Introduction

Orogenic gold systems are characterized by abundant carbonic fluid inclusions (rich in CO2)1,2,3, but the role of CO2 in gold mineralization still remains controversial and enigmatic4,5,6. Carbonic fluid inclusions have been recently observed in various types of copper deposits7,8,9,10 as well as in lode silver, lead-zinc and molybdenum deposits11,12. Therefore, there is a need to understand the relationship between CO2 and metallic mineralization.

The mutual conversion between CO2 and organic matter is common in both nature and human activity, as exemplified by photosynthesis and fossil fuel combustion13,14,15,16. Organic matter plays a significant role in metal transport and enrichment in low-temperature hydrothermal environments17. Carboxylic acids, such as acetum, have been discovered in petroleum brines18,19 and fluid inclusions of ore deposits20, and have been shown to transport Pb and Zn as complexes at temperature of <250 °C17,21. CO2 can be transformed into carboxylic by metal catalyst, such as Mn, Pd and Zn22,23,24. This encourages us to infer that, at high P-T conditions, carboxylic acids and their metallic complexes can be stable and facilitate mobilization, migration and enrichment of ore metals; and then, decompose to CO2 with decreasing pressure during upward fluid migration. Thus, the stability of carboxylic acids and their metallic complexes at high P-T conditions is the key to understand the mechanism of and the role of CO2 in mineralization processes, from a new dimension. However, nothing is known about metallic complexes with carboxylic acids at high P-T conditions, due to a shortage of experimental data.

To examine the thermal stability of metallic complexes with carboxylic acids at high P-T conditions, we have conducted experiments on copper acetate solution (7%), using a diamond anvil cell. Despite of strong fluorescence impact of diamond, the symmetry stretching vibration of C-H bond (about 2,941 cm−1), i.e. (υP)2941, was observed in copper acetate solution (Fig. 1, Table 1). In the heating process, the shape of the spectra of the copper acetate solution did not change (Fig. 1), and no new peak appeared on the Raman spectra. The volume of the copper acetate solution is constant and the system evolves along the isochore. In other words, system pressure increases with increasing temperature. This is consistent with the relationship between the Raman shift of quartz (464 cm−1) and pressure (Fig. 2, Table 1). Thus, the isochore of the copper acetate solution is defined as P = 1.89 T + 128.58 (Fig. 2), and equals to a geothermal gradient of 19 °C/km. This indicates that copper acetate is stable at temperatures up to 350 °C under low geothermal gradient conditions.

Figure 1
figure 1

(a) The hydrothermal diamond anvil cell (HDAC)25. (b) The Raman spectra of the C-H symmetry stretching vibration at different temperature and pressure, showing that the copper acetate is stable at high temperature with high pressure (PeakFit V4.12, https://peakfit.updatestar.com).

Full size image
Table 1 The Raman shift of quartz and the symmetry stretching vibration of C-H.
Full size table
Figure 2
figure 2

The relation between temperature and pressure in the experiment of copper acetate solution.

Full size image

The thermal dissociation experiment of copper acetate was conducted with a moissanite anvil cell to avoid strong fluorescence. The sample chamber was filled with copper acetate solution (7%), solid copper acetate and quartz chip (Fig. 3a). The chamber was heated step-by-step from 16 °C to 212 °C, with step interval of 6–22 °C, heating rate of 2 – 5 °C/min and pressure ranging 355–611 MPa (Table 1). Each step lasted for 10–15 minutes to achieve stable temperature and pressure, and to acquire the Raman shift of copper acetate solution. The peak symmetry stretching vibration of C-H bond (about 2,941 cm−1) shifts to higher frequency along with increasing temperature and pressure (Table 1). During heating, the solid copper acetate firstly dissolved (Fig. 3b), and then vapour bubble (Fig. 3d) and native copper grains (Fig. 3c,d,f) appeared. Under microscope, it was observed that solid copper grains suddenly formed at the conditions of 212 °C and 511 MPa, and the experiment stopped if no more copper precipitated. The vapour bubble was composed of CO2, as indicated by the Raman shift (Figs. 3e, 4). Thus, it is concluded that the copper acetate solution is stable at high P-T conditions under low geothermal gradient, and thermally dissociated when the geothermal gradient increases, in the way as below:

$$4{{\rm{Cu}}({\rm{CH}}}_{3}{{\rm{COO}})}_{2}+2{{\rm{H}}}_{2}{\rm{O}}=4{\rm{Cu}}+{{\rm{2CO}}}_{2}+7{{\rm{CH}}}_{3}{\rm{COOH}}$$
(1)
Figure 3
figure 3

The experiment was conducted in a moissanite anvil cell. The solid copper acetate (Cu(Ac)2), quartz (Qz) and copper acetate solution (7%) were enclosed in the hole of a thin rhenium metal gasket (a) With the increment of temperature and pressure, the solid copper acetate dissolved to form blue solution. (b) At 212 °C/ 511 MPa, black grains appeared (c), which were identified to be native copper under microscope (f). When the temperature reduced to 15 °C, the vapour bubble appeared (d), which was proven to be CO2 by Laser Raman (e). Unfortunately, we failed to get the pressure at 15 °C due to coating of native copper on the quartz grain.

Full size image
Figure 4
figure 4

Laser Raman spectra of the vapour and solution. (a) The C-H symmetry stretching vibration is obvious (2945.6), illustrating the copper acetate still exist. The two peaks can be observed (1382.9 and 1410.9), is the peak of CO2 and C=O of Cu(Ac)2, respectively. (b–e) The C-H symmetry stretching vibration is unobvious or vanishing, illustrating the copper acetate does not exist almost. The unique peak can be observed (1383.6, 1389.1, 1388.6 and 1389.2), showing the peak of CO2 (PeakFit V4.12, https://peakfit.updatestar.com).

Full size image

From the reaction Eq. 1 and experiment, new understandings can be drawn out: (1) the organic acids can facilitate metallic transportation via fluids during hydrothermal mineralization. (2) CO2 serves as an important buffer to maintain metallic transportation3,5, because the existence of CO2 in fluid makes the reaction 1 proceeds to the left, keeping CH3COO− stable. (3) The copper acetate solution is stable under high-pressure, and therefore, decompression causes copper acetate dissociation, CO2 escape and Cu precipitation, as similar to those revealed in previous studies1,2,3,4,5,6,7,8,9,10,11,12. (4) Wall-rock carbonation removes CO2 from the solution, and results in precipitation of metals. (5) Decreasing pH can facilitate copper acetate stability and transportation; by contrast, increasing pH accelerates copper acetate dissociation and Cu precipitation, and also causes phyllic alteration (sericite + quartz) in the way of Eq. 2:

$$3{{\rm{KAlSi}}}_{3}{{\rm{O}}}_{8}+2{{\rm{CH}}}_{3}{\rm{COOH}}={{\rm{KAl}}}_{2}[{{\rm{AlSi}}}_{3}{{\rm{O}}}_{10}]{({\rm{OH}})}_{2}+6{{\rm{SiO}}}_{2}+2{{\rm{KCH}}}_{3}{\rm{COO}}$$
(2)

Therefore, the common observation of carbonate, sericite and quartz alterations, and CO2-rich fluid inclusions in hydrothermal deposits11, such as the orogenic-type Cu lodes, corresponds well with the experimental results of stability and thermal dissociation of copper acetate solution.

Methods

The experiment was performed in hydrothermal diamond and moissanite anvil cells25,26, respectively. The sample was enclosed in the 200–400 μm diameter hole of a thin (300–400 μm) rhenium metal gasket by compressing the gasket between two diamond anvil faces27. The temperature of the diamond anvils and samples was controlled and measured using Mo resistance heaters and two attached K-type thermocouples, respectively27. Temperature measurement was corrected using the melting point of phenolphthalein and stearic acid, and the accuracy of reported temperatures is within ±5 °C. A small chip of quartz (0.18–0.20 mm) was put in cell to calibrate internal pressure. Experimental pressure was determined according to the relationship between the Raman shift of quartz and the pressure28,29.

Raman spectroscopy was performed using a Raman micro-spectrometer (Renishaw system RM-1000, Renishaw Group, Gloucestershire, United Kingdom); the slit width was set at 50 μm and the resulting resolution was ±1 cm−1 30. The objective is a Leitz 20× with a working distance of 15 mm. An argon ion laser with a wavelength of 514.5 nm operated at 20 mW was used to illuminate the sample for Raman signal generation. Each spectrum was collected within an accumulation time of 30 s and covering a wavelength of 100–4,000 cm−1 30. The initial experimental temperature was 15 °C, which was gradually increased to 350 °C. In the experiment, the Raman spectrum test was conducted 3–5 min after each change in experimental temperature to ensure that the samples firstly reach equilibrium. The results were processed using PeakFit software.