenforschung (GSI), Darmstadt, Germany. The deflector can move the beam to as many as 1000 separate locations per second (more than 5 times the speed of movement for the stage!), which will dramatically reduce the irradiation time.
It is possible to "scan" the beam into a rectangular area in the microscope view by using the two pairs of coils independently. The amount of current needed in each coil to deflect the beam to a specific location can be mapped and this mapping used as a look-up table for the target locations. The irradiation would then proceed by rapidly switching through the required sets of coil currents to steer the beam to the targets, irradiating each target with the time limited only by the beam shutter speed and beam current.
We have begun preliminary testing of the system on our permanent magnet microbeam (PMM) endstation using 5.2 MeV 4He ions. Early results show that we can deflect the beam accurately in both directions with a good linear correlation between applied coil current and deflection distance. Since the Point and Shoot system is located below the upper focusing quadrupole triplet, the deflection in the two directions is reduced by the demagnification of the focusing lens. The PMM system was designed for a demagnification ratio of 3 between the two directions in a single lens. As seen from the slopes of the curves in the graph, we measured a demagnification ratio of 4. The ideal system would be able to position the particle beam at any location in the image. While the reduction of deflection in one direction is a limitation to immediate application of this device, we are designing methods to overcome this issue.
4He beam deflection as a function of coil current in Amps (A).
4He ions from the accelerator are focused using a compound magnetic lens consisting of two quadrupole triplets. The first triplet is placed 2 m above the object aperture, with a second (identical) triplet placed 2 m above the focal plane of the first. Since each triplet does not have identical demagnifications in the x and y axes, the two lenses are rotated by 90° in the x-y plane so that a circular beam spot is obtained (Russian Symmetry). The object aperture was initially covered with a 1.8-μm thick Al scattering foil (since replaced by a phase space sweeper), used to eliminate any correspondence between angle and position for the particles in the beam. A limiting aperture is placed before the first triplet and inside the second triplet to reject ions which have very large aberrations.
The cells to be irradiated are placed at the image plane of the compound lens. The PMM is mounted on the original microbeam endstation consisting of a microscope with a particle detector mounted on one of the objective lenses and an x-y stage positioned by stepping motors.
Magnetostatic Lens
Two permanent magnet quadrupole triplets have been purchased from STI Optronics. The optimized lens, shown here, consists of two outer 4.25 cm long magnetic quadrupoles and an 8.5 cm long center quadrupole with inter-quadrupole gaps of 1.67 cm and a bore of radius 6.35 mm. It should be noted that such a small bore radius is rather difficult to obtain with standard electromagnets.
Beam Tests
Just prior to the decommissioning of the RARAF Van de Graaff accelerator in June 2005, we had attained a beam spot size of 20 µm. As a first step, the beam was imaged at the focal plane of the first quadrupole triplet, using a commercial CCD chip. The observed spot size of 50 x 150 µm was in good agreement with that expected from simulations.
As a second step, the beam diameter was measured at the endstation using the knife edge technique. The spot size was tuned by adjusting all magnets while maintaining Russian symmetry - in particular we tried to keep quadrupoles 1, 3, 4 and 6 at the same strength (A) and quadrupoles 2 and 5 at the same strength (B). The figure shows the theoretical and measured spot size and shape at the end station. While the general trends are very similar in both cases, the smallest spot size obtained experimentally was only 20 µm in diameter, two times larger than the theoretical prediction. Simulations have shown that this is probably due to residual high-order fields in the quadrupoles or due to misalignment.
In 2006, the magnetic quadrupole system had to be removed for the construction of the laboratories on the third floor and was reassembled in early 2007. Without adjusting the magnets, we measured a beam spot size of 20 microns, demonstrating the robustness of this design. Following the replacement of the scattering foil with a phase space sweeper and smaller aperture we have obtained a beam spot approximately 8 µm in diameter. The process is continuing with the aim of achieving a beam spot diameter of 6 µm or less.
For more details see:
Testing the stand-alone microbeam at Columbia University. Garty G., Ross G.J., Bigelow A., Randers-Pehrson G. and Brenner D.J. Radiat. Prot. Dosim. 122:292-296, 2006.
A single-particle / single-cell microbeam based on an isotopic alpha source. Ross G.J., et al. Nucl. Instrum. Meth.B. 231:207-211, 2005.
A microbeam irradiator without an accelerator. Garty G., et al. Nucl. Instrum. Meth. B 241:392-396, 2005.
Nearly all microbeam facilities currently employed for radiobiological applications use charged particles - from protons to heavy ions, with LETs (stopping powers) ranging from a few tenths to several hundred keV/µm. There are however considerable benefits in using soft X-ray microbeams for both mechanistic and risk estimation end-points. The higher spatial resolution achievable with modern state-of-the-art X-ray optics elements combined with the localized damage produced by the absorption of low energy photons (~1 keV) represents a unique tool to investigate the radio-sensitivity of sub-cellular and eventually sub-nuclear targets. Moreover, as low-energy X rays undergo very little scattering, by using X rays with an energy of ~5 keV it will be possible to irradiate with micron precision individual cells and/or parts of cells up to a few hundred microns deep inside a tissue sample. This will be of interest in investigating the relevance of effects such as the bystander effect in 3-D structured cell systems.
The X rays will be generated using an electrostatic quadrupole quadruplet lens system to focus protons onto a thick Ti target (best cross sections is at 4.5 MeV). The target will consist of a small plug of Ti pressed into a water-cooled copper block. The X rays generated will be demagnified using a zone plate. By using the already focused proton microbeam to generate characteristic X rays, it is possible to obtain a nearly monochromatic X-ray beam (very low bremsstrahlung yield) and a reasonably small X-ray source (~20 µm diameter), reducing the requirements on the zone plate.
Site developed by CE, page last modified by JL on April 24, 2008 .
Home | FAQ | History | Personnel
| Dissemination | Annual Reports | Experiment Scheduling | Directions | ContactMicrobeam | New Developments | Slow Neutrons | Charged Particles | Fast Neutrons | Schedule (PDF)