Introduction

It is generally assumed that determination of the content in C, H and N is sufficient to verify the composition of a compound by elemental analysis. But if hetero elements are also included in the formula their determination should be taken in consideration This gives a stronger proof of the validity of the assumed composition and enables the producer to conclude eventual deviations fom the expected composition. Primarily the simultanous analysis of C/H/N/S offers an enormous advantage as all four elements are determined within a single run and the result can be gained from as little subtance as 1.5 to 2.5 mg.

The determination of C/H/N/S is done using an "EA 1108 CHNS-O" elemental analyzer by Carlo Erba Instruments (represented by Fisons Instruments, since 1996 part of ThermoQuest).The apparatus is in use since 1991. Earlier elemental analysis on sulfur was performed offline. Two methods were in use then. On one hand mineralization by oxidation in pure oxygen followed by sulfate determination using titration method was useful for most organic material. Due to the conditions of mineralization (electrical oven or Schöninger flask) the determination of sulfur was impossible or erronous at metal containing compounds. For such compounds the reductive digestion of material according to Zimmermann was offered - quite a beautiful and somewhat unpleasant classical chemical procedure.

Method

The sample material is weighed to a tin capsule. Normally as much as 2 to 3 mg are required. But valid results can be gained from lower sample amounts under certain conditions.After folding and wrapping of the capsule (actually rather looking like a wrapped tin foil) the sample is placed into the autosampler.

The tin capsule including the sample material falls into an oven where a defined volume of pure oxygen is added. The first step of mineralization takes place at about 1.020 °C within 1 to 2 seconds. According to literature combustion of the tin capsule should produce a temperature peak of about 1.400 °C. Formation of carbon monoxide is thus probable according to the Boudoir equilibrium although excess oxygen is present.

To fully oxidize the material the products pass a catalyst zone loaded with tungstenic oxide (WO₃). Reaction products should therefore be CO₂, H₂O, NO_x, and SO_x (nothing is stated about the probablitity to produce SO₃ besides SO₂ under this conditions). Also excess oxygen O₂ passes the oxidation catalyst.

The gasous products flow through a zone packed with elemental copper kept at 860 °C. At the copper's surface excess oxygen is chemically bound to form cupric oxide CuO so no oxygen cann pass this catalyst. Oxydic species of nitrogen NO_x are also reduced as well es eventually formed SO₃. Thus this zone is left by all the analyticalle relevant species CO₂, H₂O, N₂ und SO₂ carrying all the elemental content of the sample for carbon, hydrogen, sulfur and nitrogen.

Highest purity helium (5.0 quality) is used as a carrier gas. Separation is performed by gas chromatography. Detektion and quantification are done at a Thermal Conductivity Detector (TCD).

Blank values are run at empty tin capsules.

For calibration runs with standard sample sulfanile amide, a reference material supplied by the producer of the instrument, are performed.

Working Range

All analyt elements constituing the sample material are detected by this method. It is proven that the recovery of sulfur is close to 100 % even from sulfates. Besides the common organic materials metal carbides and nitrides have been characterized successfully.

We also prefer this method instead of the common C/H/N analysis investigating small amounts of samples or accurate data on low level nitrogen content are required. As the chromatography data are recorded by home made software (S.C.A.D.A © J. Theiner) and data processing is done using manually tunable algorithms the evaluation can be confirmed more precisely compared to black box evaluation.

The elemental analysis results obtained from sample amounts below 1,5 mg are usually less valid then concluded from the method description. Nevertheless we have been able to prove that C/H/N/S analysis gives meaningful results at sample amounts as low as 100 to 300 µg.

The determination limit for carbon, nitrogen and sulfur with sample amounts of 2 to 3 mg was proven to be at about 0,05 w-% (500 ppm) with an uncertainty below 0,02 w-%.

According to the manufacturer the instrument's uncertainty in the medium concentration range is within 0,3 w-%. With carbon contents higher than 90 w-% the tolerance probably needs to be extended to 0,5 w-%. As we could prove the correctness of the elemental analysis result at a phenanthren standard material within the 0,3 w-% limit we suggest, that at such high contents hardly any material is actually as pure (min. 99,5 %!) to theoretically give the expected results.

Limits and Interferences

As it is not possible to enclose liquid and oily substances in the normal tin capsules aluminum pans with a lid are available. This supply is not supposed to be used with this instrument bas was found to perform well. The lid is tightly closed by cold .. using a special press to avoid loss of material by evaporation or leaking. As the air volume enclosed within this pans leads to a drastic increase of the nitrogen blank determination limit is increased to 0,1 to 0,2 w-% with liquid samples.

Highly viscous material ("glassy") can eventually not be handled at all.

It is known that phosphorus contained in a sample can interfere with mineralization of organic material. Literature describes the formation of glassy product ("P₂O₅.x H₂O.y C"). Also sulfur may be bound to this phase.

Significant errors can be expected with P containing samples within the C/H/N/S-analysis sometimes exceeding the 0,3 w-% limit. This phenomenon can mostly be controlled by adding vanadium pentoxide (V₂O₅).

As fluorine is reacting to form HF under the conditions of mineralization which reacts with the silica tubes a source of error is found within the analysis of F containing material. But the deviations hardly exceed the tolerance limit.

Digesting metal containing material can also lead to interferences. By methodical adaptions this source of error can mostly be handled.

Important Advice

We need as much information about your sample as available to be able to provide you with best results according to the state of the art.

Declaration of the expected elemental composition (sum formula) is required for standard elemental analysis on preparative samples. Additionally we need to be informed about remarkable instability and/or sensitivity of the material to bring your samples into the analysis in the expected state (e.g. how to store, dry etc.).

With technical products we require knowledge about the assumed amount of analyte in the sample as well as known additives. If we know about the background of analysis (e.g. verify the limit of 0,2 w-% halogene in waste material) we can guarantee results fitting your purposes best.