1. An Automated System to Mount Cryo-Cooled Protein Crystals on an SSRL Beam Line, Using Compact Sample Cassettes and a Small-Scale Robot
Paul Phizackerley, Aina Cohen, Ashley Deacon, Paul Ellis, Ross Floyd, Michael Hollenbeck, John Kovarik, Mitchell Miller and Vladimir Vinetskiy
Stanford Synchrotron Radiation Laboratory, SLAC, MS 69, 2575 Sand Hill Road, Menlo Park, CA
Ideal Hampton Pin Lengths for Cassette
Cassette Code Pin
Cassette and Sample Port Nomenclature
The Robot In Action
Abstract
Rather than place a bar code on each cassette for identification, which is difficult to read through LN2, we intend to use a code pin similar to the one in the drawing below, and place a unique one in port A1 of each cassette to identify the cassette. This code pin has multiple divisions, each being one of several standard diameters, rather like a door key in principle. After a cassette has been loaded, the robot will first load the code key on the beam line diffractometer and read the unique cassette code using the sample crystal viewing TV camera. We plan to use a number of additional divisions to provide a check sum for code verification.
A fully-automated system has been developed and tested at SSRL to mount protein crystals, that have previously been flash-cooled in cryo-loops on Hampton Research cryo-pins, on a beam line diffractometer. The driving force behind this development has been the necessity to increase the efficiency, accuracy and throughput on our crystallography beam lines, especially in light of the additional demands placed on our facilities by the new structural genomics programs. Central to the ease of shipping, storing, sorting, mounting and tracking sample crystals has been the development of a compact cylindrical sample cassette that is capable of storing up to 96 crystal samples. These cassettes are designed to fit inside a commercially available shipping Dewar and on arrival at SSRL, stored in a large capacity storage Dewar. When needed, up to three sample cassettes can be manually transferred to a LN2-filled dispensing Dewar located next to the beam line diffractometer. A robot is then used to select the required sample and (dry) mount it on the diffractometer, without interrupting the stream of N2 used for cryo-cooling. Samples can be selected for screening or data collection in any order and any number of times. A small-scale industrial pick-and-place 4-axis Seiko ES550/320 robot has been selected to make the transfer back and forth between the dispensing Dewar and the phi axis of the Kappa diffractometer. X, Y and Z sample translations on the diffractometer facilitate the remote alignment of the sample to the x-ray beam. For reliability, the robotic mounting system employs a very simple mechanical design with few moving parts and uses permanent magnets to extract and replace samples from the sample cassettes. This poster describes the design and function of the overall system.
The cassette is designed to accommodate samples mounted on off-the-shelf Hampton pins. Both CrystalCap Magnet and CrystalCap Copper Magnetic pins may be used. (The copper pins are recommended.) The drawings below show the ideal combinations of pin and Hampton Microtube length for use with the cassettes. However, small variations in these lengths can also be accommodated.
The drawing above shows a plan view of the three cassettes in the dispensing Dewar as viewed by the user when loading the cassettes. “L” Left, “M” Middle and “R” Right is used by the software to identify the appropriate cassette for sample selection. Ports “A” through “L” are identified as shown. The “transfer post” where the magnet tool is stored is shown to the back of the Dewar. A strong magnet is on the left and a weaker one is on the right. Red circles show the area used around each cassette for the sample selection process. As you can see, the three cassettes are as closely packed in the Dewar as possible.
The cryo-tongs at the transfer post. The cryo-tong / magnet tool selector is about to pick up the magnet tool and take it to the requested sample port.
The sample pin being transported inside the cryo-tong to the phi-axis of the diffractometer. The sample is loaded from above without moving the cryo-stream.
Robot and the Dispensing Dewar on BL11-1
Mount for 3 Cassettes in the Dispensing Dewar
The picture on the top shows the robot mounted on SSRL beam line In the left foreground one can see a large Dewar, with a blue lid, which is used to dispense samples from up to three cassettes (a total of up to 288 samples) located inside – shown more clearly in the bottom picture. We refer to this as the dispensing Dewar. This Dewar and the robot is supported on a rigid support stand that is attached to, and moves with, the experimental table on air bearings. The Kappa-geometry diffractometer can be seen to the right and is coloured green. The x-ray beam traverses from upper left to lower right and the x-ray area detector is out of the picture to the right - far enough away from the robot so that a collision is not possible.
The robot is a commercial unit from Seiko (Model ES550/320). It employs three vertical axes, giving a reach of 550mm in the horizontal plane, and the third axis incorporates a vertical translation of 320mm. Each of the four motions incorporates an absolute encoder providing an overall positional accuracy of better than 20 microns (3). A versatile and straightforward software language SPEL+ (which looks very similar to BASIC) is provided by Seiko for robot control. Optically isolated electronic I/O interfaces are used between the robot controller and pneumatic actuators and position sensors on the experimental apparatus. The vertical robot arm has been equipped with an SMC pneumatic gripper and a specially designed pair of tongs, that is used to select and mount the required samples. The dispensing Dewar is filled with LN2 and will shortly be equipped with an automatic top-up system. The lid on top of the Dewar is pneumatically actuated and can be driven by either the computer or by a foot switch on the floor. The Dewar is conveniently situated near the experimental table and is easily accessible to the user.
The pictures below show a close up view of the dispensing Dewar that can hold up to three sample cassettes. The picture on the left shows support platforms for the cassettes mounted on a shelf suspended well above the bottom of the Dewar to permit any ice particles that are generated in use to fall to the bottom and away from the mechanism. Each cassette is supported on a central cone and oriented about a vertical axis by means of a datum pin. Vertical rods help to guide the cassette down to the base of the mount when loading the cassette into the LN2 which covers the cassette entirely in use. Each cassette rests on 6 balls to reduce the effect of ice, should a thin layer form on the bottom of the cassette during transfer from one cryo-container to another. The picture on the right shows three cassettes loaded in place.
Description of the Robot Mounting Mechanism
A sample pin being withdrawn by the strong picker magnet on the magnet tool.
Sample Mounting procedure
After the control software receives a request to mount a particular sample in one of the cassettes (say L7C for the sample in the left cassette, port 7C), the dispensing Dewar lid is opened and the robot tool (i.e. the cryo-tong / magnet tool selector) is lifted out of the tong heater (the tong heater is a black vertical cylindrical tube attached to a heat gun that can be seen to the left of the robot in one of the pictures shown on the far left of this poster) and briefly placed in front of an air jet and given several quick bursts of high pressure air to dislodge any water droplets that have formed during heating. When the dispensing Dewar lid has been verified as fully open and locked by electronic sensors, the tool enters the dispensing Dewar and is given a few seconds to cool down to LN2 temperature. The selector then picks up the magnet tool and moves the strong permanent magnet to a datum point on cassette L opposite port L4G. Using a helical path, the strong magnet is taken to a point a short distance away from port L7C and then moves radially towards the L7C sample pin. Each sample pin is held inside the cassette port by means of a small ring magnet and when the strong magnet tool comes in contact with the sample pin, it is strongly attracted to it. Consequently, when the robot radially translates the magnet tool away from the cassette, the sample pin readily leaves the cassette. The magnet tool and sample are then transported back to the transfer post and is released by the robot. The cryo-tong is then placed around the sample pin and closed with the sample crystal embedded inside the cryo-tong. By using a twisting movement about a vertical axis through one edge of the sample pin, the sample pin is removed from the magnet tool and then transported at a moderately high speed to the beam line diffractometer and carefully placed on another permanent magnet at the center of the phi axis. This operation does not require the LN2 cryo-stream to move and when the cryo-tong opens and moves away, the sample is left in the cryo-stream. The robot then transports the cryo-tong back to the electrical heater. During the heating cycle, the dispensing Dewar lid is closed to avoid ice formation and remains closed until a dismounting operation is requested.
A diagram depicting the crystal mounting procedure is shown below.
View of the entire robot mounting system.
Project Status and Where We Go From Here
The control software, that is capable of selecting any port on any one of the three sample cassettes, and mounting (or dismounting) the sample on the beam line diffractometer, is working well and is essentially complete. What has yet to be written is the database software to track the cassette when it is shipped to SSRL, and also the high level software within BLU-ICE on the beam line control computer to interface with the robot control software. This high level software will control which sample is selected for either a quick data quality survey, or a full data collection run, as determined by a scheduling script or by “dynamic” requests sent to it remotely. The high level software will also include the capability of scheduling requests to the robot control software for sample sorting within one cassette or between cassettes. This will be useful after initial screening, to combine promising crystals from multiple cassettes into one cassette for subsequent data collection. We also intend to build a centralized system at SSRL for both sample sorting and for transferring samples from cassettes of different styles to be compatible with users from other labs. We also need to develop a number of electronic sensors and control software to guard against collisions between the robot mechanism and other experimental equipment, and to install a LN2 top up system on the dispensing Dewar. Finally, we plan to replicate the entire robot mounting system on each of our protein crystallography beam lines at SSRL.
Transferring a Cassette from Shipping to Dispensing Dewar
Close up view of the dispensing Dewar.
The Taylor Wharton CP100 transport Dewar has been adopted as our standard shipping Dewar for the cassettes. It has an ideal size for transporting a single cassette. The only modification necessary for compatibility with the cassettes is a simple angular cut to the internal sample holding sleeve. This angular cut makes it convenient for the user to load and extract the cassette from the internal sleeve with a specially designed transfer handle shown in the pictures below. This handle locks (using a 90o rotation) into a slot in the top of the cassette and is firmly attached until it is again unlocked. After the cassettes arrive at SSRL and they are required on the beam line, the user transfers the cassette into the dispensing Dewar after opening the lid using a foot switch on the floor. If, after arrival at SSRL, the cassettes that are not going to be used for a few days can be stored in a Taylor Wharton HC35 storage Dewar (pictured below) that is capable of storing up to 10 cassettes for several months without refilling.
The Standard Cassette
To provide maximum convenience and flexibility to the user and simplify transporting, storing and mounting a very large number of sample crystals, a compact “Standard” sample cassette has been developed and is shown below. Each cassette has ports for 96 samples mounted on standard Hampton Research CrystalCap Magnetic or CrystalCap Copper Magnetic Pins. There are 8 ports in the vertical direction (1 – 8) and 12 ports around the circumference (A – L). Sectional plan and side views of the cassette are shown on the right to demonstrate the internal packing of the Hampton pins. Each pin is held in place by means of a NdFeB ring magnet (shown in blue in the detailed drawing on the bottom right) and the pin is slightly recessed for protection. The bottom of the cassette has a central conical hole used to position the cassette accurately on a location platform in the dispensing Dewar. A radial slot in the bottom is used to orient the cassette on an orientation pin. The cassettes are made of 303 stainless steel because it has a high thermal capacity and is very durable. Each sample crystal is stored in a narrow tunnel within the cassette so that when LN2 drains from the cassette during transfer from one cryo-container to another, the crystals will be maintained at ~LN2 temperature. The cylindrical cassette was designed to fit snugly inside the inner cage of a Talor-Wharton CP100 dry shipping Dewar.
Sample Dismounting procedure
After the control software receives a request to dismount the sample from the diffractometer and place it in one of the cassettes, the dispensing Dewar lid is opened and as before, the robot tool is lifted out of the tong heater and placed in front of an air jet and given several quick bursts of air to dry the cryo-tong. When the dispensing Dewar lid is fully open, the tool enters the Dewar and is given a few seconds to cool down to LN2 temperature. The cryo-tong is then taken out of the Dewar and closed around the sample on the diffractometer. Using a twisting movement, the robot removes the sample pin from the permanent magnet on the diffractometer phi axis and transports it back to the transfer post inside the dispensing Dewar where the sample pin is placed on the weak permanent magnet end of the magnet tool. The sample is then moved to a position near to the cassette and port that has been selected to receive the sample and the sample pin is translated radially towards the port. Since the sample pin is only being held by a weak permanent magnet, when the sample pin gets close to the cassette ring magnet it is immediately attracted to it and readily leaves the weaker permanent magnet on the magnet tool. The robot then transports the magnet tool back to the transfer post and after releasing it, returns to it’s “home” position inside the cryo-tong heater. The heater is then switched on for a few seconds and the dispensing Dewar lid is closed.
A diagram depicting the crystal dismounting procedure is shown below.
Most of the SSRL PX Group at Their Post
SSRL is funded by:
Department of Energy, Office of Basic Energy Sciences,
The Structural Molecular Biology Program is supported by:
National Institutes of Health, National Center for Research Resources, Biomedical Technology Program
NIH, National Institute of General Medical Sciences
and by
Department of Energy, Office of Biological and Environmental Research
The authors would like to thank the members of the SSRL SMB PX Group for useful discussions during the course of this development, Henry Meier and Renato Avelar for their help with the construction and installation of the electronic components and Michael Swanson who machined some of the components of the early prototype.
Cassette being removed from
CP100 dry shipping Dewar.
Cassette being loaded into
the K3 dispensing Dewar.
Sectional plan and side views
showing Hampton pin packing.
The cassette and transfer handle.
View of the “standard” cassette.
Close up view of a single port.
The HC35 long term storage Dewar.