Since you want to analyse AsIII and AsV, you have to use ion chromatography which will separate AsIII and AsV. Then you can use ICP-MS to quantify the respective species.
Water sample can directly be used; however for other matrices you have to use extractants like strong acid.
In the process, IC-ICP-MS you can separate organic arsenic spices also.
Second Alternative: You can use spectrophotometric method to determine roughly AsIII and AsV (Ref. APHA-AWWA-WPCF, (2006) Standard Methods for the Examination of Water and Wastewater, 17th Ed.)
Be careful with chlorides in your sample, because ICP-MS uses argon gas and the argon chloride signal is identical to arsenic. So if you have chloride, you will need a collision reaction cell - it is better to keep away from chloride when preparing the samples (e.g. don't use HCl if you are doing acid extraction or titrations).
You can use the silver diethyldithiocarbamate method in which arsin is genarated by reaction with NaBH4 in acidic solution. In this method arsenate and arsenite can be identified separatly by genaraing arsine at different pHs.As say Asok, because of the severity of chloride ion interference on arsenic, HCL is not recomanded for use in preparation of any sample to be analyzed by ICP/MS.
For the water samples you could adopt solid phase extraction methodologies for separation of AsIII and AsV species and read by ICP-MS. Also helps because of the issues with sample preservation of arsenic in water samples (as arsenic species redistribution could occur in the time duration between the collection and analysis stages)
It depends of how far you want to go in speciation (the number of As species to be identified), the limits of detection required and if you need or not quantitation and a positive identification.
For As III, As V, MMA, DMA and TMA only, in rather diluted solutions, rather sensitive and sheap methods, that orovide quantiation, are (HPLC) QF-HG-AAS and (HPLC) HG-AFS.
If you are analyzing extracts with higher concentrations (biomass, soils, etc) there is a greater number of methods; the most sensitive and reliable (but expensive) available by now being HPLC-ICP-MS.
Anyway if you need a positice identification you have to complement it ,in a way or another, by MS (LC-ESI-MS)
Arsenic species in plant and sediment need to be extracted. the following may have some help.
Plant: 1-2 g of the test materials was placed in a centrifuge tube and 5 ml of 1:9 methanol/water was added. The tube was placed in an ultrasonic focalized bath for 1h (50°C). The samples were centrifuged for 15 min at 2000 g-force, the extract was filtered through a 0.22 μm membrane filter and subsequently analyzed for the As species.
Sediment: A 10 mL aliquot of the extraction solution (1 M phosphoric acid and 0.5 M ascorbic acid, with or without a standard As species solution) was added to 0.2 g (accurately weighed) of a sample in a closed microwave tube (MARS6, CEM, Matthews, NC, USA). The mixture was then maintained at 80 W during 10 min in the microwave system (MARS6). The sample was cooled to room temperature, diluted by a factor of 20 with 30 mM phosphate buffer, filtered through a 0.22 μm membrane filter, and then analyzed for the As species.
By the way, in plant organic arsenic species must be analyze due to methylation
Article Arsenic Adsorption and its Fractions on Aquifer Sediment: Ef...