1. INTRODUCTION

Experiments at synchrotron radiation facilities are often short campaigns requiring around-the- clock operation where several samples are usually investigated. It is, therefore, important to reduce the time to align newly introduced samples with the X-ray beam as efficiently as possible to optimize the time spent measuring with the beam. Particularly in the case of small samples, this task is not trivial and it is not uncommon that a significant proportion of beamtime is taken up by sample alignment through scanning of the sample.

In a protein crystallography sample, viewers are commonplace, allowing for automation of the sample change and sample alignment [1]. Traditionally, such cameras would observe the sample in a non-coaxial geometry resulting in uncertainty of alignment in one direction of the plane at right angles with the X-ray beam direction. More recently, on-axis or coaxial sample viewers have been introduced that use mirrors placed in the beam to locate the camera itself at right angles with the beam [2–5]. State-of-the-art, on-axis viewers use a microscope objective and mirror that have central hole so that a large viewing magnification with a short working distance can be achieved [6]. The limitation of such a setup is that there is a fixed field of view. A commercially available system overcomes this limitation by using a dual camera setup with different magnifications for each camera (OAV B-ZOOM—[7] and OM-1/2, Suna-Precision GmbH, Hamburg, Germany).

For these in-situ devices, the mirrors (and objectives) feature a small central hole to allow the X-ray beam to pass the mirror, which is placed at an angle of 45° with the X-ray beam unperturbed. The image of the sample taken with this setup resembles that of a system with a backstop that removes all on- and near-axis rays. A further disadvantage is that the drilled mirror acts as a fixed X-ray beam aperture. With typical X-ray spectroscopy measurements it is beneficial to expose samples with varying beam sizes to average larger sample volumes or to reduce the local radiation dose. A further requirement is to have flexibility in the exact sample position along the beam path such as to accommodate samples of variable size but also to allow samples housed in various types of environments and in operando equipment. Using a drilled microscope objective fixes the position of the sample.

Here, a flexible on-axis sample viewer is described that uses off-the-shelf components that can stay in the beam during experiments or be remotely removed. The use of a motorized zoom lens with a long working distance addresses the requirements for different size samples. Placing this lens with a long working distance on a motorized linear stage provides flexibility in the sample position along the beam path. The sample viewer is suitable for aligning samples with sizes about 5 µm and larger. I have investigated two options for the mirror that allow the camera to be placed well away from the intense X-ray beam. I show that the drilled mirror used by others results in a reduced image quality at low to moderate magnification settings. Such imaging issues are avoided by the use of a thin foil coated with a reflective layer (e.g.,  a Pellicle mirror/beam splitter) rather than a thick glass mirror. These thin foils with coating are transparent for X-rays, which removes any limitations in the X-ray beam size that can be used or blind spots due to a hole in the mirror.

2. SAMPLE VIEWER SETUP

The on-axis sample viewer presented is composed of a modern high-resolution networked camera with a long working distance zoom lens. Although the setup was developed for beamline BM14 at the European Synchroton Radiation Facility (ESRF) in Grenoble, France, it is stressed that it could equally serve well at another synchrotron beamline. A specific design choice was to allow for a flexible sample setup in which a range of sample holders and sample environmental cells can be used without changes to the sample viewer. BM14 is a beamline at the ESRF constructed specifically for X-ray spectroscopy experiments where users can conduct in-situ and in-operando experiments with a range of sample environments. For each of these experiments the distance of the sample to the first ion(ization) chamber is slightly different. The beamline provides focused X-ray radiation with a flux of approximately ${{1}}{{{0}}^{12}}\;{\rm{photons}}/{\rm{s}}$ over an energy range of 4–60 keV. The focal spot size can be varied with a minimal spot size of about ${0.05} \times {0.02}\;{\rm{mm}}^2$.

The overall setup is shown in Fig. 1. A granite tabletop that is movable with five degrees of freedom acts as a common base for sample, ion chambers and fluorescence detectors. The three ion chambers are mounted on a central, low profile LINOS FLS-95 rail ([8]) that is mounted in the middle of the granite table. Just upstream of the first ion chamber is a remote-controlled slit system AT7-F7-AIR ([9]).

 

Fig. 1. Experimental setup for X-ray spectroscopy at beamline BM14. The X-ray beam is shown as a green line and travels right to left. The second slits (1), reference sample (2), fluorescence detectors (3), sample (4), on-axis sample viewer (5), first ion chamber (6), first slits (7), common LPS95 rail (8), and granite tabletop (9) are indicated.

Download Full Size | PDF

The granite tabletop is aligned by first passing the beam (denoted by a green line) through the first set of slits (which are set to a small gap in the middle of the slit system) using movements in the $x$ and $z$ directions. The second step is to align the beam with the center of the second slit system, which is placed just upstream from the third and last ion chamber using rotations of the table around the center of the first slit system (yaw and pitch). The ion chambers are held by mechanical supports that are adjusted such that their center coincides with that of the two sets of slits when placed on the rail. Once the two-step alignment procedure is complete, the setup generally only needs small adjustments by translations of the granite table in the $x$ and/or $z$ directions. A complete realignment is only necessary if a setting of the total reflection mirrors has changed or when there is a change of the electron beam position in the synchrotron.

With the beam aligned along the rail and passing through both slit systems, the position of the sample and reference is aligned by a combination of a motorized high load $x$ and $z$ motion stage. The sample and reference rotation stage shown in Fig. 1 are optional. With the beam aligned and passing through the center of the ion chambers, the alignment of the newly introduced sample is simply a matter of moving its center along both $x$ and $z$ directions until it coincides with the beam.

The introduction of a sample viewer that looks on axis with the beam toward the sample is an ideal way to achieve fast sample alignment with the aid of the sample image. The sample viewer is mounted at the downstream side of the first ion chamber with its viewing axis carefully aligned along the axis through the center of the two slit systems. As such, it occupies about 40 mm of  beam path just downstream of the first ion chamber.

3. IMPLEMENTATION

The sample viewer is depicted in Fig. 2(a) showing its base plate, camera, zoom lens, mirror, kinematic mirror mount, and the pneumatic mirror retraction mechanism. A schematic showing the drilled mirror, X-ray path, and zoom lens is provided in Fig. 2(b). A detailed technical drawing and parts list is provided in Supplement 1.

 

Fig. 2. Sample viewer overview using a Pellicle beam splitter. (a) X-ray beam is shown as a green line traveling right to left. The GigE networked camera (1), System 6000 zoom lens (2), focus motor (3), zoom lens carriage (4), base plate (5), pneumatic cylinder (6), kinematic mirror mount (7), and Pellicle beam splitter (8) are visible. (b) Schematic cross section (side view) of the optical arrangement showing the camera (1), the Navitar zoom lens (2), and the drilled mirror (3) placed at a 45° angle with respect to the X-ray beam. The variable positioning of the zoom lens (indicated by large arrow) allows a flexible position of the sample whereas the motorized zoom function (internal movement of two lens sections) is achieved by a second stepper motor to give variable sample view magnifications.

Download Full Size | PDF

The ${6.5} \times$ zoom lens is part of the modular Zoom 6000 lens system (Navitar, Rochester, NY, USA). Depending on the lens attachment and the specific lens adapter, the working distance of the optics lies in the range from 36 to 300 mm. Since the right angle mirror will occupy a 40 mm beam path and the sample position will typically be at a distance between 40 and 100 mm from the mirror, the minimal working distance of about 90 mm should be accommodated. In Table 1, two useful working distances are listed, each with two combinations of the adapter and lens attachment (${0.5} \times$ and ${1.0} \times$). For flexibility in the exact sample position as measured along the beam, the zoom lens is mounted on a motorized linear stage.

Table 1. Zoom 6000 Lens Characteristics (Low–High Indicates the Zoom Range)

View Table

Table 1 shows the key data for the various configurations that are available for the modular zoom lens that are useful for this application. The zoom feature of the lens allows users to dynamically set the field of view. I have opted for a zoom lens with a working distance of 175 mm (second option in Table 1 indicated by italics). Of the 175 mm distance, approximately 45 mm is taken up by the tilted mirror, leaving a distance of about 130 mm that corresponds to the maximum distance between the mirror and the sample. Because the zoom lens is movable on the other side of the mirror, the sample can be placed flexibly up to a distance of 130 mm from the mirror. This range typically leaves enough distance for fairly large environmental cells such as cryostats, flow cells, and ovens.

The zoom lens is coupled with a Basler acA1440-73gc color GigE networked camera through a C-mount adapter. This type of camera uses a Sony IMX273 sensor that features ${{1440}} \times {{1080}}$ (1.6 Mpixel) 3.45 µm square pixels. I have used the setup with a working distance of 175 mm, ${0.5} \times$ lens attachment, and ${1.0} \times$ mini adapter. In this arrangement, the system is zoom lens limited since the matching camera pixel size is larger than that of the camera (see Table 1). With the high zoom setting, features down to approx. 5–10 µm can be distinguished. The depth of field ranges between 0.41 and 4.1 mm, which is a suitable range in which to locate the sample in the fixed focal point of the beam by sliding the sample along the common rail until it is sharply focused on the sample viewer.

To achieve on-axis viewing, two implementations were investigated. The first consists of a mirror that has a 1 mm diameter hole drilled at a 45° angle such that the X-ray beam can pass through its center fully unobstructed.

The second option is to use a Pellicle mirror that consists of a thin membrane with a metallic coating such as aluminum on one side that is transparent for the X-ray beam. The membrane is stretched and held by a support frame. In this case, I opted for a CM1-BP145B1 Cube-mounted Pellicle that has a coating suitable for the visible range (400–700 nm) of the electromagnetic spectrum with 45:55 reflectance:transmission ratio.

The setup allows a quick change between the two mirror options without loss of alignment due to the fact that a kinematic mirror mount is used. Finally, the mirrors are mounted on a linear rail and connected to a pneumatic cylinder that is able to remove and reinsert the mirrors from/into the X-ray beam without loss of alignment.

4. ALIGNMENT AND USE

The setup is readily integrated into an existing X-ray spectroscopy facility and takes up a small section along the beam path. Initial alignment of the sample viewer axis with that of the X-ray beam follows a few simple steps that only have to be done once and are outlined below.

With the X-ray beam path defined using the two sets of X-ray slits, the ion chambers are adjusted in height to have their centers coincide with the path of the X-ray beam. The sample viewer’s drilled mirror is aligned using a visible laser that is aligned with the X-ray beam. Before the lateral position of the complete sample viewer is set, the relative lateral position of the mirror assembly is aligned so that the hole in the mirror is as close as possible to the center of the obtained camera image. This alignment is achieved by adjusting a blocking stop of the pneumatic cylinder.

 

Fig. 3. Sample viewer image taken with the Pellicle mirror of fine strands of fiber (diameter approx. 10 µm). The X-ray beam position and size is denoted by the overlaid green circle.

Download Full Size | PDF

Changes of the pitch and yaw of the $1^{\prime \prime}$ diameter mirror are accomplished by one of two methods. Without the X-ray beam, alignment is obtained by moving the second (downstream) slit system to the sample position, which should lie within the maximal working distance of 135 mm of the sample viewer and bring the center of the slit aperture into the center of the camera image by adjusting the two mirror rotations using the 2 µm that are part of the mirror mount (KCB1C/M, Thorlabs, Inc., Newton, NJ, USA). The second method uses the X-ray beam and a small amount of fluorescent material placed in a glass capillary at the aligned sample position. Using the pitch and yaw adjustments of the mirror, the image of the fluorescent light is centered.

 

Fig. 4. Effects of the drilled mirror on sample imaging for three different zoom settings corresponding to fields of imaging of (a) 11 mm, (b) 7 mm, and (c) 2 mm as measured in the vertical ($z$) direction. Insets show the measured red, green, and blue pixel intensities along a horizontal line through the center of the image.

Download Full Size | PDF

Alignment of the sample viewer with the Pellicle mirror is similar to that of the drilled mirror with the exception that there is no requirement to align it in terms of pure translations since there are no drilled hole that must be coincident with the beam.

Figure 3 shows two images taken with the sample viewer using the Pellicle mirror with an overlay that indicates the beam position (circle). This overlay is generated in software and can be adjusted by the user if the exact position of the X-ray beam changes. Repeated sample alignments using the sample viewer gave an uncertainty in positioning better than 2 µm as measured with a separate microscope that was posited downstream of the sample and carefully aligned with the X-ray beam.

In the case of the drilled mirror setup, the central nonreflecting part of the mirror acts as a beam stop. As a consequence, the quality of the image taken from the sample changes significantly with the field of view. To illustrate this, I have taken images using three different settings of the zoom lens using a piece of white paper at the sample position. Illumination of the object was achieved by a white LED light source.

For the largest field of view, the intensity at the center of the image is reduced significantly [see Fig. 4(a)]. As the field of view decreases to a size similar to that of the drilled hole, the image shows a uniform intensity distribution. The latter situation resembles that of dark field imaging in which direct reflected light from the sample is blocked and only light scattered by the sample is collected. A Pellicle mirror features a continuous reflecting surface and, as such, provides an undistorted image.

5. DISCUSSION AND CONCLUSION

The sample viewer is a useful tool to enhance the sample-to-beam alignment and its in-situ capability allows users to observe the sample during X-ray experiments. Especially for high magnifications, additional illumination of the sample is required. Although not shown in Figs. 1 and 2, we have benefitted from using a ring LED light source ([10]) that can be placed on either side of the mirror to provide extra light.

When using the device with a tightly focused X-ray beam, the alignment of the sample in the direction along the beam can be achieved within the field of depth of the zoom lens, which lies between 0.4 and 4 mm.

Although the setup works with both mirror variants, the Pellicle mirror-based version has proven to be more popular, partly due to the fact that it avoids image artifacts such as shown in Fig. 4. The main reason, however, is that there is no chance of clipping the incident X-ray with the glass mirror if the beam is slightly misaligned or when users want to use a larger beam size. The commercial version of the Pellicle is constructed with a nitrocellulose membrane several micrometers thick. Prolonged exposure of this material will lead to radiation-induced damage. Retracting this mirror using the pneumatic actuated mechanism will avoid such deterioration of the membrane. Thin graphite, Kapton (polyimide) and SiN-membrane-based Pellicles are available that do not suffer from radiation damage for applications in which the sample must be monitored during X-ray exposure. Because BM14 is a hard X-ray beamline, the attenuation of the X-ray beam by the Pellicle mirror with coating is negligible.

Due to its on-axis viewing capability, the sample viewer provides an “X-ray view” of the sample; thus, any sample damage or misalignment that occurs during exposure is visible. In many cases, it can therefore replace traditional cameras located around the sample, allowing more free space for measurement or in-operando experiment equipment to be placed near the sample. Particularly for experiments that use environmental chambers or for experiments that use X-ray beams with low photon energy that have little opportunity to monitor the sample, the proposed sample viewer would be a welcome addition.

In the proposed solution, it is relatively straightforward to place the 45° tilted mirror into a joint vacuum chamber with the zoom lens and camera, observing the sample through a small vacuum window. An overview drawing with a parts list of the on-axis sample viewer is available in Supplement 1.