K2 Summit TM and K2 Base TM Direct Detection Camera User s Guide

Transcription

1 K2 Summit TM and K2 Base TM Direct Detection Camera User s Guide Model Number 1000, 1000.S, 1000.B and 1000.U Document Part Number: Revision 01 Gatan, Inc W. Las Positas Blvd Pleasanton, A Tel Fax

2 Table of Contents Preface 6 About this Guide... 6 Conventions... 6 Disclaimer... 6 Copyright and Trademarks... 6 Returns... 7 Support... 7 Introduction 8 Why Silicon APS and Counting are Good... 8 K2 Base and K2 Summit Configurations Super-Resolution Description and Overview of Performance Benefits Linear Accumulation Setup Mode Radiation Hardness and Expected Lifetime Getting Started with the K2 Direct Detection System 14 Power on Sequence for K2 Base Power off Sequence for K2 Base Power on Sequence for K2 Summit Power off Sequence for K2 Summit Setting Temperature Anneal Cycle K2 Summit Imaging Modes Updating Reference Images K2 Summit Reference Images K2 Base and K2 Summit Linear K2 Summit Counted and K2 Summit Super-Resolution Updating Summit Hardware Background Reference Image Only Updating Reference Images for K2 Summit Counted and Super-Resolution Modes Base Gain Reference Collection Key Points to Ensure Optimum Operation Other Recommendations

3 Operation of the K2 Direct Detection System 30 K2 Imaging Linear Mode Operation View Mode Counted and Super-Resolution Mode Operation Dose Rate Considerations Dose-Rate Monitor Shutter Delay Setup Defect Correction Binning and Subareas Dose Fractionation Setup Shutter Configuration Shutter Connections and Signal Configuration Base Camera: Linear Mode Operation Switching Between Base and Summit Modes Care of Detector Temperature Anneal Cycle Auto Retraction Insertion Indicators Dynamic Sensor Protection (DSP) Operation Software Overview: Control of K2 Camera within DM Screenshot of DM Display for K2 Camera Magnification Correction/Calibration Low Magnification High Magnification Specifications Compatibility with Other Gatan Cameras and GIF Addendum 48 Ensuring Validity of Defect Corrections

4 List of Figures figure 1-1. Monte Carlo simulations of electron interaction with direct and indirect camera system... 9 figure 1-2. Overview of the K2 Base and K2 Summit systems figure 1-3. Overview of electron detection process in K2 Base and K2 Summit figure 1-4. K2 digitizer figure 1-5. TEC Controller (top) and K2 Summit Processor (bottom) figure 1-6. K2 digitizer figure 1-7. DigitalMicrograph interface during startup of K2 Summit system figure 1-8. DigitalMicrograph interface during shutdown of K2 Summit system figure 1-9. Camera menu figure K2 Sensor temperature figure K2 Direct Detection palette: Linear mode figure K2 Direct Detection palette: Counted mode figure K2 Direct Detection palette: Super-Resolution mode figure Option to skip gain reference collection figure Suggested gain reference exposure factor figure Settings for K2 Summit linear mode gain reference figure Expert Mode settings figure Gain reference exposure factor recommendations figure Beam intensity recommendation figure Settings for intensity and number of frames to average figure Image overwrite confirmation figure Linear mode gain reference success message figure Gain reference collection option figure Hardware dark reference option figure Settings for K2 Summit linear mode gain reference figure Expert Mode settings figure Beam adjustment reminder figure Beam intensity recommendation figure Settings for intensity and number of frames to average figure Linear mode gain reference success message figure Update HW Dark Reference button figure View Mode dialog

5 figure Current dose rate on camera figure Current dose rate and total dose accumulated on the specimen per image figure Pre-Dose Delay setting figure Advanced View Options, settling delay option figure Comparison of a dose fractionated dataset acquired and summed with alignment (top right) and without per frame alignment (lower left) figure Exposure time setting figure Dose Fractionation mode settings figure Shutter settings tab figure Auto Retraction Delay setting figure Gatan logo, and Camera Inserted setting figure Typical DM setup DigitalMicrograph figure Example of marking a known distance during magnification calibration figure Magnification calibration instructions figure Calibration settings figure Calibration confirmation figure Microscope Calibration dialog figure High-resolution image of sample to be used in the magnification calibration figure Distance between peaks in the calculated diffractogram figure Calibration instructions figure Calibration settings

6 Preface About this Guide This K2 User s Guide provides a brief description of the camera s operating principles, instructions on how to use the camera, and an introduction to imaging in DigitalMicrograph. The intent of this guide is to help the user become proficient and comfortable with the camera, DigitalMicrograph software and to correctly acquire images from the camera. Additionally it should give the user an understanding of the working principles of the camera and allow for easy troubleshooting. Conventions The following typographical conventions are used for special comments: SPECIAL NOTE: Recommendations for getting the best performance from the equipment. CAUTION: This is the CAUTION symbol; it alerts you to potential hazards that require you to consult the documentation in all cases where this symbol is marked. WARNING: Alerts to indicate when something may have the potential to cause damage to the equipment or fail to operate correctly if the instructions are not followed. Disclaimer Gatan, Inc., makes no express or implied representations or warranties with respect to the contents or use of this manual, and specifically disclaims any implied warranties of merchantability or fitness for a particular purpose. Gatan, Inc., further reserves the right to revise this manual and to make changes to its contents at any time, without obligation to notify any person or entity of such revisions or changes. Copyright and Trademarks 2012 Gatan, Inc. The Gatan logo is a registered trademark of Gatan, Inc. K2, K2 Summit, K2 Base, and DigitalMicrograph are trademarks of Gatan, Inc. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of Gatan, Inc. The text and graphics are for the purpose of illustration and reference only. The specifications on which they are based are subject to change without notice. 6

7 Returns In the event that the instrument must be returned to the factory, please request a Returned Materials Authorization (RMA) number via gatan-rma@gatan.com. This RMA number must appear on all shipping documents to ensure that proper actions will be taken to repair or replace the instrument. Support Gatan, Inc. provides free technical support via phone, fax, and electronic mail. To reach Gatan technical support, contact the facility nearest you, or send an to service@gatan.com or info@gatan.com. Please consult the Customer Service section of the Gatan website at for the latest contact information. 7

8 Introduction IMPORTANT: For Regulatory Compliance and Safety information and instructions please refer to the Regulatory Pamphlet provided with this product. Review this document in full before installing and operating this product. The K2 Summit Direct Detection Camera is a new type of camera for TEM. The camera contains a direct detection transmission CMOS detector, which gives the highest resolution images of any electron-imaging sensor available today. The camera runs in a mode constantly collecting images at 400 frames per second. The camera then, through a high-bandwidth link, passes these frames at full speed to dedicated high-throughput hardware designed for the express purpose of processing these images in real time. By collecting and processing fullframe images so quickly the detector can identify and record individual electron events (counting) as they reach the sensor. By counting every single electron event, the camera can eliminate the background noise typically seen in devices that simply read out the charge deposited by an electron striking a piece of silicon. By removing this source of noise, the camera can offer higher image quality and sensitivity than previously available in an electronimaging device. This device is also capable of operating in a Super-Resolution mode. In this mode, the sensor is able to localize the electron event with sub-pixel accuracy, effectively doubling the number of pixels available for imaging (from 3838 x 3710 to 7676 x 7420). Again the processing is all done on full frames in real time as the images are collected. In addition, the camera employs a pixel design giving radiation hardness 10x greater than any other direct-detection sensor available (pixel lifetime of >5 billion electrons). K2 Summit also allows Linear read mode (used on K2 Base ) for high-dose applications. Why Silicon APS and Counting are Good Direct Detection sensors are derived from Active Pixel Sensor (APS) technology developed for digital cameras and cell phones. Like a CCD, an APS is an integrated circuit containing an array of pixels. But unlike a CCD, readout does not require pixel-to-pixel charge transfer. Each pixel contains a photo detector and active amplifier that is addressed and read-out individually. The K2 Direct Detection APS is designed using proprietary design and layout techniques that make it radiation resistant enough to detect electrons without significant damage to the sensor. K2 Direct Detection APS is based on CMOS semiconductor technology and produced at commercial semiconductor foundries with manufacturing of more advanced (smaller) nodes resulting in increased radiation resistance. Direct Detection was used in high-energy particle physics and other fields before coming to EM. A major challenge for EM uses is that energy and charge deposited by the high-energy electrons quickly damage most sensor types, causing a rapid deterioration of the image quality. CCD sensors have been used for direct detection, although only for a short period of time before failure. Early development ( first generation ) EM Direct Detectors had poor radiation hardness, meaning poor ease of use and the need for a second set up camera for 8

9 all operations other than the final acquisition of low dose data. K2 implements design and fabrication choices which significantly increase radiation hardness. figure 1-1. Monte Carlo simulations of electron interaction with direct and indirect camera system A Direct Detection sensor only records the signal deposited in its top layer; energy deposited in the sensor s support is not detected. Siliconn itself is a relatively low atomic number (Z) element, dramatically reducing the PSF relative to traditional EM cameras where phosphor scintillators are used. In a first generation bulk Direct Detector there is still backscattering of electrons from the sensor bulk back to the top (active) layer, increasing the PSF and reducing resolution. Advanced Direct Detection sensors, like K2, are back-thinned allowing the primary electrons to pass through the sensor, with no bulk support and thus minimize this backscattering. The much smaller PSF for a transmission system such as K2 allows a Direct Detection sensor to use much smaller pixels, 5 μmm in case of K2, yielding many more pixels per unit of area. Direct Detection cameras have a superior signal-to-noise ratio relative to traditional EM cameras as the signal from the number of electron hole pairs detected from each primary electron is large relative to the read noise of the sensor electronics. Another advantage of Direct Detection is the absence of a number of image artifacts such as distortion, fixed patterns, and gain variations associated with scintillators and fiber optics or lenses. The Detective Quantum Efficiency (DQE) of traditional cameras is held back by the read noise of the cameraa electronics and by large variations in the amount of energy deposited from one electron to the next. This is also truee for direct detection cameras and for non- counting models and modes; their ultimate performance is limited by the noise in the system The electron counting mode of K2 Summit replaces the analogue signal from each primary electron with a discrete count. The benefit of counting is that it rejects read noise and the natural variation in energy deposited by electrons impinging on the detector, thereby dramatically lifting the Detectivee Quantum Efficiency (DQE) of the detector across all spatial frequencies. 9

10 K2 Base and K2 Summit Configurations The K2 Summit Super-Resolution mode takes counting further and surpasses the theoretical information limit defined by the physical pixel size. The K2 sensor was carefully designed such that the PSF is slightly larger than the 5 μm physical pixel size. As a result each incoming electron deposits signal in a small cluster of pixels. The high-speed K2 Summit The K2 Direct Detection cameraa system comes in two configurations, the K2 Base which offers the performance improvement you would expect from a second-generation direct detection camera, and the K2 Summit, which includes the dedicated hardware to enable the electron counting and Super-Resolution modes of operation. The K2 Base Direct Detection camera system includes the camera itself, and digitizer unit and PC which is running DigitalMicrograph. The K2 Base system only operates in the traditional charge integration mode that we call Linear mode. figure 1-2. Overview of the K2 Base and K2 Summit systems The K2 Summit Direct Detection camera system also includes the camera, the digitizer unit and PC, but what makes the K2 Summit unique is that it includes a dedicated processing unit that is responsible for counting the individuall electrons as they strike the sensor. It is the combination of the very fast readout with the ability to count individual electrons at the same high rate that enables the unique Counted andd Super-Resolution modes available with this system. Super-Resolution Description and Overview of Performance Benefits 10

11 electronics is able to recognize each electron event (at 400 frames per second) and find the center of that event with sub-pixel precision. The net effect is a quadrupling of the effective number of pixels (pushing beyond the Nyquist information limit to even higher resolution), as well as a further improvement of the DQE (and MTF). Practically, this means that the field of view can be increased for the same end resolution allowing the researcher to capture much more data per image. As a point of interest, Counted mode uses the same processing algorithm as Super- Resolution, but bins pixels 2x2 in the camera to support faster live viewing speeds and a more compact data format. The decision of which mode to use should be based on user criteria and experiments on an application by application basis. This fact is the reason that a single Super- Resolution acquisition is all that is needed to produce software gain references for both Super-Resolution and Counted modes. The Need for Speed: Electron counting is only possible if the camera can read-out images fast enough to see the individual electrons raining down on the sensor. K2 was specifically designed to count at a typical Cryo EM dose rate of e - /pixel/s allowing a 20 e - /Å2 low dose image to be recorded in 1-2 s (at a magnification equivalent to 1 Å/pixel). To do so the K2 Summit reads out at a rate of 400 full frames/s (5.7 Gigapixels per second!) using a highly parallelized, high-speed architecture able to process image data at 80 Gb/s. Much of the discussion surrounding Direct Detection cameras focuses on only the performance improvements of the sensor relatively to traditional CCD cameras. What this misses is that it is not possible to routinely acquire images in Counted mode unless the data processing pipeline is in place to handle this massive volume of data in real time. We like to say that K2 is The One That Counts because counting is what gives the biggest performance boost to the K2 camera. K2 is the only Direct Detection EM camera specifically designed for counting and able to do so in a useful way. Counting (K2 Summit only): Individual primary electrons are counted in-line in the Summit processor on a pixel-bypixel, frame-by-frame basis. The electron Counted mode of K2 Summit replaces the analogue signal from each primary electron with a discrete count. The benefit of counting is that it completely rejects the read noise and dramatically lifts the Detective Quantum Efficiency (DQE) of the detector across all spatial frequencies. 11

12 figure 1-3. Overview of electron detection process in K2 Base and K2 Summit Super-Resolution Counting (K2 Summit only): Each primary electron s signal cloud is analyzed in the Summit processor to determine the electron s landing coordinates with sub-pixel accuracy. This technique extends the resolution beyond the number of pixels inn the sensor. The K2 Summit Super-Resolution mode takes counting further and surpasses the theoretical information limit defined by the physical pixel size. The K2 sensor was carefully designed such that the PSF is slightly larger than the 5 μm physical pixel size. As a result each incoming electron deposits signal in a small cluster of pixels. The high-speed K2 Summit electronics iss able to recognize each electron event (at 400 frames per second) and find the center of that event with sub-pixel precision. The net effect is a quadrupling of the effective numberr of pixels (pushing beyond the Nyquist information limit to even higher resolution), as well as a further improvement of the DQE (and MTF). Practically, this means that the field of view can be increased for the same end resolution allowing the researcher to capture much more data perr image. 12

13 Linear Accumulation Setup Mode Linear mode is the traditional energy-integrating read out mode that captures the total integrated signal level in each pixel just like a CCD camera does. Individual frames can be summed for an effectively unlimited dynamic range. In K2 Base and K2 Summit linear mode charge is collected, integrated during the exposure, and read out to provide the image. Sensor electron charge proportional to the total energy deposited by the incoming primary electron is accumulated. This mode benefits from Direct Detection s improved DQE arising from the transmission detector and inherent higher conversion efficiency. When not counting, there is no requirement to keep charges separated. This mode is good for higher dose images and allows for dose rates up to 400 times higher than in Counted mode, and somewhat higher than is practical even in traditional CCD cameras. Of course, if higher doses are used, the user must remain aware of the lifetime total dose capability of the camera and avoid unnecessary exposure of the camera to the beam. Radiation Hardness and Expected Lifetime Minimize exposing the CMOS sensor to the electron beam when the camera is not in use. Based on the results to date, the lifetime of the K2 detector is estimated at 5 billion e - /pixel. Considering exposure rates of 100 electrons/pixel/second, if the sensor were exposed continuously for 24 hours a day every day, the expected lifetime would be well over 1 year. The camera employs a pixel design that is 10 times more resistant to incident electron damage as compared with other direct-detection sensors. The extremely short exposure time coupled with the event discrimination of Counted mode confers an additional approximately 10 times immunity to radiation. The sensor can be damaged and marked by a bright beam. Use caution with intense beams and bright spots. Do not leave the sensor exposed to the beam when changing magnification levels or spot sizes. These can create brief intense spots. Annealing may be able to reduce or remove minor beam spots. Direct Detection cameras receive the incoming electrons directly to the imaging sensor, and in the case of K2 can detect them in several modes. This Direct Detection eliminates the need for scintillator and fiber optics or lenses and greatly reduces Point Spread Function (PSF). Furthermore, in Gatan s Counted modes, noise can be essentially eliminated, particularly because with Gatan s transmission sensor, there are no scattered electrons reentering the detector from below. 13

14 Getting Started with the K2 Direct Detection System Power on Sequence for K2 Base To start from the power off: 1. Close DM if it is open (or power on the PC if it is off). 2. Turn on the K2 digitizer (blue K2 box near the camera and TEM). figure 1-4. K2 digitizer 3. Wait at least 3 minutes after powering on the K2 digitizer. 4. Launch DigitalMicrograph (DM). 5. Click the setting button (hammer/wrench) (#1, Figure 1-7, page 16) in the K2 Direct Detection palette. 6. Click the System ON button. Wait for the system to finish turning on. 14

15 Power off Sequence for K2 Base Under normal circumstances, you can leave the K2 box running. If necessary, you can power down the K2 camera completely. To completely power down the K2 camera: 1. In DM: Click the settings button (hammer/wrench) in the K2 Direct Detection palette. 2. Click the System OFF button. Wait about 10 seconds. 3. Close DM. 4. Turn off K2 digitizer (blue box). 5. Turn off the PC if desired. Power on Sequence for K2 Summit To start from the power off: 1. Close DM if it is open (or power on the PC if it is off). 2. Turn on power to the Thermo Electric Cooler (TEC). 3. Turn on the power switch on the back of the K2 processing rack. figure 1-5. TEC Controller (top) and K2 Summit Processor (bottom) 4. Turn on the K2 digitizer (blue K2 box near the cameraa and TEM). 15

16 figure 1-6. K2 digitizer 5. Wait at least 3 minutes after powering onn the K2 digitizer. 6. Launch DigitalMicrograph (DM). 7. Click the setting button (hammer/wrench)) (#1) in the K2 Direct Detection palette. 8. Click the System ON (#2) button. Wait for the system to finish turning on (#3). figure 1-7. DigitalMicrograph interface during startup of K2 Summit system 16

17 Power off Sequence for K2 Summit Under normal circumstances, you can leave the K2 box running. If necessary, you can power down the K2 camera completely. To completely power down the K2 camera: 1. In DM: Click the settings button (hammer/wrench) in the K2 Direct Detection palette. 2. Click the System OFF button. Wait about 10 seconds (or open the logs and watch for the DONE message in digitizer log window). figure 1-8. DigitalMicrograph interface during shutdown of K2 Summit system 3. Close DM. 4. Turn off K2 digitizer (blue box). 5. Turn off K2 processing rack. 6. Turn off the PC if desired. 17

18 Setting Temperature 1. In the Camera menu, select Temperature and enter the set point value. figure 1-9. Camera menu Typical operating temperature is -20 C. 2. If the camera chamber needs to be vented, set this to +20 C. Wait for warm-up to complete before venting. The camera temperature is displayed in the Camera Monitor palette. figure K2 Sensor temperature Anneal Cycle Regular annealing (heating to +50 C) of the sensor helps maintain top performance of the sensor by reducing background levels and levels of surface contamination. Annealing can also help to repair some radiation damage, although the camera will achieve the full radiation damage specification if just a single anneal cycle is performed on the camera. We recommend that an anneal cycle be performed each time the microscope does a cryo-cycle. This can be done as often as every evening or less frequently (say once a week) during extended periods of data collection. 18

19 To perform the Annealing cycle: 1. In the Camera menu select Temperature. 2. Select a set point of +50 C, and click OK. 3. Use the Options button to select a duration other than 24 hours. Usually an annealing cycle of overnight is sufficient. 4. Click the Start button to begin the cycle. The camera cools down to whatever the set point was before cooling. If that was -20 C, then after the annealing cycle is complete, it cools back down to -20 C. K2 Summit Imaging Modes The K2 hardware can boot up in two modes, Base and Summit. K2 Base operates in charge integration mode. When the Summit camera version is selected, the three modes available from the K2 Direct Detection palette are Linear, Counted, and Super-Resolution. Linear mode is the way that the K2 Summit camera operates in charge integration mode, while Counted and Super-Resolution are highly processed modes where individual electrons are detected. In the view and acquire menus (in DM) you can select three Summit imaging modes: Linear: In this mode, enable software dark subtraction and gain normalization (selected in the Camera Acquire and Camera View palettes) and be sure the hardware corrections in the K2 Direct Detection palette (Background Subtraction, Gain Correction) are unchecked. figure K2 Direct Detection palette: Linear mode 19

20 Counted: Each electron event results in one count in the image. The raw images are dark subtracted, gain-normalized and processed in hardware to detect individual electrons. The resulting summed electron counts are output as an image. In this mode, keep the beam at low levels (20 electrons per pixel per second or less) to avoid signal loss. figure K2 Direct Detection palette: Counted mode Super-Resolution: Electron events are localized to the sub-pixel in which they occurred. Additional processing is done to further refine the electron positions resulting in a higher resolution image with the best performance possible. Raw images are dark subtracted, gain-normalized, and processed in hardware to detect individual electrons with sub-pixel accuracy giving a 4x as many pixels in each image. figure K2 Direct Detection palette: Super-Resolution mode Updating Reference Images NOTE: Power cycling the hardware does not maintain the stored hardware background reference and hardware gain reference used for Counted and Super-Resolution modes. You must re-upload your references images before starting data collection. Reference must be re-acquired or uploaded from a saved image on disk after power cycling. DigitalMicrograph detects when this has happened and prompts you to upload or re-acquire the references. 20

21 Select Camera > Prepare Gain Reference. You are led through taking all necessary gain and dark references. K2 Summit Reference Images K2 Base and K2 Summit Linear First, collect a Linear mode high beam level gain reference used for Linear mode and hardware gain correction in Counted and Super-Resolution modes. figure Option to skip gain reference collection Note that you can skip parts of this procedure if you already have a good reference and don t want to update it, or if you wish to leave the procedure part way through and come back later. For instance, at this step you could skip gathering a new linear reference if you already have one and proceed with the later steps. You may be asked if it s OK to insert the camera or switch from another mode if the camera is currently running a live view. Click Yes. In the next step, you have the option to choose gain reference collection parameters. You may see this warning: figure Suggested gain reference exposure factor Click OK to proceed to the setup window. 21

22 figure Settings for K2 Summit linear mode gain reference If you click the Expert Mode box, you have the option to restore default values or select and save your own values for linear mode gain reference collection (if you want a higher quality gain reference or your normal operating conditions are markedly different for instance). figure Expert Mode settings 22

23 Clicking the question mark button provides some guidance. Linear mode is not very sensitive to dose rate, you could increase this from 5000 to if you like. If you decrease it from the default, it s best to scale the exposure factor accordingly (for instance if you halve the dose rate, double the exposure factor). figure Gain reference exposure factor recommendations Otherwise you can use the default values which work well under most circumstances, and click OK to continue. (Note if you select a lower than default exposure factor which could result in low image quality you see the warning message pop up again and have to reconfirm your choice.) You are now asked to adjust the beam to the selected dose rate. Because short exposures are used, the requested dose (500 counts below) is a fraction of the desired dose rate (5000 counts/second above). Note the default Linear mode gain reference beam level is more than 10 times brighter than the typical Counted and Super-Resolution beam level, if you have been operating in those modes you have to brighten the beam considerably. figure Beam intensity recommendation When you adjust the beam and select Done, the script calculates input values for DigitalMicrograph s gain reference collection routine and fill them in. Do not change these values, and click OK. 23

24 figure Settings for intensity and number of frames to average Next, DigitalMicrograph informs you that previous gain reference images are not saved during this procedure. Click Yes. figure Image overwrite confirmation DigitalMicrograph collects and saves the necessary images, and calculates any needed corrections. When the success message appears, click OK. figure Linear mode gain reference success message K2 Summit Counted and K2 Summit Super-Resolution Next, if working in K2 Summit mode you are asked if you want to collect the Super- Resolution references (which are also used for counted images). Preparing a gain reference for K2 Super-Resolution mode is optional if it has been completed recently. A gain reference can last for a number of days. figure Gain reference collection option 24

25 The first step is to acquire and upload a dark reference to the processors. Preparing a hardware dark reference for K2 Super-Resolution mode is optional if it has been completed recently but is highly recommended preceding a Super-Resolution gain reference acquisition. A gain reference taken immediately following a fresh hardware dark reference will not be invalidated by a subsequent fresh dark reference. Gain references taken in this manner can last a number of days with only hardware dark reference updates performed more frequently. Hardware dark reference update should be performed at least once per day for optimum counted and super-resolution performance. figure Hardware dark reference option Next the linear gain reference is automatically uploaded. These references are used for correcting the images in the processors before electron detection and counting. The Exposure Setup window opens, as seen in previous steps, and we will proceed as before. Again you can click Expert Mode to restore the defaults or save non-default values. Note that in Super-Resolution mode the dose rate is much more tightly constrained too high of a dose rate and electrons are undercounted. Think carefully before using a dose rate other than the default value. In the following figures showing dialog boxes, the default values are given for Super- Resolution mode but the procedure will result in reference images for both Super-Resolution and Counting modes as the counted reference image is derived from the super-resolution image. The dose rates in the dialog box, multiplied by four, correspond to the per-pixel recommended dose rate for reference images. 25

26 figure Settings for K2 Summit linearr mode gain reference figure Expert Mode settings As a reminder, for default values, the beam needs to be dimmed by at least a factor of 10 from the previous step. You may need to change spot size a few steps for instance. 26

27 figure Beam adjustmentt reminder We proceed to dose setup. Super-Resolution mode is more sensitive than Linear mode to dose rate, you should try to get the beam very closee to the recommended level within a few tens of percent is best. figure Beam intensity recommendation After you set the beam, the input parameters are calculated. Do not change these, click OK and proceed through the collection. figure Settings for intensity and number of frames to average When operating at Super-Resolu ution dose rate, this gain reference takes longer to collect than the linear reference. When it is done, the references are saved. figure Linear mode gain reference success message 27

28 Updating Summit Hardware Background Reference Image Only If you wish to update only the hardware background reference image, click the Update HW Dark Reference button on the K2 Direct Detection palette. Updating Reference Images for K2 Summit Counted and Super- Resolution Modes Background reference image (hardware dark subtraction): 1. Click the Update HW Dark Reference button on the K2 Direct Detection palette. figure Update HW Dark Reference button This operation will create an average of many dark reference images and upload it into the camera. The whole process takes just a few minutes. It should be done at least once per day. A fresh hardware dark reference is recommended immediately prior to acquiring a gain reference. A gain reference taken immediately following a fresh hardware dark reference will not be invalidated by a subsequent fresh dark reference. This means that regular dark reference update can save the work of updating the gain reference, which typically takes longer due to the large exposure factor recommended for minimizing DQE derating from gain reference shot noise. The addendum provides a diagnostic for ensuring that fixed pattern noise has not grown too high. When it has, re-running hardware dark reference update will restore it close to its shot-noise limited level. Base Gain Reference Collection Follow the instructions for K2 Base and K2 Summit Linear reference images, above. Key Points to Ensure Optimum Operation Counted and Super-Resolution Dose Rate: Unlike older cameras, for K2 Summit the DOSE RATE (instead of total dose) becomes the first concern in getting good images. For the three Summit modes, the preferred dose rates are shown in the table below. Summit Mode Standard dose (counts per pixel per second) Approximate dose (electrons per pixel per second) Beam Beam counts multiplier vs. counted mode Linear 5,000 10, Bright 30 50x Counted 8 10 Very dim 1 Super- Resolution Very dim 1/4 (counts split between quarters of the sensor pixel) 28

29 In Counted and Super-Resolution modes, image quality starts to sharply decrease around twice the suggested dose rate, and is close to saturation at around three times the suggested dose rate. Note that the dose rates listed for Counted and Super-Resolution modes are the same per physical pixel. Saturation: If you saturate at one exposure length, you saturate at all exposure lengths. This is why we look at dose rate. The camera is always operating at a fixed second exposure (400 frames per second, in Summit mode). The DM exposure length determines how many of those 1/400th second images are summed into the image you see. Other Recommendations It is best to avoid saving data to the desktop. The solid state system drive is fast, but small and can fill quickly. The PC requires approximately 7 minutes to boot. The screen may remain blank during this, do not force it to power down or restart. Do not quit and immediately restart DM. DM requires a few seconds to save data when closed. 29

30 Operation of the K2 Direct Detection System K2 Imaging Linear Mode Operation View Mode Specimen viewing and single-frame acquisition are organized exactly as for traditional CCD cameras. View and Acquire can each be set up with preferences for unprocessed, darksubtracted or gain-normalized. Subareas can be selected. View mode allows the user to observe live (continuous) images on the monitor. The setup has two operating modes which are preset for easy use. Search mode displays a continuous (low resolution) image with the entire sensor field of view. This is to allow the user to quickly conduct a sample search using the camera. Focus mode displays a continuous image at higher resolution than the Search image. Fine adjustment of focus setting can be done with this operation mode. figure View Mode dialog Counted and Super-Resolution Mode Operation Dose Rate Considerations 20 electrons/sensor pixel/sec is considered the maximum dose for quality counted imaging. (16-18 counts/pixel/second) The dose recommended for high quality imaging is 10e - /pixel/second (8-9 counts/pixel/sec.). In Super-Resolution mode, this dose rate corresponds to 2.5 electrons per Super-Resolution pixel per second. (~2 counts/super-resolution pixel/second) 30

31 Contrast suffers above 20 counts/sensor pixel/sec. Due to event crowding, counts saturate just above 30 in Counted mode, so if you re near 30 you have no idea how high the beam really is. Dealing with setting the per pixel dose ratee is a common point of confusion for new users. Dose-Rate Monitor The dose rate monitor displays e - /pix/s when no pixel size calibration is in place. figure Current dose rate on camera After a pixel size calibration has been done, or when TEM communication is active so that magnification is known, the total dose is displayed in terms of electrons/specimen area (Å2). Shutter Delay Setup figure Current dose rate and total dose accumulated on the specimen per image Dose calibrations are on a per-kv basis. That is, if the camera is appropriately linked and knows the microscope kv, the calibration is correct for that kv. A shutter delay can be specified to only start image acquisition after a delay, for example, when the beam has reached a desired level or r the sample has settled. Summit systems can use dose fractionation mode to check the beam level and recommend a delay based on it. In the K2 Direct Detection > K2 Setup window you can specify a pre-dose delay for the K2 Summit camera to wait before collecting dataa and after the specimen has been exposed to the beam. This setting applies to all acquisition modes. figure Pre-Dose Delay setting 31

32 You can set a mode-specific settling delay for the Search, View, and Record modes of imaging the camera. Go to the Settings > Advanced > Advanced View Options window, select the mode from the Setup drop-down menu, and then click Advanced Settings. Different settling delays can be specified for each, but any specified pre-dose delay is added to all of them. Shutter Delay = pre-dose delay + mode specific settling time (View, Acquire, Dose- Fractionation) figure Advanced View Options, settling delay option Defect Correction Defect pixel mapping and removal are performed automatically as part of the reference acquisition process and do not require specific attention. Regular updating of references allows continued high performance of the camera as the sensor ages. For a description of how to monitor changes in the defect corrections as a method to decide how frequently to prepare a new set of reference image, see Addendum. 32

33 Binning and Subareas Dose Fractionation Binning is performed post-acquisition which means that it doesn t increase readout speed by itself in the way that binning does on a CCD. However, it does allow a large increase in dynamic range in exchange for the associated loss in spatial resolution. Because of the very low read noise of the camera, the noise floor remains low. Perhaps the biggest benefit, however, is the possibility of imaging in Counted mode with large signal levels. As an example, Counting mode binned by 6 allows 360 electrons per pixel per second instead of the binned x 1 level of 10 electrons per pixel per second. This is very useful for microscope alignment and stigmation with live diffractograms. Dose Fractionation mode provides straight-to-disk capture at 40 fps (0.025s per frame) in Summit mode, and approximately 10 fps in Base. Dose Fractionation mode is accessed as an alternate to the standard Record mode available in the camera palette. Selecting Dose Fractionation mode in the camera palette switches the system from acquiring single images to acquire a series of images. Dose Fractionation mode splits the total electron dose that would normally be used to acquire a single image into a number of separate frames. For instance a 20 e - /pixel dose in 1 second can be distributed over twenty 50 ms images each with 1 e - /pixel. These frames can be viewed independently and used for different options such as frame alignment and drift correction or in some cases advanced usages such local drift correction within regions of a single frame. 33

34 Drift correction No drift correction figure Comparison of a dose fractionated dataset acquired and summed with alignment (top right) and without per frame alignment (lower left) Setup When Dose Fractionation mode is selected for imaging in the camera palette, the exposure time for the entire series is set by the exposure time in the camera palette. The exposure time for all of the individual frames in the Dose Fractionation series is set independently using the K2 Dose Fractionation palette. You can set a mode specific delay time if a different delay is desired for Search, View, and Record modes of imaging the camera. figure Exposure time setting To configure Dose Fractionationn mode, open the Dose Fractionation Setup palette by clicking the setup icon on the right of the K2 Dose Fractionation palette. 34

35 figure Dose Fractionation mode settings Dose Fractionation mode uses a dedicated solid state raid array to store temporary data, allowing the high frame rate acquisitions enabled by this mode. Use the Dose Fractionation Setup dialog to configure the drives that are used to store temporary data, and to enable or disable multi-threaded accumulation of data from the temporary storage into frames and images. Further options in this palette allow control over the pre-exposure time used in the acquisition. This pre-exposure time can be used to delay the start of data collection until the shutter has stabilized. The final three options allow fine-grained control over output of images from Dose Fractionation mode imaging. The Show image stack checkbox enables or disables the display of the image stack that is collected during image acquisition. The Align images checkbox enables or disables the automatic alignment and accumulation of frames into a single drift corrected image. Finally the Cross Correlation Filter drop down item optimizes the automatic alignment and frame accumulation using a customized image filter. It is important to note, within the context of discussing dose-fractionated alignment and driftcorrection, that one of the most important image quality issues affecting the quality of alignments is the suppression of fixed pattern noise. The fixed pattern noise ratio diagnostic described in the addendum explains how to measure the degree to which one uniformillumination image is independent from the next. This is used to indicate when to acquire a fresh hardware dark reference or gain reference. When the fixed pattern noise ratio is maintained at a low level, the amount of self-correlation in the alignment of successive dosefractionated slices will be minimized and alignment results will be optimized. Shutter Configuration One shutter or beam blanker is required with a standard K2 and also supports a tandem setup camera. An additional shutter may be included for the configured TEM. Using the TEM shutter before the specimen is necessary for low dose work. When the K2 shutter is in use and active, the beam is not present on the TEM screen. The override signal from the TEM provides an interlock signal to disable the shutter when the 35

36 TEM screen is down in the viewable position. The user must be aware that in this case, lowering the screen can expose a beam-sensitive specimen to the beam and take measures to protect the specimen. Dual shutter is not needed for K2. Dual shutter support hass been implemented in the past to support low-dose imaging with CCD cameras. Low dose microscopy as implemented in various TEM control systems, keeps the beam off the specimen until just before the exposure. Due to the fact that the specimen typically moves when the beam is first turned on, it has been necessary to allow a certain pre-exposure of the specimen prior to any attempt to acquire a high-quality image. When using a scientificc CCD, it is necessary to blank the beam below the specimen during this pre-exposure to prevent dose from the pre-exposure from adding into the signal which is to be acquired. K2 does not need this post-specimen blanking because the image is not accumulated on the sensor. The accumulation is digital and can be started at the appropriate moment after sufficient pre-exposure has occurred. Shutter Connections and Signal Configuration To testt or confirm the settings, go to K2 Setup > Configuration tab> Shutter Confirmation. figure Shutter settings tab Base Camera: Linear Mode Operation The Gatan K2 Base camera offers the same radiation-hard high-performance sensor, set up to operate without electron counting in the linear read-out mode. The Base mode of the camera serves two purposes. It enables the camera to be configured without the high-bandwidth processor while still realizing the significant benefits of direct detection imaging. It also allows the camera to be used for high-performance imaging at low magnifications for which sparsification of the beam drive exposure times up to impractical levels. The chart below shows how dose rate recommendations for K22 Summit in counting mode combine with total dose requirements at the specimen and net system magnification to yield expected exposure times. Exposure times are given for typical signal-particle and tomography applications. When K2 Summit exposure times exceed 10s, K2 Base mode with an exposure time less than 1s is recommended. 36

37 Single particle cryo-em Tilt series cryotomography (exposure/tilt) Specimen exposure (e - /Å^2) Camera exposure rate (e - /pixel/sec. ) Nominal mag (kx) Physical pixel size at specimen (Å) SR exposure time (s) Counting exposure time (s) SR exposure time (s) Counting exposure time (s) Use K2 base or linear mode with <1 sec exposure time For linear operation, the dose rate is not tightly specified. That makes it possible to adjust the exposure rate between the single-particle and the tomography modes. The minimum exposure rate is dictated by the read noise equivalent dose, which for the Base Linear mode is approximately 0.1 primary electron equivalents per pixel per second at 300kV (the Summit camera s corresponding value in linear mode is about 1.3 electrons/pixel/second). Thus, users who do not have a processing unit can span the full range of cryo magnification and dose settings without limitation by read noise. Switching Between Base and Summit Modes To switch to Base mode (from Summit mode): 1. Open the K2 toolbox dialog. 2. Press the Switch to Base button. Follow onscreen instructions. 3. Check the acquire and view settings, your preferred settings (Dark Subtraction for instance) may be different between summit and base modes. 4. Switching between Summit and Base causes a temporary loss of temperature control, so wait until the temperature is stable at the cold set point before proceeding with imaging. To switch to Summit mode (from Base mode): 1. Open the K2 toolbox dialog. 2. Press the Switch to Summit button. Follow onscreen instructions. 37

38 Care of Detector 3. Check the acquire and view settings, your preferred settings (Dark Subtraction for instance) may be different between summit and base modes. 4. Switching between Summit and Base causes a temporary loss of temperature control, so wait until the temperature is stable at the cold set point before proceeding with imaging. 5. When the sensor is cold, open the K2 toolbox and click Update Background Reference to update the hardware dark reference. 6. If the camera does not function properly, follow the instructions above for powering down and then powering up K2. If that fails, either repeat the power cycle or contact Gatan for assistance. Temperature Typical operating temperature is -20 C. Monitor the temperature; the Camera Monitor palette indicates the camera temperature. WARNING: Never operate the camera above +15 C, or with a vacuum above 1.5E-5 torr. Check the flow of cooling water periodically. If the flow rate of the cooling water deviates significantly from the value originally set (~15 liters/hour) make sure the lines are not obstructed and adjust the pressure regulator to bring the flow back to the original level. If the water flow stops while the Peltier cooler is on, damage to the camera may result. WARNING: Ensure there are no air flow restrictions around PC, processor, TEC power supply, and monitor. Anneal Cycle Regular annealing (heating to +50 C) of the sensor extends the usable life of the sensor and minimizes contamination of the detector. If the camera is used a lot without warming, the electron beam can harden any contaminants, making the CMOS Detector difficult to clean. We recommend annealing cycle be performed each time the microscope undergoes a cryocycle. Note it is not necessary to interrupt extended data-taking to perform the anneal cycle however it is recommended to anneal the sensor about every 7 days. To perform the Annealing cycle: 1. Select Camera > Temperature. 2. Select a set point of +50 C, click OK. 3. Click the Options button to select a duration other than 24 hours. Usually an annealing cycle of overnight is sufficient. 4. Click the Start button to begin the cycle. The camera cools down to +20 C after the annealing cycle is complete. WARNING: DigitalMicrograph MUST CONTINUE RUNNING during the anneal cycle. Auto Retraction Delay setup (must be longer than longest intended image) 38

39 The camera may unexpectedly retract if you are taking several long exposures (typically 30 sec or longer). This is related to a safety feature which retracts the camera if it is not in use. Set the retraction time to 600 seconds. Insertion Indicators figure Auto Retraction Delay settingg The Gatan Logo on the camera lights up whenn the camera is inserted and the Cameraa Inserted box is checked in the Camera View pallet. figure Gatan logo, and Camera Inserted setting Dynamic Sensor Protection (DSP) Operation Protection is on only while the camera is on and functioning. Does not need DigitalMicrograph (DM) or PC to operate. 39

40 Event detection is instantaneous. System blanks the beam and retracts the camera within 5 ms. System notifies the user (if DM and PC are operational) with a pop-up window message. System releases the beam shutter with the next operation or at user command. SPECIAL NOTE: After actuation of DSP, the system operates normally whether or not the user has cleared the condition which gave rise to the overexposure. Because the K2 system does not know if the event has been cleared, it allows the camera to insert even if the beam is still able to cause damage. It is recommended that the user check the microscope status to ensure that the error state is cleared (turn down beam intensity). The system is designed to minimize damage to the sensor, but it cannot prevent damage in all cases. WARNING: The camera must be warmed up to +20 C when it becomes necessary to vent the microscope or camera chamber. WARNING: Do not operate with poor vacuum. Cleaning: Do not attempt to clean the CMOS sensor yourself as damage can result. If the surface of the CMOS sensor becomes dirty or scratched, contact Gatan for service. Sensor replacement: Recommended to only be done by a qualified Gatan service engineer. Software Install the GMS License using the Gatan license CD. Installation instructions are included with the CD. Install the GMS application software using the GMS installer CD. Installation instructions are included with the CD. Make sure K2 Direct Detection camera is selected in Camera Hardware and TEM. WARNING: Do not adjust IP configuration of K2 Private Network. Fixed IP addresses are intentional. Do not add any connections to K2 Private Network. Do not add any additional PCI cards to PC. Do not modify PC configuration. Do not use X and Y drives for data storage or scratch pad. These drives are dedicated to the sole use of the camera and must be kept empty. SPECIAL NOTE: For optimal performance do not run anti-virus applications during operation of camera. For optimal performance do not use spare 10gigE port during operation of camera. For optimal performance do not load third-party applications other than those recommended or verified by Gatan. 40

41 Overview: Control of K2 Cameraa within DM The PC requires approximately 7 minutes to boot and may display a blank screen for much of this time. Please allow it the necessary time too boot and don t force a power cycle. This may cause disk corruption essentially with the SSD RAID. Screenshot of DM Display for K2 Camera This is a typical representation of how DM should be set up, with respect to the palettes. To configure/choose your palettes, go to Window > Floating Windows. Then select the windows/palettes you would like to display. The checked items are displayed in the DM configuration. figure Typical DM setup DigitalMicrograph Magnification Correction/Calibration The displayed nominal magnification on TEM is for photographic film and has an accuracy of 5-10%. The K2 camera is located on a different plane (height wise) respect to the film camera. As a consequence, the magnification must be calibrated. The calibration is done using Reference calibration samples. At low magnifications: Use a cross grating sample or any samplee with known spacing. At high magnifications: Use graphite or any crystalline samples with known lattice spacing. Use the FFT method. It is very important to make sure DigitalMicrograph software correctly reads the TEM magnification. If the communication betweenn the computer and the TEM is established, the magnification is read automatically. Otherwise, make sure DigitalMicrograph is set to prompt 41

42 Low Magnification the user to enter a value for TEM magnification every time an image is to be acquired. This can be set by choosing the Global Microscope Info window under the Microscope menu. Record an image of a cross grating replica. figure Example of marking a known distance during magnification calibration 1. Choose Microscope > Calibrate image. 2. Follow the instructions on screen. figure Magnification calibration instructions A red line appears on the image. 3. Position it on a feature of known size. 4. Press OK on the Calibrate image window. 5. Enter the correct distance for the selected feature (for example 10 linepairs of cross grating sample where the distance = 10 x 0.463µm) in the Calibration window and select the units. Select the distance marked in the previous figure to perform the magnification calibration. 42

43 figure Calibration settings 6. Press OK. figure Calibration confirmation 7. Click Yes to complete the calibration. The calibration can be checked on the calibration table containing pairs of value, the nominal microscope magnification and the calibrated value. 8. To view the magnification table, select Microscope > Calibrations. figure Microscope Calibration dialog 43

44 High Magnification The microscope calibration dialog shows the table of magnification calibrations stored for the current imaging device. Record a lattice image of the crystalline sample. figure High-resolution image of sample to be used in the magnification calibration 1. Select Microscope > Calibrate image from Diffractogram. 44

45 figure Distance between peaks in the calculated diffractogram 2. To calculate the diffractogram, follow thee on screen instructions. A red line appears on the diffractogram indicating the distance between peaks. 3. Position the endpoints of the red line on two symmetrical diffraction peaks. 4. Press OK to specify the reciprocal unit and the d-spacing (in the corresponding real units) in the next window. figure Calibration instructions 5. Read the calibration instructions and clickk OK., then enter the known spacing between peaks in the magnification calibration in the Calibration settings window. 45

46 Specifications figure Calibration settings Specification K2 Base K2 Summit TEM operating voltage Sensor active area Sensor size in pixels Max. image size in pixels Physical pixel size Binning Sensor read-out Magnification relative to film Sensor read-out speed Transferr speed to computer Image display x full fps 8 full fps 8 full fps kv 19.2 mm x 18.6 mm 3838 x x 7420 Super-Resolution 5 μm 1 8x Any arbitrary y area x 400 full fps 40 full fps 10 full fps DQE performance (300 kv) > 0.30 (peak) > 0.25 at 0.5 of physical Nyquist > 0.7 (peak) > 0.50 at 0.5 of physical Nyquist > 0.06 at 1.25 of physical Nyquist Software Gatan Microscopy Suite including DigitalMicrograph 46