My purpose is to be sure about the closer data from ICP - OES before analysis to reduce the cost of ICP-OES. Is it possible to measure concentration of heavy metals by volumetric analysis or UV visible spectrophotometer?
Quickest and easiest way to obtain orders of magnitude prior to ICP is to run an XRF analysis (if your analytes fall within the limits of the technique, of course).
Or, you can simply run first a 1ppm solution (or 0.1 or 10 ppm depending on your expected range), then your samples and estimate target concentrations doing the intensity ratio.
If you have a number of specimens you expect some to have similar compositions, you could consider compositing those and analysing the composites to scope the element concentration ranges. However, my experience is that within the one specimen the range of concentrations of the different elements can vary widely.
Dharma, the earlier responses suffice. In addition, there are no other ways to analysing heavy metals than AAS, ICP, or XRF. You cannot determine HM concentrations by UV /VIS spect.
Beg to differ on this : limits of detection are generally worse and it is indeed not always possible for all elements, BUT older complexation methods do exist for lots of metals determination by UV-Vis (iron by o-phenantroline for instance). They are however slower than XRF.
In addition to the good suggestions already made, below is another possible approach :
It appears that you are interested in measuring samples with wide range of unknown concentrations.
Depending on whether you use an ICP-ES in house, or samples being sent out to commercial lab, you may prepare the samples at two (or more) dilutions, for example, your normal dilution (1X), and then by another 10X, etc. First analyse the diluted solutions. That will screen out any concentrations that is going to be too high for 1X, which you don't have to reanalyze. Another important advantage of measuring them at multiple dilutions is that, when you do multi-element measurements, you may choose results from both measurements to select the optimum data. For example, in some samples Ca may be too sensitive for the 1X dilution, but good with 10X, and Cr may be too low for 10X, but may be good and well above LOD with 1X. This approach also makes the measurements instrument friendly without unnecessarily overloading the instrument (unknowingly) with too high concentrations, resulting in deterioration of the torch, contamination of the sample introduction system, etc. Although it may sound like you will have to do more measurements, it pays off at the end with better data and less trouble (even more economic). Measuring at highest possible concentration is not always the best, as many tend to believe!