Fresenius' Journal of Fresenius J Anal Chem (1990) 338:41-45

@ Springer-Verlag 1990

Conception for the investigation of contaminated munition plants 2. Investigation of former RDX-plants and filling stations*

R. Haas, I. Schreiber, E. v. L6w, and G. Stork

FB Chemie der Phiiipps-Universit/it Marburg, Hans-Meerwein-Strasse, D-3550 Marburg, Federal Republic of Germany

Summary. A report is given of the photometric and gas- chromatographic determination of RDX-explosive mixtu- res. The application of these analytical methods is shown by two examples.

Introduction

The problem of environmental contamination from muni- tions residues of World War II was first evident after the investigations of the former TNT-plants at Hessisch Lichtenau [1], Clausthal-Zellerfeld and Stadtallendorf [2, 3]. At these sites a risk to the ground water and the soil from the residues of former production and processing was recog- nized.

After the inventory of the locations of former arming plants in the area of the "Deutsches Reich" [4] and specially in the federal state of"Niedersachsen" [5] a situation is given that illustrates the problems of more than 100 contaminated areas.

The explosive RDX (known in Germany as Hexogen) was produced in Bobingen (450 tons per month), Hanau- Wolfgang (270 tons per month) and in Rottweil. The maximum capacity of RDX-production in the former "Deutsches Reich" (including the "Eastern Areas") was reached in 1945 with 2800 tons per month. Besides the plants for RDX-production there were filling stations at Hessisch Lichtenau, Clausthal-Zellerfeld and Stadtallendorf [6] where RDX was mixed with TNT, ammonium nitrate, Montan- Wax and other explosives. The mixtures were filled in bombs and shells. RDX-contamination of the soil and the ground water is still conceivable at these places.

This paper describes first the methods of production of RDX and then the photometric and gas-chromatographic determination of RDX. The results of the investigations of water and soil taken from the area of the production of RDX and from the area of a fonzaer filling station are reported and discussed.

Methods of the production of RDX

The process of production and handling of RDX is shown by the example of the manufacture at "Werk Fasan" at

* Dedicated to Prof. Dr. D. Babel on the occasion of his 60th birthday Offprint requests to: G. Stork

Bobingen [7, 8]. The so-called "E-Salt Process" was started with 500 kg of acetic anhydride and 0.4% of boron(III) chloride at a temperature of 6 0 - 6 5 ° C. During a period of 6 h, 180 kg of ammonium nitrate and 63 kg ofparaformalde- hyde was added. The temperature of the reaction was held between 60 and 65 ° C. After the termination of the reaction the solution was cooled to 20 ° C. RDX precipitated and was filtered. For purification the raw RDX was recrystallized from acetone. The purified RDX contained a maximum of 10% of HMX, which has a higher explosive power than RDX. The yield was about 60%. The residue of the used solvents contained acetic acid, which was concentrated and converted back to acetic anhydride. By this process acid solutions remained, which were contaminated with residues of RDX, HMX and other by-products.

Later on the "E-Salt Process" was replaced by the "KA- Salt Process" with a higher yield.

In the "KA-Salt Process" the starting material was 262 kg of hexamethylene tetramine-dinitrate, 298 kg of a mixture of ammonium nitrate and concentrated nitric acid (in the proportion of 1 : 1) and 760 kg of acetic anhydride. To control the heat of reaction (50 ° C), the materials were added in 28 portions. After adding all components the solu- tion was stirred for 20 min and the RDX precipitated. The yield was 75 - 80% of RDX. The remaining acetic acid was reprocessed in the same way as in the "E-Salt Process". The RDX was purified from acetone solution and then formulated with Montan-Wax.

Sampling and preparation of the samples

The samples of softs were filled in brown glass containers with gas-tight caps. For the preparation of the sample 10 g of undried, homogenized soil was extracted with 25 ml acetone in an ultrasonic bath for about 30 rain. The extract was filtered through folded filter paper and then used for the following investigations.

The samples of water were also filled in brown glass containers with gas-tight caps. For the preparation of the sample 500 ml of the water was evaporated to dryness by vacuum water pressure at 60 ° C. The residue was dissolved in 5 ml of acetone over 15 min, filtered and reduced to 100 ~tl.

The tests for recovery were carried out with the gas- chromatographic method. The rate for RDX-contaminated soils lay at 95%. The recovery depended on the matrix of the sample; for very fine granulated soil the recovery could decrease.

The recovery from samples of water was 85%.

42

Diphenw famine ( coTo~less )

@NH< NH@ Diph~glbenzidine

( colourless )

Oxidation,, "Reduction

Oxida}ion~, "Reduction

O~phmqg Ibenzldine ( colourless )

N ~N'-Dipheng ]ene - diphenoch~none - diimine

( blue-violet )

Fig. 1 Scheme of reaction from diphenylamine to N N'- diphenylene-diphenoquinone-diimine

f i b s o ~ b a n c e

B'51 l

B~4" x

6,3" × "3

• 2

B,2,

8,1- •

×

8- 25 58 75 188 125 158 175 288

(,g9/2 mL)

Fig. 2 Calibration curves for the photometric determination of RDX at 310 nm (1), 370 nm (2) and 596 nm (3)

Application of a rapid test for RDX and by-products

The application of a rapid test for RDX and its by-products can be useful for the preselection of sampling points. In a contaminated area, i.e. a former RDX-plant, the samples can be taken by a screen of sampling-points (area-related sampling) or by selecting points around an object (object- related sampling). The. rapid test can be used for obtaining an initial survey O f the contamination profile during the field work.

For the rapid test a few milligrams of the soil are put on a chinaplate and a drop of a solution of 1% diphenylamine in 88% sulfuric acid is added. In the presence of RDX or other nitramines, i.e. HMX or by-products, the blue colour- ed N,N'diphenylene-diphenoquinone-diimine is developed (Fig. 1). The concentration limit for the visual indication of the blue colour lay between 5 mg/kg and 50 mg/kg depending on the matrix of the soil. The presence of nitroaromatic compounds, such as TNT, had no effect.

Photometric determination of nitrate esters and nitramines

For the photometric determination of nitrate esters and nitramines 5 ml of the acetone extract of the soil is evap- orated to dryness. The residue is dissolved in 2 ml of a solution of 1% diphenylamine in 88% sulfuric acid. After a reaction time of 5 rain at a temperature of 50 ° C the solution

is measured photometrically at 596 nm against a reference of pure reagent. Calibration curves are obtained by using solutions with increasing content of RDX (10 pg /ml - 200 p~g/ml) treated as described above. The detection limit was 5 gg/ml of RDX, using 2 cm cuvettes. Nitrate esters, for example nitrocellulose, or nitramines, for example HMX, are also determined. The presence of nitroaromatic compounds, such as TNT, does not disturb the reaction.

The reaction to the blue c01oured complex begins with the separation of NO~--ions from nitrate esters or nitramines in strong acidic solutions (pH 0). The NOr-ions oxidize the diphenylamine through an intermediate step (diphenyl- benzidine ) to N,N'diphenylene-diphenoquinone-diimine (Fig. 1). In the photometric measurement we found three absorption maxima at 310 nm, 370 nm and 596 nm. Because of interference by other compounds of the soil a measure- ment in practice is only possible at 596 nm. The calibration curves for RDX at 310, 370 and 596 nm are given in Fig. 2. The concentrations of RDX lay between 10 pg/ml and 200 gg/ml. The standard deviation of the absorbance was 0.011 at 596nm. The correlation coefficient for the calculated curve was 0.997.

Gas-chromatographic determination of RDX

For the gas-chromatographic determination of RDX we used a Packard Model 428 with a fused silica capillary DB

Intensity

43

5

J 6

7 8

ReILentiontime

Pho£ome~ry log w/(l~y/kff) 5-

Fig. 3 Gas-chromatographic separation of nitroaromatic compounds and RDX. Temperature of capillary 180 ° C isothermal (numbers see Table t)

¢>

o x x x

[ [ I - 1 , 5 -1 -B ,5

3-

2-

× × ) x

I I I I 8 , 5 1 1 , 5 2

Gaschroma~ogx'aphy log w/(mg/kg)

Table 1. Retention times and detection limits of nitroaromatic compounds and RDX. Temperature of capillary 180 ° C isothermal

No. Substance Rt/min Detection limit/(ng/gl)

1.2-Nitrotoluene 4.00 0.10 2.3-Nitrotoluene 4.19 0.10 3.4-Nitrotoluene 4.29 0.10 4.2,6-Dinitrotoluene 6.38 0.03 5.2,4-Dinitrotoluene 7.59 0.03 6.2,4,6-Trinitrotoluene 11.69 0.05 7.4-Amino-2,6-dinitrotoluene 23.26 0.10 8.2~Amino-4,6-dinitrotoluene 27.74 0.10 9. RDX 18.66 0.30

5/30N. The detector was an electron capture detector (ECD). The mobile phase was nitrogen. The temperature of the column was held at 180°C, as R D X decompose at about 185°C. The limit of detection was 0.3 ng/gl. F o r the in- vestigations we used acetone extracts of soil and water.

I 2 ,5

Fig. 4 Results of the photometric and gas- chromatographic investigations of soil samples (logarithmic values), w part of quantities; reference standard: RDX

Because R D X was mixed with TNT, the contaminat ion of soil in the area of former filling stations also included dinitrotoluenes as by-products of the TNT-product ion , and aminodini t rotoluenes as microbial metaboli tes of the T N T [9, 10]. These substances are also found with the gas-chroma- tographic method. In Table 1 the retention times and the determinat ion limits are shown. Figure 3 gives a gas- chromatogram of s tandard substances.

C o m p a r i s o n o f the photometric and g a s - c h r o m a t o g r a p h i c methods

The presented photometr ic method for the detection of R D X includes the determinat ion of nitrate esters and other nitramines. The quant i ty of R D X measured by this method may be higher than by the gas-chromatographic method showing the presence of by-products o f R D X in these samples. The gas-chromatographic method allows a single substance detection and determination. Besides RDX, detec-

44

Table 2. Composition of mixtures of RDX-explosives. Values in per cent of contents

TNT Ammonium RDX nitrate

Aluminium powder

Montan Wax

Miscel- laneous

AmmonitH5 - 50 20 - - 30 Amatol 39 50 35 15 - - - Amatol 41 - 52 10 - 2 36 Trialen 105 70 - 15 15 - -

~g/L

1,5"

B~5.

1 2 3 4 5 6 ? 8

• 2,6 I~NT

[] 2,4 DNT

[] TNT

[] RD×

[] Rrom. Amine

8ample No.

Fig. 5 Contents of aromatic amines, nitroaro- matic compounds and RDX in drinking water wells at Stadtallendorf

Intensitg

4

6

9

8

Aetentiontime

Fig. 6 Gas-chromatographic plot of nitro- aromatic compounds of a sample taken in a former bunker of storage of explosives (numbers see Table t)

tion of other ECD-detectable substances, for example PCB and nitroaromatic compounds is possible.

The results of the analyses of 34 soil samples using the photometric and gas-chromatographic methods are shown in Fig. 4. The origin of these samples was the "Werk Fasan" at Bobingen, a RDX-plant during World War II. Points labeled with "O" mark those samples in which RDX could

not be detected by gas-chromatography, but by-products could be detected by photometry. These by-products of the RDX-product ion (7 samples) contain (in relation to RDX as reference) up to 60 mg/kg. Points labeled with "X" mark samples with gas-chromatographic detection of RDX but no photometric detection of RDX or by-products (9 samples). These samples contain RDX only in low quantities below

the photometric detection limit (2.5 mg/kg). The 18 samples labeled with "." contain RDX which is indicated by both the photometric and gas-chromatographic methods. Eight of these 18 samples seem to contain RDX only, because the photometric and gas-chromatographic values are identical within the tolerance of ca. 30%. The remaining 10 samples give values by the photometric method some orders of magnitude greater than those by gas-chromatography, pointing to by-products and metabolites in these sample.

The results show that the use of a non-specific photomet- ric method together with a specific detection method gives further important information about the composition of complex mixtures from former explosives plants and filling stations, as we also showed by our investigations of former TNT-plants in comparing the results of photometric and gas-chromatographic determinations [11].

Applications

Investigation of soil and water samples from the area of a forrner filling station

During World War II 125000 tons of TNT were produced and filled with other explosives and substances in bombs, shells, mines, launchers and rocket-heads of Vls and V2s in the factory "Barbara I" at Allendorf. The following RDX- mixtures were used (Table 2): Ammonit H5, Amatol 39, Amato141 and Trialen 105. The quantity of processed RDX in 1944 was about 750 tons [3]. Besides the residues of RDX from processing, the residues of the dismantling of 30000 tons of bombs and shells from 1947 to 1948 can be expected [2, 3].

The waste water from the former filling station was collected and transported in a channel, which lay next to the former wells of process water, to the station for neutraliza- tion. Some of the former wells of process water are used as wells for drinking water today.

The water analyses of these wells in January 1988 in- dicated contamination with dinitrotoluenes, TNT, aromatic amines and RDX in ~tg/1 quantities (Fig. 5). These results demonstrate that the ground is contaminated to the depth of the wells (about 150 m).

In a soil sample taken in a former storage bunker for explosives we found 0.12mg/kg of 2,4-dinitrotoluene, 0.08 mg/kg of 2,6-dinitrotoluene, 4.2 mg/kg of TNT, 4.4 mg/ kg of 2-amino-4,6-dinitrotoluene and 8.3 mg/kg of 4-amino- 2,6-dinitrotoluene, and 0.5 mg/kg of RDX (Fig. 6). The dinitrotoluenes were by-products of the TNT and the aminodinitrotoluenes microbial metabolites of the TNT [12]. The identification of RDX suggests storage of RDX-ex- plosives, for example Trialen 105.

45

Investigation of a former RDX-plant

For an estimation of the risk of the former RDX-plant "Werk Fasan" at Bobingen we performed object related sampling [7]. Because of the large number of samples (89) a correlation of the photometric and the gas-chromatographic method was possible.

As areas of risk we localized the fields around the former RDX processing site and the incineration site for RDX- residues. With the photometric method we detected higher quantities of nitrate esters and nitramines compared with RDX as reference, than with the gas-chromatographic method. This fact pointed to contamination with by-prod- ucts besides RDX itself (Fig. 4). We also detected and deter- mined PCB in 15 % of the samples. These samples were also taken in the RDX processing area.

The investigation of 8 flat-wells on the area of the former plant showed contamination of 6 wells. The 2 wells outside the ground water stream through the plant were not contaminated.

These results show the risk of ground water contamina- tion in the areas of former RDX-production and processing.

References

1. K6nig W, Schneider U (1986) Sprengstoff aus Hirschhagen. Vergangenheit und Gegenwart einer Munitionsfabrik. Kassel

2. Haas R, v L6w E (1986) Forum Stfidte-Hygiene 37:33-43 3. PreuB J, Haas R, Koss G (1988) Geographische Rundschau

40:31-38 4. PreuB J, Haas R (1987) Geographische Rundschau 39: 578-

584 5. Interministerielle Arbeitsgruppe Riistungsaltlasten (1988) Be-

standsaufnahme und Handlungskonzept ffir Rfistungsaltlasten in Niedersachsen. Hannover

6. BIOS-Report No 1361 (1945) Visit to german high explosive and filling factories. London

7. CIOS-Report No XXXII-8 (1945) Hexogen-manufacture at Fabrik Bobingen der GmbH zur Verwertung chemischer Er- zeugnisse. London

8. Urbanski T (1964) Chemie und Technologie der Explosivstoffe, Bd 3. Leipzig

9. Neumeier W, Haas R, v L6w E (1989) Forum St/idte-Hygiene 40:22-37

10. Neumeier W, v L6w E, Kaminski L, Haas R, Steinbach K (1990) Forum St/idte-Hygiene (ira Druck)

11. Haas R, Stork G (1989) Fresenius Z Anal Chem 335:839-846 12. Haas R, PreuB J, v L6w E, Stork G (1989) Sprengstoff-

riickstfinde in Boden und Grundwasser auf dem Gebiet der ehemaligen Sprengstoffabrik in Stadtallendorf/Hessen. Dokumentation Expertengespr/ich Rfistungsattlasten, S 139- 172. Hannover.

Received December 6, 1989; revised February 22, 1990