This is a tricky question as it will be difficult to avoid mentioning brand names and producers.

I presume you are looking into a quadrupole instrument. The big players such as Thermo, Agilent, PE etc. all produce great instruments. It is certainly a matter how good your support is in your country.

You need to make sure that you have the possibility of a collision/reaction cell with He or He/H2 to get rid of interferences. Also low oxide formation rates (CeO+/Ce+) are essential for good REE data (1% and lower is needed, but this can also be reached with different sample introduction systems, such as APEX from ESI etc.

Also for whole rock analysis you would want to have a detection system that can span at least 8 orders of magnitude in intensity as you would like to analyse major and trace elements concentrations from the same solution.

In addition you would like to have a robust system that can handle high total dissolved solids content and probably you would need a HF resistant sample introduction system.

There are many more things to say. It would be better for you to ask specific questions.

I am sorry but I think that collision cell is the only trouble maker a you will never use it for 99% of your applications. What´s the reason? Chloride interferences from HCl? Add the appropriate amount of HCl to your calibration standards. ArO? OK, buy the desolvating nebulizer, APEX or ARIDUS. You will obtaine much better analytical sensitivity and not measurable oxide level. So REE without any problem, Fe on 56 without any problem. What else? Total digestion with HNO3+HF+HClO4 or HCl, eliminate SiF62- with evaporation to almost dryness and then dilute to a maximum of 2% sample matrix. Forget collision cell, do it easy. Select right isotopes and internal standard, i.e., Ge, Rh, Re...

Good luck,

Vladislav

I do agree with Thomas and Susann. I have used the Quadrupole ICP-MS for most element and isotopes. ICPMS Model X-SeriesII, Thermo-Fisher Scientific, Bremen, Germany in ‘collision cell mode’ (7% hydrogen in helium) to reduce polyatomic interferences. Internal standards included Sc (50 µg L-1), Rh (10 µg L-1) and Ir (5 µg L-1) in 2 % TAG HNO3. External multi-element calibration standards (Claritas-PPT grade CLMS-2, Certiprep) were used usually to measure multi elements in the soil digest solution in a 2% HNO3 matrix. Also the collision cell is perfect in case of the interferences with poly atoms. The standard mode is highly precise in case of measuring Pb isotopes. The problem with this kind of instrument is that it is not a smart device. In the other word you need to adjust some parameters according to your experience in order to pass the performance check to get a high count per second and high sensitivity. There are some exemption if you need to measure Fe as you need to select the isotope that has no interference with other isotopes. Also the major cations such as Mg, Ca, Na and K, the standard solution need to be in ppm and let the device tured into analog mode. Also, the multi element stable isotope, you need to take care of the quadrupole settle time between the lowest isotope ad the highest one. In general, in your case as you need it only in the giolgical samples for multi element analysis, I suggest that one or the recent version of QICPMS.

I would simply add the comment that we have been running several Agilent 7000-series quadrupole ICPMS for several years, and have found them very well suited to both solution and laser-ablation work. The detection limits are very good, the software is straightforward and easy to use. We are fortunate in having the service organization nearby, although we have not had many reasons to call on them. As Thomas says, it is important to evaluate the quality and reliability of the support available in your area.

Thanks for all your helpful comments. Two more specific questions. (1) Does collision mode show better result than reaction mode for multi-elements analysis? Or do they show similar performance? (2) Any problem can arise with 90-degree deflector lens type?

Collision or Collision/Reaction cell has been an issue in the discussion. I do not agree with Vladislav: The He mode with our Agilent 7500 is essential to obtain good data for Sc, Cr, Cu, As, and some other elements as we use Na2O2 sintering and HCl the interferences are severe. For most other elements including REE, Li, Be etc. the collision cell is not essential. In short: I would not buy an ICP-QMS without gas options!

I did some tests with the PE system (reaction cell) and was convinced by the Thermo patent with He/H2 with is also used in the Agilent system but with He only for multielement analysis.

I have no clue about the deflecto lens question.

Collision mode is better for multi element analysis to avoid any interference.