Propellants, Explosives, Pyrotechnics 27, 119 Ä… 124 (2002) 119
Analysis of ADN, Its Precursor and Possible By-Products Using Ion
Chromatography
Gudrun Bunte*, Heinz Neumann, J¸rgen Antes, and Horst H. Krause
Fraunhofer-Institut f¸r Chemische Technologie ICT, Joseph-von-Fraunhoferstr. 7, D-76327 Pfinztal (Germany)
Dedicated to Professor Dr. Hiltmar Schubert on the Occasion of his 75th Birthday
Summary nated oxidizer ammonium perchlorate (AP). ADN has Ä… in
comparison to AN Ä… the advantage of a higher energy input
In the last years ammonium dinitramide (ADN) appeared to be
combined with a reduced pressure in application. Moreover,
a promising new oxidator and a possible substitute for ammonium
ADN shows no phase transitions like AN, meaning no
nitrate (AN) and especially for the chlorinated oxidizer ammo-
critical volume/density changes under temperature stresses.
nium perchlorate. Among other main advantages of ADN are to
Compared to AP the use of ADN-based propellants is
be mentioned the higher energy input combined with a reduced
pressure in application. Furthermore, ADN shows no phase expected to have a reduced plume signature producing no
transitions like AN. For evaluating the purity of the synthesized
environmentally hazardous halogenated gases. At
and/or treated or aged pure or formulated ADN, the estimated
Fraunhofer ICT several aspects concerning the potential
ammonium nitrate content was taken into account. AN is known
use of ADN in new formulations are considered, especially
to be as well a by-product of the ADN synthesis as a possible
decomposition product of ADN. Thermally treated ADN decom- the following research topics:
poses mainly to N2O, H2O, NO2 and AN which further reacts to
*
Synthesis of ADN in a technical scale with a high purity
N2O and NH3. Determining the nitrate contents assuming the rest
*
being intact ADN must not lead to correct values especially in Recrystallization/prilling of spherical ADN(7)
cases where ADN was treated/handled at higher temperatures in
*
Stabilization of neat ADN
open systems. Concerning the technical scale synthesis of ADN,
*
Compatibility tests for ADN and new binders and/or
the precursor ammonium nitrourethane (ANU) must be elimi-
energetic plasticizers
nated in a quick but sufficient way needing a suitable analysis
*
Aging/shelf life prediction of ADN-based formulations
method for detecting nitrourethane besides nitrate and ADN. The
objective of this work was to develop a suitable ion chromato-
While ADN could, in principle, be synthesized using
graphic method for the direct analysis of the anions concerned.
ammonia and dinitrogen pentoxide at very low temper-
Different ion exchanger phases were tested with organic and/or
inorganic eluants. The ionic strength and flow rate of the eluant
atures, the more often used synthesis route involves the
was improved to get an acceptable resolution for nitrite and
urethane reaction path yielding ammonium nitrourethane
nitrate combined with a short run time for the whole analysis.
(ANU) as a precursor. In a second step, ANU is further
Detection was realized by electrical conductivity or UV absorp-
nitrated with suitable nitration agents like BF4NO2 or N2O5.
tion whereby the measurement wavelengths were optimized in
order to get a small signal-to-noise ratio and simultaneously a
For both synthesis routes AN is produced as a by-product.
suitable sensitivity especially for NO3 and nitrourethane. Under
AN as well as not converted ANU have to be eliminated in a
improved conditions (Ion Pac 11, 1 ml/min NaOH, 300 mmol),
sufficient manner in order to give a pure product, ADN. For
limits of detection (LOD) of 0.05 to 0.01 ppm were realized for
the synthesis of ADN in a technical scale, this means that the
NO3 and NO2 , respectively, measured at 214 nm. Using 220 nm
recovery should be as quick as possible but also result in a
as detection wavelength resulted in a LOD of about 0.3 ppm for
nitrate. Using a wavelength between 210 and 220 nm results in a
high enough purity of ADN. In order to qualify the recovery
LOD for ANU of about 1 ppm. The linearity range for the
procedure and to quantify the content of ANU and AN a
analysis of DN (285 nm) was found to be very broad (up to
suitable analysis method is needed.
700 ppm). All anions can be analyzed in one run taking maximally
For judging the purity of mechanically or thermally
30 minutes.
treated pure ADN or formulations containing ADN, mainly
the estimated ammonium nitrate content was taken into
account known to be a possible decomposition product of
1 Introduction
ADN(6). In former thermal decomposition studies(4Ä…5) it was
observed that the main decomposition products of ADN are
In the last years ammonium dinitramide (ADN) appeared N2O, NO2, H2O and AN which further reacts to N2O and
to be a promising new oxidator(1Ä…3) and a possible substitute NH3 at higher temperatures. Searching for the best stabilizer
for ammonium nitrate (AN) and especially for the chlori- for neat ADN as well as carrying out compatibility tests for
or aging tests of new formulations arises the problem that
* Corresponding author; e-mail: [e-mail] determining the nitrate content and the mass loss of the
Ä… WILEY-VCH Verlag GmbH, 69469 Weinheim, Germany, 2002 0721-3113/02/2701/0119 $ 17.50+.50/0
120 G. Bunte, H. Neumann, J. Antes and H. H. Krause Propellants, Explosives, Pyrotechnics 27, 119 Ä… 124 (2002)
formulation is not sufficient. For explosive formulations detect the not converted nitrourethane and nitrate, the by-
containing several ingredients also a direct measurement of product and/or the main destruction product of ADN.
the intact ADN content is needed.
4.1 Analysis of Dinitramide with ANION-R-Column and
2 Objectives Electrical Conductivity Detector
While the analysis of inorganic nitrate or nitrite by ion With a standard Wescan anion exchanger material
chromatography is a routine method at Fraunhofer ICT (ANION-R), tests were made to detect dinitramide anions
since a long time, the possibility of a direct investigation of besides nitrate and nitrite on the same column using an
the content of ammonium dinitramide (ADN) and more- electrical conductivity detector. Under standard elution
over of ammonium nitrourethane (ANU) was lacking. conditions, normally used for nitrate and nitrite, no signal
Therefore, the main objective of this work was to develop a was detected for DN in a realistic analysis time. Different
suitable analytical method for the direct characterization of ionic strengths of the eluant (p-hydroxy-benzoic acid) were
dinitramide anions parallel to nitrate and nitrourethane. also tested but nothing led to a noticeable peak for
dinitramide. A possible reason for these results could be
the good delocalization of the negative charge of DN ,
3 Experimental yielding only to a small difference in the conductivity of the
The ion chromatographic instrumentation (Alltech/
GAT) used at Fraunhofer ICT is a single column ion
chromatography (SCIC)-technique which uses no suppres-
sor in combination with the electrical conductivity detector.
Here, the ground level of the conductivity of the used eluant
is compensated electronically. Alternative systems using the
suppressor technique reduce the ground conductivity of the
eluant chemically yielding a greater usable detection range
which is needed, especially when ppb- or ppt-levels are to be
measured. For standard anions, cations and organic acids,
the used SCIC-system was tested to be a well suited
technique realizing measurements in the ppm- or ppb-level.
The used IC-system is equipped with two HPLC-pumps, a
column oven and a ten-port injection valve coupled with two
parallel injection loops going to two parallel IC-columns.
Alternatively, besides the electrical conductivity detectors
(one for each IC-channel), an UV detector (Hewlett
Packard) was installed in this study instead, or sometimes
in series, with the other ones. Registration and evaluation of
the detector signals are performed with a FISONS chroma-
tography software (Minichrom). High purities of the used
standards and chemicals for making the eluants were used,
whereby the ANU was isolated and characterized by IC,
CHN, DSC and IR. The following different IC separation
phases were used in this study:
*
ANION-R: Wescan anion exchange material, styrene-
divinyl benzene-copolymer with trimethylammonium
exchanger functions, 10 mm spherical particles
*
Ion Pac 11: Dionex anion exchange material, pellicular
latex particles, divinylbenzene/ethyl vinyl benzene co-
polymer modified with alkalonol quaternary ammonium.
4 Results and Discussion
In the first part of the study, the evaluation of an ion
chromatographic method for the direct analysis of dinitra-
mide (DN ), nitrate and nitrite is described. The second part
Figure 1. UV spectra of ADN, potassium nitrate and sodium
concerns the further developments of the method in order to nitrite.
Propellants, Explosives, Pyrotechnics 27, 119 Ä… 124 (2002) Analysis of ADN 121
Figure 2. Influence of the ionic strength and flow of the eluant (NaOH) on the retention time of the DN anion (Ion Pac 11, UV
detection).
pure and of the dinitramide loaded eluant, and moreover to were tested, no signal was recorded in a realistic analysis
a not detectable signal because of the high ground level of time. The main reason for these results is Ä… once again Ä… not
the eluant itself. This would be noticeable, especially for clear. The separation phase could generally not be suitable
high molarities of the used eluant p-hydroxy-benzoic acid for the analysis of dinitramide as well as the eluant (pHBA)
which should lead to a shorter retention time of the analyzed which has a high UV ground absorption level.
anions. Otherwise, the type and/or capacity of the used
anion exchange material (ANION-R) could be denoted as
not suitable for the retention of DN . 4.3 Analysis of Dinitramide with Ion Pac 11-Column and
UV Detector
4.2 Analysis of Dinitramide with ANION-R-Column and A special IC anion exchange-column from DIONEX
UV Detector (IonPac 11 HS) was coupled with the UV detector. This
separation phase could be used with sodium hydroxide as
For the reasons given in section 4.1, a different detection eluant also at very high molarities yielding normally
principle was tested in order to get a reasonable signal for preferable shorter retention times of large ions like dini-
the dinitramide anion. The UV spectrum of ADN shows two tramide. With this combination a peak was detected for DN
strong absorption bands with maxima at 214 and 285 nm for the first time. Optimization of the elution and detection
(Figure 1). Nitrate and nitrite show UV absorption maxima conditions were as follows.
at 202 and 211 nm but with much lower sensitivity than
dinitramide. Tests for the above mentioned ANION-R-
column were made to separate and detect dinitramide 4.3.1 Ionic Strength and Flow of the Eluant
anions using p-hydroxy-benzoic acid as eluant. As detection
wavelength 285 respectively 214 nm, was used. Although The retention time of dinitramide strongly depends on the
different molarities meaning ionic strengths of the eluant molarity and secondly on the flow rate of the eluant which is
122 G. Bunte, H. Neumann, J. Antes and H. H. Krause Propellants, Explosives, Pyrotechnics 27, 119 Ä… 124 (2002)
Figure 3. Different calibration standards (Ion Pac 11, 0.3 m NaOH, 1 ml/min, UV: NO2 and NO3 at 214 nm and DN at 285 nm).
demonstrated in Figure 2. At a flow of 0.8 ml/min DN length of 285 nm leading only to a slight loss in sensitivity.
elutes after 54 min if using 0.1 molar NaOH as eluant. So, the UV detector was adjusted at 285 nm for the detection
Doubling the NaOH concentration shifted the peak max- of DN and at 214 nm for the analysis of NO3 . Using the
imum to 28.3 min. Nearly the same retention time was high concentrated NaOH eluant (0.3 mol) the signal-to-
observed with a slightly higher flow rate of 1.0 ml/min noise ratio for nitrite as well as for nitrate was reduced, but
combined with a lower eluant molarity (0.15 m NaOH). especially the retention time of the nitrite peak was shifted
Further increase of the ionic strength of the eluant only to the edge of the injection peak. A measurement or
showed a relative small decrease of the retention time. With detection of nitrite under these conditions would still be
a molarity of 0.18 the DN elutes at about 24 min, whereas a possible but the quantitative data will be more spread.
0.3 m NaOH shifted the peak to 19.6 min. Nevertheless, a suitable separation of nitrite and the more
interesting nitrate is well achieved.
Concerning both the optimum retention time for dini-
4.3.2 Adjustment of the UV Measurement Wavelength tramide and a reasonable separation also of nitrate and
possibly of nitrite, a concentration of 300 mmol NaOH was
In order to get the optimum signal-to-noise ratios for tested to give the best results. While the resolution of nitrate
nitrate and dinitramide, the detection wavelengths were and nitrite and therefore also the possible limit of detection
varied. While a wavelength of 202 or 204 nm for the analysis was slightly better when a 0.15 m NaOH eluant was taken,
of nitrite and nitrate was tested to have the highest the higher concentration of 300 mmol NaOH yielded much
sensitivity, here, a relatively high noise of the UV base line shorter run times, especially for high concentrated ADN
signal was recorded. At 214 nm the signal-to-noise ratio was samples for which a long peak tailing was observed. Figure 3
much better for nitrate yielding also a very high sensitivity shows ion chromatograms of different standard concentra-
for the dinitramide anion. For the latter a much better tions. It could be seen that the peak maximum is slightly
signal-to-noise ratio was observed for a measuring wave- shifted between 17 and 24 min with increasing ADN
Propellants, Explosives, Pyrotechnics 27, 119 Ä… 124 (2002) Analysis of ADN 123
concentrations yielding to an entire run time of 30 min in all Figure 5 shows the ion chromatograms for a solution
cases. containing NO3 and ANU each at the same level of about
25 ppm which were measured using different UV detection
wavelengths. Differences in the peak areas/peak heights are
4.4 Analysis of Synthesized ADN During the Cleaning due to the absorptivities of the substances at the distinct
Procedure wavelength as well as to changes of the background
absorption of the eluant (0.3 m NaOH). In order to get the
The purity of synthesized ADN naturally depends on the highest sensitivity for ANU, combined with a realistic
cleaning procedure (conditions like type of washing solvent, detection possiblity for nitrate, the best UV wavelength was
number of washing steps and others). By optimizing these tested to be 220 nm.
conditions in view of the upgrade of the reaction to a
technical scale, some of the analyzed samples showed a
small further peak after the nitrate signal. Because solvents 4.5 Limits of Detection and Calibration Curves
and non-ionic organic substances should not be retained by
the IC-column, first tests to clear the identity of the peak Under improved separation conditions NO2 and NO3
were made with the precursor of ADN, meaning ammonium elute at 2.8 and 3.1 min, respectively, followed by ANU at
nitrourethane. The precursor was isolated and cleaned in a about 3.4 min. If measured at 214 nm, the limit of detection
suitable manner and measured under the IC conditions (LOD) for nitrite is about 0.5 ppm and about 0.05 ppm for
improved for nitrate and dinitramide. As expected, a peak nitrate. Using 220 nm as detection wavelength resulted in a
with the same retention time at about 3.4 min was detected. LOD of about 0.3 ppm for nitrate. Using a wavelength
In order to optimize the signal-to-noise ratio an UV between 210 and 220 nm results in a LOD for ANU of about
spectrum was measured for ammonium nitrourethane. 1 ppm. The detection limits of nitrate and nitrite are not
As Figure 4 shows, ANU absorbs with a main maximum at quite as good as observed for the standard anion analysis
about 260 nm, having a second maximum at 210 nm with method using the ANION-R-column in combination with
lower intensity. Similar to the analysis of dinitramide, the pHBA and an electrical conductivity detector. While the
most sensitive detection wavelength for nitrourethane latter method shows detection limits of about 0.1 to
should be near to the maximum of 260 nm. Using an UV 0.01 ppm, these high limits are not needed if realistic
detector with only one detection wavelength like it was used ADN samples, especially after thermal treatment, have to
in this study means that the measuring wavelength for be analyzed. For the dinitramide estimation a very broad
nitrate and ANU should be the same due to the recognized linearity of the method was observed. With a correlation
retention times of the Ä… well separated Ä… peaks. A switching factor of R ˆ 0.9992 linear proportional signal areas for DN
of the wavelength like it is used for the later eluting were recorded in the range of 0.5 and about 750 ppm for
dinitramide peak is not possible between NO3 and ANU. ADN. Also, the higher spreading and LOD for nitrite could
be accepted because in thermally treated ADN samples,
analyzed until now, mainly nitrate and only in very few cases
nitrite had been detected. The calibration of nitrate was
found to be linear up to 20 ppm and the peak area for ANU
to be linear proportional at least in the range of 3 and
150 ppm. Assuming a typical weight concentration of about
300 to 600 mg/l for an ADN sample to be analyzed, the limit
of detection enables the determination of 0.01 to 0.02% of
nitrate and of about 0.03 to 0.06% of ANU in one run
parallel to a high dinitramide content.
4.6 Analysis of Realistic Treated ADN Samples
Under improved measurement conditions also treated,
realistic samples were analyzed for their nitrate content. The
results showed that the prilling of ADN does not substan-
tially increase the low nitrate content in ADN which is
contained from the synthesis of ADN. Unstabilized ADN,
aged under different temperatures for several periods, was
also analyzed. The results showed an increased temperature
treatment and duration yielding an increased nitrate content
besides a decreasing dinitramide part while nitrite was not
detected. Assuming the nitrate to be NH4NO3 and adding its
Figure 4. Comparison of the UV spectra of potassium nitrate
and ammonium nitrourethane. content to the ADN content leads to about 95 to 96%. The
124 G. Bunte, H. Neumann, J. Antes and H. H. Krause Propellants, Explosives, Pyrotechnics 27, 119 Ä… 124 (2002)
Figure 5. Variation of the IC peak-intensities for nitrate and nitrourethane (about 25 ppm each) in dependence of the used UV
detection wavelength (Ion Pac 11 with 0.3 mmol NaOH at 1 ml/min).
"!Challenges in Propellants and Combustion+", Begell House,
rest could possibly be assumed to be water. Whether
New York, Wallingford 1997, pp. 627 Ä… 635.
ammonium nitrourethane (ANU) plays a role in the kinetics
(4) S. LĆbbecke, H. H. Krause, and A. Pfeil, "!Thermal Behavior
of thermal destruction of ADN containing formulations will
of Ammoniumdinitramide+", 27th Int. Annual Conference of
be an objective of future studies. Until now, ANU was only
ICT, Karlsruhe, Germany, June 25 Ä… 28, 1996, pp. 143/1-4.
analyzed for newly synthesized ADN to judge the cleaning (5) S. LĆbbecke, H. H. Krause, and A. Pfeil, "!Thermal Decom-
position and Stabilization of Ammonium Dinitramide
procedure.
(ADN)+", 28th Int. Annual Conference of ICT, Karlsruhe,
Germany, June 24 Ä… 27, 1997, pp. 112/1 Ä… 8.
(6) Ch. Frenck, W. Janitschek, and W. Weisweiler, "!Die Reaktion
5 Literature
von Ammoniak mit Distickstoffpentoxid+", 29th Int. Annual
Conference of ICT, Karlsruhe, Germany, June 30 Ä… July 3, 1998,
pp. 50/1 Ä… 12.
(1) V. A. Tartakovsky, "!The Design of Stable High Nitrogen
(7) U. Teipel, T. Heintz, K. Leisinger, and H. H. Krause,
Systems+", in: Th. B. Brill et al. (eds.), "!Decomposition,
Combustion, and Detonation Chemistry of Energetic Materi- "!Formation of Ammonium Dinitramide (ADN) Particles+",
29th Int. Annual Conference of ICT, Karlsruhe, Germany, June
als+", Materials Research Society, Pittsburgh 1996, Materials
30 Ä… July 3, 1998, pp. 63.1 Ä… 63.14.
Research Society, Symposium Proceedings, 418, pp. 15 Ä… 24.
(2) A. Langlet, N. Wingborg, and H. ÷stmark, "!ADN: A New
High Performance Oxidizer for Solid Propellants+", in: K. K.
Kuo (ed.), "!Challenges in Propellants and Combustion+",
Begell House, New York, Wallingford 1997, pp. 616 Ä… 626.
(3) M. L. Chan, A. Turner, L. Merwin, G. Ostrom, C. Mead, and
St. Wood, "!ADN Propellant Technology+", in: K. K. Kuo (Ed.), (Received March 25, 2002; Ms 2002/015)