5.1. General Mineralogical Remarks on Copper Minerals and Associated Phases
Various types of halides, both Cu-bearing and Cu-free, occur in the investigated samples (
Table 4). Atacamite is the most common (hydroxyl)halide in museum specimens, which span from the AD 1631 lavas (Scala, Camaldoli, Uncino;
Figure S1) to the products of the AD 1870–1872 and 1880–1906 periods (
Table 2 and
Table 4). A rare occurrence of distinct and well-crystallized individuals of atacamite has been found in a lava fragment by Russo et al. [
58] (Villa Inglese, Torre del Greco;
Figure S1), who also reported, for the atacamite-bearing 1631 lavas of Camaldoli and Villa Inglese, a more likely medieval age (AD 938 and 1037). In the studied samples, this mineral is coeval with the Cu-vanadates and silicates, azurite and some oxides (hematite, cuprite). The origin of atacamite is commonly related to the supergene oxidation zone of Cu deposits (i.e., porphyry copper, [
7]), especially under arid and saline conditions [
59,
60]. However, it is also related to fumarolic deposition and to weathering of sulfides in subsea black smoker deposits in deep seawater seafloor hydrothermal sites; in these areas, atacamite and paratacamite are the most stable copper salts at the pH and Eh of cold deep seawater, undersaturated in CaCO
3, and appear to be the ultimate sink for the Cu leached by the hydrothermal systems from the oceanic crust [
61]. At Somma–Vesuvius, this mineral has a fumarolic genesis, being deposited during the latest phases of the magma cooling mainly along the lava fractures, as described in the literature ([
35,
58] and references therein). It was also observed [
35,
53] that atacamite is associated with azurite, “
vesbine” and various oxides (hematite, hausmannite and rarely cuprite), as quite similarly observed in the present research.
Paratacamite, the trigonal polymorph of atacamite, was identified in very few occurrences and in the same atacamite-bearing samples, although in restricted parts. It occurs as tiny clusters of crystals or as alteration of tenorite in products of the 1872–1882 period of the historical activity. In agreement with Russo and Punzo [
10], the transformation of tenorite to paratacamite is due to the reaction of fumarolic HCl with the copper oxide, i.e., 2CuO + HCl + H
2O = Cu
2(OH)
3Cl. The compositional features of this basic cupric chloride show the presence of Zn in its crystal lattice; it was demonstrated by Braithwaite et al. [
62] that paratacamite is defined as a mineral with the rhombohedral structure, and composition Cu
3(Cu,M
2+)(OH)
6Cl
2, in which Cu/M
2+atomic is ~11-7, corresponding to ~1/3 to 1/2 the occupancy of the non-Jahn-Teller site by M
2+, which is an essential stabilizing non-Jahn-Teller-distorting cation of suitable radius, such as Zn
2+, Ni
2+, Co
2+, Fe
2+, Cd
2+ and Mg
2+.
Interestingly, together with the Cu-halides, several Cu-free varieties have been found in various mineral assemblages of the studied Somma–Vesuvius specimens, mainly belonging to chlorides and fluorides; halite (NaCl), sylvite (KCl) and fluorite (CaF
2) often occur together, and also in various associations with elpasolite, sbacchiite, challacolloite, thermessaite. In particular, elpasolite belongs to cubic perovskites [
63], as the mineral parascadolaite discovered for the first time at Vesuvius [
64]; other rare fluorides were also very recently found, such as the already cited sbacchite [
41] and verneite [
65]. These fluorides are formed by the reaction of HF with the surrounding rocks, indicating environments with T > 500 °C. Challacolloite was found in the 1906 fumaroles of HT [
11,
12,
35], but no temperature data were recorded; this mineral can form at a temperature range of 360 °C (Vulcano island, Italy) to 550–390 °C (Satsuma–Iwojima volcano, Japan) [
12]. Thermessaite is a rare mineral discovered at Vulcano island, southern Italy [
55] in medium temperature fumarole (around 300 °C) in sub-mm colorless prismatic crystals associated with alunite, sassolite, anhydrite, and metavoltine; it was also identified at Vesuvius in fumaroles from recent activity by Campostrini et al. [
54].
Among the cupric oxides, cuprite is restricted to one sample, labelled as “1631 lava. As many other oxidized copper minerals, cuprite is quite common in the weathered portions of many copper sulfide deposits, whereas is rare in volcanic environments. At Vesuvius, cuprite was detected in the lavas of Camaldoli and Scala [
10] (
Figure S1), in associations with atacamite. Tenorite is more widespread compared to cuprite, both in the investigated samples and in the occurrences from the literature ([
13] and references therein); in fact, in the studied samples, tenorite was detected in various rocks related to 1631 to 1906 eruptions, whereas according to the literature [
10], this oxide was identified in scoriae of the 1760 eruption and in all the eruptions from 1825 onward, with particularly abundant and large crystals in the products of the 1944 eruptive episode. Tenorite is typical of the HT fumaroles (K–Na salts bearing, T > 400 °C) and originates by interaction of gaseous CuCl
2 and water vapor, according to the reaction CuCl
2 + H
2O = CuO + 2HCl [
10,
15]. As stated before, tenorite can be subsequently altered by HCl brines to form paratacamite.
The Cu-bearing carbonates are represented by azurite and malachite, only found in two samples and mainly in association with atacamite. As also reported in the literature, these minerals are rare at Vesuvius and have been found in the so-called 1631 lavas (azurite with atacamite and “
vesbine”), in an ejectum of Lagno Macedonia with magnetite and malachite, and in fumarolic products of the 1872 eruption together with connellite, or in various ejecta (malachite) [
10]. The general scarcity of carbonates (Cu-bearing or not) in the fumarolic mineralizations is not surprising (notwithstanding the abundant CO
2 in volcanic gases), considering that the typical high acidity of the fumarolic environments makes most carbonates unstable [
35]. While azurite composition in the studied samples is close to the ideal formula, malachite is distinctly Zn-bearing, with a partial substitution of Cu
2+ by Zn
2+, giving the formula: (Cu
1−xM
x)
2(OH)
2CO
3 (M = Zn, Mg, Co, Ni, etc.) for members of the malachite–rosasite group [
66,
67].
Cu-bearing sulfates are represented by the largest variety of species in the investigated Somma–Vesuvius samples (followed by the whole of halides) and have been recognized in all the products of the recent activity. Even though these minerals belong to different systematic subgroups of sulfates, they can be gathered in main three types, i.e., (i) anhydrous sulfates (chalcocyanite, cryptochalcite, ±dravertite), (ii) anhydrous sulfates with hydroxyl or halogens (brochantite, euchlorine, linarite, ±chlorothionite), and (iii) hydrous sulfates (chalcanthite, cyanochroite, kröhnkite, leightonite, natrochalcite, ±boothite). These sulfates can be composed of Cu only, or of Cu–K, Cu–Na, Cu–K–Na, Cu–Pb and Cu–Mg (±anions, ±H
2O). As also observed by Pekov et al. [
6], the alkali cations have an important role in the chemical feature of the fumarolic minerals, just as the sulfates; indeed, according to these authors, 46 of 86 sulfates and other oxysalts plus chlorides contain K, Na and Cs, whereas only 16 of 174 copper minerals from the oxidation zones of sulfide ores show alkali elements (K and/or Na). Hence, the alkali cations represent a key factor in determining the unique crystal-chemical nature of fumarolic copper minerals. A remarkable feature was indicated by Balić-Žunić et al. [
35] for the sulfates found in fumaroles, i.e., the characteristic existence of several hydrous forms associated with the anhydrous ones. These varieties have different stability fields, that depend on the humidity and temperature conditions and hence, can appear in various zones of the same fumarole. The anhydrous sulfates are generally unstable and readily hydrate under atmospheric conditions [
35]. This is true also for the Cu-bearing sulfates, for which it is possible to observe the anhydrous phases and the hydrous counterparts; an example is chalcocyanite [(Cu(SO)
4], which rapidly changes to chalcanthite [CuSO
4·5(H
2O)] [
35]. In the studied specimens, these two sulfates were also found in the same sample (127656 E1295), probably in relation to an incomplete alteration (hydration) process of chalcocyanite. However, it is quite common to find complex and various mixtures of the above-mentioned two or three types of sulfates in the same sample, for instance (
Table 4):
- -
chalcocyanite–euchlorine (10907 D1358)
- -
cryptochalcite–euchlorine–natrochalcite (12912 E1457).
It is interesting to note that SEM studies also support the observation that hydrous Cu sulfate generally appears to be as later phases compared to the other sulfates in the paragenetic sequence. Among the identified varieties, kröhnkite and leightonite were recently reported as new occurrences among Vesuvius fumarolic mineral by Campostrini et al. [
54], whereas brochantite, cryptochalcite and natrochalcite detected in the present study are, to the authors’ knowledge, the first recorded occurrences at Vesuvius. The same is valid for boothite and dravertite, if their presence was confirmed by our ongoing investigations.
Cu-bearing vanadates characterize few studied samples, commonly indicated as
vesbine-bearing and restricted to the occurrences related to 1631 lavas. They are represented by OH- or H
2O-bearing minerals, as mottramite and volborthite, and by the anhydrous phase, starovaite. Mottramite is the prevailing cupric vanadate, at least in the investigated samples, and can be found in mineral associations, i.e., together with vanadinite + chrysocolla + wulfenite, vanadinite + chrysocolla + tenorite, or chrysocolla + starovaite + atacamite. Volborthite occurs with atacamite just in one sample. Both starovaite (the second worldwide occurrence) and vanadinite are the first occurrences at Vesuvius, to the authors’ knowledge. According to Russo and Punzo [
10] and Russo et al. [
58], it is more reliable the attribution of vesbine to volborthite, even though mottramite and vesigniéite have been also indicated as other possible mineral components; indeed, the present study pointed out that the so-called
vesbine of Vesuvius is a complex mixture of various minerals, alternatively composed of Cu, Cu–Pb, Pb and Cu–K vanadates and of Cu-bearing silicates. According to Pekov et al. [
6], Cu vanadates are important minerals in volcanic exhalations and their number is related to the fumarolic environment of two significant occurrences, i.e., Izalco (San Salvador) and Tolbachik (Kamchatka) volcanoes, is almost twice the number of supergene Cu vanadates. As for sulfates, the H-free vanadates can be considered more specific to the volcanic domain.
Tsumebite, a rare Cu–Pb phosphate, was detected in our study for the first time at Vesuvius in a complex and multimineralic assemblage; this mineral typically occurs as the secondary phase in the oxidized zone of As-bearing Cu–Pb deposits, with other secondary oxidized minerals. In fact, according to Pekov et al. [
6], copper phosphates (as well as carbonates) are even unknown in volcanic exhalations, unlike in supergene formation, where they are numerous and widespread. Moreover, Balić-Žunić et al. [
35] affirmed that phosphate group gives limited or null contribution to the mineralogy of the European fumarolic occurrences. In light of this, Vesuvius occurrence has an exceptional character.
The copper silicates identified in the studied samples are chrysocolla and litidionite, related to different type of rocks. Chrysocolla was only found in the thin encrustations on 1631 lava with atacamite or with the Cu-vanadates. Following the literature [
10], at Vesuvius, this silicate was extremely rare, and observed as light green encrustations in a lava fragment of the 1872 eruption (Le Novelle quarry, Ercolano;
Figure S1), with atacamite, apatite, hematite, magnetite and pyroxene, of presumably pneumatolytic origin. Chrysocolla is commonly found in weathered portions of many copper sulfide deposits, whereas it is very poorly represented, as Cu phosphates and carbonates, in volcanic exhalation products. Litidionite is even rarer and restricted to a unique occurrence, i.e., to thermally modified pyroclastic fragments by the fumarolic activity related to the 1872 eruption. The litidionite-bearing paragenesis is mainly composed of tridymite, calcinaksite (the second recorded worldwide occurrence), wollastonite and diopside and is typical of high-temperature alteration processes at the rock-fumaroles interface. Interestingly, the first discovery of this mineral was in a calcic xenolith hosted by an alkaline basalt of Bellerberg volcano (Eifel, Germany) as the product of contact metamorphism (metasomatism); the Bellerberg paragenesis consists of calcium silicates and CHS phases (wollastonite, gehlenite, browmillerite, tobermoreite, ettringite, etc.), which are also typical components of cement clinker and cement materials [
56,
57].
Finally, in the studied samples, we did not find Cu sulfides, such as chalcopyrite, covellite, chalcocite, or Cu sulfosalts; only sporadic grains of galena were observed. Base metal-bearing sulfides are quite rare in fumarolic products at Vesuvius [
10,
68], and can be present in products of different ages (see
Table 2 for copper sulfides). Indeed, Balić-Žunić et al. [
35] argued that metallic sulfides are confined to deeper parts of a volcanic system with its high-temperature hydrothermal conditions and rarely appear as sublimates, hence, they commonly occur in small amounts on the surface of the fumarole deposits.
5.2. Genetic Considerations
The Cu-bearing mineral assemblages observed in the investigated samples (or sub-samples) from Vesuvius are often very heterogeneous and can involve minerals that crystallized in different temperature conditions. In fact, Vesuvius can be distinguished by an oscillation in term of oxidizing vs. reducing conditions, mainly in the periods immediately following the eruptions [
35], as well as of temperature values. However, following the literature [
10,
11,
14,
35,
36,
37,
38,
39,
40,
41], most of the minerals found in the present research may be formed from high- to moderate-temperature fumaroles, at a temperature ranging from 300 °C to more than 650 °C. At higher temperatures, chlorides and fluorides (e.g., halite, sylvite, fluorite, elpasolite, sbacchiite) and alkali sulfates (e.g., thénardite) prevail. On the contrary, Cu–Cl oxyhalides and hydrohyhalides (as atacamite and paratacamite), Cu-bearing sulfates (e.g., chalcocyanite, euchlorine) and tenorite, together with a set of Pb-bearing minerals (e.g., cotunnite, pseudocotunnite, palmierite), hematite and sulfides probably formed at lower temperatures. This latter group can likely include other phases, even though with different precipitation temperatures, such as linarite, the Cu-bearing and Cu-free vanadates, chrysocolla, and tsumebite.
According to Pekov et al. [
6], similar temperature ranges can be assumed for the origin of Cu minerals of the Tolbachik volcano (Kamchatka, Russia); these authors distinguished two main groups of minerals, formed either in the hot zones of fumarolic system, corresponding to a T > 200 °C, mainly in the range ~400–700 °C, or in the moderately hot zones of fumarolic system, with a T < 200 °C and mainly in the interval of 70–150 °C. This study shows that anhydrous sulfates, as chalcocyanite, cryptchalcite, euchlorine and dravertite can be formed by hot fumarolic system. Instead, hydrous sulfates, i.e., cyanochroite, leightonite, natrochalcite and kröhnkite, can be related to lower temperature conditions and belong to the second group. In agreement with these observations, starovaite should belong to the high T group, being formed before chrysocolla, that likely had lower precipitation temperatures.
Litidionite and its peculiar paragenesis can be formed as a product of very localized high-temperature (exhalation-related) alteration processes of hosting silicates and with the introduction of other chemical components by fluids (e.g., Cu2+), also in agreement with that observed for the Eifel calcinaksite-bearing rocks; these phenomena can occur likely at a temperature of > 600 °C.
Other minerals randomly detected in the investigated samples, such as gypsum, anhydrite, alum-K, alunite, calcite and opal, are related to the last phases of mineral deposition, characterized by lower temperatures of exhalation phenomena, with a temperature rarely higher than 100 °C (see
Section 2).
In their thorough review, Pekov et al. [
6], have carried out a comparison between the crystal-chemical characteristics of Cu minerals found in volcanic fumarole environment and those formed under supergene conditions (i.e., the oxidation zone of sulfides ores, at T-P surface condition and under the influence of aqueous solutions). Even though the bulk of oxygen- and halogen-bearing Cu minerals occur primarily in the oxidation zones of sulfide ores (supergene environment), these authors report that more than one hundred new minerals were recently discovered in volcanic exhalation environment, showing great diversity and originality of motifs formed by the Cu
2+-centred coordination polyhedral for both the mineral sets. Our research confirms this statement, demonstrating that a large variety of the Cu minerals formed from fumarolic activity at Vesuvius. Pekov et al. [
6] also showed that in particular, H-free copper sulfates and vanadates of fumarolic origin are numerous and structurally diverse, but are unknown in the supergene environment, which is instead characterized by Cu
2+-based sulfates, vanadates and chlorides containing OH-bearing groups.
5.3. Crystal Structure and Complexity Considerations
A new approach for quantifying the structural complexity of a crystalline matter is based on the
TOPOS software package [
45]. The information-based complexity measures made possible to widely use the compelxity parameter as a tool for studying the evolution of mineral formation in various systems [
46,
69,
70,
71]. Mineral structures can be classified into very simple (0–20 bits, i.e., binary digits), simple (20–100 bits), intermediate (100–500 bits), complex (500–1000 bits), and very complex (>1000 bits). According to the proposed quantitative approach, the crystal structure can be viewed as a reservoir of information encoded in its complexity [
46]. Regarding this topic, we give a brief description of crystal structures of Cu-minerals from Vesuvius volcano and discuss their complexity in the context of different periods of volcanic activity.
Oxide minerals represent the simplest group from their structural complexity of Somma–Vesuvius Cu-minerals minerals. The crystal structure of tenorite
Figure 16 exhibits a framework based on CuO
4 planar groups polymerized through shared vertexes. On the polyhedral approach, the crystal structure of curpite can be described as based on the O-centered tetrahedral framework, whereas each [OCu
4] tetrahedra is connected by shared vertexes.
Vesuvius carbonates are represented by azurite and malachite. The structure of azurite based upon Cu
2O
8-chains connected to the complex layers via CuO
4 planar groups copper layers with strong magnetic properties [
72] linked by carbonate groups. In the structure of malachite, there are two independent copper sites in square-planar and octahedral coordination. The edge-shared octahedral columns connected with adjacent [CuO
4] groups form a zig-zag like pattern combined with each other by CO
3-groups. The importance of hydrogen bonding in malachite and azurite structures should be noted.
There are two copper silicate minerals in the investigated samples, but the crystal structure of chrysocolla still unsolved. The crystal structure of litidionite consists of tubular [Si
8O
20] chains with a hexagonal cross-section, which are interconnected by [CuO
4] planar groups [
56]. The structural cavities are filled by the Na and K atoms.
Hydroxyhalides include atacamite and its polymorph paratacamite. Generally, both minerals have the same net-like motif formed chains of CuO
4 planar groups connected with perpendicular chains based by [CuO
5Cl] distorted octahedra. The paratacamite structure is more ordered, one half distorted octahedra changed by [CuO
6] and the other half changed by CuO
4 planar groups, whereas Cl
- anions filled voids [
73,
74].
There are three vanadates found at the Vesuvius volcano. In the crystal structure of mottramite distorted Cu-octahedra connected by shared edges forms infinite chains along the
b axis. The V
5+ cations occupy slightly distorted tetrahedra connected by three oxygen vertexes with [CuO
6]-based chains and one oxygen involved in hydrogen bonding. The Pb
2+ cations are seven-coordinated with bonds in the range 2.45–2.80 Å [
75]. There are five independent Cu sites in the crystal structure of starovaite, which has tetragonal pyramidal or trigonal bipyramidal coordination. They form complex ∞[Cu
5O
13] sheets connected via [VO
4] tetrahedra in the framework. The cages in the framework are occupied by 10-coordinated K atoms [
76]. Volborthite is a well known type of layered structure with octahedral layers based on [CuO
6] distorted octahedra with 1/3 empty sites. Such layers are connected with [V
2O
5] nesovanadate groups that form channels populated by water molecules [
77].
The most complex Vesuvius Cu-minerals are related to the sulfates, which are represented by group of 14 minerals (
Figure 17). Structurally, they can be divided into five subgroups based on the arrangement of Cu atoms: single Cu-octahedra (kröhnkite, chalcanthite, cyanochroite, boothite, leightonite); single Cu-centered squares (chlorothionite); octahedral chains of edge-shared Cu-octahedra (dravertite, linarite, chalcocyanite, natrochalcite, tsumebite); complex chains (brochantite); complex clusters (euchlorine, cryptochalcite).
The dominant motif of the kröhnkite structure is the chains of [SO
4] tetrahedra connected with single Cu octahedra, extending along the c axis; these chains are linked together by Na atoms coordinated by seven anions at distances between 2.39 and 2.57 Å [
78]. The chalcanthite structure is based on infinite chains of distorted Cu-octahedra with an edge shared Mg-octahedra. These chains are held together by corner-shared sulfate tetrahedra and a hydrogen bonding system [
79]. In the cyanochroite structure, each Cu atom is coordinated by six water molecules and connected via hydrogen bonds with [SO
4] tetrahedra and K atoms [
80]. In this work, boothite is the second complex mineral and its structure is similar to that of cyanochroite based on Cu-octahedra with six water molecules and hydrogen-bonded with [SO
4] tetrahedra [
81]. Generally, the leightonite structure has a framework character. To date, only a disordered model has been proposed [
82]. In this model, Cu atoms have square-planar coordination, but octahedral if considering mixed K/O sites. All equatorial O atoms in the copper octahedra are connected with corner-shared sulfate groups. The Ca atoms are eight-coordinated, whereas six oxygens are shared for sulfate groups.
In the chlorothionite crystal structure, the Cu atoms are square-coordinated. In chlorothionite, Cu is coordinated by two Cl and two O atoms, whereas oxygen atoms are edge for adjacent [SO
4] tetrahedra and chlorine atoms bonded with K [
83].
The crystal structures of dravertite, linarite, chalcocyanite, natrochalcite, tsumebite consist of Cu or mixed Cu–Mg (dravertite) layers of edge-shared octahedral chains. In terms of structure, dravertite can be considered as a cation-ordered derivative of chalcocyanite, with alternating Cu- and Mg-centred octahedra in cationic chains [
50]. In chalcocyanite, the structure of octahedral chains parallel to [010] is connected to a framework by [SO
4] tetrahedra [
84]. In the linarite structure, octahedral Cu-chains are connected with vertex-shared [SO
4] tetrahedra and eight-coordinated Pb atoms [
85]. The sulfate groups are connected through Pb atoms; thus, they form a double (Pb–S) layer between Cu-layers. Similarly to other members of the brackebuschite supergroup, the tsumebite crystal structure contains a cubic closest-packed array of O and Pb atoms with infinite chains of edge-sharing distorted Cu octahedra decorated by two unique [SO
4] and [PO
4] tetrahedra [
86]. The Cu–S layers in the natrochalcite structure are close to those in chalcocyanite and are built from [Cu
2OH(SO
4)
2·2H
2O] sheets and are interconnected by Na
+ ions and hydrogen bonds [
87].
In the brochantite structure, two distorted Cu octahedra are interconnected through common edges to build infinite planar double chains running along the
c axis. These chains are interconnected by sharing vertices to build ‘zig-zag’ chains running along
c. Each [SO
4] tetrahedra is connected to three adjacent double chains [
88].
For the structures containing an “additional” oxygen atom, it is convenient to use an approach to understand the structure in terms of anion-centered coordination polyhedra [
89]. In the euchlorine structure, “additional” non-sulfate oxygen atoms are considered to form two independent oxocentered [OCu
4] tetrahedra which share a common Cu1–Cu1 edge, thus forming an [O
2Cu
6] dimer. Two S-centered sulfate tetrahedra are attached ‘face-to-face’ to the dimers. And one [SO
4] tetrahedra provides the linkage of these clusters in two dimensions to form [Cu
3O(SO
4)
3] layers parallel to the
bc plane. Potassium and sodium atoms are located in the interlayer [
90].
The crystal structure of cryptochalcite is based on the heteropolyhedral framework [Cu
5O(SO
4)
5] composed of two types of alternating Cu–S–O polyhedral layers [Cu
2(SO
4)
2] and [Cu
3O(SO
4)]
2 coplanar to the
ab plane and connected via SO
4 tetrahedra. The first layer consists of clusters formed by four edge-sharing octahedra with signifificant Janh–Teller distortion. The second layer is formed by two isolated Cu-centered tetragonal pyramids alternating with two Cu-centered trigonal bipyramids. Inside the layers, Cu-polyhedra are connected via corner-shared [SO
4] tetrahedra, interstices filled by K atoms [
49].
The complexity totals for Cu-minerals are 1449, 2101 and 550 bits for temperature ranges of 50–200 °C, 200–400 °C and 400–700 °C, respectively (
Figure 18a). At first sight, this is not in agreement with Goldsmith’s simplexity principle [
91]. The observed splash of complexity for middle temperature Cu-minerals is connected with the presence of “additional oxygen” in formula. The additional oxygen in the structure makes possible the organization of Cu-clusters, which can also be described in terms of oxy-centered crystallography [
6]. If we consider the complexity of all minerals in assemblages including Cu-free minerals, the totals would be 1863, 2245 and 1126 bits (
Figure 18b). The means of complexity for middle-temperature minerals is still higher than for low-temperature ones. This contradiction with the simplexity principle is probably due to a lack of information about Cu-free minerals, or due to destruction of some water-soluble minerals under atmosphere conditions.
The maximal values of total complexity correspond to the samples attributed, at least according to the museum labels, to the 1868–1870, 1872–1875 and 1880–1885 eruptions. The main contribution here is given by cryptochalcite and boothite minerals. According to Ivanyuk et al. [
92], a relationship between massif sizes and the number of minerals known in them can exist; similarly, the relationship between different means of total complexity for various periods of Somma–Vesuvius volcanic activity and the eruption power (size) is currently under investigation. At the surface conditions, ancient samples should contain a smaller number of water-soluble minerals than younger eruptions. This should lead to a decrease of the total number of minerals and consequently, to a decrease of total complexity during the geological time. On the other hand, some minerals could be a product of the alteration of initial chlorides at surface conditions. Such low-temperature (≤50 °C) minerals (i.e., boothite) contain a complex hydrogen bonding system, which leads to an increase of the total complexity for the low-T assemblages [
90].