Vol. 48, 01, 05ok.qxd 167 ANNALS OF GEOPHYSICS, VOL. 48, N. 1, February 2005 Key words gas geochemistry – earthquake precur- sors – greenhouse gases 1. Introduction Numerous studies on the natural sources of greenhouse gases, such as methane and carbon dioxide, have been carried out in relation to the evolution of the Earth atmosphere as well as in the fields of geological and geochemical re- search. Our primary interest lies in the effects of greenhouse gases on the global climate. It is well-known that the greenhouse gas balance is evaluated according to three major sources of emission: abiogenic (volcanoes or thermometa- morphism); non-anthropogenic biogenic (vege- tation, earth, oceans) and anthropogenic. Inorganic CO2 emissions form a significant part of the total balance. The origin of this CO2 concentration is generally attributed to volca- noes, geochemical cycles and thermometamor- phic degassing produced by contact between carbonates and magma. Thermometamorphic generation of CO2 has been discussed by Gi- anelli (1985), Marini and Chiodini (1994), Ro- gie et al. (2000), etc. Other greenhouse gases, such as abiogenic methane, are generally attrib- uted to deep degassing. Chiodini et al. (2000) estimated that ther- mometamorphic processes are responsible for about 36% of the CO2 measured in Central Italy, while 41% comes from deep sources and 23% is biological. Mailing address: Dr. Giovanni Martinelli, ARPA – Agen- zia Regionale Prevenzione e Ambiente dell’Emilia-Romagna, Sezione Provinciale di Reggio Emilia, Via Amendola 2, 42100 Reggio Emilia, Italy; e-mail: giovanni.martinelli15@tin.it Carbon dioxide and methane emissions from calcareous-marly rock under stress: experimental tests results Giovanni Martinelli (1) and Paolo Plescia (2) (1) ARPA – Agenzia Regionale Prevenzione e Ambiente dell’Emilia-Romagna, Sezione Provinciale di Reggio Emilia, Italy (2) Istituto per lo Studio dei Materiali Nanostrutturati, ISMN-CNR, Montelibretti (RM), Italy Abstract The identified emissions of abiogenic carbon dioxide, carbon monoxide and methane are generally attributed to volcanic activity or to geochemical processes associated with thermometamorphic effects. In this paper we show another possible abiogenic source of emission, induced by mechanical, and not thermal, stresses. We investigat- ed the mechanochemical production of carbon dioxide and methane when friction is applied to marly-type rock and studied the mechanisms determining the strong CO2 and CH4 emissions observed. A ring mill was used to apply friction and oriented pressure upon a synthetic calcite-clay mixture of varying proportions. We found that the CO2 and CH4 release versus the grinding action has a non-linear trend reflecting the behaviour of decreas- ing crystallinity, which indicates a close link between crystallinity and gas production. For the CO2 emission, we propose a release mechanism connected with the friction-induced fractures and the increase in structural disor- ders induced by creep in the lattice. The CH4 emission could be explained by a Sabatier reaction in which CO2 and hydrogen are involved to form CH4 and water. 168 Giovanni Martinelli and Paolo Plescia Another interesting debate on the origin of greenhouse gases concerns the relationship be- tween the greenhouse gas effect and the Creta- ceous mass extinction. Many authors consider that the mass extinction of vertebrates at the end of the Cretaceous system was due to a meteorite impacting a carbonate basin; this induced the generation of enormous quantities of CO2 through the dissociation of carbonates as an ef- fect of the shockwave (Ryder et al., 1996; Mac- Leod and Koeberl, 2001). However, the studies do not reach agreement on the mechanisms in- volved. Work on the carbonate compressibility by shockwaves shows how the CO2 dissociation threshold varies from 10 to 60 GPa (Agrinier et al., 2001; Skála et al., 2001). According to other studies, carbonate dissociation could occur at much more limited pressures, but under cer- tain stress conditions (Dickinson et al., 1991). A recent work demonstrated the possibility of obtaining an intense CO2 emission from car- bonate rocks with the contribution of mechani- cal energy by grinding alone (Martinelli and Plescia, 2004). Methane is the subject of particular study, not only because of its environmental implica- tions (greenhouse effect) but also in regard to its possible abiogenic origins of part of the oil reservoirs. A non-biogenic source of methane and more complex carbon composites could radically modify many of our current beliefs concerning the formation of oil and methane reservoirs (Beskrovny and Tikhomirov, 1968; Anders et al., 1973; Porfir’ev, 1974; Kropotkin and Valyaev, 1976, 1984; Kropotkin, 1985). Another important reason for studying the production/emission mechanisms of carbon dioxide and methane is the monitoring and pre- diction of seismicity for which there is general agreement on the importance of endogenic de- rived gases as possible precursors (e.g., Irwin and Barnes, 1980; Geller, 1997; Toutain and Baubron, 1999). The deformation processes that occurred in central Italy in 1997 and 1998 generated geo- chemical anomalies in the form of gas emissions and thermal springs within a 70 Km radius of the epicentre (Heinicke et al., 2000; Favara et al., 2001; Italiano, 2001). These geochemical anom- alies can be grouped together as an impressive degassing process due to the variations in crustal permeability generated by deformation. To study the CO2 / CH4 emission from dy- namic compression, it has been necessary to use a system that can induce a large amount of fracturing through strong compression and fric- tion and with energy sufficiently high as to re- duce crystallinity, i.e., replicating what happens in rock during a strong tectonic activity. 2. Experimental For laboratory work a ring mill was used, which is the most efficient grinding system available at present for the simulation of fric- tion and mechanical compression over a large surface. Thanks to the mill we have (Martinelli and Plescia, 2004): a high exchange surface; high pressure on the materials; an isochemical ambient with controlled elementary exchange, and a high reaction rate. The ring mill consists of a jar reinforced to contain the material and a grinding body (a steel ring and solid steel roller). A ring mill ba- sically operates through non-hydrostatic com- pression (impact of the rings on the material particles) and friction (rotation forced by the grinding body on the particles and between the particles and the jar walls). Extremely high pressures can be achieved and complete amor- phization of minerals with a Mohs hardness of less than 3-4 can be obtained within a few min- utes (Plescia et al., 2003). The exchange sur- face is very high as the pulverized material has very high specific surfaces and the quantity of material placed in the mill can exceed 10 g. For our experimental trials we used a Labtech ESSA continuous ring mill conceived for continuous production of about 15 kg/h and modified to accommodate diagnostic instru- ments (fig. 1), i.e., a konimeter to measure gas and dust and an analyzer for the internal temper- ature. Outside the grinding jar, a magnetic pick- up connected to a digital oscilloscope provides data on the internal oscillations of the metallic mass, the grinding frequency and the dynamic events occurring during grinding (fig. 2). The milled material consists of a mixture of pure calcium carbonate for analyses (Merck) 169 Carbon dioxide and methane emissions from calcareous-marly rock under stress: experimental tests results and kaolinitic clay. The clay sample, which was diffraction-analyzed before grinding, was found to be contaminated by quartz and a small quantity of illite. Two types of tests were performed: for the first, non-stop grinding was carried out for 30, 60, 120 and 180 min on fixed amounts of 5 g; for the second, fixed-time grinding for samples of 2, 2.5, 5, 10 g was carried out. The trials on samples of different grinding times and/or different weights allowed us to verify the effects of grinding with different amounts of energy per unit weight. The mixtures were analyzed by TGA and DTA thermoanalyses to determine the amount of CO2 released, the temperature and the Fig. 1. Layout of experimental apparatus. Fig. 2. Fourier analysis of mill impact frequencies ad different grinding times. Table I. Experimental conditions. Grinding device: LABTECH ESSA Jar material: steel Grinding bodies: one 1 kg internal cylinder, two 1.88 kg rings Speed: 1200 rpm FTIR - Gas cell experiment FTIR: Nicolet Avatar 360 Sample cell: TGA thermostatic transfer line and «100 m» long path cell Carrier gas: argon, 99.9% pro analysis X-ray diffraction: Rigaku Rint 1200 Anode: Cu, radiation Cu kα Monochromator: Graphite crystal Power supply: 30 kV-30 mA Differential thermal analyzer: Stanton 1500 DTA-TGA simultaneous analyzer Reference α-Al2O3 Ramp in air from 25 to 1000°C at 5°/min 170 Giovanni Martinelli and Paolo Plescia amount of energy utilized in dissociation. The mixtures had a CO2 content equal to 32% and an OH content equal to 6% (data obtained by differential thermal analysis). We analyzed the gas using a FTIR spec- trophotometer with «long path» thermostatic gas cells, similarly to what is used on GC/FTIR or TGA/FTIR systems. The instrument was a Nicolet Avatar 360. We used multiple acquisi- tion software (Nicolet Omnic Series v.6.1A) to analyse the data in real-time and reproduce the reaction behavior. To analyze the gas in the jar and to avoid contamination from the outside, the jar was flushed with argon at a pressure of 1 bar. The gas flow was 0.5/min, controlled by an analogue flowmeter. For the x-ray diffrac- tion and thermoanalyses, we used, respectively, a Rigaku Rint 1200 diffractometer and a Stan- ton STA 1000 thermal balance (table I). 3. Results Practically all the CO2 retained in marly rock was emitted in the grinding test. The quan- tity was slightly lower than that obtained by sto- ichiometry (about 20% as compared to the 22% obtained thermally), while methane, CO and H2O were emitted in measurable quantities. The phenomenon was confirmed by real- time analysis of the gas formed during grinding (figs. 3 and 4). Figures 3 and 4 show the behaviour of the FTIR spectra in the vibration region of CO2, methane and vapor as a function of grinding time for sample of a fixed weight. We can see that the CO2 emission curve shows a complex behavior. There is a rapid increase at the beginning of the phenomenon, followed by a progressive de- crease. During the experiment, the first two min- utes are blank analyses, with the mill off, and no anomalous emission observed. The CO2 emission effect is marked by peaks superimposed at a con- tinuous level, while the methane appears after prolonged treatment, together with water (fig. 3). Figure 5a,b shows a notable difference be- tween thermal curves of the treated and as-is samples. The latter show three main reactions: weight loss due to evaporation of interstitial water (100-120°C); weight loss from dehydrox- ylation of clay, between 450-550°C, and weight loss related to the dissociation of CO2 from cal- cite, between 700 and 900°C. After grinding, the dehydroxylation peak, characteristic of the crystalline phase of clay, disappears; the dissociation peak of CO2 de- creases until it nearly disappears, from 35% for the as-is samples to 6% of the 180′ grinding Fig. 3. FTIR Gas analysis: CO2, H2O, CH4 emissions during milling (1 g sample, 180 min grinding). samples, and a crystallization peak appears at very low temperatures. The vitrification tem- perature, for a mixture calcite/kaolinite clay as is, is around 1300-1400°C. In the grinded samples this temperature is placed at 1100°C. Figure 6 shows the x-ray dif- fraction analysis of grinded samples at different grinding times. The XRD shows how the treat- ment leads to amorphization of the calcite and clay minerals. In particular, calcite tends to amorphous product, instead of crystallize as aragonite, as observed in previous experiments on calcite single crystal grinding (Martinelli and Plescia, 2004). In the 120′ sample, it can be observed only the main reflections of quartz, which consti- tutes the hardest phase of the mixture. 4. Discussion of results The most noteworthy data are related to the modality and typology of emission, which can Fig. 5a,b. Thermal analysis of ground samples: a) TGA analysis; b) differential thermal analysis. a b 171 Carbon dioxide and methane emissions from calcareous-marly rock under stress: experimental tests results Fig. 4. 3D view of FTIR spectra (1 g sample). Fig. 6. XRD analysis of treated samples at different main grinding time phases: QZ – quartz; k – kaolinite; CA – calcite. 172 Giovanni Martinelli and Paolo Plescia be attributed to the mechanical, and not the thermal, phenomenon imposed on the marly material. Worth noting, firstly of all, are the CO2 emissions, which we have explained in terms of two parallel mechanisms: burst emissions, due to the release of CO2 molecules from fresh grind- ing-induced fractures; and continuous emis- sions, deriving from the bulk, i.e., produced by the material which is not yet fractured, but is undergoing compression and micro-cracking (plastic deformation, slippage of planes). Burst emissions seem to be prevalent only during the very early grinding phases, while continuous emissions become dominant after a few minutes’ treatment. The phenomenon of continuous CO2 emission brings us to conclude that the deformed lattice of the carbonate re- leases CO2 up to equilibrium. Instead water and methane emissions originate in a different sce- nario. In fact, we suggest that the methane and water derive from the Sabatier reaction CO H CH H O4 22 2 4 2"+ + . This reaction is moderately exothermic, H° = –165 kJ/mol, and could originate when hydrogen molecules become available during grinding. The protonization event coincides with dehydroxylation of the clay during the grinding of the mixture. Several authors (Agli- etti et al., 1986; Plescia et al., 2003) have clearly demonstrated that phyllosilicates are particularly sensitive to the mechanochemical action, which easily leads to the elimination of the OH groups and to a completely amorphous structure. It is also well-known that clays, par- ticularly kaolinite, are capable of exchanging protons with the external ambient, but only half of the available protons are exchangeable with a simple high-pH reaction, while acid or neutral pH exchange is not viable. The phenomenon could evolve as follows: during the first grinding phase (or friction ac- tion), the material is fractured and the calcium carbonate releases carbon dioxide with spot emission; then, the CO2 emission becomes abundant and continuous, while the chemical ambient becomes strongly alkaline; the clays are destroyed by grinding and lose hydroxyls and, hence, protons, some of which recombine with CO2 molecules to give rise to methane and water – both observable in the FTIR gas spec- tra; at the end of the grinding, the CO2 emission tends to very slowly decrease up to re-equilibri- um of the material. To verify our hypothesis, the role of energy during grinding needs to be carefully evaluated. By using the mathematical model developed to study the ring mills (Plescia, 2003), we calcu- lated that the energy provided by the mill is about 1240 kJ/h (345 J/s). This energy is subdi- vided by the amount of material put in the mill. The energy generated would be sufficient to complete the reaction, and the chemical ambi- ent is alkaline enough to guarantee de-pro- tonization of the clay. It is also worthwhile recalling a particular mineralogical fact: in some minerals present in certain metapelites, such as cordierite, signifi- cant quantities of hydrocarbons derived from methane were revealed by IR (Mottana et al., 1983; Khomenko and Langer, 1999). The pres- ence of methane and related hydrocarbons could be associated with gas formation during metamorphic events. 5. Conclusions The data obtained so far allow us to put for- ward the following hypotheses: in addition to the known mechanisms involved in CO2 and methane production, there exists a natural mechanism that works through a mechano- chemical-type phenomenon; an abundant amount of CO2 is produced during shear stress by mechanochemical molecular dissociation of calcium carbonate; the energies involved are sufficient to dissociate carbonate, so long as there is a very high grinding body/material ra- tio; the amount of shear stress necessary to ob- tain the dissociation of CO2 and production of CH4 is less then 4 kbar, the threshold value to obtain the crystallization of aragonite. The phenomenon observed in the laboratory could be tentatively extrapolated to a much larger scale so as to give us some idea of what actually happens when carbonate-rock mass undergoes tectonic activity. 173 Carbon dioxide and methane emissions from calcareous-marly rock under stress: experimental tests results REFERENCES AGLIETTI, F., P.J.M. LOPEZ and E. PEREIRA (1986): Mechanochemical effects in kaolinite grinding, Int. J. Min. Proc., 16, 135-146. AGRINIER, P., S.R. BOYD, I. MARTINEZ, U. SCHARER, M. JAVOY and A. DEUTSCH (2001): On the kinetics of CaO + CO2– – > CaCO3 and CO2 released during im- pact processes, Ann. Geophysicae, 13 (suppl. III) C738 (abstract). ANDERS, E., R. HAYATSU and M.H. STUDIER (1973): Organ- ic compounds in meteorites, Science, 182, 781-790. BESKROVNY, N.S. and N.I. TIKHOMIROV (1968): Bitumens in the hydrothermal deposits of Transbaykal, in The Gen- esis of Oil and Gas (Nedra Press, Moscow), pp. CHIODINI, G., F. FRONDINI, C. CARDELLINI, F. PARELLO and L. PERUZZI (2000): Rate of diffuse carbon dioxide Earth degassing estimated from carbon balance of re- gional aquifers: the case of Central Apennine, Italy, J. Geophys. Res., 105, 8423-8434. DICKINSON, J.T., L.C. JENSEN, S.C. LANGFORD, P.E. ROSEN- MBERG and D.L. BLANCHARD (1991): CO2 emission ac- companying the fracture of calcite, Phys. Chem. Min- erals, 18, 320-325. FAVARA, R., F. ITALIANO and G. MARTINELLI (2001): Earth- quake-induced chemical changes in the thermal waters of the Umbria Region during the 1997-1998 seismic swarm, Terra Nova, 13 (3), 227-233. GELLER, R.J. (1997): Earthquake prediction: a critical re- view, Geophys. J. Int., 131, 425-450. GIANELLI, G. (1985): On the origin of geothermal CO2 by metamorphic processes, Boll. Soc. Geol. It., 104, 575-584. HEINICKE, J., F. ITALIANO, V. LAPENNA, G. MARTINELLI and P.M. NUCCIO (2000): Coseismic geochemical varia- tions in some gas emissions of Umbria Region (Central Italy), Phys. Chem. Earth, 25, 289-293. IRWIN, W.P. and I. BARNES (1980): Tectonic relations of car- bon dioxide discharges and earthquakes, J. Geophys. Res., 85, 3115-3121. ITALIANO, F., G. MARTINELLI and P.M. NUCCIO (2001): Anomalies of mantle-derived helium during the 1997- 1998 seismic swarm of Umbria-Marche, Italy, Geo- phys. Res. Lett., 28, 839-842. KHOMENKO, V.M. and K. LANGER (1999): Aliphatic hydro- carbons in structural channels of cordierite: a first evi- dence from polarized single-crystal IR absorption spec- troscopy, Am. Min., 84, 1181-1185. KROPOTKIN, P.N. (1985): Degassing of the Earth and the origin of hydrocarbons, Int.. Geol. Rev., 23, 1261- 1275. KROPOTKIN, P.N. and B.M. VALAEV (1976): Development of a theory of deep-seated (inorganic and mixed) ori- gin of hydrocarbons (Goryuchie Iskopaemye: proble- my geologii i geokhimii naftidov i bituminoznykh porod), edited by N.B. VASSOEVICH, Akad. Nauk SSR, 133-144 (in Russian). MACLEOD, K.G. and C. KOEBERL (Editors) (2001): Cata- strophic events and mass extinctions: impacts and be- yond, Geol. Soc. Am. Spec. Pap. 356, pp. 749. MARINI, L. and G. CHIODINI (1994): The role of CO2 in the carbonate-evaporite geothermal systems of Tuscany and Latium, Italy, Acta Vulcanol., 5, 95-104. MARTINELLI, G. and P. PLESCIA (2004): Mechanochemical dissociation of calcium carbonate: laboratory data and relation to natural emissions of CO2, Phys. Earth Planet. Int., 142 (3/4), 205-214. MOTTANA, A., A. FUSI, B. BIANCHI POTENZA, R. CRESPI and G. LIBORIO (1983): Hydrocarbon bearing Cordierite from the Derivo Colico road tunnel (Como, Italy), Neues Jahrb. Mineral. Abh., 148, 181-191. PLESCIA, P., D. GIZZI, S. BENEDETTI, L. CAMILUCCI, C. FANIZZA and F. PAGLIETTI (2003): Mechanochemical treatment to recycling asbestos containing waste, Waste Management, 23, 209-218. PORFIR’EV, V.B. (1974): Inorganic origin of petroleum, Am. Assoc. Petrol. Geol. Bull., 58, 3-33. ROGIE, J.D., D.M. KERRICK, G. CHIODINI and F. FRONDINI (2000): Flux measurements of non-volcanic CO2 emis- sion from some vents in Central Italy, J. Geophys. Res., 105, 8435-8445. RYDER, G., D. FASTOWSKY and S. GARTNER (Editors) (1996): The Cretaceous-Tertiary event and other catas- trophes in Earth history, Geol. Soc. Am. Spec. Pap. 307, pp. 576. SKÁLA, R., J. EDEROVÁ, P. MATEˇJKA and F. HÖRZ (2001): Mineralogical studies of experimentally shocked dolomite: implications for the outgassing of carbonates, in Catastrophic Events and Mass Ex- tinctions: Impacts and Beyond, edited by MACLEOD K.G. and C. KOEBERL, Geol. Soc. Am. Spec. Pap. 356, 571-585. TOUTAIN, J.P. and J.C. BAUBRON (1999): Gas geochem- istry and seismotectonics: a review, Tectonophysics, 301, 1-27.