DE-Series Cameras

Our flagship DE-Series direct detection cameras for electron microscopy are revolutionizing the industry, allowing investigators to achieve results they never before thought possible…

Introduction

2010 Microscopy Today Innovation AwardTransmission electron microscopy (TEM) [1] is a powerful technique for visualizing structure at nanometer or Angstrom resolution. However, TEM performance lags considerably behind its theoretical limit based on the physics of electron scattering, especially in cases requiring low-dose imaging. Multiple factors reduce the resolution and signal-to-noise ratio (SNR) of TEM images, including the microscope instrumentation, dynamic specimen processes (e.g., drift, beam-induced motion, charging, radiation damage, etc.), and inefficient electron detectors [2].

With the goal of overcoming many of these obstacles, Direct Electron introduced the first large-format Direct Detection Device (DDD®) in 2008, as the culmination of academic and industrial partnerships working through many generations of sensor development beginning in 2001 [3]. As the pioneer in direct electron detection, our revolutionary DDD sensors were recognized with the 2010 Microscopy Today Innovation Award.

Principle of Direct Detection

Traditional TEM digital cameras use a scintillator to convert primary electrons in the microscope to photons before being detected by the imaging sensor. In contrast, the DDD directly detects image-forming electrons in the microscope without the use of a scintillator. The result is dramatically better resolution, signal-to-noise ratio, and sensitivity.

Principle of Direct Detection

The secret to the DDD’s high performance is its thin sensing layer. Incident beam electrons pass through this thin layer leaving an ionization trail that is collected and either integrated or counted in pixels. Because the layer is so thin, lateral charge spread is minimized resulting in higher resolution than other detectors. Furthermore, image distortions are minimized since the DDD directly detects electrons without having to transfer signal through fiber-optic or optical lenses.

A second innovative feature of the DDD is its high frame rate, with no dead time between frames. This high frame rate delivers intrinsic dose fractionation during image acquisition, which can be exploited for motion correction, damage compensation, and other image processing techniques. High frame rate data acquisition also enables challenging TEM applications such as in situ TEM.

Unique Features

Direct Electron’s cameras are elegantly designed with careful consideration given to each feature and component to maximize user productivity.

As funding levels tighten, productivity and throughput become increasingly critical for successful investigators. Our DE-Series Cameras are uniquely designed to deliver the most flexibility and versatility, the most productivity-boosting features, and exceptional reliability.

Direct DetectorDirect ElectronAlternative #1Alternative #2
Pixel Size (μm)6.0 - 6.5145
Array SizeVarious sizes up to
8192 × 8192
4096 × 4096
(17% smaller than DE-20;
300% smaller than DE-64)
3838 × 3710
(38% smaller than DE-20;
370% smaller than DE-64)
Pixel Design>3T
(CDS to reduce noise)
3T
(High noise)
3T
(High noise)
Frame ShutterGlobal or rollingRolling
(Introduces gradients across images)
Rolling
(Introduces gradients across images)
BackthinnedYesYesYes
Radiation HardnessExcellentExcellentExcellent
Optimal Exposure RateConsistent performance at any exposure rateWorse performance at low exposure ratesOnly very low exposure;
< 3 e-/pixel/s is optimal;
at 10 e-/pixel/s performance drops ~30%[4]
Dynamic Range> 400 e-/pixel/s~160 e-/pixel/s[5]At ~60 e-/pixel/s quantum efficiency decreases ~60%[6]
Frame RateFlexible;
Up to 30 fps, full frame;
1000+ fps for subarrays
18 fps;
Typically users may only save 7 frames
40 fps transfer rate to computer
Survey SensorIntegrated near-axis 2k × 2k survey sensorSeparately purchased and maintainedSeparately purchased and maintained
Exposure MeasurementIntegrated Faraday plate above the DDD sensorNoneNone
Sensor Protection ShutterYes, patentedNoneNone
RetractableYes, mechanicalYesYes, pneumatic
Sliding O-RingsNo, minimized vacuum spikes during insertion/retraction???Yes
Mounting PositionFilm chamber (JEOL) or bottom-mountBottom-mountBottom-mount
Magnification FactorFilm chamber: 1.0×
Bottom mount: ~1.4×
~1.4×~1.4×
Data Acquisition SoftwareClient-server model with multiple options:
Micro-Manager (open-source);
DE-IM (full-featured);
DE-StreamPix (in situ);
Open API for integration in custom software.
ProprietaryProprietary
Movie Processing SoftwareFlexible and open-source; patented algorithmsNoneProprietary

Note that the features and specifications shown above for Direct Electron cameras are typical, and are subject to change. Please CONTACT US for more specific information. The features and specifications shown above for “alternative” cameras are hypothetical examples similar to other direct detection cameras available on the market, based on information publicly available on October 25, 2014. For specific information about actual products available from other manufacturers, please contact them.

Camera Operation Modes

Images of the edge of a beamstop collected on a DE-20 camera at 300 kV in integrating and counting modes.

Integrating and Counting Modes

The high speed and ultra high single-electron sensitivity of direct detection cameras enable two different modes of operation:

  • Integrating mode – The total signal deposited on the detector by each electron is summed to form the final image. Due to its versatility and speed, this is the most popular mode of operation for a wide range of applications.
  • Counting mode – Each incident electron is individually detected, isolated, and localized. In this mode, every incident electron is consistently represented by a single count in a single pixel (or sub-pixel). While this mode represents the highest possible performance, it is only effective in very low dose applications.

Our DE-Series Cameras allow users the choice between either integrating or counting mode.

Performance

Detector performance can be evaluated in several ways, including (from most fundamental to most practical):

  • Modulation transfer function (MTF)
  • Detective quantum efficiency (DQE)
  • Carbon film Thon rings
  • Actual experimental results

Modulation transfer function (MTF)

DE-20 MTF

The modulation transfer function (MTF) of the DE-20, operated in both integrating and counting modes. The MTF curves were calculated using an image of a beamstop, using FindDQE. The theoretical maximum MTF is shown as a dashed gray curve.

Modulation transfer function (MTF) is one of the most widely used scientific methods for describing optical performance. MTF measures the contrast (magnitude response) across all spatial frequencies up to Nyquist frequency (the theoretical resolution limit of the detector). Formally, the MTF is defined as the magnitude of the Fourier transform of the point spread function (PSF) of a detector. Note that the MTF describes detector performance under absolutely ideal experimental conditions, which are not always applicable in practice.

For TEM detectors, the MTF is typically calculated by collecting an image of a straight edge (such as the microscope’s beamstop), and then using software to calculate the edge spread function (ESF), which describes the amount of blurring apparent on the straight-edge image. The Fourier transform of the derivative of the ESF is the MTF.

Simply, the MTF curve can be thought of as resolution on the X-axis and contrast on the Y-axis (higher is better).

Detective quantum efficiency (DQE)

DE-20 DQE

The detective quantum efficiency (DQE) of the DE-20, operated in both integrating and counting modes. The DQE curves were calculated by dividing the square of the MTF curve (calculated from an image of a beamstop) by the normalized noise power spectrum (calculated from the difference of two blank flat-field images). DQE(0) was calculated from the difference of two blank flat-field images using the noise binning method. The theoretical maximum DQE is shown as a dashed gray curve.

For TEM cameras, the Detective quantum efficiency (DQE) is probably the most well-known metric for assessing performance. DQE measures the combined effects of the signal (related to contrast, as described by the MTF) and noise performance of a detector. Similar to the MTF, the DQE describes detector performance under absolutely ideal experimental conditions, which are not always applicable in practice.

For TEM detectors, the DQE is typically calculated using three components: (1) the modulation transfer function (MTF), which describes the signal performance of the camera; (2) the normalized noise power spectrum (NNPS), which describes the noise performance of the camera; (3) the DQE(0) value, which describes the overall efficiency of the camera in accurately detecting each incident electron. The equation is: DQE = DQE(0) x MTF2 / NNPS.

Simply, the DQE curve can be thought of as resolution on the X-axis and relative signal-to-noise on the Y-axis (higher is better).

Carbon film Thon rings

While the MTF and DQE are useful metrics for evaluating the fundamental performance of a TEM camera, they neglect to account for many critical factors that affect practical camera performance. For example, a camera that requires a long exposure time will be much more affected by microscope and specimen instability. It is therefore instructive to examine detector performance empirically, under practical experimental conditions.

The power spectrum calculated of an underfocus image of a specimen with sufficient scattering power shows concentric Thon rings, which are an effect of the microscope’s contrast transfer function (CTF), which modulates the Fourier transform of the object in a defocus-dependent way. Even with a theoretical perfect detector, the signal-to-noise ratio (or relative intensity) of each successive Thon ring would decrease with spatial frequency, due to aberrations and imperfect coherence in the microscope. In practice, the detector used to record the image also damps the signal of the Thon rings. At some point, the combination of the microscope and the detector will dampen the signal of the Thon rings sufficiently that they disappear in the noise. The resolution limit of the microscope and detector may be approximated as the maximum spatial frequency where Thon rings remain visible.

To evaluate the performance of the microscope and detector together for collecting individual images, it is common to collect slightly underfocus images of amorphous carbon film. Carbon film has similar scattering power as biological macromolecules per unit area, but since carbon film is distributed evenly across the entire field-of-view, the total scattering power in the field-of-view is sufficient to visualize the resolution limit of the instrument.

The exceptional high-resolution performance of our DE-Series Cameras means the camera is no longer the primary limiting factor for TEM resolution. For example, the figure below shows very high resolution bright-field TEM imaging, with Thon rings out to 2.3 Angstroms resolution (92% Nyquist).

Carbon Film Thon Rings

The Fourier transform (left) and rotationally-averaged signal-to-noise ratio (SNR) of an image of carbon support film on a graphitized carbon calibration grid. The image was collected using a DE-20 operating in integrating mode, on a JEOL 3200FSC microscope (300 kV). The nominal magnification was 50kx, resulting in sampling of 1.06 Angstroms/pixel. Defocus was measured as -0.77 um. The image was motion corrected using the Direct Electron DE_process_frames.py script. To aid in visualization, the Fourier transform was normalized by the measured MTF of the DE-20. Impressively, Thon rings are visible up to 2.3 Angstroms resolution (92% Nyquist).

Sub-3 Å Resolution 3D Reconstruction

Cryo-EM density map and model of a variant of AAV (adeno-associated virus). The resolution of the full map was measured to be ~2.9 Å resolution. Data was collected on a DE-20 Camera mounted on a Titan Krios, running Leginon for automated data acquisition (which took ~3 days). Courtesy of Scott Stagg, Florida State University.

Actual experimental results

While fundamental metrics like DQE can be a helpful first step in evaluating camera performance, there is only one way to take into account the totality of a camera’s capabilities and limitations: actual experimental results. Whether you are after super high resolution, large volume imaging, or vizualization of dynamics, our DE cameras work efficiently to deliver the results you need, quickly.

One of the most challenging experiments in the field of electron microscopy is single-particle biological cryo-EM. Besides the difficulty in preparing good specimens, experimental factors such as radiation damage and beam-induced specimen motion make it very difficult to generate reproducible atomic models of biological macromolecules. The camera is absolutely critical in these experiments, since investigators must try to get the best possible data in order to push the limits of resolution. Our DE cameras have broken the 3 Ångström resolution barrier for single-particle cryo-EM, with as little as three days for data collection!

For more information about how DE cameras immediately accelerate research in across a wide variety of fields, see our applications overview and our list of recent publications.

More information

For more information about the specifics of each camera, please click the name of each camera in the list of available cameras below, or CONTACT US. We are happy to provide whatever figures or information you need to help you apply for funding, decide upon and purchase a camera, teach classes, etc.

DDD Development History

Our current Direct Detection Device® sensors are the culmination of over a decade of continuing development, in partnership with academia and the US National Institutes of Health (NIH).

Generation12345678910
Commercial Product(s)Development Systems (Academic)DE-12DE-12DE-12 (Gen2)DE-20DE-16,
DE-64
Year2002200220032005200620082010201120122014
Pixel Size (μm)205, 10, 20, 305556666.46.5
Array Size (pixels)128 × 128Various550 × 5121024 × 1024560 × 4604096 × 30724096 × 30724096 × 30725120 × 38404096 × 4096,
8192 × 8192
Pixel Design3T3T3T3T>3T>3T>3T>3T>3T>3T
Feature / Milestone4 quadrants4 sectionsSingle pixel design; MTF/DQELarger format; cryo-tomographyADC per column; electron countingLarge format; commercializedFaster; more radiation hardenedBackthinned; improved at 200 kVLarger; reduced noise; improved MTFUltra-large format; improved SNR, dynamic range, and radiation hardness
Developer & FundingUniversity of California, San Diego
NIH RR018841
Direct Electron, LP
NIH RR024964 & NIH GM103417

 

Currently Available DE-Series Cameras

As a result of our continuing development, we are proud to offer a range of direct detection cameras to the TEM field. Reflecting our focus on empowering users for a broad range of experiments, all of these cameras deliver the flexibility and versatility that a modern TEM facility demands. However, the features and specifications of each of these cameras are optimized for certain applications, to push your productivity to its limit.

For more information, please click the name of each camera below, or CONTACT US.

CameraDE-12DE-20DE-16DE-64
DDD Generation891010
Pixel Size (μm)6.06.46.56.5
Array Size (Pixels)4096 × 30725120 × 38404096 × 40968192 × 8192
Megapixels12.619.716.867.1
Maximum Frame Rate with Full Array40 fps (bin 1×)
75 fps (bin 2×)
32 fps60 fps (bin 1×)
120 fps (bin 2×)
30 fps
Integrated Survey SensorNone2k × 2k2k × 2k2k × 2k
Typical ApplicationsEconomical general-use direct detection cameraCryo-EM; DTEM; other low-dose applicationsIn situ TEM; materials scienceCryo-EM; high-throughput TEM

 

References

[1] For more information, see overviews at the Nobel Prize Educational Site and the University of Liverpool Matter Resource on TEM.

[2] Glaeser RM, Hall RJ. “Reaching the information limit in cryo-EM of biological macromolecules: experimental aspects.” Biophysics Journal 100, 2331-2337 (2011). View Publication.

[3] Jin L, Bilhorn R. “Performance of the DDD as a direct electron detector for low dose electron microscopy” Microscopy and Microanalysis 16, 854-855 (2010). View Publication.

[4] Ruskin RS, Yu Z, Grigorieff N. “Quantitative characterization of electron detectors for transmission electron microscopy.” Journal of Structural Biology 184, 385-393 (2013). View Publication.

[5] McMullan G, Faruqi AR, Clare D, Henderson R. “Comparison of optimal performance at 300keV of three direct electron detectors for use in low dose electron microscopy.” Ultramicroscopy 147, 156-163 (2014). View Publication.

[6] Li X, Zheng SQ, Egami K, Agard DA, Cheng Y. “Influence of electron dose rate on electron counting images recorded with the K2 camera.” Journal of Structural Biology 184, 251-260 (2013). View Publication.