Do pyrotechnics contain radium?

Georg Steinhauser and Andreas Musilek

Vienna University of Technology, Atominstitut der Österreichischen Universitäten, Stadionallee 2, A-1020 Wien, Austria

E-mail: georg.steinhauser@ati.ac.at

Received 23 April 2009
Accepted 31 July 2009
Published 24 August 2009

Contents

• 1. Introduction
• 2. Materials and methods
• 3. Results and discussion
• 4. Conclusions
• Acknowledgments
• References

1. Introduction

Fireworks are probably the application of chemistry with the best resonance with the general public. Nonetheless, fireworks are increasingly raising environmental concerns. Although the problem of pollution caused by fireworks (and other civil and military pyrotechnic applications) had been identified many years ago [1, 2], the number of environmental studies focusing on this problem has dramatically increased quite recently, e.g. [3–11]. Also the search for environmentally benign pyrotechnic formulations exhibits a rapidly expanding scientific field and has not hit its peak yet [5].

Pyrotechnics are thermodynamically metastable mixtures which consist of at least two basic constituents: the reductant/fuel (e.g. magnesium, aluminium, magnalium alloy, sulfur, charcoal, red phosphorus, etc) and the oxidizer (alkali metal or alkaline earth metal nitrates, perchlorates, chromates, metal oxides, etc). Several additives may find application in pyrotechnics in order to obtain a certain intended effect (e.g. colouring agents, propellants, smoke or sound generators, etc). Colours in pyrotechnics are obtained by the addition of compounds of elements with the desired flame colour. For red light, strontium nitrate is used; barium nitrate for green light; sodium oxalate or cryolithe (Na₃AlF₆) for yellow; and any copper/chlorine system (compounds or mixtures) for blue (see table 1 for some typical compositions of pyrotechnics). During combustion, very short-lived and unstable compounds, such as the monochlorides of alkaline earth metals (SrCl, BaCl) are formed, which emit light in the desired spectra [5, 12, 13]. The formation of the monochlorides thus depends on the presence of a chlorine source. If no chlorine donor is added to a pyrotechnic formulation, barium nitrate causes combustion under the emission of almost white light. This is the reason why barium nitrate is not only used as an oxidizer in green or greenish flares (with a chlorine donor, which is typically PVC powder) but also for white and yellow flares (without a chlorine donor), as shown in table 1. In the presence of chlorine, barium nitrate acts as a combined pyrotechnic oxidizer and colouring agent.

From an environmental and toxicological point of view, the formation of barium-rich aerosols following the display of a firework is a problem. The inhalation of barium-rich aerosols has adverse affects on the lungs and heart and causes muscle cramps [14, 15]. In cases of fireworks and pyrotechnics, barium compounds are set free in the form of mostly water-soluble and thus bioavailable compounds: BaO, Ba(OH)₂, BaCl₂ and undecomposed Ba(NO₃)₂. The raw material of barium compounds is generally barium sulfate (barite). In 2006, approximately 8 million tons barite have been produced by mining worldwide [16]. Only a very minor percentage is used in pyrotechnics. The major amount of barium sulfate is used as a pigment (Blanc fixe), or as a filler for paper, paint, varnish, rubber, etc. This mineral is also used as a constituent of heavy concrete for the shielding of ionizing radiation.

To the authors' knowledge, the potential hazard of fireworks due to liberated radionuclides has never been the subject of investigation in the scientific literature before. The radioactive alkaline earth metal radium has very similar chemical properties to barium (and also strontium), as they occur in the same group of the periodic table. This similarity is used, for example, in the preconcentration of radium from water by coprecipitation with barium in the form of Ba(Ra)SO₄. Natural sequestering leads to the formation of so-called radiobarite minerals. These minerals accumulate all naturally occurring radium isotopes, in particular the ²³⁸U decay chain member ²²⁶Ra (half-life T_1/2 = 1600 a) and the ²³²Th decay chain member ²²⁸Ra (T_1/2 = 5.76 a). However, the potential accumulation of ²²⁶Ra is of higher environmental significance than ²²⁸Ra, because the latter is simply too short-lived to be extremely accumulated in barium (or strontium) deposits. If young enough, these minerals and ores have remarkable ²²⁶Ra activities. On geological timescales, however, ²²⁶Ra has a relatively short half-life. If radiobarite minerals, therefore, are older than 10–20 ka and isolated from any further radium supply, they slowly lose their radioactive properties. The radiobarite-rich sludges and scalings at oil-field-production sites investigated by Zielinski et al [17], have ²²⁶Ra activities in the range between 3 and 130  Bq g^–1, with one sample as active as 4.9 kBq g^–1. In their study, the ²²⁸Ra activities have been found to be always lower than the ²²⁶Ra values. Radiobarite ores in the Ohře Rift (Bohemian Massif) have activities between 0.02 and 7.80  Bq g^–1 [18]. The scales and tailings in Polish hard coal mining sites were reported to contain radiobarites with activities in the range of 40–100  Bq g^–1 for ²²⁶Ra and 27–62 Bq g^–1 for ²²⁸Ra (barium-rich Rontok scale), and 5.3–6.4 Bq g^–1 for ²²⁶Ra and 6.4–8.5 Bq g^–1 for ²²⁸Ra (barium-poor Bojszowy tailings), respectively [19]. The ambient γ-dose rates are strongly elevated with more than 1 µSv h^–1 at both sites.

Previous studies [20, 21] have investigated the trace element content of pyrotechnics and their poisoning potential. From an economic point of view, it is clear that raw materials for the production of fireworks are usually not purified beyond the grade which is necessary for the intended effects. This explains why the fireworks investigated in those studies contained significant traces of heavy metals which do not have a pyrotechnic function. The utilization of radium-rich barium and strontium ores would, therefore, involve the risk that radium might be processed together with barium and strontium into the final product. The display of such radium-containing pyrotechnics would set the radioactive material free in the form of easily inhalable aerosols. The incorporation of α-emitting radionuclides (such as ²²⁶Ra) is a major health threat in human radiation protection. The ingestion or inhalation of α-emitters should thus be avoided under all circumstances. In order to examine this potential hazard, we applied radioanalytical methods to investigate the radioactivity of pyrotechnics purchasable in Austria.

2. Materials and methods

Fourteen samples of pyrotechnic devices (sky rockets, shell-type rockets, volcanoes) have been investigated with γ-spectrometry in this study (see table 2). The samples were weighed and filled into cylindrical polyethylene (PE) containers (comparable filling level). In principle, for the quantification of the ²²⁶Ra activity, two methods are possible: the 186 keV γ-photon emitted by the nuclide itself can be measured. Alternatively, the γ-photons of its decay products ²¹⁴Pb and ²¹⁴Bi can be measured, as they are in equilibrium with ²²⁶Ra after three or four weeks (due to the short half-lives of ²¹⁴Pb and ²¹⁴Bi as well as the intermediate ²²⁶Ra-daughter nuclides ²²²Rn and ²¹⁸Po, respectively), see scheme 1. Since radon is known to diffuse through many materials (sample vials), causing a loss of activity, the latter method appears to be the less reliable for our analytical purposes.

When the 186 keV γ peak of ²²⁶Ra is used, the possible interference of ²³⁵U, which also emits γ photons in this energy region, has to be considered. However, since we can assume that uranium in environmental samples must be present in its natural isotopic ratio, a γ spectrum showing a ²³⁵U peak should also show the γ peaks of the short-lived ²³⁸U granddaughter ²³⁴Pa (with several γ photons at 1001, 743, 786 keV, etc), as shown in scheme 1. Since we did not detect any ²³⁴Pa in our samples (detection limit approx. 20 mBq g^–1), the uranium content in pyrotechnics can be regarded as negligible. Consequently, any radium in the sample cannot be due to a contamination with uranium minerals being in equilibrium with the daughter ²²⁶Ra. Rather, it must be a significant enrichment of radium itself in one of the raw materials.

Gamma-spectrometry was performed on the novel low-level counting facility of the Atominstitut, consisting of a 226 cm³ HPGe detector (Canberra™, detector model GC5020; 2.0 keV resolution at the 1332 keV ⁶⁰Co peak; 52.8% relative efficiency), connected to a PC-based multi-channel analyser with preloaded filter. The measurement position of the sample was fixed at a distance of approximately 11 cm on top of the detector. The new detector system is characterized by only approximately one-tenth of the background of the other γ detectors of the radiochemistry group in the same institute. This is due to the improved shielding of the detector by the ORTEC™ HBLBS1 shielding (solid-cast virgin lead with steel casing, total weight 1134 kg). For calibration of the detector's efficiency for ²²⁶Ra, 50  µl of QCY48 (Amersham^® Ltd) solution in hydrochloric acidic solution (comparable bulk density) was used. The measurement times of the pyrotechnics were at least 1 week, or longer, until no significant improvement of the counting error of the most interesting peaks could be yielded by a—reasonably—longer measurement time. The standard solution was measured for 328 000 s. A background spectrum was recorded (1 816 000 s) and considered for the evaluation of the γ spectra of the pyrotechnics. For quantification, the γ photons with characteristic energies were used as listed in table 3. All nuclear data in this paper are taken from the National Nuclear Data Center [22].

3. Results and discussion

The results of the γ spectrometry are shown in table 4.

The main activity in pyrotechnics is due to ⁴⁰K (up to 20  Bq g^–1, mean value 14 Bq g^–1). The presence of ⁴⁰K in the mixtures can easily be explained by the application of oxidizing potassium salts (nitrate, perchlorate) in pyrotechnics. Black powder, for example, consists of some 75% of potassium nitrate; the increased ⁴⁰K activity, therefore, did not come as a surprise.

Much more noteworthy is the presence of ²²⁶Ra (and/or its daughter nuclides ²¹⁴Pb and ²¹⁴Bi, respectively) in almost all pyrotechnics investigated. According to the hypothesis of this study, ²²⁶Ra in the samples must be the result of the utilization of slightly active radiobarite ores for the production of the pyrotechnic raw material barium nitrate. As stated in the introduction, the use of barium salts in pyrotechnics is not restricted to green luminescent formulations, as it may be applied as an oxidizer with `neutral colour', if no chlorine donor is added to the mixture. Accordingly, this explains one of the highest specific radium activities (120  mBq g^–1) in sample R8—a rocket with a yellow effect.

The highest specific ²²⁶Ra activity was 260 mBq g^–1 (sample R11), whereas only in three samples was the detection limit not exceeded. The average value of all 14 samples was 81  mBq g^–1. Pyrotechnics thus exhibit a specific radium activity that is approximately one order of magnitude lower than the lowest active radiobarite sample investigated by Zielinski et al [17]. The specific radium activities of our samples, however, correspond to those found in the radiobarites of the Ohře Rift (Bohemian Massif). Some pyrotechnic samples exceeded the lowest active samples from the Ohře Rift (0.02 to 7.80  Bq g^–1 radium) [18].

We could also find similar levels of ²²⁶Ra activity in red luminescent pyrotechnics, where we actually expected much lower activities due to the less obvious chemical similarities of strontium (red colourant) and radium. However, it seems that both alkaline earth ore deposits, strontium and barium, respectively, similarly accumulate radium from the geological environment.

Comparing the ²¹⁴Pb activities to the respective ²¹⁴Bi activities shows very good agreement. This observation comes as expected, because the short-lived nuclide ²¹⁴Bi must be in radioactive equilibrium with its mother ²¹⁴Pb. Since the principal γ photons of both nuclides have their energy in different regions of the spectrum (352 keV for ²¹⁴Pb and 609 kev for ²¹⁴Bi), this observation leads to the conclusion that our simplified approach of `comparable bulk densities' of samples and QCY48 standard solution is a valid approach for these measurements. If the γ-photon self-absorption properties of sample and standard were significantly different, this would have led to a deviation in the evaluation of γ photons with different energies (and thus different detector efficiencies). Some occasionally higher deviations are due to higher counting errors.

The activities of ²¹⁴Pb and ²¹⁴Bi were in the same range (at least in the same order of magnitude) as the ²²⁶Ra activities. Remarkably, the ²²⁶Ra activities were a little higher than the activities of the daughter nuclides in most cases. This must be, indeed, due to the leaking from the PE sample container.

In contrast to uranium, ²³²Th and its progeny appeared to be common impurities in pyrotechnics. Since the members of the ²³²Th decay chain are generally characterized by much shorter half-lives, the thorium decay chain reaches its radioactive equilibrium within only a few decades [23]. The presence of ²²⁸Ra (and its short-lived daughter nuclide ²²⁸Ac) in pyrotechnics, therefore, can hardly be the result of a natural, selective incorporation of ²²⁸Ra into the barium or strontium ore. Radium-228 (T_1/2 = 5.75 a), without a supply from the mother nuclide ²³²Th, would decay completely after a few decades.

The specific activities of ²²⁸Ac and ²¹²Pb correlate to a high degree, although, as with the ²³⁸U chain, a gaseous decay product lies in between, namely ²²⁰Rn. In this specific case, however, we believe that the much shorter half-life of the noble gas ²²⁰Rn (T_1/2 = 55.6 s) does not allow a significant diffusion through the PE container. Consequently, ²²⁰Rn and all its decay products are trapped in the container and measured without any losses.

According to the Swiss Administration [24], 1700 tons of pyrotechnics are annually consumed in Switzerland. We can safely assume that these numbers compare to Austria as well. Assuming an average radium content in pyrotechnics of 81 mBq ²²⁶Ra g^–1 and 14 Bq ⁴⁰K g^–1, these numbers will correspond to an annual emission of 138 MBq ²²⁶Ra and 23.8 GBq ⁴⁰K year^–1 from pyrotechnics only (in countries like Switzerland or Austria).

4. Conclusions

Pyrotechnics do contain radium—not extreme activities but certainly enough to be detected in our new low-level γ spectrometer. The ²²⁶Ra activities are in the range of 16–260 mBq g^–1 and must be due to the selective incorporation of radium into the barium and strontium ore body, as no traces of the mother nuclide (²³⁸U and its early daughters) could be found.

In contrast to this, ²³²Th and its decay chain members (like ²²⁸Ra) probably is present in the form of thorium-containing minerals, which are an impurity in the raw materials of pyrotechnics. Due to its much shorter half-life, ²²⁸Ra is unlikely to accumulate significantly in the ore body of an alkaline earth metal deposit. This hypothesis is further supported by the very weak or non-existing correlation of the activity concentrations of the two radium isotopes.

Although the radium activities were relatively low, we believe that this potential hazard should be considered as well in the development of environmentally friendly pyrotechnics [5]. The incidental utilization of highly active radiobarite ores as pyrotechnic raw materials would probably exhibit a greater health hazard than the toxicity of the poisonous pyrotechnic constituents (heavy metals, perchlorate) [5]. In particular, when applied as indoor pyrotechnics, the health aspects of pyrotechnic devices should be taken seriously with respect to air pollution by toxic and radioactive substances. This finding justifies the efforts that are currently being undertaken in the search for barium-free pyrotechnics [25]. One possible method to remove radium from the radiobarite ore, however, could be to set it under reducing conditions which causes the complete mobilization of radium from the barite [26].

Acknowledgments

GS thanks the Austrian Science Fund (FWF) for financial support (Erwin Schrödinger-Stipendium, project no. J2645-N17). We also thank the Vienna University of Technology for supporting this work.

References

[1]

Bach W, Dickinson L, Weiner B and Costello G 1972 Some adverse health effects due to air pollution from fireworks Hawaii Med. J. 31 459–65 
PubMed

[2]

Smith R M and Vu D D 1975 Changes in forced expiratory flow due to air pollution from fireworks Environ. Res. 9 321–31 
CrossRefPubMed

[3]

Coghlan A 2008 Great fireworks, shame about the toxic fallout New Sci. 200 12 
CrossRef

[4]

Halford B 2008 Pyrotechnics for the Planet Chem. Eng. News 86 14–8 
CrossRef

[5]

Steinhauser G and Klapötke T M 2008 `Green' pyrotechnics: a chemists' challenge Angew. Chem. Int. Edn 47 3330–47 
CrossRef

[6]

Vecchi R, Bernardoni V, Cricchio D, D'Alessandro A, Fermo P, Lucarelli F, Nava S, Piazzalunga A and Valli G 2008 The impact of fireworks on airborne particles Atmos. Environ. 42 1121–32 
CrossRef

[7]

Steinhauser G, Sterba J H, Foster M, Grass F and Bichler M 2008 Heavy metals from pyrotechnics in New Years Eve snow Atmos. Environ. 42 8616–22 
CrossRef

[8]

Wang Y, Zhuang G, Xu C and An Z 2007 The air pollution caused by the burning of fireworks during the lantern festival in Beijing Atmos. Environ. 41 417–31 
CrossRef

[9]

Moreno T, Querol X, Alastuey A, Cruz Minguillon M, Pey J, Rodriguez S, Vicente Miro J, Felis C and Gibbons W 2007 Recreational atmospheric pollution episodes: inhalable metalliferous particles from firework displays Atmos. Environ. 41 913–22 
CrossRef

[10]

Wilkin R T, Fine D D and Burnett N G 2007 Perchlorate behavior in a municipal lake following fireworks displays Environ. Sci. Technol. 41 3966–71 
CrossRefPubMed

[11]

Drewnick F, Hings S S, Curtius J, Eerdekens G and Williams J 2006 Measurement of fine particulate and gas-phase species during the New Year's fireworks 2005 in Mainz, Germany Atmos. Environ. 40 4316–27 
CrossRef

[12]

Webster H A 1985 Visible spectra of standard Navy colored flares Propellants Explos. Pyrotech. 10 1–4 
CrossRef

[13]

Conkling J A 1985 Chemistry of Pyrotechnics (New York: Dekker) 

[14]

Reeves A L 1979 Barium Toxicity (Amsterdam: Elsevier) pp 321–8 

[15]

Hicks R, Caldas L Q d A, Dare P R M and Hewitt P J 1986 Cardiotoxic and bronchoconstrictor effects of industrial metal fumes containing barium Arch. Toxicol., Suppl. 9 416–20 
PubMed

[16]

Miller M M 2007 Barite USGS 2006 Minerals Yearbook (Reston, VA: US Geological Survey) pp 9.1–9.7 

[17]

Zielinski R A, Otton J K and Budahn J R 2001 Use of radium isotopes to determine the age and origin of radioactive barite at oil-field production sites Environ. Pollut. 113 299–309 
CrossRefPubMed

[18]

Ulrych J, Adamovic J, Zak K, Frana J, Randa Z, Langrova A, Skala R and Chvatal M 2007 Cenozoic `radiobarite' occurrences in the Ohre (Eger) Rift, Bohemian massif: mineralogical and geochemical revision Chem. Erde-Geochem. 67 301–12 
CrossRef

[19]

Leopold K, Michalik B and Wiegand J 2007 Availability of radium isotopes and heavy metals from scales and tailings of Polish hard coal mining J. Environ. Radioact. 94 137–50 
CrossRefPubMed

[20]

Alenfelt P 2000 Chemical analysis of consumer fireworks J. Pyrotech. 11–5 

[21]

Sukumar A and Subramanian R 1992 Trace elements in scalp hair of manufacturers of fireworks from Sivakasi, Tamil Nadu Sci. Total Environ. 114 161–8 
CrossRefPubMed

[22]

Brookhaven National Laboratory 2009 National Nuclear Data Center http://www.nndc.bnl.gov/ Accessed April 2009 

[23]

Poljanc K, Steinhauser G, Sterba J H, Buchtela K and Bichler M 2007 Beyond low-level activity: on a `non-radioactive' gas mantle Sci. Total Environ. 374 36–42 
CrossRefPubMed

[24]

Eidgenössisches Justiz- und Polizeidepartment 2009 http://www.fedpol.admin.ch Accessed April 2009 

[25]

Steinhauser G, Tarantik K and Klapoetke T M 2008 Copper in Pyrotechnics J. Pyrotech. 27 3–13 

[26]

Huck P M, McClymont G L, Schwartz F W, Nesbitt B E, Anderson W B and Kratochvil B 1989 Modeling of radium-226 leaching from barium-radium sulfate sludges Waste Manage. 9 157–63 
CrossRef

[27]

Jennings-White C 1993 Nitrate colors Pyrotechnica 15 23–8