How can we increase aluminum response to determine 1-10 ppb level by ICP-MS?

What is your integration time in your method? Aluminum is a fairly small ion so helium will greatly impact the sensitivity. You can either use a longer integration time, lower the He flow (try 4.3-4.5) if other elements don't need so much Helium flow to achieve low background, or a combo of both. 

Hi Cheltajc, 

Thank you for offering us these helpful suggestions. 

1. We used 1.5 second integration time for the first run, which had low responses. 

2. yesterday we run again by increasing the integration time to 2.5 seconds and lowing He flow rate to 4.4 . The responses are much better. 

3. today we will run low matrix mode to see if we can increase more . 

4. do you think we can increase integration time more than 2.5 seconds? because I don't know if the Counts will be stable enough if i increase integration time too much. 

Hi, 

why not running in NoGas ?

Aluminum is not an element with a lot of possible interferences 

Best Regards

Chris

I normally only need 1-2 sec for Aluminum when I run in He mode. However I’m also not typically calibrating as low as you are for Al.  But yes depending on your plasma conditions (low matrix, Gen purpose, HMI etc) it will affect your sensitivity. I evaluate the RSD when deciding whether to increase my integration time, if they are very poor and  I’m trying to achieve low DLs I may have to increase the time. But you will reach a point where you are just pushing how low you can actually get. 

Regarding the plasma, you want a robust enough plasma setting to handle your sample matrix, but not too much to lose sensitivity. If your samples contain > 0.1% TDS switching to low matrix will make your plasma less robust even if you have higher counts. 

Hi  Cschulz, 

Thank you fo your suggestions. we run Zinc and other elements that need He mode to remove interference, so we need to run zinc under He mode. 

Thank you 

best regards, 

Di Xu

Hi Cheltajc, 

I run one more trial with low matrix mode and found RSD% for each level was worse than before ( see below picture), additionally, the spiked sample recoveries were not good. 

So we decided to stay in Gen purpose and increase integration time to see if we can get better resutls. 

Thank you 

Best regards, 

Di Xu