Before you send back your laser, check if your  TCSPC  detection is, in your case, slow enough, some of the devices have only a detection window for 10ns or 100ns (longest scale one can measure) the rep rates in his cases are not crucial, but if you are sure that it works try to get a  laser with tunable rep rate form 2MHz to some kHz.   Check also the following paper, might be useful: "Size- and temperature-dependence of exciton lifetimes in CdSe quantum dots" in Phys. Rev. B 74, 085320 (2006) by C. de Mello Donegá, M. Bode, and A. Meijerink, they measured similar CdSe QDots

As you are not able to change the TAC range /or repetition rate, you can change the delay ( coaxial and sync, also if you have extra channel in the left part from your present IRF position) so that your decay will get more space. Good luck

I believe that, unfortunately, you need a lower repetition rate in order to record the fluorescence decay of this quantum dot. If the time between pulses is not enough for the fluorescence to decay completely, all of the decay curve (not only the longer portion) will be distorted. Hence, none of your decay times is reliable under the conditions that you are currently utilizing.  

I think too that you need an other laser with a lower repetition rate to record the full luminescence decay of our quantum dots. As rule of thumb you can say that you need a time window which is 10 times higher than the decay you have to measure to get reliable data. An extrapolation of the decay especially if you have such multiexponential decays because the obtained data is not reliable  and furthermore not comparable with others.

I do not know much about quantum dots but maybe you could quench this emission in a controlled way so you could get short enough lifetime to fit your measurement time-window?

Also, if you know the "dark" count rate of your detector for the same conditions like your experiment (in cps) + time of your experiment you should be able to predict your real background signal.

Hey Sergio!

Could you please elaborate on why would the decay curve that I am getting now would be distorted? Some extra information which could help:

I am expecting 3 lifetime values for my fluorophores. A component with lifetime around 1ns, one with 10ns and another around 30-50ns(or even longer). Thank you!

Saurabh

Hi Saurabh. I believe that the first portion of the decay curve is distorted because it is superimposed upon the last portion of the previous decay curve. Ideally, the first portion of your decay curve should correspond to the deactivation of the quantum dots that were excited by the last laser pulse. However, under your experimental conditions the first portion of the decay curve corresponds to the deactivation of the molecules that were excited by the last laser pulse plus the deactivation of the molecules that were excited by the previous laser pulse. Hence, the decay curve may seem to be more complex than what it really is (i.e., more lifetimes are needed to fit it). Nonetheless, the distortion may be negligible if the population of molecules excited by the previous laser pulse is insignificant in comparison with the population of molecules excited by the last laser pulse. If I haven't been clear enough, please let me know.   

PS: Can you upload the decay curve for us to see it?

Hi, Sourabh. I fully agree with Dennis. For your sample you need at least 500ns (some times 20 times higher is more preferable) TAC range and 2kHz (should be less than the TAC range) repetition rate laser. For higher repetition rate laser you can use reverse TAC mode & 500ns TAC range(laser trigger to STOP and signal to START). Also, you may have to change the delay to bring the decay profile to required channel.

Thank you Sergio and Subhas and everyone else fro your inputs!

Sergio! I get your point. But consider this: Dead time for my detector and electronics is around 100 ns. That means that every time a photon gets counted, I have a 100ns window where no photon is getting counted. Therefore only those photons having de-excitation time higher than 100ns actually have a chance to compete with the photon generated from the previous pulse. I still think that the decay curve I got would be distorted but the degree of distortion (because of the 50ns lifetime component) might not be that severe to completely dismiss the lifetime data. However, I know that the distortion would be much bigger if lifetime is of the order of 100ns or even bigger. And, therefore, I am planning to send my laser back to change one of the rep rate to 500KHz (2microsec TAC)or 1MHz (1 microsec TAC). 

I am attaching a representative decay profile of my QD's with the IRF and fitted decay here. Now, if you look at the decay background fitted by the software, it is much higher (580 counts) than what I get when I collect scattering from the solvent (toluene in this case) for the same amount of collection time (150 counts with some scattering), which makes me to believe that the higher lifetime component is not reliable and that is why I was wondering if tailfitting the later portion of the decay with decay background (as 150 counts) would give me somewhat qualitative lifetime value for now.

Thank you!

Saurabh

Thanks for the file, Saurabh. I think the longest lifetime is probably about 7-10 ns. Maybe the distortion is insignificant, but I can't tell for sure. I believe that the detector dead time is not really relevant in this case because the detection of a photon is a rare even in TCSPC. In order to avoid photon pile-up, your count rate should be about 1% of your excitation rate or even less than that. So, under ideal conditions, the detectors "sees" less than one fluorescence photon for every one hundred pulses. 

Hi Saurabh,

Of course the "clean" way would be to decrease the rep rate of your excitation light. There is no electronic way (TAC, delay lines) to avoid that the sample is indeed excited too soon to let its fluorescence decay relax to zero.

However, if you cannot change this, there are tricks to partly deal with such situations (i.e. extract long lifetimes, with however reduced accuracy). I developped it while using synchrotron radiation as the light source (don't know why, but the machine staff definitely refused to increase the storage ring diameter...)

1) make sure that your TAC window is sufficient to capture a little more than a complete excitation period (you must see too rising edges of the IRF/fluorescence decay in the window). This may also require decreasing electronically the frequency of the synchronisation signal that gives you the timing for TCSPC (ask some good electronician about this), which may reduce your collection efficiency (only one photon over N will have the proper synchronisation signal).

2) Collect your fluorescence decays and IRF at similar counting rates (important, because the background noise often depends on the counting rate). If you cannot, do a small study of background vs counting rate using the IRF, and use it to model the relationship between both. Your can do the same with a reference compounf in place of the IRF.

3) Get the background from the IRF (in a clean region before the rising edge) and scale it to the (counting rate & accumulation time) of your decay.

4) Analyse your fluorescence decays with IRF reconvolution using the following model :

- assume multiple excitations separated by the exact period of your excitation pulses (i.e. use multiple periodic IRFs to reconvolve your model). Use the approximate value of your sample lifetime, and the repetition period, to evaluate how many consecutive excitations you need to include

- block the stationary background to the value computed in (3). You can also try to optimise it during the fit, but the problem is highly degenerate with the value of long lifetime components, so it is unlikely that you get a good accuracy on the later.