Can you put some more light on Echelle spectrographs in case of Raman Spectroscopy?
Echelle spectrographs use the 2 Dimensions of the CCD to image a wide spectral range while maintaining very good spectral resolutions.
Traditional Echelle instruments have low throughput and lead to narrow spectral channels of a few pixels high on the detector, limiting the overall signal collection. These systems are traditionally dedicated to Light Induced Breakdown Spectroscopy (LIBS) which usually features strong optical emission.
With commercial Echelle spectrograph now giving throughput comparable to 1/4m Czerny-Turner spectrographs, these become viable tools for Raman probing when coupled to EMCCDs.
Moreover the large spectral range allows in some cases multiplexing of several Raman signals induced simultaneously or in short sequence from different lasers, increasing screening throughput capabilities.
Return to Question Index
Does Raman and Electron Multiplying CCD detectors have nanoscopic capabilities?
EMCCDs have indeed been widely used for Raman nanoscopic applications.
Nanoscopy ‘capability’ is principally given by the microscopy instrument and the data processing techniques.
In spectroscopy, Nanoscopy can mean several things: when combining Raman with Atomic Force Microscopy, technique known as Tip Enhanced Raman Spectroscopy (TERS), nanometer-range Raman spatial information can be obtained.
Raman probing of nano-scale structures such as carbon-nanotubes, nano-crystal, nanoscopic dyes or nano-wires is also called Nanoscopy. EMCCDs have been used successfully for their enhanced detection capabilities and Andor could provide references of such work on request.
During live measurements, how can I determine if I am using a light source for which the EM gain will provide a better signal and how do I determine how much of this gain I should use?
That will depend on the photon flux.For fast photon starved application, EMCCD will have no equivalent.For higher photon flux and where acquisition speed is not a constraint then CCD will be best suited.
It is to be noted that Spectroscopy EMCCD sensors have both conventional CCD and Electron Multiplying capabilities on 1 chip, allowing optimization detection performance for a wide range of signal intensity range by toggling through software between 1 mode or the other, and adjusting the gain in real-time also through software.
How much QE drop at -100 °C compared to say -20 °C, is deep cooling needed if we have high speed acquisition?
Irrespective of technology, in the UV-Vis region, CCD/EMCCD Back-illuminated Quantum Efficiency can drop by <5% between room temperature and -100 °C. Above 800nm, back-illuminated sensor Quantum Efficiency can drop by 20%, and Back-illuminated Deep Depletion by 30%.
High speed acquisition means shorter exposure time, so dark current is not going to be the main limiting factor.Typically air cooling at -70 / -80 °C is sufficient unless working in photon counting mode.
How uniform is the spectral response of the EMCCD?
The key difference between CCD and EMCCDs lies in the readout interface. The active areas of both technologies have identical properties in terms of blemishes, dark current, uniformity, sensitivity, etc.
EMCCD readout register, where gain by impact ionization is achieved, can suffer from ageing effect. This ageing will result in a slight loss of gain efficiency over the year, but this will take thousands of hours of use to experience a meaningful difference.
Is EMCCD good for acquiring Raman spectra from samples that have associated sample fluorescence?
EMCCDs will amplify any signal from the sensor area prior to passing through the gain register, so this includes both Raman and fluorescence indistinctively. The EMCCD could just lead to faster acquisition of overlapped Raman and fluorescence.
Enhanced Raman techniques can provide the needed intensity discrimination of Raman and Fluorescence signals.
In the case of pulsed Raman, Intensified CCDs, with their optical window down to nanoseconds as well as their amplifications capabilities can isolate useful signal from millisecond integration of the fluorescence signal.
Many biological species are inherently weak Raman scatterers, under nonresonant conditions. Is the EMCCD still useful for mapping if exposure times > say 5 secs are required?
It is all depending on the amount of photons captured during the 5 seconds exposure, the noise floor level of the detector and the degree of quality of the signal required (quantitative or qualitative analysis).
Taking the example of 785nm Raman, a Deep Depletion CCD will provide the best Quantum Efficiency, but however suffers from higher dark current.
At exposure times such as 5s and low photon regime, a Front-Illuminated EMCCD with moderate gain will prove as good, if not better in terms of Signal to Noise ratio as the Deep Depletion device. Deep Depletion device have higher dark current – in excess of 1 order of magnitude from non-depleted CCDs.
In the visible, a Back-thinned EMCCD will clearly outperform CCDs in low-light level conditions.
On slide #10 of the webinar, an example is shown where exposure time can be reduced by as much as x50 with an EMCCD compared to a CCD to obtain similar quality of information.
If the quantity of signal during these 5s integration is however reasonable, then conventional CCDs would be the better technology to use. It is to be noted that Spectroscopy EMCCD sensors have both conventional CCD and Electron Multiplying capabilities in 1 chip, allowing optimization detection performance for a wide range of signal intensity range.
Regarding data analysis, is there any special software with the instruments or should we go to second or higher order like MCR data analysis?
Andor can provide either software that controls the camera and spectrograph and provides raw data, or a Software Development Kit (SDK) under platforms like Matlab, Visual Basic or Labview for users own application development and data processing.
What about the dynamics of EMCCD compared to normal CCD?
Below is a graph showing a comparison of Electron Multiplying mode and Conventional CCD in terms of dynamic range (Sensor Well Depth / Detection Limit) for a readout speed of 1 MHz. For moderate gain the EMCCD has a dynamic range >15 bits where CCD performance lies at >14 bits.
At higher readout rate, say 5MHz, the CCD dynamic range will drop to ~ 13bits while the EMCCD will be >14 bits.
At 10’s of kHz readout speed the conventional CCD will provide the best Dynamic range >15bits.
To have a global picture of CCD versus EMCCD performance it is important not only to consider dynamic range versus readout speed and gain but also Signal to Noise versus photon flux level.
What are the cooling methods of the EMCCD, TE-Water? Does this effect noise?
Because EMCCDs amplify any noise or signal that comes from the active area of the CCD, and dark current is generated at the CCD level, it is important to cool the CCD.However in the case of short exposure times, the dark current contribution is going to be very low.
Andor Spectroscopy EMCCDs use Thermo-Electric cooling in a vacuum package around the sensor, to achieve sensor temperatures as low as -100 °C when assisted with water and -80 °C when simply air-cooled. Dark current figures as low as 0.0006 electrons per pixel per second for back-thinned sensors can therefore be obtained.
What are the highest spectral rates that you can achieve with these detectors?
Spectroscopy type EMCCD’s can achieve up to 1,300 to 1,500 spectrum in sub-array mode.
What are the typical "spurious charge" levels for an EMCCD, and how does SC vary with gain?
Spurious Charge, or Clocking Induced Charge (CIC) occurs as a result of impact ionization during charge transfers either in sensor pixel or shift/gain registers.
All CCDs have CIC – it is only with EMCCD technology that we are seeing them as they are amplified as any other signal and noise generated.
CIC, alongside shot noise, represent the ultimate detection limit of the CCD.
For Andor Spectroscopy EMCCD sensors, at x1000 EM Gain, minimum exposure, -70°C CCD cooling,typical figures are of the order of 0.005 events/pix – independent of time and unlike mistakenly stated during the live event. Decreasing the gain will improve this figure, especially in the case of Raman where moderate gains of a few tens to a couple of hundreds are needed.
What are the maximum dimensions for EMCCDs and how many pixels are needed for 1 channel?
Dedicated EM-CCD Spectroscopy sensors are 25.6mm wide by 3.2 or 6.4mm high with 16μm pixels.
EMCCDs are traditionally used for Imaging Life Science with square chips that are currently limited to 13 x 13mm useful field, but offer very fine spatial resolution (pixel size as low as 6-8 μm).
In terms of channel height I assume this is targeted to the Multi-track and hyper-spectral fiber optics applications.
Signal track height will essentially depend on:
- The size of the point source, either a fibre optic core or a fine focused spot at the entrance focal point of spectral instruments – typically from 50 to 200μm- The magnification of the spectral instrument - Czerny-Turner spectrographs vertical magnification like the Andor’s Shamrock 500i is 1.- The pixel size of the detector
So assuming for example the use of a 100um core fibre optic, a vertical magnification of 1 through the spectral instrument and a 16μm pixel size detector, the channel height would be 6-7 pixels.
What do you think of our website? What do you think of our website?