BFB 50NG

[chloesabrina]

Okay, this has probably been asked many times. EPA 8260 says you must tune with 50ng of BFB. How does everybody check their tune everyday - using autofind? Also what concentration to you manually inject if you don't purge and standard - also if you have a split flow do you take that into account?

[gmlvt]

We do a direct inject of our BFB solution.
Method 8260B recommends a concentration of 25 ng/ul.

8260B also states that a more dilute BFB solution can be injected if you have a more sensitive mass spec.

We make up our BFB solution at 250 ppm, and inject 1 ul. We are running a split ratio of 1:150.
You do need to take in account your split ratio. By my calculations that delivers 10 ng of BFB on column.

You can also purge and trap your BFB.
I find that I get my BFB to pass tuning criteria more easily than when purge and trapping (maybe due to excess water).

We typically use auto find to evaluate our BFB

[chloesabrina]

Okay - sorry I'm confused how does a 1:150 split of a 1ul injection of a 250ppm standard give you 10ng on column. I think my calculations must be off. We have always done 2ul of a 500ppm standard at a 1:50 split. Shouldn't this give us 20ng on column?

[James_Ball]

Okay - sorry I'm confused how does a 1:150 split of a 1ul injection of a 250ppm standard give you 10ng on column. I think my calculations must be off. We have always done 2ul of a 500ppm standard at a 1:50 split. Shouldn't this give us 20ng on column?

I was thinking the same thing. 250ng/ul / 150 = 1.667ng on column.

We currently use 50ppb BFB in our surrogate so when we purge 5ml of 50ppb and have a 50:1 split we are actually putting 5ng on column which passes easily.

250ng in 5ml sample / 50(split) = 5ng on column.

I use the auto find which is the old CLP method where it takes the center three scans averaged and subtracts one scan prior to the BFB peak. If this fails we then look at individual scans of the BFB peak or an average of the entire peak. The method doesn't state exactly how the spectra must be generated that I have seen.

Since Agilent systems actually scan from highest to lowest mass, you will notice a bias towards lower masses on the front of the peak and bias towards higher masses on the tail of the peak. This makes the average methods more representative of the actual balance of the system.

[gmlvt]

You are correct.
I was thinking oft the split we use in our 8260 method, which is 1:150.

For BFB we use 1:25 split hence 10 ng on column.