I have a very small (~200um) GaAs p-n junction device (with AlGaAs QW) and would like to collect PL images from, using an optical microscope system (300x total effective magnification is required) + a Si-CCD camera (which has a pass filter to block excitation laser light in front of CCD). The CCD is currently not cooled (plan to buy a cooled CCD later).
Any input on the following items is greatly appreciated:
1). Which laser wavelength and what power level should be used for such PL(~820 nm peak) imaging? 690 nm is preferred (30mW ~$1000) while high power 532 nm laser appears to be more difficult to get and of course is more expensive.
2) can I simply focus the excitation laser onto my sample from outside the microscope optical system, or do I have had to let the excitation laser go through the objective lens system (i.e replace the current LED light source by 690 or 530 nm laser source)?
3) If the excitation laser needs to go through the microscope optical system, will excitation laser cause damage to optical lenses?
Thank you in advance.
I have some experience in PL measurement for GaAs. Hope my inputs help.
(1) In my opinion, 690 nm laser would be a good option. Longer wavelength laser is normally better for PL because of less heat generation (from phonons). Extra heat can cause red-shift of the PL peaks. Indeed, you can use a ND filter to lower the pump power. Both red and green lasers are commonly used of PL, and 30 mW is far enough.
(The GaAs peak is normally at 875 nm, room temperature. You were referring to low-temperature PL?)
(2) Yes. This is a must. You would need a objective lens. Otherwise you cannot collect the output PL emission.
(3) The concern is that the pump laser can damage the CCD if the power is high. Saturating the CCD for a certain period of time can cause the damage of pixels. You can put a bandpass filter in front of the input slit/window of your microscope optical system.
Dingkun Ren , Thank you so much for the quick response.
1. May I know what is the minimum laser power you could recommend for 690nm laser diode excitation source? 30mW is the max output power of this laser diode from Newport, it can be user adjusted from 0 to 30mW.
(Thank you for the information on PL position, very helpful)
2. I have two possible options here:
(A) >>focus the laser spot onto the sample from outside the objective lens to excite PL (note excitation laser beam will NOT go through the microscope optical system).
>> collect the PL via objective lens (as you suggested, this is a must)
>> let PL+scattered excitation 690nm laser light pass the 800nm LP filter to filter out the scattered excitation laser light, while the PL (>800nm) will be able to pass through.
>> project the passed through PL signal onto CCD camera to see the bright/dark contrast for less defective and more defective regions.
-see the attached schematic light path
Is this possible?
The 2nd option:
(B) replace the current microscope's LED light source system with this 690nm laser beam (i.e. let the laser beam pass the beamsplitter >> go through the objective lens >> hit the ~200um GaAs sample surface >> Generate PL >> collect PL +scattered laser light via objective lens >> pass LP filter to block scattered excitation laser light>>project the PL only image onto CCD).
Since there is some hardware modification requirement for this option, it also has potential damage risk to the objective lens if 690 nm laser power is too high. I am trying to avoid this option if possible.
3. Thank you for the suggestion on CCD saturation issue.very helpful
Xilin Peng, no problem. Please let me know if anything I can help.
1. I would say tens of mW is okay since the GaAs signal should be strong enough. Of course, the signal-to-noise ratio also depends on the CCD camera you have. You probably want to check the datasheet.
You can briefly do a calculation and see if you can get a good signal from the CCD:
External quantum efficiency of PL: ~10% (probably even less since extraction efficiency is very low since most of PL emission cannot be collected)
Loss through optical path: ~5%
The laser input power 30mW --> excitation on sample ~30mW*(1-5%) = 28.5mW --> no. of photons on samples N=28mW/(hv_red/q) --> power collected by lens N*10%*(hv_GaAs/q) --> power collected by CCD N*5%*(hv_GaAs/q)*(1-5%)
Si CCD should be able to give you a strong signal at room temperature. I would say not less than 10mW since you might later want to do some power-dependent PL to study the materials etc. Meanwhile you can purchase a few ND filters, e.g. ND0.5, 1.0, 2.0, 3.0., if necessary. It's always good to use different power to calibrate the system and make sure the input power density is reasonable.
2. Both are good.
(A) People call this as side-excitation or something equivalent. This approach can give less the power density illuminated on sample and the laser beam can cover a larger area. You can definitely try it first since the setup is much simpler than (B).
(B) I double 30mW will damage the objective lens. You can check with the manufacturer.
- Badges
- Science topic
- Similar topics
- Solid State Physics
- Devices