system lost prime mystry

[meillid]

Hi, Everybody,

my colleage and I are validating one of the hplc method, method detail see below. the method come from vendor and was working no problem. But now the system keep on losing prime. Each of us use a HPLC with the same method setting, each of our system has lost prime many time from yeaterday.(different person, different system, only method same, it looks like method problem, but why it was working good before and for vendor?) Anybody experence this before? Thanks a lot.

Chromatographic system

Column XBridge C18, 2.5-m, 4.6 x 75-mm
Detector UV at 238-nm
Temperature 50℃
Flow Rate 1 mL/min
Injection Volume 10 L
Run Time 50 minutes

Buffer Solution

Dissolve 1.42 g of dibasic sodium phosphate in 1,000 mL of purified water. Adjust the pH to a 9.0 with phosphoric acid. Filter the solution through a 0.22-μm filter.


Mobile phase

gradient of buffer and methanol as follows:

Time (minutes) Buffer solution Methanol Elution
0-7 70 → 55 30 → 45 linear gradient
7-20 55 → 42 45 → 58 linear gradient
20-22 42 → 36 58 → 64 linear gradient
22-30 36 → 33 64 → 67 linear gradient
30-35 33 → 20 67 → 80 linear gradient
35-37 20 80 isocratic
37-38 20 → 70 80 → 30 linear gradient

[Bruce Hamilton]

Quick guess.

Your buffer is not soluble in 80% methanol.

Please keep having fun,

Bruce Hamilton

[HW Mueller]

Some more quickies:
According to my data phosphate at pH 9 is not really a buffer.
If it is made as described the composition can vary widely, so that it is possible that one prep does not precipitate, another does as Bruce guessed.
You didn´t degas sufficiently.
You are using "pure water" which is teeming with life as others have described in several contributions.

[meillid]

Thanks HW and Bruce,

1. I know that the buffer is not a buffer instead a salt solution. some salt solution with fixed PH vlaue also works good. During the 5 years working for industry, I saw so many method like these, which do not comply with the theory that we studied in school and advocated by HPLC/column vendor. But their daily performance are very good.

2. buffer is not soluble in 80% methanol
This looks like true. The problem is, however, how to explain that this method have been validated by vendor company and our sister company, we have also veryifed that the mehtod. Despite so many different system, different people, different location and time, it never made the system lost prime before.


more suggestion is very very welcome.

[Uwe Neue]

Dibasic sodium phosphate comes in several versions with different hydration. The consequence is that one will contain more salt than another. If you picked a version with a lower hydration than the vendor, you have more salt dissolved, and the stuff is more likely to precipitate as you add organic, which could make the check valves stick easier.

[Bruce Hamilton]

[ Edit - Uwe beat me to the hydration issue and posted whilst I was writing this ]

Some days are custard days ( everything turns to custard )..

Assuming that you have eliminated any issues arising from one site using a different hydrated form of Na2HPO4 ( choose from 0, 2, 7, or 12 water molecules ), I'd say the method was not robust, unless you can show that your Na2HPO4, H3PO4, water or Methanol do not comply.

My guess would be that your pump's mobile phase mixing system is allowing precipitates to form, then the methanol boils in the pump and loses prime. That would suggest that the method is susceptible to minor instrument or ambient temperature changes, assuming operators didn't have a brain fade.

Please keep having fun,

Bruce Hamilton.

[HW Mueller]

Again, the way you prepare this phosphate solution the concentration will be varying according to the skill, etc., of the operator. Also, some phosphate/organic mixtures take a long time to precipitate and may strongly depend on temperature. One can also imagine that some frits provide a seed for precipitation, others do not.
I am curius, why do you think that a solubility problem is causing the prime loss?

Are there people out there who are responsible for a demanding method like HPLC, who do not know how to handle different hydration of substances they use??

[lmh]

My heart sinks every time I see a method that starts "dissolve 1.42 g...". It's much better to specify buffers in molarity, always assuming that the lab staff are able to carry out simple solution calculations.

[Consumer Products Guy]

My heart sinks every time I see a method that starts "dissolve 1.42 g...". It's much better to specify buffers in molarity, always assuming that the lab staff are able to carry out simple solution calculations.

I disagree. MY heart sinks every time I see a method that starts "make 0.005 M solution..".

Most QC lab analysts at our contract manufacturing have no idea what molarity is. We find that if we want anything to be made right, we specify EVERYTHING. And that includes stuff like "weigh 1.41 - 1.43 grams..."
We have enough issues with the QC operators as it is, not knowing how to modify integration parameters or calibration tables; requiring them to make calculations to prepare a solution would be a nightmare.

[Peter Apps]

While it depends to some extent on who has to follow the instructions, I am with Consumer Products Guy on this one. When I was working in a contract lab the methods that gave the least trouble to the routine analysts were those that were explicitle and step by step in terms of what operations they had to carry out. And if a method has to be done several times (which surely is always the case) why not do the molarity etc calculations once and put masses inthe method rather than putting molarity in the method and having the calculation carried out very time the method was used ?

Peter

[HW Mueller]

And then someone puts a NaH2PO4.2H2O bottle where the NaH2PO4 bottle used to stand, comes the robot and weighs . . . .

An analytical lab filled with people who don´t know what molarity is?

[Consumer Products Guy]

An analytical lab filled with people who don´t know what molarity is?

H.W.: you're in Europe; this is not uncommon in U.S., at our own QC labs and especially at contract manufacturers' QC labs. At the latter, not uncommon to find workers who are not English-speaking, or have poor English skills, being born elsewhere, which contributes to the communications issues. Needless to say, few of those have college Chemistry or degrees. I'm in R&D, department of six, we have 2 Summa cum Laude and 1 Magna cum Laude employees, two of which are Phi Beta Kappa (no PhDs). But we need to write our procedures so they are as bulletproof as possible. And yes, many of those facilities manufacture pharmaceutical and EPA regulated products, and their employee qualifications are out of our control.

[Peter Apps]

Hello Hans

They all knew what molarity was (I suspect that they were better informed on this subject than certain others), it was just more efficient and effective to do the calculation once and specify in grams.

If there was a possibility of a mix up such as the example you give I would list it under "precautions" along the lines of "Hyperbolium sulphate exists is different forms, ensure that the reagent used is as specified". Unfortunately the idea is not to educate the analysts but to get the analysis done - and believe me you will be preaching to the converted if you throw up your hands in horror.

Peter

[meillid]

Hi, Thanks for everyone's reply,

I agree the suggestion to make the method clear for the QC operator. for my case, however, it is in R&D, I believe very body is using the unhydrous form, so no problem about this confusion.

Go back to my original question: Personally I also think the method is somehow have problem. however, since this method has been used such a long time by different person without problem, any of you have any suggestion beside the method? THANKS A LOT.

[Bruce Hamilton]

Go back to my original question: Personally I also think the method is somehow have problem. however, since this method has been used such a long time by different person without problem, any of you have any suggestion beside the method? THANKS A LOT.

I think several people have pointed out potential issues, but we can not ascertain which are relevant to your situation.

1. Reagent ( including solvents and water ) grades should be all identical ( not equivalent, but identical - meaning same product code from same producer).

2. The solution preparations should be identical ( phosphate buffers can take time to dissolve ( eg using ultrasonication ) and precipitate.

It's even possible that the mobile phases may have been pre-mixed, ( A=70/30, B20:80 ) before filtering - which could have an effect on composition.

The mobile phase filters could have been different - 0.22um would be fairly slow flow, and the buffer solution could have concentrated. Degassing of prepared solutions could have also changed composition, eg sparging.

The mobile phase solutions were allowed to standing for hours to come to laboratory temperature before use etc etc.

3. The ambient conditions are different eg the laboratory temperatures are a few degrees different.

4. The instruments are different. One uses a different brand or model to the other - One model has different degasser, or the degasser is set differently.

The pump mixer design is different, the pump solvent bottle filters ar partially blocked of of different porosity. The pump post filter is partially blocked or of different design or porosity.

5. The instruments are in need of a service. The pump check valves and seals are worn.

However, only you can investigate and ascertain which are relevant. I actually thought we had provided quite a few paths for you to explore.

Although only 10 mM, the phosphate buffer would be my starting point, unless I knew that the systems were different.

Bruce Hamilton