External Contamination of Landmines
by Organic Nitro-Compounds
DANIEL C. LEGGETT, THOMAS F. JENKINS, AUSTIN W. HOGAN,
THOMAS RANNEY, AND PAUL H. MIYARES
INTRODUCTION
formance liquid chromatography with ultraviolet
Chemical sensing of landmines requires a sufficient
detection (HPLC-UV). These data are reported here for
emission of chemical species into the local environ-
the first time. Statistical analysis indicates that the data
ment to form a detectable signature. Mines are exter-
are spatially very heterogeneous and that this is not
nally contaminated in the process of manufacture, pri-
attributable to systematic differences in laboratory or
marily during filling operations. U.S. mines are cus-
method. Most of the mines were contaminated with 2,4-
tomarily placed in wooden boxes for storage in bun-
dinitrotoluene (2,4-DNT), TNT, and RDX, and less fre-
kers. During storage, vapor transport could effectively
quently, Tetryl, octahydro-1,3,5,7-tetranitro-1,3,5,7-
contaminate all mines in proximity (Bender et al. 1992).
tetrazocine (HMX), aminodinitrotoluenes, and 1,3,5-
A study of 51 bunker-stored landmines of foreign and
trinitrobenzene (TNB). Unfortunately, 1,3-dinitroben-
domestic origin (Hogan et al. 1992) supports this con-
zene was added to all the vials as an internal standard
jecture but doesn't rule out permeation through mine
before we realized that it was a prevalent component in
casings as a source of surface contamination. Bender
some landmine signatures.
et al. (1992) also demonstrated that painted metal and
Concentrations of the major analytes ranged over
other materials contaminated with TNT vapor outgas
many orders of magnitude in the 1992 work (Appen-
at a significant rate for many days.
dix A). Some mines were very clean--for example, the
metallic U.S. M16 antipersonnel mine, in which TNT
PRIOR WORK
was detected on only one of eight mines, and in only
one of sixteen swabs, as the paired swab was blank as
well. Since these M16s were used previously in mine
In 1992 a joint study by CRREL and Thermedetec,
lane trials, they may have been cleaned prior to burial,
Inc., characterized the surface contamination of 18 mine
or burial may have effectively removed surface con-
types of seven national origins, variously encased in
tamination. Unlike plastic mines, they would not
metal, plastic, or fiberglass. Surface samples were taken
by swabbing a 5-cm2 area with a cotton swab saturated
recontaminate by permeation. All the plastic mines
examined had detectable surface contamination. The
with acetone. Each mine was swabbed in two different
TMA3 was by far the most contaminated. The Humani-
areas, and the individual swabs were analyzed inde-
tarian Demining Database (1999) describes this mine
pendently using different methods. One set was ana-
as a block of cast TNT with a thin fiberglass coating.
lyzed for 2,4,6-trinitrotoluene (TNT) and 1,2,5-triaz-
RDX/HMX accounted for more than one-third of the
ine (RDX) by gas chromatography with a thermal
signature, suggesting that what was being sampled by
energy analyzer (GC-TEA) at Thermedetec, Inc.,
Woburn, Massachusetts. These results were reported
the swab was more like a bulk explosive phase than a
deposited vapor; volatility considerations would seem
in the memo by Hogan et al. (1992).
to rule out vapor phase transport of HMX. This quan-
The other set was analyzed at CRREL by high per-
Previous Page
