Two methods are provided, the methods differing only in the preliminary treatment of the test substance and the standard. Generally, Method I is used for inorganic materials, while Method II is used for organic materials. 

Aparatus: The apparatus (see illustration) consists of an arsine generator (a) fitted in a scrubber unit (c) and an absorber tube (e) with standard-taper or ground glass ball-and-socket joints (b and d) between the units. However, any other suitable apparatus, embodying the principle of the assembly described and illustrated, may be used. 
 

Standard Arsenic Solution: Transfer 10.0 mL of Arsenic Trioxide Stock Solution to a 1000-mL volumetric flask, add 10 mL of 2 N sulfuric acid, then add recently boiled and cooled water to volume, and mix. Each mL of Standard Arsenic Standard Solution contains the equivalent of 1 mg of arsenic (As). Keep this solution in an all-glass container, and use within 3 days. 
 
METHOD I 

Standard Preparation- Pipet 3.0 mL of Standard Arsenic Solution  into a generator flask, and dilute with water to 35 mL. 
Test Preparation- Unless otherwise directed in the individual monograph, transfer to the generator flask the quantity, in g, of the test substance calculated by the formula: 

3.0/L, 
  
L is the arseníc limit in ppm, dissolve in water, and dilute with water to 35 mL. 

Procedure- Treat the Standard Preparation and the Test Preparation similarly as follows. Add 20 mL of 7 N sulfuric acid, 2 mL of potassium íodide TS, 0.5 mL of stronger acid stannous chloride TS, and 1 mL of isopropyl alcohol, and mix. Allow to stand at room temperature for 30 minutes. Pack the scrubber tube (c) with two pledgets of cotton that have been soaked in saturated lead acetate solution, freed from excess solution by expression, and dried in vacuum at room temperature, leaving a 2 mm space between the two pledgets. Lubricate the joints (b and d) with a suitable stopcock grease designed for use with organic solvents, and connect the scrubber unit to the absorber tube (e). Transfer 3,0 mL of silver diethyldithiocarbamate TS to the absorber tube. Add 3.0 g of granular zinc (No. 20 mesh) to the mixture in the flask, immediately connect the assembled scrubber unit, place the generator flask (a) in a water bath maintained at a temperature of 25 ± 3º C, and allow the evolution of hydrogen and the color development to proceed for 45 minutes,  swirling the flask gently at 10-minute intervals. Disconnect the absorber tube from the generator and scrubber units, and transfer the absorbing solution to a 1-cm absorption cell. Any red color produced by the Test Preparation does not exceed that produced by the Standard Preparation. If necessary or desirable determine the absorbance at the wavelength of maximum absorbance between 535 and 540 nm with a suitable  spectrophometer or colorímeter, using  silver diethyldithiocarbamate TS as the blank. 

METHOD II 

NOTES 
(1) Caution- Some substances may react with explosive violence when digested with hydrogen peroxide. Exercise safety precautions at all times. 
(2) If halogen-containing compounds are present, use a lower temperature while heating the test specimen with sulfuric acid,  avoid boiling the mixture, and add the hydrogen peroxide with caution, before charring begins, to prevent loss ot trivalent arsenic.
(3) lf the test substance reacts too rapidly and begins charring with 5 mL of sulfuric acid before heating, use instead 10 mL of cooled dilute sulfuric acid (1 in 2), and add a few drops of hydrogen peroxide before heating. 
  
Standard Preparation- Pipet 3.0 mL of Standard Arsenic Solution into a generator flask, add 2 mL of sulfuric acid, mix, and add the total amount of 30 percent hydrogen peroxide used in preparing the Test Preparation. Heat the mixture to strong fuming,  cool, add cautiously 10 mL of water, and again heat to strong fumes. Repeat this procedure with another 10 mL of water to remove any traces of hydrogen peroxide. Cool, and dilute with water to 35 mL. 

Test preparation- Unless otherwise dírected  in the indivídual monograph, transfer to a generator flask the quantity, in g, of the test substabce calculated by the formula: 

3.0/L 

in which L the arsenic limit in ppm.  

Add 5 mL of sulfuric acid and a few glass beads, and digest in a fume hood, preferably on a hot plate and at a temperature not exceeding 120ºC, until charrings begins. (Additional sulfuric acid may be necessary to wet some specimens completely, but the total volume added should not exceed 10 mL.) Cautiously add, dropwise, 30 percent hydrogen peroxide, allowing the reaction to subside and again heating bewteen drops. Add the first few drops very slowly  with sufficient mixing, in order to prevent a rapid reaction. Discontinue heating if foaming becomes excessive. When the reaction has abated, heat cautiously, rotating the flask ocasionally to prevent the specimen from caking on glass exposed to the heating unit. Maintain oxidizing conditions at all times during the digestion by adding small quantities of the hydrogen peroxide solution whenever the mixture turns brown or darkens. Continue the digestion until the organic matter is destroyed, gradually raising the temperature of the hot plate until fumes of sulfur trioxide are copiously evolved, and the solution becomes colorless or retaíns only a light straw color. Cool, add cautiously 10 mL of water, mix, and again evaporate to strong fuming, repeating this procedure to remove any trace of hydrogen peroxide. Cool, add cautiously 10 mL of water, wash the sides of the flask with a few mL of water, and dilute with water to 35 mL. 

Procedure- Proceed as directed for Procedure under Method I. 

Interfering Chemicals- Metals or salts of metals such as chromium, cobalt, copper, mercury, molybdenum, nickel, palladium, and silver may interfere with the evolution of arsíne. Antimony, which forms stibine, produces a positive interference in the color development with silver diethyldithiocarbamate,TS; when the presence of antimony is suspected, the red colors produced in the two silver diethyldithiocarbamate solutions may be compared at the wavelength of maxímum absorbance between 535 and 540 nm, with a suitable colorimeter, since at this wavelength the interference due to stibíne is neglígible.