Glyphosate using EPA 547

[Bigbear]

I'm looking to do glyphosate via EPA 547. I have a Waters system set up for Carbammates using EPA 531.
547 refers to using 2 reaction coils. The reaction reagent #1 is calcium hydroxide soln ( vs NaOH for carbamates), and reagent #2 are exactly the same for both methods. The carbamate method uses 1 reaction coil.

Any idea why the reference to 2 coils?

Thanks
Bear

[James_Ball]

The second is normally a room temperature coil and is normally just a long piece of tubing between the post column mixing tee for the second reactant and the detector. It just gives the OPA solution a few more seconds to react with the hydrolyzed glyphosate before it reaches the detector. I think ours is a loop of teflon tubing with about 0.3ml of volumn.

The first reactant I believe consists of sodium chloride, potassium phosphate and sodium hydroxide with a small amount of Calcium Hypochlorite added. It is best to substitute Sodium Hypochlorite for the Calcium Hypochlorite because the calcium with for calcium carbonate from any CO2 it encounters and will eventually clog up the system. Most equipment manufacturers make this suggestion and we have had permission from EPA to make the substitution for twenty years now. We just use 50ul per L of standard Chlorox Bleach. Buy the smallest bottle of bleach because you need it to be fresh to work and it will go bad after a couple months.

Also make sure your effluent pH is up near 10 for best reaction success. If it needs to be higher, just spike a little potassium hydroxide or sodium hydroxide into Reactant #2.

[Bigbear]

Thank you James.

[Steve Reimer]

When I ran that analysis (using a Pickering 5100) I kept a spare ambient reactor coil handy for fast swaps. It plugged up often enough to be annoying.

[Bigbear]

What plugged? The second Rx coil and/or the detector's flow cell?

[Steve Reimer]

The ambient reactor coil would clog, usually at the end fitting. The flow screen in the detector cell would eventually plug but that was rare.
At one point the line from the heated reactor clogged but I was able to splice in some new tubing.

[Bigbear]

Thanks
A few more questions.
Guard column Y/N?
For my Carbamate method I use Millipore HATF04700 for reagent filtering, and
Millex SLGVX13TL to filter samples.
Would these filters be acceptable for Glyphosate ?

[Steve Reimer]

Guard column? If you are going to run real samples it's better to trash an expensive guard column rather than a very expensive analytical column.
I never filtered my reagents, bought them from Pickering already filtered.
I don't remember what I filtered the samples with. I always tested new supplies to see if there were any problems.
That analysis seemed to have more problems than most. It's prone to both positive and negative interferences. I'm glad I don't have to run it anymore.

[Bigbear]

We do not see the need to run this method, except for the "feel good" aspect.
Our powers that be get quite a few questions about glyphosate.

[Bigbear]

Another question for folks that ran 547.
Is section 10.3.1 telling me to run a 250 ppb spike? I was thinking of calibrating from 10-100 or so, not beyond 250!

[Steve Reimer]

The drinking water mcl is 600 or 700 ug/L so a 250 spike is well below that.

[James_Ball]

I always calibrated from 5ppb to 100ppb just so I would have a point at the listed detection limit. I always spiked my matrix spikes at 50ppb and never had any problems from the auditors.

If you are running only finished drinking water, you probably can get by without a guard column, we normally don't run one, but we also only get a few dozen samples per year. The big thing to remember is when you finish a run, if you are not running the next day, flush the reagent lines for several hours with DI water, then with DI water/MeOH if you are not going to run for a while to prevent bacterial growth. Every few month it is also good to flush the reagent lines with 5% Nitric acid, this will help remove any salt deposits that might clog the unit, just be sure to completely flush it out before storage or running samples, low pH kills sensitivity.

It was always strange when I first began to run the method, we never saw a positive sample, the detection limit was 5 or 6ppb yet the PT providers always sent samples that were between 500ppb and 1000ppb and the pass fail was usually 200ppb to 1000ppb for a 500ppb PT sample.

Oh and the best column we found to run them was the Hamilton PRP-X400 PEEK column. It doesn't have the limitation of the Pickering column which can not tolerate any organic at all, plus the PEEK construction won't leach any iron due to the acidic mobile phase, where iron is the main thing that will poison the stationary phase.

[Bigbear]

I finally got up the gumption to run this method, and at this point I get no peaks! So far I've run from 5ppb to 100 with nothing.
I use parts of the EPA method and a Waters paper. Column Waters ion exclusion 7.8X 150m (WATO10295) at 1.5 ml/min. Making up the base as James said, and the opa reagent the same as Method 531. the Waters paper makes this reagent with 0.8g OPA and 2 ml mercaptoethanol.
I am using 340nm excitation and 455nm emission settings.

Anything jump out as incorrect?

Thanks
Bear

[Bigbear]

OOps, could it have something to do with old bleach? I had an unopened bottle about a year old that I used. Any old off the shelf bleach work?

[Steve Reimer]

We use bleach, typically Clorox, for our Chemical Warfare Analysis decon and I have noticed that recent bottles that are listed as low splash or some such thing have a surfactant or thickener. That may cause problems for post column reactors.
If I didn't see a peak in my standards the first thing I did was add more bleach.