You’re not. I’ve redone the tables but didn’t update here. Still not sure how useful that is going down to 1ng on a 50:1 split.
This is probably the lowest amount of signal I can reasonable detect
Anyway, dogs passing today so this is on hold for a few 
Sorry to hear my man. Wishing you the best, talk in a few days.
I wanted to make sure this was the case. Please by all means check my math because I could have overlooked it!
To one vial add 100 uL restek standard, allow to evaporate. Add 200 uL IS solution to this vial and seal. Then perform the following:
With each vial filled with IS solution, everything should be at the same concentration level of IS per uL. Am I mistaken? If I take 10uL of solution from the master vial above, the amount of IS should be the same, per uL as the vials with only IS solution. Maybe my brains fried, it very well could be.
I don’t mind being wrong, if I’m doing something stupid let me know!
Side question, is there any way I can change my view to match yours? I’m really at a loss at how the baselines compare with the differences in signal measurements.
I trust my brain could be fried, maybe double check my math?
Internal standard prep:
1.10919 mg per ml
11mcg per ul
1109.19ng per ul
With a 50:1 split that should be 22ng on column if I’m not mistaken?
For sample prep, if I aimed for 100mg compound to 40ml IS solution, assuming 100% potency that would be 2500ng per uL and at a 50:1 split should be 50ng on column. If thats accurate, i probably at the very least need a calibration point higher, would you agree? Maybe another point at 100uL standard evaporated and added to 100uL IS for a 1000ng/ul calibration point, and 20ng on column. Is that close enough or should there be calibration point nearer the 100% purity mark?
Seems my calibration table favors the lower end, i guess it’ll be interesting to see exactly how low the LOD will end up being. It seems kinda neat to me, this sample of flower I ran in the above looks like it has trace d8.
Maybe I’m overthinking it
Diving in, still using isolate as a base for the test
Will upload the chromatograms when done. If my bad math is right, 1ng on a 50:1 split should end up being about 20ppb, I’m very curious to see if its can be resolved.
As I sit here waiting for the injections to finish and for the next I realize my next task is getting the auto sampler running
So the issue here is you are pulling from two different concentration stocks, and changing the ratio at which they are used.
In this case:
Master Stock is @ 500µg/mL IS
IS Solution is @ 1000µg/mL
Lets say we used 0.01mL of the Master and 0.09mL of the IS Solution
0.01 * 500µg/mL = 5µg
0.09 * 1000µg/mL = 90µg
95µg in total / 0.1mL total = 950µg/mL
In the second case we will keep the final volume the same but change the ratio of fill. Let’s use 0.09mL of the Master and 0.01mL of the IS solution:
0.09 * 500µg/mL = 45µg
0.01 * 1000µg/mL = 10µg
55µg in total / 0.1mL total = 550µg/mL
So while you are effectively hitting your cannabinoid targets by changing the amount of master and then dilution with IS solution, you are unfortunately throwing off the IS concentration.
A fix for this would be to make a secondary IS solution that is the same concentration as the master IS (500µg/mL)
You just need to create a report
Edit: Note (1,2 1,4 referring to grid layout (row, Column)
Is that not corrected for by evaporation of the 100uL standard stock, and replacing it with 100uL of IS solution?
Is that for the master?
If you are only adding back in 100µL and not 200µL then the master would be at 1000µg/mL not 500. In which case the IS numbers would line up but the concentrations for all analytes would be 2x.
Edit: Unless maybe you still mean the 100µL of standard is being replaced by 100µL of IS solution but still the final volume being 200µL? In this case the cannabinoids targets are correct but the IS is at 1000µg/mL as it has not been diluted, but the IS will be consitent.
Specifically, the idea is to take 100uL of standard in methanol, evaporate off the methanol, add in 200uL of IS solution to make the 500ng reference point. Then diluting down from there.
Am I way off?
No, that works. In my calcs I was considering the 0.2mL final volume to be 0.1 from the standard and 0.1 from the IS. That was incorrect, my apologies.
The IS concentration will be 1000µg/mL and the cannabinoid concentration will be 500µg/mL. The IS will be consistent and the math lines up for everything else I have checked. So I think you are all good.
Here is a test run I did, not sure what entirely to make of the results. 10x factor dilution each time. In theory…
Using the figures from the above I expect about 1109.2 ng/uL of internal standard. I’ll take the CBD on face value and work from there.
IS triples and CBD drops by a third. Strange…
If this is to be believed, this is resolution down to 0.2ppm if I’m not mistaken
The peak in question:
I’m afraid of what I might find if I look any deeper into the signal…
I guess I have avoided cracking them open long enough…
Transferred them all, not happy with how it went. Only 1 cracked on the line (I assumed restek would include a cracker thing since I could not find an item for it in their catalog… must have forgotten it…
Moved all the standards to vials then proceeded to follow my calibration table
Setting up the sequence table for the run
I’m re-running the IS solution as it split. I’m using a 50ul syringe for this which is less than ideal and I’m not entirely used to injecting with this syringe.\
5 minute method, 3 minutes temp homogenization, its gonna take roughly 2 hours to run just the calibration standards.
i really need to setup the auto sampler…
How did this go?
If it worked well disregard the rest of this message, otherwise I will give you my two cents/tips.
If setting up a calibration curve for the first time in a method I typically run the calibrations as sample type: “sample” and not “calibration”. Usually I am not using the “calibration” sample type until my calibration curve is already established. The reason for doing so is the calibration type allows you to update retention times and response factor which helps account for drift over time. Since this is the establishing curve there is no real reason to do this, and I am not sure how the retention times will be affected since this is a multiple analyte reference standard.
The other possible issue I see is the calibration levels are all set to the same level. When creating a curve from chemstation each level should represent a different concentration.
The good thing is if something was messed up you can override all this and use the data produced to create a new cal table from scratch.
Didn’t get more than 1 run in before realizing there was a problem. For whatever reason column flow/head pressure and split vent flow all decided to not work correctly. Split vent was outputting 0ml, purge was still at 8ml/min. Column was 0.8ml/min.
I’m really tired of tearing this thing apart. Reran the gas lines that were all Frankenstein’d together only to find there’s some issue with the god damn hydrogen generator now. Hopefully its just a leak but idk. I’m so damn tired of this shit. I have no idea if I’m going to have to replace the manifold again or not.
I appreciate the notes on the sequence parameters, i wasn’t really sure what they should be. One question I had was what should be put in for sample amount? i was using the ng/ul measurement as a base but i really don’t know what better to put here.
Yeah, this was a mistake in the fill/insert that I overlooked.
Well if its not one thing, it’s another. Got the GC back together, same with the hydrogen generator. All is well, decided while I was at it I would add the auto sampler tray and injector and hopefully make my life a bit easier…
Well, good thing I had 3 trays because only one works… after getting that sorted it was quite something to see it lift up and move the vial to the injector and do its thing.
Well, that is until it crashes immediately upon injection. Based on this thread, Chemstation shuts down after injection using autosampler - Chromatography Forum The very last post, wouldn’t you know it. I’m running Chemstation B.04.01 SP1.
I’ve got an email out to Agilent to see if they’ll give up the A.09.01. If anyone else has access to that (or know someone) I’d be most grateful… @Chaboes @Cassin @SidViscous
No luck from me unfortunately, I don’t have anything related to chemstation (yet).
No worries I found A10. No luck installing it yet but maybe it needs a fresh do install
That peak at 9.956 is almost certainly d9-THC.
The method is dogshit. No resolution at all, if this is crude or distillate I’d run fast.
I was looking for CBC as the only other feasible option but my eyes hurt after staring at this mess
Auto sampler is up and running, the older version A.10.01 seems to have worked (needed a clean windows xp install). This wasn’t too bad as I wanted to move everything to VMs anyway. Now its sitting on VMware on an old mac-mini hooked up to the GC.
Not sure if my expectations are correct in what differences I should see in using the auto injector. First couple samples i tried to run didn’t seem to have any peaks at all. I watched the injection happen, seemed like it went like it was supposed to. I removed the syringe to check its function and it seemed to be good. Running a blank run now followed by a injection of IS solution.
Thankfully I don’t have to baby it at the moment…
Still to do, get vial inserts for my smaller calibration volumes so that can also be used in the auto sampler process
Going to see how the next injection works with these parameters
Been playing around with this for a few hours, replaced the syringe needle, verified it was able to draw manually, it looks like sample is being taken and dumped in the waste vial; everything relating to taking up the sample seems to work fine. Upon injection no peaks. If i manually inject, I get a peak, auto sampler? Nothing. I think next I’ll replace septa and liner, see if maybe something is clogging it?
Replaced the inlet line, o-ring and septa. Double checked flow settings, they are currently: 2.5ml/min column flow, 55ml/min apx split vent flow. 8ml/min purge vent flow. Makeup to 50ml apx with nitrogen, up to 90ml/min with hydrogen, up to 540ml/min with air. (20:1 split vent now, down from 50:1)
Setup the auto injector and ran the sequence with just solvent to see if anything would happen and I did get some peaks… after 2 minutes…
But when I re-run the samples, nothing. Strange oddities: solvent peaks showing up nearly 3 minutes into the run - usually showing up around 0.3 minutes; inconsistent peaks on samples.
Not sure whats next troubleshooting wise.
On the plus side, replacing the frakenstein tubing with Agilent 2 part furrels eliminated all the leaks i was having. My nitrogen storage tank stays pressurized at some 65 psi (up from barely holding 30psi), and my hydrogen generator currently shows a flow rate of 110, which, seems pretty close to accurate. (2.5ml column flow + 40ml/min makeup flow + 55ml/min split vent flow + 8ml/min purge vent flow = 105.5, give or take; down from 260 - though with a higher split vent flow rate).
Well, one benefit to the auto sampler is not having to baby it. I’m a fan.
Took the gold seal out, looks pretty ugly
Note to self, neutralize before injection…
Swapped the gold seal and running some simple injections to see what happens. I am getting peaks, retention time is out of whack. I’m going to do a bake tonight and see if that fixes any issues; if not, i think next step is to cut off a good portion of the injector side of the column
